布莱克威尔哲学指南 969页

\n\nTheBlackwellCompaniontoPhilosophy\nBlackwellCompanionstoPhilosophyThisoutstandingstudentreferenceseriesoffersacomprehensiveandauthoritativesurveyofphilosophyasawhole.Writtenbytoday’sleadingphilosophers,eachvolumeprovideslucidandengagingcoverageofthekeyfigures,terms,topics,andproblemsofthefield.Takentogether,thevolumesprovidetheidealbasisforcourseuse,represent-inganunparalleledworkofreferenceforstudentsandspecialistsalike.Alreadypublishedintheseries:14ACompaniontoCognitiveScienceEditedbyWilliamBechteland1TheBlackwellCompaniontoGeorgeGrahamPhilosophy,SecondEdition15ACompaniontoBioethicsEditedbyNicholasBunninandEditedbyHelgaKuhseandPeterSingerE.P.Tsui-James16ACompaniontothePhilosophers2ACompaniontoEthicsEditedbyRobertL.ArringtonEditedbyPeterSinger17ACompaniontoBusinessEthics3ACompaniontoAestheticsEditedbyRobertE.FrederickEditedbyDavidCooper18ACompaniontothePhilosophyof4ACompaniontoEpistemologyScienceEditedbyJonathanDancyandErnestSosaEditedbyW.H.Newton-Smith5ACompaniontoContemporaryPolitical19ACompaniontoEnvironmentalPhilosophyPhilosophyEditedbyRobertE.GoodinandEditedbyDaleJamiesonPhilipPettit20ACompaniontoAnalyticPhilosophy6ACompaniontoPhilosophyofMindEditedbyA.P.MartinichandDavidSosaEditedbySamuelGuttenplan21ACompaniontoGenethics7ACompaniontoMetaphysicsEditedbyJustineBurleyandJohnHarrisEditedbyJaegwonKimandErnestSosa22ACompaniontoPhilosophicalLogic8ACompaniontoPhilosophyofLawandEditedbyDaleJacquetteLegalTheoryEditedbyDennisPatterson23ACompaniontoEarlyModernPhilosophy9ACompaniontoPhilosophyofReligionEditedbyStevenNadlerEditedbyPhilipL.QuinnandCharlesTaliaferro10ACompaniontothePhilosophyofForthcomingLanguageACompaniontoAfricanAmericanEditedbyBobHaleandCrispinWrightPhilosophy11ACompaniontoWorldPhilosophiesEditedbyTommyLottandJohnPittmanEditedbyEliotDeutschandRonACompaniontoAfricanPhilosophyBontekoeEditedbyKwasiWiredu12ACompaniontoContinentalPhilosophyEditedbySimonCritchleyandACompaniontoAncientPhilosophyWilliamSchroederEditedbyMaryLouiseGill13ACompaniontoFeministPhilosophyACompaniontoMedievalPhilosophyEditedbyAlisonM.JaggarandEditedbyJorgeJ.E.Gracia,GregReichberg,IrisMarionYoungandTimothyNoone\nTheBlackwellCompaniontoPhilosophySECONDEDITIONEditedbyNICHOLASBUNNINandE.P.TSUI-JAMES\nCopyright©1996,2003BlackwellPublishersLtd,aBlackwellPublishingcompanyEditorialmatter,selectionandarrangementcopyright©NicholasBunninandEricTsui-James1996,2003Firsteditionpublished1996Reprinted1996(twice),1998,1999,2002Secondeditionpublished2003350MainStreet,Malden,MA02148-5018,USA108CowleyRoad,OxfordOX41JF,UK550SwanstonStreet,CarltonSouth,Victoria3053,AustraliaKurfürstendamm57,10707Berlin,GermanyTherightofNicholasBunninandEricTsui-JamestobeidentifiedastheAuthorsoftheEditorialMaterialinthisworkhasbeenassertedinaccordancewiththeUKCopyright,DesignsandPatentsAct1988.Allrightsreserved.Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,ortransmitted,inanyformorbyanymeans,electronic,mechanical,photocopying,recordingorotherwise,exceptaspermittedbytheUKCopyright,DesignsandPatentsAct1988,withoutthepriorpermissionofthepublisher.LibraryofCongressCataloging-in-PublicationDataTheBlackwellcompaniontophilosophy/editedbyNicholasBunninandE.P.Tsui-James.––2nded.p.cm.––(Blackwellcompanionstophilosophy)Includesbibliographicalreferencesandindex.ISBN0–631–21907–2––ISBN0–631–21908–0(pbk.)1.Philosophy.I.Bunnin,Nicholas.II.Tsui-James,E.P.III.Series.B21.B562003100––dc212002023053AcataloguerecordforthistitleisavailablefromtheBritishLibrary.1Setin10on12/2ptPhotinabySNPBest-setTypesetterLtd,HongKongPrintedandboundintheUnitedKingdombyT.J.International,Padstow,CornwallForfurtherinformationonBlackwellPublishing,visitourwebsite:http://www.blackwellpublishing.com\nForAntoniaandOliverBunninandJamiePerry\nContentsPrefacetotheSecondEditionixPrefacetotheFirstEditionxNotesonContributorsxiiContemporaryPhilosophyintheUnitedStates–JohnR.Searle1ContemporaryPhilosophy:ASecondLook–BernardWilliams23PartIAreasofPhilosophy351Epistemology–A.C.Grayling372Metaphysics–SimonBlackburn,withasectiononTimebyRobinLePoidevin613PhilosophyofLanguage–MartinDavies904PhilosophyofLogic–A.W.Moore1475PhilosophyofMind–WilliamG.Lycan1736Ethics–JohnSkorupski2027Aesthetics–SebastianGardner2318PoliticalandSocialPhilosophy–DavidArchard2579PhilosophyofScience–DavidPapineau28610PhilosophyofBiology–ElliottSober31711PhilosophyofMathematics–MaryTiles34512PhilosophyofSocialScience–MartinHollis37513PhilosophyofLaw–N.E.Simmonds40314PhilosophyofHistory–LeonPompa42815PhilosophyofReligion–CharlesTaliaferro45316AppliedEthics–JohnHaldane49017Bioethics,GenethicsandMedicalEthics–RebeccaBennett,CharlesA.Erin,JohnHarrisandSørenHolm49918EnvironmentalEthics–HolmesRolston,III51719BusinessEthics–GeorgesEnderle53120PhilosophyandFeminism–JeanGrimshawandMirandaFricker55221Ethnicity,CultureandPhilosophy–RobertBernasconi567\nCONTENTSPartIIHistoryofPhilosophy58322AncientGreekPhilosophy–RobertWardy58523PlatoandAristotle–LesleyBrown60124MedievalPhilosophy–JorgeJ.E.Gracia61925Bacon–StephenGaukroger63426DescartesandMalebranche–RichardFrancksandGeorgeMacdonaldRoss64427SpinozaandLeibniz–RichardFrancksandGeorgeMacdonaldRoss65828Hobbes–TomSorell67129Locke–R.S.Woolhouse68230Berkeley–HowardRobinson69431Hume–PeterJones70932Kant–DavidBell72533Hegel–MichaelInwood74134Marx–RichardNorman75035Bentham,MillandSidgwick–RossHarrison75936Pragmatism–SusanHaack77437FregeandRussell–R.M.Sainsbury79038Moore–ThomasBaldwin80539Wittgenstein–DavidPears81140Nietzsche–DavidE.Cooper82741HusserlandHeidegger–TaylorCarman84242Sartre,FoucaultandDerrida–GaryGutting860Glossary875Appendix893Index905\nPrefacetotheSecondEditionWethankreadersfortheirgratifyingresponsetothefirsteditionoftheCompanion.ThesecondeditionprovidesnewchaptersonPhilosophyofBiology;Bioethics,GenethicsandMedicalEthics;EnvironmentalEthics;BusinessEthics;Ethnicity,CultureandPhilosophy;PlatoandAristotle;FrancisBacon;Nietzsche;HusserlandHeidegger;andSartre,FoucaultandDerrida.TherearesignificantrevisionsorextensionstochaptersonMetaphysics,PhilosophyofLanguage,PhilosophyofMind,PoliticalandSocialPhilosophy,PhilosophyofReligion,PhilosophyandFeminism,andHobbes.ThediscussionofDescartes,SpinozaandLeibnizisnowdividedbetweentwochapters,andinanewsectionMalebrancheisconsideredalongwithDescartesinthefirstofthese.AlongerchapteronMedievalPhilosophyreplacesthechapterbyC.F.J.Martin,whowasunavailabletoextendhiswork.WewelcomeournewcontributorsandhopethatreaderswillcontinuetobechallengedanddelightedbytheCompanionasawhole.NicholasBunninE.P.Tsui-James\nPrefacetotheFirstEditionThisCompanioncomplementstheBlackwellCompanionstoPhilosophyseriesbypresentinganewoverviewofphilosophypreparedbythirty-fiveleadingBritishandAmericanphilosophers.IntroductoryessaysbyJohnSearleandBernardWilliams,whichassessthechangesthathaveshapedthesubjectinrecentdecades,arefollowedbychaptersexploringcentralproblemsanddebatesintheprincipalsubdisciplinesofphilosophyandinspecializedfields,chaptersconcerningtheworkofgreathistoricalfiguresandchaptersdiscussingnewlydevelopingfieldswithinphilosophy.Throughoutthecourseofitschapters,theCompanionexaminestheviewsofmanyofthemostwidelyinfluentialfiguresofcontemporaryphilosophy.Althoughwide-ranging,theCompanionisnotexhaustive,andemphasisisplacedondevelopmentsinAnglo-Americanphilosophyinthelatterpartofthetwentiethcentury.ApremiseunderlyingtheCompanionisthatmajorparticipantsinphilosoph-icaldebatecanprovideaccountsoftheirownfieldsthatarestimulating,accessible,stylishandauthoritative.Initsprimaryuse,theCompanionisaninnovativetextbookforintroductorycoursesinphilosophy.Teacherscanusethebroadcoveragetoselectchaptersinaflexiblewaytosupportavarietyofcoursesbasedoncontemporaryproblemsorthehistoricaldevel-opmentofthesubject.Specialistchapterscanbeusedselectivelytoaugmentstandardintroductorytopicsortopreparestudentsindividuallyfortermpapersoressays.Chap-tersincludeinitialsummaries,boxedfeatures,cross-references,suggestionsforfurtherreading,referencesanddiscussionquestions.Inaddition,termsaremarkedforacommonglossary.Thesefeaturesandtheproblem-settingnatureofthediscussionsencouragestudentstoseethesubjectasawholeandtogainconfidencethatexplo-rationswithinphilosophycanleadtounexpectedandrewardinginsights.Inthisaspect,theCompanionreflectsthecontributors’experienceofsmallgroupteaching,inwhichargumentsandperspectivesarerigorouslytestedandinwhichnosolutionisimposed.Initssecondaryuse,theCompanionwillaccompanystudentsthroughouttheirundergraduatecareersandwillalsoservethegeneralreaderwishingtounderstandthecentralconceptsanddebateswithinphilosophyoritsconstituentdisciplines.Studentsareunlikelytoreadthewholevolumeintheirfirstyearofstudy,butthosecontinuingwithphilosophywillfindtheirappreciationoftheworkdeepeningovertime\nPREFACETOTHEFIRSTEDITIONastheygaininsightintothetopicsofthemoreadvancedchapters.TheCompanionwillhelpthemtoformulatequestionsandtoseeconnectionsbetweenwhattheyhavealreadystudiedandnewterrain.Initsfinaluse,theCompanionbearsaspecialrelationshiptotheBlackwellCom-panionstoPhilosophyseries.ManyreaderswillwishtoreadtheintegrateddiscussionsofthechaptersofthepresentCompanionfororientationbeforeturningtothedetailed,alphabeticallyarrangedarticlesofthevolumesintheCompanionseries.Althoughcon-ceivedasaseparatevolume,theCompaniontoPhilosophywillserveasausefulguidetotheotherexcellentCompanionsinwhatamountstoacomprehensiveencyclopediaofphilosophy.ThegeneralreadermightbeginwiththeintroductoryessaysandturntochaptersonEpistemology,Metaphysics,EthicsandPoliticalandSocialPhilosophy,ortohistori-calchaptersfromAncientGreekPhilosophytoHume.Cross-referencesandspecialinterestswillleadreaderstootherchapters.Cross-referencesinthetextaremarkedinsmallcapitalsfollowedbyachapternumberorpagenumbersinparentheses:Ethics(chapter6)orProbability(pp.308–11).Wehaveusedourjudgementinmarkingtermsappearingmanytimesinthetextforcross-references,andhopethatwehavesuppliedguidancewithoutdistractingreaders.TheCompanionalsoprovidesaglossaryof210termsandacomprehensiveindex.Bothappearattheendofthevolume,andreadersareadvisedtousethemreg-ularlyforhelpinreadingthechapters.Whenanauthordoesnotrefertoabookbyitsfirstedition,arecentpublicationiscitedinthetext,andtheoriginaldateofpublica-tion(orinsomecasesofcomposition)willappearinsquarebracketsinthereferences.Aseditors,wearefullyawareofourgoodfortuneinattractingsuperbcontributors.ThecomplexityoftheirinsightsandtheclarityoftheirpresentationsarethechiefattractionsoftheCompanion.Weappreciatetheircareinmakingthedifficultnotonlyaccessiblebutdelightfulaswell.WealsowishtothanktheDepartmentsofPhilosophyattheUniversityofEssexandtheUniversityofHongKongfortheirsupportthrough-outthepreparationofthisvolume.WeareespeciallygratefultoLaurenceGoldstein,TimMooreandFrankCioffifortheircommentsandadvice.AversionoftheCom-panionispublishedinChinesebytheShandongAcademyofSocialSciences,andweappreciatethefriendlyco-operationofourChineseco-editors.Ourcoverillustration,R.B.Kitaj’sphilosophicallyresonantIfNot,Not,isaworkbyanAmericanartistworkinginLondonduringtheperiodthatprovidesthemainfocusofourvolume.NicholasBunninE.P.Tsui-Jamesxi\nNotesonContributorsDavidArchardisReaderinMoralPhilosophyandDirectoroftheCentreforEthics,PhilosophyandPublicAffairsattheUniversityofStAndrews.HeistheauthorofSexualConsent(1998)andco-editorofTheMoralandPoliticalStatusofChildren:NewEssays(2002).ThomasBaldwinisProfessorofPhilosophyattheUniversityofYork.HepreviouslytaughtattheUniversityofCambridge(wherehewasFellowofClareCollege)andatMakerereUniversity.HehaspublishedG.E.Moore(1990)andContemporaryPhiloso-phy:PhilosophyinEnglishsince1945(2001)inadditiontomanyarticlesonissuesinmetaphysicsandthephilosophyoflanguage.DavidBellisProfessorofPhilosophyattheUniversityofSheffield.HeistheauthorofworksonFrege,HusserlandKant.Hisinterestsincludethefoundationsofarithmetic,solipsismandthenatureandoriginsoftheanalytictradition.RebeccaBennettisLecturerinBioethicsattheCentreforSocialEthicsandPolicy,SchoolofLaw,UniversityofManchester.Sheedited(withCharlesErin)HIVandAIDS:Testing,ScreeningandConfidentiality(1999).RobertBernasconiisMossProfessorofPhilosophyattheUniversityofMemphis.HeistheauthorofTheQuestionofLanguageinHeidegger’sHistoryofBeing(1985)andHeideggerinQuestion(1993)aswellasnumerousarticlesonHegelandontwentieth-centuryEuropeanphilosophy.HehaseditedcollectionsofessaysonDerridaandonLevinasandmostrecentlyRace(2001).SimonBlackburnisProfessorofPhilosophyattheUniversityofCambridge.AformereditorofthejournalMind,hehaswrittenRulingPassions(1998),SpreadingtheWord(1984),EssaysinQuasiRealism(1993)andTheOxfordDictionaryofPhilosophy(1994).Hiscurrentworkconcernsproblemsofrealismanditsalternativesastheyhaveemergedinhistoricalandcontemporarywork.LesleyBrownisTutorialFellowinPhilosophy,SomervilleCollege,UniversityofOxford.ShehaswrittenonPlato,especiallyhismetaphysicsandepistemology,andonancientphilosophyoflanguage.\nCONTRIBUTORSNicholasBunninisDirectorofthePhilosophyProjectattheInstituteforChineseStudies,UniversityofOxfordandpreviouslytaughtattheUniversityofGlasgowandtheUniversityofEssex.Hecompiled(withJiyuanYu)theDictionaryofWesternPhilos-ophy:English–Chinese(2001)andedited(withChung-yingCheng)ContemporaryChinesePhilosophy(2002).Hismaininterestsareinmetaphysics,thephilosophyofmindandpoliticalphilosophy.TaylorCarmanisAssistantProfessorofPhilosophyatBarnardCollege,ColumbiaUniversity.Heisco-editorofTheCambridgeCompaniontoMerleau-Ponty(forthcoming)andtheauthorofHeidegger’sAnalytic:Interpretation,Discourse,andAuthenticityin‘BeingandTime’(forthcoming),andofotherarticlesonHusserl,Heidegger,andMerleau-Ponty.DavidE.CooperisProfessorofPhilosophyattheUniversityofDurhamandDirectoroftheDurhamInstituteofComparativeEthics.HisbooksincludeMetaphor(1986),Existentialism:AReconstruction(2ndrevdedn2000),WorldPhilosophies:AnHistoricalIntroduction(2ndrevdedn2002)andTheMeasureofThings:Humanism,HumilityandMystery(2002).MartinDaviesisProfessorofPhilosophyintheResearchSchoolofSocialSciences,AustralianNationalUniversity.HewasformerlyWildeReaderinMentalPhilosophyattheUniversityofOxford.Hehaspublishedwidelyintheareasofphilosophyoflanguage,mindandpsychology.GeorgesEnderleisArthurandMaryO’NeilProfessorofInternationalBusinessEthicsattheUniversityofNotreDame.HisbooksincludeBusinessStudentsFocusonEthics(1993),translatedintoPortuguese(1997)andChinese(2001).CharlesA.ErinisSeniorLecturerinAppliedPhilosophyandFellowoftheInstituteofMedicine,LawandBioethicsattheUniversityofManchester.Hehaswrittenwidelyontopicsinbioethicsandedited(withRebeccaBennett)HIVandAIDS:Testing,ScreeningandConfidentiality(1999).RichardFrancksisDirectorofUndergraduateStudiesinPhilosophyattheUniversityofLeeds.Hismaininterestsareinepistemology,thehistoryofphilosophyandthephilosophyofhistory.MirandaFrickerisLecturerinPhilosophyatBirkbeckCollege,UniversityofLondonandwaspreviouslyLecturerinPhilosophyandBritishAcademyPostdoctoralFellowatHeythropCollege,UniversityofLondon.Shehaspublishedarticlesinepistemology,ethicsandsocialphilosophy,andedited(withJenniferHornsby)TheCambridgeCompaniontoFeminisminPhilosophy(2000).Hercurrentworkfocusesontheideaofanethicsofepistemicpractice.SebastianGardnerisReaderinPhilosophyatUniversityCollege,London.HeistheauthorofIrrationalityandthePhilosophyofPsychoanalysis(1993)andKantandtheCritiqueofPureReason(1999).Hisinterestslieinaesthetics,psychoanalysisandthehistoryofphilosophy.xiii\nCONTRIBUTORSStephenGaukrogerisProfessorofHistoryofPhilosophyandHistoryofScienceattheUniversityofSydney.HeistheauthorofExplanatoryStructures(1978),CartesianLogic(1989),Descartes:AnIntellectualBiography(1995),FrancisBaconandtheTransformationofEarlyModernPhilosophy(2000)andDescartes’SystemofNaturalPhilosophy(2001).HehasalsoeditedfourcollectionsofessaysandpublishedtranslationsofDescartesandArnaud.JorgeJ.E.GraciaisaStateUniversityofNewYorkDistinguishedProfessorandholdstheSamuelF.CaponChairintheDepartmentofPhilosophy,StateUniversityofNewYork,UniversityatBuffalo.Hehaswrittenwidelyonmedievalphilosophy,metaphysics,philosophicalhistoriography,philosophyoflanguageandphilosophyinLatinAmerica.HisbooksincludeIntroductiontotheProblemofIndividuationintheEarlyMiddleAges(2ndrevdedn1988),Individuality:AnEssayontheFoundationsofMetaphysics(1988),PhilosophyandItsHistory:IssuesinPhilosophicalHistoriography(1992),ATheoryofTextuality:TheLogicandEpistemology(1995),Texts:OntologicalStatus,Identity,Author,Audience(1996)andMetaphysicsandItsTask:TheSearchfortheCategoricalFoundationofKnowledge(1999).A.C.GraylingisReaderinPhilosophyatBirkbeckCollege,London,andSupernu-meraryFellowatStAnne’sCollege,Oxford.AmonghisbooksareAnIntroductiontoPhilosophicalLogic(3rdedn1992),TheRefutationofScepticism(1985),Berkeley:TheCentralArguments(1986),Wittgenstein(1988),Russell(1993),MoralValues(1998),TheQuarreloftheAge(2000)andTheMeaningofThings(2001).HehaseditedPhilosophy:AGuideThroughtheSubject(1995)andPhilosophy:FurtherThroughtheSubject(1998).JeanGrimshawtaughtPhilosophyandWomen’sStudiesattheUniversityoftheWestofEngland,Bristol.SheistheauthorofFeministPhilosophers:Women’sPerspectivesonPhilosophicalTraditions(1986)andanumberofarticles,mainlyonfeminismandphi-losophy.Shehasedited(withJaneArthurs)Women’sBodies:DisciplineandTransgression(1999).GaryGuttingisProfessorofPhilosophyattheUniversityofNotreDame.HeistheauthorofReligiousBeliefandReligiousSkepticism(1982),MichelFoucault’sArchaeologyofKnowledge(1989)andFrenchPhilosophyintheTwentiethCentury(2001).SusanHaackformerlyProfessorofPhilosophyattheUniversityofWarwick,currentlyProfessorofPhilosophyattheUniversityofMiami,istheauthorofDeviantLogic(1974),PhilosophyofLogic(1978),EvidenceandInquiry:TowardsReconstructioninEpistemology(1993)andManifestoofaPassionateModerate:UnfashionableEssays(1998).Hermainareasofinterestarethephilosophyoflogicandlanguage,episte-mologyandmetaphysicsandpragmatism.SheisapastPresidentoftheCharlesPeirceSociety.JohnHaldaneisProfessorofPhilosophyandformerlyDirectoroftheCentreforPhilosophyandPublicAffairsattheUniversityofStAndrews.Hehaspublishedwidelyinthephilosophyofmind,thephilosophyofvalueandthehistoryofphiloso-phy.Heisco-authorwithJ.J.C.SmartofAtheismandTheism(1996)intheBlackwellGreatDebatesinPhilosophyseries.xiv\nCONTRIBUTORSJohnHarrisisSirDavidAllianceProfessorofBioethics,InstituteofMedicine,LawandBioethics,UniversityofManchester.HeisamemberoftheUnitedKingdomHumanGeneticsCommissionandoftheEthicsCommitteeoftheBritishMedicalAssociation.HewasaFounderDirectoroftheInternationalAssociationofBioethicsandafoundermemberoftheBoardofthejournalBioethics.AmonghisbooksareTheValueofLife(1985)andClones,GenesandImmortality(1998)(arevisededitionofWonderwomanandSuperman,1992),andheiseditorofBioethics(2001)intheOxfordReadingsinPhi-losophyseries.RossHarrisonteachesphilosophyattheUniversityofCambridge,whereheisalsoaFellowofKing’sCollege.AmonghispublicationsareBentham(1983),Democracy(1993)and(aseditorandcontributor)HenrySidgwick(2001).MartinHolliswasProfessorofPhilosophyattheUniversityofEastAnglia,Norwich.Hespecializedinthephilosophyofsocialscience,especiallyintopicstodowithratio-nality.AmonghisbooksareModelsofMan(1977),TheCunningofReason(1987),ThePhilosophyofSocialScience(1994),ReasoninAction(1995),TrustWithinReason(1998)andPluralismandLiberalNeutrality(1999).Thelasttwovolumeswerepublishedafterhisuntimelydeathin1998.SørenHolmisProfessorofClinicalBioethicsattheUniversityofManchester.HeistheauthorofEthicalProblemsofClinicalPractice:TheEthicalReasoningofHealthCareProfessionals(1997)andhasedited(withInezdeBeaufortandMedardHilhorst)IntheEyeoftheBeholder:EthicsandMedicalChangeofAppearance(1996)and(withJohnHarris)TheFutureOfHumanReproduction:Ethics,ChoiceandRegulation(1998).MichaelInwoodisTutorialFellowinPhilosophyatTrinityCollege,Oxford.HehaspublishedseveralbooksonHegel.HisotherinterestsincludeancientphilosophyandHeidegger.HeisespeciallyinterestedintheinterconnectionsbetweenGreekandGermanphilosophy.PeterJoneswasProfessorofPhilosophyandDirectoroftheInstituteforAdvancedStudiesintheHumanitiesattheUniversityofEdinburgh.Heistheauthorofnumer-ousworks,includingHume’sSentiments(1982).RobinLePoidevinisProfessorofMetaphysicsattheUniversityofLeeds,wherehewasHeadoftheSchoolofPhilosophy1988–2001.HeistheauthorofChange,CauseandContradiction:ADefenceoftheTenselessTheoryofTime(1991)andArguingforAtheism:AnIntroductiontothePhilosophyofReligion(1996)andhaseditedQuestionsofTimeandTense(1998)and(withMurrayMacBeath)ThePhilosophyofTime(1993).WilliamG.LycanisWilliamRandKenan,JrProfessorofPhilosophyattheUniversityofNorthCarolina.Hehaspublishedanumberofbooks,includingConsciousness(1987),JudgementandJustification(1988)andConsciousnessandExperience(1996).HeistheeditorofMindandCognition(1990).Hisinterestsareinthephilosophyofmind,thephilosophyoflanguageandepistemology.A.W.MooreisTutorialFellowinPhilosophyatStHugh’sCollege,Oxford.HeistheauthorofTheInfinite(2ndedn2001)andPointsofView(1997).Hehasalsoeditedtwocollectionsofessays:MeaningandReference(1993)andInfinity(1993).xv\nCONTRIBUTORSRichardNormanisProfessorofMoralPhilosophyattheUniversityofKent.Hispub-licationsincludeTheMoralPhilosophers(1983),FreeandEqual(1987)andEthics,KillingandWar(1995).DavidPapineauisProfessorofPhilosophyofScienceatKing’sCollege,London.Hehaspublishedwidelyinepistemology,thephilosophyofmindandthephilosophyofscience.HisbooksincludeRealityandRepresentation(1987),PhilosophicalNaturalism(1993),IntroducingConsciousness(2000)andThinkingAboutConsciousness(2002).DavidPearsisEmeritusProfessorofPhilosophyattheUniversityofOxford.HismostrecentpublicationsareTheFalsePrison:AStudyintheDevelopmentofWittgenstein’sPhi-losophy(2vols,1987and1988)andHume’sSystem:AnExaminationofBookIoftheTreatise(1991).Hisotherinterestsincludeentomologyandthevisualarts.LeonPompawasProfessorofPhilosophyattheUniversityofBirmingham.Hisresearchinterestsincludethehistoryofphilosophyandthephilosophyofhistory.Hehaspublishedanumberofarticlesontheproblemsoffact,valueandnarrativeinhistoryandonDescartes,Vico,Kant,Hegel,Marx,CollingwoodandWittgenstein.Heco-editedwithW.H.DraySubstanceandForminHistory:EssaysinPhilosophyofHistory(1981),waseditorandtranslatorofVico:AStudyofthe‘NewScience’(2ndedn1990)andistheauthorofHumanNatureandHistoricalKnowledge:Hume,HegelandVico(1990).HowardRobinsonisProfessorofPhilosophy,CentralEuropeanUniversity,Budapest.HewaspreviouslySorosProfessorofPhilosophyattheEötvösLorándUniversity,BudapestandReaderinPhilosophyattheUniversityofLiverpool.Hismaininterestsareinthephilosophyofmindandinidealism.HeistheauthorofMatterandSense(1982)andPerception(1994),andco-author(withJohnFoster)ofEssaysonBerkeley(1985).HeeditedObjectionstoPhysicalism(1991)andiscurrentlyeditingBerkeley’sPrinciplesandThreeDialoguesforOxfordUniversityPress’sWorldClassicsseries.HolmesRolston,IIIisUniversityDistinguishedProfessorandProfessorofPhilosophyatColoradoStateUniversity.Hehaswrittensevenbooks,mostrecentlyGenes,GenesisandGod(1999),PhilosophyGoneWild(1986),EnvironmentalEthics(1988),ScienceandReligion:ACriticalSurvey(1987)andConservingNaturalValue(1994).HegavetheGiffordLectures,UniversityofEdinburgh,1997–8,haslecturedonsevencontinents,isfeaturedinJoyA.Palmer’s(ed.)FiftyKeyThinkersontheEnvironmentandispastandfoundingpresidentoftheInternationalSocietyforEnvironmentalEthics.GeorgeMacDonaldRossisSeniorLecturerintheDepartmentofPhilosophyattheUniversityofLeedsandDirectorofthePhilosophicalandReligiousStudiesSubjectCentreoftheLearningandTeachingSupportNetwork.HehaswrittenextensivelyonLeibnizandotherseventeenth-andeighteenth-centuryphilosophers,andistheauthorofLeibniz(1984).R.M.SainsburyisStebbingProfessorofPhilosophyatKing’sCollege,LondonandwaseditorofthejournalMindforseveralyearsuntil2000.HehaspublishedRussell(1979),Paradoxes(1995)andLogicalForms(2000).Hismaininterestsareinphilosophicallogicandthephilosophyoflanguage.xvi\nCONTRIBUTORSJohnR.SearleisMillsProfessorofMindandLanguageattheUniversityofCaliforniawherehehasbeenafacultymembersince1959.Beforethat,hewasalectureratChristChurch,Oxford,andhereceivedallhisuniversitydegreesfromOxford.Mostofhisworkisinthephilosophyofmind,thephilosophyoflanguage,andsocialphilosophy.HismostrecentlypublishedbooksareRationalityinAction(2001)andMind,LanguageandSociety(1998).Heistheauthorofseveralotherimportantbooks,includingSpeechActs:AnEssayinthePhilosophyofLanguage(1969),ExpressionandMeaning:StudiesintheTheoryofSpeechActs(1985),Intentionality(1983),Minds,BrainsandScience,the1984ReithLectures(1989),TheRediscoveryofMind(1992)andTheConstructionofSocialReality(1995).N.E.SimmondsisaFellowofCorpusChristiCollege,Cambridgewherehelecturesinlaw.Hisinterestsincludethephilosophyoflawandpoliticalphilosophy.Hehaspub-lishedTheDeclineofJudicialReason(1984),CentralIssuesinJurisprudence(1986)andnumerousarticlesonthephilosophyoflaw.JohnSkorupskiisProfessorofMoralPhilosophyattheUniversityofStAndrews.HeistheauthorofJohnStuartMill(1989)andEnglish-LanguagePhilosophy1750–1945(1993).HismostrecentbookisEthicalExplorations(1999).ElliottSoberisHansReichenbachProfessorandHenryVilasResearchProfessorattheUniversityofWisconsin,MadisonandCentennialProfessorattheLondonSchoolofEconomics.HeistheauthorofTheNatureofSelection(1984),ReconstructingthePast(1988),PhilosophyofBiology(1993)and(withDavidS.Wilson)UntoOthers:EvolutionandPsychologyofUnselfishBehaviour(1988).TomSorellisProfessorofPhilosophyattheUniversityofEssex.HeistheauthorofHobbes(1986),Descartes(1987),MoralTheoryandCapitalPunishment(1987),Scientism(1991),(withJohnHendry)BusinessEthics(1994)andMoralTheoryandAnomaly(2000).HeistheeditorofTheRiseofModernPhilosophy(1993),TheCambridgeCom-paniontoHobbes(1995),HealthCare,Ethics,andInsurance(1998),Descartes(1999),and(withJohnRogers)HobbesandHistory(2000).CharlesTaliaferroisProfessorofPhilosophyatStOlafCollege,Northfield,MinnesotaandtheauthorofConsciousnessandtheMindofGod(1994),ContemporaryPhilosophyofReligion(1999)andtheco-editorofACompaniontoPhilosophyofReligion(1998).MaryTilesisProfessorofPhilosophyattheUniversityofHawaiiatManoa.Herinter-estsincludethehistoryandphilosophyofmathematics,scienceandtechnologyandtheirinteractionswithculture(EuropeanandChinese).ShehaspublishedLivinginaTechnologicalCulture(withHansOberdiek)(1995),AnIntroductiontoHistoricalEpiste-mology(withJamesTiles)(1993),MathematicsandtheImageofReason(1991)andBachelard:ScienceandObjectivity(1984).EricP.Tsui-JamesstudiedasapostgraduateatOrielCollege,Oxford.Hetaughtphi-losophyatStHilda’sCollege,Oxford,fortwoyearsbeforemovingtotheUniversityofHongKongin1990.Hehaspublishedworkonthemetaphysicsofmathematics,buthisresearchinterestsnowcentrearoundtheworkofWilliamJames,especiallythenineteenth-centurypsychologicalandphysiologicalcontextsofhisradicalempiricism.xvii\nCONTRIBUTORSRobertWardyteachesphilosophyandclassicsatStCatharine’sCollege,Cambridge.HehaspublishedinthefieldsofancientGreekphilosophyandrhetoric,Latinlitera-ture,thephilosophyoflanguageandChinesephilosophy.BernardWilliamsisMonroeDeutschProfessorofPhilosophy,UniversityofCalifor-nia,Berkeley,andwasWhite’sProfessorofMoralPhilosophyandaFellowofCorpusChristiCollege,Oxford.HisworksincludeMorality(1972),ProblemsoftheSelf(1973),Descartes:TheProjectofPureEnquiry(1978),MoralLuck(1981),EthicsandtheLimitsofPhilosophy(1985),ShameandNecessity(1993)andMakingSenseofHumanity(1995).R.S.WoolhouseisProfessorofPhilosophyattheUniversityofYork.HeistheauthorofLocke’sPhilosophyofScienceandKnowledge(1971),Locke(1983),TheEmpiricists(1988)andDescartes,Spinoza,Leibniz:TheConceptofSubstanceinSeventeenth-CenturyPhilosophy(1993).xviii\nContemporaryPhilosophyintheUnitedStatesJOHNR.SEARLEPhilosophyasanacademicdisciplineinAmericahasconsiderablyfewerpractitionersthandoseveralothersubjectsinthehumanitiesandthesocialsciences,suchassociology,history,English,oreconomics;butitstillshowsenormousdiversity.Thisvarietyismademanifestintheoriginalresearchpublishedbyprofessionalphilosophers,whosedifferingpointsofviewareexpressedinthelargenumberofbookspublishedeachyear,aswellasinthemanyprofessionalphilosophyjournals.ThereareovertwothousandcollegesanduniversitiesintheUnitedStates,ofwhichnearlyallhavephilosophydepartments,andthenumberofprofessionalphilosophersiscorrespondinglylarge.Becauseofthisdiversity,anygeneralizationsaboutthedisciplineasawhole,whichIamabouttomake,areboundtobemisleading.Thesubjectistoovastandcomplextobedescribableinasingleessay.Furthermore,anyonewhoisanactiveparticipantinthecurrentcontroversies,asIam,necessarilyhasaperspectiveconditionedbyhisorherowninterests,commitmentsandconvictions.Itwouldbeimpossibleformetogivean‘objective’account.Iamnotthereforeinwhatfollowstryingtogiveaneutralordisinterestedaccountofthecontemporaryphilosophicalscene;ratherIamtryingtosaywhatinthecurrentdevelopmentsseemstomeimportant.Inspiteofitsenormousvariety,therearecertaincentralthemesincontemporaryAmericanphilosophy.ThedominantmodeofphilosophizingintheUnitedStatesiscalled‘analyticphilosophy’.Withoutexception,thebestphilosophydepartmentsintheUnitedStatesaredominatedbyanalyticphilosophy,andamongtheleadingphilosophersintheUnitedStates,allbutatinyhandfulwouldbeclassifiedasanalyticphilosophers.Practitionersoftypesofphilosophizingthatarenotintheanalytictradition–suchasphenomenology,classicalpragmatism,existentialism,orMarxism–feelitnecessarytodefinetheirpositioninrelationtoanalyticphilosophy.Indeed,analyticphilosophyisthedominantmodeofphilosophizingnotonlyintheUnitedStates,butthroughouttheentireEnglish-speakingworld,includingGreatBritain,Canada,AustraliaandNewZealand.ItisalsothedominantmodeofphilosophizinginScandinavia,anditisalsobecomingmorewidespreadinGermany,France,ItalyandthroughoutLatinAmerica.IpersonallyhavefoundthatIcangotoallofthesepartsoftheworldandlectureonsubjectsincontemporaryanalyticphilosophybeforeaudienceswhoarebothknowledgeableandwelltrainedinthetechniquesofthediscipline.\nJOHNR.SEARLE1AnalyticPhilosophyWhat,then,isanalyticphilosophy?Thesimplestwaytodescribeitistosaythatitisprimarilyconcernedwiththeanalysisofmeaning.Inordertoexplainthisenterpriseanditssignificance,weneedfirsttosayalittlebitaboutitshistory.ThoughtheUnitedStatesnowleadstheworldinanalyticphilosophy,theoriginsofthismodeofphilosophizinglieinEurope.Specifically,analyticphilosophyisbasedontheworkofGottlobFrege,LudwigWittgenstein,BertrandRussellandG.E.Moore,aswellastheworkdonebythelogicalpositivistsoftheViennaCircleinthe1920sand1930s.Goingfurtherbackinhistory,onecanalsoseeanalyticphilosophyasanaturaldescend-antoftheempiricismofthegreatBritishphilosophersLocke,BerkeleyandHume,andofthetranscendentalphilosophyofKant.IntheworksofphilosophersasfarbackasPlatoandAristotle,onecanseemanyofthethemesandpresuppositionsofthemethodsofanalyticphilosophy.Wecanbestsummarizetheoriginsofmodernanalyticphilosophybysayingthatitarosewhentheempiricisttraditioninepistemology,togetherwiththefoundationalistenterpriseofKant,weretiedtothemethodsoflogicalanalysisandthephilosophicaltheoriesinventedbyGottlobFregeinthelatenineteenthcentury.Inthecourseofhisworkonthefoundationsofmathematics,Fregeinventedsymboliclogicinitsmodernformanddevelopedacomprehensiveandprofoundphilosophyoflanguage.Thoughmanyofthedetailsofhisviewsonlanguageandmathematicshavebeensuperseded,Frege’sworkiscrucialforatleasttworeasons.Firstly,byinventingmodernlogic,specificallythepredicatecalculus,hegaveusaprimarytoolofphilosophicalanalysis;and,secondly,hemadethephilosophyoflanguagecentraltotheentirephilosophicalenterprise.Fromthepointofviewofanalyticphilosophy,Frege’sworkisthegreatestsinglephilo-sophicalachievementofthenineteenthcentury.FregeantechniquesoflogicalanalysiswerelateraugmentedbytheordinarylanguageanalysisinspiredbytheworkofMooreandWittgensteinandarebestexemplifiedbytheschooloflin-guisticphilosophythatflourishedinOxfordinthel950s.Inshort,analyticphi-losophyattemptstocombinecertaintraditionalphilosophicalthemeswithmoderntechniques.Analyticphilosophyhasneverbeenfixedorstable,becauseitisintrinsicallyself-criticalanditspractitionersarealwayschallengingtheirownpresuppositionsandcon-clusions.However,itispossibletolocateacentralperiodinanalyticphilosophy–theperiodcomprising,roughlyspeaking,thelogicalpositivistphaseimmediatelypriortothe1939–45warandthepostwarphaseoflinguisticanalysis.Boththeprehistoryandthesubsequenthistoryofanalyticphilosophycanbedefinedbythemaindoctrinesofthatcentralperiod.Inthecentralperiod,analyticphilosophywasdefinedbyabeliefintwolinguisticdistinctions,combinedwitharesearchprogramme.Thetwodistinctionsare,firstly,thatbetweenanalyticandsyntheticpropositions,and,secondly,thatbetweendescriptiveandevaluativeutterances.Theresearchprogrammeisthetraditionalphilosophicalresearchprogrammeofattemptingtofindfoundationsforsuchphilosophicallyprob-lematicphenomenaaslanguage,knowledge,meaning,truth,mathematicsandsoon.2\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATESOnewaytoseethedevelopmentofanalyticphilosophyoverthepastthirtyyearsistoregarditasthegradualrejectionofthesetwodistinctions,andacorrespondingrejec-tionoffoundationalismasthecrucialenterpriseofphilosophy.However,inthecentralperiod,thesetwodistinctionsservednotonlytoidentifythemainbeliefsofanalyticphi-losophy,but,forthosewhoacceptedthemandtheresearchprogramme,theydefinedthenatureofphilosophyitself.1.1AnalyticversussyntheticThedistinctionbetweenanalyticandsyntheticpropositionswassupposedtobethedistinctionbetweenthosepropositionsthataretrueorfalseasamatterofdefinitionorofthemeaningsofthetermscontainedinthem(theanalyticpropo-sitions)andthosethataretrueorfalseasamatteroffactintheworldandnotsolelyinvirtueofthemeaningsofthewords(thesyntheticpropositions).Examplesofanalytictruthswouldbesuchpropositionsas‘Trianglesarethree-sidedplanefigures’,‘Allbachelorsareunmarried’,‘Womenarefemale’,‘2+2=4’andsoon.Ineachofthese,thetruthofthepropositionisentirelydeterminedbyitsmeaning;theyaretruebythedefinitionsofthewordsthattheycontain.Suchpropositionscanbeknowntobetrueorfalseapriori,andineachcasetheyexpressnecessarytruths.Indeed,itwasacharacteristicfeatureoftheanalyticphilosophyofthiscentralperiodthattermssuchas‘analytic’,‘necessary’,‘apriori’and‘tautological’weretakentobeco-extensive.Contrastedwiththeseweresyntheticpropositions,which,iftheyweretrue,weretrueasamatterofempiricalfactandnotasamatterofdefinitionalone.Thus,propositionssuchas‘TherearemorewomenthanmenintheUnitedStates’,‘Bachelorstendtodieearlierthanmarriedmen’and‘Bodiesattracteachotheraccordingtotheinversesquarelaw’areallsaidtobesyntheticpropositions,and,iftheyaretrue,theyexpressaposterioriempiricaltruthsabouttherealworldthatareindependentoflanguage.Suchempiricaltruths,accordingtothisview,arenevernecessary;rather,theyarecontingent.Forphilosophersholdingtheseviews,theterms‘aposteriori’,‘synthetic’,‘contingent’and‘empirical’weretakentobemoreorlessco-extensive.Itwasabasicassumptionbehindthelogicalpositivistmovementthatallmeaning-fulpropositionswereeitheranalyticorempirical,asdefinedbytheconceptionsthatIhavejuststated.Thepositivistswishedtobuildasharpboundarybetweenmeaningfulpropositionsofscienceandeverydaylifeontheonehand,andnonsensicalpropositionsofmetaphysicsandtheologyontheother.Theyclaimedthatallmeaningfulproposi-tionsareeitheranalyticorsynthetic:disciplinessuchaslogicandmathematicsfallwithintheanalyticcamp;theempiricalsciencesandmuchofcommonsensefallwithinthesyntheticcamp.Propositionsthatwereneitheranalyticnorempiricalpropositions,andwhichwerethereforeinprinciplenotverifiable,weresaidtobenonsensicalormeaningless.Thesloganofthepositivistswascalledtheverificationprinciple,and,inasimpleform,itcanbestatedasfollows:allmeaningfulpropositionsareeitheranalyticorsynthetic,andthosewhicharesyntheticareempiricallyverifiable.Thissloganwassometimesshortenedtoanevensimplerbattlecry:themeaningofapropositionisjustitsmethodofverification.3\nJOHNR.SEARLE1.2ThedistinctionbetweenevaluativeutterancesanddescriptiveutterancesAnotherdistinction,equallyimportantinthepositivistschemeofthings,isthedis-tinctionbetweenthoseutterancesthatexpresspropositionsthatcanbeliterallyeithertrueorfalseandthoseutterancesthatareusednottoexpresstruthsorfalsehoods,butrather,togiveventtoourfeelingsandemotions.Anexampleofadescriptivestatementwouldbe,‘Theincidenceofcrimesofthefthasincreasedinthepasttenyears’.Aninstanceoftheevaluativeclasswouldbe‘Theftiswrong’.Thepositivistsclaimedthatmanyutterancesthathadtheformofmeaningfulpropositionswereusednottostatepropositionsthatwereverifiableeitheranalyticallyorsynthetically,buttoexpressemo-tionsandfeelings.Propositionsofethicslookasiftheyarecognitivelymeaningful,buttheyarenot;theyhaveonly‘emotive’or‘evaluative’meaning.Thepropositionsofscience,mathematics,logicandmuchofcommonsensefallinthedescriptiveclass;theutterancesofaesthetics,ethicsandmuchofreligionfallintheevaluativeclass.Itisimportanttonotethatonthisconceptionevaluativepropositionsarenot,strictlyspeak-ing,eithertrueorfalse,sincetheyarenotverifiableaseitheranalyticorempirical.Thetwodistinctionsarecruciallyrelatedinthatallofthestatementsthatfallononesideortheotheroftheanalytic–syntheticdistinctionalsofallwithinthedescriptiveclassofthedescriptive–evaluativedistinction.Theimportancethatthesetwodistinctionshadfordefiningboththecharacterofthephilosophicalenterpriseandtherelationshipsbetweenlanguageandrealityishardtoexaggerate.Oneradicalconsequenceofthedistinctionbetweendescriptiveandevaluativepropositionswasthatcertaintraditionalareasofphilosophy,suchasethics,aestheticsandpoliticalphilosophy,werevirtuallyabolishedasrealmsofcognitivemeaningfulness.Propositionsintheseareaswere,forthemostpart,regardedasnon-sensicalexpressionsoffeelingsandemotions,becausetheyarenotutterancesthatcanbe,strictlyspeaking,eithertrueorfalse.Sincetheaimofphilosophersistostatethetruth,andsinceevaluativeutterancescannotbeeithertrueorfalse,itcannotbeoneoftheaimsofphilosophytomakeanyevaluativeutterances.Philosophersmightanalysethemeaningofevaluativeterms,andtheymightexaminethelogicalrela-tionshipsamongtheseterms,butphilosophers,quaphilosophers,canmakenofirst-orderevaluationsinaesthetics,ethicsorpolitics,asthesefirst-orderevaluationsarenot,strictlyspeaking,meaningful.Theymayhaveasortofsecondary,derivativemeaning,called‘emotivemeaning’,buttheylackscientificallyacceptablecognitivemeaning.Ifthetaskofphilosophyistostatethetruthandnottoprovideevaluations,whatthenisthesubjectmatterofphilosophy?Sincethemethodsofphilosophersarenotthoseofempiricalscience–sincetheirmethodsareaprioriratherthanaposteriori–itcannotbetheiraimtostateempiricaltruthsabouttheworld.Suchpropositionsarethepropositionsofthespecialsciences.Theaimofphilosophers,therefore,istostateanalytictruthsconcerninglogicalrelationsamongtheconceptsofourlanguage.Inthisperiodofphilosophy,thetaskofphilosophywastakentobethetaskofconceptualanalysis.Indeed,formostphilosopherswhoacceptedthisview,philosophyandcon-ceptualanalysiswerethesame.Wheretraditionalphilosophershadtakentheirtasktobethediscussionofthenatureofthegood,thetrue,thebeautifulandthejust,the4\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATESpositivistandpost-positivistanalyticphilosopherstooktheirtasktobetheanalysisofthemeaningofconceptssuchas‘goodness’,‘truth’,‘beauty’and‘justice’.Ideallytheanalysisoftheseandotherphilosophicallyinterestingconcepts,suchas‘knowledge’,‘certainty’and‘cause’,shouldgivenecessaryandsufficientconditionsfortheapplica-tionoftheseconcepts.Theysawthisasbeingthelegitimateheirofthetraditionalphi-losophicalenterprise,butanheirpurgedofthemetaphysicalnonsenseandconfusionthathaddiscreditedthetraditionalenterprise.Ifwecombinetheassumptionthatphilosophyisessentiallyaconceptual,analyticenterprisewiththeassumptionthatitstaskisfoundational–thatis,itstaskistoprovidesecurefoundationsforsuchthingsasknowledge–thentheconsequencefortheposi-tivistsisthatphilosophicalanalysistendsinlargeparttobereductive.Thatis,theaimoftheanalysisistoshow,forexample,howempiricalknowledgeisbasedon,andulti-matelyreducibleto,thedataofourexperience,toso-calledsensedata.(Thisviewiscalled‘phenomenalism’.)Similarly,statementsaboutthemindarebasedon,andthere-foreultimatelyreducibleto,statementsaboutexternalbehaviour(behaviourism).Nec-essarytruthissimilarlybasedonconventionsoflanguageasexpressedindefinitions(conventionalism);andmathematicsisbasedonlogic,especiallysettheory(logicism).Ineachcase,themorephilosophicallypuzzlingphenomenonisshowntohaveasecurefoundationinsomelesspuzzlingphenomenon,andindeed,theidealofsuchanalysiswastoshowthatthepuzzlingphenomenacouldbeentirelyreducedtolesspuzzlingphenomena.‘Phenomenalism’supposedlygavescienceasecurefoundationbecausesciencecouldbeshowntobefoundedonthedataofoursenses.Sincetheformofthereductionwasanalyticordefinitional,ithadtheconsequencethatstatementsaboutempiricalrealitycouldbetranslatedintostatementsaboutsensedata.Similarly,accord-ingtobehaviourism,statementsaboutmentalphenomenacouldbetranslatedintostatementsaboutbehaviour.Withinthecampofanalyticphilosopherswhothoughttheaimofphilosophywasconceptualanalysis,thereweretwobroadstreams.Onestreamthoughtordinarylanguagewasingeneralquiteadequate,bothasatoolandasasubjectmatterofphilosophicalanalysis.Theotherstreamthoughtofordinarylanguageashopelesslyinadequateforphilosophicalpurposes,andirretrievablyconfused.Thephilosophersofthislatterstreamthoughtthatweshouldusethetoolsofmodernmathematicallogicbothforanalysingtraditionalphilosophicalproblemsand,moreimportantly,forcreat-ingalogicallyperfectlanguage,forscientificandphilosophicalpurposes,inwhichcertaintraditionalconfusionscouldnotevenarise.Therewasneverarigiddistinctionbetweenthesetwostreams,buttherewerecertainlytwobroadtrends:onewhichemphasizedordinarylanguagephilosophyandonewhichemphasizedsymboliclogic.Bothstreams,however,acceptedthecentralviewthattheaimofphilosophywascon-ceptualanalysis,andthatinconsequencephilosophywasfundamentallydifferentfromanyotherdiscipline;theythoughtthatitwasasecond-orderdisciplineanalysingthelogicalstructureoflanguageingeneral,butnotdealingwithfirst-ordertruthsabouttheworld.Philosophywasuniversalinsubjectmatterpreciselybecauseithadnospecialsubjectmatterotherthanthediscourseofallotherdisciplinesandthediscourseofcommonsense.Afurtherconsequenceofthisconceptionwasthatphilosophybecameessentiallyalinguisticorconceptualenterprise.Forthatreason,thephilosophyoflanguagewas5\nJOHNR.SEARLEabsolutelycentraltothephilosophicaltask.Inasense,thephilosophyoflanguagewasnotonly‘firstphilosophy’;allofphilosophybecameaformofphilosophyoflanguage.Philosophywassimplythelogicalinvestigationofthestructureoflanguageasitwasusedinthevarioussciencesandincommonlife.2TheRejectionofTheseTwoDistinctionsandtheRejectionofFoundationalismWorkdoneinthe1950sand1960sledtotheovercomingofthesetwodistinctions;andwiththerejectionofthesetwodistinctionscameanewconceptionofanalyticphi-losophy–aconceptionthatemergedinthe1970sand1980sandwhichisstillbeingdeveloped.Therejectionofthesetwodistinctionsandofthefoundationalistresearchprogrammeledtoanenormousupheavalintheconceptionofthephilosophicalenter-priseandinthepracticeofanalyticphilosophers.Themostobviousproblemwithtra-ditionalanalyticphilosophywasthatthereductionistenterprisefailed.Ineverycase,theattemptstoprovidereductionistanalysesofthesortproposedbythephenomenal-istsandbehaviouristswereunsuccessful,andby1960thelackofsuccesswasobvious.Aseriesofimportanttheoreticaldevelopmentsalsotookplaceatthistime,butforthesakeofsimplicityIshallconcentrateononlyfiveofthese:Quine’srejectionoftheanalytic–syntheticdistinction,Austin’stheoryofspeechacts,Wittgenstein’scriticismoffoundationalism,Rawls’sworkinpoliticalphilosophyandthechangesinthephilosophyofscienceduetoKuhnandothers.2.1Quine’sattackontheanalytic–syntheticdistinctionPerhapsthemostimportantcriticismoftheanalytic–syntheticdistinctionwasmadebyW.V.O.Quineinafamousarticleentitled‘Twodogmasofempiricism’(Quine1953).Inthisarticle,Quineclaimedthatnoadequate,non-circulardefinitionofanalyticityhadeverbeengiven.Anyattempttodefineanalyticityhadalwaysbeenmadeusingnotionsthatwereinthesamefamilyasanalyticity,suchassynonymyanddefinition,andconsequently,theattemptstodefineanalyticitywereinvariablycircular.However,anevenmoreimportantobjectionthatemergedinQuine’sarticlewasthis:thenotionofananalyticpropositionissupposedtobeanotionofapropositionthatisimmunetorevision,thatisirrefutable.Quineclaimedthattherewerenopropositionsthatwereimmunetorevision,thatanypropositioncouldberevisedinthefaceofrecalcitrantevidence,andthatanypropositioncouldbeheldinthefaceofrecalcitrantevidence,providedthatonewaswillingtomakeadjustmentsinotherpropositionsoriginallyheldtobetrue.Quinearguedthatweshouldthinkofthelanguageofscienceasbeinglikeacomplexnetworkthatwasimpingeduponbyempiricalverificationonlyattheedges.Recalcitrantexperiencesattheedgesofsciencecanproducechangesanywherealongtheline,butthechangesarenotforcedonusbypurelylogicalconsiderations;rather,wemakevariouspragmaticorpracticaladjustmentsinthenetworkofoursentencesorbeliefstoaccommodatetheongoingcharacterofourexperiences.Language,onthisview,isnotatomistic.Itdoesnotconsistofasetofpropositions,eachofwhichcanbeassessedinisolation.Rather,itconsistsofaholisticnetwork,and,inthisnetwork,6\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATESpropositionsasgroupsconfrontexperience;propositionsindividuallyarenotsimplyassessedastrueorfalse.(ThisholismofscientificdiscoursewasinfluencedbytheFrenchphilosopherofscience,Duhem,andtheviewisfrequentlyreferredtoas‘theDuhem–Quinethesis’.)MostphilosopherstodayacceptsomeversionorotherofQuine’srejectionoftheanalytic–syntheticdistinction.Noteverybodyagreeswithhisactualargument(I,forone,donot),butnowthereisgeneralscepticismaboutourabilitytomakeastrictdis-tinctionbetweenthosepropositionsthataretruebydefinitionandthosethataretrueasamatteroffact.Therejectionoftheanalytic–syntheticdistinctionhasprofoundconsequencesforanalyticphilosophy,asweshallseeinmoredetaillater.Atthispointitisimportanttostatethatifthereisnowell-definedclassofanalyticpropositions,thenthephilosopher’spropositionscannotthemselvesbeclearlyidenti-fiedasanalytic.Theresultsofphilosophicalanalysiscannotbesharplydistinguishedfromtheresultsofscientificinvestigation.Onthepositivistpicture,philosophywasnotoneamongothersciences;rather,itstoodoutsidetheframeofscientificdiscourseandanalysedthelogicalrelationsbetween,ontheonehand,thatdiscourseanditsvocabu-laryand,ontheother,experienceandreality.Philosophers,sotospeak,analysedtherelationbetweenlanguageandreality,butonlyfromtheside.IfweacceptQuine’srejec-tionoftheanalytic–syntheticdistinction,thenphilosophyisnotsomethingthatcanbeclearlydemarcatedfromthespecialsciences.Itis,rather,adjacentto,andoverlapswith,otherdisciplines.Althoughphilosophyismoregeneralthanotherdisciplines,itspropositionsdonothaveanyspeciallogicalstatusorspeciallogicalprioritywithregardtotheotherdisciplines.2.2Austin’stheoryofspeechactsTheBritishphilosopherJ.L.Austinwassuspiciousofboththedistinctionbetweenana-lyticandsyntheticpropositions,andthedistinctionbetweenevaluativeanddescriptiveutterances.Duringthe1950shedevelopedanalternativeconceptionoflanguage(Austin1962).Hisfirstobservationwasthatthereisaclassofutterancesthatareobvi-ouslyperfectlymeaningful,butwhichdonotevensetouttobeeithertrueorfalse.Amanwhosays,forexample,‘Ipromisetocomeandseeyou’oraqualifiedauthoritywhosaystoacouple,‘Ipronounceyoumanandwife’isneitherreportingonnordescribingapromiseoramarriagerespectively.Suchutterancesshouldbethoughtnotascasesofdescribingorstating,butratherasdoing,asacting.Austinbaptizedtheseutterances‘performatives’andcontrastedthemwith‘constatives’.Thedistinctionbetweenconstativesandperformativeswassupposedtocontainthreefeatures:con-statives,butnotperformatives,couldbetrueorfalse;performatives,ontheotherhand,thoughtheycouldnotbetrueorfalse,couldbefelicitousorinfelicitous,dependingonwhetherornottheywerecorrectly,completelyandsuccessfullyperformed;andfinally,performativesweresupposedtobeactions,doingsorperformances,asopposedtomeresayingsorstatings.But,asAustinhimselfsaw,thedistinctionssodrawndidnotwork.Manyso-calledperformativesturnedouttobecapableofbeingtrueorfalse;forexample,warningscouldbeeithertrueorfalse.Andstatements,aswellasperforma-tives,couldbeinfelicitous.Forexample,ifonemadeastatementforwhichonehadinsufficientevidence,onewouldhavemadeaninfelicitousstatement.Andfinally,7\nJOHNR.SEARLEstatingisasmuchperforminganactionaspromisingororderingorapologizing.Theabandonmentoftheperformative–constativedistinctionledAustintoageneraltheoryofspeechacts.Communicativeutterancesingeneralareactionsofatypehecalled‘illocutionaryacts’.OnegreatmeritofAustin’stheoryofspeechactsisthatitenabledsubsequentphilosopherstoconstruethephilosophyoflanguageasabranchofthephilosophyofaction.Sincespeechactsareasmuchactionsasanyotheractions,thephilosophicalanalysisoflanguageispartofthegeneralanalysisofhumanbehaviour.Andsinceintentionalhumanbehaviourisanexpressionofmentalphenomena,itturnsoutthatthephilosophyoflanguageandthephilosophyofactionarereallyjustdifferentaspectsofonelargerarea,namely,thephilosophyofmind.Onthisview,thephilosophyoflan-guageisnot‘firstphilosophy’;itisabranchofthephilosophyofmind.ThoughAustindidnotlivetocarryouttheresearchprogrammeimplicitinhisinitialdiscoveries,subsequentwork,includingmyown,hascarriedthisresearchfurther.Bytreatingspeakingasaspeciesofintentionalactionwecangiveanewsensetoalotofoldquestions.Forexample,theoldquestion,‘Howmanykindsofutterancesarethere?’istoovaguetobeanswered.Butifweask‘Howmanykindsofillocutionaryactsarethere?’,wecangiveapreciseanswer,sincethequestionasks,‘Howmanypos-siblewaysarethereforspeakerstorelatepropositionalcontentstorealityintheper-formanceofactionsthatexpressillocutionaryintentions?’Ananalysisofthestructureofthoseintentionsrevealsfivebasictypesofillocutionaryact:wetellpeoplehowthingsare(Assertives),wetrytogetthemtodothings(Directives),wecommitourselvestodoingthings(Commissives),weexpressourfeelingsandattitudes(Expressives)andwebringaboutchangesintheworldthroughourutterances,sothattheworldischangedtomatchthepropositionalcontentoftheutterance(Declarations).(FordetailsseeSearle1979and1983.)2.3Wittgenstein’srejectionoffoundationalismThesinglemostinfluentialanalyticphilosopherofthetwentiethcentury,andindeed,thephilosopherwhommostanalyticphilosopherswouldregardasthegreatestphiloso-pherofthecentury,isLudwigWittgenstein.Wittgensteinpublishedonlyoneshortbookduringhislifetime,whichrepresentshisearlywork,butwiththeposthumouspublicationofhisPhilosophicalInvestigationsin1953,aseriesofhislaterwritingsbegantobecomeavailable.Now,wehaveasizeablecorpusoftheworkhedidinthelasttwentyyearsofhislife.Throughpainstakinganaly-sisoftheuseoflanguage,particularlythroughanalysisofpsychologicalconcepts,Wittgensteinattemptedtounderminetheideathatphilosophyisafoundationalenter-prise.Heasserted,onthecontrary,thatphilosophyisapurelydescriptiveenterprise,thatthetaskofphilosophyisneithertoreformlanguagenortotrytoplacethevarioususesoflanguageonasecurefoundation.Rather,philosophicalproblemsareremovedbyhavingacorrectunderstandingofhowlanguageactuallyfunctions.AkeynotioninWittgenstein’sconceptionoflanguageisthenotionofalanguagegame.Weshouldthinkofthewordsinlanguageasbeinglikethepiecesinagame.Theyarenottobeunderstoodbylookingforsomeassociatedideainthemind,orbyfollow-ingsomeprocedureofverification,orevenbylookingattheobjectforwhichtheystand.8\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATESRather,weshouldthinkofwordsintermsoftheiruse,andreferringtoobjectsintheworldisonlyoneofmanyusesthatwordshave.Themeaningofawordisgivenbyitsuse,andthefamilyofusesthatagroupofwordshasconstitutesalanguagegame.Examplesincludethelanguagegameweplayindescribingourownsensations,orthelanguagegameweplayinidentifyingthecausesofevents.ThisconceptionoflanguageleadsWittgensteintotherejectionoftheconceptionthatthetaskofphilosophicalanalysisisreductionistorfoundationalist.Thatis,Wittgensteinrejectstheideathatlan-guagegameseitherhaveorneedafoundationinsomethingelse,andherejectstheideathatcertainlanguagegamescanbereducedtocertainotherkindsoflanguagegames.Theeffect,Wittgensteinsays,ofphilosophicalanalysisisnottoalterourexistinglin-guisticpracticesortochallengetheirvalidity;itissimplytodescribethem.Languageneitherhasnorneedsafoundationinthetraditionalsense.IsaidthatWittgensteinwasthesinglemostinfluentialphilosopherintheanalytictradition,butthereisasenseinwhichitseemstomehehasstillnotbeenproperlyunderstood,norhadhislessonsbeenfullyassimilatedbyanalyticphilosophers.Iwillhavemoretosayabouthisinfluencelater.2.4Rawls’stheoryofjusticeTheconceptionofmoralphilosophyinthepositivistandpost-positivistphasesofana-lyticphilosophywasextremelynarrow.Strictlyspeaking,accordingtothepositivists,moralutterancescouldnotbeeithertrueorfalse,sotherewasnothingthatthephiloso-phercouldsay,quaphilosopher,bywayofmakingmoraljudgements.Thetaskforthemoralphilosopherwastoanalysemoraldiscourse,toanalysethemeaninganduseofmoraltermssuchas‘good’,‘ought’,‘right’,‘obligation’,etc.Itisimportanttoseethatthisconceptionofmoralphilosophywasastrictlogicalconsequenceoftheacceptanceofthedistinctionbetweenevaluativeanddescriptiveutterances.Forifevaluativeutter-ancescannotbeeithertrueorfalse,andiffirst-ordermoraldiscourseconsistsinevalu-ativeutterances,andifthetaskofthephilosopheristostatethetruth,itfollowsthatthephilosopher,quaphilosopher,cannotmakeanyfirst-ordermoraljudgements.Asaphilosopher,allheorshecandoisthesecond-ordertaskofanalysingmoralconcepts.Somephilosophersofthepositivistandpost-positivistperiodsrejectedthisnarrowconceptionofmoralphilosophy,andtherewereaseriesofattacksmountedonthedistinctionbetweenevaluativeanddescriptiveutterances,includingsomeattacksbymyselfinthemid-1960s(Searle1964).Itremained,however,forJohnRawlstoreopenthetraditionalconceptionofpoliticalandmoralphilosophywiththepublicationofhisbookATheoryofJusticein1971.Forthepurposesofthepresentdiscussion,theim-portantthingaboutRawls’sworkwasnotthatherefutedthetraditionaldichotomyofdescriptiveandevaluativeutterances,butthathesimplyignoreditandproceededtodevelopatheoryofpoliticalinstitutionsofasortthathasalongphilosophicaltradi-tionandwhichthepositiviststhoughttheyhadovercome.Rawls,ineffect,revivedthesocialcontracttheory,whichhadlongbeenassumedtobecompletelydefunct;buthediditbyaningeniousdevice:hedidnotattempt,assometraditionaltheoristshaddone,toshowthattheremighthavebeenanoriginalsocialcontract,nordidhetrytoshowthattheparticipationofindividualsinsocietyinvolvedatacitcontract.Rather,heusedthefollowingthoughtexperimentasananalytictool:thinkofthesortofsocietythat9\nJOHNR.SEARLErationalbeingswouldagreetoiftheydidnotknowwhatsortofpositiontheythem-selveswouldoccupyinthatsociety.Ifweimaginerationalbeings,hiddenbehindaveilofignorance,whoareaskedtoselectandagreeonformsofsocialinstitutionsthatwouldbefairforall,thenwecandevelopcriteriaforappraisingsocialinstitutionsonpurelyrationalgrounds.TheimportanceofRawlsforourpresentdiscussionisnotwhetherhesucceededindevelopingnewfoundationsforpoliticaltheory,butthefactthathisworkgaverisetoarenewedinterestinpoliticalphilosophy,whichwassoonaccompaniedbyarenewedinterestinthetraditionalquestionsofmoralphilosophy.Moralandpoliticalphiloso-phyhadbeenconfinedtoaverysmallrealmbythepositivistphilosophers,andforthatreasonseemedsterileanduninteresting.Verylittleworkwasdoneinthatarea,butsincethe1970sithasgrownenormously,andisnowaflourishingbranchofanalyticphilosophy.2.5Post-positivistphilosophyofscienceThroughoutthepositivistperiodthemodelofempiricalknowledgewasprovidedbythephysicalsciences,andthegeneralconceptionwasthattheempiricalsciencesproceededbythegradualbutcumulativegrowthofempiricalknowledgethroughthesystematicapplicationofscientificmethod.Thereweredifferentversionsofscientificmethod,accordingtothephilosophersofthatperiod,buttheyallsharedtheideathatscientific,empiricalpropositionsareessentially‘testable’.Initiallyapropositionwasthoughttestableifitcouldbeconfirmed,butthemostinfluentialversionofthisideaisPopper’sclaimthatempiricalpropositionsaretestableiftheyarefalsifiableinprinciple.Thatis,inorderforapropositiontotellushowtheworldisasopposedtohowitmightbeormighthavebeen,theremustbesomeconceivablestateofaffairsthatwouldrenderthatpropositionfalse.Propositionsofscienceare,strictlyspeaking,neververifiable–theysimplysurviverepeatedattemptsatfalsification.Scienceisinthissensefallible,butitisatthesametimerationalandcumulative.ThispictureofthehistoryofsciencewasverydramaticallychallengedinThomasKuhn’sbookTheStructureofScientificRevolutions(1962).AccordingtoKuhn,thehistoryofscienceshowsnotagradualandsteadyaccumulationofknowledgebutperiodicrevolutionaryoverthrowsofpreviousconceptionsofreality.TheshiftfromAristotelianphysicstoNewtonianphysics,andtheshiftfromNewtonianphysicstorelativisticphysicsarebothillustrationsofhowone‘paradigm’isreplacedbyanother.Whentheburdenofpuzzlingcaseswithinoneparadigmbecomesunbearable,anewparadigmemerges,whichprovidesnotjustanewsetoftruthsbutawholenewwayoflookingatthesubjectmatter.‘Normalsciences’alwaysproceedbypuzzle-solvingwithinaparadigm,butrevolutionarybreakthroughs,ratherthanpuzzle-solvingwithinaparadigm,aremattersofoverthrowingoneparadigmandreplacingitwithanother.JustasKuhnchallengedthepictureofscienceasessentiallyamatterofasteadyaccumulationofknowledge,soPaulFeyerabendchallengedtheconceptionoftherebeingaunitaryrational‘scientificmethod’(Feyerabend1975).Feyerabendtriedtoshowthatthehistoryofsciencerevealsnotasinglerationalmethodbutratheraseriesofopportunistic,chaotic,desperate(andsometimesevendishonest)attemptstocope10\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATESwithimmediateproblems.ThelessonthatFeyerabenddrawsfromthisisthatweshouldabandontheconstrainingideaoftherebeingsuchathingasasingle,rationalmethodthatapplieseverywhereinscience;rather,weshouldadoptan‘anarchistic’view,accordingtowhich‘anythinggoes’.ReactionstoKuhnandFeyerabend,notsurpris-ingly,differenormouslyamonganalyticphilosophers.Kuhnsometimesseemstobearguingthatthereisnotanysuchthingastherealworldexistingindependentlyofourscientifictheories,whichitistheaimofourscientifictheoriestorepresent.Kuhn,inshort,seemstobedenyingrealism.Mostphilosophersdonottakethisdenialofrealismatallseriously.EvenifKuhnwererightaboutthestructureofscientificrevolutions,thisinnowayshowsthatthereisnoindependentrealitythatscienceisinvestigating.Again,mostphilosopherswouldacceptFeyerabend’srecognitionofavarietyofmethodsusedinthehistoryofscience,butveryfewpeopletakeseriouslytheideathattherearenorationalconstraintsoninvestigationwhatever.Nonetheless,theeffectoftheseauthorshasbeenimportantinatleastthefollowingrespect.Thepositivists’con-ceptionofscienceasasteadyaccumulationoffactualknowledge,andofthetaskofthephilosopherastheconceptualanalysisofscientificmethod,hasgivenwaytoanattitudetosciencethatisatoncemorescepticalandmoreactivist.Itismorescepticalinthesensethatfewphilosophersarelookingfortheonesinglemethodthatpervadeseveryenterprisecalled‘science’,butitismoreactivistinthesensethatphilosophyofscienceinteractsmoredirectlywithscientificresults.Forexample,recentphilosophi-caldiscussionsaboutquantummechanics,oraboutthesignificanceofBell’stheoremwithinquantummechanics,revealthatitisnowimpossibletosayexactlywheretheprobleminphysicsendsandtheprobleminphilosophybegins.Thereisasteadyinter-actionandcollaborationbetweenphilosophyandscienceonsuchphilosophicallypuzzlingquestions.3SomeRecentDevelopmentsTheresultsofthechangesthatIhavejustoutlinedaretomakeanalyticphilosophyontheonehandamoreinterestingdiscipline,butontheotherhandamuchlesswell-definedresearchproject.Inthewaythattheverificationprincipleformedthecoreideologyofthelogicalpositivistsandinthewaythattheconceptualanalysisformedthecoreresearchprojectofthepost-positivisticanalyticphilosopher,thereisnownoideologicalpointofreferencethatiscommonlyagreedupon;noristhereauniversallyacceptedresearchprogramme.Forexample,conceptualanalysisthirtyyearsagowastakentobetheheartofanalyticphilosophy,butnowmanyphilosopherswoulddenythatitisthecentralelementinthephilosophicalenterprise.Somephilosophers,indeed,wouldsaythatthetraditionalenterpriseofattemptingtofindlogicallynecessaryandsufficientconditionsfortheapplicabilityofaconceptismisconceivedinprinciple.TheythinkthepossibilityofsuchanenterprisehasbeenrefutedbyQuine’srefutationoftheanalytic–syntheticdistinction,aswellasWittgenstein’sobservationthatmanyphilo-sophicallypuzzlingconceptshavenotacentralcoreoressenceofmeaning,butavarietyofdifferentusesunitedonlybya‘familyresemblance’.Manyotherphilosopherswouldsaythatconceptualanalysisisstillanessentialpartofthephilosophicalenter-prise,asindeedithasbeensincethetimeofPlato’sdialogues,butitisnolongerseen11\nJOHNR.SEARLEtobethewholeoftheenterprise.Philosophyisnow,Ibelieve,amuchmoreinterestingsubjectthanitwasagenerationagobecauseitisnolongerseenassomethingseparatefrom,andsealedofffrom,otherdisciplines.Inparticular,philosophyisnowseenbymostanalyticphilosophersasbeingadjacenttoandoverlappingwiththesciences.Myownview,whichIfeelisfairlywidelyshared,isthatwordslike‘philosophy’and‘science’areinmanyrespectsmisleading,iftheyaretakentoimplytheexistenceofmutuallyexclusiveformsofknowledge.Rather,itseemstomethatthereisjustknowl-edgeandtruth,andthatinintellectualenterprisesweareprimarilyaimingatknowl-edgeandtruth.Thesemaycomeinavarietyofforms,whetherinhistory,mathematics,physics,psychology,literarycriticismorphilosophy.Philosophytendstobemoregeneralthanothersubjects,moresynopticinitsvision,moreconceptuallyorlogicallyorientedthanotherdisciplines,butitisnotadisciplinethatishermeticallysealedofffromothersubjects.Theresultisthatmanyareasofinvestigationwhichwerelargelyignoredbyanalyticphilosophersagenerationagohavenowbecomethrivingbranchesofphilosophy,includingcognitivescience,thephilosophyofbiologyandthephiloso-phyofeconomics.Inwhatfollows,Iwillconfinemydiscussiontofivemajorareasofphilosophicalresearch:cognitivescience,thecausaltheoryofreference,intentionalis-tictheoriesofmeaning,truth-conditionaltheoriesofmeaning,andWittgenstein’sconceptionoflanguageandmindandhisresponsetoscepticism.3.1PhilosophyandcognitivescienceNowhereisthenewperiodofcollaborationbetweenphilosophyandotherdisciplinesmoreevidentthaninthenewsubjectofcognitivescience.Cognitivesciencefromitsverybeginningshasbeen‘interdisciplinary’incharacter,andisineffectthejointprop-ertyofpsychology,linguistics,philosophy,computerscienceandanthropology.Thereis,therefore,agreatvarietyofdifferentresearchprojectswithincognitivescience,butthecentralareaofcognitivescience,itshardcoreideology,restsontheassumptionthatthemindisbestviewedasanalogoustoadigitalcomputer.Thebasicideabehindcog-nitivescienceisthatrecentdevelopmentsincomputerscienceandartificialintelligencehaveenormousimportanceforourconceptionofhumanbeings.Thebasicinspirationforcognitivesciencewentsomethinglikethis:humanbeingsdoinformationprocess-ing.Computersaredesignedpreciselytodoinformationprocessing.Thereforeonewaytostudyhumancognition–perhapsthebestwaytostudyit–istostudyitasamatterofcomputationalinformationprocessing.Somecognitivescientiststhinkthatthecom-puterisjustametaphorforthehumanmind;othersthinkthatthehumanmindisliterallyacomputerprogram.Butitisfairtosaythatwithoutthecomputationalmodeltherewouldnothavebeenacognitivescienceaswenowunderstandit.Thisconceptionofhumancognitionwasideallysuitedtothetwentieth-centuryanalytictraditioninphilosophyofmindbecauseoftheanalytictradition’sresolutematerialism.Itwasanti-mentalisticandanti-dualistic.Thefailureoflogicalbehav-iourismlednottoarevivalofdualismbuttomoresophisticatedversionsofmaterial-ism.Iwillnowbrieflysummarizesomeoftherecentdevelopmentsinmaterialisticphilosophiesofmindthatledtothecomputationaltheoryofthemind.Thelogicalbehaviourists’thesiswassubjecttomanyobjections,themostimportantbeingtheobjectionthatitignoresinternalmentalphenomena.Inscienceandcommon12\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATESsenseitseemsmorenaturaltothinkofhumanbehaviourasbeingcausedbyinternalmentalstatesratherthantothinkofthementalstatesassimplyconsistingofthebehaviour.Thisweaknessinbehaviourismwascorrectedbythematerialistidentitythesis,sometimescalled‘physicalism’.Accordingtothephysicalistidentitytheory,mentalstatesareidenticalwithstatesofthebrain.Wedonotknowindetailwhattheseidentitiesare,buttheprogressoftheneurosciencesmakesitseemoverwhelminglyprobablethateverymentalstatewillbediscoveredtobeidenticalwithsomebrainstate.Intheearlyversionoftheidentitythesisitwassupposedthateverytypeofmentalstatewouldbediscoveredtobeidenticalwithsometypeofphysicalstate,butaftersomedebatethisbegantoseemmoreandmoreimplausible.Thereisnoreasontosupposethatonlysystemswithneuronslikeourscanhavementalstates;indeed,thereisnoreasontosupposethattwohumanbeingswhohavethesamebeliefmustthereforebeinthesameneurophysiologicalstate.So,‘type–typeidentitytheory’naturallygavewayto‘token–tokenidentitytheory’.Thetokenidentitytheoristsclaimedthateverypar-ticularmentalstateisidenticalwithsomeparticularneurophysiologicalstate,evenifthereisnotypecorrelationbetweentypesofmentalstatesandtypesofphysicalstates.Butthatonlyleavesopenthequestion,‘Whatisitthattwodifferentneurophysiologi-calstateshaveincommoniftheyareboththesamementalstate?’Tomanyanalyticphilosophersitseemedobviousthattheanswertoourquestionmustbethattwoneurophysiologicalstatesarethesametypeofmentalstateiftheyservethesamefunctionintheoverallecologyoftheorganism.Mentalstatesonthisviewcanbedefinedintermsoftheircausalrelationstoinputstimuli,toothermentalstates,andtoexternalbehaviour.Thisviewiscalled‘functionalism’anditisanaturaldevelopmentfromtoken–tokenidentitytheory.However,thefunctionalisthastoanswerafurtherobviousquestion:‘Whatisitaboutthestatesthatgivesthemthecausalrelationsthattheydohave?’Ifmentalstatesaredefinedintermsoftheircausalrelations,thenwhatisitaboutthestructureofdif-ferentneurophysiologicalconfigurationsthatcangivethemthesamecausalrelations?Itisatpreciselythispointthatthetraditionofmaterialisminanalyticphilosophycon-vergeswiththetraditionofartificialintelligence.ThecomputerprovidesanobviousanswertothequestionthatIhavejustposed.Thedistinctionbetweenthesoftwareandthehardware,theprogramandthephysicalsystemthatimplementstheprogram,pro-videsamodelforhowfunctionallyequivalentelementsatahigherlevelcanberealizedinorimplementedbydifferentphysicalsystemsatalowerlevel.Justasoneandthesameprogramcanbeimplementedbyquitedifferentphysicalhardwaresystems,sooneandthesamesetofmentalprocessescanbeimplementedindifferentneurophysiologi-calorotherformsofhardwareimplementations.Indeed,onthemostextremeversionofthisview,themindistothebrainastheprogramistothehardware.Thissortoffunctionalismcametobecalled‘computerfunctionalism’or‘Turingmachinefunc-tionalism’,anditcoincideswiththestrongversionof‘artificialintelligence’(StrongAI),theversionthatsayshavingamindjustishavingacertainsortofprogram.IhaverefutedStrongAIinaseriesofarticles(Searle1980a,1980b).Thebasicideaofthatrefutationcanbestatedquitesimply.Mindscannotbeequivalenttoprogramsbecauseprogramsaredefinedpurelyformallyorsyntacticallyandmindshavementalcontents.Theeasiestwaytoseetheforceoftherefutationistoseethatasystem,sayoneself,couldlearntomanipulatetheformalsymbolsforunderstandinganatural13\nJOHNR.SEARLElanguagewithoutactuallyunderstandingthatlanguage.ImighthaveaprogramthatenablesmetoanswerquestionsinChinesesimplybymatchingincomingsymbolswiththeappropriateprocessingandoutputsymbols,butnonethelessIstillwouldnottherebyunderstandChinese.However,thoughtheprojectofcomputerfunctionalismisalmostcertainlyafailure,theresultsoftheenterpriseareinmanyrespectsquiteuseful.Importantthingscanbelearnedaboutthemindbypursuingthecomputermetaphor,andtheresearchefforthasnotnecessarilybeenwasted.Themostexcitingrecentdevelopmenthasbeentothinkofmentalprocessesnotonthemodelofthecon-ventionalserialdigitalcomputer,butrathertothinkofbrainprocessesonthemodelofparalleldistributedprocessingcomputers.Themostexcitingrecentdevelopment,inmyview,incognitivesciencehasbeenthedevelopmentofsuch‘neuralnetmodels’forhumancognition.Inconcludingthissection,Iwanttopointoutthatinmyviewthechiefweaknessofanalyticalphilosophyofmind,aweaknessitshareswiththepast300yearsinthephilosophyofmind,hasbeenitsassumptionthatthereissomehowaninconsistencybetweenmentalismandmaterialism.Analyticphilosophers,alongwiththerestoftheCartesiantradition,havecharacteristicallyassumedthat‘mental’implies‘non-material’or‘immaterial’andthat‘material’or‘physical’implies‘non-mental’.Butifonereflectsonhowthebrainworks,itseemsthatbothoftheseassumptionsareobviouslyfalse.Whatthatshowsisthatourwholevocabulary,ourwholeterminologyofthementalandphysical,needswholesalerevision.3.2ThecausaltheoryofreferenceAcentralquestioninanalyticphilosophyoflanguage,sinceFrege(andindeedinphi-losophysincethetimeofPlato),hasbeen:Howdoeslanguagerelatetotheworld?Howdowordshookontothings?Inansweringthisquestion,theanalytictraditionhadchar-acteristicallyfoundaconnectionbetweenthenotionofreferenceandthenotionoftruth.Anexpression,suchasapropername,referstoorstandsforordesignatesanobjectbecauseassociatedwiththatnameissomedescriptivecontent,someconceptoftheobjectinquestion,andtheobjectinquestionsatisfiesorfitsthatdescriptivecontent.Theexpressionreferstotheobjectonlybecausethedescriptionistrueoftheobject.ThisisthestandardreadingofFrege’sfamousdistinctionbetweensenseandreference,betweenSinnandBedeutung.Expressionsrefertoobjectsinvirtueoftheirsenseandthesenseprovidesadescription,a‘modeofpresentation’,oftheobjectinquestion.Somethinganalogouslyapplieswithgeneralterms:generaltermsaretrueofanobjectbecauseeachgeneraltermhasassociatedwithitaclusteroffeatures,andthetermwillbetrueoftheobjectiftheobjectinquestionhasthosefeatures.Inthe1970sthisconceptionoftherelationbetweenlanguageandrealitywasattackedbyanumberofphilosophers,mostprominentlyDonnellan(1970),Kripke(1972)andPutnam(1975).Avarietyofargumentsweremountedagainstthetradi-tionalconceptionofmeaningandreference,butthecommonthreadrunningthroughtheseargumentswasthatthedescriptivecontentassociatedwithawordprovidedneithernecessarynorsufficientconditionsforitsapplication.Aspeakermightrefertoanobjecteventhoughtheassociateddescriptionthatheorshehadwasnottrueofthatobject;aspeakermighthaveadescriptionthatwassatisfiedbyanobjecteventhough14\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATESthatwasnottheobjecttowhichheorshewasreferring.ThemostfamousversionofthisargumentwasPutnam’s‘twinearth’example.Imagineaplanetinadistantgalaxyexactlylikeoursineveryrespectexceptthatonthisplanetwhattheycall‘water’hasadifferentchemicalcomposition.ItisnotcomposedofH2Obuthasanextremelycom-plicatedformulathatwewillabbreviateas‘XYZ’.Priorto1750,priortothetimethatanyoneknewthechemicalcompositionofwater,thepeopleontwinearthhadintheirmindsexactlythesameconceptofwaterasthepeopleonearth.Nonethelessourword‘water’doesnotrefertothestuffontwinearth.Ourword‘water’,whetherornotweknewitin1750,referstoH2O;andthisisamatterofobjectivecausalrelationsintheworldwhichareindependentoftheideasthatpeoplehaveintheirheads.Meaningsonthisviewarenotconceptsinpeople’sheads,butobjectiverelationsintheworld.Well,ifassociatedideasarenotsufficientformeaning,whatis?TheanswergivenbythethreeauthorsIhavementionedisthattheremustbesomesortofcausalconnectionbetweentheuseofthewordandtheobjectortypeofentityintheworldthatitappliesto.Thus,ifIusetheword‘Socrates’,itreferstoacertainGreekphilosopheronlybecausethereisacausalchainconnectingthatphilosopherandmycurrentuseoftheword.Theword‘water’isnotdefinedbyanychecklistoffeatures;rather,‘water’referstowhat-everstuffintheworldwascausallyrelatedtocertainoriginalusesoftheword‘water’,andtheseusessubsequentlycametobeacceptedinthecommunityandwerethenpasseddownthroughacausalchainofcommunication.Thereisaverynaturalwayofconnectingthecomputerfunctionalistconceptionofthemindwiththecausaltheoryofreference.Ifthemindwereacomputerprogram,andifmeaningwereamatterofcausalconnectionstotheworld,thenthewaythemindacquiresmeaningsisforthesystemthatimplementsthecomputerprogramtobeinvolvedincausalinteractionswiththeworld.3.3IntentionalistictheoriesofmeaningMuchofthebestworkinspeechacttheorydoneafterthepublicationofAustin’sHowtoDoThingswithWordsin1962,andmySpeechActsin1969,attemptedtocombinetheinsightsofPaulGrice’saccountofmeaningwiththeframeworkprovidedbythetheoryofspeechacts.Inaseriesofarticlesbeginninginthelate1950s(Grice1957,1968),Gricehadarguedthatthereisacloseconnectionbetweenthespeaker’sinten-tionsintheperformanceofanutteranceandthemeaningofthatutterance.Inhisoriginalformulationofthisview,Griceanalysedthespeaker’smeaningintermsoftheintentiontoproduceaneffectonthehearerbymeansofgettingthehearertorecog-nizetheintentiontoproducethatveryeffect.Thus,forexample,accordingtoGrice,ifaspeakerintendstotellahearerthatitisraining,theninthespeaker’sutteranceofthesentence,‘Itisraining’,thespeaker’smeaningwillconsistofhisorherintentiontoproduceinthehearerthebeliefthatitisrainingbymeansofgettingthehearertorecognizehisorherintentiontoproducethatverybelief.SubsequentworkbyGricealteredthedetailsofthisaccount,butthegeneralprincipleremainedthesame:meaningisamatterofaself-referentialintentiontoproduceaneffectonahearerbygettingthehearertorecognizetheintentiontoproducethateffect.Gricecombinedthisanalysisofmeaningwithananalysisofcertainprinciplesofconversationalco-operation.Inconversation,peopleacceptcertaintacitprinciples,whichGricecalls15\nJOHNR.SEARLE‘MaximsofConversation’–theyaccepttheprinciplesthatthespeaker’sremarkswillbetruthfulandsincere(themaximofquality),thattheywillberelevanttotheconversationalpurposesathand(themaximofrelation),thatthespeakerwillbeclear(themaximofmanner)andthatthespeakerwillsayneithermorenorlessthanisnecessaryforthepurposesoftheconversation(themaximofquantity).TherehasbeenagreatdealofcontroversyaboutthedetailsofGrice’sanalysisofmeaning,butthebasicideathatthereisacloseconnectionbetweenmeaningandintentionhasbeenacceptedandhasprovedimmenselyusefulinanalysingthestruc-tureofcertaintypicalspeechactphenomena.MyownviewisthatGriceconfusesthatpartofmeaningwhichhastodowithrepresentingcertainstatesofaffairsandcertainillocutionarymodes,andthatpartofmeaningthathastodowithcommunicatingthoserepresentationstoahearer.Grice,inshort,confusescommunicationwithrepresentation.However,thecombinationofanintentionalisticaccountofmeaning,togetherwithrationalprinciplesofco-operation,isimmenselyfruitfulinanalysingsuchproblemsasthoseof‘indirectspeechacts’andfigurativeusesoflanguagesuchasmetaphors.So,forexample,inanindirectspeechact,aspeakerwillcharacteristi-callymeansomethingmorethanwhatheorsheactuallysays.Totakeasimpleexample,inadinnertablesituationaspeakerwhosays‘Canyoupassthesalt?’wouldusuallynotjustbeaskingaquestionaboutthesalt-passingabilitiesofthehearer;heorshewouldberequestingthehearertopassthesalt.Nowthepuzzleisthis:howisitthatspeakersandhearerscommunicatesoeffortlesslywhenthereisabiggulfbetweenwhatthespeakermeansandwhatheorsheactuallysays?Inthecaseofmetaphor,asimilarquestionarises:howdoesthespeakercommunicatesoeffortlesslyhisorhermetaphoricalmeaningwhentheliteralmeaningofthesentenceuttereddoesnotencodethatmetaphoricalmeaning?AgreatdealofprogresshasbeenmadeontheseandotherproblemsusingtheapparatusthatGricecontributedtothetheoryofspeechacts.Oneofthemarksofprogressinphilosophyisthattheresultsofphilosophicalanaly-sistendtobeappropriatedbyotherdisciplines,andthishascertainlyhappenedwithspeechacttheory.Speechacttheoryisnowathrivingbranchofthedisciplineoflin-guistics,andtheworksofAustinandGrice,aswellasmyown,areaswellknownamonglinguistsastheyareamongphilosophers.3.4Truth-conditionaltheoriesofmeaningPhilosopherssuchasQuineandhisformerstudent,DonaldDavidson,havealwaysfeltthatintentionalistictheoriesofmeaningofthesortproposedbyGriceandSearlewerephilosophicallyinadequate,becausetheintentionalisticnotionsseemedaspuzzlingasthenotionofmeaningitselfandbecausetheycouldnecessarilyinvolvelinguisticmeaningintheirultimateanalyses.SoQuineandDavidsonattemptedtogiveaccountsofmeaningthatdidnotemploytheusualapparatusofintentionality.Themostinflu-entialversionofthisattemptisDavidson’sprojectofanalysingmeaningintermsoftruthconditions.Thebasicideaisthatoneknowsthemeaningofasentenceifoneknowsunderwhatconditionsitistrueorfalse.Thus,oneknowsthemeaningoftheGermansentence‘Schneeistweiss’ifoneknowsthatitistrueifandonlyifsnowiswhite.Nowsinceatheoryofmeaningforalanguageshouldbeabletostatethe16\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATESmeaningofeachofthesentencesofthelanguage,andsincethemeaningsofthesen-tencesofthelanguagearegivenbytruthconditions,andsincetruthconditionscanbespecifiedindependentlyoftheintentionalisticapparatus,itseemstoDavidsonthatatheoryoftruth(thatis,atheoryofthetruthconditionsofthesentences)ofalanguagewouldprovideatheoryofmeaningforthatlanguage.Inordertocarryouttheprojectofexplainingmeaningintermsoftruth,DavidsonemploystheapparatusofTarski’ssemanticdefinitionoftruth,adefinitionthatTarskihadworkedoutinthe1930s.TarskipointsoutthatitisaconditionofadequacyonanyaccountoftruththatforanysentencesandanylanguageL,theaccountmusthavetheconsequencethatsistrueinLifandonlyifp,whereforscanbesubstitutedthestructuraldescriptionofanysentencewhatever,forL,thenameofthelanguageofwhichsisapart,andforp,thesentenceitselforatrans-lationofit.Thus,forexample,inEnglish,thesentence‘Snowiswhite’istrueifandonlyifsnowiswhite.Thisconditionisusuallycalled‘conventionT’andthecorre-spondingsentencesarecalled‘T-sentences’.NowDavidsonnotesthatconventionTemploysthefactthatthesentencenamedbyshasthesamemeaningasthesentenceexpressedbyp,andthusTarskiisusingthenotionofmeaninginordertodefinethenotionoftruth.Davidsonproposestoturnthisprocedurearoundbytakingthenotionoftruthforgranted,bytakingitasaprimitive,andusingittoexplainmeaning.Hereishowitworks.Davidsonhopestogetatheoryofmeaningforaspeakerofalanguagethatwouldbesufficienttointerpretanyofthespeaker’sutterancesbygettingatheorythatwouldprovideasetofaxiomswhichwouldentailalltrueT-sentencesforthatspeaker’slanguage.Thus,ifthespeakerspeaksGerman,andweuseEnglishasameta-languageinwhichtostatethetheoryofthespeaker’slanguage,Davidsonclaimswewouldhaveanadequatetheoryofthespeaker’slanguageifwecouldgetasetofaxiomswhichwouldentailatrueT-sentencestatedinEnglishforanysentencethatthespeakerutteredinGerman.Thus,forexample,ourtheoryofmeaningshouldcontainaxiomswhichentailthatthespeaker’sutterance‘Schneeistweiss’istrueinthespeaker’slanguageifandonlyifsnowiswhite.Davidsonfurtherclaimsthatwecouldmakethisintoanempiricaltheoryofthespeaker’slanguagebyproceedingtoassoci-atethespeaker’sutteranceswiththecircumstancesinwhichwehadempiricalevidenceforsupposingthatthespeakerheldthoseutterancestobetrue.Thus,ifwehearthespeakerutterthesentence‘Esregnet’,wemightlookaroundandnotethatitwasraininginthevicinity,andwemightthenformthehypothesisthatthespeakerholdstruethesentence‘Esregnet’whenitisraininginhisorherimmediatevicinity.Thiswouldprovidethesortofempiricaldataonwhichwewouldbegintoconstructatheoryoftruthforthespeaker’slanguage.ItisimportanttonotethatwearetothinkofthisasathoughtexperimentandnotasanactualprocedurethatwehavetoemploywhenwetrytolearnGerman,forexample.Theideaistocashoutthenotionofmeaningintermsoftruthconditions,andthencashoutthenotionoftruthconditionsintermsofatruththeoryforalan-guage,whichisatheorythatwouldentailallthetrueT-sentencesofthelanguage.The17\nJOHNR.SEARLEempiricalbasisonwhichthewholesystemrestsisthatoftheevidencewecouldgetconcerningtheconditionsunderwhichaspeakerholdsasentencetobetrue.Iftheprojectcouldinprinciplebecarriedout,thenwewouldhavegivenanaccountofmeaningwhichemployedonlyoneintentionalisticnotion,thenotionof‘holdingtrue’asentence.Overthepasttwentyyearstherehasbeenquiteanextensiveliteratureonthenatureofthisprojectandhowitmightbeappliedtoseveraldifficultandpuzzlingsortsofsentences–forexample,indexicalsentences,sentencesaboutmentalstatesormodalsentences.Enthusiasmforthisprojectseemstohavewanedsomewhatinrecentyears.Inmyview,thecentralweaknessofDavidson’senterpriseisasfollows:anytheoryofmeaningmustexplainnotonlywhataspeakerrepresentsbyhisorherutterances,butalsohowheorsherepresentsthem,underwhatmentalaspectsthespeakerrepre-sentstruthconditions.Forthisreason,atheoryofmeaningcannotjustcorrelateaspeaker’sutterancewithstatesofaffairsintheworld;itmustexplainwhatisgoingoninthespeaker’sheadwhichenablesthespeakertorepresentthosestatesofaffairsundercertainaspectswiththeutterancesthatthespeakermakes.Thus,forexample,supposethatsnowiscomposedofH2Omoleculesincrystallineform,andsupposethecolourwhiteconsistsoflightwaveemissionsofallwavelengths,thenthesentence‘Schneeistweiss’istrueifandonlyifH2Omoleculesincrystallineformemitlightofallwavelengths.NowthissecondT-sentenceisjustasempiricallysubstantiatedastheearlierexample,‘Schneeistweiss’istrueifandonlyifsnowiswhite.Indeed,itisamatterofscientificnecessitythatthestateofaffairsdescribedbytheformerisidenti-calwiththestateofaffairsdescribedbythelatter.Buttheformerexamplesimplydoesnotgivethespeaker’smeaning.Thespeakermightholdtruethesentence‘Schneeistweiss’undertheseandonlytheseconditionsandnotknowtheslightestthingaboutH2Omoleculesandwavelengthsoflight.TheT-sentencegivesthetruthconditions,butthespecificationofthetruthconditionsdoesnotnecessarilygivethemeaningofthesentence,becausethespecificationdoesnotyettellushowthespeakerrepresentsthosetruthconditions.Doesheorsherepresentthemundertheaspectofsnowbeingwhite,orwhatisthesamefactintheworld,doesheorsherepresentthemundertheaspectoffrozenH2Ocrystalsemittinglightofallwavelengths?Anytheorythatcannotgivethatinformationisnotatheoryofmeaning.Therearevariousattemptstomeetthesesortsofobjections,but,inmyview,theyarenotsuccessful.Intheend,alltruthdefinitionalaccountsofmeaning,likethebe-haviouristaccountswhichprecededthem,endupwithacertain‘indeterminacy’ofmeaning.Theycannotaccountinobjectivetermsforallofthesubjectivedetailsofmeaning,andbothDavidsonandQuinehaveacknowledgedthattheirviewsresultinindeterminacy.3.5Wittgenstein’slegacyWittgenstein’sworkcoverssuchavastrangeoftopics,fromaestheticstomath-ematics,andcoversthesetopicswithsomuchdepthandinsight,thatitcontinuestobeasourceofideasandinspirationforanalyticphilosophersandislikelytocontinuetobesoformanyyearstocome.Iwillmentiononlythreeareas.18\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATES3.5.1PhilosophicalpsychologyOneofWittgenstein’smainareasofresearchwasthatofpsychologicalconceptssuchasbelief,hope,fear,desire,want,expectandsensationconceptssuchaspainandseeing.Perhapshissinglemostcontroversialclaiminthisareaisthatconcerningaprivatelanguage.Heclaimsthatitwouldbelogicallyimpossiblefortheretobealan-guagethatwasprivateinthesensethatitswordscouldonlybeunderstoodbythespeakerbecausetheyreferredtothespeaker’sprivateinnersensationsandhadnoexternaldefinition.Suchalanguagewouldbeabsurd,hesaid,becausefortheapplica-tionofsuchwordstherewouldbenodistinctionbetweenwhatseemedrighttothespeakerandwhatreallywasright.Butunlesswecanmakeadistinctionbetweenwhatseemsrightandwhatreallyisright,wecannotspeakofrightorwrongatall,andhencewecannotspeakofusingalanguageatall.‘Aninnerprocess’,saysWittgenstein(1953),‘standsinneedofoutwardcriteria’.WittgensteinishereattackingtheentireCartesiantradition,accordingtowhichthereisarealmofinnerprivateobjects,ourinnermentalphenomena,andthemeaningsofthewordsthatstandfortheseentitiesareentirelydefinedbyprivateostensivedefinitions.NoothersingleclaimofWittgenstein’shasarousedasmuchcontroversyasthe‘privatelanguageargument’.Itcontinuestobeasourceoffascinationtocontemporaryphilosophers,andmanyvolumeshavebeenwrittenaboutWittgenstein’sanalysisofpsychologicalconcepts.3.5.2FollowingaruleWittgensteinispartofalongtraditionthatemphasizesthedistinctionbetweenthemodesofexplanationofthenaturalsciencesandthemodesofexplanationofhumanbehaviourandhumanculturalandpsychologicalphenomenagenerally.Hisanalysisofthisproblemchieflydealswiththephenomenonofhumanbehaviourwhichisinflu-encedordeterminedbymentalcontents,and,mostimportantly,bythephenomenaofhumanbeingsfollowingarule.Whatisitforahumanbeingtofollowarule?Wittgenstein’sanalysisofthisstressesthedifferencebetweenthewaythatrulesguidehumanbehaviourandthewaythatnaturalphenomenaareresultsofcauses.Wittgensteinthroughoutemphasizesthedifferencebetweencausesandreasons,andhealsoemphasizestherolesofinterpretationandrulefollowing.OnthemostextremeinterpretationofWittgenstein’sremarksaboutfollowingarule,heistheproponentofacertaintypeofscepticism.AccordingtooneviewofWittgenstein,heisarguingthatrulesdonotdeterminetheirownapplication,thatanythingcanbeinterpretedtoaccordwitharule,andconsequentlythatanythingcanbeinterpretedtoconflictwitharule.Iftakentoitsextreme,thisargumentwouldhavetheconsequencethat,logicallyspeaking,rulesdonotconstrainhumanbehaviouratall.Andifthatisright,thenmentalcontents,suchasknowledgeofmeaningsofwordsorprinciplesofactionorevenbeliefsanddesires,donotconstrainhumanbehaviour,becausetheyareeverywheresubjecttoanindefiniterangeofdifferentinterpretations.Wittgenstein’ssolutiontothisscepticismistoproposethatinterpretationcomestoanendwhenwesimplyaccepttheculturalpracticesofthecommunityinwhichweareimbedded.Interpretationcomestoanend,andwejustactonarule.Actingonaruleisapractice,anditisonethatwearebroughtuptoperforminourculture.ThescepticalimplicationsofWittgenstein’saccountofrulefollowingareresolvedbyanappealtoanaturalisticsolution:wearesimplythesortofbeingswhofollowculturallyandbiologicallyconditionedpractices.19\nJOHNR.SEARLEThisinterpretationofWittgensteinislargelyduetoSaulKripke(1982)andithasarousedconsiderablecontroversy.MyownviewisthatKripkehasmisinterpretedWittgensteinincertaincrucialrespects,butwhetherornothisinterpretationiscorrect,ithasbeenasourceofcontinuingdiscussionincontemporaryphilosophy.3.5.3PhilosophicalscepticismImportantworkonphilosophicalscepticismhasbeencontinuedbyphilosopherswhoareinspiredorprovokedbyWittgenstein,notablyThomsonClarkeandBarryStroud.Thesephilosopherspointoutthatareallyseriousanalysisofouruseofepistemicdis-courseshowsthattheproblemofscepticismcannotbesimplyovercomebytheusualanalyticphilosopher’smethodsofpointingoutthatthescepticraisesthedemandforjustificationbeyondthatwhichislogicallyappropriate.ClarkeandStroudclaimthattheproblemofscepticismgoesdeeperthanthissolutionwillallow.FollowingWittgensteinininvestigatingthedepthgrammaroflanguage,theyfindthatanysolutiontothesceptic’spredicament–thatis,anyjustificationforourclaimstohaveknowledgeabouttheworld–restsonamuchdeeperunderstandingofthedifferencebetweenordinaryorplaindiscourseandphilosophicaldiscourse.Workinthislineofresearchiscontinuingatpresent.4OverallAssessmentIhavenotattemptedtosurveyallofthemainareasofactivityincontemporaryana-lyticphilosophy.Mostimportantly,Ihaveleftoutcontemporaryworkinethics.Perhapsofcomparableimportance,Ihavehadnothingtosayaboutpurelytechnicalworkinlogic.Thereis,furthermore,athrivingbranchofanalyticphilosophycalled‘actiontheory’,whichshouldbementionedatleastinpassing.Thegeneralaimofanalyticactiontheoryistoanalysethestructureofhumanactionsintermsofthecausalrela-tionsbetweensuchmentalstatesasbeliefs,desiresandintentions,andthebodilymove-mentswhichareinsomesenseconstitutiveoftheactions.Finally,itisworthcallingattentiontothefactthatamonganalyticphilosopherstherehasbeenagreatrevivalofinterestinthehistoryofphilosophy.Traditionalanalyticphilosophersthoughtofthehistoryofphilosophyasmostlythehistoryofmistakes.Someofthehistoryofthesubjectcouldbeusefulfordoingrealphilosophy;buttheoverallconceptionwasthatthehistoryofphilosophyhadnomorespecialrelevancetophilosophythanthehistoryofmathematicstomathematics,orthehistoryofchemistrytochemistry.Thisattitudehaschangedrecently,andthereisnowafeelingofthehistoricalcontinuityofanalyticphilosophywithtraditionalphilosophyinawaythatcontrastssharplywiththeorigi-nalviewofanalyticphilosophers,whothoughtthattheymarkedaradical,orindeed,revolutionarybreakwiththephilosophicaltradition.Itistooearlytoprovideanassessmentofthecontributionthatwillbemadebyworkdoneinphilosophyatthepresenttime,oreveninthepastfewdecades.Myownviewisthatthephilosophyofmindandsocialphilosophywillbecomeevermorecentraltotheentirephilosophicalenterprise.Theideathatthestudyoflanguagecouldreplacethestudyofmindisitselfbeingtransformedintotheideathatthestudyoflanguageisreallyabranchofthephilosophyofmind.Withinthephilosophyofmind,perhaps20\nCONTEMPORARYPHILOSOPHYINTHEUNITEDSTATESthekeynotionrequiringanalysisisthatofintentionality–thatpropertyofthemindbywhichitisdirectedatoraboutorofobjectsandstatesofaffairsintheworldinde-pendentofitself.Mostoftheworkdonebyanalyticphilosophersinthephilosophyofmindhastendedtoclusteraroundthetraditionalmind–bodyproblem.Myownviewisthatweneedtooverthrowthisproblem:initstraditionalversion,itwasbasedontheassumptionthatmentalpropertiesandphysicalpropertiesweresomehowdifferentfromeachother,andthattherefore,therewassomespecialproblemnotlikeotherprob-lemsinbiologyastohowtheycouldbothbecharacteristicsofthehumanperson.Onceweseethatso-calledmentalpropertiesreallyarejusthigher-levelphysicalpropertiesofcertainbiologicalsystems,Ibelievethisproblemcanbedissolved.Onceitisdissolved,however,wearestillleftwiththetaskofanalysingwhatisthecentralprobleminthephilosophyoflanguageandincognitivescience,aswellasthephilosophyofmind,namely,thewaythathumanrepresentationalcapacitiesrelatethehumanorganismtotheworld.Whatarecalled‘language’,‘mind’,‘thinking’,‘speaking’and‘depicting’arejustdifferentaspectsofthismodeofrelatingtoreality.Ibelievethatthecausaltheoryofreferencewillbeseentobeafailureonceitisrecognizedthatallrepresentationsmustoccurundersomeaspectorother,andthattheextensionalityofcausalrelationsisinadequatetocapturetheaspectualcharacterofreference.Theonlykindofcausationthatcouldbeadequatetothetaskofreferenceisintentionalcausationormentalcausation,butthecausaltheoryofreferencecannotconcedethatultimatelyreferenceisachievedbysomementaldevice,sincethewholeapproachbehindthecausaltheorywastotrytoeliminatethetraditionalmentalismoftheoriesofreferenceandmeaninginfavourofobjectivecausalrelationsintheworld.Mypredictionisthatthecausaltheoryofreference,thoughitisatpresentbyfarthemostinfluentialtheoryofreference,willprovetobeafailureforthesereasons.Perhapsthesinglemostdisquietingfeatureofanalyticphilosophyinthefifty-yearperiodthatIhavebeendiscussingisthatithaspassedfrombeingarevolutionaryminoritypointofviewheldinthefaceoftraditionalistobjectionstobecomingitselftheconventional,establishmentpointofview.Analyticphilosophyhasbecomenotonlydominantbutintellectuallyrespectable,and,likeallsuccessfulrevolutionarymove-ments,ithaslostsomeofitsvitalityinvirtueofitsverysuccess.Givenitsconstantdemandforrationality,intelligence,clarity,rigourandself-criticism,itisunlikelythatitcansucceedindefinitely,simplybecausethesedemandsaretoogreatacostformanypeopletopay.Theurgetotreatphilosophyasadisciplinethatsatisfiesemotionalratherthanintellectualneedsisalwaysathreattotheinsistenceonrationalityandintelli-gence.However,inthehistoryofphilosophy,Idonotbelievewehaveseenanythingtoequalthehistoryofanalyticphilosophyforitsrigour,clarity,intelligenceand,aboveall,itsintellectualcontent.Thereisasenseinwhichitseemstomethatwehavebeenlivingthroughoneofthegreaterasinphilosophy.ReferencesAustin,J.L.1962:HowtodoThingswithWords.Oxford:ClarendonPress.Donnellan,K.1970:ProperNamesandIdentifyingDescriptions.Synthèse,21,335–58.Feyerabend,P.1975:AgainstMethod.London:HumanitiesPress.Grice,H.P.1957:Meaning.PhilosophicalReview,66.21\nJOHNR.SEARLE——1968:Utterer’sMeaning,Sentence-Meaning,andWord-Meaning.FoundationsofLanguage,4,1–18.Kripke,S.1972:NamingandNecessity.InG.HarmanandD.Davidson(eds)SemanticsofNaturalLanguage,Dordrecht:Reidel.——1982:WittgensteinonRulesandPrivateLanguage.Cambridge,MA:HarvardUniversityPress.Kuhn,T.1962:TheStructureofScientificRevolutions.Chicago:UniversityofChicagoPress.Putnam,H.1975:TheMeaningof‘Meaning’.InhisPhilosophicalPapers,Vol.2:Mind,LanguageandReality,Cambridge:CambridgeUniversityPress.Quine,W.V.O.1953:TwoDogmasofEmpiricism.InhisFromaLogicalPointofView,Cambridge,MA:HarvardUniversityPress.Rawls,J.1971:ATheoryofJustice.Cambridge,MA:HarvardUniversityPress.Searle,J.R.1964:HowtoDerive‘Ought’from‘Is’.PhilosophicalReview,73.——1969:SpeechActs:AnEssayinthePhilosophyofLanguage.Cambridge:CambridgeUniversityPress.——1979:ExpressionandMeaning.Cambridge:CambridgeUniversityPress.——1980a:Minds,BrainsandPrograms.BehavioralandBrainSciences,3,417–24.——1980b:IntrinsicIntentionality.BehavioralandBrainSciences,3,450–6.——1983:Intentionality:AnEssayinthePhilosophyofMind.Cambridge:CambridgeUniversityPress.Wittgenstein,L.1953:PhilosophicalInvestigations(translatedbyG.E.M.Anscombe).Oxford:Blackwell.22\nContemporaryPhilosophy:ASecondLookBERNARDWILLIAMS1TheIdentityofAnalyticalPhilosophyGiventhetitleofJohnSearle’sessay,thissecondintroductionmighthavebeenexpectedtocomplementthefirstgeographically,bydealingwithpresentphilosophicaldevelop-mentsinplacesotherthantheUnitedStates,butthisisnotinfactwhatitwilltrytodo.PhilosophyintheUnitedStates,inotherEnglish-speakingpartsoftheworld,andinmanyothercountriesaswell,isnowverylargelythesame.Intheseplaces,thereisonephilosophicalculture,andinasmuchasitcontainsdifferentapproaches,andsomeofthephilosophythatisdonewithinthatcultureisdistinctfrom‘analytical’philoso-phy,thatitselfisnotamatterofgeographicalregion.Itistruethat‘analytical’philosophy,thestyleofphilosophydescribedinSearle’sessayandoverwhelminglyrepresentedinthisvolume,isoftenprofessionallydistin-guished(injobadvertisements,forinstance)from‘continental’philosophy,andthisdoesrepresent,inaclumsyway,somethingwhichuntilrecentlywastrue:thatthewaysinwhichphilosophywasdoneinFrance,GermanyandothercountriesofcontinentalEuropeweretypicallydifferentfromthe‘analytical’style.Toamuchmorelimitedextent,thatremainsso.(Chapters40–2describethesituationincontinentalEurope.)However,itisabsurdtomarkphilosophicaldifferenceswiththesetwolabels.Apartfrominvolvingastrangecross-classification–ratherasthoughonedividedcarsintofront-wheeldriveandJapanese–thelabelsareseriouslymisleading,inhelpingonetoforgetthattheoriginsofanalyticalphilosophyitselflayincontinentalEurope(notablyso,whenitsfoundingfatheristakentobeFregeanditsgreatestrepresentativeWittgenstein),andthattheinterestsof‘continental’philosophyarenotconfinedtotheEuropeancontinent.Moreover,itisnotsimplyamatteroflabelling.Itisnotthatthedistinctioninitselfisunproblematical,andonlyneedsmoreaptlychosentitlestorepresentit.Thedistinctionsinvolvedareobscure,andthetitlesservetoconcealthisfact.Theterm‘continental’servestodiscouragethoughtaboutthepossiblecontraststoanalyticalphilosophy,andsoabouttheidentityofanalyticalphilosophyitself.Atthesametime,thevaguegeographicalresonanceofthetermdoescarryamessage,thatanalyticalphilosophyisfamiliarasopposedtoexotic,andperhaps–ifsomeolderstereotypesareinplay–thatitisresponsibleasopposedtofrivolous.Thisisindeedwhatmany\nBERNARDWILLIAMSanalyticalphilosophersbelieveaboutit,andtheybelieveitnotsomuchincontrasttoactivitiesgoingoninsomeremotecontinent,but,inmanycases,asopposedtoworkdoneintheirlocaldepartmentsofliterature.Itisnottruethatworkinotherstylesdoesnotexistintheheartlandsofanalyticalphilosophy;itmerelydoesnotexistindepart-mentsofphilosophy.Thedistinctionsinvolvedarenotgeographicalbutprofessional,andwhatisatissueistheidentityofphilosophyasadiscipline.Inparticular,whatisatissueistheidentityofphilosophyasasubjectthatcansustainongoing,cumulative,research.Ifitcandothis,itcanmakeaclaimwhichthehuman-itiesdonotalwaysfinditeasytomake,excepttotheextentthattheyarebranchesofhistory:thatthereissomethingtobefoundoutwithintheirdisciplines,thattheycanaddtoknowledge.Ithasbeenpartoftheattractionofanalyticalphilosophythat,withouttheproceduresoftheexperimentalortheoreticalsciencesandwithamorehumansubjectmatter,itcanclaimtoachieveresultswhichcommand,ifnotagree-ment,atleastobjectivediscussion,andwhichrepresentintellectualprogress.Ithasachievementsthatarenotarbitrarilypersonal,andtheycomparefavourablytothoseofthesocialsciences(atleastifoneleavesasidethequitepeculiarcaseofeconomics).Idonotthinkthattheseclaimsareempty.Ithinkthattheachievementsofana-lyticalphilosophyareremarkable,andIagreewithSearlethatthesubjectisinvariouswaysmoreinterestingthanitwasfortyyearsago.Itsvirtuesareindeedvirtues.Ithinkitishardtobeingoodfaithateacherofphilosophyunlessyoubelievethatthereissomethingworthdoingthatyou,invirtueofyourexperience,candoandwhichyoucanhelpotherpeopletodo.Ithinkthatthevirtuesofgoodphilosophyaretoacon-siderableextentworkmanlike.Quitecertainly,nophilosophywhichistobeworthwhileshouldlosethesensethatthereissomethingtobegotright,thatitisanswerabletoargumentandthatitisinthebusinessoftellingthetruth.Thesethings,Ibelieve,arerepresentedbythebestofwhatiscalledanalyticalphi-losophy,andtothatextentIamcommittedtoit.Myownworkhaslargelybeeninitsstyle.Yet,havingnowworkedinitforalongtime,andhaving,likeSearle,seenitchange,IamagreatdealmorepuzzledaboutitthanIoncewas;inparticular,Iampuzzledaboutthewaysinwhichitmustunderstanditselfifitistohavethosevirtues,andalsoaboutthecostsofsustainingthosevirtues.Thereisoneunderstandingofthesevirtueswhichiscertainlywidespreadamonganalyticalphilosophers,andwhichdirectlyservesthepromiseofongoingresearch:thattheseareindeedthevirtuesofascience.Somephilosopherswhoareimpressedbythisconceptionofwhattheyaredoingritualizeitintotheformsofpresentationfamiliarfromthesciences.Sometimesthisismerescientism,butinothercasesitsignalsthefactthattheirbranchofphi-losophyisnearneighbourtosomescience,suchasquantummechanicsorcognitivescience.Butthevirtuesofworkmanliketruthfulnesswhichanalyticalphilosophytypicallycultivatesaremuchmoreimportantthananyattempttomakephilosophylooklikeascience.Withmanyotherbranchesofphilosophythereisnoplausibleversionofsharingapartywallwithscience,andyetthesevirtuesarestillregardedasvirtues.Infact,eveninthecaseofthemorescientificareasofphilosophy,itisobviousenoughthatthesevirtuesarenotrecommendedonlybecausetheyarepossessedbyitsscien-tificneighbours:theyaretakentobeintellectualvirtues,goodinthesamewayforphilosophyandforscience.24\nCONTEMPORARYPHILOSOPHY:ASECONDLOOKButhowfarcanphilosophycultivatejustthesevirtuesandremaintruetootheraspectsofthelegacyofthesubject,toaimsthatithaspursuedinthepast?Thesci-encesaimtomakeclaimsthatcanandshouldbeconveyedinwaysthatareminimallyexpressive;theyarenotmeanttoconveyfeeling,ortodisplaymuchliteraryimagina-tion,ortospeak(atleastovertly)inapersuasivemode.Butifwethinkinparticularofmoralorpoliticalphilosophy,isthisambitionactuallytruetothetraditionsofthesubject,evenasthoseareembodiedinthehistoricalcanonofanalyticalphilosophyitself?IsittruetoatraditionthatcontainsPlato,Hobbes,Hume,Kantand(cometothat)JohnStuartMill?Isitinfacttruetoanygreatfigureofthattradition,exceptperhapsAristotle?Andifwearetotakehimasourmodel,weareleftwithmanyques-tionstoconsider–whether,forinstance,theaffectlesstreatisesthatwepossessdorep-resenthisvoice;ifso,whetherthetonedoesnotrepresentaquitespecialviewofethicallife;andwhetherweshouldnotweighrathersoberlythefactthattheclosestpreviousimitationofAristotlewastobefoundinamovementcalledscholasticism.Particularlyinmoralandpoliticalphilosophy,butnotonlythere,thereisaquestionofwhattheprocedurestypicalofanalyticalphilosophymean.Therearemanyvirtuousandvaluablethingsthattheymakepossible,andatthesametimethereareresourcesofphilosophyinthepastthattheyseeminglyexclude,anditisimportantnottoassumethatthisbalanceissimplygiventous,aboveallbyanunquestionableandtransparentinterpretationoftheidealsofintellectualresponsibility.Idonotwanttosuggestthattheadoptionoftheanalyticalstyleisamereabdication,acowardlyrefusaltoadoptamoreimaginativeandcommittedmannerwhich(criticssometimessuggest)isobviouslytohand.Stilllessisitsimplyamatterofscientisticcamouflage.Itisafeatureofourtimethattheresourcesofphilosophicalwritingtypicallyavailabletoanalyticalphilosophyshouldpresentthemselvessostronglyastheresponsiblewayofgoingon,themostconvincingexpressionofaphilosopher’sclaimonpeople’satten-tion.Butthatisanhistoricalfact,andweshouldtrytounderstanditassuch.IdonotthinkthatIadequatelyunderstandit,and,forthatreason,Iwouldnotliketopredictwhatotherpossibilitiestheremayprovetobeforaphilosophythatpreservesthemeritsofanalyticalphilosophy.Intherestofthisessay,Ishalltrytogiveanoutlineofsomeprincipalconcernsofanalyticalmoralandpoliticalphilosophy.ThiswillsupplementtheaccountthatJohnSearlehasgivenofthestateoftheartinotherareas,butIhopealsothatindescribingsomeofwhatanalyticalphilosophyhasrecentlydoneforthesesubjects,itmayen-couragereaderstoaskwhatnewthingsitmightbeabletodo.2Meta-ethicsPhilosophicalstudieshaveoftenbeenunderstood,intheanalyticaltraditionbutnotonlythere,asbeinghigher-order,inthesensethatnaturalscience,forinstance,willstudynaturalphenomena,whilethephilosophyofsciencewillstudy,fromsomepar-ticularpointsofview,theoperationsofscience.Someofmoralphilosophy(or,asIshallalsocallit,ethics)iscertainlyahigher-orderstudy.Itdiscussessuchthingsasthenatureofmoraljudgements,andaskswhethertheyexpressgenuinebeliefs,whethertheycanbeobjectivelytrue,andsoforth.Suchhigher-orderquestionsaretheconcern25\nBERNARDWILLIAMSofmeta-ethics.Atonetime(thirtytofiftyyearsago)itwaswidelythoughtinanalyti-calphilosophythatethicsconsistedonlyofmeta-ethics.Apowerfulsourceofthiscon-ceptionwasthebeliefinafirmdistinctionbetweenfactandvalue,towhichSearlehasreferred.However,theideaofethicsassimplymeta-ethicsdoesnotfollowdirectlyfromthedistinctionbetweenfactandvalue,andthosewhousedthedistinctiontosupportthatideaneededtwofurtherassumptions.Onewasthatphilosophyitselfshouldbeinthebusinessof‘fact’(which,forthepur-posesofthedistinction,includedtheory)andnotofvalue.Thiswasconnectedwithacertainconceptionofphilosophy,importanttotheidentityofanalyticalphilosophy,inwhichitistakentoderiveitsauthorityfromitstheoreticalstanceasanabstractintel-lectualenquiry.Someearlierphilosophers,suchasG.E.Moore,hadindeedbelievedinthedistinctionbetweenfactandvalue,buthadsupposedthatphilosophers,inonewayoranother,couldhavequitealottosaysubstantivelyaboutvalues.Thejourneyfromthefact–valuedistinctiontoaviewofethicsasonlymeta-ethicsinvolvedtheassump-tionthatthiswasimpossibleorinappropriate.Thesecondassumptioninvolvedinthejourneywasthatmeta-ethicsitselfcouldbevalue-neutral,thatthestudyofthenatureofethicalthoughtdidnotcommitonetoanysubstantivemoralconclusions.Ayetfurtherassumption,whichwasnotnecessarytothejourneybutdidoftenaccompanyit,wasthatmeta-ethicsshouldbelinguisticinstyle,anditssubjectshouldbe‘thelanguageofmorals’.Thislatterideahasnowalmostentirelydisappeared,asthepurelylinguisticconceptionofphilosophicalstudyhasmoregenerallyretreated.Beyondthat,however,therearenowmoredoubtsabouttheextenttowhichmeta-ethicscanbevalue-neutral,and,inaddition,philosopherssimplyfeelfreerinmakingtheirownethicalcommitmentsclear.Meta-ethicsremainsapartofethics,butmostwritingsinphilosophicalethicsnowwilldeclaresubstantivemoralpositions,eitherincloseassociationwithsomemeta-ethicaloutlookorinamorefree-standingmanner.Recentmeta-ethicaldiscussionshavecarriedonthetraditionalinterestintheobjec-tivityofethics.Inthisconnection,‘moraljudgements’areoftengroupedtogetherandcomparedwitheverydaystatementsoffactorwithscientificstatements.Sometheoriesclaimthatmoraljudgementslacksomedesirablepropertythatfactualstatementscanattain,suchasobjectivityortruth,whileothertheoriesclaimthattheycanhavethisproperty.Thesedebates,particularlythoseconductedundertheinfluenceofpositivism,havetendedtoassimilatetwodifferentissues.Oneconcernstheprospectsofrationalagreementinethicalmatters.Theotherconcernsthesemanticstatusofmoraljudge-ments:whethertheyare,typically,statementsatall(asopposed,forinstance,toprescriptions),andwhethertheyaimattruth.Objectivityisbestunderstoodintermsofthefirstkindofquestion.Thereclearlyaresubstantiveandsystematicdisagreementsaboutethicalquestions,bothbetweendif-ferentsocietiesorcultures,andwithinonesociety(particularlywhen,asnow,thecultureofonesocietymaybehighlypluralist).Someofthesedisagreementsmayturnouttobeduetomisunderstandingorbadinterpretationanddissolvewhenlocalprac-ticesarebetterunderstood,butthisisnottrueofallofthem.Sinceancienttimesithasbeensuggestedthatthesedisagreementshaveastatusdifferentfromdisagreementsaboutfactsorabouttheexplanationofnaturalphenomena.Withthelatter,ifthepartiesunderstandthequestionatissue,theyseehowafterfurtherenquirytheymay26\nCONTEMPORARYPHILOSOPHY:ASECONDLOOKendupinoneofseveralpositions:theymaycometorationalagreementononeansweroranother,theymayrecognizethatsuchevidenceastheycanobtainunderdeter-minestheanswerandleavesthemwithintelligibleroomforcontinueddisagreement,ortheymayadvanceinunderstandingtoapointatwhichtheyseethatthequestioninitsoriginalformcannotbeanswered,forinstancebecauseitwasbasedonafalsepresupposition.Bycontrast,itissuggestedthatwecanunderstandanethicaldisputeperfectlywell,andyetitbeclearthatitneednotcomeoutinanyoftheseways.Disagreeingaboutanethicalmatter,thepartiesmayradicallydisagreeaboutthekindsofconsiderationsthatwouldsettlethequestion,andthesuggestionisthatattheendofthelinetheremaybenorationalwayofarrivingatagreement.Thisisthesuggestionthatethicalclaimslackobjectivity.Sometheorieshaveassociatedthispositionwithaviewaboutthesemanticstatusofmoralutterances.Emotivism,atheorycloselyassociatedwithpositivism,heldthatmoralutterancesweremerelyexpressionsofemotion,notfarremovedfromexpletives,andittookthis,reasonablyenough,nottobeanobjectivisttheory.Inthiscase,thesemanticaccountandthedenialofobjectivitywentcloselytogether.However,itisamistaketothinkthatthetwoissuesareingeneralsimplyrelatedtooneanother.AclearillustrationofthisisKant’stheory.Kantsupposedthatmoralstatements,oratanyratethemostbasicofthem,wereactuallyprescriptions,andheunderstoodthefundamentalprincipleofmoralitytobeanimperative.However,whentheissueisexpressedintermsofrationalagreementordisagreement,Kantisquitecertainlyanobjectivist:theCategoricalImperative,togetherinsomecaseswithempiricalinforma-tion,determinesforanyrationalagentwhatmoralityrequires,andallrationalagentsareinapositiontoagreeonit.Anotherexampleofobjectivitywhichisatleastnon-committalaboutthesemanticsinvolvedcomesfromvirtuetheory.Aristotlebelievedthatexperiencedanddiscriminatingagentswhohadbeenproperlybroughtupwouldreachrationalagreementinaction,feeling,judgementandinterpretation.Hebelieved,moreover,thatthispossibilitywasgroundedinthebestdevelopmentandexpressionofhumannature,andthatviewsaboutwhatcountedasthebesthumandevelopmentcouldthemselvescommandrationalagreement.Thiscertainlyoffersakindofobjec-tivity,butitdoesnotparticularlyemphasizeagreementinbelief;nodoubtsomeagree-mentinbeliefwillmatter,butsoequallywillagreementinfeelingandinpracticaldecision.However,evenifobjectivityneednotimplyrationalagreementinbelief,itmaybearguedthattheconverseholds:thatatheorywhichrepresentsmoraljudgementsasbasicallyexpressingbeliefsmustbecommittedtoobjectivity.Beliefs,thisargumentgoes,aretrueorfalse.Ifmoraljudgementsexpressbeliefs,thensomeofthemaretrueandthereissuchathingastruthinmorality.Soifpeopledisagreeaboutwhattobelieve,someonemustbewrong.Thiscertainlysoundsasthoughtheremustbeobjec-tivity.Thedifficultywiththisargumentisthatitseemstobetooeasytoagreethatmoraljudgementsadmitoftruthorfalsehood.Theyarecertainlycalled‘true’and‘false’,aseventheemotivistshadtoconcede,andtheclaimthatneverthelesstheyarenotreallytrueorfalseneedssomedeciphering.Emotivismitselfofferedasemanticanalysisintermsofwhichsuchjudgementsturnoutnotreallytobestatements,whichcertainlygivessomecontenttotheclaimthattheyarenotreallytrueorfalse.However,such27\nBERNARDWILLIAMSanalysesrunintodifficultypreciselybecausetheairofbeingastatementthatsur-roundsmoraljudgementsisnotmerelysuperficial–theybehavesyntacticallyjustasotherkindsofstatementsdo.Analternativeistoarguethatmoraljudgementscanindeedbetrueorfalse,butthatnothinginterestingfollowsfromthis.Onsometheoriesoftruth,sometimescalled‘redundancy’theories,toclaimthat‘P’istrueistodonomorethantoassertthatP.Anytheoryoftruthmustaccepttheequivalence‘‘‘P’’istrueifandonlyifP’;thepecu-liarityofredundancytheoriesistoclaimthatthisisallthereistothenatureoftruth.Ifthisiscorrect,thenthetruthorfalsehoodofmoraljudgementswillfollowsimplyfromtheirtakingastatementalform,whichallowsthemtobeassertedordenied.Objec-tivitywilltheneitherbeunderstoodassomethingthatnecessarilygoesalongwithtruthandfalsehood,inwhichcase,ontheredundancytheory,itwillbenomoreinterestingorsubstantivethantruth;oritwillbemoreinterestingandsubstantive–implyingforinstancethepossibilityofrationalagreement–inwhichcase,ontheredundancyview,itwillnotfollowjustfromthefactthatmoraljudgementscanbetrueorfalse.Itiswidely,thoughnotuniversally,agreedthatanadequatetheoryoftruthneedstogobeyondtheredundancyview,butitisdisputedhowfaritneedstogo.Somearguethatifonetakesseriouslytheclaimthatagivenpropositionistrue,thenthisdoesimplytheideathattherecouldbeconvergenceinbeliefonthepropositionunderfavourablecircumstances.Thisapproachbringstheideaoftruthitselfnearertothatofobjectiv-ityasthathasbeenintroducedhere.Othersholdthataproperly‘minimalist’theoryoftruthneednotbringinsuchastrongcondition.Ifobjectivismandthemeretruthofmoralstatementshaveoftenbeenassimilatedtooneanother,realism,equally,isoftenassimilatedtooneorbothofthem.Yetweshouldexpectrealism,ifitisanissueatall,tobeafurtherissue.Elsewhereinphi-losophy,forinstanceinthephilosophyofmathematicsorthephilosophyofscience,itcanbeagreedthatstatementsofacertainkind(aboutnumbers,oraboutsubatomicparticles)arecapableoftruth,andalsothattheycancommandrationalagreement,andyetitisthoughtbymanyphilosophersthatthisdoesnotanswerthequestionwhetherthosestatementsshouldbeinterpretedrealistically,wherethismeans(veryroughlyindeed)thatthestatementsaretakentorefertoobjectsthatexistindependentlyofourthoughtsaboutthem.Evenifitisnoteasytogiveadeterminatesensetosuchquestions,atanyrateonewouldnotexpectrealismtofollowtriviallyfromtheclaimthatmoralstatementscanbetrueorfalse.Somephilosophers,influencedbythelateJohnMackieandinthelineofHume,denyrealismbyclaimingthatthemoralpropertiesofpeople,actionsandsoforth,arenot‘intheworld’butare‘projected’ontoitfromourfeelingsandreactions.Accordingtothemostfamiliarversionofthisview,secondaryqualitiessuchascoloursarealsopro-jectedontotheworld,andthisraisesthequestionwhetherthemetaphorhasnotmis-locatedthemostsignificantissuesaboutmoralproperties.Thetheoryimpliesthatethicaloutlooksare‘perspectival’orrelatedtohumanexperienceinwaysinwhichphysicaltheory(atleast)isnot,butthisdoesnottakeusveryfar:itwillnottellusany-thingverydistinctiveaboutethicalrealismtoknowonlythatethicalconceptsareper-spectivalinasenseinwhichcolourconcepts,orperhapspsychologicalconcepts,arealsoperspectival.Ananti-realismthatgivesmoralpropertiesmuchthesamestatusas28\nCONTEMPORARYPHILOSOPHY:ASECONDLOOKcolourswillprobablysatisfymanymoralrealists.Weneedtoaskhowfarthemoralconceptsandoutlooksofvarioushumangroupscanintelligiblydifferwhiletherestoftheirwaysofdescribingtheworld,inparticulartheirpsychologicalconceptsfordescribingpeople’sbehaviour,remainthesame.Again,howfarcantheirpsychologi-calconceptsthemselvesintelligiblyvary,andhowshouldweunderstandthosevariations?Inconsideringsuchquestions,itishelpfultoabandonaverylimitingassumptionwhichhasbeenmadeuptothispointinthediscussion,namely,thatall‘moraljudge-ments’areessentiallyofthesamekindandstandinthesamerelationtosuchmattersastruthandobjectivity.Inconsideringmoraldisagreement,philosophershavecon-centratedoncasesinwhichthepartiesexpressthemselvesintermsofthe‘thin’ethicalconceptssuchas‘good’,‘right’and‘ought’.Thepartiessharethesamemoralandotherconcepts,anddisagreeaboutwhetheragivenjudgementshouldbeassertedordenied:theydisagree,forinstance,aboutwhethercapitalpunishmentiswrong.Torepresentdisagreementinthiswaymayseemtoisolateinahelpfulwayitsmoralfocus.Butalotofmoraldiscussion–todifferingdegreesindifferentsocieties–isconductedintermsof‘thick’concepts,suchas‘brutality’or‘betrayal’or‘sentimentality’,anditisamistaketosupposethatsuchconceptsaremerelyconvenientdevicesforassociatingabunchofempiricalconsiderationswithathinethicalconcept.Ithasbeenincreasinglyacceptedinrecentdiscussionsthattheapplicationofsuchconceptsisguidedbytheirevaluativepoint,andthatonecannotunderstandthemwithoutgraspingthatpoint.(Thisdoesnotmeanthatanyonewhounderstandssuchaconceptmusthaveadopteditashisorherown,butitdoesmeanthatheorsheneedstohaveimaginativelyiden-tified,asanethnographerdoes,withthosewhouseit.)Atthesametime,however,suchconceptsapplytosomeempiricalstatesofaffairsandnottoothers,andthereisroomfortruth,objectivityandknowledgetobedisplayedintheirapplication.Ifthisiscorrect,thenitmaybemorehelpfultoconsiderethicaldisagreement,notattheultimatecuttingedgeofthepracticaljudgementsaboutwhatoughtoroughtnottohappen,butfurtherback,inthenetworkofmoresubstantiveandthickerconceptsthatbackupsuchjudgements.Suchconceptswilltypicallyservemorepurposesthanexpressingbare‘moraljudgements’.Theymayplayarole,forinstance,inaschemeofpsychologicalexplanation.Thequestionwillthenbecome,rather,whyandtowhatextentdifferentculturesdifferintheirethicalconcepts,and,morebroadly,intheframe-worksofunderstandingthatgowithsuchconcepts.Seeninthislight,meta-ethicalquestionsmovefurtherawayfrombeingquestionsinthephilosophyoflanguageorthetheoryofjustificationorepistemology,andbecomemorelikequestionsinthetheoryofculturalunderstanding.Indeedtheymaybecomedirectlyquestionsofculturalunderstanding.Themostbasicquestionaboutobjectivitymayturnouttobetheques-tionoftheextenttowhichdifferenthumansocietiesshareanunderlyingdeterminateframeworkofethicalconcepts.Byturninginsuchadirection,philosophicaldiscussionbecomesmoreempiricalandhistorical,morerichlyrelatedtootherdisciplines,andmoreilluminating.Atthesametime,itmeansthatphilosophershavetoknowaboutmorethings,orpeopleinotherdisciplineshavetotakeonissuesinphilosophy.Tothatextent,philosophytendstoloseadistinctivesubjectmatteranditsidentitybecomesblurred.29\nBERNARDWILLIAMS3EthicalTheoryIhavealreadysaidthatanalyticalphilosophersarehappierthantheyonceweretorecognizethatwhattheysayinmoralphilosophyislikelytohaveasubstantiveethicalcontent,andthatevenmeta-ethicsislikelytohavesomesuchconsequences.Thereisaproblem,however,abouthowthisisrelatedtotheauthorityofphilosophy.Ifphiloso-phersaregoingtooffermoralopinions–withintheirsubject,thatistosay,andnotsimplyasanyoneoffersmoralopinions–theyneedtohavesomeprofessionalclaimtoattention.Theyarenot,asphilosophers,necessarilygiftedwithunusualinsightorimagination,andtheymaynothaveasignificantlywideexperienceorknowledgeoftheworld.Theirclaimtoattentionrestsontheircapacityfordrawingdistinctions,seeingtheconsequencesofassumptions,andsoforth:ingeneral,ontheirabilitytodevelopandcontrolatheoreticalstructure.Iftheauthorityofphilosophyliesinitsstatusasatheoreticalsubject,thephilosopher’sspecialcontributiontosubstantiveethicalissuesislikelytobefoundinatheoreticalapproachtothem.Oneofthemostcommonenterprisesinmoralphilosophyatpresentisthedevelopmentofvariousethicaltheories.Theaimofethicaltheoryistocastthecontentofanethicaloutlookintoatheoreti-calform.Anethicaltheorymustcontainsomemeta-ethics,sinceittakesoneviewratherthananotherofwhatthestructureandthecentralconceptsofethicalthoughtmustbe,thoughitneednothaveanopiniononeverymeta-ethicalissue.Itiscommit-tedtoputtingforwardinatheoreticalformasubstantiveethicaloutlook.Indoingthis,ethicaltheoriesaretodifferentdegreesrevisionary.Somestartwithasupposedlyunde-niablebasisforethics,andrejecteverydaymoralconclusionsthatconflictwithit.(Itisagoodquestionwhythebasisshouldberegardedasundeniableifithassuchcon-sequences.)Others,lessdogmatically,considerthemoralconclusionsthatwouldbedeliveredbyconflictingoutlooks,anddecidewhichoutlookmakesthemostcoherentsystematicsenseofthoseconclusionsthatwe(thatistosay,theauthorandthosereaderswhoagreewithhimorher)findmostconvincing.Unsystematizedbutcarefullyconsideredjudgementsaboutwhatwewouldthinkitrighttodoinacertainsituation,orwouldbepreparedtosayinapprovalorcriticismofpeopleandtheiractions,areoftencalledinthecontextofsuchamethod‘moralintuitions’.(Theterm‘intuition’hasapurelymethodologicalforce:itmeansonlythatthesejudgementsseemtous,afterconsideration,pre-theoreticallyconvincing,notthattheyarederivedfromafacultyofintuitingmoraltruths.)ApreferredmethodistoseekwhatJohnRawlshascalleda‘reflectiveequilibrium’betweentheoryandintuitions,modifyingthetheorytoaccom-modaterobustintuitions,anddiscardingsomeintuitionswhichclashwiththetheory,particularlyifonecanseehowtheymightbetheproductofprejudiceorconfusion.Moraltheoriesarestandardlypresentedasfallingintothreebasictypes,centringrespectivelyonconsequences,rightsandvirtues.Thefirstareunsurprisinglycalled‘consequentialist’,andthelast‘virtuetheories’.Thesecondareoftencalled‘deonto-logical’,whichmeansthattheyarecentredondutyorobligation,butthisisacross-classification,sinceconsequentialisttheoriesalsogiveprominencetoanobligation,thatofbringingaboutthebestconsequences.(Inthecaseofthemostfamiliarandpopularconsequentialisttheory,utilitarianism,thevalueoftheconsequencesis30\nCONTEMPORARYPHILOSOPHY:ASECONDLOOKexpressedintermsofwelfareorpreference-satisfaction.)Intermsofobligations,thedifferenceisratherthatpureconsequentialisttheoriespresentonlyonebasicobliga-tion,whilethesecondtypeoftheoryhasmany.Amoredistinctivemarkofdifferenceistobefoundintheideaofaright:thesecondtypeoftheorygroundsmanyofanagent’sobligationsintherightsofotherstoexpectcertainbehaviourfromthatagent,akindofconsiderationthatutilitariansandotherconsequentialistsregardasbeingatbestderivative,andatworsttotallyspurious.Anotherwayofunderstandingthedivisionintothreeisintermsofwhateachtheoryseesasmostbasicallybearingethicalvalue.Forthefirsttypeoftheory,itisgoodstatesofaffairs,andrightactionisunderstoodasactiontendingtobringaboutgoodstatesofaffairs.Forthesecondtype,itisrightaction;sometimeswhatmakesanactionrightisafactaboutitsconsequences,butoftenitisnot–itsrightnessisdeterminedratherbyrespectforothers’rights,orbyotherobligationsthattheagentmayhave.Virtuetheory,finally,putsmostemphasisontheideaofagoodperson,someonewhocouldbedescribedalsoasanethicallyadmirableperson.Thisisanimportantempha-sis,andthenotionofavirtueisimportantinethics.However,oncethetypesoftheoryaredistinguishedinthisway,itishardtoseethemasallinthesamelineofbusiness.Consequentialistandrightstheoriesaimtosystematizeourprinciplesorrulesofactioninwaysthatwill,supposedly,helpustoseewhattodoorrecommendinparticularcases.Atheoryofthevirtuescanhardlydothat:thetheoryitself,afterall,isgoingtosaythatwhatyoubasicallyneedinordertodoandrecommendtherightthingsarevirtues,notatheoryaboutvirtues.Moreover,virtuouspeopledonotthinkalways,orusually,aboutthevirtues.Theythinkaboutsuchthingsasgoodconsequencesorpeople’srights,andthismakesitclearthat‘virtuetheory’cannotbeonthesamelevelastheothertwotypesoftheory.4Morality,PoliticsandAnalyticalPhilosophyAmongmoralconcepts,thatofrightsisclosesttolawandalsotopolitics,andphi-losophicaldiscussionsofthemoftencrossthoseboundaries.Giventheserelations,itisnotsurprisingthatthekindoftheorymostoftenconstructedtoarticulatetheideaofmoralrightsiscontractualist,invokingtheideaofanagreementthatmightberatio-nallyarrivedatbypartiesinsomehypotheticalsituationinwhichtheywererequiredtomakerulesbywhichtheycouldlivetogether.Theinspirationofcontractualisttheoriesgoesback,inparticular,toKant.Kant’sownconstructionreliesonsomeideasthatarenotsharedbymanymoderntheorists,inparticularthatacommitmenttothebasicprincipleofmorality(theso-called‘CategoricalImperative’)ispresupposedbytheveryactivityofarationalagent.Italsoinvolvesaveryobscuredoctrineoffreedom.ThemoderntheoriesinspiredbyKantianideasarelesscommittedthanKantwastoshowingthatmoralityisultimaterationality,andtheyallowalsomoreempiricalmaterialintotheconstructionthanKantdid.TheleadingexampleofsuchatheoryisthatofJohnRawls.Hismodelofasetofcontractingpartiesreachinganagreementbehind‘aveilofignorance’hashadanimmenseinfluenceonthinkingaboutmorality.Itwasdesignedforapurposeinpoliti-calphilosophy,ofconstructingatheoryofsocialjustice.InRawls’stheory,theveilof31\nBERNARDWILLIAMSignoranceisintroducedtodisguisefromeachcontractorhisownparticularadvantagesanddisadvantagesandhisowneventualpositioninthesocietythatisbeingdesigned.Thepoliticaltheorythatusesthisthoughtexperimentisliberal,givingahighprioritytolibertyandatthesametimeemphasizingredistributionofresourcesintheinterestsofthedisadvantaged.ItissignificantthatwhenRawlsfirstproducedhistheoryhesawitinuniversalistterms,asofferingaconstructionofsocialjusticewhichwouldapplytoanysocietythatmettheconditions(veryroughly)ofbeingabletothinkaboutsocialidealsandhavingtheresourcestoimplementthem.Now,however,hehasmovedinthedirectionofseeingthetheoryofjusticeasonethatexpressestheaspirationsofaparticularsocialformation,themodernpluraliststate.Muchrecentpoliticalphilosophyhascentredonthisliberalproject,ofdefiningtermsofjustcoexistenceforpeoplelivinginapluralistsociety.Oneinterpretationofthataimistolookfortermsofcoexistencethatwillnotpresupposeacommonconceptionofthegood.Onsuchanaccount,citizenscanunderstandthemselvesassharingasocialexis-tencealthoughtheyhaveasindividuals,orasmembersofcommunitieslessextensivethanthestate,varyingconceptionsofagoodlife.Rawlshasinterpretedhisownpurposeinthoseterms;itisexpressed,forinstance,inthefactthatthepartiesintheoriginalpositionweresupposedtomaketheirdecisionon(broadly)self-interestedgrounds,andnotinthelightofanyantecedentconceptionofwhatagoodsocietywouldbe.Thevaluesthattheyweretakentohavewereexpressedonlyinthelistof‘primarygoods’intermsofwhichtheymadetheirchoice,alistwhichmadeitclearalreadythattheysetahighvalueonliberty,forinstance,anddidnotassesseverythingintermsofutility.Infact,Rawls’saimofmakinghistheoryasindependentaspossiblefromsubstan-tialclaimsaboutthegooddoesputitundersomestrain.Atleastinthefirstversionofthetheory,thebasicconceptionofjusticeincludedalarge-scalecommitmenttoeco-nomicredistribution,andwhilethismadeitverywelcometomanyliberals(particu-larlyintheAmericansenseofthatword),itlaiditopentothecriticismfromothersmoreinclinedtolibertarianism,suchasRobertNozick,thatthetheoryincorporatednotonlyrightfultermsofcoexistencebutasubstantialanddistinctiveconceptionofsocialjusticeaseconomicequality.InthelaterversionsofRawls’stheory,thisconcep-tionislessprominent,butevenmoreweightthanbeforeislaidontheideathatcoex-istenceinaliberalpluralistsocietyisnot‘ameremodusvivendi’butaconditionthatcallsonimportantmoralpowersoftolerationandrespectforautonomy.Thisempha-sisdoesseemtoexpressadistinctiveconceptionofthegood,ofaKantiankind,toapointatwhichitlooksasthoughtheconditionofpluralismisnotsimplyacontingencyofmodernlife,butanimportantvehicleofethicalself-expression.Others,suchasRonaldDworkin,havepursuedliberaltheory,andinpluralistterms,whileacceptingacommitmenttoadistinctiveconceptionofthegood.Others,again,haveclaimedtorejectliberalpluralismaltogetherandhaveturnedinwhatissome-timescalleda‘communitarian’direction.Itishardtotellinsomecaseswherewritersofthistendencystandinrelationtothepoliticsofliberalism.Whileafew,notablyAlasdairMacIntyre,despairingofthewholeenterprise,tryonlytodiagnoseourconditionandstoresomeethicalgoodsforabettertime,othersseemtoshareintheliberalundertakingbutprefer,inoppositiontoRawls’sKantianism,amoreHegeliantypeofdiscoursetoexpressit.32\nCONTEMPORARYPHILOSOPHY:ASECONDLOOKItisaratheroddfeatureofcommunitariantheories,atleastiftheytakeatradi-tionalistturn,thattheyrecommendapoliticsthatdoesnotsitveryeasilywiththeexis-tenceofsuchtheoriesthemselves,exceptperhapsasakindofinterimmeasure.Theyseekapoliticsinwhichpeople’srelationsareformedbysharedunderstandingswhichtoaconsiderableextentmustbeunspokenandtakenforgranted,andtheexchangeofabstractpoliticaltheoriesplaysnoobviousroleinthisconceptionofsociallife.(Hegel,manyofwhoseconcernsarere-enactedinthesedebates,thoughthehadananswertothis,inhisconceptionofasocietythatcouldultimatelyreconcileabstractunder-standingandconcretepractice,butfewcurrentdisputantsarehappytopaythepriceofadmissiontotheHegeliansystem.)Theliberals,ontheotherhand,haveaconceptionofmodernpoliticallifewhichatleastcoherentlyembracestheexistenceoftheirowntheories,sincetheyunderstandthemodernstateasaformationinwhichauthorityispeculiarlyvestedindiscursiveargument,ratherthanintraditionalorcharismaticleadership.Itistruethatmanypoliticalphilosophersintheanalyticaltradition(unlikeRawlshimself,andalsounlikeHabermas,whocomesfromasociologicaltradition)donotseetheroleoftheirtheo-riesintheseterms,butratherasadvancingtrans-historicalviewsaboutthedemandsofajustpoliticalorder.Butevenifliberalsdonotalwaysrecognizethepointthem-selves,theroleoftheoryinliberalpoliticalphilosophycanbegivenaspecialjustifica-tionintermsofcurrentpoliticalreality.Liberalscoherentlybelievethattheprojectofpoliticaltheorymakessense,becausetheyarecommittedtothinkingthatinourcircumstancesitmakessensetoengageinapoliticalactivityofexplainingthebasicprinciplesofdemocraticgovernmentinsuchterms.Inthisrespect,moralphilosophyisinanaltogetherworsesituation.Ittypicallylacksanaccountofwhytheprojectofarticulatingmoraltheoriesmakesanysenseatall.Asmanywritershavepointedout,itbearslittlerelationtothepsychologyofpeople’sethicallives,andinasmuchasitclaimsthatturningmoralityintoatheorymakesitmorerational,thereisapressingquestionofwhatconceptofrationalityisbeinginvoked.Toalimitedextent,theremaybeananswertothatquestion,inasmuchassomeethicalquestions,suchasthoseraisedbymedicalethics,arepublicquestions,closelytiedtopoliticsandthelaw.Inthosecases,weneedapublicdiscoursetolegiti-matesomeanswers,sinceitisapublicissuewhatshouldbepermitted,andmereappealstoethicalorprofessionalauthoritywillnolongerdo.Butitwouldbeamistaketosupposethatinsuchcaseswearepresentedwithapureconceptofmoralrational-itywhichwethenapplytoourhistoricalcircumstances.Rather,whatcountsasaratio-nalwayofdiscussingsuchquestionsisinfluencedbythehistoricalcircumstances,andaboveallbytheneedtogiveadiscursivejustification,insomethinglikealegalstyle,forprocedureswhichincreasinglyareadoptedinapublicdomainandcanbechallengedinit.Moreover,manyimportantethicalissuesarenotofthiskindatall.Moralityhasalwaysbeenconnectednotonlywithlawandpolitics,butalsowiththemeaningofanindividual’slifeandhisorherrelationstootherpeople.Intheseconnections,theauthorityoftheoryoverthemoralliferemainsquiteopaque.Certainly,itwillnotdotorelyontheinference:philosophymusthavesomethingtosayaboutthemorallife;themostresponsibleformofphilosophyisanalyticalphilosophy;whatanalyticalphi-losophyisbestatistheory;sophilosophy’scontributiontothemorallifeistheory.33\nBERNARDWILLIAMSInrejectingthisuninvitingargument,somewillwanttoattackthesecondpremise.However,itmaybemoreconstructive,andoffermoreofachallengetothinkingaboutwhatonewantsofphilosophy,ifonereconsidersthethirdpremise,thatwhatanalyti-calphilosophydoesbestistheory.Analyticalphilosophy’sownvirtues,suchasitsunfanaticaltruthfulness,couldencourageitinethicstoremindusofdetailratherthanbludgeonuswiththeory.Truthfulnessinpersonallife,andeveninpolitics,isnotnecessarilyopposedtotheexerciseoftheimagination.Itisrelevantherethatanimaginativetruthfulnessisavirtueinthearts.Writersinmoralphilosophysometimesurgeustoextendourethicalunderstandingbyturningtoimaginativeliterature,inparticularthenovel.Totheextentthatthisisgoodadvice,itisnotbecausenovelsareconvenientsourcesofpsy-chologicalinformation,stilllessbecausesomeofthemaremorallyedifying.Itisbecauseimaginativewritingcanpowerfullyevokethestrengthofethicalconsidera-tionsingivingsensetosomeone’slifeortoapassageofit,and,equally,presentthepos-siblecomplexity,ambivalenceandultimateinsecurityofthoseconsiderations.Goodliteraturestandsagainsttheisolationofmoralconsiderationsfromthepsychologicalandsocialforcesthatbothempowerandthreatenthem.Butthisisolationofmoralconsiderationsfromtherestofexperienceisanillusionverymuchfosteredbymoralphilosophyitself;indeed,withoutthatillusionsomeformsofmoralphilosophycouldnotexistatall.Sotherearelessonsherenotjustforphilosophy’suseofotherwriting,butforphilosophicalwritingitself.Thetruthfulnessthatitproperlyseeksinvolvesimaginativehonestyandnotjustargumentativeaccuracy.Analyticalphilosophy,orsomerecognizabledescendantofit,shouldbeabletomakearichercontributiontoethicsthanhasoftenbeenthecaseuptonow.Ifitistodoso,itwillneedtoholdontotwotruthswhichittendstoforget(notonlyinethics,butmostdamaginglythere):thatphilosophycannotbetoopure,andmustmergewithotherkindsofunderstanding;andthatbeingsoberlytruthfuldoesnotexclude,butmayactuallydemand,theimagination.34\nPARTIAREASOFPHILOSOPHY\n1EpistemologyA.C.GRAYLINGFormostofthemodernperiodofphilosophy,fromDescartestothepresent,epistemol-ogyhasbeenthecentralphilosophicaldiscipline.Itraisesquestionsaboutthescopeandlimitsofknowledge,itssourcesandjustification,anditdealswithscepticalargumentsconcerningourclaimstoknowledgeandjustifiedbelief.Thischapterfirstlyconsidersdif-ficultiesfacingattemptstodefineknowledgeand,secondly,exploresinfluentialresponsestothechallengeofscepticism.EpistemologyiscloselyrelatedtoMETAPHYSICS(chapter2),whichisthephilosophicalaccountofwhatkindsofentitiesthereare.Epistemologi-calquestionsarealsocrucialtomostoftheotherareasofphilosophyexaminedinthisvolume,fromETHICS(chapter6)toPHILOSOPHYOFSCIENCE(chapter9)andPHILOSO-PHYOFMATHEMATICS(chapter11)toPHILOSOPHYOFHISTORY(chapter14).ChaptersonindividualsorgroupsofphilosophersfromDESCARTES(seechapter26)toKANT(chapter32)discussclassicalepistemology,whileseveralchaptersaboutmorerecentphilosophersalsofollowepistemologicalthemes.IntroductionEpistemology,whichisalsocalledthetheoryofknowledge,isthebranchofphilosophyconcernedwithenquiryintothenature,sourcesandvalidityofknowledge.Amongthechiefquestionsitattemptstoanswerare:Whatisknowledge?Howdowegetit?Canourmeansofgettingitbedefendedagainstscepticalchallenge?Thesequestionsareimplicitlyasoldasphilosophy,althoughtheirfirstexplicittreatmentistobefoundinPLATO(c.427–347BC)(seechapter23),inparticularinhisTheaetetus.Butitisprimar-ilyinthemodernera,fromtheseventeenthcenturyonwards–asaresultoftheworkofDESCARTES(1596–1650)(chapter26)andLOCKE(1632–1704)(chapter29)inasso-ciationwiththeriseofmodernscience–thatepistemologyhasoccupiedcentre-stageinphilosophy.Oneobvioussteptowardsansweringepistemology’sfirstquestionistoattemptadefinition.Thestandardpreliminarydefinitionhasitthatknowledgeisjustifiedtruebelief.Thisdefinitionlooksplausiblebecause,attheveryleast,itseemsthattoknowsomethingonemustbelieveit,thatthebeliefmustbetrue,andthatone’sreasonfor\nA.C.GRAYLINGbelievingitmustbesatisfactoryinthelightofsomecriterion–foronecouldnotbesaidtoknowsomethingifone’sreasonsforbelievingitwerearbitraryorhaphazard.Soeachofthethreepartsofthedefinitionappearstoexpressanecessaryconditionforknowl-edge,andtheclaimisthat,takentogether,theyaresufficient.Thereare,however,seriousdifficultieswiththisidea,particularlyaboutthenatureofthejustificationrequiredfortruebelieftoamounttoknowledge.Competingpro-posalshavebeenofferedtomeetthedifficulties,eitherbyaddingfurtherconditionsorbyfindingabetterstatementofthedefinitionasitstands.Thefirstpartofthefollow-ingdiscussionconsiderstheseproposals.Inparallelwiththedebateabouthowtodefineknowledgeisanotherabouthowknowledgeisacquired.Inthehistoryofepistemologytherehavebeentwochiefschoolsofthoughtaboutwhatconstitutesthechiefmeanstoknowledge.Oneisthe‘rational-ist’school(seechapters26and27),whichholdsthatreasonplaysthisrole.Theotheristhe‘empiricist’(seechapters29,30and31),whichholdsthatitisexperience,prin-cipallytheuseofthesensesaidedwhennecessarybyscientificinstruments,whichdoesso.Theparadigmofknowledgeforrationalistsismathematicsandlogic,whereneces-sarytruthsarearrivedatbyintuitionandrationalinference.Questionsaboutthenatureofreason,thejustificationofinference,andthenatureoftruth,especiallynec-essarytruth,accordinglypresstobeanswered.Theempiricists’paradigmisnaturalscience,whereobservationandexperimentarecrucialtoenquiry.Thehistoryofscienceinthemoderneralendssupporttoempiri-cism’scase;butpreciselyforthatreasonphilosophicalquestionsaboutperception,observation,evidenceandexperimenthaveacquiredgreatimportance.Butforbothtraditionsinepistemologythecentralconcerniswhetherwecantrusttheroutestoknowledgetheyrespectivelynominate.Scepticalargumentssuggestthatwecannotsimplyassumethemtobetrustworthy;indeed,theysuggestthatworkisrequiredtoshowthattheyare.Theefforttorespondtoscepticismthereforeprovidesasharpwayofunderstandingwhatiscrucialinepistemology.Section2belowisaccord-inglyconcernedwithananalysisofscepticismandsomeresponsestoit.Thereareotherdebatesinepistemologyabout,amongotherthings,memory,judge-ment,introspection,reasoning,the‘apriori–aposteriori’distinction,scientificmethodandthemethodologicaldifferences,ifany,betweenthenaturalandthesocialsciences;however,thequestionsconsideredherearebasictothemall.1Knowledge1.1DefiningknowledgeTherearedifferentwaysinwhichonemightbesaidtohaveknowledge.Onecanknowpeopleorplaces,inthesenseofbeingacquaintedwiththem.Thatiswhatismeantwhenonesays,‘MyfatherknewLloydGeorge’.Onecanknowhowtodosomething,inthesenseofhavinganabilityorskill.Thatiswhatismeantwhenonesays,‘Iknowhowtoplaychess’.Andonecanknowthatsomethingisthecase,aswhenonesays,‘IknowthatEverestisthehighestmountain’.Thislastissometimescalled‘propositionalknowledge’,anditisthekindepistemologistsmostwishtounderstand.38\nEPISTEMOLOGYThedefinitionofknowledgealreadymentioned–knowledgeasjustifiedtruebelief–isintendedtobeananalysisofknowledgeinthepropositionalsense.Thedefinitionisarrivedatbyaskingwhatconditionshavetobesatisfiedifwearecorrectlytodescribesomeoneasknowingsomething.Ingivingthedefinitionwestatewhatwehopearethenecessaryandsufficientconditionsforthetruthoftheclaim‘Sknowsthatp’,where‘S’istheepistemicsubject–theputativeknower–and‘p’aproposition.Thedefinitioncarriesanairofplausibility,atleastasappliedtoempiricalknowl-edge,becauseitseemstomeettheminimumwecanbeexpectedtoneedfromsocon-sequentialaconcept.ItseemsrighttoexpectthatifSknowsthatp,thenpmustatleastbetrue.ItseemsrighttoexpectthatSmustnotmerelywonderwhetherorhopethatpisthecase,butmusthaveapositiveepistemicattitudetoit:Smustbelievethatitistrue.AndifSbelievessometruepropositionwhilehavingnogrounds,orincorrectgrounds,ormerelyarbitraryorfancifulgrounds,fordoingso,wewouldnotsaythatSknowsp;whichmeansthatSmusthavegroundsforbelievingpwhichinsomesenseproperlyjustifydoingso.Oftheseproposedconditionsforknowledge,itisthethirdthatgivesmosttrouble.Thereasonissimplyillustratedbycounter-examples.ThesetaketheformofcasesinwhichSbelievesatruepropositionforwhatareinfactthewrongreasons,althoughtheyarefromhisorherownpointofviewpersuasive.Forinstance,supposeShastwofriends,TandU.Thelatteristravellingabroad,butShasnoideawhere.AsforT,SsawhimbuyingandthereafterdrivingaboutinaRollsRoyce,andthereforebelievesthatheownsone.Now,fromanypropositionponecanvalidlyinferthedisjunction‘porq’.SoShasgroundsforbelieving‘TownsaRollsorUisinParis’,eventhough,exhypothesi,hehasnoideaofU’slocation.ButsupposeTinfactdoesnotowntheRolls–heboughtitforsomeoneelse,onwhosebehalfhealsodrivesit.AndfurthersupposethatUisindeed,bychance,inParis.ThenSbelieves,withjustification,atrueproposition:butweshouldnotwanttocallhisbeliefknowledge.Exampleslikethisarestrained,buttheydotheirwork;theyshowthatmoreneedstobesaidaboutjustificationbeforewecanclaimtohaveanadequateaccountofknowledge.1.2JustificationApreliminaryquestionconcernswhetherhavingjustificationforbelievingsomepentailsp’struth,for,ifso,counter-examplesofthekindjustmentionedgetnopurchaseandweneednotseekwaysofblockingthem.Thereisindeedaview,called‘infallibil-ism’,whichoffersjustsucharesource.ItstatesthatifitistruethatSknowsp,thenScannotbemistakeninbelievingp,andthereforehisjustificationforbelievingpguar-anteesitstruth.Theclaimis,inshort,thatonecannotbejustifiedinbelievingafalseproposition.Thisviewisrejectedby‘fallibilists’,whoclaimthatonecanindeedhavejustificationforbelievingsomepalthoughitisfalse.Theircountertoinfallibilismturnsonidenti-fyingamistakeinitssupportingargument.Themistakeisthatwhereasthetruthof‘Sknowsthatp’indeedrulesoutthepossibilitythatSisinerror,thisisfarfromsayingthatSissoplacedthathecannotpossiblybewrongaboutp.Itisrighttosay:(1)‘itisimpossibleforStobewrongaboutpifheknowsp’,butitisnotinvariablyrighttosay39\nA.C.GRAYLING(2)‘ifSknowsp,thenitisimpossibleforhimtobewrongaboutp’.Themistaketurnsonthinkingthatthecorrectwidescopereading(1)of‘itisimpossible’licensesthenarrowscopereading(2)whichconstitutesinfallibilism.Aninfallibilistaccountmakesthedefinitionofknowledgelooksimple:Sknowspifhisbeliefinitisinfalliblyjustified.Butthisdefinitionrendersthenotionofknowledgetoorestrictive,foritsaysthatScanjustifiablybelieveponlywhenthepossibilityofp’sfalsityisexcluded.Yetitappearstobeacommonplaceofepistemicexperiencethatonecanhavetheverybestevidenceforbelievingsomethingandyetbewrong(astheaccountofscepticismgivenbelowisatpainstoshow),whichistosaythatfallibilismseemstheonlyaccountofjustificationadequatetothefactsofepistemiclife.Weneedthereforetoseewhetherfallibilisttheoriesofjustificationcangiveusanadequateaccountofknowledge.TheproblemforfallibilistaccountsispreciselytheoneillustratedbytheRollsRoyceexampleabove,andotherssimilartoit(so-called‘Gettierexamples’,introducedinGettier1963),namely,thatone’sjustificationforbelievingpdoesnotconnectwiththetruthofpintherightway,andperhapsnotatall.WhatisrequiredisanaccountthatwillsuitablyconnectS’sjustificationbothwithhisbeliefthatpandwithp’struth.Whatisneededisaclearpictureof‘justifiedbelief’.Ifonecanidentifywhatjusti-fiesabelief,onehasgoneallormostofthewaytosayingwhatjustificationis;andenrouteonewillhavedisplayedtherightconnectionbetweenjustification,ontheonehand,andbeliefandtruthontheother.Inthisconnectionthereareseveralstandardspeciesoftheory.FoundationalismOneclassoftheoriesofjustificationemploysthemetaphorofanedifice.Mostofourordi-narybeliefsrequiresupportfromothers;wejustifyagivenbeliefbyappealingtoanotherorothersonwhichitrests.Butifthechainofjustifyingbeliefsweretoregresswithoutterminatinginabeliefthatisinsomewayindependentlysecure,therebyprovidingafoundationfortheothers,wewouldseemtolackjustificationforanybeliefinthechain.Itappearsnecessarythereforethatthereshouldbebeliefswhichdonotneedjustifica-tion,orwhichareinsomewayself-justifying,toserveasanepistemicunderpinning.Onthisviewajustifiedbeliefisonewhicheitheris,orissupportedby,afounda-tionalbelief.Thenextstepsthereforearetomakeclearthenotionofa‘foundation’andtoexplainhowfoundationalbeliefs‘support’non-foundationalones.Somewayofunderstandingfoundationalismwithoutrelianceonconstructionalmetaphorsisneeded.Itisnotenoughbarelytostatethatafoundationalbeliefisabeliefthatrequiresnojustification,fortheremustbeareasonwhythisisthecase.Whatmakesabeliefinde-pendentorself-standingintherequiredway?Itisstandardlyclaimedthatsuchbeliefsjustifythemselves,orareself-evident,orareindefeasibleorincorrigible.Thesearenotthesamethings.Abeliefmightbeself-justifyingwithoutbeingself-evident(itmighttakehardworktoseethatitjustifiesitself).Indefeasibilitymeansthatnofurtherevidenceorother,competing,beliefs,canrenderagivenbeliefinsecure.Yetthisisapropertythatthebeliefmighthaveindependentlyofwhetherornotitisself-justifying.Andsoon.Butwhatthesecharacterizationsareintendedtoconveyistheideathatacertainimmunityfromdoubt,errororrevisionattachestothebeliefsinquestion.40\nEPISTEMOLOGYItmightevenbeunnecessaryormistakentothinkthatitisbeliefthatprovidesthefoundationsfortheedificeofknowledge:someotherstatemightdoso.Perceptualstateshavebeenofferedascandidates,becausetheyappeartobesuitablyincorrigible–ifoneseemstoseearedpatch,say,thenonecannotbewrongthatoneseemstoseearedpatch.Anditappearsplausibletosaythatone’sbeliefthatpneedsnofurtherjustificationorfoundationthanthatthingsappeartooneaspdescribesthemtobe.Thesesuggestionsbristlewithdifficulties.Examplesofself-evidentorself-justifyingbeliefstendtobedrawnfromlogicandmathematics–theyareofthe‘xisx’or‘oneplusoneequalstwo’variety,whichcriticsarequicktopointoutgivelittlehelpingroundingcontingentbeliefs.Perceptualstateslikewiseturnouttobeunlikelycandidatesforfoun-dations,onthegroundsthatperceptioninvolvestheapplicationofbeliefswhichthem-selvesstandinneedofjustification–amongthembeliefsaboutthenatureofthingsandthelawstheyobey.Whatismostrobustlycontestedisthe‘mythofthegiven’,theideathattherearefirm,primitiveandoriginaldatawhichexperiencesuppliestoourminds,antecedenttoanduntaintedbyjudgement,furnishingthewherewithaltosecuretherestofourbeliefs.Thereisadifficultyalsoabouthowjustificationistransmittedfromfoundationalbeliefstodependentbeliefs.Itistoostrongaclaimtosaythatthelatterarededuciblefromthem.Mostifnotallcontingentbeliefsarenotentailedbythebeliefsthatsupportthem;theevidenceIhavethatIamnowsittingatmydeskisaboutasstrongasempiricalevidencecanbe,yetgiventhestandardscepticalconsiderations(suchas,forexample,thepossibilitythatIamnowdreaming)itdoesnotentailthatIamsittinghere.Iftherelationisnotadeductiveone,whatisit?Othercandidaterelations–inductiveorcriterial–arebytheirnaturedefeasible,andtherefore,unlesssomehowsupplemented,insufficienttothetaskoftransmittingjustificationfromthefoundationstootherbeliefs.Thesupplementationwouldhavetoconsistofguaranteesthatthecircumstancesthatdefeatnon-deductivejustificationdonotinfactobtain.Butifsuchguarantees–under-stood,toavoidcircularity,asnotbeingpartoftheputativefoundationsthemselves–wereavailabletoprotectnon-deductivegrounds,thenappealtoanotionoffoundationslookssimplyotiose.1.3CoherenceDissatisfactionwithfoundationalismhasledsomeepistemologiststoprefersayingthatabeliefisjustifiedifitcohereswiththoseinanalreadyacceptedset.Theimmediatetaskistospecifywhatcoherenceis,andtofindawayofdealinginanon-circularwaywiththeproblemofhowthealreadyacceptedbeliefscametobeso.Hardontheheelsofthistaskcomesanumberofquestions.Iscoherenceanegativecriterion(thatis,abelieflacksjustificationifitfailstocoherewiththeset)oraposi-tiveone(thatis,abeliefisjustifiedwhenitcohereswiththeset)?Andisittobeunder-stoodstrongly(bywhichcoherenceissufficientforjustification)orweakly(bywhichcoherenceisoneamongotherjustifyingfeatures)?Theconceptofcoherencehasitstheoreticalbasisinthenotionofasystem,under-stoodasasetwhoseelementsstandinmutualrelationsofbothconsistencyand(somekindof)interdependence.Consistencyisofcourseaminimumrequirement,andgoeswithoutsaying.Dependenceismoredifficulttospecifysuitably.Itwouldbefartoo41\nA.C.GRAYLINGstrong–foritwouldgiverisetoassertiveredundancy–torequirethatdependencemeansmutualentailmentamongbeliefs(thisiswhatsomehaverequired,citinggeometryastheclosestexample).Amorediffusenotionhasitthatasetofbeliefsiscoherentifanyoneofthemfollowsfromalltherest,andifnosubsetofthemislogicallyindependentoftheremainder.Butthisisvague,andanywayseemstorequirethatthesetbeknowntobecompletebeforeonecanjudgewhetheragivenbeliefcohereswithit.Aremedymightbetosaythatabeliefcohereswithanantecedentsetifitcanbeinferredfromit,orfromsomesignificantsubsetwithinit,asbeingthebestexplanationinthecase.Tothissomeonemightobjectthatnotalljustificationstaketheformofexplanations.Analternativemightbetosaythatabeliefisjustifiedifitsurvivescom-parisonwithcompetitorsforacceptanceamongtheantecedentset.Buthereanobjec-tormightaskhowthiscanbesufficient,sincebyitselfthisdoesnotshowwhythebeliefmeritsacceptanceoverequallycoheringrivals.Indeed,anytheoryofjustificationhastoensureasmuchforcandidatebeliefs,sothereisnothingabouttheproposalthatdistinctivelysupportsthecoherencetheory.Andthesethoughtsleaveunexaminedthequestionofthe‘antecedentset’anditsjustification,whichcannotbeamatterofcoherence,forwithwhatisittocohereinitsturn?1.4InternalismandexternalismBoththefoundationalistandcoherencetheoriesaresometimesdescribedas‘internal-ist’becausetheydescribejustificationasconsistingininternalrelationsamongbeliefs,either–asintheformercase–fromaverticalrelationofsupportbetweensupposedlybasicbeliefsandothersdependentuponthem,or–asinthelatter–fromthemutualsupportofbeliefsinanappropriatelyunderstoodsystem.Generallycharacterized,internalisttheoriesassertorassumethatabeliefcannotbejustifiedforanepistemicsubjectSunlessShasaccesstowhatprovidesthejustification,eitherinfactorinprinciple.Thesetheoriesgenerallyinvolvethestronger‘infact’requirementbecauseS’sbeingjustifiedinbelievingpisstandardlycashedintermsofhishavingreasonsfortakingptobetrue,wherehavingreasonsistobeunderstoodinanoccurrentsense.Hereanobjectionimmediatelysuggestsitself.AnyShasonlyfiniteaccesstowhatmightjustifyorunderminehisbeliefs,andthataccessisconfinedtohisparticularview-point.Itseemsthatfulljustificationforhisbeliefswouldrarelybeavailable,becausehisexperiencewouldberestrictedtowhatisnearbyinspaceandtime,andhewouldbeentitledtoholdonlythosebeliefswhichhislimitedexperiencelicensed.Arelatedobjectionisthatinternalismseemsinconsistentwiththefactthatmanypeopleappeartohaveknowledgedespitenotbeingsophisticatedenoughtorecognizethatthus-and-soisareasonforbelievingp–thatisthecase,forexample,withchildren.Amoregeneralobjectionstillisthatrelationsbetweenbeliefs,whetherofthefoun-dationalistorcoherencetype,mightobtainwithoutthebeliefsinquestionbeingtrueofanythingbeyondthemselves.Onecouldimagineacoherentfairytale,say,whichinnopointcorrespondstosomeexternalreality,butinwhichbeliefsarejustifiednever-thelessbytheirmutualrelations.42\nEPISTEMOLOGYThisuneasyreflectionpromptsthethoughtthatthereshouldbeaconstraintontheoriesofjustification,intheformofademandthatthereshouldbesomesuitableconnectionbetweenbeliefpossessionandexternalfactors–thatis,somethingotherthanthebeliefsandtheirmutualrelations–whichdeterminestheirepistemicvalue.Thisaccordinglypromptstheideaofanalternative:externalism.1.5Reliability,causalityandtruth-trackingExternalismistheviewthatwhatmakesSjustifiedinbelievingpmightnotbeanythingtowhichShascognitiveaccess.ItmightbethatthefactsintheworldareasSbelievesthemtobe,andthatindeedtheycausedStobelievethemtobesobystimulatinghisorhersensoryreceptorsintherightkindofway.Sneednotbeawarethatthisishowhisorherbeliefwasformed.SoScouldbejustifiedinbelievingpwithoutit.Onemainkindofexternalisttheoryisreliabilism,thethesis–orclusteroftheses–havingitthatabeliefisjustifiedifitisreliablyconnectedwiththetruth.Accordingtooneinfluentialvariant,theconnectioninquestionissuppliedbyreliablebelief-formingprocesses,oneswhichhaveahighsuccessrateinproducingtruebeliefs.Anexampleofareliableprocessmightbenormalperceptioninnormalconditions.Muchapparentplausibilityattachestotheoriesbasedonthenotionofexternallinkage,especiallyofcausallinkage,betweenabeliefandwhatitisabout.AnexampleofsuchatheoryisAlvinGoldman’s(1986)accountofknowledgeas‘appropriatelycausedtruebelief’,where‘appropriatecausation’takesanumberofforms,sharingthepropertythattheyareprocesseswhichareboth‘globally’and‘locally’reliable–theformermeaningthattheprocesshasahighsuccessrateinproducingtruebeliefs,thelatterthattheprocesswouldnothaveproducedthebeliefinquestioninsome‘rel-evantcounterfactualsituation’wherethebeliefisfalse.Goldman’sviewisaccordinglyaparadigmofareliabilisttheory.Anelegantsecond-cousinofthisviewisofferedbyRobertNozick(1981).Totheconditions(1)pistrueand(2)SbelievespNozickadds(3)ifpwerenottrue,Swouldnotbelievepand(4)ifpweretrue,Swouldbelieveit.Conditions(3)and(4)areintendedtoblockGettier-typecounter-examplestothejustifiedtruebeliefanalysisbyannexingS’sbeliefthatpfirmlytop’struth.S’sbelief43\nA.C.GRAYLINGthatpisconnectedtotheworld(tothesituationdescribedbyp)byarelationNozickcalls‘tracking’:S’sbelieftracksthetruththatp.Headdsrefinementsinanattempttodeflectthecounter-examplesthatphilosophersarealwaysingeniousandfertileatdevising.Ifthesetheoriesseemplausibleitisbecausetheyaccordwithourpre-theoreticalviews.Butasonecanreadilysee,thereareplentyofthingstoobjecttointhem,andacopiousliteraturedoesso.Theirmostseriousflaw,however,isthattheyarequestion-begging.TheydonotaddressthequestionofhowSistobeconfidentthatagivenbeliefisjustified;insteadtheyhelpthemselvestotwoweightyrealistassumptions,oneaboutthedomainoverwhichbeliefrangesandtheotherabouthowthedomainandSareconnected,sothattheycanassertthatSisjustifiedinbelievingagivenpevenifwhatjustifieshimliesoutsidehisownepistemiccompetence.Whateverelseonethinksofthesesuggestions,theydonotenlightenS,andthereforedonotengagethesameproblemthatinternalisttheoriesaddress.Butworstofall–soanausterecriticmightsay–thelargeassumptionstowhichthesetheorieshelpthemselvesarepreciselythosethatepistemologyshouldbeexam-ining.Externalistandcausaltheories,inwhateverguiseandcombination,arebetterdonebyempiricalpsychologywherethestandardassumptionsabouttheexternalworldandS’sconnectionswithitarepremised.Philosophy,surely,iswherethesepremisesthemselvescomeinforscrutiny.1.6Knowledge,beliefandjustificationagainConsiderthisargument:‘Ifanyoneknowssomep,thenheorshecanbecertainthatp.Butnoonecanbecertainofanything.Thereforenooneknowsanything.’Thisargu-ment(advancedinthisformbyUnger1975)isinstructive.ItrepeatsDescartes’smistakeofthinkingthatthepsychologicalstateoffeelingcertain–whichsomeonecanbeinwithrespecttofalsehoods,suchasthefactthatIcanfeelcertainthatArklewillwintheDerbynextweek,andbewrong–iswhatweareseekinginepistemology.Butitalsoexemplifiesthetendencyindiscussionsofknowledgeassuchtomakethedefi-nitionofknowledgesohighlyrestrictivethatlittleornothingpassesmuster.Shouldonecareifasuggesteddefinitionofknowledgeissuchthat,astheargumentjustquotedtellsus,noonecanknowanything?Justsolongasonehasmanywell-justifiedbeliefswhichworkwellinpractice,canonenotbequitecontenttoknownothing?Formypart,Ithinkonecan.Thissuggeststhatinsofarasthepointssketchedinprecedingparagraphshaveinter-est,itisinconnectionwiththejustificationofbeliefsandnotthedefinitionofknowledgethattheydoso.Justificationisanimportantmatter,notleastbecauseintheareasofapplicationinepistemologywherethereallyseriousinterestshouldlie–inquestionsaboutthePHILOSOPHYOFSCIENCE(chapter9),thePHILOSOPHYOFHISTORY(chapter14)ortheconceptsofevidenceandproofinLAW(seechapter13)–justificationisthecrucialproblem.Thatiswhereepistemologistsshouldbegettingdowntowork.Bycomparison,effortstodefine‘knowledge’aretrivialandoccupytoomucheffortinepistemology.ThedisagreeablepropensityofthedebategeneratedbyGettier’scounter-examples–anticipatedbeautifullyinRussell’sreviewofJames(Russell1910:95)–toproceedonachessboardof‘-isms’,asexemplifiedabove,isasymptom.44\nEPISTEMOLOGYThegeneralproblemwithjustificationisthattheproceduresweadopt,acrossallwalksofepistemiclife,appearhighlypermeabletodifficultiesposedbyscepticism.Theproblemofjustificationisthereforeinlargeparttheproblemofscepticism;whichispreciselywhydiscussionofscepticismiscentraltoepistemology.2ScepticismIntroductionThestudyandemploymentofscepticalargumentsmightinonesensebesaidtodefineepistemology.Achiefepistemologicalaimistodeterminehowwecanbesurethatourmeanstoknowledge(here‘knowledge’doesdutyfor‘justifiedbelief’)aresatisfactory.Asharpwaytoshowwhatisrequiredistolookcarefullyatscepticalchallengestoourepistemicefforts,challengeswhichsuggestwaysinwhichtheycangoawry.Ifweareablenotjusttoidentifybuttomeetthesechallenges,aprimaryepistemologicalaimwillhavebeenrealized.Scepticismisoftendescribedasthethesisthatnothingis–or,morestrongly,canbe–known.Butthisisabadcharacterization,becauseifweknownothing,thenwedonotknowthatweknownothing,andsotheclaimistriviallyself-defeating.Itismoretellingtocharacterizescepticisminthewayjustsuggested.Itisachallengedirectedagainstknowledgeclaims,withtheformandnatureofthechallengevaryingaccordingtothefieldofepistemicactivityinquestion.Ingeneral,scepticismtakestheformofarequestforthejustificationofthoseknowledgeclaims,togetherwithastatementofthereasonsmotivatingthatrequest.Standardly,thereasonsarethatcertainconsiderationssuggestthattheproposedjustificationmightbeinsufficient.Toconceiveofscepticismlikethisistoseeitasbeingmorephilosophicallytroublingandimportantthanifitisdescribedasapositivethesisassertingourignoranceorincapacityforknowledge.2.1EarlyscepticismSomeamongthethinkersofantiquity–PyrrhoofElis(c.360–c.270BC)andhisschool,andPlato’ssuccessorsinhisAcademy–expresseddisappointmentatthefactthatcen-turiesofenquirybytheirphilosophicalpredecessorsseemedtohavebornelittlefruiteitherincosmologyorethics(thislatterwasbroadlyconstruedtoincludepolitics).Theirdisappointmentpromptedthemtoscepticalviews.ThePyrrhoniansarguedthatbecauseenquiryisarduousandinterminable,oneshouldgiveuptryingtojudgewhatistrueandfalseorrightandwrong;foronlythuswillweachievepeaceofmind.AlessradicalformofscepticismovertookPlato’ssuccessorsintheAcademy.TheyagreedwithPyrrhothatcertaintymusteludeus,buttheytemperedtheirviewbyacceptingthatthepracticaldemandsoflifemustbemet.Theydidnotthinkitawork-ableoptionto‘suspendjudgement’asPyrrhorecommended,andthereforearguedthatweshouldacceptthosepropositionsortheorieswhicharemorePROBABLE(pp.308–11)thantheircompetitors.Theviewsofthesethinkers,knownasAcademicsceptics,arerecordedintheworkofSextusEmpiricus(c.150–c.225).45\nA.C.GRAYLINGInthelaterRenaissance–or,whichisthesamething,inearlymoderntimes–withreligiouscertaintiesunderattackandnewideasabroad,someofthescepticalargu-mentsoftheAcademicsandPyrrhoniansacquiredaspecialsignificance,notablyasaresultoftheusetowhichRenéDescartesputtheminshowingthattheyarepowerfultoolsforinvestigatingthenatureandsourcesofknowledge.InDescartes’sdaythesamepersoncouldbebothastronomerandastrologer,chemistandalchemist,orphysicianandmagician.Itwashardtodisentangleknowledgefromnonsense;itwasevenhardertodisentanglethosemethodsofenquirywhichmightyieldgenuineknowledgefromthosethatcouldonlydeepenignorance.Sotherewasanurgentneedforsomesharp,cleanepistemologicaltheorizing.InhisMeditations(1986)Descartesaccordinglyidentifiedepistemologyasanessentialpreliminarytophysicsandmathematics,andattemptedtoestablishthegroundsofcertaintyasapropaedeutictoscience.Descartes’sfirststepinthattaskwastoadaptandapplysomeofthetraditionalargumentsofscepticism.(Ishallcommentonhisuseofscepticismagainlater.)TheAnatomyofScepticismScepticalargumentsexploitcertaincontingentfactsaboutourwaysofacquiring,testing,rememberingandreasoningaboutourbeliefs.Anyproblemthatinfectstheacquisitionandemploymentofbeliefsaboutagivensubjectmatter,andinparticularanyproblemthatinfectsourconfidencethatweholdthosebeliefsjustifiably,threatensourholdonthatsubjectmatter.Thecontingentfactsinquestionrelatetothenatureofperception,thenormalhumanvulnerabilitytoerrorandtheexistenceofstatesofmind–forexample,dreaminganddelusion–whichcanbesubjectivelyindistinguishablefromthosethatwenormallytaketobeappropriateforacquiringjustifiedbeliefs.Byappealingtotheseconsiderationsthescepticaimstoshowthattherearesignificantquestionstobeansweredaboutthedegreeofconfidencethatweareentitledtoreposeinourstandardepistemicpractices.Scepticalconsiderationsposeproblemsforepistemologistsofboththerationalistandtheempiricistcamps.Thisdivisionintocompetingschoolsofthoughtaboutknowledgeisroughbutuseful,givingashorthandwayofmarkingthedifferencebetweenthosewhoholdthatreasonisthechiefmeanstoknowledge,andthosewhoaccordthatroletoexpe-rience.Rationalistsemphasizereasonbecauseintheirviewtheobjectsofknowledgearepropositionsthatareeternally,immutablyandnecessarilytrue–theexamplestheyofferarethepropositionsofmathematicsandlogic–andthese,theysay,canonlybeacquiredbyreasoning.Empiricistsholdthatsubstantiveandgenuineknowledgeoftheworldcanonlybelearnedthroughexperience,bymeansofthesensesandtheirextensionviasuchinstrumentsastelescopesandmicroscopes.Therationalistneednotdenythatempiricalawarenessisanimportant,evenanineliminable,aidtoreason,norneedtheempiricistdenythatreasonisanimportant,evenanineliminable,aidtoexperience;butbothwillinsistthatthechiefmeanstoknowledgeisrespectivelyoneortheother.Therefinementsofdebateaboutthesemattersmeritdetailedexaminationforwhichthisisnottheplace.Forpresentpurposes,thepointtonoteisthatscepticismisaproblemforbothschoolsofthought.Forboth,possibilitiesoferroranddelusionposeachallenge.Fortheempiricistinparticular,tothesemustbeaddeddistinctiveproblemsaboutperception.46\nEPISTEMOLOGY2.2Error,delusionanddreamsOnecharacteristicpatternofscepticalargumentisdrawnfromasetofconsiderationsabouterror,delusionsanddreams.Considertheerrorargumentfirst.Wearefalliblecreatures;wesometimesmakemistakes.If,however,weareevertobeabletoclaimtoknow(thatis,atleasttobejustifiedinbelieving)somepropositionp,wemustbeabletoexcludethepossibilitythatatthetimeofclaimingtoknowpweareinerror.Butsincewetypically,oratleastfrequently,arenotawareofourerrorsaswemakethem,andmightthereforeunwittinglybeinerrorasweclaimtoknowp,wearenotjustifiedinmakingthatclaim.Thesameapplieswhenapersonisthesubjectofadelusion,illusionorhallucina-tion.Sometimespeopleundergoingoneorotherofthesestatesdonotknowthattheyaredoingso,andtakethemselvestobehavingveridicalexperiences.Clearly,althoughtheythinktheyareinastatewhichlendsitselftotheirbeingjustifiedinclaimingtoknowp,theyarenotinsuchastate.Thereforetheyarenotjustifiedinclaimingtoknowp.Soinorderforanyonetoclaimknowledgeofsomep,theymustbeabletoexcludethepossibilitythattheyarethesubjectofsuchstates.ThispatternofargumentisatitsmostfamiliarintheargumentfromdreamingemployedbyDescartes.Onewayofsettingitoutisasfollows.WhenIsleepIsometimesdream,andwhenIdreamIsometimes–indeed,often–donotknowthatIamdream-ing.SoIcanhaveexperiencesthatappeartobeveridicalwakingexperiencesonthebasisofwhichItakemyselftobejustifiedinclaimingtoknowsuchandsuch.ButbecauseIamdreaming,Idonotinfactknowsuchandsuch;ImerelydreamthatIdo.MightInotbedreamingnow?IfIcannotexcludethepossibilitythatIamnow,atthismoment,dreaming,IamunabletoclaimknowledgeofthethingsIatthismomenttakemyselftoknow.Forexample,itseemstomethatIamsittingatadesknexttoawindowadmittingaviewoftreesandlawns.ButbecauseImightbedreamingthatthisisso,Icannotclaimtoknowit.Intheseargumentsthepossibilityoferror,delusionordreamingactsaswhatmightbecalleda‘defeater’toknowledgeclaims.Thepatternis:ifoneknowsp,thennothingisactingtosubvertone’sjustificationtoclaimknowledgeofp.Butonecanseemtooneselffullyentitledtoclaimtoknowsomep,andinfactlackthatentitlement,astheforegoingconsiderationsshow.Soourclaimstoknowledgeareinneedofbettergroundsthanwestandardlytakeourselvestohave.Wemustfindawayofdefeatingthedefeaters.2.3PerceptionBothrationalistandempiricistviewsaboutthesourcesofknowledgearethreatenedbytheargumentsjustsketched.Argumentsthatposeparticularproblemsforempiricismaresuggestedbythenatureandlimitationsofperception,thebestcurrentaccountofwhichtellsussomethinglikethefollowingstory.Lightreflectsfromthesurfacesofobjectsinthephysicalenvironmentandpassesintotheeyes,whereitirritatesthecellsoftheretinasinsuchawayastotriggerimpulsesintheopticnerves.Theopticnervesconveytheseimpulsestotheregionofthecerebralcortexthatprocessesvisualdata,wheretheystimulatecertainsorts47\nA.C.GRAYLINGofactivity.Asaresult,inwaysstillmysterioustoscienceandphilosophy,coloured‘motionpictures’ariseinthesubject’sconsciousness,representingtheworldoutsidehisorherhead.Thisremarkabletransactionisrepeatedmutatismutandisintheothersensorymodalitiesofhearing,smell,tasteandtouch,givingrisetoperceptionsofhar-moniesandmelodies,perfumesandpiquancies,smoothness,softness,warmth–andsoforth.Thismodelcanbeusedtofurnishanotherscepticalapplicationofthedefeaterargu-ment.Thecomplexcausalstorythustoldisonewhich–sothescepticcanpointout–mightbeinterruptedinproblematicwaysatanypointalongitslength.Theexperienceswhichwesayresultfromtheinteractionofoursensesandtheworldmightoccurinusforotherreasons.Theymightoccurwhen,asnotedabove,wedream,hallucinateorsufferdelusions;or,tobefanciful,theymightbeproducedinusbyagod,orbyascientistwhohasconnectedourbrainstoacomputer.Fromthepointofviewoftheexperiencingsubject,theremightbenowayoftellingthedifference.So,saysthesceptic,unlesswefindmeansofexcludingthesepossibilities,wearenotentitledtoclaimknowledgeofwhatwestandardlytakeourselvestoknow.2.4PerceptualrelativitiesThesesameconsiderationsaboutperceptioncanpromptscepticalchallengebyadif-ferentroute.AlittlereflectionofthekindtaughtusbyLocke,BERKELEY(1685–1753)(chapter30)andotherearliercontributorstothedebateshowsthatsomeoftheseprop-ertiesweseemtoperceiveinobjectsarenot‘intheobjectsthemselves’butareinfactcreaturesoftheperceptualrelation.Thequalitiesofobjects–theircolour,taste,smell,soundandtexture–varyaccordingtotheconditionoftheperceiverortheconditionsunderwhichtheyareperceived.Thestandardexamplesarelegion:grassisgreenindaylight,blackatnight;tepidwaterfeelswarmtoacoldhand,cooltoahothand;objectslooklargefromcloseby,smallfromfaraway;andsoon.Theseperceptualrelativitiesarecitedbythesceptictoraisequestionsnotjustaboutwhetherperceptionisatrustworthysourceofinformationabouttheworld,butwhethertheworldcanbesaidtoexistindependentlyofperceptionatall.Forwhatifthepropertiesbywhosemeanswedetectthepresenceofobjectscannotbedescribedapartfromtheirbeingobjectsofperception?Considertheoldconundrumwhetherasoundismadebythetreethatfallsintheforestwhennosentientbeingispresenttohearitdoso.Theanswer,onastandardtheoryofperceptioncurrentincontemporaryscience,isthatthetreefallsincompletesilence.Forifthereisnoeartohear,thereisnosound;thereareonlyatbesttheconditions–vibratingairwaves–whichwouldcausesoundtobeheardiftherewerenormallyfunctioningeardrums,auralnervesandtheresttobestimulatedbythem.Theseconsiderationssuggestascepticalpictureinwhichperceiversareinsome-thinglikethefollowingpredicament.Imagineamanwearingavisorlesshelmetwhichsoencloseshisheadthathecannotsee,hear,tasteorsmellanythingoutsideit.Imaginethatacamera,amicrophoneandothersensorsareaffixedtothetopofthehelmet,transmittingpicturesandotherinformationtoitsinterior.Andsupposefinallythatitisimpossibleforthewearertoremovethehelmettocomparethisinformationwithwhateverisoutside,sothathecannotcheckwhetheritfaithfullyrepresentsthe48\nEPISTEMOLOGYexteriorworld.Somehowthewearerhastorelyontheintrinsiccharacteroftheinformationavailableinsidethehelmettojudgeitsreliability.Heknowsthattheinformationsometimescomesfromsourcesotherthantheexteriorworld,asindreamsanddelusions;hehasdeducedthattheequipmentaffixedtothehelmetworksupontheincomingdataandchangesit,forexampleaddingcolours,scentsandsoundstoitspictureofwhatintrinsicallyhasnoneoftheseproperties(atveryleast,inthoseforms);heknowsthathisbeliefsaboutwhatliesoutsidethehelmetrestontheinferenceshedrawsfromtheinformationavailableinsideit,andthathisinferencesareonlyasgoodashisfallible,error-pronecapacitiesallowthemtobe.Givenallthis,asksthesceptic,havewenotajobofworktodotojustifyourclaimstoknowledge?2.5MethodologicalandproblematicscepticismBeforeconsideringtheseargumentsandcanvassingsomewaysofrespondingtothem,itisimportanttonotetwothings.Oneisthatscepticalargumentsarenotbestdealtwithbyattemptsatpiecemeal–thatis,one-by-one–refutation.Thesecondisthatthereisavitallyimportantdistinctiontobedrawnbetweentwowaysinwhichscepticismcanbeemployedinepistemology.Itisimportanttonotethesemattersbecauseother-wisetheprimafacieimplausibilityofmostscepticalargumentswillmisleadusintounderestimatingtheirsignificance.Itakeeachpointinturn.Attemptedrefutationofscepticalargumentspiecemealis,arguably,futilefortwogoodreasons.Assuggestedattheoutset,scepticalargumentsareattheirstrongestnotwhentheyseektoprovethatweareignorantaboutsomesubjectmatterbutwhentheyaskustojustifyourknowledgeclaims.Achallengetojustifyisnotaclaimoratheory,andcannotberefuted;itcanonlybeacceptedorignored.Sincethescepticoffersreasonswhyjustificationisneeded,theresponsemightbetoinspectthosereasonstoseewhetherthechallengeneedstobemet.Thisindeedisonegoodresponsetoscepti-cism.Wherethereasonsarecogent,thenextgoodresponseistotrytomeetthechal-lengethusposed.Thesecondreasonisthatscepticalargumentstakentogetherhavethejointeffectofshowingthatthereisworktobedoneifwearetogetasatisfactoryaccountofknowl-edge–andscepticismindicateswhatisneeded.Ifonecouldrefute,orshowtobeungrounded,oneoranotherindividualscepticalargument,otherswouldbeleftinplacestilldemandingthatsuchanaccountbesought.ThesepointscanbeillustratedbyconsideringGilbertRyle’s(1900–76)attempttorefutetheargumentfromerrorbyusinga‘polarconcept’argument.Therecannotbecounterfeitcoins,Ryleobserved,unlesstherearegenuineones,norcrookedpathsunlesstherearestraightpaths,nortallmenunlessthereareshortmen.Manyconceptscomeinsuchpolarities,afeatureofwhichisthatonecannotgraspeitherpoleunlessonegraspsitsoppositeatthesametime.Now,‘error’and‘gettingitright’areconcep-tualpolarities.Ifoneunderstandstheconceptoferror,oneunderstandstheconceptofgettingitright.Buttounderstandthislatterconceptistobeabletoapplyit.Soourverygraspoftheconceptoferrorimpliesthatwesometimesgetthingsright.Ryleobviouslyassumedthattheerrorscepticisclaimingthat,forallweknow,wemightalwaysbeinerror.Accordinglyhisargument–thatifweunderstandtheconcept49\nA.C.GRAYLINGoferror,wemustsometimesgetthingsright–isaimedatrefutingtheintelligibilityofclaimingthatwemightalwaysbewrong.Butofcoursetheerrorscepticisnotclaim-ingthis.Heorsheissimplyaskinghow,giventhatwesometimesmakemistakes,wecanruleoutthepossibilityofbeinginerroronanygivenoccasionofjudgement–say,atthispresentmoment.ButthescepticneednotconcedethemoregeneralclaimsthatRylemakes,namely,thatforanyconceptualpolarity,bothpolesmustbeunderstood,and–furtherandevenmoretendentiously–tounderstandaconceptistoknowhowtoapplyit,andforittobeapplicableisforitactuallytobeapplied(ortohavebeenapplied).Thislastmoveisquestion-beggingenough,butsoistheclaimaboutconceptualpolaritiesitself.Forthescepticcanreadilycitecasesofconceptualpolarities–‘perfect–imperfect’,‘mortal–immortal’,‘FINITE–INFINITE’(chapter11)–whereitisbynomeansclearthatthemoreexoticpolesapplytoanything,oreventhatwereallyunderstandthem.Afterall,takingatermandattachinganegativeprefixtoitdoesnotguaranteethatwehavetherebygraspedanintelligibleconcept.Thesecommentssuggestthatscepticalarguments,evenifsinglytheyappearimplausible,jointlyinviteaseriousresponse;whichiswhat,inlargemeasure,episte-mologyseekstooffer.Butthereisstillthematterofthedistinctionbetweenmethod-ologicalandproblematicscepticismtobeexplained,andhereabriefrecapitulationofDescartes’suseofscepticalargumentswillbehelpful.Descartes’sMethodofDoubtDescartes’saimwastofindabasisforknowledge,whichhedidbylookingforastartingpointaboutwhichhecouldbecertain.Tofindcertaintyheneededtoruleoutanythingthatcouldbedoubted,howeverabsurdthatdoubt,foronlyinthiswaywouldwebeleftwithwhatistrulyindubitable.InthefirstMeditationheembarksonthistaskbyborrowingsomescepticalargumentsfromtheancients.Firsthecitesthefactthatwecanbemisledinperception.Butthisisnotathoroughgoingenoughscepticism,forevenifwemisperceivethereisstillmuchthatwecanknow.Sohenextconsidersthepossibilitythatonanyoccasionofclaimingtoknowsomething,onemightbedreaming.Thisscepticalthoughtcatchesmoreinitsnet,butisstillinsufficient,forevenindreamswecanknowsuchthingsas,forexample,mathematicaltruths.So,togetassweepingaconsiderationaspossible,Descartesintroducesthe‘evildemon’idea.Herethesupposi-tionisthatwithrespecttoeverythingaboutwhichonecouldpossiblybemisled,anevildemonisindeedmisleadingone.Famously,whatsuchabeingcannotmisleadoneaboutiscogitoergosum–whenonethinks‘Iexist’,thispropositionistrue.ItisessentialtonotethatDescartes’suseoftheseargumentsispurelymethodological.TherestoftheMeditationsisdevotedtoshowingthatweknowagreatdeal,becausethefact(asDescartesunsuccessfullytriestoprove)thatthereisagoodDEITY(seechapter15)guaranteesthat,justsolongasweuseourfacultiesresponsibly,whateverisperceivedwithclearnessanddistinctnesstobetruewillindeedbetrue.Thisisbecauseagooddeity,unlikeanevilone,wouldnotwishustoembraceignorance.Descarteswasbynomeansasceptic,nordidhethinkthatscepticalarguments,leastofalltheoneemployedasadevicetosetasideasmanybeliefsaspossible,werepersuasive.The‘methodofdoubt’ismerelyatool.50\nEPISTEMOLOGYDescartes’successors,however,werefarmoreimpressedbythescepticalargumentsheemployedthanhisanswertothem.Forthetraditionofepistemologicalthinkingafterhistime,theseandalliedscepticalargumentswerenotmeremethodologicaldevices,butseriousproblemsrequiringsolution.HencethedistinctionIdrawherebetweenmethodologicalandproblematicscepticism.Itisclearthattherearescepticalconsiderationsthathavemerelymethodologicalutility,andarenotgenuinelyproblematic,becausetheydonotrepresentastableandcogentchallengetoourordinaryepistemologicalstandards.Descartes’s‘evildemon’isacaseinpoint.Sincethehypothesisthatthereissuchathingisasarbitraryandgroundlessasahypothesiscanget,itdoesnotmeritbeingtakenseriouslyotherwisethanasaploytomakeapoint.Butscepticalconsiderationsaboutperception,error,delusionanddreamsraisemoreinterestingandtroublinggeneralissues,andaccord-inglymeritexamination.AmongthemanythingsworthnotingaboutDescartes’sdiscussionarethefollow-ingtwo.Firstly,ashintedearlier,hisquestforcertaintyisarguablymisconceived.Certaintyisapsychologicalstateonecanbeinindependentlyofwhetherornotonebelievestruly.Thefalsityofabeliefisnobartoone’sfeelingcertainthatitisotherwise.Descartessoughttospecifywaysofrecognizingwhichofourbeliefsaretrue,butheledhimselfintotalkofcertaintybecause–andthisisthesecondpoint–heassumedthatepistemology’staskistoprovideonewithawayofknowing,fromone’sownsub-jectiveviewpoint,whenonepossessesknowledge.Accordingly,hestartswiththeprivatedataofasingleconsciousnessandattemptstomoveoutsideit,seekingguar-anteesfortheprocessenroute.NearlyallofDescartes’ssuccessorsinepistemology,uptoandincludingRUSSELL(1872–1970)(chapter37)andAyer(1910–89),acceptedthisperspectiveontheirtask.InthisrespectatleasttheyareallthereforeCartesians.Itislargelyforthisreason,asweshallseehintedbelow,thattheyfoundithardtomeetscepticism’schallenge.2.6SomeresponsestoscepticismThescepticalchallengetellsusthatwesufferanepistemicplight,namely,thatwecanhavethebestpossibleevidenceforbelievingsomep,andyetbewrong.Statedsuccinctlyandformally,scepticismistheobservationthatthereisnothingcontradictoryintheconjunctionofstatementssembodyingourbestgroundsforagivenbeliefpwiththefalsityofp.Aninformativerepresentationofscepticismthussummarizedisasfollows.Scepti-calargumentsopenagapbetween,ontheonehand,thegroundsaputativeknowerhasforsomeknowledgeclaim,and,ontheotherhand,theclaimitself.Responsestoscepticismgenerallytaketheformofattemptseithertobridgethisgaportocloseit.Thestandardperceptualmodel,inwhichbeliefsareformedbysensoryinteractionwiththeworld,postulatesacausalbridgeacrossthegap;butthatbridgeisvulnerabletoscepticalsabotage,sothecausalstoryatleastneedssupport.Descartes,asnoted,iden-tifiedtheepistemologicaltaskastheneedtospecifyaguarantee–callitX–which,addedtooursubjectivegroundsforbelief,protectsthemagainstscepticismandthuselevatesbeliefintoknowledge.HiscandidateforXwasthegoodnessofadeity;reject-ingthiscandidate(whilecontinuingtoaccepthisviewoftheepistemologicaltask)51\nA.C.GRAYLINGobligesustofindanalternative.IfanXcannotbefoundtosupportabridgeacrossthescepticalgap,theoptionistotryclosingit–ormoreaccurately,toshowthatthereisnogapatall.BoththequestforXandtheclosingofthegaphaveconstitutedmajorepistemologicalendeavoursagainstscepticisminmodernphilosophy.Someoftheseendeavours,inbrief,areasfollows.Descartes’simmediatesuccessorswere,asmentioned,unpersuadedbyhisattempttobridgethegapbyinvokingagooddivinitytoserveasX.LOCKE(chapter29),withoutmuchfanfare,employedaweakerversionoftheCartesianexpedientbysayingthatwecanignorescepticalthreatstothecausalstorybecause‘thelightthatissetupinusshinesbrightenoughforallourpurposes’.FromLocke’spointofviewitdoesnotmatterwhethertheinnerlightissetupbyGodornature;thepointisthatthereissomething–X,theinnerlightwhichcouldbe,perhaps,reason,empiricalintuitionornativetrustinthereliabilityofthesenses–thatgivesusgroundsforacceptingourordinaryknowledge-acquiringmeansasadequate.Others,notcontentwithsuchunsatisfactorymoves,lookforXelsewhere,andclaimtofinditinsomeversionoffoundationalism,thethought–sketchedabove–thatourepistemicsystemhasabasisinspecialbeliefsthatareinsomewayself-justifyingorself-evidentandwhich,inconjunctionwiththeevidenceweordinarilyemployinmakingknowledgeclaims,securesthemagainstscepticism.Aswesawearlier,achiefgroundforrejectingsuchtheoriesisallegedtobethatnoneofthemidentifiessatisfactorycan-didatesfor‘foundations’.Butonestimulatingwayofmakingsomethinglikeafounda-tionalistcaseisofferedbyKant,whoseattemptpromptedothers.2.7TranscendentalargumentsKANT(1724–1804)(chapter32)regardedfailuretorefutescepticismasa‘scandal’tophilosophy,andofferedhisCritiqueofPureReason(1929)asasolution.Histhesisisthatourmindsaresoconstitutedthattheyimposeaframeworkofinterpretativeconceptsuponoursensoryinput,amongthemthoseoftheobjectivityandcausalintercon-nectednessofwhatweperceive.ApplicationoftheseconceptstransformsmerepassivereceiptofsensorydataintoEXPERIENCE(pp.726–33)properlysocalled.Ourfacultiesaresuchthatwhenrawdatacomesundertheinterpretingactivityofourconcepts,theyhavealreadyhadspatialandtemporalformconferredonthembythenatureofoursensorycapacities;allourexperience,consideredasrelatingtowhatisoutsideus,isexperienceofaspatiallystructuredworld,andallourexperience,consideredasrelat-ingtoitsreceivedcharacterinourminds,isofatemporallystructuredworld.Uponthespatio-temporaldatathusbroughtbeforeourmindsweimposethecategories,thatis,theconceptsthatmakeexperiencepossiblebygivingititsdeterminatecharacter.AndhereisKant’spoint:ifthescepticasksustojustifyourclaimstoknowledge,wedosobysettingoutthesefactsabouthowexperienceisconstituted.KantclaimedHUME(1711–76)(chapter31)ashisinspirationfortheseideas,becauseHumehadarguedthatalthoughwecannotrefutescepticism–reasonwasnot,heclaimed,uptothetask–weshouldnotbetroubled,forhumannatureissoconsti-tutedthatwesimplycannothelphavingthebeliefsthatscepticismchallengesustojustify.Thosebeliefsinclude,forexample,thatthereisanexternalworld,thatcausalrelationsholdbetweeneventsintheworld,thatinductivereasoningisreliable,andso52\nEPISTEMOLOGYforth.FromthishintKantelaboratedhistheorythattheconceptsthatthescepticasksustojustifyareconstitutivefeaturesofourcapacitytohaveanyexperienceatall.Thestrategy,ifnotthedetails,ofKant’sattackuponscepticismhaspromptedinter-estinmorerecentphilosophy.Theargumentheemploysisatranscendentalargument,brieflycharacterizableasonewhichsaysthatbecauseAisanecessaryconditionforB,and,becauseBisthecase,Amustbethecasealso.Anexampleofsuchanargumentinactionagainstscepticismisasfollows.Atypicalscepticalchallengeconcernsbeliefintheunperceivedcontinuedexistenceofobjects.Whatjustifiesourholdingthisbeliefandpremisingsomuchuponit?Thetranscendentalargueranswersthatbecausewetakeourselvestooccupyasingleunifiedworldofspatio-temporalobjects,andbecauseonthisviewspatio-temporalobjectshavetoexistunperceivedinordertoconstitutetherealmassingleandunified,abeliefintheirunperceivedcontinuedexistenceisaconditionofourthinkingbothabouttheworldandourexperienceofitinthisway.Sincewedoindeedthinkthisway,thebeliefthatthescepticasksustojustifyistherebyjustified.AcontemporarythinkerwhomakesnotableuseofthisstyleofargumentisP.F.Strawson(b.1919).2.8IdealismandphenomenalismThereis,inparalleltotheseKantianwaysofrespondingtothescepticalchallenge,anotherapproach,whichdeniestheexistenceofascepticism-generatinggap.ThechieffiguresinthiscampareBerkeleyand,morerecently,thephenomenalists,who–allowingfordifferencesamongthem,andforthefactthatthetwolatterheldtheseviewsonlyforpartoftheircareers–includeMILL(1806–73)(chapter35),RussellandAyer.InBerkeley’sview,scepticismarisesfromthinkingthatbehindorbeyondoursensoryexperiencesthereliesamaterialworld.Theword‘material’means‘madeofmatter’,and‘matter’isatechnicalphilosophicaltermsupposedtodenoteanempiricallyunde-tectablesubstancebelievedbyBerkeley’sphilosophicalpredecessorstounderpinthesensorilydetectablepropertiesofthings,suchastheircolours,shapesandtextures.Berkeleyrejectedtheconceptofmatterthusunderstood–itisacommonmisreadingofhimtotakeitthathetherebydeniedtheexistenceofphysicalobjects;hedidnosuchthing–arguingthatbecausephysicalobjectsarecollectionsofsensiblequalities,andbecausesensiblequalitiesareideas,andbecauseideascanonlyexistifperceived,theexistenceofobjectsthereforeconsistsintheirbeingperceived;ifnotbyfinitemindssuchasourown,theneverywhereandatalltimesbyaninfinitemind.(WemaynotethatBerkeleythoughtthathisrefutationofscepticismwasatthesametimeapower-fulnewargumentfortheexistenceofGod.)Berkeley’shabitofsayingthatthingsexist‘inthemind’hasleduncriticalreaderstosupposehemeansthatobjectsexistonlyinone’shead,whichiswhatasubjectiveidealistorsolipsistmighttrytohold.Berkeley’sidealism,whetherornotitisotherwisedefensible,isatleastnotquitesounstableaview.His‘inthemind’shouldbereadasmeaning‘withessentialreferencetoexperienceorthought’.Forpresentpurposes,thepointisthatBerkeleysoughttorebutscepticismbydenyingtheexistenceofagapbetweenexperienceandreality,onthegroundsthatexperience53\nA.C.GRAYLINGandrealityarethesamething.(Hehadatheoryofhow,despitethis,wecannever-thelessimagine,dreamandmakemistakes.)Thephenomenalists,withoneveryimpor-tantdifference,arguedlikewise.Theirview,brieflystated,isthatallourbeliefsabouttheworldarederivedfromwhatappearstousinexperience.Whenweanalyseappear-ances–the‘phenomena’–weseethattheyarebuiltoutofthebasicdataofsense:thesmallestvisiblecolourpatchesinourvisualfields,theleastsoundsinourauditoryfields.Outofthesesense-datawe‘logicallyconstruct’thechairsandtables,rocksandmountainsconstitutingthefamiliarfurnitureoftheeverydayworld.Analternativebutequivalentwayofputtingthispoint,thephenomenalistsclaim,istosaythatstatementsaboutphysicalobjectsaremerelyconvenientshorthandforlongerandmorecomplicatedstatementsabouthowthingsseemtousintheusualemploymentofoursensorycapacities.Andtosaythatobjectscontinuetoexistunper-ceivedistosay–inMill’sphrase–thattheyare‘permanentpossibilitiesofsensation’,meaningthatonewouldexperiencethemifcertainconditionswerefulfilled.Berkeleyholdsthatthingsremaininexistencewhennotperceivedbyfinitemindsbecausetheyareperceivedbyadeity.Thephenomenalistsarguethatwhatitmeanstosaythatthingsexistunperceivedistosaythatcertaincounterfactualconditionalsaretrue,namely,thoseassertingthatthethingsinquestionwouldbeperceivedifsomeper-ceiverweresuitablyplacedwithrespecttothem.Theseconditionalsarenotoriouslyproblematic,becauseitisnotclearhowtounderstandthem.What,inparticular,makesthemtruewhentheyare(orseemquiteobviouslytobe)true?Theusualanswers,intermsofpossibleworlds,laws,idealregularitiesandsimilarexotica,dolittletohelp.ItisnotclearthatmuchofanadvanceismadeoverBerkeley’subiquitousdeitybysub-stitutingbarelytruecounterfactualsinitsplace.Berkeley’sviewhasthemodestattrac-tionthateverythingintheworldisactual–anythingthatexistsisperceived–whereasinthephenomenalist’suniversemostofwhatexistsdoessoasapossibilityratherthananactuality,namely,asapossibilityofperception.Onethingisclear,atleast:thatonedoesnotgetphenomenalismsimplybysub-tractingthetheologyfromBerkeley’stheory.Onehastodothatandthen,intheresult-ingmetaphysicalgap,substituteacommitmenttotheexistenceofbarelytruecounterfactuals,withanaccompanyingcommitmenttotheexistenceofpossibilia.BothBerkeley’stheoryandphenomenalismthusexacthighpricesforclosingthescepticalgap.2.9Scepticalepistemologyversusanti-CartesianismSomeepistemologistsdonotattempttorefutescepticismforthegoodreasonthattheythinkittrueoratleastirrefutable.Theirviewsmightbesummarizedasstatingthatscepticismistheinevitableresultofepistemologicalreflection,soweshouldaccepteitherthatweareonlyevergoingtohaveimperfectlyjustifiedbeliefs,alwayssubjecttorevisioninthelightofexperience,orthatwehavetorecognizethatscepticism,despitebeingirrefutable,isnotapracticaloption,andthereforewehavetoliveasmostpeopleanywaydo,namely,bysimplyignoringit.SomecommentatorsonHumeinterprethimastakingthislatterviewofthematter,andaccordinglycallitthe‘Humean’responsetoscepticism.InStroud(1984)andStrawson(1985),somethingliketheHumeanviewistaken.54\nEPISTEMOLOGYOthersintherecentdebatearemorecombative,amongthemDEWEY(1859–1952)(seechapter36)andWITTGENSTEIN(1889–1951)(chapter39).Despitesubstantialdifferencesinotherrespects,thesetwothinkersholdaninterestingviewincommon,whichisthatscepticismresultsfromacceptingtheCartesianstarting-pointamongtheprivatedataofindividualconsciousness.Ifinstead,theysay,webeginwiththepublicworld–withconsiderationsrelatingtofactsabouttheessentiallypubliccharacterofhumanthoughtandlanguage–adifferentpictureemerges.DeweyarguedthattheCartesianmodelmakestheepistemicsubjectamerelypassiverecipientofexperiences,likesomeonesittinginthedarkofacinemawatchingthescreen;but,hepointedout,oursisinfactaparticipantperspective–weareactorsintheworld,andouracquisitionofknowledgeistheresultofourdoingsthere.WittgensteincontestedtheverycoherenceoftheCartesianapproachbyarguingthatPRIVATELANGUAGE(pp.817–20)isimpossible.AprivatelanguageinWittgenstein’ssenseisonethatislogicallyavailableonlytoonespeaker,whichiswhataCartesiansubjectwouldneedinordertobegindiscoursingabouthisprivateinnerexperience.Hisargumentisthis:languageisarule-governedactivity,andoneonlysucceedsinspeak-ingalanguageifonefollowstherulesfortheuseofitsexpressions.Butasolitarywould-belanguage-userwouldnotbeabletotellthedifferencebetweenactuallyfol-lowingtherulesandmerelybelievingthatheisdoingso;sothelanguagehespeakscannotbelogicallyprivatetohimself;itmustbeshareablewithothers.Indeed,Wittgen-steinarguesthatlanguagecanonlybeacquiredinapublicsetting(helikenslanguage-learningtothetrainingofanimals;tolearnalanguageistoimitatethelinguisticbehaviourofone’steachers),whichsimilarlyweighsagainsttheideathattheCartesianprojectiseveninprinciplepossible.Theanti-scepticalpossibilitiesoftheprivatelanguageargumentseemnottohavebeenwhollyapparenttoWittgensteinhimself.Indraftnotesonscepticismandknowl-edgewritteninthelastmonthsofhislife–laterpublishedunderthetitleOnCertainty(1969)–heoffersaresponsetoscepticism,whichmarksareturntoamoretraditionalapproach,notunlikethatofferedbyHumeandKant.Itisthattherearesomethingswehavetoacceptinordertogetonwithourordinarywaysofthinkingandspeaking.Suchpropositionsasthatthereisanexternalworld,orthattheworldcameintoexis-tencealongtimeago,aresimplynotopentodoubt;itisnotanoptionforustoques-tionthem.Northerefore,saysWittgenstein,canwesaythatweknowthem,becauseknowledgeanddoubtareintimatelyrelated,inthattherecanonlybeknowledgewheretherecanbedoubt,andviceversa.Thepropositionswecannotdoubtconstitutethe‘scaffolding’ofourordinarythoughtandtalk,or–Wittgensteinvarieshismetaphors–theyarelikethebedandbanksofariver,downwhichthestreamofordinarydiscourseflows.Inthissensethebeliefsthatscepticismattemptstochallengearenotopentonegotiation;which,saysWittgenstein,disposesofscepticism.ThesethoughtsareassuggestiveastheyareinthephilosophiesofHumeandKant;butoneoftheproblemswithWittgenstein’swayofputtingthemisthatheusesfoun-dationalistconceptsindescribingtherelationof‘grammatical’propositionstoordinaryones,butrepudiatesfoundationalismassuch,andseemstoallowaversionofrela-tivismbydoingso–theriver’sbedandbanks,hesays,mightintimebewornaway.Butrelativismisjustscepticismindisguise–itis,indeed,arguablythemostpowerfuland55\nA.C.GRAYLINGtroublingformofscepticism,foritistheviewthatknowledgeandtrutharerelativetoapointofview,atime,aplace,aculturalorcognitivesetting:andknowledgeandtruththusunderstoodarenotknowledgeandtruth.ConcludingRemarksThereismuchonewouldliketoinsistuponintryingcorrectlytodescribetheworkthatneedstobedoneinepistemology,forthatisthenecessarypreliminarytomakingwhatprogresswecan.HereIshallsimplyunderlineacoupleofremarksalreadymadeabove.Firstly,debatesoverthedefinitionof‘knowledge’seemtometobeaside-show.Thejustificationofclaimsinthenaturalsciences,thesocialsciences(notleasthistory)andlawiswheretherealworkcriesouttobedoneinepistemology.Andthiscommentappliesonlytotheempiricalcase:whatoftheepistemologicalquestionsthatpressinETHICS(chapter6)andthePHILOSOPHYOFMATHEMATICS(chapter11)?Therecanbenoguarantee–andindeeditisunlikely–thathighgeneralitiesaboutjustificationandknowledgewillapplyunivocallyacrossallthesefields.‘Justification’isadummyconceptthatneedstobecashedoutintermsparticulartoparticularfields;somuchshouldbeobviousfromthefactthatunrestrictedlygeneralaccountsofjustificationprovehopelesslyvulnerabletocounter-example.Secondly,littleincurrentliteratureaboutscepticismmakesoneconfidentthatitsnatureisproperlyunderstood.Scepticismdefinesoneofthecentralproblemsinepistemology,namely,theneedtoshowhowjustificationofbeliefispossible.Thisisdonebymeetingthechallengetoshowthatscepticalconsiderationsdonotafteralldefeatourbestepistemicendeavoursinthisorthatspecifiedfield.Implicitinthischaracterizationaretwoimportantclaims:firstly,thatscepticismisbestunderstoodasachallenge,notasaclaimthatwedoorcanknownothing;andsecondly,thatthebestwaytorespondtoscepticismisnotbyattemptingtorefuteitonanargument-by-argumentbasis,butbyshowinghowwecomebyjustificationforwhatwebelieve.Somehowthesetwopoints,whichwereobvioustoourpredecessors,seemtohavebeenlosttosight.FurtherReadingGeneralSomeusefultextsareR.Nozick(1981);J.DancyandE.Sosa(1992);B.Williams(1978);K.Lehrer(1974);L.BonJour(1985);P.F.Strawson(1985);andG.PappasandM.Swain(1978).TheclassictextsinepistemologyincludePlato’sMenoandTheaetetus,Descartes’sMeditations,Locke’sEssayConcerningHumanUnderstanding,Berkeley’sPrinciplesofHumanKnowledgeandThreeDialoguesBetweenHylasandPhilonous,Hume’sTreatiseofHumanNatureandEnquiryConcerningHumanUnderstanding,andKant’sCritiqueofPureReason.OneofthebestshortelementarybooksremainsRussell’sTheProblemsofPhilosophy(1912,muchreprinted).56\nEPISTEMOLOGYKnowledgeThedebateaboutknowledgeandjustificationcommandsalargeliterature,ofwhichthefollow-ingaregoodexamples:W.Alston(1983);L.BonJour(1985);A.Brueckner(1988);R.Chisholm(1977);J.Dancy(1985);F.Dretske(1971);R.Feldman(1985);E.Gettier(1963);A.Goldman(1979,1980,1986);G.Harman(1973,1984);K.Lehrer(1974);P.Moser(1985);R.Nozick(1981);J.Pollock(1979,1984,1986);R.Shope(1983);E.Sosa(1981).ScepticismThebestgeneralintroductiontoscepticalargumentsremainsBertrandRussell’sTheProblemsofPhilosophy,butitisessentialtoseetheargumentsinaclassicsetting,andforthisonemustreadRenéDescartes’sMeditationsonFirstPhilosophy(translatedbyJ.Cottingham,1986),especiallytheFirstMeditation.Usefuldiscussionsofscepticalargumentsandthesense-datumtheoryaretobefoundinA.J.Ayer(1956)andJ.L.Austin(1961),althoughoneshouldalsolookatAyer’sreply(1967).FordiscussionofperceptionseeJ.Dancy(1988);T.Crane(1992);R.Swartz(1965);F.Jackson(1977);andM.Perkins(1983).ForanattemptatbeingscepticalseeP.Unger(1975).ForresponsestoscepticisminfluencedbyKantseeP.F.Strawson(1959,1985)andA.C.Grayling(1985).AlliedlinesofthoughtoccurinG.E.Moore(1959)andL.Wittgenstein(1969).MorerecentdiscussionsareB.Stroud(1984)andM.Williams(1991).Scepticism,foundationalismandcoherencetheoriesofknowledgearediscussedinK.Lehrer(1974)andinusefulpaperscollectedbyG.S.PappasandM.Swain(1978)andM.ClayandK.Lehrer(1989).ForadiscussionoftheviewsvariouslytakenbyDeweyandWittgensteinseeR.Rorty(1979).Atextbookwhichsurveysthefieldandprovidesausefulbib-liographyisJ.Dancy(1985).ForthehistoryofscepticismseeM.Burnyeat(1983)andR.Popkin(1979).ReferencesClassicaltextsBerkeley,G.1995[1710]:ThePrinciplesofHumanKnowledge(editedbyH.Robinson).PublishedwithThreeDialogues.Oxford:OxfordUniversityPress.——1995[1713]:ThreeDialoguesBetweenHylasandPhilonous(editedbyH.Robinson).PublishedwithThePrinciplesofHumanKnowledge.Oxford:OxfordUniversityPress.Descartes,R.1986[1641]:MeditationsonFirstPhilosophy(translatedbyJ.Cottingham).Cambridge:CambridgeUniversityPress.Hume,D.1978[1739–40]:ATreatiseofHumanNature(editedbyP.H.Nidditch).Oxford:ClarendonPress.——1975[1748]:EnquiryConcerningHumanUnderstanding(editedbyP.H.Nidditch).Oxford:ClarendonPress.Kant,I.1929[1781and1787]:CritiqueofPureReason(1stedn1781,2ndedn1787)(trans-latedbyN.K.Smith).London:Macmillan.Locke,J.1975[1690]:AnEssayConcerningHumanUnderstanding(editedbyP.H.Nidditch).Oxford:ClarendonPress.Plato1961:MenoandTheaetetus.InTheCollectedDialoguesofPlato(editedbyE.HamiltonandH.Cairns).Princeton,NJ:PrincetonUniversityPress.OtherwritingsAlston,W.1983:WhatisWrongwithImmediateKnowledge?Synthèse,55.57\nA.C.GRAYLINGAustin,J.L.1961:SenseandSensibilia.Oxford:OxfordUniversityPress.Ayer,A.J.1956:TheProblemofKnowledge.London:Macmillan.——1967:HasAustinRefutedtheSense-DatumTheory?Synthèse,17.BonJour,L.1985:TheStructureofEmpiricalKnowledge.Cambridge,MA:HarvardUniversityPress.Brueckner,A.1988:ProblemswithInternalistCoherentism.PhilosophicalStudies,54.Burnyeat,M.(ed.)1983:TheScepticalTradition.Berkeley:UniversityofCaliforniaPress.Chisholm,R.1977:TheoryofKnowledge.EnglewoodCliffs,NJ:Prentice-Hall.Clay,M.andLehrer,K.(eds)1989:KnowledgeandScepticism.Boulder,CO:WestviewPress.Crane,T.(ed.)1992:TheContentsofExperience.Cambridge:CambridgeUniversityPress.Dancy,J.1985:IntroductiontoContemporaryEpistemology.Oxford:Blackwell.——(ed.)1988:PerceptualKnowledge.Oxford:OxfordUniversityPress.Dancy,J.andSosa,E.(eds)1992:ACompaniontoEpistemology.Oxford:Blackwell.Dretske,F.1971:ConclusiveReasons.AustralasianJournalofPhilosophy,49.Feldman,R.1985:ReliabilityandJustification.TheMonist,68.Gettier,E.1963:IsJustifiedTrueBeliefKnowledge?Analysis,23.Goldman,A.1979:WhatisJustifiedBelief?InG.Pappas(ed.)JustificationandKnowledge.Dordrecht:Reidel.——1980:TheInternalistConceptionofJustification.InMidwestStudiesinPhilosophy5,ed.P.A.Frenchetal.Minneapolis:UniversityofMinnesotaPress.——1986:EpistemologyandCognition.Cambridge,MA:HarvardUniversityPress.Grayling,A.C.1985:TheRefutationofScepticism.London:Duckworth.Harman,G.1973:Thought.Princeton,NJ:PrincetonUniversityPress.——1984:PositiveversusNegativeUndermininginBeliefRevision.Noûs,18.Jackson,F.1977:Perception.Cambridge:CambridgeUniversityPress.Lehrer,K.1974:Knowledge.Oxford:ClarendonPress.Moore,G.E.1959:PhilosophicalPapers.London:AllenandUnwin.Moser,P.1985:EmpiricalJustification.Dordrecht:Reidel.Nozick,R.1981:PhilosophicalExplanations.Cambridge,MA:HarvardUniversityPress.Pappas,G.andSwain,M.(eds)1978:KnowledgeandJustification.Ithaca,NY:CornellUniversityPress.Perkins,M.1983:SensingtheWorld.Indianapolis:HackettPublishing.Pollock,J.1979:APlethoraofEpistemologicalTheories.InG.Pappas(ed.)JustificationandKnowl-edge.Dordrecht:Reidel.——1984:ReliabilityandJustifiedBelief.CanadianJournalofPhilosophy,14.——1986:ContemporaryTheoriesofKnowledge.Totowa,NJ:RowmanandLittlefield.Popkin,R.1979:TheHistoryofScepticism.Berkeley:UniversityofCaliforniaPress.Rorty,R.1979:PhilosophyandtheMirrorofNature.Princeton,NJ:PrincetonUniversityPress.Russell,B.1910:PhilosophicalEssays.NewYork:Longman,Green.——1912:TheProblemsofPhilosophy.London:OxfordUniversityPress.Shope,R.1983:TheAnalysisofKnowing.Princeton,NJ:PrincetonUniversityPress.Sosa,E.1981:TheRaftandthePyramid:CoherenceversusFoundationsintheTheoryofKnowl-edge.InMidwestStudiesinPhilosophy5,ed.P.A.Frenchetal.Minneapolis:UniversityofMinnesotaPress.Strawson,P.F.1959:Individuals.London:Methuen.——1985:ScepticismandNaturalism:SomeVarieties.London:Methuen.Stroud,B.1984:TheSignificanceofPhilosophicalScepticism.Oxford:ClarendonPress.Swartz,R.(ed.)1965:Perceiving,SensingandKnowing.Berkeley:UniversityofCaliforniaPress.58\nEPISTEMOLOGYUnger,P.1975:Ignorance.Oxford:ClarendonPress.Williams,B.1978:Descartes:TheProjectofPureEnquiry.Harmondsworth:PenguinBooks.Williams,M.1991:UnnaturalDoubts.Oxford:Blackwell.Wittgenstein.L.1969:OnCertainty.Oxford:Blackwell.DiscussionQuestions1Howimportantisittohaveadefinitionofknowledge?2Ifapropositionisfalse,canonebejustifiedinbelievingit?3Whydo‘Gettierexamples’raisedifficultiesforafallibilistaccountofknowledge?4Couldstatesotherthanbeliefsprovidethefoundationsofknowledge?5Iftherearefoundationalbeliefs,howaretheyrelatedtodependentbeliefs?6Isabeliefjustifiedifitcohereswithanalreadyacceptedsetofbeliefs?Dowehaveanadequateaccountofthenotionofcoherence?7Doesjustificationconsistininternalrelationsamongbeliefs?8Canabeliefbejustifiedforsomeonewhodoesnothavecognitiveaccesstowhatjustifiesthebelief?9Isknowledge‘appropriatelycausedtruebelief’?10Doesthenotionof‘tracking’helptoexplainhowbeliefsarejustified?11Ifthereisnocertainty,cantherebeknowledge?12Shouldweseekasingleaccountofjustifiedbeliefordifferentaccountstailoredtothedifferentareasinwhichepistemologicalquestionsmaybeasked?13Whatroledoesscepticismplayinphilosophy?14Howcanwedisentanglethosemethodsofenquirythatmightyieldgenuineknowledgefromthosethatcanonlydeepenourignorance?15Ourknowledgeclaimsaresometimesinerrorwithoutourknowingit.Doesthisunderminejustificationforanysuchclaims?16Couldyoubedreamingnow?17CouldtheexperiencesItaketobeoftheworldhavesomeotherorigin,withoutmybeingabletotellthatthisisso?18Canthepropertiesbywhichwedetectthepresenceofobjectsbedescribedapartfromtheirbeingobjectsofperception?19Arewelikethemaninavisorlesshelmetwhocannotchecktheinformationtrans-mittedtohimtoseewhetheritfaithfullyrepresentstheexternalworld?20Canscepticalargumentsthataresinglyimplausiblejointlyrequireaseriousresponse?21Whatistheimportanceofdistinguishingmethodologicalandproblematicscepticism?22Canepistemologyprovideawayofknowingfromafirst-personsubjectiveviewpoint?Isthereanyotherviewpointavailable?23Ifscepticismopensagapbetweenthegroundsforaknowledgeclaimandtheclaimitself,isitbettertobridgethegaportocloseit?24Couldthepossibilityofourhavingexperiencebeunintelligibletousunlessweheldacertainbelief,andyetthatbeliefbefalse?25Areexperienceandrealitythesamething?26Canweaccepttheroleofcounterfactualsinaphenomenalistaccountofphysi-calobjects?59\nA.C.GRAYLING27Whatfollowsfrombeginningouraccountofknowledgewiththepublicworldratherthanwiththeprivatedataofindividualconsciousness?28Doesthe‘privatelanguageargument’showthattheCartesianprojectisimpossible?29Iftherearepropositionsthataresimplynotopentodoubt,howcanweidentifythem?Howaretheyrelatedtopropositionsthatwecandoubt?60\n2MetaphysicsSIMONBLACKBURNWithasectiononTimebyRobinLePoidevinMetaphysicsistheexplorationofthemostgeneralfeaturesoftheworld.Weconceiveoftheworldaboutusinvarioushighlygeneralways.Itisorderly,andstructuredinspaceandtime;itcontainsmatterandminds,thingsandpropertiesofthings,necessity,events,causation,creation,change,values,factsandstatesofaffairs.Metaphysicsseekstounderstandthesefeaturesoftheworldbetter.Itaimsatalarge-scaleinvestigationofthewaythingshangtogether.Withinthisbroaddescriptiontherearetwoconceptionsofthesubject.Metaphysiciansmaythinkofthemselvesasinvestigatingthefacts,ordiscoveringthebroadstructuresofreality.Or,theymayseetheenterpriseasmoreself-reflective,gaininganunderstandingofhowwerepresentthefactstoourselves:howour‘conceptualscheme’orperhapsanypossibleconceptualscheme,structuresourownthoughtaboutreality.Oncethisdescriptioniscompleted,itmaybethateverythingpos-siblehasbeendone,forwehavenoalternativebuttocontinuetothinkfromwithintheconceptualschemewhosefeatureswehavemapped.1MetaphysicsanditsDoubtersAninfluentialdistinctionintheself-imageofmetaphysicsisduetoStrawson(1959).Metaphysicsmaybeapurelydescriptiveenterprise.Or,itmaybethatthereisreasonforrevision;thewayswethinkaboutthingsdonothangtogether,andsomecategoriesaremoretrustworthythanothers.Revisionarymetaphysicsthenseekstochangeourwaysofthoughtindirectionsitfindsnecessary.Thedistinctionbetweenrevisionaryanddescriptivemetaphysicsisnotsharp,foritisoutofthedescriptionsthattheneedforrevisionallegedlyarises,andinfactthemetaphysiciansStrawsoncitesasrevi-sionary–PLATO(c.427–347)(chapter23),DESCARTES(1596–1650)(chapter26),BERKELEY(1685–1753)(chapter30)–believedthemselvestobediscoveringnomorethanthingsthatareimplicationsofournecessarywaysofthinking.Revisionarymeta-physicsisfrequentlyassociatedwithambitious,andsometimeswild,philosophicalspeculation;descriptivemetaphysicsisintellectuallymoreconservative.Butmetaphysics\nSIMONBLACKBURNcharacteristicallyhasapracticaldimension;sincePlatoithasbeenconcernedtopromoteviewsaboutthenatureofhumanbeings,theirrelationswithnature,orGOD(chapter15),orwiththelargerSOCIALWHOLES(p.383)ofwhichtheyareapart.Hereitmaybethatdescriptivemetaphysicsisethicallyandpoliticallymoreradical,forapre-ferreddescriptionofhumanlifemayhavefar-reachingimplications,andindeeditisnotpossibleforanyseriousETHICS(chapter6)orPOLITICALPHILOSOPHY(chapter8)tobeentirelysilentaboutmetaphysics.Andradicalviewsmayhaveconservativeimpli-cations,aswhenpostmodernistviewsoftherelativeandcontingentnatureofallourjudgementsleadtofailuresofconviction,andtheretreatoftheintellectualintosocialandpoliticalinertia(Rorty1989).SometimesthesethemeshavebeenpresentedasgeneraltopicsofBeing,andsomeauthorsgivetheimpressionthatmetaphysicsisthescienceofBeing.Thisis,however,misleading,fortheremaybenothingorlittletobesaidaboutBeingassuch,evenifthereisalottosayaboutthekindsofthingsthatexist,andthecategoriesunderwhichtheyfall.AndinfactthestudyofBeingrapidlyturnsintothestudyofthingsinthesecategories,andoftherelationshipwiththemthatcharacterizesparticularthinkersatparticulartimes.ButwhatisrightintheideathatmetaphysicsisthescienceofBeingisthatthemostabstractstudyinthisabstractdisciplineconcernsthebroadnatureofreality,andthepossibilityofitsobjectiverepresentation.Whatmotivatesthestudyofmetaphysics?Centrally,problemsarisewhenwecannotseehowthingshangtogether.Wesufferfromdisquietwiththepluralityofdifferentkindsofthingsthatexist,orwiththemixtureofelementsthatwewanttokeeptogether.Paradigmaticmetaphysicalproblemsarisewhenbroadareasofourcommitmentsclash,asforinstancewhenwethinkofourselvesascomplexnaturalorganismsontheonehand,orasconsciouspurposiveorevenfreeagentsontheother,orwhenwethinkofTIME(pp.82–5)asflowingontheonehand,butrecognizethatitmakesnosensetoaskhowfastitflows,ontheother.Considerthelistofminds,physicalthings,abstractobjects,values,events,processes,dispositions,necessities,statesofaffairs,properties,factsandotherbasiccategoriesthatweseemtorecognize.Weneedastoryabouthowtheyrelatetoeachother:aresomeofthemequivalenttoothers?Aresomemadeupofothers?Aresomeofthemredundant,orotherssuspiciousinotherways?Andjustasnotably,weneedastoryabouthowwerelatetothem.Howdoweknowaboutthem,andwhyaretheyimportanttous?Suchvaguedisquietistypicallyfocusedbythedis-coveryofparadoxandinconsistency,apparentlyshowingthatourconceptionsareinadequate.AmongtheearliestexamplesofphilosophicalargumentintheWesterntraditionaretheparadoxesofmotionofZenoofElea(fifthcenturyBC),purportingtoshowthateverydaythoughtabouttheflowoftimeandmotioninspaceinvolvescontradiction.Whatweallnaturallybelievecannotpossiblybetrue.ThemoralZenowantedustodrawfromthisdiscoveryisnotquiteclear,buttheparadoxserveduntilthenineteenthcenturyasaspurtophilosophersandmathematicianstofindabetterwayofconceptualizingchangeintime.Itseemsthatsuchtopicsshouldpermitofsomedisciplined,intelligent,investigation.Butitiscontroversialhowsuchaninvestigationshouldbeconducted.Ifmetaphysicspurportstotelluswhatthingsexist,orevenjusthowtothinkaboutwhatthingsexist,thenitseemstobetrespassingonthedomainofthePHYSICALSCIENCES(chapter9),anditisunclearhowaphilosophicalstudycanhaveanythingsignificanttoaddtotheir62\nMETAPHYSICSresults.Thequestioniswhetherthereisanythingleftforphilosophytodo,oncefun-damentalsciencehastolduswhatitenablesustoknowaboutthesetopics.Theviewthatthereissomethingfurther,andthatbysomeprocessofrationalthoughtwecanobtainreliableviewsaboutthenatureofspace,time,mind,causationandtherestisoftendescribedasbeliefina‘firstphilosophy’,orphilosophyasadisciplinewithitsownmethodsandresults,andtheverypossibilityofsuchafirstphilosophyhasbeenacon-stantobjectofdoubt.Metaphysicsisthuspreoccupiedwithitsownpossibility.Itisnotuniqueamongphilosophicaltopicsinthis.Butwhereaswebelieveinourhearts,asitwere,thattheremustbesuchasubjectasethics,orthattheremustbeinterestingthingstobesaidwithintheTHEORYOFMEANING(seechapter3)orevenmind,thereisnosuchpresumptionthattherecouldbeadisciplinedmethodforachievingresultswhenthetopicsaretheabstractcategoriesIlisted.ConsiderasanexampletheproblemofrelatingMINDANDMATTER(seechapter5).Amongthefirstphilosophicalthoughtsmanypeoplehaveareonesabouttheirownconsciouslives,andthegapthereseemstobebetweenthatconsciousnessandthesimilarlivesofothers,andthemysteryofthatconsciousnessarisingatallintheworldasweotherwiseunderstandittobe.Thefactofconsciousnessseemsundeniable,butitalsoseemsmysterioushowitcanexistinaphysicalworldofthekindtowhichweknowourselvestobelong.Wecanthinkofthisasaproblemofrelatingapuzzlingorexoticfactorkindofentitytoarelativelydomesticorfamiliarclassoffactsorentities.BythisImeanthatwemaybeginbybeinghappywithsomekindsoffact,madetoobtainbysomekindsofthing,suchasthefactsofphysicalnature,orfactsaboutthepatternsintowhicheventsfall.Wefeelweunderstandfactsofthissort.Weunderstandwhatmakesthemobtain,andhowweknowthemtobetrue.Thisclassisthatofthefamiliar(homely)aspectsoftheworld:onesthatdonotpuzzleusunduly.Butthenwerealizethatwehaveotherbeliefsaswell;oneswhichonthefaceofitrelatetodifferentkindsoffact,andpositdifferentkindsofthing.Inthiscasemindsareposited,butweshallshortlymeetotherexamples.Wemaynotbeabletounderstandhowmindisevenpossibleinaphysicaluniverse(howmindscanrelatetobodies;howtheycanmakephysicalthingshappen).Thenthemainmotivetometaphysicalinquiryistoputtheexoticclassintosomesortofintelligiblerelationtothefamiliarclass.Anintelligiblerelationwillmeanthatweunderstandhowbothclassesoffactsexist,andhowtheworldisconstitutedsothateachofthemobtains,howoureverydaythoughtsabouttheirrelationshipcanbetrue.Ifwethinkofitthisway,wecanseethreemainpointsatwhichchoiceispossibleinmetaphysics.First,thereisthequestionoftheauthorityofthefamiliarclass.Atapar-ticularperiodoftimesomestartingpointsmayseemnatural.Wemaybeathome,forexample,withtheresultsandconceptsofthephysicalsciences(andinfacttheconvic-tionthattheworldisfundamentallynothingbutaphysicalworldiscertainlythemostcommonviewamongcontemporaryWesternphilosophers).Ifso,theseresultsandconceptsgiveusourclassoffamiliarfacts,andthetaskisthatofrelatingmoreexoticsubjectstothem.Buttootherphilosopherstheremaybesomethingwrong-headedorarbitraryorprejudicedaboutchoosingjustoneprivilegedfamiliarclass.Perhaps,forexample,whenweunderstandhowcomplexandstrangearethefactsandstatesofaffairspositedbycontemporaryphysics,wewillloseanyconfidencethatothersaremoreexoticbycomparison.Evendecidingwhatfactsandstatesofaffairsbelongto63\nSIMONBLACKBURNphysicsoccupiesagreatdealofPHILOSOPHYOFSCIENCE(chapter9):doesphysicaltheory,forexample,postulaterealPROBABILITIESANDCHANCES(pp.308–11),andifsohowarewetothinkoftheirexistence?Whatjustifiesourtendencytotakescienceasanobjec-tivedescriptionofindependentreality,asopposedtoseeingitasanINSTRUMENT(pp.294–5)forpredictionandcontrolofempiricalfact(andhowisthisdistinctiontobedrawn)?Butevenwhenaprivilegedclasshasbeenidentifiedanddoubtsaboutitsnatureandthesourceofitsprivilegesatisfactorilysettled,therewillbeasecondquestionofwhatcountsasasatisfactoryfitbetweentheexoticandthefamiliar.Whatkindsofrela-tionshipshouldputworriesaboutmindorvaluesorabstractobjectstorest,ifwewishtoseeourworldasfundamentallytheworldasdescribedbyphysics?Whataccommo-dationsshouldwebelookingfor?Thirdly,andfinally,therewillbedifferentattitudestotheexoticclass.Ifitseemssufficientlyexotic,sothatitsobjectsandfactsseemonlydimlyrelatedtothoseofthefamiliarclass,aradicalresponsewillbeeliminativism,orthesug-gestionthatitisnotpossiblethatthereshouldexistthingsintheexoticcategories,sothattheentireexoticareaisbestforgotten.Thetaskisnottorelateitscommitmentstothefamiliarclass,buttoteachpeopletodowithoutthem,ratherasatheistsdonotseektounderstandtalkaboutGodinsomepreferredterms,butwantinsteadtoabandonitaltogether.Inthisspiritonemightwanttoeliminatereferencetoabstractobjectssuchaspropertiesorfacts,ortovaluesoreventominds.Putlikethistheenterpriseofmetaphysicssoundseminentlyreasonable,andonecanimaginewell-conducteddisputeinallthreeareas:motivesforselectingafamiliarclass,differingviewsabouttherelationshipitbearstotheexotics,anddifferentattitudestotheexotics.Buttheproblemofamethodforconductingsuchenquirystillremains,andweshouldrecognizethatphilosophyhascontainednearlyasmanythinkerswhocountthemselvesasradicallyopposedtometaphysicsasoneswhoacceptthetitle.Inthemodern(post-seventeenthcentury)WesterntraditionthefirstphilosopherresolutelyhostiletometaphysicswasDavidHUME(1711–76)(chapter31)(Hume1978).Fol-lowinghimtherewerealwaysphilosophers,especiallyinBritain,ofamorecautiousempiricistbent,butfewweretough-mindedenoughtobelievewithHumethatbooksofmetaphysicsshouldbeconsignedtotheflames,ascontainingnothingbutsophistryandillusion.Whetherornotithasamethod,metaphysicsisswayedbylargerwindsthatblowintheprevailingculture.ItiswidelyrealizedthatbeforetheEnlightenmenttheprevail-ingtoneofphilosophywasTHEOLOGICAL(chapters15and24).Itwassupposedthattheworldwastheproductionofanintelligent,rational,caringandperfectbeing,whopos-sessedacompleteunderstandingofitsnature.Weimperfectbeingscoulddosomethingtoapproximatetothisunderstanding,buttheinsightweobtainedinsuchareasasmathematicsandlogicaffordedusonlyglimpsesoftheperfectrationalunderstandingofthewholeorderofnaturethatGodwouldpossess(Craig1987).ThefamiliarclassoffactswithwhichpeoplewererelativelycontentincludedGod’snature,purposes,valuesandrelationshiptocreation.TheabandonmentofthispictureoftheworldbyHumeandotherEnlightenmentfiguresdidnot,however,marktheendofmetaphysics.Onthecontrary,thescientificrevolutionoftheseventeenthcenturyreplacedonekindoffamiliarfactwithothers:notably,thoseofphysicalscienceandthosedescribingthecourseofperceptualexperience.Butthis‘Galileanworldview’,namedafterGalileo(1564–1642),containedtheseedsofitsownmetaphysicalthickets,namelythosesur-64\nMETAPHYSICSroundingtheissueofidealism.Whenmindissharplydistinguishedfrommatter,asitbecameinthisworldview,theuncomfortableandapparentlyimpassablegulfbetweenthemismostnaturallyclosedeitherbymakingthemindmaterial,orbymakingthematerialmental.Idealismistheumbrellatitleforphilosophiesthattakethelattercourse.Fortheidealist,factsaboutmentallifeformthefamiliarclass,andonesaboutthephysicalworldneedsomekindofcertificationfromthem.Anidealist,therefore,wouldbereceptivetothekindofinstrumentalinterpretationofphysicalsciencemen-tionedabove,inwhichscientificdescriptionsaretakenasinstrumentsforpredictionandcontrolofthecourseofperceptualexperience.Whicheverpriorityisassumed,theproblemofprovidingthiscertificationdominatedphilosophyfromthetimeofDescartesuntilthepresent.Indeed,perhapsthehigh-watermarkofmetaphysicalspeculationwasreachedwiththeattemptstomarryidealismwithareligiousworldviewinthenineteenthcentury.Accordingtotheacceptedstory,undertheinfluenceofKANT(1724–1804)(chapter32),butforgettingthetightlycriticalboundariesthatKanthimselfputuponmetaphysics,itwasbelievedthattranscendentresults–whichgobeyondthelimitsofexperienceandconcernGod,freedom,immortalityandabovealltheultimatespiritualnatureoftheuniverse–couldbeestablishedbyavarietyofaprioriarguments.Kant’smistrustoftranscendentreasoningwasbrushedasidebyFichte(1762–1814)andmostinfluentiallybyHEGEL(1770–1831)(chapter33),andthemetaphysicianbecamethespecialistinthenatureofTHEABSOLUTE(p.743),orunderlyinggroundofthecosmicorder,whichwasidentifiedvariouslywithGod,pureFreedom,orfinalSelf-consciousness,andwhichprovidedsomesortofgoaltothecumulativehistoricalprocess.Analmostreligiousbeliefinprogress,coupledwithagenuinelyreligiousemphasisonthenatureoftheSPIRIT(p.742),andaboveallitselevatedmoraltone,gaveabsoluteidealismitsholdonthemindsofEurope.Thedominantcontemporaryspiritreversesthisdirection,privilegingfactsaboutthephysicalandseekingtounderstandstatementsaboutmindandconsciousnessinitsterms.Thisisknownasphysicalism,orlessoftenmaterialism(thewordphysicalismispreferredbecausephysicsitselfassertsthatnoteverythingthatexistsismaterial;theworldincludessuchitemsasforcesandfields).Physicalismandidealismsharethegoalofrelatingmindandmatterinsomeintelligibleway,butdifferoverwhatisfamiliarandwhatexotic.Theneedforadditionaltasksofthesamekindmaybelessobviousinadvanceofcriticalreflection.Thepointtorememberisthatwhateveraphilosophermightputforwardbywayofdoubtsaboutthepossibilityofmetaphysics,thephilo-sophicalneedforatheoryrelatingcentralkindsoffactmaybedrivenbywideraspectsoftheworld-picturecharacteristicofagiventime.Humewasanopponentofmetaphysicseventhoughhebequeathedtohissucces-sorstheparadigmmetaphysicalproblemsofmodernphilosophy.Hisownattitudewassimplythatsuchproblemswereforeverinsoluble.Thedifficultyisthatitishardtobelievethatmetaphysicalquestionsareunanswerableuntilwetrytoanswerthem.Perhapstheproblemsseemformidablebecauseweareprejudicedinourchoiceofafamiliarclass(Humehimselffavouredfactsaboutthesuccessionsofourownideasor‘perceptions’asespeciallybasicorfamiliar,achoicethatfrequentlyamazescontempo-rarythinkers);perhapsweareblindtosubtleandinsightfulwaysofrelatingtheexoticstoit;perhapswedonotrealizethatweoughttogetridofexoticsthatrefusetobeaccommodated.Itisdifficulttobelievethattheentirefieldcanbeseen,inadvanceand65\nSIMONBLACKBURNfromthearmchair,tobeano-goarea.Buttherehasbeenoneschoolthiscenturywhichbelievedexactlythat.PerhapstheonethingthatisbestknownabouttheschoolknownastheViennaCircle,orthelogicalpositivists,wastheirresolutehostilitytometaphysics.Itisinstruc-tivetoseehowthishostilityflourished,buttheneventuallycrumbled.ThepositivistswereinfluencedbytheempiricismofHume,andalsobytheTractatusLogico–PhilosophicusofWITTGENSTEIN(1889–1951)(chapter39)(Wittgenstein1922),aworkwhichpurportstosetboundstothelimitsofmeaningfullanguage.Metaphysicsliesoutsidethebounds,representinglanguagethathasgoneonholiday.Thepositivistsheldthatalltruethoughtwasempirical;anythingstrayingbeyondtheuseofscientificmethodalsostrayedbeyondtheboundariesofmeaning.Butmetaphysicalthesesarenotproperlyempirical;theyseemedtypicallytobearguedaboutfromthearmchairandtobearnovisiblerelationshiptoscientificthoughtorexperiment.Hencetheydonotadmitofverificationorfalsification,andarenotonlyunscientific,butalsostrictlymeaningless(theymaybeallowedsomekindofemotionaleffect,butnothingthatpermitsofassessmentastrueorfalse).Insteadofmetaphysics,allthatwaslefttophi-losophywasdescribingthecorrectmethodsandstructuresofempiricalscience.Positivismfailedinitscrusadeagainstmetaphysics,largelybecauseofitsowninsta-bility.Forwhileinonebreathitwasproclaimingthesubjectdead,inanotheritwasitselfmakingremarkablyconfidentchoicesinallthreeoftheareasinwhichmeta-physicscharacteristicallymakeschoices.Positivismneededafamiliarclassofempiri-callysoundandbasicjudgements.Itneededviewsaboutwhatcountsasasatisfactoryaccommodationwiththatclass,notablysothatthetheoriesofphysicalsciencewereallowedtoberespectable,whereasotherswerenot.And,finally,ithadsevereviewsabouttheareasthatwerenot,bytheselights,respectable.Metaphysicssoonrevengeditself,forallthreeareasgavetrouble,andinallthreethemovementfounditselfpulledindifferentdirections,makingdifferentchoicesandindulginginclassicallymetaphysi-caldispute.Awareofthis,laterpositivistwritingsgrudginglyallowedaplaceforwhatlookedlikemetaphysicalassertionsasrecommendationsaboutwhichLANGUAGEGAMES(p.9)toplay.Thus,thedoctrinethatthereexistsmindaswellasmatterwouldbecon-struednotasanimportantdescriptionofafundamentalfactaboutreality,butasarec-ommendationtospeakintermsofmindsaswellasintermsofmatter(toplaythementallanguagegame).Therecommendationmightbeusefulornot,butcouldnotberegardedastrueorfalse.AsimilarattitudeischaracteristicofthelaterworkofWittgenstein(1953),whichisshotthroughwithwarningthatstatementsthatmightseemtobecertainkindsofdescriptionofrealityinfactfunctionindifferentways.Wittgenstein’sexactintentionsareendlesslydisputed.But,atleastasappliedtomindsandmatter,anypragmatismintheapproachisitselfacutelyuncomfortable.Consider,forinstance,thesimplepropositionthatotherpeoplebesidesmyselfareconsciousinthewaythatIam.Thisseemstobeametaphysicalthesis,althoughclassifyingitasonemayitselfrepresentaphilosophicalchoice.Itiscertainlyonethatweallbelieve.Butitisextremelystrangetosupposethatthequestionwhetherotherpeopleareconsciouscouldbeconstruedasthequestionofwhetheritwasusefulformetoenterthelan-guagegameofreactingtothemashavingminds(thisglossonthequestionhadinfactbeenputbyWilliamJAMES(1842–1910)(chapter36)muchearlier,especiallyinPrag-matism,andhasnotreceivedagenerallywarmwelcome).Weshouldalsonoticethat66\nMETAPHYSICSevenifweallowtheideathatwhatsoundlikedescriptionsoffundamentalrealityareconstruedasrecommendationsaboutlanguage,thereisstillargumentaboutwhattoselectasfamiliar,whatisexotic,whatshouldbeeliminatedandwhattherelationshipsarebetweenalltheseclasses.Soeventhepragmatistglossmaymakelessdifferencethanmightappearatfirstsight.AbsolutePresuppositionsTheirownmetaphysicalprejudiceswereinvisibletothepositivists.Thethoughtthatthisisaninevitablefeatureofthephilosophicalconditionisanother,different,reasonfordespairaboutthepossibilityofmetaphysics.ThisreasonwasvoicedmostnotablybyR.G.COLLINGWOOD(1889–1943)(pp.436–9)(especiallyinAnEssayinMetaphysics,1940),andmarkedapointatwhichhisthoughtcomessurprisinglyclosetothatofboththeearlierandlaterphilosophyofWittgenstein.Inthisconception,therearecommitmentsthatwemusthavethatarenotcapableofbeingassessedastrueorfalse,butthatnever-thelesshavethegreatestintellectualimportance.InWittgenstein’smetaphors,theyfunc-tionastheriverbedwithinwhichthoughtflows,orasthehingesonwhichordinaryjudgementanddiscourseturn;Collingwoodcalledthemtheabsolutepresuppositionsofthethoughtofatime.Becausetheyarepresupposedineveryactivityofthought,theycannotthemselvesbeassessedfortruth:inanothermetaphor,theycanbeshownbutnotsaid.ForCollingwood,indeed,theycouldonlybeshownhistorically.Thatis,attheendofanepochitwouldbepossibleforsuccessorstolookbackandtofindthataparticularmeta-physicshadstructureditsthought.Butatthetimethebasicstructureswouldthemselvesbeinvisible,becausetheywerethemselvesinvolvedinallseeing(Hegelhimselfmayhavethoughtsomethingsimilar;thefamousremarkfromtheprefacetohisPhilosophyofRight(Hegel1991)that‘whenphilosophypaintsitsgreyongrey,thenhasaformoflifegrownold.TheowlofMinervatakeswingonlywiththecomingofthedusk’canbeseenastheclaimthatphilosophicalreflectioncanonlyexistwithhindsight).Ifthisisright,thetrapthatcaughtthepositivistsisthereforeuniversal;howevermuchweaspiretocareandobjectivityinchoosingourfamiliarfacts,andawayofrelatingexoticstothem,wewillbeworkingwithinanhistoricallycontingentframeworkwhosemainstructureswillbebeyondourownvision,andincapableofourownassessment.Collingwoodwasanhisto-rian,andotherepisodesinthehistoryofphilosophysupporthisview,foritiseasiertoseeabsolutepresuppositionsatworkinthethoughtofparticularwriters(andtheircontem-poraries)whensignificanttimehaspassed,andeasytofearthatoursuccessorswillmarvelatsimilarblindnessinourownappreciationofthewaywethink.AnexampleIhavementionedalreadyisHume’schoiceofthesequenceofperceptionsasthebasicfamiliarfact:achoicethatseemsextraordinarilywrong-headedtonearlyallcontempo-raryphilosophers.Oncemore,however,manyphilosophersarereluctanttoadmitthereshouldexistanyaspectsofourownthoughtthatcannotthemselvesbeidentifiedandevencriticallyassessed.Whilesuchepisodesinthehistoryofphilosophymaynourishourmodesty,thereseemstobenoalternativetocontinuingto‘workfromwithin’,orinotherwordstodothebestwecanfromwithinourownbestunderstandings.Thethoughtthatonedaythesemaycometolooklocalandparochialislikethesimilarthought,inmoralphilosophy,thattheremightonedaybeapeopletowhomourownbestjudgementsseemedinadequate.Perhapstheremight,butwhilewecannotseeourwaytothisimprovement,wecan,andmust,simplysoldieronasbestwecan.67\nSIMONBLACKBURNTheresurfaceshereadebatethathaspreoccupiedmanyrecentwriters.Thisisthedisputebetweenrealismanditsopponents.Fortherealistitisimportantthatthereisnoresidualreferencetous(ourlanguage,oursensibilities,ourconceptualscheme)whenweconsidertheworld.Therealistwantsrealobjectivity–aworldoffactsthatarefrequentlyentirelyindependentofusandthatwouldbeastheyarewhateverourpowersofdetectingandexploringthem.Therealististhereforefundamentallyopposedtotheviewthatwhatweunderstandandinvestigateisasmuchafunctionofourconstitutionandourwaysofthoughtasitisofanythingindependentofus.Inacommonmetaphor,therealistbelievesthatagoodconceptualscheme‘carvesrealityatthejoints’;natureisconceivedaspossessingitsownstructureandarticulation,andgoodtheoryonlyreflectsthis.Thiscarnivorousmetaphoroutragesidealistsofmanyvarieties,andevenphilosopherswhowouldscarcelycountthemselvesasidealists.Thecentralproblemisthatbeliefthatwesucceedincarvingnatureinsomeespeciallyappropriatewayseemstorequireawayofcomparingourownclassificationsandthekindsofconceptweuse,withapre-existent,naturallyordainedstructureofproperties.Butnosuchcomparisoncanbemade,norisitplainwhat,apartfromordinaryscientificutility,couldjustifyusinsupposingthatsomepropertiesareintrinsicallynaturalwhereasothersarenot.ThisproblemreceivedforcefulexpressioninGoodman(1955);foranexampleofindustrial-strengthrealism,seeLewis(1983);formistrustofthemetaphorseeTaylor(1993).Thechoicebetweenrealismanditsopponentshasechoesinmanyareas.Forexample,therealistisparticularlyapttoprivilegesomefamiliarclass,usuallythatofphysicaltheory,asbeingespeciallywell-adaptedtonature’sjoints.Bycomparisonothercommitmentsmayseemtohavelesstodowiththerealwaythatthingsare.Forinstance,thesizesandshapesofthingsaroundusmayseemtobemoreobjective,moreindependentofus,thantheircolours,whichseemtobelargelyafunctionofthenatureofourvisualsystems.Idealists,impressedbythemind’scontributiontoanyschemeofthought,includingthatofphysics,aretypicallylesspartisan.Attheirmosttolerant,theymay,aspluralists,becomehappytocountenancealmostalllanguagegamesorconceptualschemesasalikereflectingtheparticularperspectiveofsomeuser;noschemeisprivilegedbyhavingauniqueandspecialrelationshipwiththewaythingsare,andallarejustifiedinsofarastheyembodyaformoflifeorwayofreactingtotheworldandcopingwithit.PluralismheremakescontactwithPRAGMATISM(chapter36),ortheviewthatwhatultimatelyjustifiesanymodeofdescriptionoftheworldisitsutilityinenablingustocopewithourproblems.Forthepluralisttheremaybenourgenttaskofrelatinganythingfamiliartoanythingexotic,forbothalikeareconceivedofasbeingnomorethanonemodeofdescriptionamongothers.Pluralism,therefore,servesasyetanotheravenuefromwhichmetaphysicscomestolooktobeachimera–anddislikeofold-fashionedmetaphysicsonthisgroundisoneofthecharacteristicsofthepostmodernistmistrustoftraditionalphilosophy.But,oncemore,itisnotsoobviousthatonecanbesatisfiedwithpluralismwithoutalreadyhavingdoneenoughmetaphysicstogainthenecessarilyrelaxedattitudetothedifferentcategoriesofbeingandtheirrelationships.68\nMETAPHYSICS2AnalysisandLogicMetaphysicsneedsamethod,butwhatmethodcantherebe?ItisgenerallysaidthatthemostinfluentialandrespectedAnglo-Americanthoughtthiscenturymarksacom-pletechangeofdirectionfromtheambitiousandspeculativemetaphysicalsystem-buildingcharacteristicofthepreviousperiod.ForthefirsthalfofthiscenturyinAnglo-Americanphilosophy,andforsometimeafterwardsinsomeplaces,theproblemofmethodformetaphysicswasgovernedbytheidealofanalysis.Facedwithameta-physicalproblemsuchasthenatureofmind,thephilosopherwouldassemblethecentraltermswithwhichwetalkaboutmind:thought,sensation,willorwhatever;carefulattentiontowhatismeantbythesetermswouldrevealthewayweconceptu-alizethenatureofmind.Themodelforsuchaprocedurewouldbethatoftheana-lyticalchemist,discoveringthenatureofasubstancebybreakingitintocomponents.Thenegativepartofthisaimisclearenough.Whenmeaningsareobscureandunclear,theinferencesthatwearepermittedtomakeareuncertaintoo,anditisthisuncer-taintythatallowsforfancifulandmonstroussystem-building.Butifmeaningwerecor-rectlylocated,inwaysfirstmadepossiblebythenewlogicaltoolsdevelopedbyFREGE(1848–1925)andRUSSELL(1872–1970)(chapter37),thenthoseinferenceswouldbesystematizedandestablished,andcorrectmethodsfinallydistinguishedfromimpos-tors.Analytictechniquewaspartlyimportantpurelyasadefenceagainstwildtheo-rizing.Inthissensethemethodofanalysiswasnotnew,butonlyanewlabelforproceduresthatareasoldasphilosophy.Inmanyofhisdialogues,SOCRATES(chapters22and23)challengeshisaudiencetostateexactlywhattheymeanbyadisputedterm;MEDIEVALPHILOSOPHY(chapter24),asmuchaslaterempiricism,isdominatedbytheaimofpreciseclarificationsofmeaning.Butanumberofconsiderationschangedtheimpactofthisworkofclarificationonmetaphysics.Onewastheexpansionoflogic,suggestingawhollynewrangeforanalytictechniques.Farthemostinfluentialexampleofthisexpansioninactionwasthetheoryofdescriptions,revealedbyRussell(1905);a‘paradigmofphilosophy’wherebyjudgementsthatweseemtobemakingaboutnon-existent‘things’arerevealedtohavequiteadifferentLOGICALFORM(p.790),andtomakenosuchcommitments.Analysisalsoprovidedtheprincipalgoalforthesatisfactoryaccommodationbetweenthefamiliarandtheexotic.ThisisthegoalofREDUCTION(p.312),inwhichtherightattentiontomeaningrevealsthatwhatseemtobeclaimsaboutapuzzlingandexoticareaareinfactclaimsofafamiliar,homelykind.Specificproblemsaboutexoticfactsandthingsdonotarise,fortheyareshowntobeordinary,familiarfactsandthings.Forexample,ifclaimsaboutbehaviourarethoughttoberelativelyintelligible,butclaimsaboutmindsseembycontrastmysterious,thenthesolutionmightbetoanalyseclaimsaboutmindsasdisguisedclaimsaboutbehaviour,ordispositionstobehaviour.Iflogicaltruthorlogicalproofarefelttobeintelligible,butclaimsaboutnumbersmysteriouslyabstractanddangerouslynon-empirical,thesolutionwouldbetoanalyseclaimsaboutnumbersasdisguisedrecipesforpurelylogicalinferences.Theprogrammesthisapproachgaveriseto–thetwoexamplesgivenareBEHAVIOURISM(pp.174–6)andLOGICISM(pp.790–1),butthereweremanymore–notonlydominatedmuch69\nSIMONBLACKBURNphilosophicalthinkinginthefirsthalfofthetwentiethcentury,butgoondirectingalargeproportionofcurrentresearcheffort.Ifitsucceeds,analysisprovidedanextraordinarilyelegantandeconomicalanswertotheproblemoftherelationshipofthefamiliartotheexotic.Forifbyanalysiswelaybarethereal‘logicalform’orstructureofourthought,anditturnsoutthatthefactsandthingsweneedtorefertoarenotwhattheyseemtobeatfirstglance,thentheproblemissolvedbyabsorptionoftheexoticintothefamiliar–thefinalandsimplestkindofsolutiontotheproblemoftheirrelationship.Weareshowntohavenocommitmentsbeyondthefamiliarandtheunpuzzling.Itisimportanttonoticethatanalyticalphilosophyarosenotprimarilyasacrusadeagainstmetaphysics,butonlyagainsttheundisciplinedperversionsofmetaphysics,whichseemedtothemajorearlyanalyststosurroundthem.ItisinfactnotablethatthepioneeringworksofMOORE(1873–1958)(chapter38)andRussellarethemselvespreoccupiedwithquestionsofwhattherereallyis.Thefamiliarclass,forFrege,MooreandRussell,includedsomebeliefsthatothersmightregardwithsuspicion,mostnotablybeliefinabstractobjects.Moorehimselfatonetimeheldthedistinctlymeta-physical-soundingthesisthatallthatreallyexistsarepropositions;Fregeremainedweddedtobutpuzzledbythedistinctivelymetaphysicalbeliefintheexistenceofabstractobjects.Russell’sempiricismledtoflirtationsbothwithdoctrinesnotunlikethoseoftheidealistBerkeley,andwithneutralmonism(thebeliefthatthesameprimi-tive‘stuff’,orderedindifferentways,makesupontheonehandthementalandontheotherhandthephysicalworld).Whatdroveanalyticalphilosophywasnotoriginallyhostilitytometaphysics,butbeliefthatthecorrectmethodforpursuingithadfinallybeenfound.Oneobjectiontothemethodusedtobemadebymorespeculativeorambitiousphilosophers,writingundertitleslikeClarityisnotEnough.WithconcernsechoingthosewhichfirstaroseinANCIENTGREEKPHILOSOPHY(chapters22and23),theyworriedthatthetruenatureofmindorothermetaphysicaltopicsmightnotberevealedbyanalysis,becausewhatisanalysedisacompendiumofcommon-senseprejudiceorfolk-lorewhichisnotparticularlylikelytoenshrinethemetaphysicaltruth.Thisworryhasbeenrevisedinadifferent,scientific,toneofvoicemorerecently.Sincecommon-senseconceptswereformedinpre-scientificdays,thereisnoparticularreasontorespectthemasasourceofmetaphysicaloranyotherkindoftruth.Indeed,somephilosophersthinktherightreactiontosuchconceptsistheresponsewementionedabove,namelyeliminativism,whichisthedoctrinethateverydaythoughtaboutsometopicissuffi-cientlyinfectedwitherrorsforitscategoriestobewhollyunreliable.Everydayopinionsaboutthenatureofmindinprincipledeservenomorerespectthaneverydayopinionsaboutthenatureofmassortheflightofprojectiles,andconceptswithinwhichevery-daythoughtframesitsopinionsmaybequiteworthless,likethoseofAristotelianphysics.Thoughtcouchedinthesetermsdoesnotneedanalysis,butelimination(Churchland1989;Stich1983).Thisiscertainlypossible.Buttoturnthepossibilityintoareallikelihood,wemustatleastcorrectlyidentifywhatthecommon-senseschemeactuallyrequires.Whatiscommonsensecommittedto,andwhichpartsofitmayneedeliminating?Herethereisaneedforworkatleastsignificantlylikethatoftheanalyticalphilosopher:wewouldneedtoknowexactlywhatthecommitmentsimplicitinthecommon-senseschemeare,andthismeanscorrectlylocatingwhatismeantbythesalientterms.70\nMETAPHYSICSLetmegiveasimpleexample.Supposesomeoneremarksthatonedoctrineofevery-daycommonsenseisthatpeopleoftenbehaveastheydobecauseofwhattheythink.Supposeitisthenpointedoutthatsciencetellsusthatphysicsisaclosedsystem:physi-calevents,suchasbodiesmoving,musthavephysicalcauses.But,goestheobjection,thedescriptionofsomeoneasthinkingsomethingisnotaphysicaldescriptionofthem.Hence,itmustbeanerrortobelievethatpeoplebehaveastheydobecauseofwhattheythink.Thisisasimplifiedversionofanargumentthathasobviouseliminativistleanings,foroncepeopleacceptthatitiswrongtothinkofpeopleasbehavingastheydobecauseofwhattheythink,thentheideaofpeopleasthinkersatallisalsofairlyimmediatelythreatened.Forifthoughtdoesnotcausebehaviour,itdoesnotseemtodoanythingelseeither.Buttheargumentisscarcelycompellingasitstands.Itraisesahostofquestions,andtheyareanalyticintheirnature.Isthe‘because’in‘peoplebehaveastheydobecauseofwhattheythink’actuallyacausalnotion?Ifitis,doesthiskindofcausalexplanationcompetewithother(neurophysiological)causalexpla-nations,orisitcompatiblewiththem?Mightitmakesensetosuggestthatmentaleventsareidenticalwithphysicalevents,andarethereforeingoodstandingasphysi-calcausesofthings?Toassesstheargumentwemustfirstclearupsuchquestions,butclearingupsuchquestionsisinvestigatingtheverytopicsthattheanalyticalphiloso-phertookasprimary.Ireturnbelowtodiscussingsomeofthemovesthathaveattractedattentioninconnectionwiththesetopics.TheFlightfromAnalysisEvenifeliminativismcannotentirelydowithoutanalysis,inthelastfortyyears,andespeciallyinthelasttwenty,analysishasofficiallylostmuchofitslustreasanideal.Partlythefailureofpositivismtaughtphilosophersthattheorycantakeusbeyondevi-dence–thatconceptsthatmaynothavemetstrictempiricistorpositiviststandardsformeaningareneverthelessperfectlyproper.Totheorizeproperlyonthebasisofdataistoinventnewconceptualstructures.Butthen,thekindsofthingsaidwhenwemakeuseofthosestructureswillnotbeequivalentinmeaningtothethingssaidwithoutthem.Newtheoreticalconceptshavetheirownmeanings,anditisfutileconservatismtodenythisbytryingtoanalysetheircontentinoldterms.Theflightfromanalysiswasalsopro-pelledfromotherdirections.Onewasthefailureofmanyofthereductionistpro-grammes,fewofwhichgaveconvincingreasonsforsupposingthattheexoticwasjustthefamiliarindisguise.Anotherwasagradualdisenchantmentwiththefoundationalroleattributedtomeaningintheanalyticparadigm.Quine(1951)convincedmanythattheequationsofmeaningdemandedbyanalysiscouldnotbeself-sufficient‘semantic’factsknownmerelybyknowingthelanguagesconcerned,butwouldthemselvesequallyrepresentdeeptheoreticalandscientificchoices.Forexample,ifstatementsaboutphysi-calobjectsareanalysedintostatementsaboutcoursesofexperience(phenomenalism)thiswillnotbetheresultofaneutralsemanticequation,butwillrepresentatheoretical(metaphysical)convictionaboutwhatthereisandthewayitistobeunderstood.Inotherwords,semanticsisdrivenbyscienceorevenmetaphysicsasmuchastheotherwayround.Itwasalsorealizedthatmetaphysicsoughttobeabletoproceedevenwheresemanticsissilent.Forexample,inthetheoryofvalueitisnotoriousthatvaluetermsresistanalysisintotermswithoutevaluativeimplications(itis,afterall,theirdistinct71\nSIMONBLACKBURNmeaningsthatdefinethemasvalueterms),butthisfailureofanalysissurelyshouldnothalttheenterpriseoftryingtounderstandthespecificnatureofvaluingasahumanactivity,notbypurelysemanticinvestigation,butbyreflectingonthepsychological,socio-logical,orevenbiologicalroleofvaluing(Gibbard1990).Valueswouldbedomesticated,butnotbyanalysis,andperhapssimilarprogresscanbemadewithotherareas.Twofinalelementscanbediscernedinthemovefromanalysis.Wehaveseenthattheanalyticparadigmwasgoodatpresentingapreferredrelationbetweentheexoticandthefamiliar:theoneisabsorbedintheother.Butitisnotsogoodattellinguswhattochooseasfamiliarinthefirstplace.Andinfactjustthispointhasengenderedkindsofconflicttowhichtheanalyticparadigmprovidesnoreadysolution.LikeHume,twentieth-centuryPHENOMENALISTS(pp.53–4)foundthestreamofperceptionsfamiliar,andthephysicalworldexotic.Othersexactlyreversethepriority.FregeandRussellfoundLOGIC(chapter4)familiar,andMATHEMATICS(chapter11)exotic;othersthinkthatlogicalrelationsareevenmoreinneedoftheorythanthepropertiesofmathematicalstruc-tures.Somethinkthatcategoricalpropertiesarefamiliar,andthatweoughttoanalyseDISPOSITIONSANDPOWERS(p.702)intermsofthem;othersagainreversethepriority,believingthatcategoricalpropertiesaredangerouslyexotic,whereassciencedealsfamil-iarlywithpowersanddispositions.Intheabsenceofanymethodofsettlingsuchques-tionsofpriority,analysisisrudderless,forwedonotevenknowinwhichdirectiontowork.PerhapsthemostsustainedandinfluentialcaseonbehalfofafundamentallyanalyticmethodinthesecondhalfofthetwentiethcenturyisthatmadebyMichaelDummett(b.1925).Inmanywritings,butmostvisiblyinTheLogicalBasisofMetaphysics(1991),Dummetthasarguedthatthefoundationformetaphysics,andindeeditsentirelegiti-matedomain,liesintheTHEORYOFMEANING(chapter3).Citingthestagnationanddead-lockthatensueswhenrivalmetaphysiciansattemptdirectdescriptionsofreality,Dummettbelievesinsteadthatwemusttakeasoursubjectnotsomuchrealityasthewaywethinkaboutreality.However,thereisnoinvestigationofthoughtofthiskindwhichisnotalsoaninvestigationoflanguageandlogic.Thismeanswemustapproachmetaphysicalproblemsfromthe‘bottom-up’,recognizingthattheproblemisthatof‘thecorrectmodelformeaningforstatementsofthedisputedclass’(Dummett1991:12).Ameaning-theory,asDummettconceivesofit,willgiveusthismodel.Itmustprovideaclearviewofhowourwordsfunction,forasthingsstandweusewordswithvariousmeanings,butwedonotknowwhatitisthatwearedoing.Wehavenoclearoverviewofourownpractices,andinthissensedonotunderstandourselves.Ameaning-theorywillgivearepresentationofwhatapersonlearnswhenlearningalanguage;itwillprovidea‘workableaccountofapracticethatagreeswiththatwhichweinfactobserve’(ibid.:14).Suchatransparentunderstandingofourmeaningswouldcertifythecorrectlogictobeused,foritismeaningthatconfersvalidityorinva-lidityonaninference;henceitwouldsettledisputessuchasthatbetweenrealistsandtheiropponents,anti-realistsindisputedareas.Onesurprisingthingaboutsuchameaning-theoryisthatitsconstructioncanproceed,accordingtoDummett,entirelyinnocentlyofmetaphysics.Itistheneutralbasisofmetaphysics,andnowhereowesanythingtometaphysicaldoctrinesorchoices.Andanothersurprisingaspectisthatitisitselfsufficientlypowerfultosettlemeta-72\nMETAPHYSICSphysicalcontroversy:byadjudicatingbetweenrivalconceptionsoftruthofferedbyrealistsandtheiropponents,‘itwillresolvethesecontroversieswithoutresidue’(ibid.).Dummettisnotquitepromotingapurelyapriorimethodformetaphysics,forthedataonwhichameaning-theorywouldbeconstructedwouldinonesensebeempirical:theywouldconcerntheinferentialpracticesofcompetentusersofthelanguage.Neverthe-lessthedataare,inprinciple,availabletoanycompetentlanguageuserpurelybyreflec-tionuponthatpractice.Thereisnoappealtogeneralscientificfactsorempiricalfactslyingoutsidethesphereoflanguageuse.Yetthereflectionwilldeliverresultsthatcansettleoutstandingmetaphysicaldispute.ThepatternofDummett’sthoughtisclearlyinfluencedbyhisownfavouriteexampleofmetaphysicalcontroversy,thatbetweenCLASSICALMATHEMATICIANSANDINTUITION-ISTS(pp.363–7).Inthisexampleitisplausiblethatthedisputeismetaphysicalinessence,withtheclassiciststhinkingofmathematicsasobjective,real,independentofus,or‘outthere’,andtheintuitioniststhinkingofmathematicaltruthasextendingnofurtherthanmathematicalproof(constructivism).Itisalsoplausiblethattheheartofthedisputeisthecorrectlogicformathematicalproofandinference.Theclassicalmathematicianbelievesinbivalence(thelogicallawthateverypropositioniseithertrueorfalse),whereasfortheintuitionist,sincethereisnoguaranteethateverypropositionisprovableordisprovable,bivalencecannotbeassumed.Andfinally,Dummettbelievesthattheissueofthecorrectlogichingesuponquestionsthatwouldbetackledbyameaning-theory,inthesensethatheconceivesofit:theintuitionistsbelievethatclas-sicalmathematiciansdeceivethemselvesintothinkingthattheycanmakesenseofnotionsthatareinfactsenseless.DecidingwhethertheyarerightaboutthatwouldbethefunctionofjustthekindofperspicuousrepresentationofmeaningthatDummettadvocates.Neverthelesstherearegravedifficultieswiththeprogrammeasitispresented.Themostobviousisthatitisinconceivablethatameaning-theorywiththepowersDummettdescribesshouldbeconstructedintheinnocentwaythathealsorequires.Descriptionsofwhatourwordsmeanarehistoricallyamongthemosttheory-ladenofphilosophicalclaims.Theyareputforward,naturally,byphilosopherswhotakethem-selvestoinhabitaworldofonekindoranother,andwhobelievethatourmentalpowerstakeoneshapeoranother.Weneedonlyrememberempiricistsandpositivistsofallkinds,whoseconceptionofwhatwecouldmeanbyatermdisallowsunderstandingofanythingnotgiveninperceptualexperience.Butontopofthisphilosophicalchoice,theconceptionofperceptualexperiencewasinturnnotpurelyempirical,orobtainedinsomeneutralscientificspirit.Rather,itwasthelocusoffiercedoctrinaldispute,forinstancebetweenperceptualatomistssuchasHumeandRussell,andmoreholisticidea-listssuchasBradley(1846–1924)orNeurath(1882–1945).Anevenmorecentralandprolongeddisputeoverthebestwaytoconceiveofunderstandingisrepresentedbytheproblemofuniversals,whichsurfacedaboveinthedisputebetweenthosewhodoandthosewhodonotthinkofscienceascarvingnatureatthejoints.Tosomereal-ists,meaningsreflectnaturalsimilaritiesamongthingswhichguideandconstraintheapplicationofconceptstonewcases;toothersofananti-realistpersuasiontheydonot,buttherulesthatgovernapplicationareafree-standinghumanconstruction.Butneithersideconceivesthemselvesasridingroughshodoverempiricalfactsaboutthenatureoflanguageuse.73\nSIMONBLACKBURNDummettisnot,ofcourse,denyingtheobviousfactthatpersonswhothinkofmeta-physics,ontology,perceptionandepistemologyonewaywillgivedifferentdescriptionsofmeaningfromthosewhothinkofitanotherway.Hisclaimisthatwhetherornotthisisinfactso,itneednotbeso.Theimpactoftheseintrudersonameaning-theoryis,somehow,capableofbeingavoided.Butsayingthatthisispossibleisfarfromshowinghowitispossible,andDummett’sownexplorationsinthisdirectionhavenotprovedreassuring.Inhisfamouspaper,‘ThePhilosophicalBasisofIntuitionisticLogic’(1978),forexample,heexplainedtheWittgensteiniandoctrinethatmeaningisuseasimplyingthatwecannotbeunderstoodtomeananythingwecannotbeobservedtomean,therebybetrayingafondnessforobservationthatisnotatallepistemologicallyormetaphysicallyneutral–onthecontrary,itisoneoftheprincipalplanksofposi-tivism(thelesscontentiousdoctrinewouldbethatwecannotbeunderstoodtomeananythingthatwecannotbethoughtofasmeaning,leavingtherelationshipofthoughttoobservationtobefoughtoveranotherday).Evenmoreremarkableisthebeliefthatasatisfactorymeaning-theorywillexhaustthefieldofmetaphysics.Wecanseethemagnitudeofthisclaimbyasimpleexample.SupposewethinkofthedomainofTHEOLOGY(chapter15)asapartofmetaphysics.Thenistherereallyaprospectofaninnocent,theologicallyneutralmeaning-theory,notonlytellinguswhatismeantbythevarioustermsusedbyreligiousthinkers,butalsotellingusthemetaphysicaltruthaboutsuchmatters?Surelythegapcannotbebridged:anythinginnocentenoughtoqualifyonthefirstscoremustleavethetruthoftheologicalclaimsundecided;anythingstrongenoughtodecidethem(forinstance,theclaimthatatermlike‘God’derivesitsmeaningfromGod,whomustthereforeexistforustothinkaswedo)isnottheinnocent,neutralstartingpointthatDummettrequiresandpromises.Thesamedilemmaisvisibleinotherareas,suchasethics.Carefulatten-tiontothenatureofevaluativelanguageiscertainlyanecessarypartofanyworth-whilephilosophyofethics,butwithoutawiderviewofthenatureofhumanchoice,desireandactionitwillnotitselfsettledisputesbetweenoneandanothermetaphysicsofvalue.Andsemanticdoctrinesthatbearonsuchdisputes,suchastheviewthatethicallanguageisessentiallyprescriptiveinfunction,ortheviewthatethicalpredi-catesworkinmuchthesamewayascolourpredicates,arenottheneutral,purelyempiricaloutcomeofameaning-theory,butrepresentinsteadwiderphilosophicalthoughtaboutthewayethicalcommitmentsactuallyfunction.Forafinalexample,considerthedescriptionsthatareofferedofthelanguageinwhichwetalkaboutneces-sityandpossibility.Thebest-knownsemanticsfortheselanguagesarethoroughly‘real-istic’,conceivingofusasreferringtopossibleworlds,anddescribingtheinferenceswemakeinclassicalquantificationalterms(ifwearetalkingaboutpossibleworlds,neces-sitybecomesequivalentto‘all’andpossibilityto‘some’,andthelogicisthenunder-stoodclassically).Thesemanticsuccessofthiskindofdescriptionisundoubted,butitsphilosophicalsignificanceiscontroversialintheextreme.Those,likeLewis,whosemetaphysicsisdrivenentirelybysemanticsembracetherealexistenceofdifferentworldstoours;othersrejectanysuchmetaphysicalimplications,andmaintainthatreferencetopossibleworldsissomekindofusefulfiction.Theexampleshowsclearlyhowanoverviewofthelogicofanareacanleaveitsmetaphysicsalmostentirelyinthedark.InspiteofDummett’simpassionedandweightyadvocacy,theprospects,then,foraneutralsemanticmethodologyformetaphysicsarenotallthatbright.74\nMETAPHYSICSThestrengthofDummett’spositionwasitsinsistencethatifmetaphysicsgivesusmorethanvaguelyagreeablepictures,thenitscontentshouldbereflectedinourpractice.Thatis,ifadoctrinesuchasrealismmeansanything,itmustmakeadifferencewhetherornotweadoptit,andthisdifferencemustbemanifestedsome-whereinourthoughtandlanguage.Themathematicalexampleisoneinwhichtheclassicalmathematicianallowsdifferentinferencesfromtheintuitionist;thedifferenceisthereforeoneoflogic.Itisnot,however,cleartowhatextentthiscasegeneralizes–certainly,manypeoplewhothinkofthemselvesasanti-realistsaboutethicshavenoobjectiontoclassicallogicasasystematizationofcorrectinferenceinthatsphere.Kantthoughtthatitwaspreciselybecausemathematicswasourownconstructionthatwehadtherighttopursueitinthecertaintythateverymathematicalpropositionistrueorfalse.Ametaphysicsisanoverallstructuredeterminingwhichexplanationsofourpracticeswefindacceptable:itdetermineswhatisfamiliar,whatexotic,whatcountsasanacceptableaccommodationandwhatisintolerabletous.Thedifferencethatrealismoranti-realismmakesisnotnecessarilyoneoflogic.Itcomesoutpri-marilyinwhatweacceptasexplanationsofourpractice,includingexplanationsofthecorrectnessorotherwiseofvariouspatternsofinferences,anditisbecauseofthisthatclassicallogicalpracticecanhappilycoexistwithananti-realistexplanationofitsappropriateness.Dummett’sconceptionofthesubjectisalmostuniqueinthecontemporarysceneinitsinsistenceonastarting-pointthatisuncontaminatedbyscientificandmetaphysi-caldoctrinesandideologies.Muchmorecommonistheviewthatsuchdoctrineswillthemselvesbehistoricallyconditionedandthereforepossiblychangeable.Itisimpor-tanttonoticethatindetailthenatureofsuchchangesisoftencontroversial.Forexample,ithasbeenclaimedbyRorty(1979)thatthephilosophyofmindanditsprob-lemsarehistoricallyquitelocal,beingmainlytheupshotoftheriseofCartesianphi-losophyintheearlyseventeenthcentury.Toothersthisisincredible,sinceitispossibletopointtotheuniversalhumanpreoccupationwithsuchpossibilitiesaslifeafterdeathorthetransmigrationofsoulsasasuresignthattherelationshipofmindtobodyhasseldombeenfullyundercontrol.Again,someseethedistinctionbetweenprimaryqualitiesontheonehand–suchasextensionandmass,whicharethoughtofasrealqualitiesofobjective,independentbodies–andsecondaryqualitiesontheotherhand–suchascolourortaste,whicharethoughtofasinherentlysubjectiveormind-dependent–asthelocalupshotofthescienceofthesameperiod.ButotherspointtothefoundationofthedistinctioninGreekthought,anditsappearanceinvariousformsinIndianorChinesethought,andarguethatthebasisisnotaparticularscien-tificideology,butisfoundinrelativelyuniversaloraprioriconsiderations.3NaturalismandIdentityTheauthorityoflogiclaybehindtheanalyticideal,forlogicalrelationshipsarethosethatarelaidbarebyanalysis.Whenthisauthorityfalters,asIhavedescribeditdoing,theauthorityofscienceisthenaturalsubstitute,anditisinitsshelterthatmetaphysicsiscurrentlymainlyconducted.Metaphysics,onthisview,isnotdiscontinuouswithscience.Itistheorizingaspursuedbythescientist,sometimesonthebasisoffamiliar75\nSIMONBLACKBURNdataandsometimesinthelightofnewscientificresults.Thisisthenaturalisticself-image,mostforciblypropoundedbytheAmericanphilosopherW.V.O.Quine(b.1908),anddominantinthemindsofmanyormostcontemporaryphilosophers.Inthisview,itisperfectlypropertoattempt,forexample,ametaphysicsofMINDANDBODY(chapter5).Itistobedoneinthespiritofscience,withanup-to-dateunderstandingofneuro-physiology,computerscience,ecology,evolutionarybiology,oranyothersubjectthatringsthearea.Thephilosophermarshalstheresultsofthesciences,andpropoundsaconceptofthemind–bodyrelationshipthatbestmakestheresultsfittogether,justasthescientistmarshalstheresultsofempiricalenquiry,andpropoundsaconceptofthenatureofthingsthatbestmakesthoseresultsfittogether.Themetaphysicians’activ-itydiffersonlyinthelevelofabstractionrequired,andindeedanydifferencefromtheactivityoftheoreticalscienceislikelytobevagueandprovisional.Successinphiloso-phizingwouldmostdramaticallybeshowniftheresultsofphilosophicalenquiryactu-allyplayedaroleindrivingscientificprogress.Itiseasytounderstandwhythenaturalisticself-imageissopopular.Firstofall,itanswersthequestionofhowmetaphysicsispossible.Itiscontinuouswithscience,and,sincescienceispossible,soismetaphysics.Secondly,itallowsthephilosophersomeoftheprestigeandgloryofthescientist.Itisreassuringtoallyphilosophicalreflectionwiththemostsecureandintellectuallyprivilegedelementsofthecontemporaryculture.Perhapsphilosophyisalwayssomethingofafree-loaderwhenitcomestocon-tinuitywiththemostprestigiousactivities.Whentheologyruledtheuniversities,phi-losophyandtheologywerecontinuous;inthefirstpartofthiscentury,andafterthespectacularsuccessesofmodernlogic,philosophywasdeemedcontinuouswithlogic;thenalittlelaterwithlinguistics,andnowphilosophymarchesintothefuturehand-in-handwithscience.(Theprocessisofcoursenotassimpleasthat.Philosophyhashadthescientificself-imageon-and-offsinceAristotle.Butotherparadigmshavealsohadtheirday.)Thirdly,theassimilationofmetaphysicstosciencesolvestheproblemofmethod.Intheabstract,itisdifficulttoknowhowtoconductmetaphysicalenquiry.Howisonetogoaboutsolvingthemind–bodyproblem,ordiscussingthenatureofnecessity?Indistanttimes,theanswermighthavebeenbyreflectiononwhatthedeitycreated,byreflectiononthelogicofmentalormodaldiscourse,orbyreflectiononthelanguageofmentalascriptionormodalembeddings.Butthecurrentanswerismorereassuring:scientifically.Withsomuchtomotivatecontentmentwiththisanswer,itmaybewelltowonderwhetheranelementofwish-fulfilmenthascreptin.Isitpossiblethattheself-imageofphilosophyasbeingcontinuouswithscienceislargelyfantasy–thatwehavebeencapturedbythemostsuperficialresemblancesbetweenphilosophicalandgenuinelyscientificactivity?Mightitbethatscience-envyhasledphilosopherstoseethemselvesthroughacomfortablemodernhaze?Ofcourse,ifapproachingasubjectscientificallysimplymeansattemptingtofollowitthroughinadisciplinedway,takingaccountoftheknownfacts,buildingonourpredecessor’slabours,usingthebestwayswehaveofdistinguishingthegoodfromthebadorthetruefromthefalse,thenphilosophyasprop-erlypractised,likeanyotherdiscipline,mustbepursuedscientifically.Butinthatsenseanovelistorpoetmayalsopractisetheircraftscientifically;theobstacleremainsthatsuchactivitiesmaybearonlythemostsuperficialresemblancetothepracticeofnaturalsciences.76\nMETAPHYSICSThereasonforthisisclearifweconsiderwhatwewantfromasatisfactoryrela-tionshipbetweentheexoticandthefamiliar.Supposewearehappywithscientificfacts:inthecaseofpeople,letussupposethesearefactsabouttheirbehaviourandtheirbrains.Thenaturalistwillbelievethatinanimportantsensebrainsandbehaviourisallthatthereis.Whatthenistheretosayaboutmind?Thesimplestansweristhatmentalstatesoreventsorpropertiesarestates,events,orpropertiesofthepreferrednaturalkind.TheequationwouldbepresentedaslikethatbetweenwaterandH2O,orbetweenlightningandelectricaldischarge.Anenormouslyimportantshifttookplacewhentheseequationsemergedasnewparadigmsofmethodinmetaphysics,largelyinthe1970s,anditwasduetotheworkofwriterssuchasPutnam(1975)andKripke(1980).Itcametobebelievedthatwecanrelateanexoticareatoadomesticoneinalesshazardouswaythanbyconcentratingonequationsofmeaning.Allwemayneedisanidentityclaimaboutthethingsandpropertiesinquestion.Wedonotalsoneedtheclaimthattheverysamethingismeantbytalkingintermsofoneandintermsoftheother.Itissufficient,forinstance,tounderstandwaterinscientifictermstoknowthatwaterisH2O;wedonotalsoneedtherequirementthatthetwotermsmeanthesame.Similarly,wewouldbringmindandconsciousnessdowntonaturalearthifwecouldidentifymentalstates,processesandevents,withphysicalstates,processesandevents.Wedonotalsohavetoclaimthatitmeansthesametotalkineitherway.Iden-tityofthingsandpropertiesbecametherulingrelationshiptosearchfor,notidentityofmeaning.However,casualacquaintancewiththemodernliteratureshowsthatphilosophicaldiscussionofsuchanequationisnotverymuchlikescientificdiscussionoftheseequa-tions.There,onewouldfindvariouskindsofempiricalevidenceandtheory.Inthephilosophicalcasewehaveageneralbeliefthatmindmustbethoughtofasfunda-mentallyphysical,andthenweconductagreatdealofarmchairtheorizingaboutthenatureofthesustainingequationsandidentities.Thistheorizingdoesnotproceedwithveryconvincingexamplesofactualidentitiestohand,nor,indeed,isitevenknownwhetherconvincing‘type–type’identitiesexist.Thus,itisalwaystruethatlightningisanelectricaldischarge,orwaterisH2O,butitmaynotalwaysbetruethatamentaleventsuchasapaincoexistswithatypeofphysicalevent,orwhetherpainmightbe‘variablyrealized’indifferentbrainsordifferentpsychologies.Itisherethatphiloso-phersdiscusswhethertheequationiscontingentornecessary,whetheritrelatestypesofmentaleventstotypesofphysicaleventsorwhetheritissufficientthateachmentaleventisidenticalwithsome–possiblydifferent–kindofphysicalevent,whetheritmattersthatdescriptionofaperson’smentalityofteninvokesrelationswiththeenvi-ronment,historyandculture,andsoon.Withoutthiskindofdiscussiontheequationisoflittleinterest,butitisthistheorizingthatisinsufficientlylikeanythingthescien-tistdoesthatcastsdoubtontheQuineanparadigm.Infact,itmaybeclaimedthattheentirediscussionofsuchidentitieswaitsuponapieceofanalysisthatis,unfortunately,notyetcompleted.Weknowhowtoassessques-tionsofidentitywhenthesubjectsarethings;wemayknowhowtoassessthemincaseslikethatofwaterorlightning,wherewhatisatquestionisthewayinwhichanaturalkindofeventisconstituted.Butdoweknowhowtoassesscross-categoryidentitieswhenthesubjectsarefacts,events,statesorproperties?Whatdoesitmean,forexample,tosaythatmystatewhenIenjoyaglassofwineisidenticalwiththestateof77\nSIMONBLACKBURNsomepartofmybrain,orthattheeventofmyenjoymentisidenticalwithsomesuchevent,orthatthepropertyofenjoyingthewineisidenticalwithsomeneurophysio-logicalproperty?Suchidentityclaimsdonotweartheirmeaningsontheirsleeve;nordoesametaphysicscentredonassertingthem.Wedonotevenknowwhethereachoftheabstractcategoriesinvolved–property,state,eventorfact–relatestoidentityinthesameway,orwhetherverydifferentcriteriaareneededineachcase.Togiveasimpleparallel,somephilosophersholdthatifapadlockopensbyhavingthreenotchesinline,thenitmakessensetosaythatthestateorpropertyofbeingunlockedis,inthispadlock(notothers),thatofhavingthreenotchesinline;otherphilosophersdoubtiftheequa-tionmakessense,orwhether,ifitdoes,itisofanymetaphysicaluse(Lewis1966;Blackburn1993).Itdoesnotfollowthatthesequestionswillbeinvestigatedbyexactlythetechniquesmodelledonanalyticalchemistry,anddevelopedbyRussell,Moore,Carnapandtheirfollowers.Butsupportforthisconceptionofthesubjectcomesfromthefollowingthoughts.Supposeweneedtolocateawayofthinkingaccurately,inordertogainanunderstandingofthecategoriesituses.Thentheprimarydataaretheinferencesthatstructurethearea.Theseareconvenientlythoughtofasofthreekinds.Therearethingsthatenableustomakeinferencestodoctrinesinthearea(sometimescalledtheassertibilityconditionsofsuchdoctrines),thereareinferencesamongsuchdoctrines,andtherearethingswemayinferfromthem,ortheirconsequences.Stayingwiththeexampleofmind,thestatementthatsomeoneisenjoyingtheirglassofwinemaybeinferredfromcertainaspectsoftheirbehaviour;itenablesustoinferthattheybelievethatthereisnothingpoisonousaboutthewine(andmanyotherthingsofthekind),anditenablesustoinfer(withlessconfidence)thattheyarelikelytostayputforsomewhile,andunlikelytostartdoingpush-upsorspitonthecarpet.Theanalyticidealsug-geststhatsuchinferences,andmanylikethem,givethestatementitsidentity.Nottomakethemwouldeventuallyshowthatsomeonedoesnotfullyunderstandwhatisinvolvedinenjoyingaglassofwine;itwouldrevealincompetencewiththenotionsinvolved.Butidentifyingtheseinferencesis,ineffect,theprojectofanalysis.Atermisanalysed,intheclassicaltradition,preciselywhenweunderstandwhatlicensesitsapplication,andwhatconsequencesmaybedrawnfromit.Theproblemthatmanyauthorsfindwiththislineofthoughtisthattheinferencesinquestiondonotseemtocomeinstatement-sizedbundles.Thementalworldisfullofsurprisesandcaveats:behaviourthatisindicativeofenjoyingaglassofwineinonepersonmaynotbesoinanother;whatelseapersonwantsorbelieveswhentheyenjoyaglassofwinemaybealmostindefinitelyvariable,andforallweknowtheremaybeculturesinwhichdoingpush-upsorspittingonthecarpetisexactlywhatwewouldexpectbywayofshowingsuchenjoyment.Thephenomenon,oftencalledthe‘holismofthemental’,meansthatstraightforwardidentificationofthepatternsofinferenceassociatedwithunderstandingamentalascriptionisnottobeexpected.Oncemore,enjoyingaglassofwineissomethingthatcanbe‘variablyrealized’indifferentpsychologies,evidencedbydifferentbehaviourandgivingrisetodifferentexpression,dependingonawholerangeofotherfactors.Identitiesneverthelessremainpopularinstrumentsforrelatingthefamiliartotheexotic.Buttherearecasesthatraisemoreobviousdifficultiesthanthatofmind.Con-sidertheprojectofgivinganaturalistictheoryofvalues.Supposeitispursuedonthe78\nMETAPHYSICSwater–H2Omodel,byfindingapropertywithwhichavaluesuchasgoodnessmayplau-siblybeidentified.Letussupposeforsimplicity’ssakethatapropertyisselected,suchasthatofcreatinghappiness.Thisisthoughtofasanatural,empiricallyrespectableproperty,itbeingafairlyordinaryfactabouttheworld(howeverhardtoassessinsomecases)thatsomethingscausehappinessandothersdonot.Sothemetaphysiciannatu-ralizesvaluesbymakingtheequation.Admittedly,sayingthatsomethingisgoodisnotjustsayingthatitcreateshappiness–Moore(1903)refutedthatidea.Nevertheless,theidentity-inclinedmetaphysicianinsists,thepropertiesarethesame.Thepuzzlewithsuchanideaisthatitisquiteunclearwhichoftheproblemsthatethicsgeneratesitactuallysolves.Itleavesuntouchedanyinvestigationoftheparticular‘take’onthecre-ationofhappinessthatishadbythosewhothinkthatitisgood,asopposedtothosewhoareindifferenttoit,orevenwhothinkthatitisbad.Itdoesnotbyitselftelluswhatkindofmistakeismadebytheseotherpeople.Isitanobjectivemistake,oramoresubjectiveone;istheirerroranempiricalerror,oroneoflogic?What,indeed,doesitmeantoseethecreationofhappinessundertheethicalheading?Whilesuchquestionsremain,theidentityseemsitselfapoorcontributiontotheoverallprojectofgivinganaturalisticunderstandingofourengagementwithvalues.TheLimitsofOntologyAlthoughdoingmetaphysicsbymeansoftheidentityofproperty,state,eventorfactispopular,itintroducesaparticularattitudetoitsproblems:oneinwhichtheprimaryquestionsareonesofwhatexist;onesofontology.Thecentralquestionbecomeswhetherwehavetwothingsorone,whetherthethingsareevents,properties,states,factsorothermorehomelyitems.ThisisalsoacontrastwiththeapproachofanalystslikeDummett.Fortheanalyst,theunitofsignificanceistheentiresentence;thereisnoself-standinginvestigationintothekindsofthingthatpartsofsentencessuchaspredicatesstandfor(however,asTaylor(1993)remarks,realistshaveoftenfeltmotivatedtoexpoundtheexistenceandnatureofsuchitemsasuniversalsandproperties).Butmanyproblems,suchasthatofvalue,suggestthatnotallmetaphysicalissuesarewellapproachedintheseterms.Anotherexamplewhereontologyscarcelyseemstobetheissueistheproblemofnaturallawornecessity.Thisproblem,broughtintoprominencebyHume(1978),isthatofgainingsomeconceptionofanynecessityunderlyingthebruteempiri-calorderofevents.Events,asweapprehendthem,happenoneafteranotherincontin-gentpatternsthatthenaturalscientistdiscovers.Descriptionsofthewaytheyfalloutmaybeallright;theyformthefamiliarclass.Whatisexoticandhardtothinkaboutisareasonwhytheymustfalloutastheydo,orwhytheuniversemustbeorderedasitis,orwhytheorderthatisdiscoveredmustcontinuetogovernitsevolutionsinthefuture.ThetheologicalworldviewcouldtakerefugeinthenecessitythoughttoattachtoGod’sexistenceandtoGod’sgoodnessincontinuingtosustainthesamenaturalorder;whensuchananswernolongerappealstopeople’smindsthereisurgentneedforsomethingtofillitsplace.Theproblemisessentiallymetaphysical,forwhatwearelookingforissomeconceptionoftheholdingofalawofnature:somethinglyingbehindorabovetheactualpatternofeventsandconstrainingthemtofalloutastheydo.Nowwillsuchaproblembeeasedbyanontologicaldoctrine?Manysupposeitcan.Theybelievethatifwecansay,sufficientlyseriously,thatnaturallawsexist,theneverythingisfine.Itwould79\nSIMONBLACKBURNnotbesufficientlyserioustosaythisbutonlytomeanthattherearesomeregularitiesthataresocentraltotheenterpriseofsciencethattheydeservecentralandprivilegedplacesinthewayswesystematizeourunderstandingofthepatternsofevents.Theproblemistogainanyconceptionofwhatitisfortheserealnaturallawstoexist,orhowthingsofthenaturetheyneedtohaveareevenpossible.Howcanthereexistastrait-jacket,whosenatureatonetimegovernshowthingsmustfalloutatlatertimesoratallothertimes(Blackburn1993:chs3,5)?Moreontologyjustprovidesmorethings,butthingsandtheircontinuedgoodbehaviourarepartoftheproblem,andsodonotseemtobeanypartofthesolution.Afinal,evenclearerexampleinwhichontologycannotbethecentralissueistheproblemoffreewill.Realizingthatweourselves,physically,aredeterminedcomplexes,oratbestarepermittedonlysuchdeparturesfromdeterminationastherandomeventsofquantummechanicsallow,philosophersfriendlytofree-willmaystruggletofindafurther‘part’ofus,suchasaCartesian,ghostly,governingagency.Butthedilemmaisnotescapedbythesemeans,foritreturnstoplaguewhateverextraisadded,foritremainstobeexplainedhowitisitselfneitherwhollydeterminednorwhollyrandom.Theadditionalontologyisnohelpwiththesolution.Problemswithidentityhaveledmanynaturaliststobackdownalittle,andstaketheirfaithonadifferent,andatfirstsightmoretractablerelation:thatofsupervenience.4SupervenienceWhetherornotitmakessensetoidentifymentalandphysicalevents,orethicalandnaturalproperties,atleastweoughttothinkthatthementalarisesoutofthephysi-cal,andtheethicaloutofthenatural.Thisiscommonlyputbysayingthatthementalsupervenesonthephysical,andtheethicalonthenatural.Whatdoesthismean?Thefundamentalideaisthatoncethephysicalaspectsoftheworldarecompletelyfixed,soarethemental.Andoncethenaturalfactsaboutaworldarecompletelyfixed,soisthequestionofitsethics–whichactionsarerightandwrong,orwhobehavedwellorbadly.Oncethefamiliarisfixed,soistheexotic.IfwethinkintermsofGodcreatingaworld,thenallhehastodo,accordingtothisidea,istocreatethephysicalworld.Hewilltherebyhavedonesufficienttofixthementalfacts,ortheethicalfacts.Thereisnosecondcreativetasktocomplete.Thisideapromisesmanyofthebenefitsofidentityclaims,withoutinvolvingitselfintheissuesoftheprevioussection.Itneedsnocontroversialmetaphysicsoffacts,statesofaffairsorproperties,noranymethodofresolvingdisputedidentityclaimsinvolvingthesenotions.Butitgivesthesamesensethatthereisinreality‘nothingbut’theunderlying,physicalornaturalworld.Atleastitremovesonekindofmetaphysicaldistancebetweenthetwocategories;thedistancethatwouldbewerethereasecondcreativetasktocomplete,sothatwecouldimaginethephysicalbeingasitis,butwithoutthementaloverlay,orthenaturalbeingasitis,butwithouttheethicaloverlay.Itisnotablethattogiveusthisadvantage,thesupervenienceclaimmustsomewhereinvolveanecessity.Therewouldbeasecondcreativetasktocompleteif,aswellascre-atingphysics,Godhadtoshopforoneamongseveralpossiblewaysinwhichphysicsisabletofixthemental(inthisworld).Ifthisrelationshipwerecontingent,capableof80\nMETAPHYSICStakingdifferentformsindifferentpossibleworlds,thenclearlytherewouldbeasecondcreativetasktocomplete.Itwouldnotbestrictlytruethatphysicsfixedeverything;onlyphysicspluswhateveritisaswellasphysicsthatlinkseverythingelsetoit.Itis,Ibelieve,worthyofnotethatthebestunderstoodexamplesofsuperveniencedonotquitegiveusthissortofnecessity.Manyexamplesinvolveareferencetoper-ception,aswhenthefaceinthepicturesupervenesuponanarrayofdots.Itiscertainlytruethattocreatethefacenomoreisnecessarythantocreatethearrayofdots.Butthatseemstobebecausetheworldcontainsperceiverssuchasourselves;relativetodif-ferentperceptualpowerstherewouldnolongerbeafaceinthepicture.So,strictly,Godhadtodomoretocreatethefaceinthepicturethanjustarrangethedots;healsohadtogeneratetheperceptualsensitivitiescapableofrespondingtojustthatarrayinjustthatway.Similarly,tocreatesecondaryqualities,suchasthoseofcolour,itisnotonlynecessarytocreateaworldinwhichsurfacesreflectlightinvariousways,butalsotocreatethekindsofperceptualsystemsthatdetectthevariationsinjustthatway.Inthesecases,insteadofthereassuring,physicalist-soundingthesisthateverythingsupervenesontheunderlyingphysics,weonlygetthatitsupervenesontheunderly-ingphysicsandtherelationsbetweenphysicsandperceivers.Granted,afull-blownphysicalistwillpromisethatthisrelationinturnsupervenesuponthephysicsofthesurfaceandthephysicsoftheperceiver,but,sofar,thisremainssimplyapromise.Wedonothave,incaseslikethis,reassuringexamplesofthenecessitythatsupervenienceclaimsrequire.Inthecaseofethicsitisquiteunsatisfactorysimplytocitethesupervenienceofvaluesonnaturalfacts,andthentohopethatthismakestheintelligiblebridgerequired.Theproblemisobviouslytoexplainthenecessityinquestion.Grantedthatanytwoworldsthatareidenticalinallnaturalfeaturesareidenticalinallethicalfeatures,whyshouldthishavetobeso?What,logicallyormetaphysically,preventstherebeingadimensionoffreedom,theextrathingthatGodhastodotofixthefact–valuelinks?Explainingsupervenienceheremarksaconstraintonsatisfactorytheory.Somebelievethatitcanbemetbycitingtheparallelwithperceptualsupervenienceclaimssuchastheonesdiscussedabove;othersdenythis,butthinkthatattentiontothepurposesforwhichwevaluethingscanonlybemetbyavaluesystemobeyingthesupervenienceconstraint.Therearisesakindofpragmaticjustificationforthisaspectofthelogicoftheactivity(Blackburn1993).Oncemoretheissueisoneofexplanation,anddiffer-entmetaphysicalchoicesareprimarilysignificantbecauseofthedifferentexplanatorypackagestheyoffer.5ConclusionIhavestructuredthissurveyroundtheproblemofmethodinmetaphysics,andsomeofthemostinfluentialsuggestionsaboutthesourceofmetaphysicalmethodandofmetaphysicalauthority.Ihavealsosketchedsomeofthewaysinwhichauthoritativesuggestions–analysis,meaning-theoryandscientificreduction–haveprovedatbestaids,butneverfinalarbitersofmetaphysicalsuccess.Myownbeliefisthatmetaphysicsisbetterregardedasanoceanintowhichallintellectualriversflow.Whentheprob-lemsareasabstractandthekindsofexplanationrequiredaresodifficulttoformulate,81\nROBINLEPOIDEVINthereisboundtobeinfluencefrommanydifferentsources:theself-conceptionofagivenperiod,thescientificparadigmsofthatperiod,themostinfluentialexamplesofintellectualsuccess,eventheethicalandpoliticalagenda,aswellasconceptionsoftheroleofthegiventime.Thisisnotacauseforregret,butitmayperhapsarmuswithacautiousscepticismwhencandidatesformetaphysicalcertaintypresentthemselves.6TimeWhatistime?Wemaybeatalosstoanswerthisratherbald,abstractquestion.Butsupposeweapproachitviathemoretractablequestionofhowwebecomeawareofthepassageoftime.Tothis,theanswerissurelythroughawarenessofchange,eitherintheexternalworld–thetickingofaclock,themovementofcloudsinthesky,thesettingofthesun–orinourownthoughts.Butifawarenessoftimeandawarenessofchangearethesamething,thenperhapsthebestanswertoourfirstquestionisthis:timejustischange.ARISTOTLE(chapter23)attributedthisanswertosomeofhispredecessors.And,aswithmanyofhispredecessors’opinions,hefoundfaultwithit.Timecouldnotbethesamethingaschange,hesaid,forfirstchangecangoatdifferentrates,speeduporslowdown,butnotsotime,andsecondlychangeisconfinedtoapartofspacewhereastimeisuniversal.Whatarewetomakeoftheseobjections?Surelytimedoesspeeduporslowdown,oratleastitappearstodoso.Forpeopleinlove,afewhoursspenttogetherwillpassalltooswiftly,whereastimewillhangheavilyduringalabourofunremittingtedium.Butsuchphenomenaareeasilydismissedasillusory.Wecanbedeceivedaboutspatialmatters,suchastheshapeorsizeofanobject,oritsdistancefromus,sowhynotalsoabouttemporalmatters?Toseewhetheritmakessensetosupposethattimeitselfcouldpassatdifferentrates,considerhowwemeasuretherateofotherkindsofchange:thespeedofapassingbus,forexample.Wemeasurethedistanceitcoversagainsttime.Orconsiderakettleonastove.Itsrateofheatingisgivenbymeasuringtheriseintemperatureagainsttime.Sorateofchangeisvariationinsomedimensioninsomanyunitsoftime.How,then,wouldwemeasuretherateofpassageoftime?Why,againsttime,presumably.Butthisleadstotheconclusionthattherateofthepassageoftimemustnevervary.Forhowlongcouldfiveminutestakeifnotfiveminutes?ButAristotle’sobjectionperhapsmissesthepoint.Itistruethattimecouldnotbeidentifiedwithparticularchanges,suchasthecrumblingofasandcastle.Buttoidentifytimewithchangeissurelytoidentifytimewithchangeingeneral.Nowitisnotatallclearthatchangeingeneral–thatis,thesumofallchangesintheuniverse–couldintelligiblyberegardedasproceedingatvaryingrates.Trytoimagineeverychangeintheworldsuddenlydoublinginspeed.Doesthatideamakesense?Aristotlewouldnothavethoughtso.Foronething,wecouldnotpossiblynoticesuchachangeinrate,forweonlynoticethechangeintherateofsomechangewhencomparingitwithotherchanges.Wenoticetheshorteningofdayswiththeonsetofwinterbymea-suringthetimebetweensunriseandsunsetagainstconventionaltimepiecesorourownbiologicalclocks.Theideathattimeistobeidentified,notwithparticularchanges,butwithchangeingeneralseemsalsotoavoidAristotle’ssecondobjection,thatchangeisconfinedto82\nMETAPHYSICSpartsofspace,whereastimeisuniversal.Onlyindividualchangesarespatiallycon-fined,butthetotalityofchangecoversthewholeofspace.Theseconsiderationsmayhaveremovedoneambiguityinthenotionofchange,butthereremainsanother.Whatkindofchangedowesupposetimetobe?Dowethinkthattimeisthesameasthesumofalltheordinarychangesofwhichwecandirectlybeaware,suchasthechangingcolourofaleaf,andalsothosewhichunderlieper-ceivablechanges,thoughnotthemselvesperceivable,suchasthemotionofmolecules?Orareweinsteadthinkingofthepassageoftimeitself,theinexorablemovementofthingsoncepresentintotheever-distantpast?Ofcourse,aphilosopherwhosaidthattimewastobedefinedasthepassageoftimewouldnotgetmuchofafollowing,sincesuchadefinitiondefinestimeintermsofitself.Weneedtohavesomewayofdefiningthepassageoftime.Thisismostvividlydescribed(thoughsomephilosopherswouldobjecttothiswayofdescribingit)asthechangeineventsastheyceasetobefuture,becomepresent,andthenincreasinglypast.Onewayofcapturingthedistinctionaboveisintermsoffirst-andsecond-orderchange.First-orderchangeischangeinthepropertiesofthingsintheworld,where‘things’areconceivedofasitemsthatpersistthroughtime,suchastrees,atomsandpersons.Second-orderchangeischangeinfirst-orderchanges,namelytheshiftingdegreeoffuturityorpastnessofsuchfirst-orderchanges.Second-orderchange,thechangingofchanges,isthepassageoftime.Sowhenwesaytimeischange,isthisfirst-orderchangeorsecond-orderchange?Supposewemeanfirst-orderchange.Now,itmightseemtousthatwecouldimagineeveryprocessintheuniversecomingtoastop–perhapsaftertheso-called‘heat-death’oftheuniverse,whereallenergyisperfectlyevenlydispersed–andyettimecontinu-ingtopass.Endlessaeonsoftimemightpassinacompletelydead,motionlessuniverse.Toputitintermsofthedistinctionabove:second-orderchangeneednotimplyongoingfirst-orderchange.Butwherewehavesecond-orderchange,wehavetime.Somaybetimecanexistintheabsenceoffirst-orderchange?Aristotledidnotthinkso,onthegroundsthat,wereall(first-order)changetocease,wewouldceasetonoticethepassageoftime.ButwemightbemorecautiousthanAristotleinmakingperceivabil-ityacriterionofintelligibility.Perhapstherearesomestatesofaffairswhichwecouldnevereveninprincipledetect.Whatifinsteadwedefinetimeassecond-orderchange?This,moresubtle,positionlooksunassailable.Howcouldtimeexistunlessitalsopassed?Toseehowthingsmightbeotherwise,weneedtointroduceatthispointanotherdistinction,duetotheCambridgephilosopherandoldercontemporaryofRussellandMoore,J.E.McTaggart(1866–1925).McTaggartdistinguishedbetweenwhathecalledtheA-seriesandtheB-series,representingdifferentwaysofdefiningpositionsintime.TodefinetheA-seriespositionofaneventistodefinethetimeoftheeventaspast(byvaryingdegrees),presentorfuture(byvaryingdegrees).TodefinetheB-seriespositionofaneventistodefinethetimeoftheeventintermsofitsrelationswithotherevents,thatiswhetheritisearlierthan,laterthanorsimultaneouswiththoseotherevents.Themoststrikingdifferencebetweenthesetwoseriesisthat,whereastheA-seriespositionofaneventisconstantlychanging,itsB-seriespositionremainsfixed.Ifanexplosiontakesplace,andisfollowedbyafire,thenthesetwoeventsremainforeverinthisrelationtoeachother:oneisearlierthantheother.Tosaythattime83\nROBINLEPOIDEVINconsistsofanA-seriesisanotherwayofstatingtheviewthattimeisthesameassecond-orderchange.ButiftimejustconsistsoftheA-series,whatbecomesoftheB-series?Thenaturalandplausibleansweristhatthedifferencebetweenthetwoseriesisoneonlyoflan-guage,notofreality.Inreality,B-seriespositionsareadirectconsequenceofA-seriespositions.Thus,ifmygettingoutofbedthismorningisnowpast,myhavinglunchispresent,andmytakingastrollintheparkisfuture(A-seriespositions),thengettingoutofbedisearlierthanthelunch,andbothareearlierthanthestroll(B-seriespositions).OncetheA-seriespositionsarefixed,soaretheB-seriespositions.Or,asonemightputit,itisA-seriesfactswhichmakeB-seriesstatementstrueorfalse.McTaggart,however,invitesustoconsideranintriguingalternative(althoughhelaterrejectsit):thattimeinrealityconsistssolelyofaB-series.Nowthisideaisnotimmediatelyintelligible,andithastobesaidthatMcTaggartdoesnothelpusverymuchtomakesenseofit,nodoubtbecausehehimselfthoughtthat,ultimately,itdoesnotmakesense.ButwecantrytodosoinwayswhichgobeyondMcTaggart’sowndiscussion.Toraiseasimilarquestiontooneaskedearlier:iftimeisjustaB-series,whatbecomesoftheA-series?NowwecannotreplyherethattheA-seriesisfixedbytheB-seriesandleaveitatthat,becausethereisaveryobvioussenseinwhichtheB-seriescannotdeterminetheA-series.ThefactthatthebattleofHastingsisearlierthanthebattleofTrafalgardoesnotcompletelyfixtheA-seriespositionsoftheseevents.TheA-seriespositionsareonlypartiallyfixed:wecansay,forexample,thatifthebattleofTrafalgarispresent,thenthebattleofHastingsmustbepast,butitspositionintheB-seriesdoesnottelluswhetherthebattleofTrafalgarispresentornot.SomephilosopherstakethisasproofthatadescriptionoftimeinpurelyB-seriesterms,intermsofwhatcomesbeforewhat,leavesoutimportantfacts.Butperhapsalltheaboveconsiderationsleadtoistheconclusionthattherelationbetweenthetwoseriesisnotstraightforward.Weneedtobringinsomefurthercomponent.Whatcouldthatbe?Considerthismorepromisingaccountofthetwoseries.Duringaparticularlyheavydownpour,Isay,withcharacteristicunderstatement,‘It’sraining’.IammakinganA-seriesstatement,inthatIamattributingpresentnesstotherain.ButontheaccountwearenowconsideringwhatmakesmystatementtrueisthepurelyB-seriesfactthatmystatementissimultaneouswiththeraining.Aftertheskyhascleared,Isay‘Itwasraining’.Again,whatmakesthisA-seriesstatement,attributingpastnesstotherain,trueisthepurelyB-seriesfactthatmystatementoccursaftertherain.SowhatB-seriesfacts–facts,thatis,abouttheB-seriespositionsofevents–areabletodoistofixthetruthorfalsityofA-seriesstatements.Thistheoryissometimescalledthetenselesstheoryoftime.Itiscapable,itseems,ofreconcilingthenotionthattimeinrealitysimplyconsistsofaB-series,withtheintuitivebeliefthatstatementssuchas‘It’snowraining’,or‘Ihadmylunchanhourago’,canbetrueorfalse.OneconsequenceoftheviewthattimeconsistssimplyofaB-seriesisthattimedoesnotflow:thereisinrealitynorecessionofeventsintothepast.Whatcouldmotivatesuchaview?Anumberofargumentshavebeenputforwardinitsdefence,thebestknownarisingfromafamousargumentofMcTaggart’s,inwhichheattemptedtoprovetheunreality,notonlyoftheA-series,butoftimeitself.Thereisnospacetodojustice84\nMETAPHYSICStothatargumenthere,butthreeotherconsiderationsinfavouroftheB-seriesview(thatthereisinrealitynoA-series,onlyaB-series)canbestatedbriefly:(a)theplausi-bilityoftheaccountgivenaboveofhowA-seriesstatementscanbemadetruesimplyinvirtueofB-seriesrelationsbetweenthosestatementsandtheeventstheydescribe;(b)thefactthatwecannotmakesenseoftherateoftheflowoftimevarying,whichinturnmakesusquestiontheideaofarateofflow,andhenceofaflowatall;(c)theapparentunanswerabilityofthequestion‘WhyisDnow?’(where‘D’standsfortoday’sdate)exceptinpurelyB-seriesterms:Discorrectlydescribableas‘now’simplybecausethequestionisaskedonthedateitmentions.TheB-series,ortenseless,theoryoftimefaceschallengesofitsown:howcanitaccountforchange,orthedirectionoftime?Suchanaccountmaybeforthcoming,butthereisatleastoneintuitivebeliefconcerningtimewithwhichitcomesintoconflict,namelythebeliefthatthefutureisunreal:therearenofuturefactsinthewaythattherearepresentfacts.OntheB-seriesview,incontrast,alltimesareequallyreal,andthisisperhapsthemoststrikingofitsconsequences.FurtherReadingThischapteremphasizestwentieth-centurymetaphysics,butreadersmightalsowishtoconsultthemetaphysicalworksofhistoricalfiguresdiscussedinotherchapters,especiallyHume,ATrea-tiseofHumanNatureandKant,CritiqueofPureReason.Craig(1987)exploresthegreatmodernchangefromadivinetoahumancontextofmetaphysicalthought.Russell(1905)waslongconsideredamodelforanalyticaltreatmentofmetaphysicalquestions.Wittgenstein(1922),whilehavingametaphysicalcharacterofhisown,inspiredthestrenuouslyanti-metaphysicalattitudeoftheViennaCircle.ThepragmatismofPeirce,James(1907),andDeweyinfluencedmanyrecentAmericanwriters,includingQuine(1951),Goodman(1955),Putnam(1975)andRorty(1979,1989).ThecriticismofpositivisminQuineandWittgenstein(1953)ledtonewperspectivesonthepossibilityofmetaphysics,especiallythatofStrawson(1959),whosefocusondescribingthecontoursofourordinaryconceptualschemehasbeenrejectedintheeliminativismofChurchland(1989)andStich(1983).Kripke(1980)andLewis(1983)haveprovideddifferentaccountsoftheimportanceofmodallogic(dealingwithnecessityandpossibility)formetaphysics.Dummett(1978,1991)hasarguedthatlogicandthetheoryofmeaningarethebasisformetaphysics,especiallyfordealingwiththefundamentalconflictbetweenrealismandanti-realism,aquestionalsoconsideredinBlackburn(1993).TimeHISTORICALLYIMPORTANTTEXTSONTIMEAnylistofthesewouldincludeBookIVofAristotle’sPhysics,Augustine’sConfessions,theLeibniz–Clarkecorrespondence,andKant’sCritiqueofPureReason(seeespeciallytheTranscen-dentalAestheticandtheFirstAntinomyofPureReason).HISTORYOFTHEPHILOSOPHYOFTIMEAnexcellentoverviewoftheoriesoftimefromthePresocraticstothelatermiddleagesisprovidedbySorabji(1983).Thereisauseful,ifbrief,discussionofAristotle,Leibniz85\nSIMONBLACKBURNandKantinVanFraassen(1985).BothancientandmodernwritersarediscussedinTuretsky(1998).INTRODUCTIONSANDANTHOLOGIESApartfromVanFraassen’sbookmentionedabove,Swinburne(1981)coversagoodrangeoftopicsinthephilosophyofspaceandtime.Alivelyintroduction,indialogueform,toissuessuchasthebeginningandendoftime,thepassageoftime,thenatureofeternity,andtherelation-shipbetweentimeandfreedomisSmithandOaklander(1995).LePoidevinandMacBeath(1993)isacollectionofimportantreadings,someofwhichareaccessibletothenon-specialist.ItincludesJ.E.McTaggart’sattemptedproofoftheunrealityoftime.Themostcomprehensiverecentintroductiontospaceandtime,whichdiscussesatlengththequestionofwhethertimeflows,isDainton(2001).SPECIFICTOPICSTherelationshipbetweentimeandchange,andissuesconcerningthetopologicalstructureoftime(e.g.whetherithasabeginning/end,whetheritisinfinitelydivisible,andwhetheritislikeanopenlineoraclosedcircle)areveryclearlydiscussedinNewton-Smith(1980).ThedirectionoftimeisthesubjectofPrice(1996).TheA-series/B-seriesdistinction,thetenselesstheoryoftime,andrelatedtopicssuchasthedirectionoftime,thenatureofchangeandthepossibilityoftime-travel,arepresentedwithconsiderabledepthandoriginalityinthesecondeditionofMellor(1998).Smith(1993)isadetailedandsustainedattackonthetenselesstheory.Acloselyrelatedissue,thatoftherealityofthefuture,isthetopicofLucas(1989)andofFaye(1989),whichalsotacklestherelationshipbetweentimeandcausality,asdoesTooley(1997).Thevariousinter-connectionsbetweentimeandethicsarepursuedinCockburn(1997).Thesignificanceofthetenselesstheoryoftime,andthatofitsrivals,foravarietyofdebatesinphilosophyisthesubjectofacollectionofreadings:LePoidevin(1998).PHILOSOPHYOFSPACE-TIMEPHYSICSTheSpecialandGeneralTheoriesofRelativityhavehadanenormousimpactonphilosophicalthinkingabouttime.Anengaging,thoughnotparticularlyeasy,introductiontotheinteractionofthephysicsandphilosophyofspaceandtimeisRay(1991).Animportantandwide-rangingtextisSklar(1974).Nerlich(1994)isanimportantcollectionofNerlich’swritingsonspace,timeandrelativity.ReferencesAristotle1983:Aristotle’sPhysics,BooksIIIandIV(translatedwithnotesbyE.Hussey).Oxford:ClarendonPress.Augustine,Saint1961:Confessions(translatedbyR.S.Pine-Coffin).Harmondsworth:PenguinBooks.Blackburn,S.1984:SpreadingtheWord:GroundingsinthePhilosophyofLanguage.Oxford:OxfordUniversityPress.——1993:EssaysinQuasi-Realism.NewYork:OxfordUniversityPress.Churchland,P.1989:ANeurocomputationalPerspective:TheNatureofMindandtheStructureofScience.Cambridge,MA:MITPress.Cockburn,D.1997:OtherTimes:PhilosophicalPerspectivesonPast,PresentandFuture.Cambridge:CambridgeUniversityPress.Collingwood,R.1940:AnEssayinMetaphysics.Oxford:ClarendonPress.Craig,E.1987:TheMindofGodandtheWorksofMan.Oxford:OxfordUniversityPress.86\nMETAPHYSICSDainton,B.2001:TimeandSpace.London:AcumenPress.Dummett,M.1978:ThePhilosophicalBasisofIntuitionisticLogic.InTruthandOtherEnigmas,London:Duckworth.——1991:TheLogicalBasisofMetaphysics.Cambridge,MA:HarvardUniversityPress.Faye,J.1989:TheRealityoftheFuture.Odense:OdenseUniversityPress.Gibbard,A.1990:WiseChoices,AptFeelings.Cambridge,MA:HarvardUniversityPress.Goodman,N.1955:Fact,FictionandForecast.London:LondonUniversityPress.Hegel,G.1991[1821]:ThePhilosophyofRight(translatedbyH.B.Nisbet,editedbyA.W.Wood).Cambridge:CambridgeUniversityPress.Hume,D.1978[1739–40]:ATreatiseofHumanNature(editedbyP.H.Nidditch).Oxford:ClarendonPress.James,W.1907:Pragmatism:ANewNameforSomeOldWaysofThinking.NewYork:Longmans.Kant,I.1933:ImmanuelKant’sCritiqueofPureReason,SecondImpression(translatedbyN.K.Smith).London:Macmillan.Kripke,S.1980:NamingandNecessity.Cambridge,MA:HarvardUniversityPress.LePoidevin,R.(ed.)1998:QuestionsofTimeandTense.Oxford:ClarendonPress.LePoidevin,R.andMacBeath,M.(eds)1993:ThePhilosophyofTime.Oxford:OxfordUniversityPress.Leibniz,G.W.andClarke,S.1956:TheLeibniz–ClarkeCorrespondence(editedbyH.G.Alexander).Manchester:ManchesterUniversityPress.Lewis,D.1966:AnArgumentfortheIdentityTheory.InPhilosophicalPapers,vol.1.NewYork:OxfordUniversityPress.——1983:NewWorkforaTheoryofUniversals.AustralasianJournalofPhilosophy,61.Lewis,H.D.(ed.)1963:ClarityisnotEnough.NewYork:HumanitiesPress.Lucas,J.1989:TheFuture.Oxford:Blackwell.Mellor,D.H.1998:RealTimeII.London:Routledge.Moore,G.E.1903:PrincipiaEthica.Cambridge:CambridgeUniversityPress.RevisededitioneditedbyT.R.Baldwin,Cambridge:CambridgeUniversityPress,1993.Nerlich,G.1994:WhatSpacetimeExplains.Cambridge:CambridgeUniversityPress.Newton-Smith,W.H.1980:TheStructureofTime.London:RoutledgeandKeganPaul.Price,H.1996:Time’sArrowandArchimedes’Point.NewYork:OxfordUniversityPress.Putnam,H.1975:TheMeaningof‘Meaning’.InMind,LanguageandReality:PhilosophicalPapers,vol.2.Cambridge:CambridgeUniversityPress.Quine,W.V.O.1951:TwoDogmasofEmpiricism.InFromaLogicalPointofView,NewYork:Harper.Ray,C.1991:Time,SpaceandPhilosophy.London:Routledge.Rorty,R.1979:PhilosophyandtheMirrorofNature.Princeton,NJ:PrincetonUniversityPress.——1989:Contingency,IronyandSolidarity.Cambridge:CambridgeUniversityPress.Russell,B.1905:OnDenoting.Mind,14.Sklar,L.1974:Space,TimeandSpace-Time.Berkeley:UniversityofCaliforniaPress.Smith,Q.1993:TheLanguageofTime.NewYork:OxfordUniversityPress.Smith,Q.andOaklander,L.N.1995:Time,ChangeandFreedom.London:Routledge.Sorabji,R.1983:Time,CreationandtheContinuum.London:Duckworth.Stich,S.1983:FromFolkPsychologytoCognitiveScience.Cambridge,MA:MITPress.Strawson,P.1959:Individuals:AnEssayinDescriptiveMetaphysics.London:Methuen.Swinburne,R.G.1981:SpaceandTime,2ndedn.London:Macmillan.Taylor,B.1993:OnNaturalPropertiesinMetaphysics.Mind,102.Tooley,M.1997:Time,TenseandCausation.Oxford:ClarendonPress.Turetsky,P.1998:Time.London:Routledge.87\nSIMONBLACKBURNVanFraassen,B.1985:AnIntroductiontothePhilosophyofTimeandSpace,2ndedn.NewYork:ColumbiaUniversityPress.Wittgenstein,L.1922:TractatusLogico–Philosophicus.London:Routledge.——1953:PhilosophicalInvestigations.Oxford:Blackwell.DiscussionQuestions1Howismetaphysicspossible?Howdometaphysicalquestionsarise?2Doesmetaphysicsdiscoverthebroadstructureofrealityordescribehowourconceptualschemestructuresourthoughtaboutreality?3Mustallpossibleconceptualschemeshavecertainfeaturesincommon?4Howmightaprogrammeofrevisionarymetaphysicsbejustified?5Canmetaphysicsbeassessedwithoutconsideringitsethicalandpoliticalimplications?6IsmetaphysicsthescienceofBeing?7Doesmetaphysicshaveanacceptablemethod?8Whyismetaphysicspreoccupiedwithitsownpossibility?9Ismetaphysicsafunctionofwhatwefindfamiliarandwhatwefindpuzzling?Canwhatwefindfamiliarchange?10Aretherelimitsbeyondwhichmetaphysicsshouldnotgo?Doesthewholeofmeta-physicsattempttogobeyondtheselimits?11Ifmetaphysicalquestionsareunanswerable,shouldmetaphysicsbeabandoned?12Aremetaphysicalclaimsmeaningless?13Areweabletodeterminewhethertheabsolutepresuppositionsofourthoughtaretrueorfalse?14Canphilosophicalreflectionexistonlywithhindsight?Whatdoesyouranswertellusaboutthenatureofmetaphysics?15Canweeliminateallresidualreferencetoourselveswhenweconsidertheworld?16Doesanyconceptualschemehaveauniqueandspecialrelationshipwiththewaythingsare?17Whatkindofjustificationcantherebeforanywayofdescribingtheworld?18Havephilosophersbeenrighttomoveawayfromanalysisasamethodformetaphysics?19ShouldweacceptRussell’stheoryofdescriptionsasaparadigmofphilosophy?20Inareductionistprogramme,howdowedeterminewhatrequiresreductionandwhatitshouldbereducedto?21Docommon-senseconceptsneedelimination,notanalysis?22Cansemanticsbemetaphysicallyneutral?23Canlogicandatheoryofmeaningsettlemetaphysicalcontroversies?24Ismetaphysicscontinuouswithscience?25Shouldmetaphysicianssearchforidentityofthingsandpropertiesratherthanforidentityofmeaning?26Whydophilosophersandscientistshavedifferentquestionstoaskaboutidentities?27Arethebasicquestionsofmetaphysicsquestionsofontology?28Shouldmetaphysicsdealwithsupervenienceratherthanidentity?88\nMETAPHYSICS29Towhatextentshouldexplanatorysuccessdetermineourchoiceofmetaphysicalposition?30Shouldweidentifytimewithchange?31Doesadescriptionoftimepurelyintermsofwhethereventsareearlierthan,laterthanorsimultaneouswithothereventsleaveoutimportantfactsabouttime?32Doestimeflow?33Arethepast,presentandfutureequallyreal?89\n3PhilosophyofLanguageMARTINDAVIESPhilosophyoflanguagedealswithquestionsthatarisefromourordinary,everydayconceptionoflanguage.(Philosophyoflinguistics,incontrast,followsupquestionsthatarisefromthescientificstudyoflanguage.)Butsayingthisdoesnotyetgiveaclearideaofthesortsofquestionsthatbelongdistinctivelyinphilosophyoflanguage.Wittgenstein(1953,§119)said,‘Theresultsofphilosophyaretheuncoveringofoneoranotherpieceofplainnonsenseandofbumpsthattheunderstandinghasgotbyrunningitsheadupagainstthelimitsoflanguage.’Onthisconception,philosophyisaboutthewaysinwhichweunderstandandmisunderstandlanguage,abouthowwecometomistakeplainnonsenseforsomethingthatisintelligible,andaboutwhatcannotbeexpressedinlanguage.So,onthisview,virtuallyallofphilosophyisconcernedwithquestionsaboutlanguage.Itis,indeed,truethatlanguagehasloomedlargeinthephi-losophyofthelasthundredyearsorso.Butthereisstillaspecific,recognizableareaofthedisciplinethatisphilosophyoflanguage.Itbeginsfromoneabsolutelybasicfactaboutlanguage,namely,thatexpressionsofalanguagehavemeaning,andcanbeusedtotalkaboutobjectsandeventsintheworld.Forphilosophyoflanguage,thecentralphenomenontobestudiedislinguisticmeaning.Thischapterintroducessomeofthewaysinwhichthatstudyproceeds.ReadersmightalsoliketolookatthecloselyrelatedchaptersonPHILOSOPHYOFLOGIC(chapter4),PHILOSOPHYOFMIND(chapter5),FREGEANDRUSSELL(chapter37),andWITTGENSTEIN(chapter39).1Introduction:QuestionsofMeaningQuestionsaboutmeaningarecentralinthephilosophyoflanguage.Thesequestionsareoftwokinds.Ontheonehand,therearequestionsaboutthemeaningsofpartic-ularlinguisticexpressions(words,phrasesandwholesentences);ontheotherhand,therearequestionsaboutthenatureoflinguisticmeaningitself.Questionsofthefirstkindbelongtosemantics(section4);questionsofthesecondkindbelongtometa-semantics(section5).Thebusinessofsemanticsincludesquestionsaboutthemeaningsofsubjectexpres-sions–includingPROPERNAMES(chapter37)(Theaetetus,Fido)andDESCRIPTIVEPHRASES\nPHILOSOPHYOFLANGUAGE(pp.798–9)(‘themaninthegabardinesuit’,‘thepresentkingofFrance’)–andofPRED-ICATE(chapter4)expressionslike‘issitting’,‘barks’,‘isaspy’and‘isbald’.Italsoincludesquestionsaboutthemeaningsofcompletesubject–predicatesentences(‘Theaetetusissitting’,‘ThepresentkingofFranceisbald’).Thereareimportantphilo-sophicalquestionsaboutthemeaningsofothersubjectterms,includingpronouns(‘I’,‘you’,‘she’,‘he’,‘it’,‘they’)anddemonstratives(‘this’,‘that’,‘thisknife’,‘thatbutter’),andalsoaboutthemeaningsofexpressionsthatgobeyondthetermsinthebasiccom-binationofsubjectandpredicate.Expressionsinthislattercategoryincludeadjectives(‘large’asitoccursin‘largeflea’,‘small’asitoccursin‘smallelephant’,‘good’asitoccursin‘goodperson’andin‘goodphilosopher’,‘false’asitoccursin‘falsesentence’andin‘falsenose’),modaladverbs(‘possibly’,‘necessarily),manneradverbs(‘slowly’,‘clumsily’),andmanymore.Stillwithinthedomainofsemantics,andcloselyrelatedtoquestionsaboutthemeaningsofwords,phrasesandsentences,arequestionsaboutthewaysinwhichthemeaningsofwordsdetermine,oratleastconstrain,themeaningsofthephrasesandsentencesinwhichtheyoccur.Evenifweknowwhatkindofthingthemeaningofasentenceis(themeaningof‘Fidobarks’orof‘Themaninthegabardinesuitisaspy’),wealsoneedtounderstandhowcomponentwordsandphrasesmaketheircontribu-tionstothemeaningsofcompletesentences.Ontheothersideofthesemanticsversusmeta-semanticsdividearequestionsaboutthenatureoflinguisticmeaningitself.Someofthesequestionsareontological.Aremeaningsentities;and,ifso,whatkindsofentitiesarethey?Oneputativeanswermightbethatthemeaningsofcompletesentencesarepropositions,andthatanswerwouldlead,inturn,toquestionsaboutthenatureofpropositionsthemselves.Analternativeanswermightbethat,ifthemeaningsofsentencesareentities,thentheyarestatesofaffairs.Someonefollowingupthisalternativemightsaythatinorderforasentencetobetrue,thestateofaffairsthatisthesentence’smeaningneedstobeastateofaffairsthatobtains.Factsmightthenbeidentifiedwithstatesofaffairsthatobtain,andtruesentenceswouldbesaidtobetrueinvirtueoffacts.Inthiscase,theanswertotheonto-logicalquestionaboutmeaningswouldleadtoaversionofthecorrespondencetheoryoftruth.Othermeta-semanticquestionsconcerntheelucidationoranalysisoftheconceptofmeaning.Canwe,forexample,giveanykindofphilosophicalanalysisoftheconceptoflinguisticmeaning;and,ifwecan,whatkindsofideascanlegitimatelybeusedintheanalysis?How,ingeneral,isthemeaningofalinguisticexpressionrelatedtoitsuse?Howistheconceptofmeaningrelatedtotheconceptoftruth?Inparticular,foracompletesentence,whatistherelationbetweenthemeaningofthesentenceandtheconditionsunderwhichanutteranceofthesentencewouldbetrue?Theeverydayideaofmeaningorsignificanceisrelatedtotheideaofwhatiscon-veyedorcommunicatedintheuseoflanguage.Inrecentphilosophyoflanguage,astandardassumptioninmeta-semanticshasbeenthatthereissuchathingastheliteralmeaningoflinguisticexpressions,andthatthetotalcommunicativesignificanceofalinguisticactistheproductjointlyoftheliteralmeaningsoftheexpressionsusedandofcontextualfactors.Accordingtothatmeta-semanticview,semanticsisthestudyoftheliteralmeaningsofexpressions,andofthewaythattheliteralmeaningsofcomplexexpressions(phrasesandsentences)aredeterminedbytheliteralmeaningsoftheir91\nMARTINDAVIEScomponentwords.Strictlyspeaking,questionsabouttheinteractionbetweenliteralmeaningandcontextualfactorsbelong,nottosemantics,buttopragmatics(section6).2TheoriesofMeaningAnswerstosemanticquestionsandanswerstometa-semanticquestionscanbegivenbypropoundingwhatmightbecalledtheoriesofmeaning;butthatphrasehastwoverydifferentsenses.Inthefirstsense,atheoryofmeaninganswerssemanticquestionsbyspecifyingthemeaningsoflinguisticexpressions.Inthesecondsense,atheoryofmeaninganswersmeta-semanticquestionsbyprovidinganelucidatoryaccountofthenatureoflinguisticmeaning.Considersemanticquestionsfirst.Thesearequestionsaboutthemeaningsofwordsandphrases,andaboutthewaysinwhichthesecontributetothemeaningsofwholesentences.Forthewords,phrasesandsentencesofaparticularlanguage,someonemightseektoanswerthesequestionsinaveryexplicitway,byprovidingacertainkindofformal,axiomatizedtheory.Theideawouldbethattheaxiomsofthistheoryshouldspecifyexplicitlythemeaningsofthewordsofthelanguageunderstudy,andthatrulesofinferenceshouldthenpermitthederivation,fromthoseaxioms,oftheoremsspeci-fyingthemeaningsofphrases,andultimatelyofwholesentencesinthatlanguage.Inthecaseofanygivensentence(say,‘Fidobarks’),thederivationofameaning-specifyingtheoremwouldmakeuseoftheaxiomsofthetheorythatspecifythemeaningsofthewordsoccurringinthatsentence(inthiscase,thewords‘Fido’and‘barks’).Thisderivationofatheoremfromaxiomscouldreasonablybesaidtodisplayhow,inthelanguageunderstudy,themeaningsofthecomponentwordscontributetothemeaningofthecompletesentence.Atheorythatshowshowthemeaningsofsentencesdependonthemeaningsoftheirpartsissometimessaidtobecompositional.Theconstructionoftheoriesofthekindenvisagedhereisnotatrivialmatter.Wemightstartoffbysupposingthatatypicaltheoremwouldsaysomethinglike:Themeaningofthesentence‘Fidobarks’isthepropositionthattheparticulardogFidoengagesintheactivityofbarking.or,avoidingtheexplicittalkaboutpropositions:Th1Thesentence‘Fidobarks’meansthattheparticulardogFidoengagesintheactivityofbarking.Wemightalsosupposethattheaxiomsfromwhichthistheoremistobederivedwouldsaythingslike:Ax1Theword‘Fido’meansaparticulardog,namely,Fido.Ax2Theword‘barks’meanstheactivityofbarking.But,eveninthecontextofthisextremelysimpleexample,wecanseethatwewouldimmediatelyconfrontatleasttwoimportantissues.Thefirstissueconcernsthediffer-92\nPHILOSOPHYOFLANGUAGEencebetweensubjecttermsandpredicateterms(here,between‘Fido’and‘barks’);thesecondissueconcernsthederivationalroutefromtheaxiomstothetheorem(here,fromAx1andAx2toTh1).2.1ThedifferencebetweensubjecttermsandpredicatetermsTheissueaboutthedifferencebetweensubjecttermsandpredicatetermsisthis.Thetwoaxiomsthatweproposedhavethesameform.Ineachcase,theaxiomsaysthatanexpres-sionstandsinthemeaningrelationtosomethingintheworld–inonecase,adog,intheothercase,anactivity.So,thoseaxiomsdonotreallyexplainwhythesetwowords–‘Fido’and‘barks’–cangotogethertomakeupasentence,whereasotherpairsofwords–suchastwonouns–cannottogetheryieldasentence,butmerelyconstitutealist.Wecanmakethispointmorevividifweconsider,foramoment,nounsthatarecloselyrelatedtopredicateterms.Consider,forexample,thenoun‘barking’.Thisnouncan,ofcourse,occurinsentences:‘Thebarkingoftheneighbourhooddogskeptmeawake’.Indeed,itcanfunctionasthesubjectterminasentence:‘Barkingisfun’.Butthenoun‘barking’cannotbejuxtaposedwiththeword‘Fido’tomakeasentence.Similarly,con-siderthenoun‘baldness’.Ifwejuxtaposethisnounwithasubjectterm,thenwhatweget–say,‘ThepresentkingofFrancebaldness’–isnotasentence.(Inordertoseethatitisjustalist,wecanimagineitasanaccountofthetopicsdiscussedatameeting.)Thepoint,inshort,isthatthereseemstobeanimportantdifferenceinmeaningbetweenpredicateterms,suchas‘barks’and‘isbald’,andothercloselyrelatedexpres-sions.Themeaningsofthepredicatetermssomehowfitthemforcombiningwithsubjecttermstoformsentences,whilethemeaningsofthoseotherexpressions(‘barking’,‘baldness’)donotsofitthem.Now,supposethatwewereaskedtoofferameaning-specifyingaxiomforthenoun‘barking’inthestyleofAx1andAx2.Wewouldbevirtuallyboundtosay:Thenoun‘barking’meanstheactivityofbarking.Butthissaysjustthesameaboutthemeaningofthenoun‘barking’asAx2saysaboutthemeaningoftheverb‘barks’.Similarly,themodelforaxiomsthatwehaveestab-lishedthusfarwouldsuggestthatfortheverbphrase‘isbald’,andforthenoun‘bald-ness’,theaxiomsshouldbe:Thephrase‘isbald’meansthepropertyofbaldness.Thenoun‘baldness’meansthepropertyofbaldness.But,aswejustsaw,while‘barks’and‘isbald’canfunctionaspredicatetermsinsen-tences,‘barking’and‘baldness’cannot.Theremustbeadifferenceinmeaningbetween‘barks’and‘barking’,andbetween‘isbald’and‘baldness’;buttheaxiomsthatwehaveproposeddonotrevealwhatthisdifferencemightbe.2.2ThederivationalroutefromaxiomstotheoremsTheissueabouttheroutefromaxiomstoatheoremisthis.ThederivationofatheoremfromaxiomsissupposedtohavethestatusofalogicalPROOF(pp.347–9),andit93\nMARTINDAVIESshouldmakeuseofwell-understood,andwell-behaved,logicalresources.So,supposethatweapplytheserequirementstothetheoremthatwehavehighlighted(Th1)andthetwoaxioms(Ax1andAx2).CertainlythetheoremdoesnotfollowfromtheaxiomsjustbywayofthelogicalresourcesofthepropositionalcalculusorthePREDICATECALCULUS(p.367),forexample.TheinferencefromAx1andAx2toTh1isnotalogicallyvalidone.Inordertoobtainavalidinference,wewouldneedtoaddanextrapremise,intheformofanotheraxiom.Clearly,ifweweretoaddthehypotheticalstatement:Iftheword‘Fido’meansaparticulardog,namely,Fido,andtheword‘barks’meanstheactivityofbarking,thenthesentence‘Fidobarks’meansthattheparticulardogFidoengagesintheactivityofbarking.thenwecouldproceedfromthisstatementplusAx1andAx2totheconclusionTh1bywayoffamiliarlogicalrulesofinference.Theformoftheinferencewouldbe:A,B,ifAandBthenC;therefore,C.whichisclearlyvalid.Buttherewouldbegoodreasonstoaspireaftersomethingmoregeneralthanthishypotheticalstatementaboutthespecificwords‘Fido’and‘barks’.Weshouldseekanaxiomthatspeaksingeneraltermsoftheeffectofputtingtogethertwoterms–subjectandpredicate–tomakeasimplesentence.Providedthatwecanfindsomeaccountofwhatasubjecttermisandwhatapredicatetermis,somethingalongthefollowinglinesmightsuggestitself:Ax3IfasubjecttermMmeanssomething,say,X,andapredicatetermNmeanssomething,say,Y,thenthesentencemadeupofMfollowedbyNmeansthatXengagesin,orexemplifies,Y.ThisispleasinglygeneralinthewaythatittalksaboutMandNandtheirmeaningsXandY,anditisageneralizationthatwecan,itseems,instantiateinordertoyieldthemorespecifichypotheticalstatementaboutthesubjectterm‘Fido’andthepredicateterm‘barks’.Ifthatisright,thentheconclusionTh1canbevalidlyderivedfromAx1,Ax2andAx3.So,bymakingexplicitthegeneralizationinAx3,wehavemadesomeprogresswiththerequirementthatthederivationofameaning-specifyingtheoremfromaxiomsissupposedtohavethestatusofalogicalproof.Furthermore,wecouldinstantiatethesamegeneralization,Ax3,togiveahypo-theticalstatementaboutthesubjectterm‘Theaetetus’,thepredicateterm‘issitting’,andthesentencethattheygotogethertomakeup.Takeninconjunctionwithspecificaxiomsaboutthemeaningsof‘Theaetetus’and‘issitting’,thishypotheticalstatementwouldenableustoderiveatheoremspecifyingthemeaningofthesentence‘Theaete-tusissitting’.But,althoughthislookspromising(apartfromtheissueaboutsubjecttermsandpredicateterms),therearestilllegitimatecausesforconcernastowhetherstatementslikeAx1,Ax2andAx3interactwithstandardlogicalresourcesina94\nPHILOSOPHYOFLANGUAGEwell-behavedway.Infact,mostworkinsemanticsproceedsontheassumptionthat,inordertobesureofthelogicalstatusofourderivations,weshouldnotmakeuseofaxiomsandtheoremsthattalkexplicitlyaboutmeaning.Weshallreturntothispointlater(section4.1).TheNotionofMeaningandStandardLogicInstandardlogicalsystemswehaveaprinciple–inessence,Leibniz’sLaw–saying,amongotherthings,thatifFido=RoverthenfromanystatementaboutFidowecaninferthecorrespondingstatementaboutRover,andviceversa.Ifwenowimaginethatoneandthesamedoghastwonames,‘Fido’and‘Rover’,wecanaskwhetherfromthepremise:Fido=Roverplus:Ax1Theword‘Fido’meansaparticulardog,namely,Fido.itfollowsthat:Ax1¢Theword‘Fido’meansaparticulardog,namely,Rover.Eitheranswer–‘yes’or‘no’–ispotentiallyproblematic.IfAx1¢doesfollow,thenwecansee–byafewmoresteps–thateverythingthatcanbesaidaboutthemeaningoftheword‘Fido’canbeequallytrulysaidaboutthemeaningoftheword‘Rover’.Furthermore,thesamewillgoforanypairofwordsthatpickoutthesameobjectintheworld,suchasthepair‘Hesperus’and‘Phosphorus’,whichbothpickouttheplanetVenus.Butweshouldwanttoleaveitopen,atthisearlystageofourenquiry,whethersuchpairsofwordscanbedistinguishedinpointoftheirmeaning.Indeed,aheadofdetailedtheoreticalconsiderations,wemightreasonablyexpectthatsuchpairsofwordswouldsometimesdifferinmeaning,despitepickingoutthesameobject.SoperhapsweshouldpreferthealternativethatsaysthatAx1¢doesnotfollowfromAx1plusthepremisethatFido=Rover.But,inthatcase,weshallhavetoadmitfranklythatwearedepartingfromwhatisfamiliar,andmovingtoakindoftheorywhoselogicalbehaviourisnotsowellunderstood.Wehavebeenconsideringtheoriesofmeaning,inthefirstoftwopossiblesensesofthatphrase.Thesearetheoriesthatseektoanswersemanticquestions.Asemantictheorymaybeatheoryinaquiteformalsense(withaxiomsandrulesofinferencebymeansofwhichtheoremscanbederived)andwehavenotedthatthereisanissueabouttheroleofthenotionofmeaninginformalderivationsorproofs.But,apartfromtheworryaboutlogicalgoodbehaviour,asemantictheorymighttaketheconceptofmeaningasanunanalysed,andunexplained,primitivenotion.Afterall,itisthejobofasemantictheorytotelluswhatlinguisticexpressionsmean.Atheoryofmeaninginthesecondsense–ameta-semantictheory–will,incontrast,setouttoexplaintheconceptofmeaninginotherterms.Ameta-semantictheorywillprovideananalysis,orsomeotherkindofphilosophicalelucidation,ofthenotionoflinguisticmeaning,perhapsbyplot-tingconnectionsbetweenthatnotionandthenotionofuse,orthenotionoftruth.95\nMARTINDAVIESexpressestalkthoughtaboutaboutworldFigure3.1Therelationshipbetweentalk,thoughtandtheworld,whichisexploredinthephilosophyoflanguage,thephilosophyofmindandmetaphysics.3Language,MindandMetaphysics:QuestionsofPriorityAsemantictheoryrelatespiecesoflanguagetopiecesoftheworld.Weuselanguagetotalkabouttheworld,andtoexpressourthoughts,whicharealsoabouttheworld.(Theaboutnessofthoughtsisoftencalledintentionality.)Talk,thoughtandworldformatriangle,andinphilosophyoflanguage,PHILOSOPHYOFMIND(chapter5),andMETA-PHYSICS(chapter2)wemovearoundthistriangle(figure3.1).Thus,forexample,wemighttrytogiveaphilosophicalaccountofsomedistinctionsinreality–say,betweenOBJECTSANDPROPERTIES(pp.726–7),orbetweenparticularsanduniversals–intermsofdifferencesamongwords,orintermsofdifferencesintherealmofthought,providedthatwealreadyhadsomeunderstandingofthoselinguis-ticormentaldifferences.Or,goingtheotherwayabout,wemightassumesomeaccountofthemetaphysicaldifferences,anduseitinourphilosophicalworkinthedomainsoftalkorthought.Weshallshortlyconsiderjustsuchaquestionofrelativeprioritybetweenphilosophyoflanguageandmetaphysics.Therearealsoimportantquestionsofprioritybetweenphilosophyoflanguageandphilosophyofmind.Indeed,anystrategyforelucidatingtheconceptoflinguisticmeaningwillinevitablydependonourgeneralviewoftheorderofpriorityasbetweentalkandthought.Weneedtobeclear,first,justwhatnotionofpriorityisatissuehere.Thenweshallconsiderthreepossibleviewsaboutlanguageandmind.3.1Philosophicalpriorities:languageandmindThekindofprioritythatconcernsushereispriorityintheorderofphilosophicalanaly-sisorelucidation.TosaythatthenotionofXisanalyticallypriortothenotionofYistosaythatYcanbeanalysedorelucidatedintermsofX,whiletheanalysisorelucida-tionofXitselfdoesnothavetoadverttoY.ThustosaythatthenotionofBELIEF(chapter1)isanalyticallypriortothenotionofKNOWLEDGE(chapter1),forexample,istosaythatknowledgecanbeanalysedintermsofbelief,whileagoodanalysisofbeliefdoesnotneedtoreintroducethenotionofknowledge.(Thisisjusttosaywhattheclaimwouldamountto,notwhetheritwouldbecorrect.)Analyticalpriorityshouldbedistinguishedfromontologicalpriorityandfromepis-temologicalpriority.TosaythatXisontologicallypriortoYistosaythatXcanexist96\nPHILOSOPHYOFLANGUAGEwithoutY,althoughYcannotexistwithoutX.Forexample,itmightplausiblybesaidthatindividualsareontologicallypriortonations.TosaythatXisepistemologicallypriortoYistosaythatitispossibletofindoutaboutXwithouthavingtoproceedviaknowl-edgeaboutY,whereasfindingoutaboutYhastogobywayoffindingoutaboutX.Soitisplausible,forexample,thatthepositionsandtrajectoriesofmedium-sizedmater-ialbodiesareepistemologicallypriortothepositionsandtrajectoriesofsubatomicpar-ticles.Wecanfindoutaboutmaterialbodieswithoutinvestigatingsubatomicparticlesbut,itmightbesaid,ourroutetoknowledgeaboutsubatomicparticleshastogoviaobservationsofmaterialbodies.Havingdistinguishedthesethreekindsofpriority,wecanmaketheworkingassumptionthattheyarelogicallyindependentofeachother.Accordingtothatassumptionwecan,forexample,supposethatthenotionofXisanalyticallypriortothenotionofY,withoutbeingobligedtoholdthatXiseitherontologicallyorepiste-mologicallypriortoY.Ourquestionabouttheorderofanalyticalpriorityasbetweenlanguageandmindrelates,particularly,tothenotionoflinguisticmeaningandthenotionofintentionality(aboutness)formentalstates.Thethreepossibleviewsthatweneedtobeawareofarethese.Mindfirst:Thisistheviewthatitispossibletogiveaphilosophicalaccountoftheintentionalityofthoughtswithoutessentiallyadvertingtolanguage,andthatthenotionoflinguisticmeaningcanthenbeanalysedintermsofthethoughtsthatlan-guageisusedtoexpress.Languagefirst:Thisistheviewthatanaccountoflinguisticmeaningcanbegivenwithoutbringingintheintentionalityofthoughts,andthatwhataperson’sthoughtsareaboutcanthenbeanalysedintermsoftheuseoflanguage.Nopriority(bothtogether):Thisistheviewthatthereisnowayofelucidatingthenotionofwhataperson’sthoughtsareaboutwithoutbringinginthenotionoflin-guisticmeaning,northeotherwayaround.Thetwonotionshavetobeexplainedtogether.(Thereisinfactafourthpossibleview,whichinvolvesadifferentkindofnopriorityclaim,namelythatthetwonotionsarequiteunrelated.Butthisview,whilepossible,isquiteimplausible.)Allthreeviewshaveadherents.ThenoanalyticalpriorityviewischaracteristicoftheworkofDonaldDavidson(b.1917),whocouplesitwithontologicalandepiste-mologicalnopriorityclaims.Thus,Davidson(1975)arguesthattherecanbenothoughtwithoutlanguagenorlanguagewithoutthought,andthatthereisnofindingoutindetailwhatapersonbelieveswithoutinterpretingtheperson’sspeech(Davidson1974).ThelanguagefirstviewfindsexpressioninMichaelDummett’s(b.1925)writ-ings(seeDummett1973,1991,1993),whilethemindfirstapproachistakenbyPaulGrice(1913–88)(Grice1989;seealsoSchiffer1972).Givenwhatwehaveassumedaboutthelogicalindependenceofthethreekindsofpriority,wecanseethatoneoptionthatisavailabletousistoagreewithDavidsonindenyingtheontologicalorepiste-mologicalpriorityofmindoverlanguage,andyettofollowGriceintryingtoanalysethenotionoflinguisticmeaningintermsofthethoughtsthatlanguageisusedtoexpress.97\nMARTINDAVIES3.2Philosophicalpriorities:languageandtheworldOurexampleofaquestionofrelativepriorityasbetweenphilosophyoflanguageandmetaphysicscomesfromtheworkofPeterStrawson(b.1919)(Strawson1959,PartII;1970a)onsubjecttermsandpredicateterms.Thedistinctioninlanguagebetweensubjectandpredicatetermsisintuitivelycloselyrelatedtothemetaphysicaldistinctionbetweenobjectsorparticulars,ontheonehand(correspondingtosubjectterms),andpropertiesoruniversals,ontheother(correspondingtopredicateterms).Canweusethemetaphysicaldistinctiontohelpusunderstandthelinguisticone?(Aswehavealreadyseen,wecertainlyneedsomeaccountofthesubject–predicatedistinction.)Orshouldwe,alternatively,seekamorepurelylogico-linguisticaccountofthedifferencebetweensubjecttermsandpredicateterms,andthentrytounderstandthemeta-physicaldistinctionintermsofthelinguisticone?Oneaspectofthesubject–predicatedistinctiontowhichStrawsongivesconsiderableattentioniswhathecallstheasymmetryofsubjectsandpredicatesregardingnegation.Wecanexplainthebasicidealikethis.Whenasubject–predicatesentence(‘Theaetetusissitting’)isnegated(‘ItisnotthecasethatTheaetetusissitting’,or‘Theaetetusisnotsitting’),thenegationcanbetakentogetherwiththepredicatetermtoyieldanewexpression(‘notsitting’)thatisofthesamekind–anewpredicateterm.Butthenegationcannotbetakentogetherwiththesubjecttoyieldanewexpression(‘notTheaetetus’)thatisofthesamekind–anewsubjectterm.Inshort,predicatetermshavenegationswhilesubjecttermsdonot.Inordertoprovide(atleastthebeginningsof)anexplanationofthisfactaboutlan-guage,Strawsoninvitesustoconsiderpropositionsinwhich,asheputsit,ageneralcharacterorkind(orpropertyoruniversal)isassignedto,orpredicatedof,aparticu-lar,orspatio-temporalindividual.Theimportantpointisthatthesegeneralcharacterscomeinincompatibilitygroupsvis-à-vissuchempiricalparticulars,whiletheparticu-larsdonotconverselycomeinincompatibilitygroupsvis-à-visthegeneralcharacters.Thus,forexample,considerthevariouscolours(generalcharacters)andthesundryitemsoffurnitureintheroom(spatio-temporalindividuals).Anitemoffurniturecannotbemorethanonecolour(allover);whereasmorethanoneitemoffurniturecanbethesamecolour.Orconsiderthevariousposturesthatpeoplecanadopt(generalcharacters)andthephilosophersinacourtyard.PerhapsTheaetetusissitting,andsonotstandingorlying,whileTheodorusisalsosittingandSocratesisstanding.Asbothexamplesillustrate,generalcharacterscomeingroupsthat‘compete’forspatio-temporalindividuals;whereastheindividualsdonotcomeingroupsthat‘compete’forthegeneralcharacters.Thereisatleasttheprospect,here,ofanexplanationofthelinguisticdistinctionbetweensubjecttermsandpredicatetermsthatappealstothemetaphysicaldistinctionbetweenspatio-temporalindividualsandthegeneralcharactersthattheymayexemplify.ThisapproachexhibitsanorderofexplanationthatistheoppositeoftheonethatisusuallyassociatedwiththeGermanmathematicianandphilosopherGottlobFREGE(1848–1925)(chapter37).OntheFregeanapproach,themetaphysicalorontologicalcategoryofobjectsistobereadofffromanantecedentlyfixedlinguisticcategoryoflogicalsubjectterms(roughly,names).Sosomeoneadoptingthisapproachwouldneedtobeabletogiveapurelylogico-linguisticcriterionforanexpressiontobeaname.It98\nPHILOSOPHYOFLANGUAGEisnotaneasymattertodischargethisobligationwithoutsmugglingbackinsomethinglikethedistinctionbetweenobjectsandproperties.Strawsonhimself,afterreviewingsomeofthelogico-linguisticmarksofthesubject–predicatedistinctionthatmightbedrawnonintheFregeanorderofexplanation,remarks:Thegeneralrelativenaturesofthetermscombinedinthefundamentalcombinationofpredication–thedifferencesorallegeddifferencesbetweensubjectsandpredicates–havesofarbeendiscussedinalmostexclusivelyformalterms,intermsbelongingtoformallogicitselfortogrammar.Tounderstandthematterfully,wemustbepreparedtousearichervocabulary,arangeofnotionswhichfalloutsidetheseformallimits.Weassumethatthesubject–predicateduality,andhencethedifferencessofarremarkedon,reflectsomefundamentalfeaturesofourthoughtabouttheworld.(Strawson1974a:13–14)Now,itwouldnot,infact,bequiteaccuratetosaythatStrawsonawardsabsolutepri-oritytometaphysics;butwecan,atleast,saythathedoesnotawardapre-eminentroletophilosophyoflanguage.Inthis,Strawson’spositioncontrastswiththatofDummett.Aswehavealreadynoted,whenitcomestothequestionofpriorityasbetweenlan-guageandmind,Dummettadoptsthelanguagefirstoption.Aswecannowadd,healsotakesaFregeanviewonthequestionaboutlanguageandmetaphysics.So,Dummettdoesregardthephilosophyoflanguageaspre-eminent.Indeed,heoncewrote:WemaycharacterizeanalyticalphilosophyasthatwhichfollowsFregeinacceptingthatthephilosophyoflanguageisthefoundationoftherestofthesubject.(Dummett1978:441)Thismaynotbeanentirelyhappycharacterizationofanalyticalphilosophyasawhole–sinceitwouldleavemanyanalyticalphilosophersonthewrongsideoftheclassification–butitdoesindicateoneinfluentiallineofthoughtwithinanalyticalphilosophy.Weshouldnot,however,thinkthattheonlypossiblereasonforinterestinphiloso-phyoflanguageisthatitmightholdthekeytoquestionsinphilosophyofmindandmetaphysics.Questionsaboutmeaningarenolessdeep,andnolessimportant,forthefactthat,inordertoinvestigatethemfully,wemaywellneedtodrawonresourcescomingfromotherareasofphilosophy.4SemanticTheories:Davidson’sProgrammeWehavealreadymettheideaofaformal,axiomatizedtheorythatpermitsthesystematicderivationoftheoremsspecifyingthemeaningsofcompletesentencesdrawnfromsomelanguage(section2).Thesentencesmightbesimpleorcomplex,andthelanguagemightbeaformallanguageoranaturallanguage.Ifwewereinterestedinactuallyconstructingtheoriesofthistype,thenwemightdowelltoproceedbystages.Thefirstandsimpleststagewouldinvolvejustafinitestockofsentencesfromaformallanguage,andwouldtreateachofthesentencesasanunstructuredunit.99\nMARTINDAVIESPerhapstherearejustthreesentences,‘A’,‘B’and‘C’,where‘A’meansthatTheaetetusissitting,‘B’meansthatFidobarks,and‘C’meansthatthemaninthegabardinesuitisaspy.Inthiscase,theideaofanaxiomatizedsemantictheoryissimpletothepointoftriviality.Forwecanjusttakethreeaxioms,onestatingthemeaningofeachofthethreesentences,andthentheprojectofderivingameaning-specifyingtheoremforeachsentenceofthelanguagequiteliterallyinvolvestakingnostepsatall.Thesecondstagewouldinvolvesubject–predicatesentencesfromaformallanguage.Wemighthavethreepredicateterms,‘F’,‘G’and‘H’andthreesubjectterms,‘a’,‘b’and‘c’,allowingustoconstructninesubject–predicatesentences,amongwhich,letussuppose:‘Fa’meansthatTheaetetusissitting‘Gb’meansthatFidobarksand‘Hc’meansthatthemaninthegabardinesuitisaspy.Ifwewanttobeabletoderiveameaning-specifyingtheoremforeachsentenceinasystematicway,thenwemighthopetoformulateatheoryinwhichthereisanaxiomtalkingaboutthemeaningofeachsubjecttermandanaxiomtalkingaboutthemeaningofeachpredicateterm(ratheralongthelinesofAx1andAx2insection2above).Atthesefirsttwostages,thetotalnumberofsentencesunderconsiderationisfinite.Thethirdstagewouldinvolveaninfinitecollectionofsentencesbuiltfromsomefinitestockbyapplyingconstructionsoverandover.Wemighttakethethreesentences,‘A’,‘B’and‘C’,fromthefirststage,andallowtheapplicationofanegationoperator‘ÿ’.Thiswouldgivethreemoresentences,‘ÿA’,‘ÿB’,‘ÿC’(meaningthatTheaetetusisnotsitting,thatFidodoesnotbark,andthatthemaninthegabardinesuitisnotaspy,respectively),andthenthreemore,‘ÿÿA’,‘ÿÿB’,‘ÿÿC’,andthenthreemore,indefi-nitely.Alternatively,wecouldallowtheapplicationofaconjunctionoperatortotwosentencesatatimetogive‘A&B’,‘A&C’,‘B&C’and(ifwemarktheorderofthecon-juncts)also‘B&A’,‘C&A’,‘C&B’and(ifweallowrepetitionofconjuncts)‘A&A’,‘B&B’,‘C&C’(thus,ninenewsentencestoaddtotheoriginalthree).Then,applyingthecon-junctionoperatoragain,weget‘(A&B)&C’,‘(A&C)&C’,‘(B&C)&C’and‘(A&B)&(A&C)’,andmanymore–infact,144newsentencestoaddtothetwelve.Wecould,ofcourse,allowtheapplicationofboththenegationoperatorandtheconjunctionoperator,togiveussentenceslike‘A&ÿB’.Supposethatwedothat,andthentrytoprovideanaxiomatizedsemantictheoryforthissimple,butinfinite,collec-tionofsentences.Alongwithaxiomsstatingthemeaningsofourthreebuildingblocks,‘A’,‘B’and‘C’,weshallneedanaxiomsayingsomethingaboutthemeaningof‘ÿ’andonesayingsomethingaboutthemeaningof‘&’.Afirstthoughtinthisdirectionmightbetosuggestthesetwoaxioms:‘ÿ’meansnegation‘&’meansconjunction.100\nPHILOSOPHYOFLANGUAGEButitisquiteunclearhow,fromthisaxiomabout‘ÿ’,say,andtheaxiomstatingthemeaningof‘A’,wecouldderiveatheoremspecifyingthemeaningof‘ÿA’.Whatlooksmorepromising,forthenegationoperator,issomethinglike:IfSisanysentenceatall,thenthesentencemadeupbyputting‘ÿ’togetherwithSmeansthenegationofwhateverSmeans.Thisbuildsintothespecificationofthemeaningoftheoperatoranindicationthatitisindeedanoperatorthatcanbeappliedtoanysentencetoyieldanewsentence.Some-thingsimilarcanbedonefortheconjunctionoperator;butweshallnotpauseanylongeroverthesedetails.(Infact,thisalsosuggestshowwemightmakesomeprogresswiththeissueaboutthedifferencebetweensubjecttermsandpredicateterms,dis-cussedinsection2.1.Wecouldbuildintotheaxiomforapredicatetermanindicationthatitisindeedapredicate–anexpressionthatcangotogetherwithasubjecttermtomakeasentence.)Aformallanguagewithsubjecttermsandpredicateterms,plusacoupleofsentence-buildingoperators,isclearlyveryfarindeedfromtherichcomplexityofnaturallan-guage.But,apartfromthedifferenceofscale,thestepfromformallanguagestonaturallanguagesmightseemsimpleinprinciple.Ifweknowhowtoprovideanaxiomatizedsemantictheoryforaformallanguagewithsubjecttermsandpredicateterms,forexample,thenwealsoknowhowtoprovideatheoryforasimilarlystructuredfrag-mentofnaturallanguagethatusesrealwordsandphrasesinsteadofletters(‘Theaete-tus’insteadof‘a’,‘issitting’insteadof‘F’).However,inmanycases,therelationshipbetweensentencesofaformallanguageandsentencesofanaturallanguageismuchlessclear.Thereisatinyindicationofthisalreadyinthecaseofnegation,wheretheformallanguageoperatorattachesatthefrontofacompletesentencewhereasnaturallanguagenegationusuallyoccurswithinthepredicateterm(‘isnotsitting’,‘doesnotbark’).And,quitegenerally,thecomplex-ityoftherelationshipbetweenformalandnaturallanguagesisshownbythedifficultyofthetask(familiartomostphilosophystudents)ofregimentingintologicalnotationargumentsthatareexpressedinnaturallanguage.So,onequestionthathastobefacedisjusthowtoregardtherelationshipbetweenthesuperficialformsofnaturallanguagesentences,ontheonehand,andtheregimentedforms(sometimescalledlogicalforms;seeSainsbury1991)towhichanaxiomatizedsemantictheorycouldbeapplied,ontheotherhand.Evenifweignorethesedifferencesbetweenformallanguagesandnaturallan-guages,itisstillthecasethatwehavesofaronlyenvisagedasemantictheoryforasimplesubject–predicatelanguagewithsentenceoperators.Subsequentstagesoftheprojectofconstructingsemantictheorieswouldinvolvebringingfurthertypesofexpressionwithinthescopeoftheenterprise.Weshould,however,pauseheretonoticethat,whileitisclearthataxiomatizedsemantictheoriescouldbeofinteresttologicians,and–totheextentthattheydisplaythemeaningsofparticularkindsofexpressionssuchaspronouns,demonstratives,adjectivesandsoon–alsotolinguists,stillitmaynotbeevidentwhytheyshouldmeritaphilosopher’sattention.Infact,theprojectofconstructingcompositionalsemantictheorieshasbeenacentralconcerninrecentphilosophyoflanguage,andDavidson’s101\nMARTINDAVIESworkinparticularrevealsseveralreasonsforfocusingphilosophicalattentiononsemantictheories(Davidson1967a;andmanyotherpaperscollectedinDavidson1984).Weshallreviewfiveofthesereasons.4.1TheformatofsemantictheoriesThefirstreasonforpayingattentiontoaxiomatizedsemantictheoriesconcernstheproperformatforthetheoremsthatsuchatheoryissupposedtoyield.Sincethesethe-oremsaresupposedtobemeaning-specifying,theinitiallyobviousformatwouldbeonethatrelatesasentencetoitsmeaning.Ifmeaningsarethemselvesregardedasentitiesofsomekind,thenwemightexpectaformatlike:ThemeaningofsentenceSism.Ifmeaningsarenotregardedasentities,thenwemightexpectinstead:SentenceSmeansthatp.Thus,onthefirstalternative,ifmeaning-specifyingtheoremsareexplicitlytorelatesentencestopropositions,say,thenanexampleofsuchatheoremmightbe:Themeaningofthesentence‘Theaetetusissitting’isthepropositionthatTheaete-tusissitting.Onthesecondalternative,similarly,anexamplemightbe:Thesentence‘Theaetetusissitting’meansthatTheaetetusissitting.If,forsomegeneralreason,thoseformatshavetoberejected–iftheconceptofmeaningcannot,afterall,figureinthetheoremsofacompositionalsemantictheory–thenthatwouldbeasignificantmeta-semanticresult.Davidson(1967a)doesindeedrejectboththoseformats,andarguesthatthetargettheoremsofasemantictheoryshould,instead,exhibittheformat:SentenceSistrueifandonlyifp(where‘ifandonlyif’(‘iff’)expressesthematerialbiconditional).Sincetheoremslikethisspecifyconditionsunderwhichasentenceistrue,Davidsonissaidtofavourtruth-conditionalsemantics.Davidson’sargumentforthisconclusioncomesintwosteps.Thefirststepisintendedtoruleouttheideathat,toeachword,eachphraseandeachsentence,thereshouldbeassignedsomeentityasitsmeaning.Inthisstep,theso-calledFregeargument(Frege1892:62–5;cf.Quine1960:148–9)isusedtoshowthat,undercertainassumptions,alltruesentenceswouldbeassignedthesameentity.Clearly,nosuchundiscriminatingassignmentofentitiescouldbeanassignmentofmeanings,sinceitiscertainlynotcorrectthateverytruesentencehasthesamemeaning.(EssentiallythesamelineofargumentallowsDavidson(1969a)toconclude102\nPHILOSOPHYOFLANGUAGEthatthereisnopointinsayingthattruesentencescorrespondtofacts.TheFregeargu-mentshowsthat,undercertainassumptions,thereisonlyonefact.Forfurtherdiscussion,seeNeale1995.)TheFregeArgumentTheFregeargumentshowsthatifanequivalencerelation,E,onsentencesmeetstwoconditionsthenitclassifiesalltruesentencesasequivalent,andlikewiseallfalsesentences.Thetwoconditionsarethese.IftwosentencesarelogicallyequivalentthentheyareequivalentaccordingtoE.Iftwosentencesarethesameexceptforthefactthat,whereonecontainsasubjecttermi,theothercontainsanothersubjectterm,j,referringtothesamething,thentheyareequivalentaccordingtoE.Supposethatwetakeanytwotruesentences.Letuschoose,say,‘Penguinswaddle’and‘Theearthmoves’.Then,firstly,thesentence(1)Penguinswaddleislogicallyequivalentto(2)Thenumberthatis1ifpenguinswaddleandis0ifpenguinsdonotwaddleis1.So,bythefirstcondition,(1)and(2)areequivalentaccordingtoE.Similarly,(3)Theearthmovesand(4)Thenumberthatis1iftheearthmovesandis0iftheearthdoesnotmoveis1areequivalentaccordingtoE.Butsecondly,thesubjecttermin(2),namely‘Thenumberthatis1ifpenguinswaddleandis0ifpenguinsdonotwaddle’,referstothenumber1,sincepenguinsdoindeedwaddle.Andthesubjecttermin(4),namely‘Thenumberthatis1iftheearthmovesandis0iftheearthdoesnotmove’,alsoreferstothenumber1.So,sentences(2)and(4)arethesameexceptforthefactthat,whereonecontainsa(ratherlengthy)subjectterm,theothercontainsanother(equallylengthy)subjecttermreferringtothesamething.Bythesecondcondition,then,sentences(2)and(4)areequivalentaccordingtoE.Puttingallthistogether,(1)isequivalentto(2),and(2)to(4),and(4)to(3).So,(1)isequivalentto(3)accordingtoE;andthatiswhatweneededtoshow.ThesecondstepinDavidson’sargumentforthetruth-conditionalformatpointsoutthatthe‘meansthatp’constructionpresentslogicaldifficulties,sothattheformalderivationsofmeaning-specifyingtheoremswillbehighlyproblematic(seesection2.2).Incontrast,thetruth-conditionalformatislogicallywellunderstood.Thewayinwhichtheoremsspecifyingtruthconditionsforcompletesentencesarederivedfromaxiomsassigningsemanticpropertiestowordsandphrasescan,toaconsiderable103\nMARTINDAVIESextent,becarriedoverfromtheworkofAlfredTarski(1902–83)oncertainformallanguages(Tarski1944,1956).Thissecondstepintheargumentistechnicalincharacter,anditisamatterofdisputewhetheracompositionalsemantictheorymakingdirectuseofthe‘Smeansthatp’formatisfeasible.(SeeSchiffer1972:162;Taylor1982;Davies1984.)Asforthestatusofthefirststepintheargument,itisnowwidelyagreedthattherearereasonsforrejectingDavidson’suseoftheFregeargument.TheFregeargumentcanonlyestablishthatalltruesentenceswouldbeassignedthesameentityastheirmeaningifitislegitimatetosupposethattheequivalencerelationofhavingthesamemeaningwouldmeetthetwoconditionsthatfigureintheargument.Butthisisveryfarfrombeingobvious.So,pendingfurtherargumenttothecontrary,wecanadmitthatitwouldbepossibleforacompositionalsemantictheorytoworkbyassigninganentitytoeachsentenceasitsmeaningorsemanticvalue.Inparticular,onepossibilitywouldbetoassigntoeachsentenceacertainkindofstructuredentity–astateofaffairswithobjectsandpropertiesasconstituents.ThesituationsemanticsprogrammeofBarwiseandPerry(1983)constitutesonedevelopmentofthispossibility(seealsoTaylor1976;Forbes1989).4.2ThecorrectnessofsemantictheoriesThesecondreasonforphilosophicalinterestinsemantictheoriesrelatestothecondi-tionsofadequacyorcorrectnessonanysuchtheory.Whateveristherightformatforasemantictheory,weneedsomeaccountoftheconditionsunderwhichatheoryinthatformatisthecorrecttheoryforthelanguageofaparticularspeakerorgroupofspeakers.Weneedtoknow,forexample,whatmakesitcorrecttosaythat,inourlan-guage,theparticularstringofletters‘p’–‘e’–‘n’–‘g’–‘u’–‘i’–‘n’–‘s’...andsoonmeansthatpenguinswaddle.Inordertobeabletoconsiderapossiblereformulationofthisquestion,weneedtomakeuseofthenotionofanabstractorpossiblelanguage.Intheabstract,alanguagecanbeconsideredtobeacollectionofsentencestogetherwithastipulatedassignmentofmeaningstothosesentences.So,forexample,therearepossiblelanguagesinwhichthereisasentencemadeupofthosesameletters–‘Penguinswaddle’–butwiththemeaningthattheearthmoves.Giventhenotionofapossiblelanguage,thequestionwhetherasemantictheoryiscorrectforthelanguageofagivengroupofspeakerscanbereformulatedasthequestionwhetherthepossiblelanguageforwhichthesemantictheoryisstipulatedtobecorrectistheactuallanguageofagivengroup(Lewis1975;Peacocke1976;Schiffer1993).Whatissometimescalledtheactuallanguagerelationisthusarelationbetweenlanguages(intheabstract)andgroupsoflanguageusers.Underthereformulationthatweareenvisaging,conditionsofadequacyonsemantictheoriesbecomeconstraintsontheactuallanguagerelation.Anyphilosophicalelucidationofthekeysemanticconceptusedinsemantictheo-ries,suchasmeaningortruth,canbetransposedintoaconditionofadequacyonthosetheories(or,equivalently,intoaconstraintontheactuallanguagerelation).Thus,supposeforexamplethatanelucidationoftheconceptofmeaningsaysthatanysentenceShasmeaningminthelanguageofagroupGifandonlyifsomeconditionC(S,m,G)holds.Thiscanbetransposedintoaconditionofadequacyasfollows:104\nPHILOSOPHYOFLANGUAGEIfasemantictheoryforthelanguageofagroupGdeliversatheoremsayingthatthemeaningofsentenceSismthenitshouldbethecasethatC(S,m,G).Similarly,itcanbetransposedintoaconstraintontheactuallanguagerelation:ApossiblelanguageinwhichShasthemeaningmistheactuallanguageofagroupGonlyifC(S,m,G).Thiskindoftranspositioncanbecarriedoutintheoppositedirectiontoo.Anyconditionofadequacyonsemantictheories(oranyconstraintontheactuallanguagerelation)canhelpuselucidatethekeysemanticconceptusedinthosetheories.Thus,forexample,considersemantictheoriesthatadoptthetruth-conditionalformat:SentenceSistrueifandonlyifp.Tarski’s(1956)ConventionTimposesaconditionofadequacyonsuchtheories;namely,theconditionthatthesentencethatfillsthe‘p’placeshouldtranslate(orelsebetheverysamesentenceas)thesentenceS.Thisconditionofadequacyontruth-conditionalsemantictheoriesconstitutesapartialelucidationofthesemanticconceptoftruthintermsoftheconceptoftrans-lation.Intuitively,theconceptoftranslationisverycloselyrelatedtotheconceptofmeaning;andwhatConventionTrequires,ineffect,isthatthesentencethatfillsthe‘p’placeshouldhavethesamemeaningasthesentenceS.Ifatruth-conditionalseman-tictheorymeetsthatcondition,thenthetruthconditionspecificationsthatityieldsareguaranteedtobecorrect.Thus,ConventionT’selucidationofthesemanticconceptoftruthinvolvesaconnectionbetweenthatconceptandtheconceptofmeaning:IfasentenceSmeansthatpthenSistrueiffp.ButConventionTprovidesnofurtherhelpwiththeconceptofmeaningitself.Onewaytoshedfurtherlightontheconceptofmeaning–andso,viaConventionT,ontheconceptoftruth–wouldbetospelloutotherconditionsofadequacyonspec-ificationsofmeanings(oroftruthconditions).Theconceptforwhichweseekelucida-tionhereistheconceptofmeaning(ortruthconditions)inthelanguageofanygroup.TheconceptofmeaningisthesamewhetherweconsideragroupofEnglishspeakersoragroupofChinesespeakers;soweexpectthataconditionofadequacyshouldrelatetogroupsinaverygeneralway.Onequitegeneralthingthatwecansayaboutspecifi-cationsofmeaningisthattheyhelpustodescribemembersofagroupasengaginginlinguisticacts.Thetheoremsofasemantictheoryforthelanguageofagroup,G,canlicensetheredescriptionofutterancesofsentencesbyamemberofGasactsofsayingorassertion.Forexample,ifthesemantictheorysaysthatasentenceSmeansthatTheaetetusissitting,thenwemightreasonablyconstrueanuttererofSassayingorassertingthatTheaetetusissitting.Construingaperson’sutterancesasparticularlin-guisticactsisoneaspectofinterpretation,andwhatwehavejustseenisthatwecanmakealinkbetweenthetheoremsofasemantictheoryandtheprojectofinterpretingthemembersofalanguagecommunity.105\nMARTINDAVIESAsemantictheorycanplayaroleinfacilitatinganoverallinterpretationofthebehaviourofmembersofagroup,byofferingawayofunderstandingtheirspecificallylinguisticbehaviour.Iftherearelegitimateconstraintsonwaysofinterpretingpeople,thenthesemayyieldconstraintsontheparticularaspectsofinterpretationthatarelicensedbyasemantictheory.Inshort,theneedsofinterpretationmayleadustocon-ditionsofadequacyonmeaningspecifications,ortruthconditionspecifications,andsotoelucidationsoftheconceptsofmeaningandtruth.InDavidson’sprogramme,thisideaofalinkbetweensemantictheoriesandinterpretationisimplementedbydescribingtheconstraintsontheprojectofradicalinterpretation.Thisistheimaginaryprojectofconstructinganoverallschemeofinterpretationforlanguageusersaboutwhomweknownothingattheoutset.Oneputativeconstraintonradicalinterpretationisthatspeakersshouldbesointerpretedthatwhattheysayabouttheworld–and,presumably,whattheybelieveabouttheworld–turnsouttobebyandlargecorrect.ThisisthePrincipleofCharity(seeDavidson1967a,1973).Thus,forexample(Davidson1984:169):Atheoryofinterpretationcannotbecorrectthatmakesamanassenttoverymanyfalsesentences:itmustbegenerallythecasethatasentenceistruewhenaspeakerholdsittobe.ThePrincipleofCharityhassignificanceoutsidephilosophyoflanguage,sinceitappearstooffersomeprospectforANTI-SCEPTICALARGUMENTS(pp.51–6),inparticularforargumentstotheconclusionthatmostofwhatweourselvessayandthinkiscorrect(Davidson1983).ButinlaterworkbyDavidsonandothers(Grandy1973;McDowell1976;Wiggins1980)thePrincipleofCharityhasgivenwaytoaprinciplethatfocusesonintelligibil-ity,ratherthanoncorrectness:speakersshouldbesointerpretedthatwhattheysayandbelieveabouttheworldturnsouttobebyandlargereasonableorintelligible.AsDavidson(1984:xvii)himselfsays:‘Theaimofinterpretationisnotagreementbutunderstanding’.Sometimesaspeaker’sbeingwrongisquiteunderstandable,whilebeingrightwouldbealmostmiraculous.Aneloquentstatementofthisprinciple–sometimescalledthePrincipleofHumanity–isprovidedbyWiggins(1980:199):Letusthenconstrainthetheory...thatprovidessentencebysentenceinterpretationsofthelanguageLbytherequirementthat[it]shouldcombinewithaplausibleanthropology...insuchawaythatinconcertthetwotheoriesmakethebestsensepossible...ofthetotallifeandconductofL-speakers.Ifthesemantictheorythatprovidessentencebysentenceinterpretationsweretomakeuseofthe‘Smeansthatp’format,thenthePrincipleofHumanitycouldbeconceivedasaconditionofadequacyonthatsemantictheory,roughlyalongthefollowinglines:IfthesemantictheorydeliversatheoremsayingthatSmeansthatpthenitshouldbethecasethatinterpretingutterancesofSasexpressionsofthepropositionthatpcontributestomakingthebestsensepossibleofthetotallifeandconductofspeakers.106\nPHILOSOPHYOFLANGUAGEThisconditionofadequacywouldhelpustoelucidatetheconceptofmeaningandthence,viathemeaning–truthconnection,thenotionoftruth.ButitisalsopossibletoregardthePrincipleofHumanityasprovidingdirectelucidationoftheconceptoftruth,withoutexplicitlygoingviatheconceptofmeaning.Onewaytodothisistointroduceapurelyformalpredicate‘T’thatappliestosentencesandthentoconsidertheoriesthatdelivertheoremsoftheform:SentenceSisTifandonlyifp.Wecanimposeaconditionofinterpretationaladequacyonsuchatheory,roughlyalongthelinesof:Ifthesemantictheory(canonically)deliversatheoremsayingthatSisTiffpthenitshouldbethecasethatinterpretingutterancesofSasexpressionsoftheproposi-tionthatpcontributestomakingthebestsensepossibleofthetotallifeandconductofspeakers.(WeneedtherestrictiontocanonicalproofsofT-conditionspecifyingtheoremsbecause,foranygivensentenceS,atheorywilldelivermanydifferenttheoremsoftherequiredform:seeDavies1981:ch.2.)Andthenwecanseethatifatheorymeetsthiscondi-tionofadequacyitspredicate‘T’willapplytopreciselythetruesentencesofthelan-guageunderstudy.Wiggins(1980)regardsthisasshowingthewaytoa‘substantialtheoryoftruth’.Truthsimplyisthepropertythatplaysthe‘T’-rolespecifiedbythecon-ditionofadequacythatthePrincipleofHumanityfurnishes(seealsoMcDowell1976:section1;Wiggins1997).4.3SemantictheoriesandmentalstatesSomeaspectsofnaturallanguagesposeparticularproblemsforasemantictheoristworkingwithintheDavidsonianframework.Oneoftheseaspectsistheuseofnaturallanguagetoreportonpeople’smentalstates.Soathirdreasonforbeinginterestedinsemantictheoriesisthatreflectiononthesetheoriesmayyieldinsightintothemean-ingsofsentencesaboutmentalstates,andso,also,insightintothenatureofthosementalstatesthemselves.Toseehowtheproblemsarise,considerabeliefreportsentencelike:(BRep)TheaetetusbelievesthatFidobarks.Intuitively,themeaningofthissentencedependsonthemeaningofthename‘Thea-etetus’,themeaningoftheverb‘believes’,andthemeaningofthecontainedsentence‘Fidobarks’.Sowemighthavesomeexpectationsaboutthewayinwhichameaning-specifyingtheoremfor(BRep)wouldbederivedinacompositionalsemantictheory.Wemightexpectthatthederivationwouldinvolve,firstly,provingameaning-specifyingtheoremforthecontainedsentence‘Fidobarks’,andthenusingthat,plusaxiomsaboutthemeaningsof‘Theaetetus’and‘believes’,inordertoprovethetargettheoremforthebeliefreportsentence.107\nMARTINDAVIESIntheframeworkoftruth-conditionalsemantictheories,similarly,wemightexpectthattheproofofatruth-condition-specifyingtheoremforthebeliefreportsentencewouldmakeuseofatheoremfor‘Fidobarks’plusaxiomsabout‘Theaetetus’and‘believes’.Butwecansee–atleast,inanimpressionistickindofway–whythisisliabletobeproblematic.Statementsaboutpeople’smentalstatesdisplaysomeofthesamelogicalpropertiesasstatementsaboutmeaning(seetheboxinsection2).Withstatementsaboutmeaning,itisnotusuallythoughttobethecasethatfrom:Fido=Roverplus:SentenceSmeansthatFidobarksitfollowsthat:SentenceSmeansthatRoverbarks.Thislogicalbehaviourisadeparturefromstandardlogicalsystems,anditraisesaques-tionaboutthelogicalresourcesthatwouldbeneededinasemantictheoryusingthe‘SentenceSmeansthatp’format.Italsohelpstomotivateuseofthealternative‘Sen-tenceSistrueiffp’format,sincethatislogicallywellbehaved.Thus,forexample,from:Fido=Roverplus:SentenceSistrueiffFidobarksitcertainlydoesfollowthat:SentenceSistrueiffRoverbarks.Butwhilethiswell-understoodlogicalbehaviourisoneoftheattractionsoftruth-conditionalsemantictheories,italsomakesithardfortruth-conditionaltheoriestogiveanadequatetreatmentofsentencesthatreportonpeople’smentalstates.Withstatementsaboutbelief,itisnotusuallythoughttobethecasethatfrom:Fido=Roverplus:TheaetetusbelievesthatFidobarksitfollowsthat:TheaetetusbelievesthatRoverbarks.108\nPHILOSOPHYOFLANGUAGEThefirstbeliefreportmightbetrue,whilethesecondwasfalse,forexample.Thisisoneaspectofwhatiscalledtheintensionalityofbeliefreports.Theproblemposedfortruth-conditionalsemanticsisthatitishardtoseehowatruth-condition-specifyingtheoremforthesentence:(BRep)TheaetetusbelievesthatFidobarkscanbederivedinthewaythatweenvisaged,withoutalsoinvolvingthedenialoftheaspectoftheintensionalityofbeliefreportsthatwejusthighlighted.Thederivationalroutethatweenvisagedgoesviaatruth-condition-specifyingtheoremforthecontainedsentence‘Fidobarks’,totheeffectthatthecontainedsentenceistrueiffFidodoesindeedbark.Giventheassumptionthat:Fido=Roverwecanmovefromthistheorem:(FF)‘Fidobarks’istrueiffFidobarksto:(FR)‘Fidobarks’istrueiffRoverbarks.Butnow,considerwhateverrouteleadsfromtheoriginaltheorem(FF)toatruth-condition-specifyingtheoremforthebeliefreportsentence(BRep),totheeffectthatitistrueiffTheaetetusdoesindeedbelievethatFidobarks.GiventheassumptionthatFido=Rover,anexactlysimilarroutewillalsoleadvia(FR)totheconclusionthatthesamesentence(BRep)istrueiffTheaetetusbelievesthatRoverbarks.So,giventheassumptionthat:Fido=Roverwearedriventotheoverallconclusionthat:TheaetetusbelievesthatFidobarksiffTheaetetusbelievesthatRoverbarks.Butthisgoesflatlyagainstthepresumedintensionalityofbeliefreports.Theissuesherearecontroversial,andtherehavecertainlybeenproposalsforbring-ingwithinthecompassoftruth-conditionalsemanticssentencesthatareusuallyreck-onedtoexhibitintensionality(seeDavidson1969b).Fornow,however,itisenoughtonoticethatattentiontocompositionalsemantictheoriescouldbringquestionsaboutmentalstatesintosharperfocus.4.4SemantictheoriesandepistemologyThefourthreasonfordirectingphilosophicalattentionontotheconstructionofseman-tictheoriesisepistemological.Wehaveindicatedthewayinwhichafinitesetofaxioms109\nMARTINDAVIEScanpermitthederivationofameaning-specifyingtheoremforeachofinfinitelymanysentences.Sointhecaseofagivenlanguagewithinfinitelymanysentences,someonewhoknewthefactsstatedbythefinitelymanyaxiomsofacompositionalsemantictheorywouldtherebybeinapositiontocometoknowwhateachsentenceofthelanguagemeans.Theexistenceoffinitelyaxiomatizedsemantictheoriesthusshedslightonanepistemologicalproblem:howisitpossibleforafinitebeingtomasteraninfinitelanguage?Sofarasthepurelyformalnotionofanaxiomatizedtheorygoes,thesetofaxiomsofatheorydoesnothavetobefinite.ButDavidson(1965)usestherequirementthatalanguageshouldbelearnable(andlearnablebyafinitebeing)tomotivateafiniteaxiomatizationconstraintonsemantictheories.SinceDavidsonfavourstruth-conditionalsemantictheories,thisyieldstheideathattheaimofasemantictheorististoconstructfinitelyaxiomatizedtheoriesoftruthconditionsfor(regimentedfrag-mentsof)naturallanguages–building,sofaraspossible,ontheworkofTarski.ThesortsoftheoriesenvisagedaresometimesspokenofasTarski–Davidsontruththeories,andthereisaveryconsiderablebodyofwork–muchofitquitetechnicalincharacter–inpursuitofthisaim.Weshouldnote,however,thatthereissomedisputeoverthefiniteaxiomatizationconstraint.Atleastoneprominentphilosopheroflanguage(Schiffer1987:ch.7)arguesthatitmaybepossibleforaspeakertomasteralanguage,eventhoughnofinitelyaxiomatizedtruth-conditionalsemantictheorycanbeprovidedforit.Evenifthereisafinitelyaxiomatizedsemantictheoryforalanguage,thereisstillapressingquestionabouttheexplanatoryrelevanceoftheexistenceofsuchatheory.Knowledgeofthefactsstatedbytheaxiomsofacompositionalsemantictheorywouldsuffice,inprinciple,forknowledgeastowhateachsentenceofthelanguagemeans.Butordinaryspeakersofanaturallanguageusuallylackconsciousknowledgeofanycompositionalsemantictheoryfortheirownlanguage.Sothequestionthatpressesishowthemereexistenceofafinitelyaxiomatizedsemantictheory,unknowntoordinarylanguageusers,cansolvetheepistemologicalproblemthattheyapparentlyface.Oneapproachtoansweringthisquestionbeginsbyconsideringawiderrangeofknowledgethatordinarylanguageusershave.Alongwithknowingwhatthesentencesoftheirlanguagemean,ordinaryspeakersofEnglishalsoknow,forexample,thatinthesentence:Nigelshavedhimtheword‘him’cannotreferbacktoNigel,whileinthesentence:NigelwantedBrucetoshavehimtheword‘him’canreferbacktoNigel.Theyknowthatthesentences:LessthantwobookswereonthetableMorethanonebookwasonthetable110\nPHILOSOPHYOFLANGUAGEsoundfine,while:LessthantwobookswasonthetableMorethanonebookwereonthetablesoundsomehowwrong–despitethefactthat‘Lessthantwobooks’isequivalentto‘Atmostonebook’,and‘Morethanonebook’isequivalentto‘Atleasttwobooks’.Andwecouldmultiplyexamplesofspeakers’linguisticknowledgeindefinitely.Ordinarylanguageusershaveconsciousaccesstothesepiecesofknowledge.Buttheydonothaveconsciousknowledgeofanysetofaxiomsorrulesfromwhichtheseparticularpiecesofknowledgefollow.Theoreticallinguistsarticulatesetsoflinguistic(especially,syntactic)rulesorprinciplescalledgrammars.Buttheexplanatoryclaimsoftheoreticallinguisticsdonotendwiththemereexistenceofgrammarsfromwhichparticularpiecesoflinguisticknowledgecouldbederived.Inafamouspassage,NoamChomsky(b.1928)says:Obviously,everyspeakerofalanguagehasmasteredandinternalizedagenerativegrammarthatexpresseshisknowledgeofhislanguage.Thisisnottosaythatheisawareoftherulesofthegrammaroreventhathecanbecomeawareofthem...Anyinterest-inggenerativegrammarwillbedealing,forthemostpart,withmentalprocessesthatarefarbeyondthelevelofactualorevenpotentialconsciousness.(Chomsky1965:8)AccordingtoChomsky(1965,1986,1995),then,ordinarylanguageuserspossessabodyoflinguisticknowledgewhichis,forthemostpart,inaccessibletoconsciousness.Forthisreason,itisoftenspokenofastacitknowledge.Itisbecausetheyhavethisbodyoftacitknowledgeoflinguisticrulesthattheyareabletoknowthevasthostofpar-ticularthingsabouttheirlanguage.WhileChomskyisconcernedprimarilywithknowledgeofsyntax,itseemsthattheideathatordinarylanguageuserspossessabodyoftacitknowledgecouldalsobeappliedintheareaofsemantics.Thus,onewayofansweringthequestionabouttheexplanatoryrelevanceoftheexistenceofcompositionalsemantictheorieswouldbetocreditordinarylanguageuserswithtacitknowledgeoftheaxiomsofsuchatheory(seebelow,section5.3).Somephilosophersoflanguagearecriticalofanyappealtothenotionoftacitknowledge,claimingthatitembodiesconceptualconfusions(BakerandHacker1984).Butaccountsofthenotionoftacitknowledge,particularlyasitappliestosemantictheories,havebeenoffered(seeEvans1981a;Davies1986,1987,1989;Peacocke1986,1989),andtheproposalthatordinarylanguageusershavetacitknowledgeoftheaxiomsoftruth-conditionalsemantictheorieshasbeendevelopedinsomedetail(seeHigginbotham1985,1986,1988,1989a,1989b).Weshouldnote,however,thatthereareotherproposalsforansweringtheepiste-mologicalquestionaboutcompositionalsemantictheoriesandordinarylanguageusers’knowledgeaboutthemeaningsofcompletesentences.Forexample,Dummett(1976,1991)makesuseofanotionofimplicitknowledgethatisimportantlydifferentfromChomsky’sideaoftacitknowledge,andWright(1986)seestheconstructionofcompositionalsemantictheories,notasanarticulationoflanguageusers’actualknowledge,butasamatterofRATIONALRECONSTRUCTION(p.386).111\nMARTINDAVIES4.5SemantictheoriesandmetaphysicsThefifthreasonforphilosophicalinterestinsemantictheoriesisthatthetaskofbring-ingspecificlinguisticconstructionswithinthescopeofacompositionalsemantictheorysometimesshedslightonissuesinotherareasofphilosophy,andparticularlyinmetaphysics.Wehavealreadydiscussedtherelationshipsbetweenphilosophyoflanguage,philosophyofmindandmetaphysics,soweshallberatherbriefhere.Semantictheoriesfornaturallanguageshavetodealwithpronouns(‘I’,‘you’,‘she’,‘he’,‘it’,‘they’),demonstratives(‘this’,‘that’,‘thisknife’,‘thatbutter’)andotherindex-icalexpressions(‘now’,‘then’,‘today’,‘tomorrow’,‘here’,‘there’).Questionsaboutthesemanticcontributionsoftheseexpressions(seeKaplan1989)inevitablyhighlightissuesinontology,aswellasinphilosophyofmind.For,whenweconsiderthementalstatesthataretypicallyexpressedusingtheseexpressions,wefindthattheyarementalstatesthatinvolvethinkingaboutpersons,things,timesandplacesinparticularways.Soanadequateaccountofthethoughtsexpressedbytheuseoftheseexpressionsseemstoinvolveacommitmenttotalkingaboutwaysofthinkingor,inFregeanterminology,aboutSENSES(pp.793–6)(seeFrege1918;Perry1977,1979;Evans1981b,1982;Peacocke1981,1983,1986,1992:ch.4;Forbes1987;Davies1982).Anotherexampleofalinkbetweenrecentworkinphilosophyoflanguageandissuesinmetaphysics–thistimenotmediatedbyphilosophyofmind–isprovidedbyworkonthesemanticsofnames(‘Theaetetus’),definitedescriptions(‘themaninthegabar-dinesuit’)andmodaladverbs(‘necessarily’),ontheonehand,andmetaphysicalissuesaboutnecessityandessentialism,ontheother(Kripke1980;Wiggins1976).Akeyclaiminthisareahasbeenthatnamesare,whiledefinitedescriptionstypicallyarenot,rigiddesignators.Asubjecttermissaidtobearigiddesignatorifitdesignatesthesameobjectwithrespecttoalldifferentpossiblesituations,solongasweholdthemeaningofthetermconstant.Toseetheimportanceofthenotionofrigiddesignationtoquestionsaboutessen-tialism,considerthesentence:Necessarily,thetallestanimalinthecourtyardishuman.Tosetthescene,supposethatTheaetetusisthetallestofagroupofpeopletalkingtogetherinthecourtyard,andthattherearenonon-humananimalsinthecourtyardotherthanadog,Fido,andsomefleas.Now,whetherthesentencecontainingthemodaladverb‘Necessarily’istruedependsonwhetherthetruthofthesubject–predicatesentence:Thetallestanimalinthecourtyardishumanisamatterofnecessityorcontingency.Thissubject–predicatesentenceistrueasevalu-atedwithrespecttothedescribedsituation;butcouldithavebeenfalse,withoutanychangeinitsmeaning?Clearly,theansweristhatthesentencecouldeasilyhavebeenfalse.IfTheaetetusandhisfriendshadnotbeeninthecourtyard,thenFidowouldhavebeenthetallestanimal,andFidoisnothuman.Sothequestionoverthemodalstatusofthe112\nPHILOSOPHYOFLANGUAGEsubject–predicatesentence(whetheritisnecessarilyorcontingentlytrue)doesnotturnonanyinterestingmetaphysicalquestionaboutwhetherhumanbeingsareessentially,ormerelycontingently,human,forexample.Nowconsiderthesentence:Necessarily,Theaetetusishuman.Whetherthissentenceistruedependsonwhetherthetruthofthesubject–predicatesentence:Theaetetusishumanisamatterofnecessityorcontingency;andthistimethequestionofmodalstatusdoesseemtoturnonaquestionaboutessentialism.Mightthisparticularhumanbeing,Theaetetus,havebeenotherthanhuman?Thedefinitedescription‘Thetallestanimalinthecourtyard’isanon-rigiddesigna-tor.Withrespecttothesituationasdescribed,itdesignatesTheaetetus,whilewithrespecttoanalternativepossiblesituation,itdesignatesFido–andwithoutanydiffer-enceinthemeaningsofthewords(‘tallest’,‘animal’andsoon)thatmakeupthedef-initedescription.Incontrast,thename‘Theaetetus’isarigiddesignator.Withrespecttothesituationdescribed,orwithrespecttotheimaginedalternative,thatnamedes-ignatesTheaetetus–solongasitretainsitsmeaning.Whatwehavejustseenisthat,wherethesubjecttermisarigiddesignator,aquestionaboutthetruthofasentencecontainingamodaladverbleadsdirectlytoaquestionabouttheessentialpropertiesofaparticularobject.Davidson’sownworkprovidesathirdexampleofalinkbetweensemantictheoriesandissuesinmetaphysics–particularlyofthewayinwhichtheconstructionofasemantictheorymayrevealthemetaphysicalcommitmentsoftheconceptualschemethatisexpressedinanaturallanguage.Sentencesthatcontainactionverbsplusadverbs,suchas:Johnbutteredthetoastslowly,inthebathroomposeachallengeforsemantictheoriesbecause,afterwedealwiththesubject(‘John’)andthepredicate(‘butteredthetoast’),itisunclearhowweshouldcopewiththeadverbs.Weshallnotpauseoverthedetailsofthechallengethatadverbspose,butsimplynotethatDavidson(1967b,1969c,1985;seealsoWiggins1986)arguesthat,inordertobringthesesentenceswithinthescopeofatruth-conditionalsemantictheory,weneedtoregimentthemasinvolvingquantificationoverevents.Roughly,theideaisthatanadverbiallymodifiedactionsentenceshouldberegimentedinlogicalnotationasbeginningwith‘($e)–’(inEnglish:‘thereisaneventwhich–’)sothat,forexample,thesentenceaboutJohnandthetoastisregimentedasequivalentto:ThereisaneventewhichisabutteringofthetoastbyJohn,isslow,andis(thatis,occurred)inthebathroom.113\nMARTINDAVIESIfthisiscorrect,thenourordinarytalkaboutpeopledoingthings,likebutteringtoast,carriesmetaphysicalcommitments,notonlytomaterialthingslikepeopleandtoast,butalsotoanontologyofevents.AdverbsandEventsDavidson’sproposalleadstoquestionsaboutthenotionoflogicalform,aboutalter-nativemetaphysicaltheoriesofevents,andabouttheroleofmaterialbodiesasthebasicparticularsinourconceptualscheme(Strawson1959:ch.1).Firstly,theproposalcanbemotivatedineitheroftwoways,correspondingtotworatherdifferentideasoflogicalform(Evans1976).Ontheonehand,thereisapatternofinferencetobeexplained.From‘Johnbutteredthetoastslowly,inthebathroom’wecanvalidlyinfereachof:Johnbutteredthetoastslowly.Johnbutteredthetoastinthebathroom.Johnbutteredthetoast.Iftheadverbiallymodifiedsentenceisregimentedasinvolvingaconjunction(‘eisabutteringofthetoastbyJohn&eisslow&eoccurredinthebathroom’)thentheseinferencescanbeseenasvalidinvirtueofthelogicalpropertiesof‘&’(Davidson1967b).Ontheotherhand,theregimentationmightbeseenasrequiredsimplyinordertobringthesentenceswithinthescopeofacompositionaltruth-conditionalsemantictheory(Evans1976).Thismotivationmightnotseemveryconvincing,sinceanalternativeschemeofregimentation(involvingoperatorsthatmodifythepredicatetermscorrespondingtonaturallanguageverbs)appearstobeavailable(Strawson1974b).Buttheapparentalternativeisnot,intheend,formallyviable(Taylor1985;Davies1991).Secondly,theproposalleadsnaturallytothequestionwhetherthedetailedsemanticbehaviourofadverbsmightprovideuswithgroundsforchoosingbetweentwoconceptionsofevents.Ontheonehand,Davidson(1967b,1969c)regardseventsasparticularsthatcanbedescribedinmanydifferentways.Oneandthesameeventmightbebothapullingofatrigger,afiringofarifleandakillingofapresident,forexample.Ontheotherhand,anumberoftheorists(Goldman1971;Kim1976;Taylor1985)proposethateventsshouldbemuchmorefinelydiscriminated–withfirings,pullingsandkillingsdistinguishedfromeachother.Thereisasubstantialmetaphysicaldifferencebetweentheseaccounts,butitturnsoutthatpurelysemanticconsiderationsdonotprovideanycompellingreasonforfavouringoneovertheother(Davies1991).Thirdly,theproposalraisesquestionsaboutanargumentforStrawson’sthesisthatmaterialbodiesare,whileeventsandprocessesarenot,basicparticularsinourconcep-tualscheme.Partofwhatthisthesiscomestoisthatourabilitytopickoutindividualeventsinourthoughtandtalkdependsonourabilitytopickoutindividualmaterialobjects,butnotviceversa.Atonepoint,Strawson’sargumentturnscruciallyontheclaimthatourmostbasicunderstandingoftheconceptuallytruesentence‘Everyanimalwasborn’involvespossessionoftheconceptofbeingborn,butnotpossessionoftheconceptofabirthasanindividualparticular.Clearly,thisargumentneedstobeassessedinthelightofDavidson’sproposalforregimentingsuchsentencesas‘Thisanimalwasbornquickly,inthebarn’.114\nPHILOSOPHYOFLANGUAGE5AnalysingtheConceptofMeaning:Grice’sProgrammeInthelastsectionweconsideredfivewaysinwhichsemantictheoriesmightrewardphilosophicalattention.Thephilosophicalsignificanceofmeta-semantictheoriesismuchmoreobvious;fortheyareattemptsatanimmediatelyrecognizablephilosophi-calproject–aprojectofconceptualelucidation,orevenconceptualanalysis.Thus,forexample,theprogrammebegunbyGrice(1957,1968,1969)aimsatananalysisoftheconceptofliteralmeaninginapubliclanguageintermsofpsychologicalconceptssuchasintentionandbelief.InthissectionweshallgiveanoutlinedescriptionofGrice’sprogrammeforanalysingthenotionofliteralmeaning.(Grice’spapersontheanalysisoftheconceptofmeaningarecollectedinGrice1989.)Theanalyticalprojectcanberegardedashavingtwostages.Thefirststageaimstocharacterizeaconceptofspeaker’smeaningthatcorresponds,roughly,totheideaofcommunicating,orattemptingtocommunicate,aproposition(tryingtogetacrossamessage).Thesecondstagethenaimstousethatconceptofspeaker’smeaning,alongwiththenotionofaconventionalregularity,toconstructananalysisoftheconceptofliterallinguisticmeaning.Thebasicideaisthatasentencehasasitsliteralmeaningthepropositionthatitisconventionallyusedtocommunicate.Sincetheprojectisananalyticalone,itisnaturallyregardedasbeingsubjecttoarequirementofnon-circularity.Theaccountsofspeaker’smeaningandofconven-tionalregularity–notionsthataredrawnonintheanalysisofliterallinguisticmeaning–shouldnotthemselvesmakeessentialappealtolinguisticnotions.Thus,Grice’sprogrammepresumesthatmindisanalyticallypriortolanguage.ButthesuccessofGrice’sprogrammeisnotsimilarlyconditionalonanyontologicalorepiste-mologicalpriorityofmindoverlanguage(Avramides1989;cf.EvansandMcDowell1976:xv–xxiii).Aswementionedinsection3.1,itislogicallyconsistenttopursueGrice’sanalyticalprogrammewhileacceptingDavidson’sclaimsaboutnoontologicalpriority,andnoepistemologicalpriority,ofmindoverlanguage(thoughitistruethatitwouldbesurprisingforsomeonetocombineGrice’sanalyticalviewswiththenoontologicalpriorityclaim–nothoughtwithouttalk).Theclaimofnoepistemologicalpriorityofmindoverlanguagesaysthatitisnotpossibletofindoutaboutaperson’smentalstatesfirst,andonlythentogoontoinvestigatethemeaningsoftheperson’sutterances.Rather,theassignmentofbothmeaningsandmentalstates,suchasbeliefsandintentions,isasingleintegratedepistemologicalprojectgovernedbythePrincipleofCharityorthePrincipleofHuman-ity(section4.2).ButDavidsongoesbeyondthisepistemologicalclaim.Hisviewisthat,byspellingouttheconstraintsoninterpretation,weprovideallthephilosophicalelu-cidationthatcanbeprovidedoftheconceptoflinguisticmeaning.Theconstraintsoninterpretationareoftwokinds.Ontheformalside,thereis,forexample,thefiniteaxiomatizationconstraintonthesemantictheorythatisused(inconcertwithaplau-sibleanthropology)forinterpretation.Ontheempiricalside,therearetheconstraintsofagreement(Charity)orintelligibility(Humanity).Buttheproblemwiththeseempiricalconstraintsisthattheyseemtoleavealacuna.Aninterpretationdeploysasemantictheory,whichgeneratesanassignmentofmeaningstosentences,inordertoprovideadescriptionoflanguageusersasengaging115\nMARTINDAVIESincertainspeechacts(sayingandassertingthings,forexample)andashavingcertainpropositionalattitudes(believingandintendingthings,forexample).ThePrincipleofHumanityconstrainsthespecificationsofmeaningbyimposingtherequirementthattheresultingoveralldescriptionofthelanguageusersintermsofmeanings,speechactsandpropositionalattitudesshouldmakethemouttobereasonableorintelligible.ButthePrincipleofHumanitydoesnotitselftelluswhichcombinationsofmeanings,speechactsandpropositionalattitudescanbeintelligiblyattributed.Onthefaceofit,anaccountofwhichcombinationsarecoherentwouldbeprovidedbyarticulatingtheanalyticalconnectionsbetweentheconceptofmeaning,thecon-ceptsofvariousspeechactslikesayingandasserting,andtheconceptsofpropositionalattitudeslikebelievingandintending.Theremight,forexample,beconceptualcon-nectionsthatrequirethatanyonewhoassertsthatpdoessobyusingasentencethatliterallymeansthatp,andthatanyonewhoassertsthatpintendsanaudiencetotakehimorher(thespeaker)tobelievethatp.Whetherthereareconnectionslikethis,andifso,whatexactlytheyare,isnotatrivialquestion;itissomethingthatrequiresdetailedinvestigation.TheboldproposalofGrice’sanalyticalprogrammeisthatthereareconnectionsofthiskindthatactuallypermittheanalysisoftheconceptoflinguisticmeaning(andtheconceptsofthevariousspeechacts;Schiffer1972:ch.4)intermsofpropositionalattitudes.Grice’sanalyticalprojectbeginsfromadistinctionbetweennaturalandnon-naturalmeaning(Grice1989:214).Wearetalkingaboutnaturalmeaning(orindicatormeaning)whenwesay,forexample:Thosespotsmean(indicate)measles.Thosecloudsmean(indicate)rain.Thisnotionofmeaning–whichisjustamatterofcausalco-variationbetweentwokindsofstatesofaffairs–isnottheconcernofGrice’sprogramme,thoughittakesonconsiderableimportanceinrecentworkinphilosophyofmind,aimedatprovidinganaccountofthecontentsofpsychologicalstatesthemselves(seeDretske1981,1986).Spotsindicatemeaslesandcloudsindicaterainwhetherornotanyonetakestheminthatway,andwhetherornotanyoneintendsthemtobetakeninthatway.ThevariousnotionsofmeaningthatGriceproposestoanalysearedistinguishedfromnaturalmeaningbythefactthattheyinvolve,inonewayoranother,theintentionsofaperpetratororutterer.Ifweareaimingatananalysisofthenotion:SentenceSmeansthatpinthelanguageofgroupGthenafirstsuggestionmightbethis:MembersofGusethesentenceStosay(toeachother)thatp.Thereareseveralreasonswhythiswouldnotbeadequate,buttheonethatconcernsusnowisthattheconceptofsayingseemsalreadytodependontheconceptofliteralmeaning.Inordertosaysomething(forexample,tosaythatpenguinswaddle),a116\nPHILOSOPHYOFLANGUAGEspeakerhastouseanexpressionthatliterallymeansthatthing(here,asentencethatliterallymeansthatpenguinswaddle).Soanyanalysisalongtheselineswouldbecircular.Itmightberepliedtothisthatthereissomethingstipulativeaboutthisconceptionofsaying–thatisnotobligatorytoequatesayingwithstrictlyandliterallysaying.Itmightalsoberepliedthatperhaps,intheend,weshallhavetobesatisfiedwithananalysisthatiscircular–acircularanalysismightbeelucidatoryprovidedthatthecircleofinter-definedtermsislargeenough.Thereissomejusticeinboththesereplies.Butstill,asecondsuggestionforananalysisoftheconceptofliteralmeaningmightbethis:MembersofGusethesentenceStocommunicatethemessagethatp.Theadvantageofthissecondsuggestionisthattheconceptofcommunicating(orgettingacross)amessagedoesnotseemtoreintroducetheconceptofliteralmeaning.Apersoncangetacrossamessageeventhoughthesoundsusedhavenoliteralmeaningatall.However,thereisstillaproblemwiththissecondsuggestion.Asoundthathasnoliteralmeaningcanbeusedbyonememberofagrouptocommunicateamessagetoanother.Theperpetratorofthesound,realizingthatthesoundhasnoliteralmeaning,mightrelyonsomeresemblancebetweenthesoundproducedandsomeothersound,forexample,thesoundofanangrydog(seeSchiffer1972:7,forthisexample).Butthen,thesamesoundmightbeusedbyeachmemberofthegroupwhenheorshewantstogetacrossthemessagethatheorsheisangry.Sotheconditionsinthesuggestedanalysiswouldbemet.Butthisstillmightnotbeacaseofliteralmeaning;and,intu-itively,thereasonforthisisthateachpersonwhousesasoundlikeanangrydogtocommunicateangermighttakehimselforherselftobemakinganinnovativeuseofaresemblancebetweensounds.Whenwehaveacaseofliteralmeaning,incontrast,itseemsthatthereasonwhyweuseaparticularsoundisjustthatitdoeshavetheappropriateliteralmeaning.Alittlemoreaccurately,wecansaythatwhenwemakecommunicativeuseofanexpres-sionwithaliteralmeaning,werelyuponasharedrecognitiononthepartofspeakerandhearerthattheexpressioncanbeused–hasbeenused–injustthatway.Communicativeuseofanexpressionthathasaliteralmeaningisarationallyself-perpetuatingpractice;andthisfitswellwiththeideathatliteralmeaningisconventionalmeaning,sinceaconventionisarguablyarationallyself-perpetuatingregularity.(ThenotionofconventionismostfamouslyexploredbyLewis1969.)5.1Speaker’smeaningWehavereachedthebasicideathatasentencehasasitsliteralmeaningtheproposi-tionthatitisconventionallyusedtocommunicate.Grice’sprogrammeaimstodevelopadetailedphilosophicalanalysisalongthelinesofthisbasicidea.Thefirststageofthatprogrammeaimstocharacterizeaconceptofspeaker’smeaningroughlycorresponding,wesaid,tothenotionofcommunicating,orattemptingtocommunicate,aproposition.However,weshouldnoticethat,giventheoverallaimoftheprogramme,itisnot117\nMARTINDAVIESessentialthattheconceptofspeaker’smeaningshouldcoincidewithanyantecedentlygivennotion.Itwouldbeconsistentwiththeoverallaimthatspeaker’smeaningshouldcometoberegardedasatheory-internalconstruct.InGrice’sownexposition,thenotionofspeaker’smeaning(utterer’soccasionmeaning)tobeanalysedis:UttererUmeantsomethingbyhisorherutterancexdirectedatanaudienceA.TheputativeanalysisinitiallyofferedbyGrice(1957)isasfollows(Grice1989:92):Uutteredxintending(1)Atoproduceaparticularresponser;(2)Atothink(recognize)thatUintends(1);(3)Atofulfilcondition(1)onthebasisofhisfulfilmentofcondition(2).Inthecasewherethespeakerisattemptingtocommunicatethemessagethatptotheaudience,theprimaryintendedresponseinclause(1)isthatAshouldbelievethatp.Ifwefeedthisintotheanalysis,andalsounpackclause(3)alittle,thenwehavetheproposalthat:UttererUmeantthatpbyhisorherutterancexdirectedatanaudienceAshouldbeanalysedas:Uutteredxintending(1)Atobelievethatp;(2)Atothink(recognize)thatUintends(1);(3)A’srecognitionofU’sprimaryintention(1)tobeatleastpartofA’sreasonforbelievingthatp.TheutterancetypethatUusesmightormightnothavealiterallinguisticmeaning;andifitdoesthenthecommunicatedmessagemightormightnotcoincidewiththatliteralmeaning.But,crucially,thisanalysisofspeaker’smeaningdoesnotitselfmakeuseofthenotionofliteralmeaning,andsoisavailablefordeploymentinanon-circularanalysisofthatnotion.AhostofrevisionsandextensionshavebeenvisitedonGrice’sinitialthree-clauseanalysis(Grice1969;Strawson1964;Schiffer1972).Variouscounter-examplesseemtoshowthatGrice’sthreeconditionsareneithersufficientnornecessaryforspeaker’smeaning.Therearetwomainkindsofreasonwhythethreeconditionsmightnotbesufficientfortheintuitivenotionofcommunicatingamessageortellingsomebodysomething.Firstly,thereisnothinginthethreeconditionstorequirethatthereshouldbesomepropertyoftheutterancethattheutterershouldintendorexpecttoguidetheaudi-encetowardsrecognitionoftheprimarycommunicativeintention(Grice1989:94).Secondly,thethreeconditionsdonotruleoutcertainkindsofconvolutedplansforinfluencingaperson’sbeliefsinwaysthatarenot,intuitively,straightforwardly118\nPHILOSOPHYOFLANGUAGEcommunicative(Grice1969;Strawson1964;Schiffer1972:17–18).Weshallnottaketimeoverthedetailshere,buttheupshotisthat–atleastsofarassufficiencyiscon-cerned–abetteranalysisof:UttererUmeantthatpbyhisorherutterancexdirectedatanaudienceAwouldbe:(1)UintendedAtobelievethatp;(2)thereissomefeatureFoftheutterancexsuchthatUintendedAtorecognizeU’sprimaryintention(theintentionin(1))inpartbyrecognizingxtohaveF;(3)UintendedA’srecognitionofU’sprimaryintentiontobeatleastpartofA’sreasonforbelievingthatp;(4)UdidnotintendthatAshouldbedeceivedaboutU’sintentions(suchasthosein(1),(2)and(3)).Thepointofclause(4)is,specifically,toruleoutcaseswhereUintendsAtobedeceivedaboutU’sintentionsastohowAshouldarriveatthebeliefthatp(Grice1989:99).ItisnotsupposedtoruleoutthepossibilitythatUmightintendAtobedeceivedaboutotherintentionsthatUhas.Forexample,clause(4)iscertainlysupposedtoallowthatanactofspeaker’smeaningmightalsobeacaseoflying.Themainreasonwhythethreeorfourconditionsmightberegardedasnotneces-saryforspeaker’smeaningisthattherearesomanythingsthatordinaryspeakersdowithsentencesthatwouldnotbeincludedunderthethree-orfour-clauseanalysis.Indeed,therearemanyquiteliteralusesofsentenceswherethespeakerdoesnotevenintendtoproduceinthehearerabeliefcorrespondingtotheliteralmeaningofthesen-tence(sonotevenclause(1)wouldbesatisfied).Grice(1989:105–9)reviewsseveralofthesecases,includingexaminationanswers,confessions,reminding,givingareviewofalreadyknownfacts,spellingouttheconclusionofanargument,andtalkingtoacounter-suggestibleperson.Davidson(1984:111)alsomentions‘stories,roterepetitions,illustrations,suppositions,parodies,charades,chants,andconspicuouslyunmeantcompliments’.Therearevarioussuggestedremediesfortheproblemsthattheseexamplesaresup-posedtopose.But,forpresentpurposes,itisenoughtonotethatthemassivevarietyofliteralusesoflanguageisnotnecessarilyproblematicfortheputativeanalysisofspeaker’smeaning,ifthefundamentalmodelforspeaker’smeaningisthecaseoftellingsomeonesomethingwiththeintentionofprovidingnewknowledge.Arguably,thatnotionofspeaker’smeaning–withstraightforwardtellingasthecentralcase–canbeanalysedroughlyalongthelinesofGrice’sfour-clausedefinition.Butifweretainananalysisofspeaker’smeaningonwhichthemajorityofusesoflanguageturnoutnottobecasesofspeaker’smeaning,thenwemusttakethatfactintoaccountinthesecondstageoftheanalyticalprogramme.5.2ConventionalmeaningGiventhebasicideathatasentencehasasitsliteralmeaningthepropositionthatitisconventionallyusedtocommunicate,thesecondstageoftheprogrammeshould119\nMARTINDAVIEScharacterizetheconceptofconvention,andmostrecentworkintheGriceantraditionmakesuseofdevelopmentsofLewis’s(1969)notionofaconventionasarationallyself-perpetuatingregularityinbehaviour.AccordingtoLewis(1975:164–6)aconventionisaregularityRinaction,orinactionandbelief,thatmeetsthefollowingconditions:(1)everyoneconformstoR;(2)everyonebelievesthateveryoneelseconformstoR;(3)thebeliefin(2)furnisheseachpersonwithareasontoconformtoR;(4)thereisageneralpreferenceforgeneralconformitytoR,ratherthanslightlylessthangeneralconformity;(5)thereisatleastonealternativeregularitywhichwouldservereasonablywell;and(6)thefactslistedin(1)–(5)aremattersofcommonknowledge.Themoststraightforwardwaytoemploythisnotionandthatofspeaker’smeaningintandeminananalysisofliterallinguisticmeaningistosay:SentenceSmeansthatpinthelanguageofgroupGiffthereisaconventioninGtouseutterancesofSinordertocommunicatethatp(thatis,inordertomeanthatp,inthesenseofspeaker’smeaning).Inaslogan:Speaker’smeaning+Convention=Literalmeaning.Thiswouldneedtoberefinedtoallowfortheambiguityandcontext-dependencechar-acteristicofnaturallanguage.But,muchmoreimportantly,thesuggestedanalysisimposesarequirementthatisfartoostricttobeanecessaryconditionforliteralmeaning.Wehavealreadyseenthatthefour-clausedefinitionofspeaker’smeaningmakesitarelativelyrarephenomenon.Sotherecancertainlybeliteralmeaningevenwherethereisnoregularityofspeaker’smeaning(aspresentlydefined).Ifwewanttogiveananalysisoftheconceptofliteralmeaningalongthelinesoftheslogan,thenweneedtoweakenthenotionofspeaker’smeaning,orweakenthenotionofconvention(orweakenboth).(TheanalysesofliteralmeaningsuggestedbyrecenttheoristsalldepartinonewayoranotherfromthatmoststraightforwardwayofpursuingthesecondstageofGrice’sprogramme.SeeBennett1976;Blackburn1984:ch.4;Davies1981:ch.1;Lewis1975;Peacocke1976;Schiffer1972:ch.5.)Letusfirstconsiderweakeningthenotionofconvention.Thetroublewiththenotionofconventiondoesnotseemtolieintheideaofrationalself-perpetuation,butratherinthecommitmenttotherebeingaregularitythatmembersofthegroupactuallyconformto.Thesentence‘Penguinswaddle’literallymeans,inourlanguage,thatpen-guinswaddle.Butthereisnotaregularityofusingthatsentencetocommunicatethemessagethatpenguinswaddle.Itisnotthecase,forexample,thatwheneveranyone120\nPHILOSOPHYOFLANGUAGEwantstogetacrossthemessagethatpenguinswaddleheorsheusesthatsentence.Therearesurelyotherwaysinwhichthatsamemessagemightbecommunicated.Butmoreimportantforpresentpurposesisthat,althoughpeopledosometimesusethesentence‘Penguinswaddle’,itisnotregularlyusedwiththeintentionsspecifiedinthefour-clausedefinitionofspeaker’smeaning.Grice(1989:126)himselfsuggestsmovingfromspeaker’smeaningtoliteralmeaningviathenotionof‘havingacertainprocedureinone’srepertoire’.Itwouldnotbeadequatetosay:SentenceSmeansthatpinthelanguageofgroupGiffeachmemberofGhasitinhisorherrepertoiretouseutterancesofSinordertocommunicatethatp(thatis,inordertomeanthatp,inthesenseofspeaker’smeaning).Thatproposeddefinitionfailstobuildintheideaofrationalself-perpetuation(cf.Schiffer1972:132–6).Butamoreplausibleanalysiscouldbebuiltfromtheconceptofspeaker’smeaning(aspresentlydefined)plusaconceptanalogoustotheconceptofconventionbutwiththenotionofregularityreplacedbythenotionofhavingacertainprocedureinone’srepertoire.ThusGrice(1989:127)suggeststhatmoreisrequiredthaneachmemberofthegroupGhavingthecommunicativeprocedureinhisorherrepertoire:Atleastsome(many)membersofgroupGhaveintheirrepertoirestheprocedure[touseutterancesofSinordertocommunicatethatp],theretentionofthisproce-durebeingforthemconditionalontheassumptionthatatleastsome(other)membersofGhave,orhavehad,thisprocedureintheirrepertoires.Ifwedonotweakenthenotionofconvention,thenwehavetointroduceaconceptweakerthanspeaker’smeaning.Someliteralusesoflanguagethatfalloutsidethescopeofspeaker’smeaningaspresentlydefinedmightwellfallwithinthescopeofarevisednotioninwhichclause(1)isreplacedby:(1)¢UintendedAactivelytobelievethatp.Arguably,examplesofremindingarecoveredbythismodification.Someotherliteralusesarecoveredbyadifferentmodification:(1)≤UintendedAtobelievethatUbelievedthatp.Arguably,thisdealswiththecaseofexaminationanswers.Butinordertomakeitplau-siblethatwehavecoveredall,oratleastmost,literalusesoflanguage,weneedtoabstractawayfromthesedetailsaboutjustwhatattitudeUintendsAtotaketothepropositionthatp.Thecompletelygeneralformwouldbe:(1)¢≤ThereissomepropositionalattitudeY,suchthatUintendsthatAshouldYthatp.121\nMARTINDAVIESWecouldtakeclause(1)¢≤,togetherwithclauses(2)–(4)ofthefour-clausedefinitionofspeaker’smeaning,todefineanewnotionofweakspeaker’smeaning.Thismightthenbecombinedwiththenotionofconventioninaputativeanalysisofliteralmeaning(Peacocke1976;Davies1981).Howeverthedetailsareworkedout,aGriceananalysisoftheconceptofliteralmeaningforsentencespresentstheuseofapubliclanguageasapsychologicallyhighlycomplexmatter.OnaGriceanaccount,everydayliteraluseoflanguageforcommuni-cationinvolvesbeliefsaboutbeliefs,intentionsaboutbeliefs,beliefsaboutintentionsaboutbeliefs,andmuchmore.Itmayseemimplausiblethatsuchcomplicatedproposi-tionalattitudesareinvolvedintheordinaryuseoflanguageandthismaymotivateanobjectiontoGrice’sprogrammeforanalysingthenotionofliterallinguisticmeaning.But,ontheotherhand,thepsychologicalplausibilityofGriceanaccountshasbeendefended(Loar1981).Thereareimportantquestionshere.ButeveniftheyareansweredinGrice’sfavour,therearefurtherdifficultiestobefaced.Forwehavenotyettakenanyaccountoftheinternalstructureofsentences.5.3ExpressionmeaningandthestructureoflanguageThewayinwhichcomponentwordsandphrasesmaketheircontributionstothemean-ingsofcompletesentencesisthestarting-pointforDavidson’sprogrammeinphiloso-phyoflanguage(seesection3).ButtheanalysisoftheconceptofliteralmeaningasitappliestoexpressionssmallerthancompletesentencesisanextremelydifficultproblemwithinGrice’sprogramme(Grice1968;Schiffer1972:ch.6;Loar1976,1981:253–60).Whetherornotitisanabsolutelyinsolubleproblemis,atbest,notclear.Theissueshereareverycomplex,andweshallonlygiveabriefindicationofthewayinwhichthedifficultyarises.Inessence,theproblemisthatthestructureoflanguagepresentsadilemmaforthegeneralideathatliteralmeaningisamatteroftherebeingconventionstouseexpres-sionswithcertainintentions.Theprimarylinkbetweenmeaningandconventionhastobemadeeitheratthelevelofwordsandphrases,andwaysofputtingthemtogether(suchassubject–predicatecombination),orelseatthelevelofcompletesentences.Eitheralternativepresentsproblems.Supposethatweoptforthefirstalternative.Then,weshallcreditlanguageuserswithhavingproceduresforusingwords,andphrases,andwaysofputtingthemtogether,andalsowithknowledgethattheyandothershavetheseprocedures.Weneedtocreditspeak-erswiththisknowledge,inordertomaintaintheideaofrationalself-perpetuation.Butitisherethattheproblemwiththisfirstalternativeispresented.Forthesemanticrulesthatgoverntheuseofsub-sententialcomponents,andwaysofcombiningcomponents,arefarfromobvious.(Thatiswhytheconstructionofcompositionalsemantictheoriesisachallengingproject.)So,onthefirstalternative,itispracticallyinevitablethattheanalysisofmeaningwillcommitustoattributingtoordinarylanguageusersbeliefsandknowledgethattheydonothave:detailedbeliefsandknowledgeaboutthecomponentsandconstructionoftheirsentences(Loar1981:256).Suppose,instead,thatwetrythesecondalternative.Inthatcase,literalmeaningwillbegroundedinrationallyself-perpetuatingregularitiesofuseofcompletesen-tences.Buttheproblemthistimeisthattherearemanyperfectlymeaningfulsentences122\nPHILOSOPHYOFLANGUAGEthatareneverusedatall(meaningwithoutuse);andtherearemanyothersthatareused,butwouldneverbeusedtocommunicatethepropositionthatistheirliteralmeaning(meaningdespiteuse).Anexampleofthelatterkindisprovidedbythesen-tence:‘Noheadinjuryistootrivialtobeignored’(Bennett1976:17).Thissentenceissometimesused;butitistypicallyusedwiththeintentionthatthehearershouldbelievethatnoheadinjuryissufficientlytrivialtobeignored.Apossibleresponsetothisdilemmawouldbetoavoidthefirstalternativeandadoptthesecondalternative,butonlyforthosesentencesforwhichthereisapracticeofliteraluse.Accordingtothisresponse,theGriceannotionofconventionalmeaningwouldnotapplydirectlytounusedsentences.Sotherewouldremaintheproblemofextendingthenotionofliteralmeaningfromthecoreofusedsentencestotherestofthesen-tencesinthelanguage.Oneplausiblesuggestionfordealingwiththisproblem–perhapstheonlyplausiblesuggestion–isthatanaccountofliteralmeaningapplica-bletoallsentences,whetherusedornot,shouldexplicitlyadverttothepsychologicalmechanismsthatunderpintheuseofsentencesinthecoreset(Loar1976:160;seealsoSchiffer1987:249–55;1993).Accordingtothissuggestion,theanalysisofliteralmeaningwouldbeextendedfromusedsentencestounusedsentencesintwosteps.Firstly,consideraspeakerwhoparticipatesinaconventionalpracticeofusingasen-tencewithacertainliteralmeaning.Weassumethatthespeaker’sassignmentofameaningtothesentenceisunderpinnedbycognitivemechanismsthatcorrespondcloselytosomeoftheaxiomsofacompositionalsemantictheoryforthelanguage.Inparticular,weassumethatthecognitivemechanismssubservingtheassignmentofameaningtoasentenceincludemechanismscorrespondingtothesemantictheory’saxiomsforthewords(andwaysofcombiningwords)inthatsentence.Secondly,consideranunusedsentencethatisbuiltfromwordsthatoccurinusedsentences(andisbuiltinwaysthatarealsofoundinusedsentences).Thenweallowtheliteralmeaningfortheunusedsentencetobedeterminedbythesemanticaxiomsthatcorrespondtocertaincognitivemechanisms;namely,themechanismsthatareimplicatedinthespeaker’sassignmentofmeaningstoregularlyusedsentencesbuiltfromthosesameresources.(Strictlyspeaking,thisdealsonlywiththeproblemofmeaningwithoutuse.Butasimilarstrategycouldbeadoptedtodealwiththeproblemofmeaningdespiteuse.)AnaturalwayofimplementingthissuggestionwouldbetomakeuseofChomsky’snotionoftacitknowledge.Insteadofsaying,rathervaguely,thattherearecognitivemechanismscorrespondingintheappropriatewaytoaxiomsofasemantictheory,wecouldsaythatthoseaxiomsaretacitlyknown(seeabove,section4.4).Indeed,Loar(1981:259)suggeststhat‘theChomskyanideaoftheinternalizationofthegenera-tiveproceduresofagrammarhasgottobeinvokedto...makesenseofliteralmeaning’(seealsoDavies2000).AppealingtothenotionoftacitknowledgeofacompositionalsemantictheorymaybethebestwayforaGriceantosolvetheproblemofmeaningwithoutuse.ButwedonothaveanydetailedprospectusforamarriagebetweenChomsky’sandGrice’sratherdifferentprojectsinthestudyoflanguage.ItiscertainlynotobviousjusthowGriceantheresultingunioncouldbe(Chomsky1976:55–77;1980:81–7;1986:19–24;seealsoLaurence1996,1998,foraChomskyanaccountofliteralmeaningpresentedasacompetitortoaGriceanor‘convention-based’account).123\nMARTINDAVIESInsummary,then,wecansaythatthereissomethingofaconsensusthatGrice’sprogrammefacesseriousproblems.Lycan(1991:84)listssomeofthem:thatmostsentencesofalanguagearenevertokenedatall;thatsincehearersinstantlyunderstand‘novel’sentences,thiscannotbeinvirtueofpre-establishedconventionsorexpectationsdirectedonthosesentencesindividually;thatsentencesareambiguousandhavemorethanonestandardlinguisticmeaning;thatsentencesareoften(notjustabnor-mally)usedwithotherthantheirstandardmeanings;andthatindeedsomesentencesarenormallyusedwithotherthantheirstandardmeanings.Inthelastfewparagraphswehavesketchedapossiblestrategyforrespondingtosomeoftheseproblems.Butweshouldalsonotethatoneofthemostauthoritativeexpo-nentsofGrice’sprogrammehas,afterconsideringjustsuchastrategy,reachedthecon-clusionthattheprojectofanalysingliteralmeaningintermsofintentionsandbeliefs(IBS:intention-basedsemantics)cannotbecarriedthrough(Schiffer1987:261):‘Ihaveexhaustedmywits...IcannotseehowtodeviseanIBSaccountofexpression-meaningthatdoesnotrequireustohaveknowledgethatweseemnottohave’.5.4MinimalismaboutmeaningandtruthSchiffer(1987:10)goesbeyondtheclaimthatGrice’sanalyticalprogrammecannotbecompletedtothemoregeneralconclusionthatthereisnocorrectandsubstantivetheoryofmeaning,inthesenseofaphilosophicaltheoryaboutthenatureoflinguisticmeaning,tobegiven.AccordingtoSchiffer,therearenosuchentitiesasmeaningstobeassignedtosentencesand,giventhatGrice’sanalyticalprojectcannotbecompleted,thereisnoprospectofaphilosophicalanalysisoftheconceptofmeaning.Thereisstillroomforatheoryabouthowlanguageunderstandingtakesplace,abouttheinformationprocessingthatgoesoninourheadswhenwehearanutterance.But,asSchiffersays,‘suchatheorywouldnotbeaphilosophicaltheory’(ibid.:269).This‘no-theory’theoryofmeaningisaspeciesofminimalismaboutmeaning.Accordingtotheminimalistview(Johnston1988),meaninghasnosubstantialnatureandsothereisnothingsubstantiveforatheoryaboutmeaningtosay.Suchgeneralelucidationofthenotionoflinguisticmeaningasmayberequiredisprovided,notbyasubstantivetheoryorconceptualanalysis,butbytruismsorplatitudes,suchas:IfasentenceSmeansthatpthenutterancesofScanbeusedtosaythatp.SomeoneworkingwithinGrice’sprogrammemighthopetoprovideananalysisofthenotionofaspeakerusinganutteranceofStosaythatpand,inthatway,toprovideasubstantive,ifpartial,accountofthenotionofmeaning.Buttheminimalistsaysthatthereisnosuchanalysistobeprovided.Theminimalistmaygoontosaythattheplatitudesaboutmeaninghighlightacon-nectionbetweenmeaninganduse:understandingS,orknowingthatSmeansthatp,isamatterofbeingabletouseutterancesofStosaythatp.Ifthatisright,thenknowl-edgeofmeaningmayappeartobesocloselyconnectedtolanguageusethatitcannotplayanyroleinthecausalexplanationoflanguageuse.Thisclaimaboutthelackof124\nPHILOSOPHYOFLANGUAGEanexplanatoryroleforknowledgeofmeaningis,indeed,oneaspectofminimalismaboutmeaning.Buttheproposalthatthereisacloseconnectionbetweenmeaninganduseseemstogoagainstthestrictestconstrualofthe‘no-theory’theoryofmeaning.Fortheproposalthatmeaningiscloselyconnectedwithuse,oreventhatmeaningisuse(Wittgenstein1953;Horwich1998),doesseemtobeasubstantivepieceofphilo-sophicaltheoryconcerningthenotionofmeaning.Itmaybeuseful,here,todistinguishtwoversionsofthe‘no-theory’theoryofmeaning.Thestrictversionsaysthatthereisnothingsubstantiveatalltobesaidaboutmeaning–noteventhatmeaningisuse.Thismightbeaccompaniedbythesuggestionthatthereisnothingmoretoknowaboutmeaningthanisconstitutedbyacceptanceofsuchtrivial-seemingstatementsaboutmeaningas:Thesentence‘Fidobarks’meansthatFidobarks.Thesentence‘Themaninthegabardinesuitisaspy’meansthatthemaninthegabardinesuitisaspy.Thesentence‘Theaetetusissitting’meansthatTheaetetusissitting.Andsoon.Themoremoderateversionofthe‘no-theory’theoryofmeaningallowsthattheremaybesomethingofphilosophicalinteresttobesaidaboutlinguisticmeaning.Itwouldbeconsistentwiththemoderate‘no-theory’theorytoofferaphilo-sophicaltheoryaccordingtowhichfactsaboutmeaningaredeterminedbyfactsaboutuse.Butthemoderate‘no-theory’theorystillinsiststhatthereisnosubstantivephilo-sophicaltheoryaboutmeaningthatcanbepackagedasananalysisoftheconceptofmeaningalongthelinesof:SentenceSmeansthatpiff_S_p_.CertainlythereisnothingofthekindthatGricewasaimingtoprovide.Theideathatmeaningdependsonuse,orthatmeaningisconstitutedbyuse,isnotobviouslyinconsistentwiththeaimofprovidingananalysisofthenotionoflinguis-ticmeaning.Afterall,Grice’sanalyticalprogrammemightberegardedasonewayofdevelopingtheideathatmeaningdependsonuse:SentenceSmeansthatpiffSisconventionallyusedtocommunicatethemessagethatp.Butitiscertainlypossibletooffera‘usetheory’ofmeaningwhileexplicitlydisavow-inganyanalyticalambitions.Theterm‘usetheoryofmeaning’isusuallyappliedtometa-semantictheoriesthatappealtothenotionofuseandalsomeettherequirementforbeingamoderate‘no-theory’theoryofmeaning.Ausetheoryofmeaning,inthissense,issometimescoupledwithminimalism(oradeflationarytheory)abouttruth(Horwich1990,1998).Themainclaimmadebyaminimalistaboutthenotionoftruthisthatthereisnothingmoretograspingthenotionoftruththanacceptinginstancesoftheschema(Horwich1990:7):(E)Itistruethatpifandonlyifp125\nMARTINDAVIESsuchas:ItistruethatFidobarksiffFidobarks.Itistruethatthemaninthegabardinesuitisaspyiffthemaninthegabardinesuitisaspy.ItistruethatTheaetetusissittingiffTheaetetusissitting.Because(E)usestheexpression‘Itistruethat–’,itdoesnotspeakofthetruthofsentences.Thereisacleardifferencebetween:Thesentence‘Fidobarks’istrueiffFidobarks.whichisaboutalinguisticexpression,‘Fidobarks’,andtheinstanceof(E):ItistruethatFidobarksiffFidobarks.whichmightbesaidtoconcernthepropositionthatFidobarks.Sowhatwehavedescribedsofarisminimalismabouttruthforpropositions,orminimalismabout‘Itistruethat’.Butaminimalistabouttruthforpropositionsislikelytomaintain,inaddition,thatthereisnothingmoretograspingthenotionoftruthasitappliestosentences(orutter-ancesofsentences)thanacceptanceofstatementsofthefollowingkind:(Anutteranceof)thesentence‘Fidobarks’istrueiffFidobarks.(Anutteranceof)thesentence‘Themaninthegabardinesuitisaspy’istrueiffthemaninthegabardinesuitisaspy.(Anutteranceof)thesentence‘Theaetetusissitting’istrueiffTheaetetusissitting.Inthesestatementsthesamesentenceoccursfirstinquotationmarksandthenwithoutquotationmarks,andstatementsofthisformaresaidtoencapsulatethedisquotationalconceptionoftruth(Field1994:250;seealsoField1986).Evenifminimalismabouttruthforpropositionsisplausible,minimalismabouttruthasitappliestosentencesmayseemtofaceachallenge.Aswesawinsection4.2,Tarski’sConventionTcanbeconceivedasapartialelucidationoftheconceptoftruthforsentencesintermsoftheconceptoftranslationortheconceptofmeaning.Wealsoobservedthatconditionsofadequacyonsemantictheories(orequivalently,constraintsontheactuallanguagerelation)canhelpustoelucidatetheconceptofmeaningandthence,viathemeaning–truthconnection,theconceptoftruthforsentences.AndwedescribedthewayinwhichWiggins(1980)offereda‘substantialtheoryoftruth’byimposingaconditionofadequacy(basedonthePrincipleofHumanity)directlyonTarski–Davidsontheoriesoftruthconditions.Byofferingtheprospectofsubstantialphilosophicalelucidationoftheconceptoftruth,theseconsiderationsseemtothreatenminimalismabouttruthforsentences.Buttheminimalistabouttruthcanrespondtothisthreatofunwantedsubstantive-ness.Hemight,forexample,focusontheapplicationofTarski’sConventionTtothe126\nPHILOSOPHYOFLANGUAGEcasewhereatheoryoftruthconditionsforalanguageisgiveninanextensionofthatsamelanguage.For,inthatcase,whatConventionTrequiresof:SentenceSistrueifandonlyifpissimplythatitshouldmeetthe‘disquotational’conditionthatthesentencethatfillsthe‘p’placeshouldbethesentenceSitself,asin:Thesentence‘Fidobarks’istrueiffFidobarks.Andhemightcombinethiswithinsistencethatthereisnothingmoresubstantivetobesaidabouttheconditionsunderwhichatheoryoftruthconditionsiscorrectand,inparticular,nothingsubstantivetobesaidabouttheactuallanguagerelation(Field1994:section6).Alternatively,aminimalistabouttruthwhoalsoholdsausetheoryofmeaning(Horwich1998)mayofferhisownaccountofthemeaning–truthconnection:IfasentenceSmeansthatpthenSistrueiffpdrawingonnoresourcesthatgobeyondtheusetheoryofmeaningplusdisquotationalequivalencessuchas:(Anutteranceof)thesentence‘Theaetetusissitting’istrueiffTheaetetusissitting.Itisplausible,inthiscase,thatthethreatofunwantedsubstantivenessflowingacrossthemeaning–truthconnectioncanbemetprovidedthattheusetheoryofmeaningisindeedamoderate‘no-theory’theory.Minimalismabouttruthavoidsapotentialproblemand,tothatextent,gainsameasureofsupport,bybeingcoupledwithamoderate‘no-theory’theoryofmeaning,suchasausetheory.Thereisalsoarelationofsupportintheoppositedirection.AsHorwich(1998:113)says:Byitself,thedeflationary[minimalist]viewoftruthdoesnotmotivatetheusetheoryofmeaning.Butitprovidesavitalpartoftheargumentbyshowingthatacommonreasonforrejectingthetheorywasbasedonamistake.Iftherewereasubstantivephilosophicaltheoryoftruththen,becauseofthemeaning–truthconnection,therewouldbeatleastasubstantivenecessaryconditionon:SentenceSmeansthatp.Thiswouldnotamounttoananalysisofthenotionoflin-guisticmeaning,butitwouldgoagainstthespiritofthe‘no-theory’theoryofmeaning.Soamoderate‘no-theory’theoryofmeaning,suchasausetheory,ismoreplausiblewhenitisaccompaniedbyminimalismabouttruth.Ausetheoryofmeaningandminimalismabouttruthforsentencesaremutuallysupporting.Eachallowstheothertoavoidaproblem.Thecombinationmightbecalled‘semanticdeflationism’(ibid.:11)–or,perhapsbetter,‘meta-semanticdeflationism’.ThisdeflationarypositiononmeaningandtruthisopposednotonlytoGrice’s127\nMARTINDAVIESanalyticalprogramme,butalsotoDavidson’sprogrammewithitsfocusontheconstructionofcompositional,truth-conditionalsemantictheories.Theevaluationofmeta-semanticdeflationismisapressingtaskforthephilosophyoflanguage.6Pragmatics:ConversationalImplicatureandRelevanceTheoryAsentencethatliterallymeansthatpcanbeusedtocommunicateamessageotherthanthepropositionthatp.Thus,supposethatsomeoneasksyou,‘Wouldyoulikemorecoffee?’andthatyouanswer:‘Coffeekeepsmeawake’.Inasuitablecontext,thisanswercanconveythemessagethatyoudonotwantanymorecoffee(SperberandWilson1995:34);inadifferentcontext(aswhenyouareobligedtostayalertthroughaboringlecture)itmightconveythemessagethatyoudowantsomemorecoffee.Butneitherofthesepossiblemessagesisanypartoftheliteralmeaningofthesentence‘Coffeekeepsmeawake’.Ineachcase,thecommunicatedmessageis,rather,somethingimplied(insomesense)inthecontext,bywhatisliterallyandexplicitlysaid.Thedistinctionbetweensemanticsandpragmaticsis,roughly,thedistinctionbetween(1)thesignificanceconventionallyorliterallyattachedtowords,andthencetowholesentences,and(2)thefurthersignificancethatcanbeworkedout,bymoregeneralprinciples,usingcontextualinformation.Grice’stheoryofconversationalimpli-catureistheseminalcontributiontorecentpragmatictheory.Itisatheory–orthebeginningofatheory–ofhowthisfurthersignificanceisgenerated.6.1ConversationalimplicatureThetheoryofconversationalimplicaturewasfirstdevelopedinthecontextofexam-plessuchasthese.and:Strawson(1952)discussesdifferencesbetweentheconnectivesofpropositionallogicandthecorrespondingwordsofEnglish,andhenotesthat‘astatementoftheform“pandq”maycarryanimplicationoftemporalorder’(ibid.:81).or:Strawsonalsonotesthat‘thealternativestatement[statementoftheform“porq”]carriestheimplicationofthespeaker’suncertaintyastowhichofthetwoitwas’(ibid.:91).but:Thestatement‘Shewaspoorbutshewashonest’carriestheimplicationthatthereissomekindofcontrastbetweenbeingpoorandbeinghonest.looks:Thestatement‘Thatlooksredtome’carriestheimplicationthatitisinsomewaydoubtfulwhethertheobjectreallyisred.Wecanaskwhethertheseimplicationsareallofthesamekind.Aretheytheproductsofspecialfeaturesofkeywordsinthesentences–‘and’,‘or’,‘but’,‘looks’–oraretheygeneratedbysomekindofinteractionbetweentheliteralmeaningsofthesentencesthatareutteredandsomeverygeneralconversationalprinciples?Grice’s(1961)firstexpositionoftheideaofconversationalimplicatureisparticu-larlyconcernedwiththeexampleof‘looks’.Hewantstomaintainthatthestatement‘Thatlooksredtome’isliterallytrue,evenwhentheobjectinquestionquiteobviously128\nPHILOSOPHYOFLANGUAGEisred,sothatthereisnoquestionofanydoubtaboutitscolour.Inthatpaper,Gricedevelopstwonotions–thedetachabilityandthecancellabilityofanimplication–thatwecanusetohelpustoidentifyconversationalimplicaturesandtodistinguishthemfromconventionalimplicatures.Intheexamplewith‘looks’,theimplicationisnotdetachable.Thismeansthatthereisnoformofwordsthatcanbeusedtostatejustwhat‘Thatlooksredtome’canbeusedtostate(thathasthesametruthconditionsas‘Thatlooksredtome’),butwithoutcarryingtheimplication.Theobviouscandidates–‘Thatappearsredtome’,‘Thatseemsredtome’–carrythesameimplicationthatthereissomedoubtabouttheobject’scolour.Ontheotherhand,theimplicationiscancellable.Itisquiteconsistenttooverridetheimplicationofadoubtbysaying:‘Thatlooksredtome...anditreallyisred–there’snodoubtaboutit’.Intheexampleswith‘and’and‘or’,theimplicationislikewisecancellablebutnotdetachable;and,indeed,thatisthetypicalprofileforconversationalimplicature.Inthecaseof‘but’,however,theprofileisreversed.Theimplicationisdetachable,butnotcancellable.Theimplicationisdetachable(asGriceusesthatterm)becausethereisanotherword,‘and’,thatmakesjustthesamecontributiontotruthconditionsas‘but’,butdoesnotgeneratethesameimplication:‘Shewaspoorandshewashonest’doesnotcarrytheimplicationofacontrastbetweenpovertyandhonesty.Theimpli-cationisnotcancellablebecauseitwouldbeextremelyodd–indeed,inconsistent,insomesense–tosay:‘Shewaspoorbutshewashonest,thoughofcourseIdonotmeantoimplythatthereisanycontrastbetweenpovertyandhonesty’.Inthecaseoftheword‘but’,then,theimplicationthatisgeneratedappearstobeafeatureoftheliteralorconventionalmeaningofthatspecificword–eventhoughitisnotafeaturethathelpstodeterminethetruthconditionsofsentencescontainingtheword.Intheothercases–‘and’,‘or’,‘looks’–theimplicationisgeneratedbysomemoregeneralprinciplesgoverningconversation.Grice(1961:94)offersaconversationalprinciplethatatleastappliestothecasesof‘or’andof‘looks’.Hesuggeststhattheimplicationsaregenerated,inthosecases,bytheassumptionthatthereisinoperationageneralprinciplesomethinglike:Oneshouldnotmakeaweakerstatementratherthanastrongeroneunlessthereisagoodreasonforsodoing.Ifsomeonemakesadisjunctivestatement,thenthehearercansuppose–giventhisconversationalprinciple–thatthespeakerisnotinapositiontomakeastrongerstatementbyusingjustoneofthedisjuncts.Similarly,‘Thatlooksredtome’isintuitivelyaweakerstatementthan‘Thatisred’,andaspeaker’suseoftheweakerstatementabouthowitlookstoher(orappearstoher,orseemstoher)generatestheimplicationthatsheisnotinapositiontomakethestrongerstatementabouthowitisinreality.Theexamplewith‘and’,however,isnotdealtwithbythisconversationalprinciple.Inalaterpaper,Grice(1975)offersamorefullyworkedouttheorythatcoverstheexampleof‘and’aswell.(Grice’spapersonpragmaticsarecollectedinGrice1989.)ThefundamentalideaisthatparticipantsinaconversationexpecteachothertoobserveaCo-operativePrinciple:129\nMARTINDAVIESMakeyourconversationalcontributionsuchasisrequired,atthestageatwhichitoccurs,bytheacceptedpurposeordirectionofthetalkexchangeinwhichyouareengaged.ThisprincipleiselaboratedbyGriceintoaseriesofmaxims.Hegroupsthesemaximsunderfourheadings:Quantity,Quality,RelationandManner.Examplesofthemaximsinclude(oneexamplefromeachcategory):Makeyourcontributionasinformativeasisrequired(forthecurrentpurposesoftheexchange).Donotsaythatforwhichyoulackadequateevidence.Berelevant.Beorderly.ThefirstofthesesamplemaximsisoneofthemaximsofQuantity,anddoestheworkoftheearlierconversationalprinciple.ThelastisoneofthemaximsofManner,andisusedtoexplaintheimplicaturegeneratedby‘and’.Thewaythatconversationalimplicaturesarediscernedinvolvesthehearerininfer-ence.Supposethataparticipantinaconversationsays–quiteliterallyandexplicitly–thatp,andgivesnoindicationthatheisnotobservingtheCo-operativePrincipleandthemaxims.Then,onGrice’stheory,aconversationalimplicaturethatqisgeneratedifthesuppositionthatthespeakerbelievesthatqisrequiredinordertomakehisliter-allysayingthatpconsistentwiththepresumptionthatheisoperatinginaccordancewiththemaxims‘oratleasttheCo-operativePrinciple’(ibid.:31).AccordingtoGrice,conversationalimplicaturescanbeworkedoutorcalculatedusingtheCo-operativePrincipleandthemaximstogetherwithinformationabouttheliteralmeaningofthesentenceused,informationaboutthecontext,andotherbackgroundassumptions.Thereisnowayoftelling,aheadoftime,whatbackgroundassumptionsmightbehelpfulinworkingoutanimplicature.Sothisaspectofcommunicationinvolvesagreatdealthatisnotspecificallytodowithlanguage.Furthermore,onGrice’sview,theCo-operativePrincipleandthemaximsthemselvesarenotspecifictolanguage,butaresimplytheinstantiationstothecaseofconversationofprinciplesandmaximsthatitwouldbereasonabletoobserveinanyco-operativeactivity–suchascookingameal,orhelpingafriendtochangethewheelonacar.Themechanismofconversationalimplicaturecanbeusedtocommunicatemorethanisliterallysaid–morethantheliteralmeaningthatisencodedinthewordsthatareused.Ifanimplicatureturnsouttobefalse,thenthehearermaywellbemisled;butthespeakerhasnot,strictlyspeaking,spokenfalsely.Thisdistinctionbetweenwhatisfalse,andwhatisonlymisleading,isofconsiderablegeneralphilosophicalimportance.IfBruceisstandingsquarelyinfrontofastandardletterboxintheUnitedKingdomingooddaylight,andsays,‘Thatlooksredtome’,thenwhathesaysmaybemisleading,totheextentthatitsuggeststhatthereissomedoubtaboutthecolouroftheletterbox.Butstill,whathesaysistrue:theletterboxdoeslookredtoBruce.(Grice(1961)usesthispointtodefendaversionofthesensedatatheoryofperception.)130\nPHILOSOPHYOFLANGUAGE6.2ThreeproblemsforGrice’spragmatictheoryAlthoughGrice’spragmatictheoryisimportantandhasbeeninfluential,therearesomequiteseriousproblemswithit.Herewementionthree.Thefirstproblemisthatthetheoryprovidesnoaccountofhowthevariousmaximsaretobeweightedagainsteachother.Thusconsiderthefollowingexample:A:WheredoesClive?B:SomewhereintheSouthofFrance.Grice’sownglossontheexampleisasfollows(Grice1989:32–3):ThereisnoreasontosupposethatBisoptingout[fromtheCo-operativePrincipleandthemaxims];hisansweris,ashewellknows,lessinformativethanisrequiredtomeetA’sneeds.ThisinfringementofthefirstmaximofQuantity[Makeyourcontributionasinfor-mativeasisrequired(forthecurrentpurposesoftheexchange)]canbeexplainedonlybythesuppositionthatBisawarethattobemoreinformativewouldbetosaysomethingthatinfringedthemaximofQuality,‘Don’tsaywhatyoulackadequateevidencefor’,soBimplicatesthathedoesnotknowinwhichtownClives.IfwelookatthisglossinthelightofGrice’saccountofthegenerationofimplicatures,thenweseethatitisimplicitthatmeasuringuptothemaximofQualityexcusesviolatingthemaximofQuantity,whenthereisaclashbetweenthem.ThemaximofQualityisthusimplicitlyrankedasmoreimportantthanthemaximofQuantity.ButthereisnothingexplicitinGrice’saccountabouttherelativerankingofthemaxims.InGrice’spragmatictheory,theCo-operativePrincipleisdevelopedinalargenumberofmaxims,butnogeneralaccountisprovidedoftherelationshipsbetweenthemaxims.Itwouldthusbetemptingtolookforsomeoneoverarchingaiminconversa-tion,whichwouldjustifyasingleoverarchingmaxim.Thesecondproblemarisesfromthefactthatwhatisconversationallyimplicatedissupposedtobepartofthemessagethatiscommunicated:itissupposedtofallwithinthescopeofspeaker’smeaning.Itisunclearhow,giventhisfact,Grice’spragmatictheorycanallowfortheopen-endedcharacterofmanyimplicatures–seenclearlyintheopen-endedinterpretationofmetaphor.WhatGriceactuallysaysaboutopen-endednessisthis(Grice1989:39–40):Since,tocalculateaconversationalimplicatureistocalculatewhathastobesupposedinordertopreservethesuppositionthattheCo-operativePrincipleisbeingobserved,andsincetheremaybevariouspossiblespecificexplanations,alistofwhichmaybeopen,theconversationalimplicatuminsuchcaseswillbedisjunctionofsuchspecificexplanations;andifthelistoftheseisopen,theimplicatumwillhavejustthekindofindeterminacythatmanyactualimplicatadoinfactseemtopossess.Butwhatisconversationallyimplicatedissupposedtofallwithinthescopeofspeaker’smeaninganditisfarfromclearthatitmakessensetosaythatthepropositionthataspeakerovertlyintendedtocommunicatewasanindefinitelyextendeddisjunction.Theproblemhereisthatitisunclearexactlywhatbeliefthespeakerwouldintendthe131\nMARTINDAVIESspeakertoendupwith(clause(1)ofthedefinitionofspeaker’smeaninginsection5.1),andequallyunclearexactlywhatintentionthespeakerwouldintendthehearertorecognizeherashaving(clause(2)).Weneedtoallowthat,althoughsometimesaspeakermayovertlyintendthehearertorecognizeaspecificconversationalimplicatureasthecommunicatedmessage,thiscaseliesatoneendofaspectrum.Furtheralongthespectrumtherearecaseswherethespeakerintendsthatthehearerwillrecognizeoneormoreimplicaturesinamoreorlesscloselydemarcatedrange.Furtheralongagain,therearecaseswherethespeakerisdoinglittlemorethantoinvitethehearertoexploreaspaceofpossibleimplicatures.Whatthissuggestsisthatwemightdobesttofocusontheways–somespecificallyintended,othersnot–inwhichahearercanexploitwhataspeakerliterallyandexplicitlysays.Conversationalimplicatures,meetingthestandardsofspeaker’smeaning,wouldthenberegardedasthelimitingcaseofamuchmoregeneralphenomenon.Thethirdproblemisthattheaccountofthegenerationofconversationalimplica-turesisunder-specified.Toseethis,considerwhatGricesaysaboutmetaphorandirony(ibid.:34).Ineachcase(metaphor:‘Youarethecreaminmycoffee’;irony:‘Xisafinefriend’)whatisliterallyandexplicitlysaidisobviouslysomethingthatthespeakerbelievestobefalse.So,onthefaceofit,thespeakerisfloutingoneofthemaximsofQuality:Donotsaywhatyoubelievetobefalse.Sothehearerhastofindapropositionqsuchthat,bysupposingthespeakertobelievethatq,thehearercanseethespeakerasneverthelessobservingtheCo-operativePrinciple.Inthecaseofmetaphor,Grice’s(ibid.)glossis:Themostlikelysuppositionisthatthespeakerisattributingtohisaudiencesomefeatureorfeaturesinrespectofwhichtheaudienceresembles(moreorlessfancifully)thementionedsubstance.Inthecaseofirony,itis:UnlessA’sutteranceisentirelypointless,Amustbetryingtogetacrosssomeotherpropo-sitionthantheonehepurportstobeputtingforward.Thismustbesomeobviouslyrelatedproposition;themostobviouslyrelatedpropositionisthecontradictoryoftheonehepurportstobeputtingforward.ButGriceprovidesnoexplanationofthewayinwhichtheparticularrelationshipbetweenliteralmeaningandcommunicatedmessageistobecalculatedineachcase(butseeGrice(1978)forsomefurtherdiscussion).6.3RelevancetheoryDanSperberandDeirdreWilson(1981,1986a,1986b,1995;WilsonandSperber1981)havedevelopedapragmatictheorythatisbasedonaquitegeneralaccountofthewayinwhichahearermakesuseofavailableinformation.Theyproposethatthekeytoatheoreticalunderstandingofcommunication–and,indeed,ofcognitioningeneral–isprovidedbythenotionofrelevance.Thus:132\nPHILOSOPHYOFLANGUAGEOursuggestionisthathumanstendtopayattentiontothemostrelevantphenomenaavailable;thattheytendtoconstructthemostrelevantpossiblerepresentationsofthesephenomena,andtoprocesstheminacontextthatmaximizestheirrelevance.Relevance,andthemaximizationofrelevance,isthekeytohumancognition.(SperberandWilson1986b:586)Inordertoexplainthenotionofrelevance,weneedtomakecleartheideaofacontextthatSperberandWilsonuse.Inrelevancetheory,acontextis‘asubsetofthehearer’sassumptionsabouttheworld’(SperberandWilson1995:15)thatprovidespotentialpremisestobeusedininferences.Theideaisthat,atanygivenpointinaconversationalexchange,thehearerhassomeassumptions‘intheforefrontofhisattention’.Whenanewpropositionisintroduced–say,astheresultofacommunicativeactonthepartofthespeaker–thenewlyintroducedpropositionandthepre-existingcontextinteractinferentially,soastoproducecognitiveeffects.Animplicationthatdependsonthenewpropositionandalsoonsomeoftheassumptionsinthecontextissaidtobeacontextualimplicationofthenewproposition;andcontextualimplicationisthesimplestexampleofacognitiveeffect.Relevanceisamatterofhavingcognitiveeffects(SperberandWilson1986b:586):Weclaimthatinformationisrelevantifitinteractsinacertainwaywithyourexistingassumptionsabouttheworld.Roughly,themorecognitiveeffectsapropositionhas,themorerelevantitis.However,thecostofderivingcognitiveeffectshastobetakenintoaccountaswell.Soweaddasecondcondition;namely,thatthelesseffortthatisrequiredinordertoderivecogni-tiveeffectsfromaproposition,themorerelevantitis.Thenotionofrelevancethenpro-videsanaccountofutteranceinterpretationviaaCommunicativePrincipleofRelevance(SperberandWilson1995:158):Everyactofostensivecommunicationcommunicatesthepresumptionofitsownoptimalrelevance.Thissays,roughly,thatanutterancecarriesaguaranteethatitcanyieldanadequaterangeofcognitiveeffectswithouttoomuchprocessingeffort.(Forfurtherdiscussionofthepresumptionofoptimalrelevance,seeibid.:266–71.)RelevancetheoryavoidsthefirstofthethreeproblemsforGrice’spragmatictheory–thatitprovidesnoaccountoftherelativeweightattachedtothevariousmaxims–byhavingasingleprincipletakeovertheworkthatisdonebytheCo-operativePrincipleandthevariousmaxims.RelevancetheoryavoidsthesecondproblemforGrice’saccount–thatitdoesnoteasilyallowfortheopen-endednessofinterpretation–becauseaspeakercancommu-nicatetoahearerthatworthwhilecognitiveeffectscanbeobtainedwithouttoomucheffort,yetnotmakeplaintothehearerjustwhatthosecognitiveeffectswillbe.Rele-vancetheorythusallowsforvaryingdegreesofdeterminacyofcognitiveeffects.AtoneendofthespectrumarethecasesthatGrice’spragmatictheoryhandleswell.Thespeakermakesitcompletelyclearwhattheintendedcognitiveeffectsare,overtly133\nMARTINDAVIESintendsthehearertorecognizeaspecificimplicatureasthecommunicatedmessage,andcanbeheldasresponsibleforthetruthoftheimplicatureasifshehadasserteditoutright.Attheotherendofthespectrumarecasesinwhichthespeakerprovidesaguaranteeofrelevance,butleavesittotallyuptothehearertoexplorearangeofcog-nitiveeffects.Suchexplorationislikelytoinvolveconsiderableprocessingeffort,butthespeakercommunicatesthattheeffortwillberepaid(seeibid.:193–202).SperberandWilsonavoidthethirdproblemforGrice’saccountbygivingdetailedandsubstantiveaccountsofmetaphorandirony.Thekeyideasarethatmetaphorisakindofloosetalkwhileironyisanechoicuseoflanguage(seeibid.:ch.4).Loosetalk–aswhensomeonewholivesjustoutsidethecitylimitsofParissays,‘IliveinParis’–isapervasivefeatureofordinarylanguageuse.Relevancetheoryaccountsforloosetalkbysayingthattheutteranceofasentencethatisstrictlyspeakingfalsemaybeahighlyeffectivewayofcommunicatingawholesetofpropositionsthatcanbeeasilyderivedaslogicalorcontextualimplicationsfromtheinitialpieceof(mis-)information.Ahearercanuseapresumptionofrelevancetoselectsomeimplicationsandignoreothers.Accordingtorelevancetheory,metaphorisakindofloosetalk.Theperpetratorofametaphorintendsahearertoderiveasetofcontextualimplicationsfromapropositionthatisliterallyfalse(SperberandWilson1986a:548):Themostcreativemetaphorsrequireoftheheareragreatereffortinbuildinganappro-priatecontext,andderivingawiderangeofimplications.Ingeneral,thewidertherangeofpotentialimplicaturesandthegreaterthehearer’sresponsibilityforconstructingthem,themorecreativethemetaphor.Theechoicuseoflanguageisalsowidespread.Thusconsiderthefollowingexchange(SperberandWilson1981:556):(a)I’vegotatoothache.(b)Oh,you’vegotatoothache.Openyourmouth,andlet’shavealook.Here,thesecondspeakerechoesthefirstspeaker’sutterance,therebyindicatingthatithasbeenheardandunderstoodandalsoexpressingareactiontoit.Often,anechoicutteranceisusedtoconveyaspeaker’sattitudetowardsthethoughtofthepersonwhoseutteranceisechoed,asinthefollowingexample(SperberandWilson1995:239):He:It’salovelydayforapicnic.[Theygoforapicnicandthesunshines.]She(happily):It’salovelydayforapicnic,indeed.Itisaveryshortstepfromheretoaclearexampleofirony(ibid.):He:It’salovelydayforapicnic.[Theygoforapicnicanditrains.]She(sarcastically):It’salovelydayforapicnic,indeed.134\nPHILOSOPHYOFLANGUAGEAccordingtorelevancetheory,ironyisakindofechoicuseoflanguage:‘Thespeakerdissociatesherselffromtheopinionechoedandindicatesthatshedoesnotholditherself’(ibid.).Ingeneral,theinterpretationofanutteranceasironicalinvolvestherecognitionoftheutteranceasechoic,theidentificationofthepersonorkindofpersonwhoseutterance(realorimagined)orthoughtisbeingechoed,andtherecognitionthatthespeaker’sattitudetowardsthisthoughtisoneofrejectionordissociation(ibid.:240).Thereisanimportantsimilaritybetweenrelevancetheory’saccountofironyandtheaccountofmetaphor.Bothmetaphorandironyaresaidtoinvolvetheuseof‘verygeneralmechanismsofverbalcommunication’(ibid.:242).Thereisnoneedforadis-tinctivetheoryoffigurativelanguage;therelevance-theoryaccountsofmetaphorandironyarecontinuouswithaccountsofutterancesthatareneithermetaphoricalnorironical.ImplicationandImplicatureWecanconveysomeofthebasicideasofrelevancetheorybylookingatanexampleinvolvingaquestion,alongwithfourpossibleanswers(basedonWilsonandSperber1981).WearetoassumethatthesetofbackgroundassumptionssharedbetweenspeakerandheareralreadyincludestheassumptionthatBruceisaphilosopher,butdoesnotincludetheassumptionthatBruceisboring.Q:AreyouinvitingBrucetoyourparty?A:(a)No,IamnotinvitingBruce.(b)Bruceisaphilosopher,andIamnotinvitinganyphilosophers.(c)Iamnotinvitinganyphilosophers.(d)Iamnotinvitinganyonewhoisboring.Eachofthesefourrepliescouldservetocommunicatethesamemessage–thenegativeanswertothequestion.Thatmessageisliterallyandexplicitlyexpressedin(a),butislessdirectlyconveyedineachof(b)–(d).Inthecaseof(b),themessagethatthespeakerisnotinvitingBrucecanbeinferredfromwhatisliterallyandexplicitlysaid.Inthecaseof(c),thatmessagecanbeinferredfromwhatisliterallyandexplicitlysaidtogetherwithasharedbackgroundassumption.Inthecaseof(d),themessagecanbeinferredfromwhatisexplicitlysaidtogetherwithanadditionalpremise,namely,thepropositionthatBruceisboring.Theideaisthat,inordertomaintainthepresumptionthatthespeaker’sutteranceisrelevantincase(d),thehearerofthatanswerwillsupplytheadditionalpremise.Togiveadescriptionofeventhissimpleexample,weneedatleastfournotions:•thepropositionthatisliterallyandexplicitlyexpressedinanutterance;•apropositionthatislogicallyimpliedbywhatisexplicitlyexpressed;•apropositionthatiscontextuallyorpragmaticallyimplied–wheretheinferencereliesonboththepropositionexplicitlyexpressedandotherpremisesdrawnfromthesetofbackgroundassumptions(thecontext);and•apropositionthatisnotalreadypartofthecontext,butissuppliedasanaddi-tionalpremise.135\nMARTINDAVIESThesefournotionscanbeusedtodescribecasesofahearer’sexploitinginformationpro-videdbythespeaker,aswellascases–suchastheexamplehere–wherethespeakerovertlyintendstocommunicateamessagethatisnotexplicitlyexpressed.ItmaynotbeimmediatelyobvioushowGrice’snotionofconversationalimplicaturemapsontothisrelevancetheoryframework.Intheexamplegiven,answer(b)wouldcertainlynotbeacaseofconversationalimplicatureonGrice’saccount,since‘thetruthofaconversationalimplicatumisnotrequiredbythetruthofwhatissaid’(Grice1989:39).Answer(c)doespassthattest,butinfacttheGriceannotionofconversationalimplicaturecorrespondsmorecloselytothatofanadditionalpremise,asinanswer(d).6.4PragmaticcontributionstoexplicitcontentGrice’spragmatictheoryoffersanaccountofthewayinwhichanimplicaturethatqcanbeworkedout,giventhatthespeakerliterallyandexplicitlyadvancestheproposi-tionthatp.Thismaysuggestthatpragmaticprinciplesarenotinvolvedindiscerningthataspeakerhasexplicitlyadvancedacertainproposition.Butthatideawouldnotbecorrect.Thus,SperberandWilson(1995:175–6):Weregardverbalcommunication,then,asinvolvingtwotypesofcommunicationprocess:onebasedoncodinganddecoding,theotheronostensionandinference.Thecodedcom-municationprocessisnotautonomous:itissubservienttotheinferentialprocess.Theinferentialprocessisautonomous:itfunctionsinessentiallythesamewaywhetherornotcombinedwithcodedcommunication(thoughintheabsenceofcodedcommunication,performancesaregenerallypoorer).Thecodedcommunicationisofcourselinguistic:acoustic(orgraphic)signalsareusedtocommunicatesemanticrepresentations.Thesemanticrepresentationsrecoveredbydecodingareusefulonlyasasourceofhypothesesandevidenceforthesecondcommunicationprocess,theinferentialone.Inferentialcom-municationinvolvestheapplication,notofspecial-purposedecodingrules,butofgeneral-purposeinferencerules,whichapplytoanyconceptuallyrepresentedinformation.Onthisview,pragmaticprinciplesareinvolvedintherecognitionofthepropositionthatisexplicitlyexpressed.Theyarealsoinvolvedwhenahearerdiscernswhatattitudethespeakertakestowardsthepropositionexpressed,orwhatkindofspeechact–anasser-tion,acommand,aquestion–thespeakerisperforming.Letusconsidertherecognitionofspeechactsfirstandthentherecognitionofthepropositionexplicitlyexpressed.Ifwearetoappreciatewhatisinvolvedindiscerningwhatkindofspeechactaspeakerisperformingthenitisimportanttodistinguishtheideaofakindofspeechactfromthegrammaticalnotionofmood.Themoodofanutteredsentence–indica-tive,imperative,interrogative–ispartofitsencodedliteralmeaning.Bututteringasentenceinaparticularmooddoesnotguaranteetheperformanceofaparticularkindofspeechact.Utteringasentenceintheindicativemood,forexample,isnotthesamethingasmakinganassertion.Anindicativesentencecanbeusedtoexpressaconjec-tureortomakeajoke;itcanbeusedironicallyorquestioningly.Noristhereanyfurtherconventionalsign,overandabovethemoodoftheutteredsentence,thatguaranteesthatanactofassertionhasbeenperformed.Utteringasentenceinaserioustoneofvoice,forexample,doesnotinevitablymaketheutteranceintoanassertion.136\nPHILOSOPHYOFLANGUAGEItistruethattherearesomeconceptualconnectionsbetweenmoodsandkindsofspeechacts,andanadequatetheoryoflanguageanditsusemustspellthoseconnec-tionsout.Davidson(1979)makestheproposalthatutteringasentenceinanon-indicativemood(forexample,theimperativemoodortheinterrogativemood)isratherlikemakingapairofutterances.Anutteranceof‘Putonyourhat’isratherlikeanutteranceof‘Mynextutteranceisimperativalinforce’(or‘Mynextutteranceconsti-tutesanactofcommanding’)followedbyanutteranceoftheindicativesentence‘Youwillputonyourhat’.WhetherornotDavidson’sproposedaccountisultimatelysatisfactory,itcertainlyestablishesaconnection‘betweenthemoodindicatorsandtheideaofacertainillocutionaryact’(Davidson1984:275).Butthecrucialpointforpresentpurposesisthateventhoughthepresenceofaparticularmoodmayconstituteaprimafaciereasonfortakinganutterancetobeaspeechactofaparticularkind,stillthatprimafaciereasonmaybeoverriddenbyallmannerofotherconsiderations.Thereareno‘special-purposedecodingrules’forcalculatingwhatkindofspeechactaspeakerisperforming.Weturnnowtotherecognitionofthepropositionexplicitlyexpressed.Itisbeyonddisputethattheliteralmeaningsofwordsandthewaysinwhichwordsareputtogethertomakesentencesmayleaveusfarshortofdeterminatetruthconditions.Afamiliarexampleisprovidedbythesentence‘Iamtired’.Thissentencedoesnot,byitself,havedeterminatetruthconditions;theconditionsforthetruthofaparticularutteranceof‘Iamtired’dependonwhoisspeaking.Butthiskindofcaseoftheunderdeterminationoftruthconditionsbyliteralmeaningisapttobemisleading.Foritisveryplausiblethatthereisaruleoflanguagethatspecifiesthewayinwhichthetruthconditionsofanutteranceof‘Iamtired’dependonaparticularfeatureofthecontextinwhichitismade,namely,thespeakerinthatcontext.(Here,weusetheordinarynotionofcontext,ratherthantherelevance-theorynotionofasetofassumptions.)Ingeneral,however,thestepfromtheliteralmeaningofasentencetothetruthconditionsofanutteranceofthatsentenceincontextisnotgovernedbyanysuchneatrulesofcontext-dependence.Forapronoun,suchas‘he’,‘she’or‘it’,thereisnorulethatassignsareferenceonthebasisofsomepredeterminedfeatureofthecontextinwhichthepronounisuttered.Itisnottrue,forexample,thatthereferenceof‘he’isalwaysthenearestmaleinthecontextorthelast-mentionedmaleinthecontext.Itmaybesaidthatthereferenceof‘he’isalwaysthemostsalientmaleinthecontext;butthereisnolimittotheinforma-tionthatmight,inprinciple,beinvolvedinanassessmentofsalienceinacontext.Whatgoesfortheassignmentofreferencetopronounsgoesalsofortheremovaloflexicalambiguity,theresolutionofvaguenessorindeterminacy,andtherestorationofellipsis.Theseareallsensitivetoglobalfeaturesofthediscoursesituation.Thus,manyaspectsoftherecognitionofthepropositionexplicitlyexpressedbyanutteranceinvolvepragmaticinterpretation;theyaremattersofinferenceratherthanofdecoding(SperberandWilson1986b:585):Pragmaticinterpretationseemstoustoresemblescientifictheorizinginessentialrespects.Thespeaker’sintentionsarenotdecodedbutnon-demonstrativelyinferred,byaprocessofhypothesisformationandconfirmationwhich,likescientifictheorizingandunlikegrammaticalanalysis,hasfreeaccesstocontextualinformation.137\nMARTINDAVIES6.5UtteranceinterpretationandpubliclanguagemeaningThefactthatevenexplicitcontent–evenaninitialpropositionfromwhichfurtherinferencesmightbedrawn–dependsonpragmatic,aswellassemantic,factorsisofthegreatestimportanceforatheoreticalunderstandingofthecommunicativeuseoflanguage.Butwemustalsonotethatthisfacthasbeenusedtocastdoubtonthetheoreticalsignificanceoftheverynotionofliteralmeaninginapubliclanguage.Ahearernotonlyusesinferenceandcontextualinformationinordertobridgethegapbetweenliteralmeaningandexplicitpropositionalcontentandthentodrawoutcontextualimplications.Ahearermayalsousethosesameresourcesinordertorevisehisassessmentofaspeaker’sliteralmeaningsthemselves.Literalmeaningprovidesanoutlinethatisthenfilledinundertheguidanceofapresumptionofoptimalrelevance.Butifnowayoffillingintheoutlinemeasuresuptothispresumption,thentheoutlineitselfmayberedrawn.Thisiswhatusuallyhappenswhenaspeakermakesaslipofthetongueorcommitsamalapropism.Itmayhappen,then,thataspeakerandahearerbringtoadiscoursedifferentassignmentsofmeaningstowords.Ahearermaybeginwithonesetofexpectationsand,findingthemunfulfilled,maysetaboutdevisinganewassignmentofmeaningstowordstoserveforthediscourseinwhichheisengaged.Thisprocessofassigningnewmeaningsislikepragmaticinterpretation.Itinvolvesnon-demonstrativeinferenceratherthandecoding,andthereisnolimittotheinformationonwhichitmightdraw.Davidsonsaysthat,indevisinganewassignmentofmeanings,ahearerwilloperateinessentiallythesamewayasaradicalinterpreter:hewilltrytomakethebestsensepossibleofthespeaker’stotallifeandconduct(section4.2).Thus(Davidson1984:278):agreementonwhataspeakermeansbywhathesayscansurelybeachievedeventhoughaspeakerandhearerhavedifferentadvancetheoriesastohowtointerpretthespeaker.Thereasonthiscanbeisthatthespeakermaywellprovideadequateclues,inwhathesays,andhowandwherehesaysit,toallowahearertoarriveatacorrectinterpretation.ThenotionofconventionloomslargeinaGriceananalysisofliteralmeaning.ButthemoralthatDavidsondrawsfromconsiderationofahearer’sabilitytorevisehisassign-mentofmeaningstoaspeaker’swordsisthatconventionsare,intheend,ofmerelypracticalsignificance(ibid.:279):Knowledgeoftheconventionsoflanguageisthusapracticalcrutchtointerpretation,acrutchwecannotinpracticeaffordtodowithout–butacrutchwhich,underoptimumconditionsforcommunication,wecanintheendthrowaway,andcouldintheoryhavedonewithoutfromthestart.Aconventionisarationallyself-perpetuatingregularityinthebehaviourofmembersofagroup.Literallinguisticmeaningis,onaGriceanaccount,meaningthatisassignedasamatterofconvention.Theassignmentofmeaningsissharedbymembersofthegroupandisperpetuatedthroughtime.But,Davidsonsays,itisnotessentialforsuccessfullinguisticcommunicationthatthereshouldbeanassignmentofmeaningsthatissharedbetweenspeakerandheareroveranextendedperiodoftime.138\nPHILOSOPHYOFLANGUAGEAllthatisneededisthatthespeakerandthehearershouldbeabletoconvergeonanassignmentofmeaningatthetimeoftheircommunicativeexchange.Utteranceinterpretationalwaysinvolvestheinferentialprocessesofpragmaticinterpretation;decodingofsharedliteralmeaningsisneversufficientbyitself.Whereliteralmeaningsarenotshared,pragmaticinterpretation(orsimilarprocessesofnon-demonstrativeinference)mayinprinciplestillallowthehearertoworkoutwhatpro-positionthespeakerwastryingtocommunicate.SperberandWilson(1995:176)saythatpragmaticinterpretationisautonomouswhiledecodingisnotautonomous.TheconclusionthatDavidsondrawsisthatallutteranceinterpretationistobeseen,fortheoreticalpurposes,onthemodelofradicalinterpretation.Sincesharedliteralmeaningsarenotcrucialforsuccessfulcommunication,aphilosophicaltheoryoflinguisticcommunicationwillnotassignacrucialroletothenotionofpubliclanguagemeaning.Itisnotclearthatthefactsaboututteranceinterpretationreallylicensethecon-clusionaboutthetheoreticaldispensabilityofthenotionofliterallinguisticmeaning.NorisitobviousthatthefactsaboutcommunicationunderminetheGriceananalyticalprogramme.Afterall,Gricehimselfbeganwiththenotionofspeaker’smeaning,anotionofcommunicationthatissupposedtobeanalyticallypriortothenotionofliteralmeaning.Butweshallnotattempttoadjudicatetheseissueshere.Howmuchimportanceattachestothenotionsofasharedpubliclanguageandofliterallinguisticmeaning?Ifthenotionofliterallinguisticmeaningistoberetained,cananythingmorebesaidaboutmeaningthanmeta-semanticdeflationismsuggests?Ifthereismoretobesaid,thenwhatconceptsmightbedrawnoninasubstantivetheoryaboutthenatureofmeaning?Doesthenotionofconventionhavearoletoplay?Shouldaphilosophicalaccountofmeaningappealtotheconceptoftacitknowledgethatfiguresintheoreticallinguistics?Theseareamongtheimportantquestionsforfuturephilosophyoflanguage.FurtherReadingPhilosophyoflanguageisanareawheremostofthereadingthatisrecommendedtostudentsisintheformofarticles.ThecollectioneditedbyMartinich(2001)contains41papers.Theseoffergoodcoverageofsomeofthemaintopicsinthischapterandalsoinclude13articlesonreferringexpressionssuchasnames,definitedescriptionsanddemonstratives.PeterLudlow’sexcellentanthology,ReadingsinthePhilosophyofLanguage(1997),offers42papers,including17onreferringexpressions.AdrianMoore’smorecompactMeaningandReference(1993)has14papers,someofwhichtakeupquestionsthatwehavenoted,buthavenotdiscussed,inthischapter.EvansandMcDowell(1976)containshigh-qualitypapersatanadvancedlevel.HaleandWright(1997)offerauthoritativeandsophisticatedsurveysofmanyofthetopicstouchedoninthischapter.Usefulrecenttextbooks,atafairlyintroductorylevel,includeDevittandSterelny(1999),Lycan(2000)andTaylor(1998).Amongolderbooks,Blackburn(1984)isengagingandchallenging,andoffersgoodcoverageofbothDavidson’sandGrice’sprogrammes.MarkPlatts’sWaysofMeaning(1979;2ndedn,1997)isusefulforanaccessible–thoughfarfromneutral–accountofDavidson’sprogramme.ForDavidson’sprogramme,hisbookofessays,InquiriesintoTruthandInterpretation(1984),isessentialreading.‘TruthandMeaning’(1967a)isagoodplacetostartandiswidelyreprinted139\nMARTINDAVIESinanthologies.ForanintroductiontoChomsky’sapproachtothestudyoflanguage,thefirstchapteroftheseminalAspectsoftheTheoryofSyntax(1965)canbecombinedwithamorerecentpaper,suchas‘LanguageandProblemsofKnowledge’(1990).Chomsky(2000)isacollectionofrecentpapersengagingwithphilosophyoflanguage.LarsonandSegal(1995)offeraverythoroughgroundingintheapproachtosemantictheorythatcombineselementsfromDavidson’sandChomsky’swork,asproposedbyJamesHigginbotham.ForGrice’sprogramme,hisStudiesintheWayofWords(1989)isarichresource.‘Utterer’sMeaningandIntentions’(1969)coversmanyoftheessentialpointsinhisattempttoanalysetheconceptofliteralmeaning.ThefirstthirtypagesorsoofSchiffer(1972)provideanauthoritativeintroductiontoGrice’sprogramme,andtheintroductiontothesecondedition(1988)setstheprogrammeagainstabackgroundofmorerecentconcerns.Strawson(1970b)offersanimportantcomparisonofDavidson’sandGrice’sprogrammes.Schiffer(1987)hasbeenmassivelyinfluentialinpersuadingmostphilosophersoflanguagethatGrice’sanalyticalprogrammecannotbecarriedthrough.Itisnot,however,abookforbeginnersinthisarea.ReadingonminimalismaboutmeaningandtruthshouldbeginwithHorwich(1998)andField(1994).Grice(1975)shouldbethestarting-pointforreadingonconversationalimplicature.SperberandWilson(1986b)providesanexcellentintroductiontotheirbookRelevance:CommunicationandCognition(1995).Forreaderswhoareparticularlyinterestedinmetaphor,thefewcommentsinGrice(1975)couldleadontoSearle(1979)andDavidson(1978),andthentoSperberandWilson’sexcellent‘LooseTalk’(1986a).Davis(1991)coversthekeytopicsinpragmatictheorythoroughly(35papersinall)withclassicpapersbyGrice,severalpapersbySperberandWilson,andafinesectiononmetaphorandirony.ReferencesAvramides,A.1989:MeaningandMind:AnExaminationofaGriceanAccountofLanguage.Cambridge,MA:MITPress.Baker,G.P.andHacker,P.M.S.1984:Language,SenseandNonsense:ACriticalInvestigationintoModernTheoriesofLanguage.Oxford:Blackwell.Barwise,J.andPerry,J.1983:SituationsandAttitudes.Cambridge,MA:MITPress.Bennett,J.1976:LinguisticBehaviour.Cambridge:CambridgeUniversityPress.Blackburn,S.1984:SpreadingtheWord.Oxford:OxfordUniversityPress.Chomsky,N.1965:AspectsoftheTheoryofSyntax.Cambridge,MA:MITPress.——1976:ReflectionsonLanguage.London:Fontana/Collins.——1980:RulesandRepresentations.Oxford:Blackwell.——1986:KnowledgeofLanguage:ItsNature,Origin,andUse.NewYork:Praeger.——1990:LanguageandProblemsofKnowledge.InA.P.Martinich(ed.)(2001)ThePhiloso-phyofLanguage,4thedn.Oxford:OxfordUniversityPress.——1995:TheMinimalistProgram.Cambridge,MA:MITPress.——2000:NewHorizonsintheStudyofLanguageandMind.Cambridge:CambridgeUniversityPress.Davidson,D.1965:TheoriesofMeaningandLearnableLanguages.InY.Bar-Hillel(ed.)Logic,MethodologyandPhilosophyofScience,Vol.2.Amsterdam:NorthHolland.ReprintedinD.Davidson(1984)InquiriesintoTruthandInterpretation.Oxford:OxfordUniversityPress.——1967a:TruthandMeaning.Synthèse,17,304–23.ReprintedinDavidson(1984),Ludlow(1997),Martinich(2001)andMoore(1993).——1967b:TheLogicalFormofActionSentences.InN.Rescher(ed.)TheLogicofDecisionandAction.Pittsburgh:UniversityofPittsburghPress.ReprintedinDavidson(1980)andLudlow(1997).140\nPHILOSOPHYOFLANGUAGE——1969a:TruetotheFacts.JournalofPhilosophy,66,748–64.ReprintedinD.Davidson(1984)InquiriesintoTruthandInterpretation.Oxford:OxfordUniversityPress.——1969b:OnSayingThat.InD.DavidsonandJ.Hintikka(eds)WordsandObjections:EssaysontheWorkofW.V.Quine.Dordrecht:Reidel.ReprintedinDavidson(1984),Ludlow(1997)andMartinich(2001).——1969c:TheIndividuationofEvents.InN.Rescher(ed.)EssaysinHonorofCarlG.Hempel.Dordrecht:Reidel.ReprintedinD.Davidson(1980)EssaysonActionsandEvents.Oxford:OxfordUniversityPress.——1973:RadicalInterpretation.Dialectica,27,313–28.ReprintedinD.Davidson(1984)InquiriesintoTruthandInterpretation.Oxford:OxfordUniversityPress.——1974:BeliefandtheBasisofMeaning.Synthèse,27,309–23.ReprintedinDavidson(1984)andMartinich(2001).——1975:ThoughtandTalk.InS.Guttenplan(ed.)MindandLanguage.Oxford:OxfordUniver-sityPress.ReprintedinD.Davidson(1984)InquiriesintoTruthandInterpretation.Oxford:OxfordUniversityPress.——1978:WhatMetaphorsMean.CriticalInquiry,5,31–47.ReprintedinDavidson(1984),Davis(1991)andMartinich(2001).——1979:MoodsandPerformances.InA.Margalit(ed.)MeaningandUse.Dordrecht:Reidel.ReprintedinD.Davidson(1984)InquiriesintoTruthandInterpretation.Oxford:OxfordUniver-sityPress.——1980:EssaysonActionsandEvents.Oxford:OxfordUniversityPress.——1983:ACoherenceTheoryofTruthandKnowledge.InE.LePore(ed.)TruthandInterpre-tation:PerspectivesonthePhilosophyofDonaldDavidson.Oxford:Blackwell.——1984:InquiriesintoTruthandInterpretation.Oxford:OxfordUniversityPress.——1985:AdverbsofAction.InB.VermazenandM.B.Hintikka(eds)EssaysonDavidson:ActionsandEvents.Oxford:OxfordUniversityPress.Davies,M.1981:Meaning,Quantification,Necessity:ThemesinPhilosophicalLogic.London:RoutledgeandKeganPaul.——1982:IndividuationandtheSemanticsofDemonstratives.JournalofPhilosophicalLogic,11,287–310.ReprintedinP.Ludlow(ed.)(1997)ReadingsinthePhilosophyofLanguage.Cambridge,MA:MITPress.——1984:TayloronMeaning-theoriesandTheoriesofMeaning.Mind,93,85–90.——1986:TacitKnowledge,andtheStructureofThoughtandLanguage.InC.Travis(ed.)MeaningandInterpretation.Oxford:Blackwell.——1987:TacitKnowledgeandSemanticTheory:CanaFivepercentDifferenceMatter?Mind,96,441–62.——1989:TacitKnowledgeandSubdoxasticStates.InA.George(ed.)ReflectionsonChomsky.Oxford:Blackwell.ReprintedinC.MacdonaldandG.Macdonald(eds)(1995)PhilosophyofPsychology:DebatesonPsychologicalExplanation.Oxford:Blackwell.——1991:ActsandScenes.InN.CooperandP.Engel(eds)NewInquiriesintoMeaningandTruth.HemelHempstead:SimonandSchuster.——2000:PersonsandtheirUnderpinnings.PhilosophicalExplorations,3,43–62.Davis,S.(ed.)1991:Pragmatics:AReader.Oxford:OxfordUniversityPress.Devitt,M.andSterelny,K.1987:LanguageandReality:AnIntroductiontothePhilosophyofLanguage.Oxford:Blackwell(2ndedn,1999).Dretske,F.1981:KnowledgeandtheFlowofInformation.Oxford:Blackwell.——1986:Misrepresentation.InR.Bogdan(ed.)Belief:Form,ContentandFunction.Oxford:OxfordUniversityPress.ReprintedinS.P.StichandT.A.Warfield(eds)MentalRepresentation:AReader.Oxford:Blackwell.Dummett,M.1973:Frege:PhilosophyofLanguage.London:Duckworth.141\nMARTINDAVIES——1976:WhatisaTheoryofMeaning?(II).InG.EvansandJ.McDowell(eds)TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.ReprintedinM.Dummett(1993)TheSeasofLanguage.Oxford:OxfordUniversityPress.——1978:TruthandOtherEnigmas.London:Duckworth.——1991:TheLogicalBasisofMetaphysics.Cambridge,MA:HarvardUniversityPress.——1993:TheSeasofLanguage.Oxford:OxfordUniversityPress.Evans,G.1976:SemanticStructureandLogicalForm.InG.EvansandJ.McDowell(eds)TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.ReprintedinEvans(1985)andLudlow(1997).——1981a:SemanticTheoryandTacitKnowledge.InS.HoltzmanandC.Leich(eds),Wittgenstein:ToFollowaRule.London:RoutledgeandKeganPaul.ReprintedinG.Evans(1985)CollectedPapers.Oxford:OxfordUniversityPress.——1981b:UnderstandingDemonstratives.InH.ParretandJ.Bouveresse(eds)MeaningandUnderstanding.Berlin:W.deGruyter.ReprintedinEvans(1985)andLudlow(1997).——1982:TheVarietiesofReference.Oxford:OxfordUniversityPress.——1985:CollectedPapers.Oxford:OxfordUniversityPress.Evans,G.andMcDowell,J.(eds)1976:TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.Field,H.1986:TheDeflationaryConceptionofTruth.InG.MacdonaldandC.Wright(eds),Fact,ScienceandMorality:EssaysonA.J.Ayer’sLanguage,TruthandLogic.Oxford:Blackwell.——1994:DeflationistViewsofMeaningandContent.Mind,103,249–85.Forbes,G.1987:IndexicalsandIntensionality:AFregeanPerspective.PhilosophicalReview,96,3–31.——1989:LanguagesofPossibility:AnEssayinPhilosophicalLogic.Oxford:Blackwell.Frege,G.1892:OnSenseandReference.InP.GeachandM.Black(eds)(1970)TranslationsfromthePhilosophicalWritingsofGottlobFrege.Oxford:Blackwell.ReprintedinLudlow(1997),Martinich(2001)andMoore(1993).——1918:TheThought:ALogicalInquiry.InP.F.Strawson(ed.)(1967)PhilosophicalLogic.Oxford:OxfordUniversityPress.ReprintedinP.Ludlow(ed.)(1997)ReadingsinthePhilosophyofLanguage.Cambridge,MA:MITPress.Goldman,A.I.1971:TheIndividuationofAction.JournalofPhilosophy,68,761–74.Grandy,R.1973:Reference,MeaningandBelief.JournalofPhilosophy,70,439–52.Grice,H.P.1957:Meaning.PhilosophicalReview,66,377–88.ReprintedinP.F.Strawson(ed.)PhilosophicalLogic.Oxford:OxfordUniversityPress;Grice(1989)andMartinich(2001).——1961:TheCausalTheoryofPerception.ProceedingsoftheAristotelianSociety,Supplemen-taryVolume,35,121–52.ReprintedinGrice(1989)andinG.Warnock(ed.)(1967)ThePhilosophyofPerception.Oxford:OxfordUniversityPress.——1968:Utterer’sMeaning,Sentence-meaning,andWord-meaning.FoundationsofLanguage,4,1–18.ReprintedinGrice(1989)andinJ.Searle(ed.)(1971)PhilosophyofLanguage.Oxford:OxfordUniversityPress.——1969:Utterer’sMeaningandIntentions.PhilosophicalReview,78,147–77.ReprintedinGrice(1989)andLudlow(1997).——1975:LogicandConversation.InP.ColeandJ.Morgan(eds)SyntaxandSemantics,Volume3:SpeechActs.NewYork:AcademicPress.ReprintedinGrice(1989),Davis(1991)andMartinich(2001).——1978:FurtherNotesonLogicandConversation.InP.Cole(ed.)SyntaxandSemantics9:Pragmatics.NewYork:AcademicPress.ReprintedinH.P.Grice(1989)StudiesintheWayofWords.Cambridge,MA:HarvardUniversityPress.——1989:StudiesintheWayofWords.Cambridge,MA:HarvardUniversityPress.142\nPHILOSOPHYOFLANGUAGEHale,B.andWright,C.(eds)1997:ACompaniontothePhilosophyofLanguage.Oxford:Blackwell.Higginbotham,J.1985:OnSemantics.LinguisticInquiry,16,547–93.——1986:LinguisticTheoryandDavidson’sPrograminSemantics.InE.LePore(ed.)TruthandInterpretation:PerspectivesonthePhilosophyofDonaldDavidson.Oxford:Blackwell.——1988:IsSemanticsNecessary?ProceedingsoftheAristotelianSociety,88,219–41.——1989a:ElucidationsofMeaning.LinguisticsandPhilosophy,12,465–518.AbridgedversionreprintedinP.Ludlow(ed.)(1997)ReadingsinthePhilosophyofLanguage.Cambridge,MA:MITPress.——1989b:KnowledgeofReference.InA.George(ed.)ReflectionsonChomsky.Oxford:Blackwell.Horwich,P.1990:Truth.Oxford:Blackwell.Secondedition(1998)Oxford:OxfordUniversityPress.——1998:Meaning.Oxford:OxfordUniversityPress.Johnston,M.1988:TheEndoftheTheoryofMeaning.MindandLanguage,3,28–42.Kaplan,D.1989:Demonstratives.InJ.Almog,J.PerryandH.Wettstein(eds)ThemesfromKaplan.Oxford:OxfordUniversityPress.Kim,J.1976:EventsasProperty-exemplifications.InM.BrandandD.Walton(eds)ActionTheory.Dordrecht:Reidel.Kripke,S.1980:NamingandNecessity.Oxford:Blackwell.ExcerptsreprintedinLudlow(1997)andMartinich(2001).Larson,R.andSegal,G.1995:KnowledgeofMeaning:AnIntroductiontoSemanticTheory.Cam-bridge,MA:MITPress.ExcerptreprintedinP.Ludlow(ed.)(1997)ReadingsinthePhilosophyofLanguage.Cambridge,MA:MITPress.Laurence,S.1996:AChomskianAlternativetoConvention-basedSemantics.Mind,105,269–301.——1998:Convention-basedSemanticsandtheDevelopmentofLanguage.InP.CarruthersandJ.Boucher(eds)LanguageandThought:InterdisciplinaryThemes.Cambridge:CambridgeUniver-sityPress.Lewis,D.1969:Convention.Cambridge,MA:HarvardUniversityPress.——1975:LanguagesandLanguage.InK.Gunderson(ed.)Language,MindandKnowledge.Min-neapolis:UniversityofMinnesotaPress.ReprintedinLewis(1983)andMartinich(2001).——1983:PhilosophicalPapers:Volume1.Oxford:OxfordUniversityPress.Loar,B.1976:TwoTheoriesofMeaning.InG.EvansandJ.McDowell(eds)TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.——1981:MindandMeaning.Cambridge:CambridgeUniversityPress.Ludlow,P.(ed.)1997:ReadingsinthePhilosophyofLanguage.Cambridge,MA:MITPress.Lycan,W.G.1991:ReviewofAvramides,MeaningandMind:AnExaminationofaGriceanAccountofLanguage.MindandLanguage,6,83–6.——2000:PhilosophyofLanguage:AContemporaryIntroduction.London:Routledge.McDowell,J.1976:TruthConditions,Bivalence,andVerificationism.InG.EvansandJ.McDowell(eds)TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.Martinich,A.P.(ed.)2001:ThePhilosophyofLanguage,4thedn.Oxford:OxfordUniversityPress.Moore,A.W.(ed.)1993:MeaningandReference.Oxford:OxfordUniversityPress.Neale,S.1995:ThePhilosophicalSignificanceofGödel’sSlingshot.Mind,104,761–825.Peacocke,C.1976:TruthDefinitionsandActualLanguages.InG.EvansandJ.McDowell(eds)TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.——1981:DemonstrativeThoughtandPsychologicalExplanation.Synthèse,49,187–217.——1983:SenseandContent.Oxford:OxfordUniversityPress.——1986:Thoughts:AnEssayonContent.Oxford:Blackwell.143\nMARTINDAVIES——1989:WhenisaGrammarPsychologicallyReal?InA.George(ed.)ReflectionsonChomsky.Oxford:Blackwell.——1992:AStudyofConcepts.Cambridge,MA:MITPress.Perry,J.1977:FregeonDemonstratives.PhilosophicalReview,86,474–97.ReprintedinPerry(1993)andLudlow(1997).——1979:TheProblemoftheEssentialIndexical.Noûs,13,3–21.ReprintedinJ.Perry(1993)TheProblemoftheEssentialIndexicalandOtherEssays.Oxford:OxfordUniversityPress.——1993:TheProblemoftheEssentialIndexicalandOtherEssays.Oxford:OxfordUniversityPress.Platts,M.1979:WaysofMeaning:AnIntroductiontoaPhilosophyofLanguage.London:RoutledgeandKeganPaul.Secondedition(1997)Cambridge,MA:MITPress.Quine,W.V.O.1960:WordandObject.Cambridge,MA:MITPress.Sainsbury,R.M.1991:LogicalForms:AnIntroductiontoPhilosophicalLogic.Oxford:Blackwell.Secondedition,2001.Schiffer,S.1972:Meaning.Oxford:OxfordUniversityPress.Secondedition,1988.——1987:TheRemnantsofMeaning.Cambridge,MA:MITPress.——1993:Actual-languageRelations.InJ.E.Tomberlin(ed.)PhilosophicalPerspectives,7:LanguageandLogic.Atascadero,CA:RidgeviewPublishing.Searle,J.R.1979:Metaphor.InA.Ortony(ed.)MetaphorandThought.Cambridge:CambridgeUniversityPress.ReprintedinS.Davis(ed.)1991:Pragmatics:AReader.Oxford:OxfordUni-versityPress.Sperber,D.andWilson,D.1981:IronyandtheUse–mentionDistinction.InP.Cole(ed.)RadicalPragmatics.London:AcademicPress.ReprintedinS.Davis(ed.)1991:Pragmatics:AReader.Oxford:OxfordUniversityPress.——1986a:LooseTalk.ProceedingsoftheAristotelianSociety,86,153–71.ReprintedinS.Davis(ed.)1991:Pragmatics:AReader.Oxford:OxfordUniversityPress.——1986b:PragmaticsandModularity.InA.M.Farley,P.T.FarleyandK.-E.McCullough(eds)TheChicagoLinguisticSocietyParasessiononPragmaticsandGrammaticalTheory.ReprintedinS.Davis(ed.)1991:Pragmatics:AReader.Oxford:OxfordUniversityPress.——1995:Relevance:CommunicationandCognition,2ndedn.Oxford:Blackwell.Stoljar,D.2001:TheDeflationaryTheoryofTruth.InE.Zalta(ed.)TheStanfordEncyclopediaofPhilosophy.Stanford,CA:TheMetaphysicsResearchLabattheCenterfortheStudyofLanguageandInformation,StanfordUniversity,URL=http://plato.stanford/archives/spr2001/entries/truth-deflationary/.Strawson,P.F.1952:IntroductiontoLogicalTheory.London:Methuen.——1959:Individuals:AnEssayinDescriptiveMetaphysics.London:Methuen.——1964:IntentionandConventioninSpeechActs.PhilosophicalReview,73,439–60.ReprintedinStrawson(1971)andDavis(1991).——1970a:TheAsymmetryofSubjectsandPredicates.InH.E.KieferandM.K.Munitz(eds)Language,BeliefandMetaphysics.NewYork:StateUniversityofNewYorkPress.ReprintedinP.F.Strawson(1971)Logico-LinguisticPapers.London:Methuen.——1970b:MeaningandTruth.InauguralLectureattheUniversityofOxford.ReprintedinStrawson(1971)andMartinich(2001).——1971:Logico-LinguisticPapers.London:Methuen.——1974a:SubjectandPredicateinLogicandGrammar.London:Methuen.——1974b:OnUnderstandingtheStructureofOne’sLanguage.InFreedomandResentmentandOtherEssays.London:Methuen.ReprintedG.EvansandJ.McDowell(eds)(1976)TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.Tarski,A.1944:TheSemanticConceptionofTruthandtheFoundationsofSemantics.Philoso-phyandPhenomenologicalResearch,4,341–75.ReprintedinA.P.Martinich(ed.)2001:ThePhilosophyofLanguage,4thedn.Oxford:OxfordUniversityPress.144\nPHILOSOPHYOFLANGUAGE——1956:TheConceptofTruthinFormalizedLanguages.InA.Tarski,Logic,Semantics,Meta-mathematics.Oxford:OxfordUniversityPress.Taylor,B.1976:StatesofAffairs.InG.EvansandJ.McDowell(eds)(1976)TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.——1982:OntheNeedforaMeaning-theoryinaTheoryofMeaning.Mind,91,183–200.——1985:ModesofOccurrence:Verbs,AdverbsandEvents.Oxford:Blackwell.Taylor,K.1998:TruthandMeaning:AnIntroductiontothePhilosophyofLanguage.Oxford:Blackwell.Wiggins,D.1976:Thedere‘Must’:ANoteontheLogicalFormofEssentialistClaims.InG.EvansandJ.McDowell(eds)TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.——1980:WhatWouldbeaSubstantialTheoryofTruth?InZ.vanStraaten(ed.)PhilosophicalSubjects.Oxford:OxfordUniversityPress.——1986:VerbsandAdverbs,andSomeOtherModesofGrammaticalCombination.Proceed-ingsoftheAristotelianSociety,86,273–304.——1997:MeaningandTruthConditions:FromFrege’sGrandDesigntoDavidson’s.InB.HaleandC.Wright(eds)1997:ACompaniontothePhilosophyofLanguage.Oxford:Blackwell.Wilson,D.andSperber,D.1981:OnGrice’sTheoryofConversation.InP.Werth(ed.)Conversa-tionandDiscourse.London:CroomHelm.Wittgenstein,L.1953:PhilosophicalInvestigations.Oxford:Blackwell.Wright,C.1986:TheoriesofMeaningandSpeakers’Knowledge.InC.Wright,Realism,MeaningandTruth.Oxford:Blackwell.DiscussionQuestions1Isthemeaningofasentenceakindofthing?2Howisthenotionofmeaningrelatedtothenotionoftruth?3Istheresuchathingastheliteralmeaningoflinguisticexpressions?4Whatrelationsexistbetweensemanticandmeta-semantictheoriesofmeaning?5Aretheregoodreasonstodevelopanaxiomatizedtheoryofmeaning?6Howdoesthedifferencebetweensubjecttermsandpredicatetermshelpustounderstandthenatureofsentences?7Iftwotermspickoutthesameobject,caneverythingthatissaidaboutthemeaningofonetermbeequallysaidaboutthemeaningoftheother?8Isthereanorderofpriorityamongphilosophyoflanguage,philosophyofmindandmetaphysics?Ifso,whatsortofpriorityisinvolved?9Doesthemetaphysicaldistinctionbetweenobjectsandpropertieshelpustounder-standthelinguisticdistinctionbetweensubjectsandpredicates?10Whydopredicateexpressionshavenegations,whilesubjectexpressionsdonot?11Doesunderstandingsentencesofaformallanguagehelpustounderstandsentencesofanaturallanguage?12Shouldtheproperformatfortheoremsofaxiomatizedsemantictheoriessaywhatsentencesmeanorgivetheconditionsunderwhichtheyaretrue?13Isitpossibletoavoidtheconclusionthatforasemantictheoryspecifyingmean-ings,alltruesentencesareequivalent?14Doesthenotionofapossiblelanguagehelptodeterminewhetherasemantictheoryiscorrectfortheactuallanguageofagivengroup?15HowdoestheconditionofadequacyimposedbyTarski’sConventionThelptoelu-cidatethesemanticconceptoftruthanditsconnectionwiththeconceptofmeaning?145\nMARTINDAVIES16Howmighttheneedsofinterpretationleadustospecifyconditionsofadequacyonspecificationsofmeaningsoroftruthconditions?17Does‘radicalinterpretation’helptoestablishalinkbetweensemantictheoriesandinterpretation?18Cantruthconditionalsemantictheoriesgiveanadequatetreatmentofsentencesaboutourmentalstates?19Doesthenotionof‘tacitknowledge’helptoestablishtheexplanatoryrelevanceofsystematicsemantictheories?20Howvaluableisthenotionofarigiddesignatorfordealingwithquestionsaboutessentialism?21Doesordinarytalkaboutpeopledoingthingscarryametaphysicalcommitmenttoanontologyofevents?22Canwesatisfactorilyanalysetheconceptofliteralmeaningintermsofspeaker’smeaningandconventionalregularity?23Dopropositionalattitudeslikebelievingandintendinghavearoleintheanalysisoftheconceptoflinguisticmeaning?24Istheclaimthat‘meaningisuse’asubstantivetheoryconcerningthenotionofmeaning?25Howmightminimalisttheoriesoftruthandmeaningsupportoneanother?26Howimportantisthedistinctionbetweensemanticsandpragmaticsforanaccountofmeaning?27Howcanwecommunicatemorethanweliterallysay?28Doesatheoryofconversationalimplicatureneedoneoverarchingmaxim?Canaprincipleofrelevancesatisfythisneed?29Arepragmatic,aswellassemantic,factorsinvolvedindeterminingtheexplicitcontentofwhatwesay?30Isasharedassignmentofmeaningsanessentialfeatureoflinguisticcommunication?31Isthereanysignificantdifferencebetweenordinaryandradicalinterpretation?146\n4PhilosophyofLogicA.W.MOORELogicisconcernedwithreasoning.Inparticular,itisconcernedwiththedistinctionbetweengoodandbadreasoning.Oneofitsaimsistoformulatecompletelygeneralcriteriaforthisdistinction.Thisislargelyamatterofmakingprecisetheconditionsunderwhichanargumentisvalid.Itisalsoamatterofsayingwhatvalidityis.Thelatteristheprerogativeofthephilosophyoflogic.Thephilosophyoflogicisthephilosophicalcoreoflogic,inwhichtheessenceoftherestofthedisciplinecomesunderscrutiny.Sincemuchofourreasoning,ifnotallofit,isineliminablyverbal,alargepartofthephilosophyoflogicisconcernedwithquestionsaboutlanguage.Thischaptertracesaroutethroughsuchquestions,amongothers,takingasitsstarting-pointacharacteri-zationoflogicproferredbyP.F.Strawson.BecauseofthenatureofStrawson’schar-acterization,twoideasservetostructuretheentirediscussion:theideaofaproposition,thatis,theideaofthatwhichistrueorfalse;andtheideaofpossibility.1IntroductionP.F.Strawsononcecharacterizedlogicas‘thegeneraltheoryoftheproposition’(Strawson1967:1).Atthislevelofabstractionwecandonobetter.Butwhatisaproposition?Andwhatissuchatheory?Aninitialanswermightbe:Apropositionisthesortofthingthatistrueorfalse.Logicdealsprimarilywithcertainrelationsthatholdbetweenpropositions,namelythosewhichdependonwhichcombinationsoftruthandfalsityarepossible.Theserelationsarecalledlogicalrelations.Twowhichareofcentralconcernareconsistencyandconse-quence.Apropositionisconsistentwithotherswhenitispossibleforthemallcol-lectivelytobetrue.(Thetruthoftheothersallowsforthetruthofit.)Apropositionisaconsequenceofotherswhenitisnotpossibleforalloftheotherstobetrueyetittobefalse.(Thetruthoftheothersdemandsthetruthofit:iftheyarealltrue,thenitmustbetruetoo.)Consequenceistherelationthatholdsbetweenthepremisesandtheconclusionofavalidargument.\nA.W.MOOREEachpartofthisanswerrequiresclarification.Inparticular,attentionsettlesonthethreekeyterms‘true’,‘false’and‘possible’.Theinterestingthingisthat,becauselogicalsohasaphilosophicalcomponent,theveryprocessofprovidingsuchclarificationisitselfpartoflogic.Moreprecisely,itispartofphilosophicallogic.Logicsubdividesintotwo:formallogicandphilosophicallogic.Formallogicisconcernedwiththecodificationoftheselogicalrelations.Philosophicallogic,orthephilosophyoflogic–unlikeotherwriters,Iintendnodistinctionbetweentheseterms–isconcernedmoregenerallywithwhatmakestheserelationstherelationsthattheyare.ThusphilosophicallogicstandstoformallogicinsomethingliketherelationinwhichthePHILOSOPHYOFSCIENCE(chapter9)standstoscience.Oneisengagedintheformerwhenoneenquiresintothecharacter,scopeandpreconditionsofthelatter.Verywell,then;letussetaboutclarifyingtheroughaccountoflogicprofferedabove,beginningwiththeideaofaproposition.2Propositions‘Apropositionisthesortofthingthatistrueorfalse.’Doesthismeanthatpropositionsarebitsoflanguage?Morespecifically,doesitmeanthattheyaredeclarativesentences?Supposewesayyes.Still,itbehovesustosayconsiderablymore.Quiteapartfromthemanygrammaticalandphilosophicalissuesthatarisefromdistinguishingdeclarativesentencesfromotherbitsoflanguage,thereare(forwantofabetterword)metaphysi-calissuesthatariseconcerningwhatstatus‘bitsoflanguage’aresupposedtohaveinthiscontext.Consider:howmanysentencesarethereintheboxbelow?THECATISONTHEMATTHEREISANAPPLEONTHETABLETHECATISONTHEMATOnecouldaswellanswertwoasthree.Thereisnospecialmysteryaboutthis,noronecorrectanswer.Itisclearthatthequestioncanbetakenintwoways,anditisreason-ablyclearalsowhatthetwowaysare.Philosophershaveintroducedsometechnicaltermstomarkthedistinction.Inthiscasetheywouldsaythattherearethreesentence-tokensandtwosentence-types:twoofthetokensareofthesametype.Philosophersspeaksimilarlyofword-tokens,letter-tokensandsoforth,andoftheircorrespondingtypes.Iftheprovinceoflogicisdeclarativesentences,thenthequestionarises:declara-tivesentence-tokensordeclarativesentence-types?Themoreattractiveanswer,inmanyways,isdeclarativesentence-types.Itseemsalientotheveryspiritofthedisciplinetosupposethatthesubjectmatteroflogicisnothingbutcertaininscriptionsornoisesthathappentohavebeenmade.Surely,whenlogiciansconsiderpropositionsoftheform‘NoAsareBs’,say,whattheymeantobeconsideringaretheindefinitelymanypropositionsofthatformthatcouldbeproduced,eventhoughmostofthemhavenevercomewithintheambitofanyone’sconscious-nessandprobablyneverwill.Typesarebettersuitedtothisrolethantokens.For,oncethetype–tokendistinctionhasbeendrawn,thereisnoobstacletoadmittingtypesthat148\nPHILOSOPHYOFLOGIClacktokens.Asentence-typecanbethoughtofasaparticularsequenceofword-types.Itistruethateachoftheword-typeswillhavetokens.(Thisissomethingofapallia-tiveforthosewhoareconcernedthatsentence-typesareundulyabstract.)Butitmaywellbethatnoonehaseveractuallyconcatenatedthosewordsinthatordertoproducethatsentence.Theproblemwithtypes,infact,ismorebasic.Theyarenotthesortofthingsthataretrueorfalse.Considerthetype‘Iamhungry’.Initself,thisisneithertruenorfalse.Atmostitistruerelativetocertainpeopleandcertaintimes,oras(potentially)utteredbycertainpeopleatcertaintimes.Ofcourse,wecouldsurrendertheideathatpropo-sitionsare,‘inthemselves’,eithertrueorfalse,andwecoulddismissthecontrarysuggestionasoneofthemanyinfelicitiesoftheroughaccountoflogicwithwhichwestarted.Butthatwouldmeanpayingaprice:therewouldbeaconsiderablelossofsimplicity.Analternativewouldperhapsbetofocusonsentencesthatdidnotcontainanyofthetroublesomewordsorconstructions,suchas‘I’,whichoccasionthiskindofcontext-dependence.Whetherthereareanysuchsentencesinnaturallanguageisdubious.PerhapstheyexistinthelanguageofMATHEMATICS(chapter11).Perhaps‘2+2=4’isanexample.Butourhope,ifwetookthisapproach,wouldbethatsuchsentencescouldatleastbeartificiallyconstructed.Afurtherhopewouldbethatsuchsentencescouldbeusedtosayanythingthatcouldbesaidbymeansofanyotherdeclarativesentences.(Weshouldhavetohopethisifwedidnotwantlogictoberestrictedinitsapplication.Anditisanimportantpointthatitwouldbeafurtherhope.Someonemightwellagreethatsuchsentencescouldbeconstructedandyetbescepticalabouthowmuchcouldbeachievedbytheirmeans.)Anexampleofsuchasentencemightbesomethingoftheform‘Sishungryatt’,wherethe‘is’wastobeunderstoodtenselessly,andwherethetermsreplacingSandtpickedoutapersonandatimerespectively.Itmightbeprotestedthattherewouldstillbeadimensionofrelativity,inthatasentencecouldmeandifferentthingsindifferentlanguages,andsobetrueinonelanguageandfalseinanother.Butwecouldstipulatethatbya‘sentence’wemeantsomethingconsideredasbelongingtoaparticularlanguage.No,therealproblemwiththisapproachisthatitwouldrepresentlogicasbeingconcernedwithentitieswhosenatureandveryexistencewereamatterofgenuinecontroversy.Someofthiscontroversywillcometotheforelater(seesection3.1below).Theoriginalobjectiontoconstruingpropositionsasdeclarativesentence-typeswasthatdeclarativesentence-typesarenotinthemselvesthesortofthingsthataretrueorfalse.Beforewetrycircumventingourcurrentdifficultiesbyrevertingtodeclarativesentence-tokens,letitbenotedthatneitherarethey,inthemselves,thesortofthingsthataretrueorfalse.Considerthethreetokensintheboxonpage148,forexample.Noneoftheseiseithertrueorfalse.Thisisbecausetheyappearthereforpurelyillus-trativepurposes.Theyarenotbeingusedtoassertanything.(Infactitwouldbeimpos-siblefortokensofthosetypestobeusedtoassertanythingwithoutagooddealofstage-settingwhichI,writingthischapter,lack.Whichcat?Whichmat?Whattable?)Evenwhenatokenisusedtoassertsomethingweshouldbewaryofclassifyingitastrueorfalsewithoutqualification.SupposethatSmith,ashesetsofftoameeting,pinsanotetohisdoorsaying‘Iamout’.Andsupposethatthefollowingday,becausehedoesnotwanttobedisturbed,hepinstheverysamenotetohisdoor.Thenthesentence-tokenonhisnoteexhibitsadisconcertinginconstancy:firsttrue,thenfalse.149\nA.W.MOORE(Oraretheretwodistinctsentence-tokenshere?Issomethingasentence-tokenonlyasusedinacertainway?)Thesedifficultiescombinetosuggestthatwedobesttostopthinkingofpropositionsasbitsoflanguageatall.Thedifficultiesalsosuggestamoresatisfactoryconstrual.Propositionsarewhatareactuallyasserted,or,moregenerally,whatcouldbeasserted,bymeansofbitsoflanguage.Smithpinshisnotetohisdooronedayandtherebyasserts,orperhapsconveys,atrueproposition.Hedoesthesamethingthefollowingdayandasserts,orconveys,aquitedifferentproposition,thistimefalse.WhenIsay‘Iamhungry’Iassertaparticularproposition.You,usingaquitedifferentsentence,sayofmealittlewhilelater,‘Hewashungry’,andasserttheverysameproposition.Propo-sitionsarethemessagecarriedbythelinguisticmedium.Buttheydonotdependfortheirexistenceonthecontingenciesandvicissitudesofactualcommunication.Therearecountlesspropositionsthatareneverasserted.Logicisthusabletoretainthatauraofabstractionwhichfirstinclinedusagainstconstruingpropositionsassentence-tokens.Ontheotherhand,propositionsareentitieswithwhichwecanhavedirectcommerce,astheyhadbetterbeiflogicistoimpingeonourordinarythinkingandreasoninginthewaythatitcertainlydoes.Indeed,theyenjoyafurtheradvantagehereoversentences(whethertypesortokens).Forourcommercewithpropositionscantakeavarietyofforms.Theycanbethought,orentertained,orhoped,evenwherenosentenceisdirectlyinvolved,andcertainlywherenosentenceisbeingusedtoassertanything.Thereisthereforenopuzzleabouthowlogiccanhaveapplicationinsuchcases,againasitcertainlycan.(Myunexpressedhopescanbeinconsistent.)Aboveall,propositions,onthisconstrual,aretrueorfalse.Whatbetterthenthantothinkintheseterms?Strawsonhimselfcertainlyseemedtobedoingthatwhen,inthecontextalreadyreferredto,hecharacterizedpropositionsaswhatwespecifyby‘that’-clauseswhenwesaythingsoftheform‘Ihopethatp’,‘Supposethatp’,‘Itfollowsthatp’,‘Itispossiblethatp’andsuchlike.Butwemustbecautious.Thisaccount,forallitsattractions,andpartly,nodoubt,becauseofthem,glidesoversomeofthedeepestproblemsinthephilosophyoflogic.Thatwhichlanguageisusedtoconvey–MEANING(chapter3)–wouldberegardedbymanyastheverystuffoflogic.Totakeitforgranted,ashasjusteffectivelybeendone,istotakeforgrantedagooddealofwhatwe,asphilosophersoflogic,oughtpreciselytobequestioning.Howexactly,forexample,doesyourutteranceof‘Hewashungry’relatetomyutteranceof‘Iamhungry’?Inwhatwaycantwoutterancesshareameaningwhenthesentence-typesconcerneddonot?Isitmerelycaptioustopointoutthatyourutterance,unlikemine,conveys(affirms?presupposes?)mysex?Arepropo-sitionscompletelyextra-linguistic?Canthesamepropositionbeassertedbysaying‘Thereisanappleonthetable’asbysaying‘Ilyaunepommesurlatable’?Ifso,couldthatpropositionhaveexistedeveniftherehadbeennolanguage?Somephilosophers,notablyW.V.O.Quine,arescepticalaboutwhetherthereisanywaythroughthismorassofquestions;andtheyturntheirscepticismonthequestionsthemselves.Quinegivesanaccountoflogicwhichisnotintermsofmeaningandwhicheschewspropositions,onthisconstrual,altogether(Quine1970:ch.1).Oneelegantwaytocutthroughtheentiredebatewouldbetostoptryingtoidentifypropositionswithanythingextra-logical.Thatis,wecouldthinkoflogicashavingitsowndistinctivesubjectmatter,akindofidealizationofanythingthatwemight150\nPHILOSOPHYOFLOGICnormallythinkofastrueorfalse.Thissubjectmattercouldthenserveasamodelforvariousaspectsofourordinarythinkingandreasoning.Propositionswouldberatherliketherationalagentsstudiedbyeconomists,orthefrictionlessplanesstudiedbyphysicists.Oneadvantageoftakingthisapproachisthatitwouldvindicatebringingtobearasimpletrue–falsedichotomyonthevagueness,indeterminacy,ambiguity,inexactitude,hyperbole,figurativenessandvividcolouringofsomuchofwhatweordinarilysayandthink;andonthecountlessusestowhichweputdeclarativesentenceswhereitisunclearwhetheranassessmentintermsoftruthandfalsityisevenappropriate,stilllesswhatprinciplesshouldgovernitifitis:wetelljokes,wewritefiction,wedefinewords,wepraise,blame,approveanddisapprove.Forthereisnodoubt,again,thatlogiccanhaveapplicationwhenwedothesethings.Ajokecanbeinternallyinconsistent.Oneexpressionofapprovalcanbeaconsequenceofothers.Onthisapproachwecouldseethesefactsasbeingnomorepuzzlingthanthefactthatacarcanmovealongaroadinaccordancewithprinciplesaboutfrictionlessplanes.Anotheradvantageofthisapproachisthatitwouldtakeanymystiqueoutofthefact,whichwasperhapsbegin-ningtoassumeaspurioussignificance,thatpropositionsarethesortofthingsthataretrueorfalse.That,now,couldbeviewedasnothingmorethanadefiningcharacteris-tic(cf.Wittgenstein1967:§§136;1961:6.111).Therewouldstillbemuchthatneededtobesaidaboutwhatpropositionsarelike.Buttherewouldbesomeprospectnowthatitcouldbesaid,indirectly,throughworkinformallogic,ratherthanaspartofsomephilosophicalpropaedeutic.Letusturnnowtotheotherkeytermthatoccurredintheoriginalroughaccountoflogic:‘possible’.3PossibilityItisafundamentalfeatureoflogicalappraisalthatitrelatestopossiblecombinationsoftruthandfalsity,andnot,directly,totruthandfalsitythemselves.Inparticular,avalidargumentcanhaveafalseconclusion,andaninvalidargumentcanhaveatrueone.Considerthetwofollowing(somewhatregimented)arguments.MothersarealwaysshorterthantheirfullygrownsonsGeorgeisfullygrownElizabethisGeorge’smotherThereforeElizabethisshorterthanGeorgeIfheisguilty,thenhisalibiisflawedHisalibiisflawedThereforeheisguiltyThefirstoftheseargumentsisvalid.ItisimmaterialifElizabethisinfacttallerthanGeorge.Thesecondargumentisinvalid.Again,itisimmaterialifthepersonconcernedisinfactguilty.Inassessinganargumentasvalidorinvalid,weareinterestedonlyinwhethertheconclusionisaconsequenceofthepremises.Itdoesnotmatterhowtheconclusionandthepremisesstandinrelationtoreality.Whatmattersishowtheystand151\nA.W.MOOREinrelationtooneanother.(ItisimpossibleformothersalwaystobeshorterthantheirfullygrownsonsifthisdoesnotincludeElizabethandherfullygrownson,George.Itisnotimpossibleforaman’salibi,whichisflawedifheisguilty,tobeflawedanyway.)Thereisnothingmoretothis,really,thanlinguisticstipulation.Itissimplyaquestionofhowwearegoingtousetheword‘valid’.Suchlinguisticstipulationdoes,however,pointtoaproblem.Sinceourchiefinter-estwithrespecttothetruthorfalsityofpropositionsistoknowwhichareactuallytrueandwhichareactuallyfalse,thequestionarises:whyislogicalappraisalofanyinterest?Theanswer,inanutshell,isthatalthoughlogicalappraisaldoesnotbeardirectlyonwhichpropositionsaretrueandwhicharefalse,itdoesdosoindirectly.Ifanargu-mentisvalid,anditspremisesaretrue,thenitsconclusionistrue.Ifanargumentisvalidanditsconclusionisfalse,thenatleastoneofitspremisesisfalse.Ifonepropo-sitionisinconsistentwithothers,thenit,oroneoftheothers,isfalse.Logicalappraisalthushasavitalnormativeroletoplay.Givenvariousthingsthatwealreadyknow,itcandeterminewhichotherthings,beyondthose,weoughttothinkorareentitledtothink.Itcansimilarlyguideuswheneverweareexploringdifferentoptionsforwhattothink,whichisitselfoftenanindispensableprecursortochoosingtherightone.Weshallmakesingularlylittleprogressinoureffortstodiscriminatebetweenthetrueandthefalsewithoutsomearticulatedawarenessofwhatthepossibilitiesare,andinpar-ticular,withoutthekindofawarenessthatenablesustotell,bydintofreasoning,howthethingsthatwealreadyknowandthingsthatwedonotyetknowarecapableoffittingtogether.Toseefurtherwhatkindofpossibilityisinvolvedhere,andhowitisinvolved,letuslookatsomeexamplesofreasoning.Astheexampleswillservetoremindus,reason-ingrarelytakestheratherstiltedformexemplifiedabove:rarelyisoneinvolvedinassess-inganargumentwhosepremisesandconclusionareallstatedexplicitly,inthatorder.Considerthefollowingthreeproblems.(1)Sevencardsarelaidoutinfrontofyouandyouseethefollowingarray.ACCBGACYouaretoldthateachcardalsohasasingleletteronthereverse.Question:Whichcards,ifany,doyounotneedtoturnovertotestthehypothesisthateverycardwithan‘A’ononesidehasa‘C’ontheother?(2)Amanneedstotransportalion,agoatandacabbageacrossariver.Hecanonlytransportoneatatime.Ifthelionisleftunattendedwiththegoat,itwilleatit.Ifthegoatisleftunattendedwiththecabbage,itwilleatit.Question:Whatshouldhedo?(3)AmanwalksfromAtoB,settingoffat2p.m.andarrivinganhourlaterat3p.m.ThefollowingdayhetakesthesameroutebackfromBtoA.Againhesetsoffat2p.m.andagainhearrivesat3p.m.Youarenottoldwhetherhisrateofprogressisuniformoneitherday.152\nPHILOSOPHYOFLOGICQuestion:Isthereguaranteedtobesomepoint,ontheroutefromAtoB,suchthatthemanreachesthatpointatpreciselythesametimeonbothdays?Herearethethreesolutions.(1)Youdonotneedtoturnoverthethreecardsshowinga‘C’.Ineveryothercaseitwouldbepossibleforaletteronthereversetoupsetthehypothesis–inthecaseofthetwocardsshowingan‘A’,bybeinganythingotherthana‘C’,andinthecaseofthecardsshowinga‘B’ora‘G’,bybeingan‘A’.Inthecaseofthethreecardsshowinga‘C’,however,thehypothesiswouldremainintactwhateverwasonthereverse.(2)Themanshouldfirsttransportthegoatover.Thenheshouldreturntopickupeitherthelionorthecabbageandtransportitover.Thenheshouldreturn,withthegoat,topickupwhicheverofthelionandthecabbagehehasnotyettransportedoveranddoso,leavingthegoatbehind.Finally,heshouldreturntopickupthegoatandtransportitover.Eachmoveiseffectivelydictated.Thisisbecause,forthepurposesofthisproblem,twopositions,whichdifferonlyinthatthelionandthecabbageareinterchanged,areequivalent.Sothemaneffectivelyhas,ateachstage,onlyoneoptionwhichwillnotleaddirectlybacktoapositionthathehasalreadybeenin.(3)Yes,thereisguaranteedtobesuchapoint.Theeasiestwaytoseewhyistothink,notofonemanundertakingthesetwojourneysonsuccessivedays,butoftwomenundertakingequivalentjourneysonthesameday.OneofthemtakesanhourtowalkfromAtoB.Theother,duringthatsameperiod,walksfromBtoA.Thequestionposedisequivalenttothequestionofwhetherthetwomenareguaranteedtopasseachother.Andofcourse,theyare.Thereisafundamentalthree-stepprocesswhichoccursatdifferentpointsinallthreesolutions.Thefirststepofthisprocessistoidentifysomerelevantrangeofpossibilities;thesecondstepistoeliminatesomeofthesepossibilitiesusingtheparametersofthequestion;andthethirdstepistoreachaverdictonsomespecifiedquestionconcerningthepossibilitiesthatremain.Applicationoftheprocessmaywellinvolvesomespecialtrickormethod:thereisnosinglemechanicalprocedurethatgovernsallourreasoning.(Wecangofurther.Thereisnosinglemechanicalprocedurefordeterminingwhetherornotanarbitraryargumentisvalid.Thisissomethingthatcanactuallybeproved:itiscalledChurch’stheorem(seeBoolosandJeffrey1989:ch.15).)Letusidentifyapplicationsoftheprocesswithineachofthethreesolutions.Problem(1)Heretheprocessisappliedtoeachindividualcard.Thefirststep,inthiscase,istoidentify,astherelevantrangeofpossibilities,thedif-ferentthingsthatcanappearoneachsideofthecard.Thesecondstepistoeliminateallthosepossibilitiesthatconflictwiththeinformationgivenandthestatedhypothesis.Thethirdstepistodetermine,ofthepossibilitiesthatremain,whethertheyexcludeanyletter’sappearingonthereverseofthecard:ifso,thisisoneofthecardsthatneedstobeturnedover.153\nA.W.MOOREProblem(2)Heretheprocessisappliedateachindividualcrossing.Thefirststepistoidentifyastherelevantrangeofpossibilitiesthedifferentcombi-nationsofman,lion,goatandcabbagewhichcanundertakethatcrossing.Thesecondstepistoeliminateallthepossibilitiesthatviolatetheconstraintsoftheproblem.Thethirdstepistodetermine,ofthepossibilitiesthatremain,whichconstituteprogressrelativetothedesiredoveralloutcome.Thekeyhereistorecognizetheequivalence,forthepurposesoftheproblem,ofsomeofthepositions.Problem(3)Heretheprocessisappliedglobally.ThefirststepistoidentifyasthepossibilitiesthedifferentwaysinwhichthemancanwalkfromAtoBandbackagain.Thesecondstepistoeliminateallexceptthepos-sibilitiesthatrespecttheconstraintsoftheproblem.Butherethethirdstepisalreadyrelevant.Thetrickistoallowthesecondsteptobeinformedbythethird.Exacttimes,wecansee,areirrelevanttotheeventualsolutiontotheproblem.Sotheuneliminatedpossibilitiescanbethoughtofsimplyasthoseinwhicheachjourneylastsanhour.Thenfurtherabstractionshowsthateventhefactthatthejourneysoccursequentiallyisirrelevant:oneofthemcanbethoughtofasbeingsuperimposedontheother.(Itisanobviousheuristicaidtothinkoftwomenundertakingsuchjourneysatthesametime.)Thethirdstep,havingdonethiswork,isnowaformality.Itistodetermine,ofthepossibilitiesthatremain,whetherthereisanythatdoesnotinvolveacross-overpoint.Allofthismayseemafarcryfrom‘thegeneraltheoryoftheproposition’.Infact,however,thesetechniquesofidentifying,eliminatingandinvestigatingpossibilities,dif-ferentlyhighlightedineachofthethreecases,arebutdifferentmeansofascertainingpossiblecombinationsoftruthandfalsityforagivenrangeofpropositions.Theiden-tificationofpossibilitiesisamatterofspecifyingsomerelevantfeaturetowhichthetruthorfalsityofthepropositionsissensitive.Theeliminationofpossibilitiesisamatterofsupposingsomeofthosepropositionstobetrueorfalse.Andtheinvestigationoftheremainingpossibilitiesisamatterofascertainingwhatfollowsconcerningthetruthorfalsityofotherpropositions.Thusifitistruethateverycardwithan‘A’ononesidehasa‘C’ontheother,anditistruethataspecifiedcardhasan‘A’ononeside,thencertainotherpropositionsconcerningthatcardmustbefalse–forexample,thattheletterontheothersideisa‘W’.Thethreepropositionsthateverycardwithan‘A’ononesidehasa‘C’ontheother,thataparticularcardhasan‘A’ononesideandthattheletterontheothersideofthatsamecardisa‘W’areinconsistent:itisnotpossibleforthemallcollectivelytobetrue.Infacttheinitialpossibilitiesarethemselves(inthelastanaly-sis)propositions.Itistheirtruthorfalsity,ultimately,whichdeterminesthetruthorfalsityofalltheotherpropositionsunderconsideration,inaccordancewithwhichcombinationsoftruthandfalsityarepossible.Butwhatkindofpossibilityisthis?Isthereoneparticularkindofpossibilitythatisthespecialconcernoflogic?Isthere,forthatmatter,oneparticularkindofpossibilitythatisinvolvedinallthreeoftheseproblems?154\nPHILOSOPHYOFLOGICThefirstthingweshouldbeclearabouthereisthatthereareindeeddifferentkindsofpossibility.Theseareofvaryingdegreesofstringency.Forexample,itisscientificallypossible,thoughnottechnologicallypossible,formetotravelroundtheworldinaminute.Itismathematicallypossible,thoughnotscientificallypossible,formetotravelroundtheworldinamicrosecond.(Itisscientificallyimpossiblebecauseitwouldmeanviolatingthephysicallawthatnothingcantravelfasterthanthespeedoflight.)Ontheotherhand,itisnotevenmathematicallypossibleformetotravelroundtheworldinfi-nitelymanytimesinanhour,providedthatmyspeedisconstant.Itismathematicallypossibleformetotravelroundtheworldinfinitelymanytimesinanhourbydoingsothefirsttimeinhalfanhourandthendoublingmyspeedaftereachcircuit.Inotherwaysthisisnotpossible.Exactlywhichtheseotherwaysareisamatterofdeepphilosophicalcontroversy.Howeverthatmaybe,itisnaturaltosupposethattheredoeseventuallycomesomeone,minimallystringent,kindofpossibilitywhichcoversalltheothersandwhichisthespecialconcernoflogic.Thus,forexample,itisabsolutelyimpossible–wemightsay,logicallyimpossible–formetotravelroundtheworldandnotdosoatthesametime.(Actually,wehadbetterbewareofdescribingthisoranyothersenseas‘minimallystringent’.Imperfectionsinourownmentalfacultiesmeanthattherewillalwaysbeanepistemickindofpossibilitywhichcutsrightacrossalltheothers.Thereisalwaysachance,givenanimpossibilityofsomeotherkind,thatitwillbeanepistemicpossibilitybecausewhateveritisaboutitwhichmakesitnotapos-sibilityistoocomplexorsubtleforustograsp.Butputtingtoonesidethatcomplication,theideathatthereisanultimate,absolutekindofpossibilityseemsright.Andsodoestheideathatlogicisconcernedwithit(seeWittgenstein1961:6.3–6.3751).)Itisextraordinarilydifficult,however,togiveanadequateaccountofthiskindofpossibility,oreventoagreeonwhethervariousproblematiccasesareinstancesofit.Isit,forexample,absolutelyimpossiblefortodaytobeThursdayiftomorrowisSaturday?Orfor2+2tobeequalto5?Orforawatermoleculetobecomposedofhydrogenandcarbon?Orforasurfacetobebothwhiteandtransparent?Or,indeed,forthemaninvolvedinproblem(3)toavoidreachingsomepointatthesametimeonbothdays?Couldhenotdosoif,say,spaceandtimewereeachquantized(insomesuitablychunkyway)?Attemptstoexplicateabsolutepossibility–orlogicalpossibility,aswemightaswellcallit–fallintofourbroadcategories.SomecastitasaMETAPHYSICAL(chapter2)notion;someasanidealizedEPISTEMIC(chapter1)notion;someasaSEMANTIC(chapter3)notion;andsomeasasyntacticnotion.Onthemetaphysicalapproach,logicalpos-sibilityisamatterofthebroadandultimatestructureofreality.Oneachoftheotherthreeapproaches,thelogicallypossibleiswhatcannotberuledoutinacertainway;orrather,toputitinitsmoreconvenientinvertedform,thelogicallyimpossibleiswhatcanberuledoutinacertainway.Ontheepistemicapproach,somethingislogicallyimpossibleifitcanberuledoutapriori,thatis,roughly,withoutdirectappealtoex-perience.Onthesemanticapproach,somethingislogicallyimpossibleifitcanberuledoutpurelybyvirtueofthemeaningofitslinguisticexpression.Andonthesyntacticapproach,somethingislogicallyimpossibleifitcanberuledoutpurelybyvirtueofthe(grammatical)structureofitslinguisticexpression.Eachoftheseapproachesfacesinnumerabledifficulties.Toassessthemonemustgrapplewithproblemsthatlieatthe155\nA.W.MOOREveryheartofphilosophicallogic.Ishallconfinemyselftodiscussionofthethirdandfourthapproaches.3.1ThesemanticapproachOnthesemanticapproach,itislogicallyimpossibleforanaunttobeaman,say,becauseitispartoftheverymeaningoftheword‘aunt’thatauntsarefemale.Nothingsimilarprecludesanauntbeing150yearsold.Tounderstandthesentence‘Someauntsaremale’isalreadytobeinapositiontoseethatitisanincorrectthingtosay.Tounderstandthesentence‘Someauntsare150yearsold’,however,isnotyettobeinsuchaposition,eventhoughittooisanincorrectthingtosay(since,asithappens,noauntisasoldasthat).Theformersentenceissaidtobeanalytic,thelattersynthetic.Solongasonelimitsone’sattentiontosimpleexamplesofthiskind,onemightbeforgivenforthinkingthatnothingmoreneedstobesaidtoestablishareasonablyfirmaccountoflogicalpossibility.Butinfactmuchmoreneedstobesaid.Wehavebeengivennorealguidancewhenitcomestoassessingtheproblematiccasescitedearlier.Canonetellthatitisincorrecttosay‘2+2=5’justbyunderstandingthesentence?Ifweweretoldmoreabout‘understanding’–indeed,ifweweretoldmoreaboutwhatitistodoonething‘justby’doinganother–wemightbeabletomakeastabatthatquestion.Asitis,thequestionisimpenetrable.Weshouldanywaybesuspiciousaboutwhetherthemeaningofasentencelike‘2+2=5’can,byitself,imposeconstraintsofthiskind.RecallQuine’shostilitytotheviewthatpropositionsarethemessagecarriedbythelinguisticmedium(seesection2above).Thathostilitywastotheveryideaofaclearlydelineatedanddeterminatemeaningattachingtoeachofoursentences.Unsurprisingly,itcarriesoverintohostil-itytothesemanticconceptionoflogicalpossibility.OnQuine’sview,if,atsomepointinthefuture,English-speakersarepreparedtodescribecertainpeopleas‘maleaunts’,inspiteoftheirearlierinsistencethatsuchathingisimpossible,thentherewillbenothingtoruleouttheirhavingchangedtheirmindratherthantheirlanguageortheirhavingcometoadifferentviewaboutwhataunts,orsomeofthematleast,arelike(Quine1961c).Oncethedistinctionbetweendifferenceinlanguageanddifferenceinviewhasbeenblurredinthewayithasbeenhere,itmayseemasmallsteptotheconclusionthatnothingisimpossibleinany‘absolute’way.Forthereisnothingwedenywhichitmightnothavebeencorrecttosay.Suppose,forexample,thatwehadcounteddifferently.Supposewehadcounted:‘1,2,3,5,4,...’.Inthatcaseitwouldhavebeencorrecttosay‘2+2=5’.Andonthecurrentreckoning,thereisnoprincipledwayofresistingthemovefromheretotheconclusionthat2+2wouldhavebeenequalto5.Butthislineofthoughtisconfused.Whatisblurred,onthecurrentreckoning,isthedistinctionbetweenasituationinwhichwecometosaydifferentthingsbecauseofchangesinourlanguageandasituationinwhichwecometosaydifferentthingsbecauseofchangesinourthinking.Thisisenoughtocastdoubtonthedeterminacyofmeaning,andinparticularonwhethermeaningissufficientlydeterminatetosustainthesemanticconceptionoflogicalpossibility.Buttheremaystillbesomeper-fectlyclearandpertinentdistinctionstobedrawnbetweenitsbeingcorrecttosay‘2+156\nPHILOSOPHYOFLOGIC2=5’and2+2’sbeingequalto5.Inparticular,ifsomeotherconceptionoflogicalpossibilitycanbesustained,thenthefirstofthesemaycountasalogicalpossibilityandthesecondnot.Ingeneral,itisimportanttodistinguishbetweentalkaboutbitsoflanguageandtalk,bymeansofthosebitsoflanguage,aboutotherthings:theformerisoftensig-nalledbytheuseofinvertedcommas.Thisdistinctionisnicelyillustratedbythefactthat,whereascatshavefourlegs,‘cats’–notethesingularverbcomingup–hasfourletters.Imyselfdoubtwhetherthisdistinctionisalwaysclearinitsapplication.Forexample,whenitissaidthatAlbert,whoknowspreciouslittlephysics,haspickedup2thate=mc,Idoubtthatitisclearwhetherwhathasbeensaidconcernsthesentence2‘e=mc’orthephysicallawitself.ButIdonotdoubtthatthedistinctionisanimpor-tantone.AndIseenothinginQuine’sviewsaboutmeaningtopreventhimfromacknowledgingit.WhereQuine’sviewsmaybeproblematicforhim,incidentally,isalongsidehisinsistencethattherearecertainprivilegedsentence-typesthatcanbeclassifiedastrueorfalsewithoutrelativization.Thisisanaptpointatwhichtopickuponthatdifficulty.Theexistenceofsuchsentence-typeswasmootedearlier(seesection2above),duringourinitialdiscussionofpropositions.Putativeexampleswere‘2+2=4’andaspeciallyconstructedsentenceoftheform‘Sishungryatt’.Itwassuggestedthatthesesentence-typesmightbesuitablecandidatesforthetitleof‘proposition’.PartofQuine’sinterestinthemisthatthisisasuggestionhefavours.Certainlyitisasuggestionthatseemstogowellwithhisgeneralscepticismaboutmeaning,whichpreventshimfromconstruingpropositionsasthemessagecarriedbythelinguisticmedium(Quine1960:208–where,itshouldbenoted,hereservestheterm‘proposition’forwhathetakestobetheobjectionableconstrual).Infact,however,thereistension.Thetensionsurfaceswhenitcomestosayingwhichtheprivilegedsentence-typesare.Itwillnotdotosaythattheyarethosedeclarativesentence-typeswhosetokensareeitheralltrueorallfalse.Sometokensofdeclarativesentence-typesarenotusedtoassertanything,aswesawinsection2above.Norwillitdotosay,coiningasuitablephrase,thattheyarethosedeclarativesentence-typeswhoseassertorictokensareeitheralltrueorallfalse.Somedeclarativesentence-typesdonothaveanyassertorictokens:indeed,some(most)donothaveanytokensatall.Norwillitdotosaythattheyarethosedeclarativesentence-typeswhosepotentialassertorictokensareeitheralltrueorallfalse.Thatincludesthesentence‘Iama150-year-oldaunt’,whichought,intuitively,tobeexcluded.(Beingfalserelativetoallpeopleandalltimesisnotthesameasbeingfalsewithoutrelativization.)Norindeedwillitdotosaythattheyarethosedeclarativesentence-typeswhosepotentialassertorictokensareguaranteedtobeeitheralltrueorallfalse.Thatincludesthesentence‘Idonotexist’.Whatwehavetosay,intuitively,isthattheyarethosedeclarativesentence-typeswhosepotentialassertorictokensareguaranteedtobeeitheralltrueorallfalsebythefact–andthisiswherethetensionsurfaces–thattheyallcarrythesamemessage,or,ifyoulike,thattheyallhavethesamecontent.Quine’sonlyrecourse,giventhistension,seemstobetodrawupalistofwords,phrasesandotherlinguisticdevices,suchas‘I’,‘tenyearsago’andtheuseofthepresenttense,andtoidentifytheprivilegeddeclarativesentence-typesasthosewhicharefreeofthesedevices.However,thiswouldscarcelybeilluminating.157\nA.W.MOORE3.2ThesyntacticapproachLetusturnnowtothefourthapproachtologicalpossibility,thesyntacticapproach.Onthisapproach,itislogicallyimpossibleforagreenapplenottobeanapple,say,becauseitispartoftheverystructureoftheexpression‘greenapple’thatitappliesonlytowhat‘apple’appliesto.Notthatthisisasstraightforwardasitseems.Itisnotpartoftheverystructureoftheexpression‘artificialapple’thatitappliesonlytowhat‘apple’appliesto.Anyonewhotookthisapproachwouldhavetoinsistthat‘greenapple’and‘artificialapple’haddifferentstructures,despiteappearances.Structurewouldsometimesshowuponlyafteranalysis.Suchanalysismightinvolveadditionalwords.Forexample,thestructureof‘greenapple’mightincorporatetheword‘and’.(Somethingisagreenappleifitisgreenandanapple.)Thisinturnwouldfitwiththemostpopularversionofthisapproach,accordingtowhichtherearecertaindesignatedwordssuchas‘and’,‘not’,‘every’and‘is’(orperhapsformalizedcounterpartsofthese)whoseroleinanexpression,orinanexpression’sanalysans,determinesthestructureofthatexpression.Bywayofclarification:supposeonestartswithanexpression,oritsanalysans.Andsupposethatonekeepsthesedesignatedwordsfixedandreplaceseveryothersubsidiaryexpressionbyaschematicsymbol(traditionallyjustaletter)whichservestoindicatewhichgrammaticalcategoryitbelongsto.Thenwhatresultswillrepresentthestructureoftheexpression.Forexample,thestructureof‘Noteveryappleisgreen’issomethinglike‘NoteveryAisB’,where‘A’indicatesasortalnounand‘B’anadjective.‘Noteverycarrotispink’presumablyhasthesamestructure.Sodo‘Noteverynumberisprime’,‘Noteverycountisbloodthirsty’andmanymore.Thelogicallyimpossibleisthenwhatcanberuledoutpurelybyvirtueofthestructure,inthissense,ofitslinguisticexpression.Ifweallowourselvestotalkaboutdeclarativesentencesasbeingtrueorfalse,wecanputitlikethis:thelogicallyimpossibleiswhatcanbeexpressedbyadeclarativesentencethatisnotonlyfalsebutsuchthatanyotherdeclarativesentencewiththesamestructureisalsofalse.Itislogicallyimpossibleforagreenapplenottobeanapple,then,becausetherearenotruesentenceswiththestructure‘ThereissomethingwhichisBandanAandnotanA’.Asbefore,thereisagooddealherethatcallsoutforclarification.Butthemostobviousandthemostpressingquestionisthis:whatissospecialaboutthedesignatedwords?Whydoes‘and’count,andnot‘carrot’?Whydoes‘not’count,andnot‘count’?Oneintuitionisthatthedesignatedwordsarewordsthatarenotalliedtoanyspecificsubjectmatter:theycanbeusedtotalkaboutanythingwhatsoever.Interestingattemptshavebeenmadetolendsomeprecisiontothisintuition.Notoriously,however,nogenerallyacceptedaccountofthedesignatedwordshasemerged,eitherfromthisdirectionorfromanyother.Thereisagenerallyacceptedlabel.Suchwordsarestandardlyreferredtoas‘logicalconstants’.Andmostpeoplewhohaveworkedinthisareahaveagreedaboutwhichwordsareclearlylogicalconstants,whichwordsareclearlynotandwhichareproblematiccases.Butitremainsanopenquestionwhat,ifanything,thisconsensusisaresponseto.Quine,characteristically,isinclinedtoplaydownthedistinctionbetweenlogicalconstantsandotherexpressions.Theclearestpossiblewayofdemarcatingtheformer,inhisview,wouldbebycompilingalist(Quine1970:especiallych.2).Butagain,thatwouldscarcelybeilluminating.158\nPHILOSOPHYOFLOGICLetus,however,grantthedistinction,forthesakeofargument.Onthesyntacticapproachtologicalpossibility,themoststraightforwardwayofdefiningconsistency,inconsistency,consequenceandotherlogicalrelationsisbyfirsttransposingtalkofthestructureofsentencesintotalkofthestructureofpropositions.Wecanthensay,forexample,thatonepropositionisinconsistentwithotherswhenanysetofpropositionswiththesamestructureasthesecontainsatleastonepropositionthatisfalse.Similarlyfortheotherrelations.Thus,takethefollowingargument:NofisharemammalsAllwhalesaremammalsThereforenowhalesarefishWecansaythatthisargumentisvalid(inotherwords,thattheconclusionisaconsequenceofthetwopremises)onthefollowinggrounds.NomatterwhattripleofpropositionsyouconsiderwiththefollowingstructureNoAsareBsAllCsareBsNoCsareAsifthefirsttwoaretrue,thensoisthethird;or,equivalently,thereisnotripleofpropo-sitionswiththisstructuresuchthatthefirsttwoaretrueandthethirdisfalse.Butthisengendersthefollowingworry.Howcanwebesurethatwearenotatthemercyoflimitedlinguisticresources?Perhapsthereasonwhythereisnosuchtripleofpropositionsisthatwedonothavethelinguisticwherewithaltoconstructthem.Itistruethatonsomeviewsofwhatpropositionsare,propositionsdonotdependfortheirexistenceonanyparticularlinguisticresources.Butthatservesonlytoredirectouranxiety:howcanwebesurethatourtalkaboutthestructureofsentencesisanadequatebaseforourtalkaboutthestructureofpropositions?Onenaturalreply,reinedtotheexampleabove,isthatitisnotjustthattherearenotriplesofthespecifiedkind,itisthattherecouldnotpossiblybe.Thatisoflittleavail,however,whenouraimispreciselytoelucidatetherelevantsenseof‘possibly’.Aquitedifferentandmoresophisticatedreplywouldinvolveallayingtheworrydirectlybyshowingthatourlinguisticresourcesareindeedadequateforthesepur-poses.Thiscanbeshown,atleastonmoststandardconceptionsofthematter:itisenough,infact,thatwecandoelementaryarithmetic(Quine1970:ch.4).However,itcanbeshownonlythankstothefactthatweareabletoprovideanindependentaccountofwhatitisthatisbeingshown.Themoreonethinksaboutthisproblem,themorelikelyitisthatone’sattentionwillbedrawnawayfromlanguagetorealityitself.(Thisshiftofattentionreflectsoneimpor-tantshiftintheevolutionfromatraditionalARISTOTELIAN(chapter23)conceptionoflogictoitsmoderndescendant.)Reconsidertheschematicsymbols‘A’,‘B’and‘C’intheexampleabove.Whatultimatelymattersisnothowthesesymbolsmightbereplacedbysortalnouns,buthowtheymightbeassigneditemsofthekindthatsortalnouns‘standfor’,touseadeliberatelyloosephrase.Itisimmaterialif,insomecases,nosortalnounsactuallyfitthebill.(Thustheexampleconcerningfish,mammalsandwhales159\nA.W.MOOREwouldhavebeenpertinenteveniftherehadneverbeenanysuchwordsas‘fish’,‘mammal’or‘whale’,noranyotherwordsequivalentinmeaningtothese.)Forthisreason,whenlogicianscircumscribeallthepossible‘interpretations’ofasymbol,whattheyusuallydoistobypasslanguageandgostraighttotheitemsthemselves.Herewearriveatasetofideas,inparticular,theideaof‘standingfor’,whicharefundamentaltophilosophicallogic,whetherornotthesyntacticapproachtologicalpossibilityiscorrect.Theyareideasabouttheveryrelationshipbetweenlanguageandreality.Toelucidatethemisabasictaskforanyphilosophyoflogic.Thisisthereforeanaptpointatwhichtoturntosomeoftheissuesassociatedwiththattask.3.3IssuesofsenseandreferenceTherearevariousmoreorlesscrudeintuitionsentangledhere,andtheyneedtobedisentangled.Theyalsoneedtoberefined.Thecrudestintuitionisthateachbitoflanguageactsasalabelforsomebitofreality:thename‘Jupiter’,forexample,actsasalabelforoneoftheplanetsinthesolarsystem.(Thisisclearlymuchlessplausibleinthecaseofawordlike‘and’.)Amuchmoresophisticatedintuitionisthateachexpres-sionofsomegrammaticallybasickindhasafeature,determinedbythatkind,whichisdirectlyrelevanttothetruthorfalsityofanypropositioninwhichtheexpressionoccurs(orwhichitcanbeusedtoexpress–howoneputsitwilldependonone’sviewofpropositions).Thisintuition,ofcourse,isdirectlyinformedbytheideasbeingcanvassedabove.Nowitwaspartofthegeniusofarguablythegreatestlogicianofalltime,GottlobFREGE(chapter37),toextricatefromtheseintuitionsadeepandpreciselyworkedoutconceptionof‘standingfor’.Fregeinventedlogicinitsmodernguise.And,inthecourseofdoingso,hefirstdistinguishedcertainbasicgrammaticalcategories,thengave,foreachone,anaccountofwhatanyexpressioninthatcategorystandsfor.Hecalledthisthereferenceoftheexpression(‘Bedeutung’intheoriginalGerman:‘Bedeutung’isusuallytranslatedas‘meaning’,butforFregeitwasservingasatechnicalterm).AtthebasisofFrege’sgrammar,alongsidedeclarativesentencesthemselves,arewhathecallednames.(Eventuallyhereckoneddeclarativesentencesthemselvesasnames(Frege1964:§2).)Thesearethesingularnounphrasesthatstandforparticu-larthings.(Theterm‘things’istobeunderstoodinthemostgenerouswaypossible.Examplesofnamesare‘Jupiter’,‘thepositivesquarerootof16’,‘thenumberofsymphonieswrittenbySchumann’and‘joy’.Thesecondandthirdofthesehavethesamereference,namelythenumber4.)OneofFrege’schiefconcerns,oncehehadintroducedtheideaofanameanditsreference,wastoshowthatthereisacrucialdifferencebetweenunderstandinganameandknowingwhatitsreferenceis.Forinstance,someonewhohasitongoodauthoritythatthereferenceofanameinUrduisthenumber4doesnottherebyunderstandthename.Again,someonewhounder-standsperfectlywellthetwonames‘thepositivesquarerootof16’and‘thenumberofsymphonieswrittenbySchumann’cannottell,withoutfurtherknowledge,thattheyhavethesamereference.Fregeaccordinglyintroducedwhathecalledsenses.Aname’ssense,hesaid,iswhatwegraspwhenweunderstandit.Twonameswiththesamereferencecanhavedifferentsenses.Buttwonameswiththesamesense(synonyms)cannothavedifferentreferences.Thename’ssensethereforedeterminesitsreference.160\nPHILOSOPHYOFLOGICIndeed,asFregealsosaid,thename’ssensecontainsthemodeofpresentationofitsreference.Ingraspingthesense,wethinkofthereferenceinacertainway.Thewholeapparatus,includingthesense–referencedistinctionitself,isthenextendedtoexpres-sionsineachoftheotherbasicgrammaticalcategories(Frege1993a,1993b).Itisalittleironicalthatthenotionofsenseshouldemergesodirectlyfromthesyntacticapproachtologicalpossibility.Forifthenotioncouldbegivensuitablephilosophicalgrounding,thenitwouldbeperfectlysuitedtoground,inturn,thesemanticapproachtologicalpossibility.SensesarepreciselythedeterminatemeaningsthatQuinerejects.GriceonMeaningTherehavebeenmanyandvariedattemptstosupplementFrege’sworkinthisareainanefforttorescuesensesfromQuine’sscepticalattacks,oratleastinanefforttoexpli-catemeaninginawaythatcould,ifsuccessful,bepittedagainstQuine.Forexample,H.P.Gricefamouslyanalysedmeaninginawaythatmadeitlook,perhapsnotcompletelydeterminate,butdeterminateenough(Grice1967).Hisanalysiswasintermsofcertaincomplexpsychologicalstates.ThefirstthingGricedidwastodelimitwhatkindofmeaningisatstake.Hecalledthisnon-naturalmeaning,incontrasttothekindofmeaningthatisinvolvedwhenanapple’sbrownskin‘means’thatitisrotten.Hethendistinguishedbetweenwhatanexpressionmeansandwhatsomeonemeansbyutteringit(onagivenoccasion).Hetookthelattertobethemoreappropriatepointofentryforanalysis:foranexpressiontomeanwhatitdoesisforpeopletomeanwhattheydobyutteringit.Finally,hearguedthatforsomeonetomeanwhathedoesbyutteringanexpressionisforhimtointendhisutterancetoproducesomeparticulareffectinanaudiencebymeansoftherecognitionofthisveryintention.ThisanalysishasnoinfluenceoverQuinehimself,however.Quine’sscepticismaboutsensessimplygetstrainedonthepsychologicalstatesthemselves.ThismaylooklikeunreasonablerecalcitranceonQuine’spart.Butconsider.HowmuchwouldGrice’sanalysisshowevenifitwerecorrect?Therewouldstillbeaquestionaboutwhetherthepsychologicalwasinanywaymorebasicthanthesemantic.ItwouldbejustasinvitingtosaythatGricehaddemonstratedcertainintimateinterconnectionsbetweenthepsy-chologicalandthesemantic.Afterall,itisveryimplausiblethatwecoulddivinewhatapersonintendsinutteringanexpressionwithoutfirstfindingoutwhattheexpressionmeans.Itmayevenbethatthepersoncouldnotsomuchashavethatintentionwithouttheexpressionfirsthavingthatmeaning.SoanyonesuchasQuine,reflectingonallofthisandscepticalabouttheveryideaofmeaninginvolved,isboundtobescepticalaboutthepsychologyinvolvedtoo.AmoreimmediateworryaboutFrege’sideaswasgivencelebratedexpressionbyBertrandRUSSELL(chapter37).RussellsharedFrege’sinterestinwhatthelatterhadcallednames.Heinsisted,however,thattheyformaveryheterogeneousclass.OnRussell’sview,itisonlybyignoringthecrucialdistinctionbetweenapparentstructureandrealstructure–adistinctionwhichFregehimself,ironically,haddonesomuchtodrawtoourattention–thatweareledtoassimilatealltheseFregeannamesinthefirstplace.IfweuseFrege’sowntechniquesofanalysis,wecanseparatesuchnamesinto161\nA.W.MOOREfundamentallydifferentkinds.ManyofFrege’sconclusionsthenbegintolookunwarranted.Russellarguedasfollows(Russell1956a,1993).SomeFregeannames,theonesthatRussellhimselfwaspreparedtocallnames,aresimpleindefinablewordswhosemean-ingsdonotdependonthemeaningsofanyconstituentexpressions.(ItishardtogiveexamplesbecauseRussellthoughtthatpropernouns,suchas‘Plato’,couldbedefined–inthiscase,presumablyassomethinglike:‘theauthorofTheRepublic’.Russell’sownviewsaboutwhatwouldcountasexamplesunderwentvariouschanges.Atonepointheheldthattheonlytwoexampleswere‘I’and‘this’(Russell1988:26).)Tounder-standoneofthesenames–aRussellianname–istoknowwhatitsreferenceis.TheFregeannamestowhichthisdoesnotapplyaredefinitedescriptions–singularnounphrasesoftheform‘theA’.Butthesefunctionverydifferently.Toseewhy,consideradefinitedescriptionthatisnotinfactsatisfiedbyanything,say‘theplanetbetweenMercuryandtheSun’.InFregeanterms,thisisanamewithasensebutnoreference.NowsupposeweunderstanditaspurportingtohaveareferenceinthesamewayinwhichaRusselliannamedoes.Whatthenarewetosayaboutasentencesuchas‘TheplanetbetweenMercuryandtheSunisgreen’?Weareforcedtosaythat,sinceitisnotaboutanythinginthewayinwhichitpurportstobe,therecannotbeanyquestionoftruthorfalsity:nogenuinepropositionisinplay.Butsurelyagenuinepropositionisinplay.Afterall,itwouldbeperfectlypossibleforsomeonetothink,mistakenly,thattheplanetbetweenMercuryandtheSunisgreen.Wemustthereforeunderstandthedescriptiondifferently.Wemustdisregardthesentence’sapparentstructure,andstopthinkingofthedescriptionasaproperconstituentofit.ThisleadstoRussell’salternativeanalysis.Onthatanalysis,‘TheplanetbetweenMercuryandtheSunisgreen’isequivalenttotheconjunctionofthefollowingthreepropositions.(1)ThereisatleastoneplanetbetweenMercuryandtheSun(2)ThereisatmostoneplanetbetweenMercuryandtheSun(3)AnyplanetbetweenMercuryandtheSunisgreenItisclearfromthisanalysisthatagenuinepropositionisindeedinplay.Moreover,giventhattherearenoplanetsbetweenMercuryandtheSun,wecanseefrom(1)thatthepropositionisfalse.Russellbelievedthatoncetheseideashadbeengeneralized,andonceasuitablydeeplevelofanalysishadbeenreached,therewouldbenoneedtoinvokesenses.SensesarejustaFregeanglossonhiddenstructuralcomplexity.Theonlysemanticrelationbetweenlanguageandreality,ultimately,isthatwhichholdsbetweenexpressionsandtheirreferences.ThisaccountisobviouslyverycongenialtoQuine.Oneofitsbonuses,whichQuinehimselfhaspursued,isano-nonsenseviewofwhatrealitymustbelikeinorderforanypropositiontobetrue–orfalse(Quine1961b).Inparticular,wearesparedthedrasticexpedientwhichmanyhavefeltitnecessarytoadopt,ofpositingreferencesfortheoffendingdescriptions(forinstance,inthecaseof‘theplanetbetweenMercuryandtheSun’,an‘unreal’planetbetweenMercuryandtheSun):thisexpedientflirtswithinco-herenceinthecaseofsentenceswhichareused,orseemtobeused,preciselytodeny162\nPHILOSOPHYOFLOGICthatsuchreferencesexist(forinstance,‘ThereisnosuchthingastheplanetbetweenMercuryandtheSun’).Allthatrealitycontains,onQuine’sview,iswhatitistrulyandclearlysaidtocontainattheendofRusselliananalysis:thethingsofwhichitissaidthatatleastoneisthusandso,oratmostoneissuchandsuch,oreachwhichisthusisalsoso,andsoforth.Strawson,meanwhile,recoilingfromwhatheseesasRussell’sover-zealouseffortstoburrowbeneaththelinguisticsurface,hastriedtorescuepartofFrege’soriginalcon-ceptionofnames,thoughwithlessconcerntorespectthedemandsofformallogic(Strawson1993).MuchofRussell’szeal,Strawsonthinks,reflectsinsufficientsensi-tivitytohowwordsareactuallyhandled,andtotherolethatattendantcircumstancesplayinthis.Suchinsensitivityisinturnduetoafailuretotakeintoaccountdistinc-tionsblurredwhenweleftbehindourinitialdiscussionofpropositionsinsection2.Inparticular,talkofwhetherornotthereisaproposition‘inplay’–thisismywayofputtingit,notRussell’sorStrawson’s–obscuresthedistinctionbetweensentence-typesandsentence-tokens.Meaning,saysStrawson,attachestothetypes,truthandfalsitytothetokens,oratleasttoassertionsmadebymeansofthosetokens.Thetwothingsarequiteseparate.Itisperfectlypossibleforameaningfulsentence-type,whichcanbeutteredinsuitablecircumstancesinmakinganassertionthatistrueorfalse,tobeutteredinunsuitablecircumstancesinfailingtomakeanassertionthatistrueorfalse.Thelatteriswhathappenswhensomeonesays‘Thatisgreen’withoutsucceedinginindicatinganything,orwhensomeonesays‘Thethingonthetableisgreen’whennothingisonthetable.Thesamethinghappens,presumably,whensomeonesays‘TheplanetbetweenMercuryandtheSunisgreen’(thoughinthiscasethereappearstobenosuchthingasasuitablecontext).TherenowseemstobenoreasonwhyweshouldnotassimilatedefinitedescriptionstootherFregeannames,inthewayinwhichFregedid.Thefactthatdefinitedescriptionscansometimesbeusedwithoutreference,insuchawaythatthereisnoquestionoftruthorfalsity,isquitecompatiblewiththeirbeingmeaningful,withthesentence-typesinwhichtheyoccurbeingmeaningful,andwithwhoeverusessuchasentence-typeinunsuitablecircumstanceshavingagenuinebutmistakenthoughtthatheorshehassaidsomethingtrue.ThisdisputebetweenRussellandStrawsonmeansthatifsomeoneweretoutterthesentence‘TheplanetbetweenMercuryandtheSunisgreen’,Russellwouldsaythattheyhadassertedsomethingfalse,whereasStrawsonwouldsaythatthequestionoftruthorfalsitydidnotarise.Again,RussellwouldsaythatitwaspartofwhathadbeenassertedthatthereisaplanetbetweenMercuryandtheSun,whereasStrawsonwouldsaythatthiswassomething‘implied’bytheutteranceaboutthesuitabilityofthecircumstances.Russell’saccountseemsthemorebarbarousofthetwo.Itcertainlyinvolvessomebutcheringofourpre-theoreticalintuitions.Butisthissuchabadthing?Suchbutcheringoftenoccursinlogic,foragaininsimplicity.Thereisausefulanalogyherewhichitisworthdigressingtoconsider.Recallthedefinitionofconsequencewithwhichwehavebeenoperating:onepropositionisaconsequenceofotherswhenitisnotpossibleforalloftheotherstobetrueyetitfalse.Veryfewpeopleflinchatthisdefi-nitionwhenfirstpresentedwithit.Yetitalreadyinvolvesconsiderablebutchering.Takeasetofpropositionssuchthatitisnotpossibleforthemalltobetrue,stilllessforthemalltobetrueandsomeotherpropositionfalse.AsimpleexamplewouldbethepairofpropositionsthatGodexistsandthatGoddoesnotexist.Thedefinitionaboveforcesus163\nA.W.MOOREtorecognizeanyarbitraryproposition–thepropositionthatgrassispink,say–asaconsequenceofthese.Hardlytheverdictofpre-theoreticalintuition!Butthisisthepricewepayforasimple,workabledefinition.Similarly,perhaps,withRussell’saccount.OnethingthatmustcertainlybesaidinfavourofRussell’saccountisthatitisespe-ciallyhelpfulwhenitcomestohighlightingtheambiguityincertainsentences,forexample‘Thistimeyesterdaythethingonthetablewasgreen’.TherearetwowaysofsubjectingthissentencetoRusselliananalysis.Either‘thistimeyesterday’governsallthreeoftheconjuncts,andthesentenceisconcernedwithwhatwasthenonthetable,or‘thistimeyesterday’ispartoftheinnerstructureofthethirdconjunct,andthesentenceisconcernedwithhowthethingonthetablethenwas.Howeverthisdisputeistoberesolved,onethingoffundamentalimportancethatemergesfromtheworkofFrege,Russell,QuineandStrawson–animportancetheywouldallacknowledge–isacertainbasicstructuralfeatureofpropositionswhichallowsustorecognizeontheonehandnames,usedtorefertoparticularthings,andontheotherhandtheirpredicativecomplements,trueorfalseofthesethingsandtherebyabletocombinewiththenamestoproducepropositionswhicharetrueorfalse,fullstop.Thequestionofwhatunderliesandexplainsthisstructureisoneofthedeepestinphilosophicallogic.Itiswellbeyondthescopeofthecurrentchaptertomakeaproperattemptatansweringit.Buthereisthesketchofapossibleanswer.Ifanyexpressionsaretohavemeaning,thensomeofthemmustsometimesbeusedassingle-expressionsentences,like‘Green’,which–tomakethepointyetonemoretime–areneithertruenorfalseinthemselves,butonlyas(potentially)utteredincertaincircumstances,intheexpressingofdifferentpropositions.Suchvariationisoftheessenceofthemeaningoftheseexpressions.Preciselytheirfunctionistodrawattentiontorecurringyettransitoryfeaturesofreality.Butiftheiruseassingle-expressionsentencesistobeemulatedinanyotherthantherelevantcircumstances(say,becauseanepisodefromthepastisunderdiscussion),thenlanguageitselfhadbetterbeabletosimulatetherelativization.Theexpressionsmustcountastrueorfalseofcertainthingstowhichreferenceismadebymeansofotherexpressions.4MarginaliaLetustakestock.Logic,Isuggestedattheoutset,isthegeneraltheoryoftheproposi-tion.Apropositionisthesortofthingthatistrueorfalse;andlogicisconcernedpri-marilywithwhichcombinationsoftruthandfalsityarepossible.SincethenIhavebeenconcernedwithclarifyingtheseideas.Butitisimportanttoseealsohowtheymightbechallengedand/orsupplemented.4.1BeyondtruthandfalsitySomepeoplehavethoughtthatlogicextendsbeyondtherealmofthetrueandthefalse.Forinstance,thereistalkoferoteticlogicandofimperativallogic:theseareconcerned,respectively,withquestionsandcommands.(Insomecasestheissueisnotwhetherlogicextendstothegiventerritory.Theissueiswhetherthegiventerritoryisoutside164\nPHILOSOPHYOFLOGICtherealmofthetrueandthefalse.Forinstance,noonedoubtsthattherecanbeincon-sistencyinethics.Butthereisfiercedebateaboutwhethertherecanbetruthandfalsitythere.)Tosaythatlogicextendsbeyondtherealmofthetrueandthefalseistosaythatlogicalrelationsapplyincertaincasesinwhichtheconceptsoftruthandfalsity(strictlyspeaking)donot.Totheextentthatthisisso,IthinkthebestexplanationisofthekindIsketchedinsection2:thoughlogicisthetheoryofthatwhichistrueorfalse,thatwhichistrueorfalsecansometimesserveasamodelforthatwhichisnot.Amoreurgentworryconcernstheinnocuousnessofthetrue–falsedichotomyitself.Thedichotomycertainlyharboursparadoxes.Themostfamiliarofthesecanbeillustratedasfollows.THEPROPOSITIONEXPRESSEDINSIDETHISBOXISFALSEIfthepropositionexpressedinsidetheboxistrue,thenitisfalse.Ifitisfalse,thenitistrue.Muchworkhasgoneintotryingtosolvesuchparadoxes(seeSainsbury1988:ch.5).Ishallnotrehearseanyofthatworknow.Butwhatseemsclearisthatasatisfac-torysolutionwillteachustobemorecircumspectinourapplicationoftheconceptsoftruthandfalsity.Fewerthingswillseemaptforassessmentastrueorfalse.Philosophersoftheanti-realistschoolthinkthatthereis,overandabovethis,amuchmoreradicalcircumspectionwaitingtobelearned.Anti-realismisarevision-aryphilosophyoflogicwhoseinvestigationhasbeenpioneeredbyMichaelDummett,thoughhehasneverhimselfgoneasfarasendorsingit(Dummett1976,1978b,1978c,1991).Themaintargetofanti-realistcircumspectionisaninveterateassumptionthatinformsallorthodoxworkonformallogic,namelythatifsomethingisaptforassessmentastrueorfalse,thenitmustbeoneortheother.Thisassumptioncomessonaturallythatitseemsbyturnsthemeresttautology,thenjustaruleofthegame,thenthedeepestmetaphysicalprincipleaboutreality,moreultimatethanlogicitself.Theanti-realistchallengetothisassumptionconcernspropositionswhosetruthorfalsitywecannotdetermine.AnexamplemightbethepropositionthatAristotlesneezedatleastonceonhisfirstbirthday.Whatistheharminourassumingthatsuchapropositioniseithertrueorfalse,evenifwecannottellwhich?Theharm,accordingtoanti-realists,isthatitblocksasatisfactoryaccountofourgraspoftheproposition.Theirargumentisasfollows.Ourgraspofapropositionisamatterofknowingwhatwouldmakeittrueand,derivatively,whatwouldmakeitfalse.Onanyviewofpropo-sitions,suchagraspisakindoflinguisticunderstanding.Assuch,itmustadmitofpublicratification.Foriftherewereanexpressionwhoseunderstandingdidnotadmitofpublicratification,thennobodycouldevershowthattheyhadunderstoodit,whichmeansthatitcouldnotbeusedforpurposesofcommunication,whichmeansthatitwouldnothaveameaning,whichmeansthattherewouldbenosuchthingasunder-standingit.Butinorderforourgraspofapropositiontoadmitofpublicratification,thepropositionhadbetternotbetruewithoutourbeingabletotellthatitis,orfalsewithoutourbeingabletotellthatitisfalse.165\nA.W.MOORETherearetwoobjectionstothisargumentongroundsofself-stultification.Thefirstrunsasfollows.Byinsistingthatourunderstandingofexpressionsmustadmitofpublicratification,anti-realistsareineffectlocatingthemeaningofexpressionsinhowweactuallyusethem.Sotheyareendorsingourstandardlinguisticpractices.Butoneofourstandardlinguisticpracticesistoregulatewhatwesaybytheassumptionthatanypropositioniseithertrueorfalse.Anti-realistswillreplythattheyareendorsingourstandardlinguisticpracticesonlytotheextentthatthosepracticesadmitofendorsement.Itisperfectlypropertorejectsomeofthosepracticesfornotharmonizingwithothers,orforbeingotherwiseinternallyincoherent.Thesecondobjectionismorecomplex.Considerthefollowingproposition.(P)Thereisatleastonepropositionwhosetruthorfalsitywecannotdetermine.Nowasfarastheanti-realistargumentgoes,itisonlyif(P)istruethatthereisanyharminourassumingthateverypropositioniseithertrueorfalse.Butis(P)true?Well,onananti-realistconception,onlyifwecantellthatitis–inotherwords,onlyifthereisatleastonepropositionsuchthatwecantellthatwecannotdetermineitstruthorfalsity.Butifthereissuchaproposition,thenwecantell,inparticular,thatwecannotdetermineitstruth.Theonlywayoftellingthat,however,againonananti-realistconception,isbydeterminingitsfalsity.Sowearriveatacontradiction.Itfollowsthat(P)cannotbetrue.Hence,asfarastheanti-realistargumentgoes,therecannotbeanyharminassumingthateverypropositioniseithertrueorfalse.Anti-realistshaveanumberofwaysofrespondingtothisobjection.Moststraight-forward,andmostheroic,istoconcedethattherecannotbeanyharminmakingtheassumption,butstillnottomakeit.Thisisitselfaninstanceofanti-realistcircum-spection.Itistoconcedethattheassumptioncannotbefalse,butstillnottoacceptitastrue.Whateverthemeritsoftakingthisstance,boththeoriginalobjectionandthestanceitselfareinmyviewoftheutmostphilosophicalsignificance.4.2BeyondpossibilityLetusnowreconsiderpossibility.Wecanproceedhereviathedistinctionbetweenanargument’sbeingvalidanditsbeinggood.Anargumentisgoodifitprovidesgoodreasonsforacceptingitsconclusion.Thisisnotofcourseadefinition.Theword‘good’hasbeenrelocatedratherthaneliminated.Moreover,theremaybedeepreasonsofprinciplewhyadefinitionisnotavailable.Howeverthatmaybe,whatIhavesaidwillsufficeforcurrentpurposes.Nowthereareallsortsofwaysinwhichavalidargumentmaynotbeagoodone.Itmaybegthequestion.Orwemayknowoneofitspremisestobefalse.Or,differently,wemayknowthatnotallofitspremisescanbetrue.Anextremeexampleofthelastofthesewouldbethebizarrecasealludedtoinsection3.3:GodexistsGoddoesnotexistThereforegrassispink166\nPHILOSOPHYOFLOGICConversely,thereareallsortsofwaysinwhichagoodargumentmaynotbeavalidone.Itspremisesmayallbeknowntobetrue,andtheymaylendtheconclusionadegreeofsupportthatisveryhigh–butstillinconclusive.Takethefollowingexample:ThisdiehasbeenrolledfiftytimesIthascomeupsixeverytimeThereforeitisnotfairSupposethatboththepremisesaretrue.Stillitisnotimpossiblefortheconclusiontobefalse.Theargumentisnotvalid.Butitiscompelling.Goodinvalidargumentsofthiskinddifferfromtheirvalidcounterpartsinrestingnotonwhichcombinationsoftruthandfalsityarepossible,butonwhichareprobable,orperhaps(insomesense)reasonable.Therearecorrespondinglyweakerversionsofthelogicalrelations.Wecancallthese‘confirmationrelations’.Weencounterthemmostfrequently,andmostinterestingly,whenwearedrawingconclusionsabouttheunobservedfrompremisesabouttheobserved(as,forinstance,whensomeoneconcludesthatthestrangerspeakingtoheronthetelephonehasonlyonehead),somethingthatwedoconstantly,unexceptionably,butagain,forthemostpart,invalidly.Ihavesaidmorethanoncethatlogicisconcerned‘primarily’withlogicalrelations,thatiswithwhichcombinationsoftruthandfalsityarepossible.Ihavenotruledoutthepossibilitythatlogicextendstotheseconfirmationrelationsaswell.Ithinkitdoes.Theycomeundertheauspicesofinductivelogic,orINDUCTION(chapter9),asopposedtothedeductivelogicthathasbeenoursoleconcernsofar.(Itisnotuncontroversialthatthereissuchathingasinductivelogic(Popper1972).)Itisafurtherquestion,however,whetherformallogicextendstotheseconfirmationrelations,orrather,whetherthereissomecodificationoftheseconfirmationrelationswhichisofapiecewiththecodificationthatformallogicsuppliesoflogicalrelations.IagreewithNelsonGoodmanthatthisisthemosturgentquestionfacingtheinductivelogician(Goodman1979:especiallych.3).Goodmanisfamousforhavingpointedinductivelogicinanewdirection.Inthepastithasbeenconcernedwiththeproblemofjustifyingourrelianceonconfirmationrelations.ThisiswhatGoodmancallstheoldproblemofinduction.Hethinksthatthatproblemcanbedissolved.Thenewproblemistodefineconfirma-tionrelationsbytakingacuefromformallogicandsettingupappropriaterules.Thisisaproblemofdescriptionratherthanofjustification.Theproblemmaywellbeintractable.ButonethingthatGoodmanhasdemonstratedquiteclearlyisthat,whateverrolestructureplaysinlogicalrelations–adefiningrole,ifthesyntacticapproachtologicalpossibilityiscorrect–ithasnoprospectofplayingananalogousrolehere.Thisisbecause,givenanygoodinductiveargument,itsstruc-ture,underanyreasonableconstrualofstructure,willalwaysbesharedbysomeotherargumentthatiscrazy.Itwillalwaysbepossibleartificiallytoconstructanexpression,oraclusterofexpressions,whichcanbeusedtoillustratepreciselythisfact.Toseehow,considerthefollowingargument:AllemeraldsobservedbeforenowhavebeengreenTherefore,allemeraldsaregreen167\nA.W.MOOREThisargumentisagoodone.Thefactthatallemeraldsobservedbeforenowhavebeengreenisagoodreasonforthinkingthatallemeraldsaregreen.ButnowwecaninvokeGoodman’scelebratedterm‘grue’,whichappliestothingsobservedbeforeacertainfuturetimeiftheyaregreenandtootherthingsiftheyareblue.Supposethat‘grue’isputinplaceof‘green’intheaboveargument.Theresultantargumentisclearlyabadone.Thefactthatallemeraldsobservedbeforenowhavebeengrue(hencegreen)isnoreasonforthinkingthatallemeraldsaregrue(hence,insomecases,blue).Theconfirmationrelationthatheldintheoriginalcasehasbeenlost.Ingivinghisownsolutiontothisproblem,Goodmandefinesconfirmationrelationsintermsofthoseargumentsthatwedoinfactendorse,orhaveendorsedinthepast.Hethereforetriestofindagroundinginouractualpracticesforwhatisessentiallynor-mative–todistinguishbetweenwhatiscrazyandwhatisnotcrazybyappealtothecommitmentsthatweareourselvespreparedtomake.Thisbringswithitashiftofphilosophicalinterestfromthemoreformalquestionsaboutmeaningwhichconcernthedeductivelogiciantosubstantivequestionsabouthowweactuallyclassifythingstogether.Self-consciousnessaboutthatforcesusintoanotoriouslydelicateposition.Forhowweactuallyclassifythingstogethercanseemontheonehandcompletelynaturalandontheotherhandcompletelyarbitrary.Severalofthedeepestproblemsinphilosophy(andnotjustinthephilosophyoflogic)areinvolvedinworkingtowardsaproperresolutionofthisdissonance.AnyonefamiliarwiththeworkofLudwigWITTGENSTEIN(chapter39)willrecognizehintsofsomeofhispreoccupationsinthatlastparagraph.ItisaptformetofinishinWittgensteinianterritory,ifonlybecauseitgivesmeachancetoexplain,half-apologetically,whyIhavemadenomentionofhimbeforenow–exceptinacoupleofreferences.Wittgenstein’swork,asiswellknown,wasdividedintotwophases.(Thelocusclas-sicusoftheearlierphasewasWittgenstein(1961),thatofthelaterphase,Wittgenstein(1967).)Intheearlierphase,hisaimwastodrawthelimitsofwhatcanbethoughtandexpressed.Inhiseffortstorealizethisaim,heassignedthemostfundamentalofrolestologicalrelations(relyingheavilyontheworkofFregeandRussell).Forhebelievedthatthelimitsofwhatcanbethoughtandexpressedweresetbylogicitself.Itwouldnotbeanexaggerationtodescribehisearlyworkasanattemptata‘generaltheoryoftheproposition’.Laterherepudiatedmuchofthisattempt.Heshruggedoffhisearlieraspirationstogenerality,andbegantolookmoreintentlyatlessformalandlessabstractissuesconcerningthedifferentwaysinwhichnormativityandmeaningresideinourdifferentpractices–issuesoftheverykindthathavejustcometolight.Heremainedconvinced,however,thatapropergraspoflogicalrelationswascrucialtoaproperunderstandingoftheproblemsofphilosophy.And,inboththeearlierphaseandthelaterphase,hemadecontributionstothephilosophyoflogicofunrivalleddepthandsignificance.YetIhavesofarlefthimcompletelyoutofaccount.Why?PartoftheexplanationliesinWittgenstein’sownambivalentandidiosyncraticattitudetothephilosophyoflogic.Inhisearlywork,thephilosophyoflogicsubservedandcontributedtoaunifiednetworkofconcernsthatweresointimatelyboundupwithoneanotherthatitisactuallysomewhatmisleadingtodescribehimaspractisingthephilosophyoflogicatall,insofarasthissuggestsoneseparableactivityamongothers.Later,inaimingforsomethingmorepiecemealandmoreexploratory,hecameto168\nPHILOSOPHYOFLOGICdeplorethekindofsystematictheorizingaboutlogicanditsfoundationsthatIhavebeenattemptingtosurveyinthischapter.Tohavereckonedwithallofthatinthecontextofsuchasurveywouldhavetakenustoofarafield.ThefactremainsthattherecanbenosatisfactoryprogressinthephilosophyoflogicthatdoesnottakeWittgenstein’sinsightsandstricturesfullyintoaccount.Thisindicatesperhapsjustonerespectamongmanyinwhichmysurveyhashadtobeselectivetothepointoftravesty.FurtherReadingThegreatpioneersofcontemporaryphilosophyoflogicareFrege,Russell,WittgensteinandQuine.Ihavegivenasmallindicationoftheirviewsinthischapter.Forfirst-handknowledge,thefollowingarerecommended:inthecaseofFrege,Frege(1952);inthecaseofRussell,Russell(1940)andRussell(1956a),inwhichessays1–4and7areespeciallyrelevant;inthecaseofWittgenstein,Wittgenstein(1961and1967),whicharethelocusclassicusofhisearlierworkandofhislaterworkrespectively;andinthecaseofQuine,Quine(1960,1961aand1970).Otherclassicsinclude:Carnap(1956),whichhashadanenduringinfluence,inparticularonQuine;Davidson(1984),acollectionofground-breakingessayswhichdeal,amongotherthings,withfundamentalquestionsaboutthenatureoftruthandreference(essays1–8,12,15and16areespeciallyrelevant);Dummett(1978a),acollectionofextremelyinfluentialessayswhichincludesomeofhismostimportantworkonanti-realism(essays1,2,7–9and14–17areespeciallyrelevant);Dummett(1981),whichnotonlyservesasaninvaluablesupplementtoanyonewishingtounderstandFregebuthasbecomeaclassicinthephilosophyoflogicinitsownright;Evans(1985),acollectionofbrilliantandinsightfulessaysonsomeofthemostdemandingissuesinthephilosophyoflogic(essays1–5,7,8and12areespeciallyrelevant);Geach(1980),afascinating,partlyhistoricaldiscussionofreferenceandquantification;andKripke(1980),well-knownforhishighlyinfluentialtreatmentofproblemsconcerningthenatureofpropernamesandnecessity.Excellentintroductionstothephilosophyoflogic,allofthemsomewhateasierthananyoftheworkcitedabove,are:Blackburn(1984),whichismoreconcernedwiththephilosophyoflanguagethanwiththephilosophyoflogic,butwhichdoeshaveplentyofrelevantmaterial;Putnam(1971),whosefocusissomewhatnarrow–itisconcernedprincipallywithabstractobjects–butwhichprovidesanengagingandstimulatingexampleofthegenre;Strawson(1952),whichoffersaveryprovocative‘ordinarylanguage’critiqueoflogic,nowslightlydatedbutstillofenormousinterest;andfinally,Haack(1978)andSainsbury(1991),bothofwhichareverywide-rangingandveryaccessibleandeitherofwhichwouldbeanidealplacetostart.Anexcellentintroductiontoformallogic,whichpaysparticularattentiontohowtheformalandphilosophicalpartsoflogicrelatetooneanother,isHodges(1977).Usefulcollectionsare:Moore(1993)(theintroductionandessays1–4,6–8,10,12and13areespeciallyrelevant);andStrawson(1967).ReferencesBlackburn,S.1984:SpreadingtheWord:GroundingsinthePhilosophyofLanguage.Oxford:OxfordUniversityPress.Boolos,G.S.andJeffrey,R.C.1989:ComputabilityandLogic,3rdedn.Cambridge:CambridgeUniversityPress.Carnap,R.1956:MeaningandNecessity:AStudyinSemanticsandModalLogic,2ndedn.Chicago:UniversityofChicagoPress.169\nA.W.MOOREDavidson,D.1984:InquiriesintoTruthandInterpretation.Oxford:OxfordUniversityPress.Dummett,M.1976:WhatisaTheoryofMeaning?(II).InG.EvansandJ.McDowell(eds)TruthandMeaning:EssaysinSemantics.Oxford:OxfordUniversityPress.——1978a:TruthandOtherEnigmas.London:Duckworth.——1978b:Truth.InTruthandOtherEnigmas.London:Duckworth.——1978c:ThePhilosophicalBasisofIntuitionisticLogic.InTruthandOtherEnigmas.London:Duckworth.——1981:Frege:PhilosophyofLanguage,2ndedn.London:Duckworth.——1991:TheLogicalBasisofMetaphysics.London:Duckworth.Evans,G.1985:CollectedPapers.Oxford:OxfordUniversityPress.Frege,G.1952:TranslationsfromthePhilosophicalWritingsofGottlobFrege(translatedandeditedbyP.GeachandM.Black).Oxford:Blackwell.——1964:TheBasicLawsofArithmetic:ExpositionoftheSystem(translatedbyM.Furth).BerkeleyandLosAngeles:UniversityofCaliforniaPress.——1993a:OnSenseandReference(translatedbyM.Black).InA.W.Moore(ed.)MeaningandReference.Oxford:OxfordUniversityPress.——1993b:LettertoJourdain(translatedbyH.Kaal).InA.W.Moore(ed.)MeaningandReference.Oxford:OxfordUniversityPress.Geach,P.T.1980:ReferenceandGenerality:AnExaminationofSomeMedievalandModernTheories,2ndedn.Ithaca,NY:CornellUniversityPress.Goodman,N.1979:Fact,Fiction,andForecast,3rdedn.Brighton:HarvesterPress.Grice,H.P.1967:Meaning.InP.F.Strawson(ed.)PhilosophicalLogic.Oxford:OxfordUniversityPress.Haack,S.1978:PhilosophyofLogics.Cambridge:CambridgeUniversityPress.Hodges,W.1977:Logic:AnIntroductiontoElementaryLogic.Harmondsworth:PenguinBooks.Kripke,S.A.1980:NamingandNecessity.Oxford:Blackwell.Moore,A.W.(ed.)1993:MeaningandReference.Oxford:OxfordUniversityPress.Popper,K.1972:ConjecturalKnowledge:MySolutiontotheProblemofInduction.InObjectiveKnowledge:AnEvolutionaryApproach.Oxford:OxfordUniversityPress.Putnam,H.1971:PhilosophyofLogic.NewYork:HarperandRow.Quine,W.V.O.1960:WordandObject.Cambridge,MA:MITPress.——1961a:FromaLogicalPointofView:Logico-PhilosophicalEssays,2ndedn.NewYork:HarperandRow.——1961b:OnWhatThereIs.InFromaLogicalPointofView:Logico-PhilosophicalEssays,2ndedn.NewYork:HarperandRow.——1961c:TwoDogmasofEmpiricism.InFromaLogicalPointofView:Logico-PhilosophicalEssays,2ndedn.NewYork:HarperandRow.——1970:PhilosophyofLogic.EnglewoodCliffs,NJ:Prentice-Hall.Russell,B.1940:AnInquiryintoMeaningandTruth.London:AllenandUnwin.——1956a:LogicandKnowledge(editedbyR.C.Marsh).London:AllenandUnwin.——1956b:Ondenoting.InLogicandKnowledge(editedbyR.C.Marsh).London:AllenandUnwin.——1988:KnowledgebyAcquaintanceandKnowledgebyDescription.InN.SalmonandS.Soames(eds),PropositionsandAttitudes.Oxford:OxfordUniversityPress.——1993:Descriptions.InA.W.Moore(ed.)MeaningandReference.Oxford:OxfordUniversityPress.Sainsbury,R.M.1988:Paradoxes.Cambridge:CambridgeUniversityPress.——1991:LogicalForms:AnIntroductiontoPhilosophicalLogic.Oxford:Blackwell.Strawson,P.F.1952:IntroductiontoLogicalTheory.London:Methuen.——(ed.)1967:PhilosophicalLogic.Oxford:OxfordUniversityPress.170\nPHILOSOPHYOFLOGIC——1993:OnReferring.InA.W.Moore(ed.)MeaningandReference.Oxford:OxfordUniversityPress.Wittgenstein,L.1961:TractatusLogico-Philosophicus(translatedbyD.F.PearsandB.F.McGuinness).London:RoutledgeandKeganPaul.——1967:PhilosophicalInvestigations,3rdedn(translatedbyG.E.M.Anscombe).Oxford:Blackwell.DiscussionQuestions1Whatkindoftheoryisatheoryofpropositions?2Whatmakeslogicalrelationstherelationsthattheyare?3Howusefulisthedistinctionbetweensentence-tokensandsentence-typesforunderstandingthenatureofpropositions?4Howdoestherequirementthatpropositionsbetrueorfalseconstrainouraccountofwhatpropositionsare?5Arepropositionsbitsoflanguage?6Howarepropositionsrelatedtoassertions?7Isithelpfultosaythatpropositionsarewhatwespecifyby‘that’-clauses?8Canouraccountoflogicdowithoutmeaningandpropositions?9Doeslogichaveitsowndistinctivesubjectmatter?10Whatkindsofpossibilityarethere?Isthereoneparticularkindofpossibilitywhichisofconcerntologic?11Whatistheroleofconsistencyinlogic?12Isthereadifferencebetweenourknowledgethatitisincorrecttosay‘Someauntsaremale’andourknowledgethatitisincorrecttosay‘Someauntsare150yearsold’?13Canthelogicalimpossibilityofaclaimbedeterminedjustbyunderstandingthemeaningofthesentenceexpressingit?14Howdeterminatedoesmeaninghavetobeinordertosustainthesemanticconceptionoflogicalpossibility?15Canwedistinguishsentence-typesthatcanbeclassifiedastrueorfalsewithoutrelativization,fromothersentence-types?16Whatkindofanalysisandwhatkindofstructureofexpressionscanhelpustounderstandlogicalimpossibility?17Canwedoanythingmorethancompilealistoflogicalconstantstoshedlightontheirspecialroleindetermininglogic?18Howcanwebesurethatourtalkaboutthestructureofsentencesisanadequatebaseforourtalkaboutthestructureofpropositions?19Couldinadequatelinguisticresourcesmisleadusaboutwhatislogicallyimpossible?20Doexpressionsofdifferentgrammaticalkindscontributedifferentlytothetruthorfalsityofpropositions?21Canmeaningbeanalysedintermsofpsychologicalstates?22Howcanwedeterminewhetherthereisonekindofnameormanylogicallyheterogeneouskindsofnames?23Istheonlysemanticrelationbetweenlanguageandreality,ultimately,thatwhichholdsbetweenexpressionsandtheirreferences?171\nA.W.MOORE24Inouraccountofpropositions,towhatshouldweassignmeaningandtowhatshouldweassigntruthandfalsity?25Whatexplainsthebasicstructureofpropositions,accordingtowhichnamesandpredicatescombinetoproducepropositionswhicharetrueorfalse?Whyisthisadeepquestion?26Cantherebepropositionsthatareneithertruenorfalse?27Ifconfirmationrelationsaredefinedintermsofouractualpracticeconcerningconfirmation,aretheydefinedintermsofsomethingcompletelyarbitrary?172\n5PhilosophyofMindWILLIAMG.LYCANThephilosophyofmindaboundsinarrestingquestions:issuesofconsciousness,per-sonalidentity,survivalofone’sphysicaldeath,freedomofthewill,mentalillness,theroleofmindinbehaviour,thenatureofemotion,thecomparisonbetweenhumanandanimalpsychologies,andmanymore.Buteachoftheseissuesultimatelydependsonasingle,fundamentalquestion,calledsimply‘themind–bodyproblem’:whatisamind,perse,andwhatisitsrelationtobody,ortothephysicalingeneral?Accordingly,thischapterwillfocusonthatquestionandsomeproposedanswers.1Mind–BodyDualismThefirstanswertothemind–bodyproblemproposedinthemodernperiodwasthatofDESCARTES(chapter26),whoheldthatmindsarewhollydistinctfrombodiesandfromphysicalobjectsofanysort.AccordingtoCartesiandualism,mindsarepurelyspiritualandradicallynon-spatial,havingneithersizenorlocation.Onthisview,anormallivinghumanbeingorpersonisaduality,amindandabodypaired(thoughtherecanbebodieswithoutminds,andmindscansurvivethedestructionoftheircorrespondingbodies).Mysteriously,despitethedrasticdistinctnessofmindsfrombodies,theyinter-actcausally:bodilyhappeningscausesensationsandexperiencesandthoughtsinone’smind;conversely,mentalactivityleadstoactionandspeech,causingthephysicalmotionoflimbsorlips.Cartesiandualismhasstrongintuitiveappeal,sincefromtheinside,ourmindsdonotfeelphysicalatall;andwecaneasilyimaginetheirexistingdisembodiedor,indeed,theirexistingintheabsenceofanyphysicalworldwhatever.Anduntilveryrecently,infact,thephilosophyofmindhasbeendominatedbyDescartes’s‘first-person’orfrom-the-insideperspective.Withfewexceptions,philosophershaveacceptedthefollowingclaims:(1)thatone’sownmindisbetterknownthanone’sbody,(2)thatthemindismetaphysicallyinthebody’sdriver’sseat,and(3)thatthereisatleastatheoreticalproblemofhowwehumanintelligencescanknowthat‘external’,everydayphysicalobjectsexistatall,eveniftherearetenablesolutionstothatproblem.Wesubjectsareimmuredwithinamovietheatreofthemind,thoughwemayhavesomedefensiblewaysofinferringwhatgoesonoutsidethetheatre.\nWILLIAMG.LYCANMidwaythroughthetwentiethcentury,allthissuddenlychanged,fortworeasons.Thefirstreasonwastheaccumulatedimpactoflogicalpositivismandtheverificationtheoryofmeaning.Intersubjectiveverifiabilityortestabilitybecamethecriterionbothofscientificprobityandoflinguisticmeaningitself.Ifthemind,inparticular,wastoberespectedeitherscientifically,orevenasmeaningfullydescribableinthefirstplace,mentalascriptionswouldhavetobepeggedtopublicly,physicallytestableverification-conditions.Sciencetakesanintersubjective,‘third-person’perspectiveoneverything;thetraditionalfirst-personperspectivehadtobeabandonedforscientificpurposesand,itwasfelt,forseriousMETAPHYSICAL(chapter2)purposesalso.ThesecondreasonwastheemergenceofanumberofpressingphilosophicalobjectionstoCartesiandualism.Hereareafew:(1)ImmaterialCartesianmindsandghostlynon-physicaleventswereincreasinglyseentofitillwithourotherwisephysicalandscientificpictureoftheworld,uncomfortablylikespooksorectoplasmthemselves.Theyarenotneededfortheexplanationofanypubliclyobservablefact,forneurophysiologypromisestoexplainthemotionsofourbodiesinparticularandtoexplainthemcom-pletely.Indeed,ghost-mindscouldnotverywellhelpinsuchanexplanation,sincenothingisknownofanypropertiesofspookstuffthatwouldbearonpublicphysicaloccurrences.(2)Sincehumanbeingsevolvedoveraeons,bypurelyphysicalprocessesofmutationandnaturalselection,fromprimitivecreaturessuchasone-celledorganismswhichdidnothaveminds,itisanomaloustosupposethatatsomepointMotherNature(intheformofpopulationgenetics)somehowcreatedimmaterialCartesianmindsinadditiontocellsandphysicalorgans.Thesamepointcanbeputintermsofthedevelopmentofasinglehumanzygoteintoanembryo,thenafoetus,ababyandfinallyachild.(3)Ifmindsreallyareimmaterialandutterlynon-spatial,howcantheypossiblyinteractcausallywithphysicalobjectsinspace?(Descarteshimselfwasveryuncomfortableaboutthis.Atonepointhesuggestedgravityasamodelfortheactionofsomethingimmaterialonaphysicalbody;butgravityisspatialinnatureeventhoughitisnottangibleinthewaythatbodiesare.)(4)Inanycaseitdoesnotseemthatimmaterialentitiescouldcausephysicalmotionconsistentlywiththeconservationlawsofphysics,suchasthatregard-ingmatter–energy;physicalenergywouldhavetovanishandreappearinsidehumanbrains.2BehaviourismWhatisanalternativetodualism?Carnap(1932–3)andRyle(1949)notedthattheobviousverification-conditionsortestsformentalascriptionsarebehavioural.Howcantherestofustellthatyouareinpain,savebyyourwincing-and-groaningbehaviourincircumstancesofpresumabledisorder,orthatyoubelievethatbroccoliwillkillyou,savebyyourverbalavowalsandyouravoidanceofbroccoli?Ifthetestsarebehavioural,174\nPHILOSOPHYOFMINDthen(itwasargued)theverymeaningsoftheascriptions,oratleasttheonlyfactsgenuinelydescribed,arenotghostlyorineffablebutbehavioural.Thusbehaviourismasatheoryofmindandaparadigmforpsychology.Inacademicpsychology,behaviourismtookprimarilyamethodologicalform.Psychologicalbehaviouristsclaimed(1)thatpsychologyitselfisascienceforthepredictionandcontrolofbehaviour,(2)thattheonlyproperdataorobservationalinputforpsychologyarebehavioural,specificallypatternsofphysicalresponsestophysicalstimuli,and(3)thatinnerstatesandevents,eitherneurophysiologicalormental,arenotproperobjectsofpsychologicalinvestigation–neurophysiologicalstatesandeventsarethebusinessofbiologists,andmentalstatesandevents,sofarastheyexistatall,arenotseriouslytobementioned.Officially,thepsychologicalbehaviouristsmadenometaphysicalclaims;mindsandmentalentitiesmightexistforalltheyknew,butthiswasnottobepresumedinpsychologicalexperimentortheorizing.Psychologicaltheorizingwastoconsist,àlalogicalpositivism,ofthesubsumingofempiricallyestablishedstimulus–responsegeneralizationsunderbroaderstimulus–responsegeneralizations.Inphilosophy,behaviourismdid(naturally)takeametaphysicalform:chieflythatofanalyticalbehaviourism,theclaimthatmentalascriptionssimplymeanthingsaboutbehaviouralresponsestoenvironmentalimpingements.Thus,‘Edmundisinpain’means,notanythingaboutEdmund’sputativeghostlyego,orevenaboutanyepisodetakingplacewithinEdmund,butthateitherEdmundisactuallybehavinginawincing-and-groaningwayorheisdisposedsotobehave(inthathewouldsobehaveweresomethingnotkeepinghimfromdoingso).‘Edmundbelievesthatbroccoliwillkillhim’meansjustthat,ifasked,Edmundwouldassenttothatproposition,and,ifconfrontedbybroccoli,Edmundwouldshunit,andsoforth.Butitshouldbenotedthatabehaviouristmetaphysicianneedmakenoclaimaboutthemeaningsofmentalexpressions.Onemightbeamerelyreductivebehaviourist,andholdthatalthoughmentalascriptionsdonotsimplymeanthingsaboutbehaviouralresponsestostimuli,theyareultimately(inreality)madetruejustbythingsaboutactualandhypotheticalresponsestostimuli.(Onthedifferencebetweenanalyticreduc-tionbylinguisticmeaningandsyntheticreductionbyaposterioriidentification,seesection3below.)Oronemightbeaneliminativebehaviourist,andholdthattherearenomentalstatesoreventsatall,butonlybehaviouralresponsestostimuli,mentalascrip-tionsbeinguniformlyfalseormeaningless.AnybehaviouristwillsubscribetowhathascometobecalledtheTuringTest.Inresponsetotheperenniallypopularquestion‘Canmachinesthink?’,AlanTuring(1964)repliedthatabetterquestionisthatofwhetherasophisticatedcomputercouldeverpassabatteryofverbaltests,totheextentoffoolingalimitedobserver(say,ahumanbeingcorrespondingwithitbymail)intothinkingitishumanandsentient.Ifamachinedidpasssuchtests,thentheputativelyfurtherquestionofwhetherthemachinereallythoughtwouldbeidleatbest,whatevermetaphysicalanalysisonemightattachtoit.BarringTuring’stendentiouslimitationofthemachine’sbehaviourtoverbalasopposedtonon-verbalresponses,anybehaviourist,psychologicalorphilosophical,wouldagreethatpsychologicaldifferencescannotout-runbehaviouraltests;organisms(includingmachines)whoseactualandhypotheticalbehaviourisjustthesamearepsychologicallyjustalike.175\nWILLIAMG.LYCANBesidessolvingthemethodologicalproblemofintersubjectiveverification,philo-sophicalbehaviourismalsoadroitlyavoidedanumberoftheobjectionstoCartesiandualism,including(1)–(4)above.ItdispensedwithimmaterialCartesianegosandghostlynon-physicalevents,writingthemoffasmetaphysicalexcrescences.ItdisposedofDescartes’sadmittedproblemofmind–bodyinteraction,sinceitpositednoimmate-rial,non-spatialcausesofbehaviour.ItraisednoscientificmysteriesconcerningtheinterventionofCartesiansubstancesinphysicsorbiology,sinceitcountenancednosuchintervention.Thusitisamaterialistview,asagainstDescartes’simmaterialism.Yetsometheoristswereuneasy;theyfeltthatinitstotalrepudiationoftheinner,theprivateandthesubjective,behaviourismwasleavingoutsomethingrealandimportant.Whenthisworrywasvoiced,thebehaviouristsoftenrepliedwithmockery,assimilatingthedoubterstoold-fashioneddualistswhobelievedinghosts,ectoplasmortheEasterBunny;behaviourismwastheonly(evenhalfwaysensible)gameintown.Nonetheless,thedoubtersmadeseverallastingpointsagainstit.Firstly,peoplewhoarehonestandnotanaesthetizedknowperfectlywellthattheyexperience,andcanintrospect,actualinnermentalepisodesoroccurrences,thatareneitheractuallyaccompaniedbycharacteristicbehaviournormerelystatichypotheticalfactsofhowtheywouldbehaveifsubjectedtosuch-and-suchastimulation.Place(1956)spokeofan‘intractableresidue’ofconsciousmentalstatesthatbearnoclearrelationstobehav-iourofanyparticularsort;seealsoArmstrong(1968:ch.5)andCampbell(1984).Sec-ondly,contrarytotheTuringTest,itseemsperfectlypossiblefortwopeopletodifferpsychologicallydespitetotalsimilarityoftheiractualandhypotheticalbehaviour,asinaLOCKEAN(chapter29)caseof‘invertedspectrum’:itmightbethatwhenyouseearedobject,youhavethesortofcolourexperiencethatIhavewhenIseeagreenobject,andviceversa.Forthatmatter,acreaturemightexhibitalltheappropriatestimu-lus–responserelationsandlackamentallifeentirely;wecanimaginebuildinga‘zombie’orstupidrobotthatbehavesintherightwaysbutdoesnotreallyfeelorthinkanythingatall(BlockandFodor1972;Kirk1974;Block1981;Campbell1984).Thirdly,theanalyticalbehaviourist’sbehaviouralanalysesofmentalascriptionsseemadequateonlysolongasonemakessubstantiveassumptionsabouttherestofthesubject’smentality(Chisholm1957:ch.11;Geach1957:8;Block1981);forexample,ifEdmundbelievesthatbroccoliwillkillhimandheisofferedsomebroccoli,hewouldshunitonlyifhedoesnotwanttodie.Therefore,thebehaviouristanalysesareeithercircularorradicallyincomplete,sofarastheyaresupposedtoexhaustthementalgenerally.Somattersstoodinstalematebetweendualists,behaviouristsanddoubters,untilthemid-1950s,whenU.T.Place(1956)andJ.J.C.Smart(1959)proposedamiddleway,aconciliatorycompromisesolution.3TheIdentityTheoryAccordingtoPlaceandSmart,contrarytothebehaviourists,atleastsomementalstatesandeventsaregenuinelyinnerandgenuinelyepisodicafterall.Theyarenottobeidentifiedwithoutwardbehaviourorevenwithhypotheticaldispositionstobehave.But,contrarytothedualists,theepisodicmentalitemsareneitherghostlynor176\nPHILOSOPHYOFMINDnon-physical.Rather,theyareneurophysiological.Theyareidenticalwithstatesandeventsoccurringintheirowners’centralnervoussystems;moreprecisely,everymentalstateoreventisnumericallyidenticalwithsomesuchneurophysiologicalstateorevent.Tobeinpainis,forexample,tohaveone’sc-fibres,ormorelikelya-fibres,firinginthecentralnervoussystem;tobelievethatbroccoliwillkillyouistohaveone’sBbk-fibresfiring,andsoon.Bymakingthementalentirelyphysical,thisidentitytheoryofthemindsharedthebehaviouristadvantageofavoidingtheobjectionstodualism.Butitalsobrilliantlyaccommodatedtheinnerandtheepisodicasbehaviourismdidnot.For,accordingtotheidentitytheory,mentalstatesandeventsactuallyoccurintheirowners’centralnervoussystems.(HencetheyareinnerinanevenmoreliteralsensethancouldbegrantedbyDescartes.)Theidentitytheoryalsothoroughlyvindicatedtheideathatorganismscandiffermentallydespitetotaloutwardbehaviouralsimilarity,sinceclearlyorganismscandifferneurophysiologicallyinmediatingtheiroutwardstimu-lus–responseregularities;thatwouldaffordthepossibilityofinvertedspectrum.Andofcoursetheconnectionbetweenabelieforadesireandtheusuallyaccompanyingbehaviourisdefeasiblebyothercurrentmentalstates,sincetheconnectionbetweenaB-orD-neuralstateanditsnormalbehaviouraleffectisdefeasiblebyotherpsychol-ogicallycharacterizableinteractingneuralstates.Theidentitytheorywastheidealresolutionofthedualist–behaviouristimpasse.Moreover,therewasadirectdeductiveargumentfortheidentitytheory,hituponindependentlybyDavidLewis(1966,1972)andD.M.Armstrong(1968).LewisandArmstrongmaintainedthatmentaltermsweredefinedcausally,intermsofmentalitems’typicalcausesandeffects.Forinstance,theword‘pain’meansastatethatistypicallybroughtaboutbyphysicaldamageandthattypicallycauseswithdrawal,favouring,complaint,desireforcessation,andsoon.(Armstrongclaimedtoestablishthisbystraightforward‘conceptualanalysis’.Moreelaborately,Lewisheldthatmentaltermsarethetheoreticaltermsofacommon-sensical‘folktheory’,andwiththeposi-tiviststhatalltheoreticaltermsareimplicitlydefinedbythetheoriesinwhichtheyoccur.Thatcommon-sensetheoryhassincecometobecalled‘folkpsychology’;seesec-tions12and13below.)Nowifbydefinitionpainiswhateverstateoccupiesacertaincausalniche,andifasisoverwhelminglylikely,scientificresearchwillrevealthatthatparticularnicheisinfactoccupiedbysuch-and-suchaneurophysiologicalstate,itfollowsstraightwaythatpainisthatneurophysiologicalstate;QED.Painretainsitsconceptualconnectiontobehaviour,butalsoundergoesanempiricalidentificationwithaninnerstateofitsowner.(AnadvancedifconvolutedelaborationofthisalreadyhybridviewisdevelopedbyLewis1980;formeticulousdiscussion,seeBlock1978;Shoemaker1981;Tye1983;Owens1986.)NoticethatalthoughArmstrongandLewisbegantheirargumentswithaclaimaboutthemeaningsofmentalterms,their‘common-sensecausal’versionoftheiden-titytheorywasitselfnosuchclaim,anymorethanwastheoriginalidentitytheoryofPlaceandSmart.Rather,allfourphilosophersreliedontheideathatthingsorproper-tiescansometimesbeidentifiedwith‘other’thingsorpropertiesevenwhenthereisnosynonymyofterms;thereissuchathingassyntheticandaposterioriidentitythatisnonethelessgenuineidentity.Whiletheidentityoftriangleswithtrilateralsholdssimplyinvirtueofthemeaningsofthetwotermsandcanbeestablishedbyreason177\nWILLIAMG.LYCANalone,withoutempiricalinvestigation,thefollowingidentitiesarestandardexamplesofthesyntheticaposteriori,andwerediscoveredempirically:cloudswithmassesofwaterdroplets;waterwithH2O;lightningwithelectricaldischarge;theMorningStarwithVenus;MendeliangeneswithsegmentsofDNAmolecules;andtemperaturewithmeanmolecularkineticenergy.Theidentitytheorywasofferedsimilarly,inaspiritofscientificspeculation;onecouldnotproperlyobjectthatmentalexpressionsdonotmeananythingaboutbrainsorneuralfirings.Sothedualistswerewronginthinkingthatmentalitemsarenon-physicalbutrightinthinkingtheminnerandepisodic;thebehaviouristswererightintheirmaterialismbutwrongtorepudiateinnermentalepisodes.Butalas,thishappysynthesiswastoogoodtobetrue.4MachineFunctionalismQuitesoon,HilaryPutnam(1960,1967a,1967b)andJerryFodor(1968b)pointedoutapresumptuousimplicationoftheidentitytheoryunderstoodasatheoryof‘types’orkindsofmentalitems:thatamentalstatesuchaspainhasalwaysandeverywheretheneurophysiologicalcharacterizationinitiallyassignedtoit.Forexample,iftheiden-titytheoristidentifiedpainitselfwiththefiringsofc-fibres,itfollowedthatacreatureofanyspecies(earthlyorscience-fiction)couldbeinpainonlyifthatcreaturehadc-fibresandtheywerefiring.Butsuchaconstraintonthebiologyofanybeingcapableoffeelingpainisbothgratuitousandindefensible;whyshouldwesupposethatanyorganismmustbemadeofthesamechemicalmaterialsasweareinordertohavewhatcanbeaccuratelyrecognizedaspain?Theidentitytheoristhadover-reactedtothebehaviourists’difficultiesandfocusedtoonarrowlyonthespecificsofbiologicalhumans’actualinnerstates,andinsodoingtheyhadfallenintospecieschauvinism.PutnamandFodoradvocatedtheobviouscorrection:whatwasimportantwasnotitsbeingc-fibres(perse)thatwerefiring,butwhatthec-fibrefiringsweredoing,whattheycontributedtotheoperationoftheorganismasawhole.Theroleofthec-fibrescouldhavebeenperformedbyanymechanicallysuitablecomponent;solongasthatrolewasperformed,thepsychologyofthecontainingorganismwouldhavebeenunaf-fected.Thus,tobeinpainisnotpersetohavec-fibresthatarefiring,butmerelytobeinsomestateorother,ofwhateverbiochemicaldescription,thatplaysthesamecausalroleasdidthefiringsofc-fibresinthehumanbeingswehaveinvestigated.Wemaycontinuetomaintainthatpain‘tokens’–individualinstancesofpainoccurringinpar-ticularsubjectsatparticulartimes–arestrictlyidenticalwithparticularneurophysio-logicalstatesofthosesubjectsatthosetimes–inotherwords,withthestatesthathappentobeplayingtheappropriateroles;thisisthethesisof‘tokenidentity’or‘token’materialismorphysicalism.Butpainitself,thekind,universalor‘type’,canbeidenti-fiedonlywithsomethingmoreabstract:thecausalorfunctionalrolethatc-fibrefiringssharewiththeirpotentialreplacementsorsurrogates.Mentalstate-typesareidentifiednotwithneurophysiologicaltypesbutwithmoreabstractfunctionalroles,asspecifiedbystate-tokens’causalrelationstotheorganism’ssensoryinputs,behaviouralresponses,andotherinterveningpsychologicalstates.Functionalism,then,isthedoctrinethatwhatmakesamentalstatethetypeofstateitis–apain,asmellof178\nPHILOSOPHYOFMINDviolets,abeliefthatkoalasaredangerous–isitsdistinctivesetoffunctionalrelations,itsroleinitssubject’sbehaviouraleconomy.Putnamcomparedmentalstatestothefunctionalor‘logical’statesofacomputer:justasacomputerprogramcanberealizedorinstantiatedbyanyofanumberofphysicallydifferenthardwareconfigurations,socanapsychological‘program’berealizedbydifferentorganismsofvariousphysiochemicalcomposition,andthatiswhydifferentphysiologicalstatesoforganismsofdifferentspeciescanrealizeoneandthesamementalstate-type.Whereanidentitytheorist’stype-identificationwouldtaketheform,‘TobeinmentalstateoftypeMistobeintheneurophysiologicalstateoftypeN’,Putnam’smachinefunctionalism(asIshallcallit)hasitthattobeinMistobemerelyinsomephysiologicalstateorotherthatplaysroleRintherelevantcomputerprogram(thatis,theprogramthatatasuitablelevelofabstractionmediatesthecreature’stotaloutputsgiventotalinputsandsoservesasthecreature’sglobalpsychology).Thephysiologicalstate‘playsroleR’inthatitstandsinasetofrelationstophysicalinputs,outputsandotherinnerstatesthatmatchesone-to-onetheabstractinput–output–logical-staterelationscodifiedinthecomputerprogram.FunctionalistLevelsofDescriptionThefunctionalist,then,mobilizesthreedistinctlevelsofdescriptionbutappliesthemalltothesamefundamentalreality.Aphysicalstate-tokeninsomeone’sbrainataparticu-lartimehasaneurophysiologicaldescription,butitmayalsohaveafunctionaldescrip-tionrelativetoamachineprogramthatthebrainhappenstoberealizing,anditmayfurtherhaveamentaldescriptionifsomementalstateiscorrectlytype-identifiedwiththefunctionalcategoryitexemplifies.Andsothereisafterallasenseinwhich‘themental’isdistinctfrom‘thephysical’.Thoughpresumablytherearenonon-physicalsub-stancesorstuffs,andeverymentaltokenisitselfentirelyphysical,mentalcharacteriza-tionisnotphysicalcharacterization,andthepropertyofbeingapainisnotsimplythepropertyofbeingsuch-and-suchaneuralfiring.Moreover,unlikebehaviourismandtheidentitytheory,functionalismdoesnotstrictlyentailthatmindsarephysical;itmightbetrueofnon-physicalminds,solongasthosemindsrealizedtherelevantprograms.5CognitivePsychologyInanotaccidentallysimilarvein,behaviourismhasalmostentirelygivenwayto‘cog-nitivism’inpsychology.Cognitivismisroughlytheviewthat(1)psychologistsmayandmustadverttoinnerstatesandepisodesinexplainingbehaviour,solongasthestatesandepisodesareconstruedthroughoutasphysical;(2)humanbeingsandotherpsy-chologicalorganismsarebestviewedasinsomesenseinformation-processingsystems;and(3)thevehicleofinformationprocessingisinternalrepresentation,arepertoireofbrainstatesthatrepresentfeaturesoftheexternalworld.Ascognitivepsychologysetstheagenda,itsquestionstaketheform,‘Howdoesthisorganismreceiveinformationthroughitssense-organs,representtheinformation,processit,storetheresult,andthenmobilizeitstotalinformationinsuchawayastoresultinintelligentbehaviour?’Theworkinglanguageofcognitivepsychologyishighlycongenialtothefunctionalist,179\nWILLIAMG.LYCANforcognitivismthinksofhumanbeingsassystemsofinterconnectedfunctionalcom-ponents,interactingwitheachotherinanefficientandproductiveway.Thiswasnowherebetterexemplified,intheearlydaysofthecognitivistrevolution,thaninChomsky’s(1959)reviewofSkinner(1957),whicharguedforafunctionalistinformation-processingaccountoflanguageuse,againstthebestbehaviouristmodel.6ArtificialIntelligenceandtheComputerModeloftheMindMeanwhile,researchersincomputersciencehavepursuedfruitfulresearchpro-grammesbasedontheideaofintelligentbehaviouras,forgiveninput,theoutputofskilfulinformation-processing.ArtificialIntelligence(AI)is,roughly,theprojectofgettingcomputingmachinestoperformtasksthatwouldusuallybetakentodemandhumanintelligenceandjudgement.Computershaveachievedsomemodestsuccessinprovingtheorems,guidingmissiles,sortingmail,drivingassembly-linerobots,diag-nosingillnesses,predictingweatherandeconomicevents,andthelike.Acomputerjustisamachinethatreceives,interprets,processes,stores,manipulatesandusesinfor-mation,andAIresearchersthinkofitinjustthatwayastheytrytoprogramintelli-gentbehaviour.AnAIproblemtakestheform,‘Giventhatthemachineseesthisasinput,whatmustitalreadyknowandwhatmustitaccordinglydowiththatinputinordertobeableto...[recognize,identify,sort,puttogether,predict,tellus,andsoon]?Andhow,then,canwestartitoffknowingthatandgetittodothosethings?’SowemayreasonablyattributesuchsuccessasAIhashadtoself-consciousrelianceontheinformation-processingparadigm.Thisencouragestheideathathumanintelligenceandcognitiongenerallyaremattersofcomputationalinformation-processing.Indeed,thatideahasalreadyfilteredwelldownintotheeverydayspeechofordinarypeople,asanyonecanverifybytalkingwithchildrenforawhile.Thistentativeandcrudecoalescingofthenotionscognition,computation,informationandintelligenceraisestwogeneralquestions,oneineachoftwodirections:(1)towhatextentmightcomputersapproximatetominds?(2)towhatextentdomindsapproximatetocomputers?Thefirstquestionbreaksdownintothreemore,whichdiffersharplyandimportantlyfromeachother.(1a)Whatintelligenttaskswillanycomputereverbeabletoperform?(1b)GiventhatacomputerperformsinterestingtasksX,YandZ,doesitdosointhesamewaythathumanbeingsdo?(1c)GiventhatacomputerperformsX,YandZandthatitdoessointhesamewayhumansdo,doesthatshowthatithaspsychologicalandmentalproperties,suchas(real)intelligence,thought,consciousness,feeling,sensa-tion,emotionandthelike?Sub-question(1a)isoneofengineering,(1b)isoneofcog-nitivepsychologyand(1c)isphilosophical;theorists’answerswilldependaccordinglyontheircommitmentsintheserespectiveareas.Butfortherecordletusdistinguishthreedifferentsensesorgradesof‘AI’:AIintheweakestsenseiscautiouslyoptimisticasregards(1a);itsaystheseengineeringeffortsarepromisingandshouldbefunded,fortheirownsake.AIinastrongersensesaysthattheengineeringeffortscanwellserveasmodellingofhumancognition,andthattheirsuccessescanbetakenaspointerstowardthetruthabouthumanfunctionalandrepresentationalorganization.AIinthe180\nPHILOSOPHYOFMINDstrongestsensefavoursanaffirmativeanswerto(1c)andsomequalifiedrespectfortheTuringTest:itsaysthatifamachineperformsintelligentlyanddoessoonthebasisofasufficientlyhuman-likeinformation-processingaetiology,thenthereislittlereasontodoubtthatthemachinehastherelevanthumanqualitiesofmindandsensation.Onthisview,anysystemororganismthat(genuinely)realizessuch-and-suchaprogramorflowchartwouldhavementalstateslikeours.Thus,ifwecouldbuildamachinetocertainfunctionalspecifications,wecouldartificiallycreateathinking,feelingmachine.Sometheoristswelcomethisconsequencewarmly,andlookforwardtotheday.Othersfindtheprospectgrotesqueandpreposterous,areductionoffunctionalismtoabsurdity.(AIinthestrongestsenseisverystrong,butnoticecarefullythatitdoesnotpresupposeaffirmativeanswerstoeither(1a)or(1b).)Theoppositeissue,thatofassimilatingmindstocomputers,isveryclosetothephilosophicalmatteroffunction-alism.Butheretoothereareimportantlydistinctsub-questions,thistimetwo:(2a)Dohumanmindsworkinverylikethewaycomputersdoascomputersarecurrentlydesignedandconstrued–forinstance,usingflipflopsgroupedintobanksandregisters,withanassemblylanguagecollectingindividualmachine-codeoperationsintosub-routinesandthesesubroutinesbeingcalledbyhigher-levelmanipulationsofreal-worldinformationaccordingtoprogrammedrules?(2b)Regardlessofarchitecture,canhumanpsychologicalcapacitiesbeentirelycapturedbyathird-person,hardware-realizabledesignofsomesortthatcouldinprinciplebebuiltinalaboratory?Sub-question(2a)iscurrentlymuchdiscussed(seesection12below),butisnotparticularlyphilosophical.(2b)istantamounttothefateoffunctionalismstrictlyso-called.Thecomputermodelofthemindhasopponentsofseveralsorts;ofthese,perhapsthebestknownareDreyfus(1979)andSearle(1980).Dreyfusisthoroughlypessimisticinregardto(1a)andgivesanemphaticallynegativeanswerto(2a).DrawingonthePHENOMENOLOGICALtraditionofHUSSERLandHEIDEGGER(chapter41),Dreyfuschal-lengesseveraloftheassumptionsunderlyingclassicalAI,anddiagnosessomeofAI’sknownfailuresasstemmingfromthoseassumptions.Inparticular,Dreyfusarguesthat,inreallifeasopposedtothetoy‘micro-worlds’constructedbyAIresearchers,suchstandardAItasksasproblemsolvingandlanguageunderstandingpresupposeunpre-dictablyrelevantandnearlyendlessknowledgeoftheworld(considerjustthegeneralproblemofunderstandingambiguoussentencesincontext);butnohumanbeing’sknowledgeoftheworldcouldeverbemadeexplicit,orevencompletelyformalizedinasetofrepresentationsandrules.Rather,humanknowledgeandunderstandinginherenon-discursivelyinourbiology,ourbodilyembeddinginourenvironment,ourinterestsandourculturalsurround–ofwhichdigitalcomputershaveatbestdistantanalogues(seealsoHaugeland1979).Morerecently,however,DreyfushasgrantedthatconnectionistAI(seesection12below)evadeshismaincriticismsofclassicalAI(seeespeciallyDreyfusandDreyfus1986).Searleoffersahypotheticalcaseinwhichahumanprogramisexecuted,notbyamechanicalorbiologicalcomputerinthenormalway,butbyasimplehumanbeingwhoisgiven‘inputs’,sayonindexcards,andfollowingsomeruleswritteninamanual,writessome‘outputs’onnewindexcardsanddropsthoseinan‘output’bin.Supposethe‘inputs’aresequencesofChinesecharacters,theprogramisthatwhichafunc-tionalistwouldsayconstitutesahumanChinesespeaker’sunderstandingofthe181\nWILLIAMG.LYCANChineselanguage,andconsequentlythe‘outputs’aregrammaticalutterancesofChinesealso;butthefunctionarywhodoesthesimplerule-followingoperationisamonoglotEnglish-speakerandonlyjustbrightenoughtofollowtheprogram’sverysimpleinstructions.ThereisherenoactualunderstandingofChinese,Searleclaims,butonlyacomputationalimitationofarealChinesespeakerandunderstander.Thus,sub-questions(1c)and(2b)areansweredfirmlyinthenegative.AnumberofdifferingrejoindersareavailabletopartisansofAIinitsstrongandstrongestsenses(theyaremadebyvariouscommentatorsintheBehavioralandBrainSciencessymposium–seeSearle1980).Thehypotheticalexampleisnotreallypos-sible,evenifweseemtobeabletoimagineit;or,thoughitispossible,itsprotagonistwould(contraSearle)reallyunderstandChinese;or,thoughSearleisrightabouttheexampleasdescribed,itcanbeaccommodatedbyanonlyslightlymodifiedfunctional-isttheory(inparticular,seesection10below).AndMorality?Itmaybewonderedwhethermaterialisttheoriesofthemindand/orfunctionalisttheo-rieshaveanyinterestingimplicationsformoralityandethics.Threematerialiststakethisupexplicitly:Smart(1963:ch.8)triestoexhibitamaterialistbasisformorals;MichaelLevin(1979:ch.7)addressesthespecificchargethatmaterialistscannotallowfreedomofthewillorwhateverelsemaybenecessarytomakeroomformoralresponsibility;Lycan(1985)exploressomemoralconsequencesofthecomputationalviewofthemind.AmainpurposeofDennett’s(1978)isalsotoshowwhymoralresponsibilityandthementalvernacularthatsupportsitarepossibledespiteDennett’sinstrumentalist–sometimesfictionalist–treatmentofthemental(seesection11below).7HomuncularFunctionalismandOtherTeleologicalTheoriesAswesaw,machinefunctionalismsupposedthathumanbrainsmaybedescribedateachofthreelevels,thefirsttwoscientificandthethirdfamiliarandcommon-sensical.Atthefirstlevel,biologistswouldmapouthumanneurobiologyandprovideneurophysiologicaldescriptionsofbrainstates.Atthesecondlevel,psychologistswould(eventually)workoutthemachineprogramthatwasbeingrealizedbythelower-levelneurophysiologyandwoulddescribethesamebrainstatesinmoreabstract,computa-tionalterms.Atthethirdlevel,psychologistswouldalsoexplainbehaviour,character-izedineverydayterms,byreferencetostimuliandtointerveningmentalstatessuchasbeliefsanddesires,type-identifyingthementalstateswithfunctionalorcomputationalstatesastheywent.Suchexplanationswouldthemselvespresupposenothingaboutneurophysiology,sincetherelevantpsychological–computationalgeneralizationswouldholdregardlessofwhatparticularbiochemistrymighthappentoberealizingtheabstractprograminquestion.Machinefunctionalismhasmorerecentlybeenchallengedonanumberofpoints,thattogethermotivateaspecificallyteleologicalnotionof‘function’(Sober(1985)speaksaptlyof‘puttingthefunctionbackintofunctionalism’):wearetothinkofa182\nPHILOSOPHYOFMINDthing’sfunctionaswhatthethingisfor,whatitsjobis,whatitissupposedtodo.Therearefivechallenges.Firstly,themachinefunctionaliststillconceivedpsychologicalexplanationinthelogicalpositivists’termsofsubsumingobserveddataunderwiderandwideruniversallaws.ButFodor(1968a),Dennett(1978)andCummins(1983)havedefendedacom-petingpictureofpsychologicalexplanation,accordingtowhichbehaviouraldataaretobeseenasmanifestationsofsubjects’psychologicalcapacities,andthosecapacitiesaretobeexplainedbyunderstandingthesubjectsassystemsofinterconnectedcom-ponents.Eachcomponentisa‘homunculus’,inthatitisthoughtofasalittleagentorbureaucratoperatingwithinitscontainingsubject;itisidentifiedbyreferencetothefunctionitperforms.Andthevarioushomuncularcomponentsco-operatewitheachotherinsuchawayastoproduceoverallbehaviouralresponsestostimuli.The‘homun-culi’arethemselvesbrokendownintosub-componentswhosefunctionsandinterac-tionsaresimilarlyusedtoexplainthecapacitiesofthesubsystemstheycompose,andsoagainandagainuntilthesub-sub-...componentsareseentobeneurophysiologi-calstructures.Thusbiologicalandmechanicalsystemsalikearehierarchicallyorga-nized,ontheprincipleofwhatcomputerscientistscall‘hierarchicalcontrol’.(Anautomobileworks–locomotes–byhavingafuelreservoir,afuelline,acarburettor,acombustionchamber,anignitionsystem,atransmissionandwheelsthatturn.Ifonewantstoknowhowthecarburettorworks,onewillbetoldwhatitspartsareandhowtheyworktogethertoinfuseoxygenintofuel;andsoon.)Butnothinginthispatternofexplanationcorrespondstothesubsumptionofdataunderwiderandwideruniversalgeneralizations.Thesecondchallengeisthatthemachinefunctionalisttreatedfunctional‘realiza-tion’,therelationbetweenanindividualphysicalorganismandtheabstractprogramitwassaidtoinstantiate,asasimplematterofone-to-onecorrespondencebetweentheorganism’srepertoireofphysicalstimuli,structuralstatesandbehaviour,ontheonehand,andtheprogram’sdefininginput–state–outputfunctionontheother.Butthiscriterionofrealizationwasseentobetooliberal;sincevirtuallyanythingbearsaone–onecorrelationofsomesorttovirtuallyanythingelse,‘realization’inthesenseofmereone–onecorrespondenceisfartooeasilycomeby(Block1978;Lycan1987:ch.3);Searle’s‘Chinese-understanding’set-upwouldbeanexample.Sometheoristshaveproposedtoremedythisdefectbyimposingateleologicalrequirementonrealization:aphysicalstateofanorganismwillcountasrealizingsuch-and-suchafunctionaldescriptiononlyiftheorganismhasgenuineorganicintegrityandthestateplaysitsfunctionalroleproperlyfortheorganism,intheteleologicalsenseof‘for’andintheteleologicalsenseof‘function’.Thestatemustdowhatitdoesasamatterof,sotospeak,itsbiologicalpurpose.(Machinefunctionalismtook‘function’initssparemathematicalsenseratherthaninagenuinelyfunctionalsense.Oneshouldnotethat,asusedhere,theterm‘machinefunctionalism’istiedtotheoriginalliberalconceptionof‘realizing’;sotoimposeateleologicalrestrictionistoabandonmachinefunctionalism.)Thirdly,ofthemachinefunctionalist’sthreelevelsofdescription,oneiscommon-sensicalandtwoarescientific,soweareofferedastarklytwo-levelledpictureofhumanpsychobiology.Butthatpictureisunbiologicalintheextreme.Neitherlivingthingsnorevencomputersthemselvesaresplitintoapurely‘structural’levelof183\nWILLIAMG.LYCANbiological–physiochemicaldescriptionandanyone‘abstract’computationallevelofmachine–psychologicaldescription.Rather,theyareallhierarchicallyorganizedatmanylevels,eachlevel‘functional’withrespecttothosebeneathitbut‘structural’orconcreteasitrealizesthoselevelsaboveit.Thisrelativityofthe‘functional’–‘structural’or‘software’–‘hardware’distinctiontoanyspecifiedleveloforganizationhasrepercussionsforfunctionalistsolutionstoproblemsinthephilosophyofmind(Lycan1987:ch.5),andforcurrentcontroversiessurroundingConnectionismandneuralmodellinginAI.Fourthly,theteleologizingoffunctionalrealizationhashelpedfunctionaliststorebutvariousobjectionsbasedonthe‘qualia’or‘feels’orexperiencedphenomenalcharactersofmentalstates(Lycan1981;Sober1985;andseesection8below).Finally,VanGulick(1980),Millikan(1984),Dretske(1988),Fodor(1990a)andothershavearguedpowerfullythatteleologymustenterintoanyadequateanalysisoftheintentionalityoraboutnessorreferentialcharacterofmentalstatessuchasbeliefsanddesires,byreferencetothestates’psychobiologicalfunctions(seesections9and10below).Beliefs,desiresandotherpropositionalattitudessuchassuspecting,intend-ingandwishingaredirecteduponstatesofaffairswhichmayormaynotactuallyobtain(forinstance,thattheRepublicancandidatewillwin),andareaboutindividu-alswhomayormaynotexist(suchasKingArthur).FranzBrentano(1838–1917)(1973)drewadistinctionbetweenpsychologicalphenomena,whicharedirecteduponobjectsandstatesofaffairs,evennon-existingones,andphysicalobjects,whicharenotsodirected.Ifmentalitemsarephysical,however,thequestionariseshowanypurelyphysicalentityorstatecouldhavethepropertyofbeing‘directedupon’oraboutanon-existentstateofaffairsorobject;thatisnotthesortoffeaturethatordinary,purelyphysicalobjects(suchasbricks)canhave.Accordingtotheteleologicaltheorists,aneurophysiologicalstateshouldcountasabeliefthatbroccoliwillkillyou,andinparticularasaboutbroccoli,onlyifthatstatehastherepresentingofbroccoliasinsomesenseoneofitspsychobiologicalfunctions.Ifteleologyisneededtoexplicateinten-tionality,andmachinefunctionalismaffordsnoteleology,thenmachinefunctionalismisnotadequatetoexplicateintentionality.Allthistalkofteleologyandbiologicalfunctionseemstopresupposethatbiologicalandother‘structural’statesofphysicalsystemsreallydohavefunctionsintheteleo-logicalsense.Thelatterclaimis,tosaytheleast,controversial.Somephilosophersdismissitashilariouslyfalse,asasuperstitiousrelicofprimitiveanimism,Panglossiantheismoratbestthevitalismofthenineteenthcentury;otherstolerateitbutonlyasausefulmetaphor;stillotherstaketeleologicalcharacterizationstobeliterallytruebutonlyrelativetoaconvenientclassificatoryorinterpretivescheme(Cummins1975).Onlyafewwriters(Wimsatt1972;Wright1973;Millikan1984;andafewothers)havetakenteleologicalcharacterizationstobeliterallyandcategoricallytrue.Thismayseemtoembarrassteleologizedfunctionalisttheoriesofmind.Fortunatelyfortheteleologicalfunctionalistthereisnowavigorousindustrywhosepurposeistoexplicatebiologicalteleologyinnaturalisticterms,typicallyintermsofaetiology.Forexample,atraitmaybesaidtohavethefunctionofdoingFinvirtueofitshavingbeenselectedforbecauseitdidF;aheart’sfunctionistopumpbloodbecausehearts’pumpingbloodinthepasthasgiventhemaselectionadvantageandsoledtothesurvivalofmoreanimalswithhearts.Actually,nosimpleaetiologicalexplication184\nPHILOSOPHYOFMINDwilldo(Cummins1975;Boorse1976;BigelowandPargetter1987),butphilosophersofbiologyhavecontinuedtorefinetheearlieraccountsandseektomakethemintoadequatenaturalisticanalysesofgenuinefunction.Functionalism,andcognitivepsychologyconsideredasacompletetheoryofhumanthought,inheritsomeofthesamedifficultiesthatearlierbesetbehaviourismandtheidentitytheory.Theseremainingobstaclesfallintotwomaincategories:qualiaprob-lemsandintentionalityproblems.8ProblemswithQualiaandConsciousnessThequaleofamentalstateorevent(particularlyasensation)isthatstateorevent’sfeel,itsintrospectible‘phenomenalcharacter’,itsnatureasitpresentsitselftoconsciousness.Manyphilosophershaveobjectedthatneitherfunctionalistmetaphysicsnoranyoftheallieddoctrinesaforementionedcan‘explainconsciousness’,orilluminateoreventol-eratethenotionofwhatitfeelsliketobeinamentalstateofsuch-and-suchasort.Yet,saythesephilosophers,thefeelsarequintessentiallymental–itisthefeelsthatmakethementalstatesthementalstatestheyare.Something,therefore,mustbedrasticallywrongwithfunctionalism.‘The’problemofconsciousnessorqualiaisfamiliar.Indeed,itissofamiliarthatwetendtooverlookthemostimportantthingaboutit:thatitsnameisLegion,foritismany.Thereisnosingleproblemofqualia;thereareatleastninequitedistinctobjectionsthathavebeenbroughtagainstfunctionalism(someofthemapplytomaterialismgenerally).(1)Earlycriticsoftheidentitytheoryarguedthatourimmediatementalaccesstoqualiamilitatesagainsttheirbeingfeaturesofanypurelyneurophysiologicalitem.(2)SaulKripke(1972)madeingenioususeofmodaldistinctionsagainsttypeoreventokenidentity,arguingthatunlessmentalitemsarenecessarilyidenticalwithneurophysiologicalones,whichtheyarenot,theycannotbeidenticalwiththematall.Kripke’sclosereasoninghasattractedconsiderablecriticalattention.(3)Block(1978)andothershaveurgedvariouscounter-examplecasesagainstfunctionalism–examplesinwhichsomeentityseemstorealizetherightprogrambutwhichlacksoneofmentality’scrucialqualitativeaspects.(Typi-callythe‘entity’isagroupofhumanbeings,suchastheentirepopulationofChinaactingaccordingtoanelaboratesetofinstructions,butSearle’stypeofset-upwouldalsoserve.Neither,itseems,wouldbefeelinganythingonitsown.)Predictably,functionalistshaverejoinedbyarguing,foreachexample,eitherthattheproposedentitydoesnotinfactsucceedinrealizingtherightprogram(forexample,becausetherequisiteteleologyislacking)orthatthereisnogoodreasonfordenyingthattheentitydoeshavetherelevantqualitativestates.(4)Farrell(1950),Gunderson(1970,1974)andNagel(1974)haveworriedoverfirst-person–third-personasymmetriesandtheperspectivalnessorsubjective-185\nWILLIAMG.LYCANpoint-of-view-nessofconsciousness.Scienceisdesignedtoabstractawayfromperspectiveandsubjectivity;materialismseemstodenytherealityofthefirst-personexperience.(5)Nagel(1974)andJackson(1982)haveappealedtoadisparityinknowledge,asageneralanti-materialistargument:Icanknowwhatitisliketohavesuch-and-suchasensationonlyifIhavehadthatsensationmyself;noamountofobjective,third-personscientificinformationwouldsuffice.Inreply,function-alistshaveofferedanalysesof‘perspectivalness’,completewithaccountsof‘whatitislike’tohaveasensation,thatmakethosethingscompatiblewithfunctionalism.NagelandJacksonhaveargued,further,fortheexistenceofaspecial,intrinsicallyperspectivalkindoffact,thefactof‘whatitislike’,whichintractablyandinprinciplecannotbecapturedorexplainedbyphysicalscience.Functionalistshaverespondedthattheargumentscommitalogicalfallacy(specifically,thatofapplyingLeibniz’sLawinanintentionalcontext;somehaveaddedthatinanycase,to‘knowwhatitislike’ismerelytohaveanability,andinvolvesnofactofanysort,while,contrariwise,someothertheoristshavegrantedthattherearefactsof‘whatitislike’butinsistedthatsuchfactscanafterallbeexplainedandpredictedbynaturalscience.(6)Jackson(1977)andothershavedefendedtheclaimthatinconsciousnesswearepresentedwithmentalindividualsthatthemselvesbearphenomenal,quali-tativeproperties.Forinstance,whenaredflashbulbgoesoffinyourface,yourvisualfieldexhibitsagreenblotch,an‘after-image’,athingthatisreallygreenandhasafairlydefiniteshapeandexistsforafewsecondsbeforedisappearing.Iftherearesuchthings,theyareentirelydifferentfromanythingphysicaltobefoundinthebrainofa(healthy)humansubject.Beliefinsuch‘phenome-nalindividuals’asgenuinelygreenafter-imageshasbeenunpopularamongphilosophersforsomeyears,butitcanbepowerfullymotivated(seeLycan1987:83–93).(7)Anumberofphilosophers,mostnotablySellars(1963),havestressedtheultra-smoothness,homogeneityorgrainlessnessofphenomenalfeels,andcon-tendedforthisreasonthatthosefeelscannotpeacefullybedissolvedintoametaphysicoflittlebruteparticlesandtheirerraticmotionthroughthevoid.(8)Ifhumanbeingsarefunctionally(eventeleologically)organizedsystemsofphysicalcomponentsandnothingmore,thentheir‘behaviour’isonlythemechanical,physicallydeterminedoutcomeofphysicalinputsandinternalenergytransformations.Theinputsarethemselvesonlythephysicalimpactsofenvironmentalcauses,whichcausesarethemselvestheresultsofeventscompletelyexternaltous.Wearemerelyautomata.(ItisnoaccidentthatPutnam’soriginalinspirationwastheTuringMachine.)Butweknowfromtheinsidethatthisisfalse.Ourconsciouschoicesandourdeliberateactionsareentirelyuptous;theyfeelentirelyfree.IfIsimplywishtoraisemyhand,thennothingcanstopmefromdoingsounlessquiteexternallyandbymakingnews(suchasamadmansuddenlypinningmetothefloororaPhantomJethurtlingthroughthewallofthebuilding).(9)Levine(1983,1993)hascomplainedofan‘explanatorygap’:thatnofunc-tionalisttheorycanexplainwhysuch-and-suchasensationfeelstoitssubject186\nPHILOSOPHYOFMINDinjustthewayitdoes.Thequestion,‘Butwhydosuch-and-suchfunctionalgoings-onconstituteorproduceasensationlikethis?’seemsalwaysopen.Somefunctionalistscontendthatafunctionalistpsychologycouldindeedexplainsuchfacts;otherstrytoshowwhyitneednot,consistentlywiththetruthoffunctionalism.Thisisaformidablearrayofobjections,and,onthefaceofit,eachisplausible.Materialistsandparticularlyfunctionalistsmustrespondindetail.Needlesstosay,materialistshaverespondedatlength;someofthemostpowerfulrejoindersareformulatedinLycan(1987,1996).Yetrecentyearshaveseensomereactionagainsttheprevailingmaterialism,includingare-emergenceofsomeneo-dualistviews,asinRobinson(1988),Hart(1988),Strawson(1994)andChalmers(1996).9ProblemswithIntentionalityTheproblemarisingfromourdiscussionofBrentanowastoexplainhowanypurelyphysicalentityorstatecouldhavethepropertyofbeingaboutor‘directedupon’anon-existentstateofaffairs.ThestandardfunctionalistreplyisthatpropositionalattitudeshaveBrentano’sfeaturebecausetheinternalphysicalstatesandeventsthatrealizethemrepresentactualorpossiblestatesofaffairs.Whattheyrepresent(theircontent)isdeterminedatleastinpartbytheirfunctionalroles.Therearetwomaindifficulties.Oneisthatofsayingexactlyhowaphysicalitem’srepresentationalcontentisdetermined;invirtueofwhatdoesaneurophysiologicalstaterepresentpreciselythattheRepublicancandidatewillwin?AnanswertothatgeneralquestioniswhatFodorhascalledapsychosemantics,andseveralattemptshavebeenmade.Theseconddifficultyisthatordinarypropositionalattitudecontentsdonotsuperveneonthestatesoftheirsubjects’nervoussystems,butareunderdeterminedbyeventhetotalstateofthatsubject’shead.Putnam’s(1975)TwinEarthandindex-icalexamplesshowthat,surprisingasitmayseem,twohumanbeingscouldbemole-cule-for-moleculealikeandstilldifferintheirbeliefsanddesires,dependingonvariousfactorsintheirspatialandhistoricalenvironments.Thuswecandistinguishbetween‘narrow’properties,thosethataredeterminedbyasubject’sintrinsicphysi-calcomposition,and‘wide’properties,thosethatarenotsodetermined.Representa-tionalcontentsarewide,yetfunctionalrolesare,ostensibly,narrow.How,then,canpropositionalattitudesbetype-identifiedwithfunctionalroles,orforthatmatterwithstatesofthebrainunderanynarrowdescription?Functionalistshaverespondedinoneoftwoways.Theformeristounderstand‘func-tion’widelyaswell,specifyingfunctionalroleshistoricallyand/orbyreferencetofea-turesofthesubject’sactualenvironment.Thelatterissimplytoabandonfunctionalismasanaccountofcontentinparticular,givingsomealternativepsychosemanticsforpropositionalattitudes,butpreservingfunctionalisminregardtoattitudetypes.(ThuswhatmakesastateadesirethatPisitsfunctionalrole,evenifsomethingelsemakesthestateadesirethatP).Theseoptionswillbeexploredbrieflyinsection10;andtherestofthischapterwillbedevotedtoproblemsraisedbythepropositionalattitudes.In187\nWILLIAMG.LYCANsection13weshallnotetheimpactofthe‘narrow’–‘wide’distinctionontheontol-ogicalstatusoffolkpsychologyitself.EmotionsInalludingtosensorystatesandtomentalstateswithintentionalcontent,wehavesaidnothingspecificallyabouttheemotions.Sincetherejectionofbehaviourism,theoriesofmindhavetendednottobeapplieddirectlytotheemotions;rather,theemotionshavebeengenerallythoughttobeconceptuallyanalysableascomplexesofmorecentralor‘core’mentalstates,typicallypropositionalattitudessuchasbeliefanddesire(andtheintentionalityofemotionshasaccordinglybeentracedbacktothatofattitudes).Armstrong(1968:ch.8,section3)tookessentiallythisline,asdoSolomon(1977)andGordon(1987).However,thereisaliteratureonfunctionalismandtheemotions;seeRey(1980)andsomeoftheotherpaperscollectedinRorty(1980).Griffiths(1997)takesagenerallyfunctionalistview,butarguesthat‘theemotions’donotconstituteasinglekind.SomepsychologicalliteratureonthetopiciscollectedinClarkandFiske(1982)andSchererandEkman(1984).10The‘LanguageofThought’HypothesisOurbasicproblemaboutintentionalityistoexplainhowapurelyphysicalsystemororganismcanbeinstateshavingsuchfeatures.Akeypointtonoteisthatintentionalorrepresentationalfeaturesaresemanticalfeatures:beliefsaretrue,orfalse;theyentailorimplyotherbeliefs;theyare(itseems)composedofconceptsanddependfortheirtruthonamatchbetweentheirinternalstructuresandthewaytheworldis;inparticulartheir‘aboutness’isverynaturallyregardedasamatterofmentalreferring.Somephilosophers,mostnotablySellars(1963),Fodor(1975,1981,1987)andField(1978),havetakenthesemanticnatureofbeliefsasastrongcluetothenatureofintentionalityitself,suggestingthatbeliefsandthoughtshavetheirintentionalityinvirtueofpropertiestheysharewithothersemanticallycharacterizeditems,thesentencesofpublicnaturallanguagessuchasEnglishandChinese.Sellarsargued(againstthebehaviourists)thatpeople’sintentionalstatesareindeedinnerandareindeedrepresentations.Theyarephysicalstatesofthecentralnervoussystem.Nonetheless(contraBrentano)theyarephysicalstatesthathavesemanticalprop-erties.Theyhavethosepropertiesinvirtueofthefunctionalrolestheyplayintheirowners’behaviouraleconomies,closelyanalogoustotheinferentialrolesthatcorre-spondinglinguistictokensplayinpublic,entirelyphysicallanguagegames.Toputthethesisslightlymoreformally:forasubjectStothinkor‘occurrentlybelieve’thatPisfortheretobeastateofS’scentralnervoussystemthatbearsthesemanticcontentthatP;thestatebearsthatcontentinmuchthesamesenseandinmuchthesamewaythatasentenceofEnglishoranothernaturallanguagemeansthatP.Letuscallthistherepresentationaltheoryofthinking.Itistemptingtoglosstherepresentationaltheorybyspeakingofalanguageofthought,anditsleadingproponentshavegivenintothattemptation.Fodorarguesthatrepresentationandtheinferentialmanipulationofrepresentationsrequireamedium188\nPHILOSOPHYOFMINDofrepresentation,nolessinhumansubjectsthanincomputers.Computersemploymachinelanguagesofvariouskinds;itisreasonabletopositoneormorehuman‘machinelanguages’inwhichhumanthoughtandcognitiontakeplace.Ontheotherhand,thereareobviousdisanalogiesbetweenprivatethoughtandpublicspeech,soifwearetotakerepresentationalismseriouslywemustspecify,inatleastapreliminaryway,whatsimilaritiesarebeingclaimed.Ontherepresentationalist’sbehalf,letussaythatphysicallyrealizedthoughtsandmentalrepresentationsare‘linguistic’inthefollowingsense:(1)theyarecomposedofpartsandaresyntacticallystructured;(2)theirsimplestpartsrefertoordenotethingsandpropertiesintheworld;(3)theirmeaningsaswholesaredeterminedbytheseman-ticalpropertiesoftheirbasicpartstogetherwiththegrammaticalrulesthathavegen-eratedtheiroverallsyntacticstructures;(4)theyhavetruth-conditions(thatis,putativestatesofaffairsintheworldthatwouldmakethemtrue),andaccordinglytheyaretrueorfalsedependingonthewaytheworldhappensactuallytobe;(5)theybearlogicalrelationsofentailmentorimplicationtoeachother.Thus,accordingtotherepresen-tationaltheory:humanbeingshavesystemsofphysicalstatesthatserveastheele-mentsofalexiconorvocabulary,andhumanbeings(somehow)physicallyrealizerulesthatcombinestringsofthoseelementsintoconfigurationshavingthecomplexrepre-sentationalcontentsthatcommonsenseassociateswiththepropositionalattitudes.AndthatiswhythoughtsandbeliefsaretrueorfalsejustasEnglishsentencesare,thougha‘languageofthought’(Mentalese,orBrainese)maydiffersharplyinitsgrammarfromanynaturallanguage.Theargumentsfortherepresentationaltheorytakeanumberofimpressivelydif-ferentforms(Fodor1981,1987;DevittandSterelny1987;Lycan1988).Thoughtheyareformidable,thetheoryhasalsocomeinforagooddealofcriticism,andtheretootheargumentstakeanumberofimpressivelydifferentforms.Someleadingobjectionsarethese:(1)Chisholm(1972)andothershavepointedoutthatthemeaningsofnatural-languagesentencesareconventional,andsodependonthebeliefsandinten-tionsofhumanspeakers.Beliefsandintentionsarepropositionalattitudes.How,then,withoutcircularityorregress,canattitudecontentbeexplicatedintermsofmeaninginthepublic-linguisticsense?(2)Dennett(1978)arguesonseveralgroundsthattheideaof‘sentencesinthehead’,implementedasinscriptionsscrawledinbrainchalkuponabrainblack-board,isfanciful,nottosaygrotesque.(Thoughhistargetseemstobeastrongerandmoreoutlandishversionofrepresentationalismthantheonesketchedabove.)(3)Dennettfurtherarguesthattacitpropositionalattitudes,suchasyouruncon-sideredbeliefthatIamlessthan12feettall,canhardlybeinternalrepresen-tations,becausetherearefartoomanyofthem.YoualsobelievethatIamlessthan12-plus-nfeettallforanyrealnumbern.(4)ChurchlandandChurchland(1983)(cf.Churchland1989)contendthatthe‘languageofthought’ideaisdistinctlyunbiological.Whenonerecallsthathumanbeingsarecard-carryingmembersoftheanimalkingdomandthatwehaveevolvedintheusualwaybynaturalselection,ourlinguisticabilities,and189\nWILLIAMG.LYCANourcognitivefunctionsonanyhighlylinguisticizedaccountofthem,seemtobeanevolutionaryafterthoughtatbest,andatinyfragmentofthepsychologythatactuallygetsusaroundintheworld.P.S.Churchland(1986)andP.M.Churchland(1986,1989)compellinglydepictabrainthatworksbyconnec-tionistnetworking(seesection12below)andbyphysicallyhardwiredco-ordinatetransformation,notbydigital-computer-likeinferentialcomputationoversyntacticallystructuredsentencesorlogicalformulas(seealsothecomplaintsraisedbyDreyfus(1979)againstclassicalAI).(5)Whilepubliclanguageis(again)conventionalineachofseveralways,thereisobviouslynothingsocialorconventionalabouttheworkingsofthebrain.The‘reference’oftheallegedlanguage-of-thought’svocabularyitemsmustbenatural.TheEnglishword‘dog’isanarbitraryvocable,whichissociallyattachedbytheEnglish-speakingcommunitytodogs,buttheMentalesewordfor‘dog’mustsomehowbenaturallyconnectedtodogs,withouthumaninter-vention.Thatishardtoswallow.(6)Thesearchforanadequatepsychosemanticsisdebilitatingatbest.IfthoughtsandbeliefscanbeaboutMargaretThatcheroraboutSantaClausbecausetheneurophysiologicalstatesthatrealizethemsomehowsemanticallyrefertoMargaretThatcherorSantaClaus,andiftheneurophysiologicalstatesdotheirreferringinvirtueofsomephysical,functionalorotherwisenaturalisticprop-ertytheyhave,whatisthatproperty?WemayimaginethatourthoughtsofThatcherstandinsomehistoricalrelationtoThatcherherself,butourthoughtsofSantaClausdonotstandinanyhistoricalrelationtoSantaClaushimself,forhedoesnotexist.Nor,intryingtosaywhatitisinvirtueofwhichsomeneurophysiologicalstate‘refersto’anything,mayweinvokeunexplicatedpropositionalattitudesorrepresentationalcontent,andthatprovestobeabitingconstraint.Psychosemanticsadmitsof(sofar)twobasiclinesofapproach,whicharereconcilablethoughentirelydistinctinorigin.Thefirstisthecausal–historicalapproach,inthespiritofKripke(1972)andPutnam(1975),accordingtowhichamental–brainitemMreferstoathingXjustincaseXfiguresappropriatelyinM’scausalhistoryoraetiology.Practitionersofthisapproachcashtheimportantqualifier‘appropriately’inanynumberofhopefulways;anysuc-cessfulwaywillhavetoaccountforreferencetonon-existentthings,whichisnosmalltaskinitselfandmultiplyhardgiventhatonemaynot,onpainofcircularity,invokeunexplicatedpropositional-attitudecontents.Thesecondapproachisteleological,asmentionedinsection7.Theteleologicalapproachbeganlifeas‘indicatorsemantics’,theideathatMrepresentsXjustincaseMisareliableindicatorofX(seeDretske1981;Stalnaker1984);butinordertoaccountforfalserepresentations,thatideawassoonteleologized.Themainhurdleforteleosemanticsistosecurefullgenerality:humanmentalstatescanbeaboutanything,butsofarastheexternalworldisconcerned,noindividualbrainstatecanhavemorethanafewpsychobiologicalfunctionsatatime;howcanneurophysiologicalstatesbeaboutanythingbutfood,shelter,predatorsandsexualpartners?Fodor(1990a,althoughwrittenearlier)wasthefirstteleologicaltheoristtosolvethatproblem–190\nPHILOSOPHYOFMINDhoweverquicklyandhoweverinadequately.(Hehassincevigorouslyrepudiatedhisoriginalsolutioninabookthat,sardonically,hastheidenticaltitle(Fodor1987,andin1990b,1994);hehasmovedtoavariationonindicatorsemanticsthathecallsthe‘asymmetricdependence’theory(Fodor1994).)ItisnotclearwhetherFodor’s(1990a)solutioncanbeextendedtothevastrangeofbeliefsandthoughtsthatareaboutthingsotherthanphysicallypresentenvironmentalobjects:mathematics,literature,philoso-phy,religion.Theteleologicalviewgivesnohintastohowsuchtopicsarejustasreadyobjectsofthoughtasarenearbycows,cats,desktopcomputersandbuildings.11InstrumentalismTheidentitytheoristsandthefunctionalists(machineorteleological)joinedcommonsenseandcurrentcognitivepsychologyinunderstandingmentalstatesandeventsbothasinternaltohumansubjectsandascauses.Beliefsanddesiresinparticulararethoughttobecausedbyperceptualorothercognitiveeventsandasinturnconspiringfromwithintocausebehaviour.IfArmstrong’sorLewis’stheoryofmindiscorrect,thisideaisnotonlycommon-sensicalbutaconceptualtruth;iffunctionalismiscorrect,itisatleastametaphysicalfact.Inrallyingtotheinner-causalstory,aswesawinsection3,theidentitytheoristsandfunctionalistsbrokewiththebehaviourists,forbehaviouristsdidnotthinkofmentalitemsasentities,asinner,orascausesinanystrongersensethanthebarehypotheti-cal.Behaviouristseitherdispensedwiththementalisticidiomaltogether,orpara-phrasedmentalascriptionsintermsofputativeresponsestohypotheticalstimuli.Morerecently,otherphilosophershavefollowedtheminrejectingtheideaofbeliefsanddesiresasinnercausesandinconstruingtheminamorepurelyoperationalorinstru-mentalfashion.D.C.Dennett(1978,1987)hasbeenparticularlyconcernedtodenythatbeliefsanddesiresarecausallyactiveinnerstatesofpeople,andmaintainsinsteadthatbelief-ascriptionsanddesire-ascriptionsaremerelycalculationaldevices,thathappentohavepredictiveusefulnessforareasonthathegoesontoexplain.Suchascriptionsareoftenobjectivelytrue,hegrants,butnotinvirtueofdescribinginnermechanisms.ThusDennettisanINSTRUMENTALIST(pp.294–5)aboutpropositionalattitudessuchasbeliefanddesire.(Accordingtoacontemporaryinterpretation,an‘instrumentalist’aboutXsisatheoristwhoclaimsthatalthoughsentencesabout‘Xs’areoftentrue,theydonotreallydescribeentitiesofaspecialkind,butonlyservetosystematizemorefamil-iarphenomena.Forinstance,weareallinstrumentalistsabout‘theaverageAmericanhomeowner’,whoiswhite,maleandthefatherofexactly1.9children.)Toascribea‘belief’ora‘desire’isnottodescribesomesegmentofphysicalreality,Dennettsays,butismorelikemovingagroupofbeadsinanabacus.(ItshouldbenotedthatDennetthasmorerecentlymoderatedhisline:seeDennett1991.)Dennettoffersbasicallyfivegroundsforhisrejectionofthecommon-sensicalinner-causethesis.(1)Hethinksitquiteunlikelythatanysciencewilleverturnupanydistinctiveinner-causalmechanismthatwouldbesharedbyallthepossiblesubjectsthathadaparticularbelief.(2)Heoffersnumerousobjectionsto‘languageofthoughtpsychology’,whichisthemostpopularinner-causetheory.(3)Hecomparesthe191\nWILLIAMG.LYCANbelief–desireinterpretationofhumanbeingstothatofloweranimals,chess-playingcomputersandevenlightning-rods,arguingthat(3a)intheircasewehavenoreasontothinkofbelief-ascriptionsanddesire-ascriptionsasotherthanmerecalcula-tional–predictivedevicesand(3b)wehavenomorereasonforthecaseofhumanstothinkofbelief-ascriptionsanddesire-ascriptionsasotherthanthat.(4)Dennettarguesfromtheverificationconditionsofbelief-ascriptionsanddesire-ascriptions–basicallyamatterofextrapolatingrationallyfromwhatasubjectoughttobelieveandwantinhisorhercircumstances–andthenheboldlyjustidentifiesthetruth-makersofthoseascriptionswiththeirverification-conditions,challenginginner-causetheoriststoshowwhyinstrumentalismdoesnotaccommodatealltheactualevidence.(5)Hearguesthatinanycaseifapurelynormativeassumption(the‘rationalityassumption’,whichisthatpeoplewillgenerallybelievewhattheyoughttobelieveanddesirewhattheyshoulddesire)isrequiredforthelicensingofanascription,thentheascriptioncannotitselfbeapurelyfactualdescriptionofaplainstateofaffairs.Stich(1981)exploresandcriticizesDennett’sinstrumentalismatlength(perhapsoddly,Stich(1983)goesontodefendaviewnearlyasdeprecatingasDennett’s,thoughclearlydistinctfromit).Dennett(1981)respondstoStich,bringingoutmoreclearlytheforceofthe‘rationalityassumption’assumption.(OthercriticismsarelevelledagainstDennettbycommentatorsintheBehavioralandBrainSciencessymposiumthatisheadedbyDennett(1988).)AclosecousinofDennett’sview,inthatitfocusesontherationalityassumption,isDonaldDavidson’s(1970)anomalousmonism.UnlikeDennett’sinstrumentalism,itendorsestokenphysicalismandinsiststhatindividualmentaltokensarecauses,butitrejectsonsimilarlyepistemologicalgroundsthepossibilityofanyinterestingmaterial-istictype-reductionofthepropositionalattitudes.12EliminativismandNeurophilosophyDennett’sinstrumentalismbreaksfairlyradicallywithcommonsenseandwithphilo-sophicaltraditionindenyingthatpropositionalattitudessuchasbeliefanddesirearerealinnercausalstatesofpeople.ButDennettconcedes–indeed,heurgentlyinsists–thatbelief-ascriptionsanddesire-ascriptionsaretrue,andobjectivelytrue,nonetheless.Otherphilosophershavetakenalessconciliatory,stillmoreradicallyuncommon-sensicalview:thatmentalascriptionsarenottrueafterall,butaresimplyfalse.Commonsenseisjustmistakeninsupposingthatpeoplebelieveanddesirethings,andperhapsinsupposingthatpeoplehavesensationsandfeelings,disconcertingasthatnihilisticclaimmayseem.Followingstandardusage,letuscallthenihilisticclaim‘eliminativematerialism’,or‘eliminativism’forshort.Itisimportanttonoteacustomaryifunexpectedalliancebetweentheeliminativistandthetokenphysicalist:theeliminativist,theidentitytheoristandthefunctionalistallagreethatmentalitemsare,ifanything,realinnercausalstatesofpeople.Theydisagreeonlyontheempiricalquestionofwhetheranyrealneurophysiologicalstatesofpeopledoinfactanswertothecommon-sensicalmentalcategoriesof‘folkpsychology’.Eliminativistspraiseidentitytheoristsandfunc-tionalistsfortheirforthrightwillingnesstostepupandtaketheirempiricalshot.Both192\nPHILOSOPHYOFMINDeliminativistsandtokenphysicalistsscorntheinstrumentalist’ssleazyevasion.(Buteliminativistsagreewithinstrumentaliststhatfunctionalismisapipe-dream,andfunctionalistsagreewithinstrumentaliststhatmentalascriptionsareoftentrueandobviouslyso.ThethreeviewsformanEternalTriangleofanotuncommonsort.)PaulFeyerabend(1963a,1963b)wasthefirsttoargueopenlythatthementalcategoriesoffolkpsychologysimplyfailtocaptureanythinginphysicalrealityandthateverydaymentalascriptionswerethereforefalse.(Rorty(1965)tookanotoriouslyeliminativistlinealso,but,followingSellars(1963),triedtosoftenitsnihilism;LycanandPappas(1972)arguedthatthesofteningservedonlytocollapseRorty’spositionintoincoherence.)Feyerabendattractednogreatfollowing,presumablybecauseofhisview’soutrageousfloutingofcommonsense.ButeliminativismwasresurrectedbyPaulChurchland(1981)andothers,anddefendedinmoredetail.Churchlandarguesmainlyfromthepovertyof‘folkpsychology’;heclaimsthathistorically,whenotherprimitivetheoriessuchasalchemyhavedoneasbadlyonscientificgroundsasfolkpsychologyhas,theyhavebeenabandonedandrightlyso.P.S.Churchland(1986)andChurchlandandSejnowski(1990)emphasizethecompar-ativescientificrealityandcausalefficacyofneurobiologicalmechanisms:giventhescientificexcellenceofneurophysiologicalexplanationandthecontrastingdiffusenessandtype-irreducibilityoffolkpsychology,whyshouldwesuppose–evenforaminute,muchlessautomatically–thattheplatitudesoffolkpsychologyexpresstruths?P.S.Churchland’sintenseinterestinneuroscienceandherdistrustofthecategoriesoffolkpsychologyarematchedwithinasectoroftheAIcommunity,inanequallyintenseupsurgeof‘neuralmodelling’.Inparticular,whatiscalled‘connectionism’orParallelDistributedProcessing(‘PDP’forshort;McClellandetal.1986)hasarousedmuchcurrentinterestasanAIresearchprogrammethatdivergesfromthestandarddeploymentof‘rulesandrepresentations’(asoutlinedinsection10above)andfromtheideaoflinearormonotonictheorem-provingfromapre-loadeddatabase.PDPemploys(althoughinpracticeitonlysimulates)anarrayof‘units’,eachbeingcon-nectedbyligaturestootherunitsandeachhavingan‘activationpotential’thatisdirectlyaffectedbythepotentialsofadjoiningunits;theobvious,andintended,allu-sionistothebrain’sneuralnets.Someunitsaredesignatedasinputs,othersasoutputs;therestare‘hidden’,andmysteriouslyregulateoutputwhentheyaregiveninputaccordingtovariousalgorithms.AmajorfocusofPDPresearchisonlearningovertime;connectionistnetworksaregoodatlearningpatternrecognitiontasks.ConnectionismisanengineeringapproachwithinAI.Butthetermhascaughtonamongbothpsychologistsandphilosophers,andisnowoftenusedneologistically,asnamingeither(1)apsychologicaltheory,roughlythatsuch-and-suchbehaviouralcapacitiesareexplainedbyconnectionistarchitectureactuallyrealizedinorganisms’brains;or(2)aphilosophicalcontentionreminiscentofRyle(1949),roughlythatintel-ligenthumancapacities,includingthinkingandrationality,aresomehowholisticallyemergentfromconnectionistarchitectureinthebrainratherthanbeingamatterofthemanipulationofinternalbeliefsorotherrepresentationsaccordingtorules.Somephilosopherstakeconnectionisminoneoranotherofitsseveralsensestorefute–oratleastembarrass–theideathathumancognitionisamatterofhostinginternalrepresentationalstatessuchasbeliefsanddesires;seeparticularlyChurchland(1989)andRamsey,StichandGaron(1991).But,logicallyspeaking,connectionism193\nWILLIAMG.LYCANseemsentirelycompatiblewithrepresentationalism(Smolensky1988;Bechtel1987),andarguablyitisaninstanceofrepresentationalism(FodorandPylyshyn1988;Lycan1991).Thematterneedsconsiderablefurtherexamination;seealsoBechtel(1992).13TheStatusof‘FolkPsychology’Intheprevioussectionwewereintroducedtothesomewhatstrangeideathatmentaltermsandmentalentitiesthemselvesareatrisk.Theideathatmentaltermsarethetheoreticaltermsofafolkscience–the‘theory’theory,asMorton(1980)callsit–wasfirsturgedbySellars,FeyerabendandQuineinthe1960s.Afteranunderstandableperiodofincredulity,the‘theory’theorybecamewidelyaccepted.Butaswehaveseen,the‘theory’theoryleadsinadisconcertingdirection.Iftheonlyreasonwehaveforacceptingtheexistenceofmentalentitiesistheutilityandpresumedtruthoffolkpsychology,andiffolkpsychologyshouldturnouttobelargelyfalseorseriouslyinfirminsomeotherway(asscientifictheoriesandespeciallyfolktheoriesoftendo),thenpresumablysomeversionofeliminativismiscorrect.Wehavealreadylookedatsomeeliminativistarguments.Inthissectionweshallturntoaspecialissueregardingpropositionalattitudecontent–theissueofmethod-ologicalsolipsisminPutnam’s(1975)phrase–andexamineitsconsequencesforthecogencyoffolkpsychology.Suppose,forthesakeofargument,thatweattributerepresentationalcontenttotheinternalstatesofcomputers–asinreallifewedo,howeveranthropomorphically.Acomputersometimesthinksthisorwantsthat;attheveryleast,itcomputesthisorcomputesthat,suchastheGNPofMonaco,theoutcomeofapresidentialelectionorthebalanceofourbankaccount.Now,totakeakeyexampleofFodor’s(1980),itisquitepossiblethattwocomput-ers,programmedbyentirelydifferentusersforentirelydifferentpurposes,shouldhappentorunphysicallyinparallel.Theymightgothroughpreciselythesamesequenceofelectricalcurrentsandflipflopsettingsandyethavetheiroutputsinter-preteddifferentlybytheirrespectiveusers,especiallyifwhattheywritetotheirscreensisallinnumericalform.Oneofthemwouldnaturallyandcorrectlybedescribedasfig-uringouttheGNPofMonaco,whiletheotherwouldjustasnaturallyandcorrectlybedescribedasfiguringoutthebattingaveragesoftheNewYorkYankees.Thepoint,plainenoughwhenwethinkaboutit,issimplythatwhatamachineiscomputingisnotfullydeterminedbythephysicaloreventheabstract-functionaloperationsthataregoingonentirelyinsidethemachine.Whatthemachineiscomputingdependstosomeextentonsomethingoutsidethemachineitself–users’intentions,causal-historicalchains(seesections9and10above),teleology,interpretationbyobserversorjusttheconvenienceofthebeholder.Thepointisnotparticularlysurprising.ButPutnam(1975)drewabroadercon-clusion,thatwasdevelopedtostartlingeffectbyFodor(1975)andbyStich(1983):asitiswithcomputers,soitiswithhumans.Therepresentationalcontentofahumansubject’spropositionalattitudesisunderdeterminedbyeventhetotalstateinsidethatsubject’shead.Aswenotedinsection9above,intentionalpropertiesarenotdeterminedbytheirsubject’sintrinsicmolecularconstitution.194\nPHILOSOPHYOFMINDThefactthatattitudecontentsare‘wide’ratherthan‘narrow’raisesseriousprob-lemsaboutthevauntedroleofpropositionalattitudesintheexplainingofbehaviour,andattendantmethodologicalquestionsforpsychology(Fodor1980;Stich1983).Butforourpurposesinthissection,themainquestionisthis:iftherepresentationalcontentofapropositionalattitudedependsonfactorsoutsidethephysicalboundaryofitsowner’sskin,whatarethosefactors,andmoreimportantly,canwestillsupposethattheattitudecontentsaregenuinepropertiesoftheowner-subject?Thespectreofeliminationreappears;perhapsitisnotreally,objectivelytrueofpeoplethattheybelievethisordesirethat.Thequestionofwhattheenvironmentalfactorsareisjustthequestionofpsy-chosemanticsagain.Butwhatabouttheontologicalstatusofpeople’sbeliefsanddesiresthemselves?Thereareseveraldifferentpossiblepositions:(1)Theexternalsemanticalinterpretationofanorganism’sinternalphysical–functionalstatesisentirelyupforgrabs:anyinterpretationthatsuitsanyone’sconvenienceisgoodenough,andiftwointerpreters’interpretationsconflict,neitheriscorrecttotheexclusionoftheother.(Schiffer(1981)dis-cussesthispossibilitysympathetically,andconstruesQuine’s(1960)famousdoctrineofthe‘indeterminacyoftranslation’asgettingatthisposition.)Ifwefallinwiththisview,wecanhardlycallitahardfactthatasubjectbelievesonethingratherthananother.(2)Thesemanticalinterpretationofbeliefsisnotupforgrabs;itisatleastlooselydeterminedbyvariouscontextualfactors,andsomeinterpretationsarecorrectwhileothersarejustwrong.Butthecontextualfactorsinquestionareintol-erablyvagueandmessy,andsocialandinterest-infested,makingthemquiteunsuitableforincorporationintoanygenuinescience.ToascribethebeliefthatPtosomeoneisnomorescientific,explanatoryorusefulthaniscallingsomething‘nice’.(2)isthepositionofStich(1983).(3)Thesemanticalinterpretationofbeliefsislooselydeterminedbyvariouscontextualfactors,andsomeinterpretationsarecorrectwhileothersarejustwrong,andthisiswhatitisforthesubjecttobelieveonethingratherthananother.Somecomplexcausal-historicaland/orteleologicalfeatureofthesubject’senvironmentmakesitobjectivelytruethatthesubjectbelievesso-and-soratherthansuch-and-such,whetherornotthefeatureisscientificallyinterestingorwell-behaved.(Lycan(1988)defendsthisview.)(4)Thereisnothingmessy,interest-relativeorwhatever,aboutthecontextualfactorsinquestion,eventhoughtheyextendoutsidetheboundaryoftheskin.Theyareamatterofsimplenomologicalorteleologicalfact.Thispositionwasendorsedby‘indicator’semanticists,amongothers.(1)isessentiallyaneliminativeview,andcertainlyfloutsthefolk-psychologicalthesisthatpropositionalattitudesquapropositionalattitudesarerealcausalcon-stituentsoftheworld.(2)allowsthatattitudeascriptionsmaybetrue(howeverinterest-infested),sothatpeopledobelieveonethingratherthananother,butstillrejectsthefolk-psychologicalviewthattheattitudesgenuinelycausebehaviour.(3)and(4)areentirelycompatiblewithfolkpsychology.(2)and(3)differfromeach195\nWILLIAMG.LYCANother,itseems,onlyindegree,sotherealissuesofaristhatofwhether(2)or(3)ismoreplausible.Thereisafurtheranddeeperissue.Evenifwegrantthat(2)istrueoreventhat(1)istrue,whyshouldwecareaboutwidepropertiesatall?Whatexplanatoryneeddotheyfulfil?Surelyitisonlynarrowpropertiesthatfigureinthecausalexplanationofbehav-iour.Amongothers,HorganandWoodward(1985),Burge(1986)andFodor(1987)trytoanswerthatdeeperquestion;Devitt(1990)investigatesfurther.Onemayofcoursedisputethe‘theory’theoryitselfasathesisaboutmentalconcepts.Ryle(1949)andWittgenstein(1953)staunchlyopposeditbeforeithadexplicitlybeenformulated.MorerecentcriticsincludeMorton(1980),Malcolm(1984),Baker(1988),McDonough(1991)andWilkes(1993).FurtherReadingBeakleyandLudlow(1992)offerclassicalandcontemporaryreadingsonthephilosophyofmind.Block(1980)containsclassicworksonthemind–bodyprobleminvol.1,whilevol.2takesupspecialtopics,includingtheimagerydebate,thephilosophyoflinguisticsandinnateideas.Campbell(1984)providesafineintroductiontothemind–bodyproblem.Churchland(1984)introducesthemind–bodyproblemfromtheneurophilosophicalpointofview.Fodor(1968b)isafoundingdocumentoffunctionalism.Haugeland(1981)containsimportantpapersfromthe1970sonArtificialIntelligence.Lycan(1990,1999)isthesuccessortoBlock(1980)andofferssomegeneralwritingsonthemind–bodyproblemandsomemorespecializedreadingsinthephilosophyofcognitivescience.Rosenthal(1991)containsclassicalandcontemporaryreadingsonmindandcognitivescience.ReferencesArmstrong,D.M.1968:AMaterialistTheoryoftheMind.London:RoutledgeandKeganPaul.Baker,L.R.1988:SavingBelief.Princeton,NJ:PrincetonUniversityPress.Beakley,B.andLudlow,P.(eds)1992:ThePhilosophyofMind.Cambridge,MA:BradfordBooks/MITPress.Bechtel,W.1987:ConnectionismandthePhilosophyofMind:AnOverview.SouthernJournalofPhilosophy,26(supplement),17–41.ReprintedinW.Lycan(ed.)1990:MindandCognition:AReader.Oxford:Blackwell.——1992:TheCaseforConnectionism.PhilosophicalStudies,71,119–54.ReprintedinW.Lycan(ed.)1999:MindandCognition:AnAnthology.Oxford:Blackwell.Bigelow,J.andPargetter,R.1987:Functions.JournalofPhilosophy,84,181–96.Block,N.J.1978:TroubleswithFunctionalism.InW.Savage(ed.)MinnesotaStudiesinthePhilosophyofScience,Vol.10:PerceptionandCognition.Minneapolis:UniversityofMinnesotaPress,261–325.ExcerptsreprintedinLycan(1990,1999).——(ed.)1980:ReadingsinPhilosophyofPsychology,2vols.Cambridge,MA:HarvardUniversityPress.——1981:PsychologismandBehaviorism.PhilosophicalReview,90,5–43.Block,N.J.andFodor,J.A.1972:WhatPsychologicalStatesareNot.PhilosophicalReview,81,159–81.ReprintedinBlock(ed.)1980:ReadingsinPhilosophyofPsychology,2vols.Cambridge,MA:HarvardUniversityPress.Boorse,C.1976:WrightonFunctions.PhilosophicalReview,85,70–86.196\nPHILOSOPHYOFMINDBrentano,F.1973[1874]:PhilosophyfromanEmpiricalStandpoint.London:RoutledgeandKeganPaul.Burge,T.1986:IndividualismandPsychology.PhilosophicalReview,95,3–45.Campbell,K.1984:BodyandMind,2ndedn.NotreDame,IN:UniversityofNotreDamePress.Carnap,R.1932–3:PsychologyinPhysicalLanguage.Erkenntnis,3,107–42.ExcerptreprintedinW.Lycan(ed.)1990:MindandCognition:AReader.Oxford:Blackwell.Chalmers,D.1996:TheConsciousMind.Oxford:OxfordUniversityPress.Chisholm,R.M.1957:Perceiving.Ithaca,NY:CornellUniversityPress.——1972:Contributionsto‘TheChisholm–SellarsCorrespondenceonIntentionality’.InA.Marras(ed.)Intentionality,Mind,andLanguage.Urbana:UniversityofIllinoisPress.Chomsky,N.1959:ReviewofB.F.Skinner.[SeeSkinner1957.]Language,35,26–57.Churchland,P.M.1981:EliminativeMaterialismandthePropositionalAttitudes.JournalofPhilosophy,78,67–90.ReprintedinChurchland(1989)andinLycan(1990,1999).——1984:MatterandConsciousness.Cambridge,MA:BradfordBooks/MITPress.——1986:SomeReductiveStrategiesinCognitiveNeurobiology.Mind,95,279–309.——1989:ANeurocomputationalPerspective.Cambridge,MA:BradfordBooks/MITPress.Churchland,P.S.1986:Neurophilosophy.Cambridge,MA:BradfordBooks/MITPress.Churchland,P.S.andChurchland,P.M.1983:StalkingtheWildEpistemicEngine.Noûs,17,5–18.Churchland,P.S.andSejnowski,T.1990:NeuralRepresentationandNeuralComputation.InW.Lycan(ed.)1990:MindandCognition:AReader.Oxford:Blackwell.ReprintedinLycan(1999).Clark,M.S.andFiske,S.T.1982:AffectandCognition.Hillsdale,NJ:LawrenceErlbaumAssociates.Cummins,R.1975:FunctionalAnalysis.JournalofPhilosophy,72,741–64.ReprintedinN.J.Block(ed.)1980:ReadingsinPhilosophyofPsychology,2vols.Cambridge,MA:HarvardUniversityPress.——1983:TheNatureofPsychologicalExplanation.Cambridge,MA:MITPress/BradfordBooks.Davidson,D.1970:MentalEvents.InL.FosterandJ.W.Swanson(eds)ExperienceandTheory.Amherst:UniversityofMassachusettsPress.ReprintedinN.J.Block(ed.)1980:ReadingsinPhilosophyofPsychology,2vols.Cambridge,MA:HarvardUniversityPress.Dennett,D.C.1978:Brainstorms.Montgomery,VT:BradfordBooks.——1981:MakingSenseofOurselves.PhilosophicalTopics,12,63–81.ReprintedinW.Lycan(ed.)1990:MindandCognition:AReader.Oxford:Blackwell.——1987:TheIntentionalStance.Cambridge,MA:BradfordBooks/MITPress.——1988:PrécisofTheIntentionalStance.BehavioralandBrainSciences,11,495–505.——1991:RealPatterns.JournalofPhilosophy,88,27–51.Devitt,M.1990:ANarrowRepresentationalTheoryoftheMind.InW.Lycan(ed.)1990:MindandCognition:AReader.Oxford:Blackwell.ReprintedinW.Lycan(ed.)1999:MindandCognition:AnAnthology.Oxford:Blackwell.Devitt,M.andSterelny,K.1987:LanguageandReality.Cambridge,MA:BradfordBooks/MITPress.Dretske,F.1981:KnowledgeandtheFlowofInformation.Cambridge,MA:BradfordBooks/MITPress.——1988:ExplainingBehavior.Cambridge,MA:BradfordBooks/MITPress.Dreyfus,H.L.1979:WhatComputersCan’tDo,2ndedn.NewYork:HarperandRow.Dreyfus,H.L.andDreyfus,S.1986:MindOverMachine.NewYork:FreePress,Macmillan.Farrell,B.A.1950:Experience.Mind,50,170–98.197\nWILLIAMG.LYCANFeyerabend,P.1963a:MaterialismandtheMind–BodyProblem.ReviewofMetaphysics,17,49–66.——1963b:MentalEventsandtheBrain.JournalofPhilosophy,60,295–6.Field,H.1978:MentalRepresentation.Erkenntnis,13,9–61.Fodor,J.A.1968a:TheAppealtoTacitKnowledgeinPsychologicalExplanation.JournalofPhilosophy,65,627–40.——1968b:PsychologicalExplanation.NewYork:RandomHouse.——1975:TheLanguageofThought.NewYork:ThomasY.Crowell.——1980:MethodologicalSolipsismConsideredasaResearchStrategyinCognitivePsychology.BehavioralandBrainSciences,3,63–73.——1981:Representations.Cambridge,MA:BradfordBooks/MITPress.——1987:Psychosemantics.Cambridge,MA:BradfordBooks/MITPress.——1990a:Psychosemantics.InW.Lycan(ed.)1990:MindandCognition:AReader.Oxford:Blackwell.——1990b:ATheoryofContent.Cambridge,MA:BradfordBooks/MITPress.——1994:TheElmandtheExpert.Cambridge,MA:BradfordBooks/MITPress.Fodor,J.A.andPylyshyn,Z.W.1988:ConnectionismandCognitiveArchitecture:ACriticalAnalysis.Cognition,28,3–71.Geach,P.1957:MentalActs.London:RoutledgeandKeganPaul.Gordon,R.M.1987:TheStructureofEmotions.Cambridge:CambridgeUniversityPress.Griffiths,P.1997:WhatEmotionsReallyAre.Chicago:UniversityofChicagoPress.Gunderson,K.1970:AsymmetriesandMind–BodyPerplexities.InM.RadnerandS.Winokur(eds)MinnesotaStudiesinthePhilosophyofScience,Vol.4:AnalysesofTheoriesandMethodsofPhysicsandPsychology.Minneapolis:UniversityofMinnesotaPress.——1974:TheTextureofMentality.InR.Bambrough(ed.)Wisdom:TwelveEssays.Oxford:OxfordUniversityPress.Hart,W.D.1988:EnginesoftheSoul.Cambridge:CambridgeUniversityPress.Haugeland,J.1979:UnderstandingNaturalLanguage.JournalofPhilosophy,76,619–32.ReprintedinW.Lycan(ed.)1990:MindandCognition:AReader.Oxford:Blackwell.——(ed.)1981:MindDesign.Cambridge,MA:BradfordBooks/MITPress.Horgan,T.andWoodward,J.1985:FolkPsychologyisHeretoStay.PhilosophicalReview,94,197–226.ReprintedinLycan(1990,1999).Jackson,F.1977:Perception.Cambridge:CambridgeUniversityPress.——1982:EpiphenomenalQualia.PhilosophicalQuarterly,32,127–36.ReprintedinLycan(1990,1999).Kirk,R.1974:Zombiesvs.Materialists.AristotelianSocietySupplementaryVolume,48,135–52.Kripke,S.1972:NamingandNecessity.InD.DavidsonandG.Harman(eds)SemanticsofNaturalLanguage.Dordrecht:D.Reidel.Levin,M.1979:MetaphysicsandtheMind–BodyProblem.Oxford:OxfordUniversityPress.Levine,J.1983:MaterialismandQualia:TheExplanatoryGap.PacificPhilosophicalQuarterly,64,354–61.——1993:OnLeavingOutWhatIt’sLike.InM.DaviesandG.Humphreys(eds)Consciousness.Oxford:Blackwell.Lewis,D.1966:AnArgumentfortheIdentityTheory.JournalofPhilosophy,63,17–25.——1972:PsychophysicalandTheoreticalIdentifications.AustralasianJournalofPhilosophy,50,249–58.ReprintedinN.J.Block(ed.)1980:ReadingsinPhilosophyofPsychology,2vols.Cambridge,MA:HarvardUniversityPress.——1980:MadPainandMartianPain.InN.J.Block(ed.)ReadingsinPhilosophyofPsychology,Vol.1.Cambridge,MA:HarvardUniversityPress.Lycan,W.1981:Form,Function,andFeel.JournalofPhilosophy,78,24–50.198\nPHILOSOPHYOFMIND——1985:AbortionandtheCivilRightsofMachines.InN.PotterandM.Timmons(eds)MoralityandUniversality.Dordrecht:D.Reidel.——1987:Consciousness.Cambridge,MA:MITPress/BradfordBooks.——1988:JudgementandJustification.Cambridge:CambridgeUniversityPress.——(ed.)1990:MindandCognition:AReader.Oxford:Blackwell.——1991:HomuncularFunctionalismmeetsPDP.InW.Ramsey,S.P.StichandD.Rumelhart(eds)1991:PhilosophyandConnectionistTheory.Hillsdale,NJ:LawrenceErlbaumAssociates.——1996:ConsciousnessandExperience.Cambridge,MA:MITPress/BradfordBooks.——(ed.)1999:MindandCognition:AnAnthology.Oxford:Blackwell.Lycan,W.andPappas,G.1972:WhatisEliminativeMaterialism?AustralasianJournalofPhilosophy,50,149–59.McClelland,J.L.,Rumelhart,D.E.andthePDPResearchGroup1986:ParallelDistributedProcessing:ExplorationsintheMicrostructureofCognition.Cambridge,MA:BradfordBooks/MITPress.McDonough,R.1991:ACulturalistAccountofFolkPsychology.InJ.Greenwood(ed.)TheFutureofFolkPsychology.Cambridge:CambridgeUniversityPress.Malcolm,N.1984:ConsciousnessandCausality.InD.ArmstrongandN.Malcolm,ConsciousnessandCausality:ADebateontheNatureofMind.Oxford:Blackwell.Millikan,R.G.1984:Language,Thought,andOtherBiologicalCategories.Cambridge,MA:BradfordBooks/MITPress.Morton,A.1980:FramesofMind.Oxford:OxfordUniversityPress.Nagel,T.1974:WhatisitLiketobeaBat?PhilosophicalReview,83,435–50.ReprintedinN.J.Block(ed.)1980:ReadingsinPhilosophyofPsychology,2vols.Cambridge,MA:HarvardUniversityPress.Owens,J.1986:TheFailureofLewis’Functionalism.PhilosophicalQuarterly,36,159–73.PlaceU.T.1956:IsConsciousnessaBrainProcess?BritishJournalofPsychology,47,44–50.ReprintedinLycan(1990,1999).Putnam,H.1960:MindsandMachines.InS.Hook(ed.)DimensionsofMind.NewYork:CollierBooks.——1967a:TheMentalLifeofSomeMachines.InH.-N.Castañeda(ed.)Intentionality,Minds,andPerception.Detroit,MI:WayneStateUniversityPress.——1967b:PsychologicalPredicates.InW.H.CapitanandD.Merrill(eds)Art,Mind,andReligion.Pittsburgh,PA:UniversityofPittsburghPress.Reprintedunderthetitle‘TheNatureofMentalStates’inN.J.Block(ed.)1980:ReadingsinPhilosophyofPsychology,2vols.Cambridge,MA:HarvardUniversityPress.——1975:TheMeaningof‘Meaning’.InK.Gunderson(ed.)MinnesotaStudiesinthePhilosophyofScience,Vol.7:Language,MindandKnowledge.Minneapolis:UniversityofMinnesotaPress.Quine,W.V.O.1960:WordandObject.Cambridge,MA:MITPress.Ramsey,W.,Stich,S.P.andGaron,J.1991:Connectionism,Eliminativism,andtheFutureofFolkPsychology.InW.Ramsey,S.P.StichandD.Rumelhart(eds)1991:PhilosophyandConnectionistTheory.Hillsdale,NJ:LawrenceErlbaumAssociates.Ramsey,W.,Stich,S.P.andRumelhart,D.(eds)1991:PhilosophyandConnectionistTheory.Hillsdale,NJ:LawrenceErlbaumAssociates.Rey,G.1980:FunctionalismandtheEmotions.InA.O.Rorty(ed.)1980):ExplainingEmotions.BerkeleyandLosAngeles:UniversityofCaliforniaPress.Robinson,W.S.1988:BrainsandPeople.Philadelphia,PA:TempleUniversityPress.Rorty,A.O.(ed.)1980:ExplainingEmotions.BerkeleyandLosAngeles:UniversityofCaliforniaPress.Rorty,R.1965:Mind–BodyIdentity,Privacy,andCategories.ReviewofMetaphysics,19.199\nWILLIAMG.LYCANRosenthal,D.(ed.)1991:TheNatureofMind.Oxford:OxfordUniversityPress.Ryle,G.1949:TheConceptofMind.NewYork:BarnesandNoble.Scherer,K.R.andEkman,P.1984:ApproachestoEmotion.Hillsdale,NJ:LawrenceErlbaumAssociates.Schiffer,S.1981:TruthandtheTheoryofContent.InH.ParretandJ.Bouveresse(eds)MeaningandUnderstanding.Berlin:WalterdeGruyter.Searle,J.1980:Minds,BrainsandPrograms.BehavioralandBrainSciences,3,417–24.Sellars,W.1963:Science,PerceptionandReality.London:RoutledgeandKeganPaul.Shoemaker,S.1981:SomeVarietiesofFunctionalism.PhilosophicalTopics,12,93–119.Skinner,B.F.1957:VerbalBehavior.NewYork:Appleton–Century–Crofts.Smart,J.J.C.1959:SensationsandBrainProcesses.PhilosophicalReview,68,141–56.——1963:PhilosophyandScientificRealism.London:RoutledgeandKeganPaul.Smolensky,P.1988:OntheProperTreatmentofConnectionism.BehavioralandBrainSciences,11,1–23.Sober,E.1985:PanglossianFunctionalismandthePhilosophyofMind.Synthèse,64,165–93.RevisedexcerptreprintedinLycan(1990,1999)underthetitle‘PuttingtheFunctionBackIntoFunctionalism’.Solomon,R.1977:ThePassions.NewYork:Doubleday.Stalnaker,R.1984:Inquiry.Cambridge,MA:BradfordBooks/MITPress.Stich,S.1978:AutonomousPsychologyandtheBelief–DesireThesis.Monist,61,573–91.ReprintedinW.Lycan(ed.)1990:MindandCognition:AReader.Oxford:Blackwell.——1981:DennettonIntentionalSystems.PhilosophicalTopics,12,39–62.ReprintedinLycan(1990,1999).——1983:FromFolkPsychologytoCognitiveScience.Cambridge,MA:BradfordBooks/MITPress.Strawson,G.1994:MentalReality.Cambridge,MA:BradfordBooks/MITPress.Turing,A.1964:ComputingMachineryandIntelligence.InA.R.Anderson(ed.)MindsandMachines.EnglewoodCliffs,NJ:Prentice-Hall.Tye,M.1983:FunctionalismandTypePhysicalism.PhilosophicalStudies,44,161–74.VanGulick,R.1980:Functionalism,Information,andContent.NatureandSystem,2,139–62.Wilkes,K.1993:TheRelationshipBetweenScientificandCommonSensePsychology.InS.ChristensenandD.Turner(eds)FolkPsychologyandthePhilosophyofMind.Hillsdale,NJ:LawrenceErlbaumAssociates.Wimsatt,W.1972:TeleologyandtheLogicalStructureofFunctionStatements.StudiesinHistoryandPhilosophyofScience,3,1–80.Wittgenstein,L.1953:PhilosophicalInvestigations(translatedbyG.E.M.Anscombe).NewYork:Macmillan.Wright,L.1973:Functions.PhilosophicalReview,82,139–68.DiscussionQuestions1Ifmindsareimmaterialandnon-spatial,howcantheyinteractwithbodies?2Couldthesuccessofneurophysiologyeliminatethemindfromexplanationsofwhatwedo?3Towhatextentshouldthephilosophyofmindbeconstrainedbyourcurrentviewofscience?4IfamachinecouldpasstheTuringTest,whatwouldbetheconsequencesforthephilosophyofmind?5Whenyouseearedobject,couldyouhavethesortofcolourexperienceIhavewhenIseeagreenobject,andviceversa?200\nPHILOSOPHYOFMIND6Arementalstatesandeventsidenticalwithstatesandeventsinourcentralnervoussystems?Whatnotionofidentityisinvolvedinthisclaim?7Arementaltermsdefinedcausally?8Doesanadequateaccountofthemindrequiredistinctlevelsofdescription,allappliedtothesamefundamentalreality?9Howshouldcognitivepsychologyandphilosophyofmindberelated?10Doesworkinartificialintelligencehelptoanswerphilosophicalquestionsabouthumanintelligence?11Canhumanpsychologicalcapacitiesbeentirelycapturedbyathird-person,hardware-realizabledesignthatcouldbebuiltinalaboratory?12Couldahumanbeing’sknowledgeeverbecompletelyformalizedinasetofrepresentationsorrules?13Doweneedateleologicalnotionoffunctionforouraccountoftherealizationofpsychologicalstatesandpsychologicalexplanation?14Howcanamentalstatebeaboutsomething?Howcanamentalstatebeaboutsomethingwhichdoesnotexist?15Cantherebeanadequatenaturalisticanalysisofteleologicalfunction?16Doesfunctionalistmetaphysicsleaveoutthemostimportantfeatureofthemental:whatitfeelsliketobeinamentalstateofanycertainsort?17Ifmentalitemsarenotnecessarilyidenticalwithneurophysiologicalones,cantheybeidenticalwiththematall?18Canweexplainconsciousness?19Ifyouhaveanafter-image,doesitshowthatfunctionalismismistaken?20Doweknowfromtheinsidethatwearenotmerelyautomata?21Howistherepresentationalcontentofastateofthebraindetermined?22Iffunctionalrolesare‘narrow’andrepresentationalcontextsare‘wide’,canpropositionalattitudesbeidentifiedwithfunctionalroles?23Areintentionalfeaturessemanticfeatures?24Whatisa‘languageofthought’?Cantherebeanadequatetheoryofthinkingwhichdeniesthatthereisalanguageofthought?25Doesthecausalhistoryofamentalitemdeterminewhatthatitemrefersto?26Cantheascriptionofbeliefsanddesiresbeobjectivelytrue,butnotinvirtueofdescribingcausallyactiveinnerstatesofpeople?27Towhatextentdoesourascriptionofbeliefsanddesirestopeopledependupontheassumptionthatpeoplewillgenerallybelieveanddesirewhattheyoughttobelieveanddesire?28Doesconnectionismoffergreaterinsightintophilosophicalproblemsthanthestandard‘rulesandrepresentations’modelofartificialintelligence?29Arementaltermstheoreticaltermsofafolkscience?Shouldwebepreparedtogivethemupinfavourofthetermsofasuperiortheory?30Iftherepresentationalcontentsofapropositionalattitudedependonexternalfactors,arethecontentsgenuinepropertiesofthesubject?Whataretheconsequencesofyouranswerfortheontologicalstatusofbeliefsanddesires?201\n6EthicsJOHNSKORUPSKIInradicallydifferentways,philosophersfromSocratestoWittgensteinhavefoundinethicsasourceofdeepphilosophicalperplexity.Virtues,principlesandconsequencesforhumanwell-beinghaveallbeenproposedasthemostimportantfocusofethicalunder-standing.Intertwinedwiththeseareconcernsaboutrationality,impartialityandmoralfreedom.Philosophershavealsoaskedquestionsaboutethics:forexample,iftherearespecialmoralfacts,howcouldweknowthem;withoutsuchfacts,whatelsecouldmakeethicaljudgementstrueorfalse?Thischapterconsidersgeneraltheoriesofethicalvalueasgeneraltheoriesofreasonsforaction,focusingonthepowerfulandilluminatingutili-tarianandKantiancontractariantheories.ReaderswillwanttoconsultchaptersonPOLITICALANDSOCIALPHILOSOPHY(chapter8),HUME(chapter31),KANT(chapter32),BENTHAM,MILLANDSIDGWICK(chapter35)andonAPPLIEDETHICS(chapter16),BIOETHICS,GENETHICSANDMEDICALETHICS(chapter17),ENVIRONMENTALETHICS(chapter18)andBUSINESSETHICS(chapter19).Chaptersaboutseveralotherhistori-calfigurescontaindiscussionsoftheirmoralphilosophy,andchaptersonEPISTEMOL-OGY(chapter1),METAPHYSICS(chapter2),PHILOSOPHYOFMIND(chapter5),AESTHETICS(chapter7)andPHILOSOPHYOFRELIGION(chapter15)canhelptounder-standissuesraisedhere.1TheScopeofEthicsReflectionaboutethicshasbeenavitalcomponentofallthetraditionsofphilosophiz-ingthatweknow.China,Greece,IndiaandmedievalandmodernEuropeallaskedbasicethicalquestions.Whatisthegood?Whatmakesalifeagoodlife?Whatarethevirtuesofahumanbeing?Isthereonegood,oroneunifiedschemeofvirtues?Whatdutiesdowehavetoeachotherortoourselves?Sofaraswehaverecordsofnon-literateculturesandtheiroraltraditions,wefindthesamequestionsbeingaskedtheretoo.Theyarephilosophicalquestionsiftheyexaminemoralhabitsandteachingsinareflectiveway.CLASSICALGREECE(chapters22and23)alsomovedtoasecondlevelofphilosophicalreflection:itaskedquestionsaboutsuchquestions.Cantheyhaveobjec-tiveanswers?Ifso,whatkindofknowledgeoftheseanswerscanwehave?Arethey\nETHICSmattersofreasonorfeeling?Thesehigher-orderissuesareoftencalledmeta-ethical.AstrikingfeatureoftheWesterntraditioninphilosophyistheurgencywithwhichithasrecurrentlyworriedaboutthem,fromitsGreekbeginnings.Andinthiscenturymeta-ethicalquestionshavebeenpursuedwithanurgencyandaperseveranceneverknownbefore.Weshallhavetosaysomethingaboutmeta-ethicsaswellasaboutethicalquestionsthemselves.Butfirst,whatisthescopeofethicsassuch?Itdoesnotdealwiththewholedomainofvalue.IfIenquireintothebeautifulorthesublimeIammorelikelytobedoingAESTHETICS(chapter7)thanethics,thoughmyinterestmightcertainlybeethicalaswell,inonewayoranother.Ethics,onemaysuggest,isconcernedwithmorality,ratherthanwithart.True,butitdoesnotstandtomoralityquiteasaestheticsstandstoart.Ithaswiderscope,sincequestionsabouttherelationsbetweenmoralityandartbelongtoit,asdoquestionsabouttherelationsbetweenmoralityandprudenceorbetweenmoralityandreason.SoifIaskwhatmakesapieceofmusicbeautiful,orwhatmakesagoodtennistechnique,Iamnotaskinganethicalquestion,oneventhewidestsenseof‘ethical’;butifIaskwhatplaceinmylife,orinotherpeople’slives,musicortennis,togetherwiththeskillsandexcellencespropertothem,shouldhave,thenIamaskinganethicalquestion.Icanaskanethicalquestionaboutmoralityitself–whatroleinlifeshouldmoralityplay?Ethicalquestionshaveacertaindetachmentandcompre-hensivenesswhichgivethemagoverningroleinthephilosophyofvalue.QuestionsofPROFESSIONALETHICS(chapter16)inbusiness,medicineorjournalismhaveasimilargoverningrole,bearingonthenatureoftheprofession,itsplaceinsociallifeandtheconsequentapplicationofgeneralethicalprinciplestoit.EthicalquestionshavethisgoverningrolebecauseethicalenquiryisconcernedquitegenerallywithreasonsforACTION(pp.734–6).Notallreasonsarereasonstoact.Therearereasonstobelieveandreasonstofeel:indeed,wedealinreasonsateveryturn.Tobealiveandawakeistobealivetoreasonsforbelieving,doing,feeling.You,forexample,havereasontobelievethatIdidnotchangetheoilinyourcar;youhavereasontofeelirritated;youhavereasontotakeaction–todeductsomethingfrommybill.Oureval-uationsofapersonregularlyturnonhowthatpersonrespondsorfailstorespondtoreasonsinoneorotherofthethreedomainsofbelief,feelingandaction.Personalityisthemannerofone’ssensitivitytoreasonsinallthreeofthem.Anirritablepersonismoreirritatedthanheorshehasreasontobe.Acredulouspersonbelieveswhenthereisinsufficientreasontobelieve.Aprecipitatepersonactswhenthereisnoreasontoact.Evenatitswidest,though,ethicsdoesnotdealwiththiswholenormativedomainofreasons.Itdoesnotdealwithwhetheraconclusionisrightlyorwronglydrawnfrompremises,withwhetherevidenceisgoodorbad,orwithwhathypothesisweoughttoadopt.SuchquestionsaretheprovinceofLOGIC(chapter4).Ethicsinitswidestsensestandstoquestionsaboutwhatthereisreasontodo,aslogicinitswidestsensestandstoquestionsaboutwhatthereisreasontobelieve.Itisthenormativetheoryofconduct,aslogicisthenormativetheoryofbelief.Throughitsconcernwithactionandreasonsforactionethicsalsobecomesconcernedwithcharacter,asitbearsonactionandreasonsforaction.(Theword‘ethics’derivesfromtheGreekethos,whichmeans‘char-acter’,or,inplural,‘manners’.)And,throughthatconcernwithcharacter,itbecomesconcernedwithquestionsaboutwhatthereisreasontofeel,andhowreasonstofeel203\nJOHNSKORUPSKIconnectwithreasonstoact.Butthebestwaytomakethiscleareristoturndirectlytosomeofthefeelingswithwhichethicsisconcerned.Ofthese,threeatleastarefunda-mental:blame,admirationanddesire.2Blame,AdmirationandDesireBlameisacentralcategoryofmorality.Somuchsothatonemaycharacterizemoralitybyreferencetoit.Forwehavemanywaysofdisapprovingofactions–holdinganactionmorallywrongisonlyone.Anactionmaybeidioticortasteless,butnotmorallywrong.Tocallitmorallywrongisamoreseriousmatter.‘Itwasworsethanimmoral,itwastasteless’,wouldbeaquip,likeBoulaydelaMeurthe’scomment(ontheexecutionoftheDucd’Enghien)–‘Itwasworsethanacrime,itwasamistake’.Thequipcouldmakeapoint,sayinthespiritofOscarWilde,butitwoulddosopreciselybyparadox.Whatisitthenthatmarksthespecialdisapprovalorhostilityconveyedincallingsomethingmorallywrong?Tocalltheactionmorallywrongistoblametheagent.Blameisanactorattitudewhosenotionalcoreisafeeling,inthewaythatfeelingsorryisthenotionalcoreofapology–Imeanthat,evenwhenthefeelingisnotactuallypresent,itisneverthelessinvokedineveryactofapologyorblame.Callthisemotionalcoreofblame‘theblame-feeling’–itdisposestopunish,asfeelingsorrydisposestomakeamends.Moreover,justasapologizingisalreadymakingamendstosomedegree,sopublicblame(andalsoself-blameintheformofguilt)alreadyisadegreeofpun-ishment.Thisoftenmakespeoplereluctantto‘apportionblame’.Thereis,itshouldbenoted,awidersenseoftheword‘blame’,inwhichwemay,forexample,blamethecar’sfaultybrakesfortheaccident–inotherwords,identifythemasitsrelevantcause.Butwecertainlyfeelnosimilarreluctancetoblameinthatsense.Therelationbetweenthiswiderandthenarrower,moral,senseofblameisafasci-natinganddeepissueinethics,butwemustpassitby.Characterizingthemorallywrongastheblameworthyinthenarrowsense,wecansaythatthemorallyrightisthatwhichitwouldbemorallywrongnottodo.Similarly,wecansay‘Xismorallyobligatory’or‘Xmorallyoughttobedone’willholdjustifthenon-performanceofXisblameworthy.Ofcourseitistrueandimportantthatweadmirepeopleforgoingbeyondthemorallyobligatory–‘beyondthecallofduty’–eventhoughwedonotblamethemfornotdoingso.Wecansaythatadmirationofsuchactionsismoraladmiration,becauseweadmirethemforthereasonsthatimpelledthem,andthosereasonsaremoralreasons.Forexample,lackofconsiderationforothers’feelings,whenitreachesacertainpoint,becomesblameworthythoughtlessness.Buttherearedegreesofcareforothers’feelingsthatgowellbeyondwhatwewouldexpect,onpainofblame,frompeopleingeneral,butwhichwestilladmirewhenwemeetwiththem.(Therearealsoexcessesofsolicitousnessofcourse.)Theoutstandinglyandtheordinarilythoughtfulpersonareimpelledbythesamereasons–considerationforothers’feelings.Andwecansaythatthosereasonsaremoralreasons–asagainst,say,prudentialoraestheticreasons,sincetheirabsencefromaperson’smindbeyondacertainpointbecomesblameworthy.Nowforanotherimportantpoint:‘blameworthy’means‘oughttobeblamed’.Isthis‘ought’itselfamoral‘ought’?Nosuchcircularityisinvolved,foraswehavenoticed204\nETHICS‘ought’ingeneralcanbedefinedintermsofreasons–inthiscasereasonstoadoptafeeling,orfeeling-basedattitude.Tosaythatanactionisblameworthyistosaythatthereisadequatereason,takingeverythingintoaccount,toblametheagentfordoingit.Thisisaspecialcaseofageneralpattern,inwhichweassessreasonsforfeelinganemotion.Sowithmanyothertermsofappraisal–‘irritating’,‘despicable’,‘frighten-ing’,‘moving’,‘tasteless’andsoon.Twothingsseemtobeinvolvedinsuchappraisals:firstly,theappraiser,intypicalcases,spontaneouslyfeelstheemotionand,secondly,heexpectsotherstofeelit.Butacertainnormativityisbuiltin:hefeelstheemotiontobeappropriate,andbelievesthatothergoodjudgeswillshareit.‘ImustsayIwasratherbored–butIwasfeelingtiredanddistracted,andIprobablydidn’tunderstandwhatthepointofitwas.’SointhiscaseIdonotjudgethatitreallywasboring.Iamdis-qualifyingmyselfasagoodjudge.Theinternalcriteriaofappropriatenessforsuchfeel-ingscanbedebatedwithgreatexactnessandsubtlety,andthe‘commonpursuitoftruejudgement’inaestheticormoralappraisalturnsonsuchdebate.Notthatcommitmenttothethoughtthatotherjudges,whosufferfromnodisqualifyingdefectorlimitation,wouldconfirmmyjudgementisspecialtothemoralandtheaesthetic:itisagenerallogicalfeatureofjudgementassuch.Judgementsabouttheblameworthyortheadmirable,then,arejudgementsaboutwhenitisreasonabletorespondwithblameoradmiration.Iftheemotionalcoreofmoralityistheblame-feeling,admirationisthefeelingfundamentaltosystemsofideals.Ihaveinmindtheidealsofcharacterandexcellencewhichplaysolargeapartinshapingthewayweliveourlives.Itisalltooeasyforethicstoignorethem,forexamplebymakinganover-simplecontrastbetweenmoralityandprudence.Admirationfor,anddebateabout,physicalcourage,style,‘cool’,imagination,presenceofmind,skill,wit,resourcefulness,sensitivity,fitness,goodlooksandmanyothersuchthingsdrivesagreatpartoflife–andofcriticismofwaysofliving.Thesearenotmoralvaluations,thoughinsomecasestheyshadeintothem.Whateveronefeelsabout,forexample,obtuseness,cravenness,leadenlackofstyleorclumsiness–perhapscontempt,derision,embarrassmentorpity–itneednotbetheblame-feeling.Theyarenotprudentialvaluationseither.Prudentialvaluationsbuildonthenotionofone’swell-being,ofwhatoneshouldpursueasanendwhenone’sownbestINTER-ESTS(pp.760–1)areexclusivelyinquestion.Inthiscasetherelevantaffectivecoreisneitheradmirationnortheblame-feeling,butdesire.Thewell-beingofanindividualcompriseswhateverisfoundinitselfdesirabletothatindividual.And‘desirable’standstowhatthereisreasontodesireas‘admirable’and‘blameworthy’standtowhatthereisreasontoadmireandblame.Questionsaboutwhatisdesirable,forpeopleingeneralorforaparticularperson,arequestionsabouttheendsoflife.Theyareasmuchamatterofdeliberationanddebateasquestionsaboutwhatisadmirableandblameworthy.WhenJohnStuartMILL(chapter35)arguedthathappinessisdesirablebecauseitiswhatpeopledesireasanend‘intheoryandinpractice’,hewasnotwastinghistime.Itisstilllesstrivialthathappiness,asheattemptedtoshow,istheonlythingdesirable.Whetherhewasrightorwrongaboutthis,hismethodwassound.ToreachaconclusionaboutwhatendsareintrinsicallydesirableImusttrytoreflect,withoutself-delusion,onwhatIdesireandwhy–andifIfindthatIdifferfromothersinmyconclusions,Imustaskwhethermy205\nJOHNSKORUPSKIdesiresmightyieldtogreaterexperienceorknowledgeand,indeed,whethermyideasaboutwhatthosedesiresreallyaremightnotbedistortedbypersonalideals,conven-tionalproprietiesorjustsheerwishfulthinking.Heretoo,then,thereisa‘commonpursuitoftruejudgement’–areflectiveexami-nationofone’sownspontaneousfeelings,carriedoutindialoguewithotherswhosimi-larlyexaminetheirs.Butthereisadifference.Thoughthepursuitmaybecommon,theconclusionneednotbe,andisnotlikelytobe,thatwhatisdesirableformeisjustthesameaswhatisdesirableforyou.Notonlywilldifferentpeopledifferonwhatmakesthemhappy;theyarealso,despiteMill’ssuggestiontothecontrary,quitelikelytodifferonhowdesirablehappinessistothem,asagainst,say,knowledge,freedomorachieve-ment.Incontrast,supposeIfindthatmanypeoplearemovedbyapieceofmusicthatIfindcacophonousordull,andthatdiscussionandfurtherlisteningdoesnotleadmetofindanythingmovinginit.Iamunlikely(philosophicalpreconceptionsaside)tocon-cludethatitis‘movingtothem’butnot‘movingtome’.Imayconclude,nodoubtsilently,thattheirappreciationisshallow–orwithgreaterhumility,thatmyownis.Whatisdesirableformemaynotbedesirableforyou–butaestheticvaluationsdonotnaturallyrelativizeinthisway;moralvaluationsstillless.Thisdifferencecertainlydeservesfurtherexamination.(Distinctionsareneeded:forexample,betweenparticularobjectsthatwedesire–thisoyster,aweekendinVienna–andthecategorialendsinvirtueofwhichwedesirethem;oragain,betweenwantingathingandlikingitwhenyougetit.)Overall,however,itremainstruethatexamina-tionofone’sownspontaneousresponses,togetherwithdiscussionwithothers,isthecriterionofallvaluation–includingtheaestheticandthemoral,andalsotheprudential.Thatisnottosaythatonecandefineanevaluativepredicate–‘tedious’,‘tasteless’,‘desirable’,‘delightful’,‘kind’–intermsoftheresponsesofgoodjudges.Howeverinformedandsensitivesuchjudgesmaybeitislogicallypossiblethattheyarewrong–thatwhatstrikesthemastastelessisnottastelessbutinnovative,andsoon.Ofcourseif‘goodjudges’aresimplydefinedas‘judgeswhogetitright’,thenitwillbealogicaltruththatathingistediousjustifgoodjudgeswouldfinditso.Butitwillthenbecirculartodefinethetediousasthatwhichgoodjudgesfindtedious.Reflectiveagreementis,itistrue,ultimate,inthesensethattheonlywayofappealingagainstitisbyforginganewreflectiveagreement–butitisneverguaranteedtobeincorrigible.Nor,whenIjudgethatsomethingistasteless,amIjudgingthatotherswillorwouldfindittasteless.Myeyeisontheobject,notonwhatotherpeoplemightthinkabouttheobject.3Character,theVirtuesandFreedomHowdovirtuesandvicesfitintothepicturethatwehavedrawn?How,inparticular,aretheyconnectedwithblame,whichwehavemadecentraltothenotionofmoral-ity?Theyareconnectedinthisway:theyaretraitsofcharacterwhichwecouldbeblamedfornotattemptingtoattainorlose.Theyarealsoconnectedinthisway:theydirectlydisposetoactionswhicharemorallyrightorwrong.Thevariousvirtuesinvolvespontaneoussensitivitytovarioustypesofreasonsforacting,andthevariousvicessimilarlyinvolveinsensitivitytovarioustypesofreasonsforacting.206\nETHICSThisisnottosaythatweareblamedforourvicesassuch,asagainstouravoidablefailurestoresistthemortotrytomendthem.Itisabasicprinciplethat‘oughtimpliescan’–thatwhatIoughttodoIcando.Anditisalsotruethatblamemakessenseonlyinrespectofthatwhichoughtnottohavebeendone.SoifIaminfactunabletomendmycharacterthenIcannotbeblamedforfailingtodoso–thoughIcanbeblamedfornotstrivinghardenoughtoresistitscharacteristicimpulsesonparticularoccasions.Thisdistinction,betweenchangingmycharacterandresistingitsimpulses,isavitalonetoordinarymoralthought.Imaybecruel,andImaynotbeabletodomuchtomakemyselflesscruel;thatis,toridmyselfofthosecharacteristicimpulsestocruelacts.ButIcanstilltrytorefrainfromactingcruellywhenIamtemptedto.Apersonwhohasnocontroloverhiscruelactionsisnotjustcruelbutpathologicallycruel.Hedoesnot,onthispointatleast,havetheimportantkindofmoralfreedomthatweattributetoanyonewhowethinkofasamoralagent,andwhichwepresupposewhenweblame.ThusifIblameyouforacrueljoke,forexample,Iassumeyoucouldhaveresistedtheimpulsetomakeit.IfIbecomeconvincedthatyouareutterlyunabletoresistsuchimpulses(even,say,whenitisgreatlytoyourpersonaladvantagetodoso)Iammorelikelytoseeyouasmad,oratanyrateasobsessed,thanasbad.Thisisthewayinwhichmoralfreedomispresupposedinattributionsofmoralresponsibility.Itisrightlypointedout,forexample,thatthefactthatIwasfreetodowhatIwantedtododoesnotsufficetoshowthatIwasmorallyfree,becausethewantitselfmayhavebeensoobsessiveastounderminemymoralfreedom.Drugaddictsarenotmorallyfreeevenwhentheydowhattheircravingcompelsthemtodo.Theyaretheslavesoftheiraddiction.Twokindsofthingmaybehappeninghere.Firstly,theircravingmaybesostrongastocloudtheirjudgement–thoughtherearegoodreasonsforthemtoresistit,thecravingitselfstopsthemfromseeingthem.Secondly,thecravingmaynotcloudtheirjudgement,butitmayimpairtheirabilitytoactonit–theyseeclearlyenoughthattheyhavegoodreasonnottodowhattheycravetodo,buttheyareunabletostopthemselves.Inboththesecasestheirmoralfreedomisattenuated,andtheirmoralresponsibil-itydiminished.Thusifwesupposecravingtohavedrivenapersontosomecrime,thenbothcases,ifestablishedontheevidence,couldreasonablyberecognizedbyajudgeasbeing,tosomedegree,extenuating.Ingeneral,moralfreedomisafunctionofthetotaleconomyofone’scharacter–thedegreetowhichone’sfeelingsareunderthecontrolofone’scapacitytorecognizereasons,insteadofoutrunningitscontrolorevenreduc-ingorcontrollingthecapacityitself.Moralfreedom,soconceived,isoneofthevirtues,whichahumanbeingcanhaveingreaterorlesserdegree.Butitalsohasaspecialroleintheeconomyofcharacter:itisapreconditionofmoralagencyassuch,anditisfacili-tatedbyalltheothervirtues,sincetheseareaffectivedispositionstoactonparticularkindsofgoodreasons.Onecanenvisage,asalogicalpossibility,thepersonwhoseestherightclearlyand,bygreateffort,isabletoactonit,despiteever-presentviciousimpulses.Butitiseasierforonewhohasvirtuousratherthanviciousimpulsesinthefirstplacetobemorallyfree.Moralfreedomisrationality,understoodasthecapacityofrationalself-government.ThisdoctrineappearsinPLATO’s(chapter23)conceptionofjusticeinthesoulandhasbeencontinuallyreassertedintheWesternethicaltradition.Butifwearenatural207\nJOHNSKORUPSKIobjects,whosebehaviourisgovernedbycausallaws,canwebeinthissensemorallyfree?Thisisthephilosophicalproblemof‘freewillanddeterminism’.FreeWillandMoralFreedomManyphilosophershavedeniedthat,ifourbehaviourisgovernedbycausallaws,wecanbemorallyfree.Butothershaveseennoinconsistency.Theissueisimportantforethics,becauseblame,andthusmorality,presupposemoralfreedom.Inparticular,blamepre-supposesthattheresponsibleagentcouldhaverefrainedfromactingonthemotiveswhichinfacthedidacton–hadtherebeenreasontodoso.Inblaming,then,wejudgethatapersonhadthecapacitytorefrain–andthedistinctionswemake(isthatpersonjustacasualsmokeroranaddict?)seemnottoturnatallonthetruthorotherwiseofanyglobalthesisaboutcausaldeterminationofhumanbehaviour.Beforeweblameanagent,wewanttoknow,amongotherthings,whetherhecouldhaverefrainedfromdoingwhathedid.Inparticular,wewanttoknowwhetherhecouldhaverecognizedthereasonsforrefraining,andrefrained,giventhathisdesiresstoodastheydid.Thisisaquestionabouthiscapacityofrationalself-control.Wearenotaskingwhetheraworldispossibleinwhichallthecausesofhisactionstoodastheydid,andtheactiondidnotoccur.Ifdeterminismistrue,thereisindeednosuchworld;buthowdoesthisbearonrationalself-control,andthusontheissueofmoralfreedom?Certainlyacloserexaminationmightyetshowsomeconnection.Itisuptothephilosopherwhothinksthatthetruthofdeterminismwouldunderminemoralfreedom–thepowerofactingornotacting,accordingtothedeterminationsofreason–toshowwhattheconnectionis.Therehavebeenplentyofattemptstoestablishaconnection.Theresilienceofthedisputebearswitnesstoitsintractability.4AnEthicalTheory:TheCaseofUtilitarianismWehavespokenofhowjudgementsaboutwhatisdesirable,blameworthyandadmirablearebasedonexaminationofone’sfeelingsanddiscussionwithothers,andofhowtheygroundtheinteractingvalue-spheresofprudence,moralityandexcellenceinactivityandcharacter.Butwehavesaidnothingabouthowsuchjudgementsgiveusreasontoact.Sinceethics,onouraccount,isthegeneralstudyofreasonsforaction,thisimportantquestionisethical.Onecouldsaythatethical‘theory’simplyconsistsofconsidered,thoughtfulanswerstoit.Thenitwouldhardlybecontroversialthatthereissuchathing.Butitiscontro-versial–because‘theory’suggestssomethingmorethanasetofpiecemealandper-sonalanswers,howeverthoughtful.Itpromisessystemandobjectivity,andmanypeopledoubtthatanswerstoethicalquestionscanachievethoseaims.Theirdoubtsarenotfactious.Bynowinthehistoryofethics–inourworldofethicalcultureswhichseeminvinciblyfragmentedanddisparate–theymustbetakenseriously.Thereisnopoint,however,inabstractclaimsandcounter-claimsabouthowambitiousethicaltheoryshouldbe.Wewillmakemoreprogressbytakingsomethingwhicheveryoneagreestobeanambitiousethicaltheory–UTILITARIANISM(chapter35)–considering208\nETHICSitsstructureandvariousobjectionstoit,andasking,finally,whenandinwhatwaythesebecomeobjectionsto‘ethicaltheory’assuch.Inthiswaywewillseewhatdifficultiesfacetheideaofsystem;difficultiesfacingtheideaofobjectivitywillbeconsideredinsection11.Utilitarianismisthethesisthatthewell-beingofeachandeveryindividualhasintrinsicethicalvalue,thatthegreaterthewell-beingthegreateritsvalue,andthatnothingelsehasintrinsicethicalvalue.Thephrase‘intrinsicethicalvalue’callsforexplanation.Letussaythatapropertyhasethicalvalueifthereisreasontodowhateverislikelytoproduceorpreserveit,increaseitsdegreeandsoon.Ithasintrinsicethicalvalueifthatreasonconsistsjustinthefactthatthepropertyisthatproperty,anddoesnotderivefromanyfurtherfactsaboutitsconnectionswithotherproperties.Whytalkabout‘properties’?Theethicalvalueofobjectsandstatesofaffairs,thereasonsforactingthattheygiveriseto,whetherintrinsicorextrinsic,alwaysseemstoresideinsomethingaboutthem–theyhavevalueinvirtueoffeaturesorpropertiesthattheyhave.Indeedthisseemstrueforvalueofanykind:thevalueofathingalwaysstemsfromsomefeatureorpropertythatithas.Acurrentlycommonwayofputtingitistosaythatthevalueofathingalwayssupervenesonsomepropertiesthatithas.Ifacertainvaluesupervenesonsomepropertiesthenanythingwhichhasjustthosepropertieswillhavejustthatvalue–adifferenceinvalueentailsadifferenceinthoseproperties.Fortheutilitarian,theonlythingthathasintrinsicethicalvalueisapropertyofindi-viduals,apropertytheycanhavetogreaterorlesserdegree:thatofbeingorfaringwell.Thegreaterthewell-beingthegreatertheethicalvalue;thatis,thegreaterthereasontoproduceorpreserveit.Utilitarianswillwishtospelloutwhatwell-beingconsistsin.Theywillthenhavetoconsiderwhatendsaredesirable,bythemethodofself-examinationanddiscussionconsideredinsection2.Butourinterestnowisinthestructureoftheutilitarianviewratherthanitsdoctrineofwhatwell-being,or‘utility’,comprises.Theutilitarianholdsthattheethicalvalueofwell-beingisnotaffectedbywhosewell-beingitis–oratleastthatistrueiftheonlyindividualsunderconsiderationarehumanbeings.Bututilitarianshavealwaysfreelygrantedthatthewell-beingofallindividualswhichcanbesaidtobesentient,orperhaps(ifthisisdifferent)tohavedesires,shouldbetakenintoaccount.Nowthisisstillasmallerclassthanthatofindi-vidualswhichcanbesaidtohavewell-beingatall.Livingthingsingeneralcanfarewellorbadly,andagreenethicistmightholdthatallsuchvarietiesofwell-beinghaveintrinsicethicalvalue.Bututilitarianshavenotheldthatview,norhavetheyheldthatthewell-beingofallsentientindividualshasequalethicalvalue.Thewell-beingofnon-humananimalshasanintrinsicethicalvalue,butitisnotasgreatasthatofhumanbeings.Deepissueslurkhere,notonlyforutilitarians,aboutwhatthesedifferencesofethicalvalue,asbetweenthewell-beingofdifferenttypesofindividual,superveneupon.Atanyrate,whatevertheclassofindividualswhosewell-beingistobetakenintoaccount,theutilitariantakesgeneralwell-beingtobeapositivefunctionofthewell-beingofallofthem,andofnothingelse–afunctioninwhichtheutilityofeveryindividualistakenimpartiallyintoaccount.Ethicalvaluethenconsistsingeneral209\nJOHNSKORUPSKIwell-beingandinnothingelse.Butnotethat‘generalwell-being’isanabstractionconstructedoutofthewell-beingofconcreteindividuals.No‘generalbeing’enjoys‘generalwell-being’.Onlyinthewell-beingofconcreteindividualsdoesethicalvalue,accordingtotheutilitarian,reallyreside.Classicalutilitarianismisusuallytakentoholdthatgeneralwell-beingisfixedbythesumofthewell-beingofallindividuals.HenrySIDGWICK(chapter35),thethirdofthegreattrioofBritishutilitarians,suggestedthatitmightalsobetakenastheaverage:thatis,thatthegreatestethicalvalueisrealizedwhenaverage,ratherthanaggregate,well-beingisatitshighest.Eitheroftheseviewsassumesthatonecanaddupthelevelsofwell-beingofdifferentindividuals–itassumesthatthatprojectmakessense,notthatitcanactuallybedone.Thepracticalquestionofmeasurementbecomesimportantifonewantstoapplyutilitarianism,forexampleinwelfareeconomics.Butevenasapureethicaltheoryclassicalutilitarianismdoesrequirethatcomparativequestionsliketheseareintelligible:wouldtheincreaseinX’swell-beingoffsetthedecreaseinY’s?WouldX’sincreasebegreaterthanthelossofY’sandZ’s?Theycertainlysometimesmakesense.IcanaskmyselfwhetherabookwillgivegreaterpleasuretoKateorJulia.IcanevenaskmyselfwhetherKatewouldgetmorepleasureoutofthisbookthanJuliaandAnnawouldfromtwootherbooksathalftheprice.Someveryroughcomparingandsummingisgoingonthere.Thenotionofasumofwell-beingmaybesufficientlywell-definedinsomeareas,forsomepurposes,tobemeaningful,evenifitisnotinall.Wedoseemtouseit.Butdosuchquestionsalways–eveninprinciple–makesense?Andhowimportanttoutilitarianismisitthattheyalwaysshould?JohnStuartMillwouldhaveconcededthattheydonotalwaysmakesense,forhethoughtthatonecandistinguishbetweenqualityandquantityofhappiness.Happi-ness,hethought,waswhatwell-beingconsistsin,butsomeformsofhappinessare‘higher’–orperhaps,deeper,truerormorevaluable–insomewayotherthanthatinwhichhavingtwochocolateicesismorevaluablethanhavingone.Soperhapstherecanbediscontinuityhere,inthesensethat,insomecontextsatleast,apersonmayreasonablybeunwillingtotradeoffahigherpleasureforanyquantityofalowerone.Hepreferstheformertothelatterbutitwillnotmakesensetosaythathefindsitntimesasgood.Forifitdid,heoughttobewillingtosacrificeaparticularhigherex-periencefornormoreofthelowerones.Thisatleastgesturestowardscomplicationswhicharegenuinelypresentinourthinkingaboutwhatmakesforagoodlife.Weshouldcertainlybearinmindthatitmaybeintelligibletoorderpossiblestates(perhapsonlypartially)bythelevelofwell-beingtheyproduce,evenifitisnotgenerallypossibletosaywhatfractionthewell-beinginstateXisofthewell-beinginstateY.Theutilitarian’sbasicideawasthatconsiderationsofwell-beingprovidetheonlyintrinsicreasonsforaction,andthatinassessingthestrengthofsuchreasonsoneshouldconsiderimpartiallythewell-beingofallindividualsaffected.Theonlythingthatthisrequiresisthatweshouldbeableinprincipletoorderdistributionsofwell-beingbytheirethicalvalue.Utilitarianismthenholdsthatethicalvalueisafunctionofsuchdistributionsandofnothingelse.Itholdsfurtherthatthefunctionispositive:ifthewell-beingofoneormoreindividualsgoesup,withoutanyotherchange,thenethicalvaluegoesup.Andfinallyitrequiresthatthefunctionbeimpartial.Butwhatisinvolvedinthisnotionofimpartiality?210\nETHICS5ImpartialityMaximizingthesumofindividuals’well-being,ifitmakessensetotalkinthiswayatall,certainlylooksimpartial.Noindividual’swell-beingisgivengreaterweightintheadditionthanthatofanyother.Noonehashisorherwell-beingmultipliedbytwoordividedbythree.Thisisonewayofimplementingtheprinciple‘Everybodytocountforone,nobodyformorethanone’–whichinUtilitarianismMillattributedtoBenthamandwhichhetooktoexpresstheprincipleofimpartiality(Mill1963,X:257).Hewas,itshouldbenoted,talkingonlyofthewell-beingofhumans–ifnon-humansarecon-sideredaswell,aswenotedintheprevioussection,itceasestobesoplausiblethatimpartialityforbidsweightingsomeindividuals’well-beingmoreheavilythanothers.Thatisapointabouttheconnectionbetweenweightingandimpartiality.Butanotherpointisthataddingup–withorwithoutweighting–isonlyonewayofbeingimpartial.NotonlyisSidgwick’sideaofmaximizingtheaverageofwell-beingequallyimpartial,soisawidevarietyofotherdistributiveprinciples.ConsidertheDIFFERENCEPRINCIPLE(pp.260–1)proposedbyJohnRawls.Accordingtooneversionofthis(Rawls1972:83),resourcesshouldbesodistributedastomaketheworst-offgroupofindividualsaswell-offasitispossibleforthemtobe;afterthat,thenextworst-offgroupshouldbemadeaswell-offaspossible,andsoonupthescale.Rawlshimselfdoesnotproposethisprincipleasacomprehensiveethicaltheory.Hepre-sentsitasanaccountof‘justiceasfairness’inthePOLITICAL(chapter8)realm,anaccountwithwhichhehopespeoplewithavarietyofcomprehensiveethicalviewswillbeabletoagree.Butwecouldalsotreatitasaproposalabouthowimpartialityshouldbeunderstoodinacomprehensiveethicaltheory,aproposalwhichwouldsharewithclassicalutilitarianismthefeaturethatittakesethicalvaluetobeapositiveimpartialfunctionofindividualwell-beingandofnothingelse.Norisittheonlypossiblepro-posalwhichsharesthisfeaturewithclassicalutilitarianism;therearemanyothers.Thedifferenceprinciplefavourstheworstoff,inthatitattachesanethicalprioritytoimprovementsintheirwell-being.Butitcanalsobesaidtocounteverybodyforone,nobodyformorethanoneinthissense:itisindifferenttowhichconcreteindividualisunderconsideration.Ittakesaccountofnopropertyoftheindividualotherthanhisorhercomparativewell-being.Adistinctivefeatureofclassicalutilitarianismisthatittakesanequalamountofwell-beingtobeofequalethicalvalue,howeveritisdistributedacrossindividuals.Thisisnottrueofthedifferenceprinciple,becauseabenefittoworse-offpeoplewillhavegreaterethicalvaluethanabenefittothebetter-off.Note,however,thatthiscontrastassumesthatitmakessensetothinkofagivenamountofwell-beingbeingdistributedinvariouswaysacrossindividuals.Whereitdoesnotmakesensetotalkaboutmovingaboutanequalamountofwell-beingitwillnotmakesensetotalkofsummingoraver-agingeither.Thedifferenceprincipledoesnotrequirethatsuchtalkshouldmakesense,thoughitdoesrequirethatweshouldbeabletocomparethelevelsofdifferentindivid-uals’well-being.Notthatthisisanethicalargumentinitsfavour;itsimplyshowsthataswegetmoresophisticatedaboutwhatitmakessensetosay,wewillhavetofindsubtlerwaysofimplementingthesecompetingdistributiveprinciplesincompetingwelfarefunctions.Butallofthemareimpartial.211\nJOHNSKORUPSKIHowthenshouldweunderstandtheterm‘utilitarianism’today?Criticshavedonetheirworstwithit,andacasecancertainlybemadeforburyingit.Butanyotherterm,suchas‘welfarism’,isatleastasmisleadingandopentodistortion.And‘utilitarianism’hasthemeritofinvokingahistoricallydefinitetradition.Isuggestthatweshouldinterpretitgenerically–asnamingaclassofethicaltheories,allofwhichholdthatoverallethicalvalueissomepositiveimpartialfunctionofindividualwell-being.Onthisweakdefinition,whatIhavecalledclassicalutilitarianismbecomesaparticular–andparticularlysimple–speciesofutilitarianethicaltheory.‘Genericutilitarianism’generalizesitbyallowing(1)distributiveprinciplesotherthantheaggregativeprincipleoftheclassicalutilitarians;and(2)differentinterpretationsofwell-beingtotheclassicalviewofitasconsistingexclusivelyofhappiness.ItgeneralizesinthewaythatisdonewhenoneshiftsfromtheEuclideanparadigmofGEOMETRY(p.348)toaclassofgeometriesofwhichEuclid’sisonlyone:byallowingcertainaxiomsoftheoriginaltheorytobevaried.Whatdistributiveprinciple,andwhataccountofhumanends,toadoptremainsopentodebateamonggenericutilitarians.6WhatCananEthicalTheorybeExpectedtoDo?Utilitarianismisafullygeneraltheoryofethicalvalueandsoafullygeneraltheoryofreasonsforacting.Buttwofurtherelucidationsarerequired.Thefirstissimplyareminderthattherearenon-ethicalformsofvalue.Utilitarian-ismisageneraltheoryofethicalvalue,notageneraltheoryofvalue,soitdoesnothaveanythingdirectlytosayaboutthoseotherforms.Forexample,whatmakesapieceofarcheologicalevidencevaluableisitsbearingonarcheologicalquestionsaboutthepast,notitseffectongeneralwell-being.Thevalueoftheevidenceliesinthereasonsitgivesforbelievingonethingratherthananother:ageneralaccountofwhatmakesevidencevaluableisataskfortheepistemologist,nottheethicist.Similarly,whenwejudgeapianorecitaltobesuperb–ordreary–weappealtotheinternalstandardsofthatactiv-ity,theaestheticsofpiano-playing.However,wecanalsoask,inoneoranotherlargercontext,whetheritisworthpur-suinganadmittedlyvaluablepieceofevidence,orputtingonabrilliantpianoperfor-mance.Thequestionmightbe,forexample,whetherweshouldfundthepianorecitalorthearcheologicaldig.Considerapianistwhoplaysonlyforhimself,oradesert-islanderwhopursuesevidenceaboutthehistoryoftheislandthatwillneverbeknowntoothers.Theaestheticvalueoftheperformance,likethearcheologicalvalueoftheevidence,isquiteunrelatedtohowmanypeoplederivewell-beingfromit.Butiftheutilitarianisright,howmanypeoplederivewell-beingfromitisveryrelevanttoitsethicalvalue.Ifthepianistenjoyshispiano-playing,thatgivesitethicalvalue.Ifothersdotoo,thatgivesitevengreaterethicalvalue.Formsofnon-ethicalvaluearetrans-latedintoreasonsforactionviatheirethicalvalue.Theutilitariansdonotreduceotherformsofvaluetoethicalvalue;theypropoundasubstantivedoctrineofwhatconnec-tionthereisbetweenthetwo.Tosaythatgreatpiano-playingshouldbepursuedorfundedforitsownsake,irrespectiveofitscontributiontowell-being,istoadoptanotherdoctrine,incompatiblewithutilitarianism.212\nETHICSThesecondelucidationcanbemadebydistinguishingbetweenatheoryofethicalvalueandatheoryofdecision.Utilitarianismisafullygeneraltheoryofethicalvaluebutthatisnotthesameassayingthatitisafullydevelopedone.Itisconcernedwithintrinsicreasonsforactionandinthatsenseoffers,astheclassicalutilitarianssaid,a‘test’,or‘standard’or‘criterion’ofconduct.Butitisanotherquestionwhetheritcanbedevelopedintorulesorproceduresfordecidingwhattodo.Itdoesnotfollowdirectlyfromtheutilitariantheoryofethicalvaluethatweshouldall,ateverymoment,followthedecisionrule:‘dothatactionwhichproducesgreatestethicalvalue’.Wemaynotknowwhatactionhasgreatestethicalvalue.Itmaybethatattemptstofollowsuchadecisionrulewouldbehighlycounter-productive.Andsoon.Alimitingpossibilityisthatethicalvalueisgreatestinaworldinwhichnoindividualevenbelievesthatutilitarianismistrue.Inthiscaseutilitarianismwouldbe‘self-effacing’.Itwoulditselfimplythatthebestworldisoneinwhichitisnotbelieved.Thesearenotobjectionstoutilitarianism.Tosaythatutilitarianismisself-effacingisnottosaythatit’sself-undermining.Utilitarianismisnotcommittedtotheviewthatthemorallyrightactionistheoptimalaction–thatis,theactionthatproducesthegreatestwell-being,oreventhat‘optimalaction’isineverysituationeveninprinciplewell-defined.Notonlyisitcon-sistentwithutilitarianismtoadopttheaccountofmoralitywhichIgaveearlier,itisinfacttheaccountthatMillhimselfgave.Oureverydaymoralthinkingisbynomeanswhollyimpartialinthewaytheutilitariantheoryofethicalvalueis.Webelievethatwehavespecialobligationstosomepeoplebecauseoftherelationinwhichwestandtothem–ofkinship,neighbourliness,fellow-citizenship,contract.Webelievethatwehavedutiesarisingfromour‘stationinlife’.Wehaveveryspecificmoralresponsestoourandothers’actionsintheircontext,responseswhicharecertainlynotpulleddownfromanethicaltheory,utilitarianoranyother.Butethicaltheoryneednotclaimthatthoseresponsesarepulleddown.Wecanrecognizethattheystemfromspontaneousmoralfeelings,asdiscussedinsection2.Certainlyutilitarianismiscommittedtothis:whereitcanbeshownthatgeneralwell-beingwouldbeimprovedifagivenwayofdoingthingswaschanged,thenitshouldbe.Thatiswhatismeantbycallinggeneralwell-beingatestofconduct.Anotherethicaltheory,suchascontractualism(seesection9below),wouldhavethesamecommitmentforthetestofconductwhichitfavoured.Ifwecouldneverplausiblyshow,ofanypractice,thatgeneralwell-beingwouldbeincreasedbychangingit,thenutilitarianismwouldhavenocuttingedge–thoughitcouldstillbetrue.Similarly,ifversionsofgenericutilitarianismdifferintheirdistribu-tivestructureortheirviewoftheendsoflifebutdonotdifferinwhatchangestheyrecommend,thenthedistinctionbetweenthemmakesnopracticaldifference.Introducingautilitarianstandardinfactmakessubtlebutveryimportantdiffer-ences,asintroducingstandardsofSCIENTIFICMETHOD(chapter9)madetoourbeliefsabouttheworld.Inneithercase,however,doestheeffectcomethrougha‘linear’derivation,ofmoralrulesfromawelfarefunction,orofspecificscientifichypothesesfromcanonsofscientificmethod.Inbothcasesthenewornewlyrefinedstandardsoperateonanexistingcosmologicalormoraltradition,changingitbutdoingsoinaholisticandconservingway.‘Whereitdoesn’titch,don’tscratch’isQuine’sformulaforthismethodofconservativeholism.Itappliestoethicalmethodtoo.213\nJOHNSKORUPSKICertainlytheclassicalutilitarianssometimestalkedina‘linear’way,buttheyalsohadtheirconservative–holistmoments.MillparticularlyemphasizedthiselementofconservativeholismbecausehewasrespondingtocriticismsofBenthamandotherearlyutilitarianradicalsonpreciselythispoint.Sidgwickthensystematizedthestrat-egyatlength.Allthismustberememberedifwearetoavoidsettingupover-simplecontrastsbetweenutilitarianandotherviewsofmorality.Theattractionofgenericutilitarianismistwo-fold.Inthefirstplaceitseemsper-fectlypossibletoaskwhyoneshouldobeyaparticularsetofmoralrules,saytheTenCommandments.Ifsuchresponsesasthattheyareself-evidentorthattheyarelaiddownbyagodseemunsatisfying,onereasonatleastisthatwewanttoknowthepointorpurposeoftherules.Andoncethequestion‘Whatpurposedoesthisruleserve?’isacceptedasproper,thegenericutilitarian’sanswer,referringustothestandardofgeneralwell-being,isverycompelling.Tobetoldthatadulteryiswrongbecauseofthehurtitcauses,directlyandthroughbreakagesoftrust,isabitmoresatisfying,thoughundoubtedlyalsoagooddealmoredebatable,thantobetoldthatitswrongnessisself-evidentorthatitisforbiddenintheBible.7ThePluralityofValueOnemayaskaboututilitarianism,oranyethicaltheory,whatitscredentialsare.Whatsortofthingcouldshowittobecorrect?Wheredoesitfitontheintellectualmap?Spontaneousimpressionsofwhenanemotionisinplace,stabilizedbyself-examinationanddiscussion,generateinternalcriteriagoverningitsreasonableness–normsofrightfeeling.Thesamegoesforepistemicnorms,whichconsistinjudgementsaboutwhatthereisreasontobelieve–normsofrightbelief.Theseappraisalsareentwinedintoourends.Findingoutaboutancientcultures,forexample–ifthatisyourobjectiveyoushouldpayattentiontovaluablepiecesofevi-dence.Evenmoreobviously,ifyouwanttoplaythepianowell,youshouldbeinterestedinwhatmakesaperformancemusical,perceptive,moving.Butthoughtheyareinthiswayentwinedinourends,theydonottelluswhatourendsshouldbe.Logicorarcheologicalmethodtellsyouwhatevidenceisimportantforgettingatthetruthabouttheancientpast,butneithermethodtellsyouthatthatoughttobeyourobjectiveorthatyououghttopursuethetruth.Internalcriteriatellyouwhatperformanceisnotmerelytechnicallyaccurate,butbeautiful,insightfulorprofound.Buttheycannottellyouhowmuchtimetospendonit,orwhatresourcestoputtoit.Exactlythesame,itwouldseem,mustapplytotheblameworthy–atleastifwethinkofthatasdeterminedsolelybytheinternalcriteriaoftheblame-feeling.Suchcriteriawilltelluswhatis‘morallyright’or‘morallywrong’,butnotwhatreasonwehavetopursuetheformerandavoidthelatter.JustasonecouldwriteabookabouttheTheArtofthePianist,oraboutGoodMethodinArcheology,withoutsayinganythingabouthowimportanttheseactivitiesarecomparedtoothers,soonecouldwriteabookaboutTheArtofMoralGoodness,withoutsayinganythingabouthowimportantitistopursuemoralgoodness.Somephilosophershavearguedasthoughtheseinternalcriteriapropertovarioustypesofactivityandexperiencearetheonlyonesweneed.Theywouldsaythatthere214\nETHICSisnospaceforwhatIhavecalledtheethicalquestiontoarise.IstherereallyaquestiontobeaskedaboutwhyIshouldavoiddoingwhatismorallywrong,shunboringorfrighteningsituations,seektoachievethequalitiesinmyselfthatIfindadmirableinothers?Emotioncanshapeone’sactionthroughthelinksbetweenemotionanddesire,anddesireandaction.Guilt,boredomandfearareemotionsthatgiverisetoadesiretoelimi-nateoravoidtheirobject.Others,suchasaestheticdelight,giverisetotheoppositedesire.Butwhataretheconnectionsatthelevelofreasons?WhyshouldIavoiddoingwhatisboringorblameworthy?Considerthis:ifathingisboring(blameworthy)thenthereisreasonnottodoit.Isthisatruthoflogic?Itdoesnotseemso.Thetruthoflogicisratherthis:ifathingisboring(blameworthy)thenboredom(blamingthepersonwhodoesit)isanappropriate,justified,reasonableresponse.Thelogicalconnectioniswithnormsoffeelingratherthannormsofaction.Doingthingsthatareboringorblameworthyislikelytomakeoneboredorguilty–statesthatmostofusfindunpleasant.Andsurelyonehasreasontoavoiddoingwhatisunpleasant.Thislastpropositionisobviousenough,andmakesthelogicallinkwithreasonsforaction.Butitalsogetsusintotheterritoryofethicaltheory.Foritisanor-mativepropositionaboutreasonsforaction,andwecanaskwhatitssourceis.Butfirstthereisanimportantpointaboutthestatusofthemoraltobeconsidered.ASpecialStatusforMorality?Moralityissurelynotonallfourswithothernormsoffeeling.Onecanbringouttheasymmetry,forexample,inthisway.IfIhonestlyfindamovieveryboringIamquitelikelytojudgeapersonwhodoesnotfinditboringratherunfavourably–perhapsasaprettyuncriticalenthusiast.OrImaybeledtoquestionmyowntasteinstead.ButevenifIremainconvincedthatthemovieisboring,Icanhardlyarguethattheuncriticalenthusiasthasreasontoavoidit.Heafterall,uncriticalfellowthatheis,willenjoyit.Canwetakethesamelineaboutimmoral,asagainstuncritical,people?Ifwedid,wecouldstillformanunfavourableassessmentofsomeonewhofeltnoguiltaboutdoingsomethingblameworthy–findinghimshamelessorblameworthy.Butwecouldnotclaimthathehadareasontorefrainfromtheimmoralact.Thereisnosymmetryhere.Ishouldavoiddoingsomethingmorallywrongbecauseitismorallywrong,notbecauseitwillmakemefeelguilty,whichisunpleasant.Evenifitdoesnotmakemefeelguilty,Istillhavereasonnottodoit.Consideraconnoisseurofsingingwhoseesthataparticularpieceofsingingisbeautifulbutnolongergetsanythingoutofit–theuninvolvedconnoisseur.Hehasnoreasontoattendtheperformance.Comparetheuninvolvedmoralconnoisseur–whoseeswhatisrightandwrongbutnolongerfeelscommittedtoit.Hestillhasreasontodotherightandavoidthewrong.Itdoesnotdependonhowinterestedhefeels.Wewanttomaintainthisethicalprinciple:IfXismorallywrongthenthereisreasonnottodoX.Andwedonotthinkitstruthdependsontheholdingofsomefurtherconditionwhichmayormaynotobtain.Forexample,whetherthereisreasonformenottodothemorallywrongdoesnotdependonwhatmyobjectivesare,orwhetheritwillbeunpleasant,orevenonwhetheritwilldegrademycharacter.Onemayputthisbysayingthatmoral215\nJOHNSKORUPSKIreasonsarecategorical,sotheethicalprinciplemaybecalledtheprincipleofmoralcategoricity.Itisbynomeansuncontroversial.Ithasbeeninfluentiallydeniedby,amongothers,PhillipaFootandBernardWilliams(Foot1978:chs7–14;Williams1981:ch.8).Onecoulddenyitonthegroundsthatmoralreasonsarenotcategorical,ormorestrongly,onthegroundsthatnoreasonsare.Butifweendorseitwemustconcludethatmoralnormscannotbegroundedsolelyinspontaneousblame-feelings.Theymustalsoberesponsivetoethicalconsiderations.Bothsidesinthedisputeaboutmoralcategoricitymustgivesomeethicalaccountofwhenitisthatonehasreasontoact.Twoideascanbeputforwardbythosewhodenymoralcategoricity:(1)Youhavereasontopursuewhatisgoodforyouandavoidwhatisbadforyou.(2)Youhavereasontodowhateverwillpromotetheobjectivesyouhave.Reasonsforactingmorallywillthendependonwhatisgoodforyouoronwhatyourobjectivesare.Butwheredotheseideascomefrom?Theyseemtobeproductsofratio-nalreflectionalone,justasmuchasisanyotherethicaldoctrineaboutreasonsforaction.So,aswellasthewholediverserangeofemotion-basedvaluesthataredataforethicalreflection,thereseemstobesomethingelse,fromwhichitstems:practicalreason.8PracticalReasonPracticalreasonconcernsreasonstoact,astheoreticalreasonconcernsreasonstobelieve.Sincewehaverepeatedlydistinguishedbetweenreasonstoact,tobelieveandtofeel,shouldwenotexpectathirddivisionofreason,dealingwithreasonstofeel?Itmightbecalled‘aestheticreason’.If‘practical’and‘theoretical’reasonweresimplylabelsforthedispositiontomakejudgementsaboutreasonstoactandreasonstobelieve,atripartitedivisionlikethatwouldberight.Butnosuchtermas‘aestheticreason’isincommonuse.Thereasonisthat‘practical’and‘theoreticalreason’donotjustlabeldispositions;theyinvokeaviewaboutthesourceofthosedispositions,theviewthatjudgementsaboutreasonstoactandreasonstobelieveareexercisesofreason.Sothenotionof‘aestheticreason’limps.Forthoughweconstantlymakejudgementsaboutwhatthereisreasontofeel,thesejudgementsarenotdeliverancesofreason.Tosaythattheywerewouldbetosaythattherearerationallyevidentprinciplesspecificallyconcernedwithreasonstofeel.Simi-larly,practicalreasonexistsiftherearerationallyevidentprinciplesyieldingreasonsforaction.Andtheoreticalreasonexistsiftherearerationallyevidentprinciplesyield-ingreasonsforbelief.But,surely,inthecaseoffeelingtherearenosuchprinciples.Wemaybereasonableinourfeelings,orunreasonable,butitisnotrationalinsightintoprinciplesofrightfeelingthattellsusso.Insofarasreasoncomesin,itdoessoinitstheoreticalguise,throughrationalassessmentofthebeliefswhichourfeelingspre-216\nETHICSsuppose.Pascalsaidthatthehearthasitsreasonswhichreasonknowsnothingof(‘Lecoeurasesraisonsquelaraisonneconnaîtpoint’:Pensées:iv,277).Hewasspeakinginadifferentcontextbuthisaphorismaptlyfitsthepresentpoint.Butistheresuchathingaspracticalreason?Wedistinguishbetweenreasonableandunreasonablefeelings,eventhoughtherearenoaprioriprinciplesofjustifiedfeeling.Sodoesthedistinctionbetweenreasonableandunreasonableactionrequirethattherebeaprioriprinciplesofjustifiedaction?Cometothat,doesthedistinctionbetweenrea-sonableandunreasonablebeliefsrequireaprioriprinciplesofjustifiedbelief?Weshalltouchonthisagaininsection11.Forthemomentwesimplynotethecontemporarytendencytogroupreasonstoactwithreasonstofeel,ratherthan,aswastraditional,withreasonstobelieve.Itassumes,rightlyorwrongly,thatwhereasrationalbeliefdoespresupposerationallyevidentprinciplesofreasoning,rationalactiondoesnot.Idea(2)intheprevioussectioniscloselyrelatedtothis,oraversionofit.Itsaysthatapersonhasreasonsforactiononlyinrelationtohisobjectives.Theonlyconceptofareasonforactingthatwehaveisonethatmakesareasonarelationbetweenanobjec-tive,anactorandanaction,thus:thereisreasonforapersonPwhoseobjectiveisOtodoX.ThatwillholdwhenP’sdoingXisawayofbringingaboutO.HowstrongthereasoniswilldependonhoweffectivedoingXis–howprobableitmakestheachieve-mentofO,andhowmuchitimpedestheachievementofP’sotherobjectives.Thismaybecalled‘instrumentalism’(becauseitsaysthatreason’sonlypracticalfunctionistoselectappropriateinstrumentsforachievingobjectives).Instrumentalismdeflatesprac-ticalreasonintoinstrumentalreason,butitdoesnotdenythattherearetruereason-statements–onlythattherearetruecategoricalones.Toholdthattherearetruecategoricalreason-statementsistotakeamoreambitiousviewofpracticalreason.AmongthosewhoaffirmthemoreambitiousviewthemostinfluentialbyfarhasbeenKANT(chapter32).Heheldthatsomereasonsforactingarecategorical–inhisterminology,categoricalandnotmerelyhypotheticalimperatives.Healso,veryinflu-entially,heldthattheonlycategoricalreasonsaremoralreasons.Butbeforeconsider-ingKantletusnotesomeotherethicaltheoriesthatmustalsoacceptthecategoricalnatureofpracticalreason.Areasonforanactortodosomethingiscategoricalifitholdsirrespectiveofthatactor’sobjectives.Thiscutsalinebetween(1)and(2).Foraccordingto(1)youhavereasontopursuewhatisgoodforyou,irrespectiveofwhetherthatisyourobjective.Thepersonwhoholdsthatthisistheonlycategoricalreasonforactingistherationalegoist.Becauseactingselfishlymightbebadforyou,rationalegoistsneednotrecommendselfishnessintheusualsense.Theirpositionisalsodistinctfromtheinstrumentalist’s.Forwhatismeantbyaperson’sobjectives?Thestrictlyinstrumen-talistviewisthataperson’sobjectivesarethosethatapersoninfacthas,andwhicharebasicratherthanderivedfrommorebasiconesthatthatpersoninfacthas.Wearenottalkingaboutbasicobjectivesthatpeoplewouldhaveiftheyhadfullrationalinsight,orwantedwhatreasontellsthemtowant,andsoon.Thesenotionsarenotavailabletotheinstrumentalist.Onemightarguethatinstrumentalismitselfmustendorseatleastonecategoricalprinciple–onpainofcollapsingintopurescepticismaboutpracticalreason:‘Yououghttodowhateverwillpromoteyourobjectives,irrespectiveofwhattheyare’.Itmustindeedendorsethatprinciple,butitisanotherquestionwhetherithastoseeitasasubstantialdeliveranceofreason(seesection11).217\nJOHNSKORUPSKIIfitcandenythatitis,thenitcaninthatsensebesaidtorejectallcategoricalprinci-plesofpracticalreason.Utilitarians,likerationalegoists,mustacceptacategoricalprincipleofpracticalreason.Theycannotbeinstrumentalists,fortheyarenotjustsayingthatifone’sobjec-tiveistopromotegeneralgoodthenonehasreasontopromotegeneralgood.Theirsisafundamentalthesisaboutethicalvalue.Itsaysthatonehasreasontopromotegeneralgoodwhetherornotonewishesto.Thewell-beingofeveryindividualisasourceofcategoricalreasonsforeveryindividual.Thecontrastbetweencategoricalandinstrumentalconceptionsofpracticalreasonisamajorwatershedinethics,andsinceutilitarianismissooftencrudelylumpedwithinstrumentalistnotionsofreasonitisworthstressingthatcogentutilitarianism,onthecontrary,likeKantianism,fallsonthecategoricalandnottheinstrumentalside.ThatwasclearertoMillthanBentham,andclearertoSidgwickthanMill.ButletusnowturntoKant’sethicaldoctrine.Kantthoughtthecompletesetofmoralprinciplescouldbeestablishedbymeansofatestwhosesoundnessisitselfderivedfrompurereason.Thistestisthecategoricalimperativeproper:Actasifthemaximofyouractionweretobecomethroughyourwillauniversallawofnature.Whenitisappliedtoanymaximofactionitissupposedtodetermineinaclearcutway–andsupposedlywithoutappealtoanyempiricaldata–whetherthatmaximisorisnotacategoricallybindingmoralprinciple.ForKant,themoralsimplyistherational.SomuchinKant’sethicsismomentousanddeepthatitiseasytoforget,ortoallowoneselftoobscure,thequiteincredibleambitiousnessofthiscentraldoctrine,whichisthebackboneofthetheory.Itremainsincredible,evenwhenwenoticethevariouswaysinwhichKantthinks‘anthropological’factsabouthumansandtheircircumstancesshapethedetailsofmorality.Oneshouldnotglossoverthisincredibility,becauseKant’sgreatnessasaphilosopher,whencombinedwithit,hasbeenruinousformoredefen-sibledoctrinesofpracticalreason,whichhavesufferedcollateraldamagefromthecol-lapseofKant’sover-ambitiousproject.BecausereasoncannothavethemonopolyindeterminingrightandwrongwhichKant,thegreatdefenderofreason,soughttogiveit,itisthoughtthatitcanplaynopartatall.Anextremeformofethicalrationalismreboundsintoextremeformsofethicalirrationalism.HenceitremainsimportanttodrawupanaccountofwhatinKant’sethicsmustbejettisonedandwhatcanberetained.9TheCategoricalImperativeandContractualismCriticsonallsidesinthenineteenthcentury(amongothers,Hegel,Mill,SchopenhauerandNietzsche)pointedoutthatthoughKant’ssubstantivemoralteachinghasaverydefiniteanddistinctivecontent,hiscategoricalimperativeistooformaltoyieldanysuchsubstantiveresults.Considerthemaxim,‘Breakyourpromisewhenitistoyouradvantagetodoso’.Kantsuggeststhatitwouldbeself-contradictorytosupposeita218\nETHICSuniversallaw,becauseifeveryonecompliedwithittherewouldbenosuchthingas‘promising’.Asitstands,thisiswrong:togetacontradictiononemustaddcontingent,non-formal,assumptions,suchasthatitisgenerallyknownthateveryonebreakstheirpromisewhenitistotheiradvantagetodoso.Butthisassumptioniscontingent.ReasonandAutonomyKantalsorequiresthatweshouldbeabletowillthatthemaximbeauniversalLAWOFNATURE(pp.645–8),andthishasseemedamorepromisinglineofthought.CanIreallywillthateveryoneshouldbreaktheirpromisewhenitistotheiradvantagetodoso?Itiseasytobemisledhere.ForKant,thewillinquestionmustbeapurelyrationalwill:itissimplythemovingforceofpurereason.Itcannotappealtoendsthatwehappen,con-tingently,tohave.Kantcertainlytalksasthoughthewillisdeterminedbythemereideaofuniversallaw,andthispartofhisdoctrinedoesfalltohisnineteenth-centurycritics.Buthealso,lessexplicitly,takesitthatthereareendsthatpurereasonrequiresustohave:persons,thatisrationalbeings,aretheonlyendscertifiableasintrinsicendsbypurereasonalone.Thisleadshimtoanother,formanypeopledeeplyresonant,versionofthecategoricalimperative:Actinsuchawaythatyoualwaystreathumanity,whetherinyourownpersonorinthepersonofanother,neversimplyasameans,butalwaysatthesametimeasanend.Resonantasthisis,itisnotwhollyclear.Whatisittotreataperson,arationalbeing,asanend?Oneview,whichseemstobeKant’s,isthatitistorefrainfromactionswhichimpedeothersoroneselffromtheexerciseofrationality.Now,forKant,rationalityismoralfreedom(thenotionwediscussedinsection3)or,inhisword,autonomy.Thisisoneofthemostimpressive,aswellasoneofthemostcentral,elementsofKant’smoralphilosophy.Hisviewisthatreasonitselfrequiresthatweshouldrespect,notinfringeupon,theautonomyofbeingswhichcanbeautonomous.Kant’slinkbetweenreasonandautonomyisasubstantialclaim,butattemptstoextractitfromconjuringtricksonstatementslike‘reasonitselfrequiresthatreasonberespected’shouldberesisted.Wedohaveanotionofrationalautonomy(asemergedinsection3)andwedo,Ibelieve,thinkthatrationallyautonomousbeingshavespecialmoralclaimsinvirtueoftheirrationality.Butwedonotthink,asKantiandoctrinewouldrequire,thatonlyrationalbeingshavemoralclaims,andwedonotthinkthattheironlyclaimonusisrespectfortheirrationality–inotherwords,abstentionfromanythingthatwouldimpedeordiminishtheirmoralfreedom.Inanycase,Kant’sfor-malismpreventshimfromgivingdefinitecontenttothenotionofpracticalrationality.Onhisaccount,rationalityconsistsinseeingwhattheuniversallawofreasonrequires,andthisuniversallawsaysonlythatoneshouldrespectrationality:nothinginthistellsuswhatrespectforrationality,andhencerationality,requires.SointheendthechargeofemptinesslaidagainstKant’scategoricalimperativestands.Themissingingredientisthenotionofanindividual’sgood.ItseemsthatKantthoughtthisnotioninappropriately‘anthropological’forapurelyrationalend.However,thenotionofanindividual’sgoodisnotitselfanthropological–onlywhenwegiveanaccountofwhatthegoodconsistsinforahumanbeingdowedescendto219\nJOHNSKORUPSKIanthropology.Ifweequippurereasonwiththeabstractnotionofindividualgood,wecanseevariouswaysofcombiningitwiththedoctrineofrationalwill.Forexample,ifrationalityitselftellsrationalagentstopursuetheirgood,theycannotrationallywillthatamaximshouldbeauniversallawofnature,wheresuchalawofnaturewouldbetotheirdisadvantage.Orwecouldtakerespectforpersonstomeanconcernfortheirgood,readingthecategoricalimperativeastheinjunctiontotreatthegoodofallpersonsassomethingtobepromotedimpartially.Thegoodofindividuals,conceivedimpartially,becomestheobjectofrationalwill,sothatamaximcouldbewilledtobeauniversallawofnature,andthusadoptedasamorallaw,onlyifuniversalcompli-ancewithitwouldpromotethegeneralgood.Eitherway,wedroptheKantianideathatrationalityitselfistheonlyrationalend,andsubstitutethenon-Kantianideathatthegoodofindividualsistheonlyrationalend.Soeitherofthesedevelopments,towardsrationalegoismor(ifgoodisidentifiedwithwell-being)towardsutilitarianism,isadecisivebreakwithKant.Letusconsideralittlefurtherthefirstidea,whichplacesthecategoricalimperativeagainstabackgroundofrationalegoism.Canarationalegoist‘will’that‘Breakyourpromiseswheneveritistoyouradvantage’shouldbeauniversallaw?Well,arationalegoistthinksitisauniversallawofreason.Butperhapsheorshecannotwillittobeauniversal‘lawofnature’,orsomethingwithwhicheveryonealwayscomplies.Itisnotthatsuchuniversalcompliancewouldbeinherentlyself-contradictory,asKantthought;rather,itisthatitlookslikelytobetohisdisadvan-tage.Andifitislikelytodisadvantagehim,hecannotwillit,ordoanythingtobringitabout.ButsupposewemodifyKant’stest,asfollows:Acceptasamoralprincipleonlythatmaximwhichyoucanwilltobeauniversallawofnature.Arationalegoistwhoendorsesthistestofmoralprincipleswillnotaccepttheprinci-pleofbreakingpromiseswhenitisadvantageoustodosoasamoralprinciple.Onthecontrary,sincehecanwillthatpeopleshoulduniversallycomplywiththeirpromises–itwouldbetohisadvantageiftheydid–heaccepts‘Keepyourpromises’asamoralprinciple.ThismodificationofKant’stestformoralprinciplesputsusontheroadtocontrac-tualism.Aformofcontractualismwhichisquiteclosetoitisthesimpleviewthatmoralrulesarethosewhichrationalegoistscould‘agree’to–promoteaspublicrulesofconduct.Anothermodificationreplacesrationalegoismbyinstrumentalism–moralrulesarethosetowhichinstrumentallyrationalagentscouldagree.ThismovesanotherstepawayfromKant,sinceitappearstodroptheideathatthereareanycat-egoricallyrationalprinciples,egoisticorother–but,bythesametoken,itappearstosomeasattractivelyhard-headed.Thoughtheseformsofcontractualismhavebeeninfluentialinrecentmoralandpoliticalphilosophy,theyhaveastraightforwardandprettyunappetizingcorollary:onehasreasontoobeyamoralprincipleonlyifitistoone’sadvantagetodoso,orifitpro-motesone’sobjectivestodoso.Theegoistorinstrumentalistconstructionofmoralitycangivemereasontoseek,inwhateverwayIcan,toinfluencepeopleintoaccepting220\nETHICSvarious‘moral’precepts,butitdoesnotgivemereason(evenincaseslikethePRISONER’SDILEMMA(pp.393–5))toactinaccordancewiththosemoralpreceptsmyselfwhenitistomyadvantage,orpromotesmyobjectives,nottodoso.Ifoneacceptstheprincipleofmoralcategoricityconsideredinsection7,suchaconclusionmustberejected,andwithitthetheorythatgivesrisetoit.AnotherrecentformofcontractualismiscloserinspirittoKant.TheKantianideaoftreatingpersonsasendscanbereadasrequiringthatwedonottakeunfairadvan-tageofthem,andacontractualistviewofmoralitybasedonthisreadingcantreatmoralrulesasthosewhichcouldbewilledbyareasonableperson–onewhopursueshisorherwell-beingorobjectiveswhileacceptingthatallmusthaveanequallyfairchancetodoso.Therulesofmoralityareruleswhichnoreasonablepersonwouldreject.ThomasScanlon(1982)providesaninfluentialdefenceofthisview.IwillcallitKantiancontractualism,thoughitsaffinitytoKant’sethicsisamatterforscholarlydebatewhichwillnotbepursuedhere.AComparisonSinceKantiancontractualismisoftenpresentedasanalternativetoutilitarianism,itisworthconsideringthesimilaritiesanddifferencesbetweenthetwoethicaltheories.Andheretheproperbasisforcomparisoniswithgenericutilitarianism,ratherthanclassicalaggregativeutilitarianisminparticular.Bothrecognizeimpartiality,or‘fairness’,asacategoricalethicalprinciplelocatedoutsidemorality–sotheybothdivergefrompureinstrumentalismandpureegoism.Moralprinciples,fortheKantiancontractualist,arethosecompliancewithwhichwouldadvantageallinafairmanner,amannerwhichrespectstheirdignityaspersons.Inparticular,theywouldbethesub-groupofsuchprincipleswhichwecouldagreetosanctionbypenaltiesofblamefornon-compliance.Nowshouldtheseprinciplessimplyestablishconstraintsoffairness,ratherminimallyconceived–alevelplaying-fieldforcompetitionamongindividualswhopursuetheirownends?OrdoesKantiancontrac-tualismleadustosomethingstronger–amaximizingprincipleofjusticesuchasRawls’sdifferenceprinciple?Theupshotofthisstrongeroptionwouldbeatestofconductcoextensivewithoneoranotherversionofgenericutilitarianism.Kantiancontractualistswouldsaythattheirwayofsettingupthetest,byreferencetoahypotheticalagreement,islikelytoleadawayfromtheclassicalutilitarians’viewsofimpartiality,whichallowsomeindividualstobedisadvantaged‘unreasonably’forthesakeofoffsettingadvantagestoothers.Theymayberightinthis.Butitremainsdebatablewhethertherightobjectiontoclassicalutili-tarianism’saggregativeprincipleisthatitisunfair.Itmaybeunreasonable,inthewaythatapplyingsimilaraggregativeprinciplesofdistributionwithinone’sownlife(allow-ingsomeperiodofone’slifetofalltolimitlesslyincreasingdepthsofmiseryforthesakeofcompensatingadvantagesinotherperiods)mayseemunreasonable.Butinthelattercasetheunreasonablenessisnotconnectedwithunfairness.OnthewholeKantiancontractualismlooksmuchclosertogenericutilitarianismthantotheegoisticorinstrumentalistformsofcontractualismwithwhichwestarted.Theutilitarianprincipleisstrictlyagent-neutral,inthatitassignsvaluetothewell-being221\nJOHNSKORUPSKIofanindividualirrespectiveofthatindividual’srelationtotheagentwhoisdeliberat-ing.Kantiancontractualism,likewise,appealstoanagent-neutralprincipleoffairness.ThereasonIshoulddowhatismorallyrightisthatitwouldbeunfairnottodoso.Whetherornotthenotionoffairnesscanultimatelycarrythisweight,itisatleastclearthatthereisnopretencehereoffoundingmoralityorjusticeonpureegoismorinstru-mentalism.Noristhereanyattempttogetasubstantivetheoryoftherightfrompurelyformalconsiderations,asKantsometimesgivestheimpressionoftryingtodo.BothgenericutilitarianismandKantiancontractualismarecommittedtotheviewthatthereareprinciplesofpracticalreasonthatarecategoricalbutnotmerelyformal.10Reason,MoralityandReligionSomepeoplefeelthattodistinguishbetweenreasonandmorality,asutilitarianismandcontractualismdo,istodrawaninvidiousline.Moraljudgements,theyfeel,areratio-nallyevidentintheirownright.Morallifehasitsowninternalrationality.Theideathatcriteriaofpracticalreasonstandoutsideitalienatesusfromit.Iftheutilitarianorthecontractualistviewalienateusfrommorality,thatisalam-entableconsequencebutnotanobjectiontotheirtruth.Thephilosophicalquestionhastocentreonwhethermoraljudgementsareindeed,contrarytoutilitariansandcon-tractualists,evidentintheirownright.Thattheyaresoistheviewofthemoralintui-tionist.Iusethistermheresimplytonametheviewthatcertainmoraljudgementsarerationallyevidentintheirownright,withoutassuminganyparticularepistemologicalaccountofwhatitisforsomethingtoberationallyevident.Wehavealreadyseenthatjudgementsaboutwhatisadmirableorblameworthy,likeotherjudgementsaboutwhatitisreasonabletofeel,haveaninternaldis-ciplinebasedonreflectionaboutone’sspontaneousfeelingsanddiscussionwithothers.Butthisinternaldisciplinedoesnotitselfconnectthemwithreasonsforaction;stilllesscoulditunderwritetheprincipleofmoralcategoricity(seesections2and7).Forthemoralintuitionist,thetruthorotherwiseofmoraljudgementsisknowndirectlybypurelyrationalreflection.Thisinitselfstilldoesnotshowwhythefactthatsomethingismorallywrong–which,bytheintuitionist’shypothesis,Iknowthroughpurerationalreflection–givesmereasonnottodoit.Butiftheintuitionistgivespurerationalreflectionasbigaroleasthis,heishardlyuppingthestakesverymuchifheaddsthatIalsoknowthatIhavereasonnottodowhatismorallywrongbypurerationalreflection.Utilitariansorcontractualistswhogranttheprincipleofmoralcategoricitycannotbeasdirectaboutitasthat.Theymust,rather,derivethecategoricityofmoralreasonsfromtheethicalprinciplesthattheyholdtobeprimitivelycategorical–thegeneralgood,orfairness.Moralreasons,theycansay,arethosereasonsforactionthat,givenhumancircumstance,peopleshouldobserveandshouldincurblameforfailingtoobserve.Iftheyareconservativeholists(section6)theywillnotclaimtoderivetheminalinearwayfromtheirethicalcriteria,butwillfeelfreetogroundtheminourspon-taneousdispositionstoblame.Inotherwords,theywillacceptthattheguidanceforactionandevaluationgivenbythesedispositionsislegitimateintheabsenceof222\nETHICSawell-madecaseagainstit;buttheywillinsistthatitiscorrigiblebyawell-madecasebasedontheirpreferredethicalcriterion.Theircriticismoftheintuitionistwillbethatthelatterconfusesspontaneousjudge-mentsaboutwhatisadmirableandblameworthywithpurelyrationalintuitions.Thatisapurelyphilosophicalcriticismandtomymindaverypowerfulone.IfIexaminemyresponsesIfindemotionsthatgiverisetojudgementsabouttheadmirableandtheblameworthy,butIfindnodetailedpurelyrationalintuitions.Itisnotreasonthatdirectlytellsme,inconcretecircumstances,whethertotellalietosaveafriend.Andtheutilitarianorcontractualtheoristcangoontodrawasocialconsequence.Whenaspontaneousresponseofadmirationandblamegettransmutedinthiswayintoasup-posedrationalintuition,heorshemaysay–withnoappealallowedtoahighercourtofethicaljudgement,nostandardofrationalcriticismindependentofourfeelingsofadmirationandblame–thenanimportantsourceforregeneratingmoralityislost.Wherethosefeelingsstronglyclusteraroundsomething,forexamplefeelingsofblamearoundhomosexuality,orfeelingsofadmirationaroundcertainkindsofmaledomi-neering,wecannotdetachfromthosefeelings,standback,andaskwhatiswrongwithhomosexualityorhowmuchwereallywantthatkindofleadershipofthepacktobeadmired.Iamnotsayingthatthisisacompellingcriticism,sinceintuitionistscanreply,firstly,thatthequestionagainstthemhasbeenbeggedwhenwhattheyclaimtoberationalmoralintuitionsaretreatedasfeelingsofadmirationandblame;and,secondly,thattheseintuitionscanbecriticizedfromwithin,whentheycomeintoconflictwitheachother,withoutanyappealtoan‘external’ethicalcriterion.Certainlythequestionwhethertherearesuchrationalintuitionsisphilosophicallythebasicone.Thecaseagainstissimplythatreasondoesnotgiveusdetailedmoralresponses.Ofcoursethedetachinglanguageoftheutilitarianorthecontractualisthasitsowndangers,ofestrangementfromourmorallegacyandmanipulativesocialengi-neering.Ideallywewouldhaveastandardofrationalcriticismwhichyetsomehowlaywithinmorality,anddidnottreatitasanobject.ThatisoneofthemostpowerfulattractionsofKant’sethics.Thecategoricalimperativepromisedtogiveastandardofrationalcriticismwhichfulfilledthisideal.Butitturnedouttobeemptyunlessitsmuggledinethicalprinciplesoffairnessorgeneralwell-being,orcovertlyconvertedmoralfeelingsintodeliverancesofpurereason.Humanreasonandhumanmoralfeelingscouldbefullyintegratedifbothflowedfromsomesourcewiserthanthehuman.Hereisaroutefrommoralitytoreligionwhichhasrealpowerformanypeople.IfbothreasonandmoralfeelingareimplantedinusasresponsestothemorallawlaiddownbyGOD(chapter15),thenwecouldsafelystaywithinthecircleofmoralitywithoutbeingforcedintoachillyexternalstandpointinwhichreasoncouldcomeintoconflictwithsentimentstransmittedbytradition.Buttwogreatdifficultiesstandinthewayofthisthought.Thefirstisthatanyargu-mentofthisshape–somethingwhichisethicallyorotherwisedesirablecanbeguar-anteedtoexistonlyifGoddoes,soGodexists–isinvalid.Wehavenoreasontosupposethattheworldisasitisdesirablethatitshouldbe.Someindependentargumentisneededtoestablishtheexistenceofawiseandbenevolentdivinity.Theseconddifficultyhastodowiththeautonomyofgood.InPlato’sEuthyphro(1961:10a)Socratesasks‘whetherthepiousorholyisbelovedbythegodsbecauseitisholy,orholybecauseitisbelovedofthegods’.Onecansimilarlyaskwhetherthe223\nJOHNSKORUPSKImorallygoodissobecauseitiscommandedbyGod,orcommandedbyGodbecauseitismorallygood.Notafewreligiousbelieverscleavetothefirstanswer.Sodosomeathe-ists:itisimplicitlyassumedinNietzsche’sassertionthat‘Godisdead’(Nietzsche1974:section125),thoughNietzscheplayedwiththefantasythat,afterthedeathofGod,newvalues,beyondmoralgoodandevil,couldbecreatedbytheimpositionofsuffi-cientlyforcefulhumanwill.Butthisvoluntarism,theologicalorNietzschean,isincompatiblewiththecat-egoricityofmorality.IfmoralityisthecommandofGod,itmakesperfectlygoodsensetoaskwhatreasonIhavetoobeyit.Punishmentandrewardintheafter-life,itmaybesaid,givemeaself-interestedreason.Toofferthatastheonlyansweristoproposeaprettyshabbyphilosophyoflife;butsettingasideitsshabbinessconsideritslogic.Ihavethatreasonbecausemywell-beinggivesmereasontoact.Butthatlocatescategoricalreasonoutsidemorality.Wemuststillacceptacategoricalreason,independentofthecommandofGod,andonthisapproachwehavesimplysucceededinlocatingitinegoismratherthanmorality.ThisistheGod-assistedtransitionfromrationalegoismtomorality.Andnotethatwehaveattachednomeaningto‘Godisgood’otherthan(perhaps)‘GodobeysGod’scommands’.ThesamelogicalpointsalsoapplytomotivesforobeyingGod’slawthatarenotshabby–loveofGod,aswellasfearofhell.Theotherresponse,inwhichGod’sgoodnessconsistsinhisrecognitionofthegood,seemsmuchsounder.Itrecognizestheautonomyofgoodandingeneraltheautonomyofreasons.ThetheistcanarguethatabenevolentGodhasimplantedourMORALSENTIMENTS(pp.716–17)inusasreliableguidestothegood,andthatgivesthetheistawayofbeingreconciledwiththem.Itisconsistentwiththistoholdthatthegoodissomethingautonomous,independentofGod’swill,andindependentlyrecognizablebyus.ThiswastheviewofKant,oneofthegreatproponentsoftheautonomyofreason:EventheHolyOneofthegospelmustfirstbecomparedtoouridealofmoralperfectionbeforewecanrecognizehimassuch...wheredowegetourconceptofGodasthehighestgood?SolelyfromtheIdeaofmoralperfectionwhichreasontracesapriori.(Kant1964:29)Ifthegoodiscategoricalitisautonomous,andifitisautonomousthenitistrans-parenttorationalbeings.God’spowerandknowledgemayallowhimtoimplantreliablerulesofconductinus,butitdoesnotgivehimgreaterinsightintoitsaprioristructure.Oncereasonhasbeenisolatedinthisway,aquestionremainsaboutitspowertoinspire.OnecanloveGod,butcanoneloveimpartialpracticalreason?Doesitmakeonelovevirtue?Canoneriseaboveself-interestandlimitedsympathywithoutbecom-ingdesiccated,ifonehasnoemotionalmediuminwhichtodoit?Virtueisnaturallyloveable,becauseitisbasedonwhatwespontaneouslyadmire.Moralexamplecanmakeuslovethegoodofothers.Butisitnotlikelythatonlyareli-giousorspiritualperspectivecansolidlyandreliablymakeusseeothers’well-beingasofnomoreorlessaccountthanourown?Well,evenifthethoughtthatthereisnothingspecialaboutmeisgiventomebyreason,itcanstillbeamovingandliberatingex-perience.Ifonewishestocallitaspiritualexperience,sobeit.Butitneedinvolvenotheisticframework.224\nETHICSWemustturnfinallytothepromisedmeta-ethicaldiscussion:ifjudgementsaboutreasonsareautonomous,whatjustifiesusinmakingthemandwhatmakesthemrightorwrong?11Meta-ethicsThejudgementswemakeaboutourreasonstoact,believeorfeel,maybecalled‘nor-mative’judgements.Itisilluminatingtokeepthemallinmind,eventhoughourconcernismeta-ethical.Forquestionswhichareoftenaskedaboutthestatusofethicaljudgementsinparticularcanusuallyberaisedingeneral,aboutallofthem.Onecanaskaboutallnormativejudgements,notjusttheethicalones,whethertheycanbeassessedastrueorfalse,correctorincorrect;andiftheycanbe,whatmakesthemso,andhowweknowthemtobeso.Framedinthisway,thequestionsgototheheartlandofphilosophy.Onlytherough-estsketchmapofitsgeographycanbeprovidedhere.Considertobeginwith,then,thisverycommonquestion:arenormativejudgements,orcantheybe,‘objective’?Thewordisvirtuallyirresistible–andvirtuallyguaranteedtosowconfusion.Itsoundsasthoughitsoppositeoughttobe‘subjective’;butassoonaswearepressedonthematterwefindthatwereallyhavenonotionofwhata‘subjectivejudgement’mightbe.Whatcausesourembarrassmenthere?AjudgementmusthaveaCONTENT(pp.843–5):onejudgesthatsomethingisso;onecanalsosupposeittobeso,wonderwhetheritis,hopethatitis,andsoon.Tojudgethatitissoistojudgethepropositionthatitissocorrect.Inotherwords,ifthereisnoproposition–nocontentofjudgementtobejudgedcorrectorincorrect–thereisnoscopeforjudgement.Theverynotionofjudgementrequiresadistinctionbetweentheactofjudginganditscontent,thepropo-sitionwhichisjudgedcorrectorincorrect.Butthatdistinctionisavailableonlyifwearealsoabletoseparateinprinciplebetweenwhatseemstothejudgertobeso,ortobecorrect,andwhatisso,orwhatiscorrect.Itmustbepossiblethatwhatseemstothejudgertobesoisnotso.Andthispossibilitysuppliesobjectivity:anobjectivediscourseisoneforwhichitcanbemaintained.Ifitcannotbemaintained,thereisnoscopeforobjectivity–butneitheristherescopeforjudgement.Whatweshouldreallybeasking,whenweaskwhethernormativejudgementsareobjective,issimplywhethertheyaregenuinejudgementsatall.ItwasjustthisquestionthatDavidHUME(chapter31)answeredinthenegative,forthemoralcase,whenhesaidthat‘Morality...ismoreproperlyfeltthanjudg’dof’(Hume1978:470).Humeallowedforpropositions,andhencejudgements,suchas‘Allfathersareparents’,whichinhisterminologymerelyexpress‘relationsofideas’.Wemaysaythattheyaretruebydefinition.Thesecasesapart,however,hetookpropositionstobefactual–theirtruthdependedonwhetherornottheycorrespondedtofacts.(Thenotionoftruthbydefinitioncertainlyrequiresverycarefulexamination,butweneednotdelveintoithere:iftherearenosuchtruthsHume’sthesiswouldsimplybethatallpropositionsarefactual.)ItisnoteasytofitnormativepropositionsingeneralintoeitherofHume’scategories.Anormativepropositionlookslikethis:225\nJOHNSKORUPSKIIfpthenthereisreasontobelieve,doorfeelso-and-so.Forexample:(1)IftheavailabledataarededuciblefromhypothesisHandnootherhypothesisavailabletousissimplerthanHthenthereissomereasonforustobelieveH.(2)IfactionApromotessomeone’swell-beingthenthereisreasonformetodoA.Itisnotplausiblethatpropositionsofthiskindaretruebydefinition.(Whatdefinition?Ofwhatwordorwords?)Soifpropositionsareeithertruebydefinitionorfactual,theymustbefactual–ornotgenuinepropositions,despiteappearances.Humehimselfthoughttheycouldnotberegardedasfactualandhewasclearthatthisconclusionappliedtoallsupposednormative‘propositions’,notjustethicalones.Thereareanumberofroutestothisconclusion.Thesimplestisthesheerimplausibil-ityoffittingthecorrespondencetheoryoftruthtosuchpurportedpropositionsas(1)and(2).Towhatfact,inwhatdomainofreality,wouldthetruthof(1)correspond?Isitsomefactofmeta-orultra-physics,beyondthefactsofphysics?Thesamequestionappliesto(2).ThefactthatdoingAwouldpromotesomeone’swell-beinggivesmereasontodoA:arewetoregardthatasafurtherfact?Asomewhatexpandedversionofthislineofthought,whichwemaycall(withasidewaysglanceinKant’sdirection)thethesisoftheautonomyofnorms,stressesthatcognitionisoffactsandreasons,sothatthedualismoffactandnormisinherenttoit.Facts‘intheworld’givereasons;butthiscannotitselfbeafactintheworld.Nofactcancorrespondtoanormativepro-position,aboutwhatthereisreasontobelieve,doorfeel.ThepointcanalsobedevelopedEPISTEMOLOGICALLY(chapter1)byappealtotheempiricistthesisthatafactualpropositioncanonlybeknownaposteriori,byempiricalenquiry.HowelsecouldIestablishtheexistenceofafactindependentofmyjudgement,yetcorrespondingtoitscontent?Butnormativepropositionsarenotknownbyempiri-calenquiry.Indeed,theyare–intheparticularcaseofpropositionsaboutwhatgivesmereasontobelievewhat–presupposedbyempiricalenquiry.IfwefollowHumewemustacceptthatnormative‘propositions’arenotgenuinepropositions,andthatnormative‘judgements’arenotproperlyspeakingjudgementsbutratherexpressionsofanaffectiveattitudeoradecisionorvolition.ResponsesofthisbroadlyHumeankindhavebeenpopularthiscentury,andhavebeendevelopedtohighlevelsofsophisticationbylogicalpositivistsinthe1930sandmorerecentlybyothers.Theyremainstrikinglyinfluentialincontemporarycultures,notleastamongmanywhoconsiderthemselvesopposedtologicalpositivism.Butitisanequallyinterestingpointthatinphilosophyatleasttheyseemforthemomenttobelosingground.Whatarethealternatives?OnemaydenyHume’sclaimthatallpropositionsareeither‘truebydefinition’orfactual.Oronemaydenyhisempiricism,holdingthatthereare,forexample,‘moralfacts’,or‘logicalfacts’,thatcanbeknownbutthatourknowl-edgeofthemisnotempirical.Orfinallyonemaytakethescepticaloption:holdingthatwhilenormativepropositionsaregenuinepropositions,purportingtocorrespondtoaspecialdomainoffacts,wecanhavenoknowledgeofanysuchdomain,ortherecanbenosuchdomain.Butthiskindofextremeposturetendstobeattractiveonlysolongasonerefusestothinkitthroughfully.226\nETHICSMyownpreferenceisforthefirstresponse,thatofdenyingHume’sclaimthatallpropositionsareeithertruebydefinitionorfactual.ThisresponseisbroadlyKantianinrespectingtheautonomyofthenormative,butitdoesnotforceusintoKant’sratio-nalism.Itcanrecognizethat‘thehearthasitsreasonswhichreasonknowsnothingof’.ButunlikeHume,itfullyacknowledgesthecognitiveorpropositionalstatusofnor-mativeclaims.Itisafalsedichotomytosaythatmoralityismoreproperlyfeltthanjudgedof.Itisjudgedofonabasisthatincludesfeeling.Thisviewdistinguishesbetweennormativeandfactualstatements–theformerobeyinganepistemologyofself-examinationanddiscussionandthelatteranepiste-mologyofcorrespondence.Unliketherealistviewthattherearenormativefacts,towhichtruenormativepropositionscorrespond,itdoesnotholdthatthereissomethingthatmakesnormativepropositionstrueorfalse,orcorrectorincorrect–intherealist’scorrespondencesense.Inparticular,thecorrectnessofanormativepropositiondoesnotcorrespondtothefactthatverdictsonitareideallyconvergent.Thatisnotits‘truthcondition’:ithasnonon-trivialone.Soitisnotmadetruebyafactaboutwhatpeoplewouldagreetoinanidealcontract,either.Theshiftbacktowards‘objectivity’–towardsrecognizingnormativejudgementsgenerally,andmoraljudgementsinparticular,asgenuinejudgements–hasbeenanotablefeatureofmoralphilosophyinthelastthirtyyears.Itisstrikingthatthisshifthasbeenhappeningevenasrejectionofnormativeobjectivityhaswidenedanddeep-enedinthegeneralcultureoftheWesternworld.Willthegeneralculturerespondtothisobjectivistturninagenerationortwo,asitrespondedtotheanti-objectiveturnofphilosophersinthe1920sand1930s?Wecanonlywaitandsee.FurtherReadingForeachsection,afewreadingsaresuggestedwhichdiscussitstopics,leadontoadditionalreadingandoftentakeaquitedifferentviewfromthatadoptedhere.(1)TheScopeofEthics:ch.5,‘TheObjectofMorality’,inMackie(1977);ch.1,‘Introduc-tion’,inSidgwick(1922);PartsIandIIofSinger(1991);andch.1,‘Socrates’Question’,andch.2,‘TheArchimedeanPoint’,inWilliams(1985).(2)Blame,AdmirationandDesire:Part1ofGriffin(1986);Hurka(1983);Skorupski‘TheDefinitionofMorality’inPhillipsGriffiths(1993);ch.10,‘Morality,thePeculiarInstitution’,inWilliams(1985);andGibbard(1990).(3)Character,theVirtuesandFreedom:Aristotle(1954);chs7–14inFoot(1978);Glover(1970);‘FreedomandResentment’inStrawson(1974);andWallace(1994).(4)AnEthicalTheory:TheCaseofUtilitarianism:Part2ofGriffin(1986);‘Utilitarianism’invol.10ofMill(1963–);andSmartandWilliams(1973).(5)Impartiality:Rawls(1972);Darwall(1983);andNagel(1991).(6)WhatCananEthicalTheorybeExpectedtoDo?:Daniels(1979);ch.6,‘EthicalPrinciplesandMethods’andch.13,‘PhilosophicalIntuitionism’,inSidgwick(1922);andPart1,‘Self-DefeatingTheories’,inParfit(1984).(7)ThePluralityofValue:MacIntyre(1981);ch.11,‘MoralityasaSystemofHypotheticalImperatives’,inFoot(1978);ch.8,‘InternalandExternalReasons’,inWilliams(1981);McDowell(1978).(8)PracticalReason:‘OftheInfluencingMotivesoftheWill’(II,III,i)and‘OfVirtueandViceinGeneral’(III,II)inHume(1978);Kant(1964);Book2,ch.1,‘ThePrincipleandMethodof227\nJOHNSKORUPSKIEgoism’andconcludingchapter,‘TheMutualRelationsoftheThreeMethods’,inSidgwick(1922);‘Autonomy,Obligation,andVirtue:AnOverviewofKant’sMoralPhilosophy’inSchneewind(1992);andAllison(1990),anadvancedstudyofKant’sethics.(9)TheCategoricalImperativeandContractualism:Gauthier(1986);‘CouldKantHaveBeenaUtilitarian?’inHare(1993);Scanlon,‘ContractualismandUtilitarianism’inSenandWilliams(1982);andO’Neill(1989).(10)Reason,MoralityandReligion:Chapter8,‘Intuitionism’,inSidgwick(1922);‘ValuesandSecondaryQualities’inMcDowell(1985);Dancy(1993);andBerg,‘HowCouldEthicsDependonReligion?’,ch.46inSinger(1991).(11)Meta-ethics:Smith(1995)andMcNaughton(1988).ReferencesAllison,H.E.1990:Kant’sTheoryofFreedom.Cambridge:CambridgeUniversityPress.Aristotle,1954:TheNichomacheanEthics.Oxford:OxfordUniversityPress.Berg,J.1991:HowCouldEthicsDependonReligion?InP.Singer(ed.)ACompaniontoEthics.Oxford:Blackwell.Dancy,J.1993:MoralReasons.Oxford:Blackwell.Daniels,N.1979:WideReflectiveEquilibriumandTheoryAcceptanceinEthics.JournalofPhilosophy,76.Darwall,S.1983:ImpartialReason.Ithaca,NY:CornellUniversityPress.Foot,P.1978:VirtuesandVicesandOtherEssaysinMoralPhilosophy.Oxford:Blackwell.Gauthier,D.1986:MoralsbyAgreement.Oxford:OxfordUniversityPress.Gibbard,A.1990:WiseChoices,AptFeelings.Oxford:ClarendonPress.Glover,J.1970:Responsibility.London:RoutledgeandKeganPaul.Griffin,J.1986:Well-Being,ItsMeaning,Measurement,andMoralImportance.Oxford:ClarendonPress.Hare,R.M.1993:CouldKantHaveBeenaUtilitarian?Utilitas,5.Harman,G.1977:TheNatureofMorality:AnIntroductiontoEthics.NewYork:OxfordUniversityPress.Hume,D.1978[1739–40]:ATreatiseofHumanNature(editedbyL.A.SelbyBigge)(2ndeditionwithtextrevisedandvariantreadingsbyP.H.Nidditch).Oxford:ClarendonPress.Hurka,T.1983:Perfectionism.Oxford:OxfordUniversityPress.Kant,I.1964[1785]:GroundworkoftheMetaphysicofMorals.InTheMoralLaw(translatedandanalysedbyH.J.Paton).NewYork:HarperandRow.McDowell,J.1978:AreMoralRequirementsHypotheticalImperatives?ProceedingsoftheAristotelianSociety,supp.vol.52.——1985:ValuesandSecondaryQualities.InT.Honderich(ed.)MoralityandObjectivity:ATributetoJ.L.Mackie.London:RoutledgeandKeganPaul.MacIntyre,A.1981:AfterVirtue:AStudyinMoralTheory.London:Duckworth.Mackie,J.L.1977:Ethics:InventingRightandWrong.Harmondsworth:PenguinBooks.McNaughton,D.1988:MoralVision:AnIntroductiontoEthics.Oxford:Blackwell.Mill,J.S.1963–:Utilitarianism,inTheCollectedWorksofJohnStuartMill,vol.10.LondonandToronto:RoutledgeandUniversityofTorontoPress.Nagel,T.1991:EqualityandPartiality.London:OxfordUniversityPress.Nietzsche,F.1974:TheGayScience(translatedwithcommentarybyW.Kaufmann).NewYork:VintageBooks.O’Neill,O.1989:ConstructionsofReason:ExplorationsofKant’sPracticalPhilosophy.Cambridge:CambridgeUniversityPress.228\nETHICSParfit,D.1984:ReasonsandPersons.Oxford:ClarendonPress.Plato1961:Euthyphro(translatedbyL.Cooper).InE.HamiltonandH.Cairns(eds)TheCompleteDialoguesofPlato.Princeton,NJ:PrincetonUniversityPress.Rawls,J.1972:ATheoryofJustice.Oxford:OxfordUniversityPress.Scanlon,T.1982:ContractualismandUtilitarianism.InA.SenandB.Williams(eds)Utilitar-ianismandBeyond.Cambridge:CambridgeUniversityPress.Scheffler,S.1992:HumanMorality.Oxford:OxfordUniversityPress.Schneewind,J.B.1992:Autonomy,ObligationandVirtue:AnOverviewofKant’sMoralPhi-losophy.InP.Guyer(ed.)TheCambridgeCompaniontoKant.Cambridge:CambridgeUniversityPress.Sidgwick,H.1922:TheMethodsofEthics.London:Macmillan.Singer,P.(ed.)1991:ACompaniontoEthics.Oxford:Blackwell.Skorupski,J.1993:TheDefinitionofMorality.InA.PhillipsGriffiths(ed.)Ethics.Cambridge:CambridgeUniversityPress.Smart,J.J.C.andWilliams,B.1973:UtilitarianismForandAgainst.Cambridge:CambridgeUniversityPress.Smith,M.(ed.)1995:Meta-Ethics.Aldershot:DartmouthPublishing.Strawson,P.F.1962:FreedomandResentment.InFreedomandResentment(1974).London:Methuen.ReprintedinG.Watson(ed.)(1982)FreeWill.Oxford:OxfordUniversityPress.Wallace,R.J.1994:ResponsibilityandtheMoralSentiments.Cambridge,MA:HarvardUniversityPress.Williams,B.1981:MoralLuck.Cambridge:CambridgeUniversityPress.——1985:EthicsandtheLimitsofPhilosophy.London:FontanaPaperbacks.DiscussionQuestions1Whatmakesaquestionanethicalquestion?2Couldweacceptmoraljudgementswhileseeingnoreasontoactonthem?Couldotherreasonsoutweighmoralreasons?3Isthereaspecialmoralsenseofblame?Canweunderstandmoralitywithoutunderstandingblame?4Ismoralitydispensable?5Iseveryexcellenceofcharacteramoralvirtue?Ifnot,whatmakesanexcellenceamoralvirtue?6Dotheendsoflifestandoutsidemoralityoraretheyobjectsofmoralassessment?7Cantherebeexpertsinmorality–orinanyotherkindofnormativejudgement?8Ifmoralfreedomisamatterofdegree,howmuchmoralfreedomisrequiredformoralresponsibility?9Whatnotionofrationalityisinvolvedintheclaimthatmoralfreedomisrationality?10Ismoralfreedomcompatiblewithourbeinggovernedbycausallaws?11Inwhatsense,ifany,cantherebeethicaltheory?12Areutilitariansrightinholdingthatonlythewell-beingofindividualshasintrinsicethicalvalue?Whatrivalviewsmighttherebe?13Ifthewell-beingofsomeanimalshaslessethicalvaluethanthewell-beingofhumans,doesthewell-beingofsomehumanshavelessethicalvaluethanthewell-beingofothers?229\nJOHNSKORUPSKI14Whatgroundsmighttherebeforchoosingamongdifferentimpartialprinciplesfordistributingwell-beingamongindividuals?15Ifstandardsofvalueareinternaltodifferenthumanactivities,isitamistaketoseekafullygeneralethicaltheory?16Isitatruthoflogicthatifathingisblameworthy,wehavereasonnottodoit?17Howcanreasonbepractical?18DoestheclaimthatKant’sformalismcannotyieldsubstantiveresultsmisunder-standtheroleofthecategoricalimperativeinhisethicalthinking?19Whatdoesitmeantotreatapersonasanend?20ShouldIdowhatismorallyrightbecauseitisunfairnotto?21Aremoraljudgementsrationallyevidentintheirownright?22Canmoralitybecompletewithoutreligion?23What,ifanything,makesethicalandothernormativejudgementstrueorfalse?24Areethicalviewsexpressionsofaffectiveattitude,forexample,ratherthanrealjudgements?230\n7AestheticsSEBASTIANGARDNERReflectiononbeautyandartgivesrisetoarichanddiversefieldofphilosophicalissues.Thefirstpartofthischapteroutlinestheviewofaestheticexperienceandjudgementwhichformsthebackgroundtocontemporarydiscussionsinaesthetics.Artisthenexaminedinthreeparts,concentratingonthekeyconceptsofrepresentation,expres-sionandmeaning.Attentionisdrawntotherelationoftopicsinaestheticstothoseinethics,metaphysicsandphilosophyofmind,andtotheplaceofaestheticsinsomeofthegreatsystemsinthehistoryofphilosophy.Theexperiencesthatwehavewhenwelistentomusic,readpoetryandlookatpaint-ingsorscenesinnature,haveadistinctiveimmediate,emotionalandcontemplativecharacter,andleadustodescribewhatweexperienceinaspecialvocabulary,andtousetermssuchas‘beautiful’,‘exquisite’,‘inspiring’,‘moving’andsoon.Philosophyemploystheterm‘aesthetic’tocircumscribethiskindofexperience.Theappreciationofartprovidesthemostcomplexandintenseformofaestheticexperience,inwhichwearesetapartfromtherealworldandourpowersofimaginationarefullyengaged,indramaticcontrastwitheverydaypracticallife.Alessmarkedformofaestheticaware-nesssuffusesourperceptionoftheworld.Somedegreeofaestheticreceptivityisshownbythequietinfluencethatarchitectureexertsonourfeelings,andthechoicesthatwemakedailyindesigningourenvironmentmanifestaestheticpreferences.Theabilitytorespondaesthetically,andtheopportunitytoappreciateart,areimportantcomponentsofwell-being.Aworldwithoutaestheticqualitieswouldbeaninferior,ifnotunin-habitable,world,andapersonwithoutanycapacityforaestheticresponse,ifimagin-able,wouldnotqualifyasafullydevelopedhumanbeing.Manyquestionsnaturallyoccurtousinthecourseofthinkingaboutanddiscussingworksofart.Whatmakesanobjectqualifyasaworkofart?Whatistherelationbetweenformandcontentinaworkofart?Doesaworkofartputusintouchwiththemindoftheartist?Whatarewetothinkwhencriticsdisagreeaboutthemeaningofawork?Whydoweattachsuchimportancetoart?Andsoon.Thephilosophyofartaimstoreflectontheselargelyfamiliarquestionsinasystematicway.Inadditiontotopicsinthephilosophyofart,thereareanumberofmoreabstract,lessfamiliarquestionsconcerningthenatureofaestheticexperienceingeneraltobeconsidered.Whatexplainsthespecialcharacterofaestheticexperience?Doesa\nSEBASTIANGARDNERjudgementthatanobjectisbeautifulreportafactabouttheobject?Ordoesitexpressafeelingofthesubject’s?Whatisbeauty?Thinkersintheeighteenthcentury,aboveallHUME(chapter31)andKANT(chapter32),tookupthesequestions,andtheirviewofaestheticexperienceprovidesthebroadframeworkwithinwhichmostcontemporary,analyticalaestheticsoperates.Thefirstpartofthischapterdescribesthisframework,andtheunresolvedphilosophicalquestionsassociatedwithit.1AestheticExperienceandJudgementHumeandKantregardaestheticjudgementasexemplifyingtaste.Taste,ineighteenth-centuryaesthetics,referstothespecialmentalfaculty,orspecialmodeofemploymentofmentalfaculties,exercisedinaestheticjudgement.Whatmorethismeanscanbegraspedbyattendingtotwoabsolutelycentralfeaturesofaestheticjudgement,insistedonbyHume(1965)andKant(1987).Thefirstisthataestheticjudgementrests,fundamentally,onafeltresponsetoanobject(Kant1987:§1).Aestheticjudgementonanobjectcannotbepassedsecond-hand:itrequirespersonalacquaintance.ThetestimonyofothersmaymakemeconfidentthatIwillfindanewlypublishednovelrewarding;butIcannotdeclareittobesountilIhavereaditmyself.Theconditionoffeltresponseisreflectedinthefactthatanaestheticjudgement,likeanavowalofemotion,canbemadesincerelyorinsincerely.Thesecondfeature,closelyconnectedtothefirst,isthatrulesorprinciplesplaynorole,oronlyahighlydiminishedrole,inaestheticjudgement(ibid.:§8).Itcannotbeinferredthatapieceofmusicisrapturoussimplybecauseitisinaparticularkeyandorchestratedforcertaininstruments;orthatapaintingisdynamicbecauseitscompo-sitionhasacertaingeometricform.Nolawsconnecttheaestheticqualitiesofanobjectwithitssensory,non-aestheticproperties.Thepointisnotjustthatweareignorantofrulesforaestheticjudgement,orunabletoagreeonthem,butthattherewouldbelittlepointintryingtoformulateasetofrules:wecouldnotuserulestoproduceinourselvesfeltresponsestoobjects;wecouldnot,incasesofdisagreement,reasonablyaskanotherpersontorelinquishhisorherjudgementonthegroundsthatitconflictedwiththerules;andwheneverourownresponsesdepartedfromtherules,wewouldquiterightlyrepudiatetherulesratherthanourresponses.Thefactthatrulesareotioseindeterminingtheaestheticqualitiesandvalueofobjectsreflectsthefactthataestheticinterestisdirectedessen-tiallytowardsparticularobjects,appreciatedfortheirownsake,andnottowardstheformulationofgeneraltruths.Tojudgeaccordingtorulesistomisstheuniquenessofobjects,andsotofailtojudgeaesthetically.Theplaceofrules,inguidingaestheticjudgement,istakenbyparticular,exemplaryaestheticobjects;‘establishedmodels’,asHumecallsthem.HumeandKantagree,further,inrejectingtheview(ofearlier,classicalandratio-nalistaesthetics)thataestheticqualitiesareobjective.Aestheticobjectivismmaintainsthataestheticqualitiesarepropertiesinheringinobjects,andthataestheticexperiencegivesusknowledgeoftheseproperties.Thesepropertiesmaybeidentifiedwiththeobject’sformalproperties,suchas‘idealproportion’;ortheymayberegardedassui232\nAESTHETICSgenerisandirreducibletoformalproperties(Moore1984:§§112–21;McDowell,inSchaper1983).(Thesevarietiesofaestheticobjectivismbearcomparisonwithnatu-ralismandintuitionisminethics,respectively.)Aestheticsubjectivismdeniesthataestheticqualitiesinhereinobjects,andmaintainsthatwhatitisforanobjecttobebeautifulisforittoyieldacertainresponseinthesubject.InaestheticexperienceIamaffectedbytheobject,andmyresponsedoesnotconsistinknowledgeofitsproperties.Theobject’sformalandothernon-aestheticprop-ertiesaremerelywhatoccasionstheresponse.Aestheticsubjectivism,wewillsee,comesinvariousdegreesofsophistication(whichagainmaybecomparedwithdiffer-entformsofsubjectivisminethics).Severalverypowerfulconsiderationsmilitateinfavourofaestheticsubjectivism.Firstly,asHumeobserves,andasweallrecognize,thetastesandverdictsofdifferentindividualsandculturesvaryenormously.(Lestthisbedoubted:JohnsonsaidofShake-spearethathehas‘faultssufficienttoobscureandoverwhelmanyothermerit’;VoltairecomparedShakespeare’sworkstoa‘dunghill’.)Ifaestheticqualitiesarereallyoutthere,inheringintheobject,whydowenotallpickuponthem?Thepointisnotjustthatdivergenceisafactofaestheticlife,butthatwelackagreedmethodsforbringingdivergentjudgementsintoline.Aswassaidearlier,rulescannotbeappealedto.Andaestheticdisagreementisunlikedisagreementaboutcolour,whereconditionsofphys-iologicalnormalityandstandardlightingcanbecheckedindependently.Aestheticqualitiesaredeeplyelusive:the‘positioning’oftheobserverneededtodiscernaestheticqualitiesinvolvesaplethoraoftemperamentalandculturallyparochialfactors,allstronglyindicativeofsubjectivity.ThesecondpointwasstressedbyKant(1987:§1).Pleasureisessentialtotheexperiencewhichgroundsanaestheticjudgement.Pleasure,likepain,is,however,amentalstatethatdoesnotrepresentapropertyofitsobject;asKantsays,pleasure‘designatesnothingwhatsoeverintheobject’.Thenecessaryconnectionofaestheticjudgementwithpleasureisreadilyexplainedbythesubjectivist,butitpresentsaestheticobjectivismwithanenormousdifficulty:howcantherebepropertiesinherentinobjectswhichnecessarilygeneratepleasuresimplythroughbeingapprehended–ifnotonthetheologicalsuppositionthattheworldhasbeenmetaphysicallytailoredtodelightus?Suchpropertieswouldbemetaphysicallyqueerintheextreme.(ThisobjectiontoaestheticobjectivismmaybecomparedwithHume’sobjectiontoethicalobjectivism,thatitfailstocapturethenecessaryconnectionofmoraljudgementwiththewill.)Thisobjectioncanbeamplified.Objectivism,becauseitholdsthataestheticqualitiesinhereinobjects,necessarilyintroducesadistinctionbetweenrealityandappearance,betweenhowthingsareandhowthingsseem.Objectivismthereforeimpliesthat,evenwhenanaestheticjudgementhaswithstoodallofthetoughesttests,wecanstillbemistakenaboutit:itisstillpossiblethattheobjectlackstheaestheticqualityattributedtoitandisreallyaestheticallyworthless.Butthisseemsmad.Itissimplynotintelligi-blethatourgreatestartcould‘really’,appearancestothecontrary,beworthless.Suc-cessful‘appearances’ofaestheticqualityareasgoodas–indeedtheyarethesameas!–therealthing:foraestheticrealitypreciselyconsistsinappearances.Putanotherway,theusualmotivationforobjectivisminphilosophy–namely,showingthatourexperiencesandtheoriesputusintouchwiththeworldasitreally233\nSEBASTIANGARDNERis,independentlyofourmeansofrepresentingit–hasnorelevancetothecontextofaestheticexperience,whichhasinsteadastrongsimilaritytothecontextofpersonalrelationships,inwhichrewardsotherthanthegainingofknowledgearesoughtprimarily.Thecognitivedemandthatmindshould‘fittheworld’doesnotapplytoaestheticinterest,whichisinthisrespectclosertodesirethanbelief–whatwewantisfortheworldtofitus.If,however,aestheticsubjectivismiscommon-sensicalandphilosophicallyplausible,itappearstocarryahighlyuncommon-sensicalandunwelcomeimplication.Ifaes-theticqualitiesdonotinhereinobjects,whatistopreventthesameobjectbeingjudgedbeautifulbyonepersonanduglybyanother–bothbeingequallyjustifiedintheirpro-nouncements?If,asHumeputsit,‘Allsentimentisright’,thereisnosuchthingasright-nessorwrongnessinaestheticmatters.Inthatcase,anaestheticjudgementmerelyreportsthespeaker’smentalstate,andaestheticpreferencesareonaparwithgusta-torypreferences.Itfollowsthataestheticjudgementscannotbecontestedorsupportedwithreasons.Thisposition,likeemotivisminethics,towhichitcorresponds,isincon-sistentwithcommonsense(Kant1987:§56).Thustheproblemoftaste:ofhowtomaintainthesubjectivefeltbasisofaestheticjudgement,withoutcollapsingintounrestrictedRELATIVISM(pp.395–7).HumeandKantrespondtotheproblemoftasteindifferentways.Humecontendsthata‘standardoftaste’–abasisforacceptingsomejudgementsoftasteascorrectandrejectingothersasincorrect–isbothphilosophicallydefensibleandvisiblyatworkinthewaythatweactuallyconductouraestheticaffairs.Thestan-dardoftasteliesnotintheobjectbutinthesensibilityofthesubject.Aestheticsensi-bilityvariesinits‘delicacy’:‘Whentheorgansaresofineastoallownothingtoescapethem,andatthesametimesoexactastoperceiveeveryingredientinthecomposition,thiswecalldelicacyoftaste’(Hume1965:11).Acorrectaestheticjudgementisonethatissuesfromadelicatesensibilityoperatingunderidealconditions,whichincludethecultivationoftastethroughpractice,theabilitytomakerelevantcomparisons,andfreedomfromprejudice.Wedefertothejudgementofthosewhosesensibilitiesweacknowledgetobesuperior,thatofthe‘critics’.(TheconceptofanIDEALOBSERVER(p.470)playsananalogousroleinHume’smoraltheory.)ItfollowsthatforHume,aestheticjudgementsdonotidentifyaestheticqualitiesinheringinobjects,butnordotheysimplyreportthesubject’sexperiences.Anobject’spossessionofanaestheticqualityconsistsinitsbeing‘fitted’togenerateacertainresponseinus.ThisdistancesHumefromthekindofunsophisticatedsubjectivismwhichendsupinrelativism.Hume’ssolutiontotheproblemoftastedependsoncontingentfactsofnature.Firstly,thefactthatsome‘particularformsorqualities[inobjects]arecalculatedtoplease,andotherstodisplease’.Secondly,thecontingentuniformityofhumansensi-bility,thesamenessin‘theoriginalstructureoftheinternalfabric’ofourminds–bywhichismeant,notthatweareallequallycompetentaestheticjudges,butthatoursensibilitiesareallofakind.Wheresensibilitiesdiffer,thedifferencesarenotarbitrary:aestheticjudgementsdivergebecauseofdifferencesindelicacy,andotherdeterminabledeficienciesoftaste;ifthesensibilitiesoftwosubjectsareequallydelicateandtheirconditionsofjudgementotherwisethesame,theirjudgementswillcoincide.234\nAESTHETICSKantandtheNormativeAspectofAestheticsKantrejectsthekindofaccountproposedbyHume.Recallthattheproblemoftasteisoneofjustification.Thequestionis,onwhatgroundscanweregardcertainaestheticjudgementsasbeingcorrect?ForKant,thereisnothinginHume’spositionwhichaccountsforthis,thenormativeaspectofaestheticjudgement.Humedoesnotvalidatetheclaim,implicitinallaestheticjudgement,thatone’sresponseisappropriatetotheobject,andthatothersoughttoconcur.Kantinterpretsthisdemandforagreementstrongly,asapplyingtoallotherpeoplewithoutexception.OnKant’sview,Humeoffersonlyacausal,psychologicalexplanationofwhywedo,asamatteroffact,defertothejudgementsof‘critics’–hedoesnotsaywhyweoughttodoso.Kant’ssolutiontothenormativeproblemis,inessence,verysimple,althoughhispre-sentationofitisextremelyintricate.Supposethat,inmakinganaestheticjudgement,weabstractfromeverythingthatmightpertaintoourcontingent,natural,individuallyvariableconstitutions,andbaseourjudgementssolelyonconditionsthatarestrictlyuni-versal,inthesenseofbeingavailableandcommontoallhumanbeings.Suppose,thatis,webaseouraestheticjudgementonthebareformoftheobjectanditsinteractionwithourbasic,universalmentalpowersofperceptionandunderstanding.(Kant’srea-soninghererecapitulateshisanalysisofmoraljudgementintermsofthecategoricalimperative.)Kantarguesthatthis‘universal’standpointcanbeachievedthroughfreeingourawarenessoftheobjectfromdesireandpracticalconcern(whathecalls‘disinter-estedness’),andfromourconceptualunderstandingofit.Whenthesestringentcondi-tionsaremet,thejudgementwemakeisvalidforeveryone.WethenhavewhatKantcallsa‘purejudgementoftaste’,whichhas‘universalvalidity’.Kantarguesthatitisindeedpossibleforthemereformofanobjecttodelightus:certainperceptualformsstimulateourmentalfacultiesoptimally,byengenderinginusa‘harmoniousfreeplayofimaginationandunderstanding’,awarenessofwhichispleasurable.Kantalsogivesametaphysicalinterpretationofaestheticexperience:itmakesusconsciousofaconnectionthatwehavewiththeworld,andwithoneanother,whichliesbeyondtheempiricalworld.Beautythereforehasasemi-religiousandmoralsignificanceforKant(hecallsbeautythe‘symbolofthegood’).EvenifwegrantthatKant’stheoryofmentalharmonyaccountsfortheexperienceofbeauty,acostattachestoKant’ssolutiontotheproblemoftaste.PurifyingaestheticjudgementinthewaythatKantrequiresleavesuswithanaustere,ifnotimpoverishedviewofaestheticexperience–asconsistingofnothingbutanestimationofform.ThislimitationshowsupinKant’sinabilitytodojusticetothepsychologicallyrichdiversityofexperiencewhichartoffers.Weareleft,then,withachoice:betweenacceptingHume’saccount,without(Kantargues)anywayofsecuringthenormativeaspectofaestheticjudgement;andaccept-ingKant’saccount,atthepriceofexcludingfromaestheticexperienceourculturalidentities,andeverythinginourpsychologicalconstitutionsthatisnotstrictlyuniver-sal(which,theHumeanwillargue,leavesalltoolittle).ThetaskofprovidingasingleaccountwhichincorporatesandcombinestheinsightsofHumeandKantremains.Anumberofotherissuesremainunresolvedwithintheframeworkofaestheticsubjectivism,includingtheformulationofthedoctrineitself.Grantedthataesthetic235\nSEBASTIANGARDNERqualitiesaresubjective,quitehowsubjectivearethey?Arethey,forinstance,onaparwithcoloursandothersecondaryqualitiesofobjects?Sometheoriesmaintainthattheyare.Othersdenythataestheticqualitiesareonaparwithcoloursandcomparethemwith‘looks’or‘aspects’ofobjects,ordescribethemas‘powers’toproduceexperiences.Another,moreradical,possibilityisthataestheticjudgementsarenotdescriptiveatall.WITTGENSTEIN(chapter39)(1978)suggeststhataestheticjudgementsaremorelikegesturesandexclamations.Onthisview,thefunctionofanaestheticjudgementisnottosayanythingaboutanobject,butto‘putacross’anexperience.(TheseoptionsareexploredinSibley1959,1965;SibleyandTanner1968;Wollheim1980:essay6;Hungerland,inOsborne1972;Meager1970;andScruton1974.)Itisalsonecessary,asbothHumeandKantsaw,toclarifythenatureandroleofreasonsandcriticisminaestheticcontexts.Thataestheticjudgementscanbejustifieddistinguishesthemfrommerelikings,andmakesthem,inabroadsense,rational.Itisthejobofthecritictoidentifythesourcesofawork’saestheticqualitiesanddeterminewhatresponsesareappropriate(afurtherroleforcriticism,tobediscussedlater,isinter-pretation).Butaestheticreason-givingdiffersfundamentallyfromreason-givinginothercontextsandhasanumberofpeculiarities,whichmakeithardtounderstand(Beardsley1982:ch.12;Scruton1979:ch.5).Itdoesnotconsistininferringonepropositionfromanother;itmakesnouseofrules,unlikemoralreason-giving;anditmakesnouseofinductionorgeneralizations,unlikereason-givinginscience.Eventhenotionofconsistencyhaslittleapplicationtoaestheticjudgements:likingWagnerdoesnotmeanthatoneoughttodislikeMozart.Theproblem,insum,isthataestheticreasonsoperateindependentlyfromalloftheconditionsthatappearintegraltotheveryconceptofareason.Twofurtherunresolvedissuesmaybepointedoutinconclusion.Thefirst,bequeathedbyKant’snotionofdisinterestedness,concernstheimportanceoftheconceptofan‘aestheticattitude’.Sometheoristshaveclaimedthataestheticexperi-enceconsistsintheadoptionofaspecialattitudeinwhichobjectsareattendedtodis-passionatelyand‘fortheirownsake’.Indeed,Schopenhaueridentifiesthepointofartwiththemetaphysicalliberationfromlifethataestheticcontemplation,whichdoesnotinvolvethewill,bringsinitswake(Schopenhauer1969,1:thirdbook).Thedistinctivephenomenologyofaestheticexperiencemakesthisideaalluring,butwhatcanbesaidabouttheaestheticattitude,otherthanthatitdiffersfromandexcludespracticalandcognitiveattitudes?Sayingthatanobjectisenjoyed‘foritsownsake’reallymeansthatitis‘notenjoyedforanyulteriorend’.Becausedescriptionsoftheaestheticattitudetendtoremainfundamentallynegative,itisunclearwhatvaluetheconcepthas,strippedofthemetaphysicalinterpretationthatmakesitsignificantforSchopenhauer(see,however,Beardsley1982:ch.16).Asecondissueconcernstheconceptofbeauty,whichhasundergoneadramaticreversaloffortunesinthehistoryofaesthetics.Classicalaestheticstakesitforgrantedthatbeautyistheonly,oratleastthefundamental,aestheticquality.Somemodernwritersproposebycontrastthat‘beautiful’ismerelyacatch-allterm,roughlyequiva-lentto‘aestheticallycommendable’,andthatthereis,asordinarylanguageimplies,alimitlesspluralityofaestheticqualities,encompassingelegance,grace,poignancyandsoon.Thereis,however,aphilosophicalissueherewhichneedstobeaddressed.Theideathatbeautyisnotjustoneaestheticqualityamongmany,butissomehowpre-236\nAESTHETICSeminentandaccountsfortheunityofaestheticexperience,iscompelling,andanyonewhodeniesthisinfavourofthe‘pluralist’viewischallengedtofindanalternativeaccountoftheunityoftheaesthetic.ArecentdefenceoftheconceptofbeautyisinMothersill(1984).2TheEssenceofArtHEGEL(chapter33)assertsthatart‘pervadeswhatissensuouswithmind’,andthatartaccordinglyhas‘ahigherrankthananythingproducedbynature,whichhasnotsustainedthispassagethroughthemind’(Hegel1993:15,34).WhetherornotHegel’sviewofthesuperiorityofartisjustified,thephilosophyofart,towhichwenowturn,sharesHegel’sconceptionofartasasynthesisofmindwithsomethingotherthanitself,inwhichthemind‘recognizesitself’;andhisrejectionoftheviewthataworkofartissimplyanartefactpossessingaestheticqualitiesofakindalsofoundinnature.Theo-riesofartattempttoelucidatethenatureoftheconnectionwiththemindwhichmakesanobjectaworkofart.Suchattemptsassumethattheconceptofartpicksoutmorethanamerelynominal,orhistoricallyaccidentalphenomenon–thatarthas,inshort,anessence.2.1DefinitionsofartAttemptshavebeenmadetocapturetheessenceofartinasimpledefinitionalformula.Tworepresentativeexamplesarethedefinitionofartastheimitationofbeautifulnature,andthedefinitionofartasthecommunicationoffeeling.Ifweconsidersuchdefinitions,itisclearthattheyarewideopentocounter-examples.Noteveryworkofartisanimitationofbeautifulnature,orcommunicatesfeeling:novelsarerarelyifeverbeautiful,instrumentalmusicisnotimitativeandmuchvisualartdoesnotcommunicatefeeling.Andeveniftheseformulaeprovidednecessaryconditionsforart,theycouldhardlyprovidesufficientconditions:someindustrialmachineryisbeautiful,waxworksareimitative,racistpropagandacommunicatesfeelings.Oneresponsetothefailureofdefinitionsofthissimplekindisscepticism.Ithasbeenclaimedthatthedifficultyofdefiningartisareflectionoftherenotbeinganessenceofart.Proponentsofthisview(discussedinTilghman1984)haveoftenappealedtoWittgenstein’snotionthatfamilyresemblancesbetweeninstancesofaconceptmaybeallthatthereistobefound.Arecent,much-discussedattempttodefineart–whichisbuiltontheWittgenstein-iandenialthatarthasanessence,butmaintainsthatadefinitionisneverthelesspossible–istheinstitutionaltheory.Theinstitutionaltheoryacceptsthatworksofartcannotbedefinedintermsoftheirintrinsic,perceptibleproperties,andturnsinsteadtotheirextrinsic,socialproperties.Thetheorysaysthatsomethingisaworkofartifamemberoftheartworldhasconferredonitthestatusofbeingacandidateforappreciation,orifithasbeencreatedinordertobepresentedtotheartworld(Dickie1984).Theartworldisanebulousentitythatincludesartists,criticsandsomeportionofthegeneralpublic.‘Workofart’is,onthisaccount,an‘honorific’term:beingaworkofartisanon-per-ceptiblepropertywhichdependsuponanobject’splaceinaculturalcontext(analogous237\nSEBASTIANGARDNERtobeing‘inpublicownership’or‘sacred’).Thisapproachreceivesanadditionalstimulusfromthewayinwhichsomeavant-gardeartofthiscentury–paradigmatically,Duchamp’sready-madeFountain,amass-producedurinalenteredforexhibitionunal-teredbytheartist–appearstohaverepudiatedalltraditionalcriteriaforart.Thereis,however,somethingmissingattheheartoftheinstitutionaltheory.Ittellsusthatmembersoftheartworldbestowthestatusofartonsomeobjectsratherthanothers;butnotwhytheydoso.Itthusleavesoutofaccountthereasonsforcallingsomethingaworkofart–theconditionstakentojustifyapplicationoftheconcept.Thisissomethingthatoughttofigureinaphilosophical,asopposedtoasociological,treatmentoftheconceptofart(Wollheim1980:essay1).Proponentsoftheinstitutionaltheoryseektodeflectthiscriticismbysayingthattheyareconcernedwiththe‘classificatory’senseofart,notthe‘evaluative’senseinwhichtocallsomethingaworkofartistorecommenditforappreciation.Butthisinvitesthefurtherobjectionthatitissimplyamistaketoseparatetheclassificatorysenseofartfromtheevaluative.Evaluationisjustasintegraltotheconceptofartasitistomoralconcepts.Wedonotfirstclassifyobjectsasart,andthendiscoverthattheyhappentobeaestheticallyrewarding:conceptually,thereisonlyonemovehere.Whatmoralistobedrawnfromtheseattemptstodefineart?Itcannotberuledoutthatacomplexdefinition,perhapscombiningtheconceptsofimitationandcommuni-cationinacomplexwayandaddingfurtherconditions,couldbedevisedtoencompassallandonlythoseobjectsthatweconsidertobeworksofart,andatthesametimerespecttheevaluativenatureoftheconceptofart.Butwhatthedifficultyofreadingoffanadequatedefinitionofartfromthesurfaceofourordinaryconceptionofartmaybetakentoshow,moreimportantly,isthatnodefinitionofartcanbeexpected,andinanycasecannotholdmuchinterest,inadvanceofatheoryofart.Atheoryofartdoesnotrequiretheretobeasimple,manifestpropertysharedbyeachandeveryworkofart;itallowsthatwhatunifiesart–itsessence–maynotbevisibleatitssurface,andprobesbeneaththesurfacetolocateit.Atheoryofartmaytakeoneoftwoforms.Itmayaimdirectlyatdevelopingasingleconceptcapturingtheessenceofart.Oritmayproceedbybuildingupfromanexam-inationofthevariousspecificdimensionsofworksofart.Thefirst,moretraditionalapproachisexploredbelow;thesecond,moremodestapproachfavouredincontempo-raryaesthetics,isexploredinthefollowingsection.2.2TheoriesofartOfthenumerousandenormouslyvariedtheoriesofartthathaveappearedinthehistoryofaesthetics,threemaybesingledoutashavinggreatestimportance.Thesearethemimetic,formalistandexpressiontheories.Eveniftheconceptofimitation,asordinarilyunderstood,doesnotstraightforwardlydefineart,itmaybesupposedthattheconceptcanbedevelopedbeyonditsordinaryscope.Onthisclaimreststhetheorythatartconsistsinmimesis–aGreektermren-deredapproximatelyby‘imitation’,‘copying’or‘representation’(seePlato1955;Aristotle1987;andchapter23ofthisvolume).Theobjectofmimesisisusuallyiden-tifiedwithnature,inclusiveofhumannature.Although,asobservedearlier,someart,suchasinstrumentalmusic,appearstobenon-representational,itisopentothe238\nAESTHETICSmimetictheoristtocontendotherwise:inantiquityitwasthoughtthatmusicimitatestheorderandharmonyofthecosmosandthesoul.ThemimetictheoryconstruestheconnectionofartwiththemindassertedbyHegelinthefollowingway:worksofartcarryoverthemind’sfundamentalfunctionofrepresentingtheworld.Grantedthattheconceptofrepresentationisopentobeingextended,themimetictheoryisopentoanobjection(Hegel1993:LXI–LXVII).Evenifweacceptthatitisnaturaltoenjoyimitationandtheskillexhibitedtherein,asAristotleobserves,whyshouldwefindimitationvaluableinthespecialwaythatartisfoundvaluable?Thevaluesconnectedconceptuallywithrepresentationarecognitive,truth-orientatedvaluessuchasaccuracyandcomprehensiveness.Thesehavesomeroleinart–verisimilitudeofplotandcharacterinliteratureforinstance–andcertainlytruthissomethingwevalue.Buttruthdoesnotcapturetherealinterestofart.Aworkofartmustcomeintoitsowninthefieldofourexperience,andnotdisappearfromourattentioninthemannerofatransparentwindowontotheworldorvehicleforcommunicatingtruthsaboutit.Themimetictheorythusappearstomisconstruethevalueofart.Themimeticist’sconceptionofrepresentationaccordinglyneedstobetightenedup.Representationwhichisartisticmustbecircumscribedinsubjectandmode:itmustbeofparticularkindsofthings,representedinaparticularway.Itisthereforenoaccidentthatmimetictheoristshavecharacteristicallygoneontoclaimthatartmustidealizeitssubjectmatter.Butoncethemimetictheoryhasundergonethismodification,theessenceofarthasbeenshiftedawayfrombarerepresentation,inthedirectionofwhateveritisaboutartthatenablesittoidealizeitssubjects.Astrongcandidateforthisroleisartisticform.Wherethemimetictheorytiesartdowntotherealworld,formalismallowstheworkofarttofloatfree,insistingthatonlyform,thecomplexarrangementofpartsuniquetoeachindividualwork,hasartisticsignificance.Onlywhatisinternaltotheworkisrelevanttoitsstatusasart;anyoutwardreferences,toarealorimaginaryworld,areirrelevant.Theconceptofformcanbemademoreorlessnarrow;atitsnarrowest,onlysensorypartsandtheirrelationsareintended.Formalistsdifferoverthecontinuityofforminartwithforminnature:Kantholdsthatartisticformmustberecognizablyofakindwithnaturalform;Bell(1914:ch.1),theboldestrecentexponentofformalism,holdsthatartisticform,labelled‘significantform’,isineffectexclusivetoart.Appreciationofartconsists,fortheformalist,notinmerelyrecognizingform,butinrespondingtoit:Kant’stheoryofmentalharmonyhasalreadybeenmentioned;Bellpositsauniquekindof‘artisticemotion’accompanyingtherecognitionofsignificantform.Accordingtoformalism,whatthemindrecognizesofitselfinworksofartisitspoweroflocatingorderinperception,exercisedtoaspeciallyheighteneddegree.Theweaknessofformalismmostfrequentlyindicatedbyitscriticsconcernstheconceptofform,whichisarguedtobetooindefinitetoplaytherolethatisaskedofit.Askedtospecifythekindofformthatmattersinart,theformalistusesnotionssuchas‘balance’or‘uniformityamidstvariety’;butitprovesimpossibletodefinetheseinawaythatpreventsthemfromapplyingtoanyobjectwhatsoever.Theformalistisconse-quentlypushedtodeclarethatartisticformisindefinable.Bellembracesthisclaim,buthistheoryisthenchargedwithvacuousness:heexplainssignificantformintermsofthe‘artisticemotion’thatitinduces,butthenrefersusbackto‘significantform’forourunderstandingofartisticemotion.239\nSEBASTIANGARDNERWhatevertheforceofthispointagainstBell,thereisanotherobjectiontobepressed.Isourinterestinformreallyasuncontaminatedwithulterior,worldlyconcernsasfor-malistssuppose?Formalvaluesmaysometimesbeofself-sufficientaestheticinterest,butmuchmorecommonlytheyservenon-formalends:formisthevehiclethroughwhichaworkarticulatesitsnon-formalmeaning,apartfromwhichformtendstobecomeartis-ticallyuninterestingandmerelydecorative.Inotherterms,theattempttodisentangleformfromcontentleaves,onthesideofform,aninsufficientlysignificantresidue.Thisobjectiontoformalismmaybepressedonbehalfofthemimetictheory;butitalsopointsinthedirectionoftheexpressiontheory.Theconnectionofartwithfeeling,wesaidearlier,cannotconsistinastraightforwardequationofartwiththecommuni-cationoffeeling.Theexpressiontheoryofartseekstoofferamoresophisticatedandpersuasiveaccountofthecentralplaceofemotioninart.TherudimentsoftheexpressiontheorylieinwritingsoftheRomantics,butitwasfirstformulatedphilosophicallybyCroce(1992),whobounditupwiththetenetsofGermanidealism.ThemoreaccessibleversionofthetheorypresentedbyCollingwood(1937)detachesitfromMETAPHYSICS(chapter2),andunderstandsartisticexpressionasaspecialformofself-expression.Artisticexpressionisaprocessinwhichtheartistbeginswithanindefiniteandinchoateemotionalstate,forwhichheorshelabourstofindauniquelyaptconcretearticulation,andinsodoingtransformshisorhermentalstateintosomethingdefinite,tangibleandintelligible.Whattheartistcreatesdoesnotdescribehisorherstateofmind,somuchasincorporateit;analogouslytothewayinwhichbodilyexpressionssuchassmilesandgrimacesembodymentallife.Sincetheproductofexpressioncannotbeknownbeforetheprocessiscomplete,expressioncannotconsistinexercisingatechnique:itmusttakewhateverparticularshapeiscommandedbytheparticularemotionwhichtheartist’smindisimpelledtoclarify.Becauseexpressionisnotundertakenwithanyfurtherendinview,artisticcre-ationcontrastswithinstrumentalactivities,inwhichmeansandendsaredistinct;theseCollingwoodcalls‘craft’,andopposesto‘artproper’.Mistaken,‘technical’con-ceptionsofart–whichincludetheconceptionofartasmimesisorcommunicationoffeeling–arisefromafailuretograspthisdistinction.Onthesideofartisticappreciation,thetheoryclaimsthattheaudienceretracesinthecourseofappreciatingaworkofarttheroutepursuedbytheartist:theirapprecia-tionre-enactstheartist’screativeprocessandtherebyretrieveshisorherpsychologi-calstate(Elliott,inOsborne1972).Therelationoftheaudiencetotheworkofartthusmirrorstheartist’sunderstandingofhisorherownmind.Thecapacityofaworkofarttotransmittheartist’spsychologicalstateisanecessaryconsequenceofsuccessfulexpression.Theexpressiontheory,bygivingprimacytotheperspectiveoftheartistratherthan(asonthemimeticandformalisttheories)thatoftheaudience,offersaverystronginterpretationofHegel’sclaimthatthemind‘recognizesitself’inworksofart:worksofartdonotmerelyexhibitmentalfeatures,they,asitwere,containmind.Theexpressiontheory’semphasisonthepsychologyoftheartistexposesittocriti-cism.Themimeticistwillobjectthatnotonlyemotions,butalsoideas,whichrefertotheworld,areexpressedbyart,whoselegitimatesubjectisnotrestrictedtotheartist’sownmind.Attheotherextreme,theformalistwillchallengetheexpressiontheorytosayhowasupremelyself-containedandself-sustainingworksuchasaMingvase,canbeconstruedasaproductofpersonalexpression.240\nAESTHETICSThefirstobjectioncanbemetbyexplainingthatwhatthetheorymeansbyemotionisanymentalstate,howeverpermeatedwithconceptsandthought,thathassomeemo-tionalcharge.Theconditionthatemotionbepresentisplausible,formostpeoplewouldagreethatmerebeliefswithoutanypsychologicalresonancedonotprovidesufficientmaterialforart.Thescopeofthethoughtsembeddedintheartist’semotionisfur-thermoreleftopen,whichmeansthattheartist’smoralandspiritualviewoftheworldandconceptionoflifearepropermaterialforexpression,andthatwhataworkofartexpressesneednotincludeapersonal,biographicalreferencetotheartist.Theformalist’sobjectioncannotbemetbyexpandingthetermsofthetheory,andobligestheexpressiontheoristtoadoptasemi-stipulativemeasure.Thetheorymustdeclarethatobjectspossessingonlyformalvirtuesdonotqualifyasfull-bloodedinstancesofworksofart.Thismoveisobviouslyacceptableonlyontheassumptionthatart’scentreofgravityliesinthepsychologyoftheartist;itwillberejectedbytheformalistasbeggingthequestion,andasbetrayingthetheory’sfundamentallyparochial,pro-Romanticbias.Ofthethreetheoriesconsidered,itisfairtosaythat–ifamonopolyistobegranted–thentheexpressiontheoryhasstrengthsthatgiveittheedgeoverthemimetictheoryandformalism.Thesestrengthsalsoaccountforthetheory’scontinuedholdonourthinkingaboutart.Itmayalsobepointedoutinitsfavour,thattheexpressiontheoryallowstheunderstandingofarttodrawontheresourcesofpsychologicaltheory(Wollheim1987);anultimateverdictonthetheorywilldependinpartonwhatvalueisfoundinsuchdevelopments.TheSemioticTheoryofArtAfourththeory,whichisnottraditionalbutdominatesmuchcontemporarydiscourseaboutart,shouldbementionedinconclusion.Thisisthesemiotictheoryofart,whichproposesthatworksandformsofartbeanalysedintermsoflogico-linguisticcategoriessuchassignification,reference,denotation,andsyntacticandsemanticrules.Goodman(1976),theprincipalexponentofthesemiotictheory,analysesrepresentationandexpressioninsuchterms.Thesemiotictheoryofartisalsoassumedbystructuralistandpoststructuralistliterarytheory,whichgraftslinguisticscienceandphilosophicaltheoriesoflanguageontothestudyofliterature.Onthesemiotictheory,artisasymbolsystemmuchlikelanguage,andnotjustinthemetaphoricalsenseinwhichanexpres-siontheoristmightgrantthatartisa‘languageofemotion’.Whattheworthofthesemiotictheoryturnsoncanbeindicatedwithreasonableassurance.Thetheoryderivesitsmotivationfrom,firstly,arejectionoftheprimacyofpsychologicalandexperientialconcepts(these,itclaims,arenotautonomousandneedtobeunderstoodintermsoflanguageandsignification);and,secondly,abeliefintheradicallyconventionalnatureofart(whichitgroundsonablanketrejectionofphilo-sophicalrealism).BothstrandsareexplicitintheworkofGoodman,andinstructural-ismandpoststructuralism.Itfollowsthatthesemiotictheory’saustereperspectiveonartmustberejectedbyanyonewhowishestomaintainthatartisessentiallyconnectedwithcertainformsofexperience;asitmayalsobe,onthesuppositionthathumanpsychol-ogysetsuniversal,non-conventionalparameterstoart.Thislastideaprovidesthebasisforafifththeoryofart,thenaturalistic,whichwillbedescribedinsection4below.241\nSEBASTIANGARDNER3TheDimensionsofArtRecentworkinanalyticalphilosophyofarthasadoptedthesecond,morepiecemealandempirical,approachdescribedearliertotheconstructionofatheoryofart.Ratherthanaimdirectlyatageneral,overarchingcharacterizationofart,ithasconcentratedonthespecificdimensionswhichconstituteworksofart,inthebeliefthatanaccuratepictureofartasawholewillemergeoutofaproperunderstandingofitsparts.Representation,expressionandmeaninghavebeencloselyanalysed,intermsofthespecificformsthattheytakeineachofthearts.Thissectiontracesthedebatessur-roundingtheseconceptsintermsoftheformofartmoststronglyassociatedwitheach:representationinpainting,expressioninmusic,andmeaninginliterature.(Issuesconcerningtheontologyofworksofart,thenatureoffiction,emotionalresponsetofiction,andmetaphor,alsobelonginthiscontext,andwouldbeincludedinafulleraccountofanalyticphilosophyofart.)3.1RepresentationWhenwelookatTitian’sTheRapeofEuropawedonotsee,ordonotjustsee,pigmentdistributedacrossthetwo-dimensionalsurfaceofacanvas:weseeawomanborneacrosstheseaaloftawhitebull.Paintingsrepresentordepictthings,bothrealandimaginary.Weareconcernedherewithwhatpaintingsrepresentinavisual,asopposedtointerpretative,sense.ApaintingofamanclutchingastonemayrepresentStJerome,andalutewithabrokenstringinastill-lifemayrepresentDiscord.Inordertomakesuchiconographicidentifications,however,ahumanfigureandamusicalinstrumentmustfirstbeseeninthepicture,andthisisaperceptualmatter;itissomethingthatarthistoricalscholarshipcouldnotdoforus.Visualrepresentationissomethingwetakeforgranted.Whenweceasetodoso,aphilosophicalquestionarises:whatmakespictorialrepresentationpossible?Themimetictheoryofart,itshouldbenoted,doesnotanswerthisquestion:itmerelyassumesthatpictorialrepresentationispossible.Thephilosophicalpuzzlesurroundingpictorialrepresentationcomesintofocuswhenonereflectsonthedifficultiesassociatedwiththecommon-sensicalviewofhowpictureswork.Commonsensesaysthatpictorialrepresentationconsistsinresemblance:apaintingrepresentsXbylookinglikeX.Thismayatfirstseemtobenotjustself-evident,buttheonlypossibleanswer.Thiscommon-senseviewofpictorialrepre-sentationisreflectedintheideathatdeceptionprovidesthetestforsuccessfulrepre-sentation–asillustratedbythestoryofZeuxis’spaintingofvines,which,accordingtoPliny,deceivedthebirdsintoflyingdowntoconsumethegrapeshehaddepicted.Itis,however,simplyandplainlyfalsethatcanvasesmarkedwithpigmentsresem-blethethingsthattheyrepresent.TheRapeofEuropadoesnotevenbegintosharethephysicaldimensionsofthescenethatitrepresents!Infact,eachcanvasresemblesnothingsomuchasothercanvases–butitdoesnot,ofcourse,representthem.Thisobjectionmayseemtooblunt,foritmayseemtohavemissedtheobviouspointthatforacanvastorepresentsomething,itmustbeseen,notasamerephysicalobject,242\nAESTHETICSbutasapicture.Butthisjustrestatestheproblem,whichnowbecomes:whatisitforacanvastobeseenasapicture?Itmaythenbeventuredthatwhattheresemblancetheoryreallymeans,isthatapaintingofXrepresentsXwhenitaffordsusanexperiencewhichisliketheexperienceofreallyperceivingX.Sometimesthisisputbysayingthatthecanvasgivesusan‘image’whichisthesameasthatwhichwewouldreceivefromtherealthing.Relocatingtheresemblanceatthelevelofanexperienceorimage,ratherthanthatofthecanvas,doesnot,however,advanceourunderstandingfor,onceagain,whatiswantedisanaccountofhowthetwo-dimensional,differentiatedsurfaceofTitian’spaintingcanbemadeto‘havethelook’–affordthevisualexperienceorcreatetheimage–ofawomancarriedonthebackofabull.Veryquicklyitcomestoseemthatresemblancedoesnotprovidetheexplanationofpictorialrepresentation.Initsplace,theoriststurntoother,lessobviousnotions.(However,forarecentdefenceofpictorialrepresentationintermsofexperiencedresemblance,seePeacocke1987.)Thecaseagainstresemblancewasfirstsetout,atlengthandwithsophistication,byGombrich(1960).Gombrichattackedinparticulartheassumption,whichaccompa-niestheresemblanceview,thatthereissuchathingasan‘innocenteye’–thatanartistcansimply‘copywhathesees’byattendingtohisvisualexperience,inadvanceofanyinterpretationoftheworld.Goodman(1976:ch.1)carriesthislinefurtherbyrejectingaltogethertheideathatpictorialrepresentationisamatterofperception.OnGoodman’saccount,depictionisaspeciesofdenotation:knowingwhatapicturerepresentsispurelyamatterofinter-pretation,whichrequiresagraspofa‘symbolsystem’.Realisminpaintingisjustafunctionofthefamiliarityofasymbolsystem,theeasewithwhichitimpartsinformation.Aswassaidearlier,themotivationforGoodman’ssemioticviewisverygeneral.Thereis,however,muchtoqueryinhisclaimthatpicturesareinterpreted.Itseemstousthatpicturesgiveusvisualexperiences,andthatthisishowtheydifferfromprosedescriptions,maps,companylogos,roadsignsandothervisualsymbolsthatneedtoberead.Goodman’stheoryimpliesthatthisimpressionofdifferenceisanillusion.Butifso,whatdoestheillusionofhavingavisualexperienceconsistin–ifnotinhavingavisualexperience?Gombrichhimselfdoesnotassimilatepictorialrepresentationtointerpretationcom-pletely.OnGombrich’saccount,apaintingrepresentsXbygivingusanillusionofX.Howillusionsarecreatedisexplainedintermsoftherepresentationalpracticesevolvedinthehistoryofart.Artdevelopsthroughaprocessof‘makingandmatching’.Visual‘schema’areoriginallyfabricatedsemi-arbitrarilyinordertorepresentobjects,withoutresemblanceplayinganykindofrole.Theyarethenmodifiedbyartistsinthelightoftheirperceptualexperience,ahistoricalprocesswhichresultsinincreasinglylife-likerepresentations(aprocesswhichGombrichmodelsonPOPPER’S(pp.287–90)concep-tionofthedevelopmentofscientifictheories).Resemblancelies,ifanywhere,attheendofthehistoryofpictorialart,notatitsbeginning.Itistruethatsomevisualrepresentations–suchasEscher’strompel’oeildrawings–createillusions,butitisnottruethatmostpaintingshavethiseffect:inlookingatapaintingwedonotfindourselveshavingtocorrectatendencytomistakeitscontentfor243\nSEBASTIANGARDNERreality,astheconceptofillusionwouldimply.Furthermore,theillusiontheoryentailsthatinseeingwhatapaintingrepresents,westopbeingawareofthematerialcanvas,astheillusiontakeshold.This,however,contradictsthefactthatappreciatingapaint-inginvolvessimultaneous,complexawarenessofitsbrushworkandwhatitrepresents(seeWollheim1974:ch.13).RepresentationandSeeingInTheweaknessesoftheviewsconsideredsofarsuggestthatweneedtopositaformofvisualperceptionwhichis,asitwere,inherentlyimaginative.Wollheim’stheorydoesthis(Wollheim1980:essay5;1987:46–77).ItproposesthatpictorialrepresentationengagesaspeciesofperceptionwhichWollheimcalls‘seeingin’.Wollheimarguesforthisclaimasfollows.Animportantclueastotheunderlyingnatureoftheexperienceofpaintingliesinthefactthatpaintingsgiveusexperiencesofsomethingabsentornon-existent.Nowthisissomethingthatoccursalsoindreams,daydreamsandhallucina-tion.Itmaybeventuredthatpictorialrepresentationexploitsandcultivatesapowerthatthemindpossessesinnately:thepowertogeneratevisualexperiencesoutofitself.Whenthiscapacityisexercisedinthecourseofperceivingtheexternalworldandfusedwithperceptionofexternalobjects–asitisinnumerouscontextsoutsidepainting:thingsareseeninRorschachink-blots,cloudsanddamp-stainedwalls–wethenhaveseeingin.Seeinginisnotinvolvedinreadingmapsorinterpretingvisualsigns;herethereisnoexperienceofanabsentobject.Onthistheory,whatitisforVermeer’sViewofDelfttorepresentatownscapeisforthecanvastobemarkedinawaythatallowsustoseeatownscapeinit.Seeingindoesnotpresupposearesemblancebetweenthecanvasandwhatthepaintingrepresents;accordingtoWollheim,nosystematicaccountcanbegivenofhowthemarkingofacanvasdetermineswhatisseeninit.Wollheim(1987)proceedstodevelopageneralaccountofpaintingwhichshowshowitsvalueasanartpresup-posesitsrootednessinseeinginandotherpsychologicalpowers.3.2ExpressionItisnotlongbeforethefuguerecedesintoitsinitialcalm.Soon,afinalcrescendoleadsustowhatappearstobeatriumphantconclusion.Butno;attheverymomentwhenvictoryseemstobeinourgrasp,themusiclosesallconfidence,comestoahaltonachordthatisfarfromfinal,andthensinksdespondentlybackintotheminorkey.Coulddisillusionmentbemoregraphicallyexpressed?Itisthecryofdespair;thehoped-forconsolationhasfailedtomaterialize.(Hopkins1982:383)ThispassagedescribesemotionsexpressedinthefinalmovementofBeethoven’sPianoSonataOp.110.Theexpressivequalitiesofworksofartareasub-classoftheiraestheticqualities,distinguishedbytheemploymentoftermswhoseprimaryuseliesindescribingpeople’semotions.Expressivequalitiesareintegraltothebeautyandmeaningofmusic.Sincethenineteenthcentury,manyhavebelievedthatmusic’sincomparablepowerofexpressionsecuresforitapositionofsupremacyamongthearts(WalterPaterclaimedthat‘allartconstantlyaspirestowardstheconditionofmusic’).244\nAESTHETICSDescribingmusicinexpressivetermsexemplifiesabroadtendencytotalkaboutworksofartasiftheyweresentiententities:theyaredescribedashaving‘organicunity’,‘vitality’andsoon.Andyetwedonot,ofcourse,reallybelieveormeantoimplythatworksofarthavefeelings.So,giventhatamusicalworkisatoneleveljustasequenceofsounds,justasapaintingisnothingbutapigmentedcanvas,thefunda-mentalproblemofexpressioninart,broughtoutmostsharplybymusic,isthefollow-ing:howisitpossibleforaworkofarttoexpressemotion?Theobviousanswerwouldseemtobethatmusicalworkshaveexpressivequalitiesbecausetheyserve,astheexpressiontheoryofartsays,asvehiclesforexpressingtheemotionsofthecomposer.Thisis,however,noanswer.Eveniftheexpressiontheoryofartisaccepted,nothinginitexplainshowitispossibleforacomposertousepatternsofsoundtoexpressemotion.Itmerelytakesthatfactforgranted.Ifitdoesnothelptothinkoftheemotionexpressedbymusicaslocatedinthemindofthecomposer,itisequallymistakentoattempttolocateitinthemindofthelistener:thatis,toanalyseamusicalwork’sexpressionofanemotionasitspowertoarousethatemotioninthelistener.Bluntly:wearenotourselvesmademelancholic,orfrozenwithgrief,bymusicthatwewoulddescribeinthoseterms–fortunately.Thelis-tenerdoesnothavetheemotionwhichthemusicexpresses,inthesamesenseasshehasemotionsinlife.Inlisteningtomusic,theemotionwhichisexpressedseemstobelocated‘outthere’inthemusic,notinoneself.Theheartoftheproblemliesinunder-standinghowemotioncanbe‘objectified’inthisfashion–howitcanbeascribedneithertothecomposernortothelistener,butrathersomehowhover,suspended,betweenthem.Furtherreflectionshowsthatbeingmovedbymusicconsists,typically,notinsimplyrecognizingorsharingtheemotionexpressedbymusic,butinreactingtoit–inthewaythatapersonmayrespondtotheemotionalconditionofanother,forexample,byrespondingtotheirdistresswithpity.Therelationisoneofsympathyratherthanempathy.Wearefamiliarwiththisstructureinthecontextofrepresentationalart,where,indrama,literatureandpainting,wefeelfortheprotagonist.Itsappearanceinmusicreinforcesthedifficultyofmusicalexpression,sinceinstrumentalmusicpreciselylackstherepresentationalcontentthatwouldmakeemotionalresponsetomusicintelligibleasacaseofparticipatinginafiction(ithasnocharactersorplot).Atthispoint,oneoptionistoreconceivetheproblemofmusicalexpressionasmerelyoneoflanguage:offindinganadequatelogicalcharacterizationofhowtermsareusedintheascriptionofexpressivequalities.Anaccountofthissort–whichappearstocutratherthanuntietheGordianknot–isagainadvancedbyGoodman(1976:85–95).Analternativewayoutoftheproblemismusicalformalism,apositiondefendedbyHanslick(1986).Hanslickgrantsthatmusichasemotionaleffectsonlisteners;buthedeniesthatthereisanygenuinesenseinwhichmusicexpressesemotion.Hanslickhasbothanargumentfordenyingthatmusiccanexpressemotion,andanaccountofwhyweshouldcometosupposefalselythatitcan.Formusictoexpressemotion,therewouldhavetobeanintrinsicrelationbetweenthem.Hanslickarguesthatthisisimpossible,onthegroundsthatemotionsareintentional:theyhaveobjectsandinvolvethoughtsandconcepts;whereasinstrumentalmusiclacksobjectsanddoesnotinvolvethoughtsorconcepts.Sincemusicisincapableofpresentinganyobject–whatitpresentsisonlyitself–itcannothaveanintrinsicrelationtoemotion.245\nSEBASTIANGARDNERWhatmusiccando,Hanslickargues,isexpresswhathecalls‘musicalideas’,whichareitsdynamicproperties:‘Thecontentofmusicistonallymovingforms’(ibid.:29).Tonalityprovidesasortofauditoryspacewithinwhichthemelody,harmonyandrhythmofmusicmove(seeScruton1983:ch.8;1997:chs2–3).Hanslickdrawsananalogywiththevisualmovementfoundinarabesqueornamentation.Likemusic,episodesofemotionhavedynamicproperties,inadditiontoandinde-pendentlyfromtheirconceptualcontent,sincetheyinvolvechangesinintensityandbodilyfeeling.Becausemusicandepisodesofemotionssharedynamicproperties,itispossibleforustofillin,arbitrarily,themusicalideaexpressedbyapieceofmusicwithsomeparticularemotionsuppliedbyourownemotionalrepertoire:givingustheillu-sionthatthemusicexpressesthatparticularemotion.ForHanslick,musicalapprecia-tionstopsproperlyatthebeautyofmusicalform.Anythingelseismerelypersonalemotionalassociation.Despiteitssubtlety,Hanslick’srepudiationofemotionisnoteasytoaccept:describ-ingmusicalworksintermsoftheirexpressivequalitiesseemstobe,notjustnatural,butindispensableforjustifyingourestimatesoftheirartisticvalue.Thethoughtthatmusicisintrinsicallyexpressiveshouldnotbegivenupunlessabsolutelynecessary.Cananythingbesalvaged?Althoughmusicdoesnotspeakorbehave,ananalogybetweentherecognitionofemotioninmusicandinpeoplecanbesuggested.Perhapsmusicderivesitsexpressivepowerfromitsresemblancetoemotionallyexpressivefeaturesofhumanbodilybehav-iour,physiognomyandspeech(Kivy1989).Thatwouldexplainthephenomenologicalsimilaritybetweenhearingemotioninmusic,andperceivingemotioninthebodyorvoiceofanotherperson:inbothcases,mentallifeseemstohavebeenmadesensuouslypalpable.ThistheoryisinapositiontodeflectHanslick’sobjectionthatmusiclackstheconceptualcomponentsnecessaryforanintrinsicconnectionwithemotion.Bodilyandvocalfeaturescanexpressemotionalstateswithoutcommunicatingtheirobjects:wecanrecognizethatapersonisdistressedfromthewaytheylookorsound,withoutknowingwhattheyaredistressedabout.Thehumananalogytheorysoonencountersdifficulties,however.Itsharestheweaknessoftheresemblancetheoryofpictorialrepresentation–namely,thedifficultyoflocatingrelevantresemblances.Whatexactlyarethefeaturessharedbymusicandhumanbodies?Justascanvasesareunlikethethingsrepresentedinpictures,aplaindescriptionofmusicintermsofkey,rhythmandsoonrevealslittlethatiscommontomusicandhumanbodies.Itistruethatbothtempoandbodilymovementcanbeslow,butsuchpropertiesarenomoreemotionallydefinitethanHanslick’s‘musicalideas’:ifasadpieceofmusicisslow,whatmakesitsslownessthatofsadnessratherthanseren-ity?Theobjectionthenarisesthatsignificantlikenessesbetweenmusicandhumanbodiescanonlybeidentifiedafterthemusichasbeenredescribedinexpressiveterms,suchas‘ponderouslyslow’.Onlyoncemusichasbeenexperiencedasexpressivedoesitbecomepossibletomodelitimaginativelyonthehumanfigure.Musicalexpressionremainsunexplained.AsecondtheorythatappealstoaresemblancebetweenmusicandemotionisthatofLanger(1942:ch.8).OnLanger’stheory,musicalworksaresymbolsofaspecialkind.Unlikethediscursivesymbolismoflanguage,musicsharesthe‘logicalform’of246\nAESTHETICSemotionand,byvirtueofdoingso,articulatesandpresentsemotion,aslanguagecannot.Hereitistheinneraspectofemotion,ratherthanitsouterbodilyexpression,whichmusicissaidtomirror(‘Musicisourmythoftheinnerlife’:ibid.:245).Asimilarproblem,however,confrontsLanger’stheory.Inwhatnon-metaphoricalsensedoesemotionhave‘logicalform’?Andinanycase,thenotionofformalsimilar-ityissoplasticandopen-endedthatitcansurelybefoundbetweenanyemotionandanypieceofmusic.Again,theallegedisomorphismofmusicwithemotionseemstobeeithernon-existent,orinsufficientlydefinitetoexplainmusicalexpression.Ifneithertheouternortheinneraspectofemotionhasaconnectionwithmusicwhichaccountsforitsexpressiveness,itmayseemthatenquirygrindstoahalt.Budd(1985)concludes,afterintensiveexaminationoftheabovetheories,andothers,thatnonesucceeds).Oneremainingavenue,however,istosupposethattheconnectionofmusicwithemotionismoredirectthanwehavehithertoassumeditmustbe:thatitdoesnottakeadetourviaresemblance.Acluetotheexplanationofexpressivequalitiesmaylieinthefactthatthemindhas,asHumeputit,anaturalpropensitytospreaditselfontotheworld,andexperi-enceobjectsasemotionallycoloured.Projection,asthismaybecalled,isawell-attestedpsychologicalphenomenon.Ifthisisjoinedwithasecond,equallyHumeanassump-tion,totheeffectthatcertainobjectsintheworldarefittedbynaturetoserveasrecipientsfortheprojectionofspecificemotionalstates,thenweareonthewaytoaconceptionofartisticexpressionasreposingon,andcultivating,anaturaltendencyofourstodyeobjectsintheworldwithouremotions(Wollheim1980:sections15–19;1987:80–7).IfHanslick’srepudiationofemotionistobeavoided,anassumptionalongtheselinesseemstobeatleastnecessary.3.3MeaningThatworksofarthavemeaning–overandabovetheirrepresentationalcontent,andaestheticandexpressivequalities–isimpliedbythefactthatworksofartallowfor,andcallfor,understanding.QuestionsofmeaninginartarebestapproachedthroughtheconceptofINTERPRETATION(pp.384–90).Interpretationaimstoidentifythemeaningsofworksofart,andisoneofthefunctionsofcriticism.Questionsofinterpretationariseofcourseinallthearts–isBeethoven’sMissaSolemnisaChristianoraRoman-ticwork?DoesTitian’sTheFlayingofMarsyuscelebratethetriumphofspiritoverbody,orthatofbodyoverspirit?–butitisliterature,andliterarycriticism,thatputtheissuessurroundinginterpretationinsharpestfocus.Ofallthearts,literatureallowsleastscopeforaestheticresponsetooperateinde-pendentlyofinterpretation:alikingforaliteraryworknotaccompaniedbysomuchasarudimentarygraspofitsmeaningfailsaltogethertoqualifyasanappreciationofit.Thisisduetothefactthatthemediumofliterature,naturallanguage,is,unlikepigment,soundorstone,inherentlymeaningful;which,plausibly,allowsliteraryworkstobearmoreconceptuallycomplexmeaningsthanworksinothermedia;forwhichreasonliteraryworksgenerategreaterinterpretativecontroversy.Itishardtofindagreatliteraryworkforwhichdifferentinterpretationshavenotbeenproposed.IsHamletcorrectlyunderstoodintermsofFreud’stheoryoftheOedipuscomplex?DoesParadiseLostrepresentSatanasamoralbeing‘farsuperiorto247\nSEBASTIANGARDNERhisGod’,asShelleyclaims?DoShakespeare’ssonnetscompriseaunifiednarrativechartingthecourseofShakespeare’sownexperience?AreKafka’swritingsreligiousorpoliticalparables,orarticulationsofthepersonalangstevidentinhisdiariesandletters?DoesaproperappreciationofYeats’spoetrypresupposeknowledgeofhismytho-logicalsystem?Toeachofthesequestionscriticshavereturnedoppositeanswers.Twoquestionsareforcedonusbydifferencesininterpretation.Firstly,canbio-graphicalinformationconcerningtheauthorserveaslegitimateevidencefororagainstacertaininterpretation?Secondly,canwetalkofoneinterpretationasbeingthecorrectinterpretationofawork?Thestarting-pointforthefirstquestionistheattackmadebyWimsattandBeards-leyonwhattheycallthe‘intentionalfallacy’(inNewton-deMolina1984).Ontheirview,criticserrinsupposingthatinformationabouttheauthor,oranykindofevidence‘external’toapoetictext(forinstance,aboutitsoriginalculturalcircumstances),canelucidateitsmeaningasaliteraryobject;allthatispertinenttothemeaningofapoemcanbegleanedbycarefulattentiontothewordsonthepage.Tosupposeotherwiseistocommitthefallacyofconfusingapsychologicalfactabouttheauthorwithafactaboutthepoem,offailingtodistinguishacausalexplanationofaworkfromaliteraryinterpretationofit.Twoopposingviewsoftherelationbetweenthemeaningofaliterarywork(textualmeaning)andtheintentionsoftheauthor(authorialmeaning)canbestakedout.Intentionalismidentifiestextualmeaningwithauthorialmeaning:whatthetextmeansiswhattheauthormeant(Wollheim1980:essay4).Anti-intentionalismdeniesthattextualmeaningisauthorialmeaning,andassertsitsautonomy:textualmeaningresidesobjectivelyinthework,‘embodiedinthelanguage’,andhasnothingtodo,con-ceptually,withwhattheauthormayhavemeant(ifthetwohappentobethesame,thatisacoincidencewithoutsignificance,fortheanti-intentionalist).Thesepositionshaveconnectionswithtraditionaltheoriesofart:theexpressiontheoryisstraightfor-wardlycommittedtointentionalism,formalismtoanti-intentionalism(themimetictheoryisuncommitted).What,then,arethemainconsiderationsintheargumentoverintentionalism?Thereismuchtobesaid,butsomeofthemostimportantpointsarethefollowing.Ontheonehand,itseemsrightthataliterarywork,consideredasaworkofart,shouldbejudgedintermsoftheexperienceofreadingit;ifmeaningsdonotshowupinanoptimallysensitivereadingofthetext,then,althoughtheauthormayhavewishedtoconveythem,theysimplydonotformpartofwhatthetextmeans.Apoemstandsorfallsonitsownmerits.ThisisWimsattandBeardsley’sstrongestargumentforanti-intentionalism.Itcan,however,becounteredbytheintentionalist.Whileitistruethattheauthor’sintentionsmustberealizedinthetextandconcretelyapprehendedthereinorderfortheworktobeartisticallysuccessful,whenaworkissuccessful,intentionalismistrue:whatwegraspistheauthor’smeaning.Theintentionalistmayaddthat,although‘external’evidencecannottransformanartisticfailureintoasuccess,attentiontosuchevidencemaymakeiteasiertoseewhatmeaningiscontainedinthewordsonthepage.Theintentionalistcanalsopointoutthat,sincewerarelyapproachliteraryworkswithoutsomeawarenessoftheauthor’shistoricalcircumstances,otherwritingsand248\nAESTHETICSsoon,thereisinfactnofirmordeepdistinctionbetweeninternalandexternalevi-dence,contrarytowhattheanti-intentionalistsupposes.Theintentionalist’scasecanbestrengthenedfurtherbydrawingattentiontothefactthatestimatesofsincerity,maturityandperceptivenessareimportantforourresponsestoliteraryworks.Literatureisflawedifitexhibitsmawkishness,self-indulgenceorcrudityofmoralvision;thefineconsciousnessoflifeandethicalcomplexityofGeorgeEliotandHenryJamesisvalued.Literaryvicesandvirtuesarequalitiesofmoralpersonalityapparentlyascribedinfull-bloodedintentionalistspirittotheauthor,insofarasheorsheismanifestinhisorherwork.(Amplifyingthisapproach,seeLeavis1986:part4.)Theanti-intentionalistwillcomplainthatheretheintentionalisthasonceagaincon-vertedanaestheticintoanon-aestheticmatter,andistreatingthetextasmaterialforliterarybiography,asifitwerethecritic’sjobtopassmoraljudgementontheauthor.Wedovaluequalitiessuchassincerityinliterature,butthesearepropertiesoftexts:theyindicatestrengthsintheworkitself.Wehave,then,intuitionsfavouringeachofintentionalismandanti-intentionalism.Neithersetofargumentsisobviouslydecisive.WimsattandBeardsleydidnot,there-fore,establishthatintentionalisminvolvesafallacy.Theydid,however,defineanti-intentionalismasatheoreticaloption,andindicatethemethodologicalrepercussionsoftheissue.Ifintentionalismistrue,thenthestudyofliteraturemaycastitsnetwideanddraw,inadditiontoliterarybiography,onotherdisciplines–history,psychologyoranthropology–concernedwithfactorscontributing,consciouslyorunconsciously,totheauthor’smeaning.Ifanti-intentionalismistrue,thenexternalinformationoughttobebracketedout.Thesecondquestionraisedbydifferencesofinterpretation,whichconcernsthenotionofcorrectness,canagainbeformulatedintermsofasharpopposition.Onthemonisticview,thereisforeachworkauniquelycorrectinterpretation.Whetheritcanbeidentifiedconclusively,andgainuniversalacceptance,isanothermatter.Monistssaythatwhereinterpretationsdiffer,andyetseemequallywell-supported,theirconflictcanbeovercomebya‘super-interpretation’,whichincorporateswhatistrueineachoftheinterpretationswhichitsupersedes.Pluralismdeniesthatthereis,eveninprinciple,auniquelycorrectinterpretationforeachwork;itmaintainsthatanumberofdifferentinterpretationscanmeetthecon-ditionof,notcorrectness,ifcorrectnessistakentoimplyuniqueness,but‘legitimacy’.Pluralistsdifferoverthecriteriaforlegitimacyofinterpretationandhowrestrictivetheseshouldbe.Theargumentsbetweenmonismandpluralismengagecloselywiththeissueofintentionalismversusanti-intentionalism.Thepluralistfocusesinthefirstinstanceonthenotionoftextualmeaning.Cor-rectnessofinterpretationpresupposesthattextualmeaningisdeterminate.WimsattandBeardsleyhelpthemselvestotheassumptionthatliteraryworkshaveobjective,determinatemeanings,butthepluralistchallengesthemonisttosaywherethisdeter-minacycomesfrom:forevenifthemeaningofeachindividualwordcomposingatextisdeterminate–aclaimwhichisitselfvulnerableinviewoftheubiquityofmetaphorinliterature–themeaningoftheworkasawholeisfarfrombeingastraightforwardfunctionofthedictionarymeaningsofthewordscomposingit.Aswewouldordinar-ilyputit,thereis‘roomforinterpretation’,andthis,accordingtothepluralist,means249\nSEBASTIANGARDNERthatthereisroomforanumberofinterpretations.Easiest,then,isforthemonisttoembraceintentionalism:since,ifintentionalismistrue,asourceofdeterminacyisreadilysuppliedbytheauthor’sintention.Thepluralist’ssecondargumentdrawsonatenetofaestheticsubjectivism:namely,thatfeltresponseisatthecoreoftheaesthetic.Literaryinterpretationshouldservethekindofinterestappropriatetoliteraryworks,andliteraryinterest,asacaseofaestheticinterest,centresonresponse.Thepropergoalofinterpretationisthereforetoenhanceourexperienceofthework,toofferareadingwhichmakesitmeanasmuchaspos-sibletous.Sincetheperspectivesofreaderswillalwaysdiffer,theinterpretationthatisoptimalforonereaderwillnotbeoptimalforanother.Multipleinterpretationsofaworkarethereforetobeexpected,andshouldbewelcomed.Tomeetthisargumentonitsownterms,themonistwouldhavetodefendthe–evidentlycontroversial–claimthatthereisforeachtextoneinterpretationwhichisuniversallyoptimal.Argumentsbetweenmonismandpluralismtendtoconvergeonthefollowingsce-nario.Themonistsaysthatifpluralismistrue,then‘anythinggoes’ininterpretation,animplicationwhichthemonistclaimsreducespluralismtoabsurdity–TheAncientMarinercannotbeinterpretedasapoemaboutissuesofpowerandgenderinlatetwentieth-centuryculture.Thepluralistobjectsthatthenotionofauniquelycorrectinterpretationfostersdogmatism,anddisregardsboththerichnessoftextsandthediversityofhumaninterestswhichmaylegitimatelybeexpressedinliteraryinterpretation.These(unresolved)philosophicalissuesgainaspecialurgencyfromtheirrelevancetothepresent,turbulentandlogomachicclimateofliterarystudies,wheretheoreticalandideologicalcommitmentsareexplicitlyadoptedandappealedtoininterpretingtexts.Thedominantdeconstructionistconceptionofliterarymeaningatworkhereisfundamen-tallyaformofpluralism,adoptedonwhollygeneralphilosophicalgrounds:theveryconceptofdeterminacyofmeaningisrejected,inall,notjustliterarycontexts.Tothisisaddedaradicallyidealistviewofliterarymeaningwhichreversestheordinaryconcep-tionoftherelationofmeaningandinterpretation:themeaningofaliteraryworkisconceivedascreated,ratherthangraspedbyinterpretation.Thepointofliterature,onthispicture,isnotaesthetic:literatureexistsforthesakeofinterpretation.4ThePointofArtWecomefinallytothequestionofthepointofart.Whatvaluedoesarthave?Whatisartfor?Thisisnotademandfora‘justification’ofartinthesamesenseasajustifica-tionofmoralityisthoughttoberequired–thatwouldbeaninappropriatedemand,sincethecreationandappreciationofartdonotsharethenecessityofmoralobliga-tion.Thequestionisratherwhetherwecanarticulateandvindicateoursenseoftheimportanceofart.Thefact,observedearlier,thatartisanevaluativeconcept,meansthatanaccountofthepointofartisnotamerecodatoaestheticenquiry:itshouldfalloutconceptuallyfromatheoryofart.Thatartdoeshavesomepointisnotbeyondalldoubt(Passmore1991).Ithasalreadybeenseenhowthemimetictheoryencountersdifficultiesonthisscore.Posi-tivereasonsforscepticismaboutthevalueofartarenothardtounearth:theyderivefromart’sessentialconnectionswithpleasure,playandimagination,itsfreedomfrom250\nAESTHETICSreasonandpracticalpurposes.Thescepticwillinsistthatarthasthesamesortofvalueasanyotherformofentertainment.Plato’snotoriouscritiqueofart,inbooktenoftheRepublic,goesfurther,bysuggestingthatart’spreoccupationwithappearanceentrenchesourignoranceofreality,andthatitseffectsmaybepsychologicallyandmorallyharmful(seeJanaway1995).Thedistinctionofendsandmeansfiguresprominentlyinmanydiscussionsofthepointofart,andisusuallyemployedcritically:someaccountsaresaidtoreducearttoa‘mere’means,otherstorecognizecorrectlythatartisanend.Thiscontrastis,however,notaltogetherfelicitous,fortworeasons.Firstly,art,solongasithassomevalue,canalwaysberedescribedasameanstowhatevervalueitrealizes.Secondly,itisnotclearwhatcanbemeantbydescribingart,oritsapprecia-tion,asanend.Certainlyindividualworksofarthavetobeapproachedintheirowntermsandcontemplated‘fortheirownsake’;butitwouldbeamistaketomovefromthisfactaboutaestheticattention,thatitterminatesinitsobjectanddoesnot‘thinkahead’,totheclaimthatartisitsownpointandcannothaveextrinsicvalue.Formal-istssuchasBellandHanslick,whodescribeworksofartasendsinthemselves,seemtomakethismistake.Whatthedenialthatartis‘merelyameans’–andtheaestheticistslogan‘artforart’ssake’–maybetryingtosay,isthatthevalueofartrealizedbyartcannotbereal-izedbyanythingelse,orcannotberealizedinthesameway.Thusinterpreted,thedoctrinemeetswithouragreement:werecoilfromhedonistic,moral,didacticorotherinstrumentalattitudestoart,totheextentthatthesesuggestthatotherthingscouldbesubstitutedforartwithoutloss,orthatthecomplexityofartisredundant.Tolstoy’s(1930)theoryofartasthetransmissionofmoralfeeling,forexample,makeslittleefforttodistinguishartfromother,potentiallymoreefficient,waysofachievingthatend.Itfollowsthatsolongasartisviewedashavinganecessaryroleinrelationtothekindofvalueinquestion,andthecomplexityofartistakenaccountof,thereisnothingnec-essarilywrongwithassigningahedonistic,moral,didacticorothergoaltoart.Schiller(1989),forinstance,arguesthatthegoalofart,asacomponentof‘aestheticeduca-tion’,istoovercomethemetaphysicalcontradictionsinhumannatureinorderthatwemayattainfull‘humanity’;thisgoalis‘extra-aesthetic’,butSchillerregardsaestheticeducationastheonlywayinwhichitcanbeachieved.Whatformsmayaccountsofthevalueofartthentake?Theymay,first,bedividedaccordingtohowcloselytheyrelatethevalueofarttothevaluesoflife.InaperspectivesuchasthatofLeavis,artisproperlyandinextri-cablyboundupwiththevaluesoflife:the‘raisond’êtreofthework’isto‘haveitsdueeffectandplayitspartinlife’;the‘essentialbusinessofcriticism’istolocate‘thecreativecentrewherewehavethegrowthtowardsthefutureofthefinestlife’(Leavis1986:283).Attheotherextreme,Bellholdsthatartisautonomousandhasvalueonlytotheextentthatitdistancesusfromlife:‘toappreciateaworkofartweneedbringwithusnothingfromlife,noknowledgeofitsideasandaffairs,nofamiliaritywithitsemotions...Inthisworldtheemotionsoflifefindnoplace.Itisaworldwithemotionsofitsown’(Bell1914:25–7).Thedeeperdistinctiontobemadeamongaccountsofthevalueofart,however,isbetweenthosethatarenaturalistic,andthosethataremetaphysical;adistinctionthatwasforeshadowedbythecontrastofHumeandKant.251\nSEBASTIANGARDNERNaturalisticaccountsgroundartinhumannature.ThecontingentfittednessbynatureofcertainobjectstoourmindsproposedbyHumeprovidesthestartingpoint.Toexhibitthepointofart,itneedstobeshownhowworksofartengagefundamentalandimportantmentalactivities(Dewey1934).Psychoanalysisprovidesoneexampleofthisapproach(Wollheim1987).Thequestionofthepointofartisthereforeansweredbysayingthatartisnatural:artexistsbecauseitisnaturalforustocreateandappreciateit.Assuch,artisa‘formoflife’(Wollheim1980),asnecessaryacomponentofhumanexis-tenceaslanguage,cultureandpolitics.Thenaturalisticviewaffirmsthatartrespondstothepsychologicalneedsofhumanbeings,butitdeniesthatthefunctionswhichartper-formscouldbespecified,orfulfilled,throughothermeans.Whenartispicturedasanec-essarypartofthehumanorder,andtheappreciationofartasanaturalcomponentofhumanwell-being,thepointofartmergeswiththatofhumanlifeitself.Thenaturalisticviewofartthereforequalifiesasafifththeoryofart,inadditiontothoseconsideredearlier(itmay,butneednot,beformulatedsoastoincorporatetheexpressiontheory).Byshowinghowthevariousdimensionsofartmapontopsycho-logicalprocesses,itaccountsfortheunityofartintermsoftheunityofthemind,andlocatestheessenceofartwithreferencetohumanpsychology.Metaphysicalaccountsofthepointofartare,ofnecessity,morespeculativeandlesspinneddowntotheempiricalfeaturesofartthannaturalisticaccounts.Theywillalso,ofcourse,dependexplicitlyonageneralphilosophicaloutlook,whichnaturalisticaccountsdonotneedtodo.Thisdoesnot,orshouldnot,meanthattheysimplysqueezeartintoapreformedmetaphysicalsystem:theymay,onthecontrary,allowartaroleinformingthesystemitself.Thisiswhatwefind,indifferentways,inKant,Hegel,Schopenhauer,SchillerandNietzsche(Germanidealismhasbeennotablymoreaccommodatingtoartthananyotherphilosophicaltradition).Thebasicdemandtowhichmetaphysicalaccountsofartanswer–andwhichnat-uralisticaccountscannotproperlysatisfy–derivesfromwhatmightbecalledthe‘trans-figurative’aspectofourexperienceofart:thesensethatthetransformationofrealitywhicharteffects,andwhichlocatesinitakindofvaluethatwefindconsoling,ismorethanafancifulembellishment.Tragedyexhibitsthetransfigurativepowerofartmostclearly.Theaspirationtoachieve,throughart,ajustificationoftheworld‘asanaes-theticphenomenon’,asNietzsche(1993:32)putit,isafurtherpartofthelegacyofRomanticism,andhasnotdisappearedfromthedemandsthatweputonart.FurtherReadingAsanintroductiontoaesthetics,RichardWollheim’sshortArtanditsObjects(1980),whichoffersbothanoverviewandadistinctiveapproachtothesubject,andMalcolmBudd’sValuesofArt(1995)arebothexcellent.Other,plainerintroductionsareCharlton(1970)andSheppard(1987).Hanfling(1992)containsaseriesofessaysspecificallywrittenasaunifiedintroductiontoaesthetics,butofsufficientlengthtopresentmanyargumentsindetail.Itisimportanttohaveagraspatfirst-handoftheclassicwritingsinaesthetics,fewofwhichmakedifficultreading.HofstadterandKuhns(1964)isafirst-rate,currentlyavailableanthology.Carritt(1931)ismorecomprehensivebuttheextractsareshorter.Twootherexcellentantholo-giesthatalsoservethisend,andintersperseselectionsfromtheclassicswithmodernwritings,areDickieandSclafani(1977)andRader(1979).Bothareorganizedthematicallyandcontainhelpfulbibliographies(asdoesWollheim1980).Thehistoricaldevelopmentofaestheticsistraced252\nAESTHETICSbyBeardsley(1966),whichisalsousefulasareferencebook.Cooper(1992)containshelpfulentriesonmostfiguresandtopics.Anumberofbooksmaybesingledoutasofferingmorecomplexdiscussionswiththestressonaparticulartopicorperspective.Danto(1981)isstimulatingandaddressesmanyproblems,butwiththeaccentfirmlyoncontemporaryart.Budd(1985)givesdetailedandlucidanalyticaltreatmentofsomecentralaesthetictheories.Wollheim(1974)bringsissuesinaestheticsinrela-tiontophilosophyofmindandpsychoanalysis.Scruton(1974)alsoemphasizesthecontribu-tionofphilosophyofmind,togetherwiththatofphilosophyoflanguage;Scruton(1983)relatesaestheticstoculturalissues.Goodman(1976)hasbeenhighlyinfluential.Savile(1982)focusesonquestionsofvalue,intheidealisttradition.Walton(1990)attemptstoprovideageneraltheoryofartisticrepresentation.ImportantarticlesarecollectedinDickieandSclafani(1977),Barrett(1965),Margolis(1978)andOsborne(1972).Withrespecttoparticulartopicsinaesthetics,asbrokendowninthischapter:Humeneedsnocommentary,butKant’sCritiqueofJudgementisextremelydifficult,anditisbesttoreadeitherKemal(1992)orMcCloskey(1987)alongside.ThreecomprehensivetreatmentsofaestheticexperienceandjudgementareBeardsley(1982),Scruton(1974),andMothersill(1984).DefinitionsofartareexploredinDavies(1991).Theoriesofart,andaccountsofitspoint,arebestapproachedthroughtheclassicwritings.Fordetailedtreatmentofthedebatessurroundingeachofthedimensionsofart,Schier(1986)dealswithpictorialrepresentation,Budd(1985)withexpression,andNewton-deMolina(1984)withliterarymeaningandinterpretation.ReferencesAristotle1987:Poetics(translatedbyS.Halliwell).London:Duckworth.Barrett,C.(ed.)1965:CollectedPapersonAesthetics.Oxford:Blackwell.Beardsley,M.C.1966:AestheticsFromClassicalGreecetothePresent.Tuscaloosa:UniversityofAlabamaPress.——1982:TheAestheticPointofView.Ithaca,NY:CornellUniversityPress.Bell,C.1914:Art.London:ChattoandWindus.Budd,M.1985:MusicandtheEmotions.London:RoutledgeandKeganPaul.——1995:ValuesofArt:Pictures,PoetryandMusic.Harmondsworth:PenguinBooks.Carritt,E.F.1931:PhilosophiesofBeauty.Oxford:ClarendonPress.Charlton,W.1970:Aesthetics.London:Hutchinson.Collingwood,R.G.1937:ThePrinciplesofArt.Oxford:OxfordUniversityPress.Cooper,D.(ed.)1992:TheBlackwellCompaniontoAesthetics.Oxford:Blackwell.Croce,B.1992[1902]:TheAestheticastheScienceofExpressionandoftheLinguisticinGeneral(translatedbyC.Lyas).Cambridge:CambridgeUniversityPress.Danto,A.1981:TheTransfigurationoftheCommonplace.Cambridge,MA:HarvardUniversityPress.Davies,S.1991:DefinitionsofArt.Ithaca,NY:CornellUniversityPress.Dewey,J.1934:ArtasExperience.NewYork:Putnam.Dickie,G.1984:TheArtCircle.NewYork:Haven.Dickie,G.andSclafani,R.J.(eds)1977:Aesthetics:ACriticalAnthology.NewYork:StMartin’sPress.Gombrich,E.H.1960:ArtandIllusion,2ndedn.NewYork:Pantheon.Goodman,N.1976:LanguagesofArt.Indianapolis:Hackett.Hanfling,O.(ed.)1992:PhilosophicalAesthetics.Oxford:BlackwellinassociationwiththeOpenUniversity.253\nSEBASTIANGARDNERHanslick,E.1986[1854]:OntheMusicallyBeautiful(translatedbyG.Payzant).Indianapolis,IN:Hackett.Hegel,G.W.F.1993[1820–9]:IntroductoryLecturesonAesthetics(translatedbyB.Bosanquet).Harmondsworth:PenguinBooks.Hofstadter,A.andKuhns,R.(eds)1964:PhilosophiesofArtandBeauty.NewYork:HarperandRow.Hopkins,A.1982:TalkingAboutMusic.London:Pan.Hume,D.1965[1757]:Ofthestandardoftaste.In‘OftheStandardofTaste’andOtherEssays.NewYork:Bobbs-Merrill.Janaway,C.1995:ImagesofExcellence:Plato’sCritiqueoftheArts.Oxford:OxfordUniversityPress.Kant,I.1987[1790]:‘Critiqueofaestheticjudgement’,PartIofCritiqueofJudgement(translatedbyW.S.Pluhar).Indianapolis:Hackett.Kemel,S.1992:Kant’sAestheticTheory:AnIntroduction.London:Macmillan.Kivy,P.1989:SoundSentiment.Philadelphia,PA:TempleUniversityPress.Langer,S.1942:PhilosophyinaNewKey.Cambridge,MA:HarvardUniversityPress.Leavis,F.R.1986:ValuationinCriticismandOtherEssays.Cambridge:CambridgeUniversityPress.McCloskey,M.1987:Kant’sAesthetics.London:Macmillan.Margolis,J.(ed.)1978:PhilosophyLooksAttheArts.Philadelphia,PA:TempleUniversityPress.Meager,R.1970:AestheticConcepts.BritishJournalofAesthetics,10,303–22.Moore,G.E.1984[1903]:PrincipiaEthica.Cambridge:CambridgeUniversityPress.Mothersill,M.1984:BeautyRestored.Oxford:ClarendonPress.Newton-deMolina,D.(ed.)1984:OnLiteraryIntention.Edinburgh:EdinburghUniversityPress.Nietzsche,F.1993[1871]:TheBirthofTragedy(translatedbyS.Whiteside).Harmondsworth:PenguinBooks.Osborne,H.(ed.)1972:Aesthetics.Oxford:OxfordUniversityPress.Passmore,J.1991:SeriousArt.London:Duckworth.Peacocke,C.1987:Depiction.PhilosophicalReview,96,383–409.Plato1955:Republic(translatedbyH.D.P.Lee).Harmondsworth:PenguinBooks.Rader,M.(ed.)1979:AModernBookofEsthetics,5thedn.NewYork:Holt,RinehartandWinston.Savile,A.1982:TheTestofTime.Oxford:ClarendonPress.Schaper,E.(ed.)1983:Pleasure,PreferenceandValue.Cambridge:CambridgeUniversityPress.Schier,F.1986:DeeperIntoPictures.Cambridge:CambridgeUniversityPress.Schiller,F.1989[1793–5]:OntheAestheticEducationofMan(translatedbyE.WilkinsonandL.A.Willoughby).Oxford:ClarendonPress.Schopenhauer,A.1969[1819]:TheWorldasWillandRepresentation,2vols(translatedbyE.F.J.Payne).NewYork:Dover.Scruton,R.1974:ArtandImagination.London:Methuen.——1979:TheAestheticsofArchitecture.London:Methuen.——1983:TheAestheticUnderstanding.London:Methuen.——1997:TheAestheticsofMusic.Oxford:ClarendonPress.Sheppard,A.1987:Aesthetics.Oxford:OxfordUniversityPress.Sibley,F.1959:AestheticConcepts.PhilosophicalReview,68,421–50.——1965:AestheticandNon-aesthetic.PhilosophicalReview,74,135–59.Sibley,F.andTanner,M.1968:Symposiumon‘AestheticsandObjectivity’.ProceedingsoftheAristotelianSociety,supplementaryvolume42,31–72.Tilghman,B.R.1984:ButisitArt?Oxford:Blackwell.Tolstoy,L.1930[1898]:WhatisArt?(translatedbyA.Maude).Oxford:OxfordUniversityPress.Walton,K.1990:MimesisasMake-Believe.Cambridge,MA:HarvardUniversityPress.Wittgenstein,L.1978:LecturesandConversationsonAesthetics,PsychologyandReligiousBelief.Oxford:Blackwell.254\nAESTHETICSWollheim,R.1974:OnArtandtheMind.Cambridge,MA:HarvardUniversityPress.——1980:ArtanditsObjects,2ndedn.Cambridge:CambridgeUniversityPress.——1987:PaintingasanArt.London:ThamesandHudson.DiscussionQuestions1Whatmakesajudgementaesthetic?2Aretheaestheticqualitiesofanobjectanylessobjectivethanitscolour?3Canargumentsaboutaestheticvalueberesolved,andwhatdoesitsignifyiftheycannot?4Canitbeshown,ormustitmerelybeassumed,thataestheticjudgementsdifferfromgustatorypreferences?5Inwhatways,ifany,canreasonsbegivenforaestheticjudgements?6Arethereanylimitsonwhatcanbeanobjectofaestheticattention?7Inwhatrespectsareaestheticjudgementslikemoraljudgements,andinwhatrespectsaretheyunlike?8Whatcancriticismofarthopetocontributetoourappreciationofart?9Inwhatrespectscanone’srelationtoworksofartbecomparedwithone’srelationstootherpeople?10Doesaestheticexperiencepresupposeanaestheticattitude?11Isitplausibletoclaimthatbeautyisthefundamentalconceptinaesthetics?12Doesitmatterifartcannotbedefined?13Whatistherelationbetweenformandcontentinart,anddoeseitherhavegreaterimportancethantheother?14Towhatextentisitappropriateandprofitabletocompareartwithlanguage?15Doestheconceptofresemblancehaveanyroletoplayinelucidatingpictorialrepresentation?16Canitbeexplainedhowmusicexpressesemotion?17Whatassumptionsareneededtoupholdtheclaimthatliteraryworkshaveobjectivemeanings,andarethoseassumptionsacceptable?18Musteachliteraryworkhaveonecorrectinterpretation?19Doesthemeaningofaliteraryworkresideinthemindofitsauthor,inthetextitself,orinneitherofthese?20Isthecorrectinterpretationofaliteraryworktheonethatmakesitmeanmosttous?21Doesthestudyofliteratureneedtobeguidedbyliterarytheory?22Doespsychologyhaveimportanceforaesthetics?23Doweneedart?24Isiteverlegitimatetoviewartasservingmoralorpoliticalends?25Isaparadoxinvolvedintheenjoymentoftragedyand,ifso,howshoulditberesolved?26Inwhatsensemightitbetruethatartisasocialratherthanindividualphenomenon?27DoesemotionalresponsetofictionrequirewhatColeridgecallsa‘willingsuspen-sionofdisbelief’?28Isitmorethanjustacontingenttruththatarthasahistory?255\nSEBASTIANGARDNER29HowcanonebemovedbytheplightofAnnaKarenina,whenoneknowshertobemerelyfictional?30Whatdistinguishesmetaphorfromnonsense?31Doessurvivalofthetestoftimeprovideacriterionforartisticachievement?32Isphotographyaformofart?33Canitbemaintainedthatanyoneformofartissuperiortotheothers?34Isarchitecture‘frozenmusic’?35Cananythingbemadeofthethoughtthatmusichas‘metaphysicalmeaning’?36Whatmakesapaintingartisticallysignificant?256\n8PoliticalandSocialPhilosophyDAVIDARCHARDSocialandpoliticalphilosophyhasexperiencedadramaticrevivalinrecentdecades.Thischapterconsidersthebackgroundofworkintheyearsafter1945,beforeexaminingthewritingsonjusticeofthemainarchitectsofthisrevival,JohnRawlsandRobertNozick.Subsequentlydiscussedaretheidealofequality,theproblemofpluralismandtheprin-cipleofneutrality,thecommunitarian,feministandMarxistcritiquesofliberalism,andthesignificanceofcommunity.Thefinalsectionconsidersthenatureofpoliticalphilosophyanditsrelationtopolitics.1IntroductionInanowcelebratedphrase,PeterLaslettannouncedthat‘forthemoment,anyway,politicalphilosophyisdead’.Hedidsoinhis1950introductiontothefirstcollectionofessaysentitledPhilosophy,PoliticsandSociety.Thesixthserieswaspublishedin1992,andtheinterveningyearsshowclearlyenoughthatthereportsofdeathweregreatlyexaggerated,oratleastthatthemomentwasmerelyapassingone.Thelastfiftyyears,andthelastthirtyinparticular,havewitnessedaquiteastonishingregenerationofsocialandpoliticalphilosophy,understoodinLaslett’sowntermsastheconcernofphilosopherswith‘politicalandsocialrelationshipsatthewidestpossiblelevelofgenerality’(Laslett1950:vii).Theworkhasdisplayedarangeofviews,fertilityofimagination,rigourofargumentation,careincriticalanalysis,andconcerntoaddresscontemporaryproblemswhichareampletestimonytotheconsiderablestrengthsofphilosophyintheEnglish-speakingworld.Inthe1950sthesestrengthswerefarmoreselectivelydeployed.Thestudyofpoliticswas,tosimplifygreatly,dividedbetweentheempiricalsocialsciencesandphilosophy.Exponentsoftheformerwereoptimisticthatpoliticalsciencecouldoffervalue-freebutwell-foundedexplanatorytheoriesofitsobject.Foritspartphilosophydisdainedsubstantivepoliticalevaluationandrestricteditselftothelogicalandlinguisticanalysisofpoliticaldiscourse.ThetitleofDanielBell’s1960book,TheEndofIdeology,conciselydefinedanhis-toricalmoment,anditssubtitle,‘OntheExhaustionofPoliticalIdeasintheFifties’,spelledoutwhatwasintended.Bell’stextwasaseriesofsociologicalessaysabout\nDAVIDARCHARDAmerica,butitsgranderclaimscarriedfurtherafield.The1930sand1940stestifiedtothesheerawfulnessofthepracticalrealizationofideologieslikefascismandStalinismwhichmadesimplifyingandabsoluteclaimstotruthabouthumanity,historyandreason.Theseyearsalsoreconciledintellectualstothevirtuesofamoremoderateandmodestpoliticalprogramme.Capitalismcouldbetamedbyawelfariststatetheeconomymixed,andfundamentalfreedomsadequatelyprotectedbydemocraticconstitutionalism.Therewasnofurtherneedforideologiesinthesenseofmillenarianvisionsofutopiayettoberealized.Theexhaustionofideologiesinthisnarrowsenseseemedalsotobetokenadisenchantmentwithpoliticsingeneral,aretreatfromthetraditionalconcernofpoliticalphilosophytoprescribethegoodsociety.Thiswasmirroredintheworkofphilosophers.The1950swasthehightideofCONCEPTUALANALYSIS(pp.2–3),andinanemblematictext,TheVocabularyofPolitics,T.D.Weldon(1953)characterizedtheenduringquestionsofpoliticalphilosophyasmis-placed,restingonfalseassumptionsandthemisuseofunanalysedfundamentalterms.Correctlyunderstood,thesequestionsaremerely‘confusedformulationsofpurelyempiricaldifficulties’(ibid.:192).Itisproperforphilosopherstoanalysethelanguageandfoundationalclaimsofideologies,buttheevaluationofpoliticalstatesofaffairsandinstitutionsismoreappropriatelyreservedforpoliticiansandpoliticalscientists.However,itwouldbeamistaketoover-simplify.From1945intothe1960stherewasseriousandimportantworkinpoliticalphilosophy.KarlPopper(1945)soughttorebuttotalizingpoliticaltheoriessuchasthoseofPLATO(chapter23)andMARX(chapter34).Buthedidsobydefendinganidealofthe‘opensociety’,andthepoliticsofpiecemealreformratherthanwholesalesocialengineering.F.A.vonHayekwasenteringuponhislifelongdefenceofapositionwhichisechoedinimportantcurrentviews(Gray1984).HayekaffirmedtheidealofjusticeasindividualfreedomundertheLAW(chapter13),anddefendedtheideathattheeconomicandsocialworlddisplaysaspontaneous,imper-sonalpattern.Itissimplyinappropriate,andultimatelydestructiveoffreedom,forgovernmentstoimposeuponthisworldaccordingtoaredundantidealof‘social’justice.TheworkofIsaiahBerlinisespeciallynoteworthy,forheanticipatessomemajorthemesinthephilosophicalliberalismthatisnowdominant.ForBerlin,aplausiblepoliticsmuststartfromarecognitionofapluralityinvalidyetconflictingendsoflife.Thereisnosingleidealofthegoodlife,andthepropergoalofpoliticsistomakepossiblethepursuitbyindividualsoftheirseveralends(Kocis1980).Berlin(1958)famouslydistinguishedtwosensesofliberty,onecharacterizedastheabsenceofobstructions,andtheothercharacterizedintermsofself-mastery.Hearguedfortheformer,‘negative’liberty,asthe‘truerandmorehumaneideal’thanthelatter,‘positive’liberty.AtthesametimeBerlinurgedaclearseparationofthenotionsoflibertyandequality.Heconcededthatthedegreeofsocialinequalitymaydeterminetheconditionsthatmakelibertyworthhaving,butinsistedthattheseconditionsdonotenterintothedefinitionofliberty.2JohnRawlsandRobertNozickonJusticeThesevariouswritingsareimportant,butitisthoseofJohnRawlsthatarenowtakentosignifythebeginningofadistinctlynewera.ThepublicationofRawls’sATheoryof258\nPOLITICALANDSOCIALPHILOSOPHYJustice(1972)hasprovedalandmark,radicallychangingthecharacterofEnglish-speakingpoliticalphilosophyandsupplyingtheineliminablebackgroundagainstwhichallsubsequentdiscussionhasbeenconducted.‘PoliticalphilosophersnowmusteitherworkwithinRawls’stheoryorexplainwhynot’isfittingtributefromonewhoseownviewsareatagreatdistancefromRawls’s(Nozick1974:183).ATheoryofJusticeismuchlaudedbutitisalsonowanextensivelycriticizedtext.Rawls’spresuppositions,methodologyandconclusionshaveallbeensubjectedtosuchextensivecriticismthatitisimportanttobeclearhowandwhyhistextsetthescene.First,ATheoryofJusticeisaworkofsubstantivepoliticalphilosophy.AlthoughRawlsengagesintheanalysisofconcepts,thisservesratherthansubstitutesfortheevaluationofpoliticalinstitutions.ATheoryofJusticeismore.Itdefendsasingle,unifiedvisionofthegoodsociety–onethatisrecognizablyliberal.RawlsgivesphilosophicalvoicetoadoctrineofdemocraticconstitutionalliberalismwhichcanwithmeritbeurgedastheonlyfeasibletheoryforcontemporaryWesternsociety.Second,RawlsseekstocombinearecommendationoftheidealpolitywithaplausibleaccountofHUMANNATURE(pp.672–3).ThisinvolvesnotjustatheoryofhumanMOTIVATION(p.392)takenfromtheSOCIALSCIENCES(chapter12).Italso–andlessnoticed–comprisesaviewofthemannerinwhichcitizenscanbebroughttorecognizeandacceptasbothfairandrealistictheformaltermsoftheircoexistence.Third,Rawlshasprovidedpoliticalphilosophywithanagenda.Hedoesnotjustofferatheoryofjustice.Hesuppliesanaccountofthepriorityofindividualliberty,thedefensiblelimitsofegalitarianism,andtheruleoflaw.Fourth,Rawls’sworkputscontemporarypoliticalphilosophybackintocontactwithitsownhistory.Theworkandconcernsofpastwriters,suchasHOBBES(chapter28),LOCKE(chapter29),Rousseau,HUME(chapter31),MILL(chapter35)andKANT(chapter32),aregivenrenewedpertinence.Fifth,ATheoryofJusticeisneverthelessaresponsetotheparticularcircumstancesofmodernity.Itisnotjustaprincipledaffirmationofthevirtuesofliberalismagainstthepostwarbackgroundofdisillusionmentwithmoreextremeideologies.Itcanbeseenasanattempttoaffirmthesevirtuesinthecontextofamorelocalcrisis,thatisthechallengeposedtotheidealofAmericandemocraticconstitutionalismbythecivilrightsmovementandthewarinVietnam.Moregenerallyagain,Rawls,asdidBerlin,insiststhatvaluepluralismisanunavoidablefeatureofmodernexistencetowhichpoliticsmustadequatelyrespond.TheideasofATheoryofJusticecanbestbelaidoutbyansweringanumberofbroadquestions.First,whyisatheoryofjusticeneeded?Becausepubliclyagreedtermsofsocialcooperationarebothnecessaryandpossible.Individualsbenefitfromlivingandworkingtogethersolongastheycanbeassuredthatsocialexistenceiswell-orderedandstable.Yetthereisapredispositiontoconflictinasmuchasindividualswantdiffer-entandnotnecessarilycompatiblethingsoutoftheirlifetogether.AgreedPRINCIPLES(pp.733–6)areneededtoregulateinteraction,anddeterminetheproperdivisionofbenefitsandcostsamongthemembersofsociety.Suchprinciplesarepossibletotheextentthatindividualscanseethemtobenecessary,andtheirparticulartermscanbeagreedunderappropriateconditions.259\nDAVIDARCHARDThePrimacyofJusticeJusticehasprimacyinthreesenses.Itisthefirstvirtueofsocialinstitutions.Thisdoesnotmeanthatsocietycannotdisplayothervirtues,onlythatwithoutatleasttheassur-anceofjusticethesewouldhavelittleornovalue.Second,truthisademandmadeofasystemofthoughtinsofarasitrelatestoreality.Justiceisdemandedofapolityinsofarasitmustrelatetotherealworld.ThecircumstancesofjusticearethosefirstnotedbyDavidHume,namelylimitedbenevolenceandmoderatescarcity.Humansarenotwell-enoughdisposedtomakesacrificesforothers,andmaterialresourcesarenotavailableinsufficientquantitytomakeformalagreementontermsofcooperationunnecessary.Rawlsaddstothisunderstandingofcircumstancesthefactsofmodernvaluepluralism.Individualshavedifferentbutnotnecessarilycompatibleaims.Mypursuitofmylife’sendswillconflictwithyourpursuitofyours.Agreementonhowtoreconcilesuchaconflictisnecessarybutpossibleonlyifnoonesetofindividualvaluesisaffirmedovertheothers.Third,anytheoryofjusticespecifiesandguaranteesanequalityofcitizenshipwhichsuppliesanassurancethatnoindividualmaybesacrificedforsociety’sgreatergood.Justicehasprimacyinthatthisassuranceisnottobecompromised.Whatisatheoryofjustice?Itisatheoryofthepubliclyagreed,finalprincipleswhichdefinethefundamentaltermsofsocialco-operation.Itisatheoryofthoseprincipleswhichregulatetheinstitutions,the‘basicstructure’ofsociety.Itdoesnotprescribepar-ticularoutcomes;nordoesitprovideacriterionforevaluatingactionsorthecharac-teranddispositionsofindividuals.Itisatheoryofsocialjustice,whichmustalsocoincidewithourconsideredandreflectiveMORALJUDGEMENTS(pp.225–7).Rawls’snotionof‘reflectiveequilibrium’allowsfor,indeedencouragesthepossibilityof,mutualadjustmentinboththeoryandjudgements.Yetanytheoryofjusticemustcapturewhatisessentialtoandsharedbydifferentaccountsofjustice,namelythat‘institutionsarejustwhennoarbitrarydistinctionsaremadebetweenpersonsintheassigningofbasicrightsanddutiesandwhentherulesdetermineaproperbalancebetweencompetingclaimstotheadvantagesofsociallife’(Rawls1972:5).WhatthenisATheoryofJustice?Rawls’stheoryassurescitizensequalLIBERTY(pp.762–3),andadistributionofallothergoodsthatmaximizestheexpectationsoftheleastwelloff.Moreformallyandcompletely:FirstPrincipleEachpersonistohaveanequalrighttothemostextensivetotalsystemofequalbasiclibertiescompatiblewithasimilarsystemoflibertyforall.SecondPrincipleSocialandeconomicinequalitiesaretobearrangedsothattheyareboth:(a)tothegreatestbenefitoftheleastadvantaged,consistentwiththejustsavingsprinciple,and(b)attachedtoofficesandpositionsopentoallunderconditionsoffairequalityofopportunity.(Rawls1972:302)Thefirstpartofthesecondprincipleisfamiliarlyknownasthe‘differenceprinciple’,andthetwoprinciplesareinlexicalorder.Thatistosaythatinequalitiesregulated260\nPOLITICALANDSOCIALPHILOSOPHYbythesecondprincipleareonlypermittedwhenequalityoflibertyunderthefirstprinciplehasbeenguaranteed.Providedthattheseprinciplesapplyorverynearlyapplytothebasicinstitutionsofsocietythenthatsocietyisjust.Andindividualshaveadutytoabidebytherulesofajustsociety.AlthoughRawlsoffersaCONTRACTUALIST(pp.672–7)argumentfortherulesthemselves,thedutytoabidebythemisnotitselfcontractuallybased.InpartthreeofATheoryofJusticeRawlsoffersanaccountoftheacquisitionofasenseofjustice.Thisgrowsoutofthebasicattachmentsandrelationshipswhichconstituteanysociety:family,friendshipandbroaderassociations.Thereisinthewell-orderedsocietyaconcordanceoffairrulesandasenseofjustice.Citizenscanrecognizethattherulesarejustandacknowledgetheirdutytoabidebythem.Fairnessandfeasibilitycoincide.Lastly,ATheoryofJusticeisatheoryofTHERIGHTANDTHEGOOD(pp.598–601).Thejustsocietyisnotonewhichrealizesthegoodlifeofitscommunity.Itisonewhichpermitsitsmemberstopursuetheirownconceptionofthegoodundercertaincondi-tions.First,thepursuitofthegoodisconstrainedbytheright,thatisthepubliclyagreedprinciplesofjustice.Second,theseprinciplesdeterminethedistributionofprimarysocialgoods–rights,liberties,powersandopportunities,incomeandwealth–whicharethedesiredprerequisitesofanyparticularpursuitofthegood.Third,peoplepursuedifferentidealsofthegoodlife.‘Humangoodisheterogeneousbecausetheaimsoftheselfareheterogeneous’(Rawls1972:554).Itwouldbeadisastrousmistakeforapolitytotrytoimposeuponitsmembersanyoneparticularidealoflife.Rawlshascontinuedtobelievethat,inanymodernsociety,therewillbeapluralityofconscientiouslysoughtends,andthatanystate,presumingorprescribingtothecontrary,wouldhavetoemployextensivecoerciontosecureitssinglevision.Thefinalquestiontobeansweredis,WhyATheoryofJustice?Thisaskswhatjustifi-cationRawlsoffersforhisownprinciples.Rawlsfamouslyemploysacontractualistmethodology:theprinciplesofjusticearethosethatrational,self-interested,butmutu-allydisinterestedindividualswouldchooseinanoriginalsituationspecifiedchieflyintermsoftheparties’selectiveknowledge.Thecontractualistmethodhasbeenthesubjectofagreatdealofcriticism:whyshouldahypotheticalcontractbind?WhywouldindividualschooseasRawlsclaimedtheywould?Wouldnotindividualschoosedifferentlyinadifferentoriginalposition?MoreoverRawlshimselfhassubsequentlydeniedthatthecontractarianargumenthasindependentjustificatoryforceandviewsitonlyasaheuristicdevicewhichservestoillustratetheforceofaclaimwhoserealjustificationlieselsewhere.However,thequestionthenpressesofwhatdoesjustifytheprinciplesofjustice,anditpressesthemoresinceRawlshasbecomeconcernedwiththeissueofhowajustsocietycanbewell-ordered,thatisbeviewedaslegitimatebyitscitizens.Beforeconsideringthisquestionitisnecessarytoexamineafundamental,andinfluential,critiqueofRawls’sunderstandingofjustice.ThisistobefoundinRobertNozick’sAnarchy,StateandUtopia(1974),whichisrightlypairedwithRawls’sATheoryofJusticewhenthecontemporaryresurgenceofpoliticalphilosophyisdiscussed.ThefoundationsofNozick’sargumentsareRIGHTS(pp.690–1)andOWNERSHIP(pp.690–1).Therightspossessedbyindividualsarefunda-mental,anddefineamoralspacesurroundingeachpersonwhoseinvasionconstitutesawrong.FollowingLocke,Nozickholdsthebasicrightstobethosetolife,liberty,and261\nDAVIDARCHARD‘estate’orproperty.ForNozick,theserightsare‘side-constraints’uponaction,andhaveinfinitemoralweight.Thatis,noamountofgoodconsequences,includingevenanoveralldiminutionofrightsviolations,canjustifyasingleviolationofoneoftheserights.Nozickdoesnotofferasystematicdefenceofhisunderstandingofrights.AtmostheclaimstheunderlyingjustificationtobetheKantianviewofINDIVIDUALSASENDS(p.736),leadingseparate,different,consciouslyanddeliberativelyshapedlives,whichshouldnotbesacrificedorusedasmeanstoothers’ends.Nozickisegalitariantotheextentofholdingthatallhumansequallypossessthesefundamentalrights.Healsoendorsesaninitialequaldistributionofownershipinrespectofourselves.Eachownshisorherownbody,anditspowers,capacitiesandabilities.Nozickbelievesthatthroughanexerciseofthisinitialself-ownershipfurtherentitlementstobitsoftheworldaregenerated.Heconsidersonlyoneformofargu-menttoshowhowthisgenerationofentitlementsmightoccur.ThisisJohnLocke’sfamousclaimthatoneacquiresrightstothatwithwhichonehasmixedone’slabour.UnfortunatelyNozickdevotesfourincisivepagestoexposeseeminglyunanswerabledifficultiesintheLockeanaccount(Nozick1974:174–8).ThisseemstoleaveNozickwithhisownproblemofunfoundedentitlements,andsosomecriticshavecharged.ButNozickisappealingtotheplausiblebackgroundassumptionthatunownedobjectsaretheretobeowned,andthatsomesortoffruitfulexerciseofone’sownpowersgroundsanentitlementtoownershipoftheseobjects.Thatis,providedcertainconditionsuponlegitimateacquisitionaremet.EntitlementsWhatoccupiesNozickislesstheprocessofacquisitionthanthelimitswhichmaybesetuponitsscope.NozickborrowsagainfromLockeandurgestheacceptanceofonefun-damentalproviso:‘Aprocessnormallygivingrisetoapermanentbequeathablepropertyrightinapreviouslyunownedthingwillnotdosoifthepositionofothersnolongeratlibertytousethethingistherebyworsened’(ibid.:178).Nozick’sconcernistoshowthatthisprovisocanbemetbyanestablisheddistributionofpropertyentitlements,whenoneconsidersthattheremaynolongerbeanyunownedobjectsforindividualstoacquire.Hisansweristhatindividualsunabletoappropriateareneverthelessbetterofflivingunderasystempermittingprivatepropertythantheywouldbeifnooriginalappropria-tionshadbeenpermitted.Theaccountofentitlementsiscompletedbyacknowledgingthatthosewhoholdbitsoftheworldareentitledfreelytotransferthemifandtowhomsoevertheychoose.Itfollowsthatthosewhoreceivesuchfreelytransferredholdingsarethemselvesnowentitledtoholdthem.Thuseverythingwhichisnotunownedislegitimatelyownedifacquiredjustly(bysomeprocesswhichdoesnotviolatetheLockeanproviso)orfreelytransferredfromonewhojustlyacquiredit.Ajustsetofholdingsisjustthatsetofholdingswhichcameaboutintherightway.InpartoneofAnarchy,StateandUtopiaNozickjustifiesaminimalstate.Hedoessobyconstructingaplausibletaleofhowastateofnaturemighthavegivenwaytoastate,thatisanorganizationlegitimatelyclaimingamonopolyontheuseofforcewithinagiventerritory,withoutatanystageviolatingrights.Thetaleturnsontheevolving262\nPOLITICALANDSOCIALPHILOSOPHYcompetitionbetweenprivateagenciesformedtoprotecttheirclients,withoneeventu-allyassumingamonopolisticroleandadequatelycompensatingtheothersforitsusurpationoftheirfunctions.ParttwoofAnarchy,StateandUtopiaisdevotedtoadefenceoftheviewthatsuchastate,restrictedtopreventingforce,theftandfraud,andenforcingcontracts,ismorallysufficient.Anygreaterroleforthestateisillegitimate.Ofcourseanobviousreasonforagreaterrolewouldbethatpeople’sholdingsneedtoberedistributedinaccordwithatheoryofjustice.Nozick’sowntheoryofjusticedoesnotrequireanysuchactionbygovernment.Histaskisthentoshowthatredistributivetheoriesofjusticearemistaken,andwhattheyrequireofthestateillicit.ThedistinctionNozickdrawsisbetweenhisownentitlementtheoryofjustice,andpatternedorend-statetheories.Hisisahistoricaltheorysincethejusticeofasetofholdingsisgivenbytheproprietyofthehistorywhichleduptoit.End-statetheoriesareunhistorical,specifyingthatadistributionmustconformtoaprescribedstructure.Moreparticularly,patternedtheoriesprescribeadistributionofholdingsaccordingtosomenaturalattributeororderedsetofsuchattributes,suchasintelligenceoreffort.Nozick’scritiqueofnon-historicaltheoriesisbothgeneralandspecific.Atagenerallevelheinsiststhatnopatterneddistributioncanbemaintainedwithoutpersistentandseriousviolationsoftherighttoliberty.Hisreasoningissimple.Voluntarytransfersofthoseholdingsinitiallyassuredunderanypatterneddistributionwillsubvertthepatternandtransgresstheprincipleunderlyingthepattern.Suchtransfersareeasytoimagine,andappeareminentlyunobjectionableinasmuchastheycanbesimple,consensual,bilateralexchanges.Nozick’sownfamousexampleisoffanswillinglypayingextratoseeanespeciallytalentedbasketballplayer,WiltChamberlain.Suchtransferscanonlybepreventedbydenyingindividualstherighttodowiththeirholdingsastheychoose.Surelynoconsistenttheoryofjusticecanbothdistributeholdingsanddenytheindividualstowhomtheycomeanyeffectivecontroloverthem?Nozick’sspecificcriticismsareofRawlsianandegalitariantheoriesofjustice.ThefundamentalweaknessofRawls’stheoryisthatitproceedsasifindividualsdidnotalreadyhaveclaimstoownershipofthemselvesandbitsoftheworld.Acakesponta-neouslygiventoagroupmightbedividedamongitsmembersinawayclosetoRawls’sprinciples;acakewhichIhavebakedorwhoseingredientsyouhaveprovidedwillbedividedinquiteanotherway.Rawlsissimplywrongtodiscountthevariousentitle-mentsindividualsbringtoanydeterminationofwhogetswhat.People,notleast,areentitledtotheirnaturalassets,and,consequently,toanyrewardsthatmayflowfromthese.Rawlsismistakentothinkthattheseassetsareacollectiveresource,orthatthedifferenceprinciplewhichreflectsthisassumptionwouldbeacceptabletothewell-offwhoarerequiredtomakeacomparativelygreatersacrificesothattheworst-offbenefit.Nozickfindsnojustificationeitherforanyegalitarianprinciple.Ademandforequalityofmaterialconditionreducestoanunsubstantiatedclaimthatthefunctionofsocietyistomeettheneedsofitsmembers,andsimplydiscountsthefactthatthingsarealreadyattachedtoparticularindividuals.Equalityisnotrequiredforself-esteem,whichinfactthrivesoncomparabledifferences.Atbasethedemandforequalityisdrivenbyenvy.Nozick’stheoryofjusticepermitsthegovernmentnointerventionistrolebeyondtherectificationofinjusticesbythetermsoftheentitlementprinciple.Itisironicthenthat,263\nDAVIDARCHARDgiventhelogisticdifficultiesintrackingbackandforwardbetweenpastwrongsandpresentholdings,Nozickshouldsuggestthedifferenceprincipleasaroughruleofthumbforrectifyinginjustices(ibid.:231).Otherwiseholdingsmustbeleftastheyhavecomelegitimatelytostand.Thatthismayresultingreatdisparitiesofwealthandlifeprospectsisregrettablebutnoinjustice.Therichmaychoosetobephilanthropistsbuttheydonowronginnotbeing.Anditwouldviolatetheirrightstorequirethattheyassistthelessfortunate.Thisseemsharshbutitisnotobviouslywrong-headed.LiberalcriticschideNozickforofferinga‘libertarianismwithoutfoundations’,butthismaynotbequitetrue.Nozick’sfoundationsareattractiveones.Individualsdoappeartohaverightsinthesensethatitwouldbefundamentallywrongtodocertainthingstotheminthenameofagreatersocialgood.Individualsdoalsohaveaclaimtoownershipoftheirselves.Thatthisissocanbeshownbyconsideringone’sreactiontotheideathatbodilypartsshouldberedistributedequitablyortobenefittheleastfortunateinthisrespect.Dotheblindhaveaclaimuponatleastoneoftheeyesofthesighted?CriticismofNozickismoreappropriatelydirectedathisrenderingofthesefounda-tionsandhowhebuildsuponthem.Thereislittlewarrantforconstruingrightsinanabsoluteandexclusivelynegativemanner,notleastwhenappealismadetoKantianunderpinnings.Withinamoralcommunity,individualsinneedmay,onKant’sownprinciples,havetherighttocontributionsfromthosewhocanhelpthem.Itisanimpos-siblystringentrequirementofrightsmerelythattheyruleoutinvasionsofourmoralspaceanddonothingtosustainusinoursatisfactionofbasicneeds.Nozickmayberighttothinkthatwe,severally,ownourselvesandmay,inconse-quenceandbysomeformofactivity,cometomakelegitimateclaimsuponwhatisunowned.But,arguably,theconditionshesetsforlegitimateacquisitionaretoolaxandtooeasilymetinwaysthatfavourfree-marketcapitalism.Itmayverywellbepossibletoconcedeaninitialequalityinself-ownership,andthelegitimacyofsomeprocessofacquisitionofunownedobjects,butthentospecifysufficientlytoughconditionsoflegitimacytoensurethatafinalequalityofholdingsisensured.Again,the‘Lockeanproviso’neednotbeaseasilysatisfiedbyhisfavouredsetofeconomicarrangementsasNozickassumes,especiallyonceambiguitiesinboththesenseof‘betteroff’andthebaselineagainstwhichcomparisonsaremadeareclarified.Nozickseemstopresumethroughoutthatownershipisfullandexclusiveownershipbyindividuals,andthatanythinglessamountstoaviolationoflibertyrights.YetNozickforgetsthat,forLocke,individualsmadeclaimthroughtheirlabouringnotuponwhatwasunowned,butwhatwascommunallyowned.InhisWiltChamberlainexampleNozickassumesthatindividualswithinthepatterneddistributionhavehold-ingswithwhichtheyarefreetodoastheychoose.Allformsofredistribution,includ-ingtaxation,areimplausiblystigmatizedascoerciveinterferenceswithfreedom.Yetitispossiblethatindividualsmightchoosetolimitinadvancedeparturesfromafavouredpatternofdistribution,anddosointhenameofaliberty-preservingequalityofcondition.Insum,Nozick’stheoryisnotsomuchlibertarianismwithoutfoundations,aslibertarianismwithunwarrantedconclusions.Hiscaseagainstanarchymaybeaccepted.Hisclaimthatstateminimalismisutopianinapositivesenseremainsunproven.264\nPOLITICALANDSOCIALPHILOSOPHY3EqualityBothRawlsandNozickareegalitarianstotheextentthattheyare,intheirdifferentways,committedtotheviewthathumanbeingsareentitledtoanequalityofregardortreatmentinsomefundamentalrespect.Rawls’stheoryassurescitizensequallibertyandNozickholdsthatallhumansequallypossessfundamentalrights.Allcontem-porarysocialandpoliticalphilosophyaffirmsthathumansareentitledtoequalityofsomething,whichsuggeststhatthecrucialquestionisnot‘Whyequality?’but‘Equalityofwhat?’(Sen1992).Ofcourseademandforequalityinrespectofsomegoodisconsistentwith,indeedmaydemand,inequalityinrespectofsomeothergood.ThisistrueofNozick’slibertarianismwhichclaimsthatanequaldistributionofindividuallibertymustleadtoanunequaldistributionofincomeandproperty.Ademandforequalityinsomegoodmaytakepriorityoverothervalues,asisthecasewithRawls’sinsistenceuponthelexicalpriorityofequalliberties.ButalthoughRawls’sfirstprincipleofjusticeformallyguaranteesequalcitizenship,thedifferenceprincipletoleratessocialandeconomicinequality.Rawlsseesnoincon-sistencyhere,thoughheconcedes–asdidBerlin–thatdifferentialaccesstoresourceswillqualifytheworthofequalliberty.Themoreyouhavethemoreyoucanmakeofafreedomsharedequallywithothersinsociety.Socialistcriticsofliberalismhavealwaysinsistedthatthisisunsatisfactory,arguingthatrealequalityoffreedomdemandsanequalizationofresources.Somefeministshavealsoarguedthatthepublicorlegalequalityofthegendersisunderminedby,andyetservestodisguise,afundamentalinequalityofsocialpower.Thischaracterizesallmale–femalerelationshipswithinpatriarchalsociety,anditseliminationwillrequiremorefar-reachingreformofsocietythansimplyinstitutingaprincipleofequalcitizenship.Walzer’s(1983)argumentisimportantinthiscontext.Hesuggeststhattherearedifferentspheresofjustice,eachbeingspecifiedbyasetofgoodstobedistributedandaconsequentprincipleoftheirfairdistribution.Theunequaldistributionofsomegoodneednotinitselfbeunjust,butifthisdistributiondeterminesthedistributionofsomeothergoodthenthatisunfair.Itisnotnecessarilywrongthatsomepeoplehavemoremoneythanothers;itisiftheirgreaterwealthbuysthempoliticalpower,officeorgreaterpersonalhealth.Walzer’sapproachhighlightsthefactthatourfavouredaccountofequalitydependsonwhatitisthatwewishtoseeequallydistributed.Contemporaryegalitarianismdefendsthreebroadfieldsofapplicationofequality.Forawelfareegalitariantheidealisa‘conditionofequalwell-beingforallpersonsatthehighestpossiblelevelofwell-being’(Landesman1983).Thecentralproblemforwelfareegalitarianismisthatitseemscom-mittedtotakingaccountofthosepleasuressomehumansmaytakefromseeingothersdolesswell,andisalsocommittedtothesatisfactionofsomeacquiredexpensivetastes.Resourceegalitarianismassignstoeachindividualabundleofgoodswhichisenviedbynootherindividual.Thecentralproblemforthisversionofegalitarianismisthatnaturalassets(skills,talentsandabilities)areunequallydistributedandinconsequencepeoplebenefitdifferentiallyfromtheiruseofequallydistributedresources.Yettoincludepersonaltalentsamongresourceswhichareequalizedmeansthatthetalentedare,ineffect,theslavesoftheuntalented.265\nDAVIDARCHARDContemporaryegalitariansarenowinclinedtoagreethatindividualsshouldnotbecompensatedfortheeffectsoffreechoices(suchas,obviously,choosingtodevelopanexpensivetaste),butshouldbecompensatedforthosefactorsaffectingthemwhicharedueto‘bruteluck’(suchas,obviously,ahandicap).OnDworkin’s(1981)influentialargument,aprincipleofliberalequalityshouldbe‘endowment-insensitive’but‘ambition-sensitive’.Thatis,itshouldpermitindividuals’livestoflourishorfounderasaresultofthechoicestheymake,butnotinconsequenceoftheirnaturalorsocialendowments.Suchaprinciplemaynotbeeasytorenderdeterminateandreliesuponaspecificationofthescopeoffreechoice.Moreoveritshouldbenotedthatindividualsmaynotchoosethecircumstanceswhichcausefreelychosenpreferencestobemoreorlessexpensive.Capabilityoropportunityegalitariansdemandtheequalizationofthecapacitytoleadthelifethattheindividualvaluesorchoosestolive.Theytherebyavoidtheproblemfacedbyresourceegalitariansofhowthesamesetofresourcesmay,accordingtothecircumstances,bedifferentiallyconvertibleintoachievablestandardsofliving.Theyalsoavoidtheproblemsofwelfareegalitarianismwhichneglectsthoseaspectsofgoodlifewhichdonotreducetowell-beingandwhichcannotacknowledgethefactthatsomepeople,subjecttoenduringconditionsofsignificantinequality,mayadapttheirpreferences(andstatesofwell-being)totheseconditionsandnot,inconsequence,expe-rienceasignificantlylesserdegreeofwelfare.Capabilityegalitarians,however,mustprovidearankingofcapacitieswhichcannotbepurelyquantitative–alifedoesnotnecessarilygobetterthemorethingsonecando.Butifsucharankingdoesnotreducetooneofopportunitiesforgreaterorlesserwelfare,itisindangerofbeinganobjectivelistrestinguponacontentiousaccountofthehumangood.Finally,anyegalitarianismmustanswercriticismsofbothpracticeandprinciple.Egalitarianismmaysimplybeinconsistentwithcertainimmutablefeaturesofhumanmotivation.Itmaynotbepossibletocombine,withintheindividual,theattitudeofuniversalimpartialityandpersonalpartiality(Nagel1991).Again,asocietyinwhichequalityisguaranteedmaylacktheincentivesnecessaryformaximizingthetotalsocialproductandtherebyimprovingthewell-beingofall.EgalitarianismmustalsomeetNozick’schallengethatthecaseforequalityisunproven,andanywaytrumpedbymorefundamentalideals,suchasfreedomorself-ownership:apriorcommitmenttolibertyunderminesanyguaranteeofequality,and,correlatively,equalitycanonlybemaintainedbythedenialofindividualfreedom.Thedisputedrelationshipbetweenequalityandlibertymayprovetobethemostenduringandpressingissueofpoliticalphilosophy.ThefactthatwehavetherebycomefullcircletotheworkofBerlinandHayekcertainlysuggestsanunderlyingcontinuityofconcerninthesubject.4PluralismandNeutralityRawlsrespondedtocriticismsofATheoryofJusticeinaseriesofarticleswhichculmi-natedinhissecondmajorbook,PoliticalLiberalism(Rawls1993).Itsconcerniswiththegoodorderofamodernliberaldemocraticsociety.Itdoesnotdefendaparticulartheoryofdistributivejustice,althoughRawlscontinuestothinkthathisowntwoprin-266\nPOLITICALANDSOCIALPHILOSOPHYciplesaresuitabletoregulatethebasicstructureofawell-orderedsociety.WhereasATheoryofJusticestartedfromHumeancircumstancesofjustice,PoliticalLiberalismstartsfromthemodernconditionofvaluepluralism,namelythatpeopleentertain,forgoodreasons,differentandprobablyincompatiblecomprehensivephilosophicaldoctrines.PoliticalLiberalismalsorepudiatesthemainargumentsforthetwoprinciplesofATheoryofJustice.Thecontractarianargumentwasonlya‘deviceofrepresentation’;nometa-physicalunderstandingoftheselfwasorneedstobepresupposed;anditisamistaketothinkagreementtotheprinciplesofjusticecouldrelyondeep-lyingacceptanceofabroadermoraldoctrineoffairness.Nevertheless,PoliticalLiberalismarguesthatapoliticalconceptionofjusticecouldcommandthesupportofan‘overlappingconsensus’ofvariouscomprehensivedoc-trinesinthesocietywhosebasicstructureitregulates.Ajustsocietyoffreeandequalcitizenscouldthusendureovertimeeventhoughdeeplydividedinitsbasicreligious,moralandphilosophicaloutlooks.SomecriticsfearthatRawls’stheoryonlyrepresentsapragmaticaccommodationtothepossibilitiesofconsensuswithinparticularsocieties.YetPoliticalLiberalismdoesnotsimplydefendthepoliticsofcompromiseandmodusvivendi.Itisanattempttospecifythetermsofco-operationthatcanwithstandthetestofpublicandrationalnegotia-tion,commandingthefreelygivenassentofequals.TheworryshouldratherbethatRawlscannolongershowwhythewell-orderedsocietymustbeoneregulatedbycertainprinciplesofjustice,andviceversa.Rawls’sideaofan‘overlappingconsensus’maybesoindeterminateastoyieldverymanydifferentoutcomes.Oritmaydeliveraspecificoutcomebyprescriptivelystipulatingwhatshallandshallnotcountasreasonabledoctrines.Moreover,itmaymisstherealnatureofnegotiatingpoliticaldifferenceinitsrepresentationofviewsanddisagreementas‘reasonable’.ThelastpartofATheoryofJusticesketchedanaccountofthedevelopmentinthecitizensofthejustsocietyoftherequisitesenseofjustice.Itofferedatheoryofmoraleducationthroughtherelationshipsofassociationwithinthatsociety.InPoliticalLiberalismRawlsforswearsthiskindofaccount.Buttothatextentanytheoryofpoliticaljusticeis‘thinner’andlessplausible.Indeedthereisageneralproblemofthebasicliberalapproach.Rawls’sconvictionremainsthat,shortofanunacceptablecoercedunanimity,adomainofpublicagreementcanbesecuredandclearlyseparatedfromthesphereoftheheterogeneousprivategood.Inthisheremainsfaithfultotheliberalvisionofasocietyinwhichallequallyandfreelypursuetheirdifferentlivesconstrainedbypublictermsoffaircoexistence.Yetsuchavisionmaybeimpaledonthehornsofadilemma.Eitherthetermsofpublicagreementaresoinsubstantialastoyieldonlyanemptyformofpoliticalcommunity.Ortheyaremadesubstantialonlybyviolatingtherequirementthatdisputedunderstandingsofthegoodberestrictedtotheprivatesphere.Aliberalpoliticalordermaydemandacultureofandeducationinliberalvalues.Thegoodofsocietymaynotbesoreadilyseparablefromthegoodofitscitizens’privatelives.Finallyitmaybeheldunreasonabletoaskpeopletoexcludetheirdeeplyheldcom-prehensiveviewsfromthetermsoftheirpoliticalactivity.Citizensarerequiredtoframeandarticulatetheirdemandsinthelanguageofapublicreasonwhichispoliticalandindependentofanyparticularcomprehensiveview.Inafrequentlyinvokedmetaphor267\nDAVIDARCHARDindividualsmustcome‘nakedtothepublicsquare’bysettingasideviewsthataredeeplyimportanttoandperhapsevenconstitutiveoftheirselves.Moreoveraliberalpoliticalcultureiscutofffromthesubstantivecomprehensiveviewsthatmayhistoricallyhavenourishedit,leavingitderacinatedanditscitizensalienated.Rawls’sstartingpointinPoliticalLiberalism,asitisnowformanyotherpoliticalphilosophers,isvaluepluralism.Thisshouldnotbeconfusedwithrelativismorscepti-cismaboutthegood.Suchpluralismisrathertakentobeaninevitableresultoftheconscientiousexercisebydistinctindividualsofasimilarcapacityforreasoning.AnexpectationofreasonabledisagreementisanacknowledgementofanevidentfactaboutmodernlifeandarejoindertoalongEnlightenmenttraditionwhichholdsreasontogenerateconvergenceonthetruth.Pluralismmustnotmerelybeexpectedbutalsotoleratedbecausevaluemonismcanonlybesecuredthroughcoercivestateinterferencewithindividuallibertyofcon-science,andbecausepluralismneednotspelldisasterifitcanstillprovidethebasisofharmonioussocialandpoliticalco-operation.Theproblemmaybethatthetolerationofvaluepluralismpresupposessomefoundationalvalue–suchastheautonomywhoseexercisebydifferentindividualsgeneratesit.Yetifpluralismgoes,asitwere,allthewaydownthereisnoreasontothinkthatanyvaluecanbeprivilegedandnotbethesubjectofreasonabledisagreementbetweenindividuals.Valuepluralismishonoured,arguably,ifandonlyifthestate,initslawsandpolicies,remainsneutralbetweenitscitizens’differentconceptionsofthegood.Thisdoctrineofofficialneutralityonthequestionofthegoodisviewedbymanyasdefinitiveorconstitutiveofcontemporaryphilosophicalliberalism(Dworkin1978;Ackerman1980:10–12).Adistinctionisnormallydrawnbetweenneutralitywithrespecttothejustificationoraimofferedbygovernmentforitsactivityandwithrespecttoitsconse-quencesoroutcome,mostpreferringtounderstandneutralityinthefirstsense.Thus,goingbacktoamuchearlierliberalthinker,JohnLockedefendedpoliticalintoleranceofCatholics(non-neutralinitsconsequences)notbecauseCatholicsweredoctrinallyinerror(non-neutralinitsjustification)butbecauseCatholics,inowingallegiancetotheforeignjurisdictionofRome,representedadangertothegoodorderofthestate(neutralinitsjustification)(Locke1689).Themannerinwhichpoliticalneutralitycanbeachievedisrepresentedeithernegatively,asthebracketingoutofwhateveristhesubjectofdisagreement,orpositively,astheoperationofasharedpublicpoliticalreason.Neutralitywillbedefendedbyappealtothevalueofequalityorthatofindividualautonomy.Onthefirstastatedoesnottreatitscitizensasequalsifitfavoursonecitizen’sviewsofthegoodoverothers.Ontheseconditismoreimportantthatindividualsleadthelivestheyseeasgoodratherthanbeledtolivethelifethatthestatethinksofasgood.Inadditionadoctrineofneutralitywillbeseenasanecessaryprotectionagainsttheexcessive,anddangerous,exerciseofstatepoweroverindividuals.Valuepluralism,andtheassociatedprincipleofneutrality,standopposedtoperfec-tionism,thedoctrinethattherearespecifiablehumanexcellencessuchthatsomeformsofhumanlifearesuperiorinthemselvestoothers.Rawlsrejectsperfectionism,butsomeliberalshaveinsistedthat,nevertheless,hisownaccountofhumangoodistoothin,andthatwecanmakejudgementsabouttherelativeworthofdifferentformsofexistence.Thisneednotbeinconsistentwithacommitmenttoegalitarianismand268\nPOLITICALANDSOCIALPHILOSOPHYtolerance(Haskar1979;Galston1980).Moreover,thereisasuspicionthatRawlsianliberalismis,atbase,inconsistent.Ifitisstrictlyneutralthenitcannotsubscribetoanynormativeunderstandingofindividuals,thattheyshould,forinstance,strivetobepurposive,autonomouscreatures.ButsuchaviewseemspresupposedbyRawls.If,ontheotherhand,liberalismdoesincludesomefoundationalmoraljudgementsabouthumanbeings,thenitcannotreasonablyclaimobjectivityforthesejudgementsandrefuseitforjudgementsaboutthegoodlife.Italsoseemsclearthataliberaldemocraticculturewillflourishtotheextentthatitscitizensacquireandpractisecertainvirtues,thoseforinstanceoftolerance,civility,respectforothersandawillingnesstomakesacrificesforthecommongood.Nowwhethersuchvirtuesarelearntinchildhood,orgotthroughgoodhabits,aliberalsocietymustsurelytakethedecisiontoencouragethosesocialformswhichfacilitatetheacquisitionofsuchvirtues.AutonomyandLiberalismInthiscontextJosephRaz(1986)isarevisionistliberal.Hedoubtswhetherlibertyshouldbethecentralvalue,anddefendsinsteadtheprimacyofautonomy.Hisunderstandingofautonomyisthatonlyalifechosenfromamongseveralmoraloptionsisautonomous.Further,Razbelievesthat,whilenotallformsoflifearevaluable,theremaybeseveralincompatibleonesthatare.Autonomyrequiresapluralityofmorallifechoices,andthisinturnrequiresthecreationandmaintenanceofsocialformsconducivetoautonomy.Thegovernmenthasadutytopreservetheseformsbut,crucially,maydosoinnon-coerciveways.Subsidiesandtaxationcaneffectivelyrendercertainchoicesattractiveorunattractiverelativetoothers.Thus,Razbitesthebullet,holdingthataliberalpoliticalcultureshouldsustainitsowncorevalues,andnotaspiretoanimplausibleidealofneutrality.Liberalismcanbetolerantofdiversity,interventionistandanti-perfectionist.InsimilartermsWilliamGalston’sliberalismembracesandsupportsasetofdistinctiveliberalpurposesthatguideliberalpublicpolicyandshapeliberaljustice.Theserequirethepractiseofandaciviceducationinliberalvirtues,andthemaintenanceofaliberalpublicculture.Liberalismhasathickenoughtheoryofthegoodlifetobeabletoruleoutcertainpracticesandencourageothers:‘itisnottheabsenceofanaccountofthegoodthatdistinguishesliberalismfromotherformsofpoliticaltheoryandpractice.Itisratheraspecialsetofreasonsforrestrictingthemovementfromthegoodtopubliccoercion’(Galston1980:180).Liberalismisnotneutral;ratheritsgovernanceisnotmorallycostlyinitsuseofcoercivestatepower.5CriticsofLiberalism:Communitarianism,Feminism,andAnalyticalMarxismTherearethreebroadmovementsdeservingconsiderationfortheircritiquesofliberalism:communitarianism,feminismandMarxism.269\nDAVIDARCHARD5.1CommunitarianismThewritersgatheredunderthistitle–RobertoUnger(1975),MichaelSandel(1982),MichaelWalzer(1983),CharlesTaylor(1985),AlasdairMacIntyre(1981,1988)andRichardRorty(1989,1991)–aredisparateanddonotconsciouslysubscribetoacommonmanifesto.Itismoreaccuratetospeakoffamilyresemblancesthanasingle,sharedprogramme.Intherespectsinwhichtheyareallcommunitariantheyoffernotsomuchanalternativepoliticalviewtoliberalismasacriticismofitspresuppositions.Communitariansinvokecommunityincriticismswhicharebothnormativeanddescriptive,althoughthedistinctkindsofcriticismarenotalwayscarefullydistin-guished(Caney1992;Taylor1989).Thecentraldescriptivecriticismconcernsthenatureoftheselforindividual,andchargesliberalismwithsubscribingtoaninadequately‘thin’understandingofthe‘self’.Thefactsspecifyingthesocialandhistoricalsituationofeachpersonconstrainthekindsofself-understandingshecanreach,andthechoicesshewillmake.TheRawlsianindividualisnot‘embedded’inanyplaceortime;sheissoemptiedofsubstan-tial,individuatingfeaturesastomakeitdifficulttodescribeheraschoosingalife.Howcanaself-lesspersonbesaidtohaveanyconceptionofthegoodormakechoicesofends?Thiscriticismmayrestonamisunderstanding.Rawls’sconcernisnottodefineide-alizedchoosersofthegoodsomuchastospecifytherelevantconsiderationsenteringintoajointlyagreeddeterminationofthepublicrulesoffairco-operation.Asheinsistsinhislaterwork,histheoryofjusticeisnotmetaphysicalbutpoliticalandassumesnoparticularunderstandingoftheself.Atthesametimethecriticismmaybeoverstated.Toclaimthatindividualsarewhollydefined,andtheiridentitycompletelyconstitutedbytheirmembershipofsomecommunityatsomehistoricalmoment,iseffectivelytodenythemanykindofmeaningfulchoiceovertheirlives.ItalsoseriouslyunderminesanyclaimtheymightmaketobeMORALAGENTS(chapter6).Andtotheextentthatthecommunitarianclaimisqualifiedbyphrasessuchas‘toalargeextent’itishardtoseewhatdistinguishesthecommunitarianfromtheliberal.Thenormativecriticismsofcommunitarianismarethreefold.First,theprioritywhichRawlsaccordsthevirtueofjusticeisallegedtoderivefromanimpoverishedunderstandingofpoliticalassociation.Rawlsclaimsthatjusticeisthefirstvirtueofsocialinstitutions,thatitisrequiredtodealwithlimitedbenevolence,moderatescarcityandmodernvaluepluralism,andthatitisneededtoprotectindividualsfrombeingsacrificedforthesociety’sgreatergood.Accordingtocommunitariancritics,however,justiceisonlytheidealofsocietieswhichdonotdisplaycommunity.Communitiesproperdonotneedtobejust,andwouldnotbecommunitiesiftheyfeltsuchaneed.MichaelSandel(1982)characterizesjusticeasaremedialvirtue,bindingupinthebestwaypossibleasecond-bestformofsocialco-operation.ThesenseinwhichjusticehasprimacyforRawlshasalreadybeenindicated,andSandel’scriticismappearsmisplaced.Resolutioninthefaceofdeathisnotlessofahumanvirtueforbeingredundantinimmortalcreatures.Sandeldoesnotclaim,ashavesomeMarxistsforinstance,thatthecircumstancesofjusticewilldisappearinthefuture.Hesaysonlythatsomecommunitiesdonotdisplaythem.Butthese–mostnotablythefamily–cannotbethemodelsforsocietyasawhole.Thefamilyhasthosecharacteristicswhichmakeitsmembersbenevolentlydisposedtooneanother–270\nPOLITICALANDSOCIALPHILOSOPHYcloseness,andaffectivityrootedinnaturalrelationships–becauseitisnotsimplyasmallerkindofsocietybutsomethingquitedifferentaltogether.Societiescannotbefamilialcommunities.Thesecondnormativeclaimofcommunitarianismisthatmembershipofapoliticalcommunityisagoodwhichliberalismneglects,ignores,orwhosesenseitcannotsuccessfullycapturebyitsownterms.PoliticalassociationisviewedbyliberalsandlibertariansasanINSTRUMENTALGOOD(pp.216–18).Itrealizesthecompromisesnecessaryforindividualstoderivemutualadvantagefromco-operating.Thereisnoothersenseofbeingtogetherascitizensthanisrequiredtobringaboutthisend.RawlsandNozickbothtalkofcommunity,orevencommunities,growingupwithinthisframework.Buttheseseeminessentialandsomehowaddedontothebasictermsofpoliticalassociation.Thiswouldnotbesoseriousacriticismwereitnotforthefurtherclaimthataliberaltheoryofjusticeneedsmore.SandelchargesthatRawlsdefendsadifferenceprinciplewithoutfoundations.Acceptanceofthisprinciplerequiresawillingnesstoseeone’snaturalassetsascommunallyowned,andyetRawls’stheoryallowsfornocommunitywhichcouldlaysuchaclaimtoownership.OnRawls’saccountIseemytalentsasforotherstoderivebenefitsfrom,yetthereisnoreasonwhyIshouldseemyselfasjoinedtotheseothers.Herewemeetagaintheproblemforliberalismofshowinghowajustsocietycanalsobewell-orderedandwhetherpoliticallegitimacymustrestuponasenseofcommunitywhichliberalismisincapableofsupplying.Thethirdnormativecommunitarianclaimisthatwhatisgoodandjustforindivid-ualsisdefinedbythecommunitytowhichtheybelong.AlasdairMacIntyre(1981,1988)isassociatedwiththeviewthatanindividual’sgoodisinseparablefromtheROLE(p.388),officeorsocialpositionhefills.MichaelWalzer(1983)makesthefurtherclaimthat,sincethegoodstobedistributedhaveparticularSOCIALMEANINGS(pp.384–90),thejusticeoftheirdistributionisrelativetothesemeanings.Asthesemeaningshaveparticularsocialandhistoricallocation,sodotheassociatedprinciplesofjustice.However,thebesettingdangerofanyappealtotheexistenceofdistinctunder-standingsofthegoodandthejustisRELATIVISM(pp.395–7).Thissocietyisjustbyitsownlights;thatsocietyisjustbyitsownlights.Andnevershallthetwobecompared.Further,sharedunderstandingsdeterminenotonlywhatisjustbutwhatisunjust.Yettheseparticularjudgementsmayviolatewhat,plausibly,oughttobeUNIVERSALMORALSTANDARDS(pp.733–6).Imagineasocietybywhoseunderstandingofthemaster–slaverelationship20lashesadayarefairandsufficient.Ifamasteradminis-ters25heisunjustandifonly15hebehavesgenerously.Butsurelyanybeatingiswrong,andany‘rules’ofslaveryareunjust,whateverthemembersofthesocietybelieve.Withregardtothecommunitarians’normativeclaimstheliberalwillrespondthatliberalismdoesnotdisputethevalueofcommunitynorneeditneglectitssignificance.Theliberalwillpressthecommunitariantoclarifypreciselywhatsortofpoliticalpro-posalsaredistinctiveofcommunitarians,worryingthat,wheretheyarespelledout,suchproposalssuggestanilliberalconcessiontowhateverhappenstobethesharedunderstandingsandpracticesofaparticulartimeandplace(Gutman1985).Atbestcommunitarianismpromptsliberalismtodemonstratehowitsconstitutivevaluesare271\nDAVIDARCHARDconsistentwiththemaintenanceandreproductionofagoodorderwhich,arguably,needsasharedsenseofcommunity.5.2FeminismFEMINISM(chapter20)offerstwoquitedistinctkindsofcriticismofphilosophicalliberalism.Thefirstconcentratesonthesilenceofliberalpoliticaltheoryand,beingconsciouslyadhominem,ofitsmaletheoristsaboutwomen’splaceinajustsociety.Moreparticularlythissilenceiscompoundedbyassumptionsaboutwhatthisplaceactuallyisandshouldremain.Aformalcommitmenttotheequalityofallisgainsaidbyanendorsementofpatriarchalism,whichisnotalwaysonlyimplicit.Patriarchalismheremeansthedoctrineofthesubordinationofwomentomalepower.ThePublicandthePrivateThecrucialassumptionunderpinningliberalpatriarchalismisthattothedistinctionbetweenpublicandprivatespheresofactivitycorrespondsadifferenceofnaturebetweenmenandwomen,andtherolesforwhichthesenaturesbestsuitthem.Thewomanisconfinedtothefamilywhereherbiologyequipshertobear,rearandcare.Thewomanisthusdoublyoppressed:excludedfromthepublic,politicalspherewheretheliberalprincipleofequalityoperates,andtheinferiorofhermaleproviderwithintheprivatehousehold.Intermsofoppressionrelatedtothedistinctionbetweenpublicandprivatespheres,SusanMollerOkin(1989)accusesRawlsofperpetuatingliberalpatriarchalism.Heneverbroachestheissueofsexualjustice,yetassumesboththecontinuedexistenceoftheinstitutionofthefamily,and,morepertinently,thetraditionalsexualdivisionoflabourwithinthisinstitution.Okincanreasonablyclaimthatafamilywhosestructureisdeeplyunjustcannot,asRawlsexpects,betheappropriateinstitutionforacquiringasenseofjustice.Okin’scritiqueofthefamilyisneverthelessnotaradicalone.ShemerelyinsiststhatRawlsianprinciplesofjusticebeuniversalinscopeandextendtoallinstitutions,includingthefamily.Thesecondlineoffeministcriticismconcernstheallegedmalenessofapreoccupa-tionwithjusticeandrights.Appealismadetoadistinctivefemaleethicswhichempha-sizesattachment,responsibility,contextandparticularityasopposedtoindependence,rights,abstractionanduniversality.Theideathatwomenspeakinanothermoral‘voice’tothatofmenisdueprincipallytotheworkofCarolGilligan(1982)inthepsychol-ogyofmoraldevelopment.Butthecontrastbetweenanethicsofcareandoneofjusticenowenjoyswidecurrencyinmoralandpoliticalphilosophy.Gilliganherselfdoesnotseethesemoraloutlooksasmutuallyexclusive;nordoessheseeeachofthemasnec-essarilyassociatedwithoneparticulargender.Indeedsheseemstofavouranaccountofmoraldevelopmentwhichemphasizesthesocializingroleofparents,whatevertheirgender.Fromthestandpointofsocialandpoliticalphilosophythisparticularlineofcriti-cismbitesonlyiftheprevailingliberalaccountsofthegoodsocietymaybejudgeddefectiveforomittingmentionofcare.Itbegstoomanyquestionstosuggestthatjustice272\nPOLITICALANDSOCIALPHILOSOPHYappliesinthepublicsphere,andcareintheprivate.Itwouldalsogrosslyover-simplifytosuggestthatsocialco-operationcouldbegovernedeitherbyrulesofjusticeorbyanethosofcare.IndeedtherearereasonstoagreewiththeRawlsianviewthatjusticehasprimacy.Carealonecannotdeterminewhoshouldbeinreceiptofwhatgoods.Andwherecarefailsorfallsshortofwhatisdesired,justicespecifieswhatcanbelegitimatelyexpectedofothers.Nevertheless,theoppositionbetweencareandjusticehighlights–asdoesthecommunitariancritique–theextenttowhichasocietygovernedsolelybyrespectforrightsanda‘senseofjustice’maynotgenerateanyrealsenseofrelatednessandinterdependenceamongitsmembers.5.3AnalyticalMarxismAsthetitleimpliesanalyticalMarxistshavedisplayedthevirtuesofanglophonephilosophyingeneral,thatisargumentativerigour,scrupulousattentiontothetext,andcarefulconceptualanalysis.WhethersuchworkisMarxistisdebatable.ItcertainlyeschewswholesalesubscriptiontoeveryaspectofMarx’swork,andprefersinsteadseparatelytoappraisethedistinctclaimsthatmaybesaidtoconstitutethistheory.Ithasrejectedacertainunderstandingofhistory,TELEOLOGICAL(pp.319–20)andHEGELIAN(chapter33)inorigin,and,inthecaseofJonElster(1985)atleast,embracedarobustMETHODOLOGICALINDIVIDUALISM(pp.397–9)thatisveryunfamiliaramongMarxists.IthasbeenwillingtojettisonwhatmanywouldseeascentralMarxistclaims,forinstancetheLABOURTHEORYOFVALUE(chapter34),andhasincreasinglyconcededtheattractivenessofnon-Marxiantheoriesinpoliticalphilosophy.However,analyticalMarxismhasinsisteduponthespecificcharacterofcapitalism,andsoughttoexplicatethemannerbothinwhichsuchasocietyisfundamentallyflawed,andinwhichsocialismrepresentsafeasibleanddesirablealternative.Yettherearedeepproblemswithsuchanapproach.InthefirstinstanceitneedstobeshownthatMarxhadatheoryofjusticebywhichCAPITALISM(pp.755–6)canbejudgedunjust.Thistaskconfrontsafamiliarparadox.Marxemploysalanguagewithappar-entmoralimport,yetexplicitlydisdainsmoralcriticismandtheory(Lukes1985).Theparadoxisnoteasytoresolve,andanyresolutionmaysimplyhavetoaccommodatethefactthatMarxwasnotconsistentinhisoutlook(Geras1985).Second,ifcapitalismisunjustthenforMarxiststhatinjusticemustinhereinsomesignificantfeatureoftherelationshipbetweencapitalistandproletarian.Twocandi-datessuggestthemselves.Thefirstisthatthecapitalistexploitstheproletarian.Yetonanyplausibleaccountofexploitationthatcanbeoffereditisarguablethatsomeformofexploitationmaycharacterizeeverysociety,includingsocialism.Moreover,thedifferenceinownershipofresourceswhichdefinesthecapitalistrelationneednotbeunfair.Itmayhaveariseninaperfectlylegitimatefashion,forinstancethroughthesuperiorefforts,talents,assiduityandprodigalityofthecapitalist.Theideaofa‘cleanlygeneratedcapitalism’conformstoRobertNozick’sprescriptionofajustdistributionthathasarisenbyjuststepsfromanoriginallyjustsituation.Thesecondcandidateforexplainingtheinjusticeofthecapitalistrelationiscoercion.Workersmaybeunfairlyforcedtoworkfortheiremployers.Thedifficultyisthatforanyoneworkerthealternativestoemploymentarenotsostarkandrestrictiveaswouldberequiredtoestablishcoercion.Thisisespeciallytrueinamodern273\nDAVIDARCHARDwelfaresocietywhereitmustalsobeacknowledgedthatavenuesofescapefromone’sclassdoexistforthetalentedandhardworking.Difficultiesinappreciatingthespecificinequityofcapitalismarecompoundedbyimprecisionintherecommendationsofsocialistjustice.NotoriouslyMarxsaidlittleaboutthelineamentsofthefuturesociety,andwhathedidsaysuggestsutopianisminthepejorativesenseofthisword.Talkofasociety‘beyondjustice’mightimplytheviewthatthecircumstancesofjusticewillbetranscended.Thisisunrealistic.Oritmightimplytheirrelevanceofanyevaluativecriteriatosuchasociety.Thisisdangerouslynaive.Marxistscanengagewithothercontemporarypoliticalphilosophersonthegroundsofequality,emphasizingthedeep,structuredinequalitiesofpresentcapitalistsocietyandthemannerinwhichtheseinequalitiescorruptanyformalequalityofcivicrights.Butiftheyaretoberadicalegalitarianstheymaydifferlittlefromsomeliberals,orbebesetbydifficultieswhichderivefromtheircommitmenttoprinciples,suchasthatofself-ownership,whichliberalsdonotshare(Cohen1990).Thereisafinalimportantpoint.Anytheorywhichindictsthepresentsocietyoffundamentalinjusticeandrecommendsafutureperfectsocietyneedsanaccountofthemeansoftransformingthefirstintothesecond.TraditionallyMarxismreliedonsomecombinationofthehistoricalguaranteeofrevolutionarychangeandtheuniquelyimportantroleoftheproletariat.AnalyticalMarxismisnotteleologicallyoptimisticandhastendedtoreticenceaboutproletarianactivism.Thisispartlyduetoitsindividualistpresuppositions,whichmakeithardtoappreciatethefactorsdisposingtoandinhibitingCOLLECTIVEAGENCY(pp.397–9).Itisalsoduetotheneedforanaccountofwhywhatisintheinterestsofthemajorityofsociety’smemberscoincideswiththemorallydesiredemancipationofall.Thisisincreasinglydifficult,forthereisnolongeragroupofindividualswhosatisfythevariousrequirementsofthatorthodoxy–beingtheexploitedbulkofcapitalistsocietywithnothingtoloseandeverythingtogainbyoverthrowingthatsocietyanditsinjustices.AnalyticalMarxismhasperformedasignalservicebybringingMarxismwithinthefoldofcontemporarypoliticalphilosophy.Butitmayhavedonesoatthepriceofexpos-ingseriousifnotfatalshortcomingsinMarxism.AndtotheextentthatanalyticalMarxistshavebecomepoliticalphilosopherstheyhavearguablyceasedtobeMarxists.6IndividualsandCommunitiesAstandardchargeagainstliberalismisthatitisindividualistic.Methodologicalindi-vidualismisdiscussedelsewhere,andthevarioussensesinwhichliberalismmightbesaidtoneglectcommunitywereconsideredinthesectiononcommunitarianism.Whatneedstobeexaminedhereisthesignificancepoliticalphilosophyshouldaccordtogroupsorcommunities,andtherelationshipofindividualstothesecollectiveentities.Forwhileeachofusisanindividualwearealsosocialcreatures,belongingtoparticulartribes,cultures,religions,racesandnations.Todenythesefactswouldbeproductiveofanimplausibletheoryandanimpoverishedpoliticalpractice.Thereareatleasttwowaysinwhichgroupidentityissignificant.Inthefirstinstancethereisthequestionofhowthestateshouldtreattheexistencewithinitsjurisdictionofstable,enduring,well-definedgroupswiththeirownhistory,cultureandwayoflife.274\nPOLITICALANDSOCIALPHILOSOPHYThisistheproblemofculturalpluralism.Inthesecondinstancethereisthequestionofwheretheboundariesofanystateorjurisdictionshouldbedrawnandwhatroleshouldbeplayedinthiscontextbynationhood.Thisistheproblemofnationalityandnationalism.6.1CulturalpluralismAlthoughpoliticalphilosophyhas,fromitsinceptionwiththeGreeks,tendedtoassumetheculturalorethnichomogeneityofthe‘people’whoseobligationalegitimatestatecommandsandtowhomasetofprinciplesofjusticemayapply,thefactisthatallmodernsocietiescomprisedistinctstablegroupswhosemembersidentifythemselves–andareidentifiedbyothers–byreferencetosomecombinationofsharedrace,religion,nationality,language,cultureorhistory.Howshouldthestaterespond?Itcouldinsistondenyingthefactofdifferenceeitherbyenforcinga‘republican’idealwhereincitizenshavenoallegiancesoridentitiesotherthanthosewhichconsti-tutethemasmembersofthepolity,orbysupportingassimilationistpracticeswherebymembersofculturalminoritiesmustacquiretheidentityofthedominantcommunity.Suchmeasuresofcompulsoryhomogenizationarewidelyperceivedasunfairindenyingtoindividualssomethingofgreatvalue,namelytheexpressionoftheirownparticularcommunalidentity.Aculturallypluraliststate,bycontrast,honourstheexistenceofpluralidentitiesbymeasureswhichmayrangefromunderwritingtherightofpersonsbelongingtominoritiestoenjoytheirownculture,to‘communalist’measureswhichpositivelyprotectandpreservethedistinctgroupsrecognizedwithinsociety.Liberalsareunwillingtoseethewarrantforsuchmeasuresintheexistenceofgrouprightswhichdonotstraightforwardlyreducetotherightsoftheindividualmembersofthegroupinquestion.Rathertheyhavearguedthattheprotectionofgroupscanbedefendedinsofarasdoingsoprotectsandadvancestheinterestsofindividualsasmembersofgroups(Raz1986:207–9).Thevalueofacultureissaidtobethatofcul-turalmembership,itsvaluetotheindividualswhoaremembers(Buchanan1991).Inanoriginalandarrestingargument,Kymlicka(1989)rejectstheideaofgrouprightsbutcommendsapolicyofactivelyseekingtopreservecultures.Hedoessobyarguingthatourculturalmembershipisagoodinsofarasitprovidesthenecessarycontextfromwithinwhichweareabletomakeandevaluatemeaningful,rational,autonomouschoicesoflife.CharlesTaylor(1992)hastriedtodefendaliberalismwhichmightpermitthestatetonourishandprotectaparticularculture,solongasitwasalsoabletosecuretherightsofthosewhodonotsharethedominantculture.Themajorproblemisthatliberalismmaybeleftwithnoothercriterionforapprais-ingdifferentculturesthantheextenttowhichtheynurturethekindofindividualsfavouredbyliberals.Shouldaliberalsocietytolerateaminorityculturewhichdoesnotrespectautonomy,evenifitsuppliesitsmemberswithanotherwiseworthwhilelife?Isculturalpluralismagoodthingonlyifthevariousculturesareallconsistentwithliberalidealsofindividuality?Moreoveranyaccountofculturalpluralismmustacknowledgethepossibledisadvantagesofdiversity.Thesemayincludetheincreasedpossibilityofsocialconflict,andtheunderminingofthesharedsenseofcommunitywhichisnecessaryforgoodpoliticalorder.275\nDAVIDARCHARD6.2NationalismandnationalityPoliticalphilosophyhas,untilrecently,remainedlargelysilentonthequestionsofnationalismornationality;orithasdismissedsuchquestionsassomehowunworthyofphilosophicalconsideration(PettitandGoodin1993:7).Yetthereisnomoresalientfactaboutthecontemporaryworld,noranymorepotentsourceofconflictandviolence,thantheexistence,actualanddisputed,ofnations.Itisalsonoteworthythatpoliticalphilosophyhasassumedtheexistenceofdistinctnation-statesandconcerneditselfwiththeapplicationtosuchentitiesofprinciplesofjustice,equalityandrights(Canovan1996).Althoughtheyarecloselyrelatedterms,oftenusedassynonyms,andfrequentlyhyphenated,‘state’and‘nation’havedistinctmeanings.Anationisacommunityofpeopleboundtogetherovertimebysomesignificant,sharedcharacteristicsuchaslanguage,raceorculture.Astateisanindependent,sovereignpoliticalassociationofpeopleinhabitingaboundedterritory.Nationalism,asadoctrine,makesthefactualclaimthathumanityisandalwayshasbeennaturallydividedintonations.Thenormativeclaimsofnationalismarethatnationsshouldbestatesandthatstatesshouldbenations.Contemporaryphilosophicaldefendersofnationalism(Miller1995;Tamir1993)havesoughttoanswerthecriticismsofclaimsofnationalism,andtoshowthatadefenceofnationalismisconsistentwith,indeeddemandedby,aproperdefenceofliberalism.Tothechargethatnationsarefictiveproductsofmodernity,defendersofnationalismwillinsistthatmodernnationsdohavepremodernethnicorigins,andthatevenfalsebeliefscanhaveinstrumentalvalueiftheysustainavaluablesenseofcommunity.Statesshouldbenationsbecauseaprincipleofnationalitymaysupplythe‘fellowfeeling’J.S.Millthoughtnecessaryforgoodgovernment(Mill1975:ch.16)orfortheacceptanceofredistributiveprinciplesofjustice(Miller1995).Nationsshouldbestatesbecausedemocraticself-governmentiscoextensivewithnationalself-determination,andbecausemembershipofanationbeingaconstitutiveelementofindividualidentityandwell-being,statehoodisanessentialmeansofprotectingnationality.Tothosewhoinsistthatthedefenceofnationalityrepresentsanunwarrantedpartialityinconsistentwiththecosmopolitanismwhichproperlyrealizestheglobalscopeofanyacceptablepoliticalphilosophy,thefriendsofnationalismwillinsistthatitisnecessarytomakeanaccommodationwiththerealitiesofthemodernworld.Furtherpartialityisnotmorallyunjustifiedsolongasitsdemonstrationremainsconstrainedbyarecognitionoftheminimumowedtothosewhoarenotone’sownco-nationals.However,criticsofnationalismwillreplythatsincethereisnowell-boundedterritorycontaininganethnicallyhomogeneouspopulationthedemandsofnationalismareproductiveofgrosslyunacceptablepoliticalresults:atbestdiscriminationagainstnationalminoritieswithinthestate,atworsttheforcibleexpulsionandslaughteroftheseminorities.Moreoverthenumberofpotential,aspirantnation-statesgreatlyexceedsthenumberofpossibleviablestates,andinternalsecessionwouldcontinuetothepointofnon-viability.Sympatheticpoliticalphilosophicaltreatmentsofnationalitymaythusstopshortofconcedingthefulldemandsofnationalism,arguing,forinstance,thatthoseends276\nPOLITICALANDSOCIALPHILOSOPHYthoughttorequirethecoincidenceofnationandstatemaybeservedbysomethingless,suchasfederalism(Buchanan1991).Unsympathetictreatmentsofnationalitymayinsistthatitispossibletoconstructanon-nationalprincipleofpoliticalcommunity,aroundforinstanceloyaltytotheconstitutiveprinciplesofaparticularpolity(Habermas1992).Bothcampsmayseektodefusethestingofnationalismbyencouragingapoliticaldisassociationof‘nation’and‘state’.Thismaybemanagedthroughthedevelopmentoftrans-andinternationalinstitutions,theprotectionofsubnationalethnicplurality,andtheprovisionfordecentralizedpoliticalrepresentation.Politicalphilosophycannotnowignorethebrutefactsofnationality.Aproperacknowledgementofthesefactsshouldconsistinarecognitionofwhatcanbechangedandwhatcannot,and,inconsequence,ofthefactthatwhatevervaluenationalidentityhasmaybesecuredonlyattheexpenseofinseparabledisadvantages.7PoliticalPhilosophyandPoliticsThislastsectionconsiderswhatispoliticalphilosophyandwhatispolitics.First,then,whatisitthoughtthatapoliticaltheoryshoulddo?Idonotmeanbythisthequestionofwhetheratheoryshouldbeaboutjustice,thestate,theclassstruggle,orwhatever.Imeanwhatisatheorytakentobedoingwhenitanswersaself-chosenquestionsuchas‘Whatisjustice?’Therearefourbroadsortsofanswer.ThefirstisthatatheoryshouldbebuiltuponFOUNDATIONS(pp.40–1)thatareuni-versallyrecognizedandevidenttruths;itsprocedureofargumentationandreasoningshouldsimilarlybeunexceptionableandacceptabletoanyrationalperson.Itislikelythatthesefoundationswillcomprisesomeunderstandingoftheindividualhumanbeing.BothRawls’sATheoryofJusticeandNozick’sAnarchy,State,andUtopiamayberep-resentedaspoliticaltheoryinthisfoundationalsense.Theproblemwiththisapproachisthatitrisksbeingemptyorcontentious.Eitherthefoundationalunderstandingofthehumanbeingcommandsuniversalagreementinvirtueofbeingemptiedofrealsubstantivecontent,orisarecognizablyinterestingviewbuttherebypartialandcontroversial.AfamiliarcriticismofRawlsisthathiscontractingindividualsarenotthoseofanytimeandplace,buthistoricallyandculturallylocatedagents–Western,liberalmen.Aresponsetotheseproblemsisself-consciouslytoadoptastancethatisrootedinone’sowncultureandhistory.BorrowingPlato’sfamousmetaphor,MichaelWalzercontrastsaphilosophicalattitudethatwalksoutofthecaveandseeksthehighgroundofuniversalobjectivitywithonethattakesitsstandinthecaveandwithitsinhabi-tants.This‘wayofdoingphilosophyistointerprettoone’sfellowcitizenstheworldofmeaningsthatweshare’(Walzer1983:xiv).Theproblemswiththisapproacharethreefold.First,ashasalreadybeennoted,itcourtsrelativism:thegoodsocietyiseachandeverysocietythatjudgesitselftobegoodbyitsownlights.Second,asociety’spractices,howeverunjusttheymayseemtothereasonableoutsider,arenotopentocriticismif,ontheinside,thatsocietylacksthesharedmeaningswhichcouldinformsuchacritique.Third,theapproachtendstoconservatism.Itimpliesthatasociety’smeaningsaredefinedbyitsexistingpractices.Butthereisnospacefromwhichtojudge277\nDAVIDARCHARDthesepractices.Or,putanotherway,themeaningsbywhichthepracticescouldbejudgedappeartofloatoutsidethesociety(Cohen1986).Rawlsappearstohavemovedfromhisoriginalfoundationalismtowardsthissecondunderstandingofpoliticaltheory.InPoliticalLiberalism(Rawls1993)heviewstheliberalconceptionofjusticeasexpressingideasimplicitwithintheinstitutionsandpubliccultureofconstitutionaldemocraticregimes.Yetthisisnotasimpleshiftfromuniversaltoparticular,fromtranscendentalrulesofjusticetotheparticularvirtuesofAmericanconstitutionalism.‘Kantianisminonecountry’isafineandwittydefinitionofRawls’spoliticalphilosophy.Itispossible,andplausible,tobelievethatmoderndemocracyisanhistoricalachievementwhichinstitutesgeneralmoralprinciples–equalityofrespectforindividualchoicesoflife,andthepublicjustifiabilityofagreedrulesofsocialco-operation.Actuallyexistentinstitutionsmaydonomorethanapproximatetotheseideals,buttheyneverthelessaspiretothem.AtthesametimeRawlswouldnowinsistthatpoliticalphilosophyisturnedtowhenasociety’ssharedunderstandingsbreakdownandcomeintoconflict.Philosophyresolvestheseconflictsbyascendingtoabstractionswhichareneverthelessalsosomehowuncoveredthroughfundamentalideasdeepandimplicitwithinthesociety’sculture(ibid.:44–6).Thisissomewhatmysterious.Itmayalsobetrayafalseoptimismthatthepluralityoffundamentalbeliefswhichcharacterizesmodernsocietyis,atsomelevel,eliminable.Ontheotherhand,iftheimplicationisthatwecanreachoutsidethetermsofourpresent,publiccultureforvalueswhichvalidateit,thentoomanycrucialquestionshavebeenbegged.ThisatleastisthechargeofRichardRORTY(pp.783–4)(1989,1991).Whetherornotdemocraticliberalismstandsattheendofideologicalhistoryisaquestionwhichanglophonephilosophyisonthewholeill-equippedtoacknowledgeoranswer,notleastbecauseitlacksthemeanstotheorizeideasofmodernityandhistoricalprogress.PhilosophyintheEnglish-speakingworldisnotoriouslyincapableof‘reflexivesocialunderstanding’;thatis,beinginapositiontointerrogateitselfaboutitsrelationtothesocietyandhistoryinwhichitissituated(Williams1980).Onewaytoopenaspacebetweenpresentlysharedmeaningsandpreferredalterna-tivesisgivenbythethirdapproachtopoliticalphilosophywhichmaybetermedCRITIQUE(chapter32).Marxfamouslyremarkedthatsciencewouldbeunnecessaryifrealityandappearancecoincided.Hisscientificconcernwastodisclosetherealworkingsofcapitalismwhichcouldnotbeapparenttothosewholivedandworkedunderit.Analogouslythetaskofpoliticalphilosophymaybetorevealwhatoursharedmean-ingsdonotandcannotsayaboutourpoliticalworld.Thefunctionofcritiqueistodisplaythesegapsandthusthedistancebetweenactualityandthemoralpretensionsgeneratedbythatactuality.TotheextentthatanalyticalMarxismsharesmanyofthepremisesandconceptsofcontemporaryphilosophicalliberalism,itmaybesaidtohaveeschewedcritiqueinfavourofcriticism.However,feministpoliticaltheorydoesaspiretocritiquewhenitseekstoshowhowunspokenassumptions–aboutwomen’snature,thepublic–privatedivide,thefamily–explainwhyliberalismisbothsilentabout,andyeteloquentinitsimplicitendorsementof,women’scontinuedsubordination.Thefourthandfinalapproachtopoliticalphilosophyishostiletoanymisjudgedrationalistambitionstheorymighthave.Itrepudiatestheideathatpoliticalassociationmustself-consciouslyrealizeanydesiredendorpurposes.Insteaditseesthetasksof278\nPOLITICALANDSOCIALPHILOSOPHYpoliticaltheoryasmoremodestandrestricted.Itisthearticulationofthecommonlyacknowledgedrulesofconductwhichinformouractualpractices.Itsalutesthehis-toricalachievementsofmodernliberalism,notleasttheemergenceof‘individuality’and‘civility’;thatis,lawandcivicorder.Yetitrefusestoseetheseeminentlypracticalachievementsasconformingtouniversal,objectiveprinciplesthatmaybelaidbare.Politicaltheorizingshouldbefaithfultotheknowledgealreadypresupposedinourpractice;itshould‘pursuetheintimations’ofestablishedtraditions.TheauthorofthisapproachisMichaelOakeshott(1962,1975).Hisscepticalandnuancedconservatismdoesnotfigurelargeinmostsurveysofcontemporarypoliticalphilosophy,perhapsbecausehespeaksatatangenttoitsmainconcerns.YethisworkhasbeeninfluentialinaBritishconservatism,ablyrepresentedbyRogerScruton(1984),whichkeepsitsdistancefrombothAmericanphilosophicalliberalismandthefreemarket,anti-statismthatoftenpassesforcontemporaryright-wingthinking.AtthesametimethereisinsuchworkaHegeliansensitivitytohistoryandpracticewhichimportantlydistinguishesitfromthesimpleappealtosharedmeaningswhichcharacterizesthesecondapproach.Oursecondimportantquestionis,‘Whatispolitics?’Thatis,whatisthescopeofpoliticalactivity,whatdistinguishesitfromotherformsofhumanactivity,andwhatisitsimportance?Hereitisbesttoconsideranumberofoppositions.First,thereisthatbetweenwhatwecouldcallinstrumentalandexpressivistunderstandingsofpolitics.Accordingtotheinstrumentalunderstanding,politicsisameansusedbyessentiallyindependentindividualstosecuretheagreementswhicharenecessaryiftheyaretoobtainthebenefitsofco-operationandavoidthecostsofnon-agreement.Itisthepoliticsofbargaining,compromise,accommodation,andachievingamodusvivendi.Thepoliticalsphereislikeamarket-placeinwhichindividualscometomaketheirseparatedealsandthenreturntolivetheirlives.Politicshasnofurtherfunctionthanisnecessarytofacilitateandprotecttheagreementsmade.OntheEXPRESSIVIST(pp.384–90)understanding,politicalactivityisitselfvaluable.Itisanimportantwayinwhichhumanbeingsexpressthemselvesassocial,co-operativecreaturesenjoyinganinterdependentexistence.Thisisthepoliticsofparticipation,communityandrepublicancitizenship.Thepoliticalsphereisaforuminwhichpeoplecometogetherascitizens,andavitalconstituentofthefulllifeledbyall.Politicsisimpoverishedtotheextentthatitfallsshortofexhibitingthischaracter.Onthewholeliberals,andcertainlylibertarians,havefavouredtheformerunder-standing.Theirpreferencederivesfromaccordingaprimaryimportancetolibertyandbelievingthatindividualpurposeswillbevarious.Politicalactivism,ofthesortenvis-agedbyexpressivists,isunlikelytobespontaneouslyuniversal,andwillbeunaccept-ableifcoerced.Perhapsinpartforthisreasonliberals,despiteaclearpreferenceforitasaregime,havehadlittletosayabouthowdemocracyshouldactuallyworkandwhatparticularformitshouldtake.DemocracytendstobeseenasmerelytheformalizedprocessbywhichSOCIALCHOICES(pp.391–5)aregeneratedfromtherationalprefer-encesofindividuals.Famousparadoxesanddifficultiescharacterizethetranslationofmanyindividualchoicesintoasinglepublicchoice.Bycontrastthereisanimportantandinfluentialdeliberativemodelofdemocraticpolitics.AkeyfigureisJürgenHabermas(1992),whohassoughttodemonstratetheprincipledpresuppositionsofeverydaycommunication,suchasthespeaker’simplicit279\nDAVIDARCHARDclaimstobetruthful,rightandsincere.Fromthesefoundationshedefendsaviewofmoralnormsasthosewhichwould,andcould,beagreeduponbythemembersofacommunicativecommunitywhorecognizeonlytheforceofthebetterargument.Therelevanceofsuchideastoanidealofdeliberativeordiscoursepoliticswherebyindividualsreachreasonedagreementonthecollectivegoodisevident.Theproblemsarealsoclear.Ifthemodelisintendedtobeadescriptionoftheactualpoliticalworlditisatbestunrealistic,andarguablyundervaluesdisagreementandcontestationasconstitutivefeaturesofpoliticalactivity.Politicsmaybeineliminablyagonistic.Ifthemodelisanidealthenitneedstobeshownthatitdoesnotimportandemployvalues,suchasequality,whichcannotbefoundsolelyinthenatureandpresuppositionsofconversationalspeech.Theconcernofsocialistsandfeministswithdemocratictheoryandpracticeisattrib-utableinparttotheexpressivistidealofpoliticalparticipationasagood(Pateman1970).Itisalsoduetoaworryaboutreconcilingtheformalequalityofpoliticalciti-zenshipwiththevariousinequalities,ofrace,class,andgender,whichcharacterizecivilsociety.Relatedtobutdistinctfromtheinstrumental–expressivistdistinctionisthatbetweenpublicandprivate,alreadyintroducedinsection5.Asthisdistinctionisnowunder-stood,thepubliccomprisesthepolitical,legalandeconomic,theworldinwhichindi-vidualswork,voteandareaccountabletotherestofsocietyfortheiractions.Theprivateisthepersonalandfamilial,thesphereofthehouseholdinwhichindividualslove,playandgenerallyretreatfromthepublicworld.Tothisdistinctioncorrespondsdifferencesofmotivation,relatednessandloyalty.Thepublicworldiscold,impersonal,governedbyabstractrules,inwhichindependentandmutuallydisinterestedindividualsmeet.Theprivateistheemotionallywarmhaveninwhichindividualsareboundbyparticularrelationshipsofaffection,loyaltyandmutualinterdependence.Liberalshave,onthewhole,refusedtoseetheprivateaspolitical.Indeedtheprivatesphereisthatwithwhichthestatehasnobusinessandintowhichitshouldnotintrude.Thisisplausibleinsofarasliberalshavecharacterizedprivateactivityasself-regardinginJ.S.Mill’ssense;thatis,whoseharmfuleffects,ifany,areconfinedtotheagent.Consensualsexualbehaviourandprocreativedecisions,accordingtoliberals,areparadigmaticallyprivateinthisregard.Itistofeminismthatweoweacritiqueofthepublic–privatedivide.Thisinsistsfirstthatthescopeandcharacteroftheprivateisdeterminedpubliclybylawandpolicy.Itisnotthattheprivatedomainpre-existsandlimitsthatofthepublic.Ratherthetermsofprivacyaresetwithinthepublicsphere.Indeedtheveryinaugurationoftheprivate,andcivilsociety,maybeduetoanunnoticedbutdeeplyillegitimate‘sexualcontract’inwhichthemalefirstsubordinatedthefemaletoservehissexualpurposes(Pateman1988).Second,theprivateispoliticalif‘political’extendstodescribeanystructureorrelationshipinwhichsomeindividualsexercisesignificantpoweroverothers.Thefamilyistheobjectofinteresthere.Thetraditionalinstitutionismarkedbyafamiliarsexualdivisionoflabourwiththehusbanddominantoverthewife.Anycharacteriza-tionoffamilialrelationsasprivateamountstoapublicendorsementoftheircontinu-inginequity.280\nPOLITICALANDSOCIALPHILOSOPHYEvenontheirownprinciplesliberalsshouldbeinterestedinthefamily.Foritcansupportorunderminebroadersocialjustice.Thisisnotonlyinsofarasthefamilyis,asnotedearlier,asiteofmoraleducation,aplacewhereasenseofjusticeislearned.Itisalsothatthefamilyisthemajormeansbywhichassets–propertyaswellasnaturalendowments–aretransferredacrossgenerations.Differencesbetweenfamilialsituationsremainacrucialobstacletosecuringequalityofopportunity(Fishkin1983).Thefinaloppositionsetsthepoliticalagainsttheeconomic.Althoughtheeconomicisincludedwithinthepublicsphere,thereisatendencyamongliberalstoberelativelyindifferenttowardsvariouseconomicarrangements.These,tosimplifytheirview,aretobepreferredongroundsofefficiencynotmorality.InATheoryofJusticeRawlsleavesopenthechoicebetweenaprivatepropertyeconomyandsocialism(Rawls1972:258).InPoliticalLiberalismhemaintainsthatthequestionis‘notsettledattheleveloffirstprinciplesofjustice’,butdependsonthecontingenciesofacountry’sparticularinstitutionsandhistoricalcircumstances(Rawls1993:338).Buttheeconomicispoliticalinthreeimportantregards.First,liberalprinciplesofjusticeapplytothedistributivesphere,thatistheydeter-minewhogetswhatgoodsoncetheyareproduced.Butprinciplesrestrictedtodistrib-utionleaveoutofaccounttheproductivesphere,concerningwhatisproducedbywhom.Decisionsmaybetakenastowhatgoodsareproducedandinwhatquantities.Itcansometimesseemasifthegoodstobedistributedinliberaltheoriesofjusticehavejustdroppedfromheaven,quiteexplicitlyinBruceAckerman’s(1980)account,wherehespeaksofmanna.Second,thereisastrongerclaimassociatedwithMarxthatproductiondeterminesdistribution;thatis,thatadistributionofgoodstobeconsumedisaconsequenceofhowproductionitselfisorganized.Ifthisdeterministthesisistruethenprinciplesofdistributivejusticeareinanimportantsensebesidethepoint.Third,theeconomicsphereisonetowhichpoliticalconsiderationsdirectlyapply.Workconstitutesasignificantpartofanindividual’slife.Initselfitcanberewardingorfrustrating,self-realizingoralienating.Itcanbeself-directedorperformedunderconditionsofsubordinationtoanother.Workisalsoinstrumentallyvaluableinsofaraspaymentforemploymentdeterminesone’slevelofsubsistenceandthecharacterofone’sworkaffectsone’ssocialstatusandself-esteem.Itisimportantthentorecognizethatemployment–anindividual’srighttoit,andthedeterminationofitsconditions–arepropersubjectsforpoliticalinquiry.Socialistshaveattendedtotheseissues.Theyhavedefendedanextensionofdemocracytotheworkplace,andhavebeenmoregenerallyconcernedabouttheproperbalancebetweenmarketandstatecontrol,privateandpublicownership.8ConclusionPoliticalphilosophyisaliveandwell.Itwasnotquitemoribundin1950butthepost-1970revivalnowmakesitappearso.Thecurrentdominanceofphilosophicalliberalismmayirkthosewhofeelthatitsapproachpre-emptivelybiasesconsiderationofimportantmattersandignorescrucialfacts.Butthosewhofeelthatwaymust281\nDAVIDARCHARDrecognizethatitistoliberalsthatweowetherevivalofpoliticalphilosophyandourpresentabilitytospeaktotheproblemsofoursocialandpoliticalexistence.Wemaynotwishtospeakinthelanguageofliberalism.Butthenphilosopherscannolonger,asWeldononceargued,besilent,andtheonusisonthosewhodissenttoperfectandpractisethealternativelanguagesofpoliticalcriticism.FurtherReadingOfthegeneralsurveysofsocialandpoliticalphilosophicalwork,Kymlicka(1990)isthebest.Itiscomprehensive,incisiveandjudicious.Plant(1991)iswell-informedandsuccessfullybringsoutmanyofthemainconcernsofcontemporarypoliticaltheory(thoughheunaccountablyignoresfeminism).Pettit(1980)andBrown(1986)arealsousefulifmorenarrowlyconcernedwiththemaintheoriesofdistributivejustice.PhilosophyandPublicAffairsremainstheflagshipjournalofpoliticalphilosophyinthepresentera.GoodinandPettit(1993)isavolumeinthepresentseries.Itcontainsextendedessaysonvariousdisciplinarycontributionstothesubject,shorterentriesonmajorideologies,andshortnotesonspecialtopics.Thecontributorsaredistinguishedpractitionersofthesubject,thewritingisalmostuniformlyexcellent,andthewholetextsuppliesanilluminatingpictureofpost-Rawlsianpoliticalphilosophy.TheOxfordPoliticalTheoryseries,editedbyDavidMillerandAlanRyan,containsexcellentbooksbydistinguishedandparticularlywellqualifiedcontributorsonparticulartopicswithincontemporarypoliticalphilosophy.OnRawlstheliteratureisvoluminous.Daniels(1975)collectsthebestoftheearliercriticalessays,andotherstandardcommentariesareBarry(1973)andWolff(1977).KukathasandPettit(1990)compares‘early’and‘late’Rawls,butbeforepublicationofPoliticalLiberalism.Paul(1981)doesforNozickwhatDanielsdoesforRawls,andWolff(1991)offersanunprejudicedbutcriticalaccountofNozick’stheory.MulhallandSwift(1996)providesafine,clearlywrittenreviewofthedebatebetweenliberalsandtheircommunitariancritics,evenifitstrugglessomewhattofindthecommonthemesincommunitarianism.InadditiontoGutman(1985),thefollowingalsosupplyusefulevaluationsofcommunitarianism:Kymlicka(1988)andBuchanan(1989).G.A.CohenhasworriedatandaboutthelegacyofMarxanditsrelationshiptocurrentpoliticalphilosophywithmoreacuityandassiduitythananyone.Cohen(1988)collectstogethersomeofhisbestpieces.KittayandMeyers(1987)givesagoodsenseofGilligan’simpactuponmoralandpoliticalthinking.Jaggar(1983)isastandardintroductiontofeministpoliticaltheory,whileGrimshaw(1986)isalsoclearlywrittenandrelevant.AnintroductiontoBritishconservatismisprovidedbyCovell(1986),andFranco(1990)offersacomprehensive,ifcautiousaccountofhissubject.Kymlicka(1995)isacomprehensivecollectionofimportantarticlesonculturalandnationalpluralism,whileBeiner(1999)offersafinecollectionofpiecesbyleadingtheoristsonthetopicofnationalism.Chambers(1996)isaverygoodevaluationoftheidealofHabermasiandeliberativedemocracywithinthecontextofRawlsianpoliticalphilosophy.ReferencesAckerman,B.1980:SocialJusticeintheLiberalState.NewHaven,CT:YaleUniversityPress.Barry,B.1973:TheLiberalTheoryofJustice.Oxford:ClarendonPress.282\nPOLITICALANDSOCIALPHILOSOPHYBeiner,R.(ed.)1999:TheorizingNationalism.Albany:StateUniversityofNewYorkPress.Berlin,I.1958:TwoConceptsofLiberty.InFourEssaysonLiberty.Oxford:ClarendonPress.Brown,A.1986:ModernPoliticalPhilosophy.Harmondsworth:PenguinBooks.Buchanan,A.1989:AssessingtheCommunitarianCritiqueofLiberalism.Ethics,99(July),852–82.——1991:Secession:TheMoralityofPoliticalDivorcefromFortSumtertoLithuaniaandQuebec.Boulder,CO:WestviewPress.Caney,S.1992:LiberalismandCommunitarianism:AMisconceivedDebate.PoliticalStudies,40,273–89.Canovan,M.1996:NationhoodandPoliticalTheory.Cheltenham:EdwardElgar.Chambers,S.1996:ReasonableDemocracy:JürgenHabermasandthePoliticsofDemocracy.Ithaca,NY:CornellUniversityPress.Cohen,G.A.1978:KarlMarx’sTheoryofHistory:ADefence.Oxford:OxfordUniversityPress.——1988:History,Labour,andFreedom:ThemesfromMarx.Oxford:ClarendonPress.——1990:MarxismandContemporaryPoliticalPhilosophy,or:WhyNozickExercisessomeMarxistsmorethanhedoesanyEgalitarianLiberals.CanadianJournalofPhilosophy,supple-mentaryvolume16,363–87.Cohen,J.1986:ReviewofWalzer,SpheresofJustice.JoumalofPhilosophy,83:8,457–68.Covell,C.1986:TheRedefinitionofConservatism.Basingstoke:Macmillan.Daniels,N.(ed.)1975:ReadingRawls.Oxford:Blackwell.Dworkin,R.1978:Liberalism.InS.Hampshire(ed.)PublicandPrivateMorality.Cambridge:CambridgeUniversityPress.——1981:WhatisEquality?Part1:EqualityofWelfareandPart2:EqualityofResources.PhilosophyandPublicAffairs,10:3,185–246,and10:4,283–345.Elster,J.1985:MakingSenseofMarx.Cambridge:CambridgeUniversityPress.Fishkin,J.S.1983:Justice,EqualOpportunity,andtheFamily.NewHaven,CT:YaleUniversityPress.Franco,P.1990:ThePoliticalPhilosophyofMichaelOakeshott.NewHaven,CT:YaleUniversityPress.Galston,W.A.1980:JusticeandtheHumanGood.Chicago:UniversityofChicagoPress.Geras,N.1985:TheControversyaboutMarxandJustice.NewLeftReview,150,47–85.Gilligan,C.1982:InaDifferentVoice.Cambridge,MA:HarvardUniversityPress.Goodin,R.E.andPettit,P.(eds)1993:ACompaniontoContemporaryPoliticalPhilosophy.Oxford:Blackwell.Gray,J.1984:HayekonLiberty.Oxford:Blackwell.Grimshaw,J.1986:FeministPhilosophers.Brighton:Wheatsheaf.Gutman,A.1985:CommunitarianCriticsofLiberalism.PhilosophyandPublicAffairs,14:3,308–22.Habermas,J.1992:CitizenshipandNationalIdentity:SomeReflectionsontheFutureofEurope.PraxisInternational,12,1–33.——1996:MoralConsciousnessandCommunicativeAction(translatedbyChristianLenhardtandShierryWeberNicholsen).Cambridge,MA:MITPress.Haksar,V.1979:Equality,LibertyandPerfectionism.Oxford:OxfordUniversityPress.Jaggar,A.M.1983:FeministPoliticsandHumanNature.Totowa,NJ:RowmanandAllanheld.Kittay,E.F.andMeyers,D.T.(eds)1987:WomenandMoralTheory.Savage,MD:RowmanandLittlefield.Kocis,R.A.1980:Reason,Development,andtheConflictofHumanEnds:SirIsaiahBerlin’sVisionofPolitics.AmericanPoliticalScienceReview,74:1,38–52.Kukathas,C.andPettit,P.1990:Rawls:ATheoryofJusticeanditsCritics.Cambridge:PolityPress.283\nDAVIDARCHARDKymlicka,W.1988:LiberalismandCommunitarianism.CanadianJournalofPhilosophy,18,2(June),181–204.——1989:Liberalism,Community,andCulture.Oxford:ClarendonPress.——1990:ContemporaryPoliticalPhilosophy.Oxford:ClarendonPress.——(ed.)1995:TheRightsofMinorityCultures.Oxford:OxfordUniversityPress.Landesman,B.1983:Egalitarianism.CanadianJournalofPhilosophy,13,27–56.Laslett,P.1950:Introduction.InP.Laslett(ed.)Philosophy,PoliticsandSociety.Oxford:Blackwell.Locke,J.1689:EpistoladeTolerantia.Englishtranslation,ALetteronToleration(introductionandnotesbyJ.W.Gough).Oxford:ClarendonPress.Lukes,S.1985:MarxismandMorality.Oxford:ClarendonPress.MacIntyre,A.1981:AfterVirtue.London:Duckworth.——1988:WhoseJustice?WhoseRationality?London:Duckworth.Mill,J.S.1975[1861]:ConsiderationsonRepresentativeGovernment.InThreeEssays.Oxford:OxfordUniversityPress.Miller,D.1995:OnNationality.Oxford:ClarendonPress.Mulhall,S.andSwift,A.1996:LiberalsandCommunitarians,2ndedn.Oxford:Blackwell.Nagel,T.1991:EqualityandPartiality.Oxford:OxfordUniversityPress.Nozick,R.1974:Anarchy,StateandUtopia.Oxford:Blackwell.Oakeshott,M.1962:RationalisminPolitics.London:Methuen.——1975:OnHumanConduct.Oxford:ClarendonPress.Okin,S.M.1989:Justice,Gender,andtheFamily.NewYork:BasicBooks.Pateman,C.1970:ParticipationandDemocraticTheory.Cambridge:CambridgeUniversityPress.——1988:TheSexualContract.Cambridge:PolityPress.Paul,J.(ed.)1982:ReadingNozick.Oxford:Blackwell.Pettit,P.1980:JudgingJustice.London:RoutledgeandKeganPaul.Pettit,P.andGoodin,R.(eds)1993:ACompaniontoContemporaryPoliticalPhilosophy.Oxford:Blackwell.Plant,R.1991:ModernPoliticalThought.Oxford:Blackwell.Popper,K.1945:TheOpenSocietyanditsEnemies,2vols.Vol.1:Plato.Vol.2:HegelandMarx.London:RoutledgeandKeganPaul.Rawls,J.1972:ATheoryofJustice.Oxford:OxfordUniversityPress.——1993:PoliticalLiberalism.NewYork:ColumbiaUniversityPress.Raz,J.1986:TheMoralityofFreedom.Oxford:ClarendonPress.Rorty,R.1989:Contingency,IronyandSolidarity.Cambridge:CambridgeUniversityPress.——1991:Objectivity,RelativismandTruth.Cambridge:CambridgeUniversityPress.Sandel,M.1982:LiberalismandtheLimitsofJustice.Cambridge:CambridgeUniversityPress.Scruton,R.1984:TheMeaningofConservatism,2ndedn.Basingstoke:Macmillan.Sen,A.1992:InequalityRe-examined.Oxford:ClarendonPress.Tamir,Y.1993:LiberalNationalism.Princeton,NJ:PrincetonUniversityPress.Taylor,C.1985:PhilosophyandtheHumanSciences:PhilosophicalPapers,Vol.2.Cambridge:CambridgeUniversityPress.——1989:Cross-Purposes:TheLiberalCommunitarianDebate.InN.Rosenblum(ed.)LiberalismandtheMoralLife.Cambridge,MA:HarvardUniversityPress.——1992:MulticulturalismandThePoliticsofRecognition(commentarybyA.Gutman,editedbyS.C.Rockefeller,M.WalzerandS.Wolf).Princeton,NJ:PrincetonUniversityPress.Unger,R.M.1975:KnowledgeandPolitics.NewYork:FreePress.Walzer,M.1983:SpheresofJustice.NewYork:BasicBooks.Weldon,T.D.1953:TheVocabularyofPolitics.Harmondsworth:PenguinBooks.Williams,B.1980:PoliticalPhilosophyandtheAnalyticalTradition.InM.Richter(ed.)PoliticalTheoryandPoliticalEducation.Princeton,NJ:PrincetonUniversityPress.284\nPOLITICALANDSOCIALPHILOSOPHYWolff,J.1991:RobertNozick.Cambridge:PolityPress.Wolff,R.P.1977:UnderstandingRawls.Princeton,NJ:PrincetonUniversityPress.DiscussionQuestions1Areequalityandlibertyincompatibleideals?2Isvaluepluralismanineliminablefeatureofmodernsociety?3Arethereanyfeasiblewaysinwhichthecircumstancesofjusticecouldbetranscended?4Howegalitarianisthe‘differenceprinciple’?5WhatrolesdoesthecontractarianargumentplayinRawls’stheoryofjustice?6Shouldatheoryofjusticebe‘metaphysical’or‘political’?7Isthedistributionofnaturalassetsmorallyarbitrary?8Arerights‘sideconstraints’onactions?9Canwebesaidtoownourselves?10Isjusticeamatteroflegitimateacquisitionofgoods,whateverthepatternoftheirdistribution?11Whatshouldtherebeequalityof?12Shouldautonomyratherthanlibertybethecentralvalueofliberalism?13Istheideaofa‘liberalcommunity’acontradictioninterms?14Doesliberalismdependonanunacceptableconceptionoftheindividual?15Isacaringsocietymorallypreferabletoajustone?16Canthefamilyeverbeaforceforjustice?17WhatforMarxistsiswrongwithcapitalism?18Is‘analyticalMarxism’Marxismatall?19Canpoliticalphilosophygobeyondwhatwehappenhereandnowtobelieveandvalue?20Isthehumanbeingapoliticalanimal?21Isanythingreallyprivateinthesenseofbeingbeyondthepublicgazeandpublicconcern?22Istheidealofofficialneutralityonthequestionofthegoodlifeeitherfeasibleordesirable?23Whyshouldminorityculturesbeprotectedorpreserved?24Doesjusticeliebeyond,betweenorwithinnations?25Whatinequalitiesinourmaterialcircumstancesarepermissible?26Isliberalnationalismanoxymoron?27Areweallliberalsnow?285\n9PhilosophyofScienceDAVIDPAPINEAUThephilosophyofsciencecanusefullybedividedintotwobroadareas.Ontheonehandistheepistemologyofscience,whichdealswithissuesrelatingtothejustificationofclaimstoscientificknowledge.Philosophersworkinginthisareainvestigatesuchques-tionsaswhetherscienceeveruncoverspermanenttruths,whetherobjectivedecisionsbetweencompetingtheoriesarepossibleandwhethertheresultsofexperimentarecloudedbypriortheoreticalexpectations.Ontheotherhandaretopicsinthemeta-physicsofscience,topicsrelatingtophilosophicallypuzzlingfeaturesofthenaturalworlddescribedbyscience.Herephilosophersasksuchquestionsaswhetheralleventsaredeterminedbypriorcauses,whethereverythingcanbereducedtophysicsandwhethertherearepurposesinnature.Youcanthinkofthedifferencebetweentheepis-temologistsandthemetaphysiciansofscienceinthisway.Theepistemologistswonderwhetherweshouldbelievewhatthescientiststellus.Themetaphysiciansworryaboutwhattheworldislike,ifthescientistsareright.ReaderswillwishtoconsultchaptersonEPISTEMOLOGY(chapter1),METAPHYSICS(chapter2),PHILOSOPHYOFMATHEMAT-ICS(chapter11),PHILOSOPHYOFSOCIALSCIENCE(chapter12)andPRAGMATISM(chapter36).1TheEpistemologyofScience1.1TheproblemofinductionMuchrecentworkintheepistemologyofscienceisaresponsetotheproblemofinduction.Inductionistheprocesswherebyscientistsdecide,onthebasisofvariousobservationsorexperiments,thatsometheoryistrue.Atitssimplest,chemistsmaynote,say,thatonanumberofoccasionssamplesofsodiumheatedonaBunsenburnerhaveglowedbrightorange,andonthisbasisconcludethatingeneralallheatedsodiumwillglowbrightorange.Inmorecomplicatedcases,scientistsmaymovefromtheresultsofaseriesofcomplexexperimentstotheconclusionthatsomefunda-mentalphysicalprincipleistrue.Whatallsuchinductiveinferenceshaveincommon,\nPHILOSOPHYOFSCIENCEhowever,isthattheystartwithparticularpremisesaboutafinitenumberofpastobser-vations,yetendupwithageneralconclusionabouthownaturewillalwaysbehave.Andthisiswheretheproblemlies.Foritisunclearhowanyfiniteamountofinfor-mationaboutwhathashappenedinthepastcanguaranteethatanaturalpatternwillcontinueforalltime.Afterall,whatrulesoutthepossibilitythatthecourseofnaturemaychange,andthatthepatternswehaveobservedsofarturnouttobeapoorguidetothefuture?Evenifallheatedsodiumhasglowedorangeuptillnow,whoistosayitwillnotstartglowingbluesometimeinthenextcentury?Inthisrespectinductioncontrastswithdeduction.Indeductiveinferencesthepremisesguaranteetheconclusion.Forexample,ifyouknowthatEitherthissubstanceissodiumoritispotassium,andthenlearnfurtherthatItisnotsodium,youcanconcludewithcertaintythatItispotassium.Thetruthofthepremisesleavesnoroomfortheconclusiontobeanythingbuttrue.Butinaninductiveinferencethisdoesnothold.Totakethesimplestcase,ifyouaretold,forpropertiesAandB,thatEachoftheAsobservedsofarhasbeenB,thisdoesnotguaranteethatAllAs,includingfutureones,areBs.Itisperfectlypossiblethattheformerclaimmaybetrue,butthelatterfalse.Theproblemofinductionseemstoposeathreattoallscientificknowledge.Allscientificdiscoveriesworththeirnameareintheformofgeneralprinciples.Galileo’slawoffreefallsaysthat‘Allbodiesfallwithconstantacceleration’;Newton’slawofgravitationsaysthat‘Allbodiesattracteachotherinproportiontotheirmassesandininverseproportiontothesquareofthedistancebetweenthem’;Avogadro’slawsaysthat‘Allgasesatthesametemperatureandpressurecontainthesamenumberofmoleculesperunitvolume’;andsoon.Theproblemofinductioncallstheauthorityofalltheselawsintoquestion.Forifourevidenceissimplythattheselawshaveworkedsofar,thenhowcanwebesurethattheywillnotbedisprovedbyfutureoccurrences?1.2Popper’sfalsificationismOneinfluentialresponsetotheproblemofinductionisduetoSirKarlPopper(1902–94).InPopper’s(1959a,1963,1972)view,sciencedoesnotrestoninductioninthefirstplace.Popperdeniesthatscientistsstartwithobservations,andtheninferageneraltheory.Rather,theyfirstputforwardatheory,asaninitiallyuncorroboratedconjecture,andthencompareitspredictionswithobservationstoseewhetheritstandsuptotest.Ifsuchtestsprovenegative,thenthetheoryisexperimentallyfalsified,andthescientistswillseeksomenewalternative.If,ontheotherhand,thetestsfitthetheory,thenscientistswillcontinuetoupholdit–notasproventruth,admittedly,butneverthelessasanundefeatedconjecture.Ifwelookatscienceinthisway,arguesPopper,thenweseethatitdoesnotneedinduc-tion.AccordingtoPopper,theinferenceswhichmattertosciencearerefutations,whichtakesomefailedpredictionasthepremise,andconcludethatthetheorybehindthatpre-dictionisfalse.Theseinferencesarenotinductive,butdeductive.WeseethatsomeAisnot-B,andconcludethatitisnotthecasethatAllAsareBs.Thereisnoroomhereforthepremisetobetrueandtheconclusionfalse.Ifwediscoverthatsomebodyfallswith287\nDAVIDPAPINEAUincreasingacceleration(saybecauseitfallsfromagreatheight,andsoissubjecttoagreatergravitationalforceasitnearstheearth),thenweknowforsurethatallbodiesdonotfallwithconstantacceleration.Thepointhereisthatitismucheasiertodisprovetheoriesthantoprovethem.Asinglecontraryexamplesufficesforaconclusivedisproof,butnonumberofsupportingexampleswillconstituteaconclusiveproof.So,accordingtoPopper,scienceisasequenceofconjecturesandrefutations.Scien-tifictheoriesareputforwardashypotheses,andtheyarereplacedbynewhypotheseswhentheyarefalsified.However,ifscientifictheoriesarealwaysconjecturalinthisway,thenwhatmakessciencebetterthanastrology,orspiritworship,oranyotherformofunwarrantedsuperstition?Anon-Popperianwouldanswerthisquestionbysayingthatrealscienceprovesitsclaimsonthebasisofobservationalevidence,whereassupersti-tionisnothingbutguesswork.ButonPopper’saccount,evenscientifictheoriesareguesswork–fortheycannotbeprovedbytheobservations,butarethemselvesmerelyundefeatedconjectures.Poppercallsthisthe‘problemofdemarcation’:whatisthedifferencebetweenscienceandotherformsofbelief?Hisansweristhatscience,unlikesuperstition,isatleastfalsifiable,evenifitisnotprovable(Popper1959a:ch.2).Scientifictheoriesareframedinpreciseterms,andsoissueindefinitepredictions.Forexample,Newton’slawstellusexactlywherecertainplanetswillappearatcertaintimes.Andthismeansthatifsuchpredictionsfail,wecanbesurethatthetheorybehindthemisfalse.Bycontrast,beliefsystemslikeastrologyareirredeemablyvague,inawaywhichpreventstheireverbeingshowndefinitelywrong.AstrologymaypredictthatScorpioswillprosperintheirpersonalrelationshipsonThursdays,butwhenfacedwithaScorpiowhosespousewalksoutonaThursday,defendersofastrologyarelikelytorespondthattheendofthemarriagewasprobablyforthebest,allthingsconsidered.Becauseofthis,nothingwilleverforceastrologiststoadmittheirtheoryiswrong.Thetheoryisphrasedinsuchimprecisetermsthatnoactualobservationscanpossiblyfalsifyit.Popperhimselfusesthecriterionoffalsifiabilitytodistinguishgenuinescience,notjustfromtraditionalbeliefsystemslikeastrologyandspiritworship,butalsofromMarxism,psychoanalysisandvariousothermoderndisciplinesthathedenigratesas‘pseudo-sciences’.AccordingtoPopper,thecentralclaimsofthesetheoriesareasunfal-sifiableasthoseofastrology.Marxistspredictthatproletarianrevolutionswillbesuc-cessfulwhenevercapitalistregimeshavebeensufficientlyweakenedbytheirinternalcontradictions.Butwhenfacedwithunsuccessfulproletarianrevolutions,theysimplyrespondthatthecontradictionsinthoseparticularcapitalistregimeshavenotyetweak-enedthemsufficiently.Similarly,psychoanalytictheoristswillclaimthatalladultneu-rosesareduetochildhoodtraumas,butwhenfacedbytroubledadultswithapparentlyundisturbedchildhoods,theywillsaythatthoseadultsmustneverthelesshaveunder-goneprivatepsychologicaltraumaswhenyoung.ForPopper,suchploysaretheantithesisofscientificseriousness.Genuinescientistswillsaybeforehandwhatobser-vationaldiscoverieswouldmakethemchangetheirminds,andwillabandontheirtheoriesifthesediscoveriesaremade.ButMarxistsandpsychoanalytictheoristsframetheirtheoriesinsuchaway,arguesPopper,thatnopossibleobservationsneedevermakethemadjusttheirthinking.288\nPHILOSOPHYOFSCIENCE1.3ThefailingsoffalsificationismAtfirstsightPopperseemstoofferanextremelyattractiveaccountofscience.Heexplainsitssuperiorityoverotherformsofbelief,whileatthesametimeapparentlyfreeingitfromanyproblematicdependenceoninduction.Certainlyhiswritingshavestruckachordwithinthescientificcommunity.PopperisoneofthefewphilosophersevertohavebecomeaFellowoftheRoyalSociety,anhonourusuallyreservedforeminentscientists.Inthephilosophicalworld,however,Popper’sviewsaremorecontroversial.Thisisbecausemanyphilosophersfeelthathisaccountofsciencesignallyfailstosolvetheproblemwithwhichhebegins,namely,theproblemofinduction(forexample,seeAyer1956:71–5;Worrall1989).Thecentralobjectiontohispositionisthatitonlyaccountsfornegativescientificknowledge,asopposedtopositiveknowledge.Popperpointsoutthatasinglecounter-examplecanshowusthatascientifictheoryiswrong.Buthesaysnothingaboutwhatcanshowusthatascientifictheoryisright.Yetitispositiveknowl-edgeofthislatterkindthatmakesscienceimportant.Wecancurediseasesandsendpeopletothemoonbecauseweknowthatcertaincausesdoalwayshavecertainresults,notbecauseweknowthattheydonot.Usefulscientificknowledgecomesintheform‘AllAsareBs’,not‘It’sfalsethatallAsareBs’.SincePopperonlyaccountsforthelatterkindofknowledge,heseemstoleaveoutwhatismostinterestingandimportantaboutscience.Popper’susualanswertothisobjectionisthatheisconcernedwiththelogicofpurescientificresearch,notwithpracticalquestionsabouttechnologicalapplications.Sci-entificresearchrequiresonlythatweformulatefalsifiableconjectures,andrejectthemifwediscovercounter-examples.Thefurtherquestionofwhethertechnologistsshouldbelievethoseconjectures,andrelyontheirpredictionswhen,say,theyadministersomedrugorbuildadam,Popperregardsasanessentiallypracticalissue,andassuchnotpartoftheanalysisofrationalscientificpractice.Butthiswillnotdo.Afterall,Popperclaimstohavesolvedtheproblemofinduction.Buttheproblemofinductionisessentiallytheproblemofhowwecanbasejudgementsaboutthefutureonevidenceaboutthepast.Ininsistingthatscientifictheoriesarejustconjectures,andthatthereforewehavenorationalbasisforbelievingtheirpredictions,Popperissimplydenyingthatwecanmakerationaljudgementsaboutthefuture.Considerthesetwopredictions:(1)whenIjumpfromthistenth-floorwindowIshallcrashpainfullyintotheground;(2)whenIjumpfromthewindowIwillfloatlikeafeathertoagentlelanding.Intuitively,itismorerationaltobelieve(1),whichassumesthatthefuturewillbelikethepast,than(2),whichdoesnot.ButPopper,sinceherejectsinduction,iscommittedtotheviewthatpastevidencedoesnotmakeanybeliefsaboutthefuturemorerationalthananyothers,andthereforethatbelieving(2)isnolessrationalthanbelieving(1).Somethinghasgonewrong.Ofcoursebelieving(1)ismorerationalthanbelieving(2).Insayingthis,Idonotwanttodenythatthereisaproblemofinduction.Indeeditispreciselybecausebelieving(1)ismorerationalthanbelieving(2)thatinductionisproblematic.Everybody,Popperaside,canseethatbelieving(1)ismorerationalthanbelieving(2).Theproblemisthentoexplainwhybelieving(1)ismorerationalthanbelieving(2),inthefaceoftheapparentinvalidityofinduction.SoPopper’sdenialof289\nDAVIDPAPINEAUtherationalsuperiorityof(1)over(2)isnotsomuchasolutiontotheproblemofinduc-tion,butsimplyarefusaltorecognizetheprobleminthefirstplace.Evenifitfailstodealwithinduction,Popper’sphilosophyofsciencedoeshavesomestrengthsasadescriptionofpurescientificresearch.Foritiscertainlytruethatmanyscientifictheoriesstartlifeasconjectures,injustthewayPopperdescribes.WhenEinstein’sgeneraltheoryofrelativitywasfirstproposed,forexample,veryfewscien-tistsactuallybelievedit.Insteadtheyregardeditasaninterestinghypothesis,andwerecurioustoseewhetheritwastrue.Atthisinitialstageofatheory’slife,Popper’srecommendationsmakeeminentsense.Obviously,ifyouarecurioustoseewhetheratheoryistrue,thenextstepistoputittotheobservationaltest.Andforthispurposeitisimportantthatthetheoryisframedinpreciseenoughtermsforscientiststoworkoutwhatitimpliesabouttheobservableworld–thatis,inpreciseenoughtermsforittobefalsifiable.Andofcourseifthenewtheorydoesgetfalsified,thenscientistswillrejectitandseeksomealternative,whereasifitspredictionsareborneout,thenscientistswillcontinuetoinvestigateit.WherePopper’sphilosophyofsciencegoeswrong,however,isinholdingthatsci-entifictheoriesneverprogressbeyondthelevelofconjecture.AsIhavejustsuggested,theoriesareoftenmereconjectureswhentheyarefirstputforward,andtheymayremainconjecturesastheinitialevidencefirstcomesin.Butinmanycasestheaccu-mulationofevidenceinfavourofatheorywillmoveitbeyondthestatusofconjecturetothatofestablishedtruth.Thegeneraltheoryofrelativitystartedlifeasaconjecture,andmanyscientistsstillregardeditashypotheticalevenafterSirArthurEddington’sfamousinitialobservationsin1919oflightapparentlybendingnearthesun.Butbynowthisinitialevidencehasbeensupplementedwithevidenceintheformofgravita-tionalred-shifts,time-dilationandblackholes,anditwouldbeaneccentricscientistwhonowadaysregardedthegeneraltheoryaslessthanfirmlyestablished.Suchexamplescanbemultiplied.Theheliocentrictheoryofthesolarsystem,thetheoryofevolutionbynaturalselectionandthetheoryofcontinentaldriftallstartedlifeasintriguingconjectures,withlittleevidencetofavourthemovertheircompetitors.Butintheperiodsincetheywerefirstproposedthesetheorieshaveallaccumulatedagreatwealthofsupportingevidence.Itisonlythosephilosopherswhohavebeenbemusedbytheproblemofinductionwhoviewthesetheoriesasbeingnobetterthaninitialhypotheses.Everybodyelsewhoisacquaintedwiththeevidencehasnodoubtthatthesetheoriesareproventruths.1.4BayesianismIfweinsist,againstPopper,thatwearefullyentitledtobelieveatleastsomescientifictheoriesonthebasisofpastevidence,thenwearecommittedtofindingsomesolutiontotheproblemofinduction.Onecurrentlypopularaccountofthelegitimacyofinduc-tionisfoundwithinBayesianism,namedafterThomasBayes(c.1701–61)(Horwich1982;HowsonandUrbach1989).Bayesiansarephilosopherswhoholdthatourbeliefs,includingourbeliefsinscien-tifictheories,comeindegrees.Thus,forexample,Icanbelievetodegree0.5thatitwillraintoday,inthesensethatIthinkthereisa50percentlikelihoodofraintoday.Simi-larly,Imightattacha0.1degreeofbelieftothetheorythatthestrongnuclearand290\nPHILOSOPHYOFSCIENCEelectro-weakforcesarethesameforce–Ithinkitunlikely,butallowthatthereisaone-in-tenpossibilityitmayturnouttrue.Astheseexamplesindicate,BayesiansthinkofdegreesofbeliefastheextenttowhichyousubjectivelytakesomethingtobePROBABLE(pp.167–8).Accordingly,theyarguethatyourdegreesofbeliefoughttosatisfytheaxiomsoftheprobabilitycalcu-lus.(SeetheboxbelowfortheDutchBookArgumentforthisthesis.)Itisimportanttorealize,however,thatwhileBayesiansthinkofdegreesofbeliefasprobabilitiesinthismathematicalsense,theystillthinkofthemassubjectiveprobabilities.Inparticular,theyallowthatitcanbeperfectlyrationalfordifferentpeopletoattachdifferentsub-jectiveprobabilitiestothesameproposition–youcanbelievethatitwillraintodaytodegree0.2,whileIbelievethistodegree0.5.Whatrationalitydoesrequire,accordingtotheBayesians,isonlythatifyouhaveasubjectiveprobabilityof0.2forrain,thenyoumusthaveoneof0.8foritsnotraining,whileifIhave0.5forrain,thenImusthave0.5foritsnotraining.Thatis,bothofusmustaccord,inourdifferentways,withthetheoremoftheprobabilitycalculusthatProb(p)=1-Prob(not-p).Atfirstsight,thiselementofsubjectivitymightseemtodisqualifyBayesianismasapossiblebasisforscientificrationality.Ifweareallfreetoattachwhateverdegreesofbeliefweliketoscientifictheories,providedonlythatwearefaithfultothestructureoftheprobabilitycalculus,thenwhatistostopeachofusfromsupportingdifferenttheo-ries,dependingonlyonindividualfadsorprejudices?ButBayesianshaveananswer;namely,thatitdoesnotmatterwhatprejudicesyoustartwith,aslongasyoureviseyourdegreesofbeliefinarationalway.Bayesiansderivetheiraccountofhowtorevisedegreesofbelief,aswellastheirname,fromBayes’stheorem,originallyprovedbyThomasBayesinapaperpublishedin1763.Bayes’stheoremstates:Prob(H/E)=Prob(H)¥Prob(E/H)/Prob(E).Thesimpleproofofthistheoremisgiveninthebox.Butthephilosophicalsignificanceofthetheoremisthatitsuggestsacertainprocedureforrevisingyourdegreesofbeliefinresponsetonewevidence.SupposethatHissomehypothesis,andEissomenewlydiscoveredevidence.ThenBayesiansarguethat,whenyoudiscoverE,youshouldadjustyourdegreeofbeliefinHinlinewiththeright-handsideoftheaboveequation:thatis,youshouldincreaseittotheextentthatyouthinkEislikelygivenH,butunlikelyotherwise.Inotherwords,ifEisinitselfverysurprising(likelightbendinginthevicin-ityofthesun)butatthesametimejustwhatyouwouldexpectgivenyourtheoryH(thegeneraltheoryofrelativity),thenEshouldmakeyouincreaseyourdegreeofbeliefinHagreatdeal.Ontheotherhand,ifEisnomorelikelygivenHthanitwouldbeonanyothertheory,thenobservingEprovidesnoextrasupportforH.Themovementofthetides,forexample,isnogreatargumentforgeneralrelativity,eventhoughitispredictedbyit,sinceitisalsopredictedbythealternativeNewtoniantheoryofgravitation.Noteinparticularthatthisstrategyforupdatingdegreesofbeliefinresponsetoevi-dencecanbeappliedtoinductiveinferences.ConsiderthespecialcasewhereHissomeuniversalgeneralization–allbodiesfallwithconstantacceleration,say–andtheevi-denceEisthatsomeparticularfallingbodyhasbeenobservedtoaccelerateconstantly.291\nDAVIDPAPINEAUIfthisobservationwassomethingyoudidnotexpectatall,thenBayesianismtellsyouthatyoushouldincreaseyourdegreeofbeliefinGalileo’slawsignificantly,foritisjustwhatGalileo’slawpredicts.Ofcourse,onceyouhaveseenanumberofsuchobserva-tions,andbecomereasonablyconvincedofGalileo’slaw,thenyouwillceasetofindnewinstancessurprising,andtothatextentwillceasetoincreaseyourdegreeofbeliefinthelaw.Butthatisasitshouldbe.Onceyouarereasonablyconvincedofalaw,thenthereisindeedlittlepointingatheringfurthersupportinginstances,andsoitistothecreditofBayesianismthatitexplainsthis.TheBayesianaccountofhowtorevisedegreesofbeliefseemstomakegoodsense.Inaddition,itpromisesasolutiontotheproblemofinduction,sinceitimpliesthatpositiveinstancesgiveusreasontobelievescientificgeneralizations.Thereare,however,problemsfacingthisaccount.Forastart,anumberofphiloso-phershavequeriedwhetherBayes’stheorem,whichafterallislittlemorethananarith-meticaltruth,canconstrainwhatdegreesofbeliefweadoptinthefuture(seetheboxbelow).Andevenifweputthisrelativelytechnicalissuetooneside,itisunclearhowfartheBayesianaccountreallyanswerstheworryraisedabove,thatthesubjectivityofdegreesofbeliefwillallowdifferentscientiststocommitthemselvesarbitrarilytodif-ferenttheories.TheBayesiananswertothisworrywasthatBayes’stheoremwillatleastconstrainthesedifferentscientiststorevisetheirdegreesofbeliefinresponsetotheevi-denceinsimilarways.But,evenso,itstillseemspossiblethatthescientistswillremainondifferenttracks,iftheystartatdifferentplaces.IftwoscientistsarefreetoattachdifferentpriordegreesofbeliefinGalileo’slaw,andbothupdatethosedegreesofbeliefaccordingtoBayes’stheoremwhentheylearntheevidence,willtheynotstillendupwithdifferentposteriordegreesofbeliefs?Thestandardanswertotheobjectionistoappealtoconvergenceofopinion.Theideaisthat,givenenoughevidence,everybodywilleventuallyendupinthesameplace,eveniftheyhavedifferentstarting-points.Thereareanumberoftheoremsofprobabilitytheoryshowingthat,withinlimits,differencesininitialprobabilitieswillbe‘washedout’,inthesensethatsufficientevidenceandBayesianupdatingwillleadtoeffectivelyidenticalfinaldegreesofbelief.Sointheend,argueBayesians,itdoesnotmatterifyoustartwithahighorlowdegreeofbeliefinGalileo’slaw–forafter1,000observationsofconstantlyfallingbodiesyouwillendupbelievingittoadegreecloseto1anyway.However,interestingastheseresultsare,theydonotsatisfactorilyanswerthefun-damentalphilosophicalquestionsaboutinductivereasoning.Fortheydonotworkforallpossibleinitialdegreesofbelief.Rather,theyassumethatthescientistsatissue,whiledifferingamongthemselves,alldrawtheirinitialdegreesofbelieffromacertainrange.Whilethisrangeincludesalltheinitialdegreesofbeliefthatseematallintuitivelyplau-sible,thereareneverthelessotherpossibleinitialdegreesofbeliefthatareconsistentwiththeaxiomsofprobability,butwhichwillnotleadtoeventualconvergence.So,forexample,theBayesiansdonotinfactexplainwhatiswrongwithpeoplewhoneverendupbelievingGalileo’slawbecausetheyarealwaysconvincedthatthecourseofnatureisgoingtochangetomorrow.Ofcourse,Bayesiansarerighttoregardsuchpeopleasirrational.Buttheydonotexplainwhytheyareirrational.Sotheyfailtoshowwhyallthinkersmustendupwiththesameattitudetoscientifictheories.Andinparticulartheyfailtosolvetheproblemofinduction,sincetheydonotshowwhyallrationalthinkersmustexpectthefuturetobelikethepast.292\nPHILOSOPHYOFSCIENCEBayesianismTheDutchBookArgumentTheaxiomsofprobabilityrequirethat(1)0£Prob(P)£1,foranypropositionP(2)Prob(P)=1,ifPisanecessarytruth(3)Prob(P)=0,ifPisimpossible(4)Prob(PorQ)=Prob(P)+Prob(Q),ifPandQaremutuallyexclusive.BayesiansappealtotheDutchBookArgumenttoshowwhysubjectivedegreesofbeliefshouldconformtotheseaxioms.Imaginethatyourdegreesofbeliefdidnotsoconform.YoubelievepropositionPtodegreey,say,andyetdonotbelievenot-Ptodegree1-y.(Youthusviolatetheconjunctionofaxioms(2)and(4),becausePornot-Pisaneces-sarytruth.)ThenitwillbepossibleforsomebodytoinduceyoutomakebetsonPandnot-Pinsuchawaythatyouwilllosewhateverhappens.Asetofbetsthatguaranteethatyouwilllosewhateverhappensiscalleda‘Dutchbook’.Theundesirabilityofsuchasetofbetsthusprovidesanargumentthatanyrationalperson’ssubjectivedegreesofbeliefshouldsatisfytheaxiomsoftheprobabilitycalculus.Bayes’sTheoremTheconditionalprobabilityofPgivenQ-Prob(P/Q)–isdefinedasProb(PandQ)/Prob(Q).Intuitively,Prob(P/Q)signifiestheprobabilityofPontheassumptionthatQistrue.ItimmediatelyfollowsfromthisdefinitionthatProb(H/E)=Prob(H)¥Prob(E/H)/Prob(E)ThisisBayes’stheorem.Asyoucansee,itsaysthattheconditionalprobabilityProb(H/E)ofsomehypothesisHgivenevidenceEisgreaterthanProb(H)totheextentthatEisimprobableinitself,butprobablegivenH.BayesianUpdatingBayesiansrecommendthatifyouobservesomeevidenceE,thenyoushouldreviseyourdegreeofbeliefinH,andsetyournewProbt’(H)equaltoyourpreviousconditionaldegreeofbeliefinHgivenE,Probt(H/E),wheretisthetimebeforeyoulearnE,andt¢after.Bayes’stheorem,appliedtoyoursubjectiveprobabilitiesatt,thenindicatesthatthiswillincreaseyourdegreeofbeliefinHtotheextentthatyoupreviouslythoughtEtobesubjectivelyimprobableinitself,butsubjectivelyprobablegivenH.ThisBayesianrecommendation,thatyoureviseyourdegreeofbeliefinHbysettingitequaltoyouroldconditionaldegreeofbeliefinHgivenE,shouldbedistinguishedfromBayes’stheorem.Bayes’stheoremisatrivialconsequenceofthedefinitionofconditionalprobability,andconstrainsyourdegreesofbeliefatagiventime.TheBayesianrecom-mendation,bycontrast,specifieshowyourdegreesofbeliefshouldchangeovertime.Bayes’stheoremisuncontentious,butitisamatterofactivecontroversywhetherthereisanysatisfactorywayofdefendingtheBayesianrecommendation(Hacking1967;Teller1973).293\nDAVIDPAPINEAU1.5InstrumentalismversusrealismAtthisstageletusleavetheproblemofinductionforawhileandturntoadifferentdifficultyfacingscientificknowledge.Muchofscienceconsistsofclaimsaboutunob-servableentitieslikeviruses,radiowaves,electronsandquarks.Butiftheseentitiesareunobservable,howarescientistssupposedtohavefoundoutaboutthem?Iftheycannotseeortouchthem,doesitnotfollowthattheirclaimsaboutthemareatbestspeculativeguesses,ratherthanfirmknowledge?Itisworthdistinguishingthisproblemofunobservabilityfromtheproblemofinduc-tion.Bothproblemscanbeviewedasdifficultiesfacingtheoreticalknowledgeinscience.Butwhereastheproblemofinductionarisesbecausescientifictheoriesmakegeneralclaims,theproblemofunobservabilityisduetoourlackofsensoryaccesstothesubjectmatterofmanyscientifictheories.(Sotheproblemofinductionarisesforgeneralclaimseveniftheyarenotaboutunobservables,suchas‘Allsodiumburnsbrightorange’.Conversely,theproblemofunobservabilityarisesforclaimsaboutunobserv-ableseveniftheyarenotgeneral,suchas‘Onefreeelectronisattachedtothisoildrop’.Inthissectionandthenext,however,itwillbeconvenienttousetheterm‘theory’specificallyforclaimsaboutunobservables,ratherthanforgeneralclaimsofanykind.)Therearetwogenerallinesofresponsetotheproblemofunobservability.Ontheonehandarerealists,whothinkthattheproblemcanbesolved.Realistsarguethattheobservablefactsprovidegoodindirectevidencefortheexistenceofunobservableen-tities,andsoconcludethatscientifictheoriescanberegardedasaccuratedescriptionsoftheunobservableworld.Ontheotherhandareinstrumentalists,whoholdthatweareinnopositiontomakefirmjudgementsaboutimperceptiblemechanisms.Instru-mentalistsallowthattheoriesaboutsuchmechanismsmaybeuseful‘instruments’forsimplifyingourcalculationsandgeneratingpredictions.Buttheyarguethatthesethe-oriesarenomoretruedescriptionsoftheworldthanthe‘theory’thatallthematterinastoneisconcentratedatitscentreofmass(whichisalsoanextremelyusefulassump-tionfordoingcertaincalculations,butclearlyfalse).Earlierthiscenturyinstrumentalistsusedtoarguethatweshouldnoteveninterprettheoreticalclaimsliterally,onthegroundsthatwecannotsomuchasmeaningfullytalkaboutentitieswehaveneverdirectlyexperienced.Butnowadaysthiskindofsemanticinstrumentalismisoutoffavour.Contemporaryinstrumentalistsallowthatscientistscanmeaningfullypostulate,say,thatmatterismadeoftinyatomscontainingnucleiorbitedbyelectrons.ButtheythentakeaSCEPTICAL(pp.45–56)attitudetosuchpostulates,sayingthatwehavenoentitlementtobelievethem(asopposedtousingthemasaninstrumentforcalculations).Aninitiallineofargumentopentorealismistoidentifysomefeatureofscientificpracticeandthenarguethatinstrumentalismisunabletoaccountforit.Oneaspectofscientificpracticeinvokedinthisconnectionhasbeentheunificationofdifferentkindsoftheoriesinpursuitofasingle‘theoryofeverything’(Friedman1984);otherfeaturesofscienceappealedtobyrealistshaveincludedtheuseoftheoriestoexplainobservablephenomena(Boyd1980),andtherelianceontheoriestomakenovelpredictions(Smart1963).For,sotherealistargues,theseaspectsofscientificpracticeonlymakesenseontheassumptionthatscientifictheoriesaretruedescriptionsofreality.Afterall,saystherealist,iftheoriesaresimplyconvenientcalculatingdevices,thenwhy294\nPHILOSOPHYOFSCIENCEexpectdifferenttheoriestobeunifiableintooneconsistentstory?Unificationisclearlydesirableifourtheoriesallaimtocontributetotheoveralltruth,butthereseemsnoparallelreasonwhyabunchofinstrumentsshouldbeunifiableintoonebig‘instru-mentofeverything’.Andsimilarly,therealistwillargue,thereseemsnoreasontoexpectamerecalculatinginstrument,asopposedtoatruedescriptionofanunderly-ingreality,toyieldagenuineexplanationofsomepastoccurrence,orareliablepre-dictionofafutureone.However,thisformofargumenttendstobeinconclusive.Therearetwopossiblelinesofresponseopentoinstrumentalists.Theycanofferaninstrumentalistaccountoftherelevantfeatureofscientificpractice.Alternatively,theycandenythatthisfeaturereallyispartofscientificpracticeinthefirstplace.Asanexampleofthefirstresponse,theycouldarguethattheunificationofscienceismotivated,notbythepursuitofoneunderlyingtruth,butsimplybythedesirabilityofhavingasingleall-purposecalculat-inginstrumentratherthanarag-bagofdifferentinstrumentsfordifferentproblems.Thesecondkindofresponsewouldbetodenythatunificationisessentialtosciencetostartwith.ThusNancyCartwrightarguesthatsciencereallyisarag-bagofdifferentinstruments.Shemaintainsthatscientistsfacedwithagivenkindofproblemwillstan-dardlydeploysimplifyingtechniquesandrulesofthumbwhichowenothingtogeneraltheory,butwhichhaveshownthemselvestodelivertherightanswertothekindofproblemathand(Cartwright1983).Similarresponsescanbemadebyinstrumentaliststotheargumentsfromexplana-tionandprediction.Instrumentalistscaneitherretortthatthereisnoreasonwhythestatusoftheoriesascalculatinginstrumentsshouldprecludethemfromgivingrisetopredictionsandexplanations;ortheycanquerywhetherscientifictheoriesreallydoaddtoourabilitytopredictandexplaintostartwith.Notalltheselinesofresponseareequallyconvincing.Butbetweenthemtheygiveinstrumentalismplentyofroomtocountertheinitialrealistchallenge.1.6Theory,observationandincommensurabilityAdifferentlineofargumentagainstinstrumentalismfocusesonthedistinctionbetweenwhatisobservableandwhatisnot.Thisdistinctioniscrucialtoinstrumen-talism,inthatinstrumentalistsarguethatclaimsaboutobservablephenomenaareunproblematic,butclaimsaboutunobservablesarenot.However,anumberofwritershavequeriedthisdistinction,arguingthatobservationreportsarenotessentiallydif-ferentfromclaimsaboutunobservables,sincetheytoodependontheoreticalassump-tionsabouttheunderlyingstructureofreality.NorwoodHanson(1958)hasargued,forexample,thatscientistsbeforeandafterCopernicussawdifferentthingswhentheylookedattheSun:whereaspre-CopernicansregardedtheEarthasstationaryandsosawtheSunrevolvingroundit,post-CopernicanscientistssawtheSunasstationaryandtheEarthasrotating.Similarly,Hanson(1963)arguesthatthephotographicplatewhichlookslikeasquigglymesstoalayobserverisseenasdisplayingawell-definedelectron–positronpairbyanexperiencedparticlephysicist.Examplesliketheseunder-minethedistinctionbetweenwhatisobservableandwhatisnot,sincetheyshowthatevenjudgementsmadeinimmediateresponsetosensorystimulationareinfluencedbyfallibletheoriesaboutreality.295\nDAVIDPAPINEAUFigure9.1TheMüller–Lyer.Althoughallthreelinesarethesamelength,thetopline,withinward-pointingarrowheads,appearstobeshorter,andthebottomline,withoutward-pointingarrowheads,appearstobelonger,thanthe‘neutral’middleline.Noristhepointrestrictedtorecherchéobservationsofastronomicalbodiesorsub-atomicparticles.Evenimmediateperceptualjudgementsaboutthecolour,shapeandsizeofmedium-sizedphysicalobjectscanbeshowntodependontheoreticalassump-tionsimplicitinourvisualsystems.Perhapsthebest-knownillustrationistheMüller–Lyerillusion(seefigure9.1),whichshowshowourvisualsystemusescomplexassumptionsaboutthenormalcausesofcertainkindsofretinalpatternstodrawcon-clusionsaboutthegeometryofphysicalfigures.Andanalogousillusionscanbeusedtodemonstratethepresenceofothertheoreticalpresuppositionsinourvisualandothersensorysystems.AsIsaid,inthefirstinstancetheunclarityoftheobservable–unobservabledistinc-tioncountsagainstinstrumentalismratherthanrealism.Afterall,itisinstrumental-ism,notrealism,whichneedsthedistinction,sinceinstrumentalismsaysthatweshouldbescepticalaboutunobservableclaims,butnotobservableones,whereasrealismishappytoregardbothkindsofclaimsasbelief-worthy,sodoesnotmindiftheycannotbesharplydistinguished.However,thereisanotherwayofrespondingtodoubtsaboutthetheory–observa-tiondistinction.Fornotethattheargumentsagainsttheobservable–unobservabledistinctiondonotinfactvindicatetherealistbelief-worthinessofclaimsaboutunob-servables;rather,theyattackrealismfromthebottomup,andunderminethebelief-worthinessofclaimsaboutobservables,byshowingthatevenobservationalclaimsdependonfallibletheoreticalassumptions.Obviously,ifthereisnoobservable–unob-servabledistinction,thenallscientificclaimsareinthesameboat.Butonreflectionitseemsthattheboattheyallendupinistheinstrumentalistboatofscepticaldisbelief,nottherealistoneofgeneralfaithinscience.Anumberofinfluentialrecentphilosophersofscience,mostprominentlyT.S.Kuhn(1962)andPaulFeyerabend(1976),haveembracedthisconclusionwholeheartedly,andmaintainedthatnojudgementsmadewithinscience,notevenobservationaljudge-ments,canclaimtheauthorityofestablishedtruth.Rather,theyargue,oncescientistshaveembracedatheoryabouttheessentialnatureoftheirsubjectmatter,suchasgeo-centrism,orNewtoniandynamics,orthewavetheoryoflight,theywillinterpretallobservationaljudgementsinthelightofthattheory,andsowillneverbeforcedto296\nPHILOSOPHYOFSCIENCErecognizethekindofnegativeobservationalevidencethatmightshowthemthattheirtheoryismistaken.KuhnandFeyerabendindependentlylitonthetermincommensu-rabletoexpresstheviewthatthereisnocommonyardstick,intheformoftheory-independentobservationjudgements,whichcanbeusedtodecideobjectivelyontheworthofscientifictheories.Instead,theyargue,decisionsonscientifictheoriesareneverduetoobjectiveobservationalevidence,butarealwaysrelativetothepresuppo-sitions,interestsandsocialmilieuxofthescientistsinvolved.Kuhn’sandFeyerabend’sblanketrelativismhasprovokedmuchdiscussionamongphilosophersofscience,butwonfewwhole-heartedconverts.Muchofthediscussionhasfocusedonthestatusofobservations.Mostphilosophersofsciencearepreparedtoacceptthatallobservationaljudgementsinsomesensepresupposesomeelementoftheory.Butmanybalkattheconclusionthatobservationsthereforeneverhaveanyindependentauthoritytodecidescientificquestions.Afterall,theypointout,mostsimpleobservations,suchasthatapointerisadjacenttoamarkonadial,presupposeatmostaminimalamountoftheory,aboutrigidbodies,say,andaboutbasiclocalgeometry.Sincesuchminimaltheoriesarethemselvesrarelyatissueinseriousscien-tificdebates,thisminimalamountoftheory-dependenceprovidesnoreasonwhyobser-vationsofpointerreadingsshouldnotbeusedtosettlescientificdisputes.Ifascientifictheoryaboutthebehaviourofgases,say,predictsthatapointerwillbeatacertainplaceonadial,anditisobservednottobe,thenthisdecidesagainstthetheoryaboutgases.Itisnottothepointtorespondthat,intakingthepointerreadingatfacevalue,wearemakingassumptionsaboutrigidbodiesandlocalgeometry.Fornothinginthedebateaboutgasesprovidesanyreasontodoubttheseassumptions.Andthisofcourseiswhyscientiststakesuchpainstoworkoutwhattheirtheoriesimplyaboutthingslikepointerreadings–sinceobservationsofpointerreadingsdonotdependonanythingcon-tentious,theywillweighwithallsidesinthescientificdebate.So,despitetheargumentsofKuhnandFeyerabend,nearlyallphilosophersarereal-istsaboutpointerreadingsandsimilarobservablephenomena.Butthisstillleavesuswiththeoriginaldisagreementbetweenrealismandinstrumentalismaboutlessdirectlyobservableentities.For,evenifclaimsaboutpointerreadingsareuncontroversiallybelief-worthy,instrumentalistscanstillarguethattheoriesaboutviruses,atomsandgravitationalwavesarenothingmorethanusefulfictionsformakingcalculations.Therealistresponse,asIsaid,isthattheobservablefactsprovidegoodindirectevi-denceforthesetheoreticalentities,evenifwecannotobservethemdirectly.However,therearetwostronglinesofargumentthatinstrumentalistscanusetocastdoubtonthissuggestion.InthenexttwosectionsIshalldiscuss‘theunderdeterminationoftheorybyevidence’and‘thepessimisticmeta-inductionfrompastfalsity’.1.7TheunderdeterminationoftheorybyevidenceTheargumentfromunderdeterminationassertsthat,givenanytheoryaboutunob-servablesthatfitstheobservablefacts,therewillbeotherincompatibletheoriesthatfitthesamefacts.Andso,theargumentconcludes,weareneverinapositiontoknowthatanyoneofthesetheoriesisthetruth.Whyshouldweacceptthatthereisalwaysmorethanonetheorythatfitsanysetofobservablefacts?Onepopularargumentforthisconclusionstemsfromthe297\nDAVIDPAPINEAU‘Duhem–Quinethesis’.Accordingtothisthesis,anyparticularscientifictheorycanalwaysbedefendedinthefaceofcontraryobservationsbyadjustingauxiliaryhypoth-eses.Forexample,whentheNewtoniantheoryofgravitationwasthreatenedbyobser-vationsofanomalousmovementsbytheplanetMercury,itcouldalwaysbedefendedbypostulatingahithertounobservedplanet,say,oraninhomogeneousmassdistribu-tionintheSun.Thisgeneralstrategyfordefendingtheoriesagainstcontraryevidenceseemstoimplythattheadherentsofcompetingtheorieswillalwaysbeabletomain-taintheirrespectivepositionsinthefaceofanyactualobservationaldata.Anotherargumentforunderdeterminationstarts,notwithcompetingtheories,butwithsomegiventheory.Supposethatallthepredictionsofsomeparticulartheoryareaccurate.Wecanconstructa‘de-Ockhamized’versionofthistheory(reversingWilliamofOckham’s‘razor’whichprescribesthat‘entitiesarenottobemultipliedbeyondnecessity’),bypostulatingsomeunnecessarilycomplicatedunobservablemechanismwhichneverthelessyieldsanewtheorywithpreciselythesameobservationalconse-quencesastheoriginalone.Bothoftheselinesofreasoningcanbeusedtoarguethatmorethanonetheoryaboutunobservableswillalwaysfitanygivensetofobservationaldata.Doesthismakerealismaboutunobservablesuntenable?Manyphilosophersconcludethatitdoes.Butthisistooquick.Forweshouldrecognizethatthereisnothingintheargumentsforalternativeunderdeterminedtheoriestoshowthatthesealternativetheorieswillalwaysbeequallywell-supportedbythedata.Whattheargumentsshowisthatdifferenttheorieswillalwaysbeconsistentwiththedata.Buttheydonotruleoutthepossibilitythat,amongthesealternativetheories,oneisvastlymoreplausiblethantheothers,andforthatreasonshouldbebelievedtobetrue.Afterall,‘flatearthers’canmaketheirviewconsistentwiththeevidencefromgeography,astronomyandsatellitepho-tographs,byconstructingfar-fetchedstoriesaboutconspiraciestohidethetruth,theeffectsofemptyspaceoncameras,andsoon.Butthisdoesnotshowthatweneedtaketheirflat-earthismseriously.Similarly,eventhoughNewtoniangravitationaltheorycaninprinciplebemadeconsistentwithallthecontraryevidence,thisisnoreasonnottobelievegeneralrelativity.Norisourabilityto‘cookup’ade-Ockhamizedversionofgeneralrelativityareasontostopbelievingthestandardversionunencumberedwithunnecessaryentities.Nevertheless,asIsaid,manycontemporaryphilosophersofsciencedomovedirectlyfromthepremisethatdifferenttheoriesareconsistentwiththeobservationalevidencetotheconclusionthatnoneofthemcanberegardedasthetruth.ThisisbecausemanyofthemaddressthisissuefromanessentiallyPopperianperspective.ForifyoufollowPopperinrejectinginduction,thenyouwillnotbelievethatevidenceeverprovidesposi-tivesupportforanytheory,exceptintheback-handedsensethattheevidencecanfailtofalsifyit.Accordinglyyouwillthinkthatalltheoriesthathavenotbeenfalsifiedareonapar,andinparticularthatanytwotheoriesthatarebothconsistentwiththeevidenceareequallywell-supportedbyit.SotheargumentsforunderdeterminationdopresentaproblemtoPopperians,sincePopperianshavenoobviousbasisfordiscriminatingamongdifferenttheoriesconsis-tentwiththedata.But,asIpointedoutabove,theseneednotworrythoseofuswhodivergefromPopperinthinkingunfalsifiedtheoriescanbebetterorworsesupportedbyevidence,forwecansimplyrespondtotheunderdeterminationargumentsby298\nPHILOSOPHYOFSCIENCEobservingthatsomeunderdeterminedtheoriesarebettersupportedbytheevidencethanothers.NowthatwehavereturnedtoPopper,itisworthnotingthattheDuhem–QuineargumentalsoraisesamorespecificproblemforPopperians.RecallthatPopper’soverallphilosophyraisedthe‘problemofdemarcation’,theproblemofhowtodistin-guishsciencefromotherkindsofconjecture.Popper’sanswerwasthatscience,unlikeastrology,orMarxismandpsychoanalytictheory,isfalsifiable.ButtheDuhem–QuineargumentshowsthatevensucheminentlyscientifictheoriesasNewtonianphysicsarenotfalsifiableinanystraightforwardsense,sincetheycanalwayssavethemselvesinthefaceoffailedpredictionsbyadjustingauxiliaryhypotheses.NotonlydoesthiscastdoubtonPopper’sdismissalofMarxismandpsychoanalysisasunscientific,butitseemstounderminehiswholesolutiontothedemarcationproblem.IfsuchparadigmaticscientifictheoriesasNewtonianphysicsarenotfalsifi-able,thenitcanscarcelybefalsifiabilitythatdistinguishessciencefromnon-science.Still,thisisPopper’sproblem,notours(seeHarding1975).Ifwedonotrejectinduc-tion,thenwedonothaveaproblemofdemarcation.Forwecansimplysaythatwhatdistinguishessuccessfulscientifictheoriesfromnon-scienceisthattheobservationalevidencegivesusinductivereasontoregardscientifictheoriesastrue.Theargumentsinthelatterpartofthissectionhavepresupposedthatacertainkindofinductiveargumentislegitimate.Thekindofinductiveargumentrelevanttounder-determinationisnotsimple‘enumerative’induction,fromobservedAsbeingBsto‘AllAsareBs’,butratherinferencesfromanycollectionofobservationaldatatothemostplausibletheoryaboutunobservablesthatisconsistentwiththatdata.Butthesearespeciesofthesamegenus;indeed,enumerativeinductionscanthemselvesbeinter-pretedastreating‘AllAsareBs’asthemostplausibleextrapolationconsistentwiththeobservedAsbeingBs.Myattitudetothismoregeneralcategoryofinductiveinferencesremainsthesameasmyattitudetoenumerativeinduction,whichIoutlinedearlier.Wedonotyethaveanexplanationofwhyinductiveinferencesarelegitimate,andtothatextentwestillfaceaproblemofinduction.Butitissillytotrytosolvethatproblembydenyingthatinductiveinferencesareeverlegitimate.Andthatiswhytheunderdeter-minationoftheorybydatadoesnotconstituteagoodargumentforinstrumentalism.Fortoassumethatweareneverentitledtobelieveatheory,ifthereareothersconsistentwiththesamedata,issimplytoassumetheillegitimacyofinduction.TheUnderdeterminationofTheorybyObservationalData(UTD)TherearetwoargumentsfortheUTD.ThefirstisbasedontheDuhem–Quinethesis,orig-inallyformulatedbytheFrenchphilosopherandhistorianPierreDuhem(1861–1916)andlaterrevivedbytheAmericanlogicianW.V.O.Quine(b.1908).Duhem(1951)andQuine(1951)pointoutthatascientifictheoryTdoesnotnormallyimplypredictionsPonitsown,butonlyinconjunctionwithauxiliaryhypothesesH.T&HfiPSowhenPisfalsifiedbyobservation,thisdoesnotrefuteT,butonlytheconjunctionofT&H.299\nDAVIDPAPINEAUnot-Pfinot-(T&H)SoTcanberetained,andindeedstillexplainP,providedwereplaceHbysomealterna-tive,H’,suchthatT&H’finot-P.ThisyieldstheDuhem–Quinethesis:anytheoreticalclaimTcanconsistentlyberetainedinthefaceofcontraryevidence,bymakingadjustmentselsewhereinoursystemofbeliefs.TheUTDfollowsquickly.ImaginetwocompetingtheoriesT1andT2.Whateverevidenceaccumulates,versionsofT1andT2,conjoinedwithgreatlyrevisedauxiliaryhypothesesifnecessary,willbothsurvive,consistentwiththatevidence,butincompat-iblewitheachother.Theotherargument,firstputforwardbyphysicistslikeHenriPoincaré(1854–1912)andErnstMach(1838–1916)attheturnofthetwentiethcentury,hasadifferentstarting-point.ImaginethatT1isthecompletetruthaboutphysicalreality,andthatitimpliesobservationalfactsO.Thenwecanalwaysconstructsome‘de-Ockhamized’T2whichpostulatesmorecomplicatedunobservablemechanismsbutmakesjustthesameobservationalpredictionsO.(Glymour1980:ch.5.)Forexample,supposewestartwithstandardassumptionsaboutthelocationofbodiesinspace-timeandabouttheforcesactingonthem.Ade-Ockhamizedtheorymightthenpostulatethatallbodies,includingallmeasuringinstruments,areacceleratingby21ft/sec.inagivendirection,andthenaddjusttheextraforcesrequiredtoexplainthis.Thistheorywouldclearlyhaveexactlythesameobservationalconsequencesastheoriginalone,eventhoughitcontradicteditattheunobservablelevel.TobringoutthedifferencebetweenthetwoargumentsforUTD,notethattheDuhem–Quineargumentdoesnotspecifyexactlywhichoveralltheorieswewillendupwith,sinceitleavesopenhowT1’sandT2’sauxiliaryhypothesesmayneedtoberevised;thede-Ockhamizationargument,bycontrast,actuallyspecifiesT1andT2infulldetail,includingauxiliaryhypotheses.Incompensation,theDuhem–Quineargumentpromisesusalternativetheorieswhateverobservationalevidencemayturnupinthefuture;whereasthede-OckhamizationargumentassumesthatallfutureobservationsareasT1predicts.1.8Thepessimisticmeta-inductionIturnnowtotheotherargumentagainstrealism.Thisargumenttakesasitspremisethefactthatpastscientifictheorieshavegenerallyturnedouttobefalse,andthenmovesinductivelytothepessimisticconclusionthatourcurrenttheoriesarenodoubtfalsetoo.(Thisiscalleda‘meta-induction’becauseitssubjectmatterisnotthenaturalworld,butscientifictheoriesaboutthenaturalworld.)Thereareplentyoffamiliarexamplestosupportthisargument.Newton’stheoryofspaceandtime,thephlogistontheoryofcombustion,andthetheorythatatomsareindivisiblewereallatonetimewidelyacceptedscientifictheories,buthavesincebeenrecognizedtobefalse.Sodoesitnotseemlikely,thepessimisticinductionconcludes,thatallourcurrenttheoriesarefalse,andthatweshouldthereforetakeaninstru-mentalistratherthanarealistattitudetothem?(SeeLaudan1981.)300\nPHILOSOPHYOFSCIENCEThisisanimportantandpowerfulargument,butitwouldbetooquicktoconcludethatitdiscreditsrealismcompletely.Itisimportantthatthetendencytofalsityismuchmorecommoninsomeareasofsciencethanothers.Thusitisrelativelynormalfortheoriestobeoverturnedincosmology,say,orfundamentalparticlephysics,orthestudyofprimateevolution.Bycontrast,theoriesofthemolecularcompositionofdif-ferentchemicalcompounds(suchasthatwaterismadeofhydrogenandoxygen),orthecausesofinfectiousdiseases(chickenpoxisduetoaherpesvirus),orthenatureofeverydayphysicalphenomena(heatismolecularmotion),arecharacteristicallyretainedoncetheyareaccepted.Norneedweregardthisdifferentialsuccessrateofdifferentkindsoftheoriesassomekindofaccident.Rather,itistheresultofthenecessaryevidencebeingmoreeasilyavailableinsomeareasofsciencethanothers.Paleoanthropologistswanttoknowhowmanyhominidspecieswerepresentonearth3millionyearsago.Buttheirevidenceconsistsofafewpiecesofteethandbone.Soitisscarcelysurprisingthatdiscoveriesofnewfossilsiteswilloftenleadthemtochangetheirviews.Thesamepointappliesonalargerscaleincosmologyandparticlephysics.Scientistsintheseareaswanttoanswerverygeneralquestionsabouttheverysmallandtheverydistant.Buttheirevidencederivesfromthelimitedrangeoftechnologicalinstrumentstheyhavedevisedtoprobetheserealms.So,oncemore,itisscarcelysurprisingthattheirtheoriesshouldremainattheleveloftentativehypotheses.Bycontrast,inthoseareaswhereadequateevidenceisavailable,suchaschemistryandmedicine,thereisnocorrespondingbarriertosciencemovingbeyondtentativehypothesestofirmconclusions.Themoralisthatrealismismoredefensibleforsomeareasofsciencethanothers.Insomescientificsubjectsfirmevidenceisavailable,andentitlesustoviewcertaintheories,likethetheorythatwateriscomposedofH2Omolecules,astheliteraltruthaboutreality.Inotherareastheevidenceisfragmentaryandinconclusive,andthenwedobettertoregardthebest-supportedtheories,suchasthetheorythatquarksandleptonsaretheultimatebuildingblocksofmatter,asusefulinstrumentswhichaccom-modatetheexistingdata,makeinterestingpredictions,andsuggestfurtherlinesforresearch.Atfirstsightthismightlooklikeavictoryforinstrumentalismoverrealism.Fordidnotinstrumentalistsalwaysacceptthatweshouldberealistsaboutobservablethings,andonlyurgeinstrumentalismforuncertaintheoriesaboutunobservablephenomena?Butourcurrentpositiondrawsthelineinadifferentplace.Instrumen-talism,asoriginallydefined,takesitforgrantedthateverythingunobservableisinaccessible,andthatalltheoriesaboutunobservablesarethereforeuncertain.Bycontrast,thepositionwehavearrivedatplacesnospecialweightonthedistinctionbetweenwhatisobservableandwhatisnot.Inparticular,itarguesthatthepessimisticmeta-inductionfailstoshowthatfalsityisthenaturalfateofalltheoriesaboutunobservables,butonlythatthereisalinewithinthecategoryoftheoriesaboutunobservables,betweenthosetheoriesthatcanbeexpectedtoturnoutfalseandthosewhoseclaimstotrutharesecure.Soourcurrentpositionisnotadogmaticinstru-mentalismaboutallunobservables,butmerelytheuncontentiousviewthatweshouldbeinstrumentalistsaboutthatsub-classoftheorieswhicharenotsupportedbyadequateevidence.301\nDAVIDPAPINEAU1.9NaturalizedepistemologyofscienceInthelastdecadeorsoanumberofphilosophersofsciencehaveturnedtoanatural-izedapproachtoscientificknowledge(Kitcher1992).Inplaceoftraditionalattemptstoestablishcriteriaforscientifictheory-choicebyaprioriphilosophicalinvestigation,thenaturalizedapproachregardsscienceitselfasasubjectforaposterioriempiricalinvestigation.Accordingly,naturalizedepistemologistslooktothehistory,sociologyandpsychologyofscience,ratherthantofirstprinciples,toidentifycriteriafortheacceptabilityofscientifictheories.Oneapparentdifficultyfacingthiskindofnaturalizedepistemologyofscienceisthatitisunclearhowempiricalinvestigationcaneveryieldanythingmorethandescriptiveinformationabouthowscientistsactuallyoperate.Yetanyepistemologyofscienceworthitsnameoughtalsotohaveanormativecontent–itoughttoprescribehowscientistsshouldreason,aswellasdescribehowtheydoreason.DavidHUME(chapter31)firstpointedoutthatthereisalogicalgapbetween‘is’and‘ought’.Anaturalizedepistemologybasedontheempiricalstudyofscienceseemsfatedtoremainonthewrongsideofthisgap.However,thereisroomfornaturalizedepistemologiststoreplytothischarge.Theycanagreethattheempiricalstudyofsciencecannotbyitselfyieldprescriptionsabouthowscienceoughttobedone.Butempiricalstudycanstillberelevanttosuchprescriptions.Supposeitisagreedthattechnologicalfertility,inthesenseofgeneratingtechnologicaladvances,isavirtueinascientifictheory.Thenthehistory,sociologyandpsychologyofsciencemightbeabletoshowusthatcertainkindsofresearchstrategiesareeffectiveatdevelopingtechnologicallyfertiletheories.Moregenerally,givenanyagreedtheoreticalendY,empiricalstudycanshowthatresearchstrategyXisaneffectivemeanstothatend.Theempiricalstudyofsciencecanthusyieldthehypotheticalprescriptionthat,ifyouwantY,thenyououghttoadoptmeansX.Itisthiskindofhypotheticalprescriptionthatnaturalizedphilosophersofscienceseektoestablish:theylooktothehistory,sociologyandpsychologyofsciencetoshowusthatscientistswhochoosetheoriesongroundsXwillingeneralachievetheorieswithcharacteristicY.Canthenaturalizedstudyofsciencetelluswhichresearchstrategiesareaneffec-tivemeanstotheoreticaltruth?Differentnaturalizedphilosophersofsciencegivedif-ferentanswerstothisquestion.Manyaresuspiciousoftheideaoftheoreticaltruth,andinsteadprefertosticktothestudyofhowtoachievemorepracticalendsliketech-nologicalfertility,simplicityandpredictiveaccuracy.However,thereseemsnogoodreasonforthisrestriction.Thereisnothingobviouslyincoherentintheideaoflook-ingtotheempiricalstudyofsciencetotelluswhichresearchstrategieshaveprovedagoodwayofdevelopingtruetheories.Indeed,thediscussionofthe‘pessimisticmeta-induction’intheprevioussectionamountedtothesketchofjustsuchaninvestigation,inthatitappealedtothehistoryofsciencetodecidewhetherornotthestandardpro-ceduresofscientifictheory-choicesucceedinidentifyingtruetheories.Itisnotdifficulttoimaginemoredetailedandspecificstudiesofthiskindofissue.LetmenowreturnbrieflytotheissuewithwhichIbegan,namely,theproblemofinduction.Itispossiblethatthenaturalizedstudyofhowtogetatthescientifictruthwillenableustomakeheadwaywiththisproblem.Foranempiricalinvestigationinto302\nPHILOSOPHYOFSCIENCEsciencemightbeabletoshowusthatacertainkindofinductiveinferenceisingeneralareliableguidetoscientifictruth.Andthiswouldthenprovideakindofvindicationofthatinductivemethod(seePapineau1993:ch.5).Itistruethatthiskindofdefenceofinductionwillinevitablyinvolveanelementofcircularity.Forwhenweinferthatcertainkindsofinductionareingeneralareliableguidetotruth,onthebasisofevidencefromthehistoryofscience,thiswillitselfbeaninductiveinference.Itisamatterofsomedelicacy,however,whetherthiscircularityisvicious.Defendersofthisnaturalizeddefenceofinductionwillpointoutthat,fromtheirpointofview,alegitimatecriterionoftheory-choiceneednotbeanaprioriguidetotruth,butonlyanempiricallycertifiableone.Giventhis,theoriginalargumentagainstinduction,thatitisnotlogicallyvalid,willnotworrynaturalizedphilosophersofscience.Inductionmaynotprovideanyaprioriguaranteeforitsconclusions;butfromthenaturalizedpointofview,thisdoesnotshowthatinductionisinanywayillegiti-mate,sinceitleavesitopenthatinductionmaybeanempiricallyreliableguidetothetruth.Andifthereisnothingtoshowthatinductionisillegitimate,naturalizedphiloso-phersofsciencecanthenargue,whyshouldwenotuseittoinvestigatetheworthofinductiveinferences?Maybethisislesssatisfyingadefenceofinductionthanwemightoriginallyhavehopedfor.Butperhapsitisdefenceenough.2TheMetaphysicsofScience2.1CausationManyissuesinthemetaphysicsofsciencehingeonthenotionofcausation.Thisnotionisasimportantinscienceasitisineverydaythinking,andmuchscientifictheorizingisconcernedspecificallytoidentifythecausesofvariousphenomena.However,thereislittlephilosophicalagreementonwhatitmeanstosaythatoneeventisthecauseofanother.ModerndiscussionofcausationstartswithDavidHume,whoarguedthatcausationissimplyamatterofCONSTANTCONJUNCTION(p.720).AccordingtoHume(1978),oneeventcausesanotherifandonlyifeventsofthetypetowhichthefirsteventbelongsregularlyoccurinconjunctionwitheventsofthetypetowhichthesecondeventbelongs.Thisformulation,however,leavesanumberofquestionsopen.Firstly,thereistheproblemofdistinguishinggenuinecausallawsfromaccidentalregularities.Notallregularitiesaresufficientlylawliketounderpincausalrelationships.Beingascrewinmydeskcouldwellbeconstantlyconjoinedwithbeingmadeofcopper,withoutitsbeingtruethatthesescrewsaremadeofcopperbecausetheyareinmydesk.Secondly,theideaofconstantconjunctiondoesnotgiveadirectiontocausation.Causesneedtobedistinguishedfromeffects.ButknowingthatA-typeeventsarecon-stantlyconjoinedwithB-typeeventsdoesnottelluswhichofAandBisthecauseandwhichtheeffect,sinceconstantconjunctionisitselfasymmetricrelation.Thirdly,thereisaproblemaboutprobabilisticcausation.Whenwesaythatcausesandeffectsareconstantlyconjoined,dowemeanthattheeffectsarealwaysfoundwiththecauses,orisitenoughthatthecausesmaketheeffectsprobable?303\nDAVIDPAPINEAUManyphilosophersofscienceduringthepastcenturyhavepreferredtotalkaboutexplanationratherthancausation.Accordingtothecovering-lawmodelofexplanation,somethingisexplainedifitcanbededucedfrompremiseswhichincludeoneormorelaws.Asappliedtotheexplanationofparticularevents,thisimpliesthatoneparticu-lareventcanbeexplainedifitislinkedbyalawtosomeotherparticularevent.However,whiletheyareoftentreatedasseparatetheories,thecovering-lawaccountofexplanationisatbottomlittlemorethanavariantofHume’sconstantconjunctionaccountofcausation.Thisaffinityshowsupinthefactthatthecovering-lawaccountfacesessentiallythesamedifficultiesasHume:(1)inappealingtodeductionsfrom‘laws’,itneedstoexplainthedifferencebetweengenuinelawsandaccidentallytrueregularities;(2)itomitstherequisitedirectionality,inthatitdoesnottelluswhyweshouldnot‘explain’causesbyeffects,aswellaseffectsbycauses;afterall,itisaseasytodeducetheheightofaflagpolefromthelengthofitsshadowandthelawsofoptics,astodeducethelengthoftheshadowfromtheheightofthepoleandthesamelaws;(3)arethelawsinvokedinexplanationrequiredtobeexceptionlessanddeterministic,orisitacceptable,say,toappealtothemerelyprobabilisticfactthatsmokingmakescancermorelikely,inexplainingwhysomeparticularpersondevelopedcancer?InwhatfollowsIshalldiscussthesethreeproblemsinorder(treatingthemasprob-lemsthatariseequallybothfortheanalysisofcausationandtheanalysisofexplana-tion).AfterthatIshallconsidersomefurtherissuesinthemetaphysicsofscience.TheCovering-LawModelofExplanationAccordingtothismodel(originallyproposedbyHempelandOppenheim(1948)andfurtherelaboratedinHempel(1965))onestatement(theexplanandum)isexplainedbyotherstatements(theexplanans)ifandonlyiftheexplananscontainsoneormorelaws,andtheexplanandumcanbededucedfromtheexplanans.Inthesimplestcase,wheretheexplanandumissomeparticularstatementtotheeffectthatsomeindividualahaspropertyE,wemightthereforehave:ahasCForallx,ifxhasC,thenxhasEahasEForexample,wemightdeducethatapieceoflitmuspaperturnedred,fromthelawthatalllitmuspaperplacedinacidturnsred,togetherwiththepriorconditionthatthispieceoflitmuspaperwasinfactplacedinacid.Themodelcanaccommodatemorecompli-catedexplanationsofparticularevents,andcanalsoallowexplanationsoflawsthemselves,aswhenwededuceKepler’slawthatallplanetsmoveinellipses,say,fromNewton’slawofuniversalgravitationandhislawsofmotion.Asappliedtotheexplanationofparticularevents,thecovering-lawmodelimpliesasymmetrybetweenexplanationandprediction.Fortheinformationthat,accordingtothemodel,sufficesfortheexplanationofsomeknowneventshouldalsoenableustopredictthateventifwedidnotyetknowofit.Manycriticshavefastenedonthisimpli-cationofthemodel,however,andpointedoutthatwecanoftenpredictwhenwedonothaveenoughinformationtoexplain(aswhenwepredicttheheightoftheflagpolefrom304\nPHILOSOPHYOFSCIENCEitsshadow)andcanoftenexplainwhenwecouldnothavepredicted(aswhenweexplainX’scanceronthebasisofX’ssmoking).Theseexamplessuggestthatgenuineexplanationsofparticulareventsneedtocitegenuinecauses,andthatthereasonthecovering-lawmodelrunsintocounter-examplesisthatitaddsnothingtotheinadequateconstantconjunctionanalysisofcausation,exceptthatitsubstitutestheterm‘law’for‘constantconjunction’.Togetasatisfactoryaccountofexplanationweneed,firstly,torecognizethatexplanationsofparticulareventsmustmentioncauses,and,secondly,toimproveontheconstantconjunctionanalysisofcausation.Thereisavariantofthecovering-lawmodelwhichallowsnon-deterministicexpla-nationaswellasdeterministicones.Thisistermedthe‘inductive–statistical(I–S)’model,bycontrastwiththeoriginal‘deductive–nomological(D–N)’model.Anexamplewouldbe:adrinks10unitsofalcoholperdiemForppercentofxs,ifxdrinks10unitsofalcoholperdiem,xhasadamagedliverahasadamagedliverHeretheexplanandumcannotbededucedfromtheexplanans,butonlyfollowswithaninductiveprobabilityofp;andtheinferenceappealstoastatisticalregularity,ratherthananexceptionlessnomologicalgeneralization.InHempel’soriginalversionofthismodel,itwasrequiredthattheprobabilityoftheexplanandumbehigh.Abetterrequirement,however,asexplainedinthesectiononprobabilisticcausationbelow,isthatthepar-ticularfactsintheexplanansneedonlymaketheprobabilityoftheexplanandumhigherthanitwouldotherwisehavebeen.2.2LawsandaccidentsTherearetwogeneralstrategiesfordistinguishinglawsfromaccidentallytruegener-alizations.ThefirststandsbyHume’sideathatcausalconnectionsaremereconstantconjunctions,andthenseekstoexplainwhysomeconstantconjunctionsarebetterthanothers.Thatis,thisfirststrategyacceptstheprinciplethatcausationinvolvesnothingmorethancertaineventsalwayshappeningtogetherwithcertainothers,andthenseekstoexplainwhysomesuchpatterns–the‘laws’–mattermorethanothers–the‘accidents’.Thesecondstrategy,bycontrast,rejectstheHumeanpresuppositionthatcausationinvolvesnothingmorethanhappenstantialco-occurrence,andinsteadpostulatesarelationshipof‘necessitation’,akindof‘cement’,whichlinkseventsthatareconnectedbylaw,butnotthoseevents(likebeingascrewinmydeskandbeingmadeofcopper)thatareonlyaccidentallyconjoined.ThereareanumberofversionsofthefirstHumeanstrategy.Themostsuccessful,originallyproposedbyF.R.Ramsey(1903–30),andlaterrevivedbyDavidLewis(1973),holdsthatlawsarethosetruegeneralizationsthatcanbefittedintoanidealsystemofknowledge.Thethoughthereisthatthelawsarethosepatternsthataresomehowexplicableintermsofbasicscience,eitherasfundamentalprinciplesthem-selves,orasconsequencesofthoseprinciples,whileaccidents,althoughtrue,havenosuchexplanation.Thus,‘Allwateratstandardpressureboilsat100°C’isaconsequence305\nDAVIDPAPINEAUofthelawsgoverningmolecularbonding;butthefactthat‘Allthescrewsinmydeskarecopper’isnotpartofthedeductivestructureofanysatisfactoryscience.Ramseyneatlyencapsulatedthisideabysayingthatlawsare‘consequencesofthoseproposi-tionswhichweshouldtakeasaxiomsifwekneweverythingandorganizeditassimplyaspossibleinadeductivesystem’(Ramsey1978:130).Advocatesofthealternativenon-Humeanstrategyobjectthatthedifferencebetweenlawsandaccidentsisnotalinguisticmatterofdeductivesystematization,butratherametaphysicalcontrastbetweenthekindoflinkstheyreport.Theyarguethatthereisalinkinnaturebetweenbeingat100°Candboiling,butnotbetweenbeinginmydeskandbeingmadeofcopper,andthatthisisnothingtodowithhowthedescrip-tionofthislinkmayfitintotheories.AccordingtoD.M.Armstrong(1983),themostprominentdefenderofthisview,therealdifferencebetweenlawsandaccidentsissimplythatlawsreportrelationshipsofnaturalnecessitation,whileaccidentsonlyreportthattwotypesofeventshappentooccurtogether.Armstrong’sviewmayseemintuitivelyplausible,butitisarguablethatthenotionofnecessitationsimplyrestatestheproblem,ratherthansolvingit.Armstrongsaysthatnecessitationinvolvessomethingmorethanconstantconjunction:iftwoeventsarerelatedbynecessitation,thenitfollowsthattheyareconstantlyconjoined;buttwoeventscanbeconstantlyconjoinedwithoutbeingrelatedbynecessitation,aswhentheconstantconjunctionisjustamatterofaccident.Sonecessitationisastrongerrela-tionshipthanconstantconjunction.However,Armstrongandotherdefendersofthisviewsayverylittleaboutwhatthisextrastrengthamountsto,exceptthatitdistin-guisheslawsfromaccidents.Armstrong’scriticsarguethatasatisfactoryaccountoflawsoughttocastmorelightthanthisonthenatureoflaws.2.3ThedirectionofcausationHumesaidthattheearlieroftwocausallyrelatedeventsisalwaysthecause,andthelatertheeffect.However,thereareanumberofobjectionstousingtheearlier–later‘arrowoftime’toanalysethedirectional‘arrowofcausation’.Forastart,itseemsinprinciplepossiblethatsomecausesandeffectscouldbesimultaneous.Moreseriously,theideathattimeisdirectedfrom‘earlier’to‘later’itselfstandsinneedofphilosophi-calexplanation–andoneofthemostpopularexplanationsisthattheideaof‘move-ment’fromearliertolaterdependsonthefactthatcause–effectpairsalwayshaveagivenorientationintime.However,ifweadoptsucha‘causaltheoryofthearrowoftime’,andexplain‘earlier’asthedirectioninwhichcauseslie,and‘later’asthedirectionofeffects,thenwewillclearlyneedtofindsomeaccountofthedirectionofcausationwhichdoesnotitselfassumethedirectionoftime.Anumberofsuchaccountshavebeenproposed.DavidLewis(1979)hasarguedthattheasymmetryofcausationderivesfroman‘asymmetryofoverdetermination’.Theoverdeterminationofpresenteventsbypastevents–considerapersonwhodiesaftersimultaneouslybeingshotandstruckbylightning–isaveryrareoccurrence.Bycontrast,themultiple‘overdetermination’ofpresenteventsbyfutureeventsisabsolutelynormal.Thisisbecausethefuture,unlikethepast,willalwayscontainmul-tipletracesofanypresentevent.TouseLewis’sexample,whenthepresidentpressestheredbuttonintheWhiteHouse,thefutureeffectsdonotonlyincludethedispatch306\nPHILOSOPHYOFSCIENCEofnuclearmissiles,butalsohisfingerprintonthebutton,histrembling,thefurtherdepletionofhisginbottle,therecordingofthebutton’sclickontape,theemissionoflightwavesbearingtheimageofhisactionthroughthewindow,thewarmingofthewirefromthepassageofthesignalcurrent,andsoon,andon,andon.Lewisrelatesthisasymmetryofoverdeterminationtotheasymmetryofcausationasfollows.Ifwesupposethecauseofagiveneffecttohavebeenabsent,thenthisimpliestheeffectwouldhavebeenabsenttoo,since(apartfromfreakslikethelight-ning–shootingcase)therewillnotbeanyothercausesleftto‘fix’theeffect.Bycon-trast,ifwesupposeagiveneffectofsomecausetohavebeenabsent,thisdoesnotimplythecausewouldhavebeenabsent,fortherearestillalltheothertracesleftto‘fix’thecause.Lewisarguesthatthesecounterfactualconsiderationssufficetoshowwhycausesaredifferentfromeffects.OtherphilosophersappealtoaprobabilisticvariantofLewis’sasymmetry.Follow-ingReichenbach(1956),theynotethatthedifferentcausesofanygiventypeofeffectarenormallyprobabilisticallyindependentofeachother;bycontrast,thedifferenteffectsofanygiventypeofcausearenormallyprobabilisticallycorrelated.Forexample,bothobesityandhighexcitementcancauseheartattacks,butthisdoesnotimplythatfatpeoplearemorelikelytogetexcitedthanthinones;ontheotherhand,thefactthatbothlungcancerandnicotine-stainedfingerscanresultfromsmokingdoesimplythatlungcancerismorelikelyamongpeoplewithnicotine-stainedfingers.Sothisaccountdistinguisheseffectsfromcausesbythefactthattheformer,butnotthelatter,areprobabilisticallydependentoneachother.2.4ProbabilisticcausationThejust-mentionedprobabilisticaccountofthedirectionofcausationisnormallyfor-mulatedaspartofamoregeneraltheoryofprobabilisticcausation.Untilrelativelyrecentlyphilosophersassumedthattheworldfundamentallyconformstodeterminis-ticlaws,andthatprobabilisticdependenciesbetweentypesofevents,suchasthatbetweensmokingandlungcancer,merelyreflectedourignoranceofthefullcauses.Theriseofquantummechanics,however,haspersuadedmostphilosophersthatdeter-minismisfalse,andthatsomeevents,likethedecayofaradiumatom,happenpurelyasamatterofchance.Aparticularradiumatommaydecay,butonanotheroccasionanidenticalatominidenticalcircumstancesmightwellnotdecay.Accordingly,anumberofphilosophersofsciencehaveputforwardmodelsofcau-sationwhichrequireonlythatcausesprobabilify,ratherthandetermine,theireffects.Theearliestsuchmodelwasthe‘inductive–statistical’versionofthecovering-lawmodelofexplanation(Hempel1965).Unlikedeterministic‘deductive–nomological’explanations,suchinductive–statisticalexplanationsrequiredonlythatpriorcondi-tionsandlawsimplyahighprobabilityfortheeventtobeexplained,notthatthiseventwillcertainlyhappen.However,eventhisseemstoostrongarequirementforprobabilisticcausation.Afterall,smokingunequivocallycauseslungcancer,butevenheavysmokersdonothaveahighprobabilityoflungcancer,inthesenseofaproba-bilitycloseto1.Rather,theirsmokingincreasestheirprobabilityoflungcancer,nottoahighfigure,butmerelyfromalowtoalesslowfigure,butstillwellbelow50percent.Somorerecentmodelsofprobabilisticcausationsimplyrequirethatcauses307\nDAVIDPAPINEAUshouldincreasetheprobabilityoftheireffects,notthattheyshouldgivethemahighprobability(Salmon1971).Thiskindofmodelneedstoguardagainstthepossibilitythattheprobabilisticasso-ciationbetweenputativecauseandputativeeffectmaybespurious,liketheprobabilis-ticassociationbetweenbarometersfallingandsubsequentrain.Suchassociationsarenotduetoacausalconnectionbetweenbarometermovementsandrain,butrathertobothofthesebeingjointeffectsofacommoncause,namely,inourexample,fallsinatmosphericpressure.Theobviousresponsetothisdifficultyistosaythatwehaveacause–effectrelationshipbetweenAandBifandonlyifAincreasestheprobabilityofB,andthisassociationisnotduetosomecommoncauseC.However,thisisobviouslyincompleteasananalysisofcausation,sinceitusesthenotionof(common)causeinexplainingcausation.Itwouldsolvethisproblemifwecouldanalysethenotionofcommoncauseinprob-abilisticterms.Itseemstobeamarkofcommoncausesthattheyprobabilistically‘screenoff’theassociationsbetweentheirjointeffects,inthesensethat,ifweconsidercaseswherethecommoncauseispresentandwhereitisabsentseparately,thentheprobabilisticassociationbetweenthejointeffectswilldisappear.Forexample,ifitisgiventhattheatmosphericpressurehasfallen,thenafallingbarometerdoesnotmakeitanymorelikelythatitwillrain;andsimilarly,iftheatmosphericpressurehasnotfallen,afaultyfallingreadingonabarometerisnoprobableindicatorofimpendingrain.(Numerically,ifCisacommoncause,andAandBitsjointeffects,wewillfindthatAandBareassociated–Prob(B/A)Pr(S/separateancestry).Ifallspeciesonearthsharecommonancestors,thequestionremainsastowhichspeciesarecloselyrelatedandwhicharerelatedonlymoredistantly.Themethodologyofthisproblemhasbeenintensivelystudiedinevolutionarybiology.Oneprinciplethatisoftenusedistermedphylogeneticparsimony.ItwasindependentlyproposedbyHennig(1966)andbyEdwardsandCavalli-Sforza(1964).Considerthetwophylogenetictreesdepictedinfigure10.1.Thetipsofthetreesrepresentpresent-dayspeciesorsuperspe-cifictaxa;interiornodesrepresentancestors.The(SR)Ctreerepresentsthehypothesisthatsparrowsandrobinshaveacommonancestorthatisnotanancestorofcroco-diles;theS(RC)treerepresentstheclaimthatrobinsandcrocsaremorecloselyrelatedtoeachotherthaneitheristosparrows.Weknowbyobservationthatsparrowsandrobinshavewings,butcrocodilesdonot.Canthisobservationbeusedtodiscriminatebetweenthetwogenealogicalhypotheses?Ifweassumethatthespeciesattherootofthetreelackedwings,thenwearesayingthatsparrowsandrobinsshareaderivedcharacter(anapomorphy)thatcrocodilesdonotpossess.The(SR)CtreeismoreparsimoniousthantheS(RC)treeinthiscase,because(SR)Cisabletoexplaintheobservationsatthetipsofthetreeby322\nPHILOSOPHYOFBIOLOGYIngroupOutgroupswinglesswingless?Figure10.2postulatingasinglechangeincharacterstate(fromwinglesstowinged)inthetree’sinterior,whereasS(RC)requirestwosuchchangestoexplaintheobservations.Ontheotherhand,ifthespeciesattherootofthetreeshadwings,thensparrowsandrobinswouldshareanancestralcharacter(aplesiomorphy),andthetwohypotheseswouldbeequallyparsimonious,sinceeachcouldexplaintheobservationsbypostulat-ingasinglechangeincharacterstate.Thus,theprincipleofphylogeneticparsimonyassertsthatderivedsimilaritiesareevidenceofrelatedness,whileancestralsimilaritiesarenot.Howisonetodeterminewhetherwingsorlackofwingsistheancestralconditionintheproblemjustdescribed?Theprincipleofparsimonyaddressesthisquestionaswell.Evenifwedon’tknowhowrobins,sparrowsandcrocodilesarerelated,westillmayknowthatthesethreegroupsaremorecloselyrelatedtoeachotherthananyofthemistovariousothergroups(forexample,daffodils).Iftheoutgroupspeciesdepictedinfigure10.2alllackwings,thenphylogeneticparsimonydictatesthatweshouldconcludethatwinglessnessistheancestralconditionfortheingroupaswell.Toinferphylogeneticrelationships,whyshouldparsimonybeused,ratherthanacriterionofoverallsimilarity,whichaccordsevidentialsignificancetoderivedandancestralsimilaritiesalike?SomehavearguedthatPopperianconsiderationscon-cerningFALSIFIABILITY(pp.287–90)sufficetojustifyphylogeneticparsimony(see,forexample,EldredgeandCracraft1980;Wiley1981);othershavesuggestedalikelihoodanalysisinstead(Sober1988).Phylogeneticinferenceprovidesaninterestingcontextforconsideringphilosophicalquestionsabouttheroleofaparsimonyprincipleinscientificreasoning.Althoughphilosophersoftencomplainthatitisunclearwhatitmeansforonetheorytobesimplerormoreparsimoniousthananother,itisclearenoughwhatitmeansforonegeneal-ogicalhypothesistoprovideamoreparsimoniousexplanationoftheobservationsthananother.Andinanswertothequestionofwhyweshoulduseparsimonytoinfergenealogicalrelatedness,itseemsunsatisfactorytoreplythatapreferenceforparsi-monioushypothesesispartofwhatitmeanstothinkscientifically.Whyisit‘unscien-tific’touseoverallsimilarityasone’sguide?Parsimonyisnotanendinitself;ifa323\nELLIOTTSOBERpreferenceforparsimoniousexplanationsisjustified,itisjustifiedbecauseitprovidesameansofachievingsomemoreultimateepistemicend.2.3AdaptationismNaturalselectionisoneamongseveralprocessesthatcaninfluencetheevolutionofatrait.Sincealineagestartsevolvingwithsomepre-existingsetofancestraltraits,thecharacteristicsexhibitedbydescendantsmayshowtheinfluenceofthatancestralcondition.Forexample,supposeanancestralbearspecieshasafurthicknessof5cm.Theclimatethengetscolder,sothattheoptimalfurthicknessforoneofitsdescendantswouldbe12cm.Ifthedescendantachievesafurthicknessof,say,10cm,onemaywanttoattributethisoutcomebothtonaturalselectionandtophylogeneticinertia,sometimescalledancestralinfluence(OrzackandSober2001).Anotherfactorthatcanpreventnaturalselectionfrommovingspeciestooptimaltraitvaluesistheunderlyinggenetics.Forexample,ifthefittestphenotypeiscodedbyaheterozygote,itwillbeimpossibleforallindividualstoexhibitthatoptimalphenotype.Insimilarfashion,randomgeneticdrift,correlationofcharacters,andotherfactorscanpreventnaturalselectionfromcausingthefittestavailablephenotypetoevolve(Sober1993).Thedebateabout‘adaptationism’thathasbeengoingonforthelasttwenty-someyearsinevolutionarybiologydoesnotconcernDarwin’streeoflifehypothesis;nordoesitconcernthemodestclaimthatnaturalselectionhasinfluencedmostofthepheno-typictraitsthatspeciesexhibit.Rather,thequestionconcernsthepowerofnaturalselection.Isselectionmerelyoneamongseveralimportantinfluencesontraitevolution,orisittheonlyimportantinfluence(OrzackandSober1994;Sober1993)?Canphylogeneticinertiaandgeneticconstraintsbeignoredinpredictingthedescendantbear’sfurthickness?Adaptationismembodiesarelativelymonisticconceptionoftraitevolution;itsalternativeisevolutionarypluralism(GouldandLewontin1979).Adaptationism,PluralismandMethodologySeparatefromthesubstantivedebateabouttheprocessesthathavegovernedevolution,thedebateaboutadaptationismhashadanimportantmethodologicaldimension.GouldandLewontin(1979)accusedadaptationistsofendorsing‘just-sostories’–ofacceptingexplanationsuncritically.Adaptivehypothesesneedtobetestedrigorously,whichmeansthattheyshouldbetestedagainstnon-adaptivealternativeexplanations.GouldandLewontinalsosuggestedthatadaptationismviolatesPopper’sinjunctionthatscientifichypothesesshouldbefalsifiable.Noticethatneitherofthesecriticisms,eveniftheywerecorrect,wouldshowthatpluralismistrueandmonismisfalseasclaimsaboutnature.Thefactthatapropositionhasbeenaccepteduncriticallydoesnotentailthatthepropositionisfalse.Thechargethatadaptationismisuntestableshouldnotbetakenatfacevalue.Itistruethatifoneadaptationisthypothesisisrejected,thenanothercanbeinvented.Thisisbecauseadaptationismisanism–itspecifiesthekindofexplanationthatallormosttraitshavewithoutsayinganythingindetailaboutwhatspecificadaptationisthypothe-sesaretrue.Consider,forexample,thefactsomespeciesreproducesexuallywhileothers324\nPHILOSOPHYOFBIOLOGYreproduceasexually.Adaptationistswillinsistthatthisvariationexistsbecausesexualreproductionisbetterinsomecircumstancesandasexualreproductionisbetterinothers.However,thisgeneralclaimdoesnottelluswhyselectionfavourssexualrepro-ductioninsomecasesandasexualreproductioninothers.AdaptationismisaresearchprogrammeinsomethinglikethesensethatLakatos(1978)described(MitchellandValone1990);thefailureofoneorseveraladaptationistmodelsdoesnotmeanthattheresearchprogrammeisbankrupt.Thesamepointappliestoevolutionarypluralism:ifonepluralisticmodelfails,anothercanbeinvented,butthisdoesnotshowthatplural-ismshouldberejected.Itisimportantthatscientistsrigorouslytestspecifichypotheses,whethertheyareadaptationistorpluralisticinform;however,theismsthatgovernresearchprogrammesarenotsoeasilytested.Onlytheaccumulationofsuccessesandfailuresoverthelongrunwilltelluswhetheradaptationismistrueasaclaimaboutnature(OrzackandSober1994).2.4TheunitsofselectioncontroversyWhetherornottheconceptoffunctionisunderstoodhistorically(section1.3),itisprettyclearthatevolutionistsusetheconceptofadaptationinthisway.Tosaythatwingsareanadaptationforflyingdoesnotmeanthatwingsnowareusefulbecausetheyhelporganismstofly;itmeansthatwingsevolvedbecausetheyhelpedorganismstofly.Therelevantconsiderationisinthepasttense,notthepresent.Wingscouldbeanadaptationforflyingevenifitnowisdisadvantageousfortheorganismsinquestiontoleavetheground.Historicaloriginandpresentadvantagearenotautomaticallylinked.Noticealsothatwingscanbeadaptationsforflyingevenifadaptationismiswronginwhatitsaysaboutwings;tosaythatnaturalselectionhelpedcausethetraittoevolvedoesnotmeanthatselectionwastheonlyimportantinfluenceonthetrait’sevolution.WhenDarwinarguedthatthisorthattraitisanadaptation,heusuallyhadinmindthatthetraitevolvedbecauseithelpedindividualorganismstosurviveandreproduce.Forexample,predatorshavesharpteethbecauseindividualswithsharpteethdobetterinthestruggleforexistencethanindividualsthathavedullteeth.Theexplanationisnotthatthetraitevolvedbecauseithelpedthewholespecies,orthewholeecosystem.TheadaptationsthatDarwindiscussedaremainlyindividualadaptations.Theseindi-vidualadaptationsevolvedbytheprocessofindividualselection,whereinindividualscompetewithotherindividualsinthesamespecies.Therewereafewoccasions,however,inwhichDarwinsawthingsdifferently.Whyaretheresterileworkersinsomespeciesofsocialinsect?Thisisnotbecausesterileindi-vidualsdobetterinsurvivingandreproducingthanfertileindividuals.Rather,Darwinarguedthatneststhatcontainsterileworkersdobetterthanneststhatdonot.Darwinintroducedtheideathatsometraitsaregroupadaptations–theyevolvedbecausetheyhelpedgroups,notbecausetheyhelpedindividualorganisms.Theseadaptationsevolvedbytheprocessofgroupselection,whereingroupscompetewithothergroupsinthesamespecies.Darwinthoughtthatgroupselectionisrequiredtounderstandimportantfeaturesofhumanmorality.HereishowheanalysedtheprobleminTheDescentofMan:325\nELLIOTTSOBERItmustnotbeforgottenthatalthoughahighstandardofmoralitygivesbutaslightornoadvantagetoeachindividualmanandhischildrenovertheothermenofthesametribe,yetthatanincreaseinthenumberofwell-endowedmenandadvancementinthestan-dardofmoralitywillcertainlygiveanimmenseadvantagetoonetribeoveranother.Therecanbenodoubtthatatribeincludingmanymemberswho,frompossessinginahighdegreethespiritofpatriotism,fidelity,obedience,courage,andsympathy,werealwaysreadytoaidoneanother,andtosacrificethemselvesforthecommongood,wouldbevictoriousovermostothertribes;andthiswouldbenaturalselection.Atalltimesthroughouttheworldtribeshavesupplantedothertribes;andasmoralityisoneimportantelementintheirsuccess,thestandardofmoralityandthenumberofwell-endowedmenwillthuseverywheretendtoriseandincrease.(Darwin1981:166)ThecharactertraitsthatDarwindescribesaredisadvantageoustoindividuals,butadvantageoustogroups.Acourageousmanwhoriskshislifeinwarfareislessfitthanacowardinthesamegroupwhoplaysitsafe;yetgroupsthatincludecourageousindividualsarefitterthangroupsthatincludeonlycowards.Biologistscametousetheterm‘altruistic’fortraitsthataredeleterioustoindivi-dualsbutadvantageoustogroups.Thisterminologyispotentiallyconfusing,sincealtruism,sodefined,describesthefitnesscostsandbenefitsofthetrait,nothoworevenwhethertheindividualthinksandfeels.SterileworkersinanestofsocialinsectsandthewarriorswhomDarwindescribedarebothaltruisticintheevolutionarysense,eventhoughthewarriorshaveviewsabouttheirownbehaviourwhiletheinsectspresum-ablydonot.Andjustasamindlessorganismcanbeanevolutionaryaltruist,italsoispossibleforabehaviourthatenhancestheactor’sownfitnesstobemotivatedbydesiresthatarepsychologicallyaltruistic(SoberandWilson1998).TheideathatevolutioninvolvesbothindividualselectionandgroupselectionwasstandardfareintheheydayoftheModernSynthesisinevolutionarybiology,roughlyfrom1930–60.Biologistsinvokedindividualselectiontoexplainsometraitsandgroupselectiontoexplainothers.Mathematicalpopulationgeneticists,suchasR.A.Fisher,J.B.S.HaldaneandSewallWright,brieflyexpressedreservationsabouttheabilityofgroupselectiontocausealtruistictraitstoevolve,butthesetheoreticalpointswerenotgivenmuchcredencebybiologistsinthetrenches;thesebiologiststhoughttheyobservedaltruisminnatureandtheysoughttoexplainaltruismintheonlywaytheyknewhow–byinvokingthehypothesisofgroupselection.Thiscomplacentpluralismabouttheunitsofselectionwasshatteredinthe1960s,whengroupselectionwasattackedbyanumberofbiologists.ThemostthoroughanddevastatingcritiquewasG.C.Williams’s(1966)bookAdaptationandNaturalSelection.Williamstouchedanerveandhisvigorousrejectionofadaptationsthatexistforthegoodofthegroupspreadquicklythroughtheevolutioncommunity.TheargumentsthatWilliamsadvancedwerepopularizedtenyearslaterbyRichardDawkins(1976)inhiswidelyreadbookTheSelfishGene.Theideathatsometraitsevolvebecausetheybenefitthegroupwasreplacedbygenicselectionism–theideathatalltraitsevolvebecausetheyaregoodforthegene.However,notallbiologistssubscribedtothisnewviewpoint;theyarguedthatgroupselectionhadbeenrejectedprematurely(forexample,Lewontin1970;Wilson1980;Wade1978).Sincethen,multi-levelselectiontheory–theviewthatselectionoccursatalllevelsofthebiologicalhierarchy–hasbeenmakingacomeback(SoberandWilson1998).326\nPHILOSOPHYOFBIOLOGYAssessingGroupSelection:AnEmpiricalorConceptualControversyThisscientificcontroversyisphilosophicallyinterestingbecausegroupselectionwasrejectedinthe1960sforanumberofreasons,andthesereasonsdiffersomuchincharacterthatitisunclearwhatinthecontroversyisempiricalandwhatisconceptual.Theschizophreniccharacteroftheargumentsthatbiologistsproducedmaybeillustratedbytheirdiscussionoftheevolutionofsexratio–themixofmalesandfemalesthatapopulationexhibits.Williams(1966)treatsthisasanempiricaltestcaseforgroupselection.Henotes,followingFisher(1957),thatindividualselectionwouldfavouranevensexratio,whereasgroupselectionwouldsometimesfavourafemale-biasedsexratio(thelatterbecausegroupsmaximizetheirproductivitybyhavingthesmallestnumberofmalesconsistentwithallthefemalesbeingfertilized).Williamsclaimsthatthesexratiosobservedinnatureareuniformlyeven,andconcludesthattheempiricalhypothesisofgroupselectionismistaken,atleastasitpertainstotheevolutionofsexratio.Ayearlater,W.D.Hamiltonpublishedapapercalled‘ExtraordinarySexRatios’.Hamilton(1967)reportedthatfemale-biasedsexratiosareverycommoninthesocialinsects.Onemightexpectthatthisempiricalfinding,interpretedthroughthelensofWilliams’sargument,wouldhaveledbiologiststoconcludethatgroupselectionisimpli-catedintheevolutionofsexratio.However,thisispreciselyhowthemodelwasnotinterpreted.ThereasonisthatHamilton’spaperprovidesamathematicaltheoryforunderstandinghowunevensexratiosevolve.SolvingHamilton’sequationsallowsonetoidentifythefittest(the‘unbeatable’)sexratiostrategythatanindividualfemaleshouldfollowindetermininghermixofsonsanddaughters;thisformalismwaswidelyinter-pretedasshowinghowindividualselectioncouldpromotetheevolutionofafemale-biasedsexratio.OneofthefewbiologistswhodidnotinterpretHamilton’smodelinthiswaywasHamiltonhimself;Hamiltonsays,thoughonlyinafootnote,thathismodelrepresentstheactionofgroupselection.AlthoughHamiltonembracedtheideaofgroupselectionin1967andlaterpublications,hisfollowersweremoreHamiltonianthanHamiltonhimself.Williams(1992)concurswithHamilton’sinterpretation.Yetmanybiologistscon-tinuetoviewgroupselectionasbeyondthepale,andtheyciteHamiltonandWilliamsashavingshownwhy.SterelnyandKitcher(1988)argueinfavourofgenicselectionismonthegroundsthatalladaptationscanbeinterpretedasevolvingbecausetheybenefitthegenesthatcodeforthem.Biasedsexratios,extremealtruismandsoon,canallbebroughtundertherubricoftheselfishgene.Thereasongroupselectionisthewrongwaytothinkaboutnatureisnotthatitisempiricallydisconfirmed;rather,theproblemisthatitisinsuffi-cientlygeneral.Althoughalladaptationscanbetreatedasgenicadaptations,itisfalsethatalladaptationscanbeinterpretedasgroupadaptations.AccordingtoSterelnyandKitcher,thesolutiontotheunitsofselectionprobleminvolvesadoptingaconvention,notansweringafactualquestionaboutthehistoryoflife.Forthem,thefactthatatraitcanbeinterpretedasagenicadaptation,nomatterwhatthetraitislike,isapointinfavourofthegenicconvention.However,thosewhothinkthattheunitsofselectionproblemisanempiricalissueviewthisasadefect,notastrength,oftheframeworkthatSterelnyandKitcherrecommend(Wimsatt1980;SoberandLewontin1982;Lloyd1988;SoberandWilson1994,1998).327\nELLIOTTSOBER2.5Aretherelawsofevolution?AlthoughDarwin’sfundamentalcontributionwastheformulation,test,andapplica-tionoftwohistoricalhypotheses(seealsoKitcher1993:ch.2),thebroadsubjectofmodernevolutionarytheoryincludesmuchmorethanhistoricalhypotheses.Thesubjectispepperedwithwhatbiologistscall‘models’.Thesearemathematicalstatementsthatdescribewhattheevolutionaryconsequenceswouldbeofspecifiedinitialconditions.Forexample,Fisher’sanalysisofsexratioisanif/thenstatement.Fisher(1957)describedasufficientconditionthatensuresthatapopulationwithanunevensexratiowillevolvetowardsanevensexratio,andremainthere.Fisher’stheoryhasalltheearmarksofalawofnature,saveone.Itisgeneral,itdoesnotrefertoanyplace,timeorindividual,andithascounterfactualforce.However,Fisher’smodelisanapriorimathematicaltruth.ThisdoesnotmeanthattheclaimthatFisher’smodelappliestoarealworldpopulationisapriori;whetherthepopulationexhibitsrandommatingandsatisfiestheotherassumptionsofthemodelisanempiricalmatter,andwhetherthepopulationexhibitsanevensexratioisanempiricalmatteraswell.Theantecedentisempiricalandtheconsequentistoo,buttheconditionalisapriori(Sober1984,1993).Empiricalgeneralizationsexistinevolutionarybiology,buttheyoftenfailtohavethemodalforcethatlawsaresupposedtohave.Cope’sRule,forexample,saysthatspeciesevolveinthedirectionofincreasedsize.Thisis,atbest,aruleofthumbthatsumma-rizesapatternfoundinthefossilrecord(Hull1974).Theproblemisnotjustthatthereareexceptions(alaw,afterall,canbeprobabilistic),butthatthereisnothinginourunderstandingoftheevolutionaryprocessthattellsusthatevolutionmusthavethissortofdirectionality.Afundamentalpropertyofnaturalselectionisthatitis‘oppor-tunistic’.Ifgrowinglargerisadvantageous,selectionwillpushthepopulationalongthattrajectory;butifgrowingsmallerisadvantageous,selectionwillfavourthatoutcome,instead.Itisanaccidentofcircumstancewhethertheenvironmentfavoursonetransitionratherthantheother.Ifthehistoryoflifeonearthhappenstocontainmoreenvironmentsofthefirstkindthanofthesecond,Cope’sRulewillbetrueasastatisticalgeneralization,butitwillnotbealaw.Thesamepointholdsaboutincreasingcomplexity.Tobesure,complexityhasoftenincreased;however,therealsoaremanycasesinwhichcomplexityhasdeclined.Forexample,theevolutionofparasitesfromtheirfree-livingancestorsofteninvolvesthelossoforgans;selectionoftenleadsorganismstolosefeaturesthattheynolongerneed.Still,iflifestartedsimple,anincreaseincomplexitywasboundtooccur.Biologistsgen-erallyregardthisasanartefactoftheinitialconditionsthatobtained,notasreflectingsomeresolutedirectionalityintheevolutionaryprocessitself.Asananalogy,consideragameplayedonaone-rowcheckerboardthathasathousandsquares.Thegameisplayedinasequenceofsteps;ateachstep,acoinistossedtodecidehowthecheckerwillbemoved.Headsmeansthecheckermovesonesquaretotherightandtailsmeansthatitmovesonesquaretotheleft.Ofcourse,ifthecoinlandstailswhenthecheckerisontheleft-mostsquare,thecheckercan’tmovetotheleft;inthiscase,onemusttossthecoinagain.Thesameappliesifthecheckerisontheright-mostsquarewhenthecoinlandsheads.Sincethecoinisfair,thereisnodirectionalityinthelawsthatgovernthechecker’strajectory.Thelawsentailnonettendencyforthecheckertomovetothe328\nPHILOSOPHYOFBIOLOGYrightortotheleft.However,anasymmetrycanbeintroducedbythechecker’sinitiallocation.Supposewebeginthegamewiththecheckerintheleft-mostsquare.Afterfiftytosses,thecheckerwillalmostcertainlyoccupyasquarethatistotherightofwhereitbegan(Maynard-Smith1988;Gould1989;Sober1994a).Isofarhavesuggestedthattherearetwotypesofgeneralizationinevolutionarybiologythatfailtobescientificlawsintheusualsense.Theconditionalstatementsthatcomprisemathematicalmodelsareapriori,andempiricalgeneralizationsaboutevolu-tionarytrendsoftenareaccidentallytrue,iftrueatall.Thisdoesnotshowthattherearenolaws.Arethereany?Beatty(1995)hasarguedthattherearenone,notjustnoneinevolutionarybiology,butnoneinbiologyasawhole.Hisreasonforthisclaimisthateverybiologicalgeneralizationistrueinvirtueofacontingentevolutionaryprocess’smakingitso.Forexample,iftheorganismsinapopulationobeyMendel’slawofinde-pendentassortment,thisisbecauseevolutionledthisregularitytoevolve.Ifselectionhadfavouredsomeotherpatternofheredity,thatpatternmighthaveevolvedinstead.ThisisBeatty’sevolutionarycontingencythesis.OnequestionthatneedstobeconsideredaboutBeatty’sargumentisdepictedinthefollowingdiagram:Efi(IfPthenQ)tt01t2Iftheevolutionaryregularity‘IfPthenQ’holdstruebetweentimest1andt2onlybecausecontingentevolutionaryeventsEhappenedtotakeplaceattimet0,thenitmakessensetosaythattheregularityiscontingent.However,thisleavesitopenthatthemorecomplexconditional‘ifEoccurs,then(ifPthenQ)willbetruelater’holdstruenon-contingently(Sober1997a).Thispointdoesnotestablishthatbiologicallawsexist,butitdoesshowthatonecannotestablishthattherearenolawsjustbypointingoutthatregularitiesdependonearliercontingencies.Furthermore,ifcausalityentailstheexistenceoflaws(ametaphysicalclaimthatshouldnotbeaccepteduncritically;Anscombe(1975),forexample,deniesit),thenthecausaldependencyof‘IfPthenQ’onEentailstheexistenceofalaw.SensitivitytoInitialConditionsGould(1989)hasemphasizedtheideathatevolutionaryoutcomesoftenexhibitagreatdealofsensitivitytoinitialconditions.Evenasmallchangeintheconditionsthatobtainedmillionsofyearsagowouldhavemadeahugedifferenceinthedistributionoflifeformsthatwefindonearthnow.Ifacertainmeteorstrikehadnotoccurred,thedinosaurswouldnothavegoneextinct;ifthedinosaurshadnotgoneextinct,mammalswouldnothaveproliferated;andifthemammalianexplosionhadnotoccurred,nomammalresem-blingHomosapienswouldhavemadeitsappearance.AccordingtoGould,‘replayingthetape’wouldrevealthatmanyoftheinterestingfeaturesofthelifeformswefindaroundusareaccidentsofhistory.Theclaimthatthereissensitivitytoinitialconditionshasthefollowingform:althoughX1leadstoY1,italsoistruethatX2leadstoY2,whereX1andX2areverysimilar,whileY1andY2areverydifferent.Theusualglossonthisclaimisthattherearetwolaws:329\nELLIOTTSOBERX1leadstoY1andX2leadstoY2.Thus,thesensitivitythesisdoesnotentailthattherearenobiologicallaws.NoticealsothatGouldisnotclaimingthatallfeaturesoflivingthingsexhibitsensitivitytoinitialconditions;heismerelycallingattentiontotheimportanceofthisphenomenon.Itisperfectlypossiblethatsomefeaturesoflifearerobust(thatis,insensitivetoperturbationsininitialconditions)whileothersarenot.Thefrequencyofevolutionaryconvergenceshowsthatsimilaroutcomesoftenarisefromratherdifferentstartingpoints.Whatissorelyneededinthisareaarebiologicallywellmotivatedtheo-riesconcerningwhichfeaturesshouldberobustandwhichshouldexhibitsensitivity.Evolutionarybiologyverymuchneedstomovethisproblembeyondthereportingofpiecemealintuitionsaboutexamples.3ThePhilosophicalSignificanceofEvolutionaryTheory3.1TheargumentfromdesignThetwopartsofDarwin’stheoryofevolutionbynaturalselectionbothconflictwithatleastonetraditionalreadingofGenesis.First,thetheorypostulatesasingletreeoflife,andthuscontradictstheclaimthatdifferentgroupsoforganismswereseparatelycreated.Second,thetheorypostulatesthemindlessprocessofnaturalselectionasthecauseoflife’sadaptivefeatures,andthuscontradictstheclaimthatadaptivefeaturesarosebytheintelligentdesignofacreator.Thereareadditional,thoughlesscentral,pointsofconflictaswell;forexample,DarwinheldthattheearthisancientwhileoneliteralreadingoftheBiblesaysthattheearthisyoung.Notonlydoesthetheoryofevolutionconflictwithonesetofreligiousdoctrines;italsoconflictswithaninfluentialargumentfortheexistenceofGod–theARGUMENTFROMDESIGN(pp.478–80).ThefifthofTHOMASAQUINAS’s(chapter24)fivewaysofprovingtheexistenceofGodcontendsthatthingsthat‘actforanend’musteitherhavemindsorbeproducedbyanintelligentdesigner.InthecenturyinBritainprecedingDarwin’spublicationoftheOrigin,thisargumentwasrepeatedlyelaborated.ItwasnotsimplyoneargumentfortheexistenceofGod;itbecamethefundamentalargument.PerhapsthemostfamousformulationofthedesignargumentisthatgivenbyWilliamPaley,inhisbookNaturalTheology(1805).Paleyproposedananalogy.Ifyouwerewalkingacrossaheathandfoundastone,youwouldnotdreamofconcludingthatthestonewasproducedbyanintelligentdesigner.However,ifyoufoundawatch,youwouldconcludewithouthesitationthatthewatchwasproducedbyawatchmaker.Thedifferencebetweenthewatchandthestoneisthatthewatchexhibitsadaptivecomplexity.Thewatchasawholemeasuresoutequalintervalsoftime,andthepartsofthewatchconspiretoallowthewatchtoperformthatfunction;wereanyofthosepartsevenmodestlydifferentfromthewaytheyare,thewatchwouldnotkeeptime.Paleythenarguesthatwhatistrueofthewatchalsoholdsformanyfeaturesoforganisms.Thevertebrateeye,forexample,alsoexhibitsadaptivecomplexity.Paleyconcludesthatorganismsaretheresultofintelligentdesignnolessthanthewatchisthehandiworkofawatchmaker.Philosophershavenotalwaysagreedonwhattheformofthedesignargumentis.Forexample,HUME(chapter31),inhisDialoguesConcerningNaturalReligion(1779),330\nPHILOSOPHYOFBIOLOGYpublishedsomethirtyyearsbeforePaley’sNaturalTheologyappeared,suggeststhatthedesignargumentisananalogicalargumentthathasthefollowingform:Watchesareproducedbyintelligentdesign.Organismsarelikewatches.Organismswereproducedbyintelligentdesign.Thedoublelineseparatingpremissesfromconclusionindicatesthattheargumentisnotintendedtobedeductivelyvalid.Humesuggeststhatthestrengthoftheanalogi-calargumentdependsonthedegreetowhichwatchesandorganismsresembleeachother–themoresimilartheyare,themorestronglythepremissessupporttheconclu-sion.Hethenpointsoutthattherearemanydissimilarities–watchesaremadeofmetalandglass,organismsbreathe,andsoon.Humeconcludesthatthedesignargumentisaveryweakanalogicalargument.Amorecharitablereadingoftheargumentfromdesignconstruesitasaninferencetothebestexplanation(pp.42–3)(whatC.S.PEIRCE(chapter36)calledanabductiveargu-ment),onethatdoesnotdependontherebeinganystrongdegreeofoverallsimilaritybetweenorganismsandartefacts.Accordingtothisreadingoftheargument,PaleyisusingtheLikelihoodPrincipletodiscriminatebetweentwohypotheses:OdOrganismswerecreatedbyanintelligentdesigner.OrOrganismswereproducedbyrandommindlessprocesses.Paleyclaimsthattheobservedadaptivecomplexityoforganismsisrenderedmuchmoreprobablebytheformerhypothesisthanbythelatter.Paleydiscussesthewatchontheheathtomakevividthepowerofthistypeofinferenceinanuncontroversialexample.Theadaptivecomplexityofthewatchfavoursthefirstofthefollowinghypothesesoverthesecond:WdThewatchwascreatedbyanintelligentdesigner.WrThewatchwasproducedbyrandommindlessprocesses.ThepointoftheanalogyisjusttoexplainhowtheLikelihoodPrincipleworks.Itdoesnotmatterhowsimilarordissimilarwatchesandorganismsareoverall,solongastheysharethesinglefeatureofadaptivecomplexity(Sober1993).WhilehestudiedatCambridge,DarwinreadPaleyandgreatlyadmiredPaley’sdiscussionofadaptivecontrivances.Darwineventuallycametorejecttheargumentfromdesign,andtomovetowardsagnosticism,becausehethoughthesawsomanyfeaturesoforganismsthatconflictwiththehypothesisthatorganismswerecreatedbyabenevolent,omniscientandomnipotentdesigner.Forexample,Darwinwasrevoltedbytheparasiticwaspsthatlaytheireggsonparalysedcaterpillars;whentheeggshatch,theyeattheirhostsalive,leavingthebrainuntillast(DesmondandMoore1991).Ifanintelligentdesignerbuiltsomuchsufferingintothelivingworld,thisdesignerdoesnotdeserveourreverence.Darwinsawthelivingworldassaturatedwithpainanddeath;thisiswhatoneshouldexpectaccordingtothehypothesisofevolutionbynatural331\nELLIOTTSOBERselection,butnotwhatoneshouldexpectaccordingtothehypothesisofintelligentdesignasDarwinunderstoodthathypothesis.ThisisaversionoftheARGUMENTFROMEVIL(pp.480–3).TheargumentisnotbestputbysayingthatthequantityofevilfoundintheworlddeductivelyentailsthatthereisnoGod;rather,theobservationsareclaimedtoplaythenon-deductiveroleoffavouringatheismovertheism.ThisDarwinianargumentisalikelihoodargument,justlikePaley’s,butDarwinconsidersahypothesisthatPaleydoesnot,andheconsidersanobservationthatdiffersfromtheoneonwhichPaleyfocused.Thehypothesisofevolutionbynaturalselectiondoesnotclaimthatlivingthingschangetheirfeaturesatrandom.Aprocessisrandomwhenitspossibleoutcomesareequiprobable,ornearlyso.Tossingacoin,spinningaroulettewheel,drawingcardsfromadeck–theseareexamplesofrandomprocesses.However,naturalselectionisaprocessinwhichsometraitshavemuchhigherproba-bilitiesofspreadingthanothers.Therulethatgovernsselectionprocessesisthatfittertraitstendtoincreaseinfrequencyandlessfittraitstendtodecline.InadditiontoelaboratingahypothesisthatPaleydidnotconsider,Darwinalsoemphasizedtheubiquityofadaptiveimperfections.Vestigialorgans,forexample,arenotwhatoneshouldexpecttofindiflivingthingswerecreatedfromscratchbyabenevolent,intelligentandpowerfulengineer.However,sincenaturalselectionworksonthearrayofvariationthatancestralpopulationshappentocontain,itisnosurprisethattheseprocessesleavetracesofancestralformsthathavenoadaptiverationale.PaleyanticipatedoneaspectofthisDarwinianargument.Hesaysthatweshouldinferawatchmakerwhenweseethewatch,evenifthewatchturnsouttobeimperfect.Paleyemphaticallydeniesthatimperfectadaptationundercutsthedesignargument.Tobesure,Paleydelightedinwhathetooktobethemanyperfectionsfoundinnature.However,itisclearthathedidnotthinkofthisaspartofhisargumentfortheexistenceofadesigner.Rather,theostensibleperfectionofnaturecomesinlater,asevidencethatbearsonthedesigner’scharacteristics.Adaptivecomplexity,evenwhenimperfect,provesthereisadesigner.Furtherdetailsthenshowthatthisdesignerisbenevolent.What,then,isthelogicalstatusofthedesignargument,andhowisDarwin’stheoryrelatedtoit?Darwinandhissuccessorshaveclaimedthatimperfectadaptationsfavourevolutionarytheoryoverintelligentdesign(see,forexample,Gould1980).Shouldtheyconcedethatperfectadaptationwouldfavourintelligentdesignoverpurelyrandommindlessprocesses?IsuggestthatthisinterpretationconcedestoomuchtoPaley.Evenaperfectwatchisnoevidenceofintelligentdesignunlesswehaveindependentlyattestedauxiliaryinformationabouttheintentionsandabilitiesthattheputativedesignerwouldhaveifheexisted.Whatprobabilitydoesthedesignhypothesisconferonthefeaturesofthewatchthatweobserve?Weobserve,forexample,thatthewatchismadeofmetalandglass.Howprobableisitthatanintelligentdesignerwouldusethesematerials?Ifintelligentdesignersneverusethesematerials,thedesignhypothe-siswillbelesslikelythanthehypothesisofrandommindlessprocesses.However,ifweassumethatintelligentdesignersoftenusemetalandglass,wereachtheoppositeverdict.Andifwesimplydonotknowwhattheputativeintelligentdesignerwouldbeinclinedtodo,wewillbeunabletocomparethelikelihoodsofthetwohypotheses.Paleyisrightthatwedonotandshouldnothesitatetoinferawatchmakerfromtheobservedfeaturesofthewatch;thisisbecauseweknowagreatdealabouttheincli-nationsandabilitiesthathumandesignershave.However,whenweshiftfromwatch332\nPHILOSOPHYOFBIOLOGYtoorganism,thesituationchanges.Whatdoweknowaboutthedesiresandabilitiesofthisputativedesigneroforganisms?Isuggestthatwedonotknowenoughtocomparethelikelihoodsofthetwohypotheses.Paley’sanalogybetweenwatchandorganisminfactconcealsadeepdisanalogy.Asaresult,thedesignargumentfailsonitsownterms;thereisnoneedtobringinDarwiniantheorytoseethis.Itisnottruethattheobservedfeaturesoforganismsputthedesignhypothesistothetestandthatwhatweobservedisconfirmsthehypothesis;rather,thepresentsuggestionisthatthedesignhypothesisisuntestable(Sober1993,1999d).Defendersofthedesignargumentarenotentitledtoassumewithoutargumentpropositionsthatdescribethefeaturesthatorganismswouldhaveiftheywereproducedbyanintelligentdesigner.Criticsoftheargumentarenotentitledtodothis,either.3.2Species,essentialismandhumannatureSpecieshavelongbeenafavouriteexamplethatphilosopherscitewhentheydiscussnaturalkinds.Forexample,MILL(chapter35)inhisSystemofLogic(1874)claimsthathumanbeingisanaturalkind,buttheclassofsnub-nosedindividualsisnot,onthegroundsthat‘Socratesisahumanbeing’allowsonetopredictmanyothercharacteristicsthatSocrateshas,but‘Socratesissnub-nosed’doesnot.Aristotelianessentialismburdenstheconceptofnaturalkindwithamoresubstantialcharacterization.Naturalkindsnotonlyhavepredictiverichness;inaddition,theyhaveESSENCES(pp.112–13).Theessenceofanaturalkindisthenecessaryandsufficientconditionthatmembersofthekindmustsatisfyinordertobemembers.Anindividual’spossessingthisessenceiswhatmakesitbelongtothekindinquestion.Inaddition,itisanecessarytruththatallandonlythemembersofthekindhavethisessentialproperty.Theessencealsodoesagooddealofexplanatorywork;thefactthatanindividualhasthisspecies-typicalessenceexplainsmanyotherfeaturesthattheindividualpossesses.Besidescitingbiologicalspeciesasexamples,philosophersoftenpointtothechemi-calelementsasparadigmnaturalkinds.Goldisakindofsubstance;itsessenceissaidtobetheatomicnumber79.Thisatomicnumberiswhatmakesalumpofmatteraninstanceofgold.Andatomicnumberexplainsmanyotherpropertiesthatgoldthingshave.Kripke(1972)andPutnam(1975)havesuggestedthatscienceisinthebusinessofempiricallydiscoveringtheessencesofnaturalkinds.Formulatedinthisway,essen-tialismisnotestablishedbytheexistenceoftrivialnecessarytruths.Itisanecessarytruththatallhumanbeingsarehumanbeings,butthisdoesnotentailthatthereisanessencethathumanbeingshave.Itisalsoimportanttoseparatetheclaimthatkindshaveessencesfromtheclaimthatindividualsinthekindhaveessentialproperties(Enç1986).Thefactthatgoldhasoneatomicnumberandleadanotherdoesnotentailthatanindividualcannotpersistthroughtimeasitchangesfrombeingmadeofleadtobeingmadeofgold.Theexampleofthechemicalelementsillustratesafurtherfeaturethatkindessencesaresupposedtohave.Gold’sessenceispurelyqualitative;‘atomicnumber79’doesnotrefertoanyplace,timeorindividual.Inprinciple,goldcouldexistatanyplaceoratanytime.Whatmakestwothingsmembersofthesamenaturalkindisthattheyaresimilarintherequisiterespect.Thereisnorequirementthattheybecausallyorspatio-temporallyrelatedtoeachotherinanyway.333\nELLIOTTSOBERAlthoughphilosopherswhoacceptthisessentialistpictureofthechemicalelementsusuallythinkthatchemistryhasalreadydiscoveredtheessencesthatvariouschemi-calkindspossess,theymustadmitthatbiologyhastodatenotdeliveredthegoodswithrespecttobiologicalspecies.Ifbiologicalspecieshaveessencesandtheseessencesarenotbeyondthekenofscience,thentheessentialistmustclaimthatbiologywilleven-tuallyrevealwhatthesespeciesessencesare.However,therearestrongreasonstothinkthatDarwin’stheoryofevolutionunderminesthisessentialistpictureofbiologicalspecies(Hull1965;Mayr1975;Sober1980).Itisnotjustthatbiologyhasnotdis-coveredthesespeciesessencesyet;rather,thewayspeciesareconceptualizedinevolu-tionarytheorysuggeststhatspeciessimplydonothaveessences.Theyarenotnaturalkinds,atleastnotontheusualessentialistconstrualofwhatanaturalkindis.Thereasonsforthisconclusionneedtobestatedcarefully.Thefactthatspeciesevolveis,perse,notaconclusiveargumentagainstessentialism.Justastheessential-istcanagreethatleadmightbetransmutedintogold,sotheessentialistcanagreethatalineagemightbetransformedfromonespeciesintoanother.Andthefactthattherearevagueboundariesbetweenspeciesisnot,initself,arefutationofessentialism,either.Ifapieceofmatteristransformedfromleadtogold,perhapstherewillbeintermediatestagesoftheprocessinwhichitisindeterminatewhetherthematterbelongstoonenaturalkindortheother(Sober1980).Unfortunately,therestillisdisagreementinevolutionarybiologyabouthowthespeciescategoryshouldbeunderstood.PerhapsthemostpopulardefinitionisMayr’s(1963,1970)biologicalspeciesconcept(forothers,seeEreshefsky1992).Itsanti-essentialistconsequencesaretoalargedegreealsotheconsequencesthatotherspeciesconceptshave,sowemayexamineitasanillustrativeexample.Mayr’sbasicideaisthatabiologicalspeciesisanensembleoflocalpopulationsthatareknittogetherbygeneflow.Theindividualswithinlocalpopulationsreproducewitheachother.Andmigrationbetweenlocalpopulationsmeansthatthereisreproductionbetweenindividualsindifferentpopulationsaswell.Thissystemofpopulationsisreproductivelyisolatedfromothersuchsystems.Reproductiveisolationcanbeasimpleconsequenceofgeographicalbarriers,oritcanmeanthattheorganismshavebehaviouralorphysiologicalfeaturesthatpreventthemfromproducingviablefertileoffspringevenwhentheyarebroughttogether.Aconsequenceofreproductiveisolationisthattwospeciescanevolvedifferentcharacteristicsinresponsetothedifferentselectionpressuresimposedbytheirdifferentenvironments.However,thedifferentphenotypesthatevolvearenotwhatmakethetwospeciestwo;itisreproductiveisolation,notphysicaldissimilarity,thatisdefinitive.Mayr(1963)initiallyallowedtwopopulationstobelongtothesamespeciesiftherewasactualorpotentialinterbreedingbetweenthem,buthelaterchangedthedefini-tionsothatactualinterbreedingwasrequired(Mayr1970).Thisraisesthequestionofwhatthetimescaleisonwhichinterbreedingmusttakeplace.Howoftenmustindi-vidualsindifferentlocalpopulationsreproducewitheachotherforthetwopopulationstobelongtothesamespecies?Indeed,thesamequestioncanbeposedaboutindivi-dualslivinginthesamelocalpopulation.Anotherdetailthatneedstobeaddedtothesketchjustprovidedconcernsindividualsthatexistatdifferenttimes.Humanbeingswhoarealivenowarenothavingbabieswithhumanbeingswholivedthousandsofyearsago.Reproductionissomethingthatoccursbetweencontemporaneous334\nPHILOSOPHYOFBIOLOGYindividuals.Sowhatmakeshumanbeingsnowandhumanbeingsthousandsofyearsagomembersofthesamespecies?Anecessaryconditionisthathumanbeingsnowaredescendedfromhumanbeingsathousandyearsago.Butthisisclearlynotsufficient;otherwise,apresent-dayspeciescouldnotbedescendedfromadifferent,ancestral,species.Finally,IshouldnotethatMayr’sdefinitionexcludesthepossibilityofasexualspecies,andthisisanotherofitsfeaturesthathasmadeitcontroversial.ThethingtonoticeaboutMayr’sdefinitionisthatqualitativesimilarityisneithernecessarynorsufficientforconspecificity.Membersofthesamespeciesmayhaveverydifferentcharacteristics.Andifcreaturesjustliketigersevolvedindependentlyinothergalaxies,theywouldnotbelongtothespeciestowhichearthlytigersbelong.Whatmakesforconspecificityarethecausalandhistoricalconnectionsthatarisefromrepro-ductiveinteractions.Biologicalspeciesandchemicalelementsareverydifferentinthisregard.Evolutionarybiologiststalkaboutspeciesinthesamewaytheydiscussindividualorganisms.Justasindividualorganismsbeargenealogicalrelationshipstoeachother,sospeciesaregenealogicallyrelated.Justasindividualorganismsareborn,developanddie,soindividualspeciescomeintoexistence,evolveandgoextinct.TheseconsiderationsledHull(1978,1988)todevelopGhiselin’s(1974)suggestionthatspeciesareindividuals,notnaturalkinds.Thereisroomtodoubt,however,thatspeciesareasfunctionallyintegratedasindividualorganismsoftenare.Thepartsofatigerdependoneachotherforsurvival;exciseanarbitrary30percentofatiger,andthewholetigerdies.However,theextinctionof30percentofaspeciesrarelycausestherestofthespeciestogoextinct.Thissuggeststhatindividuality(inthesenseoffunctionalinterdependenceofparts)comesindegrees,andthatspeciesareoftenlessindividualisticthanorganismsoftenare.Still,HullandGhiselin’smainthesisremains;perhapsitshouldbestatedbysayingthatspeciesarehistoricalentities(Wiley1981).Similarpointsapplytohighertaxonomiccategories.Althoughordinarylanguagemaysuggestthatcarnivoresalleatmeatandthatmammalsallnursetheiryoung,thisisnothowbiologistsunderstandCarnivoraandMammaliaastaxa.Thesetaxaareunderstoodgenealogically;theyaremonophyleticgroups,meaningthattheyincludeanancestralspeciesandallofitsdescendants(Sober1992).PandasbelongtoCarnivorabecausetheyaredescendedfromotherspeciesthatbelongtoCarnivora;thefactthatpandasarevegetariansdoesnotmatter.Superspecifictaxa,likespeciesthemselves,areconceptualizedasbigphysicalobjects;theyarechunksofthegenealogicalnexus.Andjustasspeciesareoftennotveryindividualistic,superspecifictaxaareevenlessso(Ereshefsky1991).Thechemicalkindsdonotcompriseanadhoclist.Rather,thereisatheory,codifiedintheperiodictableofelements,thattellsushowtoenumeratethesechemicalkindsandhowtheyaresystematicallyrelatedtoeachother.Tosaywhatthechemicalkindsare,wecanjustconsultthistheory;wedonot,inaddition,havetogooutanddofieldwork.Nosuchtheoryexistsinbiologyforspeciesandhighertaxa;fieldworkisbasicallytheonlymethodthatsciencehasforassemblingalistofspecies.Theterms‘botanizing’and‘beetlecollecting’bothalludetothisfeatureofsystematicbiology.Speciesandhighertaxaarethingsthathappentocomeintoexistenceowingtothevagariesofwhattranspiresinthebranchingtreeoflife.335\nELLIOTTSOBERItdoesnotfollowthattherearenonaturalkindsinevolutionarybiology.Perhapssexualreproductionisakind;perhapsbeingapredatorisanother.Whatmakesittruethattwoorganismseachreproducesexually,orthatbotharepredators,isthattheyaresimilarinsomerespect;itisnotrequiredthattheybehistoricallyconnectedtoeachother.Thesexualspeciesdonotformamonophyleticgroup,andneitherdothepreda-tors.Thesekindtermsappearinmodelsofdifferentevolutionaryprocesses;therearemodelsthatexplainwhysexmightevolve,andmodelsthatdescribethedynamicsofpredator–preyinteractions.AlthoughDarwin’stheoryofevolutionunderminesessen-tialistinterpretationsofspeciesandhighertaxa,itisanothermatterwhetheressen-tialismistherightwaytounderstandtheseother,non-taxonomic,theoreticalconcepts.Ifbiologicalspecieslackessences,thenHomosapienslacksanessence.Iftheessen-tialistisrightaboutchemicalkinds,thenitmakessensetoaskwhatthenatureofgoldis.Butwhatcoulditmeantotalkabout‘humannature’?Takenliterally,theideaofhumannatureistheideathatthereisanessencethatallhumanbeings,andonlythey,possess.ThisthoughtisattherootofARISTOTLE’s(chapter23)ethicaltheory.Italsoisattherootofmanynormativeclaimstotheeffectthatthisorthathumanbehaviouris‘unnatural’.Thispeculiarexpressioncanbegivenasensiblebiologicalreading,if‘natural’issimplyequatedwiththeideaofbeingfoundinnature.Understoodinthisway,everythingthathumanbeingsdo,boththegoodandthebad,isnatural.However,when‘natural’and‘unnatural’arenotusedinthisway,itisimportanttodemandthatanexplanationbegivenofwhatthetermsaresupposedtomean.Normativeethicalclaimsshouldnotbepermittedtomasqueradeasbiology.3.3EvolutionaryepistemologyEvolutionaryepistemologyisafielddividedintotworesearchprogrammes(Bradie1994).First,thereareattemptstouseideasfromevolutionarybiologytoexplainvariousfeaturesofhumancognition.Second,thereistheattempttodeveloptheoriesofculturalandscientificchangethatareanalogoustotheoriesofbiologicalevolution.Theselinesofinquiryareindependentofeachother,sothatthesuccessorfailureoftheonedoesnotautomaticallyentailthesuccessorfailureoftheother.Noriseitherresearchprogrammeparticularlyunified;thismeans,forexample,thatoneevolution-arymodelofscientificchangemightfailmiserably,whileanothersucceedsadmirably.Sociobiology(Wilson1975)andevolutionarypsychology(seethepaperscollectedinBarkow,CosmidesandTooby1992)provideexamplesofthefirsttypeofevolutionaryepistemology.Sociobiologistsusuallyfocusondifferentfeaturesofhumanbehaviourandseektoexplainthemasadaptiveresponsestoancestralenvironments.Thedebateaboutadaptationism,discussedearlier,grewoutofthevociferouscontroversythatengulfedsociobiology.Criticsofsociobiologybelievedthatsociobiology’smethodologi-caldefectswerejustthetipoftheiceberg;therealproblem,theythought,wasthepervasiveinfluenceofnaiveadaptationisminevolutionarybiologyasawhole(GouldandLewontin1979).Evolutionarypsychologydiffersfromsociobiology,notbybeinglessadaptationist,butbyshiftingitsfocusfrombehaviourtocognitivemechanismsastheitemsrequiringevolutionaryexplanation.Evolutionarypsychologiststendtothinkofthemindasahighlymodularizedsetofdevices,eachhavingevolvedasasolutiontoadifferentadaptiveproblem;theyalsotendtothinkthatcognitiveadaptationsvaryvery336\nPHILOSOPHYOFBIOLOGYlittleamonghumanbeings.Boththemodularityhypothesisandtheuniversalityhypothesishavebeenquestioned(SterelnyandGriffiths1999;Wilson1994).Thisfirsttypeofevolutionaryepistemologyhasalsobeenpursuedbyphilosophers.Forexample,therehasbeenconsiderabledebateaboutwhether,orinwhatcircum-stances,naturalselectioncanbeexpectedtofavourbeliefacquisitiondevicesthatarereliable–thatis,thatproducemostlytruebeliefs(Stephens2001).Therealsohasbeeninvestigationofwhennaturalselectionwillfavourinnateandinflexiblecharacteristics,andwhenitwillfavourtraitsthatarelearnedandplastic(Godfrey-Smith1991,1996;Sober1994b).Godfrey-Smith(1996)alsoexploresthehypothesisthatmentalityevolvedasaresponsetoenvironmentalcomplexity(Sober1997b).Examplesofthesecondkindofevolutionaryepistemologymaybefoundinpro-posalsdevelopedbyPopper(1973),Campbell(1974)andHull(1988),whichviewscientificchangeastheresultofacompetitionamongtheories,withthefittesttheorysurviving.Thisdoesnotmeanthattheoriessurvivebecausetheyenablethosewhobelievethemtohavemorebabies;nordoevolutionaryepistemologistsproposethatscientifictheoriesarepassedfromindividualtoindividualbygenetictransmission.Whereasmodelsofbiologicalevolutionstandardlymeasurefitnessintermsofrepro-ductiveoutputandtreatgenesasthemechanismofinheritance,evolutionarymodelsofscientificchangereplacegenetictransmissionwithteachingandlearning,andthinkoffitnessasanidea’spropensitytospread.Fitideasareattractive,regardlessofwhethertheyaffectbiologicalsurvivalandreproduction(Sober1993).Therearesomeobviousdisanalogiesbetweenbiologicalmodelsofevolutionbynaturalselectionandselectionistmodelsofscientificchange.Foronething,novelbiologicalvariationarisesbyrandommutation;thismeansthatwhatcausesamutationtooccurhasnothingtodowithwhetherthemutationwillbeuseful.Incontrast,itseemsclearthatnovelscientificideasareofteninventedwiththegoalofbeinguseful;scientistsdon’tmakeuphypothesesatrandom.However,thisandotherdisanalogiesdonotunderminetheclaimthattheideasinascientificcommunitycompetewitheachother;andthatthereisaselectionprocessinwhichsomeideasspreadwhileothersdisappearfromthescene.Thus,itseemsclearthatscientificchangecanbedescribedasaselectionprocess.Thesubstantivetaskistosayhowtheconceptoffitnessshouldbeunderstoodinthiscontext.Whatmakesoneideamoreattractivethananotherinascientificcommunity?Surelytherearemanyfactorsthatcaninfluenceanidea’sattractiveness,andthemixoffeaturesthatmakeanideaattractiveinonescientificcontextmaydifferfromthemixthatmattersinanother.Perhapssomeepisodesofscientificchangearedrivenlargelybyobservationalevidence,whileothersoccurbecauseofreligious,metaphysicalorpoliticalcommitments.Evolutionaryepistemologistswhowanttomodelscientificchangeasaselectionprocessmustaddressthequestionsthathistoriansofsciencedebatewhentheyconsider‘internal’versus‘external’factors.Merelysayingthatscientificchangeoccursbyaselectionprocessisnotenough.Anothervarietyofevolutionaryepistemologymaybefoundinmodelsofculturalgroupselection(BoydandRicherson1985).AsDarwinobservedinhisdiscussionofhumanmorality,groupscompetewitheachother,andgroupswithmoreadaptiveideologieswillout-competegroupswhoseideasarelessfit.Again,thereisnorequire-mentthattheseideationalelementsaretransmittedgenetically;groupsmayvarysimplybecausetheindividualsinonegenerationfaithfullyteachtheirmorestothenext.Group337\nELLIOTTSOBERcompetitionmaymeanthatpeopleinonegroupkillthemembersofothers,oritmaysimplymeanthattheconquerorsimposetheirmoresontheconquered.Indeed,thespreadofideasmaynotinvolve‘conquest’atall,ifpeopleinonegroupfreelyadoptideasfromanother.3.4EvolutionaryethicsHUME’s(chapter31)distinctionbetweenisandought–betweenclaimsthatarepurelydescriptiveandclaimsthatarenormative–mustbeborneinmindwhenweaskwhatrelevanceevolutionarytheoryhasforquestionsaboutmorality.Darwinsuggestedthatevolutionarytheoryhelpsusunderstandwhyhumanmoralityhassomeofthefeaturesthatithas.Itremainstobeseenhowmuchofthecontentofhumanmoralitycanbeexplainedinthisway,andhowmuchshouldbethoughtofasarisingbynon-adaptiveculturalprocesses.Forexample,someculturesimposerestrictivenormsonthecloth-ingthatmenandwomenmaywear,whileothersencouragepersonalinnovation.Evenifevolutionarytheoryhasnothingtosayaboutwhythispatternofvariationexists,itispossiblethatthetheoryhasanimportantbearingonotherpatterns(SoberandWilson1998).Whatcanevolutionarytheorytellusaboutwhichnormativeethicalclaims,ifany,aretrue?AlthoughmanysocialDarwiniststhoughtthatevolutionaryoutcomesarealwaysgood(becausetheythoughtthatevolutionisinherently‘progressive’),therehasalsobeenastrongtraditionthatmaintainsthatevolutionaryoutcomesareoftenmorallydeplorable.JustasDarwindidnotrejoiceinthebehaviourofparasiticwasps,soHuxley(1997)thoughtthatthepointofmoralitywastocombattheinstinctsweinherit;morerecently,andinthesamevein,G.C.Williams(1989)hassuggestedthatnatureisa‘wickedoldwitch’.Anotherlineofthinkinginevolutionaryethicsholdsthatevolutionaryconsidera-tionsshowthattherecanbenoobjectivenormativetruths.Theclaimisnotjustthatthemoralprincipleswenowhappentoholdaremistaken;theclaimisstronger–thatnomoralconvictionscanbetrue(RuseandWilson1986).Evolutionhascreatedinustheillusionthatthereareobjectivemoralstandards;thereisanadaptiveadvantageinbelievingthisfiction.Inreply,itshouldbenotedthatthemerefactthatourmoralbeliefsareproducedbyevolutiondoesnotshowthattherearenomoraltruths.Afterall,oursimplemathematicalbeliefsmaybetheresultofevolution,butthatdoesnotshowthattherearenomathematicaltruths(Kitcher1994).However,moreneedstobesaidhere,sincethesubjectivistdoesnothavetomaintainthatthenon-existenceofmoraltruthsdeductivelyfollowsfromtheclaimthatthemoralityweacceptistheproductofevolution.Abetterformulationofsubjectivismmaintainsthatitisexceedinglyimprobablethatthemoralityweacceptistrue,giventhatitisproducedbyevolution.Afterall,naturalselectioncausestraitstoevolvebecausetheyprovidereproductivebenefits.Whyshouldaprocessbentonmaximizingreproductionleadustomoralbeliefsthataretrue?Thethingtonoticeaboutthislineofthoughtisthatitdoesnotentailthattherearenomoraltruths;rather,itadvancesthemoremodesthypothesisthatthemoralbeliefswepresentlyhavearenottrue.Inaddition,thisargumentfailstoattendtothefactthatevolutionhasgivenhumanbeingsthecapacitytoreason;ifthiscapacitycanleadus338\nPHILOSOPHYOFBIOLOGYtodiscovertheoreticaltruthsinsciencethatprovidenopracticalbenefitintermsofsurvivalandreproduction,whyshouldreasonnotalsobeabletoleadustodiscoveriesaboutthegoodandtheright?Perhapstherearereasonstodoubtthis,butthemerefactthatnaturalselectionaimstomaximizereproductionshouldnotleadustoembraceethicalsubjectivism.Aslightlydifferentargumentforethicalsubjectivismmaintainsthatwecanprovideafullysatisfactoryexplanationofwhywebehavethewaywedo,andalsoexplainwhyweacceptthemoralprincipleswedo,withoutpostulatingarealmofethicaltruths(Harman1977;RuseandWilson1986).Objectivismaboutethicsisunparsimonious;asimplerexplanationofhumanthoughtandbehaviourtracesthesephenotypicfeaturesbacktothesocialcontextofhumancultureandthebiologicalcontextofhumanevolution.Therightreplytothisargumentforsubjectivismisthatnormativeethicalprinciplesarenotinthebusinessofexplainingwhypeoplethinkandactastheydo.Thepointofmoralityistoregulatebehaviour,nottoexplainit.Onemightjustaswellarguethatobjectiveepistemologicalnormsdonotexistbecausetheyarenotneededtoexplainwhypeoplethinkastheydo.Psychologyisinthebusinessofexplaininghowpeopleinfactthink;epistemology,however,isinanormativelineofwork(Sober1993).AcknowledgementsIamgratefultoDavidHullandChristopherStephensfortheirhelpinthepreparationofthischapter.ReferencesAllen,C.,Bekoff,M.andLauder,G.(eds)1998:Nature’sPurposes:AnalysesofFunctionandDesigninBiology.Cambridge,MA:MITPress.Anscombe,E.1975:CausalityandDetermination.InE.Sosa(ed.)CausationandConditionals.Oxford:OxfordUniversityPress.Barkow,J.,Cosmides,L.andTooby,J.(eds)1992:TheAdaptedMind:EvolutionaryPsychologyandtheGenerationofCulture.Oxford:OxfordUniversityPress.Beatty,L.1995:TheEvolutionaryContingencyThesis.InG.WoltersandJ.Lennox(eds)Concepts,Theories,andRationalityintheBiologicalSciences:TheSecondPittsburgh–KonstanzColloquiuminthePhilosophyofScience.Pittsburgh:UniversityofPittsburghPress.Boorse,C.1976:WrightonFunctions.PhilosophicalReview,85,70–86.Boyd,R.andRicherson,P.1985:CultureandtheEvolutionaryProcess.Chicago:UniversityofChicagoPress.Bradie,M.1994:EpistemologyfromanEvolutionaryPointofView.InE.Sober(ed.)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.Brandon,R.1978:AdaptationandEvolutionaryTheory.StudiesintheHistoryandPhilosophyofScience,9,181–206.——1990:AdaptationandEnvironment.Princeton,NJ:PrincetonUniversityPress.——1996:ConceptsandMethodsinEvolutionaryBiology.Cambridge:CambridgeUniversityPress.Buller,D.(ed.)1999:Function,SelectionandDesign.SeriesinPhilosophyandBiology.NewYork:StateUniversityofNewYorkPress.Campbell,D.1974:EvolutionaryEpistemology.InP.Schilpp(ed.)ThePhilosophyofKarlPopper.LaSalle,IN:OpenCourt.339\nELLIOTTSOBERCummins,R.1975:FunctionalAnalysis.JournalofPhilosophy,72,741–64.ReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.Darwin,C.1981[1871]:TheDescentofMan,andSelectioninRelationtoSex.Princeton,NJ:PrincetonUniversityPress.Dawkins,R.1976:TheSelfishGene,2ndedn.Oxford:OxfordUniversityPress.Desmond,A.andMoore,J.1991:Darwin–theLifeofaTormentedEvolutionist.NewYork:TimesWarner.Dobzhansky,T.1973:NothinginBiologyMakesSenseExceptintheLightofEvolution.AmericanBiologyTeacher,35,125–9.Dupré,J.1993.TheDisorderofThings.Cambridge,MA:HarvardUniversityPress.Edwards,A.andCavalli-Sforza,L.1964:ReconstructionofEvolutionaryTrees.InV.HeywoodandJ.McNeill(eds)PheneticandPhylogeneticClassification.NewYork:SystematicsAssociationPublication,6,67–76.Eldredge,N.andCracraft,J.1980:PhylogeneticPatternsandtheEvolutionaryProcess.NewYork:ColumbiaUniversityPress.Enç,B.1986:EssentialismWithoutIndividualEssences:Causation,Kinds,SupervenienceandRestrictedIdentities.MidwestStudiesinPhilosophy,11,403–26.Ereshefsky,M.1991:Species,HigherTaxa,andtheUnitsofEvolution.PhilosophyofScience,58,84–101.——(ed.)1992:TheUnitsofEvolution:EssaysontheNatureofSpecies.Cambridge,MA:MITPress.Fisher,R.1957[1930]:TheGeneticalTheoryofNaturalSelection,2ndedn.NewYork:DoverBooks.FoxKeller,E.andLloyd,E.(eds)1992:KeywordsinEvolutionaryBiology.Cambridge,MA:HarvardUniversityPress.Ghiselin,M.1974:ARadicalSolutiontotheSpeciesProblem.SystematicZoology,23,536–44.Godfrey-Smith,P.1991:Signal,Decision,Action.JournalofPhilosophy,88(12),709–22.——1993:Functions:ConsensusWithoutUnity.PacificPhilosophicalQuarterly,74,196–208.——1994:AModernHistoryTheoryofFunction.Nous,28(3),344–62.——1996:ComplexityandtheFunctionofMindinNature.Cambridge:CambridgeUniversityPress.Gould,S.1980:ThePanda’sThumb.Harmondsworth:PenguinBooks.——1989:WonderfulLife.NewYork:W.W.Norton.Gould,S.andLewontin,R.1979:TheSpandrelsofSanMarcoandthePanglossianParadigm:ACritiqueoftheAdaptationistProgramme.ProceedingsoftheRoyalSocietyofLondonB,205,581–98.ReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.Hamilton,W.1967:ExtraordinarySexRatios.Science,156,477–88.Harman,G.1977:TheNatureofMorality:AnIntroductiontoEthics.Oxford:OxfordUniversityPress.Hennig,W.1966:PhylogeneticSystematics.Urbana:UniversityofIllinoisPress.(RevisionandtranslationofHennig’s1950GrundzugeeinerTheoriederphylogenetishenSystematik.)Hull,D.1965:TheEffectofEssentialismonTaxonomy:2000YearsofStasis.BritishJournalforthePhilosophyofScience,15,314–26;16,1–18.——1974:PhilosophyofBiologicalSciences.EnglewoodCliffs,NJ:Prentice-Hall.——1978:AMatterofIndividuality.PhilosophyofScience,45,335–60.ReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.——1988:ScienceasaProcess.Chicago:UniversityofChicagoPress.Hull,D.andRuse,M.(eds)1998:PhilosophyofBiology.Oxford:OxfordUniversityPress.Hume,D.1947[1779]:DialoguesConcerningNaturalReligion.NewYork:ThomasNelsonandSons.340\nPHILOSOPHYOFBIOLOGYHuxley,T.1997[1893]:EvolutionandEthics.InA.Barr(ed.)TheMajorProseofThomasHenryHuxley.Athens,GA:UniversityofGeorgiaPress.Kitcher,P.1984:1953andAllThat:ATaleofTwoSciences.PhilosophicalReview,93,335–73.ReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.——1993:TheAdvancementofScience.NewYork:OxfordUniversityPress.——1994:FourWaysofBiologicizingEthics.InE.Sober(ed.)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.Kripke,S.1972:NamingandNecessity.InG.HarmanandD.Davidson(eds)TheSemanticsofNaturalLanguage.Dordrecht:Reidel.Lakatos,I.1978:FalsificationandtheMethodologyofScientificResearchProgrammes.InTheMethodologyofScientificResearchProgrammes:PhilosophicalPapers,vol.1.Cambridge:CambridgeUniversityPress.Lewontin,R.1970:TheUnitsofSelection.AnnualReviewofEcologyandSystematics,1,1–14.Lloyd,E.1988:TheStructureandConfirmationofEvolutionaryTheory.Westport,CT:GreenwoodPress.Maynard-Smith,J.1988:EvolutionaryProgress.InM.Nitecki(ed.)EvolutionaryProgress.Chicago:UniversityofChicagoPress.Mayr,E.1963:AnimalSpeciesandEvolution.Cambridge,MA:HarvardUniversityPress.——1970:Populations,Species,andEvolution.Cambridge,MA:HarvardUniversityPress.——1975:TypologicalversusPopulationalThinking.InEvolutionandtheDiversityofLife.Cambridge,MA:HarvardUniversityPress.ReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.Mill,J.S.1874:ASystemofLogic,8thedn.HarperandBrothers.Mills,S.andBeatty,J.1979:ThePropensityInterpretationofFitness.PhilosophyofScience,46,263–88.ReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.Mitchell,W.andValone,T.1990:TheOptimizationResearchProgram:StudyingAdaptationsbytheirFunction.QuarterlyReviewofBiology,65,43–52.Nagel,E.1961:TheStructureofScience.London:RoutledgeandKeganPaul.Oppenheim,P.andPutnam,H.1958:UnityofScienceasaWorkingHypothesis.InH.Feigl,G.MaxwellandM.Scriven(eds)MinnesotaStudiesinthePhilosophyofScience,vol.2.Minneapolis:UniversityofMinnesotaPress.Orzack,S.andSober,E.1994:OptimalityModelsandtheLong-runTestofAdaptationism.AmericanNaturalist,143,361–80.——2001:Adaptationism,PhylogeneticInertia,andtheMethodofControlledComparisons.InAdaptationismandOptimality.Cambridge:CambridgeUniversityPress.Paley,W.1805:NaturalTheology.Rivington.Popper,K.1973:ObjectiveKnowledge.Oxford:ClarendonPress.Putnam,H.1975:TheMeaningofMeaning.InK.Gunderson(ed.)Language,MindandKnowledge:MinnesotaStudiesinthePhilosophyofScience,vol.7.Minneapolis:UniversityofMinnesotaPress.Rosenberg,A.1978:TheSupervenienceofBiologicalConcepts.PhilosophyofScience,45,368–86.——1985:TheStructureofBiologicalScience.Cambridge:CambridgeUniversityPress.——1994:InstrumentalBiology,ortheDisunityofScience.Chicago:UniversityofChicagoPress.Royall,R.1997:StatisticalEvidence–ALikelihoodParadigm.London:Chapman&Hall.Ruse,M.1973:ThePhilosophyofBiology.London:HutchinsonUniversityLibrary.Ruse,M.andWilson,E.1986:MoralPhilosophyasAppliedScience.Philosophy,61,173–92.341\nELLIOTTSOBERReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.Schaffner,1976:ReductionisminBiology:ProspectsandProblems.InR.S.Cohenetal.(eds)PSA1974.Dordrecht:Reidel.Sober,E1980:Evolution,PopulationThinking,andEssentialism.PhilosophyofScience,47,350–83.ReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.——1984:TheNatureofSelection.Cambridge,MA:MITPress.(Secondedition(1994)Chicago:UniversityofChicagoPress.)——1988:ReconstructingthePast:Parsimony,Evolution,andInference.Cambridge,MA:MITPress.——1990:PuttingtheFunctionBackIntoFunctionalism.InW.Lycan(ed.)MindandCognition:AReader.Oxford:Blackwell.——1992:Monophyly.InE.FoxKellerandE.Lloyd(eds)KeywordsinEvolutionaryBiology.Cambridge,MA:HarvardUniversityPress.——1993:PhilosophyofBiology.Boulder,CO:WestviewPress.(Secondedition1999.)——1994a:ProgressandDirectioninEvolution.InJ.CampbellandW.Schopf(eds)CreativeEvolution?!Boston:JonesandBartlett.——1994b:TheAdaptiveAdvantageofLearningandaprioriPrejudice.InFromaBiologicalPointofView.Cambridge,MA:CambridgeUniversityPress.——1997a:TwoOutbreaksofLawlessnessinRecentPhilosophyofBiology.InL.Darden(ed.)PSA1996,458–67.——1997b:IstheMindanAdaptationforCopingwithEnvironmentalComplexity?BiologyandPhilosophy,12,539–50.——1999a:PhysicalismfromaProbabilisticPointofView.PhilosophicalStudies,95(1–2),135–74.——1999b:TheMultipleRealizabilityArgumentAgainstReductionism.PhilosophyofScience,66(4).——1999c:ModusDarwin.BiologyandPhilosophy,14(2),253–78.——1999d:Testability.ProceedingsandAddressesoftheAmericanPhilosophicalAssociation,73(2),47–76.Sober,E.andLewontin,R.1982:Artifact,Cause,andGenicSelection.PhilosophyofScience,47,157–80.Sober,E.andWilson,D.1994:ACriticalReviewofPhilosophicalWorkontheUnitsofSelectionProblem.PhilosophyofScience,61,534–55.ReprintedinD.HullandM.Ruse(eds)PhilosophyofBiology.Oxford:OxfordUniversityPress.——1998:UntoOthers–theEvolutionandPsychologyofUnselfishBehaviour.Cambridge,MA:HarvardUniversityPress.Stephens,C.2001:WhenisitSelectivelyAdvantageoustoHaveTrueBeliefs?SandwichingtheBetterSafethanSorryArgument.PhilosophicalStudies,105,161–89.Sterelny,K.andGriffiths,P.1999:SexandDeath:AnIntroductiontothePhilosophyofBiology.Chicago:UniversityofChicagoPress.Sterelny,K.andKitcher,P.1988:TheReturnoftheGene.JournalofPhilosophy,85,339–61.Wade,M.1978:ACriticalReviewofModelsofGroupSelection.QuarterlyReviewofBiology,53,101–14.Waters,C.K.1990:WhytheAnti-reductionistConsensusWon’tSurvivetheCaseofClassicalMendelianGenetics.PSA1990,vol.1,125–39.ReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.Wiley,E.1981:Phylogenetics:TheTheoryandPracticeofPhylogeneticSystematics.London:JohnWiley.342\nPHILOSOPHYOFBIOLOGYWilliams,G.1966:AdaptationandNaturalSelection.Princeton,NJ:PrincetonUniversityPress.——1989:ASociobiologicalExpansionofEvolutionandEthics.InJ.ParadisandG.C.Williams(eds)T.H.HuxleyEvolutionandEthics:WithNewEssaysonitsVictorianandSociobiologicalContext.Princeton,NJ:PrincetonUniversityPress.——1992:NaturalSelection:Domains,LevelsandChallenges.Oxford:OxfordUniversityPress.Wilson,D.1980:TheNaturalSelectionofPopulationsandCommunities.MenloPark,CA:BenjaminCummings.——1994:AdaptiveGeneticVariationandHumanEvolutionaryPsychology.EthologyandSociobiology,15,219–35.Wilson,E.1975:Sociobiology:TheNewSynthesis.Cambridge,MA:HarvardUniversityPress.Wimsatt,W.1979:ReductionandReductionism.InP.AsquithandH.Kyburg(eds)CurrentResearchinPhilosophyofScience.PhilosophyofScienceAssociation,352–77.——1980:ReductionisticResearchStrategiesandtheirBiasesintheUnitsofSelectionControversy.InT.Nickles(ed.)ScientificDiscovery,vol.2.Dordrecht:Reidel.Wright,L.1973:Functions.PhilosophicalReview,82,139–68.ReprintedinE.Sober(ed.)(1994)ConceptualIssuesinEvolutionaryBiology,2ndedn.Cambridge,MA:MITPress.——1976:TeleologicalExplanations:AnEtiologicalAnalysisofGoalsandFunctions.Berkeley:UniversityofCaliforniaPress.DiscussionQuestions1Whatconstitutesbeingalive?2Howshouldweunderstandfitness?3Ifbiologicalpropertiessuperveneuponphysicalproperties,canbiologybereducedtophysics?4Isreductionismcompatiblewiththemultiplephysicalrealizabilityofbiologicalproperties?5Whatisrequiredforonetheorytobereducibletoanother?6Howshouldweunderstandtheroleoffunctioninbiology?7Mustwetakeaccountofourexplanatoryintereststodistinguishbetweenthefunctionandthefortuitousbenefitofanorgan?8ArebothNewton’stheoryofgravitationandDarwin’stheoryofevolutionscientific?9Howcanwedeterminewhichspeciesarecloselyrelatedandwhicharerelatedonlymoredistantly?10Whatistheroleofparsimonyinscientificreasoning?11Canweexplainwhynaturalselectiondoesnotalwaysresultinspecieshavingoptimaltraits?12Isnaturalselectionmerelyoneamongseveralimportantinfluencesontraitevolution,orisittheonlyimportantinfluence?13Whatisthedifferencebetweentestinghypothesesandtestingresearchprogrammes?14Whyinbiologyisitimportanttorecognizethathistoricaloriginandpresentadvantagearenotautomaticallylinked?15Domorethanoneunitofselectionhavearoleintheevolutionaryprocess?16Isthechoicebetweenindividualsandgroupsastheunitsofselectionamatterofadoptingaconventionratherthanamatterofansweringafactualquestionaboutthehistoryoflife?343\nELLIOTTSOBER17Arethereanylawsofevolution?18Howcanapriorimodelshaveaparttoplayinempiricalscience?19Doestheadaptivecomplexityoforganismsjustifytheargumentfromdesign?20Couldspecies,asunderstoodinevolutionarytheory,haveessences?21Whatmakeshumanbeingsnowandhumanbeingsthousandsofyearsagomembersofthesamespecies?22Arespeciesindividualsornaturalkinds?Whatcouldcountasnaturalkindsinevolutionarybiology?23Whatfollowsfromtheclaimthatspeciesarehistoricalentities?24Isevolutionarypsychologymoresecurelybasedthansociobiology?25Isbiologicalevolutionagoodmodelforexplainingsocialandculturalchange?26Whataretheconsequencesforphilosophyofacceptingamodularityaccountofthemind?27Cannaturalselectionhelptoexplaintheexistenceandcharacterofhumanmentality?28Doesanevolutionarytheoryofthecharacterofhumanmoralityhelpustoassessmoralclaims?29Doevolutionaryconsiderationsunderminetheclaimthatthereareobjectivemoraltruths?344\n11PhilosophyofMathematicsMARYTILESSincethetimeofancientGreece,mathematicshasbeenintimatelytiedtophilosophy,bothasamodelofknowledgeandasanobjectofphilosophicalreflection.Arenumbersreal?Whatisaproof?Ismathematicsmorecertainthanotherknowledge?Canfinitemindshaveknowledgeofinfinity?Howcanmathematicsapplytotheworld?Inthischapter,changingphilosophicalconceptionsofmathematics,changesinthehistoricalcontextofmathematicalthoughtandchangesinmathematicsitselfareexploredinrelationtoabasicquestion:howcantheoreticalreasoningaboutnon-concretemathe-maticalobjectsbebothsecureandsouseful?Platonic,AristotelianandKantianapproachestomathematicsareexaminedintheirownrightandtoplaceincontexttheproblemsandproposedsolutionsofthemajormodernschoolsoflogicism,formalismandintuitionism.Recentdevelopmentswhichdonotseekfoundationsformathematicsarealsoconsidered.Manyofthediscussionsofgreathistoricalfiguresinthisvolume,especiallyFREGEANDRUSSELL(chapter37),arerelevanttothepresentchapter.ReaderswillalsowishtoconsultchaptersonEPISTEMOLOGY(chapter1),METAPHYSICS(chapter2),PHILOSOPHYOFLANGUAGE(chapter3),PHILOSOPHYOFLOGIC(chapter4)andPHILOSOPHYOFSCIENCE(chapter9).IntroductionstothephilosophyofmathematicsoftenbeginwhereKörner’s(1960)influentialintroductionbegan,outliningthreepositions:logicism,formalismandin-tuitionism.Thesewerethethreecontendingschoolstoemergefromnineteenth-centurymathematicalmovestoproviderigorousfoundationsformathematicalanalysis(includinginfinitesimalcalculus).Theproblemforthephilosophyofmathematicshasbeenthat(1)theseseemedtorepresentallthereasonablepositionsavailable,and(2)inthelightofGödel’sincompletenesstheoremsandotherresultsprovedbyTuring,Church,SkolemandTarskiinthe1930s,neitherlogicismnorformalismseemedaphilosophicallyviableposition.Intuitionism,whilsthavingitsphilosophicalcredentialsintact,wasunacceptabletomostmathematiciansbecauseitinvolveddiscardingpartsofclassicallyacceptedmathematics.Thisapparentimpassepartlyexplainsthedeclineofinterestinthephilosophyofmathematicssincethefirstpartofthiscentury,whenforawhile,withtheworkofRussellandWhitehead,andtheViennaCirclelogicalpositivists,itseemedtooccupycentrestage.\nMARYTILESHaoWang,inhisperceptiveretrospectiveanalysisofthephilosophyofmathe-maticsofthisperiod(Wang1968),explainsthistrajectoryintermsofthewidermove-mentsofanalyticphilosophy,whichwereinitiallystronglyempiricistandwhichconsequentlyinheritedahostilitytowardabstractobjects.Mathematicspresentsasignificantchallengetoempiricism,aswasmadepainfullyevidentbytheprominenceofmathematicsinEinstein’srelativitytheoriesandinthedevelopmentofquantummechanics.Howisempiricismtogiveanadequateexplanationofthecertainty,clarity,universalityandapplicabilityofmathematics?Wang’sansweristhattheabovementioneddevelopmentsinlogicandthefoundationsofmathematicsshowthatanalyticempiricismcannotmeetthischallenge.Itisthereforeonlyasphilosophyingeneralbeginstomovebeyondanalyticempiricismthatmorefruitfulapproachestophilosophyofmathematicscanbegintobeexplored.Buthowexactlyissuchamovetobemade?Becausethethreefoundationalschoolsoriginatedintheworkofphilosophicallymindedmathematicianswhowereseekingtoresolvesophisticatedconceptualproblemsthathadariseninthecontextofongoingmathematicalresearch,muchoftheworkinfoundationsisitselftechnicalandmathe-maticalincharacter.Thereisnosharplinewheremathematicsendsandphilosophybegins.Indeed,oneofthestrikingfeaturesofthedevelopmentofmathematicsinthetwentiethcenturywasthewayinwhichattemptstosolvephilosophicalandcon-ceptualproblemsgeneratednewmathematics.Thusfromthestandpointoftheearlytwenty-firstcentury,workinthefoundationsofmathematicscanbynomeansbeignored,butitmayneedtobere-evaluated.Whatisitsstatus?Whatexactlydoesittellusaboutcontemporarymathematics?Thesequestionsmustnowbeaddedtothephilo-sophicalagenda.Sooneofthewaysofmovingbeyondthepresuppositionsbuiltintophilosophyofmathematicsbyanalyticempiricistsistorecontextualizetheirworkonlogicandfoundationsinthehopeofunderstandingitsmotivations,achievementsandshortcomings.Afirststepistosketch,usingbroadstrokes,majorfeaturesintheland-scapeofWesternphilosophicalreflectiononmathematics,manyofwhichwerealreadyputinplacebytheancientGreeks.1BasicTasksforaPhilosophyofMathematicsThebasicpuzzle,ofwhichthechallengetoempiricismisbutaspecializedvariant,istoexplainhowtheoreticalreasoningabouthighlyabstractobjectscanbesouseful.Thefeaturesofmathematicswhichseemedtomarkitofffromotherdisciplinesarethatitsobjectsarenotencounteredintheworldofsenseexperience,andyet,perhapsasacon-sequence,mathematicalclaimscanbeprovidedwithproofswhichseemtoestablishthemwithanexactnessandcertaintyunparalleledinotherbranchesofknowledge.Buthowisitthatpeopleshutawayfromtheoutsideworld,thinking,calculatingandprovingtheoremsaboutthingswhichwecanneverhopetoseeortouch(numberssuchas2,1million,p,orstructuressuchascircles,ringsandvectorspaces),produceresultswhichother,verypracticalpeoplefindsousefulintheirdealingsbothwiththeevery-dayworldandwiththeesotericworldsofhigh-techscience?Whatrighthavewetobesoconfidentofthecorrectnessoftheseresults?346\nPHILOSOPHYOFMATHEMATICSAExxOOBCDFigure11.1Proofthattheinternalanglesofatriangleadduptothesumoftworightangles.Thestraightforwardanswertothislastquestionisthatmathematiciansareabletoofferproofsoftheirresults.Forexample,asgeometersbeforethetimeofAristotlereal-ized,measurementsoftheinternalanglesofplanetrianglesdonotyieldoneexactresult,butresultsthatclusterroundtworightangles,and,becausemeasuringgivesnoinsightintowhythisclusteringshouldoccur,itaffordsnoassurancethata‘rogue’trianglemightnotsomedaybefoundwhoseinternalangleshadamarkedlydifferentsum.Aproof,ontheotherhand,thatshowswhy,givenwhatitistobeatriangle,itsinternalanglesmustadduptotworightangles,assuresusthatnoroguetriangleexists.Onesuchproofisshowninfigure11.1.Suchaproofispossibleonlyifitcanbeassumedthatallthelinesinvolvedareper-fectlystraightandhavenothickness(somethingwhichisnottrueofthelinesinthediagram).Lineswhichhavethicknessdonotmeetinpointsandiftheyhavewrigglesinthemthereisnosensetobegiventoanexactmeasurementoftheanglebetweenthem.Soeventohaveinternalangleswhosesumisexactlyequaltotworightanglesispossibleonlyforaperfectplanetriangle,whichwecouldnotencounterinexperi-ence.Thetheoremprovedbelowwillholdmoreorlessexactlyforthosetrianglesweencountertotheextentthattheyaremoreorlessperfectlytriangular.Thereisanimportantassumptionmadehereabouttheapplicabilityofresultsprovedaboutidealizations:aresultprovedaboutanidealizationwillholdforsituationswhichapproximatetothatidealindirectproportiontotheclosenesswithwhichtheyapproximatetotheideal.Thisassumptionhastraditionallyunderwrittenapplicationsofmathematics,but,withrecentworkinchaostheory,ithasbeenshownnottobevalidforsystemsexhibitingsensitivedependenceoninitialconditions(Stewart1989:chs7,14).Thisworkopensupnewquestionsabouttheconnectionbetweenproofsandtheuseofresultsproved.Buthowcanwetellthatapurportedproofreallyisaproof?Allproofshavetostartfromsomeassumptions(weassumedabovesomepropertiesofparallellinesandofadditionofangles).Howcanwebecertainthatthesearecorrect?Furthermore,howcanwebesurethatallinferencesaremadecorrectly?Toaskforaproofforeveryassumptionwillleadtoaninfiniteregress.Soiftherearetobeanyproofstheremustbesomeself-evidentstarting-points,whichcanbeknowntobenecessarilytrueandwhichareneithercapableof,norstandinneedof,furtherjustification.347\nMARYTILESThismeansthatifmathematicsisadisciplineinwhichproofsarerequiredthenitmustbeonewhichexhibitsasystematichierarchicalorder.Somepropositionscon-cerningmathematicalentitieswillappearasfirstprinciples,acceptedwithoutproof:thesearethefoundationsuponwhichsubsequentknowledgemustbebuilt.Basicthe-oremsareprovedfromfirstprinciples:thesetheoremsareinturnusedtoprovefurtherresults,andsoon.Thisisthesenseinwhich,solongasmathematiciansdemandandprovideproofs,theirdisciplinemustnecessarilybeorganizedalonglinesapproximat-ingthepatternalreadytobefoundintheElementsofGeometryofEuclid(whichdatesfromaround300BC)–aworkwhichforcenturiesservedasanexemplar.Thequestionofwhatprinciplesofinferencearegenerallyvalid(rationallycom-pelling,ortruth-preserving)istraditionallyassignedtoLOGIC(chapter4).Butonemaywonderwhethertherearenotsomespecificallymathematicalprinciplesofinference.Take,forexample,Euclid’sprinciplethatifA=BandC=D,thenA+C=B+D.Isthisalogicalprinciple(aspecificapplicationoftheSUBSTITUTIVITYOFIDENTICALS(p.800))?Isitabasicprincipleofthetheoryofpartandwhole?Ifso,doesthisbelongtologicortomathematics?Orisitaspecificallymathematicalprinciplegoverningextensivemagnitudes?Atthispointitisalreadypossibletoseehowconcernforprinciplesofvaliddeduc-tiveinferenceandforthenatureandstatusoffirstprinciplesbecomesanaturalpartofanyphilosophyofmathematicsthattakesprooftobethatonthebasisofwhichmathematicalpropositionsgaintheirauraofprivilegedcertainty,ornecessarytruth.However,toequatethephilosophyofmathematicswithstudiesinlogicandfoundationswouldseemtoignoreonehalfofthecombinationoffeaturesthatmakemathematicssophilosophicallypuzzlingandintriguing.Itleavesoutquestionsconcerningtheutilityofmathematics.Sofarwehavethefollowingschemaofquestionsstructuringtheproblematicofphilosophicalreflectionsonmathematics:BasicProblemHowcantheoreticalreasoningaboutnon-concretemathematicalobjectsbebothsosecureandsouseful?Ifthequestionofsecurityispresumedtobeansweredbythefactthattheoreticalreasoningproducesproofsofmathematicalpropositions,thenwehavetodistinguishbetweenfirstprinciplesandresultsprovedfromthem,togetsomethinglikethefollowingsubsidiaryproblems:1Whatisthenatureandstatusofthefirstprinciplesofmathematics?1.1Canitsfirstprinciplesbeidentified?1.2Howcanthesefirstprinciplesbeknownwithcertainty?2Howismathematicalknowledgeacquired?2.1Discovery:Howarediscoveriesmade?Howareresultsestablishedorproofsfound?2.2Justification:Whatconstitutesaproof?Whatprinciplesofinferenceareavailable?Whatistheirnatureandwhatgroundshaveweforthinkingthemreliable?348\nPHILOSOPHYOFMATHEMATICS3Giventheanswersunder(1)and(2),howistheapplicabilityofprovedmathemat-icalresultstobeexplained?Dotheanswersunder(1)and(2)assureusthatwehavearighttoexpectmathematicalresultstohaveapplicationsandthatintheirapplicationestablishedresultswillbereliable?2BasicResponsesMathematicssincethesixteenthcenturyhasbeenahighlydynamicdiscipline.Overthisperiod,methodsofproblem-solvingandofproofhaveundergonemanychanges.Philosophershavetendedtotaketheabilitytoprovideproofstobethedistinguishingcharacteristicofmathematics.Butifproofmethodschangeandarelikelytocontinuetochange,philosophicanalysesofproofmethods,andopinionsaboutthefoundationsofmathematicalclaimsderivedfromtheseanalyses,willbehistoricallyconditioned.Similarly,thekindsofapplicationsmadeofmathematicsarecontinuallychanging.Anyaccountoftheapplicabilityofmathematicswillthusalsohavedifferentproblemstoaddressdependingontheperiodinwhichitisbeingoffered.Thismeansthataresponsetothebasicproblemwhichmaybeadequateatoneperiodofhistorymaybeinadequateatalaterperiodandsorequireconsiderablemodification.Thiscanbeillustratedbyseeinghowtheconfigurationofthebasicproblem,andhenceofviablesolutionstoit,changedintheseventeenthcenturyandthenagaininthenineteenthcentury.Inthiswaywecanindicate,albeitverysketchily,thehistoricalcontextofanalyticempiricistresponses.Forthepurposesofaschematicoverviewofthephilosophicalterrain,itispossibletoviewphilosophiesuptothelateseventeenthcenturyasproposingvariationsonwhatwemaycallPlatonicrealistorAristotelianconceptualistresponsestothebasicques-tion.(TheselabelsindicatetheclassicalancestryoftheseresponsesbutshouldnotbereadliterallyasPlato’srealismorAristotle’sconceptualism.)2.1PlatonicrealismForthePlatonicrealisttheproblemdissolvesonceitisrealizedthatthecosmoshasaneternalandunchangingunderlyingorder,orharmony,andthattheprinciplesofthisorderareessentiallymathematical.ThiskindofviewcanbetracedbacktothePythagoreanswhobasedavisionofaharmoniouslyordereduniverseontheirdis-coveryoftheassociationbetweenmusicalharmoniesandnumericalratios.ItfoundoneofitsmostexplicitexpressionsintheworkofKepler(1571–1630),thefamousastronomer.InhisHarmoniceMundi(1940)hesupposestheuniversetobetheworkofaGodwhoisnotonlyageometerbutalsoamusician.TheplanetsthereforemakeheavenlymusicastheyorbittheSun.Thesameconceptionsurfacesinmanyotherauthorsofvariousperiods.ForNewton(1642–1727)inhisOpticks(1931),thedivisionofthespectrumintosevencoloursisnumericallycorrelatedwiththeharmonicdivisionofasinglestringwhichyieldsthenotesoftheoctave.Einstein’s(1879–1955)often-quotedrejectionoftheclaimsofquantummechanicstobeafundamentalphysicaltheory–‘Goddoesnotplaydice’–349\nMARYTILESrevealsasimilarattitude.Italsofoundexpressioninthedrivetomathematicaleleganceandunitywhichresultedinhistheoryofgeneralrelativityandinhisunsuccessfulquestforagrandunifiedfieldtheory,aquestcontinuedbythoseworkingonthetheoryofsuperstrings(DaviesandBrown1988).Iftheuniverseismathematicallyordered,thiswouldexplaintheutilityofmathe-maticsbutnothowwecancometomathematicalknowledgeorthespeciallysecurestatusofthatknowledge.FortherealistwhowaseitherChristian,oraNEO-PLATONIST(p.595)(orsomeonewhowasboth),thiswasreadilyexplicable.TheChristianGodcreatednotonlytheuniverse,butalsohumanbeings;wewerecreatedwithanintel-lectthatisafiniteimageofGod’sinfiniteintellect.ThisfacultywasgiventouspreciselysothatwemightcometoasecureknowledgeoftheexistenceoftheGodwhocreatedusandtoalimitedknowledgeofGod’snature.ThusGodcreatedhumanbeingswithanintellectthatisinnatelyequippedtograspmathematicalprinciplesandtorecognizethemastruebeyondalldoubt.Thisrequiresthattheintellectbeexercisedintherightway,namely,byturningawayfromsensualdistractionstofocusonthingsintellectualandtherealmofpureabstractions(see,forexample,Descartes1985–91b;Augustine1990:BkVI,1955,BkII).Fortheneo-Platonist,thehumansoul(microcosm)wasanintegralpartoftheuniverse(macrocosm)andcapableofresonatinginsympathywiththeworldsoul.Themicrocosmandmacrocosmthusmirroreachotherinsuchawaythattheharmoniesofonewillberecognizedbyandaffecttheother.Oneitheraccountmathematicalknowledgeisamatterofinsight,ofintellectualintuition,oraseeinginthelightofreasonwhichisanalogoustoareligiousrevelation.Whenthemindisprop-erlyfocusedthefirstprinciplesofmathematicswillberevealedwithirresistibleclarityanddistinctness,fortheyareintegraltotheintellect’sownnature.Butbecausethehumanintellectislimitedandfiniteitdoesnotcomprehendallthingsimmediatelyinitsintellectualvision;itonlycomprehendsthosethingsthataresimpleandbasic.(InthemindofGodallthingsareimmediatelyintuited;Godneverneedstoengageinreasoningortomakeinferences.Neitherdoestheworldsoul;itsstructuresareintegraltoitselfandimmediatelyknowntoit.)Humanmindshavetoapproachprogressivelymorecomplexandcompleteknowledgeinstages.Reasoninganddeductiveproofdonotinthemselvesconfertheintellectualintuitionthatistrueknowledge,butreasonandintuitionare,forfiniteminds,thenecessaryroutetoit(seeSpinoza1992;Descartes1985–91a).Itishardlysurprisingthentofindthatsuchphilosophersgiveprioritytomethodsofdiscoveryandroutestoinsight,overmethodsofjustification.Proofsarelessjustificationsthanrecordsoftherealinterconnectionsofthings,theinterconnectionswhichexplain,forexample,whytrianglesmusthaveinternalanglesequallingthesumoftworightangles.Theseinterconnectionscanbegraspedinasingleactofintellectualvision,byonewhotrulyunderstands,butcanonlyberecordedsequentially.ThusforthePlatonicrealistthecosmosisanorderedwhole.Mathematicsexpressestheprinciplesofthisorderandgivesinsightintoit.Theroutetomathematicalknowl-edgeissupernaturallylinkedtoitsutility,throughtheideathatthesensesrevealtousonlythechangingphenomenalworldofsenseexperience.Toacquireknowledgeoftheeternalprinciplesunderlyingtheorderofthatworlditisnecessarytoturnawayfromthesensesandto(re-)unitethehumanintellect(orsoul)withtheintellectwhichorderstheworld(orwiththeworldsoul).350\nPHILOSOPHYOFMATHEMATICS2.2AristotelianconceptualismBycontrasttheAristotelianconceptualist’scosmos,althoughstructured,isinthefirstinstanceacollectionofindividualthingsthatundergochange,andwhoseorderispartlycontingentandpartlyaproductofthenaturesofthethingsmakinguptheuni-verse.Tocomprehendtheuniverseisthustoknowthenaturesofindividual(kindsof)things.Logicgivesussomeinsightintotheprinciplesoftheorderingofindividualthingsintospeciesandgenerabasedontheirspecificqualities(differentia).But,inaddi-tion,itisnecessary,byobservationandreflection,todiscoverthecausalprinciples,theforces,thepowersandpotentialitiesofthethingsthatgoverntheirbehaviour.Mathematicsdealswiththingsthatdonotchange.Moreover,itcanonlyeveryieldprinciplesthatconcernrelationsbetweenthings,becausenumbersareassignedonlyafterselectionofaunitofmeasureandgeometrydealsonlywithshapes,whichinturnarecharacterizedbytherelationsbetweentheirparts,andwithspatialrelationsbetweenshapes.Thismeansthatmathematicscanneveryieldinsightintothenatures,ortheprinciplesofchange,ofindividualthingsor,hence,intothecausalprinciplesthatgoverntheuniverse.Arithmeticisthescienceofdiscretemagnitudesandgeome-tryisthescienceofcontinuousmagnitudes.Magnitudesthemselvesareaproductofmeasurementandhenceofhumanmeasuringactivities.Itisinexperiencethatweencounterthevariouskindsofmagnitudesandinventmeasurementsystemsthatareappropriatetothemandtoourpurposes.Whenweinventmeasures,discriminatebetweenshapesandbecome,forpracticalpurposes,interestedinspatialrelations,forexample,ortherelativemotionsoftheplanets,mathematicscanhelpusdescribeinsystematicwaysthevariouskindsofordersthatweareinterestedinandtoco-ordinateourvariousmeasuringunits.Thusastronomyusesgeometryandarithmetictohelpconnectthevarioustemporalunits–suchasday,lunarmonthandsolaryear–inwaysthatmakeconstructionofacalen-darpossible.Geometryappliedtomeasurementsmadeonthegroundmakesbuildingplansandmapspossible.Mathematicsisusefulbecausemeasurementisusefulandthisutilityarisesoutofthepracticalcontextsinwhichmeasuresweredevised.Wecanhavemorecompleteknowledgeherethanofthenaturalworldbecauseitisahumaninven-tionandjustaswedonothavetoexaminethenaturalworldtofindouthowtousethewordsofalanguageortodeterminewhetheranactionisorisnotlegalaccordingtothelegalsystemofaspecificcityorcountry,sowewouldnotexpecttolooktotheworldtoresolvequestionsaboutmagnitudesandtheirrelations.Butinthiscasewouldtherenotbeasmanydifferentmathematicalsystemsastherearelegalsystems?Yetoneoftheexperienceswhichgivessupporttoclaimsabouttheuniversalnecessityofmathematicsisthatalthoughdifferentcultureshavedevelopeddifferentnumeralsystems,unitsofmeasurement,methodsofcalculationandmathe-maticalproblem-solving,thereisremarkableagreementatthelevelofresults.Merchantsspeakingdifferentlanguagescanneverthelessbargaintheirwaytoagreedprices,adaptingtodifferentcurrencyandmeasurementsystems.Twoplustwodoesnotmakefiveorthreeinthearithmeticofanyculture.AristotelianconceptualistsaccommodatethisinawaynotwhollydissimilarfromPlatonicrealists.Thisisnotsurprisingsincebothbelievedinauniversalhumannatureandemphasizedintellectasanimportantdistinguishingmarkofhumanity;ahuman351\nMARYTILESisarationalanimal.ThusforARISTOTLE(chapter23),LOCKE(chapter29),BERKELEY(chapter30)andevenHUME(chapter31),thehumanmindcomesintotheworlddis-posedtoimposecertainkindsofstructuresonit.Humansarebornabletostartmakingcertainbasicdiscriminationsonthebasisofqualitativesimilarityanddifferenceandtolearntomakemanymore.Infact,ifitisassumedthatallexperienceisofparticu-lars,thenallrelatingonethingtoanother,whetherwithrespecttocolour,tasteandsmell,orwithrespecttosize,shape,placeandtime,mustbeafunctionofthehumanmind.Tobeabletodothisthehumanmindmusthave,inadditiontoinnatecapacitiesforsenseperception,innaterelation-imposingcapacities.Reflectionontheseyieldsthebasicconceptsformoreabstractdisciplines,includingmathematics;conceptssuchasunity,number,size,shapeandindefinitelycontinuedrepetition.Theseabstractcon-cepts,presupposedinorderingexperience,canbesaidtobefirstintheorderofunder-standing,orsimplewithrespecttotheintellect,because,althoughtheymaybeexplicitlyacquired,ifatall,onlyafterintellectualreflection,theyarenonethelesspresupposedinallexperientialknowledge.Wheretherealistseesinthesesimple,innateintellectualconceptsindicationsofasupernaturalorder,theconceptualistseesonlystructuresofthehumanmindintermsofwhichitordersthenaturalworldofparticulars.Mathematicalfirstprinciplesthusinvolveconceptsderivedfromreflectiononinnatehumanintellectualcapacitiestodis-tinguishbetween,andtorelate,particulars.Inthissensetheyaresimpleinnateideas,ideaswhicharefirstintheorderofunderstandingalthoughnotintheorderofhumanexperience,becausetheycanonlycometobeknownbyturningawayfromtheworldandreflectingonthewayinwhichtheintellectfunctions.Theyarepotentiallyself-evidenttoallhumanbeingsbecausetheyareexplicitexpressionsofprinciplesthatareimplicitintheinnate,naturaloperationofthehumanintellect.Morecomplexconcepts,however,mustbebuiltupfromthesesimplesbyexplicitprocessesofdefinition.Notallcultureswillnecessarilydevelopthesamecomplexconcepts,althoughallstartfromthesamesimplesandallaremeasuringaspectsofthingsinthesamephysicalworld;thedefinitionofanycomplexconceptshouldthereforebeexplicabletoanyone,irrespectiveofculturallocation.Resultsconcerningmorecomplexconceptsmustbeprovedbyappealtotheirdefinitionsandtofirstprinciples,andtheymustproceedbydeductivesyllogisticreasoning.Proofs,inotherwords,aredemonstrationsinthesenseoutlinedbyAristotleinhisPosteriorAnalytics(1941a:93a16–96a19).Sincecomplexmathematicalconceptsaretheworkofhumans,thereisnogroundofproofbeyondthedefinitionofcomplexconceptsintermsofsimpleconcepts.Itcansometimesappearthatmoreisinvolvedbecauseitdoestakeworktoarriveatgooddefinitions.Thisisbecausemanymeasurement-relatedconceptsarisefirstinpracticalcontextswherecomputationordiagrammaticreasoningisinvolved.Itisonlywhenthequestionofthereliabilityofsuchreasoningarisesthatitbecomesnecessarytothinkaboutexplicitlydefiningconcepts,suchasacircle,astraightline,subtraction,multiplicationordivision.Agooddefinitionwillberequiredtojustifycorecomponentsofsuccessfulcomputationalandmeasurementpractices,butmaynotpreserve(orjustify)everycomponent.Fromthiswayoflookingatthings,justificationintheformofproofreceivesmoreemphasisthandiscovery,fortheredoesnotappeartobemuchtodiscoverybeyondgivingusefuldefinitionsanddeducingtheirconsequences.352\nPHILOSOPHYOFMATHEMATICS3NewProcedures,NewProblems:AnalyticGeometryandtheInfiniteSowehavetwoapproachestoresolvingthebasicproblem.Dowethensortoutourmetaphysicsandonthatbasistakeourpick?Ifweseetheworldasconsistingofchange-ableparticularsshouldweacceptaversionofAristotelianconceptualism,whereasifweseetheworldasaninterrelatedwholegovernedbyunchangingmathematicalprin-ciplesofharmonyshouldweacceptsomeformofPlatonicrealism?Thatwouldsurelybetoosimple,yetuptotheseventeenthcenturythechoicecouldhavelookedverymuchlikethis.Thiswasinpartbecausemathematicsitselfdidnotconstitutealargeorcomplexdomain.Itcameintheformofarithmeticandgeometry,whichdealtwithdis-creteandcontinuousmagnitudesrespectively.However,withtheRenaissancerevivalofneo-Platonismandofmathematicscamenewambitions,whichbeganwiththeextensionofthedomainofarithmeticbyalgebraandcontinuedviatheuseofalge-braictechniquesingeometrytothedevelopmentofinfinitesimalcalculusanditsuseinNewtonianmechanics.Bythelateseventeenthcentury,themathematical,scientificandpoliticalbackgroundhadshiftedinsuchawaythatneitheroftheoldoptionscouldseemunproblematic.InthecontextoftheincreasingreligiousintolerancegeneratedbytheThirtyYearsWarinEurope,neo-Platonismbecamesimplytoodangerousapositiontoholdopenlybecauseofitsobvioushereticalimplications.Tosurvive,thenewmathe-maticalnaturalphilosophyneededtodistanceitselffromthatposition.Descartes’sattempttounderwritethesecurityofhisrationalpowers(andhencethemathematicstowhichreasoningledhim)withaproofoftheexistenceofabenevolentGodwhocreatedhim,wasingenious,butitfailedtoconvincemanypeopleandalsotrespasseddangerouslyovertheborderfromphilosophytotheology,orfromthedomainofreasontothatofFAITH(pp.622–4).Thestrategyofphilosophicallysecuringanelevatedstatusformathematicsbysupernaturalmeanslookedincreasinglyunattractive,asdemarca-tionsweredrawnbetweenphilosophyandtheology,orbetweenwhatiswithinthelegit-imatescopeofhumanreasonandwhatmust,ifbelieved,beamatterorfaithorrevelation.Aristotelianconceptualismwasthusinitiallythemoreattractiveroute,anditcer-tainlyattractedempiricistphilosopherssuchasLockeandHume.Buttherewerefactorsinboththenewdevelopmentsinmathematicsandthoseinnaturalphilosophythatseemedstronglytofavoursomeformofrealism–theviewthatmathematicsdealswithasubjectmatterwhichexistsindependentlyofthehumanmind.Thedevelopmentofanalyticgeometrybroughtalgebraicmethodsintogeometryandsawthegradualemergenceofthemathematicalconceptofafunction.Thistookplaceinthecontextofdevelopingamathematicsthatwascapableofrepresentingchangeandofexpressingprinciplesofchange.Theequationofaparabola(y=ax-2bx),forexample,couldbereadbothasamathematicaldefinitionofthatparticularconicsectionandastheprincipleaccordingtowhichaprojectilemovesunderthe2influenceofgravityintheabsenceofairresistance(s=v0t-gt).Geometricalattentionshiftedtocurvesthatweredefinednotinwordsbutbyalge-braicequations,manyofwhichwerecurvespreviouslyexcludedfromgeometry353\nMARYTILESbecausetheyaregeneratedmechanicallybutcannotbedefinedintermsofrulerandcompassconstructions.Thesedevelopmentswereproblematicbecausetheymakethetacitsuppositionthatanypointonaplanecanbeassignedapairofnumbersasco-ordinates,onceco-ordinateaxeshavebeenfixed.TheancientGreeksalreadyhadproofthatthehypotenuseofaright-angledisosce-lestrianglewithunitsidesmusthavealength2,butthatthislengthcouldnotbeexpressedasanyratio.Thediscoveryoftheexistenceofsuchincommensurablemag-nitudeswasoneofthereasonswhyarithmeticandgeometryhadbeenregardedasdis-tinctmathematicaldisciplines.Onestudieddiscretemagnitudes,theothercontinuousmagnitudesanditwasnotthoughtpossibletoreducethestudyofcontinuousmagni-tudestothatofdiscretemagnitudes.Thisattitudewasreinforcedbyparadoxes,suchasZeno’sparadoxofAchillesandthetortoise,whichweretakentoshowthatacon-tinuousmagnitudecannotbemadeupofdiscreteparts.Thehistoryofmuchoftheworkinpuremathematicsintheeighteenthandnineteenthcenturiesisthehistoryofhow,inspiteofsuchparadoxes,wayswerefoundtoeffectareductionofthestudyofthecontinuoustoastudyofthediscrete.Thisreductionisnotwithoutcost.Thepriceishavingtodevelopamathematicsoftheinfiniteandadmittingtheinfiniteasafoun-dationalconcept.Thedifficultiesencounteredinthisprocessmakeithardtobelievethatcontinuousmagnitudesarehumaninventionsandthustransparenttothehumanintellect.Theinventionofinfinitesimalcalculusanditsuseinmechanicsgaveaddedurgencytophilosophicalquestionsaboutwhether,andhow,itispossibletoreasonreliablyaboutwhatisinfinitelylargeorinfinitelysmall.AnadditionalfeatureofNewtonianmechan-icsisthatitsassociatedviewofthephysicalworldisofaworldofsmallmaterialpar-ticlesmovingaroundin,butnotdenselypopulating,space.Theforcesactingonparticlesandcausingtheirmotionsareintimatelyconnectedtospaceandtimeintwoways.Firstly,theyareconnectedbyNewton’ssecondlawofmotion,whichstates:force=mass¥acceleration.ThisseemedtoNewtontorequirethatspaceandtimeformanabsoluteframeofreference,relativetowhichabsolutemotionisdistinguishedfromrel-ativemotion(changesinabsolutemotionbeingthosethatresultfromforcesactingonthemovingobject).Forrelativemotionsthesecondlawcanbefalse.Apersonstand-ingonarailwayplatformacceleratesawayfromapersoninatrain,fromwhosepointofviewthetrainisstationary.Butthereisnoforceactingonthepersonortheplat-formcausingthemtoaccelerateawayfromthetrain.Secondly,Newton’sgravitationaltheoryproposesthatthereisagravitationalattractionbetweenanytwomaterialpar-ticlesandthatthisforceisdirectlyproportionaltotheproductofthemassesofthetwoparticlesandinverselyproportionaltothesquareofthedistancebetweenthem.Inotherwords,spatialseparationisitselfacausalfactorasitdeterminesthemagnitudeofafundamentalforce.Themathematicsofspaceastheparadigmcontinuousmag-nitudemust,itseems,beregardedasdealingwithanaspectofthesamephysicalworldthatisstudiedbythenaturalphilosophers.IfonesubscribedtothevisionoftheworldproposedbyNewtonianmechanics,itwouldbehardnottobesomeformofrealistaboutmathematics.OnlybybeinginstrumentalistaboutNewtonianphysics–byregardingitasamerecalculatorydeviceforbetterpredictingandcontrollingeventsintheworldofeverydayhumanexperience–wouldonebemorelikelytobeattractedtosomekindofconceptualismaboutmathematics.354\nPHILOSOPHYOFMATHEMATICSButnotehowthegroundhasshiftedinsuchawaythatneitherformofpositionisentirelystable.Thenewformofrealismrequiresmathematicstorelatetoanindepen-dentlyexisting,naturalworldofmaterialparticlesinmotion,nottoaneternalandunchangingsupernaturalorspiritualworld.Itthereforefacesseriousproblemsaboutitsabilitytosayhowitcanbepossibleforustogetcertainknowledgeofthisworldbymathematical(asdistinctfromexperimental)methods,forthismaterialworldisexplic-itlydescribedashavingnoaffinitywiththehumanintellect.AppealtothebenevolentChristianGodcertainlyremainedapossibilityforbelievers,butitwouldbeinaclimatewherethismustberegardedasanactoffaith,andthusnotassomethingforwhichanyrational,philosophicjustificationcanbegiven.Thephilosophicalformofthebasicproblemwouldthushavetobedismissedasinsolubleandaninappropriatetopicofinvestigation.Thenewformofconceptualismalsofacedseriousproblems.Thefirstproblemareaisthenotionoftheinfinite.Ifthehumanintellectisfinitehowcanit–eitherinnately,orwhenpromptedbyitsnecessarilylimitedexperience–bethesourceofaconceptoftheinfinite?Onlysuchaconceptwouldbeadequatetoadetailedmathematicalcom-prehensionoftheinfinitecomplexityhiddenwithincontinuousmagnitudes.Secondly,theproblemsposedbythenewalgebraicmathematics,andthemethodsofinferencebeingusedtoprovidesolutions,madeitquiteclearthattheseformsofinferencecouldinnowaybeconstruedasconformingtothepatternofinferencefromexplicitdefini-tionsofconceptsviadeductivelogicalmoves(orsyllogisms)toconclusions.Becausedefinitions,eveningeometry,frequentlycametotaketheformofalgebraicequations,inferencesfromtheseinvolvedcomputationandthehandlingofequationsandin-equalitiesinacentralway.Allofthisisexplicitlyrelationalreasoning–itisnotassimilablewithintheframeworkofAristoteliansyllogistic,whichisbasicallyalogicofpredication.Twonewvariantstrategiesemergedinresponsetothechangedconfigurationofmathematicsanditsuses.One,accompanyinganINSTRUMENTALISTATTITUDETOSCIENCE(pp.294–5),wasconventionalistorformalist.Sinceinstrumentalismaboutsciencewasnotprominentintheeighteenthandnineteenthcenturies,thiswasnotawidelypursuedoption.Theother,proposedbyKANT(chapter32),isaformofconstructivismwhichrepresentsacreativesynthesisofelementsfrombothPlatonicrealismandAristotelianconventionalism.Theproblemoftheinfinite,forconceptualists,layintheirassumptionthattoattachanysignificancetoaword,thatwordmuststandforanidea.Thatideahadinsomewaytorepresent,ortopresenttothemind,somethinginfiniteandshouldhaveitsorigineitherinhumanexperienceorinthenatureofthefinite,humanintellect.Howcantherebeafiniterepresentationofsomethinginfinitewhichadequatelyexpressestheinfinityofwhatisrepresented?Berkeley(1995:§§19–20)short-circuitedthisproblembyarguingthatasign(orword)canacquiremeaningfromuse;itdoesnothavetostandforanidea.Ifaworddoesnotstandforanideaitdoesnotrepresentanyreality;however,asigncanbeofuseevenifithasnorepresentativefunction,orifitformspartofasystemthatwehavedevelopedforclassifying,orderingandanticipatingexpe-rience.Thissuggeststhatmathematiciansarefreetocreatesignssolongastheydonotclaimtobeabletoexpressanysuperiororarcaneknowledgebythesemeans.Theproblemforsuchaposition,asBerkeley’s(1947–57)astutecriticismsofNewton’s355\nMARYTILESmathematicalmethodsillustrate,ishowtoensurethatwhenwetrytomakeuseofournewsignswewillneithergetabsurdnorunreliableresults.DifficultiesoverspaceandtimeandtheusemadeofmathematicsinNewtonianmechanicsformedoneofthefocusesofKant’sCritiqueofPureReason.Hisformulationofthebasicquestionaboutmathematicsis‘Howissyntheticaprioriknowledgepossi-ble?’Hisansweristoassumethatitispossible(sinceNewtonianmechanicsisacceptedandisseentowork)andtoaskwhatarethenecessaryconditionsofitspossibility.HisanswerrejectstraditionalrelationalaccountsofspaceandtimeaswellasNewton’sabsolutespaceandtime.Heargueseffectivelythatspacecannotbereducedtospatialrelationsbutmustbethoughttobeaunitywhichhasitsowncharacteristicproper-ties.Oneargumentheuses(repeatedandstrengthenedbyNerlich1976)istopointoutthatthedifferencebetweenarighthandandalefthandcannotbecapturedbychar-acterizingtherelationsbetweenthepartsofthehand;rather,ithastodowiththewaythehandis‘inserted’intospace.Nerlichusestheverysimpleshapes(whichhecalls‘knees’)toillustratethesamepoint.Itisnotpossiblebyslidingarightkneearoundonaflatsurfacetomoveitinsuchawaythatitbecomesaleftknee.Thiscanonlybeachievedifwelifttherightkneeup,turnitoverandre-insertitintheplane,whenitwillnowbealeftknee.Thusthedifferencebetweenrightandleftisnotintrinsictotheobject,buttothewayitis‘oriented’inspace.Notallsurfaces,however,havethischar-acteristic.ThesurfaceofaMöbiusstripissuchthatbymovingarightkneeoncearoundthestripitbecomesaleftknee.Thus‘orientability’,thecapacitytoproduceright–leftdistinctions,isafeatureofspaceitself,notofthingsinit.KantonSpace,TimeandMathematicsKant’spositiononspaceandtimeisthusnon-reductionist,butitisnotrealist.Hedoesnotassertthatspaceandtimehaveanyrealityindependentofhumanperception.Rather,spaceandtimeareformsofhumanintuitionandhencestructureallhumanexperience.Wedonot,andcouldnot,gainthesenotionsfromexperiencebecausetheyarepreconditionsofandarepresupposedinallexperience.Allhumanexperienceoccursintimeandallexperienceofthingsoutsideourownmentallifeisexperiencedasoccur-ringinspaceaswellastime.Thefoundationofarithmeticistobefoundinthetempo-ralstructureofourexperience,andthatofgeometryinitsspatialstructure.Thesestructuresareoftheworldofexperience,buttheyhavethatstatusonlybecausetheformsofhumanintuitionarewhatconstituteandgiveformtothatworld.Thatmathematicsisgroundedbacktoaprioriformsofintuition,ratherthantoinnateideasorstructuresoftheintellect,issignificantandindicatesKant’sdivergencefromAristotelianconceptualism.Hedistinguishessharplybetweenlogicaldeduction,reason-ingfromconcepts,andmathematicalreasoning,whichinvolvesreasoningfromtheconstructionofconcepts.Itisourabilityimaginativelytoconstructmathematicalobjectsinpureintuitionthatgroundsourinsightintotheirproperties.Weimagineamathemat-icalpointtracingoutthecurvewhoseco-ordinatesatalltimessatisfytheequationforacircle,andtherebyconnectthealgebraicdefinitionwithwhatitconstructivelydefines.Thealgebraicdefinitionofacurveinanalyticgeometryisthusviewedasadefinitioninthesenseofarecipeforconstructinganobject,ratherthanasanAristoteliandefinitionofatermproceedingbyaspecificationofgenusanddifferentia.Itisourability,inour356\nPHILOSOPHYOFMATHEMATICSimagination,tocarryouttheoperationsthatcorrespondtofunctionalexpressions,whichformsthegroundthatmakesmathematicalknowledgepossible.Itisthefactthatspaceandtime,asformsofintuition,aresuchthatallitemsofexperienceappearasmagnitudesthatassuresusoftheuniversalapplicabilityofmathematics.Theverynotionofempiricalrealityandanobjectiveorderofexperienceisarguedtorequirethesuppositionthatthecausalorderofeventsissuchastoimposedeterminatemathematicalstructuresonspaceandtime.ThusKant,usingamodifiedconceptualiststrategy,soughttopreservetheessentialsoftheNewtonianframeworkwhilstavoidingtheproblemsthatwereencoun-teredbypostulatingabsolutespaceandtime.However,thefactthathumanintellectualconstructionformsthegroundformathe-maticalknowledgemeansthatiftheinfiniteistohaveanyplaceinmathematics,itcanonlybeasthepotentialinfinityoftheunending,never-completedseriesofwhichthenaturalnumberseriesistheprimaryexample.Noactuallyinfiniteconstructioncanbesupposedtohavebeencarriedout.Kantarguesboththathumanreasonissostructuredthatitwilltend,illegitimately,toformtheideaofacompletedinfiniteprocess,andthatthisisalwaysasourceofillusions.Ifwewanttotreatinfiniteseriesasobjectswewillnotbeabletoregardthemasthingsthatareevercompleted.Reasoningconcerningthemthuscannotobeyallthesameprinciplesasreasoningaboutfullydeterminateobjects.Inparticular,reasoningbyreductioadabsurdumcannotberelieduponinsuchcases.ItwaswithinabroadlyKantianframeworkthatmuchoftheworkinmathematicsanditsfoundationswaspursuedintothenineteenthcentury,particularlyinGermany.WorkbyWeierstrass(1815–97)andCauchy(1789–1857)seemedfurthertojustifyKant’spositionontheinfinite.Theysucceededinprovidingdefinitionsoflimitsandconvergenceforinfiniteseriesthatuseonlythenotionofapotentialinfiniteandmakenoappealtotheactualcompletionofaninfinitesequenceofapproximations.Thisworkseemedtosecurethefoundationsofanalysiswithoutinvokingtheactualinfinite.4Foundations4.1Logic,settheoryandanalyticempiricismTowardtheendofthenineteenthcentury,developmentswithinmathematicsandphysics,togetherwithwidersocio-culturalchanges,meantthattheKantianpositioncouldnolongerberegardedasadequate.Inmathematicsthediscoveryofpathologicalfunctions–functionssuchasWeierstrass’severywhere-continuous-but-nowhere-differentiablefunction•nnf(x)=Âbcos(ax)n0=clearlyindicatedthatgeometricalintuitionwasnotanadequatefoundationforpro-ducingreliablemathematics.Itwasfeltthatthetheoryofrealnumbersneededtobeputonawhollynumerical,arithmeticalfooting,onewhichdidnotatanypointappealtospatialintuition.357\nMARYTILESAsapartofattemptstodothisandtocharacterizetheorderstructurerequiredoftherealnumbers,Cantor(1845–1918)introducedtransfiniteordinalnumbers,andthen,havingprovedthattherecanbenoone–onecorrespondencebetweenthenaturalnumbersandtherealnumbers,introduceddistinctinfinitecardinalnumbers.Thatis,henotonlysupposedthattherecouldbeamathematicsoftheactuallyinfinite,but,havingarguedthatitisnecessarytodistinguishdifferentsizesofinfinitetotality,wentontodevelopanarithmeticofinfinitenumbers.Disputesoverthelegitimacyofthepro-posedintroductionofthesetransfinitenumbers,comingontopofthoseconcerningtheintroductionofGauss’s(1777–1855)imaginarynumbersandHamilton’s(1805–65)quaternions,madeclarificationoftheconceptofnumberandofthegroundsonwhichonecouldlegitimatelyclaimtohaveintroducedanewkindofnumberapressingissue.EinsteinandtheNatureofMathematicalKnowledgePerhapsmostdamagingtotheKantianpositionwasthediscoveryofnon-EuclideangeometriesandtheapplicationoftheseinEinstein’stheoriesofrelativity.InonesenseEinstein’sprocedurewasthoroughlyKantian.Hearguedfromthecausalstructuresindi-catedbytheformoflawsusedinelectro-magnetismtotheformrequiredforthoseofmechanics(assumingthatfundamentalphysicaltheoriesshouldconformtothesamespace-timeforms)andreadouttheimplicationsofthisforthegeometryofspace-time.Butthismeantthatitcouldnolongerbeplausiblyarguedthatgeometrygivesaprioriknowledgeofthespace-timestructureofthematerialworld.TheneteffectofEinstein’suseofnon-Euclideangeometryinaphysicaltheorywastodisestablishtheviewthatmathematicsisasourceofaprioriknowledgeabouttheempiricalworld.Euclideangeometrydoesnotcontain,andneverhadcontained,truthsaboutphysicalspacethatwereguaranteedtobecorrectinapplicationbytheirproofs.Thisappearedtovindicateempiricistclaimsthatsyntheticaprioriknowledgeisimpos-sible.Whenmathematicsformspartofempirical,scientifictheories,thenthecorrect-nessorotherwiseofitsresultsmustbeasmuchanempiricalquestionasthatofthecorrectnessorotherwiseofascientifictheory.Butifproofsdonotservetojustifyappli-cations,whatdotheyprove?Inotherwords,whatexactlyisthecharacterandcontentofmathematicalknowledge?Ifproofsdonotestablishclaimsabouttheworldofexperi-ence,weseemtobeleftwiththeoptionseitherofareturntoaformofPlatonicrealismorworkingtowardaviewthatlocatestheimportanceofmathematicsinitsfunctionanddoesnotseeitasmakingknowledgeclaims.TheoverthrowoftheNewtonianframeworkofclassicalmechanicswasthecombinedeffectofacceptanceofrelativitytheoryandofquantummechanics.Again,oneofitsphilosophicalimpactswastogiveaboosttoempiricism,particularlybecausequantummechanicswasacceptedonthebasisofitsempiricalsuccess,despiteconceptualprob-lemsovertheinterpretationofitsmathematicalformalismandconceptualresistancetoitsindeterminism.Nowfromanempiricistpointofview,areturntoaformofPlatonicrealismaboutmathematicswouldseemtobeanon-starter.Yet,ironically,thelogicistroutetoanempiricistaccountofmathematics(pursuedbyRussell(1919)andWhiteheadandRussell(1913))involvesadisplacedformofthePlatonicrealist358\nPHILOSOPHYOFMATHEMATICSstrategy.Twomovesarecrucialhere:(1)invokinganIDEALLOGICALLANGUAGE(p.797)(forscience);and(2)theuseofclasses,orsets,todefinenumbersintermsoftheirpotentialapplications.BothofthesemoveshadalreadybeenmadebyFREGE(chapter37)(1884,1893),whohadnocommitmenttoanempiricistepistemology.Fregewasconcernedtoprovideaphilosophicaldefenceoftheclaimsofmathematicstobeabodyofobjectiveknowl-edge.Hewasreactingontheonehandagainstwhathesawasthenaturalisticpsy-chologismofKantianandotherpositionsthattreatknowledgeasamatterofrelatingideas,andontheotherhand,hewasreactingagainstconventionalistclaimsthatmathematiciansarefreetocreatenewnumberssystems,sincethesearenomorethanformalsymbolsystems.Frege’sRealismFregeappealedtothemetaphorofgeographicexplorationwhendefendinghisclaimthatmathematicsisabodyofobjectiveknowledge,andheresistedfiercelythesuggestionthatmathematicianshavetherighttocreatenewmathematicalobjectsorconcepts.Hiscon-ceptionofobjectiveknowledgeisthusstronglyrealist.Suchknowledgeisofaninde-pendentlyexistingreality.Itisexpressedinlanguage,whichservestorepresentthatdomain,usingsentenceswhichsayeithertrulyorfalselyhowthingsare.Whethersuchsentencesaretrueorfalsedependsonthestateofthedomainreferredtoandtheresultisindependentofthestateofhumanknowledgeandofhumancognitivecapacities.Whatissaidcannot,therefore,beafunctionofideasinthemindsofindividuallanguagespeakers.Themeaningsofwordsmustbepubliclygroundedintheirlinguisticfunction,whichreferstothecontributionthattheymaketofixingthetruthconditionsofsen-tencesinwhichtheyoccur.Fromthepointofviewofknowledge-expression,thesen-tenceistheprimaryunitofmeaning.Theseviewsonlanguageandmeaningturnthefocusofphilosophicattentionawayfromideasandthoughtprocesses,tolanguageanditsfunctioninthecommunicationandexpressionofknowledge;theyconstitutewhathasbeenlabelledthe‘linguisticturn’,whichischaracteristicofanalyticphilosophy.ThisturnmadeavailableabroadlyPlatonicrealiststrategy,forlanguageishereconceivednotasanyactuallyexistingnaturallanguage,butasanidealvehiclefortheexpressionofobjectiveknowledgetowhichallnaturallanguagesmustapproximatetotheextentthattheysucceedinmakingknowledge-expressionpossible.Theideallogicallanguagebecomesthenon-naturalintermediarylinkinghumanbeingstothestructuresofreality.IttakesoversomeoftherolesofthemindofGod.Wecancometoknowthestructuralprinciplesgoverningfaithfulrepresentationinthislanguagebecausewearelanguagespeakers,andcanreflectontheconditionsofthepossibilityoflanguagefunctioninginthisway–athoughtthatfounditsclassicalexpressioninWittgenstein’sTractatusLogico-Philosophicus(1922).Logicisanexplicitcodificationoftheprinciplesoflanguageinitsrepresentative,knowledge-expressingfunction.Itslawsarenotinherentinthestructureofthehumanintellect;theyarelawsoftruth,notlawsofthought.Thestructuresofsuchanideallanguage,oneableaccuratelytorepresentreality,mustthereforealsobereflectionsofthestructureofreality.Whyshoulditbethoughtthatmathematics,thestudyofmagnitudes,canbereducedtologic,thestudyofthestructureandrelationsofsentencesintheirknowledge-359\nMARYTILESexpressingfunction?Thelinkhereisthenotionofa‘class’or‘set’.Thetheoryofclasses,theextensionsofconcepts,wastraditionallyapartoflogicbecauseAristoteliansyllo-gisticcan,ononeinterpretation,betreatedasalogicofclassesandtheirrelations.(‘Whalesaremammals’istreatedassayingthattheclassofwhalesisincludedinthatofmammals.)Thenaturalnumberscanbetreatedasmeasuresofthesizesofsets,orclasses,sothat,forinstance,thenumber2canbethemeasureofall2-memberedsets.Itcanbeshownthatnegativenumbersandrationalnumberscanbedefinedassets(ororderedpairs)ofnaturalnumbers.Dedekind(1831–1916)proposedadefinitionofrealnumbersasinfinitesets(orcuts)ofrationalnumbers(Dedekind1963).Thissuggestedthat‘set’,or‘class’,maybethesingleprimitiveconceptfromwhichallothermathe-maticalconceptscouldbedefined,ifgeometryandanalysiscouldthemselvesbereducedeithertoarithmeticortothestudyofsetsofpoints.Cantor(1895)andotherswerebasingtheirdefinitionsofnumberonaconceptionthatcanalreadybefoundinPlato.Thenumber2,forexample,consistsoftwoabstractunits;itistheidealtype(theoneovermany)towhichallpairsconform.Cantorsuggestedthatnumbersarearrivedatbyapsychologicalprocessofabstractionfromcollectionssuchastwopencilsonadesk.Wemustimaginativelysubtractallthedistinguishingqualitiesofthepencilsandtheirlocation.Ifweabstractfromeverythingbuttheirorder(theorderinwhichwecount)wegetanordinalnumber(anordertype)andifweabstracteventhiswegetacardinalnumber(acardinalitytype).Inotherwords,thePlatonicuniversalbecomesanideainanindividualmind.ButFregeruledoutsuchpsychologisticaccounts.Instead,headoptedadefinitionthatmakes2,theoneoverallpairs,simplytheclassofallpairs.Whatsavesthisfromcircularityisthathedefineswhatitistobeatwo-memberedclasswithoutusinganyreferencetothenumber2.Aclassisbydefinitionatwo-memberedclassifandonlyifthereisaone–onecor-respondencebetweenitandtheclasswhoseonlymembersare0and1.So,tosaythattherearetwopencilsonmydeskistosaythattheclassofpencilsonmydeskissuchthatthereisaone–onecorrespondencebetweenitand{0,1}.ThusifIcanfindonepencilonmydeskandpairitwith0andanotherandpairitwith1andtherearenoneleftover,Iwillknowthattheclaimistrue;otherwise,itisfalse.Ofcourse0and1havethemselvestobedefined.0isdefinedastheclassofclassesCsuchthatthereisaone–onecorrespondencebetweenCandtheextensionoftheconceptxπx.1isthentheclassofthoseclassesCsuchthatthereisaone–onecorrespondencebetweenCandtheclasswhoseonlymemberis0.Inthiswaythenumber2islinkedbydefinitiontoitspotentialapplications.Buttojustifytheclaimthatallarithmetictruthscanbeprovedbyappealonlytodefinitionsandformallyexpressedlogicallaws,thedefinitionsgivenabovemustbeexpressedinthelanguageofformallogic,andmustuseonlyconceptsbelongingtologic.Further,itmustbeshownthatallthenormallyacceptedarithmeticprinciplescanbelogicallyderivedfromthesedefinitions.Twothingswerecrucialtothisproject:firstly,treating‘=’asaprimitivelogicalrela-tionand,secondly,assuming,asaprincipleoflogic,thateveryconceptdefinedoverindi-vidualobjectshasaclassasitsextension.Itwasalsonecessarytoexpressinlogicaltermswhatitisforarelationtobeaone–onecorrespondencebetweentwoclasses.Thisideacanbedefinedasfollows–‘Risaone–onecorrespondencebetweenAandBifandonlyifforeveryxinAthereisjustoneyinBsuchthatxisR-relatedtoyandforeveryyinB360\nPHILOSOPHYOFMATHEMATICSthereisjustonexinAsuchthatxisR-relatedtoy’.Toexpressthisasalogicalconditionrequiresalogicofrelationscapableofhandlingmultiplequantificationinasinglesen-tence,andinwhichitispossibletoexpressthenumericalconcept‘thereisjustone’.Noneofthesewasavailableintraditionallogic.However,focusonthesentenceastheunitofmeaningenabledFregetotreatconceptexpressionsandrelationalexpressions(predicates)byanalogywithmathematicalfunctionsofone,twoormorevariables,assentence-formingoperators.ItwaswithinthisexpandedlogicalframeworkthatFregeclaimedthatarithmeticreducestologic.Whathadpreviouslybeenamajorstumblingblocktosuchaclaim–therelationalcharacterofarithmeticalreasoning–wasovercomebytheintegrationofalogicofrelationswiththelogicofconcepts.Thisbrilliantstrategy,however,cametogriefasaresultofRussell’sparadox,whichdemonstratedthatFrege’ssystemof‘logic’isinconsistent.Theparadoxarosebecauseitispossible,inthissystem,toexpresstheconcept‘isaclasswhichdoesnotbelongtoitself’.Sinceclassesare,forFrege,individualobjects,andeveryconceptdefinedoverindividualobjectsisassumedtohaveaclassasitsextension,thisconcepttoomustdefineaclass–sayR.DoesRbelongtoitself?Ifitdoesthenitmustbeaclasswhichdoesnotbelongtoitself;clearlyacontradiction.Butifitdoesnot,thenitisaclasswhichdoesnotbelongtoitselfandhencebelongstoR(itself).However,RUSSELL(chapter37)wasattractedtoFrege’sstrategy.Ifitwerepossibletoshowthatallmathematicsreducedtologic,thenthiswouldsubstantiallystrengthenthecasefortheempiricistdenialofanyautonomytoconceptualknowledge.Russellwasanempiricalrealist:inotherwords,heassumedthatthereisaworldthatexistsindependentlyofourexperience;itisaworldofatomicevents(or‘sensedata’).Theorderofoccurrenceofsucheventsisthuswhollycontingent.Objectiveknowledgeofthisworld,asattemptedbynaturalscience,takestheformofadescriptionofglobal,structuralfeaturesofthetotalityofeventsthatconstitutestheuniverse.Ittakestheformofknowledgeofrelationsbetweensensedata;sensedqualities,sincetheycanonlybeknownsubjectively,mustdropoutofthescientificpicture.Thussimple,atomic,sentenceswillallstatethatacertainrelationholdsbetweentwoormoresensedata.Becausetheorderofeventsiscontingent,wecanneverinfactbeinapositiontoknowthetruthofanyglobaldescription.Itstruthconditionsaresuchthatonlyanon-finitebeingapprehendingthewholespace-timeuniverseinasingleintuitioncoulddirectlyverifyitstruth.Theformthatadirectdescriptionofsuchaworldwouldtakewouldbealistofallthetrueatomicsentencesoftheideallogicallanguageofscience.Iftheuseofmathematicsinscienceistoleadtoobjectivelytruedescriptivestatementsaboutthisworld,everydescriptiongivenusingmathematicaltermsshouldbeanabbre-viationthatisreduciblewithoutresiduetosome(possiblyinfinite)complexsentencebuiltoutofatomicsentencesinsuchawaythatthetruthvalueofthecomplexisdeter-minedbythoseofitsatomiccomponents.Inotherwords,thecomplexisconstructedusing(possiblyinfinite)truth-functions.Thusthedefinitionsofallcomplexmathe-maticaltermsshouldindicatehow,inprinciple,suchaneliminationcouldbeeffected.Suchabbreviationsareresortedtoonlybecauseofthelimitedcapacityoffinitehumanminds.Mathematicaltruthsarethosethatcanbederived,logically,fromanunpack-ingofthedefinitionsofmathematicalconcepts.Theydonotsayanythingabouttheworld,buttheydohelpusmakelogicallycorrectmovesaroundtheshorthandlanguagethatwehavedevelopedfordescribingit.361\nMARYTILESHowdidRussellproposetoavoidtheproblemFregeraninto?Hisendsolutionistheso-calledno-classtheorywhichincorporateshis‘ramifiedtheoryoftypes’.Therearetwokeyideashere:(1)classesarelogicalfictions,theyarenotrealindividualobjects;(2)forthisreason,thelanguagewhichincorporatesthesefictionsmusthaveahierar-chicalstructure.Atthebaselevelwehavetheempiricalworldofindividuals.Conceptsaredefinedovertheseandforconvenienceweintroducenamesofclassesasashort-handwayoftalkingaboutallthethingstowhichagivenconceptapplies.These‘classes’clearlycannotbetreatedasbelongingtotheoriginallevel,butoncetheyhavebeenintroduceditispossibletodefinenewconceptsusingthem,andtointroducemoreclassesonayethigherlevel.Thishastheeffectofrulingoutthepossibilitythataclassshouldbelongtoitselfandalsothepossibilityofdefiningtheconcept‘isaclass’;itisonlypossibletodefine‘isaclassofleveln’.However,asRusselladmitted,therewasnohopeofdevelopingmathematicswithinthisframeworkwithouttwosubstantialassumptions:theAxiomofInfinityandtheAxiomofReducibility.TheAxiomofInfinitysaysthatthereareinfinitelymanyindividualsintheuniverse.ThisisnecessarybecauseRussellcouldnotuseFrege’sdefinitionsofthenaturalnumbersinquitetheiroriginalform.Thebasicideaisstillthesame–2istobetheclassofalltwo-memberedclassesofindividuals–butadifferentdefinitionofwhatitistobeatwo-memberedclassisrequired,namely,onewhichdoesnotreferto0and1.Thesolutionwastodefineatwo-memberedclassastheextensionofanyconceptFforwhichthefollowingistrue:‘therearedistinctobjectsxandywhichsatisfyFandeveryobjectwhichsatisfiesFisidenticaleitherwithxorwithy’.However,thismeansthatinauniversecontainingonlyoneobject,2wouldbetheemptyclass,andsowould3andeverylargernumber.Thustoensurethatthereareinfinitelymanynaturalnumbers,theremustbeassumedtobeaninfiniteuniverse.Thisillustratestheimpossibilityofsimultaneouslyrestrictingobjectstothosethatactuallyexistintheempiricalworld,insistingthatmathematicaltermsbedefinedinsuchawayastobuildinalltheirlegitimateapplicationsinadvance,andclaimingthatmathematicsdoesnotembodyanyknowledgeoformakeanysubstantiveclaimsaboutthatactualworld.TheAxiomofReducibilityisrequiredtobeabletodefinerealnumbersasinfiniteclassesofrationalnumbers.Basically,itclaimsthatallpossiblewaysofgroupingindividuals(into‘classes’)arealreadyavailableusingconceptsdefinableatthefirstlevelofthehierarchy;thatis,theintroductionofmorecomplexconceptswhosedef-initionsmakereferencetoalreadydefinedconceptsandclassesdoesnotinfactmakeanynewdiscriminationspossible;itdoesnotgenuinelyenlargeourvocabularyfordescrib-ingthestructureoftheworld.Thisassumptionis,however,equivalenttotheclaimthatthereductionofmathematicstologicwassupposedtohelpestablish–thedenialthatconceptscontributetothecontentofwhatwesayabouttheempiricalworld.Bostock(1974,1979),Field(1980),Chihara(1990)andWright(1983)havemadeinterestingattemptstofollowlogicistcoursessomewhatdifferentfromRussell’s.Butineachcase,asinRussell’s,thequestionarisesofwhatexactlyfallswithinthescopeoflogic.Aconvincingclaimtohavereducedmathematicstologicwouldrequireaclearwayofdistinguishingbetweenwhatisandwhatisnotlogic,andwell-groundedphilo-sophicalmotivationsformakingthedistinctionexactlywhereonedoes.OnewayofmakingthatdistinctioncanbefoundinQuine’swork.Fromthepointofviewofempiri-362\nPHILOSOPHYOFMATHEMATICScism,logicshouldnotintroduceanyontologyofabstractobjects.ThusQuine(1970)limitslogictofirstorderpredicatecalculus–theonlyobjectstalkedaboutareindivid-ualobjects.Modallogicandhigherorderlogicarediscounted,butifmathematicsisreducibletologicatall,atleastsomeformofoneorbothofthesekindsoflogicseemstoberequired.Thealternative,theonetakenbymathematiciansandbyQuine(1963),istosaythatmathematicsdoesnotreducetologic,butcanbereducedtoaxiomaticsettheory.Philosophicalquestionsaboutthestatusofmathematicsthenbecomequestionsaboutthestatusofsettheory,sincethisisthefoundationonwhicheverythingelsecanbebuilt.Analyticempiriciststendedtoaccept,increasinglyunquestioningly,thisfocusonfoundationswhereeverythingcomestorestonlogicandsettheory.HereagaintheoptionsofPlatonicrealismorsomeformofconceptualismareavailable.TherearethosewhodefendPlatonicrealism(seeBrown1991;Maddy1990)byadoptingtheviewthatthebestexplanationforthesuccessfulapplicationofmathematicsisthatitdoesexpressknowledgeofprinciplesunderlyingthestructureofreality.Thisneatlyturnsthebasicproblemaroundbutleavesoneaskingaboutthevalidityofthisstyleofinference.Whatotherexplanationsofsuccessinapplicationarethere?Howarewetoadjudicatebetweencompetingexplanations?ThereisacompetingexplanatorystrategyofferedbyQuine’s(1969)NATURALIZEDEPISTEMOLOGY(pp.302–3).AlthoughtheDarwinianaccountofthedescentofmanhasrenderedimplausibletheideathatGodcreatedhumanswithintellectsinnatelyequippedtodiscernthemathematicalprinciplesaccordingtowhichhecreatedtheuniverse,itcanbeusedtounderwritetheutilityofourinnatecapacities.Ourinnatecapacitiesareaproductofevolution,thatis,partlyamatterofchance(randomgeneticvariation)andpartlyamatterofnaturalselection,whichfavoursthosevariationsthathaveuptonowprovedtohaveadaptivevalue.Thuswefindcertainthingsself-evidentbecausethesethingshave,inthepast,proveduseful(theyhaveconferredsuperiorsur-vivalvalue).Adaptivevalueis,however,relativetotheenvironingexperience.Wecanhavenoguaranteethatexperiencewillnotinthefuturerequirechangesinmathe-maticaltheories,justasitmayrequirechangesinphysicaltheories.Mathematicalthe-oriesarenodifferentinkindfromotherscientifictheories;theyarejustmoregeneralandthusmoreresistanttochange,becausechangingthemmeanswholesalerevisionsinsomanyotherareas.Acknowledgingthis,Quineadoptsaninstrumentalistattitudetowardmathematics,bywhichitdoespostulatetheexistenceofabstractobjectsanditdoespostulatethemasrealexistents,butthisisseenaspartofthemythologythatwehavedevelopedfordealingwiththephysicalworld.Wearerealistaboutsuchthings,butitis(touseatermcoinedbyPutnam)an‘internalrealism’inwhichwecanreceivenojustificationthatguaranteeseitherultimatetruthorpermanentutility.4.2Intuitionism,formalismandopenquestionsWhatlogicistandsettheoreticfoundationalistapproachestomathematicssignificantlyignoreisthequestionofhowhumanbeingsactuallyacquireandusemathematicalknowledge.Theyfocusonquestionsofjustification,relegatingthestudyofmethodsofdiscoverytopsychology.WittgensteinandmorerecentlyDummett,togetherwithformalistsandintuitionistcriticsoflogicism,pointedoutthattheproposedlogicist363\nMARYTILESreduction,evenifsuccessful,begscentralquestionsabouttheepistemologicalfounda-tionsofmathematics.Thisisbecause,tocheckthatafulllogicalanalysisof,forexample,‘therearetwothingswhichareF’iscorrect,onemustcountoccurrencesofvariables.Inotherwords,agraspofarithmeticalconceptsispresupposedinbeingabletoreadaformallogicallanguageandunderstanditsdefinitionsofnumbers.Theanaly-siscannotindicatehowweacquireagraspofnumericalconceptsandhencecannotexplainhowwemightmanagetoextendourgraspoffinite,naturalnumberstoanunderstandingoftermspurportingtorefertoinfinitesets,infiniteseriesortransfinitenumbers.Thesecriticstakeitthatoneofthetasksofaphilosophyofmathematicsistoexplainwithwhatrighthumanbeingswithfinitecapacitiescanintroduceandclaimtounderstandsymbolspurportingtorepresenttheinfinite–symbolswhoseusewasintendedtofacilitateaknowledgeofinfinitestructures.Thusinvariousformstheirconcerniswiththewayinwhich,ortheextenttowhich,agraspoftheinfinitecanbegroundedinapprehensionofthefinite.Brouwer’s(1881–1966)intuitionismwasself-consciouslyKantianininspiration(Brouwer1975).Itgroundsthearithmeticofnaturalnumbersinthepureformoftemporalintuition:thebaretwo-onenesswhichisapreconditionofbeingabletoapprehendanythingasamanifold(aunityhavinginternalcomplexity),andhenceofanyknowl-edgewhatsoever.Thisisusedtoprovideafoundationforanalysis,viaatheoryofthecreativesubject.BrouwerfollowedKantineschewinguseoftheactuallyinfinite.Hethusrejectedsettheoreticapproachestothedefinitionoftherealnumbers.Potentiallyinfinitesequences,theparadigmofwhichisthenaturalnumbersequence,canbeobjectsofknowledgebutcannotbetreatedasiftheywerefullydeterminate.Suchasequencecanonlybeknownthroughtheprinciplewhichgeneratesit,theprinciplewhichassuresusthatitisunending.Thismay(foralawlesssequence)besimplytheever-availablefreechoiceofthecreativesubjecttoextendit.AsKanthadalreadyargued,reasoningconcerningincomplete,potentiallyinfinitetotalitiescannotvalidlyemployreductioadabsurdumproofs.ThisisbecausewehavenowarrantforassertingthatoneofPornot-Pmustbetrue(thePrincipleoftheExcludedMiddle)inadvanceofgeneratingthesequencetoapointwhereitcanbedirectlydeterminedeitherthatPorthatnot-Pisthecase.Forsomeclaimswehavenoguaranteethatsuchapositioncaneverbereached(forexample,regardingtheconjecturethatthereareonlyfinitelymanytimesinthedecimalexpansionofpthatadigitinanodd-numberedplaceisequaltothedigitinthefollowingeven-numberedplace).BecauseintuitionistsrejectthePrincipleoftheExcludedMiddle,andproofsbasedonit,therearemanyproofswhichclassicalmathematiciansacceptbutwhicharerejectedbyintuitionists.HavingrejectedtheFrege–Russellframeworkforlogic,intuitionistsdevelopedalogicthatdoesnottaketheprincipleoftheexcludedmiddletobeuniversallyvalid.Heyting(1956),Dummett(1973)andotherstakingthelinguisticturn,haveusedthisalter-nativelogicasameansofprovidingaframeworkforintuitionismwhichdoesnotappealtoKantianintuition.Anaccountofthemeaningofamathematicalclaimisnottobegivenbystatingitstruthconditions,butbyaspecificationoftheconditionsunderwhichonewouldbeentitledtoassertit(theproofconditions).Thislinkstheunder-standingofmathematicalquestionstoknowledgeofwhataproofconstitutingananswerwouldlooklike.Inarithmetic,thebasic,canonicalproofsarecomputations.Effectivecomputationalproceduresthusfigureprominentlyinintuitionistandcon-364\nPHILOSOPHYOFMATHEMATICSstructivistmathematics.Forthisreasonmuchofwhathasbeendevelopedhasbeenfoundtobeuseful,evenbythosewhodonotespouseintuitionistorconstructivistviewsandwhodonotrejectthemathematicsthatrequiresuseoftheactualinfinite.Hilbert(1882–1943)(1926)adoptedapositionverysimilartothatofintuitionismwithrespecttothearithmeticofnaturalnumbers.Inhiscase,however,thefoundingintuitionswerenotthoseofpureimagination,butarederivedfromtheuseoftallysystemsandtheprogressiontowrittennumerals.Hethusfocusesonmathematicsasatool,oraninstrumentforsolvingproblems,ratherthanasanelaborationofinnatehumanintellectualstructures.Thismeansthatwhenitcomestotheuseofinfinitenumbersandsymbolspurportingtorepresentinfinitisticoperations,Hilbert’sresponseisconventionalistandinstrumentalist,ratherthanconceptualist.Whileinsistingthatsuchsignscanhaveforusnodirectrepresentationalcontent,heallowsthattheymayformalegitimatepartofmathematics.Symbolshavemeaningwhichisderivedfromtheiruseinformalmathematicalsystems,andtheirintroductionisjustifiedifitmakespossiblethesolutionofpre-existingproblemsandifthetotalsystem,consistingofthenewsymbolstogetherwiththeoldsystemwhichtheyextend,canbeshowntobeconsistent.Buthowcanconsistencybeproved?Onewaytoshowthatatheoryisconsistentistoshowthatithasamodel.Thusnon-Euclideangeometrieswereprovedconsistentrel-ativetoEuclideangeometrybyshowingthattheyhadmodelsinEuclideangeometry.RiemannianplanegeometrycanbeinterpretedasgeometryonthesurfaceofaEuclid-eansphere,forexample.ButthisstillleavesopenthequestionofwhatassuresusoftheconsistencyofEuclideangeometry.Hilbert(1899)hadprovidedanaxiomatizationofEuclideangeometrywheretheaxiomstaketheformofconditionsonrelationsbetweenpointswhichmustbesatisfiedifatotalityofpointsistobeconsideredaEuclideanspace.Hefurthershowedthattherealnumberscouldbeusedtoprovideamodelfortheseaxioms.Thisreducestheconsistencyquestiontohowtoensuretheconsistencyofthetheoryofrealnumbers.ForbothHilbertandintuitioniststheconsistencyofthearithmeticofnaturalnumbersandofeffectivecomputationalprocedures(finitaryarithmetic)isguaranteedbythefactthatmodelscanactuallybeconstructed(whetherinintuitionoronpaper).Thetheoryofrealnumbers,quantificationoverthetotalityofnaturalnumbers–whentreatedasobeyingthelawofexcludedmiddle–andtheuseofsymbolsforinfinitesetsandtransfinitenumbersallfacetheproblemthat,exhypothesi,itisnotpossibletofindeffectivemethodsforconstructingmodelsoftheinfi-nitetotalitiesthattheypresuppose.Theconsistencyquestionthusreducestohowtogroundtheuseofsymbolsforinfinitaryobjects,orinfinitaryoperations,inthefinite.WhatcametobecalledHilbert’sProgrammewasthesuggestionthatoneshouldtrytotreatconsistencyasaformalpropertyofsymbolsystems.Symbolsystemsarethem-selvesfinite,eveniftheycontainsymbolsintendedtorefertoinfiniteobjectsortoinfini-taryoperations.Theyaregovernedbyruleswhosecorrectapplicationcanbecheckedbyappealmerelytoacapacitytorecognizethesymbolicformsofformulae,whichissomethingthatdoesnotrequireagraspoftheircontent.Proofsinsuchsystemsarefinitesymbolstructuresconstructedaccordingtoformalrules.Forinconsistencytoariseitwouldhavetobepossibletoprovesomesentenceanditsnegation;forexample,toproveasentencewithaformalstructureanalogousto‘Pandnot-P’.Ifitwerepos-sibletoshow,byreferencetotheformal,finitarypropertiesofthesymbolsystemthat365\nMARYTILESnosuchformulacouldbethelastlineofaproof,thenonewouldhaveademonstrationofconsistency,onewhichusesonlyfinitary,constructivemethods.GödelandConsistencyGödel(1906–78)(1931)proved(inhisFirstIncompletenessTheorem)thatitisimpos-sibletocompleteHilbert’sProgramme.Heproved,usingfinitarymethods,thatifaformalsystemScapableofexpressingarithmeticisconsistent,thenthereisasentenceGexpressibleintheformallanguageofSwhichissuchthatneitheritnoritsnegationcanbeprovedinS.(Gissoconstructedthatitcanbeinterpretedassayingofitselfthatitisnotprovable.)Hefurthershowed(inhisSecondIncompletenessTheorem)that,asaresult,aninconsistencyarisesifitissupposedthatScanbeprovedconsistentbyfinitarymethods,forthiswouldmakeitpossibletoproveinSthatGisnotprovableinS.SinceGisaformalexpressionoftheclaimthatGisnotprovableinS,thiswouldalsobeaproofofG.Hence(bytheFirstIncompletenessTheorem)Swouldbeinconsistent.Thekeyideabehindthisproofwastocodeallsymbolsoftheformallanguagebynumbers.Formulaeandproofs,beingfinitesymbolstructures,canthenalsobecodedbynumbers,insuchawaythattheirsymbolicpropertiescorrespondtoarithmeticalpropertiesoftheirnumericalcounterparts.Sincetheformallanguagewasintendedtobeonethattalksaboutnumbers,thisdeviceallowsthesymbolsystemtohaveadoubleinterpretation–intalkingaboutnumbers,italsotalksaboutitsownformulae.Thisidearevealedtheenormouspotentialofnumbersasdevicesforcodingsymbolicandotherstructures.Itisthispowerthatwassubsequentlyexploitedtodevelopdigitalelectroniccomputers.Inthisway,mathematicsconcernedwithformalsystemsandeffectivecomputationalprocedures,whichwasdevelopedinresponsetoHilbert’sProgrammeandtotheemphasisplacedoneffectiveproceduresbyintuitionism,notonlyfoundimmensepracticalapplicationbutalsoledtothedevelopmentofwholenewareasofmathematics,includingprooftheory,modeltheoryandrecursivefunctiontheory.Butwheredoesthisleavetheissueofjustifyingtheuseofinfinitarymathematics?TheexactinterpretationofthephilosophicalimportofGödel’stheoremhasbeenmuchdis-cussed(Detlefsen1986)andthereisnogeneralconsensus.Whatdoesseemtobeagreedisthatthereisnohopeofacompletejustificationalbootstrapping,usingonlyfinitarymethods,toprovideanunchallengeablelegitimationofinfinitarymethods.Formalsystemsarethemselvespotentiallyinfinite.SotoproveresultsaboutawholesystemS,suchasC,whichstates,‘ThereisnoproofinSofanyformulaoftheform“Pandnot-P”’,requiresaproofaboutapotentiallyinfinitetotality.Butwithouttheassumptionthatitislegitimatetoregardinfinitetotalitiesascompletedobjects(whichistheassumptionthatafinitaryconsistencyproofistryingtojustify),thereisnojustificationforassertingthateitherCornot-Cmustbetrue.Similarly,settheory’sabilitytodeliverdefinitiveanswersonfoundationalquestionshasbeenshowntobelimited.Whathavecometoberegardedasthecoreaxiomsofsettheory–namely,theZermelo–FraenkelaxiomandtheAxiomofChoicewhich,together,weshallcallZFC–donotmakepossiblearesolutionofthequestionthatCantorsoughttoanswerbyintroducingtransfinitenumbers,namely,thequestionofthestructureofthecontinuum.Cantorprovedthattherecanbenoone–onecorre-366\nPHILOSOPHYOFMATHEMATICSspondencebetweentherealnumbersandthenaturalnumbers.Heinterpretedthis0resultasproofthatthecardinalnumberofthesetofrealnumbers(2)isgreaterthanthatofthesetofnaturalnumbers(0).Healsoprovedthatthecardinalnumber(1)ofthesetofallwell-orderingsofthenaturalnumbersmustbegreaterthanthatofthenaturalnumbers,andthattherecouldbenoinfinitesetwhosecardinalitywasgreaterthan0butlessthan1.Thatis,1isthenextinfinitecardinalnumberafter0.Cantor0thoughtitwasthecasethat2mustbethesamenumberas1(Cantor’sContinuumHypothesis,orCH),buthewasunabletoproveit.Gödel(1940)producedamodelofZFC(theconstructiblesets)inwhichthisisthecase,butCohen(1966)producedoneinwhichitisnot.TogethertheresultsshowthatZFCleavesthequestionundecided.Thequestionremains,ifweacceptZFCanduseinfinitesets,onwhatgroundshouldweacceptorrejectCH?Whatdifference,ifany,doesitmakeinotherareasofmathematics?Thesefoundationalquestionsrequiredetailedtechnicalworkfrommathematicians.Thelargerphilosophicalquestions,however,concernevaluationofthewholefounda-tionaltrend,atrendwhichhasnot,ashadbeenhoped,providedonceandforallsolu-tionstoquestionsaboutthestatusofmathematicsandourjustificationforrelyingonit.Ithasproducedlotsofnewandinterestingmathematicsandthisinitselfmustsurelytellussomethingaboutcontemporarymathematics.Thequestionis,whatexactlydoesittellus?5BeyondFoundationalismandAnalyticPhilosophyFoundationalismisreductionist;itseekstodetermineahierarchicalorderingofagivenareaofknowledgebypushingbacktothestarting-pointsfromwhichalltherestdeduc-tivelyfollowsandhencecanberationallyjustified.Questionsofthestatusofthatwholeareaofknowledgethuscometofocusonthestatusofthesefirstprinciples.Thispre-sumesthatthereissomeabsolute,timeless,universalsenseofrationalorder,orofwhatconstitutesarationaljustification.Yetasthehistoricaloutlinepresentedherealreadyshows,ideasaboutstandardsofrigourinmathematicalproofsandaboutwhatcon-stituteslogichavechanged.Thefactthatmathematicalreasoningwasrelationalwasenoughatonetimetosuggestthatitdidnotproceedsolelyaccordingtoprinciplesofdeductivelogic.Nowthatrelationalreasoninghasbeenincorporatedintologicintheformoffirstorderpredicatecalculus,thatparticulargroundforthinkingthatmathe-maticalreasoningdoesnotreducetologicaldeductionhasbeenremoved.Butwhatarethegroundsforthinkingthatfirstorderpredicatecalculusistheultimate,canonicalformoflogic?Therearemanypotentialchallengers.Mathematicianshavestudiedanynumberofformal‘logics’,includingmodallogics,andmany-valuedlogicsandrelevantlogics.Informationtechnologyinitspursuitofthedevelopmentofexpertsystemshasinthepastfewyearsspurredagrowthofinterestinnon-monotoniclogics.Shouldwebemonistorpluralistaboutlogic?Solongasfoundationalismwastheprevailingphilo-sophicattitudetowardaccountsofknowledgeinmathematicsoranyotherarea,ithadtobepresumedthatthereisjustonesystemoflogic,thatwhichrevealstheabsolute,timeless,universal,rationalorderofknowledgeandwhichprovidesanaccountofwhatconstitutesarationaljustificationforbelief.Butfoundationalisminotherareas367\nMARYTILESofepistemologyhasinthelasthalfofthiscenturyincreasinglybeenchallenged.Theseassumptionsaboutrationalitynolongerappearself-evident.Yet,asBloor(1973)wasawarewhenlaunchingthestrongprogrammeinthesociologyofknowledge,mathe-maticswouldbeatestcaseforthatprogramme,asitwasfortheanalyticempiricists.Itmightbeabastionofafixed,eternalrationalorderaroundwhichdefendersofclaimstouniversalrationalitycanrally.Butthestandardsofrationalitycouldprovelessthanfixedandtheorderpresupposedbyfoundationalistscouldproveunnecessarytoexplainthedevelopmentofactualmathematicsoritsapplications.Ifthisweretrue,thentheclaimsoffoundationalismandofaccesstouniversalstandardsofrationalitywouldlookshakyindeed.Bloortakesmanyofhiscuesfromtheanti-foundationalistlaterphi-losophyofWITTGENSTEIN(chapter39)(1963,1967)whichemphasizesthewayinwhichformsoflife–socialpracticesofwhichmathematicsandlanguageareintegralparts–informoursenseofwhatisorisnotrational.Wittgenstein’sviewofmathe-maticsisnon-reductionist.Moreover,herepeatedlyemphasizeswhathecallsthe‘motleyofmathematics’–thevarietytobefoundinitsvariousbranchesandintheirapplications.Heraisesthequestionthatotherphilosophersdidnotatthattimedreamofraising–questionsofhowitisdeterminedwhataproofprovesandofhowthelogical‘must’acquiresitsforce.Heregardedmathematicsasanintegralpartoftheframeworkoflinguisticconventions,whichwedevelopinthecontextoftryingtolivewithothersinamaterialworld.Theseconventionsarelearntbytraining,thatis,inrepetitive,correctedpracticeincertainkindsofcontexts.Themeaning(orcontent)ofalinguis-ticsignthusderiveswhollyfromitscustomaryuses.Theseusesthemselvesderivefromformsoflifeinwhichtheyareembedded;forexample,calculatingaspartofexchang-ingmoneyforgoodsofvariouskinds.Innormalcircumstancesthequestionofjustifi-cationwillnotariseandcanreceivenoanswerbeyond‘Thatisthewaywedoit’.Thisisbecauseuseisthebedrockofunderstanding.Butasweknowfromtheoccurrencesofethicalandpracticaldilemmas,elementsofhumanpracticescancomeintoconflictwithoneanother.Itisinsuchcircumstancesthatreflectiveconceptualanalysisisrequiredtomakeexplicitwhathasbeenimplicitlypresumedinthepracticesthatconflict(forinstance,betweenrelianceonbothgeometricintuitionandonalgebraicdefinitionsinanalyticgeometry)andtoresolvetheconflictwherepossiblebyreferencetovaluesdrawnfromthewiderframeworkoftheformoflifeinwhichthepracticesareembodied.Wittgensteinhimselfseemstohavehadaverynarrowconceptionofthesourceofconceptualproblems(anideaparalleltoKant’sviewthatthehumanmindispronetodeludeitselfwithillusionsbasedonmistakingideasofreasonforrealthings).Hethoughtthatoureverydaylanguagetendstomisleadusintobeliefinabstractobjects.Wetendtobelievethatwhereverwehaveanexpressionthatfunctionsgrammaticallyasaname,theremustbesomethingthatitnames.Analysisproceedsbyrevealingthelogicalgrammarofproblematicnouns–thegrammarimplicitinthepracticesofusewhichdeterminetheircontent–andshowsthattheapparentnamesdonotfunctionasnames.InthiswayWittgensteinarguesthatneithernumbersnorexpressionsforprivatesensationsservetostandforobjects.Butsociologistsofknowledge,evenwhentakingcuesfromWittgenstein,donotnecessarilylimitthemselves,inadvance,con-cerningthepossiblesourcesofconceptualproblems.Wittgensteinthoughtthattheroleofthephilosopherinmathematicsshouldbethatofconceptualanalyst,notreviserof368\nPHILOSOPHYOFMATHEMATICSpractice.However,hisownanalysesledhimtobehighlycriticalofsettheoryandoftheuseofinfinitisticmethods.Indeed,itishardtoseehowanalysesintendedtorevealhowlanguagemisleadsuscouldfailtobecriticalintheirintent,byaimingtoreviseaspectsofpractice.Onlyifphilosopherscouldworkfromavantagepointwhollyoutsidethepracticestheyanalysecouldtheiractivityavoidbeingpotentiallyrevisionary.But,sinceparticipationinpracticeis,forWittgenstein,thegroundofunderstandinglin-guisticexpressionsemployedinthem,philosopherscannotbegintheirworkfromastanceofdistancednon-engagement.Wittgensteinlikenedtheroleofphilosophertothatofanthropologist;and,indeed,thephilosopherfacesthesameproblemsofreflex-ivityastheanthropologist–thephilosopher,andhisorherstudies,becomepartoftheobjectheorsheisstudyingandsomodifyit.Thisreflexivityisstrikinglyillustratedbytheworkofthemathematician–philosophers,suchasFrege,Russell,HilbertandBrouwer,whosefoundationalstudiesprofoundlychangedthewaymathematicsisdone.Inanefforttoresolveconceptualproblemsencounteredwithinexistingmathematicalpractices,theysoughttoclarifyandcodifytheassumptionsbuiltintothosepractices.Buttodosotheyhadtoforgenewmathematicalinstrumentsandinventnewterminologytoprovideexplicitexpressionforprinciplesandstructuresthathadpreviouslyhadonlypracticalembodiments.Evenastheyexaminedthebasesofexistingmathematicstheywereintheprocessofchang-ingit,andtheyleftasubsequentgenerationtoexaminethesebasesofthemathemat-icssoproduced;thatpatternofdevelopmentisdiscussedbyBachelard(1928).Ifwestandbackfromthefoundationalenterpriseanditstendencytofocusonlimitedsegmentsofpuremathematics,itisevidentthattherearealsoforcesforchangeinmathematicsthatderivefromitsapplications.Oncemathematicsisexaminedinthecontextoftheworldinwhichitisemployed,questionsaboutitsauthoritativestatusmaybetreatedasquestionsofsocialorpoliticalauthority.Suchauthoritywillbelinkedtoitsutilityandthepowerthatitconfers,andthefactthatitreflectsvaluesoftwenty-first-centuryculture,butnotthoseofPlato’sAthens(Latour1987:ch.6).Someresearchconductedfromthispointofviewsuggeststhatitmayindeedbefruitfultoexploretherelationsbetweenchangingstandardsofproof,thedemandsplacedonaproof,andchangingpatternsofapplication;suchchangesmayalterthekindsofreliabilityaboutwhichassuranceissought.Forexample,theincreaseduseofcomput-ersinscience(forcomputermodelling)andintechnology(particularlythemilitarytechnologyofsmartweapons)hasalreadyrenewedthequestionofwhatconstitutesaproofandhasledtothequasi-empiricalexplorationoffractalgeometryviacomputergraphics.Inwhatcircumstances,ifany,mighthundredsofpagesofcomputerprintoutconstituteaproof?MacKenzie(1993)reportsonthecase,whichnearlycametocourt,oftheVIPERcomputerchip.Itsmanufacturersclaimedtohaveaproofofitsreliability.Aprospectivepurchaserchallengedthisclaim,sayingthatwhatwastoutedasaproofwasnot.Computerchipsarephysicalembodimentsofcomplexmathematicalstructures.Assumingthatthephysicaltranscriptioniscorrect,therestillremainsthequestionofwhetherachipmadetothatdesignwillactuallyperformthefunctionforwhichitwasdesigned.Itwouldbedesirabletohaveaproofansweringthisquestioninadvanceofthechip’sdeploymentin,forexample,aweaponssystem.Moreover,sinceclaimsmadeinthiscontextarepartofcommercialtransactions,theyalsocomeunderlegaljurisdiction.Thismeansthatthequestionofwhatdoesordoesnotconstitutea369\nMARYTILESproofneedstobeagreedinsuchawaythatalegalrulingispossible.Herearetwosocialdemandsonthenotionofproof.Theyaredemandsthatarenotguaranteedtobecom-patibleorrealizable,andtheyaredemandsthatarisebecauseofthekindofapplicationgiventoaparticularsub-branchofmathematics.Thatbranchitselfowesmuchofitsdevelopmenttoitsintendedmilitaryapplications.Whether,how,ortheextenttowhichthesedemandscanbemet,isamatterofreflectivemathematicalandconceptualanaly-sis,combinedwithtechnicalresearch.Suchdemandspushpuremathematicalworkindirectionsthatitmightnototherwisehavetaken,andtheyforcechangesintheconceptofproof.Whataretheconstraintswithinwhichsuchchangecanoccur?Arethereanyobjectiveconstraints?Ifso,whatistheirorigin?Thesearethekindofquestionsthatcometotheforeiffoundationalismisdiscarded.Theygobacktothebasicproblemofthephilosophyofmathematicsandleadustolookafreshatthenexusofproof,discoveryandapplication.Thisnexushasbynowbecomeverydiverseandverycomplexbecauseoftheprominenceofmathematicsinourscienceandtechnology.Thisinturnmeansthattohaveanyunderstandingofthatscienceandtechnology,wemustalsounderstandthesourcesofthepowerofcontemporarymathematics.FurtherReadingInmanywaysFrege’sGrundlagenremainsoneofthebestintroductionstothephilosophyofmathematicsoftheearlypartofthiscentury.ThiscouldbecoupledwiththecommentarybyDummett(1991).TheselectionofreadingsinBenacerrafandPutnam(1983)isexcellentandprovidesabroad-basedintroductiontofoundationalapproachestothesubject.Crucialpapersfromtheperiod1879–1931aretobefoundinVanHeijenoort(1967).Shanker(1988)isausefulcollectionofpapersassessingthesignificanceofGödel’stheorem.Hofstadter(1979)providesanexcellentintroductiontotheconceptofaformalsystemandasenseofbothitspoweranditswiderconceptualconnections.Tiles(1989)providesanhistoricalintroductiontophilosophicalaspectsofsettheoryinrelationtothecontinuumhypothesis.AnhistoricaloverviewofdevelopmentsinmathematicsandtheirimpactonviewsaboutitsstatusisprovidedbyKline(1980).Anotherwayintothesubjectisthroughexplanationsoftheconceptualchallengesposedbyandfacedinmathematicsthatarewrittenfornon-mathematicians.Ofthese,Dantzig(1930)remainsausefulintroductiontotheintriguingaspectsofnumbers.ThemorerecentReid(1992)coverssomeofthesameground.DavisandHersch(1983)givesabroadersamplingofthekindsofissuesraisedbymathematicsandoftheirhistory,asdoesStewart(1987).Stewart(1989)isanexcellent,eminentlyreadablehistoricaltreatmentoftherelationbetweenchaosandorder,anditkeepsmathematicsinthecontextofitsrelationtophysicaltheory.ItbeginswithadiscussionoftheNewtonianvisionofaclockworkuniverseandismainlyconcernedwiththeoriginsofchaostheory.Somewhatdifferentincharacter,butaninterestingchallengetofoundationalistapproaches,isLakatos(1976).Readersmightbeinterestedintheearlyhistoryofphilosophicalreflectionsonmathematics.MostofAristotle’sremarksaretobefoundinMetaphysics(1941b:I,M,N).MuchofPlato’spositionwegetfromAristotle’sreports,theoriginalshavingbeenlost.ButtherearesomeremarksinRepublic(1961a:Books6and7).TheTimeaus(1961b)was,however,theworkwhichinspiredmanyoftheneo-Platonists,viaPlotinus.Yates(1964)andFrench(1972)givesomeideaofthecharacterofRenaissanceneo-Platonism.Caspar’s(1959)biographyofKeplergivesinsightintothewayinwhichhisvisionoftheplaceofmathematicsintheworldwasintegratedwithhisastronomicalwork.370\nPHILOSOPHYOFMATHEMATICSReferencesAristotle,1941a:PosteriorAnalytics.InR.McKeon(ed.)TheBasicWorksofAristotle.NewYork:RandomHouse.——1941b:Metaphysics.InR.McKeon(ed.)TheBasicWorksofAristotle.NewYork:RandomHouse.Augustine,St,BishopofHippo1955:TheProblemofFreeChoice(translatedbyM.Pontifex).Westminster,MD:NewmanPress.——1990:DeMusica(commentarybyPizzaniandMelanese).Palermo:EdizioniAugustinus.TranslationinL.Schoppetal.(1947–)TheFathersoftheChurch,vol.4.Washington,DC:CatholicUniversityofAmerica.Bachelard,G.1928:Essaisurlaconnaissanceapprochée.Paris:J.Vrin.Benacerraf,P.andPutnam,H.(eds)1964:PhilosophyofMathematics:SelectedReadings.Engle-woodCliffs,NJ:Prentice-Hall.Revisededition,Cambridge:CambridgeUniversityPress(1983).Berkeley,G.1947–57[1734]:TheAnalystorADiscourseAddressedtoanInfidelMathematician.InTheWorksofGeorgeBerkeley,BishopofCloyne(editedbyA.A.LuceandT.E.Jessop).London:ThomasNelsonandSons.——1995[1710,1713]:PrinciplesofHumanKnowledgeandThreeDialogues(editedandintro-ducedbyHowardRobinson).Oxford:OxfordUniversityPress.Bloor,D.1973:WittgensteinandMannheimontheSociologyofMathematics.StudiesintheHistoryandPhilosophyofScience,4,2,173–91.Bostock,D.1974:LogicandArithmetic,Vol.1:NaturalNumbers.Oxford:ClarendonPress.——1979:LogicandArithmetic,Vol.2:RationalandIrrationalNumbers.Oxford:ClarendonPress.Brouwer,L.E.J.1975:ConsciousnessPhilosophyandMathematics.Proceedingsof10thInternationalCongressofPhilosophy,Amsterdam1940.ReprintedinL.E.J.Brouwer,CollectedWorks,Part1(editedbyA.Heyting).Amsterdam:North-Holland.Brown,J.R.1991:TheLaboratoryoftheMind:ThoughtExperimentsintheNaturalSciences.London:Routledge.Cantor,G.1895:BeiträgezurBegrundungdertransfinitenMengenlehre,PartI.MathematischeAnnalen,46,481–512.TranslatedbyP.E.B.Jourdain(1915)ContributionstotheFoundingoftheTheoryofTransfiniteNumbers.Chicago:OpenCourt,andNewYork:Dover(1955).Caspar,M.1959:Kepler.London:Abelard-Schuman.Chihara,C.1990:ConstructibilityandMathematicalExistence.Oxford:ClarendonPress.Cohen,P.J.1966:SetTheoryandtheContinuumHypothesis.NewYork:W.A.Benjamin.Dantzig,T.1930:Number:TheLanguageofScience.London:AllenandUnwin.Davies,P.C.W.andBrown,J.(eds)1988:Superstrings:ATheoryofEverything.Cambridge:CambridgeUniversityPress.Davis,P.J.andHersh,R.1983:TheMathematicalExperience.Harmondsworth:PenguinBooks.Dedekind,R.1963[1888]WassindundwassollenZahlen?Braunschweig:Vieweg.TranslatedbyW.W.Beman(1901)EssaysontheTheoryofNumbers.Chicago:OpenCourt,andNewYork:Dover(1963).Descartes,R.1985–91a[1629]:RulesfortheDirectionofourNativeIntelligence.InJ.Cottingham,R.StoothoffandD.Murdoch(eds)ThePhilosophicalWritingsofDescartes.Cambridge:CambridgeUniversityPress.——1985–91b[1641]:MeditationsonFirstPhilosophy.InJ.Cottingham,R.StoothoffandD.Murdoch(eds)ThePhilosophicalWritingsofDescartes.Cambridge:CambridgeUniversityPress.Detlefsen,M.1986:Hilbert’sProgram.Dordrecht:Reidel.Dummett,M.1973:ThePhilosophicalBasisofIntuitionisticLogic.InProceedingsoftheLogic371\nMARYTILESColloquium,Bristol,July1973,ed.H.E.RoseandJ.C.Shepherdson.Amsterdam:North-Holland(1975).ReprintedinP.BenacerrafandH.Putnam(eds)1983:PhilosophyofMathematics:SelectedReadings.Cambridge:CambridgeUniversityPress.——1991:Frege:PhilosophyofMathematics.London:Duckworth.Euclid1956:TheThirteenBooksofEuclid’sElements(translatedbyT.L.Heath).Cambridge:CambridgeUniversityPress.Secondedition,NewYork:Dover.Field,H.1980:ScienceWithoutNumbers.Oxford:Blackwell.Frege,G.1884:DieGrundlagenderArithmetik.Breslau:Koebner.TranslatedbyJ.L.Austin,1959:TheFoundationsofArithmetic.Oxford:Blackwell.——1893:GrundgesetzederArithmetik,Band.1.Jena:VerlagHermannPohle.TranslatedbyM.Furth,1964:TheBasicLawsofArithmetic.BerkeleyandLosAngeles:UniversityofCaliforniaPress.French,P.J.1972:JohnDee:TheWorldofanElizabethanMagus.London:Routledge.Gödel,K.1931:UberformalunentscheidareSätzederPrincipiaMathematicaundverwandterSysteme.MonatsheftefürMathematikundPhysik,38.Translatedas‘OnformallyundecidablepropositionsofPrincipiaMathematicaandrelatedsystems’,inJ.vanHeijenoort(ed.)1967:FromFregetoGödel:ASourceBookinMathematicalLogic,1879–1931.Cambridge,MA:HarvardUniversityPress.——1940:TheConsistencyoftheContinuumHypothesis.AnnalsofMathematicsStudies,3.Heyting,A.1956:Intuitionism:AnIntroduction.Amsterdam:North-Holland.SelectionsreprintedinBenacerrafandPutnam(1964)and(1983).Hilbert,D.1899:GrundlagenderGeometrie.Stuttgart:Teubner.SeventheditiontranslatedasFoun-dationsofGeometry.Lasalle,IL:OpenCourt(1971).——1926:OntheInfinite.DeliveredbeforeacongressoftheWestphalianMathematicalSociety.OriginallypublishedinMathematischeAnnalen,95,161–90.EnglishtranslationinBenacerrafandPutnam(1964)and(1983).Hofstadter,D.R.1979:Gödel,Escher,Bach:AnEternalGoldenBraid.NewYork:BasicBooks;Hassocks,Sussex:HarvesterPress.Kepler,J.1940[1619]:HarmoniceMundi,GesammelteWerke,Vol.6.Munich:H.Beck.TranslatedasTheHarmoniesoftheWorld.Chicago:EncyclopaediaBritannica(1990).Kline,M.1980:Mathematics:TheLossofCertainty.Oxford:OxfordUniversityPress.Körner,S.1960:ThePhilosophyofMathematics:AnIntroductoryEssay.London:Hutchinson.Koyré,A.1973:TheAstronomicalRevolution.Paris:Hermann;London:Methuen.Lakatos,I.1976:ProofsandRefutations.Cambridge:CambridgeUniversityPress.Latour,B.1987:ScienceInAction.Cambridge,MA:HarvardUniversityPress.MacKenzie,D.1993:NegotiatingArithmetic,ConstructingProofs.SocialStudiesofScience,23,1,37–65.Maddy,P.1990:RealisminMathematics.Oxford:ClarendonPress.Nerlich,G.1976:TheShapeofSpace.Cambridge:CambridgeUniversityPress.Newton,I.1931[1730]:OpticksorATreatiseoftheReflections,Refractions,InflectionsandColoursofLight.London:BellandSons;NewYork:Dover(1952).Plato1961a:Republic.InE.HamiltonandH.Cairns(eds)TheCollectedDialoguesofPlato.Princeton,NJ:PrincetonUniversityPress.——1961b:Timaeus.InE.HamiltonandH.Cairns(eds)TheCollectedDialoguesofPlato.Princeton,NJ:PrincetonUniversityPress.Quine,W.V.O.1963:SetTheoryanditsLogic.EnglewoodCliffs,NJ:Prentice-Hall.——1969:EpistemologyNaturalized.InOntologicalRelativityandOtherEssays,NewYorkandLondon:ColumbiaUniversityPress.——1970:PhilosophyofLogic.EnglewoodCliffs,NJ:Prentice-Hall.372\nPHILOSOPHYOFMATHEMATICSReid,C.1992:FromZerotoInfinity.Washington,DC:TheMathematicalAssociationofAmerica.Russell,B.1919:IntroductiontoMathematicalPhilosophy.London:AllenandUnwin.Shanker,S.G.(ed.)1988:Gödel’sTheoreminFocus.Beckenham:CroomHelm;NewYork:Methuen.Spinoza,B.1992:EthicswiththeTreatiseontheEmendationoftheIntellectandSelectedLetters.Indianapolis,IN:Hackett.Stewart,I.1987:TheProblemsofMathematics.Oxford:OxfordUniversityPress.——1989:DoesGodPlayDice?TheMathematicsofChaos.Oxford:Blackwell.Tiles,M.1989:ThePhilosophyofSetTheory:AnHistoricalIntroductiontoCantor’sParadise.Oxford:Blackwell.VanHeijenoort,J.(ed.)1967:FromFregetoGödel:ASourceBookinMathematicalLogic,1879–1931.Cambridge,MA:HarvardUniversityPress.Wang,H.1968:BeyondAnalyticPhilosophy:DoingJusticetoWhatWeKnow.Cambridge,MA:MITPress.Whitehead,A.N.andRussell,B.1913:PrincipiaMathematica.Cambridge:CambridgeUniversityPress.Wittgenstein,L.1922:TractatusLogico-Philosophicus(translatedbyC.K.Ogden).London:Routledge.——1963:PhilosophicalInvestigations.Oxford:Blackwell.——1967:RemarksontheFoundationsofMathematics.Oxford:Blackwell.Wright,C.1983:Frege’sConceptionofNumbersasObjects.Aberdeen:AberdeenUniversityPress.Yates,F.A.1964:GiordanoBrunoandtheHermeticTradition.London:Routledge.DiscussionQuestions1Inwhatsense,ifany,ismathematicsobjectivelytrue?2Doesmathematicsneedfoundations?3Howcantheoreticalreasoningaboutmathematicalobjectsbebothsosecureandsouseful?4Ismathematicalknowledgedifferentfromknowledgeinotherdisciplines?5Dowehavearighttoexpectmathematicalresultstohaveapplicationsandtobereliableintheirapplications?6Doesanyaccountofmathematicsdependuponapriorchoiceofmetaphysicalperspective?Cantherebemathematicswithoutmetaphysics?7Isgroundingmathematicsinaprioriformsofintuitionanimprovementovergroundingitininnateideasorinnatestructuresoftheintellect?8Istheinfiniteinmathematicspotentiallyinfiniteratherthanactuallyinfinite?Whatdifferencewouldthismake?9DoesEinstein’suseofnon-Euclideangeometriesshowthatgeometrydoesnotgiveaprioriknowledgeofthestructureofthephysicalworld?10Isthecorrectnessofmathematicsaspartofascientifictheoryanylessanempir-icalquestionthanthecorrectnessofascientifictheory?11Whatistheimportanceofsetsorclassesinouraccountofmathematics?12Howcanwetellwhetherarealistaccountofmathematicsisthebestexplanationofthesuccessofapplicationsofmathematicstotheworld?Evenifitis,doesthisprovideagoodreasontoacceptarealistontology?373\nMARYTILES13Howcanagraspoftheinfinitebegroundedinapprehensionofthefinite?14IsitalwaysthecasethatPistrueorthatnot-Pistrue?15Whatistheplaceofeffectivecomputationalproceduresinmathematics?16Canextendingoursuccessatproblemsolvingjustifytheintroductionofnewmathematicalsymbols,eveniftheydonothaveanydirectrepresentationalcontent?17Ifwediscardfoundationalism,whataretheconsequencesforouraccountofproof,discoveryandapplication?18Arethereanyspecificallymathematicalprinciplesofinference?19Cantherebeanadequatefiniterepresentationofsomethinginfinite?20Ifspaceandtimearepartofnature,ismathematicsanaturalscience?Ifso,canweunderstandmathematicalcertainty?374\n12PhilosophyofSocialScienceMARTINHOLLISPhilosophicalproblemsarisefromourattemptstogaininsightintoourselvesandoursocialinstitutions,tocomprehendothersocieties,andtomakesenseofsocialchange.Thefollowingchapterbringstogetheracomplexarrayofquestionsinthephilosophyofsocialsciencebycriticallyexploringtwocrucialdistinctions:scientificexplanationversusinterpretativeunderstandingandindividualismversusholism.Relationsaretracedwithothermajorfieldsofphilosophy:EPISTEMOLOGY(chapter1),METAPHYSICS(chapter2),PHILOSOPHYOFMIND(chapter5),ETHICS(chapter6),POLITICALANDSOCIALPHILOSOPHY(chapter8)andPHILOSOPHYOFSCIENCE(chapter9).Discussionsofrecentworkonrationalityandsocialrelativismleadtoafinalagendaofunresolvedproblems.IftheclimateofAfricabecamesuddenlycoolerandwetter,thesocialeffectswouldbeenormous.Agriculturewouldchangecharacterassomecropsbecameeasiertogrowandothersharder.Asthedemandforbothsortsshifted,somepeopleandgroupswouldbecomericherandotherspoorer.Assomeroseinsocialesteemandothersfell,socialrolesandrelationswouldchange.Withshiftsinstatusandthedistributionofpower,politicaleffectswouldstarttoemerge.Weneednotspeculatefurthertoseethatwholesocietiescouldbetransformed.Suchchangeswouldnotsimplybedictatedbythechangeinclimate,howeverlarge.Whathappenedwoulddependonhowpeopleandgroupsresponded.Somemilletfarmerswouldfarebetterthanothermilletfarmers,andsomekindsoffarmerbetterthanotherkinds.Indifferentplaces,traders,middlemen,bureaucratsorholdersofpublicofficewouldplaydifferentparts.Again,withoutspeculatingfurther,wecanseethatanyuniformitiesintheresponsewouldhavetoliedeepbelowthesurfaceofeventsandthat,evenso,localconditionsandindividualinitiativeswouldmatter.Observerstryingtomakesenseofeventswouldhave,verybroadly,twodirectionsofapproach.Onewouldbetoconnectthevarietyofreactionwithdifferencesinsocialstructureandorganization,therebypresumingthattherearesocialsystems,whoseresponsestopressurearetransmittedtotheircomponentsinwayswhichvarywiththesystem.Thisapproachcouldbetermed‘topdown’.Theotherwouldbetotreatmacroscopicchangesasfusionsofmicroscopicchangesbyaprocesswhosefinalelementswereatleastlocalandperhapsindividual.Thisapproachcouldbetermed\nMARTINHOLLISExplanationUnderstandingHolismSystemsCulturesRationalSubjectiveIndividualismchoicesmeaningsFigure12.1Thefourkeystoanalysingsocialaction.‘bottomup’.Itisnotimmediatelyplainwhethertheseapproachesareconflictingorcomplementary.Meanwhiletheideaof‘makingsenseofevents’isnotaclearoneandwecanuse-fullyglossit,againverybroadly,intwoways.Onereliesonnotionsofexplanationthatarefamiliarfromthenaturalsciencesandtheirphilosophy,wheretheaimistoiden-tifythecauseswhichproduce,generateorperhapsaremerelycorrelatedwithwhathappensnext.Theotherappealstoideasofunderstandingthatarefamiliartostudentsofhistoryoreverydaylife,whoseektoidentifythemeaningofaction(insomesenseofthatelusivephrase).Heretooitisnotobviousinadvancewhetherexplanationandunderstandingcanfinallybecombined,howeversharplywemaywanttocontrastthematthestart.Thelasttwoparagraphsmightsuggestthat‘topdown’goeswithexplanationand‘bottomup’withunderstanding.Thatisnottheintendedmessage,however.Thischaptersetsofffromapresumptionthatanalysiscanproceedineitherdirectionbyeithermethod.Abstractlyspeaking,westartwiththetwo-by-twomatrixshowninfigure12.1,whichprovidestheframeworkforthechapter.Ignoringwhatiswritteninthefourboxesforthemoment,letustakenoteofthecolumnsandrows.Ourleadingquestionwillconcernthecolumns–theclaimsofExplanationandUnderstandingtoholdthekeytosocialscientificknowledge.Arethephenomenaofthesocialworldtobeexplainedafterthemannerofthenaturalsciencesorunderstoodinsomemannerspecialtothestudyofsociallife?Thatisacomplexquestionwiththreeaspects.Oneaspectismethodological,invitingphilosophicaldiscussionofscientificmethodandofwhetherthereisasingle,universalmethodfornaturalandsocialsci-encesalike.Oneaspectisepistemological,posingtheProblemofKnowledgeingeneralandtheProblemofOtherMindsinparticular.Thethirdaspectisontological,todowithwhatthereis(or‘thestudyofbeing’,totranslatetheancientGreektermsinvolvedliterally).Weshallbeaskingwhatgeneralkindsorcategoriesofphenomenathereareinthesocialworld,whatsortofmethodisbestabletorevealhowitworksandwhatwarrantsclaimstoknowledgeofthesematters.Therowsposequestionsofanalyticalpriority.Theybidusthinkwhetherthesocialworldistobeanalysedfinallyintoparticulars,especiallyagentsandactions(‘indi-vidualism’),orintostructures(‘holism’),orwhetherbothhaveanirreducibleclaimtofeatureinthefinalaccount.Indecidingwhetheranalysiscanbestproceed‘topdown’or‘bottomup’weshallagainbeengagedinmethodology,EPISTEMOLOGY(chapter1)andONTOLOGY(chapter2).376\nPHILOSOPHYOFSOCIALSCIENCEBut,asIhavejustsaid,thechapterischieflyorganizedroundthedistinctionbetweenExplanationandUnderstanding.Section1willexplorethecasefornaturalisticexpla-nationinthesocialsciences,thethemebeingthat,sincehumanbeingsbelongtothenaturalorder,thesocialworldpresentsnofinalobstacletothemethodsthatservethenaturalsciencessowell.Section2willmountarivalcaseforinterpretativeunder-standing,thethemeherebeingthat,sincesocialactorsviewsociallifefromtheinside,thesocialsciencesmustmakesenseofitbydoinglikewise.Section3willexaminethescopeforcollaborationbetweenexplainingandunderstanding,therebyidentifyingsomefertilephilosophicalquestionsthatremainunsettled,astheshortconcludingsectionwillpointoutbywayofsummary.1ExplanationThesolefoundationforbeliefinthenaturalsciencesisthisidea,thatthegenerallawsdic-tatingthephenomenaoftheuniversearenecessaryandconstant.Whyshouldthisprin-ciplebeanythelesstrueforthedevelopmentoftheintellectualandmoralfacultiesofmanthanfortheotheroperationsofnature?Thisbeguilingquestioncomesfromabookwrittenin1794amidtheEuropeanfermentofideasknownastheEnlightenment.TheauthoristheMarquisdeCondorcet(1743–94),mathematician,philosopherandsocialscientist,andthebookofferswhatitstitlecallsaSketchforaHistoricalPictureoftheProgressoftheHumanMind(1955).Condorcettracesthehistoricaldevelopmentofhumansocietyandrejoicesintheamaz-inglyrapidprogressmadeinthetwocenturiessincethescientificrevolution.Helooksforwardtoanenlightenedfuturewhere‘thesunwillonedayshineonlyonfreemenwhoownnootherlightthantheirreason’.Histheme,andthekeytothisprogress,isrecognitionthat‘Truth,happinessandvirtueareboundtogetherbyanindissolublechain’.Suchoptimismmayringhollowtwocenturieslater;andCondorcethimselfdiedinprisonshortlyafterfinishingthebook,avictimoftheFrenchRevolutioninwhichheferventlybelieved.Buthislinkingoftruth,happinessandvirtuestemsfromavisionofhumanityandsocialorderwhichstillinfluencesthesocialsciencesandtheirphiloso-phy.IfnatureincludesHUMANNATURE(pp.672–3),theprogressofknowledgewillfindnoradicalobstaclein‘theintellectualandmoralfacultiesofman’.Anewscienceofhumanbeingswillbringusknowledgeofhumannatureandhencethepowertoschoolandsatisfyourdesires.Virtuecomesintothestoryinsofarasweusethispowerforthebenefitofhumanity,thuscreatingafreeandenlightenedsocialorder.Thequotationservestointroducenaturalism,togiveititscurrentname,underthethreeaspectsmentionedabove.Theontologyisnaturalistic.Nature,aspicturedbyCondorcet,isanorderedrealmofobjectsgovernedbynaturallaws.Itincludes‘man’and‘theintellectualandmoralfacultiesofman’,whosedevelopmenthasbeenamong‘theotheroperationsofnature’.Althoughthisstopsshortofsayingthathumanbeingsaresolelyphysicalcreatures,itdoesimplythatthehumanmindisnotthesortofnon-physicalthinkingsubstance(rescogitans)claimedbyDESCARTES(1596–1650)(chapter26)andCartesiandualists.Asacontemporaryputit,‘Manisnotfashionedoutof377\nMARTINHOLLISamorepreciousclay;Naturehasusedonlyoneandthesamedough,inwhichshehasmerelyvariedtheleaven.’Thisexactlycapturesthemovingspiritofnaturalisminthehumansciences.Theclaimthatthelawsofnaturearegeneral,dictatethephenomenaoftheuni-verseandare‘necessaryandconstant’ushersinanaturalisticideaofscientificknowl-edgeandhowtoacquireit.ThemethodwhichCondorcettooktobethefoundationofthescientificrevolutionwasoneofidentifyingthehiddenlawsofnatureandusingthemtoexplainphenomena.Byreasonwecandiscernthelawsandthusdetecttheunderlyingcausalorderoftheuniverse.Ifthisbroadsketchwereenoughtoworkwith,wecouldproceedatoncetoques-tionsabouttheintellectualandmoralfacultiesofhumanbeings.But,asthechaptersonEPISTEMOLOGY(chapter1)andthePHILOSOPHYOFSCIENCE(chapter9)makeclear,thesketchcanbefilledinseveralways.Soobjectionstoonenaturalisticaccountofthesocialrealmmaynotbeobjectionstoall.Althoughreaderswillneedtorefertothoseotherchapterstomakefullsenseofthisone,wecanusefullypickoutsomekeyissueswhichareespeciallyrelevanttothephilosophyofsocialscience.Condorcet’sremarkthatthelawsofnatureare‘necessaryandconstant’gesturestoanolddispute,whichhasaffectedthesocialsciencesfromthestart.OntheonehandthereistheRATIONALIST(chapters26and27)contentionthatnatureisconstantbecauseitslawsholdofnecessity;ontheother,thereistheempiricistretort,stillnovelattheendoftheeighteenthcentury,thatthereisnothingnecessaryaboutnatureitselfandthatany‘necessity’isthereforeonlyamatterofhowourmindswork.Weattributenecessitytowhat,ifwereflectcarefully,weknowonlytobeconstantinitsworkings.Althoughtoday’sphilosophersusuallydeemthiscontrasttoostarkandsimple,asweshallnoteinamoment,itremainsausefulpointertosomeradicaldifferencesofapproachtosocialtheory.Therationalistcontentiongoeswithastronglydeterministvisionofnatureasasystemdrivenbyhiddenforces.ThisvisionhadinspiredDescartesandNewton(1642–1727),theleadingseventeenth-centuryphilosophersofthescientificrevolu-tion.Inapopularimageofthetime,theworldislikeaperfectwatch,with‘WheelsandSpringssooutofsightthatwehavebeenlonga-guessingatthemovementoftheUniverse’(Fontenelle1929).Inthemechanicaluniverse,whateverhappensmusthappen,giventhepreviousstateandthelawsofnature,andscienceadvancesbydemonstratingthesenecessities.Scientificknowledge,therefore,cannotdependonSENSEEXPERIENCE(chapter26),since,inthewordsofanoldadage,‘thesensesrevealnonecessities’.Itdependsinsteadonourintellectualfacultyof‘intuition’,aidedbymathematicalmodelswhichallowdemonstrativereasoning.NewtonianmechanicsandCartesiangeometricalphysicsareclassicexamples.Appliedtothesocialworld,thisambitiousschemehasstrikingimplications.Itsfirmdistinctionbetween‘phenomena’and‘laws’suggestssomekindofsocialorpsycho-logicalphysics,ofwhoseexplanatoryelementswe,asactors,maybeunaware.More-over,ifwetooarelikewatchesdrivenbyhiddenwheelsandsprings,ourcherishedfreewillmayturnouttobeanillusion.Manysocialscientistshaveacceptedtheseimplica-tions,startingwithThomasHOBBES(1588–1679)(chapter28),whoseinfluenceonsocialandpoliticalphilosophyhasbeenpervasive.InLeviathanHobbes(1990)presentshumanbeingsasmechanicalindividualsforwhomthewillismerely‘thelastappetite378\nPHILOSOPHYOFSOCIALSCIENCEindeliberating’(ibid.:ch.5).Wearedrivenbyanurgeforself-preservationand‘aper-petualandrestlessdesireforpowerafterpower,thatceasethonlyindeath’(ibid.:ch.11).Weknowthis,hemaintains,bydiscerning‘lawsofreason’whichaccountforhumanbehaviour.Yet,byharnessingthisknowledge,wecanlearntomakeandmain-tainthesocialorderwithoutwhichhumanbeingswilldestroyoneanother.‘Reasonisthepace,increaseofsciencetheway,andthebenefitofmankindtheend’(ibid.:ch.5).Leviathanremainsaleadingexampleofanindividualistaccount,inwhichsocialarrangementsemergefromtheinterplayofindividualsrationallyseekingself-preservation.Foranotableexampleofastructuralaccount,IrecommendthePrefacetoAContributiontotheCritiqueofPoliticalEconomy(1971)byKarlMARX(1818–83)(chapter34).Therewearetoldthattheeconomicstructureofsocietyis‘therealfoundation,onwhichrisesalegalandpoliticalsuperstructureandtowhichcorresponddefiniteformsofsocialconsciousness...Itisnottheconsciousnessofmenwhichdeterminestheirbeingbut,onthecontrary,theirsocialbeingwhichdeterminestheirconsciousness.’Heretootheappealtohiddenforcesgoeswithadenialoffamiliarnotionsoffreewill.YetMarxtoo,atleastinotherworksandothermoods,holdsthatmenmaketheirownhistory,eventhoughnotunderconditionschosenbythemselves,andthatsciencecanstrengthenthispowerbyrevealingwhatisfeasible.Questionsoffreedomanddeterminismarethusfarfromstraightforwardandweshallreturntothem.Meanwhile,HobbesandMarxbothbelongintheleft-handcolumnoffigure12.1andservenicelytoillustratethebottomandtopboxesrespectively.Butthisisnottosaythatitisyetclearwhatsenseattachesto‘necessity’intheseandinsimilaraccounts.Perhapsthereisindeedgoodreasontothinkthatlawsofnaturearenotmerely‘con-stant’,andthat,forinstance,whatgoesupmust,insomesense,comedownbecauseofthelawsofgravity.Butitisfarfromplainthanany‘must’involvedinthepowerofforcesortherelationofcausestoeffectsisthekindof‘must’involvedintruthsofmath-ematics.Evenintheseventeenth-centuryimageoftheworldasawatchthecompul-sionenvisagedisasmuchtodowithmechanismsaswithformulae.Twodistinctideasareinvolved,oneoflogicalnecessityandone,moremysteriously,ofnaturalnecessity.Althoughthisisnotthemomenttoexplorethelatter,theambiguityhelpstoclearthewayfortheriseofempiricismandthedenialthatnecessityofanykindisafeatureoftheworkingofnature.HeretheclassicsourceisDavidHUME(1711–76)(chapter31),especiallyATreatiseofHumanNature(Hume1978a),whichwasintendedtolaythefoundationsfor‘acom-pletesystemofthesciences’.Itisevident,Humesaidintheintroduction,thatallsci-encesrelatetohumannature.‘EvenMathematics,NaturalPhilosophy,andNaturalReligionareinsomemeasuredependentonthescienceofMAN;sincetheylieunderthecognizanceofmen,andarejudgedbytheirpowersandfaculties.’Henceascienceofhumanbeings,groundedinanempiricalstudyof‘Logic,Morals,CriticismandPoli-tics’,wouldcomprehendeverythingwhichcantendtotheimprovementofthehumanmind.Themethodwastobe‘experienceandobservation’,appliedto‘men’sbehaviourincompany,inaffairsandintheirpleasures’.Theaimwas‘explainingalleffectsfromthesimplestandfewestcauses’.TounderminethegrandpretensionsofReason,Humedrewasharpdistinctionbetween‘mattersoffactandexistence’and‘relationsofideas’.Mattersoffactare379\nMARTINHOLLISalwayscontingentandknowntousbyexperience.Any‘necessity’tothemcanonlybeduetoourmentalhabitofexpectingestablishedcorrelationstocontinue,sincegenuinenecessities,likethoseofmathematics,liesolelyinrelationsofideas.Thuswehavenowarrantinexperienceforholdingthatcausesnecessitatetheireffects,producethemormakethemhappen.Asfarasourknowledgeofobjectsandeventsgoes,acauseismerelyimmediatelypriorto,contiguouswithandregularlyconjoinedwithwhatwelabelitseffect.‘Real’connectionsonlyreflectourexpectationthatregularsequenceswillberepeatedinlikeconditions.Bythisaccountsocialscienceisconcernedwith‘men’sbehaviourincompany,inaffairsandintheirpleasures’,withoutanypresumptionthatscientificdeterminismdenieshumanfreedom.AsimilarlyempiricistandindividualistlineistakenbyJohnStuartMILL(1806–73)(chapter35)inASystemofLogic(1961),whoseBook6‘OntheLogicoftheMoralSciences’remainsthebestworkedoutmanifestoforempiricisminsocialscience.Chapter7ofBook6openswiththischallengingdeclaration:Thelawsofthephenomenaofsocietyare,andcanbe,nothingbutthelawsoftheactionsandpassionsofhumanbeingsunitedtogetherinthesocialstate.Men,however,inastateofsociety,arestillmen;theiractionsandpassionsareobedienttothelawsofindividualhumannature.Menarenot,whenbroughttogether,convertedintoanotherkindofsub-stance,withdifferentproperties;ashydrogenandoxygenaredifferentfromwater,orashydrogen,oxygen,carbon,andazote,aredifferentfromnerves,muscles,andtendons.Humanbeingsinsocietyhavenopropertiesbutthosewhicharederivedfrom,andmayberesolvedinto,thelawsofnatureofindividualman.Ifthatstillsoundsdeterminist,itismeantto.Mill,likeHume,takesthecompatibilistlinethatfreeactioniscausedactionwhosecauseslieinthedesiresandcharacteroftheagent,guidedbybeliefsaboutitslikelysuccessinsatisfyingthesedesires.Freeactionthuspresupposesaregularandpredictableworldandisgivenallthemorescopeassciencelearnstochartthepreciseregularities.Ifthisneatreconciliationoffreedomanddeterminismistenable,manyawkwardproblemsforascientificapproachtohumanbehaviourvanish.Herereadersmaywishtopauseandconsiderwhetheritisindeedtenable.Hobbes(1990:ch.21),Hume(1978a:I,8)andMill(1961:VI,2)offerausefulstarting-point.Hobbes,HumeandMilltogethergiveusastrongcaseforbelievingthattheintel-lectualandmoralfacultiesofmanarenoobstacletotheextensionofsciencetothesocialworld.Theyagreethat,inHobbes’swords,‘Reasonisthepace’,evenifHobbesisarationalistandHumeandMillempiricists.Theyagreethatscienceadvancesbydis-coveringcausallawsandthatdeterminismisnothreattohumanfreedom,properlyconceived.Allthreearesquarelyinthemainlineofmodernphilosophy,inshort,andthatisagood,clearwaytobegin.Butmatterscannotbekeptsostraightforward,Ifear,andwemustnextthinkmoredeeplyaboutscientificmethod.Aswesaw,Hume’scor-rectivetoaconfusionaboutnecessitywastodispensewithitaltogetherasareal-worldcomponentofcausallawsorcausalexplanation.Thatissplendidlybold.Butisitsatisfactory?DoesscienceneedsomestrongerdistinctionbetweenlawsandregularitiesthanHume’sempiricismoffers?Wecanbestfocusthequestionwiththehelpofarecent380\nPHILOSOPHYOFSOCIALSCIENCEexample.Itcomesfroma1970socialsciencetextbookonscientificmethod,wheretheauthorsuseittoillustratethestyleofexplanationwhichtheyrecommend(PrzeworskiandTeune1970).TheyintroduceatypicalFrenchworker,MonsieurRouget,andenquire,‘WhydoesM.Rouget,agetwentyfour,blondhair,browneyes,aworkerinalargefactory,voteCommunist?’Theythenremarkthat‘toexplainthevoteofM.Rouget,onemustrelyonprobabilisticstatementsthatarerelevantforvotingbehav-iourandhavebeensufficientlyconfirmedagainstvarioussetsofevidence’.Whenthisprocessofevidencegatheringiscomplete,theysuggest,anexplanationwillemergeinthefollowingform:1M.Rougetisayoungmaleworkeremployedinalargefactoryinasocialsystemwherethechurchplaysanimportantrole.2YoungworkersinlargefactoriesvoteLeftwithaprobabilityof.60to.70,andinthosesystemsinwhichtheroleofthechurchisstrong,menvoteLeftmoreoftenthanwomen;thereforeitishighlylikely(probabilityof.80)that3M.RougetvotesforapartyoftheLeft.Theguidingideahereisthattoexplainbehaviouristoarriveatreliablegeneraliza-tionsfromwhichitcouldhavebeenpredictedwithhighPROBABILITY(pp.308–10).Whethersuchgeneralizationscountascausaldependssolelyonhowwellthepredic-tionsderivedfromthemfare.InHumeanspirit,thereisnothingmoretoitbywayofcausalconnectionsorhiddenmechanisms.Aso-calledlawofnatureissimplyawell-enoughconfirmedhypothesisandtheonlytestofahypothesisisitspredictivesuccess.Thisaccountofexplanationisusuallyknownasthecoveringlawmodelanditsruleofapplicationasthehypothetico-deductivemethod.Figure12.2presentsitindia-gramaticform,taken(slightlysimplified)fromanothertextbook,RichardLipsey’sIntroductiontoPositiveEconomics(1972).Thediagramshowsaprocessofformulatinghypothesesortheories,deducingtheirimplicationsandtestingthemagainstexperience.Whenobservationconfirmsthetheory,‘noconsequentactionisrequired’;whenitdoesnot,thetheoryistobeamendedorreplaced.Again,thereisnosuggestionthatcausallawshavemoretothemthanwhatisinvolvedinthisfilteringprocess,which,inbothtextbooks,hastodosolelywithdegreesofprobability.Howsatisfactoryisthisideaofexplanation?Onthecreditside,itisassimpleandelegantasonecouldwish,especiallyifoneisimpressedbytheempiricistcaseforholdingthatonlyexperiencecanjustifyclaimstoknowledgeoftheworld.Butthereareseverallargeobjections.IfwefocusthesedoubtsonMonsieurRouget,wesoonwonderwhethertheproposedformofexplanationevenbeginstotelluswhyhevotesCommunist.Thathebelongstoagroupofwhichsome80percentvotesCommunistmayberelevant,butishardlyrevealing.Whydoage,sexandthesizeofthefactoryandthestrengthofthechurchmatter,whereascolourofhairandeyesdonot?Herearethreekindsofanswer,eachopeningfurtherlinesofenquiry.(1)Post-empiricistTheexplanationofferedisinadequatenotbecauseitfailstounearthhiddencausesbutbecauseitistoosketchy.Weneedmoreofawebofcorrelations,compactlystated381\nMARTINHOLLISDefinitionsandhypothesesaboutbehaviour(oftencalledassumptions)AprocessoflogicaldeductionTheoryPredictionsisamendedinlight(oftencalledimplications)ofnewlyacquiredfactseitherAprocessofempiricalobservationTheoryisdiscardedinfavourofasuperiorcompetingtheoryConclusion:thatthetheoryappearstobeoreitherinconsistentwiththefactsorconsistentwiththefactsIftheoryisrejectedIftheorypassesthetestnoconsequentactionisrequiredFigure12.2Thecoveringlawmodelofexplanation,whichusethehypothetico-deductivemethodtoformhypotheses,deduceimplicationsandtestthemagainstevidence.andelegantlyorderedsothattheyhavegreaterpredictivepowerforalargersetofcases.Empiricismisrighttomaintainthatexplanationsstandorfallsolelybythetestofexperienceandthusrelywhollyonwhatexperiencecantellus.Butexperi-enceisneverbruteanditsverdictsneverunambiguous.Italwaysinvolvestheappli-cationofconcepts,inthespiritofKant’sremarkthatperceptswithoutconceptsareblind.Themindisnotablanktabletonwhichexperiencewrites,butanactive382\nPHILOSOPHYOFSOCIALSCIENCEinterpreterbringingtheoriestobear.Hencetherearenofactsthatareneutralbetweenallinterpretationsand,inexplaininganything,wearechoosingwhichtheorytoprefer.Thisfirstanswerhastwoprincipalsources.OneisKarlPopper’sideaofscienceas‘ConjecturesandRefutations’,tocitethetitleofthefamouspaperinPopper(1969).Popperarguesthatobservationisnevertheory-freeandcanneverestablishorevenconfirmthetruthofahypothesisorconjecture.Butitcaninprinciplerefutethegenuinelyempiricalones(asopposedtothe‘pseudo-scientific’,which,likeMarxismorpsychoanalysis,inhisview,aresoorganizedthattheycanbesquaredwithallpossibleexperience).Butprogressispossiblebecauserefutationisamomentofobjectivetruth,evenifconfirmationisnot.Thisisnotradicalenoughfortheothermainsource,PRAGMATISM(chapter36),especiallyQuine(1953),wheretherearenomomentsofunvarnishedtruthindependentofallwebsofinterpretation.Refuta-tionsaredecisiveonlyifwechoosetoregardthemassuch.Explanationbecomesholistic,withconsequencesthatIcannottraceinthischapter.MeanwhilethereasonsforpickingoutM.Rouget’sageandsexbutnothishairandeyesareprag-matic.Theyinstancereliableregularitieswhichmeshwellwithwiderpatternsinourinterpretationofsociallife.(2)RealistNoexplanationhasyetbeenofferedeitherforM.Rouget’svoteorfortheleftistten-dencyofthegroupstowhichhebelongs.Toask,‘Why?’bidsusidentifycauses–psychologicalorsocialprocesses,orboth.Indeeditmaybethatprocessesdependinturnoncausalmechanisms–powerfulparticularswhosepowersdonotsomehowderivefrom‘lawsofnature’but,insum,aretherealityofthoselaws.Inanycase,theoldideaofstructuresandforcesexternaltotheenquiringmindremainsanec-essarypresuppositionofscience.Ifitisaskedhowwecanknowthatthereareany,andhencehowweareentitledtosuchanontology,theansweristhatsciencecanandshouldproceedbyinferencetothebestexplanation.Whetherthebestexplanationispsychologicalorsocial(orboth)raisesthe‘ver-tical’questionofindividualismandholism.Theindividualistcasehastraditionallydepended,asinHobbes,HumeandMill,onconsciousorunconsciousmotivesinher-entinhumannatureandonanaccountofhowthesemotivescangeneratesocialorganizationsthroughtheinterplayofindividuals.Insofarasthemotivesarecon-scious,thereisscopeforrationalreconstructionsofactionandsoforthekindofinterpretativeapproachthatisdiscussedbelow.Butanysuchinternalstoryisfinallysubordinatetoanaturalisticcanonofcausalexplanation.Theleadingrealistanaly-sisofcausation,whichisdescribedinchapter9onthePhilosophyofScience,willservenicely.Theholistcaseofteninvolvesafurtherclaimonbehalfoffunctionalexplanation.Liketheplanetarysystem,whichismechanical,oratermitecolony,whichisorganic,socialsystemshavevariousequilibriumconditions.Whentheseconditionsarebreached,thesystemtendstocorrectthedisturbanceorshifttowardsafreshequilibrium.Thebehaviourofpartsofthesystemcanthusbeexplainedbyfunctionaldemandsarisingfromthepropertiesofthesystemasawhole.Thetreat-mentoftheclassstruggleinacrudeversionof‘vulgar’Marxism,wherethe383\nMARTINHOLLISconsciousnessofhumansiswhollydeterminedbytheirsocialbeing,isaclassicexample.HereM.Rouget’smembershipoftheindustrialproletariatiscrucial(althoughtheideologypurveyedbythechurchmightmutetheconflictbetweentheeconomicforcesandrelationsofproduction).ClassictooaresomeoftheinstancesofsocialexplanationgivenbyEmileDurkheim(1858–1917)inthecourseofinsist-ingthatsocialfactsbetreatedasthings.LikeMarx,hedoesnotbelongsquarelyinthetop-leftquadrantoffigure12.1,ifhisworksarereadasawhole.ButthebriefargumentinTheRulesofSociologicalMethod(1964)thattherewillalwaysbecrimebecauseallsocietiesneedcrimestostigmatizeforthesakeofincreasingsocialintegrationisagem.Althoughfunctionalthinkingisnolongerascommonandexplicitasitwasuntilthe1960s,ithasbynomeansvanished.Itoccurslatently,forinstance,eveninthatmostindividualistofsocialsciences,neo-classicaleconomics.Herethepictureofindividualfirmspursuingtheirself-interestbyrationalcalculationistemperedbyaconvictionthat,iftheyfalter,theywillbedrivenoutofbusinessbytheoperationofmarketforces.Theseforcesapparentlyconstituteaneconomicsystemthatdemandsefficiencyand,unlesssubjectedtopoliticalinterference,achievesit,tothegeneralbenefitoftheeconomy.Withoutplungingintoadiscussionofhowseriouslyneo-classicistsreallytaketheideaofmarketforceswithinaneconomicsystem,wecanplausiblysuggestthisasanexampleofexplanatoryrealismwithbothpsychologicalandsocialcomponentsandfunctionalistaspects.M.Rouget’sinterestsneednotbeidenticalwiththoseofthefirmforwhichheworks.(3)InterpretativeItisstrangetosuppose,asPrzeworskiandTeuneseemto,thatwecantrytoexplainM.Rouget’svotewithoutenquiringhowheseeshissituation.Nodoubtwecanmakesenseofwhattermitesdowithoutinterviewinganytermites.ButM.Rougetisnotatermiteandhissocialworldisnotanexternalstructureinwhichheisenclosed.Or,ifitislargelyexternaltohim,thatdoesnotmakeitexternaltoallitsinhabitantscollectively.Inthatcaseexplanationsneedtostart,andperhapseventofinish,byunderstandingthesituationfromwithin.Thatpromisesapotentiallydifferentaccountofwhyageandsexmatterbuthairandeyesdonot,andonewhichisnotdirectlyconsonantwithnaturalism.Thisbringsustotheideaofunderstanding.2UnderstandingMeaning,accordingtoWilhelmDilthey(1833–1911),is‘thecategorywhichispecu-liartolifeandtothehistoricalworld’.DiltheywasoneofagroupofGermanidealistsinterestedinthestudyofhistory.PromptedinpartbytheworkofHEGEL(chapter33),Diltheycametotheviewthathumanlifecanbeunderstoodonlybyapplyingcategories,like‘purpose’and‘value’,whichhavenoplaceinourknowledgeofthephysicalworld.Thesecategoriesaddresstheconnectednessoflifeandmakeusrealizethatthemeaningofitsepisodesmustbeunderstoodbyseeingitasawhole.Butnothingexternaltohumanityisinvolved.‘Lifedoesnotmeananythingotherthanitself.Thereisnothinginitwhichpointstoameaningbeyondit’(Dilthey1926:VII,24).384\nPHILOSOPHYOFSOCIALSCIENCEThisstrikesaboldnoteonwhichtoopenourdiscussionofINTERPRETATIVE(pp.247–50)understanding.Itnotonlysetsthehistoricalworldapartfromthenaturalbutalsobidsustothinkholistically.Leavingthemeritsofholismuntillater,letusstartwiththeideaofmeaning.VarietiesofMeaningWhenM.Rougetcastshisvoteinanelection,hisactionhasmeaninginseveralsenses,threeofwhichareespeciallyrelevant.Firstly,wecanaskwhathemeansbyit.Whatishisintention,purpose,reasonormotive?Howdoesheexpectotherpeopletointerprethisaction?Thesearequestionsaboutwhatwemightterm‘subjectivemeaning’–whatisgoingoninM.Rouget’sconsciousmind.Secondly,wecanaskwhattheactionitselfmeans–exactlywhatactionisbeingperformed,consideredfromapublicpointofview.Actionsareusuallytakenfromstock,asisclearestifwethinkaboututterances.Thewords‘pandasareblackandwhite’haveapublicmeaningwhichcanbefound,forinstance,byuseofadictionaryorphrasebook.Theydonotchangetheirpublicmeaningifthespeakeruttersthembymistakeorbelievingthattheymeansomethingelseorsup-posingthatapandaisakindofbird.Similarlywhatmoveonemakesinagamedependsontherulesofthegameandcommonlyhasnomeaningoutsidethoserules.Thesearemattersofwhatwemightterm‘intersubjectivemeaning’.Thirdly,wecanaskthesignificanceofactions,seekingperhapstosettheminawidercontextortotreatthemasasignofanunderlyingprocess.Therearemanyvariants,rangingfrombackgroundarticlesoncurrentaffairsinnewspaperstoambitiousattemptstoprobewhatishidden,aswhenFreudianstreatslipsofthetongueascluestotheunconsciousorwhenhistoriansseektotracetheunderlyingmovementofhistory.Thecommonelementisareadinesstofindmeaningswhichtheactorsgrasponlydimlyorarenotawareofatall.Wemighttermthese‘objectivemeanings’,whilenotingthatweshallneedtoenquirewhetheraccountsofunderlyingsignificancecouchedintermsveryremotefromtheactors’ownbelongundertheheadingofcausalexplanations.Philosophically,theneatestwaytomakemeaningcentralistoposetheproblemofotherminds.Howdoesonepersonorgroupofpersonsknowwhatisinthemindofanother?Thistooisaproblemofknowledgebutitdiffersfromthoseraisedearlierinthatitinvolvesa‘doublehermeneutic’(‘hermeneutic’comesfromtheancientGreekhermeneus,aninterpreter).Section1gaveusgoodreasontoacceptthatfactsdonotsimplyobtrudebutinvolveinterpretation.Hence,wemightsay,thereisasinglehermeneuticinallourknowledgeoftheworld,includingourobservationofhumanbehaviour.Butwhenweinterpretbehaviourasaction,weareinterpretingtwice:onceinidentifyingthebehaviourandagaininascribingitsmeaningasaction.Toknowwhatisinotherminds,wemustbeabletomanagethisdoublehermeneutic.Theproblem,astraditionallyposed,isthateachofushasakindofself-knowledge,whichisdeniedusinourknowledgeofothers.Weunderstandtheirthoughts,emo-tionsanddesires,themeaning(inthefirstsenseabove)oftheirwordsandactions,byinferenceswhosestarting-pointistheirbehaviour.Theseinferencescrossadividewhichitisnaturaltodescribeasonebetweenphysicalandmentalorbetweentheoutwardandinwardaspectsofaperson.Yet,unlikestandardinductiveinferences,385\nMARTINHOLLISthereisnoindependentwayofcheckingconclusionsinsomecases,soastohaveareli-ablewarrantforothers.How,then,dowemanage,epistemologically,toprogressfromasingletoadoublehermeneutic?Letusfocustheproblemontheinterpretationofbehaviourasaction.Consider,forexample,blinkingandwinking.Thereisnoobviousorimmediatephysicaldifference.Yetblinksbelongwhollytoaphysiologicalstoryofresponsetostimulus,whereaswinksareavehicleforinformation–hints,reservations,conspiracy,warnings;theyare,inshort,speech-acts.Howdoweknowablinkfromawinkandhowdoweidentifyexactlywhatawinkconveys?ThisisagoodmomenttoturntoMaxWeber(1864–1920)andthemethodologicalessayatthestartofEconomyandSociety(1922).Webersaystherethat‘thescienceofsocietyattemptstheinterpretativeunderstandingofsocialaction’.In‘action’heincludes‘allhumanactionwhenandinsofarastheactingindividualattachessub-jectivemeaningtoit’.By‘socialaction’hemeansaction‘whichtakesaccountofthebehaviourofothersandistherebyorientedinitscourse’.Thuscyclistsenteringtrafficengageinsocialaction,whereassomeoneputtingupanumbrelladoesnot.Theformertakeaccountofotheractors,thelattertakesaccountonlyoftheweather.Thereissub-jectivemeaninginbothcases,sincethatmerelyrequiresanintentionofsomekind,whereasinsocialactiontheintentionrelatestotheintentionsofothers.Webersuggeststhatidentifyingintentionsisatwo-stepprocess.Thefirstis‘directunderstanding’or‘empathy’,bywhichoneknows,forinstance,thatamanswinginganaxeischoppingwoodorthatamarksmanisaimingarifle.Inotherwords,thereisaprocessakintoperception,whichletsusidentifyintentions,andhenceactions,directly.Thesecondstepis‘explanatoryunderstanding’bywhichwemightcometoknowthatthewoodcutterisearningalivingorthemarksmantakingpartinavendetta.Thisisdonebyassigningtheactionto‘acomplexofmeanings’.Itcanbedone‘historically’,wherewetracetheexacthistoryandmotiveofaparticularaction,or‘sociologically’,wherewefillinasocialcontextlikethatofrevenge-killing,or‘ideal-typically’,whereweconductaRATIONALRECONSTRUCTION(p.436)ofwhatisgoingon.Ishallpursuethislastsuggestioninamoment.Noticethattheappealto‘empathy’effectivelyby-passestheother-mindsproblembyrefusingtotreattheidentificationofsomeactionsastheconclusionofaninferencefrombehaviour.Toreverttoourfirstexample,wedonotinferthatsomeonewinkedfrompremisesaboutthemovementsofoneeyelid;weperceivethewink,eventhoughanignorantalienobservermightperceivenothingofthesort.Thismaybeashrewdmove,sincetheproblem,astraditionallyposed,threatenstobeinsoluble.Ifonecangetawaywithit,thereismuchtobesaidforsimplyintroducingadatumlevelofactionsasadistinctivefeatureofUnderstanding.Butitclearlyrisksbeggingthequestionandcallsforfurtherthoughtabouttherelationbetweenbehaviourandaction.Meanwhile,themovemaybevulnerablefromadifferentdirection.Weberisassum-ingthatwecan,sometimesatleast,identifyintentionslike‘aimingarifle’withoutpre-supposingsocialpractices.Thisisanindividualistassumptionand,ifitiswrong,holismcomesswiftlyintoplay.ItneednotbeholismonanygrandHegelianscale.Thecounter-suggestionisonlythattoby-passthetraditionalother-mindsproblemistogodirectlytothepublicmeaningofaction.WhereasWeberstartsfromwhatanagentmeansbyanaction(ourfirstsenseofMeaningabove),aholisticaccountstartsfromoursecond386\nPHILOSOPHYOFSOCIALSCIENCEsense–whattheactionmeans.Thismayseemaminorshiftbutitgoeswithawide-spreadrecentrejectionoftheCartesiancogitoasthefoundationofourknowledgeoftheworldandofotherminds.Thebroaderthemeisthatsubjectivemeaningsarepos-sibleonlybecauseweshareaneverydayworldwithotherpeople.ThekeyphilosophicaltexthereisperhapsPhilosophicalInvestigations(1953)byLudwigWITTGENSTEIN(1889–1951)(chapter39).Insofarasactionsaretypicallymovesinsome‘game’ofsociallife,theycannotbeidentifiedindependentlyofthegametowhichtheybelong.Thisisplainestforactionsthatareliterallymovesinagame,forinstancetheplayingofatileinmah-jong,wherethesubjectivemeaningcannotbeana-lyticallypriortotheintersubjectivemeaning.Wittgensteinbidsusregardsocialactioningeneralinthissortofway.Sociallifeismadeupofinstitutionsorpractices,bestunderstoodbyanalogywithgames.Crucially,gameshavetwosortsofrules.Onesortisregulativeandgovernsthebestchoiceofmovesateachturn.Forinstance,successinmah-jongdependsonknowingsomeusefulrulesofthumbandinknowingwhethertheotherplayersknowthemtoo.Theothersortisconstitutiveandgovernstheverypossibilityofmakingmovesatall.Theconstitutiverulesofmah-jongdefinethepro-cedures,piecesandpurposeofthegameandmustbefollowedbytheplayers,onpainoffailingtotakeanypartinthegame.Importanttothisfertileanalogyisthefactthatthereisnothing(exceptthepresenceofconstitutiverules)whichallgameshaveincommon:Don’tsay:Theremustbesomethingcommon,ortheywouldnotbecalled‘games’–butlookandseewhetherthereissomethingcommontoall.Forifyoulookatthemyouwillnotseesomethingthatiscommontoall,butsimilarities,relationships,andawholeseriesofthematthat.(Wittgenstein1953:§66)Thereisonlyacomplexnetworkofsimilarities,overlappingandcriss-crossing,or‘familyresemblances’.Thereisnofurthercommoncoreto‘boardgames,...Olympicgamesandsoon’.Soweshouldnotexpectsocialpracticestodisplayanyuniversalcommoncoreeither.Theyvarywithinandbetweensocietiesandthereisnothingmoreglobaltoit.‘Whathastobeaccepted,thegiven,is–onecouldsay–formsoflife’(ibid.:226).Theprimaryunderstandingofaction,then,isnotasanexpressionofsubjectivemeaningbutasthefollowingofarulebyanagentwho,inafamousWittgensteinianphrase,‘knowshowtogoon’.Thismaysounddeterministic.Itcanseemasmallstepfrompresentingsocialactorsasrulefollowerstopresentingthemascreaturesoftheruleswhichtheyfollow.Butthatcanbeavoided,iftherulesofthegamesofsociallifearenottakentobesocomplete,detailedandpervasivethattheycovereverypossiblesituationinadvance.OnoneinterpretationofWittgenstein,rulesarealwaysopen-endedandareconstructedtocovernewcasesinthecourseofplay.(Foraplausibleexampleofthisprocess,thinkhowlegalrulesevolvethroughtheinterplayofjudicialinterpretationsincourtsoflaw.)Inthatcasewehaveafertileholisticapproachwhichdoesnotturntheactorsintoculturaldopes.Instead,theactorsretainapositiveinter-pretativerole.Evenso,itdoesnotleaveasmuchroomforindividualinitiativeasanindividualistmightwish.Itsuggestsnotonlythatthe‘games’ofsociallifeareanalyticallypriorto387\nMARTINHOLLISthe‘moves’madewithinthemandthatallsocialactionsare‘moves’,butalsothatsocialactorsareessentiallyplayers.Ifso,itconflictswiththecommonbeliefthatweareindi-vidualswhochoosewhethertoplayandwhattoplay.Furthermore,thisbeliefextendstotheconvictionthatwhatwecannotdoseparatelywecanoftendocollectively.Forexample,manytheoriesofpoliticalfreedomanddemocracypresumethattheinstitu-tions,rulesandpracticesofasocietycanbechangedbycollectivedecision.Ananaly-sisthatworkedwholly‘topdown’intheUnderstandingcolumnoffigure12.1wouldthusbepoliticallycontentious,evenifitleftsomeroomforplayerstomanoeuvrewithinthegame.Ontheotherhand,ananalysiswhichworkedwholly‘bottomup’intheUnderstandingcolumnwouldseemwilfullyblindtothepointthatreasonsforsocialactionarecommonlyboundupwiththerolesthatpeopleplay.Soletusnexttryembeddingsocialactorsintheirsocialroleswithouttherebydenyingthattheyareindividuals.Ifsocialnormsareindeedasmuchconstructedinthecourseofinter-actionaslaiddowninadvanceofit,thereshouldberoomforanindividualiststoryaboutthisprocessofconstruction.Totellit,weneedtostressthatrolesenableaswellasconstrainaction.RolesA‘role’isstandardlydefinedasasetofnormativeexpectationsattachedtoasocialposi-tion.Thusaprofessorshipisasocialpositionandprofessorsareexpectedtodoresearch,teachstudentsandsoon.Theseexpectationsarenormativeinthatthosedealingwithprofessorsareentitledtopresumethattherolewillbeplayedasrequiredandtocom-plainifitisnot.Buttherequirementsoftherolearenotexpressibleinacompletelistofexplicitrulestocovereverysituation.Nosuchcompletelistcouldbespecifiedandroleplayershavetobelefttointerprettheroleinsomegeneralspirit.Professorsarerequiredtobescrupulousaboutgivingcreditforideasborrowedfromothersandtobeconscien-tiousandfair-mindedwhenjudgingstudents’work;butwhatcountsasscrupulous,con-scientiousandfair-mindedcannotbespecifiedindetail.Thatdependstoomuchontheexactcircumstancesofeachnewcaseinwaysthatcallforjudgementbytheactors.Rolesthustakeonshapeastheyareplayedand,singlyandcollectively,wechoosemuchofthatshape.Itisaconstrainedchoicebutstillachoiceand,althoughthechoicessumtodefinetherolemoreclosely,theshiftingcharacterofthesocialscenecontrivestounde-fineitatthesametime.Therequirementsofrolesbothconstrainandenabletheroleplayers.Idonotwishtoassertdogmaticallythatthisistherightwaytoconceiveofroles.But,ifitisaplausibleway,thenitoffersscopeforasocialindividualismthatisphilosophi-callychallengingonseveralgrounds.Althoughthereisspaceonlyforwhatmayseemaverybaldlist,thinkofwhatfollowsasapointertoacompromiseintheUnderstand-ingcolumnoffigure12.1betweenthepureindividualismofthebottomboxandthethoroughgoingholismofthetopone.Firstly,weneedtothinkfurtheraboutthenotionofareasonforaction.Therearetwoideasofpracticalreasonincommonuseinthesocialsciences.Oneisinstrumen-388\nPHILOSOPHYOFSOCIALSCIENCEtalorwhatWeber,intheessaycitedearlier,termsZweckrationalität–therationalchoiceofthemosteffectivemeanstotheagent’sends.This,elaboratedindecisiontheory,isbasicforeconomicsand‘economic’theoriesofhumanbehaviour,aswillemergeinthenextsection.Itconfinespracticalreasontodeliberationaboutthelikelyconsequencesofalternativeactions.TheotherisexpressiveorwhatWebertermsWertrationalität,wheretheagentattachessuchvaluetothegoalortotheperformanceoftheactionthatconsequencesdonotmatter.Thissoundslikethekindofrationality–orlackofit–thatturnssocialactorsintowhatIhavecalledculturaldopes.Butthatneednotbeitsimplication,ifwetreatitasrule-following,asintheparagraphsabove,andsosetforamorehumanhomosociologicusthetaskofdeliberatingaboutwhatisrequiredofhimorher.Yetneitheroftheseaccountsseemstoyieldananalysisofpracticalreasonthatdoesjusticetotheconceptofapersonassomeonecapableofmoralconduct.Norcantheybereadilycombined,asweshallseepresently.Secondly,mentionofMORALCONDUCT(chapter6)invitesthethoughtthatexpressiveactionoftenexpressesthecharacteroftheagent,ratherthanthedemandsofasocialposition.Istheselfnotmorethanthesumoftherolesthatitplays?ThatleadstotangledquestionsofpersonalidentityanddeepintothePHILOSOPHYOFMIND(chapter5).Canpersonsbekeptdistinctfrombothpersonalitiesandpersonae?If‘personality’referstopsychologicalcharacterand‘persona’(asindramatispersona,acharacterinaplay)referstopeopleasmanifestedinsocialrelations,isthereanythingfurtherandnon-physicalthatisspecialtoaperson?Neither‘personality’nor‘persona’capturestheelementsofself-awarenessandself-directionwhichmanythinkersholdtobecrucialforwhatitistobeaperson.Apromisingstarting-pointistoinvokethedistinc-tionmadebyWilliamJAMES(chapter36)betweenthe‘I’andthe‘me’:‘theselfasknower’and‘theselfasknown’(James1958,1962).Butthereisnoavoidingdeepermattersthanwecanpursuehereand,itmustbesaid,theymaketheother-mindsproblemalltheharder.Thirdly,questionsaboutthenatureofhumanfreedomremainurgent.Thosecon-cerningmoral,politicalandsocialfreedomaretouchedonbelow,althoughchieflydealtwithelsewhere(seechapters6and8).Meanwhile,wenotedearlierthattheproblemoffreewillthatfacesanyoneapplyingadeterministicphilosophyofnaturalsciencetothesocialworldisnotindisputablysettledbyadoptingcompatibilism.OneattractionofamethodofUnderstandingthatallowsforself-directiononthepartofsocialactorsisthatitseemslessdeterministic.Theideathatwehelptocreatethesocialordertowhichwebelongisintriguingbutdoesnotabolishtheproblemeither.Theinitialholisticsug-gestionthatsocialactorsarethetransmittersratherthanthecreatorsofthemeaningoftheiractionsstillneedsverycarefulamendmentifitisnottosuckusintoaresidualsocialdeterminism.Norshouldweforgetthatwehavenotaddressedthedifficultiesofrelatinganactive,autonomousselftothephysicalworldthatitinhabits.Anyoneseekingacompromisebetweenindividualismandholismtoconcludethissectionthereforehasagreatdealofunfinishedbusiness.SoitseemswisertoendthesectionwithsomesummaryremarksaboutUnderstanding.Itstartsbymakingsenseofaction‘fromwithin’bydisclosingboththemeaningoftheactionandwhattheactormeantbyit.Thatcandemandtwosortsofaccount,oneidentifyingtherelevantprac-ticesandformsoflife,theotherseekingtheactors’aimsandreasons.Forbothitishelpfultosuggestthatweproceedontheassumptionthattheactionisrational.But389\nMARTINHOLLISthisformofwordsconcealsadeepdivision.Incallingactionrationalweoftenmeanthattheagenthasparticularreasonsforit,whicharegoodreasonsoratleastseemtohimorhertobeso.Thatistostresstheactor’smeaning.Butwemayalsomeanthattheactionisrationalinthesenseofbeingappropriateandinaccordwiththerelevantrules,especiallyifweareseekingtounderstandactionsthatstemfromthedemandsofroles.Thatistostresstherule-governedcharacterofsocialactions.Itisnotobvioushowthesetwosortsofaccountfinallyrelatetooneanother.Yetitiscertainlyfertiletoask.TheremayalsobedeeperquestionsaboutthemeaningofHISTORY(chapter14)thatshowthemselveswhenhistoriansandsocialtheoriststrytoidentifyepochsandmove-ments.Suchattemptscontinuetoinspiremanyphilosophers,especiallythoseinflu-encedbyHegel.Ontheotherhand,theyinfuriateothers,andPopper’sfulminationsinThePovertyofHistoricism(1957)arenottobemissed.Thereisnothingbrieftosayaboutwhetherthehistoricalworldisopentoadialecticalinterpretationthatcallsforavaultinglyambitiousconceptofunderstanding.Butitisworthnotingthat,evenifPopperwererighttodenyit,nothingwouldfollowaboutthesuccessofnaturalism.TheWeberianandWittgensteinianreflectionsdeployedinthissectionstandwithoutpresupposinganyhiddendialecticsandwehaveyettodiscusstherelationbetweenunderstandingandexplanation.3ExplainingandUnderstandingWithoutadequacyonthelevelofmeaning,ourgeneralizationsremainmerestatementsofstatisticalprobability,eithernotintelligibleatalloronlyimperfectlyintelligible...Ontheotherhand,fromthepointofviewofsociologicalknowledge,eventhemostcertainadequacyonthelevelofmeaningsignifiesanacceptablecausalpropositiononlytotheextentthatthereisaprobability...thattheactioninquestionreallytakesthecourseheldtobemeaningfullyadequate.ThisisWeber’s(1922)viewofhowsections1and2relate.(IttoocomesfromtheopeningessayinEconomyandSociety,withhisitalics.)Itsoundsentirelyreasonable.Ontheonehand,wehavestatisticsaboutpeoplelikeM.Rouget;butitmakesscarcelymoresensetolearnthat65percentofyoungworkersinlargefactoriesvoteLeftthanthat65percentofmenwithbrowneyesridebicycles.Sowealsoneedameaningfulstory,whichpresentsM.Rouget’sworldfromtheinsidebymeansofarationalrecon-struction.Ontheotherhand,thereisnoinherentguaranteethatarationalrecon-structionyieldsatruestory.Toensurethat,weneedahighempiricalprobabilitythatthestoryiscausallyadequatetoo.Onreflection,however,thisresolutionisnotsoneat.WeberhasadoptedtheHumeanthesisthatcausalpropositionsaretobeconstruedasstatementsofstatisticalprob-ability.But,aswesawinsection1,therearegroundsforrejectingthisanalysisinfavourofoneinvolvingrealconnectionsorcausalpowersorboth.Inthatcase,statisticalgen-eralizationsdonotrelateadequacyonthelevelofmeaningdirectlytowhatever‘really’movesM.Rougetatacausallevel.ThisinturnrevealsanambiguityinWeber’sreso-lution.Ishesayingthatthesocialworlddoesindeedworkasameaningfullyadequate,390\nPHILOSOPHYOFSOCIALSCIENCErationallyreconstructedaccountclaims,providedonlythatpeoplelikeM.Rougetarestatisticallycommonenough?Orishesayingthatsuchaccountsprovideanessentialcluetoacausallevel,wheretheelementsthatdothe‘real’explanatoryworkmaybeofaverydifferentsort?Itisthusopentoustoconcludeeitherthattheactorsareusuallyself-propelledandknowwhattheyaredoing,atleastwhentheiraccountstally,orthattheirsocialbeingcommonlydeterminestheirconsciousness;eitherthattheirsocialbeingisdistinctfromthefabricofmeaningsandcausesitorthatcausalexplanationsmustfinallybecouchedinculturalterms.Inotherwords,thetalecouldstillendwithanyofthepossibilitiesinfigure12.1.Inwhatfollows,Ishallfocusontwoseparatecentresofcurrentdebate.Onecon-cernstherightsortofwaytoreadthe‘economic’theoriesofbehaviourmentionedearlier,thuscontinuingourdiscussionofrationalityasanindividualistkeytosocialanalysis.Theotheristhevexedquestionofrationalityandrelativism,wherethedisputebetweenExplanationandUnderstandingisbroadlyholisticonbothsides.Theconclu-sionsofthischapter,giventhepresentstateofphilosophyandsocialtheory,willconsistlargelyofanagenda.3.1Rationalityandthetheoryofgames‘Thefirstprincipleofeconomicsisthateveryagentisactuatedsolelybyself-interest.’Thismuch-quotedremarkcomesfromF.Y.Edgeworth’sengaginglytitledbookMathematicalPsychics(1886)andcaptureswhatiswidelytakentobethekeytothemostforcefulindividualistanalysisofsocialaction,institutionsandpractices.Asnotedearlier,oneofWeber’sroutestoexplanatoryunderstandingisby‘ideal-typicalrationalreconstruction’ofaction.Thismeansinterpretingactionsasthechoicesmadebyrationalagentsinordertoachievetheirendsortosatisfytheirpreferences.‘Ideal-typical’referstowhatisnowadaysoftencalleda‘rationalchoice’or‘economic’approach,whichworksbywayofanidealizedmodelofaworldwhereallagentsareperfectlyrational.Wecanbestbeginwithabriefoutlineofsuchamodel.Startbydefiningaperfectlyrationalagentasanagentwithcompleteandconsistentpreferences,perfectinformationandperfectpowersofcalculation.Hisorherprefer-encesrangeovertheconsequencesofeachactionopentohimorher.Informationincludesknowingwhattheseconsequenceswouldbeandhowlikelytheyaretooccur.Calculationthenidentifiestherationalchoiceinthecircumstances.Forexample,Jackisafarmerchoosingwhethertobreedgoatsorpigs.Hewouldratherhavepigsbuttheycostmoreandaremorepronetodisease.Sohemustcalculatewhetherthegreater‘utility’ofpigsismorethanoffsetbythegreatercostsandrisks.Intheoryatleast,hecancomparetheexpectedutilityofgoatsandpigsbyrankingtheirutility,netofcosts,afterdiscountingitfortheriskthatitwouldnotmaterialize.Arationalagentneverchoosesacourseofactionwhoseexpectedutilityislessthanthatofanotheravailabletohim.TheconceptofUTILITY(chapter35)hereowesmosttoJeremyBENTHAM(1748–1832)(chapter35)andtheutilitarians.Benthamheldthatallactionsaresubjecttoa‘principleofutility’,definedas‘thatpropertyinanyobject,wherebyittendstoproducebenefit,advantage,pleasure,goodorhappiness...tothepartywhoseinter-estisconsidered’.Theprincipleguideseachrationalagenttodowhat,asmodern391\nMARTINHOLLISeconomistsputit,maximizeshisorherexpectedutility.Whetheranactionisrationalisfinallyamatterofthementalsatisfactionwhichitsconsequencesbringtotheagent(Bentham1970:ch.1).ItmaysoundasifBenthamassumedhumanbeingstobeselfish,self-seekingcrea-tures.ThesameimpressionisconveyedbyEdgeworth’sremarkthat‘thefirstprincipleofeconomicsisthateveryagentisactuatedsolelybyself-interest’.But,evenifpeopleareinfactalltoooftenself-seeking,theprincipleofutilitydoesnotsayso.Itimpliesnothingaboutthespecificcharacterofpreferencesandtellsusonlythatpeoplearemotivatedtoseekwhatwillbestsatisfywhatevertheyhappentodesire.SimilarlyHume,whileinsistingthatactionisalwaysmotivatedbypassion,madeitclearthatnotallpassionsareselfish:Ambition,avarice,self-love,vanity,friendship,generosity,publicspirit:thesepassions,mixedinvariousdegrees,anddistributedthroughsociety,havebeen,fromthebeginningoftheworld,andstillare,thesourceofalltheactionsandenterprises,whichhaveeverbeenobservedamongmankind...Mankindaresomuchthesame,inalltimesandplaces,thathistoryinformsusofnothingneworstrangeinthisparticular.(Hume1978b:VIII,I,65)Thussaintscanbeasrationalassinners.Whetheractionisrationaldependssolelyonwhetheritisinstrumentalinfurtheringtheagent’spassionsorpreferences.Thisbecomesstillmoreexplicitinrecentutilitytheory,whereaconsciouseffortismadetoremovealltraceofpsychologyandtotreatrationalityasconsistencybetweenpreferences,actionsand(expected)outcomes(seeespeciallySavage1954).WhereasHumethoughtintermsofmotivatingpassions,modernexponentsofthepuretheoryofrationalchoicedispensewithpassionsandrequireonlythatarationalagenthasaconsistentpreference-orderoverpossibleoutcomes.Whatmakesrationalchoicetheory(ordecisiontheory)complexisnotthecomplexityofhumanpsychologybuttheprob-lemsofdealingwithriskanduncertainty.WemightwishtodoubtHume’sclaimthatmankindare‘somuchthesame,inalltimesandplacesthathistoryinformsusofnothingneworstrangeinthisparticular’.Itisatleasthardertodoubttheblanderclaimthatallrationalactionshaveasingle,universallogic.Wenextapplythismodelofindividuallyrationalactiontomorethanoneagent.WhetherJackdoeswelltochoosepigsovergoatsmaydependonwhetherJillchoosestobreedgoatsorpigsontheneighbouringfarm.Ifthepay-offtohisactiondependsonherstrategyandviceversa,thentheyareplayinga‘game’,inasensedefinedbythetheoryofgamesandfarremovedfromWittgenstein’s.Perhapstheyintendtoselltheiranimalsinasmallvillagemarketwhichwillabsorbsomegoatsandsomepigsbutnotaglutofeither.Thentheymaybeplayingaco-ordinationgame,asshowninfigure12.3.Hereeachhasachoicebetweenthetwostrategies,GoatsandPigs,andthegamehasfour(2¥2)possibleoutcomes.Eachoutcome(orcellofthematrix)yieldsapay-offtoeachplayer,shownasutilitynumbers(Jack’sfirstandthenJill’s).Whatexactlythesenumbersrepresentisaphilosophicallyinterestingquestion;but,forpresentpur-poses,letustakethemtoregisterhoweachplayerrankstheoutcomes,withahighernumbersignifyingagreaterpreference.Thusfigure12.3showsbothplayersequallywellsatisfiedbyeitherwayofco-ordinating,whereoneoptsforpigsandtheotherforgoats.392\nPHILOSOPHYOFSOCIALSCIENCEJillPigsGoatsPigs1,12,2JackGoats2,21,1Figure12.3Theco-ordinationgame:strategiesfacedbyJackandJillforbreedingpigsorgoats.Thisparticulargamehastwopure‘solutions’:ifJackchoosespigs,Jill’sbeststrat-egyisgoats,inwhichcaseJack’sbestchoiceisindeedpigs;ifJackchoosesgoats,Jill’sbeststrategyispigs,inwhichcaseJack’sbestchoiceisindeedgoats.Sothereisnosinglebeststrategyforeither,ornonebetterthantossingacoin.Ifthegameistobeplayedseveraltimes,however,JackandJillwouldbothwelcomethehelpofaconventiontosteerthemtowardsoneofthepuresolutions.Ifitwerethepractice,forinstance,thatpigsaremen’sworkorgoatsarewomen’swork,andifeachknewthattheotherknewthis,theneachwouldpresumablyoptaccordinglyforwhatgametheoriststermthe‘salient’strategy.Atanyrate,assoonasarelevantconventionisestablished,bothplayersarewellservedandneitherhasanyreasontodepartfromit.Theideaissimplebutcanbeusedtopromptapowerfulindividualistanalysisofinsti-tutions.Wearebeinginvitedtoregardthebasicrulesornormsofanysocietyascon-ventionsthatareofmutualbenefittorationalindividualswithseparateinterests.Itseemsveryplausibletosuggestthatrulesornormswhichsolveco-ordinationproblemsexistandpersistbecausetheydoso.Moregenerallystill,italsoseemsplausibletosuggestthatsocietyanditsbasicinstitutionscanbeanalysedasanassociationofindi-viduals,regulatedtomutualadvantage.IsitnotsimplerstilltosupposethatJackandJill,insteadoflookingroundforacon-ventionorwaitingforonetoemerge,willmakeadirectagreement?Yesitis;butthatpresupposestheexistenceoflanguageandapracticeofmakingagreements.Inthemostambitiousversionsofthetheoryofthesocialcontractnothingispresupposedandevenlanguageistreatedasasetofconventionsandpracticeswhichletindividualsgainthemutualbenefitsofcommunication.Gametheorythereforestartswithgameswhereinitialchoicesaretobemadewithoutpriorcommunicationandwhereanyemergentconventionsexistwithoutbeingenforcedbyanykindofsanctions.Evenifco-ordinationisabasicgame,itdoesnotholdthekeytoallnormsorinsti-tutions.Nolesscrucialisthegameknownastheprisoner’sdilemma,showninfigure12.4.Here,letussuppose,JackandJillhaveamutualinterestingivingbackanyanimalsthatstrayacrosstheboundarybetweentheirfarms.(JackhasspecializedinpigsandJillingoats;sostrayscanbeeatenbutnotusedforthemoreprofitablebusi-nessofbreeding.)But,sincethereisnowayofprovingwhathappenstoamissinganimal,eachdoesbetterstilltokeeptheother’sstraysanddoesworstofallbybeingtheonlyonetoreturnthem.Ifeachisarationalagentwiththepreferenceordergiveninfigure12.4,theywillfailtoachievetheirmutualinterest.Jackwillreasonthathe393\nMARTINHOLLISJillKeepReturnKeep2,24,1JackReturn1,43,3Figure12.4Theprisoner’sdilemma:strategiesfacedbyJackandJillfordealingwithanimalsthatstrayontotheirfarms.doesbettertokeepJill’sstrays,ifshedoesnotreturnhis,andbettertokeepherstrays,evenifshedoesreturnhis;sohedoesbettertokeephers,whatevershedoes.Jillreasonssimilarly.Theoutcomeisthereforeworseforboththanitwouldbeifbothactedco-operatively.Itmightseemthattheycanimproveonthisinferioroutcomebymakinganagree-mentandpromisingtokeepit.Butthatwillnothelpthem,iftheyhavethesameinstru-mentalviewofagreementsandpromisesastheysofarhaveofothermatters.Supposethateachnowprefersthatbothkeeptheagreementthanthatneitherdoes,butmostpreferstobreakit(secretly,nodoubt)whiletheotherkeepsit,andleastpreferstokeepitwhiletheotherbreaksit.Thenthesituationisstillthatinfigure12.4,andeachwillbreaktheagreement,thusagainproducingamutuallyinferioroutcome.Thelastparagraphiscontentious.IttakesalinemostfamouslyputforwardbyHobbes,whowascitedaboveforhisviewofhumanbeingsasmechanicalindividualswith‘aperpetualandrestlessdesireforpowerafterpower,thatceasethonlyindeath’.HobbesgoesoninLeviathan,chapter13,toarguethatmencanliveatpeaceonlyiftheycancreate‘acommonpowertokeepallinawe’,whichcanthenenforceanyagree-mentsmade;andinchapter17hesays:‘Covenants,withouttheSword,arebutWords,andofnostrengthtosecureamanatall’.Thisargumentliesatthecoreofhisversionofthesocialcontract,whichisthusstrikinglydifferentfromversionsgroundedintherationalityofco-operatingforsimplemutualadvantage.Herelietheoriginsofanenduringdisputebetween‘conflict’and‘consensus’modelsofsociety.Thedifferencemightseemtoturnonopposingviewsofhumannatureandwhetherwehaveenoughgenialpassions,likeHume’s‘friendship,generosity,publicspirit’,tomakeco-operationnatural.But,althoughthismayberelevantifweaskhowwide-spreadexamplesoftheprisoner’sdilemmagameare,itdoesnotaffectthelogicofthegame.Rationalchoicetheoryimpliesthatplayerswiththepreferencesshowninfigure12.4willchoosenottoco-operate,regardlessofhowtheyacquiredthepreferences.Reasonsforactionalwayslookforwardtotheconsequencesofactionandneverbacktohowthesituationcameabout.Ifthefactthatapromisehasbeenmadeistoserveasareasonforkeepingit,thetheoryneedsrecasting.Meanwhileenoughhasbeensaidtoshowwhy‘Rationalchoices’belonginthebottom-leftquadrantoffigure12.1.Initsstandardform,rationalchoicetheoryslotstheagentintoacausalsequencebetweenwhatevergivesrisetohisorherpreferencesandtheactionswhichhisorherinternalcomputerdirects.Thatthreatensitsstatusasanindividualisttheory.Butifitisnotdefenceenoughtopointoutthatactionsresult394\nPHILOSOPHYOFSOCIALSCIENCEfromindividualpreferences,thereisstillthecaseforholdingthatinstitutionsaretheoutcomeofpreviousinteractionsamongindividuals.Ontheotherhand,itisatleastinterestingtotryrecastingthetheorysothatagentscandistancethemselvesfromtheirpreferencesandcanactonotherthanconsequentialistreasons.Thatwouldmakeitacandidateforthebottom-rightquadrant,aformofindividualismwhichisadequateatthelevelofmeaningalongthelinesindicatedwhenwewerediscussingrole-playingintheprevioussection.3.2RationalityandrelativismAlthoughnothinghasyetbeensaidtoproveit,itishardtoresisttheideathatsocialactioncannotbeidentifiedanddescribedwithoutseeingtheactors’situationfromwithin.Soletusnowassumeforthesakeofargumentthatthefirststepintheanaly-sisofanysocialphenomenonbelongstoUnderstanding.Evenso,thelaststepmaystillbelongtoExplanation.Wemayneedunderstandingtoidentifywhatisgoingonandyetalsoneedanexplanationofhowandwhyitoccurs.TheimplicationsofthisthoughtleadustoacurrentdisputeaboutrationalityandRELATIVISM(pp.295–7).Itusedtobecommontomeetbehaviouristswhoregardedsocialactorsasveryinfe-riorguidestotheirownbehaviour.Scientists,theymaintained,haveanaccesstothebrain,thebody,thehumanenvironmentandthemechanicsofinteraction,whichmereagentscannotaspireto.Thereforequantitative,andhenceobjective,evidencealwaystrumpsqualitative,andhencemerelysubjective,reportsfromtheinside.Moreover,itisnoobjectionthatbehaviouralexplanationsoftenuseconceptsalientothoseinwhichtheactorsthink.Afterall,biologydoesnotneedtotakeaworm’seyevieworworryaboutanysubjectivemeaningattachingtowhatwormsdo;thereisnothinginherentlyspecialabouthumanlifeamongtheotheroperationsofnature.Thisattitudehasrecededandmostsocialscientistsorphilosophersarenowwillingtostartbydescribingsocialphenomenaintermswhichtheactorswouldunderstandandassentto.Thisiseffectivelytograntthatsocialeventsaresignificantlymorehis-toricallyandculturallylocatedthanthosestudiedbyphysics.Butnaturalistsdonotregardthisasadamagingconcession,providedthataninterpretativeaccountofwhathappensisfinallysubordinatetoanaturalisticaccountofhowandwhyithappens.Thenactioncanbeidentifiedfromwithinandexplainedfromwithout:acompromisefavouringnaturalism.Amajorcomplication,however,istherecentgrowthinthesociologyofscience,encouragedbyThomasKuhn’sworkonthestructureofscientificrevolutions(Kuhn1970).TheEnlightenmentviewofsciencehasalwaysbeenthatitisasearchaftertruthundertheguidanceofreason.Kuhnputupastrongcaseforbelievingthatreasoncannotdoeverythingaskedofit,becauseallsystemsofthoughtrelyon‘paradigms’whichgovernwhatcountsasreasonable.APARADIGM(pp.10–11)hastwokeyfea-tures.Oneiscognitive–asetofverygeneralpresuppositionsaboutthenatureofthings,forexamplethePtolemaicworldviewortheCopernicanonewhichreplacedit.Theotherissocial–asetofinstitutionalpracticesthatentrenchthepresuppositionsinthelifeofasociety,forexamplethosethatmaintaintheprofessionalconduct,statusandinfluenceofneo-classicaleconomists.Sincebothfeaturesregulateacanonofrationalbeliefandaction,paradigmsarenotsubjecttorationalcriticismfromthesortof395\nMARTINHOLLISexternal,absolutestandpointtowhichtheEnlightenmentaspired.Therecanbenosuchstandpoint.Ifreasondoesnotdetermineourmostgeneralbeliefs,whatdoes?Oneinfluentialanswerhasbeentoextendthesociologyofknowledgetoincludethestudyofwhateverpassesfor‘knowledge’–scienceandepistemologynolessthan,say,religionandtheol-ogy.Insomeversions,thethemeisthatcognitivesystemsaregeneratedbytheworkingofsocialsystems,withthesecausalrelationsbeingasubjectforanobjectivesocialscience(BarnesandBloor1982).Butthisseemstoleaveanallegedlyobjectivesocialscienceasanexceptiontotheclaimthatallcognitivesystemshavesocialdeterminants.Otherversionsarecarefultoincludethesociologyofknowledgeinthesociologyofknowledge.This,however,promptsastillmoreradicalthought.Ifscience,naturalandsocial,isultimatelyapracticeamongpractices,thenexpla-nationcannotfinallybecontrastedwithunderstandingbyopposinganexternalstand-pointtoaninternalone.Explanationisavitalpartofthegameofscience,oreventhepurposeofthatgame,butitcannotbeadevicefordoingtheimpossible,namelystep-pingoutofthegamealtogetherinordertorevealthenatureofreality.So,althoughitremainsimportanttounderstandtheclaimstoobjectivitymadeinthepracticeofscience,thiscanonlybeanexerciseinseeingfromwithin.Hencealinewhichsetouttogiveunderstandingthefirstturnandexplanationthesecondpresumablyendsbygivingafinalturntounderstanding.Meanwhile,thereissomethingoddaboutthissharpcontrastbetweencognitivesystemsandsocialsystems.Theformerdonotfloatinthinairandthelatterdonotconsistofunthinkingbodies.Althoughitisusefultodrawdistinctionsamongthemanyactivitiesthatcompriseasocialworld,itisnotasifwewereunpickingafabricwhosethreadswereeithersocialorcognitivebutnoneofthemboth.Weber’sdistinctionbetweenthelevelofmeaningandthecausallevelcannotbetakenasonebetweensepa-rateordersofbeing.Inrelatingexplanationtounderstanding,wearetryingtomakesenseofasingleworld.Theprecedingparagraphsraiseanurgentquestionaboutrelativism,whichweshouldnextmakeexplicit.TheEnlightenmentstoryofsciencereliesonnotionsofrationalitywhichconnectreasondirectlywithtruth.Thebroadestideaisthatthereisarationalorderinnatureandthatreasoncandiscernthetruthaboutit.Ifthereislessofarationalorderinthehumanworld,thatisbecausehumansareimperfectlyrational,atleastwhenitcomestoconstructingsocialarrangements.Butthatsimplygivesreasonthefurthertaskofdiscerninghowtoimprovethem.Itwillbeclearbynowthatthisstoryhascomeunderheavyfireoflate.Ineffect,criticshavebeentryingtorelativizereasonbysuggestingthatrationalityisalwaysamatterofconformingtotherulesofaparticularpractice.Ifpractices,inturn,arerelativetoparticulargroupsandsocieties,thentruthtoobecomesunattainableor,ifyouprefer,amatterofconformitytolocalcustom.ItisironictofindEnlightenmentthinkers,whosetouttodestroydogmawiththeaidofreason,beingattackedformakingreasonadogma.But,ifitisindeedone,thenitisliberatingtohaveitexposedassuch.This,accordingly,hasbecomeathemeinrecentpoliticalphilosophy,especiallywherethetargetisthestateastheembodimentofadomineeringrational–legalorder.SomeLIBERTARIANS,FEMINISTSandCOMMUNITARIANS(chapters8and20)inparticularhavesoughttobaseideasof396\nPHILOSOPHYOFSOCIALSCIENCEautonomyandrespectforpersonsonarejectionofclaimstojustifycentralizedpowerbyappealtorationalauthority.Theircaseofteninvolvesadefenceofrelativism,appar-entlystrengthenedbyinsistingthat,torecallthequotationfromDilthey,‘lifedoesnotmeananythingotherthanitself.Thereisnothinginitwhichpointstoameaningbeyondit.’Butotherlibertarians,feministsandcommunitariansfindthisadangerousline.Ifthereisnothingbeyondformsoflifebywhichtojudgeformsoflife,therecanbenoexternalstandpointforcriticizingthestate.Indeed,iflifehasnomeaningoutsideexist-ingformsoflife,thenwepresumablyneedtolove,honourandobeytheoneswehave.Inthatcaserelativism,alliedwithhermeneutics,isnotliberatingatall.Sincethisisnotachapteronpoliticalphilosophy,Ishallnottrytoumpirethedispute.Butitsexist-enceshouldmakeuswaryofrelativisminepistemologytoo,ifthatweretoclearthewaytomaintainingthataclosedsocietyisasentitledtoitswayoflifeasanopenone.YetweshouldnotassumethattoendorsetheclaimsofUnderstandingistherebytointernalizereasonandtruthtopractices.Thatwouldbetorunintodeeptroublewiththenotorious‘hermeneuticcircle’.Hereliesaproblemtooramifiedforthischapter,butthereisjustroomforapointertoit.Supposethateachcultureor‘formoflife’wereself-containedinasensewhichincludesthepossessionofrulesdefiningwhatisreal.Thussuppose,forinstance,thatwhetherwitchesexistis,atheart,aquestionaboutwhethertherearecultureswithrulesforidentifyinganddealingwithwitches,sothattheref-erenceaswellasthemeaningof‘witch’isdeterminedinternally.Suppose,roughlyspeaking,thatwitchesexistwhereverbeliefinwitchesiscognitivelyandsociallyentrenched.Thengeneralizetheidea,sothatnatureatlargeceasestobeindependentofrulesdeterminingwhatitisrationaltobelieveabouttheworld.Thatwouldmakeitplausibletosuggestthateachculturedetermineswhatisrealandwhatisrationalforitself.Thissoundsanattractivesuggestion,evenifittakesaboldimaginationtofancythatwhatholdsforwitchesalsoholdsforrabbits.Butitposestheother-mindsprobleminanacuteform.Inordertopenetrateanotherculture,wewouldhavetoestablishwhatitheldtoberationalbeforewecouldknowwhatitheldtobereal;andwewouldhavetoestablishwhatitheldtoberealbeforewecouldknowwhatitheldtoberational.Thatisthe‘hermeneuticcircle’.Toescapeit,Ithink,onemusteitheravoidinternalizingreasonandtruthtoprac-ticesorinsistthatsomepracticesareuniversal.Thatisnodoubteasiersaidthandone.Inmentioningit,Imeanonlytosuggestthatdefendersofobjectivereasoninhermeneuticsneednotgiveupwithoutastruggle.4ConclusionsWhatcanweconcludeaboutthefourkeystoanalysingsocialactionofferedinthequadrantsoffigure12.5?Itwouldmakeforatidyendingifoneofthemwereamasterkey.Butthatisunlikely.Startwiththeleastpromisingclaimant,‘systems’(topleft),andtrythinkingofactionswhollyasresponsestochangesintheactors’environment.Foreachactorsingly,theenvironmentincludesotheractors;forallcollectively,itisanexternalstructure–‘society’orperhaps‘nature’.Forthistobeatallplausibleasanexplanatorymasterkey,itneedsthebackingofanambitiousfunctionalor397\nMARTINHOLLISExplanationUnderstandingHolismSystemsCulturesIndividualismRationalSubjectivechoicesmeaningsFigure12.5Conclusions:thefourkeystoanalysingsocialactionrevisited.EVOLUTIONARYTHEORY(pp.320–30).Suchtheoriesdocertainlyhavechampionsinsomerealmsofscience.Forinstance,beesadaptandevolveastheirenvironmentchanges.Theirnormalbehaviour,singlyandcollectively,canbepresentedasbeingwhatisrequiredifthehiveistofunctioninitsenvironment.But,evenso,thiscanhardlybethewholestory.Beesaretoofascinatinglyparticular,andhumansmuchmoreso.Curiosityswiftlybeckonsusintothebottom-leftquadrant.Individualbeesarewonderfullyintricatecreatures,whoseinteractionsareamaz-inglycomplex.Forinstance,ascoutwillreturntothehiveandperformanelaboratedancewhosemovements,itseems,communicatethedistance,direction,quantityandqualityofthenectarwhichithaslocated.But,eventhoughwemightthereforedeembeestohavealanguage,theirbehaviourisalwaysaneffectofenvironmentaleventsonacomplexmechanism.Isthisalsohowwewanttoconstruetheexplanatoryindi-vidualismofthebottom-leftquadrant,asappliedtohumanbeings?Thatisunclear.Ontheonehand,rationalchoicetheory,extendedtoincludegametheory,certainlysetsouttoanalysecollectivephenomenaintotheirindividualcomponents.Socialconven-tionsandhenceenduringsocialinstitutionscanbepresentedasthesumsanddepositsofindividualchoices.Ontheotherhand,the‘individuals’involvedareautomatictrans-latorsofgivenpreferencesintoexpectedutilities,nomoreparticularthandifferentdatasetsfedintothesamecomputerprogram.Whateverestablishesthepreferencescouldplausiblybesaidtodictatethechoices.Herecompatibilismisphilosophicallycrucial.Iftheconditionsforfreedomandindi-vidualityaresatisfied,providedthatactionsarerationallyconsistentwithparticularpreferences,thenrationalchoicetheoryisanindividualisttheoryofthesortwanted.Whilepreferencesremaingiven,andsoexogenouslydetermined,itcannotclaimtoprovideasolemasterkey;butitcanclaimtosupplyanessentiallinkbetweensocialstructureandsocialaction.Yettheindividualityofhumanbeingsstillseemsdifferentinkindfromtheparticularityofindividualbees.Ifso,wearedrawnintothebottom-rightquadrantinsearchofwhatisspecialaboutpersons.AswecrossintotheUnderstandingcolumn,thenatureofactionchanges.Itceasestobeindependentlydescribablebehaviourcausedbyamentalstateandbecomeswhattheagentmeansbyit.‘Subjectivemeanings’arethestuffofaction,whichisnodoubtwhyWeber’smethodofunderstandingbeginswithempathy.Actorsinterpret398\nPHILOSOPHYOFSOCIALSCIENCEtheirsituationcreativelyandexpressthemselvesinbringingtheirintentionstobearonit.Intersubjectiverulesandmeaningsareproposed,negotiatedandleftopentorenegotiation.Yet,whateveritsmeritsasanaccountofsocialexperienceatfirsthand,thisistooanarchicforevenahermeneuticallyinclinedsocialscience,andmakesforahopelesslyone-sidedanalysisofrolesandrole-play.Wecannotresistaskingwhatthe‘Cultures’ofthetop-rightquadranthavetooffer.Thekeyideahereisthatrulesenableaswellasconstrain.Constitutiverulesengen-derthe‘games’inwhichsocialactionsare‘moves’;regulativerulesgiverisetoreliableexpectationsaboutwhichmoveswillbeplayed.Theideaconveyedbycallingtherules‘cultural’isthatacultureisawebofsharedmeaningsexternaltoeachofitsmembers,yetinternaltoallcollectively.Thatissuggestiveforthedifferencebetweennaturalfactsandsocialfacts,forinstancebetweenAfricanweatherandAfricanpolitics.ItrecallsWittgenstein’sremarkthat‘Whathastobeaccepted,thegiven,is–onecouldsay–formsoflife’,apparentlyoffering‘formsoflife’asamasterkey.Butitishardtobelievethattheyaresoself-propelled,althoughIregretthattherehasnotbeenroomtodis-cussrecentattemptstotracetheprogressofthehumanmind(torecallthetitleofCondorcet’sbook)inunifiedtheoriesaboutthedynamicsofculturalhistory.Mean-while,mycommentaboutAfricanweatherandAfricanpoliticsisdouble-edged.Theexternalworldofnaturestillconstrainsandenablesthelifeofsocieties,providingcommonproblemsandscarceresources.Wearebackwiththetop-leftquadrant.Presumablythen,thereisnouniquemasterkey.Humanbeingsmakesomeoftheirownhistoryincircumstancespartlyoftheirownchoosing,withgreaterscopeiftheyknow‘howtogoon’.Thisknowledgemixesexplanationandunderstandinginwaysthateludesocialscientistsandphilosophers,oratleastdefysummary.Sothechaptercanbestendwithanagendaofquestionsaboutthebordersbetweenthequadrants.Arethereasonsthatpromptrole-playerscausesofactionafterthemannerofnaturalcauses?Canplayersofthegamesofsociallifebeseenconsistentlybothasfollowersofrulesandasmakersofchoices?Inmakingsenseofsocialfacts,doweneedconceptsbothpriortothoseofindividualpsychologyandalientothoseofnaturalscience?Whatformofdeterminism,ifany,shallasocialsciencebringtobearonaworldwhoseinhab-itantsclaimtobeitscreators?Howpeculiar,inshort,aretheintellectualandmoralfacultiesofhumanbeings?NorshouldweforgettheoriginalEnlightenmenthopethatthesocialscienceswouldbringmoralandpoliticalprogress.Afinalquestion,then,iswhethertruth,happinessandvirtueareindeedboundtogetherbyanindissolublechain.FurtherReadingForageneralbackgroundinthephilosophyofscienceseetheFurtherReadingforchapter9.ButIwilljustrecommendChalmers(1982)asamasterlystart.Amongmanygeneralintroductionstothephilosophyofsocialscience,Ryan(1970)hasproveddurable.Itisrichininstructiveexamples,althoughmorepositivistinapproachthanonemightwishfornow.KeatandUrry(1975)isharder,butcastsitsnetwider.Threeinteresting,recentintroductions,whichgivescopetoUnderstanding,arethosebyDoyleandHarris(1986),Rosenberg(1988)andBraybrooke(1991).AnyonewantingtoseethethemeofthischapterdeployedinfullmighttryHollis(1994).399\nMARTINHOLLISPositivismremainsimportantandso,therefore,doesMill(1961)BookVI,latelyeditedbyA.J.Ayer.MiltonFriedman’s(1953)paperspeaksvolumes.Butthetidehasturned.Winch(1958)hasbecomeclassic.ItmakesafertileWittgensteiniancaseforthedistinctnessofthesocialsciences.AcontraryviewistakenbyBhaskar(1979).LikehisScientificRealismandHumanEmancipation(1986),itisdeepbutdifficultreading.Thatgoesgenerallyformuchanti-Positivistwork;butGeuss(1981)isaclearinitialguide.Severaloftheissuesraisedinthesereferencesleadtoheatedargumentaboutrelativism,asintheessaysroundedupinWilson(1970)andHollisandLukes(1982).Thereisclearerwaterfordisputesbetweenindividualismandholism,atanyrateontheindividualistside.ThatowesmuchtoJonElster’sseriesoflucidandimaginativebooksinsistingthat‘theelementaryunitofsociallifeistheindividualhumanaction’.IsuggeststartingwithExplainingTechnicalChange(1983)chs1–3,LogicandSociety(1978)chs4and5,andUlyssesandtheSirens(1984),ratherthanthemorerecent(Elster1989).Thereisanaccessibleliteratureonrationalchoiceandgametheory;forinstance,HargreavesHeapetal.(1992).Holismishardergoing,butRuben(1985)putsanexactanalyticalcase.ReferencesBarnes,B.andBloor,D.1982:Relativism,RationalismandtheSociologyofKnowledge.InM.HollisandS.Lukes(eds)RationalityandRelativism.Oxford:Blackwell.Bentham,J.1970[1789]:AnIntroductiontothePrinciplesofMoralsandLegislation.London:AthlonePress.Bhaskar,R.1979:ThePossibilityofNaturalism.Hassocks:HarvesterPress.——1986:ScientificRealismandHumanEmancipation.Hassocks:HarvesterPress.Braybrooke,D.1991:PhilosophyofSocialScience.EnglewoodCliffs,NJ:Prentice-Hall.Chalmers,A.F.1982:WhatisThisThingCalledScience?2ndedn.MiltonKeynes:OpenUniversityPress.Condorcet,M.de1955[1795]:SketchforaHistoricalPictureoftheProgressoftheHumanMind(translatedbyJ.Barraclough).London:WeidenfeldandNicolson.Dilthey,W.1926:GesammelteWerke(editedbyB.Groethuysen).Stuttgart:Teubner.Doyle,L.andHarris,R.1986:Empiricism,ExplanationandRationality.London:RoutledgeandKeganPaul.Durkheim,E.1964[1895]:TheRulesofSociologicalMethod(translatedbyS.SolovayandJ.Mueller).NewYork:FreePress.Edgeworth,F.Y.1886:MathematicalPsychics.London:RoutledgeandKeganPaul.Elster,J.1978:LogicandSociety.NewYork:Wiley.——1983:ExplainingTechnicalChange.Cambridge:CambridgeUniversityPress.——1984:UlyssesandtheSirens.Cambridge:CambridgeUniversityPress.——1989:NutsandBolts.Cambridge:CambridgeUniversityPress.Fontenelle,B.de1929[1688]:ThePluralityofWorlds(translatedbyJ.Glanvill).London:NonesuchPress.Friedman,M.1953:MethodologyofPositiveEconomics.InEssaysinPositiveEconomics,Chicago:UniversityofChicagoPress.Geuss,R.1981:TheIdeaofaCriticalTheory:HabermasandtheFrankfurtSchool.Cambridge:CambridgeUniversityPress.HargreavesHeap,S.,Hollis,M.,Lyons,B.,Sugden,R.andWeale,A.1992:Choice:ACriticalGuide.Oxford:Blackwell.Hobbes,T.1990[1651]:Leviathan(editedbyR.Tuck).Cambridge:CambridgeUniversityPress.Hollis,M.1994:ThePhilosophyoftheSocialSciences.Cambridge:CambridgeUniversityPress.400\nPHILOSOPHYOFSOCIALSCIENCEHollis,M.andLukes,S.(eds)1982:RationalityandRelativism.Oxford:Blackwell.Hume,D.1978a[1739]:ATreatiseofHumanNature.Oxford:OxfordUniversityPress.——1978b[1748]:AnEnquiryConcerningHumanUnderstanding.Oxford:OxfordUniversityPress.James,W.1958:ThePrinciplesofPsychology.NewYork:DoverBooks.——1962:Psychology:BrieferCourse.London:CollierBooks.Keat,R.andUrry,J.1975:SocialTheoryasScience.London:RoutledgeandKeganPaul.Kuhn.T.S.1970:TheStructureofScientificRevolutions,2ndedn.Chicago:ChicagoUniversityPress.Lipsey,R.1972:IntroductiontoPositiveEconomics.NewYork:HarperandRow.Marx,K.1971[1859]:PrefacetoAContributiontotheCritiqueofPoliticalEconomy.London:LawrenceandWishart.Mill,J.S.1961[1843]:ASystemofLogic.London:Longworth.BookVI,editedbyA.J.Ayer,London:Duckworth(1988).Popper,K.1957:ThePovertyofHistoricism.London:Routledge.——1969:ConjecturesandRefutations.London:Routledge.Przeworski,A.andTeune,H.1970:TheLogicofComparativeSocialEnquiry.NewYork:J.WileyandSons.Quine,W.V.O.1953:TwoDogmasofEmpiricism.InFromaLogicalPointofView,Harvard,MA:HarvardUniversityPress.Rosenberg,A.1988:PhilosophyofSocialScience.Oxford:ClarendonPress.Ruben,D.-H.1985:TheMetaphysicsoftheSocialWorld.London:RoutledgeandKeganPaul.Ryan,A.1970:ThePhilosophyoftheSocialSciences.London:Macmillan.Savage,L.J.1954:TheFoundationsofStatistics.NewYork:J.Wiley.Weber,M.1922:EconomyandSociety,ed.G.RothandC.Wittich.Berkeley:UniversityofCaliforniaPress.Wilson,B.(ed.)1970:Rationality.Oxford:Blackwell.Winch,P.1958:TheIdeaofaSocialScience.London:RoutledgeandKeganPaul.Wittgenstein,L.1953:PhilosophicalInvestigations.Oxford:Blackwell.DiscussionQuestions1Doesthestudyofsocietyrequirebothcausalexplanationandinterpretativeunderstanding?2Dopersonsandtherolestheyfillbothhaveimportanceforsocialexplanation?3Canasocialsciencecopewithhumanautonomy?4Howaresocialfactsrelatedtofactsaboutindividuals?5Doesanyaccountofnaturalscienceprovideamodelonwhichsocialsciencemightreasonablybebased?6Howimportantisthenotionofa‘formoflife’totheunderstandingofsociety?7Ifweareconcernedabouthumanfreedom,shouldwebemoreworriedbyasocialsciencewhichbindsindividualsbysociallawsorasocialsciencewhichreducesindividualstomerecarriersofculture?8Isitamistaketoseekauniversalsocialsciencetocoverallsocieties?9Doweneedmorethanrationalchoicetheoryandgametheorytoanalysesocialinstitutions?10Ifsocialexplanationisintermsofbaseandsuperstructure,howcanwedeter-minewhatisthebaseandwhatisthesuperstructure?401\nMARTINHOLLIS11Whataretheimplicationsoftheclaimthatsocialunderstandingisessentiallyhistorical?12Howdodifferentconceptionsofrationalityaffectouraccountofsocialscience?13Isindividualismorholismabetterapproachtounderstandingsocialinstitutionsandpractices?14Domoralnotionslikejusticehavealegitimateroleinsocialexplanation?15Canidealtypeshelptoexplainsocialaction?16Howisitpossibletounderstandothersocieties?17Inwhatsenseshouldsocialscienceseekobjectivity?18Aretheregoodreasonstoconfineaccountsofbehaviourandinstitutionsinasocietytoconceptsavailabletomembersofthatsociety?19Cananaccountofsocietybasedonhermeneuticsexplainsocialchange?20Doesananalysisofsocialactionintermsofrule-followingimplythatsocialactorsfollowrulesautomatically?21Whatconceptofthepersonismostsuitableforsocialscience?402\n13PhilosophyofLawN.E.SIMMONDSThephilosophyoflawliesattheintersectionofanumberofproblems.Theseconcerntherelationshipbetweenlawandmorality;thelogicalstatus(asdescriptiveorpre-scriptive)ofpropositionsoflaw;thepossibilityofseparatingadjudicationfrompolitics;andthedistinctionbetweenlawandorganizedforce.Legalpositivistshaveinsistedonamorallyneutralconceptoflaw,whereinalegalsystemconsistsofpositedrulesascer-tainablebyreferencetofactualcriteria.Thenaturallawtradition,bycontrast,drawsattentiontolaw’srootednessinnotionsofjusticeandthecommongood.Moderndebateshaveoftencentredontheprocessofadjudication,inparticularuponthequestionoftheextenttowhichjudgesmayproperlyrelyupontheirpersonalmoraljudgement,andhowfartheyarelimitedtotheapplicationoftechnicallegalrules.Formostofus,mostofthetime,lawseemstobenothingmorethanasetofgovernmentalarrangementswherebytransactionsareregulatedandundesirableformsofconductdiscouraged.Whyshouldthisgiverisetophilosophicalproblems?Orrather,whyshoulditgiverisetophilosophicalproblemsotherthantheproblemsofAPPLIEDETHICS(chapters16–19),whichmayariseinrelationtoanysetofhumanarrangements?Aconvincinganswertothisquestionwillrequireustooutlineanumberofpuzzlingfeaturesofthephenomenonoflaw.Thephilosophyoflawliesattheintersectionofthesepuzzles,andseekstoconstructacoherentunderstandingofthenatureoflawinsuchawayastoresolvethepuzzles.Someproblemsareofakindthatmightoccurtoanythinkingperson,whileothersspringfromthetechnicalunderstandingandex-perienceoflawyers.Someproblemsaregeneratedbythespecialrolethatlawenjoyswithinmodernpoliticallife,whileothersariseonlyagainstthebackgroundofspecificphilosophicalcommitmentsimportedfromareasofenquirysuchasEPISTEMOLOGY(chapter1)orETHICS(chapter6).Philosophersoflawhopetooffersolutionsthatarebothenlighteningandintegrated,inthesensethatasingletheoryoflawshouldprovideabasisfortheresolutionofallthecentralproblems.Unsurprisingly,adiversityoftheorieshasbeenoffered.Thedebateissignificantlycomplicated,however,bythefactthatdifferentlegalphilosophersandjuristsconceiveofthedebateinquitedifferentways:theydisagreeaboutthenatureandstatusoflegal\nN.E.SIMMONDStheoryjustasmuchastheydisagreeaboutthenatureandstatusoflaw.Sometheoriespurporttobepurely‘analytical’or‘clarificatory’,whileothersarguethatanadequatesolutiontothephilosophicalproblemsoflawrequiresatheorythathasaprescriptiveaspectandthatisrootedinPOLITICALPHILOSOPHY(chapter8);sometheoriesdefendclaimsaboutthe‘necessary’constituentsofthe‘concept’oflaw,whileothersseekareflectiveunderstandingofspecificsocialpractices.Thesedisagreementswillinevitablycomplicateoursurvey.1ProblemsofJurisprudence1.1LawandforceLawmayattimesstrikeusasbeingnothingmorethanasetofgovernmentalarrange-mentsworthstudyingforpracticalpurposes,butdevoidofphilosophicalinterest.Wedonotalways,however,thinkoflawinthismundaneguise.Wetreattheideaoflawasamoralaspirationthatissystematicallyconnectedtoarangeofothervalues.Wecontrasttheruleoflawwithpoliticalregimesthatarebasedsolelyonforce;wethinkoflawashavingdeepconnectionswithindividualLIBERTY(pp.258–64)andwithJUSTICE(chapter8);wespeakoflawasimposingobligationsuponus,andasconferringrights.Thelawseemstobebothmundaneandaspirational:amundanesetofarrange-ments,andatthesametimeanembodimentofaspirationsthathaveacentralplacewithinourpoliticalmorality.Thisambivalenceoflawmanifestsitselfateveryturn.Wemightthink,forexample,oftherelationshipbetweenlawandforce.Wecontrastformsofgovernmentbasedon‘theruleoflaw’withgovernmentbasedmerelyonorganizedforce.Indeedtheobjectoflawmightbethoughttobetheeliminationofforcefromhumanrelations,anditsreplacementbypeacefulcompliancewithsharedrules.The‘ruleoflaw’seemstorep-resentasituationwherecitizensandofficialsregulatetheirconductbypubliclyrecog-nizedrulesthatareacceptedasbindingonallconcerned.Yetarenotlegalsystemssimplyverywellorganizedandsystematicstructuresofviolence?Isthebindingforceoflawanymorethanitsbackingbycoercivesanctions?Unlikeagangofrobbers,thestateauthoritiescandeployoverwhelmingforceagainstthosewhodefythem,andtheywillgenerallyusethatforceinregularandpredictableways.Butdoesthatmakethemtheembodimentofavaluetoberadicallycontrastedwithrulebyforce?Theattemptedanswerstosuchquestionsareoftendividedintotwodifferentgroups.Ontheonehandaretheories(generallycalled‘naturallawtheories’)claimingthatlawisinitsessentialnatureconnectedwiththevaluesofjustice,rightandthecommongood.Agoverningregimecantruthfullyclaimtogovernbylawonlyifitsactionsanddecreesaredevisedtoservethecommongoodortodojustice.Ifagovernmentsimplyexploitsasubjectpopulationwithouteventhepretenceofseekingtodojustice,itgovernsbyforce,notbylaw,howevermanyofthetrappingsoflegitimacyitmayassume.Ontheotherhand,so-called‘legalpositivist’theoriesinsistonamorallyneutralconceptoflaw.Suchtheoriesareinclinedtoanalyselawintermsofpurelyformalfea-tures,suchasthepresenceoforganizedsanctions,orthepublicationofblack-letterrules(namely,ruleswhichareestablishedinadefiniteverbalformulation,suchasthe404\nPHILOSOPHYOFLAWprovisionsofastatuteoranexplicitlyformulatedrulelaiddownbyajudge)whicharequiteindependentofthegoodorbadcontentofthelaw,andthegoodorbadintentionsofthelaw-makers.Likenaturallawtheories,positivisttheoriesvarywidelyintheirreasonsforadvocatingsuchaneutralconceptoflaw,aswellasintheirconceptionsofthenatureofthephilosophicalenquiry.Thecontrastbetween‘naturallaw’and‘legalpositivism’isthereforeaslipperyonewhichneedstobehandledwithcare,butitcanserveasapreliminaryorganizingdevice.1.2PositivismandreductionismLawtellsuswhatwe‘ought’todo;itconfersrights;itimposesobligations.Donotthesetermsimportadimensionofmoralauthority?Howcanwespeakoflawasimposingobligationsandconferringrightsunlessweregardthelawasmorallybinding?Iflawissimplyanapparatusofrulesestablishedbythosewhohappentohaveextensivecoer-civeforceattheirdisposal,howcoulditeveralterourrightsandobligations?Themerepossibilityofaskingsuchquestionsdemonstratesthat,eveniftheposi-tivistcouldconvinceusthatalllawsarepositedrulesemanatingfromauthoritativesources,thiswouldnotsufficetoestablishtheseparationoflawandmorals.KANT(chapter32),forexample,arguedinTheMetaphysicsofMoralsthat,eveninasystemofwhollypositedlaws,onewouldstillrequireabasicnaturallawthatestablishedthemoralauthorityofthelawgiver.Intheabsenceofsomesuchacknowledgedmoralclaimuponourconduct,howcouldweviewthelawasasourceofobligation?Thepositivistreplytothislineofargumentconsistsintheclaimthatalegalobli-gationisnotaspeciesofmoralobligation,norisalegalrightaspeciesofmoralright.Lawmayormaynotimposemoralobligationsonus,thepositivistclaims:moraloblig-atorinessisnotalogicallynecessaryfeatureoflaw.Lawdoesimposelegalobligationsonus,butthesearenotmoralobligationsandtheyinvolvenocommitmenttothelaw’smoralauthorityormoralbindingness.Havingclaimedthatlegalrightsandobligationsarenotaspeciesofmoralrightorobligation,thepositivistisfacedwiththetaskofexplainingwhattheyare.Earlierver-sionsofpositivismtendedtorespondtothischallengebyadoptingaREDUCTIONIST(pp.312–13)stance.Theyclaimedthattheapparentlyprescriptivelanguageoflawcouldbeanalysedreductivelyasadisguisedformofdescriptiveorpredictivelanguage.WhenIstatethatyouhavealegalobligationtopayme£100,Iseemtobemakingapre-scriptiveclaimabouthowyououghttobehave.Butinfact(itwasargued)allthatmyassertionactuallymeansisthatifyoudonotpayme£100youarelikelytosufferasanction.Inthisway,propositionsoflawconcerninglegalrightsandobligationscouldbetreatedasfactualstatementsabouttheissuingofcertaincommands,predictionsofthelikelyreactionsofcourtsorthelikelihoodofsufferingsanctions.Thistypeofreductionistanalysisappealedtopositivistsbecauseitseemedtotrans-latelegalconceptsintotermsthatwerefreeofallmoralconnotations.Theanalysisencounteredsomeratherobviousdifficulties,however.Forexample,ImaynotbelikelytosufferasanctionifIfailtopayyou£100(Imaybeoutofthejurisdictionandhavenointentionofreturning,oritmaybeclearthatyouwouldnotdreamofsuingforsuchasmallsum),butitmayneverthelessbethecasethatIhavealegalobligationtopay.Similarly,courtsgenerallyinvoketheexistenceofcertainlegalobligations405\nN.E.SIMMONDSincumbentonthelitigantsasareasonfordecidingthecaseonewayratherthananother,andasajustificationforimposingasanction.Thiswouldmakenosenseif,cor-rectlyanalysed,statementsaboutlegalobligationswerepredictionsofthereactionsofcourtsortheimpositionofsanctions.Moremodernpositivistshavethereforesoughttoanalysepropositionsoflawinawaythatpreservestheirprescriptivecharacterwithoutequatingthemwithmoraljudgements.Thishasprovedtobeaverydelicatebalancingact.ThelegalpositivistHansKelsen(1881–1973)soughttoavoidreductionismbytreatinglegaldiscourseasbasedonafundamentalpresuppositiontotheeffectthattheregimewhichisineffectivecontrolwithinacertainterritoryoughttobeobeyed.Hecalledthispresupposition‘thebasicnorm’.Toinvokesuchabasicnormlooksatfirstlikeanabandonmentofpositivism,sincetheneedforsomesuchbasicassumptionseemstobepreciselythepointthatisbeingmadebyKantinhisabovemen-tionedremarkfromTheMetaphysicsofMorals.YetKelsenarguedthatthisbasicpresuppositionoflegalsciencecanbedeployedsimplyforthepurposeofgroundingacognitionofpositivelaw,andwithoutcollapsingquestionsoflegaldoctrinebackintomoralquestionsaboutjustice.Hedescribedthebasicnormastheminimumelementofnaturallawwithoutwhicha(non-reductionist)cognitionoflawisimpossible.Otherapproacheshavebeenexploredbythoselegaltheoristswhohavefoundthenotionofabasicnormativepresuppositionunattractive.H.L.A.Hart(1907–93),forexample,arguesinhisbookTheConceptofLaw(1961)thatreductionisttheoriesoflawproceedfromtheexternalviewpointofsomeoneobservingtheworkingsofalegalsystem,ratherthantheviewpointofaparticipantinthatsystem.Toavoidtheerrorsofreductionism,legaltheoryneedstoreproducetheviewpointofaparticipantwithinthesystem.Suchaparticipantregardsthelegalrulesasstandardsthatoughttobecompliedwith.Statementsaboutlegalrightsandobligationsdonotpredictsanc-tions,butdrawconclusionsabouttheapplicabilityoflegalrulestoparticularcircum-stances.The‘internal’pointofviewfromwhichsuchlanguageisdeployedneednot,Hartargues,beanattitudeofmoralapprovalofthelaw;norneeditbeexpressiveofajudgementastolaw’smoralbindingness.Theinternalpointofviewrestsonajudge-mentthatthelawoughttobeobeyed,butthis‘ought’neednotbeamoral‘ought’.Agreatmanydiverseconsiderationsmightunderpintheconclusionthatone‘ought’toobeythelaw.Legaltheorycanbeneutral,itisargued,onwhatthoseconsiderationsmightbe.Theproblemwiththiswayofavoidingreductionisminlegaltheoryisthatverylittlemiddlegroundseemstobeleftbetweentheviewpointoftheexternalobserverandthejudgementthatlawismorallybinding.Someonewhoobeysthelawpurelyfromfearofsanctionswouldseemtotakeanexternalview,sincehisonlyconcernwillbewithiden-tifyingthoseformsofconductthatareinfactlikelytoincursanctions.Ifthepartici-pant’sinternalpointofviewisnotoneofsimplesanction-avoidance,whatcanitbeotherthanabeliefinlaw’smoralbindingness?Thelegaltheoristcouldsimplyexcludesuchquestionsasirrelevant,therebytreatingthe‘ought’asfoundationalforlegaldis-course,whilebracketingoutallquestionsaboutwhatreasonsunderpinthe‘ought’judgement.Thisapproach,however,seemstodependuponafundamentalnormativepresuppositioninverymuchthesamewaythatKelsen’stheorydoes.Itistherefore406\nPHILOSOPHYOFLAWdoubtfulwhetherHartsucceededinofferingagenuinealternativetoKelsen’sapproach.1.3LawandpoliticsThedebatebetweenlegalpositivismandnaturallawiscommonlythoughtofasadebatewithanancientandcontinuoushistory.PhilosophersfromSOCRATES(chapter22)tothepresentdayhavereflectedontherelationshipbetweenmightandright,andhaveaskedwhetherlawissimplyabodyofdecreesissuedbythoseenjoyingamo-nopolyofcoercivepower,orwhetheritisanexpressionofmoralrightwhichservestoconcretizetheabstractrequirementsofjustice.Yetinmorerecenttimesthesefunda-mentalissueshaveacquiredfromthecircumstancesofmodernityapeculiarsalienceandadistinctivecharacter.Newquestionshavearisen,andoldoneshaveacquirednewresonances,asaresultofthecentralrolethatconceptionsoflawplayinmodernpoliticallife.Theancientunderstandingofpoliticsplacedquestionsaboutthenatureofthegoodlifeatthecentreofpoliticalenquiry.Politicalinstitutionswerethoughttoexistinordertoinculcateandmakepossibleawayoflifethatwouldbeworthyandexcellent.Suchafullyvaluablelifewastobeattainedwithinpubliclifebyparticipationinthepoliticalaffairsofone’scollectivity.Theprivaterealm,whichconsistsoflabour,thefamilyandthehouseholdeconomy,wasconsideredtobearealmofnecessitythatwasofvalueonlyinsofarasitmadeapublicpoliticallifepossible.Inthemodernworld,however,theseprioritieshavebeenlargelyreversed.PoliticalphilosopherssinceHOBBES(chapter28)havetendedtosearchforaformofpoliticsthatwouldleaveeachindividualfree,sofaraspossible,topursuehisorherownconceptionofagoodandexcellentlife.Onthisapproach,theprivaterealmisseenasacriticallyimportantareaofautonomywithinwhichtheindividualcanchooseandpursuewhollypersonalgoalsandprojects.Theexistenceofsuchprivateareasofautonomynecessitatesthecleardemarcationandenforcementofindividualentitlements.Lawandthestateexistprimarilytosustaintheclearareasofentitlementthatcomprisetheprivaterealm.Totheextentthatitemphasizestheroleofindividualentitlements,togetherwiththeneedforPLURALISM(pp.266–9)andautonomy,themodernconceptionofpoliticstakesonaprofoundlyjurisprudentialcharacter.Foritislawthatdefinesandenforcesentitlements,andthepossibilityofcoherentanddeterminateprinciplesoflawisanecessaryfoundationforthemodernview.Yet,ifmodernLIBERAL(chapter8)politicalthoughttendstopromoteindividualautonomyabovecollectiveparticipation,thelattervalueneverthelesshasavitalparttoplay,andapartthataddsafurtherdimensiontotheproblemsoflegalphilosophy.Forthemodernworldnolongerregardsthestruc-turesofcivilsociety,theeconomyandthefamilyasimmutablestructuresforminganaturalhorizonwithintheparametersofwhichpoliticallifemustbeconducted.Wearedeeplyawareofthemutabilityofsuchstructuresandoftheiropennesstotrans-formationthroughpolitics.Invaluingdemocracy,wevalueourabilitytodecidecol-lectivelyuponthegeneralprofilethatoursocialstructuresshouldexhibit.Inthissense,wevalueourabilitytopursuecollectiveprojectsofsocialreform.Liberalpolitics,therefore,valuesbothcollectiveprojectpursuitand(intheformoftheliberal’sconcernforlibertyorautonomy)individualprojectpursuit.Butthesetwo407\nN.E.SIMMONDSvaluescompete.Asocietywhichalloweditsoverallcharactertobedeterminedsolelybytheuncoordinatedactionsandtransactionsofindividualspursuingtheirownpro-jectswouldforfeitanypowerofcollectivecontroloveritsgeneralprofileandstructure.Ontheotherhand,asocietywhichsoughttoregulatecollectivelyeveryfacetofitsexis-tencewouldinevitablyeraseanyscopeforautonomousindividualprojectpursuit.Itisthepossibilityofastablelegalsystemthatrenderstheliberalschemeofvaluescoherent.Iftheruleoflawispossible,collectiveprojectpursuitcanfinditssphereofexpressionwithinpoliticaldecisionsconcerningthegeneralruleswhicharetobeenactedinlaw,andthegeneralentitlementswhicharetobeconferred.Actionwithinthescopeofsuchrulesandentitlementswillthenbeprotectedwhetherornotitservestoadvancethegoalsofthecollectivityasawhole.Thisconceptionoftheruleoflawassumes,however,thatonecanidentifyandapplytherulesandprinciplesofexistinglawwithoutreopeningthepoliticalquestionsofcol-lectiveprojectpursuitthatmayhaveledtotheenactmentofthoserulesandprinciples.Itassumesthatajudgecandeterminewhetherornotacitizenhasactedwithinthescopeofhisorherlegalrightswithoutaskingwhetherthatcitizen’sactionsservetoadvancedesirablecollectivegoals.Iftheseassumptionsarefalse,thenone’sentitle-mentswillextendonlysofarasone’sactionsservecollectivegoals.Inotherwords,thevalueofindividualautonomywillbewhollysubordinatedtothedemandsofcollectiveprojectpursuit.Amajorareaofinvestigationinthephilosophyoflawconcernstheattempttosustainaviabledistinctionbetweenlawandpolitics.Thisdebatehaslargelybeencon-ductedinthecontextofgeneraltheoriesoflawandadjudication,butithasalsoinvolveddetailedstudiesofparticularareasoflegaldoctrinewithaviewtoassessingtheextenttowhichtheycanresolvedisputeswithoutcollapsingallquestionsbackintobasicissuesofcollectivegoalsanddistributivepolicies.1.4LegalpositivismandthecriteriaofvalidityLegalpositivismoffersamorallyneutralconceptionoflaw.Itclaimsthatvalidlawsmaybeidentifiedbyreferencetocriteriaofapurelyfactualnature.H.L.A.Hart,forexample,arguesthateverylegalsystemcontainsabasic‘ruleofrecognition’,acceptedbyofficials,whichidentifiescertainsources(suchasdecisionsofcourtsorenactmentsofthelegislature)assourcesoflaw.Arulethatemanatesfromanappropriatesourceisavalidruleregardlessofitsjusticeorinjustice.Correspondingly,arulethatdoesnotemanatefromanappropriatesourceisnotavalidrule,nomatterhowjustorreason-ableitmaybe.Hart’slegalpositivismhasmuchtorecommendit.Itofferspleasinglyclear-cutanswerstoanumberofquestions,andseemstoprovideasimplewayinwhichadis-tinctionbetweenlegaldoctrineandpoliticaldisputemaybemaintained.If,however,thetheoryisconstruedasanattempttoelucidatethecriteriaofvalidityactuallyemployedbylawyers,thetheoryseemslessconvincing.Expertlawyersfrequentlydis-agreeaboutwhattheexistinglegalrulesare.Forexample,manylawyersbelievethattheEnglishlawofcontractcontainsadoctrineof‘commonmistake’,holdingthatacontractisvoidifitwasenteredintoonthebasisofafundamentalmistakesharedby408\nPHILOSOPHYOFLAWbothparties.Otherlawyersdenytheexistenceofanysuchdoctrine.Butiflawyersemploycriteriaofafactualkindinordertoestablishthecontentofthelaw,howisthistypeofdisagreementpossible?RatherthanprovidingasimplerefutationofHart’slegalpositivism,however,suchexamplesraisethequestionofwhetherhistheoryaimssimplytorecordthecriteriaofvalidityimplicitinlawyers’practices,orwhetheritmightnotbeseenasproposingarevisedconceptionoflaw,constructedinthelightofmoregeneraltheoreticalconsid-erations.HartwasProfessorofJurisprudenceatOxfordUniversityduringthe1950sand1960s,andheworkedinclosecollaborationwithmembersoftheOxfordordinarylanguageschoolofphilosophy.Thisleadsmanytoassumethathisobjectinlegaltheorywastheunearthingandarticulatingofdeepcriteriathatlawyersandcitizensuncon-sciouslyemployinmakingjudgementsaboutlegalvalidity.YetweshallseelaterthattherearegoodgroundsforrejectingthisinterpretationofHart’swork.1.5LawandadjudicationWhencourtscometoapplylegalrules,theyarefrequentlyconfrontedbydifficultques-tionsofinterpretation.Factualsituationsdonotcomeneatlypackagedandlabelledinlegalvocabulary,andtherecanbeamplescopefordoubtaboutwhetheraparticularlegalruleshouldorshouldnotbeappliedintheindividualcase.Cansuchquestionsberesolvedbyreferencetoformalsemanticrules?Generally,themeaningofanutterancewillbeafunctionofformallinguisticrulesinconjunctionwiththeimplicationsofthepragmaticcontextoftheutterance.Butlegalrulesfrequentlylackanyobviousprag-maticcontext:theymayhavebeenenactedlongagowithintentionsthatarenowobscure,buttheyneverthelesspurporttoregulatethepresent.Itmaybetemptingtointerpretlegalrulesinthelightofdesirablepolicyobjectives,butthisthreatenstocol-lapsefundamentallythedistinctionbetweenlawandpolitics.Differenttheoristshavedrawndifferentconclusions.Hartsoughttosketchoutamiddlecoursebetweentwoextremepositionswhichhecalled‘formalism’and‘rule-scepticism’.By‘formalism’hemeanttheviewthatallconceivablecasescanbedecidedbyapplyingpre-existingrulesoflaw,withoutanyneedtoaskquestionsaboutnon-legalconsiderationssuchasjusticeorsocialpolicy.‘Rule-scepticism’,ontheotherhand,istheviewthatjudgesareneverreallyboundbyrules,sothatthedecisivefactorsineachcasearealwaysextra-legalconsiderationsofsocialpolicy.Hartarguedthatformalismignoresthefactthatlanguageexhibitsan‘opentexture’:whilewordshavea‘coreofsettledmeaning’theyalsopossessa‘penumbraofuncertainty’whereitwillbeunclearifthewordisproperlyapplicableornot.Insofaraslegalrulesarelinguisticentitiestheytoowillexhibitthis‘opentexture’.Casesofpenumbraluncertaintywillinevitablyariseinanyconceivablesystemoflaws,andsuchcasescanonlyberesolvedbythecourtexercisingitsdiscretion.Ontheotherhand,rule-sceptics(Hartargued)overesti-matetheextentofpenumbraluncertaintyandignorethefactthatthegreatmajorityofcasesfallwithinthe‘coreofsettledmeaning’oftherule.Hart’sresponsetotherule-scepticsisnotentirelysatisfactory.Inthefirstplacehisargumentplacesgreatrelianceontheassumptionthatlegalruleshaveadefiniteverbalformulation.Thisis,ofcourse,trueofstatutoryrules,butwithinthecommonlaw409\nN.E.SIMMONDSdevelopedbythecourtsitisnottrue.Thebindingpartofaprecedentiscalledtheratiodecidendi,aphrasewhichreferstothatpartofthecourt’sreasoning(aspublishedinitsjudgement)whichwasnecessarytojustifytheconclusionreached.Allotherobserva-tionsmadeinthejudgementsaresaidtobeobiterdicta.Thedistinctionbetweenratioandobiterisextremelypliable,sincealatercourtmaytakeadifferentviewfromtheoriginalcourtaboutwhatexactlywasnecessarytothedecision.Evenwhereacourtdoesofferadefiniteverbalformulationoftherulethatitisapplying,sucharuleneverbecomeslawsimplybyvirtueofhavingbeensopronounced,foritisalwayspossiblethatalatercourtmaytreattheruleasobiter.Inthisway,courtsconstantlymodifytherulesinthecourseofapplyingthem.Supposethatwe,asjudges,havetodecideacasethatissomewhatsimilartoanearlierprecedent.Thecourtintheearliercaselaiddownaprinciple,thewordingofwhichappearstoapplyexactlytothecaseinfrontofus.Itwillneverthelessbepos-sibleforustoavoidapplyingtheprincipleasformulatedintheearliercaseifwecanidentifyfactsinthepresentcasethatwereabsentfromtheearliercase,andthosefactsappeartojustifydecidingthetwocasesdifferently.Wewillsaythattheprincipleasformulatedintheearliercasewas‘toowide’,wasnotnecessaryforthedecisionintheearliercase(sinceanarrowerprinciple,containinganexceptionforcaseswithfactslikethoseofthepresentcase,wouldhaveservedequallywell),andwasthereforeobiter.Wewillthendecideourcasebyreferencetoarevisedversionoftheearlierprinciple,nowincludinganexceptionforcasesexhibitingfactsofthekindfoundinthepresentcase.Theapparentpliabilityofthedistinctionbetweenratioandobiter,combinedwiththeresultingtendencyofcourtstomodifyrulesinthecourseofapplyingthem,raisesaproblemofrule-scepticismthatisnotaddressedbyHart.Forifcourtscanaltertheruleswhenevertheythinkitbest(ongroundsofjustice,forexample)thentheyarenotreallyboundbyrules:tobeboundbyaruleinvolveshavingtoapplyitevenwhenyouthinkthatitsapplicationisnotforthebest.Isitpossibletoarguethatcourtsmayalterthelawwhilestillbeingboundbythelaw?Hart’ssuccessoratOxfordUniversity,RonaldDworkin(b.1931),hasdevelopedatheorywhichseemstoexplainhowthatmightbepossible,butonlybyradicallybreak-ingwiththeassumptionsoflegalpositivism.Dworkinarguesthatlawdoesnotconsistsolelyofrulesdeliberatelyestablishedinprecedentsandstatutes.Inhisview,lawalsoincludesgeneralprincipleswhichareimplicitwithintheestablishedblack-letterprovi-sions.Judgeshavethetaskofconstructingacoherentmoraltheorythatprovidesanappropriateabstractjustificationfortheestablishedrulesandinstitutions.Theymayinterpretandmodifyestablishedrulesinawaythatbringsthemmorecloselyintolinewiththeoverarchingabstractjustification.Thus,evenwhenjudgesmodifyestablishedlegalrulestheyaredoingsointheapplicationofdeeperlegalprinciples.Ontheonehand,Dworkin’stheoryresemblesrule-scepticisminitsdenialthatblack-letterrules,viewedinabstractionfromconsiderationsofjustice,providereliableguidestojudicialdecisioninmostcases.Ontheotherhand,thetheoryresemblesformalism,sinceitclaimsthatthejudgeneedneverstepoutsidethelawtodecidethecaseonextra-legalgroundsofsocialpolicy.Everycase,inDworkin’sview,canbedecidedbyrefer-encetothelaw:butlawcomprisesmuchmorethantheblack-letterrulesallowedforbythepositivist.410\nPHILOSOPHYOFLAW2TheEnterpriseofLegalTheoryEvenafterwehavebecomefamiliarwiththerangeofproblemsoutlinedsofar,wemayremainuncertainaboutthetypeofsolutionthattheyrequireorinvite.Weneedageneraltheoreticalaccountofthenatureoflawthatwill,forexample,explaintherela-tionshipbetweenlawandforce,elucidatethecriteriaoflegalvalidityandclarifythenatureofadjudication.Yetwhattypeoftheoryisbeingsoughthere?Wouldsuchatheorybedescriptive,astalkofinvestigatingthe‘nature’oflawseemstosuggest?Descriptiveofwhat?Whatfactsaboutlawareaccessibleonlytophilosophicalenquiryandunknowntoconventionallegalorsocialscientificscholarship?Orwouldsuchatheorybeprescriptive?Woulditbeofferingmoralidealsandcriti-calstandardsfortheevaluationoflegalsystems?And,ifso,isitnotmisleadingtoexpressthesearchforsuchatheoryinthelanguageofaninvestigationintolaw’s‘nature’?Thephilosophyoflawisoftensaidtobedividedintotwofields,knownas‘analyti-caljurisprudence’and‘normativejurisprudence’.Analyticaljurisprudenceclaims,nottoofferprescriptions(whichisthejobofitsnormativepartner),butto‘analyse’con-cepts,includingtheconceptoflaw.Differenttheoristshaveinevitablyhadsomewhatdifferentconceptionsofwhatsuchanalysismightamountto.TheformofanalysisdevelopedbythelegalpositivistJohnAustin(1790–1859)soughttodispelproblemsbybreakingdowntheconceptof‘law’intosimplerelementsthatwouldbemoretransparent.ThusAustinclaimedthatlawswerethecommandsofasovereignpersonorbodyinanindependentpoliticalsociety.Eachoftheseterms(‘sov-ereign’,‘command’andsoon)wasthendefinedinmorebasicand,Austinhoped,moretransparentterms.The‘sovereign’inasocietywastobeidentifiednotbyanylegalcri-terion(thiswouldrenderthetheorycircular),norbyamoralstandard(whichwouldconvertthetheoryintoanaturallawtheory),butbydirectreferencetoobservablefactsaboutpatternsofbehaviour.Thusthe‘sovereign’wassaidtobethatpersonorbodythatishabituallyobeyedbythebulkofthepopulation,andisnotitselfinthehabitofobeyinganyotherdeterminatepersonorbody.DespitetheconsiderableinfluencethatAustin’stheorycametoexertinlatenineteenth-centuryEngland,itexhibitedanumberofseriousdefects.InmanylegalsystemsitisnotpossibletoidentifyanypersonorbodysatisfyingAustin’sdefinitionofsovereignty:powerandobediencearedividedaccordingtosharedconstitutionalrules.Moreover,Austin’stheoryisthoroughlyreductionistincharacter,claimingthatpropo-sitionsaboutlegaldutiesarestatementsaboutthelikelihoodofsufferingsanctionsincertaincircumstances.HartendeavouredtodevelopaversionoflegalpositivismthatwouldovercomethelimitationsofAustin’sapproachwhileremainingfaithfultothebasicobjectivesoflegalpositivism.ThusHartsoughttodemonstratethatvalidlawscouldbeidentifiedbycri-teriaofafactual,non-moralkind,withoutrunningintothedifficultiesencounteredbyAustin’saccountofthesovereign;andthatpropositionsoflaw(aboutlegalobligationsandrights)couldbedistinguishedfrommoraljudgementswithoutadoptingareduc-tionistanalysisthatdeniedtheirprescriptivecharacter.HarttooktheviewthatAustin’saccountofthesovereignwasmisconceivedinsofarasitattemptedtoidentifyafundamentalsourceoflawbydirectreferenceto411\nN.E.SIMMONDSobservablebehaviouralpatterns.Whatwasrequired,inHart’sview,wasnotthenotionofabasic‘habitofobedience’butthenotionoftheacceptanceofarule.Anacceptedrulediffersfromamerehabitofobedienceinthat,inadditiontoaregularpatternofconformingconduct(whichHartcalledthe‘externalaspect’),aruleinvolvestheexist-enceofacriticalreflectiveattitudeonthepartofparticipants(the‘internalaspect’):thepatternofconformingconductisthoughtofasastandardthatoughttobecom-pliedwith;thosewhodeviatearecriticized;andthecriticismisthoughtofasjustified.HartonPrimaryandSecondaryRulesAustin’stheoryalsolackedanotherdistinctiondevelopedbyHart:thedistinctionbetweenprimaryandsecondaryrules.Primaryrulesarerulesdirectlyregulatingconduct,suchasrulesprohibitingtheftorviolence.Secondaryrulesarerulesthatregu-latetheidentification,modificationandapplicationofotherrules.Hartclaimedthattheinventionofsecondaryruleswasasimportantastheinventionofthewheel,anddescribeditasrepresentingthestepfromthepre-legaltothelegalworld.Asocialordercontainingonlyprimaryruleswouldexhibitvariousdefectssuchasinflexibilityandinef-ficiency.Mostimportantly,ifsuchasocietywereofanycomplexityitwouldbelikelytoexhibitgreatuncertaintyaboutthescopeandcontentofitsprimaryrules.Uncertaintyofthiskindcouldbeovercomebyadoptingabasic‘ruleofrecognition’whichstipulatedsomecriterionbywhichthesociety’sprimaryrulesmightbeidentified.Averysimpleruleofrecognitionmightprovidethatalltherulescarvedonacertainstonetabletweretobeobeyed.Developedlegalsystemswillhavehighlycomplexrulesofrecognition,iden-tifyingsourcesoflawsuchasenactmentsofthelegislatureanddecisionsofthecourts,andregulatingtherelationsbetweenthosevarioussources.Aswenotedearlier,HartsoughttoavoidthereductionismofAustin’stheoryandtopreservetheprescriptivecharacterofpropositionsoflawbyreproducingthe‘internalpointofview’oftheparticipants.Hart’sobjectwastoelucidatethemeaningofpropo-sitionsoflawbyexhibitingtheirrolewithinaspecificsocialcontext.Propositionsoflaw(reportingthecontentoflegalrights,duties,powersandsoforth)derivetheirsignifi-cancefromasocialcontextcharacterizedbytheexistenceofabasicruleofrecognition.Buttheyarenotdescriptionsofthatsocialcontext.Theyexpressconclusionsabouttheapplicabilityoftheruleofrecognitionandtherulesderivedfromit.Theyare,asitwere,prescriptivejudgementsmadefromtheviewpointofaspecificsystemofrules.Onemightcomparethem,byanalogy,withjudgementsaboutwhatisrequiredbytherulesofagame.Suchjudgementsare,withintheparametersofthegame,prescriptive;buttheybracketoutallquestionsaboutwhetherthereisgoodreasonforplayingthegameinthefirstplace.Hart’stheorydoesnotpurporttooffermoralprescriptions:itrepresentsanenterpriseofconceptualclarification,whichHartconceivesasbeingpriortosubstantivemoralorempiricalinquiriesintolaw.ButwhatexactlydoesHarttake‘conceptualclarification’inthiscontexttobe?ThecentralobjectiveofHart’spositivismisadefenceofwhathecallsthe‘separa-tionoflawandmorals’.This‘separation’hastwomainfacets.Oneaspectinvolvestheclaimthatlegalvalidityisamatterofarule’sderivabilityfromabasicruleofrecogni-412\nPHILOSOPHYOFLAWtioninalegalsystem:arulethatemanatesfromanappropriatesourceisvalidirre-spectiveofitsjusticeorinjustice(and,similarly,thejusticeofaruledoesnotmakeitlawifitdoesnotemanatefromanappropriatesource).Aregimeexhibitingcertainformalfeatures(acceptancebyofficialsofabasicruleofrecognition,combinedwithgeneralcompliancewiththelawsflowingfromthatbasicrule)isalegalsystemirre-spectiveofwhetheritpursuesjusticeorexploitation.Theotheraspectoftheseparationoflawandmoralsistheclaimthatknowledgeofthecontentofthelawdoesnot,initself,haveanybearingonwhatoneshoulddofromamoralpointofview.Theremayormaynotbeamoralobligationtoobeythelaw.Ifthereis,thenknowledgeofthelawwillgiveprecisecontenttothatobligation,butsuchanobligationisnotentailedbytheconceptoflegalvalidity.Itisnotpartoftheconceptoflawthatitismorallybinding.WhydidHartbelievethatweshouldaccepttheseclaimsabouttheconceptoflaw?HislongassociationwiththeOxfordordinarylanguageschoolofphilosophyhasledsometoassumethatHartsawhimselfassimplyunearthingandlayingbeforeusthecriteriathatareimplicitinourordinaryusageoftheword‘law’.Itshouldbeobvious,however,thatordinaryusageisunlikelytoprovideanenlighteningguideontheseissues.Toalargeextentordinaryusagesimplyexpressesandevincestheambivalencesarticulatedwithintheoreticaldebate.Dworkinhasdrawnadistinctionbetween‘borderline’disputesand‘pivotal’disputes.Inaborderlinedispute,wehaveasharedcriterionbutdisagreeabouttheextentofitsapplicability:forexample,weagreeonsomegeneralconceptof‘art’butdisagreeaboutwhetherthatconceptisapplicabletophotography.Inapivotaldispute,wehavenosharedcriterion:hereourdisagreementaboutphotographyismerelysymptomaticofadeeperdisagreementabouttheessentialnatureofart.DworkinappearstosuggestthatHartviewedlegaltheoryasbeingconcernedwithborderlinedisputes,whereas(inDworkin’sview)legaltheoryisconcernedwithapivotaldispute.ThisclearlymisrepresentsHart’sconceptionoflegaltheory.Harthimselfpointedout,intheintroductiontohisEssaysinJurisprudenceandPhilosophy(1983),thathisanalysisoftheconceptof‘law’wasnotintendedasapieceofordinarylanguagephilosophy.InHart’sview,theuncertaintiessurroundingtheconceptoflawrequireachoicetobemadebetweenrivaltheoreticalstandpoints:theycannotberesolvedsimplybycloseattentiontoordinaryusage.Nordotheyrequireasimpledecisionontheapplic-abilityoftheconcept’s‘penumbra’.The‘coreofsettledmeaning’oftheconceptof‘law’mayreflect,notasharedcriterion,somuchasagreementonarangeofinstanceswherediversetheoreticalpositionsandcriteriahappentooverlap.ThuswemayallagreethattheenactmentsoflegislaturesinFrance,theUnitedKingdomandCanadaarecentralinstancesof‘law’,yetwemayhavequitedifferentreasonsforsoregardingthem.Inyourviewthecriticalfactmaybethemoralauthorityofthoselegislatures,orthejusticeoftheirenactments,whileinmyviewitmaybetheeffectivecontrolthattheyexertovertheirterritory.Hart’sreasonforadoptingapositivisttheoryoflawappearstobehisbeliefthatpositivismreflectsthemostimportantfeaturesoflawasasocialphenomenon.Suchajudgementobviouslyrequirescriteriaof‘importance’and,ifthetheoryistoplayaclar-ificatoryroleasapreliminarytosubstantiveenquiry,thecriteriashouldbedrawnfromvaluesthatareasweakandwidelysharedaspossible.ReconstructionofHart’s413\nN.E.SIMMONDSargumentsinTheConceptofLawsuggeststhathispointisasfollows.Giventheveryminimalandwidelysharedgoalofsurvival,andgivencertainverygeneralfactsabouthumannatureandcircumstances(Hartspeakshereofa‘minimumcontentofnaturallaw’)itcanbedemonstratedthatsocietiesrequirerulesofconduct.Inasocietyofanycomplexity,sharedruleswillbesustainableonlyiftheyarepubliclyidentifiablebymeansofaruleofrecognition.Theprovisionofsuchasetofpubliclyidentifiablerulesisafunctionoflaw.Hart’stheorydevelopsanaccountoflaw’snaturethatemphasizesitscentralroleinprovidingpubliclyascertainablerules.OnlyrulesidentifiablebyreferencetothebasicruleofrecognitionareacknowledgedbyHarttobeinstancesof‘law’.Moreopen-endedcriteriaofjusticeorfairness,eventhoughadmittedinforensicdebatesandregardedbylawyersaslegaldoctrinalarguments,aretreatedbyHartasextra-legalconsiderationsthatmaybelookedtowhenthelegalrulesfailtoyieldadeterminateanswer.Thejus-tificationforsemanticregimentationofthiskindisnotthatitreflectsordinaryusage,northatitadjudicatesonpenumbraluncertainties,butthatitreflectsthemostimpor-tantanddistinctivefeaturesoflaw,judgedfromtheviewpointofhumansurvival.OncetheminimalelementofevaluationinvolvedinHart’sviewoflawisexposed,however,itcanbedifficulttopreventitspreadingsofarastoendangerthepositivismofthetheory.Forexample,Hartclaimsthatpositivismclarifiestheissuethatisfacedbyacitizenwhohastodecidewhethertoobeyanunjustlaw.InHart’sviewitisbetterandmorehonesttosay,‘Thisislaw,butbecauseitisunjustIshouldperhapsdisobeyit’,ratherthan,‘Becausethisruleisunjust,itisnotlaw’.Yetwhy,wemightask,shouldonewayofexpressingthepointbethoughtclearerthantheother?ThemostplausibleanswerisonethatinvolvesarejectionofHart’sstrictseparationoflawandmorals.Ifpubliclyascertainablerulesarenecessaryforsurvival,thatseemstogiveusamoralreasonforcomplyingwiththemevenwhentheyareunjust.Afterall,ifpeoplewerepreparedtocomplyonlywhentheyconsideredtherulestobeper-fectlyjust,therewouldbenopointinhavingasystemofpubliclyascertainablerulesinthefirstplace.Thesemoralreasonsforcompliancecontinuetoapplyevenwhentheyareoutweighedbyconflictingconsiderations(theruleis,letussay,grosslyunjust).Thebestreasonforinsistingoncontinuingtoapplythelabel‘law’eventounjustrulesisthatthisusageacknowledgesthecontinuedforceandapplicabilityofthemoralcon-siderationsstemmingfromtheneedforasystemofpubliclyascertainablerules.Torefusetoacknowledgethelegalvalidityofanunjustrulewouldbetotreatitasamerenon-entitylackingallmoralsignificance,andsowouldbetodenythecontinuedrele-vanceoftheneedforpubliclyascertainablerules.Hartisconcernedtodefendtwotheses,whichwemightcall‘theexclusivityofsources’and‘themoralneutralityofpropositionsoflaw’.‘Theexclusivityofsources’istheideathatlegalvalidityisamatterofderivabilityfromtheruleofrecognition,andnotamatteroftherule’sjustorunjustcontent.‘Themoralneutralityofpropositionsoflaw’istheideathatknowledgeofone’slegaldutiesdoesnotinitselfentailanyjudge-mentsabouthow,fromamoralpointofview,oneshouldbehave.Yetthemostpower-fulreasonsforadvocatingtheexclusivityofsourcesrequiretherejectionofthe‘moralneutrality’thesis.Thiswouldbetotallyunacceptableifitforcedonustheviewthatlawismorallyconclusiveofhowoneshouldbehave.Nosuchmoveisrequired,however.The‘moralneutrality’thesisshouldbeabandonedinfavourofamuchweakerclaim,that414\nPHILOSOPHYOFLAWcanbelabelled‘themoralnon-conclusivityoflaw’.Thisthesisclaimsthatpropositionsoflawareindeedaspeciesofmoraljudgement:theyarejudgementsabouttheapplic-abilityofthosemoralreasonsflowingfromtheneedforabodyofpubliclyascertain-ablerules.Yetsuchjudgementscanneverbemorallyconclusive,sincetheyembodyonlyonerangeofmoralconsiderationswhichwillneedtobebalancedagainstmanyotherfactors.3FromPositivismtoNaturalLawLegalpositivismoriginatedinoppositiontotheoriesofnaturallaw.Somepositivistsrejectedtheclaimstoobjectiveknowledgeofmoralitythatformedthebedrockofthenaturallawtradition.Kelsen,forexample,adoptedanETHICALNON-COGNITIVIST(p.440)stancewhichdeniedthepossibilityofknowledgeofmoralnormsorvalues.Thecentralcoreofhisworkwastheattempttodemonstratethatscientificknowledgeofpositivelegalnormswasinfactpossible.Itisinthecontextofthisprojectthatweshouldinter-prethisremark(mentionedearlier)aboutthebasicnormastheminimumelementofnaturallawwithoutwhichacognitionoflawisimpossible.Hart’spositivismwasclosertoatraditionofthoughtinitiatedbyJeremyBENTHAM(1748–1832)(chapter35).Benthamdidnotwishtodenythatwecanhaveknowledgeofmoralrightandwrong.Infact,hislegalpositivismwasinmanyrespectstheconse-quenceofhisviewthatthePRINCIPLEOFUTILITY(chapter35)(requiringthatweshouldactsoastomaximizethegreatesthappinessofthegreatestnumber)isthesupremeprincipleofmorality.Thisutilitarianstanceledhimtoregardtheoriesofnaturallawasbeingfoundedonconfusionandposingathreattoorderedlibertyunderlaw.TheopeningsalvoofBentham’spositivistassaultonnaturallawwasdirectedagainsttheworkoftheEnglishjudgeandjuristSirWilliamBlackstone(1723–80).BlackstonewastheauthorofthefirstsystematicexpositionofEnglishlaw,publishedastheCommentariesontheLawsofEnglandin1765.PriortoBlackstone,oneacquiredaknowledgeofEnglishlawbyimmersioninitschaosofdetails.Lawbooksweremostlyhugecompendiaoflegalknowledgearrangedonnoveryrationalprinciple.Adoptingmodelsdrawnfromcontinentallegalscholarship,BlackstonearrangedtheinstitutionsofEnglishlawaroundabasicschemeofnaturalrights.InaccordancewiththeviewofnaturallawwriterssuchasJohnLOCKE(chapter29)(1632–1704),BlackstonetreatedEnglishlawastheconcretizationandenforcementofpeople’snaturalrights.Thelawwasnotsimplyamassivelistofdisorderlyrules,butasystematicbodyofprin-ciples.Itwascapableofsystematicstudyandexpositionpreciselybecauseandinsofarasitwasbasedonprinciplesofnaturallaw.Consequently,Blackstone’saccountofEnglishlawwasinclinedtopasssmoothlyfromexpositionofthelawtojustification,sincetheveryschemeofexpositionwasstructuredbynaturallawprecepts.InBentham’sopinion,thistypeofnaturallawtheoryconfusedquestionsaboutwhatthelawiswithquestionsaboutwhatitoughttobe.Theexpositionofexistinglegalrulesisonething,heargued,whiletheircriticalevaluationisanother.Itdoesnotfollowfromtheexistenceofalegalrulethatitismorallyright.Nordoesitfollowfromtheimmoralityofarulethatitisnotlaw.Benthambelievedthatconfusionofthesediffer-entquestionswaslikelytoinduceeithersupineconformitywithoutanypreparedness415\nN.E.SIMMONDStocriticizethelaw(sinceitwouldbeassumedthatthelaw,beinglaw,mustberight)or,alternatively,irresponsibleanarchism(sinceitcouldbeclaimedthatanyparticularrulewhichisconsideredtobeunjustisthereforenotlawatall).ItisdoubtfulifBlackstonewasreallyguiltyofthecrudeconfusionsthatBenthamclaimedtodiscoverinhiswork.MuchofBentham’sfirewasdirectednotsomuchatBlackstone’sviewoftherelationshipbetweenlawandmoralsasatBlackstone’semphasisonnaturalrightsasthebasisofmorality.Benthamconsideredalltalkofnaturalrightstobe‘nonsense’andhebelievedthatsuchtalkmerelyservedtodivertourattentionawayfromtheonlyrelevantconsiderations:considerationsofutility.IndeedBentham’swholetheoryoflawwasfundamentallyshapedbywhathetooktobetherequirementsofutility.Hedevelopedanelaborateanalyticalframeworkfortheintellectualreconstructionofpositivelaw,bymeansofwhichhesoughttoexplaintheentireapparatusoflegalconcepts(includingrights,duties,powersandproperty)intermsofageneraltheorythatviewedlawasasetofcommandsbackedbysanctions.ForBentham,suchreconstructionservedanessentiallypracticalpoint,bymakinglawandlegalthinkingmorecloselyfittedtoitsroleasaninstrumentofutility,andbyeliminatingtheconfusionsthatsooften(inBentham’sview)obscuredthetruerequire-mentsofutilityandinsulatedthelawfromthedemandsofreform.Hisaimwasnottoexplicatelegalthinkingsomuchastoreviseit.ThemoveawayfromBentham’sclearUTILITARIAN(chapter35)objectives,andtowardsamorepurely‘analytical’jurisprudence,commencedwiththeworkofBentham’sowndisciple,JohnAustin,whomwementionedearlier.Austin’stheoryoflawisinmanyrespectsasimplifiedversionofBentham’s.But,whereasBenthammighthavepresentedthetheoryasadeliberateproposaltoreviseordinaryassumptionsaboutlaw,Austinseemstointendhistheoryasanexplicationofthoseassumptions.Indeed,farfrommountinganassaultonthevestedinterestsoflawyersinthewaythatBenthammighthavewished,Austin’sversionoflegalpositivismprovidedapowerfulunderpinningforthelawyer’sclaimtobeanexpertinatechnicalandself-containedareaofknowledgedistinctbothfrommoralityandfrompolitics.Austinandhisdisciplesthusintroducedintolegaltheoryabasicuncertaintycon-cerningthestatusofinvestigationsintothenatureoflaw.Thisuncertaintycontinuestodogthesubject.Superficialexaminationofthewritingsoflegalpositivistsmightleadonetobelievethattheirtheoriestreatlawinentirelyformalterms,involvingnojudgementaboutlaw’spurposeorvalue.Yetifthelegalpositivististoofferanintegratedsolutiontothevariousproblemsoutlinedatthestartofthischapter,somefeaturesmustbeselectedfromtheimmenselycomplexsocialphenomenaoflawandgivenacentralrolewithinthetheory,sothattheymayserveasapointoffocusaroundwhichotherfeaturescanbearranged.ThusHartchoosestoemphasizethelawyer’srelianceuponrulesema-natingfromalimitednumberofauthoritativesources(especiallylegislativeenactmentsandjudicialdecisions).Otherfeaturesoflawareeitherreinterpretedormarginalized.ThisselectionoffocusexpressesHart’sjudgementaboutthefundamentalpurposeorfunctionoflaw:theprovisionofabodyofpubliclyascertainablerulesasabasisfortheorderingofconductinaworldcharacterizedbydisagreement.YettheghostlyinfluenceofAustinseemstoinduceinHartadesiretoleavethisbasicfoundationofhistheoryunderstatedandrelativelyunexplored.AsthecontemporarynaturallawyerJohn416\nPHILOSOPHYOFLAWFinnis(b.1940)aptlyputsit,‘Hart’smethodpointsoutalandwhichislefttohisreadersandhearerstohazardtoenter.’FinnisonNaturalLawInFinnis’sview,thetraditionofnaturallawthoughthasbeenmisunderstoodandmis-representedbylegalpositivists.Positivistshavetendedtoassumethatthecentralthesisofnaturallawtheoryistheclaimthatlegalvalidityisamatterofmoralbindingness(or,perhaps,ofjustice)sothatarulethatisnotmorallybinding(ornotjust)cannotbesaidtobelegallyvalid.Asitistraditionally,ifsomewhatmisleadingly,put:lexinjustanonestlex(‘anunjustlawisnotalaw’).Positivistsregardthisclaimassimplybeingconfusing,andprefertoapplythelabel‘law’toalltherulesemanatingfromparticularformalsourceswithinalegalsystem,irrespectiveoftheirmoralstatus.Yet,aswehaveseen,theclaimthatthiswayofthinkingisclearer,andreflectsthemostdistinctiveandimportantfeaturesofthesocialphenomenaoflaw,itselfrequiressomeevaluativestancefromwhichwecanmaketherequiredjudgementof‘importance’.Finnisclaimsthatnaturallawyershaveneverreallywishedtodenyvalidityinthelawyer’ssensetounjustorimmoralrules.Rather,theirobjecthasbeentoexplorethelandtowhichHart’smethodpoints.Inotherwords,theybelievethatacoherenttheoryoflawcanonlybetheconsequenceandexpressionofadeepermoralandpoliticaltheory.Infact,Finnisclaims,weunderstandthenatureoflawbygraspingthewayinwhichlawisthesolutiontoaproblemofpracticalreason.Withthisendinview,Finnisdevelopsanaccountoftheobjective‘goods’or‘formsofhumanflourishing’whichwewillpursueifwearefullyrational.Thesegoodscanbepursuedonlywithinacommunity,andthemaintenanceofacommunitythatsustainstheabilityofeachindividualtopursuetheobjectivegoodsamountsto‘thecommongood’.Thecommongoodisthereforenotanaggregativeconception,involvingthe‘maximization’ofindividualgoods,butaframeworkofinstitutionsandconditionsmakingindividualpursuitofagoodlifepossible.Theexistenceofanappropriatecom-munityrequirestheco-ordinationofconductaroundsharedrules,andtheprovisionofsuchasetofrulesisthecentraltaskoflaw.Inonesense,therefore,Finnis’sconceptionoflawisnotdissimilartoHart’s.Theybothviewlawasasetofpubliclyascertainablerules.Finnis,however,believesthatthisconclusionistobearrivedatonlybymeansofanunderstandingoflaw’srelationshipwiththerequirementsofpracticalreason.Inits‘focal’sense,lawservesthecommongood,andthisrelationshiptothecommongoodinevitablyshapesourthinkingaboutlaw,andthevocabulary(usingtermssuchas‘rights’and‘duties’)weusetodescribelaw.Legalinstitutionsmayattimesbeusedforpurposesofoppressionorexploitation,butthesearedegenerateinstancesoflawwhichwecanonlyunderstandbygraspingthewaysinwhichtheyresemble,andyetdivergefrom,lawinits‘focal’instances.Thetra-ditionalmaximlexinjustanonestlex,Finnisconcludes,wasintendedtohighlightthedegeneratenatureofsuchusesoflaw.Finnishasperformedagreatserviceinoverturninghithertoprevailingcaricaturesofnaturallawtheory.YetheseemstounderestimatethesignificanceofHart’sapparentdesiretoconstructhistheoryoflawfromastandpoint(oftheaimof‘survival’)whichisbothweakandwidelyshared.Hart’srelianceuponsuchaminimalevaluative417\nN.E.SIMMONDSstarting-pointreflectsanunderstandingofhisenterpriseasaclarificatoryonewhichispriortosubstantivemoralenquiryorempiricalinvestigation.Harthopestoconstructacoherentconceptoflawthatmaybeemployedasatoolinmoralandempiricaldebate,therebyovercomingtheproblemswhicharisewhendifferentparticipantsindebateemploydifferentconceptionsoflaw;or,stillworse,whencontroversialmoralorempiri-caltheoriesaboutlawareintroducedintheguiseofdefinitionsof‘law’.FinnisappearstoviewHart’sminimalismasafailuretoexplorefullytheunderpinningsofhisowntheory;butitisprobablybettertoregarditasarefusaltobuildtheclarificatoryenter-priseonarichlycontroversialfoundationwhenaweakandminimalfoundationwillserveequallywell.4TheoryasInterpretationPhilosophyisoftenthoughttobecharacterizedbyaconcernforthenecessaryanduni-versal.Particular,contingentfactsarethoughtofasmattersforempiricalenquiry,notphilosophicalreflection.Thusthephilosophyoflawtraditionallyconcernsitselfwithfeaturesoflawthatarelogicallynecessary,asdefiningfeaturesofasupposedlyuni-versalconceptoflaw.Thepeculiarcharacteristicsofthisorthatparticularlegalsystemaremattersforsocialscienceratherthanphilosophy.WriterssuchasHartandFinnisofferconceptsoflawthatpurporttobeuniversal.Theydonotsimplydescribecharac-teristicfeaturesofthelegalsystemsoftheUnitedKingdomorAustralia:rather,theyofferuscriteriaforsayingofanyformofsocialorderingwhetherornotitshouldberegardedasaninstanceoflaw.Familiarthoughthisconceptionofphilosophymaybe,itisnotbeyondchallenge.Wemayreflectuponthemeaningandsignificanceofpracticeswithinourownsocietyandculture,withoutanypretensiontouniversality,andsuchreflectionsmaybearmanyofthehallmarksofphilosophicalthought.Indeed,wemaydiscoverthatitwasthedesirereflectivelytounderstandourownpracticesthatdroveusintothesearchforuniversalandnecessaryfeaturesinthefirstplace.ThetheoryoflawdevelopedbyDworkinadoptsafocusthatisculturallyspecificratherthanuniversal,andthatseeksareflectiveunderstandingofourownpractices.AccordingtoDworkin,thephilosophicaldebateconcerningthenatureoflawisnot,whenproperlyunderstood,asearchforasemanticdefinitionof‘law’,orasearchforauniversalconcept.Ratheritarisesoutoftheadoptionofan‘interpretiveattitude’inrelationtoourownpractices.Theinterpretiveattituderaisesaquestionaboutthegeneralpointorrationaleofapractice,notasatheoreticalbutasapracticalissue:theultimatequestionishowbesttocontinuethepractice.Questionssuchasthoseconcerningthenatureoflawreflectdisputesaboutthemeaningorpointofourlegalpractices.Suchdisputescannotberesolvedbyreferencetotheintentionsorunderstandingsoftheparticipantsinthepractice,becausetheparticipantswillthemselvesbeembroiledinthedispute.Thusasocietymighthavepracticesofcourtesy,involvingraisingone’shattopeopleincertaincircumstances,oropeningdoorsforthem.Yettheparticipantsinthesepracticesmightdisagreeabouttheunderlyingpointofcourtesy.Istheobjecttoexpressajudgementofrespectabouttherecipientofcourtesy?Or,quitetothecontrary,istheobjecttoprovideaneutral,non-418\nPHILOSOPHYOFLAWjudgemental,wayofdealingwithpeopleregardlessofwhatyouthinkofthem?Thechoicebetweenrivalinterpretationsofthiskindmightmakeadifferencetowhatwewouldseeasanappropriatewayofcontinuingthepractice.Shouldwe,forexample,makeaspecialefforttoraiseourhatstosoldiersreturningfromawar?Ifwecannothaverecoursetotheintentionsandunderstandingsoftheparticipants,howarewetochoosebetweenrivalinterpretationsofapractice?Dworkinoffersageneraltheoryofinterpretationwhichhecalls‘constructiveinterpretation’.Onthisview,thebestinterpretationwillsatisfytwocriteria,of‘fit’,andof‘appeal’.Criteriaoffitconcerntheabilityoftheinterpretationtoaccommodatetheuncontroversiallyobservablefeaturesofthepractice.Aninterpretationneednotbeaperfectfit,butitmustcomeuptoacertainthresholdofadequacy.Withintheconstraintsoffit,wemustchoosetheinterpretationthatmakesthepracticeintothemostappealingpracticeitcanbefromamoralpointofview.Wenotedabovethatthereisroomforsomeuncertaintyaboutwhetherphilosophi-calenquiryintothenatureoflawisintendedtobedescriptiveorprescriptive.Dworkin’stheoryofinterpretationpresentslegaltheoryashavingbothadescriptiveandapre-scriptiveaspect.Interpretivetheoriesoflawwillbedescriptivetotheextentthattheymustsatisfytheconstraintsoffit;theywillbeprescriptiveinsofarasonemustchoosethemorallymostappealinginterpretationamongstthosethatsatisfytheconstraintsoffit.Atheoryoflaw(oranyotherinterpretationofapractice)will,Dworkinholds,involvethreestages.Atthe‘pre-interpretive’stagewemustachievearoughprelimi-naryidentificationofthepracticeswearediscussing.Dworkinconcedesthateventhepre-interpretivestageinvolvesinterpretation(everydescriptionofhumanconductinvolvesinterpretation),buttheinterpretationsinvolvedatthisstagemaybewhollyuncontroversial.Nextcomesthe‘interpretive’stage,wherewemustarriveatageneralinterpretationofthemeaning,pointorrationaleofthepractice,choosing(withintheconstraintsoffit)theinterpretationthatmakesthepracticeintothebestpracticeitcanbefromamoralpointofview.Finallycomesthe‘post-interpretive’stage,wherewereviseouroriginalunderstandingofwhatthepracticerequiresinthelightoftheinter-pretationnowadopted(forinstance,wedecidethat,sincecourtesyisanexpressionofone’sjudgementsofrespect,itrequiresustoraiseourhatstosoldiersreturningfromawar,butnottothelocalsquire).Thisaccountofthenatureoflegaltheory,andtheviewofinterpretationonwhichitdepends,raisealargenumberofquestions.Somepositivistsfeel,forexample,thatimportantquestionsarebeggedwhenaprescriptive,justificatoryaspectisbuiltintoone’sconceptionofthelegaltheoreticalenterprise.Isitnotthewholeobjectofposi-tivism,theyask,topresentlawasbeingmorallyneutral?Ontheotherhand,onecanarguethatDworkinactuallyoverestimatesthecontrastbetweenhisconceptionoflegaltheoryandthatheldbyHartandothers.For,aswehaveseen,Hart’stheoryalsoadoptsaninterpretationofthepointoflaw(theprovisionofasetofpubliclyascertainablerules)andthenrevisesourordinaryunderstandinginthelightofthatinterpretation.Dworkinalsoseemsoblivioustothewayinwhichaninterpretiveandculturallyspe-cificviewofthetasksoflegaltheoryfundamentallyerodesthedistinctionbetweenphi-losophyandsociology(formerlypolicedbythedistinctionbetweentheuniversalorthenecessaryandtheparticularorthecontingent).Inresistingattemptstocollapsethe419\nN.E.SIMMONDSphilosophyoflawintosociology,Dworkinmaintainsthattheinterpretationofaprac-ticeisnecessarilypriortoastudyofitssocialorhistoricalcontext.Wemaywellwishtochallengethisclaim,however.Whenviewedinisolation,apracticemayhaveanappar-entmeaningthatissubvertedonceitisrelocatedinitssocialcontext.Practicesofmalecourtesytowardswomen,forexample,maytakeonadifferentmeaningwhenlocatedinthecontextofamale-dominatedsocietywithinwhichwomenaregenerallysub-ordinated.Equally,anunderstandingofthehistoricalcontextmaywellbenecessaryifwearetograspthegenesisandcharacteroftheinterpretivedisputesthatarise:suchdisputesare,afterall,notfree-floating,butariseoutofhistoricallyproducedfracturesandtensionswithinformerlyunproblematicpractices.Anhistoricalandsociologicalenquirymaythereforebenecessaryifphilosophyistoattainafullytransparentandself-consciousunderstandingoftheinterpretivedisputesinwhichithopestoengage.Dworkin’sneglectoflaw’ssocialcontextisparticularlyironicsince,asweshallsee,heviewslawasadeepexpressionofequalityandinterpretsinthislightthelawyer’sconcernforconsistencyandprinciple.Radicalcriticsoflaw,suchasMARXISTS(chapter34)andFEMINISTS(chapter20),mightwellagreethatlawinvolvesa‘discourse’ofequality;buttheywouldpointoutthattheapparentmeaningofthisdiscourseissubvertedoncewelocatelawinthecontextofadeeplyhierarchicalsocietyfoundedonoppressionanddomination.Thelawthenappearsnotastheexpressionofadeepconcernforequality,butasanattempttomystifyandmisrepresentsocialrelationsbypresentingthemasbeingfoundedonequality.5LawasIntegrityHavingoutlinedDworkin’sviewsonthenatureoflegaltheory,whatofhissubstantiveconclusionsaboutthenatureoflaw?Whatinterpretationofourlegalpracticesdoesheadopt?Dworkinconsidersthreerivalinterpretationsortheoriesoflaw,whichhecalls‘con-ventionalism’,‘pragmatism’and(hisownpreferredtheory)‘lawasintegrity’.Eachtheoryisevaluatedagainstthedescriptivecriteriaof‘fit’andtheprescriptivecriteriaofmoral‘appeal’.Conventionalismresembleslegalpositivisminitsemphasisonthepublicascertain-abilityoflegalrules,butitproffersthisviewinthecontextofawiderinterpretationofthepointoflaw.Accordingtotheconventionalist,thepointoflawistogivecitizensfairwarningofthecircumstancesinwhichthestate’scoercivepowerswillbeused,andtoprotecttheexpectationsformedinrelianceonthiswarning.InDworkin’sview,thistheoryfailstofitthefactsoflaw,sinceitcannotexplaintheabilityofcourtstoinno-vatebygrantingnewremediesincasesthatarenotclearlycoveredbyanyexistingrules.Whatiftheconventionalistrespondsbysayingthatthevalueofprotectingexpec-tationsmustbetradedoffagainstthedemandsofflexibility,sothatsomejudicialinno-vationincasesnotcoveredbyexistingrulesispermissible?Inthatcase,Dworkinargues,conventionalismwillfailprescriptivelysinceitproposesanirrationalwayoftradingoffflexibilityagainsttheprotectionofexpectations.Ifsuchatrade-offisthepointoflaw,itisbesteffectedcase-by-case,sothattheexpectationsengenderedineachcasecanbeidentifiedandweighed.420\nPHILOSOPHYOFLAWPragmatismisatheorythatproposespreciselysuchacase-by-caseapproach.Thepragmatistisreallyaspeciesofrule-sceptic,sinceheorsheholdsthatjudgesshouldnotviewthemselvesasbeingboundbypre-existinglegalrules.Rather,thejudge’staskshouldbetogivethedecisionthatwillhavethebestconsequences,allthingsconsid-ered.Theexpectationsthathavebeencreatedbythepublicationoflegalrulesmustbetakenaccountofbythepragmatistjudgewhenheorshetriestodeterminewhichdeci-sionwillbeforthebest.Itmayfrequentlybethecase,therefore,thatthepragmatistjudgehasverypowerfulreasonsforapplyingpre-existinglegalrules.Yetsuchreasonsareneverconclusiveinadvanceoftheparticularfactsofthecase:theymustalwaysbeweighedagainstotherconflictingconsiderations.PragmatisminitsturnisrejectedbyDworkin,andheturnstothedevelopmentofhisowntheory.Thishingesonthevalueof‘integrity’inlaw.Dworkinclaimsthatweexpectthelawtoexpressandembodysomemoreorlesscoherentmoralvision.Whilethelawmaycontaincompetingmoralvalueswewouldnotexpectittostrikewhollyarbitrarycompromisesbetweenrivalmoralpositions,evenifsuchcompromisesgaveexpressiontotheprevailingbalanceofpower(intermsofnumbersofvotes)betweentheadvocatesofrivalpositions.Wevaluethemoralcoherenceorintegrityofthelawquiteindependentlyofthevalueweputonthelaw’sjustice,forweexpectthelawtoexhibitacertaindeepmoralcoherenceevenwhenwepersonallysubscribetosomerivalmoralvision.Whenthestateemploysitscoerciveapparatusinaccordancewithacoherentmoralposition,ittreatsitscitizensasequals.Thepointoflaw,Dworkinclaims,istoensurethatgovernmentstreattheircitizensasequalsinthatsense.Thelawyer’sconcernwithprinciplesandconsistencyisaconcernnotwiththepublicascertainabilityofrulesbutwiththemoralcoherencethatisadeepexpressionofequality.DworkinonAdjudicationHowthenshouldajudgesetaboutthetaskofadjudication?AtthispointDworkinbringshistheoryofinterpretationintoplayonceagain.Ajudgeistodecidecasesbyreferencetothemorallybestinterpretationofthematerialsoflaw.Ratherthanthinkingofthejudgeasconfrontedbyafinitenumberofblack-letterruleswhich,onceexhausted,throwusbackontoextra-legalconsiderationsofsocialpolicy,thejudgeconfrontsanenormousarrayofstatutesandcaseswhichmustbesubjectedtoaprocessofinterpretation.Thejudgemustfindinterpretationsofthesematerialsthatdiscoverwithinthemsomecoher-entmoralvision.Suchinterpretationsmustsatisfytheconstraintsoffit,andwithinthoseconstraintsmustmakethelawasgoodasitcanbefromamoralpointofview.Infact,Dworkinclaims,thisprocessofinterpretationisofapiecewiththeinvestigationsoflegaltheorysothatjurisprudence(orthephilosophyoflaw)isbutthefirstandmostabstractstageofadjudication.Thistheoryoffersanelegantsolutiontoanumberofintractableproblems.Forexample,wenotedearlierthewayinwhichitenablesustoexplainhowjudgescanmodifytheestablishedrulesoflawwhilestillbeingboundbylaw,afeatureofadjudicationwhichposesseveredifficultiesformostothertheories.Someseriousproblemsremain,421\nN.E.SIMMONDShowever.OnDworkin’stheory,judgesmustdecidecasesnotbyreferencetowhattheyconsidertobejustintheabstract,butbyreferencetotheconceptionofjusticethattheyfindembodiedintheexistinglaw.Yet,indecidingwhatthatconceptionis,theymustrelyupontheirpersonalmoralviewpoint.Sinceitisquitelikelythatthe‘constraintsoffit’willprovenotsoconstrainingafterall,istherenotamplescopeherefordifferentjudgestoreachdramaticallydiverseconclusionsaboutwhattheexistinglawis?Indeed,Dworkininsiststhatitisuptoeachcitizentodecideforhimselforherselfwhatisrequiredbylaw,anditisinprinciplepossibleforacitizentobecorrectwhenallthejudgesarewrong.Ofcourse,convergenceinlegaljudgementsmaybesecuredbypre-existingconvergenceinmoraljudgement.Butisitnotplausibletosuggestthatamajorobjectofhavinglawistomakeitpossibleforustohavesharedruleseveninaworldcharacterizedbyseriousandfar-reachingmoraldisagreements?Theperceivedneedforauthoritativerulesinacontextofmoraldiversitymakesconventionalismseemattractive.Dworkin’scriticismsofconventionalismareinfactsomewhatmisconceived,fortheyassumethattheconventionalistisconcernedwiththeprotectionofexpectationsthathavebeenformedinrelianceonpublishedrules,andthistypeofconcerncouldbeadequatelyaccommodatedbyapragmatistapproach.Butabetteraccountofconventionalismwouldseeitasbeingconcernedtoencouragetheformationofstablemutualexpectationsaboutconduct,andapragmatistapproachtoadjudicationwouldbeinferiortoconventionalismwhenjudgedbythisobjective.SomeproblemswithDworkin’sanalysisstemfromthecharacterofamodernlegalorder.WhenBlackstonesystematizedhisexpositionofEnglishlawaroundahandfulofbasicnaturalrights,thelawwasmainlyconcernedtoenforcerightsofpropertyandtodealwithobviouswrongssuchastheftorassault.Also,thelawhadbeenmainlydevelopedbyjudges,whowereconcernedtopreservethecoherenceoftheirdecisionswiththeexistingfabricoflaw.Oncethelawislargelyaproductofstatute,andthestatutesareenactedinpursuitofahostofverylooselyrelatedorevenconflictingsocialpolicies,isthereanygoodreasontoexpectittoexhibitahighdegreeofmoralcoherence?Thegrowingmuddleandcomplexityofmostmodernlegalordersmight,infact,provideapowerfulexplanationofDworkin’sowntheory.Lawyersregularlyconstructtheoriesasabasisforthedecisionofcases,inthesensethattheyneedtogeneralizefromcertainspecificinstances,ortofindaperspectivefromwhichtointerpretspecificstatutorywords.Yetthelawyer’stheoriesoperateataverylowlevelofabstraction.WhatissostrikingaboutDworkin’svisionoflawistheabstractphilosophicalplaneontowhichhedriveslegaldoctrinalargument.Suchabstraction,however,mayitselfbearesponsetothegrowingcomplexityandincoherenceofthelaw.Ifoneisdeterminedtoseekcoherence,onemustseekitatevermoreabstractlevels.Legalpositivismmightbeseenastheoppositeresponsetothesameproblem:facedbygrowingincoherenceatthelevelofgeneralprinciples,onefocusesmoreandmorenarrowlyandexclusivelyuponblack-letterrules.Dworkin’stheorycanbecomparedandcontrastedwiththeworkofLonFuller(1902–78).Fuller’sworkseemstohaveinfluencedDworkininseveralfundamentalrespects,althoughtheinfluenceisnotacknowledged.LikeDworkin,Fullerseesthejudgeasbeingconfrontedbyataskthatcannotbereducedtotheapplicationofabasicruleofrecognition.Adjudicationis,inFuller’sview,apurposiveenterpriseguidedby422\nPHILOSOPHYOFLAWthecomplexmoralaspirationof‘fidelitytolaw’.WhereasDworkinconnectsthevalueoflawtoaleft-liberalconcernforequality,Fullerinterpretsthecentralvalueoflawintermsofamoreconservativeconcernforlibertyandorderunderthegovernanceofgeneral,prospectiveandpublishedrules.Inmanycases,fidelitytolawwillrequireadherencetopositedrules;but,Fullerclaimed,toimaginethatthisrepresentstheulti-mateguidingprincipleofadjudicationistoconfusefidelitytolawwith‘subserviencetoauthority’.Thus,apowertomodifytherulesinsomecasesmaybeconsistentwiththejudge’sfundamentalduty.Onappropriateoccasions,fidelitytolawmayrequirearefusaltoapplytherulesestablishedbyauthority(aswhere,forexample,thoserulesaresovagueastodefeattheideaofgoverningconductbyprospectiverules:anideathatFullertakestobeanintegralpartoftheconceptof‘law’).Insomesituations,thejudge’sdutymaygiverisetointractablemoraldilemmas:enforcingahopelesslyvaguestatutemayunderminetheidealsofclaritythatareinherentintheruleoflaw,butassumingapowertostrikedownsuchstatutesmayitselfgiverisetogreatuncertainty.Acomplexprocessofmoralandphilosophicalreflectionisthereforeinternaltotheapplicationoflaw.6UngerandtheCriticalLegalStudiesMovementWesawearlierthatliberalismrequiresafirmdistinctionbetweenlawandpolitics.Po-liticaldebateanddecisionresultsintheenactmentofcertainrulesandtheconfermentofcertainentitlements.Actionwithinthescopeofsuchrulesandentitlementsshouldbeprotectedevenwhenitdoesnotservethegeneralwelfare,oradvancethepoliticalobjectivesofthecollectivity.Yetcanweapplytherulesanddecideupontheexactscopeoftheentitlementswithoutreopeningthepoliticalquestionsofsocialpolicyordis-tributivejusticethatresultedintheirenactment?ThesuggestionthatwecannotsuccessfullyseparatelegaldoctrinefrompoliticsliesattheheartoftheargumentsputforwardbytheCriticalLegalStudiesmovement(CLS),alooseassociationofleftistlawyerscentredintheUnitedStates.MuchoftheworkofCLStakestheformofdetailedstudiesofparticularareasoflegaldoctrine,withaviewtodemonstratingthedoctrine’sopen-endedness,incoherence,oritstendencytocollapsebackintobasicpoliticalissues.WhenCLSadherentsseektooffermoregeneraltheoreticalstatements,theyfrequentlyexhibitahighdegreeofconfusionandalackofclaritywhichposessevereobstaclestointerpretation.Thisprovidesuswithajustifica-tionforfocusingontheworkofoneleadingfigurewhohasexertedapowerfulinfluenceonCLSevenifhisviewsarenotaltogethertypicalofthemovementgenerally.TheworkofRobertoUngerexhibitsamuchhigherdegreeofsophisticationthanmostotherCLSwritings,andmeritsconsiderationeveninaverybriefsurveyoflegalphilosophy.InUnger’sfirstbook,KnowledgeandPolitics(1976),heidentifiedtwoproblemsthatareinternaltoliberalpoliticaltheory,whichhecalled‘theproblemoflegislation’and‘theproblemofadjudication’.Theproblemoflegislationconcernsthequestionofhowliberalpoliticaltheorycanfindabasisforgeneratingjustrulesfortheconductofsociallifegivenitscommitmenttoneutralitybetweenconceptionsofthegoodlife.Thisproblemisacentraltopicwithinpoliticalphilosophy,andweshallleaveitononesideinordertofocusonaproblemthatisspecifictothephilosophyoflaw,namely,‘theproblem423\nN.E.SIMMONDSofadjudication’.HereUngerarguedthat,evengiventheenactmentofauthoritativelegalrules,thoserulescouldnotbeappliedinspecificcaseswithoutcollapsingtheissuebackintopoliticalvaluejudgements.Hesawthisclaimasfollowingfromliberalism’srejectionof‘essentialism’,whichisunderstoodasthethesisthatwordshavestablemeaningsbyvirtueofrepresentingfundamental‘essences’.Liberalism’smoralneutrality(asUngercharacterizedit)requiredarejectionofessentialism,andyetlib-eralism’sfaithinlawcouldnotbesustainedwithoutessentialism.Ineffect,Ungerwasarguingthat,oncelegaltheoryabandonedformalism(whichistheclaimthatallcasescanberesolvedsimplybyapplyinglegalruleswithoutreferencetoanyotherconsiderations),itwouldhavenowaytoavoidcollapsingadjudicationintopolitics.Inhislaterwork,TheCriticalLegalStudiesMovement(1986),Ungeremployedtheterm‘formalism’inawidersensetorefertoalltheoriesthatasserttheexistenceofadistinctionbetweenlegaldoctrineandopen-endedideologicalcontroversy:heclaimedthat‘formalism’initsmorenarrowandconventionalsenseissimplytheextremeendpointofthesamegeneralthesis.Ungerarguedthatwhenwecometoapplylegalrules,weareforcedbackintodeeperquestionsabouttheprinciplesunderpinningtherules:weareforcedtoassumethattherulesrepresentamorallycoherentanddefensibleschemeofhumanassociation(anassumptionthatUngerterms‘objec-tivism’,andwhichwemayfindbestexemplifiedbyDworkin’stheoryoflaw).Yetthisassumption,heclaims,isunsustainable.Lawalreadyrepresentswithinitselfthecon-flictbetweenrivalsocialvisions.Foreveryprinciplethatwediscoverinthelegaldoc-trine,therefore,wewillalsofindacounter-principlethattendstosubvertit.Theremaybeawayinwhichtheprincipleandcounter-principleareconventionallybalancedagainsteachother,butthisbalancewillnotitselfbegroundedinanydeeperprinciple.Rather,itisanexpressionoftheerroneousassumptionthatsocialarrangementshavetotakemoreorlesstheformthattheypresentlydo.Oncewegrasptheessentiallyplasticandrevisablenatureofourformsofassociation,Ungerargues,wewillseethatthereisnoreasonwhythecounter-principlesshouldnotbeextendedattheexpenseoftheprinciples.This‘deviationist’formofdoctrinalargumentrevealsthehollownessoflegaldoctrine’sclaimtobeseparablefrompolitics,whilesimultaneouslyrevealingthedoc-trine’spotentialasaninstrumentofrevolutionarysocialtransformation.ItisdoubtfulwhetherUnger’sargumentissuccessful.AsJohnFinnishaspointedout,ontheassumptionthatthelawispursuingapluralityofdifferent‘goods’ormoralvalues,thereisnoreasonwhytheparticularbalancestruckbetweenthosegoodsshoulditselfberequiredbyadeeperprinciple;theremaybedifferent,equallyreason-able,waysofstrikingthebalance.Alternatively,amoreconservativeviewmightpointoutthat,ifapartoflaw’sobjectistostabilizemutualexpectationsaboutconduct,onewouldexpectthelawtoreflectthoseformsofhumanassociationthatcurrentlyexist;themerefactthattheycouldbeotherwisewould,fromthisperspective,beirrelevant.7PhilosophicalReconstructionofLegalDoctrineThejuxtapositionofDworkin’semphasisontheinternalcoherenceoflawwithUnger’sassertionofitsconflict-riddenincoherencehasspawnedanupsurgeofphilosophically424\nPHILOSOPHYOFLAWinformedstudiesofparticularareasoflegaldoctrine.Theaimhasbeentoidentifythegeneralmoraltheoriesorvaluesthatappeartounderpinmajorlegalinstitutionssuchascontract,tort,liabilityandcriminallaw.Themostwell-establishedfieldforthistypeofphilosophicalresearchhaslongbeencriminallawandthepenalsystem.Benthamofferedasweepingjustificationofpun-ishmentinutilitarianterms,backedupbycorrespondingjustificationsforsuchfeaturesofcriminallawasthementaldefences(criminalresponsibilitygenerallyrequireseitherintentionorforesightonthedefendant’spart),thedefenceofduress,andtheexclusionofliabilityfortheinsane.Suchutilitarianaccountshavebeenrelentlesslyopposedbyrivaltheories,oftendrawingtheirinspirationfromKant.Morerecently,theoristshaveturnedtheirattentiontoprivatelaw,particularlycontractandtort.Herethemajorsplithasbeenbetweenthosetheoristswhoseeprivatelawdoctrinesasbeinginextricablyconnectedtowiderconceptionsofdistributivejusticeoraggregatewelfare,andrivaltheoristswhoseektoconfinesuchdistributiveoraggregativeconcernstothetaxandwelfaresectorsofthelaw,andwhoseektoexplainprivatelawintermsofthevalueof‘correctivejustice’.Amongsttheformergroup,scholarsof‘theeconomicanalysisoflaw’havebeenprominent,generallyseekingtoexplainandjustifylegaldoctrinesasinstrumentsofefficientresourceallocationorwealthmaximization.Inevitably,thesedebatesareconnectedbothwithgeneralquestionsaboutthenatureoflawandwithwiderissuesofpoliticalphilosophy.FurtherReadingCotterrell(1989),Lyons(1984)andSimmonds(1986)provideelementaryintroductionstothesubject.Afterthat,itisprobablybesttotackleHart(1961),Fuller(1969),Finnis(1980)andDworkin(1986).Dworkin’searlierbooksalsorepaydetailedstudy.Hart(1983)containsanumberofveryimportantandimmenselylucidessays.Kelsen’sworkprovidesaverydifferentformoflegalpositivismfromthatofHart,andthebestplacetobeginiswithKelsen(1992).The1992EnglishtranslationcontainsaveryvaluableintroductionbyStanleyPaulson.JosephRazisoneofthemostimportantcontemporarywritersonthephilosophyoflaw,andhehasdevelopedasophisticateddefenceofpositivismthatisstronglyinfluencedbybothHartandKelsen,inTheAuthorityofLaw(1979),andinnumerousotherpublications.Hisworkisalsonotablefortheconnectionsthatitdevelopsbetweenphilosophyoflawandtheoriesofpracticalreason.Forvariousviewsontherelationshipbetweenlawandpracticalreason,seeRaz(1975),MacCormick(1978),BeyleveldandBrownsword(1986)andFinnis(1980).RecentdebatesonthemoralunderpinningsoflegaldoctrinesmaybeexploredbyreadingFried(1981),Coleman(1992)andHart(1968).Unger(1976)isaratherwaywardbookthatisinclinedtoresorttocrudecaricatures.Nevertheless,itcontainssomeargumentsthatareinteresting,andwhichhaveexertedapowerfulinfluenceonthecriticallegalstudiesmovement.Unger(1986)containsapowerfulcritiqueofDworkin(referredtoas‘therightsandprinciplesschool’)initsearlypages.ThismightbecomparedwiththeessaybyFinnis,inEekelaarandBell(1987).Kelman(1987)andAltman(1990)provideusefulguidestotheconfuseddebatesurroundingcriticallegalstudies.425\nN.E.SIMMONDSReferencesAltman,A.1990:CriticalLegalStudies:ALiberalCritique.Princeton,NJ:PrincetonUniversityPress.Austin,J.1955[1832]:TheProvinceofJurisprudenceDetermined.London:WeidenfeldandNicolson.Beyleveld,D.andBrownsword,R.1986:LawasaMoralJudgment.London:SweetandMaxwell.Coleman,J.1992:RisksandWrongs.Cambridge:CambridgeUniversityPress.Cotterrell,R.1989:ThePoliticsofJurisprudence.London:Butterworths.Dworkin,R.1978:TakingRightsSeriously,revdedn.London:Duckworth.——1985:AMatterofPrinciple.Cambridge,MA:HarvardUniversityPress.——1986:Law’sEmpire.London:Fontana.Eekelaar,J.andBell,J.(eds)1987:OxfordEssaysinJurisprudence,3rdseries.Oxford:ClarendonPress.Finnis,J.1980:NaturalLawandNaturalRights.Oxford:ClarendonPress.Fried,C.1981:ContractasPromise.Cambridge,MA:HarvardUniversityPress.Fuller,L.1969:TheMoralityofLaw,2ndedn.NewHaven,CT:YaleUniversityPress.Hart,H.L.A.1961:TheConceptofLaw.Oxford:ClarendonPress.——1968:PunishmentandResponsibility.Oxford:OxfordUniversityPress.——1983:EssaysinJurisprudenceandPhilosophy.Oxford:OxfordUniversityPress.Kelman,M.1987:AGuidetoCriticalLegalStudies.Cambridge,MA:HarvardUniversityPress.Kelsen,H.1992[1934]:IntroductiontotheProblemsofLegalTheory(translatedbyB.L.PaulsonandS.L.Paulson).Oxford:ClarendonPress.Lyons,D.1984:EthicsandtheRuleofLaw.Cambridge:CambridgeUniversityPress.——1993:MoralAspectsofLegalTheory.Cambridge:CambridgeUniversityPress.MacCormick,N.1978:LegalReasoningandLegalTheory.Oxford:ClarendonPress.Raz,J.1975:PracticalReasonandNorms.London:Hutchinson.——1979:TheAuthorityofLaw.Oxford:ClarendonPress.——1980:TheConceptofaLegalSystem,2ndedn.Oxford:ClarendonPress.Simmonds,N.E.1986:CentralIssuesinJurisprudence.London:SweetandMaxwell.Unger,R.M.1976:KnowledgeandPolitics.NewYork:FreePress.——1986:TheCriticalLegalStudiesMovement.Cambridge,MA:HarvardUniversityPress.DiscussionQuestions1Canasingletheoryoflawdealsatisfactorilywithallthemajorpuzzlingfeaturesoflegalthoughtandpractice?2Isitmoreimportantforphilosopherstoanalysetheconceptoflaworthesocialpracticesofwhichlawisapart?3What,ifanything,distinguishestheruleoflawfromaregimebasedsolelyonforce?4Arelegalrightsandobligationsspeciesofmoralrightsandobligations?5Canareductionistprogrammedealwiththeapparentlyprescriptivelanguageoflaw?6DoesKelsen’sbasicnormallowlawtobeprescriptivewithoutbeingabsorbedintomoralquestionsaboutjustice?7Howminimalisthe‘minimumelementofnaturallaw’requiredforapositivistaccountoflaw?426\nPHILOSOPHYOFLAW8ShouldweacceptHart’sclaimthatanadequatepositivistaccountoflegalobli-gationcanbeprovidedbytakingaccountofthe‘internal’viewpointofaparticipantofalegalsystem?9Whatimportancedoesthemodernnotionofautonomyhaveforthetheoryoflaw?10Discusstheclaimthatrulesarelawsiftheyemanatefromanappropriatesource,whetherornottheyarejust.11Mustwechoosebetweenthe‘formalist’positionthatallconceivablecasescanbedecidedbypre-existingrulesoflawandthe‘rule-sceptical’positionthatalllegalcasesaredecidedbyextra-legalfactors?12Dojudgeshavethetaskofconstructingacoherentmoraltheory?Candifferentjudgesconstructdifferenttheories?13Whatistheimportanceofaruleofrecognitioninthetheoryoflaw?14Doeslegalpositivismprovideamiddlewaybetweensupineacceptanceofthelawandirresponsibleanarchism?15Shouldphilosophyoflawbeamatterofreflectivelyunderstandinglawinapar-ticularsocietyratherthanthedeterminationofauniversaltheory?16IsitafaultofDworkin’sinterpretiveandculturallyspecifictheorythatiterodesthedistinctionbetweenphilosophyandsociology?17Shouldwevaluemoralcoherenceorintegrityinlawevenifwesubscribetoadif-ferentmoralview?18HowdoesFuller’s‘fidelitytolaw’differfromsubserviencetoauthority?19Whataretheconsequencesifwecannotseparatelegaldoctrinefromopen-endedideologicalcontroversy?20Doesthepossibilityofstrikingdifferentbalancesamongthepluralityofmoralvaluespursuedbythelawunderminetheauthorityoftheactualbalancestruckbythelaw?427\n14PhilosophyofHistoryLEONPOMPAHistoriansareconcernedwiththehumanpast,withwhathappened,andhowandwhyithappened.Butthepastisadifficultobjectofstudy.Itisbeyondthereachofpercep-tionandrequiresspecialmethodsofenquirywhichmayleadtoresultsthatfailtosatisfythecriteriaofknowledgethatperceptionsatisfies.Itscontentsarealmostlimitlessandcanbestudiedingreaterandlesserdegreesofdetail,requiringdecisionsofselectionwhichmayaffecthistorians’claimstotruth.Itcontainsthedeedsandpracticesofpastindividualsandcommunitieswhichmaybesufficientlydifferentfromourownthatwecannotbesurethatwecanreachanytrueexplanationorunderstandingofthem.Finally,historiansmustexpresstheirresultsinformsthatmayalsoaffecttheepiste-mologicalstatusoftheirclaims.Modernphilosophyofhistoryhasengagedwithallofthesedifficulties.Anunderlyingconcernhasbeentoestablishhowfarthecapacityofhistorianstoreachobjectivetruthisaffectedbythedifficultieswhichariseineachoftheseareasofenquiry.Thiswillbethecentralquestioninthischapter,startingwithwhathistoriansthemselvesconsidertheleastphilosophicallypuzzlingaspectsoftheiractivityandproceedingtothosethattheyconsidermoreproblematic.1KnowledgeofHistoricalFactHistoriansnormallyadoptarealistviewofknowledge,involvingthreeclaims:thatindividualevents,actionsandoccurrencesreallytookplaceinthepast;thattruehis-toricalstatementsorhistoricalfactsarestatementsaboutsomeoftheseoccurrences;andthatthesestatementsareknownbecausesufficientevidenceforthemisavailableinthepresent.Thus,whenweknowthatNapoleonlosttheBattleofWaterloo,weknowthatacertainevent,Napoleon’slosingthebattle,tookplaceinthepast.Weknowthisbecauseweknowthatthestatementthatitdidtakeplaceistrue,andweknowthatthestatementistruebecausethereissufficientevidenceinthepresenttojustifyourclaimtoknowit.Butthisthreefoldconceptionisdeceptivelysimpleandhasbeenchallengedateverystage.Ishallconsidereachoftheseinturn.Itisclearthatthiswholeviewwouldbeatriskifwehadreasontodoubtthattherewasarealpastatall.SuchareasonwasfirstofferedbyBertrandRUSSELL(chapter37),\nPHILOSOPHYOFHISTORYwhopointedoutthatitislogicallypossiblefortheworldtohavecomeintoexistencefiveminutesago,withapopulationhavingacompleterangeof‘memories’ofappar-entlypriortimes(Russell1921:159–60).Butiftheworldcameintoexistencefiveminutesago,notonlywouldmostofourmemoriesbefalsebutsowouldalmostallofourhistoricalknowledge.Itmightbethoughtthatwecanavoidthisconclusionbyappealing,ashistoriansnormallydo,totheevidencepresentlyavailabletous.Butthiswillnothelpfor,accordingtothehypothesis,everythingwouldnowbeasitis,includ-ingallthathistoriansthinkofasevidenceforthepast.Hence,weretherenopastbeforefiveminutesago,nothingthatwetakeasevidenceforthatpastcouldreallybeevidenceforit.Russelltookhishypothesisseriouslyenoughtosuggestthatoursupposed‘knowl-edge’ofthepastshouldbetranslatedintostatementsaboutthepresentcontentsofmindbecausehebelievedthat,throughperception,wearedirectlyacquaintedwiththerealitythatmakesstatementsaboutthepresenttrueinawayinwhichwecannotbewithregardtothepast.Thehypothesishasbeensubjecttoanumberofobjections(Danto1965:83–4),butIshallconcentrateononewhichdependsuponimportantfeaturesofthenatureofhistoricalevidence.Accordingtothisobjection,itdoesnotfollowfromthelogicalpossibilityofthehypothesiseitherthattheworldbeganfiveminutesagoorthatwecannotknowwhetherornotitdid,because‘presentevidencewouldnotbewhatitisifnopasthadexisted’(Gorman1982:72).Tounderstandtheforceofthisobjectionitisnecessarytonotetwopoints.Firstly,itisanimportantfeatureofhistoricallanguagethatitcontainswhatarecalled‘past-entailing’or‘past-referring’terms.Theseareterms‘whosecorrectapplicationtosomepresentobjectoreventlogicallyinvolvesareferencetosomeearlierobjectorevent’.Thustodescribesomebodyashavingascarentailsthatatsomeearliertimeheorshesufferedawound(Danto1965:71–2).Secondly,theuseofsomeartefactasapieceofhistoricalevidencerequiresuseofsuchpast-entailingterms.ThuswhenanhistoriandescribessomedocumentasaletterwrittenbyGladstone,heorsheismakingapast-entailingassertion.Forheorsheisapplyingtoadocumentwhichexistsinthepresentadescriptionwhichentailsthatsomeeventhappenedinthepast,namelythatGladstonewrotethisletter(Gorman1982:56–8).Thus,giventhepast-entailingnatureofthedescriptionsunderwhichdocumentscountashistoricalevidence,therefutationtakestheformofassertingthatwecannotmaintainourpresentpast-entailingconceptionofhistoricalevidenceandadmitthatthehypothesiscouldbetrue.Itmightseem,however,thatthisfailstorefuteRussell’shypothesis,sinceitispartofthelatterthat,eveniftheworldcameintoexistencefiveminutesago,everythinginitwouldremainthesameforus,includingthereforeourpresentconceptionofhistor-icalevidence.Sowewouldcontinuetodescribepresentartefactsinpast-entailingterms,eventhoughtheevents‘entailed’nevertookplace.Infact,however,therefuta-tionissubstantiallycorrect.ForRussellmustallowfortherealityofsomeperiodoftime–letussayhis‘fiveminutes’–beforewhichthepastmaynothaveexisted,inordertoallowustounderstandthedoubtexpressedinthehypothesis.Thismeansthathemustallowthatsomepast-entailingdescriptions,namelythosethatrelatesomethingnowpresenttosomethinginthepastfiveminutes,aretrue.Butifthisisso,hemustshowhowitispossiblethatsomepast-entailingtermsreallyhavetheimplicationsforthepastthatwetakethemtohavewhileothers,whichareofpreciselythesamekindandsatisfy429\nLEONPOMPApreciselythesamecriteria,donot.Butthiscannotbeshown.Fortheconsistentuseofcriteriaisnecessarynotonlyforhistoricalthoughtbutforanycoherentthoughtwhat-soever.Hencethescepticalchallengecouldbemaintainednotasachallengeabouttherealityofsomepartofthepastbutonlyasachallengeabouttherealityofthewholeofthepast.Butthiswouldbeself-defeatingsincethecontrastbetweenarealpast(the‘fiveminutes’)andanunrealpast,whichisneededtomakethehypothesiscoherent,wouldthenbelost.Sofar,therefore,thereisnoreasontodoubtthattherewasarealpast.Butrealisminvolvestheviewthatthepasttowhichhistoriansarecommittedisalsoanindepen-dentpast;inotherwords,itisapastfilledwitheventsthatoccurredirrespectiveofwhetherornotweshallevercometoknowofthem.Historiansconceiveoftheirtaskasdiscoveringthecontentsofthispastbutnotofcreatingit.Thisviewhas,however,beenchallengedfromanidealistpointofview,viaatheorynowgenerallyknownas‘con-structionism’(Nowell-Smith1977).Theconstructionistthesisacceptsthathistorianscometoknowtruthsaboutthepast.Butthepasttowhichtheserelateisnottherealpast,aswenormallyunderstandit–thatis,asanindependentpastofrealeventswhichawaitdiscovery.Itisseenasapeculiarly‘historical’pastthatexistsassomethingthatweconstructfrompresentevidence.Accordingly,werewetocomeintopossessionofnewevidenceandconstructitdifferently,notonlywouldourknowledgealterbutsoalsowouldthehistoricalpastitself.Butifthisisso,thestandardhistoricalbeliefinrealism,andtheconceptionofknowledgewhichpresupposesit,wouldbefalse.Theprincipalreasongivenforthisclaimisthatsincetherealpastnolongerexists,itcanhavenoparttoplayintheverificationofthehistorian’saccount.Accordingly,itisclaimed,thepastthatthehistorianbringsintoviewisnotthepastasitmighthavebeeninitselfbutonethatexistsrelativeonlytoitscapacitytoexplainourbeliefsabouttheevidence.Theideathathistoriansdiscoverapastthatexistedindependentlyofourknowledgeofitisthereforerejectedinfavouroftheideaofapastthatexistsonlyasaconstructiontoexplainpresentevidence(Goldstein1976,1977).Itshouldbenoted,althoughitcannotbepursuedhere,thatthisviewhasaffinitieswithsomeformsofscientificANTI-REALISM(pp.294–5).ConstructionisminvolvesthesameassumptionasRussellmade:thatbecausehis-toricalstatementsareaboutapastwhichwecannolongerperceive,thereissomethingproblematicaboutitsontologicalstatus.Butthisisaverydoubtfulassumption.Mostkindsofknowledge,botheverydayandscientific,areaboutwhatiscurrentlyunob-servable.Ourmemories,forexample,areaboutwhatwecannotnowperceive.So,too,areourbeliefsaboutthespeedoflightortheexistenceofneutrons.Wereitthecase,therefore,thatastatementaboutsomethingthatisunobservablecannotbeaboutanindependentreality,therangeofstatementsaboutwhatwetaketoberealentitieswouldneedtobereplacedbyarangeofstatementsaboutconstructedentities.Totakeasimplecase,oneperson’sreasonforclaimingtoknowthecolourofQueenElizabethII’scoronationdressmaybethatheremembershavingseenitonCoronationDay,whileanother’smaybethatshehasconsultedtheorderbooksoftheroyaldressmakers–shehas,thatis,undertakensomesimplehistoricalresearch.Iftheconstructionistaccountiscorrect,thetwoclaimscannotbeaboutthesamedress,foreachwillbeaboutadressthatexistsonlyinrelationtothemodeinwhichourknowledgeofitiscon-structed.Thustherewillbetwodresses,oneexistingaspartofan‘historical’pastand430\nPHILOSOPHYOFHISTORYtheotheraspartofa‘memorial’past,ratherthan,aswewouldnormallythink,onedresswhichwecanknowaboutineitheroftwodifferentways.Thisconclusion,more-over,willapplytotheconstituentsofallofourdifferentkindsofknowledgethatgobeyondsimplereportsofwhatispresenttoperception.ThusthebookinthepastthatIrememberwilldiffernotonlyfromthatwhosehistoryIreconstructbutfromthatwhosepastphysicalorchemicalpropertiescanbeanalysedintermsoftheoriespropertochemistryorphysics.Butinthiscaseweshallbeunabletouseourknowledge,say,ofthechemicalpropertiesofsomepastobjecttohelpusinourhistoricalthinkingaboutthatobject,perhapsbyhelpingustodateit,sincetheobjectwhosepastchemicalpropertiesweconstructwillbeadifferentobjectfromthatwhosehistoricalexistenceweconstruct.Thustherewouldbeanumberofdifferentobjects‘existing’inanumberofdifferentrealities,eachrelativetoaparticularmodeofresearch,ratherthan,astherealistbelieves,objectsexistinginoneandthesamereality,whosedifferentpropertiescanbeknownbydifferentkindsofresearch.Thisthesiscertainlyconflictswithoureverydayviewoftherealworld.Butthisisperhapsnottooimportantsince,whilephilosophyshouldnotignoreeverydaybelief,itcannotbewhollyconstrainedbyit.Moreimportantly,itconflictswiththerequirementthatwemustthinkofallknowledgeasbeingaboutonesinglerealitybecause,unlesswedoso,weshallbeunabletorelatetheproductsofourdifferentdisciplinestooneanother.Constructionism,andthestrongverificationismuponwhichitdepends,mustberejectedbecauseitleadstoresultsthatconflictwiththisrequirement.Thesepointsdonot,however,ruleouthistoricalknowledgebeingreachedbytheuseofargumentinthewayinwhichtheconstructionistssuggest.Theyshowonlythatthereisnoreasontobelievethattheargumentsresultinstatementsaboutsomepecu-liarly‘historical’pastratherthanaboutarealpastwhichwecanknowinavarietyofdifferentways.Howtheydothishasbeenshowninanaccountthatemphasizestheneedforhistoricalevidencetobedescribedinpast-entailingtermsinthewayindicatedabove(Gorman1982:56–68).Whenthehistorianhascarriedouthisresearchcor-rectlyandcheckedforauthenticity,hewillbeinpossessionofanevidentialstatementsuchas‘Here,beforeus...isaletterwrittenbyGladstone’,thetruthofwhichbothentailsandrequiresthatGladstonewrotetheletter.Thusthereisarelationshipofmutualentailmentbetweenpast-referringevidentialstatementsandhistoricalstate-ments,invirtueofwhichwecantrulyassertthelatteronthestrengthoftheformer.Butwhilethisshowshowevidentialstatements,iftrue,entailtruehistoricalstate-mentsaboutarealpast,itleavesroomforathirdandfinalscepticaldoubt.ForinordertoarriveatanevidentialstatementsuchasthataboutGladstone’sletter,weneedbothtoemployatheoryofinterpretationandtobesatisfiedthatitisthecorrectone.BecausetheexamplegivenisinEnglish,thisisapointthatisoftenoverlookedbyEnglishspeak-ers,wherewejustassumethatweareusingthecorrectlanguage.Butthefactthatatheoryofinterpretationisemployedbecomesevidentifweconsiderthecasewheretheevidentialstatementisinsomelanguagewhichwedonotunderstand.Thenweneedtoturntodictionariesandgrammarbookstoenableustotranslatethelanguageofthedocument.This,ofcourse,israrelydifficult,althoughitcanbecomesoinpracticewhenhistoriansaredealingwithdocumentsinoldandrelativelyunknownlanguages.Butitillustratesthefactthatatheoryofinterpretationisalwaysinuse,evenifnotexplicitly.431\nLEONPOMPAThescepticaldoubtthatnowarises,however,concernsthejustificationofthetheoryofinterpretationthatisemployed.ForrecentworkinthePHILOSOPHYOFLANGUAGE(chapter3)hasshownthatthereisalwaysthepossibilityofmoretheoriesofinterpre-tationthananywhichweemploy.Itfollowsthatwecannot,especiallywhendealingwiththepast,simplyrestcontentwiththebeliefthatweareemployingthecorrecttheoryofinterpretation,theoneenablingustojustifytheclaimthatweareinposses-sionoftruepast-entailingstatements.Werequiresomereasontosupportthisbelief.Thus,toreverttotheexampleofGladstone’sletter,itisafactthatwehaveatheoryofinterpretationunderwhichtheword‘Gladstone’istakentobethenameofaparticu-larhistoricalindividual.Moreover,ouruseof‘Gladstone’islinkedtotheuseofmanyothersimilarlyfunctioningnames,suchas‘Disraeli’and‘QueenVictoria’.Ifwearewronginourinterpretationoftheword‘Gladstone’,weshallbewrongalsoaboutalltheseconnectednames.Thedoubtisnotthereforesimplyaboutourrighttointerpretasinglenameinacertainwaybutaboutourrighttorelyuponthewholetheorygov-erningtheinterpretationofthatandmanyothernames.Thisdoubtaboutjustifyinganytheoryofinterpretationweemployprovidesathreatnotonlytoourknowledgeofthepastbuttoaconsiderableamountofourknowledgeofthepresent,sinceouruseofpast-entailingtermsmeansthatmuchofwhatwetaketobeknowledgeofthepresentinvolvestruthsaboutthepast.Wethinkofourselves,forexample,ashavingpersonalandcommunalhistoriesthatinterlinkininnumerableways.Indeed,suchbeliefscompriseanimportantcomponentinourconceptionofper-sonalidentity.Thusifitwerenottrue,say,thattherewasa1914–18war,allthatpartofpeople’spersonalhistorieswhichinvolvereferencetothewarwouldbefalse.Insofar,therefore,asmuchknowledgeofthepresentinvolvesclaimsaboutthepast,thescepticalthreattohistoricalknowledgealsothreatensmuchofourknowledgeofthepresent.Onepossibleresponsetothisdoubtistopointtothefactthatourpresentprocedureleadstocoherenceinourclaimsaboutthepast.ClaimsaboutGladstonearelinkedtoclaimsaboutDisraeli,theIrishproblemandsoon.Butanappealtocoherencealoneisinadequatetojustifyatheoryofinterpretation.Fortherearemanyotherpossibletheoriesofinterpretationwhichcouldleadtoequallycoherentsetsofclaimswhichareincompatiblewiththosethatwenowaccept.AFrameworkofBasicHistoricalFactsAnotherwayofcounteringthedoubtconcerningjustificationforourtheoryofinter-pretationwouldbebysupportingthetheorybylinkingittosomeknowledgeofthepastreachedbyadifferentroute.Ifitcouldbeshown,forexample,that,priortohisexami-nationoftheevidence,thehistorianalreadyknowsthatGladstoneexisted,wecouldusethatknowledgetojustifyhisorheruseofatheoryofinterpretationinwhichtheword‘Gladstone’isthenameofthathistoricalindividual.Theproblemthenwouldbenothowtodefendthetheoryofinterpretationthatallowsthehistoriantoconcludethatanindi-vidualcalledGladstoneexistedandwrotetheletterinfrontofhimorher,butthemuchsimplerproblemofdecidingwhetherthehistoriancanshowthattheletterwaswrittenbytheGladstoneofwhoseexistenceheorshealreadyknowsorwhetheritwaswritten,perhapsforged,bysomeotherperson.432\nPHILOSOPHYOFHISTORYButhowisitpossibletosupportthisrefutationwhich,atthemoment,amountstolittlemorethanassertingwhatthescepticisquestioning:thatweknowsomethingsatleastaboutwhathappenedinthepast?Whatweneedhereisanaccountofhow,inde-pendentlyofargumentfromevidenceunderatheoryofinterpretation,wecanhavesomesuchknowledge.Sinceothershavenotwrittenonthissubject,Imustherefallbackuponmyownsuggestionthatsomeofourknowledgeofthepasthasactuallycomefromthepast–inotherwords,thatitistransmittedwithinsocietieswithappropriateinstitutionalmechanisms,fromthepasttothepresent,andthat,invirtueofexistinginsuchasociety,thehistorianisinpossessionofknowledgeofaframeworkofbasicfactswithwhichtosupporthisorhertheoryofinterpretation(Pompa1993).Thus,insteadofarguingfromevidenceunderatheoryofinterpretationto,say,Gladstone’sexistence,thehistoriantakeshisorherinheritedknowledgeofGladstone’sexistenceasjustificationforatheoryofinterpretationinwhichthatexistenceispresupposed.Onthisview,anyattempttointerprettheevidenceinsuchawayastothrowdoubtuponGladstone’sexistencewouldbedisallowedandanytheoryofinterpretationthatseemstoallowsuchdoubtwouldberejectedasunjustified.Thesuggestion,therefore,isthatourtheoryofinterpretationisjustifiedbecauseitpresupposessomehistoricallyacquiredknowledgeofthepast.Thehistorianisthenentitledtouseittodiscovernewfactsaboutthepastthatarenotpartofourinheritedknowledge.Itshouldbenoted,however,thatthissuggestionconflictswiththewidelyheldFALLIBILIST(pp.39–40)theoryofknowledge,accordingtowhichnobeliefisinprincipleimmunefromthepossibilityofrevision,includingthereforethemostcentralofourpresenthistoricalbeliefs,suchasthattherewasaCrimeanWar.ButthereplytothefallibilistmustbethatalthoughitisaLOGICALPOSSIBILITY(pp.151–6)thattheCrimeanWarneveroccurred,itisnotanepistemologicalpossibility.Forwecanallowthatitisanepistemologicalpossibilityonlyifwecanenvisagesomewayinwhichsomefuturehistoriancouldtrytojustifytheclaimthatthewarneveroccurredwithoutacceptingthesortoftheoryjustoutlined.Butifheisunabletodothis,hewillbeopentoscepticaldoubtabouthisorhertheoryofinterpretationandthereforeabouthisorherclaimstoknowledgeaboutthepast.2ExplanationandUnderstandingThediscussionsofarlendssupporttothebeliefthathistorianscanproduceknowledgeofwhatactuallyhappenedinthepast.Butalargepartoftheirworkisconcernedwithexplainingwhathappened.Theanalysisoftheproperwayinwhichtodothishasbeenthesourceofamajorcontroversyinvolvingtwoopposingviewpoints:thattheyshouldoffer‘scientific’explanationsorthattheyshouldfocusuponahumanisticformofunderstanding.TheviewthatexplanationsinhistoryeitherareorshouldbeofthesameformasthoseinthenaturalscienceswasadvancedinaparticularlystrongwaybyCarlHempel(b.1905)who,buildingupontheworkofKarlPopper(1902–94)inthePHILOSOPHYOFSCIENCE(chapter9),advancedageneralthesis(Hempel1959)aboutwhatitisforsomethingtobeanexplanation.Thethesis,subsequentlyknownastheCOVERINGLAWMODEL(pp.304–5)orthedeductive–nomologicaltheory,involvestwopoints:thattoexplainaneventistoshowittobepredictable;andthattobeshowntobepredictable,433\nLEONPOMPAtheeventmustbeshowntobesubsumableundersomesetofcausallaws.Thus,althoughhistoriansnormallyexplaintheoccurrenceofparticularevents,suchasthe1914–18war,intermsofacombinationofparticularcauses–suchasthedevelop-mentofnationalrivalriesforpoliticalsupremacyinEuropeinthenineteenthcenturyorthecommonneedfornewsourcesofrawmaterialsandnewmarketstomaintaintheindustrializationofWesternEurope–thiscanbeanexplanationonlyifthepar-ticularcausescitedareinstancesofacausallaworlawstotheeffectthatwheneverthereisabuildupofnationalrivalriesforpoliticalsupremacyandacommonneedfornewsourcesofrawmaterialsandnewmarkets,awarbreaksout.Hempelrealizedthatthesortsofgeneralizationsmentionedinthisexamplewerenotlawsinthesensethattheyheldwithoutexceptionand,indeed,thatproperlysupportedcausallawswererarely,ifever,usedbyhistorians.Forthisreasonhereferredtotheexplanationsthathistoriansactuallygiveas‘explanationsketches’,claimingthattheyofferedmoreorlessvagueindicationsoftherelevantlawsandpointedthewayforthefurtherempiri-calresearchnecessarytoestablishthemasrequiredinafullysupportedexplanation.Onthisaccount,therefore,historicalexplanationsareincompleteexplanationsofthesameformasthosetobefoundinthephysicalsciences.Astrengthofthecoveringlawmodelisthatitsatisfiesonefeaturethatonewouldexpectofanyexplanation,namely,thatitshouldexplainwhyoneeventoccurredratherthansomeother(Donagan1966).Itdoesthisbecause,bysubsuminganeventundercausallaws,itshowsthattheeventhadtooccur.Thismeans,however,thatthekindofexplanationitoffersisdeterministicandthis,asweshallsee,hasbeenthemainunderlyingreasonforthestrongestobjectionstoit.Objectionstothetheorydivideintotwokinds.Firsttherearethosethatemphasizethathistoricalexplanationsdonot,asamatteroffact,conformtoit.Therehavebeenmanyobjectionsofthissort(White1965:14–104,offersareformulationofthetheory;seealsoDonagan1966:142–6),butthesecannotbeconclusive.Ifthetheoryclaimsthatthisistheformthatanyfullysupportedexplanationmusttake,objectionsthathistoricalexplanationsdonottakethisformcanbemetbythereplythathisto-riansarefailingtoofferfullysupportedexplanations.Forthisreason,Ishallconcen-trateonobjectionsofthesecondkind,whichclaimthatthesubjectmatterofhistoryissuchthatthetheoryiswronginprinciple.Thefirstoftheseobjections,whichhasweighedparticularlystronglywithhisto-rians,isthatthetheoryfailstorecognizetheuniquenessoftheeventsandindividualsfeaturinginhistory(Oakeshott1933:154).Therewas,andcanbe,onlyone1914–18war.Hence,ifhistorianssucceedinexplainingwhatcausedthisuniqueevent,itcannotbebyconformingtotherequirementsofthecoveringlawmodelsincethelawsinques-tionholdbetweendifferentkindsofeventsratherthanuniqueevents.Buttheclaimtouniquenessmustbetreatedwithcare,formuchdependsuponthesortofuniquenessinquestion.Thecoveringlawmodeldoesnotdeny,forexample,thattheeventsthatitofferstoexplain,suchastheoccurrenceofthe1914–18war,areunique.Itmustinsist,ofcourse,thattheseevents,althoughunique,aredescribableingeneralterms,inordertobesubsumableunderlawsaboutkindsofthings.Butitcanhardlybedeniedthatalthougheventsareuniqueinbeingtheparticulareventsthattheyare,theymustalsobedescribableingeneralterms,forthisisnecessaryinorderforustobeabletosayany-thingatallaboutthem.The1914–18warwasauniqueevent,buttotalkaboutitwe434\nPHILOSOPHYOFHISTORYmustbringitunderthegeneralterm‘war’,atermthatcanbeappliedtomanyotheruniqueevents.Hencethemereappealtotheuniquenessofhistoricaleventsisaninsuf-ficientobjectiontothecoveringlawmodel.Whathistorianssometimesmean,however,whentheyappealtouniqueness,isthewayinwhichtheoccurrenceofeventsdependsupontheactivitiesofuniqueindivid-ualsaboutwhosebehaviourtherecanbenolaws.Iftherecanbenolawsaboutthebehav-iourofindividuals,theexplanationofanyeventsthattheirbehaviouraffectscannotinvolvesubsumptionundergenerallaws.Nevertheless,whilesomecoveringlawtheo-ristshaveacceptedthatdifferentindividualsareunique,theyhaverejectedtheclaimthatthisprecludestheapplicationofthemodel,onthegroundsthattherecanbelawsaboutuniqueindividuals.Wecanexplainwhataparticularindividualdoes,forexample,byassertingthatheorshealwaysactslikethat,whichinvolvesreferringtohisorherdispositions(White1965:47–53).Thereis,however,animportantdiffer-encebetweencausallawsandthedispositionsofindividualstoactincertainways.Forcausallawsnecessitateeventsandhencewillsupportdeductivepredictionsaboutthem.Thedispositionsofindividuals,ontheotherhand,aresimplypatternsofbehaviourthatarecharacteristicofthem;explanationsandpredictionsbasedonknowledgeofthemlacktherigorouscharacterthatthecoveringlawmodelrequires.Onecanexplainsomebody’sactinginacertainwaybyreferencetohisorhergeneraltendencytodoso,butthatdoesnotmeanthatheorshemaynotuponoccasionactinanentirelydif-ferentwayinthesamesortofsituation:afactthatweacknowledgewhenwesaythatwhatheorshedidwasoutofcharacter.Wecannotbothadmitthispossibilityandmaintainthatexplanationintermsofdispositionscarriesthesameimplicationsoflaw-governedpredictabilityasisrequiredbythecoveringlawmodel.Theissueofindividualityisgivenincreasedimportancebythefactthathistoriansarenotusuallyconcernedwiththeindividualassuchbutwiththeindividualinparticularhistoricalroles,suchasthatoftheKingorQueenofEngland.WecanbeinterestedinElizabethTudorastheindividualshewas,butthisislargelytounderstandherbehaviourasQueenElizabethIofEngland,whichinvolvesheractinginacertainroleorsetofroles,andhow,astheuniqueindividualshewas,sheaffectedthecharacteroftherolesthatsheoccupied.ThisraisestheissueofROLETHEORY(p.388).Veryroughly,wemaynotethattheoccupancyofaroleinvolvesthepossessionofpowersandresponsibilitiesthatrangefromthosethatarefairlyspecific,evenlegallydefined,toothersthatarerelativelyindeterminate–andeveninareasoflegaldefini-tionthereisconsiderableflexibilityinthewaysinwhichtheindividualmayproperlybehaveintherole.Asaresult,thebehaviouroftheindividualswhooccupytherolescanhaveasmucheffectonthenatureoftherolesastherequirementsoftheirroleupontheirbehaviour.Tracingthisinterplayisamongthemostimportantwaysinwhichhistoriansseektoexplainthehistoricaldevelopmentofroles.Thus,therehasbeenongoingdiscussionastohowfarchangesintherespectivepowersofParliamentandMonarchyinsixteenth-centuryEnglandwereduetoElizabeth’sconductinherrole.Thisdoesnotmeanthatthesechangesmaynothavebeenaffectedalsobyreligious,socialandeconomicchangesbut,sincethesefurtherchangesalsoinvolvereferencetoindividualsintheirroles,therelationshipbetweenindividualandrolestillremainsatthecentreofthediscussion.Givenitsflexibility,itseemsthattherecouldbenolawsabletoexplaineitherthisrelationshiporthehistoricaldevelopmentofrolestowhich435\nLEONPOMPAithascontributed(Berlin1966:23–4).Thusthecoveringlawmodelisinprincipleinap-plicablehere.Thesedifficultiessuggestthatthefailureofthecoveringlawmodelisduetoanincor-rectassimilationofexplanationinhistorytoexplanationinthenaturalsciences.Analternativeviewpoint,whichstressestheprimacyofunderstandingoverexplanation,hasbuiltupontheworkofR.G.Collingwood(1889–1943),whodistinguishedbetweenthesubjectmatterofscience,whichisthephysicalworld,andthatofhistory,whichistheworldofhumanactivity.Accordingly,heargued,historicalexplanationsdifferfromscientificexplanationbecausehumanactivityisalwaysanexpressionofthought,whichitisthehistorian’stasktounderstand.‘Whenanhistorianasks“WhydidBrutusstabCaesar?”hemeans“WhatdidBrutusthink,whichmadehimdecidetostabCaesar?”’(Collingwood1993:213–15).Thusweexplaintheagent’sactionbyestablishinghisorherreasonfordecidingtodowhatheorshedid.Themistakeofthinkingthathistoricalexplanationsarethesameinkindasthoseinthenaturalsciencesstemsfromafailuretorecognizethisdistinctioninsubjectmatter.RationalReconstructionCollingwood’sclaimshavebeenveryinfluentialinthedevelopmentofanalternativemodelofhistoricalexplanation,involvingwhatisoftencalledaRATIONALRECONSTRUC-TION(p.386).Thecentralideahereisthatthehistorian’sprincipaltaskistounderstandhumanACTION(pp.384–8)inaspecificallyhumanway,whichheorshedoesbyshowingsomeeventtobeanINTENTIONALACT(p.386)undertakeninaccordancewithbeliefsthatrendereditsensibletotheagentinquestion(Dray1957,1980,1993;Martin1977).Theactionisexplainedwhenweunderstandwhytheagentdecidedtoactinthatway.This,however,isonlyapartoftheexplanationsince,asCollingwoodinsisted,historiansareinterestednotonlyinwhypeopleactedincertainwaysbutalsointhesuccessorfailureoftheiractions.Hence,oncetheagent’sreasonsandbeliefsareestablished,theymustundergofurthercriticalexamination,sincethesuccessorfailureofactionsoftendependsuponthemeritsordefectsofthebeliefsandreasonsbehindthem.Thus,itmaybethatalthoughtheagent’sactionsmadegoodsensefromhisorherpointofview,someoftherelevantbeliefswerefalse,aswerethebeliefstheRomansheldaboutHannibal’spositionattheBattleofTrasimeno.Once,however,thehistorianhasestablishedwhattheRomansbelievedandthatthosebeliefswerefalse,heorshecanusethisknowledgetoexplaintheirdefeat.Ifhistoriansattempttounderstandactionsintermsofrationalreconstruction,theexplanationwillbetrueonlyiftheycanestablishwhattheagent’sreasonforactingreallywas.Ithasbeenargued,however,thattodothistheymustestablishtheaddi-tionalfactsthattheagentwasrationalandthat,intheagent’skindofsituation,ratio-nalagentsinvariablydothekindofthingwhichheorshedid.Theneedforacoveringlawisthusre-introduced,thistimerelatingthebeliefsofrationalagentsincertainkindsofsituationstocertainkindsofactions(Hempel1974:98–105).Thisisausefulobjec-tiontoconsider,forithighlightsthedifferencebetweenthedeterministicassumptionsofthecoveringlawmodelandthevoluntaristassumptionsoftherationalreconstruc-tionmodel.Theobjectionassumesthatanappealtoanagent’srationalitywillsuffice436\nPHILOSOPHYOFHISTORYonlyifhisorherbehaviourwasaninstanceofsomesortofcausallawgoverningthebehaviourofrationalagents.Butthisiswhattherationalreconstructionmodeldenies.Accordingtothelatter,whenwesaythatanagentwasrationalwearesayingthathisorheractionwascausedbyadecisionbaseduponanappraisalofhisorhersituationundertakeninthelightof,butnotcausedby,hisorherknowledgeofcertainstandardsofappropriateness.Whenweactwemusthavesomeideaofwhatitisappropriatetodotoachieveourends.Butstandardsarenotlaws.Theyaresimplybeliefsaboutthesortsofthingsoneoughttodo,eitherfortheirownsake,asinthecaseofmoralstan-dards,orinordertoachievecertainends,andtherecanbemanydifferentbeliefsofbothkinds.Thefactthatwemakeuseofstandardswhenconsideringwhattododoesnotnecessitatethatsomeparticularactionbedoneonaparticularoccasion–hencethefactthatanagent’sactionisaconsequenceofadecisiontakeninthelightofstandardsdoesnotmeanthatitwascausedornecessitatedbythem.Thecrucialweaknessinthecoveringlawmodeltowhichthispointsisthusnottheassumptionthatthehistoriancanexplaintheoccurrenceofsomethingonlybyshowingwhythatactionoccurredratherthansomeotheraction;rather,theweaknessisthefurtherassumptionthatheorshecandothisonlybyshowingthatthatactionhadtooccur.Butthisisarequirementthatneithercannorneedstobemet.Toshowwhyoneactionoccurredratherthananother,itissufficienttoshowwhythatactionratherthantheotherinfactoccurredandtodothisthehistorianneedonlyshowwhy,fromamongthehistoricallyavailableoptions(theoptionsavailabletoagivenindividualinaspecifichistoricalsociety),onecourseofactionseemedpreferabletotheothers.Theimportanceofstandardsinthisconception,however,givesrisetoaproblemwhichdoesnotbesetthecoveringlawtheorist.Forstandardsvaryindifferenthistori-calsocieties.TheGreeks,forexample,consultedtheoraclesforadviceincertainsitua-tions.Becausewedonotourselvesnormallyconsultoracles(althoughsomepeopledo),wemaywonderwhethertheGreekstandardofrationalitywasthesameasoursand,ifitwasnot,whetherwecanunderstandrationalreconstructionsofGreekactionsinthewayinwhichwecanunderstandthoseofourcontemporaries.Thisdifficultyhasbeenexacerbated,moreover,bytheclaimthatstandardsarepeculiarto,andpartof,thelivedexperienceofspecificsocieties.Weunderstandthestandardsofourownsocietybecausetheyareruleswhichwearetaughthowtofollowbylivinginoursociety.Butsincewecannotsharethelivedexperienceofpastsocietiesandbetaughttheirrulesofconduct,wecanhavenowayofunderstandingthem.Evenifwemanagetomakestatementsaboutthem,theywillnotcomeoutasstandardstherationalityofwhichwecangrasp.Thisposition,sometimescalledhistoricism,relativizesthenotionofratio-nalitytoparticularhistoricalsocietiesandindoingsochallengesourrighttoclaimtounderstandwhypeopleinthosesocietiesactedastheydid.Onemighttrytocounterthisdifficultybypointingoutthat,despiteit,historiansdosucceedinmakingtheactivitiesofpastsocietiesintelligibletous.Butthismightseemtobegthequestion,whichconcernshowfarwereallyunderstandthekindofratio-nalityindispute.Amoreeffectiveresponseistoquestionwhetherhistoricalsocietieshavetheself-containedcharactersuggested.Fortotalkofanhistoricalsocietyisoftenjustawayoftalkingaboutearlierphasesofthecareerofsomepresentsociety.Itissometimesthoughtthathistoricismisprovidedsomesupportbythedifficultiesthat437\nLEONPOMPAanthropologistsencounterwhentryingtograsptheruleswhichcanmakesenseofthepracticesofprimitivesocieties.Butthisisanunhelpfulparallelfor,whileanthropolo-gistsoftenstartwithnoknowledgewhatsoeverofthelanguageofthesocietiestheystudy,thisisnotsointhehistoricalcase.Asarguedintheprevioussection,thehisto-rianisnotcompletelyalientothepastwhichheorsheseekstounderstand,forthesocietyinwhichanhistorianfindshimselforherselfisaproductofthatpast.Itcarrieswithinitmanysocialandlinguisticpractices,someofwhicharedevelopments,andothersremnants,ofearlierpractices.Ifpresentstandardshavearisenfromearlierones,thereseemsnoreasoninprinciplepreventingthehistorianfromworkingbackwardstotheirearlierstages.Ifthisisso,wecanrefutethisversionofanhistoricistargumentbyappealingtothefactthatwedonotfindunderstandingallstandardsinearlierpartsofhistoryasimpossibleastheobjectionsuggests.CertainlythereseemstobenomoredifficultyinunderstandingtheGreeks’recoursetotheDelphicOraclethanacontem-porarynon-believermightfindinunderstandingabeliever’srecoursetoprayer.Itcouldbeargued,ofcourse,thatnon-believershavenounderstandingatallofreligiousprac-tice.Butthiswouldlimittheconceptofunderstandingtowhatwebelieveandthatisfartoosevereastandard.Wecan,surely,understandearlycosmologicaltheoriesevenifwedonotbelievethem.Moreover,theattempttolimitwhatwecanunderstandtowhatwebelievewouldhavescepticalimplicationsformuchsharedeverydayunder-standingthatwetaketobequiteunproblematic.Butinvestigationofthispossibilityisbeyondthescopeofthischapter.Thescopeofrationalreconstructionextendsbeyondactioninthestrictsensetoincludewhathasbeenmade–includingthedifferentpoliticalandeconomicinstitu-tions,technologiesandartisticcreationsofthepast–byenablingustounderstandthepointofthesecreationsand,throughthis,tounderstandwhytheyhavechanged.Inhistoriesofart,forexample,pastinfluencesplayaveryimportantpart,asinthecaseoftheeffectwhichtherediscoveryofclassicalarchitectureandsculpturehaduponRenaissanceartandarchitecture.Butrationalreconstructionsarerequiredherepre-ciselybecausetheinfluenceoftherediscoveryoftheclassicalworldlayinthewaytheRenaissanceartistsandarchitectsassimilatedclassicalthoughtintotheirownandthroughthatsawtheexpandedrangeofpossibilitieswhichtheycouldincorporateintotheiraims.Nevertheless,rationalreconstructionsarenottheonlykindofexplanationstobefoundinhistory.Collingwoodclaimedthatallhistoricalexplanationshouldbeintermsoftheagent’sthought,butthiswascertainlyanoverstatement.ThepoliticalsupremacyofAthensdependeduponafleetofveryexpensivetriremes,afleetwhichAthenscouldnothavebuiltwithoutthediscoveryofnewsourcesofsilver.EvenifanexplanationofAtheniansuccessinvolvesestablishingthereasonswhythecitytooktheunusualdecisiontoinvestthiswealthinbuildingthetriremes,theexplanationcannotomitthisdiscovery,withoutwhichAthenscouldnotevenhaveconsideredthedecision.Humansliveinanaturalenvironmentandmanyoftheirpracticesderivefromtryingtocopewiththisenvironment.Referencetoitcanthereforeneverbewhollyomitted.Buttorecognizethisisnottoreadmitthecoveringlawmodel.Itissimplytoacceptthatwhilehistoriansareprimarilyconcernedwithunderstandinghistoricalactions,theyoftenneedtorefertothenaturalworldtoachievethis.438\nPHILOSOPHYOFHISTORYItmustbementioned,however,thatmanyphilosophersdonotacceptthataperson’sreasonfordoingsomethingcanbeacauseofdoingit.Theyarguethatanactionisanevent,albeitofaspecialkind,andeventscanbebroughtaboutonlybyotherevents.ThisisanissuemuchdebatedinthePHILOSOPHYOFMIND(chapter5).Here,however,weneednoteonlythat,evenifthiswerecorrect,itwouldgivenosupporttothecoveringlawtheoryaswehavebeenconsideringit.Forifreasonscanbecausesonlyiftheyareinsomesensephysicalevents,theywillneedtobestrippedofthedescrip-tionsunderwhichtheyarecitedasreasons,including,forexample,‘ambitiousforlit-eraryfame’or‘feltmaltreated’,whereasthecoveringlawtheoryrequiresthattherebelawsaboutthebehaviourofagentsunderthesedescriptions,includinglawsaboutwould-beliteratiormaltreatedsubjects.3ObjectivityandValueTheprevioussectionssupporttheclaimthathistoriansareprimarilyconcernedwithwhatisofinterestinthepastfromahumanpointofview.Inthiscase,however,theymustalsobeconcernedwithhumanvalues.Thisraisesnoproblemfortherationalreconstructionmodelitself,providedthatthehistoricistdifficultycanbeovercome,sincemanyactionsareundertakeninthelightofanagent’svalues.Ithasbeenargued,however,thatthehistorian’sownvaluejudgementsmustaffecthisorheraccountinanumberofwaysanditispossiblethatthismayaffecttheclaimthathistorianscangivetrueaccountsofwhatreallyhappened.PastandPresentViewpointsWemayseeonewayinwhichtheissueconcerningvaluejudgementsarisesbyaskingwhosepointofviewthehistorianisinterestedinrecapturing.Ifthehistorianwerecon-finedtoseeingthingsintermsofrationalreconstructionsitwouldseemthatitmustbethepointofviewofhistoricalagents,sinceitistheirviewoftheirsituationthatthereconstructionisconcernedtoelucidate.Buthistoriansdonotconfinetheiraccountsinthiswayandoftentheyexplainthingsintheirhistoricalcontextbyreferencetofactorsunknownandevenunknowablewithinthatsociety.Whethertheyshoulddothishasbeenasourceofcontroversy,reflectingdifferentviewsaboutthevalueofstudyinghistory.Oneview(Lovejoy1959)isthatitsmostimportantvalueliesinitscapacitytoenableustograspthepossibilitiesofotherwaysoflivingand,hence,ofothersetsofvaluesthanourown.Thissuggeststhatthehistorianshouldbepri-marilyconcernedwiththepastinitsowntermsandasacontrasttothepresent.Others,however,haveclaimedthatitsvalueliesinitscapacitytoenableustounderstandcertainaspectsofthepresent.Butthismeansthatourconcernwillbewiththosepartsthatlinkupwithcurrentinterests.Ithassometimesbeenheldthatthesetwoviewpointsareincom-patiblealthoughhistoriansusuallyinterweavethemboth.Theyfrequentlymakeclaimsaboutthetruththatconflictwithwhatwasbelievedatthetime,oftenbecausetheirknowl-edgeofwhatsubsequentlyoccurredenablesthemtoshowthattheagent’sviewsweremis-taken.Moreimportantly,however,theyoftendescribeandgivesignificancetothingsinwayswhich,becausetheyrequirehindsight,wereunavailabletopastagents.439\nLEONPOMPATheroleofhindsightcanmosteasilybeshownbyattendingtoaparticularkindofsen-tencewhichhistoriansuse,nowknownasa‘narrativesentence’(Danto1965:143ff.).Anarrativesentenceisoneinwhichaneventorpersonisdescribedinthelightofaconnectionwithsomethingthatoccurredlater.Itstruthcanthereforebeknownonlybysomeonewhoknowsthetruthaboutthelatereventandnotbyanyoneatthetimeoftheearlierevent.Theearliereventisthusseenaspartofatemporalwhole.TouseoneofDanto’sownexamples,anhistorianmayassertthatwhenPetrarchclimbedMountVentouxheopenedtheRenaissance,toemphasizePetrarch’ssignificanceinthesetofeventsthatconstitutetheRenaissance.ThetruthofsuchanassertioncouldnothavebeenknownbyanybodywhowitnessedPetrarch’sascentsinceheorshecouldnotknowthefuture.Nevertheless,thedescriptionisnotincompatiblewithsomethingwhichtheeye-witnesscouldhaveknown,namely,Petrarch’sascentofMountVentoux.Allthatishappening,therefore,isthatthesameeventisbeingdescribedfromtwodif-ferenttemporalperspectivesandinthelightofitsconnectionwithdifferentevents.Sincethestatementsarecompatible,however,itisnotthecasethatifstatementswhichreflectonereasonforstudyingthepastaretrue,thosereflectingtheothermustbefalse.Thusthesedifferentviewsofthevalueofstudyingthepastgiverisetonoproblemsabouttheobjectivityoftheaccountsoffered.Thereare,however,twootherwaysinwhichithasbeenclaimedthatvaluejudge-mentshaveadetermininginfluenceuponhowthehistorianconstruesthepast,andthesemayhaveconsequencesfortheobjectivityoftheaccountsoffered.Toseewhythisisso,itisnecessarytoalludebrieflytothecurrentdebateaboutthecognitiveornon-cognitivestatusofVALUEJUDGEMENTS(pp.4–6).Cognitivistsholdthatvaluejudgementsareeithertrueorfalse.Ifthisisso,andifhistoricalaccountspresupposesuchjudge-ments,thenthetruthorfalsityofhistoricalaccountswilldependuponthatofthepar-ticularjudgementsmade.Non-cognitivists,however,claimthatvaluejudgementsareexpressionsofattitudeoremotionand,assuch,thattheyhavenotruthvalues.Ifthisissoandifhistoricalaccountspresupposethem,theywillturnouttobeaccountsexpressiveofcertainattitudestowardsthepastbutnotaccountswhichcanbesaidtobetrueorfalse.Thus,althoughtheindividualstatementsandexplanationstheycontaincanstillbetrue,theaccountasawholecannotbetrue.Thereremainsthefurtherpossibility,however,thathistoricalaccountsdonotpresupposevaluejudge-ments,inwhichcasetheirobjectivitywillnotbeaffectedbythem.Thefirstargumenttobeconsideredconcernsthequestionwhether,sincehistorianscannotincludeeverythingintheiraccounts,theymustusevaluejudgementsintheirselectionofwhattoinclude.Manyreasonshavebeenofferedtoshowthattheymust,buthereIcanconsideronlyone:thatthehistorianmustdecidewhicheventsweremoreimportantthanothersand,sincetherearenoobjectivecriteriaofimportance,thisrequireshistorianstomaketheirownvaluejudgementsaboutthem(Dray1993:42–6).Whethertheiraccountsaretruewillthendepend,atleastinpart,uponwhethertheirjudgementsofimportancearetrue.Attemptstomeetthisclaimbyarguingthatthereareobjectivecriteriaofimpor-tancehavenotbeengenerallyaccepted.Themostplausiblesuggestionisthathisto-riansshouldchoosethoseeventsthathavethewidestrangeofcausalconsequences.Butsincenotallthecausalconsequencesofaneventcanbetakenintoaccount–indeed,itisnotclearhowtheycanevenbecalculated–itseemsthatonlythemost440\nPHILOSOPHYOFHISTORYimportantshouldbetakenintoaccount.Inthiscase,however,furtherjudgementsofimportancewillberequired.Ithasbeennotedalsothatthiscriterionhaslittlerele-vancetohistorieswhichdescribeandinterpretthemaincharacteristicsofanage,suchasG.M.Young’sVictorianEngland:PortraitofanAge.‘What’,ithasbeenasked,‘wouldwethinkofavolume-length‘‘portrait’’ofVictorianEnglandthatignoredtheworking-classmovement?’Suchanomissionwouldneedjustificationandtheonlyoneavailableseemstobethatwhatwasincludedwasmoreimportantthanwhatwasexcluded(Dray1993:42).Itisnotclear,however,thatthisargumentmustleadtotheconclusionthatthehis-toriancannotavoidmakinghisorherownvaluejudgement.Toseethis,itisworthcon-sideringtwopossibleobjectivistresponses,relatingtothetwodifferentreasonsforstudyinghistorymentionedabove.Iftheportraitismeanttobeaportraitofapeopleastheysawthemselvestobe–andthisiswhatYoungclaimsinhisIntroduction–thehistorianwillnotgetveryfarinhisorherresearchwithoutmakingtheempiricaldiscoverythatthestateandactivitiesandconditionoftheworkingclasswerewidelyperceivedasamatterofimportancebyverylargepartsofthepopulation,includingworkers,factoryowners,politicians,philanthropistsandmanyothers.Itistruethatthehistorianneedstoknowwhichsourcesheorsheshouldconsulttolearnthis,butthisispreciselywhatawell-trainedhistorianistaughttoknow.AnhistorianstartingresearchwiththehypothesisthattheworkingclasswasnotacauseofconcerntotheVictorianswouldsurelyneedtoabandonthisifthesourcesindicatedthatitwaswrong.Thisimpliesthatthesourceshave,inacertainsense,acapacitytospeaktothehisto-rian.Theywerenot,ofcourse,producedtospeaktohistoriansbuttheywereproducedtospeaktosomebody,andithardlyrequiresavaluejudgementtohearwhattheyweresaying,whatworriestheywereexpressingor,ifmanyareaboutthesamething,tocon-cludethatthatwasamatterofwidespreadconcern.Thehistorianmay,therefore,needtoabandonaninitialhypothesisinviewofwhatthesourcesrevealand,ifthisisso,thereisnoreasontobelievethatdoingsoinvolvesmakingavaluejudgement.Itistrue,ofcourse,thattheremaybeareasoflifewheretherewasnosuchconsensusofimpor-tance,butifthisisso,itissurelyanhistorian’sdutytoreportthisfact.But,asnotedabove,historiansoftenwritefromthepointofviewofthebearingofthepastuponpresentinterestsandforanhistorianofthissortthefactthattheVicto-riansthoughtthedevelopmentoftheworkingclasswasimportantwouldnotprovideacriterionforincludingitinanaccountofVictorianEngland.Butalthoughthisistrue,analternativeobjectivecriterionfortheinclusionofthedevelopmentoftheworkingclassinthissortofaccountcanbefoundintheimportancethatthehistoriancanshowthatdevelopmenthadforcertainlaterchanges.Wouldit,onemightask,havebeensomisleadingforanhistorianofthissorttohaveomittedanaccountofthedevelopmentoftheworkingclassinfavourofsomethingelsehadthatdevelopmentnotcometotransformalmostthewholesubsequentstructureofBritishsocial,economic,politicalandconstitutionallife?Whatmustberememberedhereisthathistoricalrealityisaprocessofongoingbutcontinuouschangeandthat,astheworldsinwhichhistoriansfindthemselvesdiffer,differentaspectsofthepastbecomerelevanttoanunderstand-ingofthepresent.Onthisview,whatdeterminesanhistorian’sselectionofmaterialisnotwhatheorshepersonallyholdsimportant,butwhatthathistoriancanfindtoexplainthethingsinvolvedinasociety’sperceptionofitself.Thereasonwhya441\nLEONPOMPAcontemporaryhistorianofthiskindwouldgivemoreprominencetoanaccountoftheworkingclassthan,say,anhistorianinthelatenineteenthcentury,mustsurelybethattheworkingclassnowaffectssomanymoreaspectsofthestructureofourpresentworld,andissomuchapartofwhatwetakethatworldtobe,thataportraitoftheperiodinwhichitfirstarosethatomittedthatdevelopmentwouldnolongerbeacceptable.Thisisaconsequentialistbutnotanarrowlycausalcriterion.Itcouldbeargued,however,that,sinceitappealstowhatwethinkimportantaboutthepresent,allthatitdoesismakejudgementsofimportanceaboutthepastconsequentuponothersaboutthepresent.Butthisseemsimplausible.Foritissurelyanobjectivefactthattheworkingclassnowoccupiesacertainplaceintheeconomy,thatithasinstitutionswhichclaimtospeakforitandtoprotectitsinterests,andthatthisispartofoursharedunder-standingofourpresentworld.Accordingly,historiansmustincludeintheiraccountswhatevertheycanfindtoexplainthisfactiftheirintentionistoexplainthepresentintermsofthepast.Moreover,inremovingthebasisofselectionfromtheindividualhis-torian’sownjudgementofimportancetotheself-perceptionofthesocietyforwhichthehistorianwrites,thiscriterionprovidesasharedbasisonthestrengthofwhichindividualhistorianscanreasonablyaskoneanothertojustifytheirselectionofsomethingsinpreferencetoothers.Thesecondobjectionthatraisesquestionsabouttherelationbetweenvalueandobjectivityistheargumentfromcharacterization.Theclaimhereisthatbeforedecid-ingthatacertainitemissufficientlyimportanttobeincludedinanaccount,inthecaseofcertainkindsofhistoryatleast,thehistorianmustfirstdecidewhethertheitemshouldbecharacterizedinsuchawayastomakeittherightkindofcandidateforinclu-sionandthatsuchadecisionagainrequiresavaluejudgementbythehistorian.Thesortsofhistorieswhichmostobviouslyseemtoraisethisdifficultyarehistoriesofartandreligion,whereitmustbedecidedwhetheracertainworkisgenuinelyaworkofartoracertainactiongenuinelyreligiousbeforedecidingwhethertoincludeitintherelevanthistory.Butthecasecan,ithasbeenargued,beextendedalsotopoliticalhistory,whereitcanbeaskedwhethertheactionsoftheIRAorthePLOareactsofter-rorismoractsofwar.Thejudgementsmadewilldecidethekindsofhistoryforwhichtheyarecandidatesforinclusion(Dray1980:45;1993:49–51).Inconsideringthisviewitisnotnecessarytodenythateventsfindtheirwayintotheirrelevanthistoriesonlyascharacterizedorundercertaindescriptions.Theques-tionatissueissolelythebasisofthecharacterization.Thefirstpointtobenotedisthatwewouldloseourcapacitytotalkaboutanything,pastorpresent,weretherenotawidespreadbasisofsharedcharacterizations.TheIRA,forexample,maydescribeitsactionsasexecutionsratherthanmurders,butthisclaimcannotbetakenasliterallytrueintheabsenceoftheinstitutionallegalandjudicialproceduresthatmustoperatebeforesomethingcanbeanexecution.IndescribingitsactsasexecutionsratherthanmurderstheIRAmaybeexpressingitsrejectionofthelegitimacyofBritishruleinUlster,butthatisnotsufficienttoturnmurdersintoexecutionsunlesstheactsissuefromalternativelegalandjudicialprocesses,thelegitimacyofwhichisestablished.Thisexample,ofcourse,raisesthequestionofpoliticallegitimacy,whichistoolargetobediscussedhere.Nevertheless,itremainsthecasethatwemusthaveasharedbasisofcharacterization,andhencesharedcriteriaofcharacterization,beforewecandiscuss442\nPHILOSOPHYOFHISTORYanything.Thisdoesnoteliminatetheexistenceofborderlinecases,butitdoesmeanthatthemajorityofcasesmustbeunarguableand,indeed,thatitisbyreferencetothemthatwecanidentifyanddiscussborderlinecases.Itmustbenoted,however,thatcriteriaarenormallyimplicitandoftencannotbemadeexplicitwithoutreferencetoobjectsthatexemplifythem.Thisisevidentwhenweturntohistories,evenofthoseareaswhichseemmostappropriateforthecharacterizationargument.Forhereitisnoticeablethatthehistoricalprocessitselfhasproducedthecharacterizationasworksofartorasreligionsofthethingsthatconsti-tutethesubjectmatterofmosthistoriesofartorreligion.Butthesecannotsimplybedisregardedinfavourofnewcharacterizationswhichanhistorianmightwishtopropose.ItwouldsurelybeveryoddifanhistorianwhowasalsoadevoutChristian,weretoexcludetheritualsatDelphiasacandidateforinclusioninahistoryofreligionsimplybecausetheyinvolvednobeliefswhichaccordedwiththoseoforthodoxChris-tianityornonewhichheorshepersonallywaspreparedtoevaluateas‘genuinely’reli-gious.Theclaimthattheyshouldbeexcludedonthesegroundscannotbecorrectsince,byparityofargument,anhistorianbelongingtoanyotherfaithwouldbeentitledtowriteahistoryofreligionforwhichChristianityisnotacandidate.Indeed,onthislineofreasoning,therewouldbenothingtoblocktheclaimthatonereligionwastheonlyappropriatecandidateforahistoryofallreligion.Itwouldbeequallyunacceptable,forthesamereasons,ifanhistorianofItalianartweretoexcludetheworksofMichelan-geloascandidatesforhisorheraccount.Surelysuchcasesraisequestionsabouttheclaimthatahistoryofartorreligionhasbeenproducedratherthanquestionsabouttheentitlementoftheworksandinstitutionstobecharacterizedasappropriatecandi-datesforinclusioninthesehistories.Thus,thereiswhatmightbecalleda‘verdictofhistory’butthisisnotaverdictthathistorianshavereached;insteaditisonethathasarisenthroughtheinternalhistoricaldevelopmentoftheirsubjectmatter.Thisisnottodenythatcriteriamaychange,but,iftheydo,thiscanonlybeasaconsequenceofmajornewdevelopmentsintheirsubjectmatterandnotofhistorians’valuejudgements.Aswiththeargumentfromselection,itmightberepliedthatthisdoesnotshowthatvaluejudgementshavenoparttoplayinthecharacterizationofhistoricalphenom-ena,sinceitmerelyreplacesthepartplayedbythehistorian’svaluejudgementbysharedcharacterizationswhichhavearisenthroughpastvaluejudgements.Butiftheargumentiscorrect,thereisasignificantdifferencebetweenthetwopositions,foritmeansthatthehistorianmustrespectthestatethatthehistoricaldevelopmentofthesubjectmatterhasreached,withallthecharacterizationofmaterialthatthatinvolves,andcannotbethoughtofasbeinginapositiontoreconstructitabinitio,asitwere,accordingtohisorherownvalues.Asnotedabove,borderlinecaseswillinevitablyremain,butthereisnoharminthisprovidedthattheyaresorecognized.Anhistorianisunlikelytocommandmuchsupport,however,inattemptstomakethesecentralandtorearrangeallotheritemsofcharacterizationaccordingly.Itisdifficulttoimaginewhatsortofhistorymightbepro-ducedbyanhistorianwhoacceptedanIRAkillingasacentralcaseofanexecutionandtriedtowriteahistoryofpolitics,withattendantrevisionsofthenotionofpoliti-calauthority,onsuchanassumption.Itisevenmoredifficulttobelievethatsuchahistory,ifwritten,wouldbeacceptedasatrueaccountofthepast.443\nLEONPOMPA4NarrativeandRealismHistorianshavetraditionallypresentedtheresultsoftheirresearchintheformofanar-rative.Thatthisisasitoughttobehasbeenchallengedbythedevelopmentofnewschoolsofhistorians.Theseinclude,forexample,econometrichistorianswhousequantitativeeconomictechniquestoestablish‘objective’accountsofthecausesofsocialandpoliticalchangeortheFrenchannalisteschool,whichgivespreferencealsotoquantitativetechniquesandtreatsindividualsandeventsasrelativelysuperficialconsequencesofdeeperunderlyingstructuresandprocesses,theexistenceofwhichcanbediscernedonlyoverdifferentlengthsoftime.Nevertheless,despitecallsforfurtherdevelopmentsofthesekindsofhistory,narrativeshaveremainedthemostchar-acteristicwayofpresentinghistoricalknowledgeandtherehasbeenastrongdebateabouttheirepistemologicalstatusandtheirontologicalimplications.Thebasicissuearisesfromclaimsaboutthenatureofnarrativeitself.Narrativesmusthavesomeforminvirtueofwhichwecanfollowwhattheysayandvariousanaly-sesoftheirtypicalformhavebeenproposed(Gallie1964;White1965;Danto1965;White1973).Inalloftheseanalysesitisagreedthat,althoughanarrativepresupposesachronologicalsequenceofevents,itgoesbeyondthisinvariousways;forinstance,byconnectingdifferenteventsandactionsthroughtheirnecessaryconditions(Gallie1964)orbyexplaininglateraspectsoftheobjectofenquiryintermsofearliercauses(White1965).But,drawinguponmodernliterarytheory,twochallengingclaimshavebeenmadeaboutnarrativesingeneral:firstly,thatnarrativesareliteraryartefactsand,assuch,constructionsoftheliteraryimagination;secondly,thattherearemanydifferentformsthatanarrativemighttakeandnothingtodeterminethatanauthor,whetherhistorianornovelist,shouldadoptoneratherthananother.Thus,thereisindeterminacyatthelevelofnarrativeconstruction.Applyingtheseconsiderationstohistoricalnarratives,ithasbeenclaimedthatwecannotthinkofthepastasadeter-minaterealityofwhichanarrativeasawholeistrueforthereisnoonerightwayofcomposinganhistoricalnarrativeaboutthepastand,hence,ofclaimingthatthroughitwehavecapturedthemeaningofthepast.Rather,thereareanumberofdifferent,irreconcilablewaysofcreatingnarrativesaboutthepastand,throughthem,ofimposinguponitavarietyofmeaningsthatitdoesnotpossessofitself.Sincemeaningisnotanobjectivefeatureofthepast,itcannotbeclaimedthatthehistoricalnarra-tives,whichprojectmeaninguponit,areexpressionsofobjectivetruth(White1973:1–42;Mink1978).Tomentiononepointinillustrationofthisview,itisanecessaryfeatureofastorythatitshouldhaveabeginning,middleandend.Butrealityhasnobeginnings,middlesandends.Hence,whenwethinkaboutthepastintermsofbeginnings,middlesandends,asnarrativesinviteustodo,weareimposinguponitastructurewhichitdoesnotitselfpossess.Onemustnote,however,thattheclaimallowsthatsomethingsinthenarrativecanbetrue,suchasthefactsthatitcitesortheexplanationsthatitoffers.Forwerethisnotpossible,thedistinctionbetweenfactandfictionwouldbedestroyed.Buttheanti-realist’sintentionisnottochallengethisdistinction.Itis,rather,topreservethehistorian’srighttopresentuswithavarietyofdifferentbutincompatiblewaysofthinkingaboutthepast.Itisonlythewayinwhichthenarrativeorganizestheseintoacoherentwhole,itisclaimed,whichcannotexpress444\nPHILOSOPHYOFHISTORYafurthertruthaboutrealconnectionsinthepast.Thus,tothink,astherealistdoes,thatthenarrativeasawholeassertssomethingthatistrueofthepast,inthewayinwhichsomeofitsconstituentpartsdo,isamistake.Itisthemistake,asithasbeenput,ofthinkingofthepastas‘anuntoldstorywaitingtobetold’(Mink1978).Thecontroversyisthereforeaboutwhether,inconstructinganarrativeaboutthepast,thehistorianmustinevitablyfalsifyit.Althoughthisclaimhasbeenappliedonlytonarrativehistories,ithasimplicationsforotherformsofhistory,manyofwhichhavenarrativeelementsinthem.AninterpretativehistorysuchasYoung’sVictorianEnglandcontainsmanypassagesofnarrative.Indeed,itisdifficulttoseehowanyformofhistorywhichinvolvestracingandexplainingatleastsomedevelopmentscanfailtocontainnarrativeelements.Moreover,althoughtheclaimhasnotbeenextendedtootherformsofhistorysuchaseconometrichistory,itwouldseemthat,insofarasallkindsofhistoricalaccountsrequiresomeformofliteraryexpression,objectionsofthesamekindmustapplytothem.Beforediscussingthisposition,oneotherreasonthathasbeenadvancedforitshouldbementioned,namelythatanarrativecontainsmuchthatinvolveshindsight.This,aswesawearlier,istrue.Thus,anhistoriancantalk,retrospectively,of‘hopesunfulfilled,plansmiscarried,battlesdecisiveandideasseminal’.But,itisclaimed,theseareaspectsofstoriesandnotoflife,forinlifethereareonly‘hopes,plans,battlesandideas’(Mink1970).Hence,byusinghindsight,thehistorianfalsifiesthepast.Butthisargumentcannotbecorrectfor,althoughitistruethatwecanonlyknowcertainthingsatcertaintimes,thereisnoreasonwhyweshouldtakewhatweknowonlyatcertaintimes,forinstance,whiletheeventisoccurring,asthesolewayofcharacterizingreality.For,aswesawwhendiscussingnarrativesentences,thedescriptionsthathistoriansofferret-rospectivelyarenotincompatiblewiththosethatcanbegivenatthetimeoftheevent.Therecanonlybea‘hopeunfulfilled’iftherewasa‘hope’inthefirstplace.Butthefactthattheformerpresupposesthelatterisnotareasonforbelievingthatonlythelattercanbetrue.Norcantheroleofretrospectivedescriptionsbedeniedbyconfiningtheconceptofrealitytolifeasweexperienceit,foritispartoflifeasweexperienceitthatweseeoursituation,atanyparticularmoment,intermsoffulfilledorunfulfilledhopes,successfulorunsuccessfulplansandsoon,andabandonorreviseourfurtherhopesandplansinthelightofthisknowledge.Thus,itisfalsethat,asthisobjectionasserts,retrospectivityisafeatureofnarrativesbutnotoflife.Giventhestronglyanti-realistimplicationsoftheseclaimsaboutnarratives,itishardlysurprisingthattherehasbeenastrongrealistreactionagainstthem.Oneofthesebuildsuponthepointjustmade:that,contrarytowhattheanti-realistsassert,webothliveoutnarrativesinourownlivesandunderstandourlivesintermsofthosenarratives.Hence,itisappropriatetounderstandthelivesofothersinthesameway(MacIntyre1981:197ff.).Butwhilethisisanadequateresponsetoanti-realistobjec-tionstoretrospectivity,itcannotbeanadequateresponsetothemorefundamentalclaimsabouttherelationofnarrativestoreality.Foranarrativehistoryisnotmerelyanaccountofthewayinwhichindividualshavelivedouttheirlives.Thefocusmay,aswesawearlier,beuponthehumanpast,butthelatterrequiressocial,economicandnaturalcontexts,manyofwhichwillimpingeuponindividualsinwaysunknowntothem.Thehistorianwill,ofcourse,drawupontheseingivinganarrativeaccount,butindoingsoheorsheisgoingwellbeyondthesortofunderstandingwhichwehave445\nLEONPOMPAwhenwetrytounderstandourselvesinnarrativeterms.Historicalrealitydoesnotconsistsolelyofhumanagentsandthereisthereforenoreasonwhyweshouldcon-cludethatbecauselivedexperiencehasanarrativestructure,thesamemustbetrueofhistoricalreality.Astrongerandmuchsimplerrealistdefenceistochallengetheanti-realist’swayofsettinguptheproblem.Wecandothisbyacceptingtheclaimthatstoriesneedstruc-turesandauthors.If,therefore,therealistiscommittedtothenotionofpastrealityas‘anuntoldstorywaitingtobetold’,heorsheisindeedcommittedtoanincoherentcon-ception,sincepastrealitywouldthenbothneedbutnothaveanauthor.However,thiscanbeshowntobeanunacceptablewayofpresentingtheproblembythefactthat,ifthisisso,allclaimsaboutthepast,includingthosethattheanti-realistwishestoaccept,wouldbeincoherent.Toseethis,letusconsiderthecaseofsomestatementmadebyanhistorianaboutthepastwhichwasnevermadebyanyoneinthepast.Statementsarelesscomplexthannarrativesbuttheyalsorequirealiterarystructure,usuallyasentence,andanauthor.Hence,if,astheanti-realistclaims,thenarrativerealistiscommittedtotheconceptionofthepastasan‘untoldstorywaitingtobetold’,astate-mentrealistwouldbecommittedtotheconceptionthatpartofthepast,thatwhichtheanti-realistneedstoaccepttopreservethedistinctionbetweenfictionandhistory,involvestheequallyincoherentnotionof‘unstatedsentenceswaitingtobestated’.Onthisview,then,theideaofanyindependenthistoricalrealitywhatsoevermustbeaban-donedasincoherent.Butthiswouldproduceascepticismthatappliesnotmerelytonarrativehistories,buttoanyclaimsofanysortaboutthepast.Since,however,aswesawinthefirstsection,ascepticismofthissortwoulditselfbeincoherent,theanti-realistargumentwhichentailsitcannotbecorrect.Theuseofnarrativesthuscannothavetheimplicationsthattheanti-realistclaims.Nevertheless,torejecttheanti-realistviewisnotthesameastounderstandhownarrativesrelatetothepast.Fromconsiderationspresentedearlier,itmightseemthattheydosobytellinguswhathappenedandbyshowingushowtounderstandwhyithappened.Butthisseemstodoscantjusticetotheroleofthenarrativeitselfwhich,astheanti-realistshaverightlyinsisted,involvesrelatingindividuals,eventsandcontextsofmanydifferentkindsandfromdifferentpointsofview,includingthoseoftheagentsandthehistorian.Furthermore,neitherwoulditexplainthefact,troublingforrealists,thathistoriansdisagreesomuchintheiroverallaccounts,norwoulditidentifythesourceoftheirdisagreements.Onepromisingsuggestioninvolvesthenotionofanhistoricalnarrativeasprovid-ingaspecialformof‘configurationalunderstanding’whichischaracteristicofhistory(Mink1987;Dray1993).Therearetwoaspectstothisidea,whichmaybetakenseparately.Thefirstisthatthenarrativeintegratesdifferentpartsofthepastaspartsofatemporalwhole.Thiscanbequiteeasilyunderstoodifwereturntotheideaofanarrativesentence.Heresomeeventorinstitutionisdescribedinthecontextofalarger-scaleeventorprocesswhichdidnotenduntilaftertheterminationoftheoriginaleventorinstitution.Itisclear,however,thathistorianscouldnotmaintainclaimstothetruthofsuchsentencesiftheyweretoconnectearlierandlatereventsinanarbitraryway.WecandescribePetrarch’sascentofMountVentouxastheopeningoftheRenaissance,buthardlyastheopeningoftheIndustrialRevolution.Noristhisjustamatteroftime-scale,foreventsverydistantfromoneanothercanberelatedinnarrativesentences.446\nPHILOSOPHYOFHISTORYWhatisrequirediswhathasbeencalleda‘colligatoryconcept’ofanappropriatetype(Walsh1967).Acolligatoryconceptisonethatreferstosomelarge-scaleunderlyingmovementorsomedevelopingidea,withinwhichtheeventsorindividualsinwhomthehistorianisinterestedwillbeseeninvarioussetsofrelationshipsandthusacquiredifferentkindsofsignificance.Butwhatdetermineswhethersuchaconceptisoftheappropriatetype?Why,forexample,istheRenaissancebutnottheIndustrialRevolu-tionappropriateforPetrarch?Therearetwopartstotheresponsetothisquestion.Thefirstisthatthehistorianmustbeabletoshowthatthenewrelationshipswhichheorshecantracebyadoptingthisviewpointareempiricallytrue.Thereisnoreasoninprinciplewhyanycolligatoryconceptmaynotbeusedintheattempttounderstandthesignificanceoftheactivitiesofanyindividual.Butaminimumconditionofappro-priatenessmustbethatthehistoriancanprovideempiricalsupportforthepossibilitiesopenedup.Thesecondisthatacolligatoryconceptmustnotbeonewhichmerelyallowseventstobereclassifiedingeneralterms–amurderasanexecution,forexample.Itmustbeonethatreferstospecifichistoricalitems,suchastheRenaissanceortheFrenchRevolution.Thus,colligatingunderappropriateconceptsisaspecificallyhis-toricalformofunderstandingbecauseitinvolvesunderstandingspecifichistoricalphe-nomenainthelightoftheirrelationshipstootherspecificphenomena,thecharacterofwhichcanbeascertainedonlywithhistoricalhindsight.Ifthisisoneofthemaincharacteristicsofhistoricalnarratives,itmightseemthatitcanexplainwhatisaproblemforrealists,theexistenceofapparentlyirresolvablelarge-scaledisagreementbetweenhistorians.Forthesameeventsandpeoplecanbebroughtunderdifferentappropriatecolligatoryconcepts.Toseethis,onemerelyneedstothinkofthesignificanceofLuther’sNinety-FiveThesesagainsttheabuseofindul-genceinrelation,forexample,tothehistoriesoftheRomanandProtestantchurches,totheology,tochangesofrelationshipsbetweenchurchesandstatesandtothedevel-opmentofthemodernProtestantconscience.Infact,however,whiletheuseofdiffer-entcolligatoryconceptscanexplainhowhistorianscometogivedifferentnarrativeaccountsofthesameevents,itcouldshowthatthesedifferencesareirresolvableonlyifthecolligatoryconceptswerecontradictory.Butsincetheyareconceptsofspecificlarge-scalehistoricalphenomena,whichareknowntohaveexisted,theycannotbecontradictory.Whensuchdisagreementoccurs,therefore,itcanbeputdownonlytodisagreementabouttherelativeimportanceofthethemesreferredtobythecolligatoryconceptsinquestion.Butthereisnoreasonwhyhistoriansshouldthinkthattheseaccountsareincompatible,unlesstheybelievethatthereisoneandonlyonewayinwhichhistoricalrealitycantrulybedescribed.But,aswesawearlier,arealistcanacceptthatthereisonlyonereality,topartsofwhichalltruestatementsandaccountsrefer,withoutneedingtodenythattherearemanydifferentwaysofreferringtoanddescribingthatonereality.Itisnecessarytoturntothesecondaspectofthisconceptionofhistoricalnarrative:thenotionthatitisaconfiguration.Thisconcepthasbeenmuchusedintryingtochar-acterizehownarrativesexpresstheirtruth.Itarisesfromareluctancetoaccepteitheroftwoalternatives.Thefirstisthatnarrativesstatethetruth,forinthiscaseitmightseemthattheirtruthwaslittlemorethanafunctionofthestatementsandexplana-tionsthattheycontain.Thealternativeisthatanarrativeissimplyaliterarywayofdressinguptheresultsofthehistorian’sresearchand,inthiscase,onecannotascribe447\nLEONPOMPAtothenarrativeassuchanytruthvalueatall.Themainclaimmadeinconnectionwiththeconceptofconfiguration,however,isthathistoricalnarrativesgobeyondtheexplanatoryvalueoftheircontentstoaformofunderstanding,inwhichthepartsarerelatedtoawhole,asdescribedabove,butinsuchamannerthatthehistorian’scon-clusionscannotbegraspedotherthanbythewayinwhichtheyareexpressedinthenarrative.Thus,ratherthanstatethetruth,thenarrativedisplaysit.Oneanalogyusedhereiswithmusic,whereitisoftenclaimedthatthereisnowayofunderstandingwhatapieceofmusicisaboutotherthanbylisteningtoit.Inthesameway,itissuggested,itisnotpossibletounderstandthehistoricalnarrativeotherthanbygraspingtherela-tionshipsbetweenpartsandwholeinjustthewayinwhich,bytheend,ithasdisplayedthem.Theanalogywithmusic,however,isbothhelpfulandunhelpful.Itishelpfulinthesensethatonecancontrastlisteningtoapieceofmusicsequentially,asitwere,notinghowonethemeisrelatedtoordevelopsoutofanother,andgrasping,bytheend,thepieceasawhole.Inthelattercase,onecangainamuchmorecomplexunderstandingofthesamepartsofthepiecebyseeing,forexample,howsomeforeshadowwhatcomeslater.Similarly,sincenarrativespresupposechronologicalsequences,theyhaveasequentialstructure–yetthehistorianenablesustounderstandthepartsofachrono-logicalsequenceinmanydifferentways,byrelatingthemtodifferentsetsofrelationsandcontextsthroughtheuseofcolligatoryconceptsandthroughacapacitytoseerealityretrospectively.Theanalogywithmusicisunhelpful,however,inthatitisnotclearthatmusicisaboutsomethinginthesamesensethatanhistoricalnarrativeisaboutsomething.Theinterrelationshipsthatonecancometograspinthemusicalexampledonotrefertoanythingthatstandsoutsidethepieceofmusic.Butthosewhichoneshouldcometograspinanarrativehistory,evenifonecandosoonlybythewayinwhichtheyarepresentedinthenarrativeaccount,mustbefeaturesofthatwhichthenarrativeisabout,thatis,historicalreality.Moreover,thisattempttogivethenar-rativeauniquecognitivevaluebypresentingitasawayofdisplayingthecharacterofawholeratherthanstatingit,runsintothedifficultythathistorianscananddodetachtheirconclusionsfromtheirnarrativesandarguetheirmeritsinawaythatisincom-patiblewiththesuggestionthatconclusionsareundetachablefromthenarrative.Forwecan,anddo,askwhethertwodifferentnarrativeaccountsofthesamehistoricaleventaresubstantiallyinagreementornot,and,havingallowedfordifferencesthatmightarisebecausetheyareaddressedtodifferentreaders,wecanfairlyeasilydecide.Thesedifficultiesdonotrefutetheaccountweareconsideringofwhatnarratives,invirtueofbeingnarratives,relate,buttheycastdoubtupontheclaimthatnarrativesexhibitthisinvirtueoftheirowninternalconfiguration.Thereseemsnoreasonthere-foretodenythatnarrativesareimplicitstatementsoftruthsaboutthepastwhicharedifferentinkindfromthosestatedbyanyoftheirparts,butwhichcanneverthelessbemadeexplicitshouldthatberequired.Itremainstobeasked,however,whethersomehistoricaldisagreementsareinprin-cipleirresolvable.ThesuggestionthattheyarehasbeenpowerfullydefendeduponthegroundsthatultimatelytheymaydependupondifferentMETAPHYSICAL(chapter2)orevaluativepresuppositions(Walsh1967,1986).ThiscanbeillustratedbyreferencetothedifferentwaysinwhichMarxistandWhighistorianshavewrittennarrativeaccountsofthenineteenthcentury.FortheMarxist,theunifyingthemeisthecollapse448\nPHILOSOPHYOFHISTORYofcapitalism,whilefortheWhigitisthegrowthoffreedom.Butbehindtheseobviousdifferences,thereliedifferentmetaphysicalconceptionsofthenatureofhumanbeings.FortheMarxist,humanbeingsarepre-eminentlyproducersoftheirownmeansofsubsistence,andallthatfollowsistobeexplainedinthelightoftheneedsoftherelatedtechnological,economicandideologicalsystemsthatdevelopnecessarilytoenablethistobedone(seechapter34).FortheWhig,ontheotherhand,humansareessentiallyfree,andtheunderlyingthemeofhumanhistoryisthecreationofthesetofinstitutionsthatcouldbestsatisfytheirinalienablerighttoliberty.Hence,wheretheMarxisthistorianseesthedevelopmentofso-called‘liberal’institutionsinthenineteenthcenturyascon-cealedexpressionsoftheneedsofcapitalisminaparticularphase,theliberalseesthemasthecollectiveconstructionsofpeoplebecomingevermoreawareoftheirinalienablerights(Mandelbaum1971).Theseare,ofcourse,incompatiblemetaphysicalviewsandthereisnowayofresolvingthemotherthanbymetaphysicaldiscussionofthesortinwhichhistoriansdonotnormallyengage.Itcouldbearguedthatifthisisso,theyoughttoengageinit.Butthisistoooptimistic,fordisagreementcouldberemovedonceandforallonlyifthepossibilityexistsofestablishingthetruthofsomemetaphysicalsystemonceandforall.Butthisseemsanimpossibledream.Intheend,therefore,claimstotheobjectivityoflarge-scalehistoricalaccountsmustberelativizedtothemetaphysicalper-spectiveuponwhichtheyrest.Butalthoughthisimposesanultimatelimituponthekindofobjectivitytowhichhistorycanaspire,itdoesnotshowthathistoricalknowledgeisanylessreputablethananyotherkindofknowledge.Forthereisnobranchofknowl-edgewhoseconclusionsarenotsubjecttothesamequalification.Thefactthatthereappearstobelessdisagreementin,say,thenaturalsciencesthaninhistory,issimplyaconsequenceofthefactthatthetheoriesofnaturalscienceinvolveagreedassumptionsaboutthenatureofreality,butnotthattheyinvolvenoneatall,whereashistoriansoftendonotagreeaboutatleastsomeoftheirassumptions.5ConclusionHistorianslivewithin,sharetheinterestsof,andwritefor,specifichistoricalsocieties.Theyseekobjectivetruthbut,ifthemainelementsinthepreviousdiscussionarecorrect,thedegreeofobjectivitythatcanbeclaimedfortheirfinalaccountsmustberelativizedinvariousways:totheself-perceptionoftheirsocieties,totheirownimplicitmetaphysicalassumptionsandalso,perhaps,totheirevaluativestandards.Butsuchrelativizationdoesnotmeanthattheyarenotwarrantedwaysofthinkingofthepast.Itcouldonlymeanthisifthehistoriancouldescapethetemporalandconceptualcon-straintsofhisorhersituationtoachievesomeGod’s-eyeviewpointfromwhichtoseethewhole.Butitisnotclearthattheideaofsuchawholeandsuchaviewpointisintel-ligible.Thefactthathistoricalknowledgemustbefromaperspectivecannotthereforevitiateit.Itdoes,however,putanendtotheoldhopeofaonce-and-for-alluniversalhistory,forhistoricalknowledgeisnotsimplyaggregativeinthewaythatsuchahoperequires.Thepresentisnotsimplyanadditiontothepastbutadevelopmentofit;and,asnewpresentscomeintobeing,somustnewtemporalstandpointsfromwhichwecandiscern,iftheyexist,newrelationshipsbetweendifferentpartsofthepastorbetweenpresentandpast,requiringnewcolligatoryconcepts.Butthisalonewouldnot449\nLEONPOMPAmeanthatwhatcannowbediscernedmustbefalsefor,aswehaveseen,colligatoryconceptscannotbecontradictory.Theultimatereason,therefore,whyanyaccountasawholemustbethoughtrevisableliesinthefactthattheperspectivefromwhichitoperatesinvolvesrevisablemetaphysicalassumptions.FurtherReadingTherearethreeexcellentgeneralintroductions.Gorman(1992)isexplicitlywrittenasanintroductiontoanalyticphilosophythroughphilosophyofhistory,butincludesanaccountofeconometrichistorywhichIhavehadtoomithere.Walsh(1967)remainsaclearandhelpfuldiscussionofsomeoftheaboveissues.Dray(1993)ismorecomprehensiveandmoredetailed,containinganexcellentassessmentofthevarietiesofcausaljudgementinhistory,whichIhaveagainhadtoomit,whilealsoputtingthecasefortheineliminabilityofvaluejudgementsatitsstrongest.Foranexcellent,extendeddefenceofCollingwood’sapproachtohistoricalunderstanding,seeDray(1995).McCullagh(1984),whiletoodifficulttobeanintroduction,hasexcellentchaptersonhistoricalgeneralizations,historicalcausesandcausalsignificance,togetherwithawealthofexamples.ConstructionismretainsitsappealandreadersshouldconsultGoldstein(1976)andtheessaysinHistoryandTheory,1977,4,forthestartofthedebateand,forlaterdevelopments,MichaelKrausz’s‘Historyanditsobjects’inTheMonist,74.TheproblemofinterpretationhasreceivedgreaterattentionbyEuropeanwriters.Foracomprehensivebutdifficultdiscussion,leadingtoresultsdifferentfromthoseoutlinedhere,Gadamer(1975)isindispensable.Thescientificconceptionofcausalexplanationhasremainedimportant.Foradevastatingattackonthecoveringlawmodelitselfandanimpressiveargumentforitsreplacementbyamorecomparativeconceptionofcausalexplanation,whichneverthelessremainsnaturalistic,seeMiller(1987).Foracompletelyanti-positivistview,seeHartandHonoré(1985).TothewritingsonUnderstandingdiscussedinthechaptershouldbeaddedVonWright(1971)andWinch(1958).Theontologicalandepistemologicalimplicationsofnarrativecontinuetobeasourceofmuchdispute.Forbothsidesofthedebate,seethearticlesinHistoryandTheory,1986,4,and1987,4.Forastrongdefenceofitsepistemologicaladequacy,andanequallystrongdenialofit,seeCarr(1986)andAnkersmit(1983),respectively.ReferencesAnkersmit,F.R.1983:NarrativeLogic.TheHague:MartinusNijhoff.Berlin,I.1966:TheConceptofScientificHistory.InW.H.Dray(ed.)PhilosophicalAnalysisandHistory.NewYork:HarperandRow.Carr,D.1986:Time,NarrativeandHistory.Bloomington:UniversityofIndianaPress.Collingwood,R.G.1993:TheIdeaofHistory,revdedn.Oxford:OxfordUniversityPress.Danto,A.C.1965:AnalyticPhilosophyofHistory.Cambridge:CambridgeUniversityPress.Donagan,A.1966:ThePopper–HempelTheoryReconsidered.InW.H.Dray(ed.)PhilosophicalAnalysisandHistory.NewYork:HarperandRow.Dray,W.H.1957:LawsandExplanationinHistory.Oxford:OxfordUniversityPress.——1980:PerspectivesonHistory.London:RoutledgeandKeganPaul.——1993:PhilosophyofHistory,2ndedn.EnglewoodCliffs,NJ:Prentice-Hall.——1995:HistoryasRe-enactment:R.G.Collingwood’sIdeaofHistory.Oxford:OxfordUniversityPress.Gadamer,H.-G.1975:TruthandMethod.London:SheedandWard.Gallie,W.B.1964:PhilosophyandtheHistoricalUnderstanding.London:ChattoandWindus.450\nPHILOSOPHYOFHISTORYGoldstein,L.J.1976:HistoricalKnowing.Austin:UniversityofTexasPress.——1977:HistoryandthePrimacyofKnowing.HistoryandTheory,16,4.GormanJ.L.1982:TheExpressionofHistoricalKnowledge.Edinburgh:EdinburghUniversityPress.——1992:UnderstandingHistory:AnIntroductiontoAnalyticalPhilosophyofHistory.Ottawa:UniversityofOttawaPress.Hart,H.L.A.andHonoré,T.1985:CausationandtheLaw,2ndedn.Oxford:ClarendonPress.Hempel,C.G.1959[1942]:TheFunctionofGeneralLawsinHistory.InP.Gardiner(ed.)Theo-riesofHistory.NewYork:FreePress.——1974:ReasonsandCoveringLaws.InP.Gardiner(ed.)ThePhilosophyofHistory.Oxford:OxfordUniversityPress.Krausz,M.1974:HistoryanditsObjects.TheMonist,74,2.Lovejoy,A.O.1959:PresentStandpointsandPastHistory.InH.Meyerhoff(ed.)ThePhilosophyofHistoryinOurTime.NewYork:Doubleday.McCullagh,C.B.1984:JustifyingHistoricalDescriptions.Cambridge:CambridgeUniversityPress.MacIntyre,A.1981:AfterVirtue.NotreDame,IN:UniversityofNotreDamePress.Mandelbaum,M.1971:History,ManandReason:AStudyinNineteenth-CenturyThought.Baltimore,MD:JohnsHopkinsUniversityPress.Martin,R.1977:HistoricalExplanation:Re-enactmentandPracticalInference.Ithaca,NY:CornellUniversityPress.Miller,R.W.1987:FactandMethod.Princeton,NJ:PrincetonUniversityPress.Mink,L.O.1970:HistoryandFictionasModesofComprehension.NewLiteraryHistory,I.——1978:NarrativeFormasaCognitiveInstrument.InR.H.CanaryandH.Kozicki(eds)TheWritingofHistory:LiteraryFormandHistoricalUnderstanding.Madison:UniversityofWiscon-sinPress.——1987:HistoricalUnderstanding(editedbyB.Fay,E.O.GolobandR.T.Vann).Ithaca,NY:CornellUniversityPress.Nowell-Smith,P.1977:TheConstructionistTheoryofHistory.HistoryandTheory,16,4.Oakeshott,M.1933:ExperienceandItsModes.Cambridge:CambridgeUniversityPress.Pompa,L.1993:ThePossibilityofHistoricalKnowledge.SupplementaryProceedingsoftheAristotelianSociety.Russell,B.1921:TheAnalysisofMind.London:GeorgeAllenandUnwin.VonWright,G.H.1971:ExplanationandUnderstanding.London:RoutledgeandKeganPaul.Walsh,W.H.1967:AnIntroductiontoPhilosophyofHistory,2ndedn.London:Hutchison.——1986:FactandValueinHistory.InM.C.DoeserandJ.N.Kraay(eds)FactsandValues.Dordrecht:MartinusNijhoff.White,H.1973:Metahistory:TheHistoricalImaginationinNineteenth-CenturyEurope.Baltimore,MD:JohnsHopkinsUniversityPress.——1978:TheHistoricalTextasLiteraryArtefact.InR.H.CanaryandH.Kozicki(eds)TheWritingofHistory:LiteraryFormandHistoricalUnderstanding.Madison:UniversityofWiscon-sinPress.White,M.1965:FoundationsofHistoricalKnowledge.NewYork:HarperandRow.Winch,P.1958:TheIdeaofaSocialScience.London:RoutledgeandKeganPaul.Young,G.M.1936:VictorianEngland:PortraitofanAge.Oxford:OxfordUniversityPress.DiscussionQuestions1Isitlogicallypossiblethattheworldcameintoexistencefiveminutesago?Ifitis,whatimplicationswouldithaveforourknowledgethattherewasarealpastmorethanfiveminutesago?451\nLEONPOMPA2Isthepastinwhichhistoriansbelieveapastthatcouldexistinitsownright,orcanitexistonlyasaconstructionfrompresenthistoricalevidence?3Whatisapast-entailingexpression?Whatisitsfunctioninenablingustoinferfromhistoricalevidencetohistoricalstatements?4Canwejustifyourbeliefthatweareinpossessionofatheoryofinterpretationwhichgivesusaccesstohistoricaltruths?5Isthecovering-lawtheoryadequateasanaccountofthelogicalformofhistori-calexplanations?6Cantherebelawsaboutthebehaviourofindividuals?Iftherecan,howwouldthisaffectthecharacterofhistoricalexplanations?7Cantherebelawsaboutthedevelopmentofhistoricalroles?8Ishistoricalunderstandingdifferentinkindfromhistoricalexplanation?Ifitis,whatdoesitenableustograspwhichhistoricalexplanationsdonot?9Arehistoricalexplanationsdeterministic?10Howcanthetransmissionofinfluenceinhistorybemadeintelligible?11Whatishistoricism?Doesitrendertheunderstandingofthepastinitsowntermsimpossible?12Isthereanincompatibilitybetweenunderstandingthepastasacontrasttothepresentandunderstandingitinrelationtothepresent?13Whatarethelogicalfeaturesofnarrativesentences?Whatistheirfunctioninhis-toricalaccounts?14Ifhistoricalaccountsdependuponvaluejudgements,howdoesthisaffecttheirtruth?15Musthistoriansmakejudgementsofimportancewhendecidingwhattoincludeintheiraccounts?Iftheymust,onwhatwouldthesejudgementsdepend?16Eventscanbecandidatesforspecifickindsofhistoryonlyiftheyarecharacter-izedincertainways.Onwhatdoestheircharacterizationdepend?17HowwouldahistoryofpoliticsbeaffectedbythejudgementthatakillingbytheIRAwasacentralcaseofanexecutionratherthanamurder?18Musthistoriansappealtotheirownreligiousbeliefswhendecidingwhatmightbeappropriateforinclusioninahistoryofreligion?Whatwouldbetheconsequencesforhistoriesofreligioniftheymust?19Mustnarrativehistoriesimposeonthepastmeaningsthatitcannotpossess?20Whatisacolligatoryconcept?Doesitintroduceaspecificallyhistoricalformofunderstandingintohistory?21Howdohistoricalnarrativesrelatetowhattheyareabout?22Aretheresomedisagreementsinhistorythatcannotberesolved?Ifso,whatisthesourceoftheirirresolvability?23Canhistoricalaccountsreachadegreeoftruthsuchthatallfuturehistoriesmerelyaddlatertruthstothem?Ifnot,howdoesthefutureofrealityaffectthem?452\n15PhilosophyofReligionCHARLESTALIAFERROPhilosophyofreligionexploresphilosophicalissuesthatarisefromreflectiononthenatureandtruthofreligiousbeliefandthemeaningofreligiouspractices.Ifthefieldisdefinedbroadlyenough,thenthehistoricalrecordofthebeginningofphilosophymarksthebeginningofphilosophyofreligion.Thefieldincludesphilosophicalargu-mentsforandagainstbeliefinaCreatorofthecosmos,comparativetreatmentsoftheDivine,accountsofthemeaningofreligiouslanguageandfaith,theethicalimplicationsofreligiouscommitments,therelationbetweenfaith,reason,experienceandtradition,conceptsofthemiraculous,theafterlife,thesacredrevelation,mysticism,prayer,salvationandotherreligiousconcerns.Thischapterreviewsthecurrentpracticeofthephilosophyofreligion,highlightingargumentsonthenatureofreligiousbelief,theintelligibilityofcompetingconceptionsofGodandthedebateovertheroleofevidenceinshapingreligiousconvictions,anditexaminessomerecentworkonargumentsforandagainstbeliefthatthereisaGod.Philosophyofreligionisintegrallyrelatedtometaphysics,epistemology,ethics,thephilosophyofmindandtootherareas,nottheleastofwhichisthehistoryofphilosophyitself,asvirtuallyalltheseminalhistoricalfiguresofphilosophyheldviewsthatbearonreligiousmatters.1OverviewoftheCurrentStateoftheFieldPhilosophyofreligionhasgrownsubstantiallysincethe1970stobecomeoneofthelargestfieldswithinphilosophy.Fruitfulnewworksinphilosophyofreligionhavelaunchedahostofnewjournals,societies,institutesandwebsitesdedicatedtoit.(ForasamplingoftheseresourcesseetheFurtherReadingattheendofthischapter.)Thepopularityofthefieldmaybedueinparttoincreasedawarenessofthediversityofreli-gioustraditionsandcommunities.ItislesslikelythanbeforeforsomeoneintheWesttogrowupwithonlyonereligiousalternativetosecularlife.Itisthereforequitenaturalformanytoemploythetoolsofphilosophyintheirinvestigationandassessmentofdif-ferentreligiousandsecularvalues.PhilosophyofreligionisthesecondmostrequestedphilosophycoursebystudentsintheUnitedStates.Anotherreasonforgrowthinthe\nCHARLESTALIAFERROfieldistheretreatofarestrictive,empiricaltheoryofmeaningandevidencecalledlogicalpositivism(sometimesreferredtoas‘verificationism’).Logicalpositivismhasdeeprootsinempiricism(seeHUME,chapter31),butitsepi-centreinthetwentiethcenturywasagroupofphilosopherswhometinthe1920sand1930sattheUniversityofViennacalledtheViennaCircle.Logicalpositivistsinsistedthatmuchoftraditionalmetaphysics,includingthebeliefthatthereisaGod,isinca-pableofevidentialtestingandthereforemeaningless.Thereweremanyversionsoftheempiricismpromotedbylogicalpositivism,butthefollowingempiricistprincipleisrepresentative:forapropositionalclaimtobemeaning-fulitmusteitherbeaboutthebareformalrelationsbetweenideassuchasthoseenshrinedinmathematicsandanalyticdefinitions(‘AisA’,‘trianglesarethree-sided’),ortheremustinprinciplebeperceptualexperienceprovidingevidenceofwhethertheclaimistrueorfalse.Ostensiblyfactualclaimsthathavenoimplicationsforourexperienceareemptyofcontent.Metaphysicalclaimslike‘TheAbsoluteisoutsideoftime’and‘TherearePlatonicproperties’weretherebysweptawayasnonsense.Inlinewiththisformoflogicalpositivism,A.J.Ayer(1910–89)andothersclaimedthatreligiousbeliefsweremeaningless.HowmightoneempiricallyconfirmthatGodisomnipresentorloving,orthatKrishnaisanavatarofVishnu?Inanimportantdebateinthe1950s,philosophicalargumentsaboutGodwerelikenedtodebatesabouttheexistenceandhabitsofanunob-servablegardener.Theideaofagardenerwhoisnotjustinvisiblebutwhoalsocannotbedetectedbyanysensoryfacultyseemednonsense.Usingthisgardenanalogyandotherscraftedwiththesamedesign,AntonyFlew(seehisessayinMitchell1971)madethecasethatreligiousclaimsdonotpasstheempiricaltestofmeaning.Thefieldofphilosophyofreligioninthe1950sand1960swaslargelyanintellectualbattlefieldwherethedebatescentredonwhetherreligiousbeliefsweremeaningfulorconceptuallyabsurd.Empiricalverificationismisbynomeansdead.Somecriticsofthebeliefinanincor-porealGodcontinuetoadvancethesamecritiqueasthatofFlewandAyer,albeitwithfurtherrefinements.MichaelMartin(1990)andKaiNielsen(1982)arerepresentativesofthisapproach.Despitetheseefforts,empiricistchallengestothemeaningfulnessofreligiousbeliefarenowdeemedlessimpressivethantheyoncewere.Thechargethatpositivismisitselfmeaninglesshasbeenadvancedonthegroundsthattheempiricistcriterionofmeaningitselfdoesnotseemtoinvolvetheformalrela-tionbetweenideasaswithtautologies,nordoesitappeartobeempiricallyverifiable.Howmightoneempiricallyverifywhethertheprincipleiscorrect?Atbest,theprinci-pleofverificationseemstobearecommendationastohowtodescribethosestatementsthatlogicalpositivistsarepreparedtoacceptasmeaningful.Butthen,howmightadisputeaboutwhichotherstatementsaremeaningfulbesettledinanon-arbitraryfashion?Toreligiousbelieversforwhomtalkof‘Brahman’and‘God’isatthecentrestageofmeaningfuldiscourse,theuseoftheprincipleofempiricalverificationwillseemarbitraryandquestion-begging.Iftheprincipleistighteneduptoofar,itseemstothreatenvariouspropositionsthatatleastappeartobehighlyrespectable,suchasscientificclaimsaboutphysicalprocessesandeventsthatarenotpubliclyperceptible.Forexample,whatarewetothinkofstatesoftheuniversepriortoallobservation,orphysicalstrataofthecosmosthatcannotbeobserveddirectlyorindirectlybutonlyinferredaspartofanoverridingSCIENTIFICTHEORY(chapter9)?Orwhataboutthementalstatesofotherpersons,whichmayordinarilybereliablyjudged,butwhich,454\nPHILOSOPHYOFRELIGIONsomeargue,areunderdeterminedbyexternal,publicobservation?Aperson’ssubjec-tivestates–howonefeels–canbeprofoundlyelusivetoexternalobserversandeventothepersonconcerned.Alsoworrisomewasthewholesalerejectionofethics.Arewewillingtogiveupanormativeaccountofethicsandallvaluejudgements?ThelatterquestionhadsomeforceagainstanempiricistlikeA.J.Ayer,whoregardedethicalclaimsaslackinganytruthvalueandyetatthesametimeconstruedempiricalknowl-edgeintermsofhavingtherighttocertainbeliefs.Cananethicsofbeliefbepreservedifonedispenseswithethicsaltogether?Thestrictempiricistaccountofmeaningwasalsochargedasmeaninglessonthegroundsthatthereisnocoherent,clear,basiclevelofexperiencewithwhichtotestpropositionalclaims.Theexperiential‘given’issimplytoomalleable,oftenreflectingpriorconceptualjudgementsand,onceoneappreciatestheopen-texturedcharacterofexperience,itmaybeproposedthatvirtuallyanyexperiencecanverifyorprovidesomeevidenceforanything.Amysticmightwellclaimtoexperiencetheunityofatimelessspiriteverywherepresent.A.J.Ayerallowedthatinprinciplemysticalexperiencemightgivemeaningtoreligiousterms.Thosewhoconcedethisappearedtobeonaslipperyslopeleadingfromempiricalverificationismtomysticalverificationism.Agrowingnumberofphilosopherswereledtoconcludethattheempiricistchallengewasnotdeci-sive.Criticalassessmentsofpositivismcanbefoundinworkby,amongothers,AlvinPlantinga(1967,1974,1993),RichardSwinburne(1977,1979),andJohnFoster(1985).RonaldHepburnsummarizesawidelyheldconviction:‘Therecanbenoshort-cutinthephilosophyofreligionpastthepainstakingexaminationandre-examinationofproblemsintheentirefield...Nosingle,decisiveverification-test,nosolemnDeclarationofMeaninglessness,canrelieveusofthelabour’(Hepburn1963:50).Onceverificationismwascalledintoquestion,thefloodgatesopenedandthephiloso-phyofreligioncametoincludehighlyspeculativeaccountsofGod’snature,power,knowledgeandgoodness.Inaddition,debategotunderwayinthecross-culturalphilosophyofreligion(comparingreligiousconceptionsoftheself,theafterlife,theexperienceoftheDivineandNirvana)inwaysthatwouldhaveseemedentirelyunworthyofattentionearlier.TheabandoningofstrictempiricismclearedthewayforthereturnofmanyoftheprojectsthatwereattheheartofMEDIEVAL(chapter24)orANCIENT(chapters22and23)philosophyofreligion.Anotherfactorshapingcontemporaryphilosophyofreligionandalsocontributingtoitsgrowthhasbeenthebeliefthatphilosophyneednotbepractisedinastrictlyuniformfashion,withcommonstartingassumptionsandmethodology.Asinglemodelofphilosophymadesenseinatightempiricistframeworkorinawidelyupheldratio-nalism,butitisfrequentlyarguedthatneitherofthesepositions(noranysubstitute)demandsuniversalassentbyallintellectuallyresponsibleenquirers,whichmeansthatthepathisclearforavarietyofphilosophicalprojects.Theterm‘evidence’isderivedfromtheLatinexvideremeaning‘fromseeing’and,accordingtosomephilosopherstoday,itislessthanobviousthatasecularwayofseeingrealityshouldreceiveanypriorevidentialprivilegeoverareligiouswayofseeingreality.Thismorefree-wheeling,con-textualizedtreatmentofevidencehasledsomephilosopherstodotheirworkfromwithindifferentreligioustraditions(seeBeaty(1990)asanexampleofthisapproachtothephilosophyofreligion).Thereis,forexample,ahugecontemporaryphilosophi-calliteratureondifferentviewsspecifictotheChristianfaith(topicsliketheTrinity,the455\nCHARLESTALIAFERROIncarnationandatonement)andinterfaithdialogue(philosophicalcomparisonsbetweentheconceptsofGodacrossreligions)thathascurrencyamong‘thebelievers’,butdoesnothavebuiltintoittheneedtojustifyallitsprojectstoasecular,externalsceptic.Thisoutlookreceivessomesupportfromphilosopherswhorepudiateanyover-archingtreatmentoftruthwhichisindependentofhistoricalconceptualschemes.Withphilosophyofreligionrevivified,anumberofphilosopherswhoseprimaryworkisinotherareashavebeendrawntoitinordertoarticulatetheirviews.Sodif-ferentargumentsandconceptsaboutspaceandtime,freewillanddeterminism,whichareathomeinMETAPHYSICS(chapter2),thePHILOSOPHYOFMIND(chapter5)andthePHILOSOPHYOFMATHEMATICS(chapter11),havebeenexploredinreferencetotheideaofGod–forinstance,itisnotunusualtoseecontemporarydiscussionsonthenatureoftruthtobeputintermsofadebateaboutwhethertherecanbeaGod’s-eyepointofviewofreality.2ReligiousBeliefsandReligiousFormsofLifeAnaturalwaytobeginthephilosophicalexplorationofreligionistooutlinethevariousmetaphysicalclaimsaboutthecosmosandGod,tochecktheclaimswithreferencetostandardmovesinmetaphysics(thewayonewouldargueaboutthenatureofthemate-rialworldoraboutnumbers),andthentoexaminetheclaimsintermsofepistemology(whydowethinkthemetaphysicalclaimsaretrue?),andethics(isthetheoryofvalueselaboratedinthereligioncredible?).Thiskindofphilosophicalinquirycustomarilyemploysthefollowingtermsandcategories.Theism,commontotraditionalJudaism,ChristianityandIslam,istheviewthatthecosmosiscreatedandkeptinexistencebyanomnipresent,omniscient,omnipotent,supremelygoodbeing.Itpreservessomedis-tinctionbetweenGodandthecreation,accordingtowhichthetwoarenotidenticalhoweverinterwoventheymightbe.Typically,theterm‘theism’isusedtodesignatemonotheism,thoughcontextsmayarisewhenitwillbequiteimportanttonoteexplic-itlythedifferencesbetweenmonotheism(thereisoneGod),polytheism(therearemanygods)andtritheism(therearethreeGodsorgods–thisusagecomesintoplayinphilo-sophicalworkontheTrinityinChristianity).HenotheismistherecognitionofoneGodforpurposesofdevotionandworshipwithoutdenyingtheexistenceofotherDivinebeings.Theterm‘theism’canbeusedtodaywithsomeflexibility,sothatitwouldstillbepossibletocalloneselfatheistifonewereamonotheistwhobelievedGodwasomnipresentandalltherestexceptforbeingomnipotent.Someself-described‘theists’claimGodis‘almighty’(verypowerful)withoutbeingomnipotent.PantheismistheviewthatallisGod,whilePanentheismoccupiesapositionmidwaybetweentheismandpan-theism.Forpanentheists,whileitisnotstrictlytruethateverythingisGod,everythingislodgedorembeddedwithinGod,makingthetwointerdependent.ThismetaphysicalapproachtoreligionallowsfortherelativelyneatdivisionbetweenvariousformsofbeliefintheDivineasdistinctfromatheism(fromtheGreekatheos,‘withoutgod’,atermcommonlyusedtodenytheismaswellasallothernotionsoftheDivine)andagnosticism(fromtheGreekagnosis,meaning‘withoutknowledge’anddesignatingthethesisthattheDivineisunknownor,inanextremeform,thattheDivinecannotbeknown).456\nPHILOSOPHYOFRELIGIONThreeobstaclestothiswayofproceedingshouldbenoted,thefirsttwoonlybriefly.Firstly,theterm‘religion’isnotatalleasytodefineandsotheboundariesofthephi-losophyofreligionarelikewisedifficulttolayoutwithprecision.Anecessarycondi-tionforbeinga‘religion’cannotbethebeliefthatthereisaGod(TheravadaBuddhismisareligionwhich,inthemain,isatheistic).Definitionsthatrequirethepromotionofreverenceoraweforwhatisbelievedtobesacredaredifficulttousewithmuchconfi-dencebecauseofthevaguenessoftheterms‘reverence’and‘awe’,andthereisalsotheproblemthatsomereligionsmaynotenshrinethese‘religiousemotions’.Consider,forexample,religiouspracticesthatappeartotreattheDivineonlyasavehicleforensuringhealthycrops,muchlikeEuthyphro’seconomicapproachtoGodinPLATO’s(chapter23)dialogueEuthyphro,andnotetheprevalenceofmanytodaywhorepudi-ateallreligionandyetseektolivewithreverenceforlife.Debatesoversuchmatters(Iscapitalismareligion?IsMarxism?)canveryquicklyappearphilosophicallysterile.Whatfollowsofphilosophicalinterestifcapitalismisdefinedasareligion?Thewayphilosophersusuallyhandletheproblemofdefiningreligionisbynotingthemajortraditionsrecognizedtodayas‘religions’andcastingthefieldofphilosophyofreligionasenquiryintothosespecificreligionsandthetraditionsandpracticesresemblingthem.Onthisview,religionsincludeJudaism,Christianity,Islam,Hinduism,Buddhism,Taoism,Sikhism,andthosetraditionsthatresembleoneormoreofthem.Ifitisplau-siblyarguedthatasecular,atheisthumanismhasmanyreligiouscharacteristicsandshouldbeconsideredasonereligionamongtheothers,thismayhaveimportantpoliti-calandsocialconsequences(perhapsintermsofgovernmentprotectionorpersecu-tion),butsettlingthematterdoesnotseemtoinvolveveryhighstakesphilosophically.Allitmeansforphilosophersofreligionisthattheprimarytopicofwhatisconsideredtheirfieldhaswidenedsomewhat.Asecondworryisthatthesystematiccategorizingofreligionsbyregisteringtheirmetaphysicalcommitmentswillinvariablydepreciatetheenormousdiversitywithinanygivenreligioustradition.Forexample,thereareimportantfigureswithinthegreattraditionsofJudaism,ChristianityandIslamwhosometimesadheretoviewsthatseemeitheratheistic,agnostic,pantheistorpanentheist.Thereisalsoconsiderablecross-referencingtodayinwhichreligiouspractitionerswilldescribethemselvesasBuddhistChristiansandthelike.Allonecandohereissimplyexercisecaution,beingreadyheavilytoqualifyone’sdescriptionofthedifferentreligioustraditionsasoneworksatcloserangeonreligiousfiguresandmovements.AthirdobstacletoconsiderismorepotentphilosophicallyandemergeslargelyfromtheworkofLudwigWITTGENSTEIN(1889–1951)(chapter39).Inhisearlyphilosophi-caldevelopment,WittgensteindeeplyinfluencedtheempiricismoftheViennaCircle,whilehismatureworktookaverydifferentturnwithitsopen-endedapproachtothemanydifferentwaysinwhichlanguagemaybeused.FollowingWittgenstein’slaterviewsontruth,beliefandformsoflife,somephilosophersofreligionhavegonesofarastocontendthatallreligionsareprincipallymattersofpracticeandthatmetaphysi-calconcernsarelatentandmarginal.Wittgensteinlaunchedanattackonwhathasbeencalledthepicturetheoryofmeaning,accordingtowhichstatementsmaybejudgedtrueorfalsedependinguponwhetherrealitymatchesthepicturerepresentedbythebelief.Thisunderstandingoftruthandbeliefs–essentiallythecorrespondencetheoryoftruthinwhichthe457\nCHARLESTALIAFERROstatement‘Godexists’istrueifandonlyifGodexists–seemedtoWittgensteintobemisguided.Itgivesrisetoinsolublephilosophicalproblemsanditmissesthewholepointofhavingbeliefs,whichisthattheirmeaningistobefoundinthelifeinwhichtheyareused.Byshiftingattentionfromthereferentialmeaningofwordstotheiruse,Wittgensteinpromotedtheideathatweshouldattendtowhathecalled‘formsoflife’.Asthismovewasappliedtoreligiousmatters,anumberofphilosophershaveeitherdeniedoratleastplayeddowntheextenttowhichreligiousformsoflifeinvolvemeta-physicalclaims.NormanMalcolm(1975),B.R.Tilghman(1993)andD.Z.Phillips(1970)haveallpromotedthisapproachtoreligion.Itmaybeconsiderednon-realistinthesensethatitdoesnottreatreligiousbeliefsasstraightforwardmetaphysicalclaimsthatcanbeadjudicatedinarealistmanneraseithertrueorfalse.Bytheirlights,thetraditionalmetaphysicsoftheismgotwhatitdeservedwhenitcameunderattackinthemid-twentiethcenturybyradicalempiricistsandundertheearlierphilosophicalattacksintheeighteenthcenturybyDavidHUME(1711–76)(chapter31),becausethetraditional,cognitiveoutlookprofoundlymisconstruedwhatitmeanttobereligious.Thischallenge,then,appearstoplaceincheckmuchofthewayphilosophersintheWesthaveapproachedreligion.When,forexample,LOCKE(1632–1704)(chapter29),BERKELEY(1685–1753)(chapter30),Hume,Butler(1692–1752)andReid(1710–96)arguedforandagainstthejustificationofbeliefinGod,metaphysicswasatthefore-front.AndthesameislargelytrueofancientormedievalphilosophicalreflectionabouttheDivine.WhenStThomasAQUINAS(1224–74)(chapter24)articulatedhisproofsforGod’sexistencehewasengagedinfullyfledgedmetaphysics.Intermsofthehistoryofthefield,theWittgensteinianstanceisclosesttoKantianism,radicallyrecastingandyetcontinuingtheKantianmoveoflocatingtheroleofreligiousbeliefintermsofethicalpractices.KANT(1724–1804)(chapter32)opposedmetaphysicalargumentssupportingreligioustraditions,claimingthatnonewassuccessful,buthethenwentontoarguethattherewerepractical,moralgroundsforactingonthebasisofareligiousviewoflifeandhistory.Severalpointscanbemadeonbehalfoftwentieth-centurynon-realism.Firstly,ithassomecredibilitybasedonthesociologyofreligion.Inthepracticeofreligionitselfitappearsthatwehavesomethingmore(onemightwellsaysomethingdeeper)than‘mere’metaphysicaltheorizing.Religionseemspre-eminentlytobefocuseduponhowwelive.Phillipshaslookedatdifferentreligiouspracticessuchasprayerandthebeliefinanafterlife,concludingthatbothareintelligibleandthatthemotivesbehindeachcanbeheldintactwithoutanyofthemetaphysicalbaggagetraditionallylinkedwiththem.PrayertoGodbyparentsfortherecoveryofachild’shealthmaybeunderstoodasanexpressionoftheiranguishandanefforttocentretheirhopeonthechild’sgettingbetterandnotasanattempttoinfluenceGod’swill.Asecondreasonthatmightbeoffered(consideredingreaterdetailinsection4)isthattheclassicalandcontemporaryargumentsforspecificviewsofGodhaveseemedunsuccessfultomanyphilosophers(thoughnottoall).Tilghmantakesthislineandarguesthatifthetraditionalargu-mentsforGod’sexistencearereinterpretedaspartofreligiouslifeandnottreatedasiftheywerescientifictruth-claims,thentheyhaveanintelligibilityandforcethattheyotherwiselack.So,ifoneexpectstoargueconvincinglytotheexistenceofaDivinedesigneronthebasisoftheappearanceofdesigninthecosmos,oneisinphilosophi-caltrouble.OneisinfarbettershapecastingreligiousreferencestoGod’sdesignofthe458\nPHILOSOPHYOFRELIGIONcosmosintermsofreligiouspracticeslikelivingwithrespectforhumanity.Itmightappearthatthisnon-realismwouldleadreligiousadherentsintosomehighlypara-doxicalpositions.DoesitmakesensefortheretobeChristianatheists,forexample?Butthisworrymightbemet,atleastinpart.Afterall,iftheWittgensteiniancritiqueholds,isnottheatheistasmuchcommittedtotheunacceptablepicturetheoryasthetheist?Ifso,thenthenon-realistcanunderstandreligiouslifeasneitheratheistnortheist.Finally,onbehalfofnon-realism,itmaybenotedhowsuchastancereceivessupportfrommorethantheWittgensteinianmovement,butalsobythecurrentrevivalofPRAGMATISM(chapter36),especiallyinNorthAmerica,andtheappropriationofsomeoftheworkofHEIDEGGER(1889–1976)(chapter41).ThereiscurrentlyarenewedinterestinJohnDEWEY’s(1859–1952)(chapter36)emphasisonpracticeovertheorizing,andwithHeidegger’sstressoncopingandactivityasopposedtoconscious,detachedreflection.Practice,then,ratherthanmetaphysicaltheoriesseemspredomi-nantinthesephilosophicaloutlooks.Non-realistviewshavetheircriticsfromthevantagepointbothofatheists,suchasMichaelMartin(1990)andtheists,suchasRogerTrigg(1989).Bywayofaprelimi-naryresponseitmaybepointedoutthatevenifanon-realistapproachisadoptedthiswouldnotmeanaltogetherjettisoningthemoretraditionalapproachtoreligiousbeliefs.Ifoneofthereasonsadvancedonbehalfofnon-realismisthatthetraditionalprojectfails,thenthetraditionalprojectwillstillneedinvestigatingand,assumingthatifithasfailedthefailureisnotoverwhelminglyobvious,constantrethinking.ThisissimilartoamoveinreplytoaKantianphilosophyofreligion.AsKantbuilthiscaseagainsttraditionalmetaphysics,inpart,onthebasisthattheoretical,metaphysicalreasonleadstocontradictorypositions(forexample,thattheworldhasanoriginandthattheworlddoesnothaveanorigin),theongoingassessmentofthetenabilityofaKantianstancerequiresrethinkingofthemetaphysicalapproachandcheckingtoseeifthecontradictionsareinfactunavoidable.AmoresubstantialreplytoWittgensteiniannon-realismhasbeenthechargethatitunderminestheveryintelligibilityofreligiouspractice.Letusconcedethatpracticeisantecedenttotheory,whichisaconcessionnotsharedbyall.Ifonehasareligiouspractice,suchasprayertoGodorBuddhistmeditationtoseethroughtheillusionofhavingasubstantialego,thedevelopmentofsomesortoftheoryofthecosmostomakesenseofthispracticeseemsinevitable.Oncesuchatheoryisinplace,itisintelligibletoraisethequestionofitstruth.WhileMalcolmhasproposedthatitmakessensetobelieveinGodwithoutbelievingthatGodexists,othershavesubmittedthatlackofbeliefthatGodexistsmakesbeliefinGodmeaningless.BeliefthatXispriortobeliefinX.Onemayhopethatsomethingwilloccur(achildrecoversfromillness)withouttheaccom-panyingbeliefitwilloccur,butitismorepuzzlingtosupposeonecantrustinachildrecoveringoraDivinebeing,withoutbelievingtherewassomerealitytheretorelyupon.Whilenon-realismmightseemtolaythegroundworkforgreatertolerancebetweenreligions(andbetweenreligionsandthesecularworld)becauseitsubvertsthebattleoverwhichreligionhasatruepictureofthecosmos,criticshavelamentedthelossofanormativewayofchoosingbetweenreligions,waysthatseemtobeusedincommonplacereflectiononthemeritsofreligion.Sotodayitisstillnotatallunusualforpeopletoclaimtheyhavechangedreligions(orstayedwiththeirownorabandonedallreligion),forreasonsliketheappealtoreligiousexperience,answeredorunanswered459\nCHARLESTALIAFERROprayer,miraclesorthelackofthem,moralandculturalrelativism,anoverwhelmingsenseoftherealityofgoodandevil,andsoon.Whilerealistsandnon-realistsareatoddsindebate,eachsidecanlearnfromtheother.Non-realistscanconsidertherealistapproachtodivineattributesandaphiloso-phyofGodasreflectionsofareligiousformoflife.AphilosophicaltreatmentofGod’sgoodnessmayrevealimportantinsightsaboutpracticalreligiousformsoflife.Ontheotherhand,thenon-realistapproachtoreligionmaycautionrealistsagainstapproach-ingreligionasameretheoretical,abstractenterprise.3DivineAttributesMostphilosophyofreligionintheWesthasfocusedondifferentversionsoftheism.Ancientphilosophyofreligionwrestledwiththecredibilityofmonotheismandpoly-theisminoppositiontoscepticismandveryprimitivenaturalisticschemes.Forexample,shouldPlato’sviewthatGodissingularlygoodbepreferredtotheportraitofthegodsthatwasarticulatedinGreekpoetictradition,accordingtowhichtherearemanygodsandtheyareoftenimperfectandsubjecttoviceandignorance?TheemergenceanddevelopmentofJudaism,ChristianityandIslamonaglobalscalesecuredthecentral-ityoftheismforphilosophicalenquiry,buttherelevanceofaphilosophicalexplorationoftheismisnotlimitedtothoseinterestedinthesereligionsandtheculturesinwhichtheyflourish.Whiletheismhasgenerallyflourishedinreligioustraditionsamidreli-giouspractices,onemaybeatheistwithoutadoptinganyreligionwhatever,andonemayfindtheisticelements(howeverpiecemeal)inConfucianism,Hinduism,somever-sionsofMahayanaBuddhism,aswellasinthereligionsofsomesmaller-scalesocieties.ThedebateovertheismalsohascurrencyforsecularhumanismandreligiousformsofatheismasinTherevadaBuddhistphilosophy.Inaddition,theismmaybedeemedphilosophicallyinterestinginsofarasitprovidesanarenaforstudyingthenatureandlimitsofcertainconceptsusedtodescribehumanlife,suchashumanknowledge,power,creativityandgoodness.AlltheseareutilizedintheismwhenGodisdescribedasall-knowing,powerful,creativeandgood.Thisgeneratesmuchphilosophicalwork,analysinghowthesetermsfunctiononatranscendentlevel,notinginparticularhowouruseofthemmaybeinformedorunderminedbydifferenttheoriesofhumanlife.AsGodisalsodescribedinsometheistictraditionsasnecessarilyexisting,immutable(unchanging),eternalandincorporeal,thisinvitesphilosophicalstudyofthemeaningofsuchnotionsandthejustificationoftheensuingideaofabeingwitheitherallormostofthesesupremeattributes.ItwillbeusefulfirsttogetaclearerideaaboutthegeneralwaysinwhichthelanguageaboutGodhasbeentreatedandthentoconsidersomespecificcasesofphilosophicalreflectionontheattributesofGod.TermsappliedbothtoGodandtoanyaspectoftheworldhavebeenclassifiedaseitherunivocal(sharingthesamesense),equivocal(usedindifferentsenses)oranal-ogical.Thereisarangeofaccountsofanalogouspredication,butthemostcommon–andtheoneassumedhere–isthattermsareusedanalogouslywhentheiruseindifferentcases(Johnlimpsandtheargumentlimps)isbasedonwhatisbelievedtobearesemblance(John’slegisimpairedandtheargumentisweakinsomerespect).It460\nPHILOSOPHYOFRELIGIONseemsclearthatmanytermsusedtodescribeGodintheistictraditionsareusedanalogously,aswhenGodisreferredtoasafountain.MoredifficulttoclassifyaredescriptionsofGodasgood,personal,knowingandcreative.Heatedphilosophicalandtheologicaldisputescentreonunpackingthemeaningofsuchdescriptions,disputesthatareoftencarriedoutwiththeuseofthoughtexperiments.Inthoughtexperiments,hypotheticalcasesaredescribed–casesthatmayormaynotrepresentthewaythingsare.Inthesedescriptions,termsnormallyusedinonecontextareemployedinexpandedsettings.Thus,inthinkingofGodasomniscient,onemightbeginwithanon-controversialcaseofapersonknowingthatapropositionistrue,takingnoteofwhatitmeansforsomeonetopossessthatknowledgeandofthewaysinwhichtheknowledgeissecured.Atheisticthoughtexperimentwouldseektoextendourunderstandingofknowledgeaswethinkofitinourowncase,workingtowardtheconceptionofamaximumorsupremeintellectualexcellencebefittingthereligiousbelievers’understandingofGod.Variousdegreesofrefinementwouldthenbeinorder,asonespeculatesnotonlyabouttheextentofamaximumsetofpropositionsknownbutalsoabouthowthesemightbeknown.Thatis,inattributingomnisciencetoGod,wouldonetherebyclaimGodknowsallinawaythatisanalogoustothewaywecometoknowtruthsabouttheworld?Toocloseananalogywouldproduceapecu-liarpictureofGodrelyingupon,forexample,induction,sensoryevidenceorthetesti-monyofothers.OnemoveinthephilosophyofGodhasbeentoassertthattheclaim‘Godknowssomething’employstheword‘knows’univocallywhenreadaspickingoutthethesisthatGodknowssomething,whileitusestheterminonlyaremotelyanalogicalsenseifreadasidentifyinghowGodknows(Swinburne1977).Hereamedievaldistinctioncomesintoplaybetweentheressignificata(whatisasserted–forinstance,thatGodknowsX)andmodussignificandi(themodeormannerinwhichwhatissignifiedisrealizedorbroughtabout–forinstance,howGodknowsX).Wemighthaveagoodgraspofwhatismeantbytheclaimthatabeingisomni-scientwhilebeingscepticalaboutgraspinghowabeingmightbeso.Thoughtexperi-mentsaimedatgivingsomesensetotheDivineattributeofomnisciencehavebeenadvancedbydrawingattentiontothewayweknowsomethingsimmediately(bodilypositions,feelingsandintentions),andthenbyextendingthis,coaxingusintocon-ceivingabeingthatknowsallthingsaboutitselfandthecosmosimmediately(seeBeaty(1990)foraconstructiveview,andBlumenfeldinMorris(1987b)forcriticism).Utilizingthoughtexperimentsandlanguageinthisway,philosophicaltheologyhasastakeinthesoundnessandrichnessoftheimagination,picturingthewaythingsmightbe‘inone’smind’seye’,whetherornotthisreliesonanyactualimagery.Philoso-phersarenowmorecautiousaboutdrawingsuchinferencesasweareincreasinglyawareofhowsomefeaturesofanimaginedstateofaffairsmightbemisconceivedoroverlooked.Evenso,ithasbeenpowerfullyarguedthatifastateofaffairsappearstoonetobepossibleaftercarefulreflection,checkingitagainstone’sbackgroundknowl-edgeinotherareas,thenthereisatleastsomewarrantinjudgingthestateofaffairstobeabonafidepossibility(Sorensen1992).UnderstandingthecurrentdebateoverthelimitsofourthinkingaboutGodandtheDivineattributescanbesharpenedbydistinguishingbetweenthreeimportantcases:acaseofclaimingtoconceiveofapossiblestateofaffairs,acaseinwhichoneconceivesoftheimpossibilityofastateofaffairs,andacaseinwhichonesimplyfailstoconceiveofastateofaffairs,perhaps461\nCHARLESTALIAFERROaccompaniedbyanargumenttotheeffectthatthestateofaffairscannotbeconceivedregardlessofwhetheritispossibleorimpossible.AstrongatheiststrategyistoarguethateventhebestcandidatesforconceivingofGodrepresentsomethingimpossible;theideaofGodisasincoherentastheideathatoneequalsthree.Notallatheiststakethistack,however;someallowthatwhileitispossiblethatGodexists,Goddoesnotexist.SometheistshavebeenfairlyagnosticwhenitcomestoconceivingofGod’sverybeing,assumingthatvirtuallyalllanguageaboutGodishighlyanalogicalatbestandthatunivocallanguageisappropriatechieflywhenmakingclaimsaboutwhatGodisnot(forinstance,thatGodisnotamaterialobject,notwickedorcontingent).TheideaofGodcanbeapproachedthroughtheuseofimagination(theviaimaginativa)andvariestotheextentthatitemployspositiveattributions(theviapositiva)ornegative(thevianegativa).WhileBrianDavies(1993)adoptsamorequalified,vianegativaphilosophyofGod,Swinburne(1977)adoptsatheismthatismoreinformedbytheviapositiva.Asthesestrategiesbecomemoreexplicititbecomeseasiertonotethewaysinwhichphilosophyofreligionintersectswithotherareasofphilosophy.Considerthiswithrespecttothephilosophyofmind.Accordingtocertainformsofeliminativematerial-ism,theverynotionofpossessingsuchthingsasbeliefs,desiresorloverestsonconfu-sions;onlyphysicalobjects,statesandentitiesexist.Adoptingthisradicaleliminativismwillprobablyleadtoscepticismabouttheintelligibilityoftherebeinganall-knowing,loving,non-physicalbeinginlinewithclassicaltreatmentsofGod.If,ontheotherhand,oneadoptsadualistunderstandingofthehumanpersonaccordingtowhichwearenon-physicalminds,soulsorpersonswhoaremateriallyembodied,thisscepticismwillbetempered.Supposingthatthereisanon-physicalsuprememindwillnotseemintellectuallyconfusedfromtheverystart.Arangeofphilosophersattributetheirreser-vationsabouttheismtotheirscepticismaboutdualism(Kenny1979;Barnes1972)andforthisreasonasubstantialnumberoftheistshavelinkedtheirdefenceoftheismtoadefenceoftheintelligibilityofdualism(forinstance,Swinburne1977;Taliaferro1994).Sometheisticphilosophersdonotthinkoneneedgoquitesofartodefendtheintelligibilityofdualism;allthatisrequiredistodefendthepossibilitythatGodisnon-physicalandpossessesthedivineattributesregardlessofhowwehappentobecon-stituted.Onthisview,Godmightbenon-physicalandperson-likeevenifallcreatedpersonsareexclusivelyphysical.Thus,WilliamAlston(1991a)defendsthecoherenceoftheismonthebasisofaneutral,functionalistreadingofhumanpersons.Alstonthinksthatsolongaswearenoteliminativistsandwerecognizementaltermsasintel-ligibleastheyfunctioninhumanlife,theismcanbedefendedwithouthavingrecoursetodualism.Suchthree-waydebateisquitecommonbetweenthoseclaimingtohavegoodreasontobelievethatagivenideaofGodisincoherent,thoseclaimingequallygoodreasontoconcludeitisconceptuallycoherent,andthoseclaimingthattheques-tionofGod’sexistenceisleftunresolveddespiteallsuchproandcontraargument.GodandIneffabilityBeforeconcentratingonspecificattributes,considertheprospectsforareligiouslymoti-vatedrobustagnosticismaccordingtowhichGodorultimaterealityiscompletelybeyonddescriptionorconception.StAnselm(1033–1109)heldthatGodisgreaterthanthatwhichcanbeconceived.Whatifallconceptsmustbytheirverynaturefailinmaking462\nPHILOSOPHYOFRELIGIONanyreferencetotheDivine,eveninhighlynuanced,analogousways,notbecausethereisnoGod,butbecausethereisone?ThisproposalissometimesputastheclaimthatGodisineffable(fromtheLatin,ineffabilis,for‘notbeingcapableofbeingexpressed’).TheinsistenceonGod’sineffabilityhasplayedaroleinthemysticismandthoughtofimpor-tantfiguresinthegreatmonotheistictraditions,andinBuddhistandHindutraditionsaswell.QualificationsabouttheadequacyofourconceptstorepresenttheDivinearecommoninreligiouslysympatheticphilosophicalsystems.AchievingphilosophicalproficiencyinmappingoutarobustunderstandingoftheDivinemayrequireanorderofexperiencefarbeyondwhatwenowhaveorcouldeverhave.Perhaps,asJosephAlbo(c.1380–c.1444)putitinthefifteenthcentury,‘ToKnowGod’snatureonewouldhavetobeGodHimself’.Itdoesnotseemobviouslywrongtosupposethatcertainconceptualschemesallowustoobtainonlyaverythinunderstandingofultimatereality,butprob-lemsarisewhenitisclaimedthatultimaterealityorGodarealtogetherbeyondallcon-ceptualdescriptionsnomatterhowtenuousandcircuitous.PhilosopherslikePlantinga(1980)haveobjectedthatthiskindofstrictineffabilityisincoherent,because,firstly,tobedescribedasineffableisitselfadescription;secondly,descriptionslike‘ifitis,itis’areirresistible;and,thirdly,wehavegoodreasonforthinkingtherecanbenosuchineffableX‘outthere’.Evenifthesechargesfail,itisverydifficulttoseewhatusepositinganulti-matelyineffableXwouldhave,religiously,practically,philosophicallyorscientifically.DenyingtheexistenceofXmightmakeasmuch(oraslittle)senseasaffirmingit.For-tunatelyorunfortunately,veryfewreligionsruleoutinprincipleallconceptualizingofGod,eveniftheyarequitewaryofsuchtheorizing,andsophilosophersofreligioncanstillengageinconceptualanalysiswithouthavingtoadmitfromtheoutsetthattheyarealreadycutofffromthereligionstheyareattemptingtoexplorephilosophically.Inproceedingtotreatdivineattributessomeselectivityisinorder.Ishallcoverthosethathaveoccasionedthemostintensephilosophicalscrutinyinrecentyears.3.1OmnipotenceSomeofthemostvigorousphilosophicaldebatehascentredonwhetheritmakessensetothinkofGodasasupreme,maximallypowerfulbeingresponsiblefortheoriginandcontinuedconservationofthecosmos.Beginwithaclaimlike‘Godcandoanythingwhatever’or‘AllthingsarepossibleforGod’andoneisboundtorunintoobstaclesthatmaybeclassifiedasexternalorinternal.Externaldifficultieswouldhinderanybeingwhateverfrombeingomnipotent,whileinternalproblemswouldpreventusfromrecognizingasomnipotentGodasconceivedinthethreemainmonotheistictraditions.Considerexternalproblemsfirst.Theexternalbarrierheremightbepurelylogicaloritmaytakeonbroadermeta-physicaldimensions.Thereareproblemswithsupposingthatanythingoranyonecandowhatislogicallyimpossible.CanGodmake2+2=6?Itwouldappeartobenon-sensetosupposethatthiscouldbedonebyanybeing.Theproblemhasanadditionalmetaphysicaledgeifoneassumesthattherearecertaintruths,suchas‘timeisone-directional’,whicharenecessarybutnotduetobarelogicalrelationships(itsdenialdoesnotinvolveacontradiction).Grantthattherearenecessarymetaphysicaltruths463\nCHARLESTALIAFERROanditseemsthatnothingcanbeomnipotent,becausenothingcanaltersuchmeta-physicallynecessarytruths.Theseproblemsareoftentreatedatonestrokebyinsistingthatwhatisorshouldbemeantbytheclaimthatthereisanomnipotentbeingisthatifabeingisomnipotent,itcandoanythingthatislogicallyormetaphysicallypossible.Thisapproachhastheimplicationthatthemathematicalandtime-bendingtasksmentionedaboveareidlepseudo-tasksorabsolutenonsense.Thisisnoworsethan‘con-ceding’thatanomnipotentbeingcannotmaketheconceptofjusticejumparoundthenumber7oreathappiness.ThisstrategyofdefiningthepossibilityofDivineomnipotenceisfairlycommonandhasantecedentsgoingbacktomedievalphilosophyofreligion.Itissometimesrefinedtotakeintoaccountdifferentlevelsandsequencesofaction.ConsiderthequestionofwhetherGodcouldmakesomethingthatGodcouldnotcontrol.Itwouldappearso,givenacertainconceptionoffreedomofthewill.Arguably,abeingisfreeonlyifitisnotcompletelyunderthecontrolofanotherbeing.IfGodcreatesafreecreature,hasGodceasedtobeomnipotentbecausenowthereisabeingGodcannotcontrol?Presumably,thesesortsof‘limits’onDivinefreedomcanbecastintermsofGod’sbeinglimitedbythelimitsGodintroduces.Toexpectanomnipotentbeingtobeabletomakeacreaturebothfreeandnotfreeatthesametimewouldbeakintoexpectingthatanomnipotentbeingcouldmakesomethingthatissquarewhilelackingfourrightangles.Canweenvisageamaximalpowerthatisdeeperthan,orbehind,thelawsoflogic?Perhapsthereissomeroomtomoveifoneadoptsananti-realisttreatmentofabstractobjectsandofLOGIC(chapter4),accordingtowhichlogicallawsaremerelythecodifi-cationofhumandesignsandlanguage.Alternatively,onemightfollowDESCARTES(1596–1650)(chapter26)whothoughtthatGodwasresponsibleforthecreationofnecessarytruths.Thishasnotbeenwidelyacceptedbytheistsonthegroundsthatthe-isticvoluntarismwouldunderminethenecessityoflogicalandmetaphysicaltruths.IfGodcouldmakeA=A,thenGodwouldhavethepowertomakeÿA=Aandthusthelawofidentitywouldbecontingentandnotnecessary.AgroupofcontemporaryphilosophershassoughttobolsterDescartes’sstance,however.Plantinga(1974),MenzelandMorris(1986)andothersproposepreservingthenecessityoflogicalandmetaphysicaltruthsandyetinsistthattheyaretrueinvirtueofGod’screativewill,acreativewillwhichcouldnotbeotherwise.ThenecessityofA=AandothersuchtruthsistherebyaffirmedandaccountedforbyadeeperDivinenecessity.ThisrevivalofDescartes’spositionmaybearticulatedusingtheconventionoftalkingofpossibleworlds.Apossibleworldmaybesaidtobeawaythingscouldbe.Theactualworldisonepossibleworld;inadifferentpossibleworldyouwrotethischapterandIdidnot.Usingthisterminology,onecouldsaythat,inallpossibleworlds,GodwillsthatA=A,andsoonforallnecessarytruths.ThisrevisedversionofDescartes’sstancehasbeencommendedonthegroundsthatitpreservesthedesirednecessities,thatitofferssomeaccountofthetruthsatissueratherthanleavingthenecessitiessuspendedinPlatonicstrata,andthattheexplanationitoffershasgreatersimplicitythanrivalaccounts.Isthereanultimatelimittowhat,inprinciple,wecouldcredittoamaximallypow-erfulbeing?Couldtherebeabeingthatcouldtakecreditevenforitsownexistenceandconstitution?CouldGodhavecreatedGod?Theideathatanybeingcouldcreateitself(causasuimeans‘causeofitself’)hasbeendismissedonthegroundsthattocreateitself464\nPHILOSOPHYOFRELIGIONabeingwouldhavetoexistalready,whichisanabsurdity;butithasrecentlybeensuggestedthatifcreatingitselfdidnotrequireafirstmomentofexisting,theprospectforcausasuimightimprove.OneproposalbyMenzelandMorrisisthatinanimportantrespectGodcouldbelikeamachine.Imagineamachine–theMenzel–Morrismachine–thatcreatesmatterfromnothing.Ithasalwaysexistedandhascontinuouslyreplacedallitspartssystematically,notallatonce(inwhichcaseitwouldhavetoceasetobeandthencreateitself)butincrementally(forinstance,onMondaysitwouldreplaceitsrightside,onTuesdaysitsleft).Witheverycycleofreplacement,woulditnothavecreateditself?ThestandardmoveinthephilosophyofGodistoresistthesepos-sibilities(perhapsitisabsurdtothinkofGodasamachineorhavingpartsorbeingintime),andtothinkinsteadthat,ifthereisaGod,God’sbeingGodisnotsomethingGodcouldhavebroughtaboutoraltered.PerhapsGoddidnotmakeGodomnipotent,butthatdoesnotmeanGodisharnessedbysomethingexternaltoGodorissomehowdefective.Considernowsomeinternalobstaclestoimaginingomnipotence,especiallyastheseareborneoutinthemonotheistictraditions.Godisbelievedtobenon-physical.Ifso,canGodremainGodandyetswim?Godisthoughtofasexistingnecessarily.Ifso,canGodceasetobe?Godisthoughtofasnecessarilygood.DoesthismeanthatGodcouldnotbringaboutanevilstateofaffairs,withouttherebeingsomeoverridingethicalreasontodoso?WouldthismeanthattherecouldbesomethingmorepowerfulthantheGodofthesereligioustraditions,namelyabeingthatcouldswim,ceasetobe,commitevilactsandsoon?SomewaysofhandlingthesepuzzlesincludeproposingthatGodcandosuchthingsasswim(Godcanassumeabody),bringaboutevilandsoon.Alternatively,onecanconcedethatGodcannotdothesethings,butdenythatomnipotencerequiresdoinganyactpossibleforsomeconceivablecreature;ratheranomnipotentbeinghasmaximalpower,suchthattherecanbenobeingmorepowerful.PerhapsGodcannotcommitsuicide,butanomnipotentGodwouldnonethelesspossessascopeofpowerthatcannotbesurpassedbybeingscapableofself-annihilation.Anotherresponsetotheseparadoxeshasbeentore-examinetherelationbetweenDivineattributesandthemotivationforascribingthemtoGod.Schlesinger(1988)andMorris(1987a),amongothers,havecontendedthattherationalebehindascribinggreat-makingqualitiestoGodistomakeexplicittheideathatGodisperfect.Whatoneisafter,then,islocatingthosepropertiesthat,together,formthegreatestcompossiblesetofproperties.Onthisschema,twopropertiesarecompossibleiftheycanbothbeinstantiatedsimultaneouslyofthesamebeing.IfGod’sbeingperfectlygoodinanywaylimitsGod’sbeingall-powerful,itdoessoonlyinafashionthat,overall,contributestotheexcellenceofGod.Godwouldbelessgreat,lessworthyofworshipifGodcouldcommitevil.Thisshiftingofattentionfrombarepowertothinkingofpowerintermsofvalueshasbeenheraldedbysomefeministandsocialcriticsofreligionasahealthydeparturefromamoremale-basedcelebrationofapowerthatonlytooquicklyturnsintooppressive,suffocatingpicturesoftheDivine.TheMorris–Schlesingerstrategyisinthesamespiritofearliertactics,suchasthatofBoethius(c.480–524)inthesixthcentury,whoconcludedthattheabilitytocommitevilwasfarfrombeingapositivepowerorsignofgreatness.InhisdialogueTheConsolationofPhilosophy(Boethius1969)thereisthefollowingexchange:465\nCHARLESTALIAFERRO‘ThereisnothingthatanomnipotentGodcouldnotdo.’‘No.’‘Then,canGoddoevil?’‘No.’‘Sothatevilisnothing,sincethatiswhatHecannotdowhocandoanything.’3.2OmniscienceThebeliefthatGodisall-knowinghashadavitalroleinahostofreligiousconvictionsaboutGod’sroleasthecreatorofthecosmosandrelianceuponGod’swisdom,loveandprovidence.Philosophically,omniscienceistypicallydefinedintermsofknowingallpropositions,or(withsomeelaboration)knowingwithrespecttoalltruepropositionsthattheyaretrueandwithrespecttoallfalsepropositionsthattheyarefalse.Someofthedebateonomnisciencestemsfromphilosophicaldebateoverthenatureofpro-positions.Somepropositionscontainindexicalterms(termshavingtheirreferencedeterminedaccordingtowhichpersonisusingthem).Ifyouthinkthatsometruepropositionscontainingfirst-personindexicalterms,like‘Iaminpainnow’,canonlybeknownbytheindividual‘I’whoisentertainingthem,thenyouwouldnotbeabletocastomniscienceintermsofknowingalltruepropositions.Onthisview,indefinitelymanypropositionscouldbetruebutknownonlybyparticularindividuals.MyknowingthatSmithisinpainwouldinvolvemyknowingadifferentpropositiontobetruethanthatknownbySmith.Heknows‘Iaminpain’andIknow‘Smithisinpain’.Thispuzzleissometimesdealtwithbydenyingtheexistenceofsuchhyper-privatepropositions(Kvanvic1986).Moresustaineddebate,withlesspromiseofresolution,hasoccurredoverthescopeofomniscience.OmniscienceandFreedomAreomniscienceandfreedomcompatible?ImaginethereisaGodwhoknowsthefuturefreeactionofhumanbeings.IfGoddoesknowyouwillfreelydosomeactA,thenitistruethatyouwillindeeddoA.Butifyouarefree,wouldyounotbefreetoavoiddoingA?GiventhatitisforeknownyouwilldoA,itappearsyouwouldnotbefreetorefrainfromtheact.Initiallythisparadoxseemseasytodispel.IfGodknowsaboutyourfreeaction,thenGodknowsthatyouwillfreelydosomethingandthatyoucouldhaverefrainedfromit.God’sforeknowingtheactdoesnotmakeitnecessary.Doesnottheparadoxonlyarisebecauseweconfusetwopropositions:‘Necessarily,ifGodknowsX,thenX’with‘IfGodknowsX,thennecessarilyX’?AUGUSTINE(354–430)(chapter24),Boethius,Anselm,AquinasandotherssoughttopreservetherealityoffreedomalongwithGod’sforeknowledge,andthisisastancewidelyrepresentedtoday.Theproblemisretained,however,whenthepointispressedconcerningthegroundsforforeknowledge.IfGoddoesknowyouwillfreelydoX,thenitappearsthattheremustnowbeafactofthematteraboutwhatyouwillandwillnotdo,andthussomeresidualsenseinwhichyourfreelydoingXisnotsomethingthatcanbealtered.Iftheproblemisputinfirst-persontermsandoneimaginesGodforeknowsyouwillfreelyturnthenextpage,thenaneasyresolutionoftheparadoxseemselusive.ImagineGodtellsyouwhatyouwillfreelydo.Undersuchconditions,isitstillintelligibletobelieveyouhavetheabilitytodootherwiseifitisknownbyGodaswellasyourselfwhatyouwillindeedelecttodo?466\nPHILOSOPHYOFRELIGIONVariousreplieshavebeengiven,ofwhichInotethree.Someadoptwhatiscalledcom-patibilism,affirmingthecompatibilityoffreewillanddeterminism,andconcludethatforeknowledgeisnomorethreateningtofreedomthandeterminism.Asecondpositioninvolvesadheringtotheradicallylibertarianoutlookofinsistingthatfreedominvolvesaradical,indeterministexerciseofpower,andconcludesthatGodcannotknowfuturefreeaction.WhatpreventssuchphilosophersfromdenyingthatGodisomniscientisthattheycontendtherearenotruthsaboutfuturefreeactions.Priortosomeone’sdoingafreeaction,thereisnofactofthematterthatheorshewilldoagivenact.ThisisinkeepingwithARISTOTLE’s(chapter23)philosophyoftimeandtruth.Aristotlethoughtitwasneithertruenorfalsepriortoagivenseabattlewhetheragivensidewouldwinit.Sometheists,suchasSwinburne(1977),adoptthislinetoday,holdingthatthefuturecannotbeknown.Ifitcannotbeknownformetaphysicalreasons,thenomnisciencecanbereadasknowingallthatitispossibletoknow.ThatGodcannotknowfuturefreeactionisnomoreofamarkagainstGod’sbeingomniscientthanGod’sinabilitytomakesquarecirclesisamarkagainstGod’sbeingomnipotent.Otherparties,includingLinville(1993),denytheoriginalparadox.TheyinsistthatGod’sforeknowledgeiscompatiblewithlib-ertarianfreedomandseektoresolvethequandaryeitherbyclaimingthatGodisnotboundintime(Goddoesnotsomuchforeknowaseternallyknow)orbyarguingthattheuniquevantagepointofanomniscientGodpreventsanyimpingementonfreedom.Godcansimplyknowthefuturewithoutthishavingtobegroundedonthefuturesomehowbeingalreadyfixed.SomeadvocatesofthispositionholdthatGodknowsnotjustwhatwilloccur,butwhatwouldoccurgivenallpossiblefutureevents.Godknowswhetheryouwillfreelyturnthepageandalsoknowswhetheryouwouldfreelydosoifanynumberofothereventsweretooccurpriortotheact(forinstance,Godknowswhatyouwoulddoifyousawanelephantjustnow,eventhoughyouarenotseeingone).Thiskindofknowledgeofpossiblefreechoicesisreferredtoasmiddleknowledge,asitdesignatesknowledgemid-way,asitwere,betweenwhatwilloccurandwhatcouldoccur.Middleknowledgeinvolvesknowingwhatwouldhappenunderallpossibleconditions.3.3Eternaloreverlasting?Couldtherebeabeingthatisoutsidetime?Inthegreatmonotheistictraditions,Godisthoughtofaswithoutanykindofbeginningorend.Godwillnever,indeed,cannever,ceasetobe.SomephilosophicaltheistsholdthatGod’sTEMPORALITY(seechapter2)isverymuchlikeoursinthesensethatthereisabefore,duringandanafterforGod–orapast,presentandfutureforGod.ThisviewissometimesreferredtoasthethesisthatGodiseverlasting(Wolterstorff1982).ThoseadoptingamoreradicalstanceclaimthatGodisindependentoftemporality,arguingeitherthatGodisnotintimeatall,orthatGodis‘simultaneously’atorinalltimes.ThisissometimescalledtheviewthatGodiseternalasopposedtoeverlasting(Leftow1991).Whyadoptthemoreradicalstance?Onereason,alreadynoted,isthatifGodisnottemporallybound,theremaybeareso-lutiontotheproblemofreconcilingfreedomandforeknowledge.AsAugustineputitinTheCityofGod:‘ForHedoesnotpassfromthistothatbytransitionofthought,butbeholdsallthingswithabsoluteunchangeableness;sothatofthosethingswhichemergeintime,thefuture,indeed,arenotyet,andthepresentarenow,andthepast467\nCHARLESTALIAFERROnolongerare;butallofthesearebyHimcomprehendedinHisstableandeternalpresence’(Augustine1972:XI,21).IfGodisoutsidetime,theremayalsobeasecurefoundationexplainingGod’simmutability(changelessness),incorruptibilityandimmortality.Furthermore,theremaybeanopportunitytouseGod’sstandingoutsideoftimetolaunchanargumentthatGodisthecreatoroftime.ThoseaffirmingGodtobeunboundedbytemporalsequencesfaceseveralpuzzleswhichInotewithouttryingtosettle.IfGodissomehowatorinalltimes,isGodsimul-taneouslyatorineach?Ifso,thereisthefollowingproblem.IfGodissimultaneouswithRomeburningin410,andalsosimultaneouswithyourreadingthisbook,thenitseemsthatRomemustbeburningatthesametimeyouarereadingthisbook.Adifferentproblemariseswithrespecttoomniscience.IfGodisoutsideoftime,canGodknowwhattimeitisnow?Arguably,thereisafactofthematterthatitisnow,say,midnighton1July1996.AGodoutsideoftimemightknowthatatmidnighton1July1996certainthingsoccur,butcouldGodknowwhenitisnowthattime?TheproblemisthatthemoreemphasisweplaceontheclaimthatGod’ssupremeexistenceisinde-pendentoftime,themoreweseemtojeopardizetakingseriouslytimeasweknowit.Finally,whilethegreatmonotheistictraditionsprovideaportraitoftheDivineassupremelydifferentfromthecreation,thereisalsoaninsistenceonGod’sproximity.Forsometheists,describingGodasapersonorperson-like(Godloves,acts,knows)isnottoequivocate.ButitisnotclearthataneternalGodcouldbepersonal.3.4ThegoodnessofGodAllknownworldreligionsaddressthenatureofgoodandevilandcommendwaysofachievinghumanwell-being,whetherthisbethoughtofintermsofsalvation,libera-tion,deliverance,enlightenment,tranquillityoranegolessstateofNirvana.Notwith-standingimportantdifferences,thereisasubstantialoverlapbetweenmanyoftheseconceptionsofthegoodaswitnessedbythecommendingoftheGoldenRule(‘Dountoothersasyouwouldhavethemdountoyou’)inmanyreligions.SomereligionsconstruetheDivineasinsomerespectbeyondourhumannotionsofgoodandevil.InHinduism,forexample,Brahmanhasbeenextolledaspossessingasortofmoraltranscendence,andsomeChristiantheologianshavelikewiseinsistedthatGodisonlyamoralagentinahighlyqualifiedsense.TocallGodgoodisforthemverydifferentfromcallingahumanbeinggood(Davies1993).BeliefinthesupremegoodnessofGodhas,however,beenamainstayofthegreatmonotheistictraditions,andbeliefinthegoodnessofGodhasalsohadconsiderableforceindependentofthesetraditions.DisputesoverthegoodnessofGodlayattheheartoftheinitialquarrelbetweenphilosophyandpoetry.PlatowassorepelledbythemoralincoherenceofthepolytheismdisplayedinthepoetryofHesiodandHomer,withthecomplicityofthegodsininfanticide,patricide,murder,rapeandadultery,thatheproposedbanningpoetsfromhisidealrepublic.HereInotesomeofthewaysinwhichphilosophershavearticulatedwhatitmeanstocallGodgood.Intreatingthematter,therehasbeenatendencyeithertoexplainGod’sgoodnessintermsofstandardsthatarenotGod’screationandthus,insomemeasure,independentofGod’swill,orintermsofGod’swillandthestandardsGodhascreated.Thelatterviewistermedtheisticvoluntarism.Acommonversion468\nPHILOSOPHYOFRELIGIONoftheisticvoluntarismistheclaimthatforsomethingtobegoodorrightsimplymeansthatitiswilledbyGodandforsomethingtobeevilorwrongmeansthatitisforbiddenbyGod.Theisticvoluntaristsfaceseveraldifficulties:morallanguageseemsintelligiblewithouthavingtobeexplainedintermsoftheDivinewill.Indeed,manypeoplemakewhattheytaketobeobjectivemoraljudgementswithoutmakinganyreferencetoGod.Iftheyareusingmorallanguageintelligibly,howcoulditbethattheverymeaningofsuchmorallanguageshouldbeanalysedintermsofDivinevolitions?NewworkinthePHILOSOPHYOFLANGUAGE(chapter3)maybeofusetotheisticvoluntarists.Accordingtoacausaltheoryofreference,‘water’necessarilydesignatesH2O(itisnotacontin-gentfactthatwaterisH2O),notwithstandingthefactthatmanypeoplecanusetheterm‘water’withoutknowingitscomposition.Similarly,coulditnotbethecasethat‘good’mayrefertothatwhichiswilledbyGodeventhoughmanypeoplearenotawareoforevendenytheexistenceofGod?Anotherdifficultyforvoluntarismliesinaccount-ingfortheapparentmeaningfulcontentofclaimslike‘Godisgood’.ItappearsthatincallingGodgoodthereligiousbelieverissayingmorethan‘GodwillswhatGodwills’.Ifso,mustnottheverynotionofgoodnesshavesomemeaningindependentofGod’swill?Alsoatissueistheworrythatifvoluntarismisaccepted,thetheisthasthreatenedthenormativeobjectivityofmoraljudgements.CouldGodmakeitthecasethatmoraljudgementswereturnedupsidedown?CouldGodmakecrueltygood?Arguably,themoraluniverseisnotsomalleable.Inreply,somevoluntaristshavesoughttounder-standthestabilityofthemorallawsinlightofGod’simmutablyfixed,necessarynature.RecallthesimilarmoveintheCartesianstrategyofanchoringthenecessityoflogicandmetaphysicsinGod’sverybeing.Forsympathetictreatmentsofvoluntarism,seeAdams(1987)andOwen(1965).ByunderstandingGod’sgoodnessintermsofGod’sbeing(asopposedtoGod’swillalone),wecomeclosetothenon-voluntariststand.AquinasandothershaveheldthatGodisessentiallygoodinvirtueofGod’sverybeing.Allsuchpositionsarenon-voluntaristinsofarastheydonotclaimthatwhatitmeansforsomethingtobegoodisthatGodwillsittobeso.ThegoodnessofGodmaybeexplainedinvariousways,eitherbyarguingthatGod’sperfectionrequiresGodbeinggoodasanagentorbyarguingthatGod’sgoodnesscanbearticulatedintermsofotherDivineattributessuchasthoseoutlinedabove.Forexample,becauseknowledgeisinitselfgood,omniscienceisasupremegood.GodhasalsobeenconsideredgoodinsofarasGodhascreatedandconservesinexistenceagoodcosmos.Debatesovertheproblemofevil(ifGodisindeedomnipotentandperfectlygood,whyisthereevil?)havepoignancypreciselybecausetheychallengethischiefjudgementoverGod’sgoodness.Thisdebateisconsideredinthelastsection.Thechoicebetweenvoluntarismandamoreinternalistpictureisrarelystrict.Sometheistswhoopposeafull-scalevoluntarismallowforpartialvoluntaristelements.Accordingtoonesuchmoderatestance,whileGodcannotmakecrueltygood,Godcanmakesomeactionsmorallyrequiredormorallyforbiddenwhichotherwisewouldbemorallyneutral.ArgumentsforthishavebeenbasedonthethesisthatthecosmosandallitscontentsareGod’screation.Accordingtosometheoriesofproperty,anagentmakingsomethinggoodgainsentitlementsovertheproperty.ThecrucialmovesinargumentsthatthecosmosanditscontentsbelongtotheirCreatorhavebeentoguard469\nCHARLESTALIAFERROagainsttheideathathumanparentswouldthen‘own’theirchildren(theydonot,becauseparentsarenotradicalcreatorslikeGod),andtheideathatDivineownershipwouldpermitanything,thusconstruingourdutiesowedtoGodasthedutiesofaslavetoamaster(aviewtowhichnotalltheistshaveobjected).TheoriesspellingoutwhyandhowthecosmosbelongstoGodhavebeenprominentinallthreemonotheistictraditions.Platodefendedthenotion,asdidAquinasandLocke.(SeeBrody(1992)foracurrentdefenceofLocke’sversionofDivineownership.)OneothereffortworthnotingtolinkjudgementsofgoodandevilwithjudgementsaboutGodreliesupontheidealobservertheoryofethics.Accordingtothistheory,moraljudgementscanbeanalysedintermsofhowanidealobserverwouldjudgematters.Tosayanactisrightentailsacommitmenttoholdingthatiftherewereanidealobserver,itwouldapproveoftheact;toclaimanactiswrongentailsthethesisthatiftherewereanidealobserver,itwoulddisapproveofit.ThetheorycanbefoundinworkbyHume,AdamSmith(1723–90),SIDGWICK(1838–1900)(chapter35),Hare(1996)andFirth(1970).Theidealobserverisvariouslydescribed,buttypicallyisthoughtofasimpartial,omniscientregardingnon-moralfacts(factsthatcanbegraspedwithoutalreadyknowingthemoralstatusorimplicationsofthefact–forinstance,‘Hedidsomethingbad’isamoralfact;‘HehitSmith’isnot),andasomniper-cipient(Firth’stermforadoptingapositionofuniversalaffectiveappreciationofthepointsofviewofallinvolvedparties).Thetheoryreceivessomesupportfromthefactthatmostmoraldisputescanbeanalysedintermsofdifferentpartieschallengingeachothertobeimpartial,togettheirempiricalfactsstraight,andtobemoresensitive–forexample,byrealizingwhatitfeelsliketobedisadvantaged.Thetheoryhasformidablecriticsanddefenders.Iftrue,itdoesnotfollowthatthereisanidealobserver,butifitistrueandmoraljudgementsarecoherent,thentheideaofanidealobserveriscoherent.GivencertainconceptionsofGodinthethreegreatmonotheistictraditions,Godfitstheidealobserverdescription(andmorebesides,ofcourse).Shouldanidealobservertheorywork,atheistwouldhavesomereasonforclaimingthatatheistscommittedtonormative,ethicaljudgementsarealsocommittedtotheideaofaGodoraGod-likebeing.3.5OtheraspectsofthephilosophyofGodInoteverybrieflysomeadditional,recentworkinthephilosophyofGod.Withrespecttotranscendence,thereisconsiderabledebateoverwhetherGodisbestthoughtofasanindividualbeingorasBeingitself.SomeresistthinkingofGodasanindividualbecausetheythinkofindividualsintermsofonethingamongothers.Thatis,ifthereisanindividual,itmustbethoughtofaspartofagenus.But,byhypothesis,Godissuigeneris(oneofakindoragenusuntoitself)andhencenotanindividual.Alsohighonthelistoftranscendentproperties,somephilosophershaverekindledthenotionthatGodisultimatelysimple,withoutpartsandnotpossessingdistinguishableattributes.ItmakessensetodistinguishtheDivineattributes,aswehavedoneabove,butthisshouldbeseenasanartificialdevice,notreflectinganyrealdistinctionwithintheGodhead.Theappealofthisphilosophicaltheismwilldependuponone’stheoryofproperties,justastheearlierdebateonbeingandindividualwilldependsuponone’soverall470\nPHILOSOPHYOFRELIGIONontology.ForadefenceofGodasBEING(p.62)seeDavies(1993),andforDivinesimplicityseeKretzmannandStump,inMorris(1987b).OnDivineimmanence,muchofthedebatehasturnedonclaimsaboutwhetherGodisimpassable(notsubjecttopassions).Ifoneadoptsahigh,non-temporalistviewofGod’seternity,onewillbemorelikelytoaffirmtheimpassabilityofGodas,arguably,undergoingpassionrequireschange.OnemightstillallowforcertainchangeforGod,sometimescalled‘Cambridgechange’(becauseitsatisfiesthecriterionofchangesug-gestedbyRussellandothersinCambridgewithoutinvolvingrealchangeinthesubject),butthiswillnotbeasubstantialchangeonGod’sside.AnexampleofCambridgechangewouldbethechangeinJoneswhenhebecomesshorterthanSmithasaresultofSmith’sgettingtallerwhileJonesremainsthesamesize.ACambridgechangeregard-ingGodmightbeforGodtobethoughtaboutbyJonesononedaybutnotonanother.Theonly‘real’alterationisonJones’sside,notGod’s.WouldrecognizingthatGodcanundergoCambridgechangebeenoughtomakephilosophicalsenseoftheideathatGodsuffersandfeelspleasureoverthedifferentillsandgoodsofthecosmos?PerhapsGodcansufferandfeelpleasureincreation,butdoessoinsomenon-temporalfashionorpossiblyateachtimeGodisindifferentaffectivestates.ThosewhobelieveGoddoesfeelpassiontendtobelieveGodiseverlastingandthusintime.Howonedecidessuchissueswillhaveanimpactonhowoneconstrueshigher-orderDivinepropertieslikeGod’somnipresence.Thus,ifoneacceptsthatGodissubjecttopassions,God’sbeingevery-wherepresentmightinvolveafullhostofDivineproperties–forexample,Godwouldknowwhatisoccurringeverywhere,Godcouldexercisemaximalpowerwithoutlimit(except,perhaps,bythelimitsoflogicandnecessarymetaphysicaltruths),noplacewouldexistwithoutGod’sconserving,creativepower,andtherewouldbetheadditionalDivineattributeofaffectiveresponsivenesstovalues.Thus,forGodtobepresentinthecosmoswouldmean,inpart,thatGodfeelssorrowandpleasureinresponsetoitsillsandgoods.4Evidence,ReligiousExperienceandSecularExplanationsThereiscurrentlysomephilosophicaluneaseaboutwhethermetaphysicsorepiste-mologyshouldhavepriority.ModernWesternphilosophymaybeseentohavefollowedDescartesinchampioningepistemologyovermetaphysics.OnCartesiangrounds,eachindividualmuststartfromhimselforherselfandthenworkoutward,eventuallyachievingaknowledgeofGod,otherpeopleandtheexternalworld.Itisahallmarkofsomephilosophynowthatittakesdeliberateissuewiththisapproach.Thus,ithasbeenheldthatanystarting-pointofanepistemologicalkindwillalreadyhousewithinitcertainassumptionsaboutwhatexists.Areweallabletoagreeonacommonstarting-point?AcommoncriticismofDescartesisthatifyoubeginwithhisMeditations,youwillneverescapeanexasperatingscepticism.Wesimplylacktheself-evidentaxiomsandtheclearandevidentideasrequiredtohaveanysortofconstructivephilosophyoftheworld.Astheefforttobuildupone’sunderstandingoftheworldfromaCartesianfoundationislosingitsappeal,therehavebeenwidespreadrepercussionsinthephilosophyofreligion.471\nCHARLESTALIAFERROOnewayofcarryingoutphilosophyofreligionalongnon-foundationalistlineshasbeentobuildacaseforthecomparativerationalityofareligiousviewoftheworld.Ithasbeenarguedthattheintellectualintegrityofareligiousworldviewcanbesecuredifitcanbeshowntobenolessrationalthantheavailablealternatives.Itneedonlyachieveintellectualparity.JohnHick(1990)andothersemphasizetheintegrityofreli-giouswaysofseeingtheworldthatareholistic,internallycoherent,andopentocriti-cismalongvariousexternallines.Onthelatterfront,ifareligiouswayofconceivingtheworldisatcompleteoddswithcontemporaryscience,thatwouldcountasgoodgroundsforrevisingthereligiousoutlook.Thecaseforreligionneednot,however,bescientificorevenanalogoustoscience.IfHickisright,religiouswaysofseeingtheworldarenotincompatiblewithscience,butcomplementary.IndependentofHickbutinthesamespirit,Plantinga(1993)hasproposedthatbeliefinGod’sexistencemaybetakenasproperlybasicandfullywarrantedwithouthavingtobejustifiedinrelationtostandardargumentsforGodfromdesign,miraclesandsoon.PlantingaarguesthatthetendencytobelieveinGodfollowsnaturaltendenciesofthehumanmind.ThisstancecompriseswhatiscommonlyreferredtoasReformedEpistemologybecauseofitsleaningonworkbytheReformedtheologianJohnCalvin(1509–64),whomaintainedthatwehaveasenseofGod(sensusdivinitatus)leadingustoseeGodintheworldaroundus.Plantingahastherebycouchedthequestionofjustificationwithinthelargerarenaofmetaphysics.Byadvancinganintricate,comprehensivepictureofhowbeliefscanbewarrantedwhentheyfunctionasGoddesignedthem,hehasprovidedwhatsomebelievetobeacombinedmetaphysicalandepistemiccasefortherationalityofreligiousconvictions.Whohastheburdenofproofinadebatebetweenatheistandanatheist?AntonyFlew(1984)thinksitisthetheist.Byhislights,thetheistandatheistcanagreeonawholebaselineoftruths(suchasthefindingsofthephysicalsciences).Thequestionthenbecomes,whygoanyfurther?FlewwieldsaversionofOckham’srazor,arguingthatifonehasnoreasontogofurther,onehasreasonnottogofurther.Hischallengehasbeenmetonvariousfronts,withsomecriticsclaimingthatFlew’sburdenofproofargumentisweddedtoanoutmodedfoundationalism,thatanyburdenofproofissharedequallybyatheistsandtheists,orthatthetheisthasanarrayofargumentstohelpshoulderagreaterburdenofproof.Thepositionoffideismisafurtheroption.Fideismistheviewthatreligiousbeliefdoesnotrequireevidenceandthatreligiousfaithisself-vindicating.KarlBarth(1886–1968)advocatedafideisticphilosophy.HickandPlantinganeednotbeconsideredfideistsbecauseofthemajorroleeachgivestoexperi-ence,coherenceandreflection.(Forasuperbstudyoffideism,seePenelhum1983.)Perhapsthejustificationmostwidelyofferedforreligiousbeliefconcernstheoccur-renceofreligiousexperienceorthecumulativeweightoftestimonyofthoseclaimingtohavehadreligiousexperiences.Puttingthelattercaseintheisticterms,theargu-mentappealstothefactthatmanypeoplehavetestifiedthattheyhavefeltGod’spresence.DoessuchtestimonyprovideevidencethatGodexists?ThatitisevidencehasbeenarguedbyGellman(1997),Yandell(1993),Alston(1991a,b),Davis(1989),Gutting(1982),Swinburne(1979)andothers.Thatitisnot(orthatitsevidentialforceistrivial)isarguedbyMartin(1990),Mackie(1985),Nielsen(1982)andothers.Inanefforttostimulatefurtherinvestigation,Ishallbrieflysketchsomeofthemovesandcounter-movesinthedebate.472\nPHILOSOPHYOFRELIGIONTheImportanceofReligiousExperience:ObjectionsandRepliesObjection:ReligiousexperiencecannotbeexperienceofGodforexperienceisonlysensoryandifGodisnon-physical,Godcannotbesensed.Reply:Thethesisthatexperienceisonlysensorycanbechallenged.Yandell(1993)marksoutsomeexperiences(aswhenonehas‘afeeling’someoneispresentbutwithouthavinganyaccompanyingsensations)thatmightprovidegroundsforquestioninganarrowsensorynotionofexperience.Objection:TestimonytohaveexperiencedGodisonlytestimonythatonethinksonehasexperiencedGod;itisonlytestimonyofaconviction,notevidence.Reply:TheliteratureonreligiousexperiencetestifiestotheexistenceofexperienceofsomeDivinebeingonthebasisofwhichthesubjectcomestothinktheexperienceisofGod.Ifreadcharitably,thetestimonyisnottestimonytoaconviction,buttoexperiencesthatformthegroundsfortheconviction.Objection:Becausereligiousexperienceisunique,howcouldoneeverdeterminewhetheritisreliable?Wesimplylacktheabilitytoexaminetheobjectofreligiousexperienceinordertotestwhetherthereportedexperiencesareindeedreliable.Reply:AswelearnedfromDescartes,allourexperiencesofexternalobjectsfaceaproblemofuniqueness.Itispossibleinprinciplethatalloursensesaremistakenandwedonothavethepublic,embodiedlifewethinkwelead.Wecannotstepoutofourownsubjectivitytovindicateourordinaryperceptualbeliefsanymorethaninthereligiouscase.Objection:Reportsofreligiousexperiencedifferradicallyandthetestimonyofonereli-giouspartyneutralizesthetestimonyofothers.ThetestimonyofHinduscancelsoutthetestimonyofChristians.ThetestimonyofatheiststoexperienceGod’sabsencecancelsoutthetestimonyof‘believers’.Reply:Severalrepliesmightbeofferedhere.TestimonytoexperiencetheabsenceofGodmightbebetterunderstoodastestimonynottoexperienceGod.FailingtoexperienceGodmightbejustificationforbelievingthatthereisnoGodonlytotheextentthatwehavereasontobelievethatifGodexistsGodwouldbeexperiencedbyall.Theistsmightevenappealtotheclaimbymanyatheiststhatitcanbevirtuoustoliveethicallywithatheistbeliefs.PerhapsifthereisaGod,Goddoesnotthinkthisisaltogetherbad,andactuallydesiresreligiousbelieftobefashionedunderconditionsoftrustandfaithratherthanknowledge.Thediversityofreligiousexperienceshascausedsomedefendersoftheargu-mentfromreligiousexperiencetomutetheirconclusion.Thus,Gutting(1982)contendsthattheargumentisnotstrongenoughfullytovindicateaspecificreligioustradition,butthatitisstrongenoughtooverturnananti-religiousnaturalism.Otherdefendersusetheirspecifictraditiontodealwithostensiblycompetingclaimsbasedondifferentsortsofreligiousexperiences.Theistshaveproposedthatmoreimpersonalexpe-riencesoftheDivinerepresentonlyoneaspectofGod.Godisapersonorisperson-like,butGodcanalsobeexperienced,forexample,assheerluminousunity.HindushaveclaimedtheexperienceofGodaspersonalisonlyonestageintheoveralljourneyofthesoultotruth,thehighesttruthbeingthatBrahmantranscendspersonhood.Howonesettlestheargumentwilldependonone’soverallconvictionsinmanyareasofphilosophy.CurrentlyitispopulartoseetheindividualargumentsfavouringandopposingGod’sexistenceaspartofcumulative,mutuallysupportivelinesofreasoning.Thisviewstressestheholisticcharacterofconcepts,suchastheviewthatconcepts473\nCHARLESTALIAFERROcannotbeisolatedbutmustbeseenaspartofconceptualframeworks.Theholistic,interwovennatureofboththeisticandatheisticargumentscanbereadilyillustrated.Ifyoudiminishtheimplicationsofreligiousexperienceandhaveahighstandardregard-ingtheburdenofproofforanysortofreligiousoutlook,thenitishighlylikelythattheclassicalargumentsforGod’sexistencewillnotbepersuasive.Moreover,ifonethinksthattheismcanbeshowntobeintellectuallyconfusedfromthestart,thentheisticargumentsfromreligiousexperiencewillcarrylittleweight.Testimonytohaveexperi-encedGodwillhavenomoreweightthantestimonytohaveexperiencedaroundsquare,andnon-religiousaccountsofreligiousexperience–likethoseofFreud(aresultofwish-fulfilment),Marx(areflectionoftheeconomicbase)orDurkheim(aproductofsocialforces)–willincreasetheirappeal.If,ontheotherhand,youthinkthetheisticpictureiscoherentandthatthetestimonyofreligiousexperienceprovidessomeevidencefortheism,thenyourassessmentoftheclassicaltheisticargumentsmightbemorefavourable,fortheywouldservetocorroborateandfurthersupportwhatyoualreadyhavesomereasontobelieve.Fromsuchavantagepoint,appealtowish-fulfilment,economicsandsocialforcesmighthavearole,buttheroleistoexplainwhysomepartiesdonothaveexperiencesofGodandtocounterthechargethatfailuretohavesuchexperiencesprovidesevidencethatthereisnoreligiousreality.Someinterconnectionsamongtheisticargumentsandamonganti-theisticargumentswillbenotedbelow.5FourArgumentsabouttheExistenceofGodWehavealreadycoveredconsiderablegroundthatistraversedinargumentssupport-ingandopposingtheism.Iwillmapsomeofthestrategiesthatareemployedinfouradditionalareaswhererecentphilosophicalactivityhasbeenfocused.Iwillrecountwhatappeartobestrongandlivelyversionsofthevariousargumentsandleavereaderstoexploremattersfurther,ratherthanofferthedozenorsoversionsofeachargumentthatarenowunderdebate.5.1OntologicalargumentsThereisahostofargumentsunderthistitle;allofthemarebasedprincipallyoncon-ceptual,apriorigroundswhichdonotinvolveaposterioriempiricalinvestigation.Ifaversionoftheargumentworks,thenitcanbedeployedusingonlytheconceptofGodandsomemodalprinciplesofinference,thatis,principlesconcerningpossibilityandnecessity.Theargumentneednotresistallempiricalsupport,however,asIshallindicate.Thefocusoftheargumentisthethesisthat,ifthereisaGod,thenGod’sexistenceisnecessary.God’sexistenceisnotcontingent–Godisnotthesortofbeingthatjusthappenstoexist.ThatthisisanaccuratepictureofwhatismeantbyGodmaybeshownbyappealingtothewayGodisconceivedinJewish,ChristianandIslamictraditions.Thiswouldinvolvesomeaposteriori,empiricalresearchintothewayGodisthoughtofinthesetraditions.Alternatively,adefenderoftheontologicalargumentmighthopetoconvinceothersthattheconceptofGodistheconceptofabeingthatexistsnecessar-474\nPHILOSOPHYOFRELIGIONilybybeginningwiththeideaofamaximallyexcellentbeing.Iftherewereamaxi-mallyexcellentbeingwhatwoulditbelike?Ithasbeenarguedthatamongitsarrayofgreat-makingqualities(omniscienceandomnipotence)wouldbenecessaryexistence.Oncefullyarticulated,itcanbearguedthatamaximallyexcellentbeing,whichexistednecessarily,couldbecalled‘God’.TheontologicalargumentgoesbacktoStAnselm,butIshallexploreacurrentversionrelyingheavilyontheprinciplethatifsomethingispossiblynecessarilythecase,thenitisnecessarilythecase(or,toputitredundantly,itisnecessarilynecessary).Theprinciplecanbeillustratedinthecaseofpropositions:1+1=2doesnotseemtobethesortofthingthatmightjusthappentobetrue.Rather,eitheritisnecessarilytrueornecessarilyfalse.Ifthelatter,itisnotpossible,iftheformer,itispossible.Ifonehasgoodreasontobelieveitispossiblethat1+1=2,thenonehasgoodreasontobelievethat1+1doesindeed,necessarilyequal2.DowehavereasontothinkitispossibleGodexistsnecessarily?Insupportofthis,onecanalsoappealtoaposteriorimatters,notingtheextantreligioustraditionsthatupholdsuchanotion.TheredoesnotappeartobeanythingamissintheirthinkingofGodasexisting;ifthebeliefthatGodexistsisincoherentthisisnotobvious.Indeed,anumberofatheiststhinkGodmightexist,butconcludeGoddoesnot.Ifwearesuccessfulinestablishingthepossi-bilitythatGodnecessarilyexists,theconclusionfollowsthatitisnecessarilythecasethatGodexists.Therehavebeenhundredsofobjectionsandrepliestothisargument.PerhapsthemostambitiousobjectionisthattheargumentcanbeusedwithoneminoralterationtoarguethatGodcannotexist.Assumealltheargumentabovebutalsothatitispos-sibleGoddoesnotexist.AtheistscanpointoutthatmanytheistswhobelievethereisaGodatleastallowforthebarepossibilitythattheycouldbewrongandthereisnoGod.IfitispossiblethatthereisnoGod,thenitwouldnecessarilyfollowthatthereisnoGod.Repliestothisobjectionemphasizethedifficultyofconceivingofthenon-existenceofGod.ThebattleoverwhetherGodisnecessaryorimpossibleisoftenfoughtoverthecoherenceofthevariousdivineattributesdiscussedinsection3.Ifyouthinktheseattributesarecompossible,involvenocontradictionsandviolatenoknownmetaphysicaltruths,thenyoumaywellhavegoodgroundsforconcludingthatGodispossibleandthereforenecessary.However,ifyouseeacontradiction,say,indescribingabeingwhoisatonceomniscientandomnipotent,youmaywellhavegoodgroundsforconcludingthatGod’sexistenceisimpossible.Anotherobjectionisthatitmakesnosensetothinkofabeingexistingnecessarily;propositionsmaybenecessarilytrueorfalse,butobjectscannotbenecessaryorcon-tingent.Somephilosophersreplythatitmakesnolesssensetothinkofanindividual(God)existingnecessarilythanitdoestothinkofpropositionsbeingnecessarilytrue.Afurtherobjectionisthattheontologicalargumentcannotgetoffthegroundbecauseofthequestion-beggingnatureofitspremisethatifthereisaGod,thenGodexistsnecessarily.Doesadmittingthispremiseconcedethatthereissomeindividualthingsuchthatifitexists,itexistsnecessarily?Replieshaveclaimedthattheargumentonlyrequiresonetoconsideranostensiblestateofaffairs,withouthavingtoconcedeinitiallywhetherthestateofaffairsispossibleorimpossible.ToconsiderwhatisinvolvedinpositingtheexistenceofGodisnomorehazardousthanconsideringwhatisinvolvedinpositingtheexistenceofunicorns.Onecanentertaintheexistenceof475\nCHARLESTALIAFERROunicornsandtheirnecessaryfeatures(thatnecessarilyiftherewereunicorns,therewouldexistsingle-hornedbeasts)withoutbelievingthatthereareunicorns.Finally,considertheobjectionthat,ifsuccessfulinprovidingreasonstobelievethatGodexists,theontologicalargumentcouldbeusedtoestablishtheexistenceofawholearrayofotheritems,likeperfectislands.Repliestothissortofobjectionhavetypicallyquestionedwhetheritmakessensetothinkofanisland(aphysicalthing)asexistingnecessarilyorashavingmaximalexcellenceonaparwithGod.Doestheimaginedislandhaveexcellenceslikeomniscience,omnipotence(apowerwhichwouldincludethepowertomakeindefinitelymanyislands)andsoon?Classical,alternativeversionsoftheontologicalargumentarepropoundedbyAnselm,SpinozaandDescartes,withcurrentversionsbyPlantinga(1974),HartshorneandMalcolm(bothinHickandMcGill1967),andDore(1984);classicalcriticsincludeGauniloandKant,andcurrentcriticsaremany,includingRowe(1993),Barnes(1972)andMackie(1985).5.2CosmologicalargumentsArgumentsinthisveinaremorefirmlyplantedinempirical,aposteriorireflection,butsomeversionsemployapriorireasonsaswell.Therearevariousversions.Somearguethatthecosmoshadaninitialcauseoutsideit,aFirstCauseintime.Othersarguethatthecosmoshasanecessary,sustainingcausefrominstanttoinstant.Thetwoversionsarenotmutuallyexclusive,foritispossibleboththatthecosmoshadaFirstCauseandthatitcurrentlyhasasustainingcause.Thecosmologicalargumentreliesontheintelligibilityofthenotionofsomethingwhichisnotitselfcausedtoexistbyanythingelse.Thiscouldbeeithertheall-outneces-sityofsupremepre-eminenceacrossallpossibleworldsusedinversionsoftheonto-logicalargument,oramorelocal,limitednotionofabeingthatisuncausedintheactualworld.Ifsuccessful,theargumentwouldprovidereasonforthinkingthereisatleastonesuchbeingofextraordinarypowerresponsiblefortheexistenceofthecosmos.Atbest,itmaynotjustifyafullpictureoftheGodofreligion(aFirstCausewouldbepowerful,butnotnecessarilyomnipotent),butitwouldnonethelesschallengenaturalisticalternativesandbringoneclosertotheism.Bothversionsoftheargumentaskustoconsiderthecosmosinitspresentstate.Istheworldasweknowitsomethingthatnecessarilyexists?Atleastwithrespecttoourselves,theplanet,thesolarsystemandthegalaxy,itappearsnot.Withrespecttotheseitemsinthecosmos,itmakessensetoaskwhytheyexistratherthannot.Inrelationtoscientificaccountsofthenaturalworld,suchenquiriesintocausesmakeabundantsenseandareperhapsevenessentialpresuppositionsofthenaturalsciences.Someproponentsoftheargumentcontendthatweknowapriorithatifsomethingexiststhereisareasonforitsexistence.Sowhydoesthecosmosexist?Ifweexplainthecontingentexistenceofthecosmos(orstatesofthecosmos)onlyintermsofothercon-tingentthings(earlierstatesofthecosmos,say),thenafullcosmicexplanationwillneverbeattained.Atthispointthetwoversionsoftheargumentdivide.ArgumentstoaFirstCauseintimecontendthatacontinuoustemporalregressfromonecontingentexistencetoanotherwouldneveraccountfortheexistenceofthecosmos,andtheyconcludethatitismorereasonabletoaccepttherewasaFirstCausethantoaccept476\nPHILOSOPHYOFRELIGIONeitheraregressortheclaimthatthecosmosjustcameintobeingfromnothing.Argumentstoasustainingcauseofthecosmosclaimthatexplanationsofwhysomethingexistsnowcannotbeadequatewithoutassumingapresent,contempora-neoussustainingcause.Theargumentshavebeenbasedonthedenialofallactualinfinities(Craig1980)orontheacceptanceofsomeinfinities(forinstance,thecoherenceofsupposingtheretobeinfinitelymanystars)combinedwiththerejectionofaninfiniteregressofexplanationssolelyinvolvingcontingentstatesofaffairs(Taylor1963).Thelatterhasbeendescribedasaviciousregressasopposedtoonethatisbenign.Thereareplausibleexamplesofviciousinfiniteregressesthatdonotgenerateexplanations:forinstance,imaginethatIexplainmypossessionofabookbyreportingthatIgotitfromAwhogotitfromB,andsoontoinfinity.ThiswouldnotexplainhowIgotthebook.Alternatively,imagineamirrorwithlightreflectedinit.Wouldthepresenceoflightbesuccessfullyexplainedifoneclaimedthatthelightwasareflectionoflightfromanothermirror,andthelightinthatmirrorcamefromyetanothermirror,andsoontoinfinity?Considerafinalcase.Youcomeacrossawordyoudonotunderstand;letitbe‘ongggt’.Youaskitsmeaningandaregivenanotherwordwhichisunintelligibletoyou,andsoon,forminganinfiniteregress.Wouldyoueverknowthemeaningofthefirstterm?Theforceofthesecasesistoshowhowsimilartheyaretotheregressofcontingentexplanations.VersionsoftheargumentthatrejectallactualinfinitiesfacetheembarrassmentofexplainingwhatistobemadeoftheFirstCause,especiallysinceitmighthavesomefeaturesthatareactuallyinfinite.Inreply,Craigandothershavecontendedthattheyhavenoobjectiontopotentialinfinities(althoughtheFirstCausewillneverceasetobe,itwillneverbecomeanactualinfinity).Theyfurtheracceptthatpriortothecreation,theFirstCausewasnotintime,apositionrelyingonthetheorythattimeisrelationalratherthanabsolute.Thecurrentscientificpopularityoftherelationalviewmayoffersupporttodefendersoftheargument.Ithasbeenobjectedthatbothversionsofthecosmologicalargumentsetoutaninflatedpictureofwhatexplanationsarereasonable.Whyshouldthecosmosasawholeneedanexplanation?Ifeverythinginthecosmoscanbeexplained,albeitthroughinfi-nite,regressiveaccounts,whatislefttoexplain?Onemayreply,eitherbydenyingthatinfiniteregressesactuallydosatisfactorilyexplainorbychargingthatthefailuretoseekanexplanationforthewholeisarbitrary.Thequestion‘Whyisthereacosmos?’seemsaperfectlyintelligibleone.Ifthereareaccountsforthingsinthecosmos,whynotforthewhole?Theargumentisnotbuiltonthefallacyoftreatingeverywholeashavingallthepropertiesofitsparts.Butifeverythinginthecosmosiscontingent,itseemsjustasreasonabletobelievethatthewholecosmosiscontingentasitistobelievethatifeverythinginthecosmosisinvisible,thecosmosasawholewouldbeinvisible.Anotherobjectionisthatratherthanexplainingthecontingentcosmos,thecos-mologicalargumentintroducesamysteriousentityofwhichwecanmakeverylittlephilosophicalorscientificsense.HowcanpositingatleastoneFirstCauseprovideabetteraccountofthecosmosthansimplyconcludingthatthecosmoslacksanultimateaccount?Intheend,thetheistseemsboundtoadmitthatwhytheFirstCausecreatedatallwasacontingentmatter.If,onthecontrary,thetheisthastoclaimthattheFirstCausehadtodowhatitdid,wouldnotthecosmosbenecessaryratherthancontingent?477\nCHARLESTALIAFERROSometheistscomeclosetoconcludingthatitwasindeedanessentialfeatureofGodthatcreationhadtooccur.IfGodissupremelygood,therehadtobesomeoverflowingofgoodnessintheformofacosmos(seeKretzmannandStumpinMorris(1987b)ontheideasofDionysiustheAreopagite).ButtheiststypicallyreservesomeroleforthefreedomofGodandthusseektoretaintheideathatthecosmosiscontingent.Defend-ersofthecosmologicalargumentstillcontendthatitsaccountofthecosmoshasacomprehensivesimplicitylackinginalternativeviews.God’schoicesmaybecontin-gent,butnotGod’sexistenceandtheDivinechoiceofcreatingthecosmoscanbeunderstoodtobeprofoundlysimpleinitssupreme,overridingendeavour,namelytocreatesomethinggood.Swinburne(1977)hasarguedthataccountingfornaturallawsintermsofGod’swillprovidesforasimple,overarchingframeworkintermsofwhichtocomprehendtheorderandpurposivecharacterofthecosmos.Atthispointwemovefromthecosmologicaltotheteleologicalarguments.DefendersofthecosmologicalargumentincludeSwinburne(1979),Taylor(1963),Meynell(1982),Reichenbach(1972)andRowe(1993);prominentopponentsincludeFlew(1984),Martin(1990)andMackie(1985).5.3TeleologicalargumentsTheseargumentsfocusoncharacteristicsofthecosmosthatseemtoreflectthedesignorintentionalityofGodor,moremodestly,ofoneormorepowerful,intelligentGod-likeagents.Partoftheargumentmaybeformulatedasprovidingevidencethatthecosmosisthesortofrealitythatwouldbeproducedbyanintelligentbeing,andthenarguingthatpositingthissourceismorereasonablethanagnosticismordenyingit.Asinthecaseofthecosmologicalargument,thedefenderoftheteleologicalargumentmaywanttoclaimonlytobegivingussomereasonforthinkingthereisaGod.Notethewaythevariousargumentsmightthenbebroughttobearoneachother.Ifsuc-cessful,theteleologicalargumentmayprovidesomereasonforthinkingthattheFirstCauseofthecosmologicalargumentispurposive,whiletheontologicalargumentpro-videssomereasonforthinkingthatitmakessensetopositabeingthathasDivineattrib-utesandnecessarilyexists.Behindallofthemanargumentfromreligiousexperiencemayprovidesomeinitialreasonstoseekfurthersupportforareligiousconceptionofthecosmosandtoquestiontheadequacyofnaturalism.Oneversionoftheteleologicalargumentwilldependontheintelligibilityofpurpo-siveexplanation.Inourownhumancaseitappearsthatintentional,purposiveexpla-nationsarelegitimateandcantrulyaccountforthenatureandoccurrenceofevents.Inthinkingaboutanexplanationfortheultimatecharacterofthecosmos,isitmorelikelyforthecosmostobeaccountedforintermsofapowerful,intelligentagentorintermsofanaturalisticschemeoffinallawswithnointelligencebehindthem?Theistsemployingtheteleologicalargumentwilldrawattentiontotheorderandstabilityofthecosmos,theemergenceofvegetativeandanimallife,theexistenceofconscious-ness,morality,rationalagentsandthelike,inanefforttoidentifywhatmightplausi-blybeseenaspurposivelyexplicablefeaturesofthecosmos.Naturalisticexplanations,whetherinbiologyorphysics,arethencastasbeingcomparativelylocalinapplicationwhenheldupagainstthebroaderschemaofatheisticmetaphysics.Darwinianaccountsofbiologicalevolutionwillnotnecessarilyassistusinthinkingthroughwhy478\nPHILOSOPHYOFRELIGIONthereareeitheranysuchlawsoranyorganismstobeginwith.Argumentssupportingandopposingtheteleologicalargumentwillthenresembleargumentsaboutthecos-mologicalargument,withthenegativesidecontendingthatthereisnoneedtomovebeyondanaturalisticaccount,andthepositivesideaimingtoestablishthatfailingtogobeyondnaturalismisunreasonable.Inassessingtheteleologicalargument,wecanbeginwiththeobjectionfromunique-ness.Wecannotcompareourcosmoswithothers,determiningwhichhavebeendesignedandwhichhavenot.Ifwecould,thenwemightbeabletofindsupportfortheargument.Ifwecouldcompareourcosmoswiththoseweknewtobedesignedandifthecomparisonwerecloserthanwiththoseweknewtobeundesigned,thentheargu-mentmightbeplausible.Withoutsuchcomparisons,however,theargumentfails.Repliestothislineofattackhavecontendedthatwerewetoinsistthatinferencesinuniquecaseswereoutoforder,thenwewouldhavetoruleoutotherwiseperfectlyrespectablescientificaccountsoftheoriginofthecosmos.Besides,whileitisnotpos-sibletocomparethelayoutofdifferentcosmichistories,itisinprinciplepossibletoenvisageworldsthatseemchaotic,randomorbasedonthelawsthatcrippletheemer-genceoflife.Nowwecanenvisageanintelligentbeingcreatingsuchworlds,but,throughconsideringtheirfeatures,wecanarticulatesomemarksofpurposivedesigntohelpusjudgewhetherthecosmoswasdesignedratherthancreatedatrandom.Somecriticsappealtothepossibilitythatthecosmoshasaninfinitehistorytobolsterandreintroducetheuniquenessobjection.Giveninfinitetimeandchance,itseemslikelythatsomethinglikeourworldwillcomeintoexistence,withallitsappearanceofdesign.Ifso,whyshouldwetakeittobesoshockingthatourworldhasitsapparentdesignandwhyshouldexplainingtheworldrequirepositingoneormoreintelligentdesign-ers?Repliesrepeattheearliermoveofinsistingthatiftheobjectionweretobedecisive,thenmanyseeminglyrespectableaccountswouldalsohavetofallbythewayside.Itisoftenconcededthattheteleologicalargumentdoesnotdemonstratethatoneormoredesignersarerequired;itseeksrathertoestablishthatpositingsuchpurposiveintelli-genceisreasonableandpreferabletonaturalism.DefendersoftheargumentthiscenturyincludeSchlesinger(1977)andSwinburne(1979).ItisrejectedbyMackie(1985),Martin(1990)andothers.Onefeatureoftheteleologicalargumentcurrentlyreceivingincreasedattentionfocusesonepistemology.Ithasbeencontendedthatifwedorelyonourcognitivefac-ulties,itisreasonabletobelievethatthesearenotbroughtaboutbynaturalisticforces,forcesthatareentirelydrivenbychanceoraretheoutcomeofprocessesnotformedbyanoverridingintelligence.Anillustrationmayhelptounderstandtheargument.Imaginecomingacrosswhatappearstobeasignreportingsomeinformationaboutyourcurrentaltitude(somerocksinaconfigurationgivingyouyourcurrentlocationandpreciseheightabovesea-levelinmetres).Ifyouhadreasontobelievethatthis‘sign’wastotallytheresultofchanceconfigurations,woulditstillbereasonabletotrustit?Sometheistsarguethatitwouldnotbereasonable,andthattrustingourcognitivefac-ultiesrequiresustoacceptthattheywereformedbyanoverarching,good,creativeagent.ThisrekindlesDescartes’spointaboutrelyingonthegoodnessofGodtoensurethatourcognitivefacultiesareingoodworkingorder.Objectionstothisargumentcentreonnaturalisticexplanations,especiallythosefriendlytoevolution.Inevolu-tionaryepistemology,onetriestoaccountforthereliabilityofcognitivefacultiesin479\nCHARLESTALIAFERROtermsoftrialanderrorleadingtosurvival.Arejoinderbytheistsisthatsurvivalaloneisnotnecessarilylinkedtotruebeliefs.Itcould,inprinciple,befalsebeliefsthatenhancesurvival.Infact,someatheistsdothinkthatbelievinginGodhasbeencrucialtopeople’ssurvival,thoughthebeliefisradicallyfalse.Martin(1990)andMackie(1985),amongothers,objecttotheepistemicteleologicalargument;Plantinga(1993),Creel(1995)andTaylor(1963)defendit.Amoresustainedobjectionagainstvirtuallyallversionsoftheteleologicalargu-menttakesissuewiththeassumptionthatthecosmosisgoodorthatitisthesortofthingthatwouldbebroughtaboutbyanintelligent,completelybenevolentbeing.5.4ProblemsofevilIfthereisaGodwhoisomnipotent,omniscientandcompletelygood,whyisthereevil?TheproblemofevilisthemostwidelyconsideredobjectiontotheisminbothWesternandEasternphilosophy.Therearetwogeneralversionsoftheproblem:thedeductiveorlogicalversion,whichassertsthattheexistenceofanyevilatall(regardlessofitsroleinproducinggood)isincompatiblewithGod’sexistence;andtheprobabilisticversion,whichassertsthatgiventhequantityandseverityofevilthatactuallyexists,itisunlikelythatGodexists.Thedeductiveproblemiscurrentlylesscommonlydebatedbecauseitiswidelyacknowledgedthatathoroughlygoodbeingmightalloworinflictsomeharmundercertainmorallycompellingconditions(suchascausingachildpainwhenremovingasplinter).Moreintensedebateconcernsthelikelihood(orevenpos-sibility)thatthereisacompletelygoodGodgiventhevastamountofevilinthecosmos.Considerhumanandanimalsufferingcausedbydeath,predation,birthdefects,rav-agingdiseases,virtuallyuncheckedhumanwickedness,torture,rape,oppressionand‘naturaldisasters’.Considerhowoftenthosewhosufferareinnocent.Whyshouldtherebesomuchgratuitous,apparentlypointlessevil?Inthefaceoftheproblemofevil,somephilosophersandtheologiansdenythatGodisall-powerfulandall-knowing.MILL(1806–73)(chapter35)tookthisline,andpanen-theisttheologianstodayalsoquestionthetraditionaltreatmentsofDivinepower.Bytheirlights,Godisimmanentintheworld,sufferingwiththeoppressedandworkingtobringgoodoutofevil,althoughinspiteofGod’sefforts,evilwillinvariablymarthecreatedorder.AnotherresponseistothinkofGodasbeingverydifferentfromamoralagent.Davies(1993)andothershavecontendedthatwhatitmeansforGodtobegoodisdifferentfromwhatitmeansforanagenttobegood.ThosewhothinkofGodasBeingasopposedtoabeinghavesomereasontoadoptthisposition.Amoredesperatestrat-egyistodenytheexistenceofevil,butitisdifficulttoreconciletraditionalmonothe-ismwithmoralscepticism.Also,insofaraswebelievetheretobeaGodworthyofworshipandafittingobjectofhumanlove,theappealtomoralscepticismwillcarrylittleweight.Theideathatevilisaprivationofthegood,atwistingofsomethinggood,mayhavesomecurrencyinthinkingthroughtheproblemofevil,butitisdifficulttoseehowitalonecouldgoveryfartovindicatebeliefinGod’sgoodness.Searingpainandendlesssufferingseemaltogetherrealeveniftheyareanalysedasbeingphilo-sophicallyparasiticonsomethingvaluable.Thethreegreatmonotheistictraditions,withtheirampleinsistenceontherealityofevil,offerlittlereasontotrytodefusetheproblemofevilbythisroute.Indeed,classicalJudaism,ChristianityandIslamareso480\nPHILOSOPHYOFRELIGIONcommittedtotheexistenceofevilthatareasontorejectevilwouldbeareasontorejectthesereligioustraditions.WhatwouldbethepointofJudaicteachingabouttheExodus(GodliberatingthepeopleofIsraelfromslavery)orChristianteachingabouttheIncar-nation(ChristrevealingGodasloveandreleasingaDivinepowerthatwill,intheend,conquerdeath),ortheIslamicteachingofMohammed(theholyprophetofAllahwhoisall-justandall-merciful)ifslavery,hate,deathandinjusticedonotexist?Inpart,themagnitudeoftheproblemofevilfortheismwilldependuponone’scom-mitmentsinotherareasofphilosophy,especiallyethics,epistemologyandmetaphysics.Ifinethicsyouholdthatthereshouldbenopreventablesufferingforanyreason,nomatterwhatthecauseorconsequence,thentheproblemofevilwillconflictwithacceptingtraditionaltheism.Moreover,ifyouholdthatanysolutiontotheproblemofevilshouldbeevidenttoallpersons,thenagaintraditionaltheismisinjeopardy,forclearlythe‘solution’isnotevidenttoall.Debatehaslargelycentredoverthelegitimacyofadoptingsomepositioninthemiddle:atheoryofvaluesthatwouldpreserveaclearassessmentoftheprofoundevilinthecosmosaswellassomeunderstandingofhowthismightbecompatiblewiththeexistenceofanall-powerful,completelygoodCreator.CouldtherebereasonswhyGodwouldpermitcosmicills?Ifwedonotknowwhatthosereasonsmightbe,areweinapositiontoconcludethattherearenoneorthattherecouldnotbeany?Exploringdifferentpossibilitieswillbeshapedbyone’smetaphysics.Forexample,ifyoudonotbelievethereisfreewill,thenyouwillnotbemovedbyanyappealtothepositivevalueoffreewillanditsroleinbringingaboutgoodasoffsettingitsroleinbringingaboutevil.Theisticresponsestotheproblemofevildistinguishbetweenadefenceandatheod-icy.AdefenceseekstoestablishthatrationalbeliefthatGodexistsisstillpossible(whenthedefenceisemployedagainstthelogicalversionoftheproblemofevil)andthattheexistenceofevildoesnotmakeitimprobablethatGodexists(whenusedagainsttheprobabilisticversion).Accordingtothedefenceresponse,nocreatureshouldexpecttobeabletosolvetheproblemofevil;itisbeyondourepistemiccapacitiestostandinjudgementhere.SomehaveadoptedthedefencestrategywhilearguingthatweareinapositiontohaverationalbeliefsintheexistenceofevilandinacompletelygoodGodwhohatesthisevil,eventhoughweareunabletoseehowthesetwobeliefsarecom-patible.Atheodicyismoreambitious,andistypicallypartofabroaderproject,arguingthatitisreasonabletobelievethatGodexistsinlightofthegoodaswellastheevidentevilofthecosmos.Inatheodicy,theprojectisnottoaccountforeachandeveryevil,buttoprovideanoverarchingframeworkwithinwhichtounderstandatleastroughlyhowtheevilthatoccursispartofsomeoverallgood–forinstance,theovercomingofevilisitselfagreatgood.Inpractice,adefenceandatheodicyoftenappealtosimilarfactors,thefirstandforemostbeingwhatmanycalledtheGreaterGoodDefence.EvilandtheGreaterGoodIntheGreaterGoodDefenceitiscontendedthatevilcanbeunderstoodaseitheranecessaryaccompanimenttobringingaboutgreatergoodsoranintegralpartofthesegoods.Thus,inaversionoftencalledtheFreeWillDefence,itisproposedthatfreecrea-tureswhoareabletocareforeachotherandwhosewelfaredependsoneachother’sfreelychosenactionconstituteagood.Forthisgoodtoberealized,itisargued,theremust481\nCHARLESTALIAFERRObethebonafidepossibilityofpersonsharmingeachother.Thefreewilldefenceissome-timesusednarrowlyonlytocoverevilthatoccursasaresult,directorindirect,ofhumanaction.ButithasbeenspeculativelyextendedbythoseproposingadefenceratherthanatheodicytocoverotherevilswhichmightbebroughtaboutbysupernaturalagentsotherthanGod.AccordingtotheGreaterGoodcase,evilprovidesanopportunitytorealizegreatvalues,suchasthevirtuesofcourageandthepursuitofjustice.Reichen-bach(1982),Tennant(1930)andSwinburne(1979)havealsounderscoredthegoodofastableworldofnaturallawsinwhichanimalsandhumanslearnaboutthecosmosanddevelopautonomously,independentofthecertaintythatGodexists.SomeatheistsaccordvaluetothegoodoflivinginaworldwithoutGod,andtheseviewshavebeenusedbytheiststobackuptheclaimthatGodmighthavereasontocreateacosmosinwhichDivineexistenceisnotoverwhelminglyobvioustous.IfGod’sexistencewereover-whelminglyobvious,thenmotivationstovirtuemightbecloudedbyself-interestandbythebarefearofoffendinganomnipotentbeing.Further,theremayevenbesomegoodtoactingvirtuouslyevenifcircumstancesguaranteeatragicoutcome.JohnHick(1978)soarguedandhasdevelopedwhatheconstruestobeanIrenaeanapproachtotheproblemofevil(namedafterStIrenaeusofthesecondcentury).Onthisapproach,itisdeemedgoodthathumanitydevelopsthelifeofvirtuegradually,evolvingtoalifeofgrace,maturityandlove.ThiscontrastswithatheodicyassociatedwithStAugustine,accordingtowhichGodcreatedusperfectandthenallowedustofallintoperdition,onlytoberedeemedlaterbyChrist.HickthinkstheAugustinianmodelfailswhereastheIrenaeanoneiscredible.Somehavebasedanargumentfromtheproblemofevilonthechargethatthisisnotthebestpossibleworld.Iftherewereasupreme,maximallyexcellentGod,surelyGodwouldbringaboutthebestpossiblecreation.Becausethisisnotthebestpossiblecreation,thereisnosupreme,maximallyexcellentGod.FollowingAdams(1987),manynowreplythatthewholenotionofabestpossibleworld,likethehighestpossiblenumber,isincoherent.Foranyworldthatcanbeimaginedwithsuchandsuchhappiness,goodness,virtueandsoon,ahigheronecanbeimagined.Ifthenotionofabestpossibleworldisincoherent,wouldthiscountagainstbeliefthattherecouldbeasupreme,maximallyexcellentbeing?IthasbeenarguedonthecontrarythatDivineexcellencesadmitofupperlimitsormaximathatarenotquantifiableinaserialfashion(forexample,Divineomnipotenceinvolvesbeingabletodoanythinglogicallyormeta-physicallypossible,butdoesnotrequireactuallydoingthegreatestnumberofactsoraseriesofactsofwhichtherecanbenomore).ThoseconcernedwiththeproblemofevilclashoverthequestionofhowoneassessesthelikelihoodofDivineexistence.SomeonewhoreportsseeingnopointtotheexistenceofevilornojustificationforGodtoallowitseemstoimplythatiftherewereapointtheywouldseeit.Notethedifferencebetweenseeingnopointandnotseeingapoint.Inthecosmiccase,isitclearthatiftherewereareasonjustifyingtheexistenceofevil,wewouldseeit?Rowe(1993)thinkssomeplausibleunderstandingofGod’sjustificatoryreasonforallowingtheevilshouldbedetectable,butthattherearecasesofevilthatarealtogethergratuitous.DefenderslikeHasker(1989)andWykstra(1984)replythatthesecasesarenotdecisivecounter-examplestotheclaimthatthereisagoodGod.Thesephilosophersholdthatwecanrecognizeevilandgraspourdutytodoallin482\nPHILOSOPHYOFRELIGIONourpowertopreventoralleviateit.ButweshouldnottakeourfailuretoseewhatreasonGodmighthaveforallowingeviltocountasgroundsforthinkingthatthereisnoreason.Someportraitsofanafterlifeseemtohavelittlebearingonourresponsetothemag-nitudeofevilhereandnow.DoesithelptounderstandwhyGodallowsevilifallvictimswillreceivehappinesslater?Butitisdifficulttotreatthepossibilityofanafterlifeasentirelyirrelevant.Isdeaththeannihilationofpersonsoraneventinvolvingatrans-figurationtoahigherstate?Ifyoudonotthinkthatitmatterswhetherpersonscon-tinuetoexistafterdeath,thensuchspeculationisoflittleconsequence.Butsupposethattheafterlifeisunderstoodasbeingmorallyintertwinedwiththislife,withoppor-tunityformoralandspiritualreformation,transfigurationofthewicked,rejuvenationandoccasionsfornewlife,perhapsevenreconciliationandcommunionbetweenoppressorsseekingforgivenessandtheirvictims.Thentheseconsiderationsmighthelptodefendagainstargumentsbasedontheexistenceofevil.Insofarasonecannotruleoutthepossibilityofanafterlifemorallytiedtoourlife,onecannotruleoutthepossibilitythatGodbringssomegoodoutofcosmicills.5.5OtherargumentsIhavenotbeenabletocoverthemanyotherargumentsforandagainsttheexistenceofGod,butIshallnowmentionafewofthese.Theargumentfrommiraclesstartsfromspecificextraordinaryevents,arguingthattheyprovidereasonsforbelievingtheretobeasupernaturalagentor,moremodestly,reasonsforscepticismaboutthesufficiencyofanaturalisticworldview.Theargumenthasattractedmuchphilosophi-calattention,especiallysinceDavidHume’srejectionofmiracles.Thedebatehasturnedmainlyonhowonedefinesamiracle,understandsthelawsofnature,andspecifiestheprinciplesofevidencethatgoverntheexplanationofhighlyunusualhistoricaloccurrences.Therearevariousargumentsthatareadvancedtomotivatereligiousbelief.OneofthemostinterestingandpopularisawagerargumentoftenassociatedwithPascal(1623–62).ItisdesignedtoofferpracticalreasonstocultivateabeliefinGod.ImaginethatyouareunsurewhetherthereisorisnotaGod.Youhaveitwithinyourpowertoliveoneitherassumptionandperhaps,throughvariouspractices,togetyourselftobelieveoneortheother.TherewouldbegoodconsequencesofbelievinginGodevenifyourbeliefwerefalse,andifthebeliefweretrueyouwouldreceiveevengreatergood.TherewouldalsobegoodconsequencesofbelievingthatthereisnoGod,butinthiscasetheconsequenceswouldnotalteriftherewerenoGod.If,however,youbelievethatthereisnoGodandyouarewrong,thenyouwouldrisklosingthemanygoodswhichfollowfrombeliefthatGodexistsandactualDivineexistence.Onthisbasis,itmayseemreasonabletobelievethereisaGod.Indifferentformstheargumentmaybegivenaroughedge(forexample,imaginethatifyoudonotbelieveinGodandthereisaGod,helliswaiting).Itmaybeputasanappealtoindividualself-interest(youwillbebetteroff)ormoregenerally(believerswhoselivesareboundtogethercanrealizesomeofthegoodscomprisingamaturereli-giouslife).Objectorsworryaboutwhetheroneeverisabletobringchoicesdowntojustsuchanarrowselection–forexample,tochooseeithertheismornaturalism.Some483\nCHARLESTALIAFERROthinktheargumentistoothoroughlyegotisticandthusoffensivetoreligion.ForarecentdefenceoftheargumentseeRescher(1985).6NewDirectionsintheFieldGrowthinthephilosophyofreligionshowsnosignofabating.Therearecurrentlyeffortstoexplorecross-culturalphilosophyofreligion,toarticulatefeministchallengestotraditionalreligions,toaddresspostmodernversionsofreligion,andtoconsiderahostofpractical,moralandsocialproblemsfromthestandpointofaphilosophicallyarticulatereligiousethic,focusingonenvironmental,medicalandpoliticalconcerns.Specificissuesinternaltoreligioustraditionsarealsoreceivingincreasedattentionwithphilosophicalspeculationontheafterlife,heaven,hell,thesacraments,themeaningofrevelationandthenatureofreligiouscommunities.AcknowledgementsTheauthorgratefullyacknowledgescommentsonearlierversionsofthischapter,especiallyfromM.Okerlund,C.Gossett,G.Gentry,M.Lazenby,C.GrannemanandtheColumbiaUniversityPhilosophersofReligion.FurtherReadingReadersnewtothephilosophyofreligionmaybeginwithanyoftheintroductorytextsmentionedinthefirstsectionofthereferences,supplementedbyoneormoreoftheseusefulanthologiescontainingimportantpapersorexcerptsfrombooks:QuinnandTaliaferro(1996)includes78originalessaysspanningthehistoryofthefieldinEasternandWesternphilosophy.Mitchell(1971)containspapersontheearlydebatesaboutthemeaningofreligiouslanguage,andincludesearlypapersbyFlewandMitchell.Morris(1987)isanextremelyusefulcollectionontheconceptofGod,withpapersbyKretzmann,Stump,BlumenfeldandRowe.OthervaluablecollectionsareeditedbyPojman(1987),RoweandWainwright(1989),Brody(1992),Hick(1970),GeivettandSweetman(1992),Hudson(1991),Beaty(1990)andMacDonald(1991).Readersinterestedinexploringparticulartopicsdiscussedinthechaptercanturntothebooksandarticlesinthesecondsectionofreferences.Journalsthatareparticularlystronginaddressingthephilosophyofreligionare:TheInternationalJournalforPhilosophyofReligion,ReligiousStudies,FaithandPhilosophy,Sophia,TheJournalofReligion,TheJournalofReligiousEthics,TheHeythropJournalforPhilosophyandTheology,TheAnnualoftheSocietyofChristianEthics,AmericanCatholicPhilosophicalQuarterlyandtheThomist.PapersinphilosophyofreligionarepublishedinvirtuallyallthemainphilosophyjournalsintheEnglish-speakingworld.ThereareregularsessionsaboutphilosophyofreligionontheprogrammesoftheannualmeetingsofthethreedivisionsoftheAmericanPhilosophi-calAssociation,aswellasontheprogrammeoftheannualmeetingoftheAmericanAcademyofReligion.Societiesinterestedinthephilosophyofreligioninclude:theSocietyforPhilosophyofReligion(intheUK),theSocietyofChristianPhilosophers,thePhilosophyofReligionSociety,theSocietyforPhilosophyofReligion(intheUSA),theAmericanCatholicPhilosophicalAssociation,theAmericanHumanistAssociation,theAmericanMaritainAssociation,the484\nPHILOSOPHYOFRELIGIONFellowshipofReligiousHumanists,theJesuitPhilosophicalAssociation,theSocietyforMedievalandRenaissancePhilosophy,andtheSocietyforPhilosophyandTheology.AddressesandwebsitesformanyoftheseorganizationsarefoundintheDirectoryofAmericanPhilosophers,apublicationofthePhilosophyDocumentationCenter,BowlingGreenStateUniversity.ReferencesIntroductorytextsAbraham,W.1985:AnIntroductiontothePhilosophyofReligion.EnglewoodCliffs,NJ:Prentice-Hall.Anderson,P.S.1997:AFeministPhilosophyofReligion.Oxford:Blackwell.Clark,B.andClark,B.R.1998:ThePhilosophyofReligion.Oxford:Blackwell.Davies,B.1993:AnIntroductiontothePhilosophyofReligion.Oxford:OxfordUniversityPress.Evans,S.C.1985:PhilosophyofReligion.DownersGrove,IL:IntervarsityPress.Hick,J.1990:PhilosophyofReligion,4thedn.EnglewoodCliffs,NJ:Prentice-Hall.Morris,T.V.1991:OurIdeaofGod.DownersGrove,IL:IntervarsityPress.Peterson,M.etal.1991:IntroductiontothePhilosophyofReligion.Oxford:OxfordUniversityPress.Rowe,W.1993:PhilosophyofReligion.Belmont,CA:Wadsworth.Taliaferro,C.1998:AnIntroductiontoContemporaryPhilosophyofReligion.Oxford:Blackwell.Tilghman,B.R.1993:AnIntroductiontoPhilosophyofReligion.Oxford:Blackwell.Wainwright,W.1988:PhilosophyofReligion.Belmont,CA:Wadsworth.OtherbooksandarticlesAdams,R.M.1987:TheVirtueofFaith.Oxford:OxfordUniversityPress.Alston,W.1991a:PerceivingGod.Ithaca,NY:CornellUniversityPress.——1991b:TheInductiveArgumentfromEvil.PhilosophicalPerspectives,5,29–68.Augustine1972[426]:TheCityofGod(translatedbyH.Bettenson).Harmondsworth:PenguinBooks.Barnes,J.1972:TheOntologicalArgument.London:Macmillan.Beaty,M.(ed.)1990:ChristianTheismandtheProblemsofPhilosophy.NotreDame,IN:UniversityofNotreDamePress.Beaty,M.andTaliaferro,C.1990:GodandConceptEmpiricism.SouthwestPhilosophyReview,6:2.Boethius1969[524]:TheConsolationofPhilosophy(translatedbyV.E.Watts).Harmondsworth:PenguinBooks.Brody,B.(ed.)1992:ReadingsinthePhilosophyofReligion.EnglewoodCliffs,NJ:Prentice-Hall.Brown,D.1987:ContinentalPhilosophyandModernTheology.Oxford:Blackwell.Clayton,J.1987:Religions,ReasonsandGods.ReligiousStudies,23.Craig,W.1979:TheKalamCosmologicalArgument.NewYork:BarnesandNoble.——1980:TheCosmologicalArgumentfromPlatotoLeibniz.NewYork:BarnesandNoble.Craig,W.L.andSmith,Q.1993:Theism,Atheism,andBigBangCosmology.Oxford:ClarendonPress.Creel,R.1995:DivineImpassibility.Cambridge:CambridgeUniversityPress.Davis,C.1989:TheEvidentialForceofReligiousExperience.Oxford:OxfordUniversityPress.Dore,R.1984:Theism.Dordrecht:D.Reidel.Evans,S.1996:TheHistoricalChristandtheJesusofFaith:TheIncarnationalNarrativeasHistory.Oxford:ClarendonPress.485\nCHARLESTALIAFERROFirth,R.1970:EthicalAbsolutismandtheIdealObserver.InW.SellarsandJ.Hospers(eds)ReadingsinEthicalTheory.EnglewoodCliffs,NJ:Prentice-Hall.Flew,A.1984:God,FreedomandImmortality.Buffalo,NY:PrometheusBooks.Forrest,P.1996:GodWithouttheSupernatural:ADefenseofScientificTheism.Ithaca,NY:CornellUniversityPress.Foster,J.1985:Ayer.London:RoutledgeandKeganPaul.Geivett,R.andSweetman,B.(eds)1992:ContemporaryPerspectivesonReligiousEpistemology.Oxford:OxfordUniversityPress.Gellman,J.1997:ExperienceofGodandtheRationalityofTheisticBelief.Ithaca,NY:CornellUniversityPress.Griffiths,P.1994:OnBeingBuddha.Albany:StateUniversityofNewYorkPress.Gutting,G.1982:ReligiousBeliefandReligiousSkepticism.NotreDame,IN:UniversityofNotreDamePress.Hare,J.E.1996:TheMoralGap.Oxford:OxfordUniversityPress.Hasker,W.1989:God,Time,andKnowledge.Ithaca,NY:CornellUniversityPress.——1999:TheEmergentSelf.Ithaca,NY:CornellUniversityPress.Helm,P.1988:EternalGod.Oxford:OxfordUniversityPress.Hepburn,R.W.1963:FromWorldtoGod.Mind,72.Hick,J.1966:RationalTheisticBeliefWithoutProof.London:Macmillan.——(ed.)1970:ClassicalandContemporaryReadingsinthePhilosophyofReligion.EnglewoodCliffs,NJ:Prentice-Hall.——1978:EvilandtheGodofLove.NewYork:HarperandRow.Hick,J.andMcGill,A.(eds)1967:TheManyFacedArgument.NewYork:Macmillan.Howard-Snyder,D.(ed.)1996:TheEvidentialArgumentfromEvil.Bloomington:IndianaUniversityPress.Hudson,Y.(ed.)1991:ThePhilosophyofReligion.London:MayfieldPublishing.Hughes,G.1995:TheNatureofGod.London:Routledge.Hume,D.1947[1779]:DialoguesConcerningNaturalReligion(editedbyN.KempSmith).London:Nelson.Kant,I.1960:ReligionWithintheLimitsofReasonAlone(translatedbyT.M.GreeneandH.H.Hudson).NewYork:HarperandBrothers.Kenny,A.1979:TheGodofthePhilosophers.Oxford:ClarendonPress.Kerr,F.1988:TheologyAfterWittgenstein.Oxford:Blackwell.Kvanvic,J.1986:ThePossibilityofanAll-KnowingGod.London:Macmillan.Leftow,B.1991:TimeandEternity.Ithaca,NY:CornellUniversityPress.Levinas,E.1961:TotalityandInfinity(translatedbyA.Lingis).Pittsburgh:DuquesneUniversityPress.Linville,M.1993:DivineForeknowledgeandtheLibertarianConceptionofHumanFreedom.JournalofthePhilosophyofReligion,33.MacDonald,S.1991:BeingandGoodness.Ithaca,NY:CornellUniversityPress.Mackie,J.1985:TheMiracleofTheism.Oxford:ClarendonPress.Malcolm,N.1975:TheGroundlessnessofReligiousBeliefs.InS.Brown(ed.)ReasonandReligion.Ithaca,NY:CornellUniversityPress.Martin,M.1990:Atheism.Philadelphia,PA:TempleUniversityPress.Menzel,C.andMorris,T.V.1986:AbsoluteCreation.AmericanPhilosophicalQuarterly,23,353–62.Meynell,H.1982:TheIntelligibleUniverse.Totowa,NJ:BarnesandNoble.Mitchell,B.(ed.)1971:ThePhilosophyofReligion.Oxford:OxfordUniversityPress.——1973:TheJustificationofReligiousBelief.London:Macmillan.——1994:FaithandCriticism.Oxford:OxfordUniversityPress.486\nPHILOSOPHYOFRELIGIONMorris,T.V.1986:TheLogicofGodIncarnate.Ithaca,NY:CornellUniversityPress.——1987a:AnselmianExplorations.NotreDame,IN:UniversityofNotreDamePress.——(ed.)1987b:TheConceptofGod.Oxford:OxfordUniversityPress.Neville,R.C.1995:Religions,Philosophies,andPhilosophyofReligion.InternationalJournalforPhilosophyofReligion,38.Nielsen,K.1982:AnIntroductiontothePhilosophyofReligion.NewYork:StMartin’sPress.——1996:NaturalismWithoutFoundations.Buffalo,NY:PrometheusPress.Oppy,G.1995:OntologicalArgumentsandBeliefinGod.Cambridge:CambridgeUniversityPress.Owen,H.P.1965:TheMoralArgumentforChristianTheism.London:GeorgeAllenandUnwin.Padgett,A.1992:God,EternityandtheNatureofTime.NewYork:StMartin’sPress.Penelhum,T.1983:GodandSkepticism.Dordrecht:D.Reidel.Phillips,D.Z.1970:ConceptofPrayer.London:RoutledgeandKeganPaul.Plantinga,A.1967:GodandOtherMinds.Ithaca,NY:CornellUniversityPress.——1974:TheNatureofNecessity.Oxford:OxfordUniversityPress.——1980:DoesGodHaveaNature?Milwaukee,MN:MarquetteUniversityPress.——1993:Warrant,2vols.Oxford:OxfordUniversityPress.Pojman,L.(ed.)1987:PhilosophyofReligion:AnAnthology.Belmont,CA:Wadsworth.Proudfoot,W.1976:GodandSelf.Lewisburgh:BucknellUniversityPress.——1985:ReligiousExperience.Berkeley:UniversityofCaliforniaPress.Quinn,P.andTaliaferro,C.(eds)1996:ACompaniontoPhilosophyofReligion.Oxford:Blackwell.Reichenbach,B.1972:TheCosmologicalArgument.Springfield,IL:ThomasPress.——1982:EvilandaGoodGod.NewYork:FordhamUniversityPress.Rescher,N.1985:Pascal’sWager.NotreDame,IN:UniversityofNotreDamePress.Rhees,R.1969:WithoutAnswers.NewYork:SchockenBooks.Rowe,W.andWainwright,W.J.(eds)1989:PhilosophyofReligion:SelectedReadings.NewYork:HarcourtBraceJovanovich.Schlesinger,G.1977:ReligionandScientificMethod.Dordrecht:D.Reidel.——1988:NewPerspectivesonOld-TimeReligion.NewYork:OxfordUniversityPress.Sessions,L.1994:TheConceptofFaith:APhilosophicalInvestigation.Ithaca,NY:CornellUniversityPress.Sharma,A.1990:AHinduPerspectiveonthePhilosophyofReligion.NewYork:StMartin’sPress.Smart,J.J.C.andHaldane,J.J.1996:AtheismandTheism.Oxford:Blackwell.Sorensen,R.A.1992:ThoughtExperiments.Oxford:OxfordUniversityPress.Soskice,J.M.1984:MetaphorandReligiousLanguage.Oxford:OxfordUniversityPress.Stiver,D.1996:ThePhilosophyofReligiousLanguage.Oxford:Blackwell.Swinburne,R.1977:TheCoherenceofTheism.Oxford:ClarendonPress.——1979:TheExistenceofGod.Oxford:ClarendonPress.——1996:IsThereaGod?Oxford:OxfordUniversityPress.——1998:ProvidenceandtheProblemofEvil.Oxford:OxfordUniversityPress.Taliaferro,C.1994:ConsciousnessandtheMindofGod.Cambridge:CambridgeUniversityPress.Taylor,R.1963:Metaphysics.EnglewoodCliffs,NJ:Prentice-Hall.Tennant1930:PhilosophicalTheology.Cambridge:CambridgeUniversityPress.Tracy,T.F.(ed.)1994:TheGodWhoActs.UniversityPark:PennsylvaniaStateUniversityPress.Trigg,R.1989:RealityatRisk.London:Harvester.——1993:RationalityandScience:CanScienceExplainEverything?Oxford:Blackwell.Wainwright,W.1981:Mysticism:AStudyofItsNature,CognitiveValue,andMoralImplications.Madison:UniversityofWisconsinPress.——(ed.)1996:God,Philosophy,andAcademicCulture.Atlanta,GA:ScholarsPress.487\nCHARLESTALIAFERROWestphal,M.1984:God,Guilt,andDeath.Bloomington:IndianaUniversityPress.Whitehead,A.N.1978:ProcessandReality.NewYork:FreePress.Wierenga,E.1989:TheNatureofGod.Ithaca,NY:CornellUniversityPress.Wolterstorff,N.1976:ReasonWithintheBoundsofReligion.GrandRapids,MI:Eerdmans.——1982:GodEverlasting.InS.M.CahnandD.Shatz(eds)ContemporaryPhilosophyofReligion.Oxford:OxfordUniversityPress.Wykstra,S.1984:TheHumeanObstacletoEvidentialArgumentsfromSuffering.InternationalJournalforPhilosophyofReligion,16,73–93.Yandell,K.1993:TheEpistemologyofReligiousExperience.Cambridge:CambridgeUniversityPress.Zagzebski,L.T.1991:TheDilemmaofFreedomandForeknowledge.NewYork:OxfordUniversityPress.DiscussionQuestions1Onthemeritsandproblemsofverificationism:HowmightanempiricistgoaboutassessingthemeaningfulnessoftheclaimthatGodisnon-physical,all-goodandcreative?IfGodisnon-physical,doesitfollowthatGodcannotbeperceived?Ifitisallowedthatwecanperceive(orsomehowexperience)God,canGod’sbeingomnipotentorbeingeverywherepresentbesomethingperceivable?WhatabouttheotherDivineattributes?AssuminganempiricisttheoryofmeaningandtheexistenceofGod,howwouldGodknowaboutthecosmos?Tobeomniscient,wouldGodhavetohaveamassivenumberof(orperhapsinfinitelymany)sensoryexperiences?2Assessnon-realistapproachestoreligiousbelief.Mostversionsofnon-realismassumeasubstantially‘allornothing’position.Couldamodifiedversionofnon-realismbedefendedinwhich,say,thepracticeofprayeristreatedcognitively,whilebeliefintheafterlifeistreatedalongnon-realistlines?IsitintelligibletohavefaithinGodwithoutbelievingthatGodexists?Doesitmakeanymore(orless)sensetohavehopethatGodwilldosomethingwithoutbeliefthatthereisaGod?3Givenyourunderstandingofthefollowingterms,asusedtodescribehumanbeings,wouldusingthemaspredicatestodescribeasupremelyexcellent,non-physicalreality,requireusingthemanalogicallyorcouldtheybeusedunivocally?:knowledge-able,powerful,creative,good,intelligent,lovingandwise.Alternatively,doyouthinkanyofthemwouldhavetobeusedequivocally?4Howmightone’stheoryofpersonsaffectone’stheoryaboutGod?CoulditbethatallcreatedpersonsarephysicalwhileGodisapersonthatisnon-physical?Howmanyofthetraditional,theisticDivineattributescouldbepreservedontheassumptionthatGodisphysical?Inwhatrespectscanorshouldone’sviewofGodbeinfluencedbyone’sunderstandingofgenderandrace?5Illustratesomeofthewaysinwhichone’sphilosophyofreligionisinfluencedbyotherareasofphilosophy,includingmetaphysics,epistemology,ethics,thephilosophyoflanguageandthephilosophyofhistory.Alternatively,consideracasewhereone’sviewsinthephilosophyofreligionwillhaveaninfluenceonone’sviewoftheseotherareas.6Criticallyanalysethefollowingargument.IfGodisall-powerful,Godcanmakeastonetooheavyforanyonetolift.IfGodisall-powerful,Godcanliftanystone.Godisnotall-powerful,becauseGodcannotcreateastonetooheavyforanyonetolift488\nPHILOSOPHYOFRELIGION(becauseGodcanliftanystone)orGodcannotliftanystone(ifGodsucceededincreatingastonetooheavyforGodtolift).7Istheexistenceoffreedomcompatiblewithforeknowledge?Ifnot,canevenGodbefree?8AnalysesofomnisciencetypicallyfocusonthescopeofGod’sknowledge.ShouldthewayinwhichGodknowsbeconsideredaspartofthemeaningofwhatitistobeomniscient?Ifabeingknewalltruepropositionsbutonlyhadthisknowledgebyvirtueofthetestimonyofothersorthroughreliablebutcircuitousmeans(havingtorelyonhighlytechnicalequipment)wouldyouconsiderthisbeingtohavethedivineattributeofomniscience?9Ifabeingisomnipotent,mustitalsobeomniscient?Towhatextent,ifany,canabeingcreateitself?WhatisyourviewoftheMenzel–Morrismachine?10DoyoufindcoherentthethesisthatGodiseternal?IfGodwereeternal,couldGodknowwhattimeitisnow?CouldGodhavecreatedtime?11WhatisyourassessmentoftheidealobservertheoryandhowthismightarticulateaGod’s-eyeethicalpointofview?Cantheisticvoluntarismbedefended?12CanthefailuretoexperienceGodbeemployedinaplausibleargumentprovidingevidencethatthereisnoGod?WhatisthedifferencebetweenexperiencingtheabsenceofGodandnotexperiencingGod?13AssesstheargumentsforandagainstGod’sexistenceinthischapter,including:argumentsfromreligiousexperience,ontologicalarguments(considerboththetradi-tionaltheisticversionandtheversioncitedaboveinwhichonearguesagainstGod’sexistence),cosmologicalarguments,teleologicalarguments,theproblemofevil,argu-mentsfrommiraclesandwagerarguments.Howmighttheseargumentsbeinter-connected,sothattheforceofoneargumentdependsontheothers?14Doesoneneedasuccessfultheisticargument,suchastheteleological,cosmolog-icalorontologicalargument,inorderforthebeliefthatthereisaGodtobejustified?Isthereaburdenofproofthatisbornemorebythereligiousbelieverthanthenaturalist?15Inwhatway,ifany,maytraditionalorotherconceptionsofanafterlifebeemployedinaddressingtheproblemofevil?16WhataretheethicalimplicationsoftherebeingagoodGodwhoisaffectivelyresponsivetothecosmos–grievedbytheillsandtakingpleasureingoodssuchasjusticeandfriendship?489\n16AppliedEthicsJOHNHALDANERecentphilosophyhasreturnedtoatraditionalconcernwithprovidingcriticalassess-mentofconductandinstitutions.Undertheterm‘appliedethics’areasasdiverseashealthcare,warfare,businessandtheenvironmenthavebeensubjecttoclosescrutinyintermsofrivalethicaltheoriesorwithinthescopeofaspecificethicaltheory.Thechapterexaminestheriseofappliedethicsagainstthebackgroundoftwentieth-centurydevelopmentsinethicsandpoliticalphilosophyandsuggeststhattherearedangersinitscurrentpopularity.Itconsidersaprecedentforappliedethicsandexaminesitspos-siblefuturedevelopment.Readersmaywishtoreadthischaptertogetherwiththechap-tersonBIOETHICS,GENETHICSANDMEDICALETHICS(chapter17),ENVIRONMENTALETHICS(chapter18)andBUSINESSETHICS(chapter19).1WhatisAppliedEthics?Intryingtounderstandwhatappliedethicsis,andhowithasdevelopedwithinEnglish-speakingphilosophyoverthelasttwentytothirtyyears,itisusefultobeginbymakingathreefoldorthree-leveldistinction,betweenMORALITY(chapter6),MORALTHEORY(pp.212–14)andMETA-ETHICS(pp.225–7).Thefirstoftheseconsistsofindividualclaimsaboutwhatisofmoralvalue,suchashonestyandloyalty,andaboutwhatoughttobedoneoravoidedingeneralandonparticularoccasions:forinstance,alwayshelpthoseingreatneed,orneverinflictunnecessarypain.Sometimesclaimsofthesesortscomeingroupsandconstituteamoraloutlookorsystem.Whenthisissoitbecomespos-sibletospeakofamoralityoranethic–suchasaChristianmoralityoranecologicalethic.Amoralityinthissenseisabodyofmoralclaimsusuallyexpressingacertainkindofconcernorcommitment.Itislikely,butnotlogicallyrequired,thatamoralitywillbeaccompaniedbyoneormorePRINCIPLES(pp.733–6)byreferencetowhichparticularclaimsarejustified.Thisbringsustothesecondofthethreelevels,forthearticulationofsuchprincipleswithinasystematicstructureiswhatconstitutesamoraltheory.Hereitisimportanttonotetwopoints:first,thatonecanmakemoralclaims,andmoregenerallypossessamoral-ity,withouthavingamoraltheory;and,second,thatthesamemoralclaims,andeventhesamemoralitiesorethics,arecompatiblewithdifferentmoraltheories.Thus,for\nAPPLIEDETHICSexample,onemayholdthatlyingiswrongwithouthavingasystematicaccountofwhatmakesitwrong,andmoreoverthismoralclaimmightbejustifiedbyquitedistincttheoriessuchasthoseofconsequentialism,deontologyorDIVINELAW(pp.223–4).Theimmoralityoflyingmight,forexample,betakentoderivefromitstendencytocauseunhappiness,itsintrinsicwrongnessoritsbeingatvariancewithadivinecommandment.Thethirdlevelisdifferentfromtheothertwoinasmuchasitisnotconcernedwithwhatisrightorwrong,goodorbad,andwithwhythingsareso,butratherconcernsthelogicalstatusofmoralclaimsandmoraltheories.Meta-ethicsisthemostphilo-sophicalandabstractformofthinkingaboutmorality.ItaddressessuchMETAPHYSICAL(chapter2)andEPISTEMOLOGICAL(chapter1)questionsaswhethervaluesareobjective,andwhatthatmightmeanandwhetherthoughtsaboutthemcanconstitutegenuineknowledge.Whileanumberofthingsnowgobythedescriptionof‘appliedethics’thecentralaspectofthisconsistsinthemoreorlesssystematicapplicationofmoraltheorytopar-ticularmoralproblems(intermsofthethreefolddivision,theapplicationofsecond-leveltofirst-levelthinking).Initsearlierstagesthesewereprincipallylifeanddeathissues,suchasthoseofabortion,euthanasia,suicideorwarfare.Morerecently,however,therehasbeenaconsiderableincreaseofinterestamongcollegeanduniver-sitystudents,professionalphilosophersandthoseoutsidehighereducationintheexaminationofmoralissues,andthisinteresthasledtoanextensionofmedicalethicsfromquestionsofabortionandeuthanasiaandtothedevelopmentofnewfieldsofappliedethicssuchasbusinessethics,computerethics,environmentalethics,genderethics,journalismethics,reproductiveethicsandsoon.Indeed,thepointhasnowbeenreachedatwhichthereareveryfewifanyareasofhumanactivitythatarenotthesubjectofsomenamedbranchofappliedethics.Whetherthisisagoodthingisitselfaninter-estingandimportantquestion.2TheRiseofAppliedEthicsTheriseandexpansionofthisareaofphilosophyisduetoanumberofcauses,somemorecreditabletoitspractitionersthanothers.TobeginwiththereisthefactthatforthefirsthalfofthetwentiethcenturyEnglish-speakingmoralphilosophyconsistedalmostexclusivelyofmeta-ethics.Themainconcernsofthiswere,andremain,whethervaluesandrequirementsarefeaturesoftheworldormerelyproductsofhumanpreferenceandcommitment,andwhetherthelanguageofmoralityisdescrip-tive,expressiveorprescriptive.Discussionsofthesequestionsbecameevermoredistantfromactualmoralthinking,inpartbecausebytheirnaturetheywereabstractedfromparticularmoraloutlooks,butalsobecausethelanguageinwhichtheywerepresentedwasatechnicalphilosophicaloneowingnothingtofamiliarethicalvocabularies.Althoughduringthisperiodthestudyofmeta-ethicsproducedsomeverygoodana-lyticalphilosophy,moralproblemswereleftentirelyunaddressedandtheintermediarylevelofmoraltheorywasalsoneglected.Thiscannowseemveryoddbecausethroughoutthesameperiodtherehadbeentwoworldwars,theriseofideologicaltotalitarianism,widespreadattemptsat491\nJOHNHALDANEgenocide,andthedevelopmentanduseofweaponsofmassdestruction.Buthoweveritwasthatphilosophersjudgeditnoneoftheirprofessionalbusinesstodiscussthesematters,thepressureincreasedforsystematicmoralthinkingaboutfundamentalvaluesandthebasisofconduct.Bythemid-1960sachangeinattitudesandpracticewasbecomingapparent.TheexperienceofstudentsandacademicsintheUnitedStateswasperhapsthemostimportantsocialfactorintheriseofappliedethics.Questionsofcivilrights,ofsexualethics,ofthemoralityofwarfareandofbioethicsbecamepromi-nentthemesofpublicdebateandthesocietybegantodividebetweenconservativesandradicals,absolutistsandrelativists.Settingasidethequestionofparticularviewpoints,whatbecameclearandwasfelttobesomethingofascandalwasthatprofessionalmoralphilosophyappearedtohavenothingtosayabouttheseimportantmoralissues.Inearliertimes,aprofessionalresponsemighthavebeenthataphilosopherhasnomorebusinesssayinghowoneshouldactthanachemisthassayingwhetherthereshouldbemoreorfewersynthesizedcompoundsintheworld–ineachcasethepro-fessionaldutyistounderstandthenatureofreality,nottochangeit.Butthissortofreplycametoseemunacceptable.First,fromitsearliestdays,philosophyhasbeenconcernedwithboththeoreticalandpracticalreasoning.Aswellasaskingwhatoneshouldbelieveabouttheworld–thatis,whatistrue?–ithasaddressedthequestionabouthowoneshouldactinit–thatis,whatisitgoodtodo?Second,developmentsinmeta-ethicsthemselvessteadilyweakenedthecasefornon-involvement.Onewidelyheldreasonforthinkingthatmoralphilosophycouldnotprovideaguidetoactionwasthebeliefthattotheextentthatsystematicreflectionshowedanythingaboutmoralityitrevealedittobeamatterofsubjectiveattitudesorcommitmentsaboutwhichtherecouldbenorationalargument.Onemightperhapsshowthatsomeonewasactingirra-tionallygiventhemoralvaluesandoutlookheorshepossessed(forinstance,bybeinginconsistent),butonecouldnotdemonstratethatitwascontrarytoreasontohavethosevaluesandoutlook,fortheyweresimplymattersofsubjectivepreference.Bythelate1960s,however,meta-ethicalsubjectivismhadbeguntolosegroundtomoreobjec-tivistviews.IntheUnitedKingdomtherewasarevivalofethicalnaturalismintheformofthethesis,advancedbywriterssuchasElizabethAnscombe,PhilippaFootandPeterGeach,thatmoralclaimsrelatetohumanwell-beingandareassessableastrueorfalse,reasonableandunreasonable,byreferencetothisobjectivenorm.IntheUnitedStates,meanwhile,otherearliertraditionswerebeingrevived,inparticularcertainformsofethicalrationalismwhicharguethatquestionsofconductcanbeansweredbyreferencetostandardsofwhatareuniversallyreasonablerulesofaction.SomeofthemostinterestingdevelopmentsinNorthAmericawhichinfluencedthecourseofappliedethicsthereandelsewhereemergedoutofPOLITICALPHILOSOPHY(chapter8)andinturnaretraceabletocertainstrandsinmodernepistemology.ClearandhistoricallyimportantexamplesofthisaretobefoundinJohnRawls’sbookATheoryofJustice(1971).ThereheisconcernedtoarticulateprinciplesofJUSTICE(pp.258–64)thatmightunderliethepoliticalinstitutionsandpracticesofaliberalstate.Inordertoarriveatthese,heemploysaversionofsocialcontracttheoryinvitingustoconsiderwhatprincipleswewouldrationallychoosetohaveregulateourlivesifwewerenottoknowinadvancewhatoursocialcircumstanceswouldbe.Thisproceduretherebygivesaformofrationaljustificationforsuchprinciples.However,inthinkingatatheoreticallevelaboutwhatprinciplesdoorwouldguideourconduct,itsoon492\nAPPLIEDETHICSbecomesapparentthatsuchrationalesmaynotentirelysquarewithourmoralincli-nationsandpre-reflectivejudgements.Herewefaceaproblemofwhethertorevisetheprinciplesorabandonthemoralintuitions.Rawls’sdiscussionofthisissueinvolvestheideaofreflectiveequilibrium.Thegeneralprinciplesimplicitinourattitudestorealorimaginedsituationsmayneedtobeamendedinthefaceofcontraryjudgementsoffurther,oftenborderline,specificcases.Bythesametoken,whereaprincipleitselfisshowntobewell-attestedinothercircumstanceswemaythenneedtoreviseourintuitivejudgementofthetroublesomecase.Thepointhereisthatweshouldseektomaximizecoherenceinourmoralandsocialoutlook,andsuch‘equilibrium’isplausiblyarationalrequirementofreflectiononquestionsofvalueandconduct.Whatevertherightsandwrongsofthissortofmoralepistemology,it,andtherevivalofotherbroadlyobjectivistethicaltheories,providedbothanexampleandanimpetustoappliedethics.Ingeneralonefindstwomethodsbeingemployed:firstly,thatjustdescribed–thatis,thederivationofprinciplesfrompre-theoreticalmoraljudge-mentsandtheirextensiontoothercases,constrainedbydemandsofconsistencyandcoherence;and,secondly,thedirectapplicationofwell-establishedmoraltheories,suchasUTILITARIANISM(chapter35),todeterminegeneralpoliciesandtoresolveparticularproblems.Sincethe1970stheemploymentofthesemethods,bywriterssuchasRichardHare,AnthonyKenny,ThomasNagel,PeterSinger,MaryWarnockandBernardWilliams,hasproducedsomeinterestingworkonethicalandsocialquestionswhichwillsurvivejustaswellascontemporaryresearchinmoretheoreticalareasofphilosophy.Earlier,however,Isuggestedthatnotallfeaturesofthegrowthinappliedethicsaretobewelcomed.WhatIhadinmindwastheunseemlyrushofsomephilosophersandthoseonthefringesofthesubjecttobecomepartofwhathasproventobearapidlyexpandingandinstitutionallysuccessfulareaofacademicpractice.Itisnotjustthatopportunismisundignified.Thepointisratherthatthroughpoor-qualityproductstheveryactivityrisksfallingintodisrepute,andtherebyanopportunityforpracticalphilosophytore-establishitselfasamajoraspectofseriousstudymaybelostasotherphilosophersturnagainstwhattheyperceivetobeintellectuallyshallowwork,andthepublicpatronsofappliedethicscometofeeldissatisfiedwithit.TheMoralDangerofAppliedEthicsThislatterprospectalsoraisesamoralconcernakintothatwhichpartlymotivatedSocratesinhisattackontheSophists,thepurveyorsofmoralandpoliticalwisdomintheGreekcity-statesofthefifthcenturyBC.Althoughitisoftensupposedotherwise,evenbythosewhoshouldknowbetter,moralphilosophyisnoeasiertopractisethananyotherareaofthesubject.Unlikephilosophyoflanguage,say,itsinitialstarting-pointsareusuallycommonplacesofreflectiveexperience:forexample,thequestionoftruth-tellingofteninvolvesdilemmas,buttheattempttostructuresatisfactoryargumentsaboutsuchmattersinvolvesdeepanddifficultproblems,withtheconsequencethaterrorandcon-fusionarealltoocommon.ThesuggestionoftheSophistsandofsomecontemporaryappliedethicists–thatthereareskillsthemoreorlessmechanicalemploymentofwhichwillyieldanswerstomoralquestions–isbothadisservicetophilosophyanda493\nJOHNHALDANEcorruptinginfluenceontheconsciencesofthosewhomitclaimstobeassisting.More-over,asSOCRATES(chapter22)saw,thereisadangerthatwhensophistryisexposed,people’sreactionisnottoseekbettercounselbuttoconcludethattherecanbenosoundreasoningaboutquestionsofvalueandconduct,supposinginsteadthatthesearejustmattersofpersonalpreferences.Itwouldbeaterribleironyiftherecentriseofappliedethicsledinduecoursetothespreadofmoralirrationalism.3ANeglectedPrecedentAlthoughthetermisarecentone,thereisinfactnothingnewinthegeneralideaofappliedethics.Whileitistruethattherehavebeenperiods,suchastheearlierhalfofthetwentiethcentury,whenphilosopherswerealmostexclusivelyconcernedwithspeculativequestions,Inotedearlierthatfromantiquitytheimportanceofpracticalaswellastheoreticalphilosophyhasbeenrecognized.Itisworthobserving,therefore,thattherewas(andtosomeextentremains)averylongandwell-developedtraditionofthinkingaboutpracticalproblemsinthelightofgeneralprinciples,namelythatofmoralcasuistry.Itshouldbesaidimmediatelythattheterm‘casuistry’isnowmostoftenusedperjorativelytocondemnsomepieceofthinkingasattemptingaspeciousjustificationorasinvolvingequivocationorhypocrisy.Interestingly,thiscriticalusagearosefromperceivedorimaginedabusesofpracticalreasoningsimilartothoseattributedtotheSophists.Inthisconnectionsee,forexample,theridiculingofJesuitcasuistrybyPASCAL(1623–62)(pp.483–4)inhisProvincialLetters(1967).However,theauthentictradi-tionderivesfromthesincereefforttoapplyrigorousstandardsofcriticalargumenttotheresolutionofquestionsofconduct.Historically,itdevelopedthroughmoraltheologyasChristianstriedtoworkoutthedetailedimplicationsofscripturaltextsandtheteachingsoftheearlyChurchFathers.Inthemiddleages,writerssuchasThomasAQUINAS(1225–74)(p.619)advancedquitecomplextheoriesofmoralreasoninginwhichvariouselementsofethicaljudgement(synderesisandconscientia)werecon-nectedwithanaccountofrightreasoning(rectaratio).ThistraditionsurvivedtheReformationandwasfurtherdevelopedbybothCatholicandProtestantauthors,suchasFranciscoSuarez(1548–1617)andHugoGrotius(1583–1645),andlaterstill,inincreasinglysecularversions,byImmanuelKANT(1724–1804)(chapter32)andJeremyBENTHAM(1748–1832)(chapter35).Aswellasbeingconcernedwiththeformsofmoralreasoning,theSCHOLASTIC(pp.621–2)casuistswereinterestedinresolvingmajordilemmasinparticularfieldsofhumanactivity,suchasinsexualmatters,economicaffairsandtheconductofwarfare.Theirwritingsontheformandsubstanceofmoralitycontainmanyinsightsthatarestillworthrecovering.Oneaspectofthesewritingsrelevanttotheassessmentofappliedethicsasitisnowpractisedisthewayinwhichthecasuistsrelatedmoralproblemsandmoraltheory.Thepresent-dayexpressions‘appliedethics’and‘appliedphilosophy’suggestatwo-stageprocess.Firstly,oneworksoutsomegeneralphilosophicalprinci-ples,thenoneappliesthemtoagivenissue.Onthisaccountthereneedbenothingphilosophicalaboutthesecondstage;ratheritshouldbeanalgorithmicormechanical494\nAPPLIEDETHICSprocess.Forthemedievalandmoderncasuists,bycontrast,thetreatmentofparticu-larcasesisitselfanexerciseofmoralreasoningnolessthanistheworkingoutofgeneralprinciples.Indeed,thereneednotbetwodistincttasksinvolved.Rather,thedeterminationofthegeneralandtheparticularmaybeaspectsofasinglecomplexproceduresimilarinsomerespectstotheestablishmentinRawlsianmethodologyofastateof‘reflectiveequilibrium’.Afurthermarkofthedifferencebetweencasuistryorpracticalmoralphilosophyandappliedethicsisrevealedinthefactthatwhereascontemporarywritersusuallysupposethattherearenotopic-specificmoralvaluesbutonlytopic-indifferentgeneralprinciples,themedievalsandmodernstendedtodevelopaccountsofvirtuesandvaluesspecifictothedifferentdepartmentsoflife.Thus,forexample,chastitywasthoughtofasadistinctlysexualvirtueandnotaninstanceof,forexample,prudence.Similarly,themedievalprohibitiononusuryderivedfromreflectionsonthevaluesimplicitindis-tinctiveformsofsocialrelationships.Perhapsthemostimportantareaofcasuisticalthought,however,wasinrelationtolarge-scaleviolence,concerningwhichitdevel-opedthedoctrineofjustwarwithitsvarioustopic-specificmoralvalues,suchasthatofnon-combatantimmunity.4TheFutureoftheSubjectContemporarymoralphilosophersconcernedwithpracticemightthereforedobettertoadoptsomethingoftheapproachofcasuistryandpracticalphilosophyratherthantothink,asatpresent,inthesomewhatlimitingtermssuggestedbythephrase‘appliedethics’.Thereiscertainlyneedandconsiderablescopeforthephilosophicalstudyofthepracticeofthinkingaboutpractice–thatistosay,thereisarequirementfora‘meta-philosophical’accountofmoralreasoning.Ifsuchanenquiryweretoleadtoareconceptionofthesubjectofnormativeethicsalongthelinessuggestedinconnectionwithearlierstylesofthought,thatwouldcer-tainlyhaveimplicationsforitsfuturedevelopment.Asitis,therecanbelittledoubtbutthatthegrowthofpublications,collegeanduniversitycourses,in-serviceprofessionaltrainingsessions,andotherformsofactivityconcernedwiththephilosophicaltreat-mentofmoralandsocialquestionswillcontinueatmoreorlessitspresentrateuntilatleasttheendofthetwentiethcentury.Itisimpossibletobeconfidentaboutdevelopmentsbeyondthatpoint,butitseemslikelythatappliedethicistswilllookfornewfieldsofoperation.Forthemostpart,thatwillinvolveseekingoutexistingareasoflifethathavenotyetbeenmadethesubjectofacademicmoralenquiry;butitwillalsoleadtogreaternoveltyastechnologyextendsthelimitsofthepossibleandtherebychangespatternsofbehaviour,andasotherfactorsbearuponthestructureofpersonalandsocialrelationships.Wehaveseenhowthedevelopmentofelectronicdata-gatheringanddata-recordinghasgivenrisetoincreasedconcernsaboutprivacy,andhowreproductivetechnologyhaspromptedworriesabouttheproperlimitsofcontroloverthecreationoflife.Doubtlessatthisverymomentsomeoneiswritingabookontheethicsof‘virtualreality’technology,andtherewillsoonbeotherwritingsonthemoraldimensionsoftechnologiesnotasyetgenerallyknownof,noreven,perhaps,yetinvented.495\nJOHNHALDANEAstheseinterestsdevelopsowillthepublishingandotherservicesthatcaterforthem.Thirtyyearsagotherewerenoacademicphilosophyjournalsconcernedpri-marilyorevenlargelywithnormative,first-ordermoralandsocialquestions.Today,suchjournalsaboundandmorearecreatedeveryyear.Thepresentrangeincludesthefollowing:AgricultureandHumanValues,Bioethics,BusinessEthics,BusinessandProfes-sionalEthicsJournal,CriminalJusticeEthics,EnvironmentalEthics,EnvironmentalValues,EthicsandAnimals,EthicsandMedicine,InternationalJournalofAppliedPhilosophy,Inter-nationalJournalofMoralandSocialStudies,InternationalReviewofEconomicsandEthics,JournalofAgriculturalEthics,JournalofAppliedPhilosophy,JournalofBusinessEthics,JournalofEnvironmentalEthics,JournalofMedicalEthics,JournalofMedicineandPhi-losophy,JournalofReligiousEthics,JournalofSocialPhilosophy,PhilosophyandPublicAffairs,PhilosophyandTechnology,PublicAffairsQuarterly,SocialPhilosophyandPolicyandSocialTheoryandPractice.Howeverthesubjectdevelops,therewillbenoshortageofjournals–such,forgoodandill,isacademicprofessionalism.FurtherReadingHare(1981)offersanadvancedbutclearlypresentedaccountoftheauthor’s‘prescriptivist’theoryofmoralitydevelopedinrelationtodifferentlevelsofthinkingaboutmoralproblems.Kenny(1985)providesaseriesofclearlyarguedandwell-balancedessays,thecoreofwhichconcerntheethicsofwarandnucleardeterrence.Nagel(1979)haswrittenacollectionofsubtlyarguedessaysmanyofwhichdevelopdeontologicalwaysofthinkinginconnectionwithvariousmoralproblems.Singer(1979)isaclearanddirectaccountofutilitariantheoryandofitsapplicationtoarangeoflifeanddeathissues.Itisveryusefulforseeingtheattractionsandtheproblemsofconsequentialism.Warnock(1992)presentsaselectionofwritingsbyaphilosopherwhoisprominentinpubliclifeintheUnitedKingdom,whichexplorethepossibilitiesandlimitationsofphilosophyinrelationtopracticalproblems.Williams(1981)offersacollectionofinfluentialessaysbyanotherprominentpublicphilosopherdevelopingtheideaoftheimportanceofpersonalcharacterinrelationtomorality.ReferencesMostofthefollowingbooksarecollectionsofessaysbyoneormoreauthors.However,Rachels(1979),Singer(1986)andVelasquezandRostankowski(1985)areanthologiescontaining,inwholeorinpart,someofthemostinfluentialpapersinvariousfieldsofappliedethicspublishedinthelasttwentytothirtyyears.GeneralworksinappliedethicsDeMarco,J.P.andFox,R.M.(eds)1986:NewDirectionsinEthics:TheChallengeofAppliedEthics.London:RoutledgeandKeganPaul.Evans,J.D.G.(ed.)1987:MoralPhilosophyandContemporaryProblems.Cambridge:CambridgeUniversityPress.Griffiths,A.P.(ed.)1985:PhilosophyandPractice.Cambridge:CambridgeUniversityPress.Hare,R.M.1981:MoralThinking.Oxford:ClarendonPress.——1989:EssaysinEthicalTheory.Oxford:ClarendonPress.Kenny,A.1985:TheIvoryTower:EssaysinPhilosophyandPublicPolicy.Oxford:Blackwell.Louden,R.B.1992:MoralityandMoralTheory.Oxford:OxfordUniversityPress.496\nAPPLIEDETHICSNagel,T.1979:MortalQuestions.Cambridge:CambridgeUniversityPress.Nuttall,J.1993:MoralQuestions.Cambridge:PolityPress.Pascal,B.1967[1656]:ProvincialLetters(translatedbyA.J.Krailsheimer).Harmondsworth:PenguinBooks.Phillips,D.Z.1992:InterventionsinEthics.London:Macmillan.Rachels,J.(ed.)1979:MoralProblems.NewYork:HarperandRow.Singer,P.1979:PracticalEthics.Cambridge:CambridgeUniversityPress.——(ed.)1986:AppliedEthics.Oxford:OxfordUniversityPress.——(ed.)1991:ACompaniontoEthics.Oxford:Blackwell.Spaemann,R.1989[1982]:BasicMoralConcepts(translatedbyT.J.Armstrong).London:Routledge.Velasquez,M.andRostankowski,C.(eds)1985:Ethics:TheoryandPractice.EnglewoodCliffs,NJ:Prentice-Hall.Warnock,M.1992:TheUsesofPhilosophy.Oxford:Blackwell.Williams,B.1981:MoralLuck.Cambridge:CambridgeUniversityPress.Winkler,E.andCoombs,J.(eds)1993:AppliedEthics:AReader.Oxford:Blackwell.AnimalrightsFrey,R.1983:Rights,KillingandSuffering.Oxford:Blackwell.Regan,T.1983:TheCaseforAnimalRights.Berkeley:UniversityofCaliforniaPress.BusinessethicsBeauchamp,T.andBowie,N.1979:EthicalTheoryandBusiness.EnglewoodCliffs,NJ:Prentice-Hall.Velasquez,M.1982:BusinessEthics.EnglewoodCliffs,NJ:Prentice-Hall.EnvironmentalethicsAttfield,R.1991:TheEthicsofEnvironmentalConcern,2ndedn.Athens,GA:UniversityofGeorgiaPress.Rolston,III,H.1988:EnvironmentalEthics:DutiestoandValuesintheNaturalWorld.Philadelphia,PA:TempleUniversityPress.GenderissuesGilligan,C.1982:InaDifferentVoice:PsychologicalTheoryandWomen’sDevelopment.Cambridge,MA:HarvardUniversityPress.Jaggar,A.1983:FeministPoliticsandHumanNature.Totowa,NJ:RowmanandLittlefield.MedicalethicsHarris,J.1985:TheValueofLife:AnIntroductiontoMedicalEthics.London:RoutledgeandKeganPaul.Sommerville,A.etal.1993:MedicalEthicsToday.London:BritishMedicalAssociation.PunishmentHonderich,T.1984:Punishment:TheSupposedJustifications.Harmondsworth:PenguinBooks.Primoratz,I.1989:JustifyingLegalPunishment.AtlanticHighlands,NJ:HumanitiesPress.SexScruton,R.1986:SexualDesire.London:Methuen.Soble,A.(ed.)1980:PhilosophyofSex.Totowa,NJ:RowmanandLittlefield.497\nJOHNHALDANEWarGlover,J.1987:CausingDeathandSavingLives.Harmondsworth:PenguinBooks.Teichman,J.1986:PacifismandtheJustWar.Oxford:Blackwell.Walzer,M.1977:JustandUnjustWars.Harmondsworth:PenguinBooks.DiscussionQuestionsQuestionsaboutappliedethics1Shouldallareasofhumanactivitybesubjecttoabranchofappliedethics?2Howcloseshouldthemainphilosophicalproblemsofethicsbetoactualmoralthinkingandfamiliarethicalvocabularies?3Howcanwejudgewhetheritistheprofessionalbusinessofmoralphilosopherstodiscussmatterssuchastotalitarianism,genocide,warfare,civilrightsandsexualethics?4Doeseitheraconceptionofhumanwell-beingoranotionofuniversallyreason-ablerulesofactionprovideasatisfactorybasisfortheobjectiveassessmentofconductandinstitutions?5Howcanwedeterminewhethertherearerightmethodsinappliedethics?6Aredebatesinappliedethicsmorelikelytoentrenchortochallengewell-establishedmoraltheories,likeutilitarianismorKantianethics?7Doesappliedethicscorruptourconsciences?Howcanthisdangerbeavoided?8Howshouldpracticalphilosophyrelatethoughtaboutgeneralprinciplesandthoughtaboutparticularcases?Whataretheconsequencesofyouranswerforthepossibilityofappliedethics?9Aretheredifferentvaluesanddifferentvirtuesspecifictodifferentdepartmentsoflife?Howshouldweunderstandappliedethicsinthelightofyouranswer?10Howmightameta-philosophicalaccountofmoralreasoninghelpustochoosebetween‘appliedethics’and‘casuistry’asthemainformofmoralphilosophyaboutpractice?11Inwhatnewfieldsmightappliedethicsfindafieldofoperation?Questionswithinappliedethics12Whatdistinguishesjustwarsfromunjustwars?13Isdestroyingacultureawronginadditiontokillingthepeoplewhoarethebearersofthatculture?14Isthereanymoraldifferencebetweenabortionandinfanticide?15Canwerespecttheautonomyofmentalpatients?16Doanimalshaverights?17Doesthefactthathumansliveinanenvironmentgiverisetomoralconstraintsonhowweshouldtreatthatenvironment?18Doesthemarketanswerallquestionsofbusinessethics?19Shouldpornographybebanned?498\n17Bioethics,GenethicsandMedicalEthicsREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLMBioethicsinvestigatesethicalissuesarisinginthelifesciences(medicine,healthcare,genetics,biology,researchandsoon)byapplyingtheprinciplesofMORALPHILOSOPHY(chapter6)totheseproblems.Medicalethicsandgenethics(ethicalissuesarisingfromthedisciplineofgenetics)aresubsetsofbioethics.1Bioethics1.1MethodologyThereisoftenmuchconfusionaboutwhatthemethodologyofbioethicsis.Bioethicsischaracterizedasamultidisciplinarymodeofenquiry.Healthcareprofessionals,lifescientists,philosophers,theologians,lawyers,economists,psychologists,sociologists,anthropologistsandhistoriansareamongthosewhoaretypicallyinvolvedinbioethi-calenquiry.However,whileawiderangeofdisciplinesareactivelyinvolvedinbioethics,thecentralmethodofbioethicsismoralphilosophicalenquiry.Bioethics,ratherthanbeingamultidisciplinarymodeofenquiry,isabranchofAPPLIEDETHICS(chapter16),whichischaracteristicallyinformedbymultidisciplinaryexpertiseandfindings.AsRonaldGreenputsit:‘whileethicsandmoralphilosophymaysometimesrepresentarelativelysmallpartoftheactualworkofbioethics,theyforminasensetheconflu-encetowhichallthelargerandsmallertributarieslead,and,morethananyothersingleapproach,themethodsofethicsandphilosophyremainindispensabletothisdomainofinquiry’(Green1990:182).Appliedethicsinvolvestheapplicationoftheprinciplesandmethodsofmoralphilosophytopracticalproblems.Bioethics,abranchofappliedethics,appliestheseprinciplesandmethodsofmoralphilosophytoissuesarisinginthelifesciences.1.2TheoreticalapproachesWithinbioethicsanumberoftheoreticalapproacheshavebeenusedasthebasisforethicalanalysisofconcretebioethicalproblems.Allthemajorapproachesinmoralphilosophyhavealsobeenappliedwithinbioethics,althoughofteninformsthatare\nREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLMsomewhatsimplifiedincomparisontothemostadvancedcontemporarydevelopmentsoftheseapproaches.KANTIANETHICS(chapter32)isforinstanceoftenappropriatedinbioethicsasifitconsistedonlyofthetwoformulationsoftheCATEGORICALIMPERATIVE(pp.735–6),andbioethicalconsequentialismoftenneglectstheproblemsofpreciselyspecifyingthemaximandinapreferenceCONSEQUENTIALIST(pp.30–1)theory.Thepopularityofthedifferentapproacheshaswaxedandwanedovertime,andatleastoneapproachspecifictobioethicshasbeendeveloped,theso-calledprinciplistorFourPrinciplesapproach.Becausedeontology,consequentialismandvirtueethicsarecentraltheoreticalapproachesinmodernmoralphilosophytheywillnotbepresentedhere,butinthelistofreferencesthemainworksofthemostimportantproponentsofeachoftheseframe-workswithinbioethicsarepresented.Itisalsoimportanttonotethatphenomenologi-cal,existentialist,discourse-ethics,andpostmodernapproachesareprevalentinthatpartofthebioethicsliteraturewhichisnotwritteninEnglish,andwhichisthereforeoftenneglectedintheAnglo-Americanliterature.Threeapproachesthatarenotespeciallycommon,orentirelyabsent,withinmoralphilosophyingeneralhavegainedsomecurrencywithinbioethics:(1)intuitionism;(2)theFourPrinciples;(3)Pluralistviews.Aswillbecomeapparent,allthreeapproachescanbeseenasthreedifferenttypesofreactiontotwoperceiveddeficienciesinmoralphilosophy:1Thelackofagreementatthebasiclevelofmoralphilosophy.Thisisexemplifiedintheinterminabledebatebetweenconsequentialistsanddeontologists.2Theperceptionthattheoreticalelegancehasbeenprioritizedtothedetrimentofrealisticapplicabilitytoreal-lifeethicalproblems.IntuitionisminbioethicsisoftenderivedfromthewritingsofthelaterWITTGENSTEIN(chapter39),especiallyhisconsiderationsaboutrulesandrule-followinginPhilosophi-calInvestigations.Itentailsthemostradicalrejectionofthemaintraditionsinmoralphilosophy,andclaimsthatthereasonsbehindethicaldecision-makingcannotbeformulatedinstatablerules,andevenifsuchrulescouldbestatedtheywouldunderdeterminethedecisiontobemadeineveryrealisticsituationcallingforethicaljudgement.Bioethicalintuitioniststhereforecriticizemainlinebioethicsforbeingmuchtoosimplistic,andforneglectingtheimportanceoftheagents’pre-reflectiveethicaljudgements.Intuitionistswillmaintainthatifagentsimmediatelyfeelrevulsionatthethoughtofinfanticide,thenthisisamoralfactwhichcannoteasilybesetaside.A‘primitive’intuitionismwillplaceallemphasisonimmediate‘gutfeelings’asthebedrockofmoraljudgement,andwillthusstandindangerofremovingthepossibilityofmoraldebate,butmoresophisticatedintuitionistswilltrytoavoidthisdanger.TheFourPrinciplesapproach(4PA)istheonlycommonlyusedapproachthatistrulyindigenoustobioethics.TheapproachwasinventedbyThomasBeauchampandJamesChildressinthefirsteditionoftheirveryinfluentialThePrinciplesofBiomedicalEthicsin1979.ItwaspopularizedinEuropebyRanaanGilloninaseriesofarticlesintheBritishMedicalJournal,latercollectedinthebookPhilosophicalMedicalEthics(Gillon1986).The4PApositsfourprinciplesthatarethoughttooccupyamiddlelevelandto500\nBIOETHICS,GENETHICSANDMEDICALETHICSmediatebetweenanupperlevelofconflictingmoraltheoriesandalowerlevelofcommonmorality.Itisclaimedthatthefourprinciplesaresupportedbybothdeonto-logicalandconsequentialistmoraltheories,andthattheyalsoemergefromarationalreconstructionofcommonmorality(i.e.theimmediatemoralintuitions/judgements)ofordinarypeople.Thefourprinciplesare:RespectforautonomyNon-maleficenceBeneficenceJusticeThefourprinciplesarenotrank-orderedinthe4PA.Intheconsiderationofamoralproblemanagentisfirsttoconsiderwhichofthesefourprinciplesareapplicableinthesituationbylookingatthecontentandscopeofeachprinciple.Iftwoormoreprinci-plesareapplicable,andiftheydonotsupportthesamedecision,theagentmustspecifythecontentofeachprincipleandbalancetheprinciplesagainsteachotherinordertoreachafinalsolution.The4PAhasattainedgreatpopularityinmedicalethics,althoughitsimpactonacademicbioethicsseemstobewaning.Theattractionsofthe4PAareundoubtedly(1)itssimplestructure,(2)itsapparentabilitytomakeadeeperunderstandingof,andengagementwith,moraltheoryunnecessary,and(3)itsappar-entcongruencewithcommonfeaturesofmoralthinking.Thediminutionoftheimpactofthe4PAonacademicbioethicscanbetracedtosixkindsofcritiquespointingoutthat(1)thederivationofthefourprinciplesfrommoraltheoryisproblematic;(2)thedegreetowhichcommonmoralitycanbesaidtosupportthefourprinciplesisquestionable;(3)thefourprinciplesmusteitherbeseenascontentlesslabelsortheircontentmustbeseenasculturallydeterminedwithinaspecificculture;(4)althoughBeauchampandChildressdenyarank-orderingoftheprinciples,theirexamplesofapplicationgiverespectforautonomymuchgreateremphasisthanforinstancejustice;(5)thefourprinciplesdonotcoverthewholemoraldomain;and(6)theguidelinesforspecificationandbalancingaresovaguethatthe4PAinmanycircumstancesradicallyunderdeterminethesolutiontoagivenproblem.Thelabel‘Pluralistviews’isusedheretodenoteaclassofapproachestobioethicsthatattempttocombinethe‘best’insightsfromdeontology,consequentialismandvirtueethicsinsomeformofcoherentframework.Theimpetusfordevelopingapluralistviewisoftenthatallofthe‘pure’theoreticalapproachesleadtostrikinglycontra-intuitiveresultswhenappliedtocertainbioethicalproblems.Thepluralistviewstrytosolvethisproblemby,forinstance,advocatingconsequentialismbutwithcertaindeontologicalside-constraintsthatcanremovethe‘unwantedside-effects’ofpure,unbridledconsequentialism.CertainpluralistviewsstandinthetraditionofAmericanPRAGMATISM(chapter36)(followingfiguressuchasPeirceandDewey),butothermodernkindsofpragmatismseemtobeveryclosetopureconsequentialism.Themainproblemforproponentsofpluralistviewsisthatthedeepertheoreticalfoundationsofthepluralistprojectareoftenverydifficulttodefendbecausetheproponentsof‘pure’theoriescanoftenexploitthecracksinthefoundationwherethedifferenttheoriesinthepluralityarejointedtogether.501\nREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLMEthicsandLawManyareasof‘bioethics’,suchasartificialreproduction,areheavilyregulatedbyLAW(chapter13),andthisraisesthequestionoftherelationshipbetweenlawandethicsinthisfield(seeSimmonds1986:chs5–7).Itwouldbealltoosimpletosupposethatwhatethicsprovidesinthewayofprinciplesandmoralcalculusformsthefoundationoflegislationandjudicialreasoninganddecision-making,butthereis,ofcourse,someoverlapbetweenmoralphilosophyandjurisprudence.Thiskindofapproachisembodiedintheworkofso-called‘naturallawyers’whosebasictenetholdsthatthejobofthelawistoenforcenaturalrights.ThroughthisexplicitcommitmenttoatheoryofNATURALRIGHTS(p.769),theexpositoryisconflatedwiththejustificatory,anditistheviewthatthisconflationismis-takenanddangerousthatledtothedevelopmentof‘legalpositivism’,originatingintheworkofJeremyBENTHAM(chapter35).ForBentham,itistheprincipleofutility,andnottraditionalfaithinnaturalrights,whichshouldconstitutethetouchstonebywhichthelawistobejudged.WhilstlegalpositivistssuchasH.L.A.Hartwoulddisclaimneithertheimportanceofmoralitynorthefactthatmoralitycaninfluencelaw-making,theyadvo-catethedivorceofexpositoryjurisprudencefromcensorialjurisprudence.Thatis,discov-eringwhatthelawisshouldbeseenasadistinctenterprisefrommoralevaluationofthelaw.Inprinciple,wehavewellestablished‘RULESOFRECOGNITION’(p.412)fordecidingwhatlegalrulesexistandwhatrightsanddutiestheyconfer,andeverylegalsystemmustpossessafundamentalruleofrecognition.Positivistswouldclaimthatanyrelationshipbetweenlegalrightsanddutiesandmoralrightsanddutiesisatmostcoincidental.‘Whatisthelaw?’istobeviewedasaverydifferentquestionto‘Shouldthelawbeobeyed?’and,indeed,followingBentham,Hartwouldchargethatitisthroughthisseparationthatweareputinapositiontoanswerthelatter.Now,asSimmondshasargued,legalpositivismitselfreliesonaparticularmoralapproach:‘Positivistswishtoportraythelawasabodyofrulesthatcanbeascertainedinsomemoreorlessuncontroversialway,quiteindepen-dentlyofourdifferingmoraljudgements.But...thisapproachitselfrestsonanunder-standingofthedistinctivesignificanceoflaw.Itispreciselythepublicnatureoflawthatgivesititsmoralclaimonourconduct’(Simmonds1986:95).TheforemostcritiqueoflegalpositivisminmoderntimesistobefoundintheworkofRonaldDworkin(1977).Dworkinarguesthatlawencompassesnotonlyrules,butprinciples.Whilstvalidrulesareappliedinan‘all-or-nothing’fashionandcannotcon-flict,principlesmayconflictand,whentheydo,resolutionisachievedbybalancingoneagainstanother.(Tofindoneprincipledecisiveinaparticularcase,however,doesnotimplythatthatprinciplewill‘trump’inallcases.)Itisthroughthisunderstandingofthefunctionofprinciplesthatwecanexplainhow,forexample,legalprecedentcanbeamendedoroverturned.NowwhereDworkin’stheoryisexclusiveoftheclaimsofpositivismisinhowDworkinclaimslegalprinciplesaretoberecognized.Thisisnotbyreferencetoaruleofrecognition,butthroughmoraland/orpoliticalanalysis.ForDworkin,whilstapolicyisastandardwhichexpressessomegoalofthecommunity,aprincipleis‘astandardthatistobeobserved...becauseitisarequirementofjusticeorfairnessorsomeotherdimensionofmorality’(ibid.:22).Andaprinciplecountsasalegalprincipleifitisaconstituentofthe‘soundesttheoryoflawthatcouldbeofferedasajustificationfortheestablishedlegalrulesandinstitutions’(Simmonds1986:95).502\nBIOETHICS,GENETHICSANDMEDICALETHICSThus,whilstlegalpositivistswanttosaythatlawisseparatefrommoralitybecauselawhasaruleofrecognitionthatprovidesforastraightforwarddelineationofwhatstand-ardsapply,Dworkin’scounter-argumentrestsonthefactthatdevelopingthe‘sound-esttheoryoflaw’iscertainlynotstraightforwardandmustincludemoralandpoliticaltheoreticalconsiderations.IfDworkinisright,thenthelegalregulationofmedicineandthebioscienceswillultimatelyinvolvetheappealtomoralandpoliticalprinciples.1.3ThehistoryofbioethicsModernbioethicscanbetracedbacktotwodistinctroots.Thefirstrootistraditionalmedicaletiquetteandethicsandthesecondisacademicmoralphilosophy.Considera-tionsabouttheethicalimplicationsofmedicalpracticehavealwaysplayedaroleintheself-understandingoftheWesternmedicalprofessionandintheeducationofmedicalapprenticesandlatermedicalstudents.ThiswasfirstgivenexpressioninthefamousHippocraticOath,whichprobablyoriginatedinaminorityPythagoreanschoolinearlyGreekmedicine,butwhichbecameimportantbecauseitresonatedwithlaterChristianvalues.Itis,however,importanttonotethattheHippocratictraditionhasneverbeentheonlytraditioninmedicalethics.Therehavealwaysbeenseveralcompetingtraditions.Duringcertainhistoricalperiodswhentherehasbeennoreflectiononthecontentoftherulesofmedicalethicstheyhavedegeneratedtopurerulesofprofessionaletiquettetobelearnedbyrote,butinotherperiodstheruleshavebeenquestionedandthephilosophyofthedayhasbeenemployedinthecriticismandreconstructionofmedicalethics.Itispossible,intheEnglish-speakingworld,tolocatethebeginningofmodernprofessionalethicstothe1827publicationofMedicalEthicsbytheManchesterphysicianThomasPercival.Percival’sideasformedthebasisoftheearlyethicscodesoftheAmericanandBritishmedicalassociations,andalsoinfluencedthecodesoftheWorldMedicalAssociationaslateasthe1940sand1950s.Thecornerstoneofthisconceptionofethicswasthedualresponsibilityoftheprofessionaltowardsthepatientandtowardstheprofession.Intheearly1960sthisunderstandingofethicscameunderpressureintheincreasinglycomplexhealth-caresystemsintheWesternworld.Theotherrootofbioethicsisacademicmoralphilosophy.IntheAnglo-AmericantraditionmoralphilosophyencounteredabarrenperiodinthemiddleofthetwentiethcenturywhenthemostprominentphilosophersturnedawayfromanengagementwithrealmoralproblemstodiscussionsofMETA-ETHICS(pp.225–7).Moralphilosophythuslostitsimportanceasacorrectivetothemorallyproblematicaspectsofsocialpolicy.Duringtheearly1960smanyyoungphilosophersbeganagaintoturntothemoralproblemsofmodernsociety,utilizingthephilosophicaltoolsofmodernAnglo-Americananalyticphilosophy.Thismovecanbeseenaspartofthegeneralsocialunrestresultingintheanti-war,nucleardisarmamentandhippiemovements.Someofthisphilosophicalattentionwasturnedtowardsthefieldwenowcallbioethics,andformedthesecondmajorrootofmodernbioethics.1.4ThescopeofbioethicsWhilemedicalethicsmakesupalargeandcentralsubsectionofbioethics,bioethicsextendsbeyondtherealmsofmedicineandhealthcaretoencompassthoseofthe503\nREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLMbiologicalsciencesmoregenerally,andareasincludingenvironmentalethics,ethicalissuesofsexualityandreproduction.Consider,forexample,genethics,theethicsofgeneticchoiceandmanipulation.2GenethicsGenetics,perhapsmorethananyotherscience,createdbothacuteethicaldilemmasandacuteexistentialanxieties.Itisgeneticsthatmakesusrecallnotsimplyourrespon-sibilitiestotheworldandtooneanother,butourresponsibilitiesforhowpeoplewillbeinthefuture.Itisgeneticsthatisenablingusmoreandmoretodeterminenotsimplywhowillliveandwhowilldie,butwhatallthosewhowillliveinthefuturewillactu-allybelike.Genesarespecialbecausetheyhaveencodedwithinthemalltheinforma-tionrequiredtomakenotonlythepartsbutalsothewholeorganismandtheyhaveoneotherveryspecialcharacteristic:theyareessentiallyimmortal,orrathertheinfor-mationthattheycontainisessentiallyimmortal,passedfromgenerationtogenerationindefinitelyandsharedbetweendifferentspecies.TheHumanGenomeProjectaimedatidentifyingandmappingalloftheonehundredthousandorsogenesthatmakeupthehumangenomeandisnowessentiallycomplete.Locatingandunderstandingthefunctionofallthegeneswillhelpusindevelopingtechniquesormodifyingtheminvariousways.Sofartheimpetusforthisresearchhasbeenlargelytherapeutic.Themoreweknowaboutdiseasethemoreweunderstandtheinfluenceofgenes.Insinglegenedisorders,suchassicklecelldisease,thepresenceorabsenceofaparticulargeneiswhatdetermineswhetherornottheindividualwillcontractthedisease.Inothercases,forexample,thegenesforbreastcancer(BRCA1and2genes),itisnotthepres-enceorabsenceofasinglegeneorgroupofgenesthatisdecisivebutrathertheirpres-enceorabsencewillactasanimportantpredisposingfactor.Itisobviousthattheabilitytoinfluencethesegenesandtheiroperationisofimmensepotentialbenefit.However,genesdetermineorinfluencemanyotherfeaturesofhumanbeingsaswell,whichrangefromfactorswiththerapeuticimportanceliketheabilitytofightdisease,towhatonemightthinkofasmorallyneutraltraits,likeheight,weight,build,haircolour,eyecolourandsoon.Thetherapeuticmanipulationofgenes(genetherapy)can,inprinciplethoughperhapsnotalwaysinpractice,becarriedouteitheronthesomaticlineoronthegermline.Somaticchangestogenesareone-offchangesthataffectonlytheindividualwhosegeneshavebeenmanipulated;suchchangesarenotpassedonthroughnormalrepro-ductionanddiewiththeindividual.Manipulationstothegermline,ontheotherhand,affectthosegeneswhicharepassedonthroughreproductionandconsequentlyiftheyarechangedthosechangescanbepassedonindefinitelyfromgenerationtogeneration.Atfirstsightitseemsobviousthatifwecanmakebeneficialchangeswhichwillremoveaninheriteddiseaseormitigateitsdisastrouseffects,thenbothhumanityandeconomywouldseemtoindicateapreferenceforthegermline,sincethebeneficialchangeswillbepassedonwithoutfurtherinterventionsbeingrequiredfromgenerationtogenera-tion.However,thereisadownsidetothis:anymistakesorunforeseenconsequenceswillalsobebuiltpermanentlyintothegeneticmake-upofanindividual,whichwillinturnbepassedfromgenerationtogeneration.Thatiswhymanyhavebeenanxiousto504\nBIOETHICS,GENETHICSANDMEDICALETHICSconfinegenetherapytosomaticchangesatleastuntilweunderstanditfarbetterthanwedocurrently.Oneproblemthatoccurshereisthattheremaybesomebeneficialchangesthat,iftheyaretobemadeatall,canonlybemadeinthegermlineandthisraisesthequestionofwhetherornotitcouldbeethicaltodenypresentgenerationspossibleeffectivetherapiesforfearthatthegeneticchangesmayexpressthemselvesindisastrouswaysperhapsmanygenerationshence.2.1GenetictestingandscreeningSincegenesareassociatedsostronglywithmanydiseasesoradverseconditions,itispossiblebyusinggenetictestseithertopredictwhetherindividualsortheirprogenyarelikelytobeaffectedbythesediseasesorindeedtoscreenwholepopulationsorpartialpopulations.Thesetestsraisemanyimportantethicalissues.Wherethetestswillbeperformedonembryos,eitheratthepre-implantationstage(pre-implantationgeneticdiagnosis:PIGD)orinutero(pre-nataldiagnosis),manyproblemsarise.Oneconcernstheethicsofattemptingtousetestresultstofilteroutthepresenceofparticularcon-ditionsinfuturegenerations.Tomanythisseemsunproblematicallyadvantageous,sincedamagingordisablingconditionsmaybeeradicatedandhumanlifemadebetter.Toothersthisseemstoconstitutediscriminationagainstthosewithdisabilityandaninsulttopeoplecurrentlylivingwithdisability.Thesetechniquesalsoraiseissuesaboutthemoralstatusoftheembryoandwhetherweareentitled,forexample,toterminatepregnanciesonthebasisofparticularfeaturesthatweknowtobepresentintheembryo.Whilemanythinkthatterminatingpregnancyforsignificantdisabilityisjustifiable,genetictestswillalsorevealthepresenceofnon-disablingcharacteristics,forexample,genderorskincolour,onthebasisofwhichsomemightalsowishtoterminatepregnancies.Separatequestionsconcerntheethicsofmakinggeneticinformationavailablethroughtestingandscreeningsothatpeoplemayknowwhattoexpectintheirchil-drenandsothatsocietiesmayknowwhattoexpectinfuturegenerationsandtheethicsofactingonthatinformation.2.2CloningFinally,cloningtechnologyofthesortthatwasusedtocreatethenowworldfamous‘Dollythesheep’maybeusednotonlytocopythegenotypeofthewholeorganismandproduceageneticallyidenticalcopyofexistingadults,butalsocanbeusedatthelevelofcellstocreatecellsandcelllinesthataregeneticallycompatiblewiththegenomeofparticularindividualssothatthesemaybeinsertedwithoutfearofrejectionandalsotocreate,inprinciple,limitlessnumbersofcopiesofparticulargenesthatmightbeusefulanddifficulttoproduceortofind.Whilecloningwholeorganismshasgreatdra-maticappeal,ithaslimitedtherapeuticoraltruisticutility.Ontheotherhand,cloningcellsandcelllinesisoneofthemostpromisingavenuesofscientificresearchwhichmayleadtothepowertohelptissuerepairitselfandregenerateitselfinlivinghumanorganismsandeventuallymayleadtothe,perhapsindefinite,postponementofdeath.Whiletheethicsoftherapyseemunproblematic,theethicsofincreasinglifeexpectancyarecomplicatedandraiseissuesnotonlyofindividualbutalsoofglobaljustice.505\nREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLM3MedicalEthicsMedicalethicsisasubsetofbioethicsandinvestigatesethicalissuesarisinginmedicineandhealth-careprovisionbyapplyingtheprinciplesofmoralphilosophytotheseprob-lems.Whiletheterms‘bioethics’and‘medicalethics’areoftenusedinterchangeably,theydonotrepresentidenticaldisciplines.Althoughappliedethicsrelatingtomedicineandhealthcaremaybecategorizedaseitherbioethicsormedicalethics,theterm‘bioethics’representstheapplicationoftheprinciplesofmoralphilosophytoallthelifesciences,notjusttomedicineandhealth-careprovision.Someofthemajorissuesinmedicalethicsarediscussedbelow.3.1Whendoeslifebegin?Manypeoplehavethoughtthattheproblemofwhenlifebecomesmorallyimportant,intheultimatesense,isansweredbyknowingwhenlifebegins.Manyofthemostpopularaccountsofpersonhood,forexample,concentrateonattemptingtoanswerthequestionofwhenlifebegins,treatinglifeunproblematicallyinthiscontextashumanlife.Theeventmostpopularlytakentomarkthestarting-pointofhumanlifeiscon-ception.Butconceptionisunhelpfulasathresholdofmoralimportanceforanumberofreasons.First,conceptioncanresultinahydatidiformmole,acancerousmulti-plicationofcellsthatwillneverbecomeanythingbutapalpablethreattothelifeofthemother.Second,evenifhumanlifedoesbeginatconception,itisnotnecessarilythelifeofanindividual;twinsmayformatanypointuptoapproximatelyfourteendaysfollowingconception.Cloningalsohasraisedproblemsforourunderstandingofwhenlifebegins.Ifonehasapre-implantationembryointheearlystagesofdevelopmentwhenallofthecellsaretotipotent,thatis,whereanyofthecellscouldbecomeanypartoftheresultingindividual,andonesplitsthisearlycellmass(anythinguptothe100-cellstage)into,say,fourclumpsofcells,eachofthefourclumpswouldconstituteanew,viableembryothatcouldbeimplantedwitheveryhopeofsuccessfuldevelopmentintoadulthood.Eachclumpisthecloneoridenticaltwinofeachoftheothersandcomesintobeingnotthroughconception,butbecauseofthedivisionoftheearlycellmass.Moreover,thefourclumpscanberecombinedintooneembryo.Thiscreatesasituationwhere,withoutthedestructionofasinglehumancell,onehumanlife,ifthatiswhatitis,canbesplitintofourandcanberecombinedagainintoone.Thosewhothinkthatensoulmenttakesplaceatconceptionhaveaninterestingproblemtoaccountforthesplittingofonesoulintofour,andforthedestructionofthreesoulswhenthefourembryosarerecombinedintoone,andtoaccountforthedestructionofthreeindividualswithoutasinglehumancellbeingremovedorkilled.Thesepossibilitiesmightgiveuspauseinattributingabeginningofmorallyimportantlifetoapointlikeconception.However,ifwewanttoknowwhenlifebeginstomattermorallythenitcanneverbeenoughtoknowwhenlifebegins;wehavetoknowwhylifeofaparticularsort,wheneveritbegins,isimportant,andmoreoverwhyitismoreimportantthanothersortsoflivestowhichcare,respectandresourcesmightalsobedevoted.506\nBIOETHICS,GENETHICSANDMEDICALETHICS3.2SpeciesismandnaturalkindsSomepeoplehaveattemptedtoovercome,orratherside-step,thisproblembysimplystipulatingthatitishumanbeingsthatmatter.Althoughthismovecertainlyavoidstheproblem,itdoessoatsomecost.Itisdifficulttoimaginehowonewoulddefendamoraltheorythatwasfoundedonthestipulationofanarbitrarypreferenceforonekindofcreatureoveranother,particularlywhenthispreferenceisassertedbyself-interestedindividualsonbehalfoftheirownkind.Wearealltoofamiliarwiththehistoryofsimilarclaimsinwhichthemoralpriorityandsuperiorityof‘ourownkind’hasbeenassertedonbehalfofoneracial,religiousorgendergroupattheexpenseofanother.Simplystipulatingarbitrarilythesuperiorityofourownkind,whetherdefinedbyspeciesmembership,race,gender,nationality,religion,oranyothernon-moralcharacteristic,withoutfurtherargumentisinsufficienttoestablishanymoralclaim.Membershipofanaturalkind,orofanethnic,religiousorothergrouping,isnotofitselfamoralproperty.3.3PotentialityTheproblemistodistinguishinsomemorallysignificantrespect,humanembryosfromtheembryosandindeedtheadultmembersofanyotherspecies.Theonethinghumanembryoshavethatmembersofotherspeciesdonotistheirpotentialnotsimplytobebornandtobehuman,buttobecomethesortofcomplex,intelligent,self-conscious,multifacetedcreaturestypicalofthehumanspecies.Thereare,however,twodifficul-tiesforthepotentialityargument.Thefirst,logicaldifficultyisstraightforwardbuttelling.Weareaskedtoacceptthathumanembryosorfoetusesarepersons,morallyimportantbeingswhoseintereststrumpthoseofothersortsofbeings,invirtueoftheirpotentialtobecomeanothersortofbeing.Butitdoesnotfollowlogicallywithoutfurtherpremises,evenifweacceptthatwearerequiredtotreatxincertainways,andevenifawillinevitablybecomex,thatwemusttreataasifithadbecomex,atatimeoratastagepriortoitshavingbecomex.Theseconddifficultywiththepotentialityargumentinvolvesthescopeofthepoten-tialforpersonhood.Ifthehumanzygotehasthepotentialtobecomeanadulthumanbeingandissupposedlymorallyimportantinvirtueofthatpotential,thenwhatofthepotentialtobecomeazygote?Somethinghasthepotentialtobecomeazygote,andwhateverhasthepotentialtobecomethezygotehaswhateverpotentialthezygotehas.Cloningbynucleartransfer,whichinvolvesdeletingthenucleusofanunfertilizedegg,insertingthenucleustakenfromanyadultcell,andelectricallystimulatingtheresult-ingnewlycreatedeggtodevelop,can,intheory,produceanewhuman.Thismeansthatanycellfromanormalhumanbodyhas,ifappropriatelytreated,thepotentialtobecomeanew‘twin’ofthatindividual.PersonhoodThequestforthosefeaturesthatmakeindividualsmorallyimportantisoftenconsideredtobethesamequestionasthequestion‘Whatisaperson?’Towardtheendoftheseven-teenthcenturyinhisEssayConcerningHumanUnderstanding,thephilosopherJohnLOCKE507\nREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLM(chapter29)attemptedtoanswerthisquestioninawaythatscarcelyhasbeensurpassed.Hewrote:Wemustconsiderwhatpersonstandsfor;whichIthinkisathinkingintelligentbeing,thathasreasonandreflection,andcanconsideritselfthesamethinkingthing,indifferenttimesandplaces;whichitdoesonlybythatconsciousnesswhichisinseparablefromthinkingandseemstomeessentialtoit;itbeingimpossibleforanyonetoperceivewithoutperceivingthathedoesperceive.(Locke1924:ch.27,bookII)Locke’sdefinitionformsthebasisofmostaccountsofpersonhoodgiventoday.Onequestionthatimmediatelyarisesis:‘Cansomeonebemoreorlessofaperson?’AlloftheelementsinLocke’sdefinition–intelligence,theabilitytothinkandreason,thecapac-ityforreflection,self-consciousness,memoryandforesight–arecapacitiesthatadmitofdegrees.Doesthisleadusintoahierarchyofpersonsandhenceofmoralimportanceorvalue?Theanswertothisquestionturnsonwhetherornotthedegreetowhichthesecapacitiesarerequiredforpersonhoodformsa‘threshold’oracontinuum.Ifathresh-old,thenonceacrossthethresholdtheindividualisaperson.Ifnot,theaboveproblemsforpersonhoodaccountsofthemoralstatusofindividualsremain.3.4AbortionandmoralstatusIncontemporarydebatetheissueoftheethicsofabortionhasbeenapproachedinthreemajorways.Thefirsttreatstheethicsofabortionasturningonthemoralstatusoftheembryoorfoetus.Broadly,ifthehumanindividualinuteroqualifiesasapersonthenithasamoralstatuswhichprotectsitslife.Adifficultyhereisabsenceofagreementonthedefiningcharacteristicsofpersonhood.Thesecondapproachtakesa‘women’srights’viewoftheethicalissueandsuggeststhatitisawoman’srighttochoosewhathappensinandtoherbodyandthattheethicsofabortionmustbeseenasadimensionofthemoralentitlementsofwomen.Adiffi-cultywithsuchaviewisthatiftheindividualenventresamereisapersonthenitisunclearwhytheredoesnotresultanirresolvableconflictorrightsorinterests.Thethirdapproachattemptstoside-steptheirreconcilabilityofthesetwoap-proachesbysuggestingthatevenifthefoetusisapersonorhasmoralimportancecomparabletothatofthemother,itmaybeseenasmakinganunjustifiableclaimonthemotherthatsheisundernoobligationtomeet(seeThompson1971).3.5EndoflifedecisionsEndoflifedecisionsoccurwhenadecisionismade,theeffectofwhichistoendthelifeofahumanindividualwhichcould,inprinciple,havecontinued.Therearemanydif-ferentinstancesofthis:suicide,whereacompetentindividualendsherownlifebyherownhand;assistedsuicide,wheresuchanindividualishelpedtodie;physician-assistedsuicide,wherethehelpinghandisthatofaphysician;andvoluntaryeuthanasia,wherethehelpinghandisthatofanon-medicallyqualifiedthirdparty.Theethicsofallthesecasesturnprincipallyonthequestionsofwhetherornotacompetentindividualisenti-tledtocontrolherowndestiny,includingthetimingandmannerofherdeath,andifsheisnotwhollyentitledtoexercisethiscontrolwhetherornottherearesomecir-cumstancesinwhichthismaybelegitimate.Somepeopleregardasmorallysignificant508\nBIOETHICS,GENETHICSANDMEDICALETHICSwhetherthestepsrequiredtobringaboutdeathareactiveorpassive,distinguishingbetweenactiveeuthanasia(sometimesreferredtoasdirectkilling)andpassiveeuthanasia(sometimesreferredtoasindirectkilling).Involuntaryandnon-voluntaryeuthanasiaInvoluntaryeuthanasiaisthekillingofanotheragainsttheirwill.Thisisofcoursetantamounttomurderandwillbecondemnedwherevermurderiscondemned.Non-voluntaryeuthanasiaistheendingofthelifeofanindividualinanyofthemannersabovedescribedwherethatindividualhasnotexpressedapreferenceandisatthetimeofthedecisionnolongercapableofexpressingpreferencesonewayortheother.Thiscaseusuallyconcernsthoseindividualswhohaveinsufficientcompetence,forwhateverreason,tomakeanautonomousdecisionabouttheendingoftheirlifeandwhoareinsuchseverepainordistresswhichisirremediablebynormalpalliativemeans,butthequestionofendingtheirexistencearises.Thesecondclassofcasesarethosewhohavepermanentlylostconsciousnessandwhorequirelife-sustainingmeasures,whethertheyareintheformofmedicaltreatmentorintheformofassistedfeedingandhydration.ClassiccasesofthistypeareaffordedbytheconditionknownasPersistentVegetativeState(PVS).Thisistheconditionwhichwasatissueintwocrucialcasesofhighestjuris-dictionintheUnitedStatesandintheUnitedKingdom,thoseofNancyCruzanandTonyBland.ForbrevitywewillconcentrateonthedetailsoftheBlandcase.BothNancyCruzanandTonyBlandhadbeenleftinapersistentvegetativestatefollowingaccidents.PVSisanunconsciousstatewhich,afterayear’sduration,isgenerallyacceptedasper-manentandirreversible.PeopleinPVSdonotrequirelifesupportasthisisusuallyunder-stood.Theyarenotonventilators,forexample,althoughtheydorequiretubefeedingandhydration.Theyarenot,norwithoutassistancewilltheybecome,‘dead’accordingtoanyofthecurrentcriteriaoraccepteddefinitionsofdeath.Theyarenotbrain-stemdeadforexample,andbecausetheyhavespontaneousheartbeatandrespirationarenotdeadbyanynormalconceptionsofdeatheither.TonyBland’sparentsaskedthecourtstorulethathisdeathcouldbebroughtaboutbywithdrawaloffeedingandwithholdingofotherlife-sustainingmeasures,includingantibiotics,becauseafterwithdrawaloffeedinghewasexpectedtosuccumbtoinfectionsfromwhichhewoulddiewithoutantibiotics.Inadvanceofthevarioushearings,ithadbeenexpectedthattheissueincourtwouldturnonwhetheritwaslawfultowithdrawfeedingandstarvesomeonetodeath.Therewerelegalprecedentsfordecisionstowith-drawlife-sustainingmedicaltreatment,butnooneconsideredfeedingtobea‘treatment’andhencesomethingthatdoctorscouldwithdrawonthebasisoftheirjudgementsastowhetherthemeasureatissuewasinthepatient’sbestinterestsorcouldbeaffordedbythehealth-caresystem.Totheircreditthecourtsdidnotattempttostretchthemeaningof‘medicaltreatment’tocoverfeeding,thusgivingthedoctorsclinicaldiscretioninthematter,butsquarelyfacedtheissueofwhetherornotTonyBlandshouldcontinuealive.Despitedisclaimers,theHouseofLordsdecisionintheBlandcaseisregardedbymanyasineffectonepermittingnon-voluntaryeuthanasia.SinceTonyBlandwasnotdead,andwouldnotdieunlesstheLawLordspermittedadefinitecourseofactionwhichwouldresultinhisdeath,theirdecisiontotheeffectthatitwaspermissibletoendhislifewhenitotherwisewouldhavecontinuedindefinitely,effectivelybroughthislifetoanend.Andindeed,suchadecisionwassoughtbyTonyBland’sparentsforpreciselythatreason.509\nREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLMIfweaskwhatjustifiedthedecisiontotakestepsthatwouldinevitablyresultinhisprematuredeath,theanswerraisesinterestingissuesaboutthelegitimacyofendinglives.Forifitisbelievedthatthesanctity–thesacredness–oflifeattachestothelivesofhumans,thenasalivehumanTonyBlanddoesnotrelevantlydifferfromotherswhoselivesitiswrongtoend.If,however,itislivesofaparticularcharacter,livesthatareinsomesenseworthleading,wemightarriveatadifferentanswer.TonyBland,atthetimeofthecourts’deliberations,hadalifethathecouldnotleadinanymeaningfulsenseatall.InthewordsofLordKeithofKinkelinhisjudgementinthatcase,‘Itis,however,perhapspermissibletosaythattoanindividualwithnocognitivecapacitywhatever,andnoprospectofeverrecoveringanysuchcapacityinthisworld,itmustbeamatterofcompleteindifferencewhetherhelivesordies’.Itiswidelythoughtthatdeathistheonlyeventthatcanbringtoanendavaluablelifeandreleaseothersfrommost,thoughnotall,oftheobligationstheyowetootherpeople.Ifthisisright,thenofcourseadefinitionofdeathisveryimportantanditisnowwidelyacceptedthatthedeathofthebrainstem(socalled‘brain-stemdeath’)isthebestcriterionoforevidencefordeathintheordinary-languageortraditionalsense.However,incasessuchasthatofTonyBland,itmaybearguedthatitisnotbrain-stemdeaththatindicatestheendofhumanlifebutthelossofthecapacityforconsciousnessandthought.Itisthehigherbrain,thecerebralcortex,thatisgenerallyassociatedwithcon-sciousnessandthought,andinPVSitisthissectionofthebrainthathasceasedtofunc-tion.Thishasledsometosuggestthatnotonlycanweconsiderhigherbraindeathassynonymouswiththedeathofthathumanlifeinthebiographicalsense,butfurtherthatdeathshouldbeequatedwiththelossofhigherbrainfunction.However,itisonethingtoclaimthatwemaynothavethesameobligationsnottokillandtohelpsustainthelifeofthosewhosehigherbrainisdead;itisquiteanothertoredefinedeathtoincludeapermanentlyunconsciousbutneverthelessspontaneouslybreathingindividual.3.6HealthcareandjusticeThemainproblemofDISTRIBUTIVEJUSTICE(pp.258–64)discussedinbioethicsistheproblemofthedistributionofscarcehealth-careresources.Thisquestionmayoccurinconnectionwithindivisibleresourceslikeorgansfortransplant,butalsomorecommonlyinconnectionwithdivisibleresourceslikemoney.Thegeneralproblemistodistributearesourceinsituationswhereanumberofpersonseither(1)haveclaimsontheresourcethatcannotbefulfilledsimultaneously,or(2)canbenefitfromtheresourcebutwherethereisnotsufficientforeverybody.Ifthepossibilityofbenefitingfromaresourceissufficienttohaveaclaimonthatresource,problems(1)and(2)areidentical,otherwisetheyarepotentiallydifferentproblems.Foreaseofexpositionwewillassumethatpossiblebenefitcanbethebasisforavalidclaimonhealth-careresources.Iftwoormorepeoplehaveaclaimonthesameresourcehowshouldwethendecidewhoshouldhavetheresource?Withinaconsequentialistframeworkthegeneralanswermustbethatweshouldmaximizegoodconsequencesbyourallocationofhealth-careresources,butthisraisesthefurtherquestionoftheappropriatespecifica-tionofthemaximand.Apopularansweristhatweshouldtrytomaximizethepro-510\nBIOETHICS,GENETHICSANDMEDICALETHICSductionofQALYs(QualityAdjustedLifeYears),oroneofthesimilarmeasureslikeHLYEs(HealthyLifeYearEquivalents)orDALYs(DisabilityAdjustedLifeYears)(althoughfortechnicalreasonsDALYsshouldbeminimized,sincetheymeasuretheburdenofdisability).Thebasicideabehindallthesemeasuresisthathealthcareshouldmaximizehealthoutcomesandthathealthoutcomesarebestmeasuredastheextralifetimegainedasaresultoftheuseofresourcesadjustedbysomefactordesignedtoreflectthataperiodwithoutdisability,distress,etc.isworthmorethanaperiodwiththesefactors.SeveralproblemshavebeenraisedconcerningQALYmaximization,includingthefollowing:(1)itdiscriminatesagainsttheelderlyandthealreadydisabled;(2)itfailstotakeaccountofanimportantdistinctionbetweenlife-savingandnon-lifesavinginterventions;and(3)itfailstotakeaccountofthefactthathealthcaremayhavemorethanonegoal.Whetherornotcertainkindsofmeritordesertcangroundaspecialclaimonresourcesisadifficultquestion.Forinstance,coulditbethecasethatafiremaninjuredinthelineofdutyhasaspecialclaimtohealthcare,evenifitwillnotbringhimorherbacktoactiveduty(andtherebyopenroomforanon-desertexplanationoftheclaim)?Suchaclaimseemstohaveintuitivemerit,butitturnsouttobeverydifficulttoexplicateexactlywhatkindsofdesertcanjustifyhealth-careclaims.3.7PublichealthversusindividualrightsThelasttwodecadeshavewitnessedthegradualevolutionofadoctrineofpatientautonomyfullyendorsedinthewritingsofmostethicistsandatleastpartiallyrecog-nizedandenforcedbythelawandobservedinmedicalpractice.Patientautonomyisincreasinglyandrightlyperceivedasamanifestationoftheindividual’srightsofself-determinationandprivacy,universallyregardedasapillarofcivilliberty.Itisacknowl-edgedthatthestateshouldrefrainfrominterventioninprivatelivessavewheretheindividual’shealth-stateorlifestyleendangersothers.Health,moreandmore,isamatterofprivatechoice,asisthedecisionastowhetherornottoundergomedicaltreatment.Whileitisheldimportantthatindividualsareabletoexercisetheirownchoiceswithregardtotheirhealthandlifestyle,itisalsoassumedthatnation-statesareresponsiblefortheprotectionofcitizensagainstthreatstotheirlives,includingthreatsfromillnessordisease.Thus,whenconsideringsocialpolicyintheareaofpublichealth,wearefacedwitha‘balancingact’betweenrespectingindividualautonomyandthestate’sresponsibilitytoprotectothers.ThedevelopmentofsocialpolicyrelatingtoHIV/AIDSillustratesthistensionbetweenindividualautonomyandpublichealthresponsibilities.ThethreatposedtothecommunitybyHIV/AIDShasforcedareconsiderationofanyconceptofabsoluteautonomy.HIV/AIDSisoftenseenasadiseasecentrallyrelatedtolifestyleforitsasso-ciationwithdruguseandsexualconduct.Atthesametimeasage-oldproblemssuchasinfectionsurfaceoncemore,newdifficultiesarisetocomplicatetherelationshipbetweenpublichealthandprivatelives.Theemphasisonautonomyinrecentyearshastendedtopre-emptanalysisoftheresponsibilitiesofthestateandoftheindividualforpersonalandpublichealth.A511\nREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLMnumberofimportantquestionsemerge.Wheredoestheinterfacebetweenthestate’sresponsibilitytoprotectandpromotepublichealthandtheindividual’scivilrights(privacy,freedom,etc.)lie?Whereshouldalinebedrawnbetweenpublichealthandprivatelives?Towhatextentcanthestaterelyonindividualresponsibilityinhealthpromotion?Suchquestionspervadeeverylevelofhumanlife.Forexample,if,bythechoiceshehasmade,anindividualisresponsible,atleastinpart,forherownill-health,ishermoralclaimtocare,ortopriorityincare,somuchtheless?Howarerightsandrespon-sibilitiesforsafetyandhealthtobefairlydistributedintheworkplace?Howshouldlifestyleconsiderationsanddiseasestatebeaccountedforininsurance(lifeandhealth)?3.8ProfessionalethicsAcornerstoneofmodernprofessionalethicswithinthehealthprofessionsistheconceptofinformedconsent.Itisgenerallyaccepted,andinmanyjurisdictionsalsolegallyenforced,thatnomedicalinterventioncanbeperformedonacompetentpersonunlessthatpersonhasgiventheirfreeandinformedconsent.Inorderforaconsenttocountasvalidinformedconsent,twoinformationconditionsandtwoconsentcondi-tionshavetobefulfilled:Informationconditions1Informationaboutthenature,purpose,expectedeffect,andexpectedside-effectsoftheinterventionhasbeengiven.2Theinformationmentionedin(1)hasbeenunderstoodtoasufficientdegree.Consentconditions3Thepersoniscompetenttoconsent.4Theconsentisgivenfreelyandwithoutcoercion.Theethicaljustificationoftherequirementforinformedconsenthasthreepartlyindependentstrands,anddependingontheoreticalorientationonemayplacetheemphasisexclusivelyononeoftheseoracceptmorethanone.Thefirstjustificationcomesfromconsiderationsaboutbodilyintegrityandtheethi-callyproblematicnatureofunwantedbodilyinvasion.Thesecondjustificationcomesfromconsiderationsabouttherighttohaveautonomousdecisionsrespected.Andthethirdjustificationcomesfromconsiderationsaboutthepluralityofvaluesandprefer-ences,andtheproblemstherebycreatedbymakingdecisionsnotbasedontheperson’sownvalues.Theothermajorareainprofessionalethicsisconcernedwiththeclassicaltopicsofconfidentiality,privacyandthemoderntopicofdataprotection.Theideathatconfi-dentialityisanimportantcomponentoftherelationshipbetweenthedoctorandthepatienthasbeenupheldinmedicalethicseversincetheHippocraticOathinancientGreece.Itis,however,importanttonotethattheethicaljustificationforconfidential-ityhaschangedovertime.Thetraditionaljustificationwasbasedontheideathatthereputationofthemedicalprofessionwouldsufferifdoctorspubliclyspreadthesecretstheyknewabouttheirpatients.Thiswasthemainjustificationforconfidentialityuntil512\nBIOETHICS,GENETHICSANDMEDICALETHICSthelate1940s,whenamoremoderntwo-prongedjustificationbegantoemerge.Thecurrentargumentforconfidentialityhastwostrands,onegenerallyconcernedwiththepassingofinformationbetweenhealth-careprofessionalsorfromhealth-careprofes-sionalstooutsiders,andonespecificallyconcernedwiththepassingofinformationtolawenforcementagenciesandsimilarbodies.Thefirststrandofargumentproceedsfromoneofthefollowingpremises:(1)thereisarighttoprivacy,(2)thereisarighttocontrolpersonalinformation(asanextensionofamoregeneralrighttoself-determination),or(3)personalinformationisakindofpersonalproperty.Fromallofthesepremisestheconclusioncanbederivedthatthepatienthasdecisioncontroloverwhathealthinformationshouldbegeneratedandtowhomitcanbetransmitted.Thiscontrolisnotabsoluteandnon-defeasible,butitisstrongandcanonlybedefeatedbysubstantialharmtoothersincasesofnon-disclosureofconfidentialinformation.Thesecondstrandofargumentproceedsinmuchthesamewayastheclassicaljus-tificationofprofessionalconfidentiality.Itpositstwopremises:(1)itisimportantthatinjuredandillpersonsseekmedicalattention;(2)ifitisgenerallyknownthathealth-careprofessionalsroutinelydivulgeinformationtolawenforcementagencies,itwilldetersomepersonsfromseekingmedicalattentionwhentheyneedit.Theconclusionwhichthisstrandoftheargumentseekstoestablishisthusthathealth-careprofes-sionalsshouldneveroronlyinveryrarecircumstancesbeforcedtodiscloseinforma-tiontolawenforcementagencies.Withtheincreasinguseofinformationtechnologyinhealthcare,issuesconcern-ingdataprotectionhavebecomemoreandmoreimportant.Whatisatissuehereisnotthedeliberatedisseminationofconfidentialinformation,buttherisksofnon-deliberatedisclosureandpossiblemisuseofinformationwhenitisstoredineasilyretrievableandlinkableform.Therearegreatpossiblebenefitsassociatedwiththeuseofinformationtechnology,andtheproblemisthustobalancethesebenefitsagainsttheinherentrisks.Somehavetriedtoanalysetheseproblemsintermsofadistinctionbetweensensitiveandnon-sensitiveinformation,butthisisunhelpfulinmostbioethicscontextsbecauseallkindsofhealthinformationareorcanbecomesensitiveinarangeofquitenormalcircumstances.3.9ResearchethicsTheethicsofresearchgenerallyandresearchonhumansubjectsinparticularisofincreasingimportance.FollowingtheNazidoctors’trialatNurembergattheendoftheSecondWorldWar,therewaswidespreadunderstandingoftheterribleabuseofsciencethathadoccurredundertheNazis.Thisunderstanding,coupledwithadeterminationtoattempttopreventsuchabusesinthefuture,culminatedintheformulationofmanyinternationalprotocolsgoverningresearch,particularlyresearchonhumansubjects,mostprominentofwhichisperhapstheWorldMedicalAssociationDeclarationofHelsinkianditsvariousamendments.However,inthelightofrecentmassivechal-lenges,nottohumanrightsbuttohumanhealth,thestringencyofthesafeguardsthathavebeenproposedhasincreasinglybeencalledintoquestion.Forexample,theWorldMedicalAssociationDeclarationofHelsinkistates(Article1Para.5):‘Concernfortheinterestofthesubjectmustalwaysprevailovertheinterestofscienceandsociety’.Ifscienceandsocietyareconstruednarrowlyasthepersonalorprofessionalinterestof513\nREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLMscientistsandtheinterestsofsocietyaredefinedintermsofaparticularideologythisprotocolseemssoundenough.However,theAIDSpandemichasdrawnattentiontothefactthatboththeinterestsofindividualsandtheinterestsofsocietymayalsorequirethepursuitofscienceandthattosomeextentifindividualsaretoprotectthemselvestheymustbewillingtoco-operateinthisendeavour.Thistensionislikelytoremainwithusandisasyetunresolved.AsecondprincipleoftheHelsinkiDeclaration,andonemaintainedinmanyotherresearchethicsprotocols,requiresthatsuchsubjectsshouldnotbecoercedorinducedtoparticipatebyoffersofreward,particularlyfinancialreward.However,inanincreas-inglyconsumeristworld,individualsmoreoftenthannotexpecttoberewardedoratleastcompensatedfortheircontribution.Hereagainthereisatensionbetweenwhatisrequiredtosafeguardtherightsoftheindividualandwhatmaybefairorreasonablebothinordertoavoidexploitationandtoenablesciencetocontinue.Manyhavetriedtoresolvethistensionbydistinguishingsharplybetweenthereimbursementofrea-sonableexpensesandpaymentoffees.However,thisdividinglineisoftendifficulttodrawbecauseifreasonableexpensesincludelostopportunitycostsofusingone’stimeinanotherwayitisdifficulttoseehowthedistinctionmaybemaintained.Afurtherissue,whichisofincreasingimportanceinthecontemporaryworld,istheethicsofusinganimalratherthanhumansubjectsinmedicalresearch.Manywhotakeastandontheimportanceofsentience(thecapacitytofeelpainorsensation)notethathumansandanimals,insharingthiscapacity,sharevulnerabilityandareentitledtoequalprotection.Others,bearinginmind,perhaps,considerationsofthesortadvancedbyLordKeithofKinkel(above),drawadistinctionbetweenmosthumanindividualsandmostanimalswhichwouldpermitresearchonthelatterbutnottheformer.Finally,someperceiveamorallysignificanthierarchyinanimalsandwoulddistinguish,forexample,primatesfromotheranimalsintermsoftheirappropriatenessasresearchsubjects.Theresolutionoftheseissuesdependscruciallyontheaccountonefeelscompelledtogiveoftheacquisitionandmaintenanceofmoralstatusoronthemoralsignificanceofsentience.ReferencesBeauchamp,T.L.andChildress,J.F.1994:PrinciplesofBiomedicalEthics,4thedn.NewYork:OxfordUniversityPress.Bennett,R.andErin,C.A.(eds)1999:HIVandAIDS:Testing,ScreeningandConfidentiality.Oxford:OxfordUniversityPress.Dworkin,R.1977:TakingRightsSeriously.London:Duckworth.Dyson,A.O.andHarris,J.1991:ExperimentsOnEmbryos.London:Routledge.Englehardt,Jr,H.T.1996:TheFoundationsofBioethics,2ndedn.NewYork:OxfordUniversityPress.Gillon,R.1986:PhilosophicalMedicalEthics.London:JohnWileyandSonsonbehalfoftheBritishMedicalJournal.Glover,J.1977:CausingDeathandSavingLives.Harmondsworth:PenguinBooks.Green,R.M.1990:MethodinBioethics.TheJournalofMedicineandPhilosophy,15,2,April.Harris,J.1991:TheValueofLife:AnIntroductiontoMedicalEthics.London:RoutledgeandKeganPaul.514\nBIOETHICS,GENETHICSANDMEDICALETHICS——1998:Clones,GenesandImmortality.Oxford:OxfordUniversityPress.Harris,J.andHolm,S.(eds)1998:TheFutureofHumanReproduction:Ethics,ChoiceandRegula-tion.Oxford:ClarendonPress.Hursthouse,R.1987:BeginningLives.Oxford:Blackwell.Locke,J.1924:AnEssayConcerningHumanUnderstanding(editedbyE.S.PringlePatterson).Oxford:ClarendonPress.Oderberg,D.S.2000:AppliedEthics:ANon-ConsequentialistApproach.Oxford:Blackwell.Rachels,J.(ed.)1986:EthicalTheory.Oxford:OxfordUniversityPress.Simmonds,N.E.1986:CentralIssuesinJurisprudence:Justice,LawandRights.London:SweetandMaxwell.Singer,P.(ed.)1986:AppliedEthics.Oxford:OxfordUniversityPress.Steinbock,B.1980:KillingandLettingDie.EnglewoodCliffs,NJ:Prentice-Hall.Thompson,J.J.1971:ADefenseofAbortion.PhilosophyandPublicAffairs,1,1.ReprintedinJ.Feinberg(ed.)1984:TheProblemofAbortion,2ndedn.Belmont,CA:WadsworthPublishing.DiscussionQuestions1Shouldweattempttoeradicatedisability?2Istheselectionofthesexorappearanceofone’soffspringmorallyunacceptable?3Doespre-implantationgeneticdiagnosisimplythatthelivesofindividualswithdisablingconditions,suchasdeafness,arenotworthliving?4Couldcloningbecomeamorallyacceptablepracticeifcertainrestrictionsweremadeonitsuse?5Whatisitthatgiveshumanlifevalue?6Isitpossible,inthecaseofabortion,tobalancetherightsofthecarryingmotheragainsttherightsoftheunbornchild?7Isitconsistenttoarguethattheremaybebothcaseswhereabortioniscompletelyunjustifiableandalsocaseswhereabortioniscompletelyjustifiable?8Whatmoralobligationsdoesapregnantwomanhavetowardsthefoetussheiscarrying?Whatimplicationsdothesemoralobligationshaveforlegislationinthisarea?9Iskillingpeoplealwayswrongandifso,why?10Doesthewithdrawingoftreatment,suchasinthecaseofTonyBland,constituteeuthanasia?11Isthereadefensiblemoraldistinctiontobedrawnbetweenkillingsomeoneandlettingthemdie?12AdvanceDirectivesaimtorepresentourinterestswhenwelosethecapacityforconsent.ShouldAdvanceDirectivesalwaysbefollowed?13Examinethejusticeofprioritizinghealthcareaccordingtodesert.Inparticularwhateffectshouldapatient’sresponsibilityforhisowndiseasehaveondeterminingaccesstotreatment?Isitfairthatsomeonewhohascontributedtotheirownill-healthshouldbetreatedwiththesameurgencyasan‘innocentvictim’?14DoindividualswhobelievethemselvestobeHIV-positivehaveamoralobligationtoforewarnothersoftheirpositivestatus?15Dopeoplehavearighttoremaininignoranceoftheirhealthstate?Isthis‘right’affectedifthepersonhasacommunicabledisease?515\nREBECCABENNETT,CHARLESA.ERIN,JOHNHARRISANDSØRENHOLM16Whyisconsenttotreatmentimportant?Underwhatcircumstanceswoulditbeappropriatetotreatapersonwithouttheirconsent?17Shouldpatientsalwaysbetoldthetruthabouttheircondition?18Shouldmedicalconfidentialitybeabsoluteoraretherecircumstancesinwhichhealth-careprofessionsshouldbreachpatientconfidentiality?19Isitacceptabletoconductresearchonhumansubjectswithouttheirinformedconsentifitisveryunlikelythattheywillbeharmedbytheresearch?516\n18EnvironmentalEthicsHOLMESROLSTON,IIIEnvironmentalethicsistheoryandpracticeaboutappropriateconcernfor,valuesin,anddutiesregardingthenaturalworld.Byclassicalaccounts,ethicsispeoplerelatingtopeopleinjusticeandlove.Environmentalethicsstartswithhumanconcernsforaqualityenvironment,andsomethinkthisshapestheethicfromstarttofinish.Othersholdthat,beyondinter-humanconcerns,valuesareatstakewhenhumansrelatetoanimals,plants,speciesandecosystems.Accordingtotheirvision,humansoughttofindnaturesometimesmorallyconsiderableinitself,andthisturnsethicsinnewdirections.1TheEnvironmentalTurnHumansaretheonlyself-reflective,deliberativemoralagents.Ethicsisforpeople.Butarehumanstheonlyvaluable,valuingagentsinanotherwisevalue-freeworld?Humansco-inhabitEarthwithfivetotenmillionspecies.NaturehasequippedHomosapiens,thewisespecies,withaconscience.Perhapsconscienceislesswiselyusedthanitoughttobewhen,asinclassicalEnlightenmentethics,itexcludestheglobalcom-munityoflifefromconsideration,withtheresultingparadoxthattheself-consciouslymoralspeciesactsonlyinitscollectiveself-interesttowardalltherest.Environmentalethicsclaimsthatwehumansarenotso‘enlightened’asoncesupposed,notuntilwereachamoreconsiderateethic.Ifsomeonehadbeenattemptingtoforeseethefutureofphilosophyatthemiddleofthetwentiethcentury,oneofthemostsurprisingdevelopmentswouldhavebeentheriseofenvironmentalphilosophy.Environmentalethicsremainedunknownuntilthemid-1970s.Thatwastochangerapidly.Philosophershavepublisheddozensofanthologiesandsystematicworksinthefield,andcoursesaretaughtinseveralhundreduniversitiesandcollegesonmanycontinents.Therearefourprofessionaljour-nals.TheInternationalSocietyforEnvironmentalEthics(ISEE)has400membersin20countries.TheWorldCongressofPhilosophy(1998)devotedfoursectionstoenvironmentalphilosophy,withdozensofotherrelatedpapers.ThewebsitebibliographyoftheISEEcontains8,000articlesandbooksnotonlybyphilosophers,ethicistsandtheologians,butalsobypolicy-makers,lawyers,envi-ronmentalprofessionals,foresters,conservationandwildlifebiologists,ecologists,\nHOLMESROLSTON,IIIeconomists,sociologists,historians,developersandbusinesspersons–allwithanethicalconcernabouthumanusesofthenaturalenvironment.Althoughthefirsteditionofthisbookcontainednochapteronenvironmentalethics,thissecondeditionincludesthisethicalrevising.Philosophershavethoughtaboutnatureformillennia.Althoughthereisanethicimplicitinmanyoftheseworldviews,thiswasnevermuchdevelopedintheWest.FollowingtheEnlightenmentandthescientificrevolution,insecularphilosophiesnaturecametoberegardedasavaluelessrealm,governedbymechanisticcausalforces.Valuesaroseonlywiththeinterestsandpreferencesofhumans.IntheprevailingJudeo-Christiantheologies,GodcreatedagoodEarthwithmyriadsofcreatures,andsubjectedthesetohumandominion.Forfourcenturies,WesternphilosophyandTHEOLOGY(chapter15)werebothdominantlyhumanistic,or,incurrentvocabulary,anthropocentric.Environmentalethicsappliesethicstotheenvironment,analogouslytoethicsappliedtoBUSINESS(chapter19),MEDICINE(chapter17),engineering,LAW(chapter13)andtechnology.Suchhumanistapplicationsmaybechallenging:limitingpopulationgrowthordevelopment,questioningconsumerismandthedistributionofwealth,advocatingtheinclusionofwomenoraboriginalpeoples,orfearingglobalwarming.Environmentalqualityisnecessaryforqualityofhumanlife.Humansdramaticallyrebuildtheirenvironments;still,theirlives,filledwithartefacts,arelivedinanaturalecologywhereresources–soil,air,water,photosynthesis,climate–aremattersoflifeanddeath.Cultureandnaturehaveentwineddestinies,similarto(andrelatedto)thewaymindsareinseparablefrombodies.Soethicsneedstobeappliedtotheenvironment.Atdepth,however,environmentalethicsismoreradicalin‘applyingethics’(somanyadvocatesclaim)outsidethesectorofhumaninterests.Contemporaryethicshasbeenconcernedtobeinclusive:thepooraswellastherich,womenaswellasmen,futuregenerationsaswellasthepresent.Environmentalethicsisevenmoreinclusive.Whalesslaughtered,wolvesextirpated,whoopingcranesandtheirhabitatsdisrupted,ancientforestscut,Earththreatenedbyglobalwarming–theseareethicalquestionsintrinsically,owingtovaluesdestroyedinnature,aswellasalsoinstrumentally,owingtohumanresourcesjeopardized.Humansneedtoincludenatureintheirethics;humansneedtoincludethemselvesinnature.Somewhatironically,justwhenhumans,withtheirincreasingindustryandtech-nology,seemedfurtherandfurtherfromnature,havingmoreknowledgeaboutnaturalprocessesandmorepowertomanagethem,thenaturalworldhasemergedasafocusofethicalconcern.Humanpowertoaffectnaturehasdramaticallyescalated,aswithspecieslossorglobalwarming.Explodingpopulationsraiseconcernsthathumansarenotinasustainablerelationshipwiththeirenvironment.Norhavetheydistributedthebenefitsderivedfromnaturalresourcesequitably.Norhavetheybeensensitiveenoughtothewelfareofthemyriadsofotherspecies.Theplanhereistooutlinesixlevelsofconcern:humans,animals,organisms,species,ecosystems,Earth.Thesewillbecriss-crossedwithoveradozendifferingapproachestoenvironmentalethics:humanisticethics,animalwelfareethics,biocentrism,deepecology,landethics,theologicalenvironmentalethics,ethics518\nENVIRONMENTALETHICSofecojustice,communitarianethicswithcirclesofconcern,environmentalvirtueethics,axiologicalenvironmentalethics,politicalecology,sustainabledevelopmentethics,bioregionalism,ecofeminism,postmodernenvironmentalethics,andanethicsofplace.2Humans:PeopleandtheirWorldHumansarehelpedorhurtbytheconditionoftheirenvironment,andthatthereoughttobesomeethicconcerningtheenvironmentcanbedoubtedonlybythosewhobelieveinnoethicsatall.Ethicswillhaveaconcernforwhathumanshaveatstakethere–benefits,costs,andtheirjustdistribution,risks,pollutionlevels,rightsandtorts,envi-ronmentalsustainabilityandquality,theinterestsoffuturegenerations.Ananthro-pocentricethicsclaimsthatpeopleareboththesubjectandtheobjectofethics.Humanscanhavenodutiestorocks,rivers,nortowildflowersorecosystems,andalmostnonetobirdsorbears.Humanshaveseriousdutiesonlytoeachother.Anthro-pocentristsmaywishtosavethesethingsforthebenefitstheybring.Buttheenviron-mentisthewrongkindofprimarytargetforanethic.Natureisameans,notanendinitself.Manisthemeasureofthings,saidProtagoras,anancientGreekphilosopher,settingthetoneofphilosophysince.Humansdeliberatelyandextensivelyrebuildthespontaneousnaturalenvironmentandmaketheruralandurbanenvironmentsinwhichtheyreside.Wecareaboutthequalityoflifeinthesehybridsofnatureandculture.Ethicsarisestoprotectvariousgoodswithinourcultures;this,historically,hasbeenitsprincipalarena.Asphiloso-phersfrequentlymodelthis,ethicsisafeatureofthehumanSOCIALCONTRACT(pp.622–7).Peoplearrangeasocietywheretheyandtheotherswithwhomtheylivedonot(oroughtnot)lie,steal,kill.Thisisright,andonereasonitisrightisthatpeoplemustco-operatetosurvive;andthemoretheyreliablyco-operatethemoretheyflour-ish.Onewayofenvisioningthisistheso-calledORIGINALPOSITION(p.261),whereoneentersintocontract,figuringoutwhatisbestforapersononaverage,oblivioustothespecificcircumstancesofone’stimeandplace.Thisiswhereasenseofuniversality,oratleastpan-culturalism,inmoralityhasaplausiblerationalbasis.Agreatdealoftheworkofenvironmentalethicscanbedonefromwithinthesocialcontract.Mostofenvironmentalpolicyisofthiskind.Humansneedtobehealthy.Health,however,isnotsimplyamatterofbiologyfromtheskin-in.Environmentalhealth,fromtheskin-out,isequallyasimportant.Itishardtohaveahealthycultureonasickenvironment.Morethanthat,humansdesireaqualityenvironment,enjoy-ingtheamenitiesofnature–wildlifeandwildflowers,scenicviews,placesofsolitude–aswellasthecommodities–timber,water,soil,naturalresources.Supportingenvi-ronmentalhealthandaqualityenvironmentcancertainlybecountedasdutieswithinasocialcontract.Environmentalethics,bythisaccount,isfoundedonwhatwecancallahumanrighttonature.TheWorldCommissiononEnvironmentandDevelopmentclaims:‘Allhumanbeingshavethefundamentalrighttoanenvironmentadequatefortheirhealthandwell-being’(1987b:9).Thisincludesthebasicnaturalgivens:air,soil,water,func-tioningecosystems,hydrologiccyclesandsoon.Thesecouldpreviouslybetakenfor519\nHOLMESROLSTON,IIIgranted.Butnowtherightmustbemadeexplicitanddefended.Notethatisnotanyclaimagainstorfornatureitself;ratheritisaclaimmadeagainstotherhumanswhomightdepriveusofsuchnature.Thefourmostcriticalissuesthathumanscurrentlyfacearepeace,population,developmentandenvironment.Allareinterrelated.Humandesiresformaximumdevelopmentdrivepopulationincreases,escalateexploitationoftheenvironmentandfueltheforcesofwar.Thosewhoexploitpersonswilltypicallyexploitnatureasreadily–animals,plants,species,ecosystemsandtheEarthitself.Ecofeministshavefoundthistobeespeciallytruewherebothwomenandnaturearetogetherexploited.Theinter-estsofenvironmentalethicsdonefromperspectivesofpoliticalecology,sustainabledevelopment,bioregionalism,ecojustice,fromanethicsofstewardship,orhumanvirtuesincaring,orasenseofplace–allthesetendtobehumanisticandtorecognizethatnatureandculturehaveentwineddestinies.BryanG.Norton(1991)claimsthatfullyenlightenedanthropocentristsandmorenaturalisticenvironmentalistswillalmostentirelyagreeonenvironmentalpolicy,whathecallsa‘convergencehypothesis’.3Animals:BeastsinFleshandBloodEthicsisforpeople,butisethicsonlyaboutpeople?Wildanimalsdonotmakemanthemeasureofthingsatall.Thereisnobetterevidenceofnon-humanvaluesandvaluersthanspontaneouswildlife,bornfreeandonitsown.Animalshuntandhowl,findshelter,seekouttheirhabitatsandmates,carefortheiryoung,andfleefromthreats.Theysufferinjuryandlicktheirwounds.Animalsmaintainavaluedself-identityastheycopethroughtheworld.Theydefendtheirownlivesbecausetheyhaveagoodoftheirown.Thereissomebodytherebehindthefurorfeathers.Ananimalvaluesitsownlifeforwhatitisinitself,withoutfurthercontributoryref-erence,althoughofcourseitinhabitsanecosystemonwhichitslife-supportdepends.Animalsarevalue-able,abletovaluethingsintheirworld,theirownlivesintrinsicallyandtheirresourcesinstrumentally.Sotherecanandoughttobeananimalwelfareethic;or,someprefertosay,ananimalrightsethic.Suchethicistsmaystillsaythatvalueexistsonlywhereasubjecthasanobjectofinterest,onlynowrecognizingthatthepleasuresandpainsofnon-humansubjectsmustbeconsidered.Atleastsomeofwhatcountsinethicsisgenerictoourkinshipwithanimals,notjustspecifictoourspecies.Commonsensefirstandsciencelaterteachesthatwehumananimalshavemanysimilaritieswithnon-humananimals.Noonedoubtsthatanimalsgrowhungry,thirsty,hot,tired,excited,sleepy.TheproteincodingsequencesofDNAforstructuralgenesinchimpanzeesandhumansaremorethan99percentidentical.Confrontedwithsuchfacts,wehavetophilosophizeoverthem.Theconclusionseemstofollowthat,whateverouruniquedifferencesasHomosapiens,thereisalsoakinshipwithothers.Byparityofreasoning,itseemsthatwhathumansvalueinthem-selves,iftheyfindthiselsewhere,theyoughtalsotovalueinnon-humanothers.Wevaluewhatdoesnotstanddirectlyinourlineagebutisenoughlikeourselvesthatwearedrawnbyspillovertosharedphenomenamanifestinothers.Theprincipleof520\nENVIRONMENTALETHICSuniversalizabilitydemandsthatanethicistrecognizecorrespondingvaluesinfellowpersons.Growthinethicalsensitivity,orvirtue,hasoftenrequiredenlargingthecircleofneighbourstoincludeotherracesandcultures.Butthesewideningcirclesdonotendwithreciprocatingmoralagents.Acommunitarianethicsfindsenlargingcon-centriccirclesaroundthemoralself:family,localcommunity,nation,humankind,and–inasurroundingthoughmoreremotecircle–animals.Amoosedoesnotsufferthewintercold,aswemight(humanshavingevolvedinthetropics).Perhapsthewarblerisnotgladwhenitsings.Butonemustnotcommitthehumanisticfallacyofsupposingnonaturalanaloguestowhathumansplainlyvalue.Wehaveeverylogical,biologicalandpsychologicalreasontovaluepositivedegreesofkinship.Therewillariseconflictsofinterest.Theremightevenbebadkinds(rat-tlesnakes?),butprimafacie,atleast,thesekindredlivescount.Theyaregoodadaptedfitsintheirplaces,co-evolvedwithothers.Presumptively,animallifeisanEVOLUTION-ARYSUCCESS(pp.320–30)andagoodthing.Somemaythinkitlogicallyorpsychologicallyimpossibletovaluekindsofexperi-encethatwecannotshare(thoseofthesnakes).True,animallivesdonotcoincidewithourown,andtherearerealmsofexperiencethatwecannotreachandwhicharedifficulttoevaluate.Butneithershouldweunderestimatethehumangeniusforthoughtfulappreciationandconsiderablerespectforalienforms.Meanwhile,theclaimsofkindredanimalsoughttocountinenvironmentalethics.4Organisms:RespectforLifeAbiocentricethicsasksaboutappropriaterespecttowardalllivingthings,notonlythewildlifeandfarmanimals,butnowthebutterfliesandthesequoiatrees.Otherwise,mostofthebiologicalworldhasyettobetakenintoaccount:loweranimals,insects,microbes,plants.Over96percentofspeciesareinvertebratesorplants;onlyatinyfractionofindividualorganismsaresentientanimals.Consideringplantsmakesthedifferencesbetweenbiocentrismandananimalrightsethicclear.Aplantisaspontaneouslifesystem,self-maintainingwithacontrollinggeneticprogramme(thoughwithnocontrollingcentre,nobrain).Aplantisnotasubject,butneitherisitaninanimateobject,likeastone.Plants,quitealive,areunifiedentitiesofthebotanicalalthoughnotofthezoologicalkind;thatis,theyarenotunitaryorgan-ismshighlyintegratedwithcentredneuralcontrol,buttheyaremodularorganisms,withameristemthatcanrepeatedlyandindefinitelyproducenewvegetativemodules,additionalstemnodesandleaveswhenthereisavailablespaceandresources,aswellasnewreproductivemodules,fruitsandseeds.Plantsdonothaveends-in-view,andinthatfamiliarsensetheydonothavegoals.Yettheplantgrows,reproduces,repairsitswoundsandresistsdeath,maintainingabotanicalidentity.Allthis,fromoneperspective,isjustbiochemistry–thewhirandbuzzoforganicmolecules,enzymes,proteins–ashumansaretoo,fromoneperspective.Butfromanequallyvalid–andobjective–perspective,themorphologyandmetabolismthattheorganismprojectsisavaluedstate.Vitalisamoreamplewordnowthanbiological.Wecouldevenarguethatthegeneticsetisanormativeset;itdistinguishesbetweenwhatisandwhatoughttobe–notofcourseinanymoralor521\nHOLMESROLSTON,IIIconscioussense–butinthesensethattheorganismisanaxiologicalsystem.Thegenomeisasetofconservationmolecules.Alifeisspontaneouslydefendedforwhatitisitself.Anobjectorcansay,‘Theplantsdon’tcare,sowhyshouldI?’Butplantsdocare–usingbotanicalstandards,theonlyformofcaringavailabletothem.Theplantlifeperseisdefended–anintrinsicvalue.Thoughthingsdonotmattertotrees,agreatdealmattersforthem.Weask,‘What’sthematterwiththattree?’Ifitislackingsunshineandsoilnutrients,wearrangeforthese,andthetreegoestoworkandrecoversitshealth.Suchorganismsdo‘takeaccount’ofthemselves;andweshouldtakeaccountofthem.Forclassicalethicists,allthisseemsodd.Plantsarenotvaluerswithpreferencesthatcanbesatisfiedorfrustrated.Itseemscurioustosaythatwildflowershaverights,ormoralstanding,orneedoursympathy,orthatweshouldconsidertheirpointofview.Butbiocentristsclaimthatenvironmentalethicsisnotmerelyanaffairofpsychology,butofbiology.Theconcentriccircleskeepexpanding.Everyorganismhasagood-of-its-kind;itdefendsitsownkindasagoodkind.Perhapsmanistheonlydeliberativemeasurerofthings,butmandoesnothavetomakehimselftheonlymeasureheuses.Lifeisabettermeasure.AdaptiveFitsandInclusiveEthicsEthicsandbiologyhavehaduncertainrelationsoverrecentcenturies.Anoften-heardargumentforbidsmovingfromwhatisthecase(adescriptionofbiologicalfacts)towhatoughttobe(aprescriptionofduty);anywhodosocommittheNATURALISTICFALLACY(pp.805–6).Ontheotherhand,ifspontaneousnaturallivesareofvalueinthemselves,andifhumansencounterandjeopardizesuchvalue,itwouldseemthathumansoughtnottodestroyvaluesinnature,notatleastwithoutoverridingjustificationproducinggreatervalue.Perhapssomeoftheseplantkindsarebadkinds(poisonoak?),butagain,aswiththeanimals,intheirplacetheyareadapted,theyarepresumptivelywellsuitedforlifeintheirniches.Thecounter-riskisafallacyofmislocatedvalue,ahumanisticmistaketakingvaluetolieexclusivelyinthesatisfactionofourhumanpreferences.Theproblemisthat,despitetheexcellenceofourincreasinglyscientificaccountsinbiology,naturehasbeenmappedphilosophicallyasamoralblankspace,asvalue-freeinandofitself.Theologically,weforgotGod’sgoodcreation.Ethicsissignificantlyamatterofrespectingothersforwhattheyareinthemselves,apartfrommyself-interests.Thatisaltruism.Butahumanisticethicisnotreallyyet‘altruistic’towardanynon-humanothers;evenananimalrightsethicfindsvalueonlyinouranimalcousins.Environmentalethics,themostaltruisticofethics,takesaccountsofallotherlivingorganisms.Thisnowheredeniestrade-offsanddegreesofsignificanceandvalue.Givenourownbiologicalneeds,humanstoohavetomakeawaythroughtheworld,andthisrequiresdefendingourselves(againstpoisonoak)andcapturingvaluespresentinplantsandanimals,forfoodandshelter.Humansdosonotonlyasbiologicalagentsbutasmoralagents.Wehave,ifyoulike,arighttoeat;wealsohavearesponsi-bilitytorespectthevitalitiesofthefaunaandfloraaroundus.Afullethicsisinclusiveofeverylivingorganism.522\nENVIRONMENTALETHICS5SpeciesandBiodiversity:LifelinesinJeopardyAtthespecieslevel,responsibilitiesincrease.Sodoestheintellectualchallengeofdefendingdutiestospecies.Whatarespecies?Thequestionisscientific,onetobeansweredbybiologists.Havehumansdutiestothem?Thequestionisethical,tobeansweredbyphilosophers.Onabiologicallevel,speciesarehistoricallineages.Ursusarctos(thegrizzlybear)isadynamicongoingbear-bear-bearsequence,aspecificformoflifehistoricallymaintainedovergenerationsforthousandsofyears.Thesowdevotesherlifetohercubs.Theindividualrepresents(re-presents)aspeciesineachnewgeneration.Itisatokenofatype,andthetypeismoreimportantthanthetoken.MoralConcernforSpeciesLinesAswithplants,classicalethicistswillfindspeciesoften(thoughnotalways)tobeusefulnaturalresources.Buttheyfindspeciesobscureobjectsofdirectmoralconcern.Species,thoughtheycanbeendangered,cannot‘care’–soreturnstheobjectionweheardbefore.Theyjustcomeandgo.Around98percentofthespeciesthathaveinhabitedEarthareextinct.Mostethicistssaythatoneoughtnotneedlesslytodestroyendangeredspecies;virtuouspersonsarenotvandals.Butmanywillgivehumanisticreasons,andthinkthisenough.Moreradicalenvironmentalethicistsclaimthatoneoughttorespecttheselifelines.Biologicalidentityneednotattachsolelytotheindividualcentredormodularorganism,ananimaloraplant.Biologicalidentitycanbereassertedgeneticallyovergenerations,persistingasadiscrete,vitalpatternovertime.Thelifethattheindividualhasissome-thingpassingthroughtheindividualasmuchassomethingitintrinsicallypossesses,andacomprehensiverespectforlifefindsitappropriatetoattachdutydynamicallytothespe-cificformsoflife.Thevalueresidesinthedynamicform;theindividualinheritsthis,exemplifiesitandpassesiton.Theappropriatesurvivalunitistheappropriatelocationofpersistentvaluing,wherethedefenceoflifegoesoninregeneration,asindividualmembersofaspeciesaregivenovertosurvivaloftheirkind.Plantsandanimalsnotonlydefendtheirownlives;theydefendtheirkinds.Suchkindsarethedynamismoflife.AshutdownofthelifestreamonEarthisthemostdestructiveeventpossible.InthreateningEarth’sbiodiversity,thewrongthathumansaredoingisstoppingthehistoricalvitalityoflife.Everyextinctionisanincrementaldecayinthisstoppinglife.‘Oughtspeciesxtoexist?’isadistributiveincrementinthecollectivequestion,‘OughtlifeonEarthtoexist?’SincelifeonEarthisanaggregateofmanyspecies,whenhumansjeopardizespecies,theburdenofprooflieswiththosewhowishdeliberatelytoextinguishaspeciesandsimultaneouslytocareforlifeonEarth.Fewpastphilosophershaveevenraisedthequestionofdutiestospecies,muchlessansweredit.Nowsuchdutyisbecomingclearer.Ifitmakesanysensetoclaimthatoneoughtnottokillindividualswithoutjustification,itmakesmoresensetoclaimthatoneoughtnottoextinguishspecieslines,withoutextraordinaryjustification.Thisisakindofsuper-killing.523\nHOLMESROLSTON,III6Ecosystems:TheLandEthicAldoLeopold,aforester–ecologistandprophetofenvironmentalethics,claimed,famously:‘Athingisrightwhenittendstopreservetheintegrity,stability,andbeautyofthebioticcommunity.Itiswrongwhenittendsotherwise.’‘Thatlandisacommunityisthebasicconceptofecology,butthatlandistobelovedandrespectedisanextensionofethics’(Leopold1969:224–5,viii–ix).Inaholisticethic,thisecosystemiclevelinwhichallorganismsareembeddedalsocountsmorally–insomerespectsmorethananyofthecomponentorganisms,becausethesystemicprocesseshavegenerated,continuetosupport,andintegratetensofthousandsofmemberorganisms.Theappropriateunitformoralconcernisthefundamentalunitofdevelopmentandsurvival.That,wewerejustsaying,isspecieslines.Butaspeciesiswhatitiswhereitis,encircledbyanecology.Inanaxiologicalethics,hereissystemicvalue,aswellasinstrumentalandintrinsicvalue.Valueliesinprocessesaswellasinproducts.Tovalueindividualsamongthefaunaandfloraandnottheevolutionaryandecologicalprocessesislikevaluingtheeggsthatthegoldengooseproducesmorethanthegooseabletoproducethem.Itwouldbeamistaketovaluethegooseonlyinstrumentally.Agoosethatlaysgoldeneggsissystemi-callyvaluable.Howmuchmoresoisanecosystemthatgeneratesmyriadsofspecies,oreven,aswesoonsee,anEarththatproducesbillionsofspecies,ourselvesincluded.Alandethicmightseemanaturalisticethic,butpeoplearelivingonthisland,andsonatureandculturesoonmix.Tryingtomapthehumanenvironments,wearevaluingthreemainterritories:theurban,theruralandthewild–allthreeofwhicharenecessaryifwearetobethree-dimensionalpersons.Natureismuchpresentinthehybridhabitatsofrurallandscapes;weneedanethicforagro-ecosystems.Wildlifecanextensivelyremainonlandscapesputtomultipleuse;andsoweneedanethicofwildlifemanagement.Weneedanethicforforestsandfarmlands,forthecountryside.Natureispresentin,andasupportof,ourcitiesaswell.PeopleonLandscape:EnvironmentalPolicyandManagingNatureEnvironmentalethicshastobedirectedtohumandominated,managed,disturbed(andoftendegraded)landscapes.Suchalandethicmustbeinformedaboutecosystemhealth,butmorefocusedonhumanecology,onpoliticalecology.Governmentandbusinessarelargeinfluencesinourlives;bothhavevastamountsofpowertoaffecttheenvironmentforgoodorill.Socialsystemsmakehumansbehaveastheydotowardtheirenvironment,andanyeffectivereformationwillhavetobeworkedoutinreformed,moreenviron-mentallysensitivesocialinstitutions.Environmentalethicscannotbeanecosystemethicpureandsimple;thereisonlyanethicabouthumansrelatingtotheirecosystems,intheeconomiesinwhichtheylive.Environmentalethicsmustbecorporate;actionmustbetakeninconcert:greenpol-itics,greenbusiness.Thenaturalenvironmentiscruciallya‘commons’,apublicgood.Policieswillneedtorelatesuchacommonstocapitalism,ownershipofthemeansofpro-duction,marketforces,theconcernsoflabour,realestatedevelopmentpolicies,propertyrightsofindividuals,populationcontrol,equitabledistributionoftheproductsmadefromnaturalresources.Thereisno‘invisiblehand’thatguaranteesanoptimalharmonybetweenapeopleandtheirlandscape,orthattherightthingsaredoneinencounterwithfauna,flora,species,ecosystems,orregardingfuturegenerations.524\nENVIRONMENTALETHICSHumansaremostlymovedtoactintheirself-interest;andtheywilldosotothedegra-dationoftheenvironment–unlessenvironmentalpolicygivesthemincentiveother-wise.Short-termself-interestwillgetoutofhand,especiallywhencoupledwithsocialpower.Thus,torespectecosystemsandkeepthemhealthy,toensureenvironmentalqualityeveninahumanisticethics,thereisaneedforlawstoregulateprivateandbusi-nessuse;theseregulationsareimposedinthelonger-rangepublicinterestbytheforcesofdemocracy.Whataboutspontaneouswildnature?Wildernessareasandnaturereservesarepartofourglobalenvironment,andyetnotourhumanhabitat.Thewildisanenvironmentthathumansneedandoughttorespect;theymayliketovisitthere.Butthewildisnotanenvironmentinwhichwecanresideandstillbehuman.‘Manisbynatureapoliti-calanimal’,saidAristotle–theanimalwhobuildsandinhabitsapolis,atown.Thatiswhy,somesay,ethicsarisestogovernconductinthepolis,withitssocialcontract,ori-entingbehaviourtoprotectthegoodsofhumannatureandculture.Hence,theysay,ethicsdoesnotbelonginthewild.Itisforpeople,inurbanorruralenvironments.But,again,amoreradicalenvironmentalethics,resolvingtobequiteinclusiveandcomprehensive,holdsthathumanscanandoughttosetasidewildareasforwhattheyareinthemselves,areaswhichwetrytomanageaslittleaspossible,ortomanagehumanusesofthemsoastoletnaturetakesitscourse,asfaraswecan.Virtuouspersonsoughttorespecttheintegrity,thefreedomoflifeinallitswildness.True,humansarethedominantspeciesonthelandscape,whichtheymustmanage.Buthumansarealsoamoralspecies,whocanandoughttorespectevolutionaryecosys-tems–atleastonrepresentativepartsofthelandscape.An‘ethicsofplace’hasatendencytoseevalueslargelyasthosethathumaninhab-itants‘place’onnature.TheEnglishlovetheircountryside.Butoughtthisethicnotalsobesensitivetovaluesthatarealready‘inplace’beforewehumansarrivetodwellthere(warblersdefendingthegoodoftheirkind)?Partoftheneededethicdoesdemandaconstructedsenseofplace;apersonneedsanembodiedsenseofresidence.Anotherpartrespectsnatureasitisfoundtobeinitself,oblivioustothespecificcircumstancesofaperson’sparticulartimeandplace.Thepersonalviewmustbecomplementedbyaregionalview,indeedaglobalview.Bioregionalismemphasizeslivingonregionallandscapes.Themostworkableethiciswherepersonsidentifywiththeirgeography.Peoplearelikelytobemostmotivatedbywhattheyhaveatstakeontheirat-homelandscapes.True,oneoughttohaveconcernforendangeredspecies,vanishingwildlife,intrinsicnaturalvaluesorwilder-nessconservation;butthatisnotwhatorientsday-to-daybehaviour.Whatispoliticallypossibleisconcernaboutthecountrysideofeverydayexperience.Afterall,ecologyisaboutlivingathome(Greek:oikos,‘house’).Thatiswherethelandethicreallyoper-ates.Thatiswherepeoplecanact,wheretheyvoteandpaytaxes.Theyneedtobe‘natives’,asmuchas‘citizens’.MichelSerres(1995:20)arguesthat‘theoldsocialcontractoughttobejoinedbyanaturalcontract’.Abioregion,saysKirkpatrickSale(1985:43),is‘aplacedefinedbyitslifeforms,itstopography,anditsbiota,ratherthanbyhumandictates;aregiongovernedbynature,notlegislature’.Afocusonbioregionspermits‘ecosystemmanagement’,amuchlaudedgoal.Bioregionalismappealstogeographers,landscapearchitects,developers,statelegislators,countycommissioners–allthosechargedwithdecisionsabouta525\nHOLMESROLSTON,IIIqualityenvironment.Humansneedtolearnto‘reinhabit’theirlandscapes.Thisisenvironmentalethicsonahumanscale.AldoLeopoldconcludeswithalandethicthatherecommendsuniversally.Itisnoaccident,however;rather,itisessentialthattheearlierpagesofhisSandCounty(Wisconsin)AlmanacrememberaJanuarythaw,thespringfloweringofDraba,theAprilmatingdanceofthewoodcock.Leopold’sbiographicalresidenceisthepersonalbackingtohisethic.Anenvironmentalethicneedsrootsinlocality.Takingamodelfromecology,thedeepecologymovementemphasizesthewaysinwhichhumans,althoughindividualselves,canandoughttoextendsuchselvesthroughawebworkofconnections.Thehuman‘self’isnotsomethingfoundfromtheskin-in,anatomisticindividualsetoveragainstotherindividualsandtherestofnature.Ecologydissolvesanyfirmboundarybetweenhumansandthenaturalworld.Ecologydoesnotknowencapsulatedegosoveragainsttheirenvironment.Ecologicalthinkingisakindofvisionacrossboundaries.Humanshavesuchentwineddestinieswiththenaturalworldthattheirrichestqualityoflifeinvolvesalargeridentificationwiththesecommunities.Suchtransformationofthepersonalselfwillresultinanappropriatecarefortheenvironment.Ecofeministsmayaddthatwomenarebettersuitedforsuchcaringthanmen–atleastmentoomuchdominatedbythe‘dominion’view,toomuchinclinedtobemanagers.7Earth:EthicsontheHomePlanetViewsofEarthfromspacearethemostimpressivephotographsevertaken.Theyarethemostwidelydistributedever,havingbeenseenbywelloverhalfthepersonsonEarth.Fewarenotmovedtoamomentoftruth,atleastintheirpensivemoods.ThewholeEarthisaestheticallystimulating,philosophicallychallengingandethicallydisturbing.‘Iremembersovividly’,saidMichaelCollins,‘whatIsawwhenIlookedbackatmyfragilehome–aglistening,invitingbeacon,delicateblueandwhite,atinyoutpostsuspendedintheblackinfinity.Earthistobetreasuredandnurtured,some-thingpreciousthatmustendure’(Collins1980:6).ThereisavisionofanEarthethicinwhathesees.But,replytheanthropocentrists,thatthisis‘ourhomeplanet’revealstherealfocusofethicalconcern:humansandtheirsustainablefuture.HumanscanandoughttobeheldresponsibleforwhattheyaredoingtotheirEarth,whichistheirlife-supportsystem.But–sothisargumentgoes–thesearedutiesowedbypeopletootherpeople;caringfortheplanetisameanstothisend.Environmentalethicsonglobalandregionalscalesisinextricablycoupledwithdevelopmentethics.TheRioDeclarationbegins:‘Humanbeingsareatthecentreofconcernsforsustainabledevelopment.Theyareentitledtoahealthyandproductivelifeinharmonywithnature’.TheUnitedNationsWorldCommissiononEnvironmentandDevelopmentdeclares:‘Sustainabledevelopmentisdevelopmentthatmeetstheneedsofthepresentwithoutcompromisingtheabilityoffuturegenerationstomeettheirownneeds’.Thatappliestoagriculture,forestry,wateruse,pollutionlevels,indus-try,resourceextraction,urbanization,nationalpoliciesandstrategies.‘Sustainable’526\nENVIRONMENTALETHICScoupledwith‘development’expectscontinuedgrowth,butnotsuchasdegradesoppor-tunitiesandenvironmentsforthefuture.Ethics–thisargumentclaims–oughtnottoconfusepeopleandtheirEarth.Earthisabigrockpilelikethemoon,onlyoneonwhichtherocksarewateredandillumi-natedinsuchawaythattheysupportlife.Earthisnodoubtpreciousasameansoflifesupport,butitisnotpreciousinitself.Thereisnobodythereinaplanet.Thereisnoteventheobjectivevitalityofanorganism,orthegenetictransmissionofaspeciesline.Earthisnotevenanecosystem,strictlyspeaking;itisaloosecollectionofmyriadsofecosystems.Soanyethicistmustbetalkingloosely,perhapspoeticallyorromantically,ofvaluingEarth.Earthisamerething,abigthing,aspecialthingforthosewhohappentoliveonit,butstillathing,andnotappropriateasanobjectofintrinsicorsystemicvaluation.Wedonothavedutiestorocks,air,ocean,dirtorEarth;wehavedutiestopeople,orsentientthings.Wemustnotconfusedutiestothehomewithdutiestotheinhabitants.Nature,notultimatelyimportant,is(intheliteralsense)provisionallyimportant.Anyconditionofnaturethatsuppliesandsustainssuchopportunitieswillbeacceptable.Theradicalenvironmentalethicfinds,however,thatthishumanisticaccountfailstorecognizethegloballyrelevantsurvivalunit:Earthanditsbiosphere.Thebottomline,trans-culturalandnon-negotiable,isasustainablebiosphere.Thatistheultimateexpandingcircle:thefullEarth.Theus-and-our-sustainable-resourcesviewisnotasys-temicanalysisofwhatistakingplace.Theplanetisaself-organizingbiosphere,whichhasproducedandcontinuestosupportalltheEarthboundvalues.Earthisthesourceofvalue,andthereforevalue-able,abletoproducevalueitself.Thisgenerativityisthemostfundamentalmeaningoftheterm‘nature’,‘togivebirth’.Donothumanssome-timesvalueEarth’slife-supportingsystemsbecausetheyarevaluable,andnotalwaystheotherwayround?True,humansaretheonlyevaluatorswhocanreflectaboutwhatisgoingoninanimals,plants,specieslines,overevolutionaryhistory,oratglobalscales,orwhocandeliberateaboutwhattheyoughttodoconservingit.Whenhumansdothis,theymustsetupthescales;andhumansarethemeasurersofthings.Animals,organisms,species,ecosystems,Earth,cannotteachushowtodothisevaluating.Buttheycandisplaywhatitisthatistobevalued.Theaxiologicalscalesweconstructdonotconstitutethevalue,anymorethanthescientificscalesweerectcreatewhatwetherebymeasure.Humansarenotsomuchlightingupvalueinamerelypotentiallyvaluableworld,astheyarepsychologicallyjoiningongoingplanetarynaturalhistoryinwhichthereisvaluewhereverthereispositivecreativity.AnaxiologicalethicsoughttooptimizethevaluelevelsanddiversityonEarth,bothnaturalandcultural.Toputthistheologically,humansaretrustees,aswellasstewardswithdominion.Atdepth,suchanEarthethicsaskswhethertheEuropeanEnlightenmentiscompat-iblewiththeemergingecologicalmovement,boththeoreticallyandpractically.Science,technology,industry,democracy,humanrights,freedom,preferencesatisfaction,max-imizingbenefitsovercosts,consumerism–allthese‘managementethics’areoutcomesoftheEnlightenmentworldview.Andtheyareallseriouslyimplicatedascausesoftheenvironmentalcrisis.MuchoftheenthusiastichumanismthattheEnlightenmentstoodforhasbeenagoodthinginmoderntimes;buttoday,withanenvironmentalturn,itneedstobeecologicallychastened.Ethicsneedstobecomepostmodern.527\nHOLMESROLSTON,IIIDevelopmentintheWesthasbeenbasedontheEnlightenmentmythofendlessgrowth.ButintheUnitedStatesandEurope,whetheroneconsidersagriculturaldevel-opment,forestscut,riversdammedanddivertedforwater,landsfenced,mineralsextracted,orhighwaysandsubdivisionsbuilt,thenexthundredyearscannotbelikethelasthundredyears.Noneofthedevelopednationshaveyetsettledintoasustain-ablecultureontheirlandscapes.Onthesescales‘sustainable’alsomeans‘fair’or‘just’,anethicsofeco-justice.OnEarth,thedevelopednationsholdaboutone-fifthoftheworld’s5billionpersons,andtheyproduceandconsumeaboutfour-fifthsofallgoodsandservices.Theunderdevel-opednations,withfour-fifthsoftheworld’speople,produceandconsumeone-fifth.Ofthe90millionnewpeopleonEartheachyear,85millionappearintheThirdWorld,thecountriesleastabletosupportthem,andtheresultispovertyandenvironmentaldegradationinafeedbackloop.Meanwhile,the5millionnewpeopleintheindustrialcountrieswillputasmuchstrainontheenvironmentasthe85millionnewpoor.Theover-consumptionprobleminthedevelopednationsislinkedwiththeunder-consumptionprobleminthedevelopingnations,andthisresultsinincreasingenvi-ronmentaldegradationinbothsetsofnations.Sustainabledevelopmentmustclosethegapbetweentherichandthepoor,betweenandalsowithinnations.Eveniftherewereanequitabledistributionofwealth,thehumanpopulationcannotgoonescalatingwithoutpeoplebecomingmoreandmorepoor,becausethepiehastobeconstantlydividedintosmallerpieces.Eveniftherewerenofuturepopulationgrowth,consump-tionpatternscannotgoonescalatingonafiniteEarth.Therearethreeproblems:over-population,over-consumption,andunder-distribution.Oncethemarkofaneducatedandethicalpersoncouldbesummedupascivitas,theprivileges,rightsandresponsibilitiesofcitizenship.Peopleoughttobeuprightandmoral,productiveintheircommunities,leadersinbusiness,theprofessions,govern-ment,church,education.Thatwastheresponsibilitythatwentwithone’srights.Themarkofavirtuouspersontoday,increasingly,issomethingmore–soenvironmentalethicistsclaim.Itisnotenoughtobeagood‘citizen’,forthatisonlyhalfthetruth;weare‘residents’dwellingonlandscapes.Acenturyago,acallforcommunitywastypi-callyphrasedasthebrotherhoodofmanandthefatherhoodofGod.Formostofthetwentiethcenturythecallwasphrasedasjusticeandhumanrights.Inthiscenturysuchacallmustbemoreecologicalandlesspaternalistic,lesshumanisticandmoreglobal.Weareexpandingethics:itisnotjustwhatasocietydoestoitsslaves,women,blacks,minorities,handicapped,childrenorfuturegenerations,butwhatitdoestoitsfauna,flora,species,ecosystemsandlandscapesthatrevealsthecharacterofthatsociety.WehumansareEarthlingsandcarefortheEarthisadevelopingandanultimatehumanvirtue.ReferencesCompletelistsofanthologiesandsystematicworksmaybefoundontheInternationalSocietyforEnvironmentalEthicswebsitebibliography,under‘Anthologies’and‘SystematicWorks’.Thereisalsoalistof‘IntroductoryArticles’andsyllabusesofcoursesatcollegesanduniver-sities:http://www.cep.unt.edu/ISEE.htmlJournals:EnvironmentalEthics,EnvironmentalValues,EthicsandtheEnvironment,JournalofAgriculturalandEnvironmentalEthics.528\nENVIRONMENTALETHICSBotzler,R.G.andArmstrong,S.J.(eds)1998:EnvironmentalEthics:DivergenceandConvergence,2ndedn.NewYork:McGrawHill.Collins,M.1980:Foreword.InR.A.Gallant,OurUniverse.Washington,DC:NationalGeographicSociety.DesJardins,J.R.1997:EnvironmentalEthics:AnIntroductiontoEnvironmentalPhilosophy,2ndedn.Belmont,CA:Wadsworth.Johnson,L.E.1991:AMorallyDeepWorld:AnEssayonMoralSignificanceandEnvironmentalEthics.Cambridge:CambridgeUniversityPress.Leopold,A.1969[1949]:ASandCountyAlmanac.NewYork:OxfordUniversityPress.Norton,B.G.1991:TowardUnityAmongEnvironmentalists.NewYork:OxfordUniversityPress.Pojman,L.P.(ed.)2001:EnvironmentalEthics:ReadingsinTheoryandApplication,3rdedn.Belmont,CA:Wadsworth.Sale,K.1985:DwellersintheLand:TheBioregionalVision.SanFrancisco:SierraClub.Serres,M.1995:TheNaturalContract.AnnArbor:UniversityofMichiganPress.UnitedNationsConferenceonEnvironmentandDevelopment1992:TheRioDeclaration.UNCEDDocumentA/CONF.151/5/Rev.1,June13.UnitedNationsWorldCommissiononEnvironmentandDevelopment1987a:OurCommonFuture.Oxford:OxfordUniversityPress.UnitedNationsWorldCommissiononEnvironmentandDevelopment1987b:EnvironmentalPro-tectionandSustainableDevelopment:LegalPrinciplesandRecommendations.London/Dordrecht:GrahamandTrotman/MartinusNijhoff.DiscussionQuestions1Arehumanstheonlyvaluable,valuingagentsinanotherwisevalue-freeworld?2Dophilosophershaveaspecialroleinthedevelopmentofenvironmentalethics?3Dovaluesariseonlywiththeinterestsandpreferencesofhumans?4Ishumanisticphilosophynecessarilyanthropocentric?5Howsimilaristheinseparabilityofcultureandnaturetotheinseparabilityofmindsandbodies?6Dohumansneedtoincludenatureintheirethicsandneedtoincludethemselvesinnature?7Dohumanshaveseriousdutiesonlytoeachother?8Arethereethicalconcernsthatgobeyondthescopeofasocialcontract?9Isthehealthoftheenvironmentofethicalimportanceaswellasthehealthofhumanbeingswithintheenvironment?10Whatweightshouldenvironmentalethicsgivetothevaluesofanimals?Whoshouldrepresenttheseconcerns?11Whatdistinguishesananimalrightsethicfromananimalwelfareethic?12Doeskinshiptohumansgiveanimalsmoralstanding?13Howcanwedeterminewhetherthewideningcirclesofethicalconcernendwithreciprocatingmoralagents?14Doesevolutionarysuccessmakeanimallifeagoodthing?15Canwevaluekindsofexperiencethatwecannotshare?16Whatjustifiesgivingmoralvaluetonon-sentientlivingthings?17Isthebiologicalidentityofindividualsorspeciesenoughtojustifymoralconcern?18Plantscanthriveorfallill,andtheyfunctiontopreservetheirlifeortheirspecies.Doesthisgivethemmoralstanding?529\nHOLMESROLSTON,III19Arespontaneousnaturallivesofvalueinthemselves?20Dowehavedutiestospecies?21Ifthetypeismoreimportantthanthetokenforanimalandplantspecies,isthesametrueforthehumanspecies?Whatwouldbethesocialandpoliticalconsequencesofthisview?22Isthefundamentalunitofdevelopmentandsurvivaltheappropriateunitformoralconcern?23Howwouldgreatermoralsensitivitytotheenvironmentaffecthumansocial,politicalandeconomicsystems?24Whowouldbepartiestoa‘naturalcontract’thatmightsupplementasocialcontract?25Isanecologicalethicamatterofargumentorvision?26IfEnlightenmentvaluesareincompatiblewiththeecologicalmovement,whichshouldbesupported?27Howmightjusticeandsustainabledevelopmentberelated?530\n19BusinessEthicsGEORGESENDERLEBusinessandeconomicshavebecomemajordrivingforcesincontemporarysocieties,inboththenationalandtheinternationalcontext.Combinedwithmoderntechnologies,theyshapethoughtandbehaviourandhaveincreasingimpactonmanydomainsoflife:researchanddevelopment,telecommunications,biotechnology,politics,education,culture,religionandthefamily.Thegreatertheinfluenceofbusinessandeconomics,thegreatertheurgencytoensurethattheydevelopintherightdirection.Guidanceforbusi-nessandeconomicsmustcomefromboth‘outside’and‘inside’:fromoutsideintermsofpoliticalpressure,legalregulations,socio-culturalcustomsandlearning;frominsideintermsofproactivebehaviourofbusinessorganizationsandbusinesspeople,self-regulationofindustriesandbusinessalliances.Theoutsideapproachalonecannotachievethisguidancebecauseitlackstheinnercommitmentofbusiness,buttheinsideapproachisalsoinsufficientbecausebusinessisonlyasinglepartofsocietyandneedsadditionaloutsidecontrolandguidance.However,theinsideapproachisbecomingmoreimportantaseconomicactorsenjoymorefreedomandthusbearmoreresponsibility.1IntroductionThatbusinessandeconomicsshoulddevelop‘intherightdirection’isawaytoexpresstheethicaldimensioninherentinthisdomain.Atstakeisnotonlymoralpracticebutalsoethicalreflectionandtheory,or‘businessandeconomicethics’asacademicinquiry.Thefurtherthatglobalizationadvances,thegreaterandmorecomplexthepracticalandtheoreticalchallengesbecome.Untilnow,theachievementsofbusinessethicshavefallenconsiderablyshortofwhatisrequired.Globaltransformationsarehighlycomplexprocessesthatareincompleteandarepoorlyunderstoodintheirdepthandconsequencesforbothglobalandlocalsocieties.Fromtheethicalperspectiveaswell,thechallengesareawesome.Thequestionsconcernnotonlyintersocietalrela-tions(seeRawls1999),butalsoperson-to-personrelationsinmultipleinstitutionsacrossnationalborders(seeSen1999b).Businessethicsmustcometogripswithculturalandreligiouspluralismworldwidewithincountriesandculturesandstriveforacommonethicalgroundinfacingcommonchallenges(seeAGlobalEthic1993;Küng1998).\nGEORGESENDERLEGivenourcurrentsituation,itisadvisabletoapproachbusinessandeconomicethicsinaglobalcontext.Howevervaluabletheapproachesinparticularcountriesandcul-turesmaybe,theyhavenotmatchedtheneedsofaglobalizingworld.Forexample,Frederick’simpressiveACompaniontoBusinessEthics(1999)overwhelminglyreflectsaUSapproach,andHandbuchderWirtschaftsethik(Korffetal.1999)isstronglyshapedbyGermanthinking.Thesecontributionsareimportant,buttheiruseisseriouslylimitedoutsidetheirhomeenvironment.Incontrast,aworldwidesurveyfortheFirstWorldCongressofBusiness,Econom-ics,andEthics1996inTokyosoughtthematerialfromwhichtodevelopacompara-tiveconceptualframeworktodeterminethechallenges,initiativesandachievementsintheemergingfieldofbusinessethics(JournalofBusinessEthics,October1997).Theneedforsuchaframeworkhasbeenespeciallyfeltbythoseconductingresearchandteachingbusinessethicssimultaneouslyindifferentcountriesandcultures.Thischapterwillintroduceaconceptualframeworkforbusinessethicsandthenchartthecurrentcontoursofthefield.Finally,itwilldiscussdifferenttheoreticalapproachestobusinessethicsbeforeafewconcludingremarks.2AConceptualFrameworkforBusinessEthicsThedevelopmentofaconceptualframeworkthataccommodatesdifferentapproachestobusinessethicscanovercomewidespreadconfusioninthefieldandpromoteusefuldialogue.Byconcentratingontheinterfacebetweenempiricalandtheoreticalstudies,itcanavoidmanytheoreticalcontroversies,offerstrategiesforempiricalinvestigationsandprovidegroundingforoperationalizationandmeasurement.Aconceptualframeworkisneitherafull-fledgedbodyofknowledgenorablueprint.Nevertheless,itprovidesaconsistentorganizationofthekeyelements,relationsandtermsofthefieldunderinvestigation.Itcanbeusedbydifferenttheoreticalapproachesandcanbespecifiedinawidevarietyofways.Ithastheoreticalimplicationsconcern-ing,forexample,therelationshipbetweentheoryandpractice,thephilosophicalfoun-dationsofECONOMICS(pp.388–9),themoralstatusofeconomicorganizationsandanthropologicalassumptionsabouttheMIND–BODYRELATIONSHIP(chapter5).BusinessethicsisakindofAPPLIEDETHICS(chapter16)andsharescommonfeatureswithotherkindsofappliedethics,suchasBIOETHICSANDMEDICALETHICS(chapter17),legalethics,engineeringethics,mediaethicsandcomputerethics.Atthesametime,approachestoappliedethicsandbusinessethicsvaryconsiderablyaccordingtodifferentconceptionsofethics,thefieldsofapplicationandthewaysofconceivingtherelationshipbetweentheethicsandthefield.Theconceptualframeworkdevelopedhererelatesbusinessethicstodecision-makingandactionanddistinguishesamongdescriptiveethics,normativeethicsandmeta-ethics,withaspecialfocusonNORMATIVEETHICS(pp.201–4).Asaconceptualframework,itdoesnotadvanceaparticularethicaltheory,butplacesmuchemphasisonthestructuringofthefieldofbusinessethicsandusesatwo-leggedapproachtocharacterizetherelationshipbetweenethicsandbusiness.‘Businessethics’isnotclearlydefinedbecause‘business’itselfhasvariousmean-ings.Thiscanleadtoseriousmisunderstandings,especiallywhentranslatedintoother532\nBUSINESSETHICSTable19.1Questionnaireforcountry-andregion-relatedreportsonbusinessethics1SemanticsofbusinessethicsWhatarekeytermsinyourcountry/regiontoexpressethicalissuesinbusiness?Whatistheprevailingunderstanding,definitionandfieldofbusinessethics?2Whataremajorchallengesofbusinessethicsinyourcountry/region?Broadissues?Specificissues?Relevanceofculturalandreligioustraditions?Impactofinternationalbusiness?Whataremajorreasonstoaddressthesechallenges?3Businessethicsactivitiesinthebusinessworld•Whatendeavoursofbusinesspeople,companiesandbusinessassociationshavebeenundertaken?Settingupanddisseminationofcodesofconduct?Ethicstrainingprogrammes?Positionsofethicsofficersorthelike?Publicstatementsandpublicationsofbusinessleaders?Legislation?•Are,andifsohoware,domesticandforeigncompaniesevaluatedfromanethicalpointofview?Byconsumers,workers,investorsandothergroupsandorganizations?•Whatexpectations,criticismsandrequirementsareraisedbyabroaderpublic,e.g.governments,politicalandeducationalorganizations,non-governmentalorganizations,churches,etc.?•Whataremajorreasonsforconductingthesebusinessethicsactivities?Whatmajorfactorshindersuchactivities?4Businessethicsactivitiesinacademia•Whatinstitutionalformsofbusinessethicshavebeenestablished?Chairs,institutes,centres,networks,associations,etc.?Surveys,publicationsaboutthisinstitutionalaspect?•Whatteachingactivitieshavebeenundertaken?Electiveandmandatorycourses,interdisciplinaryseminars,integrationinthestudents’curricula?Surveys,publicationsaboutteachingbusinessethics?•Whatresearchactivitieshavebeenconducted?Surveys,publicationsaboutresearchinbusinessethics?•Whatconsultingactivitieshavebeencarriedout?Surveys,publicationsonconsultingactivities?•Whatparadigm(conception)ofbusinessethicsisinvolvedintheseactivities?5Whatarecommontasksofbusinessandacademiainthefieldofbusinessethics?Dotheyfacecommonchallenges?Whatcommonchallenges?Whatkindofcompetencescananddoeseachsidebringin?6RelationshiptobusinessethicsinothercountriesandregionsWhatexperiencesin,andinsightsfrom,businesspracticeandacademiacanyourcountry/regionoffertopeopleworkinginbusinessethicsinothercountriesandregions?Whatlessonscanyourcountry/regionlearnfromothercountriesandregions?Source:JournalofBusinessEthics,October1997,1477.languagesandcultures.Forinstance,thefamoussaying‘Thebusinessofbusinessisbusiness’ispuzzlingforpeoplewhoarenotconversantwithEnglish.Because‘business’standsfor‘thetaskorjob’,‘theeconomicorganization’and‘tomakeprofit’,theslogancanread‘thetaskofthecompanyistomakeprofit’.However,intermssuchas‘busi-nesscycle’,‘business’isalmostequivalentto‘theeconomy’,sothat‘businessethics’533\nGEORGESENDERLEcanbeunderstoodasaverybroadnotionthatiscomparableto‘economicethics’,theGermanWirtschaftsethikandtheSpanishéticaEconómica.Theincreasinglyusedexpres-sion‘corporateethics’concernstheethicsofbusinessorganizationsaswellasethicsinbusinessorganization(Berenbeim1987).Inthischapter,weshallunderstand‘businessethics’inabroadsense,coveringthewholeeconomicdomainoflife.Understoodinthisway,businessethicsdealswiththeindividualdecision-makingofeconomicactors(suchasmanagersandemployees),theshapingandconductofeconomicorganizations,business-relatedpublicpolicies,eco-nomicsystems,andglobaleconomicandfinancialinstitutions.Weshallassumethatthisdomaincanbeidentifiedanddistinguishedreasonablywellfromotherdomainsoflife(suchasthepolitical–legalandthesocio-culturaldomain)andthateachdomainhasacertainautonomyandtypeofRATIONALITY(pp.395–7).Giventhisunderstanding,wecanseethatbusinessethicsaimstoimprovetheethicalqualityofdecision-makingandactingatalllevelsofbusiness.Fromacross-culturalperspective,itisimportanttodistinguishthreemodesofunderstandingbusinessethics:semantics,practiceandtheory.Attitudestowardspeak-ingaboutbusinessethicsandthetermsuseddiffersignificantlyfromoneculturalsettingtoanother.Moreover,talkingaboutbusinessethicsisusuallynotidenticalwithethicalconduct.Tothinksystematically,ortheorize,aboutbusinessethicsisimportantnotonlyfortheoreticalreasons,suchastheneedforconsistency,criticalscrutinyandindependentevaluation,butalsoforpracticalreasons,suchastheneedtoclarifyconceptualissuesinpractice,toofferimpartialassessmentofbusinessconductandtoprovideseriousconsultationtocompanies.Therefore,althoughthesethreemodesareinterrelated,theyaredistinctandcannotreplaceeachother.However,thedefinitivetestofthevalueofbusinessethicsispractice.2.1Structuringthefieldofbusinessethics:amultiple-levelaction-orientedapproachFundamentalquestionsofethicsare‘WhatshouldIdo?’and‘Whatshouldwedo?’Weareunderpressuretoactandcannotavoidmakingdecisions.Decision-makingandactingareanessentialpartofhumanexistence.AsAlanGewirthstates,‘theindepen-dentvariableofallmoralityishumanaction’(Gewirth1984:12).Inourpluralisticsocieties,wefaceahostofdifferentethicalbeliefsandtheories.WhetherweholdstrongconvictionsonHUMANRIGHTS(pp.31–4),believeintheoverridingpowerofSELF-INTEREST(pp.761–3),adoptapositionofETHICALRELATIVISM(pp.395–7)orarescep-ticalaboutethicsasawhole,wemustmakedecisionsandtakeactions.Weproposeanaction-orientedapproachtobusinessethics,whichtakesaccountofboththeactorsandthestructuresinwhichtheyareembedded.ActioninvolvesFREEDOMOFCHOICE(pp.206–8).Wecanchoosefromamongdiffer-entcoursesofactionwhilebeingsubjecttoconstraints.Actionsareneverlimitless,withoutboundaries,andthelimitationsareoftentheresultsofpreviouschoices.Forexample,marketconditionsandlawsandregulationsconstraintheactionsofeco-nomicactorswhofacethemandcannotchangethemimmediately.Butthesehuman-madeconstraintsarelargelytheoutcomeofprecedingdecision-makingprocesses.Freedomrelatesbothtochoicewithinconstraintsandtochoiceofconstraints(how534\nBUSINESSETHICSChoicewithinConstraints:areasSpaceofResponsibilityfreedomChoiceofConstraints:linesFigure19.1Spaceoffreedomandresponsibility.actorsshapetheconditionsandlimitationsoftheirfutureactions).Bothsetsofcir-cumstancesareessentialtoaconceptualframeworkforbusinessethics(seefigure19.1).Decisionsandactionsalwaysareconcrete,meaningthattheyareexposedtothecomplexitiesoflife.Decision-makersfacecomplexchoiceswithfar-reachingconse-quencesthataredifficulttoforeseeandevaluate.Theycannotcontentthemselveswithanalysingdifferentoptionsanddeterminingtheconditionsofgooddecisionsbecausetheymust(astheChinesesayfordoingbusiness)‘jumpintothesea’:makedecisionsandtakeactions.Yetconcretehumanactionscharacteristicallyneedtheoryinreach-ing‘real’decisionsandactions.‘Goodpractice’needs‘goodtheory’inbusinessethicsaswellasinotherdomains.Choiceswithinconstraintsandchoicesofconstraintsnecessarilyinvolveanethicaldimensionthatcanbearticulatedintermsofethicalbeliefsandethicaltheories.Forthepurposeofourconceptualframework,weshallintroducethekeyterm,‘responsibility’.Althoughresponsibilitydoesnotcoverthefullrangeofethicalvaluesandnorms,itisakeynotionofcontemporarymoralityandhascomplexpracticalandtheoreticalaspects(seeFischer1986,1999;French1984;Glover1970;Jonas1984;SocialPhilosophyandPolicy1999).Weshallassumethattheextentofresponsibilityisafunctionoftheextentoffreedom:thebiggerone’sspaceoffreedom,thebiggerone’sresponsibility(seefigure19.1).Reflectingthetraditionalethicalprinciple‘oughtimpliescan’,thedeterminationofthespaceoffreedomofeconomicactorsiscrucialforallocatingtheirresponsibilities.Thisconcretenotionof‘spaceoffreedom’comesclosetoSen’s(1999a)conceptofasetofrealfreedomsor‘capabilities’andtothedefinitionof‘humandevelopment’as‘enlarg[ing]people’schoices’(UNDP1990:10).535\nGEORGESENDERLEInordertoidentifythesubjectsofresponsibilityasconcretelyaspossible,threedifferentlevelsofactingareproposed,eachofwhichincludesactorswiththeirrespe-ctiveobjectives,interestsandmotivations:themicro-level,meso-levelandmacro-level.Atthemicro-levelthefocusisontheindividual,thatis,whatactors,asemployeeoremployer,colleagueormanager,consumer,supplierorinvestor,do,candoandoughttodoinordertoperceiveandassumetheirethicalresponsibility.Atthislevelwealsoconsidergroups,composedofsmallnumbersofindividualsandwithoutorganizationalstructures,makingcollectivedecisionsandtakingcollectiveactions.Atthemeso-levelweconsiderthedecision-makingandactionofeconomicorganizations,chieflybusinessfirms,butalsotradeunions,consumerorganizationsandprofessionalassociations.Finally,atthemacro-levelweconsidertheeconomicsystemassuchandtheshapingoftheoveralleconomicconditionsofbusiness:theeconomicorderwithitsmultipleinsti-tutionsandeconomic,financialandsocialpolicies.Ateachleveltheactorsareassumedtohavespacesoffreedomfordecision-makingwithcorrespondingethicalresponsibilitiesandtobelimitedbyconstraintsthattheycannotchange,atleastforthetimebeing.Nolevelcansubstituteforanother.Evenifallproblemsatonelevelweresatisfactorilysolved,problemsattheotherlevelscouldremain.Thisthree-levelconception,adoptedbyvariousbusinessethicsscholars(forexample,Goodpaster1992;Solomon1993),contrastswiththecommondistinctionbetweenthemicro-levelandmacro-levelineconomicsandsociology(see,forexample,Coleman1990).Accordingtothethree-levelconception,theindividualpersonisexplicitlyaddressedasmoralactor,differingfromthedecision-makerinmicro-eco-nomicsandmicro-sociology.Thebusinessorganizationisconsideredtobeamoralactor,too,althoughofaspecialnature.Theemphasisofthemeso-levelexpressestheenormousimportanceoforganizationsinmodernsocietiesandissupportedbytheNewEconomicsofOrganization,whichdrawsabasicdistinctionbetweenmarketsandorga-nizations(seeWilliamson1985,1990).Notealsothedistinctionbetweeninstitutionsandorganizations.Forinstance,theinstitutionoftheAmericancorporationatthemacro-levelincludestheessentialfeaturesofallAmericancorporations(Bellahetal.1992:3–18),whileanindividualAmericancorporation,withitsparticularidentity,cultureandconduct,isanorganizationatthemeso-level.Thecentralpointofthisthree-levelconceptionistoperceivethelinksbetweendecision-making,actingandresponsibilityasconcretelyaspossibleandtoprovideconceptualroomforaddressingthedifferencesandconflictsofobjectives,interestsandmotivationswhicharelocatedattheinterfacesbetweendifferentlevels.Wecanaskateachlevelwhatcanbedoneandwhatoughttobedone.Underpressuretoact,singleactorscannotpushtheirresponsibilityontootheractors,norcanresponsibilitybedelegatedfromoneleveltoanother.Whencorporateethicsareatstake,theproblemmustnotbepersonalizedorpushedtothesystemlevel.Ifwetakeconcretedecision-makingandactingseriously,wedonotaskinthefirstplacehowotherindividuals,companiesoreconomicsystemsoughttobehave.Ratherthequestionmustbe:howI,howmycompanyandhowmyeconomicsystemcanperceiveandassumeethicalresponsibility.Nevertheless,actingresponsiblyatalllevelscanrequireethicaldisplacement,atech-niqueofresolvingadilemma,orsometimessolvinganethicalproblem,byseekingasolutiononalevelotherthantheoneonwhichthedilemmaorproblemappears(DeGeorge1993:97).Forexample,inordertopreventsexualharassment,anexplicitcor-536\nBUSINESSETHICSporatepolicyandasustainedcorporateculture(atthemeso-level)mightbenecessaryifachangeofattitudeandbehaviourbyindividualactors(atthemicro-level)isnotsufficient.Ifsuchorganizationsandinstitutionsathigherlevelsdonotexist(asisthecaseindealingwithmanyinternationalproblems),itmightbenecessarytocreatethem.Untilnowthethree-levelconceptionhasbeenappliedtothenationaleconomyoreconomicdomainoflife.Giventheincreasingimportanceofinternationalissues,wecanchangethisconception,forexamplebyaddinganinternationallevelorbysuper-imposingagloballevelunderwhichallissuesaresubsumed.Wepropose,however,anextendedthree-levelmodeltoprovidesufficientconceptualroomforseveraltypesofinternationalrelationsatdifferentlevelsofactingintheemergingworldeconomyandtoidentifytheactorsandtheirresponsibilitiesintheinternationalcontext.Indevelopingourextendedconceptionwemustunderstandthebordersbetweennationalandinternationalmattersatallthreelevels.Atthemicro-levelspecialatten-tionispaidtopersonal(inner-group)relationsandresponsibilitiesacrossnationalborders;forexample,cross-nationalgroupsofmanagersandemployeesorcross-nationalfamilies.Atthemeso-levelthefocusisoninner-organizationalrelationsandresponsibilitiesacrossnationalborders;forexample,multinationalcorporations,internationaltradeunionsorconsumerorganizations.Themacro-levelincludesinner-systemicrelationsandresponsibilitiesacrossnationalborders,incorporated,forinstance,inbilateralagreements,regionaltreatiesorglobalinstitutionsliketheWorldTradeOrganization.Borderscanbepermeableinvariousdegrees.Atoneextremeishermeticseclusion,whichcharacterizedtheformercommunistAlbaniavis-à-visitsneighbours;attheotherextremeisthecompleteabolitionofallbordersandtotalopenness,asproposedbysomeproponentsofglobalization.Intermediatestatesgiveadifferentiatedpictureofinternationalrelationsandencompassallcross-nationalvariants,includingbothimminentconflictsandopportunitiesforco-operationbetweenvariousactors.Thesevariantscanbeclassifiedinfourtypesofinternationalrelations:theforeigncountrytype;empiretype;interconnectiontype;andglobalizationtype.Althoughthesetypescanbefoundatallthreelevels,theexplanationsbelowrefermainlytothemacro-level.Theforeigncountrytypeisexemplifiedbytherelationshipbetweenasmalleconomyorsmallcompanyandaforeigncountry,forexampleSwitzerlandorSchläpferEmbroi-derieswithNigeria.Theinternationalrelationsdiffersignificantlyfromdomesticrela-tionsandhavenorelevantrepercussionsonthem.Theinternationalrelationsareaddedtothenationalframeworkandcanberelativelyeasilydetachedfromit.Eachcountryisdifferent.Foreignershavetoadaptthemselvestothehostcountry,andnationalbordersarerelativelyimpermeable.ExamplesoftheempiretypearetherelationshipbetweenGreatBritainandIndiaduringBritishcolonialism,andtheUnitedFruitCompanyinCentralAmerica.Thistypecharacterizesinternationalrelationsasapurecross-nationalexpansionofdomesticrelationswithoutmodification.Fromthehostcountry’sperspective,theasymmetricpowerrelationshipofteninvolvesmisunderstanding,exploitationandrepression.Repercussionsonthehomecountryarenegligible,sincenationalbordersaremuchmorepermeablefromthehometothehostcountrythanintheoppositedirection.537\nGEORGESENDERLEMicro-levelMeso-levelEmpiretypeGlobalizationtypeMacro-levelInterconnectiontypeForeigncountrytypeNationalInternationalFigure19.2Theextendedthree-levelconceptionofbusinessethics.TheinterconnectiontypecanbeillustratedbytherelationshipbetweenItalyandtheEuropeanUnion.Internationalrelationsdiffersignificantlyfromdomesticrelations,butareintrinsicallyinterconnectedwiththem.Whatisbeyondnationalbordershasimpactondomesticrelationsanddomesticrelationshaveimpactontheinternationalrela-tions.Interdependenceblursthenotionofanationalinterestthatdisregardstheinter-estsofothernationsandsupranationalentities.Althoughtheyarestillimportant,nationalbordersarepervioustosomeextentinbothdirections.Intheglobalizationtype,exemplifiedbyglobalwarming,internationalrelationsaresoimportantthatnationalbordersbecomeirrelevant.Citizensbecomecosmopolitan;multinationalfirmsbecomeglobal;andnation-statesfadeaway.Thistypecaninprin-ciplecomprehendthewholeEarth,althoughuntilnowitismainlyconfinedtopartsofthenorthernhemisphere.Theproposedtypologyofinternationalrelations(seeEnderleandKamm2000)canbevisualizedinfigure19.2.Itiseasytofigureoutthemultiplepossiblecombinationsoftypesandlevelswiththeirspatio-temporaloverlapsandconflicts.Theextendedthree-levelconceptionofbusinessethicsprovidesaframeworktolocatethepluralaffiliations(Sen1999b)oftheeconomicactorsinthecurrentglobalcontext.2.2TherelationshipbetweenethicsandbusinessBusinessethicsasappliedethicshastoreflectontherelationshipbetweenethicsandthefieldofapplication.Fromtheperspectiveoftheaction-orientedapproach,equalimportanceshouldbegiventounderstandingbusinessandeconomicsandunder-standingthereasonedethicalorientation.Thecognitiveandthenormativedimensionsofthesubjectmattershouldbedistinguished,butiftheyaresplitintotwoseparatereal-ities,businessethicsfrom‘inside’wouldbeimpossible.Eithertherelationshipwould538\nBUSINESSETHICSremainexternal,orthedistinguishingboundarybetweenthecognitiveandnormativewouldbeblurred,renderingbothfactualandnormativestatementsmeaningless.Inbusinesspracticethetwodimensionsaregenerallyaccepted.Whenamanagerfacestheproblemofsoilpollution,forinstance,hehastoknowthenature,causesandpossibleremediesofthisENVIRONMENTALHARM(chapter18),alongwiththecostsinvolvedandlegalrequirements.Inaddition,heappliesnormativestandards:whethertofollowtheprincipleofsustainablebusiness,whethertorespectenvironmentalreg-ulations,andsoon.Similarly,whenacompanystruggleswithacorruptbusinessenvi-ronment,itneedssoundethicalguidelines,butalsoneedstounderstandthekindsandmechanismsofcorruptioninordertochooseaneffectivestrategyfordealingwiththiscorruptenvironment.Inacademia,however,thesituationseemstobedifferent.Whilethedescriptive–ana-lyticalperspectiveisextensivelyscrutinized,normative–ethicalquestionshaveattractedlessattention.ThesequestionsrelatetotheVALUESANDNORMS(pp.202–4)whichoughttoguidebusinessdecisionsandactionsandthewaystoproviderationaljustificationforthesevaluesandnorms.Normativeethicsisaphilosophicalundertaking(seeKagan1998)thatfacesurgentissuesinthecontextofcomplexpluralisticsocieties.Incoun-teringwidespreadacceptanceofETHICALRELATIVISM(pp.395–7),itgoesbeyondtheinstrumentalnotionsofthenormativecommonlyusedinacademicbusinessdis-ciplines,suchasdevelopingthemostefficientstrategytoachievethenormofprofitmaximizationormaximizingthesocialwelfarefunctionwithinagivenutilitarianframework.TheTwo-LeggedApproachThetwo-leggedapproachgivesequalimportancetothecognitiveandnormativedimen-sionsandstrivestointegratetheminabalancedway.Bydoingso,itaffectstheunder-standingofbothbusinessandethics.Structuringthefieldofbusinessethicsinthiswayposesnumerouscomplexquestionstoethics,suchasthemoralstatusoforganizationsandsystems,therelationshipbetweenpersonal,organizationalandsystemicethics,thelegitimacyofparticularethicaltheoriesintheglobalcontext,andthefoundationofaglobalethic.Businessethicschallengestheparadigmsofthebusinessdisciplines.Ifman-agementtheory,marketing,accountancy,economicsandfinancearevalue-free,aninte-grationofcognitiveandnormativedimensionswithinthedisciplinesisexcluded,andethics,atbest,hasanexternalroletoplay.Iftheseparadigmsincorporateanormativedimensionthatisopentoethicalreasoning,theintegrationcanbeeffectedfromwithin.Ofcourse,businessdisciplinesarenottherebytransformedintoethicaldisciplines,buttheyprovidebridgeheadsintoethics.AmartyaSen’s(1987)distinctionbetweentheengineeringapproachandtheethics-relatedapproachcanhelpustounderstandthevalueofthetwo-leggedapproach.Theengineeringapproachprimarilyfocusesonlogisticalissues:whatmeansshouldonechooseinordertoachieveasefficientlyaspossible,underverysimplebehaviouralassumptions,goalsthataregivenfromelsewhere?Theethics-relatedapproachinvolvesabroaderunderstandingofeconomicsandotherbusinessdisciplines.Italsocompre-hendsproblemsofhumanmotivationandthejudgementofsocialachievementsthat539\nGEORGESENDERLEcannotbedisconnectedfromtheethicalquestionsofthegoodandthejust:howshouldoneliveandwhatisajustsociety?Ifbusinessdisciplinesembracethesequestions,theintegrationofthecognitiveandnormativedimensionsbecomespossiblewithinthem.TheworkofAmartyaSen,the1998NobelLaureateinEconomics,showsthatsuchanapproachcanbeconsistentlydeveloped(see,particularly,Sen1999b).Hisscholarshipinbothethicsandeconomicsisoutstanding,buthealsoexploreswithgreatsophisticationtheinterfacesbetweenthemandthebuildingofbridgesthathelpthedifferentperspectivesinteract.Anotherprominentscholar,ArthurRich,haspioneeredbusinessandeconomicethicsinGerman-speakingcountries.TheguidingprincipleofhisWirtschaftsethikis:‘Thatwhichisnoteconomicallyrationalcannotreallybehumanlyjust,andthatwhichconflictswithhumanjusticecannotreallybeeconomi-callyrational’(Rich1984:81).3ChartingtheEmergingFieldofBusinessEthicsTopresentthestateofbusinessethicstodayisadifficultundertaking.Comparedtothesituationin1990,hostsofpublicationsinmanylanguages,fromscholarlyworkstocorporatestatements,haveappearedandcannotbefullysurveyed.Nevertheless,manyrecentsurveysandencyclopediasareavailable,andourextendedthree-levelconcep-tionmayhelptoorganizeourpresentation.ThemostextensiveoverviewingeographictermscanbefoundinthespecialissueoftheJournalofBusinessEthics,October1997,whichincludes13region-andcountry-relatedreportsonbusinessethicswithextendedlistsofliterature.Thereportsshowthatbusinessethicsisanemerginganddynamicfield,dependingoneconomicfactors,politicalchangesandanawarenessofvalue-conflictsandethicalandenvironmentaldemands.Becauseitisanemergingfield,itismuchmoredifficultthaninanestab-lishedfieldtocapturethelastingfeaturesofbusinessethicsandtoforeseetheirlikelydevelopment.Eachcountryandeachregionhasitsownbusinessethicshistory.Some,liketheUSandCanada,areold-timers(if25yearscanbecalledold),andsome,likeChinaandSouthAfrica,are‘newcomers’.Ofcourse,suchanobservationimpliesacertainunderstandingofbusinessethicsanddoesnotmeanthatethicalissuesinbusinessandtheeconomyhadnotarisenbefore.Fromtherichfindingsofthereports,onlyafewstrikingfeaturescanbehighlightedhere.Afirstcharacteristicistherelevanceofsemantics.Becauseethicalissuesinbusi-nessarenotmerelyrationalpuzzles,butaredeeplyrootedinemotionsandcultures,thewaysofspeakingaboutethicsandtheuseandmeaningofethicaltermsarestronglyaffectedbyemotionalandculturalfactors.Forinstance,theJapanesetermsforeconomy(keizai)andbusiness(keiei)arenotvalue-free,butalreadycontainanorma-tiveethicalcomponent,namelygoverningtheworldinharmony(kei)andmakingceaselesseffortstoachieve[thesepurposes](ei).Asaconsequence,Japanesebusinessethicscanhaveaverybroadscopeandincludesdifferentmoralagentsatvariouslevelsofeconomicactivity.Inanarrowersense,preferredbymanybusinessethicistsinJapan,thetermrelatestocorporateethics.Asecondfeatureconcernstheenormousvarietyofbusinessenvironmentsthatareshapedbythehistoricalandsocietalconditionsofcountriesandregionsand,inmany540\nBUSINESSETHICScases,byrecentdramaticchanges.Thetransitionfromcommunistregimestodemoc-raciesandmarketeconomiesinCentralandEasternEuropeandRussia,thenearlytwentyyearsofeconomicreforminChina,theabolitionofapartheidruleandthebuild-ingofdemocracywithoutdiscriminationinSouthAfrica,aredevelopmentswhichhavefar-reachingimplicationsforthespecificroleofbusinessethicsintheseenvironments.Ifweignorethesesystemicdifferencesatthemacro-level,thechallengeofbusinessethicsforcorporationsatthemeso-levelcannotbeunderstood.Amongthelargenumberofspecificbusinessethicsissuesmentionedinthereports,corruption,leadershipandcorporateresponsibilityattractparticularattention.Theglobalizationoftheeconomyhasbeenaccompaniedbyaglobalizationofcorruption.Itseliminationisconsideredthemostsignificantchallengeformanycountriesbecausecorruptionprofoundlyaffectsbusinessrelations(meso-level),individualdecision-makers(micro-level)andwholeeconomies(macro-level).Numerousreportsemphasizetheneedforanewgenerationofbusinessleaders.Theyareconsideredtobeacrucialinstrumentforimplementingorganizationalintegritythroughmanagement-drivenself-governanceprogrammesinwhichcompanyvaluesandaspirationsplayacriticalrole.Leadershavetoincorporatethisvision.Inspiteofbroadsystemicchallenges,thereportersfromthecountriesofMahatmaGandhiandNelsonMandelastressedtheimportanceofindividualethics,personalcharacterandintegrity.Notsurprisingly,thereportsdiscussawiderangeofissuesregardingcorporateethicsinthenationalaswellasinternationalcontext.Theyraiseissuesofcorporaterespon-sibilityconcerningtheconduct,cultureandstructureofbusinessorganizations(atthemeso-level),andtheseclearlydifferfromissuesatthesystemicandindividuallevels.OneofthemostcommonthemesinthebusinessethicsliteratureofNorthAmericaconcernsstakeholderobligationsandstakeholdertheoryandhasgainedwideintern-ationalinterest.Anotherexportofideas,fromtheCadburyReportintheUnitedKingdom,concernscorporategovernanceandhasinfluencedAustraliaandSouthAfricaindifferentways.Finally,deregulationandagrowingcivilsocietyhasin-tensifiedthecallforeffectivewaystoholdcompaniesaccountable,nationallyandinternationally.TopicareasofinternationalbusinessethicswereidentifiedanddiscussedattheFirstandSecondISBEEWorldCongressofBusiness,Economics,andEthics,1996inTokyo,and2000inSãoPaulo(seeEnderle1999andISBEEwebsiteatwww.isbee.org).Also,manyoverviewshaveappeared.BusinessEthics:AEuropeanReviewhaspublishedanumberofEuropeancountryreports.TheJournalofBusinessEthicshasdedicatedawholeissuetoSpain(1999).BusinessEthicsQuarterlyhascelebratedits10thanniversarywithacollectionofoverthirty,mostlyUScontributions(January2000).WerhaneandSinger(1999)presentcontributionsfromAsiaandNewZealand.Enderle(1996)comparesbusinessethicsinNorthAmericaandcontinentalEuropeandsurveysbusinessandcorporateethicsintheUSA(Enderle1999).BarkhuysenandRossouw(2000)reportonbusinessethicsasanacademicfieldinAfrica.And‘Trendsofbusi-nessethicsinLatinAmerica’werepresentedattheISBEECongressinSãoPaulo(Arruda2000).Whilesurveysandoverviewscanonlyintroduceonetothefield,encyclopedicworkscandigdeeperandreflectmorecomprehensiveandelaborateconceptionsofthefield.The1990shavebroughtseveralmajorworks:intheUSA,theEncyclopedicDictionary541\nGEORGESENDERLEofBusinessEthics(WerhaneandFreeman1997),ACompaniontoBusinessEthics(Frederick1999)andtheEncyclopediaofEthics(BeckerandBecker1992),withmanyentriesonbusinessandeconomicethics;inGerman-speakingcountries,theLexikonderWirtschaftsethik(EncyclopediaofBusinessEthics)(Enderleetal.1993)andthefour-volumeHandbuchderWirtschaftsethik(HandbookofBusinessEthics)(Korffetal.1999).AlthoughwecannotprovideathoroughcomparisonbetweentheUSandGermanapproachesincorporatedintheseworks,afewremarksmightbeofinterest(seealsoEnderle1996).FourdifferencesbetweentheUS(andCanadian)approachandtheGerman(andcontinentalEuropean)approachstandout.First,businessethicsintheUSAchieflydealswithissuesattheindividualmicro-level,whereasthemainemphasisinGerman-speakingcountries(andcontinentalEurope)isonthesystemicmacro-level.NorthAmericansaddressanincreasingnumberofissuesattheorganizationalmeso-level,butrarelydiscussissuesatthemacro-level.Incontrast,Germansareonlybeginningtobeinterestedinthemeso-leveloftheorganization.Secondly,theUSapproachemphasizesthefreedomandresponsibilitiesofdecision-makingandacting,withthetendencytooverlooktheirlimitations.TheGermanapproachunderlinestheimportanceofshapingbusinessconditionsinanethicallyresponsiblewayandtendsnottomakefulluseofexistingspacesoffreedom.Thirdly,theUSapproachtobusinessethicsismuchmorepracticalthantheGermanone,atleastwithregardtothemicro-levelandthemeso-level,whereascontinentalEuropeanstendtofocusontheoreticalissuesbeforeaddressing,ifever,practicalchallenges.Fourthly,thereisamultitudeofviewsonanappropriateparadigmofbusinessethics,andneitherNorthAmericanorcontinentalEuropehasawell-establisheddisci-plineofbusinessethics.WhileacademicsinGerman-speakingandScandinaviancoun-triesarelikelytobemoreconcernedaboutthecognitivedimensionofbusinessethicsthatincorporatesthecontributionsofbusinessdisciplines,NorthAmericanacademicsdealmuchmoredirectlyandopenlywithnormativeissues.4TheoreticalApproachestoBusinessEthicsThediscussionofvarioustheoreticalapproachestobusinessethicscanbeorganizedaccordingtotheextendedthree-levelconceptionandthecognitiveandnormativedimensionsofthetwo-leggedapproach.4.1Foundationalissues:productiveinterdisciplinarycommunicationContrarytoawidespreadEuropeanperception,foundationalissuesofbusinessandeconomicethicshavebeenextensivelydiscussedinEnglish-speakingcountriestoo,thoughindifferentformsandbeyondnarrowlydefinedbusinessethicscircles.Testi-monytothisdiscussionarecontributionstothereviewEconomicsandPhilosophy,largepartsofSen’swork,thesurvey‘TakingEthicsSeriously:EconomicsandContemporaryMoralPhilosophy’intheJournalofEconomicLiterature(HausmanandMcPherson1993)andsubstantiveintroductionslikeEconomicAnalysisandMoralPhilosophy542\nBUSINESSETHICSTable19.2DissimilaritiesofbusinessethicsinNorthAmericaandcontinentalEuropeNorthAmericaContinentalEurope1SemanticsOnelanguage(basically)MultiplelanguagesEasytalkaboutBEReluctanttalkaboutBE2Level(actor)-focusMICRO,mesoMACRO,micro3Outreach-dimensionLessinternationalMoreinternational4ChoiceparametersFreechoiceofactorsConstraintsofactors(choicewithinconstraints)(choiceofconstraints)5ArgumentsforBEArgumentsofscandalsandArgumentofeconomization‘goodbusiness’6Practice-focusStrongWeak7ParadigmofBEClearnormativeConnectionwithsocialorientationsciences8TeachingBEWideacceptanceinbusinessBeginningefforts;nogeneralschools;beginningintegrationacceptanceandnointocurriculumintegrationintocurriculum9CorporateinitiativesWidespreaddisseminationofFair,increasingdisseminationcodesofconduct;increasingofcodesofconduct;fewpositionsofethicsofficersethicstrainingprogrammes;andethicstrainingnoethicsofficersprogrammes10RoleofbusinessTeachingmainlyfutureTeachingstudentsinvariousethicistsmanagers;morepractice-disciplines;lesspractice-orientedresearch;increasingorientedresearch;consultingconsultingactivitieshasjustbegunSource:Enderle(1996).(HausmanandMcPherson1996).FollowingSen’snotionofanethics-relatedapproach,threebroadtopicareascanbedistinguished.Thefirsttopicarea,dealingwithhumanmotivation,concernsrationalityandmorality:rationalityandutilitytheory;rationalityinpositiveandnormativeeconomics;andrationality,normsandmorality.Thesecondtopicarea,dealingwithjudgementsaboutsocialarrangements,concernsevaluatingeconomicandsocialinstitutions,policiesandoutcomes:theinfor-mationalbasisofevaluation;welfare,efficiency,utilitarianismandconsequentialism;andliberty,rights,equalityandjustice.Athirdtopicarea,usinganengineeringapproach,dealswithmoralmathematics,particularlysocialchoicetheoryandgametheory.InGerman-speakingcountries,thediscussionoffoundationalissueshasalonghistorygoingbacktothe‘HistorischeSchule’inthe‘VereinfürSocialpolitik’[Histori-calSchoolofThoughtintheGermanEconomicAssociation]andtheoreticalstrug-glesintheearlytwentiethcentury.Inthemid-1980sthisprofessionalassociation543\nGEORGESENDERLEestablisheditsownEconomicsandEthicsDivision,whichhaspublishedaseriesofbookswithcontributionsonfoundationalproblems.Moreover,theHandbuchderWirtschaftsethik(Korffetal.1999)dedicatesitsentirefirstvolumetodeterminingtherelationshipbetweenbusiness/economicsandethics.Itcoverstopicsfromtheessentialcomponentsofmodernbusinessethics(suchaswork,propertyandtechnology)topolitical–structuralimplicationsofmoderneconomies.ItalsopresentsvariousapproachesofcontemporarybusinessethicistsinGerman-speakingcountries.Despitethevarietyofapproachesinbothscholarlygroups,theoverallsituationcanbesummarizedasfollows:‘Inthelasttwentyyears,economistsandmoralphilosophershaverenewedaconversationthatwasinterruptedduringtheheydayofpositivistmethodologyinbothdisciplines.Althoughthereremainconsiderablegapsbetweenbothmodesofexpressionandhabitsofthoughtofmoralphilosophersandeconomists...wehopeourpresentationhasalsoshownthatthereisconsiderableroomforpro-ductiveinterdisciplinarycommunication’(HausmanandMcPherson1993:723).4.2FeaturesoftheeconomicsystemAtthemacro-levelwecanpresentafewkeyfeaturesfromthefocusontheeconomicsystem.Thesefeaturesapplynotonlytoclosedeconomieswithoutinternationalrelations,butalsotoeconomieswithdifferenttypesofinner-systemicrelationsacrossnationalborders.Accordingtotheprevailingviewinmoderntheoryofeconomicsystems(Kromphardt1990),wecannotproperlycharacterizeaneconomicsystembyonesinglecriterion,suchascapitalasthepropertyregimeorthemarketastheallocationregime.Rather,threesetsofcriteriaarenecessary:1Ownershipanddecision-making:whoparticipatesintheprocessofeconomicdeci-sion-making?Whoplansandcontrolsproduction,distributionandconsumption,forexample,throughahighconcentrationofeconomicpowerorabroadpartici-patoryeconomy?2Informationandco-ordination:withthehelpofwhatinformationsystemaretheindividualdecisionsco-ordinated,forexample,bydecentralizedmarketsorcentralizedplanning?3Motivation:whatobjectivesmotivatethevariousdecision-makers;forexample,self-interest,thecommongoodorloyalty?Whichwaysarechosentoimplementeconomicdecisionsandwhattypeofbehaviourisexpected?Onlyifallthreecriteriaaretreatedinabalancedmanner,cantheeconomicsystembeadequatelyunderstoodandethicallyevaluated.Thestudyofeconomichistorycanfurtherclarifythenotionoftheeconomicsystem.AsNorthstates,twodimensionsmakeuptheeconomyfromthehistoricalperspective:‘(1)theoverallgrowthoftheeconomyovertimeandthedeterminantsofthatgrowth(orstagnationordecline)and(2)thedistributionofincomewithinthateconomyinthecourseofitsgrowthordecline’(North1972:468).Bothdimensions,theproduc-tiveandthedistributive,areequallyimportantandcloselyrelatedtoeachother.Thisholdstrueatthemacro-leveland,withappropriatemodifications,alsoatthemeso-levelandmicro-level.Countrieswhichachievedthe‘EastAsianMiracle’(WorldBank1993)544\nBUSINESSETHICSpaidmuchattentiontotheinterplaybetweenproductiveanddistributiveaspects,andmanydifficultiesofeconomicdevelopmentinLatinAmericaandtheCaribbeancanbeexplainedbytheneglectofthedistributivedimension,thatis,byignoringpovertyandinequalityasimpedimentstogrowth(ECLAC1999).Thattheeconomyisnotonlyaproductivesystembutalsoincludesdistributionasanintegralparthasraisedfunda-mentalissuesinregionalandglobaleconomicaffairs;forexample,thehotlydebateddistributionalproblemsinthecontextoftheNorthAmericanFreeTradeArea,theEuropeanUnion,theWorldTradeOrganizationandglobalizationingeneral.Inaddition,thecleararticulationoftheeconomicsystemcanpreventtheshort-sightedequationoftheeconomywithmarketsandprivategoods.Thetheoryofpubliceconomics(AuerbachandFeldstein1987)goesfarbeyondmarketsandprivategoodsbydealingwithpublicgoods,taxation,incomemaintenance,socialinsuranceandotherissues.Intheinternationalcontexttoo,publicgoodsarebecomingacrucialissue(Kaul,GrunbergandStern1999;EnderleandKamm2000).Therefore,businessethics,ifrestrictedtomarketmorality,isdoomedtomissanessentialpartoftheeconomy.Afinalremarkconcernstheroleoftheeconomicsystem(macro-level)anditsrela-tionshipwiththeeconomicactorsattheorganizational(meso-level)andindividuallevels(micro-level).Ifoneassumesthattheeconomicsystemfullydeterminesallactorsandactionsintheeconomicdomain(asinthemechanisticparadigmandinLuhmann1988),thereisnoneedtopayattentiontothemeso-levelandmicro-level.However,ifweassumethatthereareextendedspacesoffreedomatalllevels,theactorshaverelativeautonomyandresponsibility.Thisdoesnotminimizetheimportanceofthemacro-level,becauseatthislevelthemaininstitutionsandgroundrulesoftheeconomyaredefinedandanchored.Withtheinternationalizationofbusiness,systemicissuesoftheempire,interconnectionandglobalizationtypesbecomemoreimportantandmustbealignedwithsimilarinternationaldevelopmentsatthemeso-levelandmicro-level.4.3DifferentapproachestocorporateethicsBusinessorganizationshavebecomepowerfulenginesofeconomicandsocialchangeandhaveextendedtheirstructuresandactivitiesinternationallyinmultipleforms.Theimportanceofethicsinbusinessenterpriseshasincreasedaccordingly.Oneimportantapproachtocorporateethicsisthestakeholderapproach.SinceEdwardFreeman’sseminalworkin1984,thisapproachhasgatheredmuchmomen-tuminNorthAmericaandbeyond(DonaldsonandPreston1995;Näsi1995;Clarkson1998).Althoughitisaworldviewratherthanacoherentandelaboratedtheory,ithasfoundalargeconsensusovertherejectionofanarrowconceptionofcorporateresponsibilitygearedonlytotheshareholdersoftheenterprise(Friedman1970)byincludingotherstakeholdersaswell.Thusthestakeholders,definedasanygroupandindividualwhocanaffectorareaffectedbytheachievementofacorporation’spurpose,includecustomers,employees,suppliers,competitors,localcommunities,governmentsandothers.Becausebusinessesareinvolvedinrelationswithmultipleinternalandexternalstakeholders,theybearethicalaswellaseconomicresponsibilityforshapingtheserelations.Accordingtosupportersofthisview,thisimpliesthatthestakeholders’voicesshouldberecognizedashavingsomeintrinsicvaluebeyondtheir545\nGEORGESENDERLEinstrumentalvalueforthecorporation.Still,manyquestionsremainopen:whatspecificresponsibilitieshavecorporationstowardsvariousstakeholders?Howcantheseresponsibilitiesbejustifiedandhowshouldtheybebalanced?(EnderleandTavis1998).Moreover,theincreasingimportanceofbusinessenterpriseshasgivenrisetothedevelopmentandapplicationofvarioustheoriesofcorporateethics.TheyincludeaKANTIAN(chapter32)perspective(Bowie1999),UTILITARIANISM(chapter35)(SnoeyenbosandHumber1999),VIRTUEETHICS(pp.206–8)(Solomon1992),aSOCIALCONTRACT(pp.672–7)approach(Donaldsonetal.1999),discourseethics(Ulrich1993,1997),dialogueethics(SteinmannandLöhr1992)andothers.OverviewsofmanyoftheseapproachesareprovidedbyFrederick(1999).Thedevelopmentofcorporateethicsisarelativelyrecentundertakinginthehistoryofethicsandwillrequirefurtherefforttoachieveawell-establishedandbalancedview.Sometheories(likevirtueethics)concernmoretheindividualsinthecorporationthanthecorporationitself,ordrawnocleardistinctionbetweenthemicro-levelandmeso-level.Othertheories(liketheKantianperspectiveandsocialcontracttheory)arefairlywelldevelopedatthemeso-level,buthavelittlefoundationinthesystemiclevel.Othertheories(likediscourseethics),withstrongfoundations,paytoolittleattentiontoorganizationalissues.Inaddition,onlyafewtheoriesexplicitlyaddresstheinternationaldimension(DeGeorge1993;Donaldsonetal.1999;Bowie1999).5ConcludingRemarksThisintroductiontotheongoingdevelopmentofbusinessethicsinvariouspartsoftheworldshowsthatbusinessethicsfacesenormouspracticalandtheoreticalchallenges.Thefurtherthatglobalizationadvances,thegreaterandmorecomplexthesechallengesbecome,butstrugglingandcomingtogripswiththemisexcitingaswellasurgentanddemanding.Developmentsinrecentdecadesinmanyplacesjustifyreasonablehopeforthefutureofthefieldratherthanscepticismorcynicism.However,differentapproachestobusi-nessethicsoftenstandaloneandbarelylearnfromeachother.Corporateethicsandanappropriateethics-relatedconceptoftheenterpriseintheglobalcontextneedmuchmoreelaboration.Systemicissuesshouldbeaddressedandintegratedintothebusinessethicsapproaches.Theimportanceofindividualdecision-makingandactionshouldbetakenseriously.Andtheunderstandingofthemultipleformsofinternationalinvolve-mentofeconomicactorsshouldbecomemoresophisticated.Businessethics,too,needsglobalnetworkingandglobalization.Theeffortwillyieldaricherunderstandingthatisclosertothecomplexityofmodernbusinessandbetterabletoprovideethicalguidanceforbusinesspractice.FurtherReadingInthischapterseveralencyclopedicworksarementionedwhichprovideeasyaccesstoalargevarietyoftopicsinbusinessandeconomicethics,alongwithrichbibliographicalinformation(Enderleetal.1993;WerhaneandFreeman1997;Frederick1999;Korffetal.1999).Onecanfindvaluablemonographsandtopic-relatedcollectionsofarticlesinanumberofseries:the546\nBUSINESSETHICSRuffinSeriesinBusinessEthicsbyOxfordUniversityPress(since1989),IssuesinBusinessEthicsbyKluwerAcademicPublishers(since1990),StudiesinEconomics,EthicsandPhilosophybySpringer(since1992);andinGerman,seriesinSchriftendesVereinsfürSocialpolitikbyDunckerandHumblot(since1985)andSt.GallerBeiträgezurWirtschaftsethikbyHaupt(since1986).Leadingjournalsofbusinessethicsare:BusinessandProfessionalEthicsJournal(since1981),JournalofBusinessEthics(since1982,withacontentanalysisof1,523issuesintheissuedated1July2000),BusinessEthicsQuarterly(since1991)andBusinessEthics:AEuropeanReview(since1992).Inaddition,journalsofmanagementandsocialscienceshavebegunpublishingarticlesonbusinessandeconomicethics.Abibliographyondisketteincludesallbusinessethicsarticlespublishedfrom1992toJuly2001insixmajorjournals(EnderleandKamm2001).Thetextbookliterature,withaprimaryfocusoncorporateethics,hasdevelopedconsiderably,particularlyintheUSA.SeeBoatright(1997),BuchholzandRosenthal(1998),DeGeorge(1999),Donaldsonetal.(1999),HoffmanandFrederick(1995),Shaw(1996),Weiss(1994)andVelasquez(1998).TextbooksinothercountriesareSorellandHendry(1994)(UK);Rossouw(1994)(SouthAfrica);VanLuijk(1986)(Netherlands);Kreikebaum(1996)andSteinmannandLöhr(1992)(Germany);Staffelbach(1994)(Switzerland).DifferentculturalandinterculturalperspectivesareaddressedbyKumarandSteinmann(1998),Moorthyetal.(1998),GasparskyandRyan(1996),Cavanagh(1998),Frederick(1995),Pesqueux(2000),Becker(1996),YuXuanmengetal.(1997)andChakraborty(2001).CodesofconductarediscussedfromabehaviouralresearchperspectivebyMessicketal.(1996)andfromaglobalperspectivebyWilliams(2000).Agreatdealofliterature,especiallyfromNorthAmericaandvaryinggreatlyinquality,dealswiththeethicsofandformanagersandleaders.OutstandingbooksareJackall(1988),Levi(1986)andCiulla(1998).ReferencesAGlobalEthic:TheDeclarationoftheParliamentofWorld’sReligions1993:London.Arruda,M.C.(ed.)2000:TheEthicalChallengesofGlobalization:ProceedingsLatinAmerica.SãoPaulo:EAESP/FGV(seewww.isbee.org).Auerbach,A.J.andFeldstein,M.(eds)1987:HandbookofPublicEconomics,vol.2.Amsterdam:North-Holland.Barkhuysen,B.andRossouw,G.J.2000:BusinessEthicsasAcademicFieldinAfrica:ASurveyofItsCurrentStatus.PaperpresentedattheSecondWorldCongressofBusiness,Economics,andEthics.Website:www.nd.edu/~isbee.Becker,G.K.1996:EthicsinBusinessandSociety:ChineseandWesternPerspectives.Berlin:Springer.Becker,L.C.(ed.)andBecker,C.B.(associateed.)1992:EncyclopediaofEthics.NewYork:GarlandPublishing.Bellah,R.N.,Madsen,R.,Sullivan,W.M.,Swidler,A.andTipton,S.M.1992:TheGoodSociety.NewYork:RandomHouse.Berenbeim,R.E.1987:CorporateEthics.NewYork:TheConferenceBoard.Boatright,J.R.1997:EthicsandtheConductofBusiness.UpperSaddleRiver,NJ:Prentice-Hall.Bowie,N.E.1999:BusinessEthics:AKantianPerspective.Oxford:Blackwell.Buchholz,R.A.andRosenthal,S.B.1998:BusinessEthics:ThePragmaticPathBeyondPrinciplestoProcess.UpperSaddleRiver,NJ:Prentice-Hall.BusinessEthicsQuarterly.2000:Januaryissue.Cavanagh,G.F.1998:AmericanBusinessValueswithInternationalPerspectives,4thedn.UpperSaddleRiver,NJ:Prentice-Hall.Chakraborty,S.K.2001:TheManagementandEthicsOmnibus.ManagementbyValues.EthicsinManagement.ValuesandEthicsforOrganizations.NewDelhi:OxfordUniversityPress.547\nGEORGESENDERLECiulla,J.B.(ed.)1998:Ethics:TheHeartofLeadership.Westport,CT:QuorumBooks.Clarkson,B.E.(ed.)1998:TheCorporationandItsStakeholders:ClassicandContemporaryReadings.Toronto:UniversityofTorontoPress.Coleman,J.S.1990:FoundationsofSocialTheory.Cambridge,MA:BelknapPress.DeGeorge,R.T.1993:CompetingwithIntegrityinInternationalBusiness.NewYork:OxfordUniversityPress.——1999:BusinessEthics,5thedn.EnglewoodCliffs,NJ:Prentice-Hall.Donaldson,T.andDunfee,T.W.1999:TiesThatBind:ASocialContractsApproachtoBusinessEthics.Boston:HarvardBusinessSchoolPress.Donaldson,T.andPreston,L.E.1995:TheStakeholderTheoryoftheCorporation:Concepts,Evidence,andImplications.AcademyofManagementReview,20,1,65–91.Donaldson,T.andWerhane,P.H.(eds)1999:EthicalIssuesinBusiness:APhilosophicalApproach,6thedn.UpperSaddleRiver,NJ:Prentice-Hall.Enderle,G.1996:AComparisonofBusinessEthicsinNorthAmericaandContinentalEurope.BusinessEthics:AEuropeanReview,33–46.——(ed.)1999:InternationalBusinessEthics:ChallengesandApproaches.NotreDame,IN:UniversityofNotreDamePress.Enderle,G.andKamm,D.1998:BusinessandCorporateEthicsintheUSA:PhilosophyandPractice.InB.KumarandH.Steinmann(eds)EthicsinInternationalManagement.Berlin:WalterdeGruyter.——2000:WhoseEthosforPublicGoodsinaGlobalEconomy?AnExplorationinInternationalBusinessEthics.BusinessEthicsQuarterly,131–44.FirstpublishedinGermanas‘WelchesEthosfüröffentlicheGüterinderWeltwirtschaft?’InH.KüngandK.-J.Kuschel(Hg.)1998:WeltethosundWissenschaften.Munich:Piper.——2001:BibliographyofBusinessEthicsArticles(ondiskette).Articlespublishedfrom1992toJuly2001inBusinessEthics:AEuropeanReview;BusinessEthicsQuarterly;BusinessandSociety;BusinessandSocietyReview;JournalofBusinessEthics;TeachingBusinessEthics.NotreDame,IN:ISBEESecretariat(seewww.isbee.org).Enderle,G.andTavis,L.A.1998:ABalancedConceptoftheFirmandtheMeasurementofitsLong-termPlanningandPerformance.JournalofBusinessEthics,17,1121–44.Enderle,G.,Homann,K.,Honecker,M.,Kerber,W.andSteinmann,H.(eds)1993:LexikonderWirtschaftsethik.Freiburg:Herder.Fischer,J.M.(ed.)1986:MoralResponsibility.Ithaca,NY:CornellUniversityPress.——(ed.)1999:RecentWorkonMoralResponsibility.Ethics,110,1,93–139.Frederick,R.E.(ed.)1999:ACompaniontoBusinessEthics.Oxford:Blackwell.Frederick,W.C.1995:Values,Nature,andCultureintheAmericanCorporation.NewYork:OxfordUniversityPress.Freeman,R.E.1984:StrategicManagement:AStakeholderApproach.Boston:Pitman.French,P.A.1984:CollectiveandCorporateResponsibility.NewYork:ColumbiaUniversityPress.Friedman,M.1970:TheSocialResponsibilityofBusinessIstoIncreaseItsProfits.NewYorkTimesMagazine,13September.Gasparski,W.W.andRyan,L.V.(eds)1996:HumanActioninBusiness:PraxiologicalandEthicalDimensions.NewBrunswick,NJ:TransactionPublishers.Gewirth,A.1984:TheEpistemologyofHumanRights.SocialPhilosophyandPolicy,1,2,1–33.Glover,J.1970:Responsibility.London:RoutledgeandKeganPaul.Goodpaster,K.E.1992:BusinessEthics.InL.C.Becker(ed.)andC.B.Becker(associateed.)EncyclopediaofEthics.NewYork:GarlandPublishing.Hausman,D.M.andMcPherson,M.S.1993:TakingEthicsSeriously:EconomicsandContem-poraryMoralPhilosophy.JournalofEconomicLiterature,31,671–731.548\nBUSINESSETHICS——1996:EconomicAnalysisandMoralPhilosophy.Cambridge:CambridgeUniversityPress.Held,D.,McGrew,A.,Goldblatt,D.andPerraton,J.1999:GlobalTransformations:Politics,EconomicsandCulture.Stanford,CA:StanfordUniversityPress.Hoffman,W.M.andFrederick,R.E.1995:BusinessEthics:ReadingsandCasesinCorporateMorality.NewYork:McGraw-Hill.Homann,K.andBlome-Drees,F.1992:Wirtschafts-undUnternehmensethik.Göttingen:UTB.Jackall,R.1988:MoralMazes:TheWorldofCorporateManagers.NewYork:OxfordUniversityPress.Jonas,H.1984:TheImperativeofResponsibility:FoundationsofanEthicsfortheTechnologicalAge.(OriginalinGerman:PrinzipVerantwortung.)Chicago:UniversityofChicagoPress.JournalofBusinessEthics1997:Country-andRegion-RelatedReports.October.JournalofBusinessEthics1999:BusinessEthicsinSpain.11November.Kagan,S.1998:NormativeEthics.Boulder,CO:WestviewPress.Kaul,I.,Grunberg,I.andStern,M.A.(eds)1999:GlobalPublicGoods:InternationalCooperationinthe21stCentury.PublishedfortheUnitedNationsDevelopmentProgramme(UNDP).NewYork:OxfordUniversityPress.Korff,W.,Baumgartner,A.,Franz,H.,Genosko,J.,Homann,K.,Kirchner,C.,Kluxen,W.,Küpper,H.-U.,Picot,A.,Rendtorff,T.,Richter,R.,Sautter,H.andSchlecht,O.(Hg.)1999:HandbuchderWirtschaftsethik,4vols.Gütersloh:GütersloherVerlagshaus.Kreikebaum,H.1996:GrundlagenderUnternehmensethik.Stuttgart:Schäfer-Poeschel.Kromphardt,J.1990:KonzeptionenundAnalysendesKapitalismus,vol.3:Auflage.Göttingen:VandenhoeckandRuprecht.Kumar,B.N.andSteinmann,H.(eds)1998:EthicsinInternationalManagement.Berlin:Springer.Küng,H.1998:AGlobalEthicforGlobalPoliticsandEconomics.NewYork:OxfordUniversityPress.Levi,I.1986:HardChoices:DecisionMakingunderUnresolvedConflict.Cambridge:CambridgeUniversityPress.Luhmann,N.1988:DieWirtschaftderGesellschaft.FrankfurtamMain:Suhrkamp.Messick,D.M.andTenbrunsel,A.E.(eds)1996:CodesofConduct:BehavioralResearchIntoBusinessEthics.NewYork:RussellSageFoundation.Moorthy,R.S.,DeGeorge,R.T.,Donaldson,T.,Ellos,W.J.,Solomon,R.C.andTextor,R.B.1998:UncompromisingIntegrity:Motorola’sChallenge.Shaumberg,IL:MotorolaUniversityPress.Näsi,J.(ed.)1995:UnderstandingStakeholderThinking.Helsinki:LSR-Publications.North,D.C.1972:EconomicHistory.InD.L.Sills(ed.)InternationalEncyclopediaofSocialSciences.NewYork:MacmillanandFreePress.Pesqueux,Y.2000:LeGouvernementdel’entreprisecommeidéologie.Paris:EllipsesÉdition.Rawls,J.1999:TheLawofPeopleswith‘TheIdeaofPublicReasonRevisited’.Cambridge,MA:HarvardUniversityPress.Rich,A.1984/1990:Wirtschaftsethik,vols1–2.Gütersloh:Mohn.TranslatedasTheEthicsofEconomicSystems,forthcoming.Rossouw,D.1994:BusinessEthics:ASouthernAfricanPerspective.Pretoria:SouthernBookPublishers.Sen,A.1987:OnEthicsandEconomics.Oxford:Blackwell.——1999a:DevelopmentasFreedom.NewYork:AlfredA.Knopf.——1999b:GlobalJustice:BeyondInternationalEquity.InI.Kaul,I.GrunbergandM.A.Stern(eds)GlobalPublicGoods:InternationalCooperationinthe21stCentury.PublishedfortheUnitedNationsDevelopmentProgramme(UNDP).NewYork:OxfordUniversityPress.Shaw,W.H.1996:BusinessEthics.Belmont,CA:Wadsworth.Snoeyenbos,M.andHumber,J.1999:UtilitarianismandBusinessEthics.InR.E.Frederick(ed.)ACompaniontoBusinessEthics.Oxford:Blackwell.549\nGEORGESENDERLESocialPhilosophyandPolicy1999:Responsibility.Summerissue.Solomon,R.C.1992:EthicsandExcellence:CooperationandIntegrityinBusiness.NewYork:OxfordUniversityPress.Solomon,R.T.1993:BusinessEthics.InP.Singer(ed.)ACompaniontoEthics.Oxford:Blackwell.Sorell,T.andHendry,J.1994:BusinessEthics.Oxford:Butterworth-Heinemann.Staffelbach,B.1994:Management-Ethik.AnsätzeundKonzepteausbetriebswirtschaftlicherSicht.Bern:Haupt.Steinmann,H.andLöhr,A.1992:GrundlagenderUnternehmensethik.Stuttgart:Poeschel.Ulrich,P.1993:TransformationderökonomischenVernunft.FortschrittsperspektivendermodernenIndustriegesellschaft.3:Auflage.Bern:Haupt.——1997:IntegrativeWirtschaftsethik.Grundlageneinerlebensdienlichenökonomie.Bern:Haupt.UncompromisingIntegrity:Motorola’sGlobalChallenge.Schaumburg,IL:MotorolaUniversityPress.UnitedNationsDevelopmentProgramme(UNDP)1990:HumanDevelopmentReport1990.NewYork:OxfordUniversityPress.UnitedNations,EconomicCommissionforLatinAmericaandtheCaribbean(ECLAC)1999:LatinAmericaandtheCarribeanintheWorldEconomy.Santiago,Chile:UnitedNations.VanLuijk,H.1986:Inhetbelangvandeonderneming;aantekeningenvooreenbedrijfsethik,2ndedn.Delft:Eburon.Velasquez,M.G.1998:BusinessEthics:ConceptsandCases,4thedn.UpperSaddleRiver,NJ:Prentice-Hall.Walton,C.C.1988:TheMoralManager.NewYork:HarperBusiness.Weiss,J.W.1994:BusinessEthics:AManagerial,StakeholderApproach.Belmont,CA:Wadsworth.Werhane,P.H.andFreeman,R.E.(eds)1997:EncyclopedicDictionaryofBusinessEthics.Oxford:Blackwell.Werhane,P.H.andSinger,A.E.(eds)1999:BusinessEthicsinTheoryandPractice:ContributionsfromAsiaandNewZealand.Dordrecht:KluwerAcademicPublishers.Williams,O.F.(ed.)2000:GlobalCodesofConduct:AnIdeaWhoseTimeHasCome.NotreDame,IN:UniversityofNotreDamePress.Williamson,O.E.1985:TheEconomicInstitutionsofCapitalism.NewYork:FreePress.——(ed.)1990:OrganizationTheory:FromChesterBarnardtothePresentandBeyond.NewYork:OxfordUniversityPress.WorldBank1993:TheEastAsianMiracle:EconomicGrowthandPublicPolicy.NewYork:OxfordUniversityPress.YuXuanmeng,LuXiaohe,LiuFangtong,ZhangRulunandEnderle,G.1997:EconomicEthicsandChineseCulture.ChinesePhilosophicalStudies,14.Washington,DC:CouncilforResearchinValuesandPhilosophy.DiscussionQuestions1Howhasglobalizationshapedtheproblemsaddressedbybusinessethics?2Arethereadvantagestohavingacomparativeconceptualframeworkwithwhichtodiscussbusinessethics?3Whatistherelationshipbetweentheoreticalandempiricalstudiesinbusinessandeconomicethics?4Howdodifferentconceptionsofbusinessaffectournotionofbusinessethics?5Howarethesemantic,theoreticalandpracticalaspectsofbusinessethicsrelated?6Shouldbusinessethicsbe‘action-oriented’?Howcanwerelatefreedom,choice,constraintandresponsibilityinbusinessethics?550\nBUSINESSETHICS7Whataretheconsequencesforbusinessethicsofthe‘concreteness’ofactionanddecision-makinginbusiness?8Howshouldweunderstandtherelationsamongproblemsofethicsthatariseatdifferentlevelsofbusinessandeconomiclife,fromtheindividualactortothewholeeconomicsystem?Shouldonelevelhavepriorityovertheothers?9Howshouldatheoryofbusinessethicsrespondtotheinternationalizationofeco-nomicrelationsacrossnationalandculturalborders?10Canthecognitiveandnormativeaspectsofbusinessethicsbeseparated?Whataretheconsequencesofyouranswerforbusinessandeconomicdisciplines?11Whatisthemoralstatusoforganizations?12Howshouldbusinessethicsrespondtothediversityofhistoricalandsocialcon-texts,businesspracticesandproblemsthatariseindifferentcountries?13Whatshouldbethecharacterofcorporateethics?Howcanwedecidewhethercorporationsshouldberesponsibletostakeholdersotherthanshareholders?14TowhatextentshouldweseektoreconcilethedifferentapproachestobusinessethicsinNorthAmericaandcontinentalEurope?15Whatnotionsofrationalityshouldbeusedinbusinessethicsandotherbusinessdisciplines?16Canbusinessethicshelpustoevaluatesocialandeconomicinstitutions?17Whatistheroleof‘moralmathematics’inbusinessethics?18Whatcriteriashouldbeusedtocharacterizeaneconomicsystem?19Whatislikelytobethecharacterofbusinessethicsasitbecomesamaturediscipline?551\n20PhilosophyandFeminismJEANGRIMSHAWANDMIRANDAFRICKERRecentdecadeshaveseenintensedebatesamongfeminists,especiallyinFrance,theUnitedKingdomandtheUnitedStates,abouttheexclusionofwomenfromphilosophy,therepresentationofwomeninphilosophicalworks,varyingreasonsthesubjectitselfmightbeconsidered‘masculine’,anddifferentproposalsforafeministphilosophy.Dis-cussionshaverangedfromareassuchasethicsandpoliticalphilosophy,toareasinwhichquestionsofgenderariselessspontaneously,likeepistemology,metaphysicsandphi-losophyofscience.Thischapterexaminessomemainfeaturesofcontemporaryfeministphilosophyandconsidersproposalsforthefuture.JeanGrimshawconsidersfeminismandphilosophy(sections1–4),andMirandaFrickerdiscussesfeminisminphilosophy(sections5–6).1FeminismandPhilosophy:IntroductionForwomenwhohavebeeninanywaysensitizedtotheissuesconcerninggenderandthesituationofwomen,whichhavebeencentraltocontemporaryfeminism,tobeawomanstudyingphilosophyistobefacedwithasetofproblemsaboutone’sownloca-tionwithinthediscipline.Aboveallotheracademicdisciplines,perhaps,philosophy,initshistoricalaspect,appearsasahistoryof‘GreatMenofIdeas’,andintherollcallof‘great’philosophersfromthepast,womenseemtobesingularlyabsent.Womenphilosopherswhosenamesareknowntoushavetended,asMichèleLeDoeuff(1991)pointsout,toberestrictedtotheroleofdiscipleof,orcorrespondentwith,the‘master’;asinthecaseofDESCARTES(chapter26)andPrincessElizabethofBohemia(1618–80).Theformalbarrierstotheindependentstudyofphilosophythatexistedinthepastforwomenhavenowlargelygone.Quitealotofwomenstudyphilosophy;manycometothinkofitastheir‘intellectualhome’.Yetrelativelyfewofthem‘makeit’intoaca-demicposts,andthoseforwhomphilosophycontinuestobecentraltotheirlivesmayalsofeelthattherearewaysinwhichphilosophycanbealienandinhospitabletowomen.Beingawomanandaphilosopher,asLeDoeuffnotes,isalwayssomethingthatisproblematic,thathastobenegotiated.Inthemajorityofmainstreamcoursesinphilosophy,issuesconcerningphilosophyandgendercontinuetobelargelyinvisible.Thiswasoftenthecaseforthosewomen\nPHILOSOPHYANDFEMINISMphilosopherswhoarenowtryingtotheorizetherelationshipbetweenfeminismandphilosophy.Forthem,theinitialrecognitionthatmalephilosophershadwrittenagreatdealaboutwomen,andthatmuchofwhattheyhadwrittenwasriddledwithsexismandmisogyny,cameasashock.Howwasitpossiblethatonecouldhavestudiedphi-losophyforalongtime,yetfailedtonoticethis?How,forexample,couldoneberequiredtostudyRousseau(1712–78)withoutone’sattentionbeingdrawnatanypointtowhathewroteaboutSophieandthesubordinationofwomen?HowcouldonepossiblyhavestudiedNIETZSCHE(1844–1900)(chapter40)withoutnotingthecrucial,ifproblem-atic,rolethattheconceptof‘woman’playsinhisthought?Alongwiththeshocktherefrequentlycameasenseofdislocation.Thisarosefromtherecognitionthatwhereasthestudyofphilosophyhadseemedtobesomethingtowhichone’sgenderwasirrel-evant,muchofphilosophyinfactassumedandwasaddressedtoamalesubject.Thefirst(andessential)stageinthinkingaboutphilosophyandgenderconsistedinhighlightingtheendemicmisogynyofmanymalephilosophers;innotingthewaysinwhichphilosophycommonlyassumedtheexclusionofwomen(ascitizens,asrationalsubjects,orasMORALAGENTS(chapter6),forexample),andinchartingthewaysinwhichwomenweredevaluedorseenasinferior.Thishistoryofsexismandmisogynyhaslargelybeeninvisibletotheacademicphilosophicalmainstream.Ifnoted,ithasbeendismissedasunimportant,akindoflocal‘accident’whoserecognitionneednototherwiseaffectphilosophicaldiscussionandtheorizing.Andindeed,themereexpo-sureofsexism,howeverimportant,is‘negative’inthesensethat,byitself,itdoesnottheorizethewaysinwhichsuchsexismmightaffectthesubstanceofphilosophicaltheories,whatrelationshipfeminismandphilosophymightcontinuetohave,orhowwomenmightrelocatethemselveswithinphilosophy.Onestrategyadoptedbythosewhowereconcernedaboutthisexclusionofwomenhasbeentoattempttoincludewomeninphilosophicaltheories.Aparadigmofsuchattemptsatinclusioncanbefoundinthewritingsoftheeighteenth-centuryfeminist,MaryWollstonecraft(1759–97).InAVindicationoftheRightsofWoman(1975)Wollstonecraftinsistedthatthepossibilityofrealizingidealsofrationalityandvirtue,ofwhichitwascommonlythoughtthatonlymenwerecapable,shouldbeextendedtoincludewomen.ShenotedthatRousseauregardedwomenasincapableofreason.Notso,sheargued;totheextentthatwomenwereunabletoreasonadequately,itwasbecausetheyhadbeenpreventedfromsodoingbyapernicioussystemoffemaleedu-cation.Rousseauarguedthatthe‘virtue’ofamanwasentirelydifferentfromthe‘virtue’ofawoman.Wollstonecraftrepliedthatvirtueshouldbethesameregardlessofsex.Buthoweverimportantithasbeentoinsistthatwomenarenot‘inferior’tomen,andarenolesscapableofreasonorvirtue,thestrategyofsimplyincludingwomeninphilosophicaltheoriessoonbreaksdown.Firstly,itisfrequentlythecasethatonecannotsimply‘include’womenwhilstleavingtherestofatheoryintact.Rousseauisacaseinpoint.Histheoryofthestateandtheeducationofthecitizenwaswhollypremisedonaprivate,patriarchalandruralfamilylife,inwhichwomenweretobecompletelysubordinatedtomen.Secondly,theattemptedinclusionofwomenfailstoaskquestionsaboutthatfromwhichtheyareexcluded;isthereasense,forexample,inwhichsomeidealsofreasonandvirtuemightthemselvesbeseenas‘masculine’?Muchfeministphilosophicalworkhasbeendevotedtoaskingwhethertherearedeeper553\nJEANGRIMSHAWANDMIRANDAFRICKERsensesinwhichthecontentofphilosophicaltheoriesmightbeseenas‘masculine’.Canphilosophysometimesatleastbeseenas‘gendered’evenincaseswheresuchthingsasthenatureofwomenandmenisnotapparentlyunderdiscussion?Doestheenterpriseofphilosophyitselfhavesomekindof‘deepstructure’thatmightalsobeseeninthisway?Theanalysisofthiskindof‘masculinity’reachesfardeeperthanmerelydetect-ingoranalysingovertinstancesofmisogynyorsexism,andithasbeenapproachedinmorethanonesortofway.2PhilosophyandMasculinityThefirstessentialpremise,onwhichtheanalysisofmasculinityinphilosophyispredi-cated,isasfollows.Perhapsaboveallotherdisciplines,philosophyhashistoricallylikedtothinkofitselfasbasedonreason;butareasonwhichisassumedtobeuniversalandobjective.Thephilosophicalsubjecthasoftenbeenassumedtobecapableofa‘God’s-eyeview’,a‘viewfromnowhere’;a‘rationalmind’thathasnoclass,gender,race,socialandhistoricallocation,perhapseven,insomephilosophicaltheories,nobody.Discus-sionofthesociallocationofthosewhopractisephilosophyhasbeenseenincontem-poraryterms(oftendismissively)asaformofsociologyofknowledgewhichhasnorelevancetoquestionsaboutthetruthoradequacyofphilosophicaltheoriesthem-selves.Butfromafeministperspective,questionsaboutthesociallocationofknowers,orthosewhoclaimtoknow,arenotextrinsictophilosophy.Rather,theyshouldbeatitsveryheart,sincethegenderingofphilosophy,anditsfrequentexclusionofwomen,arebasedonthedenialofthislocation,ontheassumptionofa‘universality’whichisinfactaconcealedpartiality.The‘human’subjectinphilosophyhasoftenturnedouttobeamalesubject.Theprocessofattemptingtotheorizethekindsofsociallocationthatareofrel-evancetothestudyofphilosophyhas,however,producedsomedifferingviewsofthewaysinwhichphilosophymightbeseenas‘masculine’.InherbookSpeculumoftheOtherWoman(1985),theFrenchfeministphilosopherLuceIrigaray,whoisalsoaprac-tisingpsychoanalyst,offeredareadingofthemythofthecaverninPlato’sRepublic.Shearguedthatthemythrestedonaprofounddenialandrepressionofthemother,thewomb,andthefactofmaledependenceonwomen,andsheinterpretedthisdenialbyofferingareinterpretationoftheFreudiantheoryofthedeathdrives.Apsycho-analyticapproachhasbeenusedbyotherphilosophersaswell.Inanimportantarticlepublishedin1983,JaneFlaxarguedthatphilosophicaltheoriesaroseoutoftypicallymalesocialexperience,interests,fearsandconcerns.Drawingontheworkofthefemi-nistpsychoanalyticwriterNancyChodorow(1978),whousedanapproachderivednotfromorthodoxFreudianismbutfromobject-relationstheory,shearguedthattherearetypicaldifferencesin‘genderpersonality’betweenwomenandmen,arisingfromthefactthatitiswomenwhoperformthemainworkofchild-rearing.Men,Flaxsuggested,typicallyfeelaneedtodistancethemselvesfromthe‘worldofwomen’,todevalueit,toassociatewiththeworldofmasculinity‘outside’thehome,andtostressadefensiveautonomyandseparatenessagainstthefearofbeingengulfedbywomen.FlaxnotedfeaturesinthephilosophyofPLATO(chapter23),HOBBES(chapter28),DescartesandRousseauwhich,sheargued,couldbeinterpretedinthelightofthesetypicalmascu-554\nPHILOSOPHYANDFEMINISMlinepsychicstructures.Suchfeaturesincludedindividualism,fearofthebodyandsenses,adenialofhumaninterdependenceandinterconnectedness,andastressontheradicalisolationofthehumanknowerandthehumanself.Thereare,however,considerableproblemsinthiskindofanalysisof‘masculinity’inthewritingsofphilosophers.Malephilosophershavesaidverydifferentthings;theyhavecertainlynotallespousedthekindsofatomisticorindividualistictheoriesthatFlaxseesasparadigmaticallymasculine.Inaddition,therearegreatproblemsintryingtogivesuchageneralorahistoricalaccountofmalesocialexperienceorofthemalepsyche.Thesekindsofproblemsmightseemtosuggestthattryingtolookforanyverygeneralfeaturesofphilosophythatmightbeseenas‘masculine’issomethingofahope-lessquest,giventheimportanceofhistoricalvariabilityanddifference.Eventhoughmanymalefiguresinthecanonhaveexpressedderogatoryviewsaboutwomeninonewayoranother,thesewayshavevariedgreatly.Itiscommonlythecasethatwhateverisvaluedinaphilosophicaltheorycanbefoundtobeassociatedwithanidealofmasculinity;nevertheless,theseidealsvaryhistoricallyagreatdeal.However,despitethisdifference,oneisfacedwiththefactthatmisogynyandtheexclusionofwomenseemtobeendemicinphilosophy,eveniftheymayattimestakeverydifferentforms.Isitpossibletogiveanysortofgeneralaccountofthefeaturesofphilosophythatseemtounderpinthisconstantanddrearyhistoryofthedevaluationofwomen,whichdoesnotrunquitethisriskofhistoricalover-generalization?Otherapproachesthatwomenworkinginphilosophyhaveadoptedinattemptingtoanswerthisquestiondependnotsomuchonlookingatthemoreparticularorspecificfeaturesofindividualtheories,oronpsychoanalyticspeculationaboutthemalepsyche,butontryingtotracesomeofthestructuresofrepresentationandideologythroughwhichphallocentrismandtheexclusionanddevaluationofwomengetwrittenintophilosophy.3Dichotomies:DerridaandFeminismFeministwritershavenotedtheconstanttendencyinphilosophicaltheoriestowardsformsofbinarism,tothinkingintermsofdichotomies.Thesearelegioninthehis-toryofphilosophy;theyincludeman–woman,culture–nature,reason–emotion,mind–body,public–privateandproduction–reproduction.Theydonotalwaystakethesameforms,andnoneofthemisuniversal,butthereareneverthelessimportanthis-toricalcontinuities.Withineachdichotomyorbinarismthereisahierarchy,aprivi-legedterm,andthisprivilegedtermisassociatedwithmenratherthanwithwomen.Therecognitionoftheimportanceofsuchbinarismshasnotbeenuniquetotheworkoffeministphilosophers.Ithascharacterizedmuch‘postmodern’philosophyaswell.JacquesDERRIDA(chapter42),forexample,arguesthatmostofWesternmeta-physicshasbeenconstructedonasystemofdifferencesorbinarismsinwhichonetermisprivilegedattheexpenseoftheother.Ithasbeen‘phallogocentric’,basednotonlyontheassumedlogocentricauthorityandprimacyofspeechand‘presence’overwriting,butalsoontheassumedphallocentricprimacyofthetranscendentalmalesubject.Derrida’sprojectofdeconstructionaimstodestabilize;toprivilegethesub-ordinateterminsuchmetaphysicalbinarisms,andtoshowhowtheprivilegedtermis555\nJEANGRIMSHAWANDMIRANDAFRICKERinfactdependentonthesubordinateterm.Derridaaims,bycomplexstrategiesoftextualanalysisandargument,topreventtheoriginalbinarismfromreassertingitself.Derridaisoneoftherelativelyfewcontemporaryphilosophersinwhoseworktheconceptof‘woman’playsacentralrole.‘Woman’,inDerrida’swork,becomesafigureforthatwhichisundecidable,thatwhichcanevadetheillusoryfixityoftraditionalmetaphysicalstructuresofmeaning.ButfeministreactionstoDerrida’sworkhighlightsomeofthemostinterestingtensionsanddifficultiesintheprojectoffeministphi-losophy.Derridahasbeennotoriouslycriticaloffeminism(andrarelyinhisownworkreferstofeministwritings).Thereasonseemstobeasfollows.Theprojectof‘decon-struction’aimstosubvertandunderminedualismsanddichotomiesandtocreateaspaceformultiplicityanddifference,includingsexualdifference.But,accordingtomostinterpretationsofDerrida,hearguesthatifthefeministprojectincludesanykindofassertionoffemalespecificity–ofwomen’srighttothesubjectivitythatmenhavebeenabletoassumeastheirs–thenitsimplyrepeatstheolddichotomies.Women,infact,cometo‘speaklikemen’.Derrida’sowndesireto‘speaklikeawoman’,then,seemstoamounttoadesiretoperformavertiginousandelusivetextualdanceatthemarginsorintheintersticesoftraditionalphilosophicaltexts.Itseemsverydifferentfromthefeministprojectoffindinga‘voice’forwomeninphilosophy,ofallowingwomentospeakintheirownright.SomecriticsofDerridahavearguedthathisvalorizationof‘speakinglikeawoman’effectsyetanothererasureofwomen.To‘speaklikeawoman’isnotthesameas‘speakingasawoman’.IrigarayandWomen’sVoiceinPhilosophyOfcoursetherearedangers,evidentinsomefeministwriting,ofsimplyrecapitulatingoldandoppressivedichotomies.Simplytoproposethat‘thefeminine’shouldbevalued,ifitinvolvesnocritiqueofthewaysinwhich‘thefeminine’isconstructedastheshadoworinversionofthemasculine,mayreinstatethatfromwhichmuchfeministtheoryhastriedtoescape.Theproblemistoseehowwomencanspeakwithinphilosophy,assumethestatusofphilosophicalSUBJECTS(pp.846–7),withoutthisreificationofoldercat-egories.LuceIrigaraywritesthatitistheobjectofherworkto‘jamthetheoreticalmachinery’;agreatdealofrecentfeministphilosophicalwritinghasexploredwaysinwhicholdphilosophicaldichotomiesmightbedeconstructedsothattheyarenolongerseenaspolarized,asexclusiveofeachother;rather,theyareredefinedasmutuallyinter-dependent.Nevertheless,Irigarayisalsoconcernedthatwomenshouldfindavoicewithinphilosophy.Inherwork,itisnotaquestionoffindingan‘essence’ofwoman,oroftheorizingaboutwomen’s‘nature’.Shedoesnotassumethatwecouldknowinadvancewhatwomenmightsay,weretheytospeak‘intheirownvoice’.4FeminismandPhilosophyButhowmightwesupposethatphilosophywouldchangeiftheinsightsoffeministphilosophicalwritingweretakenseriously?Atthepresenttime,whilstthereisagreatdealoffeministphilosophicalwritingbeingpublished,ithasseeminglymadelittleimpactinmanyplacesonthetraditionalacademicteachingofphilosophy,oron‘main-556\nPHILOSOPHYANDFEMINISMstream’publication.‘Feminism’or‘feminismandphilosophy’areoftenatbestan‘option’onthemargins,a‘specialinterest’,inwhichitisusuallyimplicitlyassumedthatonlywomenwilltakearealinterest.Thisisreflectedinsuchthingsasthewaysinwhichbooksareclassifiedinlibrariesandinbookshops.Philosophicalbooksbywomenarefrequentlynotincludedintheshelveslabelled‘philosophy’;theyareplacedunder‘gender’or‘women’sstudies’.Thelong-termquestionabouttherelationshipbetweenfeminismandphilosophyconcernswhatmightbecalled‘thenatureofthehyphen’,the‘and’whichappearsinthetitleofthischapteritself.AsIsuggestedearlier,thecrucialpremiseoffeministphilosophicalwritingisthatquestionsaboutthesocialandhistoricallocationofphilosophersareintrinsictounderstandingthediscipline,andtheycannotbedismissedasextraneousorirrelevant.Thereisaneedtorecognizethatphilosophicaltheoriesarisefromandspeaktotheconcernsofparticularsocialgroupsatparticularperiodsinhistory.Thisviewshouldnotbeequatedwithcrudekindsofreductionism–thekind,forexample,whichassumesthatdiscussionofphilosophicaltheoriescanbereducedtodiscussionofthepsychesofthephilosopherswhoholdthem.Butitdoesimplythatunderstandingphilosophicaltheorieswillnecessitateunderstandingsomethingaboutthesocialinter-estsandconcernswhichgaverisetothem–andthepowerrelationsbywhichtheymayhavebeenshaped.Feministphilosophershavearguedthatthisisastrueofthelessapparently‘gen-dered’areasofphilosophy,suchasEPISTEMOLOGY(chapter1),METAPHYSICS(chapter2)andthePHILOSOPHYOFSCIENCE(chapter9),asitisofthosewhereissuesofgenderhaveamoreobviousplace,suchasETHICS(chapter6)orPOLITICALPHILOSOPHY(chapter8,especiallypp.272–3).Philosophicalenquiryshouldrejectthe‘falseuniversalism’ofsupposingthatphilosopherscanspeakfromnowhere,orconstructtheoriesthatareuniversallyapplicable.Therecognitionofthishastheconsequencethattheparametersofphilosophicalenquirywillinevitablychange.Boundariesbetweenwhatisandisnot‘philosophy’willshift;oneconsequenceofdevelopingafeministperspectiveonphilosophyisthatonebecomesfarlessworriedbythechallenge:‘Butisthisphilosophy?’–achallengethathasoftenbeenusedtotrytodismissfeministquestioningattheoutset.Butthecritiqueof‘falseuniversalism’alsoimpliesthatfeministenquiryitselfshouldrecognizeitsownlocatedness,andtakeintoaccountthefactthatthedifferencesofsociallocation,experienceandperspectiveamongwomenthemselvesmaybeasimportantasthosebetweenwomenandmen.Butifthemostimportantclaimofafeministapproachtophilosophyisthatphilo-sophicaltheoriesaresociallylocated,andhavetendedtoanswertotheinterests,con-cernsandexperiencesofmenratherthanofwomen,thisstillleavesopentheimportantquestionofwhatthefutureofphilosophymightbelikewerefeministperspectivesonthemselvestobecomepartofthemainstream,ratherthana‘specialinterest’onthesidelines.WomenphilosopherssuchasIrigarayhavebeenconcernedthatwomenshouldbeable,inphilosophy,toassumethepositionofsubjectsandspeakintheirownvoice.Butthequestionremainswhethertheinsistenceonthisvoiceshouldbeseenasastrategicnecessitybornofcircumstancesinwhichwomenhaveoftenbeenexcludedandsilenced,orwhethertheconcernsofwomenandmenarelikelytoremainirre-duciblydifferentinwaysthatwillimplyacontinuingneedforphilosophywrittenfor,andmainlyby,womenthemselves.557\nJEANGRIMSHAWANDMIRANDAFRICKERApproachestothisquestionvary.Thosewhobelievethatwomen’sexperiencesandconcernswillalwaysdivergeinsomerespectsfromthoseofmenaremorelikelytocon-tinuetoinsistontheneedfor‘women’sphilosophy’.Ontheotherhand,therearethosesuchasMichèleLeDoeuffwhobelievethattheprojectoffeministphilosophy,inthelongterm,shouldbeanewkindofuniversalism,differentfromtheolderkindofuni-versalismwhichassumedthepossibilityofa‘God’s-eyeview’.LeDoeuffarguesthatwhereasitiscommonlyassumedthatphilosophyitselfis‘universal’andfeministenquirymerelypartial–aformof‘specialinterest’–itwouldbemoreappropriatetosaythatphilosophyhasbeen‘partial’inthesensethatithascommonlyexcludedwomenbothfromthepossibilityoftheproductionofphilosophicaltheoriesandfrombeingcapableofexemplifyingtheidealstowhichphilosophicaltheorieshavebeencom-mitted.LeDoeuffseesitasessentialtotheenterpriseofphilosophythatitposits‘objects’ofthoughtwhichareassumedtobeofinteresttoanaudiencenotdifferentiatedbyextra-intellectualcriteriasuchasgender.Inotherwords‘feministphilosophy’shouldnotbeseen,intheend,asa‘specialinterest’.Itissomethingthataimstochangeandinformthemainstreamofphilosophy,suchthatquestionsthatnowtendtobemar-ginalorinvisiblewillbecomenormal,visibleandpartoftherepertoireofallofthosewhostudyphilosophy.Butiftheproductionofphilosophyarisesfromsocialinterestsandconcerns,soalsodoesitsreception,andLeDoeuff’sconceptionofanaudienceundifferentiatedbyanyextra-intellectualcriterionmayseemtobeaformofidealismwhichistooclosetoolderphilosophicalidealsof‘universality’.Nevertheless,ithighlightswhatisperhapsthemostinterestingquestionaboutthefutureoffeministphilosophicalenquiry.Thereisanongoingneedtobringthemanydifferingperspectivesofwomenintophilosophy,andtocontinuetoaddressquestionsabouttheinvisibilityandmarginalizationofthese.Butfeministenquiryalsoaddressesissuesofgeneralhumaninterestandrelevance;itisnotanorthodoxycommittedtoanarroworblinkeredsetofpreconceptionsaboutwhatshouldbeseenas‘women’sissues’.Thetensionbetweentheprojectofcreatingframeworkswithinwhichwomencanspeakandthinkaswomen,andtheprojectofchangingthefaceofphilosophysothatwomen’sperspectivesarenolonger‘marked’asmarginaltothemainstream,isbothanessentialoneincurrentcircumstances,andonewhoseresolutionisperhapsthemajortaskfacingfeministphilosophyatthemoment.5FeminisminPhilosophy:TwoConceptionsOnoneconceptionoffeminism’srelationtophilosophyitisappropriatetothinkintermsofaconjunction–feminismandphilosophy–inwhichfeministconsciousnessisbroughttobearfromtheoutsideuponthephilosophicaltradition.Onthisconception,acentralquestionarisingisthequestionhowfarphilosophyhasbeendistinctivelymas-culine(hence‘masculist’)initsapproach,andwhethertheremightbesuchathingasadistinctivelyfemininephilosophicalvoiceasyetbarelyheardinthehistoryofthesubject(hencethepossibilityofacertain‘feminist’approach).Thekeycommitmentarisingfromexplorationsofthisquestiontendstotaketheformofaninsistenceonhis-toricisminourunderstandingofphilosophersandtheirwork.ThisisborneoutinJean558\nPHILOSOPHYANDFEMINISMGrimshaw’saccountinsections1–4offeminismandphilosophy,inwhichsheoutlinesfeministideasaboutboththeliterallyandthesymbolicallymasculinecharacterofphi-losophy.Inconnectionwiththefirst,feministshaveconcernedthemselveswiththeexpressionofprejudicialordownrightmisogynistattitudestowardswomenbymalefiguresinthecanon;andinconnectionwiththesecond,theyhavetriedtorevealsomeofthesubtlerwaysinwhichphilosophymightbeconsideredhistoricallytohaveexpressedamasculinepointofview.Thesubtlerwaysinwhichamasculinepointofviewisarguedtobeimplicitinagivenphilosophicaltextmightberevealedthroughapsychoanalyticanalysisofthepsychologyexpressedthere,orthroughamoresocio-logicalanalysisdesignedtorevealthatthesocialexperiencewhichinformsthewritingisamasculineone.Grimshaw’saccountcloseswiththeaspirationthatfeministworkshouldimpactonthemainstreamratherthanremaining‘marked’asmarginal.Inthespiritofthataspiration,Iwouldliketotakeupmyaccountwherehersleavesoff,byoutlininganon-conjunctiveconceptionoffeminism’srelationtophilosophysoastosuggestamodelforhowandwhyfeminismmightcometohaveitsdeservedimpactwithinthemainstream:feminisminphilosophy(alsoseetheintroductiontoFrickerandHornsby2000).Thisso-to-speakinternalconceptionisperfectlycompatiblewiththeconjunctiveconception–itsubstantiallyoverlapswithit,andindeedishistoricallydependentuponit.Todifferentiatetoosharplybetweenthetwo,bypretendingforinstancethatallfeministphilosophyfallsneatlyintooneorothercategory,wouldbegrosslyartificial.Withthiswarningbellsounded,however,theinternalviewseemsworthtreatingasdistinctivewhenoneisaimingtomapfeministworkintermsofitsambitionvis-à-visthemainstream.Theambitionmightconsist,broadlyspeaking,inoneorotheroftwoverydifferentaims:theaimtoachieveintegrationintomainstreamdebates,or,bycontrast,theaimtomaintainindependencefromthemainstreameitherbywayofasustainedmarginalityoramoreextremeintellectualseparatism.Inthisrespect,abooksuchasHipparchia’sChoice(LeDoeuff1991)mightrepresentarevisionistspirit,whileSpeculumoftheOtherWoman(Irigaray1985)mightberegardedasofferinga‘separatist’projectthatpositstheideaofanasyetunheardfemininephilosophicalvoice.Whereastheconjunctiveconceptionfosterseitherambitionwithequalencouragement–thisisitsstrength–theinternalconceptionwilltendmoreexclusivelytonurturethesortofworkwhichfacilitatesdialoguenotonlywithotherfeministsbutalsowithcolleaguestowhomtheideaoffeministphilosophymaystillbealienorevensuspect–this,inturn,isitsstrength.Hereadesirableby-productofdoinginternal-stylefeministworkmightbetoconvincesuchcolleaguesthatatleastsomeofthesocializingconcernsoffeminismmightproperlybetheconcernsofanyphilosopher(hencetheappropriate-nessoftheintegrationistambition).Consequently,thosewhostartfromtheinternalconceptionarefarlesslikelytoproduceworkwhichengagesinwholesaleCRITIQUE(p.278)ofphilosophyperse,andmorelikelytoproduceworkwhichgoesinforpiecemeallocalizedcriticalparticipa-tioninthephilosophicaldiscoursesoftheday.Bycertainradicalistlights,ofcourse,suchapiecemealapproachthreatenstheintegrityofthecriticalimpulseinfeminism.Thefearisthatengagementwiththemainstreamwillerodetheradicalpotentialofthefeministstanceinphilosophy.Buttheintegrationistambitioncanbeentertainedwithoutfearofanysuchthing,solongasoneisentitledtotrustahistoricistviewof559\nJEANGRIMSHAWANDMIRANDAFRICKERwhatconstitutesthediscipline:ifphilosophyisanevolvinghistoricalentity,thenfemi-nistconsciousnesscanhelp(re)constituteitbyincludingitselfamongthecriticalener-gieswhichinfluencethediscipline’songoinghistory.Philosophypurportstoaspiretoanunconstrainedself-criticism,andworkofthefeminisminphilosophykindaspirestoprovethepoint.Integrationisthusanactivecriticalambition,notadesireformereacceptance.Thusinsofarascriticalengagementbringstransitiontobothparties,itcanbefreelyadmittedthatgenuinedialoguebetweenagivensetoffeministideasandagivenpatchofmainstreamphilosophywilltransformthethinkingonbothsides;butitwouldbeundulypessimistictothinkthatthismustbeabadthingforthefeministideasatstake.Inanyevent,iftherewaseveracasefortheseparatistapproachtofemi-nistphilosophy,thentheencouragementofintegrationistfeministworkdoesnotchangethat.Thereisroomforallkindsoffeministphilosophy.Thepointhereisonlytocharacterizeanapproachwhichiswell-designedtofulfiltheundeniablyvalidaspirationthatatleastsomefeministworkshouldcometoloseitsmarkofmarginalitybywayofthemainstream’sgainingthemarksoffeministinsight.Likeitsconjunctivesister-conception,theinternalconceptionincorporatesacom-mitmenttohistoricism(understoodinclusivelytocoversocialspecificityofallkinds).Butwhiletheconjunctiveconceptiontendsinparticulartoproblematizethemas-culinityofthecanonicalphilosopherwithaviewtodiagnosingtheextentofthemasculinityofphilosophyperse,theinternalconceptiontendsmoreexclusivelytoproblematizetheseeminglycompulsoryasocialityofphilosophy’sideaofthebeingsittreats.Thusinworkofthefeminisminphilosophykindthereislikelytobeamoreexclu-siveuseofthesortofHISTORICISM(pp.437–9)whichhighlightsnotsomuchthesocialidentityofthephilosopher,butratherthesocialspecificityofthehumansubjectwhosepractices–conceptual,linguistic,epistemic,ethical,political–philosophypurportstoaccountfor.Thisisstrictlyadifferenceonlyinemphasisbetweenthetwoconceptions,fortohistoricizetheHUMANSUBJECT(pp.864–8)inthiswayisreallyjusttoinsistuponwhatfeministsofalmostallstripeshaveforyearscalled‘difference’.AsIhavesaid,thetwoconceptionsofhowfeminismrelatestophilosophyaresubstantiallyoverlapping,andtheconcernwithDIFFERENCE(pp.868–71)isnaturallyamajorareaofoverlap.Indeed,hereagainthegeneticdependenceoftheinternalupontheconjunctivecon-ceptionisapparent:feminism’sroutetoproblematizingtheasocialityofphilosophy’shumansubjectscomestousbywayofthosedebateswhichproblematizedthesupposedlyasocialpersonaadoptedbythephilosopherhimself(sic).6PhilosophicalCommitmentsSometimesitissupposedthatacommitmenttofeminismentailsacommitmenttoopposecertainfundamentalcommitmentsofphilosophy,itsmethodsandaspirations.Thepossibilityofdoingworkwhichflowsfromtheinternalconception,however,showsthatthisisnotso.Theinternalconceptionremainssternlyneutralaboutcertaintraditionalstylesofphilosophicalargumentinawaythathasnotbeentypicalofaconsiderableproportionoftheworkproducedbytheconjunctiveconception,asthatconception–beingthemoreinclusive–givessomesignificantnourishmenttotheappetiteforawholesalecritiqueofphilosophy.Perhapsoneofthedefinitivelytypical560\nPHILOSOPHYANDFEMINISMaspirationsofphilosophyistheaspirationtouniversaltruth,anditisoftenassumedthatthereissomethinginfeminismwhichimpliesacommitmenttotherejectionofthatuniversalistaspiration.Certainlythereisagreatdealoffeministworkwhichdoesrejectitandwhichcharacterizesitasapieceofmasculismparexcellence.Wecan,forexample,readRosiBraidotti’s(1991)feministreworkingoftheDeleuzianideaof‘nomadism’asafeministanti-universalism.Butinfactthereisnothinginfeminismwhichforceseitherthischaracterizationortherejection.Onthecontrary–andhappilyforfeminism’scredentialsasanon-dogmaticintellectualattitudeunderwhoseinfluencephilosophicaldebatemayflourish–thereisroombetweenfeministphilosophers,indeedbetweenphilosophersofallstripes,forargumentaboutthemeritsanddemeritsoftheaspiration.Whilesomefeministthinkersmaybepersuadedbyargumentsagainsttheveryideaofuniversaltruthsabouthumanpractices,andperhapsagainstthemethodsdesignedtoissueinthemtoo(transcendentalorothersortsofaprioriargumentation,forinstance),nonethelesstheinternalconception’semphasisonthesocialityofthehumansubjectneednotdiscreditsuchsourcesofuniversalism.Onthecontrary,itsveryarticulationcallsuponatleastoneinstanceofuniversalism,namelytheuniversalclaimthatallhumanbeingsaresociallyidentifiedindividuals.Andthepossibilitythattherearesomebasicuniversal,evennecessary,truthsabout,say,thenatureofsensoryexperienceisperfectlycompatiblewiththeirreduciblesocialityofthehumanbeingsthathavetheseexperiences.Thekeycorrectiveimpactoffeminisminphilosophyoverthispointisthatrelationsofpowerandidentityimpingeinmanymoreofthebasichumanpracticeswhichphilosophyaimstotheorizethanhastraditionally–complacently–beenassumed.Hereweoughttorecognizeonesourceofjustificationforamorewholesalesuspicionofphilosophythatmighttypicallybecountenancedonthisconception:ifhistoryhadwitnessedanexpressprojectofcreatingaphilosophicaldiscoursedesignedtomakequestionsofpowerapparentlyunaskable,becauseapparentlyunphilosophical,thenitcouldhardlyhavehopedtohavecomeupwithanythingbetterthantheAnglo-Americanparadigm.Thereisroom,then,fortheexpressionofwholesalesuspicionwithintheinternalconception.Butsuchsuspicionswillnotinspireseparatismthere,onlyagreaterwilltoparticipateinthetransformationofthetradition.Asecondfundamentalcommitmentofphilosophybesidestheaspirationtouniver-saltruthisthecommitmenttoconductingdebatesataveryhighlevelofabstraction.Itisinthenatureofphilosophytotradeinabstractions,andthatisasitshouldbeforanon-empiricaldiscipline.Sometimesitissupposedthatthereissomethinginfemi-nismthatimpliestherejectionofabstraction,asifabstractionmustbereplacedbysomethingelse–somethingwhichmightbegesturedatintheideaofsocialandcul-turalspecificity.(Thisistheformthatthecommitmentto‘difference’sometimestakes,asiftheorizingaboutpeopleassociallyspecificsubjectswerenecessarilyanenterpriseinabandoningabstraction.)Again,thereisinfactnothinginfeminismwhichrequirestherejectionofabstractionperse.Iftherewere,thentherewouldbenoprospectforfeministphilosophyremotelytoresemblephilosophyasweknowit,andsotherewouldbenofutureforfeministworkoftheinternalsort–butthentherewouldbelittleprospectformuchfeministtheorizingofanykind.Feministphilosophywouldhavetobetransformedintoanutterlyempiricalenterprise,sheddingitsapriorimethodologi-caloriginsaltogether.Thereareadmittedlypostmodernformsoffeministphilosophy561\nJEANGRIMSHAWANDMIRANDAFRICKERwhichmightembracesuchanoutcome,orsomeversionofit.Perhapsthevalueofintellectualpluralismmeansthisisasitshouldbe;thepresentpointissimplythatnothinginfeminismrequiresit.Insteadthefeministattitudemaybringaverydiffer-entproblematizationofphilosophy’sfavouritelevelofabstraction–thelevelatwhichhumanbeingsareconceivedas,simply,rationalbeings.Itmayquestiontheobligationtoconductdebatesatthislevel,bringingareleasefromthisseemingobligation,notbyrejectingabstractionassuch,butratherbyinsistingontheappropriateness,forsomeareasofphilosophyatleast,ofanabstractionwhichissotospeakonelevellower,sothathumanbeingsareconceivednotasrationalbeingssimplybutasrational–socialbeings.Thiswouldbeaconceptionthatrelatesnottoanysociallocationinparticular(thatwouldindeedinviteempiricalquestions),butratheronewhichpictureshumanbeingsasalwayslocatedsomewhereorotherinrelationsofSOCIALIDENTITYANDPOWER(pp.864–8).Suchasocializedconceptionofthehumansubjectliberatesthephilo-sophicalsensibilityinsofarasitushersinanewrangeofquestionswhich,thoughonemighthavebeencuriousaboutthem,hadseemedunaskableinthephilosophicalidiom:questions,forinstance,abouthowasubject’slocationinthesocialworld–theirplace,whateveritmaybe,inrelationsofidentityandpower–mightimpingeupontheirpar-ticipationinvariousfundamentalpractices.Thusitalsoallowsinquestionsabouthowfarthepracticesthemselvesmaydependuponthesesocialdimensionsfortheirverystructure–questions,then,abouthowfartheinvolvementofidentityandpowerisanecessaryfeatureofthepracticesunderscrutiny.Herewerevisitthequestionoftheappropriatenessofauniversalistaspirationinphilosophy,andwearriveattheheartofthataspiration:theideathatphilosophyisspeciallyplacedtorevealthenecessaryaspectsofhumanpractices.Ifaphilosopherdiscoversthatagivenpractice–anepistemicoranethicalpractice,forinstance–exhibitsanecessarydependenceonoperationsofpower,thenaneminentlyrespectablesortofphilosophicaltruthwillhavebeenrevealed:anecessarytruthaboutthatsortofpractice(Fricker1998).Thusthesocializedconceptionofthehumansubjectasarational–socialbeing(alwaysanabstraction)whichisencouragedbytheinternalconceptionoffeminism’srelationtophilosophy,isabletoproduceuniversalistconclu-sionsofakindtowhichthedisciplinehasquintessentiallyaspired.Here–aseverinthisenterpriseofdistinguishingdifferentapproachesinfeministphilosophy–Iriskoversimplification.Theoversimplificationthatnowlooms,however,concernsnotsomuchoneorotherfeministapproach,butratherthecharacterof‘the’philosophicaltradition–thequestionhowfar‘it’maybedescribedasespousingastrictlyaprioriquestforuniversaltruths.Inthecontinentaltraditionthereisanobvi-ouslyinfluential,evendefinitive,heritageofthoughtwhichspecificallyrejectsuniver-salism,orwhichotherwiselaunchesitsownwholesalecritiqueofphilosophy(seeDerrida1976).SimilarlyinfluentialanddisruptivecurrentsaretobefoundtoointheAnglo-Americantradition,thoughtheymightnotplausiblybecastasdefinitiveofit.TheWittgensteinianideaofphilosophyasatherapeuticenterpriseisanobviouscaseinpoint;amorerecentexampleisRichardRorty’sPhilosophyandtheMirrorofNature(1980).Inconnectiontoowiththeasocialconceptionofthehumansubject,itshouldbeacknowledgedthattherearesub-traditionswithinanalyticphilosophythatarein562\nPHILOSOPHYANDFEMINISMsomedegreesocialized:NATURALIZEDEPISTEMOLOGY(pp.302–3),forinstance;orPRAGMATISM(chapter36);orsocial-historicistcurrentsofthoughtinphilosophyofscience,andinethics.Inthesesortsofwork,thesocialityofhumanbeings–thefactthattheyalwaysoperateasmembersofsomecultureorother,orthinkanddeliberatefromthecontextofsometraditionorother–isverymuchinthepicture.However,whattheylackwhichfeminismintroducesisafullysocializedconceptionofthissociality:thedifferencethattherelationsofidentityandpowermaketohumansubjects’participationintherelevantpracticesofunderstanding,enquiryorevaluation.Thisisnottosaythattherearenopractices–thatofsensoryperception,forinstance–accountsofwhicharequiteproperlyconductedatthelevelofrationalbeings,simply.Torepeat,partoftheimpactoffeminisminphilosophyshouldbetoencouragenon-complacencyaboutwhich,ifany,thesepracticesare,sothatthedifferentlevelsofabstractionwhichphilosophymakesavailablecometobeemployedmorediscrimi-natingly,moreimaginativelyandmoreappropriately.Giventheasocializingnatureofthetradition,thismayjustifyaheuristicofgeneralsuspicionregardingclaimstosociality’sirrelevance;butitdoesnotrequireanyblanketpronouncementtotheeffectthatitisalwaysmistakenorotherwiseinappropriatetoconductdebatesatthehighestlevelofabstraction.Atitsmostpositive,feminisminphilosophyisaforward-looking,relentlesslyinterrogativeenergydirectedatopeningupnew,moresocialized,morepoliticizedquestions.Asidefromacertainqualifieddiscouragementtowholesalecritique,nothingphilosophicallysubstantiveisdictatedinadvance,andtheinternaliststancesignalsabaselineneutralityregardingphilosophy’straditionalcommitmentsandaspirations.Thisiswhatmakestheinternalconceptionausefuloneforthosewiththepositive,criticalintegrationistambitionforfeministworkwhichispointedtoinJeanGrimshaw’saccountoffeminismandphilosophy,andforwhichIhaveattemptedtosuggestapossiblemodelinthiscompanionaccountoffeminisminphilosophy.FurtherReadingAnthonyandWitt(1993)isanimportantcollectionofaccessibleessaysintheanalyticstyle.Braidotti(1991)isnotaneasybooktoread,butitisveryusefulinparticularforitsdiscussionofFrenchphilosophyaswellasAnglo-Americanfeministwriting,andforitsperspectivesontherelationshipbetweenfeminismandpostmodernism.Coole(1993)providesageneralsurveyofissuesconcerninggenderinpoliticalphilosophy.FrickerandHornsby(2000)isacompendiousrecentcollectionputtogetherspecificallywiththe‘internal’conceptionoffeministphilosophyinmind.Gatens(1991)isarecentandwide-rangingdiscussionofphilosophicaltextsfromafeministpointofview,andoffeministperspectivesinphilosophy.GriffithsandWhitford(1988)offersacollectionofessaysonfeminismandphilosophycoveringawiderangeoftopics.Grimshaw(1986)isanaccessibleintroductiontoquestionsconcerningwomenandphilosophy,withaparticularfocusonethics.Grosz(1990)surveysfeministapproachestophilosophy,andoffersaveryusefulaccountofwhatitmightmeantosaythatphilosophyis‘masculine’.HardingandHintikka(1983)presentaground-breakingcollectionofessayswhichlaidthefoundationsforagreatdealofsubsequentfeministenquiryintophilosophy.Haslanger(1995)isavaluablecollectionofessaysontopicsinphilosophyoflanguage,epistemologyandmetaphysics.Irigaray(1985)establishedthetermsofdebateformuchcurrentFrenchfeministthoughtandhasalso563\nJEANGRIMSHAWANDMIRANDAFRICKERbeeninfluentialintheUnitedKingdomandtheUSA.Kourany(1998)isahelpfulcollectionofessaysonarangeofsubjects.ThefurtherreadingcollectedinLarrabee(1993)providesexcellentbackgroundtotheissueofwhetherwomenspeakwitha‘differentvoice’inethicsandtheproblemssurroundingtheideaofan‘ethic’ofcare.LeDoeuff(1991)iswrittenintheformofaseriesof‘notebooks’,anditincludesanexcellentdiscussionofJean-PaulSartreandSimonedeBeauvoir,aswellasraisingbroadissuesaboutthelocationofwomeninphilosophyandtheobjectivesoffeministphilosophy.LennonandWhitford(1994)offerausefulcollectionfromafeministperspective,addressingbothtraditionalepistemologyandpostmodernviews.Lloyd(1993)isa‘classic’discussionofwaysinwhichconceptionsof‘reason’inWesternphilosophyhavebeenalignedwithnotionsof‘masculinity’.Nicholson(1990)isaninvaluablecollectionoffeministessays,aimingtoshowwhyfeminismhasbothanaffinitywithandyetalsoadeepambivalenceaboutmanyofthecentralthemesofpostmodernism.Nye(1988)surveysmisogynyinawiderangeofphilosophicaltheoriesandoffersananalysisofpossiblefeministresponses.PatemanandGrosz(1986)containsalivelyandinterestingcollectionofessaysbothongeneralissuesconcerningtheplaceofwomeninphilosophy,andonmorespecificsocialandpoliticalquestions.Scheman(1993)offersawide-ranging,accessibleandengagingcollectionofessaysonfeminismandphilosophy,providinganexcellentintroductiontosomecentralissues.Spelman(1988)discusseshowsomefeministthinkinghasexcludedmanywomeninwaysthatarenotunliketheexclusionofwomeninmuchofthephilosophythatiswrittenbymen.Tronto(1993)defendsan‘ethicofcare’,butarguesthatitshouldnotbecontrastedsharplywithanethicbasedonnotionsof‘rights’or‘justice’.Forfeministworksthatarecontinuouswithsocializedsub-traditionsinanalyticphilosophy,seeNelson(1990)(naturalizedepistemology);Seigfried(1996)(pragmatism);Harding(1991)(historicisminthephilosophyofscience);andGilligan(1982)(historicisminethics).ReferencesAlcoff,L.andPotter,E.1993:FeministEpistemologies.London:Routledge.Antony,L.andWitt,C.(eds)1993:AMindofOne’sOwn.Boulder,CO:WestviewPress.Benhabib,S.,Butler,J.,Cornell,D.andFraser,N.1995:FeministContentions:APhilosophicalExchange.London:Routledge.Braidotti,R.1991:PatternsofDissonance.London:Routledge.Chodorow,N.1978:TheReproductionofMothering.Berkeley:UniversityofCaliforniaPress.Code,L.1995:RhetoricalSpaces:EssaysonGenderedLocations.London:Routledge.Coole,D.1993:WomeninPoliticalTheory,2ndedn.Brighton:Wheatsheaf.Derrida,J.1976:OfGrammatology.Baltimore,MD:JohnsHopkinsUniversityPress.Dwyer,S.(ed.)1995:TheProblemofPornography.Belmont,CA:WadsworthPublishing.Flax,J.1983:PoliticalPhilosophyandthePatriarchalUnconscious.InS.HardingandM.Hintikka(eds)DiscoveringReality:FeministPerspectivesonEpistemology,Metaphysics,andthePhilosophyofScience.London:D.Reidel.Fricker,M.1998:RationalAuthorityandSocialPower:TowardsaTrulySocialEpistemology.ProceedingsoftheAristotelianSociety,vol.98,part2.Fricker,M.andHornsby,J.(eds)2000:TheCambridgeCompaniontoFeminisminPhilosophy.Cambridge:CambridgeUniversityPress.Garry,A.andPearsall,M.1989:Women,KnowledgeandReality:ExplorationsinFeministPhilosophy.London:UnwinHyman.Gatens,M.1991:FeminismandPhilosophy.Cambridge:PolityPress.Gilligan,C.1982:InaDifferentVoice:PsychologicalTheoryandWomen’sDevelopment.Cambridge,MA:HarvardUniversityPress.564\nPHILOSOPHYANDFEMINISMGriffiths,M.andWhitford,M.(eds)1988:FeministPerspectivesinPhilosophy.London:Macmillan.Grimshaw,J.1986:FeministPhilosophers.Brighton:Wheatsheaf.Grosz,E.1990:Philosophy.InS.Gunew(ed.)FeministKnowledge,CritiqueandConstruct.London:Routledge.Harding,S.1991:WhoseScience,WhoseKnowledge?Buckingham:OpenUniversityPress.Harding,S.andHintikka,M.(eds)1983:DiscoveringReality:FeministPerspectivesonEpistemol-ogy,Metaphysics,andthePhilosophyofScience.London:D.Reidel.Haslanger,S.(ed.)1995:PhilosophicalTopics:FeministPerspectivesonLanguage,Knowledge,andReality,vol.23,no.2,fall.Held,V.(ed.)1995:JusticeandCare:EssentialReadingsinFeministEthics.Boulder,CO:WestviewPress.Irigaray,L.1985:SpeculumoftheOtherWoman.Ithaca,NY:CornellUniversityPress.Jaggar,A.andYoung,I.M.(eds)1998:ACompaniontoFeministPhilosophy.Oxford:Blackwell.Kourany,J.(ed.)1998:PhilosophyinaFeministVoice:CritiquesandReconstructions.Princeton,NJ:PrincetonUniversityPress.Larrabee,M.J.(ed.)1993:AnEthicofCare.London:Routledge.LeDoeuff,M.1991:Hipparchia’sChoice:AnEssayConcerningWomen,Philosophy,etc.Oxford:Blackwell.Lennon,K.andWhitford,M.(eds)1994:KnowingtheDifference:FeministPerspectivesinEpis-temology.London:Routledge.Lloyd,G.1993:TheManofReason:MaleandFemaleinWesternPhilosophy,2ndedn.London:Methuen.Nelson,L.H.1990:WhoKnows:FromQuinetoaFeministEmpiricism.Philadelphia,PA:TempleUniversityPress.Nicholson,L.(ed.)1990:Feminism/Postmodernism.London:Routledge.Nye,A.1988:FeministTheoryandthePhilosophiesofMan.London:CroomHelm.Pateman,C.andGrosz,E.(eds)1986:FeministChallenges:SocialandPoliticalTheory.London:AllenandUnwin.Rorty,R.1980:PhilosophyandtheMirrorofNature.Oxford:Blackwell.Scheman,N.1993:Engenderings.London:Routledge.Seigfried,C.H.1996:PragmatismandFeminism:ReweavingtheSocialFabric.Chicago:UniversityofChicagoPress.Spelman,E.1988:InessentialWoman:ProblemsofExclusioninFeministThought.Boston:BeaconPress.Tronto,J.1993:MoralBoundaries.London:Routledge.Wollstonecraft,M.1975[1792]:AVindicationoftheRightsofWoman.Harmondsworth:PenguinBooks.DiscussionQuestions1Why,ifatall,shouldbeingawomanandaphilosopheralwaysbe‘somethingthatisproblematic,thathastobenegotiated’?2Shouldwomenseektobeacceptedascitizens,rationalsubjectsandmoralagentsjustlikemenorshouldtheyseektoalterourconceptionsofcitizenship,rationalityandagency?3Areclass,genderandethnicoriginofequalimportanceassocialandhistoricallocationsinshapingourformsofrationality?565\nJEANGRIMSHAWANDMIRANDAFRICKER4Doeshistoricalvariabilityunderminethepossibilityoftherebeinganygeneral‘masculine’featuresinphilosophy?5Doesthefeministproject,inassertingfemalespecificity,repeatratherthanunder-mineexistingdichotomies?6Whatistheimportanceofdistinguishing‘speakinglikeawoman’and‘speakingasawoman’?7Canwomenspeakwithinphilosophywithoutreifyingthecategoriestheyseektoescape?8Whymightfeministsseekto‘jamthetheoreticalmachinery’?9Howmightphilosophychangeiftheweightoffeministphilosophicalwritingsweretakenseriously?10Mustfeministphilosophersaccepttheclaimthatknowingthesocialandhistori-callocationofphilosophersiscrucialtounderstandingthediscipline?11Comparetheroleoffeministphilosophyinanareasuchasepistemology,meta-physicsorphilosophyofscience,withitsroleinanareawherequestionsofgenderhaveamoreobviousplace,suchasethicsorpoliticalphilosophy.12Isuniversalisminphilosophynecessarily‘falseuniversalism’?Whataretheimpli-cationsofyouranswerfortheroleoffeminisminphilosophy?13Aredifferencesamongwomenasimportantasdifferencesbetweenwomenandmenforfeministenquiry?14Arethereadvantagesindistinguishingbetween‘feminismandphilosophy’and‘feminisminphilosophy?15Doeslocalpiecemealparticipationwithinmainstreamphilosophythreatentheintegrityoffeminism’scriticalimpulse?16Shouldfeminismchallengephilosophy’sasocialaccountofwhatitstudies?Ifso,howcanthisbestbeachieved?17Towhatextent,ifatall,shouldfeminismseektoalterthemethodsandaspira-tionsofphilosophy?18Doesfeministcritiqueaddanythingvaluabletootherradicalcritiquesofphilosophy?19Doestheorizingaboutpeopleassociallyspecificsubjectsrequireabandoningabstraction?20Howmightfeminismhelpustoaskpreviouslyunaskablequestionsinphilosophy?566\n21Ethnicity,CultureandPhilosophyROBERTBERNASCONIThequestionoftheplaceofethnicitywithinphilosophycanbeunderstoodasthequestionofreconcilingtheconceptionofphilosophyasauniversalquestforuniversaltruthswiththefactthatitsvarioushistoricalinstantiationsarealwaysmarkedbytheirspatio-temporallocalization.Inotherwords,itisthequestionofwhetherphilosophicalpluralismattheculturallevelshouldbeconsideredavalue,ratherthanadefect,asitwasforHEGEL(chapter33).Ifitisgrantedthatthedifferentperspectivesthatdiffer-entculturesbringtobearenrichphilosophy,thisnotonlyreopensthedebateaboutthenatureofphilosophy,butalsoopensWesternphilosophytokindsofcriticalreflectionthathavesofarbeenforeigntoit.Thequestionofwhatisandwhatisnotphilosophyisnotsimplyaquestionofclassification.Thelabel‘philosophy’hashistoricallybeenanameforoneofthenoblestactivitiesofthehumanmind,sothattoacknowledgeaformofthoughtasaphiloso-phyistoaccorditastatus;itisawayofacknowledgingtheseriousnessofthatthought.Today,eventhoughphilosophyintheuniversitiesarguablytakesitselflessseriouslythanhitherto,itremainsthecasethattoexcludeaformofthoughtfromphilosophy,ortomarginalizeitwithinphilosophy,isusuallymeanttodisparageit.Nevertheless,toacknowledgeaformofthoughtasphilosophywhilekeepinginplacethecriteriawhichhithertowereusedtoexcludeitcanbeevenmorediminishing,becauseitisnolongertreatedonitsownterms.Theideaofethnicphilosophydoesnotimplythatphilosophymustrenouncetheuniversalismofhavingsomethingtosaytoeveryone.Theproblemarisesonlywhenaphilosophywantstospeakforeveryoneandintheprocesssilencesorignoresalternativevoices.Althoughitisnolongerascustomaryasitoncewastoreferdifferencesinphilo-sophicalstyleandcontenttotheculturesfromwhichanygivenphilosophyarises,thequestionoftherelationofphilosophyandculturehasresurfacedwithinphilosophyasabelatedresponsetomulticulturalism.Multiculturalismisanacademicmovementwithinthehumanitiesthat,bycelebratingculturaldiversity,challengestheculturalhegemonyclaimedbycertainethnicgroups.Itisalsoafactoflife,particularlyinmajorcitiesacrosstheworld.Multiculturalisminphilosophychallengestheestablishedcanonofphilosophybyinsistingontheneedforagreaterawarenessoftraditionsofphiloso-phybeyondso-calledWesternphilosophywithitsorigininGreece.Althoughmany\nROBERTBERNASCONIphilosopherstendtoassociatethereferenceofphilosophytoculturewiththespectreofrelativism,itisapparentthattherearedistinctcultures,thattheseareoftenreflectedinphilosophy,andthatcross-culturaldialoguedoestakeplace.Oneofthetasksofthephilosophyofcultureistoexaminetheconditionsunderlyingsuchadialogueandtheobstaclestoit,giventhatagreatdealofmisunderstandingandincomprehensionstillpersistsbetweenvariousphilosophicaltraditions.InthefirstsectionofthischapterIrecallthefactthatphilosophyintheWestonlyrecentlycametolocateitsoriginsinGreeceandtothatextentidentifyitselfasGREEK(chapters22and23).ManyWesternphilosophersdonotunderstandwhydiscussionsaboutmulticulturalismwithinphilosophyoftenquicklycometofocusontheallegedGreekoriginsofphilosophy,butonereasonisbecausetheGreektraditionofphiloso-phyisstilloftenusedasthestandardintermsofwhichdisputesaboutwhatisandwhatisnotphilosophyarejudged.InthesecondsectionIpasstoaconsiderationofAfricanphilosophy.Itisimpossibleinabriefessaytosurveythedifferentkindsofphilosophythathaveaclaimtobeingincludedinthecanon.IfocusonAfricanphilosophybecauseithasbeenpreoccupiedforalmostfiftyyearswiththequestionofitsstatusasphiloso-phy.ThatthereisanIndianphilosophyoraChinesephilosophytendstobetakenforgrantedinacademicdiscussionsacrosstheworld,althoughatthesametimetheyareusuallymarginalized.ByexaminingtheobstaclestoarecognitionofAfricanphiloso-phy,itispossibletohighlighthowthedominantconceptionofphilosophyresistschallengestoit.Inthethirdsection,Ibroadentheperspectivetotakeaccountofavarietyofdifferentmodelsofcross-culturalorinterculturaldialogue.Finally,inthefourthsection,Iconsideroneofthemostvibrantfruitsofinterculturaldialogueincontemporaryphilosophy:thecriticalre-examinationoftheWesternphilosophicalcanonfromviewpointsithaslargelyexcluded.Insteadofseeingthequestionofso-callednon-WesternphilosophyasaquestionoffindingapproximationstoWesternphilosophy,whereitremainstheunquestionedstandard,WesternphilosophersmightseeitasawayofrenewingWesternphilosophy.Itwouldbeabsurdtotrytoprovideinthisspaceasurveyofthedifferentkindsofphilosophythroughouttheworld,showingtheirrelationtothespecificculturefromwhichtheyarise.Instead,Ihaveprovidedanintroductiontosomeofthedebatesabouttherelationofphilosophyandculturethataretakingplacebothwithinandoutsidetheacademy.Inevitablythosedebates,insofarastheyaddressthequestionofwhetherthebalanceofwhatistaughtinuniversitiesacrosstheworldasphilosophyshouldbechangedtotakeaccountofabroaderrangeofcultures,takesitsstarting-pointinthestatusquo.Thatshouldnotbetakentomeanthattheuniversityhasbecometheprivilegedorsolesiteofphilosophy.Nevertheless,theinstitutionalrealitiesneedtobeacknowledged,evenasonerecognizesso-calledWesternphilosophyasoneformofethnicphilosophyamongothers.1HistoryoftheProblemintheContextoftheWesternPhilosophicalTraditionIntheseventeenthcentury,Europeanhistoriansofphilosophytendedtotreatthesubjectalongethniclines.JohannesGerardusVossius(1577–1649)beganhispost-568\nETHNICITY,CULTUREANDPHILOSOPHYhumouslypublishedDephilosophorumsectis(1657)withabriefsurveyoftheBarbarianphilosophiesofAsia(theChaldeans,theJews,thePersians,theIndians,thePhoe-niciansandthePhrygians),andofAfrica(theEgyptians,theEthiopiansandtheLibyans),beforepassingtothephilosophiesofEurope(theThracians,theDruidsand,finally,theGreeks).Acenturylater,JacobBruckercontinuedthesamepractice,althoughhislistwassomewhatlonger,andincludedtheancientArabs,theEtruscans,theancientRomansand,inanappendixonso-called‘exotic’philosophies,theBuddhists,theChinese,theJapaneseandtheCanadians(NativeAmericans).However,Brucker(1742:458)identifiedThalesasthefirstphilosopher,insofarasThalesandhisstudent,Anaximander,introducedtheSCIENTIFICMETHOD(chapter9).TheEgyptians,accordingtoBrucker,arrivedattheirknowledgebycustomandchance.BruckeracknowledgedthattheGreeksthemselvesplacedahighvalueonEgyptianlearning,but,onhisaccount,thiswasduetotheinfluenceofOrpheus,andsofarasBruckerwasconcernedOrpheusbroughttheGreeksonlyEgyptianmythology,notphilosophy.ThiswasinsharpcontrasttoMarsilioFicino,whoovertwohundredyearsearlier,temporarilyputasidethetaskoftranslatingPlotinustotranslatetheHermeticainstead(Yates1964:12–17).DuringtheItalianRenaissancethereturntoEgypttookprecedenceoverthereturntoGreece.Subsequently,therichersourcesavailableforthestudyofGreekthought,thefactthatEgyptianhieroglyphicshadnotbeendeciphered,andthedoubtplacedontheauthenticityoftheHermetica,ledtoEgyptbeinggivenanincreasinglylimitedroleinaccountsofthehistoryofphilosophy.Nevertheless,itwasnotuntiltheendoftheeighteenthcenturythatfull-lengthhis-toriesthatpurportedlycoveredallofphilosophywerewrittenreflectingunambiguouslythisbeliefinphilosophy’sGreekorigins.DietrichTiedemanninGeistderspekulativenPhilosophie(published1791)seemstohavebeenthefirsttowriteafull-lengthhistoryofphilosophythatbeginswithThaleswithoutanyacknowledgementofhisprecursorselsewhere.AtthesametimeEuropeanphilosophers,particularlybutbynomeansexclusivelytheGermans,regardedthemselvesasthelegitimateheirsoftheGreeks,therebyestablishingtheideaofaunifiedtradition.VariousexplanationscanbegivenforthisrestrictionofphilosophytotheWestatthispointinitshistory.Onecanpoint,firstly,totheemergenceofanarrowerconcep-tionofphilosophy.Brucker,whoperhapsdidasmuchasanyonetoestablishthenewcanon,wastheproponentofaneclecticismthatheidentifiedasoneofthedominantcharacteristicsofmodernphilosophy,butthiseclecticismwasnotopentoallphiloso-phies.KANT(chapter32)playedamajorroleinrestrictingtheideaofphilosophyfurther,therebyhighlightingthephilosophicalsectarianismofWesternphilosophyattheexpenseofnon-Westernphilosophy.Therewasalso,secondly,intheeighteenthcenturyasecularizationofphilosophythatledtoarenegotiationofthedistinctionbetweenmythandreason,sothatwhathadpreviouslybeenregardedasphilosophywasdisplacedintoRELIGION(chapter15).ThisisnowheremoreevidentthaninthecaseofChinesephilosophy,whichhadbeencriticizedasatheistinresponsetoChristianWolff’sattempttopromoteit,butwhichwassubsequentlyexcludedfromphilosophyasreligion.Thirdly,thehistoryofphiloso-phycametobewritteninconformitywiththenewphilosophyofhistorywithitsbeliefinprogress.Toshowprogressinhistory,historywaswrittenasanarrative.Whatdidnotfitintothecontinuousnarrativewasregardedasrelativelyinsignificantoran569\nROBERTBERNASCONIaberration.Fourthly,partlyforpoliticalmotives,theconvictiongrewthatDEMOCRACY(pp.277–81)wasaconditionforthepossibilityofaflourishingphilosophicaldebate.Thisargumentcanbefound,forexample,inbothAdamSmithandVoltaire.Fifthly,andperhapsmostimportantly,therewasthedevelopmentinEuropeofavirulentchauvinism,whichbytheendoftheeighteenthcenturywasreinforcedbythenewconceptionofrace.Theexistenceofhighlydevelopedoldercivilizations,suchasthoseofEgyptorIndia,contradictedthechronologyofcreationthatEuropeanshadcal-culatedonthebasisoftheBible.Similarly,theexistenceofhighlydevelopedphilo-sophicalsystemsinChinaandIndiathreatenedEurope’sself-conception,particularlyamongthosewhobelievedthatREVELATION(pp.622–4),notreason,wastheultimatesourceofwisdom.AlloftheseforceswereatworkinHegel,whodidasmuchasanyonetoestablishthephilosophicalimportanceofthehistoryofphilosophy.Hegelexplicitlyraisedthequestionoftherelationofethnicitytophilosophywhenheaskedwhyphilosophy,whichsupposedlywastheteachingofabsolutetruth,appearedrestrictedtoasmallnumberofindividualsandtoparticularpeoplesatparticulartimes(Hegel1985:12).Theformu-lationclearlyechoes,howeverunconsciously,theanxietyofChristianstryingtomakesenseofthefactthattherevelationofGod’slawhadapparentlybeenconfinedtoafew.Unlikesomeothercultures,whichmaintainthateachculturehasthereligionappro-priatetoitandsoisnotcommittedtomissionaryactivity,theWesthastendedtoseereligionintermsofastarkalternative:eitheroneisinthetruthoroutsideit.ItisnotnecessarytoexplorehereHegel’saccountoftherelationofphilosophytoreligioninallitscomplexity,butHegeltookthepositionthatonlyafewpeopleshadsucceededinraisingthemselvestothelevelofphilosophy.Theresthaddepositedtheirviewsoftheworldinreligion(ibid.:28).Hegelraisedtheconcernthatthepluralityofphilosophiesmadephilosophyappearpointless(ibid.:16).Pluralismwasaconditiontobeovercomethroughthesystematicpresentationofaselectivehistoryofphilosophy.Onlyifonemadeacarefulselectionofwhatshouldbeincludedandrestrainedtheurgetocompletenesswouldtheprogressiveunfoldingoftruthbecomevisible.Hegelwasalsoinvolvedinanimportantbutlargelyforgottendebateonthecanon.AtthebeginningofthenineteenthcenturysomeEuropeansbecameawareoftherichnessofIndianphilosophy.Mostnotably,FriedrichSchlegelinOntheWisdomandLanguageoftheIndians(1809)attemptedtogiveHinduphilosophyaplacewithinthephilosophicalcanon.ItistellingthatheattemptedtodosobylocatingIndianphilosophyattheoriginoftheWesterntradition.Thatistosay,hepresentedIndianphilosophyastheforgottensourceofacontinuousnarrative.Heproposedakindofrenaissance,somewhatonthemodeloftheItalianRenaissance.Schlegel’seffortsdidnotsucceed,inpartbecausehehimselfconvertedtoCatholicismandsodirectedhisinterestselsewhere.Hegel,whowasastaunchopponentofSchlegel’sconceptionofphilosophy,neverthelesstookIndianphilosophysufficientlyseriouslytodiscussitonoccasioninhislecturesonthehistoryofphilosophy,althoughherejecteditsclaimtobeing‘philosophyproper’(Hegel1994:346–7).TodaytheideathatphilosophybeganinGreeceisunderattack.Firstly,thereisaquestionastowhytheupholdersofphilosophyasauniversalhumanpossibilityshouldbesokeentoinsistthatphilosophyhadasingle,identifiable,beginning.Thisseemstobeanunwarranteddogmatism,particularlyinthefaceofthefactthateventhestrict570\nETHNICITY,CULTUREANDPHILOSOPHYcriteriausedtoprivilegetheplaceofGreekphilosophycanbemetelsewhere,forexample,inthediscussionoflogicattainedinIndia.Secondly,thereisalongstandingdisputeaboutwhymoreattentionisnotpaidinthestandardhistoriestothenon-GreeksourcesonwhichtheGreeksclearlydrewandwhichprovideevidenceofphilosophiesthatprecededthem.ThepointismostoftenmadewithreferencetotheclaimsofEgyptandhascometobeparticularlyassociatedwithcertainAfricanphilosophers,suchasCheikhAntaDiopandTheophileObenga,andAfrican-AmericanphilosopherslikeGeorgeJames(1954).However,someestablishedclassicalscholars,suchasBurkert(1992)andWest(1971),havealsoinsistedontheroleofZoroastrianism,therebyrenewinganotherconnectionthatwasreadilyacknowledgedintheseventeenthcenturybutthathastendedtobeneglectedoverthelasttwohundredyears.ItwouldbeamistaketoattendonlytothosemomentswhensomeofthecentralfiguresoftheWesternphilosophicalcanondismissedtherepresentativesofnon-WesternphilosophieswithoutalsoacknowledgingthosemomentswhenWesternphilosophersprovedmoreopentophilosophiesfromelsewhere.ThroughoutthehistoryofWesternphilosophytherehavebeenstrikingexamplesofthinkersbeingdrawntowhatwasforeigntothem.Indeed,oneshouldprobablycounttheenthusiasmofThomasAQUINAS(chapter24)forGreekphilosophyanditsArabcommentatorsamongthem,asitisonlythesubsequentconstructionofWesternphilosophyasaunitythatconcealstheradicalnatureoftheencounter.Recognitionoftheexistence,ifnotthesignificance,ofsuchdialoguesisacknowledgedevenwithinfairlystandardaccountsofthehistoryofWesternphilosophy.LeavingasidethecontroversyoverGreekphiloso-phy’sdebttotheEgyptiansandthePersians,aswellasthedisputedcaseofPlotinus’sallegeddebttoIndianphilosophy,onemightlookatMarsilioFicino’stranslationoftheHermetica,LEIBNIZ’s(chapter27)responsetotheIChing,Hegel’scommentsontheBhagavad-Gita,Heidegger’senthusiasmforLaoTsu,andsoon.Indeed,althoughHEIDEGGER(chapter41)dogmaticallymaintainedaconceptionofphilosophyasessen-tiallyWestern(Heidegger1958:31),heneverthelessintroducedaconceptionofpoeticthinkingthatopenedthewaytoafruitfuldialoguewithcertainaspectsofIndianandChinesephilosophy.Westernphilosopherswhoarenewtodebatesaboutthemeritsofso-callednon-WesternphilosophiessometimesexpresssurpriseabouthowquicklythefocusoftheargumentshiftsfromthequestionofthemeritsofthosephilosophiestothequestionoftheconstructionofthehistoryofphilosophyasexclusivelyWestern.TheirtendencyistoseethisthesisaboutthebeginningofphilosophyasacontingentHISTORICALTRUTH(pp.000–00)whichisirrelevanttothequestionofwhether,forexample,Confucianismisphilosophy.However,proponentsofthebroader,moreinclusive,viewofphilosophyhavemadethequestionofthehistoryofphilosophyoneofthemainsitesofcontesta-tionbecausetheysuspectthatthisaccountofhistoryisbeingemployedtolegitimateanundulynarrowdefinitionofphilosophyandsocannotbeleftunexamined.2TheExampleofAfricanPhilosophyThemostsustaineddiscussionofthequestionofethnicphilosophyinrecenttimeshastakenplaceinthecontextofAfricanphilosophy,althoughsimilardebatescanbefound571\nROBERTBERNASCONIwithreferenceto,forexample,LatinAmericanphilosophy.Thedebateoverthenature,andeventhepossibility,ofAfricanphilosophymaybedifferentincharacterfromthatoccasionedby,forexample,IndianorChinesephilosophy,butitindicatesveryclearlywhatisatstakewhenaphilosophy’sethnicityismarked.Indeed,itisinpartthevehe-mencewithwhichphilosophersintheWestlikeHUME(chapter31),KantandHegeldismissedtheintellectualpotentialofAfricansonracialgroundsthathasledtotheten-dencyforthediscussionofAfricanphilosophytobepreoccupiedwiththequestionofitspossibility.IthasbecomecustomarytofollowOderaOruka’sdistinctionbetweenfourschoolsofthoughtorkindsofAfricanphilosophy(Oruka1990).Thefirst,theethnographicalschool,describestheworldviewofaparticularAfricanpeopleorcollectionofpeoples.Itsbest-knownexemplarisBantuPhilosophy,aworkbyaBelgianpriest,PlacideTempels(1959).HeconsidereditthetaskofWesternphilosopherstowritethephilosophyoftheBantuforthem,onthegroundsthattheywereincapableofdoingitforthemselves.Thisattitudehascontributedtothedisreputeintowhichthisapproachhasfallen.Asecondapproach,thatofsagephilosophy,focusesonindividualindigenousthinkers.ItisoftenthoughttobeexemplifiedbyOgotemelli,whobecamewidelyknownthroughtheworkofMarcelGriaule(1965),butOrukapreferredtofocusonsageswhoillus-tratedagreatercapacityforcriticalthinking(Oruka1990:83–162).Athirdapproach,sometimescallednationalist–ideological,wasoriginallyassociatedwiththeAfricanpoliticianswholedtheircountriestoindependence.JuliusNyerere,LeopoldSedarSenghorandKwameNkrumahareprimeexamples.Ithasnowbecomemorebroadlyidentifiedtoincludeotherthinkerswhocontributedtothecritiqueofcolonialismandthedevelopmentofpostcolonialphilosophy,suchasSékouTouréandWalterRodney.Fourthly,someAfricanphilosophershavearguedthatAfricanphilosophyissimplyAfricansdoingphilosophyaccordingtothesameuniversalistviewofphilosophythatiswidespreadamongacademicallytrainedphilosophers.Orukareferredtothisasthe‘professional(critical)trend’.HislistofexamplesranfromAUGUSTINEOFHIPPO(chapter24)andAntonWilhelmArno,thefirstAfricantostudyphilosophyataGermanuniversity,tocontemporaryfigureslikePaulinHountondjiandKwasiWiredu.TotheextentthatAfricanphilosophershaveallowedWesternphilosopherstoprovidethedefinitiondeterminingwhatisandwhatisnotphilosophy,theyseemtobeplacedinadoublebind:whenAfricanphilosophytakesWesternphilosophyasitsmodel,thenitseemstomakenodistinctivecontributionandsoeffectivelydisappears,butwhenitsspecificityisemphasizedthenitscredentialstobeconsideredgenuinephi-losophyareputinquestionanditisdismissedeitherasreligionoraswisdomliterature.Africanphilosophy’spreoccupationwithestablishingitscredentialsasphilosophy,usuallyaccordingtoWesternstandards,hasclearlyimpededitsdevelopment.Butitisataskwhoseimportanceisobviousinthecontextofcontemporaryacademicinstitu-tions.KwasiWireduandValentineMudimbohaveeachintheirownwaynegotiatedtheimpasse.Wiredu(1996),employingthetoolsofANALYTICPHILOSOPHY(pp.2–6),hasfashionedasynthesisoftraditionalandmodernideastoarticulatetheconceptualityoftheAkan.Mudimbe(1988),bycontrast,usedFOUCAULT(chapter42)toquestionthediscoursesthroughwhichtheacademyhastheorizedAfricanidentities.TheproblemidentifiedinthecontextofAfricanphilosophycanbestatedmoregenerally:doesWesternphilosophyestablishthetermsaccordingtowhichanyother572\nETHNICITY,CULTUREANDPHILOSOPHYphilosophyisjudgedorcanthelatterberecognizedasphilosophyonitsownterms?ToaccepttheformeralternativewouldmaintaininforcethestructuralmechanismbywhichtheachievementsofthesecultureshavehistoricallybeendiminishedwithintheWest.ThequestionofethnicphilosophythusnecessarilypresentsachallengetothedominantphilosophythathaslargelydefineditselfintermsofUNIVERSALREASON(pp.395–7).Thedecisivequestionisthusthequestion‘whatisphilosophy?’Itisworthnoting,however,thatthedefinitionofphilosophyiswidelyemployedasaninstrumentofexclusionbetweenthevarioussectarianphilosophiesintheWestandnotonlytoconfinephilosophytotheWest.Ithas,forexample,beenusedinanalyticphilosophy’spolemicagainstCONTINENTALPHILOSOPHY(chapters40,41,42).Andyet,althoughphilosophersliketoproclaimtheirrighttopoliceorotherwiseinterrogateotheraca-demicdisciplines,onthegroundsthatthosedisciplinesareincapableofinvestigatingtheirowncentralconcepts–forexample,theconceptoflifethatsustainsBIOLOGY(chapter10)–philosophersarerarelyas‘unphilosophical’aswhentheyarelayingdownthelawaboutwhatisandwhatisnotphilosophy.3InterculturalisminPhilosophyThereisaquestionabouttheidentitiesusedtodescribevariousphilosophies,buttheyarenotasproblematicasissometimesclaimed.OnedoesnothavetobecommittedtosomeformoflinguisticorculturalRELATIVISM(pp.395–7)tobelievethatlabelslike‘Arabicphilosophy’,‘theScottishMoralists’or‘contemporaryFrenchphilosophy’areinformative.EvenafterHegel’srestrictionof‘philosophyproper’totheGreeksandtheirheirs,thequestionoftheculturalidentityofphilosophywasnotregardedasclosed.HegelhadbothabroadandanarrowunderstandingoftheGermanic,butinwhicheversensehemeantit,philosophywasfirmlylocatedalongaGreek–Germanaxis.ItwasthepracticethroughoutthenineteenthandtwentiethcenturiestodivideEuropeanphi-losophyalongnationallinesandwrite,forexample,specificallyofFrenchorScottishphilosophies.Culturalfactorsweremoreoftenappealedtoasawayofexplainingwhyanothernationorgroupdidnotrecognizethetruththatwaspresentedtothemonaplatter,butinsteadpreferredtoadheretofalseopinionsinlinewiththeirnationalchar-acteristics.Ofcourse,termslike‘Africanphilosophy’arelesslikelytobeinformativethanlabelslike‘Bantuphilosophy’becauseoftheirlevelofgenerality,buttheyneverthelessretainacertainusefulness.Itisdisingenuoustoquestion‘African’asanidentity,evengivenitsquestionablegenealogy,andyetleavetheideaofWesternphilosophyintact.TheconstructionofAfricarepresentsaproblem,butnolessofaproblemthantheconstructionof‘theWest’asemployedinthephrase‘Westernphilosophy’,particularlyiftheGreeksarethoughtofasWestern,whichis,ofcourse,nothowtheyunderstoodthemselves.Thereisalsoaquestionastotheresourcesavailablefortheorizingtherelationofaspecificethnicitytothephilosophiesthataresaidtoexpressorrepresentit.Reflectionaboutphilosophyandethnicityhasthetaskofestablishingtheconditionsofagenu-inelyopendialogue.Agreementaboutthenatureofphilosophyisnotoneofthoseconditionsinthewaythatcareabouttranslationis.Oneofthewaysinwhichphiloso-phyhappensisthroughthetranslationofthedefiningexperiencesofonecultureinto573\nROBERTBERNASCONIalanguagethatmakesthoseexperiencesaccessibletopeoplewhohavenotundergonethem.InsteadofWesternerstellingtheBantuwhattheirphilosophywas,asTempelsproposed,oneofthetasksofBantuphilosophersis,onthismodel,toarticulatethemostcentralideasoftheircultureinawaythatrendersitintelligibletomembersofverydifferentculturalgroups.Philosophy,inthisconception,isnotmerelytheexpres-sionofworldviews.Itisavehiclenotjustforlearningabout,butalsoforlearningfrom,othercultures.ThecriticalfunctionofphilosophytherebycomestobeunderstoodDIALECTICALLY(pp.746–7)asthatmomentwithinanencounterthatcanbetrans-forming.Therearethosewholeavesuchanencounteruntouched;butothers,ifnotconverted,areattheveryleast,inT.S.Eliot’sphrase,‘nolongeratease,hereintheolddispensation’(Eliot1963:110).Onedangerofanyattempttodrawethnicdivisionsamongphilosophiesisthatitlendsitselftobiologicalessentialism.Anotherdangeristhataculturecancometobeidentifiedwithwhatbelongstoitspast.Inbothcasesculturecancometobethoughtofasconfinedtoastaticsetofbeliefsalreadyinplace,givingrisetotheimpositionofthestrait-jacketofauthenticity.However,anadequateunderstandingoftheoperationofhowcross-culturaldialoguetakesplaceinpracticeexposesthesedangersmoreastrapsthattheoriesfallintothanrealpossibilities.Someconceptionsofphilosophyarebetterabletoaccommodatephilosophicalpluralismthanothers.FromthisperspectivethedebatebetweenKantandHerderinthe1780srepresentsadecisivemomentinWesternphilosophy’sunderstandingofitsrelationto‘rival’traditions,althoughtheissuebetweenthematthetimewasthePHILOSOPHYOFHISTORY(chapter14)ratherthanthehistoryofphilosophy.Kant’sideaofacosmopolitanismthattookitscuefromEuropeclearlyhadmoreimpactonthewaythehistoryofphilosophycametobewritteninthenineteenthcentury,butonecanfindinHerdertheresourcesforacertainmulticulturalism.Indeed,Herderwasunusualinthewayhenotonlygaveaplacetoeverycultureintheformationoftheideaofhumanity,buthealsoallowedforoneculturetochallengeanother.PerhapsbecauseacertainviciousformofcosmopolitanismwasgaininggroundamongtheEuropeanpowers,theHerderianapproachwaslargelyneglected.Thedecisivemomentwasperhapswhen,followingHegel,themodeladoptedforunderstandingwasthatof‘pureself-recognitioninabsoluteotherness’(Hegel1977:14).ThismodelremainsaliveinGadamerianHERMENEUTICS(p.397),wheretheideaoftraditionstillhasacontrollingforce.Thepossibilityofadifferentapproachissketchedbystudentsofthestudyofreligion.TheveryconceptionofreligionintheWesthasbeenenrichedandtrans-formedbyanunderstandingof‘non-Western’religions.ItisnolongeraquestionofseeingwhatinotherculturesresemblesorpreparesforChristianity,aswasoncethecase.Cross-CulturalDialogueThequestionofphilosophyandethnicityis,therefore,bynomeansrestrictedtoques-tionsaboutthephilosophicalcanon,butembracesaconcernforestablishingthecondi-tionsforacross-culturalgenuinedialogue.Itisatleastarguablethattreatingtheexistenceofphilosophyasaculturaluniversal,presentinsomedegreeinallculturesatleastasanimpulse,withoutchallengingthestatusofWesternphilosophyforphiloso-phyingeneral,hasdonemoretodiminishIndianandChinesethoughtthantheirtotal574\nETHNICITY,CULTUREANDPHILOSOPHYexclusiondid.Itmeansthattheyarenolongerapproachedontheirterms,butarejudgedonthebasisofalienstandards.Theroleofmulticulturalismwithinphilosophygoesbeyondrecognizingsomeaspectsof,forexample,Indianthoughtaslegitimatephiloso-phyaccordingtoanalreadyestablisheddefinitionofphilosophy.Forthisreasonthephrase‘cross-culturalconversation’canbeusedtocharacterizetheprocessbywhichcultures,whichareinanycasenotdiscreteunits,canengagewitheachotherinawaythatismutuallyempowering(Balslev1996:17).InGermandiscussionstheterm‘inter-culturalism’(Kimmerle1994)hasservedasimilarpurposeanditisperhapspreferabletotheterm‘multiculturalism’,whichsuggestsamorestaticconception.TheexistenceofWesternphilosopherswhoincorporatetheirknowledgeof,usually,IndianorChinesephilosophyintotheirworkshowscross-culturalconversationatwork,buttofocusonthemwouldbetoretainacertainone-sidedness.SarvepalliRadhkrishraninIndiaandKitarôNishidainJapanareamongthebest-knownexam-plesofcross-culturalconversationoutsideofEuropeandNorthAmerica.Nishida’sphilosophicaleffortwasarguablydevotedto‘thedevelopmentandarticulationofthelogicoforientalexperience’(Inagaki1993:293),withastrongemphasisontheBud-dhisttradition,buthisextensivediscussionsofWesternphilosophy,particularlyofKantandHegel,wereintegraltohisperformanceofthistaskandinnowayadistractionfromit.However,forthequestionoftherelationbetweenphilosophyandethnicity,theworkofLeopoldSenghor,theformerpresidentofSenegal,isparticularlyinstructive.SenghorwasoneofthefoundersoftheNegritudemovement,whichhadariseninParisinthe1930samongstudentsfromtheFrenchcolonieswhohaddiscoveredtheirracialidentityinlargemeasureasaresultofthewaytheyexperiencedthemselvesasseeninracialterms.ButfarfromtryingtobeexclusivelyAfrican,whichishowsomeofitscriticsrepresenttheNegritudemovement,Senghoraspiredtoakindofculturalmis-cegenationtowhichallculturescouldcontribute.IndirectcontrasttothecivilizationoftheUniversalthatitwasofferedbytheWest,andwhichseemedonlyanothernameforexplicitlyWesternideas,Senghorproposedauniversalcivilization.Thegoalwas‘toassimilate,nottobeassimilated’(Senghor1964:39).Amonghisextensiverefer-encestoAfricanphilosophy,onefindsfrequentappealsto,forexample,HenriBergson,PierreTeilharddeChardinandMartinHeidegger.Attentiontotheprocessoftranslationhelpsclarifyhowunderstandingacrosstra-ditionstakesplace.Manykeyphilosophicalterms,likelogos,yoga,tao,zenor,forthatmatterGeistorDasein,areusuallyleftuntranslated.ThesewordshavenowpassedintotheEnglishlanguagebutwithoutlosinganyoftheirforeignness.Thisshowshowtheoften-hearddemandthateveryphilosophyexpressitselfclearlyinEnglishistoosevere.Thelexiconofanylanguageexpandstoaccommodatewhatisnew,especiallyintheencounterbetweencultures.OneseesthisveryclearlyinthecaseofJapanese,whichinitiallylackedawordfor‘being’anddevelopedoneinthecourseoftranslatingsomeoftheclassicsofWesternphilosophy.However,thefactthattheGreekword‘philoso-phy’hasitselfproveddifficulttotranslateintomanyoftheworld’slanguagesdoesnotmeanthatitmakesnosensetotalkofphilosophyinrelationtotheculturesthatspeakthoselanguages.Philosophyisnotstatic.TheWesterntraditionofphilosophyisfamiliarwiththewaygreatphilosophersorgreatmovementsinphilosophyredefinethefield.Aparticularly575\nROBERTBERNASCONIfruitfulconceptionoftherelationofphilosophytocultureisnowevidentinpost-HolocaustJewishphilosophy.Forexample,EmmanuelLevinasnotonlytranslatedaHebraicwisdomthathedrewfromtheTalmudintothelanguageofGreekphilosophy,buthealsoenrichedthatlanguagefurtherbyintroducingintoitarangeofexperiencesthathadpreviouslybeenabsentfromthattradition.Inparticular,Levinasclaimedthattheexperienceofbeingpersecutedgaveaccesstodepthsofsubjectivitytowhichpre-viousphilosophyhadnothadaccess.AlthoughLevinashimselfhadreservationsaboutlabellinghiswork‘Jewishphilosophy’,therewasaclearcontinuitybetweenhiscom-mentariesontheTalmud,whichheexplicitlydescribedastranslationsfromHebrewintothelanguageoftheuniversities,andhisdirectlyphilosophicalworks,whichneededafurthermomentoftranslationinsofarasphilosophycalled,inhisview,fordifferentkindsofevidence.Levinashimselfhadasomewhatconservativeconceptionofthephilosophicalcanon,andhewasproneininterviewstodismissnon-Westerncul-turesinoutrageousterms(Mortley1991:18).Nevertheless,hisuseoftheexperienceofbeingpersecutedasevidencecanserveasamodelforotheroppressedgroupscom-mittedtothephilosophicalarticulationandcommemorationofthesufferingthathashelpedforgetheiridentityandculture,andwho,likeLevinas,arecommittedtoresist-ingthetendencyofthetraditionalaspectsofaculture’sexistencefromsinkingtotheleveloffolklore(Levinas1994:197).Asimilarrelationbetweenphilosophyandex-periencecanbefoundinAfrican-Americanphilosophy,whichhasbeenaptlydescribedbyLeonardHarris(1983)asa‘philosophybornofstruggle’.4Re-examiningtheEurocentrismoftheCanonandRenewingPhilosophyInadditiontobroadeningthephilosophicalcanonandseekingwaystoadvancethedia-loguebetweenthevariousphilosophiesincludedinsuchanexpandedcanon,intercul-turalisminphilosophyengagesinare-examinationofthemajorfiguresoftheWesternphilosophicaltradition.Muchoftheenergyanimatinginterculturalismwithincon-temporaryphilosophyhascomefromthewidespreadtendencyofcontemporaryWesternphilosopherstodismissaslackinginphilosophicalsignificancethefactthatmanyofthemostprominentphilosophersofthetraditionwerelesscriticaloftheWest’sexploitationofothersocietiesthanmanyoftheircontemporaries.Inotherwords,theethnocentriccharacterofWesternphilosophyisillustratednotonlybytheconstructionofthecanon,butalsobythefactthat,forexample,theracismofmanyofthecentralfiguresfromthecanonis,liketheirsexism,oftenignored.FEMINISTPHILOSOPHERS(chapter20)haveforsometimeinsistedonconfrontingthephilosophicalcanonwithitsblindnessanditscomplicityinoppression,butsimilarcriticismshavealsolongbeenwidespreadamongotheroppressedorneglectedgroups,anditismerelyamatteroftimebeforetheytoobecomemorewidelydisseminatedthroughtheacademy.ThatWesternphilosophyoftentriestotakecreditforthenobleideasithasarticulated,whilerefusingresponsibilityfortheevilthatithasunderwritten,sometimesinthenameofthosesameideals,isoneofthefocalpointsofpostcolonialphilosophy.Thishasledsomeofthemajorphilosophersofthecanontobescrutinizedafresh.Forexample,LOCKE’s(chapter29)discussionofrightsisnowseeninthecontextofhisinvolvementin576\nETHNICITY,CULTUREANDPHILOSOPHYtheAfricanslavetrade,inthesamewaythathisaccountoftheoriginofprivatepropertyisnowseeninthecontextofhissupportforthecolonizationofNorthAmerica.Similarly,Kant’scosmopolitanismiscontrastedwithhisadvocacyoftheconceptofraceandhisoppositiontoracemixing.AndNIETZSCHE’s(chapter40)writingsareminedforhisremarksonwomenandBlacks.AlthoughithasoftenbeenAfrican-Americanphiloso-pherswhohaveinsistedinthefirstinstanceonaskingaboutJohnLocke’sinvolvementinAfricanslavery,JewishphilosopherswhohaveinsistedonrecallingKant’santi-semi-tism,andwomenphilosopherswhohaveinitiallyraisedquestionsaboutNietzsche’ssexism,theseconcernsshouldnotbeandhavenotbeenconfinedtomembersoftheoffendedgroup.However,manystandardcommentariesstilldismisstheseapproachesasirrelevanttoanunderstandingofthespecificphilosophyunderconsideration.WhatsignificanceshouldbegiventothefactthatLockeandKantfailedtoquestionthechattelslaveryinAfricansofwhichtheywerewellaware,eventhoughtheyhadattheirdisposaltheintellectualresourcestochallengeit?ThisfailureisregardedbymanyWesternphilosophersasmerelyacontingentfactthattellsusnothingsignificantabouttheirphilosophies,stilllessthetraditiontowhichtheybelong.WhereasLocke’sandKant’sdefendersexplainthattheyaremerelymakingthelogicalpointthatthereisnothingatthecoreoftheirphilosophiesthatentailsrace-basedslavery,criticscom-plainthatthatdefinestheirphilosophiestoonarrowly.Thisexampleshowshowattheheartofthedebateisthequestionofthedefinitionofphilosophy.Therecognitionthatsuchcontroversiesprovideanotheraccesstothequestionofthenatureofphilosophygivesadditionalimportancetothem.Thewidespreadtendencytoconfinethephilo-sophicalstudyofmajorfiguresinthehistoryofphilosophytoconsiderationoftheirontologicalorepistemologicalarguments,eventhoughthesephilosophershadwide-spreadintereststheyconsiderednolessphilosophical,isdefendedasawayoffocusingonwhatismostimportantintheirthought,butatthesametimeitdistortstherelationofphilosophytoitshistoricalcontext.Foronething,itmakesitseemthatphilosophyisunmotivatedorunconcernedabouttheprevailingcontext.ExtendingtheRoleofPhilosophyWhatseems,particularlyfromtheoutside,asdefensivenessonthepartofWesternphi-losophyaboutitsownhistoryisalsoareflectionofitsacceptanceofareducedroleforphilosophy.Westernphilosophyaspractisedinuniversitiesacrosstheworldhasbecomereconciledtoitsroleasasomewhatmarginalactivity,atleastcomparedwithitsformercentrality.Nevertheless,understandingatextinvolvesanappreciationofwhatisbeingattemptedinthattext.Thisisnotalwayseasy,asattemptstoestablishthelegitimacyofatextasphilosophyofteninvolveminimizingsomeofitscentralelements.Forexample,withinIndianphilosophyonefindsKarman,karmamârgaintheGita,astheworkingoutofone’sownsalvation,notonlyasanideabutalsoassomethingtorealize.Pannikarhasarguedthatbyappealingtothisnotiononecanupsetacertainprevalentconcep-tionofphilosophyandtherebydeepenthenotionofphilosophy.PannikarconcedesthatthisworksbyremindingtheWestofthefactthatthereisanentiretraditionwhichpro-claimsthephilosophicallifewithoutarupturebetweentheoryandpraxis(Pannikar1993:25).HetherebygrantsthattheinsightstillwouldbeconductedwithoutreferencetoIndiaatall!Nevertheless,theargumenthelpstoestablishhowlimitedtheconception577\nROBERTBERNASCONIofacademicphilosophyhasbecome.Similarly,oneoftheargumentsoftenmadeastowhatWesternphilosophycanlearnfromChinesephilosophyisthattheformercanredis-coverthroughthelatterarecognitionoftheimportanceofthequestion:howshouldonelive(VanNorden1996)?Althoughtherehasbeenamarkedincreaseinworkcon-trastingthediscussionsofthevirtueswithintheConfuciantraditionwiththosefoundinclassicalGreekphilosophy,forthemomentthisstillseemsadistantpossibility.5ConcludingRemarksEvenifonerestrictsoneselftothedominantconceptionofphilosophythatiscurrentwithinAnglo-Americanphilosophyandtakesthatasone’sstandard,onewillnever-thelessbeabletofindthatstandardmetelsewhere,forexample,inselectmomentsinChineseorIndianphilosophy.Onewouldalsohavetoacceptthatmuchthat,bycon-sensus,passesforphilosophywouldalsonotmeetthatcriterion.Thepointisthatthereisnosingledefinitionofphilosophythatcoverseveneverythingthatisregularlyrec-ognizedasphilosophywithintheacademy.Theflexibilityoftheconceptionofphiloso-phyiswhatallowseachmajorphilosophertoredefinethediscipline,ofteninawaythatbelittlesorevenexcludesfromphilosophymuchthathadpreviouslybeenincluded.Nevertheless,ourfamiliaritywiththatproceduredoesnotgiveKantianstherighttodenythat,forexample,pre-Kantianmetaphysicsbecalledphilosophy.Norshouldweconfusethelegitimacyofaspecifictraditionofphilosophywithallofphilosophy.Itistruethatphilosophicaldebateinthetwentiethcenturyhasoftenbeenconductedbyrefusingtosharethetitle‘philosophy’withone’sintellectualopponents,buttherela-tiveflexibilityoftheconceptionofphilosophydoesnothavetobeusedonlyasawayofnarrowingwhatcountsasphilosophy.Itcanequallywellservetoexpanditsbound-aries.JustasthestudyofreligionintheWestlargelyceasedduringthecourseofthetwentiethcenturytobeawayofpromotingChristianityattheexpenseofnon-Westernreligions,sothestudyofphilosophyintheuniversitiescanberedesignedtooffergreateropportunityforstudyingdifferenttraditions.Untilthishappenstherewillalwaysbeasuspicionthat,inspiteofthepretensionsofWesternphilosophytowarduniversalism,forthemostpartWesternphilosopherswilltendtobemoreprovincialintheirintereststhan,forexample,manyAfrican,Chinese,IndianorJapanesephilosophers,whooftenhavemasteryofmorethanonetradition.FurtherReadingThechallengetotheGreekoriginsofphilosophyisoftenassociatedwiththefirstvolumeofMartinBernal’sBlackAthena(1987),butalthoughthatbookusefullydirectsattentiontothehistoryofthehistoryofphilosophyintheeighteenthcentury,hisaccountisnotalwaysreliable.ForthehistoriographicalquestiononewoulddobettertoconsulttheessaysbyBlackwellandSchneiderinKelly(1997).ThecaseagainsttheGreekoriginsofphilosophyhasbeenmadewithreferencetoAfricanandparticularlyEgyptianphilosophy,byCheikhAntaDiop(1991)andTheophileObenga(1992).ThecaseforfindingtheoriginsofphilosophyintheNearEasthasbeenmadebyBurkert(1992)andWest(1971).Foranexcellentsurveyofmanyofthedifferentkindsofphilosophythatareatissueindiscussionsofphilosophy’srelationtoethnicity,seethecollectioneditedbyDeutschand578\nETHNICITY,CULTUREANDPHILOSOPHYBontekoe(1997).ThereisagoodreviewofrecentdebatesonthephilosophicalstatusofAfricanphilosophyinMasolo(1994).Oruka(1990)notonlyincludessomeofhismostimportantessays,butalsoclassicessaysbyBodunrinandOutlaw.Halbfass(1988)hasprovidedasurveyoftherelationsbetweenIndianandEuropeanphilosophythatfurthersourunderstandingofintercul-turaldialogue.ThebestintroductiontoHerderasanearlyexponentofaformofmulticultural-ismisBerlin(1976),althoughBerlindoesnotusethattermspecifically.Foranintroductiontocross-culturalconversation,seeBalslev(1996).Finally,foranexcellentexampleofaphilosophereducatedbothintheEastandtheWest,steeringacoursebetweenuniversalismandrelativism,seechapters20and26ofMohanty(1993).ReferencesBalslev,A.1991:CulturalOtherness:CorrespondencewithRichardRorty.Shimla:IndianInstituteofAdvancedStudy.——1996:Cross-CulturalConversation.(Initiation.)Atlanta,GA:ScholarsPress.Berlin,I.1976:HerderandtheEnlightenment.InVicoandHerder.London:HogarthPress.Bernal,M.1987:BlackAthena,Volume1:TheFabricationofAncientGreece,1785–1985.NewBrunswick,NJ:RutgersUniversityPress.Bernasconi,R.1997:AfricanPhilosophy’sChallengetoContinentalPhilosophy.InE.C.Eze(ed.)PostcolonialAfricanPhilosophy.Oxford:Blackwell.——2000:Krimkrams:HegelandtheCurrentControversyabouttheBeginningofPhilosophy.InC.E.ScottandJ.Sallis(eds)InterrogatingtheTradition.Albany:StateUniversityofNewYorkPress.Brucker,J.1742:HistoriaCriticaPhilosophiae.Leipzig:Weidemann.Burkert,W.1992:TheOrientalizingRevolution(translatedbyM.E.PinderandW.Burkert).Cambridge,MA:HarvardUniversityPress.Deutsch,E.andBontekoe,R.1997:ACompaniontoWorldPhilosophies.Oxford:Blackwell.Diop,C.A.1991:CivilizationorBarbarism(translatedbyYaa-LengiMeemaNgemi).NewYork:LawrenceHillBooks.Eliot,T.S.1963:JourneyoftheMagi.InCollectedPoems1909–1962.London:Faber.Gadamer,H.-G.1989:TruthandMethod(translatedbyJ.WeinsheimerandD.G.Marshall).NewYork:Continuum.Gracia,J.J.1992:PhilosophyandItsHistory.Albany:StateUniversityofNewYorkPress.Griaule,M.1965:ConversationswithOgotommêli(translatedbyR.Butler).Oxford:OxfordUniversityPress.Halbfass,W.1988:IndiaandEurope.Albany:StateUniversityofNewYorkPress.Harris,L.(ed.)1983:PhilosophyBornofStruggle:AnthologyofAfro-AmericanPhilosophyfrom1917.Dubuque:IowaStateUniversityPress.Hegel,G.W.F.1977:PhenomenologyofSpirit(translatedbyA.V.Miller).Oxford:OxfordUniversityPress.——1985:IntroductiontotheLecturesontheHistoryofPhilosophy(translatedbyT.M.KnoxandA.V.Miller).Oxford:OxfordUniversityPress.——1994:VorlesungenüberdieGeschichtederPhilosophie.Teil1(editedbyP.GarnironandW.Jaeschke).Hamburg:FelixMeiner.Heidegger,M.1958:WhatisPhilosophy?German–English(translatedbyW.KlubackandJ.T.Wilde).London:VisionPress.Herder,J.G.1800:OutlinesofaPhilosophyoftheHistoryofMan(translatedbyT.Churchill).NewYork:BergmannReprint.Hountondji,P.J.1996:AfricanPhilosophy:MythandReality(translatedbyHenriEvans).Bloomington:IndianaUniversityPress.579\nROBERTBERNASCONIInagaki,B.R.1993:TheConceptofCreationinthePhilosophyofKitaroNishida.InG.Fløistad(ed.)ContemporaryPhilosophy:ANewSurvey,vol.7.Dordrecht:Kluwer.James,G.G.M.1954:StolenLegacy.NewYork:PhilosophicalLibrary.Kelly,D.R.1997:HistoryandtheDisciplines.Rochester,NY:UniversityofRochesterPress.Kimmerle,H.1994:DieDimensiondesInterkulturellen.Amsterdam:Rodopi.Leibniz,G.W.1977:DiscourseontheNaturalTheologyoftheChinese.Honolulu:UniversityPressofHawaii.Levinas,E.1994:BeyondtheVerse(translatedbyGaryD.Mole).London:AthlonePress.Masolo,D.A.1994:AfricanPhilosophyinSearchofIdentity.Bloomington:IndianaUniversityPress.Mohanty,J.N.1993:EssaysonIndianPhilosophy.Delhi:OxfordUniversityPress.Mortley,R.1991:FrenchPhilosophersinConversation.London:Routledge.Mudimbe,V.Y.1988:TheInventionofAfrica.Bloomington:IndianaUniversityPress.Nishida,K.1966:IntelligibilityandthePhilosophyofNothingness(translatedbyR.Schinzinger).Honolulu:East–WestCenterPress.Obenga,T.1992:AncientEgyptandBlackAfrica(translatedbyS.MartinonandA.Shelk).London:Karnak.Oruka,H.O.(ed.)1990:SagePhilosophy:IndigenousThinkersandModernDebateonAfricanPhilosophy.Leiden:E.J.Brill.Panikkar,R.1993:Satapathaprajñâ.ShouldWeSpeakofPhilosophyinClassicalIndia?ACaseofHomeomorphicEquivalents.InG.Fløistad(ed.)ContemporaryPhilosophy:ANewSurvey,vol.7.Dordrecht:Kluwer.Santinello,G.etal.1993:ModelsoftheHistoryofPhilosophy,Volume1:FromitsOriginsintheRenaissancetothe‘HistoriaPhilosophica’.Dordrecht:Kluwer.Schlegel,F.1849:OntheIndianLanguage,Literature,andPhilosophy.InTheAestheticandMiscellaneousWorksofFrederikvonSchlegel(translatedbyE.J.Millington).London:Bohn.Schneider,U.1990:DieVergangenheitdesGeistes.Frankfurt:Suhrkamp.Senghor,L.1964:Liberté.Paris:ÉditionsduSeuil.Tempels,P.1959:BantuPhilosophy(translatedbyColinKing).Paris:PresenceAfricaine.VanNorden,B.W.1996:WhatShouldWesternPhilosophyLearnfromChinesePhilosophy?InP.J.Ivanhoe(ed.)ChineseLanguage.Chicago:OpenCourt.Vossius,J.G.1657:DePhilosophorumSectis.TheHague.West,M.L.1971:EarlyGreekPhilosophyandtheOrient.Oxford:OxfordUniversityPress.Wiredu,K.1996:CulturalUniversalsandParticulars:AnAfricanPerspective.Bloomington:IndianaUniversityPress.Wolff,C.1985:RedeüberdiepraktischePhilosophiederChinesen.Latin–German(translatedbyM.Albrecht).Hamburg:FelixMeiner.Yates,F.1964:GiordanoBrunoandtheHermeticTradition.Chicago:UniversityofChicagoPress.DiscussionQuestions1Inwhatsenseisphilosophytranscultural?2Ifphilosophyisnottranscultural,isitcondemnedtoculturalrelativism?3Whyhasphilosophyasadisciplinelargelyresistedthepronouncedtendencytowardmulticulturalisminneighbouringdisciplines?4Doesitmakesensetotalkaboutthebeginningofphilosophy?Ifso,whatdiffer-encedoesitmakewherephilosophybegan?5Isthereprogressinphilosophy?Howwouldthisbeshown?580\nETHNICITY,CULTUREANDPHILOSOPHY6Howisthestandardestablishedbywhichwhatisandwhatisnotphilosophycomestobedetermined?7IsWesternphilosophyoneformofethnicphilosophyamongothers?8WhatconsiderationsledtotheidentificationofWesternphilosophyasphilosophyproper?9Whatdifference,ifany,doesitmakeforthequestionofdeterminingwhatcultureshavephilosophythattheword‘philosophy’isderivedfromGreek?10Whatproblemsaretobeguardedagainstwhenconstruingphilosophicaldivisionsalongethniclines?11Whatistherelationbetweenphilosophyandexperience?Canculture-specificexperienceslegitimatelyprovidethebasisforaphilosophy?12WhatsignificanceshouldbegiventotheinstancesofracismwithinthehistoryofWesternphilosophy?Underwhatconditionsshould,forexample,Locke’sorKant’sracismcometoalterourviewsoftheirphilosophy?13Doesthefactthatamodeofthinkingisattachedtoacultureorawayoflivingcountasevidencethatitisnotphilosophical?581\nPARTIIHISTORYOFPHILOSOPHY\n22AncientGreekPhilosophyROBERTWARDYWesternphilosophyhasitsoriginsinancientGreece,withspeculation,problemsandargumentsstretchingfromthePresocraticstothemajorHellenisticschools.Insteadofconcentratingontherichvarietyofphilosophicaldoctrinesexploredanddefendedinthisbrilliantformativeperiod,thefollowingchapterfocusesonthedeeplyoriginalconceptionsofmethoddevelopedbySocrates,Plato,AristotleandtheHellenisticStoicsandEpicureans.Thisstudyofmethodprovidesaframeworkforunderstandingconcep-tionsofphilosophy,thegoalsofphilosophersandthepointofparticularphilosophicalquestionsnotonlyinancientGreece,butalsoinsubsequentphilosophicalperiods,includingourown.WhyreadancientGreekphilosophy?Themostobviousansweristhatifthehistoryofphilosophyisaworthyobjectofstudy,thenanyseriousphilosophermustknowsome-thingoftheGreekcontributiontothesubject.Thispriorityisnotsimplyamatterofancestralpiety.ItisnotmerelythatWesternphilosophybeganinancientGreece.Becausemanyofthedefiningtechniques,characteristicpreoccupationsandperennialdifficultiesofphilosophyaroseduringitsbrilliantfirstphase,anyoneworkinginigno-ranceofthoseachievementsandfailureslaboursunderasignificanthandicap.Thislargeclaimcanbegivenboth‘historical’and‘philosophical’specifications.MuchMedieval(chapter24)andlaterphilosophyisunintelligibletothoseunacquaintedwiththeGreekmaterialthatinlargemeasuresettheagendaforsubsequentthinkers;whileviewsofthedistinctivegoalsandmethodsofphilosophicalinvestigationwillbeseriouslyimpoverishediftheinitialGreekconceptionsareleftoutofaccount.Greekphilosophyisasgoodasanyphilosophyeverdone.Duringthetwentiethcenturyitattractedmanyoftheverybestphilosophicalhistorians,andtheexplosioninscholarshipwithinrecentdecadesmakesevenabriefoverviewofworkonanumberofcentralissuesimpossible.NewinterpretationsofPresocraticphilosophycontinuetoappear,accompaniedbyrenewedreflectiononwhatfactorsdistinguishthemomentousnewmovementofrationalisticinvestigationfromolder,mythologicalpatternsofthought(Lloyd1979:10–58).MuchinsighthasbeengainedabouttheGreekinven-tionofaxiomatics,andabouthowphilosophersandmathematiciansmighthaveinter-acted(Barnes1975;Lloyd1979:59–125).Revitalizedinterestinthetheoryofthevirtues,practicalreasonandthenaturalisticturninEthics(chapter6)hassharpened\nROBERTWARDYtheappreciationthatAristotle(384–322bc)isperhapstheWest’smostimportantethicalthinker(MacIntyre1981;Williams1985).AdvancesinthehistoryofEpiste-mology(chapter1)haveestablishedthatHellenisticscepticalcontroversyplayedavitalpartintheevolutionoftheoriesofknowledge(Burnyeat1980;AnnasandBarnes1985).AlthoughAristotle’sLogic(chapter4)continuestobenefitfrominnovativenewinterpretations(Lear1980),thereisalsoasteadyincreaseintheilluminatingstudyofhithertoneglectedancientlogicalsystems(Frede1974;Brunschwig1980;Barnes1980;Denyer1991).AmassiveprojectoftranslationandexegesisiscurrentlymakinggenerallyavailablethevastresourcesofthelatecommentatorsonAristotle,whosephilosophicalcreativityhadhithertobeenneglectedontheerroneouspresumptionthattheexegeticalformatprohibitsoriginalityandbecauseofthesheerdifficultyofthelargelyuntranslated,unexploredsources(Sorabji1987–).1SocratesandDialecticalMethodThislistisonlyaveryincompletecatalogueofongoingresearch.Apartfromafewpassingillustrations,Ishallconcentrateexclusivelyononlyoneespeciallyvigorousfield,thestudyofdialecticalmethodinancientGreece.Mydiscussionwillstillbeheavilycurtailed,butitwillsummarizesomeofancientthought’smostfruitfulachieve-mentsandindicatecurrentproblemswhichshouldexerciseallphilosophers.Philosophyhasbeenprovocativelydescribedas‘footnotestoPlato’.Inthecontem-porarysceneitwouldbemoreaccurate,ifnolessprovocative,tocharacterizephilosophyasanswerstoSocrates(c.470–399bc).Whyisheofsuchoverwhelmingimportance?AlthoughtheargumentativeexplorationofrealityandillusionwhichisphilosophywasundertakenforthefirsttimeintheGreekcoloniesofAsiaMinorroughlyacenturybeforehistime,theindomitablyrationalisticSocratesofthePlatonicdialogueshascompletelydominatedtheWesternportraitofthearchetypicalphiloso-pher,ascanbeseenduringtheRenaissanceinRabelais(c.1494–1553)andMontaigne(1533–92),andduringthenineteenth-centuryintellectualupheavalinKierkegaard(1813–55)andNietzsche(1844–1900)(chapter40).Whetherasheroorasvillain,Socratesepitomizesthepromiseandthethreatofphi-losophy.ThattheseevocationsofSocratesappearasoftenonthemarginsofphiloso-phyasinitsmaintextincreasesratherthandiminisheshisstature:theSocraticimageconnectsphilosophywiththegreatmovementsinliterature,politicsandreligionwhichcollectivelygivephilosophyitsplaceinculture.TheSocratesofPlato’sSymposiumisafigureofalmostgrotesquephysicaluglinessconcealingunequalledintellectualgifts,whodevotesunceasing,single-mindedefforttothediscoveryofvitalethicaltruthsthroughdisciplinedargument.Hiseffectonalmosteveryoneisunsettling:heshakestheirmostcherishedconvictions,andarousesshameatfailuretothinkrationallyandliveconsistentlybythelightofreason.Heinspiresapassionbesidewhichordinaryeroticfeelingsdwindleintoinsignificance.PhilosophyduringthetwentiethcenturyintheEnglish-speakingworldemergedfrominfatuationwithanexceedinglynarrowself-conceptionasthehandmaidofthenaturalsciences;butitremainsaself-contained,highlytechnicalacademicdisci-plinelargelyinaccessibleeventohighlyeducatednon-specialists.Inconsequence,586\nANCIENTGREEKPHILOSOPHYAnglo-Americanphilosophyhassadlyalmostcompletelylostthestrengthitgavetoandreceivedfromthecultureatlarge,andtooeasilylapsesintoesotericscholasticism.Someresponsibilityforthisisolationmustrestwithalongstandingfeatureofphilosophy,itsclaimtotranscendmerelypersonalpersuasionbyconformingsolelytothedictatesofobjectivecanonsofargument;peopleoutsidephilosophyarenotreadilyinclinedtolistentounattractivelyimpersonal,difficultmessages.ThecentralenigmaofSocratesisthatwhereasthestandardsofargumenttowhichheadheredareamodelofrigour,heworkednotinanacademy,butinthemarket-placeandstreetsofAthens.InthePlatonicimageofSocrates,irresistiblepersonalityanddialecticalforce,subjectiveandobjectiveappeal,inextricablycombine.ContemporaryphilosophersworriedaboutthemarginalizationandinaccessibilityoftheirsubjectdowelltoponderSocrates’suniversalimpact;philosopherscomplacentabouttheirspecialismdoevenbetterifforcedtoawakenfromtheirdogmaticslumbersbytheSocraticgadfly(Plato,Apology,30E).InwhatsensewasSocratesastreet-philosopher?AsportrayedintheearlierPlatonicdialogues,hecandophilosophynotbyhimself,butratherwithandagainstaninter-locutor.Heexaminesanyonewhomsoeverwithareputationforwisdom(ibid.:23B),includingbutnotlimitinghimselftoprofessionalthinkers,anddeliberatelyseeksoutthepoliticians,poetsandartisansofAthens(ibid.:21B–22E).(SincethewomenofancientGreecedidnotengageinpublicaffairs,Socratesdoesnotarguewiththem,butthatisnotonaccountoflackofinterest;whenhedescribeshisattitudetowardsexis-tenceafterdeath,ifsuchtherebe,heindicateshisenthusiasmattheprospectofcon-tinuinghischaracteristicdebateswithfamousmenandwomenintheafterlife(ibid.:41C).)WhydidSocratescarryoninthisfashion?Ashefamouslydeclares,‘theunex-aminedlifeisnotworthlivingforahuman’(ibid.:38A).Althoughheabstainedasmuchaspossiblefrominvolvementinpoliticalaffairsasconventionallyunderstood,hisconvictionwasthattheonlywaytoachievevirtuewastounderstandit;Socraticphilosophy,sofarfrombeinganacademicavoidanceoftheactivelife,wastheessen-tialmeansofconductinglifeproperly,andchallengedeveryonewithoutexception.TheSocraticmethodofdoingphilosophy,adversarialco-operation,soundspara-doxical;howdiditwork?ItsGreekname,elenchos,occursoutsidephilosophicalcon-textsandmeans‘testing’or‘examination’.NowheredoesSocratesformallydelineate,letalonejustify,hisprocedure,butitispossibletoderiveadescriptionofhisstandardpracticefromaconsiderationofthePlatonicdialogueswhereitisputtowork.Socratesgetshisinterlocutortoofferananswertothequestion‘WhatisX?’,wherethevalueof‘X’isusually,ifnotalways,avirtuesuchascourage,justiceorpiety.Inthesedialogues,thatthereissuchathingas,say,courageistakenforgranted;andwhatsortofthingitmightbe–anitemintheworld,aspeopleandcoloursareintheirdifferentways,oramentalitem,orafeatureoftheGreeklanguage–isalsonevermadeanobjectofenquiry.ThusSocraticelenchosisamatterofdefinition,yetvitalquestionsconcerningthenatureandimportofdefinitionarenotraised.Oncehispartnerhasofferedacandidatedefinition,Socratestypicallyproceedsbyelicitinghissupportforfurtherpropositions.Sometimestheconnectionbetweenthesepropositionsandtheoriginaldefinitionisimmediatelyobvious,sometimesnot;buttheeventualresultisalwaysanargumentusingthedefinitionwhoseconclusioneitherdirectlycontradictsit,oratleastconflictswithassumptionstheinterlocutorisnot587\nROBERTWARDYwillingtosurrender.Theoutcomeisconsistentlynegative:Socratic‘examination’isrefutation,establishingthataproposeddefinitionofsomevirtueisunacceptablebecauseitentailscontradictiondirectlyorindirectly.Socratesconcludesthathispartnerhasnobettergraspofthenatureofthatwhichisbeingdefined,thedefiniendum,thanhehimself,andurgesthattheypersevereinthedefinitionalquest,sincenothingcouldbemoreimportantthanknowledgeofwhattheydemonstrablydonotknow.ThelogicoftheSocraticelenchoshascontinuouslyfascinatedworkersonGreekphi-losophy,andprovokedaseriesofextremelyimportantstudiesfromthelateGregoryVlastos(1907–91),theleadingSocraticscholarofthetwentiethcentury.Vlastosisexercisedbytwocloselyinterrelatedquestions.Theelenchosasdescribedcanevidentlyatbesthopetoestablishtheconsistency,notthetruth,oftheinterlocutor’sbeliefs,butitisoverwhelminglyobviousthatSocratesaspirestoobjectivevalidity;howcanhesupposethathisphilosophicalmethodmightgetbeyondmerelyhumanagreement?Then,intheso-called‘middle’dialogues,thePlatonicSocratesismadetopropoundacollectionofTranscendental(chapter32),Metaphysical(chapter2)andEpistemo-logical(chapter1)doctrinesstrikinglydifferentfromanythingencounteredbeforeinthecorpus.ItisnowcommongroundtoalmostallscholarsthatthesepronouncementsrepresentPlato’sownlaterphilosophy,supplementingbutalsolargelysupplantingtheelenchos,whichisitselfperhapstobeidentifiedwiththemethodofthehistoricalSocrates.SoVlastostriestounderstandhowtherealSocratescouldjustifiablyhaveusedtheelenchosasaninstrumentreachingbeyondconsistencytotruth,andwhyPlatoultimatelyrejectedhishero’sposition.In‘TheSocraticElenchus’Vlastos(1983)makesaproposalwhichhasexcitedthegreatestinterest.HeemphasizesthattheprocedureofelicitingcontradictionsketchedaboveisasignificantlyincompletedescriptionofSocrates’sprocedure.Inaddition,itisessentialthatwerecognizethatSocratesdemandsthathisinterlocutor’sresponsesbesincere:theelenchosisexamination,notofneutralpropositionswhichmightbeheld,butratherofassertedpropositions,thesebeingthesincerebeliefsofSocrates’sdialecti-calpartner.ThisfeatureofthemethodisofvitalimportancetoVlastosbecausehearguesthatiftheelenchosworksoncommitmentsratherthanbare,unassertedpropo-sitions,itcanyieldsubstantiveresultsonone,admittedlyremarkable,assumption:namely,thatthebelief-setsofallhumanbeingsalwayscontaintruepropositionsonthebasisofwhichitisalwayspossibletorefutetheirfalsebeliefs.Socratescantranscendmereinconsistencybecauseinconsistencywiththetruthisnotjustinconsistency–itisfalsehood.Finally,wearetounderstandPlato’seventualrenunciationoftheelenchos–andwithitthepopulismcharacteristicofauthenticSocraticstreet-philosophy–astheoutcomeofhisdissatisfactionwithSocrates’sstartlingepistemologicalfaith,adissatisfactionissuingeventuallyinthephilosophytranscendingordinarybeliefsandexperienceintimatelyassociatedwithhisname.Vlastos’sproposalhasproducedintensedebateratherthaneasyconsent.Histhesisisintrinsicallyunstable.Theleadingthemeofhislaterbookisthattheproceduresandconvictionsofthe‘real’Socratesandthelater,Platonic‘Socrates’‘contrastassharplywithoneanotheraswithanythirdphilosophyyoucaretomention’(Vlastos1991:46).Oneaspectofthisnearschizophreniaisthetotalepistemologicalnaivety,oratleasttotalsilenceonepistemologicaltopics,ofthe‘real’Socrates:Vlastos’sanalysisofthelogicoftheelenchosisinferredfrom,ratherthanseenin,theactualtextofthedia-588\nANCIENTGREEKPHILOSOPHYlogues.ThushemustattributetoSocrateshighlycontentiousandundefended,ifnotindefensible,philosophicaltenets,whilemaintainingthattheyremainwhollyimplicitinthepracticeoftheiroriginalproponentandonlybreakoutinordertoberejectedinPlato’ssubsequentreflections.Thisscenariostrainscredibility,andhasnotwongeneralsupport;thenatureandimplicationsofSocraticquestioningcontinuetotroublereadersofGreekphilosophy.ItisnoweasytoseewhyIlabelledtheelenchos‘adversarial’:propositionsareassertedbySocrates’spartnerinthecourseofthesearchforadefinition,andarerefutedonthebasisofhisfurtheradmissions.Buthowisthismethodalso‘co-operative’?Anticipatingorreactingtohisinterlocutors’irritationatbeingworsted,Socratesdeclarestimeandagainthatexaminationofhistypeisnotapersonalmatter:forexample,whenintheGorgiasheisabouttoelicitfurtherdamagingadmissions,heinsiststhathedoesthis‘notforyoursake,butfortheargument,sothatitwilladvanceinthefashionbestabletorenderwhatisunderdiscussioncleartous’(Plato,Gorgias,453C).Dialecticalargumentdoesnotfallwithinthescopeofapersonalpronoun;itisneither‘yours’nor‘mine’;rather,itprogressesthroughthephasesofcriticalexami-nationtotheintellectualandmoralbenefitofquestionerandanswereralike.Dialecticisultimatelyforthesakeofknowledge.Ifitappearstoattackorsparetheinterlocutor,thatisamereappearance.Theargumentitselfisnotonlyourchiefbutoursoleconcern:weinteractwithourpartneronlybecauseandinsofarashecontributestotheinvestigation.Bythesametokenwedonotcareaboutourowndialecticalfateassuch,thatis,whetherwhateverfragmentorfigmentoftruthemergingfromthedis-cussionis‘ours’.Truth,onthisSocraticconception,isnotacommodityaccessibleatsomepositionswithinahierarchyattheexpenseoftheoccupantsofotherpositions;allparticipantsinthediscussionsharesuccess–arrivingatthetruth–communally.PerhapsthemostsalientcharacteristicofancientGreeksocietyisthatitwasago-nistic,orcompetitive:whetherinpolitics,artorathletics,Greekscompeted.Intellectuallifewasnoexceptiontothisgeneralization.TheprototypesofwhatwenowdistinguishasWesternphilosophy,science,medicineandevenmathematicswereinventedbytheGreeksnotbydispassionate,disinterestedinvestigationinacalminstitutionalsetting,butratherinanarenaofstrident,oftenhighlypersonal,debate.OfthemanymetaphoricaldescriptionsofphilosophyitselfthatappearinPlato’swritings,thoselikeningittocombatpredominate:philosophyisastruggleinwhichoneopponentisworstedbyanotherwithargumentativeweapons.Thisgladiatorialconceptionsurvivestothisday–witnesstheeasewithwhichcontemporaryphilosophersspeakadmiringlyofa‘knock-downargument’–andislargelyresponsibleforthesuspicion,hardeningintoactivedislike,inwhichphilosophyisnowoftenheld.Theviewisprevalentthatphilosophicalreasonis,togetherwithscience,amereintellectualmanifestationofdis-tinctivelymasculineaggressionconcealingitsviolencebeneaththeideologyofobjec-tivity.ThislargelyFeminist(chapter20),occasionallynon-Western,critiqueofthearroganceofreasondeservestobetreatedwiththeutmostseriousness,andwecannotdobetterthantoturntoSocrates,notforadefinitiveanswer,butforhighlysuggestiveindicationsofwhereananswermightlie.Theelenchosdoesnotabjureaggressiveargu-ment;itdoes,however,insistthatrationalenquiryresidesinaspecialformofdisci-plinedcompetitionaimingatmutualbenefitratherthanatvictimizationforitsownsake.Thereisgoodreasontoquerythetenability,perhapseventhecoherence,of589\nROBERTWARDYSocrates’selenchos,butphilosopherseagertoconfrontthemodernattackontheirratio-nalisticpretensionsstillhaveeverythingtolearnfromhisideal.IhavementionedVlastos’sinsistencethatunderdiscussioninSocraticdialecticaresincerebeliefs;giventheimportanceofthedefinienda,thevirtues,whatareexaminedareaperson’sdeepestcommitments,thatperson’slife,notacademicopinionsdefendedonlyinschool.Thisassertionraisesafinalproblemforus.Vlastosisemphaticallyinter-estedinthehistoricaltruth;heisanythingbutinsensitivetothefactthatinformationaboutSocratesisfilteredtousthroughthemediumofPlato’swritings,amongothers,buthestillavowedlyconceivesofhistaskasthehistoricalexerciseofextractingfactualinformationfromdocumentarysources.ThinkingAboutMethodVlastoschampionsPlatoasthehistorical,notjustthephilosophical,authorityonSocrates.Leavingasidethepersistentcontroversyoverthemeritsofthisjudgement,whatweshouldrecognizeisthatthe‘sincerebeliefs’underpinninginterlocutors’‘lives’inPlato’swritingsarewordsgiventocharactersindialogues.SocratesandPlatotogethershouldencouragethestudenttothinkhardaboutphilosophicalmethod,notnowinthesenseoftheelenchosasopposedtosomeotherparticularmodeofenquiry,butinthesenseoforalasopposedtowritten,aporeticasopposedtodemonstrative,and‘scientific’asopposedto‘literary’.TherealSocrateswrotenothingdown;thePlatonicSocratesisconstantlyengagedinargumentativediscussion,butisnotrepresentedasgoingofftorecordhisnegativeresultsinanysortoftreatise.Doubtlessinlargemeasureasaconse-quenceoftheuniformfailureofcandidatedefinitions,Socraticphilosophyisnotawrittenproduct,butanoralprocess;itissomethingthathappensbetweenpartnersindialectic,notabodyofformulizabledoctrinetobelearntbythepassivereader.ItisalmostimpossibletoexaggeratethesignificanceofthisfeatureofSocraticpractice,whichsur-viveswithmodificationsthroughoutancientphilosophy.ThecoreofGreekphilosophyremainsaliveencounterbetweendialecticiansdebatingathesisaccordingtomoreorlesswell-articulatedrulesofargumentation,andthisundoubtedlyledtoanespeciallysophisticatedsensitivitytoandexploitationofthelogicofself-refutation,mostablyexpoundedbyBurnyeat(1976a,1976b,1990).Thefirst,mostdifficultandmostimpor-tantlessonforanyaspiringstudentisthatphilosophyisproperlyunderstoodnotasabodyofideas,nomatterhowinteresting,butasalivingactivityofargumentundertakenbypractitionerswho,ideallyatleast,sharestandardsofrationaldebateratherthananofficialsetofopinions.ReadingPlato’sdescriptionsofSocratesatworkisnoguaranteethatthislessonwillbelearned,butitcanprovidematchlessinsightintohowphilosophy,asopposedto‘thehistoryofideas’,isactuallydone.2PlatoPlatonicdialogueiswrittenphilosophy,butit,nolessthantheSocraticmodel,challengesunthinkingassumptionsabouthowaphilosophershouldcommunicate.ItisacommonplaceofscholarshipandcultureatlargethatPlatoranksasauniquelygiftedliteraryartistaswellasaphilosopherofgenius;yet,allthebestintentionsnotwithstanding,adisastroustendencypersiststoconceiveofhisdialoguesashard590\nANCIENTGREEKPHILOSOPHYphilosophicalcontenttrickedoutinattractivebutdisposableliteraryornament.Atleastsofarashisownwritingsareconcerned,wehavenoideaofwhatPlatohimselfthoughtaboutanything.Thecharacter‘Socrates’isnomorePlatothanHamletisShakespeare.Thedialoguesconfrontuswithargumentswhoseconstructionwecanobserveinfic-tionaldebatesbetweenfictionalcharactersandwhichshouldthereforeforceustoreacttothemasproductsofafabricatedmise-en-scène.Some‘literary’factorsarepotentiallyasinfluentialasconventionally‘philosophical’ones,andconclusionsarealwaysadhominem,ratherthanofuniversalimport.ThePlatonicdialogueisaninvitationtoargue,notarecordofendorsedpropositions.Itoffersanalternative–devastatinginitsseductiveandirritatingcharm–tothemodeloftheexpositorytreatise,onewhichconstitutesanoutstandingprimafaciecounter-exampletoourowntime’sprevalentdogma:thatphilosophyachievesaproba-tiveforcebeyondfictionbyausterelysacrificingthemeretriciousattractionsofliteraryartifice.Toaconsiderableextent,thepotencyofPlato’schallengetophilosophers’comfortablebutlimitingpresumptionsaboutwhattheyshouldbedoingandhowtheyshouldbedoingit,derivesfromthemethodologicalself-reflectionwithwhichthedialoguesabound.ConsiderthisstrikingexchangefromPlato’sProtagoras(334C–E):Socrates:‘Being,asithappens,aratherforgetfulsortofperson,Protagoras,Itendtoforget,facedwithalengthystatement,theoriginalpointoftheargument.Now,supposeIhappentobehardofhearing:ifyoumeanttoholdaconversationwithme,youwouldthinkitnecessarytospeakmoreloudlythannormal;sonowthatyouarefacedbyamanwithapoormemory,pleasecutyouranswersdownandmakethemshortenoughformetofollow.’Protagoras:‘Whatdoyoumean“makemyanswersshort”?AmItomakethemshorterthanisnecessary?’Socrates:‘Certainlynot.’Protagoras:‘Asshortasisnecessary,then?’Socrates:‘Yes.’Protagoras:‘AmIthentomakemyanswersasshortasIthinknecessary,orasshortasyouthinknecessary?’Socrates’srepresentationofhisdissatisfactionwiththewayhisinterlocutorhasbeengoingonat(whathefeelstobe)evasivelengthasmerelytheupshotofafeeblememoryis,ofcourse,aparticularlyabrasiveinstanceofhiscelebratedirony.Platoisindirectlyinvitingustopondertherelativemeritsofextended,narrativeexpositionandsharp,argumentativedialecticalinterchange.ButwemustnotjumptotheconclusionthatthisisnomorethanacrypticinvitationtoendorseuncriticallytheprocedureoftheSocratic‘hero’,andrejectthatofthe‘sophistical’villain,Protagoras.Protagoras–whoisalsoaPlatoniccharacter–isgivenanexcellentripostebythephilosophicaldramatistwhenheismadetoaskwhoseauthorityarbitrates,whodecidesjusthowmuchisenough.Thelastthingweattentive,perhapsevenargumentative,readersofPlatonicdialoguesarehereencouragedtodoistotakeanythingforgranted.Theclassicdilemmaconfrontingenemiesofcomplacencyishowtheymightattackthetyrannyofunthinkingconventionwithoutthemselves,ifsuccessful,becomingyetfurtherfiguresofbalefulauthority.Platohascertainlybecomesuchafigure;butacarefulreadingofthedialoguesshould,onthecontrary,persuadeusthatheexpendedthewealthofhis591\nROBERTWARDYphilosophicalandliterarygeniusonthemostingeniouseffortstofrustrateourmostlikelyexpectations–thatis,tomakeusthinkphilosophically.Contemporaryscholarsmorereadilyadmittheprofoundconsequencesofthedialogueformthaninthepast,butcontinuetostruggletocometotermswithitsdauntingimplications.HubbardandKarnofsky(1982)isahighlyexceptionalefforttohelpthestudentcatchthereper-cussionsofPlato’sformat;their‘Socraticcommentary’strivestoovercometheself-defeatingdogmatismoftraditionalPlatonicinterpretationbyrestrictingitselfexclusivelytoquestions.3AristotleAswithSocratesandPlato,sotooinAristotelianstudiesitisthecasethatespeciallyimportantrecentworkfocusesonmethodology.Amajorconcernthroughoutthehistoryofphilosophyhasbeenthediscovery,analysisandevaluationofconvictionsdeeplyentrenchedwithinagivensocietyor,inthecaseofespeciallyambitiousphilo-sophicalprojects,heldbyhumanityatlarge.WehavealreadyseenhowSocrates’sdefi-nitionalprojectskirtsthequestionofwhatsortofthingisbeingdefined.Plato’sdevelopmentofSocraticpracticewasananswerwithavengeance:postulationoftheexistenceofunqualified,perfectexemplarsbeyondspaceandtime,theso-calledPlatonicForms.Aristotle,however,pursuesdialecticalinvestigationbyworkingfromwhathecallstheendoxaandthephainomena,whichitwasoncestandardpracticetotranslateas‘commonbeliefs’and‘phenomena’,respectively.But,largelyundertheinfluenceofG.E.L.Owen(1922–82)(seeOwen1986;andsupplementedbyNussbaum1986),itisnowgenerallyagreedthattheserenderingsareunacceptableforreasonscentraltotheproperunderstandingofAristotle’smethod.Endoxaarenotmerelycommonbeliefs,althoughtheymightbewidespread;theyarereputablebeliefs,espousedeitherby‘themany’or‘thewise’.Similarly,phainomenaarenotlimitedto,althoughtheyinclude,observable‘phenomena’;theyare‘appearances’whichembracewhatpeoplearein-clinedtoacceptwithoutprompting,butacrossaspectrummuchbroaderthanthelatter-dayempiricistconnotationsoftheword‘phenomena’mightsuggest.Aristotle’scharacteristicmethodwhenaddressingquestionsinthephilosophyofnatureaswellasinethicsandpoliticsisfirstofalltogathertogethertheseappearancesandreputableopinions.Oftenwhencollectedtheyseemtoconflict,eithersuperficiallyorinsomemoreseriousfashion.This(ostensible)conflictiswhathecallsanaporia,apuzzle.Forhimtheroleofthephilosopheristotakeonnotanyrandomproblem,butpuzzlesgeneratedinthiswaybydisharmonyinsuggestionswhichdeserveahighdegreeofinitialcredence.Thephilosopher’sjobisessentiallyconciliatory,toexplainthattheconflictisamatterofmereappearancewhereverthisisthecase,and,wherethedifficultyrunsdeeper,toadjudicatebetweentheopposingviews,effectingtheminimaladjustmentsnecessarytoresolvetheconflict.InAristoteliandialectic,Greekphilosophyattainedadegreeofanalyticalsubtletyandbreadthofscopewhichhasbeenequalledbutneversurpassed.Yetthemethodhastroublingimplications.Theendoxaalsoincludewhatwe,atleast,areinclinedtoviewaslinguisticphenomena:Aristotledrawssweepingconclusionsfrom‘whatwesay’–thatis,whatordinarypeoplesayinGreek.ThescepticalmodernreactionistoaskhowhisclaimsaboutGreekmorphology,syntaxandusagesupportsubstantialpropositions592\nANCIENTGREEKPHILOSOPHYvalidbeyondtheconfinesofhislinguisticcommunity,eveniftheyarecorrect.Furthermore,thefearthatamethodotherwiseadmirablyinclinedtotreatuntutoredopinionwithconsiderablerespectmightyetspilloverintoideologicaljustificationforperniciousbeliefsandpracticesfindsamplejustificationinAristotle’snotorioustheo-riesthatcertainpeople–slaves–are‘livinginstrumentalproperty’(Aristotle,Politics,I.2)andthatthefemalesexisa‘naturaldeformation’ofthemalestandard,definedbyincapacity(Aristotle,DeGenerationeAnimalium,A.20–adevastatingcritiqueistobefoundinLear1988).ThereisnocomparisonbetweenAristotle’smethodologicalself-consciousnessandSocrates’ssilenceabouthowtheelenchosworks,butpreciselythesameproblemwhichimpelledVlastostoattributetoSocratesanundefendedconvic-tionthateveryone’sbelief-setcontainssometruebeliefsaboutvirtuere-emergeswhenwereflectonAristotle’sphilosophicalpractice:howcanhebeconfidentthathisdialec-ticeverachievesobjectivetruth,ratherthanjustcoherent(andperhapscoherentlyfalse)belief?IrwinonAristotle’sDialecticIrwin(1988)proposesanevolutionaryanswertotheproblem.HearguesthatAristoteliandialecticpassesthroughtwostages.Aninitial‘weak’formdoesindeedaspiretonomorethancoherence,butitissupersededby‘strong’dialectic,whichtriestoestab-lishresultsthatare‘true’bythestandardsofmetaphysicalrealismandvalidbyrefer-encetotheworld,notourattitudes.ForIrwin,thisevolutionispossiblebecause‘strong’dialecticproceedsfromaprivilegedsubsetofbeliefs,thosewhichcannotberejectedonpainofforfeitingrationality,suchastheprincipleofnon-contradiction.Thisprinciplerejectsthepossibilityofapropositionanditsnegationbothbeingtrue.Irwin’sargumenthasexcitedmuchinterest,butisvulnerableontwomajorfronts.Firstly,evenifso-called‘strong’dialecticcanbeisolated,itishardtoseehowanythingbeyondaveryfewfun-damentalprinciplescouldmeetitsforbiddingstandardsandthusyield‘strong’results;sinceAristotle’sphilosophycoverssomuchmorethanthefoundationsofmetaphysicsandepistemology,mostofhisdialecticalworkwouldremain‘weak’.Secondly,Irwin’sreadingofAristotleisunashamedlyKantian(chapter32):hissolutionattributestohimatypeoftranscendentalargumentwhichgainedcurrencyinthehistoryofphilosophylongaftertheGreekperiod.Furthermore,transcendentalargumentsthemselvesmightfacethesameproblemofmovingbeyondcoherencetotruth.AnalternativeapproachsearchesforananswerconstructedexclusivelyfromAristotelianmaterials(Wardy1991).TheAristoteliannaturalworldisteleological:thatis,many,perhapsall,characteristicsofthestructureandbehaviourofnaturalkindsaretobeexplainedintermsofwhatisgoodforthem.AlthoughtheimportofAristotle’sargumentsforteleologyisamatterfordebate(Sedley1991;Wardy1993),itisperfectlyclearthatAristotelianteleologicalexplanationsareirreducible.Theycanoftenbesupplementedbycausalaccountswhichcite,say,materialprocesses,buttheycannotbereducedtonon-teleologicalexplanation.Ateleologicalexplanationconsti-tutesaperfectlylegitimatereason,completeinitself,forsomestateofaffairstobeasitis.ThealternativeinterpretationofAristotle’sdialecticissimplyanapplicationofthisglobalteleologytothehumanspecies.ForAristotle,weareessentiallyknowers;a593\nROBERTWARDYhumanbeingisarationalanimalendowedwithinnatespeculativecuriosity.SinceAristotle’steleologicalphilosophydictatesthatnoanimalkindcanberadicallyfrus-tratedinfulfillingitsnature,itfollowsthattheworldmustbeaccessibletohumanenquiry.Justastheloweranimalsareobligedtostruggletosurvive,sotoowecannotuncriticallyacceptthephainomenaandendoxawithoutmodification–buttherealtruthistheretobefound.Aristotle’sconfidenceintheutilityofhisdialecticalmethodasaninstrumentfordiscoveringabsolutetruthreliesonhisphilosophicalbiology.Sincehisbiologyinturnisderivedfromanexerciseinthatverydialectic,thepurported‘justifica-tion’seemstobegthequestion.Butitmightbethatdissatisfactionwiththiscircularityistheproductofimposingpost-KantianepistemologicalconcernsonAristotle,withtheinevitableresultthathisthoughtiseitherseverelydistortedorfoundseriouslywanting.TheargumentbetweenthesecandidatesolutionstotheriddleofAristotle’smethodremainsunresolved.Quiteapartfromtheobviousintrinsicinterestofcomprehendinghisprocedure,thedebateisaparticularlyaptillustrationwithinGreekphilosophyofaninescapabledilemmainallphilosophicalhistoriography.OnemightfeelthattheKantianreadingisanachronistic,whiletheteleologicalexplanationlandsAristotlewithastrangepositionthatistotallyinaccessibletocontemporaryphilosophers.Allhistoricalstudyentailscommitmentstakingtheinvestigatorfarbeyondmundanefactualreportage.Inthehistoryofphilosophythiscommitmentistodoingphilosophywhendoingthehistoryofphilosophy.Ifphilosophersaretolearnfromtheirpast,thentheymustabandonhopeofaneasyformuladictatinginterpretationswhichsucceedinengagingwitholdargumentswhileavoidinganachronisticdeformation.Oneofthemostvaluablegiftsthehistoryofphilosophyhastoofferuscanbeaglimpseofhowdifferentphilosophywasatothertimes.By‘different’Imeannotjustthatearlierthinkerspropoundedarichvarietyofdoctrinesnowunfamiliar,althoughthereisnodoubtthathistoricalexpertisemakesavailablearangeofpotentialideaseasilyover-lookedbytoonarrowaconcentrationoncontemporaryviews.Themorechallengingsenseof‘difference’accessiblethroughhistoryistherealizationthattheverytermsinwhichtheapparentlyconstantproblemsofphilosophyareopposed,thelogicalandheuristicconsiderationstowhichthinkersaresensitive,thestandardstowhichtheysubmittheirarguments,mightthemselvesbesusceptibletohistoricalchange(notnec-essarilyprogress).ThedebateoverAristotelianmethodexemplifiesthismostfruitfuldilemma,andtheexpansionofphilosophicalperspectiveitencourages.4HellenisticMethodologyIfonehadtoselectthesinglemostimportantdevelopmentinworkonancientGreekphilosophyinrecenttimes,itwoulddoubtlessbetherenaissanceinHellenisticstudies.UntilafewyearsagotheprevailingattitudetowardstheHellenisticperiodevenamongancientGreekspecialistswasdismissive.TheusualviewleftPlatoandAristotleinsplen-didisolation,thefirstfiguresinaparadeofgreatphilosophersalltoooftenfollowedimmediatelybyacontingentfromtheseventeenthcentury.AlthoughlaterGreekphilosophershadtheirableadvocates,therelativeinaccessibilityofthescatteredsourcematerials,alltoooftenintheformofbiasedindirecttestimony,exacerbatedtheneglectintowhichHellenisticphilosophyhadlapsedsincethetimeoftheEnlightenment.The594\nANCIENTGREEKPHILOSOPHYpublicationofamajornewstudy(LongandSedley1987)utterlytransformedthesituation.TheirinvaluablecollectionofthesourcesbothintheoriginallanguagesandintranslationwithextensivecommentarypermitsallphilosopherstopenetratetherelativelyunknownterritoryoftheHellenisticperiodwithassurance.AnimmediateadvantagefromsomefamiliaritywithHellenisticphilosophyisproofthatonehallowednotion,thatofthe‘timeless’problemsofphilosophy,crumblesoninspection.Tonamebutthemostobviousexamples,Hellenisticphilosophersinventedtwoissues,determinismandscepticism,whichsocametodominatelaterpreoccupa-tionsthattheunwaryaretemptedintosupposingthatthesetopicsareeternallypresentinphilosophy.Thismistakeatonceinstilsfalseexpectationsofwhatcanbereasonablyexpectedwhenevaluatingearlierperiods–forexample,bewilderedsurpriseattheinca-pacityofAristotle’stheoryofvoluntarinesstocopewithdeterminism,orgravemis-construalsofPlato’sepistemologicalconcepts–andprohibitsrecognitionofthesheercreativityoftheHellenisticcontribution.But,asonsomanyotherpointswehavecanvassed,thereisahealthylackofcon-sensusononeparticularlyimportantquestion.Inatleastsomerespects,theGreekchapterinphilosophicalhistoryissurprisinglyunified.TheperiodstretchingfromthePresocraticstothelaterNeoplatonistsismorethanamillennium,buteventheverylatestworkfallingwithinitdemonstratesanacuteappreciationofinheritedinsightsandalargelysharedproblematic.TheremarkableforceofthiscoherencebecomesapparentwhenwereflectthatAquinas(1224–74)isclosertousintimethanfiguresatthetwoendsoftheGreekmillenniumweretooneanother.Thisintellectualcohe-sionlendsGreekphilosophymuchofitsappeal,butreadilyshadesintoahighlyques-tionablemethodologicalprinciple:namely,thatancientphilosophersreacttotheworkoftheirmoreorlessimmediatepredecessorsinobediencetooursenseofchronologyandimportance.ApplicationofthisprinciplehasproducedtheassumptionthatsinceAristotleisatoweringfigureforus,hemustsimilarlyhavedominatedtheformationofHellenisticphilosophy–whichisindeedaguidingassumptionofmuchrecentschol-arship.CentralHellenisticdoctrineshavebeenregularlyinterpretedasproductsofreflectiononAristotelianinsightsoraseffortstocircumventsomeofhiscriticisms.ButanalternativeschoolmaintainsthatAristotelianphilosophywentintorapideclipseshortlyafterhisdeathanddidnotregainprominenceuntillongaftertheformativephaseofHellenisticphilosophy(Sandbach1985).LongandSedleythemselvesdifferoverthisissue,onlytheformerroutinelyexplainingfeaturesofHellenisticpositionsbyadducinganAristotelianforerunner.WhatdidtheHellenisticphilosophershavetosayaboutourmajorthemeregardingmethod,therecognition,assessmentandexploitationofnon-philosophicalbelief?IstressedthattranslationofthecardinaltermendoxoninAristotleas‘commonbelief’isagraveerror;reputableopinionishisdialecticalmaterial.Hellenisticmethodworkedwithaverydifferentsetofassumptions:itsdialecticdidindeedoperateongeneralopinions,designated‘commonpreconceptions’.Theleadingschoolsactiveinthisperiodsharedtheconcept‘commonpreconception’,butdifferedsharplyoveritssigni-ficationand,ofcourse,whatthepreconceptionsactuallywere.TheStoics,ononeside,arguedthatNature,identifiedasadivine,providentialreasoninformingtheentirecosmos,leadsustoformcertainnotionsonwhichwemightultimatelybuildacom-prehensivephilosophytoguideusthrougheveryaspectoflife.Theirarch-opponents,595\nROBERTWARDYtheEpicureans,deniedtheexistenceofanysuchdivinerationality,whilearguingthathumanbeingsnaturallyconstructreliablepreconceptionsonthebasisofsenseperception.Althoughtheschoolsdisagreedaboutthemechanismresponsibleforournaturalintellectualendowment,theylargelyagreedonthereality,reliabilityandphilosophicalsignificanceofourpreconceptions.BothcampsalsopropoundedNaturalistic(pp.805–6)ethicaltheories:lifehasagoalwhichisdeterminedbynatureitself,andsuccessinlivingdependsentirelyonperceivingandconformingtotheconditionswhichmightpermitustoachievethegoodaccordingtonature.FortheEpicureans,thisgoalwaspleasureandtheabsenceofpain;fortheStoics,itwas‘livingaccordingtoreason’.Howdo‘commonpreconceptions’haveabearingontheargumentbetweenEpicureansandStoicsconcerningtherealgoaloflife?Neitherschoolasserted,implau-sibly,thatpeopleapprehendedtheirfavoureddoctrineinallitsdetailasamatterofnaturallyacquiredknowledge:thatextremepositioncouldnotevenaccommodatetheexistenceofwrong-headedStoics(orEpicureans)withoutembarrassment.Instead,preconceptionsweremadetoprovideafundamentallysoundinitialindicationofthetruthwhichis,however,vulnerabletosubsequentcorruption,andrequireselaborationanddefencefrom,ofcourse,thecorrectphilosophy.Theresultwasabraceoffasci-natingdevelopmentsindialecticaltechnique.TheHellenisticthinkerswerethefirstphilosopherstoadducetheevidenceofinfantilebehaviourinsupportoftheirdoctrines:Epicureansclaimedthatbabiesnaturallyseekpleasureandshunpain,Stoicsthattheystriveforself-preservationandavoidextinction(theseargumentsreceivedmagisterialattentionfromBrunschwig1986).Secondly,thecommonallegiancetoaformofinnatistepistemologyledproponentsofrivalphilosophiestocriticizehostilecitationofostensible‘preconceptions’foravarietyofintriguingreasons.ItiseasytoadoptajaundicedattitudetowardssuchHellenisticdebate.Foranyissue,itmightseemtousthatifitappearedtofavourEpicureanism,thenitschampionswouldsaythatitisatrustworthyexpressionofournaturalintellectualequipment,incorrigi-blewithinthelimitsofitsonlypartialarticulationofthetruth;whiletheStoicswouldprotestthat,althoughtherecertainlyarereliableconceptionsrealizablebygraceofprovidentialreasonandharmonizingwithStoicism,inthisinstancesocialcontamina-tionhasinsomewaybastardizedtheopinion.If,ontheotherhand,theStoicsclaimedthesupportofacommonpreconception,theEpicureanscouldreadilyreplythatinthiscasespontaneous,trueinstincthasbeenadulteratedbytheadmixtureofunhealthysocialconditioning.Theever-presentavailabilityofthechargeofcorruption-by-societyensuredthatthetestimonyofcommonpreconceptionscouldnevercontroltheexcessesofinter-schooldialectic,whichthreatenstodegenerateintomerepolemic.ItisclearfromBrunschwig’sstudythatalthoughHellenisticphilosophersbelievedthathumannatureismanifestatapre-culturalage,theyevincedlittleinclinationtoengageinthescrupulousobservationofinfantsevidentlycalledforbythisbelief;andthesearchforsuchdatawasanythingbutsystematic.Nevertheless,weshouldavoidtooharshajudgement.ThedoctrineofpreconceptionencouragedtheStoicsinparticulartoturntonon-philosophicalwitnesses,especiallyliteraryones,whichmadetheirphilosophyreceptivetoanenrichingdiversityofinfluences.IfHellenisticphilosophyisvulnerabletoheavycriticismoftheopportunitiesitoftenmissedbymanipulating,ratherthanreactingcircumspectlyto,formulationsof596\nANCIENTGREEKPHILOSOPHYcommonbelief,wemustrememberthatthisphilosophicalproblemdoeshaveagoodclaimtofigureonacautiouslistof‘perennial’challengestophilosophers.Allbutthemostobstinatelyscientisticandincreasinglyold-fashionedmodernthinkerswouldconcedethatphilosophycannotavoidderivingsubstantialimpetusfromsourcesclearlyrelatedtobothAristotelianendoxaandHellenisticcommonpreconceptions.We,ofcourse,subscribeneithertotheproto-KantianismdubiouslyattributedtoAristotle,nortotheepistemicteleologytowhichheundoubtedlysubscribed,noragaintotheHellenisticphilosophers’ambitiousepistemicnaturalism.Butwhetherornotwegosofarastocharacterizeouractivityasconceptualorlinguisticanalysis,thefactremainsthatweareoftenintheuneasypositionofarmchairtheoreticiansobligedtopronounceonwhatpeople‘really’believe.IfcontemporaryphilosophersdoratherbetterthantheirpioneeringHellenisticancestorsinassimilatingthediscoveriesofamultiplicityofempiricaldisciplines,theystillfacetheuncomfortabletaskofdistinguishinginawell-motivated,convincingfashionbetweeninnocentbeliefandtheconstructionputonitbyphilosophy.Iwarnedattheoutsetthatmytreatmentofancientdialecticwouldbeobligedtoomiteventhemention,letalonethesustaineddiscussion,ofcrucialepisodesinthehistoryofGreekphilosophicalmethodology.ButitwouldbeverywrongtoleavethetopicwithoutencouragingthereadertoconnectpositiveHellenisticepistemicnatu-ralismwiththeperiod’snegativeepistemologicalinvention,scepticism.Theprimarychallengeconfrontingancientscepticswaswhetheritwaspossibleconsistentlytomaintainthatalltheschoolsofdogmaticphilosophyhadfailedtoreachthetruth,oratleasthadfailedtoprovethattheyhadreachedit,whileneverthelessavoidinganironicallyself-refutingdogmaticcommitmenttothetruthofscepticismitself.Further,thevigorousdialecticoftheHellenisticerahelpedtocreateandfeedanotherimpor-tantdevelopmentinthephilosophicaldeploymentofcommonbeliefs,adevelopmentwhichhasalsoonlyjustbeguntoachievethescholarlyrespectitdeserves.Philosophersoftheso-calledscepticalschools,Academicsandneo-Pyrrhonists,werenotonlyadeptinpittingordinarybeliefsagainstthephilosophicaldoctrinesoftheir‘dogmatist’rivals–theStoic,Epicurean,PeripateticorPlatonist–andinallegingintolerableinconsis-tenciesbetweenwhatopposingschoolsclaimedas‘natural’beliefs,tendenciesorprac-tices.Theyalsodisplayedtheordinarybeliefsandpracticesofdifferentplaces,timesandculturesasdifferent,butasequallyvalid–orequallyinvalid.Allsuchmanipulationsofcommonconceptionswereintendedtocontributetothedesiredscepticalend:thesuspensionofbeliefinthefaceofasyetunresolved,andperhapsirresolvable,univer-salconflictsbetweenideas,beliefs,perceptionsandarguments.Oneofthemostinter-estingcurrentdebatesaboutancientscepticismpreciselyconcernsthequestionswhethersuchsuspensionextendstoallbeliefs,andwhether,ifso,anyhumanlife,letaloneadesirableone,ispossibleinsuchastate(Burnyeat1980;Frede1984).TheaimofthislimitedexercisehasbeentodemonstratehowevenanarrowlyfocusedsurveyofsomeaspectsofthephilosophicalmethodsofSocrates,Plato,AristotleandtheHellenisticschoolsimmediatelyrevealsaprovocativemixtureofargu-mentandassumptionabouthowphilosophycouldandshouldbeconducted;theexer-cisecouldeasilyberepeatedagainandagainformanyotheressentialtopics.NophilosophercanaffordtoignorethehistoryofGreekphilosophy;but,thenagain,nophilosopherwouldwantto.597\nROBERTWARDYFurtherReadingBarnes(1980)andBrunschwig(1980)makefascinatingcompanion-pieces,whileBrunschwig(1986)isamodelofphilosophicallyilluminatinghistoricalreconstruction.TheworksofBurnyeat(1976a,1976b,1980,1990)haveallbeendeservedlyhailedasmasterpieces(thelastisquitepossiblythefinestmoderncommentaryonaPlatonicdialogue).Frede(1974)remainstheclassicstudyofStoiclogic.Lear(1988)isthebestofthemanysingle-volumeintroductionstoAristotleforphilosophers.Nostudentshouldbeunacquaintedwiththeuniformlybrilliant,unfailinglyprovocativeessayscollectedinOwen(1986).ReferencesPlatoandAristotleTherearemanyvaluabletranslationsofPlatoandAristotle,butreadersmightfindithelpfultousetwostandardcollections:Aristotle1984:TheCompleteWorksofAristotle,2vols(editedbyJ.Barnes).Princeton,NJ:PrincetonUniversityPress.Plato1961:TheCollectedDialoguesofPlato(editedbyE.HamiltonandH.Cairns).Princeton,NJ:PrincetonUniversityPress.Inaddition,OxfordUniversityPresshaspublishedtwoseriesoftranslationswithphilosophicalcommentary:Aristotle,ClarendonAristotle1975–:Oxford:OxfordUniversityPress.Plato,ClarendonPlato1973–:Oxford:OxfordUniversityPress.OtherwritersAnnas,J.andBarnes,J.1985:TheModesofScepticism:AncientTextsandModernInterpretations.Cambridge:CambridgeUniversityPress.Barnes,J.1975:Aristotle’sPosteriorAnalytics,translatedwithnotes.Oxford:ClarendonPress.——1980:ProofDestroyed.InM.Schofield,M.F.BurnyeatandJ.Barnes(eds)DoubtandDogmatism:StudiesinHellenisticEpistemology.Oxford:ClarendonPress.Brunschwig,J.1980:ProofDefined.InM.Schofield,M.F.BurnyeatandJ.Barnes(eds)DoubtandDogmatism:StudiesinHellenisticEpistemology.Oxford:ClarendonPress.——1986:TheCradleArgumentinEpicureanismandStoicism.InM.SchofieldandG.Striker(eds)TheNormsofNature:StudiesinHellenisticEthics.Cambridge:CambridgeUniversityPress.Burnyeat,M.F.1976a:ProtagorasandSelf-refutationinPlato’sTheaetetus.PhilosophicalReview,85,172–95.——1976b:ProtagorasandSelf-refutationinLaterGreekPhilosophy.PhilosophicalReview,85,44–69.——1980:CantheScepticLiveHisScepticism?InM.Schofield,M.F.BurnyeatandJ.Barnes(eds)DoubtandDogmatism:StudiesinHellenisticEpistemology.Oxford:ClarendonPress.——1990:TheTheaetetusofPlato(witharevisedtranslationbyM.J.Levett).Cambridge:HackettPublishing.Denyer,N.C.1991:Language,ThoughtandFalsehoodinAncientGreekPhilosophy.London:Routledge.Frede,M.1974:DieStoischeLogik.Göttingen:VandenhoeckandRuprecht.——1984:TheSceptic’sTwoKindsofAssentandtheQuestionofthePossibilityofKnowledge.InR.Rorty,J.B.SchneewindandQ.Skinner(eds)PhilosophyinHistory:EssaysontheHistoriographyofPhilosophy.Cambridge:CambridgeUniversityPress.598\nANCIENTGREEKPHILOSOPHYHubbard,B.A.F.andKarnofsky,E.S.1982:Plato’sProtagoras:ASocraticCommentary.London:Duckworth.Irwin,T.H.1988:Aristotle’sFirstPrinciples.Oxford:ClarendonPress.Lear,J.1980:AristotleandLogicalTheory.Cambridge:CambridgeUniversityPress.——1988:Aristotle:TheDesiretoUnderstand.Cambridge:CambridgeUniversityPress.Lloyd,G.E.R.1979:Magic,ReasonandExperience:StudiesintheOriginsandDevelopmentofGreekScience.Cambridge:CambridgeUniversityPress.Long,A.A.andSedley,D.N.1987:TheHellenisticPhilosophers:TranslationsofthePrincipalSources,withPhilosophicalCommentary(vol.1)andGreekandLatinTextswithNotesandBibliography(vol.2).Cambridge:CambridgeUniversityPress.MacIntyre,A.1981:AfterVirtue.London:Duckworth.Nussbaum,M.C.1986:SavingAristotle’sAppearances.InTheFragilityofGoodness:LuckandEthicsinGreekTragedyandPhilosophy.Cambridge:CambridgeUniversityPress.Owen,G.E.L.1986:Tithenaitaphainomena.InLogic,ScienceandDialectic:CollectedPapersinGreekPhilosophy(editedbyM.C.Nussbaum).London:Duckworth.Sandbach,F.H.1985:AristotleandtheStoics.Cambridge:CambridgePhilologicalSociety.Sedley,D.N.1991:IsAristotle’sTeleologyAnthropocentric?Phronesis,36,179–96.Sorabji,R.(ed.)1987–:TheAncientCommentatorsonAristotle.London:Duckworth.Vlastos,G.1983:TheSocraticElenchus.OxfordStudiesinAncientPhilosophy,1,27–58.——1991:Socrates,IronistandMoralPhilosopher.Cambridge:CambridgeUniversityPress.Wardy,R.1991:TranscendentalDialectic.Phronesis,36,88–106.——1993:AristotelianRainfallortheLoreofAverages.Phronesis,38.Williams,B.1985:EthicsandtheLimitsofPhilosophy.London:Fontana.DiscussionQuestions1DophilosopherstodayhaveanyreasontostudyancientGreekphilosophy?2HowdidGreekphilosophyhelptoformthedistinctivegoalsandmethodsofphilosophy?3WhatistheSocraticmethodinphilosophyandwhyisitimportant?4Candisciplinedargumentreachethicaltruths?5Howshouldphilosophyberelatedtothecultureatlarge?6Inwhatsensecouldaphilosopherbea‘gadfly’?7Howcanwedecidewhetherphilosophyisbestconductedindialogue?8Inordertoachievevirtueinourlife,mustweunderstandwhatvirtueis?9CantheSocraticmethodreachbeyondconsistencytotruth?10IfPlatorejectedtheSocraticelenchus,whydidhedoso?11Whatisthedifferencebetweenphilosophyasadisciplinedcompetitionaimingatmutualbenefitandphilosophyasindividualpursuit?12Shouldeitheroralphilosophyorwrittenphilosophyhavepriorityovertheother?13Inwhatsenseshouldareaderofphilosophybeactive?14Whatdistinguishesphilosophy(andthehistoryofphilosophy)fromthehistoryofideas?15Isthereanyreasonforphilosophytostartfromreputablebeliefsandfromwhatpeopleareinclinedtoaccept?16CanAristotle’sdialecticeverachieveobjectivetruth,ratherthanmerecoherence?599\nROBERTWARDY17Canweagreethatwehaveknowledgebecauseweareessentiallyknowers?18Are‘reputableopinions’or‘commonpreconceptions’abetterstarting-pointforphilosophicalargument?19DidtheStoicsortheEpicureansprovideabetterbasisfordeterminingwhatareourcommonpreconceptions?20DidtheHellenisticuseofcommonpreconceptionsrequireanepistemicnaturalism?21Caninfantilebehaviourtellusanythingimportantaboutphilosophicaldoctrines?22Howcanweknowwhendialecticdegeneratesintomerepolemic?23Howcandiscoveriesinotherdisciplineshaveconsequencesforphilosophy?24Ifscepticismistrue,isitself-refuting?25Aretheordinarybeliefsandpracticesofdifferentplaces,timesandculturesequallyvalidorequallyinvalid?26Whatisthescopeofthesuspensionofbelieflicensedbyscepticism?600\n23PlatoandAristotleLESLEYBROWNChapter22highlightsissuesofphilosophicalmethodinthetriumvirateofGreekphilosophers:Socrates(whowrotenothingbutwhosetthephilosophicalagenda),Plato(c.429–347BC)andAristotle(384–322BC).RobertWardyremindsreadersthatPlato’sworksarewrittenasdialogues,fictionaldebates,andrightlycautionsagainstatoo-readyassumptionthatwecandiscernPlato’sviewsinthedictaofthefictionalSocrateswhoistheleadingspeakerinalmostallthedialogues.Yetphilosophersthroughtheageshave–rightlyorwrongly–thoughtitpropertotrytodiscerninthesedramaticfictionsPlato’sownfavouredargumentsandconclu-sions.InsupporttheycanclaimtheprecedentsetbyAristotle,whocametoAthensfromStagirainnorthernGreeceandstudiedfortwentyyearsunderPlatointheAcademybeforefoundinghisownschoolofphilosophy.Aristotle’sworks–bestthoughtofaslectures,somemorepolishedthanothers–betraynoqualmsininterpretingPlato’swritingsascontainingthesesacceptedandarguedforbyPlato,onesAristotleoftencriticizesandsometimespraises.Insimilarlyrashvein,thischapterextractssomethemeswhicharecommontoPlatoandAristotle’sphilosophy,oftenonesonwhichAristotleoffersamoreorlessdirectresponsetowhathetookPlatotobearguing,andwhichhavesetamajorpartofthephilosophicalagendainsubsequentWesternphilosophy.Butimportantareasoftheirwritingsareomitted,includingthecontribu-tionbyboth,butespeciallybyAristotle,tologicandphilosophyoflanguage.Sincechapter22discussesmethod,thischaptersayslittleabouttherichnessofdialecticalargumentthateachphilosopher,inhisdifferentway,presents,andthatissuchanimportant,andattractive,featureoftheirphilosophizing.1Plato1.1KnowledgeandrealityKnowledgeisrecollection:soSocratesarguesinseveraldialogues.ThethesisislaunchedinaremarkablesceneinPlato’sdialogue,Meno,aworkinwhichSocratesandtheself-assuredMenodiscusswhatvirtueisandwhetheritisteachable.Toshow\nLESLEYBROWNMenothat,despitemanyfalsestarts,itispossibletoinquireintowhatvirtueis,Socratessummonsayounguneducatedslaveandquestionshimaboutaproblemingeometry,aproblemtheboyhasevidentlyneverencountered.ThankstoSocrates’squestioningandindeedprompting,theslavecomestoseethesolution,whereuponSocratescon-vincesMenothat,sincetheboyhadnotbeentaughtgeometry,hewasrecollectingthecorrectanswers.Indeed,theslave’ssoul–arguesSocrates–knowsthesetruthsforalltime,hencemustbeimmortal.ThoughtheconclusionsPlatodrawsfromthisepisode–thatoursoulscanrecollectinnateknowledge,andareimmortal–aredaringandnotultimatelyconvincing,wefindherePlatoexpoundingamatterofcrucialimportance:one’sabilityto‘drawouttruthsfromwithinoneself’.SocratesinsiststhattheknowledgeheandMenoseek,whatthenatureofvirtueis,nolessthanmathematicaltruthsasintheexperiment,istobegainedbythis‘inwardinquiry’,labelledrecollection.Theemphasisonwhatcametobecalledaprioriknowledge,togetherwiththeideathatsuchknowledge,intheincarnatestate,isrecollectedfromapreviousdiscarnateexistence,becomesevenmorepromi-nentinthePhaedo(and,withouttheexplicitclaimofrecollection,intheRepublic).Butinthesedialogues,thispictureaccordingtowhichrecollectedknowledgeistobehadbypurethought,aidedbydialecticalquestioning,isaccompaniedbyafurtherthesis:Plato’sfamoustheoryofForms.Leibniz(chapter27)discernedintheMeno’stheoryaprecursorofhisownbeliefinlatentinnateknowledgepresentinallhumanbeingsandactualizedineverydaycognitiveactivity.Recollection,onthis‘optimistic’view,isopentoall,especiallywiththeprobinghelpofaSocrates.ButisthisPlato’svieworisit,asScott(1995)hasargued,amore‘pessimistic’accountaccordingtowhichrecollectingisthepreserveofphilosophersalonethroughhardintellectuallabour?TheMeno’schoiceofaslave-boyasthesubjectofrecollectionsuggestsLeibniz’soptimisticview,butScott’sviewhasmoreplausibilityforthePhaedo,wheretheobjects‘recollected’aretheForms.TheTheoryofFormsWhatarePlato’sForms?Weapproachthisbylookingfirstathowtheinquiryproceedsinsomedialoguesfromwhichthefull-blowntheoryofFormsisabsent.Inthese,Socratesseeksknowledgeofwhatqualitiessuchaspiety,self-controlandcourageare.Thisisthefamous‘WhatisX?’question(seechapter22).Attemptstoanswerthequestionfailinvariousinstructiveways,andthereaderlearns,fromthewaySocratestreatsthefailedattempts,whatisrequiredofacorrectanswer.ItmustencompassallandonlycasesofX(forinstance,ofcourage),anditmustshowXtobeasinglequality,andonewhichexplainswhyallthecasesofX(the‘manyX’s’)areX.Finally,itshouldhelptheinquirertosettle,indisputedcases,whethersomethingisorisnotX.Inanewphaseofthesearchintoknowledgeanditsobjects,Plato’sPhaedoandRepublicgofurther,anddeveloptheideathattoknowwhat,forexample,beautyis,onemustgettoknowtheFormbeauty,labelled‘thebeautifulitself’or‘thatwhichisreallybeautiful’.This‘Xitself’,theoneX,iscontrastedwiththemanyXswhicharechangeable,areliabletobetheoppositeofXaswellasX(uglyaswellasbeautiful,forinstance)and,unliketheForm,areknownwiththehelpofthesenses.TheForms,asSocratesarguesforthem,arerealbutimmaterialobjects,knownbythesoul’sreasoning,notbysight,hearingoranyofthebodilysenses.602\nPLATOANDARISTOTLETheyexistindependentlyofthematerialworldandofthepersonswhoaspiretoknowthem.AndtheFormofthebeautifulisitselfbeautifulinamannerevidentlydifferentfromthatofthe‘manybeautifuls’whicharebeautifulthroughtheirrelationtotheForm,specifically,throughwhatiscalled‘participationinthebeautifulitself’.Beforewelookmorecloselyatthetheory,awarning.Thetitle‘Plato’stheoryofIdeas’(oftenusedasanalternativeto‘theoryofForms’)isbestavoided.Thoughtheyareimmaterial,Plato’sFormsarenotideasinthesensemadefamousbyseventeenth-centuryandeighteenth-centuryphilosophy,noristhetheoryidealistinBerkeley’s(chapter30)sense.UnlikeBerkeleyanideas,Formsdonotdependfortheirexistenceonminds,notevenonthedivinemind,althoughsomelaterthinkers(so-called‘neo-Platonists’)developedthetheoryinthatdirection.But,asweshallsee,thoughFormsaredecidedlynotmind-dependententities,Formsandmindsarepeculiarlyfittedtooneanother.TheForms,then,aretoserveasobjectsofknowledgeandunderstanding.Buttheyhaveseveralotherroles:ontological,semanticandethical,aswellasepistemological.TocomplementandunderpintheirEpistemological(chapter1)role,theOntological(chapter2)statusofFormsiscrucial.Formsplayafundamentalroleinanaccountofhowthingsare–even,somehow,inanaccountofhowthingsintheempiricalworldare,sincetheX’sintheempiricalworldareXbyparticipationintheFormX.Knowledge,asSocratesofteninsists,isofbeing,ofwhattrulyandfullyis,andthatmeans:ofForms.ButhowshouldweunderstandtheclaimthatonlyForms‘trulyandfully’are,orhavebeing?Doesthismeanbeingasexistence,orbeingasbeingsome-thingorother?SomehaveaccusedPlatoofconfusionhere,whileothers(suchasVlastos)insistthatthesuperiorbeingoftheFormsshouldnotbereadasasuperiorkindofexistence,butratherasuperior(because‘cognitivelyreliable’)wayofbeingX–forexample,ofbeingbeautiful.Perhapstoposethequestionthusistofoistontoancientthoughtananachronisticdistinction.InelaboratingthedifferentrolesofForms,itisusefultoemployArmstrong’sdis-tinctionbetweentwoconceptionsofuniversals,realistandsemantic.Ontherealistconception,universalsarepositedasexplanatoryentities(usuallyasproperties),suchthatafullaccountoftheworldwillmakereferencetothem.Thediscussioninthepre-viousparagraphmakesitclearthataPlatonicFormissomekindofuniversalontherealistconception:universalbecausetheFormFisthatbyvirtueofwhichallthemanyF’sareF,andrealistbecauseitispartofthestructureoftruereality.The‘semantic’conceptionofuniversals,bycontrast,seesthemasthemeaningsofgeneralterms,ofexpressionssuchas‘beautiful’or‘circular’or‘man’or‘horse’,andthetheoryseemstoaccordFormsthisrolealso.ThisisshownbyafamouspassageinRepublic10(596),whereSocratesremarksthat‘wecustomarilyhypothesizeasingleForminconnectionwitheachofthemanythingstowhichweapplythesamename’.Butthissemanticrolemustbesubordinatetotheontologicalone,inparticularbecause,onoccasion,naturallanguagefailsto‘dividenatureatthejoints’(Phaedrus).Anexampleistheterm‘barbarian’,towhichnoFormcorresponds,sincethedivisionofmankindintoGreeksandbarbarians(unlikethatintomaleandfemale,forinstance)failstoreflectareal,603\nLESLEYBROWNnatural,division.Insuch(rare)cases,thetermwillnothaveacorrespondingForm,but,ingeneral,investigationofLanguage(chapter3)isanothertoolforthephiloso-pheraspiringtodiscerntheroleofFormsinstructuringreality.Thisisespeciallyso,asthelatedialogueSophistshows,fortheinvestigationofFormswhichfigureprominentlyinLogic(chapter4)andDialectic(chapter22),including,forinstance,sameanddifferent,orlikeandunlike.Finally,intheirethicalrole,Formssuchasjustice,virtueand,especially,theFormofthegood,whichisprominentintheRepublic,serveasguarantorsofobjectivityinEthical(chapter6)andAesthetic(chapter7)questions.ConfidenceintheexistenceofethicalandaestheticFormsisabulwarkagainstthosewhowouldassertthatjusticeandothervaluesarenothingmorethanthebeliefsorconventionsofagivenpersonorgroup.Formoreonthis,seesection4.Arethevariousrolescompatible?CouldtherebeentitiessuchasPlatonicForms,oristhetheoryimpossiblyambitious,andintheendlogicallyincoherent?AcrucialassumptionseemstobethataForm–saytheForm,beautiful–isitselfbeautifulinsomespeciallyprivilegedway,awaywhichisbothsuperiortothatinwhichthemanyinstancesofbeautifulthingsarebeautifulandwhichexplainsorgroundstheirbeingbeautiful.(PerhapsthisideaisanearlyversionofthecausaladequacyprinciplefavouredbyDescartes(chapter26),thatifAcausessomethingtobeF,thenAmustitselfbeF.)WardycallsthisthenotionthatFormsare‘IDEALPERFECTEXEMPLARS’(p.592).RatherthantheinnocuousideathattheFormoruniversal‘beautiful’issimplywhatallbeautifulthingshaveincommon,thisseemstosuggest–incoherently–thattheFormcanhaveitselfasaproperty;henceitislabelledtheself-predicationassump-tion.AnargumentsketchedinPlato’sowndialogueParmenides,andlabelledbyAris-totle‘theThirdManargument’,claimstofindaregresshere.If(1)wepostulateaFormFtobewhatthegroupofthemanyF’shaveincommon,andif(2)theFormFisitselfF,then(3)wehaveanewgroupofF’s,themanyF’splustheFormF,forwhichwemust1postulateanewForm,F,toexplainwhatthenewgrouphaveincommon.Andsooninaninfiniteregress.Torebutthischallenge,thetheorycantrytoappealtoaspecialwayinwhichitistruethattheFormFisF,awaywhichavoidstheneedtoclassit,asanotherFthing,togetherwiththemanyF’satthethirdstepoftheargument.Thus,‘TheFormbeautifulisbeautiful’maybereadas‘TheFormbeautifuliswhatitistobebeautiful’,oralternatively,as‘TheFormbeautifulexplainswhyordinarythingsarebeautiful’.Butitishardtoacceptthateitherofthesegivesusanacceptablewayofreadingthepredication‘isbeautiful’.Indeed,itiseasytosympathizewiththoseimag-inedopponents(RepublicBook5)whodeniedthatthereisasinglethingthatexplainswhyallbeautifulthingsarebeautiful,letalonethattheputativesingleexplanansmustitselfbebeautiful.1.2ThesoulPlato’stheoryofthesoulisdevelopedintandemwithhistheoryofForms.Forms,thoughimmaterial,arenotmind-dependent,butitisthetaskofthemind,orthesoul,orreason,tocometoknowledgeoftheForms.WenowturntoPlato’spsychology,hisdiscussionsofthehumansoul(Greek:psyche).604\nPLATOANDARISTOTLEInthePhaedo,wherethetheoryofFormsisapparentlylaunched,aviewofthesoullatercalledCartesiandualism(pp.650–1)findsvividexpression,anditislinkedwiththeexistenceoftheForms.‘Assurelyasthoseobjects[theForms]exist,sosurelydooursoulsbeforeweareborn’(Phaedo76e).Socratesgoeson(79–80)todevelopawholestringofaffinitiesbetweensoulsandForms,ontheonehand,andbodiesand‘themany’ontheother.Thebody,likethemanysensibleobjects,isvisibletothesenses,ever-changing,many-facetedandperishable,whilethesoulis‘verylike’Forms,whichareinvisible,perceivedbyintellect,notthesenses,unchanging,simpleandimperish-able.Noteespeciallythatasoul,likeaForm,issaidtobesimple,thatisuncompounded,lackingparts.Unsurprisingly,itisinthePhaedothatwefindtheclaimthatitisthetaskofthephilosophertotrytoapproximate,inlife,tothestateofdeath,thatis,tothatstateinwhichthesoulisseparatedfromthechainsofthebody.Thebeststateforyou,asahumansoul,istobediscarnate,independentofthebody,thestateyouwereinbeforeandwillreturntoafterthisexistenceofyoursasahumanbeing.SuchisSocrates’sclaim,andthePhaedocontainsthreedaringargumentstotrytoshowthatthesoulisimmortal.ButtheRepublicintroducesamorenuancedandmorepsychologicallyconvincingpicture,thefamousaccountofthetripartite(orthree-part)soul.‘Dowelearnwithonepartofus,feelangrywithanotheranddesirethepleasuresofeatingandsexwithanother?Ordowedosowiththesoulasawhole?’asksSocrates(Republic436a).Theyreachagreementthatthefirstiscorrect:thateachofushasathree-partsoul,withareasoningelement,aspiritedelement(thumos)andathirdwhichistheseatofthebodilydesires.Evidenceforthiscomesfrominnerpsychologicalconflict,andPlatoquotesHomer’sdescriptionofOdysseusrebukinghisthumostoillustratehowaperson’sthumosisseenasdistinctfromthepartwhichreflectsonwhatisbestfortheagent,thereasoningpart.Wheneverwefindourselveswithoppositeimpulses–asOdysseusdid,oraswhen,thoughthirsty,werestrainourselvesfromdrinkingthroughconcernforhealth–inallsuchcaseswecandiscernopposingelementsinthesoulresponsiblefortheseopposingmotivations.LikeFreudtwothousandyearslater,Platosuggeststherearenofewerbutnomorethanthreedistinctpsychicparts–hencethelabel‘thetripartitesoul’.WeaknessofWillInplaceofthesimpledualismofbodyandsoulfromthePhaedo,theRepublicpresentsahighlysuggestivepictureofacomplexhumansoulwithconflictingmotivations,apicturewhichallowsandaccountsforirrationalbehaviour,suchaschoosingwhatyouknowisbadforyou.ThispicturediffersmarkedlynotjustfromthatinthePhaedobutfromviewsSocratesdefendsinPlato’sdialogueProtagorasagainstthecommon-sensebeliefinsuchweaknessofwillorakrasia.ThereSocrateshadarguedthatonlythosewhomiscalculateormakesomeothermistakecanchooseanddowhatisbadforthem.OntheProtagorasaccount,aperson’smotivationalwaysmirrorstheirevaluationoftheoptions;ifIfailtodowhatIclaimedtoknowwasbetter,thissimplyshowsthat,atthe605\nLESLEYBROWNvitalmomentofchoosing,Iinfactvaluedtheworseoptionmorehighly,akindofself-deception.Aristotleregardedthisasadenialofakrasia,butitmightbetterbethoughtofasexplainingakrasiainaspecialway,onewhichdeniesso-called‘clear-eyedakrasia’,butpostulatesakindofself-deceptiontoexplainsuccumbingtotemptation.ButnowinRepublicitisrecognizedthatdesirescananddosometimesgettheupperhandinaquitestraightforwardmanner:theyexertmoremotivationalforcethantheyshould,andtherationalpartcanbepowerlessagainstthem.(Thepowerofdesiresisrevealedinourdreams,saysSocrates–anotheranticipationofFreudiantheory.)Henceeducation–aprominentthemeinRepublic–mustfirstseektotameandcultivatethetwonon-rationalpartsofthesoul.Despiteitsgreaterpsychologicalrealism,includingallowing‘clear-eyedakrasia’,thetripartitesoultheoryinvitesmanyquestions.Whyonlythreepartsofthesoul?Consideringthehugenumberofpotentialinnerconflictsmightforceustopostulateinnumerableparts.Again,whyspeakofpartsatallratherthanasingleperson’smanyandconflictingsourcesofmotivation?How,inparticular,shouldweenvisagetheratio-nalpart?Hume(chapter31)objectedtoPlato’sideathatreason’sroleinvolvesbothrecognizingthegoodandpursuingit,forHumethoughtofreasonascapableonlyoffindingmeanstoends,notofrecognizingthevalueoftheendsthemselves.CentraltoPlato’s(andAristotle’s)conceptionofreasonistheideaHumerejected,thatreasonhasitsownendsanddesires.1.3EthicsandpoliticsInPlato’sworkswefindahighlyinfluentialandimportantdiscussionofmanymajorquestionsstillhotlydisputedinethicstoday.Herearesomeofthem.Thequestionofmoralrealismorobjectivism:aretheremoraltruthswhosetruthisindependentofanyindividual’sorsociety’sbeliefsand/orcustoms?Thequestionofmoralmotivation:whyshouldIbemoral?ThequestionoftherelationbetweenwhatGodwillsorcommandsorapprovesofandwhatismorallyright(seechapter15).Moralrealism.Platoisunswervinginhisoppositiontothosewhoholdthatjusticeandothermoralvaluesarenomorethanthebeliefsofgivenindividualsorsocieties.Indeedthisrealismissodeeplyentrenchedthatitisnotexplicitlyarguedfor.Ratherthantrytoprovethatmoraltruthsorfactsaretheretobeknown–ahopelessendeav-our–Plato’sdialoguesshow,ortrytoshow,thereaderhowtoseekmoralknowledge.Butthedifferentanswersareintheirowndifferentwaysunconvincing.FortheideathattheSocraticelenchoscanyieldmoralknowledge,seechapter22.The‘new’claim,thatmoralknowledgeistheapprehension(perhapsthroughrecollection)oftranscen-dentForms,islesslikelytoconvincereaders.Moralmotivation.Herethescepticalchallenge:‘whyshouldIbemoral?’isskilfullyconstructedandequallyboldlyrebutted.SpeakingforthescepticsinRepublic,Socrates’spartnerinconversation,Glaucon,putsthechallenge:whatreasondoIhavetobejust(thatis,tobemoral)ifIcangetthebenefitsofthemoralbehaviourofothersandofareputationformoralitymyself?Actuallybeingmoralbenefitsothers,notmyself,says606\nPLATOANDARISTOTLEthesceptic.GlauconreportsaSocialcontract(pp.393–4)theoryofmorality,onethatstressesthecostsandbenefitstoalloflivinginasocietywheremoralrestraintsencour-agethesociety’smembersto‘liveandletlive’.Thistheoryseesmybeingmoralasacosttome,butabenefittoothers,acostcompensatedfor(innormalcircumstances)bythebenefitsIderivefromtherestraintshownbyothers–theykeeptheirhandsoffmeandmypropertyinreturnformeleavingthemandtheirsalone.ButjustsupposeIcouldgetthebenefitsatnocost,withouthavingactuallytobemoralorconsistentlyactmorallymyself–if,forinstance,IhadtheringofGyges,whichmakesthewearerinvisibleatwill.Wouldnotanysensiblepersonchoosetoactimmorally,insuchascenario?Yes,saythesceptics.No,insistsSocrates.Whyso?Socrates’sansweristoarguethattobejust(thatis,moral)istohaveaharmoniouslyorderedsoul,thatis,oneinwhichreasonrules;thumosbacksuptheruleofreasonandthelowest,appetitive,partiskeptfirmlyincheck.Withsuchawell-orderedsoulthejustpersonisfarhappierthantheunjust,whatevermaterialbenefitstheunjustpersonenjoys,andwhateverhardshipsorpoorreputationthejustpersonsuffers.InthiswaySocratestriestoshowthatbeingmoralisofsupremevaluetothemoralagent,whothushaseveryreasontobemoral.Doesthisgiveasatisfyinganswertotheproblemofmoralmotivation?Herearesomeresponses.A:‘ThissaysIshouldbemoralbecauseitisgoodforme,butthatjustreducesmoralitytoself-interest,whichcannotberight.’B:‘No,itmakesmoralityanintrinsicgood,notsomethingwhichbringsotherbenefitstotheagent.’C:‘Evenso,insayingmoralityisvaluableinthatitisanintrinsicgoodtotheagent,thetheorydivorcesmoral-ityfromitseverydayconnectionwithinterpersonalactionsandwithother-regardingmotivationsandsentiments.’1.4PoliticaltheoryWefindnotonebuttwodifferentUtopias–sketchesofanidealpoliticalcommunity–inPlato;firstintheRepublic,andanotherinhislastwork,calledLaws.Bothemphasizetheimportanceofonlyproperlyqualifiedexpertshavingauthority.TheidealpolisinRepublicissketchedasananalogytotheperfectlyjusthumanbeing:eachhasthreeparts,andeachisperfectwhenthatpartwhichisnaturallyfittedtodosocontrolsthewhole.Inthepolisthispartfittedtoruleisasmallclassofhighlyedu-cated‘guardians’,withtherightnature,educationandlivingconditions(noprivatepropertyorfamilyties)tofitthemforabsolutepowerwieldedforthegoodofthewholepolis.KarlPopper’s(1966)famouscritiqueofPlato’spoliticalphilosophy,wherehechargeditwithtotalitarianism,hasbeenhotlydisputed.WherePoppersawanauthor-itarianelevationofthegoodofthestateoverthehappinessofthecitizens,morerecentwritershaveseentheproposalsasmerelypaternalistic.Thatis,thepoliticalaimissimplythehappinessofallthecitizens,thoughthisrequiresbenevolentcompulsionbythewiseguardiansofthemajoritywhoareruledbytheirpassions,notreason.Whichinterpretationiscorrect?Popper’scriticsarerightthatSocrates,inRepublic,neversug-geststhatthehappinessofthewholepoliscouldcomeaboutwithoutthatofalltheclasses.ButcanSocratesreallyallowthatmembersofthelowestclasscanenjoyahap-pinessotherthanthatofmakingtheirlowlycontributiontothehappinessofthepolis?607\nLESLEYBROWNIfnot,thenPopper’schargethatSocratesregardsthesecitizensaslittlemorethancogsinamachinehassometruthinit.TwokeythemespervadePlato’swritingonpolitics:theideathatfreedomisnotagoodwhenexercisedbythosewholackreasonandself-control,andtheideathatpowershouldonlybeentrustedtothosewiththeknowledgeofhowtouseitforgoodends.Thoseespousingliberalismmusthaveananswertoboth.2AristotleItisacommonplacetocontrastAristotle’sphilosophywiththatofPlato,forwhomtoinvestigaterealityisnottomakeanempiricalinquiryintothenaturalworldofevery-dayobjectsbuttoseekoutPlatonicForms,knowntotheintellect,nottosense-perception.AtfirstblushAristotle’sthoughtcanseemprofoundlyanti-Platonic.2.1Metaphysicsandepistemology2.1.1Aristotle’sempiricismOurword‘metaphysics’derivesfromthetitleofamajorworkofAristotle’s,thoughitsimplymeant‘whatcomesafterthePhysics’;thatis,afterhisinvestigationsconcern-ingthenaturalworld.TheMetaphysicsopenswiththeremarkthatthesenses,andsightinparticular,areavitalsourceofknowledge.Learningandknowledgerequiresense-perception,memoryandexperience–progressivestepstotheformationoftheuniver-salgraspofsimilarthingsrequiredforknowledge.Aswellasthisemphasisontheneedforsense-perception,memoryandexperience,anotherapparentlydirectconfrontationwithPlatoisAristotle’sdown-to-earthinsistence,inhisearlyworktheCategories,that‘primarysubstances’areeverydaythingslikeaparticularmanoraparticularox–adirectrebuttalofPlato’srelegationofthesetoaninferiorstatustotheForms,andofPlato’sclaimthatparticularthingsgettheirrealityby‘sharingin’theForms.Indeed,Aristotlemakesfrequentattacksontheclaimsof‘thePlatonists’aboutForms.Hearguesthatin‘separating’Formsfromtheeverydaythingswhicharesup-posedto‘sharein’them,PlatoandhisfollowersmakethemistakeoftreatingFormsbothasparticulars(thatis,forAristotle,asthingswhichcannotbepredicatedofmanythings)andasuniversals.(SeeespeciallyMetaphysics7,ch.13.)ButonPlato’sbehalfwecouldreplythatwhatAristotlecriticizesasPlato’sobjectionablethesisofthe‘separation’oftheFormsismerelytheplausibleclaimthatuniversalsarepriortoparticularsinthesensethattheycouldexistuninstantiated.Toarguethisisnottomakeauniversalatthesametimeaparticular,itcouldbeurged.ButwhileAristotlefirmlybelievesinuniversals–aswesaw,hebelievedknowledgemustbeofwhatisuniversal–herejectedtheideathattheycouldexistuninstantiatedandthusindependentlyoftheirparticularinstances.2.1.2Aristotle’srationalismDespitehisrejectionofaspectsofPlato’smetaphysics,AristotlesharesmanytenetswithPlato.Knowledge,forboth,isofwhatisuniversal,and,moreimportantly,requires608\nPLATOANDARISTOTLEknowledgeofthe‘why?’–ofthecauseorexplanationofwhateverisbeinginvestigated,whethernaturalobjectslikeakindofanimaloritsparts,orphenomenasuchaseclipsesorlightning.Ininvestigatingcausesitisvitaltodistinguishwhatisessentialfromwhatismerelyaccidentalorcoincidental.Athing’sessentialpropertiesarethoseithasinitsownright,onwhichitsotherpropertiesdepend.Andinhisinsistenceonthelinkbetweenknowledge,explanationandessence,Aristotlegivesamajorroletoform,incontrasttomatter,andoften,likePlato,equatesformwithessence.2.1.3MatterandforminAristotleItiseasytograspthedistinctionbetweenmatterandforminthecaseofsomethinglikeabronzesphere,wherethematterormaterial–bronze–acquirestheform–asphericalshape–whenpouredintoamouldbythecraftsman.Bothmatterandformcancountascauses,inthateachexplainssomepropertiesofobjects:itsmatterexplainswhythebronzesphereringsoutwhenstruck,itsshapeexplainsitsrollingdownanincline.Butgenerally,Aristotleholds,formismoreimportantinacausalexplanationthanmatter,andhefrequentlycriticizeshispredecessors(notincludingPlato)forfocus-ingonmaterialcauses–thebasiselementsofthings–andignoringform.Thisisespe-ciallysoforlivingthings,forwhomtheirformismorethantheirshapeandiscloselyrelatedtoFunction(pp.319–20).Awooden‘hand’isnotahand,andthatofadeadpersonisonlybycourtesycalledahand.ForAristotletheformofalivingthingisitsorganization,thebasisoftheactivities(suchasflying,hunting,eatingmice)characteristicofathingofthatkind.Nowinlivingthingstheirmatterisreplacedovertimeanditistheform,thusunderstood,whichpersists,andwhichexplainstheorganism’sbehaviour;henceformismoreproperlycauseandessencethanmatter.ButAristotlerecognizesthatacleardemarcationbetweenformandmatterseemstobreakdown;attimesheallowsthattheformoflivingthingsmustincludethe‘proximatematter’–thesortoffleshandbonesneededforcreaturesofagivenkindtofunctionastheydo.Andthereisafurther,grave,probleminunderstandingAristotle’smetaphysics:thatofwhetherheregardsformsasuniversalorasparticular.WhileinhisearlyworkCategorieshemadeconcreteindividualssuchasaparticularmanoroxtheprimarysubstances,herevisitsthequestion‘Whatissubstance?’intheMetaphysicsandtheregivesashisconsideredanswer,nottheconcreteindividual,composedofformandmatter,butform.Now,ifformistheuniversal,thespeciesform‘man’(thatis,whatSocratesandCalliasshare),thenitcancertainlybetheobjectofknowledgeanddefini-tion,sincethesemustbeofsomethinguniversal,Aristotleinsists.Butnothinguniver-salcanbeasubstance–thiswasineffectAristotle’sobjectiontoPlato.Soif,onthenewview,formistobesubstance,thenitmustbeparticularforms,oneforSocratesandoneforCallias,whichAristotlehereintends.SoAristotle’smetaphysicsandepistemologyhavemuchincommonwithPlato’s.Despitehismoredown-to-earthapproachtothenaturalworld,andhiscriticismsoftheexcesses,ashesawit,ofPlato’stheoryofForms,AristotlesharesPlato’sfocusoncauses,essencesandform,andhisinsistencethatrealitymustbeintrinsicallyintelli-gibleandwell-ordered.Indeed,itmightbethoughtthatinsomeofhistheses–notablyinthefamousargumentthatseekstoshowthatiftherearetobechangingsubstances609\nLESLEYBROWNsuchashorses,treesandhouses,theremustbeanunmovedmover,God,whoispureformwithoutmatter–AristotleequalsorsurpassesPlato’smetaphysicalflights.2.2ThesoulHerethecontrastbetweenAristotle’sviewandPlato’sismoremarked,and,fromacon-temporaryperspective,Aristotle’spsychology(theoryofthesoul)containsfarmorepromisinginsights.InopposingPlato’sview(aprecursorofCartesiandualism,aswesaw)thatthehumansoulissomethingimmaterial,independentofthebodyandcapableofsurvivingitsdeath,Aristotleplaceshisinvestigationfirmlyinthecontextofnaturalscience.Soul,forAristotle,issimplythatinvirtueofwhichlivingthingsarealive,andhecastigatesthosewhofocustheirinquiriesintosoulonhumansoulandneglectnon-humananimals,whosharemany‘psychicfunctions’withhumanbeings.ForAristotle,thisincludesfunctionsoflivingthingsnotnowthoughtofaspsychic,suchasbreathing,growthandnutrition,inadditiontothought,perceptionandmemory.Thisextensionof‘psychological’inquiriestoalllivingthingsandtothecapacitiestheyhaveaslivingthingsgroundsmanyofthedistinctivefeaturesofAristotle’saccountofsoul.Asinhisoverallaccountofreality,hesteersapathbetweenthelegacyofPlato–here,theaccountofthesoulasanindependententity,thesubjectofpsychologicalpropertiessuchasthinkingandmemory,andcapableofsurvivingthedeathoftheorganism–andthatofmaterialistpredecessorswhohadsoughttoexplainallpsychicfunctionsintermsofthematerialconstituentsoflivingthings.Famously,Aristotledefinedthesoulastheformofanaturalbody,thatis,ofabodyofalivingthingwithorgans.Heinsiststhatpsychicphenomenasuchasperception,emotions(forexample,anger),memoryandthinkingcannotoccurexceptinanem-bodiedorganism(thoughheoccasionallyqualifiesthiswithreferencetoonekindoffaculty–labelledactiveintellect–whichiscapableofindependent,thatisnon-embodied,existence).Theinsistencethatpsychicfunctionsrequireabodygoesbeyondaclaimthatthesoulcanonlyoperatewiththehelpofabody.Itisnotthesoul,buttheorganismitself,whichisthesubjectofmentalphenomena.‘Tosaythatthesoulisangryislikesayingthatthesoulisweavingorisbuildingahouse.It’sbettertosay,notthatthesoulisfeelingpity,orthinking,orlearning,butthatthehumanbeingisdoingsoinvirtueofthesoul’(DeAnima1.4).Theseremarks,togetherwiththeinsistencethatpsychicphe-nomenaarenecessarilyenmattered,mayseemtosuggestthatAristotlesideswiththematerialistswhoseektoaccountforpsychicphenomenapurelyintermsofthemate-rialconstituentsofthelivingorganism.Buttocallsoultheformofthenaturalbodyisverydifferentfromthinkingofitasthearrangementofthematerialparts.Aswesawabove,theforminthecaseofalivingthingistheprincipleoftheorganizationofthatthing,itsdispositiontobehaveinallthewayscharacteristicofthatkindofthing.Certainly,sinceperceiving,beingangry,pityingandsoonarefunctionsofaneces-sarilymaterialthing,thesoulastheformofanaturalbodywillincludedescriptionsoftheminapparentlymaterialterms.Aristotledistinguishestwocurrentdefini-tionsofanger.One,givenbyonekindofstudentofnature,focusesonthematerialaspectanddefinesangeras‘theboilingofbloodaroundtheheart’,whileanother,the610\nPLATOANDARISTOTLEdialectician’s,givestheform,definingangeras‘thedesiretotakerevenge’.Butatruestudentofnature,hetellsus,shouldincludebothaspects.Thereasonforthisispre-sumablynotjustthateachcontributesanessentialpartoftheaccountofanger,butalsothateachalreadyincorporatesaspectsofthepartnerterm.Theallegedlymaterialdefinition–theboilingofbloodaroundtheheart–isalreadyanaccountwhichpre-supposesthatalivingorganismisbeingdescribed–henceonewhoseformisevidentlyvitaltoitsbeing–whilethedesiretotakerevengeisitselfnecessarilyenmattered,thatishasaphysiologicalbasis.Aristotle’smiddlepathbetweendualismandmaterialism,hisdiscussionofpsycho-logicalphenomenawithreferencetonon-humanaswellastohumananimals,andhisforegroundingofthewholelivingorganismasthesubjectofpsychologicalpropertiesgivehispsychologyconsiderableappeal.Andmanywillapprovehisrefusal,nonethe-less,toacceptaReductionist(pp.174–8)accountofmentalproperties.Hugeques-tionsremainastohowAristotlethinksweshouldunderstandtherelationbetweenaccountswhichfocuson,say,thechemicalpropertiesofanorganismandaccountswhichmentionsuchthingsasthedesireforrevenge.Suchquestionsremainatthefore-frontofmodernphilosophyofmind,thoughAristotleisinnocentofthemodernemphasisonthedifficultyofcapturinginphysiologicaltermsthefirst-personperspec-tive,andthephenomenaofConsciousness(pp.650–2).2.3EthicsEthicsistheinquiryintothehumangood,thegoodforhumanbeings.Everyoneagreesthatthehumangoodiseudaimonia(happinessorflourishing),butbeyondthatthereismuchcontroversy.Settingoutreputableopinionsonthesubjectyieldsthesuggestionsthateudaimoniaispleasure,orvirtueorcontemplation.Suchopinionsalsorevealthatitsformalpropertiesincludethefollowing:eudaimoniaisactuality,notpotentiality(andthisisonereasontoexcludetheidentificationofvirtue–apotentiality–witheudai-monia,thoughvirtueisintrinsicallyconnectedwithit);itissoughtforitsownsakeanditcannotbemadebetterbytheadditionofothergoods(forthisreasonitcannotbeequatedwithpleasure).TofleshouttheaccountAristotleturnstotheideaofthefunc-tionofahumanbeing,andlocatesthisinactivitiesinaccordancewithreason(broadlyunderstood),concludingthateudaimonia,beingsupremelygood,consistsinactivitieswhichmanifestthebeststates,thatisthevirtues,oftherationalaspectofhumanbeings.Soanaccountofthesevirtuesisneeded;theyfallintotwocategories:virtuesofintellectandvirtuesofcharacter(oftencalledmoralvirtues).Thiscorrespondstoanimportantdivisionoftherationalpartofthesoulintotwo:thepartwhichisfullyrational,theintellect,andthepartwhichcan‘obeyreason’,whichAristotlelabelsthedesiderativepart.Thispartisthesourceofdesiresofmanykinds,bothphysicalappetitesanddesiresforfame,honourandrevenge,togetherwiththeemotionsandfeelingsassociatedwiththesedesires.Thevirtuesofcharacter,excellentstatesofthispartofthehumanbeing,aredefinedasinvolvingameanbetweenexcessanddefi-ciency;theyincludecourage,temperance,justiceandahostofothervirtuessuchasgentlenessandgenerosity.Thevirtuesofintellectaredividedintothoseconnectedwithpracticalthinking,andthoseconnectedwithpurelyintellectualthought.Forthe611\nLESLEYBROWNmajorityoftheworkitseemsthateudaimoniaistobeidentifiedwithalifemanifestingvirtuesofallthreetypes,butaratherdifferentpictureemergesattheendoftheworkwherecontemplation,theactivityofthetheoreticalintellect,isarguedtobethebestcandidateforeudaimonia.Anattractivewaytoresolvethistensionistorecognizethatcontemplationwouldbethebestlifeforthosewithadivinenature,butcannotbesoforhumanbeings,who,unlikethegods,haveneeds,desiresandfeelings,andareessen-tially‘politicalanimals’,thatiswhoflourishincommunitieswheresocialattributesarecrucial.EthicsastheSearchfortheHighestHumanGoodInputtingeudaimoniaattheheartofethicalinquiry,Aristotle’stheorymayseemtoresembleUtilitarianism(chapter35),buttherearetwomajordifferences.Aswehaveseen,hisaccountofeudaimoniaintermsofactivitiesmanifestingthevirtuesmakesitafarcryfromtheutilitarianaccountofhappinessaspleasure.Secondly,utilitarianismmakeshappinessthesoleultimategoodandproceedstoexplainmoralgoodness,espe-ciallyrightaction,asthatwhichpromotesthegood.Thusmoralgoodnesshasaderiva-tiveworth,derivedfromitscontributiontohappiness,whichisconceivedinnon-moralterms.Bycontrast,incharacterizingeudaimoniaasactivitiesmanifestingthevirtues,Aristotleisoperatingwithaunifiedconceptionofhumangood,oneinwhichethicalormoralgoodnessisanintegralpartofthegoodlife.AndinthisrespectAristotle’sstarting-pointdiffersalsofrom,andisarguablysuperiorto,thatofKant(chapter32)andneo-Kantianswholiketodistinguishthegoodorvalueofahumanbeingfromwhatisgoodforahumanbeing,insistingthatethicsshouldconcernitselfwiththefirstandnotthesecond.ItisastrengthofAristotle’sapproachthathisinquiryinto‘thehumangood’doesnotallowthisbifurcationand,afortiori,privilegesneitherthenon-moralgood(asutilitarianismdoes)nor,withtheKantians,themoralgood.Insearchingforthehumangood,AristotledistanceshimselffromPlato’sinquiryintotheFormoftheGood,anallegedtranscendentgood.Aristotlerejectsthisonthegrounds,first,thattherecouldbenosuchthing,andsecond,thateveniftherewereitwouldbeofnohelpinanethicalinquiryintothegoodforhumanbeings.2.3.1Eudaimonia,virtue,functionandthegoodofothersTheideathathumanbeingshaveasharedfunctionisobjectionableifitimpliesthathumanswerecreatedforapurpose.ButAristotle’sappealtothehumanfunctionissimplyanappealtothepremisethattherearecapacitiesandactivitiescharacteristictohumanbeings,withoutwhichalifeislessthanfullyhuman,andtotheclaimthatthisessentiallyinvolvestheexerciseofreason.Thatthebesthumanlifeatleastinvolves(thoughitneednotbeexhaustedby)activitiesmanifestingthebeststatesofarationalnatureisrelativelyuncontroversial.FarmoreproblematicareAristotle’sassumptions(a)thatthisseeminglyvaguecharacterizationpicksoutasufficientlydeterminateendand(b)thatthese‘beststatesofarationalnature’includethemoralvirtues,astradi-tionallyunderstood:courage,temperance,generosity,justice.Forwhatthetheorytakestobegoodaboutthesestatesisthattheyarewhatmakestheindividual’slifegowell,612\nPLATOANDARISTOTLEbutitisalargeassumptionthatjustice,inparticular,isintherequiredwaygoodfortheindividual.WhereasPlatohadtackledheadonthescepticalchallenge:whyshouldIbemoral?Aristotlequietlyavoidsit,bysimplyincludingjusticeamonghislistofmoralvirtues,aninclusionthatascepticwouldchallenge,onthegroundsthatjusticeisnotevidentlyoneofthosedispositionsthatitisbestfortheagenttopossess.OnAristotle’sbehalfwecouldreply,first,thathisworkwasaddressedtowell-brought-upyoungmen,whowillalreadybedisposedtoviewjusticeasadesirablequalityforanindividualtopossess;second,thatgoodgroundsforthisassumptioncanbefoundinthenatureofhumanbeings,whoflourishincommunitiesandwhenengagedinco-operativeactivitieswithfriendsandassociates.2.3.2Moralvirtues,themean,and‘practicalwisdom’:Aristotle’sversionofmoralrealismThefamousdoctrineofthemean–thethesisthatthemoralvirtuesaremeanstateswherethemeanis‘relativetous’–involvesthefollowingclaims.Tohaveandtomani-festamoralvirtuesuchastemperanceinvolvesfeelinganappropriatedegreeofitsasso-ciatedemotion(desireforbodilypleasures,inthecaseoftemperance)andmakingtheappropriatechoiceofsuchpleasures,avoidingexcessanddeficiencyinfeelingandaction.Themeaniswhatisappropriate,neithertoomuchnortoolittle,anditis‘rela-tivetous’ashumanbeings,notasdifferentindividuals–therearenotdifferentvirtuesaccordingtoourdifferentsusceptibilitiestosuchpleasures.Butwhatdetermineswhatisappropriate?Forthisoneneeds,asacomplementtoagivenmoralvirtue,theintel-lectualvirtueofphronesis,practicalwisdom.Aristotle’saccountofthisimportantvirtueisfraughtwithdifficulties,butitisclearthatheviewsitasakindofmoralknowl-edge,agraspofamoralreality(thattheappropriatewaytorespondinthesecircum-stancesistofeellikethisandtoactinthisway).Itisevidentthathisviewprivilegesknowledgeofwhatismorallyappropriateinaparticularsetofcircumstancesoveruni-versalmoraltruths.Somegofurtherandclaimthathisviewisparticularistinrefus-ingtoacknowledgeanyimportantroleforuniversalmoralprinciplesinmoralknowledge.Aristotlenevermakescleartherespectiverolesofdeliberationandinsightinphronesis,butitisclearthatheseesacrucialinterdependencebetweenthefullyfledgedmoralvirtueanditsintellectualcounterpart,practicalwisdom,suchthatonecannotfirstacquiretheoneandthentheother:learningtofeelandactintheappro-priatewaygoeshandinhandwithacquiringagraspofwhatthatappropriatewayis.Andno‘external’criterionofwhatisappropriateshouldbedemanded;whatisappro-priateisevidenttoapersonofmoralvirtueandpracticalwisdom,butthismustnotbetakentomeanthatitistheirsay-sothatmakesitappropriate.Thistheoryofmoralknowledgeandrealityavoidsthetwinextremesofreducingthegoodnessofagoodactiontoamatterofconvention,andoffindingitinthestrictapplicationofuniversalprinciples.Andifsomeonecomplainsthattheaccountisnotveryinformative,thatitdoesnottellus,casebycase,wherethemeanlies,thenAristotlewouldreplythatthecomplaintmissesthepoint:histheoryismeanttoshowwhynosuchaccountispossi-ble,andwhysuchknowledgeisavailableonlytoagoodperson(whoseresponseshavebeenproperlytrainedandwhohasdevelopedpracticalwisdomintandemwiththis)placedintherelevantcircumstances.613\nLESLEYBROWN2.4PoliticaltheoryWhatarewetomakeofAristotle’sfamousdictumthat‘manisapoliticalanimal’?Certainlythismeansthathumanbeingsarebynaturesuitedtolivinginapoliticalcommunity,butdoesAristotlethinkthathumanbeingsexistforthepolis,orthatthepolisexistsforhumanbeings?Iftheformer,hisviewwouldresemblethattakeninPlato’sRepublic(atargetforAristotle’ssharpcriticism),atleastonPopper’sinterpreta-tionofit.Aristotle’sclaim(Politics1253a:19–29)thatthepolisispriortothefamilyandthecitizenasawholeorganismistoapartsuchasthehandcertainlysuggeststhisdisturbingview.Butthisismisleading.Aristotleholds,indeed,thatahumanbeing’sfunctionorcharacteristicexcellencecanbeperfectedonlyinapolis,buthedoesnotthinkthepolisdefinesthatfunction,orthatman’sfunctionistoservethepolis.Rather,thepointofapolisistocreatecondi-tionsinwhichitscitizenscanlivethebestlifepossible.Thiswillincludesocialandpoliticalinteractionaswellasvariousintellectualpursuits,allpursuedfortheirownsake.Aristotle’spoliticalphilosophydoesnot,then,subordinatethecitizentothepolis.Butithasaverydifferentdefect,thatofholdingthatcitizenshipshouldnotbeextendedtothosehumanbeingsincapableoflivingthebesthumanlife,andthis,forAristotle,meanswomen,manualworkersandthosehenotoriouslylabels‘naturalslaves’.ItisanabidingchallengetopoliticaltheorytoseeifitispossibletomarryamoreinclusiveanddemocraticconceptionofcitizenshiptoAristotle’scentral,andattrac-tive,tenetthatthegoalofthepolis–thatis,theaimofpoliticalarrangements–shouldbenotmerelytoenableitscitizenstolive,butto‘livewell’,thatistoprovidetheconditionsunderwhichcitizensacquireandexercisethemoralandintellectualhumanvirtues.FurtherReadingThebestintroductiontoPlatoistoreadsomeofhisdialogues,whichareliteraryaswellasphilosophicalmasterpieces.TheyarecollectedinEnglishtranslationinasinglevolumeinHamiltonandCairns(1961)andalsoinCooper(1997).ManyofthetranslationsinthelatterarealsopublishedseparatelybyHackett.Anexcellent‘taster’collectionisTheLastDaysofSocrates,containingthreeshortworks(Euthyphro,on‘Whatispiety?’,Apology,Plato’srepresen-tationofSocrates’sdefenceathistrial,andCrito,onpoliticalobligation)andthemuchlonger,moreprofoundanddemandingPhaedo.SeealsoGrube(1981),whichhasthesefourandMeno.EthicalissuesarediscussedinthesparklingProtagorasandinGorgias,oneofwhosecharacters,Callicles,wastheinspirationforNietzsche’srevaluationofvalues.Republic,Plato’smostfamouswork,coversethics,politics,philosophyofmind,metaphysicsandtheoryofknowledge,aswellasphilosophyofart.Despiteitshyperboliccriticism,Popper’s(1966)attackonRepublicisaclassic.Annas(1981)isanexcellentintroductiontothewholework.Theaetetus,whosetopicisknowledgeanditsrelationtoperceptionandtoopinion,rivalsorsurpassesRepublicandPhaedoinitsbrillianceandtoughnessofargument,butunlikethemdiscussesknowledgeandrealitywithoutanappealtotheForms.Threefineworksonit(byMcDowell,BostockandBurnyeat)arelistedbelow.StillonPlato,Kraut(1992)containsafullbibliographyandrecentarticles,includingdiscus-sionsoftheorderanddatingofPlato’swritings,andofthedebateoverwhetherweshouldseek,inthedialogues,anaccountofPlato’sownphilosophy,andifso,whether‘unitary’or614\nPLATOANDARISTOTLE‘developmental’.Kahn(1996)favourstheformer,butmostrecentwritersassumedevelop-mentbothinPlato’sethicalviews,forexampleIrwin(1995),andinhismetaphysics.ValuablecollectionsofarticlesonarangeofPlatonicissuesbyvariouswritersareAllen(1965),Kraut(1992)andFine(1999);single-authoredcollectionsareOwen(1986)andVlastos(1981,1991).Rutherford(1995)isvaluableforliteraryandhistoricalaspectsofPlato.Aristotle’scollectedworksareavailableinatwo-volumetranslation(Barnes1984).SubstantialselectionsareavailableinAckrill(1987)andintheveryusefulIrwinandFine(1995),withnotes,glossaryandbibliography.Itcontainspartsorallofthefollowingworks:onlogic:Categories*,DeInterpretatione*,PriorandPosteriorAnalytics*,Topics;onnaturalphilosophy:Physics*anddeGenerationeetCorruptione;onpsychology,deAnima*(ontheSoul)anddeMotuAnimalium;Metaphysics*,NicomacheanEthics*;Politics*,RhetoricandPoetics.ClarendonAristotletranslationspluscommentaryareavailableforallorpartsofthoseasterisked.ExcellentshortintroductionstoAristotleareAckrill(1981)andBarnes(1982).HelpwiththedifficultMetaphysicscanbegainedfromseveralvolumesoftheClarendonAristotleseries(translationandcommentary)andfromWitt(1989)orLoux(1991).ForNicomacheanEthics(themostimportantoftheethicalworks)seetranslationsbyCrisp(2000)orIrwin(1999,withusefulnotesandglossary),andstudiesbyUrmson(1988,agoodintroduction)andBroadie(1991).Rorty(1980)isavaluablecollectionofessaysonAristotle’sethics.Barnes(1994)hasarticlesonmanyaspectsofAristotle’sthought,andagoodbibliography.Ross(1964)isstillaveryusefulgeneralwork.ReferencesPlato:translationsCooper,J.M.(ed.)1997:Plato:CompleteWorks.Indianapolis,IN:Hackett.Grube,G.M.A.1981:Plato:FiveDialogues.Indianapolis,IN:Hackett.Hamilton,E.andCairns,H.(eds)1961:TheCollectedDialoguesofPlato(includingtheLetters).Princeton,NJ:PrincetonUniversityPress.Tredennick,H.(trans.)1993:TheLastDaysofSocrates,revdedn(revisedbyH.Tarrant).Harmondsworth:PenguinBooks(containsEuthyphro,Apology,Crito,Phaedo).ThetranslationswithphilosophicalcommentaryintheClarendonPlatoseriesincludethoseonPhaedo(D.Gallop),Protagoras(C.C.W.Taylor),Gorgias(T.H.Irwin)andTheaetetus(J.H.McDowell).TheTheaetetusofPlato,translatedbyLevett(1990:Hackett),hasalongintroductionbyBurnyeat.ManytranslationsofRepublicareavailable;forLaws,thereisT.J.Saunders(1970:PenguinBooks),withintroductionandsummariesaswellasbibliography.WorksonPlatoAllen,R.E.1965:StudiesinPlato’sMetaphysics.London:Routledge.Bostock,D.1988:Plato’sTheaetetus.Oxford:ClarendonPress.Fine,G.1999:Plato1MetaphysicsandEpistemology,Plato2EthicsPoliticsReligionandtheSoul(OxfordReadingsinPhilosophy,alsoavailableinasinglevolume).Oxford:OxfordUniversityPress.Kahn,C.H.1996:PlatoandtheSocraticDialogue.Cambridge:CambridgeUniversityPress.Kraut,R.1992:TheCambridgeCompaniontoPlato.Cambridge:CambridgeUniversityPress.Owen,G.E.L.1986:Logic,ScienceandDialectic(collectedpapers).Ithaca,NY:CornellUniversityPress.Popper,K.R.1966[1945]:TheOpenSocietyanditsEnemies,vol.1:TheSpellofPlato.London:Routledge.615\nLESLEYBROWNRutherford,R.1995:TheArtofPlato.London:Duckworth.Scott,D.1995:RecollectionandExperience:Plato’sTheoryofLearninganditsSuccessors.Cambridge:CambridgeUniversityPress.Vlastos,G.1981:PlatonicStudies.Princeton,NJ:PrincetonUniversityPress.——1991:Socrates,IronistandMoralPhilosopher.Cambridge:CambridgeUniversityPress.Aristotle:translationsAckrill,J.L.(ed.)1984:ANewAristotleReader.Oxford:OxfordUniversityPress.Barnes,J.(ed.)1984:TheCompleteWorksofAristotle,2vols.Princeton,NJ:PrincetonUniversityPress.Irwin,T.andFine,G.1995:Aristotle:Selections.Indianapolis,IN:Hackett.WorksonAristotleAckrill,J.L.1981:AristotlethePhilosopher.Oxford:OxfordUniversityPress.Barnes,J.1982:Aristotle.Oxford:OxfordUniversityPress.——(ed.)1994:TheCambridgeCompaniontoAristotle.Cambridge:CambridgeUniversityPress.Ross,W.D.1964[1923]:Aristotle.London:Methuen.ONMETAPHYSICSLoux,M.J.1991:PrimaryOusia.Ithaca,NY:CornellUniversityPress.Witt,C.1989:SubstanceandEssenceinAristotle.Ithaca,NY:CornellUniversityPress.ONPSYCHOLOGYNussbaum,M.C.andRorty,A.O.(eds)1992:EssaysonAristotle’sDeAnima.Oxford:OxfordUniversityPress.ONETHICSCrisp,R.2000:Aristotle,NicomacheanEthics(introductionandtranslation).Cambridge:CambridgeUniversityPress.Irwin,T.H.1999:Aristotle,NicomacheanEthics(translation,notesandglossary).Indianapolis,IN:Hackett.Rorty,A.O.(ed.)1980:EssaysonAristotle’sEthics.Berkeley:UniversityofCaliforniaPress.Urmson,J.O.1988:Aristotle’sEthics.Oxford:Blackwell.DiscussionQuestions1Ifknowledgeisrecollection,mustthesoulbeimmortal?2Whatisvirtue?Isvirtueteachable?3CanordinarythingsparticipateinaForm?4AretheregoodreasonstothinkthatFormsandmindsarepeculiarlyfittedtooneanother?5CananythingsatisfyalltherolesthatPlatoassignstoForms?6DoestheFormofXhaveasuperiorkindofexistenceorasuperiorwayofbeingX?7Whichshouldhavepriority:thesemanticroleofFormsortheirontologicalrole?616\nPLATOANDARISTOTLE8HowdoesaninvestigationoflanguagehelptodiscerntheroleofFormsinstructuringreality?9CanthetheoryofFormsguaranteetheobjectivityofethicsandaesthetics?10DoesthetheoryofFormsinvolveusinanendlessregress?11Canasinglethingexplainwhyallbeautifulthingsarebeautiful?12Dosoulshaveparts?Ifso,whataretheyandhowcantheybeidentified?13Howisweaknessofwillorakrasiapossible?14Whatshouldbeourconceptionofreason?Canreasondetermineendsaswellasmeanstoachieveends?15WhyshouldIbemoralifIcouldgetthebenefitsofmoralitywithouthavingactuallytobemoral?16Isthejustpersonhappierthantheunjust,evenifthebalanceofmaterialbenefitsandhardshipsfavourstheunjust?17Whoshouldhaveauthority?Istheanswerthesameinapolisandinamodernsociety?18Isfreedomagoodwhenexercisedbythosewholackreasonandself-control?19IsAristotle’sthoughtanti-Platonic?20Isaparticularmanasubstance?21Howshouldwedrawthedistinctionbetweenparticularsanduniversals?Areformsparticularsoruniversals?22Canuniversalsexistiftheyareuninstantiated?23Doesknowledgerequireanaccountofthecauseorexplanationofwhatisbeinginvestigated?24Canwedistinguishbetweenessentialpropertiesofathinganditsaccidentalproperties?25Istheessenceofathingitsform?26Isformmoreimportantthanmatterincausalexplanation?27Whatissubstance?Shouldcontemporaryphilosophybeconcernedwiththisquestion?28Mustrealitybeintrinsicallyintelligible?29Canallmentalfunctionsbeexplainedintermsofthematerialconstituentsoflivingthings?30Howshouldweassesstheclaimthatthesoulistheformofanaturalbody?31Dopsychicfunctionsrequireabody?32Isthesoulortheorganismitselfthesubjectofmentalphenomena?33Areconsiderationsofbothphysiologyanddesiresnecessaryforanaccountofanger?34Canweavoidreductionisminouraccountofmentalphenomena?35Whatconstituteshumanhappinessorflourishing?Dophilosophershaveaspecialroleinansweringthisquestion?36Canwedistinguishbetweenvirtuesofintellectandvirtuesofcharacter?37Dovirtuesinvolveameanbetweenexcessanddeficiency?38Isalifeofcontemplationthebestcandidateforeudaimonia?39Mustagoodlifebeethicallygood?40WouldknowingtheFormoftheGoodhelpustounderstandthegoodforhumanbeings?617\nLESLEYBROWN41Doesknowledgeofuniversalmoralprincipleshaveanyimportantroleinmoralknowledge?42Couldwehavemoralvirtuewithoutmoralwisdomormoralwisdomwithoutmoralvirtue?43Whatistheroleofcitizenshipinthepolis?Shouldcitizenshiphavethesameroleinmoderndemocraticsocieties?618\n24MedievalPhilosophyJORGEJ.E.GRACIATheconcerntointegraterevealeddoctrineandsecularlearningdistinguishesmedievalphilosophyfromancient,Renaissanceandmodernphilosophyanddeterminestoagreatextentthephilosophicalproblemsthemedievalsaddressedandthesolutionstheyproposedforthoseproblems.Thischapterexaminesthewaythemedievalsapproachedthismainthemeandillustrateshowitaffectedtheirchoiceofphilosophicalproblemsandhowtheydealtwiththem.Inparticular,itpaysattentiontosevenproblemswelldiscussedthroughouttheage:therelationoffaithandreason,theexistenceofGod,thesignificanceofnamesusedtospeakaboutGod,theobjectoftheologyandmetaphysics,thewayweknow,universals,andindividuation.C.F.J.Martinhascontributed‘UnderstandingMedievalPhilosophy:ArgumentsfromAuthority’.Theuseoftheexpression‘medievalphilosophy’torefertophilosophyinthemiddleagesisparadoxicalbecauseitishardtofindanyoneduringtheperiodwhoconsideredhimselfaphilosopher,whoseconcernswerepurelyphilosophical,orwhocomposedpurelyphilosophicalworks.MedievalauthorsfromtheLatinWestthoughtofthem-selvesratherastheologians,wereprimarilyinterestedintheologicalissues,andveryseldomcomposedpurelyphilosophicalworks.Forthem,thephilosophersweretheancients,PlatoandAristotle(chapter23),andsomeoftheIslamicauthors,likeAvicennaofBaghdad(IbnSina,980–1037)andAverroesofCordoba(IbnRushd,1126–98).Thereareveryfewworksproducedintheperiodthatcanbeclassifiedstrictlyspeakingasphilosophical.Mostofthephilosophythatwefindiscontainedinbooksoftheologyandusedtoelucidatetheologicaldoctrine.Whencethewell-knownphrase,coinedbyThomasAquinas(c.1225–1274)inreferencetophilosophy,ancillatheologiae,‘servantoftheology’.Theexpression‘medievalphilosophy’,moreover,hasadisparagingconnotationderivedfromtheterm‘middleages’,usedfirstbyRenaissancehumaniststorefertowhattheythoughtwasabarbaricanddarkperiodofWesternhistoryfoundbetweenthetwocivilizedandenlightenedagesofclassicalantiquityandtheRenaissance.Inspiteofthelackofphilosophers,theabsenceofpurelyphilosophicalworks,andtheprejudicesofRenaissancehumanists,themiddleagesisnotonlythelongestperiodofphilosophicaldevelopmentintheWest,butalsooneoftherichest.Indeed,inintensity,sophisticationandachievement,thephiloso-phicalfloweringinthethirteenthcenturycouldberightlysaidtorivalthegoldenage\nJORGEJ.E.GRACIAofancientphilosophyinthefourthcenturybcinGreeceandtheextraordinarydevelopmentsinthetwentiethcenturyinEuropeandAmerica.Thetemporalandterritorialboundariesofthemiddleagesareasubjectofcon-troversyamongscholars.Nomatterwhichdatesarepicked,however,itisclearthatbothAugustine(354–430)andJohnofStThomas(1589–1644)wereengagedinthesameintellectualprogrammeandthereforebelongtogether.BeforeAugustine,theintellectuallifeoftheWestwasdominatedbypaganphilosophy,andDescartes(1596–1650)(chapter26),generallyregardedasthefirstmodernphilosopher,wascontemporaneouswithJohn.Territorially,weneedtoincludenotonlyEurope,butalsotheMiddleEast,whereimportantGreekOrthodox,JewishandIslamicauthorsflourished.Aperiodthatextendsformorethanamillenniumisbynomeansuniformoreasilybreaksdownintosmallerunits.ThefirstofthesemightbecalledPatristic,andbeganinearnestwithAugustine,althoughitsrootswentbacktothesecondcenturybc.Itextendedtotheseventhcentury,andclosedwiththedeathofIsidoreofSeville(c.560–636),authoroftheEtymologies,thefirstofmanymedievalencyclopedias.BetweenthistimeandtheCarolingianRenaissancenothingofphilosophicalimportancetookplace.ThankstotheeffortsofCharlemagne(742–814)toestablishschools,regularizewriting,andgatherinhiscourtallthegreatmindsofthetimesinordertoencouragelearningandtoreplicatethegreatnessofRome,therewassomeimportantintellectualactivityattheendoftheeighthandthebeginningoftheninthcenturies,whichculminatedintheworkofJohnEriugena(c.810–75).Thisperiodwasfollowedbyadarkagewhichendedwithanother,morelastingrevivaloflearningintheeleventhandtwelfthcenturies.Thetwelfth-centuryrenais-sance,asitisoftencalled,producedsomeofthegreatestofallmedievalthinkers:Anselm(1033–1109),GilbertofPoitiers(c.1076–1154),PeterAbelard(c.1079–1142)andtheSchoolofChartres.Theperiodthatwentfrom1150toabout1225isofparamountimportance.Atthistimemanyoftheworksoftheancientsbecameavail-abletothemedievalsforthefirsttime,thankstotheconquestofterritorybyChristiansinSpain,andWesternscholarsengagedinafeverishattempttoassimilatethem.SomeoftheseworkshadbeentranslatedfromGreekintoSyriacintheMiddleEast,andlaterweretranslatedintoArabic.FromArabic,theyweretranslatedintoLatinwiththehelpofSpanishJews.AfewotherworkswererenderedintoLatindirectlyfromGreekoriginalsbyscholarsworkinginSicilyandSouthernItaly.Priorto1150,themedievalshadarathermeagregroupoftechnicalphilosophicalworksfromAristotleandhiscommentators,knownasthelogicavetus.ButinafewyearsnotonlythewholeOrganon,butmostotherworksofAristotle,withcommentariesbyIslamicauthors,andmanyscientificworksfromantiquity,becameavailable.Therenaissanceofthetwelfthcenturyandthefermentcreatedbythenewlyavail-abletextsgaverisetowhatisusuallyknownasscholasticism.Thisisamethodofteach-ingandlearningusedinvariousdisciplines,particularlyphilosophyandtheology.Theoriginofthetermistobefoundinmedievalschools,wherealecturer,particularlyonewhotaughttheliberalarts(triviumandquadrivium)wascalledscholasticus.TheaimofthemethodwastoyieldknowledgeconcordantwithbothhumanreasonandtheChristianfaith,aconcordiadiscordantiumofopinionswhichthemedievals620\nMEDIEVALPHILOSOPHYregardedasauthoritative.ThemethodwaspractisedinthemedievaluniversityandusedasatoolofAristotelianlogic.Asaresult,theliterarygenresusedbyscholasticsreflectuniversityactivitiesandsettings.Thecommentaryis,generallyspeaking,theproductofclassroomlecturesontexts;thequaestioistheproductofuniversitydisputations;andthesummaewerethetextbooksoftheage.AmongthefirstscholasticsofnotewereRogerBacon(c.1214–92)andAlberttheGreat(c.1200–1280),buttheywerefollowedbyahostoftoweringfigures:Bonaventure(c.1217–1274),ThomasAquinas,JohnDunsScotus(c.1265–1308)andWilliamofOckham(c.1285–1347).Inthemiddleofthefourteenthcentury,however,scholasticismsufferedanearlyirreversiblesetbackthroughtheBlackDeath(c.1347–51),whichdecimatedtheuniversitiesofEurope.IttookmorethanahundredyearstorecoverandgenerateasecondperiodofgreatnessundertheleadershipofSpanishscholasticsofthesixteenthcentury,suchasFranciscoSuárez(1548–1617)andFranciscodeVitoria(c.1483–1546).Thedistinguishingmarkofmedievalphilosophyistobefoundinitsdoubleaim:theunderstandingofChristianfaithanditsdefenceagainstthosewhoattackedit.Theeffortatunderstandingproducedtheologicalworks;theeffortatdefenceproducedapologeticworks.Thisdoesnotmean,however,thatthemedievalswerenotinterestedinpurelyphilosophicalproblems.Theywere,butmostoftenthereasonfortheirinter-estwasthatthesolutionstotheseproblemshadimportantimplicationsforChristiandoctrine;indeed,thesolutionsadoptedwereoftengovernedbythedoctrinalprinciplestheywishedtodefend.Inthissense,philosophywasgenerallysubordinatedtotheologyandapologetics.ThisattitudeseparatesthephilosophyofthemiddleagesfrombothancientandRenaissancephilosophy.TheapproachofmedievalphilosophyseparatesitfromancientphilosophybecausebothinclassicalGreeceandRome,philosophyenjoyedalargelyindependentstatusandapredominantposition.Philosophywasapursuitunsubordi-natedtoanyotherintellectualactivity,whosemaingoalwastheunderstandingoftheworldandman’splaceinit.Ontheotherhand,themedievalattitudeisquitedis-tinctfromthatoftheRenaissance,becausethehumanistslookedupontheclassicalpastasamodeloftheiractivityand,therefore,restoredmantothecentreofattentionandchannelledtheireffortstotherecoveryandemulationofclassicallearning,particularlyinthephilosophyofPlato.Incontrast,philosophyinthemiddleageswassubordinatedtotheology,andthecentreofintellectualattentionwasGod(chapter15)andhisrevelationratherthanhumanbeings;humanbeingswerestudiedonlyascreaturesofGodmadeinhisimageandlikeness.Themodelemulatedbythemedievalswasnottobefoundinthelivesandtheoriesofancientphilosophers,butinsteadinthelivesofsaintsandtheirprayers.Thecharacterofmedievalphilosophyisevidentinthephilosophicalproblemsmedievalschosetoaddress,thewaytheyinterpretedphilosophicalproblemstheyfoundinancienttexts,andthesolutionstheygavetomostoftheseproblems.Threeofthemostimportantproblemsthemedievalsinheritedfromtheancientsweretheproblemofhowweknow,theproblemofGod’sexistence,andtheproblemofuniversals.Threeproblemstheyraisedasaresultoftheirtheologicalconcernsandcommitmentsweretheproblemoftherelationbetweenfaithandreason,the621\nJORGEJ.E.GRACIAproblemofindividuation,andtheproblemconcernedwiththelanguageusedtotalkaboutGod.1FaithandReasonNootherissueconcernedthemedievalsmorethantherelationoffaithtoreason,forthesuccessoftheprogrammeadoptedintheagedependedtoalargeextentinturnonthesuccessinworkingoutthisrelationship.Forancientphilosophers,thishadnotbeenaconcern,formostofthemwerenotreligious,sotherewasnoneedtoreconcilereasontofaith,ortruthsderivedfromthestudyoftheworldindepend-entlyoffaithtoabodyofrevealedtruthsknownbyfaith.Underthisrubric,severalanddifferent,ifinterrelated,issuesarecontained.Theproblemisfirstexplicitlyfor-mulatedinthesecondcenturyoftheChristianera,whensomeearlyFathersofthechurchquestionedthemeritofusingsecularlearningbythosetowhomthetruthhasbeenrevealedbyGod.Twosidesareeasilyidentifiable.Somerejectedthevalueofsecularlearningaltogether;thispositionisoftencalledfideismbecauseofitsexclu-sivepreferenceforfaith.Othersfoundaplaceforsecularlearningintheunderstand-ingoffaith.Tertullian(c.160–220)arguedthatthereisnoplaceforthelearningofinfidelsinChristianity,andhecoinedaphrasethathasmadehistory:‘Ibelievebecauseitisabsurd’(Credoquiaineptum).AmongthosewhosawsomemeritintheuseofsecularlearningandtriedtobringittogetherwithrevealedtruthwasJustinMartyr(d.c.165).AugustinefollowedinthefootstepsofJustinMartyrandprovidedtheparametersforfuturediscussionsofthisissue.Forhim,alltruthisone,regardlessofthesource,sotheChristiancanandshouldmakeuseofsecularlearning.However,itisonlyintheChris-tianfaiththatonecantrulyunderstandtheworldandtheplaceofhumanbeingsinit.Christiandoctrinecompletes,illuminatesandtransformssecularlearning,provid-inganswerstoquestionswhicharemostimportantandforwhichnon-Christianshavenoanswers.Moreover,itsuppliesuswithaninfalliblecriterionoftruth.AnythingfoundinsecularlearningthatcontradictsChristiandoctrineisfalseandmustberejected;anythingconcordantwithitmaybeusedaslongasitisdoneinthecontextoffaith.Thecontroversybetweentheapproachofthosedenyingthevalueofsecularlearn-ingandthoseadvocatingitsusesurfacedagainintheeleventhandtwelfthcenturies.Thistimethefocuswasupontheuseoflogic,knownthenasdialectic,intheunder-standingofscriptures.Amongtheanti-dialecticianswasPeterDamian(1007–72),whowentsofarastorejectnotjustlogic,butevengrammarbecause,asheputit,theDevilbecamethefirstgrammarianwhenhedeclinedthewordDeusintheplural.Hisirrationalismwassostrong,andhisfaithinGod’spowersogreat,thathearguedthatGodcouldbringitaboutthatthepastneverhappened.Themostoutspokendialecti-cianwasAbelard,knownasthePeripateticfromPalletbecauseofhisuseofandpredilectionforAristotelianlogic.Inacontroversialbook,entitledSicetnon(YesandNo),AbelardshowedthatChristianauthoritiescontradicteachother,andthereforeanunderstandingofChristianfaithrequirestheuseoflogic.AmoremoderatepositionwasadoptedbyAnselm.InspiredbyAugustine,hearguedforameasureduseoflogic,622\nMEDIEVALPHILOSOPHYinwhichunderstandingbeginswithfaithbutisachievedwhenthedoctrinesrevealedinscripturesarearticulatedinlogicalform.Hisviewisencapsulatedintwofamousformulas:Credoutintelligam(IbelieveinorderthatImayunderstand)andFidesquaerensintellectum(Faithseekingunderstanding).TherelationbetweenfaithandreasonwasalsoofconcerntoIslamicandJewishthinkersduringthisperiod.Oneofthemostcontroversialviewsonthetopicwaspro-posedbyAverroes.AdoptingastrictAristotelianmodelofknowledgeasdemonstra-tion,hearguedthattheunderstandingofscripturescanneverreachthelevelofknowledge,forknowledgeisbasedondemonstrativereasoning,andreasoningfoundedonpremissesthatarenotself-evidentcanneverbeconsidereddemonstrative.Theologydoesnotyieldknowledgeproperlyspeaking,andthereforemustbesubordinatedtophi-losophy,whichdoes.Averroes’sposition,aswellasthepositionofthosewhopreferredreasonoverfaith,isusuallyreferredtoasrationalism.Inthethirteenthcentury,bothBonaventureandAquinasrespondedtoAverroes.BonaventurerejectedtheuniversalityoftheAristotelianmodelofknowledge,thoughheadmitteditscompetencewithinitsownsphere.Sinceallthingsinthecreatedorderare,forBonaventure,signsoftheUncreatedWisdom,eachsphereofrealitymustbeseeninitsconnectiontothatWisdom.Asaresult,althoughinanyonescienceknowl-edgecanbeacquiredwithoutrevelation,eachscienceanditssubjectneedstobetracedback(reducere)totheUncreatedWisdomforproperappreciationofitsrolewithinhumanlifeandthought.Hence,BonaventureprivilegesAugustinianwisdomoverandagainstAristotelianscience,rejectingthelatterasthehighestcanonofjudgementregardinghumanknowledge.IncontrasttoBonaventure,AquinasdidnotrejecttheAristotelianmodelusedbyAverroes,butratherarguedthatnotallknowledgeisofthesamesort.Someknowledgehaspremisseswhichareself-evidentprinciples–asisthecasewithmeta-physics–butsomehavepremisseswhichhavebeendemonstratedinotherbranchesofknowledge–aswithoptics,whichtakesitsprinciplesfromgeometry.Theologyisbasedonfaith,butitcanbeconsideredknowledgebecauseitrestsonGod’sownknowledge,whichisthehighestonetherecanbe.Aquinas,moreover,maderoomforboththeologyandphilosophyinthebodyofallknowledgebyarguingthatsometruthscanbeknownonlythroughfaith(forexample,ChristisGod),somecanbeknownonlythroughreason(forexample,allmaterialsubstancesarecomposedofmatterandform),andsomecanbeknownthrougheitherfaithorreason(forexample,Godexists).InspiteoftheeffortsofBonaventure,Aquinasandothers,theinfluenceofAverroescontinuedtobefeltwellintothesixteenthcenturyandpromptedrepeatedcondemna-tionsfromvariousquarters.Themostfamousoftheseoccurredin1277,andincludedevensomeviewswhichAquinashimselfhadheld.ThepopularityofAverroeswasmoststronglyfeltinthefacultyofartsratherthantheology.AmongthoseinthethirteenthcenturyaccusedoffollowingAverroestoocloselywasSigerofBrabant(c.1235–84).Hewaschargedwithholdingadoctrineofdoubletruth,accordingtowhichthereisatruthoffaithandatruthofreason,andthesetruthscanandoftendocontradicteachother.Clearly,thiswasunacceptabletomostmedievals,foritunderminedtheoverallprogrammeoftheage,thatis,theintegrationofrevelationandsecularlearningintoaconsistentbodyofdoctrine.623\nJORGEJ.E.GRACIAUnderstandingMedievalPhilosophy:ArgumentsfromAuthorityLetustakeacaseinpoint:thatofthemedievalattitudetoargumentsfromauthority.Atypicalmedievaltextwillcontainfrequentappealstoauthoritativestatements,drawnfromscripture,fromAugustine,say,orfromAristotleorCicero.Thisisnotafeatureofcontemporaryphilosophy:wequotephilosopherschieflytodisagreewiththem,orbecauseapreviousauthorhasmadeourpointbetterandmoreconciselythanwecan.Wedonot–or,atleast,weclaimwedonot–bringindictaof,forexample,Wittgensteintosupportourviews,withtheimplicationthatifWittgensteinsaidsoitmustbeso.Butthemedievalsseemtoactinjustsuchaway.Onereactiontothisistosaythatthereforethemedievalswerenotdoingwhatwewouldcallphilosophy,andthatthestudyoftheirphilosophymustbeprincipallyofhis-toricalinterest–‘merely’historicalinterest,somewouldsay,withtheimplicationthatmedievalphilosophyisasuitablestudyforthehistorianofideas,butnotforthephiloso-pher.Someonewhotakesthislinewillbeabletolearnfrommedievalphilosophy,atbest,whatthehistorianofsciencemightlearnfromtheworksof,forexample,thephlogistontheorists.Analternativereactionwouldbetoignoretheappealstoauthorityasirrelevant–toskipthem–andtoconcentrateonwhatacontemporaryphilosopherwouldregardasthe‘real’arguments.Suchaphilosopherwouldlearnsomethingofphilosophicalvaluefromthemedievals,butnotmuch:heorshewouldlearnfromamedievalauthorsolelywhatcouldhavebeenlearnedfromacontemporaryphilosopherwhohappenedtouseargumentsroughlyformallyparallel.Ifsuchreaderslearnanythingnewfromthemedievals,itwillonlybebecauseofthemerechancethatnocontemporaryauthorhappenstohavehitonroughlyparallelarguments.Neitherkindofreaderwilllearnwhatareadingofthemedievalstrulycanteachinthiscase:torethinkourcurrentpreconceptionsandprejudicesabouttheuseofauthor-ity.Medievaluseofauthoritywasconsciousandarticulate,notblindandall-trusting.Asithappens,weprofess,inourday,ablindandall-doubtingrejectionofauthoritywhich,bycontrastwiththenuancedviewsofthemedievals,canbeseenasverycrude.Themedievalsacceptedthefact,whichmostcontemporaryphilosopherswouldimplicitlyorexplicitlydeny,thatmostoftheinterestingandimportantthingsthatweknow,weknowfromauthority.Sincethemedievalsrecognizedthisfact,theywereabletoexamineit,whatweighttogivetoauthority,andwhentogiveit.Sincewerefusetorecognizethefact,weareallatsea;weblindlyacceptauthority,oftenenough,withoutevenknowingthatwearedoingso;and,atmomentswhereaconsciousacceptanceofauthoritymightbeofgenuineusetous,weconspiretorejectit.Byreadingthemedievalsintelligently,avoidingtheextremesmentionedabove,beingawareofhowdifferentwastheirthoughtfromours,butnotforthatreasonabandoningitasof‘merelyhistoricalinterest’,wecangenuinelylearnfromthemtocorrectthemistakesthatourageistoopronetomake.2God’sExistenceProvingthatGodexistswasimportantforthemedievalsbecauseGod’sexistence(chapter15)istheangularstoneonwhichtheChristianfaithrests.Itwasimportant624\nMEDIEVALPHILOSOPHYinorderbothtolaydownthefoundationofallChristiantheologyandtoestablishabaseforapologeticeffortsdirectedtowardMuslimsandJews.TheancientshadalreadyprovidedsomeargumentsfortheexistenceofGod,butitwasthemedievalswhoformulatedtheseinelegantandparsimoniousways.Theseargumentsbreakdownintotwotypes:argumentsbasedontheanalysisofconceptsandargumentsbasedonexperience.Ofthefirst,themostfamousaretheargumentsofAnselmintheProslogionandJohnDunsScotusinOntheFirstPrinciple.Bothhavecometobeknownasversionsoftheso-calledontologicalargument,atermfirstusedbyKanttodesignatethem.Ofthesecondtype,themostfamousaretheFiveWayspresentedinAquinas’sSummatheologiae,whichcomprisebothcosmologicalandteleologicalarguments.Anselm’sargumentderivesGod’sexistencefromtheconceptionofGodasthatthanwhichagreatercannotbethought.Godexists,forifhedidnot,thatthanwhichagreatercannotbethoughtwouldnotbethatthanwhichagreatercannotbethought.Anselmassumes,inlinewithhisAugustinian–Platonicframework,thatsomethingthatexistsisgreaterthansomethingthatdoesnot,thatthenotionofthatthanwhichagreatercannotbethoughtisintelligible,andthatlogicalnecessityhasabearingonexistence.Hehasbeencriticizedforallthreeassumptions.Buttothisdaytherearestrongsupportersofthesoundnessoftheargument.EachofAquinas’sFiveWaysbeginsbytakingnoteofafactgiveninexperience,suchasthatsomethingschange.Fromthistheygoontopointout,throughvarioussteps,thattheseexperiencescannotbeexplainedwithoutrecoursetoabeingwhoisulti-matelyresponsibleforthem,andthisbeingisGod.Thefirstwayarguesfromthefactthatthereischangeintheworldtoafirstcauseofthechange.Thesecondarguesfromtheefficientcausalityweexperienceintheworldtoafirstefficientcause.Thethirddistinguishesbetweennecessaryandcontingentbeings,andfromthisconcludesthattheremustbeonenecessarybeingwhosenecessaryexistenceisnotderivedfromanyotherbeing.Thefourtharguesfromthegradationfoundinthingstoabeingwhoisboththemaximumandthecauseofthosethings.Andthefiftharguesthatallthings,intelligentornot,actforanend,andtheremustbeanintelligentbeingwhodirectsthemtowardtheirend.3TheNamesofGodShowingthatwecanknowGodwasasimportanttothemedievalsasprovingthatheexists.Indeed,becausethelatterimpliesknowingsomethingaboutGod,onemightsaythatthetaskofshowingthatwecanknowGodlogicallyprecedesthetaskofprovingheexists.SeveralphilosophersfromantiquityhadtalkedaboutGod.TextsaboundinPlato,AristotleandtheStoicsthatspeakaboutasingledivinity.Inallthesecases,however,Godseemstohavebeenconceivedaspartoftheworld.KnowingGod,then,wasnotessentiallydifferentfromknowinganythingelse,evenifperhapsmoredifficult,forthetermsweusetotalkabouttheworldareinprincipleapplicabletoGodaswell.TheChris-tianconceptionofGod,however,changedthis.IfGodiswhollyotherthancreationandtranscendsit,thenitisquestionablethatthetermsweusetospeakabouttheworldcanalsobeappliedtohim.625\nJORGEJ.E.GRACIAUnderstandingtheNamesofGodThebackgroundofthecontroversyoverunderstandingtermsusedforGodisfoundinbothAugustine’swritingsandananonymoustreatiseprobablywrittenbyafifth-centurySyrianmonkwhoposedasDionysustheAreopagite,entitledOntheDivineNames.Con-troversiesoverthewaystounderstanddivinenamesheatupinthetwelfthandthirteenthcenturieswithMosesMaimonides(MoshebenMaimon,c.1135–1204),AquinasandScotus.Theissueconcernstheapplicationandunderstandingoftermswhichexpressperfections,suchas‘good’and‘just’;nooneheldthattermsexpressingimperfections,suchas‘bad’and‘unjust’,areapplicabletoGod.IftermsofthefirstsortdonotsignifyanythingaboutGod,thenitappearsthatwhenweusethemwedonotunderstandany-thinginparticularaboutGod;andiftheydo,thenitappearsthatweunderstandsome-thingaboutGodbutthatheisnotfundamentallydifferentfromtheworld.ThefirstmakesGodunknowableandthescripturesunintelligible;thesecondmakesGodpartoftheworldandthereforenotdivine.BothareunacceptabletoanorthodoxMuslim,JeworChristian.Almosteveryoneinthemiddleagestriedtofindasolutiontothisdilemma.MaimonidesarguedthattherearetwokindsoftermsapplicabletoGod.First,termsthatstandforattributesdonotsignifyanythingaboutGodhimself,butratheraretobeunderstoodnegatively,asdenyingsomethingofGod.TosaythatGodisgoodistosaythatheisnotevil,andtosaythatheisjustistosaythatheisnotunjust.Second,termsthatstandforactionsdoconveyinformation,buttheinformationtheyconveyisnotaboutGodhimselfbutaboutwhatGodhasdoneforothers.Attheotherextreme,Scotusarguedthat,inorderforthelanguagewepredicateofGodtobeeffectiveinproducingunderstanding,theremustbeatleastonetermthatisusedunivocally(thatis,withthesamemeaning)ofGodandcreatures,andproposed‘being’assuchaterm.TheunivocityofthistermgroundsourknowledgeofGodandmakesitpossibletospeakintelligiblyabouthim.Aquinasadoptedamiddleposition,betweenMaimonidesandScotus,withthedoctrineofanalogy.ThetermswepredicateofGodarenotusedequivocally(i.e.withdifferentmeanings)orunivocally,butanalogi-cally.‘Godisgood’doesnotmeanthatheisgoodlikeweare,orthatheisnotbadinthesenseweare;itmeansthatheisgoodinproportiontohisnatureandthusbetterthanweare,inasuperlativedegree,asthePseudo-Dionysushadalreadystated.4TheologyandMetaphysicsBecauseGodisatthecentreofourunderstanding,theremustbeadisciplinedevotedtohisstudy.Butwhichisthisdiscipline?Ontheonehand,itisclearthatthescripturesarethesourcewherewecanfindrevealedknowledgeofGod.But,ontheotherhand,theworldalsocontainsinformationaboutGodbecause,ascreator,hehaslefthisimprintonit.Indeed,thirteenth-centurytheologiansfoundtextsofAristotleintheMetaphysicsthatspokeofascienceconcernedwithGod.ThisgaverisetoaheatedcontroversyconcerningwhetherGodisstudiedintheologyorinmetaphysics.IntheIslamicworldalreadywefinddifferingviewswithrespecttothisissue.Avi-cennarejectedtheviewthatGodisstudiedinmetaphysicsbecausenoscienceproves626\nMEDIEVALPHILOSOPHYtheexistenceofwhatitstudiesandmetaphysicsprovestheexistenceofGod.Onthecontrary,AverroesarguedthatGodisstudiedinmetaphysics,becausehisexistenceisnotproveninthissciencebutinphysics.OntheLatinside,AquinasdistinguishedbetweenSacredDoctrine,thatistheologybasedonscriptures,andwhatwenowcallNaturalTheology,thatistheologybasedonthestudyoftheworld.Moreover,hecon-trastedbothofthesedisciplineswithmetaphysics.Ontheonehand,bothSacredDoc-trineandNaturalTheologystudyGod:thefirststudiesGodasrevealedinthescripturesandthesecondstudiesGodasrevealedincreation.Ontheotherhand,metaphysicsdoesnotstudyGodprimarily,butratherstudiesbeingquabeing,thatis,beinginsofarasitisneitherthiskind(e.g.human,divine)northisindividualbeing(e.g.Socrates,God).MetaphysicsstudiesGodonlysecondarily,astheFirstCauseofbeing.ScotusagreeswithAquinastotheextentthathetoobelievesthattheproperobjectofstudyoftheologyisGod,whereasthatofmetaphysicsisbeingquabeing.Thisapparentagreementbetweentwotoweringfiguresdidnothelptosettlethematter,however,fortheveryunderstandingofbeingquabeingwasatissue.Aquinasandhisfollowersarguedthatbeingquabeingistobeunderstoodasthelastact(esse)andperfectionofanessenceinanindividualentity,anddistinctinrealityfromtheessence.ButbothScotusandOckhamrejectedthisconceptionofbeing.Indeed,Ockhamevenrejectedthenotionthatanysciencehasasingleobjectofstudy.Accordingtohim,sci-encesaremerelycollectionsofmentalpropositionsandbecausethesepropositionshavedifferentsubjects,onecannotsaythatanysciencehasonlyonesubject.5HowWeKnowTheproblemofhowweknowbeingsotherthanGodwasintroducedintothemiddleagesbyAugustine’sdialogueOntheTeacher.Theostensiveproblemraisedinthisworkisthepurposeoftheuseofwords,buttherealunderlyingconcernistheoldPlatonicissueofwhetherwecanbetaught.Plato’sanswertothisquestionhadbeennegative:wecannotbetaughtbecausetheobjectsofknowledgeareimmaterialIdeas,andtheonlywaytoknowtheseisthroughadirectencounterwiththeminapreviouslife,whenwewerenotfetteredtothebody.OuronlyhopeforacquiringknowledgeinthislifeistoberemindedthroughlanguageoftheIdeasweonceknew.AugustinefollowedcloselyonPlato’sfootstepsbutbecause,asaChristian,hecouldnotacceptthepre-existenceofthesoul,hemodifiedthePlatonicscheme.ChristbecomestheteacherwhoplacesIdeasinourmemoryanditistherethatweencounterthembybeingremindedofthemthroughwords.Augustine’sviewbecameknownastheDoctrineofIllumination,becauseheusedthePlatonicmetaphoroflighttodescribehowChristmakesusseeideas:ChristisliketheSunwhichilluminesourmindswithknowledgeofintelligiblerealities.ThisdoctrineturnedintooneofthemostimportantbattlegroundsbetweenAugustiniansandAristoteliansinthelatermiddleages.AlmosteveryoneacceptedAugustine’smetaphor,butthatiswheretheagreementended.BonaventureandHenryofGhent(c.1217–1293),amongothers,triedtoanswersomeofthequestionsraisedbythedoctrineandtoresolvesomeofitsambiguities,butAquinasandScotusopposedtheseinterpretations.ThefirstarguedthatthelightaboutwhichAugustine627\nJORGEJ.E.GRACIAwasspeakingisnoneotherthanthenaturallightofreason,sothatilluminationisanaturalratherthanasupernaturalprocess.Scotus,althoughaFranciscan,opposedhisbrothersinthis.HearguedthatHenryofGhent’sinterpretationofAugustineleadstoscepticism,andknowledgeispossiblewithoutilluminationunderstoodinasupernaturalway.6UniversalsBothAristotleandPlatohadmadeclearthatknowledgeproperlyspeakingisoftheuniversal,andtheauthorityofAugustinehadaddedfurthersupportforthisview.Knowledge,inastrictsense,isnotaboutthisorthatcat,butaboutcat,notaboutthismanorthatman,butaboutmaningeneral.Themedievalsgenerallyacceptedthis,butatthesametimemostofthemheldthatnotjustsubstancesintheAristoteliansense(forexample,thiscat,thisman),butalsothefeaturesofsubstances(forexample,acat’sblackfurcolour,aman’shumanity)wereindividual.Thisposedahostofepistemologicalandmetaphysicalproblems,oneofwhichisknownastheproblemofuniversals.TheProblemofUniversalsIntheearlypartofthemedievalperiod,theproblemofuniversalswasframedintermsofthreequestionsPorphyrythePhoenician(c.232–304)hadaskedintheIsagogecon-cerninggeneraandspecies,andwhichthemedievalsfoundinBoethius’stranslationofthatwork:(1)Arethingslikeanimalandmansomethinginthemindonlyoralsosome-thingoutsidethemind?(2)Iftheyaresomethingoutsidethemind,aretheymaterialorimmaterial?And(3)aretheysomethingseparateanddifferentfromindividual,sensiblethings,orsomethinginthemandlikethem?Boethiushimselfgaveratherambiguousanswerstothesequestions,whichleftmuchforotherstodo.Roughly,heheldthatanimalandmanarebothsomethinginthemindandsomethingoutsidethemind.Theyareunderstoodinonewayinthemindandexistinanotherwayinthingsoutsidethemind;inthemindtheyareunderstoodasuniversal,whereasoutsidethemindtheyareindividualandsensible.Moreover,explicitlyadoptinganAristotelianstance,whichhejustifiedbecausehewascommentingonaworkdealingwithAristotle,herejectedtheviewthatgeneraandspeciesexistseparatelyfromindividualthingsoutsidethemind.ChallengedbyBoethius’sanswerstohisquestionsaboutgeneraandspecies,subse-quentauthorsdevelopedmanypositionsintheearlymiddleages.TheyrangedfromtheextremerealismofEriugena,accordingtowhomgeneraandspeciesarePlatonicIdeas,totheextremenominalismofRoscelin(c.1050–1120),whoheldtheyaremereindi-vidualutterances.ThemostsophisticatedviewwasofferedbyAbelard,whoarguedthatuniversalsarewordswhicharecreatedtobepredicatedofseveralthings.Althoughthesewordsdonotcauseanunderstandingofanyindividualthinginparticular,but628\nMEDIEVALPHILOSOPHYratherofaconceptioncommontomanyofthemwhichthemindcontrives,thecauseoftheirimpositionistobefoundinthestatusofindividualthings.Thestatusitselfisnotathing,oranykindofreality,butmerelywhatthingsare.ThestatusofSocratesandPlatoisman,butmanisnoentityotherthanSocratesandPlato.InspiteofthesophisticationofAbelard’stheory,thereweremanyquestionsthatitleftunansweredandwhichweretakenupbysubsequentauthors.InthethirteenthcenturythetermsofthecontroversychangedsomewhatbecauseoftheintroductionofnewterminologyfoundintherecenttranslationsofAristotleandthecommentariesonthembyAverroesandAvicenna.Insteadofspeakingaboutgenera,speciesoruniversals,thetalkchangedtonatures.Moreover,thequestionwasframedintermsoftheirunityandbeing:whatkindofbeingandunitydonatureshave?TheclassicmoderatepositionwastakenbyAquinas,whoarguedthatnaturescanbeconsideredabsolutelyorinrelationtothemindorindividualthings.Absolutely,onlywhatisincludedintheirdefinitionsbelongstonatures.Therefore,theycannotbesaidtohavebeingorunity,butneithercantheybesaidtolackthem.Becausethedefinitionofthenature‘man’is‘rationalanimal’,onlyanimalityandrationalitycanbesaidtobelongtomanconsideredabsolutely.Andbecausebeingandunity,justlikewhiteness,arenotpresentinthedefinition,thesecannotbesaidtobelongtomanconsideredasman,butneitheraretheysupposednottobelongtoit.Thenature‘man’isasneutralwithrespecttobeingandunity,asitiswithrespecttowhiteness.Beingandunitybelongtonaturesonlywhentheyareconsideredinrelationtothemindortoindivid-ualthingsoutsidethemind.Inrelationtothemind,naturesareconceptsproperlyspeakingand,therefore,areuniversalandhavementalbeing.Inrelationtoindividualthings,naturesareindividualandhaveindividualbeing.Man,whenunderstood,hasbothbeingandunity,thebeingpropertothemind,whereitisfoundasaconcept,andtheunitypropertouniversals,becauseitcanbeusedtothinkaboutnotanymaninparticularbutabouteachandeveryman.Man,consideredinrelationtoindividualmen,hasbothindividualbeingandunity,thebeingandunityofeachmanwhereitisfoundastheirnature.BothScotusandOckhamdevelopedviewsthatdisagreedwiththatofAquinas,butinoppositedirections.ScotusmovedclosertorealismandOckhamclosertonominal-ism.ForScotus,naturesconsideredabsolutelyhaveabeingandunitypropertothem-selves.Thus,inindividuals,natureshaveadoubleunityandadoublebeing,theirownandthatofindividuals.Manhasabeingandunitypropertonatures,sothatinthismanthereisadoublebeingandunity:thebeingandunityofthenatureandthebeingandunityoftheindividual.OckhamwasquitedissatisfiedwiththisviewandappliedtoithisfamousRazor,accordingtowhichexplanationsshouldnotmultiplyentitiesbeyondnecessity.Forhim,thereisnosuchathingasanatureconsideredabsolutely;thereareonlyuniversalconceptsinthemindandindividualthingsoutsidethemind.Thenotionofanatureconsideredabsolutely,whetherthatnatureisconceivedneutrallyasAquinasdid,orashavingsomebeingandunityasScotusdid,issuperfluous.Theexistenceofuniversalconceptsinthemindcanbeexplainedintermsofthenaturalcapacityofthemindtoformageneralconceptbasedontheparticularexperienceofindividuals.629\nJORGEJ.E.GRACIA7IndividuationThoseauthorswhoattributedsomestatustonaturesinthingsoutsidethemindnaturallyaskedthemselvesthequestionofwhatitisinthingsthatmakesthemindi-vidual.Ifallthetermswepredicateofindividualthingsindicatesomethinguniversalorcommoninthem,whatisindividualinthings?Thiswasaparticularlyimportantquestionformedievalauthors,andonewhichhadbeengenerallyneglectedbytheancients.BothPlatoandAristotlehadtalkedaboutindividuals,buttheirprimaryconcernwaswithuniversalsandtheirstatus.Forthemedievals,theorderofimpor-tancewasreversed,becauseforthemGodwasnotuniversalandhadevenbecomeanindividualpersonintheworld.Moreover,God’screationwasconceivedasindividualandendowed,asAugustinehadpointedout,withavaluehigherthantheideasthroughwhichweknowit.ThefirstauthortoraisequestionsconcerningindividuationwasBoethiusinOntheTrinity,atreatisedevotedtotheexplanationofhowGodcanbebothonesubstanceandthreepersons.Forhim,individualityistheresultofthebundleofaccidents(thatis,offeatureswhicharenotnecessarytothething)substanceshave,andultimately,iftheyhaveallotheraccidentsincommon,oftheplacetheyoccupy.Althoughthisviewiscontroversial,itenjoyedenormouspopularitythroughouttheearlymiddleages.AfterAbelard’schallengeinthetwelfthcentury,however,itwasgenerallyrejected.Hearguedthataccidentscannotindividuateasubstancebecauseasubstanceispriortoitsaccidentsinsofarasparticularaccidentsarenotnecessaryforthesubstance.Ockhamandotherconceptualistsandnominalistsdidnotthinktheyneededtofindaprincipleofindividuationbecausetheyheldthatonlyindividualthingsexistanduni-versals,ornatures,arenothingbutconceptsproducedbymentalprocesses.Realists,however,whohelduniversalsornaturesaresomethingrealoutsidethemind,hadtoidentifyaprincipleofindividuation.Apopularviewwastoholdthatsubstancesareindividualowingtotheirmatter.InanAristotelianframework,wheresubstancesarecomposedofmatterandform,andformiscommon,thisviewmakessenseprimafacie.Uponfurtheranalysis,however,itappearsthatmatteralsoiscommonandthismakesitdifficultforittoindividuate.Aquinas’sresponsewastoproposethatitisnotmatterbyitselfthatindividuates,butrathermattertakentogetherwithquantity,whichheunderstoodasdimensions.ThiswasunsatisfactorytoScotus,whopointedoutthatquantityisascommonasmatterandthereforethecombinationofthetwocannotexplainindividuality.Instead,heproposedasuigenerisprincipleofindividuation,afor-malityhecalledthisness.Thisisanunanalysableandindefinableprinciplewhoseonlyfunctionistoindividuate.Eachindividual,then,hasacommonnaturewithaunityandbeingpropertoitself,andalsoaprincipleofindividuationwhichmakesitathis.Thisprincipleandthecommonnaturearedistinguishedmorethanconceptsare,butlessthanrealthingsare;theyaredistinguishedformally.8ConclusionTheproblemsdiscussedaboveprovideonlyasmallsampleofthemanythatthemedievalsaddressed.Indeed,exceptforproblemsonlysubsequentlyraisedbecauseof630\nMEDIEVALPHILOSOPHYadvancesinscienceandtechnology(forexample,artificialintelligence),themedievalsseemtohavetoucheduponmostofthephilosophicalproblemsofperennialinterest.Althoughmedievalphilosophyissignificantlydifferentfromcontemporaryphiloso-phyinsofarasitisprimarilyconcernedwiththeintegrationofrevelationandsecularlearning,nonethelessithasmuchincommonwithit.Forexample,itshareswithanalyticphilosophyanemphasisonlinguisticprecision,theuseoftechnicallanguage,anargumentativespirit,andtheviewthatphilosophicalproblemscanbesolvedbydrawingdistinctions.Anditshareswithcontinentalphilosophyaconcernwithbeingandtheexistentialissuesthataffecthumans.Muchcanbefoundinmedi-evalphilosophy,therefore,thatshouldbeofinteresttocontemporaryphilosophersnotjustasamatterofantiquariancuriosity,butalsoasasourceofphilosophicalunderstanding.FurtherReadingGoodgeneralintroductionstomedievalphilosophyareCopleston(1946–75,1972),Knowles(1988)andMarenbon(1983,1987).Forgeneralinformationaboutparticularauthors,periodsandtopics,seeGraciaandNoone(2002).ImportantandfairlyaccessibletextsaretheConfes-sions(1991)andOnFreeChoiceoftheWill(1964)ofAugustine,TheConsolationofPhilosophyandTheTheologicalTractatesofBoethius(1968),andtheMonologionandProslogionofAnselm(1973).ForanintroductiontoAquinas,readCopleston(1976)orKenny(1980).Spade(1999)isagoodintroductiontoOckham.ReferencesMedievalphilosophersAbelard,P.1954:TheStoryofAbelard’sAdversities(translatedbyJ.T.Muckle).Toronto:Pontifi-calInstituteofMediaevalStudies.Anselm,St1973:ThePrayersandMeditationsofStAnselm(translatedbyB.Ward).Harmondsworth:PenguinBooks.Aquinas,StThomas1963–75:Summatheologiae(editedbyT.Gilby).London:EyreandSpottiswoode.——1988:ThePhilosophyofThomasAquinas:IntroductoryReadings(editedbyC.E.J.Martin).London:Routledge.Augustine,St1964:OnFreeChoiceoftheWill(translatedbyAnnaS.BenjaminandL.H.Hackstaff.Indianapolis,IN:Bobbs-Merrill.——1991:Confessions(translatedbyH.Chadwick).Oxford:OxfordUniversityPress.Averroes1961:OntheHarmonyofReligionandPhilosophy(translatedbyGeorgeF.Hourani.London:Luzac.Boethius1968:TheTheologicalTractatesandTheConsolationofPhilosophy(translatedbyH.F.StewartandE.K.Rand).Cambridge,MA:HarvardUniversityPress.DunsScotus,J.1962:PhilosophicalWritings:ASelection(editedandtranslatedbyA.Wolter).London:Nelson.Maimonides1956:TheGuideforthePerplexed,2ndrevdedn(translatedbyM.Friedlander).NewYork:DoverPublications.Ockham,Williamof1957:PhilosophicalWritings:ASelection(editedandtranslatedbyP.Boehner).London:Nelson.Pseudo-Dionysus1966:OntheDivineNamesandtheMysticalTheology(editedandtranslatedbyC.E.Rolt).NewYork:Macmillan.631\nJORGEJ.E.GRACIAOtherwritersBroadie,A.1993:IntroductiontoMedievalLogic,2ndedn.Oxford:OxfordUniversityPress.Chadwick,H.1986:Augustine.Oxford:OxfordUniversityPress.Copleston,F.A.1946–75:HistoryofPhilosophy,vol.2.Westminster,MD:NewmanBookshop.——1972:HistoryofMedievalPhilosophy.London:Methuen.——1976:Aquinas.London:SearchPress.Gibson,M.M.T.1981:Boethius:HisLife,ThoughtandInfluence.Oxford:Blackwell.Gilson,E.1961:TheChristianPhilosophyofStAugustine(translatedbyL.E.M.Lynch).London:Gollancz.Gracia,J.J.E.andNoone,T.(eds)2002:TheBlackwellCompaniontoPhilosophyintheMiddleAges.Oxford:Blackwell.Kenny,A.J.P.1980:Aquinas.Oxford:OxfordUniversityPress.Kirwan,C.1989:Augustine.London:Routledge.Knowles,D.1988:TheEvolutionofMedievalThought,2ndedn.London:Longman.Kretzmann,N.andStump,E.(eds)1993:TheCambridgeCompaniontoAquinas.Cambridge:CambridgeUniversityPress.Kretzmann,N.,Kenny,A.J.P.andPinborg,J.(eds)1982:TheCambridgeHistoryofLaterMedievalPhilosophy.Cambridge:CambridgeUniversityPress.Marenbon,J.1983:EarlyMedievalPhilosophy(480–1150):AnIntroduction.London:Routledge.——1987:LaterMedievalPhilosophy(1150–1350):AnIntroduction.London:Routledge.Spade,P.V.(ed.)1999:TheCambridgeCompaniontoOckham.Cambridge:CambridgeUniversityPress.DiscussionQuestions1‘Ibelievebecauseitisabsurd.’Doesthisleaveanyroleforphilosophy?2‘IbelieveinorderthatImayunderstand.’Canbeliefenableustounderstand?3Canwegainknowledgethroughrevelation?Howwouldknowledgegainedthroughrevelationberelatedtootherhumanknowledge?4Iftheologyhadalegitimateroletoplayinmedievalphilosophicalthought,howcanwedecidewhetherithasasimilarroletoday?5Amonglogic,metaphysicsandtheology,doesonedisciplinehavepriorityovertheothers?6Isthereasinglesystemforunderstandingandassessingallclaimstoknowledge?7Cansometruthsbeknownonlythroughfaith,somebeknownonlythroughreasonandsomebeknownthrougheitherfaithorreason?8Whatexplanationsarepossibleifreasonandfaithappeartocontradictoneanother?9Whatisnecessaryexistence?Doesabeingthanwhomnogreatercanbeconceivednecessarilyexist?10CanweusefactsgiveninexperiencetoprovethatGodexists?11CanthetermsweusetospeakabouttheworldalsobeappliedtoawhollytranscendentGod?12MustGodbeeitherunknowableorpartoftheworldandthereforenotdivine?13Canwedeterminewhetheratermisusedequivocally,univocallyoranalogically?14Canwestudy‘beingquabeing’?15Mustanysciencehaveasingleobjectofstudy?16Doesthemetaphorofilluminationexplainhowweacquireknowledge?632\nMEDIEVALPHILOSOPHY17Isknowledgeofmanaboutmaningeneralratherthanaboutthismanandthatman?18Couldmanbebothsomethinginthemindandsomethingoutsidethemind?19Coulduniversalsbewordsthatarecreatedtobepredicatedofseveralthingsratherthanbeingaspecialkindofentity?20Isitusefultodistinguishbetweenabsoluteandrelativeunderstandingofnatures?21Mustnatureseitherhavebeingandunityorlackbeingandunity?22Howshouldweunderstandtheclaimthatsomethinghasmentalbeing?23Whichexplanationsmultiplyentitiesbeyondnecessity?24Donaturesinindividualshaveadoubleunityandadoublebeing,theirownandthatofindividuals?25Doesournaturalcapacitytoformgeneralconceptsbasedontheparticularexperienceofindividualsexplaintheexistenceofuniversalconceptsinthemind?26Whatmakessomethingindividual?27Canthisnessexplainindividuation?28Whatcancontemporaryphilosophylearnfrommedievalphilosophy?633\n25BaconSTEPHENGAUKROGERFrancisBacon(1561–1626)wasinstrumentalineffectingamajorshiftinmentalityfromacontemplativetoanempiricalapproachtothenatureofthephysicalworld.Hisachievementwastwofold.First,hetransformedthedisciplineofphilosophyfromsomethingcontemplativewhichfocusedaboveallonmoralquestionsintosomethingpracticalwhichfocusedcentrallyonquestionsinnaturalphilosophy(whatisnowcalledscience).Secondly,hesetoutanaccountofproceduresforscientificenquirywhichhasresultedinhisbeingconsideredoneofthefoundersofmodernSCIENTIFICMETHOD(chapter9).Applyingideasforreforminitiallydevelopedintheareaoflawtonaturalphilosophy,hismethodtakestheformofinduction(aprocedurethatmovesinferentiallyfromobservableeffectstodeeperunderlyingcauses)whichproceedsbymeansofeliminationofvariouspossibleexplanationsbytestingtheirconsequencesagainstexperimentorobservation.1IntroductionBaconwasbroughtupinaRenaissancehumanistcontext,andwasemployedthrough-outhislifeinseniorlegalpositionsingovernment,endingupasLordChancellor,untilhisimpeachmentin1621.Humanisteducationcentreduponrhetoric,anditwasrhetoricandthelawthatguidedhisthought.Whatwasunusualabouthisapplicationofpreceptslearnedfromrhetoricandlawtonaturalphilosophywasthatheusedthemtoproposeafundamentalreformofphilosophy.Philosophyasadisciplinewasatalowebbinsixteenth-andearlyseventeenth-centuryEngland.Acontrastwasoftendrawninclassicaltermsbetweenthelifeofcon-templation(otium)andthelifeofpractical,productiveactivity(negotium),andtherewasadecisiveshiftinfavourofthelatterinsixteenth-centuryEngland.Therewasastressonpracticalquestions,andthepracticalusesoflearning,andphilosophywaswidelyregardedasauselessdisciplinewhichfosteredargumentforitsownsake,nevergettinganywhereandneverproducinganythingofvalue.Moreover,moralitywaswidelyseenasthekeyphilosophicaltopic(followingtheCiceronianmodelcurrentinRenaissanceEuropegenerally),andanumberofElizabethanthinkers,mostnotably\nBACONthepoetSirPhilipSidney,werearguingthatpoetrywassuperiortophilosophy,inthatphilosophycouldonlydiscourseonthenatureofgoodness,whereaspoetrycouldactuallymovepeopletogoodness,whichwasthepointoftheexercise.Bacondidtwothings:heshiftedphilosophyfromotiumtonegotium,andhemadenaturalphilosophyreplaceMoralphilosophy(chapter6)asthecentreofthephilo-sophicalenterprise.Thecombinationofthesetwo(andtheyareintimatelyconnected)isaradicalmovethatmarksadecisivebreakwithearlierconceptionsofphilosophyand,ratherimportantly,withearlierunderstandingsofjustwhatthejobofthephiloso-pherwas.2TheReformofPhilosophyanditsPractitionersNaturalphilosophyexistedinanumberofformsinthesixteenthandseventeenthcen-turies,andthereweretwoextremeforms.Thefirstwasexemplifiedinalchemy,whichwasanesotericbutpracticaldisciplinewhichhadlittleconnectionwithtraditionalphilosophicalpracticeandwhichsuffered,inBacon’sview,fromalackofstructure,sothatwhatfewresultswereachievedwereachievedbychance.Attheotherextremewasscholasticnaturalphilosophy,anintenselytheoreticaldisciplinewhich,inBacon’sview,producednothingatall,despiteitsgreatsophistication,whichturnedouttobealmostexclusivelyverbal.Baconwantedsomethingthatcoulddelivertheadvantagesofeachofthesewithoutanyofthedisadvantages.Hewantedsomethingthatwouldprovideadetailedtheoreticaloverviewofthenaturalrealmsuchthatnaturalprocessescouldnotonlybeunderstoodbut,moreimportantly,transformedonthebasisofthisunderstanding:thisisthecontextofhisfamousdictumthat‘knowledgeispower’.Theultimateaimwastotransformnaturalprocessesforthecommongood(wherethecommongoodwasverymuchsomethingtobedecidedbythesovereign,onBacon’sview),anditwasthis,ratherthansomecontemplativeunderstandingofnature,thatprovidedtherationalefornaturalphilosophy,andbyextensionphilosophyperse.Bacon’sfirstattemptsatreformwereintheareaofLaw(chapter13)ratherthannaturalphilosophy,wherehewasconcernedtosystematizethelaw,provideregularrecordsandreviewsoflegaldecisions,andthentrytodiscoversomefirmfoundationsforlegalpractice.Thelawworkedwithelaborateproceduresforgathering,assessingandtestingevidence.Moreover,itwasanareaoftheoreticalsophisticationwhollydevotedtopracticalends.ThiswasexactlythekindofthingthatBaconhadinmindfornaturalphilosophy,althoughnaturalphilosophywasinafarworsestatethanthelaw.3AMethodofDiscovery:FromRhetorictoScienceThelawdidnotactasamodelinitsownright,however:itsimportancearosefromthefactthat(especiallyinareformedstate)itexemplifiedarhetoricallymotivatedaccountofdiscovery.ThisholdsthekeytoBacon’senterprise.Hiseducation,likethatofanyotherschoolboyintheWestintheearlymodernera,wasintheliberalarts,thestudyofwhichunderlayabroadrangeofareas,includinglawandPolitics(chapter8),on635\nSTEPHENGAUKROGERtheonehand,andtheissuesofscientificdemonstrationanddiscoveryontheother,andthemostcrucialpartoftheliberalartsinthisrespectwasrhetoric.Atitsmostgenerallevel,thetaskofrhetoricwastheformulation,organizationandexpressionofone’sideasinacoherentandcompellingway.Itwasdesignedtohelponefindone’swayaroundthecomprehensivebodyoflearningbuiltupfromantiquity,torecognizewhereappropriateevidenceandargumentsmightbefound,toprovidemodelswhichweredesignedtogiveoneasenseofwhatwasneededifaparticularquestionwastobeinvestigated,oraparticularpositiondefended,modelsthatwouldbesharedwiththosetowhomonewasexpoundingordefendingone’scase.Itwasdesignedtohelponefocusone’smentalpowersinvariousways,toorganizeone’sthoughtsinthemosteconomicalfashion,aswellasprovidingmodelstoshowonehowparticularkindsofcasewerebestdefended,dependingonsuchfactsastheavailabilityandcomplexityoftheevidence,andthestateoftheknowledge,opinionsandpreju-dicesoftheaudiencetowardswhichonewasdirectingone’sarguments.Atagenerallevel,rhetoricwasindifferentastosubjectmatter,inthatverycomprehensivepro-cedureswererecommendedthatwouldaidone’sinvestigationsorone’scaseirrespec-tiveofwhetheronewereconductingascientificinvestigationoralegalone,althoughataspecificleveltherewouldbesimilaritiesoranalogies(asregardsthestandingofvariouskindsofevidence,forexample)anddissimilarities(asregardsthemeansbywhichonecollectedevidence,forexample)betweenlegalcasesandthoseinnaturalphilosophy.Thelaw,takeninabroadsense,wasverymuchaparadigmcaseforrhetori-calwriters:rhetoricaltreatiseswereoftenseenexplicitlyasbeingdirectedtowardslawyersandlegislators,andexamplesweregearedaroundthekindsofproblemcasethataroseinlaw.Inthelightofthis,itisonlytobeexpectedthatusingarhetoricalmodelforknowledge–thatis,amodelthatgivesdirectiononhowtocollectandassessevidenceforaview,howtomakeajudgementonthebasisofthatevidence,andhowtoestablishthecorrectnessofone’sjudgement,usingpreceptsderivedfromthestudyofrhetoric–isinmanyrespectsusingalegalmodel.TheseconnectionsareparticularlystronginthecaseofBacon’sattemptstoreformnaturalphilosophy.4TheDoctrineofIdolsIfrhetoricisthefirstingredientinBacon’saccountofmethod,thesecondisadis-tinctiveunderstandingofwhytheneedformethodarises.HereBacon’sstressonapsychologicaldimensiontoknowledgeisimportant:questionsofpresentationofknowledgearenotonlyrecognizedtobeimportant,buthavetobeunderstood,wheresuchanunderstandingisnotsupplementarytoEpistemology(chapter1)butactuallypartofit.Thereisnothingnewinthisatonelevel,foritissimplypartofalongtradi-tionwhichbeginsinearnestwiththeRomanrhetoricians;butalthoughitborrowsfromGreekwriters,itisratherdifferentfromtheapproachtoepistemologicalquestionsthatwefindintheclassicalGreekphilosophers(chapters22–3)andHellenisticphilosophers(chapter22).WhenonethinksofBacon’sgeneralprojectinthiscontext,itbecomesclearthatthereissomethingnovelhere.Fornaturalphilosophyhadgener-allybeenthepreserveofGreekphilosophy,andhadbeenpursuedinasimilarwaybyScholasticphilosophers(chapter24).TheRomantradition,withtheexceptionof636\nBACONLucretius,hadgenerallyspeakingnotconcerneditselfwithspeculativenatural-philosophicalquestions,dealinginsteadwithpracticalmoral,politicalandlegalquestions.Inthinkingofpersuasionintermsofapsychologicaltheory,inthinkingofpsychologicaltheoryaspartofepistemology,andinthinkingofepistemologyasbeingdirectedprimarilytowardsnaturalphilosophy,Baconcanprovidehimselfwithsomeoftheresourcestostartthinkingthroughnaturalphilosophy,notasaspecula-tivebutasapracticaldiscipline.IdolsoftheMindThepsychologicaldimensiontoepistemologyisbroughtoutfullyinBacon’sdoctrineofthe‘IdolsoftheMind’.Thesecondpartofthe‘GreatInstauration’,whichaimsattherenewaloflearning,isdevotedtothe‘Inventionofknowledge’,andhastwocomponents,oneofwhichaimstoridthemindofpreconceptions,whiletheotheraimstoguidethemindinaproductivedirection.Thesecomponentsareinterconnected,foruntilweunder-standthenatureofthemind’spreconceptions,wedonotknowinwhatdirectionweneedtoleaditsthinking.Inotherwords,variousnaturalinclinationsofthemindmustbepurgedbeforethenewprocedurecanbesetinplace.Bacon’sapproachhereisgenuinelydifferentfromthatofhispredecessors,asherealizes.Logicormethodinthemselvescannotsimplybeintro-ducedtoreplacebadhabitsofthought,whichBaconidentifiesas‘Idols’,becauseitisnotsimplyaquestionofreplacement.Thesimpleapplicationoflogictoone’smentalprocessesisinsufficient.InhisdoctrineofIdols,Baconprovidesanaccountofthesystematicformsoferrortowhichthemindissubject,andthisisacrucialpartofhisepistemology.Itisinhistreatmentofinternalimpediments,theIdolsofthemind,thatthequestionisraisedofwhatpsychologicalorcognitivestatewemustbeintobeabletopursuenaturalphi-losophyinthefirstplace.BaconbelievesanunderstandingofnatureofakindthathadneverbeenachievedsincetheFallispossibleinhisowntimebecausethedistinctiveobstaclesthathaveheldupallpreviousattemptshavebeenidentified,inwhatisinmanyrespectsanoveltheoryofwhatmighttraditionallyhavebeentreatedunderatheoryofthepassions,onenowdirectedspecificallyatnatural-philosophicalpractice.TheIdolsoftheTribederivefromhumannatureitself,andaffecteveryoneequally.Theyaremanifestedineagernesstosupposethatthereismoreorderandregularityinnaturethanthereactuallyis;inthetendencytoneglectorignorecounter-examplestoone’stheories;inthetendencytoextrapolatefromstrikingcaseswithwhichoneisfamiliartoallothercases;intherestlessnessofthehumanmind,whichmeansitisnotsatis-fiedwithperfectlygoodfundamentalexplanations,mistakenlyandconstantlyseekingsomemorefundamentalcauseadinfinitum;andinthetendencytobelievetruewhatonewouldliketobetrue.TheIdolsoftheCave,wearetold,‘taketheirriseinthepecu-liarconstitution,mentalorbodily,ofeachindividual;andalsoineducation,habit,andaccident’.Theyincludefascinationwithaparticularsubject,whichleadstoover-hastygeneralization;thereadinessofsomemindstofocusondifferences,andsometofocus637\nSTEPHENGAUKROGERonsimilaritiesandresemblances,whileabalanceisdifficulttoattainnaturally;thefactthatsomemindsareoverlyattractedtoantiquityandsometonovelty;finally,therearethosewhoareconcernedwhollywithmaterialconstitutionattheexpenseofstructure(theancientatomists),andthosewhoareconcernedwhollywithstructureattheexpenseofmaterialconstitution.TheseexamplesbringtolightaverysignificantdifferencebetweentheIdolsoftheTribeandIdolsoftheCave.Thereseemstobeasetofroutineproceduresonecangothroughtoremedythesituationinthelattercase,procedureswhichareprovidedbythepositivepartofBacon’sdoctrineof‘eliminativeinduction’,whereasthecaseofIdolsoftheTribeis,inmostcases,muchmoredifficulttoremedy.TheIdolsoftheMarket-placederive,inessence,fromthefactthatwehavetoexpressandcommunicateourthoughtsbymeansoflanguage,whichcontainssystematicdeficiencies:itprovidesnameswhichrefertothingsthatdonotexist,suchas‘Fortune’and‘PrimeMover’.Thesolutionhereissimplytogetridofthetheoriesthatgiverisetothesefictitiousentities.Asecondkindofcaseisnotsostraightforward.Itarisesbecausewordshavemultiplemeanings,and/orill-definedmeanings,orboth,andthisisespeciallysointhecaseoftermssuchas‘humid’,whichhavebeenabstractedfromobservation.Finally,thefourthkindsofimpediment,the‘IdolsoftheTheatre’,areneitherinnateinthemindnorinlanguage,butareacquiredfromacorruptphilosophicalcultureanditsperverserulesofdemonstration.Hereageneralremedyisavailable,namelyfollowingBacon’spositivemethodologicalprescriptions.OneofthegreatvaluesofBacon’saccountoftheIdolsisthatitallowshimtomakethecaseformethodinaparticularlycompellingway.Indeed,neverhastheneedformethodbeensetoutmoreforcefully,forBacon’sadvocacyofmethodisnotsimplyasanaidtodiscovery.Wepursuenaturalphilosophywithseriouslydeficientnaturalfaculties,weoperatewithaseverelyinadequatemeansofcommunication,andwerelyonahopelesslycorruptphilosophicalculture.Inmanyrespects,thesearebeyondremedy.Thepractitionersofnaturalphilosophycertainlyneedtoreformtheirbehav-iour,overcometheirnaturalinclinationsandpassionsandsoon,butnotsothat,indoingthis,theymightaspiretoanatural,prelapsarianstateinwhichtheymightknowthingsastheyarewithanunmediatedknowledge.Thistheywillneverachieve.Rather,thereformofbehaviourisadisciplinetowhichtheymustsubjectthemselvesiftheyaretobeabletofollowaprocedurewhichisinmanyrespectsquitecontrarytotheirnaturalinclinations,whichisatoddswithtraditionalconceptionsofthenaturalphilosopher,andwhichisindeedsubversiveoftheirindividuality.5EliminativeInductionWhatBaconisseekingfromamethodofdiscoveryissomethingthatmodernphiloso-pherswoulddeemimpossiblystrong:thediscoveryofcauseswhicharebothnecessaryandsufficientfortheireffects.WhyplacesuchstrongconstraintsonCausation(pp.303–12),sothatweonlycallsomethingacausewhentheeffectalwaysoccursinthepresenceofthisthingandneverinitsabsence?Inthefinalanalysis,whatBacon–likeAristotle(chapter23)beforehim–isafteraretheultimateexplanationsofthings,638\nBACONanditisnaturaltoassumethatultimateexplanationsareunique.WhatBacon’smethodisdesignedtodoisprovidearoutetosuchexplanations,andthisroutetakesusthroughanumberofproposedcausalaccounts,whicharerefinedateachstage.Theprocedureheelaborates,eliminativeinduction,isoneinwhichvariouspossiblycon-tributoryfactorsareisolatedandexaminedinturn,toseewhethertheydoinfactmakeacontributiontotheeffect.Thosethatdonotarerejectedandtheresultisaconver-genceonthosefactorsthataretrulyrelevant.Thekindof‘relevance’thatBaconisafteris,ineffect,necessaryconditions:theprocedureissupposedtoenableustoweedoutthosefactorsthatarenotnecessaryfortheproductionoftheeffect,sothatweareleftonlywiththosethatarenecessary.TheCaseofColourBaconprovidesanexampleofhowthemethodworksinthecaseofcolour.Wetakeasourstarting-pointsomecombinationofsubstancesthatproduceswhiteness,thatiswestartwithwhatareineffectsufficientconditionsfortheproductionofwhiteness,andthenweremovefromtheseanythingnotnecessaryforthecolour.First,wenotethatifairandwateraremixedtogetherinsmallportions,theresultiswhite,asinsnoworwaves.Herewehavethesufficientconditionsforwhiteness,butnotthenecessarycon-ditions,so(second)weincreasethescope,substitutinganytransparentuncolouredsub-stanceforwater,whencewefindthatglassorcrystal,onbeingground,becomewhite,andalbumen,whichisinitiallyawaterytransparentsubstance,onhavingairbeatenintoit,becomeswhite.Third,wefurtherincreasethescope,andaskwhathappensinthecaseofcolouredsubstances.Amberandsapphirebecomewhiteonbeingground,andwineandbeerbecomewhitewhenbroughttoafroth.Thesubstancesconsidereduptothisstagehaveallbeen‘moregrosslytransparentthanair’.Baconnextconsidersflame,whichislessgrosslytransparentthanair,andarguesthatthemixtureofthefireandairmakestheflamewhiter.Theupshotofthisisthatwaterissufficientforwhiteness,butnotnecessaryforit.Hecontinuesinthesamevein,askingnextwhetherairisnecessaryforwhiteness.Henotesthatamixtureofwaterandoiliswhite,evenwhentheairhasbeenevaporatedfromit,soairisnotnecessaryforwhiteness,butisatransparentsub-stancenecessary?Bacondoesnotcontinuewiththechainofquestionsafterthispoint,butsetsoutsomeconclusions,namelythatbodieswhosepartsareunequalbutinsimpleproportionarewhite,thosewhosepartsareinequalproportionsaretransparent,pro-portionatelyunequalcolours,andabsolutelyunequalblack.Inotherwords,thisistheconclusionthatonemightexpectthemethodofsiftingoutwhatisnecessaryforthephenomenonandwhatisnottotake,althoughBaconhimselfdoesnotprovidetheroutetothisconclusionhere.OnecanaskwhatBacon’sconfidenceinhisconclusionderivesfromifhehasnotbeenabletocompletethe‘induction’himself.Theansweristhatitderivesfromtheconse-quenceshecandrawfromhisaccount.Therearetwowaysinwhichthejustificationfortheconclusionscanbeassessed:bytheprocedureofeliminativeinductionthathehasjustsetout,andbyhowwelltheconsequencesoftheconclusionssogeneratedmatchotherobservations.Inotherwords,thereisatwo-wayprocess,fromempiricalphenomenatofirstprinciples(induction),andthenfromfirstprinciplestoempiricalphenomena.639\nSTEPHENGAUKROGER6TruthCloselytiedupwithBacon’saccountofmethodishistreatmentofthequestionofTruth(chapter36).Bacongoesthroughanumberofwhatheconsiderstobeinade-quatecriteriathathavebeenusedtoestablishtruth.Herejectscriteriadependingonantiquityorauthority,thosederivingfromcommonlyheldviews,andthoserelyingupontheinternalconsistencyorthecapacityforinternalreductionoftheories,pre-sumablyonthegrounds,amongothers,thatsuchcriteriadonotbearonthequestionofwhetherthereisanycorrespondencebetweenthetheoryandreality.Healsorejects‘inductionswithoutinstancescontradictory’,thatis,inductionswhichrestrictthem-selvestoconfirmingatheory,aswellas‘thereportofthesenses’.Noneofthese,hetellsus,are‘absoluteandinfallibleevidenceoftruth,andbringnosecuritysufficientforeffectsandoperations’.Thathetiesinevidenceforthetruthofatheoryanditsuse-fulnesshereisnoaccident,fortheseareintimatelyconnected,tellingusinValeriusTer-minusthat‘thediscoveryofnewworksandactivedirectionsnotknownbefore,istheonlytrialtobeacceptedof’.IsBaconprovidingaglossontruthhere,maintainingthatithasbeenmisconstrued,thattosaysomethingistrueisexactlythesameassayingthatitisuseful?Orishesayingthatsomethingistrue,intheordinaryacceptedsense,onlyifitisuseful?Whichever,itisaverystrongclaimonBacon’spart.Aretherenouselesstruths,andaretherenofalsehoodswhichhavepracticalapplication?Itisnotsimplythatfalsepremissesmayleadtotrueconclusions,buttherearecaseswhereapproximationswhich,whilefalse,mayhavemorepracticalvaluethanthetruthsofwhichtheyaretheapproximation.Thesolutionbecomesclearwhenweconsiderthatsinceantiquitydebatesonmethodsofgeneratingtruthshadhingedonthequestionofgeneratinginformativetruths:theaimistodiscoversomethingwedidnotalreadyknow.Inparticular,therewasaconcernamongAristotleandhisRenaissancefollowerstoshowthatformalmodesofreasoningsuchastheSyllogism(pp.151–3)werenottrivialorcircular,because,atthestartoftheinferentialprocess,wehaveknowledgethatsomethingisthecase,whereasattheendofitwehaveknowledgewhyitisthecase.Inparticularwhattheysoughttoshowwasthatthekindofknowledgeofanobservedphenome-nonwehavethroughsensationisqualitativelydifferentfrom,andinferiorto,thekindofknowledgewehaveofthatphenomenonwhenwegraspitintermsofitscauses.ThisisalsowhatBaconwasseeking.Ifwethinkintermsof‘informativetruths’,Bacon’spositionmakesalittlemoresense.Heissayingthattheonlywayinwhichwecanjudgewhethersomethingisinformativelytrueistodeterminewhetheritisproductive,whetherityieldssomethingtangibleanduseful.Andifsomethingdoesconsistentlyyieldsomethingtangibleanduseful,thenitisinformativelytrue.(The‘consistently’hereisimportantifwearetobeabletoruleoutcaseswherefalsepremissesjusthappenonparticularoccasionstoyieldtrueconclusions,forwecanassumethat,unliketruths,theywillnotcontinuetodothisindefinitely.)Andthecaseofapproximationscanperhapsbedealtwithbysayingthatthesederivetheiruseful-nessnotfromtheirfalsitybutfromtheirproximitytothetruth,althoughthecaseswheretheapproximationismoreusefulthanthetrueaccountcannotbehandledsoeasily.640\nBACONThequestionofthepracticalityoftruthturnsonitsinformativeness,butthereisanotherdimensiontothisquestionwhich,althoughitisnotexplicitlymentionedbyBacon,isofimportanceinunderstandinghisgeneralorientation.InthehumanistthoughtwhichisthesourcefromwhichBaconderivesmuchofhisinspiration,moralphilosophyfiguresverypredominantly.Nowinthisphilosophy,beingvirtuousandactingvirtuouslyarethesamething:thereisnoseparatepracticaldimensiontomoral-ity.Thisisallthemoreinterestingbecausemoralphilosophyisacognitiveenterprise,oneinwhichthepracticaloutcomeisconstitutiveofthediscipline,somethingBaconstressesintheAdvancementofLearning.Ifweseenaturalphilosophyasbeinginsomerespectsmodelledonmoralphilosophy,somethingwhichisnaturalenoughinahumanistcontext,andwhichisreinforcedintheshiftfromotiumtonegotium,thenwemaybeabletomakealittlemoresenseoftheideathattruthisnottruthunlessitisinformativeandproductive.Moreover,wemayalsobegintoapproachthequestionofwhythenaturalphilosopherisnotsimplysomeonewithaparticularexpertise,forBacon,butsomeonewithaparticularkindofstanding,aquasi-moralstand-ing,whichresultsfromthereplacementoftheidea(prevalentthroughoutancientthought)oftheSageasamoralphilosopherwiththeideaoftheSageasanaturalphilosopher.FurtherReadingForageneraloverviewofBacon’sthoughtseeZagorin(1998)andPeltonen(1996).OnhisnaturalphilosophyandmethodseeRossi(1968)andGaukroger(2000).Anderson(1971)providesadetailedsummaryandassessmentofhisnaturalphilosophy.ReeshasdealtwithspecificaspectsofBacon’snaturalphilosophyinanumberofarticles,andhisintroductiontovolume6ofthenewOxfordeditionofBaconprovidesagoodsummaryofhispioneeringworkonsuchtopicsasBacon’smattertheoryandcosmology.Urbach(1987)providesthemostcomprehensivediscussionofBacon’saccountofmethod.Jardine(1974)andPez-Ramos(1988)concentrateontheroleofrhetoricinBacon’swork.Martin(1992)providesagoodaccountofBacon’sthoughtonlegalquestions.Webster(1975)looksattheinfluenceofBaconbetweenthetimeofhisdeathandthefoundingoftheRoyalSociety.ReferencesBaconThestandardeditionis:Bacon,F.1857–74:TheWorksofFrancisBacon,14vols(editedbyJ.Spedding,R.L.EllisandD.DenonHeath).London:Longmans,Green,Reader,andDyer.Thiswilleventuallybereplacedbythefollowing,asthevolumesappear:Bacon,F.1996–:TheOxfordFrancisBacon,12vols(generaleditorsG.ReesandL.Jardine).Oxford:ClarendonPress.OtherwritersAnderson,F.H.1971:ThePhilosophyofFrancisBacon.NewYork:OctoganBooks.Briggs,J.C.1989:FrancisBaconandtheRhetoricofNature.Cambridge,MA:HarvardUniversityPress.Eamon,W.1994:ScienceandtheSecretsofNature:BooksofSecretsinMedievalandEarlyModernCulture.Princeton,NJ:PrincetonUniversityPress.641\nSTEPHENGAUKROGERFarrington,B.1964:ThePhilosophyofFrancisBacon:AnEssayonitsDevelopmentfrom1603to1609withNewTranslationsofFundamentalTexts.Chicago:UniversityofChicagoPress.Fattori,M.1997:IntroduzioneaFrancisBacon.Rome:EdizioneLaterza.Findlen,P.1997:FrancisBaconandtheReformofNaturalHistoryintheSeventeenthCentury.InD.R.Kelley(ed.)HistoryandtheDisciplines:TheReclassificationofKnowledgeinEarlyModernEurope.Rochester,NY:UniversityofRochesterPress.Gaukroger,S.2000:FrancisBaconandtheTransformationofEarlyModernPhilosophy.Cambridge:CambridgeUniversityPress.Jardine,L.1974:FrancisBacon:DiscoveryandtheArtofDiscourse.Cambridge:CambridgeUniversityPress.Jones,R.F.1982:AncientsandModerns:AStudyoftheRiseoftheScientificMovementinSeventeenth-centuryEngland.NewYork:Dover.Kuhn,T.S.1977:MathematicalversusExperimentalTraditionsintheDevelopmentofPhysicalScience.InTheEssentialTension,2ndedn.Chicago:UniversityofChicagoPress.Laudan,L.1981:ScienceandHypothesis:HistoricalEssaysonScientificMethodology.Dordrecht:D.Reidel.Leary,J.E.,Jnr1994:FrancisBaconandthePoliticsofScience.Ames:IowaStateUniversityPress.Malherbe,M.1985:Bacon,l’EncyclopédieetlaRevolution.ÉtudesPhilosophiques,3,387–404.Martin,J.1992:FrancisBacon,theState,andtheReformofNaturalPhilosophy.Cambridge:CambridgeUniversityPress.Milner,B.1997:FrancisBacon:TheTheologicalFoundationsofValeriusTerminus.JournalfortheHistoryofIdeas,58,245–64.Peltonen,M.(ed.)1996:TheCambridgeCompaniontoBacon.Cambridge:CambridgeUniversityPress.Pez-Ramos,A.(1988)FrancisBacon’sIdeaofScienceandtheMaker’sKnowledgeTradition.Oxford:ClarendonPress.Rees,G.1975a:FrancisBacon’sSemi-ParacelsianCosmology.Ambix,22,81–101.——1975b:FrancisBacon’sSemi-ParacelsianCosmologyandtheGreatInstauration.Ambix,22,161–73.——1977:MatterTheory:AUnifyingFactorinBacon’sNaturalPhilosophy?’Ambix,24,110–25.——1979:FrancisBacononVerticityandtheBowelsoftheEarth.Ambix,26,202–11.——1980:Atomismand‘Subtlety’inFrancisBacon’sPhilosophy.AnnalsofScience,37,549–71.——1981:AnUnpublishedManuscriptbyFrancisBacon:SylvasylvarumDraftsandOtherWorkingNotes.AnnalsofScience,38,377–412.——1984a:Bacon’sPhilosophy:SomeNewSourcesWithSpecialReferencetotheAbecedariumnovumnaturae.InM.Fattori(ed.)FrancisBacon:terminologiaefortunanelXVIIsecolo.Rome:Edizionedell’Ateneo.o——1984b:FrancisBaconandspiritusvitalis.InM.FattoriandM.Bianchi(eds)Spiritus:IVColloquioInternazionaledelLessicoIntellettualeEuropeo.Rome:Edizionedell’Ateneo.Rossi,P.1968:FrancisBacon:FromMagictoScience.Chicago:UniversityofChicagoPress.Urbach,P.1987:FrancisBacon’sPhilosophyofScience:AnAccountandReappraisal.LaSalle,IL:OpenCourt.Vickers,B.(ed.)1968:EssentialArticlesfortheStudyofFrancisBacon.Hamden,CT:ArchonBooks.Webster,C.1975:TheGreatInstauration:Science,MedicineandReform,1626–1660.London:Duckworth.Wormald,B.H.G.1993:FrancisBacon:History,PoliticsandScience,1561–1626.Cambridge:CambridgeUniversityPress.Zagorin,P.1998:FrancisBacon.Princeton,NJ:PrincetonUniversityPress.642\nBACONDiscussionQuestions1Isaphilosophythatdoesnothavepracticalconsequencesdeficient?Isthereaplaceforapurelycontemplativeformofphilosophy?2WasBaconbeingrealisticinthinkingthatonecouldtransformacontemplativeandspeculativedisciplineintoanempiricalone?3Cantheunderstandingofnaturebeseparatedfromthetransformationofnature?4WhatdoesBaconmeanwhenhemaintainsthatknowledgeispower?5Couldtherebeageneralmethodofdiscoverywhichwasindifferentastoarea(politics,law,naturalscience)?6Couldtherebeamethodofdiscoveryatall,eveninarestrictedarea?7Aretherepsychologicalobstaclestolearningwhichneedatleasttoberecognized,andatbestpurged,beforewecanattaintoknowledge?8TowhatextentcantheIdolsoftheTribebeovercomebyBacon’smethodofeliminativeinduction?9IsthereanythingthatcanbedoneabouttheIdolsoftheCave?IsBaconbeingundulypessimisticinhisaccountofthem?10IsthewayinwhichBaconcharacterizeslanguageinhisaccountoftheIdolsoftheMarket-placesatisfactory?11IslanguagereallytheobstacletothoughtthatBaconbelievesittobe?12TowhatextentcantheIdolsoftheTheatrebepurgedbyfollowingBacon’smethodology?13CouldsimplyfollowingBacon’smethodofeliminativeinductioneverenableanyonetodiscoveranything?14Isthereanywayofavoidingcircularityinabstractinggeneralprinciplesfromasetoffacts,andthendemonstratingthosefactsfromtheabstractedprinciples?15Istheresomewayofmakingsurethatthemethodsorproceduresoneusesyieldinformativetruths?16Howdoesthefact(ifitisafact)thatapproximations,whicharestrictlyspeakingfalse,aresometimesmoreusefulthanthetruthsofwhichtheyaretheapproximations,bearontheclaimthatusefulnessisasignoftruth?643\n26DescartesandMalebrancheRICHARDFRANCKSANDGEORGEMACDONALDROSSTheseventeenthcenturywasaperiodofprofoundtransformationinphilosophy,scienceandtheology,andintheconceptionofhumanlife.Thescholasticphilosophyoftheuniversitieswaschallenged,andmanythemesstillatthecentreofphilosophicaldebatewereinitiated.Thisandthenextchapteraredevotedtofourmajorfigures,Descartes(1596–1650),Malebranche(1638–1715),Spinoza(1632–1677)andLeibniz(1646–1716).Allwerepowerfulmetaphysicalthinkers,whousedreasontocriticizeboththephilosophyoftheirmedievalpredecessorsandcommon-senseviewsoftheworld.Therivalaccountsofsubstanceofferedbytheseauthorsareexploredinthecontextofattemptstoestablishmodernexplanatoryscience.TheseaccountsincludeDescartes’sdivinelycreatedextendedsubstanceandthinkingsubstances;Malebranche’s‘occasional’causes;Spinoza’ssinglesubstance,GodorNature;andLeibniz’ssoul-likemonads,eachexistinginisolationbutreflectingthewholesystemofrealityaccordingtoadivinelycreatedpre-establishedharmony.WiththeexceptionofMalebranche,whosephilosophyraisesspecialissues,thefocusisontheimplicationsfornature,Godandhumanbeings.Readerswillwishtoconsultseveralofthehistoricalchaptersinthisvolume,butespeciallythosedealingwithotherseventeenth-centuryphilosophers,HOBBES(chapter28)andLOCKE(chapter29).TheywillalsowishtoreadchaptersonEPISTEMOLOGY(chapter1),METAPHYSICS(chapter2),PHILOSOPHYOFMIND(chapter5),PHILOSOPHYOFSCIENCE(chapter9)andPHILOSOPHYOFRELIGION(chapter15).1DescartesRenéDescarteshasbeenreferredtosincetheeighteenthcenturyasthe‘fatherofmodernphilosophy’,butasconceptionsofphilosophyhavealteredduringthattime,sotoohasthemeaningofthetitle.Inthenineteenthcenturyandformostofthetwen-tiethcentury,English-speakingphilosophershaveseenDescartesprimarilyasanEpistemologist(chapter1),whoseworksettheagendaforthecentralphilosophicaltaskofshowingwhywearejustifiedinbelievingwhatwebelieve.AtthesametimeinContinentalphilosophy(chapters41and42)Descartes’sconcentrationontheindividualthinkingselfanditsrelationtoitssurroundingshasledtohisbeing\nDESCARTESANDMALEBRANCHEregardedprimarilyastheoriginatorofthephenomenologicalandexistentialisttradi-tions.NoonewoulddenyDescartes’simportanceinthosetwoareas,butmorerecentworkhasgivenusaratherdifferentDescartes:bylookingmorecloselyatthehistori-calcontextwithinwhichDescarteswasworking,recentphilosophershaveofferedadif-ferentperspectiveonhiswork,andinsodoinghaveallowedhimtokeephistitle,whilechangingitsmeaningbacktosomethingnearertowhatthosewhooriginallygaveittohimmighthavehadinmind.‘Philosophy’intheseventeenthcenturywasnotthenameofaspecificacademicdis-cipline,butmeantmoregenerallysomethinglike‘learning’or‘knowledge’,construedasanyattempttounderstandtheworldbypurelynatural,rationalmeans.Inthatcontext,tosaythatDescarteswasthefatherofmodernphilosophyistosaythathewasthegreatchampionoftheself-proclaimed‘modern’movementoftheseventeenthcentury.Thatmovementwasverydiverse,verydiffuseandverywidespread,andwasunitedbylittlemorethantheconvictionthatthetraditionallearninghadfailed,butthatasecureandpracticallyusefulunderstandingoftheworldcouldbeachievedifonlywewouldrenouncethemistakenmethodsandassumptionsofthepastandrefoundknowledgeonanewbasis.Themovementwasvariouslyreferredtoasthe‘NewPhilosophy’,the‘MechanicalPhilosophy’andthe‘ExperimentalPhilosophy’,andtheperceivedsuccessofitsadherentsintakingoverorreplacingtheintellectualestablish-mentofthedayisacentralpartofthecomplexprocesswhichlaterhistorianscametorefertoasthe‘ScientificRevolution’.Descartesdidnotstartthemodernistmovement,norwashisworkthecauseofitssuccess.Hisreputationasthefatherofthemodernviewderivesfromthefactthatformanyyearsbothbeforeandafterhisdeathhewasconsidereditsleadingtheoristandpublicist.Hiscarefullyworkedoutandbrilliantlypresentedphilosophicalvisionoftheworldandhumanity’splaceinitplayedalargepartinestablishingthepossibilityofaviablealternativetothetraditionalaccount,andinlegitimizingbeliefinit.ThelifeofDescarteswasaclassicstoryofamanwithamessageinwhichhepassionatelybelievesbutwhichheknowswillbeunpopularwithmuchoftheintellectual,religiousandpoliticalestablishmentofhistime.Hewasconstantlyworkingtoavoidpersecution,towininfluentialpeopleovertohisside,andtodisguiseorsuppresshistrueopinionswhereverhethoughtitnecessary,whileatthesametimeinventingaseriesofsuperbliteraryandconceptualdevicestopropagatehismessagewheneveritwassafetodoso.Topresentmodernismasaviablealternativetotraditionallearning,Descartesessen-tiallyhadtodothreethings:todescribeinverygeneraltermsthenatureoftheworldaccordingtohisversionofmodernism,toshowhowthatworldrelatedtoaChristianGod(chapter15),andhowitrelatedtohumanity.Wewillsetoutthissketchofhisworkunderthosethreeheadings.1.1NatureDescartessawthewholenon-humannaturalworldasasingle,deterministic,mechani-calsystem.HeheldthatjustasKepler(1571–1630)hadreducedtheapparentlywan-deringpathsoftheplanetstosimplemathematicallaws,orasGalileo(1564–1642)hadfoundsimilarlysimplemathematicallawsunderlyingallthediversephenomenaoffallingobjectsonearth,soinfactallnaturalphenomena,fromtheshiningofthe645\nRICHARDFRANCKSANDGEORGEMACDONALDROSSSuntothebitingofaflea,could,ifproperlyunderstood,beseentobegovernedbyunchangingLawsofnature(pp.305–6)suchthat,ifonlyweknewenoughaboutthem,wewouldseethemnotasrandomandisolatedeventsbutastheonlypossibleoutcomesofthetimelessfactsofnature.Andthoseunchangingfacts,heheld,wereultimatelythelawsofmechanics.Thelawsofmeteorology,forexample,andofzoologymightappearverydifferent;butunderlyingbothofthemaremorefundamentallawsgoverningthemechanicalinteractionofparticlesofinertmatter.Thelawsofoptics,forexample,heheldtobeaconsequenceofthelawsofmechanics,becausehesawlightasastreamofhighlyenergeticparticleswhicheitherpenetrateorbounceoffsur-facesinthesamewayasarrowswilleitherpenetrateorbounceoffasurfaceofcustardorofsteel.Kepler’slawsofplanetarymotionheheldtobederivablefrommechanicsbecausetheplanetswerepushedalongintheirorbitsbyaswirlor‘vortex’oftinymate-rialparticlesswingingaroundthesun,injustthesamewaythataleafisswirledaroundbyacurrentofair,oracorkbyaneddyinastream.Inbiology,too,mechanismruled:theheartofalivingcreaturebeatsbecauseofaconstantrepetitionofheatingandcooling–themselvesmechanicalphenomenatobeexplainedbytheacceleratinganddeceleratingofparticles–whichthebloodundergoesinmovingthroughthebody;aseedgrowsbecauseofthegradualaccretionofsomeoftheparticleswhicharecon-stantlypushedthroughitbyits(mechanical)digestiveprocesses–andsoon.Thusthewholeofnaturecomestobeseenasagiantclockwork–theseventeenthcentury’sfavouriteanalogy–woundupbyGodatcreation,andtickingalonguntiltheendoftheworld,poweredbyanunendingseriesofpushesandpulls,bumpsandtwists,shakingsandswirlings.AtthispointthemodernreadermaybeinclinedtotakearatherpatronizingviewofDescartes’swork:thispurelymechanicalviewseemschildishlynaive,anobviouslyinadequateresourceforexplainingallthephenomenaofnature.Andindeed,Descartes’sbiologywasaresearchprogrammethatwasneversuccessfullycarriedout;hisanatomywasrejectedinfavourofthatofHarvey(1578–1657);withinagenera-tionhiscosmologyandhisopticswerebeingreplacedbyNewton’s(1642–1727);andevenhismechanicsitselfwasincreasinglyabandoned.Twopoints,though,shouldbemadeinhisdefence.Firstly,Descarteswasapioneer,andhoweverunsuccessfulhisphysicaltheorymayhavebeenindetail,thephilosophicalvisionthatinspireditwasenthusiasticallyadoptedevenbythosewhoabandonedhisowncherisheddiscoveries,andinfactstillformsthebasisofthemodernviewoftheworld.ThesecondpointindefenceofDescartes’smechanismismorecomplex,andgivesabetterviewofDescartes’srelationbothtohispredecessorsandtohissuccessors.Descartes’sworldischaracterizedaboveallbyitsintelligibility.Aswithallthoseofthemodernistparty,hisconstantcomplaintagainstsupportersofthedebasedAristotelianismoftheuniversitieswasthattheylefttheworldmysterious,bypostulat-ingincomprehensiblefaculties,powers,virtuesandgoalsinnature.Insteadofexplain-ingsuchwell-knownandpoliticallyandeconomicallyimportantphenomenaasthatofthecompass,forexample,theymerelyredescribeditsbehaviourintechnicaltermswhichhadnocontentbeyondtheveryphenomenatheywereinventedtoexplain.Descartes’smechanism,bycontrast,succeededinreducingtheunknowntotheknown,byshowinghowinitiallypuzzlingphenomenacanbeexplainedastheoperationofanunperceivedbutperfectlywellunderstoodcause,namelytheimpactofinvisible646\nDESCARTESANDMALEBRANCHEparticles.Significantly,thisemphasisonexplicabilitywaswhatdistinguishedtheCartesianworldpicturefromthoseofmanyofhiscontemporariesandsuccessorswhoproclaimedthemselveshisfellowmodernists.ThedisagreementwithHarvey,forexample,overtheoperationoftheheartderivesfromarejectionbyDescartesofHarvey’sacceptanceofthelivingpoweroftheheartassomethingnotitselfexplained;whereasforHarveytheheartisimbuedwithlife,asbefitsthecentralorganofthebodyaroundwhichthebloodcirculateslikecourtiersaroundaking,forDescartesitismerelyalumpofmatter,alternatelydilatedandcontractedbythemechanicalpressureoftheblooditcontains.Inthesameway,laterCartesianssoughtunsuccessfullytodefendhissystemofvorticesandetheragainsttheNewtoniantheoryforpreciselythesamereason:Newton’sviewcontainedasagiventhefactthatallbodieshavethepowertodrawotherbodiestowardsthemwithnointerveningmediumthroughwhichsuchforceistransmitted.AscomparedtoDescartes’stheory,thisisaclearbetrayalofthemodernistideal,andare-introductionoftheoccultpowersoftheScholastics(chapter24),inthatitclaimstoexplainthefactthatheavyobjectsfalltoearthnotbymeansofaflowofinvisibleparticleswhichpushthemdownward,butbyimbuingthephysicalworldwiththepseudo-propertyof‘gravitas’.VitalpowersandgravitationalattractionwerenottheonlycategoriesofthingsthatDescartes’sbareOntology(chapter2)ofmatterinmotionruledout.Andherewereturntoanaspectofhisworkwhichsubsequentgenerationshaveforthemostpartaccepted.Thebarematerialparticlesofhismechanicalworldareequippedonlywiththemechanicalpropertiesofshape,sizeandmotion,andalltheirotherpropertiesmustbeafunctionofthosefew.Thuswhereasanordinarypersonmightsaythatasword,forexample,wasshiny,sharpanddangerous,Descarteswouldsaythatinitselftheswordisnoneofthosethings:initself,asitreallyis,theswordisanextendedobjectofacertainsizeandshape;allitsotherProperties(p.686)areReducible(pp.312–13)eithertothosebaremechanicalproperties,ortotheinteractionofthosemechanicalpropertieswiththehumansensorysystem.Colourisaclassicexample,whichDescartesandhiscontemporariesinsistedagainandagainmustberemovedfromthelistofanobject’srealproperties.Thecolourofanobjectisnotapropertyitpossesses,asitsshapeoritssizeis,butratherafunctionofthewayinwhichtheshape,sizeandmotionofitsconstituentpartsinteractwiththoseofparticlesthatmakeupraysoflighttooperateonoursensorysystem.‘Silver-coloured’and‘shiny’,therefore,arewordsthatrefernottopropertiesintheobject,butonlytoexperiencesthathumanstypicallyhavewhenconfrontedbyit.Descartes’sWorldTheworldofDescartesisabare,mechanicalplaceoperatinginaccordancewithtime-lesslawsofmotion,and,initself,itisdevoidofmostofthepropertiesthatcouldmakeusfeelathomeinit.Infact,eventhemechanicalpropertiesthatitdoespossessturnouttobeimportantlydifferentfromtheshape,sizeandmotionwithwhichwearefamiliar.TheirimportanceforDescartes,andtheirsuitabilitytoexplaineverythingthathappensinthenaturalworld,liesnotintheirfamiliarity,butinthefactthattheyaremeasurable,andsocanbeexpressedandmanipulatedmathematically.Galileohadalreadywritten647\nRICHARDFRANCKSANDGEORGEMACDONALDROSSthatthebookofnatureiswritteninthelanguageofmathematics,andtheideaiscentraltoDescartes’smorefullyworkedoutmetaphysicalandepistemologicalposition.ForDescartes,whatweseewhentwobilliard-ballscollideisonlyanappearance;therealityoftheeventisappreciatednotbytheeyeofthebody,butbytheeyeofthemind.Thus,theonlyreallyaccurateaccountoftheevent–theonewhichmostfullycorrespondstowhatisreallythereintheworld,purgedofanyreferencetoitseffectsonoursensoryapparatus–isamathematicaldescription,inwhichsize,shape,weight,speedanddirectionareallexpressed(withthehelpofwhatwestillrefertoasCartesianco-ordinates)asnumbers.Theoutcomeofthatcollisionisthenexpressibleasanewsetofsuchnumericalvalues,andthetransformationfromonesettotheotherisgovernedbylawsofmotion,centralamongwhichisthatoftheconservationofthetotalamountofmotioninthesystemasawhole.ItisatestimonytothesuccessofDescartesandhisfellowmodernistsinpropagatingtheirviewsofnaturethatmuchofthatstoryseemssofamiliar,indeedobvious,toustoday.Ourcontemporaryviewisadirectdescendantofthephilosophicalinnovationsoftheseventeenthcentury,tosuchanextentthatitishardforustoseehowanyonecouldeverhaveseenthingsotherwise.Inparticular,wetakeforgrantedDescartes’sfunda-mentalmetaphysicalclaimthattheworldofexperience,theworldoftheordinaryperson,isnottheworldasitreallyis,butonlythewaythatrealworldappearstous.Theworldasweexperienceitisarichandvariedsuccessionofqualitiesandvalues;theworldasitreallyis,isastatelytransformationofnumbersequencesinaccordancewithimmutablelaws.Theviewoftheexpertisnotafullerandmoredetailedversionoftheviewofthecommonperson,butaradicallydifferentview,whichisexpressiblenotinthelanguageofcommonsensebutinthelanguageofmathematics;therealworldissimplyunavailabletothesensoryapparatusoftheordinaryperson,andisdiscoverableonlybythearcaneinvestigationsofthe‘philosopher’.Thiscontrastbetweenthecommon-senseviewandtheviewoftheexpertisoneofthemoststrikingfeaturesofDescartes’swriting.Timeandagainhehastotrytoper-suadehisaudiencethattheycannottrustthewaytheworldappearstobetotheirsenses;theymustworkoutrationallywhatitisreallylike.Itmayappeartothecasualobserverasifthebloodinthearteriesandthatintheveinsisofadifferentkind,butthephiloso-phercanshowthatinrealitytheyarethesame;itmaylooktothesensesasiftheearthstandsstillwhiletheheavenlybodiesmovearoundit,butthephilosophercanshowthatinrealitytheearthmovesaroundthesun.1.2GodDescarteswasacutelyaware,especiallyafterthepubliccondemnationofGalileoin1633,thatmanywouldregardhisphysicalandmetaphysicaltheoriesastantamounttoheresy,buthehimselfsawthemasperfectlycompatiblewith,andindeedinsepa-rablefrom,hisversionofChristianity.Toseetherelation,weneedtolookmorecloselyathisaccountofmaterialsubstance.Descartes,forallhismodernity,wasnotanatomist.Althoughhemaintainedthatthepropertiesofobjectswereafunctionofthewaytheirminutepartswereputtogether,hedidnotthinkthatallmatterwasmadeupintheendofimpenetrableand648\nDESCARTESANDMALEBRANCHEirreducibleatoms.ForDescartes,remember,therealityoftheworldiswhatcanbedescribednumerically,soitfollowsthatjustasthereisnosmallestpossiblenumber,sotherecanbenosmallestpossibleparticleoratom.Matter,therefore,mustbeinfinitelydivisible.Forthesamereason,Descartesdeniedthepossibilityofavacuum.Ifthetruedescriptionoftheworldisitsmathematicaldescription,andmatterjustisitsdimen-sions(‘extension’),thenaspacebetweenobjects,sinceitpossessesshape,size,durationandmotion,iseverybitasrealaphysicalthingasaretheobjectsoneithersideofit.Thereisthereforenosuchthingasgenuinelyemptyspace:justasaroomwhichthecommonpersonwouldsaywasemptyisknownbythephilosophertobefullofair,sothespacebetweentheplanetsandthatabovethecolumnofmercuryinTorricelli’sfamousexperimentisheldbyDescartestobeinrealityfilledwithafine,fluid,‘subtle’matter,throughwhichlightandgravityaremechanicallytransmitted.Therelationbetweenobjectsandthespacesaroundthemmustthereforebeseenasanalogoustothatbetweenawhirlpoolandthewaterthatsurroundsit.Awhirlpoolcanbeseen,andcanbetouched;ifitisfastenough,itcanfeelquitehard,evenimpenetrable.Andyetitismadeupoftheverysamestuffasthewateraroundit:thereisnothinginthewhirlpoolthatisnotinthewater,andthereisnomoreofitinthewhirlpoolthanthereisintherestofthesea.Theonlydifferencebetweenthetwoisrelativemotion:allthewaterthatweconsideraspartofthewhirlpoolismovinginunison,andmovingatahighspeedrelativetotheotherwiseindistinguishablewateraroundit.ThatisDescartes’smodelfortheindividuationofphysicalobjectsintheseamlesscontinuumofextendedmatter:whatwecallanobjectiswhatwemightnowdescribeasahigh-pressureareainthematerialcontinuum.ForDescartes,theordinaryperson’splural-istviewofaworldmadeupofanindefinitenumberofseparatephysicalobjectsisachildisherror,fromwhichthemanofsciencemustescape.Thoseseparatephysical‘objects’areinreality,hesays,onlyadjectivalinnature.Justastheshapeofanappleorthefallofaleafarenotseparatethingsoverandabovetheappleandtheleaftowhichtheybelong,butonly‘modes’or‘modifications’oftheirrespectivesubstances,so,accordingtoDescartes,theappleandtheleafthemselves,likeeveryotherphysicalobjectinexistence,arenotinrealityseparatethingsorsubstances,butonlymodesoftheonesingleextendedsubstance:theyareareasofacontinuumdifferentiatedbyus‘modally’onthebasisofrelativemotion,butarenot‘really’or‘substantially’distinct.TherelationofGodtothatphysicalcontinuumisadoubleone.Ontheonehand,Godcreatedthematerialworldexnihiloandsetitspartsinmotion,andthatoriginalmotionhasbeenconservedthroughallthemechanicalinteractionswhichhavetakenplacefromthatdaytothis.WiththeexceptionofoccasionalMiracles(p.483),there-fore,thewholehistoryofthenaturalworldcanbeexplainedpurelynaturalistically,purelyscientifically,throughthelawsofmechanics.Butthatisonlypartofthestory.Inadditiontocreatingtheuniverseandsettingitinmotion,Godalsosustainsitinbeingfrommomenttomoment.Thereforethereasonwhy,forexample,someparticularbeamoflightisreflectedfromasurfaceatanangleof30degreesisnotonlybecauseithitthesurfaceatanangleof30degrees,butalsobecauseGodwasatthattimemaintainingthelawofreflectionwherebytheangleofincidenceequalstheangleofreflection.Inthesameway,thereasonwhyastonelyinginafieldwillcontinuetolieinthatfieldisnotonlybecausenoonecomesalongtomove649\nRICHARDFRANCKSANDGEORGEMACDONALDROSSit,butalsobecauseGodkeepsinoperationduringthisperiodthelawswhichdetermineitsnatureanditsstructure.InthiswayGod’s‘concourse’or‘concurrence’isinvolvedateverymomentofhistory;withoutthecontinualinvolvementofGod,theclockworkmechanismoftheuniversewouldnottickalongregardless,butwouldfallapartintochaosandnothingness.Inthissense,therefore,themodalrelationwhichwehaveseentoexistbetweenthepropertiesofanobjectandtheobjectitself,andalsobetweenanygivenobjectandextendedsubstanceingeneral,ismirroredintherelationbetweenGodandthatextendedsubstanceitself.Justastotalkabouttheshapeofanobjectisnottotalkaboutaseparatething,buttotalkaboutoneaspectoftheobjectitself,andtotalkofanindi-vidualobjectisinrealitytotalkaboutoneparticularaspectofmaterialsubstance,sototalkaboutmatteratallis,whencorrectlyunderstood,onlyawayoftalkingaboutoneaspectofGod.Strictlyspeaking,Godistheonlytruesubstance,theonlyrealthing;thematerialworldisonlyanexpression,oraconsequence,ofGod’snature.ThesetwostrandsinDescartes’saccountofGodandtheworld–thedivineclock-makerandtheonetruesubstance–obviouslytendtopullindifferentdirections,andtheyweredevelopedindifferentdirectionsbyhissuccessors.TheircombinationinhisworkmeansthatDescarteshadaverystrongdefenceagainstthosewhotendedtoseethenewScience(chapter9)ofnatureastantamounttoatheism:Descartescouldclaimthattheinvestigationofnature,farfrombeingincompatiblewithChristianduty,wasactuallyco-extensivewiththediscoveryofGod’struenatureandintentions.1.3HumanityDescartes’smechanistviewofnaturecoverstheentirephysicalworld,includingthecreationoftheEarthitselfasaconsequenceofthebreak-upofalargervortexofmatter.Itextendsalsototheentireplantworld,andtheentirenon-humananimalkingdom.Thelifeofananimalisnotamysteriousfacultyoravitalprincipleinfusedintomatter,butsimplyaseriesofmechanicalinteractions,exactlyliketheworkinglifeofamachine:ananimalhasmechanicallyproduceddrivesandappetites,anditmakesitswayaroundtheworldseekingbymechanicalmeanstoachievethoseendswiththehelpofitsmechanicalsensorysystem–itisa‘beastmachine’,averycomplexandsophisticatednaturalrobot.Thehumanbodyisanothermechanismofexactlythesamekind,withjustthesamekindofmechanicallifewhichisresponsibleforallofitsnaturalprocesses,andmostofitsdesiresandaversions,itsactionsandreactions.Buttherewereachthelimitsofmechanicalexplanation:thereismoretoahumanbeingthanthosemechanicalpropertiesalonecanexplain.Unlikeanyothercreatureonearth,wecanconsiderthepast,thefutureandthedistantpresent,farbeyondthescopeofourphysicalsenses;wecansometimesstandbackfromourdesiresandchoosewhichonestosatisfyandwhichtodeny;andwecanhaveknowledgeofthingswhicharenopartofthephysicalworldrevealedbythephysicalsenses,notablythingssuchasthetruthsofmathematics,themindsofotherpeople,andGod.Descartes’sexplanationfortheseremarkablehumancapacitiesisthatthereismoretoahumanbeingthantheareaofthephysicalcontinuumwhichishisorherbody;inaddition,humanbeingsalsopossessasoul,animmaterialsubstancewhichcanunderstandandchoose,andwhichisafiniteimitationoftheinfiniteimmaterialsubstancewhichisGod.650\nDESCARTESANDMALEBRANCHEThis,then,isthefamous‘CartesianDualism’,thesuggestionthatahumanbeingisanareaofthephysicalcontinuumtogetherwithanindividualthinkingmindorsoul.Thestatusofeachsuchindividualthinkingsubstanceisthesameasthatoftheonematerialsubstance:itisdependentonGodbothforitsinitialcreationandforitscontinuedbeing.Descartesholdsthatthistheoryisobviousfromthefactsofnaturethatwehavesur-veyed,butinadditionhebroughtforwardadramaticproofofitstruth.InperhapsthemostfamoussinglemomentinthewholeofWesternphilosophy,heclaimedthattheexistenceoftheindividualthinkingsubstanceisindubitableforanythinker.Itmakesperfectsense,heclaimed,formetopretendthatIhavenophysicalbody,andindeedthattheentirephysicalworldthatmysensesrevealtomeisnothingbutadream,oranillusioncreatedbysomeevilpower.ButnomatterwhatmistakesImaymake,andnomatterhowdeceivedImaybe,Imuststillexistinordertobemistaken.ThemerefactthatIamabletomakejudgementsatall,evenmistakenones,provesthatIexist:cogitoergosum.AndthefactthatIcanthinkawaytheentirephysicalworldbutcannotthinkawaymyownexistencesurelyprovesthatImyself,thethingthatisdoingthisthinking,cannotbemerelyapartofthatphysicalworld.TherelationsbetweenthatMindandbody(chapter5)havealwaysbeenperceivedasproblematic.Descartesproposedasystemwherebythehydraulicfluidswhichoperatethemusclesofthebodyareallroutedthroughthepinealglandinthecentreofthebrain,sothatbymakingtinyadjustmentsinthepositionofthatgland,thesoulisabletocontroltheflowofthosefluidsandinthatwaytoredirecttheactionsofthebodyinaccordancewithitsdecisions.Howeverquaintthemechanicsofthatstorymightseem,theviewofhumannaturewhichitembodiesisonethatisverynaturalandverypowerful.Thesoul,becauseitisimmaterial,isnotsubjecttotheeffectsofphysicaldecay,andsoisimmortal.ItisGod-likeinitsimmateriality,andthusisnotpartofthecrude,basematerialworld,althoughitistiedtoitbyitsincarnation.Theactionsofthesoularehumanversionsofthedivineactivitiesofunderstandingandwilling,theformeraperfectbutfinitecopyofGod’sdivineintelligence,thelatterpronetomisuseandcorruptedbytheFall.Inmakingjudgements,theunderstandingpresentsanideatothemind,whichthewillthenaffirmsordenies.InthiswayDescartesisabletoshowthatintellectualmistakesandmoralandreli-giousdeviancearealike,inthatbothresultfromthemisuseofthefreewill.Descartesisthusabletorepresenthisprojectforarefoundingofhumanknowledgeasnotonlyascientificinvestigationwithimportantpracticalbenefits,butalsoasareligiousandmoralreformationofhumansocietyandofthehumanspirit:bycomingtoseetheworldasitreallyis,throughtheeyeofreason,weareliberatingourtrue,activeimma-terialthinkingselvesfromtheirsubjectiontothedeceitfulphysicalsenses,andweareallowingthepure,immaterial,divinesoultotakeprecedenceoverthecorruptphysicalbodyandachieveaviewoftheworldwhichisclosertothatofGod.This,then,isthemetaphysical,psychologicalandtheologicalunderpinningofDescartes’scelebratedworkinepistemology.Histaskherewastoshowthatdespitethefailureoftheestablishedauthoritiesasembodiedintheuniversities,anddespitethefashionablepessimismofphilosopherswhowerescepticalastothepossibilityofanygenuinehumanknowledge,itisindeedpossibleforustoescapefromtheworldofappearancesandestablishastableandusefulbodyofknowledgeastohowtheworld651\nRICHARDFRANCKSANDGEORGEMACDONALDROSSreallyis,ifonlywecanabandontraditionallearningandadoptthenewmethodsofthemodernistparty.Again,thedeviceofthecogitotakescentrestage:becauseitbeginsfromDescartes’s‘hyperbolical’doubtabouttheultimatedeception,andshowsthateveninthoseextremecircumstancesgenuineknowledgewouldbeattainable,itmeansthatnomatterhowextremethescepticaldoubtsmaybe,thepossibilityofcertaintycannotcoherentlybedenied,providedonlythatweescapefromtheillusoryappearancesofcommonsensetotheclearunderstandingoftherationalsoul.Ourcommon-senseviewsarepartofouranimalnature,usefulforfindingourwayaroundintheworld,butdeeplymisleadingastoitstruenature;thecogitoprovesthatbyreasonandreflec-tionwecanovercomethatdeceptionandworkoutrationallywhatisreallygoingon.Whatwemustdoistolookbehindtheapparentcontingencyofeventsintheworldtofindtheunderlyingstructureswhichdeterminethem,andouraimistoworktowardsaunifiedscienceofnaturewhichcouldbelaidoutlikeEuclid’sElementsasasingleaxiomaticsystemofnature,inwhichallnaturalphenomenaarerepresentedastheinevitableoutcomeofunchanginguniversallaws.2MalebrancheCartesianismbecamethedominantphilosophyinEuropeduringthesecondhalfoftheseventeenthcentury,andwellintotheeighteenth.NotallCartesiansfol-lowedDescartesslavishly,andtherewereanumberofindependent-mindedthinkerswhomodifiedhisphilosophyinvariousways,inordertoovercomeitstensionsandambiguities.ByfarthemostinfluentialofthesewasFatherNicholasMalebranche.Indeed,hewassoinfluentialthatallthemajorfiguresofthelateseventeenthandearlyeighteenthcenturies(Locke,Leibniz,Berkeley,HumeandReid)founditnecessarytoattackhimindetail,despite,orperhapsbecauseof,theextenttowhichtheyhadcomeunderhisinfluence.Itisaremarkablefactaboutthehistoryofphi-losophy,that,outsideFrance,duerecognitionhasbeengiventohisimportanceonlyrecently.TherearethreemainareasinwhichMalebranchedivergedsignificantlyfromDescartes;namelyperception,causation,andourknowledgeoftheselfandofGod.Weshallexaminetheseinturn.2.1PerceptionWehaveseenhowDescartesmadeasharpdistinctionbetweentheworldasrepresentedtousbyoursenses,andtheverydifferent,realworld,whichisknownonlytophiloso-phersbytheuseofreason.InordertounderstandMalebranche’scriticisms,weshallneedtolookmorecloselyatDescartes’stheoryofperceptionanditsbackground.ThedominanttheoryinDescartes’sdaywasthescholastictheorythatmaterialobjectsareacompoundofmatterandform.Themattercontainsalltheparticularqual-itiesofanobjectwhichmakeittheindividualthingitis;andtheformcontainsallthegeneralattributeswhichmakeitanindividualofacertainkind.Forexample,ifa652\nDESCARTESANDMALEBRANCHEcarpenterconstructsasetofchairs,eachchairwilldifferfromeveryotherchairinthegrainandcolourofthewood,minordifferencesinshapeandsize,andsoon;butallthechairsshareoneandthesameformofachair,byvirtueofwhichtheyareallequallychairs,andnottablesoranyotherkindofthing.Whenweperceiveoneofthesechairs,whathappensisthatitemitsetherealsurfaces,called‘species’,whicharecap-turedbyoneormoresenseorgans(itscolourbysight,itshardnessbytouch,etc.).Thesearetransmittedtothe‘commonsense’,whereallthesensoryinformationisbroughttogethertomakeupthesinglecompositeimageofthechairofwhichwearedirectlyaware.Sofarthereisnodifferenceinkindbetweenwhathappensinthebrainofananimal,andwhathappensinthebrainofahuman.Inbothcases,thereareimageswhichcloselyresemblethesurfacesofphysicalobjectsastheyareinthemselves.Thebigdifferenceisthat,inadditiontohavingasensoryimageoftheparticularobject,humanscan,throughtheuseoftheirreason,abstractitsuniversalform.Soalthoughmycat’ssensoryimageofthechairandmyimageofitmaybemuchthesame(allow-ingfordifferencesinoursenseorgans),onlyIcanrecognizeitashavingtheuniversalformofachair.BothDescartesandMalebrancherejectedthescholastictheoryofspecies,andmaintainedinsteadthatsensationsbecomepresenttothemindastheresultofmotionsemanatingfrombodies(lightwavesandsoundwaves,forexample),whichgiverisetomotionsinthesenseorgansandthenceinthebrain.Thesensorycharacteristicsofwhichweareconscious,suchascoloursandsounds,arewhollyunlikethepropertiesoftheobjectswhichcausethem.Inthemselves,objectsarenothingotherthanparcelsofextension,movinginvariousways.Althoughthegeometricalpropertiesofobjectsgiveninsensationaresimilarinkindtothepropertiesofobjectsthemselves,eventhesecanbeverydifferent.TouseDescartes’sexample,theSunappearstooursensesasaflatdiscaboutafootacross,andjustbeyondthehorizon;whereasourscientificideaisofabodyfarlargerthentheearth,andmillionsofmilesaway.Again,thesensoryimageofacoinisusuallyelliptical,eventhoughourtrueideaofthecoinisofsomethingcircular.WhereMalebranchepartscompanywithDescartesisoverthequestionofhowwehaveintellectualknowledgeofthetruenaturesofthings.Descartesmadeacom-pleteseparationbetweensensoryimagesandintellectualideas,sothatthesensoryimageoftheSunisatotallydifferententityfromtheintellectualideaofit,andonlythelattergivesustrueknowledge.Butwhatgroundsdowehaveforsupposingthatourintellectualideascorrespondtorealityanymorethanoursensoryimages?Descartes’sanswerwastosaythatGodequippedoursouls,atconception,witharangeofinnateideas,whicharefaithfulcopiesoftheideasheusedwhencreatingtheuniverse.SinceGodisnotadeceiver,theseideasmustbecorrect,andanyuniversalprinciplewhichwe‘clearlyanddistinctly’perceivetobetruemustactuallybetrueoftheworld.MalebranchecriticizedDescartes’stheoryonanumberofgrounds.Inparticular:(1)itisliabletoleadtoacompleterelativism,sinceGod’snotbeingadeceiverisinsufficientguaranteethatwhateachindividual‘clearlyanddistinctlyperceives’correspondstoasingle,objectivereality;(2)thereareinfinitelymanyintellectualideaswhichwemightneedatsometimeorother,anditisimpossibleforafinitemindto653\nRICHARDFRANCKSANDGEORGEMACDONALDROSScontainthemall;and(3)someoftheseideasarethemselvesinfinite(e.g.theideaofinfiniteextension),andagaintheycannotbecontainedwithinafinitemind.Hissolutionwastosaythattheseideasarenotinourminds,butinGod’smind–inotherwords,wearegrantedtheprivilegeofperceivingasubsetofGod’sideas,withouttheirbeingtransferredintoourownminds.Thismayseemastrangenotion;butMalebranchewasoperatinginaChristiancontext,accordingtowhich‘welive,andmove,andhaveourbeing’inGod(Acts17.28);or,asheputit,‘Godisthelocusofspirits’,orthemetaphoricalplacewheretheyexist.Sothevalidityofourclearanddis-tinctperceptionsofintellectualideasisguaranteedbythefactthatweareindirectcontactwiththemindofGod.Asforourperceptionsofexternalobjects,Malebrancherevertedtothescholasticviewthatourperceptionscontainamixtureofsensoryimageryandintellectualform,withtheCartesianprovisothatimagesdonottellushowthingsareinthemselves,butonlyhowourbodiesareaffectedbythem.Ourknowledgeofhowthingsareinthem-selvesdependsontheideaswhichweseeinGod–hencehissloganthat‘weseeallthingsinGod’.Nevertheless,thecreateduniverseisseparatefromGod,anditisonlyitsgeneralfeatureswhichweseeinGod.Theparticulardifferencesbetweenonephysicalobjectandanotheraregiventousonlyinsensation,whichisprovidedbyourphysicalbodies.2.2CausationAswesawearlier,Descarteshadlittletosayaboutthenatureofcausation,sincehetookitforgrantedthatGodistheonlycause.Asfarasthematerialworldisconcerned,heheldthatGodconstantlyre-createsitfromonemomenttothenextinaninfinitesi-mallydifferentform,sothatthereisanillusionofcontinuousdevelopment;butinfactitsstateatoneinstantisnomorethecauseofitsstateatthenextinstantthanoneframeofamovieisthecauseofthenext.Hisaccountofcausationintherealmofthoughtismoreproblematic,sinceheascribedtohumanscompletefreedomofwill,despitetheirtotalsubjectiontoGod.Similarly,hehadgreatdifficultyaccountingforhowmattercouldinfluencemind,orviceversa,sincebothwereofutterlydisparatenatures,anditwasimpossibletoconceiveofanyrelationshipbetweenthetwo.Hisulti-mateanswertobothquestionswasthatweseemtobefacedwithacontradiction,sinceinthislifeweareboundbythelawsoflogic;butGodisnotsobound,andallwillberevealedintheafterlife.Malebrancheanalysedtheconceptofcausationmoredeeply.Matterisessentiallypassive,sinceitmerelytransfersforceswhichareimpressedonit.Ifitweretobeagenuinecause,itwouldneedtohaveitsownimmaterial,activepower,ofwhichwearenotdirectlyawareinperception.Causalpoweristotallybeyondourcomprehen-sion,anditcanonlybeattributedtoGod.Totreatphysicalobjectsashavingcausalpowersistoconsiderthemasdeities,whichisabsurdandheretical.Asbefore,Malebranchediscussestheissueintheologicalterms;yetthereremainstheessentialphilosophicalpointthatoursensoryexperiencegivesusonlyaregularsuccessionofevents,andtheideaofanecessaryconnectionbetweencauseandeffecthastobesuppliedfromsomealternativesource.Subsequentphilosophers,suchasHumeandKant,didnotagreethatthealternativesourcewasGod;buttheywereheavily654\nDESCARTESANDMALEBRANCHEinfluencedbyMalebranche’sinsightthatnecessarycausalconnectionsarenotgiveninexperience.GiventhatGodistheonlycause,thereasonwhyoneeventfollowsanotherliesinGod’smind,whichisbeyondourcomprehension.AllwecansayisthatGodwillsthesecondeventontheoccasionofthefirstevent,notthatthefirsteventcausesthesecond.ThisisMalebranche’sfamousdoctrineof‘occasionalcauses’.Therelationbetweenthemindandthebodyposesaspecialproblem(ofwhichDescarteswasacutelyaware),since,unlikematter,themindisessentiallyactive.HereMalebrancheusesadifferentargumenttoprovethattherecanbenocausalinterac-tionbetweenthetwo.Thisisthat,sincetheybelongtocompletelydifferentcategoriesofbeing,wecannotcomprehendanyrelationshipbetweenthem,letaloneacausalone.Forexample,ifIwilltomovemyarm,themovementofmyarmimmediatelyfollowsthewilltomoveit.ButIhaven’ttheslightestideahow.EvenifIhadacompleteknowl-edgeofphysiology,Istillwouldn’tknowwhattodo,asapurethinkingsubstance,inordertosetinmotionthechainofeventswhichwouldresultintheraisingofmyarm.Consequently,theraisingofmyarmmustbecauseddirectlybyGodontheoccasionofmywillingit.ThisisnottosaythatGodisinanywayinfluencedbymyactofvolition,butratherthatonlyGodcanunderstandthereasonwhytheraisingofmyarmshouldfollowmyactofwilling.2.3TheselfandGodForDescartes,theonlythingwecannotpossiblydoubtistheexistenceoftheselfasathinkingthing.Hebelievedthatbybracketingeverythingwhichcouldconceivablybedoubted,wearriveatadirectintuitionoftheessentialnatureofthethinkingself.ItisonlyafterthishasbeenestablishedthatwecanreinstateabeliefinGod,bydeducinghisexistencefromtheideaofGodwefindinourminds.MalebrancheheldthatDescarteshadgotthingsthewrongwayround.Whenweintrospect,allwearedirectlyawareofareourinnersensationsandthoughts.Wedonothaveanyimmediateintuitionofouressentialnature.Theideaofourselvesasthinkingthingsisanabstractidea,andthereforeonewhichweperceiveinGod.SotheexistenceofGodisaprerequisiteofourthinkingourownexistence,notaconsequencetobededucedfromit.Conversely,DescartesheldthatwehaveanideaofGod,which,althoughitrevealsGodtous,neverthelesscomesbetweenusandGod.ForMalebranche,thereisnoideaofGod,andthereisnoneedofone,sinceGodisthelocusofallspirits.Wehaveanimmediateanddirectvisionofhisinfiniteessenceandexistence,evenif,asfinitebeingswecannotfullycomprehendit.Consequently,weknowGodbetterandwithgreatercertaintythanweknowourselves.Inconclusion,forallitsCartesianism,Malebranche’sphilosophycanbeseenasareversaloftheseventeenth-centurymovetowardsagreaterseparationbetweenphi-losophyandtheology,andareturntothealmostmysticalworldviewoftheChristianPlatonist,StAugustine.Nevertheless,themeansbywhichhearguedforhispositionwerephilosophicallyrigorous,andhehadaprofoundinfluenceonthesubsequentdevelopmentofseventeenth-andeighteenth-centuryphilosophy,whetherornothisoverallphilosophywasaccepted.655\nRICHARDFRANCKSANDGEORGEMACDONALDROSSFurtherReadingForDescarteswefirstsuggestamajorsourceoftheirwritingstranslatedintoEnglish:Descartes(1985),andthenamoreconvenientselectionofhis:Descartes(1988).Readersoftenconcen-trateontheDiscourseonMethod(published1637)andMeditationsonFirstPhilosophy(published1641),butwewouldsuggestalsoreadingpartsofPrinciplesofPhilosophy(published1644),RepliestoObjectionstotheMeditations(published1641–2)andsomeofDescartes’scorrespond-ence.WedonotdiscussPassionsoftheSoul(published1649),butthatisworthreadingaswell.Garber(1992)showstherelationshipbetweenDescartes’sphilosophicalandscientificaims.ThearticlesinCottingham(1992)provideacomprehensiveandup-to-dateexplorationofcurrentphilosophicalresponsestoDescartes.Hooker(1978)andMoyal(1991)offerextremelyusefulcollectionsofarticlesaboutDescartes.Malebranche’smostimportantworkisTheSearchafterTruth,whichwasfirstpublishedin1674–5,andwentthroughanumberofrevisededitionsduringhislifetime.ThestandardEnglishtranslationisMalebranche(1997).RecentcommentariesinEnglishincludeRome(1963)andNadler(1992).TherearetwobookswhichdiscussMalebranche’sinfluenceaswellashisphilosophyitself:McCracken(1983)andBrown(1991).OtherbookscontainchaptersonMalebrancheaswellasotherseventeenth-centuryphilosophers,forexampleYolton(1983)andJolley(1990).ReferencesDescartesDescartes,R.1985:ThePhilosophicalWritingsofDescartes,2vols(editedandtranslatedbyJ.Cottingham,J.R.StoothoffandD.Murdoch).Cambridge:CambridgeUniversityPress.——1988:Descartes:SelectedPhilosophicalWritings(editedandtranslatedbyJ.Cottingham,R.StoothoffandD.Murdoch).Cambridge:CambridgeUniversityPress.WritersonDescartesCottingham,J.(ed.)1992:TheCambridgeCompaniontoDescartes.Cambridge:CambridgeUniversityPress.Garber,D.1992:Descartes’sMetaphysicalPhysics.Chicago:UniversityofChicagoPress.Hooker,M.(ed.)1978:Descartes:CriticalandInterpretiveEssays.Baltimore,MD:JohnsHopkinsUniversityPress.Moyal,G.J.D.(ed.)1991:RenéDescartes:CriticalAssessments.London:Routledge.MalebrancheMalebranche,N.1997:TheSearchafterTruth;ElucidationsoftheSearchafterTruth(translatedandeditedbyT.LennonandP.Olscamp).Cambridge:CambridgeUniversityPress.WritersonMalebrancheBrown,S.(ed.)1991:NicolasMalebranche:HisPhilosophicalCriticsandSuccessors.Assen:VanGorcum.Jolley,N.1990:TheLightoftheSoul:TheoriesofIdeasinLeibniz,Malebranche,andDescartes.Oxford:ClarendonPress.McCracken,C.1983:MalebrancheandBritishPhilosophy.Oxford:ClarendonPress.Nadler,S.1992:MalebrancheandIdeas.Oxford:OxfordUniversityPress.656\nDESCARTESANDMALEBRANCHERome,B.1963:ThePhilosophyofMalebranche.Chicago:HenryRegnery.Yolton,J.1983:PerceptualAcquaintancefromDescartestoReid.Oxford:Blackwell.DiscussionQuestions1Whatistheimportanceofmethodinphilosophy?2Isthenon-humanworldasingle,deterministic,mechanicalsystem?3Whatwouldmakesomelawsofnaturemorefundamentalthanothers?4WasDescartesrighttodemandthatscientifictheoriesbeintelligible?5Howcanwedeterminewhenanaccountexplainsandwhenitmerelyredescribesaphenomenon?6Mustpropertiesbemeasurableifwearetousetheminscientificexplanation?7Isthebookofnaturewritteninthelanguageofmathematics?Whataretheimplicationsofyouranswer?8Ifwhatweseeisonlyanappearance,howcanwegainknowledgeofwhatreallyhappensintheworld?9ShouldDescarteshavebeenanatomist?10Cantherebegenuinelyemptyspace?Howareobjectsrelatedtothespacesaroundthem?11IfGod’sconcurrenceisneededateverymomenttosustaintheuniverse,howcanthelawsofnaturebelaws?12Ifanimalsareexplicableinpurelymechanicalterms,canhumansbeexplainedentirelyinthiswayaswell?13DoesthefactthatIcanmakejudgements,evenmistakenones,provethatIexist?14IfIcanthinkawaytheentirephysicalworld,butcannotthinkawaymyownexistence,doesitfollowthatIamathinkingthing,andnotmerelypartofthephysicalworld?15ShouldwefollowDescartes’smethodofdoubt?16WhydidMalebranchethinkthatourintellectualideasareasubsetofGod’sideas?17WhydoesMalebrancheusetheslogan‘weseeallthingsinGod’?18ExplainwhyMalebranchedeniedthatphysicalobjectshavecausalpowers.19WhatisMalebranche’sdoctrineof‘occasionalcauses’?Whatproblemisitmeanttosolve?20ExplainMalebranche’scriticismofDescartes’sviewsoftheexistenceofGod.21InwhatwaysdoesMalebranche’sphilosophyshowagreaterintegrationofphilosophyandtheologythanDescartesandotherseventeenth-centurythinkers?657\n27SpinozaandLeibnizRICHARDFRANCKSANDGEORGEMACDONALDROSSThischaptercontinuesthediscussionbeguninchapter26oftheseventeenth-centurytransformationinphilosophy,scienceandtheology,andintheconceptionofhumanlife.ItexploresthereasoningthatledtothegreatmetaphysicalsystemsofSpinozaandLeibniz,focusingontheimplicationsofSpinoza’ssinglesubstanceandLeibniz’smonadsforphilosophy,scienceandtheology.1SpinozaIfDescartes(chapter26)wastheleadinglightofthemodernistmovement,BaruchSpinoza(BenedictdeSpinoza)wasitsmostperfectexpression.BydevelopingDescartes’svisionofanew,scientificunderstandingofthingstoitsultimateconclu-sion,heproducedanaccountofGod,theuniverseandhumanity’splaceinitwhichwasofunparalleledbreadth,consistencyandbeauty.Asaresulthewasexcommuni-catedfromhisJewishcommunity,hisbookswerepubliclyburned,andhisnamebecameabywordforwickednessandatheismuntilahundredyearsafterhisdeath,whenhewasrediscovered–andmisinterpreted–bytheRomantics.IncontrasttoDescartes,whowasconstantlytryingoutnewwaysofre-presentingandre-expressingdifferentaspectsofhismessage,Spinozaworkedforyearstopolishandperfectasingle,comprehensiveanddefinitiveaccountofhowthingsare.Hisworkisthereforeverybrief,verydenseanddeeplyobscure.OursurveyofSpinoza’sthoughtwillusethesamethreeheadingsofNature,GodandHumanity;butintalkingofSpinoza,wemustbeginwithGod.1.1GodThesimplestsummaryofSpinoza’smetaphysicsisthathetakesDescartes’soutlineofamechanisticscienceofnon-humannatureanddevelopsitintoaunifiedtheoryofthewholeofbeing,whichhecalls‘GodorNature’.ForDescartes,aswesaw,Godwastheonetruesubstance,whichunderliesandsupportsallcreatedsubstances.ButDescartes’sGodwasalso,andsimultaneously,apre-existing,independentimmaterialbeingwhofreelychosetobringmatterandmindsintoexistenceforGod’sownpurposes.Spinozaseesthosetwoelementsasinconsistent,\nSPINOZAANDLEIBNIZandwithcharacteristicruthlessnesshesimplyeliminatesthesecond:GodorNatureistheonesubstancewhichexists,thesingleunchangingrealitywhichunderliesallphenomenaofanykind,andwhichaperfectsciencewouldreveal.Thispantheisticdoctrinewas,andremains,startlingtomanypeople,andpuzzlingtoevenmore.God,forSpinoza,isthecauseoftheworldandallitcontains–butnotinthewaythatthetermisusuallyunderstood.Godisnotaseparateentity,whosecausationoftheworldisacreativeactwhichgivesititsbeginning,inawayanalogoustothatinwhichapottercreatesapot.Rather,Godistheworlditself,asitreallyis;Godstandsastheindwellingorimmanent‘cause’–orexplanation–oftheworld,inthesenseinwhichwemightsaythatthefundamentallawsofphysics‘cause’–theyunderlie,giveriseto,orexplain–allphysicalobjectsandevents.Godisthereforeiden-tifiedwithNature,notinthesensethatGodissimplytheaggregateofeverythingthatexists,butinsofarasGodisthefundamentalfactsofnaturewhichexplainalltherest,theunchangingrealitybehindallappearanceswhatsoever.GodorNatureisthereforetheatemporalrealitywhichscienceseekstodiscover,andofwhichtheworldasweknowitisatemporalexpression.Theideastrikesusasanoddmixtureofreligiousvisionandscientificresearchprogramme,anddifferentreaderswillproducedifferentSpinozasastheyemphasizeoneorotherofthoseseeminglyirreconcilableaspects.Spinozahimself,though,isinsis-tentthatthetwoareunited,andthatitisonlytheweak-mindednessandanthropo-morphismofDescartesandtheCartesians(andthereforeofusastheirdescendants)whichpreventsthem(andus)fromseeingit.JustashetriestoreinterprettheideaofGod’scausalitybyremovingthemisleadinganalogywithhumancreativity,andinsodoingclaimsthathisGod,andonlyhisGod,istrulythecauseoftheworld,soheinsiststhatalmostallthetraditionalattributesofGodcanintruthbelongtohisGod,andtohisGodalone.GodorNatureisinfinite(limitless)andeternal(atemporal);itisself-caused(self-explanatory),andnecessarilyexists(cannotwithoutcontradictionbesaidnottoexist).Itisomnipotent(theexplanationofeverythingthatispossible)andomnipresent;itisalltruthandallknowledge,andasweshallseelater,intheknowl-edgeofGodorNatureliestrueblessednessandouronlyhopeofimmortality,andtoloveGodorNatureistheonlyroutetoperfectfreedom.ItwasGodorNaturewhochosetheJewsandinspiredtheprophets,becausetheprophetsweremenofgreatunder-standing,whounderstoodthetruthsthatSpinozaisnowrevealing,butwhoexpressedthatunderstandingforanignorantpublicinpicturesqueandanthropomorphiclan-guage.Infact,theonlycharacteristicoftraditionalgodswhichSpinoza’sGodlacksispersonality:GodorNatureisnotaperson,andisnotanythinglikeone.Itdoesnothavedesiresorfeelingsorintentions,anditcannotloveorbeangry,orbeinterestedinthefateofhumanbeings.TothosewhomaintainthatanyconceptionofGodmusthavethesefeatures,Spinoza’sreplyisuncompromisingandclear:wenaturallytendtoseetheworldinourownhumanterms,andnodoubtinjustthesamewayiftrianglescouldspeak,theywouldinsistthatGodmustbeeminentlytriangular.1.2NatureGod,then,isthefundamentalfactsoftheworld,fromwhichallthephenomenaofnaturenecessarilyflow.Inthematerialworld,thatbecomesastraightforwardly659\nRICHARDFRANCKSANDGEORGEMACDONALDROSSCartesianstoryinwhichallphysicalphenomenaarestrictlydeterminedbyunchang-inguniversallaws,andallphysicallawsdependultimatelyonlawsofmotion.LikeDescartes,Spinozadeniedthepossibilityofavacuum,andofatoms,andlikeDescartesheheldthatwhatcommonsensetakestobeindividualphysicalobjectsareinrealityonlyconventionallyindividuatedmodesofanextendedcontinuum,whichSpinozacalledtheAttributeofExtension.Allofthatis,andwas,familiar.Inhisaccountofmentalphenomena,though,Spinozawasmuchmoreradical.WhereasforDescartesthementalworldconsistsofanindefinitenumberofindividualimmaterialthinkingsubstances,someofwhicharetemporarilylinkedbyGodtospecificareasofthematerialcontinuum,Spinozapositsasecond,non-material,mentalcontinuum–theAttributeofThought–ofwhichallmentalphenomenaaremodes,justasallphysicalphenomenaaremodesoftheAttributeofExtension.Thetwoattributesareco-extensive,andexactlyparallel:everythingthatexistsinonehasanexactcounterpartintheother.Infact,Spinozaclaims,thosecounterparts,andtheattributesonwhichtheydepend,areinrealitythesamething:thereisonlyasinglesubstance,whichisexpressedinthosetwodifferentways(andincountlessothersunknowntous).Whatarewetomakeofsuchastory?Ithasbeenreadinavarietyofways,asequivalenttocertainvarietiesofBuddhistthought,asareturntosomekindofprimitiveanimism,andasamysticalpoeticvisionofaWorldSoulunderlyingallthephenomenaofnature.Itmaybealloranyofthosethings;butwhatitisalsoisanattempttoextendDescartes’smodernismintothementalrealm,andbysodoingtocreateauniversalexplanatoryframeworkforacompletescienceofnature.Descartes’sviewofthemindhadbeenasurprisinglyconservativeone,inthattheappearance/realitydistinctionwhichsodominatedhisaccountofmatterwasalmostcompletelyabsentfromit.Whereasheheldthatthecommon-senseviewofthemate-rialworldiscompletelytransformedwhenwecometoseeitwiththeeyeofreason,thecommon-senseviewofthemindturnsoutonhisviewtobesurprisinglyaccurate.Theuneducatedlayman’sbeliefthatthemindissomeextra,immaterialthingwhichisunitedwiththebodyandwhichleavesitatdeathissharedbythephilosopher.Theexpertmayhaveamorecompleteandamoresystematicknowledgeofthemind’soperationsthandoesthepersoninthestreet,butthatknowledgeisstillexpressedintermsofthecategoriesoftheordinaryperson’sexperience,suchaswilling,under-standing,feelingandperceiving.Themindisknown,asitwere,entirelyfromwithin:whatitistothinkjustistohavecertainkindsofexperience,andthereisnodeeperlevelofunderstandingwhichexplainsthatexperienceinthewayinwhichaknowledgeofCartesianphysicsexplainstheobservablephenomenaofthenaturalworld.ThoughtSpinoza’sAttributeofThoughtisanattempttoprovidethatdeeperlevelofunder-standing,andbysodoingtogiveanobjective,scientificexplanationofthephenomenonofthoughtpreciselyparalleltothatwhichDescarteshadgivenofthematerialworld.JustasDescartesextendstheconceptofmaterialityfromactualsolidobjectstoauni-versalextension,andinsodoingexplainsawaythosesolidobjectsasmerelylocalcon-centrationsofauniversalstuff,soSpinozatriestoextendtheconceptofthoughtfrom660\nSPINOZAANDLEIBNIZindividualthinkingagentstoauniversalmentality,andbysodoingseekstoexplainawayourindividualmindsasmerelylocalconcentrationsofauniversalcharacteristicofallthings.Everythingisextended,evenso-calledemptyspace,andwhenextensionisarrangedappropriately,theresultiswhatweexperienceasasolidobject.Similarly,every-thingismental,evenso-calledinertobjects,andwhenmentalityisarrangedappropri-ately,theresultiswhatweexperienceasathinkingthing.Theideaisastrangeone,butperhapsnomoresothantheparallelmoveinthematerialworld,whichtodaywefindquiteeasytoaccept.Ifitcanbemadetowork,itoffersagenuinelynon-materialistalternativetoCartesiandualism–abasisforascienceofnaturewhichseeshumanbeingsasnaturalobjectswhicharetobetreatedinthesamewayasanyothernaturalobject,butwhichatthesametimepreserveshuman(andpotentiallyanimal)subjectivityasanineliminablefeatureoftheworld.Thepricehepaysforthoseadvantages,though,isonewhichmanypeopleareunwillingtopay.Herulesoutanypossibilityoffreewill,andinsiststhatlikeallothernaturalphenomenahumanbeingsareentirelydeterminedeffectsoftheunderlyingfactsofnature.Thestructureofhistheoryalsodeniesthepossibilityofanycross-attributecausation:mentalactscanneverbegivenphysicalexplanations,orviceversa,buteachmustbeunderstoodthroughthelawsofitsownnature.Morefundamentallystill,hedeniestheultimateindividuality,andevenidentity,ofthehumanmind.Amind,forSpinoza,ismerelyalocalconcentrationofthought,conventionallyindividuatedfromotherconcentrationsaroundit.Itisnotaseparate,individualthingwithitsowniden-tityandindependence,butmerelyaconvenientwayoftalkingaboutaconstantlychang-ingbodyofinterrelatedthought,whichcomestogether,developsandchangesovertime,andeventuallydissipates,likeawhirlpoolinthebathwater.Whetherornotwefindthepictureacceptable,itsmotivationisclear:whenseenthroughtheeyeofreason,allthephenomenaofnatureareseenaspreciselythat–merephenomena,generatedbytheunchangingfeaturesofnaturewhichatruesciencewouldreveal.Thetaskofunderstandingistherefore,asDescartessaid,toriseabovethelevelofexperience,atwhichweperceivetheworldasamultiplicityofobjectsintimeandspace,andtoseethingsfromthepointofviewofeternity(subspecieaeternitatis),astemporarylocalmanifestationsoftheunchangingfactsofnature.ButwhereasDescartesandhisfollowerssawthattaskasapplyingonlytomaterialobjects,Spinozaseekstoextendittoeverythingthatexists.1.3HumanityTheaccountofhumanitywhichemergesfromthismetaphysicisamixtureofausterityandpassion.Aswehaveseen,Spinoza’soverridingconvictionthathumanbeingsarepartsofnature,naturalphenomenalikeanyother,andthereforetobeunderstoodinthesamedispassionate,naturalisticway,leadshimtoanoutrightdenialoftheCartesianfreewill:humanactionsarenaturalevents,liketheactionsofcloudsorrainstorms,andtounderstandthemistoseehowtheyfollownecessarilyfromtheunchangingfactsofGodorNature.Theconventionalobjectiontoanysuchdeter-minism,ofcourse,isthatitmakesmoraljudgementsimpossible,becauseitmakesnosensetopraiseortoblamewhatcouldnothavebeenotherwise,andwithtypical661\nRICHARDFRANCKSANDGEORGEMACDONALDROSSremorselesslogicSpinozathereforeacceptsthatmorality,atleastasconventionallyunderstood,isimpossible.Termslike‘right’and‘good’,heclaims,aremerelywordswhichweapplytothingsthatwewant,andwedeceiveourselvesifwethinktheyareanythingelse.Similarly,politicsandthelegalsystemarepurelypragmaticdevicesfororganizinganddirectinghumanaction,andtoseekanyhigherpurposeorclaimanyhigherjustificationforourdecisionsisabsurd.Thestorydependsonhisunderstandingofhumanbeingsaspurelynatural,appetitivecreaturesinaworldoffellowstrugglers,bothhumanandnon-human.Likeanyotherpartofnature,whetheritisonewenormallythinkofasconsciousornot,westrivetomaintainourexistence,todevelopournaturesandtorealizeourpotentials.Insodoingweperceivehelpsandhindrances:somethingswefindtobebeneficialtoourdevelopmentandenableustosurviveandtoprogress;othersthwartandobstructourgrowthandpreventusfromrealizingouraims.Thosebasicfactsproducethethreefun-damentalemotionalstatesofDesire,PleasureandPain,andhethendevelopsarigorousandsystematicanalysisofallotheremotionsasbeingbuiltupoutofcombinationsofthosethree,inawaywhichispreciselyparalleltothatinwhichallchangeinthemater-ialworldcanbeseentobebuiltupoutofthebasicfactsofmotionandresistance.Yetoutofthisaustere,rational,mechanicalaccountofhumanexistencetherearisesanunderstandingofhumanlifeandbehaviourwhichissurprisinglyrich,andwhichprovidesanotionofthegoodlifeforhumanbeingswhichmakesclearwhyhisonegreatworkiscalledtheEthics.Weareallcapable,tosomeextent,ofunderstand-ingtheworldandourselves,ofseeingthingsastheyreallyare.Andthatunderstand-ingcanbeenhancedandextended.Totheextentthatwedounderstandourselvesandtheworldaroundus,tothatextentweactrationally,andweknowwhyweact–ouractionsaregroundedinwhatisinternaltous.Inthatsense,whenwethinkclearlyweareincontrolofourownaction,whereaswhenweactirrationallyweareatthemercyofexternalforcesoperatingonus.Thoseforcesmaybeinternalorexternal:whenwegazeattheSunlikeafool,andjumptotheconclusionthatitmovesaroundtheEarth,wearepassive,enslaved:ourbelieffoistedonusbycircumstances.Bycontrast,whenweunderstandwhythingsappearthatway,weareactive,free,rationalagents–ourbeliefsaretheresultofourownthinkingprocesses.Inexactlythesameway,whenwereactangrilytoaninsultweareatthemercyofourownpassiveemotions,drivenhitherandthitherbyalienforceswithinourselves.Butwhenweunderstandwhytheinsultwasoffered,andunderstandwhywefeelaswedoinresponse,wedonotmagicallyloseourappetitivenatureandriseabovetheemotionallifetoalifeofpurerationality,butwearenever-thelessabletochoosehowtoreact,andinthatsense,eventhoughouractionremainsasrigorouslydeterminedasanyother,wearefree,self-determiningagents.Thepictureisnowalmostcomplete.Thisnotionoffreedom,ofself-determination,istheonlytruegoodforhumanbeings,becausetheactive,rationallifeistheonewhichasamatteroffactinthelongtermisbestcalculatedtoproducethestable,prosperousandhappylife,anditisthatfactofnaturewhichmoralistshavetriedintheirmislead-ingwaytoteachbytheirsystemsofgoodandevil.Becauseitisanemotionallife,andbecauseitisbasedonatrueunderstandingofoneselfandone’ssurroundings,SpinozacanalsoclaimthatitistheIntellectualLoveofGod,andthereforethatitisalsothelessonwhichthegreatreligiousteachersoftheworld,ofallfaiths,havebeentryingto662\nSPINOZAANDLEIBNIZputacross,oncewepurgetheirmessageofitsanthropomorphicandfigurativelan-guage.Andbecauseself-determining,rational,freeagentswillnecessarilytendtoagreewithothersofthesamekind,therequirementsofcivillifeandpoliticalobligationwillalso,whenproperlyunderstood,turnouttobefoundedonthesameconceptionofhumanbeingsintheireternalandunbreakabledependenceonGodorNature.InthiswaywhatSpinozaoffersisindeedanaccountofthegoodlifewhichisfirmlygroundedinanobjective,scientificaccountofhumanbeings,andwhichisinturnlocatedwithinanall-embracingmetaphysicalstructureofallreality.Ifitweresuc-cessfulitwouldbringtogetherscienceandreligion,reasonandemotion,objectivityandmoralityinawaywhichisuniqueinthemodernworld.Butthepriceisperhapstoohighformostofustopay.AccordingtoSpinoza,allthatiswrongwithcruelty,allthatiswrongwithselfishness,isthatalifeofthatkindisnotinthelongtermbeneficialeithertotheagentortoherneighbours.Similarly,allthatiswrongwithdespotismandwithtotalitarianismisthatsystemsofthatkindtendinthelongrunnottolast,andnottobeconducivetothebestinterestsoftheirmembers.Manypeoplestillfeeltheneedforsomestrongerweaponagainsttheirenemies.2LeibnizPhilosophershavealwaysbeennotoriousforthestrengthoftheirdisagreements,andfortheirinabilitytoresolvethembyrationaldebate.Leibnizwasadiplomatbyprofes-sion,andheappliedthetechniquesofdiplomacytobringaboutpeaceandharmonyintheworldofphilosophy.Althoughhisphilosophicalsystemhadmanyoriginalfeatures,hepresenteditnotasonenewsystemamongmany,butasincorporatingthecommongroundonwhichthewarringfactionscouldagree.Hisapproachcanbesummedupinhisdictumthatphilosophersaregenerallyrightinwhattheyassert,andwronginwhattheydeny.Helikedtoillustratethiswithananalogy:differentviewsofatowngiveprominencetosomefeaturesandhideothers,sothatobserversatvariousviewpointswillgivedifferentdescriptionsofit.Butthesedescriptionsaremutuallycompatible,aslongasindividualobserversdonotdenytheexistenceoffeaturestheycannotthemselvessee.Leibniz’sambitionwastoachieveaGod-likeperspectivewhichencompasseseverything,andrevealsallotherperspectivesaspartial,buttrueasfarastheygo.Unliketheothermodernphilosophers,whosawthemselvesasmakingamoreorlesscompletebreakwiththepast,Leibnizacceptedallphilosophersasmakingpositivecontributionstophilosophicaltruth–notmerelycanonicalfiguressuchasPlato(c.429–347bc)(chapter23)andAristotle(384–322bc)(chapter23),butmoreexoticthinkerslikePythagoras(b.c.570bc)andtheNeoplatonists,themuch-deridedscholasticphilosophers,andmysticsandalchemists.Again,hewaslessEurocentricthanmostofhiscontemporaries,andhelearnedasmuchashecouldaboutChinesephilosophy,toshowhowitcomplementedtheWesterntradition.Butdespitehisreconciliatoryprogramme,hewasassharpinhiscriticismsofhispredecessorsasanyone.InoutliningthemainthemesofLeibniz’spositivephilosophyunderthesameheadingsofNature,God,andHumanity,weshalltakenoteofhisargumentsagainstDescartes,Malebranche,andSpinozainparticular.663\nRICHARDFRANCKSANDGEORGEMACDONALDROSS2.1NatureDescartes,MalebrancheandSpinozaallheldthattherealmofnaturewasaunitarywhole,whetherdistinctfrom,ormerelyanaspectofGod.Twoalternativeviewsoftheworldwhichtheyconsciouslyrejectedweretheatomistviewthatitderivesallitsrealityfromthediscreteatomsofwhichitiscomposed,andthevitalistviewthatallphysicalobjectsareeitherlivingbeings,orcollectionsoflivingbeings.Leibnizheldthatallthreeviewscontainedpartofthetruth,andthattheycouldbereconciledbyrevivingtheAristotelian–scholasticconceptofasubstantialform.Firstly,Descartes,MalebrancheandSpinozawererighttoinsistonthesystematicwholenessofnature–itismorethanjustachaoticjumbleofatoms.ButSpinozainparticularwaswrongtosaythatonlytheuniverseasawholeisarealsubstance,sinceitisacompoundofitsparts,andthepartsmustbeatleastasrealasthewhole.Thequestionis,whataretheultimatepartsofwhichtheuniverseisacompound?Everyoneagreedthateverydayphysicalobjectslikesticksorstonesdonotqualify,sincetheyareonlytemporarycollectionsofsmallerparts.Theatomistsheldthatthereareverysmallpartsofmatterofvariousshapesandsizes,whichcannotbedividedintofurtherparts,andwhicharetheultimatesubstances.ForLeibniz,sizewasirrelevant,sincewhethersomethingisbigorsmallisrelativetotheobserver–agrainofsandtoosmalltobenoticedbyanelephantwouldbeamountaintoamicrobe.Fromamathe-maticalpointofview,thereisnolesspossibilityofsubdividingasphericalatomthanacannonball;anditissimplyarbitrarytoassertthatatoms,unlikecannonballs,canneverbebrokendownintotheircomponentparts.Inshort,thereisnosuchthingasthesmallestconceivablematerialatom.TheconclusionLeibnizdrewwasnot(likeDescartes,MalebrancheandSpinoza)thattherearenofundamentalunitsofnature,butthattheseunitsarenotmaterialparti-cles.Intheirenthusiasmforthemechanisticviewoftheworld,themodernphiloso-phershadoverlookedthecrucialdistinctionbetweenthingswhicharemerecollectionsofparts,andthingswhichareorganicwholes.ForDescartes,aclock,acatandthelivebodyofahumanbeingwouldallbetransitorymanifestationsofmatterofthesamekind.Themaindifferencebetweenacatandaclockisthatthemechanismismorecom-plicated;thedifferencebetweenacatandahumanbodyisthatthehumanbodyhasatemporaryspecialrelationshipwithanimmaterialsoul.ForLeibniz,theclockisamerecollectionofparts,whereasboththecatandthelivehumanbodyareorganisms.Anorganismisabeingofacompletelydifferentkind.Atthecommon-senselevel,ithasthecapacitytoinitiatemotion,toactpurposefully,toreacttostimuliandtoreproduce.ButofspecialsignificancetoLeibnizwashisbeliefthatanorganismisaunitarywhole,inthatitspartsarepeculiartoitself–everyorganandeverycell(indeed,everypartofeverycell)bearsuniquecharacteristicsthatlinkittothebodytowhichitbelongs.Soevenifthepartsofthebodybecomephysicallyseparated,theyarestillpartsofthesameorganism(ratherasafamilyisasinglefamily,evenifitsmembersdonotliveunderthesameroof).ForLeibniz,then,theonlyfundamentalunitsofnatureareorganisms,sincetheyalonepossesstrueunity.Inthis,Leibnizsidedwiththevitalists,whoseunderstandingofnaturewasmodelledonthepurposeful,vitalforcesoflivingbeings,andwhosawlifeeverywhere.Eventhoughthemajorityofmacroscopicphysicalobjectsweseearoundusarenot664\nSPINOZAANDLEIBNIZthemselveslivingorganisms,theyareinfactcomposedoftinyorganisms.Leibnizappealedtotheevidenceoftherecentlyinventedmicroscope,whichrevealedappar-entlyhomogeneoussubstances,suchaschalk,bloodorsemen,ascontainingmassesoftinycreatures.Indeed,sincematterisinfinitelydivisible,hemaintainedthateveryorganismisacolonyofsmallerorganisms,andsoonadinfinitum.However,althoughLeibnizacceptedthemetaphysicsofthevitalists,hecouldnotgoalongwiththeirrejectionofthenewmechanisticscience.Inordertoachievearecon-ciliationofthetwoworldviews,hedrewonthetraditionalAristoteliandistinctionbetweenmatterandform.ForAristotle,everybeinghastwoaspects:itsform(theactiveprinciplewhicheachindividualshareswithothermembersofthesamespecies)anditsmatter(thepassiveprinciple,whichbearsthecharacteristicsbywhichoneindi-vidualisdistinguishedfromanother).Inthecaseofanimatebeings,theformwasiden-tifiedwiththesoul.PerhapsinfluencedbySpinoza’stheorythatextensionandthoughtaretwoattributesofthesamesubstance,Leibnizmaintainedthatnaturehadtwoaspects:thematerialaspect,consistingofthebodiesofalltheorganismsofwhichitiscomposed;andthevitalaspect,consistingofalltheformsorsouls.Thematerialaspectcouldbeunderstoodonlyintermsofthemathematicallawsgoverningtheinteractionsofpiecesofmattercharacterizedbydifferentshapes,sizesandmasses.Thevitalaspectcouldbeunderstoodonlyintermsofthelifeprincipleofstrivingforperfection.AswithSpinoza,therewasaperfectone-to-onecorrelationbetweenwhatistrueofmatter,andwhatistrueofforms,or‘monads’(unities),ashelatercalledthem.So,forexample,abulletfiredfromaguncanbeseenequallyasblindlyobeyingthelawsofmechanics,andasstrivingforthebestpossiblestatewiththegreatesteconomyofeffort.2.2GodIftheuniverseconsistsinaninfinityofindividualsubstances,howdoesitformasys-tematicwhole?Leibniz’sanswerwasthatGodisneededtoensurethatthecomponentsoftheuniverseinteractasharmoniouslyaspossible–indeed,heoftenreferredtohisphilosophicalsystemasthe‘systemofuniversalharmony’.Thisdoesnotmeanthattheuniverseisperfectinallitsparts,otherwiseitwouldbeanabsolutelyperfectbeing,andthereforeidenticalwithGod.LikeDescartesandMalebranche,butunlikeSpinoza,LeibnizinsistedthatGodmustbedistinctfromtheuniversehecreated.Butasaperfectbeing,Godmusthavecreatedthebestuniversewhichitislogicallypossibletocreate,sothatifitweremadebetterinsomerespect,itwouldbeworseoverall.Theevil,painanduglinessthatweseearoundusarerealenough;butthesolutiontotheproblemofevilistorecognizethattheycontributetoagreatergood.MonadsandtheHarmonyoftheUniverseSowhatisitfortheuniversetobeaharmonioussystem?Monads,beingsoul-like,havecharacteristicsanalogoustothoseofhumansouls.Inparticular,theyhavesomethingakintoaperceptualstateinhumansouls,onlylackingconsciousness.Ineffect,theyrep-resenteverythingthathasanyinfluenceonthem,withgreaterorlessclarity,dependingonthestrengthoftheinfluence.Forexample,themotionofaparticleisaffected,howeverremotely,byallotherparticlesintheuniverse;andinorderto‘know’which665\nRICHARDFRANCKSANDGEORGEMACDONALDROSSdirectiontomovein,itmustregisteralltheseinfluences.ThisledLeibniztodescribemonadsas‘livingmirrorsoftheuniverse’,eachrepresentingallothermonadsfromadistinctandindividualpointofview.Godperceiveseverythingwithequalclarity,andwithoutanypointofview;whereascreatedmonadsarelimitedbytheirperspectivalpoint.Theharmonyofthetotalsystemconsistsinthefactthattheperceptionsofeachindividualmonadinterlocktoformasingle,consistentviewoftheuniverseasperceivedbyGod,likeajig-sawpuzzleofinfinitelymanydimensions.Moreover,sincethepercep-tualstateofeachmonadisconstantlychanging,theharmonyoftheuniversehastobere-establishedfromoneinstanttothenext.ButhowdoesGodbringitaboutthatthewholeuniverseiskeptinperfectharmony?Leibnizheldthattherewerethreepossibletheories,whichheillustratedwiththethen-popularmetaphorofGodastheperfectclock-maker.Tokeeptheaccountassimpleaspossible,supposethatGodhasmadejusttwoclocks,whichkeepinperfecttimewitheachothertoeternity.Onetheorywouldbethathehasdevisedaninvisiblemechanismwhichphysicallycompelstheclockstokeepintime(e.g.rodsconnectingtheirhands).Thiswouldbeanalogoustosayingthattheuniverseisharmoniousbecausetherearenecessarycausalconnectionsbetweenitscomponents.But,asMalebranchehadshown,wecannotperceiveanysuchconnections,andwecannotconceivehowonepieceofmattercantransferitsenergytoanother.Moreover,giventhattheultimaterealitiesarespiritualmonads,itisevenmoredifficulttoconceivehowanythingcouldpassfromonemonadtoanother.AsLeibnizhimselfputit,monadshavenowindowsthroughwhichanythingcanpassinorout.Thesecondtheoryisthat,assoonastheclocksbegintodiverge,theclock-makeradjustsoneofthemsothattheykeeptellingthesametime.LeibnizattributedthistheorytoMalebranche,sincehetookhimassayingthat,whenIwilltoraisemyarm,nothingwouldhappenunlessGodintervened.Inthenormalcourseofbodilyevents,myarmwouldhaveremainedstationary,ormovedinsomeotherwayforpurelyphysiologicalreasons.So,accordingtoLeibniz,Malebranche’stheoryofoccasionalcausesimpliesthatGodperpetuallyadjuststhemechanismofnatureinordertokeepitinharmonywithhumanvolitions;andanydivineinterventionofthissortisamiracle.LeibnizrejectedoccasionalismonthegroundsthatGodwouldnotbeperfectifhehadtomakeinnumer-ablemiraculousinterventionsintotheworkingofhiscreation.Aswillbeevidentfromouraccountofoccasionalisminthepreviouschapter,thisisprobablyamisinterpreta-tionofMalebranche,whosepositionwasclosertoLeibniz’sthanthelatterrealized.Theonlyalternativeisthethirdtheory,whichissimplythattheclocksweremadesoperfectly,thattheykeepintimeforeverwithoutanyinvisibleconnectionsormiracu-lousadjustments.Inotherwords,eachmonadhasbeenpre-programmedwithsuchaccuracy,thatitspontaneouslyevolvesinperfectharmonywitheveryothermonad.ForLeibniz,theessenceofeachindividualmonadisauniquelaworprogramme,whichcontrolsitsdevelopment,andspecifieswhatstateitwillbeinatanygiventime.God’sperfectionconsistsinhiscreatinganinfinitenumberofmonadswhichwillevolveinperfectharmonywitheachother.Inusingtheanalogyoftheclocks,Leibnizwasrestrictedbythetechnologyofhisday.Inmodernterms,abetteranalogymightbetwospaceshipswhicharepre-programmedtodocktogetheratacertainpointinspace,withoutreceivinganyinformationabouteachother’spositions,andwithoutany666\nSPINOZAANDLEIBNIZadjustmentsfrombase.Itisconceivablethatthiscouldbedone;butitwouldrequirealmostdivinepowersofpre-programming.SofarwehaveconsideredGod’screativeactasanalogoustoacraftspersonmakingamachine.Butatadeeperandmoremetaphysicallevel,Leibnizexplaineditasoneofendowingaconceptwithreality.SinceGodisomniscient,hismindcontainsallpos-sibleconcepts,inalltheirpossiblepermutationsandcombinations.Everylogicallyconsistentsetofpossibilitiesconstitutesapossibleworld.Thesepossibleworldscanbegradedastotheirperfection–theirfullnessofbeing,coherence,variety,orderliness,harmonyorbeauty,forinstance.Inthecompetitionforexistence,thebestofthepossibleworldsistheonewhichisactualized.OneofLeibniz’sprojectswastodeviseanideallanguageforeverypossibleconcept,whichwouldmaketheirlogicalrelationshipspatent(his‘universalcharacteristic’).Ifconceptsweregivennumbersratherthannames,andthemainlogicaloperatorsreducedtoarithmeticaloperators,itwouldthenbepossibletoreplacefalliblehumanreasoningbypurelymechanicalcomputation–andhedesignedandbuiltaprototypemachinefordoingjustthis.Inhismoremysticalmoments,hespeculatedthatallconceptsweregeneratedoutofthebinarynumbers1and0,with1representingpurebeing,orGod,and0representingabsenceofbeing,orthevoid.IndividualsubstancesaredependentonGodlikethelightradiatingfromtheSun;andtheyaredistinctfromGodandfromeachotherbecauseoftheirvaryingproportionsofdarknessornon-being.Ineffect,theuniverseisthemostperfectnumber,otherthan1itself.2.3HumanityLeibniz’saccountofhumankindfallssomewherebetweenthoseofDescartesandMalebrancheontheonehand,andofSpinozaontheother.HeagreedwithDescartesthateachhumanisadistinctsubstanceinitsownright;buthealsoagreedwithSpinozathathumansarepartofnature,andthattherearenospecialsoul-substanceswhichexemptthemfromthelawsofnature.LeibnizmademanydetailedcriticismsofDescartes’sabsolutedistinctionbetweenmindandbody,ofwhichoneexamplewillsufficehere.Inanumberofhiswritings,Descartesproposedatheorytoaccountforhowtheimmaterialsoulcouldinfluencewhathappensinthebody.Inessence,histheorywasthatbodilymovementsarecon-trolledbytinyparticlesinthebrain(‘animalspirits’),whichemergefromthenervesattachedtothesenseorgans,andaredeflectedbythepinealglandintonervesleadingtothemuscles.Theprecisepositioningofthepinealglandisdeterminedbyemotionalfactors,suchaswhetherthebodyisinastateofangerorfear;andthisinturndeter-mineswhethertheperceptionofsomethingdangerousresultsinfightorflight,forexample.However,thesoul(whichhasitsseatinthepinealgland)canmakeinfinitesi-maladjustmentstoitsposition,sothatthebodyfightsinsteadoffleeing,orviceversa.Leibnizcriticizedthistheoryontwogrounds.First,thelawsofmotionwhichDescarteshadformulatedwerewrong,becausehemaintainedthatthetotalquantityofmotioninthematerialuniversewasconstant,withouttakingintoaccountthedirec-tionofmotion.Hethoughtthatthesoulcouldalterthedirectionofmotionofanimalspiritsinthebrainwithoutcontraveningthelawsofmaterialnature.Butinfactitisforceorenergywhichisconserved,andittakesenergytoalterthedirectionofmotion667\nRICHARDFRANCKSANDGEORGEMACDONALDROSSofamaterialparticle.Soanyinterventionbythesoulwouldhavetoaddenergytothematerialworld,andcontravenethelawoftheconservationofenergy.Secondly,Descartesleavesitentirelyunexplainedhowthesoulcaninfluencethedirectionofmotionofparticlesinthebrain.Itisnotenoughtosaythattheamountofdeflectionissoinfinitesimallysmallthatitcanbebroughtaboutevenbyanimma-terialsubstance,sincethesoulisofacompletelydifferentnaturefrommatter,andcannothaveanyrelationshipwithit.Itisnotevenpropertosaythatithasitsseatinthepinealgland,sinceanimmaterialsubstancecannothaveanylocationinspace.LeibnizsuggestedthatDescarteswasthinkingofthesoulasliketheriderofahorse.Thehorsesuppliesalltheenergyforitsmotion,withoutanycontributionfromtherider;buttheridercontrolsthedirectioninwhichthehorsemoves.However,themeansbywhichtheridercontrolsthehorsearephysical–pullingatthereins,ordigginginthespurs.Butthesoul,asanimmaterialsubstance,hasnoneofthesephysicalmeansatitsdisposal,anditiswhollyunintelligiblehowitcanexertaphysicalinfluenceonthefunctioningofthebrain.Underlyingpenetratingcriticismssuchasthese,thereisadeeperdifferencebetweenLeibnizandDescartes.Descarteshadrejectedtheprevailingscholastic–Aristoteliantheorythatthesoulisnomorethantheformofthebody,andhadrevertedtothePlatonicandAugustiniantheorythattheyareentirelyseparatesubstancesintheirownright(andwithsomedoubtsastowhetherbodyexistsatall).LeibnizwasmoreAristotelian,andheheldthatsoulsareinextricablyboundupwithmatter,asitsform.Indeed,justastherecanbenoformwithoutmatter,therecanbenomatterwithoutform–whichiswhyallgenuinesubstancesarelivingbeings,withasoulanalogoustothatofhumans.However,althoughLeibnizmaintainedthatthereisnoontologicaldifferencebetweenhumanbeingsandothersubstances,herecognizedthattheydiffersufficientlyindegreetogivehumanbeingsaspecialstatusintheschemeofthings.Althoughmostofourperceptionsare,likethoseofothermonads,belowthethresh-oldofconsciousness,someofthetime,someofourperceptionsaresufficientlyclearanddistincttogiverisetoconsciousness–astateinwhichwedonotmerelyperceive,butareawareofourselvesasperceiving.Itisthiswhichdistinguisheshumanbeingsfromothersubstances.Similarly,ourdesiresandambitionsarenotmereunconsciousgoal-directedbehaviour,butsetinacontextofself-improvement,rationalityandmoralchoice.DrawingonStAugustine’s(354–430)(chapter24)greatwork,TheCityofGod,LeibnizheldthathumanssharedwithangelsandothersuperiorconsciousandrationalbeingstheprivilegeofbeingmembersoftheCityofGod,orthe‘realmofgrace’.Justasformandmatteraredifferentaspectsofthesamereality,soaretherealmsofgraceandofnature.Aspartofnature,allhumanactionsareexplicableinpurelydeter-ministic,mechanicalterms,asmotionsofparticlesofmatter.Butsincewearealsociti-zensoftherealmofgrace,ouractionsarealsodescribableastheoutcomeofcompletelyfree,rationalchoices.Asscientists,weinvestigatethecausaldeterminantsofhumanbehaviour,butassubjectsofGod’skingdomweareeachcontributingtothefulfilmentofGod’splanforthemoralperfectionoftheuniverse.Thankstothegiftofreason,wecanandshouldmakeitourgoalinlifetoraiseourselvesabovealimitedandparochialviewoftheworld,toonewhichisascloseaspossibletothedivineperspective.668\nSPINOZAANDLEIBNIZFurtherReadingForthetwophilosophersfeaturedinthischapterwefirstsuggestamajorsourceoftheirwritingstranslatedintoEnglish:Spinoza(1985)andLeibniz(1969)andthenamorecon-venientselectionoftheirworks:Spinoza(1992)andLeibniz(1995).ReadersmightseeksomeunderstandingofSpinoza’sdemandingmasterpiece,theEthics(published1677),beforeturningtohisotherwritings.ThefullestaccountsofmanyofLeibniz’sideasappearinhisletters,butanoverviewmaybegainedfromsuchworksasDiscourseonMetaphysics(writtenin1686,published1846)andthematureaccountsinNewEssaysonHumanUnderstanding(written1704,published1765),Theodicy(published1710)andtheMonadology(written1714).HisCorrespondencewithArnauld(publishedposthumouslyin1846)andTheLeibniz–ClarkeCorrespondence(publishedposthumouslyin1717)areclassicexchangeswhichhelptoexplainthegroundsforhissystem.Readersmightbegintheiradditionalreadingwithbookscomparingallthreeauthors:Cottingham(1988)andWoolhouse(1993).FreemanandMandelbaum(1975),Grene(1979),andGreneandNails(1986)arecollectionsofferingwide-rangingcriticalresponsestoSpinoza.Donagan(1988)exploresSpinoza’sdoctrineswithbalanceandsophistication,whileBennett(1984)examinesthedetailsofSpinoza’ssystemwithengagingvigour.Leibniz’slifeofbrilliantachievementinmanyfieldsisexploredinAiton(1985).MacDonaldRoss(1984)providesahelpfulintroductoryaccount.Woolhouse(1993)andJolley(1995)arecollectionsofferingguidanceformoreadvancedwork.ReferencesSpinozaSpinoza,B.1985:TheCollectedWorksofSpinoza,vol.1(editedandtranslatedbyE.Curley).Princeton,NJ:PrincetonUniversityPress.——1992:EthicswiththeTreatiseontheEmendationoftheIntellectandSelectedLetters(editedbyS.Feldman,translatedbyS.Shirley).Indianapolis,IN:Hackett.——1994:ASpinozaReader(editedandtranslatedbyE.Curley).Princeton,NJ:PrincetonUniversityPress.WritersonSpinozaBennett,J.1984:AStudyofSpinoza’sEthics.Cambridge:CambridgeUniversityPress.Donagan,A.1988:Spinoza.London:HarvesterWheatsheaf.Freeman,E.andMandelbaum,M.(eds)1975:Spinoza:EssaysinInterpretation.LaSalle,IL:OpenCourt.Garrett,D.1996:TheCambridgeCompaniontoSpinoza.Cambridge:CambridgeUniversityPress.Grene,M.(ed.)1979:Spinoza:ACollectionofCriticalEssays.NotreDame,IN:UniversityofNotreDamePress.Grene,M.andNails,D.(eds)1986:SpinozaandtheSciences.Dordrecht:D.Reidel.Lloyd,G.1996:SpinozaandtheEthics.London:Routledge.LeibnizLeibniz,G.W.1969:G.W.Leibniz:PhilosophicalPapersandLetters(editedandtranslatedbyL.E.Loemker).Dordrecht:D.Reidel.——1995:G.W.Leibniz:PhilosophicalWritings(editedbyG.H.R.ParkinsonandtranslatedbyM.MorrisandG.H.R.Parkinson).London:Everyman.669\nRICHARDFRANCKSANDGEORGEMACDONALDROSSWritersonLeibnizAiton,E.J.1985:Leibniz:ABiography.Bristol:AdamHilger.Jolley,N.(ed.)1995:TheCambridgeCompaniontoLeibniz.Cambridge:CambridgeUniversityPress.MacDonaldRoss,G.1984:Leibniz.Oxford:OxfordUniversityPress.Woolhouse,R.S.(ed.)1993:G.W.Leibniz:CriticalAssessments.London:Routledge.WritersonDescartes,SpinozaandLeibnizCottingham,J.1988:TheRationalists.Oxford:OxfordUniversityPress.Woolhouse,R.S.1993:Descartes,Spinoza,Leibniz:TheConceptofSubstanceinSeventeenth-CenturyMetaphysics.London:Routledge.DiscussionQuestions1DoesSpinozaofferabetteruniversalexplanatoryframeworkforascienceofnaturethanDescartes?2Shouldtheappearance–realitydistinctionapplytothemind?3Canweexplainawayindividualmindswhileretainingmentality?4Whatwouldthingsbelikeifseenfromthepointofviewofeternity?Couldtherebesuchaviewpoint?5HowcanwedecidewhetherSpinozaiscorrectinclaimingthatthereisonlyonesubstance:GodorNature?6ShouldweacceptSpinoza’saccountofthelinkbetweenrationalityandfreeaction?7HowvaluableisSpinoza’saccountoftheemotions?8Shouldaphilosophicalsystemseektoincorporateratherthantorejectpastphilosophicalthought?9Whatisatstakeindecidingwhethernatureisasystematicwhole?10Couldthefundamentalunitsofnaturebeorganismsratherthanmaterialparticles?11Whataremonads?Inwhatsenseiseachmonadmeanttoreflectalltheothers?12Wouldpre-establishedharmonyorcausalinteractionamongmonadsbeabetterexplanationoftheharmonyoftheuniverse?13WhatistheroleofLeibniz’sconceptofpossibleworlds?Howcanwedeterminewhetherthisisthebestofallpossibleworlds?14CouldLeibniz’sprojectofdevisinganideallanguageberealized?15Doesthenotionofsubstancehelporhinderphilosophicalunderstanding?16DoesDescartes,SpinozaorLeibnizhavethebestunderstandingofhumanfreedom?17DoesDescartes,SpinozaorLeibnizgivethebestinsightintohowthemindandthebodyarerelated?670\n28HobbesTOMSORELLModernpoliticaltheoryoriginateswithThomasHobbes(1588–1679).InLeviathanandotherworks,Hobbespresentedableakpictureofviolenceanddisorderastheinevitableconditionofhumansinthestateofnature.Peace,themainconcernofpolitics,couldberealizedonlybyastrongsovereignestablishedthroughasocialcontract.Subjectswouldhavenorationalgroundstochallengetheruleofthesovereignsolongaspeacewasmaintained.Hobbes,whowroteduringtheturmoiloftheEnglishCivilWar,hashadwideinfluenceamongsubsequentpoliticalandsocialtheorists.Hismaterialistmethodisworthexamininginhisdiscussionofmanyothertopics.1IntroductionThomasHobbesisbestknownasaphilosopherofHUMANNATURE(p.259)andHUMANSOCIETY(pp.690–1).Heisfamousformaintainingthatthenaturalconditionofpeopleisoneofwar,inwhichlifeis‘solitary,poor,nasty,brutishandshort’,andhewasanearlysocialcontracttheorist:hebelievedthattheSTATE(pp.261–2)couldbeunder-stoodastheoutcomeofanagreementbetweenfreehumanbeingstosubmittogovernment.Leviathan(1651),publishedinLondon,iswidelyregardedashismasterpiece,anditisthemostcommonlyquotedsourceofhisviewsabouthumanityandaboutthestate.Itisnot,however,Hobbes’sonlyfull-lengthworkofpoliticalphilosophy–thereweretwoothers–anditisnottheworkinwhichhethoughtthathehadgiventhebestormostrigorousstatementofhistheoryofthestate.NorwaspoliticsHobbes’sonlypreoccupationasaphilosopher.Hewasaveryforthrightandthoroughgoingmaterialist,and,perhapsaboveall,acontributorto,andsystematizerof,someofthenewNATURALSCIENCE(chapters25–7)oftheseven-teenthcentury.InhisownopinionhedeservedtogodowninthehistoryofphilosophyastheoriginatorofatruescienceofnaturalJUSTICE(chapter8)orpolitics,andasaninnovatorinoptics.Healsoclaimedsomedistinctionasthefirsttoboilthewholeofphilosophyorsciencedowntoits‘elements’–theconceptsnecessarytounderstandbodyormatter,manandcitizen.HisElementsofPhilosophy,comprisingDeCive(1642),DeCorpore(1656a),DeHomine(1658),publishedinthreevolumesbetween1655and1658,wasthelargestintellectualprojectheundertook,occupyingasitdidnearly\nTOMSORELLtwodecadesofthoughtandwriting.InsofarasHobbeshadacompletesystemofphilosophy,itisexpoundedinthistrilogy.2Hobbes’sPoliticsTherearetworatherdifferentwaysinwhichHobbes’spoliticalwritingscanbeunder-stood.Ontheonehand,theycanbeinterpretedaspartisaninterventionsinadebatebetweenEnglishparliamentariansandroyalistsbetween1628and1650,thatis,theperiodthatclimaxedintheEnglishCivilWaroftheearly1640s.InthisdebateHobbes’sownpositionwasdeterminedbytheroyalistallegiancesofhisaristocraticemployers.HesidedwithCharlesI,andapprovedofmanyofhisactions–todowithtaxation,theforcedbilletingoftroopsinprivatehousesandthecurtailmentofparliamentarypowers–thatwerehighlyunpopularinEnglandintheprewarperiod.WasittheseactionsthatHobbes’stheoreticalworksofpoliticswereintendedtolegitimatize?ThesuggestionfitstheearlyElementsofLaw(1640)betterthanLeviathan,fortheElementswereactuallywrittenasasourcebookorbriefforaristocratswhospokeindefenceofthekinginparliamentarydebatesin1640,whileLeviathanseemstojustifytheactionsofanydefactosovereignpower,including,explicitly,theregimethatreplacedthemonarchyinEnglandafterthedeathofCharlesI.ThesecondreadingofHobbes’spoliticsisasacontribution,albeitarevolutionarycontribution,toaloftydebateconcerningquestionsofpoliticsintheabstract,ques-tionsdiscussedinalonglineoftreatisesgoingbacktotheGreeksandtheHebrews.ThisreadingfitsneithertheElementsofLawnorLeviathanaswellasitdoesthethirdofHobbes’sthreepoliticaltreatises,DeCive(1642).DeCivehasacertainpre-eminenceamongHobbes’sthreepoliticaltreatises:itwasonthestrengthofwritingthisworkthatHobbesclaimedtohavedonesomethingoriginalinPOLITICALPHILOSOPHY(chapter8).DeCivewasthefirstbook,accordingtoHobbes,todealscientificallywiththerecog-nizedcentralquestionsofpoliticalphilosophy.Otherworks,heconceded,hadtakenupthosequestions,butnotbyamethodthathadanychanceofputtingconclusionsbeyonddispute.DeCive,ontheotherhand,purportedtodoforitssubjectmatter–com-monwealthsorbodiespoliticandtheirmaintenance–whathadonlyeverbeendonebeforeforlinesandfigures.Inshort,DeCivewassupposedtoexpoundascienceofpoliticscomparabletogeometry,withconclusionsascompellingtoanopenmindasthoseofEuclid(fl.300BC).3HumanNatureandtheStateofWarDeCivewaspublishedin1642,twoyearsaftertheElementsofLaw,andnearlynineyearsbeforetheappearanceofLeviathan.Itopens(ch.1,ii)withastrongdenialofoneofthecardinalprinciplesofARISTOTLE’s(chapter23)politics,thathumansarebynaturecutoutforlifeinthepolis.AccordingtoHobbes,humanbeingsarenotnatu-rallymadeforthepoliticallife.Theythinktoomuchofthemselves,puttoomuchvalueonpresentgratification,andtheyarebadatpredictingtheconsequencesoftheir672\nHOBBESactions.Oneeffectofthesetendenciesisforpeopletocomeintoconflictwithoneanother,especiallywhentheyfeelthattheyareundervaluedinotherpeople’seyes.Theconflictcanconsistofthedenigrationofonepersonintheconversationofanotherperson,oritcantaketheformofoutrightquarrellingandevenviolence.AndifthesearefamiliarpatternsofbehaviourwhenthereisLAW(chapter13)andcustomandgoodmannerstorestrainpeople,howmuchmoreextrememustthehostilitybewhenthesethingsareabsentandnakedhumannatureisallowedtoexpressitselfwithoutinterference?WhatwouldLifebeLikewithoutLaw?Hobbes’sanswer(Hobbes1640:PtI,ch.14;1642:ch.1;1651:ch.13)isthat,lefttodowhatcomesnaturally,humanbeingswouldquicklyfindthemselvesinastateofwar.Theywouldhavetoseetotheirownsurvivalandhappinessguidednotbyagovernment’ssenseofwhatwasconducivetosurvivalandhappiness,butonlybytheirconflicting,fluctuatingandirremediablydiverseprivatejudgementsaboutwhatwasbest.Notbeingrequiredtodefertoanyoneelse’sjudgements,butnotbeingable,either,torelyonotherstoseekthegoodco-operativelyorbenevolently,peoplewouldhaverationallytoexpecttheworstfromoneanotherand,iftheythoughtthattheirsafetydemandedit,takepre-emptiveaction.Thus,eventhosewhowerenotnaturallygreedyorviolentmightbegintoseetheirownsurvivalandwell-beingintakingasmuchofwhattheywantedastheycould,orinkillinganyonewhomightbeapotentialenemy–whichcouldmeananyone.Asforthosewhowerenaturallyviolentandgreedy,andwhowerenormallyabletogetawaywithit,theywouldfightandappropriategoodsuntilstoppedbypeoplestrongerthanthemselves.Whoeverwasapotentialvictimwouldhaveareasontotakeactionagainstthemfirst.Inthiswaylifewithnoholdsbarredwouldquicklydegenerateintofearfulinsecurityandapermanentstruggleforsurvival.Or,asHobbesputsit,lifeinthestateofnatureiswar–theanti-socialconditionparexcellence.Inrelationtotheseedsofwarinhumannature,theargumentofDeCiveisnotverydifferentfromthatoftheElementsofLaworLeviathan.Hobbesmaintainsinallthreetreatisesthatlifewouldbeextremelyunpleasantiflawandcoercivegovernmentweretodisappear;healsoholdsthatanygovernmentthatdoesexistisconstantlyindangerofbeingsubverted,thatis,forcedbyrebelsintoastateofcollapsethatre-introducesthewarofallagainstall.Theseedsofthiswararenotonlythedangerouspassionsofindividualpeople,whocanbearousedbytheircircumstancesortheircorporealcon-stitutions.Theseedsalsoincludethelackofscienceandcapacityforreasoning,foritisonlyintheabsenceofthesethingsthatthePASSIONS(pp.716–17)succeedindeter-miningindividuals’behaviour.Again,Hobbesbelievedthatpeoplewereeasilyswayedbycleverpoliticiansandputinfearofhellbyamanipulatingclergy.Theywerethuspronetoresenttherestrictionsoflifeundergovernment,ortoimaginethatobediencewouldcostthemsalvation.Itwasashortstepfromtheseresentmentsandfearstocivilwar.673\nTOMSORELL4TheLawsofNatureandtheRationalefortheStateHobbes’srulesforavoidingcivilwarifacommonwealthexists,andforendingwarinthepre-politicalstateofnature,arecalledlawsofnature.Thoughstatementsofother‘lawsofnature’existedlongbeforetheElementsofLaw,DeCiveandLeviathan,Hobbes’sownlist(Hobbes1640:PtI,chs16–17;1642:chs2–3;1651:chs14–15)wasunusual,andheclaimedhewasthefirsttofindanuncontroversialbasisforlawsofnature.Whatismore,thebasiswouldmakeitirrationalforanyonetoignorethem.Itispartlybecausehewasabletoidentifyasingle,uncontroversialgoodthatthelawsofnatureallpromotedthatHobbesbelievedhisstatementofthelawsofnatureamountedtoamoralscience.Thebasicgoodthatheidentifiedwasthegoodofpeaceortheabsenceofwar;or,ashesometimesputit,thegoodofsafety.Becausewarthreatenedthepossessionofallgoodthings,notleastlifeitself,anyactionthathelpedtoavoidwarortoputanendtoit,andthatdidnotitselfthreatentoomanygoods,wasarationallycompulsorythingtodo.ThereweretwolawsofnaturethatHobbesputbeforetherest:firstly,toseekpeacebyallmeansiftodosowasnottoounsafe;and,secondly,toadoptoneparticularmeansofseekingpeaceifotherpeopleagreedtodothesame:namely,tolaydownrightsthatgetinthewayofpeace.Inthestateofnature,therightthatpeoplehavetogiveupinordertosecurepeaceisthe‘rightofnature’–therighttobeone’sownjudgeofwhatwillmakeonesafeandhappy.Solongaseachishisorherownjudgeofthesethings,thereisnoinsuranceagainstwar.Therightmustthereforebegivenup.Hobbeswasnotthinkingofasimplerenunciationofthisright:hehadinmindatransferofit(Hobbes1640:Pt1,ch.15,iii;1642:ch.2,iv;1651:ch.14)bymostinthestateofnaturetoafewortooneinthestateofnature,whowouldthendecideforthemanywhatwouldmakeeachassafeandhappyaspossible.TheoneorthefewwouldbecomeaSOVEREIGNPOWER(pp.261–4),andthemanywouldbecomesubjectsofthatpower,andbeobligedtoliveinkeepingwiththesovereign’sjudgementsaboutsafetyandwell-being,asexpressedinthesovereign’slaws.Thetransferoftherightofnaturetoasovereignpoweriswhatcreatesastateorcommonwealth.Oncetransferred,therightofnaturecannotbetakenback,exceptforthesakeofapersonalsafetywhichaweakorincompetentsovereignisnolongerabletosecure.Inotherwords,onegetsbackone’srightofnatureifobediencetothesov-ereignisnotafterallameansofavoidingwar(Hobbes1640:Pt2,ch.1,v;1642:ch.6,iii;1651:ch.17).Ontheotherhand,ifacommonwealthiskeepingonesafe,oneisobligedtoobeythesovereign’slaws,eitherbytheactthattransferredone’srightofnature,andbythedutytoabidebyone’sagreements;orbysomeactofsubmissionifonedidnotenterthecommonwealthfromthestateofnature.Drawingbackfromthedetails,onecanseethatHobbeshasbothasecularconceptionofthestate,andonethattriestoshowthattheexistenceofthestateisreasonable,notarbitrary.Thereisnodivinerightofthesovereigntorulethatobligesonetoobey,noryetisitamatterofthesovereign’ssimplybeingmorepowerfulthanoneisoneself.Theobe-diencespringsfromatransferofrightforthesakeofpeace.Thestateisthusseenasasecurityorpeace-keepingdevicewhoseexistenceisintheinterestofthemany,ratherthanasafacsimileonearthofadivineorderingofthingsbynobilityorrankorstation.674\nHOBBESLeavingtheStateofNatureasaPrisoner’sDilemmaHowcanitberationalforanyoneinthestateofnaturetolaydownhisrighttoprotecthimself,ifbydoingsohelayshimselfopentotheattackofsomeonewhodecidesnottodoso,orwhopretendstodoso?Doesn’titmakemoresensetostayinthestateofnaturewhenothersdecidetoleaveit?Hobbesconsidersthequestionintheformofscepticismabouttheneedtokeepone’sagreements.Heanswersthatinthelongrunitpaystokeepone’sagreements.Ifonedoesn’t,onewillbemistrustedbyothers,whowon’tenteranyfurtheragreementswithyou,orwhomaydecidetogettheirrevengeforyourtakingadvantage.Anyshort-termadvantageiscancelledoutbythesubsequentinsecurity.FollowingGauthier(1969),onemayrepresentthepracticalchoicesfacingtwowould-bepartiestothesocialcontractasaprisoner’sdilemma(seechapter12):BKeepsthecontractBreaksthecontractKeepsthecontract2,24,1ABreaksthecontract1,43,3ThismatrixdepictstherankingsplacedbyAandBondifferentoutcomesintheshortterm.AandBeachlikemost(givetheranking1to)thecaseinwhichtheythemselvesbreakthecontractbuttheiroppositenumberkeepsit.Thencomesmutualadherence,followedbymutualviolationofthecontract.Overthelongtermthingschange:BKeepsthecontractBreaksthecontractKeepsthecontract1,14,2ABreaksthecontract2,43,35TheObligationsofSubjectsandtheRightsofSovereignsThesovereign’ssubjectsareobligedbyasortofagreementamongthemselvestoobeythesovereign’slaws;butthesovereign,thoughheisabeneficiaryoftheagreement,isnotpartytoit.Themanymakeagifttoapersonorbodyofpeopleoftheirsubmission;butthesubmissiondoesnotobligethesovereigntopleasethemanyinreturn.Thesov-ereignwouldperhapsbeimprudentifhedidnotdohisbesttoarrangeforthepeaceandwell-beingofhissubjects;forifheallowsthestatetodescendintolawlessness,hissubjectstakebacktheirrightofnature,actoncemoreaslawsuntothemselvesandregardnooneastheirsovereign.Insuchacaseaformersovereign,beingonlysolitaryorasmallgroup,isnomoresecurethananyoneelse.Soitisinthesovereign’sinterest675\nTOMSORELLnottodisappointhissubjects.Butheisnotobligedtodoso,andifhislawsonlybringaboutapeaceinwhicheveryoneissafebutlivingausterely,thatdoesnotamounttoaninjusticeagainsthissubjects,eveniftheyarehugelydissatisfiedwiththeirlot.Onthecontrary,thesovereignisleftfreebyhissubjectstoactasheseesfittosecurepeaceandwell-being;itisthesubjectswhoactunjustlyiftheytrytoreinterprettheagreementasoneinwhichtheydecidehowthepeaceandwell-beingaretobegot,andofwhattheywillconsist.Thesovereign,then,hasveryextensivefreedomsorrights(Hobbes1640:Pt2,ch.1;1642:ch.6;1651:ch.18).Hedecideswhatislegalandillegal,whoownswhat,whoistoholdtheofficesofstate,whoistobeallowedtopreachandmakepoliticalspeeches,howpeoplearetobeallowedtotradeandassociate.Hehaspowerstolegislateandtotaxandofmakingwar.Nothinglessthantheserightswouldenablethesovereigntomakeandkeepthepeace.Butitisonethingforasovereigntohavetheseextensiverights;itisanotherforhimtoexercisethem;andHobbeslistsasoneoftheprincipalcausesofcivilwarthefactthatsovereignsfailtoexercisethem.Thisfailure,justasmuchasthefailureofsubjectstoacknowledgetherightsofsovereignty,candisturbthepeace.ItisnothardtoworkouttheconsequencesofHobbes’stheoryinrelationtothequestionofwhetherCharlesIwaswithinhisrightstoforcealoanfromhisrichsubjects,tobilletsoldiersandtocircumventparliamentintheperiodbeforetheCivilWar.Hewaswithinhisrightsinalloftheseareas;andthatcivilwarensuedwasasmuchtraceabletoamistakenunderstandingoftherightsofsubjects,astoineptitudeintheexerciseofsovereignrights.WehavecomefarenoughtoseethatbothreadingsofHobbes’spoliticsfitafterall.Therecanbenodoubtthat,inhisthreemajorworksonthecommonwealth,Hobbesframeshisaccountoftherightsofsovereigns(andofthecausesofcivilwar)withtheeventsofEnglandinthe1630sand1640sinhismind.ButitisalsotruethatthetheorytakesupabstractquestionsthattranscendhishistoricalperiodandwhichgobacktoPlatoandAristotle.Hobbeshasananswertothequestionofwhatthestateisfor,ofwhathumanhappinesscanconsistinandwhatgoodsarepromotedbymorality;heisnotonlyoperatingthroughhiswritingsasapolicyadvisertoacertainEnglishkingorkingsandtheirentourage.6StrengthsandWeaknessesofHobbes’sPoliticsHowsuccessfulisHobbes’stheoryofsubjectionandsovereigntyasapieceofpoliticalphilosophy?Oneofitsmostdistinctiveideasisthatpeaceratherthanhappinessiswhattheinstitutionofthestateisfor.Hobbesthoughtthattheconceptofpeacewasclearerthantheconceptofhappiness,andalsothat,whilepeoplehaddifferent,andsometimesirreconcilable,conceptionsofhappiness,peoplecouldreachagreementaboutwhatconstitutedpeace.Healsobelievedthatpeoplecouldnotrationallyrepudiatethedemandsofpeace.Againstthisbackground,atheoryofthestateandofmoralitygearedtopeacewassupposedtomakeagreatadvanceonprecedingtheories.Unfor-tunately,allofthesupposedadvantagesoftheconceptofpeacecanbequestioned.Firstly,theconceptofpeaceisnotveryclearinHobbes’swritings.Itisdefinedasany676\nHOBBEStimewhenthereisnotwar.ButbecauseHobbesthinksthatwarisnotonlyopenfight-ingbutalsoincludesthevaguephenomenonwenowcall‘coldwar’,theconceptoftheabsenceofwarisboundtohavethevaguenessthattheconceptofwaritselfhas.Secondly,Hobbessometimesseemstoidentifythechiefaimofthestatewiththatofsecuringpeaceandsometimeswiththatofachievingpublicsafety(Hobbes1651:ch.30).Thepublicsafety,however,hedefinesinbroadandinnarrowsenses.Thenarrowsenseofsafetyisprotectionfromphysicalattack;thebroadsenseissafetyinthenarrowsenseplusasmuchwell-beingasiscompatiblewithsafetyinthenarrowsense.Plainly,iftheadvisabilityofobediencetothestateistobejudgedbythestate’ssuccessinmakingpeoplesafeinthebroadsense,then,sincesafetyinthebroadsenseisopentomanyinterpretations,thereisperhapstoomuchscopefortheconclusionthatthestateisnotfulfillingitsfunctionsandthatobedienceisnotdue.Theexpectationofsafetyfromthestatestartstobecomeasdestabilizingastheexpectationofhappiness,whichisnotsurprising,sincesafetyinthebroadsenseoverlapswiththeconceptofhappiness.Hobbes’sproblemsdonotdisappeariftheconceptofsafetyisnarroweddowntotheconceptofsafetyfromphysicalattack,forindividualsinthestateneverlosetheirrighttojudgewhentheyareunderthreatofphysicalattack,andtheyneverlosetheirrighttodowhattheyliketoprotectthemselvesiftheyreallydothinktheyareundersuchthreat.Stuckinthemiddleofariot,forexample,theymaytakeuparmsandkilliftheythinkthatishowtosurvive.Buteveninwhatisoutwardlyapeacefulstatesomeonemaysincerelybelievethatanundergroundparamilitarymovementissecretlyworkingitswaytowardsasuccessfultakeoverofthestateandafuturepolicyofgenocide;itisunclearthatthisbelief,evenifparanoid,wouldnotfreeanindividualfromthetiesofthesocialcontract.BecauseHobbesinsistsontheinalienabilityoftherighttobeone’sowninterpreterofpresentdangers,hemayleavetoomuchroomforpretextsfortakingthelawintoone’sownhands:preciselywhatheprideshimselfonhavingexcluded.7TheRestofHobbes’sPhilosophyHobbes’spoliticsisnotunrelatedtotherestofhisphilosophy–tohismaterialisticmetaphysics,forinstance,orhistheoriesofoptics.Onthecontrary,Hobbesthoughtthathismetaphysics,hispolitics,hisoptics,aswellasotherphilosophers’findingsinotherbranchesoflearning,couldbearrangedintooneorderlybodyofscienceingeneral.Scienceingeneralwassupposedtobeginwith‘firstphilosophy’.ThiswasnotaCartesianfirstphilosophy,noranAristotelianone.HobbeswasacriticofDESCARTES(chapter26),onwhoseMeditationshewasoneofthefirsttocomment.Hedeniedthattherewereirreduciblyimmaterialsubstances,eitherfiniteorinfinite;therewerenoimmaterialsoulsinparticular,andnothingintheworldlikeaCartesianrescogitans.InhisrepliestoBishopBramhall(1594–1663)inacontroversyoverfreewill,HobbessaidthatevenGOD(chapter15)wasacorporealspirit;andheinsistedthatMENTALACTIVITY(chapter5)wasmotionintheheadandinthenervousandcirculatorysystems.AnotherkindofCartesianimmaterialismthatHobbesdeniedwasthebeliefinabstractobjects,suchasthenatureofatriangle.Descarteshadinsistedonthereality677\nTOMSORELLofsuchthingsinhisfifthMeditation;Hobbes,forhispart,thoughtthateverythingimportantabouttrianglescouldbeunderstoodasflowingfromlinguisticconventionsgoverningtheword‘triangle’,ratherthanfromtriangularity.Hobbes’spositivefirstphilosophywasasetofdefinitionsenablingdifferentkindsofmotioninmattertobedistinguished.Thedifferentbranchesofthesciencesweredis-tinguishedbythedifferentkindsofmotiontheystudied.Geometrystudiedmotionintheabstractwithmanyspecificpropertiesofmovingbodiesleftoutofaccount;physicsincludedthemotionsofpartsofbodiesandtheinteractionbetweeninanimatebodiesandthesenses.Inoptics,Hobbesadvocatedthetheorythatweseethingsasaresultofthedisplacementofamediumandtheagitationofthesenseorgansbythemotionofaluminousobject,liketheSun.Whenitcametopoliticsandethics,Hobbes’sclaimthatallthesciencesaresciencesofmotionishardtounderstand.Thelawsofnatureandtheargumentsforcertainkindsofdistributionsofrights,whicharecentraltothetheoryofthecommonwealth,arenotmechanistic,thoughtheyareconnectedtoamechanisticpsychology.Hobbes’smechanisticgeometryalsopromptedcriticisminitsdayfrommathematiciansmuchmoreablethanHobbeshimself.Thethreadrunningthroughallofhisowncontributionstothesciencesisastrongantagonismtotheoldlearningofthescholastics,andespeciallytheheroofthescholas-tics,Aristotle.HobbesdevelopedhistheoryofvisioninoppositiontotheAristoteliantheorythatweseethingsbyaprocessoftransferofformsorspeciesfromtheseenobjecttotheinterveningmediumtotheeye.Hedevelopedhispolitics,aswehaveseenearlier,partlyinoppositiontotheAristoteliantheorythathumanbeingsareanimalsthataresuitedtothepolity.AndhearrivedathismaterialismpartlyinoppositiontoAristotle’stheorythatthemainbranchesofscienceallhavediscreteanddistinctivesubjectmatterswhichexistintheirownways.ForHobbes,toexistwastobematerial,andtheexplanationofeverythinginnaturewastobereferredtodifferenttypesofmotioninmatter.Onthesurface,HobbesfollowedAristotleindividingupsciencesaccordingastheirsubjectmatterwasnaturalorartificial;butevenherehebrokethemould,forneitherofthescienceshereckonedtobeconcernedwithartefacts,namelygeometryandpolitics,wascountedasartificialbyAristotle,andtobeartificialforHobbeswasstilltobematerialandtheresultofmotion.Tobesure,themotionthatgivesrisetoartefactsismotioninitiatedbythehumanwill,butthewillitselfismechanicallyexplicable;sotherewasnofundamentaldistinctionforHobbesbetweenthesubjectmattersofthenaturalsciencesandthesciencesofartefacts:bothweresciencesofmotionormechanisticsciences.Inthecaseofpolitics,thesubjectmatterwassomethingmadebyandoutofhumanbeings,namelyacommonwealth.Thepurposeofthesciencewasnot,however,primarilyexplanatory,butproductive:Hobbesclaimedhewouldteachhowtoconstructalastingcommonwealth,muchasEuclidhadtaughthowtoconstructcircles.PreceptsdominateHobbes’spolitics;statementsaboutthepossiblecausesofobservedeffectsdominatehiscontributionstothenaturalsciences.SoaroughwayforHobbestodistinguishbetweenthepoliticalormoralsciencesandthenaturalsciencesisbywayofthedifferencebetweenoughtandis;or,moreprecisely,betweenstatementsofwhatoneoughttodo,backedbyreasons,andconjecturesofcausesofobservedeffects,backedbyreasons.678\nHOBBESFurtherReadingThoseworkingonHobbesshouldbeginbyreadingLeviathan(1651).Peters(1956)offersaclearlywrittenstudyofthewholeofHobbes’sphilosophy,whileSorell(1986)providesastudyofHobbes’smetaphysicsandpoliticsagainstthebackgroundofhisphilosophyofscience.Gauthier(1978),Oakeshott(1975),Macpherson(1962)andWarrender(1957)areclassicappraisalsofHobbes’spoliticalthought.Mintz(1970)dealswiththecontextandearlyreceptionofLeviathan,andSommerville(1992)isaveryclearandcomprehensiveguidetothehistoricalcontextofHobbes’spoliticalwritings.Johnston(1986)exploresthetensionsbetweenthescientificpretensionsofsomeofHobbes’spoliticalwritingsandtheirpersuasivepurpose.AshortstudybyTuck(1989)arguesfortheimportanceofphilosophicalscepticisminHobbes’sthought.MostofthepapersinRogersandRyan(1988)areaccessible,andthereiscoverageofHobbes’smetaphysics,philosophyofscienceandpolitics.Sorell(1996)containsanauthoritativebiographicalaccountandarticlesonallaspectsofHobbes’sthought,includinghismathemat-icsandopticsandhistheoriesoflawandreligion,aswellashismetaphysics,politicsandphilosophyofscience.ReferencesHobbesThestandardeditionofHobbes’sworksisdividedbetweenhisEnglishandLatinwritings:Hobbes,T.1966a:TheEnglishWorksofThomasHobbes,11vols,2ndreprint(editedbySirW.Molesworth).Aalen:ScientiaVerlag.——1966b:OperaLatina,5vols,2ndreprint(editedbySirW.Molesworth).Aalen:ScientiaVerlag.Individualworks,unlesscitedotherwise,aretobefoundinthecollectedworks.Leviathanisavailableinseveralrecenteditions.Hobbes,T.1641:ThirdSetofObjectionstoDescartes’sMeditations.InE.HaldaneandG.Ross(eds)1978:ThePhilosophicalWorksofDescartes.Cambridge:CambridgeUniversityPress.——1642:DeCive,shorttitleforElementorumPhilosophiae,SectionTertia,DeCive(translatedbyHobbesintoEnglish(1651)asPhilosophicalRudimentsConcerningGovernmentandSociety).——1646:AMinuteorFirstDraughtonOptics.——1651:Leviathan,ortheMatter,FormandPowerofCommonwealth,EcclesiasticalandCivil.——1654:OfLibertyandNecessity.——1656a:DeCorpore,shorttitleforElementorumPhilosophiae,SectoPrimadeCorpore,(1655),translatedintoEnglishasElementsofPhilosophy,thefirstsectionconcerningBody,1656.——1656b:TheQuestionsConcerningLiberty,NecessityandChance,ClearlyStatedandDebatedbetweenDr.Bramhall,BishopofDerry,andThomasHobbesofMalmsbury.——1656c:SixLessonstotheProfessorsofMathematics...intheUniversityofOxford.——1658:DeHomine,siveElementorumPhilosophiaeSectioSecunda(editedbyB.Gert,translatedbyC.T.Wood,T.Scott-CraigandB.Gert,1962,asOnManinManandCitizen.NewYork:HumanitiesPress).——1662a:SevenPhilosophicalProblems.——1622b:ConsiderationsoftheReputation,Loyalty,Manners,andReligion,ofThomasHobbesofMalmsbury.——c.1666:DialogueBetweenaPhilosopherandaStudentoftheCommonLawsofEngland.——c.1668:Behemoth,ortheLongParliament.DialogueoftheCivilWarsofEngland.——1678:DecameronPhysiologicumorTenDialoguesofNaturalPhilosophy.679\nTOMSORELL——1889[1640]:TheElementsofLawNaturalandPolitic(editedbyF.Tonnies).London:SimkinandMarshall.——1976[1641or1642]:ThomasWhite’s‘DeMundo’Examined(translatedbyH.W.Jones).Bradford:BradfordUniversityPress.OtherwritersBoonin-Vail,D.1994:ThomasHobbesandtheScienceofMoralVirtue.Cambridge:CambridgeUniversityPress.Brandt,F.1928:Hobbes’sMechanicalConceptionofNature.London:Hachette.Brown,S.M.1959:Hobbes:TheTaylorThesis.PhilosophicalReview,68.Dewey,J.1974:TheMotivationofHobbes’sPoliticalPhilosophy.InR.Ross,H.SchneiderandT.Waldman(eds)ThomasHobbesinHisTimes.Minneapolis:UniversityofMinnesotaPress.Gauthier,D.1969:TheLogicofLeviathan:TheMoralandPoliticalTheoryofThomasHobbes.Oxford:Blackwell.——1978:TheLogicofLeviathan.Oxford:ClarendonPress.Goldsmith,M.1966:Hobbes’sScienceofPolitics.NewYork:ColumbiaUniversityPress.Johnston,D.1986:TheRhetoricofLeviathan.Princeton,NJ:PrincetonUniversityPress.Levin,M.1982:AHobbesianMinimalState.PhilosophyandPublicAffairs,11.Lloyd,S.A.1994:IdealsasInterestsinHobbes’sLeviathan.Cambridge:CambridgeUniversityPress.McNeilly,F.S.1968:TheAnatomyofLeviathan.London:Macmillan.Macpherson,C.B.1962:ThePoliticalTheoryofPossessiveIndividualism.Oxford:ClarendonPress.Mintz,S.1970:TheHuntingofLeviathan.Cambridge:CambridgeUniversityPress.Nagel,T.1959:Hobbes’sConceptofObligation.PhilosophicalReview,68.Oakeshott,M.1975:HobbesonCivilAssociation.Oxford:Blackwell.Peters,R.1956:Hobbes.Harmondsworth:PenguinBooks.Raphael,D.D.1977:Hobbes:MoralsandPolitics.London:GeorgeAllenandUnwin.Robertson,G.C.1886:Hobbes.Edinburgh:W.Blackwood.Rogers,G.A.J.andRyan,A.(eds)1988:PerspectivesonThomasHobbes.Oxford:ClarendonPress.Sacksteder,W.1982:HobbesStudies(1879–1979).BowlingGreen:PhilosophyDocumentationCenter.Shapin,S.andSchaffer,S.1985:LeviathanandtheAirPump.Princeton,NJ:PrincetonUniversityPress.Skinner,Q.1966:TheIdeologicalContextofHobbes’sPoliticalThought.TheHistoricalJournal,9.——1996:ReasonandRhetoricinthePhilosophyofHobbes.Cambridge:CambridgeUniversityPress.Sommerville,J.1992:ThomasHobbes:PoliticalIdeasinHistoricalContext.London:Macmillan.Sorell,T.1986:Hobbes.London:Routledge.——(ed.)1996:TheCambridgeCompaniontoHobbes.Cambridge:CambridgeUniversityPress.Taylor,A.E.1938:TheEthicalDoctrineofHobbes.Philosophy,13.Thomas,K.1965:TheSocialOriginsofHobbes’sPoliticalThought.InK.C.Brown(ed.)HobbesStudies.Oxford:Blackwell.Tuck,R.1989:Hobbes.Oxford:OxfordUniversityPress.Warrender,H.1957:ThePoliticalPhilosophyofHobbes.Oxford:ClarendonPress.Zarka,Y.1987:LaDecisionmetaphysiquedeHobbes–Conditionsdelapolitique.Paris:Vrin.DiscussionQuestions1CanweunderstandHobbes’scontributionstoabstractpoliticaltheorywithoutgraspinghisroleinthepoliticalcontroversiesofhistime?680\nHOBBES2WhydidHobbeswishtodealscientificallywiththequestionsofpoliticalphiloso-phy?Didhesucceed?3WhatfollowsfromHobbes’sdenialthathumanbeingsaremadeforpoliticallife?4Howcanwedecidewhetherpeoplearepeaceableorbynaturepronetoconflictandviolence?5Islifeinthestateofnaturewar?6AssessHobbes’smoralpsychologyinastateofnatureandunderacoercivegovernment.7Howdoestheprisoner’sdilemmasupportHobbes’sclaimthatitisrationalforustoleavethestateofnature?8DidHobbeshavegoodreasonsforassertingthatevenstronggovernmentsarevulnerabletorebellionandcollapse?9Doscienceandacapacityforreasoningpreventthepassionsfromdetermininganindividual’sbehaviour?10DidHobbesfindinpeaceasingle,uncontroversialgoodonwhichtobaselawsofnature?11Inrelinquishingrightsinordertosecurepeace,doHobbeseanactorsgiveuptoomuch?12CanweacceptHobbes’saccountofhowoneorafewmightbecomeasovereignpower?Howelsemightasovereignpowerbeestablished?13Whatisastateorcommonwealth?Wouldatransferofrightsfromindividualsbesufficienttoestablishastate?14Aretherecircumstancesotherthanrenewedhazardsofwarwhichwouldjustifyrejectingthesovereign’slaws?15Canweacceptthatthesovereignhasnoobligationtomaintainthepeace,butdoessoonlyasamatterofprudence?16Wouldsubjectshaveareasonablecomplaintagainstasovereignwhoselawsleftthemsafe,buthugelydissatisfiedwiththeirlot?17Dosubjectsactunjustlyiftheytrytodecidewhatconstitutespeaceandwell-beingandhowtheyaretobeobtained?18WasHobbesrighttoseedangerinasovereign’sfailuretoexerciserights?19Doesitmakebettersensetosaythattheinstitutionofthestateisforpeaceorforhappiness?20CanHobbes’stheorybealteredtogiveagreaterroletowell-being?21Istheexpectationofsafetyfromthestateasdestabilizingastheexpectationofhappiness?22DoesHobbes,contrarytohisintention,leavetoomuchroomfortakingthelawintoone’sownhands?23Canweavoidanycommitmenttoimmaterialthings?24Caneverythingimportantabouttrianglesbeunderstoodascomingfromlinguisticconventionsabouttheword‘triangle’?25DoesHobbessucceedinestablishingamaterialisttheoryofthemind?Considerespeciallyhisaccountofthewill.681\n29LockeR.S.WOOLHOUSEJohnLockeisoneofthemostimportantEuropeanphilosophersofthe‘earlymodern’orimmediatepost-Cartesianperiod.BorninSomerset,England,in1632andeducatedinOxford,hediedinEssexin1704.BesidesbeinginvolvedinProtestantpoliticsandpublicaffairsthroughhisemployer,LordAshley,laterEarlofShaftesbury,hefoundtimetowriteonawidevarietyofbroadlyphilosophicaltopics:religion,education,econ-omics,politicalphilosophy.Heisknownprimarilyforhisepistemologicalmasterpiece,AnEssayConcerningHumanUnderstanding(published1690),butalsoforhisTwoTreatisesofGovernment(published1690)andTheReasonablenessofChristianity(published1695).ReaderswillfindithelpfultolookatchaptersonEPISTEMOLOGY(chapter1),METAPHYSICS(chapter2),POLITICALANDSOCIALPHILOSOPHY(chapter8),PHILOSOPHYOFRELIGION(chapter15),DESCARTESANDMALEBRANCHE(chapter26),SPINOZAANDLEIBNIZ(chapter27),HOBBES(chapter28),BERKELEY(chapter30)andHUME(chapter31).1TheEssayConcerningHumanUnderstandingLocke’sEssaywasbegunsometwentyyearsbeforeitwaspublished,anditwentthroughvariousdrafts.Itembodiesanddrawstogethermanyofthethought-currentsthatwerecommonplaceinthelatterhalfoftheseventeethcentury:negativeones,suchasanti-Aristotelianism,andpositiveones,suchasthatofthemechanicalphilosophyandthenewexperimentalphilosophy.Ahintofthesethemesisgiveninthe‘EpistletotheReader’oftheEssay,whereLockecontraststheimpedimentstoknowledgethrownupbySCHOLASTICPHILOSOPHY(chapter24),withthelastingmonumentsproducedbycontributorssuchasRobertBoyle(1627–91)andIsaacNewton(1642–1727),to‘thecommonwealthoflearning’.ThechemistRobertBoylewasafoundermemberoftheRoyalSocietyofLondonfortheAdvancementofExperimentalKnowledge,andthemathematicianandtheoreticalphysicistIsaacNewtonwasoneofitsfirstpresidents.LockewaselectedaFellowtotheSocietyin1668,notlongbeforehebeganworkontheEssay.Besidesthusbeingaproductofitscentury,theEssayhadagreatinfluenceontheEnlightenmentofthenextcentury,whoseRATIONALISM(chapters26and27)wearsthe\nLOCKEcoloursoftheindividualismwhichpervadesbothLocke’sexpressionandhisexplicitdoctrine.RunningthroughtheEssayisaninsistencethatopinionsarecarefullytobeweighed,impartiallyconsideredandjudgedontheirownmerits,byindividualsforthemselvesindependentlyofwhatotherssay,particularlythoseinthemajorityorinauthority.ThoughthereareMETAPHYSICALANDONTOLOGICAL(chapter2)elementsintheEssay,itsmainconcernistheEPISTEMOLOGICAL(chapter1)oneoftheextenttowhichthehumanmindhasthecapacitytoacquireknowledge.‘Mypurpose’,Lockesays,is‘toenquireintotheoriginal,certainty,andextentofhumanknowledge;togetherwiththegroundsanddegreesofbelief,opinionandassent’(Locke1975:I,I,ii).Itisplain,more-over,thatoneofLocke’sconcernsinsettingout‘totakeasurveyofourownunder-standing,[to]examineourpowers,andtoseetowhatthingstheywereadapted’(ibid.:I,I,vii)wastomakearesponsetoscepticismabouttheverypossibilityofknowledge.Hespeaksofthescepticaldespair,andthetendencytodisclaimallknowledge,whichcanarisewhenpeoplerealizethat‘somethingsarenottobeunderstood’(ibid.:I,I,vi):‘Men,extendingtheirenquiriesbeyondtheircapacities...’tisnowonderthattheyraisequestions...whichnevercomingtoanyclearresolution,areproperonlyto...confirmthematlastinperfectscepticism’(ibid.:I,I,vii).Hisinterestinthismatterwasnotpurelytheoretical.Hehadapracticalinterestinthehumanpredicamentingeneral,inthequestionofwhatkindofcreatureGOD(chapter15)hasseenfittomakeusandinthequestionofhow,asaconsequence,weshouldarrangeandmanageourlivesandthoughts.Hewished,therefore,‘tofindoutthosemea-sures,wherebyarationalcreatureputinthatstate,whichmanisin,inthisworld,may,andoughttogovernhisopinions,actionsdependingthereon’(ibid.:I,I,vi).Locke’sgeneralposition,asoutlinedattheveryoutsetofhisdetailedinvestigation,isthatthereissometruthinscepticism:ourknowledgedoesindeedhaveitslimits.Therearethingsofwhichwearecompletelyignorant,andtherearethingsaboutwhichwecanmerelyopineandbelieve.Butthisisnocausefordespairfor,soLockebelieved,ourcapacitiesforknowledgearesuitedtoourneeds.‘Howshortsoever...[people’s]knowledgemaycomeofanuniversalorperfectcomprehensionofwhatso-everis,ityetsecurestheirgreatconcernments’(ibid.:I,I,v).Yetwhatareourneedsandconcernments?InLocke’sviewweareputherebyGodinthisworld,withsomehopeofalifehereafterinanother.Ourneedsaremet,there-fore,ifweknowenoughfortheeverydaypracticalitiesoflifeinthisworldand,intermsofourdutiesandobligationstoeachotherandtoGod,foroursalvationinthenext.And,inLocke’sview,theseneedsaremet,or,withsomereasonableeffortonourpart,canbemet.Peopleshouldbe‘wellsatisfiedwithwhatGodhaththoughtfitforthem,sincehehath...putwithinthereachoftheirdiscoverythecomfortableprovisionforthislifeandthewaythatleadstoabetter’(ibid.:I,I,v).ThisisnottosaythatGodhasactuallygivenuswhatknowledgeisnecessaryanduseful;thewholeofBookIoftheEssayisdevotedtorejectingroundlyanysuggestionthatanyofourknowledgeorideasisinnate,andpossessedfrombirth.ItisratherthatGod,byequippinguswithanunderstanding,hasenabledustoacquirewhatisneces-sary;and,withoutstrivingagainstorcomplainingaboutitslimits,weshouldmakeuseofthatability.Idonot,hesays,‘disesteem,ordissuadethestudyofnature’,butonly‘thatweshouldnotbetooforwardlypossessedwiththeopinion,orexpectationofknowledge683\nR.S.WOOLHOUSEwhereitisnottobehad’(ibid.:IV,XII,xii).Thisability,moreover,hasbeengiventoeachoneofus,andeachoneofusmustautonomouslymakehisownindividualuseofit.Lockeispassionatelydismissiveofthosewho‘takingthingsupontrust,misimploytheirpowerofassent,bylazilyenslavingtheirmindstothedictatesanddominionofothers,indoctrineswhichitistheirdutycarefullytoexamine’(ibid.:I,IV,xxii).Havingrejectedanysuggestionofaninnateendowmentofknowledge,Lockesetsouttoshowhowits‘fountain’isexperience.Itisinexperiencethat‘allourknowledgeisfounded;andfromwhichitultimatelyderivesitself’(ibid.:II,I,ii).Onaccountofstatementslikethis,Lockehasbeencharacterizedasanempiricist.(Indeed,hisplaceinthehistoryofphilosophyhasoftenbeenseenasthefounderoftheso-calledschoolofBritishEmpiricism,aschoolofwhichBERKELEY(chapter30),HUME(chapter31)and,later,MILL(chapter35),aresupposedtobemembers.)Butitisimportanttoseethathedoesnotholdthatknowledgeisimmediately‘madeouttousbyoursenses’(Locke1990:75)bythemselves.Whatisimmediatelyderivedfromexperienceare‘ideas’,andideasinthemselvesarenotknowledge.Theyare,rather,‘thematerialsofknowledge’(Locke1975:II,I,ii).BesideshavingSENSES(pp.000–00)whichgivethemideasfromexperience,humanshavereasonandunderstandingtoo,anditisbytheirmeansthatknowledgeisultimately,asLockecarefullysays,derivedfromideas:‘reason...byarighttraceingofthoseideaswhichithathreceivedfromsenseorsensationmaycometo...knowledge...whichoursensescouldneverhavediscovered’(Locke1990:75).ThefrequentclassificationofLockeasanempiricistmustnotbeallowedtohidethatheclearlyholdsthatreasontooisessentiallyinvolvedintheultimatederivationofknowledge.WhattheclassificationasanempiricistdoescorrectlycaptureisLocke’sviewthat,priortoexperience,thehumanmindis‘whitepaper,voidofallcharacters,withoutanyideas’(Locke1975:II,I,ii),andthat‘allthosesublimethoughtswhichtowerabovetheclouds,andreachashighasheavenitself,taketheirriseandfooting’(ibid.:II,I,xxiv)inexperience.Buthowisourknowledgederivedfromthisexperientialmaterialandhowdoesourreasoncometoknowledgeonitsbasis?Locketakesknowledgetobe‘theperceptionoftheconnexionandagreement,ordis-agreementandrepugnancyofanyofourideas’(ibid.:IV,I,ii).Histhoughtisthatsomeideashavenecessaryconnectionswithothers.Knowledgeofthetruthsabouttheseconnectionsconsistsinthe‘perception’,therecognitionbyourunderstanding,oftheconnections.So,ourknowledgethatatriangle’sthreeanglesareequaltotworightanglesconsistsinourintellectual‘perception’oftheconnectionbetweentheideaofthisequalityandtheideaofthetriangle’sthreeangles.Weintellectually‘see’that‘thatequalitytotworightones,doesnecessarilyagreeto,andisinseparablefromthethreeanglesofatriangle’(ibid.:IV,I,ii).Suchconnectionsaresometimesdirectandimmediate,sometimesindirectandmediatedbyconnectionswithotherideas.When‘themindperceivestheagreementordisagreementoftwoideasimmediatelybythemselves,withouttheinterventionofanyother’,wehaveintuitiveknowledge,aswhenwe‘perceive’directly‘thatthreearemorethantwo,andequaltooneandtwo’(ibid.:IV,II,i).Whentheconnectionisindirect,wehavedemonstrativeknowledge,aswhenwecannotknowthattheanglesofatrian-gleareequaltotworightangles‘byanimmediateviewandcomparingthem’,butneedinsteadto‘findoutsomeotherangles,towhichthethreeanglesofatrianglehave684\nLOCKEanequality;andfindingthoseequaltotworightones,comestoknowtheirequalitytotworightones’(ibid.:IV,II,ii).BesidesintuitiveanddemonstrativeknowledgeLockerecognizessensitiveknowledge:knowledgeof‘theexistenceofparticularextendedobjects,bythatperceptionandcon-sciousnesswehaveoftheactualentranceofideasfromthem’(ibid.:IV,II,xiv),thatis,knowledgebyoursensesofwhatisnowgoingonbeforeoureyes.SensitiveknowledgeseemsnottofitLocke’sofficialdefinitionofknowledge:itdoesnotpertaintosomecon-nectionbetweentwoideas,buttothepresentexistenceofsomethingintheworldcor-respondingtoourpresentsensoryperceptionsorideas.Thesethree‘degrees’ofknowledgecutacrossaclassificationLockeproposesoftheagreementorconnectionbetweenideas,theperceptionofwhichconstitutesknowl-edge,intofour‘sorts’:‘Identity,ordiversity’;‘relation’;‘co-existence,ornecessarycon-nection’;and‘realexistence’(ibid.:IV,I,iii).Thesefoursortsofconnectiongiverise,respectivelyandroughly,tointuitivelyknownpropositionssuchasthat‘whiteiswhite’orthat‘threeismorethantwo’;tointuitivelyordemonstrativelyknownpropositionssuchasthoseaboutgeometricalfigures;topropositionsaboutthepropertiesofsubstancessuchasgold;andtothepropositionsthatmakeupsensitiveknowledge.Itisplausibletothinkthatourknowledgeofthepropertiesoftrianglesisbasedonourgrasp,bymeansofourunderstanding,ofconnectionsbetweenideas.Butwhatofourknowledgeofthepropertiesofsilver(thatitdissolvesinnitricacid,forexample),orofgold(thatitdoesnotsodissolve)?Here,asLockefreelysays,thereisnoconnec-tion,discoverablebyandintelligibletoourunderstanding,betweenourideas;here,heacknowledges,weare‘leftonlytoobservationandexperiment’(ibid.:IV,III,xxviii).Hispositiononthisis,simply,thatthesearenotcasesofknowledge.Theyarecasesofwhathecalls‘belief’or‘opinion’(ibid.:IV,XV,iii).KnowledgeandBeliefLocke’snotionofKNOWLEDGE(chapter1)ashedefinesitisrelatedtothemorerecentoneofapriorior‘conceptualknowledge’.Hisstatementthat‘insomeofourideastherearecertainrelations,habitudes,andconnexions,sovisiblyincludedinthenatureoftheideasthemselves,thatwecannotconceivethemseparablefromthem,byanypowerwhatsoever’(ibid.:IV,III,xxix)parallelswhathassometimesbeensaidinthiscenturyabouttheaprioriknowledgesuchasthatwhichwehaveinmathematics.Similarly,hisnotionof‘belief’,whichisbasedon‘observationandexperiment’becauseof‘awantofadiscoverableconnectionbetweenthoseideaswhichwehave’(ibid.:IV,III,xxviii),canberelatedtothenotionoftheaposterioriknowledgeofthekindwhichwehave,inasys-tematicform,intheempiricalsciences.YetjustasLockewouldcallsuchempiricalknowl-edgenot‘knowledge’assuchbutrather‘belief’,sohewouldnotcallanorganizedbodyofit,asinchemistry,‘ascience’.Geometryandarithmeticaresciencesforhim,and(sohesays)MORALITY(chapter6)couldbecomeone,but‘naturalphilosophyisnotcapableofbeingmadeascience’(ibid.:IV,II,x).Locke’sviewis,then,quiteclearly,thatthereisahorizonandlimittoourknowledge.Moreover,thequestionwemightaskofhim–namely,whythisisso,whyitisthatinsomecasesthereareintelligibleandgraspableconnectionsbetweenourideas,andinothercasestherearenot–isonetowhichhegivesastructuredanswer.Alongthewaytouncoveringthatstructurewewillreviewsomeother,butrelated,aspectsofhisphilosophytoo.685\nR.S.WOOLHOUSEAsahelptosupportinghisclaimthatallourideas,allthematerialsofknowledge,stemfromexperience,Lockemakesanappealtoadistinctionbetweensimpleandcomplexideas–anappealwhichhadbeenmadeearlierbyGassendi(1592–1655)andwastobemadebysomelaterphilosophers,includingHume.So,forexample,theideasofacentaur,ofGodandofinfinity(ideasofwhoseobjectswecanhardlyhavehadexperi-ence)arecomplexideas,whosecomponentsimplesorpartsmust(perhapsaspartsofothercomplexes)havecomefromexperience.Atthesametimehedividescomplexideasintovariouscategories,ofwhichsubstancesandmodesaretwo.‘Substances’,saysLocke,areideasof‘thingssubsistingbythemselves’(ibid.:II,XII,vi);theyare,asisplainfromhisexamples,ideasofnaturallyoccurringkindsofmaterialthing–justthekindofthing,infact,aboutwhosepropertiesweformbeliefsonthebasisofexperienceandobservation.Lockedistinguishesthesepropertiesthatmakeupourideaofgoldinto‘primaryandoriginal’,and‘secondary’.Thisfamousdistinction,whichhesharedwithotherearlymodernphilosophersandscientists,suchasGalileo(1564–1642),DESCARTES(chapter26)andBoyle,ismadeagainstthebackgroundofcorpuscularianism–atheorycommonintheseventeenthcentury,butonederivedfromtheclassicalGreekatomists,Democritus(c.460–c.370BC)andEpicurus(341–270BC),whoheldthatmatteriscom-posedoftiny,imperceptibleparticles.Primaryqualitiesarequalitiesthatbelongnotonlytosubstancessuchasgold,butalsototheinsensiblecorpusculeswhichmakethemup.Thus,apieceofgoldissolid,hasextensionandshape,andcanbemoved;and,accordingtothecorpusculartheory,itscorpusculeshavethesequalitiestoo.Secondaryqualities,suchascolourandtaste,belongtoapieceofgoldbutnottoitscomponentcorpuscules.Itisbecausethesesecondaryqualitiesweresupposedtoarisefromthearrangementofthesolid,shapedandmovablecorpusculeswhichconstitutegold,thatis,fromtheprimaryor‘original’(originating)qualitiesoftheparts,thattheywerenottobeattributedtothecorpusculesthemselves.Afurtherpointtobenoted,andonewhichhasledtocontinuingcontroversy,isthatLockesaysaboutourcomplexideasofsubstancesthattheirmaincomponentofthecomplexis‘thesupposed,orconfusedideaofsubstance’,somethingwhichhealsocalls‘substratum’or‘puresubstanceingeneral’(ibid.:II,XXIII,ii).Lockeissometimestakentobesayingherethatmaterialthingsaremorethan‘bundlesofcoexistingproperties’(asBertrandRUSSELL(chapter37)laterheld),buthave,inadditiontotheirproperties,apropertyless‘substratum’which‘supports’theirproperties.Accordingtoanotherinterpretation,however,Lockeisnotinterestedinabstractquestionsaboutthediffer-encebetween‘things’and‘properties’orintheontologyunderlyingthedistinctionbetweenSUBJECTANDPREDICATE(pp.93–6).Rather,histalkofa‘substratum’supportingpropertiesisareferencenottosomefeaturelessI-know-not-what,buttomatterasunderstoodbythecorpusculartheory,andintermsofwhichthepropertiesofasubstancearetobeexplained.Ourcomplexideaofgold,whichwehavederivedfromexperience,comestobethatforwhichtheword‘gold’stands–Lockespeaksofitas‘nothingbutthatabstractideatowhichthenameisannexed’(ibid.:III,VI,ii),and,assuch,hereferstoitasa‘nominalessence’.Butthisnominalessenceofgold,composedasitisfromvariousqualitieswhichexperienceandobservationhaveshownusgotogetherinthematerialstuffwecall‘gold’,isdifferentfromwhatLockecallsits‘realessence’.Thislatteristhat‘upon686\nLOCKEwhichdependsthisnominalessence,andallthepropertiesofthatsort’(ibid.:III,VI,ii);itisthat‘onwhichallthepropertiesofthespeciesdepend,andfromwhichalonetheyallflow’(ibid.:III,V,xiv).Inthecaseofsubstancesitis,inshort,thatarrange-mentofprimary-qualitiedimperceptiblecorpusclesonwhichdependtheprimaryandsecondaryqualitiesofthesubstanceasweexperienceit.Differentmaterialsubstanceshavedifferentobservablequalitiesbecausetherearedifferencesintheshape,size,arrangementandmotionofthe‘insensiblecorpuscules’(ibid.:IV,III,xxv)whichmakeuptheirrealessencesorrealconstitutions.Thisaccountofasubstance’srealessenceintermsofthecorpusculartheoryofmatterisoneoftheanti-AristotelianfeaturesoftheEssaywhichwerementionedearlier.‘Concerningtherealessencesofcorporealsubstances’,hesays,‘thereare...twoopinions’.His,the‘morerational’,takesitthatsubstances‘haveareal,butunknownconstitutionoftheinsensibleparts,fromwhichflowthosesensiblequalities,whichserveustodistinguishthemonefromanother’.Theotherbelongstothehylo-morphictheoryoftheAristoteliansaccordingtowhichamaterialsubstanceisacom-positeof‘form’(morphe)and‘matter’(hyle).Accordingly,itsupposes‘acertainnumberofformsormoulds,whereinallnaturalthings,thatexist,arecast,anddoequallypartake’(ibid.:III,III,xvii).LikeRobertBoyle,theRoyalSocietychemist–whomadethepointinhisbookTheOriginofFormsandQualities(published1666)–Lockeandother‘modern’philosophersoftheseventeenthcenturyfeltthatthisaccountofrealessenceshad‘verymuchperplexedtheknowledgeofnaturalthings’(Locke1975:III,III,xvii).Materialsubstanceswerebetterunderstood,theythought,intermsofanotherclassicalGreektheory,thatofatomism.Nowthereasonwhynaturalphilosophy,thestudyofsubstancessuchasgoldorlead,is‘incapableofbeingmadeascience’issimplythatthenominalessenceofthosesub-stances,ourideasofthem,arenotideasoftheirrealessences;evensupposingthatthecorpusculartheoryistrue,wedonotknowtheactualdetailsofthecorpuscularcon-stitutionofanysubstance.Asaconsequence,ourideasornominalessencesofthevarioussubstancesrelateinsteadtosomecollectionofthepropertieswhicharesup-posedtoflowfromanddependontheirrealessences.Nowthough,itmaybesupposed,gold’smalleabilityisnecessarilyconnectedwithitsreal,atomicconstitution;becauseourideaofgoldisnotofthatconstitution,thereisnonecessaryconnectionbetweengold’smalleabilityandourideaofgold.Naturalphilosophyisnotcapableofbeingmadeasciencebecausewedonothavetherightideas;wedonothaveknowledgeofthecorpusculardetailsoftherealessencesofthesubstanceswhichitstudies.ThecasewithGEOMETRY(chapter11)andthemodesitstudiesismuchthesame,butdifferentinatleastoneveryimportantrespect.Likesubstances,geometricalfigures,suchasthetriangle,havearealessencetoo,arealessenceonwhichtheirpropertiesdepend.Ofcourse,sincethetriangleisnotamaterialthingbutratherashape,orawayinwhichmaterialthingsmaybearranged,itsrealessenceisnotacorpuscularconstitutionorarrangementofatomicparticles.However,justasgoldis,initsrealessence,atomicparticlesarrangedinacertainway(thedetailsofwhichareunknowntous),atriangleis,quiteobviously,‘afigureofthreelinesenclosingaspace’.Thisis,asLockesays,‘notonlytheabstractideatowhichthegeneralnameisannexed’(ibid.:III,III,xviii),notonlythenominalessenceoftriangularityandwhatatriangleistous,butalsotherealessence,whatatrianglereallyis:itis‘theveryessentia,orbeing,of687\nR.S.WOOLHOUSEthethingitself,thatfoundationfromwhichallitspropertiesflow,andtowhichtheyareallinseparablyannexed’(ibid.:III,III,xviii).Sothepropertiesofatriangle,suchasitshavingexternalanglesequaltoitsinter-nalopposites,flowfrom,andsoarenecessarilyconnectedwith,itsrealessence,andourideaofthetriangleisanideaofitsrealessence.Itfollowsthatthereisanecessaryconnection,inthiscase,betweenourideaofatriangleanditsvariousproperties.Thefollowingpassagepointsupthiscontrastbetweenthewayourknowledgeislimitedinnaturalphilosophy,duetoawantoftherelevantideas,andthewayitisnotlimitedinthiswayingeometry:Substancesaffordmatterofverylittlegeneralknowledge;andthebarecontemplationoftheirabstractideas,willcarryusbutaverylittlewayinthesearchoftruthandcertainty...Experienceheremustteachme,whatreasoncannot:anditisbytryingalone,thatIcancertainlyknow,whatotherqualitiesco-existwiththoseofmycomplexidea,v.g.whetherthatyellow,heavy,fusible,body,Icallgold,bemalleable,orno;whichexperience...makesmenotcertain,thatitisso,inalloranyotheryellow,heavy,fusiblebodies,butthatwhichIhavetried...Becausetheotherpropertiesofsuchbodies,dependingnotonthese,butonthatunknownrealessence,onwhichthesealsodepend,wecannotbythemdiscovertherest.(Ibid.:IV,XII,ix)Ourknowledgeinnaturalphilosophyis,therefore,severelylimitedbythenatureofourideas,bythefactthattheyarenotideasoftherelevantrealessences.Butgeome-tryandmathematicsarenottheonlyareaswhereitisnotsubjecttosimilarlimits.Onthegroundsthattheideasofmoralityare,likethoseofmathematics,modes,whoserealessenceswemightcometoknow,Lockesuggeststhepossibilityof‘plac[ing]moral-ityamongstthesciencescapableofdemonstration:whereinIdoubtnot,butfromself-evidentpropositions,bynecessaryconsequences,asincontestableasthoseinmath-ematics,themeasuresofrightandwrongmightbemadeout’(ibid.:IV,III,xviii).The‘moralscience’LockeenvisageswouldhaveasoneofitsbasestheideaofGod,‘infiniteinpower,goodness,andwisdom’(ibid.:IV,III,xviii),whosedictatesare‘theonlytruetouchstoneofmoralrectitude’(ibid.:II,XXVII,viii).ItwouldhaveasanothertheideaofourselvesasdependentcreaturesofGodwhohavesufficientunderstandingandrationalitytounderstandhiswill:fromthiswe‘ascertainlyknowthatmanistohonour,fearandobeyGod,as...thatthree,four,andseven,arelessthanfifteen’(ibid.:IV,XIII,iii).Thoughnotsupposingthissystematicmoralsciencetobemorethanapos-sibilityyettoberealized,Lockeneverthelessthoughtthatwehadalreadyreasonedourwaytosomeitemsofmoralknowledge.Hemakesclear,however,particularlyintheReasonablenessofChristianity,thatwhathecallsreason’s‘greatandproperbusinessofmorality’(Locke1873:140)isnotaneasyone.Sothosewhoareimmersedinthenecessitiesofeverydaylife,andhaveneitherthetimenorabilitytouncovermoraltruthsbyreasoncanfinditinthegospels.The‘lawwhichGodhassettotheactionsofmen...[is]promulgatedtothembythelightofnature[reason],orthevoiceofRevelation[thegospels]’(Locke1975:II,XXVIII,viii).ButeventhoughmoralitycanbeshowntousbyREVELATION(pp.622–4)andwithoutrecoursetoreason,thevoiceofrevelationisstillanswerabletothelightofnature.Foronething,weneedtobeassuredthatthevoiceisgenuine,andreasonmustbethejudgeofthis.Moreover,wecannotbeascertainofthisasweareofknowledgewhichisbaseddirectlyontheuseofourownreason.‘Theknowledge,wehave,thatthisrevelationcame688\nLOCKEatfirstfromGod,canneverbesosure,astheknowledgewehavefromtheclearanddis-tinctperceptionoftheagreement,ordisagreementofourownideas’(ibid.:IV,XVIII,iv).Aswehaveseen,itisaGod-givendutycarefullytoexamineanypropositionbeforeautonomouslygivingorwithholdingourindividualassent:it‘wouldbetosubverttheprinciples,andfoundationsofallknowledge...[if]whatwecertainlyknow,givewaytowhatwemaypossiblybemistakenin’(ibid.:IV,XVIII,v).Reasonissuperioroverrevelationinanotherwaytoo.Thoughsomethingsarediscoverableonlybyrevelation,anyknowledgethatisnecessaryforsalvationcanbereachedbyournaturalfaculties.Godhas‘givenallmankindsosufficientalightofreason,thattheytowhomthiswrittenword[theBible]nevercame,couldnot(whenevertheysetthemselvestosearch)eitherdoubtofthebeingofaGod,oroftheobedienceduetoHim’(ibid.:III,IX,xxiii).PersonalIdentityInthecourseofreachingconclusionsabouttheoriginandlimitsofknowledge,Lockehadoccasiontoconcernhimselfwithtopicswhichareofphilosophicalinterestinthemselves.Oneoftheseisthequestionofidentity,whichincludes,morespecifically,thequestionofpersonalidentity:whatarethecriteriabywhichapersonatonetimeisnumericallythesamepersonasapersonatanothertime?Herehemakesthepoint,takenupbylaterwriters,thatthereisarelativityaboutidentity.Whenwepointandaskwhether‘this’iswhatwasherebefore,itmatterswhatkindofthing‘this’ismeanttobe.If‘this’ismeantasamassofmatterthenitiswhatwasbeforesolongasitconsistsofthesamematerialparticles;butifitismeantasalivingbodythenitsconsistingofthesameparticlesdoesnotmatterandthecaseisdifferent.‘Acoltgrownuptoahorse,sometimesfat,sometimeslean,isallthewhilethesamehorse:though...theremaybeamanifestchangeoftheparts’(ibid.:II,XXVII,iii).So,whenwethinkaboutpersonalidentity,weneedtobeclearaboutadistinc-tionbetweentwothingswhich‘theordinarywayofspeakingrunstogether’–theideaof‘man’,andtheideaof‘person’.Aswithanyotheranimal,theidentityofamanconsists‘innothingbutaparticipationofthesamecontinuedlife,byconstantlyfleetingparticlesofmatter,insuccessionvitallyunitedtothesameorganizedbody’(ibid.:II,XXVII,vi).Buttheideaofapersonisnotthatofalivingbodyofacertainkind.Apersonisa‘thinking,intelligentbeing,thathasreasonandreflection’(ibid.:II,XXVII,ix)andsuchabeing‘willbethesameselfasfarasthesameconsciousnesscanextendtoactionspastortocome’(ibid.:II,XXVII,x).Lockeisatpainstoarguethatthiscontinuityofself-consciousnessdoesnotnecessarilyinvolvethecontinuityofsomeimmaterialsubstance,inthewaythatDescarteshadheld.Forallweknow,saysLocke,consciousnessandthoughtmaybepowerswhichcanbepossessedby‘systemsofmatterfitlydisposed’(ibid.:IV,III,vi);andevenifthisisnotsothequestionoftheidentityofpersonisnotthesameasthequestionoftheidentityofanimmaterialsubstance.Forjustastheidentityofahorsecanbepreservedthroughchangesofmatter,anddependsnotontheidentityofacontinuedmaterialsub-stancebuton‘theunityofonecontinuedlife’(ibid.:IV,XXVII,x),sotheidentityofapersondoesnotdependonthecontinuityofanimmaterialsubstance.Theunityofonecontin-uedconsciousnessdoesnotdependonitsbeing‘annexedonlytooneindividualsubstance,[andnot]...continuedinasuccessionofseveralsubstances’(ibid.:II,XXVII,x).ForLocke,then,personalidentityconsistsinanidentityofconsciousness,andnotintheidentityofsomesubstancewhoseessenceitistobeconscious.689\nR.S.WOOLHOUSE2Locke’sPoliticalPhilosophyLocke’sanonymousTreatisesofGovernment,whichcontainhispoliticalphilosophy,werepublishedinthesameyearashisEssay,andtwoyearsaftertheGloriousRevolu-tioninwhichWilliamofOrangereplacedJamesIIontheEnglishthrone.Buttheircom-positiondatesfromsometenyearsearlier,andtothetimeoftheExclusioncrisisduringwhichLordShaftesbury,Locke’spatron,sought,withothers,toexcludeJamesfromsuccessiontothethrone.TheTreatisessupportedtheirargumentsforgovernmentbyconsent,andfortherighttoreligiousdissent.TheFirstTreatisecontainscriticismoftheoriesofabsolutemonarchy,andofthedivinerightofkings,theoriesforwhichRobertFilmer(c.1588–1653)hadarguedinhisPatriarchia(published1680).LockefoundquiteunworkablethetheorythatAdamhadbeengivenabsoluteandtotalpoliticalauthoritybyGod,authoritywhichwasthenpassedontohisheirs.ThetheorycannotshowofanyparticularrulerthatheisagenuineheirtoAdam’soriginalauthority,andsoitisincapableofjustifyinganyactualpoliticalauthority.IntheSecondTreatiseLockesoughtsomeother,andmoreworkable,foundationforpoliticalauthority.SubjectsdohaveadutytoGodtoobeytheirruler,hedidnotdeny.Butrulersthemselvescorrespondinglyhavedutiestotheirsubjects.Rulersarenotabsolute,andiftheircommandsdonotdeserveobediencethenresistancetothemmightbejustified.LockebeganwiththeideaoffamiliesandinformalgroupsofpeoplelivinginaSTATEOFNATURE(p.574).Insuchanaturalstatetherearenopoliticalauthoritiesandpeople’sdutiesstemdirectlyfromGod;theyaredutiesnotto‘harmanotherinhislife...liberty...orgoods’(Locke1960:II,§6).Butwhatmakesthe‘goods’,towhichapersonhasaright,hisgoodsinthefirstplace?How,inthestateofnature,canitcomeaboutthatapersonhasarighttothelandhecultivates?AccordingtoFilmer,justasabsolutepoliti-calauthorityispasseddownfromAdamtohisheirs,soispossessionoftheearthanditsfruits.AccordingtoLocke’slabourtheoryofproperty,bycontrast,justasallmenweremadefreeandequalinGod’seyes,withnoonedirectlysetinpoweroverthem,soGodgavetheearthanditsfruitstoallmenequally.ClearlyitfollowsfromthisthatAdamandhisheirscouldnotbealoneinhavingprivateproperty,butthisisnotbecausenoonehasprivatepropertyandbecausethereisworldwidecommunalownership.Thelandsomeonecultivates,anditsfruitsandcrops,arehis,Lockeexplains,becauseeachindi-vidualhasaGod-givenrighttohisownlifeandlabour,andconsequently,anindividualrighttowhathe‘mixes’thatlabourwithandproducesbyit.‘Whatsoeverthen,heremovesoutofthestatethatnaturehathprovided,andleftitin,hehathmixedhislabourwith,andjoinedtoitsomethingthatishisown,andtherebymakesithisproperty’(ibid.:II,§27).Therightsandduties,concerninglife,libertyandpossessions,whichpeoplehaveinthestateofnature,maynotnecessarilyberespectedandobeyed.Peoplemaylackthepowertodefendtheirrights,orgotoofarintheirdefence.TheythereforeagreetocomeoutofthestateofnatureandtoformPOLITICALSOCIETY(chapter8).Theyagreetounite,andto‘enterintosocietytomakeonepeople,onebodypolitic,underonesupremegov-ernment’(ibid.:II,§89).Theyagreetoempoweranauthorityoverthem;they‘setup690\nLOCKEajudgeonearthwithauthoritytodetermineallthecontroversiesandredresstheinjuriesthatmayhappentoanymemberofthecommonwealth’(ibid.:II,§89).Butthisauthorityintowhosehandspeopleputpowerisnot,inLocke’sanalysisofcivilsociety,absolute.Heisanswerableto‘thewillanddeterminationofthemajority’(ibid.:II,§96).Theviewsandwishesofthemajorityareacourtofappealagainstthedecisionsoftherulingauthority.Lockeanpoliticalsocietyisnotonlybegunby‘theconsentofanynumberoffreemencapableofamajority’(ibid.:II,§99),itiscontinuedandsustainedbyittoo.Peopleunderanabsoluteauthority(whetherdivinelyestablishedasinFilmer’sview,orsetup,asintheequallyabsolutistaccountofHOBBES(chapter28)bythepeople)are,forLocke,notinacivilsociety;theyareeffectivelystillinastateofnature.Ofcourse,unlikethosewhoforsometimehavebeeninastateofnature,mostpeoplearesimplybornintoanalreadyexistingcivilsociety.Itisnotbyanychoiceoftheirsthattheycomeunderitslawsandauthority;theyenteredintonoagreementtouniteandsetuptheauthority.Butjustascontinuingconsentiswhatunderwritesanauthor-ityaswellaswhatcreatedit,soLockedistinguishesbetweentacitandexplicitconsent.Apersonmaynotchoosewheretobeborn,andsowhoseauthoritytobesubjectto;butbyremainingtherehegiveshistacitconsent.Fromtheideathatitisthe‘willanddeterminationofthemajority’(ibid.:II,§96)thatistheultimatebasisofpoliticalauthority,therefollowsthepossibilityoflegitimateresistancetothatauthority,apossibilitywhichdoesnotfollowfromFilmer’sdivinerighttheory:‘Thecommunityperpetuallyretainsasupremepowerofsavingthem-selvesfrom...theirlegislators,whenevertheyshallbesofoolish,orsowicked,astolayandcarryondesignsagainst[their]libertiesandproperties’(ibid.:II,§149).Justasitisinthewillofthepeoplethatpowerresides,soitisthat‘thePeopleshallbeJudge’(ibid.:II,§240)ofwhenthatpowershouldbeused.FurtherReadingJamesGibson’sLocke’sTheoryofKnowledge(1917)andR.I.Aaron’sJohnLocke(1937)aretwovaluableclassicaccountsofLocke’sepistemologicalthought.TherearemorerecentgeneralaccountsbyDunn(1984),Mabbot(1973),Woolhouse(1983)andYolton(1985).Moredetailedandspecialized(whetherhistoricallyorphilosophically)treatmentscanbefoundinAyers(1992),Bennett(1971),Mackie(1976),Woolhouse(1971)andYolton(1970).Ashcraft(1987)andDunn(1969)dealspecificallywithLocke’spoliticalphilosophy,Colman(1983)withhismoralphilosophy,andAshcraft(1969)andCragg(1950)withhisphilosophyofreligion.Thelastfull-lengthbiographyofLockeisthatofCranston(1957).AnexhaustivelistofbooksandarticlesonLockeisprovidedbyHallandWoolhouse(1983).ReferencesLockeLocke,John1873[1695]:TheReasonablenessofChristianity.Vol.7ofTheWorksofJohnLocke.London.——1960[1690]:TwoTreatisesofGovernment(editedbyPeterLaslett).Cambridge:CambridgeUniversityPress.——1975[1690]:EssayConcerningHumanUnderstanding(editedbyP.H.Nidditch).Oxford:ClarendonPress.691\nR.S.WOOLHOUSE——1990[1671]:DraftAofAnEssayConcerningHumanUnderstanding(editedbyPeterH.NidditchandG.A.J.Rogers).Oxford:ClarendonPress.OtherwritersAaron,R.I.1937:JohnLocke.Oxford:ClarendonPress.Alexander,P.1985:Ideas,QualitiesandCorpuscles:LockeandBoyleontheExternalWorld.Cambridge:CambridgeUniversityPress.Ashcraft,R.1969:FaithandKnowledgeinLocke’sPhilosophy.InJ.W.Yolton(ed.)JohnLocke:ProblemsandPerspectives.Cambridge:CambridgeUniversityPress.——1987:Locke’sTwoTreatisesofGovernment.London:AllenandUnwin.Ayers,M.1992:Locke,2vols.London:RoutledgeandKeganPaul.Bennett,J.1971:Locke,Berkeley,Hume:CentralThemes.Oxford:OxfordUniversityPress.Chappell,V.1994:TheCambridgeCompaniontoLocke.Cambridge:CambridgeUniversityPress.Colman,J.1983:JohnLocke’sMoralPhilosophy.Edinburgh:EdinburghUniversityPress.Cragg,G.R.1950:TheReligiousSignificanceofJohnLocke.InG.R.Cragg(ed.)FromPuritanismtotheAgeofReason.Cambridge:CambridgeUniversityPress.Cranston,M.1957:JohnLocke:ABiography.London:Longmans.Dunn,J.1969:ThePoliticalThoughtofJohnLocke.Cambridge:CambridgeUniversityPress.——1984:Locke.Oxford:OxfordUniversityPress.Gibson,J.1917:Locke’sTheoryofKnowledgeanditsHistoricalRelations.Cambridge:CambridgeUniversityPress.Hall,R.andWoolhouse,R.1983:80YearsofLockeScholarship:ABibliographicalGuide.Edinburgh:EdinburghUniversityPress.Mabbot,J.D.1973:JohnLocke.London:Macmillan.Mackie,J.L.1976:ProblemsfromLocke.Oxford:ClarendonPress.Martin,C.B.andArmstrong,D.M.(eds)1968:LockeandBerkeley:ACollectionofCriticalEssays.London:Macmillan.Tipton,I.(ed.)1977:LockeonHumanUnderstanding:SelectedEssays.Oxford:OxfordUniversityPress.Woolhouse,R.S.1971:Locke’sPhilosophyofScienceandKnowledge.Oxford:Blackwell.——1983:Locke.Brighton:Harvester.Yolton,J.W.1970:LockeandtheCompassofHumanUnderstanding.Cambridge:CambridgeUniversityPress.——1985:Locke:AnIntroduction.Oxford:Blackwell.DiscussionQuestions1Didthemechanicalphilosophyandtheexperimentalphilosophyoftheseven-teenthcenturyimproveuponAristotelianphilosophy?2WasLocke’smethodwellsuitedtohisprojectofsurveyingourunderstanding?3Ifourknowledgehaslimits,howcanwedeterminewhattheyare?4Isourknowledgebestunderstoodasansweringtheoreticaldemandsorasansweringthepracticalneedsofeverydaylife?5Howisknowledgerelatedtoexperience?6TowhatextentwasLockean‘empiricist’?7WhatdoesLockemeanby‘idea’?8Mustsenses,reasonandunderstandingallbeinvolvedinourexplanationofhumanknowledge?692\nLOCKE9Priortoexperience,istheminda‘whitepaper,voidofallcharacters,withoutanyideas’?10Howadequateisanaccountofknowledgeintermsofperceivingtheconnectionordisagreementofideas?11WasLockejustifiedinthinkingthatempiricalsciencecannotyieldknowledge?12Howcanwetellwhenanideaissimple?13Shouldweacceptthedistinctionbetweenprimaryandsecondaryqualities?14WasLockecorrectinclaimingthatourideasofphysicalthingsarenotideasoftheirrealessences?15What,forLocke,istheroleofrealessencesingeometricalknowledge?16Canmoralitybeascience?17Isreasonsuperiortorevelation?18Arequestionsofidentityrelativetothekindstowhichthethingsaskedaboutbelong?Whataretheimplicationsofyouranswerforquestionsofpersonalidentity?19Whatproblemsmightarisefortheclaimthatpersonalidentityconsistsinthecontinuityofconsciousness?Whatotherviewsareworthconsidering?20Wouldbeinganheirtooriginallegitimateauthorityjustifyoneinholdingcurrentpoliticalauthority?21Whatroledoesthe‘stateofnature’playinLocke’spoliticalthought?Woulditmatterifarivalviewofthestateofnatureturnedouttobemoreaccurate?Orifthereneverwasastateofnature?22Howcansomethingbethepropertyofaperson?23Whatiscivilsociety?Whatistheultimatebasisofauthorityinsuchasociety?Whatlimitstheauthorityofitsrulers?24Is‘tacitconsent’consentatall?693\n30BerkeleyHOWARDROBINSONAsacriticofLockeandaprecursorofHume,Berkeleyofferedaprofoundlyoriginalperspectiveoncentralquestionsinepistemologyandmetaphysics.Thischapter,informedbyrecentcriticaldiscussionofBerkeley’sidealism,focusesonhisimmaterialism:hisclaimthatthephysicalworldanditscontentscannotbeunderstoodasexistingindepen-dentofthemind.Berkeley’sdiscussionofabstractideas,whichisintimatelyrelatedtohisimmaterialism,isexaminedalongwithotherimportantsupportingarguments.ThechapterendswithanexplorationoftwofeaturescompletingtheBerkeliansystem,namelyhisdoctrineofspiritsandtheroleheassignedGodinexplainingourexperience.Insections1–4ofthischapterIconsidertwogroupsofargumentsthatBerkeley(1685–1753)usedinhisattempttoproveidealism.Withregardtothefirstgroup,afterinvestigatingBerkeley’sattackonabstractideas,Iconcludethatarevisedversionoftheargumentshowsthatwecanhavenopositiveconceptionofmatter.Thisisstrengthenedbyaproofthatnosensiblequality–nor,probably,anythinganalogoustoasensiblequality–couldbemind-independent.Asmind-independentmatterwouldhavetobemorethananemptystructure,andasonlysomethinglikeasensiblequalitycouldfillsuchastructure,theconceptofmind-independentmatterisdefective.Withregardtothesecondgroup,Berkeley’soverstrongimplicationthat,becausemythoughtisinmymind,soisitsobject,canbereformedtotheweakerclaimthat,becauseallourconceivingofobjectsisofthemastheyarefromviewpoints,itfollowsthatwecon-ceiveofthem,notastheyare‘inthemselves’butastheyaretoapotentialperceiver.Iraisedoubtsaboutwhetherthisiscorrect,butaugmentitwithBerkeley’sargumentthatsizeisperception-dependent,forbodiescannotbemind-independentiftheirdimensionsarenot.Insection5IbrieflydiscussBerkeley’suseofhisdoctrineofspiritandhisinvocationofGodtoexplainexperienceintheabsenceofamind-independentworld.1Introduction:Berkeley,CommonSenseandthe‘NewPhilosophy’Commonsenseisusuallythoughttoaffirmthefollowingtwoprinciples:(CS1)Innormalvisualandtactileexperiencewearedirectlyawareofphysicalobjects.\nBERKELEY(CS2)Theexistenceofphysicalobjectsdoesnotdependon,norconsistin,theirbeingperceived:theyaremind-independent.The‘newphilosophy’oftheseventeenthcenturywastakenbymanyphilosophersandscientistsasrefutingthefirstoftheseprinciples:theobjectsofourawarenessarenotobjectsintheexternalworld.Thishadtobesobecauseobjectsintheexternalworlddonotthemselvespossesssecondaryqualitiesintheforminwhichthosequalitiesareperceived.Theexternalphysicalworldiscolourless,odourlessandsilent;secondaryqualitiesintheformweknowthemareideasinthemind.Allweareawareofareourideas;thephysicalworld,whichisagreatmachinecomposedofinsensibleatoms,liesconcealedbehindtheveilofperceptionthatisconstitutedbythoseideas.Berkeleyacceptedtheargumentsthatweresupposedtoshowthatwecouldnotbedirectlyawareofamind-independentphysicalworld.Insteadofconcluding,likehismoreconventionalcontemporaries,thatwearenotdirectlyawareofthephysicalworld,heconcludedthatthephysicalworldisnotmind-independent:hechose,thatis,tosaveCS1ratherthanCS2.Hehadtwomotivesforthispreference.Firstly,hethoughtthatputtingthephysicalworldbehindthe‘veilofperception’openedthewaytoSCEPTICALDOUBT(p.50)aboutitsexistence;secondly,hefearedthatthemechanisticconceptionoftheworldbehindtheveilfavouredbythenewscience,smackedofamaterialistand,therefore,anatheistmetaphysic(Berkeley1713).Berkeley’sidealistictheoryofthephysicalworldisexpressedinthemaximthatesseestpercipi–tobeistobeperceived.ThisfamousmaximisequivalenttotheaffirmationofCS1withoutCS2,foritistheaffirmationthatthephysicalworldisthatofwhichwearedirectlyaware,togetherwiththedenialthatitismind-independent.Bymakingthephysicalworlddirectlyaccessibletoexperience,thisstrategyseemstoavoidthethreatofscepticism,andbymakingtheworldmentalitavoidsmaterialism.Ofcourse,ifCS1andCS2arebothpartsofcommonsense,Berkeley,inrejectingCS2,setshimselfagainstcommonsense.AccordingtoBerkeley,however,theclaimthatthephysicalismind-independentisnottrulyapartofcommonsense,forhebelievesthatitrestsonaconfusionbetweenthetruththatthephysicalworldismorethantheexperienceofanygivenindividualandhenceindependentofanygivenmind,andthefalsehoodthatittranscendsalltheexperiencesthatmightbehadofitbyallmindstakentogether(Berkeley1713:III,230–1).AccordingtoBerkeley,thislatterkindofmind-independenceisnotmerelynotcommonsense,itisstrictlyunintelligible.Soitisnotjustthatwhenfacedwithanexclusivechoicebetweenacceptingthedirectperceiv-abilityofthephysicalworldandacceptingitsmind-independence,hepreferstheformer;rather,hedeniestheintelligibilityofafullymind-independentphysicalexistence.Berkeleyhasavarietyofargumentsagainstthepossibilityofmind-independentorunperceivedobjects.Mystrategyinthischapteristoexaminetheseargumentssym-pathetically,whichinvolvesdevelopmentaswellasexposition.IshallorganizeBerke-ley’smoreorlessindependentargumentssothattheyformtwogroups,eachgroupconstitutinganextendedlineofargument.OnelineofargumentstartsfromBerkeley’sattackonabstractideas.Byreformingthatattack,andaugmentingitwithaclusterofargumentscentringonthemaximthat‘nothingcanbelikeanideabutanidea’,wecanfindinBerkeleyaplausiblecaseforbelievingthatwecanformnocontentful695\nHOWARDROBINSONconceptionofMATTER(pp.647–8).Theothergroupofargumentsstartsfromtheclaimthatwecannotconceiveofunperceivedobjects;bybuttressingthiswithhisargumentsaboutthesense-dependenceofsize,wecandevelopreasonsforthinkingthatallthesupposedlyobjectivepropertiesoftheexternalworldarereallyperception-relative.2AbstractIdeas,RelativeIdeasandImmaterialismHUME(chapter31)remarkedthatBerkeley’sdismissalofabstractideaswas‘oneofthegreatestandmostvaluablediscoveriesthathasbeenmadeoflateyearsintherepublicofletters’(Hume1739:BkI,VII).Berkeleyhimselfregardedhisattackonabstractideasasavitalcomponentintheargumentforimmaterialism,for‘[i]fwethoroughlyexamine[thebeliefinunperceivedobjects]itwill,perhaps,befoundatbottomtodependonthedoctrineofabstractideas’(Berkeley1710:§5).Butmanyphilosophershavebeenpuzzledastotheexactconnectionbetweenabstractionandmatter,andtendtoleaveitonthesidelineswhenexpoundingBerkeley’sargumentsforidealism.Itwillbehelpful,therefore,tolookcarefullyatwhattheattackonabstractionisabout.Berkeley’sargumenthasanimmediatetargetandaremotetarget.TheimmediatetargetisLOCKE’s(chapter29)accountoftheideasofdeterminableorgenericthings,suchasredasopposedtoanyparticularshadeofred,forexample,ortriangleasopposedtoanisoscelestriangleoraright-angledtriangle.Theremotetargetisanydoc-trineofconceptsthatallowsthattheyare,insomesense,intrinsicallyuniversal:heis,thatis,tryingtocarrytocompletiontheideathateverything,whetheroutsidethemindorwithinit,isparticular,andtherebytofreehisphilosophyoftheperennialproblemofuniversals,which,insomeformorother,hadbeentheheartofmostscholasticmeta-physics.Accordingtothisambitioustheory,notonlyaretherenoforms,speciesoruni-versalsintheworlditself,butevenconceptsare,inthemselves,purelyparticular.Locke,too,believedinparticularism,but,accordingtoBerkeley,headoptedanincoherentformofit.Theparticularistprogrammefacesadilemma,expressedbyLockeandquotedbyBerkeley:‘Sinceallthingsthatexistareonlyparticulars,howcomewebygeneralterms?’(Locke1689:III,III,vi;quotedinBerkeley1710:§11).Howcanaconcept,forexampletheconceptofred,whichbydefinitionissomethingthatcoversmanythings,simplybeaparticular,ratherthanbeinguniversal?Evenifonecanmanagetoavoidrealismaboutuniversalsinthings,surelyonemustbeaconceptualist,thatis,arealistaboutuniversalityinthemind,intheconceptsthroughwhichwethink?Lockegivestheanswer:‘Wordsbecomegeneralbybeingmadethesignsofgeneralideas’.Lockegroupedallmentalcontents,includingconcepts,togetherasideasandtreatedideasasmentalimages(seeYolton1984andAyers1991:I,60–9for,respectively,anattackonthisimagistinterpretationandadefenceofit).Takeninaliteralwayasakindofmentalpicture,amentalimageisprettyclearlyaparticular,buthowcanithaveasitscontentagenericordeterminableconceptsuchasredortriangle?AccordingtoBerkeley,Locke’sansweristosaythatthiskindofimagehasanindefinitenesstoit:doesitnotrequiresomepainsandskilltoformthegeneralideaofatriangle...foritmustbeneitherobliquenorrectangle,neitherequilateral,equicrural,norscalenon,butallandnoneoftheseatonce.Ineffect,itissomethingimperfectthatcannotexist,anideawherein696\nBERKELEYsomepartsofseveraldifferentandinconsistentideasareputtogether.(Locke1689:IV,VII,ix;quotedinBerkeley1710:§13)Sothereisanindefinitenesswhichisexpressedintwodifferentways.Ontheonehandtheimageistriangularwithoutbeingscaleneorequilateral;ontheotherhanditisalloftheseincompatiblepropertiesatonce,becauseitincludesthemall.Inthiswayonehasaparticularobject–theimage–whichbothleavesoutallspecificforms,andisasuperimpositionofthespecificforms:inthesewaysitisdeemedtoencompassjusttherightgeneralityofthings.Berkeley’sobjectionisthatonecanhaveneitherinde-terminateparticulars,lackingaspecificform,norcontradictoryones,containingincompatibleforms.Locke’sattempttocombineimagismwithintrinsicgenerality,therefore,fails.ManycommentatorshavearguedthatBerkeleymisinterpretsLocke’saccountofabstraction,buttheexegeticalquestionisnotveryimportant.ThetheoryofgeneralityascribedtoLockebythosewhothinkBerkeleymisinterpretshimisthetheoryBerkeleyhimselfadopts;and,althoughsomephilosophersallowthatimagescanbegenuinelyvaguetosomeextent,itisgenerallyagreedthattheycannotbevaguetotheextentandinthewaythatthetheoryattributedtoLockerequires.SoeveryoneagreesthattheviewBerkeleyattacksisfalse:theyalsoagreethattheaccounthewishestoreplaceitwithistheonlyotheroneavailabletotheimagist.Therealissue,therefore,iswhetherBerkeley’sowntheoryisadequate.Berkeley’stheoryofgeneralityhastwoparts.Thecentralclaimisthat‘anidea,whichconsideredinitselfisparticular,becomesgeneral,bybeingmadetorepresentorstandforallotherparticularideasofthesamesort’(Berkeley1710:§12).Soanentirelyparticularimageofatriangle,forexample,istakenasstandingfortrianglesasawhole.Thisaccountexplainsthemeaningofanideaintermsexternaltoit–thatis,bymakingitconsistintheidea’srelationstootherideas.Butthemeaningoftheideaswethinkissomethingweexperienceinourthinking–weknowthatinacertainactofthinkingwemean‘triangle’.ThissubjectiveelementingeneralideasBerkeleyexplainsbythesecondpartofhisaccount,whichisthedoctrineofselectiveattentionorpartialconsideration:amanmayconsiderafiguremerelyastriangular,withoutattendingtotheparticularqual-itiesoftheanglesorrelationsofthesides...InlikemannerwemayconsiderPetersofarforthasman,orsofarforthasanimal,withoutframingtheforementionedabstractidea,eitherofmanorofanimal,inasmuchasthatallthatisperceivedisnotconsidered.(Berkeley1710:§16)Sothesignificanceofaparticularideaconsistsinthewayitrepresentsotherparticularsandinthewayafeatureofitcanbegraspedwhileneglectingtherest.Theconnectionbetweenthisaccountofthoughtandidealismisquitestraight-forwardandcanbestatedinasimpleargument.1Ideasareimages.Therefore,2abstractideaswouldbeabstractimages.3Therearenoabstractimages.Therefore,697\nHOWARDROBINSON4therearenoabstractideas.5Iftherewereanideaofmatter–orofunperceivedexistenceormind-independentexistence–itwouldbeanabstractidea.Therefore,6thereisnoideaofmatter–norofunperceivedexistencenormind-independence.(1)assertsBerkeley’simagism;(2)followsbysimplelogic;(3)isthecoreoftheargu-mentagainstLocke’ssupposedposition;(4)followsfrom(2)and(3)bysimplelogic(namely,modustollens).Thecrucialclaimisin(5).Wehaveseenthat,forBerkeley,anon-abstractideaisonethatcanbeidentifiedwithanimageoranaspectofanimage.Itisdifficulttoseehowtheideasofmatter,unperceivednessormind-independencecouldbeactual,realizedfeaturesofimages.Imagescanoccurindreamsandhalluci-nations,sotheycannotactuallypossessthepropertiesofmateriality,unperceivednessormind-independence,forsuchimagesarepurelymentalobjectsofawareness.(6)followsfrom(4)and(5),alsobymodustollens.ObjectionstoBerkeley’sImagismTheargumentisvalidbutunsoundbecauseBerkeley’saccountofgenerality–explainedaboveandenshrinedin(1)–issubjecttotwomassiveobjections.Firstly,hepresentsnoaccountofwhathemeansby‘representing’or‘standingfor’.Thisisnotmerelyagapintheargument;itsuggestsacompletefailuretounderstandtheproblem.Theproblemofuniversals,classicallyreferredtoastheproblemofthe‘oneoverthemany’,isessentiallyaboutwhatisinvolvedinsomethingcomprehendingorstandingformanyinstances.Toinvokeasanunexplainednotiontheideaofstandingforis,therefore,tonametheproblem,nottoproposeasolutiontoit.ThesecondfeatureofBerkeley’stheoryisopentoanadditionalobjection.What,withinBerkeley’sparticularistONTOLOGY(chapter2),isittoattendtoanaspectofanidea?Attendingmightseemtobeapurelysensorything,analogoustofocusingacamera,but,becausethefocusingisonanaspect,notaregion,oftheimage,thefocusingisaconceptualone.Onecannotfocusonthetriangularityofatriangle–asopposedtoitscolourorsize,forexample–exceptbydeployingtheconcept‘triangle’.Selectiveattentionisofnohelpinexplaininghowtheconceptcouldbenomorethantheimage,fortheselectioninquestionisnomorethanbringingacertainconcepttobearontheimage.Ifoneisseekinganaccountofconcepts,thenotionofselectiveattentionwillnothelp,anymorethanwillthenotionof‘standingfor’.GiventheinadequacyofBerkeley’simagism,cananythingbesavedoftheargumentforimmaterialismthatrestsuponit?Toseethatsomethingcan,wemustconsiderthedistinctionbetweenpositiveandrelativeideas,whichBerkeleyadoptedfromLocke(Locke1689:II,XXIII,iii;Berkeley1710:§80;andforafullcontemporarydiscussionseeFlage1987:134–42).Tohaveapositiveideaofsomethingistoknowwhatitislikeinitself–toknowitsintrinsicnature.Tohavearelativeideaofsomethingisnottoknowitsintrinsicnature,butonlythatitstandsinsomespecificrelationtosomethingofwhichwehaveapositiveidea.Divestedofimagismandofthemisconceivedattemptstosolvetheproblemofuniversalsbygivingaparticularistanalysisofconcepts,Berkeley’sargumentscanbeusedtoshowthatwedonothaveapositiveideaofmatterorofunperceivedormind-independentexistence.Thiswouldbedonebyreplacingthe698\nBERKELEYhopelesstheorythatideasareimageswiththemoreplausibleclaimthatthecontentsofpositiveideascanberepresentedinimages.Thisisnotatheoryaboutwhatideasorconceptsareinthemselves,butafarmoremodesttheoryabouttheconditionsunderwhichwecanknowwhatthatofwhichitisaconceptorideaislikeinitself.Theclaimisthat,asfarasaspectsoftheempiricalworldareconcerned,wecanonlyknowwhattheyarelikeinthemselvesiftheycanfeatureasaspectsofsensoryideas.Thisisthecomplementoftheclaimoftenexpressedbysayingthatourtheoreticalorpurelyscientificknowledgeoftheworldistopic-neutral,meaningthatittellsusaboutthestruc-tural,relationalandcausalpropertiesofthings,butnotabouttheirintrinsicorquali-tativenatures:onlydirectsensorycontactwillgiveyouthelatter(Lockwood1989).Thisthengeneratesanewargument:7Anyfeatureofthephysicalworldofwhichwecanhaveapositiveideacanbespecificallyrepresentedinasensoryidea;thatis,inanimage.Therefore,8ifwehaveapositiveideaofmatter–orunperceivedexistenceormind-independence–thenitcanbespecificallyrepresentedinanimage.9Matter–orunperceivedexistenceormind-independence–cannotbespecificallyrepresentedinanimage.Therefore,10wedonothaveapositiveideaofmatter–orunperceivedexistenceormind-independence.Ihavealreadysuggestedtheplausibilityof(7),and(8)ismerelyaninstantiationof(7).Theargumentsfor(9)arethesameasthosefor(5),forthereisnofeatureofanimagewhichjustis,orisdedicatedtotherepresentationof,materialityorunperceivednessorextra-mentality:anyfeatureofanimagemustbeasensiblefeatureofitandsomustrepresentsomethingmorespecificandsensiblethanmateriality,unperceivednessorextra-mentality.Thereareinfacttwodifferentquestionsintheairhere.Firstly,whatsortofproper-tiesismind-independentmattersupposedtopossess?Secondly,whatkindsofideas(iftheyarerealideasatall)aretheideasofmateriality,unperceivedexistenceandmind-independence?Ihavearguedthatwecannothaveapositiveideaoftheselatter,becausewecannotexperienceorimagethepropertiesinquestion.Berkeleyconsidersthesug-gestionthatwehavearelativeideaofmateriality–thatis,ofthatcomponentinmatterwhichismorethanitsqualities.Wemaynotethatthequestionoftheintelligibilityofmind-independenceandofunperceivedexistenceismorecomplicatedthanthequestionofmatter.Humeinhistourdeforce‘Onscepticismwithregardtothesenses’(Hume1739:I,IV,ii)showedhowwecome,throughaseriesofmistakes,toattachsensetotheformertwo,whichisconsistentwithderivingallourideasfromsense.ThisiswhytheBerkelianneedsaseparateargumentagainstthepossibilityoftherebeingexternalqualities.Returningtothemainargument,wecanseethattohavearelativeideaofonerelatedthing(relatum),onemusthaveapositiveideaofboththeotherrelatumandtherelationitself.Inthecaseofmatterwecannotformaproperconceptionoftherela-tionitissupposedtostandintothatofwhichwehavethepositiveidea.TheexampleBerkeleyconsidersisthatmatteristheSUBSTRATUM(p.686)forthequalityofexten-sion,whichheexplainsasmeaningthatmattersupportsextension.Thuswehavea699\nHOWARDROBINSONrelationRxy,withybeingidenticaltothequalityofwhichweknowthenaturedirectly(thatis,someinstanceofextension),Rbeingtherelationofsupporting,andx,matter,beingknownonlyasthatwhichstandsinthesupportingrelationtotheextendedquality.Berkeleychallengesthisbydenyingthatwehaveaclearideaoftherelationinquestion.Pillarssupportroofsandbridges,butitisnotinthatsensethatmattersupportsqualities.Itisdoubtfulwhetheraproperty-lesssubstratumcanbesaidtodoanything:norisitclearthatitmakessensetosaythatanunextendedsubstratumstandsunderandsupportsthequalityofextension.Thisshowsthatitishardtoseehowonecouldhavearelativeideaofmateriality(Berkeley1713:I,197ff.).Therearecomplicationshere,however.Berkeleyisdiscussingthepossibilityofarelativeideaofmatteraccordingtowhichthereisnomoretomatterthanthebarepropertyofbeingasupport.Heisdiscussing,thatis,theideaofmaterialityassuch.Thisisnotthesameasdiscussingthequestionofthenatureofmatter.Lockehasusuallybeentakenasclaimingthatmatterconsistsofsubstratumandprimaryquali-ties.Berkeleyshowsthatwecannotmakemuchsenseofthesubstratumelement,butonemighttrytogetbyonthequalitiesalone.ThisbringsustoanotherbatteryoffamousBerkelianargumentswhichareintendedtoshowthattherecannotbeextra-mentalqualities.Thesecanbestbeapproachedthroughtheclaimthatanideacanbelikenothingbutanidea;whichmeansthatnothingexternalcouldbelikeafeatureofamentalimage;whichmeansinturnthatwecouldhavenopositiveideaofanyexternalquality.3Qualities,IdeasandSensationsTheclaimthatanideacouldbelikenothingbutanideahasbeengivenfairlyshortshriftonthegroundsthatitissimplyquestion-beggingagainstrepresentationalism(Berkeley1710:§8).Whatismore,anyforcethatitseemstopossesscomesfromtheconfusionarisingfromLocke’sextensionoftheordinarynotionof‘idea’.When‘idea’meanssomethinglike‘concept’or‘thought’thenitdoesseemtobeacategorymistaketohavereplicasoftheminthephysicalworld,butonceideaisextendedtoincludesense-contents,theissueseemsquitedifferent.Buttheargumentcanbeexpandedandmadequiteplausible,asfollows:1Asensationcouldbelikenothingbutasensation;mostespecially,nothingextra-mentalcouldbelikeasensation.2Thereisnodistinctioninkindbetweensensationsandideas.Therefore,3Anideacouldbelikenothingbutanidea;mostespecially,nothingextra-mentalcouldbelikeanidea.Thefirstpremiseisentirelyintuitive:theideaofanextra-mentalpainoritchseemstomakenosenseand,hence,thesuggestionthattherecouldbesuchcouldplausiblyberegardedasacategorymistake.ThesecondpremiseiswhatBerkeleyisdefendingwhenhearguesthatthefeelingofheatandacertainkindofpainarecontinuous,sothatthelatterisnothingmorethananintenseversionoftheformer(Berkeley1713:I,176ff.).700\nBERKELEYTwokindsofobjectionaremadetothisargument’ssecondpremise.Oneistoaffirmthatwecandistinguishtheperceptionofheatfromthefeelingofpain,evenwhentheyoccurtogether,sothattheyarenotmerelydifferencesofdegree.TheotheracceptsthatBerkeley’sargumentistrueforthesensesthatoperatebycontact,butdeniesthatitappliestovision.Firstly,Ishallprovideaclarificationoftheargument.Berkeleydoesnotneedtodenythatthepainfulaspectofasensationisdistinguishablefromitshotaspect;heneedsonlytoaffirmthattheyareaspectsofthesamething,suchthatonecannotexistinadifferentwayfromtheother.Thetimbreandpitchofsounds,orthebrightnessandhueofcolours,wouldbeexamplesofthis.Itisnotnecessary,therefore,toclaimthattheydifferonlyindegree,andBerkeleycanallowthatthepainisanextrafeatureoftheheatphenomenonitacquiresthroughintensity.Theissuethenconcernswhetherthepainfulness–oranysimilarsensationalfeaturethatcanaccompanytactileexperience–isexperiencedascausedbybutseparablefromsomeotherfeatureoftheexperience,orasjustanotheraspectofit.Itseemstometobehighlyplausibletoclaimthat(1)thepurelysensationalelementintactileexperienceand(2)thosepartsoftactileexperiencethoughttobecandidatesforbeingthedirectperceptionofsomethingexternalarenotseparable.Infact,thispointisbestbroughtoutbyconsideringavisualcase(Berkeley(1709)islargelydevotedtoshowingthatvision,initself,doesnotpresentexternality).IfIlookatanunpleasantlydazzlinglight,threethingsaretrue:(1)thelight,includingitsdazzlingness,seemstobeoutsideme;(2)Iexperiencetheunpleasantnessasakindofsensation,thatis,assomethinginternaltome,afeatureofhowIamaffected;(3)theunpleasantnessoftheexperienceisnotseparablefromthedazzlingqualityofthelight.Theparadoxisthattheunpleasantsensationalqualityisbothanexternalfeatureoftheworldandasensationalfeatureofmyexperience.Itismoreclearphenomeno-logicallythaninthecaseofheatandpainthatlight,brightnessandunpleasantdazzleareonlyamatterofdegree,forthereisnotarecognizableseparatesensationofpain,distinctfromthedazzle.Someonemighttryarguingthatthereisadifferencebetweenthedatumofextremebrightnessandthefeelingofitsunpleasantnessandthatthelatterismerelycausedbytheformer,notanaspectofit.Thatthisisnotplausibleisbroughtoutbythefollow-ingthoughtexperiment.Imaginethatsomeoneblindfrombirthweregivenadazzling-lightexperiencebydirectstimulationofthecortex:doyouthink(1)thattheywouldexperiencethelightasputativelyexternalormerelyasasensationnolessinternaltothemselvesthanabodilysensation?Assumingtheanswerto(1)isthattheywouldnotexperiencethelightasexternal,(2)doyouthinkthattheywouldexperiencetheunpleasantnessasotherthanafeatureofthebrightness?Theanswerseemstometobeplain,anditisthattheywouldthinkofthembothasintegralfeaturesofthesen-sation;thisstronglysuggestsaBerkelianinterpretation.Thisinterpretationisthatwhetherornotsomethingisexperiencedasexternalisamatternotofitsintrinsicphenomenologybutofhowitisinterpretedintoone’spictureoftheworld.Suchinterpretationdoes,inonesense,alterthephenomenology,butitdoesnotaltertheontologicalstandingofthecontent:learningtoseecoloursasifexternaldoesnotmakethemexternal.Thisseemstometobeastrongargumentfortheassimilationofideastosensationsand,therefore,forsayingthatnothingextra-mentalcouldbelikethem.Inreplyitmight701\nHOWARDROBINSONbearguedthatitonlyshowsthattheideasofsecondaryqualitiesareassimilabletosen-sations,notthatthoseofprimaryqualitiesare.Butevenifthisissoitdoesnotsubverttheargument.OneBerkelianresponsewouldbetopointoutthatthereisnoideaof,say,shapeindependentofthesecondaryqualitiesofagivensense(Berkeley1710:10).Sovisualshapeisinseparablefromcolourandtactileshapefromfeltsensation,§hencetheassociatedprimaryqualitycannotbeextra-mental.Evenifshapescouldexistindependentlyofthesesecondaryqualities,itisgenerallyagreedthatsuchprimaryqualitiesalonecouldnotconstituteaworldbecausetheyaretooformal–merelyshapeswithoutanyfilling.Somethinganalogoustoasecondaryqualityisnecessarytogivebodytothisconception.Althoughtheargumenthasonlyshownthatouractualsecondaryqualitiesareassimilabletosensations,itstronglysuggeststhatanyinten-sivequalityofakindlikethatofoursecondaryqualitieswouldbeofthatkindbybeingamodeofexperienceandthat,therefore,theideaofsomethinglikeasecondaryquality,yetextra-mental,isofdubiousintelligibility(Foster1982:84–107).Soallideasareessentiallymodesofsensation,andtheargumentissound;nothingextra-mentalcouldbelikeanidea.Thisdiscreditsthesuggestionthatanexternalworldcouldpossessorconsistofanythinglike,orbeanalogousto,thequalitiesweexperience.WestartedbylookingatBerkeley’sattackonabstractideas,bothtounderstanditinitselfandtoseehowitrelatedtohisimmaterialism.WediscoveredthatBerkeley’santi-abstractionismasheconceivesofitisatissueofconfusions,butitiscloselyasso-ciatedwithargumentsthatareverypowerful.Iftheimagisttheoryofthoughtistrans-formedintoanimagistcriterionforpositiveempiricalcontentandcombinedwithargumentsthatsuggestthatallmonadicqualitiesaresensational(‘anideacanbelikenothingbutanidea’),thenBerkeley’sargumentsshowthatitisdoubtfulwhethermind-independentmattercouldpossessanyintrinsicpropertiesatalland,hence,whethersuchathingcouldexist.Therearestandardconceptionsofmatteraccordingtowhichitdoesnotpossessanyintrinsicmonadicproperties,butonlyrelationalanddispositionalones.Theideathatitconsistssolelyofpowers,forcesorfieldsissuchaconception.SoisaLockeanviewthatmatterisjustimpenetrabilitydistributedinspace.Thereisnoroomtoconsiderthishere,butreadersmightwishtolookatrecentdiscussions(Foster1982:67–72;Robinson1982:108–23).4Conceivability,PerceivabilityandIntrinsicPropertiesBerkeley’sMasterArgumentInwhathasbeencalledhis‘masterargument’Berkeleyclaimsthatwecannotconceiveofunperceivedthings.Butsayyou,surelythereisnothingeasierthantoimaginetrees,forinstance,inapark,orbooksexistinginacloset,andnobodybytoperceivethem.Ianswer,youmay[say]so,thereisnodifficultyinit.Butwhatisallthis,Ibeseechyou,morethanframinginyourmindcertainideaswhichyoucallbooksandtrees,andatthesametimeomittingtoframetheideaofanyonethatmayperceivethem?Butdoyounotyourselfperceiveorthinkof702\nBERKELEYthemallthewhile?Thisthereforeisnothingtothepurpose.Itonlyshowsthatyouhavethepowerofimaginingorformingideasinyourmind;butitdoesnotshowthatyoucanconceiveitpossible,theobjectsofyourthoughtmayexistwithoutthemind.Tomakeoutthis,itisnecessarythatyouconceivethemexistingunconceivedorunthoughtof,whichisamanifestrepugnancy.Whenwedoourutmosttoconceivetheexistenceofexternalbodies,weareallthewhileonlycontemplatingourownideas.(Berkeley1710:§23)Theremustbesomethingwrongwiththisargumentbecause,byparityofreasoning,onecouldprovetheimpossibilityofconceivingofideasinotherpeople’smindsorofideashadatothertimes:onecouldprove,thatis,solipsismofthepresentmoment.Theargu-mentmovesfromthepremisethatwhateverIthinkof,thethoughtisinmymind,totheconclusionthatwhatIthinkofisinmymind.Byparityofreasoning,whenIformtheideaofanideainsomeoneelse’smind,thisthoughtisinmymind:andifIformthethoughtofanideainmyownmindatsomeothertime,thisthoughtisinmymindnow:sothesethoughtscanonlybeaboutobjectsthatarenowinmymind.Thispatternofargumentrestsondenyingthatideasareintrinsicallyintentional–thatis,ondenyingthattheymakeessentialreferencebeyondthemselves.Thisfollowsfromtreatingthemsimplyasimages,foranimageisonlyapatternofsensiblequalitiesandsuchqualitiesarenotaboutanythingbutthemselves.Supposethatweaccepttheintentionalityofthought,doesthisremoveallforcefromtheargument?Itdoesnot,ifoneacceptssomeconnectionbetweenwhatwecanCONCEIVE(pp.726–33)andwhatwecanimagine.Wheneverweimagineanobject,weimaginewhatitwouldbelikeperceivedfromacertainviewpoint:wedonotimaginewhatitislikeinitself,inabstractionfromsomepossibleperception.Anyattempttoconceiveoftheobjectasitisindependentlyofsomepossibleperceptualperspectivewouldhavetobemoreabstractthanaconcreteimagination.Asaphysicalobjectisanempiricalobject,withempiricalproperties,itmightseemthattherewassomethingpeculiarabouttheideathatitpossessesamodeofexistencethatcouldnotberepresentedimagistically,thatis,inaforminwhichthoseempiricalpropertiesareactualized.Thenaturalreplytothisisthatagoodperspectiveonanobjectenablesonetoformaconceptionoftheobjectasitisinitself.Thisismostsimplyrepresentedbyaclearviewofaflatsurface,whichenablesonetoseeitnotmerelyfromaperspectivebutasitisinitsownplane.Ourvisualperceptioncomestobestructuredinthreedimensions,soitshavingaperspectivedoesnotforceusintohavingamerelyabstractconceptionoftheobjectinitsownspace,asitwoulddoifvisionweretwodimensionalanddistancewasonlyinferred.AnotherofBerkeley’sargumentscanbedeployedhere.Theconceptionofhowtheobjectisinitself,independentofanyperceptualperspective,includestheideaofwhatsizeitisinitself.Berkeleyarguesthatthereisnosuchnotionofsize:toamiteevery-thingislargerthantoahuman,whichistosaythat,ultimately,sizeisnotseparablefromtheproportionofasensoryfieldthatanobjectoccupies(Berkeley1713:I,188).Itis,ofcourse,noanswertothistosaythatthereareobjectiveunitsofmeasurement–thefoot,theinch,themetre–forwhatkindofcontentisultimatelygiventotheseexceptbytheirconnectionwithperception?Theobviousansweristhatsizeispurely703\nHOWARDROBINSONrelative.Onceoneunderstandsthecomparativesizeofsomethinganinchlongwhenconstrastedwithsomethingafootlong,andonehasthesesetinacontextrelatingthemtothesizeofhumanbeings,cats,mountains,thedistancetotheSun,andsoon,thenthereisnothingmoretoknowaboutsize.Itis,afterall,generallyacceptedthatitmakesnosensetothinkofeverythingdoublinginsize,andthiscouldonlybebecausethereisnothingmoretosizethanwhatcanbeanalysedrelationally.Thisanswer,thoughplausible,isnot,Ithink,adequate.SPACE(p.729)isnotpurelyrelational,inthesensethatitisnotsimplyarelationofnumericalmagnitudes;anintuitivenotionofspaceorextendednessisalsorequiredtogivetheseformalrelationsanempiricalinterpretation(Foster1982:73–88).Thingswouldlooklargertoanintelligentmitethantoahumanbeing,eventhoughtheyagreedinalltheirrelationaljudgements.Themiteandthehumandisagreepreciselyinrespectoftheintuitiveelementthatisessentialfortheempiricalinterpretationofspace.Theyare,therefore,disagreeingaboutsomethingthatisessentialtothenatureofexperiencedspace.Thisdisagreementcouldbedefusedonlyiftheintuitiveelementcouldbewhollyprojectedontoandabsorbedbytherelationalphysicalstructure,sothattheonlywayofcharacterizingthesizeoftheintuitive,phenomenalcomponentinspatialexperiencewasintermsofthephysicalmeasurementsthatareessentiallyrelational.Thatthisisnotsofollowsfromthefactthatobjects‘lookhuge’tothemite.Thesituationisasfollows.Physicalsizecanbetreatedaspurelyrelationalbutonlyatthecostofnotbeingrealistaboutit.Anyrealspacemustincludetheintuitiveelementaswellastherela-tionofmagnitudes,andtheintuitiveelementisnotseparablefromhowitseemstoacertainkindofsubject,thisnotbeingcollapsibleintotherelational.ItwouldbewrongtothinkthatEinstein’srelativitytheorycanbeinvokedagainstBerkeleyhere.Einsteinisconcernedwiththevariabilityofsizeindifferentframesofreference,whichmeansframesofreferencethatareinmotioninrelationtoeachother.Inasense,theBerkelianpointisanextensionofEinstein’s,foritaddsafurtherframeofreferencetothedeterminantsofsize,namelythephenomenalfieldofthesubject;thoughitisnotadeterminantofphysicalsize,whichisanabstractionfromthephenomenal.5FromPhenomenalismtoTheismRatherthantryinginthischaptertogiveaguidedtouraroundthewholeofBerkeley’ssystem,Ihaveconcentratedonshowingtotheirbesteffecthisargumentsfordenyingtheexistenceofmind-independentmatter,foritisthisfeatureofhissystemthatisofmostfundamentalinterestandtowhichthereismostnaturalresistance.Myapproachsofar,therefore,emphasizesthenegativethrustofhissystemandtherebyignoresthethingsthatmainlydistinguishitfromthescepticalPHENOMENALISM(pp.53–4)ofHume.TorectifythisimbalanceIshallendbysketchingtherationaleformovingfromimma-terialismtotheism.Thismovehastwoelements,bothofwhichdistinguishBerkeleyfromHume.Firstly,thereisthenotionofspiritwhichisessentialforBerkeley’sconceptionsoftheSELF(pp.650–2)andofGOD(chapter15);secondly,therearehisreasonsforclaimingthatourexperienceiscausedbyGod.Althoughhehadoriginallytoyedwithconceivingofthe704\nBERKELEYselfasmerelyacollectionofideas,thusanticipatingHume,bythetimeofwritingthePrincipleshehaddecidedthatweareawareofourselvesasactiveaswellasbeingawareoftheobjectsontowhichouractivitiesaredirected.Becauseideasarepassiveandtheselfisactivewecannothaveanideaofspirit.Wecan,however,understanditbecauseweareconsciousofourselvesinouractivities,thoughnotasanobjectofthem;hecallsthishavinganotion,ratherthananidea,ofourselves:‘Isay...thatIhaveanotionofspirit,thoughIhavenot,strictlyspeaking,anideaofit.Idonotperceiveitasanideaorbymeansofanidea,butknowitbyreflexion’(Berkeley1713:III,233).Inresponse,Hylas,amaincharacterinBerkeley’sThreeDialogues,immediatelysug-geststhatifspiritcanbeletinwithoutacorrespondingidea,socanmatter.Buttheasymmetrywithmaterialsubstanceispreserved;wecannothaveanotionofmatterbecausethatisnotdirectlyavailabletoconsciousness,evenintheformofreflection.Sotheassociationbetweenmeaningfulnessandconsciousacquaintanceisnotlost.Havingestablishedtheontologyofspiritsandideas,allthatisneededistoshowthatGodisresponsibleforthoseideasthatarenottheproductsofourimagination–thatis,forthosethatconstitutethephysicalworld.WecanconceiveofGodbecauseGodisspiritinthesamesenseasweare,withallthelimitationsandimperfectionsthoughtaway.Itisnaturaltobelieve–andBerkeleybelieves–thatourexperiencehassomekindofsystematiccause.Thesimplesthypothesisisthatitiscausedbymind-independentphysicalobjects,butthisexplanationhasbeenruledoutbytheargumentsforimma-terialism.Thisleavesuswithachoicebetweenpostulatingsomekindofimmaterialcauseorofabandoningthenaturalbeliefthatthereisacause.ThelatterisHume’sscepticalphenomenalism.Berkeleychoosestheformer.Becausewedonotseemtobethecausesofourownexperiences,eitherassingleindividualsorasaspecies,weareforcedtopostulatesomeotherimmaterialagentandBerkeleynaturallycastsGodinthisrole.Onecould,inprinciple,optforapaganBerkelianism,asdidW.B.Yeats,andpostulateawholetribeofgodsanddemons,butBerkeley’sIrishnesswasofamorerationalistickindthanYeats’sandhestucktotheorthodoxpicture.TherealrivaltoBerkeliantheismcomes,onceidealismhasbeenestablished,fromthedoubleHumeanchallenge:firstly,thatexperienceingeneralrequiresnoexplanation;and,secondly,thatnothingcouldconstitutesuchanexplanationevenifitwererequired.ForHume,some-thingrequiresanexplanationonlyifitsoccurrencewithoutanexplanationisimprob-able,andsomethingisimprobableonlyifexperiencesuggeststhatitwouldnotnaturallyhappen.Consequently,howanaeroplanecanflyrequiresanexplanation.Thatexperience(seeminglyofaphysicalworld)continuestoflowinitsnormalcourse,however,isthestandardrunofeventsandneedsofnoexplanation:thatisjustthewaytheworldis.Theopinionthatthewaytheworldisneedsexplanationcannotbebasedonexperience,forexperiencecannotshowhowthingsinfactusuallyaretobeimprob-able.Forsuchthingstobejudgedimprobable,onemustemployastandardofaprioriPROBABILITY(pp.308–10),andHumedidnotbelievethattherewasanysuchstandardforempiricalmatters.Furthermore,askingforanexplanationinthiscontextisaskingwhatbroughtabouttheworldofexperience.Thisinvolvesaconceptofcausationasrealpowerorforce,whichHumerejects.ForHume,acausalexplanationoftheworldmerelyinvolvesmarshallingphenomenaintoconstantlyconjoinedpairingssothatonecanhavethesatisfactionofseeingtheworldasanorderedstructure.Allthisisnomore705\nHOWARDROBINSONthantheregimentingofactualandpossibleexperiencesandhasnothingofageneralkindtosayaboutwhatbringsexperienceabout;itiswhollyinternaltotheworldofphenomena.Berkeley’spositionismorecommonsensicalthanHume’sandisexpressedinsec-tions25–6ofthePrinciples:25Allourideas,sensationsorthethingswhichweperceive,bywhatsoevernamestheymaybedistinguished,arevisiblyinactive,thereisnothingofpoweroragencyincludedinthem...26Weperceiveacontinualsuccessionofideas,someareanewexcited,othersarechangedortotallydisappear.Thereisthereforesomecauseoftheseideaswhereontheydepend,andwhichproducesandchangesthem.Thatthiscausecannotbeanyqualityorideaorcombinationofideas,isclearfromtheprecedingsection.Itmustthereforebeasubstance;butithasbeenshewnthatthereisnocorporealormaterialsubstance:itremainsthere-forethatthecauseofideasisanincorporealactivesubstanceorspirit.(Berkeley1710:25–6;seealsoBerkeley1713:III)§ThereisnothinginBerkeleytosuggestthathewasawareofthepossibilityofscepti-cismaboutaprioriprobabilities.Hetreatsitasobviousthatexperiencemustcomefromsomewhere.HehadentertainedHumean-stylescepticismaboutcausalefficacyandheldtosuchaviewconcerningcausationintherealmofideasand,hence,thephysi-calworldasheunderstoodit.Buthebelievedthatreflectionmadeusawareoftherealcausalefficacyofthewill.Bysuchefficacy,heheld,God’swillproducesinourmindstheideaswhichconstitutetheexternalworld.FurtherReadingUrmson’s(1982)excellentbookonBerkeleyinthePastMastersseriesisanelegantandbriefintroduction.Tipton(1974),Pitcher(1977),Dancy(1987),Grayling(1986)andWinkler(1989)aremoredetailedandmoreconcernedtoarguetheirowntheses,butallcanbeusedasgeneralguides.Bennett(1971)relatesBerkeleytoLockeandHumeinaphilosophicallystimu-latingway,thoughrathertoomuchtoBerkeley’sdisadvantageand,somethink,withtoolittleregardtothehistoricalcontext.ContemporarydefencesofroughlyBerkelianformsofidealismcanbefoundinJohnFoster’scontributiontoRobinson(1993),whichisamuchsimplerversionofthedifficultFoster(1982):therearedefencesalsoinRobinson’scontributiontoFosterandRobinson(1985)andinthefinalchapterofRobinson(1982).AnidealismthatowesmuchtoBerkeley,butwhichfinallytakesaHegelianform,isdefendedbySprigge(1983).ReferencesBerkeleyThestandardeditionis:Berkeley1948–55:WorksofGeorgeBerkeley,BishopofCloyne,7vols(editedbyA.A.LuceandT.E.Jessop).London:Nelson.Easilyavailableone-volumepaperbackeditions,containingatleastthePrinciplesandtheThreeDialoguesinclude:706\nBERKELEYBerkeley1962:ThePrinciplesofHumanKnowledgewithotherwritings(editedandwithanintroductionbyG.J.Warnock).Glasgow:Fontana.——1988:PrinciplesofHumanKnowledge/ThreeDialogues(editedandwithanintroductionbyR.Woolhouse).Harmondsworth:PenguinBooks.——1993:PhilosophicalWorks:includingtheworksonvision(editedandwithanintroductionbyM.R.Ayers).London:Everyman,Dent.——1995:PrinciplesofHumanKnowledgeandThreeDialogues(editedandwithanintroductionbyH.Robinson).Oxford:OxfordUniversityPress.Withoneexception,myreferencestoBerkeley’stextsareeithertoThePrinciplesofHumanKnowl-edge(1710)ortoThreeDialoguesbetweenHylasandPhilonous(1713).Referencestotheformergivethesectionnumberalone,forthesectionsaregenerallyshorterthanapageandthisfacili-tatestheuseofanyedition.Thereisnosuchshort-cutwiththeDialogues,andhereIcitethenumberofthedialogueandthepagereferenceinvolumetwoofLuceandJessop’sstandardeditionofBerkeley’sworks.ThereisonereferencetoAnEssayTowardsaNewTheoryofVision(1709).Locke,AnEssayConcerningHumanUnderstanding(1689)andHume,ATreatiseofHumanNature(1739)arealsomentioned.OtherwritersAyers,M.1991:Locke,2vols.London:RoutledgeandKeganPaul.Bennett,J.1971:Locke,Berkeley,Hume:CentralThemes.Oxford:OxfordUniversityPress.Dancy,J.1987:Berkeley:AnIntroduction.Oxford:Blackwell.Flage,D.E.1987:Berkeley’sDoctrineofNotions.Beckenham:CroomHelm.Foster,J.1982:TheCaseforIdealism.London:RoutledgeandKeganPaul.Foster,J.andRobinson,H.(eds)1985:EssaysonBerkeley.Oxford:OxfordUniversityPress.Grayling,A.C.1986:Berkeley:TheCentralArguments.London:Duckworth.Lockwood,M.1989:Mind,BrainandtheQuantum:TheCompound‘I’.Oxford:Blackwell.Martin,C.B.andArmstrong,D.M.(eds)1968:LockeandBerkeley.London:Macmillan.Pitcher,G.1977:Berkeley.London:RoutledgeandKeganPaul.Robinson,H.1982:MatterandSense.Cambridge:CambridgeUniversityPress.——-1993:ObjectionstoPhysicalism.Oxford:OxfordUniversityPress.Sosa,E.(ed.)1987:EssaysonthePhilosophyofGeorgeBerkeley.Dordrecht:D.Reidel.Sprigge,T.1983:TheVindicationofAbsoluteIdealism.Edinburgh:EdinburghUniversityPress.Tipton,I.C.1974:Berkeley:ThePhilosophyofImmaterialism.London:Methuen.Turbayne,C.(ed.)1982:Berkeley:CriticalandInterpretiveEssays.Manchester:ManchesterUniversityPress.Urmson,J.O.1982:Berkeley.Oxford:OxfordUniversityPress.Warnock,G.J.1953:Berkeley.Harmondsworth:PenguinBooks.Winkler,K.P.1989:Berkeley:AnInterpretation.Oxford:OxfordUniversityPress.Yolton,J.W.1984:PerceptualAcquaintancefromDescartestoReid.Oxford:Blackwell.DiscussionQuestions1HowisBerkeley’sattackonabstractideasrelatedtohisargumentsforimmaterialism?2WhataretheadvantagesanddisadvantagesofBerkeley’sownaccountofthegeneralityofideas?3DoesrepresentativerealismofthekindfoundinLockereallyopenthewaytoscepticismabouttheexistenceeitheroftheexternalworldorofGod?707\nHOWARDROBINSON4Isthedoctrinethat‘anideacanbelikenothingbutanidea’betterthanquestion-beggingasanobjectiontorepresentativerealism?5Inwhatsense,ifany,isthephysicalworldmind-independent?6Shouldweagreethat‘tobeistobeperceived’?7Isittruethatapainfulheatandapleasantheatdiffernotinkind,butonlyindegree?Howcanonesetabouttryingtoanswerquestionsofthiskind?8Doesthefactthatwecannotimageaninstanceofaprimaryqualitywithoutasecondaryqualityshowthatitis–ormightbe–impossibleforaprimaryqualitytoexistwithoutasecondaryone?9Wouldsomeonewhohadgainedsightafterbeingblindfrombirthbeinapositiontorecognizeparticularshapesvisuallyonthestrengthofhavingpreviouslyfeltthem?Ifnot,woulditmeanthatvisualandtactileshapesaredifferentqualities?10Doesthefactthatwheneverweimagineaphysicalobjectweimagineitasitwouldappearfromaparticularperspectivemeanthatwecannotformaproperconceptionofhowanythingisinitself?11Musttherebeanintuitiveelementinanyadequateaccountofrealspace?12Howcanwedeterminewhetherexperienceasawholerequiresexplanation?13TowhatextentareBerkeley’sdoctrinesmattersofcommonsense?14Isitpossibletohaveanotionofspiritifwecannothaveanideaofit?Isthecasedifferentformatter?15IfGodisdirectlyresponsibleformyexperience,doIhaveanyreasontothinkthatthereareanyotheragents–humanminds,forexample–connectedwiththethingsthatIexperience?16Howdoesanidealistaccountofthephysicalworldaffectnaturaltheology?708\n31HumePETERJONESDavidHume(1711–76)iswithoutdoubtoneofthegreatestphilosopherstowriteinEnglish.Hiselegantstylecomplementedthedepthandoriginalityofhisphilosophicalthought.Therangeofhisworkiswide,butheisbestknowntodayforhisviewsoncau-sation,induction,perception,personalidentityandonthenatureofmorality.Heagreedwithhispredecessorsthatunderstandinghowandwhythingschangeistheonlywaytoexplainthepastandtoplanforthefuture,andthatonlyaknowledgeofcausescanhelptodispelourmostnaturalfears.Hebelieved,however,thatmanyhadmisunder-stoodtheprecisenatureofsuchknowledge,andhadthusfailedtobenefitfromit.Healsoheldthatitwasnotenoughtoexaminethenaturalworld;astheinvestigator,humanbeingssetouttoknow,butofteninignoranceoftheirowncontributionstoatask.Philosophyshouldbegin,therefore,withaninvestigationintothenatureofhumanity,forthiswouldalsoenableustounderstandthenatureandlimitsofourknowledge.Humearguedthatweareallgovernedmuchlessbyreasonthanhasoftenbeenclaimed,andaremotivatedessentiallybyourpassions.Moreover,ineverydaylifewemustlearntolivewithprobabilitiesratherthancertainties.Thepublic,andintenselypractical,benefitswhichcouldresultfromsuchanunderstandingareprofound.1BiographyDavidHumewasborninEdinburghin1711.HeenrolledattheUniversityofEdinburghattheageof11,andstudiedthereforthree,orpossiblyfour,years.Aswasusualatthetime,heleftwithoutgraduating.Afterabriefstudyofthelaw,heembarkedonalongandprivatestudyofphilosophy.From1734until1737helivedinFranceandwroteATreatiseofHumanNature(Hume1978a),whichwaspublishedanony-mouslyin1739–40.In1740,toboostdisappointingsalesofthebook,Humepublishedananonymoussummaryofitsmaintheses.Thedocument,nowsimplycalledtheAbstract(Hume1978b),emphasizedpreciselythosetenetshechosetodiscusswhenhelaterrewrotepartsoftheTreatise.Humewasover40yearsofagebeforeheobtainedhisfirstfull-timejob.Hehadtwicefailedtobecomeaprofessorofphilosophy,inEdinburghin1745,andGlasgowin1752.\nPETERJONESIn1741and1742hehadpublishedtwovolumesofessays,underthetitleEssays,Moral,Political,andLiterary(Hume1985a),andthereafterrewrotepartsofthefirstandthirdbooksoftheTreatise;theseappearedin1748and1751andarenowknownastheEnquiryConcerningHumanUnderstanding(Hume1975b),andtheEnquiryConcerningthePrinciplesofMorals(Hume1975c).In1752hepublished,underhisownname,averysuccessfulsetofessaysonpoliticalandeconomictopics,underthetitlePoliticalDiscourses(Hume1985b).Inthesameyear,hewasappointedKeeperoftheAdvocates’LibraryinEdinburgh.Usingtheirmagnificentcollection,whichlaterbecametheNationalLibraryofScotland,hepublishedbetween1754and1762thesixvolumesofhisHistoryofEngland(Hume1985c).whichwerealsoimmenselysuccessful.From1763until1766HumeservedinParis,firstasSecretaryandthenasChargéd’affairesattheBritishEmbassy.Hebecamefriendlywithalmostallthegreatphilosophersandwritersoftheday,suchasDiderot(1713–84),D’Alembert(1717–83)andRousseau(1712–78),buttotheirmutualregrethenevermetVoltaire(1694–1778).OnreturningtoLondonHumebecameUnderSecretaryofState,handlingdiplomacywithforeignpowerstothenorthofFrance,includingtheRussiaofCather-inetheGreat.HewasalsoinchargeofhomeaffairsforScotland,andofficialpatronoftheScottishChurch.In1769HumeretiredtoEdinburgh,renewedoldfriendships,particularlywithAdamSmith(1723–90),andplayedhosttoalltheimportantvisitorstothecity,suchasBenjaminFranklin(1706–90).InretirementHumecontinuallyrevisedneweditionsofhisHistory,andworkedonhisDialoguesConcerningNaturalReligion,modelledonCicero(106–43BC)andpublishedposthumouslyin1779.2PhilosophyGreatphilosophersveryoftenarenotthefirstpeopletoexpresstheideasforwhichtheythemselvesbecomefamous;theirdistinctionliesintheroutestheytakeinwhattheyseeasthecontextoftheirthought,inthesynthesestheyoffer,andaboveallinthepreciseargumentstheyofferfortheirconclusions.Moreover,allphilosopherswriteinthefirstplacefortheircontemporaries,andareinfluencedbyotherthinkerswhoseworkmaybeunknowntolatergenerations.HumehimselfwasprofoundlyinfluencedbytheRomanpoliticianMarcusTulliusCicero,andbytheseventeenth-centuryFrenchphilosopherNicholasMalebranche(1638–1715).Moreover,thehistoricalcontextinwhichhelivedwasverydifferentfromourown.Forexample,whenhewaswriting,nothingwasknownaboutlatentheatorthecompositionofairandwater,aboutanypowersotherthanlightandheat,aboutanimalgenerationorevolution,abouttheageoftheEarthorthesizeoftheuniverse;andsuchcommonplacetermsas‘police’,‘mag-istrate’,‘government’,‘agriculture’,‘art’and‘city’carrieddifferentmeaningsfromtoday.Ifconceptsliketheseirradiateourthoughttoday,thenweneedtoidentifywhatwemustunthinkinordertounderstandHume’sthought.Itmustberemembered,therefore,thatalthoughmanycontemporaryproblemshaveaHumeanancestry,inaveryobvioussenseHumewasnotdiscussingtwentieth-centuryissues.710\nHUME2.1CausationHumewasinterestedintheprocessesbywhichweacquireKNOWLEDGE(chapter1);theprocessesofperceivingandthinking,offeelingandreasoning.Herecognizedthatmuchofwhatweclaimtoknowderivesfromotherpeoplesecondhand,thirdhandorworse;moreover,ourperceptionsandJUDGEMENTS(pp.726–9)canbedistortedbymanyfactors–bywhatwearestudying,aswellasbytheveryactofstudyitself.Themainreason,however,behindhisemphasison‘probabilitiesandthoseothermeasuresofevi-denceonwhichlifeandactionentirelydepend’(Hume1978b)isthis:itisevidentthatallreasoningsconcerningmatteroffactarefoundedontherelationofcauseandeffect,andthatwecanneverinfertheexistenceofoneobjectfromanotherunlesstheyareconnectedtogether,eithermediatelyorimmediately.(Ibid.)Whenweapparentlyobserveawholesequence,sayofoneballhittinganother,whatexactlydoweobserve?Andinthemuchcommonercases,whenwewonderabouttheunobservedcausesoreffectsoftheeventsweobserve,whatpreciselyarewedoing?QuestionsaboutCausalityInconsideringcausalityHumeseparatedimportantlydifferentissues:1Questionsaboutthecausalrelation:whatexactlyisitforoneeventtobethecauseofanother?Whatistheevidenceofcausalrelationsbetweenevents?2Aquestionaboutthecausalinference:whatentitlesustoinferfromtheoccurrenceofoneeventthatanothereventisitscauseoritseffect?3Questionsaboutthecausalprinciple:whydoweinsistthateveryeventhasacause,andwhatentitlesustodoso?4Threeextraquestionsaboutnecessity:(4a)whydoweinsistthatwhatdistinguishesagenuinecausalrelationfromamerelycasualrelationisthenecessityofthatparticularpairofeventsoccurringtogether?(4b)whatentitlesustoinferthatifoneeventisthecauseoreffectofanother,thenitmusthaveoccurredwithit,orimmediatelybeforeorafterit?(4c)whydoweinsistthateveryeventmusthaveacause?Humerecognizedthatanotionof‘must’ornecessityisapeculiarfeatureofcausalrelations,inferencesandprinciples,andchallengesustoexplainandjustifythenotion.Hearguedthatthereisnoobservablefeatureofevents,nothinglikeaphysicalbond,whichcanbeproperlylabelledthe‘necessaryconnection’betweenagivencauseanditseffect;eventssimplyare,theymerelyoccur,andthereisno‘must’or‘ought’aboutthem.However,repeatedexperienceofpairsofeventssetsupthehabitofexpectationinus,suchthatwhenoneofthepairoccursweinescapablyexpecttheother.Thisexpectationmakesusinfertheunobservedcauseorunobservedeffectoftheobservedevent,andwemistakenlyprojectthismentalinferenceontotheeventsthemselves.Thereisnonecessityobservableincausalrelations;allthatcanbeobservedisregularsequence:thereisnecessityincausalinferences,butbecauseinferencesarewhatobserversdo,wecansaythatthisnecessityisonlyinthemind.OncewerealizethatcausationisaRELATION(p.800)betweenpairsofevents,wealsorealizethatoftenwe711\nPETERJONESarenotpresentforthewholesequencewhichwewanttodivideinto‘cause’and‘effect’.Ourunderstandingofthecausalrelationisthusintimatelylinkedwiththeroleofthecausalinference,becauseonlycausalinferenceentitlesusto‘gobeyondwhatisimmediatelypresenttothesenses’(Hume1978a:I,III,ii).Butnowtwoveryimpor-tantassumptionsemergebehindthecausalinference:theassumptionthat‘likecauses,inlikecircumstances,willalwaysproducelikeeffects’,andtheassumptionthat‘thecourseofnaturewillcontinueuniformlythesame’–or,briefly,thatthefuturewillresemblethepast(Hume1978b).Unfortunately,thislastassumptionlackseitherempiricaloraprioriproof;thatis,itcanbeconclusivelyestablishedneitherbyexperiencenorbythoughtalone.Humefrequentlyendorsedastandardseventeenth-centuryviewthatallourideasareultimatelytraceable,byanalysis,tosensoryimpressionsofaninternalorexternalkind.Accordingly,heclaimedthatallhisthesesarebasedonEXPERIENCE(p.52),understoodassensoryawarenesstogetherwithmemory,sinceonlyexperienceestab-lishesmattersoffact.Butisourbeliefthatthefuturewillresemblethepastproperlyconstruedasabeliefconcerningonlyamatteroffact?AsBertrandRussellremarked,earlierthiscentury,therealproblemthatHumeraisesiswhetherfuturefutureswillresemblefuturepasts,inthewaythatpastfuturesreallydidresemblepastpasts.Humedeclaresthat‘if...thepastmaybenoruleforthefuture,allexperiencebecomesuselessandcangiverisetonoinferenceorconclusion’(Hume1975b:IV).Andyet,heheld,thesuppositioncannotstemfrominnateideas,sincetherearenoinnateideasinhisview,norcanitstemfromanyabstractformalreasoning.Foronething,thefuturecansurpriseus,andnoformalreasoningseemsabletoembracesuchcontingencies;foranother,evenanimalsandunthinkingpeopleconducttheirlivesasiftheyassumethefutureresemblesthepast:dogsreturnforburiedbones,childrenavoidapainfulfire,andsoon.HumeisnotdeploringthefactthatwehavetoconductourlivesonthebasisofPROBABILITIES(pp.308–10);andheisnotsayingthatinductivereasoningcouldorshouldbeavoidedorrejected.Rather,heacceptedinductivereasoningbuttriedtoshowthatwhereasformalreasoningofthekindassociatedwithmathematicscannotestab-lishorprovemattersoffact,factualorinductivereasoninglacksthe‘necessity’and‘certainty’associatedwithmathematics.Hisposition,therefore,isclear:because‘everyeffectisadistincteventfromitscause’,onlyinvestigationcansettlewhetheranytwoparticulareventsarecausallyrelated;causalinferencescannotbedrawnwiththeforceoflogicalnecessityfamiliartousfromapriorireasoning,but,althoughtheylacksuchforce,theyshouldnotbediscarded.Inthecontextofcausation,inductiveinferencesareinescapableandinvaluable.What,then,makes‘pastexperience’‘thestandardofourfuturejudgment’?Theansweriscustom:itisabrutepsychologicalfact,withoutwhichevenanimallifeofasimplekindwouldbemoreorlessimpossible.‘Wearedeter-minedbycustomalonetosupposethefutureconformabletothepast’(Hume1978b);nevertheless,wheneverweneedtocalculatelikelyeventswemustsupplementandcorrectsuchcustombyself-consciousreasoning.2.2Someimplications:freewill,historyandtheuniformityofnatureHumewaswellawareofthepotentialimpactofhisanalysis,andhebeganhisexpla-nationofitsimplicationsbytryingtoreconciletheargumentsofthosewhoholdthat712\nHUMEallouractionsaredetermined,andthosewhoholdthatwearegenuinelyfreetodowhatwechoose.Thedeterministsareright,hesaid,toinsistthateveryeventhasacause,butmistakenintheviewtheytakeofnecessity,failingtoseethatitbelongsonlytoaninferenceinthemind;advocatesoffreewill,ontheotherhand,arerighttosaythatresponsibilitymakessenseonlyifwearetrulytheagentsofouractions,butmis-takeninholdingthat‘free’inthissensemeans‘uncaused’.Hume’sviewisthatourACTIONS(comparepp.733–6),likeallotherevents,havecauses,namely,thechoicesanddecisionsthatimmediatelyprecedeouractions,andforwhichweacceptresponsi-bility.However,althoughthesementalhappeningsfunctionastheMOTIVES(p.673)ofouractions,theythemselveshavecauses,ofwhichwearenormallyunawareandforwhichwearenotheldresponsible.Hispointindiscussingthefree-willissueistoshowthathumanactionscanbeunderstood,andareproperlyunderstandable,intermsofwhathehassaidaboutcausationingeneral–wenoticepatternsofbehaviour,weinfermotivespreciselyinordertooffer‘interpretations’ofactions:weretherenouniformityinhumanactions,andwereeveryexperimentwhichwecouldformofthiskindirregularandanomalous,itwereimpossibletocollectanygeneralobservationsconcerningmankind,andnoexperience,howeveraccuratelydigestedbyreflection,wouldeverservetoanypurpose.(Hume1975b:VIII)SuchthoughtswereimportantwhenhecametoreflectonthenatureofHISTORY(chapter14).Here,however,twofeaturesofthediscussionshouldbenoticed.Firstly,humanbehaviouriscapableofcausalexplanation,andthereforealsoofprediction,hewassaying,albeitnotyetwiththesamedegreeofprecisionorcertaintythatweexpectinthenaturalsciences.Thesecondfeatureisrelatedtothis.Soimportantisourbeliefthatsimilarcausesproducesimilareffects,thatweholdtoitevenwhenwecannotfindtherequireduniformities;andwhenwefindsomethingpuzzlingorinexplicableweneithersupposethatthelawsofnaturehavechanged,northatwehaveencounteredacauselessevent.Absenceofadequateevidenceforacauseisnottakenasadequateevi-dencefortheabsenceofacause.ExternalObjectsIn‘Ofscepticismwithregardtothesenses’,Hume(1978a:I,IV,ii)discussedproblemsthathaveparticularlyinterestedtwentieth-centuryphilosophers,namely,thosecon-cerningourknowledgeofexternalobjectsandpersonalidentity.Humeregardedbothexternalobjectsandtheself,inanimportantsense,asimaginativeCONSTRUCTIONS(p.000).Thetexts,togetherwiththeirvariousmoderncommentaries,areunusuallycomplexandrewardcloseexamination.Everyoneassumesthattheworldgoesonaroundus,whetherweareawakeorasleep,unconsciousordead.Butthisviewistroublesometophilosopherswhoheldthatweareonlyreallyconsciousofourinternalperceptions,andthatultimatelysensoryexperienceconsistsofunconnectedatomicinstants(forinstance,Hume1978a:I,IV,vi).Humeasked‘whetheritbethesenses,reasonortheimagination,thatproducestheopinionofacontinu’dorofadistinctexistence’ofobjects(ibid.:I,IV,ii).Hearguedthatimmediatesenseexperiencescannotgenerateanynotionofunsensedobjects,and‘nevergiveus713\nPETERJONEStheleastintimationofanythingbeyond’themselves(ibid.:I,IV,ii).Thesolutiontosuchproblemsliesinthepoweroftheimaginationtodisguisethedisjointedandinterruptedcharacterofsensoryexperience,byunitingorbondingourideasofitinthemind(ibid.:I,IV,vi).‘Theopinionofthecontinu’dexistenceofbodydependsonthecoherenceandconstancy’(ibid.:I,IV,ii)ofsomeofourimpressions,namelythosewhichourmemoryassuresuspreservesomedegreeofresemblanceandrepetitionovertime;thisfeatureofourperceptionsisthenprojectedontotheexternalworld.Whenwedo‘inferthecon-tinu’dexistenceofobjectsofsensefromtheircoherence’itis‘inordertobestowontheobjectsagreaterregularitythanwhatisobserv’dinourmereperceptions’(ibid.:I,IV,ii).PhilosopherssinceHumehavemademuchofthetheoreticalandpracticaladvantagesofprojectingontotheobjectswesenseagreatercoherencethanseemstobestrictlywar-rantedbythesensoryevidenceitself.2.3PersonalidentityHumeanchoredhisdiscussionofexternalobjectsinreflectionsonidentity,orsame-nessandchange:hetoldusthathisaimwas‘toprove,thatallobjects,towhichweascribeidentity,withoutobservingtheirinvariablenessanduninterruptedness,aresuchasconsistofasuccessionofrelatedobjects’(ibid.:I,IV,vi).Itishardlysurprising,then,thathestated:‘theidentity,whichweascribetothemindofman,isonlyaficti-tiousone,andofalikekindwiththatwhichweascribetovegetablesandanimalbodies’(ibid.:I,IV,vi).(Fiction,itmustbenoted,meant‘hypothesis’or‘invention’,inHume’sday.)Infact,however,Humeaskeddifferentquestionsaboutpersonalidentity,andcon-sequentlysaiddifferentthingsaboutit.Heasked:‘howfarweareourselvestheobjectsofoursenses’?Whatis‘thenatureoftheunitingprinciple,whichconstitutesaperson’(ibid.:I,IV,ii)?What‘impressiongivesrisetotheideaofself’(ibid.:I,IV,vi)?Heheldthat‘whatwecallamind,isnothingbutaheaporcollectionofdifferentperceptions,unitedtogetherbycertainrelations,andsuppos’dthofalsely,tobeendow’dwithperfectsimplicityandidentity’(ibid.:I,IV,ii;seealsoI,IV,vi).Helikenedthemindtoa‘repub-licorcommonwealth’(ibid.:I,IV,vi)butalsotoa‘theatre,whereseveralperceptionssuccessivelymaketheirappearance’;theidentity‘isnothingreallybelongingtothesedifferentperceptions,andunitingthemtogether;butismerelyaquality,whichweattributetothem,becauseoftheunionoftheirideasinimagination’(ibid.:I,IV,vi)–andthecapacityofthememorywhichissaidto‘notsomuchproduceasdiscover[thatis,reveal]personalidentity’(ibid.:I,IV,vi).CommentatorshavepointedoutthatHumeconfinedhismainlyfirst-personaccounttomentalphenomena,andinthesecontextsdidnotconsidertherolesofphysicalcriteriaorsocialbehaviourinourideaoftheself.Hisreflectionsinotherpassages,however,onmoral,socialandpoliticalstructuresandbehaviour,considerablyenrichedhisformalconsiderationofpersons.Nevertheless,hisdiscussionsofidentity,especiallyhisreflectionson‘numericalandspecificidentity’(ibid.:I,IV,vi),needradicalrevisionandsupplementation.2.4MiraclesandreligionThecasesthatmostclearlyconflictwithwhatHumehassaidaboutcausationarethoseofallegedmiracles.Hisowndiscussionwaseventuallypublishedinthe1748Enquiry714\nHUME(Hume1975b),butitisnotprimarilyaboutwhethermiraclesoccur–heratherbrusquelydefinesthemasimpossible;rather,itisaboutthenatureandreliabilityofevi-denceforfactualclaimsofanykindwhatsoever.Therecurrentterminhisdiscussionis‘testimony’,thatis,reportsofevidence.Humereadilyconcedesthatallofushavetorelyonreportsfromotherpeople;knowledgeisasocialphenomenoninthesensethatagreementaboutwhathappenedisimportantindefiningwhatwecountasknowledge.Moreover,noneofushasenoughfirst-handexperienceofourowntomakeveryextensiveclaims.Humeholdsthat,asageneralrule,testimonyisreliable,andhedoessofortworeasons:ourmemoriesaremoderatelyreliable,andhumanshaveanaturalinclination,stronglyre-enforcedbysocialsanctions,totellthetruth,andtofeelshameifdetectedinfalsehood.Thesegeneralassurances,however,cannothelpusinparticularcasesofdoubt,soHumesug-geststhatinweighingtheprobabilityofevidencethecontentandthecontextofwhatawitnesssaysmustbothbeexamined.Weneedtoconsidernotonlythetypeandquan-tityofevidence,butalsothemannerandmotiveingivingit.Intheendwemustrelyon‘experienceandobservation’.Butsurelywemustbetoldwhoseexperienceistocount,andwhyandhowthetestsofevidencearetobeperformed.Exhypothesi,wecannotappealtoourownexperienceinordertotestreportsbyothersofexperiencesunlikeourown;andinweighingwhatpeoplesayweneitherappealindiscriminatelytoanyone,norrequiretheassentofeveryone.Humereliesonanotionofexperts,orqualifiedobservers,towhomwecanappealforhelp.Typicallywelearnwhocountastheexpertswhilelearningtheproceduresformakingreportsofthekindinquestion;thelearningmaybequiteinformal,evenunconscious,andtheexpertsrangefromthesublimetotheridiculous–atitslowest,anyonemightcountasanexpertwhoseemstoknowmorethanoneself.Parentsandteachersarethefirstexpertschildrenencounter,buttheirownfriendsalsoplaythisroleforchildrenonmanyoccasions.Humeusesthenotionofexpertsandagreementtosecureboththesocialdimensionofourjudgements,andtheirobjectivity–althoughthattermwasnotusedintheeighteenthcentury.Humetakesstoriesofmiraclestobereportsofuniqueevents,thecauseofwhichisallegedlyknownbyinference,notbyexperience,eventhoughnoneoftheconditionsforestablishinganinferredcauseissatisfied.Weareentitledtoinferacause,heholds,withagreaterorlesserdegreeofprobability,whentheeventtobeexplainedresemblesotherswhichwehavedirectlyexperiencedinconjunctionwiththeircauses.Manyquestionsarisehere.Howmuchresemblanceisneededtowarrantsuchaninference?Whatpropertiesorcharacteristicsisoneentitledtoassigntosuchaninferredcause?Hume’sextendedreflectionsonsuchissueswerepublishedposthumouslyin1779inhisbrilliantDialoguesConcerningNaturalReligion(Hume1947).ProtestanttheologiansinHume’sdaywereeagertoembracewhatlittletheyunder-stoodofNewtonianscienceanditsmethods,andtheypopularizedanancientargu-mentknownastheargumentfromdesign.ThiswassupposedtoestablishthenatureofGodbymeansofinferencesfromtheobservableworldaroundus–inHume’swords,itwas‘anargumentdrawnfromeffectstocauses’(Hume1975b:XI).Schematicallyitranlikethis:itisimpossibletodescribeallthediversephenomenaintheknownuni-versewithoutimplyingsomeoverallplanbehindit,whichcausallyexplainsthenatureofthephenomenaandtheirinterconnections;suchaplan,however,wouldneedtobe715\nPETERJONESofsuchcomplexityitself,thatitcouldbeattributedonlytosomebeingwithcapacitiesfarexceedinghumancapacities–inbrief,itcouldbelongonlytoGod.Sophisticatedversionsofthisargumentarestillpopular,butHumemaintainedthattheyinherentlylackadeterminatebaseinexperience,thattheyareuntestable,unrevisableandclaimadogmaticfinalityanduniquenessinimicaltogenuinescientificenquiry.Theneces-saryrelianceoninductiveargumentsinsuchenquiry,afterall,meansthatinanygivencasewecouldturnouttobemistaken.Noevent,consideredbyitself,pointsbeyonditselftoitscausesoreffects;thesehavetobediscoveredbyexperience,andwhatismore,nofeaturesofacauseneedresembleanyoftheeffect–theycouldbe‘totallydifferent’(ibid.:IV).Factualinferencesrestontheassumptionofsimilarityofcases,andcanbeassignedvariousdegreesofprobability.Intheabsenceofdirectexperienceweareallowedonlytopostulatewhatissufficienttobringabouttheallegedeffects.Humeisquicktoseethatthisfailstosatisfyhisowndefinitionofwhatacausalrelationis,andofhowwecanknowone,sincethecrucialnotionofnecessityislacking.Heconcludesthatallegedly‘unique’causescannotlegitimatelybeinferredfrom,orusedtoexplain,allegedly‘unique’effects.Whenuniquenessisclaimedforanevent,wecanneverknowwhetherwehaveadequatelyidentifiedorcircumscribedit.Inaddition,ifwehavenoexperienceofathing’sattributeswecannotpickitoutfromamongotherthings,andifwecannotdothisevenvaguely,wehavenomeanstoconfirmthatitexists.Itisuselesstopostulateacausewhoseattributes,bydefinition,areunknown.Finally,thereistoomuchpain,sufferingandchaosintheworldasweknowit,tojustifyanyinferencetoafullycompetentcauseofitall.2.5MoralphilosophyHumeregardedhisownmoralviewsasthoseofacommon-senseman,albeitwithastrictlysecularoutlook.Heheldthatahumanbeingcannotbeadequatelydefinedorunderstoodasanessentiallyrationalanimal,asARISTOTLE(chapter23)andsomanyotherphilosophersandtheologianshadproclaimed.Indeed,purethought,orabstractreasoningofthekindassociatedwithpuremathematics,isitselfinert,andcannotmotivateanyonetodoanything;allthatsuchthoughtcandoisenableustorecognizerelationsbetweenthings.Thisrecognitioncansomehowinfluencewhatdoesmotivateus,namely,oursentiments,whichareformsofthemostbasicdesirestoavoidpainandattainpleasure.IntheTreatiseHumedramaticallysummarizedhisviewthatthoughtplaysonlyasubordinateroleinthemorallifebydeclaringthat‘reasonis,andoughtonlytobe,theslaveofthepassions’(Hume1978a:II,III,iii).Ithelpsintheformula-tionofourgoals,butnotinourmotivationtowardsthem.Whenhecametorewritethispassagein1751,however,hetoneditdown:‘nodoubtreasonandsentimentconcurinalmostallmoraldeterminationsandconclusions...[but]itisprobable...thatthisfinalsentencedependsonsomeinternalsenseorfeeling,whichnaturehasmadeuniversalinthewholespecies’(Hume1975c:I).Forsuchviews,HumedrewinspirationfromShaftesburyandfromFrancisHutcheson,ProfessorofMoralPhilosophyinGlasgowuntilhisdeathin1746.Oneofthefeatureswhichmostdis-tinguishedthephilosophersoftheScottishEnlightenmentfromtheirsuccessorsintheFrenchandGermanEnlightenment,isanemphasisontheessentialpartplayedinourlivesbythepassions;theirinfluenceisabrutepsychologicalfact,tobeneithercon-716\nHUMEdemnednorcondoned,butafactneverthelessthattheoriesofhumanityandmoralityignoreattheirperil.Veryoften,asithappened,theFrenchandGermanwritersregrettedthefact,thoughtittobeavoidable,andproclaimedthesovereigntyofreason.Indecidingwhattodo,weallweighthepossibleconsequences,butthemeritofanaction,Humeholds,derivesfromthemotivebehindit.ItisworthrecordingthatHume’slanguageissometimesconfusing:heusestheterm‘sentiment’tomeanbothfeelingandthought,andthismakeshisclaimthatpassionsorsentimentsarethesolecausesofactionmoreplausiblethanitmightotherwiseappear.Hedrawsadistinctionbetween‘artificial’and‘natural’virtues.‘Artificial’virtuesarethosewhichwemightnowadaysdescribeassociologicallyconditioned,inthesensethattheyareproceduresweadopt,asindividualsorgroups,andoftenunselfconsciously,tohelpthingsrunsmoothlyandtotemperourself-seekinganddivisivetendencies.Forexample,oursenseofJUSTICE(chapter8)arises‘artificially,thoughnecessarily,fromeducationandhumanconven-tion’;indeed,atoneplace,Humestatesthat‘publicutilityisthesoleoriginofjustice’(Hume1975c:III).‘Natural’virtues,bycontrast,aresaidtobegroundedinouruni-versaltendencytoshare,quiteliterally,thefeelingsofothers.Such‘sympathy’(andthenotionisnotequivalenttothemodernideaofcompassion)ultimatelyexplainseventheartificialvirtues.Bothartificialandnaturalvirtues,then,areunderstoodasqual-itieswhichhumanbeingsfindusefuloragreeablebothintheirowncase,asownersofthosequalities,andinotherpeoplewhomtheyobserveasspectators.Hume’snotionof‘sympathy’provedfruitful,andwastakenupanddevelopedbyAdamSmith,aswastheideaoflearningtojudgeourselvesimpartially.2.6Art,tasteandaestheticjudgementHume’sbriefaccountofAESTHETICJUDGEMENTS(chapter7)buildsonwhathesaysaboutmoralgoodnessandmoraljudgement.Bytalkingaboutthethingthatcurrentlypleasesordispleasessomeone,wecanchangethatperson’sresponsetoit;thatis,wecanchangehowtheyperceiveit,thinkofit,orreacttoit,andinthiswayinducethemtofeeldifferentlyaboutit.Suchemphasisonfeeling,however,shouldnotbetakentoimplythatourdiscussionsareincurablypersonalorsubjective.Onthecontrary,Humetriestoshowthatagreementispossible,andofakindthatisquitestrongenoughtosatisfyourdeepestyearningforso-calledobjectivity.Threeconditionsmustbemet:firstly,theconventionsoflanguagemustbeobservedbyalldisputantsand,ifneces-sary,clarifiedintheparticularcontext;secondly,itmustbepossibletoestablishpubliclyshareableviewpointsandreferents;andfinally,wemustassumeasharedpsychologicalmake-upamongthedisputants.Personalidiosyncrasiesduetoageorbackgroundshouldberecognizedasidiosyncrasiesanddiscounted.Andtheoldsayingthattherecanbenodisputinginmattersoftasteshouldbeseenasbeingharmlessly,becauseonlytrivially,true.Iftheexerciseoftasteinvolvesnojudgement,then,indeed,therecanbenodisputeordiscussion;butifitdoesinvolvejudgement,asmostpeoplethink,thendiscussionisnotonlypossiblebutdesirable.Moreover,judgement,under-stoodaspublicdiscourse,isalwayscapableofreachingapragmatic,ifnotaformal,resolutionofdifficulties.Hume’sremarksonartmostlyoccurwhenheisdiscussingfeaturesofoursociallife.Thatiswhyheconsidersartfromthestandpointofhumanactions:theartististrying717\nPETERJONEStogetsomethingdone,andspectatorsaretryingtoestablishthewhyandthewhere-foreofwhattheartistdoes.Worksofart,inthisway,arepleasurablemeansofcom-munication,andtheiracceptancedependsontheirmakingsensetospectatorswhocanagreeintheirresponses.Humeheldthatasanhistoricalphenomenon,artoccursonlywhenapersonhastimetoturnasidefromthebrutenecessitiesofliving.Nevertheless,inthemoreleisurelycontextofHume’sownepoch,theartscouldbeheldtorefinethetemperandmakepeoplemoresociable,becauseanartisttryingtopleasespectatorshadtoconsidertheirneedsandinterests.Thispointhasanimportantconsequenceforcriticismofthearts:becauseartisahumanactivity,oneofthegeneralcriteriaforunderstandinganyhumanactionappliestoit–namely,knowledgeofhowtheagentviewedthecontextandappropriatenessofwhatheorshewasdoing.Soweshouldcon-siderworksofartintermsoftheaudienceforwhichtheywereintended,aswellasintermsoftheirinternalcharacter,theirgenre,styleandtradition.Humeheldthatonecanachievethe‘propersentiment’towardsmost,andcertainlytowardsthebest,worksofart.Therearetwoelementsinhisclaim.Tobecausallyaffectedbyawork’sproper-ties,wemustadopttherightframeofmindorviewpoint,aswellastherightphysicalstandpoint.Thesecondelementisevenmoreimportant.Whatdistinguishesaproperresponsefromamerelypassivereaction,aswhenweenjoybaskinginthehotsun,istheactivecontributionofthespectator.ThisinvolvesanINTERPRETATION(pp.247–50)ofthework’smeaningandvalue,overandaboveidentificationoftheaspectswefindaffecting;meaning,afterall,cannotbedetectedbythefivesensesalone,anymorethannecessity.Interpretationinvolvestheexerciseofthemind.Hume’sviewsonartandcriticismreceivetheirmostsustainedtreatmentinasingleessay,‘OftheStandardofTaste’,publishedin1757,andtheybelongtoaperiodwhenthemodernnotionofaestheticswasreceivingonlyitsfirsttentativeformulation.Publicconcertswerebeginningtotakeplace,easelpaintingwasbecomingpopular,andthefirstnovelswereappearing;exceptforthosewhotravelledtoEurope,however,mostpeoplesawveryfewpaintings,hadaccesstonomuseums,hadlittleknowledgeofGreek–asopposedtoRoman–sculptureandarchitecture,andknewnothingoftheartofothercultures.2.7ViewsonhistoryandpoliticsHume’sviewofthenatureofhistoricalwritingandunderstanding,whichbringstogetherhisthoughtsonevidenceandhumantestimony,aswellasonagencyandmoti-vation,restsontwofundamentaltenets:firstly,agentsactwithcertainintentions,butarenecessarilyignorant,whentheysetout,oftheoutcome;secondly,observersknowtheoutcome,butaredoomedmerelytoconjecturetheintentionsnecessaryforunder-standingit.Theunbridgeablegapbetweenforesightandhindsightofthe‘actor’and‘spectator’,asHumecalledthem,canbafflethemostdiligentenquirer.Oneimportantpointhereisthetensionbetweentheagent’spassions,whichalonecanmotivatehimorher,andthespectator’sreason,whichfunctionsinhisorherunderstanding:‘Mens’viewsofthingsaretheresultoftheirunderstandingalone:Theirconductisregulatedbytheirunderstanding,theirtemper,andtheirpassions’(Hume1985c:LIII).Specta-torsdohaveaccess,however,todatalogicallyunavailabletotheagent:inparticular,tothelonger-termconsequencesoftheagent’sactsandtothevaryingsignificance718\nHUMEascribedtothemfromdifferentperspectives,andinrelationtodifferentsetsofissues.Thejudgementsofhindsightaretransient,nevertheless,becausethepossibilityofreinterpretationiseverpresentinthelightofsucceedingevents.AcentralthemeofHume’shistoricalreflectionsconcernstheunavoidablestrugglebetweenLIBERTY(pp.258–61)andauthority.HeisprimarilyconcernedwithlibertyundertheLAW(chapter13)andwithconstraintsuponitexercisedbyvariouskindsofauthority;toalesserdegreeheconsiderslimitsonpersonallibertyimposedbyhabitoreducation.ForHume,adiscussionoflibertyisnotultimatelyseparablefromadiscus-sionofvirtue,andthusofhowtoliveasasocialbeing;therealthreattosocietystemsfromfaction.OnonepointHumewasadamant:‘violentinnovationsnoindividualisentitledtomake’(Hume1975b:‘OntheOriginalContract’).Andpreciselybecausetheconsequencesofexcesscannotbeforeseen,thereisneveracaseforinstigatingarevolutionorresistingthelaw.Thestabilityandverystructureofsocietydependontheupholdingoflaw,andsocietyfacesitsgreatestperilwhenresistancetolawisitselfproclaimedas‘lawfulorcommendable’.Gradualchangeshouldbesought,intherecognitionalso,thatthecomplexityofpoliticalquestionsmeansthat‘therescarcelyeveroccurs,inanydeliberation,achoice,whichiseitherpurelygood,orpurelyill.Consequences,mixedandvaried,maybeforeseentoflowfromeverymeasure:andmanyconsequences,unforeseen,doalways,infact,resultfromeveryone’(ibid.:‘OntheProtestantSuccession’,published1752).Throughouthiswritings,Humeobservesthatpersonalinclinations,politicalinter-ests,religiouszeal,areconstantthreatstosociety.Itis‘notenoughforlibertytoremainonthedefensive’,notleastbecause,howevercarefullyframed,nolawscould‘possiblyprovideagainsteverycontingency’(Hume1985c:XLVIII,XLVII).HumeheldthatTOLERATION(p.690)developedhistoricallynotasaresultofpositivepolicy,butfromaslowrecognitionthatattemptsatrepressionmerelystiffenedresistance.Hedidnotthink,however,thattolerationalwayswinsthrough.Inthiscontext,Humefacedadeeprecurringconflictinhisphilosophybetweenupholdingmoderationandstabilizingtheunavoidabletensionbetweenlibertyandauthority.Sincethereisnocriterionofmoderationinadvanceofaparticularcontextwithclearboundaries(formoderationisdefinableonlybyreferencetolimits),therecanbenofullyintelligibleprincipleofmoderationindependentofcases:buthowaresuchcasestobeidentified?Moreover,itisunclearhowanattitudeofmoderationcould,inpractice,motivateanyonetodisplacedeeplyengrainedhabits,ontheonehand,orfanaticism,ontheother.Theaimofgovernment,inHume’sview,istoestablishandthentopreservejustlaws.Butthelawsmustnotsupposethathumannaturecanbeverymuchchanged.Thelaw-makers,inotherwords,whetherinthesphereofgovernmentormorality,mustfirstseektounderstandHUMANNATURE(pp.672–3).Theymustalsograspthatthreatstolibertystemfromvarioussources.OntheeconomicfrontHumesawthenationaldebt,withitsattendanttaxationlevies,togetherwiththeprospectofnationalizationofproperty,asespeciallydangerous.Heconsidersthatlowinterestrates,controlledlabourcosts,andcompetitivepricingareallessentialtoensurebothsuccessfulforeigntrade,andabuoyanthomeeconomywhichmotivatesthelabourforce.Heattackedallisola-tionisminpoliticsoreconomics,andrecognizedthatdiversificationwouldbenecessaryforaneconomyrestingonafewstapleproducts,becausesuchaneconomyisalwaysvulnerabletodecreasingdemand,andtolowerpricingbysuccessfulcompetitors.It719\nPETERJONESmightbeaddedthathedisapprovedofstateownershiponthegroundsthatastatecanalwayssecureitselfagainstenquiryandaccountability.TheScotlandofHume’sdaywaspoor,sparselypopulatedandentirelyagrarian;howfarhisreflectionsareadapt-abletomodern,urban,industrialsociety,andforpopulationsofimmenselygreatersize,ishotlydebated.2.8ScepticismHumewasbynomeansashamedofhisscepticalstanceandhisscepticalarguments,althoughhisopponents,usingthesamelabel,regardeditextremelyunfavourably.Asageneraltermofabuse,‘sceptic’coveredallthosewhochallengedorthodoxviews,askeddifficultquestions(evenwhennooneelseclaimedtoknowtheanswer),omittedmentionofGodorcanvassedviewsthatleftlittleornoroomforaGod,deniedtheexistenceofthingsforwhichwordsneverthelessexisted,ormerelyconfessedigno-rance.OnallthesegroundsHumequalifiesforthelabel.Atdifferenttimesheexplicitlydeniesthatwehaveinnateideas;thatthecausalrelationisobservablyanythingotherthanconstantconjunction;thatthereareobservablenecessaryconnectionsanywhere;thatthereiseitheranempiricalordemonstrativeprooffortheassumptionsthatthefuturewillresemblethepast,andthateveryeventhasacause;thatthereisanirresolvabledisputebetweenadvocatesoffreewillanddeterminism;thatformalrea-soningiscruciallyinvolvedineverydaylife;thatthereisacaseforbelievingreportsofmiracles;thatthereisexplanatoryforcetothedesignargument;thatextremescepti-cismiscoherent;andthathecanfindtheexperientialsourceofourideasofself,substanceorGod.Inspiteofthislist,whichcaneasilybesupplemented,Humethoughtofhisownpositionasoneofonlymoderatescepticism.Inthis,hewascarryingforwardthesci-entifictemperamentoftheseventeenthcenturyandalsofollowinghismentorCicero,whourgedthat,evenifwelackafoolprooftestfordistinguishingtruthfromfalsehood,inpracticallifeprobabilityisenough;thingshavetogetdone,andwedonotneedthelastounceofevidenceinordertodothem.Humewasfrequentlymisunderstood,however,asdenyingthetruthofcertainclaims,whenhewasaskinghowweknowthemtobetrue.2.9InfluenceanddisputedquestionsHowevermuchwemaythinkthathiscontemporariesmisunderstoodHume’sviews,thereisnodoubtthattheyreactedstronglytothem.AdamSmithderivedmanyofhisnotionsinmoral,economicandpoliticalphilosophyfromHume,extendingthemwhereneedbe–asinthecaseof‘sympathy’and‘theimpartialspectator’.AdamFerguson(1723–1816),likewise,developedhisconceptionofcivilsocietyagainsttheback-groundofHume’sreflections.InAmerica,JamesMadison(1751–1836)andAlexanderHamilton(1755–1804)studiedHume’spoliticalwritingsandHistorywithgreatcarewhileformulatingtheirviewsonthefuturestructureanddirectionoftheirnation.InFrance,too,Hume’spolitical,economicandhistoricalwritingswerewidelyadmiredanddiscussedbytheleadingintellectualsbeforetheRevolution.InScotland,720\nHUMEThomasReid(1710–96)devotedagreatdealoftimetorebuttingHume’sviewsonthenatureofbothknowledgeandmorality.OneofthemosttroublesomenotionsformanyofHume’scontemporaries,however,wasthatofcausation:inGermany,ImmanuelKANT(1724–1804)(chapter32),havingcreditedHumewithwakinghimfromadog-maticslumber,devotedaconsiderableportionofhisphilosophytoformulatingacogentalternativetoHume’sposition.AgreatdealofWesternphilosophysincetheendoftheeighteenthcenturycanbeassociatedwiththerivalviewsofHumeandKant.IntermsofsalesthemostpopularofHume’sworkswastheHistory.Itwentthrough150editionswithinacenturyofpublication,whereastheTreatisehadtowaitnearlyaslongforasecondprinting.InthenineteenthcenturyafewphilosophersclaimedHumeastheirmentor,suchasAugustComte(1798–1857)inFrance,andJ.S.MILL(chapter35)andT.H.Huxley(1825–95)inEngland;WilliamJAMES(chapter36),inAmerica,drewmuchinspirationfromhim.TherehasbeenanastonishingresurgenceofinterestinHume’sworksincethe1960s,althoughintheearlierpartofthiscenturyhisnamehadbeenkeptaliveinScotland.Inthefirstplace,disputehascentredontheappropriatecontextsforunderstandingHume’sworkandthedueweighttobegiventohisvariousclaims,whichweremadeovera40-yearspaninworksofdifferentkinds.Humewastryingtodisplacemanyofhisreaders’viewsonthenatureofhumanityandsociety,andhisownknowledgeoftheirbeliefsprofoundlyinfluencedthestyle,contentandemphasisofhisapproach.InthiscontextthereisdisputeoverHume’sconceptionofphilosophyanditsroles,andoverthenatureandscopeofhisscepticism.Onmoreparticulartopics,thereisconsiderabledisagreementabouttheneedfor,orthepossibilityof,modifyinghisaccountofcausationtocopewithdevelopmentsinmodernscience,especiallyatthelevelofverysmallorverylargephenomena.DoappealstostatisticalfrequenciesavoidthescepticalelementsinHume’sposition?Arehisownviewsoninductiveargumentandtheweighingofprobabilitiesfundamentallysound?Ifoursensoryexperienceisaseamlesscontinuum,assomepeoplemaintain,doesitmakesensetodivideitupintoeventswithbeginningsandends?Ifwecannotexplainhowweidentify,separateandclassifyevents,canwetalkofseparatecausesandeffectsatall?Humeemphasizedtheimportanceofresemblanceorsimilarityinthistheoryofcausation:similarcausesareassumedtoproducesimilareffects.Thingsaresimilar,however,onlyincertainrespects.Howdowedecideontherelevanceandexistenceofsimilarities?Hisviewsonthenatureoftheselfprovokelengthyargumentinlegalandmedicalspheres,aswellasamongphilosophers;andhisviewsonfreedomanddeterminism,therolesofreasonandsentiment,andthenatureofjustice,aremuchdebated.HiseconomicviewsarenowcloselycomparedwiththoseofAdamSmith.Studiesofseventeenth-centuryandeighteenth-centuryScottishlegalthought,andtheprecisetheologicalissuesoftheday,togetherwithattemptstoestablishthenatureofHume’sratherslightscientificknowledge,arelikelytoleadtorevisedinterpretationsofhiswork.Onesignofagreatthinkeristhatlaterreadersdeemitaculturaldutytodefinetheirpositioninrelationtothatthinker,andtotreathimorher,howeveranachronis-tically,asacontemporary.Withtheattentionnowbeinggiventothewholerangeofhisworks,includingtheEssaysandtheHistory,itisunlikelythathisviewswillsoonsufferneglect.721\nPETERJONESFurtherReadingThereisnostandardoruniformeditionofHume’sworks,butmostofthemareavailableinpaper-back.ConvenienteditionsofHume’sEssaysandalsoofhisHistoryarepublishedintheLibertyClassicsseries(Hume1985aand1985crespectively).TherearemanyeditionsofAnEnquiryConcerningHumanUnderstandingandAnEnquiryConcerningthePrinciplesofMorals,andtheyareprintedtogetherintheClarendonPresseditioneditedbyNidditch(1975a).Nidditch(1978a)hasalsoeditedATreatiseofHumanNature,whichispublishedalongwiththeAbstract.ThebesteditionoftheDialoguesConcerningNaturalReligion(1779)iseditedbyN.KempSmith,whoalsoprovidedalongintroduction.Mossner(1980)hasbecomethestandardbiography,withtheScottishEnlightenmentandothercontextsexploredinDaiches,JonesandJones(1986),Jones(1989,1982)andSher(1985).ClassicstudiesofHumeincludeChurch(1934),Laird(1932),Passmore(1952),Price(1940)andSmith(1941).OfrecentgeneralstudiesofHume’sphilosophy,IcanmentionCapaldi(1974),Norton(1982),Noxon(1973),Penelhum(1975),Stroud(1977)andWright(1983).Bennett(1971)exploresmajorthemesinLocke,BerkeleyandHume.Ardal(1966),Baier(1991),Fogelin(1985),Flew(1961)andPears(1991)examinedifferentaspectsoftheTreatise.BeauchampandRosenberg(1981)andStrawson(1989)considerHume’saccountofcausation,whichisalsoacentralconcernofMackie(1974).Stove(1973)discussesHumeoninduction.Hume’sphilosophyofreligionisdiscussedbyGaskin(1978),Hurlbutt(1965),Jeffner(1966),andTweyman(1986),andhispoliticsisdiscussedbyMillar(1981).ReferencesHumeHume,D.1947[1779]:DialoguesConcerningNaturalReligion(editedbyN.KempSmith).London:Nelson.——1975a[1748–51]:EnquiriesConcerningHumanUnderstandingandConcerningthePrinciplesofMorals,3rdedn(editedbyP.H.Nidditch).Oxford:ClarendonPress.——1975b[1748]:EnquiryConcerningHumanUnderstanding.InHume(1975a).——1975c[1751]:EnquiryConcerningthePrinciplesofMorals.InHume(1975a).——1978a[1739–40]:ATreatiseofHumanNature,2ndedn(editedbyP.H.Nidditch).Oxford:ClarendonPress.——1978b[1740]:AnAbstractofaTreatiseofHumanNature.ReprintedinHume(1978a).——1985a[1741–2]:Essays,Moral,Political,andLiterary.Indianapolis,IN:LibertyFund.PublishedwithPoliticalDiscourses(Hume1985b).——1985b[1752]:PoliticalDiscourses.Indianapolis,IN:LibertyFund.PublishedwithEssays,Moral,Political,andLiterary(Hume1985a).——1985c[1754–62]:HistoryofEngland,6vols.Indianapolis,IN:LibertyFund.OtherwritersArdal,P.S.1966:PassionandValueinHume’sTreatise.Edinburgh:EdinburghUniversityPress.Baier,A.C.1991:AProgressofSentiments:ReflectionsonHume’sTreatise.Cambridge,MA:HarvardUniversityPress.Beauchamp,T.L.andRosenberg,A.1981:HumeandtheProblemofCausation.Oxford:OxfordUniversityPress.Bennett,J.F.1971:Locke,Berkeley,Hume:CentralThemes.Oxford:OxfordUniversityPress.722\nHUMECapaldi,N.1974:DavidHume:TheNewtonianPhilosopher.Boston:Twayne.Church,R.W.1934:Hume’sTheoryoftheUnderstanding.London:AllenandUnwin.Daiches,D.,Jones,E.J.andJones,P.1986:AHotbedofGenius.Edinburgh:EdinburghUniversityPress.Flew,A.1961:Hume’sPhilosophyofBelief.London:Routledge.Fogelin,R.J.1985:Hume’sScepticismintheTreatiseofHumanNature.London:Routledge.Gaskin,J.C.A.1978:Hume’sPhilosophyofReligion.London:Macmillan.Hall,R.1978:FiftyYearsofHumeScholarship.Edinburgh:EdinburghUniversityPress.Hendel,C.W.1963:StudiesinthePhilosophyofDavidHume.Indianapolis,IN:Garland.Hurlbutt,R.H.1965:Hume,NewtonandtheDesignArgument.Lincoln:UniversityofNebraskaPress.Jeffner,A.1966:ButlerandHumeonReligion.Stockholm:Diakonistyrelsensbokforlag.Jones,P.1982:Hume’sSentiments:TheirCiceronianandFrenchContext.Edinburgh:EdinburghUniversityPress.——(ed.)1989:TheScienceofManintheScottishEnlightenment:Hume,ReidandOtherContem-poraries.Edinburgh:EdinburghUniversityPress.Laird,J.1932:Hume’sPhilosophyofHumanNature.London:Methuen.Livingston,D.W.1984:Hume’sPhilosophyofCommonLife.Chicago:UniversityofChicagoPress.Livingston,D.W.andKing,J.T.1976:Hume:ARe-evaluation.NewYork:FordhamUniversityPress.Mackie,J.L.1974:TheCementoftheUniverse.Oxford:ClarendonPress.——1980:Hume’sMoralTheory.London:Routledge.Michaud,Y.1983:Humeetlafindelaphilosophie.Paris:PressesUniversitairesdeFrance.Millar,D.1981:PhilosophyandIdeologyinHume’sPoliticalThought.Oxford:ClarendonPress.Mossner,E.C.1980:TheLifeofDavidHume,2ndedn.Oxford:ClarendonPress.Norton,D.F.1982:DavidHume:CommonSenseMoralist,ScepticalMetaphysician.Princeton,NJ:PrincetonUniversityPress.——(ed.)1993:TheCambridgeCompaniontoHume.Cambridge:CambridgeUniversityPress.Norton,D.F.,Capaldi,N.andRobison,W.L.(eds)1979:McGillHumeStudies.SanDiego,CA:AustinHillPress.Noxon,J.1973:Hume’sPhilosophicalDevelopment.Oxford:ClarendonPress.Passmore,J.1952:Hume’sIntentions.Cambridge:CambridgeUniversityPress.Pears,D.1991:Hume’sSystem.Oxford:OxfordUniversityPress.Penelhum,T.1975:Hume.London:Macmillan.Price,H.H.1940:Hume’sTheoryoftheExternalWorld.Oxford:ClarendonPress.Sher,R.B.1985:ChurchandUniversityintheScottishEnlightenment.Edinburgh:EdinburghUni-versityPress.Smith,N.K.1941:ThePhilosophyofDavidHume.London:Macmillan.Stove,D.C.1973:ProbabilityandHume’sInductiveScepticism.Oxford:OxfordUniversityPress.Strawson,G.1989:TheSecretConnexion:Causation,RealismandDavidHume.Oxford:ClarendonPress.Stroud,B.1977:Hume.London:Routledge.Tweyman,S.1986:ScepticismandBeliefinHume’sDialoguesConcerningNaturalReligion.TheHaig:MartinusNijhoff.Wright,J.P.1983:TheScepticalRealismofDavidHume.Manchester:ManchesterUniversityPress.DiscussionQuestions1Aretherecausalrelationsbetweenevents?Whatgroundsarethereforbelievingso?723\nPETERJONES2Ifnecessitycannotbeobserved,howcancausalityinvolvenecessity?3Whyshouldwebelievetheclaimthateveryeventhasacause?4Willthefutureresemblethepast?Canweproveit?Arewejustifiedinbelievingit?5Whatareideas?Dotheyallhavetheirorigininimpressions?6Doesexperienceprovideasufficientbasisforknowledgeofmattersoffact?Whatelsemightbeneeded?7Doescustomjustifyinductiveinferenceorexplainwhywemakesuchinferenceseventhoughtheyarenotjustified?8Doweactfreelyevenifallouractionsarecaused?9Couldweeverbejustifiedinsupposingthatthelawsofnaturehadchangedorthatwehaveencounteredaneventwithoutacause?10Doeseachofourperceptionsexistindependentofanyotherperception?11Arewejustifiedinbelievingthattherearephysicalobjects?12WhydidHumecometobelievethathisaccountofpersonalidentitywasunsatisfactory?Ifthereisadeficiency,canitberemedied?13Shouldouraccountofpersonalidentitydifferfromouraccountoftheidentityofphysicalobjects?14Whatistheroleofagreementinouraccountofknowledge?Isknowledgepossiblewhereexpertsdisagree?15Whatphilosophicalproblemsarisefromreportsofmiracles?Howcanthesephilosophicalproblemsberesolved?16CanthecomplexityandorderoftheworldtellusanythingabouttheexistenceofGod?CantheytellusanythingaboutthenatureofGod?17DoesHume’snaturalismanswerhisscepticism?18Whatroledoesreasonhaveinexplainingactions?19Whatmakesaqualityavirtue?20Howisitpossibletojudgeoneselfimpartially?21Canaestheticjudgementsbeobjective?Canaestheticdisputesberesolved?22Musttheintentionoftheartistorrelevancetoanaudienceenterintojudgingaworkofart?23Inhistoricalknowledge,howcanwecombinetheknowledgeoftheactorandthatofthespectator?24Howisthereroomforreinterpretationinhistoricalknowledge?25Canlibertyandauthoritybereconciled?26Isthereeverarighttoresistthelaw?Istherearighttoinstigatearevolution?27Whatarethemainthreatstoliberty?28IsHume’sthoughtconcerninglaw,governmentandsocietyapplicabletoday?724\n32KantDAVIDBELLThereishardlyanareaofphilosophytowhichImmanuelKant(1724–1804)didnotmakemassiveandprofoundcontributions,includingmetaphysics,philosophyofmind,epistemology,philosophicallogic,thefoundationsofmathematics,ethics,aesthetics,philosophyofscienceandpoliticalphilosophy.Thereare,however,twocentresofKant’sphilosophicalthought,whichcomprisethefoundationforthewhole.Ontheonehand,thereisthe‘critical’or‘transcendental’metaphysicsoutlinedinhisCritiqueofPureReason(firstpublishedin1781),andontheotherhand,thereistheethicaltheorythathepublishedinanumberofworksafter1784.Intheformerhedevelopedanaccountofthemindanditsplaceinnaturethatwasdesignedtoovercomethelimitationsinherentinthethoughtofhisrationalistandempiricistpredecessors.Inhisworksonethicsheformulatedacompellingvisionofhowmorallaw,autonomy,rationality,freedom,dutyandvirtuearerelatedtooneanotherinmorallife.ThefollowingchapterconcentratesonthesetwocentralareasofKant’sthought.1LifeImmanuelKantwasborninKönigsberg,aflourishingPrussianportontheBalticSea.Hewenttothelocalgrammarschool,byallaccountsanunwholesomeplace,wherethepietistreligiousatmospherewasenforcedbythe‘pedanticandgloomydisciplineoffanatics’,butwhereheneverthelesssucceededinsupplementinghislimitedmeanswithhiswinningsfrombilliardsandcards.Attheageof16hebeganhisstudiesatKönigsbergUniversity.Duringthenextdecadehestudied,amongotherdisciplines,mathematics,theology,physicsandphilosophy.In1755KantgainedthequalificationsnecessaryforhimtobecomeaPrivatdozent(anindependentteacher)attheuniversity.AsaPrivatdozenthewasnotentitledtoasalary,butinsteadreceivedasmallfeefromeachstudenthetaught.InordertosurvivefinanciallyKantwasforcedtolectureforupto36hoursperweek,onsubjectsasdiverseaslogic,metaphysics,geography,ethics,physics,pedagogics,arithmetic,geometry,trigonometry,naturaltheology,astronomy,meteorologyandanthropology;andin1764hewasoffered,butdeclined,thepostofProfessorofPoetry.Inspiteofthis\nDAVIDBELL‘monstrousacademicworkload’,Kantwasanythingbutarecluse.OnecontemporaryrecalledthatKant‘spentalmosteverymiddayandeveningoutsidehishouseinsocialactivities,frequentlytakingpartalsoinacardparty,andonlygettinghomearoundmidnight’.Duringthis‘pre-critical’periodKantpublishedanumberofminorworksonavarietyofsubjects.(ThoughimportantforafullunderstandingofKant’sthought,thesepre-criticalwritings(Kant1968)willnotbeexaminedinthepresentchapter.)KantbecameProfessorofLogicandMetaphysicsinKönigsbergin1770,butitwasnotuntilsomeelevenyearslater,attheageof57,thatheproducedthefirstintheseriesofincomparablygreatandmassivelyinfluentialworksthatwastoembodyhisso-calledCriticalPhilosophy:thefirsteditionoftheCritiqueofPureReasonappearedin1781.Itwasfollowed,withremarkablerapidity,bytheProlegomena(publishedin1783),theFoundationsoftheMetaphysicsofMorals(publishedin1785),theMetaphysicalFoundationsofNaturalScience(publishedin1786),thesecond,substantiallyrevisededitionoftheCritiqueofPureReason(publishedin1787),theCritiqueofPracticalReason(publishedin1788),andtheCritiqueofJudgement(publishedin1790).TwomonthsbeforehiseightiethbirthdayKantdiedinKönigsberg.Hisfuneralwasattendedbymanythousandsofmourners,anditisclearthathewasuniversallyreveredasmuchforhishumanityandmoralintegrityasforhisphilosophicalgenius.Onhistombstonewasinscribed,appropriately,amottotakenfromthepassagewithwhichhehadconcludedtheCritiqueofPracticalReason:‘Twothingsfillthemindwithevernewandincreasingadmirationandawe,theoftenerandmoresteadilywereflectonthem:thestarryheavensaboveme,andthemorallawwithinme’.Asthislastsentencehints,thetwogreatissuestowhichKant’scriticalphilosophyisaddressedare,firstly,thenatureofourunderstandingoftheuniverseand,secondly,thenatureofourmorallife.Theseissuesarethetopics,respectively,ofthefollowingtwosections.2TheMetaphysicsofExperienceHowcanweunderstandtheuniverseandthethingsitcontains?Howisthisevenpossible?Asthisisahuge,andhugelycomplexproblem,wemightbeginbybreakingitdownintomoremanageableparts.Wemightask:inwhatwaysdowe,asamatteroffact,understandtheworldandthethingsitcontains?Andforaninitial,veryroughanswer,wemightsaysomethinglikethis.Ontheonehand,ourunderstandingoftheworldiscouchedverylargelyintermsofobjects–objectsthatpossessproperties,par-ticipateinevents,occupyspace,persistthroughtime,interactcausallywithoneanother,aresubstantialandobeythelawsoflogic,arithmetic,geometryandphysics.Wemightcallthesetheobject-relatedaspectsoftheproblem.Ontheotherhand,ourunderstandingoftheworldisbasedonourobservationsandperceptions,ourconcepts,thoughts,judgementsandinferences.Wemightcallthese,incontrast,thesubject-relatedaspectsoftheproblem.Theobject-relatedaspectscanbemademoreexplicitinanumberofMETAPHYSICAL(chapter2)questions.Forexample:whatisanOBJECT(p.713)?WhatisaPROPERTY(pp.96–7)?Whatisanevent?WhatareTIME(pp.82–5)andSPACE(pp.358–9)?WhatisCAUSALITY(chapter9)?Howislogic,ormathematicsornaturalscience,possible?726\nKANTCorrespondingly,thesubject-relatedaspectsyieldanumberofquestionsthatbelong,broadlyspeaking,toEPISTEMOLOGY(chapter1)andthePHILOSOPHYOFMIND(chapter5):whatisittoperceivesomething?Whatisaconcept?WhatisaJUDGEMENT(p.797)?Howareconceptsrelatedtojudgements?Howareconceptsrelatedtoperceptualexperi-ences?Howareperceptualexperiencesrelatedtosensations?ThequestionsmentionedinthisparagraphareamongthemostimportantthatKantaddressesintheCritiqueofPureReason,andtheyarequestionstowhich,indeed,hethereprovidesspecificandoftenprovocativeanswers.Thereis,however,averygeneral,overarchingphilosophicalproblemwhichcanitselfbecharacterizedintrinsicallyasneitherobject-relatednorsubject-related;foritis,precisely,thegeneralproblemofhowthingsthatareobject-relatedcanpossiblyfittogetherwiththosethataresubject-related.Again,thisproblemcanbespelledoutinanumberofmorespecificquestions.Forexample,howdoourconceptsrefertoitemsinreality?Howdooursubjectivesensationsandexperiencesgiveusobjectiveknowl-edgeofthematerialuniverse?Howcanourinnerthoughtsandjudgementsbeobjectivelytrueoftheexternalworld?Kant’scriticalphilosophymakesthisoverarchingmetaphysicalproblemitsmostfundamentalconcern;andthatphilosophyasawholeischaracterized,aboveallelse,byhisinsistencethatinanyacceptablesolutiontothemetaphysicalproblem,object-relatedandsubject-relatedphenomenamustbetakentobemutuallydependentand,ultimately,inseparable.Inotherwords,nocoherenttheoryofeitheroneispossiblethatisnot,already,atheoryoftheother.Inparticular,Kantclaimed,nocoherenttheoryof(objective)properties,objects,events,causalrelations,substance,timeorspacecanbeprovidedthatisnotalreadyalsoanaccountofour(subjective)perceptions,conceptsandjudgementsconcerningsuchthings.Inthelastanalysis,thatis,thenatureoftherealityweknowisinseparablefromthenatureofthemindthatknowsit.ThisishowKantfamouslyexpressedthisdoctrineintheprefacetothesecondeditionoftheCritiqueofPureReason:Hithertoithasbeenassumedthatallourknowledgemustconformtoobjects.Butallattempts[forinstance,toaccountforthepossibilityofobjectiveknowledge]have,onthisassumption,endedinfailure.Wemustthereforemaketrialwhetherwemaynothavemoresuccessinthetasksofmetaphysics,ifwesupposethatobjectsconformtoourknowledge.(Kant1929:Bxvi)Kantcalledthischangeinmetaphysicalperspectivehis‘Copernicanrevolution’.Theresultingtheoryisaformoftranscendentalidealism.TheCopernicanRevolutionCopernicus,asixteenth-centuryastronomer,pointedoutthattheapparentmotionoftheSunandstarshadhithertobeenassumedtoberealmotion:itwasclaimedthatthereasontheSunappearstoriseintheeast,travelacrossthesky,andsetinthewest,issimplythatitreallyismovingroundastationaryEarth.Copernicus,however,arguedthatpreciselythesameappearanceswouldresultif,instead,theSunwereinfactsta-tionaryandtheEarthwerespinningonitsaxis.Copernicusreplacedthenaivetheory,whichtooktheapparentmotionoftheSuntoberealmotion,withatheoryaccording727\nDAVIDBELLtowhichtheapparentmotionoftheSunisineffectaproductoftherealmotionoftheobserver:itisbecausewearespinningthattheSunseemstomoveacrossthesky.Kant’sso-calledCopernicanrevolutionisanalogous.Ithadhithertobeenassumedthatthereappeartobespatio-temporalobjectsthatexistindependentlyofusbecausetherereallyaresuchthings.Kantreplacedthisnaiverealismwithatheoryaccordingtowhichtheapparentnatureandindependenceoftheobjectiveworldisaproductofourperceptions,conceptsandjudgements;inthelastanalysis,itisbecauseweperceiveandthinkaswedothattheworldseemstobeasitis.Kant’sidealismismotivatedbyanumberofconsiderations,oneofwhichisasfollows.Therearecertainverybasicprinciplesandjudgementswhichformthefoundationofourunderstandingoftheuniverse.Theseprinciplesandjudgementsinclude,forexample,thetruthsofarithmetic,thetruthsofgeometry,theprinciplethateveryeventhasacause,thejudgementthatobjectsneversimplygooutofexistence,andlikewisethejudgementthatobjectsneversimplycomeintoexistence,outofnowhereandoutofnothing.AccordingtoKantsuchclaimsasthesehaveanumberofinterestingandproblematiccharacteristics:1Theyarenecessarilytrue,andcannotbeeitherjustifiedorfalsifiedbyappealtocontingentfactsorperceptualexperience(thatis,theyareapriori).2Theyarenotmerelylogicaltruths,ortruthsbydefinition(thatis,theyaresynthetic).3Theyareessentialtoourunderstandingofreality.Kantcalledtheknowledgewehaveoftruthsofthiskind‘syntheticaprioriknowledge’,andhisidealismismotivatedbyacombinationoftwobeliefs:(1)thatwithoutsyntheticaprioriknowledge,noknowledge,orunderstanding,ormeaningfulexperiencewouldbepossibleatall;and(2)syntheticaprioriknowledgeisnotknowledgewecanpossi-blyreceive‘fromwithout’,asaresultofourexperienceoftheworld.Onthecontrary,syntheticaprioriknowledgeisonlypossibleifitencodesthejudgementsandprinciplesweadoptinmakingsenseofourexperience.TheCopernicanrevolutioninphilosophyisnecessary,inotherwords,becausewithoutitwecannotmakeintelligiblethepossi-bilityofsyntheticaprioriknowledge–andintheabsenceofthatpossibility,Kantbelieved,theverypossibilityofanyknowledgeorunderstandingwhatsoeverwouldbecomeunintelligible.SyntheticKnowledgeapriori‘Analytic’isshortfor‘analyticallytrue’,andajudgementisanalyticifitstruthissolelyaconsequenceoftheconceptsitcontains.So,forexample,thetruthsoflogicareana-lytic,asaresuchconceptualtruthsas,forinstance,thefactthatallbachelorsareunmar-ried,orthatredisacolour.‘Synthetic’simplymeans‘notanalytic’.Thetermaposterioriisshortfor‘acquiredaposteriori’,andsomethingisknownaposterioriiftheknowledgeisacquiredonthebasisofsensoryexperience,thatis,ifitisinprinciplefalsifiablebyexperience.Thetermapriorisimplymeans‘notaposteriori’.Sosyntheticaprioriknowl-edgeisknowledgethatisnecessarilytrue,whichcannotbefalsifiedbyexperience,butwhichisnotmerelytrivially,logicallyoranalyticallytrue.728\nKANTInconformitywiththisidealistperspective,theCritiqueofPureReasonanalysesboththeobject-relatedandthesubject-relatedelementsofknowledgewiththeaimofshowingthatallapriorielementsmustultimatelybeexplainedbyappealtoouressentialsubjectiveconstitution–thewaysinwhichwenecessarilyperceive,concep-tualizeandthink–andcannotbeexplainedbyappealtothewaytheworldisinde-pendentlyofanyaccesswemayhavetoit.Itisnotbecausereality,inandofitself,isspatial,temporalorcausalthatweexperienceitassuch.Rather,theexplanationrunsintheoppositedirection:itisbecauseoftheessentialnatureandstructureofourmindsthatourexperienceis,precisely,ofaspatial,temporal,causalreality.Kantputsitbluntly:‘wecanknowaprioriofthingsonlywhatweourselvesputintothem’(Kant1929:Bxviii).TheoverallstructureofKant’scriticalphilosophy–and,aswewillsee,theorgani-zationoftheCritiqueofPureReasonitself–dependsuponhisdistinctionbetweenthreefundamental,irreduciblepowersofthemind.Firstly,wehavethecapacitytoreceiveorregistersensoryitemssuchassensations,impressions,sense-data,perceptsandthelike.Kant’stermforsuchsensoryitemsis‘intuitions’,andhecallsthecapacitytoregisterthem‘sensibility’.Secondly,wehavecapacitiesofanessentiallyintellectualkind,involvingthepowertoconceptualize,tothinkandtojudge.Kantassignssuchabilitiestowhathecalls‘theunderstanding’,afacultywhichisresponsibleforourabilitytouseconcepts.Thirdly,weareabletoinferlogically,todrawvalidconclusions–thatis,toreason.Sensibility,accordingtoKant,islargelypassive:sensationsandintuitionsarethingsweundergoorthathappentous.Understanding,however,andreasonareessen-tiallyactive:conceptsarethingsweuse,andofcoursethinkingandreasoningarethingswedo.Kantisadamantabouttwoclaims.Thefirstisthatsensibilityandunderstandingarequitedistinct.Theyhavetheirownoperations,principlesandfunctions.Thesecondisthatinallknowledgewhatsoeverthatisavailabletous,bothsensibilityandunder-standing–bothintuitionsandconcepts–mustbeinvolved.Kantcategoricallydenies,inotherwords,thatwecanhaveanyknowledgethatispurelysensory,or,equally,thatwecanhaveanyknowledgethatisexclusivelyconceptual;for‘thoughtswithoutcontentareempty,intuitionswithoutconceptsareblind’(ibid.:B75).TheorganizationoftheCritiqueofPureReasoniscomplex,andcanindeedbebaf-fling.Thedistinctionbetweensensibility,understandingandreasoncan,however,helpustomakesomesenseofit.Figure32.1(whichisnotexhaustive)representssomeofthemostimportantdivisionswithinthestructureofKant’stext.Thefollowingremarksmaybeofsomehelpinnavigatingthismaze.ThemaindivisionoftheCritiqueofPureReasonintotheTranscendentalAestheticandtheTranscendentalLogiccorrespondstoKant’sdistinctionbetweensensibility,ontheonehand,andunderstandingandreasonontheother.TheTranscendentalAestheticisso-called–fromtheGreekwordaisthesis,meaningperceptionorsensation–becauseitdealswiththeapriorielementsthataretobefoundinsensory,perceptualexperience(sothispartofthetextinfacthasnothingtodowith‘aesthetics’asthatwordisusednow).OneofKant’smainobjectivesinthispartoftheCritiqueistodemonstratethatspaceandtimeareaprioriformsofsensoryexperience.Inotherwords,asatranscen-dentalidealist,heaimstoshowthatitisbecause(1)ourintuitionshaveapriorispatialandtemporalformsthat(2)empiricalobjectsintheexternalworldarespatio-temporal.729\nDAVIDBELLTHECRITIQUEOFPUREREASONTRANSCENDENTALTRANSCENDENTALAESTHETICLOGICTimeSpaceTRANSCENDENTALTRANSCENDENTALANALYTICDIALECTICANALYTICOFANALYTICOFCONCEPTSPRINCIPLESMetaphysicalTranscendentaldeductiondeductionSchematismPrinciplesConceptsofDialecticalinferencespurereasonofpurereasonTheparalogismsTheantinomiesTheidealofofpurereasonofpurereasonpurereasonFigure32.1ThemaindivisionsofKant’sCritiqueofPureReason.Hebelievesnoattempttoexplain(1)intermsof(2)canbecoherent.TheTranscendentalLogicisso-called–fromlogos,theGreekwordforpropositionormeaning–becauseitdealswiththeapriorielementsthatarepresentinourconceptual,intellectualmake-up(sothissectionisonlypartiallyandindirectlyrelatedto‘logic’asthetermisusedtoday).WithintheTranscendentalLogicitself,theTranscendentalAnalyticconcernsitselfprimarilywiththenatureofourunderstanding,thatis,withthenecessaryfeaturesonwhichourentireabilitytojudgeandapplyconceptsultimatelyrests.HereKantargues,amongotherthings,thatifwedidnotpossessandapplycertainfundamental,primitive,aprioriconcepts(hecallsthemthe‘categories’),thennounderstandingorthoughtwouldbepossible.Thecategoriesincludesuchbasic,aprioriconceptsasthosewhichwetypicallyexpress,forexample,withthewords‘all’,‘one’,‘not’,‘is’,‘maybe’,‘exists’and‘causes’.(Kantrespectivelylabelsthesecate-goriesTotality,Unity,Negation,Reality,Possibility,ExistenceandCausality.)IntheTranscendentalDialectic,ontheotherhand,Kantaddressesthephilosophicalproblemsassociatedwithourcapacitytouse–and,moreimportantly,withourtendencytoabuseandmisuse–ourpowersofreason.(Kantcallsreasoning‘dialectical’whenitisinvalidorfallacious.)TheoverallaimoftheTranscendentalDialecticistoshowthatpurereasonalone,whendivorcedfromanyinputfrom,orapplicationto,sensoryexpe-rience,canprovideuswithnogenuineknowledgewhatsoever.That,inessence,isKant’scriticismor‘critique’ofpurereason.WithintheTranscendentalDialectic,thesectionentitledtheParalogismsofPureReasonaimstoshowthatanyattempttodeducefacts730\nKANTaboutthemindorself,onpurelyrationalgroundsandindependentlyofallreferencetosensoryexperience,isdoomedtofailure.(Aparalogismisacertainsortofinvalidargument.)IntheAntinomiesofPureReason,Kantattemptstodemonstratetheimpossibilityofpurelyconceptual,sense-independentknowledgeofthematerialworld.(Anantinomyisaparadoxorcontradiction.)AndintheIdealofPureReason,heattemptstoprovethatpurereasonalonecanyieldnoknowledgeofGod.ATranscendentalEnquiryAsKanttypicallyusesit,theterm‘transcendental’isintendedtocontrastsharplywithtwootherterms,namely‘immanent’and‘transcendent’.Somethingisimmanentwithrespecttocertainboundsorlimitsifitlieswithinthem.Somethingis‘transcendent’ifitliesbeyondthoseboundsorlimits.Somethingis‘transcendental’,however,ifitliesneitherwithinnorwithoutthoselimits,butis,rather,amatteroftheessentialnatureofthoseverylimitsthemselves.Kantwasconcernedtoestablishtheessential,apriorilimitsofhumanknowledge.Nownothingcanpossiblyprovideinformationaboutwhatistran-scendent,thatis,whateverbydefinitionliesbeyondthelimitsofknowledge.Scienceandcommonsenseprovideuswithinformationaboutwhatlieswithinthelimitsoftheknow-able.Butaspecial,non-scientific,peculiarlymetaphysicalinvestigationisrequiredifwearetoacquireanyknowledgeoftheverylimitsoftheknowablethemselves.Kantcallssuchaninvestigation‘transcendental’:‘Ientitletranscendentalallknowledgewhichisoccupied,notsomuchwithobjects,aswiththemodeofourknowledgeofobjectsinsofarasthismodeofknowledgeistobepossibleapriori.Asystemofsuchconceptsmightbeentitledtranscendentalphilosophy’(Kant1929:B25).TheheartofKant’smetaphysicsiscontainedintheTranscendentalAnalytic,foritisherethatheprovideshisanalysisnotonlyofhowthemindmustworkifunderstand-ingorknowledgearetobepossible,butalsoofhowtheworldmustbeifitistobecapableofbeingknownorunderstood.TheremainderofthissectionisdevotedtoanexaminationofKant’saccountintheTranscendentalAnalytic.Oursensespresentuswithamultiplicityofdifferentsensationsor,inKant’stermi-nology,withamanifoldofintuitions.Theseintuitionsareextendedintime,andsomeofthemhavespatialproperties,buttheirintrinsicnatureguaranteesthemnoperma-nence,stability,determinacy,objectivityorconceptualarticulation.Amanifoldofintu-itionsiswhatHUME(chapter31)hadinmindwhenhewroteofconsciousnessascontainingsensorycontents‘whichsucceedeachotherwithaninconceivablerapidity,andareinaperpetualfluxandmovement’.Nowifourawarenessisevertobe,say,anawarenessofanintelligible,predictableworld,containingdeterminate,stableobjectsthatpersistthroughtheirchangesandinteractcausallywithoneanother,thenwemustbecapableofunifyingandinterpretingthismanifoldofintuitions.Thatis,wemusthavetheabilitytofindourwayaboutwithinthediversityofourever-changingsensoryexperience.Andwedothisbyconstruingthosesensorystatesasexperiencesofobjectsintheexternalworld.Kanttakesitasaxiomaticthat‘thecombinationofamanifoldingeneralcannevercometousthroughthesenses’,andheconcludesthat‘allcombination,beweconsciousofitornot...isanactofunderstanding’.Hethenadds:‘Tothisactthe731\nDAVIDBELLgeneraltitleof“synthesis”maybeassigned’(ibid.:B130).Inshort,ifwearetobeabletomakeanysenseofthediversityofsensationsweundergo,wemustbeabletocombineorunitethosesensationssothattheycometorepresenttousaspectsofobjectsintheworld.Forinstance,Imaytakeamanifoldcomprisingcertainparticulartactilesensations,sensationsofshape,colourandmovement,sensationsofsound,smellandthelike,asbelongingtogether,bytakingthemtobevariousformsofawarenessofoneandthesameobject–mymotorbike,say.Totakeavarietyofdisparatesensoryintuitionsasbelongingtogether,invirtueoftheirbeingdifferentexperiencesofoneandthesamething,justistosynthesizethoseintuitions.Tounifyone’ssensorystatesistounderstandthemtoberepresentationsofobjectsandpropertiesintheexternalworld.So,intheabsenceofsuchsynthesizingactivities,Kantclaims,wecouldhavenoawarenessofanexternalworld,andindeednoexperiencethatwouldmakeanysensetous.Synthesiscannot,ofcourse,beamerelyrandomorarbitraryprocess,butmustproceedinconformitywithrules.Moreover,themostfundamentalrulesgoverningsyn-thesiscannotbelearnedoracquiredonthebasisofexperience,fortheyarepreciselytheruleswhichmustalreadybeinplaceifthereistobeanyknowledgeorcoherentexperience.Kantthereforeconcludesthatthebasicrulesgoverningsynthesisareapriorirules–andthemostfundamentalofallaretheaprioriruleswhichenableanumberofdifferentsensorystatestobereferredtooneandthesameobject.NowaccordingtoKant,conceptsarenothingbutrulesforsynthesis,andthecategoriesarethemostbasic,apriorisuchrules.Sothecategoriestogetherspecifytherulestowhichhumanunderstandingmustconformifwearetobeabletotakeourvarioussensorystatesasrepresentingobjectsintheworld.InKant’sownterminology,thecategoriestogethercomprise‘theconceptofanobjectingeneral’.Kantwritesthatthecategories‘serveastheantecedentconditionsunderwhichaloneanythingcanbe...thoughtasobjectingeneral’(ibid.:B125).IntheMetaphysicalDeductionKantattemptstoidentifyjustthoseconceptswhicharecategories.(Therearetwelveofthem:Unity,Plurality,Totality,Reality,Negation,Limitation,Substance,Causality,Community,Possibility,ExistenceandNecessity.)InthesectioncalledtheTranscendentalDeductionoftheCategories,Kantoffersacomplexsetofargumentsaimedatprovingtheclaimthatintheabsenceofsynthesisinconformitytoapriorirules(thecategories),noobjective,intelligibleexperiencewouldbepossibleatall.OnecentraltrainofthoughtthatrunsthroughtheTranscendentalDeductionis,inbriefestoutline,asfollows.Thereisaninescapableconditionthatmustbemetbyanypluralityofmentalstatesandmentalcontentsiftheyarealltobelongtoone,single,unifiedconsciousness–if,forexample,theyarealltobemystatesandcontentsofcon-sciousness.Thenecessarycondition,accordingtoKant,isthattheconsciousnessormindinquestionshouldpossessself-consciousness:Imustnotmerelyhavethecapac-itytobeinanarbitrarymentalstate,S,ImustalsobeabletoknowthatIaminthatstate.Andthisrequires,inturn,thatIhavetheabilitytoascribestateStomyself,roughlyspeakingbybeingabletomakethejudgement,‘IamawareofS’.Kantwrites:‘Itmustbepossibleforthe“Ithink”[or“Iamaware”]toaccompanyallmyrepresen-tations’(ibid.:B131);‘forthemanifoldofrepresentations...wouldnotbeoneandallmyrepresentations,iftheydidnotallbelongtooneself-consciousness’(ibid.:B132).732\nKANTInotherwords,theunityofconsciousnessrequiresself-consciousness,andself-consciousnessinturnrequirestheabilitytoascribeone’smentalstatestooneself.TheprinciplewhichgovernstheunityofconsciousnessKantcallsthetranscendentalunityofapperception.TheTranscendentalUnityofApperceptionThistermdoesnotstandforanobjectorentityofanysort,andinparticularitisnotthenameasanythinglikeanego,self,soulormind.Rather,itisthenameoftheformalprin-ciplewhichisresponsiblefortheunityofconsciousness.‘Apperception’isanyformofconsciousnessthatinvolvesself-consciousness;andtheunityofapperceptionis‘tran-scendental’becauseitisanaprioriconditionforthepossibilityofintelligibleexperience.Sothetranscendentalunityofapperceptionisaconditionthatconsciousnessmustmeetifitistocompriseacoherent,integratedunityorwhole–thecondition,namely,thatitbeaformofself-consciousness.Kantexpressestheconditionthus:‘Itmustbepossibleforthe“Ithink”toaccompanyallmyrepresentations’.Ifthisconditionweretofail,heclaims,thenself-consciousnesswouldbeimpossible,inwhichcasetheunityofcon-sciousnesswouldbeimpossible,inwhichcaseonecouldhavenointelligibleexperienceofanykind.InthefinalstageintheTranscendentalDeductionKantexaminesinmoredetailtheconditionsthatmustobtainiftheascriptionofdifferentexperiencestooneselfistobepossible:whatexactlymustIbeabletodo,ifIamtohavethecapacitytobeself-conscious?P.F.Strawsonwrites:‘Kant’sanswermustbeadmittedtohaveacertainsublimity.Whatisrequiredforaseriesofexperiencestobelongtoasingleconscious-nessisthattheyshouldpossesspreciselythatrule-governedconnectednesswhichisalsorequiredforthemcollectivelytoconstitute[the]experienceofasingleobjectiveworld’(Strawson1966:92–3).Inotherwords,theunityofmyconsciousnessrequiresmetosynthesizemydifferentsensoryexperiencesinconformitywiththerulesembodiedinthecategories.Butformetosynthesizemyvariousexperiencesinthiswayjustisformetoexperienceasingleobjectiveworld–aworldmadeupofstable,materialobjectsthatpossessproperties,persistthroughtimeandinteractcausallywithoneanother.OnthebasisofthisargumentKantfeelsjustifiedinconcludingthatknowledge,intelligibleexperience,eventheunityofconsciousnessitself,wouldbeimpossibleifwedidnotpossessandapplytheaprioricategoriesinwaysthatprovideuswithexperienceofanobjective,externalworld.‘Inotherwords’,asKanthimselfsays,‘theconscious-nessofmyexistenceisatthesametimeanimmediateconsciousnessoftheexistenceofotherthingsoutsideme’(Kant1929:B276).3AutonomyandtheMoralLawKant’smajorcontributionstoETHICS(chapter6)arecontainedintwoworks:TheFoundationsoftheMetaphysicsofMorals(1785)(whoseoriginaltitle,Grundlegung733\nDAVIDBELLzurMetaphysikderSitten,isvariouslytranslatedintoEnglish,notonlyasTheGrounding,TheGroundwork,TheFoundationsorasTheFundamentalPrinciplesoftheMetaphysicsofMorals,butalsounderthetitleTheMoralLaw),andthesecondCritique,entitledCritiqueofPracticalReason(1956).Thefirstoftheseworksisbeyonddoubtoneofthemostprofoundandinfluentialworksofmoralphilosophyeverwritten.Thefirstsectionopensresoundinglywiththefamouswords:‘Itisimpossibletoconceiveofany-thingintheworld,orindeedanythingbeyondtheworld,whichcanbeconsideredgoodwithoutqualification,exceptagoodwill.’And,asKanthimselfputsit,‘thesoleaim’ofthebookisto‘seekoutandestablishthesupremeprincipleofmorality’.Kant’saim,inotherwords,istoidentifyanddefendthemostbasicmoralprincipleorlawwhichdirectlydeterminestheETHICALVALUE(p.203)possessedbyactsofWILL(p.219),andwhich,asaconsequence,indirectlydeterminesallothermoralvalueswithoutexception.HecallsthissupremeprincipleofmoralitytheCategoricalImperative.Theunderlyingpicture,intermsofwhichwecanbegintomakesenseofKant’smoraltheory,isasfollows.Likeallanimals,humanbeingsaregovernedtoagreatextentbytheirdesires:muchofourbehaviourisintelligiblewhenitisseenasdirectedtowardsendsthataredeterminedbyourwishes,passions,appetites,desiresandthelike.Inthisrespect,asHumesaid,‘reasonisandoughttobetheslaveofthepassions’.Heretheroleofpracticalreasonismerelytosupplyrational,efficientmeansfortheattainmentofthethingswewant.Ontheotherhand,andunlikemanyanimals,wearecapableincertaincircumstancesofactingonprinciple,andoffollowingthedictatesofreason–evenwhenthisinfactgoesagainstourappetitesanddesires.Wedothiswhen,forinstance,wekeepapromiseatsignificantcostorinconveniencetoourselves,whenwevolunteerforsomeunpleasanttaskpurelyoutofasenseofduty,orwhenwedecidewemust‘dotherightthing’inspiteofastronginclinationtodootherwise.Whenweactinthisway,Kantsays,themoralworthofouractiondoesnotreside‘inthepurposetobeattainedbyit,butinthemaximinaccordancewithwhichitisdecidedupon’.Amaximisageneralruleorprinciplewhichapersoncanuseasaguidetoaction:‘Alwaystellthetruth’;‘Ifyouthinkitwillrain,carryanumbrella’;‘Don’talwaystakethebiggestpieceofcake’;‘NevereattroutonaThursday’,andsoon.Astheseexamplesshow,however,noteverymaximisamoralone;fortherearemaximsthatexpressprinciplesofprudence,say,oretiquette,orevenwhimandsuperstition.Foranactiontobeapurelymoralone,Kantdemandsthatthreeconditionsmustbemet:1Themaximgoverningtheactionmustbeagenuinelymoralmaxim(andnotoneofmereprudence,say,oretiquette).2Theactionitselfmustnotmerelyconformtothemaxim,butrathermustbedoneforthesakeofthemoralprinciplethatthemaximembodies.Apurelymoralaction,inotherwords,isonemotivatedexclusivelybyadesiretoimplementamoralprin-ciple.Itisonethatisperformed,asKantsays,‘exclusivelyfromasenseofduty’and‘forthesakeofduty’.3Theactionmustbeperformedfreely,byanautonomousagent.734\nKANTAutonomyKantbelievedthatmature,adulthumanbeingsareautonomous,andthatitisforthisreasonthattheypossessmoraldignity,deservemoralrespect,andoughtnevertobetreatedasmeremeanstoanend,butalwaysasendsinthemselves.Possessionofgenuineautonomydependsuponthepresenceofthreethings.Firstly,anautonomousagentmustbefree,thatis,capableofactsofwillthatarenotcausallydetermined,eitherbyinnerforceslikeinclinations,desiresandpassions,orbyexternalforcesintheouterworld.Sec-ondly,autonomyrequiresthatagentscanexerciserationalself-controlovertheirdeci-sionsandtheiractions.And,thirdly,autonomyrequiresthattherationalprinciplesgoverningsuchself-controlarethesoleresponsibilityoftheindividualagent:nogen-uinelyautonomousagentcanrelinquishresponsibilityfortheprinciplesonwhichheorsheactstoanyexternalauthority–whetheritbethechurch,thestate,thelaw,society,family,teachersorfriends.Onthecontrary,allsuchprinciplesmustbechosenandimposedbytheagentalone.Andthesupremeprincipleofrational,moralautonomyistheCategoricalImperative.Atthispointweneedtoask:whendoesamaximembodyanintrinsicallymoralprinciple?How,thatis,arewetotellagenuinelymoralmaximfromonethathasnointrinsicmoralcontentorrelevance?Theanswertothisquestionisstatedbelow,atthebeginningofthefollowingparagraph.Beforeprovidingtheanswer,however,somepreliminaryscene-settingisrequired.Wecanpredict,inadvance,thattheanswerwillbehighlyabstractandgeneral–afterall,weareaskingfornothinglessthanaprinci-plewhichwillenableustoidentifyallandonlythosemaximsthataremorallygood,irrespectiveofwhere,whenandbywhomtheyareformulated,andregardlessoftheparticularacts,orgoals,oragentstheyareintendedtogovern.Wecanalsopredictthattheprinciplewillbebothsyntheticandapriori.Itmustbeapriori,forinsofarasitistodeterminethemoralvalueofanypossibleact,performedbyanypossiblerationalagent,inanypossiblecircumstance,itmustclearlybeindependentofallthatismerelycontingentlythecase.Inparticular,itmustbeindependentofwhatevercontingentdesiresandinclinationsanagentmightpossess,andofwhatevercontingentconse-quencesanactionmighthaveinparticularcircumstances.Theprinciplemustalsobesyntheticbecauseamerelytrivial,analyticanswertoourquestionwouldbeincapableofprovidingthesubstantiveknowledgeweseek.Andfinally,theanswertoourques-tionwillitselfbeamoralprinciple,albeitaverygeneral,abstractone.Whatwerequireatthispointisnotadescriptiveassertionspecifyingwhatisthecaseandthewaysinwhichwedoinfactact,butratheranormativerulewhichwillspecifywhatshouldbethecaseandhowweoughttoact.Moralprinciples,thatis,havetheprimaryfunctionofregulatingandevaluatingbehaviour,andassuch,Kantbelieved,theyhaveimpera-tiveforce.Sowhatweseek,inanswertoourquestion,isa(second-level)moralimperativethatspecifieswhichmaxims(first-levelimperatives)weoughttoadoptasreasonsforaction.AmaximcomprisesamorallygoodbasisforactionwhenitconformstotheCategoricalImperative;andtheCategoricalImperativeispreciselyasecond-level,universallyapplicable,abstract,normative,apriori,syntheticrulefortheevaluationof735\nDAVIDBELLfirst-levelmaxims.InitsmostfamousformulationtheCategoricalImperativereads:‘Actonlyonthatmaximthroughwhichyoucanatthesametimewillthatitshouldbecomeauniversallaw.’TheCategoricalImperativeAnimperativeisacommand,instructionorrulegoverninghowoneshouldact.Animperativeis‘categorical’whenitisexceptionless,thatis,whenitisbindingonallratio-nalagents,inallcircumstances,atalltimes.Kantbelievedthatwhathecalled‘thesupremeprincipleofmorality’wasjustsuchacategoricalimperative,andheprovidedanumberofdifferentformulationsofit.Themostimportantare:1Actonlyonthatmaximthroughwhichyoucanatthesametimewillthatitshouldbecomeauniversallaw.2Soactastousehumanity,bothinyourownpersonandinthepersonofeveryother,alwaysatthesametimeasanend,neversimplyasameans.3Soactasifyouwerealwaysthroughyourmaximsalaw-makingmemberofakingdomofends.TheCategoricalImperativeis,forKant,anapriori,abstractlawwhichgovernsthemoralvalueofthemaximsonwhichweact–maximswhich,inturn,determinethemoralvalueofthoseactsthemselves.Soanactismorallygoodifitisperformedforthesakeofamorallygoodmaxim;andamaximismorallygoodifitconformstotheCategoricalImperative.TherelationofKant’smoraltheorytohismetaphysicsofexperienceiscomplexandobscure.Onething,however,isclear.IntheCritiqueofPureReasonKanthadattemptedtoestablishthelimitsofgenuineknowledge,thatis,todiscovertheaprioriconditionsthatmustobtainifknowledgeistobepossible.Oneofhisconclusionswasthat,forus,therecanbenoknowledgeofanysuper-sensible,non-empiricalreality.Inotherwords,anecessaryconditionofthepossibilityofhumanknowledgeisthatitinvolvenotonlyintellectualconcepts,butalsosensoryintuitions.ForKant,therefore,therecouldbenoknowledgeofGod,theimmortalityofthesoul,orthefreedomofthewill.Tosuchmattersasthese,hemaintained,conceptslikeobservation,discovery,evidence,truth,science,explanationandthelikemustremainforeverinapplicable.Butaboutsuchmatterswecanneverthelesshavefaith.FaithinGod,commitmenttothemorallaw,respectforotherpersons,appreciationofthedignitythatautonomybestows–thesewereforKantamongthehighestidealsthathisphilosophy,asawhole,wasintendedtoprotect.Andinparticular,hismetaphysicsofexperienceandhisaccountofthelimitsofknowledgewereintendedtoshowthatnoempirical,naturalisticorscientifictreat-mentofthesetranscendentmatterscouldpossiblybeintelligible.AshesaysintheprefacetothesecondeditionoftheCritiqueofPureReason:‘Ihavethereforefounditnecessarytolimitknowledge,inordertomakeroomforfaith’(Kant1929:Bxxx).FurtherReadingForthosewhoarenewtoKant,anumberofgeneral,introductoryworksareavailablewhichcoverallormostofthemajoraspectsofKantianthought.Copleston(1964)presentsaconsci-736\nKANTentious,accessibleanddetailed,thoughratherunexciting,overview.Scruton(1982)isbrieferandmorespirited.Kemp(1968)andKörner(1955)arebothreliableintroductions,thoughtheyareslightlylesselementarythanthosementionedabove.Walker(1978)isanotherexcellent,thoughnon-elementary,introductiontoKant’sthoughtasawhole.AmongintroductoryworksthatconcentratespecificallyonthemesfromtheCritiqueofPureReasonthefollowingarenoteworthy.Broad(1978)presentsathorough,helpfuldiscussionofthefirstCritique.Wilkerson(1976)issometimeshelpfulasanintroduction,thoughitistoooftenunsympatheticandimpatient.Walsh(1975)writeswithexemplaryclarityandsophistication,asdoesBird(1962),agoodintroductiontoKantianepistemology.Forrathermoreadvancedreaders,Bennett(1966and1974)areidiosyncraticandiconoclastic,butcanassistnotonlyinunderstandingKant’sthought,butalsoinfindingitphilosophicallyexciting.Schwyzer(1990)providesasympatheticandaccessibletreatmentofsomeofthemostcentralconcernsofthefirstCritique.Allison(1983)isclear,scholarlyandinsightful.AlthoughStrawson(1966)hasbeenhugelyinfluential,itisperhapsmostsuitableforadvancedreaders.PaulGuyer’s(1987)exhaus-tivestudy,thoughdoubtlessimportant,istoodenseandtoolongtoberecommendedtoanybutthemostdeterminedreader.Onindividualphilosophicaltopics,importantandhelpfulmaterialcanoftenbefoundincol-lectionsofarticles(thoughthismaterialisrarelyofanelementaryorintroductorynature).AmongthemostaccessibleandusefulofsuchanthologiesareWalker(1982),Wolff(1967),Guyer(1992)andSchaperandVossenkuhl(1989).Forthoseseekinghelpinunderstandingparticularsections,passages,orevenwordsandphrasesintheCritiqueofPureReason,anumberofcommentariesareavailablewhichprovidedetailedtextualclarification.AmongthemostimportantareKempSmith(1923),Paton(1936)andEwing(1938).AnintroductiontotheoverallphilosophicalcontextofKant’smoralphilosophycanbeobtainedfromSinger(1961)andBroad(1930).Sympathetic,scholarlyandcleartreatmentsofKantianethicsaretobefoundinAllison(1990),Sullivan(1989),Paton(1947)andActon(1970).Thereare,inaddition,anumberofexcellentcommentariesonKant’sethicalwritings.AmongthebestareBeck(1960),Ross(1954)andWolff(1973).ReferencesKantKant,I.1783:ProlegomenatoAnyFutureMetaphysics.Invarioustranslations:translatedbyL.W.Beck,Indianapolis,IN:Bobbs-Merrill(1950);translatedbyP.G.Lucas,Manchester:ManchesterUniversityPress(1953);translatedbyJ.Ellington,Indianapolis,IN:Hackett(1985).——1785:TheGroundworkoftheMetaphysicsofMorals.Invarioustranslations:translatedbyH.J.Paton,London:Hutchinson(1949);translatedbyL.W.Beck(asTheFoundationsoftheMetaphysicsofMorals),Indianapolis,IN:Bobbs-Merrill(1959).——1790:CritiqueofJudgement.Invarioustranslations:translatedbyJ.Meredith,Oxford:ClarendonPress(1952);translatedbyJ.H.Bernard,NewYork:Haffner(1968);translatedbyW.S.Pluhar,Indianapolis,IN:Hackett(1987).——1929[1781,1787]:CritiqueofPureReason(1stedn(A),1781,2ndedn(B),1787)(translatedbyN.KempSmith).London:Macmillan.——1956[1788]:CritiqueofPracticalReason(translatedbyL.W.Beck).Indianapolis,IN:Bobbs-Merrill.——1968:SelectedPre-CriticalWritingsandCorrespondencewithBeck(translatedbyG.B.KerfordandD.E.Walford).Manchester:ManchesterUniversityPress.737\nDAVIDBELLCambridgeUniversityPressiscurrentlypreparingnewEnglishtranslationsofallKant’spub-lishedworks,undertheseriestitleTheCambridgeEditionoftheWorksofImmanuelKant.ItsgeneraleditorsarePaulGuyerandAllenW.Wood.Kant’slifeandhistoricalbackgroundBeck,L.W.1969:EarlyGermanPhilosophy:KantandHisPredecessors.Cambridge,MA:HarvardUniversityPress.Cassirer,E.1951:ThePhilosophyoftheEnlightenment(translatedbyF.C.A.KoellnandJ.P.Pettegrove).Princeton,NJ:PrincetonUniversityPress.——1981:Kant’sLifeandThought(translatedbyJ.Haden).NewHaven,CT:YaleUniversityPress.Gulyga,A.1987:ImmanuelKant:HisLifeandThought(translatedbyM.Despalatovic).Boston:Birkauser.GeneralsurveysofKant’sthoughtBeck,L.W.1965:StudiesinthePhilosophyofKant.Indianapolis,IN:Bobbs-Merrill.Copleston,F.1964:AHistoryofPhilosophy,vol.6,partsIandII.NewYork:ImageBooks.Guyer,P.(ed.)1992:TheCambridgeCompaniontoKant.Cambridge:CambridgeUniversityPress.Kemp,J.1968:ThePhilosophyofKant.Oxford:OxfordUniversityPress.Körner,S.1955:Kant.Harmondsworth:PenguinBooks.Scruton,R.1982:Kant.Oxford:OxfordUniversityPress.Walker,R.C.S.1978:Kant.London:RoutledgeandKeganPaul.Wolff,R.P.(ed.)1967:Kant:ACollectionofCriticalEssays.London:Macmillan.MetaphysicsandepistemologyAllison,H.E.1983:Kant’sTranscendentalIdealism.NewHaven,CT:YaleUniversityPress.Ameriks,K.1982:Kant’sTheoryofMind:AnAnalysisoftheParalogismsofPureReason.Oxford:ClarendonPress.Aquila,R.E.1983:RepresentationalMind:AStudyofKant’sTheoryofKnowledge.Bloomington:IndianaUniversityPress.Bennett,J.1966:Kant’sAnalytic.Cambridge:CambridgeUniversityPress.——1974:Kant’sDialectic.Cambridge:CambridgeUniversityPress.Bird,G.1962:Kant’sTheoryofKnowledge.London:RoutledgeandKeganPaul.Broad,C.D.1978:Kant:AnIntroduction.Cambridge:CambridgeUniversityPress.Ewing,A.C.1938:AShortCommentaryonKant’sCritiqueofPureReason.Chicago:UniversityofChicagoPress.Friedman,M.1992:KantandtheExactSciences.Cambridge,MA:HarvardUniversityPress.Guyer,P.1987:KantandtheClaimsofKnowledge.Cambridge:CambridgeUniversityPress.KempSmith,N.1923:ACommentarytoKant’s‘CritiqueofPureReason’.London:Macmillan.Kitcher,P.1990:Kant’sTranscendentalPsychology.Oxford:OxfordUniversityPress.Paton,H.J.1936:Kant’sMetaphysicofExperience:ACommentaryontheFirstHalfoftheKritikderreinenVernunft.London:GeorgeAllenandUnwin.Pippin,R.B.1982:Kant’sTheoryofForm:AnEssayontheCritiqueofPureReason.NewHaven,CT:YaleUniversityPress.Schaper,E.andVossenkuhl,W.(eds)1989:ReadingKant:NewPerspectivesonTranscendentalArgumentsandCriticalPhilosophy.Oxford:Blackwell.Schwyzer,H.1990:TheUnityofUnderstanding:AStudyinKantianProblems.Oxford:ClarendonPress.Strawson,P.F.1966:TheBoundsofSense.London:Methuen.Walker,R.C.S.(ed.)1982:KantonPureReason.Oxford:OxfordUniversityPress.738\nKANTWalsh,W.H.1975:Kant’sCriticismofMetaphysics.Edinburgh:EdinburghUniversityPress.Wilkerson,T.E.1976:Kant’sCritiqueofPureReason:ACommentaryforStudents.Oxford:OxfordUniversityPress.Wolff,R.P.1963:Kant’sTheoryofMentalActivity:ACommentaryontheTranscendentalAnalyticoftheCritiqueofPureReason.Cambridge,MA:HarvardUniversityPress.MoralphilosophyActon,H.B.1970:Kant’sMoralPhilosophy.London:Macmillan.Allison,H.E.1990:Kant’sTheoryofFreedom.Cambridge:CambridgeUniversityPress.Beck,L.W.1960:CommentaryonKant’sCritiqueofPracticalReason.Chicago:UniversityofChicagoPress.Broad,C.D.1930:FiveTypesofEthicalTheory.London:RoutledgeandKeganPaul.O’Neill,O.1989:ConstructionsofReason:ExplorationsofKant’sMoralPhilosophy.Cambridge:CambridgeUniversityPress.Paton,H.J.1947:TheCategoricalImperative:AStudyinKant’sMoralPhilosophy.London:Hutchinson.Ross,D.1954:Kant’sEthicalTheory:ACommentaryontheGroundworkoftheMetaphysicsofMorals.Oxford:ClarendonPress.Singer,M.1961:GeneralizationinEthics.NewYork:AlfredA.Knopf.Sullivan,R.J.1989:ImmanuelKant’sMoralTheory.Cambridge:CambridgeUniversityPress.Williams,T.C.1968:TheConceptoftheCategoricalImperative.Oxford:ClarendonPress.Wolff,R.P.1973:TheAutonomyofReason:ACommentaryonKant’sGroundworkoftheMetaphysicsofMorals.NewYork:HarperandRow.DiscussionQuestions1How,ifatall,doesKantreconciletheclaims(1)thatallouractionsarecausallydetermined,and(2)thatasmoralagentswecanactfreely?2WhichofKant’svariousformulationsoftheCategoricalImperativeisethicallythemostdefensible?3‘Thoughtswithoutcontentareempty,intuitionswithoutconceptsareblind.’Whatdoesthismean?Isittrue?4HowconvincingareKant’scentralargumentsintheTranscendentalDeduction?5DoesKant’streatmentoftheTranscendentalUnityofApperceptioncompriseasolutiontoHume’sproblemsconcerningtheself?6DidKantshowthattherecouldbeno‘ontologicalproof’ofGod’sexistence?7HowadequateanaccountofcausalitydoesKantprovideintheSecondAnalogy?8WhatareKant’saimsintheSchematismchapterofthesecondCritique?Howsuccessfullydoesherealizethoseaims?9Istimeanecessaryformofallsensoryexperience?Ifso,doesthisgiveusanygroundsforclaimingthattimeistranscendentallyideal?10HowgoodisthecasethatKantmakesfortheclaimthatwehaveknowledgethatisbothsyntheticandapriori?11Whatistherelation,ifany,betweenmoralityandrationality?12Arethereanynecessarylimitstowhatcanbeknown,andifso,howcouldweestablishwhattheyare?13Whichismoremorallysignificant:themotivesfromwhichagivenactionisperformed,ortheconsequencesthatthatactionhas?739\nDAVIDBELL14Canoneconsistentlybeatranscendentalidealistandatthesametimeanempiricalrealist?15Arethereanyaprioriconcepts?Ifso,howdidweacquirethem?16Arethetruthsofgeometry(orarithmetic)syntheticandapriori?17Cananobjectsimplycomeintoexistence,outofnowhereandoutofnothing?18Ifwecannotperceiveidentity(ornecessity;orsubstance;orcausality),howcanwehaveanyempiricalknowledgeofit?19Couldamoralagentlackautonomy?20Whatismetaphysics?740\n33HegelMICHAELINWOODHegelcreatedagreatanddeeplyinfluentialsystemofphilosophy,rangingovermetaphysics,epistemologyandlogic;ethics,political,socialandlegalphilosophy;andaesthetics,religionandthenatureofphilosophyitself.Hewasfundamentallyconcernedwithrationality,freedomandself-consciousness,butsawthemashistoricalphenom-ena,developingthroughanintelligiblesuccessionofforms,movedby‘spirit’andstruc-turedby‘dialectic’.Hegel’sresponsetoKantprovidesaninitialbasisforunderstandinghisdifficulttexts.HisworkilluminatesthethoughtofhiscontemporariesFichteandSchellingandlateridealistsaswellasfeaturesofMarx’swritings.ManyEnglish-speakingphilosophersturnedawayfromHegelianismaftercriticismbyRussellandMooreatthebeginningofthetwentiethcentury,butthelasttwentyyearshasseenastrongrevivalofinterestinmanyaspectsofHegel’sthought.1IntroductionGeorgWilhelmFriedrichHegel(1770–1831)wasborninStuttgart,Germany,in1770.HetrainedasatheologianattheTübingentheologicalseminary,wherehisroom-mateswereHölderlin(1770–1843)andSchelling(1775–1854),andthroughouthislifeheremainedadevoutLutheran.Afteraperiodasahouse-tutorinBern(1793–7)andFrankfurt(1797–1800),hebegantolecture,in1801,attheUniversityofJena,whereSchellingwasalreadyprofessorofphilosophy;herehepublishedhisfirstbook,ontheDifferencebetweenFichte’sandSchelling’sSystemofPhilosophy(Hegel1977a),andco-editedwithSchellingTheCriticalJournalofPhilosophy(published1802–4).Napoleon’svictoryatJenain1806interruptedHegel’scareerandin1807hebecameeditorofaNapoleonicnewspaperinBamberginBavaria.Hisfirstmajorwork,thePhe-nomenologyofSpirit(Hegel1977b),wascompletedin1806andpublishedin1807.In1808hebecamerectorofahighschoolinNuremberg,wherehemarriedMarievonTucherin1811andpublishedtheScienceofLogic(Hegel1969)inthreepartsbetween1812and1816.In1816hebecameaprofessoratHeidelbergandin1817publishedhisEncyclopaediaofthePhilosophicalSciences(Hegel1991a),whichwasoriginallyintendedasabareoutlinetoaccompanyhislecturesbutwasgreatlyexpandedinfurthereditionsin1827and1830.From1818untilhisdeath(ofcholera)in1831hewaspro-fessorofphilosophyatBerlin,wherehepublishedThePhilosophyofRight(Hegel1991b)\nMICHAELINWOODin1821.HewasburiedinBerlinnexttoFichte,hispredecessorasprofessorofphiloso-phyatBerlin.Hisfriendsandpupilseditedhisworksposthumously,includinghislec-turesonaesthetics,history,thephilosophyofreligionandthehistoryofphilosophy.Hegel’sfirstsignificantwritings,producedbetween1793and1801butnotpublisheduntil1907,dealtwithreligiousquestionssuchas‘HowdidChristianitydegeneratefromareligionofloveintoa“positive”religionofrulesanddogmas?’and‘Canthecomplex-ityandfluidityoftheworldasreligionconceivesitbeadequatelycapturedinphilosoph-icalandconceptualterms?’Atthisstagehetendedtodisfavourthemethodicalrigouroftheconceptualunderstanding.Theworldistobeconceivedintermsof‘life’and‘love’,notionswhich,heheld,cannotbecapturedinphilosophicalconcepts;hepreferredthenaturalsimplicityofearlyChristianity,oroftheGreekcity-state,tothecomplexorderofmodernbourgeoissocietyanditsreligion.ButonmovingtoJenahebegantodevelopaphilosophyadequatefortheconceptualizationoftheworldasawhole,andespeciallyofthemodernworld.Thistaskoccupiedhimfortherestofhislife.Hegel’sResponsetoKantHegel’sknowledgeofthephilosophyofthepast,especiallyGREEKPHILOSOPHY(chapters22and23),wasbothwideanddeep.ButtheimmediatecontextofhisphilosophicalactivitywasthethoughtofKANT(chapter32)andthephilosophersinspiredbyit,notablyFichteandSchelling.Likethem,HegelwasdissatisfiedwithKant’sviewthatwecannotknow‘things-in-themselves’:Kant’sattempttoassignlimitstoourknowledgeisself-defeating,sincetobeawareofalimitistotranscendit.HegeloftenspokeofKantasa‘subjectiveidealist’andarguedthatthecategoriesorthoughtsthatstructureourexpe-riencearenot,asKantheld,simplyimposedbyusonoursensoryintuitions,butareembeddedinthenatureofthings.HefeltthatKant’sthoughtwasinsufficientlysystem-atic,andthatthiswaslargelyduetoKant’sfailuretoexamineourcategories,andtheirsystematicinterconnections,insufficientdepth.Thecategoriesarenot,asKantsup-posed,thesameforallpeopleatalltimes;theydevelopoverHISTORY(chapter14),nothaphazardly,butinawaythatbroadlyreflectstheirlogicalorder.WhileKantfocusedonmathematicsandthenaturalsciencesastheprimaryformsofourcategorialknowledge,Hegelregardedeveryspecificallyhumanenterprise–morality,socialinstitutions,politi-callife,art,religionandphilosophy–asawayinwhichwecometounderstandourworldandourselves;theenterprisesarenotallofequalstatus–philosophy,forexample,issuperiortoart–buttheyallhelptodomesticatetheinitiallyalienworldandtomakeusathomeinit.Ingeneral,Kant,asHegelviewedhim,acceptedthestarkoppositionsoftheunderstanding:betweenthoughtandreality,conceptandintuition,dutyandincli-nationandbetween‘is’and‘ought’;Hegelattemptstoovercometheseoppositionsinahighersynthesisof‘speculativereason’.2PhenomenologyofSpiritHegeldevelopedtheseideasreflectinghisresponsetoKantinhislecturesatJena,buthefirstgavethempublic,ifsomewhatchaotic,expressioninthePhenomenologyofSpirit(Hegel1977b).TheworkbeginswithanEPISTEMOLOGICAL(chapter1)problemstem-mingfromKant:ifweregardcognitionasaninstrumentthatweapplytotheobject,742\nHEGELorelseasamediumbywhichthe‘ray’fromtheobjectistransmittedtous,wecannotbesurethattheobjectisnotdistortedbycognitionandthataccessto‘theabsolute’(thatis,thingsastheyareinthemselves)isnotforeverclosedtous.Hegelproposedtomeetthisdifficultybyexaminingnotobjectsdirectly,butthe‘forms’ofourconsciousnessofobjects,sincewecannotcoherentlysupposethatourknowledgeoftheformofourconsciousnessisdistortedinthesamewayasourknowledgeofobjectsmaybe.Inthiswaywecan,hebelieved,surveytheriseofknowledgeor‘science’,sinceaformofconsciousnessdevelopsintoanother,higherformasaresultofitsinternaldefects.Thuswebeginwith‘sensecertainty’,whichattemptstopickoutindividualitemsinitsexperiencebymeansofsuchtermsas‘this’,‘here’and‘now’.Itfindsthatitisunabletodothis,sincesuchtermsare‘universal’:anythingwhatsoever,forexample,canbereferredtoas‘this’.Sensecertaintyisthustransformedinto‘perception’,whichdealsinexplicitlyuniversalterms,regardingthingsasbearersofuniversalproperties.Plaguedbyitsinabilitytogiveacoherentaccountoftherelationbetweenathinganditsproper-ties,perceptiongiveswayto‘understanding’,whichviewsthesurfacequalitiesofthingsasthemanifestationofinnerforcesandthelawsthatgovernthem.Theformsofconsciousnessconsideredsofararesituatedinnoparticularsocialorhistoricalcontext.Butfromnowon–beginningwiththeaccountof‘self-consciousness’,inwhichtwopeoplefighttosecurerecognitionfromeachotherandthustoconfirmorrealizetheirself-consciousness–theworkbecomesareviewofhumanculturaldevelopment,examiningphasesofourhistory(suchasmedievalChristianityandtheFrenchRevolution)whichareneverthelessofgeneralsignificance.Hegelbelievedthatknowledgeisnotwhollyseparatefromourreligious,moralandpoliticalpractices;hebelievedthatitisasocial,interpersonalenterprise–theworkof‘spirit’orofthe‘IthatisaweandthewethatisanI’,ratherthanofaloneindividual–andthatitchangesinanintelligiblewayfromoneepochtoanother.SinceHegel’sownsystemwasthelatestphaseofthisevolution,thePhenomenologyattemptedtorecordthehistoricalpreconditionsofitsemergence.Atthesametimeitwasintendedasanintroductiontothissystem.Thereaderisledbysuccessivestagestothe‘absoluteknowledge’withwhichtheworkconcludes.AbsoluteknowledgeiswhatHegelhimselfhad;itenabledhimtoreview,inawaythatnoearlierformofconsciousnesscould,alltheformsbywhichitwasreachedandtodistilthelogicthatunderlaytheirdevelopment.3HegelianLogicThePhenomenologywasoriginallyconceivedasthefirstpartofHegel’ssystem,tobefollowedbylogicandthe‘sciences’ofnatureandspirit.Hedidnotfulfilthisplan,andthelogicsthatheeventuallyproduced,theScienceofLogic(Hegel1969)andthefirstvolumeoftheEncyclopaedia(Hegel1991a),wereconceivedindependentlyofthePhe-nomenology.(TheEncyclopaediaLogicisprecededbyanalternativeIntroduction,includ-ingacritiqueofpreviousGermanphilosophy;aphilosophicalsystemcannotdispensewithanintroduction–howeversimpleandapparentlyself-evidentourstarting-pointmaybe,weneedtoexplainwhyweareabletoadoptthatstarting-pointandwhatisthepointofdoingso,andthisexplanationformsanintegralpartofthesystem.)Logicisanexaminationoftheabstractthought-determinationsorcategoriesthatconstitute743\nMICHAELINWOODthecoreofthehumanmindandareinvolved(withvaryingdegreesofexplicitness)inourordinaryexperienceanddiscourse,inourpracticalactivitiesandinstitutions,inoursystematicwaysofunderstandingtheworld–suchasreligion,thesciencesandphilosophy–andintheverystructureoftheworlditself.ThefirstsectionoftheLogic,the‘DoctrineofBeing’,examinesthecategoriesinvolvedinthesurfacefeaturesofthings:itbeginswithpurebeing(theobviousstarting-point,sinceitisanallbutemptyconcept,littlemorethantheblankspaceofourthinking),which‘passesinto’nothing,andthusgivesriseto‘determinatebeing’,thedeterminatequalitythatmakessome-thing‘other’thansomethingelse.Fromthereitproceedsto‘quantity’.Thequantita-tivefeaturesofsomething,suchasthesizeofafield,areinitiallyseenasindependentofitsquality,ofthefactthatitisafield,butinthethirdphaseofthissection,‘measure’,qualityandquantityareinterdependent:if,forexample,afieldincreasesinsizesuffi-cientlyitbecomesaplainorprairieratherthanafield,andthemeasurementof‘inten-sive’quantities,suchastemperature,presupposesabruptqualitativechanges(suchasthatofwaterintoice)atcertainnodalpointsonacontinuousscale.Thesecondsection,the‘DoctrineofEssence’,considersthepairedconceptsintermsofwhichscientistsandtraditionalMETAPHYSICIANS(chapter2)haveunderstoodtherelationbetweentheoutersurfaceofthingsandtheirinnernature:thesepairedcon-ceptsincludeessenceandappearance,thingandproperties,substanceandaccident,causeandeffectandthereciprocalinteractionofsubstances.Thefinalsection,the‘DoctrineoftheConcept’,beginswiththe‘concept’,inpartasaresponsetothedifficultyofexplaining,forexample,thegrowthofaplantincausalterms,withoutreferencetotheconceptencodedintheseed.Butitexpandsintoanaccountoftheformsof‘subjective’logic:universal(suchas‘coloured’),particular(suchas‘red’)andindividualconcepts;thetypesofjudgement;andthetypesofinfer-enceorsyllogism.(Hegeltendedtoregardthesetooasformsnotonlyofourthinking,butalsoofobjectiveentitiessuchasthesolarsystemandthestate.)Next,Hegelturnedto‘objectivity’andconsideredthreeincreasinglysophisticatedwaysinwhichobjectsareconceived:mechanism,chemismandteleologyorpurposive-ness.Finally,hecomestothe‘idea’,theunificationofsubjectivityandobjectivity.Themostprimitiveversionofsuchunityis‘life’.Butfromthereheproceedstoourattemptstoachieveunityonahigherplane,incognitionof‘thetrue’andinrealizationof‘thegood’.TheclimaxoftheLogicisthe‘absoluteidea’,whichrepresentstheultimateunifi-cationofsubjectandobject,andservesseveralpurposesinHegel’ssystem:itrepresents,forexample,theenterpriseoflogicitself,inwhichthethinkerisnotdistinctfromthethoughtsthatheorshethinksabout–likethePhenomenology,theLogicconcludeswithasurveyofthestepsbywhichtheconclusionwasreached.Pastlogics(suchasARISTO-TLE’s(chapter23)orKant’s)were‘finite’:theyfailedtoaccommodatetheirownthoughtsaboutthethoughtsthattheydescribed.ButHegel’slogicis‘infinite’:theabsoluteideacontainsallthethoughtsrequiredtothinkadequatelyaboutitself.4NatureandSpiritTheabsoluteideaalsorepresentstherelationshipbetweenthepurethoughtswithwhichlogicdealsandtheconcreteworldsofnatureandspirit.ItistothesethatHegel744\nHEGELturnedinthesecondandthirdpartsofhisEncyclopaedia(Hegel1970,1971).Nature,inHegel’sview,doesnotdevelopovertime:itschangesarecyclicalandrepetitive.Itdoes,however,haveacomplexhierarchicalstructure,thesuccessivephasesofwhichembodysuccessivelyhighercategories:space,forexample,withwhichthePhilosophyofNature(Hegel1970)begins,approximatescloselytopurebeing,whiletheanimalorganism,withwhichitconcludes,canonlybeexplainedinconceptualterms.Hegeldidnotattempttodeducethedetailsofnaturefromlogic,buttriedtosupplyfromthenaturalsciencestheempirical‘content’oftherelativelyabstractframeworkcertifiedbylogic.Hewasnotloath,however,tocriticizethetheoriesofthedayanddistinctlypreferredthemorequalitativetheoriesofKeplerandGoethetoNewtonianismandatomism.ThethirdmainsectionoftheEncyclopaedia,thePhilosophyofMind(Hegel1971),is,liketheothersections,subdividedintothreeparts,dealingrespectivelywith‘subjectivespirit’–theINDIVIDUALMIND(chapter5);‘objectivespirit’–theinterpersonalSOCIALSTRUCTURES(chapter12)thathumanscreate;and‘absolutespirit’–ART(chapter7),RELIGION(chapter15)andphilosophy.ThePhilosophyofRight(Hegel1991b)givesafulleraccountofobjectivespiritor‘right’.Righthasthreemainphases:‘abstractright’,inwhichwebecomepersonsinvirtueofowningPROPERTY(pp.261–4)andaregovernedbyLAWS(chapter13)protectingourpersonsandpropertyandsanctionedbyPUNISHMENT(p.760)(whichisjustifiednotbydeterrentorUTILITARIAN(chapter35)considerations,butsolelybytheoccurrenceofaninfringementofright);‘MORALITY’(chapter6),primarilytheindividual(Kantian)moralityofconscience,accordingtowhich,forexample,oneisresponsibleonlyfortheintendedconsequencesofone’sacts;and‘ethics’or‘SOCIALMORALITY’(chapter8)(Sittlichkeit),theinstitutionsandnormsthatconferobjectiverightsanddutiesonusandmouldusintofullyfledgedhumanbeings.Ethicsinturninvolvesthreephases:thefamily,whosemembersareunitedbylove;‘civilsociety’,thefieldofeconomiccompetitionbetweenindividuals,whoseacquisitivedrivesareneverthelessmoderatedbylaws,policingand‘estates’(classes)and‘corporations’(tradeguilds);andtheSTATE(pp.261–4),whichrestoreshumanstounityonahigherlevelandmakesthemcitizensratherthansimplyeconomicagents.SpiritandtheStateHegel’sstateisafairlyliberal(butnotverydemocratic)constitutionalmonarchy,designedtoreconciletherightssecuredbytheFrenchRevolutionwithsocialstability.Itaccordsitscitizensboth‘objective’freedom,whichinvolvesaconformitybetweentherationalityofthesocialorderandtherationalitythatconstitutesthedeepestleveloftheirselfhood,andadegreeof‘subjective’freedom,whichinvolvesthefreedomtomakeper-sonalandmoralchoices,andwhichwasunknownintheGreekcity-state.AstateisessentiallyoneofasystemofstatesanditmustengageinwarfaretomaintainitsunitybydrawingthecitizenstogetheroutoftheirprivateINTERESTS(p.761).Stateswhichfailtodothisfallintodecay.Apeopleornationformingastatehasa‘spirit’ofitsown,awayofthinkingandactingthatpervadesitssocialandpoliticallife.Thesuccessiveriseandfallofstatesconstitutes‘worldhistory’,ahistorythatisnothaphazard,butratio-nal,eachnationalspiritbeingintelligiblyrelatedtoitspredecessors.Hegelmarksthisbyviewinghistoryastheworkofasinglespirit,the‘world-spirit’,whichcircuitously745\nMICHAELINWOODrealizesitsfinalgoal,thefreedomandself-consciousnessofhumanity,byexploitingthepassionsofhumanagentswhoareonlydimlyawareofthepurposestheyserve.Thespiritofapeopleinformsnotonlyitshistoricdeeds,butalsoitsart,religionandphilos-ophy,thoughathigherlevelsthesetendtobecomelessmarkedbynationalpeculiarities.Theyarethemainwaysbywhichapersongainsinsightintotheworldandintohimselforherself;inartapersondoessoinasensoryform,inreligionintheformofimagina-tionorVorstellung,andinphilosophybypurethought.5DialecticandSystemHegel’ssystemisknittogetherbydialectic,anddialecticalsooperatesinthenaturalandspiritualrealmsinvirtueofthethoughtsembodiedinthem.‘Dialectic’derivesfromtheGreekforthe‘artofconversation,dialogue’.Hegel’sdialecticdoesnot,however,involveadialoguebetweenphilosophersorbetweenthephilosopherandhisorhersubjectmatter.Thedialectical‘movement’isintrinsictothesubjectmatter,suchasformsofconsciousnessorcategories,andthephilosophersimplywatchestheirdevel-opment.Often,ifnotinvariably,thisinvolvestheresolutionoftwooppositesinahigherunity:forexample,thesimpleunityofthefamilyprovestobeinadequate(or‘contra-dictory’)onitsownandthusgivesrisetoitsoppositeor‘negation’,civilsociety,whichischaracterizedbytheindividualismthatenablesitsmemberstodeveloptheirownper-sonalitiesandinterests;butthisinturnprovesdefective,andrequiresforitscompletionthehigherunityofthestate,the‘negationofthenegation’which‘sublates’,oratoncedestroys,preservesandelevates,thetwoopposites.Historicalprocessesalsoexemplifythisdialectic:forexample,theunifiedcity-stateofantiquitygiveswaytotheindividu-alismofmodernity,butHegel’sproposedstatereconcilesthetwo,givingfreeplaytoindividualismwhilecontainingitwithinthedifferentiatedunityofthestate.AnotherfeatureofdialecticisthatSpiritessentiallyreflectsonitself:forinstance,logicisthinkingaboutthoughts,thephilosopherreflectsonpastphilosophiesandthehistorianreflectsonhistoricalevents.Butinreflectingonitself,Spiritchangesitself:tothinkaboutathoughtweneedmorecomplexthoughts;inthinkingaboutotherphilosophies,wedevelopanewphilosophy;historicalwritingraiseshumanstoanewlevelofawarenessandthuschangesthepatternsoftheirthoughtandconduct.ThusSpiritadvancesbycontinuallyreflectingonitself.Thiswayofproceedingisoftencom-binedwiththeresolutionofoppositions:forexample,Hegel,inreflectingonapparentlyopposedphilosophies(suchasSPINOZA’smonism(chapter27)andLEIBNIZ’spluralism(chapter27)),attemptstosublatetheminanoverarchingunity.ThisfeatureofHegel’sthoughtraisesadifficulty.Doreflectionandtheresolutionofoppositionscometoanendorproceedtoinfinity?Forexample,arehislogicandhisphilosophycompleteinthesensethatfurtherreflectiononthemwillgeneratenonewcategoriesnotalreadycontainedinthem,andthatlogicandphilosophyarethusatanend?Hashistorycometoanendinhisowntimeoristheremoretocome(perhaps,asheoccasionallysuggests,inRussiaorAmerica),evenifwecannotpredictwhatitwillbe?OnsuchquestionsastheseHegelwasineradicablyambiguous.Humanbeingsthinknotonlyaboutthemselvesandtheirthoughts,butaboutwhatis,initiallyatleast,otherthanthought,namely,‘nature’.Thoughtsareimplicitin746\nHEGELnature,butonlygraduallydohumansmakethemexplicitbytheirtheoreticalandprac-ticalactivities,domesticatingnatureandcomingtoseeitasaprojectionorextensionofthemselves.ThisiswhyHegelregardedhimselfasanidealist:hebelievednotonlythatmind(orminds)exist,but,firstly,thatthethoughtsthatconstitutethecoreofthehumanmindarealsoimplicitlyembeddedinthings,and,secondly,thatnatureispro-gressivelysublatedor‘idealized’bythespiritualactivitiesofhumanbeings.Wetherebyachievefreedom,sincefreedomconsistsinself-determination,innotbeingdeterminedbyan‘other’,andnatureisnowseenasbeingnolongerstarklyotherthanourselves.Freedomisalsoenhancedbyaperson’sself-understanding,whichadvancesalongwithhisorherunderstandingofnature.Theworkingsofhistoryandofthesocietiesweinhabit,aswellasthelogicalstructuresthatunderliethem,becomeincreasinglyperspicuoustous.Thusthegrowthofourself-consciousnessaccompaniesthatofourfreedom.Hegelheldthathisphilosophicalsystemexpressesintheformofthoughttheessen-tial‘content’ofTrinitarianChristianity.Inreligiousimagery,GodtheFather‘alienates’orexternalizeshimselfbyproducinganaturalworld(the‘son’)astheobjectofhis‘con-sciousness’,andbecomesself-consciousinthespiritualreclamationthatpeoplemakeofnatureandofthemselves(the‘holyspirit’).Inconceptualterms,the‘logicalidea’,thesystemofconceptsstudiedbylogic,‘freelyreleasesitself’intonaturewhichisthenspiritualizedbyhumanbeings.Thequestionwhetherthisamountstoatheismornotisanotherofthoseambiguities(alongwithhisattitudestothepoliticalstatusquoandtothefuturecourseofhistory)thatareintrinsictohisthought.Notsurprisinglyhisfollowersdividedinto‘right’,‘left’and‘centre’Hegelians,accordingtowhethertheywereconservatives,radicalsormoderatereformersinreligiousandpoliticalmatters.Hegel’simmenseinfluence–inGermany(onD.F.Strauss,FeuerbachandMarx),Britain(onBradley,BosanquetandMcTaggart),Italy(onGentileandCroce)andinFrance(onKojève,SartreandDerrida)–dependsinlargemeasureonthecomprehen-siverichnessofhisthoughtanditsaccompanyingambiguity.FurtherReadingForareliableandinterestingaccountofHegel’slifeandworks,Kaufmann(1965)isstillunsur-passed.ThebestbriefgeneralstudiesofHegel’sthoughtareMure(1965)andSoll(1969).ThephilosophicalbackgroundtohisthoughtislucidlypresentedbyCopleston(1965),andtheessaysinPriest(1987)exploreinmoredetailhisrelationshiptoKant.ThecomplexitiesofHegel’ster-minologyareunravelledinInwood(1992).LongergeneralstudiesincludeFindlay(1958),Taylor(1975)andInwood(1983).OnparticularaspectsofHegel’sthought,O’Brien(1975),Wood(1990),Pinkard(1994)andForster(1998)aretoberecommended.EssaysonvariousaspectsofhisthoughtanditsinfluencearetobefoundinInwood(1985),Beiser(1993)andStern(1993).ReferencesHegelHegel,G.W.F.1892–6:LecturesontheHistoryofPhilosophy,3vols(translatedbyE.S.HaldaneandF.H.Simson).London:KeganPaul.AlsotranslatedbyT.M.KnoxandA.V.Miller,IntroductiontotheLecturesontheHistoryofPhilosophy.Oxford:ClarendonPress,1985.747\nMICHAELINWOOD——1948[1793–1800]:EarlyTheologicalWritings(translatedbyT.M.Knox).Chicago:UniversityofChicagoPress.——1969[1812–16]:ScienceofLogic(revdedn1832)(translatedbyA.V.Miller).London:AllenandUnwin.——1970[1817]:PhilosophyofNature(Pt2oftheEncyclopaedia)(translatedbyA.V.Miller).Oxford:ClarendonPress.AlsotranslatedbyM.J.Petry,3vols.London:AllenandUnwin.——1971[1817]:PhilosophyofMind(Pt3oftheEncyclopaedia)(1830edntranslatedbyW.WallaceandA.V.Miller).Oxford:ClarendonPress.——1975a[1835]:Aesthetics:LecturesonFineArt(revdedn1842),2vols(translatedbyT.M.Knox).Oxford:ClarendonPress.AlsotranslatedbyB.Bosanquet,editedbyM.Inwood,asIntroductoryLecturesonAesthetics.Harmondsworth:PenguinBooks,1993.——1975b[1837,1840]:LecturesonthePhilosophyofWorldHistory:Introduction:ReasoninHistory(translatedbyH.B.Nisbet).Cambridge:CambridgeUniversityPress.AlsotranslatedbyL.Rauchfromalessrecentlyeditedtext,asIntroductiontothePhilosophyofHistory.Indianapolis,IN:Hackett,1988.——1977a[1801]:DifferenceBetweenFichte’sandSchelling’sSystemofPhilosophy(translatedbyH.S.HarrisandW.Cert).NewYork:StateUniversityofNewYorkPress.——1977b[1807]:PhenomenologyofSpirit(translatedbyA.V.Miller).Oxford:ClarendonPress.——1983[1805–6]:HegelandtheHumanSpirit:ATranslationoftheJenaLecturesonthePhiloso-phyofSpirit(1805–6)withCommentary(translatedbyL.Rauch).Detroit,MI:WayneStateUniversityPress.——1984–6[1832,1840]:LecturesonthePhilosophyofReligion(translatedbyP.Hodgson,R.F.BrownandJ.M.Stewart).Berkeley:UniversityofCaliforniaPress.AlsotranslatedbyE.B.SpeirsandJ.B.Sanderson,3vols,fromalessrecentlyeditedtext,London:KeganPaul,1895.——1986[1804–5]:TheJenaSystem1804–5:LogicandMetaphysics(translatedbyJ.S.BurbidgeandG.diGiovanni).Montreal:McGill-Queen’sUniversityPress.——1991a[1817]:TheEncyclopaediaLogic(Pt1oftheEncyclopaediaofthePhilosophicalSciences)(revdedns1827,1830,withposthumousedns1840–5)(translatedbyT.F.Geraets,H.S.HarrisandW.A.Suchting).Indianapolis,IN:Hackett.AlsotranslatedbyW.WallaceasHegel’sLogic.Oxford:ClarendonPress,1975.——1991b[1821]:ThePhilosophyofRight(translatedbyH.B.Nisbet,editedbyA.W.Wood).Cambridge:CambridgeUniversityPress.AlsotranslatedbyT.M.Knox,Oxford:ClarendonPress,1942.OtherwritersBeiser,F.(ed.)1993:TheCambridgeCompaniontoHegel.Cambridge:CambridgeUniversityPress.Copleston,F.1965:AHistoryofPhilosophy,vol.7,PartI:FichtetoHegel.NewYork:Doubleday.Findlay,J.N.1958:Hegel:ARe-examination.London:AllenandUnwin.Forster,M.1998:Hegel’sIdeaofaPhenomenologyofSpirit.Chicago:UniversityofChicagoPress.Inwood,M.1983:Hegel.London:RoutledgeandKeganPaul.——(ed.)1985:Hegel.Oxford:OxfordUniversityPress.——1992:AHegelDictionary.Oxford:Blackwell.Kaufmann,W.1965:Hegel:Reinterpretation,TextsandCommentary.London:WeidenfeldandNicolson.Mure,G.R.G.1965:ThePhilosophyofHegel.London:OxfordUniversityPress.O’Brien,G.D.1975:HegelonReasonandHistory.Chicago:UniversityofChicagoPress.Pinkard,T.1994:Hegel’sPhenomenolgy:TheSocialityofReason.Cambridge:CambridgeUniversityPress.Priest,S.(ed.)1987:Hegel’sCritiqueofKant.Oxford:ClarendonPress.Soll,I.1969:AnIntroductiontoHegel’sMetaphysics.Chicago:UniversityofChicagoPress.748\nHEGELStern,R.(ed.)1993:G.W.F.Hegel:CriticalAssessments.London:Routledge.Taylor,C.1975:Hegel.Cambridge:CambridgeUniversityPress.Wood,A.W.1990:Hegel’sEthicalThought.Cambridge:CambridgeUniversityPress.DiscussionQuestions1DoesHegelrefuteKant’sviewthatthings-in-themselvesareunknowable?2Inwhatsense,ifany,isHegelanidealist?3DoesthePhenomenologyofSpirithaveacoherentplan?4Canwehaveself-consciousnesswithoutconflict?5Whatisdialectic?6IsHegel’s‘logic’reallylogic?7DoesHegelhaveanygoodreasonforbelievingthatphilosophymustbesystematic?8CanHegelgiveacoherentaccountofthetransitionfromlogictonature?9WhatusedoesHegelmakeofthenotionofinfinity?10Whatisspirit?11Doweneedobjectivefreedomaswellassubjectivefreedom?12Ishistoryrational?13IsHegelcommittedtotheviewthathistoryhascometoanend?14WasHegelanatheist?15Isitplausibletosupposethatart,religionandphilosophyhavethesame‘content’?749\n34MarxRICHARDNORMANThewritingsofKarlMarxspanmanydisciplinesbesidesphilosophy,suchashistory,economics,andpoliticalandsocialtheory.Thischapterreviewsrecentworkonthephilo-sophicalaspectsofhistheoryofhistory,andonhisviewofethics,inordertosuggestthattheconnectionsbetweenthosetwothemesareessentialtoanunderstandingofMarx’sphilosophicalsignificance.1IntroductionForaperiodofabouttwentyyearsafterthe1939–45war,Marx’sideasweregivenlittleseriousacademicattention.Therewereexceptions,suchasPopper’sinfluentialbookTheOpenSocietyanditsEnemies(1945),butonthewholeMarxwasignoredbyphilosophyashewasbyotherintellectualdisciplines.Sincethen,therehasbeenaveritableexplosionofseriousinterestinhim,beginninginthelate1960swiththeperiodofpoliticalradicalismintheuniversitiesofWesternEuropeandNorthAmerica.Ahugeamounthasbeenwrittenonallaspectsofhisthought–farmorethanIcandojusticetointhischapter.IshallconcentrateonjusttwothemeswhichhavebeenprominentinrecentphilosophicaldiscussionsofMarx:histheoryofHISTORY(chapter14)andhisviewofETHICS(chapter6).TherevivalofseriousinterestinMarx’sphilosophyinitiallytookaformwhichthrewnewlightonhisethics.Thiswasthediscoveryofhisearlywritings,unpublishedinhislifetimeandformanyyearsthereafter.InparticularhisEconomicandPhilosophicalManuscripts,writtenin1844,firstpublishedin1932andfirsttranslatedintoEnglishin1959,revealawholenewdimensionofhisthought.Akeyconcepthereisthatof‘alienation’,atermwhichMarxtookoverfromHEGEL(chapter33)andusedtodescribetherelationofworkerstotheirproductandtotheirproductiveactivitywithinacapi-talistsystemofproduction.Theworkers’ownlabour,andtheproductsofthatlabour,cometodominateandenslavethem.Becausetheirworkisimposedonthembyothersanditsproductsareappropriatedbyothers,worklacksallmeaningfortheworkersthemselves.Insteadofbeingasourceoffulfilmentthroughtheemploymentofmentalandphysicalenergy,it‘mortifieshisbodyandruinshismind’.Ittherebyalienatestheworkerfromhisorherveryhumanity.\nMARXMarx’stheoryofalienationconfirmstheviewthat,atleastinhisearlywritings,Marxcondemnscapitalismnotjustforitshistoricalobsolescencebutbecauseofitseffectsonthelivesofthosewholiveandworkwithinsuchasystem.Theconceptofalienationalsoservesasamoraltoolwithwhichtocriticizemorethanjustthework-ingsofacapitalisteconomy.Economicalienationisonlyoneaspectofalienation.Otherformsofitarereligiousalienation,whichinvolvesahumanbeing’sself-abasementbeforebeingsthataretheproductofhisorherownimagination,andpoliticalalien-ation,whichinvolvesthedominationofhumanbeingsbythestatewhichiscreatedandsustainedthroughtheirownactivity.OneobviousreasonwhyMarx’searlywrit-ingsreceivedsomuchattentionwhentheywerediscoveredisthattheyshowedveryclearlythemoraldistancebetweenMarx’sownideasandthecharacterofself-proclaimed‘Marxist’regimessuchasthatintheSovietUnion.Thephilosophicalinterestoftheearlywritingswasalsothattheyrevealedadis-tinctiveandpositiveethicaltheorylyingbehindtheconceptofalienation.ItisatheoryinthesametraditionasARISTOTLE’s(chapter23)accountofeudaimoniaandMILL’s(chapter35)accountofthe‘higherpleasures’:theviewthatthegoodlifeforhumanbeingsconsistsinthefullactualizationoftheirdistinctivelyhumancapacities.ForMarx,however,thesecapacitiesarenotjustintellectualandtheoretical;theyare,aboveall,practical–thecapacityforfreecreativeworkwhichtransformstheexternalworldandgivesitahumansignificance.Ahumanbeing‘duplicateshimselfnotonlyintel-lectually,inhismind,butalsoactivelyinrealityandthuscanlookathisimageinaworldhehascreated’(Marx1977:82).Alienatedlabourthwartsthesecapacitiesforself-objectificationandself-realization.Itde-humanizeshumanbeingsandtheworldtheyhavecreated.ThediscoveryoftheseearlywritingsgeneratedcontroversyabouttheirrelationtoMarx’slaterwork.Somecommentatorssawanessentialcontinuity.TheycouldpointtothepassagesinCapital(1867)andothereconomicwritingswhereMarxdescribeshow,‘withinthecapitalistsystem...allmeansforthedevelopmentofproductiontransformthemselvesintomeansofdominationover,andexploitationof,thepro-ducers’,and‘mutilatethelabourerintoafragmentofaman’(Marx1977:482).Othercommentators,however,regardedthetheoryofalienationandMarx’slatertheoryofhistoryasincompatible.TheforemostadvocateofthisviewwastheFrenchphiloso-pherLouisAlthusser,whose‘structuralistMarxism’positedadecisive‘epistemologicalbreak’betweenMarx’searly‘humanism’andthelater‘scientific’writings.Althusser-ianMarxismwasforatimetheheightofintellectualfashion.Idonotintendtodiscussithere,anddonotfindmuchofvalueinit,butitdidserveasareminderthatthecoreofMarx’sworkistobefoundinhismaturesocialandhistoricaltheory,andthatanyphilosophicalengagementwithMarxmustcometotermswiththatwork.2MarxandAnalyticalMarxismAlandmarkinthestudyofMarx’smatureworkwasthepublicationin1978ofCohen’sbookKarlMarx’sTheoryofHistory:ADefence.CohendrewexplicitlyonthetechniquesandstyleofanalyticalphilosophyintheattempttointroduceanewclarityandrigourintotheexpositionanddefenceofMarx’stheory.Thethreemainachievementsof751\nRICHARDNORMANCohen’sbookare:firstly,theprovisionofaclearandpreciseconceptualapparatusforMarx’stheory;secondly,ageneralexpositionanddefenceofthethesisof‘theprimacyofproductiveforces’;andthirdly,theclaimthatthekindofexplanationemployedinMarxiantheoryis‘functionalexplanation’.CohentakesasadefinitivetextMarx’s1859prefacetohisCritiqueofPoliticalEconomy.Thetheorysketchedthereidentifiesthreelevelsofsociallife:(1)the‘pro-ductiveforces’(toolsandmaterials,skillsandtechniques),towhichtherecorrespond(2)‘relationsofproduction’whichconstitutethe‘economicstructure’andwhichgiveriseto(3)‘alegalandpoliticalsuperstructure’and‘formsofsocialconsciousness’.EarliercriticsofMarx,suchasActonandPlamenatz,hadarguedthatthetheorywasincoherentbecausethesethreelevelswerenotinfactdistinct.CohenthereforeprovidescleardefinitionsoftheseconceptsinordertoshowthatthethreelevelscaninfactbedistinguishedandthatitispossibleforMarxtomakepreciseanddefiniteclaimsabouttheexplanatoryrelationsbetweenthem.Theessentialclaimsarethatthecharacterofnon-economicinstitutionsislargelyexplainedbythenatureoftheeconomicstructure,andthatthenatureoftheeconomicstructure(thatis,thesetofrelationsofproduc-tion)islargelyexplainedbythelevelofdevelopmentoftheproductiveforces.Cohenreferstothesecondclaimas‘theprimacyoftheproductiveforces’,and,afterprovid-ingtextualevidencethatMarxmadethisclaim,heproceedstoofferhisowngeneraldefenceofit.ThefirststageisthedefenceofapreliminaryclaimwhichCohencallstheDevelop-mentThesis:thattheproductiveforcestendtodevelopthroughouthistory.Hethensug-geststhatthestrongerPrimacyThesiscanbedefendedonthebasisoftheDevelopmentThesistogetherwithafurtherfact:thatagivenlevelofproductivepoweriscompatibleonlywithcertaintypesofeconomicstructure.Forexample,asocietybasedoncom-putertechnologycouldnotbeaslavesociety.Therefore,thoughproductiveforcescandevelopwithinagivensetofproductionrelations,thatdevelopmentisliabletoreachapointatwhichitisnolongercompatiblewiththoseproductionrelations.Whenthishappens,itistobeexpectedthattheincompatibilitywillberesolvedbythealterationoftheproductionrelations,sinceotherwisefurtherdevelopmentoftheproductiveforceswouldbeimpossible,andthisisdeniedbytheDevelopmentThesis.Cohendoesnotofferanysimilargeneralargumentfortheclaimthattheeconomicstructureinturnlargelyexplainsthecharacterofnon-economicinstitutions,andIshallcomebacklatertothisomission,butIturnnowtohisthirdimportantcontribu-tion:hisaccountofthekindofexplanationemployedbyMarx’stheory.MarxandFunctionalExplanationMarx’sexplanations,Cohenargues,arefunctionalexplanations.Theseareexplanationsoftheform‘FoccursbecauseitbringsaboutE’.Suchexplanationsappeartobeemployedinotherfields,forexampleinbiology,instatementssuchas‘Birdshavehollowbonesbecausehollowbonesfacilitateflight’.Suchstatementshaveacomplexstructure.TheinitialexplanatoryclaimisthatFbringsaboutE,butthewiderexplanatoryclaimisthenthatFitselfoccursbecauseitbringsaboutE.AttributingtoMarxtheuseoffunctionalexplanationshastheattractionofremovingapuzzleinMarxiantheory.Marxisclearlycommittedtotheclaimsthatproductionrelationsfacilitatethegrowthofproductive752\nMARXforces,andsimilarlythatsuperstructuralinstitutionshelptosustaintheeconomicstruc-ture.Thisseemstosuggestthattherelationsexplaintheforces,andthatsuperstructuresexplaineconomicstructures,whichseemstorundirectlycountertothecentralclaimsofthetheory.Thoseclaimsare,however,thattherelationsofproductionobtainbecausetheyfacilitatethegrowthofproductiveforces,andthatthesuperstructuralinstitutionshavetheparticularcharacterthattheyhavebecausetheytherebyhelptosustaintheeco-nomicstructure.Theuseoffunctionalexplanationsis,however,contentious,andtheirvalueisstronglycontestedbyElster,anotherphilosopherassociatedwithANALYTICALMARXISM(pp.273–4).Themerefactthatsomethinghasbeneficialconsequencesisnot,accordingtoElster(1985),sufficienttoexplainit.ConsiderasanexamplethecrudeexplanationsofferedbysomeMarxistsfortheexistenceofracism:itdividestheworkingclass,andtherebyservestheinterestsoftherulingclass.Elsterwouldsayofthisexamplethat,thoughitmaybetruethatracismservestheinterestsoftherulingclass,thisdoesnotinitselfprovideanygroundsforassertingthatthatiswhyracismexists.Noexplanationhasbeengivenunlesswecanciteamechanism,anaccountofhowracismisexplainedbyitstendencytopromotetheinterestsoftherulingclass.Suchamechanismmightbeidenti-fiedby,forinstance,a‘conspiracytheory’whichclaimedthatracismwasdeliberatelyfos-teredbymembersoftherulingclassinordertopromotetheirinterests.Butevidenceforthiswouldhavetobecited,andintheabsenceofsuchevidencenoexplanationhasbeengiven.ElsterdoesnotthinkthatthemisuseoffunctionalexplanationisconfinedtocrudeversionsofMarxism.Marx’sownwork,hesuggests,isflawedbytherelianceonunsup-portedfunctionalexplanations.Marxoffersnoaccountofhowtheneedtopromotethedevelopmentofproductiveforcesexplainstheemergenceofnewproductionrelations.Themechanismissupposedtobefoundintheoperationofclassconflict,butMarxwouldhavetoexplainhowtheexistingconstraintsonthegrowthofproductiveforcesmotivatethemembersofaclasstochangetherelationsofproduction,andthishefailstodo.Moreseriouslystill,Marx’stheoryofhistoryreliesnotjustonunsupportedfunctionalexpla-nationsthatinvokebeneficialconsequencesforgroupsofhumanbeings,butalsoonvaguespeculativeexplanationsofsocialchangeandsocialinstitutionsthatpromotethegoalsofsupposedimpersonalagentssuchas‘History’or‘Capital’.CohenwouldagreewithElsteronthedesirabilityofspecifyingamechanisminordertobackupafunctionalexplanation.TherealdisagreementbetweenthemisthatElsterthinksthatfunctionalexplanationsareworthlesswithoutsuchbacking,whereasCohenthinksthatevenifwecannotspecifyamechanism,functionalexplanationsmaystillexplain,thoughtheywouldexplainmorefullyiftheycouldbeelaborated.WhatCohendoesrecognizeisthat,thoughfunctionalexplanationsmaylackanelaboration,theydo,likeordinarycausalexplanations,havetobebackedbyageneralization.ThemerefactthatFhasafunctionaleffectEinaparticularcasedoesnotconstituteanexplanationofF;butitbecomesarelevantexplanationifitcanbebackedbyageneralizationoftheform‘WheneverFwouldbringaboutafunctionaleffectE,Foccurs’.3ExplainingSuperstructureandIdeologyIthinkitisfairtosaythatCohendoesnotspecifyinanydetailthekindsofgeneral-izationthatwouldbackthefunctionalexplanationsinMarx’stheoryofhistory,and753\nRICHARDNORMANthatCohen’sowndefenceofthattheorydoesnotappealtosuchgeneralizations.Thatdefence,asoutlinedabove,concentratesprimarilyontheroleofproductiveforcesinexplainingtherelationsofproduction,butIwantnowtolookbrieflyatthesecondlevelofMarxianexplanation,theroleoftheeconomicstructureinexplaininglegalandpoliticalinstitutions(the‘superstructure’)andformsofconsciousness(‘ideology’).Itisherethatthequestionofwhatkindofgeneralizationwearelookingforintersectswithanotherimportantproblem:thequestionofthescopeofMarx’s‘economicdeterminism’.Marxsometimesappearstoclaimthatalllegalandpoliticalinstitutionsandallformsofconsciousnesscanbeexplainedbyreferencetotheeconomicstructure.Thisclaimiswildlyimplausible.Ishallnotarguethepointhere;Ishallsimplyassumethattheclaimhastobequalifiedinsomeway.Theproblemisthenthatthetheoryisindangerofbecomingvagueorvacuous.FromEngelsonwards,Marxistshavetakenrefugeinsuchphrasesasthattheeconomicstructureisdeterminant‘ultimately’or‘inthelastinstance’,anditisnotoriouslydifficulttopindownwhatsuchclaimsmeanorhowtheycouldbeassessed.Cohenhastackledtheprobleminhismorerecentwriting,andhasdistinguishedbetween‘inclusive’and‘restricted’construalsofhistoricalmaterialism.Whatevertheappropriategeneralformulation,Itakeitthattheexplanatorypowerofhistoricalmaterialismmustbelookedforinitscapacitytoexplainspecificsuperstruc-turalandideologicalphenomena:thisparticularinstitution,thisparticulartheoryorbelief,areastheyarebecausetheyhelptosustainaparticularstructureofrelationsofproduction.Thisnowreturnsustothequestionofthekindofgeneralizationthatcouldbacksuchaclaim.Theclaimwouldbe,forinstance,thatliberal-democraticpoliticalinsti-tutions,orliberal-individualisttheoriesofpoliticalfreedom,prevailbecausetheyhelptolegitimateandsupportmarketeconomies,andthiswouldbebackedbythegeneral-izationthatsuchinstitutions,orsuchtheories,emergeandbecomedominantpreciselywhentheycanhelptosustainmarketeconomies.Thereisacertainplausibilityinthis.Inthecaseofmytwoexamples,itmaywellbepossibletoshowthatliberal-democraticinstitutionsandliberaltheoriesoffreedomareparticularlywellsuitedtomarketeconomies,anditisthenastrikinggeneralfactthatthoseinstitutionsandthe-oriescometotheforeinjustthosehistoricalcircumstanceswheretheycanservethatfunction.This,however,doesnotsettlethequestionofexplanatoryprimacy.Analter-native‘idealist’accountwouldbethattheseinstitutionsorthesetheorieshavebeenadoptedbecausehumanbeingshavegoodreasonsforadoptingthem,becausetheyareinsomeappropriatesense‘good’institutionsortheories;iftheyalsohappentosupportcertainkindsofeconomicstructures,itisnotsurprisingthattheyhavedoneso,butthatisnotwhytheyhavebeenadopted.Tosupportthefunctionalexplanation,then,somethingmorethanthegeneralizationisneeded.Wehavetobeabletoclaimthattheinstitutionorthetheorywouldnothavebeenadoptedunlessitservedtosustaintheeconomicstructure.Theproblemofhowwecansupportcounterfactualhistoricalstatementsofthisformisanotoriouslydifficultone,butIwanttosuggestthatinsuchacontextaplausiblecandidatetosupportitwouldbesomekindofnormativejudge-ment:theinstitutionortheorycanbeshowntobeflawedordefectiveinsomeway,andthatiswhyweareinclinedtosaythatitwouldnothavebeenadoptedunlessitservedthisfunctioninrelationtotheeconomicstructure.754\nMARX4MarxistEthicsItisinthelightofthissuggestionthatIwanttoreturnfinallytothequestionofMarx’sviewofethics,atopicthathasbeenthesubjectofvoluminousdebateinthepasttwentyyears.Marxregularlyspokeofmoralityasatypicalexampleofideology,implyingthatallmoralbeliefsarebytheirverynatureaproductofclassinterestsandclassstandpoints,thattheyarerelativetoparticularsocialandhistoricalconditionsandthattheycanhavenoindependentrationalvalidity.However,Marxhimselfalsoseemedtobecommittedtohisownmoraljudgementsabouttheoppressiveandexploitativecharacterofclasssoci-eties,andinparticularaboutthedegradinganddehumanizingconditionsofworkinthecapitalistsocietiesofhisownday.AsLukesputsit,wethenseemdriventotheconclusionthat‘Marxism’sattitudetomoralityisparadoxical’(Lukes1985:1).Themostpromisingwayofattemptingtoresolvetheparadoxistodistinguishbetweendifferentformsofmorality.LukesproposesadistinctionbetweenthemoralityofRecht,thatisofJUSTICEANDRIGHTS(chapter8)andthemoralityofemancipation,andsuggeststhat‘theparadoxinMarxism’sattitudetomoralityisresolvedonceweseethatitisthemoralityofRechtthatitcondemnsasideologicalandanachronistic,andthemoralityofemancipationthatitadoptsasitsown’(ibid.:29).Thisdistinction,however,hastobequalifiedinvariousways.Ontheoneside,itisnotclearthatMarxwouldregardhisownuseofthelanguageof‘emancipation’asanappealtomorality.Notallcondemnationofaspectsofhumanlifecountsas‘moral’condemnation,andMarxmayhavebeenquiteconsistentinmaintainingthatmoralformsofcondemna-tionareintrinsicallyideologicalwhileotherformsofcondemnationarenot.Whatisobviouslyneededhereissomemorepreciseaccountofthedifferencebetween‘themoralpointofview’andotherformsofevaluation.TheotherqualificationneededtoLukes’sdistinctionisthat,asherecognizes,Marx’sattitudetothemoralityofRechtisnotsimple.Wood(1981)andothershavearguedthatMarxdidnotregardcapitalismasunjust.ForMarx,theconceptofjusticeisessentiallyajuridicalconcept.Assuch,itisacomponentinasetofrulesbywhichaparticularmodeofproductionisregulated.EachhistoricalmodeofproductionhasitsownappropriatesetofJURIDICALINSTITUTIONS(chapter13)andconcepts.Therefore,fromthestandpointofthejuridicalframeworkappropriatetothecapitalistmodeofproduction,particularactsmaybeunjustbutcapitalismitselfcannotbe.Othercom-mentators,however,havepointedtopassagesinwhichMarxdidappeartocriticizeinjusticefromastandpointwhichtranscendsthatofcapitalism.Lukessuggeststhat‘Marx’sviewofcapitalism’sjusticewasbothinternallycomplexandhierarchicallyorganized’(Lukes1985:58).Marxdididentifyaconceptionofjusticewhichisinter-naltothecapitalistmodeofproduction,buthealsocriticizeditfromanexternalstand-point.Thisisthestandpointof‘thefirstphaseofcommunistsociety’,asMarxcalleditinhisCritiqueoftheGothaProgramme(1875).Itisthestandpointdefinedbytheprin-ciplethateachindividualshouldberewardedinproportiontothelabourheorshehascontributed.Thatstandpoint,however,willitselfbetranscendedinthe‘higherphaseofcommunistsociety’,embodyingtheprinciple,‘Toeachaccordingtohisneeds’.Thelatterisnotitselfaprincipleofjustice;itistheexpressionofasocietyofabundanceinwhichjuridicalprinciplesregulatingthesocialallocationofgoodsarenolonger755\nRICHARDNORMANnecessary.LukesgoesontocriticizeMarx’sassumptionthatprinciplesofjusticecaneventuallybedispensedwith.Thisassumption,hethinks,hasdivertedtheMarxisttra-ditionfromthenecessarysearchforanadequateaccountofjusticeandrights,andhasleftthewayopenfortheappallinginjusticesandviolationsofhumanrightscommit-tedby‘Marxist’politicalmovementsandregimes.Marx’sethicaltheory,then,iscomplexandcontentious.Idoubtwhetheritispossi-bletoarriveatanydefinitiveconclusionaboutit.Hisremarksonthesubjectaretoobrief,scatteredandcontradictoryfortheretobesuchathingas‘whatMarxreallythought’aboutethics.MyownconcludingsuggestionissimplyaviewaboutwhatisphilosophicallyfruitfulinMarx’sapproachtoethics,anditisaviewwhichlinksitwithmycommentsabouthistheoryofhistory.Thattheory,Ihaveindicated,isneitherplau-siblenorhelpfulwhenformulatedasageneraltheorywhichpurportstoexplainallpoliticalandlegalinstitutionsandallformsofconsciousnessasproductsofaneco-nomicstructure.Ialsosuggestedthat,wheninterpretedasamoreselectivetheorywhichexplainsspecificinstitutionsandformsofconsciousness,itdependsforitsplausibilityonbeingalsoacriticaltheory.Particularinstitutions,beliefsortheoriesarecriticizedasbeingflawedordefectiveinwaysthatcanbeexplainedwhenitisseenhowtheyservetosustainaparticulareconomicstructureandthepowerofaparticularclass.ThesamejudgementnowhastobemadeofMarx’sethicaltheory.WhateverMarxhimselfmayhavesaid,histheoryisself-defeatingifitisunderstoodastheclaimthatallvaluesaresociallyandhistoricallyrelativeandaretheexpressionofspecificclassinterestsandpositions.Itismostfruitfulwhenunderstoodasacriticalaccountofhowparticularvaluesandethicalbeliefscometobedistortedbyreflectingalimitedclassstandpointandclassinterests.AndthereisofcourseaparticularreasonwhyMarx’sethicaltheoryneedstobeinterpretedasbeingselective,notasall-inclusive;forifwhatMarxoffersisacriticaltheory,itmustitselfappealtovalues,andcannotthere-foretreatallvaluesastheideologicalby-productsofeconomicstructures.Marx’stheoryneedsitsownvalues.FurtherReadingNooneworkbyMarxencapsulateshiscontributionstophilosophy,andtheuseofananthologyisessential.AusefulcollectionistheoneeditedbyMcLellan(Marx1977).AgoodbriefgeneralintroductiontoMarx’sthoughtisthatbySinger(1980).Whencompleted,thebestscholarlyresourceinEnglishforadvancedworkwillbeMarxandEngels’sCollectedWorks(1975–).Popper’sinfluentialcriticismsofMarx,andinparticularofMarx’shistoricism,canbefoundinvolume2ofTheOpenSocietyanditsEnemies(1945).OtherimportantcriticaldiscussionsofMarx’stheoryofhistoryarethosebyActon(1955)andPlamenatz(1954,1963).Cohen(1978)isstronglyrecommended.SomeofCohen’smorerecentwork,bothamplifyingandmodifyinghisearlierdiscussion,iscollectedinhisHistory,Labour,andFreedom(1988).Elster’sinterpreta-tionandassessmentofMarxissetoutatlengthinhisMakingSenseofMarx(1985),andmuchmorebrieflyandreadablyinhisAnIntroductiontoKarlMarx(1986).ThereisgoodmaterialonthetheoryofhistoryinMcMurtry(1978),Shaw(1978)andWood(1981);WoodisalsothemostconvenientgeneralguidetoMarxfromaphilosophicalpointofview.ThebestintroductiontothedebateonMarxandethicsisLukes(1985).EssaysonthesubjectarecollectedinCohen,NagelandScanlon(1980),inNielsenandPatten(1981)andinWareandNielsen(1989).Nielsen(1988)isanotherusefulguide.Wood’spositionissetoutinanarticleinthecollection756\nMARXbyCohen,NagelandScanlon(1980)andinWood(1981).OnMarx’sconceptofalienation,seePlamenatz(1975),Ollman(1971)andMeszaros(1972).ReferencesMarxDatesgiveninsquarebracketsaredatesoffinalcomposition,exceptforCapital,vol.1,whichwascompletedin1865–6andpublishedin1867.Marx,K.[1844]:EconomicandPhilosophicalManuscripts.InMarx(1977).——[1859]:PrefacetoACritiqueofPoliticalEconomy.InMarx(1977).——[1864]:Capital,vol.3(translatedbyE.Untermann).London:LawrenceandWishart.——[1867]:Capital,vol.1(translatedbyS.MooreandE.Aveling).London:LawrenceandWishart.——[1875]:CritiqueoftheGothaProgramme.InMarx(1977).——1971[1857–8]:Grundrisse(translatedbyM.Nicolaus).Harmondsworth:PenguinBooks.——1977:SelectedWritings(editedbyD.McLellan).Oxford:OxfordUniversityPress.Marx,K.andEngels,F.(1975–):CollectedWorks,58vols.London:LawrenceandWishart.OtherwritersActon,H.B.1955:TheIllusionoftheEpoch.London:CohenandWest.Reprintedin1972byRoutledgeandKeganPaul.Cohen,G.A.1978:KarlMarx’sTheoryofHistory:ADefence.Oxford:OxfordUniversityPress.——1988:History,Labour,andFreedom.Oxford:OxfordUniversityPress.Cohen,M.,Nagel,T.andScanlon,T.(eds)1980:Marx,Justice,andHistory.Princeton,NJ:PrincetonUniversityPress.Elster,J.1985:MakingSenseofMarx.Cambridge:CambridgeUniversityPress.——1986:AnIntroductiontoKarlMarx.Cambridge:CambridgeUniversityPress.Lukes,S.1985:MarxismandMorality.Oxford:OxfordUniversityPress.McMurtry,J.1978:TheStructureofMarx’sWorld-View.Princeton,NJ:PrincetonUniversityPress.Meszaros,I.1972:Marx’sTheoryofAlienation.London:MerlinPress.Nielsen,K.1988:MarxismandtheMoralPointofView.Boulder,CO:WestviewPress.Nielsen,K.andPatten,S.C.(eds)1981:MarxandMorality.CanadianJournalofPhilosophy,supplementaryvolume7.Guelph:CanadianAssociationforPublishinginPhilosophy.Ollman,B.1971:Alienation.NewYork:CambridgeUniversityPress.Plamenatz,J.1954:GermanMarxismandRussianCommunism.London:Longman.——1963:ManandSociety,vol.2.London:Longman.——1975:KarlMarx’sPhilosophyofMan.Oxford:OxfordUniversityPress.Popper,K.R.1945:TheOpenSocietyanditsEnemies,vol.2.London:RoutledgeandKeganPaul.Shaw,W.1978:Marx’sTheoryofHistory.Stanford,CA:StanfordUniversityPress.Singer,P.1980:Marx.Oxford:OxfordUniversityPress.Ware,R.andNielsen,K.(eds)1989:AnalyzingMarxism.CanadianJournalofPhilosophy,supplementaryvolume15.Calgary:UniversityofCalgaryPress.Wood,A.1981:KarlMarx.London:RoutledgeandKeganPaul.DiscussionQuestions1Whatisalienation?DoesMarx’saccountsucceedinexplainingwhypeople,insomesocietiesatleast,arealienated?757\nRICHARDNORMAN2Areintellectualcapacitiesorpracticalcapacitiesmoreimportantforrealizingthegoodlifeforhumanbeings?3CanthestyleandtechniquesofanalyticalphilosophyhelpustounderstandandassessMarx’shistoricalmaterialism?4Shouldweaccepttheprimacyofproductiveforcesinexplainingsocialphenomena?5Iftherearelevelsofsociallife,howcanwedeterminewhichaspectsofsocietybelongtoeachlevel?6Howcanweassesstheclaimthattheeconomicstructureisdeterminant‘inthelastinstance’?7Ifhistoricalmaterialismcannotexplainallthemainfeaturesofsociety,canitexplainanyofthem?8Canfeaturesofthelegalandpoliticalsuperstructureandideologyofasocietybeexplainedbythenatureofitseconomicstructure?Whatotherhypothesesareworthconsidering?9Whatarefunctionalexplanations?CanweinterpretMarx’sexplanationsasfunctionalexplanations?10Arefunctionalexplanationsjustifiedonlyforinstitutionsortheoriesthatareflawedordefectiveinsomeway?11WasMarx’sattitudetowardmoralityparadoxical?Ifso,cantheparadoxbeovercome?12Canweeventuallydispensewithprinciplesofjusticeandrights?13IfMarx’stheoryneedsitsownvalues,canittreatallvaluesastheideologicalby-productsofeconomicstructures?758\n35Bentham,MillandSidgwickROSSHARRISONJeremyBentham(1748–1832),JohnStuartMill(1806–73)andHenrySidgwick(1838–1900)aretheleadingrepresentativethinkersoftheBritishutilitariantradition.Benthamdeclaredatthestartofhisfirstwork,AFragmentonGovernment(publishedin1776),thatthe‘fundamentalaxiom’isthat‘itisthegreatesthappinessofthegreatestnumberthatisthemeasureofrightandwrong’.Happiness(or,asBenthamotherwisecalledit,utility)ishencetakenastheuniquemeasureofvalue.Benthamwaschieflyconcernedwiththeapplicationsofthisfundamentalprinciple.Hewasapracticalthinker,centrallyinterestedinmakingimprovementsintherealworld.WithJ.S.Millwegetamorenuancedstatementofthecentralprincipleaidedbyamoresubtlesurroundingpsychology.HischiefworkinthisareawashisUtilitarian-ism(whichappearedasasequenceofarticlesin1861andasabookin1863).Finally,SidgwickinhisMethodsofEthics(1874)criticizedMillbutalsoaddedamorecompleteepistemologicalbaseforutilitarianism.RecentworkonthesethreehasbeenmuchmoreonBenthamandMillthanonSidgwick,andthisbalancewillbefollowedhere.Inparticular,inthelastthirtyyearstherehasbeenasustainedrevisionaryinterpretationofMill,whichhasremovedhimfromthecriticismsofpreviousstandardcommentators.Sincethecentralfocusofattentioninthischapteriswiththisrecentwork,theorderofexpositionwillnotbechronological.Instead,itmovesbetweentheoriginaltextsandrecentcommentators,andalsobackandforwardbetweenBenthamandMill.1TheCentralIdeaItissimplesttotakethecentralideaofutilitarianismfromBentham.Inthefirstchapterofhismajorwork,AnIntroductiontothePrinciplesofMoralsandLegislation(1968–),Benthamsaysthat‘byutilityismeantthatpropertyinanyobject,wherebyittendstoproducebenefit,advantage,pleasure,good,orhappiness(allthisinthepresentcasecomestothesamething)or(whatcomesagaintothesamething)topreventthehappeningofmischief,pain,evil,orunhappiness’.ThekeypointisthatthereisasinglesourceofVALUE(p.203),whichcanbeindifferentlycalledutility,happinessorpleasure.Thisistobemaximized.\nROSSHARRISONUtilitarianism,Bentham’sethic,isaconsequentialistethic.Thingsaretobemea-suredandevaluatedbytheiractualandpossibleconsequences.Rightactionisjustifiedbyfuturestatesofaffairsratherthanbypastevents.Forexample,Bentham’saccountofthejustificationofpunishmentisadeterrentaccount.PUNISHMENT(p.645)isjusti-fiedifitdeterspeoplefromcommittingundesirableactions.Itisjustified,thatis,bythepreventionoffutureharmsratherthanbyretributionforpastwickedacts.Morepre-cisely,forBentham,theproperaimofpunishment,asofanythingelse,istoproducepleasureandtopreventpain.Yetallpunishmentisinitselfunpleasant,apain.Itcanthereforeonlybejustifiedifthisparticularpainisthecauseofagreaterreductioninpainelsewhere.Soifpeoplearedeterredbyapunishmentfromdoingthingswhichwouldotherwiseproducemorepain(suchas,forexample,rape,theftormurder),thenthatpunishmentwillbejustified.Iftherearetwoalternativepunishments,eachofwhichwouldproducelesspainthantheactionstheywoulddeter,thentheleastpainfuloptionistheappropriateone.Reductionofpain,andsuccessindeterringpain-causingbehaviour,areallthatcount.Nootherquestionsabouttheappropriatenessofaparticularpenaltyarerelevant.Similarly,Bentham’saccountofthejustificationofobediencetothestatelookstoactualandpossiblefutureconsequencesratherthantopastactionsorevents.Anexampleofacontrastingjustificationwhichlookstothepastisthatoftheoriginalcontract,whichsupposesthatobedienceisjustifiedbyan(actualortacit)agreementtoobeythestate.SuchadevicewasusedbytheleadingcontemporarydefenderofBritishlawinBentham’sday,WilliamBlackstone(1723–80).BenthamattacksBlackstonevehementlyinhisFragmentonGovernment(1968–)andridiculesthiskindofjustification.ForBentham,justificationofobediencetogovernmentdependsuponthecalculationofconsequences.AsheputsitintheFragment,itdependsuponcalculationofwhetherthe‘probablemischiefsofobediencearelessthantheprobablemischiefsofresistance’.Again,thejustificationisnotbasedonapastact(apastpromise),butontheavoidanceofpossiblefutureharms.Theseexamplesshouldsufficetogivethecentralideaofutilitarianism,atleastasitwasexpoundedbyBentham.Itcan,ofcourse,becriticizedonmanyaccounts;andtherehasbeencontinuouscriticismsinceBentham’sownday.BothJ.S.MillandSidgwick,althoughtheycontinuedtheutilitariantradition,werealsocriticsofBentham.However,beforeMill’scriticismandextensionofBenthamismisexamined,moreshouldbesaidabouttheusewhichBenthamhimselfmakesofutilitarianism.2Bentham’sUseofUtilityItisnotaccidentalthatthetwoexamplesgivenofutilitarianismintheprevioussectionconcernedtherelationsbetweenanindividualandthestate.Benthamwasuninterestedinprivateethics.Insteadhewasinterested,asheputsitintheIntroduction,in‘thatsystem,theobjectofwhichistorearthefabricoffelicitybythehandsofreasonandoflaw’.Theendisfelicity(happiness);theutilitarianend.Themeanstobeused,however,arereasonandLAW(chapter13):reason,becausethemethoddeployedwillnotbeappealtohistory,prejudiceorcustom;andlaw,becausetheprojectisaPOLITICAL(chapter8)project.Itistheprojectofdesigningaperfectsystemoflawandgovernment.760\nBENTHAM,MILLANDSIDGWICKSuchaprojectneedstwokindsoffundamentalassumptions:assumptionsaboutwhatoughttobethecaseandassumptionsaboutwhatisthecase.Theformersettheevaluativeends,thelatterdescribetheactualmaterialswhichhavetobeworkedwithinordertoachievetheseends.TheformeristheVALUETHEORY(p.212),andforBenthamthisisclearlyutilitarianism.Thelatterisapsychology,explaininghowpeopleactuallyact.Unlessthisisknown,thenitcannotbeknownwhatshouldbedonetopeopletoproducethedesiredend.ThepsychologywhichBenthamdeploysisthatpeopleactintheirownINTERESTS(p.205).ForBentham,thisisalsounderstoodintermsofpleasureandpain.Peoplearetakentobeseekersafterpleasureandavoidersofpain.ItisofutmostimportanceinunderstandingBenthamthatthesetwoprinciplesarenotconfused.Thestatementaboutwhatoughttohappen(thatthegeneralhappinessoughttobemaximized)isquiteseparatefromthestatementaboutwhatpeopleactu-allydo(attemptingtomaximizetheirownhappiness).TheyareclearlydistinguishedbyBenthamand,indeed,nosensecanbemadeoftheoverallprojectunlesstheyaredistinguished.Forthisdependsupontakingpeopleastheyare(seekersaftertheirownpleasure)andsoconstructingasystemoflawandgovernmentthatthey,actinginthisway,willalsodowhattheyought(promotethegeneralpleasure).OneexampleofaBenthamitesolutionistheaccountofpunishmentgivenintheprevioussection.Peoplemovedbytheirowninterest(thedesireofavoidingthepainofpunishment)areledtodowhattheyoughtbyacorrectsystemofthreatenedpunishments.Theyare,thatis,ledbysuchSELF-INTEREST(p.261)toconsideralsotheinterestsofothersandactinamannerwhichpromotesthegeneralhappiness.ThesameprinciplecanbeappliedtothedesignofinstitutionsandwasappliedbyBenthamtoeverythingfromthedesignofgovernmentsdowntothedesignsofprisons.Thekeyprinciple,inhiswords,wasthe‘dutyandinterestjunctionprinciple’.Thingshadtobecunninglyarrangedsothatdutyandinterestcoincided.ThisisthekindofprojectforwhichJ.S.MillworkedasayoungmanwhenhewasanunreconstructedBenthamite.Laterhebecamemoredetached.Ifwenowmoveontoinvestigatethis,andthecriticismofBenthamthatitsupplies,weshouldrememberthatBenthamismwasneverintendedtobeanythingotherthanasystemofpoliticsandgovernment.3TraditionalInterpretation:MillAlthoughMilltookhimselftobeautilitarian,hewantedtobeautilitarianwithadiffer-ence.HehimselfwasoneofBentham’scritics,andincriticizingBenthamitisarguablethathecriticizedthecentralcoreofutilitarianismitself.Such,atleast,ispartofthesub-stanceofthetraditionalinterpretationofMill;anditisthistraditionalinterpretationwhichwillbeexaminednext,concentratingonUtilitarianism(1963–),whichisthetextthateverybodyreadsandwhichMillhimselftooktobeadefenceofthedoctrine.Thefirstpointofcriticismconcernschapter2,whichisentitled‘WhatUtilitarianismIs’.Millherebeganbysaying:Thecreedwhichacceptsasthefoundationsofmorals,Utility,ortheGreatestHappinessPrinciple,holdsthatactionsarerightinproportionastheytendtopromotehappiness,761\nROSSHARRISONwrongastheytendtoproducethereverseofhappiness.Byhappinessisintendedpleasure,andtheabsenceofpain;byunhappiness,pain,andtheprivationofpleasure.ThisisexactlylikeBentham’sremarksquotedabove:thereisthesameuseoftheword‘tend’andthesamepromiscuousmovementbetween‘utility’,‘happiness’and‘pleasure’.However,Millshortlyafterwardsclaimedthat‘itisquitecompatiblewiththeprin-cipleofutilitytorecognizethefact,thatsomekindsofpleasurearemoredesirableandmorevaluablethanothers’.Here,atleastonthetraditionalinterpretation,thetroublestarts.Forthecentralpointofutilitarianismisitsclaimthatthereisasinglesourceofvalue(whichmaybecalledbydifferentnames).Ifthisisso,thenitwouldseemthatdistinctionsinthisvaluecanonlybequantitativeratherthanqualitative.YetMillexplicitlywantsqualitativedistinctions.Hefamouslysaysthat‘itisbettertobeahumanbeingdissatisfiedthanapigsatisfied;bettertobeSocratesdissatisfiedthanafoolsatisfied’.Thismaywellbetrue;butthetraditionalobjectionholdsthatthisisbecauseMillhasintroducedvaluesinadditiontosatisfaction(orhappinessorutility).ThehumanbeingandSocrates,thatis,areonlybetterbecausetheyhavemoreoftheseothervalues,compensatingfortheirlessersatisfaction.In1859,justbeforeMillwroteUtilitarianism,hewroteafamousdefenceofpoliti-calliberty,OnLiberty(1963–).Thiscontainsastrongsupportofthevalueofsuchthingsastruth,individualityandautonomy.Itsobjectis,asMillsays,‘toassertoneverysimpleprinciple’;whichis‘thatthesoleendforwhichmankindarewarranted,indi-viduallyorcollectively,ininterferingwiththelibertyofactionofanyoftheirnumber,isself-protection’.Millholds,thatis,thatthereshouldbeapurelyprivateareabeyondthecontroloflaworpublicopinion.Hesaysspecificallythatsomeone’s‘owngood’is‘notasufficientwarrantforinterference’.Thesemayallbeadmirablesentiments,butthequestionis,again,whetherMill’sassertionofthemiscompatiblewithutilitarianism.Millhimselfsaysthatitis.Shortlyaftertheassertionofhis‘simpleprinciple’hesaysthatheregards‘utilityastheultimateappealonallethicalquestions’.However,thestandardinterpretations,startingwiththenineteenth-centuryattackbyJ.F.Stephen(1829–94),againholdthatMillcanonlydefendlibertybyrelyingonotherindepen-dentvalues;andsoagainbybreakingwithutilitarianism’ssinglevalue.Conversely,ifmaximizingutilityisthesolepermittedgoal,thenpaternalistinterferencewithotherswouldsometimesbepermitted;foritwouldbepermittedpreciselyonthoseoccasionsonwhichsuchinterferenceincreasedtheirutility.Utilitarianismisconcernedwithmaximizingthegood,irrespectiveofwhosegooditis.Hence,contrarytowhatMillasserts,someone’sgoodwouldbeasufficientwarrantforinterference.Herearetwoversionsoftheobjectionthat,bytryingtoincludemoreintheideaofhappinessthanBenthamdoes,Millinfactbreakswiththecentralspiritofutilitarianism.Thegreatestnegativeenergyofthecommentatorshasbeenreservednotforchapter2butforchapter4ofUtilitarianism.Thisshortchapter,entitled‘OfWhatSortofProofthePrincipleofUtilityisSusceptible’,hasproducedanenormousamountofcriticalcommentary.MostcommentatorssinceF.H.Bradley(1846–1924)andSidgwickhaveheldthatMillheremadeelementaryandembarrassingerrors;sothatithasbeen762\nBENTHAM,MILLANDSIDGWICKsuggested(Berger1984:45)thatMill’scurrentreputationwouldbeverymuchhigherifthisshortchapterhadnotbeenwritten.Itseemsthatthechapterintendstoprovideaproofoftheutilityprinciple.Thesup-posedfallaciesarehencefallaciesofinference.Thetwomostfamousonesoccurinthethirdparagraph.HereMillstartsbysuggestingananalogywithprovingthatsomethingisvisible.‘Theonlyproofcapableofbeinggiventhatanobjectisvisible’,hewrites,‘isthatpeopleactuallyseeit.’Similarlyforsomethingbeingaudible:theproofisthatpeoplehearit.‘Inalikemanner’,hecontinues,‘thesoleevidencethatitispossibletoproducethatanythingisdesirable,isthatpeopledoactuallydesireit.’ThestandardcriticismofMillisthatheherederivesthedesirable(whatoughttobethecase)fromthedesired(whatisthecase).Thatis,heconfusesoughtandis;orintheterminologyofG.E.MOORE(1873–1958)(chapter38)hecommitsthenaturalisticfallacy.Noristhistheendoftheinfamoushowlersthathavebeenfoundinthisshortpara-graph.Foraftershowing,ashethought,thateachperson’shappinesswas‘agoodtothatperson’,Millconcludedthat‘thegeneralhappiness’was‘therefore,agoodtotheaggregateofallpersons’.Herebysleightofhandheseemstomovefromegoismtoaltruism;fromtheclaimthateveryoneisinterestedintheirownhappinesstotheclaimthattheyarethereforeinterestedineveryoneelse’shappiness.4Reinterpretation(1):TheArtofLifeThiscatalogueofapparentdisastersmightbethoughttospelltheendofMillasaseriousthinker;andsoalsotheendofhisrevisedversionofutilitarianism.Ontheotherhand,Mill’sowncriticismsofBenthammightbeheldtospelltheendoftheoriginalutilitarianproject.However,Millhasrecentlybeensubjecttoasustainedprocessofreinterpretationanddefence.A‘newMill’hasemergedtowhomitishardertoattachthetraditionalcriticisms;andperhapsthatreinterpretationcanprovidesustenanceforarevisedformofutilitarianism.Itwasnotedinsection3thatmuchofthecriticismhasbeendevotedtoasingleparagraphinashortchapterofasingleoneofMill’sworks.OnewaythatMillhasbeenreinterpretedisbygettingabroaderview.So,inUtilitarianismitself,considerationhasbeengiventochaptersnotsofarmentioned,particularlychapter5onJUSTICE(chapter8).ButithasalsobeennotedthatUtilitarianismisarelativelysmallpartofMill’stotaloutput.Mill’slargestworks–whetherthisismeasuredbybulk,bycontemporaryesteem,orbytheamountofdescriptionhehimselfaccordstheminhisAutobiography(1963–),whichdescribeshisintellectualdevelopment–areworksoflogicandmetaphysics.MillmadehisnameamonghiscontemporarieswithhisSystemofLogic(1963–),whichwaspublishedin1843,andtowardstheendofhislifehewrotehismassivemetaphysicalworkAnExaminationofSirWilliamHamilton’sPhilosophy(1963–),whichwaspublishedin1865.ItisthereforecompletelyappropriatethatJohnSkorupski’s(1989)largestudyandreinterpretationofMillpaysmuchmoreattentiontothelogicthananythingelse.Thisbalancewillnotbefollowedinthepresentchapter,sincestudentsreadUtilitarianismandOnLibertymuchmorethanMill’sotherbooks.However,itdoessuggestthatabetterviewcanbegainedofthesefamiliartextsiftheyareplacedinthecontextofMill’soverallthought,andhenceoftheLogic.763\nROSSHARRISONThepartoftheLogicthatismostrelevantfortheunderstandingofMill’sethicalandpoliticalthoughtisthelastchapter,whichisabout‘moralityandpolicy’.Init,Millsaysthat‘apropositionofwhichthepredicateisexpressedbythewordsoughtorshouldbe,isgenericallydifferentfromonewhichisexpressedbyisorwillbe’.Inotherwords,Millhereveryclearlydistinguishesbetweenisandought.Furthermore,hesaysthateveryarthasitsleadingprinciple.Theleadingprincipleofarchitecture,forexample,isthatitisdesirabletohavebuildings.StandingabovealltheseparticularartsistheArtofLifeitself.Itsleadingprincipleisthatwhatis‘conducivetohappiness’isdesirable.Milldividestheartoflifeinto‘threedepartments,Morality,PrudenceorPolicy,andAesthetics;theRight,theExpedient,andtheBeautifulorNoble’.FromthisitcanbeextractedthatMillnotonlyclearlydistinguishesbetweenisandought,butthathealsodistinguishesbetweenwhatisgoodandwhatitisrighttodo.Otherwiseput,heonlytakesmoralityaspartofwhattheprincipleofutilitywouldpromote.If,withthisclue,wenowreturntoUtilitarianism,buttothelessfamiliarchapteronjustice,wefindthereMillsayingthat‘wedonotcallanythingwrong,unlesswemeantoimplythatapersonoughttobepunishedinsomewayorotherfordoingit;ifnotbylaw,bytheopinionofhisfellowcreatures;ifnotbyopinion,bythereproachesofhisownconscience’.Somoralityseemstoconcernonlythosepartsoflifeinwhichthereoughttobesanctionsdeterringspecifickindsofconduct.Fortherestoflife,eventhoughtheremaybeprinciplesofexpediency,ormattersofaesthetictaste,thereisnotinthesamewayobligationsaboutwhatoughttobedone.Theimportanceofthisisthattherootideaofutilitarianismasapersonalethicseemedtobethateveryoneoughtallthetimetobedoingwhatproducedthegreatestutility;animpossiblystrenuousobligation.InsofarasBenthamavoidsprivateethics,healsoavoidsthisproblemofstrenuousness.ButMill,unlikeBentham,isconcernedwithprivateethics.Furthermore,thestatementsofthe‘creed’quotedinsection3above(‘actionsarerightinproportionastheytendtopromotehappiness,wrongastheytendtoproducethereverseofhappiness’,currentlydubbedthe‘proportionalitycriterion’)doesseemtoadvisesuchastrenuousperformance.Itsaysthat,inpro-portionasutilitycanbeincreased,sosomethingoughttobedone.However,ifweconsiderinsteadthestatementjustquotedfromchapter5(currentlydubbed‘thepun-ishmentcriterion’),thenthisstrenuousnesscanbeavoided.Millcanbetakentoholdthatonlycertainverycentralhumaninterestsneedprotectionbytheimpositionofsuchsanctions.Onlythesecountasmorality;andhencemoralityneednotbecomeoverstrenuous.Assumingthatautonomyandsecurityareamongthesevitalhumaninterests,wenowseehowwecangetacoherentaccountofmoralityandlaw,aswellasofMill’sproposalsinOnLiberty.Forwhensuchvitalinterestsareconcerned,theyhavetobeprotectedeitherbylaw(usingthethreatoflegalpunishment)orelsebymorality(usingthethreatofpublicdisapproval).Otherwise,inparticularwiththeexerciseofmattersthatarepurelyprivate,thereshouldbenosuchsanctions.TheaccounthereislikeBentham’stheoryofpunishment.Certainkindsofconductaresubjecttosanctions,eitherlegalormoral.Thereasonpeopledonotperformtheseactions(inthemain)isbecauseoftheirdesiretoavoidthesesanctions.So,thecon-nectionbetweenactionandutilityisindirect.Therefore,ifMillisstillautilitarianonthisinterpretation,itisoftheindirectratherthanthedirectkind.764\nBENTHAM,MILLANDSIDGWICK5Reinterpretation(2):HappinessandIndirectUtilitarianismTheideaofamoralsanctiondidnotoriginatewithMill.Benthamalsotalkedofa‘moral’sanction.Inchapter3ofhisIntroductionhedescribesfoursanctions:physical,political,moralandreligious.Hethinksthatpeoplecanbedeterredbythesimpleideaofpain(whichmightbecausedphysically,legally,sociallyorbyGod).Millisalsointerestedinhowpeoplecanbemadebysanctionstododesirablethings.ButwhatMilladdstoBenthamisaricherpsychology,involvinguseoftheassociationofideas.WhenMilldescribesinhisAutobiographyhisseparationfromtheoriginalutilitariancreedhesays:‘Inever,indeed,waveredintheconvictionthathappinessisthetestofallrulesofconduct,andtheendoflife.ButInowthoughtthatthisendwasonlytobeattainedbynotmakingitthedirectend.’Asheputsit,‘theonlychanceistotreat,nothappiness,butsomeendexternaltoit,asthepurposeoflife’.Thisdescribesindividualmotivation.Yetthesamepointsapplyaswithmorality.Inbothcases,happinessistheultimateend.Butinbothcasestheapproachisindirect.Theimmediateorapparentendofactionmaybenothappinessbutsomethingelse.Anditisonlythisotherthingthatisexplainedorjustifiedbyhappiness.Mill’smainconcernintheinfamousfourthchapterofUtilitarianismwasnottoshowthathappinessisanendofaction(whichhepassesoverquicklyinthemuch-discussedparagraph)but,rather,toshowthatitistheonlysuchend.Forthisheneedshisfulleraccountofhappinessinwhichthetheoryoftheassociationofideasispressedintoexplanatoryservice.Ashesaysthere,somethingwhichbecomesassociatedwithsome-thingasameanstoanendnaturally‘comestobedesiredforitself’.So,forexample,virtuecanstartasameanstohappiness,andthenbecomeanapparentlyself-sufficientendofaction.Theexplanationofactionbyhappinessis,again,indirect.Thereis,therefore,considerablecurrentsupportfortheinterpretationofMillasanindirectutilitarian.Thismightseemtodeterminealsoanolderinterpretativeissue,thequestionofwhetherMillisaruleutilitarian.ThiswasproposedbyJ.O.Urmsonin1953(seeSchneewind1968:179–89).However,themainweightofthesubsequentdebatehasrunagainstthisinterpretation;andthemorerecentinterpretersofMillasanindirectutilitarianalsodonotusuallythinkthatthismakesMillaruleutilitarian.Mill,RulesandConductOnanyinterpretation,Millclearlygivesaspecialpositiontorules.Moralityisnotonlydefinedas‘therulesandpreceptsforhumanconduct’;Millalsoemphasizedthespecialpositionof‘secondary’principles.Hethoughtthatthepastexperienceofhumankindhasshownuswhichkindsofactionpromotehappinessandwhichdonot.‘Duringallthattime’,hesaid,‘mankindhavebeenlearningbyexperiencethetendenciesofactions.’Allthisis,however,compatiblewithMillnotmeaningrulestoapplyinmorethanashorthandmanner.Therules,thatis,canbeheldtobepracticalrules,givingknowledgeaboutthetendenciesofactionswhennobetterinformationisavailable.Althoughitispracticaltousesuchrules,thisdoesnotinitselfmeanthattheactionisonlyjustifiedbytherule;andhenceitisnotruleutilitarianisminthenarrow,orstrict,sense.Furthermore,765\nROSSHARRISONoncetheartoflifeconceptionistakenseriously,moreisjustifiedbyutilitythanmorality.Evenif,thatis,moralityconsistsonlyofrulestherearealsotheotherpartsoftheartoflife,suchasprudenceandaesthetics.Theyescapetheanalogywithlawandpunish-ment;andsoescapewhateverpressurethisanalogymayasserttowardsaruleutilitarianinterpretation.TheproposedinterpretationofMillasanindirectutilitarianissometimesmadebycon-trastwithBentham.ButBenthamcanalsobemadeouttobeanindirectutilitarian.Hewascentrallyconcernedwiththecreationoftherightsystemoflaw.Therefore,whysomeoneactsinsucharightsystemcanalsobeinterpretedasbeingindirectlycon-nectedwithutility.Theyactdirectlybecauseitisthelaw;orbecausetheyareafraidofthepunishment.Butthelawistherightlawbecauseitpromotesutility.Benthamalsolaysdownsecondaryendsorgoals,whichareparticularlyprominentinhiswritingsoncivillaw.ThefoursecondaryendshelistsinhisCivilCode(1843b)aresubsistence,abundance,equalityandsecurity.Thesearewhatagoodsystemoflawshouldpromote.Theyagain,asinMill,standinthepositionofbeingintermediateprinciplesofjustifi-cation.Again,asinMill,theycanbejustifiedbyreferencetothepast,ortothegeneral,experienceofhumankindbecausetheyareendsthatpromotethefundamentalprincipleofmaximizingutility.6Mill’sMetaphysicsandLogicItwasnotedabovethatrecentworkonMillhasstressedthebreadthofhisprojectandhasrevealedthatitscentreofgravityismoreinthewritingsonlogicthaninthemorefamiliarethicsandpolitics.Noproperconsiderationcanbegivenhereofthelogicandmetaphysics;butitwillbeusefultohaveaglanceatMill’soverallprojectbeforereturningfinallytohissupposedproofofutility.InbothmoralsandmetaphysicsMillcanbeseenasworkinginthecentraltraditionofBentham.Theyboththoughtthattherearebasicallytwokindsofphilosophicalposi-tion,whichcanbedistinguishedbythemethodstheyuse.OntheonehandisthemethodofrelyingonINTUITIVE(pp.726–33)oraprioriknowledge.Ontheotheristheempiricistmethodofrelyingonexperience.Millalignshimselfwiththelatter.Thismethodholdsthatmoralandmetaphysicalconclusions,andhencepoliticalends,arenottobeestablishedbysomeinnateintuitivefaculty,ormoralsense;theyaretobeestablishedaccordingtotheeffectsofvaryingcoursesofpossibleactionondemon-strableandunmysteriousexperience,namelythehappinesstheyproduce.ThisdistinctionnotonlyexplainsMill’spositioninmoralandpoliticalphilosophybutalsohiswiderintellectualproject.ItgivesthecentralpurposeofhisLogicaswellashisExaminationofHamilton.ThechiefobjectoftheLogicistoshowhowtherecanberightreasoninganddiscoveryoftruthwhennobasisisusedexceptforexperience.Millhereattemptstoshowthatallcandidatesfortruthsknownintuitivelyandindependentofexperience(suchasMATHEMATICALTRUTHS(chapter11)orthebeliefthateveryeventhasacause)areinfactfoundedonexperience.TheLogicalsodealswithMEANING(chapter3)andwithDEDUCTIVEARGUMENT(p.287)(whereheholdsthatthereal766\nBENTHAM,MILLANDSIDGWICKinferenceinasyllogismistheestablishmentofthemajorpremiseandisthereforeinductive).However,itscentralfocusistoproduceanempiricistPHILOSOPHYOFSCIENCE(chapter9),showinghowknowledgemaybeacquiredfromexperience.Mill’sexperientialtreatmentoflogicandmathematicsisinstarkoppositiontotheprevailingtoneofthiscentury’sphilosophyoflogic.However,asrecentworkhasbroughtout,ithassimilaritieswiththeworkofQUINE(b.1908)(chapter4).NeitherMillnorQuineultimatelyholdlogicormathematicstobespecial.Intheend,theytooaresubjecttothetribunalofexperience.SoMill’scanalsobeseenasanaturalizingproject;givingnaturalexplanationratherthanapriorijustification.Thingsthoughttobenecessaryareexplainedastheproductofthenaturaleffectsoftheenvironmentonthemind.Thisappliesforlogicandmathematicswiththeirsupposedintuitive,ornec-essary,truths.Italsoappliesformorality,analogouslysupposedbymanytobeintuitiveornecessary.BothSkorupski(1989)andScarre(1989),whobringoutthisanalogywithQuine,findatensionbetweenthisnaturalismandMill’sownmetaphysics(whichischieflyfoundinhislongExaminationofHamilton).IntheExaminationMillpresentsanaccountofrealityinwhichallitemsareconstructedfromthebasicatomsofexperience.Suchconstructionsareaperson’sownMIND(chapter5),OTHERMINDS(p.692),aswellasOBJECTSINTHEEXTERNALWORLD(pp.726–33)(whicharefamouslyclassifiedas‘perma-nentpossibilitiesofsensation’).TheExaminationiscurrentlynotmuchreadorstudied,butoneinterestofitisthatitcontains(inchapter12)thefirstappearanceoftheargumentbyanalogyforthenatureofotherpeople’sminds,theattackonwhichhasdonesomuchtoshapecurrent,post-Wittgensteinianphilosophyofmind.ThebasisofSkorupskiandScarre’sattackisthatthephilosophyofsciencedemandsarealistattitudetotheworld,whichisundercutbythePHENOMENALISM(pp.53–4)oftheMETAPHYSICS(chapter2).7ProofofthePrincipleofUtilityIthasbeennotedthatMill’sdefenceoftheuseofexperienceagainsttheuseofaprioriintuitionextendsalsotohiswritingsinethics.ThechiefobjectofUtilitarianismistodemonstratethatethicalprinciplesdonotneedtorestonsuigenerisethicalintuitions.Hencethepointofthelongchapteronjustice.Justiceisshownnottobeaseparateintuition,buttobefoundedontheobservableutilityoftheactionsthatjusticerecommends.Hencealsothepointoftheso-calledproof.Itisnotintendedasadeductiveproofofutility.Itisnotintendedtoshowthatutilityisanintuitiveoraprioriprinciple.Rather,itisintendedtorevealitsempiricalsupport;toshowhowtheprincipleconnectswithobservationsofwhatactuallyhappens.Itappealsnottomoralsense,ormoralin-tuitions,but,rather,toactual,observable,desires.ItisquiteclearfromthetextitselfthatMillneverintendeditasadeduction.Earlierinthework(inchapter1)hesaysthat‘questionsofultimateendsarenotamenabletodirectproof’;theinfamouschapteritselfisentitled‘Ofwhatsortofproof...’;anditstartswithMillrepeatinghisclaimthat‘questionsofultimateendsdonotadmitofproof’.Soinsteadofproofwehavesuggestions;notadeduction,butexperience.767\nROSSHARRISONSomethingexactlysimilarhadhappenedpreviouslyinBentham,whoalsoclaimsthatthefirstprinciple,beingafirstprinciple,cannotbeprovedbysomethingmoreulti-mate.SoBenthamprovides,insteadofaproof,somethingwhichmayleadthereaderto‘relish’theprinciple.Mill,similarly,providessomethingwhichmayleadthereadertolookontheprinciplemorefavourably.Asanempiricaltask,theclaimisthat,toseewhatmightbestserveasanultimateendofconduct,somecontrolcanbegotbyseeingwhat,asamatteroffact,peopleactu-allydotreatasends.Thedesirableisnottakentobethesameinmeaningaswhatisdesired.Rather,byobservingwhatpeopleactuallydesire,wegetsomekindofempiri-calcontrol,ortest,aboutwhatitispossibletodesire.And,byseeingwhatitispossibletodesire,wegetananalogousempiricalcontrol,ortest,onwhatshouldbedesired.Itisanecessarycondition,notasufficientone:thereisnopointinanethicsthatproposessomethingasdesirablewhen,asamatterofempiricalfact,itisnotpossibletodesireit.WhatMillistryingtoshowisthat,asamatteroffact,happinessistheonlythingthatisactuallydesired(asanend);hencethatitisreasonabletosupposethatitistheonlythingthatitispossibletodesire(asanend);hencethatitistheonlyproperobjectofmorality,prudenceandlife.Theinitialempiricalobservationheremaybeheldtobehighlydubious;althoughitisobviouslyhelpedbyMill’senlargedsenseofhappiness,whichwasdiscussedabove.Followingfromthisenlargedsense,itwillnotbearefuta-tionthatpeopletakethemselvestobeaimingatsomethingquiteotherthanhappiness(aslongas,thatis,happinesshastoentertheexplanationofthis).8BenthamonClarificationIthasbeenseenhowMillwishedtoconnectsignificantthoughtwithexperience.Benthamalsotookexperienceasaguide,butwasmorepreparedthanMilltotalkaboutthisgivingmeaning.Withouttheprincipleofutility,hesaid,wehave‘soundsinsteadofsense’;andhisconstantaimistoextirpatenonsense.Theconnectionofboththecentralevaluativeprincipleandthecentralpsychologicalprinciplewithpainsandplea-sures–thatis,withtherawdataofexperience–endowsboththesecentralprincipleswithmeaning.Bentham’spracticalprojectofproducingtheperfectsystemoflawledhimintoaninvestigationofthenatureoflaw.Law,hethought,hadtobeclarifiedandunderstood(sothat,onceunderstood,itcouldbereformed).Understandingthelawmeansunder-standingrightsandduties.Understandingisgainedbyplacingthingsincontactwithexperience.Yet,sincerightsanddutiesseemtobeneitherobservableentities,norcom-plexesofobservableentities,thiswouldnotseemtobepossible.However,Benthamsolvedthisproblembyinventingatechniquehecalled‘paraphrasis’,whichanticipatestwentieth-centurytechniquesofanalysis.Theideaisnottounderstandproblematicwordsinisolationbut,rather,‘somewholesentenceofwhichitformsapartistranslatedintoanothersentence’.Sointheanaly-sisofwhatBenthamcalled‘fictionalentities’(suchas‘right’,‘duty’,‘property’,‘obliga-tion’,‘immunity’and‘privilege’:thewholelanguageofthelaw),heuseshistechniqueofparaphrasistoplacethesetermsinsentencesforwhichhethengivessubstitute768\nBENTHAM,MILLANDSIDGWICKsentencesnotcontainingtheoffendingterm.Forexample,sentencesaboutrightsareexplainedbyBenthamintermsofsentencesaboutduties.Aparticularrightisforhimthebenefitthatisconferredonsomeonebytheimpositionofdutiesonothers.Withdutieswestill,ofcourse,havefictionalentities.Butthese,inturn,canbeplacedinsen-tencesthataretranslatedintosentencesaboutthethreatofpunishment.Punishmentis,forBentham,thethreatoftheimpositionofpain.Sohere,atlast,wereachwhatBenthamcallsrealentities;thatis,thingsthatcanbedirectlyunderstoodbypercep-tion.Wereachpainandpleasure,whicharetheultimateclarifiersbothofhowthelawisandalsoofhowthelawoughttobe.NaturalRightsandNonsenseBenthamwasopposedtotheexistenceofnaturalrights.Theyformedpartoftheoppos-ingrhetorictoutilityandwerethoughtbyhimtobenonsense.Atfirstsightitmightbethoughtthattheycouldbesavedbythesametechnique.However,comparinganaturalrightwithalegalrightexposesthedifference.Bothcanbeanalysedintermsofcorre-spondingduties.Yet,aswesawabove,Benthamanalysesalegaldutyintermsofthelaw(orthreatofpunishment)whichcreatesit.Thereisnocorrespondinglaw,heholds,withrespecttosupposednaturalduties.Henceheholdsthatnaturalrightsarejustimaginaryrights.Bentham’smostfamoussloganexpressingthisviewisthattalkofnaturalrightsis‘nonsenseonstilts’.ThiscomesinaworknormallycalledAnarchicalFallacies(1843a),whichisacriticalanalysisoftheFrenchDeclarationofRights,whichwasissuedin1791.BenthamthinksthattheFrencharenotreallydeclaringexistingrightsbut,rather,givingreasonswhythereoughttoberights.Yet,asheputsit,‘areasonforwishingthatacertainrightwereestablished,isnotthatright;wantisnotsupply;hungerisnotbread’.Sotosupposethatsuchrightsactuallyexistisnonsense.Furthermore,tosupposethattheyareunalterable–thatis,tosupposeinadvanceofexperiment,thattheyarerightforalltime–isevenworsethannormalnonsense.Itisnonsenseonstilts.9SidgwickFinally,acodaonSidgwick.Bentham’sCivilCodewasoneofSidgwick’sfavouritebooksandhestartedhisintellectuallifeasafollowerofMill.However,justlikeMill,bebrokewithandcriticizedhismentors.Hewasstillpreparedtosaythathewasautilitarian,but,again,theutilitarianismiswornwithadifference.Sidgwickproducessofunda-mentallydifferentabaseforutilitarianismthat,again,itmightbethoughtthatheremovesitscentralpoint.IthasbeenseenhowbothBenthamandMilltookthemselvestobeopposingschoolsofthought(orpoliticalmovements)thatreliedupondirectintuitionoftherightanswer.Againstthistheybothposedobservationandexperience.YetalthoughSidgwickarguedforutilitarianism,hedidsopreciselyonanintuitionistbase.Hesawhimselfashealingthesplitinthepreviousgeneration:thewarbetweentheutilitariansandtheintuitionists.Sidgwickwasboth.Healingsplits,orunderstandingdifferentpointsofview,wasverymuchSidgwick’spersonalandphilosophicaltemperament.HiscentralworkiscalledTheMethodsof769\nROSSHARRISONEthics(1874).Thisis,asthetitlesays,astudynotjustofonebutofseveralmethods.Intuitionism,utilitarianism,commonsenseandegoismareallexamined.Thehistoryofthesubjectisfullydeployed.Theproblemishow,beingcompletelyimpartialinthefaceofallthismaterial,tofindtherationalbasisforaction.TheMethodsishenceastudyofgoodreasons.Themaximsofcommon-sensemoral-ity,hethinks,donotmeetthecriteriahelaysdownforsomethingbeinganintuitivelygoodreasonforaction.However,thesecriteriaaremetbycertain‘absolutepracticalprinciples’ofamoreabstractnature,suchasthatfuturegoodisasimportantaspresentgood.Anothersuchprincipleisthat,asSidgwickputsit,‘thegoodofanyoneindivid-ualisofnomoreimportance,fromthepointofview(ifImaysayso)oftheUniverse,thanthegoodofanyother’.Thisinvolvesutilitarianism;henceSidgwickprovidedutilitarianismwithanintuitionistfoundation.However,Sidgwickthinksthategoismisalsoanintuitiveprincipleofaction.Sowearecaught,hethinks,ina‘dualismofpracticalreason’withtwoindependentprinci-plesthatarebothrationallycompelling:anegoisticprincipletellingmetomindmyownhappinessandanaltruisticprincipletellingmetomindthegeneralhappiness.ThissplitwasthreatenedinMillandBentham,buteluded(forBentham,bymakingtheegoisticprincipleapsychologicalprincipleratherthananevaluativeone;forMill,byallowingthembothtobeevaluativebutgivingthemdifferentspheres).ForSidgwick,however,theproblemofcollisionisharsh.Itwouldonlybeavoided,hethinks,ifGod(workingasasortofdivineBenthamitelegislator)keptthetwoinstep.However,Sidgwick,whohadconsiderablereligiousdoubts,wasreluctanttointroduceGodforthispurpose.ThefamouslastwordsofthefirsteditionoftheMethodsstatethat‘theprolongedeffortofthehumanintellecttoframeaperfectidealofrationalconductisseentohavebeenforedoomedtoinevitablefailure’.(Sidgwickhimselfsaidthathewantedtowriteabookinwhichthefirstwordwas‘ethics’andthelastword‘failure’;inthis,atleast,hesucceeded.)MillwagedwaroncontemporaryCambridgeprofessors,likeWhewell,whousedintuition.Mill,likeBentham,foughtfromoutsidetheacademy.Theybothproposedasecularethic,incompetitionwiththechurch-supportedEnglishuniversitiesandes-tablishment.Sidgwick,bycontrast,wasaCambridgeprofessor.Likehispredecessor,Whewell,heusesintuition,albeitwithdifferentresults.HetriestodispensewithGod;butthisgiveshimproblems.However,writingasaprofessor,hesucceedsingivingathoroughlyprofessionaltreatmentofthesubject,recognizableassuchinthiscentury.Afterall,itwasSidgwickwhoproducedthecritiquethatG.E.Moore(alaterCambridgeprofessor)usedtotheeffectthatMillconfusedthedesiredwiththedesirableandsocon-fusedisandought.Fromthiscritique,muchtwentieth-centuryethicalthoughtfollows.FurtherReadingReadersmightreasonablybeginwithBentham,AFragmentonGovernmentandAnIntroductiontothePrinciplesofMoralsandLegislation;withMill,UtilitarianismandOnLiberty;andwithSidgwick,TheMethodsofEthics,beforeturningtosecondarysources.Dinwiddy(1989)providesagoodbriefintroductiontoBentham,withamorecomprehensiveexaminationinHarrison(1983).Hart(1982)offersengagingessayswhichhavesetthetoneofcontemporaryBenthamstudies.Bentham’stheoriesofjustice,government,lawanddemocracyare(respectively)consideredbyKelly(1990),Lyons(1991),Postema(1986)andRosen(1983).770\nBENTHAM,MILLANDSIDGWICKRyan(1974)givesanexcellentandclearintroductiontoMill,butsomereaderswillwanttomovequicklytothemoredemandingrecentstudiesbyScarre(1989)andSkorupski(1989).Schneewind(1968)andThomas(1985)arealsovaluableandSkorupski(1998)isanauthoritativesurveyofallaspectsofMill’sthought.Gray(1983),Ten(1980)andRiley(1998)providedifferentperspectivesonMill’saccountofliberty.Robson(1968)exploresMill’ssocialandpoliticalthought,whileaspectsofhisethicalandpoliticalthinkingarediscussedbyBerger(1984)andSemmel(1984).Crisp(1997)isaguidebookspecificallydevotedtoMill’sUtilitarianism.Schneewind(1977)placesSidgwickinhistoricalcontext,andSchultz(1992)reflectsarecentrenewalofinterestinSidgwick’sphilosophy.ReferencesBenthamBentham,J.1776:AFragmentonGovernment.InBentham(1968–).——1789:AnIntroductiontothePrinciplesofMoralsandLegislation.InBentham(1968–).——1843a:AnarchicalFallacies.InBentham(1962),vol.2.——1843b:PrinciplesoftheCivilCode.InBentham(1962),vol.2.——1962[1838–43]:TheWorksofJeremyBentham,11vols(editedbySirJ.Bowring).NewYork:RussellandRussell.——1968–:CollectedWorks(editedbyJ.H.Burns).London:AthlonePress.WritersonBenthamDinwiddy,J.1989:Bentham.Oxford:OxfordUniversityPress.Harrison,R.1983:Bentham.London:Routledge.Hart,H.L.A.1982:EssaysonBentham.Oxford:OxfordUniversityPress.Kelly,P.J.1990:UtilitarianismandDistributiveJustice.Oxford:OxfordUniversityPress.Lyons,D.1991:IntheInterestoftheGoverned.Oxford:OxfordUniversityPress.Postema,G.J.1986:BenthamandtheCommonLawTradition.Oxford:OxfordUniversityPress.Rosen,F.1983:JeremyBenthamandRepresentativeDemocracy.Oxford:OxfordUniversityPress.MillMill,J.S.1843:ASystemofLogic.InMill(1963–),vols7–8.——1859:OnLiberty.InMill(1963–),vol.18.——1863:Utilitarianism.InMill(1963–),vol.10.——1865:AnExaminationofSirWilliamHamilton’sPhilosophy.InMill(1963–),vol.9.——1873:Autobiography.InMill(1963–),vol.1.——1963:CollectedWorks,33vols(generaleditorJ.M.Robson).London:UniversityofTorontoPressandRoutledgeandKeganPaul.WritersonMillBerger,F.R.1984:Happiness,JusticeandFreedom.Berkeley:UniversityofCaliforniaPress.Crisp,R.1997:MillonUtilitarianism.London:Routledge.Gray,J.1983:MillonLiberty:ADefence.London:Routledge.Riley,J.1998:MillonLiberty.London:Routledge.Robson,J.M.1968:TheImprovementofMankind:TheSocialandPoliticalThoughtofJohnStuartMill.Toronto:TorontoUniversityPress.Ryan,A.1974:J.S.Mill.London:Routledge.Scarre,G.1989:LogicandRealityinthePhilosophyofJohnStuartMill.Dordrecht:Kluwer.771\nROSSHARRISONSchneewind,J.B.(ed.)1968:Mill.London:Macmillan.Semmel,B.1984:JohnStuartMillandthePursuitofVirtue.NewHaven,CT:YaleUniversityPress.Skorupski,J.1989:JohnStuartMill.London:Routledge.Skorupski,J.(ed.)1998:TheCambridgeCompaniontoMill.Cambridge:CambridgeUniversityPress.Stephen,J.F.1967:Liberty,Equality,Fraternity,newedn(editedbyR.J.White).Cambridge:CambridgeUniversityPress.Ten,C.L.1980:MillonLiberty.Oxford:OxfordUniversityPress.Thomas,W.1985:Mill.Oxford:OxfordUniversityPress.SidgwickSidgwick,H.1874:TheMethodsofEthics.London:Macmillan.WritersonSidgwickSchneewind,J.B.1977:Sidgwick’sEthicsandVictorianMoralPhilosophy.Oxford:OxfordUniversityPress.Schultz,B.1992:EssaysonHenrySidgwick.Cambridge:CambridgeUniversityPress.DiscussionQuestions1Aretheregoodreasonstobelievethathappinessistheuniquemeasureofvalue?2Couldtheutilitarianprinciplebeacceptedevenifpeopledidnotacttomaximizetheirownutility?3Shouldweequateutility,pleasureandhappiness?4Shouldouractionsandinstitutionsbeevaluatedbytheiractualandpossiblecon-sequences?Whatotherwaymightwehaveofassessingthem?5Whatjustifiespunishment?6Isutilitarianismequallysuitedtodesigningpublicinstitutionsandtoguidingprivateethics?7WasBenthamjustifiedinrejectingaroleforhistory,prejudiceandcustominhisdesignofasystemoflawandgovernment?8Isitreasonabletoattempttodesignaperfectsystemofpublicinstitutions?9Caninstitutionsbearrangedtoreconciledutyandinterest?10Doesutilitarianismfailifsomekindsofpleasureare‘moredesirableandmorevaluablethanothers’?11Onwhatgroundsarewejustifiedininterferingwiththelibertyofanotherperson?12DoesBenthamorMillprovideamoreacceptableaccountofhappiness?13Whatsortofproof,ifany,canbegivenfortheutilityprinciple?14ShouldweacceptMill’sclaim‘thatthesoleevidencethatitispossibletoproducethatanythingisdesirable,isthatpeopledoactuallydesireit’?15Doesthe‘punishmentcriterion’protectutilitarianismfrommakingimpossibledemandsonouractions?16IsitmorehelpfultoconsiderMillasaruleutilitarianorasanindirectutilitarian?WasBenthamanindirectutilitarian?17Whatfollowsfromtheclaimthattheartoflifeincludesprudenceandaestheticsaswellasmorality?772\nBENTHAM,MILLANDSIDGWICK18Shouldweseekanaturalexplanationwhereveranapriorijustificationisoffered?19Inwhatsense,ifany,mightone’sownmind,themindsofotherpeople,andobjectsintheexternalworldbesaidtobeconstructed?20IsMill’sphenomenalismcompatiblewiththerequirementsofthephilosophyofscience?21Isthefactthatpeopleoftentakethemselvestobeaimingatsomethingotherthanhappinessanempiricalrefutationoftheclaimthathappinessistheonlythingthatisactuallydesired?22HowsuccessfulwasBenthaminshowingushowtorecognizeandavoidnonsense?23Ifrightsanddutiesarefictionalentities,howcantheybeunderstood?24Are‘naturalrights’imaginary?25HowshouldweassessSidgwick’sprojectofprovidinganintuitionistbaseforutilitarianism?26Ifthereareabsolutepracticalprinciples,howcanwedeterminewhattheyare?773\n36PragmatismSUSANHAACKPragmatism,astyleofphilosophyinitiatedbyCharlesSandersPeirceandWilliamJames,isbestcharacterizedbythemethodexpressedinthepragmaticmaxim,accord-ingtowhichthemeaningofaconceptisdeterminedbytheexperientialorpracticalcon-sequencesofitsapplication.Peirce’saspirationtoareformed,scientificphilosophyfreeofmetaphysicalexcessesissustainedbytherealistelementsofhisviewofperceptionandofnaturallaws;mutedbynominalistleaningsinJamesandHegelianleaningsinDewey,pragmatismistransformedbySchillerintoarevolutionaryrelativism–and,inourowntime,byRichardRortyintoaconceptionofphilosophyas,notaformofinquiry,butonly‘carryingontheconversation’ofWesternculture.1Introduction:FromReformisttoRevolutionaryPragmatism‘Ithasprobablyneverhappened’,Peircewrotein1905,‘thataphilosopherhasattemptedtogiveageneralnametohisowndoctrinewithoutthatname’ssoonacquir-ingincommonphilosophicalusage,asignificationmuchbroaderthanwasoriginallyintended.’His‘pragmatism’,hecontinued,hadbythenacquiredasignificationsomuchbroaderthanhisoriginalintentionthatitwas‘timetokisshischildgood-by’and‘toannouncethebirthoftheword“pragmaticism,”whichisuglyenoughtobesafefromkidnappers’(Peirce1931–58:5.412:subsequentreferencesaretovolumeandparagraphnumber).Notunexpectedly,therefore,itisscarcelypossibletogiveanaccuratebriefstatementofthephilosophicaltendenciescharacteristicofpragmatism.Theneedtoacknowledge,forinstance,thesignificanceforhisconceptionofpragmatismofPeirce’sshiftfromearlynominalistsympathiestohismaturerealism,oroftheinfluenceonJames’sconceptionofpragmatismofhisdoctrineoftheWilltoBelieve,makesithardenoughtospecifywhatimportantphilosophicalideasweresharedbyPeirceandJames,thefoundersofpragmatism.AnditisharderyettofindacharacterizationthatwouldalsocomfortablyaccommodateDewey,SchillerandMead,letalonetoextendittoincludemorerecentpragmatists,neo-pragmatistsandsympathizersasdiverseasRamsey,Lewis,Sellars,Quine,Putnam,Rescher,Rortyandsoon.\nPRAGMATISMSchiller’sobservationthatthereareasmanypragmatismsaspragmatists,shrewdasitis,suggestsatworstonlyaformidablediversity;butthefactisthattherearenotmerelydifferent,butradicallyopposed,pragmatisms.Thescientificphilosophyenvis-agedbyPeirce’spragmatismcallsfor‘inquiryintothetruthfortruth’ssake’,andmain-tainsthat‘inordertoreasonwellitisabsolutelynecessarytopossess...arealloveoftruth’,andthatthetruth‘isSO,whetheryouorIoranybodythinksitissoornot’(Peirce1.44,2.82,2.135).Theliterarypost-philosophyenvisagedbyRorty’spragma-tism,instarkestcontrast,aspiresonlyto‘carryingontheconversation’,anddeclaresthat‘true’meansnothingmorethan‘whatyoucandefendagainstallcomers’,and‘rationality’nothingmorethan‘respectfortheopinionsofthosearoundone’(Rorty1979:377–8;1991a:32,37).Asapreliminarytounderstandinghowpragmatismcametobesostartlinglytransmuted,itwillbeusefultodistinguishtwobroadstyles:thereformistandtherevolutionary(adistinctionadaptedfromMigotti1988).Thesehaveincommontheaspirationtofreephilosophyofmetaphysicalexcessesand,morespecifically,oftheartificialconfinesoftheCARTESIANTHEORYOFKNOWLEDGE(chapter26).Relinquishingtherequirementofabsolutecertaintyinknowledge,moreinclinedtoNATURALISM(p.767)thantoapriorism,emphasizingthesocialaspectsofinquiry,pragmatismofareformiststripeacknowledgesthelegitimacyoftraditionalquestionsaboutthetruth-relatednessofourcognitivepractices,andsustainsaconceptionoftruthobjectiveenoughtogivethosequestionsbite.Pragmatismofarevolutionarystripe,bycontrast,relinquishingtheobjectivityanddenyingthevalueoftruth,canacknowledgenolegitimateepistemologicalquestionsnotstrictlyinternaltothecognitivepracticesofthisorthatcommunity.And,sinceitisatautologythatinquiryaimsatthetruth,theverypossibilityofbonafideinquiry,andafortioritheconceptionofphilosophyasakindofinquiry,istherebyundermined.Thatthisdistinctionofreformistversusrevolutionarydoesnotmapperfectlyontothelistofpragmatists,neo-pragmatistsandsympathizers,sofarfromdetractingfromitsusefulness,preciselyfitsittoserveasaframeworkforunderstandingthetransfor-mation–or,onemightbettersay,thevulgarization–ofpragmatism.ThediagnosiswillbethatthedelicatebalanceofobjectivityandfallibilismattemptedinPeirce’sreformistpragmatism,somewhatdisturbedbyJames’snominalistictendenciesandbyrevo-lutionary,orrevolutionary-sounding,HegelismsinDewey,isdestroyedwhenSchiller(mis)interpretsJamesasrelativizingtruth,andRorty(mis)interpretsDeweyasre-pudiatingthetheoryofknowledgeandasaspiring,nottothereconstruction,butthedeconstruction,ofphilosophy.2C.S.Peirce(1839–1914)AttheheartofPeirce’spragmatismliesthepragmaticmaxim:‘ifonecandefineaccu-ratelyalltheconceivableexperimentalphenomenawhichtheaffirmationordenialofaconceptcouldimply,onewillhaveacompletedefinitionoftheconcept,andthereisabsolutelynothingmoreinit’(Peirce5.412).Meaningisamatteroftheconceivableexperimental,experientialconsequencesofaconcept’sapplying;pragmatisch,intheKantiansense.Thepragmaticmeaningof‘xisF’isgivenbyalistofconditionals;as775\nSUSANHAACKthematurePeirceovercameyouthfulnominalistleanings,heinsistedonthesubjunc-tiveformulation:‘ifyouweretodoA1,experientialconsequenceE1wouldresult’,‘ifyouweretodoA2,experientialconsequenceE2wouldresult’,andsoon.Thus,adiamondreallyishardifitwouldscratchothersubstancesifitwererubbedagainstthem–evenifitneverisrubbed.AndPeircestressesthatthelistofmeaning-specifyingconditionalsisopen-ended;meaninggrows,asheputsit,asourknowledgegrows.Pragmaticismis,asPeirceacknowledges,akindofpositivism;oneroleoftheprag-maticmaximistorevealthat‘almosteverypropositionofontologicalmetaphysics...isgibberish’(Peirce5.423).ButthemaximisnotintendedtoruleoutMETAPHYSICS(chapter2)altogether,butrathertodiscriminatetheillegitimate,thepragmaticallymeaningless,from‘scientific’metaphysics,whichusesthemethodofscience,observa-tionandreasoning,andwhichisundertakenwiththescientificattitude,thatis,fromthedesiretofindouthowthingsreallyare–andnot,ashappenswhenphilosophyisinthehandsoftheologians,fromthedesiretomakeacaseforsomedoctrinewhichisalreadyimmovablybelieved.Scientificphilosophy,asPeirceconceivesit,isanobserva-tionalscience,differingfromtheothersciencesnotinitsmethodbutinitsrelianceonaspectsofexperiencesofamiliar,soubiquitous,thatthedifficultyistobecomedistinctlyawareofthem.(SoitwouldbeamisunderstandingtothinkofPeirce’saspirationtomakephilosophyscientificasinanywayscientistic;Peirceexpresslydeniesthatphilosophicalissuescouldberesolvedwithin,andcertainlyneversuggeststhatphilosophyoughttobereplacedby,thenaturalsciences.)ItisasscientificmetaphysicsthatPeircepresentshisownmetaphysicaltheories.Forexample,Peirceoffersatheoryofcategories–firstness,potentiality;secondness,particularexistence,reaction;andthirdness,generality,connection,law–basedbothphenomenologicallyand,likeKant’scategories,onlogic,butonPeirce’smoreadequatelogic,capable,unlikeAristotle’s,ofexpressingrelations.Amonghisotherimportantcontributionstometaphysicsarehistychism(thereischance,objectiveindeterminacy,aswellaslaw,intheuniverse);hissynechism(theubiquityofcontinuity);agapism(thegradualevolutionoftheworldfromchaostowardsorder,‘concretereasonableness’);andhisscholasticrealism(therealityof‘generals’,naturalkindsandlaws).ButPeirce’sconceptionofareformed,scientificphilosophyisperhapsmostdirectlyseeninhisthoroughgoingcritiqueofCartesianepistemology,andinthemorenaturalistictheoryofinquirywithwhichheproposestoreplaceit.WithDescartes,philosophy‘putoffchildishthings,andbegantobeaconceitedyoungman’(Peirce4.71).Descartes’smethod,accordingtoPeirce,isasham,amatteroffeigned,‘paper’doubtswhichinevitablyledtotheeventualreinstatementofthebeliefssupposedlydoubted.ThereisnosuchfacultyastheintuitiononwhichDescartes’scriterionofclearnessanddistinctnessrelies,andnosuchintuitiveself-consciousnessashisrelianceonthecogitoastheindubitablestarting-pointforthereconstructionofknowledgerequires.Descartes’saspirationtocertaintyismisplaced,hissubjectivecriterionoftruthviciouslyindividualistic.UnlikeDescartes,andinstrikinglyDarwinianspirit,Peirceseeshumanbeliefascontinuouswithanimalexpectation,humaninquiryascontinuouswithanimals’explorationsoftheirenvironment.FollowingAlexanderBain,Peirceconceivesofbeliefasahabitofaction,adispositiontobehave,andofdoubtastheunsettledstate776\nPRAGMATISMresultingfromtheinterruptionofabelief-habitbyrecalcitranceonthepartofexperi-ence.Realdoubt–unlikeDescartes’spaperdoubt–isthusinvoluntaryandunpleas-ant.Theprimitivebasisofthemostsophisticatedhumancognitiveactivity,scientificinquiry,isahomeostaticprocessbywhichtheorganismstrivestoreturntoequilibrium,aprocesshaltedwhenanewhabit,arevisedbelief,isreached.Thishomeostaticprocessisitsprimitivebasis,butitisnotyetscientificinquiry.Thebeliefsthus‘settled’arelikelyonlytemporarilyfixed;sothemostsophisticatedinquirers,realizingthat,unlessabeliefistrue,thoughitmaybetemporarilysettled,itcannotbepermanentlyso,andaspiringtoindefeasiblysettledbelief,willalwaysbemotivatedtofurtherinquiry,neverfullysatisfiedwithwhattheypresentlyinclinetothink.Theprimitivehomeostaticprocessisthustransmutedintoscientificinquiry.Amongpossiblemethodsforthe‘fixationofbelief’,theSCIENTIFICMETHOD(chapter9),Peirceholds,isdistinguishedbyitsappropriatenesstotheendofinquiry.Unliketheapriorimethodtraditionallyfavouredbymetaphysics,thescientificmethod,ifitweresufficientlypersistedin,wouldenableinquirytocometorestwithbeliefswhichareindefeasiblystable,permanentlysafefromrecalcitrance.Thescientificmethod,accord-ingtoPeirce,isthemethodofOBSERVATION(pp.295–7)andreasoning:moreprecisely,itisthemethodofaccommodatingtheperceptualjudgementsforcedupononebyexpe-rienceintoanexplanatoryframework,bymeansofreasoningofthreetypes:abduc-tion,thepostulationofhypothesestoexplainsomepuzzlingphenomenon;deductionofconsequencesfromsuchabductivehypotheses;andtheinductivetestingofsuchhypotheses.AsPeirceputsit,wewishourbeliefstoconformtohardfacts;anditisinthebrutecompulsivenessofexperiencethatthehardnessoffactconsists.EvenMATHEMATICS(chapter11),accordingtoPeirce,isanobservationalscience.Thoughmathematicaltruthsarenecessary,mathematicalknowledgedependsonexperience,butoninnerexperience,theconstruction,manipulationandobservationofimagined‘icons’ordiagrams.Thedesiretolearnthetruthrequiresanacknowledgementthatonedoesnotsatis-factorilyknowalready.Thescientificinquireristhereforea‘contritefallibilist’whoisreadyto‘dropthewholecartloadofhisbeliefs,themomentexperienceisagainstthem’(Peirce1.14,1.55).Inthecriticalcommon-sensismofPeirce’smaturephilosophy–anattemptedsynthesis,asthephrasesuggests,ofKANT’s(chapter32)andReid’sresponsestoHUME(chapter31)–thescientificinquirerisseenassubmittingtheinstinc-tivebeliefsofcommonsensetocriticism,refinementandrevision.Ascientificdoubt,asPeirceputsit,‘cannevergetcompletelysettorestuntil,atlast,theverytruthaboutthatquestiongetsestablished’(Peirce7.77).Itcouldwithsomejusticebeprotestedthatthecriticalcommon-sensist’spolicyofdeliberatecriticalscrutinyandattemptedfalsificationofthebeliefswithwhichheinitiallyfindshimselfbearsamorethanpassingresemblancetoDescartes’smethodofdoubt.AndinanycasePeirce’sobjectionstothecriticalphaseofDescartes’sprojectseemlargelytomisstheirmark.However,hiscritiqueofDescartes’sassumptionthatanyknowledgeworthyofthenamemustbeabsolutelycertain,ofthesubjectivityofhiscriterionoftruth,andofhisfirst-person,strictlyindividualisticapproach,ismoresustainable–andnotonlycompatible,butintimatelyconnected,withthekeythemesofPeirce’sowntheoryofinquiry.777\nSUSANHAACKThattheoryisfallibilisticand,inasense,thoroughlysocial:inthesensethatPeircethinksofthescientificinquirerasjustonecontributortoavastenterpriseextendingbothwithinandacrossgenerations.Scientificinquirymaybeundertakenbyindividu-als;buttheindividualinquirerwillmakehisworkfreelyavailabletoothersand,evenifhefails,hiswillbeoneofthecarcassesoverwhichfuturegenerationsofinquirersclimbastheyfinallystormthefortressofknowledge.Fallibleandimperfectasscientificinquiryis,ifitweretocontinuelongenough–Peirceisawarethatthereisnoguar-anteethatitwill–eventuallyafinal,indefeasible‘ultimateopinion’wouldbeagreed.WhatsupportsPeirce’sconfidencethat,ifinquirycontinuedlongenough,eventu-allyconsensuswouldbereached?Theideathatevolutionaryadaptationhasgivenhumanbeingsaninstinctforguessingrightwhichenablesthemtocomeupwithsuc-cessfulabductions,andthethesisthatinductiontendstobeself-corrective,bothplayobviousroleshere.Lessobvious,butnolessimportant,istheroleoftheunusualcom-binationofthedirectandtheinterpretativeinPeirce’stheoryofperception;and,lessobviousagain,theroleofthemetaphysicaldoctrinePeircesometimescalls‘extremescholasticrealism’.Peircedistinguishesthepercept,theperceptualeventorpresentation,fromtheper-ceptualjudgement,thebeliefpromptedbytheexperience.Theperceptualjudgement,hemaintains,isthoroughlyinterpretative–indeed,butforitscompulsivecharacter,itisanalogoustotheconclusionofanabduction–andhence,thoughincorrigible,thor-oughlyfallible.ButPeirce’sacknowledgementoftheinterpretativenessofthepercep-tualjudgementiscombinedwithaninsistencethattheperceptisnotaREPRESENTATION(p.732)butapresentation,andthatwedirectlyperceive,notimagesormentalrepresentations,butreal,externalobjects.Hallucinationsandillusionsmaybeindistinguishablephenomenologicallyfromgenuineperception;theymayshareitsqualityofinsistency,ofbruteness;theymayevenbeexperiencedbymorethanoneindividual;buttheyaredistinguishablebythefactthatthebehaviourofthereal,exter-nalobjectsofgenuineperceptionispredictableonthebasisofourknowledgeoftheLAWSOFNATURE(pp.305–6).Peirce’s‘extremescholasticrealism’isthethesisthattherearerealgenerals;thatis,thattherearenaturalkindsandlawswhichareindependentofhowwethinkthemtobe.(Peircedistinguisheshisscholasticrealismnotonlyfromnominalism,whichmain-tainsthatgeneralsarenotrealbutfigments,butalsofromPLATONISM(chapter23)–whichhedubs‘nominalisticplatonism’becauseitmaintainsthatgeneralsexist;exis-tencebeingforPeircethecharacteristicmodeofbeingofseconds,notthirds.AndhedistinguishesitfromScotisticrealism,itsclosestancestor,because,unlikeScotus,heholdsthatwhichgeneralsarereal,whichgeneraltermsrepresentrealkinds,isamatterforscientificinquirytodiscover.)ScholasticrealismisintendedbyPeirceasahypoth-esisofscientificmetaphysics.Iftherewerenorealgenerals,heargues,scientificinquirywouldbeimpossible;fortheverypossibilityofPREDICTION(pp.289–90),inductionandexplanationdependsontherealityofkindsandlaws.Thereisapatternofgenerals,ofnaturalkindsandlaws,underlyingtheparticularobjectsandeventsweperceive,whichis‘independentofhowyouorIoranynumberofmenthink’.Andsoitcanbeexpectedthatthe‘arbitrary,accidentalelement’ininquiryintroducedbythepeculiarcircum-stancesandidiosyncrasiesofindividualinquirerswouldeventuallybeovercomeasinquiryproceeds,andconsensuseventuallyreached(Peirce8.13).778\nPRAGMATISMPeircecontrastsscholasticrealismnotonlywithnominalismandnominalisticplatonism,butalsowithwhatonemightcall‘noumenalism’,theideathatthereallyrealisinprincipleinaccessibletohumancognition.Thisisacharacteristicallyprag-matistattitude,forthepragmaticmaximdisqualifiesasmeaninglessanyhypothesisdevoidofallpossibleexperientialconsequences,anyquestionwhichwouldresistsettlementnomatterhowlongscientificinquiryweretocontinue.Thethesisthatthegoalofinquiryispermanentlysettledbelief,andthethesisthatthescientificattitudeisadisinteresteddesirefortruth,areunitedbyPeirce’sdefinitionof‘true’.Hedoesnotthinkitfalsetosaythattruthiscorrespondencetoreality,butshallow–amerelynominaldefinition,givingnoinsightintotheconcept.Hispragmaticdefinitionidentifiesthetruthwiththehypotheticalidealtheorywhichwouldbethefinalupshotofscientificinquirywereittocontinueindefinitely.‘Truthisthatconcor-danceof...[a]statementwiththeideallimittowardswhichendlessinvestigationwouldtendtobringscientificbeliefs’;‘anytruthmoreperfectthanthisdestinedcon-clusion,anyrealitymoreabsolutethanwhatisthoughtinit,isafictionofmetaphysics’(Peirce5.565,8.13).Thesequotationsrevealsomethingbothofthesubtlety,andofthepotentialfortension,withinPeirce’sphilosophy.Hisaccountofrealityaimsatadelicatecompromisebetweentheundesirableextremesoftranscendentalismandidealism,hisaccountoftruthatadelicatecompromisebetweenthetwindesiderataofobjectivityand(in-principle)accessibility.3WilliamJames(1842–1910)‘Therecanbenodifferenceanywherethatdoesn’tmakeadifferenceelsewhere’(James1907:30):hisversionofthepragmaticmaximliesattheheartofJames’spragmatismtoo.UnlikePeirce,however,JamesthoughtthatphilosophywoulddowelltogoroundKant,ratherthanthroughhim;andthisisreflectedinhisinterpretationofthemaxim,whichstressespraxis,thepracticalconsequencesofaconcept’sapplying.ThepointisnotthatPeircedoesnot,likeJames,holdthatmeaningispurposive,thatthemeaningofaconceptliesinitsapplication;itisratherthatthemature,realistPeirce,unlikeJames,regardsthiskindofformulationasunfortunateifitencourages‘subordinat[ing]theconceptiontotheact,knowingtodoing’(Peirce1900:332).ThisdifferenceofemphasisismagnifiedbyJames’sreadinesstoconstrue‘theconsequencesofabelief’inawaythatincludesnotonlytheconsequencesofthetruthofthepropositionbelieved,butalsotheconsequencesoftheperson’sbelievingit.Itisnotsurprising,inviewofthis,thatreadersofJames,andevenJameshimself,sometimeshavedifficultydisentanglinghispragmatismfromhisdoctrineoftheWilltoBelieve,his‘defenceofourrighttoadoptabelievingattitudeinreligiousmatters,inspiteofthefactthatourmerelylogicalintellectmaynothavebeencoerced’(James1897:1–2).Whenahypothesisbyitsnaturecannotbedecidedonthebasisofevidence;whenitisliveforus,thatis,itappealstousasarealpossibility;whenthechoicebetweenbelievingitanddisbelievingitisforced,thatis,unavoidable,andmomentousforourlives–then‘ourpassionalnature...lawfullymay...decide’(ibid.:11).RELIGIOUSBELIEFS(chapter15)whichinprinciplecannotbeverifiedorfalsifiedmaybelegitimatedbytheirsalutaryeffectonthebeliever’slife.779\nSUSANHAACKPeirce–towhomTheWilltoBelieveisdedicated–thefollowingyearisfounddescrib-ingthescientificattitudeasthe‘WilltoLearn’,andobservingthat,forhimself,he‘wouldnotadoptahypothesis,andwouldnoteventakeitonprobation,simplybecausetheideawaspleasing’tohim;thattodosowouldbe‘acrimeagainsttheintegrityofthereasonthatGodhaslent’him(Peirce5.583,5.598).Heputsit,onemightthink,ifanythingtoomildly;forbythestandardsofthepragmaticmaximasheinterpretsit,anyhypothesiswhichisinprincipleincapableofsettlementbyobservationandrea-soning,wouldbenogenuinehypothesisatall.AlsoassociatedwithJames’spragmatismishismature‘radicalempiricism’,bymeansofwhichhehopestoescapethedualismofCartesianmetaphysicsandthesub-jectivityofCartesianepistemology.WhatJamescalls‘pureexperience’is,heholds,ofitselfneithermentalnorphysical;theseparationintoconsciousnessandcontent,knowerandOBJECT(p.794)known,isamatterofrelationsamongdifferentbitsof‘pureexperience’.(Howcanoneandthesamethingbebothinspaceandinthemind?Jamesanswers,asoneandthesamepointmaybeontwolines,ifsituatedattheirintersec-tion.)Theknowingrelation,Jamessuggests,is‘successfulleading’;onecountsasknowingathingifone’sideaofitenablesonetoactsuccessfullytowardsit.Onemightreasonablyfeelthatthe‘cashvalue’ofradicalempiricismscarcelymatchesthatofPeirce’sreconciliationofthedirectandtheinterpretativeelementsinperception,andthatJames’stalkof‘leading’isatbestapromissorynote.Itsimpor-tanceforthepresentargumentlies,however,initsclearconnectionwithJames’saccountofthetrueas‘thesatisfactoryinthewayofbelief’.Radicalempiricismismonistic;itwas,indeed,influentialinthedevelopmentofRUSSELL’s(chapter37)neutralmonism.Nevertheless,itiscompatiblewiththeideasofJames’sHibbertlectures,entitled‘APluralisticUniverse’.Forradicalempiricism,asJamesputsit,‘laystheexplanatorystressuponthepart,theelement,theindividual’;itisa‘mosaic’philosophy(James1912:24).Thereisonlyonekindofstuff,butmanyportionsorindividualbitsofit.ThisrevealsJames’smarkedpredilectionfortheparticular,theconcrete.LikePeirce,Jamesthinksitnotexactlyfalsebutinadequatetosaythattruthiscor-respondencewithreality.AgainmuchlikePeirce,hecharacterizes‘truthabsolute’as‘anidealsetofformulationstowardswhichallopinionsmaybeexpectedinthelongruntoconverge’(James1909:143).Thedifferencebetweentruebeliefsandfalseonesisthattheyareverifiable,theyleadussuccessfully.Thetrue,Jamessays,isthesatis-factory,theuseful,theexpedientinthewayofbelief.Truebeliefswork.CriticslikeMooreandRussellwerescandalizedbywhattheysawasacrassidentificationoftruthandutility–acriticismJamesdescribesasa‘slander’(ibid.:147);for,thoughindeedhehadwrittenthat‘thetrue...isonlytheexpedientinthewayofbelief’,hehadgoneontoexplain,‘expedientinthelongrunandonthewholeofcourse;forwhatmeetsexpedientlyalltheexperienceinsightwon’tnecessarilymeetallfurtherexperienceequallysatisfactorily.Experience...haswaysofboilingover,andmakinguscorrectourpresentformulas’(James1907:106).Distinguishinghis‘radical’empiricismfromearliervariants,Jamesdisassociateshimselffromthedisconnectednessofearlierempiricisms;and,specifically,fromthethesesthatsimilarshavenothingreallyincommon,andthatthecausaltieisnothingbuthabitualconjunction.But,unlikePeirce,hedoesnotappealtotherealityofkinds780\nPRAGMATISMandlawstoexplainwhy,inthelongrunofexperience,opinionscanbeexpectedtoconverge.Rather,hispredilectionfortheparticularleadshimtoapreoccupationwithspecifictruthsratherthantruthassuch.Hemanifestssomediscomfortwiththenotionofverifiability,preferringtoconcentrateonparticulartruthsactuallyverified.Thisleadstoapparentinconsistencies,asJamesseemssometimestoallowthatbeliefswhichareverifiedaretherebyshowntohavebeentrueallalong,sometimestosuggestthatbeliefsbecometruewhentheyareverified.ConsistencycanberestoredbymeansofadistinctionJamessometimesmakes,butdoesnotalwaysmark,between‘abstract’or‘absolute’Truthand‘concrete’or‘relative’truths,identifyingabstractTruthwiththeverifiable,andconcretetruthswiththeverified.Ofcourse,whatJamescalls‘concretetruths’arenotreallytruthsatall;hisdistinctionisreallybetweenabelief’sbeingtrueanditsbeingshowntobetrue–or,moreaccurately,sinceheallowsthatwhatisatonetime‘verified’maylaterturnouttobefalse,betweenabelief’sbeingtrueanditsbeingconfirmed.James’sstressontheparticular–whichthematurePeircewouldnodoubthavefeltbetraysnominalistleanings–manifestsitselfinhistendencytodownplayabstractTruthandtoemphasizeconcretetruths.Andhisuseof‘concretetruth’for‘propositionnowconfirmed’encouragestheideathattruth(andnotmerely‘concretetruths’)ismadebyus,andchangesovertime.4JohnDewey(1859–1952)DeweydescribesoneofPeirce’sbest-knowncharacterizations–‘[t]heopinionwhichisfatedtobeagreedbyallwhoinvestigate’(Peirce5.407)–as‘thebestdefinitionoftruth’(Dewey1938:345n.).LikeJames,however,hetendstostresstheconcreteovertheabstract,actualverificationoverpotentialverifiability,truthsoverTruth.Inmuchofwhathewritesonehearsacertainambivalenceaboutjusthowradicalhemeanstobe.Theintellectualist’shorrorofacknowledgingthatverificationis,asetymologysuggests,amatterofmakingabelieftrue,hedeclares,is‘largelysentimental’;buthegoesontoexplainthattomakeanideatrue‘istomodifyandtransformit’(thatis,presumably,tochangethepropositionconcerned)untilitisabletosurvivetesting(Dewey1910:139).Abeliefthatisverified,hesays,istherebyshowntohavebeentrueallalong,buthegoesontoexplainthatthismeansonlythatitwasgoingtobeverified(ibid.:142–4).Theprag-matistdoesnotdeny,Deweyinsists,thattruthiscorrespondenceofthoughttoexistence,thoughhedoesdenythatitiscorrespondenceofthoughttounknowablethings-in-themselves;buthegoesontoexplainthatthiscorrespondencemeansco-respondence,amatterof‘interadjustment’ofourideastoproblematicsituations(ibid.:158–9).Notsurprisingly,perhaps–giventhatconcretetruthisnotreallytruthatall–Deweyisalsodrawntotheideathatitmightbeaswelltostopusingtheword‘true’,andtoworkinsteadwiththeconceptofwarrantedassertibility.LikePeirceandJames,Deweyrepudiatesthe‘questforcertainty’.Unlikethem,hegoesontoofferapsycho-sociologicaldiagnosisofthemotivationforthatquest:itarose,hesuggests,fromthesharpdichotomyoftheoryversuspractice,andthedistasteforthepractical,thechangeable,theuncertain,embodiedintheslave-owningcultureofancientGreece.ThemostHEGELIAN(chapter33)ofthepragmatists,Deweyissuspiciousoftraditionalphilosophicaldualisms;andthisisreflectedinhisepistemological781\nSUSANHAACKwritings,whicharecriticalofthewholetraditionfromPlatothroughDescartestohisowncontemporaries,becauseofitsrelianceonthedichotomiesofobject–subject,fact–value,mind–body,theory–practice.‘Specialtheoriesofknowledgedifferenor-mouslyfromoneanother.Theirquarrels...filltheair.Thedinthuscreatedmakesusdeaftothewayinwhichtheyallsayonethingincommon...Theyallholdthattheoperationofinquiryexcludesanyelementofpracticalactivitythatentersintotheconstructionoftheobjectknown’(Dewey1929:22).Dewey,bycontrast,insiststhatknowingisnotisolatedfrompractice,butisitselfakindofpractice–tobejudged,likeotherpractices,byitspurposivesuccessratherthanbysomesupposedstandardofaccuracyofreflectionofitsobjects.Theobjectofknowl-edgeisnotanimmutable,independentreality,butischanged,andisinpartconstitutedby,ourcognitiveinteractionswithit;forinquirytransformsaproblematic,indetermi-natesituationintoadeterminateone.LikePeirce,Deweyconceivesoflogicbroadly.Itisthetheoryofinquiry,ofwhichFORMALDEDUCTIVELOGIC(chapter4)–logicintheusualmodernsense–isonlyapart.Peircewasanimportantcontributortothedevelopmentofmodernformallogic,discov-eringquantifiersandthelogicofrelationsindependentlyofFREGE(chapter37),antici-patingPrincipiaMathematica’sdefinitionofnumberbymorethanthirtyyears,anddevisingtruth-tablesandventuringintothree-valuedlogicmorethanadecadebeforePostandL¢ukasiewicz.Dewey’slogicalworkisnotformal,butthisisnotwhyPeirceiscriti-calofit;Peirce’sobjectionisthatitistoopurelydescriptive,amountingonlytoa‘naturalhistory’ofinquiry,notanormativemethodology.ButbythetimeofLogic,TheTheoryofInquiry,atanyrate,Deweyisclearlyacknowledgingthatlogicisnormative–stressing,however,thatitsrecommendationsabouthowinquiryshouldproceedcanonlybebasedonscientificstudyofwhatprocedureshaveproventhemselvestowork.ThecumulativeeffectoftheshiftsofemphasisfromPeircethroughJamestoDeweyis,itmightbefairtosay,fromanessentiallyreformistattitudetothephilosophicaltradi-tiontoapotentiallyrevolutionaryone.Peirce’stheoryofinquiryisbehaviouristic,nat-uralisticandfallibilist;itisthoroughlyanti-Cartesian,butclearlyreformistinitsapproach.SotooisJames’s,buthisstressontheso-far-satisfactoryinthewayofbelief,especiallygiventhemisleadingterminologyof‘concretetruths’,hintsatsomethingmorerevolutionary.Deweyisperhapsleastmisleadinglydescribedashalf-straddlingthereformist–revolutionarydistinction.Forsidebysidewithreformistelements–hisdis-tinctionofstateandcontentsensesof‘belief’,hisinsistenceonaconceptionofexperi-encemuchricherthantheoldsensationalistone,and,especially,hisaspirationtotranscendtheolddichotomyofrationalismversusempiricismandtoacknowledgetheinterlockingofexperienceandreason–arerevolutionary,oratleastrevolutionary-sounding,suggestionsaboutthemakingoftruth,themodificationoftheobjectsofknowledgethroughinquiry,andthereplacementofthetraditionaltheoryofknowledgebyacriticalscrutiny,internaltoscience,ofthemethodsofinquiryscienceemploys.5F.C.S.Schiller(1864–1937)Schillerisuncompromisinglyrevolutionary.Inhiswritingsonefindsastraightforwardidentificationoftruthwithverification,andanunequivocalcommitmenttothe782\nPRAGMATISMmaking,andthemutability,oftruth.Jamesacknowledgesthathisaccountofconcretetruthscouldnotstandalone:‘toadmit,aswepragmatistsdo,thatweareliabletocorrection...involvestheuseonourpartofanidealstandard’(James1909:142).ButSchilleroffersatheoryofconcretetruthsasacompletetheoryoftruth.Hedeniesoutrightthattruthiscorrespondencewithreality,anideawhichhedescribesasnotonlyworthlessasacriterionbutabsurdinitself.Truthispracticalworking.‘True’,hesays,means‘valuedbyus’.Apropositionistrueifit‘forwardsourends’.Truthismutable,sincepropositionsbecometrueonlywhensuccessfullyapplied;‘atruthwhichwillnot...submittoverification,isnotyetatruthatall’(Schiller1907:8).Realityisalsomutable,growingastruthgrows.Truthisdependentonus,relativetoourpurposes.Andsoisreality;factsarenotdiscoveredbutselected,evenmade,byus.Inaspirit,nodoubt,ofdeliberateprovocation,SchillerlikenshisviewstothoseofPROTAGORAS(chapter22).Certainlyhisrevolutionary,relativistichumanismcouldscarcelybefurtherremovedfromPeirce’srealisticpragmaticism.‘TheapostleofHumanism’,PeircewroteofSchillerin1905,‘saysthatprofessionalphilosophists“haverenderedphilosophylikeuntothemselves,abstruse,arid,abstractandabhorrent.”ButIconceivethatsomebranchesofsciencesarenotinahealthystateiftheyarenotabstruse,aridandabstract’;andin1908:‘[i]tseemstomeapitythat[MrSchillerandthepragmatistsoftoday]shouldallowaphilosophysoinstinctwithlifetobecomeinfectedwiththeseedsofdeathinsuchnotionsas...themutabilityoftruth’(Peirce5.537,6.485).Theseobservationsarenotonlypoignantbutprophetic.Peirce’swarningsofthedangersofSchiller’shumanisticpragmatismbringtomindRussell’sdismissalofprag-matismasan‘engineers’philosophy’,boundtoleadtocosmicimpiety,oratleasttofascism;Schiller’sdisdainforthemummifiedpedantryofthephilosophicaltraditionbringstomindRorty’sunkindwordsaboutthelogic-choppingofanalyticphilosophy.6RecentandContemporaryPragmatismsRorty,mostradicalofcontemporaryself-styledneo-pragmatists,andtheclosesttoSchiller,uses‘pragmatist’asthecontrastto‘realist’.Hetellsusthattruthisnotthekindofthingoneshouldexpecttohaveaphilosophicallyinterestingtheoryabout,thattocallapropositiontrueisjusttogiveitarhetoricalpatontheback(Rorty1982:xxi,xiii).Hetransmutesinquiryinto‘carryingontheconversation’ofWesternculture(Rorty1979:377–8)andexperientialfalsificationinto‘conversationalobjections’(Rorty1982:165–6);andreferstothosewhodescribethemselvesasseekingthetruthas‘old-fashionedprigs’(Rorty1991b:86).TheaffinityofRorty’sandSchiller’sconceptionsofpragmatism,itshouldbesaid,isnottheresultofSchiller’sinfluenceonRorty;perhapsitis,inpart,theresultratherofRorty’s(mis)readingJamesmuchasSchillerdid.Neither,itshouldalsobesaid,isRortyprepared,asSchilleris,toadmittorelativism.Indeed,itisnotquiteunambiguouslyclearthatRortyisproperlyconstruedasarelativist;butonlybecauseheseemstoshift,asoccasiondemands,betweenacon-textualist,andhencerelativist,conceptionofepistemicjustification(Aisjustified-relative-to-community-Cinbelievingthatpjustincasehecandefendthebelief783\nSUSANHAACKthatpagainsttheobjectionsofmembersofC),andatribalistone(Aisjustifiedinbelievingthatpjustincasehecandefendthebeliefthatpagainsttheobjectionsofourcommunity).But,eitherway,Rorty’spositionisthoroughlyanti-epistemological;forheisunambiguouslycommittedtothethesisthatcriteriaofevidencearenotobjectivelygroundedintheirrelationtothetruth,butentirelyamatterofconventional,conversationalconstraints.Indeed,asRortycheerfullyacknowledges,thisisnotonlythoroughlyanti-epistemological,butanti-philosophical,underminingthetradi-tionalconceptionofphilosophyasaformofinquiry.Still,hethinksthereisafuturefortheex-philosopher,butitistobe‘hermeneutic’ratherthanepistemological,‘edifying’ratherthansystematic,literaryratherthanscientific,amatterof‘carryingontheconversationofWesternculture’ratherthanofinquiringintohowthingsare.RortyappealstoDewey’srepudiationoftheepistemologicaltradition.Buttheappealisstrained,forDeweylookstothe‘naturalizationofintelligence’,toascientificapproachtoreplacethe‘spectatortheoryofknowledge’(Dewey1929:195,196),whileRortydeclaresthatepistemologyneedsnosuccessor-subject,and,farfromwelcomingashifttowardsscience,anticipatesthefutureofpost-philosophyasagenreofliteratureorofliterarycriticism.True,Jameswritesoftheaffinityofsomemoralphilosophywith‘novelsanddramasofthedeepersort’(James1897:210).Thisisbestconstrued,however,notasdenyingthatmoralphilosophyisaformofinquiry,butassuggestingtherolethatimaginativeempathymightplayinsuchinquiry;and,soconstrued,neednotconflictwithDewey’svisionofanormativeethicsinwhichscientificinquiryintothenatureoftruehumanflourishingwouldplayanimportantrole.Peircewrites,withcharacteristicverve,oftheneed‘torescuethegoodshipPhilosophyfortheserviceofSciencefromthehandsofthelawlessroversoftheseaofliterature’;‘[a]sforthatphrase,“studyinginaliteraryspirit”’,heremarks,‘itisimpos-sibletoexpresshownauseatingitistoanyscientificman’(Peirce5.449,1.33).Perhapsitisnowonder,then,thatRortyreferssodismissivelytoPeirce’s‘undeservedapotheo-sis’,andtellsusthatPeirce’sonlycontributiontopragmatismwastogiveitaname(Rorty1982:160–1).ButitwouldbeamistaketoreadPeirce’sharshremarksaboutliteraryphilosophyasscientistic,orashostiletoliterature;theyexpress,rather,hisinsistencethatphilosophy,likescience,isaformofinquiry,oftruth-seeking,andthat,inthispursuit,testabilityisparamount,andclarityandprecisionoflanguagemusttakepriorityovereuphonyoreleganceofstyle.Thispointisdoublyimportantbecausetherevolutionarywingofcontemporarypragmatismalsoshelters,besidesRortyandotheradmirersofliteraryphilosophy,othersofanotablyscientisticstripe.PaulChurchlandargues,fromthenon-propositionalworkingsofthegangliaofthesea-slugandthesuccessesofconnection-istartificialintelligenceintrainingupacomputertodistinguishaminefromarockontheoceanfloor,thatcognitiveprocessingisnotpropositional,andtruthnotthegoalofinquiry.Anothercontemporaryproponentofvulgarpragmatism,Stich,maintainsthatthegoalofinquiryis,nottruth,butbeliefsthatconducetowhateveritisthesubjectvalues;ironicallyenough,herestshiscaseinpartonadenialthattruthisinstrumen-tallyvaluable–andcites,asjustificationforcallingthis‘pragmatism’,apassageinwhichJamesisarguingtheutilityoftruebeliefs.784\nPRAGMATISMOthercontemporarieshavesoughthabitablegroundbetweenthereformistandtherevolutionarywingsofthepragmatisttradition.Animportantthemehereisthehopeofsomekindofrapprochementofthe‘analytic’andCONTINENTAL(chapters41and42)camps.Thehopeisreasonableenough:thepragmatisttradition,afterall,antedatesthesplit;bothPeirceandJamesmakesignificantcontributionstophenomenology(or,asPeircesometimescallsit,‘phaneroscopy’);andDewey’sstressonthehistorical,thesocial,andonpraxishaveledsometoseehimasAmerica’s(thoroughlydemocratic)answertoMarx.Margolis,forexample,believesitpossibletodevelopa‘pragmatismwithoutfoundations’inwhichtheplausibleelementsofrealismcanbereconciledwiththeplausibleelementsofrelativism.Bernsteinwants,instead,tomovebeyondthesup-posedoppositionofrealismandrelativismbyovercomingwhathecallsthe‘CartesianAnxiety’,thequestforfixedconstraints.Pragmatistsofaplainerreformiststripewouldrespond–perhapsafterprotestingtheimplicationthattheirphilosophyisdrivenbytheneuroticneedsuggestedbytheterm‘anxiety’–thatitisnotnecessarytomovebeyondobjectivity,ortoreconcileitwithrelativism,provideditistemperedbyasufficientlythoroughgoingfallibilism.Anditistomymindwithinthetraditionofreformistpragmatismthatthemostenduringcontributionsaretobefound:Mead’stheoryofthesocialconstructionoftheself,inspiredbyPeirce’scritiqueoftheintuitiveself-consciousnessassumedbyDescartes;Lewis’spragmaticaccountoftheapriori,itselfaninspirationforQuine’scallfor‘amorethoroughpragmatism’;Ramsey’sBEHAVIOURISTIC(pp.65–7)approachtobelief,andQuine’s,also,tomeaning(QuinecitesDewey:‘meaningis...apropertyofbehavior’);Quine’sassociationofnaturalkinds,inductionandevolutionaryepistemology;Reichenbach’s‘pragmaticvindication’ofinduction;Hanson’sdefenceoftheideaofanabductivelogicofscientificdiscovery,Sellar’sappealtothenotionofexplanatorycoher-ence,andHarman’stoinferencetothebestexplanation;Putnam’sexplorationsofcon-ceptionsoftruthintermediatebetweenmetaphysicalrealismandrelativism;Apel’softheneedforanaccountofthepragmaticdimensionstosupplementTarski’ssemantictheoryoftruth;Rescher’sinvestigationsofcriteriaofsuccessandimprovementofcog-nitivemethods,Jardine’sofscientificprogress;andmanymore.Inmyownwork,suchkeyideasasexplanatoryintegration(centraltothearticulatedquasi-holismofmyaccountofevidentialsupport),thedistinctionofthestateandthecontentsensesof‘belief’,andthecombinationofdirectandinterpretativeelementsinperception(bothcentralelementsofmyaccountoftheroleofexperienceinjustification),weremined,asIcantestify,fromthesamevein.Thislistillustratestheformidablediversityaswellastherichnessofthereformisttraditioninpragmatism.Thisdiversityisunifiedbywhatonemightcalltheongoingprojectofreformistpragmatism:theaspirationtofindamiddlegroundbetweendog-matismandscepticism;aconceptionoftruthasaccessibleenoughtoberealisticallyaspiredto,yetobjectiveenoughtobeworthyofthename;anarticulationoftheinter-playbetweentheworld’scontributiontoknowledge,andours.Thisistheessentialspiritofreformistpragmatism,succinctlysummedupbyJames:‘Pleaseobserve...thatwhen...wegiveupthedoctrineofobjectivecertitude,wedonottherebygiveupthequestorhopeoftruthitself’(James1897:17).Soconceived,thetraditionofreformistpragmatismstillflourishes;and,thoughveryfarasyetfromthe‘catholicconsent’Peircesaw(Peirce8.13)astheendofinquiry,itis,indeed,‘instinctwithlife’.785\nSUSANHAACKFurtherReadingDeweywriteson‘TheDevelopmentofAmericanPragmatism’inhisPhilosophyandCivilization,NewYork:Minton,Balch,1931.UsefulselectionsofthewritingsoftheclassicalpragmatistsareH.S.Thayer,ed.,Pragmatism:TheClassicWritings(detailsunderLewisinReferences),andA.Rorty,ed.,PragmaticPhilosophy,GardenCity,NJ:DoubledayAnchor,1966(theformerincludesaninterviewwithWilliamJamespublishedintheNewYorkTimesin1907,entitled‘Pragmatism–WhatItIs’;thelatterincludesexcerptsfromcriticismsofpragmatismbycontemporariessuchasRussellandMoore,andReichenbach’s‘pragmaticvindicationofinduction’).Forabibliogra-phyofprimarymaterials,seePragmatism:AnAnnotatedBibliography1898–1940,ed.JohnR.Shook,Amsterdam:Rodopi,1998.ThayerhasafinearticleonpragmatismintheEncyclopediaofPhilosophy,ed.P.Edwards,NewYork:MacMillan,1967(updatedbyHaackinthenewedition,publishedin1996);seealsoThayer’sMeaningandAction:ACriticalHistoryofPragmatism,Indianapolis,IN:Bobbs-Merrill,1968.I.Scheffler,FourPragmatists,London:RoutledgeandKeganPaulandNewYork:HumanitiesPress,1974,isaclearintroductiontoPeirce,James,MeadandDewey.A.J.Ayer,TheOriginsofPragmatism,SanFrancisco:Freeman,Cooper,1968,isillu-minatingontherelationofpragmatismandAyer’sown,logicalpositivist,approach.TheTransactionsoftheCharlesS.PeirceSocietyisaquarterlyjournalofAmericanphilosophy,whichcarriesarticlesoneveryaspectofpragmatismaswellasreviewsofsignificantbooksinthearea.T.Goudge,TheThoughtofC.S.Peirce,Toronto:TorontoUniversityPress,1950,andNewYork:Dover,1966,andC.Hookway,Peirce,London:RoutledgeandKeganPaul,1985,areusefulintro-ductionstoPeirce’swork.M.Murphey,TheDevelopmentofPeirce’sPhilosophy,Cambridge,MA:HarvardUniversityPress,1961,andIndianapolis,IN:Hackett,1993,hasbecomeaclassic.TwocollectionsofpapersthathavestoodthetestoftimeareStudiesinthePhilosophyofCharlesSandersPeirce,firstseries,ed.P.P.WienerandF.Young,andsecondseries,ed.E.C.MooreandR.S.Robin,Amherst:UniversityofMassachusettsPress,1952and1964.OnPeirce’slogic,Putnam,‘PeircetheLogician’,HistoriaMathematica,9(1982)290–301,isagoodstarting-point.SeeK.L.Ketner,AComprehensiveBibliographyofthePublishedWorksofCharlesSandersPeircewithaBibliographyofSecondarySources,BowlingGreen,OH:PhilosophyDocumentationCenter,1986,forabibliographyofworkonPeirceuptothatdate.Anew,chronologicaleditionof(selected)worksofPeirce,WritingsofCharlesS.Peirce,isbeingpublishedbyIndianaUniversityPress,undertheeditorshipofMaxFischetal.,thefirstfivevolumes(upto1886)beingavailableatthetimeofthiswriting.AlsowithIndianaUniversityPressaretwovolumesofTheEssentialPeirce(papersfrom1867–93,publishedin1992;andpapersfrom1893–1913,publishedin1998).James’sWorksarepublishedbyHarvardUniversityPressundertheeditorshipofF.BurkhardtandF.Bowers.WilliamJames:SelectedWritings,ed.G.Bird,Everyman,London:Dent,Dutton,andVermont:CharlesE.Tuttle,iscompactandhandy.RalphBartonPerry,ed.,TheThoughtandCharacterofWilliamJames,2vols,Boston:Little,Brown,1936,and(abridged,onevolume)Cambridge,MA:HarvardUniversityPress,1948,isunparalleledasanintellectualportraitofthemanandhiswork.SeealsoG.Myers,WilliamJames,HisLifeandThought,NewHaven,CT:YaleUniversityPress,1986,andhisbibliography;W.R.Corti,ed.,ThePhilosophyofWilliamJames,Hamburg:FelixMeiner,1976;E.Fontanell,God,SelfandImmortality:AJamesianInvestigation,Philadelphia,PA:TempleUniversityPress,1986;T.L.S.Sprigge,JamesandBradley:AmericanTruthandBritishReality,LaSalle,IL:OpenCourt,1993.S.Hook,JohnDewey:AnIntellectualPortrait,NewYork:JohnDay,1939,isafinebriefintro-duction.J.Tiles,Dewey,NewYork:RoutledgeandKeganPaul,1988,willalsobefounduseful.JohnDewey:TheManandhisPhilosophy,Cambridge,MA:HarvardUniversityPress,1930,includesanessaybyG.H.Mead.ThereisavolumeintheLibraryofLivingPhilosophersdevotedtoDewey’swork(ThePhilosophyofJohnDewey,ed.P.A.Schilpp,Evanston,IL:NorthwesternUniversityPress,1939);thisincludesabibliographyofDewey’swritings.R.W.Sleeper,The786\nPRAGMATISMNecessityofPragmatism,NewHaven,CT:YaleUniversityPress,1986,astudyofDewey’slogic,alsoincludescriticalbibliographies.SeeJ.BoydstonandK.Poolos,eds,AGuidetotheWorksofJohnDewey,Carbondale,IL:SouthernIllinoisUniversityPress,1970,andChecklistofWritingsAboutJohnDewey,1887–1977,Carbondale,IL:SouthernIllinoisUniversityPress,1978,forfurtherreferences.Dewey’scompleteworksarepublishedin17volumesbySouthernIllinoisUniversityPressundertheeditorshipofJoAnnBoydston.TwovolumesofTheEssentialDewey,thefirstonpragmatism,educationanddemocracy,thesecondonethics,logicandpsychology,editedbyL.HickmanandT.Alexander,werepublishedbyIndianaUniversityPressin1998.OnSchiller,seeR.Abel,HumanisticPragmatism,NewYork:FreePress,andLondon:Collier-MacMillan,1966,andThePragmaticHumanismofF.C.S.Schiller,NewYork:King’sCrossPress,ColumbiaUniversityPress,1955.OnMead,seeA.J.Reck,ed.,SelectedWritings,Indianapolis,IN:Bobbs-Merrill,1964,andW.R.Corti,ed.,ThePhilosophyofGeorgeHerbertMead,Winterthur,Switzerland:AmriswilerBucherei,1973,bothofwhichincludebibliographiesofMead’swritings.ThereisavolumeintheLibraryofLivingPhilosophersontheworkofC.I.Lewis(ThePhilosophyofC.I.Lewis,ed.P.A.Schilpp,LaSalle,IL:OpenCourt,1968);thisincludesabiblio-graphyofLewis’swritings.SeealsoS.Rosenthal,ThePragmaticAPriori,StLouis,MO:WarrenH.Green,1976.NicholasRescher’sthree-volumeASystemofPragmaticIdealismwaspublishedbyPrincetonUniversityPress(Princeton,NJ:1992–4).HilaryPutnam’sPragmatism:AnOpenQuestionwaspublishedbyBlackwell(Oxford:1995).J.J.McDermott,R.W.SleeperandA.Edel,inasymposiumpublishedinTransactionsoftheC.S.PeirceSociety,21,1(1985),1–38,arguethatRortymisrepresentsDewey.Haack,‘Philosophy/philosophy,anUntenableDualism’,TransactionsoftheC.S.PeirceSociety,29,3(1993),411–26,arguesthatRortymisrepresentsPeirce.Haack’s‘“WePragmatists...”;PeirceandRortyinConversation’,PartisanReview,1/1997,91–107,isadialoguecompiledfromPeirce’sandRorty’sownwords;her‘Asforthatphrase“studyinginaliteraryspirit”...’,ProceedingsoftheAmericanPhilosophicalAssociation,70,2,November1996,57–75,contrastsPeirce’sandRorty’saccountsoftherelationofphilosophytoscienceandtoliterature(botharereprintedinHaack,ManifestoofaPassionateModerate:UnfashionableEssays,Chicago:UniversityofChicagoPress,1998).H.O.Mounce,TheTwoPragmatisms:FromPeircetoRorty,London:Routledge,1997,presentspragmatismassharplydividedintotwowingseversincePeirceandJames.Somerecentcollectionsreflectthegrowinginfluenceofrevolutionarypragmatism.R.Goodman,Pragmatism:AContemporaryReader,NewYork:Routledge,1995,includespapersonpragmatismandliterature;L.Menand,Pragmatism:AReader,NewYork:VintageBooks,1997,presentingthepragmatisttraditionasculminatinginRorty,includespapersbycontemporaryliterary,historicalandlegalwriterssympathetictorevolutionaryneo-pragmatism(butseealsoHaack’scriticalnotice,‘VulgarRortyism’,TheNewCriterion,November1997,67–70).ReferencesApel,K.-O.1983:C.S.PeirceandPost-TarskianTruth.InTheRelevanceofCharlesPeirce,ed.E.Freeman.LaSalle,IL:HegelerInstitute,MonistLibraryofPhilosophy.Bernstein,R.J.1983:BeyondObjectivityandRelativism.Philadelphia:UniversityofPennsylvaniaPress.Churchland,P.M.1989:ANeurocomputationalPerspective:TheNatureofMindandtheStructureofScience.Cambridge,MA:BradfordBooks,MITPress.Dewey,J.1910:TheInfluenceofDarwinonPhilosophy.NewYork:HenryHolt.——1925:ExperienceandNature.LaSalle,IL:OpenCourt,1958.787\nSUSANHAACK——1929:TheQuestforCertainty.NewYork:Putnam,1960.——1938:Logic,theTheoryofInquiry.NewYork:HenryHolt.Haack,S.1993:EvidenceandInquiry:TowardsReconstructioninEpistemology.Oxford:Blackwell.Hanson,N.R.1958:PatternsofDiscovery.Cambridge:CambridgeUniversityPress.Harman,G.1965:TheInferencetotheBestExplanation.PhilosophicalReview,71,86–95.James,W.1897:TheWilltoBelieveandOtherEssaysinPopularPhilosophy.NewYork:Dover,1956.——1907:Pragmatism,ed.F.BurkhardtandF.Bowers.Cambridge,MA:HarvardUniversityPress,1975.——1909:TheMeaningofTruth,ed.F.BurkhardtandF.Bowers.Cambridge,MA:HarvardUniversityPress,1975.—-1912and1909:EssaysinRadicalEmpiricism(1912)andAPluralisticUniverse(1909),ed.R.J.Bernstein.NewYork:Dutton,1971.Jardine,B.1986:TheFortunesofInquiry.Oxford:ClarendonPress.Lewis,C.I.1923:APragmaticConceptionoftheAPriori,JournalofPhilosophy,20,169–77;reprintedinH.S.Thayer(ed.)Pragmatism:TheClassicWritings.NewYork:Mentor,1970,andIndianapolis,IN:Hackett,1983.Margolis,J.1986:PragmatismWithoutFoundations.Oxford:Blackwell.Mead,G.H.1934:Mind,SelfandSociety,ed.C.W.Morris.Chicago:UniversityofChicagoPress.Migotti,M.1988:RecentWorkinPragmatism:RevolutionorReformintheTheoryofKnowledge?PhilosophicalBooks,29,65–73.Peirce,C.S.1900:ReviewofClarkUniversity,1889–1899:DecennialCelebration.Science,620–2;reprintedinP.P.Wiener(ed.)1958:CharlesS.Peirce:SelectedWritings.NewYork:Doubleday.——1931–58CollectedPapers,ed.C.Hartshorne,P.WeissandA.Burks.Cambridge,MA:HarvardUniversityPress(referencesbyvolumeandparagraphnumber).Putnam,H.1978:MeaningandtheMoralSciences.London:RoutledgeandKeganPaul.Quine,W.V.1953:TwoDogmasofEmpiricism.InFromaLogicalPointofView.Cambridge,MA:HarperTorchbooks.——1969:OntologicalRelativityandOtherEssays.NewYork:ColumbiaUniversityPress.Ramsey,F.P.1931:BeliefandPropositions.InTheFoundationsofMathematics,ed.R.B.Braithwaite.London:RoutledgeandKeganPaul.Reichenbach,H.1951:TheRiseofScientificPhilosophy.Berkeley:UniversityofCaliforniaPress.Rescher,N.1977:MethodologicalPragmatism.Oxford:Blackwell.Rorty,R.1979:PhilosophyandtheMirrorofNature.Princeton,NJ:PrincetonUniversityPress.——1982:ConsequencesofPragmatism.Brighton:HarvesterPress.——1991a:Objectivity,RelativismandTruth.Cambridge:CambridgeUniversityPress.——1991b:EssaysonHeideggerandOthers.Cambridge:CambridgeUniversityPress.Russell,B.1910:Pragmatism;andWilliamJames’sConceptionofTruth.InPhilosophicalEssays.NewYork:Longman’s,Green.Schiller,F.C.S.1907:StudiesinHumanism.London:Macmillan.Sellars,W.1973:GivenessandExplanatoryCoherence.JournalofPhilosophy,61.——1979:MoreonGivenessandExplanatoryCoherence.InJustificationandKnowledge,ed.G.Pappas.Dordrecht:Reidel.Stich,S.P.1990:TheFragmentationofReason:PrefacetoaPragmaticTheoryofCognitiveEvaluation.Cambridge,MA:BradfordBooks,MITPress.DiscussionQuestions1ExplainPeirce’sversionofthePragmaticMaxim,andtheroleitplaysinhisattitudetometaphysics.788\nPRAGMATISM2WhatdoesPeircemeanbysayingthat‘pragmaticismcouldhardlyhaveenteredaheadthatwasnotalreadyconvincedtherewererealgenerals’(Peirce5.503)?Washerightaboutthis?3ExplainPeirce’sdefinitionoftruthasthe‘UltimateOpinion’.Doyouthinkitisadefensibleaccountoftheconceptoftruth?4WhatdoesPeircemeanbysayingthatphilosophyshouldbeundertakenwith‘thescientificattitude’?5‘Ourpassionalnature...lawfullymay...decideanoptionbetweenpropositions,wheneveritisagenuineoptionthatcannotbyitsnaturebedecidedonintellectualgrounds’(James1897:11).Doyouagree?6WhatarethemainthemesofJames’s‘radicalempiricism’?HowdoesitdifferfromtheempiricismofLockeorHume?7‘“[T]hetrue”...isonlytheexpedientinthewayofbelief.’WhatdidJamesmeanbythis?Isitplausible?8CriticallycompareJames’sobservationthat‘booksuponethics...sofarastheytrulytouchthemorallife,mustmoreandmoreallythemselveswith...novelsanddramasofthedeepersort’(James1897:210),withDewey’sthat‘[e]xperimentalempiricisminthefieldofideasofgoodandbadisdemanded’(Dewey1929:258).9CriticallydiscussDewey’sattitudetotheideathat,whenweverifyaproposition,wediscoverittohavebeentrueallalong.10ExplainanddiscussthemeritsofDewey’scritiqueof‘theSpectatorTheoryofKnowledge’.11WhatdoesSchillermeanbysayingthattruthisrelativetoourpurposes?Isheright?12CriticallydiscussMead’stheoryofthe‘socialconstructionoftheself’.13ExplainLewis’sdoctrineofthepragmaticapriori;isitaplausibleapproachtothequestionofalternativelogics?14‘PragmatiststhinkthatthehistoryofattemptstoisolatetheTrue,ortheGood,ortodefinetheword“true”or“good,”supportstheirsuspicionthatthereisnointerest-ingworktobedoneinthisarea’(Rorty1982:xiv).Isthatwhatpragmatiststhink?15‘Accordingtothepragmatists,tosay“itistruethatotherpeopleexist”means“itisusefultobelievethatotherpeopleexist”...Nowifpragmatistsonlyaffirmedthatutilityisacriterionoftruth,therewouldbemuchlesstobesaidagainsttheirview’(Russell1910:119–20).IsthisafaircommentonJames’saccountoftruth?789\n37FregeandRussellR.M.SAINSBURYGottlobFrege(1848–1925)andBertrandRussell(1872–1970)bothcametophi-losophythroughmathematics.Theybothcontributedtoadoctrineknownaslogicism,accordingtowhichmathematicsisapartoflogic,andsoisasunquestionableaslogic.Inpursuitofthisaim,eachdevelopedthebasisofwhatwenowthinkofasmodernlogic.Eachregardedhislogicallanguageasbetterabletorepresentthoughtthanordi-narylanguage.Thisviewgeneratedthenotionof‘logicalform’,whichhasbeencrucialinthedevelopmentofphilosophyforthelastcentury.Eachwasledintofundamentaldiscussionsaboutmeaning.Intheareasinwhichtheirworkoverlapped,theyagreedaboutwhichproblemswerecentral,butdisagreedinmanycasesabouthowtosolvethem.Allcurrentworkinlogic,thephilosophyofmathematicsandthephilosophyoflanguageeithertakesforgrantedsomethingowedtooneorbothofthem,orelsetakestheirviewsasthemostimportantonestodefendorattack.Thischapterwillfocusonthenotionofmeaningasitoccursintheworksofthesetwophilosophersandwillexpoundtwocentralideasfromeach:Frege’snotionofaconcept,andhisdistinctionbetweensense(Sinn)andreference(Bedeutung);andRussell’saccountofnamesandhistheoryofdescriptions.Inbothcases,andespeciallyRussell’s,whoseworkrangedmorewidelythanFrege’s,thisisaverysmallproportionoftheideasforwhichtheyarefamous.1IntroductionFregewasbornin1848,andspentmostofhisadultlifeinthemathematicsdepart-mentoftheUniversityofJena.Hisfirstmajorwork,theBegriffsschriftor‘concept-script’,inwhichhesetouthisformallanguage,waspublishedin1879,whenRussellwasachildofseven.InhisGrundlagen(GroundworkofArithmetic)of1884hegaveaninformalexpositionofhislogicistposition,definingnumbersintermsofextensionsofconcepts,whichhetooktobeapurelylogicalnotion.ThetwoworkswerebroughttogetherinhisGrundgesetze(BasicLawsofArithmetic,1893),inwhichthefor-maldevicesoftheBegriffsschriftwereappliedtothelogicistprogrammesetoutintheGrundlagen.Closetothistime,hewrotethetwoarticlesuponwhichthischapter\nFREGEANDRUSSELLfocuses:‘FunctionundBegriff’of1891(‘FunctionandConcept’,1984)and‘UberSinnundBedeutung’(‘OnSenseandMeaning’,1892).Russell’sfirstexpositionoflogicism,andoneofhisearliestworks,wasinhisPrinciplesofMathematics(1903).Atalatestageinitspreparation,hediscoveredthatapparentlytrueassumptionsledtoacontradictionconcerningclasses,adiscoveryknownasRussell’sparadox.HewrotetoFregein1903,showingthatFrege’slogicalsystemintheGrundgesetzepermittedaproofofthecontradiction.TheblowhadaparalysingeffectuponFrege.Anappendixtovolume2oftheGrundgesetze(whichwasaddedin1903)suggestedaweakeningoftheaxiom(BasicLawV)whichheregardedastherootofthetrouble,buthecametoseethatthiswouldpreventderivationswhichwereneededforthelogicistprogramme,andDummetthassuggestedthatitislikelythat,by1906,Fregehaddespairedofcarryingthroughhislogicism(Dummett1981:21–2).Russell’sownfinalresponsetothecontradiction,takingtheformofatheoryoftypes,waspublishedin1908,andthefullstatementofhislogicism,incorporatingtypetheory,ishismonumentalPrincipiaMathematica(1910–13),writtenjointlywithA.N.Whitehead.ThefirstdecadeorsoofthetwentiethcenturywasaveryfruitfulperiodforRussell,andthetwoarticlesIshallmainlyfocusondatefromthen.Theyare‘OnDenoting’(1905)and‘OnKnowledgebyAcquaintanceandKnowledgebyDescription’,whichwaspublishedin1911andwhichisreprintedalmostunchangedinhisProblemsofPhilosophy(1912).ThisrepresentsonlyafragmentofRussell’sphilosophy,omittingnotonlyhisworkinthePHILOSOPHYOFMATHEMATICS(chapter11),butalsohislogicalatomism(seeespeciallyRussell1918–19),histheoriesofmind(seeespeciallyRussell1921)andhislateworks(whichincludeInquiryintoMeaningandTruth,1941,andHumanKnowledge,1956).Inadditiontoworkinphilosophy,Russellhadwide-ranginginterests,andneverheldanestablisheduniversitypost.Hewroteanumberofpopularbooks(forexample,MarriageandMorals,1929),wontheNobelPrizeforliteraturein1950,wasacivilrightsactivistinthe1950sand1960s,andbecameawell-knownpublicfigure,espe-ciallyasamemberoftheCampaignforNuclearDisarmament(hewasarrestedforpar-ticipatinginoneoftheirprotestdemonstrations).HisAutobiographycausedastirbyitsselectivefrankness,andbytheratherunattractivepictureitconveyedofthegreatman’stardyyetintenseemotionaldevelopment.Hediedin1970,outlivingFregeby45years.2FregeonFunction,ConceptandObjectManysentencessaysomethingtrueorfalseabouttheworld.Whatmakesthispossi-ble?Perhaps,asRussellthought,acrucialpointisthatthewordsofthesentencestandforpartsorfeaturesoftheworld.Perhaps,in‘Londonisbeautiful’,‘London’standsforLondon,and‘beautiful’forbeauty.Evenifthisistherightstart,itcannotbetheendofthestory,fornothinghasbeensaidtodistinguishthesentencefromthemerelist‘London,beauty’:thewordsinthelistalsostandforpartsorfeaturesoftheworld.AsRussellputit,theprocessofanalysis,itemizingwhatthepartsofasentencestandfor,791\nR.M.SAINSBURYseemstodissolvetheunityoftheproposition:thefeaturesthatmakeaproposition,unlikeamerelist,somethingthatrepresentstheworldasbeingthus-and-so,andsomethingthatcanbeaffirmedordenied.Russellhimselfnevergaveanydetailedsolutiontothisproblem.However,oneofthecornerstonesofFrege’sapproach,setoutindetailinhis‘FunctionandConcept’(Frege1891),suggestsawayofapproachingit.2Instandardarithmeticalnotation,anexpressionlike‘x’issaidtostandforafunc-tion.Ifwereplace‘x’by‘3’wehaveanexpressionwhichstandsfor9.Thefunctionfor2which‘x’standsyields9asitsvaluewhen3isitsargument;ingeneral,thisfunction2yieldsasvaluethesquareoftheargument.Anexpressionlike‘x=9’isinsomewayssimilar,inhavingtheundeterminedpart‘x’.Canweviewsuchanexpressionasalsostandingforafunction?Ifwereplace‘x’by‘3’wegetasentencewhichistrue;ifwereplaceitbyanexpressionforanyotherpositivenumber,wegetasentencewhichis2false.Fregesuggestedthatthisrevealshowwecanthinkof‘x=9’asstandingforafunction:itstandsforafunctionwhichyieldsthevaluetruefortheargument3,andthevaluefalseforanyotherpositivenumberasargument.Thisinvolvessupposingthatthereareabstractobjects,theTrueandtheFalse,andequatingbeingtruewithbeingappropriatelyrelatedtotheTrue,andsimilarlyforbeingfalse.Afunctionisincomplete.Justasanexpressionforafunctioncontainsanundeter-minedpart,say‘x’,callingforcompletionbyadeterminateexpression,say‘3’,sothisfeatureismirroredinthenatureofthefunctionitself.Itneedsanobjectasargumenttoyieldavalue.Aconceptisafunctionofaspecialkind:onewhosevalue,foreveryargument,isatruthvalue.Ifaconcept-wordiscompletedbyinsertingawordforanobject,whichFregecallsapropername,intotheblankspace,theresultisasentence.Theconceptexpressedby‘()isbeautiful’hasanincompletenesswhichmirrorsthatoftheexpres-sion:thereisanundeterminedpart,heremarkedby‘()’.Ifweinsertapropername,say‘London’,intothespace,theresultisthesentence‘Londonisbeautiful’.Althoughhedidstressthepredicativenatureofconcepts(see,forexample,Frege1892a:187),Fregehimselfdidnotexplicitlyusethisaccounttoaddresstheproblemoftheunityoftheproposition.However,onecanseeintheaccountatleastthebeginningofasolu-tion.Itisonethingtodevisealist,listingsayanobjectandaconcept;itisanothertoinserttheobjectintotheslotwhichconstitutestheincompletenessoftheconcept.FregethushasroomforthedifferencewithwhichRussellwasconcerned.Frege’simmediateuseoftheanalysisofsentencesintermsofconceptsandobjectswastointroducehistheoryofgeneralityintermsofquantifierandvariable.‘Allmenaremortal’istobeunderstoodasineffectclaimingthatthevalueofthefunction‘ifxisamanthenxismortal’istheTruewhateverobjectisargument.Histheoryofquantificationisoneofhisgreatestachievements,and,whileithasbeenrefined,andalternativesdeveloped,itisinnodangerofbeingsupplanted.Russellproducedasimilartheory(Russell1905).Heclaimedthat,althoughhehadreadFrege’sBegriffsschriftinthe1890s,hehadnotunderstoodit.ThepoorqualityofRussell’sfirstpresentationcomparedtoFrege’ssup-portshisclaimtohaverediscoveredthetheoryforhimself(Russell1967–9:68).Russell’slateraccountsofthetheory(particularlyRussell1908,1918–19;RussellandWhitehead1910–13),thoughsomuchlaterthanFrege’saccount,wereprobablymorewidelyreadandinfluential.792\nFREGEANDRUSSELL3Sinn(Sense)andBedeutung(Reference)Compareatruthoftheform‘ais(identicalto)b’withoneoftheform‘ais(identi-calto)a’.Wearetoimagine‘a’and‘b’replacedbypropernames,forexampleby‘Hesperus’and‘Phosphorus’respectively.ThesewerenamessupposedlygivenbytheancientstoVenus:‘Hesperus’wasusedforaplanetwhichwasfirsttobeseenintheeveningatcertaintimesoftheyear;‘Phosphorus’wasusedforaplanetwhichwaslasttofadeinthemorningatcertainothertimesoftheyear.Theydidnotrealizethatbothappearanceswereofthesameheavenlybody.Eachpropernamestandsfor(bedeutet)Venus;VenusistheBedeutungofbothnames.InordinarycontextsBedeutungisbesttranslatedas‘meaning’.However,ithasbeencommontomarkFrege’ssomewhattech-nicaluseofthewordbyadifferenttranslation,mostpopularly‘reference’,andIshallfollowthisprecedent.Thataisbcanbeavaluableextensionofourknowledge:underappropriatecircumstances,comingtoknowthishas‘cognitivevalue’.However,thataisaseemstrivialanduninformative,andoflittleornocognitivevalue.Thiscontrast,whichitisnaturaltotreatasinvolvingtwodistinctitemsofknowledge,cannotbeexplainedintermsofreference,sincethetwosentencescoincidepoint-by-pointinreference.Fregecorrectlyinferredthatacompleteaccountofanamemustalludetootherpropertiesthanitsreference.Forallthathasbeensaid,thesefurtherpropertiescouldbeidiosyncratic(varyingfrompersontoperson),orcouldbeSYNTACTIC(pp.158–60)(forexample,residingmerelyinthefactthatintheonecasetwodifferentnamesareinvolved).However,Frege’shypothesiswasthatthedifferenceisoneofsense(Sinn),wherethisisconnectedwithwhatmustbegraspedincomingtolearnhowtouseaname,andwithhowsomethingispresentedorthoughtof.Senseisthusneitheridio-syncratic,sinceitisimplicatedincomingtolearntheproperpublicuseofaword,normerelysyntactic,sinceitisimplicatedinhowthingsarethoughtof.Thesenseofanexpressionisanabstractobject,notamereidea,andisinprincipleavailabletoanyone.Whetherasentenceexpressessomethingofcognitivevaluetouswill,onthistheory,dependnotonlyuponthereferenceofeachofitsnamesbutalsoupontheirsense.Ifwethinkthatcomingtoknowthataisbcanextendourknowledgeinawaythatcomingtoknowthataisacannot(supposingwecansomuchasmakesenseof‘coming’toknowthelatter),thenwearecountingthethingsweknow,objectsofknowledge,asobjectsindividuatedbysense.Whereasabrief,andinmyviewdecisive,argumentleadstotheconclusionthatthereismoretoanamethanitsreference,weshouldnotthinkofFregeasofferinganydirectargumentfortheconclusionthatthissomethingissense,regardedaspublicandsemantic;rather,themeritoftheviewissupposedtoemergeasitiselaborated,extendedtoexpressionsotherthannames,andappliedtovariousproblems.Thedistinctionbetweensenseandreferencewasfirstintroducedforpropernameslike‘Hesperus’.Dothesenotionsalsoapplytowholesentences?Fregeaffirmedthatanassertoricsentenceexpressesathought,andarguesthatthisshouldbetakenasthesentence’ssenseandnotasitsreference.Heheldittobeobviousthat,forexample,‘HesperusisHesperus’expressesadifferentthoughtfrom‘HesperusisPhosphorus’.Sincenamesinthesentencesagreeintheirreference,itwouldseemthatwecannot793\nR.M.SAINSBURYidentifythethoughtsthesentencesthemselvesexpresswiththeirreference.So,ifthethoughtasentenceexpressesiseitheritssenseoritsreference,itmustbeitssense.ThisdoesnotgiveFregewhatheneedsunlesstheantecedentoftheconditionalcanbeestab-lished.Giventhatsensewasoriginallyintroducedtoexplainadifferencebetweennameswhichaffectswhatisknown,andtoindividuateOBJECTSOFKNOWLEDGE(p.782),andgiventhatthoughtsarepossibleobjectsofknowledge,itisplausibletostipulatethatthesenseofasentenceisthethoughtitexpresses.Cananassertoricsentenceasawholebecreditedwithareference?Referencewasintroducedintermsoftheseeminglytransparentrelationholdingbetweenanameanditsbearer.Ifwethinkofreferenceaswhollydeterminedbythisrelation,thenitwouldseemthatthereferenceofacompletesentencelike‘Hesperusisaplanet’wouldsimplybeHesperus.However,Fregeinvitedustoconsideranotherfeatureofthereferenceof‘Hesperus’:itexhauststheword’scontributiontothedeterminationofthetruthorfalsehoodoftypicalsentencesinwhichthewordoccurs.In‘Hesperusisaplanet’,allthat‘Hesperus’contributestowhetherthesentenceexpressessomethingtrueorsome-thingfalseistointroduceHesperusasitsreference.Sofarastruthandfalsitygo(asopposedtocognitivevalue)onlyreferenceseemstomatter.Toestablishthisconnectionbetweenthereferenceofthepartsofasentence,andwhetherthewholesentenceistrueorfalse,Fregeinvitedustothinkofaspecialcase:fiction.Inthesentence‘OdysseuswasputashoreatIthacawhilesoundasleep’wehaveasentencecontainingapropername,‘Odysseus’,whichlacksreference(Frege1892b:162).Thesentenceasawholeisneithertruenorfalse.Soitlooksasifwhenapartlacksreference,thewholesentencelacksthepropertyofbeingtrueandthepropertyofbeingfalse.Frege’ssuggestionwasthatweshouldseewholenon-fictionalsentencesashavingoneorotherofthetruthvaluesastheirreference.Wesawthatalmostthisviewwassuggestedtohimbyhistreatmentofconceptsaskindsoffunctions.Iftheconceptexpressedby‘isaplanet’takesthevalueTrueforargumentHesperus,thenthesen-tence‘Hesperusisaplanet’mustberelatedinsomespecialwaytothisvalue.Havingthevalueasitsreferenceseemsjusttherightrelation.Theapplicationofthenotionsofsenseandreferencetopartsofspeechotherthanpropernameswasguidedbycompositionalityprinciples.Fregeineffectstipulatedthatthesenseandreferenceofasentencearedeterminedbythesenseandreferenceofitspartsinatleastthefollowingsense:substitutionsofpartswhichagreeinsense(reference)leavethesense(reference)ofthewholeunchanged.Giventhis,thatthepropernames‘Hesperus’and‘Phosphorus’differinsensebutarealikeinreferenceestablishesthatthesentences‘HesperusisHesperus’and‘HesperusisPhosphorus’differinsense(expressdifferentthoughts)butarealikeinreference(havethesametruthvalue).4IdentityStatementsandBearerlessNames:Russell’sViewofNamesasAssociatedwithDescriptionsFrege’sfirstcharacterizationofsenseis‘themodeofpresentationofthethingdesig-nated’.Theideaisthatthesameobjectcanbepresentedinmorethanoneway(asVenusispresentedinonewayinconnectionwithusesof‘Hesperus’andanotherin794\nFREGEANDRUSSELLconnectionwithusesof‘Phosphorus’).Sense‘determines’referenceinthatexpressionsthatagreeinsenseagreealsoinreference.However,Fregeseemedtoallowthatsensewithoutreferenceispossible:‘Theexpression“theleastrapidlyconvergentseries”hasasensebutdemonstrablynothingisitsreference’(Frege1892b:159).ThoughthisistheorthodoxinterpretationofFrege,ithasbeenchallengedbyGarethEvans(1982),whopointsoutthatFregeveryoftenusedthecaseoffictionwhendiscussingthepos-sibilityofsensewithoutreference.Hisremarksaboutfiction,Evanssuggests,couldbeinterpretedasclaimingthatinfictionwepretendthatthereisasense(andalsothatthereisareference).Onedoesnothavetosupposethattherereallyisasense.Evans’sviewreceivessupportfromsuchremarksofFrege’sas:‘Asentencecontainingapropernamelackingreferenceisneithertruenorfalse;ifitexpressesathoughtatall,thenthatthoughtbelongstofiction’(Frege1979:194).Inotherwords,ifasentencecon-tainsanamewithoutreference,thenitlackssense(exceptforthepretendthoughtsoffiction);whichcanonlybeexplainedbythefactthatanamelackingreferencelackssense.NotalltheevidencesupportsEvans’sview,asthesentencequotedabovemakesplain.Ifwereturntotheorthodoxinterpretation,weshouldnotholdthat,forFrege,sense‘determines’reference,ifthisisunderstoodtoentailthateachsenseisassociatedwithareference.Fregesaysthatanexpressionexpressesitssenseanddesignatesitsreference(weshouldpresumablyadd‘ifany’).Russell,using‘meaning’forsenseand‘denoting’forreference,arguedthatatheoryofthiskindleadsto‘aninextricabletangle’(Russell1905:50).TheexegesisofRussell’sargumentisnoeasymatter,andcommentatorsdisagreebothonhowitshouldbeunderstoodandonwhetheritisofanysignificanceforFrege’sviews(seeHochberg1976;BlackburnandCode1978;Pakuluk1993).WhatcannotbedisputedisthatRussell’sownaccountoftheworkingsofthoughtandlanguageusedasinglenotion,whichheusuallycallsmeaning,inplaceofFrege’stwonotionsofsenseandreference.Accordingtothissinglenotion,simpleexpressionslike‘this’havemeaningbystandingforsomething.Complexexpressionsarecapableofbeingunderstoodsolongastheyarecomposedofsimpleexpressionshavingmeaning.Toremindusofthedistinctionbetweenhowtheaccounttreatssimpleandcomplexexpressions,Russellsaidofthelatterthattheyhavenomeaningontheirownaccount.OnewaytoformaconceptionoftheadequacyofFrege’stwo-notiontheoryistocompareitwithRussell’scontrastingsingle-notionone.HowcouldRussellhaveallowedthatitisonethingtobelievethatHesperusisHesperus,andanothertobelievethatHesperusisPhosphorus,withoutallowingsomethinglikeFrege’sdistinctionbetweensenseandreference?Theanswer,accordingtoRussell,wasthatnameslike‘Hesperus’and‘Phosphorus’cannotbeusedtogivethebestaccountofwhatisgoingoninathinker.Rather,‘thethoughtinthemindofapersonusingapropernamecorrectlycangenerallyonlybeexpressedexplicitlyifwereplacethepropernamebyadescription’(Russell1967–9:29).By‘description’,inthiscontext,Russellmeansadefinitedescription,thatis,aphraseoftheform‘theso-and-so’.Perhapssomeoneassociates‘Hesperus’with‘theEveningStar’and‘Phosphorus’with‘theMorningStar’.Doesnottheoriginalproblemariseagain?Since‘theEveningStar’and‘theMorningStar’standforthesamething–or,inFrege’sterminology,havethesamereference–yet‘theMorningStaristheMorningStar’differsincognitivevaluefrom‘theMorningStaristheEveningStar’,mustonenotrecognizesomethingover795\nR.M.SAINSBURYandabovewhatexpressionsstandfor?Russell’sansweristhatoneneednot,forinsuchcases,wherethephraseiscomplex,onecanexplainthedifferenceintermsofthedif-ferentpartsofthecomplexes.Onecanexplainthedifferencesbetween‘theEveningStar’and‘theMorningStar’intermsofthefactthatonephrasecontainsawordwhichstandsforthemorningandnottheevening,andtheothercontainsawordwhichstandsfortheeveningandnotthemorning.Thereisnoneedtoappealtoanythingcor-respondingtoFregeansense.(FormoredetailsonhowRusselldevelopedthisview,seesection6below.)Somenameshavenobearers,andyethaveasystematicroleinlanguage(suchas‘SantaClaus’or‘Vulcan’–theplanetpostulated,needlesslyasitturnedout,toexplaintheorbitofMercury).OntheconventionalinterpretationofFregeadoptedhere,hecanaccountforthissimplyasacaseofsensewithoutreference.Howcouldasingle-notiontheorylikeRussell’scopewithsuchcases?Russellagainappealedtothelinkbetweennames(ofthekindinquestion)anddescriptions.Weassociate‘SantaClaus’with,forexample,‘thebeardedmanfromLaplandwhobringsChristmaspresents’,and‘Vulcan’with‘theplanetlyingbetweenMercuryandtheSun’.Eachcomponentofthesedescriptionscan,onRussell’ssingle-notionview,beregardedashavingmeaningbystandingforsomething;andthisaccounts,atoneremove,fortheintelligibilityofthenameswithouttheirhavingtostandforanything.Suchnamesandtheirassociateddescriptionshavenomeaningontheirownaccount,unlikegenuinelysimpleexpressions.5NamesandCommunicationHowshouldwecomparethesetheories?Onewayistoasktowhatextenteachprovidestheresourcesforacorrectaccountoftheworkingoflanguage.Frege’snotionofsenseisfirmlyconnectedtolanguage:‘thesenseofapropernameisgraspedbyeverybodywhoissufficientlyfamiliarwiththelanguage’(Frege1892b:158).Hestressedthatwemustnotconfusethesenseofanexpression,whichisobjec-tiveandinprincipleavailabletoall,withtheideasinthemindsofpeoplewhentheyuseit,which,beingpsychological,aresubjectiveandconfinedtoasingleperson.Heheldthat,unlikethepublicsenseofanexpression,‘everyideahasonlyoneowner;notwomenhavethesameidea’(Frege1918:361).OneintuitionisthatFrege’sinsistenceiscorrect:languageslikeEnglishorGerman,ortheformallanguageswhichwereFrege’smainconcern,areindeedpublic,andacorrectaccountmustbeinpublictermslikesense,ratherthanidiosyncraticoneslikeideas.Wecommunicateinapubliclan-guage,andthestraightforwardFregeanaccountofthisisthatwedosobysharingthoughts:foraspeakertosucceedincommunicatingisforhimtoarouseinthemindofthehearerthethoughtthatheorsheexpresses.Russell’saccountassofardescribed,appliedtonameslike‘Hesperus’,isbasedonpossiblyidiosyncraticfeaturesofvariousspeakers.Wearetofindthedescriptionsinthemindofsomeonewhousessuchaname,andRussellwasexplicitthat‘thedescrip-tionrequiredtoexpressthethoughtwillvaryfordifferentpeople,orforthesamepersonatdifferenttimes’(Russell1967–9:29).Onthefaceofit,thisseemsill-adaptedtoprovideagoodaccountofthefunctioningofapubliclanguage.796\nFREGEANDRUSSELLBeforeinvestigatingthis,itisimportanttostressthatRussell’stheoryconnectingnamesanddescriptionswasnotadvancedprimarilyasatheoryofapubliclanguage,butratherasanaccountofthoughtorJUDGEMENT(pp.726–9).ItwasthusnotadvancedundertheconstraintofpublicitywhichFregetookuponhimself.However,inabriefpassage,Russellshowedhowhisideascouldbedevelopedtoprovideanaccountofcommunicationinapubliclanguage.TheaccountisofinterestbecauseitmovesawayfromtheFregeanmodelofcommunicationasthought-sharing.Letusconsiderthetwophilosophers’reactionstoaproblemaboutnames.Whentryingtosaywhatthesenseofanamelike‘Aristotle’or‘DrLauben’is,Fregefounditnaturaltoproduceadefinitedescription.However,heobservedthatinthecaseofanactualpropernamesuchas‘Aristotle’opinionsastothesensemaydiffer.Itmight,forinstance,betakentobethefollowing:thepupilofPlatoandteacherofAlexandertheGreat.Anybodywhodoesthiswillattachanothersensetothesentence‘AristotlewasborninStagira’thanwillamanwhotakesasthesenseofthename:theteacherofAlexandertheGreatwhowasborninStagira.(Frege1892b:158)Fregeputthisvariabilitydowntoimperfectionsofordinarylanguage,andclaimedthatitoughtnottooccurinaperfectlanguage.Thisis,ineffect,anadmissionthatsense,thoughappropriatetoanaccountofapubliclanguageinvirtueofthegeneralfeatureofbeingpublic,fitsuneasilywiththefactthatnamesareassociatedwithdefinitedescriptionsinavariable,andtothatextentnon-public,way.Thevariabilityofassoci-ateddescriptions,theapparentlyunequivocalcharacterofouruseofnames,andthesuppositionthatdescriptionsareoftenconstitutiveofsense,shouldhaveledFregetodenythatpropernamesinordinarylanguagehavesense;butthiswouldhaveunder-minedtheexplanationofthecognitivevalueof‘HesperusisPhosphorus’.Russellsteeredthroughthisdifficultybysayingthatthefeatureofanamelike‘Hesperus’thatmattersincommunication,andthefeaturethatfixesthename’scon-tributiontowhetherwhatissaidistrueorfalse,isnotthedescriptionassociatedwithitbythisorthatuser,butratherwhatthenamestandsfor,wherethisisdeterminedbythevariousassociateddescriptions.Supposeaspeaker,S,uttersthewords‘Bismarckwasastute’.Insodoing,hewill,onRussell’stheory,thinkathoughtmoreexplicitlyexpressibleas,perhaps,‘ThefirstChancellorofGermanywasastute’.Scommunicatessuccessfullywithhisinterlocutor,H,justonconditionthat,forsomedescription,perhaps‘theugliestmaninEurope’,whichalsodenotesBismarck,HrealizesthatShassaidsomethingabouttheugliestmaninEuropewhichistrueifandonlyifheisastute.ItisnotnecessarythatSandHshouldshareathought.Allthatisrequiredisthatthepossiblyvariousdescriptionstheyassociatewiththename‘Bismarck’shouldstandforthesamething;forthisiswhatitmeanstosaythatthename‘Bismarck’hasa(public)referenceinthelanguagecommunitytowhichSandHbelong(Sainsbury1993).Theupshotisthatifanaccountofnamesistobecouchedintermsofdefinitedescriptions,onedoesbettertoviewtherelationshipinRussell’swaythantoviewthedescriptionsasspecificationsofFregeansense.AlthoughFregeoffereddefinitedescrip-tionsasexamplesofaname’ssense,hewasnotcommittedtotheviewthatthesenseofapropernamecantypicallybegivenbyadescription,andaFregeandoesbesttodenythis(seeMcDowell1977;Kripke1972).797\nR.M.SAINSBURY6Russell’sTheoryofDescriptionsWehaveseenthatRussellprovidedanaccountoftheintelligibilityofdefinitedescrip-tionswithoutsupposingtheyhadtostandforanything;itisenoughthattheirpartsbeintelligible,whichistypicallyamatterofthepartsstandingforsomething.Thenotionof‘standingfor’,orasRusselloftensays‘meaning’,isthusthesinglenotionwhichissupposedtogiveabetteraccountthanFrege’stwinnotionsofsenseandreference.Inmoredetail,Russellthoughtgenuinelysimplesingularandgeneraltermshadmeaningbystandingforsomething:aparticularinthecaseofsingulartermslike‘this’,auniversalinthecaseofgeneraltermslike‘red’.Understandingsuchexpressionsinvolvesacquaintancewithwhattheystandfor.Thereisjustoneothersourceofmeaning,andthatderivesfromthelogicalconstants,like‘and’,‘not’,‘isidenticalto’and‘all’.Russellnevergaveasatisfactoryaccountofhowsuchexpressionshadmeaning,butitisobviousthattheycontributetotheintelligibilityofcomplexexpressions.Russelladmittedaverynarrowclassof‘logicallypropernames’,singulartermswhichalogician,asopposedtoanordinarypersonoragrammarian,shouldcountaspropernames.ThesearegenuinelySIMPLE(pp.813–18)singularterms,andunder-standingtheminvolvesacquaintancewithwhattheystandfor.Russellsupposedthatiftwosuchterms,say‘a’and‘b’,stoodforthesamething,anyonewhounderstoodbothwouldbeawareofthetruthof‘aisb’.Sincemostofwhatweordinarilycallpropernames,suchas‘Hesperus’and‘Aristotle’,donotsatisfythisrequirement,heexcludedthemfromthecategoryoflogicallypropernames,andexplainedtheirfunctioningintermsofassociateddescriptions,inthefashionalreadydescribed.TheonlylogicallypropernamesinordinaryEnglish,accordingtoRussell,are‘this’and‘that’.DefiniteDescriptionsRussellisfamousforhistheoryofdefinitedescriptions.Hesaidthataphraseoftheform‘themoon’isaquantifierphrase,like‘nomoons’and‘allmoons’,andshouldnotbelikenedtoalogicallypropernamelike‘this’.Inparticular,‘the’istheuniquenessquan-tifier,so‘themooniscold’istrueif,andonlyif,thereisexactlyonemoonanditiscold,and‘themoonexists’istrueif,andonlyif,thereisexactlyonemoon.ThisenabledRusselltoexplainhowasentencelike‘thegoldenmountaindoesnotexist’canbetrue:itistrueif,andonlyif,eithernothingormorethanonethingisbothgoldenandamountain.Iftheintelligibilityofthephrase‘thegoldenmountain’requiredthatitstandforsomethingthen,unlesstherearethingswhichdonotexist,thesentence‘thegoldenmountaindoesnotexist’couldnotbetrue:if‘thegoldenmountain’stoodforsomething,thesentencewouldbefalse,andif‘thegoldenmountain’didnotstandforanything,thesentencewouldbeunintelligible.Russellexplainedtheintelligibilityofthephraseintermsoftheintelligibilityofitsparts:‘the’isineffectalogicalconstantlike‘all’(sinceitisdefinedintermsofsuchexpressions)andsohasmeaninginwhateverwaylogicalconstantsdo;‘golden’and‘mountain’canhavemeaningbystandingfortheuniversalsofgoldennessandmountainhood.798\nFREGEANDRUSSELLFregethoughtthateventhoughthereiscertainlyaconceptcorrespondingtoanexpres-sionlike‘squarerootof-1’,oneshouldnotwithoutcarepermitoneselftousetheexpression‘thesquarerootof-1’:onewouldneedtoassureoneselfthattheexpres-sionhadareferencebeforeadmittingitintoaperfectlanguage.Ifbyanychanceonedoesfindoneselfwithexpressionswithoutreference,onecanintroducethespecialstip-ulationthattheunrulyexpressionistohavethenumber0asitsreference.AsRussellsaid,thisisan‘artificial’andunsatisfactorysuggestion,andoneoutofkeepingwithotherFregeantheses.Toprovideareferencebyadhocstipulationistoforgetthat,forFrege,anexpression’ssensedeterminesitsreference(ifany),andthesenseofanexpressionisdeterminedbythatofitsparts.Frege’sviewonthismatterisnoteasytounderstand,sinceitisveryhardtoseehowaperfectlanguageadequateformathe-maticscouldfailtohavetheresourcestopermitsuchphrasesas‘theprimenumberbetween5and7’.Fregedoesindeedexplainhowthesentence‘thereisnosuchthingastheprimenumberbetween5and7’canbetrue,withoutattributingareferenceto‘theprimenumberbetween5and7’.SimilarlytoRussell,hewouldhaveusregardthesentenceassayingthatnoobjectuniquelyfallsundertheconceptofbeingaprimenumberbetween5and7.WhatishardtoseeishowFregecouldsupposethatthesyntaxforalanguagecouldproperlydiscriminatebetweenconcept-wordswhichmay,andthosewhichmaynot,beprefixedby‘the’.7IndirectDiscourseFrege’sdoctrineofsenseandreferencehasacomplicationwehavenotyetmentioned.Hethoughtitobviousthatinindirectspeech,forexamplewhenwesaythattheancientastronomersdidnotrealizethatHesperuswasPhosphorus,‘itisquiteclearthatinthiswayofspeakingwordsdonothavetheircustomaryreferencebutdesignatewhatisusuallytheirsense’(Frege1892b:159).Inshort,inindirectspeech,wordshavean‘indirect’reference:itistheircustomarysense.Ratherthantakingthisasobvious,Fregecouldhaveusedhiscompositionalprin-ciplestoproveit.Ifwereplaceoneoccurrenceof‘Hesperus’by‘Phosphorus’inTheancientsbelievedthatHesperusisHesperusweturntruthintofalsehood.Since,bycompositionality,replacingapartbyanexpres-sionwiththesamereferencecannotaffectthereferenceofthewholesentence,thatis,itstruthvalue,thisshowsthat‘Hesperus’,asitoccursinthedisplayedsentence,doesnothaveHesperus(thatis,Phosphorus)asitsreference.Thehypothesisthat,inanindirectcontext,‘Hesperus’hasasitsreferencethesenseithasinordinarycontextsissuggestedbythefactthat,inanindirectcontext,wearetryingtoidentifysomeone’sthought,andathoughtisasense.Itwouldbeconfirmedifwefoundthatexchangingexpressionsinsuchcontextsbyoneshavingthesamecus-tomarysensewasguaranteedtoleavethereferenceofthewholesentenceunchanged.Itwouldberefutedifwecouldfindexpressionswiththesamesense,andasentencewhosetruthvaluewaschangedjustbyreplacingoneofthembytheother.Perhaps‘pail’and‘bucket’havethesamesense,yetreplacingonebytheotherin‘Johnsaidthat799\nR.M.SAINSBURYapailisabucket’altersthetruthvalue(Wiggins1976).ThisobservationmightinsteadbeusedtoshowthatonFrege’stechnicalconceptionofsense,‘pail’and‘bucket’differinsense.ThisputspressureuponustoexamineFrege’sconceptionmorecarefully.Theindirectreferenceofanameisitscustomarysense;butwhatisitsindirectsense?Fregehimselfsaidthat‘wedistinguish...customarysensefrom...indirectsense’.Buthemakesnomoreofthepoint,andDummetthasargued,convincinglyinmyview,thathewouldhavedonebettertoidentifycustomaryandindirectsense.Inotherwords,hewouldhavedonebettertodenyanydistinctionbetweensenseandreferenceinin-directcontexts.ThisconclusionwouldfollowfromFrege’scompositionalityprinciples,ifwecouldtakeasapremisethattheconditionsthatgovernSUBSTITUTIVITYSALVAVERITATE(p.95)inthedisplayedsentenceabovecoincidewiththeconditionsgoverningitinasentencelike:SomephilosophersdisputethattheancientsdidnotrealizethatHesperusisPhosphorus.Frege’sdistinctionbetweensenseandreferencearguablyhelpsusunderstandindi-rectspeech.Theshiftinreferenceasbetweenordinaryandindirectcontextsmirrorsthefactthatinordinarycontexts,truthisamatterofhowthingsare,whereasinindi-rectcontextstruthistypicallyaffectedbyhowthingsarethoughttobe.GiventhatRussellmadenodistinctionlikethatbetweensenseandreference,howcanhedealwithsuchcontexts?Russell’stheorydoesnotexplicitlydealwiththedifferentsubstitutivityprinciplesrequiredbyindirectcontexts.HisgoalistogiveanaccountofsentenceslikeOthellobelievesthatDesdemonalovesCassiowhichdoesnotrequireanysingleentitycorrespondingtoaFregeanthought.Russellheldthat,althoughonemightallowthattherearesuchthingsastruths,onecouldnotallowthattherearesuchthingsasfalsehoods;henceonecouldnotanalyseasentenceliketheonejustdisplayedasarelationbetweenOthelloandsomesingleobject(Russell1967–9:72).Tounderstandwhyheheldthis,oneneedstoappreciatewhatonemightcallhis‘non-conceptualism’(Sainsbury1986).Onpainofidealism,hebelievedthatthemindhadtostandindirectrelationstotheworld,notrelationsmediatedbyideasorotherrepresentations.Atruepropositioncouldbethoughtofasafact;butafalsepropositionwouldbesomekindoffalserepresentation,and,forRussell,therearenosuchthingsasrepresentations.ItisapitythatRussellneverconsideredFrege’sviewofrepresentations(thoughts)asabstractratherthanmental.Russell’sanalysisofthesentence‘OthellobelievesthatDesdemonalovesCassio’isthatitexpressesafour-termrelationofbeliefholdingbetweenOthello,Desdemona,loveandCassio.Usingthestandardconventionsoffirst-orderlogic,wemightwritethisasBelieves(Othello,Desdemona,love,Cassio).Suppose‘N’issomeothernameforCassio;supposeOthellohasencounteredtheuseofthisname,butwronglybelievesthatitdesignatessomeoneelse.Thenwecouldnotinfer800\nFREGEANDRUSSELLthatOthellobelievesthatDesdemonalovesN.YetonRussell’saccount,itwouldfollowthatsuchaninferenceisvalid.Inotherwords,Russell’stheoryfailstodojusticetothesubstitutionprinciplesappropriatetoindirectcontexts.Russellwouldacceptthisconsequence,providedthatthenamesinquestionarelogi-callyproper.Forthese,heheld,unqualifiedsubstitutionsalvaveritateispermissible.Whatmakesusfindthesubstitutionsunacceptableisthatwearethinkingofthesenamesnotaslogicallyproperones,butasonesforwhichacorrectaccountisofthekind,alreadysketchedinsection6above,thatinvolvesassociateddescriptions.PerhapsOthelloasso-ciates‘Cassio’withonedescription,say‘theF’and‘N’withanother,say‘theG’.Theanalysisofsentencesascribingbeliefwouldhavetobecomplicatedtodealwiththis,andRussellgivesnohintabouthowtoproceedinthisdirection.However,thereisnoques-tionofproblemsforsubstitutionprinciples,fortheserequirefortheirpremisesidentitystatements.AccordingtoRussell,expressionsoftheform‘theFistheG’arenotidentitystatements,fortheexpressionsflankingtheidentitysignarenotreallyreferringexpres-sions.Thelogicalformofsuchasentenceissimply:somethingisuniquelyFandisuniquelyG.Thisdoesnotevenlookasifitwouldlicencesubstitutionsalvaveritate.(Russellallowedthatincertaincontextswecantreatsomethinglike‘theFistheG’asifitwereanidentity:seePrincipiaMathematica14.16andtherelateddiscussion.)8ConclusionHowshouldwecompareFrege’sandRussell’scontributions?Fregehassoleclaimtothetheorythatconceptsarefunctions,andtotheuseofthisinexplainingquantifiersandvariables.Hewasalsothefirsttointroducethesense–referencedistinction.Russellattackedhimforthis,andinventeddistinctivetheoriesofnames,communicationanddefinitedescriptionstoexplainphenomenawhichFregeexplainedintermsofsenseandreference.SupposeoneaddedRussell’stheoryofdescriptionstoFrege’sview.OnecouldthenimagineFregeinsistingthateverysimpleexpressionhasasensewhichconsistsinthemodeofpresentationofanobject,wherethiscouldbeunderstoodinsuchawayastoensuretheexistenceoftheobjectpresented.Thesenseandreferenceofthesimpleexpressionswouldthendeterminethesense,andreferenceifany,ofthecomplexestheycompose,andtherewouldbenothingimperfectaboutsensewithoutreference.Theques-tionwouldthenremain:doweneedsenseaswellasreferencetoexplaintheworkingofthesimpleexpressions?ItisessentialtoanythinglikeFrege’sviewthattheanswershouldbeaffirmative,andtoanythinglikeRussell’sviewthatitshouldbenegative.FurtherReadingThefirstthingistoreadRussellandFrege.Isuggestthefollowingfourarticlestobeginwith:Frege(1891,1892b)andRussell(1905)(orperhapsbetter,inthefirstinstance,thelaterandclearerdiscussioninchapter16ofRussell1919),andRussell(1967–9,esp.ch.4).AgoodcommentaryinvolvingbothphilosophersisSalmon(1986).HecriticizesFrege’sdis-tinctionbetweensenseandreferencefromaRussellianperspective.OnFrege,thebeststarting-pointisKenny(1995)orBeaney(1996).Forfurtherstudy,Dummett’swritingsareessential:801\nR.M.SAINSBURYDummett(1973)supplementedbyDummett(1981).ForanindicationofthecurrentstateofthoughtaboutFregeonsenseandreference,seetheFregeissueofMind,vol.101,1992.OnRussell,youmighttrySainsbury(1979).RecentcollectionsincludeWadeSavageandAnderson(1989)andIrvineandWedekind(1993).ReferencesFregeandRussellFrege,G.1972[1879]:Begriffsschrift.TranslatedbyT.W.Bynum,ConceptualNotationandRelatedArticles.Oxford:ClarendonPress.——1884:DieGrundlagenderArithmetik.Breslau:W.Koebner.TranslatedbyJ.L.Austin,TheFoundationsofArithmetic.Oxford:BlackwellandMott,1950.——1891:FunctionandConcept.InB.McGuinness(ed.)CollectedPapersonMathematics,LogicandPhilosophy.Oxford:Blackwell,1984.——1892a:OnConceptandObject.VierteljahrsschriftfürwissenschaftlichePhilosophie,16,192–205.TranslationinB.McGuinness(ed.)CollectedPapersonMathematics,LogicandPhilosophy.Oxford:Blackwell,1984.——1892b:OnSenseandMeaning.ZeitschriftfürPhilosophieundphilosophischeKritik.Transla-tioninB.McGuinness(ed.)CollectedPapersonMathematics,LogicandPhilosophy.Oxford:Blackwell,1984.——1893:GrundgesetzederArithmetik,begriffsschriftlichabgeleitet.BandI.Jena:HermannPohle.PartstranslatedbyM.Furth,TheBasicLawsofArithmetic:ExpositionoftheSystem.LosAngeles:UniversityofCaliforniaPress,1964.——1918:LogicalInvestigations:Thoughts(translatedbyP.GeachandR.H.Stoothoff).InB.McGuinness(ed.)CollectedPapersonMathematics,LogicandPhilosophy.Oxford:Blackwell,1984.——1979:PosthumousWritings.Oxford:Blackwell.Russell,B.1903:ThePrinciplesofMathematics.Cambridge:CambridgeUniversityPress.——1905:OnDenoting.Mind,14,479–93.ReprintedinR.C.Marsh(ed.)LogicandKnowledge.London:RoutledgeandKeganPaul.——1908:MathematicalLogicasBasedontheTheoryofTypes.AmericanJournalofMathematics,30,222–62.ReprintedinR.C.Marsh(ed.)LogicandKnowledge.London:RoutledgeandKeganPaul.——1912:ProblemsofPhilosophy.Oxford:OxfordUniversityPress.——1918–19:LecturesonthePhilosophyofLogicalAtomism.Monist,28,29.ReprintedinR.C.Marsh(ed.)LogicandKnowledge.London:RoutledgeandKeganPaul.——1919:IntroductiontoMathematicalPhilosophy.London:GeorgeAllenandUnwin.——1921:TheAnalysisofMind.London:GeorgeAllenandUnwin.——1929:MarriageandMorals.London:GeorgeAllenandUnwin.——1940:InquiryintoMeaningandTruth.London:GeorgeAllenandUnwin.——1948:HumanKnowledge.London:GeorgeAllenandUnwin.——1967–9:Autobiography.London:GeorgeAllenandUnwin.Russell,B.andWhitehead,A.N.1910–13:PrincipiaMathematica.Cambridge:CambridgeUniversityPress.OtherwritersBeaney,M.1996:FregeMakingSense.London:Duckworth.Blackburn,S.andCode,A.1978:ThePowerofRussell’sCriticismofFrege:‘OnDenoting’.Analysis,38,65–77.802\nFREGEANDRUSSELLDummett,M.1973:Frege:PhilosophyofLanguage.London:Duckworth.——1981:TheInterpretationofFrege’sPhilosophy.London:Duckworth.Evans,G.1982:TheVarietiesofReference.Oxford:ClarendonPress.Hochberg,H.1976:Russell’sAttackonFrege’sTheoryofMeaning.Philosophica,18,2.ReprintedinhisLogic,OntologyandLanguage.Munich:PhilosophiaVerlag.Irvine,A.andWedekind,G.(eds)1993:RussellandAnalyticPhilosophy.Vancouver:UniversityofBritishColumbiaPress.Kenny,A.1995:Frege.Harmondsworth:PenguinBooks.Kripke,S.1972:NamingandNecessity.Oxford:Blackwell.McDowell,J.1977:OntheSenseandReferenceofaProperName.Mind,86,159–85.Pakuluk,M.1993:TheInterpretationofRussell’sGray’sElegyArgument.InA.IrvineandG.Wedekind(eds)RussellandAnalyticPhilosophy.Vancouver:UniversityofBritishColumbiaPress.Sainsbury,R.M.1979:Russell.London:RoutledgeandKeganPaul.——1986:RussellonAcquaintance.InG.Vesey(ed.)PhilosophersAncientandModern.Cambridge:CambridgeUniversityPress.——1993:RussellonNamesandCommunication.InA.IrvineandG.Wedekind(eds)RussellandAnalyticPhilosophy.Vancouver:UniversityofBritishColumbiaPress.Salmon,N.1986:Frege’sPuzzle.Cambridge,MA:MITPress.WadeSavage,C.andAnderson,A.(eds)1989:RereadingRussell:EssaysinBertrandRussell’sMetaphysicsandEpistemology.MinnesotaStudiesinthePhilosophyofScience.Minneapolis:UniversityofMinnesotaPress.Wiggins,D.1976:Frege’sProblemoftheMorningStarandtheEveningStar.InM.Schirn(ed.)StudiesonFrege.Stuttgart:FreidrichFrommannVerlag.DiscussionQuestions1Whatfeaturesmakeapropositionsomethingotherthanamerelist?2DiscussthefollowingobjectiontotheclaimthatFrege’snotionofaconceptcanbeusedtoexplaintheunityoftheproposition:Thereisnoaccountofincompletenessexceptintermsoftheunityofthepropo-sition:incompletenessisjustwhatgeneratestheunity.Soitcannotexplaintheunity.3Canexpressionsotherthannameshavesenseandreference?4Whydoessensematterforcognitivevalueandreferencematterfortruthandfalsity?5Ifapartofasentencelacksreference,doesthewholesentencelacktruthandfalsity?6Fregeheldthatfictionalsentenceswereneithertruenorfalse.Russell,bycontrast,heldthatfictionalsentenceswerefalse.Athirdviewisthatsomefictionalsentencesaretrue(like‘Hamletwasaprince’)andothersfalse(like‘Hamletwasdecisive’).Whatreasonsaretheretochoosebetweentheseviews?7Howdoweexplainwhythereispointtosaying‘HesperusisPhosphorus’,butnopointinsaying‘HesperusisHesperus’?8ShowthatFrege’scompositionalityprincipleforreferenceentailsaversionofLeibniz’slawoftheindiscernibilityofidenticals(ifx=y,anythingtrueofxistrueofy).803\nR.M.SAINSBURY9Discussthefollowingargument:Frege’stheoryusestwobasicnotions(senseandreference),Russell’sonlyone(standingfor);soRussell’stheoryshouldbepreferred.10AccordingtoRussell,ordinarynamesareintelligiblethroughtheirassociationwithdescriptions.Thisviewinvolvesregardinganexpressionlike‘Vulcan’asnotgen-uinelysimple,becauseitisassociatedwithadescription,whichiscomplex.Soyoucannottellbylookingwhetheranexpressionissimpleornot.WhattestforgenuinesimplicitycanRusselluse?11Arethereanylogicallypropernames?Woulditmatteriftherewerenone?12DiscussthefollowingobjectiontotheargumentthatFregeshoulddenythatpropernamesinordinarylanguagehavesensebecausetheyareassociatedwithvariableand,hence,non-publicdefinitedescriptions:Tobepublicisonething,tobeinvariantacrossawholecommunityisanother.Fregecouldwellholdthatsensevariesfrompersontoperson,withoutprejudicetothepublicityofsense,providedheholdsthatsuchanidiosyncraticsenseispublicinthesenseofbeinginprinciplegraspablebyanyone(evenallthoseotherpeoplewhoinfactdonotgraspit).Bycontrast,anideaofmineisprivateinthatitcannoteveninprinciplebehadbyyou.13Canwedistinguishbetweenthoughtsthatareobjectivelogicalentitiesandideasthataresubjectivepsychologicalentities?14WhatarethephilosophicalconsequencesofadoptingRussell’stheoryofdefinitedescriptions?15Howdologicalconstantshavemeaning?16Whatcanbesubstitutedsalvaveritateinindirectcontexts?Whatarethephilosophicalimplicationsofyouranswer?17CanFregerespondadequatelytotheobjectionthatonhistheory,onewhoreportsabeliefreferstothingsotherthanthosethatthebeliefitselfrefersto?(Youneedfirsttoamplifytheobjection.)804\n38MooreTHOMASBALDWING.E.Moore(1873–1958)wasoneofthemostinfluentialBritishphilosophersofthetwentiethcentury.AtanearlystageinhiscareerheestablishedhisreputationbyhiscriticalrejectionoftheidealistphilosophythatwasdominantinBritainattheendofthenineteenthcentury.Healsorejectedtheempiricistphilosophythatappearedtobethealternativetoidealism,maintainingthatitsapplicationinethicsleadstoafallacy–the‘naturalisticfallacy’.InplaceofbothidealismandempiricismMooreadvancedanextremeformofrealismwhichcentredonthehypothesisofadomainof‘propositions’asbothobjectsofthoughtandpossiblestatesofaffairs.InhislateryearsMoorerefinedhistechniquesof‘philosophicalanalysis’,combiningthemwithasustaineddefenceof‘thecommonsenseviewoftheworld’againsttheassaultsofscepticalandcriticalphilosophers.HespentmostofhiscareerinCambridge,wherehewasafriendandcolleagueofRUSSELL(chapter37)andWITTGENSTEIN(chapter39).1EthicalTheoryMoore’sfirstwritingsconcernedthefoundationsofETHICS(chapter6),inparticular,KANT’s(chapter32)ethicaltheory.MoorearguedthatKant(1724–1804)waswrongtosupposethatthefundamentalprinciplesofethicshaveanyspecialconnectionwithhumanreasonortheWILL(p.734).Instead,theyconcernquestionsofvaluewhichareasobjectiveandmind-independentasanyordinarymatteroffact.Generalizingfromthisrejectionofanidealistfoundationforethics,Mooreisequallycriticalofempiricisttheorieswhichsuggestthatquestionsofvaluecanberesolvedbyreferenceto‘nature’,includingHUMANNATURE(chapter35).Inoppositiontobothidealismandempiricism,therefore,MooreproposedinhisfamousethicaltreatisePrincipiaEthica(Moore1903)thatthefundamentaltruthsofethicshaveastatuscomparabletothetruthsofarith-metic:theyareabstractnecessarytruthsconcerningtheintrinsicvalueofstatesofaffairs.Insupportofthisthesis,hemaintainedthatthereisafallacy,the‘naturalisticfallacy’,inalltheorieswhichseektoresolvequestionsofvaluebydefiningvaluesintermsoffurtherfacts,suchasfactsconcerninghumannature.Hismainargumentforthischargewasthatanysupposeddefinitionofgoodness,whichhetooktobethe\nTHOMASBALDWINfundamentalvalue,canbeseentobeincorrectwhenweseethatitincorporatesasubstantiveethicalthesis.Forexample,supposeoneweretodefinegoodnessasthatwhichwedesiretodesire:then,Mooreargues,wewouldhavetoacceptthatwhateveronedesirestodesireisipsofactogood.Yetwecannotbutobservethatitisasubstantialquestionwhetherthisisthecase,andthusthattheproposeddefinitionisincorrect.Thisargumentisinmanywaysproblematic.Mooreappearstoassumethatdefinitionsmustbeobvious,iftrue;yetthisisnotitselfobvious,asMoorerecognizedinhislaterwritings.Furthermore,becauseMoore’sargumentisonlyanappealtothewayinwhichwethinkaboutquestionsofvalue,itisvulnerabletoscepticswhoofferanalternativesetofethicalconceptswithoutclaimingthattheirnewconceptscapturethewholecontentofourordinaryones.Fortheycanarguethatevenifourordinaryconceptsinvolvesomepurelyabstractconceptionofvalue,thensomuchtheworseforthem;wewouldbebetteroffwithoutthem.HenceMoore’spositionneedstobeembeddedinabroaderMETAPHYSICS(chapter2)whichelucidatesethicalconceptsandexplainswhytheyareindefinable.MoorehimselfoffersaPLATONIST(chapter23)metaphysicsofgoodnessasasimple,non-natural,propertywhoseapriorirelationshipstothenaturalpropertiesofthingsweareabletodiscernthroughreflectiveintuitions.Butthishaslittlepersuasivepowerforthosenotantecedentlyfavourabletoit,andmanyofthosewhoacceptedhisclaimthatvaluesarenotdefinableinnon-evaluativetermssoughttojustifyitbyarguingthatthedistinctivecontentofethicalstatementsarisesfromthefactthatthesestatementshaveaspecialroleasexpressionsofemotion,orasrecommendationsandprescriptions.Fromthislatterperspective,Mooreseemedrightinthinkingthatethicalconceptsarenotdefinableintermsofnon-ethicalones,butmistakeninconcludingthatethicalvaluesconstituteasuigenerisdomainofbeing.This‘linguistic’reformulationofMoore’scritiqueofethicalnaturalismnolongerfindsmuchfavour,becausetheimpliedconceptionofethicalthoughtasjustaspecialtypeofnon-descriptivediscourseappearsrathershallow.Atthesametime,versionsofethicalnaturalismwhichlinkhumanvaluestohumannaturehavereturnedtofavour.SuchpositionsconflictwithMoore’stheory,butbecauseMooredoesnotconsiderthehypothesisthatgoodnessisbothindefinableandnatural,hisdiscussiondoesnotapplytonaturalistictheoriesthatdonotofferareductivedefinitionofvalue.Sincetheoriesofthiskindimplythatethicalvalueshaveanirreduciblerolewithinhumanpsychologyasreasonsforaction,thedebateaboutMoore’streatmentofethicalnaturalismconnectswithdebatesincurrentPHILOSOPHYOFMIND(chapter5)concern-ingreductionismandnaturalisminpsychology.Moore’sethicaltheoryisnowstudiedprimarilybecauseofhisdiscussionofnaturalism;butitisalsofamousasaversionofidealutilitarianism,accordingtowhichtheidealendofactionisthemaximization,notofpleasure,butofgoodness–wheregoodness,thoughindefinable,issaidtobefoundespeciallyintheenjoymentofArtandLove.Moore’sstraightforwardcommendationofthesevalues,whicharecelebratedwithoutanyoftheweightymetaphysicalhypothesesconcerningtheirsignificancethatarecharacteristicofRomanticthought,foundareadyaudienceamonghisyoungerfriends.Inparticular,themembersofthe‘BloomsburyGroup’(J.MaynardKeynes,LeonardWoolf,CliveBell,LyttonStracheyandothers)acknowledgedtheirintellectualdebttoMoore,takingfromhimthethesisthatitisonlycertainrefinedstatesofconsciousnessthathavepositiveintrinsicvalue.806\nMOORE2TheRejectionofIdealismJustasMooredevelopedhisethicalvalueinthelightofhiscriticismsofidealisttheory,inotherearlypapershedevelopedhisconceptionofa‘proposition’,astheobjectofthought,inthecourseofhiscriticismsofidealistmetaphysics.Tobeginwith,MoorearguesagainstBERKELEY(chapter30)thatitisamistaketotreatthe‘object’ofperceptionasifitweremerelyasubjective‘content’or‘idea’;instead,heargues,experienceisessentiallyawarenessofthingswhoseexistenceisindependentofus.Hethengeneralizesthispointtoapplytothoughtaswellasperception,arguingthateventhetruthofaprioripropositionsdependsuponusnomorethandoestheexistenceoftheobjectsofexperience.Sincehetakesitthattruepropositionscoincidewiththestatesofaffairsthatobtainjustwhereapropositionistrue,hetakesitthatpropositions,theobjectsofthought,arestructuresofobjectsandpropertieswhoseexistenceisindependentbothoftheirtruthandofourthoughtofthem.TheresultingTHEORYOFJUDGEMENT(p.796)ismanifestlyincompatiblewithidealistmetaphysics;and,justlikeMoore’sethicaltheory,itisequallyincompatiblewithanempiricisttheoryofjudgement.Yetfairlysoonhecametofeelthatthisextremerealisttheorywassimplynotcredible.Inhiswritingsfrom1911onwards,heabandonedbeliefintheexistenceofpropositionsbecausehecouldnolongeracceptthatrealityincludedfalseproposi-tionsaswellastrueones.Helikewiseabandonedhisunquestioningrealismconcern-ingtheobjectsofexperience,andinclinedtowardstheviewthattheobjectsofsenseexperience,whichhecalled‘sensedata’,arenon-physicalrepresentationsofthephys-icalworld.Finally,inthelightofthislastpoint,heacknowledgedthatitisagooddealmoredifficulttofindfaultwithtraditionalscepticalargumentsconcerningknowledgeoftheexternalworldthanhehadearliersupposed.3DefendingCommonSenseTheseproblemslargelysettheagendaofMoore’slaterwritings.InordertohandleSCEPTICALARGUMENTS(pp.45–6)Mooreintroducedhisappealtocommonsense.Thisappealconcernsthestatusofphilosophicalarguments.Mooremaintainedthatthegeneralprincipleswhichphilosophersemployareanswerabletoourcommon-sensejudgementsconcerningparticularmattersoffact.Hencescepticalphilosophersunderminethereasonsonemighthaveforacceptingtheirgeneralprinciplesiftheydisputeourcommon-sensejudgements.Wearemorecertainthatweknowsuchthingsasthat‘Thisisapencil’thanwecanbeofthepremisesofanyscepticalargumentthatcastsdoubtonsuchknowledge.Moore’scriticsmaywellwanttourgethatthisistoosimple,inthatscepticalargumentstypicallyexploittensionsalreadyinherentinourordinarycognitivepracticesandMoorehimselfcametorecognizethatmorehadtobesaid.Inhisfamouslectureof1939,‘ProofofanExternalWorld’(Moore1959),hearguedthat,contrarytoidealistclaims,onecouldprovetheexistenceofanexternalworldsimplybydemonstratingtheexistenceofone’shands.ButMooreatthistimeexplicitlydeniedthatonecouldsimilarlyprovetheexistenceofknowledgeofanexter-nalworld;inordertoeffectsuchaproof,hemaintained,onehadtorefutesceptical807\nTHOMASBALDWINarguments,whichonecouldnotaccomplishbywavingone’shandsatthem.Moorethenattemptedthistaskofrefutation,buthadtoacknowledgebytheendofitthathewasunsuccessful.Onereasonforthis,Ithink,isthathelackedtheconfidencetorethinkthefullimplicationsforEPISTEMOLOGY(chapter1)ofhisappealtocommonsense.ItwasthereforelefttoWITTGENSTEIN(chapter39)totakeoverthistask,whereMoorehadleftoff,inhisnotesOnCertainty(Wittgenstein1969)whicharoseoutofreflectionspromptedbyMoore’swritings.4PhilosophicalAnalysisMoore’sdiscussionsofhisproblemsconcerningtheobjectsofthought(propositions)andtheobjectsofexperience(sensedata)makeessentialuseofhisothercharacteris-ticphilosophicalmethod–philosophicalanalysis.Aswesaw,Moore’searlytheoryofjudgementinvolvedthehypothesisofpropositionsascomplexstructuresinvolvingobjectsandpropertieswhichcoincidewithactualstateofaffairswhenapropositionistrue.Sincethephilosophicalanalysisofapropositionwassupposedtorevealthestruc-tureofitsconstituentobjectsandproperties,itfollowedthatphilosophicalanalysishadanimportantMETAPHYSICAL(chapter2)role.Thisrolewas,however,greatlyenhancedasMooreappreciatedthesignificanceofRUSSELL’s(chapter37)workinLOGICALTHEORY(chapter4).ForhesawthatRussell’sconceptionofalogicalfictionenablesonetotalkasiftherewereentitiesofacertaintype,whilepresentingananalysisofthemeaningofwhatistherebysaidinwhichthereisnoreferencetosuchentities.Mooreusedthisideainhisdiscussionsofpropositions.Foralthoughhestandardlyrepresentedhimselfasconcernedwiththetruthof‘propositions’,hedidnotwantthistalktobetakenascarryingwithitacommitmenttotheexistenceofpropositionsasgenuineentitiestowhichreferenceismadeintheanalysisofwhatismeantbysentencesinwhichweattributethoughtstoeachother.Whetherhispositiononthismatterwasaltogetherconsistentcanbedoubted.Moore’sParadoxThereissomethingveryoddaboutthesentence‘Itisraining,butIdonotbelievethatitis’.Thereisnoexplicitcontradictionhere;andindeed,whenutteredinasuitablecontext,thesentencemightwellbetrue.Yetifweimagineourselveshearingsomeoneutterit,wewouldimmediatelyfeelthattheirutterancewasabsurd.Thesamesituationarisesforanysentenceoftheform‘pbutIdonotbelievethatp’anditischaracteristicofMoore’sanalyticacumenthatheshouldhaveidentifiedthispuzzle,whichWittgensteincalled‘Moore’sParadox’.Mooreproposedthatthesenseofabsurdityherearisesfromaconflictbetweenwhatspeakersassertbytheirutterance,andwhattheyimplybyit.WhenJonesutters‘Itisraining’heassertsthatitisrainingandtherebyimpliesthathebelievesthatitisraining;whenhethengoesontosay‘butIdonotbelievethatitis’heassertsthathedoesnotbelievethatitisraining.Sowhatheimpliesbythefirstpartofhisutterancecontradictswhatheassertsinthesecondpart.Thetruthorfalsityofhisutterancedependsonlyonwhatheasserts;sohisutterancemightbetrue.Buttheconflictbetweenwhatheimpliesandwhatheassertsnonethelessrenderstheutteranceabsurd.808\nMOOREMoore’sexplanationappearslargelycorrect,althoughitremainsdisputedjustwhataccountshouldbegivenoftheimplicationsoftheutterance.Wittgenstein’sdiscussionsofMoore’sParadox,whichoccurinhislaterwritings,derivedfromhissensitivitytheretowide-rangingdifferencesbetweenfirst-personandthird-persondiscourseandthought,andconnectwithhisearlierviewsabouttheMETAPHYSICALSUBJECT(p.816).Moorealsoappliedhisconceptionofphilosophicalanalysistohisdiscussionsofper-ceptionanditsobjects:heheldthatalthoughweallknowforcertainsuchpropositionsas‘Thisisahand’,theiranalysisisdeeplypuzzling.Hetookitthatthedemonstrative‘This’refersheretoasensedatum,typicallyanobjectofvisualexperience,andthatthereallydifficultquestionconcernstherelationshipbetweensuchasensedatumandone’shand.Moorediscussedseveralalternativeshere,thoughintheend,Ithink,itwasaPHENOMENALIST(pp.53–4)positionthathewasmostinclinedtofavour.IndeedonecanbestappreciatetheroleofanalysisinMoore’sphilosophyfromthefactthathefavouredaphenomenalistanalysisofpropositionssuchas‘Thisisahand’attheverytimethathewasadvancinghisproofofanexternalworldbydemonstratingtheexistenceofhishandstohisaudience.Forthisshowsthattheappealtocommonsensedoesnottellusmuchaboutthepropositionsofwhosetruthitassuresus;wealsorequireananalysisofthesepropositions,andthisanalysiscanbeasdisconcertingasonelikes,aslongasthetruth-valuesofourcommon-sensejudgementsarerespected.ThusMoore’suseofanalysisunderminestheapparentconservatismofhisdefenceofcommonsense.Mooreistodayregardedasaparadigmofan‘analyticphilosopher’.Indeed,hisearlytheoryofjudgementplayedacrucialpartinthedevelopmentofanalyticphilosophy,sinceitimpliedthattheanalysisofpropositionsisofdirectphilosophicalsignificance.ButbecauseMooreremaineddetachedfromtheall-encompassingdoctrinesoflogicalatomismandlogicalpositivism,hisuseofphilosophicalanalysiswasalwayspiecemealandidiosyncratic.Nonetheless,particularlybecausehisworkcombinessophisticatedtechniquesoflogicalanalysiswiththedefenceofcommonsense,itprovidesaclassicexampleof‘analyticphilosophy’.FurtherReadingMoore(1993)isaselectionofwritingsdrawnfromallstagesofMoore’sphilosophicalcareer.Schilpp(1942)isaclassiccollectionofpapersconcernedwithMoore’sphilosophy,whichalsocontainsMoore’sautobiographyandhisreplytohiscritics.Regan(1986)attemptstosetthedevelopmentofMoore’sethicaltheoryinthecontextoftheBloomsburyGroup.Baldwin(1990)isacriticalstudyofallaspectsofMoore’sphilosophy.ReferencesMooreG.E.Moore1903:PrincipiaEthica.Cambridge:CambridgeUniversityPress.RevisededitioneditedbyT.R.Baldwin.Cambridge:CambridgeUniversityPress,1993.——1912:Ethics.London:WilliamsandNorgate.Revisededition,Oxford:OxfordUniversityPress,1966.809\nTHOMASBALDWIN——1922:PhilosophicalStudies.London:Routledge.——1953:SomeMainProblemsofPhilosophy.London:GeorgeAllenandUnwin.——1959:PhilosophicalPapers.London:GeorgeAllenandUnwin.——1986:EarlyPapers(editedbyT.Regan).Philadelphia,PA:TempleUniversityPress.(ThisvolumecontainsMoore’spapersfromtheperiod1898–1903.)——1993:SelectedWritings(editedbyT.R.Baldwin).London:Routledge.(ThisvolumecontainssomeofMoore’searlypapers,selectionsfromMoore1922,1953,1959andfurtherlatewritings.)OtherwritersBaldwin,T.R.1990:G.E.Moore.London:Routledge.Keynes,J.M.1949.MyEarlyBeliefs.InTwoMemoirs.London:Hart-Davies.ReprintedinJ.M.Keynes,CollectedPapers,vol.10.London:Macmillan.Regan,T.1986:Bloomsbury’sProphet.Philadelphia,PA:TempleUniversityPress.Schilpp,P.A.(ed.)1942:ThePhilosophyofG.E.Moore.Evanston,IL:NorthwesternUniversityPress.Thirdedition,LaSalle:OpenCourt,1968.Stroud,B.1984:TheSignificanceofPhilosophicalScepticism.Oxford:ClarendonPress.Wittgenstein,L.1969:OnCertainty.Oxford:Blackwell.DiscussionQuestions1Shouldthedefinitionofafamiliarwordappeartousasobviousifitiscorrect?2Isitamistakeinprincipletosupposethatethicalvaluescanbedefined–forexample,byreferencetoreligiousbeliefs?3Must‘ethicalnaturalism’includeacommitmenttothethesisthatethicalvaluescanbedefined?4Canonerefuteidealismbydemonstratingtheexistenceofapairofhands?5Canonerefutescepticismbydemonstratingtheexistenceofapairofhands?6Whatcouldyoudotovindicateyourknowledgethatyouhavetwohandsinthefaceofscepticaldoubttothecontrary?7Isthoughtarelationshipwithaspecialkindofobject–a‘proposition’?Ifso,whatarefalsepropositions?Ifnot,whataccountshouldonegiveofthecontentsofthought?8WhenMacbethhasavisionasofadagger,doesheseeanything?Ifso,what?9Doestheterm‘sensedatum’haveanyworthwhileuseinphilosophy?10Isanalysisthemethodofphilosophy?810\n39WittgensteinDAVIDPEARSWittgensteinhasbeenadominantfigureintwentieth-centuryphilosophy,histhoughtalteringthetermsofphilosophicaldebateandchallengingtraditionalconceptionsofphilosophy.Hisearlywork,culminatingintheTractatusLogico-Philosophicus,dealtwithquestionsraisedbyFREGEANDRUSSELL(chapter37)butalsopursuedhisowndistinctiveagendaconcerninglogic,languageandtheirimplicationsforphilosophy.AtransitionperiodbeganwithdissatisfactionovertheTractatusandledtohissecondmajorwork,theposthumousPhilosophicalInvestigations.BoththeTractatusandtheInvestigationshavefascinatedreadersbytheirunusual,butdifferingstylesaswellasbytheirphilosophicalcontent.ThischapterconcentratesonWittgenstein’searlydiscussionsoflogicalnecessityandfactualpropositions,histransitionalaccountsofmeaningandtheself,andhislaterexaminationsofthepossibilityofaprivatelanguage,thenatureofrulesandthefoundationsofmathematics.Italsoconsidersthedevelopmentofhisviewsonphilosophicalmethod.ReadersmightwishtoreadchaptersonPHILOSOPHYOFLOGIC(chapter4),PHILOSOPHYOFLANGUAGE(chapter3),PHILOSOPHYOFMATHEMATICS(chapter11),DESCARTESANDMALEBRANCHE(chapter26),SPINOZAANDLEIBNIZ(chapter27),KANT(chapter32),FREGEANDRUSSELL(chapter37)andMOORE(chapter38).1LifeLudwigWittgensteinwasborninViennain1889anddiedinCambridgein1951.Hestudiedengineering,firstinBerlinandtheninManchester,andhesoonbegantoaskhimselfphilosophicalquestionsaboutthefoundationsofmathematics.Whatarenumbers?Whatsortoftruthdoesamathematicalequationpossess?Whatistheforceofproofinpuremathematics?Inordertofindtheanswerstosuchquestions,hewenttoCambridgein1911toworkwithRUSSELL(chapter37),whohadjustproducedincollaborationwithA.N.Whitehead(1861–1947)PrincipiaMathematica(1910–13),amonumentaltreatisewhichbasesmathematicsonlogic.Butonwhatislogicbased?Wittgenstein’sattempttoanswerthisquestionconvincedRussellthathewasagenius.Duringthe1914–18warheservedintheAustrianarmyandinsparemoments\nDAVIDPEARScontinuedtheworkonthefoundationsoflogicwhichhehadbegunin1912.Hiswar-timejournal,Notebooks1914–16(1961),revealsthedevelopmentofhisideasmoreclearlythanthefinalversion,TractatusLogico-Philosophicus,whichhepublishedintheearly1920s.Hethenabandonedphilosophyforalifeofaverydifferentkind.HetaughtinavillageschoolinAustriaandafterthatworkedinthegardenofamonastery.Hereturnedtophilosophyinthelate1920s,drawnbackintoitbydiscussionswithsomeofthemembersoftheViennaCircleandwiththeCambridgephilosopherFrankRamsey(1903–30).TheywantedhimtoexplaintheTractatus,buttheirrequestforelucidationssoonproducedsecondthoughts.WhenhereturnedtoCambridge,hedevelopedadifferentphilosophywhichmadeitsfirstpublicappearancein1953inhisposthumousbook,PhilosophicalInvestigations.Beforeitspublication,directacquaintancewithhisnewideashadbeenconfinedtothosewhoattendedhislecturesandseminarsinCambridge.Inthesecondperiodofhisphilosophy,asinthefirst,hisnotesarethekeytohispub-lishedwork.Writtencontinuouslyfrom1929untilhisdeath,theywouldoccupymanymetresofshelvesiftheywerealleditedandpublishedasbooks.Philosophersusuallyintendwhattheywritetobereadbyothers,butthesenotesareakindofthinkingonpaper.Itistruethatsomeofthesetsareevidentlybeingsteeredtowardspublication,butinmostofthemWittgensteinisfacingproblemsalone.Sincehisdeathmanybookshavebeenextractedfromthismaterial.2TheGeneralCharacterofWittgenstein’sPhilosophyWittgenstein’sphilosophyisdifficulttoplaceinthehistoryofideaslargelybecauseitisanti-theoretical.ItistruethatinhisearlyworkhedidproduceatheoryofLOGIC(chapter4)andLANGUAGE(chapter3),butitwasatheorywhichdemonstrateditsownmeaninglessness.Thatwasaparadoxwhichhepresented,appropriatelyenough,inametaphorborrowedfromtheGreeksceptic,SextusEmpiricus(c.150–c.225):‘Anyonewhounderstandsmeeventuallyrecognizes[mypropositions]asnonsensical,whenhehasusedthem–assteps–toclimbupbeyondthem.(Hemust,sotospeak,throwawaytheladderafterhehasclimbedupit)’(Wittgenstein1922:6.54).After1929hecompletelyavoidedtheorizing.Thetaskofphilosophy,ashenowsawit,wasnevertoexplainbutonlytodescribe.SinceWesternphilosophyhadmainlybeenconceivedasasearchforexplanationsataveryhighlevelofgenerality,hisworkstoodtoonesideofthetradition.WittgensteinwasnotaSCEPTIC(pp.46–56).Thereasonwhyherejectedphilosophi-caltheorizingwasnotthathethoughtittooriskyandliabletoerror,butbecausehebelievedthatitwasthewrongwayforphilosopherstowork.Philosophycouldnot,andshouldnottryto,emulatescience.ThatisapointofaffinitywithKANT(chapter32),butwhileKant’scritiqueresultedinasysteminwhicheachofthemanyformsofhumanexperiencefoundaplace,Wittgensteinattemptednosuchthing.Hismethodwastoleadanyphilosophicaltheorybacktothepointwhereitoriginated,whichmightbesomeverysimpleroutine,observableeveninthelifeofanimalsbutrenderedunintelligiblebythedemandforanintellectualjustification.Oritmightstartfromthe812\nWITTGENSTEIN‘crossingoftwopictures’–forexample,weconstrueSENSATIONS(pp.700–2)asobjectsthatarenotessentiallydependentontheirlinkswiththephysicalworld,andsoweattributetothemabasicindependencemodelledonthebasicindependenceofphysicalobjects.Hisaimwastocurethiskindofillusionbyatherapythatwouldgraduallyleadthesufferertorecognize,andalmosttorecreateitsorigin,andsotoescapefromitsdomination.Philosophersareexpectedtobeabletoabstractthegeneralfromtheparticular,butWittgenstein’sgiftwastheopposite–arareabilitytoseetheparticularinthegeneral.Hecoulddemolishatheorywithafewappropriatecounter-examples.Hismethodwastodescribeaneverydaysituationwhichbringsaphilosophicalspeculationdowntoearth.Whenheusedimagery,itwascarefullychosentorevealthestructureoftheproblemunderexamination.Allthishelpstoexplainwhyhislaterphilosophicalwrit-ingshavebeenreadandappreciatedbypeoplewithverylittlephilosophicaltraining.However,theexplanationofthewideappealofhislaterworkisnotjuststylistic.Heisevidentlytakingapartaphilosophicaltraditionthatgoesbacktoantiquity.Thatisawayoftreatingthepastwhichcanbefoundinmanyotherdisciplinestoday,andevenwhenthescenethatheisdismantlinghasnotbeenpreciselyidentifiedhecanstillbereadwithsympathyandwithintuitiveunderstanding.3Wittgenstein’sEarlyPhilosophyInspiteofthewiderappealofhislaterwork,Wittgensteinwaswithoutadoubtaphilosopher’sphilosopher.IntheTractatushedevelopedatheoryoflanguagethatwasdesignedtoexplainsomethingthatRussellhadleftunexplainedinPrincipiaMathematica,thenatureoflogicalnecessity.Themarginalstatusoftheoriesinhisearlyphilosophydidnotdeflectthistheoryfromitsmaingoal,whichwastoshowthatlogicallynecessarypropositionsareakindofby-productoftheordinaryuseofpropositionstostatefacts.Afactualproposition,accordingtoWittgenstein,istrueorfalsewithnothirdalternative.ForheagreedwithRussell’stheoryofdefinitedescriptions:failureofacomplexreferencesimplymakesapropositionfalse.Soiftwopropositionsarecombinedtoformathird,compound,proposition,itstruthorfalsitywillsimplydependonthetruthorfalsityofitstwocomponents.Nowsupposethatwewanttofindoutifitreallyisacontingent,factualproposition,likeitstwocomponents.Whatwehavetodoistotakethetwocomponentsandrunthroughalltheircombinationsoftruthandfalsity,andwewillfindthattherearethreepossibleoutcomestothistest.Thecompoundpropositionmaybetrueforsomecombinationsofthetruth-valuesofitscomponentsbutfalseforothers,inwhichcaseitisacontingentfactualproposition.Oritmaycomeoutfalseforallcombinations,inwhichcaseitisacontradiction.Or,finally,itmaybeatautology,trueforallcombinations.Contradictionsandtautologiessaynothing.However,theyachievethisdistinctionintwooppositeways,theformerbyexcluding,andthelatterbyallowing,everystateofaffairs.Sopropositionswhosetruthorfalsityareguaranteedbylogicalnecessityarelimitingcases,extremedevelopmentsoftheessentialnatureoffactualpropositions.ThisexplanationoflogicalnecessityansweredaquestionthattheaxiomatizationoflogicbyFregeandRussellhadleftunanswered:whyshouldweaccepttheaxiomsand813\nDAVIDPEARSrulesofinferencewithwhichsuchcalculistart?Infact,ifeverypropositioncanbetestedindependentlyforlogicalnecessity,theredoesnotseemtobeanyneedtopromotesomeofthemasaxiomsandtodeducetheothersfromthem.Schopenhauer(1788–1860)hadtakenthesameviewoftheaxiomatizationofgeometry,whichseemedtohimtoberenderedsuperfluousbyspatialintuition.Wittgenstein’searlyaccountofthefoundationsoflogicreliesonsemanticinsight.Ifaformulaislogicallynecessary,wecanseethatitis.Thereisnoneedtoproveitsstatusfromaxioms,becauseatruth-functionalanalysiswillrevealit.Ifonefactualpropositionfollowsfromanother,wecanseethatitdoes,andthereisnoneedtodemonstratethatitdoesbyaproofthatstartsfromalogicallynecessaryformula.Thetwopropositionswillsimplyshowthattheirsensesareconnectedinawaythatvalidatesourinference.Thistreatmentoflogicleavesproofwithoutanyobviousutility–aproblemthatWittgensteintookuplaterinhisphilosophyofmathematics.PropositionsasPicturesWittgenstein’searlytheoryoflanguagewasalsodevelopedinanotherdirection.Ifitthrewlightonthefoundationsoflogic,itoughtalsotothrowlightonthestructureofordinaryfactualdiscourse.Inordertounderstandthisdevelopment,wehavetogobackonestepandaskwhyhethoughtthatfactualpropositionsmustbetrueorfalsewithnothirdalternative.Hisreasonwasthathetookthemtobeakindofpicture.Ifthepointsonthecanvasofalandscape-painterwerenotcorrelatedwithpointsinspace,nopicturethathepaintedwouldsucceedinsayinganything.Similarly,ifthewordsinafactualpropositionwerenotcorrelatedwiththings,nosentenceconstructedoutofthemwouldsayanything.Inbothcasesaliketheconstructionswouldlacksense.But,giventhenecessarycorrelations,thepaintingandthepropositionhavesenseandwhattheysaycanonlybetrueorfalse.Nowthisrunsupagainstanobviousobjection.Manywordsdesignatecomplexthings,whichdonothavetoexistinorderthatthepropositionscontainingthemshouldmakesense,andthiscastsdoubtontheanalogybetweenpointsonacanvasandwords.Theobviousresponsewouldbetoclaimthatsuchwordsarecomplexandthatthesimplerwordsoutofwhichtheyarecompoundeddohavetodesignatethings.Wittgensteinwentfurtherandarguedthatitmustbepossibletocontinuethiskindofanalysistoapointatwhichnomoresubdivisionwouldbepossible.Hisargumentforthisextremeversionoflogicalatomismhadnothingtodowithempiricism.Whathearguedwasthat,ifanalysisstoppedshortofthatterminus,thesenseofapropositioncontainingawordwhichdesignatedsomethingcomplexwoulddependonthetruthofafurtherpro-position.Thisfurtherpropositionwouldsaythatthingshadbeencombinedtoformthecomplexbutitwouldnotbepartofthesenseoftheoriginalproposition.That,heargued,wasanunacceptableresultbothforpicturesandpropositions.Theirsensesmustbecomplete,self-containedandindependentofoneanother.ThepictorialcharacterofpropositionsisathemewithmanydevelopmentsintheTractatus.Notonlyisitusedtoexplainthefoundationsoflogicandtheinternalstructureoffactuallanguage;italsohasimplicationsforSCIENTIFICTHEORIES(chapter9),ityieldsatreatmentoftheselfwhichcarriestheideasofHUME(chapter31),Kant814\nWITTGENSTEINandSchopenhaueronestagefurther,anditallowsWittgensteintodemonstratethatfactuallanguagehasdefinitelimits.WittgensteinbelievedthatETHICAL(chapter6),AESTHETIC(chapter7)andRELIGIOUS(chapter15)discourseliebeyondthoselimits.Thishasstartedalong-runningcon-troversyabouttheimplicationsofhisplacingofthe‘softer’kindsofdiscourse.Washeapositivist,asthephilosophersoftheViennaCirclelaterassume?OrcouldtheoppositeconclusionbedrawnfromwhathesaidabouttheTractatusinalettertoL.Ficker:‘Thebook’spointisanethicalone...Myworkconsistsoftwoparts:theonepresentedhereplusallthatIhavenotwritten.Anditispreciselythissecondpartthatistheimportantone.Mybookdrawslimitstothesphereoftheethicalfromtheinside,asitwere’–thatis,frominsidefactuallanguage.Ifthisepigramistakenliterally,itappearstoputalowvalueonhissolutionstotheproblemsoflanguageandlogic.However,nearlyallhislaterworkisconcernedwiththesameproblems,andthereisreallynoneedtosubjecttheTractatustosuchasimpledilemma.ItspointisKantian:ethics(andthewholesoftersideofdiscourse,includingphilosophyitself)mustnotbeassimilatedtoscience.4TransitionThefirstmodificationofthesystemoftheTractatusappearedin1929inWittgenstein’sarticle,‘SomeRemarksonLogicalForm’(theonlyotherpieceofworkthathepublished–everythingelseisposthumous).Henolongerbelievedintheextremeversionoflogicalatomismforwhichhehadarguedin1922.Therequirement,thatelementarypro-positionsbelogicallyindependentofoneanother,nowstruckhimasexcessive.Thereasonforhischangeofmindwassimple:singularfactualpropositionsalwayscontainpredicatesbelongingtorangesofcontraries.Socolour-predicatesareincompatiblewithoneanotherandthereisnohopeofanalysingthemintosimplerpredicatesthatwouldnotbeincompatiblewithoneanother.Position,length,velocityand,ingeneral,allmeasurableproperties,showthesamerecalcitrancetotheanalysisrequiredbytheTractatus.Hethereforedroppedtherequirement.Therearetwothingsthatmakethischangeofmindimportant.Firstly,thoughtheTractatuscontainsanatomistictheoryoflanguage,therearepassagesthatrevealanunderlyingholism.Forexample,hesays:‘Apropositioncandetermineonlyoneplaceinlogicalspace:neverthelessthewholeoflogicalspacemustalreadybegivenwithit’(Wittgenstein1922:3.42).Butthenewviewofelementarypropositionsisanopenmovetowardsholism.Whatisnowsaidtobe‘laidagainstrealitylikearuler’(ibid.:2.1512)isnotasingle,independent,elementaryproposition,but,rather,asetoflogicallyincompatibleelementarypropositions.Forwhenonepredicateinagroupofcontrariesisascribedtoathing,theothersarenecessarilywithheld.ItisplausibletoregardthisholismasthenaturaltendencyofWittgenstein’smind,andtheatomismassomethingthathetookoverfromRussellandeventuallyrepudiated.Thechangeofmindalsohasamoregeneralimportance.TheatomismoftheTractatuswasofferednotasatheorythatwassupposedtofittheobservablesurfaceoffactuallanguage,butasatheoreticaldeductionaboutitsdeepstructure.Indeed,Wittgensteinwassoconfidentofthevalidityofthedeductionthathewasnotworriedbyhisinabilitytoproduceasingleexampleofalogicallyindependentelementary815\nDAVIDPEARSproposition.Thisdogmatismevaporatedwhenitoccurredtohimthatthelogicalstructureoflanguagemightbevisibleonitssurfaceandmightactuallybegatheredfromtheordinaryusesthatwemakeofwordsinordinarysituations.Thiswasthepointofdepartureofhislaterphilosophy.5HisLaterPhilosophy:TheBlueBookThemostaccessibleexpositionoftheleadingideasofWittgenstein’slaterphilosophyistobefoundintheBlueBook(1958),asetoflecture-notesthathedictatedtohisCambridgepupilsin1933–4.Whathethendidwassofaroutoflinewiththetraditionthatwemayatfirstfeelinclinedtoquestionwhetheritreallyisphilosophy.Hisanswerwasthatitmightbecalled‘oneoftheheirsofthesubjectwhichusedtobecalled“philosophy”’.Theconspicuousnoveltyistheabsoluterefusaltoforceallthemultifariousvarietyofthoughtandlanguageintothemouldofasingletheory.Hecriticizedthe‘contemptfortheparticularcase’thatanysuchattemptwouldinvolve,andhesystematicallyrepressedthecravingforgeneralitythathascharacterizedWesternphilosophysinceSocratesfirstinstigatedthesearchfortheessencesofthings.Wemay,ofcourse,asksuchquestionsas‘Whatisknowledge?’,butwemustnotexpecttofindtheanswerwrappedupintheneatpackageofadefinition.Therewillbemanydifferentcases,andthoughtheywillshowafamilyresemblancetooneanother,theywillnotbelinkedbythepossessionofasinglesetofcommonproperties.Socratesaskedforaconjunctionofproperties,butwemustbecontentwithadisjunction.ThatisanaccurateplacingofWittgenstein’snewphilosophyinthehistoryofideas,butitleavesanimportantquestionunanswered.Whyshouldacatalogueofexamplesberegardedasasolutiontoaphilosophicalproblem?Isitnotjustacollectionofthekindofmaterialthatposestheproblem?Thepointofthesequestionsisthat‘theheirtophilosophy’needstobesomethingmorethanwell-documentednegativeadvicenottotheorize:itoughttoteachustoseephilosophicalproblemsfromtheinsideandtofindamorepositivewayoflayingthemtorest.Thereare,infact,twodiscussionsintheBlueBookthatdemonstratethatWittgenstein’slaterworkwasapositivecontinuationofthephilosophyofthepast.OneisthelonginvestigationofMEANING(chapter3)andtheotheristhetreatmentoftheSELF(pp.732–3).Bothareveryilluminating.ThediscussionofmeaningisadevelopmentofapointmadeintheTractatus:‘Inordertorecognizeasymbolbyitssign,wemustobservehowitisusedwithasense’(Wittgenstein1922:3.326).Thisremarkconsortsuneasilywiththepicturetheoryofpropositions,whichderivesmeaningfromtheoriginalactofcorrelatingnamewithobject.Thetheoryimpliesthatmeaningisrigid,becauseitisbasedonasingle,self-containedconnectionwhich,oncemade,remainsauthoritative,withoutanyneedforinterpretationoranypossibilityofrevision.Theremarkpointsthewaytoamoreflexibleaccountofmeaningwhichwillaccommodateallthedifferentusesthatwemakeofwordsandleavesroomforplasticity.Thisisthedifferencebetweentreatinglanguageasafossilandtreatingitasalivingorganism.ThediscussionofmeaningintheBlueBookdevelopstheisolatedremarkintheTractatusandcriticizestherigidityofthetheoryofferedelsewhereinthebook.816\nWITTGENSTEINOstensivedefinition,whichwassupposedtoattachawordtoitsobject,isshownupasaveryinscrutableperformance,compatiblewithmanydifferentinterpretationsofaword’smeaning:theunderlyingassimilationofalldescriptivewordstonamesdesig-natingobjectsisrejected;andsotooistheassumptionthatthemeaningofawordissomethingthatbelongstoitintrinsically,and,therefore,independentlyofitsuse.Thislastpointprovedtobeimportant.Forifmeaningneverbelongstoawordintrinsically,itwillneverbepossibletoexplaintheregularityofaperson’suseofawordbycitingtherulethatheorsheisfollowing.Forthemeaningsofthewordsinwhichtheruleisexpressedwillthemselvesneedtobeinterpreted.ThislineofthoughtisdevelopedinPhilosophicalInvestigations(1953).ThetreatmentoftheselfintheBlueBookisveryclearandstronglyargued.AsintheNotebooksandtheTractatus,itispresentedaspartofanexaminationofsolipsism,butitismucheasiertodiscernthestructureofthelaterversionoftheargument.Thecentralpointisthatthesolipsist’sclaim,‘OnlywhatIseeexists’,isnotwhatitseemstobe.Thesolipsistseemstobereferringtohimselfasaperson,butreallyheorsheisusingthepronoun‘I’torefertosomethingentirelyabstractwhichisintroducedmerelyas‘thesubjectwhichislivingthismentallife’or‘thesubjectwhichishavingthesevisualimpressions’.Butifthesubjectisnotgivenanyindependentcriterionofiden-tity,thereisnopointfromwhichthereferenceto‘theseimpressions’canbemade.Thesolipsistconstructssomethingwhichlookslikeaclock,exceptthathepinsthehandtothedial,sothattheybothgoroundtogether.Sayingwhatexists,liketellingthetime,mustbeadiscriminatingperformance.Theideathatthesubjectisavanishingpoint,whichwasdevelopedbyHume,KantandSchopenhauer,ishereputtoanewuse.6PhilosophicalInvestigations:ThePrivateLanguageArgumentTheso-called‘privatelanguageargument’ofPhilosophicalInvestigationsiscloselyrelatedtotherejectionofasolipsismwhichisbasedonanegowithoutacriterionofidentity.WhatthetwocritiqueshaveincommonisarequirementwhichwaslaterexpressedveryconciselybyW.V.O.Quine,‘Noentitywithoutidentity’.Thesolipsist’segolacksanycriterionofpersonalidentity,andsimilarly,ifthequalityofasensoryexperiencewerecompletelydisconnectedfromeverythinginthephysicalworld–notonlyfromanystimulusbutalsofromanyresponse–itwouldlackanycriterionofTYPEIDENTITY(p.178).Theparallelismofthetwocritiquesisverycloseinthelecture-notesinwhichWittgensteinfirstdevelopedtheso-called‘privatelanguageargument’.Againstthesolipsistwhosays,‘ButIaminafavouredposition.Iamthecentreoftheworld’,heobjects,‘supposeIsawmyselfinamirrorsayingthisandpointingtomyself,woulditstillbeallright?’(Notesforlectureson‘PrivateExperience’and‘Sense-data’,1968:299).AgainstthePHENOMENALIST(pp.53–4)whoarguesfordetachedsen-sation-typesandsays,‘Butitseemsasifyouareneglectingsomething’,heobjects,‘WhatmorecanIdothandistinguishthecaseofsaying“Ihavetooth-ache”whenIreallyhavetoothache,andthecaseofsayingthewordswithouthavingtooth-ache?Iamalso(further)readytotalkofanyxbehindmywordssolongasitkeepsitsidentity’(ibid.:297).817\nDAVIDPEARSTheinterpretationofthisimportantlineofthoughtinhislaterworkisdifficult.Theparallelismbetweenthetwocritiquesisalwaysahelpfulclue,athreadwhichwemustneverrelinquishinthelabyrinthofconfusingindications.OnesourceofconfusionishardlyWittgenstein’sfault.Hehimselfneverusedthephrase‘privatelanguageargument’.Itishiscommentatorswhouseitand,bydoingso,theyhavecreatedtheillusionthatasingle,formalargumentoughttobeextractablefromthetextofPhilosophicalInvestigations.Butthatisnotthestructureofhiscritique.Heargueddialectically,andwhenhisadversarytriestointroducethe‘neglectedx’behindthewordsreportingasensation,healwaystriestoshowhisadversarythat,ifthisxisnotcoveredbyordinarycriteriaofidentitybasedonthephysicalworld,itwillnothaveanycriterionofidentityatall.Atthispointitisagoodideatoaskwhohisadversaryis.Evidently,hisadversaryisaphilosopherwhosupportssomekindofsense-datumtheory.Butwhatkind?Onesug-gestionthathasbeenmadeisthatitisthesense-datumtheorywhichhehimselfadoptedintheTractatus(notveryopenly).Sinceheextendedhiscritiquetoincludeothermentalentitiesbesidessensations,thissuggestionhasbeengeneralized,andhehasbeentakentobecriticizingthe‘mentalistictheoryofmeaningoftheTractatus’(Malcolm1986:ch.4).Butabriefreviewofthedevelopmentofhisphilosophyofmindwillshowthatthesesuggestionsaremistaken.AnyonewhocompareswhatWittgensteinsaidaboutsimpleobjectsintheNotebooksandintheTractatuswillseeimmediatelythathewasuncertainoftheircategoryintheformerandinthelatterwasconvincedthathisuncertaintydidnotendangerhislogicalatomism.MaybetheywerematerialparticlesorperhapstheywereRusselliansensedata.Hedidnotcare,becausehisargumentfortheirexistencedidnotdependontheircategoryandhedidnotthinkthatitneededverificationbytheactualdiscoveryofexamples.However,thepossibilitythattheyweresensedatawasworthexploring,espe-ciallyafterhisabandonmentin1929oftheextremeversionoflogicalatomism.Sowhenhereturnedtotheseproblemsafterthelongintervalthatfollowedthepublica-tionoftheTractatus,heworkedouttheconsequencesofidentifyingthemwithsensedata.ThisfittedinverywellwiththeprogrammeoftheViennaCirclephilosophers.Theywereinterestedinthephilosophyofscienceandpredisposedtoacceptasimplestratificationoflanguage,withthephenomenalvocabularyonthebasiclevelandthephysicalvocabularyontheupperlevelandcompleteinter-translatabilitybetweenthetwolevels.Itisnotoriousthatthiskindofphenomenalismlooksimpregnableuntilweexaminetheroutethatledusintoit.ThatiswhatWittgensteindid,andhefoundanobstaclewhichseemedtohimtomaketherouteimpassable.Theoriginalpositionfromwhichitstartedwassupposedtobeoneinwhichpeoplespokeaphenomenallanguagewithavocabularycompletelydisconnectedfromthephysicalworld.Itisobviousthatsuchalanguagewouldbenecessarilyunteachable,andsinceWittgensteinusedtheword‘private’tomean‘necessarilyunteachable’,itwouldbeaprivatelanguage.Whatisnotsoobviousishowhethoughtthathecouldshowthatsuchalanguagewouldbeimpossible.Theprimarytargetofhiscritiqueofprivatelanguageisthesense-datumlanguagethatphenomalistsclaimedthateachofuscouldsetupindependentlyofanythinginthephysicalworldand,therefore,inisolationfromoneanother.Butthescopeofhis818\nWITTGENSTEINattackismuchwider,becauseitwouldshowthatnomentalentityofanykindcouldeverbereportedinsuchalanguage.However,neitherinitsnarrownorinitswidescopeisitdirectedagainstanythingintheTractatus.Forjustastherewasnocommitmenttophenomenalismintheearlywork,sotootherewasnocommitmenttothethesisthatthemeaningofasentenceisderivedfromthemeaningofthethoughtbehindit.IfthecritiqueofprivatelanguageisrelatedtoanythingintheTractatus,itistothecritiqueofego-basedsolipsism,butpositively,asafurtherapplicationofthesamegeneraldemandforacriterionofidentity.ItisnecessarytodistinguishtwomovesthatWittgensteinmadeinhisdialecticalcritiqueofanecessarilyunteachablesensation-language.Thephenomenalistbelievesthatwecansetupthislanguageanduseittoreportoursensedataincom-pleteindependencefromanythinginthephysicalworld.Againstthis,Wittgenstein’sfirstmovewastheonethathasalreadybeendescribed:heaskedforthecriterionofidentityofthesupposedlyindependentsensation-types.Thatisapurelydestructivedemand.Hissecondmovewastopointoutthatareportofasensationwillusuallycontainanexpressionofthesensorytypeandseldomadescriptionofit.Thismovewasthebeginningofareconstructionofthesituation,designedtoleadtoabetteraccountofsensation-language.ThedestructivemoveismademostperspicuouslyinPhilosophicalInvestigations.Supposethatawordforasensation-typehadnolinkswithanythinginthephysicalworldand,therefore,nocriteriathatwouldallowmetoteachanyoneelseitsmeaning.Evenso,Imightthinkthat,whenIappliedittooneofmyownsensations,IwouldknowthatIwasusingitcorrectly.But,accordingtoWittgenstein,thatwouldbeanillusion,becauseinsuchanisolatedsituationIwouldhavenowayofdistinguishingbetweenknowingthatmyuseofthewordwascorrectandmerelythinkingthatIknewthatitwascorrect.Noticethathedidnotsaythatmyclaimwouldbewrong:hispointismoreradical–therewouldbenorightorwronginthiscase(Wittgenstein1953:§258).Thecommonobjectiontothiscriticismisthatitsimplyfailstoallowfortheabilitytorecognizerecurringtypesofthings.This,itissaid,isapurelyintellectualabilityonwhichweallrelyinthephysicalworld.Sowhatistheretostopasinglepersonrelyingonitintheinnerworldofhismind?PerhapsCarnapwasrightwhenhechose‘remem-beredsimilarity’asthefoundationofhisLogicalStructureoftheWorld(1967).HereWittgenstein’ssecondmoveisneeded.Iftheabilitytorecognizetypesreallywerepurelyintellectual,itmightbeusedinthewayinwhichCarnapandothershaveusedit,anditmightbepossibletodismissWittgenstein’sobjectionbysaying,‘Wehavetostopsomewhereandwehavetotreatsomethingasfundamental–sowhynotourabilitytorecognizesensation-types?’ButagainstthisWittgensteinarguesthatwhatlookslikeapurelyintellectualabilityisreallybasedonnaturalsequencesofpredicament,behaviourandachievementinthephysicalworld.Painmayseemtobeaclearexampleofasensation-typewhichisindependentlyrecognizable,butthewordisreallyonlyasubstituteforthecrywhichisanaturalexpressionofthesensation(Wittgenstein1953:§§244–6).Or,totakeanotherexample,ourabilitytorecognizelocationsinourvisualfieldsisconnectedwiththesuccessofourmovementsinphysicalspace.Ourdiscriminationsintheinnerworldofthemindare,andmustbe,answerabletotheexigenciesofthephysicalworld.819\nDAVIDPEARSAtthispointwemightbegintoregretWittgenstein’srefusaltotheorize.Ifhehadofferedamoresystematicaccountofthedependenceofoursensorylanguageonthephysicalworld,theso-called‘privatelanguageargument’mighthavecarriedmoreconviction.Infact,manyphilosophershavebeenconvincedbyit,butthereisalargeopposition,containingfewdoubtersandconsistingalmostentirelyofphilosopherswhofeelsurethattheargumentisinvalid.ThedialecticalcharacterofWittgenstein’sargumenthascontributedtothisresult.7PhilosophicalInvestigations:MeaningandRulesAnother,similar-sounding,butinfactverydifferent,questionisdiscussedinPhilosophicalInvestigations.Couldapersonspeakalanguagethatwasneverusedforcommunicationwithanyoneelse?Suchalanguagewouldbeprivateintheordinarysenseofthatword,becauseitwouldbeunshared;butitwouldnotbenecessarilyunteachable,becauseitwouldbealanguagefordescribingthephysicalworld,andsoitwouldnotbeprivateinWittgenstein’ssense.Thequestionisimportant,butwehavetogobacktothetheoryofmeaningoftheTractatusinordertoseewhyitisimportant.Arough,butuseful,distinctioncanbedrawnbetweentwokindsoftheoryofmeaning,therigidandtheplastic.ThetheoryofferedintheTractatusisrigid.Oncenameshavebeenattachedtoobjectseverythingproceedsonfixedlines.Theapplica-tionofthenamesissettledonceandforall,andpropositionsandtruth-functionalcombinationsofpropositions,includingthetwolimitingcases,namelytautologyandcontradiction,allunfoldwithoutanymorehelpfromus.Thetheorydoesnotactuallytreatthemeaningofanameassomethingintrinsictoit,becausewedohavetocorre-latethenamewithanobject.Ifwewantanexampleofatheorythatdoestakethefurtherstepandtreatsmeaningasanintrinsicfeatureofasymbol,thereisthetheorythatamentalimageautomaticallystandsforthingsthatitresembles.Thatillustratestheextremedegreeofrigidity:wewouldhavenooptions.Aplastictheoryofmeaningwouldrejecttheanalogybetweenadescriptivewordandaname,anditwoulddenythatthemeaningofadescriptivewordcanbefixedonceandforallbyostensivedefinition.BoththesemovesaremadeinPhilosophicalInvestigations.Itis,ofcourse,notdeniedthatouruseofadescriptivewordwillexhibitaregularity:whatisdeniedisthatitisaregularitythatisanswerabletoanindependentauthority.Wedohaveoptions.Wemaysay,ifwelike,thatwearefollowingarule,butthatwillnotbeanexplanationoftheregularityofourpractice,becauseitisourpracticethatshowshowweareinterpretingthewordsinwhichtheruleisexpressed.Arethere,then,noconstraints?Istheuseofdescriptivelanguagepureimprovisation?Evidently,theremustbesomelimittoplasticity,andoneobviouspossibilityisthatitisimposedbytheneedtokeepinstepwithotherpeople.Thatishowthequestion‘Doeslanguagerequireexchangesbetweenmembersofacommu-nityofspeakers?’comestobeimportantinPhilosophicalInvestigations.WhenwetracethelineofdevelopmentfromWittgenstein’searlytohislatetheoryofmeaning,weneedtoknowhowfarhemovedinthedirectionofplasticity.Hisfirsttwostepsweretakensoonafterhisreturntophilosophyin1929.Themeaningofasymbolcanneverbeoneofitsintrinsicfeatures,andevenanostensive820\nWITTGENSTEINdefinitioncannotsaddleitwithasingle,definitemeaning,becauseanostensivedefinitionisalwayscompatiblewithmanydifferentsequels.Meaning,then,mustdependonwhatwedonextwithaword–onouruseofit.TheAuthorityofRulesWittgenstein’snextstepinthedevelopmentofhistheoryofmeaningwastoarguebyreductioadabsurdum.Giventhatthemeaningofawordisnevercontainedintheworditself,eitherintrinsicallyorafteranostensivedefinition,itmustbeamistaketoholdthatsomeonewhofollowsalinguisticruleisobeyinganindependent,externalauthority.Foranybasisthatwemightproposefortheso-called‘authority’willalwaysleaveitopenwhatthespeakershoulddonext,and,whatismore,openbetweenmanydifferentalternatives.Ifwetrytoremedythissituationbyofferingamoreexplicitstatementoftherulethatheissupposedtobefollowing,hewillstillbeabletointerpretthatstatementinmanydifferentways.Sowhenwetrytofixtherightuseofawordpurelybypreceptandpastapplications,wefail,becauseweendbyabolishingthedistinctionbetweenrightandwrong.Wefeelthatthedistinctionrequiresarigidexternalauthorityandsoweeliminateallplasticity,but,whenwedothat,wefindwehavelostthedistinction.Itwouldbeabsurdtosuggestthatthemeaningofaparticularinstructionisdeterminedbywhatapersondoeswhenhetriestoobeyit.Iftheinstructiondidnotalreadystandtherecompletewithitsmeaning,therewouldbenoquestionofobedience.Butwhenthesamesuggestionismadeaboutageneralinstruction,orrule,itisnotabsurdbutonlyparadoxical.InordertounderstandWittgenstein’sargument,wehavetoelucidatetheparadox.Itshocksus,becauseindailylifethereisnodoubtaboutwhatcountsasobeyingtheinstruction‘Alwaystakethenextleftturn’.However,thereisanimportantgrainoftruthintheparadox.Forthereasonwhythereisneveranydoubtaboutwhatcountsasobeyingthisinstructionisthatintheordinarycourseofourlivesnobodyeverdoestakeittomeananythingbizarre,like‘Alwaystaketheoppositeturntotheoneyoutooklast’.Butifsomeonedidunderstanditinthiseccentricway,averbalexplanationofwhatitreallymeantmightwellfailtoputhimright.Forhemightgiveourverbalexplanationanequallyeccentricinterpretation.Nowbecausethissortofthingneverhappensinreallife,wefindWittgenstein’sargumentparadoxical.Itsimplydoesnotfitourpictureoftheindependentauthorityofarule.However,hispointisthatitcouldhappen,andthatindi-catessomethingimportant.Itindicatesthatouruseoflanguagetogivegeneralinstruc-tionsandstaterulesdependsonoursharedtendencytofindthesameresponsesnatural.Wehavetoagreeinourpracticebeforerulescanhaveanyindependentauthority.Theindependentauthorityislimitedbytherequirementthatmakesitpossible.Wittgenstein’sargumentisnotconcernedwiththerealpossibilityoflinguisticcrankinessbutwiththelogicalstructureofthesituation.Itappears,then,thattheintellectualperformanceoffollowingalinguisticruleisbasedonsomethingoutsidetherealmoftheintellect.Itsbasisisthefactthatwe,likeotheranimals,finditnaturaltodivideandclassifythingsinthesamewayasothermembersofourspecies.ThislineofthoughtrunsparalleltothelinethatWittgensteintookaboutsensation-language.Fortheretootheintellectualachievementofreportingsensationswasbasedonpre-establishednaturalresponsesandbehaviourofamoreprimitivekind.821\nDAVIDPEARSWhatlimittoplasticitydidWittgensteinrecognize?Didhetreatagreementwiththeresponsesofotherpeopleasanabsoluteconstraint?Anddidherecognizeanyotherabsoluteconstraints?Itisdifficulttoextractdefiniteanswerstothesequestionsfromhislaterwritings.BothinPhilosophicalInvestigationsandinRemarksontheFoundationsofMathematics(1978)hesaysthatagreementinjudgementsisrequiredifpeoplearegoingtocommunicatewithoneanother.Butthatdoesnotruleoutthepossibilitythatawolf-childmightdevelopalanguagesolelyforhisownuse.Itwould,ofcourse,bealanguageofwrittensignswithwhichhewouldcommunicatewithhimselfacrossinter-valsoftime.Butitwouldnotrequiretheco-operationofotherpeople.SoitlooksasifWittgensteinwentnofurtherthanmaintainingthat,ifthereareotherpeoplearoundandifthelanguage-userisgoingtocommunicatewiththem,theplasticityofhislanguagewillbelimitedbytheexigenciesofagreementwiththem.Thissquareswiththefactthatinseveralofhistextsdatingfromthe1930sheallowsthatapersonmightsetupasystemofsignssolelyforhisownuse.However,thoughWittgensteindoesnotdenythisinPhilosophicalInvestigationsorinRemarksontheFoundationsofMathematics,hedoesnotreassertiteither.Thereseemstobesomeambivalence.Inanycase,thereisanother,moreimportantconstraintonplasticitywhichreallyisabsolute.Ifwearegoingtodiscoverregularitiesinnature,ourlanguagemustexerciseacertainself-discipline:itmustfollowroutineswhichallowustocollectevidence,tomakepredictionsand,later,tounderstandthem.Sheerimprovisationwillnotputusinaposi-tiontodothesethings.ThisobviousconstraintisoverlookedbythosewhoattributetoWittgensteina‘communitytheoryoflanguage’.Itishardlylikelythatheoverlookedit.8Wittgenstein’sPhilosophyofMathematicsWittgenstein’sPHILOSOPHYOFMATHEMATICS(chapter11)isanotherdevelopmentofthesamegeneralidea.Forheretooourmostelaborateintellectualconstructionsaresaidtobefoundedonbasicroutinesthatcannotbejustifiedintellectually.Whenwecount,wefeelthatweareusingfootholdsalreadycarvedforusinarock-likereality.Butcounting...isatechniquethatisemployeddailyinthemostvariousoperationsofourlives.Andthatiswhywelearntocountaswedowithendlesspractice,andmercilessexactitude...‘Butisthiscountingonlyause,then?Isn’ttherealsosometruthcorre-spondingtothissequence?’...itcan’tbesaidoftheseriesofnaturalnumbers–thatitistrue,butthatitisuseful,and,aboveall,itisused.(Wittgenstein1978:I,4)Theapplicationofthisideatothephilosophyofmathematicshasprovedlessfruitfulthanitsapplicationtothephilosophyoflanguage.Thereasonforthismayonlybethegreaterdistancebetweenthesuperstructureandtheproposedbasisinthiscase.Orsomewouldarguethatmathematicsisreallynotamenabletothistreatment.Itis,forexample,questionablewhetheritcanyieldaconvincingaccountofproofinmathematics.9ConclusionInevitablypeopleaskwhatmessagecanbeextractedfromWittgenstein’sphilosophy.Ifamessageisatheory,then,aswehaveseen,themessageisthatthereisnomessage.822\nWITTGENSTEINLikeanyotherphilosopher,hepushedthequestforunderstandingbeyondthepointatwhichtheordinarycriteriaforunderstandingaresatisfied.However,unlikeothers,hebelievedthatphilosophicalunderstandingismoreliketheexperienceofajourneythantheattainmentofadestination.HeregardedphilosophyasanactivitythatislikeFreudiantherapy.Youreliveallthetemptationstomisunderstandandyourcurerecapitulatesthestagesbywhichitwasachieved.Soitisnottherestorationofastateofunreflectivehealth.Infact,ratherthanaimingtoreestablishanykindofstate,philosophyisconcernedwiththeprocess.Ifthereisasinglestructurediscernibleinhisphilosophy,itishisrejectionofallillu-sory,independentsupportforourmodesofthought.Rigidtheoriesofmeaningtreatlinguisticrulesasindependentauthoritiestowhichwewhofollowthemaresupposedtobewhollysubservient.Buthearguesthatthisisanillusion,becausethesystemofinstructionandobedienceinvolvesacontributionfromeachindividualandpresup-posesabasiclike-mindedness.Similarly,thenecessityofmathematicsissomethingwhichweprojectfromourpracticeandthenmistakenlyhailasthefoundationofourpractice.Evidently,hewasrejectingrealism,buthistreatmentofrulesshowsthathewasnotrecommendingconventionalisminitsplace.Certainly,hisinvestigationshaveastructure,butitisnotthestructureoftraditionalphilosophy.FurtherReadingThereareseveralwaysinwhichWittgenstein’swritingsmaybeapproachedforthefirsttime.SomeoneunfamiliarwithphilosophymightreadCultureandValue(1977).Thisisacollectionofremarksthatarenotdirectcontributionstophilosophy;theyareaboutart,religion,Wittgenstein’sownlifeandespeciallyaboutwhatitisliketobeaphilosopher.AcloserapproachwouldbetoreadhisCambridgeLectures1930–1932ortheBlueBook(1958),whichwaswrittenin1933–4.IntheformerheexplainstheleadingideasoftheTractatustoanaudienceofCambridgestudents;thelatterdealswiththenatureofhislaterphilosophyandwithtwospecificphilosophicalproblems:meaningandtheself.Bothareeasiertofollowthanhisotherworks,becausetheyarewritteninthecontinuous,consequentialwayinwhichtreatiesareusuallywritten,unlikethemajorpartofhisoeuvre,whichconsistsofindependentremarks,whicharecarefully,butnotalwaysperspicuously,arranged.IntheprefacetoPhilosophicalInvestigationshecompareshisremarkstosketchesmadebyanartistonmanydifferentjourneys,crossingandre-crossingthesameterrain.Thisworkcanbeappreciatedonlybysomeonewhotravelsonthesameroutes,followinginhisfootsteps–thatis,onlybysomeonewhocanreadhisremarksascontributionsofthephilosophicalproblemswithwhichtheyareconcerned.Readingabookofthiskindisnotlikeordinaryreading.Itis,ofcourse,partlyamatteroftakinginwhathewrites,butitisalsomuchmorethanthat.Youhavetostopandthinkafterreadingaremark,andsometimesyouhavetofollowupalongtrainofthoughtbeforegoingontohisnextremark.ButhowcanyougetintoapositiontoreadWittgenstein’stextsinthisway?Thatistheproblem.OnesolutionistouseAWittgensteinWorkbook(compiledbyCoope,Geach,PottsandWhite;Oxford:Blackwell,1970).Thishelpfulcompaniontakes18topicsofcentralimportanceinWittgenstein’sthought,andforeachofthemgivesreferencestohiswritingsandtorelateddiscussionsbyotherphilosophers,andalistofquestionstowhichthereadercanworkoutanswers.AnothersolutionistouseoneofthemanybooksaboutWittgenstein’sphilosophyinasimilarway.Insteadofreadingitthroughfrombeginningtoend,takeachapterdealingwitha823\nDAVIDPEARSparticulartopic,readitandthenlookupallthereferencestoWittgenstein’sworkinthefootnotes.ThechapterwillhelpyoutounderstandWittgenstein’sremarksand,inthereversedirection,theremarkswillserveasacheckontheaccuracyoftheinterpretationofferedinthechapter.Ingeneral,hisearlierphilosophyismoredifficulttopenetratethanhislaterphilosophy,becauseitisdirectlyandimmediatelyrelatedtotheworkofFregeandRussell.ButthoughitiseasiertofindpointsofentryintothedialecticaldiscussionsofPhilosophicalInvestigations,afullunderstandingofthemrequiresfamiliaritywiththetheoriesundercriticism,whethertheyarethoseofotherphilosophersorofWittgenstein’searlierpersona.ReferencesWittgenstein:generalWittgenstein,L.1977:CultureandValue(editedbyG.H.vonWrightandH.Nyman,translatedbyP.Winch).Oxford:Blackwell.Wittgenstein’searlyperiodWittgenstein,L.1922:TractatusLogico-Philosophicus(translatedbyC.K.Ogden).London:Routledge.AlsotranslatedbyD.F.PearsandB.McGuinness,London:Routledge,1961.——1961:Notebooks1914–1916(editedbyG.H.vonWrightandG.E.M.Anscombe,translatedbyG.E.M.Anscombe).Oxford:Blackwell.Wittgenstein’stransitionalperiodWittgenstein,L.1966:SomeRemarksonLogicalForm.ProceedingsoftheAristotelianSociety,supplementaryvol.9.ReprintedinR.BeardandI.Copi(eds)EssaysonWittgenstein’sTractatus.London:RoutledgeandKeganPaul,1966.——1975a:PhilosophicalRemarks(editedbyR.Rhees,translatedbyR.HargreavesandR.White).Oxford:Blackwell.——1979:LudwigWittgensteinandtheViennaCircle:ConversationsRecordedbyFriedrichWaismann(editedbyB.McGuinness,translatedbyJ.SchulteandB.McGuinness).London:Blackwell.——1980:Wittgenstein’sLectures,Cambridge,1930–1932(editedbyD.Lee).Oxford:Blackwell.Wittgenstein’slaterperiodWittgenstein,L.1953:PhilosophicalInvestigations(translatedbyG.E.M.Anscombe).Oxford:Blackwell.——1958:TheBlueandBrownBooks(editedbyR.Rhees).Oxford:Blackwell.——1967:Zettel(editedbyG.E.M.AnscombeandG.H.vonWright,translatedbyG.E.M.Anscombe).Oxford:Blackwell.——1968:NotesforLectureson‘PrivateExperience’and‘Sense-data’.PhilosophicalReview,77.ThenotestakenontheselecturesbyR.RheesarepublishedinthejournalPhilosophicalInvestigations,7,January1984.——1969:OnCertainty(editedbyG.E.M.AnscombeandG.H.vonWright,translatedbyD.PaulandG.E.M.Anscombe).Oxford:Blackwell.——1975b:Wittgenstein’sLecturesontheFoundationsofMathematics(editedbyC.Diamond).Hassocks:HarvesterPress.——1978:RemarksontheFoundationsofMathematics(editedbyG.H.vonWright,R.RheesandG.E.M.Anscombe,translatedbyG.E.M.Anscombe).Oxford:Blackwell.OtherwritersBaker,G.andHacker,P.M.S.1984:Scepticism,RulesandMeaning.Oxford:Blackwell.Bolton,D.E.1979:AnApproachtoWittgenstein’sPhilosophy.London:Macmillan.824\nWITTGENSTEINBudd,M.1989:Wittgenstein’sPhilosophyofMind.London:Routledge.Carnap,R.1967[1928]:LogicalStructureoftheWorld(translatedbyR.A.George).London:Routledge.Coope,C.,Geach,P.,Potts,T.andWhite,R.1970:AWittgensteinWorkbook.Oxford:Blackwell.Kenny,A.1973:Wittgenstein.Harmondsworth:PenguinBooks.Kripke,S.A.1982:WittgensteinonRulesandPrivateLanguage:AnElementaryExposition.Oxford:Blackwell.McGuinness,B.1988:Wittgenstein:ALife:YoungLudwig1889–1921.London:Duckworth.Malcolm,N.1986:NothingisHidden.Oxford:Blackwell.Monk,R.1990:LudwigWittgenstein:TheDutyofGenius.London:JonathanCape.Pears,D.F.1987–8:TheFalsePrison:AStudyoftheDevelopmentofWittgenstein’sPhilosophy(vol.1,1987;vol.2,1988).Oxford:ClarendonPress.Wright,C.1980:WittgensteinontheFoundationsofMathematics.London:Duckworth.DiscussionQuestions1Doeslogichaveabasis?2Whatkindoftheoryisphilosophicaltheory?Istheorizingthewrongwayforphilosopherstowork?3Howcan‘crossingtwopictures’leadtophilosophicalillusions?4Canweacceptatherapeuticmethodofdealingwithphilosophicalproblems?5Whatisthenatureoflogicalnecessity?Howarelogicallynecessarypropositionsrelatedtofactualpropositions?6Mustapropositionbeakindofpictureinordertohavesense?Whatkindofpictureisinquestion?7Mustelementarypropositionsbecomplete,self-containedandindependentofoneanother?Cantherebesuchpropositions?8IsWittgenstein’slogicalatomismcompatiblewithanempiricistinterpretation?Areotherinterpretationsavailable?9Whataretheimplicationsofplacingethical,aestheticandreligiousdiscoursebeyondthelimitsoffactuallanguage?10HowdoesWittgenstein’sthoughtdisplayholism?Discussinrelationtoatleasttwostagesofhisdevelopment.11Shouldthelogicalstructureoflanguagebeadeepstructureorvisibleonthesurface?12IsWittgenstein’sthoughtacontinuationofphilosophyoritsheir?13Whyshouldacatalogueofexampleslinkedbyfamilyresemblanceberegardedasasolutiontoaphilosophicalproblem?14Howisostensivedefinitionpossible?Whatdoesyouranswerindicateaboutthenatureoflanguage?15Howisthemeaningofawordrelatedtoitsuse?16Isthereanythingwrongwiththesolipsist’suseof‘I’?Whataretheimplicationsofyouranswerforouraccountoftheself?17Whatconstraintsdoestheneedforacriterionofidentityplaceonouraccountofsensoryexperience?Whatsortofidentityisinvolved?18WhataretheconsequencesifthereisnowayofdistinguishingbetweenknowingthatmyuseofawordiscorrectandmerelythinkingthatIknewthatitwascorrect?825\nDAVIDPEARS19Ifwemusttreatsomethingasfundamental,whynottreatourabilitytorecognizesensation-typesasfundamental?20Arewhatlooklikepurelyintellectualabilitiesreallybasedonnaturalinvolvementsinthephysicalworld?21Doesthe‘privatelanguageargument’succeed?22Isthemeaningofawordcontainedintheworditself?23Canthemeaningofadescriptivewordbeunderstoodonanalogywiththemeaningofaname?24Isfollowingalinguisticruleobeyinganindependent,externalauthority?25Howcantheauthorityofalinguisticruledependonagreementinthepracticeofobeyingit?26CanWittgenstein’sapproachtomathematicsaccountforproof?826\n40NietzscheDAVIDE.COOPERWithinafewyearsofhisdeathin1900,FriedrichNietzschewaswidelyrecognizedasathinkerandwriterofgenius,adevastatingcriticofreligion,anacutediagnosticianoftheculturalillsofEurope,andamasterofGermanprose.Itwaslater–firstamongcontinentalEuropeanphilosophers,thenamongEnglish-languagephilosophers–thattheoriginalityofNietzsche’streatmentsofperennialquestionsinMETAPHYSICS(chapter2),EPISTEMOLOGY(chapter1),PHILOSOPHICALPSYCHOLOGY(p.19)andETHICS(chapter6)wasappreciated.Bytheendofthemillennium,itwasclearthatnoothernineteenth-centurythinkerhadsodecisivelyshapedthecontoursofcontemporaryphilosophicaldiscussion.This‘professional’appreciationdidnoteclipsetheearlierreputation,formuchoftheexcitementinreadingNietzscheowestoaninterplaybetweenphilosophicalspeculationanddiagnosisoftheparlousmodernconditionofhumanity.Nietzscheissometimestreatedasa‘playful’,unsystematicthinkerrevellingin‘masks’andcontradictions.Inthischapter,adifferentviewistaken.Althoughnotwritteninasystematicstyle,theworksofNietzsche’smatureyearsarticulateacohe-sivegeneralposition,onewhich,arguably,flowsfromconvictionsexpressedinhisveryearlyessays.WhilethischapterwillfocusonNietzsche’scontributionstophilosophy,totherelativeexclusion,therefore,ofmore‘empirical’onestopsychology,sociologyandhistory,nosharpdistinctionisintendedhere.Nietzschehimselfcertainlydeniedsuchadistinction.1LifeandWritingsAnothersharpdistinctionNietzschedeniedwasthatbetweensomeone’sphilosophyandtheirlife.‘Everygreatphilosophy’,hewrote,hasbeen‘thepersonalconfessionofitsauthor’(BGE6).Sinceanygreatphilosophywill‘commandandlegislate’,will‘makeandcreate’concepts,notjust‘accept[them]asgifts’(WP409),itsauthormustbesomethingofanoutsider–aneagleratherthanastarling(WP989)–uncomfortablewiththeintellectualhabitsofhistimes.CertainlyNietzsche’slifewas,inthemain,closertothatoftheeaglethanthestarling.\nDAVIDE.COOPERFriedrichWilhelmNietzsche(1844–1900)wasthesonofaLutheranpastorwhodiedwhenFriedrichwasonlyfour.EducatedatGermany’sbest-knownschool,theSchulpforta,andattheuniversitiesofBonnandLeipzig,theprecociousNietzschebecameaProfessorofClassicsinBaselattheageof25.TheyearsatBaselweremarkedbyintoxicationwiththewritingsofSchopenhauerandwithboththemusicandper-sonalityofWagner–influenceswhichacceleratedNietzsche’sdisillusionmentwithacademicscholarship.Illnessandthehostilereceptionofhisfirstbook,TheBirthofTragedy,furtherinducedhimtoabdicatefromuniversitylife.For12yearsthepatternofNietzsche’slifewasoneoflonelywandering–fromhoteltohotel,intheSwissmoun-tainsorNorthernItaly–occasionallypunctuatedbyintense,usuallydifficultmeetingswithfriends.(HeendedrelationswithWagnerin1876,appalledbythephilistineatmosphereattheBayreuthfestivalandbythegushingreligiosityofWagner’slastopera.)In1888Nietzsche’shealthandmindcollapsed,theresultofexcessiveworkandperhapsofsyphiliseitherinheritedorcontractedthrough,possibly,hissolesexualencounter.Fortheremainderofhislife,Nietzschewasavegetable,achildlikemannursedbyhismotherandsister.ItisfamiliartodivideNietzsche’swritingsinto‘early’,‘middle’and‘late’periods.Themostsubstantial‘early’worksarehisreappraisal,imbuedwithaSchopenhauer-ianvisionofablindcosmic‘will’,ofGreekthoughtandart,TheBirthofTragedy,andanumberofUntimelyMeditationscriticalofcontemporarycultureandeducation.Attention,however,isnowdeservedlypaidtosomeunpublishedsketchesofthe1870sinwhichNietzschedevelopsadistinctive,radicalaccountoftherelationbetweenthoughtorLANGUAGE(chapter3)andtheworld.Duringthe‘middle’or‘positivist’period,inworkssuchasHuman,AllTooHuman,Nietzsche’sprimaryconcernisthemore‘scientific’oneofexposing,oftenwithwitandvenom,thefactsofhumanpsychologywhich,hebelieved,wouldbothexplainanddiscreditthepretensionsofRELIGION(chapter15),metaphysicsandART(chapter7)torevealan‘eternal’realmof‘absolutetruth’.The‘late’period,duringwhichNietzschedevelopssuchfamousnotionsas‘eternalrecurrence’,‘willtopower’,‘perspectivalknowing’and‘theOverman(Übermensch)’,beginswiththelatersectionsofTheGayScienceandThusSpokeZarathustra.Whilethatlatterwork–aphilosophicalfantasywovenaroundthelifeofanimaginarywanderingsage–maybeNietzsche’sliterarymasterpiece,itsmainthemesaremadeclearerbythetwogreatworkswhichshortlyfollowed,in1886–7,BeyondGoodandEvilandOntheGenealogyofMorals.Thewritingsofthefinalyearoflucidity,suchasTheAntichristandEcceHomo,whilefullofperceptivematerial,aremarredbyshrillnessandexcess,symptomsofNietzsche’simpendingdescentintomadness.Throughouthiscareer,Nietzschewrotenotesandplansforbookswhichdidnotmaterialize(hisNachlass).Commentatorsdifferontheweighttobeputuponthesewritings,especiallythosefromthe1880ssubsequentlyassembledbyNietzsche’ssisterunderthetitleTheWilltoPower.Inthischapter,andinoppositiontosomerecentcommentaries,theimmenselyinterestingmaterialfoundinthosenotesisfreelydrawnupon.Wheretherearetensionswiththepublishedworks,Idonotpre-emptivelysettletheissueinfavourofthelatter.828\nNIETZSCHE2‘Catastrophe’Likeothernineteenth-centurythinkers,includingMARX(chapter34),NietzschethoughtthatWesterncivilizationwasatacriticaljuncture,indeedthatitwasfacinga‘catastrophe’whichhecalled‘theadventofnihilism’(WPpreface).Thesymptomsofthecrisiswerevarious,evenanomalous:politicalanarchism,revolution-arysocialism,world-wearyapathy,undiscriminatingtolerance,vulgarhedonism,reli-gioushypocrisy,andsoon.Allthese,forNietzsche,weresymptomaticoftheerosionofbeliefswhichhad,forcenturies,given‘meaning’tocivilizedlife.Inafamouspassagehewrotethat‘Godisdead...wehavekilledhim’(GS125.cf.ZPrologue2).Enlight-enmentrationalism,thenaturalsciencesandmodernpsychology,withtheir‘cultiva-tionof“truthfulness”’(WP3),hadmadeitincreasinglydifficulttomaintainreligiousbeliefand,therefore,tosubscribetomoralvalueswhichpresupposedtheexistenceofGod.By‘thedeathofGod’,however,Nietzschehasmuchmoreinmindthantheerosionofspecificallyreligiousbeliefsandvalues.Godisonlyanespeciallyvividinstanceofabeingimaginedtoexistina‘trueworld’setagainstan‘apparent’worldofeverydaysense-experience.OtherinstanceswouldbePLATO’s‘forms’(chapter23)andKANT’s‘thingsinthemselves’(chapter32).‘Theadventofnihilism’spellsthelossofbeliefinanysuch‘higher’or‘trueworld’and,consequently,inthe‘asceticideal’whichhasgroundedourhitherto‘highestvalues’onthenatureofthatworld–onthewillofGod,say,or‘theFormoftheGood’.Despitethecatastrophicupheavalsitoccasions,Nietzschelargelywelcomesthis‘waron...atrueworld’(WP583),sinceitisonewagedby‘truthfulness’onillusion.Hewelcomesit,moreover,despitehisappreciationofjusthowradicalitmustbe.Tobeginwith,itisawarwhichmusteventuallybewagedagainstmanyofthose–includingscientists–whoarebusilydispellingtheoldillusions.Thisisbecausetheytooareguiltyofsettingupa‘trueworld’–ofnaturallaws,particlesandsoon–setagainst,andallegedlyunderlying,an‘apparent’worldof‘becoming’.Indeed,itmustbewagedagainstallthose‘articlesoffaith’–including‘bodies,lines,planes,causesandeffects,motionandrest’–withoutwhichallofus,andnotjustscientistsandphilosophers,mustfindithardto‘endurelife’,soengrainedaretheyinourfamiliarwaysofthink-ingandspeaking(GS121).ForNietzsche,these‘articlesoffaith’areasmuch‘fictions’asGodorplatonic‘forms’.Further,thewarcannotbeendedbyhittinguponnewbeliefsevenremotelyanalogoustothediscreditedones.Ifnihilismistobeovercome,andlifereaffirmedandinvestedwithmeaning,thiswillbebecausehumanbeingsareabletodispensewithidealsofthekindhithertoembraced.Whethertheycandothisandstill‘endurelife’isuncertain.So,therefore,isourfuturehistory.TounderstandhowNietzschearrivesathisperceptionofthemodern‘catastrophe’,weneed,first,tograspbothwhyheconcursinthe‘assassination’oftheillusionof‘thetrueworld’(section3),andhowheexplainsourpronenesstothisillusion(section4).Second,weneedtoappreciatewhyNietzschethinksthatscience,despite–orbecauseof–itsurgeto‘truthfulness’,fallsvictimtothesameillusion,andtounderstandhowhetriestoexoneratehisown‘truths’fromsuchacriticism(section5).Special829\nDAVIDE.COOPERattentionmustbepaid,third,toreligionandmorality(section6):foritisherethattheillusionhasbothitsoriginandmostdamagingimpact.WewillthenbeinapositiontorevisitNietzsche’svisionofthemodernconditionandtheprospectsfor‘overcoming’it(section7).3Concepts,WorldandLifeTheconfluenceoftwolinesofthoughtresultedinNietzsche’stotalrejectionof‘thetrueworld’–ofanystructured,‘OBJECTIVE’REALITY(pp.726–33)deemedtoexistindependentlyofhumanconcerns,judgementsandvaluations.Thefirstoftheseisalreadyvisibleinhisearliestessays,wherehearguesthatanyaccountwecangiveoftheworldisindelibly‘anthropomorphic’.Therecanbenoreasontosupposethattheconceptsandjudgementsemployedtodescribetheworldcaptureitsantecedent,objec-tivelayout,sincethesearetheoutcomeofvarioushumanlywrought‘transferences’or‘metaphors’.Inparticular,itiswe,notNature,whodividethefluxofsensoryexperienceintoclassesorspecies,suchasleavesordogs–somethingwedobyimposinga‘sameness’amongthedatawhichtheydonotthemselvespossess.‘Everyconceptarisesfromtheequationofunequalthings’(PT,p.83).Thatthese,butnotthose,objectsfallunderacertainconcept,isduetous,nottothepre-givenstructureofreality.This‘nominalist’or‘constructivist’viewinspiresNietzsche’smuch-quotedremarkthat‘truthsareillusionswhichwehaveforgotten...metaphorsthathavebecomewornout’(PT,p.84).Sohabitualhaveourconceptsbecomethatweimaginethem,andthejudgementswhichemploythem,torecordtheobjectivelayoutoftheworld.Thatisanillusion,andnoneofthosejudgementsistrue,therefore,inthetraditionalsenseofcorrespondingtothewaythingsobjectivelyare.Theformationanduseofconceptsisnot,then,duetothedemandsoftheworld.Itowes,rather,topracticalhumaninterests–inorganizingtherelativechaosofexperi-ence,inpredictingandcontrollingthecourseofexperience.FortheyoungNietzsche,theseclaimswerecompatiblewiththepostulationof‘thingsinthemselves’:it’sjustthat,asKantrightlyemphasized,wecanhavenoknowledgeofwhatthesearelike.Inasecondlineofthought,however,thematureNietzschecomestorejecttheveryintel-ligibilityofthingsinthemselves,ofadomainofrealityinaccessibletohumanbeings.‘The“trueworld”finallybecameafable’,hewrites,withthemovefromtheKantianviewofitas‘unattainable’totherecognitionthatitis‘superfluous’–that,indeed,theverycontrastbetweena‘true’worldandamerely‘apparent’onemustbeabolished(TIIV).Ifthe‘apparent’worldofleaves,dogsandpeoplecontrastswithanything,thisisthe‘chaosofsensations’whichwehavebroughttoorder,tosomethingwecancopewith,inthat‘apparent’world.Itcannotcontrastwithsomemore‘real,trulyexisting’orderofthings,fornosensecanbemadeofsuchanorder.Thisisbecausetheverymeaningoftermslike‘real’and‘exist’istiedtowhat‘concernsus’,whathas‘efficacy’forus,whatengageswithourlifeandexperience(KGWVIII1.5.19).Thequestionearlierallowed,‘whatoftheworldwouldstillbethere’ifthe‘humanhead’throughwhichitisviewedwere‘cutoff’(HAH1,9),isnowproscribed.Astheseremarkssuggest,Nietzsche’sfirstlineofthoughtdidnotexpire.Hecon-tinuestoemphasizethemannerinwhichourconceptsareactively‘constructed’or830\nNIETZSCHEmouldedbyusinkeepingwithourinterestsandpracticalconcerns.Forexample,theconceptofpunishmentisreallya‘wholesynthesisof“meanings”’inwhichhas‘crystallized’awhole‘historyof[the]employment[ofpunishment]forthemostvariouspurposes’(GMII13).Tosupposethatconceptscouldmirroranindependentrealityistoignoretheirsubjectionto‘form-givingforces’atworkinour‘fundamental...activity’(GMII12),theirroleina‘generaleconomyoflife’whichNietzschecomestoequatewith‘willtopower’(BGE23,WP675).Nietzsche’s‘abolition’of‘thetrueworld’meansthatheis,inrecentparlance,arobust‘anti-realist’,whodeniesthatthereisawaytheworldanywayisindependentofhumaninterests,perspectivesandjudgements.Somecommentators(forexample,Clark1990,1998)suggestthat,inhislateworks,Nietzscheretractedthispositionandembraceda‘commonsenserealism’whichrulesoutonlya‘metaphysicalrealism’,accordingtowhichtheremayexistarealityclosedto‘anypossibleknower’.Thissug-gestion,however,requiresonetoignoremanylateremarksintheunpublishednotes,suchas‘wecancomprehendonlyaworldthatweourselveshavemade’(WP495).ItalsorequiresonetosupposethathisprolongedcriticismsofrealisticconceptionsofTRUTH(pp.74–5)andKNOWLEDGE(chapter1)–onesofwhich,incidentally,hethinks‘commonsense’isguilty–werelevelledagainstapositionwhich,arguably,almostnoonehaseverheld(seePoellner1995).If‘commonsenserealism’istheviewthattherearetruestatementswhichcorrespondtoarealityindependentofhumanperspectives,itisonewhichNietzscheconsistentlyrejected.Nietzsche’srejectionof‘thetrueworld’indeedraisesthequestionofthestatusas‘truths’bothofperfectlyacceptableeverydaystatements,suchas‘It’sraining’,andofthephilosophicalclaimswhichhehimselfadvances.Nietzschedoesnotwanttodenytheavailabilityoftruthandknowledgeinsomesenseofthoseterms.Hisrelativelysketchyremarksonthisquestionwillbeconsideredinsection5.Onereasonhisremarksaresketchyisthatheislessinterestedinthisquestionand,hence,intheanaly-sisoftruth,thaninthequestionofwhypeoplevaluetruth(seePippin1998).Why,especially,havetheyneededtheillusionofa‘trueworld’fortheirbeliefstocorrespondto?TounderstandNietzsche’sanswertothatquestion,weturntohisphilosophicalpsychology.4Psychologyand‘Genealogy’Nietzsche’saccountofhumanpsychologyisdoublyimportant.First,itmountsarobustattack,oftenprescientofWITTGENSTEIN’s(chapter39),onatraditional,entrenchedconceptionofmind.Second,itplaysakeyroleinexplainingtheillusionof‘thetrueworld’.Attackandexplanatoryrolearecloselyconnected,sinceitisNietzsche’sviewthattheconceptionherejectsisitselfamodelinstanceof‘thetrueworld’illusion.Hence,understandingwhypeoplearesoattractedtothemistakenconceptionofmindwillaidinunderstandingpeople’spronenesstothebroaderillusion.If,inparticular,wecanaccountforbeliefinasubstantialmental‘subject’,wewillhavedonemuchtoexplaintheviewthatrealityconsistsofsubstantialobjectsdistinctfrom‘themedleyofsensations’,since‘itisonlyafterthemodelofthesubjectthatwehaveinvented[that]reality’(WP552).831\nDAVIDE.COOPERThementalsubjectis,infact,Nietzsche’scentralcriticaltarget.Whetherreferredtoasasubject,self,ego,I,mindorsoul,itis‘afable,afiction’,theproductofa‘crudefetishism’whichpostulatessomeentityasthehiddencauseofthoughts,feelingsandactions(TIVI3,III5).Tosupposethattheremustbeadoer‘behind’thedeed,athinker‘behind’thethought,islikesupposingthatthelightningissomethingdistinctfromitsflashing.Thesewrongsuppositions,Nietzscheremarks,areencouragedbyverbalformslike‘Idid/thought/feltX’or‘Thelightningflashed’,whichtemptustopostulateasubstantialsubjectcorrespondingtothegrammaticalsubject.Nietzsche’srejectionoftheself,subjectoregoas‘onlyaword’(TIIII5)registershis‘NOMINALIST’(pp.628–9)hostility,akintoHUME’s(chapter31),tothinkingthattheremustbesomeentityincommonto,orcausing,thediversethoughts,feelingsandactionsattributedtoaperson.Moreimportantly,hethinksthatthenotionofselforsubjecthasbecomeirretrievablyinvestedwithwrong-headedconceptionsofwhathumanbeingsarelike.ItisnotonlyphilosopherslikeDESCARTES(chapter26),butedu-catedcommonsense,thatconceiveofapersonasbeing,essentially,arational,con-scious(andself-conscious),self-directingagentpossessedoffreewill,onlycontingentlyconnectedtoabody.Nietzscherejectsthiswholeconception.Peoplehardlyeveractrationally,inthesenseofactingforreasons–thesebeing,typically,‘rationalizations’aftertheeventwhich‘causenothing’.Neitherreasonnorconsciousthoughtplaysany-thinglikeaslargearoleinbringingaboutbehaviourasdoes‘thenervoussystem’(WP529,476,526).Moregenerally,thereisno‘helmsman’servingasthe‘directingforce’behindourbehaviour(GS360).Asforself-consciousness,farfrombeinganessentialpropertyofhumanbeings,itissomethingwhich‘developedonlyunderthepressureoftheneedforcommunication’(GS354).Self-reflectiveconcernowestothepracticalneedtoletothersknowofone’scondition,andhencepresupposesthedevelopmentoflanguage.Freedomofthewill,towhichournotionofselforsubjectisespeciallyinhock,isanotherfableorfiction(TIVI3),incompatiblewithanhonestrecognition–itselfatoddswiththetraditionalconception–oftheinseparabilityofmentalandbodilylife.‘BodyamIentirely’,proclaimsZarathustra,and‘soulisonlyawordforsomethingaboutthebody’(ZI4).Nietzschedoesallowforthepossibilityofcertainindividualsachievingwhatdeservetobecalledfreedom,self-directionand,hence,selfhood.Preciselybecausethiswouldbeanachievement,however–arareanddifficultone,atthat–itiswrongtodepicttheactuallivesofordinaryhumanbeingsasthoseoffree,self-directingselves.Why,then,isthisdepictionsoentrenched?Nietzschewillusetheresourcesofthephilosophicalpsychologyhepitsagainstthetraditionalonetosupplyananswer.Thepsychologyor,ashesometimesprefers,‘physio-psychology’whichNietzscherecommendsisa‘minimalist’onewhichseekstounderstandall‘doingandwilling’,believingandvaluing,intermsofasingle‘universalandbasicinstinct’(WP675).Atonelevel,apersonisa‘multiplicity’or‘socialstructure’ofvariousdrives,instinctsandaffects,butallofthesemaybesubsumedunderthe‘universalinstinct’for‘life’whichNietzschecalls‘willtopower’(BGE12,23)–an‘instinct’manifestedingrowthandexpansion,inovercomingobstacles,inadaptationtocircumstances,andsoon.Alljudgementsandevaluationsare‘intheserviceof’and‘expression(s)’ofwilltopower(WP675).832\nNIETZSCHENietzsche’s‘physio-psychology’isatthecentreofhis‘genealogical’accountoftheillusionof‘thetrueworld’.Forwhilegenealogy–‘theattempttoshowushowwehavebecomewhatweare,sothatwemayseewhatwemightyetbecome’(Ridley1998:8)–alsoincorporatesreflectionontherolesof,say,historicalprocessesinbringingaboutbeliefs,itisclearthat,forNietzsche,theserolesarenotautonomous.Thehistoricaldevelopmentstowhichhealludes,suchas‘theslaverevolt’inmorality,arethemselvesexplicableintermsof‘physio-psychology’.Nietzsche’sgenealogyisintended,notonlytoexplainbeliefsandevaluations,buttobreakpeople’sattachmenttothem.Itcandothis,despitethefactthat‘thequestionof[their]origin...isnotatallequivalenttotheircritique’(WP254).FromHuman,AllTooHumanon,aprominentthemeisthatimplicitinmanybeliefsandevaluationsisafurtherbeliefabouttheirorigins.Especiallyinthecaseofour‘highestconcepts’–moralandreligiousones,say–theimplicitconvictionisthat‘thehighermaynotgrowoutofthelower’(TIV4).Todemonstratethatsuchconceptsarenottheproductsofreasonorintuition,butdevelopoutofsomethingashumbleor‘shameful’asnaturalneeds,isthereforeboundtodiscreditthem.TheultimatetargetofNietzsche’sgenealogyor‘physio-psychology’isthewholeillu-sionof‘thetrueworld’.Since,aswehaveseen,theentitieswithwhichpeoplepopu-latethatworld–God,atoms,Formsandsoon–aremodelledonthesubstantialselforsubject,debunkingthelatterwillstrikeamajorblowatthelargerillusion.Sowhatisthegenealogicalexplanationofbeliefintheselforsubject?ForNietzsche,itisnoacci-dentthatthepropertieswithwhichtheselfisinvestedareonesofmoralsignificance–freewill,self-directionandthelike.Freedomofthewill–that‘foulestofalltheolo-gians’artifices’–isconcoctedsothat‘men...mightbejudgedandpunished’(TIVI7).Theimageofourselvesasrational‘helmsmen’,byseparatingusfrommereanimals,isintegraltomoralesteem(GS360).ThetraditionalphilosophicalpsychologyNietzscheattacksis,inshort,arequirementofmorality.Hencewecanonlyunder-standNietzsche’sexplanationoftheillusionof‘thetrueworld’,includingtheillusionoftheselforsubject,byattendingtohisgenealogyofmorality(section6).Itistherethathis‘physio-psychology’ofdrives,affectsandwilltopowerdoesitscorrosivework.5Science,PerspectiveandPowerTheterm‘physio-psychology’maysuggestthat,forNietzsche,theproperantidoteto‘trueworld’metaphysicsisanaturalscientificaccountoftheworld.Certainly,headmiresconscientiousscientistsfortheirrespectfortheevidenceofthesensesandtheircommitmentto‘truthfulness’.‘Hurrayforphysics!’,thatitteachesustoobserve,heproclaimsinTheGayScience(340).Foursectionslater,however,hewritesthatpeoplewith‘faithinscience’infact‘affirm...anotherworldthanthatoflife,nature,andhistory’,andthereby‘deny...thisworld,ourworld’(GS344).Science,itemerges,isthelatestformoftheillusionof‘thetrueworld’,of‘faithintheasceticidealitself’(GMIII24).Thescientist,nolessthanthetheist,isincapableofaccepting‘ourworld’of‘chaos’and‘becoming’astheonlyoneandopposestoitamore‘real’worldofBeing–onepopulated,notbygods,butbyforces,substances,lawsofnatureandsoon.833\nDAVIDE.COOPERNietzscheisan‘instrumentalist’withrespecttosuchtheoreticalentities.They‘simplydon’texist’(GS112),beingatbest‘regulativefictions’requiredforcertainpurposes,suchaspredictingthecourseofexperience,butinnowayexplanatoryofexperience.Physicsis‘onlyaninterpretation...oftheworld(tosuitus...)andnotaworld-explanation’(BGE14).Ifitwererecognizedassuch,therewouldbenoobjec-tion:indeed,itisan‘interpretation’whichmaybe‘imperativefor...machinistsandbridge-builders’(ibid.).Notonly,however,doestheasceticidealimpelpeopletothinkthatscienceisexplanatoryofexperience,theyprivilegescientificdescriptionsoverallothers.Thatisaterrible‘prejudice’whoseeffectistoreducetheworldtosomething‘essentiallymeaningless’.Music,forinstance,getsreducedtowhatcanbe‘counted,calculated,putinto[the]formulas’ofmathematicalphysics(GS373).Sincefaithinscience,likebeliefintheselfandfreewill,isaproductoftheasceticideal,withitsrelegationof‘ourworld’,explanationofitbelongs,ultimately,tothegenealogyofmorality.Withthescientific‘willtotruth’,we‘standonmoralground’(GS344),sincethisisthebreeding-groundfortheasceticidealandtheillusionof‘thetrueworld’.Beforeturningtothat,however,weneedtodiscusstheproblemswhichNietzsche’sattitudetoscienceposesforhisownclaimsabouttheworldandhumanbeings.Relatedproblemsloomedinsection3asaresultofNietzsche’sdenialoftheexistenceoftruthinthetraditionalsenseofcorrespondencewith‘objective’facts.Thoseproblemsbecomemoreacutewhenhewritesthat‘thereis...onlyaperspec-tive“knowing”’(GMIII12)andthat‘factsarepreciselywhatthereisnot,onlyinterpretations’(WP481).Everyone,notjustphysicistsandtheists,isadvancingonlyinterpretationsfromacertainperspective.Forsomecritics,theclaimthatthereareonlyperspectivesorinterpretationsisplainparadoxical.Iftrue,itiseitherfalseornobetterwarrantedthananyotherclaim–sinceititselfexpressesjustonemoreperspectiveorinterpretation.Thiscriticismmissesitsmark.WhenNietzschecallsajudgementperspectival,hedoesnotmeanthatitismis-taken,butthatitisnottrueinthesenseofcorrespondingtoreality.Andthatatheoryorjudgementisaninterpretationdoesnotprecludetherebeingreasonstopreferit,incertaincontextsatleast,overarivalone(seeNehamas1985).Thephysicist’sinterpretation,recall,maybe‘imperative’for‘bridge-builders’.Ifitiswrongtochargeperspectivismwithparadox,soitis,atanoppositeextreme,toconstrueitastheuncontentiousepistemologicalclaimthatknowledgepresupposessomestandpoint,thatknowersmusthavesomecognitiveinterests.ForNietzsche,thestandpointsandinterestswhichitisimpossibletotranscendarepractical,‘alltoohuman’and‘biological’ones.Thatthesearepresupposedbyallenquiryisfarfromtruistic.Moreover,Nietzschepersistentlyconnectshisperspectivismwithmetaphysicalassertionstotheeffectthatanystructuredworld–anyfacts–ofwhichsensecanbemadeisone‘invented’or‘made’.Perspectivismisofapiecewithhisanti-realism.Nietzscheisoftenunderstoodtoregardhisownclaimsasperspectivalinterpreta-tionswhicharetrueinthesenseofbeingsuperioronPRAGMATIC(chapter36)groundstorivalones.Manyofhisremarkssupportthisreading.Hesaysofhisdoctrineofwilltopower,‘supposingthatthisalsoisonlyinterpretation...well,somuchthebetter’(BGE22).AndtherearemanypassagesinTheWilltoPowerwhereheurgesthatitis‘valueforlife’whichisthe‘finaldeterminant’oftruthintheonlyviablesensewhich834\nNIETZSCHEremainsoncethetraditionalnotionisabandoned(WP493).Eveninthecaseoflogicalandmathematicalpropositions,‘theirutilityaloneistheir“truth”’(WP514).IfthisisNietzsche’sposition,itisnotanobviouslyincoherentone.Buthisperspec-tivalaccountofthesciencesstillposesarealproblem.Thisisbecausehisown‘physio-psychology’lookslikeascientifictheory.Now,science(psychologyandbiologyincluded)inquiresintotherelativelyorderedempiricalworldwhich,forNietzsche,isthe‘product’ofthedrives,affects,etc.,subsumedunderourwilltopower.Theproblemis,howcanoneexplaintheveryexistenceofanorderedworldonthebasisofphe-nomenawhichbelongwithinit?Nietzscheisawarethatonecannot:itisabsurdtoproposethat‘theexternalworldistheworkofourorgans’,ofanythinginvestigatedbyphysiologyorbiology,forthenourorgans,as‘partofthisexternalworld...wouldbe–theworkofourorgans!’(BGE15).Ifcoherenceistobemaintainedthen,asHEIDEGGER(1987:46)(chapter41)urges,Nietzschecannotbe‘thinkingbiologically’aboutourdrives,instinctsand‘life’.Thatthesearenottheprocessesstudiedbythebiologicalsciencesissupportedbydescrip-tionsofthemas,interalia,‘interpreting’and‘havingperspectives’(WP643,581).The‘rulingdrives’whichcompriseourwilltopowerarenotnaturalprocesses,butbelong,asHeidegger(1987:72)putsit,to‘thepraxisoflife’.Theyarethebasic,purposiveactivitieswithoutwhichintelligentthoughtandactionareimpossible–thosewhich‘imposeuponchaos’,throughschematizing,simplifying,ordering,‘subduing’dataandthelike,the‘regularityandform’requiredbythe‘practicalneeds’ofallbutthecrudestkindofhumanexistence(WP515).Itistheimpositional,orderingcharacterofthispraxiswhichinvitesthelabel‘willtopower’.Thedrivesetc.subsequentlyinvestigatedbybiologyandphysiologybelong,nottothis‘form-giving’praxis,buttothenaturalorderwhichisits‘product’.AsnagwiththisinterpretationisthatwhileNietzschesometimesequateswilltopowerwith‘life’,inotherplaceshespeaksofitaspresent‘inallevents’(GMII12),asthe‘innermostessence’oftheworldatlarge(WP693).(Insodoing,hepartlyresurrectstheposition,inspiredbySchopenhauer,ofTheBirthofTragedy.)Apossiblewaytoexplainthisproclamationofa‘cosmic’willtopoweristhefollowing.‘Thetotalcharacteroftheworld’,hewrites,isoneof‘chaos’,inthatitlacks,‘initself’,the‘arrangement’intoobjects,species,causalprocessesandsoonwhicharea‘humancontribution’(GS109).Still,the‘chaos’isarelativeone,fortheworldmustdisplaythatdegreeof‘direction’and‘organization’(WP561)whichmakesitpossibleforthecon-ceptsandschemasweimposetogainanypurchase(seeSchacht1983:194–9).Ifwearetocharacterize,howeverinadequatelyandmetaphorically,thisquasi-chaoticorganizationof‘forces’,itcanonlybeby‘employ[ing]manasananalogy’(WP619),byinvokingtheterminologyofpower–ofwill,subjectionandindeedforceitself.Ifthisisright,thenNietzsche’sownmetaphysicsis‘onlyinterpretation’or‘onlyper-spectival’inaspecialway.Itisnotarival,onthesamelevel,toperspectivesherejects,suchasthereligiousand‘mechanistic’interpretationsoftheworld.Unlikethese,itisnotguiltyoftheillusionofastructured,‘trueworld’of‘Being’.Ontheotherhand,Nietzscherecognizesthatthemetaphorswhichthecharacterizationoftheworldaswilltopowerinvokesmight‘eventuallyseemunsuitable’and‘toohuman’(BGE22).Nietzsche’scharacterizationis‘onlyinterpretation’tothedegreethathedoesnotexcludethepossibilityofmorefittingandresonatingmetaphors.835\nDAVIDE.COOPER6MoralityandReligionThedomaintowhichNietzschemostresolutelyapplieshis‘physio-psychology’ofwilltopowerisreligiousmorality.Thisispartlybecause,asnoted,thevariousmetaphysi-calillusionsheexposes,suchastheselfor‘faithinscience’,turnedoutto‘standonmoralground’.Itisalsobecause,inNietzsche’sview,itisJudaeo-ChristianmoralitywhoseemergencewasthedecisiveeventofWesterncultureandwhoseatrophyistheprimaryreasonfor‘theadventofnihilism’.Agenealogyofmoralityismadeimperativebytheimpossibilityofconcurringinpeople’sownestimateoftheirmoraljudgementsasstatementsofmoralfact.‘[T]herearealtogethernomoralfacts’,onlya‘moralinterpretation’ofphenomena(TIVII1).Theneed,then,istoexplainhow,froma‘pre-moral’conditionoflife,properlymoralinterpretationsemerge.‘Properly’needsemphasizing,sinceNietzschedistinguishesbetweenabroad,thinnotionofmoralityas,roughly,‘obedienceto...laworcustom’(HAH1,96),andthenarrower,richernotionnowsuggestedbytheterm‘moral’.Morality,inthelattersense,comprisesseveralcomponentsmissingfromthethinnernotion:aconcernforthemotivesofactionsmorethanfortheactionsthemselves;anemphasisonconscienceandasenseofguilt;adeterminationtoholdresponsible,blameandpunish;apremiumon‘disinterested’,altruisticactions;theideathatmoraldemandsareuniversal,bindingonallhumanbeingsassuch;andtheaimofminimiz-ingsuffering.Nietzschehashisobjectionstoeachofthesecomponents.Forexample,theuniver-salprescription‘Thisishoweveryoneshouldact’ignoresboththethe‘rank-order’ofhumanbeingsandtheuniquecontextofeachaction(GS335).Again,moralblamepresupposestheuntenableconceptoffreewill.Crucially,however,healsohasacentralcriticismofthewholepackage–ofmoralityitself.Itis‘againstlife’(GMIII13),‘directedagainst...theinstinctsoflife’(TIV4),‘themorbidsoftening...throughwhich...“man”finallylearnstobeashamedofallhisinstincts’(GMII7).Inexplaininghowmoralityproperemergesfrommoralityquacustomormores,genealogymustalsoexplain,therefore,theapparentparadoxof‘life[turning]againstlife’,ofaninstitutionwhich,likeeverythinghuman,isinthrallto‘basicinstincts’comingtobedirectedagainstthem.Nietzsche’sgenealogytakestheformofanhistoricalreconstructioninwhichthedecisiveeventis‘theslaverevoltinmorality’againsttheethicalcodesofthe‘masters’or‘warrior-nobles’whohavehithertodominatedthe‘slaves’.Therevolutionaryresultwasthat‘pre-moral’conceptionsofgood,rightanddutywerehijackedandtrans-formedbythe‘weak’inaccordancewithan‘asceticideal’successfullypromotedbythosefriendsandexploitersoftheweak,the‘asceticpriests’.Aftertherevolt,forexample,thepre-moralterm‘bad’,whichthewarrior-nobleshadearlierappliedtothose‘inferiorinnature’,isgiventhemoralmeaningof‘evil’andappliedbytheslavestotheirerstwhilemasters(GMI11).Onemotivefortheslaverevoltisobvious:resentmentagainstmasterswhosenaturalaggression,especiallywhenexternalenemieswerelacking,turnedtowardstheslavesthemselves.Theslave’sadvantageinpersuadinghisoppressorstosubscribetosuchvirtuesashumilityandcharityisapparent.This,however,canhardlyexplainthe836\nNIETZSCHEsuccessoftherevolt–themasters’self-emasculation–andanyway,Nietzscheargues,ignoresadeeperdimensionoftheslaves’ressentiment.Thisistheirrancourtowardsthemselves,theresultinpartofdirectinginwardsanimpotentresentment‘deniedthetruereaction,thatofdeeds’(GMI10),andinpartofself-loathing.Thelatterisduetoasenseofinferioritywithrespect,especially,tothepowerexercisedbythenoblesof‘creationandimpositionofforms’,ofdetermininghowthingsshouldbecalledandactionsvalued(GMII17).Thegeniusofthat‘repulsivecaterpillar’,theasceticpriest,istoconvertthiscorro-sive,innerresentmentintosomething‘joyful’.Hedoessobyinventingandsuccessfullymarketingtworelateddualisms:betweenthemundaneworldandthe‘trueworld’ofGod,andbetweenmaterialexistenceandthatoftheimmortalsoul.Thefirstdistinc-tioniselaboratedsoasto‘devaluate’anddenyseriouspointto‘ourearthlyreality’(EHIV8),andtoconstruesufferingassomethingdeserved,through‘originalsin’.Thisstrategy,itmightseem,couldhardlyalleviatetheslaves’suffering,butinavirtuosodisplayofspeculativepsychology,Nietzschearguesthatitdoes.Forpeopleasceticallyfixedonanideal,divinerealm,thetribulationsofanearthlylifenowperceivedasa‘mistake’becometrivial.Moreover,itwasnever‘sufferingassuchbutthesenselessnessofsuffering’whichwasunbearable,anditispreciselytheideathatthereis‘anysuchthingassenselesssuffering’whichChristianityhas‘abolished’(GMII7).Theseconddualism,withthepromiseitaffordsofthesoul’simmortalbeatitude,reinforcestheasceticideal’sdevaluationoftheearthly,andservesintwofurtherwaystobolstertheweak.First,theidentificationofthepersonwithaninnerselforsoulfa-cilitatesadoctrineofequality,sinceovertdifferencesinstrengthandbeautybetweenmastersandslavesmaynowbedismissedassuperficial.Thewayisopenforprinciplesofequalhumanrights,ofKantian‘respectforpersons’assuch.Second,thesoulismadethelocusofafreedomofthewillwhichisdeployed,notonlyforthepurposeofholdingthemastersculpablyresponsiblefortheirdeedsbut,moresubtly,toerasetheslaves’senseofinferiority.Iftheybehavemeeklyandunaggressively,thisisnotbecausetheyareweak,butbecausetheyhavechosensotobehave.Itisasif,Nietzscheobserves,lambsweretoconvincethemselvesthattheirdifferencefrombirdsofpreyconsistsinelectingnottopredate(GMI13).Nietzsche’sstoryoftheemergenceofmoralityproperviathedualisms,illusionsandidealspromotedbyasceticpriestsisofacomplexitythatnobriefaccountcanaccom-modate.Itis,insum,thestoryof‘lifeturnedagainstlife’.Inthis,forNietzsche,thereisnoparadox.Allhumanlifeiswilltopower,whose‘natural’expressioniscreativecontrolof,‘form-giving’impositionon,theworld.Among‘botchedandbungled’peopledeniedsuchexpression,thewillmustbe‘sublimated’intootherstrategies,oneswhich‘devaluate’thekindsofexpressionofwhichsuchpeopleareincapable.If,intoto,thereligiousmoralityoftheasceticpriestsandtheirclientèle,theweak,representsanet‘diminutionoflife’,ithasneverthelessbeenanexpressionoflife–oftheirlife.7OvercomingNihilismThatfinalpointindicatesacertainambivalenceofNietzsche’stowardstheerosionofreligiousmorality,forherecognizesthegenuinebenefitsithasbroughtto‘theherd’–837\nDAVIDE.COOPERindeed,toallofus,sinceitscivilizingeffectshavebeenresponsiblefor‘thesuperiority...ofmenoverotheranimals’(GMI6).Certainlyhedoesnotadvocateareturntothemoresofthewarrior-nobleswho,foralltheircourageandabsenceofrancour,were‘stupid’and‘barbaric’.Onbalance,nevertheless,Nietzschewelcomesthedemiseofreligiousmorality,‘theendofitstyranny’overthosecapableofrisingabovetheherd(WP361)andtheaboli-tionofthe‘trueworld’illusionwhichhassustainedit.Thisisdespitetheupheavalsthisdemisemustcause,theentrenchmentofthe‘trueworld’illusioninnewforms(notably‘faithinscience’),andNietzsche’sprediction,indarkermoods,thatnihilismwilllastfortheforeseeablefuture.Thiswillnotbethe‘active’,andtransitory,nihilismofbomb-chuckingrevolutionaries,butsomethingmoreenduringanddepressing–the‘passive’,‘decadent’,‘sickly’nihilismofpeoplewithoutidealsandpurpose,indeedwithoutwill.Atleastthe‘active’nihilistwilledsomething,ifonlysomethingnegativeanddestructive,anditisbetterthat‘manwillnothingnessthannotwill’atall(GMIII28).Thetruly‘nau-seating’spectreisthatofthe‘maggot-men’,‘littlemen’or‘lastmen’ofZarathustra’sPrologue,withtheireasy-goinghedonism,liberaltoleranceofallopinionsandtastes,‘thicklypaddedhumanity’(TIIX37),andlackofcommitmenttoany‘decisiveandhard’YesorNo–menforwhomtheonlysin,perhaps,istothinkanythingasin.Thisremarkableanticipationofa‘postmodern’climateofthoughtissometimescountered,oraccompanied,byNietzsche’sconfidenceintheemergence,ifonlyastheoccasional‘luckyhit’(GMI12),ofa‘redeeming’kindofpersoninwhomnihilismanddecadenceareovercome.Likethe‘lastman’,andunliketheasceticpriest,thisredeem-ingtypewillbewithout‘articlesoffaith’,liberatedasheisfromtheillusionofa‘trueworld’ofvaluestheretobediscovered.Butliketheasceticpriest,andunlikethe‘lastman’,hewillbecommittedtovaluesandideals,albeitoneswhichare‘moraline-free’and‘created’.Assuch,this‘Overman’,asNietzschesometimescallstheredeemingtype,willindeedbe‘new,unique,incomparable’,forneverpreviouslyhavepeopleattempted,letalonesuccessfullymanaged,tolivewithcommitmenttoidealsofwhichtheyrecognizethemselvestobethesolesource.Wherefaithina‘transcendental’sourcehasatrophied,so,aswiththe‘lastman’,hasanysenseofmeaningfulpurpose.NietzschenowherepaintsadetailedportraitoftheOverman(orhisnear-relatives–the‘highertype’,‘sovereignindividual’,‘freespirit’andsoon).Afterall,‘therehasneverbeenanOverman’(ZII4),andsincetheOvermancreateshisownvalues,indeed‘createshimself’,noblueprintofhislifecanbepreparedinadvance.Nevertheless,wearetoldofsomeofthequalitiestheOvermanmusthaveifheistobethe‘supremetype’ofhumanbeingthroughmaximallyembodyingwilltopower.Tobeginwith,hemust,asthenamesuggests,overcome,forin‘thedegreeofresistance...continuallyover-come’isameasureof‘freedomunderstood...aswilltopower’(WP770).Unliketheasceticpriest,whattheOvermanovercomesarenotourinstinctsand‘rulingdrives’,buttheasceticidealitselfandotherobstaclestotheexpressionofthosedrives.Second,hisattitudetolifeasawholemustbethatofthe‘Yea-sayer’,ofunqualifiedaffirma-tion,tothepointindeedofcelebratingthethoughtof‘eternalrecurrence’,oflife–one’sownincluded–‘return[ing]toyou,allinthesame...sequence’innumerabletimes(GS341;cf.ZIII2).Thisisbecausetheultimateexerciseofpoweriscontrolofthepast,‘torecreateall“itwas”intoa“thusIwilledit”’(ZII20),which,giveneternalrecurrence,issomethingonecanindirectlydointhesenseofwillingpasteventsto838\nNIETZSCHEreoccur.(WhetherNietzscheregardedeternalrecurrenceasaplausiblecosmologicalhypothesis,andnotsimplyasathoughtexperimentfortestingpeople’sstrengthofaffirmation,ismuchcontested.See,forexample,Danto1965.)Finally,theredeeminghumantypemustpossessthe‘greatandrareartof“givingstyle”toone’scharacter’,inthespecificsenseofincorporatingeverythinginhisorherlife–drives,affects,values,ideals–intoan‘artisticplan’(GS290).Tobeweakistobedisintegrated,tornapart,withoutacentre.Tobepowerfulistoachieve,likeGoethe,‘totality’,anintegrationof‘reason,senses,feeling,andwill’.ItisbecauseGoethe‘dis-ciplinedhimselftowholeness’thathecanbesaidtohave‘createdhimself’andtohaveapproximated,atleast,totheOverman(TIIX49).Ineffect,theOverman,an‘artistofhislife’,combinesthetwogreat‘artdrivesofnature’,DionysianandApollonian,vividlydepictedinNietzsche’sfirstbook(BT1).LiketheApollonianartist,theOvermangivesformandstructuretohislife,butliketheDionysians,withtheirinsightintothe‘primaloneness’oftheworld,thisisdoneinfullrecognitionthat‘allisredeemedandaffirmedinthewhole’(TIIX49),thatformsareimposedbythe‘form-giver’.Thisiswhy‘style’,hencepower,hencetheredeemingtypeofhumanbeing,arepossibleonlyinthewakeofthedeathsofGodandtheillusionsofanalreadystructured‘trueworld’.Whethersuchtypeswillbeafew‘luckyhits’inaworlddominatedby‘lastmen’,orwhethertheycanbe‘bred’insufficientnumbertobringaboutthedemiseofnihilismtoo–thisisanissueNietzscheleavestothefuturetodecide.FurtherReadingUntilveryrecently,theonlyreliabletranslationsofNietzsche’smainworkswerethosebyKaufmannandHollingdale,listedbelow.Currently,however,nolessthanthreeuniversitypresses–Oxford,CambridgeandStanford–arepublishinggood,well-editedtranslationsoftheseworks.ForbriefsurveysofNietzsche’slifeandwritings,seeHollingdale(1999)andTanner(1994).ThebestmoredetailedandgeneralaccountofhisthoughtremainsSchacht(1983),thoughthepioneeringworksbyKaufmann(1974)andDanto(1965)arestillworthconsulting.BooksonparticularareasofNietzsche’sphilosophyhavemushroomedinrecentyears.Forcontrastingaccountsofhisepistemology,metaphysicsandtheoryoftruth,seeClark(1990)andPoellner(1995).(Thepresentchapteriscloserinapproachtothelatter’s.)Forilluminatingdiscussionsofhismoralphilosophyandgenealogy,seeRidley(1998)andMay(1999).ThreeadvancedandoriginalinterpretationsofNietzsche’sphilosophyareNehamas(1985),Staten(1990)andMueller-Lauter(1999).OnNietzsche’scomplicatedrelationshiptoSchopenhauer,seeJanaway(1998);onhisviewsonart,Young(1992)ishelpful;onhisviewsonpoliticsandeducation,seeAnsell-Pearson(1994)andCooper(1991),respectively.Craig(1987)andPippin(1998),thoughnotsolelydedicatedtoNietzsche’sphilosophy,containusefulchaptersonit.Among‘continental’contributionstoNietzschescholarship,Deleuze(1983)andKofman(1993)havetheiradmirers.Heidegger(1987)–oneoffourvolumestranslatedfromhislecturesonNietzsche–isinaclassofitsown:onegreatthinker’sruminationsonanother.ReferencesNietzscheReferencesinthetextaretosectionnumbersofNietzsche’sworks,notpagenumbers,unlessotherwisespecified.Titlesoftheworkshavebeenindicatedbyacronymswhichareexplainedinthefollowinglistofworkscited.839\nDAVIDE.COOPERF.W.Nietzsche1960:ThePortableNietzsche(translatedbyW.Kaufmann).NewYork:Viking.Includes:ThusSpokeZarathustra[Z](1883–5),TwilightoftheIdols[TI](1889),TheAntichrist[A](1889).——1967–84:KritischeGesamtausgabe:Werke[KGW],ed.G.ColliandM.Montinari.Berlin:deGruyter.——1968a:BasicWritingsofNietzsche(translatedbyW.Kaufmann).NewYork:ModernLibrary.Includes:TheBirthofTragedy[BT](1872),BeyondGoodandEvil[BGE](1886),OntheGenealogyofMorals[GM](1887)——1968b:TheWilltoPower[WP](1880s)(translatedbyW.KaufmannandR.J.Hollingdale).NewYork:Vintage.——1974:TheGayScience[GS](1882;PartV,1887)(translatedbyW.Kaufmann).NewYork:Vintage.——1986:Human,AllTooHuman[HAH](1878)(translatedbyR.J.Hollingdale).Cambridge:CambridgeUniversityPress.——1993:PhilosophyandTruth:SelectionsfromNietzsche’sNotebooksoftheEarly1870s[PT](translatedbyD.Breazeale).AtlanticHighlands,NJ:HumanitiesPress.OtherwritersAnsell-Pearson,K.1994:AnIntroductiontoNietzscheasaPoliticalThinker.Cambridge:CambridgeUniversityPress.Clark,M.1990:NietzscheonTruthandPhilosophy.Cambridge:CambridgeUniversityPress.——1998:Nietzsche,Friedrich.InTheRoutledgeEncyclopediaofPhilosophy,ed.E.Craig.London:Routledge.Cooper,D.E.1991[1980]:AuthenticityandLearning:Nietzsche’sEducationalPhilosophy.Aldershot:Gregg.Craig,E.J.1987:TheMindofGodandtheWorksofMan.Oxford:ClarendonPress.Danto,A.C.1965:NietzscheasPhilosopher.NewYork:Macmillan.Deleuze,G.1983[1962]:NietzscheandPhilosophy(translatedbyH.Tomlinson).NewYork:ColumbiaUniversityPress.Heidegger,M.1987[1961]:Nietzsche,Vol.3:TheWilltoPowerasKnowledgeandasMetaphysics(translatedbyJ.Stambaugh,D.F.KrellandF.A.Capuzzi).SanFrancisco:HarperandRow.Hollingdale,R.J.1999:Nietzsche:TheManandHisPhilosophy.Cambridge:CambridgeUniversityPress.Janaway,C.(ed.)1998:WillingandNothingness:SchopenhauerasNietzsche’sEducator.Oxford:ClarendonPress.Kaufmann,W.1974[1950]:Nietzsche:Philosopher,Psychologist,Antichrist,4thedn.Princeton,NJ:PrincetonUniversityPress.Kofman,S.1993[1972]:NietzscheandMetaphor(translatedbyD.Large).London:AthlonePress.May,S.1999:Nietzsche’sEthicsandHisWaronMorality.Oxford:ClarendonPress.Mueller-Lauter,W.1999[1971]:Nietzsche:HisPhilosophyofContradictionsandtheContradictionsofHisPhilosophy(translatedbyD.Parent).Urbana:UniversityofIllinoisPress.Nehamas,A.1985:Nietzsche:LifeasLiterature.Cambridge,MA:HarvardUniversityPress.Pippin,R.B.1998:IdealismasModernism:HegelianVariations.Cambridge:CambridgeUniversityPress.Poellner,P.1995:NietzscheandMetaphysics.Oxford:ClarendonPress.Ridley,A.1998:Nietzsche’sConscience:SixCharacterStudiesfromthe‘Genealogy’.Ithaca,NY:CornellUniversityPress.Schacht,R.1983:Nietzsche.London:RoutledgeandKeganPaul.Staten,H.1990:Nietzsche’sVoice.Ithaca,NY:CornellUniversityPress.840\nNIETZSCHETanner,M.1994:Nietzsche.Oxford:OxfordUniversityPress.Young,J.1992:Nietzsche’sPhilosophyofArt.Cambridge:CambridgeUniversityPress.DiscussionQuestions1IsNietzscherighttoseeacloseconnectionbetweenaperson’slifeandtheirphilosophy?2WhatdoesNietzschemeanby‘nihilism’andhowdoesheexplainits‘advent’?3WhatareNietzsche’sobjectionstotheideasofa‘trueworld’and‘thingsinthemselves’?4WhydoesNietzscheholdthatalltruthsare‘metaphors’?5OnwhatgroundsdoesNietzscheregardtheselforsubjectasa‘fiction’?6WhatisNietzschean‘genealogy’?Howcanitbeusedtodiscreditpeople’sbeliefsandvalues?7What,forNietzsche,isthestatusofnaturalscience?8WhatisNietzsche’s‘perspectivism’?Isitaparadoxicaldoctrine?9HowshouldoneunderstandNietzsche’sclaimthateverythingis‘willtopower’?10HowdoesNietzscheexplainthetransitionfromcustomandmorestomorality‘proper’?11Whatrolesdothe‘asceticpriest’andhis‘asceticideal’playinthe‘slaverevoltinmorality’?12Inwhatways,accordingtoNietzsche,isthenotionoffreedomofthewillcrucialtomorality?13WhatarethemainfeaturespossessedbyNietzsche’s‘redeeming’typeofperson,theOverman?14WhatroleinNietzsche’sphilosophyisplayedbythethoughtof‘eternalrecurrence’?841\n41HusserlandHeideggerTAYLORCARMANEdmundHusserl(1859–1938)wasthefounderofphenomenology,aphilosophicalmovementthatexertedenormousinfluenceonEuropeanthought,especiallyduringthefirsthalfofthetwentiethcentury.Hisassistant,MartinHeidegger(1889–1976),nowwidelyrecognizedasoneofthemostimportantphilosophersinrecenthistory,radicallyredirectedphenomenologybyapplyingittothequestionofthemeaningofbeingandthestructureofhumanexistence.ItwasHusserl’sdescriptiveapproachtotheproblemofintentionality,avoidingtheoreticalconstructionandmetaphysicalspecula-tionasmuchaspossible,thatinitiallyinspiredtheyoungHeidegger.Yettheirapproachestophilosophysoonprovedtobedeeplyatodds,inbothstyleandsubstance.ForwhereasHusserlidentifiedintentionalitywith‘pure’consciousness,ortranscendentalsubjectiv-ity,Heideggertraceditbackinsteadtothepragmaticcontextandthetemporalstruc-tureofoureveryday‘being-in-the-world’,whichhethoughtprecededanydistinctionbetweensubjectivityandobjectivity.InhislaterwritingsHeideggermovedbeyondanythingHusserlwouldhaverecognizedasgenuinephenomenologicalinquiry,reflect-ingonthenatureofart,poetry,science,technology,andthenihilismhecametoregardasinherentinallmetaphysicalthinking.1HusserlEdmundHusserlwasbornin1859inProssnits,Moravia(nowProstejov,CzechRepublic).HestudiedinLeipzig,theninBerlin,whereheworkedwiththemathemati-cianKarlWeierstrass.AfterreceivinghisPh.D.inmathematicsin1881,heattendedFranzBrentano’slecturesinphilosophyandpsychologyinViennafrom1884to1886andthereuponchosetodevotehimselfentirelytophilosophy.HetaughtatHalle(1887–1901),Göttingen(1901–16),andfinallyFreiburg(1916–28),wherehewassucceededbyhisapparentprotégé,MartinHeidegger.Inretirement,Husserl,whowasJewish,sufferedtheeffectsofanti-SemiticlegislationforabriefperiodwhentheNazisrosetopowerin1933.Hegrewincreasinglyisolated,bothprofessionallyandpersonally,untilhisdeathin1938.\nHUSSERLANDHEIDEGGER1.1Psychologism,intentionalityandcategorialintuitionInhisbook,PhilosophyofArithmetic(published1891),Husserltriedtoworkoutanempiricisttheoryofarithmeticalconceptsbytracingthembacktotheirpsychologicalorigins.Theconceptofmultiplicity,heargued,isrootedinourconcreteintuitionofaggregates,orclustersofthings.Whenweintuitaggregates,wegroupobjectstogetherinanactof‘collectivecombination’.Itisthismentalactofcombiningthatunderliesourconceptofthecardinalnumbers,regardlessofwhatsortsofobjectswecombine;whethertheyarerealorimaginary,abstractorconcrete.Ourintuitionsofconcreteensemblesarethemostbasic,however,andfromthemweacquireanotionofabstractwholesandthepurelyformalconceptssomethingandone,aswellasmultiplicityandnumber.Weunderstandthenumber3,forexample,asadeterminatemultiplicityofcollectivelycombined1s:‘somethingandsomethingandsomething,’or‘oneandoneandone’(Husserl1970a:335).Theconceptofnumber,Husserlconcludes,isderivedfromareflectiononthementalactofcombiningparticulars.Latein1894HusserlwasstillworkingontheprojectedsecondvolumeofPhilosophyofArithmetic,butby1896hehadabandonedtheprojectandrenouncedthepsychologismonwhichhistheoryrested.Hischangeofheartwasverylikelyprompted,atleastinpart,byareviewGottlobFREGE(chapter37)wroteofthefirstvolumein1894.FregeobjectedstrenuouslytoanyblurringofthelinesbetweenLOGIC(chapter4)andpsychology,betweenobjectiveconceptsandmeresubjective‘ideas’(Vorstellungen).InpsychologistictheorieslikeHusserl’s,Fregecomplains,‘everythingisturnedintoideas’(Frege1984:197).By1896Husserlbeganworkonhisfirstproperlyphenomenologicaltreatise,LogicalInvestigations(published1900–1).TheentirefirstvolumeoftheInvestigations,the‘ProlegomenatoPureLogic’,isanextendedrefutationofpsychologism.AlthoughHusserldidnotatthetimementionanydirectinfluence,helateracknowledgedthat‘Frege’ssignificancewasdecisive’(quotedinFøllesdal1982:55),andthat‘Frege’scriticism...hitthenailonthehead’(Spiegelberg1972:66).AllHusserlsaysintheforewordtotheInvestigationsisthathebegantohave‘doubtsofprinciple,astohowtoreconciletheobjectivityofmathematics...withapsychologicalfoundationforlogic’,andthatthesedoubtsforcedhimtorethinktherelation‘betweenthesubjectivityofknowingandtheobjectivityofthecontentknown’(Husserl1970b:42).LikeFrege,Husserlwouldnowinsistthatthenormativityoflogiccannothaveitstheoreticalfoundationsinpsychology.Forlogicallawsareexact,whilepsychologicallawsareinexact.Moreover,whereaslogicallawsareknowableaprioriandpurelybyinsight,psychologicaltheoryisnecessarilyempiricalandinductive.Andlogicaffordscertainty,whileknowledgeinpsychologyisalwaysmerelyprobable.Finally,thelawsoflogicyieldnormsgoverningourreasoning,whilethelawsofpsychologyaremeredescriptionsofcausalregularities.(ForHusserl’sexplicitrepudiationofhisearliertheoryofnumber,seeibid.:784.)ThisdistinctionbetweentheobjectivecontentsoflogicandMATHEMATICS(chapter11)andthesubjectivestuffofexperience,betweenmattersofessenceandmattersoffact,wouldfigureprominentlyinallofHusserl’ssubsequentwork.Mostimportantly,itinformshistheoryofintentionality,whichisthecentrepieceofhismaturephenomenology.‘Intentionality’,atechnicaltermthatBrentanoimportedintomodern843\nTAYLORCARMANphilosophyfromthescholastictradition,referstothedirectednessofconsciousness.Consciousnessalwayshasanaccusative;itisalwaysof(orratherasifof)anobject.Wedonotjustseeorremember;weseeorremembersomething.Intentionality,then,isthe‘of-ness’or‘about-ness’ofourmentalstates.Drawingasharpdistinctionbetweentheidealcontentsofexperienceandthereal(temporallyextended)experiencesthemselves,Husserlnowattributestheintentionalityofanattitudetoitsidealcontent,orwhathecallsits‘act-matter’,relegatingitspsychologicalcharactertoits‘act-quality’(ibid.:586–90).Eachofthesetwocomponentsofanintentionalstatecanvarytosomedegreeindependentlyoftheother:perceptionsallhavethesameact-qualityquaperceptions,forexample,thoughtheymayintenddifferentobjects(orindeedoneandthesameobject)bymeansofdifferentact-matters;sotoo,actsofanticipation,perceptionandmemoryallhavedifferentact-qualities,thoughtheymaybedirectedtowardoneandthesamethinginvirtueofthesameact-matter.EvenmoresignificantisthedistinctionHusserldrawsbetweentheinternalcontentsandtheexternalobjectsofintentionalacts,forphilosophershaveperpetuatedanumberoflong-standingproblemsabouttheontologicalstatusofintentionalobjectsbyconflatingthelatterwiththeformer.Ifallconsciousnessisconsciousnessofsome-thing,forexample,whatshouldwesayaboutdreams,hallucinations,falsememoriesandexpectations,andnon-veridicalactsgenerally?Doweinthosecasesstandinrelationto‘non-existentobjects’,asAlexiusMeinong(1853–1920),whowasalsoastudentofBrentano’s,proposed?Traditionalepistemologytendedtoobscurethephenomenonofintentionalityaltogetherbydescribingmentallifenotasadirectednesstowardexternalobjects,butasanimmediatepossessionof,orinnerconfrontationwith,ideas,fromwhichwewouldthenhavetoinfertheexistenceoftheexternalworldandotherminds.Husserl’stheoryunderminestheindirectrepresentativepicture,forhedistinguishestheidealcontentthatorganizesourawarenessfromtheobjectstowhichweareputativelyrelatedinvirtueofthatcontent.Intentionalityisnotarealexternalrelationbetweenthemindandobjects,Husserlargues,butamentaldirectednessthatobtainsandhascontentwhetherornottheobjectsofourattitudesthemselvesexist,hencewhetherornotwestandinanyrealrelationtothem.Ithasbeennoted(seeFøllesdal1969)thatHusserl’sdistinctionbetweencontentandobjectbearsastrikingresemblancetoFrege’snotionsofthesense(Sinn)andreference(Bedeutung)oflinguisticexpressions.Inthiscase,however,itwasalmostcertainlynotFregewhoinspiredHusserl,butratherearlierfiguressuchasBernardBolzano(1781–1848)andJ.S.MILL(1806–73)(chapter35),whohaddrawnverysimilardistinctions.IntheSixthInvestigationHusserlarguesforanotionofintellectualor‘categorial’intuition,overandbeyondsenseperception.Perceptualconsciousnessinvolvestwodistinct,sometimescoinciding,intentionalacts:anemptyintendingorsignifyingact,andanintuitiveorfulfillingact.Perceptionisnotwhollypassive,then,sinceweexperienceoursensuousintuitionsassatisfying(orfailingtosatisfy)priorsignifyingacts,oranticipations.Noteverythinganticipatedinoursignifyingacts,however,canbegiveninsenseperception.Anactthatmerelysignifiesorintendswhitepaper,say,maybefulfilledinasensuousintuitionofwhitepaper,butonedoesnotliterallyseeitsbeingwhite.Thecopula‘is’intheproposition‘Thepaperiswhite’doesnotitselfcorrespondtoanythinggivensensuously,astheexpressions‘white’and‘paper’do.Nevertheless,844\nHUSSERLANDHEIDEGGERHusserlmaintains,anintentionwhosesignifyingcontentisthatthepaperiswhitecanbeadequatelyfulfilledinanon-sensuous,categorialintuition.Categorialintuitionsfulfilactswhosecontentsincludethemeaningsofpurelyformaltermslike‘is’and‘not’,thelogicalconnectives‘and’and‘or’,andquantifierslike‘all’,‘some’and‘none’.ContrarytoempiricistslikeLOCKE(chapter29)andHUME(chapter31),Husserlinsiststhatourunderstandingofstatesofaffairsassatisfyingpropositionalattitudescannotbebasedonamereabstractionfromourperceptionofsensuousparticularsansweringtoindividualideasorlinguisticterms.Ahigher-levelintuition,heinsists,affordsusdirectinsightintothestructuredstatesofaffairsintheworldthatmakeourbeliefstrueorfalse.1.2PurephenomenologyandtranscendentalsubjectivityHusserlwouldelaborateandmodifyhistheoriesofintentionalityandintuitioninhismiddleperiod,whichbeganaround1905andculminatedinhismagnumopus,theFirstBookofIdeasPertainingtoaPurePhenomenologyandPhenomenologicalPhilosophy(published1913).Husserlnownolongerreferredtohisprojectas‘descriptivepsycho-logy’,anexpressionheinheritedfromBrentano,butas‘purephenomenology’,whichhesaysisan‘eidetic’science,ascienceofessences.Phenomenologythuspromisestoelevatephilosophybeyondthemereexpressionofaworldview(Weltanschauung)tothelevelofarigorousscience,whosewatchwordis,asHusserloftenputit,‘Tothethingsthemselves!’(seeHusserl1965:96ff.).Bythelate1920sHusserlhadmovedsofarfromhisearlier,morestrictlypsychologicalapproachtointentionalitythathecametocharacterizephenomenologyitselfasaformof‘transcendentalidealism’(Husserl1931:18–19;1960:83–6).Heregrettedthisterminologicalchoicewithinafewyears,however,complainingin1934that‘Noordinary“realist”haseverbeenasrealisticandasconcreteasI,thephenomenological“idealist”(awordwhich,bytheway,Inolongeruse)’(lettertoAbbéBaudin,quotedinKern1964:276n.).ThisTRANSCENDENTAL(pp.731–3)phaseofHusserl’sdevelopmentismarkedbytwophilosophicalinnovations:the‘phenomenologicalreduction’,amethodologicalnotionhedevisedinthesummerof1905;andthenewconceptionofintentionalcontentheformulatedby1907,whichhewouldnowcallthenoema,incontrasttotherealmentalact,ornoesis.ThesenotionsspecifywhatHusserlregardsastheprivilegedsiteforanyinquiryintointentionality,namely‘transcendentalsubjectivity’,or‘pure’con-sciousness,abstractedfromallrealpsychologicaldeterminations.Husserlfrequentlyemphasizestheradicalbreakbetweenourordinary‘naturalattitude’commontoeverydaylifeandempiricalscience,priortothereductionandthereflectiononpureconsciousness,andthephenomenologicalattitude,whichrevealstranscendentalsubjectivity,whichHusserlcalls‘thewonderofallwonders’(Husserl1980:64).Thismethodologicalgapbetweennatural(andnaturalistic)cognitionandphenom-enologicalreflectionmirrorswhatHusserlinsistsisanessentialdiscontinuitybetweenconsciousnessandreality.‘Insofarastheirrespectivesensesareconcerned’,hewrites,‘averitableabyssyawnsbetweenconsciousnessandreality’(Husserl1983:111);heevengoessofarastosaythatalthoughconsciousness‘wouldindeedbenecessarilymodifiedbyanannihilationoftheworldofphysicalthingsitsownexistencewouldnotbetouched’(ibid.:110).PassagessuchasthesehavegeneratedcontroversyconcerningHusserl’smetaphysicalandepistemologicalcommitments.Inanyevent,845\nTAYLORCARMANalthoughherejectssubstancedualism(Husserl1970c:212),HusserlhimselfisfrequentlyatpainstostressthespiritofCARTESIANISM(chapter26)animatinghisphilosophicalorientationatlarge.ThePhenomenologicalReductionsandtheNoemaHusserldescribesanumberofdifferent‘reductions’,buttwoinparticulararecentraltohisconceptionofphenomenologicalmethod:theeideticreductionandtranscendentalreduction.Takentogether,Husserlsuggests,thetwoservetoidentifythepureintentionalcontentofconsciousnessassuch,orwhathecallsthenoema.Theeideticreductionconsistsinignoringreal(temporal)particularsandfocusinginsteadongeneralandideal(atemporal)featuresofthings.So,forexample,onecanabstractfromalltheothercontingentpropertiesofrosesandfiretrucksandgrasptherednessinstantiatedinboth.Husserlcallssuchgeneralproperties‘essences’–hence‘eidetic’fromtheGreekeidos(Plato’s‘form’).Sotoo,phenomenologicalreflectiononcon-sciousnessabstractsfromtherealfeaturesofconcretepsychologicalepisodesoccurringintimeandconcentratesinsteadontheiridealstructuresandcontents(Husserl1983:xx).Thetranscendentalreduction,whichHusserlalsocallstheepoche¯(atermborrowedfromancientscepticism),consistsinsettingaside,or‘bracketing’out,allobjectstranscendenttoconsciousness,focusinginsteadontheintentionalcontentsimmanentwithinit(ibid.:§31–4,56–64).Anobjectis‘transcendent’,inHusserl’ssense,ifonly§onesideoraspectofitcanbeimmediatelypresenttousatanyonetime;suchthingsarenecessarilygivenperspectivally,orin‘adumbrations’.Anobjectis‘immanent’ifitisgiventoconsciousnessallatonce,transparently,sothatnoperspectivalvariationmediatesourapprehensionofit.Physicalbodiesandstatesofaffairsaretranscendentobjects,forexample,andsotooaretheabstractentitiesofmathematicsandformalontology.Thecontentsofconsciousnessareimmanent,bycontrast,sinceweeachhaveimmediate,transparentaccesstoourownthoughtsandexperiences.Theinwardreflectionoftheepoche¯,then,firstpresentsamentalstateasaconcreteparticular,ornoesis–includingsensation,whichHusserlcallshyle¯–andtheeideticreductionthensetsasideitsconcretepsychologicalfeaturesinfavouroftheidealintentionalstructuresandcontentsitinstantiates.Thoseidealstructuresandcontentsconstitutethenoemaofthementalstate,whichincludesa‘core’ofrepresentationalcontent,or‘sense’(Sinn),aswellastheideal‘positingcharacter’inwhichthatsenseisputforwardinone’smindaseither(say)perceived,judged,remembered,anticipated,imaginedorwishedfor.Thetranscendentalandeideticreductionstogetherpurporttoisolatethesphereof‘pure’phenomenologicalinquiry,namelytranscendentalsubjec-tivity.Ananalogybetweenthecoreor‘sense’componentofthenoemaandFrege’snotionoflinguisticsense(Sinn)findssupportinHusserl’sremarkthat‘thenoemaingeneralisnothingotherthanthegeneralizationoftheideaof[linguistic]meaning(Bedeutung)toallact-domains’(Husserl1980:76).AnothermajorshiftinHusserl’sthoughtduringthisperiodconcernshisaccountofthephenomenologicalstatusoftheself.InLogicalInvestigationshedefendsaversionofHume’s‘bundletheory’,rejectingKantianappealstoapureego,supposedlyneededtounifyourvariousintentionalacts.‘Imustfranklyconfess’,hewritesintheFifthInvestigation,echoingHume,‘thatIamquiteunabletofindthisego,thisprimitive,846\nHUSSERLANDHEIDEGGERnecessarycentreofrelations’(Husserl1970b:549).Inthesecondedition,publishedin1913,however,heretractsthedenialinafootnote,writingsimply,‘Ihavesincemanagedtofindit’(ibid.:note).Whilestillrepudiatingwhathecalls‘corruptformsofego-metaphysic’,Husserlnowconsidersitphenomenologicallyevidentthatpureconsciousnessexhibitsastructureofownership,centredaroundapure,transcen-dental‘I’,whichis‘essentiallynecessary’andremains‘absolutelyidentical’throughthewholeofone’sexperience(Husserl1983:132),butwhich‘isnotapieceoftheworld’(Husserl1960:25).Husserl’sdoctrineofthetranscendentalegolaterdrewcriticism,firstfromHeidegger,thenfromJean-PaulSARTRE(1905–80)(chapter42)andMauriceMerleau-Ponty(1908–61),allofwhomfeltthatitviolatedphenomenology’scommitmenttoadescriptionofexperienceasweactuallyliveit,priortoalltheoreticalembellishment.TheircommoncomplaintisthatHusserl’snotionofapureortranscendentalselfconstantlyinhabitingallourthoughtsandactionsamountstoareflectivedistortionofourconcreteengagementwiththeworld.WhenIamactivelyabsorbedinwhatIamdoing,nosuchabidingcentralizedegomanifestsitselfinmyawareness.Husserl’spositionisalsovulnerabletoWittgenstein’scritiqueofsolipsisminTractatusLogico-Philosophicus,intheBlueBook,andinthecelebrated‘privatelanguageargument’inPhilosophicalInvestigations.LikeHusserl,WITTGENSTEIN(chapter39)distinguishestheembodiedpersonfromthepurelyformal‘subject’ofexperience;heevendeniesthatthelatterisanobjectintheworld.ButwhereasWittgensteinconceivesofthesubjectasnothingmorethananideallimitorvanishingpoint,borrowingSchopenhauer’smetaphoroftheeyethatdoesnotappearinitsownfieldofvision,Husserlregardsthepureego‘assomethingabsolutelyidentical’acrossallone’sexperience(Husserl1983:132).Mytranscendentalego,then,turnsouttobeakindofenduringprivateobject,afterall,internaltomyconsciousness.Butifitremainsidenticalacrosstime,whatarethecriteriaofitsidentity?Apparentlytherearenone,atleastnoneIcanspecify,eveninprinciple,inwhichcaseonecanhardlyassertitstemporalpersistence,asHusserldoes.1.3IntersubjectivityandthelifeworldWhetherornotonecanspeakofathirddistinctphaseinHusserl’sthought,hislaterworkatleastplacesnewemphasisonpracticalasopposedtotheoreticalattitudes,onintersubjectivity,thebody,andtheculturalandsocialconstitutionofwhathecallsthe‘lifeworld’(Lebenswelt).Atthesametime,inhislastworksHusserlcontinuestorevisitandelaboratethebasicmethodologicalprinciplesofphenomenology.CartesianMeditations(basedonthe‘ParisLectures’hedeliveredattheSorbonnein1929andfirstpublishedinFrenchtranslationin1931),forexample,bearsthesubtitle‘AnIntroductiontoPhenomenology’.ThebookisfarfromamerereiterationofHusserl’spreviouswork,however,forintheFifthMeditationhesketchesanoriginalaccountofintersubjectivity,partlyinordertoquellsuspicionsthathistranscendentalmethodentailssolipsism.Husserlintroducesanewreduction,areductiontomysphereof‘ownness’,thatis,anabstractionfromeverythingreferringovertlyorcovertlytootherselves.Ithendiscover,inaprimordial‘here’,myownbodyastheuniquelocusofmyownwillandsensations.Otheroutwardlysimilarbodiesoccur‘there’inmyperceptualenvironment,butIneithercontroltheirmovementsnorlocatemysensationsinthem.847\nTAYLORCARMANThankstoakindof‘pairing’associationofthoseotherbodieswithmyownbody,Irecognizethemaslinkedtotranscendentalegospheresoftheirown,whichcannotinprinciplebegivendirectlytomeinintuition.Ithusperceivethemasotheregos,notjustobjects,andIseeourrespectivesubjectivitiesasconstitutinganintersubjectivity.HusserlisnotattemptingtoreconstructouractualpsychologicalacquisitionofourconceptofOTHERMINDS(pp.817–20),butinsteadwhathetakestobethestructureofourmatureconsciousnessofothersquaother.Healsoinsiststhatthe‘pairing’associationsupposedlyunderlyingmyrecognitionofothersisnotaninference,butanapperception.Thatis,justasIneitherdirectlyintuitnormerelyinferthebacksidesofphysicalobjects,butinsteadseethemaswholeobjectswithbacksides,sotooIneitherintuittheinnercontentsoftheconsciousnessofothersnormerelyinferthemfromasinglecase,myown.DoesHusserladequatelydescribeourintentionalrelationtoothers?Isasystematicreductiontoasphereofownnessitselfaplausiblenotion?Heretoo,aswiththedoctrineofthepureego,theexistentialphenomenologistsremaineddissatisfiedwithHusserl’saccount.In1934HusserlbeganwritingthetexteventuallypublishedinitsentiretytwentyyearslaterasTheCrisisofEuropeanSciencesandTranscendentalPhenomenology,alsosubtitled‘AnIntroductiontoPhenomenologicalPhilosophy’.Thoughthetermoccursinhismanuscriptsasearlyas1917,Husserl’sconceptofthe‘lifeworld’makesitsfirstappearanceinprintinCrisis(thefirsttwopartsofwhichwerepublishedinthejournalPhilosophiain1936).Thetermisroughlyequivalentto‘naturalconceptoftheworld’,aphraseHusserlhadborrowedinhislecturesof1910–11fromRichardAvenarius(1843–96),whichinturnrecallsthedistinctionhehaddrawnin1907betweenthenaturalandthephenomenologicalattitudes.Thelifeworldisnotjustacollectionofphysicalobjects,butincludessuchthingsasculturalandhistoricalartefactsandsocialinstitutions.Husserlequivocatesaboutwhethertherearemanylifeworlds,oronlyone,andaboutwhetherthetermreferstotheimmanentsubjectivecontentofourconsciousnessoftheworldortotheculturalworlditselfinitstranscendence.Husserlsometimesdrawsacontrastbetweenthelifeworldandtheworldasdescribedbythesciences,particularlyinthewakeoftheGalilean‘mathematizationofnature’(Husserl1970c:23).ItisthereforetemptingtoassimilateHusserl’spointtothedistinctionWilfridSellars(1912–89)drawsbetweenthe‘manifestimage’ofcommon-sensebeliefandthe‘scientificimage’ofadvancedtheory.Butwhereas,forSellars,thescientificimageistheonlytrueimage,andsomayinprinciplesupplantourcommon-sensebeliefsaltogether,Husserlinsistsontheprimacyofthelifeworld,towhichscienceitselfbelongs(ibid.:380),andtowhichitisforeverbeholden.ForHusserl,thatis,scientifictheoriesacquiremeaningandjustificationonlybyreferringbacktotheworldasitisgiventousinordinaryexperience.All‘theoreticalresultshavethecharacterofvaliditiesforthelife-world’,Husserlwrites.‘Theconcretelife-world,then,isthegroundingsoilofthe“scientificallytrue”world’(ibid.:131).2HeideggerMartinHeideggerwasbornin1889inthetownofMesskirchinBaden.HeattendedtheBertholdgymnasiumandtheTheologicalSeminaryattheUniversityofFreiburg,848\nHUSSERLANDHEIDEGGERintendingeventuallytoenterthepriesthood.HavingdiscoveredtheworkofBrentanoandHusserl,however,hechoseinsteadtostudylogic,mathematicsandphilosophy.HewroteadissertationonpsychologismandthetheoryofjudgementunderHeinrichRickert(1863–1936),aNeo-Kantian,andbeganlecturinginFreiburg.His1915Habilitationsschrift,‘TheDoctrineofCategoriesandMeaninginDunsScotus’,isastudyofascholastictextsinceshowntohavebeenwrittenbyThomasofErfurt.In1916HusserlsucceededRickertattheUniversityofFreiburg,andHeideggerbecamehismostpromisingassistant.By1919HeideggerhadabandonedhisCatholicismandbegunlecturingonphenomenology.HewasappointedprofessoratMarburgUniversityin1923,whereheworkedonthemanuscriptthatwouldbecomehismagnumopus,BeingandTime(published1927).In1928HeideggertookoverHusserl’schairatFreiburg,andin1933hejoinedtheNaziPartyandassumedthepositionofrectoroftheuniversity.Contrarytoapopularrumour,HeideggerdidnotbanHusserlfromuseoftheuniversitylibrary.Husserl’stemporarysuspensionwasinsteadtheresultoflegislationenactedbeforeHeideggertookofficeandsubsequentlyrevokedduringhisterm.Still,afterhisarrivalinFreiburg,HeideggercutoffalmostallpersonalcontactwithHusserlandin1941accededtohispublisher’sdemandthatheremovethededicationtoHusserlfromthefiftheditionofBeingandTime.Heresignedasrectorin1934,afterlessthanayearinoffice,butneverexplicitlyrenouncedNationalSocialism.AfterthewarhewasbarredfromteachinginGermanyuntil1951,thoughhewaslecturingagainprivatelyby1949,togreatacclaim.Hediedin1976.2.1TheanalyticofDaseinandthequestionofbeingHeideggerwasdeeplyimpressedbytheaimofphenomenologytoforgoabstracttheoreticalconstructsandgetbacktoaconcreteaccountof‘thethingsthemselves’,asHusserlhadurged.Husserl,forhispart,sawinHeideggeraprotégéworthyoftakingoverandcontinuinghisworkandreportedlyoftensaidtoHeidegger,‘YouandIarephenomenology’(Cairns1976:9).YearslaterHusserlwouldcometorealize,tohischagrin,thattheirrespectiveapproachestophilosophyhadinfactalwaysbeenprofoundlyatodds.First,perhapsaboveall,HeideggercriticizesHusserl’sstrictdistinctionbetweenimmanenceandtranscedence,betweentheinnerandtheouter:‘thereisnooutside’,Heideggersays,‘forwhichreasonitisalsoabsurdtotalkaboutaninside’(Heidegger1982:66).Repudiatingwhathecallsthe‘worldlesssubject’ofDescartesandHusserl,Heidegger(1962:254)insiststhathumanexistenceisessentially‘being-in-the-world’.Intentionalityisnecessarilytiedtoitsworldlycontext,andtheideaofisolatingasphereofpuretranscendentalsubjectivityfromallworldlyobjectsandstatesofaffairsbymeansofanepoche¯mustbefundamentallymisconceived.Second,Heideggerobjectstowhathecalls‘theontologicallyunclarifiedseparationoftheRealandtheideal’(ibid.:259),onwhichthesubstanceandmethodofHusserl’sphenomenologybothdepend.AlthoughhereportsthatthedoctrineofcategorialintuitioninLogicalInvestigationsinspiredsomeofhisearlyreflectionsonthequestionofthemeaningofbeing(Heidegger1972:78),HeideggerexplicitlyrejectsHusserl’stheory,alongwithitslaterincarnationintheconceptofeideticinsight.Bymaintaining849\nTAYLORCARMANinsteadthatallformsofperceptionandinsightareparasiticonabackgroundofcommon-senseunderstanding,Heideggerdeclares,‘wehavedeprivedpureintuitionofitspriority...Eventhephenomenological“intuitionofessences”isgroundedinexis-tentialunderstanding’(Heidegger1962:187).HeideggerconsequentlyabandonsHusserl’sidealofapuretheoreticaldescriptionofintentionalphenomenaand,adopt-ingthehermeneuticalstanceofWilhelmDILTHEY(1833–1911)(pp.384–5),declares,‘themeaningofphenomenologicaldescriptionasamethodliesininterpretation’(Heidegger1962:61).Farfromrestingonpurecategorialintuitionoreideticinsight,Heideggerwrites,‘ThephenomenologyofDaseinisahermeneutic’(ibid.:62).Third,whereasHusserldrawsasharpdistinctionbetweenphenomenologyandontology,Heideggerinsiststhatthetwoareinseparable:allphenomenologyhasitsontologicalpresuppositions,and‘Onlyasphenomenology,isontologypossible’(ibid.:60).Indeed,Heidegger’soriginalconceptionofphenomenologycannotbeunderstoodapartfromhisdeepercommitmenttoaquestionthatremainedforeigntoHusserl,andindeedtomuchofthehistoryofWesternphilosophy,accordingtoHeidegger:thequestionconcerningthemeaning(Sinn)ofbeing(Sein).Whereastraditionalontologyconcernsitselfwithentities,orwhatis(dasSeiende),Heideggeraskswhatitmeansforanythingtobe.Hesetsouttoshedlightonthequestionbyinvestigatingphenomenologicallyourunderstandingofbeing,whichisconstitutiveofhumanbeings,whichHeideggercalls‘Dasein’(literallybeing-there).Thequestionofbeing,then,boilsdowntothequestion‘whatdoweunderstandwhenweunderstandthatandwhatentitiesare,includingourselves?’ModellingBeingandTimelooselyonKant’sTranscendentalAnalyticintheCritiqueofPureReason,then,Heideggercallshisownprojectan‘analytic’ofDasein,whichwillconstitutea‘fundamentalontology’.Fundamentalontologyisfundamentalrelativetoordinarythoughtandtraditionalontology,bothofwhichsystematicallyobscureanddistortourpre-theoreticalunderstandingofexistence.Forthemetaphysicaltraditionandcommonsensebothtendtoconstrueallentitiesasobjectsorsubstancesoccurringinapresentmoment,thehorizonofthenow.Heidegger,bycontrast,baseshisinterpretationofbeing-in-the-worldonthephenomenonofpurposive,future-directedpracticalunderstanding.Weunderstandourselves,thatis,inlightofthepossibilitiesintowhichweproject,thealreadyconstitutedworldinwhichwefindourselvessituated,orthrown,andtheenduringpresentinwhichweencounterotherentities.Dasein’sbeing-in-the-worldthushasatemporalstructureHeideggerdescribesas‘thrownprojection’.Theentitiesweencounter,bycontrast,showupforusinthehorizonofthepresent,eitherasinconspicuously‘available’(or‘ready-to-hand’)inourpracticalactivityoras‘occurrent’(or‘present-at-hand’)asobjects.The‘Availability’ofEquipmentandthe‘Worldliness’oftheWorldInanefforttogetbackto‘thethingsthemselves’,asHusserlinsisteditwasthetaskofphenomenologytodo,HeideggerdeliberatelyfocuseshisanalyticofDaseinnotonthekindofreflectivecognitiveattitudethatphilosopherssooftentakeastheirparadigm,butonourbackgroundimmersioninmundaneactivity.WhenweseeDaseininits‘averageeverdayness’,Heideggersuggests,weseethatwhatsituatesusinaworldmostfundamentallyisnotoursubjectiveexperiencesormental850\nHUSSERLANDHEIDEGGERstates,butourexternallysituatedskillsandpractices.Sotoo,theentitieswetypicallyencounterandconcernourselveswithinoureverydaypracticalactivitypresentthemselvesnotas‘occurrent’(vorhanden)objects,butastransparently‘available’(zuhanden)equipment,whichwerelyonandtakeforgrantedincarryingoutourtasks.WhenIgraspadoorknoborwieldahammer,Iamneitherperceivingnorthinkingaboutitasanobjectwithproperties,butavailingmyselfofitunthinkingly,skilfullytreatingitasanelementintheoverallpurposivestructureofthesituation.Moreover,individualpiecesofequipmentdonotjustoccuralongsideoneanotherinobjectivespaceandtime,butinsteadformanorganizedequipmentaltotality,whichholisticallyassignstheparticularitemstheirrespectivepracticalmeanings.Thatequipmentaltotalityisinturnimplicatedinabroaderintelligiblenetworkofpragmaticrelationsassigningtoolstocontexts,totasks,togoals,andtotheultimateunderlyingpointofwhatwearedoing,whichHeideggercallsour‘for-the-sake-of-which’.Hedescribesthesepracticalrelationsas‘signifying’(be-deuten),andcallstheentireintelligiblenetworkofsignifyingrelations‘significance’(Bedeutsamkeit).Thepragmaticstructureofsignificanceisaconditionofthepossibilityoflinguisticmeaningandmentalcontent,bothofwhichoccuronlyagainstabackgroundofpracticalunderstanding.Moregenerally,itisourprimordialfamiliaritywiththestructureofsignificancethatconstitutesourbeing-in-the-world.For‘being-in’consistsnotinentertainingcognitiveattitudes,butinbeingcompetentlyorientedandinvolvedinintelligiblesituations;similarly,whatHeideggercallsthe‘worldliness’oftheworldisnotjustasumtotalofobjects,propertiesandrelations,butameaningfullystructureddomainofpracticesandinstitutions.Inspiteoftheirmethodologicalandstylisticdifferences,Heidegger’saccountofsignificanceinBeingandTimewarrantscomparisonwithWittgenstein’sideaintheBlueBookandPhilosophicalInvestigationsthatlinguisticmeaningistiedessentiallytoouruseofwords,andwithhisclaiminOnCertaintythatallknowledgeisembeddedinaparticular‘formoflife’.AsHeideggerputsit,knowledgeorcognition(Erkennen)isalwaysonly‘afoundedmodeofbeing-in’(Heidegger1962:86).ThecentralthesisofBeingandTimeisthatbeingisintelligiblealwaysonlyintermsoftime,andthattraditionalontologyfromPlatotoKant(andincludingHusserl)hasinter-pretedthebeingofallentitiesasoccurrentnessbecauseofitsfixationonthetemporalpresent.SinceDasein’stemporalityhastheformnotofpresencebutofthrown-projection,however,wecannotmakesenseofourselvesintermsofthemetaphysicalCATEGORIES(pp.594–8)thatstructureourknowledgeofobjects.Heideggerthereforeintroducesanumberoffundamentalconcepts,or‘existentials’,thatdefineDasein’suniquetemporalexistence.Heideggeralsomaintainsthatourunderstandingofentitiesismediatedbyananonymoussocialnormativity,whichhecalls‘theone’(dasMan),atermalludingtoeverydaylocutionsdescribingandprescribingwhatisproper,thatis,what‘onedoes’.HeideggermaintainsthatdasManisaprimitiveexistentialstructureofbeing-in-the-world,buthealsoassociatesitwithDasein’stendencytolapseintoan‘inauthentic’ordisowned(uneigentlich)modeofexistence,inwhichwefailtocometogripswiththeconcreteparticularityandfinitudeofourindividualexistence.Toexistauthentically,bycontrast,istorecognizethe‘groundlessness’ofone’sbeingandtoanticipateone’s851\nTAYLORCARMANeventualdeathinamoodofanxiety(Angst),butwithanattitudeofopennessandresolve(Entschlossenheit).2.2TheworkofartandthescientificimageoftheworldAccordingtoHeidegger’soriginalplan,BeingandTimewastoconsistoftwoparts,eachcomprisingthreedivisions(see1962,p.64).Ofthetotalsixdivisions,Heideggerwroteonlythefirsttwo,atwhichpointheabandonedtheprojectoffundamentalontology,apparentlybecausehecametodoubtthatananalyticofDaseincouldinprincipleopenthewaytoamoregeneralinquiryintothemeaningofbeing.Itisalsoplausibletosuppose,however,thatheabandonedtheassumptionthatbeinghasaunifiedahistoricalmeaningatall,andsoturnedinsteadtoagenealogicalaccountofthesuccessiveunder-standingsofbeingthathaveinformeddifferentepochsinthehistoryofWesternthought.Ithasthusbecomecustomarytocontrastthehermeneuticalphenomenologyofthe‘early’HeideggerduringtheperiodofBeingandTimewiththereflectionsonpoetry,science,technologyandthehistoryofmetaphysicswhichonefindsinthe‘later’Heidegger,muchasonedistinguishesearlyandlaterstagesintheworkofLudwigWITTGENSTEIN(1889–1951)(chapter39).ThetransitionfromtheearlytothelaterHeideggerisoftenreferredtosimplyas‘theturn’(dieKehre),anditsprecisemotivesremainasubjectofspeculationanddebateamongscholars.Atanyrate,forwhateverreason,bytheearly1930sHeideggerhadrenouncedtheahistoricalsystem-buildingapproachtophilosophythatheinheritedfromKANT(chapter32)andtheneo-Kantiantradition.Nolongercomposingambitioustreatisesinthegrandstyle,henowwroteessaysandlectures,collectedinvolumesbearinghomelytitleslike‘ForestTrails’(Holzwege)and‘Pathmarks’(Wegmarken).ThesubstanceandstyleofHeidegger’sthinkinghadnowbrokenawaycompletelyfromHusserl’sconceptionofphenomenol-ogyasa‘rigorousscience’.In1935,inwhatisprobablyhismostinfluentialessay,‘TheOriginoftheWorkofArt’,Heidegger(1971a)supplementsthetwo-folddistinctionhehaddrawninBeingandTimebetweenoccurrentobjectsandavailableequipment,arguingthatworksofartcannomorebeunderstoodaspracticalequipmentwithsuperveningaestheticproper-tiesthanequipmentcanbeunderstoodassomanybareobjectswithutilitypredicatesaddedon.Consequently,alongsidehisearlierconceptof‘world’,heintroducesacomplementarynotion,whathecalls‘earth’.Whereastransparentlyusefulthingsmakeupthepurposivestructureofaworld,anearthalwaysharbourssomethingrecalcitrantandanomalous,somethingordinarilyhiddenineverydaylife,butwhichweconstantlyrelyon,andwhichworksofartbringtothefore.Ininstitutinganopenbattleor‘strife’betweenearthandworld,theworkofartemergesasaculturalparadigmthatinauguratesthehistoryofapeople.Heidegger’sexamplesincludeanancientGreektemple,aGothiccathedral,and(somewhatproblematically)VanGogh’spaintingofapairofshoes.Likethefoundingofastate,thepresenceofGod,andgenuinephilosophicalthinking,Heideggersays,worksofartlettruth‘happen’intheirownwaybydisclosingthegroundsandlimitsofintelligibilityinanhistoricalworld.Heideggerappealshere,ashedoesinBeingandTime(§44),towhathetakestobethearchaicmeaningoftheGreekwordfortruth,alethe¯ia.Intheessay‘Plato’sDoctrineofTruth’(written1931–2)hearguesthatwhereasthewordorginallymeant‘what852\nHUSSERLANDHEIDEGGERhasbeenwrestedfromhiddennness’(Heidegger1998:171),inthecaveallegoryoftheRepublicittakesontheadditional,bynowcustomary,senseof‘correctnessofapprehendingandasserting’(ibid.:177)andthereafterremainsambiguous,eveninAristotle.Heideggerwouldlateracknowledgethehistoricalincorrectnessofthisinterpretation,thoughhecontinuedtoinsistthatcorrespondenceinfactpresupposes‘theopeningofpresence’(Heidegger1972:70).Inthe1930sHeideggeralsolecturedextensivelyonthephilosophyofFriedrichNIETZSCHE(1844–1900)(chapter40)andonthepoetryofFriedrichHölderlin(1770–1843).InHeidegger’seyes,Hölderlinisthepoetparexcellenceofthepresentage,whoseworkembodiesareflectiononthenatureandsignificanceofpoetryitself,andwhostandsasawitnesstotheculturalspiritandmissionthatHeideggerbelievestheGermanpeoplehaveinheritedfromtheGreeks.Heidegger’sreadingsofHölderlin’spoemsandfragments,likehisinterpretationsofcanonicalphilosophicaltexts,areoftenidiosyncratic,attimesplainlyuntenable.Nonetheless,hislecturesonHölderlinremainessentialtoanunderstandingofhisthought,notleastofallbecauseHölderlin’spoetryturnsouttobethesourceoftheconcreteexamplesandtheterminologythatHeideggerinvokesthroughouthislaterwork,particularlyinthe‘WorkofArt’essay.TheoriginoftheWORKOFART(p.231),Heideggerconcludes,isnottheartistbutartitself,whichhesaysisessentiallypoetry,anditisHölderlin’spoetryaboveallthatspeakstousofthenatureofpoetry,andsoofthemeaningandprospectsofartinatechnologicalage.Inhistwo-volumeNietzsche(publishedin1961,butdrawinglargelyonhislecturesofthe1930sand1940s),HeideggerpresentsNietzscheasthelastgreatthinkeroftheWesternmetaphysicaltradition,whosedoctrinesofthewilltopowerandtheeternalreturnreiteratetheconceptsofessentiaandexistentia,thatis,thewhatandthethatofentitiesingeneral.ForNietzsche,willtopoweriswhateverythingis,whiletobeatallistorecureternally.Moreover,Heideggerargues,Nietzsche’sthoughtbelongsessentiallytothesubjectivismofmodernmetaphysics,thoughhisnihilismalsoanticipatesthe‘technological’understandingofbeingthatholdsswayinthepresentage.Heidegger’sinterpretationofNietzsche,too,isascontroversialasitisoriginal,inthiscaseowinginparttohisrelianceonthemiscellaneouscollectionofnotespublishedposthumously,andmisleadingly,asNietzsche’smagnumopus,underthetitleTheWilltoPower.YetHeideggeralwaysdeniedthathisinterpretationswereintendedascontributionstoscholarlyresearch.Rather,hisreadingsofphilosophersandpoetsofthepastamounttoproductive,ifinevitably‘violent’,confrontationsandconversationswithfellowthinkers.Hisaim,heinsists,isnottoascertainobjectivehistoricalfacts,buttoopenupnewpathsofthinking.Unlikeart,religionandphilosophy,Heideggercontends,‘scienceisnotanoriginalhappeningoftruth,butalwaysthecultivationofadomainoftruthalreadyopened’(Heidegger1971a:62).AnticipatingthesocialconceptionofscienceadvanceddecadeslaterbyThomasKuhn(1922–96),Heideggerarguesinhis1938essay,‘TheAgeoftheWorldPicture’,thatthemodernscientificimageoftheworldisaproductofrigorousresearch,whichrestsontheprojectionofanunderlying‘groundplan’(Heidegger1977:118).Thegroundplanprovidesakindofapriorischemaornormativeframeworkfortheprocedural,industriouscharacterofscientificpractice.Consequently,notonlyisourmodern‘worldpicture’peculiartous,buttheveryidea853\nTAYLORCARMANofaworldpictureisauniquelymodernphenomenon,aconstructoftheresearchsketchedoutinadvanceandprescribedbyanunderlyingplan.Strictlyspeaking,then,previousagesdidnothaveworldpicturesdifferentfromours,sincetheydidnotconceiveoftheworlditselfasarepresentationpresentto,indeedconstructedby,anautonomoussubjectatall.2.3TechnologyandtheforgetfulnessofbeingModernsciencethusrestsonanunderstandingofbeingthatdiffersradicallyfromunderstandingsoperativeintheancientandmedievalworlds.Forexample,theancientGreeksunderstoodbeingasphusis,whichwastranslatedintoLatinasnaturaandbecameourword‘nature’,butwhichHeideggerinterpretstomean‘self-blossomingemergence’and‘openingup,unfolding’(Heidegger1959:14).TheChristianmiddleages,bycontrast,conceivedoftheworldasenscreatum,amadethingseparatedbyachasmfromitsmaker,God(ibid.:106).Themetaphysicsofthemodernperiod,bycontrast,aswehaveseen,movesustoregardtheworldasanobjectivepictureorrepresentationstandingoveragainstathinkingsubject.Inhis1955lecture,‘TheQuestionConcerningTechnology’,Heideggerarguesthatwearenowinthegripofa‘technological’understandingofbeing,definedintermsofefficientordering,or‘Enframing’(Ge-stell),andthatwetendtotreatentitiesingeneral,includingourselves,asresourcematerial,or‘standing-reserve’(Bestand).Thetechno-logicalagediffersfundamentallyfromthemodernage,theageoftheworldpicture,for‘Whateverstandsbyinthesenseofstanding-reservenolongerstandsoveragainstusasobject’(Heidegger1977:17).ThehydroelectricplantontheRhineisHeidegger’sfavouriteexampleofthenearlytotalbutinconspicuoustechnologicalmanipulationofentities,whichatonceflattenstheworldoutandobscuresmeaningfuldifferences.Othermanifestationsofourcurrenttechnologicalunderstandingofbeingincludehigh-speedtransportationandinformationtechnology,abovealltelevision,allofwhichabolishdistancesandcorrodeoursenseofthingsbeingnearorfar,nobleorbase,importantortrivial.Worksofart,bycontrast,disclosehistoricalworldsandmarkdifferences,whilemundaneartefactssuchasawindmill,awoodenbridgeorajugletordinarythingsbewhattheyarebyrevealingthemintheirlocalworlds,asopposedtofittingthemintomoreandmoreefficient,totalizing,homogeneousorderings.Likeeveryhistoricallyrealizedunderstandingofbeing,Heideggersays,‘Technologyisawayofrevealing’(ibid.:12).Indeed,‘Solongaswerepresenttechnologyasaninstrument,weremainheldfastinthewilltomasterit’(ibid.:32),andsofailtounder-standitsessenceasawayforentitiesasawholetoshowupasintelligible.Unlikeearliermodesofrevealing,then,technologytendstoconcealthefactthatitisone;moreoveritconcealsthatveryconcealing.Asaresult,ourcurrenttechnologicalworldissystematicallyresistanttothekindofmeditativequestioningthatHeideggercalls‘thepietyofthought’(ibid.:35).Indeed,thetechnologicalunderstandingofbeingrepresentstheculminationofthenihilistic‘forgetfulnessofbeing’(Seinsvergessenheit)thathethinkshasfuelledthemetaphysicaltraditionsincethePresocratics(Heidegger1959:18).Whatheenvisionsinitsplace,however,isneitherarenunciationofourtechnicaldevicesandhabitsnoraregressiontoanearlierunderstandingofbeing,butanappreciationoftechnologyasthewaythingsshowupforusandarecognitionthat854\nHUSSERLANDHEIDEGGERwearenotthemastersofourmodeofrevealing,butitswitnesses.Heideggerthereforeconcludesthe‘Technology’essaywithHölderlin’sgnomicdeclaration,‘wheredangeris,grows/Thesavingpoweralso’.ForHeidegger,‘preciselytheessenceoftechnologymustharbourinitselfthegrowthofthesavingpower’(Heidegger1977:28).ThesavingpoweriswhatHeideggercallsthe‘freeingclaim’(ibid.:26)thattechnologywillmakeonuswhenwecometounderstanditsessenceasawayofdisclosingentities,thatis,asamodeoftruth,orunhiddenness(ale¯theia).FurtherReadingThebestwaytoapproachHusserlinitiallyistoreadthefirstfewCartesianMeditationsandtheessay‘PhilosophyasRigorousScience’.IthasbeensaidthatHusserlishardertoreadthantounderstand,andthelifelessnessofhiswriting,especiallyintranslation,makesaseriousstudyofIdeasIasomewhatforbiddingtask.Fortunately,therearetwoparticularlyusefulcollectionsofcriticalessays.DreyfusandHall(1982)containsclassicarticlesbyAronGurwitschandDagfinnFøllesdal(seeespeciallyFøllesdal’s1969‘Husserl’sNotionofNoema’);essaysbyJ.N.Mohanty,DavidWoodruffSmith,RonaldMcIntyreandIzchakMillerrelatingHusserl’sviewstorecentdevelopmentsinphilosophyofmind,languageandlogic;andfinallytwopapers,whichmakenomentionofHusserl,bycontemporarytheoristsofintentionality,JohnSearleandJerryFodor.Theothervaluablecollection,coveringawiderangeoftopics,isSmithandSmith(1995).AnexcellentpresentationoftheFøllesdalinterpretation,andanilluminatingaccountofHusserl’stheoryoftimeconciousness,canbefoundinMiller(1984).Bernet,KernandMarbach(1993)isalsoavaluableguide,andSpiegelberg(1982)providesausefulhistoricalcatalogueofpersonalitiesandideas.Finally,forabrilliantclosereading,andtoseehowHusserl’sworkhasinfluencedcontemporaryFrenchthought,seeJacquesDerrida’searlyessaysinDerrida(1989,1973).TherearetwoEnglishtranslationsofBeingandTime.MacquarrieandRobinson’s(1962)classicisgenerallyexcellent.Stambaugh’s(1996)versionistruerstylisticallytoHeidegger’speculiarblendofcolloquiallanguageandneologism,butisnotasaccurateorreliable.ThebestavailablesecondarysourceisDreyfus(1991),whichconcentratesalmostexclusivelyonDivisionIofBeingandTime.ThemostthoroughtreatmentofHeidegger’sconceptionoftemporalityinDivisionII,butwhichalsooffersasuperbaccountofthefirsthalfofthebook,isBlattner(1999).Therearetwoexcellentvolumesofcriticalessays:DreyfusandHall(1992)andGuignon(1993).Mulhall(1996)isagoodintroduction,andGuignon(1983)andRichardson(1986)arehelpfulinanalysingBeingandTimeasacritiqueofCartesianism.Kisiel(1993)providesapainstakingrecordofHeidegger’sveryearlydevelopment,whileOkrent(1988)offersaprovocativeinterpretationofHeideggerasaverificationist.Lafont(2000)offersaninterpretationandacritiqueofHeidegger’sconceptionoflanguage,relatingitbothtothehermeneutictraditioninGermanyandtocontemporaryanalyticphilosophy.Finally,furtherelaborationandanalysisofmanyofthethemesIhavediscussedinthischaptercanbefoundinCarman(forthcoming).ThreeparticularlygoodstudiesofHeidegger’slaterthoughtareSchürmann(1987),Haar(1993)andYoung(2001).Ott(1993)andSafranksi(1998)arebothadmirablebiographies.ThebestdiscussionsofthephilosophicalsignificanceofHeidegger’sNazism,however,canbefoundinSluga(1993)andYoung(1997).ReferencesHusserlHusserl,E.1931:Author’sprefacetotheEnglishedition.Ideas:GeneralIntroductiontoPurePhenomenology(translatedbyW.R.BoyceGibson).NewYork:HumanitiesPress.855\nTAYLORCARMAN——1960[1931]:CartesianMeditations:AnIntroductiontoPhenomenology(translatedbyD.Cairns).TheHague:MartinusNijhoff.——1965[1911]:PhilosophyasRigorousScience.InPhenomenologyandtheCrisisofPhilosophy(translatedbyQ.Lauer).NewYork:HarperandRow.——1970a[1891]:PhilosophiederArithmetik(editedbyL.Eley).TheHague:MartinusNijhoff.——1970b[1900–1]:LogicalInvestigations,2vols(translatedbyJ.N.Findlay).NewYork:HumanitiesPress.——1970c[1954]:TheCrisisofEuropeanSciencesandTranscendentalPhenomenology:AnIntroductiontoPhenomenologicalPhilosophy(translatedbyD.Carr).Evanston,IL:NorthwesternUniversityPress.——1980[1952]:IdeasPertainingtoaPurePhenomenologyandPhenomenologicalPhilosophy.ThirdBook.PhenomenologyandtheFoundationsoftheSciences(translatedbyT.E.KleinandW.E.Pohl).TheHague:MartinusNijhoff.——1983[1913]:IdeasPertainingtoaPurePhenomenologyandtoaPhenomenologicalPhilosophy.FirstBook(translatedbyF.Kersten).TheHague:MartinusNijhoff.——1989[1952]:IdeasPertainingtoaPurePhenomenologyandtoaPhenomenologicalPhilosophy.SecondBook(translatedbyR.RojcewiczandA.Schuwer).Dordrecht:KluwerAcademic.——1999:TheEssentialHusserl:BasicWritingsinTranscendentalPhenomenology(editedbyD.Welton).Bloomington:IndianaUniversityPress.WritersonHusserlBernet,R.,Kern,I.andMarbach,E.1993:AnIntroductiontoHusserlianPhenomenology.Evanston,IL:NorthwesternUniversityPress.Cairns,D.1976:ConversationswithHusserlandFink(editedbyR.M.Zaner).TheHague:MartinusNijhoff.Derrida,J.1973[1967]:SpeechandPhenomena,AndOtherEssaysonHusserl’sTheoryofSigns(translatedbyD.B.Allison).Evanston,IL:NorthwesternUniversityPress.——1989[1962]:EdmundHusserl’s‘OriginofGeometry’:AnIntroduction(translatedbyJ.P.Leavey,Jr.).Lincoln:UniversityofNebraskaPress.Dreyfus,H.L.andHall,H.(eds)1982:Husserl,Intentionality,andCognitiveScience.Cambridge,MA:MITPress.Føllesdal,D.1969:Husserl’sNotionofNoema.JournalofPhilosophy,66,20,680–7.ReprintedinH.L.DreyfusandH.Hall(eds)1982:Husserl,Intentionality,andCognitiveScience.Cambridge,MA:MITPress.——1982:Husserl’sConversionfromPsychologismandtheVorstellung–Meaning–ReferenceDistinction:TwoSeparateIssues.InH.L.DreyfusandH.Hall(eds)1982:Husserl,Intentionality,andCognitiveScience.Cambridge,MA:MITPress.Frege,G.1984[1894]:ReviewofE.G.Husserl,PhilosophiederArithmetikI.CollectedPapersonMathematics,Logic,andPhilosophy(editedbyB.McGuinness).Oxford:Blackwell.Kern,I.1964:HusserlundKant:EineUntersuchungüberHusserlsVerhältniszuKantundzumNeukantianismus(Phenomenologica16).TheHague:MartinusNijhoff.Miller,I.1984:Husserl,Perception,andTemporalAwareness.Cambridge:MITPress.Smith,D.W.andMcIntyre,R.1982:HusserlandIntentionality:AStudyofMind,Meaning,andLanguage.Dordrecht:D.Reidel.Smith,D.W.andSmith,B.(eds)1995:TheCambridgeCompaniontoHusserl.Cambridge:CambridgeUniversityPress.Spiegelberg,H.(ed.)1972:FromHusserltoHeidegger:Excerptsfroma1928FreiburgDiarybyW.R.BoyceGibson.JournaloftheBritishSocietyforPhenomenology,2,1.——1982:ThePhenomenologicalMovement:AHistoricalIntroduction,3rdrevdedn.TheHague:MartinusNijhoff.856\nHUSSERLANDHEIDEGGERHeideggerHeidegger,M.1959[1954]:AnIntroductiontoMetaphysics(translatedbyR.Manheim).NewHaven,CT:YaleUniversityPress.——1962[1927]:BeingandTime(translatedbyJ.MacquarrieandE.Robinson).NewYork:HarperandRow.AlsotranslatedbyJ.Stambaugh.Albany:StateUniversityofNewYorkPress,1996.——1968[1954]:WhatIsCalledThinking?(translatedbyJ.G.Gray).NewYork:HarperandRow.——1971a:Poetry,Language,Thought(translatedbyA.Hofstadter).NewYork:HarperandRow.——1971b[1959]:OntheWaytoLanguage(translatedbyP.D.Hertz).NewYork:HarperandRow.——1972[1969]:OnTimeandBeing(translatedbyJ.Stambaugh).NewYork:HarperandRow.——1977:TheQuestionConcerningTechnologyandOtherEssays(translatedbyW.Lovitt).NewYork:HarperandRow.——1979–82[1961]:Nietzsche,4vols(editedbyD.F.Krell).SanFrancisco:HarperandRow.——1982:TheBasicProblemsofPhenomenology(translatedbyA.Hofstadter).Bloomington:IndianaUniversityPress.——1988[1975]:TheBasicProblemsofPhenomenology,revdedn,Lectures,1927(translatedbyA.Hofstadter).Bloomington:IndianaUniversityPress.——1993:BasicWritings,revdedn(editedbyD.F.Krell).SanFrancisco:HarperandRow.——1994[1984]:BasicQuestionsofPhilosophy:Selected‘Problems’of‘Logic’.Lectures,1937–8(translatedbyR.RojcewicaandA.Schuwer).Bloomington:IndianaUniversityPress.——1998[1967]:Pathmarks(editedbyW.McNeill).Cambridge:CambridgeUniversityPress.WritersonHeideggerBlattner,W.1999:Heidegger’sTemporalIdealism.Cambridge:CambridgeUniversityPress.Carman,T.forthcoming:Heidegger’sAnalytic:Interpretation,Discourse,andAuthenticityin‘BeingandTime.’Cambridge:CambridgeUniversityPress.Dreyfus,H.L.1991:Being-in-the-World:ACommentaryonHeidegger’s‘BeingandTime’,DivisionI.Cambridge:MITPress.Dreyfus,H.L.andHall,H.(eds)1992:Heidegger:ACriticalReader.Oxford:Blackwell.Guignon,C.B.1983:HeideggerandtheProblemofKnowledge.Indianapolis,IN:Hackett.——(ed.)1993:TheCambridgeCompaniontoHeidegger.Cambridge:CambridgeUniversityPress.Haar,M.1993[1990]:HeideggerandtheEssenceofMan(translatedbyW.McNeill).Albany:StateUniversityofNewYorkPress.Kisiel,T.1993:TheGenesisof‘BeingandTime’.Berkeley:UniversityofCaliforniaPress.Lafont,C.2000:Heidegger,Language,andWorld-Disclosure(translatedbyG.Harman).Cambridge:CambridgeUniversityPress.Mulhall,S.1996:RoutledgeGuidebooktoHeideggerand‘BeingandTime’.London:Routledge.Okrent,M.1988:Heidegger’sPragmatism:Understanding,Being,andtheCritqiueofMetaphysics.Ithaca,NY:CornellUniversityPress.Ott,H.1993[1988]:MartinHeidegger:APoliticalLife(translatedbyA.Blunden).NewYork:BasicBooks.Pöggeler,O.1987[1963]:MartinHeidegger’sPathofThinking(translatedbyD.MagurshakandS.Barber).AtlanticHighlands,NJ:HumanitiesPress.Richardon,J.1986:ExistentialEpistemology:AHeideggerianCritiqueoftheCartesianProject.Oxford:ClarendonPress.857\nTAYLORCARMANSafranski,R.1998[1994]:MartinHeidegger:BetweenGoodandEvil.Cambridge,MA:HarvardUniversityPress.Schürmann,R.1987[1982]:HeideggeronBeingandActing:FromPrinciplestoAnarchy(trans-latedbyC.-M.Gros).Bloomington:IndianaUniversityPress.Sluga,H.1993:Heidegger’sCrisis:PhilosophyandPoliticsinNaziGermany.Cambridge,MA:HarvardUniversityPress.Young,J.1997:Heidegger,Philosophy,Nazism.Cambridge:CambridgeUniversityPress.——2001:Heidegger’sPhilosophyofArt.Cambridge:CambridgeUniversityPress.DiscussionQuestions1Whatistherelationbetweenourmathematicalconceptofnumber,ontheonehand,andourperceptionofaggregates,ontheother?Husserlobservesthatmostpeoplecandifferentiateonlyuptoaboutadozenparticularsatatimewithoutresortingtocounting.Isthisfactrelevanttoarithmetic?2Whatistherelationbetweenlogicandpsychology?Dobothdisciplinesstudythewaywethink?Dotheybothdescribe‘laws’ofthought?Dologicalnormsstandinneedoftheoreticalfoundationsatall,whetherinpsychologyorinsomeotherpurelytheoreticaldiscipline?3Whatis‘intentionality’?Ifallconsciousnessisconsciousnessofsomething,whatshouldwesayaboutdreams,hallucinations,andfalsememoriesandexpectations?Aresuchattitudesnotintentionalafterall,thoughtheyseemtobe?Areallourattitudesdirectedtowardmentaltokensinternaltoourminds?4WhydoesHusserlinsistthatamerelypsychologicaltheoryofmentalactswillalwaysfailtocapturetheintentionalcontentofthoughtandperception?Whydoeshedistinguishtheintentionalcontent,or‘actmatter’,ofamentalstatefromitspyschologicalmode,or‘actquality’?Dothetwocomponentsvarycompletelyindependentlyofoneanother?5WhydoesHusserldistinguishbetweenthecontentandtheobjectofanintentionalstate?Howdoesthatdistinctionundercutpotentialproblemsaboutouraccesstotheexternalworldandabouttheontologicalstatusofso-called‘intentionalobjects’?6Whatis‘categorialintuition’?Isitnecessarytosupposethatwecanintuitanythingfulfillingthesignifyingsenseoftermslike‘is’,‘not’,‘and’,‘or’,‘all’,‘some’and‘none’?7Whatisthe‘eideticreduction’?Whatistheepoche¯,or‘transcendentalreduction’?Howdothetworeductionspurporttorevealpureconsciousness,ortranscendentalsubjectivity?Whatisthenoema?8Isallofourexperiencecentredaroundatranscendentalego?WhatmotivatedHusserl’sinitialdenialthatreflectionrevealsanysuchpureego?Whatthenmotivatedhislateradmissionthatallconsciousnessmustbeownedbyan‘I’?9HowdoesHusserldescribeourconsciousnessofothersubjectiveselves?Howdowecometounderstandoursubjectivityasconstitutinganintersubjectivity?Ishisaccountplausible?10Whatisthe‘lifeworld’?Howarewetounderstandthesenseandjustificationofscientifictheoriesthatdepictaworldradicallydifferentfromtheworldgiventousinordinaryexperience?Canourordinaryexperienceeverbepartlywrongaboutthewaytheworldis?Canitbewhollymistaken?858\nHUSSERLANDHEIDEGGER11WhataretheprincipaldifferencesbetweenHeidegger’sphenomenologyandHusserl’s?IsHeideggerrighttorejectthedistinctionsbetweenimmanenceandtrans-cendence,betweentherealandtheideal,andbetweenphenomenologyandontology?Whatdoesitmeantosay,‘ThephenomenologyofDaseinisahermeneutic’?12HowdoesthetemporalityofDasein’sbeing-in-the-worlddifferfromthatoftheavailableequipmentandoccurrentobjectsweencounterinoureverydayexperience?WhydoesHeideggermaintainthattheavailabilityofusefulartefactsandtoolscannotbeanalysedintermsoftheiroccurrentpropertiesandrelations?13Howdoesananxious,resoluteanticipationofdeathallowoneto‘ownup’toone’sexistenceauthentically?Isauthenticityagoodthing?Why?14Whatisaworkofart?Whatdoesaworkofartaccomplish?IsHeidegger’snotionof‘art’consistentwithourordinaryuseoftheword?15How,accordingtoHeidegger,isthemodernconceptionoftheworldasa‘picture’rootedintheprojectionofasystematicgroundplanforrigorousscientificresearch?Howdoesthenormativestructureofscientificpracticepresupposethegivennessoftheworldasarepresentationtoanautonomoussubject?16Whatisthe‘technological’understandingofbeing?Howdoesthetechnologicalagedifferfromtheageofthe‘worldpicture’?Whatdoesitmeantotreateverythingasresourcematerial,or‘standing-reserve’?Isitwrongtodoso?Isitpossibletobeopentotechnologyasanunderstandingofbeing,asawayofrevealingentities,whilecontinuingtomakeuseoftechnologicaldevices?859\n42Sartre,FoucaultandDerridaGARYGUTTINGDuringthesecondhalfofthetwentiethcentury,FrenchphilosophywasdominatedbyJean-PaulSartre,MichelFoucaultandJacquesDerrida.Sartrewastheleadingfigureofexistentialphenomenology,anapproachthat,withimportantdebtstotheGermansHUSSERLANDHEIDEGGER(chapter41),triedtodevelopacomprehensivephilosophicalvisionofhumanbeingsandtheirplaceintheworldthroughthecarefuldescriptionofourconcreteexperience.FoucaultandDerridarepresent,indifferentways,whathascometobecalledthepoststructuralistreactionagainstphenomenology.1SartreSartrewasimportantnotonlyasaphilosopherbutalsoasaliteraryfigureandapoliticalactivist.Ourdiscussionwillberestrictedtohisphilosophyand,withinphilosophy,towhatisbyfarhismostimportantandmostinfluentialachievement,BeingandNothingness.Sartre’spositionderivesfromhistwofundamentalclaimsaboutconsciousness:thatitisalwaysofsomething,butthatititselfisnotsomething.CONSCIOUSNESS(pp.185–7)isalwaysofsomethinginthesensethatformetobeconsciousimpliesthatthereissomeobject–typicallysomethingreal,thoughinsomecasessomethingimaginaryorillusory–thatIamconsciousof.Consciousnessis,inthelanguageofHusserl’sphenomenology,essentiallyintentional–directedtowardsomethingelse.Intentionalityisarelation,butitcannotbeunderstoodonanalogywithordinaryrelationsbetweenthingsintheworld;forexample,therelationwherebyaboxisontopofatableorafishisinastream.Thisisbecauseconsciousnessisnotathing,notamaterialthingbutalsonotanimmaterialthingsuchasasouloraspiritualsubstance.Itisnotathingbecauseitsentireexistenceisexhaustedbyitsrelationtoitsobjects.Ithasnocontentorstruc-tureofitsown.Nordoesittakeoncontentorstructurebysomehowincorporatingitsobjects.Ourordinarytalkofwhatweexperienceorthinkabout‘beinginthemind’ismisleading.Intypicalcases(forexample,senseperception,non-deceptivememory)theobjectexperiencedexistsoutsidethemindintherealworld.Butevenwhen,asinimaginationorillusions,thereisnorealobject,theobjectisnotliterallyinthemind.\nSARTRE,FOUCAULTANDDERRIDAConsciousnessisatotally‘transparent’intendingofitsobjectsandnothingmore.Inviewofthis,Sartreispreparedtosaythatconsciousnessisnothing.Consciousnessisalsotransparentinanothersense.Itisalwaysaware,directlyandimmediately,ofitselfasconsciousness.AsSartreputsit,tobeconsciousistobeself-conscious.Buthereweneedtobecareful.Self-consciousnessmightseemtomeanconsciousnessofself,where‘self’referstoconsciousnessasanobjectofawarenesslikeanyother.Butthenconsciousnesswouldalsobeathinglikeanyother,andthisofcourseSartrefirmlydenies.Becauseconsciousnessisalwaysself-aware,Sartresaysthatithasbeing-for-itself:itsveryexistenceinvolvesaninternalrelationtoitself.Theobjectsofconsciousnessarenotself-aware(evenwhenweareawareofotherpeople,theyarenotconsciouspreciselyasobjectsofourawareness).Ontheotherhand,unlikeconsciousness,objectsofconsciousnessarethings,withthepresenceandsolidityofintrinsiccontent;theyhave,inSartre’sterminology,being-in-itself,whichconsciousnessofcourselacks.Sartre’saccountseemstoignoreanobviousfeatureofourexperience:theperson(orpsychologicalself)asthesubjectoftheproperties(habits,charactertraits,beliefs,inclinations)thatdefineusasindividuals.Infact,Sartredoesnotignorethisobviousreality,buthedisplacesit.Ratherthanidentifyingitwith(orsituatingitwithin)con-sciousness,hemaintainsthattheselfexistsonlyasanobjectofconsciousness,thatitisapartoftheworld,likeanyotherthing.Weneed,finally,tosayabitmoreabouttherelationofbeing-in-itselftobeing-for-itself.Ourordinaryexperienceisofobjectsthathaveabasicintelligibility;forexample,patternedorfunctionalbehaviourinnature,apurposeinourlivesthatgivethemameaning.InSartre’sview,however,thisintelligibilitydoesnotbelongtotheobjectinvirtueofitsmostbasicrealityasbeing-in-itself.Onthefundamentallevelitisabrute,unstructuredgiven,merelyexistingwithnointrinsicmeaning.Onthislevelweshouldnot,infact,speakofdifferentthings,sincethestructurenecessaryfordifferentiationisnotpresent.Thereisjustsheerindistinctbeing-in-itself.IntheliteraryphenomenologyofhisnovelNausea,Sartrecharacterizestheunintelligibilityofbeing-in-itselfintermsofsuperfluity,absurdityandcontingency.Being-in-itselfismeaningful(anddividedintodiscrete,intelligibleobjects)onlyinsofarasitistheobjectofconsciousness.Consciousnessis,therefore,theultimatesourcenotoftherealityofbeing-in-itselfbutofitsmeaning.Sartreunderstandstherelationofbeing-for-itself(consciousness)tobeing-in-itselfintermsofnegativity(nothingness).ConsideraparadigmcasehediscussesinBeingandNothingness:Ienteracafé,lateforanappointmentwithPierre,whoismeticulouslypunctual.Iatfirstseemtoencounterafullnessofbeing;everywhereIlookthereareobjectsoractivities.But,sinceIamlookingforPierreandworriedthathemayhavealreadyleft,everyelementofthisscenefallsback,assoonasitbeginstopresentitself,becauseIseeitasnot-Pierre.NothingnessSartreintroducestheneologism‘nihilation’(néantisation)todenotetheprocesswherebynegationisintroducedontheconcretelevelofimmediateperception(asopposedtothereflectivelevelofintellectualjudgement).Correspondingly,nothingnessishistermforthe861\nGARYGUTTINGontologicalrealityofnegationthatisintroducedbynihilation.ThefactthatmyexperienceofthePierrelesscaféinvolvesnihilation–ofboththeabsentPierreandofthecaféthatlackshim–showsthatnegationexistsasaconcreterealityandthatthereforenothingnessisrequiredasadistinctontologicalcategory,inadditiontobeing-in-itselfandbeing-for-itself.Nothingnessmust,therefore,existlike‘awormintheheartofbeing’.Buthowdoesitarise?Certainlynotfromitself,throughsomeincoherentself-nihilation,norfrombeing-in-itself,whichisentirelyenclosedinitsinertgivenness.Nothingnessmustthereforebesomehowderivedfromconsciousness(andhencefrombeing-for-itself),aconclusionsupported,moreover,byourexperienceofnegationsasarisinginthefaceofourexpectationsandfears.Butconsciousnessisalwaysofbeing-in-itselfandsoitmustgiverisetonothingnessbynegatingbeing-in-itself.Thisnegationisnot,ofcourse,amatterofliterallydestroying(annihilating)being-in-itself.Rather,consciousnessnegatesbywithdrawingfrombeing.Sartreidentifiesthiswithdrawalofconsciousnessfrombeingwithfreedom,sincefreedomisatranscendingofthedeterminismofcausallawsandtheselawsexistonlyasstructuresofbeing-in-itself.FREEDOM(pp.206–8)existsasconsciousness’sabilitytowithdrawfrom(reviseorevenreject)theself.BecauseIamfree,IcandenywhatIam(thatis,whatIhavebeenuptonow)andconstituteatanymomentanewmeaningformyexistenceasaself.Negationenterstheworldinvirtueofconsciousness’schoicetomakeitsselfthisandnotthatinrelationtotherestoftheworld.AnotherwaytoputSartre’spointistosaythatconsciousnessisentirelyfreeinthesensethatitistheultimatesourceofallmeaning.Freedomisnotamatterofdeterminingwhathappensbutofdeterminingthemeaningofwhathappens.Consciousnessobviouslyhasnocontroloverthecausesthat,say,cripplemeordestroyatownbyaflood.Butitdoes,Sartrethinks,absolutelydeterminethemeaningofsuchevents.WearelikelytorespondtoSartrethatwedonottypicallythinkofourselvesaspossessingsuchradicalfreedom.Surelymuchofwhattheworldmeansformeisbeyondmycontrol!Butthis,hewouldreply,resultsfromoureffortstofleefromthetruth,topretendtoourselvesthatwearenotreallyfree,despiteourimplicitawarenessofourfreedom.Thisflightfromfreedomexpressesthewidespreadbehaviourofbadfaith(mauvaisefoi,sometimestranslatedas‘self-deception’).Sartredoesnotmaintainthatbadfaithisinevitable,butheclearlydoesregarditasverycommon,atleastinourcurrentsocialsituation,andtakeselaboratepainstodescribethemanyvarietiesandsubtletiesofitsmanifestations.Hisoverridingpoint,however,isthatbadfaithisbasedonthedistinctionofbeing-for-itselfandbeing-in-itselfandtheessentialrelationoftheformertothelatter.Badfaithispossiblebecausebeing-for-itselfischaracterizedbyafunda-mentaldualitywherebyitisbothseparatedfromandidentifiedwithbeing-in-itself.Sartre’sradicalunderstandingofthefreedomfromwhichbadfaithfleesisattheheartofhisexistentialism.Heacceptsthetraditionalideathatfreedomrequiresfreedomfromastandardsetofevils:theliesandvaluesofpopularmaterialisticculture(inSartre’stermsbadfaithandbourgeoismorality).Butherejectsthetraditionalviewthatfreedommustalsoberegardedasfreedomforsomeobjectivesummumbonum(truth,happiness,virtueandsoon).Althoughaccountsofthesummumbonumhavedifferedconsiderably,theyallseefreedomasdirectedtowardanobjective,pregivenorder(PlatonicForms,humannature,scientificlaws,divineprovidence).Bycontrast,862\nSARTRE,FOUCAULTANDDERRIDASartredeniesthatthereisanyobjectiveorderofthingsthatisnormativeforhumanfreedom.Accordingly,thecentralpositivefeaturesofthetraditionalconceptionoffreedom(freedomfor)becomenegativefeatures(freedomfrom)forSartre.Specifically,SartreanfreedomrequiresfreedomfromGod(anti-theism)andfreedomfromobjectiveethicalvalues(anti-moralism).ForSartre,theanti-theisticdimensionisaxiomatic.HesimplyfindstheveryideaofaGodentirelyincredible.Heoccasionallysuggestsargumentsforhisatheism.Butheplaceslittleweightonargumentandmostlytreatshisatheismassimplyabasicstarting-point.Atthemost,wemightsaythatSartrestartswithafundamentalexperienceofhumanbeingsassoradicallyfreethattheideaoftheirbeingcreatedbyaGodmakesnosense.Givenatheism,Sartrethinksthatanti-moralityfollowslogicallybecauseoftheDostoyevskianpremise:ifGoddoesnotexist,everythingispermitted.Sartrefurthermaintainsthathumanactionsareentirelyfreeinthesensethatthereisnothingoutsidemychoicethatconstrainsit.Thisdoesnot,however,meanthatfreedomhasnolimits,onlythatanylimitsfreedomhasarethemselvestheresultoffreechoices.Morefully,theideaisthatanyobstacletoorconstraintonmyfreeactissuchonlyasaresultofmydecisiontoact.Forexample,amountainisanobstacleonlyonceIhavedecidedtoclimbit.Mydecisionmakesanobstacleofwhatwouldotherwisebemerely,say,anobjectofmycuriosity,indifferenceoraestheticsensibility.(Sothemountainisnoobstacletothetouristwhoismerelyinterestedinadmiringitsgrandeur.)Wemightacceptthispointaboutthingsoutsidethemind,butwhataboutinternal,psychologicalcausesofbehaviour?Surelysuchthingsasanoverpoweringfeeling,anobsessivedesire,adeeplyingrainedcharactertraitdeterminemetoactincertainways?Imay,forexample,reallywanttokeepclimbingthemountain,butmyfeelingoffatiguemaybesogreatthatIsimplycannotgoon.Sartre,however,suggeststhatwhetherornotthefeelingis‘sogreat’astostopmedependsonwhetherornotIreallywanttogoonnomatterwhat.Infact,Icouldkeeppushingmyselftothepointofcollapseorevendeath.ThefatigueisanobstacleonlyifIallowittobe.AsradicalasSartre’sfreedomis,itwouldseemtohavenosignificanceforusifitisnotdeployedforsomepurpose.Butaswehaveseen,Sartreexcludesalltraditionalgoalsforhumanlife.How,then,canheseeourexistenceasmeaningful?Theanswer,putbriefly,isthatheexcludesallexternalgoalsforfreedombuttakesfreedomitselfasitsowngoal.Inotherwords,freedomceasestobeameanstosomethingelse(freedomfor)andbecomesanabsoluteendinitself.Asaresult,Sartre’sfundamentalethicalcategorybecomesauthenticity,thatis,acceptanceofone’sstatusasradicallyfree.Authenticityisopposedtobadfaith(self-deception),whichinvolvestheefforttodenyorfleefromone’sfreedom.Sartre’sexistentialphenomenology(andrelatedviewsdevelopedbyhisassociates,SimonedeBeauvoirandMauriceMerleau-Ponty)dominatedFrenchthoughtfromtheendoftheSecondWorldWaruntilabout1960.Afterthat,therewasasharpreactionagainsthisviews,ledbythinkersoftenlabelledasstructuralistsorpoststructuralists.Structuralismwasashort-lived,mostlyFrenchphenomenonofthe1960sthatclaimedtogivevarioussocialscientificandhumanisticdisciplinesascientificbasisbyofferingrigorousaccountsoftheobjectivestructuresunderlyingtheirdomains.Itrejectedtheessentiallysubjectiveviewpointofphenomenologybyfocusingonstructuresthatlieoutsidethescopeofconsciousawareness.Poststructuralismretainsthestructuralist863\nGARYGUTTINGstyleofobjective,technical,andevenformaldiscourseaboutthehumanworld,butrejectsthestructuralistclaimthatthereisanydeeporfinaltruththatsuchdiscoursecanuncover.SincethelatterclaimischaracteristicnotonlyofstructuralismbutofthehistoricalmainstreamofWesternphilosophy,poststructuralismhasamuchwidersignificancethanitsnamesuggests.Aspoststructuralists,FoucaultandDerridabothrejectthetraditionalprojectofphilosophy,althoughtheyalsoproposealternativeemploymentsforitsintellectuallegacy.Theyseek,wemightsay,notacontinuationofphilosophybyothermeansbutacontinuationofphilosophy’smeansforotherends.ButFoucault’sworkiscentrifugalinrelationtophilosophy,movingawayfromitstraditionalaporiaetowardsuccessorprojects(‘archeologies’and‘genealogies’ofknowledge).Derrida’swork,bycontrast,iscentripetal,relentlesslycomingbacktodissectthebodyoffailedphilosophicalknowledge.2FoucaultFoucaultrejectsthestandardphilosophicalprojectofdiscoveringnecessaryoressentialtruthsaboutourselvesandourworldinfavouroftheinverseprojectofdiscoveringcasesinwhichwhatarepresentedasnecessarytruthsaboutourconditionareinfactonlycontingentproductsofourhistoricalsituation.Hiswritings,therefore,donotofferacoherentvisionofwhatwemustbe,butratheraseriesofhistoriesdesignedtoshowhowwemightbedifferent.Wewilllookbrieflyatthebooksinwhichhedevelopshisfourmajorhistories:TheHistoryofMadnessintheClassicalAge,TheOrderofThings,DisciplineandPunishandTheHistoryofSexuality.FoucaultseesthehistoryofmadnessinEuropeascharacterizedbytwomajor‘breaks’:oneinthemid-seventeenthcenturywhichsharplyseparatesclassicalviewsofmadnessfromthoseofthemiddleagesandRenaissance;anotherattheendoftheeighteenthcenturythatinauguratesthemodernviewofmadness.Histreatmentofpre-classical(medievalandRenaissance)madnessiscursory.Buthedoesmakeoneclaimcrucialforhisargument:priortotheclassicalage,madnesswasseenasanintegrallyhumanphenomenon.Madnesswasopposedtoreason,butasanalternativemodeofhumanexistence,notasimplerejectionofit.IncontrasttothemedievalandRenaissanceviews,theclassicalagesawmadnessasmerelythenegationoftheessentialhumanattributeofreason.Itwasregardedasunreason(déraison),aplungeintoananimalitythathadnohumansignificance.Therewas,accordingly,aconceptualexclusionofthemadfromhumansociety.HereFoucaultcitesasaprimeexampleDESCARTES’s(chapter26)rejectionintheFirstMeditationofthepossibilityofhisownmadnessasagroundsofdoubt.Correlativetothisconcep-tualexclusion,therewasaphysicalexclusionofthemadeffectedbytheirconfinementininstitutionsthatisolatedthemfromordinaryhumanlife.Theconceptualandphysicalexclusionofthemadreflectedamoralcondemnationofthem.Themoralfault,however,wasnottheordinarysort,wherebyamemberofthehumancommunityviolatesoneofitsbasicnorms.Rather,madnesscorrespondedtoaradicalchoicethatrejectedhumanityandthehumancommunityintotoinfavourofalifeofsheer(non-human)animality.Ontheclassicalview,theanimalityofthemadwas864\nSARTRE,FOUCAULTANDDERRIDAexpressedintheirdominationbypassions,adominationthatledthemtoadeliriuminwhichtheymistooktheunrealforthereal.Passionatedeliriumthusresultedinafundamentalblindnessthatcutthemadofffromthelightofreason.Soconstrued,classicalmadnessisanafflictionofthemind–bodycomposite,notsomethingspecificallypsychologicalorsomatic.Inparticular,itisnotconceivedasa‘mentalillness’.Withthemodernage(roughly,fromthenineteenthcenturyon),themadareonceagainregardedaswithinthehumancommunity(notasanimalsbeyondthepaleofhumanity).But,withinthehumancommunity,theyarenowseenasmoraloffenders,violatorsofspecificsocialnorms,whoshouldfeelguiltattheirconditionandwhoneedreformoftheirattitudesandbehaviour.Correspondingtothisnewconceptionofmadnessisthecharacteristicmodernmodeoftreatingthemad:notmerelyisolatingthembutmakingthemtheobjectsofamoraltherapythatwouldsubjectthemtosocialnorms.Thisisthemovefromthemerelycustodialconfinementoftheclassicalagetothetherapeuticasylumofthemodernperiod,foundedbytheTukesinEnglandandbyPinelinFrance.Althoughthetherapeuticasylumiswidelyregardedasanunquestionableadvanceinhumanitarianism,Foucaultseesitasmerelyamoresubtleandthoroughmethodforcontrollingthemad.Ratherthanatrueliberationofthemad,itisa‘giganticmoralimprisonment’.Wereadilyseethat,forFoucault,theidentificationofmadnessasmentalillnessisnotprimarilyanobjectivescientificdiscovery.Rather,itwasintroducedasameansoflegitimatingtheauthorityofphysiciansintheasylumoncetheideaofadistinctivelymoraltherapywasabandoned.Morefully,thefactthatphysicianscametobeinchargeofasylumsinitiallyhadlittletodowiththeirmedicalexpertise.ThemoraltreatmentrecommendedbyTukeandPinelwasnotessentiallymedicalandcouldbecarriedoutbyanypersonwithmoralauthority.However,asthenineteenthcenturydeveloped,medicinebecamedominatedbytheidealofobjective,value-freeknowledge,whichleftnoroomforvalue-ladenmoraltherapies.Theideaofadistinctivelymentalsortofillnesswasintroducedprimarilytojustifythecontinuingauthorityofdoctorsoverthemad,notbecauseofitsscientifictruthorcurativesuccess.TheOrderofThingscoversroughlythesamechronologicalperiodsasTheHistoryofMadness:theRenaissance,theclassicalageandthemodernage.Foreachperiod,Foucaultsketchesthegeneralepistemicstructure(theepistemé)underlyingitsthoughtandthenshowshowthedisciplinesthatarethecounterpartsoftoday’shumansciencescanbeunderstoodintermsofthisbasicepistemicstructure.This,hemaintains,showsthattheseearlierdisciplinesarenothaltinganticipationsofthemodernhumansciencesbutautonomousalternativesforconstruinghumanreality.FromthishewillargueforthecontingencyandreplaceabilityoftheHUMANSCIENCES(chapter12).Foucault’sArcheologyMethodologically,TheOrderofThingsisthefullfruitionofthearcheologicalmethodtowardwhichFoucaultwasgropinginTheHistoryofMadness(andalsoTheBirthoftheClinic).Archeologyemergesasamethodofanalysisthatrevealstheintellectualstructuresthatunderlieandmakepossibletheentirerangeofdiverse(andoftenconflicting)concepts,methodsandtheoriescharacterizingthethoughtofagivenperiod.865\nGARYGUTTINGThelevelofconcepts,methodsandtheoriescorrespondstotheconsciouslifeofindividualsubjects.Byreadingtextstodiscovernottheintentionsoftheirauthorsbutthedeepstructureofthelanguageitself,Foucault’sarcheologygoesbeneathconsciouslifetorevealtheepistemic‘unconscious’thatdefinesandmakespossibletheknowledgeofindividuals.AsFoucaultexplainsinhissubsequentmethodologicaltreatise,TheArchaeologyofKnowledge,archeologyissimilartologicandgrammarinthatitdiscoversrulesgoverningourdiscursivebehaviourofwhichwemaywellnotbeaware.Grammarformulatestherulesdefiningthedomainofsheermeaningfulnessandlogictherulesofsheerconsistency.Butitisobviousthattherearemanygrammaticallyandlogicallyacceptablestatementsthatareneverutteredinagivendomainofdiscourse.Wemaybeinclinedtoattributethissimplytothefactthatnoindividualshappentowanttomakethesestatements.ButFoucaultmaintainsthat,inadditiontotherulesofgrammarandlogic,therearefurtherunderlyingruleslimitingtherangeofpermissiblestatements.(Forexample,therulesofclassicaldiscourseaboutmadnessdidnotpermitthestatement:‘Madnessissimplyadiseaseofthemind’.)Foucault’sarcheology,thearcheologyofknowledge,isthehistoricalmethodthatuncoverssuchrules.WithDisciplineandPunishFoucault’sworkreturnstotheexplicitlyethicalmotivationsofTheHistoryofMadness,butnowwithafirmsenseofhowtoincorporatecausalaccountsofchangesinknowledgesystems.Hisparticularfocusistheemergenceofthemoderndisciplineofcriminologyandotherrelatedsocialscientificdisciplines.WithabowtoNIETZSCHE(chapter40),hecharacterizeshisnewapproachasgenealogical.Thegeneralideaishardlyoriginal:thatshiftsinthepowerstructuresofasocietyproducechangesinepistemicformations.Whatisnewishisunderstandingofthenatureofpowerandoftheprecisewayinwhichpowerandknowledgearerelated.Iwillherecommentbrieflyonthelatterpoint,butdefertheformertoourdiscussionofFoucault’shistoryofsexuality.Genealogydealswiththeconnectionbetweennon-discursivepracticesandsystemsofdiscourse(bodiesofknowledge).Inthisregard,Foucault’scentralclaimisthatthereisaninextricableinterrelationofknowledge(discourse)andpower(expressedinnon-discursivepractices,inparticular,thecontrolofbodies).Thisiswhyanunderstandingofhisgenealogicalapproachrequiresanunderstandingofhisviewoftherelationofknowledgeandpower.Negatively,FoucaultdoesnothaveinmindthestandardBACONIAN(chapter25)idea,whichseesknowledgeasfirstexistingasanautonomousachievement,whichisthenusedasaninstrumentofaction(forexample,puresciencevs.technology).Hemaintainsthatknowledgesimplydoesnotexistincompleteindependenceofpower,thatthedeploymentofknowledgeandthedeploymentofpoweraresimultaneousfromthebeginning.Ontheotherhand,Foucaultdoesnotgosofarastoidentifyknowledgewithpower;forexample,tomakeknowledgenothingmorethananexpressionofsocialorpoliticalcontrol.Ashehimselfnoted,ifhethoughttherewasnodifferencebetweenknowledgeandpower,hewouldnothavehadtotakesuchpainstodiscovertheprecisewaysinwhichtheyarerelatedtooneanother.Hispositiveviewisthatsystemsofknowledge,althoughexpressingobjective(andperhapsevenuniversallyvalid)truthintheirownright,arenonethelessalwaysmoreorlesscloselytiedtotheregimesofpowerthatexist866\nSARTRE,FOUCAULTANDDERRIDAwithinagivensociety.Conversely,regimesofpowernecessarilygiverisetobodiesofknowledgeabouttheobjectstheycontrol(butthisknowledgemay–initsobjectivity–gobeyondandevenultimatelythreatentheprojectofdominationfromwhichitarises).Foucaultprettyclearlyintendsthisinterconnectionofknowledge/powerasageneralthesis,buthedefendsitandisspecificallycommittedtoitonlywithrespecttoparticularmoderndisciplines.InDisciplineandPunishFoucaultdealswithknowledge/powerwithrespecttotheconnectionbetweenthedisciplinarypracticesusedtocontrolhumanbodiesinthemodernperiodandmodernsocialscientificdisciplines.Hisprimaryexampleisthepracticeofimprisonmentasawayofpunishingcriminalsinitsrelationtocriminologyandothersocialscientificdisciplinesrelevanttocrimeandpunishment(socialpsychologyandsoon).Buthedevelopshisdiscussionofimprisonmentinthecontextofadiscussionofmoderndisciplinarypracticesingeneral(asemployed,forexample,inschools,factoriesandthemilitary);andheshowshowtheprisonservedasamodelandcentreofdiffusionforthiswholerangeofdisciplinarypractices.Asalways,Foucaultbeginsbycontrastingthemodernagewiththeclassicalageimmediatelypreceding.Henotesthatthemoststriking(andessential)differencebetweenthetwoperiodswastheviolentandflamboyantlypublicnatureofthepunishmentofcriminalsofthepremodernperiodincontrasttothephysicallymuchmilderand‘lowprofile’modernpunishmentofimprisonment.Whereasstandardaccountshaveattributedthisdifferenceprimarilytoamorehumane,compassionatemodernattitude(basedonnewphilosophicalideasandascientificunderstandingofcriminality),Foucault’spower/knowledgehypothesissuggeststhatthereissomethingelsegoingon(althoughheagreeshumanenessmaybeasecondaryfactor).Specifically,heexplorestheideathatpunishmentbecomesmildernotsimplyforthesakeofmildnessbutforthesakeofnew,moreeffectiveandmoreextensiveformsofcontrol.Asheputsit,thepointwasnotsomuchtopunishlessastopunishbetter.Althoughmoderndisciplinarypracticesoriginatedinisolated,enclosedinstitutionssuchasprisonsandasylums,theyrapidlyspreadthroughoutsociety,toschools,factories,medicalclinics,welfareagencies,andsoon.Inthisexpansion,thesepracticesbecamepositiveaswellasnegative,notjustpreventingdisapprovedbehaviourbutincreasingproductionandknowledge.Detachedfromparticularinstitutions,thepracticesalsofellunderthecontrolofthenationalstateandexertedtheirinfluenceonsocietyasawhole.Thegeneralresultwasthetransformationofsocietyfromoneofspectacletooneofsurveillance.InthefirstvolumeofhisHistoryofSexualityFoucaultshowshowmoderncontrolofsexualityparallelsmoderncontrolofcriminalitybymakingsex,likecrime,anobjectofallegedlyscientificdisciplines,whichsimultaneouslyofferknowledgeanddomina-tionoftheirobjects.However,itbecomesapparentthatthereisafurtherdimensioninthepowerassociatedwiththesciencesofsexuality.Notonlyistherecontrolexercisedviaothers’knowledgeofindividuals;thereisalsocontrolviaindividuals’knowledgeofthemselves.Individualsinternalizethenormslaiddownbythesciencesofsexualityandmonitorthemselvesinanefforttoconformtothesenorms.Thus,theyarecontrollednotonlyasobjectsofdisciplinesbutalsoasself-scrutinizingandself-formingsubjects.ThisisthereasonFoucaultseesourapparentlyliberatingfocusonoursexualityasjustareinforcementofthemechanismsofsocialcontrol.Intryingtodiscoverourdeep867\nGARYGUTTINGsexualnaturethroughself-scrutinyandtoexpressthisnaturebyovercomingvarioushang-upsandneuroses,wearemerelyshapingourselvesaccordingtothenormsandvaluesimplicitinmodernsciencesofsexuality.Wemaybreakwithcertainsocialconventionsandconstraints,butwedosoonlybyputtingourselvesjustasfirmlyinthecontrolofanothersystemofconstraints.ThesecondvolumeofFoucault’shistoryofsexualitywasprojectedasastudyoftheoriginsofthemodernnotionofasubjectinpracticesofChristianconfession.Foucaultwrotesuchastudy(TheConfessionsoftheFlesh)butdidnotpublishitbecausehedecidedthataproperunderstandingoftheChristiandevelopmentrequiredacomparisonwithancientconceptionsoftheethicalself.ThisledtotwovolumesonGreekandRomansexuality:TheUseofPleasure(1984)andTheCareoftheSelf(1984).ThesefinalwritingsmakeexplicittheethicalprojectthatinfactinformsallofFoucault’swork:theliberationofhumanbeingsfromcontingentconceptualconstraintsmaskedasunsurpassableapriorilimitsandtheadumbrationofalternativeformsofexistence.3DerridaAlthoughDerridadisavowstraditionalphilosophy,heremainsitsmostassiduousreader.Andheis,indeed,aboveallaremarkablereaderwithadistinctivetalentforclose,subtleandimaginativescrutinyoftexts.Whyshouldaphilosopher,particularlyoneatDerrida’shistoricalsite,besoobsessedwithwhatothershavewritten?Because,asDerridaseesit,writingrevealstheessentialpeculiaritiesandlimitationsofhumanthought.Awrittentextwillalwaysescapetotalclarification.Therewillalwaysbetextualambivalencesthatremainunresolvableandpreventusfromunderstandingfully‘whattheauthorreallymeans’.Wemaythink,asPlatosometimessuggests,thattheproblemisduesimplytothemediumofwriting.Ifwecoulddirectlyspeaktotheauthor,ourperceptionofintonationsandgesturalnuances–alongwiththepossibilityoffollow-upquestions–wouldeliminateallambivalence,allundecidability.Butofcourseevenface-to-facespeakingwillnotconveyamessageperfectly.Theinevitabledifferences(inpastexperience,inexpectations,inidiolect)betweenspeakerandhearermaintainpermanentpossibilitiesofmisunderstanding.Suppose,then,thattoeliminatethesedifferences,Iconsiderjustthecaseofmyowninternalformulationofmythoughts.Butevenhere,Derridamaintains,thelinguisticformulationwillnotbetotallyadequate.Thegeneralityofanylinguisticexpressionwillmakeitalessthanperfectexpressionoftheprecisedetailsofmythoughtortheexactnuancesofmyfeelings.Itwouldseemthatperfectadequacyisachievedonlyintheimmediate,pre-linguisticpresenceofmythoughttoitself.ButDerridaarguesthatthereisnosuchpurepresenceofthoughtstotheself.Allthoughtismediatedthroughlanguageandcanneverattainthetotalclarityofpurepresencetotheself.Thereisalwaysadifferencebetweenwhatisthought(orexperiencedorsaidorwritten)andtheidealofpure,self-identicalmeaning.TheabovelineofargumentisaprototypeofDerrida’srepeateddemonstrations,indifferentcontextsandterms,thattheapparentlycontingentandremediabledefectsofwritingareinfactinevitablefeaturesofallthought,allexpression,allreality.Derrida’s868\nSARTRE,FOUCAULTANDDERRIDAphilosophicalprojectisanunendingextrapolationofthereader’sinabilitytomasteratext.Despiteourrelentlessfailuretoattainperfectmeaningandtruth,allourthoughtandlanguageisbasedontheassumptionofanddriveforsuchmeaningandtruth.ThisassumptionanddrivecanbeformulatedbythreeprinciplesthatarecentralintheWesternphilosophicaltradition.(Derridahimselfneverformulatestheprinciplesintheseterms,buttheycatchwhathehasinmindby‘logocentric’thinking.)First,thebasicelementsofthoughtandlanguagearepairsofopposingconcepts,suchaspres-ence/absence,truth/falsity,being/nothingness,same/other,one/many,male/female,hot/cold.ThiswecancallthePrincipleofOpposition.Next,theopposingpairsareregardedasexclusivelogicalalternatives,governedbytheprinciplesofidentity(A=A)andnon-contradiction(nothingisbothAandnot-A).ThiswecancallthePrincipleofLogicalExclusion.Forexample,beingpresentexcludesbeingabsent;thepresentissimplywhatitis(present)andisinnowaywhatitisnot(absent).Finally,eachfundamentalpairisasymmetricalinthesensethatonetermhasinsomecrucialsensepriorityovertheother(forexample,ismorefundamental,morereal,morallybetterthantheother).ThisisthePrincipleofPriority.DeconstructionAtypicalDerridianreadingwillrevealtheextenttowhichagiventextdoesnotfitthemodelofalogicalsystemasdefinedbythePrinciplesofOpposition,LogicalExclusionandPriority.Specifically,itwillshowthatthebinaryoppositionsonwhichthetextisbasedarenotsustained,thattheallegedrelationsoflogicalexclusionandprioritycannotbecoherentlyformulatedandareimplicitlydeniedbytheverytextthatformulatesthem.Derridacallsthistechniquedeconstruction.DeconstructionisthusamethodofshowinghowtextsbasedonbinaryoppositionsthemselvesviolateboththePrincipleofExclusionandthePrincipleofPriority.Thus,adeconstructivereadingofatextrevealspointsatwhichitintroducesoneoftheopposingtermsintothedefinitionoftheotherorreversestheorderofprioritybetweenthetwoterms.Derridacharacterizesinvariouswaystheviewsputintoquestionbydeconstructiveanalysis.Thegeneralprojectofdeconstructingthefundamentaldichotomiesbuiltintothoughtyieldsacritiqueoflogocentrism.Thedominanttermsofthestandardpolaroppositionsalwayscorrespondtosomesortofpresence,arealitythatispositive,complete,simple,independentandfundamental(Plato’sForms,Aristotle’ssubstances,Aquinas’sGod,Hegel’sAbsolute).Thispresenceisalwaysunderstoodasthepolaroppositeofsome-thingthatisnegative,incomplete,complex,dependentandderivative(matter,creatures,appearanceandsoon).Derrida’sdeconstructiveanalysesshow,however,thatthepurityandpriorityofpresenceisneversustainedinthetextsofthegreatmetaphysicians.Forexample,PlatodiscoversthattheFormsparticipateinnon-being,orChristiansthinkofGodassomehowhumanlyincarnate.Theresultisacritiqueofmetaphysicalpresence.Derrida’sdeconstructivereadingsarecomplementedbyamorepositiveand,insomeways,evensystematicphilosophicalproject.Thisiscarriedoutthroughhisrepeatedeffortstointroducevocabulariesthatattempttoadumbratetheontologicallevelatwhichdichotomiesdissolveandtheiroppositionsreverseandslideintooneanother.Thisprojectissystematicbothinthecomprehensiveapplicabilityofeachofthe869\nGARYGUTTINGvocabulariesandinthecomplexinterconnectionsthatbindtogethertermsfromdif-ferentvocabularies.Asanexample,wewilldiscussthevocabularyofdifférance,theoneDerridahasmostdevelopedandmostoftendeploys.‘Differance’,atransliterationoftheFrenchneologismdifférance,ismeanttoevoketheinstabilityofthebinaryoppositionsfundamentaltologicalsystems.Derridadesignedthetermtohavetwobasicconnotations:differenceanddeferral.(TheFrenchverbfromwhichitderives,différer,meansbothtodifferandtodefer.)Thefirstconnotationcorrespondstothewayinwhichanypairofbinaryoppositesalwaysfailstomatchexactlythedomaintowhichitissupposedtoapply.Therearealwaysirreducibledifferencesbetweenthestructureoftheactualphenomenon(ahistoricalevent,atext,apersonality)andthebinarydivisionsrequiredbyalogicalsystem.Forexample,Rousseau’sactualuseoftheconceptsspeechandwritingdoesnotcorrespondtothesharpdivisionheclaimstomakebetweenthem.Further,effortstoimposethesharpdistinctionrequiredbybinaryoppositionsmustalwaysbe‘putoff’(deferred)inthefaceoftherecalcitranceofthephenomenon.Forexample,whenweseethatPlatohasviolatedthespeech/writingdichotomybydefiningthethoughtexpressedbyspeechasawritinginthesoul,wemaytrytofallbacktoadistinctionbetween‘goodwriting’–whichislikespeech–and‘badwriting’.Theideaisthat,evenifthedistinctionfailsatonelevel,itcanberevivedatanotherlevel(wemerely‘defer’drawingit).But,Derridamaintains,furtheranalysiswillshowthateventhe‘fall-back’distinctioncanbeunderminedandthatatrulysharpdistinctionwillremainelusive,willhavetobeindefinitelydeferred.Wesee,then,Derrida’spointinforminganounfromdifférer.Butwhydoesheintroducethe‘misspelling’différanceinsteadofthestandarddifférence?First,althoughhewantshistermtorecallthestandardone,healsowantstoemphasizethedifferencebetweenhisuseofthetermandthatofotherthinkers(Hegel,Saussure,Heidegger).Further,theainthefinalsyllablefollowsthepatterninFrenchforformingverbalnouns(gerunds),sothatdifférancemaintainsapointedambivalencebetweenanaction(amakingdifferent)andthestateresultingfromthisaction(adifference).Inthisway,Derrida’stermsuggestsarealitynotcaughtbysuchstandardmetaphysicaldichotomiesasactive/passive,event/state,action/passion.Finally,thefactthatthereisawrittenbutnospokendifferencebetweendifféranceanddifférenceevokesDerrida’sdiscussionofthespeech/writingdichotomy,andgivesaprioritytowriting(whichalonecanexpressDerrida’smeaning)thatsubvertstheforceofthestandarddistinction.Theabovewayoftalkingaboutdifferanceisneitherincoherentnormerelycute.ItreflectsratherDerrida’sconvictionthat,despitetheintrinsiclimitationsofthestandarddichotomies–signalledbytheir‘production’byadifferancethateludesthem–wehavenowayofthinkingapartfromthem.Thereisnostandpointoutsideofthedichotomiesfromwhichwecanoverlookandmasterthem.Differanceisnot,likeHegel’sAbsolute,asynthesisofalloppositesintoafullyintelligiblewhole.Itisitselfcaughtintheendlessplayofdifferences,neithercontrollingnorcontrolled,alwaysgeneratingnewparadoxes.Wecanuse‘differance’toindicatethelimitationsofourconceptsandlanguagebutnottoovercomethem.Derrida’squestioningofthedistinctionsonwhichthinkingisbasedisnotinthenameofanewsetofdefinitiveanswers(thatis,anewsetofdichotomizedconcepts),butinthenameoftheperpetualneedtobeawareofthelimitsofanyanswers.870\nSARTRE,FOUCAULTANDDERRIDAUpuntilabout1980Derridahadlittletosayaboutspecificallyethicalissues;and,incontrasttoSartreandFoucault,forexample,hedidnottiehisphilosophicalreflectionstopoliticalactivity.Butsubsequentlyhehasmadeethics(andreligion)amajortheme.SomeattentiontothisthemewillgiveusasenseofDerrida’smorerecentwritings.Derrida’sviewofethicsisbestdevelopedthroughhistreatmentofthecentralethicalconceptofjustice.Herehebeginsfromanessentialtension:justiceisinonesensenothingotherthantheruleoflaw;butinfactthemere,literalapplicationofthelawisgenerallyunjust.Thepointisnotmerelythetraditionalonethatthereisadistinctionbetweenanypositivelaw(legislatedbysomeparticularhumansociety)andtheabsolutestandardsdefinedby,say,divinecommand,theFormofJustice,orhumannature.Derrida’sclaimismuchstronger,thatanysystemoflaws–human,natural,divine–willneverbeableinitselftospecifyadequatelytheconductthatisjustinagivensituation.Accordingly,anysystemoflawsissubjecttodeconstruction.Nomatterhowstable,consistentandcoherentitmaybeinitsownterms,therewillbepointsofapplicationwhereitbecomesincoherent.Atsuchcriticalpointsofapplication,weseethetensionbetweenlawandjustice.Derridadoesnot,however,meantosuggestthatwehaveaccesstojusticethroughanythingotherthanlaw.Thereis,asalways,noquestionofsomespecialaccessthroughaprivilegedexperienceorinsighttowhatliesbeyondthelaw.Justaswecannevergetbeyondourconcepts,wecannotgetbeyondourlaws,whicharejustconceptualizationsofethicalobligations.Rather,wemovetowardjusticesimplybyremainingeversensitivetopossiblelimitationsoflaws,bybeingalwaysreadytodeconstructlawsthatareworkingagainstjustice.ThisisonereasonthatDerridacansuggestthat‘deconstructionisjustice’.Thisdeconstructionisnotascepticalrejectionofthelawbutaclearingofthegroundforanewjudgementofhowweshouldbehave.Thisnewjudgement(decision)doesnotitselfdrawauthorityfromanysystemoflaw–howcouldit,whenitpresupposesadeconstructionofprevioussystems?Butneitherdoesitdrawauthorityfromsomeprivilegedinsightbeyondthelaw(sayaPlatonicintuitionofJustice).Thejudgementisaleap,atakingofastandwhenthereisnoadequatejustificationfortakingastand.Inthisregard,DerridacitesKierkegaard:‘Theinstantofdecisionisamadness’–themovementoftheindividualassuchbeyondtheuniversal.Butthisleapdoesnottakeusbeyondtherealmoflawassuch.Itmovesbeyondpreviousformula-tionsoflawbutmustinturnjustifyitselfbyconstructinganew,moreadequatesystemoflawthatwill,ofcourse,itselfbesubjecttodeconstruction.Incontrasttodangerousirrationalisms(forexample,fascism,religiousfanaticism)thatleavereasonbehindinafloodofmerewilloremotion,Derrida’sdeconstructiveapproachalwayssubjectsour‘leaps’beyondonesystemofrationalthoughttotheconstraintofconstructinganewsystemofrationalthought.Inthisway,hebalancesreasonandasenseofitslimitationsagainstoneanotherinaconstantplayoftensions.FurtherReadingTheprimarytextforSartre’sexistentialismis,ofcourse,BeingandNothingness.Hisfamouspopularlecture‘ExistentialismIsaHumanism’remains,despitehisowndisavowal,thebest871\nGARYGUTTINGsingleintroductiontothedifficultideasofBeingandNothingness.HisearlycritiqueofHusserl,TheTranscendenceoftheEgo,givesagoodviewofSartre’sdistinctivenotionofconsciousness.Sartre’slaterphilosophicalthoughtisprimarilydevelopedinTheCritiqueofDialecticalReason,alargeanddifficultworkdevotedtosynthesizingexistentialismandMarxism.AfascinatingportraitofbothSartre’sthoughtandpersonalityemergesintheextensiveinterviewswithSimonedeBeauvoirprintedinherAdieux(whichalsoincludesheraccountofthelastdaysofSartre’slife).Foucault’sHistoryofMadnessisonlypartiallyavailableinEnglish:MadnessandCivilizationtranslatesalittleoverhalftheoriginaltext.TheOrderofThings(theFrenchtitleisLesMotsetleschoses),DisciplineandPunish(SurveilleretPunir)andthethreevolumesofTheHistoryofSexuality(I:Introduction,II:TheUseofPleasure,III:TheCareoftheSelf)areallfullytranslated.AsuperbselectionofFoucault’sessays,lecturesandinterviews(basedonthecomprehensiveFrenchcollectionDitsetrits)hasbeenpublishedinthreevolumesasEssentialWorksofFoucault,1954–1984,editedbyPaulRabinow.PerhapsthemostaccessibleintroductiontoDerrida’sthoughtisthroughthethreeinterviewspublishedunderthetitlePositions.TheessayscollectedinWritingandDifferenceandMarginsofPhilosophygiveagoodsenseofhisinterestsandmethods.Forhismorerecentworkonethicsandreligion,seeTheGiftofDeath.Astosecondaryliterature,thebestbriefdiscussionofthewholeofSartre’sworkisCaws(1979),whileDanto(1975)providesaveryaccessibleguidetoSartre’sexistentialism.ThebestbiographyisCohan-Solal(1987).OnFoucault,agoodgeneralsurveyisMcNay(1994),andthebestbiographyisMacey(1993).FormoredetailedanalysisofFoucault’searlierworks(throughTheOrderofThingsandTheArchaeologyofKnowledge),seeGutting(1989).TwogoodandaccessibleintroductionstoDerridaareNorris(1987)andHowells(1998).OnDerrida’slaterwork,CaputoandDerrida(1997)isinvaluable–boththeopeningdiscussionwithDerridaandCaputo’sfollowinganalysis.ReferencesSartreSartre,J.-P.1946:ExistentialismIsaHumanism.Paris:Nagel.——1957a[1943]:BeingandNothingness:AnEssayonPhenomenologicalOntology(translatedbyH.Barnes).London:Methuen.——1957b:TheTranscendenceoftheEgo:AnExistentialistTheoryofConsciousness.NewYork:NoondayPress.——1976:TheCritiqueofDialecticalReasonI.London:NewLeftBooks.WritersonSartredeBeauvoir,S.1984:Adieux:AFarewelltoSartre.London:AndréDeutsch;WeidenfeldandNicolson.Caws,P.1979:Sartre.London:RoutledgeandKeganPaul.Cohan-Solal,A.1987:Sartre:ALife.NewYork:PantheonBooks.Danto,A.1975:Jean-PaulSartre.NewYork:VikingPress.FoucaultFoucault,M.1967[1961]:MadnessandCivilization:AHistoryofInsanityintheAgeofReason(translatedbyR.Howard).London:Tavistock.——1970[1966]:TheOrderofThings:AnArchaeologyoftheHumanSciences.London:Tavistock.——1977:DisciplineandPunish:TheBirthofthePrison.London:AllenLane.872\nSARTRE,FOUCAULTANDDERRIDA——1981–:TheHistoryofSexuality.Harmondswoth:PenguinBooks.——1988–:EssentialWorksofFoucault,1954–1984,3vols(editedbyPaulRabinow).London:AllenLane.WritersonFoucaultGutting,G.1989:MichelFoucault’sArchaeologyofScientificReason.Cambridge:CambridgeUniversityPress.Macey,D.1993:TheLivesofMichelFoucault.London:Hutchinson.McNay,L.1994:Foucault:ACriticalIntroduction.Cambridge:PolityPress.DerridaDerrida,J.1978[1967]:WritingandDifference(translatedbyA.Bass).London:RoutledgeandKeganPaul.——1981[1972]:Positions(translatedbyA.Bass).Chicago:UniversityofChicagoPress.——1982[1972]:MarginsofPhilosophy(translatedbyA.Bass).Chicago:UniversityofChicagoPress.——1995:TheGiftofDeath.Chicago:UniversityofChicagoPress.WritersonDerridaCaputo,J.andDerrida,J.1997:DeconstructioninaNutshell:AConversationwithJacquesDerrida.NewYork:FordhamUniversityPress.Howells,C.1998:Derrida:DeconstructionfromPhenomenologytoEthics.Cambridge:PolityPress.Norris,C.1987:Derrida.London:Fontana.DiscussionQuestions1Isconsciousnessnothing?2Istobeconsciousalwaystobeself-conscious?3Ifconsciousnessisnotanobject,howcantheselfbepartoftheworldlikeanyotherthing?4CanweacceptSartre’saccountofhowbeing-in-itselfandbeing-for-itselfarerelated?5Doweneedanontologicalcategoryofnothingnessaswellasalogicalconceptofnegation?6Isfreedomamatterofdeterminingwhathappensorofdeterminingthemeaningofwhathappens?7Howisself-deceptionorbadfaithpossible?8Canfreedomitself,ratherthananyexternalend,bethegoaloffreedom?9Arewhatappeartobenecessarytruthsaboutourconditiononlycontingentproductsofourhistoricalsituation?10Doestheexistenceofdifferenthistoricallylocatedepistemicstructuresshowthatthehumansciencesarecontingentandreplaceable?11WhataretheimplicationsforphilosophyofFoucault’sarcheologicalmethod?12Doesagenealogicalaccountoftherelationsbetweenknowledgeandpowerlimittheclaimsofscientificknowledgeinmodernsociety?13IsFoucault’sgenealogyanimprovementonhisarcheologyasamethodinthehumansciences?873\nGARYGUTTING14Howshouldweassesstheclaimthatthespreadofmoderndisciplinarypracticesresultedinthetransformationofsocietyfromoneofspectacletooneofsurveillance?15Doesaliberatingfocusonoursexualitymerelyplacemodernhumanbeingsinthecontrolofanothersystemofconstraints?16Shouldwerejecttheideaofaperfectlycleartext,thought,expressionandreality?WhataretheimplicationsofthisrejectionforDerrida’sconceptionofphilosophy?17Doesthemethodofdeconstructionyieldinsightsaboutphilosophy?18WhataretheconsequencesofDerrida’saccountofdifferancefortheemploymentofbasicdichotomiesinphilosophy?19Whataretheconsequencesforethicsifapplyinganysystemoflawwillproduceinjustice?20Canwelivewithparadox?874\nGlossaryapriori–aposterioriAnepistemologicaldistinction.Aprioripropositions,unlikeaposterioripropositions,donotrequireexperiencetoestablishtheirtruth.Wederiveorjustifyaprioriconcepts,unlikeaposterioriconcepts,independentofexperience.abstractideasLocke’sattempt,rejectedbyBerkeley,toexplainhowanideacanstandforindividualsofagivenkind,eventhoughtheindividualsvaryintheirproperties.Lockeheldthatabstractionfromthedifferentpropertieswouldproduceageneralideacoveringtherightindividuals.abstractobjectsObjects,suchasnumbersoruniversals,thatdonotexistasspatio-temporalparticulars.Philosophersdisagreeaboutwhethertherecanbesuchobjectsor,iftheydoexist,howtheyarerelatedtoconcretephysicalobjects.adhominemargumentAfallaciousargumentattackingtheholderofaviewratherthanthepositionitself,orasoundargumentshowinganinconsistencybetweenaviewheldbyapersonandaconsequenceofthatview.Thepersonpointingouttheinconsistencyneednotholdtheinitialview.akrasiaTheGreektermforweaknessofwill.Inakrasiaonedoesnotdowhatoneknowstobebestordoeswhatoneknowsnottobebest.SocratesandAristotleinitiatedattemptstodeterminewhetherakrasiaispossible.alienation/estrangementHegelianconcept,alsousedbyMarxandlaterEuropeanphilosophers,tostandforastateofbeingcutofffromsomethingofimportance,suchasoneself,others,natureortheproductofone’slabour.Theanalysisandinterpretationofalienationvariesaccordingtothephilosopher.altruismTheviewthatthewell-beingofothersshouldhaveasmuchimportanceforusasthewell-beingofourselves.Somearguethataltruism,evenifitisdesirable,isnotpossible,andthatourethicsmustbebasedonegoism.analysisThecentralmethodofanalyticalphilosophy,shapedbythedevelopmentofmodernlogicandfoundintheworkofFrege,Russell,MooreandWittgenstein,accordingtowhichphilosophicalproblemscanbeovercomethroughreplacingtheapparentstructureofstatementsbytheirreallogicalstructure.Manyphilosophers,whilestillconsideringthemselvesanalyticalphilosophers,havealteredorevenabandonedthisprogramme.analytic–syntheticAccordingtoKant’sformulationofthedistinction,inananalyticpropo-sitiontheconceptofthepredicateiscontainedintheconceptofthesubject,andwecantellthatthepropositionistruebyanalysis.Inasyntheticproposition,theconceptofthepredicateaddssomethingnewtotheconceptofthesubject,andthetruthorfalsityofthepropositioncannotbedeterminedbyanalysis.Therehasbeenmuchdisputeovertheadequacyofthisaccount,butthereisgeneralagreementthatsyntheticpropositionstellussomethingabouttheworld.\nGLOSSARYTogetherwiththemetaphysicaldistinctionbetweennecessaryandcontingentpropositionsandtheepistemologicaldistinctionbetweenaprioriandaposterioripropositions,thislogicaldistinctionsetstheframeworkformuchmodernphilosophy.Kantfamouslyarguedthatsomeapriorinecessarypropositionsaresynthetic.ContemporarydiscussionhasdevelopedfromQuine’scriticismofthedistinctionasadogmaofempiricism.anthropocentrismLiterally,‘human-centred’.Henceinethics,forexample,theviewthathumanactionsandhumanwelfarearethesoleproperfocusofmoralconcern.aporiaApuzzleorperplexity.IntheearlyPlatonicdialoguesSocratesraisedproblemswithoutofferingsolutionstothemandshowedthatthosehequestionedcouldnotofferanacceptablesolutioneither.Thisaporeticmethodledtothedevelopmentofthedialecticalmethod,bywhichSocrateselicitedtruththroughquestioning.Thetermaporia(‘nowaythrough’)wasintroducedbyAristotleforpuzzlesconcerningincompatibilitiesthatariseamongviewsweholdwithoutpromptingoramongreputablebeliefsadoptedcommonlyorbythewise.Hisapproachwastoseektheminimaladjustmentsneededtoreconciletheseconflictingviews.archeologyFoucault’smethodfordeterminingthedeepstructureorformoftheconditionsofthepossibilityofknowledgeinaparticularhistoricalperiod.argumentfromdesignTheargumentfortheexistenceofGod,disputedbyHume,accordingtowhichthecomplexandintricateorderoftheworldcanonlybeexplained(orcanbestbeexplained)bypositinganintelligentandpowerfulcreator.artificialintelligence(AI)Theuseofprogramstoenablemachinestoperformtaskswhichhumansperformusingtheirintelligence.EarlyAIavoidedhumanpsychologicalmodels,butthisorientationhasbeenalteredbythedevelopmentofconnectionism,whichisbasedontheoriesofhowthebrainworks.Inconnectionism,complexfunctions,includinglearning,involvethetransmissionofinformationalongpathwaysformedamonglargearraysofsimpleelements.AIraisesquestionsabouttheconditions,ifany,inwhichwewouldbejustifiedinascribingmentalattributestopurelyphysicalsystems.A-seriesandB-seriesoftimeMcTaggart’stermsforthetemporalorderingofeventsaccordingtowhethertheyarepast,presentorfuture(A-series)orearlierorlaterthanoneanother(B-series).associationofideasAview,especiallyimportantinHume,explainingthepatternedoccurrenceofourideasaccordingtolawsofassociation.PhilosopherstodaygenerallyseektomaintainwhatisimportantinHumewhilerejectingthismechanism.autonomyAnautonomousbeingisonethathasthepowerofself-direction,possessingtheabilitytoactasitdecides,independentofthewillofothersandofotherinternalorexternalfactors.axiomsPropositionsselectedasthefoundationsofafield–classicallygeometry–which,togetherwithmethodsofproof,allowotherpropositionstobeprovedinanorderedway.Theaxiomaticmethodhaspowerfullyinfluencedphilosophy,althougheachfeatureofthemethodhasbeencriticizedasinappropriateforphilosophy.beingAccordingtoPlato,onlyFormshavebeing.ForAristotle,whodistinguisheddifferentsensesofbeing,beingquabeingisthecentralconcernofmetaphysics.Incontemporaryphenomenology,HeideggerconsideredthemeaningofBeingandcharacterizedhumanBeingasDasein(‘beingthere’)andbeing-in-the-world.Sartredistinguishedthebeing-for-itselfofconsciousnessfromthebeing-in-itselfofobjects.CategoricalImperativeThefundamentalformaldemand(orsetofdemands)whichKantplacedonourchoiceofprinciplesonwhichtoact.Itiscontrastedwithhypotheticalimperatives,whichhaveforceonlyifwehavecertaindesiresorinclinations.FormulationsoftheCategoricalImperativeseemtoberadicallydifferentfromoneanother,andsomecriticsarguethattheCategoricalImperativeproducesanemptyformalism.Sympatheticcommentatorsbelievethatbothoftheseproblemscanbeovercome.Theformulationstesttheprinciplesonwhichweact876\nGLOSSARYaccordingtowhethertheycanbeuniversallawsorlawsofnature,whetherwetreathumanityinourselvesandothersneversimplyasmeansbutalsoasends,whetherwetreateveryrationalbeingasawill-makinguniversallaw,andwhetherwetreatoursharedmorallifeastakingplacewithinakingdomofends.NoneoftheprincipalnotionsusedinexpressingtheCategoricalImperativeiseasytounderstand.categoriesThebasicgeneralconceptsofthought,languageorreality,sometimesclaimedtohaveanoriginorjustificationdifferingfromthoseofordinaryconcepts.AristotleandKantprovidetheclassicaldiscussionsofcategories,althoughcategoriesplaydifferentrolesintheirthought.categorymistakeAnascriptionofsomethingtoonecategorywhenitbelongstoanother.ForRyle,whointroducedtheterm,discriminationsamongcategorieswerenotconfinedtothegrandAristotelianorKantiancategories,butworkedtheirwaywithgreatsubtletythroughthewholeofourlanguage.Theinterestingmistakesarethosewithphilosophicalconsequences.causaltheoryofreferenceTheviewofKripkeandothersthatnames,andperhapssomeotherterms,gainmeaningfromaninitialactofnamingandthenpreservemeaningthroughsuitablecausallinks.causationIncausalrelationsbetweenevents,ifaneventofthefirstkindoccurs,aneventofthesecondkindwillormustoccur,andthefirsteventwillexplaintheoccurrenceofthesecondevent.Possiblyitemsotherthaneventscanenterintocausalrelations.SinceHume,wehavebeenpuzzledaboutwhethercausalrelationsarerealorarejustmattersofourimposingourhabitsupontheworldandoverthenatureofcausalnecessity.certaintyDescartessoughttobuildknowledgeonthebasisofcertainty,withnoroomfordoubt.Althoughtheprojectasawhole,aswellasitsdetail,hasbeencontested,certaintyremainsanidealformanyphilosophers.chaostheoryThetheoryofnon-linearfunctions,suchthatsmalldifferencesintheinputofthefunctioncanresultinlargeandunpredictabledifferencesintheoutput.classAcollectionofentitiessatisfyingaconditionformembershipintheclass.Toavoidproblemsarisingifclassesgettoolarge,orbelongtootherclasses,orarenotcompleted,settheorydistinguishesclassesfromsets.clearnessanddistinctnessDescartes’scriteriaofindubitabletruthderivedfromhisreflec-tionontheimpossibilityofdoubtinghisownexistence(seecogitoergosum).Clearperceptionsare‘presentandaccessible’.Distinctperceptionsare‘sharplyseparated’fromotherperceptionsandcontainonlywhatisclear.Wecanhopetospecifyclearnessanddistinctnessinanillumi-natingway,butthismightinvolvereplacingperceptualcharacterizationsbyconceptualones.cogitoergosumDescartes’scrucialclaim‘IthinkthereforeIam’providesastandardofcer-taintyfortherestofhisphilosophyandleadsontotheclaimthatwhatheisisathinkingthing.concept‘Concept’canbetakenpsychologicallyorlogicallyforwhatwegraspinunderstand-inganexpression,butsinceFregethelogicalsidehashadprimaryimportance.ForFrege,therewasacrucialdistinctionbetweenobjects(referredtobynamesorsubjects)andconcepts(referredtobypredicates).Differentaccountsoflogicalformmightchallengethisclaim.Ifconceptsarethoughtofascomponentsofpropositions,scepticismaboutpropositionscanproducescepticismaboutconceptsaswell.consciousnessAphilosophicalexplanationofwhatconsciousnessisorhowitmightbeexplainedeludesus.Ifwesticktowhatitisliketobeaconscioushumanbeing,wehavenoexplanation;ifwetrytoexplainconsciousnessintermsofwhatgoesoninourbrains,thesheerfeelofconsciousnessitselfisleftaside.consentFromLocke,theliberaltheoryofgovernmenthasrequiredtheconsentofthegovernedforpoliticalruletobelegitimate.Becauseexplicitconsentisnotalwaysavailable,Lockeintroducedanotionoftacitconsent,butitisnotalwayseasytodistinguishtacitconsentfromnon-consenttodeterminewhethertacitconsentisconsent.877\nGLOSSARYconsequentialismTheviewthatthevalueofanactionisdeterminedbythevalueofitscon-sequencesratherthanbytheprincipleonwhichtheactionisperformedorthevirtueitexpresses.Utilitarianismisaconsequentialisttheory,wheretherelevantvalueisindividualhappinessorwell-being.conservativeInpolitics,alooselydefinedtermindicatingadherencetooneormoreofafamilyofattitudes,includingrespectfortraditionandauthorityandresistancetowholesaleorsuddenchanges.consistencyPropositionsareconsistentiftheycanallbetrue.Asystemofpropositionscanbeshowntobeinconsistentifitcontainsacontradiction(apropositionanditsnegation).Consistencyandcompletenessaretwokeyconcernsofmodernlogic.constructivismInthephilosophyofmathematics,abroadposition(encompassingbothintuitionismandformalismbutalsogoingbeyondthem)whichholdsthatmathematicalentitiesexistonlyiftheycanbeconstructedandthatproofandtruthinmathematicsareco-extensive.Constructivistsopposetherealist(orPlatonist)viewthatmathematicalobjectsortruthexistindependentlyofhumanprocedures.ThishastheconsequencethatcertainclassicalresultswhoseproofsrelyonPlatonicassumptionsarenotconstructivelyvalid.contentsofconsciousnessMentalstates,likestatementsandotherlinguisticitems,havecontents,butitisaphilosophicalproblemhowthiscanbeso.Furthermore,therearedisputesovertheextenttowhichinternalfactorsandexternalenvironmentalfactorsrespectivelycontributetodeterminingthecontentsofmentalstates.contingent–necessaryContingentpropositionshappentobetrueorfalsebutcouldbeotherwise.Necessarypropositionsmustbetrue.Itisnotclearthatthereareanynecessarypropositionsor,ifthereare,thattheyarerestrictedtoanalyticpropositionsorotherpropositionstruebecauseoftheirlogicalform.Acontingenteventisonethatdoesnotnecessarilytakeplace.Iftherearenecessaryevents,naturalratherthanlogicalnecessityisinvolved.continuumAcollectionofpoints,suchthatbetweenanytwopointstherearedistinctpoints.Classicalexamplesofacontinuumarealine,planeorspace.continuumhypothesisTheclaimthatthereareonlytwoclassesintowhichanythinkablecollectionofinfinitelymanydistinctrealnumbersmayfall.contradictionAconjunctionofapropositionanditsnegation,which,accordingtotheprincipleofnon-contradiction,cannotbetrue.Aristotlepointedoutthedangersofacceptingcontradictions.Exceptinsomespeciallydesignedlogics,anythingcanfollowfromacontradiction.conventionalismTheviewthathumanconventionsratherthanindependentrealitiesornecessitiesshapeourbasicconceptsoftheworld,scientifictheories,ethicalprinciplesandthelike.Onthisview,wecouldhavechosenotherconventions,whichwouldhavebeenassatisfactoryastheconventionsentrenchedinouractualaccountoftheworldormorality.Someconventionalistpositionsallowforacontributionbyrealityaswellasbyconventions,butitisdifficulttodistinguishthesecontributions.counterfactualconditionalAconditional(ifp,thenq)inwhichthe‘ifclause’iscontrarytofact;forexample,‘ifthewaterhadbeenboiling,youwouldhavebeenscalded’.Thereisnogenerallysatisfactoryanalysisofcounterfactualconditionals,althoughsomephilosophersbelievethatweneedthemtodealwithmanyimportantphilosophicalproblems.criterionAtestorstandardbywhichtruth,existence,identityormeaningcanbedetermined.Questionsariseoverthechoiceofcriteriaandovertherelationbetweencriteriaandthatforwhichtheyarecriteria.critiqueKantintroducedthetermforthecriticalexaminationofreasonbyitself.LaterEuropeanphilosophershavepursuedamethodofcritique,butsomehaverelinquishedKant’scommitmenttoreasonasthekeyelementoftheirreflectivemethod.878\nGLOSSARYDarwinismOriginallyreferringtotheviewsofthenaturalistCharlesDarwinconcerninghistheoryofspecieschangeanddevelopmentthroughthefilterofnaturalselection.Nowcommonlymoreloosely(andperhapsmisleadingly)usedasasynonymforevolutionarytheoryingeneral.dasManForHeidegger,abasicstructureofanonymoussocialnormativitythatisassociatedwithourtendencytolapseintoinauthenticity.DaseinInHeidegger’spursuitofwhatheconsideredtobethecentralquestioninphilosophy–thatis,thequestionofthemeaningofBeing–hespokeofourBeingashumanbeingsasDasein(‘beingthere’).Intalkingofourselvesinthisway,hemeantDaseintobreakwiththewholehistoryofontology,includingtheinheritedCartesianconceptionoftheself,andtoreplaceitwithaneworientation.deconstructionDerrida’smethodaimingtoovercomecrucialmetaphysicaldichotomies.Byshowingthatonetermofanoppositionisunjustifiablyprivilegedwithrespecttotheother,decon-structionallowsustousethepairoftermsfreedfrommetaphysicaldistortion.Hisapproachtodeconstructionemployssophisticatedandsurprisingresponsestolanguage,cultureandsocietyderivedfromFreudandothers.defeasibleThestandardcriteriaforthecorrectapplicationofadefeasibleconceptallowforthatapplicationtoberetractedinthelightoffurtherevidence.Verificationofclaimsusingdefeasibleconceptsisneverconclusive,inprinciplebeingalwaysopentorevision.Forexample,inepistemology,adefeasibleknowledgeclaimisonemadeconfidently,butinrecognitionofthepossibility(nomatterhowapparentlyunlikely)thatfurtherevidencecouldgivereasonfortheclaimtobewithdrawn.definitedescriptionAdescriptionpickingoutsomethingasthesoleindividualhavingacertainproperty.Russell’stheoryofdefinitedescriptionsanalysessentencescontainingdefinitedescriptionstoremovetheburdenoffindingobjectstowhichtheseexpressionsseemtorefer.deontologyAnethicsbasedonactingaccordingtodutyordoingwhatisright,ratherthanonachievingvirtueoronbringingaboutgoodconsequences.Itistoocrudetomakesharpdivisionsortodenyaplaceformorethanoneapproachtoethics.Kantisthemostimportantdeontologicaltheorist.dialecticTheSocraticmethodofdiscoveringtruththroughquestioninganddebate,alteredanddevelopedbyhisGreeksuccessors,andstillamodelofoverwhelmingimportanceinphilosophy.Kantiandialecticexpressedreason’scapacitytoreachcontradictoryconclusionsfromapparentlysoundpremises.Hegel’sdialecticdrovethenecessaryunfoldinganddevelopmentofconceptsinhistory.Marx’sdialecticexplainedthehistoricaldevelopmentofsocietythroughclassconflictandtherelationsbetweentheforcesandrelationsofproductionandthebaseandsuperstructure.didacticOfferedwithanintentionofinstructingorteaching.différanceDerrida’stermmeaningboth‘todiffer’(relatedtospatialityandthebasisofallconceptualoppositions)and‘todefer’(relatedtotemporalityandtheperceptionofchangeintherelationshipdeterminingmeaning).dualismTheviewthateachpersonistwoentities,amindwithmentalattributesandabodywithphysicalattributes,insteadofasingleentitywithattributesofbothsorts.egoismTheviewthatwearealwaysmotivatedbyself-interestorthatwealwaysshouldbesomotivated.Contemporaryrationalchoicetheoristsattempttounderstandhowactualsocialinstitutionscanbebasedonthechoicesofindividualsactingaccordingtoegoistprinciples.Theprisoner’sdilemmaandotherproblemcasesshowdifficultieswiththisapproach.elenchusTheSocraticmethodofseekingtruthbycross-examiningpeopletoshowconflictsintheirbeliefs.eliminativematerialismTheviewthatourmentalconcepts,suchasbeliefanddesire,areinappropriateforaseriousscientificaccountofhumanbeingsandshould,orwillbe,eliminated.879\nGLOSSARYempiricismTheclaimthatallknowledgeorallmeaningfuldiscourseabouttheworldisrelatedtosensoryexperienceorobservation.Logicalempiricism(orlogicalpositivism)combinedmodernlogicalanalysiswiththedemandsofempiricismandwasmostfamousforitsverificationisttheoryofmeaning.EnlightenmentAbroadintellectualmovementineighteenth-centuryEurope,particularlyBritain,FranceandGermany,characterizedbyarejectionofsuperstitionandmysteryandanoptimismconcerningthepowerofhumanreasoningandscientificendeavour(henceitsalternativename:theAgeofReason).entailmentTheintuitivenotionofstrictlogicalimplication,suchthatnecessarilyifp,thenq(thatis,itisimpossiblethatpandnotq).epoche¯AtermusedbythescepticsforthesuspensionofbelieforjudgementandbyHusserlfor‘bracketingout’allobjectsthattranscendconsciousness.essenceForAristotle,thatwhichremainsthenatureofathingthroughoutitschangefrompotentialitytoactuality.Moregenerally,thenecessarydefiningcharacteristicofathing,suchthatwithoutthatcharacteristicthethingwouldnotbethethingitis.eternalrecurrenceNietzsche’stermfortheinfinitecyclicalprocessofthewilltopowerwhichprovidesmeaningwithoutteleologybymovingfromthesimpletothecomplexandthecomplextothesimple.ethicalintuitionTheimmediateawarenessofanethicalpropertyoranethicaltruth,butforRawlsintuitionismistheclaimthatthereisanirreduciblebodyoffirstprincipleswhichcannotbeorderedintermsofpriority.ethicalnaturalismTheview,criticizedbyG.E.Moore,thatethicalpropertieslikegoodnesscanbedefinedintermsofthenaturalpropertiesthatjustifytheiruse.Moore’salternativeaccountintermsofnon-naturalintuitionofgoodnesshasseemedimplausible,butnewwaysofconceivingrelationsamongpropertieshavealsobeenimportantintherevivalofethicalnaturalism.eudaimonia‘Flourishing’or‘well-being’,thecentralaimofAristotelianethics,therealizationofwhichisacompleteandself-sufficientcombinationofvirtueanditsrewardsinhappiness.Itisimportanttorealizethecomplexityofthenotionandnottoreduceittooneofitscomponents.eventWecanthinkofeventsaschangesinobjectsorinrelationsamongobjects,sothatthebasicentitiesinouraccountoftheworldareobjectsandrelationsratherthanevents.However,somephilosophersseeadvantagesinrecognizingeventsasbasic,althougheventsmighthavetotakeonsomeofthecharacterofobjectsforthistobeacceptable.evolution(theoryof)Darwin’stheoryofevolutionaccountsforthechangesinspeciesoflivingthings,accordingtowhichallorganismsarederivedfromcommonancestorsthroughnaturalselection.Formodernbiologists,evolutionisidentifiedwithchangesinthegeneticcompositionofpopulations.evolutionaryepistemologyAnapproachtothedevelopmentofhumanknowledgeinevolutionaryterms,eitherasanintegralpartofnaturalselectionorasanindependentprocessmodelledonbiologicalnaturalselection.Evolutionaryepistemologyispartofabroaderpro-grammeofnaturalizedepistemology.Ratherthanseekingtosecureourknowledgeclaimsagainstscepticaldoubts,naturalizedepistemologytriestoexplainmajorfeaturesofourknowledgeasnecessaryorinevitablefeaturesofourselvesasnaturalbeings.expertsystemAcomputerwith‘built-in’expertise,which,usedbyanon-expertinaparticularsubjectarea,canevaluateormakeotherdecisionsconcerningthatsubject.explanationAnaccountcharacteristicallytellinguswhysomethingexistsorhappens,ormustexistorhappen.ThecoveringlawmodelofexplanationproposedbyHempelhasbeenwidelyinfluential,buthasmanycritics.Therearecontroversiesoverthenatureoffunctionalorteleologicalexplanation,overthelegitimacyofinferringtothebestexplanation,andoverDilthey’scontrastbetweenscientificexplanationandhistoricalunderstanding.880\nGLOSSARYfallibilismPeirce’sviewthatnoneofourbeliefs,eventheapparentlymostfundamental,iscertainandthatanyofourbeliefscanberevised.Peircebelievedthat,rightlyunderstood,relinquishingcertaintydoesnotopenthewaytoscepticaldoubt.familyresemblanceWittgenstein’sterminhislaterphilosophyforthewayinwhichexpressionsapplytothingsorkindsofthingsnotsharingacommondefiningproperty,butinsteadsharingsomeofaninterwovencomplexoflikenesses,asinthefacialfeaturesoffamilymembers.Hemadethesamepointregardingtheoverlappingfeaturesofgames.fictionalentitiesNotallexpressionspickoutthingswhichexist.Non-existingorfictionalentitiescanplayhavocwithouraccountoflanguageandreality.Iftheydonotexist,wearetemptedtoenrichourontologywithanexistence-likestatustoallowreferencetothem.Analysis,inoneformoranother,mightavoidsuchtemptationsbyshowingthatwecanmeanwhatweneedtomeanwithoutcommittingourselvestooddontologicalitems.Problemsremainforrealasopposedtonotionalfictions.WecansaywhatistrueorfalseaboutAnnaKareninaorSherlockHolmes,yetwecannotbecertainhowtounderstandourabilitytodoso.firstphilosophyInAristotle,thestudyofthegeneralcharacteristicsofalltypesofexistenceortheprinciplesofbeing.Moregenerally,especiallysinceDescartes,thepositionthatthereisanessentialroleforphilosophy,priortoanyscience.Suchaclaimischallengedbyscientisticviews,whichreversethepriority.formoflifeWhatWittgensteintakestobefundamentalinhislateraccountofunderstandinglanguageasvariouslyembeddedinsharedhumanactivity.Thecrucialnotionof‘form’shiftsfromlogicalforminhisearlywritingtoformoflife.Whatisfundamentalinunderstandingthepossibilityoflanguageshiftsfromobjectstolanguagegamesandformsoflife.formalIngeneral,formalconsiderationshavetodowiththeabstractstructure,orpattern,ofasubject,ratherthanwithitscontent.Henceformallogic,forexample,isconcernednotwiththecontentofparticularsentencesinanargument,butonlywiththeirstructureoftruthvalues.FormsAccordingtoPlato’stheoryofForms,objectivelyexistingimmaterialentitiesthataretheproperobjectofknowledge.OrdinarythingsexistbyparticipatinginForms.Formsarealsoheldtobenecessaryfortheobjectivityofethicsandaestheticsandthemeaningoflanguage.foundationalismTheviewthatknowledgeispossibleonlyifsomeitemsserveasacertainfoundationfortherest.Specialattentionisdrawntotheallegedcertaintyoftheproposedfoundationsandtotherelationbetweenthefoundationsandtherestofknowledge.freewillanddeterminismTherearethreatstofreedominvolvingtheapparentdetermina-tionofhumanactionindependentofourwill.Theseincludedivineforeknowledgeandinmodernphilosophythepossibilitythatouractionsaredeterminedbycausallaws.Onewayoutwouldbetoarguethatfreedomandaparticularformofdeterminismarecompatiblebyshowingthatdeterminismhasweakerimplicationsthanatfirstseemsthecase.Onecouldalsoarguethatwhatonevaluesinfreedomisnotriskedbydeterminismandisnotaidedbyindeterminism.Nevertheless,somephilosophersstillclaimthat,onaproperunderstandingoffreedomandforeknowledgeorfreedomandcausallaws,wecannothaveboth.functionAfunctionisarelationbetweenthevalueofvariablesandthevalueofthefunctionasawhole.Givingadefinitevaluetothevariablesyieldsadefinitevaluetothewholefunction.functionalexplanationAlsoteleologicalexplanation,inwhichanitemisexplainedbytheroleorfunctionithasinproducingsomethingwhichpromotesorpreservestheentityorkindofentityofwhichitisapart.Functionalexplanationisprominentinbiologyandinsomeofthesocialsciences.Thereisdisagreementwhetherfunctionalexplanationisarivaltocausalexplanationoraformofcausalexplanation.genealogyNietzsche’sattempttoexplainourbasicbeliefsandevaluationsandtobreakourattachmenttothem.ForFoucault,theexplorationofthemultipleinterrelationsbetweenknowledgeandpower.881\nGLOSSARYgoodManyapproachestoethicsarecentredonachievingwhatisgood,althoughothersarebasedondoingwhatisright.Prioritytoonegoalneednotexcludetheother,butmightshapethecontentsorlimithowwepursuetheother.Wecanaskabouttherelationshipbetweenhumangoodnessandthegoodnessofotherthings,suchasaknifethatcutswell.Onsomeviews,goodnessisreducedtoonequality,likehappinessorpleasureorsatisfyingdesire,butothersthinkofgoodnessasinherentlycomplex.Ineithercase,goodnesshasaplaceinmoralpsychology,motivatingouractionsandexplainingouremotions.happinessAccordingtodifferentethicalviews,happinessmightbeonegoalinlife,theonlypossiblegoal,orafortunateby-productofthepursuitofothergoals.Happinessmightconcernone’saggregateofpleasureorrequireacomplexbalanceinvolvingvirtue,pleasure,achievementandgoodfortune.hedonismThebeliefthatpleasureisthegreatestgoodandhighestaspirationofhumankind.Inearlyutilitarianthinking,thisbeliefprovidedtheinterpretationof‘utility’or‘good’.hermeneuticsAmethodofinterpretation,initiallyofbiblicaltextsbutlaterextendedtoothertextsandwithDiltheytowholecultures.Characteristically,themethodimprovesourunder-standingofwhatisobscure,corruptorincompletebyplacingitinthecontextofawhole.Thehermeneuticcircleisaproblemwhichclassicallyarisesbecausetheunderstandingofaniteminatextdependsonourunderstandingofthewholetext,whiletheunderstandingofthewholetextdependsonourunderstandingofthatandotheritems.Thisinterdependenceofinterpreta-tion,however,neednotbevicious,althoughitmightcallformodestyratherthandogmatisminadvancinganyinterpretation.historicismTheviewthatconcepts,beliefs,truthsandevenstandardsoftruthcanbeunderstoodonlyinrelationtothewholemoral,intellectual,religiousandaestheticculturesofthehistoricalperiodsinwhichtheyariseorflourish.Thispositionislinkedtodemandsforahermeneuticmethodtoachieveappropriateunderstanding.ThetermisalsousedbyPopperfortheviewthathistoryisgovernedbynecessarylawsofdevelopment.holismTheviewthatwholeshavesomepriorityovertheelements,members,individualsorpartscomposingthem.Socialholismclaimsthatindividualscanbeunderstoodonlyintermsofthepracticesorinstitutionsinwhichtheytakepartandisarivaltosomeaspectsofindividual-ism.Methodologicalholismandmethodologicalindividualismproposedifferentmethodologicalconstraintsonthestudyofphenomenawithoutpronouncingontheirrealconstitution,whilemetaphysicalholismclaimsthatwholesaredistinctentities,whoseexistencecannotbereducedtothatoftheitemscomposingthem.Holisticviewsinthephilosophyofscienceandthephilos-ophyoflanguageproposethatthemeaningandtruthofourclaimscannotbeassessedonebyone,butmustbeassessedaspartoftheories,bodiesoftheory,orallwebelieveabouttheworld.humanismAnintellectualmovementofRenaissanceItalywhicharguedthattheGreekandLatinclassicscontainedallthewisdomneededtoleadamoralandeffectivelife.Suchaviewcontributedtothedevelopmentofarigorouskindofclassicalscholarship,whichattemptedtocorrect,andbetterunderstand,theancienttexts.Acorollaryofthisviewwasanincreasingintellectualconfidenceinthepowerofhumankindtodiscovertruthandfalsehood.iconographicDerivedfromtheGreekwordfor‘image’:hence,anartefactofsomekind(suchasapainting)thatvisuallyresemblestheobjectitrepresents.idealismTheviewthattheexistenceofobjectsdependswhollyorinpartonthemindsofthoseperceivingthemorthatrealityiscomposedofmindsandtheirstates.Therearemanyvarietiesofidealism,rangingfromPlato’sdoctrineofindependentlyexistingideasorformstoBerkeley’ssubjectiveidealismandHegel’sabsoluteidealism.Kantattemptedtocombineempiricalrealismwithtranscendentalidealism.ideasForPlato,analternativetermforForms,theunchangingindependentlyexistingbasesoftheperceivedworldandthoughtabouttheworld,andinseventeenth-andeighteenth-centuryphilosophy,thevehiclesofsensoryrepresentationofexternalobjectsandofthought.Lockeand882\nGLOSSARYLeibnizinitiateddisputeswhetherideascouldbeinnate,inusindependentofsensoryexperience.BerkeleyrejectedLocke’sdistinctionbetweentheideasofprimaryqualities(whichresemblethequalitiesthatproducethem)andtheideasofsecondaryqualities(whichareproducedbyqualitiesthattheydonotresemble).Humearguedthatideasoriginatedinimpressions,butstillretainedsensoryandintellectualfunctionsforideas.Kantseparatedthesefunctions(andusedtheterm‘idea’forotherpurposes).identityInordertouseaconceptwemustbeabletoindividuatedifferententitiesfallingunderthatconceptandtoidentifytheseindividualentitiesovertime.Itisunclearwhetherornotindi-viduatingandidentifyingentitiesmustrefertothekindconceptsunderwhichtheyfall.Aspecialcaseofidentity,discussedbyLocke,Humeandmanycontemporaryphilosophers,ispersonalidentity.Boththememoryandbodilycriteriaforpersonalidentityhaveencountereddifficulties,leadingParfittoquestiontheimportanceofpersonalidentity.Theidentitytheoryofmind,accordingtowhichmentalstatesareidenticalwithstatesofthecentralnervoussystem,ledtomuchdiscussion,althoughotherformsofmaterialismhavesupplanteditincurrentcontroversy.Kripke’srejectionofthenotionofcontingentidentityinfavourofnecessaryidentityhelpedtorenewdiscussionofthenatureoftheidentityrelationitself.idolForFrancisBacon,anillusionorfalseappearancethathinderstheacquisitionofknowledgethroughprejudiceorfalsewaysofthinking.illocutionaryactInJ.L.Austin’stheoryofspeechacts,whatonedoesinutteringwhatoneutters.Aperlocutionaryactiswhatonedoesbyutteringwhatoneutters.imaginationTheabilitytorepresentobjectsorstatesofaffairswhichcannotexist,whichdonotexistorwhichdonotexisthereandnow.Imaginationisbothcondemnedforitslinkwithfalsityandprizedforitsroleinartisticcreativityandhumanunderstanding.InKant’saccount,imaginationperformsindispensablerolesinperceptionasanintermediarybetweenoursensibilityandunderstandingwhichallowsustohaveknowledgeofaunifiedworld.immaterialsubstanceBerkeleyarguedthatthenotionofmaterialsubstancecouldnotbesustained,butthatimmaterialsubstances,humanmindsandGod,werecrucialtoouraccountofreality.immutableLiterally,notabletochange.individualismAnapproachtoethics,socialscienceandpoliticalandsocialphilosophywhichemphasizestheimportanceofhumanindividualsincontrasttothesocialwholes,suchasfamilies,classesorsocieties,towhichtheybelong.Indifferentcontexts,individualismiscontrastedtoholismandcollectivism.Metaphysicalindividualismclaimsthatsocialobjectslikesocietiescanbereducedtoindividuals.Methodologicalindividualismdoesnotmakemetaphysicalclaims,butratherconstrainsthewaysweexplainsocialaction.inductionAprocessofreasoningcontrastedwithdeductioninwhichconclusionsaredrawnthatallindividualsofakindhaveacertaincharacteronthebasisthatsomeindividualsofthekindhavethatcharacter.inferenceTheprocessofreasoningwherebyonestatement(theconclusion)isderivedfromoneormoreotherstatements(thepremises).infiniteregressAninfiniteregressinaseriesofpropositionsarisesifthetruthofpropositionP1requiresthesupportofpropositionP2,andforanypropositionintheseriesPn,thetruthofPnrequiresthesupportofthetruthofPn+1.TherewouldneverbeadequatesupportforP1,becausetheinfiniteseriesneededtoprovidesuchsupportcouldnotbecompleted.intensionalityAfeaturecharacterizingsentencesaboutmentalstates,accordingtowhichthetruthvalueofthesentencemaybealteredbyreplacingexpressionsinthesentencebyotherexpressionsreferringtothesameobjects.Inextensionalcontexts,unlikeintensionalcontexts,suchsubstitutionsdonotaffectthetruthvalueofthesentence.intentionalityAcharacteristicfeatureofmentalandlinguisticactsorstatesaccordingtowhichtheyhaveanobjectorcontentandarethusaboutsomething.Anintentionalobjector883\nGLOSSARYcontentisthatwhichamentalorlinguisticstateoractisabout.Problemsarisebecauseintentionalobjectsneednotexistandmentalcontentsneednotbetrue.intersubjectivityForHusserl,thestructureallowingustoseeothersasegosratherthanasobjectsandthustoescapethesolipsisticimplicationsthatsomecriticsderivefromhisaccountofoneselfasatranscendentalego.introspectionPopularly,any‘inward-looking’atone’sownmentalstates,easilylampoonedasaninadequatesourceofevidence.Morespecifically,inthehistoryofpsychology,atechniquedevelopedtoasophisticatedexperimentallevelattheendofthenineteenthcentury.irreducibleToreduceX’stoY’swouldbetoshowhowX’swere,inreality,onlyY’sorinalinguisticguisetoshowthatX-talkcouldbesystematicallyeliminatedinfavourofY-talkinawayinvolvingnolossofcontent.Propertiesaresaidtobeirreduciblethen,iftheyresistsuchreductions.Leibniz’sLawIfA=B,foranytruestatementaboutAtherewillbeacorrespondingtruestatementaboutB,andviceversa.Therearedisputesoverthewaytodealwithstatementsforwhichthislawdoesnotseemtohold.lifeworldOurnaturalconceptionoftheworld,includingsuchthingsasphysicalobjects,culturalandhistoricalartefactsandsocialinstitutions.ForHusserl,thelifeworldhaspriorityoverthescientificaccountoftheworldthatisgroundedinit.logicalatomismTheviewheldforatimebyRussellandWittgensteinthatforlanguagetohavemeaningitmustbeanalysableintomutuallyindependentpropositions,theatomicelementsofwhichcorrespondtoelementsinstatesofaffairs.ForWittgensteinatthisstage,propositionshadmeaningbylogicallypicturingpossiblestatesofaffairs.Russell’slogicalatomism,unlikeWittgenstein’s,wastiedtoanempiricalinterpretation.logicalpositivismAgeneralphilosophicalposition,alsocalledlogicalempiricism,developedbymembersoftheViennaCircleonthebasisoftraditionalempiricalthoughtandthedevelop-mentofmodernlogic.Logicalpositivismconfinedknowledgetoscienceandusedverificationismtorejectmetaphysicsnotasfalsebutasmeaningless.Theimportanceofscienceledleadinglogicalpositiviststostudyscientificmethodandtoexplorethelogicofconfirmationtheory.logomachyAcontroversyofnorealsubstance,dependingonamerelyverbaldispute.materialbiconditionalThebiconditionalifandonlyif(‘iff’)isarelationbetweentwostatementspandq,suchthatpimpliesqandqimpliesp.Thebiconditionalismaterialiftheimplicationsarecontingentandstrictiftheimplicationsarenecessary.materialismThedoctrinethatallitemsintheworldarecomposedofmatter.Becausenotallphysicalentitiesarematerial,therelateddoctrineofphysicalism,claimingthatallitemsintheworldarephysicalentities,hastendedtoreplacematerialism.matterandformAristotle’smetaphysicaldistinctionusedtoexplainthepropertiesofobjects.Matter,formandthecompositeofmatterandformareAristotle’scandidatesforthestatusofsubstance.mauvaisefoiForSartre,self-deceptioninwhichoneattemptstoevadeone’sfreedombydenyingone’sbeing-for-itselfinfavourofthebeing-in-itselfofanobject.Hisaccountraisesthegeneralquestionofhowself-deceptionispossible.maximAstatementexpressingageneraltruthorruleofconduct.methodPhilosophicalmethodsarecombinationsofrules,proceduresandexamplesdeterminingthescopeandlimitsofphilosophyandestablishingacceptablewaysofworkingwithinthoselimits.Thequestionofphilosophicalmethodisitselfamatterforphilosophyandconstitutesamajorfeatureofphilosophy’sreflectivenature.Philosophersdisagreeaboutwhatisanappropriatephilosophicalmethodandabouttherelationshipbetweenphilosophicalmethodandthemethodsofotherdisciplines,especiallyscientificmethod.methodologicalConcerningthestudyofmethod,particularlyscientificmethod.Questionsaskedincontemporarymethodologyconcernnotonlythecharacterizationofscientificmethod,884\nGLOSSARYbutalsowhetherasinglesuchcharacterizationisnecessary,orcandojusticetothemultitudeofapproachesanddevicesactuallyused.minimalismInphilosophyoflanguage,theviewthatmeaninghasnosubstantialnatureandsothereisnothingsubstantiveforatheoryaboutmeaningtosay.modelInscience,arepresentationsuchthatknowledgeconcerningthemodeloffersinsightabouttheentitymodelled.Whethermodelsareheuristicdevicesoressentialfeaturesofscientificexplanationisamatterofdebate.Mathematicalmodelsareinterpretationsofaformalsystemassigningtruthvaluestotheformulaeofthesystem,thustestingthesystemforconsistency.modernForphilosophicalpurposes,theperiodofphilosophyandgeneralintellectuallifefollowingDescartesandhiscontemporaries.Thebroadcharacteristicsoftheperiodareoftentakentoincludeanemphasisonindividualism,theintellect,theuniversalityofjudgement,andtheconsequenceofadoptingthesefeaturesasstarting-points.modusponensandmodustollensModusponens,oraffirmingtheantecedent(theifclauseinanif-thenproposition),isanargumentofthevalidform,‘Ifp,thenq;p;thereforeq’.Modustollens,ordenyingtheconsequent(thethenclauseinanif-thenproposition),isanargumentofthevalidform,‘Ifp,thenq;notq;thereforenotp.’naturalkindAnaturallyexistingspeciesofthing,suchascats,oranelement,suchasgold.Thereisphilosophicaldisputeoverhownaturalkindtermsarerelatedtonaturalkinds.naturalselectionAcentralthesisofthebiologistCharlesDarwinwhichsuggeststhatwithineverypopulationoflivingorganismstherearerandomvariationswhichhavedifferentsurvivalvalue.Thosewhichaidsurvival(orenhancereproductivecapacity)are‘selected’bybeinggeneticallytransmittedtosucceedinggenerations.necessarySeecontingent–necessary.neo-PlatonistAnymovementwhichseekstoreinstatePlato’sdoctrinesasacentralphilosophicalframework.MorenarrowlythetermreferstotheschoolinitiatedbyPlotinusinthird-centuryGreece,andtoalaterRenaissanceformwhoseleadingmemberswereNicholasCusanus,MarsilioFicinoandPicodellaMirandola.neutralmonismApositionaccordingtowhichthedifferencebetweenmindsandbodiesderivesfromdifferentarrangementsofthesameneutralentities.Theentitiesareneutralbecausetheythemselvesareneithermentalnorphysical.Thispositionproposedasolutiontothemind–bodyproblem,buttherearedifficultieswiththeneutralstatusofthatwhichconstitutesmindsandbodiesandwithhowarrangementsofwhatisneutralcanissueinmindsandbodies.Ifexperiencesareproposedastheneutralentities,itisnotclearwhetherneutralmonismclarifiesorobscuresthenatureofexperience.nihilismForNietzsche,thelossofbeliefinahigherworldandintheidealsandvaluesgroundedinsuchaworld.noemaandnoesisThatwhichisthoughtaboutandtheactofthinking.ForHusserl,thematerialandtheformalaspectsofintentionalexperience.nominalismTheviewthattheonlyfeaturethatparticularsfallingunderthesamegeneraltermhaveincommonisthattheyarecoveredbythesameterm.Nominalismisopposedtorealism,forwhichuniversalsarerequiredtoexplainhowgeneraltermsapplytodifferentparticulars.Fornominalism,language,ratherthanindependentreality,underliesperceivedlikeness.Manyphilosophersareattractedtotheontologicalausterityofnominalism,butproblemsremainconcerninghowlanguage,especiallypredication,worksonnominalistprinciples.nomotheticLiterally,‘constitutingthelaw’,butgenerallyusedtorefertothephilosophicalaspectsoflawortoscientifictheoriesconstitutedbylaws.non-EuclideangeometryEuclid’sFifth‘Parallels’Postulate(orAxiomXI)isrenderedinformallyas:‘throughagivenpointPnotonalineL,thereisoneandonlyonelineintheplaneofPandLthatdoesnotmeetL’.Non-Euclideangeometriesexploresystemsinwhichtwodifferent885\nGLOSSARYdenialsofthispostulateareused:LobachevskyangeometrycontainsaninfinitenumberofparallelsthroughP;ReimanniangeometrycontainsnoparallelsthroughP.Reimanniangeometryhasplayedacrucialroleinthedevelopmentofthegeneraltheoryofrelativity.non-monotoniclogicsInlogic,deductivevalidityiscumulative,ormonotonic,inthatnomatterwhatfurtherpremisesareaddedtoanoriginallyvalidargument,theargumentremainsvalid.Non-monotoniclogics,usedinartificial-intelligenceresearch,explorelogicalsystemsinwhichmonotonicitydoesnothold.nothingnessForSartre,theontologicalrealityofnegationthatisderivedfromconscious-ness’snegationofbeing-in-itself.occasionalismMalebranche’saccountofcausationaccordingtowhichratherthanoneeventcausinganother,Godwillsthesecondeventontheoccasionofthefirstevent.Ockham’srazorTheprincipleenunciatedbythemedievalnominalistWilliamofOckhamthatentitiesarenottobemultipliedbeyondnecessity.Appliedtosystemsofontologyorbodiesofscientifictheory,theprincipleencouragesustoaskwhetheranyproposedkindofentityisnecessary.Thisprincipleofmetaphysicaleconomyretainsinfluenceincontemporaryphilosophy,althoughinjudgingrivalsystemsitisnotalwaysclearwhichbestmeetstherequirementsofOckham’srazor.omnipotenceandomniscienceAttributesofbeingall-powerfulandall-knowing,tradit-ionallyascribedtoGod.Thetwocombinedwithdivinegoodnessgiverisetotheproblemsofexplaininghowtherecanbeevilintheworld.Puzzlesaboutomnipotencehaveledtoexplanationsthatitisconstrainedbylogicalandmetaphysicalpossibility.Divineomniscience,includingforeknowledge,hasbeenseenasathreattohumanfreewill,andotherphilosophershavebeenconcernedaboutwhatanomniscientbeingwithoutabodyinspaceandtimecouldknow.ontologyThestudyofthebroadestcategoriesofexistence,whichalsoasksquestionsabouttheexistenceofparticularkindsofobjects,suchasnumbersormoralfacts.operatorAnexpressionwhichaltersthelogicalpropertiesofotherexpressionstowhichitisapplied.Asententialoperatorcanbeappliedtosentencestoyieldnewsentences.OvermanForNietzsche,personsofanewtypewhorecognizethattheyarethesolesourceofthevaluesandidealsbywhichtheylive.paradoxAnargumentwhichseemstojustifyaself-contradictoryconclusionbyusingvaliddeductionsfromacceptablepremises.perceptionOurawarenessoftheworldanditscontentsthroughsensoryexperience.Theanalysisofperceptionandtheattempttodealwithscepticalargumentsaboutperceptualknowledgearecentralphilosophicaltopics.Perceptioninvolvesbothourcapacitytobesensorilyaffectedbyexternalobjectsandourabilitytobringtheseobjectsunderconcepts,althoughothercapacitiesmightalsohavearoletoplay.Whatweperceiveandhowtheseobjectsofperceptionarerelatedtousandtophysicalobjectsaremattersofcontinuingconcern.perspectivalismTheviewthattherecanbenonon-perspectivalknowledgeandthatallknowledgeisfromaperspective.phenomenologicalreductionForHusserl,themethodofignoringrealtemporalparticularsandfocusinginsteadonthegeneralandidealatemporalfeaturesofthings(theeideticreduction)andsettingasideallobjectstranscendenttoconsciousnesstofocusontheintentionalcontentsimmanentwithinconsciousness(theepoche¯ortranscendentalreduction).phenomenologyForHegel,thestudyofthedialecticaldevelopmentofSpiritthroughstagestowardsrational,self-consciousfreedom;forHusserl,aphilosophicalmethodbasedonthereflectiveanddescriptivestudyofconsciousnessfocusedontheintentionalityofmentalstates.Thestructureofconsciousnessrevealed,whichincludesanegothatexistsabsolutely,aimedtoprovideasurefoundationforknowledge.InresponsetoFrege’searlycriticism,Husserlattemptedtodrawasharpboundarybetweenphenomenologyandpsychology.Inhislaterwriting,Husserl886\nGLOSSARYalteredmanyfeaturesofhisnotionofphenomenology,andotherwriters,especiallyHeidegger,usedtheterminradicallydifferentways.phronesisForAristotle,theintellectualvirtueofpracticalwisdom,akindofmoralknowl-edgeguidingustowhatisappropriateinconjunctionwithmoralvirtue.picturetheoryofmeaningWittgenstein’sviewintheTractatusLogico-Philosophicusthatapropositionhasmeaninginvirtueofsharingaformwithanactualorpossiblestateofaffairs.Thepropositionprovidesalogicalpictureofthestateofaffairs,andistrueifitselementsstandinthesamerelationastheobjectsinthestateofaffairs.PhilosophersdisagreeovertheextenttowhichWittgensteinmovedawayfromthistheoryinhislaterwritings.pluralismInmetaphysics,thebeliefthatthereismorethanonekindoffundamentalrealityoroffundamentalexistents.Hence,pluralismstandsincontrasttomonism(onekindoffundamentalrealityorexistent)anddualism(twokindsoffundamentalrealityorexistent).Inethics,thebeliefthatthereismorethanonekindoffundamentalgoodorsupremeethicalvalue.polisTheGreektermforcity-state,fromwhichtheterm‘politics’isderived.Thecharacterofthepoliswasexaminedinancienttimes,andmanylaterthinkersyearnedforitsre-establishmentasanidealofpoliticallife.possibleworldTheanalysisofstatementsintermsof‘possibleworlds’comesfromthesemantictreatmentofmodallogic–thelogicofpossibilityandnecessity.Apossibleworldisawaytheworldmighthavebeen.Anecessarystatementisonethatistrueinallpossibleworlds;acontingentstatementisonethatistrueinatleastonepossibleworld.postmodernismIfDescartesisseenasthefatherofmodernism,thenpostmodernismisavarietyofculturalpositionswhichrejectmajorfeaturesofCartesian(orallegedlyCartesian)modernthought.Hence,viewswhich,forexample,stressthepriorityofthesocialtotheindividual;whichrejecttheuniversalizingtendenciesofphilosophy;whichprizeironyoverknowledge;andwhichgivetheirrationalequalfootingwiththerationalinourdecisionproceduresallfallunderthepostmodernumbrella.prescriptivestatementAstatement,forexampleinethics,whichsayshowthingsshouldbe,asopposedtoadescriptivestatement,whichsayshowthingsare.presuppositionAgenerallyimplicitassumption(thoughitcanbemadeexplicit)underlyingaclaimoraprocessofinference.primaryandsecondaryqualitiesSeventeenth-centuryscientistsandphilosophersattemptedtodistinguishqualitieslikesizeandshape,whichproduceinusideas(ideasofprimaryqualities)thatresemblethequalitiesthemselves,fromqualitiesthatproduceinusideas(ideasofsecondaryqualities),likecolourandtaste,whichdonotresemblethequalitiesthemselves.Theformerideas,unlikethelatter,offeredsomethingthatcouldbemeasured,andwerethusconsideredasuitablebasisforscientificexplanation.Berkeleyarguedagainstthedistinction.principleofnon-contradictionThelogicalprinciplerejectingthepossibilitythatpropositionsoftheform‘pandnotp’aretrue,thatasubjectcanbeandnotbe,orthatwecanascribeanddenythesameattributetothesamesubject.propositionThatwhichischaracteristicallystatedbyadeclarativesentenceandcanbetrueorfalse.Understandingthenatureandstructureofpropositionsisoftenseenasthecentraltaskofthephilosophicalexaminationoflogic.Philosophersconsidertheapparentlydifferentfunctionsofcomponentsofpropositions(names,predicatesandlogicalconstants)andhowtheyareunifiedintosomethingcapableofhavingatruthvalue.Theyaskabouthowtheform,meaninganduseofpropositionsarerelatedandhowdifferentpropositionshavelogicalrelations.Theyaskhowlinguisticorpsychologicalstatescanhavepropositionalcontents.Theexistenceofpropositions,asopposedtosentences,ischallengedbythosesuspiciousoftheirabstractnature.psychologismAprogramme,criticizedbyFregeandthenbyHusserl,toexplainlogicintermsofhumanpsychology.887\nGLOSSARYquantifierModernpredicatelogicusesquantifierexpressionssome(|)andall(...)insentenceswithvariables(x,y,...),predicates(F,G,...),relations(R,...),identityandlogicalconstants(and,or,not,if-then).(|x)(|y)(ifFxandGy,thenx=y)reads‘Forsomexandforsomey,ifxisFandyisG,thenxisidenticalwithy’.Theindividualormultipleuseoftheuniversalquantifier‘all’andtheexistentialquantifier‘thereexists’tobindvariablesinsentenceshasbeenseenasthekeytothedevelopmentofapowerfulandflexiblesystemofmodernpredicatelogic.quantummechanicsAmodernphysicaltheory(muchdevelopedandrefinedsinceNeilsBohr’sground-breakingworkin1913)whichdealswiththestructureandbehaviourofsubatomicparticles.Ithasgivenrisetophilosophicalproblemsofitsown(somequantumphenomenaseemtorequireanon-classicallogic)andhasalsobeenusedbybothsidesinthephilosophicaldisputebetweenrealistandinstrumentalistconstrualsofscientifictheories.rationalnumberAnynumberoftheformx/y,wherexcanbeanypositiveornegativeintegeror0,andyisapositiveinteger.Anirrationalnumberisanyrealnumberwhichisnotrational.realnumberArealnumberisanynumberwhichcanberepresentedasanon-terminatingdecimal.realismAvarietyofdoctrinesindifferentareasofphilosophyholdingthatentitiesorfactsofcontestedsortsexist.Thereare,ofcourse,differentargumentsconcerningtherealityofnumbersinmathematics,therealityofmoralfactsinethics,andtherealityoftimeinphysicsormetaphysics.Thekindofrealityascribedtouniversalsdiffersfromthekindofrealityseenasbelongingtocommon-sensematerialobjectsortotheoreticalentitiesinscience.Variousrealismsarehenceopposedbynominalism,idealism,instrumentalism,reductionism,eliminativism,conventionalism,constructivism,relativismandanti-realism.Kantarguedforbothempiricalrealismandtranscendentalidealism.InrecentyearsmanyphilosophershavediscussedMichaelDummett’sargumentforanti-realism,inwhichherejectstheclaimthateverypropositionmustbeeithertrueorfalseandarguesthatrealismmustbefalsebecauseitimpliesthisclaim.reasonAnabilitytomovefromthetruthofsomebeliefstothetruthofothers.Somephiloso-phershaveseenthiscapacityasmoreorlesssufficienttodetermineonecorrectsystematicaccountofreality,whileothershavearguedthatsuchanaccount,ifpossibleatall,mustbebasedprimarilyuponexperience.Kant,followingAristotle,sawreasondividedbetweentheoreticalandpracticalreason,thelatterissuinginactionsratherthanbeliefs,butheldthatatadeeplevelthetwocapacitieswerethesame.Hegelsawreasonandmuchelsealteringatdifferentstagesofhistoricaldevelopment.Humerestrictedpracticalreasontofindingmeanstoobtainendssetoutbythepassions;othershaverejectedthemeans–endsaccount.Reasonenterstheaccountofinstitutionsthroughmodelsoftheinteractionofthechoicesofindividualsandthroughthedirectassessmentofpracticesandsocieties.recursivefunctionAfunctionforwhichthevalueofthefunctionforanyargumentXn+1isafunctionofthevalueofthefunctionfortheargumentXn.reductionismTheviewthattheconceptsortheoriesofallsciencescanbereducedtophysics.reflectiveequilibriumAtermusedbyNelsonGoodmanandJohnRawlsforatwo-wayreconciliationbetweenjudgementsandprinciples.Judgementsaboutindividualcasesareguidedbyprinciples,butprinciplescanbemodifiedinlightofjudgements.Equilibriumisreachedwhenprinciplesandjudgementsfitoneanotherwithoutfurtheralteration.Anequilibriumisalwaysliabletobeupsetbynewcases,butthatisalsotrue,althoughlesstransparentlyso,ifonetriestodetermineprinciplesbyotherprocedures.relativity(theoryof)ThespecialtheoryofrelativityisamodernphysicaltheoryduetoEinstein,givinganaccountinwhichneitherspacenortimehasanindependentabsolutevalueorexistencebutiseachrelativetotheother.Thustheclassicalviewofspaceandtimeisreplacedwithoneinwhichthetwoareaspectsofthesameunderlyingreality:space-time.Thegeneraltheoryofrelativityextendedthespecial(fromconsideringframeworksinuniformrelativemotion888\nGLOSSARYtoconsideringframeworksinarbitraryrelativemotiontooneanother)andisthecurrentlyacceptedbasisforourtheoryofgravitation.rightsLegalormoralcapacities,oftencorrelatedwithduties,whichmaybeexercisedwithoutinterferencebyothers,includingthestate,orinsomecaseswiththeassistanceofothers.Rightscanregardsuchmattersasbelief,actions,relationships,property,orthesafetyandintegrityofoneself.Somerightsinvolvelimitingtherightsofothersorthecreationofduties.Rightscanbeconsideredonebyoneorfromthestandpointofasystemofrightsandduties.Philosophersdiscussthestatusofrightsinmorality,inparticularwhetherrightscanprovidethebasisofmoralorpoliticalphilosophyormustbeunderstoodwithinthecontextofothernotions,suchasprinciples,virtues,utilityorasocialcontract,fromwhichtheirlegitimacymightderive.Philosophershavediscussedtheclaimthatrights‘trump’otherconsiderations,theclaimthattherearecollectiveaswellasindividualrights,andtheclaimthatanimalshaverights.Exploringthislastquestioncanhelptoseewhatonemustbelikeinordertoholdrights.Philosophershavealsoconsideredtherelationshipbetweenlegalandmoralrightsand,inthefaceofBentham’sattackonnaturalrights,haveexaminedthenotionofnatural,universalorhumanrights.rigiddesignatorAsubjecttermdesignatingthesameobjectinalldifferentpossiblesituations(orpossibleworlds)inwhichtheobjectexists,solongasthemeaningofthetermisheldconstant.Anon-rigiddesignatorwithconstantmeaningcandesignatedifferentobjectsindifferentpossiblesituations(orpossibleworlds).RorschachtestApersonalitytest,bearingthenameofitsSwissinventor,inwhichasetofink-blotsisshowntoapatient,whohastodescribewhattheyresembleorsuggest.ruleAprincipleguidingaction.ForKant,conceptsarerules,theunderstandingisthefacultyofrules,andouruseofrulesiscentraltoouraccountofobjectivity.MuchofWittgenstein’sdiscussionoffollowingarulerunsparalleltotheKantianinsightthatrulesdonotdeterminetheirownapplication.ThequestionofwhatgivesaruleauthorityandholdsitanditsapplicationinplaceledtotherejectionofthepossibilityofaprivatelanguageandWittgenstein’semphasisonpracticesandformsoflife.Interrelationsamongrules,rolesandpracticesarecentralconcernsinthephilosophyofsocialscience.Discussionsofwhatconstitutesalegalruleandhowsuchruleshavenormativeforcearemainfeaturesofthephilosophyoflaw.ruleutilitarianismAversionofutilitarianisminwhichgeneralrulesratherthanactsareassessedforutility,thusshiftingconcernfromindividualstopracticesandinstitutions.Actsareendorsednotintheirownright,butbecausetheyaccordwithpracticesorinstitutionswhichmeetthetestofmaximizingutility.Russell’sparadoxAparadoxbasedonthenotionofclassmembershipdiscoveredbyBertrandRussellandunderminingthecrucialnotionofclassorsetinFrege’sfoundationsofmathemat-ics.Theparadoxledtoimportantchangesinsettheoryandinthenotionofaset,inpartalsotopreventfurtherparadoxesfromarising.Theparadoxarisesfromaskingwhethertheclassofallclassesthatarenotmembersofthemselvesisamemberofitself.sanctionsNegativeandpositivesanctionsarepunishmentsorrewardsforbehaviourthattransgressesorisinaccordwitharule.scopeInlogic,theshortestpropositionalfunctioninwhichalogicaloperatoroccurs.Scopeambiguitiesarecommoninordinarylanguagebutareeliminableduringformalization.semanticsInthestudyoflanguage,semanticsisconcernedwiththemeaningofwords,expressionsandsentences,ofteninrelationtoreferenceandtruth.Semanticsiscontrastedwithsyntax(thestudyoflogicalorgrammaticalform)andpragmatics(thestudyofthecontributionofcontextualfactorstothemeaningofwhatlanguageuserssay).Meta-semantictheoriesstudykeysemanticnotionssuchasmeaningandtruthandhowthesenotionsarerelated.sensedataInempiricisttheoriesofperception,whichwerepopularearlyinthetwentiethcentury,thatwhichisgivenbythesenses.Questionsariseconcerningwhetheranythingispurely889\nGLOSSARYgiveninperception,whatmightbegiven,andhowwhatisgivenmightberelatedtotheexternalobjectsofcommonsense.setAsetisacollectionofdefinitedistinguishableentities.Settheory,however,allowsforthenullset:thesetthathasnomembers.socialcontractAnactualorhypotheticalcontractprovidingthelegitimatebasisofsovereigntyandcivilsocietyandoftherightsanddutiesconstitutingtheroleofcitizen.Thecontractcanbeagreedbetweenpeopleandaproposedsovereignoramongthepeoplethemselves.solipsismTheviewconfiningrealitytooneselfandone’sexperiences.subspecieaeternitatisLiterally,‘fromtheaspectofeternity’,henceaclaimforanabsolutelynon-perspectivalaccountofknowledge.subjective–objectiveThesubjectcontributeswhatissubjectivetosuchthingsasperceptual,moralandaestheticjudgementandexperience;theobjectsofsuchjudgementsandexperiencecontributewhatisobjective.Thesubjectiveseemspronetovariationamongsubjects,whiletheobjectiveappearstoprovideabasisforuniversalagreement.Thereisdisagreementoverthecontributionofthesubjectandtheobjecttosuchjudgements.Differentnotionsofobjectivitymightbesuitableindifferentdomains.Historicaljudgements,forexample,mightbeobjectiveifthehistorianmakingthemisunbiasedratherthanthroughhavingafavouredrelationtorelevantobjects.sublationAchangewhichbothcancelsandpreservesanentityorconceptbyraisingittoahigherlevel.Thefullcomplexmeaningofthisterm(aufhebeninGerman)wasestablishedbyHegel.substanceSomethingwhichcanexistbyitself,isthesubstrateunderlyingtheexistenceofotherthings,andisthesubjectofwhichotherthingsarepredicated.InhisMetaphysicsAristotleconsideredwhatcanbesubstance:matter,form,oracombinationofmatterandform.Accordingtovariouscriteriaheused,differentanswersseemedplausible,althoughhefinallypreferredform.Seventeenth-centuryphilosophers,includingDescartes,SpinozaandLeibniz,cametostrikinglydifferentsolutionstotheproblemofwhatqualifiesasbeingsubstance.Locke’saccountseemstosuggestanunknowablesubstratumwhichfallsoutofanyaccountofknowl-edge,butitcanalsobeseentoofferacorpuscularsubstratelinkedtohisdoctrineofprimaryqualitiesandtheirexplanatoryroleinscience.suigenerisLiterally,‘ofitsownkind’,orunique.supervenienceApropertyFsupervenesonapropertyG,firstly,ifanythingwhichhaspropertyFhasitinvirtueofhavingpropertyGand,secondly,ifsomethinghaspropertyFinvirtueofhavingpropertyG,thenanythingelsehavingpropertyGwouldalsohavetopossesspropertyF.Supervenienceisintendedtoallowfornon-reductiverelationsamonghierarchiesofproperties.syllogismAnargumentaccordingtoAristotle’slogicaltheoryinvolvingamajorpremise,aminorpremise,andaconclusion.teleologyTheexplanationofathingorfeatureofathingintermsofitsfunctionorgoal.theodicyOftenrefersonlytoLeibniz’stextofthesamename,butcanrefertoanyphilosoph-icalpositionwhichseekstoprovetheexistenceofGod,ortojustifyabeliefinsuchexistenceinthefaceoftheproblemofevil.things-in-themselvesKantdistinguishedappearances(phenomena)fromthings-in-themselves(noumena).Things-in-themselvesaremeanttoexistindependentlyofhowweexperiencethem,inparticularindependentlyofspace,timeandthecategories.thirdmanargumentPlato’sargument,endorsedbyAristotle,thattherelationbetweenaFormandthegroupofthingshavingthatFormincommonleadstoaninfiniteregress.thoughtForFrege,athoughtisthesenseofasentencewhichcanbeusedtomakeanassertionortoaskaquestionthatisanswerablebyeither‘yes’or‘no’.Thecontentsofthoughtscanbetrueorfalse.Thoughtsinthissensearelogicalorconceptualratherthanamatterof890\nGLOSSARYindividualpsychology.Differentindividualsmaysharethesamethought,althoughtheycannotsharethesameactofthinking.thoughtexperimentAnattempttoconceivetheconsequencesofaninterventionintheworldwithoutactuallyintervening.Insomecases,anactualexperimentwouldbepreferablebutisimpossibleinpracticeorperhapsevenintheory.three-valuedlogicClassicallogicallows‘true’and‘false’astheonlyvaluesforpropositions,asexpressedbythelawofexcludedmiddle(whichstatesthateverypropositioniseithertrueorfalse).Three-valuedlogiccanrejectthislawbyaddinganadditionalvaluelike‘indeterminate’orcanreplacethetwoinitialvalueswiththreeothervalueslike‘necessarilytrue’,‘necessarilyfalse’and‘contingent’.TorricelliTheearlyseventeenth-centurynaturalphilosopherwhowasthefirstpersontocreateasustainedvacuum,thusapparentlythreateningDescartes’sclaimthattherecouldbenovacuuminnature.transcendentalForKant,‘transcendental’iscontrastedwith‘transcendent’.Somethingtranscendentgoesbeyondthelimitsofexperience,whilesomethingtranscendentalrelatestotheconditionsofthepossibilityofexperience.truthPropositions,statements,sentences,assertionsandbeliefshavebeenofferedasappropriatebearersoftruthorfalsity.Understandingtruthisfilledwithdifficulty.Philosophershaveexploredthepossibilitythattruthis:acorrespondencebetweenwhatwesayandhowthingsare;amatterofcoherencebetweenstatementsandabackgroundofsettledbeliefs;anideallimitwhichenquirywillapproach;afeatureofassertionswhichfunctionwellinenquiryorinlifemoregenerally;amatterofgivingatruthdefinitionforalanguage;aredundancy,because‘Itistruethatp’isequivalentto‘p’;ordisclosednessofbeing.Someofthesetheoriesarecompatibleandmightbeintegratedinamorecomprehensivetheory.Onsomeaccounts,eachpropositionistrueorfalseonitsown,whileothersadoptaholisticview.Therelationbetweenmeaningandtruthisofcentralphilosophicalconcern.truthfunctionThetruthvalueofacombinationofpropositionswhichdependsonlyonthetruthvaluesoftheconstituentpropositionsandthelogicalconstants(and,if-then,or,not)bywhichtheyarecombined.Truthfunctionscanbesetoutintruthtables:pqporqTFTTTTFFFFTTNotallpropositionsaretruthfunctionsoftheirconstituents.Forexample,thetruthvalueof‘Ibelievethatitwillrain’isnotdeterminedbythetruthvalueof‘Itwillrain’.understandingThecapacitytouseconceptsandtobringindividualsunderthem.Kantdis-tinguishedsensibility,understandingandreasonasfundamentaltoourcapacityforexperienceandknowledge.Heunderstoodconceptsasrulesandsawtheunderstandingasthefacultyofrules,includingbothempiricalconceptsandthecategoriesaspureconceptsoftheunderstand-ing.Inthesocialsciencesandhistory,DiltheyandWeberhavecontrastedunderstanding(verstehen)andexplanation,withexplanationprovidingthecausalaccountsofscienceandunderstandingofferinginsightintosuchthingsashumanlives,cultureandhistoricalperiods.Hermeneuticshasbeenproposedasthemethodappropriatetounderstanding.Philosophershavedisagreedovertheclaimeddifferencebetweenexplanationandunderstanding,aboutthecharacterofunderstanding,andaboutthemethodologicalimplicationsofrecognizingunderstandingasadistinctivemodeofknowledge.891\nGLOSSARYuniversalsAbstractobjectsintendedtoexplainhowgeneraltermshavemeaningandhowtheyapplytoindividuals.Thereweremedievaldisputesoveruniversalsinvolvingrealism,whichacceptedtheirexistence,andnominalism,whichdeniedit.Somephilosophersseeproblemsinembracinganyabstractobject,especiallyinthiscasebecausetheirrelationtoindividualsseemsopaque,yetitisalsodifficulttoseehowanaccountofhownameshavemeaningandrefertoindividualswillexplainthecontributionofpredicatesandrelationstothemeaningandtruthorfalsityofsentencesorpropositions.Explainingtheserolesmightnotrequireclassicuniversals,derivedfromindependentlyexistingPlatonicForms,butthereareotherpossibilitiesshortofnominalism.verificationprincipleAcentraldoctrineoflogicalpositivismaccordingtowhichthemeaningofapropositionisitsmethodofverification.Claimswithoutamethodofverification,suchasthoseofreligionandmetaphysics,aremeaningless.Verificationismthusofferedacriterionofmeaningfulness.verisimilitudeLiterallyandasusedinaestheticcriticism,theappearanceofbeingtrueorreal.Inphilosophy,asurrogateforthetruthofscientifictheoriesofferedbyKarlPopperaspartofhisvisionofthenatureofscientificprogress.viciousregressAnattempttosolveaproblemwhichreintroducedthesameproblemintheproposedsolution.Ifonecontinuesalongthesamelines,theinitialproblemwillrecurinfinitelyandwillneverbesolved.Notallregresses,however,arevicious.ViennaCircleAnintellectual(mostlyphilosophical)group,ledbyMoritzSchlick,whichmetfrom1924–36,thoughitsinfluencecontinuedformuchlonger.Theirgeneralposition,pro-scienceandhostiletospeculativemetaphysics,gaverisetothedoctrinesoflogicalpositivism.virtueAnexcellenceofmoralorintellectualcharacter.Plato,Aristotleandmanysubsequentphilosophersexploredthenatureofthevirtues,theirrelationsamongthemselvesandtonon-virtuousstates,theirplaceinourpsychologyandtheirroleinachievinghappiness.VirtuesofferabasisforethicalliferivallingthoseprovidedbyKantianprinciplesorautilitariancalculationofhappiness,althoughanaccountofethicsmightreasonablyincludeprinciples,consequencesandvirtues.Arecentrevivalofvirtueethicshasbeenmotivatedinpartbydissatisfactionwiththeabstractuniversalnatureofthemainalternativeviews.Theemphasisoncultivatingvirtuesinconcretehumanindividualscouldcorrectthis,butitisnotclearthataperfectionistconcernforindividualexcellenceissatisfactorytogroundethics.vitalismAnypositionwhichholdsthatlifeisanon-reducible,non-physicalattributeoflivingthings.Specificformsofvitalismhaveariseninthehistoryofphilosophyasreactionstoaperceivedoveremphasisondeterministandmaterialistmetaphysics.willtopowerForNietzsche,thebasichumandrivetoattainahigherandmoreperfectstate,manifestedinform-givingpracticalactivitythatexplainstheworldandthenatureofreality.892\nAppendixManyoftheauthorsdiscussedinthisCompanionarelistedbelowinchronologicalorderwithsomeoftheirmainworks.Thelistisnotcomprehensive,butitshouldbeahelpfulguidetothediversityofphilosophicalquestionsandtothecomplexhistoryofphilosophicalthinking.Plato427–347BCProtagoras;Gorgias;Meno;Timaeus;Philebus;Phaedrus;Symposium;Phaedo;Republic;Theaetetus;Sophist;Parmenides;Cratylus;LawsAristotle384–322BCMetaphysics;NicomacheanEthics;Politics;Categories;DeInterpretatione;PriorandPoste-riorAnalytic;Topics;Physics;DeAnima;Rhetoric;ArtofPoetryMarcusTulliusCicero106–43BCOntheState;OntheLaws;OnDutiesPlotinusAD204–70EnneadsPorphyrythePhoenicianc.232–304IntroductiontoAristotle’sCategoriesStAugustineofHippo354–430Confessions(397–400);CityofGod(413–26)Pseudo-DionysustheAreopagite,fifthcenturyOntheDivineNamesandtheMysticalTheologyBoethiusc.480–524ConsolationofPhilosophyJohnEriugenac.810–75DePraedestinatione;DeDivisioneNaturaeAvicennaofBaghdad(IbnSina)980–1037Healing;TheDirectivesandRemarks;Deliverance\nAPPENDIXStAnselmofCanterbury1033–1109Monologion;ProslogionPeterAbelard1079–1142DialecticaAverroesofCordoba(IbnRushd)1126–98TheIncoherenceoftheIncoherenceMosesMaimonidesc.1135–1204TheGuideforthePerplexedRobertGrossetestec.1170–1253DeLuce;DeMotuCorporalietLuce;HexameronAlberttheGreatc.1200–80OperaOmniaRogerBaconc.1214–92OpusMaiusStThomasAquinas1224–74SummaTheologiaeJohnDunsScotusc.1265–1308Ordinatio;QuaestionesQuodlibetalesWilliamofOckhamc.1285–1349SummaLogicaeJosephAlboc.1360–1444TheBookofPrinciple(1425)NicolausCopernicus1473–1543OntheRevolutionsoftheHeavenlyOrbs(1543)JeanCalvin1509–64InstitutesoftheChristianReligion(1536)MichelEyquemdeMontaigne1533–92EssaysFranciscoSuárez1548–1617DisputationesMetaphysicae(1597)FrancisBacon1561–1626NovumOrganum(1620)GalileiGalileo1564–1642DialogueConcerningtheTwoChiefWorldSystems(1632);DialoguesConcerningTwoNewSciences(1638)HugoGrotius1583–1645OntheLawofWarandPeace(1625)894\nAPPENDIXThomasHobbes1588–1679DeCive(1642);Leviathan(1651);DeCorpore(1656);DeHomine(1658)RenéDescartes1596–1650DiscourseonMethod(1637);MeditationsonFirstPhilosophy(1641);RepliestoObjectionstotheMeditations(1641–2);PrinciplesofPhilosophy(1644);PassionsoftheSoul(1649)BlaisePascal1623–62Pensées(1670)RobertBoyle1627–91TheOriginofFormsandQualitiesaccordingtotheCorpuscularPhilosophy(1666);ADisquisitionabouttheFinalCausesofNaturalThings(1688)BenedictdeSpinoza1632–77ThePrinciplesofDescartes’sPhilosophy(1663);TreatiseontheImprovementoftheIntel-lect;ShortTreatiseonGod,ManandhisWellBeing;TractatusTheologico-Politicus(1670);Ethics(1677)JohnLocke1632–1704EssayConcerningHumanUnderstanding(1690);TwoTreatisesonGovernment(1689);ALetterConcerningToleration(1689)NicolasMalebranche1638–1715TheSearchafterTruth(1674–5);DialoguesonMetaphysicsandReligion(1688)IsaacNewton1642–1727PhilosophiaeNaturalisPrincipiaMathematica(1687)GottfriedWilhelmLeibniz1646–1716DiscourseofMetaphysics(1686);NewEssaysonHumanUnderstanding(1704);Theodicy(1710);Monadology(1714);Leibniz–ClarkeCorrespondence(1717);CorrespondencewithArnauld(1846)GeorgeBerkeley1685–1753EssaytowardsaNewTheoryofVision(1709);ATreatiseConcerningthePrinciplesofHumanKnowledge(1710);ThreeDialoguesbetweenHylasandPhilonous(1713)FrancisHutcheson1694–1746InquiryintotheOriginsofOurIdeasofBeautyandVirtue(1725)ThomasBayes1702–61AnEssaytowardsSolvingaProblemintheDoctrineofChances(1763)ThomasReid1710–96EssaysontheIntellectualPowersofMan(1785);EssaysontheActivePowerofMan(1788)DavidHume1711–76ATreatiseofHumanNature(1739);EnquiryConcerningHumanUnderstanding(1748);DialoguesConcerningNaturalReligion(1799)895\nAPPENDIXJeanJacquesRousseau1712–78DiscourseonInequality(1755);TheNewHéloïse(1761);mileÉ(1762);TheSocialCon-tract(1762);Confessions(1782–9)WilliamBlackstone1723–80CommentariesontheLawsofEngland,8thedn(1778)AdamSmith1723–90TheoryofMoralSentiments(1759);TheWealthofNations(1776)AdamFerguson1723–1816EssayontheHistoryofCivilSociety(1767)ImmanuelKant1724–1804CritiqueofPureReason(1781,2ndedn1787);GroundworkoftheMetaphysicsofMorals(1785);MetaphysicalFoundationsofNaturalScience(1786);CritiqueofPracticalReason(1788);CritiqueofJudgement(1790);religionwithintheLimitsofReasonAlone(1793);MetaphysicsofMorals(1797);OpusPostumumMarquisdeCondorcet1734–94EssayontheApplicationofAnalysistotheProbabilityofMajorityDecisions(1785);TheSketchforaHistoricalPictureoftheProgressoftheHumanMind(1795)JohannGottfriedHerder1744–1803OutlinesofaPhilosophyoftheHistoryofMan(1784–91)JeremyBentham1748–1832AFragmentonGovernment(1776);AnIntroductiontothePrinciplesofMoralsandLegis-lation(1789)MaryWollstonecraft1759–97AVindicationoftheRightsofMen(1790);VindicationoftheRightsofWomen(1792)FriedrichSchiller1759–1805LettersontheAestheticEducationofMankind(1794–5)JohannGottliebFichte1762–1814TheScienceofKnowledge(1794andlaterrevisededitions);TheScienceofEthicsasBasedontheScienceofKnowledge(1796);TheVocationofMan(1800);TheCharacteristicsofthePresentAge(1806)GeorgeWilhelmFriedrichHegel1770–1831TheJenaSystem1804–5:LogicandMetaphysics(1804–5);ThePhenomenologyofSpirit(1807);ScienceofLogic(1812–16);EncyclopaediaofthePhilosophicalSciences(1817);ThePhilosophyofRight(1821);Aesthetics(1835);LecturesonthePhilosophyofReligion(1832,1840);LecturesonthePhilosophyofWorldHistory(1837,1840);LecturesontheHistoryofPhilosophy(1892–6)FriedrichvonSchelling1775–1854IdeasforaPhilosophyofNature(1797);TheSystemofTranscendentalIdealism(1800);OntheEssenceofHumanFreedom(1809)896\nAPPENDIXBernardBolzano1781–1848Wissenschaftslehre,4vols(1837)WilliamHamilton1788–1856LecturesofMetaphysicsandLogic(1859–60)ArthurSchopenhauer1788–1860OntheFourfoldRootofthePrincipleofSufficientReason(1813);TheWorldasWillandIdea(1818)JohnAustin1790–1859TheProvinceofJurisprudenceDetermined(1832)AugusteComte1798–1857Coursdephilosophiepositive,6vols(1830–42)JohnStuartMill1806–73Bentham(1838);Coleridge(1840);SystemofLogic,2vols(1843);PrinciplesofPoliticalEconomy,2vols(1848);OnLiberty(1859);Utilitarianism(1861);anExaminationofSirWilliamHamilton’sPhilosophy(1865);SubjectionofWomen(1869);Autobiography(1873)CharlesDarwin1809–82TheOriginofSpeciesbyMeansofNaturalSelection(1859);TheDescentofMan(1871)SørenKierkegaard1813–55OntheConceptofIrony(1841);Either/Or(1843);FearandTrembling(1843);Philosoph-icalFragments(1844);TheConceptofDread(1844);StagesonLife’sWay;ConcludingUnscientificPostscript(1846);TheSicknessUntoDeath(1849)KarlMarx1818–83EconomicandPhilosophicalManuscripts(1844);TheCommunistManifesto(withEngels,1848);Grundrisse(1857–8);PrefacetoACritiqueofPoliticalEconomy(1859);Capital(1867–94)EduardHanslick1825–1904OntheMusicallyBeautiful(1854)WilhelmDilthey1833–1911IntroductiontotheHumanSciences(1883);FormationoftheHistoricalWorldintheHumanSciences(1910)FranzBrentano1838–1917PsychologyfromanEmpiricalStandpoint(1874),augmentedin2ndedn(1911)and3rdedn(1925);TheOriginofOurKnowledgeofRightandWrong(1889);TruthandEvidence(1930)HenrySidgwick1838–1900TheMethodsofEthics(1874)ErnstMach1838–1916TheAnalysisofSensations(1914)897\nAPPENDIXCharlesSandersPeirce1839–1914CollectedPapers(8vols,1931–58)WilliamJames1842–1910ThePrinciplesofPsychology,2vols(1890);TheVarietiesofReligiousExperience(1902);Pragmatism(1907);TheMeaningofTruth(1909);EssaysinRadicalEmpiricism(1912)FriedrichNietzsche1844–1900TheBirthofTragedy(1872);Daybreak(1881);TheGayScience(1882);ThusSpokeZarathustra(1883–5);BeyondGoodandEvil(1886);OntheGenealogyofMorals(1887);TheTwilightoftheIdols(1889);EcceHomo(1908)FrancisHerbertBradley1846–1924AppearanceandReality(1893)GottlobFrege1848–1925Begriffsschrift(Concept–Script)(1879);TheFoundationsofArithmetic(1884);TheBasicLawsofArithmetic(1893);‘TheThought’inStrawson,ed.,PhilosophicalLogic(1967);TranslationsfromthePhilosophicalWritingsofGottlobFrege(1980)FerdinanddeSaussure1857–1913CourseinGeneralLinguistics(1916)ÉmileDurkheim1858–1917TheRulesofSociologicalMethod(1895);Suicide(1897);TheElementaryFormsofReli-giousLife(1912)JohnDewey1859–1952ExperienceandNature(1925);TheQuestforCertainty(1929);ArtasExperience(1933);LiberalismandSocialAction(1935)EdmundHusserl1859–1938LogicalInvestigations,2vols(1900–1);Ideas:GeneralIntroductiontoPurePhenomenology(1913);ThePhenomenologyofInternalTimeConsciousness(1928);CartesianMedita-tions(1931);TheCrisisofEuropeanSciencesandTranscendentalPhenomenology(1954)DavidHilbert1862–1943FoundationsofGeometry(1899)MaxWeber1864–1920TheProtestantEthicandtheSpiritofCapitalism(1904–5);EconomyandSociety(1922)FerdinandCanningScottSchiller1864–1937Humanism:PhilosophicalEssays(1903)BenedettoCroce1866–1953AestheticsastheScienceofExpressionandGeneralLinguistics(1902)JohnMcTaggart1866–1925NatureofExistence,2vols(1921–7)898\nAPPENDIXBertrandRussell1872–1970PrinciplesofMathematics(1903);PrincipiaMathematica(withWhitehead,1910–13);ProblemsofPhilosophy(1912);TheTheoryofKnowledge(1913);OurKnowledgeoftheExternalWorld(1914);ThePhilosophyofLogicalAtomism(1918);IntroductiontoMathematicalPhilosophy(1919);TheAnalysisofMind(1921);TheAnalysisofMatter(1927);AnInquiryintoMeaningandTruth(1940);HumanKnowledge(1948)GeorgeEdwardMoore1873–1958PrincipiaEthics(1903);PhilosophicalStudies(1922);SomeMainProblemsofPhilosophy(1953);PhilosophicalPapers(1959)AlbertEinstein1879–1955TheMeaningofRelativity(1921);CollectedPapers,2vols(1987–9)LuitzenEgbertusJanBrouwer1881–1966CollectedWorks(1975–6)HansKelsen1881–1973GeneralTheoryofLawandState(1949);PrinciplesofInternationalLaw(1967);PureTheoryofLaw(1967);GeneralTheoryofNorms(1991)ClarenceIrvingLewis1883–1964AnAnalysisofKnowledgeandValuation(1946);TheGroundandNatureoftheRight(1955)KarlBarth1886–1968TheEpistlestotheRomans(1919);ChurchDogmatics(1932)RobinGeorgeCollingwood1889–1943ThePrinciplesofArt(1938);AnEssayonMetaphysics(1940);TheIdeaofHistory(1946)MartinHeidegger1889–1976BeingandTime(1927);TheBasicProblemsofPhenomenology(1927);KantandtheProblemofMetaphysics(1929);BasicQuestionsofPhilosophy:Selected‘Problems’ofLogic.’‘Lectures(1937–8);AnIntroductiontoMetaphysics(1954);WhatisCalledThink-ing?(1954);ThePrincipleofReason(1957);OntheWaytoLanguage(1959);Nietzsche,4vols(1961);Pathmarks(1967);OnTimeandBeing(1969);Poetry,Language,Thought(1971);TheQuestionConcerningTechnologyandOtherEssays(1977)LudwigWittgenstein1889–1951TractatusLogico-Philosophicus(1922);TheBlueandBrownBooks(1933–5);Philosophi-calInvestigations(1953);Notebooks1914–16(1961);RemarksontheFoundationsofMathematics(1966);Zettel(1967);OnCertainty(1969);PhilosophicalRemarks(1975);CultureandValue(1980)RudolphCarnap1891–1970TheLogicalStructureoftheWorldandPseudoproblemsinPhilosophy(1928);TheLogicalSyntaxofLanguage(1934);MeaningandNecessity(1947);LogicalFoundationsofProba-bility(1950)899\nAPPENDIXFriedrichAugustvonHayek1899–1992TheRoadtoSerfdom(1944);TheConstitutionofLiberty(1960);Law,LegislationandLiberty,3vols(1973–9)GilbertRyle1900–76TheConceptofMind(1949);Dilemmas(1954);CollectedPapers,2vols(1971)Hans-GeorgGadamer1900–2002TruthandMethod(1960)MichaelOakeshott1901–92ExperienceandItsModes(1933);RationalisminPolitics(1963);OnHumanConduct(1975);OnHistory(1983)FrankPlumptonRamsey1902–30TheFoundationsofMathematicsandOtherLogicalEssays(1931)LonFuller1902–78TheMoralityofLaw(1969)KarlPopper1902–94TheOpenSocietyandItsEnemies,2vols(1945);TheLogicofScientificDiscovery(1959);ConjecturesandRefutations(1963);ObjectiveKnowledge(1972)AlfredTarski1902–83Logic,Semantics,andMetamathematics(1956)TheodoreAdorno1903–69DialecticoftheEnlightenment(withHorkheimer,1947);NegativeDialectics(1966)AlonzoChurch1903–95IntroductiontoMathematicalLogic(1956)EmmanuelLevinas1905–96TotalityandInfinity(1961);OtherwisethanBeingorBeyondEssence(1974);EthicsandInfinity(1982)CarlHempel1905–97AspectsofScientificExplanation(1970)Jean-PaulSartre1905–80SketchforaTheoryoftheEmotions(1939);ThePsychologyofImagination(1940);BeingandNothingness(1943);CritiqueofDialecticalReason,2vols(1960,1985)KurtGödel1906–78CollectedWorks(1986–)NelsonGoodman1906–98TheStructureofAppearance(1951);Fact,FictionandForecast(1954);TheLanguagesofArt(1968);WaysofWorld-Making(1978)LeopoldSenghor1906–2001Liberté(1964)900\nAPPENDIXHerbertLionelAdolphusHart1907–92CausationintheLaw(withHonoré,1959);TheConceptofLaw(1961);EssaysonBentham:StudiesinJurisprudenceandPoliticalTheory(1982)SimonedeBeauvoir1908–86TheSecondSex(1949)ClaudeLévi-Strauss1908–StructuralAnthropology(1958);TheSavageMind(1962)MauriceMerleau-Ponty1908–61PhenomenologyofPerception(1945);TheAdventuresoftheDialectic(1955);Signs(1960);TheVisibleandtheInvisible(1964)WillardVanOrmanQuine1908–2000FromALogicalPointofView(1953);WordandObject(1960);SetTheoryanditsLogic(1963);TheWaysofParadox(1966);OntologicalRelativity(1969);TheRootsofRefer-ence(1974);TheoriesandThings(1981)IsaiahBerlin1909–97FourEssaysonLiberty(1969);VicoandHerder:TwoEssaysintheHistoryofIdeas(1976);AgainsttheCurrent(1980);TheCrookedTimberofHumanity:ChaptersintheHistoryofHumanIdeas(1991)AlfredJulesAyer1910–89Language,TruthandLogic(1936);FoundationsofEmpiricalKnowledge(1940);TheProblemofKnowledge(1954);TheCentralQuestionsofPhilosophy(1973)JohnLangshawAustin1911–60HowtoDoThingswithWords(1961);PhilosophicalPapers(1961);SenseandSensibilia(1962)AlanTuring1912–54CollectedWorksofA.M.Turing(1990)WilfredSellars1912–89Science,PerceptionandReality(1963);ScienceandMetaphysics:VariationsonKantianThemes(1968);NaturalismandOntology(1980)HerbertPaulGrice1913–90StudiesintheWaysofWords(1989)WilliamHerbertWalsh1913–96AnIntroductiontoPhilosophyofHistory(1967)MonroeBeardsley1915–85TheAestheticPointofView(1982)DonaldDavidson1917–EssaysonActionsandEvents(1980);InquiriesintoTruthandInterpretation(1984)LouisAlthusser1918–90ForMarx(1965)901\nAPPENDIXGertrudeElizabethMargaretAnscombe1919–2001Intention(1957);AnIntroductiontoWittgenstein’sTractatus(1959);CollectedPhilosoph-icalPapers,3vols(1981)RichardMervynHare1919–2002TheLanguageofMorals(1952);FreedomandReason(1963);MoralThinking(1981)PeterFrederickStrawson1919–IntroductiontoLogicalTheory(1952);Individuals(1959);TheBoundsofSense(1966);Logico-LinguisticPapers(1971);FreedomandResentment(1974);SubjectandPredicateinLogicandGrammar(1974);SkepticismandNaturalism:SomeVarieties(1985);AnalysisandMetaphysics(1992)PhilippaFoot1920–VirtuesandVices(1978)JohnJamiesonCarswellSmart1920–PhilosophyandScientificRealism(1963);EssaysMetaphysicalandMoral(1987);OurPlaceintheUniverse(1989)WilliamAlston1921–DivineNatureandHumanLanguage(1989);EpistemicJustification(1989);PerceivingGod(1991)WilliamDray1921–LawsandExplanationinHistory(1957);PerspectivesonHistory(1980)JohnRawls1921–ATheoryofJustice(1971);PoliticalLiberalism(1993);TheLawofPeoples(1999);Col-lectedPapers(1999)ThomasKuhn1922–97TheCopernicanRevolution(1957);TheStructureofScientificRevolutions(1962);TheEssentialTension(1977)JohnHick1922–FaithandKnowledge(1957);EvilandtheGodofLove(1966);PhilosophyofReligion(1966)PaulFeyerabend1924–94AgainstMethod(1974);PhilosophicalPapers,2vols(1981)ArthurDanto1924–AnalyticPhilosophyofHistory(1965);AnalyticPhilosophyofAction(1973);TheTrans-figurationoftheCommonplace(1981)MichaelDummett1925–Frege:PhilosophyofLanguage(1973);TruthandOtherEnigmas(1978);TheLogicalBasisofMetaphysics(1991);OriginsofAnalyticPhilosophy(1993)MichelFoucault1926–84MadnessandCivilisation(1961);TheOrderofThings:AnArchaeologyoftheHumanSci-ences(1966);TheArchaeologyofKnowledge(1969);DisciplineandPunish(1975);TheHistoryofSexualityI–III(1976–84)902\nAPPENDIXDavidArmstrong1926–AMaterialistTheoryofMind(1968);UniversalsandScientificRealism(1978)HilaryPutnam1926–Mathematics,MatterandMethod(1975);Mind,LanguageandReality(1975);RealismandReason(1983);RepresentationandReality(1988)NoamChomsky1928–SyntacticStructures(1957);CartesianLinguistics(1966);KnowledgeofLanguage(1986);DeterringDemocracy(1992);LanguageandThought(1993),TheMinimalistProgram(1995)JürgenHabermas1929–KnowledgeandHumanInterests(1968);TheTheoryofCommunicativeAction,2vols(1981);ThePhilosophicalDiscourseofModernity(1985)AlasdairMacIntyre1929–AShortHistoryofEthics(1966);AfterVirtue(1981);WhoseJustice?WhichRationality?(1988);ThreeRivalVersionsofMoralEnquiry(1990)BernardWilliams1929–ProblemsoftheSelf(1973);Descartes:TheProjectofPureEnquiry(1978);MoralLuck(1981);EthicsandtheLimitsofPhilosophy(1985);ShameandNecessity(1993);MakingSenseofHumanity(1995)JacquesDerrida1930–SpeechandPhenomena(1967);OfGrammatology(1967);WritingandDifference(1967);MarginsofPhilosophy(1972);Positions(1972);TheGiftofDeath(1995)RonaldDworkin1931–TakingRightsSeriously(1977);AMatterofPrinciple(1985);Law’sEmpire(1986)RichardRorty1931–PhilosophyandtheMirrorofNature(1979);ConsequencesofPragmatism(1982);Con-tingency,Irony,andSolidarity(1989);Objectivity,Relativism,andTruth(1991)CharlesTaylor1931–TheExplanationofBehaviour(1964);Hegel(1975);PhilosophicalPapers,2vols(1985);SourcesoftheSelf(1989);TheEthicsofAuthenticity(1991);MulticulturalismandthePol-iticsofRecognition(1992)LuceIrigaray1932–SpeculumoftheOtherWoman(1974);Je,Tu,Nous:TowardsaCultureofDifference(1990)AlvinPlantinga1932–GodandOtherMinds(1967);TheNatureofNecessity(1974)JohnSearle1932–SpeechActs(1969);Intentionality(1983);TheRediscoveryoftheMind(1992);TheCon-structionofSocialReality(1995)RichardWollheim1932–ArtandItsObjects(1968);TheThreadofLife(1984);TheMindanditsDepths(1993)903\nAPPENDIXJerryFodor1935–TheLanguageofThought(1975);TheModularityofMind(1983);Psychosemantics(1987);ATheoryofContentandOtherEssays(1990);Holism:AShopper’sGuide(withLapore,1992)MichaelWalzer1935–SpheresofJustice(1983)ThomasNagel1937–ThePossibilityofAltruism(1970);MortalQuestions(1979);TheViewfromNowhere(1986)RobertNozick1938–2002Anarchy,StateandUtopia(1974);PhilosophicalExplanations(1981);TheNatureofRatio-nality(1993);Invariance:TheStructureoftheObjectiveWorld(2001)JosephRaz1939–PracticalReasonandNorms(1975);TheAuthorityofLaw(1979);TheConceptofaLegalSystem(1980);TheMoralityofFreedom(1986)SaulKripke1940–NamingandNecessity(1980);WittgensteinonRulesandPrivateLanguage(1982)DavidLewis1941–2002Convention(1969);Counterfactuals(1973);PhilosophicalPapers(1983–7);OnthePlu-ralityofWorlds(1986)GeraldAlanCohen1941–KarlMarx’sTheoryofHistory:ADefense(1978);History,LabourandFreedom:ThemesfromMarx(1988)DanielDennett1942–ContentandConsciousness(1969);Brainstorms(1978);TheIntentionalStance(1987);ConsciousnessExplained(1991)GarethEvans1946–80TheVarietiesofReference(1982);CollectedPapers(1985)PeterSinger1946–AnimalLiberation(1975);PracticalEthics(1979);RethinkingLifeandDeath(1995)MicheleleDoeuff1948–Hipparchia’sChoice:EssaysConcerningWomenandPhilosophy(1990)904\nIndexaprioriknowledge602,685,728,729,socialaction397731,736,766three-levelconception536,537Abelard,Peter620,622,628–9Acton,H.B.752abortion508Adams,R.M.theabsolute65,743,744bestpossibleworld482absoluteidea744–5theisticvoluntarism469absoluteknowledge743adaptionism324–5,325,336absolutemonarchy690cognitive336–7absolutepresuppositions67groupadaptations325abstractobjectsindividualadaptations325Frege792,793seealsonaturalselectionHobbes’sdenialof677–8addiction207philosophyofmathematics346,363,adjectives91368adjudication409–10,421–3,424abstraction561,562,563admiration,ethicsof204,205,223anti-abstractionism694,695–6,702adverbs113Academics597andevents114Ackerman,Bruce281manneradverbs91actionmodaladverbs12,91analyticactiontheory20aestheticjudgement234,236,717ethicsand203,213Humeon234,235events439Kanton235freedomofchoice534–5andpleasure233Humeon713rules232Ideal-typicalrationalreconstructionWittgensteinon236391aesthetics4,231–56intentionality8,436aestheticattitude236maximsof218,734–5,736aestheticexperienceandjudgementmoralactions734,735232–7,717–18motives259,392,713aestheticobjectivism232–3philosophyof8aestheticpreferences234politicalactivism279aestheticqualities233,234,235–6rational390aestheticreceptivity231reasonsfor203,208,213,216,217,aestheticsubjectivism234,235–6,250388–9art237–52905\nINDEXaestheticscont.metaphysicsand69–75conceptofbeauty236–7Moore808–9discussionquestions255–6origins2,23Hume232,233,234,247post-positivistphilosophyofscienceKant232,233,234,23510–11Mill764,766Rawls’stheoryofjustice9–10,31–2,taste232,234,235,718492aetiology184recentdevelopments11–20Africanphilosophy568,571–3,575rejectionoffoundationalism2,3,6,8–9,African-Americanphilosophy576,57740,52,55,367–8,472afterlife483andscience24,25agapism776self-critical2agnosticism456speechacttheory6,7–8AIseeartificialintelligencetruth-conditionaltheoriesofmeaningakrasia605–612,16–18AlberttheGreat621truthfulness24,34Albo,Joseph463virtuesof24,34alchemy193,635weakness14alienationanalytictechnique69alienatedlabour751analyticalbehaviourism175economic751analyticalMarxism273–4,278,751–3Marxon750,751analyticalpriority96,97political751anarchy,Nozickand261–4religious751Anaximander569Alston,William462ancestralinfluenceseephylogeneticinertiaAlthusser,Louis,structuralistMarxism751anger610–11altruism326,522,763animalsambiguity164animalexperimentation514analysis81animalrightsethic520–1,522flightfrom71–2mechanistviewof650andlogic69–75annalisteschool444analyticempiricism346,357–63anomalousmonism192analyticgeometry353,356Anscombe,Elizabeth329,492analyticparadigm71,72Anselm620analyticphilosophy1–21,23–5,33–4divineomniscience466achievements24existenceofGod625actiontheory20ineffabilityofGod462analysisofmeaning2logic622–3analytic–syntheticdistinction2,3,4,anthropocentrism518,519,5266–7,11anthropology437–8causaltheoryofreference12,14–15,anti-abstractionism,Berkeleyon694,21,469695–6,702cognitivescience12–14,24anti-realism28–9,75,165,166,430,444,descriptive–evaluativedistinction2,4–6,445,4467,9Dummett165development2–3historicalnarratives444–8identityof23–5scientific294–5intentionalistictheoriesofmeaning12,anti-Semitism57715–16AntinomiesofPureReason731linguisticturn359Apel,K.-O.785906\nINDEXaporia592philosophicalpractice593appliedethics403,490–8,499,532phronesis613businessethicsas538–40polis614description490–1politicaltheory614discussionquestion498onpsychicphenomena610futureof495–6rationalism608–9metaphilosophicalaccountofmoralthesoul610–11reasoning495teleologicalphilosophy593–4moralcasuistry494–5andTheoryofForms608,609moraldanger493–4timeandtruthphilosophy82,467riseof491–4universals608scopeof490–1,499virtuetheory27Aquinas,Thomas619,621arithmetic353,354,356,361casuistry494Armstrong,D.M.divinenames626behaviourism176divineomniscience466emotions188anddivineownership470identitytheory177existenceofGod330,458lawsandaccidents306FiveWays625Arno,AntonWilhelm572goodnessofGod469art231–2,237–52Greekphilosophy571analyticphilosophyof231,237,242individuation630artisticform239knowledge623,627–8avant-gardeart238metaphysics627classificatorysense238moralreasoning494criticism717–18universals629definitions237–8Archard,David,politicalandsocialDewey252philosophy257–85dimensions242–50Aristotle82,318,333,352,525,586,andemotions241,244–7592–4,595,608–14essenceof237–52akrasia606ethicsand203,206onanger610–11evaluativesense238aporia592expressiontheory240–1,248dialectic592,593expressivequalities244–7empiricism608formalism239,240,241,248endoxa592,594,595Hegel237essentialism333Heidegger852,854ethicaltheory336,611–13Hume232–7,251–2,717–18eudaimonia611,612,751illusiontheory243–4Kantianreading593,594,597institutionaltheory237onknowledge628intentionalfallacy248logic586,621interpretation247–8,249,250,718materialism610,611literaryworks34,248–50mathematicalproofs352meaning247–50matterandform609–10,665metaphysicalaccountsof252andmedievalphilosophy620mimetictheory238–9,241,242,248,metaphysics608,609–10,626250methodologicalself-consciousness593music244–7mimesis238naturalisticaccountof252phainomena592,594pictorialrepresentation239,242–4907\nINDEXartcont.authority33Plato251argumentsfrom624pointof250–2metaphysical81Schiller251tacitandexplicitconsent691‘seeingin’244autonomy764semiotictheory241,243ofgood223theoriesof238–41individual407Tolstoy251Kanton219,224,733–5transfigurativeaspect252lawandpolitics407valueof239,250–2andliberalism269Wittgenstein237patientautonomy511,512worksofart237–8,245,251,853,854private407seealsoaestheticsrational219artificialintelligence(AI)12,13,180–2seealsofreewillcomputermodelofmind180–2Avenarius,Richard848connectionism184,193–4AverroesofCordoba(IbnRushd)619,623,grades180–1627neuralmodelling184,193AvicennaofBaghdad(IbnSina)619,ParallelDistributedProcessing(PDP)193626–7standardtasks181axiomatics585strongAI13,180–1,182axiomatizationofgeometry365–6,814asocialconceptionofthehumansubjectaxioms562–3AxiomofChoice366assertibilityconditions78AxiomofInfinity362assertives8AxiomofReducibility362assetoricsentences157,793,794semantics17,92–6,99–102,103,107,astrology288108,110,123atheism332,456,459,460,462,472,Zermelo–Fraenkelaxiom366,367475,480,747Ayer,A.J.51,454,455burdenofproof472Sartre863Bachelard,G.369atomism648–9,664,686,687Bacon,Francis634–43corpuscularianism686,687colour639logical791,814,815,818discussionquestions643perceptualatomism73DoctrineofIdols636–8AttributeofExtension660eliminativeinduction639AttributeofThought660knowledge636AugustineofHippo572,620,622,626,naturalphilosophy635,636,637668rhetoricallymotivatedaccountofdiscoverydivineomniscience466635–6DoctrineofIllumination627–8scientificmethod635–40eternalGod467–8systematizationoflaw635faithandreason622truth640–1onknowledge627Bacon,Roger621theodicy482badfaith862,863Austin,J.L.Bain,Alexander776legalpositivism416Baldwin,Thomas,Moore805–10onsovereignty411Bantuphilosophy572,574theoryoflaw411,416Barnes,J.462theoryofspeechacts6,7–8Barth,Karl472908\nINDEXBayes,Thomas290religiousbeliefs779Bayesianism290–3,308self-justifying40,41Bayesianupdating293semanticalinterpretation188,195Bayes’sTheorem293settledbelief777,779DutchBookArgument293,308tracking44Beardsley,M.C.236WilltoBelieve779attackonintentionalfallacy248,249seealsofaith;foundationalism;probabilityBeatty,L.329belief-ascriptions191,192Beaty,M.455,461Bell,C.239,251Beauchamp,Thomas,FourPrinciplesBell,Daniel257–8approach(4PA)500Bell,David,Kant725–40beauty236–7Bell’stheorem11aestheticsubjectivism233Bennett,J.120,123Kanton235Bennett,Rebecca,bioethics,genethicsandBechtel,W.194medicalethics499–516BedeutungseereferenceBentham,Jeremy759–61,768,770behaviourasaction386casuistry494behaviourism5,6,13,69,174–6onclarification768–9analyticalbehaviourism175criticismofBlackstone760eliminativebehaviourism175discussionquestions772–3instrumentalismand191egoisticprinciple770logicalbehaviourism12fictionalentities768metaphysicalform175happinessandpain759,760,761,769methodologicalform175indirectutilitarian766philosophicalbehaviourism176legalpositivism415,502psychologicaltheorizing175Millon763TuringTest175,176,181naturallawtheory415–16,769Beingonobediencetothestate760Being-for-itself861,862paraphrasis768–9Being-in-itself861,862sanctions764,765Being-in-the-world849,850,851theoryoflaw416,425,760,761,766,andessence627768–9HegelianDoctrineofBeing744theoryofpunishment425,760,761,scienceof62764,769seealsoDasein;existenceofGodutilitarianism391–2,415,416,425,belief96502,759–61,768–9behaviouristicapproach785Berger,F.R.763coherence41–2Berkeley,George2,48,694–708contentsensesof785abstractideas,relativeideasandendoxa592,594,595immaterialism696–700externalism43,44anti-abstractionism695–6,702ashabitofaction776beliefinGod53,458indefeasibility40Britishempiricism684intensionality109causation706internalism42commonsenseandthe‘newphilosophy’justifiedbelief37–8,39–41,42,44,56,694–6,706217conceivability,perceivabilityandintrinsicandknowledge685properties702–4normsofrightbelief203,214conceptionsoftheself704–5reliabilism43discussionquestions707–8909\nINDEXBerkeley,Georgecont.biologicalidentity523esseestpercipi695definitionoflife317–18generalitytheory697,698discussionquestions343–4idealism53–4,694–6fitness319,337imagism698mechanism646immaterialism53,694,696,698,700,ModernSynthesis326705reductionism319theinfinite355superveniencethesis80–1,318intentionality703teleology184–5,319–20onLocke696–7vitalism318masterargument702–3seealsoevolutionarytheorymind-independentmatter694,695,699,bioregionalism525–6704bivalence73particularistontology696–7,698BlackDeath621perception53–4,694–6,700–2Black-letterrules404–5,410,421,422phenomenalism700–6Blackburn,S.120philosophyofreligion458metaphysics61–89qualities,ideasandsensations700–2Blackstone,SirWilliamrefutationofscepticism53–4Bentham’scriticismof760revisionarymetaphysics61naturallawtheory415–16,422selectiveattention697blame204–5,208,214,215,223theism704,705Block,N.J.176,177,185Berlin,IsaiahBloomsburyGroup,Moore’sinfluenceonliberty258806politicalphilosophy258Bloor,D.368Bernasconi,Robert,ethnicity,cultureandBoethius628philosophy567–81divineomniscience466Bernstein,R.J.785evil465–6Bestand854onindividuation630Bhagavad-Gita571Bolzano,Bernard844Bigelow,J.185Bonaventure621,623,627binarismBoolos,G.S.153feminismand555Boorse,C.185,320metaphysical555–6,869Bostock,D.362biocentrism521–2Boyd,R.294biodiversity523Boyle,Robert682,687bioethics499–504,532Bradley,FrancisHerbert73discussionquestions515–16Braidotti,Rosi561FourPrinciplesapproach(4PA)500–1Brentano,Franz184,187,843historyof503BritishEmpiricism684intuitionism500Brouwer,L.E.legalregulation502intuitionism364methodology499reflexivity369moralphilosophicalenquiry499Brown,Lesley,PlatoandAristotle601–18pluralism500,501Brucker,Jacob569scope503–4Brunschwig,J.596theoreticalapproaches499–503Buchanan,A.275,277seealsogenethics;medicalethicsBudd,M.247biology,philosophyof12,301,312,Buddhism457,460317–44,573Burkert,W.571910\nINDEXBurnyeat,M.F.590exploitation273businessethics531–51inequity274action-orientedapproach534–8juridicalframework755asappliedethics532,538–40ownershipofresources273challenges531care,ethicsof272,273cognitiveandnormativedimensions539,andjustice273542Carman,Taylor,HusserlandHeideggerconceptualframework532–40842–59incontinentalEurope542,543Carnap,R.174–5,819corporateethics534,536,541,545–6carnivores335corruption541Cartesianco-ordinates648decision-making536,542,544,546Cartesiandualism173,174,318,605defining532–4non-materialistalternativeto660–1discussionquestions550–1objectionsto174,176economicsystem544–5Cartwright,Nancy295emergingfield540–2casuistry494–5foundationalissues542–4moral494globalization531,538,541,546scholasticcasuists494international537–8,541CategoricalImperative27,31,500,734,Japanese540735–6leadership541andcontractualism218–22macro-level542,544,545categories62meso-level542,545Hegel742,743–4micro-level542,545Kant730,742overview540Peirce’stheoryof776productiveinterdisciplinarystudyofBeing62communication542–4causallaws434,435,437publiceconomics545causalmechanisms383responsibility535,536causalpropositions390semantics533,534,540causalrelations396,711,712,716,720stakeholdertheory541,545–6mentalstatesand13theoreticalapproaches542–6objective15three-levelconceptionofaction536,538causaltheoryofreference12,14–15,21,two-leggedapproach539–40469intheUSA542,543causation638varietyofbusinessenvironments540–1asymmetryofoverdetermination306–7Butler,Joseph458Berkeley706causalinference711,712Calvin,John472causalprinciple711Cambridgechange471constantconjunctionaccount303,304,Campbell,D.720behaviourism176coveringlawmodelofexplanation307,evolutionarytheory337381,382,434,435,436,437,439Canovan,M.276Descartes654Cantor,G.,mathematics358,360,366–7directionof306–7capitalism258,264,273–4,278doctrineof‘occasionalcauses’655,666alienatedlabour750,751Hume303,304,306,711–12,721analyticalMarxismon273–4,278,intentional21751–3lawsandaccidents305–6coercion273mental21911\nINDEXcausationcont.citizenship260,261,614metaphysicsofscience303–5rightsandresponsibilities528necessity711civilsociety,Lockeon690–1probabilistic307–8Clarke,M.S.188rationalreconstruction111,386,436,Clarke,Thomson20437,438,439classinterests755Cavalli-Sforza,L.322classstruggle383–4,753certaintyclassesDescartes50,51,652,776logicof360Dewey781mathematical360–3questfor50,51,781,785metaphysical63–4,65Chalmers,D.187no-classtheory362change82–3Russell’sparadox361,791awarenessof82cloning505,506,507first-orderchange83co-ordinationgame393second-orderchange83coexistence32chaostheory347cognition14,742–3charactertraits206,207,326cognitiveeffects133,134characterizationargument442–3cognitivepsychology179–80Charlemagne620cognitivescience12–14,24CharlesI676ideology12chauvinism570interdisciplinarycharacter12chemicalelements333,335neuralnetmodels14Chihara,C.362Cohen,G.A.,analyticalMarxism751–2,Childress,James,FourPrinciplesapproach753–4(4PA)500coherence41–2Chinesephilosophy569,571,574–5,578,collectiveagency274663colligatoryconcepts447,449,450Chisholm,R.M.176Collingwood,R.G.67representationalism189absolutepresuppositions67Chodorow,Nancy554expressiontheoryofart240–1Chomsky,Noamhistoricalexplanations436,438functionalism180colonialism537,572languagephilosophy111,123colour28ChristiantheologyBacon639casuistry494Descartes647criterionoftruth622commissives8evil480–3commonsense3,4,70–1andmathematicalknowledge350,Berkeley694–6355folkpsychology177,192,193,194–6andsecularlearning622Moore’sdefenceof807–8,809seealsotheismPeirce’spragmatism777Christianity457,460,480World-view648Churchland,P.S.andP.M.commonwealth674,676,678eliminativism193communitarianism32,33,270–2,521folkpsychology193critiqueofliberalism270–2pragmatism784normativeclaims270–1representationalism189–9communitytheoryoflanguage822Church’stheorem153compatibilism389,398,467Cicero710computerfunctionalism13,14,15912\nINDEXcomputermodelofmind180–2conservatism277,279computerscience12conservativeholism213,214,222seealsoartificialintelligenceconsistency41,147Comte,August721Hilbert’sProgramme365,366conceivability,Berkeleyon702–3mathematical365–6,367conceptsconstatives7abstract352constructionism430–1Berkeley696–7,702–4constructivism73,355colligatory447,449,450contextualimplications133ethical29continentalphilosophy23,562,573,631Frege790,792contingency112,113Hegel742,744contractualism31,213,220–2,261Nietzsche830–1agent-neutralprincipleoffairness222andobject791–2CategoricalImperative218–22conceptualanalysis4,5,11,258Kantian221–2conceptualpolarities49,50moralcategoricity216,221,222,224conceptualism355Rawlson261,267Aristotelianconceptualism351–2,353conventionT17,105Kant355–7conventionalregularity115,117,120–2mathematics351–2,353,355–7,368–9conventionalism5,420,823Condorcet,Marquisde,naturalsciencesAristotelian355377,378legal420,422confirmationrelations167,168conversationConfucianism460,571,578Co-operativePrinciple129–30,131connectionism184,193–4conversationalco-operation15–16connectionistartificialintelligence184,maximsofManner130193–4maximsofQuality131,132conscience,Hegelon745maximsofQuantity130,131consciousness63,185,611,743conversationalimplicatures128–30Hegel743,747cancellabilityofimplications129Hume731conventionalimplicatures129Husserl844–5,847,848detachabilityofimplications129intentionality860Grice’stheory128–30Kant732–3irony132,134,135Leibniz668metaphor16,132,134,135linguisticknowledge111open-endedcharacter131Locke689relevancetheory132–5objectsof861Cooper,DavidE.,Nietzsche827–41perceptual844–5CopernicanRevolution727–8problemsof185–7Cope’sRule328andreality,discontinuitybetween845corporateethics534,536,541,545–6Sartre860–2corpuscularianism686,687seealsomentalstates;mind,philosophyof;correspondencetheoryoftruth226,perception;self-consciousness457–8consent,tacitandexplicit691corruption541consequentialism30,31,147,163,501,cosmicimpiety783510cosmology301,312bioethical500cosmopolitanism276,574,577historical442cosmosseealsoutilitarianismanalyticgeometryandtheinfinite353–7913\nINDEXcosmoscont.Davies,Martin,philosophyoflanguageAristotelianconceptualism351–2,35390–146Divineownership470Dawkins,Richard,genicselectionism326existenceofGod53,79,163,223,330,deBeauvoir,Simone863353,458,459,462DeGeorge,R.T.536explaining477deathFirstCause476,477,478afterlife483infinities477brain-stemdeath510Platonicrealism349–50,353,363definitionof510uniqueness479deathdrives554coveringlawmodelofexplanationdecisiontheory389deterministicassumptions436deconstructioninductive–statisticalversion307Derridean868philosophyofhistory433–9literaryinterpretation250philosophyofscience304–5,307,311metaphysicalbinarisms869,871philosophyofsocialscience381,382Dedekind,R.,numberstheory360Craig,W.477deductiveargument287,348,766–7criminology384,866deductiveinferencesCriticalLegalStudiesmovement(CLS)423mathematics348Croce,B.,expressiontheoryofart240science287cruelty207,469deductive–nomologicaltheoryseecoveringculturalmembership275lawmodelofexplanationculturalpluralism275deepecology526culturalunderstanding29definitedescriptions,Russell’stheoryof69,culture,philosophyof568798–9seealsoethnicphilosophydelusion46,47,49Cummins,R.democracy258,570biologicalteleology185,319,320Democritus686machinefunctionalism183demonstrativereasoning623demonstratives91,112DALYs(DisabilityAdjustedLifeYears)511Dennett,D.C.Damian,Peter622instrumentalism191–2Danto,A.C.machinefunctionalism183knowledgeofhistoricalfact429moralresponsibility182narrativesentences440representationalism189Darwin,Charles,evolutionarytheory321,deontology30,500325–6,330,331–2,334,336,338,Derrida,Jacques868–71363binarisms555–6,869Dasein850–1,852deconstruction555–6,869,871dataprotection495,513différance870Davidson,Donalddiscussionquestions873–4anomalousmonism192ethics871moodsandspeechacts137andfeminism556semantictheories97,102–4,106,110,justice871113,114,115,138poststructuralism864Tarski–Davidsontruththeories110,thoughtandlanguage868–9126Descartes,René37,46,620,644–52,864Truth-conditionaltheoriesofmeaningaccountofGodandtheworld648–50,16–18653,654,658Davies,Brian,philosophyofGod462,480accountofmentalphenomena660914\nINDEXCartesiancogito50,377,387,651,652,economicdeterminism281,379,752,776754Cartesiandualism173,174,176,605,andfreewill380,467650–1Hume721causation654social389oncolour647Dewey,John459,781–2discussionquestions657,670art252anddreaming47knowledge782epistemology644,651–2,776logic782‘evildemon’idea50,51pragmatism775,781–2,784,785explanationofhumancapacities650scepticism55fatherofthemodernview620,645truth781freewill661dialecticgeometricalphysics378Aristotle592,593God50,353,464,471,479Hegel746humanmind377,644,660Hellenistic594–7humannature650–2Kant730immaterialism174,176,677–8medieval622individuation649Socrates586–90knowledgeofselfandGod655Wittgenstein817–20lawsofmotion667dialogueethics546masculinepsychicstructure554–5dichotomies,feminism555–6mechanism646–7,650,664Dickie,G.237mind–bodyrelations173,651,655,667différance870modernism645,660differenceprinciple211,260–1,263,264,natureoftheworld645–8,664265,271ontology647feminismand560,561perception646–8,653Nozick264physicaltheory646Rawls211,260–1,263,265,271properties647,648,653Dilthey,Wilhelm397,850questforcertainty50,51,776meaning384revisionarymetaphysics61Diop,CheikhAnta571scepticism46,50–2,471,652directives8onthesoul667–8disagreementdescriptionsethical29causaltheoryofreference12,14–15,moral2921,469discourseethics546definitedescriptions69,112,798–9,indirectdiscourse799–801813disinterestedness236functionalistlevelsof179,183–4dispositionsRussell’stheoryof69,798–9,813ofindividuals435descriptivemetaphysics61,62andpowers72,702–3descriptivephrases90,798–9divineattributes460–71,659descriptiveutterances4–6,9eternity467–8,659desire191,205–6goodness460,468–70,479,480desire-ascriptions191,192immanence471,480,659despotism663impassability471determinatemeaning156,250ineffability462–3determinism208,307,379,399,595,modussignificandi461713,720omnipotence463–6,482,659915\nINDEXdivineattributescont.legalobligations405–6omniscience461,466–7,468,667Locke690ressignificata461andrights769simplicity471utilityprinciple768transcendence470Dworkin,Ronalddivinelaw224,491onadjudication421divinerightofkings690,691critiqueoflegalpositivism502–3DNA317liberaltheory32,266DoctrineofIdols636–8theoryoflaw410,413,418,419–20,dogmatism250421–2Donagan,A.434Donnellan,K.14Earthethic526–8doubt776–7ecofeminism520,526Dray,W.H.econometrichistorians444historicalexplanation436economicdeterminism281,379,752,754historicalnarratives446economicsvaluejudgements440,441,442andbusinessethics544–5dreaming46,47,49,50,651economicanalysisoflaw425Dretske,F.184Hume719–20,721Dreyfus,H.L.,computermodelofthemindMarxisteconomicdeterminism281,379,181752,754dualism12,462,611,837neo-classical384Cartesiandualism173,174,176,318,andthepoliticalsphere281605,660–1publiceconomics545intelligibilityof462ecosystemmanagement525neo-dualism187Edgeworth,F.Y.391,392Duhem,Pierre7Edwards,A.322Duhem–Quinethesis7,298,299,300egalitarianism265–6,268,274Dummett,Michaelcapabilityegalitarianism266anti-realism165criticismofpracticeandprinciple266onFrege791,800opportunityegalitarianism266andimplicitknowledge111resourceegalitarianism265,266mathematics364welfareegalitarianism265,266metaphysics72–4egoismsemantictheories97,99,111Bentham770DunsScotus,John621Mill763,770divinenames626rationalegoism217,220,224existenceofGod625Sidgwick770individuation630Egyptianlearning569knowledge628Einstein,Albertmetaphysics627non-Euclideangeometry358universals629quantummechanics349Durkheim,Emilerelativitytheory290,350,358,704crime384Ekman,P.188religiousexperience474elenchos587–90,606socialexplanation384eliminativebehaviourism175DutchBookArgument293,308eliminativeinduction639dutyeliminativism64,70,71,192–3,195,462dutyandinterestprinciple761andfunctionalism192–3Hobbes674–6Elizabethanphilosophy634–5916\nINDEXElster,Jon273animalrightsethic520–1,522functionalexplanation753anthropocentrism518,519,526embryoresearch505axiologicalethics524,527emotionsbiocentrism521–2aestheticsand241,244–7biodiversity523ethicsand204–5,215bioregionalism525–6andfunctionalism188deepecology526intentionality188,245discussionquestions529–30logicalform247Earthethic526–8materialism610ecofeminism520,526mentalstates188environmentalhealth519emotivism27,234environmentalquality518empathy386ethicsofplace525empiricalpsychology44holisticethic524empiricism2,3,38,46,380,382,454humansandtheworld519–20analyticempiricism346,357–63inclusiveness518,522,525Aristotelian608landethic524–6BritishEmpiricism684riseof517–18Dewey782species,responsibilitiesto523andfalsification10sustainabledevelopment526–7,528Hume66,379,380,684,705–6,712,Epicureanism596714–15Epicurus686James780epistemologicalpriority96,97Kantian726–33,736epistemology37–60andknowledgeoftheworld46Cartesian644,651–2,776Locke684definition37–8andmathematics346,357–62discussionquestions59–60Mill684,766–7evolutionaryepistemology336–8,Moore’scriticismof805479–80problemsfor47–8naturalizedepistemology302–3,363,radical780563andreligion454–5,471–4,714–15ReformedEpistemology472andscience38,302scepticalepistemology45–56sensedata5,54,130,361,807,818semantictheoriesand109–11andsocialscience378,379,380,382seealsobelief;knowledgeend-statetheories263equality263,265–6Enderle,G.,businessethics531–51analyticalMarxism274endoxa592,594,595ofcitizenship260EnglishCivilWar672,676gender265Enlightenment64,377legaldiscourseof420,421Locke682–3liberalequality266rationalism682–3andliberty258,266andscience396Nozick262,263worldview527ofopportunity281entities191,850,851Rawls211,261,263,265assignmentof102Ereshefsky,M.335meanings91Erin,CharlesA.,bioethics,genethicsandentitlementprinciple262,263medicalethics499–516environmentaldegradation525,528,539Eriugena,John620,628environmentalethics517–30eroteticlogic164917\nINDEXerror47,49professional203vulnerabilityto46reason,moralityandreligion222–5errorsceptic50reasonsforaction203esseestpercipi695scope202–4essence333,334,336,424,686–7,744,secular770846utilitarianismseeutilitarianismbeingand627ethnicphilosophy25–34,567–81formand609Africanphilosophy568,571–3,575HegelianDoctrineofEssence744culturalidentity573ofnaturalkinds333discussionquestions580–1essentialism112,113,333–4interculturaldialogue568,574–5Aristotelian333interculturalism573–6biological574lexicon575chemicalelements334–5multiculturalism567,568,574,575liberalrejectionof424Euclideangeometry212,348,365ethicalconceptsaxiomatizationof365–6,814‘thick’29inferenceprinciples348‘thin’29Schopenhauerand814ethicalsubjectivism338–9eudaimonia611,612,751ethicaltheory30–1,212–14euthanasiaaim30involuntary509consequentialist30non-voluntary509–10deontological30TonyBlandcase509–10virtuetheories27,30,31,500,546voluntary508ethicalvalue203,209–10,211,212,213,evaluativeutterances4–6,9218,734Evans,G.111,795ethics4,26,80,81,202–30events,metaphysicaltheoriesof114appliedethics403,490–8evilAristotelian611–13andtheafterlife483ofbelief455argumentfrom332blame,admirationanddesire204–6FreeWillDefence481,482ofcare272Godand465–6,469CategoricalImperativeandcontractualismGreaterGoodDefence481,482218–22Irenaeanapproach482discussionquestions229–30problemsof480–3emotions204–5,215theisticresponses481evolutionaryethics338–9evolutionarycontingencythesis329femaleethics272evolutionaryepistemology336–8andfunctionalism182culturalandscientificchange336,337idealobservertheory470evolutionarypsychology336–7impartiality211–12,221humancognition336,337intrinsicethicalvalue209–10sociobiology336Kantian500,733–6evolutionarypluralism324,325andlaw502,503evolutionarypsychology336–7meta-ethics25–9,30,203,225–7,491,evolutionarytheory320–39,397–8492,503adaptionism324–5,336normative532,539convergence330objectivity26,27Cope’sRule328practicalreason216–18Darwinianaccount321,325–6,331–2,private764334,336,338,363918\nINDEXdesignhypothesis330–3naturalistic478empiricalgeneralizations329purposive478ethics338–9socialscience376,377–84,395,396evolutionarycontingencythesis329andunderstanding390–7evolutionaryepistemology336–8,explanatoryintegration785479–80expressiontheoryofart240–1,248evolutionarypluralism324,325expressives8lawsofevolution328–30externalobjects713,726–33,767mutation337externalism43,44philosophicalsignificance330–9phylogeneticinertia324fact–valuedistinction26phylogeneticinference322–4faith355phylogeneticparsimony322,323–4Augustine622andreligiousdoctrines330–3badfaith862,863sensitivitythesis329–30Kant736species333,334,335andreason353,622–3unitsofselectioncontroversy325–6,fallibilism39,40,433,775,777,778,785326,327falsificationism10seealsonaturalselectionfailingsof289–90ExcludedMiddle,Principleof364Popper287–90,298,299,323,324existenceofGod53,79,163,223,330,thefamiliarandtheexotic69,70,77,80353,458,459,462family280–1Aquinas330,458assettransfer281argumentfromdesign330–3,715–16,communitarianism270–1720familialrelations280argumentfrommiracles483feminismand272cosmologicalarguments476–8,625Hegel746medievalarguments624–5Locke690ontologicalarguments474–6,625public–privatedivide280problemofevil480–3sexualbehaviour280teleologicalarguments478–80Farrell,B.A.,functionalism185wagerargument483fascism258,783existentialism1,645federalism277exoticclass64,65,69,70,80femaleethics272experience38,52,77,712feministphilosophy552–66,576James780concernwithdifference560metaphysicsof726–33,736conjunctiveconception559,560Mill’sdefenceof767critiqueofliberalism272–3,278relationwithphilosophy576critiqueofreason589religious472–4dichotomies555–6sensoryexperience714,729,731,732discussionquestions565–6seealsoempiricismecofeminism520,526‘ExperimentalPhilosophy’645femininephilosophicalvoice272,556,explanation557,558,559causalexplanations311–12feminismandphilosophy552–8coveringlawmodelof304–5,307,311,feminisminphilosophy558–63433,434,436,437,438andhistoricism559–60functionalexplanations311–12,383,integrationistambition559,560,563752–3,754internaliststance559,560,563historicalexplanations433–4,436,438masculinityanalysis554–5919\nINDEXfeministphilosophycont.disciplineandpunishment867andpatriarchalism272discussionquestions873–4philosophicalcommitments560–3genealogicalmethod866politicaltheory278knowledge/powerhypothesis866–7postmodern561–2onmadness864–5powerandidentityrelations561,562,poststructuralism864563onsexuality867–8public–privatedivide272,280foundationalism40–1,52,367–8,472andsexism553,555,559,577justifiedbeliefs40–1sociallocatedness557Kant2Ferguson,Adam720philosophyofmathematics357–68Feyerabend,Paulpoliticalphilosophy277folkpsychology193rejectionof3,6,8–9,55,368–9scientificphilosophy10–11,296–7Wittgensteinand8–9Fichte,JohannGottlieb65,742FourPrinciplesapproach(4PA)500–1Ficino,Marsilio569,571Francks,Richardfideism472,622DescartesandMalebranche644–57Field,H.127,188,362SpinozaandLeibniz658–70Filmer,Robert690,691freewill80,208,389,481Finnis,John,legaltheory242,417Cartesian651,661firstphilosophy6,8,63,677andevil481–2Firth,R.,idealobservertheory470FreeWillDefence481,482Fisher,R.A.Godand464,466–7groupselection326Hume713,720sexratioanalysis328misuse651Fishkin,J.S.281andmoralfreedom208Fiske,S.T.188Nietzsche832fitness319,337andontology80Flax,Jane554Spinoza661Flew,Antony,onreligiousbeliefs454,472seealsoautonomy;determinism;freedomFodor,J.A.176freedom31,65,260,379,662machinefunctionalism178–9,183Berlin258representationalism188–9anddeterminism380,467teleology184andequality258,261,266teleosemantics190–1equalityof265folkpsychology177,192,193,194–6flightfrom862Fontenelle,B.de378Hegel747Foot,PhilippaHume719,721ethicalnaturalism492liberaltheoriesof754moralcategoricity216Mill’sdefenceof762formalismmoralfreedom207,208,219inart239,240,241,248natureof389legal409,424negativeliberty258mathematical345,363–7andomniscience466–7musical245–6Sartre862,863Forms,Theoryof592,602–4,609self-mastery258Foster,John455andself-understanding747Foucault,Michel572,864–8spaceoffreedom535archeologicalmethod865threatsto719criminology866underthelaw719920\nINDEXfreedomofchoice534–5machinefunctionalism178–9Frege,Gottlob2,23,69,70,72,98,objectionsto185–7790–801,843teleological184abstractrepresentations800Turingmachinefunctionalism13accountofnames160–1,163,793,796,797Gaillie,W.B.444communicationasthought-sharingGalileanworldview64–5797Galileo645,647–8compositionalityprinciples800Galston,William269concept792gametheorydiscussionquestions803–4businessethics543foundationalism369co-ordinationgame393function791–2prisoner’sdilemma393–4,675indirectdiscourse799–801rationalityand391–5language2,98,359Gardner,Sebastian,aesthetics231–56logicalanalysis2,160Gassendi,Pierre686logicism791Gaukroger,Stephen,Bacon634–43mathematics2,359,360,361Geach,Peter176numberstheory360ethicalnaturalism492objects70,792genderequality265quantificationtheory792genderpersonality554realism359generalitytheory697,698semantics102–3,104genethics504–5sense–referencedistinction14,160,cloning505,506,507793–4,795,799,800,801,844embryoresearch505‘standingfor’160genetherapy504–5Fregeargument102–3,104genetictestingandscreening505frequencytheory308–9,310geneticsFreud,Sigmund474geneticdrift324Fricker,Miranda,philosophyandfeminismnaturalselection324552–66geometry351Friedman,M.294analyticgeometry353,356Fuller,Lon422–3axiomatizationof365–6,814functionEinstein358andform609Euclidean212,348,365Frege791–2fractalgeometry369mathematicalconcept353Hobbes678functionalexplanation311–12,383,Locke687–8752–3,754mechanistic678functionalthinking384motion678functionalism13,193,318non-Euclidean358,365computerfunctionalism13,14,15proofsandidealizations347andeliminativism192–3Riemannianplanegeometry365andemotions188Gestell854andethics182Gettierexamples40,43,44etiologicalaccount319–20Gewirth,Alan534homuncularfunctionalism182–5Ghiselin,M.335andinstrumentalism191GilbertofPoitiers620andintentionality184Gilligan,Carol,femininephilosophicalvoicelevelsofdescription179272921\nINDEXGillon,Ranaan,FourPrinciplesapproachGod79,224,468–70,479,480(4PA)500GreaterGoodDefence481,482globalization531,538,541,546justsociety261,267,277Glymour,C.300Kant219–20God65,74Moore’sethicaltheory805–6Augustineon467–8Platonistmetaphysicsof806asBeing480Rawls’stheoryofjustice9–10,31–2,beliefin458261,493Berkeleyon704,705valuepluralism259,267,268causality659seealsovirtue;virtueethicsclock-makermetaphor650,666Goodpaster,K.E.536andcreation64,80,81,456,469–70,Gordon,R.M.188478,518,665,667Gorman,J.L.429,431deathofGod224,829Gould,S.Descarteson50,649–50,653,654,658adaptionism324andevil465–6,469,480–3evolutionarytheory329existence53,79,163,223,330,353,sociobiology336458,459,462,621,624–5Gracia,JorgeJ.E.,medievalphilosophyformsofbeliefin456619–33andfreewill464,466–7grammars111,123goodness79,224,468–70,479,480depthgrammaroflanguage20identificationwithNature659Foucault866ineffability462–3gravitationaltheoryintellect350Newtonian298,321,354Leibniz665–7scholastic647Locke470,683,688,689Grayling,A.C.,epistemology37–60medievalphilosophicalfocus621Greekphilosophy568,569,571,585–600,andmetaphysics626–7636andmorality224axiomatics585namesof625–6dialecticalmethod586–90relationtothephysicalcontinuumdiscussionquestions599–600649–50gladiatorialconception589sorrowandpleasure471Hellenisticmethodology594–7Spinoza658–9Hellenisticscepticalcontroversy586seealsodivineattributesintellectualcohesion595Gödel,K.logicofself-refutation590incompletenesstheorems345non-Greeksources571mathematicalconsistency366,367Presocraticphilosophy585Godfrey-Smith,P.337seealsoAristotle;Plato;SocratesGoldman,Alvin43greenpolitics524Goldstein,L.J.430Green,Ronald499Gombrich,E.H.243Griaule,Marcel572Goodman,NelsonGrice,Paulconfirmationrelations168analysisofmeaning15–16,115–28,inductivelogic167–8161musicalexpression245conversationalimplicaturetheorysemiotictheoryofart241,243128–30,136goodness/thegood805–6pragmatictheory128–32autonomyof223semantictheories97communitarianismand270–2Griffiths,P.188922\nINDEXGrimshaw,Jean,philosophyandfeminismHegelianlogic743–4552–66historyofphilosophy570Grotius,Hugo494idealism747groupidentity274Indianphilosophy571Gunderson,K.185individualism746Gutting,G.natureandspirit744–6religiousexperiences473perception742–3Sartre,FoucaultandDerrida860–74PhenomenologyofSpirit741,742–3responsetoKant742Haack,Susan,pragmatism774–89Heidegger,Martin459,571,835,848–55Habermas,Jürgen33analyticofDasein850–1,852onnationality277art852,854politicalphilosophy279–80being-in-the-world850,851Haldane,J.B.S.,appliedethics490–8Bestand854Hamilton,Alexander720biography848–9Hamilton,W.D.327discussionquestions858–9Hanslick,E.245–6forgetfulnessofbeing854Hanson,Norwood295Gestell854abductivelogic785intentionality849happiness79,377interpretationofNietzsche853Bentham759,760,761phenomenology849,850,852GreatestHappinessPrinciple415,761–2onpoetry853Hobbes673,674,676religionandphilosophy853–4Mill205,206,210,761–2,763,765,technology854–5768Hellenisticmethodology594–7Socrates607–8Hellenisticscepticalcontroversy586Hare,RichardHelsinkiDeclaration513–14appliedethics493Hempel,Carlidealobservertheory470coveringlawmodelofexplanation307,Harman,G.785311,433,434Harris,Leonard576historicalexplanations433Harrison,Ross,Bentham,MillandSidgwickHempel,P.304759–73Hennig,W.322Hart,H.L.A.henotheism456legalpositivism408–9,411,412–14,HenryofGhent627415,416,417–18,502Hepburn,Ronald455legalreductionism406Herder,J.G.574Hart,W.D.187hermeneutics385,397,574Harvey,William646,647doublehermeneutic385,386Hasker,W.482hermeneuticcircle397Hausman,D.M.544Heyting,A.364Hayek,F.A.von258Hick,JohnHegel,GeorgWilhelmFriedrich33,65,67,problemofevil482741–9religiousworldview472art237,240Higginbotham,J.111biography741–2Hilbert,D.dialecticandsystem746–7axiomatizationofEuclideangeometrydiscussionquestions749365doctrineoftheconcept742,744foundationalism369ethnicityandphilosophy570,573Hilbert’sProgramme365,366923\nINDEXHinduism460,468,473,570passions673Hippocratictradition503,512peace676–7historicism437–8politicalphilosophy671,672,676–7feminism559–60socialcontracttheory394,671,675history,philosophyof390,428–52,574thestate671,674–6anti-realism430war671,672–3,674characterizationargument442–3Hôlderlin,Friedrich853,855chronologicalsequence448holism383,386–7colligatoryconcepts447,449,450andindividualism389consequentialistcriterion442ofthemental78constructionism430–1scientificdiscourse6–7discussionquestions451–2socialscience376,386–7,398evidentialstatements431andutilitarianism213–14explanationandunderstanding433–9Hollis,Martin,philosophyofsocialsciencehistoricaldisagreements448–9375–402historicism437–8,559–60Holm,Søren,bioethics,genethicsandindividualityissue435medicalethics499–516knowledgeofhistoricalfact428–33homogenization275narrativeandrealism444–9homuncularfunctionalism182–5narrativesentences440Horwich,P.125,127objectivityandvalue439–43Hountondji,Paulin572pastandpresentviewpoints439Hubbard,B.A.F.592past-entailingterms429,431,432Hull,D.335,337rationalreconstruction436–8HumanGenomeProject504retrospectivity440,445humannature259,377,379roletheory388,435–6Descartes650–2theoryofinterpretation431,432,433Hobbes672–3uniquenessofhistoricalevents434–5Hume379–80,721valuejudgements440Leibniz667–8verificationism431mechanistaccount662,668HIV/AIDSsocialpolicy511,514pre-culturalage596HLYEs(HealthyLifeYearEquivalents)511Rawls259Hobbes,Thomas671–81Spinoza661–3Aristoteliantheory678humanism457,460commonwealth674,676,678environmentalethics518DeCive672,673Renaissance634denialofCartesianimmaterialism677–8Hume,David2,73,302,654,709–24discussionquestions680–1aesthetics232,233,234,235,247duty674–6art,tasteandaestheticjudgement717–18humannatureandsociety378–9,394,atomism73671,672–3onBerkeley696lawsofnature674biography709–10Leviathan378–9,394,671,672bundletheoryofmind846materialistmetaphysics677,678causation303,304,306,711–12,721mechanisticgeometry678consciousness731onmentalactivity677designargument330–1onmotion678discussionquestions723–4naturaljustice671economicviews719–20,721naturalscience671empiricism66,226,380,684,705–6,optics671,678712,714–15924\nINDEXexternalobjects713idealutilitarianism806freewill713,720idealism53,65,68historyandpolitics713,718–20Berkeley53–4,694–6hostilitytometaphysics64,65,66Hegel747humannature379–80,721Kant726–33idealobservertheory470Moore’srejectionof805,807inductivereasoning712,721physicalismand65influence720–1transcendentalidealism727,728,845innateideas712,720ideasandinterpretation713,718absoluteidea744–5isandoughtdistinction338abstract696–8justice260,717,721assimilationtosensations701–2liberty719,721complex686metaphysics64,65,66conceivability703miracles483,714–15innate352,683–4,712,720moralphilosophy225,716–17modes686naturallawtheory79positive698,699objectiontoethicalobjectivism233relative698passions392simple686perception711,714–15substances686onpersonalidentity713,714identityclaims80phenomenalism704,705–6identitystatements794–6philosophy710–21identitytheory77–8,176–8philosophyofreligion458,714–16biologicalidentity523propositions225,226,227common-sensecausalversion177reasonandsentiment606,716–17,721cross-categoryidentities77scepticism52,54,699,705,720andeliminativism192–3toleration719Hume713,714Husserl,Edmund842–8Locke689categorialintuition844,845naturalismandidentity75–80conceptofnumber843personalidentity689,713,714consciousness844–5,846,847,848personhood507–8discussionquestions858–9powerandidentityrelations561,562,intentionality843–4563intersubjectivity847,848relativity689lifeworldconcept847,848tokenidentity178,313perception844–5token–tokenidentitytheory13phenomenology843,845,846–7type–typeidentitytheory13,77psychologism843Wittgenstein819transcendentalidealism845ideology,exhaustionof257–8transcendentalsubjectivity845,847,illocutionaryacts8,16848assertives8Hutcheson,Francis716commissives8Huxley,T.H.338,721declarations8hylomorphictheory687directives8hypothetico-deductivemethod381expressives8illusionIChing571illusiontheoryofart243–4ideallogicallanguage359,796Muller–Lyerillusion296idealobservertheory470Peirce778925\nINDEXimaginativewriting34metaphysics78seealsoliteraryworksnaturalizeddefenceof303imagism697,698phylogeneticinference322–4immanence471,480,659,731,846problemofstatisticalinference310immaterialismrationalinference38Berkeley53,694,696,698,700,705refutations287Descartes174,176,677–8inscience286–7Locke689andsemantics114impartiality211–12,221theinfinite355–6,357imperativallogic164AxiomofInfinity362implication135Berkeley355cancellability129cosmos477detachability129Godand659implicature135Hegelianlogic744seealsoconversationalimplicaturesmathematics354,355,357,362,364implicitknowledge111infinitesequences310,357incommensurability295–7information-processing180inconsistency62innatecapacities363indeterminacyoftranslation195innateideasIndianphilosophy570,571,574–5,577Aristotelianconceptualism352indirectdiscourse799–801Hume712,720indirectspeechacts16Locke683–4individualism383,746inner-causethesis191,192Hegel746institutionaltheoryofart237andholism389instrumentalgood271liberal274instrumental–expressivistpolitics279,280Locke683instrumentalism191–2,193,217,220methodological273,274scientific294,295,296,297,301,355social376,388,395semantic294individuation630integrity,inlaw420–3Descartes649intellectualcapacities352induction52,167,640,712,721,778,intelligibility106,115785intentionality21,386,436,860eliminativeinduction639aesthetics248–9falsificationism10,323,324anti-intentionalism248,249pessimisticmeta-induction300–1,312Berkeley703problemof286–7,289,290–3,294,emotions188,245302–3andfunctionalism184underdeterminationoftheorybyevidenceofGod478297–300Heidegger849ineffabilityofGod462–3Husserl843–4infallibilism39–40identification386infantilebehaviour596languageofthoughthypothesisinference785188–91deductiveinferences287,348ofmentalstates184Humeand715problemswith185,187–8identificationofpatternsof78speechacttheory15–16inductiveinferences286–7,303,385ofthoughts96,97justificationof38interculturalism573–6mathematics348interests205,745,761926\nINDEXinternalism42pragmatism774,779–81,785interpretation19,105–6,115–16,radicalempiricism780247–50,713self-awareness389art247–8,249,250,718truth780–1axiomatizationconstraint115Japanesebusinessethics540constraintsofagreement115Japanesephilosophy575history431,432,433Jardine,B.785Hume713,718Jeffrey,R.C.153literaryworks247–8Jewishphilosophy575–6,623open-endedness133JohnofStThomas620philosophyoflaw418–20,421Johnston,M.124pragmaticinterpretation137,139Jones,Peter,Hume709–24radicalinterpretation106Judaism457,460,480,481semantictheories106,115–16,139judgements225intersubjectiveverifiability174aesthetic232,234,235,236,717intersubjectivity847,848content225intuition38,223,350,356,378,729,ofhindsight719731,732,766,767,770Moore’stheoryof807,808categorialintuition844,845moral26,27,28,222,227,260,269,Kant729,732470Locke685normative225,226,227Mill766subjective225moral30,222,223theoryof726–9,797Peirce776jurisprudenceSidgwick769,770censorial502spatialintuition357ethicsand502seealsoperceptionexpository502intuitionism73,345,364–5,500justsociety261,267bioethical500justwardoctrine495mathematics345,350,356,364–5justice‘primitive’500agreedprinciples259,261Inwood,Michael,Hegel741–9Derrida871Irigaray,Lucedistributivetheories263,266,266–7,cavernmythinPlato’sRepublic554276,281,510femininephilosophicalvoice556,557egalitariantheories263ironyentitlementtheory263echoicuseoflanguage134,135andethicsofcare273relevance-theoryaccounts132,134,Hume260,717,721135liberalconception278Irwin,T.H.593Marxon755–6IsidoreofSeville620Nozick’stheoryof263Islam457,460,480,481politicaljustice267Islamicphilosophy623,626primacy260punishmentseepunishmentJackson,F.186Rawls’stheoryof9–10,31–2,258–61,James,George571265,266–7,270–1,272,273,277,James,William721,779–81278,281,492doctrineoftheWilltoBelieve779senseof,acquisition261languagegames66sexualjustice272nominalistleanings775socialistjustice274927\nINDEXjustification39–41,42,43,44,56Kenny,Anthonyfallibilisttheories40appliedethics493problemofscepticism45theism462JustinMartyr622Kepler,Johannes,planetarymotion349,645,646Kamm,D.538Kierkegaard,Søren586,871Kant,Immanuel2,65,75,654,721,Kirk,R.176725–40Kitcher,P.302,338aesthetics232,233,234,235genicselectionism327anti-Semitism577Kivy,P.246onbeauty235knowledge12,96,396biography725–6aprioriknowledge602,685,728,729,casuistry494731,736,766CategoricalImperative219,500,734,absoluteknowledge743735–6acquisition38constructivism355Aristotelianmodel608–9,623contractualism31,221–2Bacon636andtheCopernicanRevolution727–8andbelief685cosmopolitanism574,577Cartesiantheory652,775CritiqueofPureReason52,279,356,conditionsfor39726,727,729,730,736defining38–9,40,44,56discussionquestions739–40demonstrativereasoning623,684disinterestedness236Dewey782foundationalism2empirical685Hegel’sresponseto742andexperience38idealism726–33fallibilism433individualgood219–20,262foreknowledge466,467theinfinite357historical428–33intuition729,732howweknow627–8Kantianethics500,612,733–6illumination627–8knowledge728,729,731,736,742implicitknowledge111maximofaction217,218innateknowledge602,683,684mentalharmonytheory239intuitive684,685metaphysicsofexperience726–33,asjustifiedtruebelief37–8,39736Kant728,729,731,736,742moralphilosophy218–19,223,224knowledge/powerhypothesis866–7andmoralstatements27Locke683–5naturallawtheory405middleknowledge467philosophicalsectarianism569moralknowledge606,613philosophyofreligion459object-relatedelements729properties726–7Plato601–4reasonandautonomy219,224,733–5problemof376religiousphilosophy458propositionalknowledge38–9scepticism52–3psychologicaldimension636spaceandtime356–7andreason38transcendentalidealism2,727,728rhetoricalmodelfor636transcendentalunityofapperceptionself-knowledge385733sensitiveknowledge685Karnofsky,E.S.592socialphenomenon715Kelsen,Hans,legalpositivism406,415sociologyof368,396928\nINDEXsubject-relatedelements729naturallanguage101syntheticknowledge728philosophicalpriorities96–9tacitknowledge111,123philosophyofaction8seealsobelief;epistemologyplasticity820,822Kôrner,Stephan345predicateterms93,94Kripke,Saul14,77PrincipleofCharity106,115functionalism185PrincipleofHumanity106,107,115,naturalkinds333116onWittgenstein20privatelanguageargument817–20,847Krophardt,J.544questionsofmeaning90–2Kuhn,Thomas853Russell’spubliclanguagetheory797denialofrealism11semantictheories99–114scientificphilosophy10,11,296–7sensation-language818,819onscientificrevolutions10,11,395subjectterms93,94,98Kung,H.531tacitknowledge111,123Kvanvic,J.466theoremderivation93–4Kymlicka,W.275theoriesofmeaning92–6Wittgensteinon8–9,817–20Lakatos,I.325languageofthoughthypothesis188–91landethic524–6LaoTsu571Langer,S.246–7Laslett,Peter257discussionquestions145–6Laudan,L.300languagegames8,9,66Laurence,S.123language,philosophyof5–6,90–146,469law,philosophyof403–27acquisition55adjudication409–10,421–3,424actuallanguagerelation104ambivalenceoflaw404authorityofrules821analyticaljurisprudence411,416communitytheoryoflanguage822aspirations404compositionaltheory92Bacon635contextualfactors91andbioethics502conventions138black-letterrules404–5,410,421,422conversationalimplicatureandrelevanceborderlineandpivotaldisputes413theory128–39constructiveinterpretation419depthgrammarof20conventionalism420,422echoicuseof134–5convergenceinlegaljudgements422ethicallanguage74criminallaw425evaluativelanguage74discourseofequality420,421factual815discussionquestions426–7formallanguage101doctrineof‘commonmistake’408–9Frege2,98,359economicanalysisoflaw425grammars111,123exclusivityofsources414Grice’sprogramme115–28,136fidelitytolaw423implicitknowledge111formalism409,424limitsofmeaningfullanguage66growingcomplexityandincoherenceliteralmeaning91422logicalstructure816integrity420–3loosetalk134interpretativetheories418–20,421meaning-specifyingtheorem92,94,lawandforce404–5100,102legalpositivism404,405–7,408–9,mindandmetaphysics96–9411,412–14,415,416,417–18,422929\nINDEXlaw,philosophyofcont.Leibniz’sLaw95,186legaltheory411–15,418–20Leopold,Aldomoralcoherence421,422,424environmentalethics524moralneutralitythesis404–5,408,414landethic526moralnon-conclusivitythesis415Levin,Michael182moralobligatoriness405Levinas,Emmanuel575–6naturallawtheory404,405,406,Levine,J.,functionalism186–7415–16,417Lewis,C.I.785normativejurisprudence411Lewis,Davidobiterdicta410causation305,306,307penumbraluncertainty409conventionalregularity120andpolitics407–8,423–4identitytheory177pragmatism421Lewontin,R.privatelawdoctrines425adaptionism324problemsofjurisprudence404–10sociobiology336propositions412,415liberalequality266ratiodecidendi410liberalpatriarchalism272reductionistanalysis405–6Rawls260,265andrhetoric636–7liberalpluralism32rulesseerulesliberalismseparationoflawandmorals412–13,andautonomy269,407–8414collectiveprojectpursuit407,408socialcontext420communitariancritique270–2sovereignty411criticsof269–74systematizationoflaw635andculturalpluralism275utilitarianaccounts425andtheeconomicsphere281andutility766feministcritique272–3,278validitycriteria409andgrouprights275lawsofmotionindividualism274Descartes667liberalequality266Hobbes678liberalvirtues269Newton’slawof354Marxistcritique273–4paradoxesof62moralneutrality424lawsofnature219,220,305–6,321,378,philosophical33,281–2379,381,645–6,674,778politicaltheory423–4Hobbes674public–privatedivide280LeDoeuff,Michèle,feminism552,558rejectionofessentialism424leadership541socialistcriticsof265Leavis,F.R.251libertarianism32,264,265Leftow,B.467libertyseefreedomlegalpositivism404,405–9,411,412–14,life415,416,417–18,422,502–3afterdeath75Leibniz,GottfriedWilhelm571,663–8definitionof317–18discussionquestions670endoflifedecisions508–10God665–7startingpoint506humannature667–8LikelihoodPrinciple322,331innateknowledge602Lipsey,Richard381monads665–6literaryworks34,248–50Nature664–5authorialmeaning248reconciliatoryprogramme663intentionalism248,249930\nINDEXinterpretation247–8inductivelogic16,167textualmeaning248,249mathematical5,69,345,358,359,Loar,B.122,123361,362–3,790,791Locke,John2,37,48,264,682–93non-monotoniclogics367accountofabstraction687–8,696–7normativetheoryofbelief203Aristoteliantheory687philosophicallogic148civilsocietyanalysis690–1possibility151–64corpuscularianism686,687andtheprinciplesoflanguage359discussionquestions692–3propositionsseepropositionsdivineownership470relationalreasoning367duty690standardlogic95empiricism684Strawson’scharacterization147,150,epistemology683163EssayConcerningHumanUnderstandingsymboliclogic2,5682–9Wittgenstein168–9geometry687–8logicalbehaviourism12God470,683,688,689logicalconstants158onideas686,696,700logicalform69,790immaterialism689logicalpositivism2,3,11,66,174,454individualism683logicalrelations147,159,167,168innateideas683–4consequence147intuitiveknowledge685consistency147knowledge683–5logicism5,69,345,358,359,361,naturallaw415362–3,790,791particularism696Frege791perception684Russell791andpersonalidentity689Long,A.A.595personhood507–8loosetalk134philosophyofreligion458Lovejoy,A.O.439politicalintolerance268Lukes,S.755,756politicalphilosophy690–1Lycan,W.G.124properties686–8philosophyofmind173–201rationalism682–3,684rights262,576Mach,Ernst300scepticism52,683machinefunctionalism178–9andslavery577challenges182–3substance686,687,688levelsofdescription183–4logic3,38,41,46,72,147–72,351machinelanguages189abductivelogic785MacIntyre,Alasdair32,271,445Aristotelian586,621Mackie,John28authorityof75McPherson,M.S.544dialectic622–3McTaggart,J.E.83–4discussionquestions171–2Madison,James720eroteticlogic164madness,Foucaulton864–5firstorderpredicatecalculus367Maimonides,Moses(MoshebenMaimon)formaldeductivelogic148,167,782626Frege2,160,791Malcolm,Norman458Frege–Russellframework364Malebranche,Nicholas652–5,710Hegelianlogic743–4causation655imperativallogic164discussionquestions657931\nINDEXMalebranche,Nicholascont.Aristotelianconceptualism351–2,353doctrineof‘occasionalcauses’655,666arithmetic353,354,356,361knowledgeofselfandGod655challengetoempiricism346mind–bodyrelation655computationalandmeasurementpracticesnature664352perception652–4conceptualism351–2,353,355–7,manneradverbs91368–9manner,maximof16consistency365–6,367Margolis,J.785discussionquestions373–4Martin,Michaels454,459firstprinciples350,352Martin,R.436formalism345,363–7Marx,Karl750–8foundationalism357–68alienation750,751globaldescription361discussionquestions757–8idealizations347ethicaltheory750,755–6theinfinite354,355–6,357,362,364functionalexplanation752–3,754infinitenumbers358justice755–6infinitesimalcalculus354politicalphilosophy258,750–6intuitionism345,350,356,364–5Popperon750Kantianframework355,356–7religiousexperience474logicism5,69,345,358,359,361,superstructureandideology753–4362–3,367,790,791theoryofhistory750,751mathematicalknowledge350,357,358,Marxism1363–4analyticalMarxism273–4,278,751–3mathematicaltruths361capitalism278numberstheory351,357–8,360,364,critiqueofliberalism273–4365,843economicdeterminism281,379,754observationalscience777historicalnarratives448–9paradoxes354natureofhumanbeings449pathologicalfunctions357productionanddistribution281Platonicrealism349–51,358–9,363structuralistMarxism751principlesofinference348utopianism274proofs347–9,350,352,364,365,masculinityinphilosophy554–5,559369–70materialbodies114reductiontologic362–3materialism12,13,14,65,186,699settheory365,366anti-materialism186ofspace354,356Aristotle610,611symbolsystems365–6Descartes660theoreticalreasoning348eliminativematerialism192–3,462Wittgenstein364,368–9,822historical754matterHobbes678abstractionand696materialisttheoriesofmind182conceptof53physicalistidentitytheory13corpusculartheoryof687tokenmaterialism178divisibility665seealsoimmaterialism;physicalismandform609–10,665mathematics,philosophyof3,38,41,46,hylomorphictheory68772,149,345–74mind-independentmatter694,695,699,analyticempiricism357–63704analyticgeometry353,356mauvaisefoi862applications346–9maxims734–5,736,775,776,779,780932\nINDEXMayr,E.334,335Wittgenstein816–17,820Mead,G.H.785worksofart247–50mean,doctrineofthe613meaning-specifyingtheorems92,100,102meaning150,384meaning-theories12,16–18,72,73,74,analysisof2,15–16,91,115–2881,92–6,125,127,174assignment138‘MechanicalPhilosophy’645causaltheoryofreference12,14–15,medicalethics33,491,503,506–14,53221,469abortion508cognitivemeaning4dualresponsibility503conventionalmeaning119–22,123endoflifedecisions508–10conventionalregularity115,117,FourPrinciplesapproach(4PA)500–1120–2healthcareandjustice510–11determinacy156,250personhood507–8Dilthey384potentiality507–8emotivemeaning4professionalethics203,499–516entities91publichealthversusindividualrightsexpressionmeaning122–4511–12flexibleaccountof816researchethics513–14Grice’sanalysisof15–16,115–28,161speciesism507intentionalistictheoriesof12,15–16startingpointofhumanlife506intersubjective385,399medicaletiquette503Leibniz’sLaw95medievalphilosophy69,585,619–33literal117,120,121,122,123,138apologeticworks621meaningdespiteuse123argumentsfromauthority624meaningwithoutuse123Aristotelianlogic621meaning–truthconnection126,127casuistry494–5meta-semantictheory95characterof621,631minimalismabout124–6dialectic622natural116discussionquestions632–3natureoflinguisticmeaning91doubleaim621no-theorymeaningofmeaning124,existenceofGod624–5125,127faithandreason622–3non-natural116Godascentreofintellectualattention621objective385individuation630ontologicalquestionof91logicavetus620ostensivedefinition817namesofGod625–6picturetheoryof457,459patristictradition620plasticity820philosophicalproblems621–2propositions91problemofhowweknow627–8questionsof90–2problemofuniversals628–9,630semantictheory95,96scholasticism620–1,636,647speaker’smeaning115,117–19temporalandterritorialboundaries620specificationsof116theologicalworks621statesofaffairs91theologyandmetaphysics626–7subjective385,398–9Meinong,Alexius844truth-conditionaltheoriesof12,16–18mentalascriptions192,193usetheoryofmeaning125,127mentalstates13,77,176,177,179,182,varieties385184,192verificationtheory174causalrelations13weakspeaker’smeaning122semantictheories107–9,112,115933\nINDEXmentalterms177,192,194extensionofBenthamism760,761mentalism14,21happiness205,206,210,751,761–2,mentality,evolutionof337763,765,768Menzel,C.465immanenceofGod480Merleau-Ponty,Maurice847,863indirectorruleutilitarian765–6meta-ethics25–9,30,203,225–7,491,intuition766492,503metaphysicsandlogic766–7value-neutral26nationalism276meta-induction300–1,312reality767meta-semanticdeflationism127–8,139reinterpretation763–6meta-semantics90,91,115sanctions764,765metaphor16species333conversationalimplicature132utilitarianism211,761–3,764,765–6,asloosetalk134767–8relevance-theoryaccounts135values762metaphysics61–89,96Millikan,R.G.184analysisandlogic69–75mimesis238–9,242,248,250andart252mind,philosophyof8,20–1,75,96,assimilationtoscience76173–201,389,439authority81behaviourism5,6,13,69,174–6descriptivemetaphysics61,62cognitivepsychology179–80discussionquestions88–9computationaltheory12foundation72computermodelofmind15,180–2andGod626–7Descartes173,377,660hostilityto64,65,66discussionquestions200–1Hume64,65,66DoctrineofIdols636–8medieval626–7eliminativismandneurophilosophymetaphysicalproblems62192–4methodproblem69–75,76,81folkpsychology177,192,193,194–6naturalismandidentity75–80homuncularfunctionalism182–5andphilosophyoflanguage98–9identityseeidentitytheorypracticaldimension62instrumentalism191–2preoccupationwithownpossibility63intentionality21revisionarymetaphysics61languageofthoughthypothesis188–91scientific776machinefunctionalism178–9self-image61,76materialisttheories12,182semanticmethodology74andmetaphysics63–7andsemantictheories112–14mind-independentmatter694,695,699,supervenience80–1704time82–5mind–bodydualism21,173–4,176,Mill,JohnStuart721,761–8,770,844177artoflife764neurophilosophy193conservativeholism214andphilosophyoflanguage8,20–1,criticismsofBentham76396–7defenceofexperience767problemswithintentionality187–8defenceofpoliticalliberty762problemswithqualiaandconsciousnessdiscussionquestions772–3185–7egoisticprinciple763,770andreligion462empiricism380Spinoza661ethicalandpoliticalthought764superveniencethesis80–1,318934\nINDEXteleology182–5moralagents270,553truenatureofmind70moralcategoricity216,221,222,224seealsoartificialintelligence;emotionsmoralclaims490,491mind–bodyrelations21,75,76,173–4,moralconduct389176,177,532moralfreedom207,208,219Descartes173,651,655,667moralintuitions30,222,223Malebranche655moralirrationalism494minimalismmoraljudgements26,27,28,222,227,aboutmeaning124–6260,269,470abouttruthforpropositions126–8moralmotivation606–7Mink,L.O.446,447moralnorms216miracles483,649moralphilosophy212–14,490,491,493,Hume483,714–15733–6Mises,Richardvon,frequencytheory308–9deficiencies500misogyny553,555,559positivistconceptionof9modaladverbs91,112startingpoints493modernism259,276seealsoappliedethics;bioethics;ethicsCartesianversion645moralprinciples221,735monads665–6moralrationality33monarchymoralresponsibility182,207absolute690moralsentiments224,716–17divinerightofkings690,691moralutterances,semanticstatus27Locke’scritiqueof690moralvalue490monism324morality33,203,208,215–16,490,anomalousmonism192745andintentionalism249,250andart203neutralmonism70,780contractualistview221andpluralism250ofemancipation755valuemonism268andevolution338monotheism456,460,468andGod224Montaigne,MichelEyquemde586Hume225Moore,A.W.,philosophyoflogic147–72Judaeo-Christian836Moore,G.E.2,26,70,763,770,780,languageof491805–10Marx755common-sensism807–8,809materialistbasis182discussionquestions810Mill764ethicalnaturalism806andreason222ethicaltheory805–6Recht755extremerealism805,807religiousmorality836–7,838idealutilitarianism806separationoflawand412–13,414Moore’sParadox808–9Sidgwick770naturalisticfallacy805–6seealsoethicsoppositiontoidealismandempiricismMorris,T.V.465805,807Mothersill,M.237perception807motivation259,392,713phenomenalism809Mudimbo,Valentine572philosophicalanalysis808–9Müller–Lyerillusion296propositions805,807,808multiculturalism567,568,574,575rejectionofidealism807musictheoryofjudgement807,808appreciation246935\nINDEXmusiccont.opportunism328expressivequalities244–7sexratiostrategy327formalism245naturalism233,390,395,766,775mysticalexperience455atomistview664mythofthegiven41Descartes645–8,650,664deterministview378Nagel,E.319empiricistcontention378Nagel,Thomas266epistemic597appliedethics493ethicalnaturalism492,596,806functionalism185,186Hegel744–5namesHellenistic595–6,597bearerlessnames796Hobbes671,672–5andcommunication796–7andidentity75–80Frege160–1,163,793,796,797Leibniz664–5propernames90,162,794–7,798Malebranche664reference160,162natureandspirit744–6Russell161–2,794–6,797,798Peirce778semanticsof112rationalistcontention378sense160,161,162socialscience377–84Strawson163Spinoza659–61,664usedforGod625–6stateofnature674,675,690nationsystematicwholenessof664defined276vitalistview664,665nation–statedisassociation277naturalisticfallacy522,763,805–6nationalism275,276–7necessitation306defenceof276necessity74,79,112,113,380,711normativeclaims276causalrelations711unsympathetictreatmentsof277logicalnecessity813naturallawtheory79,80,404,405,406,socialscience380415–16,417negation98,101naturalphenomena645–8Negritudemovement575naturalphilosophy635,636,638,641,neo-Platonism350,353686–7,688neo-Pyrrhonists597alchemy635Nerlich,G.356Bacon635,636,637neuralmodelling184,193scholastic635neuralnetmodels14seealsoscience,philosophyofNeurath,Otto73naturalscience38,671neurophilosophy193seealsoBacon;Descartes;Leibniz;neurophysiology182Malebranche;Spinozaneutralmonism70,780naturalselection174,319,324,332,neutrality,political268337‘NewPhilosophy’645adaptivecomplexity328,332Newton,Isaac646,682adaptiveimperfections332gravitationaltheory321,354altruistictraits326mechanics354,358,378culturalgroupselection337Opticks349genicselectionism326,327secondlawofmotion354groupselection325,326,327,337–8Nielsen,Kai454individualselection325,326,327Nietzsche,Friedrich586,827–41multi-levelselectiontheory326accountofhumanpsychology831–2936\nINDEXaesthetics252differenceprinciple264anti-realism831,834entitlementprinciple262,263andtheasceticpriest837,838equality262,263biography827–8justice263onconcepts830–1libertarianism264,265deathofGod224,829minimalstate262–3discussionquestions841ownership261,262,264eternalrecurrence828,838–9politicalphilosophy261–2freewill832numberstheory351,357–8,360,364,genealogicalaccount833,836365,843Heidegger’sinterpretationof853Nyerere,Julius572metaphysics835nihilism829,836,838,839,853Oakeshott,Michael279Overman828,838,839Obenga,Theophile571perceptionofmodern‘catastrophe’objectivity28,67,68,227,233–4,744,829–30775perspectivism828,834,835aestheticobjectivism232–3physio-psychology832,833,835,836ethicalobjectivism233,339redeeminghumantype838,839ethics26,27rejectionoftheself832Hegel744onreligiousmorality836–7,838historical439–43scientificaccountoftheworld833–4,normative227835objects96,99,780,793sexism553,577abstractobjects70,792,793‘trueworld’illusion830,831,833intheexternalworld713,726–33,767willtopower828,832,834,835,836,Frege70,792837,838,853objectsofknowledge782,794willtotruth834ontologicalcategory98nihilation861–2perceptualrelativities48nihilism829,836,838,839,853qualitiesof48Nishida,Kitarô575spatio-temporalobjects52,53Nkrumah,Kwame572unperceivedcontinuedexistenceof53no-classtheory362observation777noema845,846occasionalism655,666nomadism561Ockham’srazor298,472,629nominalism778,779,832Ogotemelli572nomotheticscience321Okin,SusanMoller272non-cognitivism415,440omnipotence463–6,482,659non-physicalevents174,176omniscience461,466–7,468,667Norman,Richard,Marx750–8andfreedom466–7normativeethics532,539ontologicalpriority96,97North,D.C.544ontology851Norton,BryanG.520Descartes647nothingness(Sartre)861–2entities850noumenalism779existenceofGod474–6,625Nowell-Smith,P.430fundamental850Nozick,Robert32,261–4limitsof79–80anarchy261–4andphenomenology850basicrights261–2,264religion474–6belief43socialscience377937\nINDEXoperators101peace,Hobbeson676–7Oppenheim,C.304Peacocke,C.111,120,122optics623,646,671,678Pears,David,Wittgenstein811–26oracles437,438Peirce,C.S.331,774,775–9,780,784Oruka,Odera,Africanphilosophy572agapism776other-mindsproblem376,386,389,397,beliefanddoubt776–7848common-sensism777overdetermination,asymmetryof306–7metaphysics776Overman828,838,839pragmatism774,775–9,784Owen,G.E.L.592reality779Owen,H.P.469onSchiller783Owens,J.177scholasticrealism776,778,779ownershipscientificphilosophy776,777,780divineownership470synechism776ofresources273theoryofcategories776self-ownership261,262,264,274theoryofinquiry782theoryofperception778pain177,178,760,769truth779Paley,Williamtychism776designargument330–1,332Penelhum,T.472LikelihoodPrinciple331perception41,47–8,81,703panentheism456Berkeley53–4,694–6,700–2Pannikar,R.577Descartes646–8,653pantheism456,658–9Hegel742–3Papineau,David,philosophyofscienceHume711,714–15286–316Husserl844–5Pappas,G.193Locke684paradigms395–6Malebrancheon652–4analyticparadigm71,72Mooreon807cognitivefeature395natureof46socialfeature395Peirce’stheoryof778paradox62‘seeingin’244ParallelDistributedProcessing(PDP)193sense-datatheory130paralogisms730–1Wollheim’stheory244paraphrasis768–9seealsoexperience;sensationPargetter,R.185perceptualatomists73participation,politicsof279perceptualrelativities48–9particularism696–7,698Percival,Thomas,medicalethics503particulars96,98perfectionism268Pascal,Blaise217performatives7casuistry494PersistentVegetativeState(PVS)509,wagerargument483510passions716–17,734,865personality389Hobbes673personhood507–8Hume392perspectivism186,828,834,835motivation392pessimisticmeta-induction300–1,312Pateman,C.280phainomena592,594Pater,Walter244phallocentrism555patientautonomy511,512phaneroscopy785patriarchalism272phenomenalindividuals186938\nINDEXphenomenalism5,6,53,54,71,72,704,critiqueofart251809,817,818anddivineownership470Berkeley700–6ethicsandpolitics606–7Hume705–6Euthyphro223,457Moore809God460sense-data5,54,130,361,807,818justiceinthesoulconcept207phenomenology1,645,785knowledgeandreality601–4,627,628eideticreduction846,849Laws607existential863Meno601–2Heidegger849,850,852mimesis238Husserl843,845,846–7nominalisticPlatonism778,779andontology850Parmenides604Sartre863Phaedo602,605transcendentalreduction846Platonicdialogue590–2Phillips,D.Z.458Platonicrealism349–51,353,358–9,philosophicalmethod590363philosophypoliticalphilosophy258,607–8aim4,5polytheism468andexperience576Protagoras591flexibleconceptionof578renunciationofelenchos588historyof20,568–70Republic554,603,604,605,607secularization569revisionarymetaphysics61shiftfromotiumtonegotium635,641Symposium586subjectmatter4Theaetetus37task4,5,8TheoryofForms592,602–4,608,609astherapeuticenterprise562theoryofthesoul604–6phronesis613pleasurephylogeneticinertia324andaestheticjudgement233phylogeneticinference322–4Epicureanism596phylogeneticparsimony322,323–4seealsohappinessphysicalism13,65,81,178,192,193,Plotinus571318pluralism32,68,249,250,500,570seealsomaterialismbioethics500,501physics64,81,321culturalpluralism275Cartesian660evolutionarypluralism324,325geometricalphysics378Leibniz746Newtonianphysics354liberalpluralism32Nietzsche834andmonism250picturetheoryofmeaning457,459valuepluralism259,260,267,268place,ethicsof525Poincaré,Henri300Place,U.T.176polarconceptargument49Plamenatz,J.752polarities49Plantinga,Alvinconceptualpolarities49,50beliefinGod’sexistence472polisGod’screativewill464Aristotelianconception614ineffabilityofGod463ethics525positivism455goal614plantlife521–2Platonicconception607Plato45,590–2,601–8seealsocitizenshipcavernmyth554politicalactivism279939\nINDEXpoliticalassociation271failureof71politicalintolerance268hostilitytometaphysics66politicalneutrality268legalpositivism404,405–9,411,politicalscience257412–14,415,416,417–18,422,politicalandsocialphilosophy4,10,20,502–325,257–85,404,492logicalpositivism2,3,11,66,174,454discussionquestions285andmeaning-theory74economicsphere281meaningfulpropositions3equalityseeequalitypossibility74,151–64,166foundations277absolutepossibility155,433Hobbes671,672,676–7Church’stheorem153Hume718–20elimination153,154individualsandcommunities274–7epistemicpossibility155instrumental–expressivistdistinctionidentification153,154279,280investigation154jurisprudentialcharacter407logicalappraisal152justiceseejusticemetaphysicalapproach155liberalismseeliberalismsemanticapproach155,156–7Locke690–1senseandreference160–4Marx258,750–6syntacticapproach155,158–60Nozick261–4truth–falsitycombinations151–2,154,pluralismandneutrality266–9164andpolitics277–81possibleworlds464,667Popper258post-positivistphilosophyofscience10–11Rawls6,258–61,265,266–7,270–1,postcolonialism572272,273,278,281postmodernism62,838scopeof279feminism561–2task278poststructuralism863–4politicsofparticipation279powerpolytheism456,460,468andidentityrelations561,562,563Pompa,Leon,philosophyofhistoryknowledge/powerhypothesis866–7428–52practicalreasonPopper,Karlcategoricalconceptions218critiqueofPlato’spoliticalphilosophyethics216–18607instrumentalconceptions218evolutionarytheory337Kant734falsificationism10,287–90,298,299,socialscience388–9324Weber388–9historicism390pragmatics92onMarx750conversationalimplicatureandrelevance‘opensociety’258theory128–39politicalphilosophy258Grice’stheory128–32propensitytheoryofprobability309semantics–pragmaticsdistinction128scientificphilosophy298,299,383pragmatism1,66–7,68,383,459,501,populationgrowth518,520563,774–89PorphyrythePhoenician628andconsequentialism501positivism5,26,66,67,74,454,455,776contemporary783–5absolutepresuppositions67Dewey775,781–2,784,785andempiricism454–5discussionquestions788–9evaluative–descriptivedistinction4–6humanisticpragmatism782–3940\nINDEXlegal421propositionalknowledge38–9neo-pragmatism783–4propositions39,91,147,148–51,164Peirce774,775–9,784analytic–syntheticdistinction3,6pragmaticmaxims775,776,779,780cognitiveeffects133,134reformist775,785asconsequence147,163revolutionary775consistency147Schiller775,782–3declarativesentence-types148–9preconception,doctrineof596descriptive4predicate91,93,94,98elementary815predicatecalculus2,94,367empiricalenquiry226prediction289–90,778ofethics4PrincipleofCharity106,115evaluative4PrincipleofHumanity106,107,115,extra-linguistic150116factual813,814principles490,734–6falsifiability10agreedprinciples259Hume225,226,227legal502hyper-privatepropositions466prisoner’sdilemma393–4,675legal412,415privacyconcerns495,512–13logicallynecessarypropositions813privatelanguageargument19,55,mathematical348817–20,847minimalismabouttruthfor126–8probability45,291,308–11,381,712Moore805,807,808apriori705,706nonsensical3,4ofevidence715normative225–6,227frequencytheory308–9,310pictorialcharacter814mathematical291Quine156,157,162,163objectiveprobabilities309,310Strawson150probabilisticcausation307–8true–falsedichotomy3,7,151,163,propensitytheory309–10165,166quantummechanics310verification3,6single-caseprobabilities309Wittgenstein813,814,815statistical390Pseudo-Dionysus626subjective291psychicphenomena610weighingof721psychoanalysis252professionalethics512–13psychologicalphenomena184confidentiality512–13psychologism843dataprotection513psychosemantics187,190informedconsent512causal–historicalapproach190privacy513teleologicalapproach190pronouns91,112publiclanguagetheory797proofs347–9,352,364,365,369–70public–privatedistinctionpropensitytheory309–10feministphilosophy272,280propernames90,162,793–7philosophyoflaw280properties79,96,98,99,686,687punishmentDescartes647,648,653Bentham760,761,764ethicalvalue209deterrence760Kant726–7Foucault867Locke686–8Hegel745propertyentitlements262Nietzsche831propositionalattitudes116properaim760941\nINDEXPutnam,Hilary14,77ethical218,492machinefunctionalism178–9knowledgeoftheworld46naturalkinds333Locke682–3,684representationalism194medieval623truth785Spinoza661–2twinearthexample15,187seealsoreasonPyrrhoofElis45rationality33,207,219,220,368Pyrrhonians45,46businessethics543moralrationality33QALYs(QualityAdjustedLifeYears)511rationalityassumption192qualiaproblems185andrelativism395–7qualities,Berkeleyon700–2scientific291quality,maximof16andthetheoryofgames391–5quantificationtheory792universalstandards368quantity,maximof16Wertrationalität389quantummechanics11,24,80,307,Rawls,John258–61310–11,321,358contractualism261Bell’stheorem11differenceprinciple211,260–1,263,interpretation310–11265,271probability310–11andequality211,261,263,265Schrôdinger’sequation310humannature259wavefunction310,311intersocietalrelations531Quine,W.V.O.76,383,767,785justice258–61,265,266–7,270–1,analysisofmeaning16,161272,273,278,281analyticparadigm71liberalpatriarchalism272Duhem–Quinethesis7,298,299,300liberalism268–9logic150,363liberty260,265logicalconstants158overlappingconsensus267naturalizedepistemology363politicalphilosophy6,278propositions156,157,162,163reflectiveequilibrium30,260,493,494rejectionofanalytic–syntheticdistinctionsocialcontracttheory9–10,4926–7socialjustice32theoryofjustice9–10,31–2,277,492Rabelais,François586valuepluralism260,267,268racism753Raz,JosephRadhkrishran,Sarvepalli575autonomyandliberalism269Ramsey,F.305,306,785,812grouprights275rationalchoicetheory392,394,398realism28,68,75rationalegoism217,220,224anti-realism28–9,75,165,166,430,rationalinference38444,445,446,464rationalreconstruction111,386,436–8,Aristotelianconceptualism351–2439‘commonsenserealism’831historicalexplanation436–8denialof11ideal-typicalrationalreconstruction391ethicalrealism28voluntaristassumptions436extremerealism807rationalself-perpetuation120,121,122,Frege359138historical429–30rationalism38,46industrial-strengthrealism68Aristotelian608–9mathematical355,356Enlightenment682–3moralrealism606942\nINDEXnon-realism458,459,460causaltheory12,14–15,21,469Platonicrealism349–50,353,358–9,names160,162363sense–referencedistinction14,160,161,scholasticrealism776,778,779793–4,795,799,800,801,844scientific294,297,301reflectiveequilibrium30,260,493,495Scotisticrealism778ReformedEpistemology472realityReichenbach,B.482Mill767Reichenbach,H.785Peirce779Reid,Thomas458,721Schiller783relation,maximof16ultimate463relativism55–6,234,568,653,783realization183ethical539functionalrealization184justiceand271,277reasonandrationality391,395–7aestheticreason216,236relativitytheory290,298,321,350,358,andautonomy219704andfaith622–3relevancetheory132–5feministcritiqueof589reliabilism43Kant729,730religion,philosophyof453–89andmorality216,222contemporarypractice453–6natureof38definingreligion457overrevelation689discussionquestions488–9practicalreason216–18andevolutionarytheory330–3role606metaphysicalapproach456,457,458speculativereason742Nietzsche836–7,838theoreticalreason216non-foundationalistlines472universal573andphilosophyofmind462seealsorationalismproblemsofevil480–3reasoning152religiousbeliefs,natureof456–60demonstrative623religiousexperience472–4dialectical730Wittgensteiniannon-realism458–9,460inductive52,712,721seealsodivineattributes;faith;Godmoral494,495religiousbeliefsrelational367justification472syllogism640,767meaningfulnessof454transcendent65religiousexperiences472–4redistribution264non-religiousaccounts474reductionism69,81,557,611religiousworldview65,472analyticreduction175Renaissance621AxiomofReducibility362representation732,778,800biological319,321representationalism188–9,194,195failureof6,71reproduction324–5,334–5historical312,313asexualandsexual324–5,336legal405–6reproductiveisolation334metaphysical312,313reproductivetechnology495syntheticreduction175Rescher,N.785theoretical312–13responsibility535,536redundancytheories28revelation570,688–9referenceRey,G.188aspectualcharacter21rhetoric636–7943\nINDEXRich,Arthur540rule-scepticism409,410,421rights31,261–2,264Wittgenstein19,500,821abstractright745Ruse,M.338Bentham769Russell,Bertrand2,69,70,72,73,164,defenceof690–1712,790–801,811grouprights275atomism73Locke262,576discussionquestions803–4moralrights31empiricism70,361naturalrights416,502,769foundationalism369Nozick261–2,264identitystatements794–6andpublichealthpolicy511–12knowledgehypothesis428–9ofsovereigns675–6logicalatomism791violationof262logicism791right–leftdistinctions356mathematics361,362rigiddesignators112names161–2,794–6,797,798RioDeclaration526neutralmonism780RNA317no-classtheory362Robinson,Howard,Berkeley694–708non-conceptualism800Robinson,W.S.187publiclanguagetheory797Rodney,Walter572quantificationtheory792rolestheoryofdescriptions69,798–9,813definition388onthoughtandlanguage795normativeexpectations388Russell’sparadox361,791roletheory388,435–6Ryle,GilbertRolston,Holmes,environmentalethicsbehaviourism174–5517–30polarconceptargument49Romantradition636–7Rorty,Richard278,562SacredDoctrine627eliminativism193sagephilosophy572,641emotions188Sainsbury,R.M.,FregeandRussellneo-pragmatism783–4790–804philosophyofmind75Sale,Kirkpatrick525politicalphilosophy278sanctions764,765pragmatism775Sandbach,F.H.595RoscelinofCompiêgne628Sandel,Michael270,271Ross,GeorgeMacdonaldSartre,Jean-Paul847,860–4DescartesandMalebranche644–57anti-morality863SpinozaandLeibniz658–70atheism863Rousseau,Jean-Jacques870being-for-itself861,862masculinepsychicstructure554–5being-in-itself861,862sexism553consciousness860–2Rowe,W.482discussionquestions873–4rulesexistentialphenomenology863actingon19,214freedom862,863aestheticjudgement232mauvaisefoi862,863fidelitytolaw423nothingness861–2andhabitofobedience412Scanlon,Thomas221primaryandsecondaryrules412Scarre,G.767ruleofrecognition412,502scepticism38,45–56,683ruleutilitarianism765–6Academicsceptics45,46944\nINDEXanatomyof46naturalizedepistemology302–3ancientscepticism595,597observation295–7,776,777anti-scepticalarguments51–6,106pessimisticmeta-induction300–1,312Berkeley53–4theory–observationdistinction296characterizing45underdeterminationoftheorybyDescartes46,50–2,471,652observationaldata(UTD)297–300Dewey55unobservabilityproblem294earlyscepticism45–6science,metaphysicsof303–13error,delusionanddreams47causation303–5errorsceptic50coveringlawmodelofexplanationHumeanresponseto52,54,720304–5andjustification45lawsandaccidents305–6Kant52–3probability308–11Locke52,683teleology311–12methodological49–51theoreticalreductionism312–13perceptionandperceptualrelativitiesscience,philosophyof286–31647–9,51assimilationofmetaphysicsto76philosophicalscepticism20authorityof75problematic49–51Bayesianism290–3Pyrrhonians45,46discussionquestions315–16refutationofscepticalarguments49Enlightenment395,396scepticalepistemologyversusanti-fallibility10Cartesianism54–6historicalsciences321transcendentalarguments52–3instrumentalinterpretation64,65,355Wittgenstein19,20,55instrumentalismversusreason294–5Scherer,K.R.188mechanistic665Schiffer,S.117,118,119,123,124,195nomotheticsciences321Schiller,F.C.S.251,782–3observationalscience295–7,776,777humanisticpragmatism775,782–3post-positivistphilosophy10–11reality783probability308–11truth783researchethics513–14Schlegel,F.,Indianphilosophy570sociologyof395Schlesinger,G.465third-personperspective174scholasticism25,587,620–1,636,647seealsonaturalscience;science,casuistry494epistemologyof;science,metaphysicsofnaturalphilosophy636,647scientificmethod10,11,569,777SchoolofChartres620Bacon635–40Schopenhauer,Arthur814,828,835Peirce777,778science,epistemologyof286–303scientificrevolutions11,645anti-realismarguments294–5,296,scientism24300–1seealsopositivismBayesianism290–3ScottishEnlightenment716demarcationproblem299Scruton,Roger236,279empiricalstudy301–2Searle,J.R.falsificationism287–90,298,299computermodelofthemind181–2,183incommensurability297contemporaryUSphilosophy1–21inductionproblem286–7,289,294,illocutionaryacts8302–3secularizationofphilosophy569instrumentalismversusrealism294–5,Sedley,D.N.595296,301Sejnowksi,T.193945\nINDEXselectiveattention697sense–referencedistinction14,160,161,self793–4,795,799,800,801,844Nietzsche’srejectionof832senses48,112,378,684,699,701Wittgenstein’saccountof817sensibility729self-awareness389sentence-types148–9,163self-consciousness65,689,732,733,743,ambiguity164747,785,861assertorictokens157,793,794self-control207,208,258declarative148,149,157,158,160self-determination276,662,663,747single-expressionsentences164self-interest525,761subject–predicatesentences91,100self-knowledge385T-sentences17–18,105,107self-ownership261,262,264,274type–tokendistinction148selfhood832word-types149Sellars,Wilfrid848Serres,Michel525functionalism186settheory365,366representationalism188sexratioanalysis327,328semantics96,99–114sexism553,555,559,577axiomatizedsemantictheories101,102,SextusEmpiricus45,812110sexualjustice272businessethics533,534,540sexualreproduction324–5,336ConventionT105,107,126–7sexuality,Foucaulton867–8correctnessofsemantictheories104–7Shoemaker,S.177Davidson’sprogramme99–114Sidgwick,Henry769–70drivenbyscience71discussionquestions772–3andepistemology109–11egoism770finiteaxiomatizationconstraint110idealobservertheory470Fregeargument102–3,104intuitionism769,770indicatorsemantics191morality770intention-basedsemantics124utilitarianism210,211,214,769,770interpretationaladequacy107Sidney,SirPhilip635andmentalstates107–9,112SigerofBrabant623meta-semantics90,91,115,127–8,139Simmonds,N.E.,philosophyoflawandmetaphysics112–14403–27negation98,101Singer,Peter493questionsaboutmeaning90–1Sinnseesensesemantics–pragmaticsdistinction128Skinner,B.F.180theoryformat102–4Skorupski,John763,767truth-conditionalsemantics102,103,ethics202–30105,108,109,113,114slavery577,593,836–7Sen,Amartya,businessethics531,539,Smart,J.C.C.176,182,294540,542,543Smith,Adam570,720,721Senghor,Leopold572,575idealobservertheory470sensation185–7,700–1,729,813Smolensky,P.194assimilationofideasto701–2Sober,E.,philosophyofbiology317–44sensation-language819socialactionsenseanalysis376customary800Weber386indirect800Wittgenstein387names160,161,162socialchoices279sense-data5,54,130,361,807,818socialco-operation273946\nINDEXsocialcontracttheory492,519,525,546,soul650,651607Aristoteliantheory610–11environmentalethics525Cartesiantheory667–8Hobbes394,671,675ensoulment506andnaturalcontract525neo-Platonistview350originalposition519Platonictheory604–6Rawls9–10,492scholastic–Aristoteliantheory668socialinequality258,260–1,264,265transmigrationof75socialjustice31–2tripartitesoul605,606economicequality32sovereignty411,674economicredistribution32absolutemonarchy690socialmeanings271divinerightofkings690,691socialphilosophyseepoliticalandsocialrightsofsovereigns675–6philosophyspace729–30socialscience,philosophyof375–402Kant356–7discussionquestions401–2spatio-temporaldata52,53explaining–understandingcollaborationspecies333,334–5377,390–7biologicalspeciesconcept334,335explanation376,377–84chemicalelementsasparadigmnaturalholism376,383,386–7,389,398kinds333individualism379,389historicalentities335naturalism377–84,390individuality335rationalityandgamestheory391–5monophyleticgroups335rationalityandrelativism395–7naturalkinds333socialroles388scholastictheoryof653structuralaccount379speciesism507understanding376,384–90speechacts116socialityofhumanbeings563Austin’stheoryof6,7–8societycontext-dependence137conflictmodel394descriptiveutterances9consensusmodel394evaluativeutterances4–6,9‘opensociety’258Grice’stheoryof15–16sociobiology336illocutionaryacts8Socrates69,493,494,586–90indirectspeechacts16akrasia605–6mood136–7dialecticalmethod586–90performative–constativedistinction7–8elenchos587–90,606recognition136moralmotivation607truthconditions137Platonicdialogues587,590,591,602,utteranceinterpretation139603,605,606–7Sperber,Dan,pragmatictheory132–3,136Platonicimage587Spinoza,Benedictde658–63Socraticmethod587–90AttributeofExtension660streetphilosophy587,588AttributeofThought660solipsismdiscussionquestions670ego-based819God658–9Wittgenstein’scritique817,819,847humannature661–3Solomon,R.188,536mentalphenomena660Sophists493,494metaphysics658–9Sorensen,R.A.461onthemind661Sorrell,Tom,Hobbes671–81Nature659–61,664947\nINDEXSpinoza,Benedictdecont.subjectterms90,93,98,112pantheism658–9subject–predicatedistinction98,99,686rationalism661–2subject–predicatesentences91,100spiritsubjectivismBerkeley705aesthetic234,235–6,250Hegel744–6ethical338–9natureof65,744–6meta-ethical492andthestate745–6transcendentalsubjectivity845,847,848stakeholdertheory541,545–6substanceStalinism258Locke686,687,688‘standingfor’160seealsoessence;naturalphilosophystatesubstitutivityauthority33ofidenticals348,800culturalpluralism275salvaveritate800,801defined276substitutivityprinciples800,801expectationofsafetyfrom677substratum686,699,700Hobbes671,674–5successcriteria785interventionistrole263suicide508minimalstate262–3,264assisted508nation–statedisassociation277physician-assisted508Nozick262–3summumbonum862obedienceto760superstrings350obligationsofsubjects675–6supervenience80–1,318pluraliststate32Swinburne,Richard482rightsofsovereigns675–6foreknowledge467role263positivism455secularconception674theism462spiritand745–6syllogism640,744,767stateownership720synechism776transferoftherightofnatureto674synonymy6,160statisticalinference,problemof310syntacticapproachtologicalpossibilityStephen,J.F.762155,158–60Stephens,C.337synthesis732Sterelny,K.327stewardship,ethicsof520,527T-sentences17–18,105,107Stich,S.Taliaferro,Charles,philosophyofreligioninstrumentalism192453–89pragmatism784Talmud576Stoicism595,596Tarski,Alfred104Strawson,P.F.53,118,119,187semantictheoryoftruth17,785Kant733Tarski–Davidsontruththeories110,126logic147,150,163taste232,234materialbodies114Hume234,718metaphysics61Kant235names163Taylor,Charles275philosophyoflanguage98,99Taylor,R.477propositions150technology,Heideggeron854–5Stroud,Barry20teleologicalcharacterizations184structuralism863teleologySuárez,Francisco494,621Aristotle593–4948\nINDEXbiologicalteleology184–5,319–20transcendence470,731existenceofGod478–80TranscendentalAesthetic729homuncularfunctionalism182–5TranscendentalAnalytic730,731metaphysicsofscience311–12TranscendentalDeduction732,733teleosemantics190–1TranscendentalDialectic730Tempels,Placide,Bantuphilosophy572TranscendentalLogic729,730Tertullian622transcendentalunityofapperception733Thales569Trigg,Roger459theism184,460,462,473,474TrinitarianChristianity747definition456tritheism456metaphysicsof458true–falsedichotomy150,165,166,813problemsofevil480–3truth12,28,91,95,640–1seealsoexistenceofGodabsolutetruth780,781theisticvoluntarism464,468–9Bacon640–1theodicy481,482Christiancriterion622theologicalphilosophy64,74,623concretetruths783casuistry494–5correspondencetheoryof226,457–8naturaltheology627Dewey781seealsomedievalphilosophy;religion,disquotationalconceptionof126,127philosophyofelucidation107theologicalworldview79historicaltruth571TheoryofForms592,602–4,609informativetruths640aestheticforms604James780–1epistemologicalrole603mathematical361,777ethicalforms604meaning–truthconnection126,127ontologicalstatus603metaphysicaltruths463–4separationofForms608minimalismabouttruthforpropositions‘theory’theory194,196126–8thoughts,intentionality96,97,703minimalisttheoryof28Tiedemann,Dietrich569natureof38Tiles,Mary,philosophyofmathematicsnecessarytruth38345–74Peirce779Tilghman,B.R.458practicality641time62,82–5,356–7redundancytheories28A-series83–4,85relativetruth781Aristotle82semanticdefinitionof17awarenessof82Tarski–Davidsontruththeories110B-series83–4,85truth-conditionaltheoriesofmeaningandchange82–316–18Kant356–7universal561,562measuringthepassageof82truth-conditionalsemantics102,103,105,tenselesstheoryof84,85108,109,113,114tokenidentity178,313truth-functions361tokenmaterialism178Turingmachinefunctionalism13token–tokenidentitytheory13TuringTest175,176,181toleration690,719twinearthexample15,187Tolstoy,LeoNikolayevich251tychism776totalitarianism607,663Tye,M.177Touré,Sékou572typeidentity178–9,817tracking44type–typeidentitytheory13,77949\nINDEXunderdeterminationoftheorybywell-being209,210,211,212,213,observationaldata(UTD)297–300221–2understandingutilityprinciple762,763configurational446,447proofof767–8empathy386utopianism274andexplanation377,390–7explanatory386vacuum649,660interpretative377,385value72,759,761Kant729cognitive793Unger,Roberto423–4ethical203,209–10,211,212,213,UnitedNationsWorldCommissionon218,734EnvironmentandDevelopment526labourtheoryof273UnitedStatesMill762Americancorporations536moral734,735businessethics542,543naturalistictheoryof78contemporaryphilosophy1–21non-ethicalformsof212ethicalrationalism492pluralityof214–16universalmoralstandards271utilitarianism762universalism561,562,567valuejudgements4–6,440–1,442,falseuniversalism557443andfeministphilosophy558cognitivestatus440universals73,79,96,98,608non-cognitivestatus440Aristotle608valuemonism268problemof73,621,628–9,630,valuepluralism259,260,267,268698VanGulick,R.184realist603verification3,6,11,174,192,781semantic603intersubjectiveverification174,176Urmson,J.O.765verificationism455utilitarianism30,31,208–12,218,493,empirical454,455612,806historical431aggregrativeprinciple221mystical455Bentham391–2,759–61,768–9vices206–7classical210,211,212,214,221ViennaCircle2,66,454,457,812,818cogent218virtue206,224,377,611–12consequentialism760acquisition269conservativeholism213,214,222‘artificial’717disciplineandpunishment425liberalvirtues269ethicalvaluetheory209,210,212,213,moralvirtues611,612,613218‘natural’717generic212,213,214,221,222virtuesofcharacter611happinessseehappinessvirtuesofintellect611idealutilitarianism806virtueethics27,30,31,500,546indirectutilitarianism765–6vitalism184,318,664,665intuitionistfoundation770Vitoria,Franciscode621law415,425Vlastos,Gregory,Socraticmethod588–9,Mill211,761–3,764,765–6,767–8590personalethic764Voltaire570self-effacing213voluntarism224Sidgwick210,211,214,769,770theistic464,468–9standards213Vossius,JohannesGerardus568–9950\nINDEXWalsh,W.H.448Wittgenstein,Ludwig2,18–20,23,66,Walzer,Michael265,277399,500,811–26Wang,Hao346accountoftheself817Wardy,Robert,ancientGreekphilosophyactiontheory387585–600aestheticjudgements236warfare491,494art237coldwar677biography811–12Hobbes671,672–3,674BlueBook816–17justwardoctrine495discussionquestions825–6Warnock,Mary493lifeprocesses318wavefunction310,311logic,philosophyof168–9,814Weber,Maxlogicalatomism814,815,818causalpropositions390–1logicalnecessity813socialaction386mathematics364,368–9,822Wertrationalität389onmeaning816–17,820Weierstrass357Moore’sParadox808–9Weldon,T.D.258philosophicalpsychology19welfareegalitarianism265,266philosophicalscepticism20welfarism212philosophyofmathematics822well-being209,210,211,212,213,privatelanguageargument19,55,221–2817–20,847West,M.L.571propositions814,815Westernphilosophicalcanon568rejectionoffoundationalism6,8–9,55ethnocentrism576rulesandrulefollowing19,500,821Eurocentrism576–7scepticism19,55extendingtheroleofphilosophy577–8solipsism817,819,847origins568–71,585theoryoflanguage8–9,814,815,816relationtoreligion569,570TractatusLogico-Philosophicus66,359,sectarianism569812,815White,M.434,435,444truthsandbeliefs457–8Wiggins,D.106,113,126Wollheim,R.,onart244,248,252wildernessareas525Wollstonecraft,Mary553will219,734Wolterstorff,N.467weaknessof605–6Woolhouse,R.S.,Locke682–93WilltoBelieve779work281willtopower828,834,835,836,837,worldpictures853–4838,853Cartesian645–8WilliamofOckham621,627common-senseview648individuation630religious65,79,472Ockham’srazor298,472,629Wright,C.362universals629semantictheories111Williams,BernardWright,L.,biologicalteleology319,320appliedethics493Wright,Sewall326contemporaryphilosophy23–34Wykstra,S.482moralcategoricity216Yandell,K.473Williams,G.C.326,327,338Williamson,O.E.536ZenoofElea354Wilson,Deidre132–3,136paradoxesofmotion62Wilson,E.338Zermelo–Fraenkelaxiom366,367Wiredu,Kwasi572Zoroastrianism571951\nREVELATION\n