【剑桥哲学指南】皮尔士

【剑桥哲学指南】皮尔士

cherylmisak1CharlesSandersPeirce(1839–1914)1.introductionCharlesSandersPeircewasthefounderofpragmatism–theviewthatourtheoriesmustbelinkedtoexperienceorpractice.Hisworkisstaggeringinitsbreadthandmuchofitliesinahugebulkofmanuscriptsandscraps.Hisfewpublishedpapersincludethoseofthe1870sseriesinPopularScienceMonthlycalled“IllustrationsoftheLogicofScience,”mostnotably“HowtoMakeOurIdeasClear”and“TheFixationofBelief.”HisLowellLecturesin1898and1903andhisHarvardPragmatismLecturesin1903alsocontainessentialmaterial.Butmuchofwhatisimportantisonlynowbeingpublishedinthedefinitivechronologicaledition:TheWritingsofCharlesSandersPeirce.Peircewasadifficultmanandthiswasnodoubtpartlyrespon-sibleforhisbeingfrozenoutofwhathemostdesired:apermanentacademicposition.1HeworkedinsteadfortheU.S.CoastSurvey–hisscientificandmathematicalendeavorstherehadasignificantinfluenceonhislogic,onhisworkinstatisticalinference,andonhisepistemologyandmetaphysics.Heisperhapsbestknowntodayforhistheoryoftruthandhissemeiotics,aswellasforhisinflu-enceonWilliamJamesandJohnDewey.Butbecauseofthescat-terednatureofhisworkandbecausehewasalwaysoutoftheaca-demicmainstream,manyofhiscontributionsarejustnowcomingtolight.AsPhilstrom’sessayinthisvolumemakesclear,oneofthemostimportantinfluencesonPeircewasKant.Thereisalsoastronggustofmedievalphilosophyblowingthroughouthiswrit-ing.ItisfromherethatPeircegetshisScholasticrealism,whichis1CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n2cherylmisaksetagainstthenominalismoftheBritishempiricists.(SeeBoler’scontributiontothisvolume.)Buttherearealsoclearaffinitiesbe-tweenPeirceandtheBritishempiricists.Forinstance,Peircecred-itsBerkeley’sargumentsthatallmeaningfullanguageshouldbematchedwithsensoryexperienceastheprecursorofpragmatism:Berkeleyonthewholehasmorerighttobeconsideredtheintroducerofpragmatismintophilosophythananyotheroneman,thoughIwasmoreexplicitinenunciatingit.2Ithasseemedtomanythat,despitePeirce’sclaimstobeputtingto-getheragrand‘architectonic’system,therearesubstantialtensionsinhiswork.Goudge(1950)declaredthatthereweretwoincompati-blePeirces.Oneisahard-headedepistemologist/philosopherofsci-enceandtheotherisasoft-headedreligiousthinkerpronetometa-physicalspeculation.MisakandAndersonargueinthisvolumethatthetwoPeircescanandoughttobebroughttogether.WhetherornotPeirce’sworkcanbebroughtintoaharmoniouswhole,thereaderofthiscollectionwillbestruckbytheenormousrangeofdebatestowhichPeircewasaseriouscontributor.Inthisintroductoryessay,awhirlwindtourofthosecontributionswillbeconducted.32.thepragmaticmaximPeircetookthe‘spirit’ofpragmatismtobecapturedinthefollow-ingmaxim:“wemustlooktotheupshotofourconceptsinorderrightlytoapprehendthem”(CP5.4).Thereisaconnectionbetweenunderstandingaconceptandknowingwhattoexpectifsentencescontainingtheconceptweretrueorfalse.Ifaconcepthasnosuchconsequences,thenitlacksanimportantdimensionwhichwewouldhavehadtogetrightwerewetofullyunderstandit.ThiscriterionoflegitimacyliesattheheartofPeirce’swork.Notonlydoeshecriticisecertainphilosophicalpositionsaspragmati-callyspurious,buthearrivesatmanyofhisownviewsbyfocussingontheconsequencesof,say,“Pistrue”or“xisreal.”Thepragmaticmaxim,thatis,servesbothasastandardfordeterminingwhichex-pressionsareemptyandasamethodologicalprincipleforformulat-ingphilosophicaltheoriesoftruth,reality,etc.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)3In“HowtoMakeOurIdeasClear,”Peircepublicallyunveilsprag-matismandsetsoutthemaximasfollows:Considerwhateffects,whichmightconceivablyhavepracticalbearings,weconceivetheobjectofourconceptiontohave.Then,ourconceptionoftheseisthewholeofourconceptionoftheobject.(W3,266)Peircesuggestsinthispaperthatknowingthemeaningofanex-pressionisexhaustedbyknowingits“practical”effects,whichhecharacterizesas“effects,directorindirect,uponoursenses”(W3,266).Theseeffectscanbedescribedbyconditionalsofthesort:ifyouweretodoA,youwouldobserveB.Hesays:Wecomedowntowhatistangibleandpractical,astherootofeveryrealdistinctionofthought,nomatterhowsubtileitmaybe;andthereisnodis-tinctionofmeaningsofineastoconsistinanythingbutapossibledifferenceofpractice.(W3,265)Asanexampleofhowthepragmaticmaximoperates,Peirceex-aminesthemeaningof“thisdiamondishard.”Hesaysthatitmeansthatifyoutrytoscratchit,youwillfindthat“itwillnotbescratchedbymanyothersubstances”(W3,266).Noticethatthepracticaleffecthereisformulatedasanindicativeconditional,asamatterofwhatwillhappen.Peirceseesthatifheformulatespracticaleffectsinthismanner,itmakeslittlesensetodescribeadiamondwhichisinfactneverscratchedasbeinghard.Heseemstobecontentwiththisconclusionin“HowtoMakeOurIdeasClear.”Butwhenheconsidersthematterlater,heinsistsonasubjunctiveformulation.Hechideshimselfformakingthenom-inalistsuggestionthathabits,dispositions,or“would-bes”arenotreal.AScholasticrealismaboutdispositionsandsubjunctivecondi-tionalsmustbeadopted:adispositionismorethanthetotalofitsrealizationsandasubjunctiveconditionalcanbecorrectorincor-rect,whetherornottheantecedentisfulfilled.Thepracticaleffectswhichconcernpragmatismarethosewhichwouldoccurundercer-tainconditions,notthosewhichwillactuallyoccur.Hisconsideredviewabouttheunscratcheddiamondisthat“itisarealfactthatitwouldresistpressure”(CP8.208).ThiswasnotPeirce’sonlyamendmenttothepragmaticmaxim.Inhisstruggletoarriveatasuitableaccountofunderstanding,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n4cherylmisakwesometimesfindhimsuggestingsomethingverysimilartowhatwefindlaterinlogicalpositivism.Thepositivists’criterioneffec-tivelyrestrictedmeaningtostatementsaboutphysicalobjects–tostatementsaboutthatwhichisdirectlyobservableorverifiable.Statementsaboutanythingelse–metaphysicsorethicsforexam-ple–wereliterallymeaningless.But,infurtherimprovementstothepragmaticmaxim,Peircemakesitclearthatheisconcernedtogiveamuchmoregenerousaccountofwhatisinvolvedinunderstanding.First,Peircehimselfinclinedtowardmetaphysicsandhedidnotwanttodoawaywithitaltogether.Inmetaphysics“onefindsthosequestionsthatatfirstseemtooffernohandleforreason’sclutch,butwhichreadilyyieldtologicalanalysis”(CP6.463).Metaphysics,“initspresentcondition,”is“apuny,rickety,andscrofulousscience”(CP6.6).Butitneednotbeso,formanyofitshypothesesaremean-ingfulandimportant.Itisthejobofthepragmaticmaximtosweep“allmetaphysicalrubbishoutofone’shouse.Eachabstractionisei-therpronouncedtobegibberishorisprovidedwithaplain,practicaldefinition”(CP8.191).Second,Peircefrequentlyclaimsthatthepragmaticmaximcap-turesonlyapartofwhatitistoknowthemeaningofanexpression.Inordertograspaterm,heargues,athreefoldcompetenceisrequired.Theinterpretermustbeableto(1)pickoutwhatobjectsthetermreferstoorknowtheterm’sdenotation,(2)giveadefinitionofthetermorknowtheterm’sconnotation,and(3)knowwhattoexpectifhypothesescontainingthetermaretrue.Hetakesthesethreeaspectsofunderstandingtospelloutcompletelywhatsomeonemustbeabletodoifshegraspsaconceptorknowsthemeaningofanexpression.Amuch-neglectedimplicationofthisviewisthatdefinitionisnotthemostimportantprojectforphilosophers:“DefinitioncannolongerberegardedasthesuprememodeofclearApprehension”(MS647,p.2).Thatis,wemustbealerttothefactthatwhatPeircearrivesatwhenheappliesthepragmaticmaximtoaconceptisnotCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)5adefinitionoftheconcept,butrather,apragmaticelucidation.Heexaminesaconceptthroughitsrelationswithpracticalendeavors.Thatisoneroutetounderstandingaconcept,theroutePeircetakesashisowncontributiontodebatesaboutwhatitistounderstandsomething.Third,Peircetriestodivertthephilosopherfromthinkingthatsensoryexperienceisall-important.Aperceptualbelief,heargues,ismerelyabeliefthatiscompelling,surprising,impinging,unchosen,involuntary,orforceful.Suchbeliefsneednotarisefromthesenses.Peirce,unlikehisverificationistsuccessors,wantsallhypothesestobeexposedtothepragmaticmaxim;hedoesnotexemptformal(or“analytic”)sentences.Logicalandmathematicalhypothesescanmeetthecriterionbecausethereisakindofexperiencerelevanttothem–youcanmakemanipulationsinproofsordiagramsandob-serveunexpectedresults.Andsomemetaphysicalhypothesesmeetthecriterionaswell.Theymusthaveconsequences,Peirceargues,forordinary,everydayexperience.SeethecontributionsherefromWigginsandMisakforadiscussionofhowmathematicsandmoralsfitinthispicture.3.truthandrealityPeirceappliesthepragmaticmaximtothedebateonthenatureoftruthandreality.Thephilosophermustlooktoourpracticesandseewhataccountoftruthwouldbebestsuitedforthem:“Wemustnotbeginbytalkingofpureideas,–vagabondthoughtsthattrampthepublicroadswithoutanyhumanhabitation,–butmustbeginwithmenandtheirconversation”(CP8.112).AsWiggins’sessayinthisvolumemakessoclear,theupshotisasubtleandcompellingview.Peirce’sroutetotheconceptoftruthisthroughbelief,inquiry,anddeliberation:thepracticeslinkedtotruthandtotheseekingoftruth.Peircesuggeststhatweconcernourselveswithpropositionswehavearrivedat,expressed,affirmed,orbelievedandthoseweshallarriveat,express,affirm,orbelieve.4Bymakingthisourfocus,wewilldiscoversomethingaboutwhatitisatwhichweaim:truth.Thisdoesnotmeanthattruthisanepistemologicalnotion.Rather,thisexemplifiesoneroutetofindingoutsomethingabouttruth:theroutethroughourepistemologicalpracticesofbelieving,inquiring,anddeliberating.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n6cherylmisakThecorrespondencetheory,Peirceargues,canhavenoconse-quencesforourpractices.Itholdsthatatruehypothesisisonewhichisinagreementwithanunknowable“thing-in-itself.”But:Youonlypuzzleyourselfbytalkingofthismetaphysical“truth”andmeta-physical“falsity”thatyouknownothingabout.Allyouhaveanydealingswithareyourdoubtsandbeliefs....Ifyourterms“truth”and“falsity”aretakeninsuchsensesastobedefinableintermsofdoubtandbeliefandthecourseofexperience...wellandgood:inthatcase,youareonlytalkingaboutdoubtandbelief.Butifbytruthandfalsityyoumeansomethingnotdefinableintermsofdoubtandbeliefinanyway,thenyouaretalkingofentitiesofwhoseexistenceyoucanknownothing,andwhichOckham’srazorwouldcleanshaveoff.Yourproblemswouldbegreatlysimplified,if,insteadofsayingthatyouwanttoknowthe“Truth,”youweresimplytosaythatyouwanttoattainastateofbeliefunassailablebydoubt.(CP5.416)Peirce’sthoughthereisthatifoneofferedanaccountof“Pistrue”intermsofitsconsequencesfordoubt,belief,andperceptualdisappointment,onewouldbeofferingapragmaticelucidationoftruth.That,ifitwereacorrectspecificationoftheconsequences,wouldtellussomethingabouttruth.Butadefinitionoftruthwhichmakesnoreferencetobelief,doubt,andexperienceisempty.Itisameredefinition–usefulonlytothosewhohaveneverencounteredthenotionoftruth.Peircesometimesstatesthisobjectiontothecorrespondencethe-orybylabelingita“transcendental”accountoftruth(CP5.572).Suchaccountsregardtruth“asthesubjectofmetaphysicsexclu-sively”–spuriousmetaphysics,notpragmaticallylegitimatemeta-physics.Onthecorrespondencedefinition,truthtranscends(andthushasnoconsequencesfor)belief,experience,andinquiry.Hesays:TheDingansich...canneitherbeindicatednorfound.Consequently,nopropositioncanrefertoit,andnothingtrueorfalsecanbepredicatedofit.Therefore,allreferencestoitmustbethrownoutasmeaninglesssurplusage.(CP5.525)Thecorrespondencetheoryhasitthatthereisanunbridgeablegapbetweenabeliefwhichissupportedbyexperienceandabeliefthatcorrespondstoreality.Wecouldhavethebestpossibleevidenceforahypothesisandyetthathypothesismightfailtobetrue.TheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)7correspondencetheorydoesnottelluswhatwecanexpectofatruehypothesisandsoitisnotcapableofguidingusinouractionsandinquiries.Iftruthistheaimofinquiry,thenthecorrespondencethe-oryleavesinquirerscompletelyinthedarkastohowtheyshouldconducttheirinvestigations.Theaimisnot,Peircesays,“readilycomprehensible”(CP1.578).Howcouldanyoneaimforasortoftruththattranscendsexperience?Howcouldaninquirercomeupwithameansforachievingthataim?Inanticipationofcertainkindsofminimalistaccountsoftruth,Peircefocusesonwhathethinksthetranscendentalisthaslostsightof–theunseverablelinkbetweentruthontheonehandandasser-tion(andbelief)ontheother.ToassertPistoassertthatPistrueandtoassertthatPistrueistoassertP.(Alternatively,tobelievePistobelievethatPistrueandtobelievethatPistrueistobelieveP.)Thenotionoftruthisboundupwiththenotionsofassertionandbelief.ButPeircetakesastepfurtherthantheminimalist.Onceweseetheinternalconnectionbetweentruthandassertion/belief,wemustlooktothepracticeofassertion/beliefandtothecommitmentsincurredinit,sothatwecansaysomethingmore.Whatweknowabouttruthisthatitiswhatweaimatwhenweassert,believe,ordeliberate.Werewetoforeverachieveallofourlocalaimsinasser-tion,belief,anddeliberation(prediction,explanatorypower,andsoon),thenthebeliefinquestionwouldbetrue.Thereisnothingoverandabovethefulfillmentofthoselocalaims,nothingmetaphysical,towhichweaspire.Werewetogetabeliefwhichwouldbeasgoodasitcouldbe,thatwouldbeatruebelief.Peircesumsupthematterthus:“Atruepropositionisapropo-sitionbeliefinwhichwouldneverleadto...disappointment”(CP5.569).Thisisanaccountofwhatwecanexpectofatruebelief:ifweweretoinquireintoP,wewouldfindthatPwouldencounternorecalcitrantexperience.Wecanpredictthatifwewerediligentlytoinquire,itwouldnot,intheend,beoverturnedbyexperienceorargument.Analternativewayofmakingthepointistosaythatwewouldexpectthefollowing:ifinquirywithrespecttoPweretobepursuedasfarasitcouldfruitfullygo(i.e.,farenoughsothatthehypothesiswouldnolongerbeimprovedupon),Pwouldbebelieved.Atruebeliefisapermanentlysettledorindefeasiblebelief.Peirce’sviewofrealityisconnectedtohisviewoftruthinthatheoftensaysthatrealityisthe“object”oftruebeliefs–itiswhatCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n8cherylmisaktruebeliefsareabout.ChrisHookwayhasrecentlyimprovedourunderstandingofhowPeircesawthisconnectionandthereaderisadvisedtoturntohiscontributiontothisvolumeforasummaryofthatnewunderstanding.4.semeioticsPeircewasapioneerinsemeiotics.Notonlyisheresponsibleforthedistinctionbetweentype(‘human’asageneralterm)andtoken(‘hu-man’asappliedtovariousindividuals),buthedevelopedacomplexmapofsixty-sixkindsofsigns,fromwhichsprout59,049varieties.Thedetailsofthismaparestillofgreatinteresttosemeioticians,buttheywillnotconcernmehere.Short’sandSkagested’spapersinthisvolumeconveymanyoftheimportantpoints.ShortshowshowPeirceeventuallyabandonedhisearlytheoryofsignsandsub-stitutedforitamuchlessparadoxicaloneandSkagestedshowshowPeirce’stheoryofsignsconnectstoissuesaboutintentionalityandthephilosophyofmind.ItisimportanttonoticeforthisbroadoverviewofPeirce’sworkthathistheoryofsignshasinterpretationatitscenter.Peirceholdsthatthesign–referentrelationisnotable,onitsown,tosustainacompleteaccountofrepresentation.Representationistriadic:itinvolvesasign,anobject,andaninterpreter.EachaspectofthisrepresentationrelationcorrespondstooneoftheelementsinPeirce’sdivisionofsignsintoicons,indices,andsymbols.Andineachofthese,oneoranotheraspectofthelinguisticcompetencealludedtoinSection2ismostprominent.Iconsaresignsthatexhibittheirobjectsbyvirtueofsimilarityorresemblance.Aportraitisaniconofthepersonitportraysandamapisaniconofacertaingeographicalarea.Peircearguesthatthemeaningoficonicsignsliesmostlyintheirconnotation:whatmakesapaintingoramapaniconisthatitsqualitiesorattributesresemblethequalitiesorattributesofitsobject.Indicesaresignsthatindicatetheirobjectsinacausalmanner:anindex“signifiesitsobjectsolelybyvirtueofbeingreallyconnectedwithit”(CP3.360).Asymptomisanindexofadiseaseandsmokeisanindexoffire.Theessentialqualityofanindexisitsabilitytocompelattention.Apointingfinger,aknockonthedoor,oraCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)9demonstrativepronoun,suchas‘there’or‘that,’drawsattentiontoitsobjectbygettingtheinterpretertofocusontheobject.Soanindex,bybeingobject-directed,hasitsdenotationorextensionasits“mostprominentfeature”(CP8.119).Anindexpicksoutorindicatesitsobject;itpointsto‘that,that,andthat’asitsextension.Asymbolisaword,hypothesis,orargumentwhichdependsonaconventionalorhabitualrule:asymbolisasign“becauseitisusedandunderstoodassuch”(CP2.307).Symbolshave“principle”orpragmaticmeaning;theyhave“intellectualpurport.”Peircecontrastspragmaticmeaningwith“internal”meaning(whichherelatestoiconsandconnotation)andwith“external”meaning(whichherelatestoindicesanddenotation).Hesuggeststhatthepragmaticmeaningofsymbolshastodowitha“purpose”(CP8.119).Asymbolhaspragmaticmeaningbecauseiftheuttererknowshowinterpretershabituallyinterpretasign,shecanusethesigntocauseaspecificeffectintheinterpreter.AndPeircecallsthiseffectthe“interpretant”ofthesign.If,forinstance,Iwrite‘dog,’Iintendthesigntocauseacertaineffectintheinterpreter(perhapsIwanttheinterpretertothinkofadog),whereasifIwrite‘odg,’Idonot,as‘odg’isnotaconventionalsign.OrifIassert‘Thatbridgehasalooseplank,’Imightwanttheinterpretertobecarefulwhencrossingthebridge.Peircecharacterizesanassertionastheattempttoproduceadispositioninaninterpreter;itis“thedeliberateexer-cise,inutteringtheproposition,ofaforcetendingtodetermineabeliefinitinthemindofaninterpreter”(NE4,249).Noticethatifpragmaticmeaningisaboutthissortofeffect(hav-inganeffectonthebeliefsoftheinterpreter),itisnolongerabout“effects,directorindirect,uponoursenses.”Pragmaticmeaning,rather,involvesconsequencesforactionorthought.In1905wefindPeirceofferingthisversionofthepragmaticmaxim:Theentireintellectualpurportofanysymbolconsistsinthetotalofallgeneralmodesofrationalconductwhich,conditionallyuponallthepossibledifferentcircumstancesanddesires,wouldensueupontheacceptanceofthesymbol.(CP5.438)Peircethinksthat“rationalconduct”willeventuallymanifestit-selfinamodificationoftheinterpreter’sdispositiontobehave.And“rationalconduct”includestheconductofone’sthought.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n10cherylmisakThistwistinthepragmaticmaxim–thattheacceptanceofahypothesismusthaveeffectsonaninterpreter’strainofthought–coincideswithadevelopmentintheearly1900sinPeirce’stheoryofsigns.HerePeircearrivedatacomplextheoryofinterpretantsandhelocatespragmaticmeaningwithinthistheory.Hedistinguishesthreetypesofinterpretants.The“immediate”interpretantisthefitnessofasigntobeunderstoodinacertainway;the“dynamical”interpretantistheactualeffectasignhasonaninterpreter;andthe“final”interpretantistheeffectwhicheven-tuallywouldbedecidedtobethecorrectinterpretation.Pragmaticmeaning,Peircesays,liesinakindofdynamicalinterpretant:the“ultimatelogicalinterpretant”.Asign,Peirceargues,sparksasub-sequentsign(aninterpretant)inthemindoftheinterpreter,andsinceaninterpretantisitselfasign,aninfinitechainofinterpre-tation,development,orthoughtisbegun.Peircestopstheregressbyintroducingthenotionofan“ultimatelogicalinterpretant”ora“habit-change”.HefollowsAlexanderBainintakingabelieftobeahabitoradispositiontobehave.Andsothisnewhabitisabe-lieforamodificationoftheinterpreter’stendenciestowardsaction.Thepragmaticmeaningofanexpression,accordingtoPeirce’sthe-oryofsigns,istheaction(whichincludestheactionofsubsequentthought,andwhichendsinadispositiontobehave)thatarisesafteraninterpreteracceptsit.5.theoryofinquiryThenotionofinquiryoccupiesacentralplaceinPeirce’sthought.Philosophy,heinsisted,mustgetalongwithotherbranchesofin-quiry.Indeed,thefollowingmotto“deservestobeinscribeduponeverywallofthecityofphilosophy:Donotblockthepathofin-quiry”(CP1.135).In“TheFixationofBelief,”Peircecharacterizesinquiryasthestruggletoridourselvesofdoubtandachieveastateofbelief.Aninquirerhasabodyofsettledbeliefs–beliefswhichare,infact,notdoubted.Thesebeliefs,however,aresusceptibletodoubt,ifitispromptedbysome“positivereason,”suchasasurprisingexperience(CP5.51).WehaveseenthatPeircetakesexperiencetobethatwhichimpingesuponus–experience,hesays,teachesus“bypracticalCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)11jokes,mostlycruel”(CP5.51).Whenexperienceconflictswithaninquirer’sbelief,doubtisimmediatelysparked.Anddoubt“essen-tiallyinvolvesastruggletoescape”(CP5.372n2).Inquiryisthatstruggletoregainbelief.Thepathofinquiryisasfollows:belief–surprise–doubt–inquiry–belief.Peircedoesnottakethesepointstobemereobservationsabouthumanpsychology;hethinksthatpsychologyshouldbekeptoutoflogicandthetheoryofinquiry.Doubtandbelief,althoughtheydohavepsychologicalaspects,suchasmakingtheinquirerfeelcom-fortableoruncomfortable,arebestthoughtofintermsofhabits.A“belief–habit”manifestsitselfinanexpectation:ifwebelieveP,thenwehabituallyexpecttheconsequencesorthepredictionswederivefromPtocomeaboutwhentheappropriateoccasionarises.Inquir-ersarethrownintodoubtwhenarecalcitrantexperienceupsetsordisruptsabelieforexpectation.Therearethreestancesaninquirermayhavewithrespecttoahypothesis:believeit,believeitsnegation,orconsiderthematteropentoinquiry.Onlyinthethirdstanceareweleftwithoutahabitofexpectationandthusitisagnosticism,whichistheundesirablestate.Thatis,doubtingwhetherahypothesisistrueisnotequivalenttobelievingthatitisfalse–rather,doubtingisnotknowingwhattobelieve.Whatiswrongwiththisstateisthatitleadstoparaly-sisofaction.Aninquirerhassomeendinview,andtwodifferentandinconsistentlinesofactionpresentthemselves,bringingthein-quirertoahalt:“hewaitsattheforkforanindication,andkickshisheels...Atruedoubtisaccordinglyadoubtwhichreallyinterfereswiththesmoothworkingofthebelief-habit”(CP5.510).Doubtisnotknowinghowtoact.Andaction,forPeirce,includesactionindiagrammaticandthoughtexperiments.Peirce’stheoryofinquiryhasacertainkindofempiricismatitscore.Inquirersaimforbeliefsthatfitwithexperience,inPeirce’sbroadsenseofthatword.Whenwereplaceabeliefwhichhascomeintodoubt,thatnewbeliefstandsuptoexperiencebetterthantheoldone.Soweacceptit,actonit,andthinkforthetimebeingthatitistrue.Butweknowverywellthatiteventuallymightbeover-thrownandshownbyexperiencetobefalse.Peirceaddsthemorecontentiousclaimthatwhatweaimforispermanentlysettledbe-lief.WhenwehaveabeliefthatwouldforeverwithstandthetestsCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n12cherylmisakofexperienceandargument,hearguesthatthereisnopointofre-fusingtoconferuponitthetitle“true.”Onlyaspuriousdesirefortranscendentalmetaphysicswillmakeonewanttodistinguishper-fectlygoodbeliefsfromtruebeliefs.Butin“TheFixationofBelief”Peircesaysthatapermanentlyfixedbelief,nomatterhowitisfixed,istrue.Aproblemofcourseloomslargehere.Ifbeliefscouldbesettledbyareligiousauthority,orbyacharismaticguru,orbyastrology,sothattheywereperma-nentlyresistanttodoubt,hisaccountwouldgiveusnoreasonforcriticisingthem.Peircetrystosolvethisproblembyconsideringvar-iousmethodsoffixingbeliefandarguingthatitishardreallytoendtheirritationofdoubt.Themethodoftenacity,orholdingontoyourbeliefscomewhatmay,willnotwork,hesays,becausedoubtwillbesparkedwhenonenoticesthattheopinionsofothersdifferfromone’sown.Be-liefsproducedbythemethodofauthority(fixingbeliefsaccordingtothedictatesofastateorreligion)willsimilarlybesubjecttodoubtwhenonenoticesthatthoseinotherstatesorreligionsbelievediffer-entthings.Beliefsproducedbytheapriorimethod(adoptingbeliefswhichareagreeabletoreason)willeventuallybedoubtedwhenitisseenthatwhatwetakeasbeingagreeabletoreasonshiftslikeapendulumandisreallyamatterofintellectualtaste.Noneofthesemethodswillproducepermanentlysettledbeliefbecausetheyhaveaself-destructivedesign:thebeliefssettledbythemeventuallywouldbeassailedbydoubt.TheagentofdestructionwhichPeirceseesineachofthespeciousmethodsseemstobeapurportedfactaboutourpsychologicalmakeup:ifaninquirerbelievesahypothesis,andnoticesthatotherinquirersdonotbelieveit,thatfirstinquirerwillbethrownintodoubt.Thisimpulse,Peircesays,is“toostronginmantobesup-pressed,withoutdangerofdestroyingthehumanspecies”(W3,250).Ifthispsychologicalhypothesisexpressesauniversalfactaboutus,thentheunsatisfactorymethodswillindeedproveunreliableinthelongrun.Theywillnotproducepermanentlysettledbeliefandweshouldrefrainfromusingthem.Thepsychologicalhypothesis,however,seemstobefalse.Ihavesuggested(Misak1991)thatthewaytoresolvethisdifficultyistofocusonPeirce’sthoughtthatbeingresponsivetooranswerabletosomethingisoneofthe“essentialsofbelief,withoutwhichitwouldCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)13notbebelief”(MS673,p.11).Theaimofinquiryistogetbeliefswhicharenotmerelyfixed,butfixedinsuchawaythattheyfitwithandrespondtotheevidence.Theymustbe,inPeirce’swords,“causedbycircumstancesnotextraneoustothefacts.”Wigginsoffersushereaground-breakinganalysisofthisthoughtofPeirce’sandofhowitneednotleadtoauniformlycausalpicture.5Therequirementcanbemetbyallsortsofbelief.TherearetwoothercornerstonestoPeirce’stheoryofinquiry:criticalcommonsensismandfallibilism.Criticalcommonsensismisapositionabouthowweoughttoregardthosebeliefswhicharesettled.Itholdsthattherearemanythingswhichinquirersdonotdoubtandthatinquirymuststartwithabackgroundofbeliefswhicharenotdoubted.Abodyofsettledbeliefispresupposedfortheoper-ationofinquiryinthattherehastobesomethingsettledforsurprisetostirup.ThisdoctrinearoseasPeirce’sresponsetohisconceptionofDescartes’project–asystematicattempttobringintodoubtallhy-pothesesaboutwhicherrorisconceivable.Peircearguedthatsuchdoubtswouldbe“paper”doubts.Theyarenotgenuineandtheycan-notmotivateinquiry.Themerepossibilityofbeingmistakenwithrespecttowhatonebelievesisneverareasontorevisethosebeliefs.Anyofourbeliefsmightbefalse,butitwouldbeabsurdtodoubtthemallbecauseofthis.Ifwedid,wewouldnotpossessabodyofsta-blebeliefbywhichtojudgenewevidenceandhypotheses,andhence,wewouldblockthepathofinquiry.Wecandoubtonebeliefandin-quire,butwecannotdoubtallofourbeliefsandinquire.Peirce’spointagainstDescartesisthatifweweretosettherequirementsonknowledgeashighasDescartesdoes,wewouldhavenothinglefttogoon:...thereisbutonestateofmindfromwhichyoucan“setout,”namely,theverystateofmindinwhichyouactuallyfindyourselfatthetimeyoudo“setout”–astateinwhichyouareladenwithanimmensemassofcognitionalreadyformed,ofwhichyoucannotdivestyourselfifyouwould...Doyoucallitdoubtingtowritedownonapieceofpaperthatyoudoubt?Ifso,doubthasnothingtodowithanyseriousbusiness.(CP5.416)SoPeirceisnotconcernedwithscepticalquestionsaboutfounda-tionsforcertaintyandhisargumentsarenotaddressedtothosewhoare.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n14cherylmisakButheisalsoa“contritefallibilist,”holdingthatallourbeliefscanbedoubted;thatis,thatnoneofthemarecertain.Thereisatensionhere:howcanitbethatallourbeliefsarefallible,orsubjecttodoubt,butnevertheless,someofourbeliefsmustnotbedoubtedifinquiryistobepossible?Peirce’sreconciliationoffallibilismwithcriticalcommonsensismismadeintermsofhisnotionoftruth.Hethinksthatmanyofourbeliefsareindeedthosewhichwouldsurviveinquiry,butsincewecannotknowforanygivenbeliefwhetherornotitwouldbeinde-feasible,wecannotknowthatitistrue.Thatis,wedonotknowiftheantecedentofthissubjunctiveconditionalisfulfilled:“ifin-quirywerepursuedasfarasitcouldfruitfullygo,thenPwouldbebelieved.”InquirymayormaynothavebeenpursuedfarenoughwithrespecttoP,andsowecannothavecertaintywithrespecttoanybelief.Buttheuncertaintyorfallibilitythatinprincipleaccompaniesev-eryoneofourbeliefsdoesnotmeanthatweshoulddoubtoursettledbeliefs.“Practicallyspeaking,”hesays,manythingsare“substan-tiallycertain”(CP1.152);wedonotdoubtthem.While“itispossi-blethattwicetwoisnotfour...itwouldbedifficulttoimagineagreaterfollythantoattachanyseriousimportancetosuchadoubt”(CP7.108).“Substantialcertainty,”however,isdifferentfromthe“absolutecertainty”whichwouldresultfromknowingthatwehaveperma-nentlysettledbelief.Wemayhavethissettledopinionaboutmanyquestions,butwemustnotinferthatwe“perfectlyknowwhenweknow.”Again,wecannotknowthatanygivenhypothesisisper-manentlysettleduponortrue–wecannothaveabsolutecertainty.Nevertheless,ineverystateofintellectualdevelopmentandinfor-mation,therearethingsthatseemtoussure“sothateventhoughwetellourselvesthatwearenotsure,wecannotclearlyseehowwefailofbeingso”(CP4.64).Practically,wemusttreatsomehy-pothesesascertain.Settledbeliefsmustberegardedasinfallible,inthesensethattheinquirerdoesnotdoubtthemforthepurposesofinquiry;sciencehas“establishedtruths”tobeusedaspremissesinfurtherdeliberation(CP1.635).Inthissense,wedonotdoubtwhatwebelieve,butinanothersense,eachofourbeliefscan,orcould,bedoubted.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)15Peirce’stheoryofinquiryprovidesthekeytounderstandinghisviewofthegrowthofknowledgeandtheprogressofscience.HispositionanticipatesNeurath’smetaphorofbuildingaboatatsea,replacingdefectiveplanksonebyone.Science,Peircesays,isnotstandinguponthebedrockoffact.Itiswalkinguponabog,andcanonlysay,thisgroundseemstoholdforthepresent.HereIwillstaytillitbeginstogiveway.(CP5.589)Acceptedhypothesesandtheoriesarestableuntiltheyareupsetbyexperience.Theyareasgoodastheycanbe,giventhestateofev-idence,technology,argument,etc.Knowledgeisrebuiltbitbybitwhenexperienceforcesinquirerstorevisetheirbeliefs.Wehavesomereasontobelievethatweareadvancingorgettingclosertothetruth,forthenewbeliefswillgetalongwithexperiencebetterthantheoldones.Truebeliefsarethosewhichwould,intheend,getalongwithexperienceandoneexplanationofourbeliefsachievingmoreandmorefitwithexperienceisthatagoodnumberofthemaretrue.Agoodnumberofthemwouldbepermanentlydoubt-resistant.ButPeirce’spictureisnotoneofplacingindubitablebuildingblocksuponeachotherasweprogresstowardthetruth.Rather,thepictureisoneofdoubt(recalcitrantexperience)forcingustoinquireuntilwereachanothertentativedoubt-resistantbelief.Thegrounduponwhichinquirywalksistenuousanditisonlythedangeroflosingourfootingthatmakesusgoforward.Doubtanduncertaintyprovidethemotiveforinquiry.Allourbeliefsarefallibleandwhensomeoneacceptsabelief,shedoessowiththeknowledgethatitmightverywellsuccumbtothesurpriseoffurtherexperience.Butifsheknowsthatthebeliefistheresultofamethodwhichtakesexperienceseriously,thensheiswarrantedinacceptingit,assertingit,andactinguponit.Inaddition,Peirce’stheoryofinquiryinvokestworegulativehopes:assumptions,suchthat,withoutmakingthem,thepartici-pantsinapracticecouldmakenosenseofthatpractice.Wemust,Peircesays,hopeorassumethatthecommunitywillcontinueindef-initelyandwemusthopethattherewouldbe,ifinquirywerepursuedfarenough,afinalsettledanswerto“theparticularquestionswithwhichourinquiriesarebusied”(CP6.610).Wemusthope,thatis,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n16cherylmisakthatbivalenceholdsforthequestionathand;wemusthopethatPor−P.Hesays,Areasonabledisputantdisputesbecausehehopes,oratleast,goesupontheassumptionthatthedisputewillcometosomething;thatistosay,thatbothpartieswillatlengthfindthemselvesforcedtoacommonbeliefwhichwillbedefinitiveandfinal.Forotherwise,whydispute?(CP2.29)Inquiryistheaskingofquestions,andapresuppositionofinquiryisthatthequestionerhopesforananswer.Wehave,Peircesays,somegroundforthishopebecauseallsortsofquestionsthatseemedatonetimetobecompletelyresistanttoresolutionhavebeenresolved.6.logic:deduction,induction,abductionPeircedescribedhimselfasfirstandforemostalogician.HedespairedofthestateofphilosophyinAmericaattheturnofthelastcen-tury;philosophers,hesaid,foundformallogictoodifficult.Heclas-sifiedinferenceintothreetypes,deduction,induction,andabduction(whichhealsocalledretroductionorhypothesis)andmadesignifi-cantcontributionstothestudyofeach.Indeed,theveryideaofab-duction,whatistodayknownas“inferencetothebestexplanation,”isduetoPeirce.AsismadeclearinDipert’sessayinthisvolume,Peirce’scontri-butionstodeductivelogicaremostimpressive,althoughtodayitisFrege,notPeirce,whoisregardedasbringingmodernlogicintotheworld.Peircedevelopedalogicofrelationsandquantifiersindepen-dentofandatroughlythesametimeasFrege,discoveredtheShefferStroketwentyyearsbeforeSheffer,andinventedanotation(utilizingnormalforms)verysimilartotheonestillinuse.Inmathematics,heanticipatedDedekindonthedifferencebetweenfiniteandinfinitesetsandindependentlydevelopedargumentsaboutinfinitysimilartoCantor’s.6Peirceisalsoknownforhisworkoninduction.SomeseeinhiswritingananticipationofReichenbach’sprobabilisticresponsetoHume’sscepticismaboutinduction,whileothersseeananticipa-tionoftheNeyman–Pearsonconfidenceintervalapproachtotestingstatisticalhypotheses.7Whatweusuallythinkofasinductiveinference(thatwhichcon-cludesthatallAsareBsbecausetherearenoknowninstancestotheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)17contrary)iswhatPeircecalled“crudeinduction.”Itassumesthatfutureexperiencewillnotbe“utterlyatvariance”withpastexpe-rience(CP7.756).This,Peircesays,istheonlykindofinductioninwhichweareabletoinferthetruthofauniversalgeneralization.Itsflawisthat“itisliableatanymomenttobeutterlyshatteredbyasingleexperience”(CP7.157).Theproblemofinduction,asHumecharacterizesit,isaboutcrudeinduction;itisaboutthelegitimacyofconcludingthatallAsareBsorthatthenextAwillbeaBfromthefactthatallobservedAshavebeenBs.PeirceassumesthatHume’sproblemisstraightforwardlysettledbyfallibilismandcriticalcommonsensism.Wearerighttobelievethatthesunwillrisetomorrow,yetitisbynomeanscertainthatitwill.Toshowthatinductionisvalid,weneednotshowthatwecanbecertainaboutthecorrectnessoftheconclusionofacrudeinductiveinference.Fallibilismholdsthatthisisapipedream.Whatwehavetoshow,rather,isthatinductionisareliablemethodininquiry.Peirceholdsthatitisamistake,anyway,tothinkthatallinduc-tivereasoningisaimedatconclusionswhichareuniversalgeneral-izations.Thestrongestsortofinductionis“quantitativeinduction”anditdealswithstatisticalratios.Forinstance:Case:Thesebeanshavebeenrandomlytakenfromthisbag.Result:2/3ofthesebeansarewhite.Rule:Therefore2/3ofthebeansinthebagarewhite.Thatis,onecanarguethatif,inarandomsamplingofsomegroupofSs,acertainproportionr/nhavethecharacterP,thesamepropor-tionr/noftheSshaveP.Oneconcludesfromanobservedrelativefrequencyinarandomlydrawnsampleahypothesisabouttherel-ativefrequencyinthepopulation.SeeLevi’scontributiontothisvolumeforacarefulanalysisofPeirceandquantitativeinduction.Peirceisconcernedwithhowinductiveinferenceformsapartofthescientificmethod:howinductiveinferencescanfulfilltheirroleasthetestinggroundforhypotheses.Quantitativeinductioncanbeseenasakindofexperiment.WeaskwhattheprobabilityisthatamemberoftheexperimentalclassoftheSswillhavethecharacterP.TheexperimenterthenobtainsafairsampleofSsanddrawsfromitatrandom.ThevalueoftheproportionofSssampledthatarePapproximatesthevalueoftheprobabilityinquestion.Whenwetest,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n18cherylmisakweinferthatifasamplepassesthetest,theentirepopulationwouldpassthetest.Orweinferthatif10%ofthesamplehasacertainfeature,then10%ofthepopulationhasthatfeature.Peircetookthethreetypesofinferencetoformthescientificmethod.Theroleplayedbyinductionistotesthypotheses.Thejobofabductiveinferenceistoprovidehypothesesfortesting.Inab-ductiveinference“wefindsomeverycuriouscircumstance,whichwouldbeexplainedbythesuppositionthatitwasacaseofacertaingeneralrule,andthereuponadoptthatsupposition”(W3,326).Theformittakesis:Thesurprisingfact,C,isobserved;ButifAweretrue,Cwouldbeamatterofcourse,Hence,thereisreasontosuspectthatAistrue.(CP5.189)PeircearguedwithPaulCarusaboutwhenanexplanationiscalledfor.CarusclaimedthatirregularitydemandsanexplanationandPeircedisagreed.Nobody,hesays,is“surprisedthatthetreesinaforestdonotformaregularpattern,orasksforanyexplanationofsuchafact”(CP7.189).Peircesuggeststhatirregularityis“theover-whelminglypreponderantruleofexperience,andregularityonlythestrangeexception.”Amereirregularity,wherenodefiniteregular-ityisexpected,hesays,createsnosurprise;itexcitesnocuriosity.Anditissurpriseoranomalywhichthrowsusintodoubt–whichdemandsaninquirytoexplainthesurprisingphenomenon.Anun-expectedregularityorthebreachofanexistingregularitymakesademandforexplanation.Theinterruptionofahabitofexpectation(abelief)callsforanexplanation.Abductionis“theprocessofforminganexplanatoryhypothesis”(CP5.171)forsuchunexpectedregularitiesorbreachesofregular-ities.Thesehypotheses,however,aremerelyconjectures;wemust“holdourselvesreadytothrowthemoverboardatamoment’snoticefromexperience”(CP1.634).Foranabductiveinference“commitsustonothing.Itmerelycausesahypothesistobesetdownuponourdocketofcasestobetried”(CP5.602).Sothefirststageofinquiryisarrivingataconjectureoranex-planatoryhypothesis.Peircearguedthatabductionandinductionare“ampliative”anddeductionis“explicative.”Inexplicativein-ference,theconclusionfollowsfromthepremissesnecessarily;inampliativeinference,theconclusionamplifiesratherthanexplicatesCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)19whatisstatedinthepremisses.Hearguesthatampliativeinferenceistheonlykindthatcanintroducenewideasintoourbodyofbelief.Beingaformofampliativeinference,abductionallowsustoinfer,oratleastconjecture,fromtheknowntotheunknown.Wecaninferahypothesistoexplainwhyweobservedwhatwedid.Thesecondstageistodeduceconsequencesorpredictionsfromthehypothesis.The“purpose”ofdeductionis“thatofcollectingcon-sequentsofthehypothesis.”Thethirdstageisthatof“ascertaininghowfarthoseconsequentsaccordwithExperience”(MS841,p.44).Byinductionwetestthehypothesis:ifitpasses,itisaddedtoourbodyofbelief.Peirceseesthatthevalidityofabductiveinferenceisatrickymat-ter.Itsconclusionisnotevenassertedtobetrue:Thehypothesiswhichitproblematicallyconcludesisfrequentlyutterlywronginitself,andeventhemethodneednoteverleadtothetruth;foritmaybethatthefeaturesofthephenomenawhichitaimstoexplainhavenorationalexplanationatall.Itsonlyjustificationisthatitsmethodistheonlywayinwhichtherecanbeanyhopeofattainingarationalexplanation.(CP2.777)Thereasonwearejustifiedinmakingabductiveinferencesisthat,ifwearetohaveanyknowledgeatall,wemustmakethem.Alogician,Peircesays,shouldhavetwogoals–heshould“bringouttheamountandkindofsecurity...ofeachkindofreasoning”andheshouldbringoutthe“uberty,orvalueinproductiveness,ofeachkind”(CP8.384).Abductionissuchthat“thoughitssecurityislow,itsubertyishigh”(CP8.388).Itistheothertwokindsofinferencetowhichthenotionsofsecurityandvaliditymoreaptlyapply.7.thecategoriesPeirceexpendedagreatdealofintellectualenergyengaginginaprojectwhichabsorbedAristotleandKant–thecategories.Peirce’subiquitousclassificatoryscheme–thecategoriesofFirstness,Sec-ondness,andThirdness–isdesignedtocoveranyobjectofthought.Itisaclassificatoryschemethattakeseachcategorytobean“inde-pendentanddistinctelementofthetriuneReality”(CP5.431).Thedoctrine,whichpermeatesPeirce’swork,isextremelycomplexanddifficult.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n20cherylmisakPeircehadthreemethodsforarrivingathislistofcategories.Thefirstandearliestoneisfoundinthe1867“OnaNewListofCate-gories.”TheprojectisaKantianone–tofindoutwhat“is”or“hasbeing”by“reducingthemanifoldofsenseimpressionstounity”viaananalysisoftheproposition.Thesecondmethodisanargumentfromphenomenology,which“ascertainsandstudiesthekindsofelementsuniversallypresentinthephenomenon”or“whateverispresentatanytimetothemindinanyway”(CP1.186).Bothofthesemethodsaimtoshowthateverythingthatweexperienceoridentify,thatis,anythingthat“is,”hasanelementofeachofthethreecate-goriesinit,andthatwedonotexperienceanythingthatgoesbeyondthethreecategories.BoththeKantianandthephenomenologicalderivationsofthecategoriesrestontheAristotelian/Scholasticmethodofprescis-sion.Thismethodseparatesordistinguishesdifferentelementsofaconceptsothat,althoughwecannotimagineasituationinwhichoneofthemisactuallyisolated,wecantellthattheelementsaredistinct.Wecan‘suppose’onewithouttheother,forwecan,byattendingtoonefeatureandneglectingothers,isolatefeaturesofphenomenawhicharenotinfactseparable.Wecan,forinstance,supposespacewithoutcoloureventhoughcolourlessspaceisnotimaginable.Prescission,however,isnotreciprocal,asitisamatterofdiscerningalogicalpriorityofnotions.Hence,wecannotpre-scindcolourfromspace–wecannotsupposecolourwithoutspatialextension.Withrespecttothecategories,Peircearguesthatwecanabstractorprescindcertainnotionsfromexperienceandclassifythemasbe-longingtooneoranotherofthecategories.WecanprescindFirstnessfromSecondnessandwecanprescindbothfromThirdness,butwecannotprescindintheotherdirection.Sothecategoriesaredesignedtodescribethegeneralfeaturesofeachoftheclassesofelementsthatcomebeforethemindorareexperienced.Eachclassisdistinct,butitsmemberscannotstandinisolation.Eachofthecategoriesispresentineverythingweex-perience,buttherearemanycasesinwhichoneortheotherofthecategoriesisemphasisedorpredominant:“althoughtheyaresoinex-tricablymixedtogetherthatnoonecanbeisolated,yetitismanifestthattheircharactersarequitedisparate”(CP1.284).Andthelistofthreeisallthatisneeded.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)21PerhapstheeasiestwaytosetoutPeirce’sdoctrineofcategoriesistoconcentrateonhisthirdderivation,thatwhichrestsonthelogicofrelations.(Thismethod,however,isdiscussedbyPeirceasbeingpartofphenomenology.)Herethecategoriesarerepresentedbyn-placerelations.Peircearguedthatallrelationsfallintooneofthreefundamentalclasses:monadic,dyadic,andtriadic.Eachisirreducibletotheothers,andallpredicateswithmorethanthreeplacesarereducibletotriadicones.Forinstance,“aisred”ismonadic,“ahitb”isdyadic,and“agivesbtoc”istriadic.Afour-placepredicatesuchas“aputbbetweencandd”isreducibletotwothree-placeones:“aputbinspote”;“spoteisbetweencandd.”Athree-placerelationsuchas“gives,”ontheotherhand,isnotreducibleto“aputbdown”and“cpickedbup,”asthesefailtoexpresstheintentionofathatchaveb.Theresultsofthethreewaysofinquiringintotheultimatecat-egoriesaresimilar.Hereisabriefdescriptionofthoseresults,onewhichdoesnotundertaketheintimidatingtaskofsortingoutthere-lationshipsbetweenallofthethingsthatsupposedlymanifesteachcategory.Thethirdcategoryinvolvesamediumorconnectinglinkbetweentwothings;irreduciblytriadicactionissuchthataneventAproducesaneventBasameanstotheproductionofaneventC.Thirdnessisoftenmanifestedinpsychologicalconcepts.Wecannotgraspwhatitisforatogivebtocwithoutthenotionofintentionmediatingbe-tweenaputtingbdownandcpickingupb.Similarly,Peircearguesthatrepresentationissuchthataninterpretingthoughtmediatesbetweensignandobject.(OneroutetoPeirce’sclaimthatallexpe-rienceisamatterofThirdnessisviahisargumentthateverythingweexperienceisofthenatureofasignorrepresentation.Thereisnoexperienceindependentofourrepresentationofit.)Peircealsosaysthatlawandnecessitymanifestthirdness.Alaw,oranecessaryconnection,mediatesbetweentheactionofonethinguponanother,makingitmorethananaccidentthattheybehavedinthewayinwhichtheydid.ContinuityandgeneralityareotherexamplesPeircegivesofthirdness.WecancognitivelyisolateSecondnessasthedualityofactionandreactionwithoutanymediatingforce.Itisbruteexistenceandhenceisthemodalityofactuality.Itisfound(byprescission)mostclearlyinthenotionsofstruggle,action/reaction,cause/effect,andbruteCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n22cherylmisakforce.Thesecondcategoryisone“whichtheroughandtumbleofliferendersmostfamiliarlyprominent.Wearecontinuallybumpingupagainsthardfact”(CP1.324):Wecanmakenoeffortwhereweexperiencenoresistance,noreaction.Thesenseofeffortisatwo-sidedsense,revealingatonceasomethingwithinandanothersomethingwithout.Thereisbinarityintheideaofbruteforce;itisitsprincipalingredient.(CP2.84)AFirstisasimplemonadicelement.Peircesaysthatitsuggestsspontaneity,anditisreal“regardlessofanythingelse.”Invirtueofitsverynature,itisindescribable;itcanonlybegraspedbyprescission:Itcannotbearticulatelythought:assertit,andithasalreadylostitscharac-teristicinnocence;forassertionalwaysimpliesadenialofsomethingelse.Stoptothinkofit,andithasflown!...thatisfirst,present,immediate,fresh,new,initiative,original,spontaneous,free,vivid,conscious,andevanescent.Only,rememberthateverydescriptionofitmustbefalsetoit.(CP1.357)These“qualitiesoffeeling”aremerepossibilities:Idonotmeanthesenseofactuallyexperiencingthesefeelings...thatissomethingthatinvolvesthesequalitiesasanelementofit.ButImeanthequalitiesthemselveswhich,inthemselves,aremeremay-bes,notnecessar-ilyrealized.(CP1.287)Sothefirstcategoryisthatofpossibility.OneupshotofPeirce’sdoctrineofcategoriesisthathethinksthatrealitycomesinthreegrades.Heisa“realist”withrespecttoallofthecategories–possibility,actuality,andgeneralityarereal.Heinsiststhat“thewillbe’s,theactuallyis’s,andthehavebeensarenotthesumofthereals.Theyonlycoveractuality.Therearebesideswouldbe’sandcanbe’sthatarereal”(CP8.216).AndhisScholasticrealismhasitthatlawsorthirdsarereal;theyarenotmerementalconstructions.Peircetakesnominalism–thedoctrinethat“lawsandgeneraltypesarefigmentsofthemind”(CP1.16)–tobepernicious.Hesays:...theproperty,thecharacter,thepredicate,hardness,isnotinventedbymen,asthewordis,butisreallyandtrulyinthehardthingsandisoneinthemall,asadescriptionofhabit,disposition,orbehavior....(CP1.27n1)CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)23Peircethinksthatthefactthatwecanpredictthingsoughttoconvinceusofrealismaboutgenerals.Scholasticrealismexplainspredictionbyholdingthatlawsanddispositionshavecausalefficacy:“ifthereisanywouldbeatall,thereismoreorlesscausation;forthatisallthatImeanbycausation”(CP8.225n10).Ifapredictionhasatendencytobefulfilled,itmustbethecasethatfutureeventshaveatendencytoconformtoageneralrule.Peirceconcludesthatsomelawsorgeneralsarereal.Lawsanddispositionsmediatebetweenpossibility(Firstness)andactuality(Secondness)–itisthelawthatmakesthepossibleactual,forlawsorgeneralpatternscausetheirinstances.ButPeircedoesnotthinkthatpossibilitiesandgeneralsactuallyexist;universalsorgeneralsarenot“things.”Therealmofexistenceisthesecondcategory,andsopossibilitiesandgeneralsarerealbutnotexistent.8.metaphysicsThedoctrineofcategoriesisnotPeirce’sonlymetaphysicalven-ture.Heissetagainstdeterminism,whichhetakestobethepositionthat“everysinglefactintheuniverseispreciselydeter-minedbylaw”(CP6.36).His“Tychism”hasitthatthereisabso-lutechanceintheuniverse–thereisspontaneousdeviationfromthelawsofnature.PeircetakesacorolloryofTychismtobethatphysicallawsarestatistical,somethingwhichphysicsnowtakesforgranted.TychismistiedtoPeirce’sviewofevolutionarycosmology,forTychismhasitthatthereisatendencytowarddiversificationintheuniverse.Laws,Peircethinks,evolvedfrom“purepossibility.”Thestartingpoint“wasnotastateofpureabstractbeing.Onthecontraryitwasastateofjustnothingatall,notevenastateofemptiness,forevenemptinessissomething”(CP6.215).HeusuallysaysthatitwaspureFirstness–recallthatspontaneityisparadigmaticofFirstness.Itisastatewhichhasnoexistingthings(Secondness),compulsion(Secondness),orlaw(Thirdness):itisastateofpurechanceorpossi-bility.Fromthisstateofpossibilitycameaccidental“flashes”(CP1.412)which,againaccidentally,reactedwithoneanother.Thatis,Sec-ondnessemerged.Andfromthesereactionsaroseahabit-takingCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n24cherylmisaktendencyorThirdness.Peircesaysthatitisthenatureofhabittoeverstrenghtenitselfandthus,lawscameintobeing.Evolutionistheprocessofgrowth;theworldbecomesmoreandmorerationalandlaw-bound.AnotherofPeirce’smetaphysicaldoctrinesis“Synechism,”whichhasitthatthenotionofcontinuityisthekeytophilosophy.Some-timeshesaysthat“Synechismisnotanultimateandabsolutemeta-physicaldoctrine;itisaregulativeprincipleoflogicprescribingwhatsortofhypothesisisfittobeentertainedandexamined”(CP6.173).Butatothertimeshepresentsitashighlymetaphysical.LikeAristotle,Peirceholdsthatacontinuousseriesisnotacol-lectionofdiscretepoints.Acontinuousseriesisratherapossibilityofendlessfurtherdetermination.Acontinuumhasnoexistingparts,butonlyapotentialforbeingdividedintoparts.Theinfinitenumberofpointsonacontinuouslinearereallyplacesatwhichapointcouldbelocated;theyaremerelypossiblesorFirstsratherthanactualsorSeconds.ContinuityitselfisaninstanceofThirdness;itisakindofultimatemediation.Foracontinuousseriesisapathwherewecanalwaysfindonethingbetweentwoothers.Peircecharacteristi-callytriestolinkupthisexampleofThirdnesswithothers,mostparticularly,withlawsandgenerality.AnothermetaphysicaldebatewhichPeircejoinedisthedebateaboutreality.SometimeshewritesofrealitynotinthewaydescribedinSection3,whererealityistheobjectofperfectlystablebeliefs.Butsometimesheplaceshisviewofrealitywithintheidealism–materialismdebateandsidesforakindofidealism.Reality,hesays,isnothingbut“effetemind”–“whatwecallmatterisnotcom-pletelydead,butismerelymindhideboundwithhabits”(CP6.158).ItisunclearwhetherthisidealismcanbereconciledwiththeviewofrealityelucidatedwithinPeirce’saccountoftruth.Anditisun-clearwhetheridealism,alongwiththeothermetaphysicaldoctrinestoucheduponhere,canpassthepragmatictest,whichrequiresmeta-physicaltheoriestohaveconsequencesforpractice.89.influenceThepragmatictheoryoftruthisstillagoingconcern.SomeofthecurrentbrandsarenottheonePeircehimselfoffered,butclosertothoseofWilliamJamesandJohnDewey,bothofwhomCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCharlesSandersPeirce(1839–1914)25acknowledgedtheirdebttoPeirce.(SeePihlstrom’sessayforanac-countoftheserelationships.)RichardRorty’spragmatism,forinstance,hasitthattheveryno-tionoftruthismetaphysicalandoughttobeabandoned.Peirce,ontheotherhand,thinksthattruthnotonlyisasensiblenotion,but,giventhatitiswhatinquirersaimat,itisanotionwhichisessentialforinquiry.W.V.O.Quine(insomemoods)andHilaryPutnamaremoreclearlytheinheritorsofPeirceanpragmatism.AnotherareawherePeirce’sinfluenceisstillfeltisinthefieldofsemeiotics,wheremanyofhisdistinctions,classifications,andterminologystillreign.Hisinfluenceinthefieldoflogicwasim-pededbyhisisolationandbythefactthattheBooleanschoolwaseventuallyedgedoutofthemainstreambytheFregean.Schroder¨adoptedPeirce’snotation,andsomewell-knownresultsarewrit-teninit.9AndWhiteheadseemstohavelearntquantificationfromPeirce.Butdespitethequantityandqualityofhisworkinformallogicandstatisticalinference,heisprobablybestrememberedinlogicforintroducingabductiveinference,somethingwhichbyitsverynaturecannotbeformalized.Unfortunately,Peirce’slackofsuccessinsecuringanacademicposition,hisperhapsabrasivepersonality,andhispenchantforcum-bersometerminologycombinedtorenderhisviewsprettymuchin-accessibleduringhisownlifetime.Hediedpennilessandunappre-ciated.Ithasonlybeenrecentlythathisworkhasfoundtheinterestitdeservesandtheexcavationitrequires.notes1.SeeBrent(1993)foranaccountofPeirce’swoesintheacademy.Menand(2001)is,Ibelieve,anunreliableaccount:itishighlyspeculativeaboutPeirce’scharacterandbizarreinitsanalysisoftheriseandfallofprag-matism’spopularityintheUnitedStates.2.LettertoWilliamJamesin1903,quotedinPerry1936:vol.II,425.SeeMisak(1995)foranaccountofPeirce’splaceintheempiricisttradition.3.MuchofthematerialthatfollowsrestsonMisak(1991)and(1995).4.SeealsoHookway(2000).5.SeealsoMisakinthisvolumeand(1991:59ff).6.SeeDauben(1982),Dipert(1981a),andPutnam(1982)fordetails.7.Forthelatter,seeLevi(1980),Levi’scontributiontothisvolume,andHacking(1980).SeeWigginsinthisvolumeforabetterPeirceanCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n26cherylmisakresponsetoHumethantheReichenbachianonesometimesattributedtohim.8.Peircedidargueagainstsomekindsofidealismonpragmaticgrounds:“Verywell;anidealist...isloungingdownRegentStreet...whensomedrunkenfellowunexpectedly...letsflyhisfistandknockshimintheeye.Whathasbecomeofhisphilosophicalreflectionsnow?”(CP5.539).9.Forinstance,Lowenheim’stheoremandZermelo’saxioms:seePutnam(1982).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nsamipihlstr¨om2Peirce’sPlaceinthePragmatistTraditionYourintenselymathematicalmindkeepsmynon-mathematicaloneatadistance.Butsomanyofourcategoriesarethesamethatyourexistenceandphilosophizinggivemethegreatestcomfort.Perry1935/1936:I,224;James’slettertoPeirce,March27,1897Yourmindandmineareaslittleadaptedtounderstandingoneanotherastwomindscouldbe,andthereforeIalwaysfeelthatIhavemoretolearnfromyouthanfromanybody.CP8.296;Perry1935/1936:II,431;Peirce’slettertoJames,October3,19041.introduction“Whooriginatedthetermpragmatism,Ioryou?Wherediditfirstappearinprint?Whatdoyouunderstandbyit?”CharlesPeirceaskedhisfriendWilliamJamesinaletteronNovember10,1900(CP8.253;Perry1935/1936:II,407n5).OnNovember26,1900,Jamesreplied:“Youinvented‘pragmatism’forwhichIgaveyoufullcreditinalec-tureentitled‘Philosophicalconceptionsandpracticalresults.’”1Thepublishedversionofthatlecture(1898)isverylikelytohavebeenthefirstplacewheretheterm“pragmatism”wasusedinprint,andJameswasthefirstphilosopherknownasapragmatist.Theprag-matistmovementwaslargelydevelopedbyJames,althoughDewey,Royce,andevenSchillermayhavehadanoriginalandindependentroletoplayinitsformation.Nonetheless,JamesreferredtoPeirce’searlierunpublishedusageofthetermandacknowledgedPeirceas27CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n28samipihlstr¨omthefirsttoformulateapragmatisticdoctrineinthediscussionsoftheCambridge“MetaphysicalClub”intheearly1870s.Thepurposeofthisessayisnottodeterminetheoriginofpragmatism.2Itis,rather,myaimtosituatePeirce’sversion(s)ofpragmatism3intheircontext;thatis,toinvestigatePeirce’splaceinthetraditionofpragmatistthoughtthatextendsfromthe1870stotherecentneopragmatismsofthe1980sand1990s.Weshouldremem-berthatPeirceinfluencedtwentiethcenturyphilosophersmainlyposthumously.ThecollectionChance,Love,andLogicwaspub-lishedin1923,nineyearsafterhisdeath,andtheCollectedPaperswerepublishedineightvolumesin1931–1958.Inanycase,Peirce’sworksdideventuallyhaveaninfluenceinthephilosophicalcommu-nity,anenormousinfluencewithoutwhichtherewouldbenothinglikepragmatismasweknowit.Peircebegantocallhisview“pragmaticism”afterhavingper-ceivedhowthenotionofpragmatismhadbeenusedafterhisoriginalcoinageoftheterm.Thekeypassagefromhis1905Monistpaper,“WhatPragmatismIs,”isworthquoting:[James]firsttook[theword“pragmatism”]up,seeingthathis“radicalempiricism”substantiallyansweredtothewriter’sdefinitionofpragma-tism....Next,theadmirablyclearandbrilliantthinker,Mr.FerdinandC.S.Schiller...lit...uponthesamedesignation“pragmatism,”whichinitsorig-inalsensewasingenericagreementwithhisowndoctrine....Sofarallwenthappily.Butatpresent,thewordbeginstobemetwithoccasionallyintheliteraryjournals,whereitgetsabusedinthemercilesswaythatwordshavetoexpectwhentheyfallintoliteraryclutches....Sothen,thewriter,findinghisbantling“pragmatism”sopromoted,feelsthatitistimetokisshischildgood-byandrelinquishittoitshigherdestiny;whiletoservetheprecisepur-poseofexpressingtheoriginaldefinition,hebegstoannouncethebirthoftheword“pragmaticism,”whichisuglyenoughtobesafefromkidnappers.(CP5.414/EP2:334–35)4PeircedidnotclaimJamesorSchillertohave“kidnapped”his“pragmatism.”Itistheuseofthenotion“intheliteraryjournals”thatwasthecauseofhisanger;hedidnotwanttoreplace“pragma-tism”assuchwith“pragmaticism,”butapparentlyintendedhisnewcoinagetorefertoasubdivisionofpragmatism.(SeeHaack1998:55;Kilpinen2000:35.)ThisisnottosaythatPeircewouldhaveagreedwithJamesandSchiller,butitperhapsshowsthatthereissuchCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition29athingasthepragmatisttradition,originatedbyPeirceandcon-tinuedbyJames,Dewey,Schiller,Mead,andtheirfollowers.Thereisnoneedtoinsist,assomescholarsdo,thatthebroadermove-mentknownaspragmatismissomethingessentiallydifferentfromPeirce’sownpragmaticism.AggressivelyorthodoxPeirceanswhothinkonlyPeirce’sviewsdeservephilosophicalattentiontendtooverlooktheremarkableintegritywefindamongthepragmatists,de-spitetheiroccasionalprofounddisagreements.Therearebothunityanddifferences-in-unityinthepragmatisttradition.Weshouldbeskepticalaboutallattemptstofindjusttwoformsofpragmatism(e.g.,Peirce’sandallothers’)opposedtoeachother.5Boththeintegrityandthedisagreementsamongpragmatistsareworthdiscussing.Sinceitisimpossibletomakeanydetailedcom-parisonsbetweenPeirceandotherpragmatistsinasinglearticle,IshallfocusonJames(Sections2–4),offeringonlygeneralremarksonPeirce’srelationstoDeweyandSchiller(Sections5–6),whileRoyce,Mead,andotherclassicalthinkerscanhardlybemorethanmentioned.6Finally,IshallcomparePeircetoneopragmatistssuchasPutnamandRorty(Section7),beforeconcludingwithreflectionsonPeirce’sandotherpragmatists’relationtotherealismvs.idealismdispute(Section8).Thesecomparisons,briefastheymustremain,areintendedtoplacePeirceinhispositionintheextremelyrichtraditionhefounded.2.peirceandjames:realismandtruthIthasbeensuggested,plausibly,thatthebasicdifferencebetweenPeirceandJamesintheirpartlyconflictingcharacterizationsofprag-matismwasthattheformerdevelopedastrictlylogicalmethodthatwouldhelpusunderstandthemeaningofscientificconcepts,whereasthelatterwasinterestedinawiderapplicationofthepractice-orientedmethodofpragmatisminhumanconcerns(Hook-way1997).Thisdifferenceintheir“philosophicaltemperaments”–touseJames’sterm–andintheiroverallphilosophicalprojectsisreflectedinanumberofmoredetaileddifferences,7which,how-ever,shouldnotconcealtheirsimilarities.ItisamistaketointerpretJames’spragmatismasameremisunderstandingormisapplicationofPeirce’s.Jameswasanindependentthinker.Hedidnotsimplymis-understandPeircebutemployedpragmatismmorebroadly,partlyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n30samipihlstr¨ombecausehehadadifferentconceptionofscienceandthepracticalusesofinquiry(cf.Hookway2000).Theoppositionbetweenrealismandnominalismhasbeenrec-ognizedasoneoftheissuesdividingPeirceandJames.Peirceal-waysresistednominalism,thinkingthatitcommittedtheworstofphilosophicalsins,viz.,blockingtheroadofinquiry(cf.CP1.170,c.1897).Peirceevencametoresistsomeofhisownearlyformu-lationsofpragmatismastoonominalistic,anddescribedhimselfas“ascholasticrealistofasomewhatextremestripe”(CP5.470,c.1906).Scholasticrealismisessentiallythedoctrinethatthereare“realgenerals”(universals,dispositions,laws,habits).Thisview,Peircethought,isrequiredinanyadequateformulationofscientificphilosophyandmetaphysics,includingpragmatismitself.Ifuniver-salityandgeneralitywere“dependentuponwhatwehappenedtobethinking,”science“wouldnotrelatetoanythingreal”(CP8.18,1871).8James’spragmatismismorenominalisticallyinclined.Al-thoughitwouldbeanexaggerationtocallJamesa“nominalist,”itistruethathefocusedonparticularexperiencesandpracticalconse-quencesofactions,whereastheconsequencesPeircewasinterestedinweregeneralpatternsandhabits(Hookway1997:152).Anotherdifference,relatedtoscholasticrealism,isthis:whileinsomesenseJameswent(orwantedtogo)“roundKant”whereasPeirce’sviewsweredeveloped“throughKant”(Fisch1986:288),itturnsoutthatJames,contrarytohisownself-image,wasthemorethoroughgoingKantian.James’sconstructivisticpragmatismcanbeinterpretedasaformoftranscendentalidealism,whereasinPeirce’scasesuchaKantian(re)interpretationismoredifficult,becauseofhisextremerealism(cf.Pihlstrom¨1998a).OneofthepointswhereJameshasbeentakentohavedistortedPeirce’spragmatismisthetheoryoftruth.Butratherthaninter-pretingJames’spragmatisttheoryoftruthasamisunderstandingofPeirce,wemayseeitasa“substantialextension”ofPeirce’sview,accordingtowhichtruthissomethingthatissatisfactory,useful,expedient,orgoodforustobelieve,somethingthatis“safefromoverthrowbysubsequentexperience”(Haack1976:233–4).Becauseofhismorenominalistbias,Jamesfocusedonindividual,concretetruthsthatweretobepracticallyusedinthecourseofexperience,ratherthanonanythinglikethe“Truth,”orthefinalopinionofthescientificcommunity(Haack1976:234).PeircementionsJames’sCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition31doctrineofthe“mutabilityoftruth”asoneofthe“seedsofdeath”withwhichhisoriginalpragmatismbecameinfectedinthehandsoflaterpragmatists(CP6.485,1908).Yetthepragmatisttheoryoftruthis,accordingtoHaack(1976:236,247),a“cosmopolitan”the-ory,containingbothcorrespondenceandcoherenceelementsandre-ceivingdifferentemphasesindifferentauthors.Itneednotbearivalofthecorrespondencetheory,butitismeaningfultosaythatthereisonesinglepragmatisttheory,differentlydevelopedbyPeirce,James,Dewey,andothers.Hookway(2000:82,89)alsonotesthatJames’stheoryoftruth,insteadofcompetingwiththecorrespondencethe-ory,wasdesignedtoelucidatewhatagreementwithrealitymeans–andso,thoughdifferently,wasPeirce’s.9James,aswellasDewey,endorsedratherthanrejectedormisun-derstoodPeirce’sformulationthattruthistobeequatedwiththeeventualoutcomeofinquiry,orwiththeconvergenceofbelief.AsHookway(2000:44)putsit,Jamesacceptedtheconnectionbetweenconvergenceofopinionandtruth“asanaccountof‘absolutetruth,’”whereasDewey“agreedwithitasananalysisoftruthbeforecon-cludingthatlogicandepistemologywoulddowelltoabandonthisnotioninfavourof‘warrantedassertibility.’”James(1907[1975]:107)treats“absolutetruth”asaregulativenotion,andDewey(1938:345)referstoPeirce’sdefinitionas“thebestdefinitionoftruth,”fromthelogicalpointofview.10ItisDewey’sconclusionthatthenotionoftruthhasnosignificantroletoplayinlogicorinquirythatPeircedidnotdraw.Theproblemsof(Scholastic)realismandtruthonlygiveusprelim-inaryanswerstothequestionofwhatdistinguishesJames’spragma-tismfromPeirce’s.Wehavenoticeddifferencesofemphasisratherthanofprinciple–butimportantdifferencesnevertheless.Furtherelucidationisneeded.3.theoryandpracticeTurningmorecloselytotheopinionsPeirceexpressedaboutJames’sviews,wecaneasilyseethatPeircewascriticalofJames’swaysofde-velopingthepragmatistideashehadhimselfpresented,whilealsoadmittingthathisearlyformulationswererelativelyclosetothepragmatismJamesdeveloped.11PeircealsoexploredJames’sviewsincontextsnotdirectlyrelatedtopragmatism.Forexample,heCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n32samipihlstr¨omreviewedJames’sThePrinciplesofPsychology(1890)inTheNationin1891(CP8.55ff.).PeircerespectedJamesasathinker.HeadmittedthatJameswasa“perfectloveroftruth”(CP6.183,c.1911;Perry1935/1936:I,540)andagreatpragmatist:“Youareofallmyfriendstheonewhoillus-tratespragmatisminitsmostneedfulforms.Youareajewelofprag-matism”(Perry1935/1936:II,427;Peirce’slettertoJames,March16,1903).Therewere,however,significanttemperamentaldifferencesbetweenthetwo,whichPeircerecognized:“Hiscomprehensionofmentotheverycorewasmostwonderful.Who,forexample,couldbeofanaturesodifferentfromhisasI?Hesoconcrete,soliving;Iameretableofcontents,soabstract,averysnarloftwine.YetinallmylifeIfoundscarceanysoulthatseemedtocomprehend,nat-urally,[not]myconcepts,butthemainspringofmylifebetterthanhedid.Hewasevengreater[inthe]practicethaninthetheoryofpsychology”(CP6.184,c.1911).ThesedifferencescanbehighlightedbytakingalookatwhatPeircesaysaboutJames’s(1897[1979])doctrineofthe“willtobe-lieve.”InPeirce’sview,thisdoctrine,assumingthat“theendofmanisaction,”pushesthepragmaticmethod“tosuchextremesasmusttendtogiveuspause”(CP5.3,1902).James’spragmatismis“ex-treme,”implyingthat“Doingistheultimatepurposeoflife”(CP8.115,c.1900).Earlier,Peircehadremarkedthat“faith,”though“highlynecessaryinaffairs,”is“ruinousinpractice,”ifitmeansthat“youarenotgoingtobealertforindicationsthatthemomenthascometochangeyourtactics”(CP8.251,1897;Perry1935/1936:II,222;seealsoCP6.485,1908).12Later,commentingontheBergsonianconceptionofphilosophymanifestedinJames’sAPluralisticUni-verse(1909b[1977]),Peircewasevenmorecritical:“IthoughtyourWilltoBelievewasaveryexaggeratedutterance,suchasinjuresaseriousmanverymuch,buttosaywhatyounowdoisfarmoresuicidal....[P]hilosophyiseitherascienceorisbalderdash...”(Perry1935/1936:II,438;lettertoJames,March9,1909).13PeirceinsistedthatpragmatismisnotaWeltanschauungbut“amethodofreflexionhavingforitspurposetorenderideasclear”(CP5.13n1,c.1902).InalettertotheItalianpragmatistMarioCalderoni,Peirce,havingmadethedistinctionbetweenpragmatism(amongwhoserep-resentativeshementionedSchiller,James,Dewey,andRoyce)andpragmaticism,notedthatpragmaticismis“notasystemofphiloso-phy”but“onlyamethodofthinking”(CP8.205–6,c.1905).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition33Itwasalreadyin1897,afterhavingreceivedJames’sTheWilltoBelieve,dedicatedtohim,thatPeircereflectedontherelationbe-tweenhisoldandmorerecentconceptionofpragmatisminalettertoJames(March13,1897;cf.alsoCP8.255–6,1902):Thateverythingistobetestedbyitspracticalresultswasthegreattextofmyearlypapers;so,asfarasIgetyourgeneralaim...Iamquitewithyouinthemain.Inmylaterpapers,IhaveseenmorethoroughlythanIusedtodothatitisnotmereactionasbruteexerciseofstrengththatisthepurposeofall,butsaygeneralization,suchactionastendstowardregularization,andtheactualizationofthethoughtwhichwithoutactionremainsunthought.(CP8.250)Thiscontains,innuce,thedifferencebetweenJames’sandPeirce’spragmatisms,asPeircesawit.WhileitisnotclearthatJamesshouldbeinterpretedashavingfavoredmere“bruteexerciseofstrength,”itisfairlyaccuratetosaythatheconsideredactionor“doing”themainpurposeoflife.ThisissomethingthatPeirce,impressedmorebyself-reflectivehabitsandregularizedactionthanindividualac-tions,couldnotaccept.“[T]heendofthought,”hewrote,“isactiononlyinsofarastheendofactionisanotherthought”(CP8.272,1902).Thus,Peircethoughtthathisfellowpragmatists,overempha-sizingwhathecalled“secondness,”didnotreallyunderstandwhathiscategorieswereallabout(CP8.263,1905).HealsoconsideredJames’sterminologyunclear:inadditiontoaccusingJamesofhavingmisdescribed“pragmatism,”heremarkedthatJames’s“pureexpe-rience”(James1912[1976])“isnotexperienceatallandcertainlyoughttohaveaname,”becauseitis“downrightbadmoralssotomisusewords,foritpreventsphilosophyfrombecomingascience”(CP8.301,1904).Butthenagain,Jameshardlywantedphilosophytobecomeascience.AmetaphilosophicaloppositionbetweenPeirceandJamescanbeobservedintheirconceptionsoftheroleofphilosophyinhumanlife.WhilesomePeirceans–e.g.,Misak(1994,2000)–havefoundsup-portfromPeirce’snotionsoftruthandinquiryindefendingmoralrealism,thereissomeevidenceforthecontentionthatPeircedidnotconsiderour“practicalaffairs”ormattersrelatedto“theconductoflife”philosophicallyimportant.14Hecondemned,inhisCambridgeConferencesLectures(1898),“withthewholestrengthofconvic-tiontheHellenictendencytominglePhilosophyandPractice,”andremarkedthatinphilosophy,“theinvestigatorwhodoesnotstandCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n34samipihlstr¨omalooffromallintenttomakepracticalapplications,willnotonlyobstructtheadvanceofthepurescience,but[...]willendangerhisownmoralintegrityandthatofhisreaders”(RLT:107).Heclaimedthatpuresciencehasnothingtodowithaction,thatnothingis“vi-tal”forscience,that“puretheoreticalknowledge,orscience,hasnothingdirectlytosayconcerningpracticalmatters,”andthatwecannotserve“thetwomasters,theoryandpractice”(RLT:112–13;cf.CP1.642).15Yet,asimpletheory/practicedistinctionistoocrudetohavebeenPeirce’sconsideredview.Wemustrememberthecon-textofPeirce’sclaims:HeprotestedagainstJames’ssuggestionthatheshouldgivelecturesabout“vitallyimportanttopics”ratherthantechnicallogicalquestions.16Whilepointingoutthatthere“appearstobenoslighttheoreticaldivergence”betweenJames’sdefinitionofpragmatismandhisown,Peircesaidthatthatdivergence,“forthemostpart,becomesevanes-centinpractice,”andthat“thediscrepancies[betweenJamesandhim]resideinotherthanthepragmatisticingredientsofourthought”(CP5.466,c.1906).HeremarkedthatJamesdoesnotrestrict“mean-ing,”or“theultimatelogicalinterpretant,”toahabit,ashedoes,butallowsperceptstoplaythisrole;andthat,ifhe(James)iswill-ingtodothis,heneednotgiveanyroomtohabit.“Butpractically,hisviewandminemust[...]coincide,exceptwhereheallowscon-siderationsnotatallpragmatictohaveweight”(CP5.494,c.1906;seealsoEP2:421,1907).Now,inasense,practiceiswhatpragmatismisallabout.Ifthereisno“practical”differencebetweenPeirceanandJamesianpragmatisms,thenthereisallthemorereasontoseeprag-matismasonesingletraditionwithsomewhatdifferentovertones.Itis,then,overhastytoregardPeirce’sandJames’spragmatismsasfundamentallyopposedtoeachother.EventhestandarddivisionbetweenJames’s“nominalistic”andPeirce’s“realistic”pragmatismturnsouttobeproblematic,asHaack(1977:392–393)shows:thedifferenceisnotthatPeirceacceptedandJamesdeniedtherealityofuniversalsbutthatPeircedeniedthatrealuniversalscanbereducedtoparticulars,whileJamesthoughtthattheycan.Perry(1935/1936:I,547)observesthatJames,recognizingthesignificanceof“generalideas,”“wasnever(inspiteofPeirce’sstrictures)athoroughgoingnominalist”andeven“approachedthe‘realistic’position”inhismaturewritings,especiallyinAPluralisticUniverse.James“neverbecameanominalist,”forhealwaysfoundsomeway“toprovideforuniversals,generalsandconcepts,howevermuchhemightdisparageCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition35them”(Perry1935/1936:II,407).Thisisanimportantpoint,stillin-sufficientlydiscussed.AsSeigfried(1990:267)alsonotes,Jamesdidnotexclude“themodalityofpossibility,”forheaffirmedtheneedfor“generalrules,”eventhoughtheemphasiswasonthepartic-ularconsequencesexperiencedinthefuture.“Peirce’swell-knowncriticismofJamesasanominalistratherthanarealistcouldnotbefurtherfromthetextualrecord,”sheconcludes,“andyetitisuncrit-icallyrepeatedtothisday”(399n5).17AsRosenthal(2000:94)putsit,Peirceopposedanominalisticpluralismof“discreteunits,”whileJames’spluralismwasclosertoPeirce’sownsynechism,thedoctrineofcontinuity.Jamesarguedthatphilosophicalabstractionsmustdosomerealwork:pragmatism“hasnoobjectionwhatevertotherealizingofabstractions,solongasyougetaboutamongparticularswiththeiraidandtheyactuallycarryyousomewhere”(James1907[1975]:40).“Wearelikefishesswimmingintheseaofsense[sensiblefacts],boundedabovebythesuperiorelement[abstractideas],butunabletobreatheitpureorpenetrateit”(64).Itisquestionablewhetherthisevenamountstoareductionistconceptionofabstractionsandgener-alitiesinrelationtoconcretefacts.Jamesseemstohavemaintainedthatweneedabstractionsinordertoactintheworldofparticularexperientialfactsandthatthisisallweneedthemfor,buthedidnotsay,atleastnotexplicitly,thattheformerarenothingbutcom-plexesofthelatter.PerhapsthemoreimportantconflictisbetweenPeirce’sstrictantipsychologismandJames’smorepsychologicallyorientedadmissionofgeneralideas.ForJames,generalideaswerehumanbeings’classificationsofrealitythroughtheirpractices,andthusdependentonoremergingfromhumanpurposiveaction,notanythingready-madeinrealityitself.ForPeirce,undoubtedly,thiswaslittlemorethannominalism,becausetheindependent,nonpsy-chologicalrealityofgeneralswasnotacceptedbyJames.Inanycase,thedifferencesPeircefoundbetweenhisviewsandJames’s,thoughgenuineandimportant,shouldnotbeoveremphasized.18Moreover,PeirceandJamesbothheldanextremelyrich,inclu-siveconceptionofexperience,accordingtowhichweexperience“externalthingsasexternal,”interactionsbetweenthem,theirsen-soryimpactsuponus,and“law-governedinteractions–mediatedtransitions–betweenthingsweexperience,andrealcontinuityinthewaysthatprocessesdevelop”(Hookway2000:292;seePape2000).Whilebothwereempiricists,urgingthatourknowledgeisCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n36samipihlstr¨ombasedonexperience,theyrejectedthepassive,atomisticconceptionofexperienceconsistingofscatteredindividualsensationsassumedinmuchoftheempiricisttradition.4.interpretationsofthepragmaticmaximWhilePeircedistancedhispragmatismfromJames’s,Jamestendedtodiminishthedifferences.SpecificreferencestoPeircebyJamescanbefoundinTheWilltoBelieve(1897b[1979]),theVarieties(1902[1985]),Pragmatism(1907[1975]),andTheMeaningofTruth(1909a[1978]),aswellasinmanuscriptsandlecturenotes.Theseareinmostcasestothepragmaticmaxim,thoughJamesdidteachPeirce’sevo-lutionarymetaphysicsinhiscoursesatHarvard,ashisManuscriptLectures(1988)show.IntheVarieties,James(1902[1985]:351)men-tioned“theprincipleofPeirce,theprincipleofpragmatism,”re-ferringto“HowtoMakeOurIdeasClear”(1878)andapplyingtheprincipletoadiscussionofGod’smetaphysicalattributes.ThesamearticlebyPeircewasalreadyquotedinJames’s“TheFunctionofCog-nition,”readbeforetheAristotelianSocietyin1884andpublishedinMind(vol.10,1885).ThatpaperlaterformedthefirstchapterofTheMeaningofTruth.19LaterJamesreports:Theterm[“pragmatism”]isderivedfromthesameGreekword[πραγµα],meaningaction,fromwhichourwords‘practice’and‘practical’come.ItwasintroducedintophilosophybyMr.CharlesPeircein1878....Mr.Peirce,afterpointingoutthatourbeliefsarereallyrulesforaction,saidthat,todevelopathought’smeaning,weneedonlydeterminewhatconductitisfittedtoproduce:thatconductisforusitssolesignificance....Toattainperfectclearnessinourthoughtsofanobject,then,weneedonlyconsiderwhatconceivableeffectsofapracticalkindtheobjectmayinvolve–whatsensationswearetoexpectfromit,andwhatreactionswemustprepare.Ourconceptionoftheseeffects,whetherimmediateorremote,isthenforusthewholeofourconceptionoftheobject,sofarasthatconceptionhaspositivesignificanceatall....TotakeintheimportanceofPeirce’sprinciple,onemustgetaccustomedtoapplyingittoconcretecases.(James1907[1975]:28–9)Peirce’soriginaltextreadsasfollows:“Considerwhateffects,whichmightconceivablyhavepracticalbearings,weconceivetheobjectofourconceptiontohave.Then,ourconceptionoftheseCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition37effectsisthewholeofourconceptionoftheobject”(CP5.402/W3,266,1878).20WhenpresentingPeirce’sprincipleinhisCaliforniaaddressin1898,Jamessaid“itshouldbeexpressedmorebroadlythanMr.Peirceexpressesit”(James1898:124).Attemptingtodothis,heappearstoslidefromacknowledgingPeirce’snotionsofpos-sibledifferencesandconceivableeffectstothestrongerrequirementthatthosedifferencesoreffectsshouldbeactualizedinourconcreteexperiencesorpractices.Jamesdemandedthepracticalconsequencesofourconceptionstobe,aboveall,particular(James1909a[1978]:124;Perry1935/1936:I,458;II,410–11).This,thoughlittlemorethanacorollaryofhisinsistencethatabstractideasoughttobeputtoworkamongtheactualfactsofourworld,conflictswithPeirce’sfocusongeneralityandhabits,asPeirceconsistentlyemphasized–insteadofanypar-ticular,actualizedbearings–the“conceivablypracticalbearings”inwhich“theentiremeaningandsignificanceofanyconception”lies(EP2:145,1903).ThePeirceanformulationallowsthatconcep-tions,thoughalwaysconceptionsof“conceivablepracticaleffects,”“reachfarbeyondthepractical”;itisonlyrequiredthatwemain-tainaconnectionwithsomepossiblepracticaleffect(CP5.196/EP2:235,1903).Thus,Scholasticrealism,theprinciplethatgeneralityisoperativeinnature(andthatmodalitiesarethusinterpretedrealis-tically),isacentralbackgroundassumptionofpragmatism.Itisnotrequiredthatcertainspecific,particularconsequencesbeactualized;itisenoughthatsomegeneralhabitualpatternscanbeconnectedwithallofourmeaningfulideas.PeirceremarkedinaletterinDecember,1904,thatJames’s“Hu-manismandTruth”(reprintedinTheMeaningofTruth)haddistortedhisviews:Youhaveaquotationfrommewhichgreatlyastonishesme...:“Theseriousmeaningofaconceptliesintheconcretedifferencetosomeonewhichitsbeingtruewillmake.”21...IdonotthinkIhaveoftenspokenofthe“meaningofaconcept”whether“serious”ornot.Ihavesaidthattheconceptitself“is”nothingmorethantheconcept,notofanyconcretedifferencethatwillbemadetosomeone,butisnothingmorethantheconceptoftheconceivablepracticalapplicationsofit.(Perry1935/1936:II,432–3)PeircewassomewhathappierwiththewayJamesinterpretedhiminPragmatism,thoughhewishedthatJameshadlearnedtothinkCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n38samipihlstr¨om“withmoreexactitude”(Perry1935/1936:II,436–7).Itis,clearly,inthe“applications”thatJames’spragmatismtakesaturnawayfromPeirce’s.JamesdidnotpaymuchattentiontoPeirce’slaterdevelop-mentsofpragmatism;thelogicalspiritofPeirce’sthoughtremainedalientohim.ThisissomethingthatJamesadmitted,referringtohis“non-mathematical”mindand“slightinterestinlogic”(Perry1935/1936:II,427;lettertoPeirce,June5,1903).Hewrote:“Yourmindinhabitsatechnicallogicalthicketofitsownintowhichnoothermindhasasyetpenetrated”(Perry1935/1936:II,427n7;let-tertoPeirce,July10,1903;seealsoPerry1935/1936:II,680).PeirceagreedthatJames’sfailuretoappreciatehis(Peirce’s)pragmatismre-sultedfromhis(James’s)weakmathematicalandlogicalcapacities:James“hadnoheadforlogicatall”andthus“madethemanin[the]streetgetsomenotionsofwhatpragmatismwas”(NE3/1:192,1911).22TheoppositionsbetweenPeirceandJamescanbeseenasemerg-ingfromtheirdifferentformulationsandapplicationsoftheprag-maticmaxim.Peirce’sScholasticrealism,emphasisoncommunity,antipsychologisticviewoflogic,andemphasisonpragmatismasalogicalprincipleconflictedwithJames’snominalism,individual-ism,psychologicalorientation,andpsychologisticinterpretationofpragmatism.23Theseconflictsarenotunrelatedtohowtheyviewedthenotionofpracticalconsequences:forinstance,inanoteaddedin1893tothe1878paper(CP5.402n2),Peirceremarkedthatthemaxim,understoodasanapplicationoftheBiblicalrule,“Yemayknowthembytheirfruits,”oughttobeinterpretedcollectively,notindividualistically.Theemphasisonthecollectivenatureofscience,andofthehabituallyevolvingrationalitythathumanactionman-ifests,extendsthroughvirtuallyeverythingthatPeircewrote.TheindividualisticovertonesofJames’spragmatismwereasalientohimasJames’spsychologism.ThesedifferencesareespeciallyclearinPeirce’s1903HarvardLectures,oneofthemostsignificantdocu-mentsofhowPeirceresistedthepsychologizationofpragmatism.24Hesaidthathisownformulationsofthe1870sweretoopsycholog-icalandthathenolongerconsidersitsatisfactory“toreducesuchfundamentalthings[asthepragmaticmaxim]tofactsofpsychol-ogy,”because“mancouldalterhisnature”(EP2:140;seealsoCP5.28).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition39Intheselectures,Peircewasconcernedwithdemonstratingthetruthofpragmatismasamethodofthoughtandinquiry,connectingthemaximwithalmostallotherbranchesofhisphilosophy(i.e.,phe-nomenology,thecategories,logicofrelatives,theoryofprobability,thenormativesciences–logic,ethics,andaesthetics,theoryofin-ference,semiotics,andscholasticrealism).25Regardingthe“truth”ofpragmatism,James’sviewmay,however,havebeenmorecon-sistentlypragmaticthanPeirce’s.Arguably,Jamesappliedpragma-tismtoitself,treatingthepragmatistprincipleaspragmaticallytrue(cf.Conant1997,Pihlstrom¨1998a).Nologicaldemonstrationofitstruth,independentlyofpragmatism,wasneededorevenpossibleforhim;thepragmaticefficacyandthetruthofpragmatismwere(paceTurrisi1997b:28)prettymuchthesamethingforJames,thoughnotforPeirce.Themaximthatideasoughttobetestedpracticallyinthecourseofexperiencecoversthispragmatistideaitself.ThismetaphilosophicaldifferenceoverthestatusandprovabilityofthepragmaticmaximwasacorollaryoftheoppositionbetweenthelogicalandpsychologicalorientationsofPeirceandJames,respec-tively.WemaysaythatforJamestheevaluationofthephilosophicalroleofgeneralitiesorabstractideaswasamongtheapplicationsofthepragmaticmaxim,whereasforPeircetherealityofgeneralswasapresuppositionmakingpragmatismpossible.Jamescouldhavear-guedthatanysuchpresuppositionmustagainbepragmaticallyas-sessed.Peircealsothoughtthatthepragmaticmaximhadpragmaticconsequences;he,too,inhisownwayappliedpragmatismtoitself.ButthepointisthatJameswaswillingtoletpracticalconsequences–whichforhimconstitutedamoreopenandinclusiveclassthanthescientificallyfocusedconsequencesPeirceemphasized–determinethephilosophicalvalueofpragmatisminapragmaticmanner,inde-pendentlyofanypriorlogicaldemonstration.Peirce’spragmatismwassubordinatedtologic;accordingtoJames,whateverphilosophi-calvaluelogichaditwastobeexplainedonapragmaticbasis.Someofthesedifferencesmayhideamorebasicsimilarity.Thefactthat,inPeirce’sview,theorymustbedistinguishedfromprac-ticeandphilosophycannothelpusin“mattersofvitalimportance,”mightbeconsideredakeydifferencetoJames,butitmightalsoex-pressapartialagreement.Peircethought,withJamesratherthanagainsthim,thatvitallyimportantissuesshouldberesolvedbyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n40samipihlstr¨ominstinctandsentimentratherthanmereintellectualreflectionortheorizing.Evenso,thedistancefromJamesisconsiderablehere.ForJames(aswellasforDewey),mattersofvitalimportancedorequiresomethinglike“inquiry,”because“inquiry”isdefinedinhighlygeneralterms,morebroadlythan“scientificinquiry.”Ontheotherhand,evenPeircemaybeinterpretedashavingheldtheviewthat“themethodofscience”canbeappliedto“allrespectablesubject-matters”(Hookway2000:76–7).ThereisperhapsatensioninPeirce’spositioninthisrespect.26Anotherinterestingcomparison,notunrelatedtothepragmaticmaxim,resultsfromthequestionofwhetherPeirce’spresupposi-tionsofinquiry–e.g.,thattherearerealthingsindependentofwhatwethinkaboutthem–shouldbeinterpretedastranscendentallyes-tablishedtruthsormerehopes(cf.Hookway1998:§10;2000:6–7,39,109–10,185–6,190,296).Hookwayobservesthat,fromPeirce’spointofview,thefactthatsomethingisapresuppositionofinquiry,experience,orthoughtonlyprovidesareasonforhoping,notforbelieving,thatitistrue.Now,James’spragmatismmightleadustorejectthedistinctionbetweenthesetwoattitudesaspracticallyidle.Whatwehavetoadoptasasincerehopeonthebasisofwhatourinquiriesorexperiencespresupposeis,Jameswouldhaveurged,forusipsofactopragmaticallytrue.Thereis,inJames’spragma-tism,nopragmaticallysoliddistinctiontobedrawnbetweenhopesandbeliefsinthePeirceanway.Thisisespeciallyclearinthe“willtobelieve”doctrineandinJames’s“faithladder”(asformulatedinAPluralisticUniverse):thestatusofsincerehopesispragmat-icallyindistinguishablefromtheirstatusasconvictionsweneedinourlives,convictionsthatare,forthisreason,pragmaticallytrueforus.HereJameswasamoreradicalpragmatistwithrespecttotruththanPeirce.Onemightargueagainsthimbysayingthathopesorregulativeassumptionsarenottrueorfalseandshouldbedistin-guishedfrombeliefs.Callingsomethingaregulativeassumptionis“tomakeastatementaboutapractice,”aboutsomepractice(e.g.,inquiry)requiring“foritssensiblecontinuation”certainassump-tion(s)bythosewhoengageinit(e.g.,inquirers);thisisnottoclaimthatsuchassumptionsaretrue(Misak1991:140).ButoneoftheargumentscharacterizingJames’spragmatismasawholeisthattheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition41boundarybetweentheconceptsofbeliefandhopeisvague.ItispartofJames’s“humanizing”oftheconceptoftruthtoinsistthatwhatweneedtohopeinourlivesistrueinthepragmaticsense.Whatwe,quaagentsengaginginapractice,cannothelpassumingis,forus,true.Asourneedsandhopesmaychange,anelementofmutabilityisintroducedintothepragmatistconceptionoftruth–somethingthathorrifiedPeirce.ThisdisagreementcanperhapsbeexpressedbysayingthatPeirceendorsed,whileJamesdenied,Kant’sdistinctionbetweenpraktischandpragmatisch.27Theformer,Kantthought,isconcernedwithapriorimorallawsestablishedthroughthepracticaluseofreason;thelatter,insteadofbeingassociatedwithmorality,relatestothepurposivenatureofcognitioninrelationtosensibilityandisclosertowhatPeircehadinmindindiscussingtheexperimentalproce-duresofinquiry.Jamessawnopragmaticallymeaningfuldifferencehere.Heappliedthesamepragmaticmethodthatheusedinvariousphilosophicalproblemsmoremetaphilosophicallytothedissolutionofthecontrastbetweenpragmatischescientificexperimentaloper-ationsandpraktischemorallymotivatedconsiderations.FromtheJamesian(butsurelynotfromthePeircean)pointofview,moral(practical)issuesarealwaysalreadyatworkinourpragmaticassess-mentsoftheconceptionsofrealityweoperatewithinourpractices,scientificconceptionsincluded.Itispreciselythoseethicalconse-quencesofouractionsorhabitsofactionthatmustbetakenseriouslyinpragmaticevaluations.Weshouldnot,accordingtoJames,relyonanysciencevs.ethicsdichotomyifweattempttounderstandwhatpragmatismisallabout.Foranyonewillingtodefendtheroleofphilosophyinarationalconsiderationofethicalandpoliticalissues,theJamesianroute–inheritedbyDeweyandhisfollowers–isamaturationratherthanadistortionofpragmatism.Thepragmaticmaximremainstoonarrowifconfinedtoscientificmethodology.ThisistherelevantpracticaldifferencebetweenPeirce’sandJames’sapplicationsoftheirmethod.Wecanusethepragmaticmethoditselfreflexivelyandmetaphilo-sophicallyinordertodeterminewhatthedifferenceis.ItmaybesuggestedthatbythusapplyingpragmatismtoitselfweadoptamoreJamesianthanPeirceanapproach.SincethisproposalisvulnerabletoaPeirceancounterargumentemphasizingthescientificneedtoCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n42samipihlstr¨omstatepragmatismmoresharplyasalogico-semioticprinciplebasedonscholasticrealism,nobottomlineofthedebatecaneasilybereached.5.peirceanddeweyOurconceptionoftherelationbetweenPeirceandJamescanbeen-richedbystudyingtheviewsofsomeotherpragmatists.Theobvi-ousplacetobeginisDewey’sphilosophy,variouslylabelednotonlyas“pragmatism”but,moreoften,as“instrumentalism”or“exper-imentalism.”AsinJames’scase,IshallfocusnotonthebulkofDewey’swritingsbutonwhatDeweysaidaboutPeirceandonwhatPeircesaidabouthim.Dewey’s(1923)essay“ThePragmatismofPeirce,”supplement-ingPeirce’sChance,Love,andLogic,isstilloneofthebestbriefcharacterizationsofPeirce’spragmatism.28DeweycomparesJamesandPeirce,notingthestandarddifferences(nominalismvs.realismaboutgenerals,individualityvs.emphasisonthesocial).Peirce,ac-cordingtoDewey,emphasized“themethodofprocedure”morethanJames(307)andrejectedtheJamesian“appealtotheWilltoBelieve–under...themethodoftenacity”(308).InanotherpaperdiscussingPeirceandJames,Dewey(1922)pointedoutthatJames,beinga“hu-manist”ratherthanalogician,bothexpandedthepragmaticmethodbyapplyingittothetheoryoftruthandrestricteditbyemphasiz-ingparticularinsteadofgeneralconsequences.Later,Dewey(1946:156–7)referredfavorablytoPeirce’swayoflinkingtruthwiththedynamicsofscientificbelief–againsttheideaoftruthasa“fixedstructure”–andcalledPeirce“themanwhomorethananyothersinglepersonisthebegetterinphilosophyofanattitudeandoutlookdistinctivelyAmerican.”ThereareissuesonwhichPeirceandDeweywereclosertoeachotherthaneitherofthemwastoJames–inparticular,thesocialorientationofpragmatismandtheadvancementofscientificknowl-edge.29However,regardingtheissueofrealism,DeweywasclosertoJamesthantoPeirce.NeitherJamesnorDeweycouldacceptscholas-ticrealism;nordidtheyacceptPeirce’slogical,nonpsychologicalin-terpretationofpragmatism.30AsinJames,PeircefoundinDeweytheunfortunatetendencytopsychologizewhathehadpresentedaslogicalandnormativeprinciplesofscientificinference.OnJune9,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition431904,hewrotetoDewey:“YouproposetosubstitutefortheNor-mativeSciencewhichinmyjudgmentisthegreatestneedofouragea‘NaturalHistory’ofthoughtorofexperience....Idonotthinkanythinglikeanaturalhistorycananswertheterribleneed...”(CP8.239).Sincepragmatismwas,forPeirce,amethodforclarifyingideasand,becauseofitsrelationtothetheoryofinference,amaximoflogic,andsincelogicwasanormativescience,JamesandDeweywerefromPeirce’sperspectiveguiltyofaconflationoflogicaland(socio)psychologicalissues.Still,pragmatistslikeDeweyandMeadcanbeseenasdevelopingfurthersomebasicallyPeirceanthemes,particularlythereflexivityofhabitsofactionandofrationality(Kilpinen2000:ch.3).DeweydidnotentirelyrejectPeirce’srealismofgenerality:“...Peircehaslaidthebasisforavalidlogicaltheoryofuniversals.Itisthebusinessofleadingprinciples,asformulaeofoperations,toguideusinthedrawingofinferences.Theyaccomplishthistaskbyindicatingwhatqualitiesofthingsarecharacteristicofthepresenceofaspecifiedkindofobjectorevent”(Dewey1936:532).Butheinsistedthattheproblemoftherelationbetweenuniversalsandindividualsislog-icalratherthanontological(533),resistingthemetaphysicsofrealgenerality.Dewey(1946:228)alsoapprovinglyremarkedthatPeircewasthefirsttodrawattentiontotheimportanceoftheprinciplethat“[t]hegenericpropositionsoruniversalsofsciencecantakeeffect...onlythroughthemediumofthehabitsandimpulsivetendenciesoftheonewhojudges”andthattheyhave“nomodusoperandioftheirown.”OneofthemajordifferencesbetweenPeirce’sandDewey’scon-ceptionsofinquiryisrelatedtotheiraccountsoftruth.Aswasob-served,Dewey(likeJames)approvedofPeirce’s1878definitionoftruthastheultimateopinionofinquiry;yetDeweydidnotrelyontheideathattheremustbeauniquelimittoinquiry(Tiles1988:107).HeconceivedofthetasksofinquirymorepluralisticallythanPeircedid,remainingclosertoJames.InsteadofPeircean“purescience,”Deweyfavored“sociallyresponsiblescience”(Tiles1988:160).Thisbasicpositionregardingthesocialand,moregenerally,humanrel-evanceofinquirycanbefoundinvirtuallyallofhiswritings.Fur-thermore,Dewey(likeJames)wasmoreidealisticallyorconstruc-tivisticallyorientedthanPeirceinhisquiteexplicitviewthattheactionsofinquirersconstitutetheobjectsofknowledgeinsteadofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n44samipihlstr¨ombeinganswerabletopre-existingrealthings(cf.Dewey1929;seeShook2000).ItisundeniablethatPeirce’scommunity-drivenconceptionofin-quirywasacrucialbackgroundofDewey’s“instrumentalism”(cf.Dewey1922);moreover,evenwithinaDeweyan,morepluralisticconceptionofwhatourinquiriesaimatonemayretainthePeirceanviewthatthereisonedefiniteanswertobearrivedatregardinganyparticularquestion,providedthatinquirycouldbecarriedoutlongenough.ThepluralismassociatedwithJames’spragmatismandhisdoctrineofthemutabilityoftruthseemstobemoreextremethanthepluralismwecanreadintoDewey’saccountofinquiry.6.peirceandotherearlypragmatistsAmongtheinitialpragmatists,JosiahRoycewasanimportantcriticofJamesanddevelopedamixtureofpragmatismandHegelianideal-ism(“absolutepragmatism”)thatwasclosertoPeirce’sviewsthanweremostotherclassicalformulationsofpragmatism.G.H.MeadwasperhapstheoneclosesttoPeirceamongtheearlyfiguresofthetradition,especiallybecauseofhisinterestinsemiotics.C.I.Lewis,sometimesdescribedasthelastclassicalpragmatist,wasalsoclosertoPeircethantoJamesorDewey.Lewis’s“conceptualisticpragma-tism,”developedinMindandtheWorld-Orderin1929,perhapsliesbetweenPeirce’sandRoyce’sviews(Fisch1986:300–1).Theseprag-matistsremainoutsidethescopeofthepresentinquiry.31Ishall,inthissection,focusonF.C.S.Schiller,themostradicalsubjectivistamongtheclassicalpragmatists.PeircedidnotapproveofSchiller’smanneroftransformingprag-matismanymorethanheapprovedofJames’s:“...I,bynomeans,followMr.Schiller’sbrilliantandseductivehumanisticlogic,ac-cordingtowhichitispropertotakeaccountofthewholepersonalsituationinlogicalinquiries.”HisreasonfordismissingSchillerre-sembleshiscritiquesofJamesandDewey:“...Iholdittobeveryevilandharmfulproceduretointroduceintoscientificinvestigationanunfoundedhypothesis,withoutanydefiniteprospectofitshas-teningourdiscoveryofthetruth”(CP5.489,c.1906;cf.also5.494,c.1906).Schillerwasirresponsiblyunclearaboutwhathemeantby“thereal”(CP5.533,c.1905;cf.alsoCP8.319,undated),aswellasabouthisdefinition(influencedbyJames)oftruthassomethingthatis“satisfactory”(CP5.552,1906).InhisreviewofthebookCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition45PersonalIdealism(Sturt1902),towhichSchillerhadcontributed,PeircenotedthatSchiller“doesnotbelievethatthereareanyhardfactswhichremaintrueindependentlyofwhatwemaythinkaboutthem”(CN3,127).Althoughhedidnotcriticizethispositioninanydetailinthereview,mostofhiswritingsonpragmatismandthesci-entificmethoddefendsuch“hardfacts.”“Humanism,”inparticular,remainedunclearandunscientificinPeirce’seyes:[Schiller]doesnotwishustodevoteanyattentiontotheeffectsofconditionsthatdonotoccur,oratanyratenottosubstitutethesolutionofsuchaproblemforthetrueproblemsofnature....Ithinksuchtalkshowsgreatignoranceoftheconditionsofscience.[As]Iunderstandit,thisHumanismistobeaphilosophynotpurelyintellectualbecauseeverydepartmentofman’snaturemustbevoicedinit....Ibegtobeexcusedfromhavinganydealingswithsuchaphilosophy.Iwishphilosophytobeastrictscience,passionlessandseverelyfair.(CP5.537,c.1905)Toignoretheconditionsofscience–especiallyscholasticreal-ism,whichdrawsattentiontounactualizedgeneralities–was,forPeirce,toignorethecentralteachingsofhispragmati(ci)sm.AshewrotetoJames:“Thehumanisticelementofpragmatismisverytrueandimportantandimpressive;butIdonotthinkthatthedoc-trinecanbeprovedinthatway.Thepresentgenerationlikestoskipproofs....YouandSchillercarrypragmatismtoofarforme.Themostimportantconsequenceofit,byfar,...isthatunderthatconcep-tionofrealitywemustabandonnominalism”(CP8.258,1904;Perry1935/1936:II,430).32Apparently,Schiller,likeJames,appliedprag-matism(orhumanism)toitself,findingitapragmaticallyvaluablephilosophyinhumanaffairs,insteadofseekingaproofavailablefornonpragmatistsandpragmatistsalike.Apartfromthismetaphilosophicaldifference,Peirce’sdisagree-mentswithJamesandSchillerwerepartlyterminological.Inan-otherlettertoJames,Peircenotedthathewouldprefertheterm“anthropomorphism”toSchiller’s“humanism,”especiallyifitim-pliestheism(thoughherejectedtheideathatthetheisticGodmightbefinite).33Furthermore,“[p]luralism,”hesaid,“doesnotsatisfyeithermyheadormyheart”(CP8.262,1905;Perry1935/1936:II,434).Later,hementioned“pluralismgenerally,”alongwiththe“willtobelieve”andthe“mutabilityoftruth,”asanimplicationofJames’sandSchiller’spragmatismhedidnotaccept(EP2:457,1911).AsinthecaseofJamesandDewey,PeircefeltthatSchiller’sCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n46samipihlstr¨ompsychologismandnominalismweretheoppositeofthetruespiritofpragmatism:“WhenyousaythatLogicalconsequencescannotbeseparatedfrompsychologicaleffects,...youaremerelyadoptingamodeofexpressionhighlyinconvenientwhich...canonlyconfuse,anysoundargumentation.Itisapartofnominalismwhichisut-terlyantipragmatistic...”(CP8.326;lettertoSchiller,September10,1906).GivenPeirce’sremarksontheindistinguishabilityofhisviewsfromJames’s,itseemsthatPeircewasmorecriticalofSchillerthanofJames.Why?IsthereadifferencebetweenJames’spluralisticprag-matismandSchiller’spersonalistichumanism,althoughJamesoftenappearedtoendorseSchiller’sviewsontruthandontheconstitutionofrealitythroughhumanpractices?ThisissuemustbeleftforJamesandSchillerscholarstosolveonanotheroccasion.WecansaythatSchiller,evenmoreradicallythanJames,distancedhimselffromPeirce’slogical,scientificprag-matism.HeadmittedthatPeircewastheonewhoinventedpragma-tism,butaddedthat“itwouldseemtofollowfrompragmatistprinci-plesthatadoctrinebelongstohimwhomakesaneffectiveuseofit”(Schiller1903:27n1).Schiller(1907:ix–x)ignoredPeirce’scriticismofJames’sandhisownviewssimplybyremarkingthatPeirce’s1905Monistpapers“haveshownthathehadnotdisavowedthegreatPrag-maticprinciplewhichhelaunchedintotheworldsounobtrusivelynearlythirtyyearsago.”Schiller(1907:5)thoughtthisprinciplewas“thegreatesttruism”:itisclearthattheconsequencesofaclaimareusedtotestthetruthoftheclaim.“Humanism”isabroaderdoctrinethanpragmatism(1907:5n1).Schilleradded,though,thatPeircehadprivatelyassuredhimthat“fromthefirsthehadperceivedthefullconsequencesofhisdictum.”NeitherJamesnorSchillerwasresponsivetothecritiquePeircelaunchedagainstthem,althoughthey,asleadingfiguresofthemove-mentfoundedbyPeirce,perhapsoughttohavebeen.This,onemightspeculate,mayhavebeenoneofthereasonsPeirce’spragmatismwasonlyslowlyreceivedinthephilosophicalcommunity.7.peirceandlaterpragmatismPeirceandotherclassicalpragmatistsinfluencedlaterthinkersinmanyways.Amongcentraltwentiethcenturyphilosophers,LudwigCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition47Wittgensteinisoneofthemostinterestinginrelationtotheprag-matisttradition,althoughhewasinfluencedmorebyJamesthanbyPeirce.Peirce’sinfluenceonWittgensteinhasbeenshowntogopri-marilythroughFrankRamsey.34UnlikeWittgenstein,postpositivistphilosophersofscience,especiallyscientificrealists,havebeenlessaffectedbyJamesandDeweyandmoreattachedtoaPeirceandoc-trineofthefinalopinionofthescientificcommunityasthemeasureoftruth(cf.Niiniluoto1999).Thereare,furthermore,contemporarypragmatists(e.g.,Haack1998;Rescher2000)whoseviewscanberegardedas“Peircean,”butdespitethegrowingindustryofPeircescholarship,itseemsthatthemostoriginalthinkerstobeclassifiedaspragmatiststodayhavebeenmorestronglyinfluencedbyJamesandDeweythanbyPeirce(e.g.,Putnam,Rorty,andothers).YetwecanfindconflictingattitudestoPeirceevenamongtheseJamesian–Deweyanneopragmatists:thereisagreatgulfseparatingPutnam’s(1990:ch.18)appreciationofPeirce’sroleasoneofthefoundersofmodernlogicfromRorty’sinfamouswayofrestrictinghiscontribu-tiontothepragmatisttraditiontohishavinggivenitthenameandhavingstimulatedJames(seeRorty1982:160–161).Putnam(1994,1995a),likeRorty,seesJamesandDeweyasthetwogreatpragmatistshewishestofollow.Hereferstohimselfasonewhoattemptstorevivetheideathattruthis,“insomeway(notinPeirce’sway,butinamorehumanlyaccessible,modestway),anidealizationofthenotionofwarrantedassertibility”(Putnam1990:223),andpointsoutthat“Peircewascertainlywronginthinkingthattruthcanbedefinedaswhatinquirywouldconvergetointhelongrun”(Putnam1994:152).Still,therearePeirceanelementsinPutnam’spragmatism:hisattempttodefinetruthinepistemicterms(Putnam1981,1990)isnotunlikePeirce’snotionoftheideallimitofscien-tificopinion.35InRorty’sneopragmatism,suchPeirceanelementshavedisappeared,sinceinRortywecanhardlyfindanysincerecon-cernwithtruthorinquiry.RortyalsomisusesPeirceanideasbyre-gardingthepragmatisttraditionasbasedonwhathecalls“antirep-resentationalism.”Itisoddtoclaimthatthefounderofsemioticsalsofoundedanantirepresentationalistphilosophy.YetRorty(1998)maintainssomethingfromthePeirceanaccountoftruth:insistingonthe“cautionary”useof“true,”hecomesclosetothekindofreg-ulative“absolute”truththatJamesandDeweyconsideredvaluableinPeirce’sphilosophy,viz.,anotionoftruthwhosepointisthat“itCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n48samipihlstr¨omisalwayspossible(andfrequentlylikely)thatfurtherinquirieswillexercisetheirpowersof‘retroactivelegislation’andthusrequireustoabandonourcurrentconclusions”(Hookway2000:69).Weusethenotionoftruthpartlyinordertoremindourselvesofourfallibility,sincethenotionoferrorseemstopresupposethenotionoftruth(seeMisak2000).Wemayalwaysbemistakeninouropinions,andsince(asPutnam,Rorty,andmanyothershaveargued)wecannotdirectlycompareourbeliefsandtheoriestoanunconceptualizedpractice-,perspective-,anddiscourse-independentreality(totheworldinit-self),thereisnohigherauthoritythan“ourfutureselves”(touseoneofRorty’sfavoriteexpressions)todeterminewhetherwehavebeenmistakenornot.ThedifferencebetweenPeirce’sandRorty’spragmatismisclear,however,whenthePeirceaninquirerpointsoutthatourfalliblebe-liefsshouldaddressanunlimitedcommunityofinquirers(Hook-way2000:70).Rortyhasnouseforsuchanotion,asheinsistsonthelimitedandcontextualnatureofhumanprojects,includingin-quiries.HereRortyismuchclosertoJamesandDewey.Sciencewas,forJames,essentiallyinstrumental,andthepracticalusetowhichscientifictheoriesaretobeputdoesnotrequirethatthosetheoriesbeinterpretedintermsof“absolutetruth”(Hookway2000:73–74).Rortyappearstoholdanequallyinstrumentalistconception.AmongcontemporaryPeirceans,Haack(1993,1998)hasmostvigorouslyattackedRorty’sversionofpragmatism.ShearguesthatRorty’sneopragmatismamounts,inPeirce’sterms,toapseudo-inquirycarriedonina“literaryspirit,”ora“fakereasoning”ratherthangenuinetruth-seeking.Thus,RortyfailstofollowPeirce’s“firstruleofreason,”therulethat“inordertolearnyoumustdesiretolearn”(seeCP1.135/EP2:48,1898).FromHaack’sperspective,Rorty’spragmatismisavulgarizationofPeirce’s.36PeircewouldhardlyhaveanydifficultiesinjudgingRortyasoneoftheabusersoftheword“pragmatism,”asoneofthosewhomisapplytheterm–andthedoctrine–in“literaryjournals.”Haack’sandotherPeirceans’critiquesofRortyareamongthemostimportantrecenttwistsinthepragmatisttradition.Yet,hadPeirce’soriginalviewsneverbeenex-tended,reinterpreted,andperhapsinsomecasesevenmisapplied,hadpragmatistideasconcerningtruthandrealityneverbeencar-riedintotheRortyanantirealistandethnocentristextremes,theCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition49pragmatisttraditionmightbepoorerthanitis–althoughwecannotknowforsure.8.conclusion:realismandidealismOnlythefuturecanshowhowmuchthePeirceanconceptionofphilosophyasinquirywillberespectedinthepragmatisttradition.Peirce’spragmatismisoflastingvalue,butJamesandDeweyde-velopedindependent,thoughcontroversial,versionsofpragmatismthatarelessrealisticallybiased.Theirconstructivisticandhuman-isticviewscan–contrarytowhattheythemselvesclaimed–beinterpretedasvariationsofKantianidealism,whichperhapscannotbeconsistentlydoneinPeirce’scase.Theideathattheobjectsofknowledgeareinasenseconstructionsbytheknowingsubject,orbythesubject’sactionsinthecourseofinquiry,anideathatPeircerejectedbutJames,Dewey,andSchillerinsomesenseendorsed,isafundamentallyKantianidea.Inneopragmatism,itistheJamesian–DeweyanstandpointthatdominatesoverthePeirceanone,althoughPeirce’sthoughtismoreinfluentialinthephilosophyofscience,especiallyinthetraditionofscientificrealism,aswellasinsemioticsandcommunicationstudies.37Insofaraspragmatismisconsideredanimportanttradi-tiontoday,itislargelybecauseofitspromisetotakeseriouslythevitalquestionsofhumanlife,ratherthanmakingthedistinctionbe-tweentheoryandpracticethatPeircemade.Forexample,James’spragmatismoffersamorepromisingagendaforphilosophersofre-ligionseekingtounderstandreligiousexperiencesandthepossible“pragmatictruth”(orwarrant)ofreligiousbeliefsthanPeirce’sevo-lutionarymetaphysics.AndalthoughtherelevanceofPeirce’sprag-matism,especiallyitshabitualconceptionofrationality,tosocialtheoryhasbeenemphasized(Kilpinen2000),itiseasiertofinddi-rectlyrelevantsocial-theoreticalviewsinDeweythaninPeirce.Peirce’sandhisfollowers’interpretationsofpragmatismareunitedbycertainquestionstheirviewsseemtoleaveunsettled.Inparticular,theproblemofrealismvs.idealismisunavoidableinthepragmatisttradition.Itislegitimatetoobjectthatthiscontrastisnotappropriateinadiscussionofpragmatism,aspragmatistshaveattemptedtotranscendtheoscillationbetweenrealismandidealismCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n50samipihlstr¨ominsteadofdefiningtheirviewsintermsofit.Butitisequallylegiti-matetousethistraditionaloppositiontouncoverthetensionsthatremaininpragmatists’peculiarcombinationsofrealismandideal-ism(evenifwemayintheendagreethatthecontrasthasbeentran-scended).Whatmakespragmatismphilosophicallyinterestingisitstendencytoresultinfruitfulalbeitnoteasilyresolvablestrugglesbe-tweenrealismandidealism.38NeitherPeirce’s,James’s,norDewey’s(northeirmorerecentfollowers’)viewscanbesimplydescribedasrealisticoridealistic.TheyareascomplexdoctrinesasKant’s,whocombinedtranscendentalidealismwithempiricalrealism.InhisessayonPeirce,Deweyconcluded:“Donotalargepartofourepistemologicaldifficultiesarisefromanattempttodefinethe‘real’assomethinggivenpriortoreflectiveinquiryinsteadofasthatwhichreflectiveinquiryisforcedtoreachandtowhichwhenitisreachedbeliefcanstablycling?”(1923:308)Thissuggestion–thatthe“real”shouldnotbedefinedas“somethinggivenpriortoreflectiveinquiry”–leadstotheelusivenessofthecontrastbetweenrealismandidealismthatcanbefoundthroughoutthepragmatisttradition.Doesinquiryproducetherealbybeingforcedtoreachforit?Howindependentlydoestherealexistbeforeinquiry,ifitisnot“given”priortoit?Andhowmeaningfulisthisworryit-self?Althoughweshouldnotconfusetheproblemsweencounterinformulatingtherealismquestionwiththeopennessoftheques-tionitself,thefactthatacertainissueishardtoformulateisanindicatorofitsgenuineopenness.Throughpragmatists’writings,theproblemofrealismiscontinuouslytransformed,butneverfullysettled.Forexample,Putnam(1992a:73)classifiesPeirce’sscholas-ticrealismasaspeciesofmetaphysicalrealism,theunpragmatisticviewthatwecandiscoverNature’sown“joints”–aviewwhoserejectionheregardsasavirtueratherthanaviceinJamesandDewey.Peirceandotherpragmatistswerepresumablyawareoftheirdiffi-cultiesinreconcilingtheprimafacieconflictingdemandsofrealismandidealism.Peircecharacterizedtruthas“[t]heopinionwhichisfatedtobeultimatelyagreedtobyallwhoinvestigate”andreality,or“thereal,”as“theobjectrepresentedinthisopinion”(CP5.407/W3,273,1878).Butthereal,healwaysemphasized,mustbethoughtofassomethingthatis“independentofthevagariesofmeandyou”(CP5.311/W2,239,1868;seealsoCP5.405/W3,271,1878;CP5.430,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition511905).Traditionalrealistsrequirethatthenatureofrealitybeabso-lutelyindependentofour–evenourmostconsidered,collective,or“final”–opinions.Peircethoughtthatreality“dependsontheulti-matedecisionofthecommunity”(CP5.316/W2,241,1868).Claimslikethisseemtomakehispragmati(ci)smambiguousbetweenreal-ism(connectedwithacorrespondenceanalysisoftruth,accordingtowhichthefinalopinionofinquirycorrespondstothewaythingsare)andidealism(connectedwithacoherenceorconsensusaccountoftruth).Peircealsosaidthatreality,whilebeingindependentof“whatyouorIoranyfinitenumberofmenmaythinkaboutit,”maynotbeindependent“ofthoughtingeneral”(CP5.408/W3,274,1878;cf.alsoCP7.336,1873).39Thisreferenceto“thoughtingeneral”intheconstitutionofrealityinsomesensemakeshimanidealist.Re-alizingtheinstabilityofhisposition,Peirceremarkedthattheclaimthat“[t]heobjectoffinalbeliefwhichexistsonlyinconsequenceofthebelief,shoulditselfproducethebelief”soundsparadoxical,butthatthisisnottosaythattheobjectofthebelief“beginstoexistfirstwhenthebeliefbeginstoexist”(CP7.340,1873).EventhoughthePeirceanpragmatistcharacterizesinquirynonpsychologicallyinilluminatingthenotionsofrealityandtruthintermsofthefinalout-comeofinquiry,itisnoteasytomakesenseoftheideaofinquiryasagenuinediscovery,ifinquiry,fatedtoleadtoaconsensusofopin-ioninthelongrun,constitutesthewaytheworldis(Hookway1985:37–9).Thesecondaryliteratureisfullofattemptstoreconcilethetensionbetweenrealismandidealism.Forexample,CarlHausman(1993)endorsestheideathatPeircewasa“metaphysicalrealist”(althoughpreferablytobecalledan“evolutionaryrealist”),anddefendsthisviewagainstphilosopherslikePutnam.WhilePeircerejectedthe“spectatortheoryofknowledge”(asallpragmatistsdid),heinsistedthatthereareconditionsofinquirythatwerenevermadebyus,thatthereis“resistance”inourexperience(224–5).Butisitpossibletorejectthespectatortheory,denyingthattheobjectofknowledgeis“given”tous,andyetclaimthatthereisanexternal,independentworldthatistheobjectofknowledge?Iftheobjectofknowledgeisconstitutedasthefinaloutcomeofinquiry,iftruthistobeequatedwithbeliefthatcannotbeimprovedonthroughfurtherinquiry,itishardtoseehowtheworldcanbetotallyindependentofusinthesenseinwhichrealistsclaimittobe.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n52samipihlstr¨omItwouldbetoosimpletosaythattheprogressivedynamicsofscience–scientificinquirers’collectivebelief-fixation–decides,de-termines,orconstructstheworld.Itwouldcertainlybetoosimpletoascribethisviewtoanyofthepragmatists.Butitwouldalsobetoosimpletosaythatrealityexistsinaready-madeform,asa“thinginitself,”independentoftheinquirers’habitsofaction.Peirce’sprag-matism,andthecountlesspost-Peirceanversionsofpragmatism,allthewayuptoandincludingcontroversialcontemporaryfigureslikePutnamandRorty,dealwithortrytounderminethisoppositionbetweenrealismandidealism.Perhapsthequestion,“IsPeirceas-suminganexternal,objectiveworldindependentofinquiry,oristheworldconstitutedthroughtheprocessofinquiry?”isabadquestion,butitremainstobedeterminedexactlyinwhatsenseitisabadques-tionandwithwhichquestionsitshouldbereplaced.Forinstance,onemayaskwhetherPeirceheldanonepistemicoranepistemicconceptoftruth.40Truthisepistemicinthesenseofbeingneces-sarilytiedtoourinquiriesbutnonepistemicinthesenseofbeingaboutarealitywedidnotbuildup.Accordingtophilosophersoper-atingwithtraditionalnonpragmaticdichotomies,thisishopelesslyambiguous;accordingtopragmatists,wedonothaveambiguitiesherebutcomplexitythatcannotbeavoided,ifwewishtoobtainanadequateconceptionoftruthandrealism.Roycemadeanimportantpointin1881,wheninalettertoJamesheasked,“Doyouordoyounotrecognizethisrealityofwhichyouspeakas...independentoftheknowingconsciousness?”observingthesamehesitationandambiguityinPeirce’s1877and1878papers:“[He]seemstoregardrealityasforusmerelytherepresentativeofourdeterminationstoactsoorso,andofourexpectationsthatweshallsucceedifwedoso....Yet[he]isnotcontentwiththis,butcontinuallyappealstothetranscendentrealityasjustifyingourde-terminationandourexpectation”(Perry1935/1936:I,792).TheissueRoyceidentifiedis,essentially,aKantianone,reflectingtheKantianbackgroundofthepragmatisttradition.InawayPeirce,likemostotherpragmatists,wasanempiricalrealistaboutthe“realthings”thataretheobjectofthefinalscientificopinion,whileremainingatranscendentalidealistabouttheconstitutionofthesethings,andoftheirobjectivity,groundedintheintersubjectiveactionofthesci-entificcommunity.41CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition53Although(orbecause)no“solution”toourKantianissuehasbeenreached,IhopeIhavebeenabletoproduceamodestcontributiontothepragmatisttraditioncharacterizedbytheirreduciblecomplexityoftherealismvs.idealismopposition.42notes1.SeeCP8.253,editors’note.James’slecture,oftenconsideredthebe-ginningofthepragmatistmovement,waspublishedintheUniversityofCaliforniaChronicle1(1898)andismosteasilyfoundas“ThePrag-maticMethod,”inEssaysinPhilosophy(1978:123–39)orasanappendixtoPragmatism(James1907[1975]:257–70).2.NoramIconcernedwiththeMetaphysicalCluborwiththebroaderhistoricalbackgroundofpragmatism;cf.Menand(2001).3.Thisisnot,however,ahistoricalstudyonthechangesthattookplaceinPeirce’sphilosophy.SuchdevelopmentalquestionsaredealtwithelsewhereinthisCompanion.4.Anotherinteresting,somewhatbitterpassageisthis:“Tospeakplainly,aconsiderablenumberofphilosophershavelatelywrittenastheymighthavewrittenincasetheyhadbeenreadingeitherwhatIwrotebutwereashamedtoconfessit,orhadbeenreadingsomethingthatsomereaderofminehadread.Fortheyseemquitedisposedtoadoptmytermpragmatism....Icannotfindanydirerfaultwiththenewpragmatiststhanthattheyarelively.Inordertobedeepitisrequisitetobedull.//Ontheirside,oneofthefaultsthatIthinktheymightfindwithmeisthatImakepragmatismtobeameremaximoflogicinsteadofasublimeprincipleofspeculativephilosophy.”(EP2:134,1903;cf.CP5.17–18.)SeealsoCP6.482,6.490,1908.5.Forthe“twopragmatisms”image,seeApel(1981),Mounce(1997),Haack(1998),Rescher(2000),andMisak(2000).Accordingtothesecom-mentators,Peirce’spragmatismwasgradually,throughmisapplicationsanddistortions,transformedintoRorty’scompletelyun-Peirceanneo-pragmatism.6.Inordertoobtainagoodoverallpictureofpragmatism,itisadvisabletofocusonthosepragmatists(James,Dewey,Schiller)whoseviewsweredifferentfromPeirce’sratherthanonthose(Royce,Mead,Lewis)whomoreorlessagreedwithhim.7.Onthesedifferences–realismvs.nominalism,truth,formulationsofthepragmaticmaxim,etc.–seePerry(1935/1936:II,ch.75),Thayer(1968),andHookway(2000).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n54samipihlstr¨om8.ForPeirce’sformulationsofScholasticrealism,seeCP5.430–3,1905;5.453ff.,1905;5.470,c.1906;5.528,c.1905;8.7–38,1871,aswellasthe1898lectures,ReasoningandtheLogicofThings(RLT).OnlyapartofPeirce’simportant1905–1907writingsonpragmaticism(inwhichScholasticrealismisamajortopic)canbefoundintheCollectedPa-pers;amorecomprehensiveselectionisincludedinEP2:chs.24–8.TheequallyimportantearlyBerkeleyreview(1871)canalsobefoundinW2,462–87,andinEP1:ch.5.OntheroleofScholasticrealisminPeirce’sthought,seeApel(1981),Skagestad(1981),Margolis(1993),Haack(1998),andPihlstrom(¨1998b).9.FordiscussionsofPeirce’stheoryoftruth,seeMisak(1991)andHook-way(2000).10.SeeHookway(2000:68–69);onthe“Peirceanstrain”inJames’stheoryoftruth,seePutnam(1997:167–71);onDewey’sapprovalofPeirce’sdefinition,seeTiles(1988:106)andShook(2000:130).11.Cf.,e.g.,CP5.504n1,c.1905.Peircereferstohis1868writingsintheJournalofSpeculativePhilosophy(cf.CP5.213ff.;thesecanalsobefoundinW2,chs.21–3,andinEP1:chs.2–4;seealsoFisch1986:118).12.Suchalertnesswas,however,hardlydeniedbyJames(cf.Pihlstrom¨1998a:ch.6).13.James(1909b[1977]:153–4)referredfavorablytowhatheregardedasaffinitiesbetweenPeirceandBergson.ThismusthaveannoyedPeirce(seealsoNE3/2:836,1909).14.See,however,Misak’scontributiontothisCompanion,“C.S.PeirceonVitalMatters.”AlessPeirceanversionofpragmaticmoralrealismisdefendedinPihlstrom(¨2003).15.Asimilar–ratherunpragmatic–theory/practicedistinctionisatworkinPeirce’s1903lectures.Cf.alsoPutnam(1992a:55–8).16.Forrelevantcorrespondence,seePerry(1935/1936:II,418–21).PeircenotedonJanuary4,1898,thathisfirstlecturewouldbeabout“vitallyimportanttopics,”“showingthatwheretheyare‘vital’thereislittlechanceforphilosophyinthem”(421).Peirce’slectureswerestimulatedbyJames’swilltobelievetheory(Houser1998:xxi).17.Seigfried’sreferenceistoJames(1907[1975]:18)and(1909a[1978]:28).18.Thisextendstotheirviewsonreligion.Peircemayhavethought,withJames,thatwehaveahumanlynaturaltendencytobelieveinGod(seeCP6.487,1908;Roth1965).InalettertoJames’ssonHenryafterWilliam’sdeathin1910,PeircesaidthatTheVarietiesofReligiousEx-periencewasthebestofJames’sbooks(Perry1935/1936:II,286).Theremayevenbeaversionofthe“willtobelieve”doctrineinPeirce(seeCP5.60/EP2:156,1903;cf.Gavin1980;Hookway2000:19;Kilpinen2000:117).Gavin(1980)arguesthatPeirceemployedsuchadoctrineinCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition55hisidentificationoftherealandtheknowable–inhisrejectionofanincognizableDingansich(CP5.257/W2,208ff.,1868).19.SeeJames(1909a[1978]:31).20.SeealsoCP5.422,1905;5.438,1905;5.468,c.1906;6.481,1908;8.191,c.1904.Alongerformulationisthefollowing:“Pragmatismistheprin-ciplethateverytheoreticaljudgmentexpressibleinasentenceintheindicativemoodisaconfusedformofthoughtwhoseonlymeaning,ifithasany,liesinitstendencytoenforceacorrespondingpracticalmaximexpressibleasaconditionalsentencewithitsapodosisintheimperativemood”(CP5.18/EP2:134–5,1903).FordiscussionsofPeirce’smaxim,seeApel(1981:ch.4),Skagestad(1981:ch.3),andHookway(1985:ch.8);onJames’sinterpretation,seeHingst(2000).Recentscholarshipap-pearstoshowthatPeircehadenunciatedthepragmaticprincipleattheMetaphysicalClubnotlaterthanNovember1872(seetheeditors’introductiontoW3,xxixff.).21.AsPerry(1935/1936:II,432n11)notes,Jamesdoesnotinfactquotethispassage.Itisaparaphrase,thoughinaccuratebyPeirce’slights.22.SomeofPeirce’slongletterstoJameswerefulloflogicalandmathe-maticalformalisms–apparentlyPeircetriedtoteachhisfriendsomemathematics(seeNE3/2:788–878).TheselectionofPeirce’sletterstoJamesinEP2:492–502isfocusedonsemiotics.23.Onpragmatismasalogicalmethod,cf.furtherTurrisi(1997a,1997b)andHookway(2000:286ff.).24.Turrisi(1997a:9)remarksthatthetitleofthelectures,PragmatismasaPrincipleandMethodofRightThinking(seeCP5.14–212/EP2:chs.10–16),wasgivenbyJames.JamesprobablyauthoredtheHarvardCrimsonannouncementonPeirce’slectureonMarch26,1903,whichdefinedpragmatismasaphilosophicalsystemviewingphilosophicalquestions“primarilyfromthestandpointoftheirpracticalbearinguponlife”(Turrisi1997a:10;1997b:23).25.OnthequestionofwhetherPeircewasableto“prove”pragmatism,seeHouser(1998)andHookway(2000:ch.12).26.See,again,Misak’scontributiontothisvolume,whichseekstoshowhowPeirce“buildsinstinctintothescientificmethod.”Peircearguednotonlythatweshouldnottrustscienceinvitallyimportantmattersbutalsothatbelievinghasnoplaceinscience(CP5.60/EP2:156,1903).Itisproblematictofitsuchaviewwithhisownbelief/doubttheoryofinquiry.Cf.Hookway(1998:§5;2000:ch.1).27.SeeKant(1781/1787:A800/B828,A823–4/B851–2);forPeirce’swayofmakingthedistinction,seeThayer(1968:138–139).28.DeweyalsooccasionallyreviewedPeirce’swritings,forexample,thefirstvolumeofPeirce’sCollectedPapersinNewRepublic68(1932).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n56samipihlstr¨om29.Dewey’sandPeirce’saffinitieswerenotedearly.ThepsychologistJamesRowlandAngellwrotetoJamesin1898thatPeirce’spragmatismis“surprisinglylikewhatDeweyisdrivingat.”(Seetheeditors’notestoJames1907[1975]:146.)30.Deweyalsocriticized(inalettertoJamesin1903)Peirce’smetaphysical“hypostatizingofchance”(Perry1935/1936:II,523).31.Cf.thediscussionsofPeirce’srelationtolaterpragmatistsbyThayer(1968),Kilpinen(2000),andRescher(2000).Theinfluenceofpragmatismbecame,afteritsmajorearlyclassics,alsogeographicallysodispersedthatitwouldbeimpossibletogiveanyevennearlyexhaustivesurveyhere.Forexample,inItaly,therewerebothJamesianpragmatists(e.g.,Papini)andPeirceanones(VailatiandCalderoni)(Fisch1986:295–6;seePerry1935/1936:II,ch.84;Shook1998;andPeirce’sownnote,N3:233–4,1905).32.HerePeirceimpliesthatpragmatismisa“conceptionofreality”(andnotameremethodofthought).FromJames’sorSchiller’sperspective,thesemaybepracticallyindistinguishable.33.“WilliamJamesandF.C.S.SchillermaintainthatGodandeverythingelseisfinite–adoctrinesomepeoplecallpragmatism.Tomeitisasabhorrentasitisincredible.”(NE3/2:786,1906.)PeirceremarkedtoJamesthatpragmatismdoesnotrequirerenouncingideasabouttheAbsolute(NE3/2:871,1909).OneofJames’sapplicationsofpragmatismwashiscriticismofthenotionoftheAbsolute.34.OntherelationbetweenPeirceandWittgenstein,includingRamsey’sinfluence,seeThayer(1968:304–5),Bambrough(1981),Gullvag(˚1981),Nubiola(1996),andCrocker(1998).InadditiontohisconversationswithRamsey,WittgensteinmusthavebeenacquaintedwithPeircethroughhisreadingofJames’sVarieties.35.ForacomparisonbetweenPeirceandPutnam,seeHookway(2001).Hookwaypointsout(1)thatitisnotnecessarytointerpretPeirceassubscribingtotheideaofan“absoluteconceptionoftheworld”;(2)thatPeircemaybeseenassharingJames’s(andPutnam’s)viewthatrealitycanberelativetohumanthought,interests,ordesires,sincetheconceptsbymeansofwhichweclassifythingsare“sensitivetoadistinctivehumanperspective”;and(3)thatPutnam’s(1994)“naturalrealism”iscomparabletoPeirce’s“criticalcommonsensism.”ItisanopenquestionwhetherPutnam’sdefenseofcommonsensewouldbesufficiently“critical”byPeirce’slights.36.Thayer(1996)suggeststhataneopragmatismwhichseesobjectsas“so-cialconstructs”mighthavebeenregardedasanexampleofthe“apriorimethod”ofbelief-fixationbyPeirce–asoneofthemethodsPeircefoundinferiortothescientificmethod(CP5.382ff./W3,252ff.,1877).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPlaceinthePragmatistTradition5737.Amongmajorphilosophersofscience,IsaacLevi(1991),inhisstudiesonthedynamicsofscientificbelief,hasbeenoneofthemostimportantfollowersofPeirce.OnPeirce’srelevanceforcommunicationstudies,seeBergman(2000).38.Tensionsliketheonebetweenrealismandidealismmaybeconsid-eredunfruitful.Ibelieve,however,thatsuchtensions,dilemmas,andopenissuesareextremelyimportantinphilosophy.Theykeepourphilo-sophicalwonderalive.Thisattitudetophilosophicalquestionsrequiresthatonevaluesthequestionsthemselves,theiropennessandeventheirunclarity,morethanthe“results”thatmaybeachieved,inawayresem-blingscientificinquiry,inthecourseofphilosophizing.SeePihlstrom¨(1998a).39.SeethedraftsonthenotionofrealityinPeirce’s1872–1873investiga-tionsoflogic(W3,28–61).OnPeirce’sattempttocombine“semeiotic”or“discursive”realismwithidealism,seeHouser(1992).40.Cf.Putnam(1981:ch.3)and(1990).41.WhilePeircemovedfromaviewresemblingtranscendentalidealismtoamorerealisticposition,hemayhavecomeclosertotranscendentalidealisminhislatestthought(Hookway1985:117).IhavediscussedtheKantiannatureofthepragmatisttraditionelsewhere(Pihlstrom¨1996,1998a,1998b,2003).ThecommonKantiantensionsharedbythepragmatismsofPeirce,James,Dewey,Putnam,Rorty,andothersisagoodreasontorejectthepopulardualismsbetween“twopragma-tisms.”SomescholarswhorecognizetheKantianbackgroundofPeirce’sthought–e.g.,Christensen(1994),influencedbyApel–arecommittedtothissimplisticpicture,assumingthatPeirce’spragmati(ci)smisfun-damentallydifferentfromtheJames–Rortylineage.42.IamgratefultoCherylMisakforhavinginvitedmetocontributetothisCompanionandforherenormouslyusefulcommentsonearlierdrafts.IalsowishtothankMatsBergman,SusanHaack,LeilaHaaparanta,PeterH.Hare,ErkkiKilpinen,IlkkaNiiniluoto,JaimeNubiola,SamiPaavola,RichardS.Robin,andKennethR.Westphal,allofwhomhavetaughtmealotaboutPeirceandpragmatism.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\njohnboler3PeirceandMedievalThought1IIntroductionPeirce’sknowledgeofandattitudetowardmedievalthoughtwasclearlyunusualamonghispeers,2anditcontainssomeinterestingsurprises.Hiscriticalremarks,ofcourse,whilemorecolorfulthanmost,arenotunexpected.3Itisnotworthourwhile...toascertainwhattheschoolmastersofthatdegenerateageconceivedmathematicstobe.(CP3.554,1898)[A]beastlikesuperficialityandlackofgeneralizingthoughtspreadslikeapalloverthewritingsofthescholasticmastersoflogic....(CP1.561,1907)Moreover,heseemstothinktheentireerawas,withthepossibleexceptionofRogerBacon,lackinginascientificappreciationorout-look.4Peirce’scriticismofthelaterdecadentscholasticismisofspecialinterestforitsreferencetothefollowersofScotuswhohadgainedcontroloftheuniversitiesandweregiven(bythehumanists)thesarcastictitleof“dunses”or“dunces”(CP1.17–18,1903;2.166–8,1902).They“setuptheiridlelogicaldistinctionsasprecludingallphysicalinquiry”(CP6.361,1902).Andwhiletheywereontherightsideoftherealist–nominalistissue,“theirdunsicaloppositiontothenewlearningandtheirdreadfulcorruptionoftheuniversitydisgustedthenewmen”(CP7.666,1903).5Atthesametime,Peirce’sputdownofthehumanists’reactiontothescholasticsisifpossibleevenmorerude.TheDunsesdefendedtheirposition“withalogicalaccuracy,bornofcenturiesofstudy,58CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought59withwhichthenewmenwereutterlyincapableofcoping”;theyneededtoformulateobjectionstotheDunses’positions,butitwas“abusinessforwhichtheywereutterlyunfitted”(CP6.361).6Thehumanists,hesays“wereweakthinkers”(CP1.18).“Therenais-sance...condemnedthescholastictermsasnotbeingCiceronian,withtheresultofmakingrenaissancephilosophyassoftandsavor-lessasasagepudding”(CP7.494n9,c1898).But,Peircesays,oneshouldnomoreconfusethedecadentscholas-ticswiththeworkoftheprecedinghighScholasticperiodthanlink“thehumanists”withthemodernphilosophyandsciencethatfol-lowedthem(CP8.11,1871).Infact,hispraiseofthehighScholasticscanbeextravagant.DunsScotusisagenius(CP2.166,1902),7“oneofthegreatestmetaphysiciansofalltime”(CP4.28,1893),andatleastthegreatestdefenderofrealism,whileOckhamisthegreatestnominalist(CP1.29,1869).8AndalongwithsomelaterBritishlo-gicians,ScotusandOckham“canbeusedtolayasolidfoundationonwhichtoerectanewlogicfitforthelifeoftwentiethcenturyscience”(CP7.161,1902).ScotusisevensingledoutasinaclasswithAristotleandLeibniz–andPeirce(!).9WhenPeircespeaksofScotusandOckhamasgreatlogicians(CP1.29),whatheadmiredwasnottheirlogicaltheorybuttheirrigorousapplicationofalogicalmethod.10But[their]logic,relativelytothegeneralconditionofthought,wasmar-vellouslyexactandcritical.Theycantellusnothingconcerningmethodsofreasoningsincetheirownreasoningwaspuerile;buttheiranalysesofthoughtandtheirdiscussionsofallthosequestionsoflogicthatalmosttrenchuponmetaphysicsareveryinstructiveaswellasverygooddisciplineinthatsubtlekindofthinkingthatisrequiredinlogic.(CP1.15,1903)11And:[A]boveallthings,itisthesearchingthoroughnessoftheschoolmenthataffiliatesthemwithmenofscienceandseparatesthem,worldwide,fromso-calledphilosophers.ThethoroughnessIalludetoconsistsinthis,thatinadoptinganytheory,theygoabouteverywhere,theydevotetheirwholeenergiesandlivesputtingittotestsbonafide–notsuchasshallmerelyaddanewspangletotheglitteroftheirproofsbutsuchasreallygotowardsatisfyingtheirrestlessinsatiableimpulsetoputtheiropinionstothetest.Havingatheory,theymustapplyittoeverysubjectandtoeverybranchofeverysubjecttoseewhetheritproducesaresultinaccordancewiththeCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n60johnboleronlycriteriatheywereabletoapply–thetruthoftheCatholicfaithandtheteachingofthePrinceofthePhilosophers.(CP1.33,1869)ThemixofcriticismandpraiseinPeirce’sattitudetowardsme-dievalthinkersisespeciallyintriguingwhenitcomestoauthority;forhethinksthewholeerawascharacterizedbyitsrespectforau-thority:“Themoststrikingcharacteristicofmedievalreasoning,ingeneral,istheperpetualresorttoauthority”(CP5.215n,1893).Thebackhandedcomplimentheoffersonthisscoreoffersnorelief:theweighttheyattachedtoauthority“wouldbeexcessivewerenotthehumanmindatthetimeinsouneducatedastatethatitcouldnotdobetterthanfollowmasters,sinceitwastotallyincompetenttosolvemetaphysicalproblemsforitself...”(CP1.31).12GivenwhatPeircehastosayaboutthemethodofauthorityin“HowToMakeOurIdeasClear”(CP5.379ff,1878),onemustwonderthathedidnotdismisstheperiodentirely.13SoitisasurprisewhenhegivesaspecialtwisttotheScholastics’dedicationtoauthority:ThegreatobjectofthemetaphysicsofDunsScotusissotostatetheresultsofordinaryexperience,thatitshallnotcloseanypositiveexperimentalin-quiry,orpronounceanythingpossiblyobservabletobeaprioriimpossible.InScotusthisnaturallyledtoloyaltytoAuthority,thentherecognizedfoun-tainoftruth;inourdayitwillproduceunfalteringfaithinObservation.(CP7.395,1893)Thisloyaltytoauthoritymeantthatthemedievalswerelessinter-estedinoriginalitythaninconsistencyofinterpretation;andtheywereremarkablyfreeof“thevanityofcleverness”(CP1.31,1859).Peircecontinues:Allthesecharactersreminduslessofthephilosophersofourdaythanofmenofscience.Idonothesitatetosaythatscientificmennowthinkmuchmoreofauthoritythandometaphysicians;forinscienceaquestionisnotregardedassettledoritssolutionascertainuntilallintelligentandinformeddoubthasceasedandallcompetentpersonshavecometoacatholicagreement....(CP1.32,1869)14Itisclear,Ithink,thatPeircehasamoraltopreachhereasmuchaboutmodernasaboutmedievalthought.WhilehispeersintheUnitedStatesandEuropesawthemselvesas(critically)advancingthecauseofmodernphilosophicalthought,Peirceincreasinglysawhimself,ifnotinopposition,atleastasproposingaradicaloverhaul.AndperhapsnothingwouldgettheattentionofhisreaderssomuchCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought61asflagrantlylininghimselfupwiththethoughtofa“backward”ageandoutlook.15Buthisremarkscarryasubstantivephilosophicalpointaswell.Consider:ThelogicalupshotofthedoctrineofScotusisthatrealproblemscannotbesolvedbymetaphysics,butmustbedecidedaccordingtotheevidence.Ashewasatheologian,thatevidencewas,forhim,thedictaofthechurch.Butthesamesysteminthehandsofascientificmanwillleadtohisinsistinguponsubmittingeverythingtothetestofobservation.(CP4.28,1893)Peircewaswritingatatimewhengrandsystem-buildingwaspromi-nent,anditwouldbepartofhispragmatistoutlooktoinsistthattherealworlddoesnotrevealitselftoarmchairtheorizing.16Whatherecognizedinthemedievals’respectforauthoritywasacheckonthepenchantofphilosopherstolettheirtheorizingdictatewhattheworldisreallylike.IIInanyevent,itisfromPeirce’sexplicitremarksthatwehavetheclearestindicationofhisreadingsandthepossibleinfluenceofme-dievalthoughtuponhisown.IshalltakeupsomeofthemoreobviousaspectsofthatinSectionIII.Buttherearetwoothersourcesofevi-denceforpossiblemedievalinfluences.TheonehastodowithwhatwecanidentifyfromreferenceshemakestomedievaltextsandfromhintsforareconstructionofPeirce’sownlibrary.Theotherhastodowithsimilarities(acknowledgedornot)topositionsofmedievalthinkers.Noneofthethreesortsofevidenceiswithoutitsproblems,asIshalltrytoexplainasIgoalong.Infact–thatis,asIseeit–wemaynotyetbeinapositiontoprovideadefinitiveaccountofmedievalinfluencesonPeirce.Itisextremelyunlikelythatthereisasmokinggunyettobediscoveredintheunpublishedmanuscripts,buttherearesomefamiliarenoughfactorsthatcomplicatetheenterprise.One,ofcourse,isthecon-stantdevelopmentwithinPeirce’sownthinking.AnotherhastodowithimprovementsinourownunderstandingofbothPeirceandthemedievals.Whatwe(should)havelearnedfromthegoodworkthathasbeendoneinrecentyearsisthatthebetterweunderstandthemthemorewerealizethereistoknow.Fortunatelyformypresentpurposes,wealreadyknowenoughtoallowsomeplausibleremarksaboutwherewearetoday.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n62johnbolerA.ALittleHistory17Actually,Peirceseemstohavehadaninterestinandaflairforthehistoryofphilosophy,18thoughitwasmotivated,Ithink,morebyphilosophicalthanhistoricalconcerns:Thechiefvalueofthestudyofhistoricalphilosophyisthatitdisciplinesthemindtoregardphilosophywithacoldandscientificeyeandnotwithpassionasthoughphilosopherswerecontestants.(CP1.28;cf.,CP8.9)Atarelativelyearlyage,hehadbeenforciblyimmersedinKantbyhisfather,BenjaminPeirce(CP3.405).AndduringhisHarvardCollegeeducation(andbeyond)heseemstohavereadthetextsinmodernphilosophythatwouldhavebeenstandard(Hookway1985:4–6,12).But,sometimearoundthemid–1860s,perhapsdissatisfiedwiththelogictextsofhistime,heundertookaconcentratedstudyofthehistoryoflogic.19Itwasthatprojectthatledhimtoreadextensivelyin(ancientand)medievalsourcesandintroducedhimtothebroaderrangeofphilosophicalanalysisinthelaterMiddleAges.PeircegraduatedfromHarvardCollegein1859andhadalreadywrittensomemetaphysical(butunpublished)essaysaround1860(Hookway1985:4).In1863,hereceived(ineffect)agraduatedegreeinchemistry.Asapromisingscholar(stillonly25yearsold),hewasinvitedtogiveaseriesoflecturesonthelogicofscience.20Hookwaythinkstheymayhavecontainedmaterialthatwouldappearintheanti-Cartesian/intuitionarticlesof1868–1869(Hookway1985:6).JustwherePeircefoundthetextsnecessaryforhisearlystudyofthemedievalsisnotaltogetherclear(tome).21Inanyevent,bythetimeofhisteachingcareeratJohnsHopkins(1879–1884),Peircehadamassedanextensivelibraryofmedievaltexts;andin1880,heof-feredtosellittotheuniversitylibrarythere.22Ofcourse,Peircehadbeentravelingwidelybeforethis,andmostofthebookswerepurchasedbetween1866and1871invariouscitiesinEuropewhichwouldhaveprovidedaplentifulmarketforobtainingmedieval(andother)texts.23B.Peirce’sLibraryWeknowfromtheenthusiasticsupportforthepurchasebythelibrar-ianatJohnsHopkinsthatthecollectionwasanoteworthyone,appar-entlyovertwohundreditems,thoughnotallonmedievalsources.24CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought63Fischgivesasketchydescriptionoftheworksonthelibrary’sac-cessionslistand,moreimportantly,oftherathersadhistoryofthecollection(Fisch,1986:52–3).Fortunately,thebulkofthecollectionisstillintheJohnsHopkinslibrary,includingthirty-fourincanabula.Butinsomewhattypicalfashion,Peirceborrowedbackanumberofthebooks.Hemeanttorepurchasetheentirecollection,butneverhadthemoneytodoso.SomeofthebookshekeptwenttoHarvardwhentheypurchasedPeirce’spapers.25ButothersmayhavebeenburnedwhenPeirce’swidowdiedin1934(Fisch,1986:54).C.Citations(ofMedievalAuthors)WhenweturntoPeirce’sownwritings,wefindagoodnumberofexactquotationsandevenmorespecificreferencestoancientandmedievalwriters.JustworkingfromtheindicesoftheCollectedPapers,andleavingasidehisreferencestoAristotleandtoStoicandEpicureanlogic,thelistwouldincludeAugustine,Boethius,Cassiodorus,ScotusEriugena,Anselm,Abelard,JohnofSalisbury,AlexanderofHales,PeterofSpain,WilliamofAuvergne,RogerBacon,AlberttheGreat,AvicennaandAverroes,ThomasAquinas,HenryofGhent,DunsScotus,WilliamOckham,andPaulofVenice.26MostofthesecitationsappearedasentriesinBaldwin’sDictionaryofPhilosophyandPsychology,whichwaspublishedin1901–1902.Peircehadbeforethen“retired”toArisbeinMilford,Pennsylvania;andwhilehemayhaveretainedaccuratenotesfromearlierreadingsandconceivablyeventraveledtoalibrarysuchasJohnsHopkins,themorelikelyimplicationisthathehad(manyof)thetextswithhim.Anumberofgeneralallusionstomedievalwritersappearevenearlierasintroductions,bywayofthehistoryofthefield,indraftsofhiseffortstoconstructacomprehensivetextonLogic.27Therearealsosomecaseswhere,evenwithoutexplicitcitation,Peirceisclearlyworkingwithamedievaltext:forexample,inthere-viewofFrazer’sBerkeley,hehasinfrontofhimScotus’streatmentofuniversalsfromBookVII,q.18,oftheQuestionsontheMetaphysics(DunsScotus1997:II,337–56).28D.SimilaritiesAlanPerriahhassuggestedafurthersourceofevidenceforpossi-blemedievalinfluencesonPeirce’sthought:similaritiesindoctrineCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n64johnbolerevenifnotexplicitlyidentifiedbyPeirce(Perriah1989:41–9).Hesuggests(butdoesnotmeanasexhaustive):theideaofmodesofbeingandmodesofpropositions,thenormativecharacteroflogic,thepriorityofdialecticalreasoning,logicandprobability,seman-ticsofsigns(e.g.,significationandsupposition),andofcoursethenominalist–realistdebate.ItisaninterestingideabutIthinkitde-servesawordofcaution.OnewouldlikeideallytodistinguishcasesofinfluenceonPeirce’sthoughtfromthosewherehesimplyfoundinthemedievalspositionsthatheliked.29Ittakesatrainedeyetodistinguishappearanceandrealityhere(aselsewhere).Ineffect,onehastobefamiliarwithbothPeirceandtherelevantmedievals,andasIhavesuggestedearlier,thatwouldinvolvekeepingupwiththebestofrecentcommentary.Peirce’sallusionsto“speculativegrammar”seemtomeacaseinpoint.30The1639WaddingeditionofScotus’sworkscontainedanumberofworkswenowknowwerenotwrittenbyScotus,31oneofwhichisGrammaticaSpeculativabyThomasErfurt.“Speculativa”simplymeans“theoretical,”andmedievalspeculativegrammarwasanattempttoprovideaformalgrammar,inpartfromastudyofnaturallanguagesandinpartfromlogicalstructure.ItisnothardtoseewhyPeircemightnotmakegreatuseofit,32forits“formal”characterreliesheavilyonasubject–predicateanalysisofLatinandGreekmodels.33Butitshouldalreadybesuspicious,Ithink,thatwhilePeircereferstothemedievalworkfrequently,heneverquotesfromitorevengivesexactcitations.34ThekeytoPeirce’sinterestinthetopicliesinhisidentifyingspec-ulativegrammarwithwhathecalls“Erkenntnislehre”orsometimes“Elementarlehre,”35anideahemorelikelygotfromKantthanfromScotus/Erfurt.Wecanseewhathehadinmindfromanimportantearlyarticle“OnaNewListoftheCategories”(CP1.545–59,1867).36ForKant,veryroughlyput,categoriescanbederivedfromthelogi-calformsofjudgmentsandrepresentthewaysthemindstructuresexperience.Peircewassympathetictoboththeseideas,37objectingonly(!)toKant’shavinginsufficientlygeneralizedthebasicnatureoflogicalform;andthe“NewList”washisfirstefforttocorrectthat(CP1.560–4).The“NewList”isaveryoriginalworkexpressedintraditionalter-minology.ItsthreecategoriesofQuality(referencetoaground),Re-lation(referencetoacorrelate),andRepresentation(referencetoanCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought65interpretant)mediatebetween“Substance”and“Being”(CP1.555).Intheprocess,however,Peirceisabletosetouthisbasictriadofsigns,likeness(latericon),index,andsymbol(CP1.558),andtode-lineatehisthreetypesofinference,deduction,hypothesis(laterab-duction),andinduction(CP1.559).Heevenmanagestohintathislatermostabstractcharacterizationofcategoriesintermsofvalen-cies:i.e.,monadic,dyadic,andtriadic.Inthefinalanalysis,however,the“NewList”provesinadequatetoPeirce’spurposes,inlargepartbecausethewayhederivesthecategoriesthereessentiallydependsuponasubject–predicate(or“SisP”)forminitsanalysisofpropositions.38Hookway(1985:80–117)hasanextensivediscussionofthe“NewList”andmostimportantlyofthetransformationinPeirce’sapproachtothederivationofcate-gories,bothinthemoreabstractdeviceofvalenciesandinthemoreconcrete“phenomenological”approach.IhavealittlesomethingtosayaboutthelattertowardtheendofsectionIII;but,fortunatelyforme,theverycomplexandcontroversialtopicofPeirce’smaturecat-egorytheoryimpingesonlymarginallyonthequestionofmedievalinfluences.Toreturntotheimmediatetopicathand,Iamnotsuggestingthatspeculativegrammar,akaErkenntnislehre,isnotimportantforPeirce,orthatastudyofthemedievaltreatiseswouldnotbevaluableinitsownright.ButIdonotthinkonewouldlearnalotabouttheformerbyexaminingthelatter.ThatPeircesawinitaforerunnertohisowninterestsseemstomeamorelikelyhypothesisthanitshavingbeenaninfluenceonhisthought.IIIPeirce’sScholasticRealismThebestknownandmostcommenteduponcaseofmedievalinflu-enceonPeirceishisself-ascribed“scholastic/Scotisticrealism”:39IshouldcallmyselfanAristotelianofthescholasticwing,approachingSco-tism,butgoingmuchfurtherinthedirectionofscholasticrealism.(CP5.77n1,1903)NoonedoubtsPeirce’sclaimthathewassignificantlyinfluencedbyreadingScotusandotherScholastics,butthereissomecontroversyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n66johnboleraboutwhen,howmuch,andwhatPeirceultimatelymadeofit.Theanswertothosequestions,however,iscomplicatedby,amongotherthings,Peirce’sowndevelopment,bothinhislogicaltheory(e.g.,wherehemovedawayfromthesubject–predicateformasbasic)andinhismature,three-categoriedontology.40Still,onecanmakeoutthegenerallinesofananswer.Asforthedetails,however,IshallnottrytogiveafullaccountofeitherPeirce’sorScotus’srealismbutwilldeveloponlyenoughoftheirpositionstodealwithquestionsofsimilarityandinfluence.41ForreasonsthatshouldbecomeclearinSectionIV,IshallconcentrateonPeirce’searlierformulations.Thereareinterestingandsignificantdifferencesinthewayme-dievalshandledvarioussemanticissues,42butfromearlyontherewasnoconfusionaboutwhatmakesatermuniversal:i.e.,notthatitsignifiesageneralentitybutthatitsignifiesmanyindividuals.Afterall,weretheterm“donkey”tostandforsomegeneralentity,itwouldfunctionratherasapropername.Wecancallthisone–manycharacterofthesignificationofgeneralterms“universality”;itisclearlyapropertyofsignsandnotofthethingstheysignify.Butthedisputethatcametobecalledthe(sic)problemofuniversalshadtodonotwithuniversalitybutwiththeobjectivestatusofnaturalkinds.43Andinthemedievalcontroversyatleast,twoconditionsprovidethecontextofthatdiscussion;forwhiletheScholasticsheldthatscienceisoftheuniversalandnecessarytheytookindividualsubstancestobeontologicallyprior.44Theyreadilytalkedofessencesand/ornaturesaswellasofabstractentitiessuchasjustice,buttheirtendencywastorejectanyrealmof“separate”entities,anattitudereinforcedbyAristotle’scriticismofPlato.EvenwheretheyfoundaroleforPlato’sIdeasasexamplarsinthemindofGod(i.e.,patternsaccordingtowhichGodcreatedthings),theseweredistinguishedfromformsinherentincreatures.AlongwithotherScholasticrealists,Scotusproposeswhatwemightcalla“metaphysical”compositionwithinthings(orfirstsubstances),45inhiscasethatofcommonnature(s)andhaecceity(or“thisness”).46Hemarshallsasetofargumentstoshowthatthenaturesofthings,ascommon,havearealbut“lessthannumericalunity”whichisrequiredtogroundtheobjectivityofourscientificknowledge.47Andheisrenownedformaintaining(againstAquinasandothers)haecceityorthisnessasapositiveprincipleofindividua-tion.Moreover,whileheemphasizestherealityofcommonnatures,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought67heholdsthatthenatureis“contractedtothemodeoftheindividual”inactuallyexistingthings(CP8.208,1910).48Thisisimportantforanumberofreasons,asweshallsee,butforthemomentitisenoughtonotethat,forScotus,theanalysisintocomponentsisnotreduc-tive:theyarenotmerelyfoundinsubstancebutareimportantlydependentuponsubstanceasprimarybeing.WilliamOckham,whoisPeirce’sfavoritenominalist(CP1.29),isequallycommittedtothepriorityoffirstsubstance,andheisperfectlyclearthatScotus(likeotherrealistssuchasAquinas)wastryingtoprotecttheobjectivityofknowledgewithoutpostulatinganextramentalgeneralthing.Buthethinksthattheappealtospecialcomponentsofthings,forms(Aquinas)orcommonnatures(Scotus),cannotavoidblurringacrucialboundarybetweenwhatisrealoroutsidethemindandwhatisconceptualorinthemind,thusmakingrealthingssomehowrelativetoordependentuponthemind.49Wemightassume,then,thatweareonfamiliargroundwheninhisearliestextendedaccountofthenominalist–realistdispute,whichoccursinhisreviewofFrazer’sBerkeley(CP8.7–38,1871),Peircesays:Thequestion,thereforeiswhetherman,horse,andothernamesofnatu-ralclasses,correspondwithanythingwhichallmen,orallhorses,reallyhaveincommon,independentofourthought,orwhethertheseclassesareconstitutedsimplybyalikenessinthewayourmindsareaffectedbyin-dividualobjectswhichhaveinthemselvesnoresemblanceorrelationshipwhatsoever.(CP8.12)AndinanearlierbriefreferencetotheScholasticoutlook:Objectsaredividedintofigments,dreams,etc.,ontheonehand,andrealitiesontheother.TheformerarethosewhichexistonlyinasmuchasyouorIorsomemanimaginesthem;thelatterarethosewhichhaveanexistenceindependentofyourmindormineorthatofanynumberofpersons.Therealisthatwhichisnotwhateverwehappentothinkofit,butisunaffectedbywhatwemaythinkofit.(CP5.311,1868)50ButthereaderisinforasurprisewhenPeircegoesontoexplainwhatthedisputewasreallyabout.Hebeginsitthisway:Thecurrentexplanationsoftherealist–nominalistcontroversyareequallyfalseandunintelligible51...Yetitisperfectlypossiblesotostatethematterthatnooneshallfailtocomprehendwhatthequestionwas,andhowthereCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n68johnbolermightbetwoopinionsaboutit.Areuniversalsreal?Wehaveonlytostopandconsideramomentwhatwasmeantbythewordreal,whenthewholeissuesoonbecomesapparent.(CP8.12)Asithappens,Peirce’ssenseoftheobviousisitselfsomethingtowonderat.Forhedevotesthenextfourorfivepagestoadiscussionofthe“twoviewsofreality”inawaywhichatfirstglanceseemstohavemoretodowithidealismthanrealism.Butwhileonecannotdismisstheidealistidiom,Ithinkitmaybesomethingofadistrac-tion.PeirceisnotembarrassedaboutfavoringidealismsohehasnoreasontohidethatbehindtheScholastics.Itwillbethemainburdenofmyaccountheretoshowhowthisearly(1871)accountexhibitsthecentralelementsofwhat,forPeirce,isScholasticrealism.Inordertoseethat,oneshouldbeginwithPeirce’sdescriptionthereofnominalism.Onthenominalistview,hesays,realitylieswholly“outsidethemind”(orextraanimam);itcausesoursensa-tionsandthroughthemourconceptions(whichareinanima).Thesearenotgeographicallocations,ofcourse;whatis“inthemind”isnotamentalact,butsomethingmorelikeitscontent:being“in”themindisratherlikebeing“in”apictureor“in”astory.Thisinsistenceonasharp“internal–external”distinctionis,forPeirce,anominalistfixation.52ItisalsosomethingPeircecouldwellhavelearnedaboutfromOckhamforwhom“theworsterrorinphiloso-phy”istoconfusethepropertiesofourrepresentativesystemwiththepropertiesofrealthings.53Thepoint,forthenominalist,isthattherealhastobewhollyindependentoftheconceptual.54AsPeircewillsaylater:Theheartofthedisputeliesinthis.The[nominalists]...recognizebutonemodeofbeing,thebeingofanindividualthingorfact,thebeingwhichconsistsintheobject’scrowdingoutaplaceforitselfintheuniverse,sotospeak,andreactingbybruteforceoffact,againstallotherthings.Icallthatexistence.(CP1.21,1903)55Peircecomestoidentifynominalismwiththeclaimthat“reality”and“existence”aresynonymous(CP5.503,1905).Whilethisde-scriptionappearsonlyafterPeircehasdevelopedanexplicittheoryof(three)modesofbeing,56thedoctrineitselfcanbeseeninthe“nom-inalisticPlatonism”hehadidentifiedalreadyintheBerkeleyre-view(CP8.10):philosopherswhowantedtopreservetheobjectivityCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought69ofscientificlawsbutrecognizedonlytheonemodeofbeinghadtosupposethatrealgeneralswerethemselvesindividualthings.57TheScholastics’rejectionofthatmove,infavorofspecialconstituents,isnotwhatmadethem“moderate”inPeirce’seyes,butwhatmadethemrealists.Itiseasytomissthisimportantpoint.58Inanyevent,PeircethinksOckham’s(oranynominalist’s)efforttomaintainanontologicalpurityultimatelycreatesanunbridgeablegapbetweenthewaythingsreallyareandthewayweconceiveofthem:notjustbetweenwhatisindividualandwhatisgeneral,butbetweentherealconditionofexternalthingsandthepropertiesweconceivethemtohave.59Fromhisnineteenthcenturyperspective,Peircedescribesthisasmakingtherealanunknowablething-in-itself(e.g.,CP8.13).Theidealistidiomcontinuesinhisdescriptionofthecontrasting,realistconception,wheretherealislocatednotinwhatstartsthethoughtprocessbutinitsresult,whichistheopinionthatanyinquirer(giventhepropercircumstances)wouldarriveat:Thisfinalopinion,then,isindependent,notindeedofthoughtingeneral,butofallthatisarbitraryandindividualinthought.(CP8.12)60Andthen:Itisplainthatthisviewofrealityisinevitablyrealistic;becausegeneralconceptionsenterintoalljudgments,andthereforeintotrueopinions...Itisarealwhichonlyexistsbyvirtueofanactofthoughtknowingit;butthatthoughtisnotanarbitraryoraccidentalone...butonewhichwillholdinthefinalopinion.(CP8.14)Andfinally:[W]hatKantcalledhisCopernicanstepwaspreciselythepassagefromthenominalistictotherealisticviewofreality.Itwastheessenceofhisphilos-ophytoregardtherealobjectasdeterminedbythemind.(CP8.15)Onemightwellwonderhowanice,youngAristotelianScholasticlikeScotuscouldgetmixedupwithallofthis!BeforelookingatScotusmoreclosely,however,acommentonPeirce’saccounthereisinorder.Foritseemsclearenough(ifonlyfromwhathesayslater)thatPeirceisnotproposingasimplecommitmenttoidealism,asiftryingtomakeScotusaprecursortothepost-Kantians.Afterall,onecouldeasilymakeouta“nominalisticidealism”onthemodelof“nominalisticPlatonism.”WhatmakesforScholasticrealismisCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n70johnbolertheappealto“constituents,”61nottorealthingsthatarethoughtlikebutrathertoanaspectofthingsthatmustberealifourknowledgeisobjective.WhatPeirceseesintheScholastics,Ithink,isanap-pealtoastructureinthingsthatisanalogoustothestructureofthought.62Finally,Peircewillnotmaintainthatacommitmenttoanexistenceindependentofthoughtmustbeeliminatedbutthatitbecomplementedwithanothermodeofreality(CP7.339,1873).ToreturntoScotus,then,whatisuniquetohisposition(amongscholastics)isthathebringsthetwocomponents,commonnaturesandhaecceity,underthe“formaldistinctionaparterei,”anotoriousdevicethatatleastcreatesarealdistinctionamongcommentatorsandalonecouldhaveearnedhimthetitleof“MostSubtleDoctor.”Histheoryof“formalities”wasthesublest,exceptperhapsHegel’slogic,everbroached,andhewasseparatedfromnominalismonlybythedivisionofahair.(CP8.11)63ItisunderstandablethatScotuswouldwantadistinctionherethatissomethingofanentredeux:i.e.,notmerelyalogicaldistinction(asbetweentheauthorofRobRoyandtheauthorofIvanhoe),butalsonotastrongrealdistinctionasbetweentwothings(resetres).ButtherearefeaturespeculiartotheformaldistinctionthatwouldbeattractivetoPeirceeveniftheycomplicatethedescriptionofScotus.64Theformaldistinctionrequiresaspecialsortofterm:notforms,Scotussays,butformalitates,notresetresbutrealitates.65Whatwehavethenisnotjustasomewhatstrangedistinctionbetweenotherwisefamiliarsortsofthingsbutrather(ormore)aspecialdis-tinctionbetweensomewhatstrangesortsofthing.Incashingthisin,moreover,onemustbalancetheideathatthedistinctionisobjective(thatis,aparterei)withanunavoidablereferencetoconceivability.Hereisarecentinterpreter’sefforttocharacterizeit:Aformaldistinctionisadistinctionfromthenatureofthethingoccurringbetweentwoormorereallyidenticalformalities,ofwhichone,beforetheoperationoftheintellect,isconceivablewithouttheothersthoughinsepa-rablefromthemevenbydivinepower.(Grajewski1944:93)Thatis,thedistinctionholdspriortotheactofanyintellect;butitisevensoamatterofwhatwouldbeconceivedbyatrulyknowingmind.66ItisnotthatformalitiesaretobeconfusedwiththoughtsCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought71aboutthem;buttheirbeingrelativetothoughtgoesbeyondapsy-chologicalorepistemologicalclaimthatthisistheonlywaywecandescribethem.67IndescribingtheformaldistinctioninScotus,IhavefocusedonthestatusofformalitiesorrealitiesaswhatwouldbeconceivedbecauseIthinkthatisthebackgroundforPeirce’searly(andcontin-uing)approachtothenominalist–realistdisputeintermsofviewsof“reality.”Thatistosay,despitetheidealistidiom,theemphasisintheBerkeleyreviewonarealistconceptionofrealityseemstomeidentifiably“Scotistic.”Moreover,thegroundforPeirce’slateridentificationof(three)modesofbeing68canbefound,eveninthisoriginalpresentation,inthegenuinely“Scholastic”appealtocon-stituents,i.e.,toastructureinthingsanalogoustothestructureofpredication.ThesebroadScotistic/ScholasticelementsareneverabandonedbyPeirce,thoughsomethingneedstobesaidaboutthetransformationtheyundergoinhisdevelopingmetaphysics.69Fromsomeofhislaterformulationsofthecategories,onemightbeledtothinkthatPeircemeanstobetalking,asScotusis,aboutaspectsorfeaturesofthesortsofindividualsthatthemedievals(andmostoftherestofus)takeasmakingupthefamiliarworldinwhichwelive.Onthisapproach,Firstness,Secondness,andThird-nesscouldbeseenas,say,thequality,bruteexistence,andlawlikebe-haviorexhibitedbyfirstsubstances.Properlyunderstood,however,familiarphysicalobjectsforPeircearelawlikeprocesses,systems,constitutedbyFirstness,Secondness,andThirdnessratherthanbe-ingsupportiveofthem.70Butthereareotherchangesaswell.Haecceity,tobeginthere,71maylookasifitgoesdirectlyintoSecondnessforPeirce.Butitsprimaryontologicalroleforhimisnotatalltoreflectthepriorityoffamiliarindividualthingsasfirstsub-stances.Secondness,whichishaecceitytransformed,72issomethingofasurd:itisthebrutefacticityofourencounterwithanexternalworld.Itisnotthecontentorintelligibilityofthatencounterbutthepureresistanceof,say,tuggingatastuckdoor(CP1.324)ortheshockofhavingone’sreverieinterruptedbybeingknockedtotheground(CP1.431);henceitsdyadiccharacter.ItisimportantforPeirce,73butnotasafeatureofphysicalobjects,asifthelatterwereontologicallyprior.FirstnessisvariouslydescribedbyPeirce(Hookway1985:106–7).Itis“themodeofbeingwhichconsistsinasubject’sbeingpositivelyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n72johnbolersuchasitisregardlessofallelse[whichis]apositivequalitativepos-sibility”(CP1.25).Inthiscase,themoreabstractcharacterizationsmayhelp:itismonadic(andsoisindependentofanyinstantiation);anditispurepossibilityandsodifferentfrompotentialitywhichbelongstoThirdness.74ThirdnessisIthinkbasic,asitwerethefirstamongthreeequalsinPeirce’smaturemetaphysics.AndthoughhewilllocatetherealgeneralitythatconcernedmedievalrealistsunderThirdness,itisnotsomuchthecommonnessofnaturesthatPeircethinksmustberecognizedintheobjectivediscoveriesofscienceasitistherealityof“would-be’s”thatarenotexhaustedby(orreducibleto)anyoneormoreactualeventsorSeconds(CP1.422).ItistothisthatPeirceultimatelyascribeshisown“extremerealism”:ImyselfwenttoofarinthedirectionofnominalismwhenIsaidthatitwasamereconvenienceofspeechwhetherwesaythatadiamondishardwhenitisnotpressedupon,orwhetherwesayitissoftuntilpressedupon.[Cf.CP5.403,1878]...Itisarealfactthatitwouldresistpressure,whichamountstoextremescholasticrealism.(CP8.208,1905)Asithappens,thereisaScotisticbackgroundforthatdevelop-mentaswell,thoughitisnotsoclearlyacknowledgedassuchbyPeirce.ItistobefoundlessinScotus’saccountofrealgeneralsthaninhisaccountofpotentiality.75Afterall,whilethenature“loses”itscommonnessintheindividualsubstanceforScotus,potentialitycantranscenditsexercise.Thepotentialhouseis“replaced”bytheactualhousethatisbuilt.Butwhileheisanactualandnolongerapotentialbuilderwhenheisonthejob,thebuilder’scapacityforbuildinghousesisnotdisplacedbythatactivity.76LetmeconcludethissectionwithwhatisperhapsthemostbasicdifferencefromScotus,reflectedinPeirce’scomplaintabout“con-traction”:EvenDunsScotusistoonominalisticwhenhesaysthatuniversalsarecon-tractedtothemodeofindividualityinsingulars,meaningashedoes,bysingulars,ordinaryexistingthings.Thepragmaticistcannotadmitthat.(CP8.208,1905)77Thereare,Ithink,threethingsgoingonhere.First,contractionforScotuspreservestheontologicalpriorityoffirstsubstance;andPeirceCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought73meanstodenythat.WhileFirstness,Secondness,andThirdnessareneverfoundapartfromoneanother,theyeachhaveanindependentstatussothattheyarepriortoandnotbroughttogetherinorsup-portedbyasubjectthatismetaphysicallyprior.ThesecondroleofcontractionforScotusisthatitgroundstheactivityofsubstancesintheirindividualnatures.ForPeirce,aswehaveseen,therealityofThirdnessislocatedinobjective“would-be’s”whicharenotex-haustedinanyactoractualconditionofthings.Thescholastics,ofcourse,reliedheavilyonpotenciesintheirexplanationoftheactivi-tiesofthings,buttheystillsawthegroundforthatactivityincertainactualconditions(e.g.,formsornatures)ofindividualsubstances.78ThethirdproblemwithScotus’scontractionisnotsomethingthatPeircecallsexplicitattentionto,butitisbasicforhim;anditreturnsustoPeirce’searlydescriptionofnominalismintheBerkeleyreview:contractionreintroducesagapbetweenthewaythingsareandthewayweconceivethemtobe.AnditisthatratherthanarestrictionongeneralitythatIthinkPeircefinds“nominalistic.”ThegapisperhapsnarrowerinScotusthaninAquinas,butitissomethingPeircewouldbeespeciallysensitiveto.79Conclusions“[G]eneralprinciplesarereallyoperativeinnature.Thatisthedoc-trineofscholasticrealism”(CP5.101,1903).TheformthistakesinPeirceisacommitmentnottoanewsortofthing(individual)buttoanewsortofconstituent,theanalogueofthepredicatefunctionintheexpressionofourthought.ThestatusofthatsortofconstituentinhisScotisticrealismisthatofaformalityorreality,whichiswhatatrulyknowingmindwouldconceive.IfIamright,thisgen-eralstructureofScholasticrealism,asPeirceseesitincontrasttoanominalisticPlatonism,isalreadyinplaceintheBerkeleyreview.WhethertheimportantconceptofpotentiashouldbeseenaspartofScholasticrealismorsimplyasanotherScotistic/Scholasticinflu-enceisnotcrucial.WhatneedsemphasisisthatthecontributionofScholasticre-alismisonlyoneelementinthedevelopmentofPeirce’sulti-mateposition.80Forexample,thatreality,properlyunderstood,issomehowthought-relativeisnotsomethingnewforPeirceeveninCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n74johnboler1871.Itisearlyonassociatedwiththeideaofacommunityofin-quiry;anditisnotunfairtosaythatPeirceisstillstrugglingattheendofhislifewithaschemeforgroundingtheobjectivityofthesethought-relativedata.81Then,theScholasticconstituents,ofcourse,areradicallytransformedasmodesofbeingthatconstituteratherthandependuponindividualsubstances.Eventheidentifica-tionofgeneralsin“would-be’s”goesbeyondtheScholasticideaofpotentiality.Giventheextentofthesedevelopments,isPeircerighttopersistincallinghimselfaScholasticrealist?Commentatorswhosesimpleanswerstothatquestionaredifferentmaynotalwaysdifferaboutthefactsofthecase.82Onceoneisclearaboutthedetailsofhis“Scholasticrealism”andhis“extremerealism,”however,Idonotseeanyprofitinworryingoverthelabels.IVInmakingoutacaseforPeirce’s“Scholasticrealism,”Ihavecon-centratedonitsearlyappearancewhilenotingsomeoftheimportanttransformationsthataccrueinthecourseofhisdevelopinglogicandmetaphysics.Butsomethingneedstobesaidabouttheplaceofreal-ismitselfinPeirce’sdevelopment,briefandsketchythoughitwillhavetobe.Itcanperhapsserveasakindofreview.In1891,Peircesaysthat“never,duringthethirtyyearsinwhichIhavebeenwritingonphilosophicalquestions,haveIfailedinmyalle-giancetorealisticopinionsandtocertainScotisticideas”(CP6.605).Butinaveryimportantarticle,MaxFischclaimsthatPeircehaslo-catedthatallegianceaboutfiveyearstooearly(Fisch1986:197–8).HeplacesPeirce’sintensivestudyofthemedievalsin1868–1869,83thoughthatwouldnotprecludePeirce’sholdingarealistpositionwithoutyetknowingthatitwas“Scholastic.”Moreimportantly,Fischclaimsthat,untilhewas29yearsold,Peircewaswillingtocallhimselfandbecalledanominalist.84FischthenarguesthatPeircefirstrunsuptherealistflagtowardtheendof“SomeConsequencesofFourIncapacities”(CP5.312,1868),oneofhisanti-Cartesian(oranti-intuition)articles.Butthis“firststeptowardsrealism,”Fischclaims,wasactuallyonlya“rideronhisearlyidealism”(Fisch1986:193).EventhemorefullydevelopedaccountintheBerkeleyreviewseemstoFischonly“asecondsteptowardsrealism”(Fisch1986:CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought75188,italicsmine),whereFischisevidentlythinkingofthelaterreal-ismthatisfullydevelopedinPeirceonlyaftertheturnofthecentury(Fisch1986:187–88).85Fischthenpointsoutthat“fornearlytwodecades,”from1872to1890,Peirceisalmostsilentonthenominalism–realismissuewhileconcentratingonmajordevelopmentsinthelogicofrelations,arevisionofhistheoryofcategories,hispragmaticistcommitmenttowould-be’s,anddoingworkontransfinitenumbersthatformedapreparationforhistheoryofrealcontinuity(Fisch1986:188).FischseesthesechangesasamatterofPeirce’sprogressivelyaban-doningrecalcitrant“nominalistic”positions(e.g.,onthematerialconditionalinhisearlypragmaticmaxim,andthedefinitionofthepossible).86AndhesuggeststhatitmayhavebeenwhilepreparingforhiscontributionstotheCenturyDictionarythatPeirceundertookareviewofthehistoryofphilosophythatre-energizedhisinterestinthemedievaldebate(Fisch1986:192).Inanyevent,after1890,especiallyindraftsfortheGrandLogic,referencestorealismbecomeprominentagain,forexampleinthefirstfullyexplicitaccountofthreemodesofbeing,wherehaecceitygetsanewemphasis87andpragmatismisassociatedwithrealism(Fisch1986:195).Finally,FischclaimsthatalateconversiontoadoctrineofImmediatePerception(connectedinpartwiththeem-phasisonScotus’shaecceitiesasSecondness)constitutesarejectionofidealisminfavorofwhatisfinallyPeirce’sultimatepositionon“realism”(Fisch1986:192–6).Thesechanges,headds,openedupwholenewlinesofdevelopmentwhichPeircestruggledwithbutneverbroughttocompletionbeforehisdeathin1914(Fisch1986:196–7).Asisevidentfromthelastsection,IfindamoresubstantialScholasticrealisminPeirce’s1871formulationsthanFisch(interalia)does.ButIhavenoquarrelwithFisch’saccountofthemanydevelopmentsthattookplacebetweenthatandPeirce’slaterreturntothetopicofrealism.Afterall,itonlyreinforcesmyownemphasisonthelimited(ifimportant)influenceofhisScholasticrealismonthelaterrealism.ButIhavemydoubtsaboutthe“nominalismtorealism”storyasthebestformatforunderstandingPeirce’soveralldevelopment.Ofcourse,theremaybenoonestorythatwoulddothejob.88ButIwillclosemyaccountwithtwocommentsonthepictureFischoffers.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n76johnbolerTobeginatthebeginning,ayouthfulespousalofnominalismdoesnotseemtometoprovideaverysolidbasefromwhichtoviewPeirce’ssubsequentdevelopment.Atthattime,beinganominalistwasthepoliticallycorrectstanceforsomeonewhowantedtogivetheimpressionofhavingahard-headedandscientificoutlook.Itisunlikely,moreover,thatevenifPeirceheldanumberof“nominal-istic”positions,hewascommittedatthattimetoanythinglikeadevelopedtheoryofnominalism.89TheendofFisch’sstorystrikesmeasevenmoreproblematic.FischclaimsthatPeirceultimatelyrejectshisidealismbecauseoftherealismofhislatedoctrineofImmediatePerception.90WhatPeircesaysis:EveryphilosopherwhodeniesthedoctrineofImmediatePerception–includingidealistsofeverystripe–bythatdenialcutsoffallpossibilityofevercognizingarelation.(CP5.56,1903)Butthisis,Ithink,misleading.IfthedoctrineofImmediatePer-ceptionissupposedtobeaformofDirectRealism,itisopposedtorepresentationalism,whichisnotapeculiarlyidealistposition.Peircemayhaveatonetimemaintainedaformofrepresentation-alism(CP8.12),91butheseemstomeallalongtofavorthebroadersortof“objectiveidealism”whicharguesnotthattheimmediateobjectofknowledgeisourownideasbut(roughly)thatifknowledgeispossible,therealastheobjectofknowledgemustbeidea-like:cf.(CP5.553,1906).92Infact,Peirce’sdoctrineofImmediatePerceptionisnotamatterofperceivingordinaryphysicalobjectsassuch,asanallusiontoDi-rectRealismmightsuggest.IthastodowiththebruteencounterofSecondness;andwestillneedtointroducetheaspectsofFirstnessandThirdnesstoprovideanyintellectualcontenttotheexperience.Butanyexperiencethatincludesthelatterwillbewoefullyunderde-terminedbythebruteencounterwithSecondness,leavingusagainwithanappealtotheobjectofthefinalopiniontoestablishtheobjectivityofanyempiricalclaimbaseduponit.93Andthat,tomymind,couldevenleadusintotheissueofrealismandantirealism.94Insum,totrytobringrealism-versus-nominalismandrealism-versus-representationalism(andrealism-versus-antirealism)underasingleheadingismorelikelytobeconfusingthanilluminating.95CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought77Formypresentpurposes,itmaybeenoughtonotethatittakesuswellbeyondthetopicofmedievalinfluences.notes1.Theratherblandtitlereflectsmycautionaboutclaimsofinfluence(onwhichmorelater).AllmydirectreferencestoPeircearefromtheCollectedPapers.Theyare,however,chronologicallychallenged,soIhaveoftenincludedthedatesofpassagesIcite.2.HewasinfactaperceptivereaderofthehistoryofideasintheMid-dleAgesandhisextendedaccounts,toolongtoquotehere,areworthlookingat:e.g.,CP4.1,1829,4.21–37,1893,8.9–11,1871;seealsoCP2.166–8,1902and6.312,1891.ThoughheshouldhaveknownbetterthantolaythecanardaboutangelsdancingontheheadofapinontoAquinas(CP8.11).HealsoanticipatesPanofsky’sanalogyofgothicar-chitectureandScholasticism(1957)with(tomymind)equallylittleeffect(CP4.27and8.11).3.Theentireeradoesnotgetoffunscathed,buthisharshestremarksarerarelydirectedatthefiguresofhighScholasticismandseemtobere-servedmainlyforthebeginningandendoftheperiod:i.e.,beforethetwelfthcenturyresurgenceoflearningandthenthedecadentScholas-ticismofthesixteenthcentury.SeeCP8.11and1.27n1.4.“Theschoolmen,whoregardedAristotleasallbutinfallible,yettowhomtheideasofanaturalistwereutterlyforeign...”(CP6.357);andseeCP6.361,1902(thisisaparagraphworthreadinginitsentirety);thoughforcontrast:CP1.32–3,1869,CP7.161,1902,CP7.395,1893.ForRogerBacon,CP1.29,1869;CP5.360,1877.5.SeealsoCP2.166–8,1902,CP3.509,1896,CP6.348,1909.6.“[Nominalists’protests]againstmuchoftheemptydisputationsofthemedievalDunces[amountedto]aprotestagainsttheonlykindofthink-ingthathaseveradvancedhumanculture.”(CP3.509,1896)7.Cf.CP1.3ff,1890,1.6,1897,6.328,1909.Aquinasisthe“psychologist”ofthetrio(CP4.27).8.OneshouldnotunderestimatewhatPeircelearnedfromOckham,pos-itivelyandnegatively;seemyremarksinSectionIIIandBoler1980.F.Michael,whothinksPeircebeganasanominalist,isclearthattheearlynominalismwasnotfromtheScholastics(Michael1988:317).ButIamnotpersuadedbyMichael’sattempttomakeoutthatPeircethendevelopedamoresophisticatednominalismunderOckham’sinfluence(Michael1980:179–85).9.“TheonlylogicianswhoareinthesamerankasIareAristotle,DunsScotusandLeibniz.”(MSL482,quotedinFisch1986:250.)CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n78johnboler10.Onmethod:deWaal(1996:440n11).Ontheory:“Duringthemiddleages,purelyformalsyllogisticmadenoprogressworthmentioning”(CP1.567,1805).Thoughheknowstheyworkedmorewiththetheoryofconsequence:CP4.45;andonthattopic,seeMoody(1953).Ontheirrespectforlogic,see,e.g.,CP1.29–33,1869andCP2.12,1902.11.“Logicandmetaphysicswerestudiedwithaconsiderabledegreeofminutenessandaccuracy;sothatinspiteofabarbariccivilizationandotherunfavorableinfluences,sufficientlyobvious,theyreachedanex-cellencewhichourgenerationhasnotbeenabletoappreciate”(CP2.12,1902).12.Andalittlefurtheron:“theschoolmen,however,attachedthegreatestauthoritytomenlongsincedead,andtheywereright,forinthedarkagesitwasnottruethatthelaterstatusofhumanknowledgewasthemostperfect...”(CP1.32).13.“Whenthemethodofauthorityprevailed,thetruthmeantlittlemorethantheCatholicfaith.Alltheeffortsofthescholasticdoctorsaredi-rectedtowardharmonizingtheirfaithinAristotleandtheirfaithinthechurch,andonemaysearchtheirponderousfoliosthroughwithoutfindinganargumentwhichgoesanyfurther”(CP5.406,1878).14.Seenote11,above,andthewholeofCP1.29–33,1869for“thespiritofscholasticism.”TheconnectionwithPeirce’sideaofacommunityofinquirersisclear:“Thereal,then,isthatwhich,soonerorlater,in-formationandreasoningwouldfinallyresultin,andwhichisthereforeindependentofthevagariesofmeandyou.Thus,theveryoriginoftheconceptionofrealityshowsthatthisconceptionessentiallyinvolvesthenotionofaCOMMUNITY,withoutdefinitelimitsandcapableofadefiniteincreaseofknowledge”(CP5.311,1868).15.“Themedievaluniversitieswereplacesoflearningwhereoursarein-stitutionsforteaching”(CP5.582,1898).16.ManleyThompson(1952:note9)refersonthispointto5.436.17.SeeFisch,Kloesel,andHouser(1982)andmoregenerallyHookway(1985:Introduction).Peircegiveshisownaccount:CP1.3–14,1897,CP1.560,1907,andCP4.2–4,1898.18.Peirceisanastutereader:e.g.,herecognizesthatAristotle’sagentcausalityisnotaboutevents(CP6.66);heisfamiliarwiththeprac-ticecalledobligationes(CP5.340n,8.118);hehasaccuratethingstosayaboutAquinasonsignatematter(CP6.359–60),onangels(CP3.403I),andon“species”(CP8.18).AndhehasinterestingbitsonGreekandLatintermsforformandmatter(CP6.353),andontheshiftinlatermedievallogicfromthesyllogismtothetheoryofconsequence(CP4.45).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought7919.PeircesaysthatinhisstudyofKant’scategories,herealizedthatwhatwasneededwasabetterlogicalbasis:CP1.560,1905andCP1.563,1898;andthatseemstohaveledtohisstudiesinthehistoryoflogic.FischputsPeirce’s“mostintensive”studyofthemedievalsin1868–9(Fisch1986:188);PopkinandMeyersthinkhecouldhavebeenreadingmedievaltextsasearlyas1865(PopkinandMeyers1993:610–11).ThebestaccountofthedatingisE.Michael(1976:48).20.ThiswasprobablytheHarvardlecturesof1864–5and/ortheLowelllecturesof1866–7(Hookway1985:6,8).21.HecouldhaveusedtheHarvardlibrary;weknowhecametomakeconsiderableuseofPrantl(1955)andCousin(1836).AmongtheScotusvolumesIexaminedatJohnsHopkins,onehadthenameofhisfather,BenjaminPeirce,onthefly-leaf.22.AsinsomanyotherfacetsofPeircestudies,thebasicworkonthisincidenthasbeendonebyMaxFisch(Fisch1986:51ff).Ifollowhisaccount.23.Peircekeptarecordofwhatheboughtandwhere(Fisch1986:52).24.Evengrantingthathewastryingto“sell”thehigheradministrationonbuyingthebooks,thelibrarian’senthusiasmisgenuine:“Itisdoubt-fulwhetherasimilarcollectionexistsinanylibraryinthiscountry.”(QuotedinFisch1986:52.)25.Goodwin(1961:479–80)listsfivevolumesofScotusthatPeirce’swidowlatersoldtoJohnsHopkins.26.Peircemadeconsiderableuseofcertainsecondarysources,especiallyPrantl(seenote21,above),ofwhomheissometimescritical:CP2.218,CP323n,CP361,CP364,CP391n,CP393;CP5.4,CP83;CP6.312.27.CP1.28ff,LecturesonBritishLogicians,givenatHarvardin1869.SeealsoCP4.21–37.28.CP5.312nisanexplicitquotefromQuaestionesinMetaph.,VII,q18,n8.Itisnotamatterofinfluence,butMurphey(1961:117)pointsoutthatCP5.213ff.takesaftertheformofascholasticdisquisition.Thesecasesaredifferentfromthesimilaritiesdiscussedbelow.29.Also,helearnedalotaboutmedievallogicandlogicalandothertermi-nology,butIdonotcountthatasinfluence:seeCP2.225,CP3.159,CP7.494n9,CP7.395.30.E.Michael(1976)andKloesel(1981)discussPeirce’sappealtospecula-tivegrammar.ForworkonThomasErfurt,seeBursill-Hall(1971).31.McKeon(1952:241n5)notesthatPeircereferstootherbitsofpseudo-Scotus:e.g.,atCP1.549n1.32.EvenKloesel(1981:32)saysErfurt’sworklacksthebreadthandcom-prehensivenessofPeirce’stheoryofthenatureandmeaningofsigns.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n80johnboler33.SeeCP4.48forPeirceonEuropeangrammarandthesubject–predicateform.Imeanbythelatter,asIthinkotherswhousethephrasedo,abasicpropositionalformconsistingoftwocategorematicterms(subject-andpredicate-terms)withacopulaandthevarioussyncategorematicsthatmakeupthefamiliarAEIOformsofAristotelianlogic.The“logicofrel-atives”whichPeircedevelopsuseswhatwewoulddesribeaspredicate-functionswithinaquantificationalscheme.34.CP1.191,CP444,CP559,CP2.83,CP206,CP229,CP332,CP432,CP438,CP4.9,CP8.342.35.Seeforexample,CP2.60,CP62,CP64,CP83,CP229,CP232,CP3.432.SeeCP2.206for“Kant’sElementahrlehre.”36.Indiscussingtheneedtoupdateit,Peircedescribesthe“NewList”asspeculativegrammar(CP2.332,ca.1895).Murphey(1961)isinlargepartadetailedanalysisofthearticleanditsbackground.37.Thelistofcategoriesis“atableofconceptionsdrawnfromthelogicalanalysisofthoughtandregardedasapplicabletobeing”(CP1.300,ca.1894).Whateverthemedievalsmayhavemeantbyspeculativegram-mar,itisthisanalysisofthestructureofthoughtthroughthestructureofitsexpressionthatPeircemeansbyErkenntnishlehre.38.Seenote33,above.Asweknow,itiseasyenoughtotranslatefrom“AllSisP”to“Forallx,ifxisSthenxisP”andviceversa.Theproblemariseswithrelationalpredicateswhosestructureis“buried”inthecorrespondingmonadicpredicatesinS-is-Pform.Thelatter,there-fore,cannotsupporttheclassificationofmonad,dyad,andtriadthatisthestructureofPeirce’smaturecategorytheory.SeeHookway(1985:97ff.).39.Allthebook-lengthstudiestakeupthetopic.Amongarticlesdevotedtoit,McKeon(1952),Moore(1952),Bastian(1953),andGoodwin(1961)aretheearliest,thoughthereareahostthatfollow.Mineistheonlybooksofarspecificallyonthistopic,thoughitneedsqualification,asyoucanseeifyoureadthesearticles.AmongthosethatIdonotrefertoinothernotesspecifically,IwouldmentionE.andF.Michael(1979),Forster(1992),H.Lee(1982),R.Lee(1998),Nesher(1981),Rohatyn(1983),andRosenthal(1968)40.“Forassoonasyouhaveoncemountedthevantage-groundofthelogicofrelatives...youfindthatyoucommandthewholecitadelofnomi-nalism,whichmustfallalmostwithoutanotherblow”(CP4.1,1898).Anumberofcommentatorshavenoticedthisandtheimportanceofhisthreemodesofbeing:e.g.,F.Michael(1988:329–35)andRaposa(1984:151f.)41.AmongrecentstudiesofPeirce,seeHookway(1985).ForScotus,seeKing(1992:60and67)and(2001:especiallySection2.2);MarilynCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought81Adams’saccountofthedisputebetweenScotusandOckhamisverygood(Adams1987:chs.1–2).42.Forexample,onwhatispredicated;whetherwordssignifythingsdi-rectlyoronlythroughconcepts;whether“white”primarilysignifiesapropertyorthebearer.43.Peircerecognizesthedifference:“nogreatrealistheldthatauniversalwasathing”(CP1.27n).Andthismaybewhyhedistinguishestheear-lierdiscussionsfromthoseofthethirteenthandfourteenthcenturies:CP8.11,andseenote3,above.44.SeePeirceCP8.18.Ockhamshowsthecourageofhisconvictionsinholdingthatsincescienceisoftheuniversal,itisstrictlyaboutcon-cepts,althoughtheystandforrealindividuals(Ockham1990:11–12).45.ItisstandardformostaccountsoftheScholasticstodescribethisas“moderaterealism”preciselybecauseithastodowithspecialcompo-nents(dependentonandposteriortofirstsubstance)ratherthanwithacommitmenttogeneral“things.”Platoisfortheman“extreme”re-alist.AsIshallexplainbelow,PeircethinkstheappealtoconstituentsconstitutesScholasticrealismasopposedtonominalisticPlatonism;sothatisnotwherehelocatesScotus’smoderateness(i.e.,hisbeing“toonominalistic”).46.Iofferapluralon“natures”becauseScotusallowsapluralityofsub-stantialforms(seeCP7.580)andincludesamonggeneralsvarious“ac-cidents”:onformalitiesandmultiplehabitsseeGoodwin(1961:482).47.Hemarshallssevenarguments(King1992:n6)totheeffectthatthereisarealbutless-than-numericalunitythatgroundsourascriptionofcommonterms.48.Kingoffersaninterpretationthatallowsforbothacontractedandanuncontractednatureinthings(King1992:54–6),thoughIdonotthinkitwouldhavesatisfiedPeirceevenso.49.PeircegivesabriefbutaccurateenoughaccountofthepositionsofScotusandOckhamatCP8.18–21,1871.Itisimportanttokeepinmind,whiletalkingaboutthechangesPeircemakestoScholasticrealism,thathedoesnotdistortormanipulatethemedievals’positiontosupporthisown.(Thoughseenote72,below.)50.ItistypicalofPeircetoputthequestionofnominalismandrealismas“whetherlawsandgeneraltypesarefigmentsofthemindorarereal”(CP1.16).Anditiseasytoassumehemeanstocontrastthingsinsidethemindwithwhatisunequivocallyoutside.Butthecontrasthastodowiththeobjectsoftwokindsofthought;andthereal,ashesayselsewhere,isnotindependentofthought.51.“As[usually]stated,thequestionwaswhetheruniversals,suchastheHorse,theAss,theZebraandsoforthwereinreorinrerumnatura.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n82johnbolerButthatthereisnogreatmeritinthisformulationofthequestionisshownbytwofacts:first,thatmanydifferentanswersweregiventoit,insteadofmerelyyesandno,andsecond,thatallthedisputantsdividedthequestionintodifferentparts”(CP4.1,1898).NotetheparallelinCP8.12,quotedjustabove,whereonemighthavethoughtPeircemeantthisasjustthewaytodescribethedispute.52.“Thegistofallthenominalist’sargumentswillbefoundtorelatetoaresextraanimam,whiletherealistdefendshispositiononlybyassumingthattheimmediateobjectofthoughtinatruejudgmentisreal”(CP8.17).53.SeeBoler1985:121–3.54.“Roughlyspeaking,thenominalistconceivedthegeneralelementofcognitiontobemerelyaconvenienceforunderstandingthisandthatfactandtoamounttonothingexceptforcognition,whiletherealists,stillmoreroughlyspeaking,lookeduponthegeneralnotonlyastheendandaimofknowledge,butalsoasthemostimportantelementofbeing.Suchwasandisthequestion”(CP4.1,1898).55.“[T]hesectatorsofindividualism,theessenceofwhosedoctrineisthatrealityandexistencearecoextensive...goalongwithyouinholdingthat“real”and“existent”havethesamemeaning,orInhalt”(CP5.503,1905).56.ThethreemodesofbeingarethecategoriesofPeirce’slatertheory:Firstness,Secondness,andThirdness.SeethediscussioninHookway(1985:80–117).AsIsuggestbelow,theyarewhattheScholastics’“con-stituents”becomeinPeirce’smaturemetaphysics:seeCP1.22–3.AsFisch(1986:194)pointsout,Peircedoesnotuse“modeofbeing”talkuntilmuchlater(around1896,forexampleatCP1.432ffandCP1.51ff);andhecitesThompson(1953:182)whofindsonlytwomentionsofmodesofbeingbefore1902.Oncenominalismisunderstoodasthede-nialofallbutonemodeofbeing,itislesssurprisingthatPeirceshouldextendhispejorativeuseofthelabelsothatthemarkofthebeastisonanyfailuretoacknowledgethefullrange(viz.,three)ofmodesofbeing(seeCP5.79).57.CP5.502–4,1905.ForafulldiscussionofnominalisticPlatonism,seeAndersonandGoff(1998:165–78).Itsoundsabitodd,ofcourse,butitisnotunreasonabletoseePlatoasanominalist.TheIdeasarenotgenerals.Theyarenotpredicatesand“participatesin,”whichisthecanonicalpredicate,hasnocorrespondingIdea.58.Seenote45,above.59.F.Michaelalsoemphasizesthe“gap”(1980:185).60.SeeCP5.311,quotedinnote14,above.61.Goodwin(1961:489)recognizesformalitiesasconstituents.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought8362.“Thequestionofrealismandnominalism[is]thequestionofhowfarrealfactsareanalogoustologicalrelations”(CP4.68).And:“Thatwhichanypropositionassertsisreal,inthesenseofbeingasitisregardlessofwhatyouorImaythinkaboutit.Letthispropositionbeageneralconditionalpropositionastothefuture,anditisarealgeneralsuchasiscalculatedreallytoinfluencehumanconduct;andsuchthepragmaticistholdstobetherationalpurportofeveryconcept”(CP5.312).F.Michaelemphasizesthis(1988:330–4).63.SeealsoCP8.19–20andCP1.549n1.64.IninterpretingScotus,onemightwellwanttoplaydownthetalkofconceivability:seeKing(1992:60;and2001:Section2.2).Indescrib-ingScotusfromPeirce’spointofview,however,Iwanttobringoutthe“reality–conceivability”connectionaslyingbehindhisaccountinBerkeley.65.PeirceclaimsScotusbroughttheterm“real”intocommonuse(CP8.319),andIthinkthatisrelatedtorealitates:seeCP4.28,CP5.430,CP6.328andCP495,andCP8.14–18.66.Foracarefulaccountoftheformaldistinction,seeAdams(1987:22–9).67.Thatrealitydoesnothavetodowithconcepts,forPeirce,isclearinhisrejectionofconceptualismwhichheseesasjustaconfusedformofnominalism:1.27.68.Seenote56,above.69.AsIsaidearlier,IcannotdojusticeheretoPeirce’stheoryofthecate-goriesormodesofbeing.His“phenomenological”approachcanbeseeninCP1.284–353andCP5.41–65.ThethreecategoriesarealsodescribedinHookway(1995:chs.3and4).70.For“systems,”seeRaposa(1984:161).Forsome,thedifferenceisenoughtojustifytheclaimthatPeircegivesuponScholasticrealism:McKeon(1952:247ff),Thompson(1952:133,136f),Bastian(1953:246–9),withMoore’sreply(Moore1953:250–1),Murphey(1961:138ff.,401),Pihlstrom(1998d:n47).Goodwin(1961:509)seemstometohaveasensibleapproachtotheproblem.71.Peircedoesnot:1890ishisfirstuse(Murphey1965:131).McKeon(1952:245)citesCP1.458,1896,butsaysitisdifferentfromScotus.SeeCP6.95,1903,onScotusandKant;andCP6.319,1908on“whatScotusshouldhavemeant.”IthinkPeircecomesbacktohaecceityfromhis(later)interestinSecondness.In1901,Fraserproducedanewversion(notmerelyare-edition)oftheworksofBerkeley,andPeircerevieweditintheNation.WhilehementionsScotusinconnectionwithhaecceity,hisconcernhereistocriticizethesortofabsoluteidealismthattakesinsufficient(orno)noticeofthebruteencounterwiththeexistent(N3,36,1901).DiLeodevotesanarticletohaecceity(1991:79–107).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n84johnboler72.Forotherreferencestothisness/haecceity:CP1.405,CP408,1890,CP458;CP3.434,CP460–2,CP475,CP479,CP535,CP543,1896.Thereisaterminologicalproblemhere.IamcallingattentiontothedifferencebetweenScotus’shaecceityandPeirce’sSecondness.ButPeircetendstocoopt“haecceity,”makingitsynonymouswith“Secondness”;seeCP6.319quotedinnote71,above.73.Especiallyinhiscriticismofabsoluteidealism:seenote71,above.74.FischfindsFirstnessmoreimportantinPeirce’sdevelopment,thescholasticrealisminvolvedinThirdnesshavinggotPeirceonlyhalfwayoutoftheclutchesofnominalism(Fisch1986:228;andseeCP6.201).IadmittonothavingaveryfirmgriponFirstness.75.TimothyPottspointedthisoutinanearlyreviewofBoler(1963):(Potts1965:362).Peircesaysthat“thegeneralhasanadmixtureofpotential-ity”(CP1.420).“Itsmodeofbeingisesseinfuturo”(CP2.148;cf.,CP5.48andMcKeon(1952:239)).Itiscentral,ofcourse,inPeirce’snotionofthewould-be:e.g.,CP3.527,CP5.77n1,CP428&CP527,andCP6Bk.1,chs.1&2.Onthethingasabundleofhabits/powers:CP1.414.See:Goodwin(1961:497–507),Olshewsky(1981:87),Engel-Tiercelin(1992:51–82);anditisthethemeofRaposa(1984:147–68).ForScotus,seeKing(1992:67).76.AmoreradicalmodificationinPeirce’srealismistheemphasisoncon-tinuity:itiswhatgeneralitybecomesinthelogicofrelatives(CP5.436,CP6.172),asItriedtoexplaininmybook(Boler1963:ch.3).SeealsoHookway(1985:174–80).Ontheimportanceofthechangefromasubject–predicatelogictothelogicofrelatives,seeMurphey(1961:401),Olshewsky(1981:87,92),F.Michael(1988:329–34),Raposa(1984:151f.,andnote68),andShort(1996:420).77.PresumablythisiswhatPeircemeansbyScotus’s“haltingrealism”inCP6.175,1905.78.Fr.BastianmaynothavepaidenoughattentiontothewayinwhichanAristotelianformresemblesahigherlevelpower:apowerofpowers(as,e.g.,intheintellectualsoulwhichgroundsthemanypowersthathumansexhibit).ButheissurelytraditionalininsistingthatinthehylomorphismoftheScholastics,potentialityisgroundedintheessenceorsubstantialform(Bastian1953:246–9,espec.249).McKeon(1952:249)citesCP6.361,wherePeircerejectsform(andpresumablyanythingotherthanThirdnessitself)asthegroundfordynamicalpowerinagents.79.Aquinassaysexplicitlythatthemodeofunderstandingdoesnothavetomatchthemodeofbeing:SummaTheologiaeI,q.84,a.1.80.Peirce’srealismisnotasimpledoctrine(F.Michael1988:336).WhilePeircehimselfseemstoattributehisrealismtoaclosereadingofDunsScotus,theremayhavebeenotherfactorseveninhisearlierideaofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirceandMedievalThought85“Scholasticrealism.”Thereisevidence,forexample,thatFrancisAb-bottmayhavebeenastrongandmoreproximateinfluencethanthescholasticsatsomepointintheprocess(O’Connor1964:543–64).ForaverypositiveaccountoftheimportanceofPeirce’sextremerealismalongwithitsscholasticelements,see(Haack1992).81.Short(1996a:419)andMurphey(1961:esp.123–50)intheirownwaysmakeasimilarcomplaint.AndseeThompson(1952:138).82.Seenote70,above.83.Seenote19,above.84.RobertsdisputesFisch’sevidence,holdingthatPeircewasarealistallalong(Roberts1970:67–83).F.MichaeldefendsFischontheearlynom-inalism(1988:339)andShortreturnstoitscriticism(Short1996b:espe-cially416–20).BothMichaelandShortthinkFisch’s“firststep”isreallyonlyachangeoflabels,though(ifIhavethemright)MichaelclaimsthatPeirce’s“earlyrealism”wasnominalistic(1988:339)whileShortholdsthathis“earlynominalism”wasrealist(1996b:416–22).85.IwouldhavesaidthatPeirce’sScholasticrealism,alreadyevidentintheBerkeleyreview,wasonlyafirststeptowardhislaterrealism.ButthismightupsetFisch’sstoryline.HeisclearthatPeircedoesnotretreatfromthepositionintheBerkeleyreview(Fisch1986:198n3).86.Theremaybeaspectrumstretchingfromnominalismtorealism,whereonecouldseePeirceasbasicallyanominalistgraduallygivinggroundinthefaceofrealistconcernsorasbasicallyarealistgraduallyworkingouttheimplicationsofrealism.Onthe“half-fullorhalf-empty”anal-ogy,thelabelsdonotmakeallthatmuchdifference.Theissueisdoc-trines,ofcourse,andnotlabels.Theproblemwiththelabelsisthatdifferentpeople(Peirceincluded)havedifferentideasaboutwhatthevarious“isms”amountto.Hausman(1991:475–500)triestosortoutsomedifferentsensesofrealismandidealism.87.Seenote71,above.88.Fischrecognizesotherstorylines:(1986:390).Amongleadingcon-tenders,Iwouldhavepickedthedevelopmentsinhislogicaltheory,thetroubledjourneyfromobjectiveidealismtotherealasobjectofa(realorfictional)communityofinquiry,orperhapsthetransformation(orreplacementof)theNewListwiththethreemodesofbeing.89.Hisadvicethatoneshouldstartwithnominalismandacceptrealismonlywhenforcedto(CP4.1)suggestsmorethatatheoryisbeingdevel-opedthanoneabandoned.Hesaysthatnominalismisan“undevelopedstateofmind”(CP5.121,1903).Here,asinmoralassessment,thereisanimportantdifferencebetweenayoungpersonstillintheprocessofacquiringamoraloutlookandamaturepersonundergoingachangeofheart.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n86johnboler90.“Thenoveldoctrineoftheselectures,sofarasconcernsrealism,isthetheoryofImmediatePerception.Itisinthisconnectionthathefirstmakesitquiteclearthathisrealismisnowopposedtoidealismaswellasnominalism”(Fisch1986:195.)ForadetailedaccountofthedoctrineinPeirce,seeBernstein(1964).91.Olshewsky(1981:87,90)findsPeirce’smaturesemeiotictheorytoberepresentationalist;andifthatisright,theissueofhisepistemologicalrealismiscomplicated.Myonlypointisthatitdoesnotinvolvetherealismopposedtonominalism.92.‘[T]hattowhichtherepresentationshouldconformisitselfsomethinginthenatureofarepresentationorsign–somethingnoumenal,intelli-gible,conceivableandutterlyunlikeathing-in-itself”(CP5.553,1906).ThisisthreeyearslaterthanCP5.56(quotedjustabove)anditseemstomestilltohaveanidealistring.93.Seenote81,above.94.Fortunately,thelatterissueisnotmytopic,forIamnotclearaboutjustwherePeircefinallystandsonthematter.ButbothAltshuler(1982:esp.38–46)andHookway(1985:37ff.)seehimasanantirealist;thoughapparentlythelatterhaschangedhismindonthat(Hookway2000).95.O’Connor(1964:552ff.)claimsthatAbbottconvincedPeircethatepiste-mologicalandontologicalrealismstandorfalltogether.Engel-Tiercelin(1992:52)triestotiethetwo“realisms”together.Istillseeitaschangingthesubject.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\ndavidwiggins4ReflectionsonInquiryandTruthArisingfromPeirce’sMethodfortheFixationofBeliefMypaperofNovember1877,settingoutfromthepropositionthattheagitationofaquestionceaseswhensatisfactionisattainedwiththesettlementofbelief...goesontoconsiderhowtheconceptionoftruthgraduallydevelopsfromthatprincipleundertheactionofexperience,beginningwithwillfulbelieforself-mendacity,themostdegradedofallmentalconditions;thencearisingtotheimpositionofbeliefsbytheauthorityoforganizedsociety;thentotheideaofsettlementofopinionastheresultoffermentationofideas;andfinallyreachingtheideaoftruthasoverwhelminglyforceduponthemindinexperiencesastheeffectofanindependentreality.CP5.564,“BasisofPragmatism.”1906.(italicsnotinoriginal)Thethirdphilosophicalstratagemforcuttingoffinquiryconsistsinmaintainingthatthis,that,ortheotherelementofscienceisbasic,ultimate,independentofaughtelse,andutterlyinexplicable–notsomuchfromanydefectinourknowingasbecausethereisnothingbeneathittoknow.Theonlytypeofreasoningbywhichsuchaconclusioncouldpossiblybereachedisretroduction.Nownothingjustifiesaretroductiveinferenceexceptitsaffordinganexplanationofthefacts.Itis,however,noexplanationatallofafacttopronounceitinexplicable.That,therefore,isaconclusionwhichnoreasoningcaneverjustifyorexcuse.CP1.139“TheFirstRuleofLogic.”189987CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n88davidwigginsAbductionconsistsinstudyingfactsanddevisingatheorytoexplainthem.Itsonlyjustificationisthat,ifweareevertounderstandthingsatall,itmustbeinthatway.CP5.145“HarvardLecturesonPragmatism.”1903[Scientificprocedure]willattimesfindahighprobabilityestablishedbyasingleconfimatoryinstance,whileatothersitwilldismissathousandasalmostworthless.Frege1884:16I“TheFixationofBelief”waspublishedin1877asapopularessay.ButPeircemusthaveattributedtoitnotsimplytheliteraryfelicitythatwefindinit,buthighphilosophicalimportance.Forintheen-suingdecadesheconstantlyreturnedtothispaperasafocusfortheclarificationofhisthoughts,eitherenteringcorrectionsandamplifi-cationsorelseadaptingittonewphilosophicalinitiatives.Someoftheamendmentsweredesignedtoadjusttheessaytotheprojectsof“TheGrandLogic”and“TheSearchforaMethod.”Ourchiefcon-cernherewillbewiththeessayasPeircecametorereadandrewriteit,ratherthanwiththeessayinitsoriginalcondition.Thefirstofourepigraphs,whichisdated1906,issurelytheproductofoneoftheserereadings.Notonlydoes“Fixation”appearatleastasimportantasPeircesupposed.Thereradiatefromitsomeofthegrandestthemesofmod-ernphilosophy–thenatureoftruth,forinstance,andtherelationtruthhastomeaningwhenmeaningisoperationallyorpragmati-callyconceived;inquiryandtheethicsofbelief;theepistemicstatusofperceptualexperience;andtheproperaspirationoraspirationsofhypothesis.OnceitisseeninproperconjunctionwithotherPeirceanclaimsintowhichitleads,theessaywillevenpromisealineofre-sponsetoHume’sdoubtsabouttherationalbasisofoureffortstoarguefromtheknowntotheunknown.WhathaspreventedphilosophersfrominvestingthepaperwiththesortofimportanceinconnectionwithtruththatwefindPeirceattributingtoitinourfirstepigraph?Maybethetendencytoreadthepaperitselfasaphaseinaone-issuephilosophicalcampaigntodemystifytheideaoftruthbyredefiningitastheeventual,ifCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth89notpredestinate,opinionofthosewhoopen-endedlyandresolutelypursuethebusinessofinquiry.Onlyforasmallminoritywhostillespousesomesortofverificationismor“prope-positivism”(Peirce’sterm)couldsuchacampaignbeinterestingorconvincing.Recently,Peirceanscholars1havepointedtotheimplausibilityofattributingtotheexponentofatheoryofsignsandsignificationasspecialasthatofPeircetheprojectofofferingananalyticalde-compositionoftheconceptoftruth(orofanyotherconcept).Inthelightofthisdoubt,itwillnolongerdotosupposethat,atanytime,letalonein1906,bywhichpointhehadrecantedtheworstexag-gerationsof“HowtoMakeOurIdeasClear”(1878),Peircewouldhaveapprovedareport,giveninourlanguageasusedbyus,totheeffectthatCharlesSandersPeircethoughtthatanopinion’sbeingtrueanditsbeingtheeventualopinionweresimply,analytically,ornecessarilyoneandthesamething.Notonlyisthataquestionablereport.Unlesssomequitepeculiarsenseisattachedto“theeventualopinion,”itappearsinconsistentwiththatwhichwereadtowardtheendof“TheFixationofBelief”andreadagaininPeirce’ssubsequentreportsofitscontent(e.g.,thatquotedfrom1906).Itistimetosu-persedetheformofwordsthatPeircetooktheriskofusingwhenhewrote:“Theopinionwhichisfatedtobeultimatelyagreedbyallwhoinvestigateiswhatwemeanbytruth,andtheobjectrepresentedinthisopinionisthereal”(CP5.407,1877).For,ifPeirce’sideasaretoreachagainintothebloodstreamofphilosophy,thenweneednotonlyfreshstudiesofhistextsbutspeculativetranspositionsoftheseideas–transpositionsrecognizedasspeculativebutgiveninlanguagethatcanbeunderstoodwithoutanyreferencetospecialorpeculiarstipulations.(Totheextentthatourchosenvehicleofexpressionisphilosophicallanguage,letusconfineourselvestosuchportionsofitasaresecurelycantileveredfromtheeverydaylanguageofthosewhoarewonttolistentowhattheyaresaying.)UnderthetranspositionIshallproposehere,theexactmeaningsof“ultimately”and“endofinquiry”willnolongerbeanygreatissue,andtherelationbetweentruthandinquirywillbeadegreeortwomoreindirectthanPeircewasaccustomedtoallow.AnotherPeirceanpreoccupationweshallabandonisPeirce’sthoughtthatproperconductofinquirynotonlycomprisesthemotiveandmeansforcorrectingitsownconclusionsbutisboundinthelongruntoironouteveryerror.ThisperilousclaimwillbenopartofthecoreCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n90davidwigginsPeirceanismthatisheretobeconsolidatedanddefended.Whatthenwillremain?Aboveall,theideaofinquiry,seenalwaysasaprocessthatgathersrationalstrengthasitgathersforceandgathersforceasitgathersrationalstrength,aprocessatoncecommunalandpersonal,inwhichparticipantsreceivebenefitsthatareindefinitelydivisibleamongthemandreciprocate,inthelightoftheirownexperiencesandreflections,howevertheycanorarepermittedtodo.IIPeircesaysin“Fixation”that,withrespecttoanyquestionthatcon-cernsus,belieforopinionisthestateweseektoattainanddoubt(notknowingwhattothinkaboutthisorthat)isthedisquieted,dissatis-fiedstatethatweseektoend.Theessayreviewsfourdifferentbutde-velopingmethods:themethodofdogmatismortenacity,themethodofauthority,theapriorimethod,andthemethodofexperience,whichPeircehimselfapprovesandcommendstohisreader.Thislastmethodembraceslogic,inthebroadnineteenth-centurysenseoftheterm.The“distinctionbetweengoodandbadinvestigation...isthesubjectofthestudyoflogic....Logicisthedoctrineoftruth,itsnatureandthemannerinwhichitistobediscovered”(CP7.320–1,1873).InPeirceanusage,logicisthegeneralartofreasoning–nothinglessthaneverythingwhich“TheFixationofBelief”opensoutinto–anditsubsumestheartofmakinginferencesfromtheknowntotheunknown.Logicembracesnotonlydeduction,notonlyinduction,whichisthetestingofhypotheses,butalsoabduction,whichistheframingofexplanatoryhypotheses.“Reasoningisgoodifitbesuchastogiveatrueconclusionfromtruepremissesandnototherwise[good],”(CP5.365,1877)Peircewrote.Later,heamendedthissentencetosay“Reasoningisgoodifitbedominatedbysuchahabitasgenerallytogiveatrueconclusionfromtruepremises”(CP2.11,1902).Nineteenth-andtwentieth-centuryconceptionsoftheprovinceoflogicareinterestinglydifferent.Eachconceptioninsists,however,ontheincompletenessofthecanonthatweshallrecognizeatanypointforgoodreasoning.Twentieth-centurylogicianshavestressedtheessentialincompletenessofcanonsofprovability,anincom-pletenessdemonstratedforpurposesofstrictlydeductivelogicbymeta-mathematicalinvestigationinitiatedinthenineteen-thirties.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth91ForPeirce,theincompletenessoflogicinhisbroadersenseismademanifestinconstantextensionsofthemethodsofscientificargu-mentandtheconcomitantrenewaloftheabductiveinitiativeofacommunityofinquirers.“Eachchiefstepinsciencehasbeenalessoninlogic”(Fixation,W3,243,1877).”2InsofarasaPeirceanphilosophyoftruthwillelucidatetruthbyreferencetoinquiryitselfandinquirybyreferencetoastruggleagainstdoubt(=notknowingwhattothinkaboutthisorthat)whichfindsitsfinalfruitioninthedispassionofapurescienceofunboundedaspiration,itisnottobeexpectedthattheelucidationwillhavetheeffectofcircumscrib-ingtruthitselforlimitingittothatwhichisdiscoverablebyanyparticularresearchmethodoraggregationofresearchmethods.Theinsatiabilityoftheinquiringmentality,liketheordinarydis-comfort(fromwhichthescientificoutlookoriginates)ofnotknow-ingwhattothinkaboutsomeparticularquestion,isonepartofthebackgroundfortheveryideaofbelieforopinion.Sotooisthecalm-nessandsatisfactorinessofknowingwhattobelieve.ChristopherHookway3hasbeentroubledthatPeirceshouldatoncehavecon-demnedpsychologisminlogicandintrudedpsychologicalfactsintohisaccountofinquiry.Butifweseethesebackgroundfactsascon-ditioningtheemergenceoffullyfledgedopinionorbelief–ifweseetheconcernfortruthaslatentalreadywithinthenatureofopinionandbeliefthemselves,inquirybeingtheexpressionofthatverycon-cern–thenIhopewecanexemptPeircefromthechargethatheallowsmerecontingenciestocorrupthisconceptionsoflogicandtruththemselves.Thesearenotmerecontingencies.Theyaretheenablingfactsfortheexistenceofbeliefitself–andofanynorma-tivescienceofthatwhichdeserves(howevertentativeandfallible)belief.IIIItmightbequestionedhowexactlyandfaithfully,datingfrom1906,ourfirstepigraphreflectsPeirce’sintentionsof1877.4Butitmat-tersfarmoreforpresentpurposes–anditmattersespeciallyforthePeirceanconceptionoftruth–howPeircehimself,inhisfullmatu-rity,wantedtoreadorrereadorrewrite“Fixation”andwhatplacehecametowantittooccupywithinthecontextofhismatureposition.If“Fixation”readwiththeemphasesPeircesuggestsin1906canCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n92davidwigginshelpustoseetheconceptionoftruth“gradually”and“undertheac-tionofexperience”emergefromtheabandonmentofdogmatismandauthoritarianism,andmakeusseeitemergethencebyvirtueoftheworkingsoftheprinciplethattheagitationofaquestionceaseswhensatisfactionisattainedwiththepropersettlementofbelief,thenthenexttaskforthephilosopherofinquirywillbetospeculatewhatitisaboutthenotionsoftruthandbeliefthatfitsthemtocohereandconsistwithoneanotherinthisway.Forbeliefs,truthmustbethefirstdimensionofassessmentoftheirgoodnessandbadness(oftheireligibility,sotospeak),evenastrueopinionmustbeourpreeminentaspirationifweask“WhatshallIbelieveaboutsuchandsuchorsoandso?”Thesearenormativeclaimsofasort,conceptuallyfounded.Underbothaspects,theyaremoreorlessindispensabletothecor-rectnessofPeirce’sconceptionofinquiry–and,insofarasplausible,supportiveofit.IVOurfirstepigraphrecapitulates“Fixation,”butalongercommentaryisneeded.ThismayusefullybeginwithPeirce’sclaimthatbelieforopinionisthestateweseektoattainanddoubtthestateofirrita-tionweseektoend.Thelatentcomplexityofthissimple-seemingdeclarationmaybebroughtoutbyananalogy.Supposethatsomeonehasappealedtomeformyhelp,Imakeanexcuse,andthenIfeelashamedoflettingthemdown.Findingithardtolivewiththisfailure,Itryatfirsttoforgetallaboutthematter.(Afterall,Ididn’towethepersonanyhelp,ratherthereverseperhaps.Andthereareallsortsofotherpeopletheycouldhaveappealedto.)Supposethat,withinmyownmind,thisdoesn’tworkandIstarttowonderwhyIhaven’thadasimpler,morestraightforwardthought:oughtn’tItogobacktothepersonandseeifthereisanythingleftthatIstillcando?Supposethat,pursuingthatverythought,Iministerdirectlytotheobjectofdisquietandthen,likeacloud,thedisquietitselfdisappears.Mutatismutandis,comparenowthedisquiet/dissatisfactionofnotbeingsurewhattothinkaboutwhether...(someparticularques-tion,thatis).Shouldsuchdisquietberemediedbyattentiontothestateofmind,vexatiousasitis,ortotheobjectofthestate?IfPeircehadbeenaskedthisquestion–andwhatapityhewasn’t–thenhowCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth93wouldhehavereplied?Idon’tknow.ButIthinkhisreplyoughttohavebeenthis:thefirsttwoofhismethodsoffixationseektoworkdirectlyuponthestate;thethirdmethodistransitional;thefourthworksdirectlyupontheobjectofthedisquiet.Aswereviewthefourmethods,however,letthereaderverifythisforhimselforherself.Concerningthemethodoftenacityordogmatism,Peircesaysthatthesocialimpulse,whichcomprisestheinnercompulsiontopayanxiousheedwhereverothersthinkdifferentlyfromoneself,practi-callyguaranteesthetotalineffectivenessofthismethodtoimplantormaintainconvictionorforestallthedisquietofnotknowingwhattobelieve.Thesecondmethodisthemethodofauthority,consistingofdog-matismsupportedbytherepressionofsocialimpulsesthatunsettleprescribedopinion.HerePeirce’sprescriptionisthis:Let[men’s]passionsbeenlisted,sothattheymayregardprivateandunusualopinionswithhatredandhorror.Then,letallmenwhorejecttheestablishedbeliefbeterrifiedintosilence.(W3,250“Fixation,”1877)[L]etitbeknownthatyouseriouslyholdatabooedbelief,andyoumaybeperfectlysureofbeingtreatedwithacrueltylessbrutalbutmorerefinedthanhuntingyoulikeawolf.(W3,256“Fixation,”1877)Eventhoughthismethodholdsbetterpromisefortheendofdoubtthanthefirst,anditspasttriumphsaremanifest,Peircethendeclares–inpassageswhichseeminthelightofrecenteventsinEasternEuropenotonlypropheticbutvividlyillustrativeofwhy,underthosekindsofconditions,thingswillalmostinevitablytendtoworkoutinacertainway–thatsuchapolicywillbepowerlessintheendtocountertheirritationofdoubtortostabilizeopinion:[N]oinstitutioncanundertaketoregulateopinionsuponeverysubject.Onlythemostimportantonescanbeattendedto,andontherestmen’smindsmustbelefttotheactionofnaturalcauses.(W3,251“Fixation,”1877)Forthatreason,Peirceholds,oncesomepeopleareledbyunreg-ulatedconvictionstorejectthatwhichisofficiallyprescribedforgeneralbelief,moreandmoreotherswillcometothinkthattheirownadherencetothisorthatapprovedopinionmaybeowedto“themereaccidentofhavingbeentaughtas[they]have.”Wherepeoplearealreadytendingtowarddoubt,thebeliefsthattheythinktheyowetothissourceareboundtocomeadrift.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n94davidwigginsEvidentlythen,anewmethodofsettlingopinionsmustbeadopted,whichshallnotonlyproduceanimpulsetobelieve,butshallalsodecidewhatpropositionitiswhichistobebelieved.Lettheactionofnaturalpreferencesbeunimpeded,then,andundertheirinfluenceletmen,conversingtogetherandregardingmattersindifferentlights,graduallydevelopbeliefsinharmonywithnaturalcauses.(W3,252“Fixation,”1877)Peircecallsthisthirdmethod,theapriorimethod,new.Butallthat“new”needstomean(Isuggest)isthathehasputitnextaftertenac-ityandsubmissiontoauthorityinhisenumerationofremediesfordoxasticdisquiet.Forhesaysoftheapriorimethod“solongasnobettermethodcanbeapplied,itoughttobefollowed”because“itistheexpressionofinstinct,whichmustbetheultimatecauseofbeliefinallcases.”Underthisaspect,theapriorimethodisonlyaresumptionofprotorationalwaysofinformationgathering.Self-evidently,then,evenwhendignifiedasamethodfor“thefermenta-tionofideas,”(CP5.564,1906)themethodcanonlyrestorethestatewherewewerebeforeweturnedtotheseotherexpedients.Itisnotsurprisingthenif(asPeirceclaims)itsfailurehasbeenthemostmanifest.Itmakesofinquirysomethingsimilartothedevelopmentoftaste;buttaste,unfortunately,isalwaysmoreorlessamatteroffashion...[And]Icannothelpseeingthat...sentimentsintheirdevelopmentwillbeverygreatlydeterminedbyaccidentalcauses.Now,therearesomepeople,amongwhomImustsupposethatmyreaderistobefound,who,whentheyseethatanybeliefoftheirsisdeterminedbyanycircumstanceextraneoustothefacts,willfromthatmomentnotmerelyadmitinwordsthatthatbeliefisdoubtful,butwillexperiencearealdoubtofit,sothatitceasestobeabelief.(W3,253,“Fixation,”1877myitalics)ThelastsentenceisoneofthemostimportantsentencesinPeirce’swholeessay.Itsuggestsinteraliathatthosewhopracticethefirstorsecondmethodshavemisunderstoodthenatureofthedisquietorir-ritationofnotknowing.Oncetheyunderstandthisbetter,theywillreturntotheobjectoftheirdisquiet,namelytheparticularthingnotknown.WouldthatPeircehadsaidmorehere,andwouldthathehadattendedseparatelytotheaspectsofthematterthatappeartothein-dividualinquirerandthosethatappearunderacollectiveaspect.ButthelastsentencequotedisthepointoftransitiontoPeirce’sfourthCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth95methodofcounteringourdisquietatnotknowingornotknowingforsure:Tosatisfyourdoubts,therefore,itisnecessarythatamethodshouldbefoundbywhichourbeliefsmaybecausedbynothinghuman,butbysomeexternalpermanency–bysomethinguponwhichourthinkinghasnoef-fect....[Thatexternalpermanency]mustbesomethingwhichaffects,ormightaffect,everyman.And,thoughtheseaffectionsarenecessarilyasvar-iousasareindividualconditions,yetthemethodmustbesuchthattheultimateconclusionofeverymanshallbethesame.Suchisthemethodofscience.Itsfundamentalhypothesis,restatedinmorefamiliarlanguage,isthis:Therearerealthings,whosecharactersareentirelyindependentofouropinionsaboutthem;thoserealitiesaffectoursensesaccordingtoregularlaws,and,thoughoursensationsareasdifferentasareourrelationstotheobjects,yet,bytakingadvantageofthelawsofperception,wecanascertainbyreasoninghowthingsreallyandtrulyare,andanyman,ifhehavesuf-ficientexperienceandreasonenoughaboutit,willbeledtotheonetrueconclusion.Thenewconceptionhereinvolvedisthatofreality.(W3,253–4“Fixation,”1877)Here,insofarasweareinfluencedbytheglossof1906,weshallunderstandPeircetosaysomethinglikethis:thatanyonewhohastheideathattheproperresponsetotheirdoubtordisquietaboutthisorthatistoworkontheparticularobjectoftheirdissatisfactionandinitiatesomecarefulinquiryisinapositiontodiscoverthatthisapparentlysimpleaimmust,iftheyaretoescapetheirdisquiet,comprehendwithinitnothinglessthanthis:thatanyopinionorbelieftheyarriveattotheeffectthatpshouldbedeterminedbycircumstancesthatarenotextraneoustothefactthatp.Meanwhile,thephilosopherofinquiry,reflectingontheaimthatanordinaryinquirerwillhaveifhefeelsthedissatisfactionofnotknowingwhattobelieve,mustbetemptedtothinkthatthisneedoftheinquirer’s,thiswantinghisorherbelieftobedeterminedinjustsuchaway,isexactlywhatisneededforustobegintounderstandtheideaof“areality”towhichthebeliefthatpisanswerable.Hereithelpstodistinguishverydeliberatelytherolesofinquirerandofphilosopherofinquiry.Normally,whenweengageasinquir-ersinsomeinvestigation,wedonotthink,intheabstract,aboutmethodology.ThePeirceanphilosopherofinquiryknowsthat.Butif,evenasinquirerssubmittoexperience,theydoreflectabstractlyabouttheirproceduresandtherationaleofwhattheydo,then,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n96davidwigginsaccordingtoPeirce,thethingtheyareboundtofindtheyhavedis-coveredistheideasoftruth,offact,andofarealityor(asPeircerewrotesomepassagesof“Fixation”tosay)“aReal.”Thecorre-spondingroleofthephilosopherofinquiryistomakethehoweverinexplicitworkingideasofinquirersmoreexplicit,thusharvestingtheoutcomeoftheapparentlymeagerresourcesthatthemethodol-ogyofinquirystartsoutwith.Amidthisharvest,onceitisproperlyexamined,alongwiththeideasoftruth,fact,andreality(Real),thetheoryorphilosophyofthepracticeofinquirywillalsofind“thefundamentalhypothesis,”whichspeaksofourtakingadvantageofourperceptionsandthe“lawsofperception,”inordertoascertain“byreasoninghowthingsreallyandtrulyare.”5Considerationofthishypothesiswillforceuponaphilosopherofinquirythetaskofarrivingatapropercon-ceptionofexperience.Forexperienceisthatbywhichwecananddoexposeourmindstorealities/Realsandmakeourbeliefsanswerabletorealities/Reals.TheforcibleelementinourexperienceiswhatPeircecalls“secondness.”6“Itmaybeasked,”Peircenotes,“howIknowthatthereareanyrealities[Reals].”TothisquestionPeircegivesfourreplies,ofwhichthemoststriking,interesting,andcon-clusiveisthisone:Thefeelingwhichgivesrisetoanymethodoffixingbeliefisadissatisfactionattworepugnantpropositions.Butherealreadyisavagueconcessionthatthereissomeonethingtowhichapropositionshouldconform.Nobody,therefore,canreallydoubtthattherearerealities[Reals],or,ifhedid,doubtwouldnotbeasourceofdissatisfaction.Thehypothesis,therefore,isonewhicheverymindadmits.Sothatthesocialimpulsedoesnotcausemetodoubtit.(W3,254“Fixation,”1877)Thereismoretosayaboutrealities(Reals)(seeSectionIX)andrealityandthereismoretobesaidalsoabouttheextentofPeirce’scommitmenttocausalrealism(seeSectionVI),butsuchinbareout-lineisPeirce’sdoctrine.Beforewecanconvertanyofthisintoadis-tinctivelyPeirceancontributiontothephilosophyoftruth,however,somefurtherexplanationanddefenseisneededofPeirce’sconcep-tionofbelief,oftheabductivecolorationthathegivestotheideaofexperience,andofhisideaofabductionitself.Weshallattendtoeachofthesethings,inSectionsV,VI,andVII.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth97VAttheoutset,Peircesaysthatbeliefinaparticularpropositionisacalmandsatisfactorystate.Itisastate“wedonotwishtoavoid,ortochangetoabeliefinanythingelse.Onthecontrary,weclingtenaciouslynotmerelytobelievingbuttobelievingjustwhatwedobelieve.”(CP5.372,1902–03)Inthesametenorhewrites,Withthedoubt...thestrugglebegins,andwiththecessationofdoubtitends.Hence,thesoleobjectofinquiryisthesettlementofopinion.Wemayfancythatthisisnotenoughforus,andthatweseek,notmerelyanopinion,butatrueopinion.Butputthisfancytothetest,anditprovesgroundless;forassoonasafirmbeliefisreachedweareentirelysatisfied,whetherthebeliefbetrueorfalse.Anditisclearthatnothingoutofthesphereofourknowledgecanbeourobject,fornothingwhichdoesnotaffectthemindcanbethemotiveforamentaleffort.Themostthatcanbemaintainedisthatweseekforabeliefthatweshallthinktobetrue...and,indeed,itisameretautologytosayso.(W3,248“Fixation,”1877)Fromthisconclusionaboutthethoughtsfirst-orderinquirerscanhavePeircederivesthreeattractivecorollaries,corollariesthatmain-taintheimpossibilitycumpointlessness,intheabsenceofrealandlivingdoubt,ofanygeneralprojectorplanofexposingeverythingtoquestion.(ContrastDescartes.)Ilikethatandshallreturntooneas-pectofitsmostdistinctiveimportanceinSectionXV.Butoneoughtnottopermittheclaimthatthesoleobjectofinquiryisthefixationofbelieftoescapecriticismjustbecauseitdeliversconclusionsthatwehaveotherreasonstofindattractive.Sotheclaimneedsmoreextendedcomment.Indeeditrequiresemendation.Thedirective“seekatruebelief,”Peirceseemstoargue,hasnomorepracticalcontentthan“seekabeliefyouthinktrue.”Andthenhecontinues,“wethinkeachoneofourbeliefsistrue.Itisameretautologytosayso.”Ifdoubtirritatesus,Peirceseemstosay,hisadvicetousis“seekforabelief,”not“seekforatruebelief.”Thisdoesnotlookright.Ifitiswrong,moreover,roomcanandmustbemadeinourtranspositionofPeirce’stheoriesofinquiryandtruthtocorrectthemistake.ThedefectofPeirce’swayofarguingbecomesveryevidentassoonaswerecallthatsimilarclaimsusedtobemadetotheeffectthatthereisnopracticaldifferencebetweenthedirectives“doyourduty”and“dowhatyouthinkisyourduty.”CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n98davidwigginsMakersofsuchphilosophicalclaimsalwayspaidtoolittleattentiontothefactthatitisnotwithoutconsequencewhichofthesedirec-tivesyousecond.Onewhofailstothinkhardwhathisdutyisbutdoeswhathetakestobehisdutyobeysthedirectivetodowhathethinksishisduty,buthedoesnotnecessarilydohisduty.InPeirce’smannerofargumentaboutseekingabelief,thereisacloselyparal-leloversight.Onewisheshehadnotsaidwhathesays.Itwillmissthepoint,though,toinsisttoomuchuponit.For,aswehaveal-readyseen,itbecomesclearslightlylaterinPeirce’sessay–seethetransitionfromthethirdtothefourthmethod–thatheisdeeplyimpressedbyaparticularandspecialpointaboutbeliefandthecon-ditionsthatareconstitutiveofbelief,namelythatthebeliefthatp,oncechallenged,isastatewhichneeds,onpainofextinction,toseeitselfasastatenot“determinedbycircumstancesextraneoustothefacts[concerningwhetherornotp].”Byitsnature,beliefisatouchy,uncomplacentconditionofthemind,adispositionwhichwillnotandcannotstayaroundonjustanyoldterms.Thisisoneofthethingsthatbringintobeingthenormativescienceoflogic.Onceweabsorbthesepoints,charitywillsuggestthatweshouldseePeirce’sinsistencethatthesoleobjectofinquiryisthesettlementofopinionastantamounttohissayingthis:“Believewhatyouwill–endtheirritationofdoubthoweveryoulike–onlyprovidedthatthebeliefwithwhichyouconquerdoubtwillstick,provideditreallywillconquerdoubt.”7Sounderstood,theprescriptionsuggeststhat,giventheexigencethatPeircefindslatentinthestateofbeliefandgiventheobject-directednessofthedisquietofnotknowing,theinjunctiontogetoneselfabeliefinordertoendtheirritationofsomedoubtcanneverbesatisfiedbypossessingoneselfofjustanyopinionorjustanysubstituteforabeliefthatgivesdissatisfactionordisquiet.Onceaquestionarisesthatonecaresabout,onecanonlybesatisfiedbyananswertoitthatonetakesoneselftohaveacquiredinamannerthatispropertothecontentoftheanswer.8WhatthenoughtPeircetohavesaidwasthewholeaimofinquiry?Itwouldhavebeenbetter,andmuchlessopentomisunderstanding,ifhehadsaidthatthewholeaimofinquirywastoendtheirritationofnotknowing(whether/who/when/what/....)bybringingintobe-ingtheproperconditionsforthesettlementofopinionwithrespecttothematterthatisinquestion.Thewholeaimistosecureevery-thingthatittakestoobtainthissettlement.Oncewesaythat,letitCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth99benoted,thereisscarcelyanytemptationatalltomakePeirce’smis-takeofsayingthatthereisnodifferencebetweenseekinganopinionandseekingatrueopinion.Nodoubtthenotionoftruthlurkswithinthenotionofthe“properconditionforthesettlementofopinion.”Butwearealreadypreparedforanydisappointmentsthisholdsfortheprojectofphilosophicalanalysis.Analysisassuchisnottheonlypossibleaim.SeeSectionsIandXII.VISomuchforbelief.NowletuscontinuethedescriptionofPeirce’sfourthmethod,andfillouttheaccountofthefundamentalhypothe-sis,ofrealities(Reals),andofthat“externalpermanencyuponwhichourthinkinghasnoeffect”whichwillundertherightconditionspromptbeliefstous.WhenPeircespeaksof“realities/Realsaffect[ing]oursensesac-cordingtoregularlaws”orspeaksofanyonewithsufficientexperi-enceandwillingnesstoreason“takingadvantageofthelawsofper-ception”inorderto“ascertainbyreasoninghowthingsreallyandtrulyare,”thecasesthatfirstcometomindasillustrationsofthisstrange–seemingdoctrinearesingularempiricaljudgmentsrelatingtothepastandpresentastreatedbythecausaltheoriesofmem-oryandperception.ItwouldbeapityiftheseweretheonlycasesthatweretractablebyPeirce’stheoryofinquiry.Nevertheless,letusstartintheareathatiseasiestforthedoctrineandbeginbyaskingwhatkindofreasoningitisthatleadstheretothatascertaining,andhow,inthemoststraightforwardperceptualcase,Peirceenvisagesitsworkings.Theanswertothequestionisthatthisreasoningisabductiveorretroductive,thoughinaspecialway.Evenforthenormalcase,ofperceptionormemory,Peirceoffersnosystematicaccountoftherelationbetweenperceptionandabductionormemoryandab-duction.Wedo,however,getsomeindicationsofthelinkhesawbetweenhypothesisorabductionandstatessuchasmemoryorperception:IoncelandedataseaportinaTurkishprovince;andasIwaswalkinguptothehousewhichIwastovisit,Imetamanuponhorseback,surroundedbyfourhorsemenholdingacanopyoverhishead.AsthegovernoroftheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n100davidwigginsprovincewastheonlypersonageIcouldthinkofwhowouldbesogreatlyhonoured,Iinferredthatthiswashe.Thiswasanhypothesis.Fossilsarefound;say,remainslikethoseoffishes,butfarintheinteriorofthecountry.Toexplainthephenomenon,wesupposetheseaoncewashedoverthisland.Thisisanotherhypothesis.NumberlessdocumentsandmonumentsrefertoaconquerorcalledNapoleonBonaparte.Thoughwehavenotseentheman,yetwecannotexplainwhatwehaveseen,namely,allthesedocumentsandmonuments,withoutsupposingthathereallyexisted.Hypothesisagain.Asageneralrule,hypothesisisaweakkindofargument.Itofteninclinesourjudgmentsoslightlytowarditsconclusionthatwecannotsaythatwebelievethelattertobetrue;weonlysurmisethatitmaybeso.Butthereisnodifferenceexceptoneofdegreebetweensuchaninferenceandthatbywhichweareledtobelievethatweremembertheoccurrencesofyesterdayfromourfeelingasifwedidso.(W3,326–7,“Deduction,Induction,andHypothesis,”1878,myitalics.)Thisisthecaseofmemory.Forthecaseofperception,wehavethefollowing:...abductiveinferenceshadesintoperceptualjudgmentwithoutanysharplineofdemarcationbetweenthem;or,inotherwords,ourfirstpremisses,theperceptualjudgments,aretoberegardedasanextremecaseofabductiveinferences,fromwhichtheydifferinbeingabsolutelybeyondcriticism.Theabductivesuggestioncomestouslikeaflash.Itisanactofinsight,thoughofextremelyfallibleinsight.(CP5.181,1903)WhatdidPeircehaveinmindwhenheclaimedthatperceptionandmemorywereabductive?AnyadequateanswerforthecaseofperceptionwouldneedtocoherewithtwootherPeirceandoctrines(thefirstnot,inthelightofourearliermentionof“secondness,”unexpected):...thisdirectconsciousnessofhittingandgettinghitentersintoallcogni-tionandservestomakeitmeansomethingreal(CP8.41,c.1885)and,second,ThechairIappeartoseemakesnoprofessionsofanykind,essentiallyem-bodiesnointentionsofanykind,doesnotstandforanything.Itobtrudesitselfuponmygaze;butnotasadeputyforanythingelsenor“as”anything.(CP7.619,1903)CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth101AfullreconstructionofPeirce’sdoctrinewouldnotonlyhavetoco-herewiththesecluesandwithhisfallibilism(whichsuggeststhatwhatis“beyondcriticism”mustbenotaperceptualbeliefbuttheperceptualstateitselfwhichmayormaynotsustainabelief).ItwouldalsoneedtocoherewithPeirce’snumerousbutsketchyhintsaboutthedistinctrolesinperceptionofpercept,percipuum,andper-ceptualjudgment.Inlieuofsuchareconstruction,Iofferaninterimstatement.Itisintendedtorespectmostoftheseconstraints,butitisnotgiveninPeirceanlanguage.Supposeobjectandperceiverencounteroneanotherinperception.Thenindependentlyofwillorreason,theperceivermaybemovedtoreportwhatheseesbyutteringthewords“Sixwindowsobtrude,itseems,uponmygaze.”Noabductionyet.Butfortheperceivertotakewhatheisconfrontedwithforsixwindowsjustis–whetherheknowsitornot–forhimtotakeitthatthebestexplanationofhisperceptionisthattherearesixwindowsthere.Mutatismutandisitwillbethesamefortherememberingcase.Fromremembering(oritsbeingasifoneremembers)themessengergivingonealetteryes-terday,oneconcludesthatthemessengerdidindeedgiveonealetteryesterday.Nothingelse(here)willexplain(here)one’sconvictionthathedid.Again,whetheroneknowsthisornot,theconclusionisabductive.Ratheritisalimitingcaseofabduction,orsoPeircesupposes.Insofarasonetakesoneselftoremember,oneiscommit-tedtoaccepttheconclusionofanabductionofcourse,thisisathirdpersonremarkaboutthelegitimacyofwhattheinquirerdoes,notareconstructionofhisthoughts.Canwegeneralizethis?Well,itseemstherelationofexperienceandbeliefmustbethis:thattheexperiencecreates,byitsnatureasexperience,afalliblepresumptionthatwhatwearemovedtore-portthatweseeorrememberisthatwhichaccountsforourbeingsomovedtoreport.Ratherthanattributethoughtsofthiskindtoordinarypercipientsorintellectualizethatwhichneedsnottobeintellectualized,onemightsaythattheacceptabilityofabductionisquietlyandtacitlyinstitutionalizedinourexerciseofourfaculties,inourpracticeandinthetitlethatperceiverscouldclaimthattheuseofsensesormemoryaffordsforthemtomakeempiricalclaims.EchoingaformulationthatappealedatonetimetoA.J.Ayer,thephilosopherofinquirycansaythatitistheoutcomeoftheexer-ciseofthesefacultieswhich,withrespecttocertainindispensableCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n102davidwigginsjudgments,givesordinaryinquirersinordinarycircumstancestherighttobesure.Suchnormativeclaimsasthisareundergirdedbynonnormativelawswhosespecialdependabilitywouldlegitimateanabduction.VIIHereendstheinterpretationandexplicationof“TheFixationofBe-lief”(atleastwithrespecttojudgmentsconformingtotheeasiestempiricalparadigm).Indeed,inoneway,wearewellbeyondtheendofthepaperitself.Butwearenotyetattheendofexpoundingthefourthmethod,themethodofexperience,whichisstillinthecon-ditionofaprogramneedingtobeworkedout.Thefourthmethoddependsonabduction,notonlyinthe(limiting)perceptualcasebutforalmosteverythingelsethatwecanthenbuilduponperception.Inordertoenlargeuponthemethod,wenowhavetosetoutcertaindetailsthatPeircegivesinotherwritingsthathedevotedtologicandhistheoryofinference.Peirceclassifiesinferencesasdeductive/analytic/explicativeandassynthetic/ampliative.Andthesynthetic/ampliativehesubdividesinto(1)abduction,hypothesis,orretroduction(thesetermsareclosetosynonymousinPeirce)and(2)induction.Letusbeginwithinduction:Inductioniswherewegeneralizefromanumberofcasesofwhichsomethingistrue,andinferthatthesamethingistrueofawholeclass.Or,wherewefindacertainthingtobetrueofacertainproportionofcasesandinferthatitistrueofthesameproportionofthewholeclass.(W3,326,“Deduction,InductionandHypothesis,”1878)Hypothesis,ontheotherhand,iswherewefindsomeverycuriouscircumstances,whichwouldbeex-plainedbythesuppositionthatitwasacaseofacertaingeneralrule,andthereuponadoptthatsupposition.Or,wherewefindthatincertainrespectstwoobjectshaveastrongresemblance,andinferthattheyresembleoneanotherstronglyinotherrespects.(W3,326,1878)OrasPeircedescribesabductivethoughtelsewhere:Thefirststartingofahypothesisandtheentertainingofit,whetherasasimpleinterrogationorwithanydegreeofconfidence,isaninferentialstepCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth103whichIproposetocallabduction.Thiswillincludeapreferenceforanyonehypothesisoverotherswhichwouldequallyexplainthefacts,aslongasthispreferenceisnotbaseduponanypreviousknowledgebearinguponthetruthofthehypotheses,noronanytestingofanyofthehypotheses,afterhavingadmittedthemonprobation.Icallallsuchinferencebythepeculiarname,abduction...(CP6.525“HumeonMiracles,”1901)HeretherestrictionsweseePeircestarttodraftmayneedverycarefulstatement.Forweshallalsoneedtopreventthisformofinferencefromallowingintotheplaceofahypothesis–intotheplacemarkedby“A”inournextcitation–suppositionsthatarecontrarytothingsinthereasoner’sevidentialbackgroundorthataregratuitousrelativetothatbackground.Thethoughtthatsomesuchpreclusionisneededbecomesevenmoreevidentwhenabductionissetoutasstarklyasitishere:Thehypothesiscannotbeadmitted,evenasanhypothesis,unlessitbesup-posedthatitwouldaccountforthefactsorsomeofthem.Theformofinference,therefore,isthis:Thesurprisingfact,C,isobserved;ButifAweretrue,Cwouldbeamatterofcourse;Hence,thereisreasontosuspectthatAistrue.Thus,Acannotbeabductively...conjectureduntilitsentirecontentisal-readypresentinthepremise,‘IfAweretrue,Cwouldbeamatterofcourse.’(CP5.189,1903)Whentheformofthisreasoningissetoutinthisway,thequestionthattakesshapeiswhether(subjecttotherestrictionsPeircegivesin6.525,cited)justanysupposition,anysuppositionatallwhichwouldmake“C”amatterofcourse,shouldbepermittedtocountasahypothesis,andassomethingreadytomoveuptothenextstageofbeingsubjectedtoconfirmation/disconfirmation.9Musttherenotbecriteriafortheinterrogationandselectionofthingsthatshallcountashypotheses?10Andwheredotheyspringfrom?Dotheyentirelyspringfromtheneedtostabilizebeliefonbelief’sownterms,etc.?Howmuchdoesithelptoreflectthat,inhypothesizing,“mandivinessomethingofthesecretprinciplesoftheuniverse,becausehismindhasdevelopedasapartoftheuniverseandundertheinfluenceofthesesamesecretprinciples”?Ishallnotanswerthesequestionshere.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n104davidwigginsAccordingtoPeirce’sdoctrine,retroductionorabduction(how-everweenlargeuponit)isadistinctivemodeofthinking.Itisre-ducibleneithertodeduction,whoseroleistheancillaryoneofdraw-ingouttheconsequencesofhypotheses,noryettoinduction,towhichPeirceassignsthespecialroleoftesting(refutingorsupport-ing)thehypothesesthataresubmittedtoitbyabduction.Inductionitself,asPeirceseesit(andnotethatPeircedoesnotdenythatthereisanysuchthingasreasonableinduction),cansupportgeneraliza-tionsbut,paceNicod,itdoesnotlicenseus,inorofitself,togofrompositiveinstancesofanarbitraryputativegeneralizationtowardtheassertionofthatgeneralization.Beforethatcanhappen,thegener-alizationhastoenjoythestatusofahypothesis.Itcanonlyattainthatstatusif,intherightway,itrenderslesssurprisingsomethingelsethathasseemedsurprisingorwantedexplaining.FromthisitfollowsthatnomethodologicalparadoxessuchasHempel’s(oftheravens,etc.)orGoodman’s(of“grue,”etc.)cangainanypurchaseonthePeirceanaccountofinquiry.ForthereisnothinginthataccountthatcorrespondstoNicod’spostulate.Ifawhiteshoereallydidcon-firmtosomedegreethat“allnonblackthingsarenonravens”–thiswouldbetheeffectofNicod’spostulate–thenitwouldhavetocon-firmtothesamedegreeitscontrapositiveequivalent“allravensareblack.”Andthat,inanynormalsetup,isabsurd.Hereathousandconfirmatoryinstancesareworthless.InPeirce’sconceptionofin-quiry,appostioningworkinthewayitdoesbetweeninductionandabduction,thereisnoplaceforNicod’spostulate.11VIIIHowdoesaputativesubjectmatterneedtobeifPeirceistoallowthatitconstitutesaproperfieldofgenuineinquiry?Asimilar(orequiv-alent?)question:towhatstandardmustaputativesubjectmatterattain,andwhatmustbeitscondition,forthejudgmentsitthrowsuptocountasproperlyanswerabletoPeirceanrealities/Reals?Supposethereisamodeofthinking,neitherpurelyperceptualnorrelatingonlytowhatisremembered,thatiswellenoughmanage-ableforthefollowingtohold:ifyouengageintheformofthinkinginquestion,then“secondnessjabsyouperpetuallyintheribs”(CP6.95,1903).Supposethat,practicingthismodeofthinking,youcanreachbypatientlaboracomplexandmany-layeredstateofreadinessCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth105andthenarrive,whenjabbed,atabelief.Thisistosaythat,atsomecrucialpointinyourthoughtsorexplorations,somethingthatisnotuptoyoubutisoftherightsorttodothiscanbringitaboutthatyouareconvinced,falliblybutfully.Supposethatinthisfieldyoucanarriveatabelief(asLeibnizwouldsay)malgrevous´.Then,whateverthedistanceatwhichthisformofthinkingliesfromtheperceptualcaseorthememorycase,yoursearchcannothelpbutrepresentagen-uineformofinquiry–aformwithinwhichthejudgmentthatyouarriveatcanbeanswerableforitscorrectnesstosomereality/Real.Orsoitseems.IfPeirce’saccountsofthefourthmethodandofsec-ondnesshaveanygeneralityatall,thentheonlydoubttherecanbeconcerningwhetherthereisanysuchreality/RealisadoubtrelatingtothecredentialsthemselvesoftheformofthinkingthatpurportstoinvoketheRealinquestion.SuchisthedistancethatitappearsonecanputbetweenPeirce’stheoryofinquiryandanyuniformlycausalpicture.SupportmightalsobemusteredherefromPeirce’sphilosophyofmathematics,wherePeircedescribesthesortofsecondnessthatcanarisefromexperimentingbypencilandpaperwitharepresentativediagram,runningthroughallpossiblecasesandfinding(say)thatsomeappar-entpluralityofalternativesreducestoonecase.See,forinstance,CP4.530,19053.516,1896.Nevertheless,itmaybesaid,thereisanob-jection.ConsiderPeirce’sownphrase“determinedbycircumstancesnotextraneoustothefacts.”Whatcanthesewordsmean,itwillbeasked,unlessRealsareitemswithadistinctivelycausalrole?Ifthisobjectionisright,theneitherwemustabandoneverykindofthinkingthattrespassesoutsidetheparadigmfurnishedbythecausaltheoriesofmemoryandperception(asarithmeticalthinkingsurelydoes)orelsewemusttrytounpackthephrase“determinedbycircumstancesnotextraneoustothefacts.”Thesecondresponseseemsmorepromising.Norarewethefirsttothinkthis.Inthecourseofoneofhisrereadingsof“Fixation,”Peircemadeanan-notationagainstthewords(alreadyquotedinSectionIV)“Tosat-isfyourdoubts,itisnecessarythatamethodbefoundbywhichourbeliefsmaybecausedbynothinghuman,but....bysomethinguponwhichourthinkinghasnoeffect.”Peirce’sannotationrequirestheword“caused”tobereplacedbytheword“determined.”12Itsuggeststhathewantedtoconstrue“[beliefsoropinions]determinedbycircumstancesnotextraneoustothefacts”inawaythatallowedCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n106davidwigginsbutdidnotrequiresuchdeterminationtobesimplecausaldetermi-nation.InordertosetoutsomeoftheoptionsthatthiscreatesforthedifferentkindsofcasethatPeirceneedstoaccommodatehere(theyarefartoomanyforcomfort,butletusseewhetheranythingatallcanbesaidatthislevelofgenerality),wemustbeginwithaconcessiontocausality.Opinionsarisefromthoughtsandthoughtsareproducedbyearlierthoughts.“Ifwemountthestreamofthoughtinsteadofdescendingit,weseeeachthoughtcausedbyapreviousthought”(W3,34,1872).Takingourcuefromthisdictumandtracingthesequencefromlatertoearlier,letusacceptthattheopinionorconvictionthatathinkerreachesattheendwillbetheproductoreffectofsomesecondnessexperience(asonemightsay).Letusallowtoothat,attheearlierpoint,thesecondnessexperienceitselfmustbetracedbacktoitsproperancestryinsomerealitythatitpresents.Theseancestrieswillcomeindifferentvarieties,however.Inanordinarycausalcase,thereisacausal-cum-perceptualtrans-actionbetween(say)theCathedralatChartresandaconscious,prop-erlyrecipientsubjectS;andthen,onthestrengthofthisevent,Sbe-lievesjustifiablyandcorrectlythattheCathedralatChartreshastwospires.HereitisbyvirtueofthecausalperceptualtransactionthattheRealconsistinginthecathedral’shavingtwospiresdeterminesS’sbeliefthatthecathedralhastwospires.(Inafullertreatmentonewouldattendseparatelyofcoursetothecasesofseeingxandofseeingthatxisϕ.)Thatisthefamiliarcase.Butnowsupposethattheinitiatorofbeliefwasnotperceptionbutsome“elaborativeprocessofthought”(W3,42),oneleadingintoagradualaccumulationofreasonsthatculminatedatthemomentofsecondnessinthethinker’sfindingnothingelsetothinkbutthat(...).Herethethingwhichbroughtthethinkertothepointofconvictionwasnotjustanycausaleffective-ness.Stilllesswasthethinker’sfindingthattherewasnothingelsetothinkbutthat(...)theoutcomeofsomereality’s/Real’scausallyeffectiveagency.13Ratherthereasonwhythethinkerwasunabletofindanythingelsetothinkbutthat(...)wasthatthereisnothingelsetothink.Ifthereisnothingelsetothink,nowonderthethinkerthoughtthat!Youcansay,ifyouwish,thatsomereasonablebeing’sfindinghimselfunabletodiscoveranythingelsetothinkcausallyexplainshisfinallyarrivingattheopinionthat(...).ButattheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth107temporallyfirstlinkinthechain,thereasonablenessofthethinkerandthereasonedcharacterofhisthoughtisessentialtotheexplana-tion.ItisinthisessentiallynormativewaythatwesatisfyPeirce’sre-quirementthattheinquirer’sopinionthat(...)shouldbedeterminedbyacircumstancenotextraneoustothefacts.Itissatisfiedbecausethecircumstanceoftherebeingnothingelsetothinkbutthat(...)isnotsomethingextraneoustothefacts.Rather,thiscircumstancebearsa(sotospeak)constitutiverelationtothereality/Realthatconsistsinthefactthat(...).14IXIhopethattheproposaljustofferedisinthespiritofPeirce’sanno-tationandcorrection.Itshowshowthepurelycausalcaseneedonlybeoneamongmanyothers.Elsewhere,Ihavetriedtoillustratetheformalpatterngivenintheprevioussection.Ishallgiveagainheretwoexamples,doubtfulthoughitisthatPeircewouldhaveapprovedofthesecond.(A)Peterbelievesthat7+5=12.HehaslearnedthisneitherbyrotenoryetbyreadingthatfamouspassageofKant’sCritiqueofPureReasonwhere7+5=12servesasanexample.Whythendoeshebe-lieveit?Well,theexplanationbeginswiththefactthatalltheotheranswerstothequestion“whatisthesumof7and5?”areblockedorexcluded.Inafullversionoftheexplanation,thisexclusioncouldbeprovedbyreferencetothecalculatingrules.Inamaximalver-sion,onewouldalsorehearsetheirresolubledifficultiesattachingtoproposalsfordifferentrules.Oncesomuchwassetdown,theex-planationmightcontinueasfollows.Peterknowsthosecalculatingrules.Moreover,inespousingtheanswer12,Peterisgoingbytherules.Sonowonderitishisopinionthat7+5=12.So,inthiscase,Peter’sbeliefthat7+5=12isdetermined(asPeircerequires)byacircumstancenotextraneoustothefactthat7+5=12.Hisreasoningsummarilyrecapitulatestheveryreasonwhysevenplusfiveistwelve.IndeedthefullexplanationofPeter’sbeliefpreciselyvindicatesPeter’sbelief.(B)Paulbelieves,letussuppose,thatslaveryisunjustandinsup-portable.Supposethat,inseekingtoexplainwhyPaulbelievesthis,weinquireintohisreasonsforthinkingthis,andsupposewethenlookforfurtheramplificationsandelucidationsofthosereasons,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n108davidwigginsdrawingonthewholeethicalbackgroundthatwesharewithPaul.Thiswilltakealongtime,butsupposethat,asweproceed,itappearsmoreandmoreclearlythattheonlywaytothinkanythingatvari-ancewiththeinsupportabilityandinjusticeofslaveryistooptoutaltogetherfromanymoralviewpointthatcanmakesenseofaskingthequestion“Whatisonetothinkofthesupportabilityorjusticeofslavery?”Forsupposethatatsomepoint,inheapingconsiderationontoconsideration,wefindwehaveenoughanditbecomesapparentthatthereissimplynoroominwhichtoformanotheropinion.Nodoubttherewillbemanyethicalcaseswherewedonotreachthispointandwedonotknowhowtocloseoffeveryavenue.But,inthecasewherewereallycanseePaul’sbeliefasdownwindofreasonsliketheconvincingonesthatwehaveimaginedsomeone’seventuallyre-hearsingaboutslavery,surelywecansay“NowonderPaulbelieveswhathebelieves!Thereisnothingelsetothink.”Inotherwords,Paul’sbeliefaboutslaveryisdeterminedbycircumstances(namelytheconsiderationsthatwearesupposingtohavebeenrehearsedandtoimpingeononewhounderstandsthemoralquestion)notextra-neoustothefactthatslaveryisunjustandinsupportable.ForPaul’sreasonsforthinkingwhathethinksdosummarilyrecapitulatethatinvirtueofwhichslaveryiswrongandinsupportable.Peircewouldhavebeensceptical,Ifear,whetherourexample(B)couldbeworkedoutinthewayIhaveimagined.Hecouldnotob-jectinprinciple,though,totheideathatamassofconsiderationscanculminateinconviction.Forhespeaksinotherconnectionsofreasons“notform[ing]achainwhichisnostrongerthanitsweak-estlinkbutacablewhosefibersmaybeeversoslenderprovidedtheyaresufficientlynumerousandintimatelyconnected”(W2,213,1868).XItisevident–andasearchformoreexampleswouldmakeitevenmoreevident–thatthegeneralitytowhichPeirceaspiresinhistheoryofinquiryinvolvesusinabewilderingandindefinitevarietyofdifferentwaysinwhichthinkersindifferentareasofconcerncansatisfythePeirceanrequirementonwhichwehavelaidsuchstress.Itisnolessevident,though,thatinsofaraswewanttopersistatthatlevelofgenerality,theanswerwegivetothequestionproposedCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth109atthebeginningofSectionVIII,willhavetobeasfollows:thethingthatisminimallyrequiredinordertosecurepragmaticcontenttoasubjectmatteristhis:thatthere,inthatsubjectmatter,abelieftotheeffectthatpcanbedeterminedbycircumstancesnotextraneoustothefactthatp.Moregenerally,theconclusiontowhichwearedrawnisthatforanygenuinebelief,whethertrueorfalse,therehastobesomethingitisanswerabletoandsensitiveto.ThislastmayaswellbecalledaReal.ButinsteadofrushingintoanewontologyofReals,letuslookcarefullytothestatusofourfamiliarform:whoeversincerelyinquireswhetherpseekstoensurethatanybeliefoftheirstotheeffectthatpbedeterminedbycircumstancesnotextraneoustothefactthatp.Thisisonlyaschema.Realsarenothereobjectsquanti-fiedover.Thesentenceletter“p,”beingnotavariable,functionsbyholdingaplaceforasentenceinuse.Ontheseterms,theminimalclaimabouttheformationofthebeliefthatpisanotionallysimulta-neousassertionofallinstancesoftheitalicizedsentenceformwithallpossiblesententialfillingsfortheletter“p.”Inputtingforwardthisschema,wegesture(ifyoulike)atsomethingentirelygeneral,somethingthatwouldbegesturedatbythesecountlessassertions.Butweonlygesture.For,strictlyspeaking,thereisonlyapatternhere,nothingmore.Ifwedoproceedinthisway,though,thereisan-otheradvantage.Wecanalsogivenoticethatnotalltheseassertionsworkinthesameway.Thesecondnessrequirement,thenonextrane-ousnesscondition,andtheotherrequirementsonthedeterminationoftheinquirer’sbeliefaretobeunderstoodinthediverswaysthatareappropriatetodifferentexamples.IfwhatPeircesaysaboutRealsisinterpretedorelucidated,logicallyspeaking,inthisway,thenthephilosophicaleffectisthattheschemaisgrammaticallyandphilo-sophicallyfilledoutfordifferentkindsofcasesaccordingtothesub-jectmatter–andinthelightofwhateverPeirce’slogiccanaddtohischaracterizationofthefourthmethod.XIIfthefourthmethodasnowexplainedandenlargeduponistheonlymethodofsatisfactorilysettlingopinion(albeitfallibly,alwaysfal-libly),whatconceptionoftruthdotheoristsofinquiryhavetoseeasanimatingandconstrainingtheepistemiceffortsofthosewhoCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n110davidwigginspracticethemethod?Andhowaretheoristsfurthertoelaborateorelucidatethisconception?Letusbeginwithsomeofthematerialsof“Fixation”itself.InafootnotetoapassagethatIquotedinSectionV,continuingthatpassageintoanafterthoughtdated1903,Peircewrites:1.CP5.375:[T]ruthisneithermorenorlessthanthatcharacterofapropositionwhichconsistsinthis,thatbeliefinthepropositionwould,withsufficientexperienceandreflection,leadustosuchconductaswouldtendtosatisfythedesiresweshouldthenhave.Tosaythattruthmeansmorethanthisistosaythatithasnomeaningatall.Thisisthekindofstatementthathasgivenpragmatismsuchabadname.Apart,however,fromthewarningsalreadyurgedinsectionI,thereareotherreasonstobecarefulhere.IfIsaythatthecharacterofbeingredisnothingmorenorlessthanthecharacterofbeingthecolorthoughtbyblindpeopletobewellgraspedbyacomparisonwiththesoundofatrumpet,doesmyclaimhavetobeinterpretedasadefinition?If,inaddition,wetakethecharacterizationinpassage(1)tobeadefinitionoftruththatisintendedtobringoutwhatissogoodabouttruth,thenwemisunderstandPeirceevenmoreunfairly.Notonlydoweattributetohimacynicalinstrumentalismthatisutterlyalientohisactions,hischaracter,andhisexpressedviewsofscienceandlifeitself.Weinterpolateintohistheoriessomethingthatisentirelyalientothelatersectionsof“Fixation.”Itistruethat,inotherplaces,Peircegivesapragmaticreinterpretationofthenotionsof“reality”/“Real”/“externalpermanency”thatplaysuchanimportantrolein“Fixation.”Butthatreinterpretationispreciselynotintendedtoblunttheforceofthelatersectionsof“Fixation.”Theintention(whethersuccessfulorunsuccessful)israthertoexplicatethesenotions–inthespiritof“look[ing]totheupshotofourconceptsinorderrightlytoapprehendthem”(CP5.3,1901).WeneedmorePeirceantestimony,testimonybeyondthatalreadydisplayedinSectionIV,abouttheideaoftruth:152.CP2.135,1902:YoucertainlyopinethatthereissuchathingasTruth.Otherwisereasoningandthoughtwouldbewithoutapurpose.WhatdoyoumeanbytherebeingsuchathingasTruth?YoumeanthatsomethingisSO...whetheryou,orI,oranybodythinksitissoornot....TheessenceoftheopinionisthatthereissomethingthatisSO,nomatteriftherebeanoverwhelmingvoteagainstit.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth1113.CP5.553,1905:Thattruthisthecorrespondenceofarepresentationwithitsobjectis,asKantsays,merelyanominaldefinitionofit.Truthbelongsexclusivelytopropositions...thepropositionisasign...thoughtisofthenatureofasign.Inthatcasethen,ifwecanfindouttherightmethodofthinkingandcanfollowitout–therightmethodoftransformingsigns–thentruthcanbenothingmorenorlessthanthelastresulttowhichthefollowingoutofthismethodwouldultimatelycarryus.Inthatcase,thattowhichtherepresentationshouldconform,isitselfsomethinginthenatureofarepresentation,orsign–somethingnoumenal,intelligible,conceivable,andutterlyunlikeathing-in-itself.CP554:Truthistheconformityofarepresentamentoitsobject,itsobject,ITSobject,mindyou....Hereisaviewofthewriter’shouse:whatmakesthathousetobetheobjectoftheview?Surelynotthesimilarityofappearance.Therearetenthousandothersinthecountryjustlikeit.No,butthephotographersetupthefilminsuchawaythataccordingtothelawsofoptics,thefilmwasforcedtoreceiveanimageofthishouse....So,then,asign,inordertofulfilitsoffice,toactualizeitspotency,mustbecompelledbyitsobject.Thisisevidentlythereasonofthedichotomyofthetrueandthefalse.Forittakestwotomakeaquarrel,andacompulsioninvolvesaslargeadoseofquarrelasisrequisitetomakeitquiteimpossiblethatthereshouldbecompulsionwithoutresistance.4.CP5.565,1901,“TruthandFalsityandError“:Truthisthatconcor-danceofanabstractstatementwiththeideallimittowardswhichendlessinvestigationwouldtendtobringscientificbelief....Realityisthatmodeofbeingbyvirtueofwhichtherealthingisasitis,irrespectivelyofwhatanymindoranydefinitecollectionofmindsmayrepresentittobe.ThetruthofthepropositionthatCaesarcrossedtheRubiconconsistsinthefactthatthefurtherwepushourarchaeologicalandotherstudies,themorestronglywillthatconclusionforceitselfonourmindsforever–orwoulddoso,ifstudyweretogoonforever.Anidealistmetaphysicianmayholdthatthereinalsoliesthewholerealitybehindtheproposition;forthoughmenmayforatimepersuadethemselvesthatCaesardidnotcrosstheRubicon,andmaycontrivetorenderthisbeliefuniversalforanynumberofgenerations,yetul-timatelyresearch–ifitbepersistedin–mustbringbackthecontrarybelief.Butinholdingthatdoctrine,theidealistnecessarilydrawsthedistinctionbetweentruthandreality.5.CP5.416,1905:[Atruthis]thattoabeliefinwhichbeliefwouldtendifitweretotendindefinitelytoabsolutefixity...Letusbeginwith(2).(2),likethebeginningof(3),effectivelyreinforcesthemanifestpurportofPeirce’srationaleforthefourthmethod.Itreinforcestheconcludingmessageof“Fixation”butdoesnotcarryusbeyond.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n112davidwigginsIn(3),theobjectofarepresentamenmustsurelybetheverysamethingasitsReal.Theanalogywithonevieworaspect,asgiveninthephotographofPeirce’sownhouse,echoesasentence(CP5.549,1905)wherePeirceclaimsthatafactissomething“sohighlyprescissive16thatitcanbewhollyrepresentedasasimpleproposition.”But,how-everstrikingtheanalogymayappearthatwefindin(3),andhoweverhelpfulthismaypromisetobe,theontologythatitimportsisfullofdifficulty.Theproblemisfamiliar.Wethinkapropositionshouldbetruebysatisfyingitstruth-condition.But,ifwethinkthat,thenweneedtobeabletostatethetruth-conditioninsuchawaythatthepropositioncanbefalseaswellastrue.Peirceclaimsthat“apropo-sitionistrueifitconformstoitsobject,”andheindicatesinthepassagecitedin(3)(whichIhaveabbreviated)thathereheisreach-ingbeyondthecorrespondencetheory.Butthenwemustaskwhattosay,accordingtotheaccountheisdeveloping,ifthepropositionisfalse.ForinthatcasetherewillbenosuchthingasITSobject.17Therewillbenothingthepropositionhasfailedtobe“compelledby.”Ontheotherhand,ifthepropositionistrue,itsobjectwillex-istandtherewillbenoneedtoenterfurtherintoquestionsaboutconformityorcompulsion.Foritsobjectexistsandthatalonewillbeenough.“Correspondstothefacts,”asthecorrespondencetheoryhasit,where“facts”isplural,appearstobeamerelystylisticvarianton“true.”Understandably,thisprovokesustotrytofindaproperre-lationherebetweenapropositionandsomeonething,itsrepresen-tatum,sotosay(CP5.384).Buttheauguriesarenotgood,ithasalreadyappeared,forthetheorywhichweseePeircepickinghiswaytowardinourcitation(3).Thecounterattractionswillbefurtherev-identoftheapproachweadoptedinSectionX.Onthatapproachtheschematicletter,bystandinghostageineachcaseforasentencethatisreality-involving,doesasortofjusticetotherealismofPeirce’sviewofthesearchfortruth.Butitdoesthiswithoutontologicalcom-mitmenttofacts,realities,orRealsanditdispensesentirelywithallrelationsofconformityandcompulsionbetweenabeliefandaparticularitem,whetherfactorobject.ThismayseemtosuggestthepossibilityofreplacingPeirce’seffortinCP5.554(citedearlier)withsomeschemefortruththatismoreanodyne,lesstroublesome,andentirelygeneral.ButIhavetriedtoshowinanotherplacethat,heretoo,nogeneralaccountoftruthitselfistobehad–atbestCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth113athoughtwhich,consistentlyandsincerelypursued,leadsstraightintoparadox.18Theproposals(4)and(5)bringusclosertotheformulationfrom“HowtoMakeOurIdeasClear”whichwerejectedinSectionI.Proposals(4)and(5)areintendedtogivetheeffectiveorpragmaticmeaningofthemanifestlycorrect“merelynominal”conceptionsexpressedinproposals(2)and(3).In(4),suchpragmaticproposals,eventheidealistones,aredefendedfromthechargeoflosingthedis-tinctionbetweentruthandreality.Sofar,sogood;andnodoubttheseproposalsalsohaveothermerits.Nevertheless,whenreadliterally,theyallseemtodependfortheiracceptabilityuponthesuppositionthatnoinformationofthekindthatwouldbeneededtotestplau-sibleguessesalreadymadeordiscovertruthsasyetunknown(e.g.,concerningthatwhichispastorispresentlyhidden)everperishesorbecomesunavailabletoinquiry.Forif,alwaysandconstantly,suchinformationisbeinglost,thenitisneitherherenortherethatinquirycanbeconstantlyrenewed,constantlycorrected,andopen-endedlyprolonged.Moreover,suchperishing,asHilaryPutnampointsout,isnotonlyafactbutafactthatisimpliedbymodernphysics.19Peirceansmayrespondtothiscripplingobjectionbyreadingpro-posals(4)or(5)lessliterally.Butthentheconstrualwillneedtoleanheavilyonourunderstandingofthatwhich(4)and(5)purporttodefine/explain/elucidate.Itisalsoworthremarkingthat,oncethereferenceof“thattowhichinquirywouldtend,”etc.,issufficientlycarefullydistinguishedfromanyparticularsetofpropositionsthathasbeenredactedorwillhavebeenredactedatanyparticularpointinthefuture,thephrase“thattowhichinquirywouldtend,”sofarfromdistillinganeffectiveorpragmaticmeaningfromthetruismsthatfigurein(2),isaformofwordsthatstandsinradicalneed(asradicalaneedasanyexpressionevercould)ofpragmaticelucidation!XIIIsallthenentirelylostsofarastruthisconcerned?Canitbethattruthwaitsinthewings,islatentintheinquirer’sprojectofdecid-ingwhattothink(seeagainourfirstepigraph),canbeclearlyseenemerginginthethoughtsofsomeonewhomovesthroughthefirst,second,andthirdmethodsintothemethodofexperience,abduction,andtherest–andyetisacharacterthatdefiesallidentificationorCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n114davidwigginselucidation?No.SurelywecanfindforPeircesomeformofwordsthatfastensdownandpromisesinduecoursetohelpelucidate,intermsthatessentiallyinvolvethebusinessofinquiryandthemethodof“experience,”thenatureofthatproperty,namelytruth,which(un-lesswearecompletestrangerstoopinionordoubt)isalreadyfamiliartoanyorallofus.Onceweallowourselvestospeakofapropertythatisalreadyknowntous,andoncewedissociateourselvesfromPeirce’snumerousanduniformlyunsuccessfulattemptstoarriveatthepropertyfrompragmatistwould-bedeterminationsoftheex-tensionof“true,”severalsuitableformsofwordsstareusintheface:6.Truthisthepropertythatitistheaimofinquiryassuchtofindbeliefspossessedof.7.Truthisthecharacterwhich,ifonlywefollowthefourthmethodofinquiry,wemayjustifiablyhopewillbeenjoyedbybeliefsthatsurvivehoweverlongorfarinquiryispursuedorprolonged.8.Truthisthepropertythatanyonewillwantforhisorherbeliefswhosincerelyinquireswhetherp(ornot)andwhoseekstoensurethatanybe-liefofhisorherstotheeffectthatp(ornotp)shouldbedeterminedbycircumstancesnotextraneoustothefactthatp.SuchformulationsmightnothavepleasedPeirce,buttheyholdaplaceforaviewtowhichhecouldlayclaimifhewanted.XIIISupposethat,inthecauseoffurtherelucidating20thepropertyoftruth,weweretodeploytheidentitiesgivenin(6),(7),and(8)andweweretoelaboratethepluralityoflinkagesholdingbetweentruth,ontheoneside,andinquiry,experience,secondness,hypothesis,...,ontheother.Supposethat,proceedinginthisway,weweretopresentourfindingsasthemarks,inFrege’ssense,21oftheconcepttrue,andsupposethat,inthesameeffort,wetriedtoexplorethelogi-calpropertiesoftheconceptoftruth(ascertainwhatpropertiesthepropertyoftruthimplies,excludes,etc.inathingthoughtorsaid).Thenwhatwouldfollowfromthefactthewholebasisonwhichthiselucidatoryexercisewasconductedwasalinkbetweenano-tionoftruthawaitingfurtherspecificationandthenotionofinquirythatisalreadydeveloped(cp.IV,V,VI)andispartiallydefinitiveofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth115pragmaticismasaphilosophicalposition?If,proceedinginthewayindicatedandadducingourunderstandingofinquiry,welookinthisspiritofpragmaticism“totheupshotofourconception[oftruth]inorderrightlytoapprehend[it],”thenwhatdowelearnaboutthatconception?Willourfindingshavetheeffectofsubvertingtheordi-nary(“realist”)presumptionthatthetruthisperfectlyindependentofus(except,ofcourse,insofarassomejudgmentthatisinques-tionrelatestodoingsofours,orrelatestotheeffectsofsuchdoings)?Willthepragmaticistoutlookhavetheeffectofunderminingtheor-dinaryidea–comparecitation(2)–thatthetruthis“thereanyway,”definitelyanddeterminately?WillitmoveustowardthepositionthatMichaelDummetthascalledantirealism?22ItishardtofindverymuchinPeirce’stextsthatconformstotheseexpectations.Noristhereanyanticipationinanylogicalwrit-ingsbyPeirceoftheclassicantirealistpositiondevelopedbyMichaelDummettundertheinfluenceofmathematicalintuitionismandthestrongemphasisthatintuitionismplacesontherelationbetweengraspingaproofofapropositionandunderstandingit.(Theintu-itionistemphasisisphilosophicallyakintothelogicalpositivists’foundationalideathattounderstandaSatzistoknowthemethodofitsverification.)TheantirealistwhosepositionDummettdevelopsisonewhoaffirmsthelawsofnoncontradiction(nostatementistrueandfalse)andoftertiumnondatur(nostatementisneithertruenorfalse)whilewithholdingassentfromtheprincipleofbivalence(everystatementiseithertrueorfalse).Suchassentiswithheldbyvirtueoftheabsenceofanyassurancethat,withregardtoeverywell-formedassertion,eitheritoritsnegationcanbeprovedorestablishedtobetrue.(Dummettpointsoutthat,forthesamesortofreason,thepositivistswouldhavebeenwelladvisedtowithholdthatassent.)ThingsseemverydifferentwithPeirce.Inallhislogicalexplo-rations,heneverraisesdoubtsorquestionsofprincipleaboutthestatusofthelawofdoublenegationelimination.Sincedoublenega-tioneliminationelidesthesubtledifferencebetweentertiumnondaturandbivalence,committinganyonewhoacceptstheformertothelatter,itwouldappearthatPeircecanhavehadnopremonitionatallofanantirealismsuchasDummett’s.Itistruethat,inamorephilosophicalcontext,Peircewrites(usingthenameexcludedmid-dlewheremodernantirealistsmightpreferbivalence)that“Logicrequiresus,withreferencetoEACHquestionwehaveinhand,toCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n116davidwigginshopesomedefinitiveanswertoitmaybetrue.ThatHOPEwithreferencetoEACHcaseasitcomesupis,byaSALTUS,statedbylo-giciansasalawconcerningALLCASES,namelythelawofexcludedmiddle.Thislawamountstosayingthattheinversehasaperfectreality.”23ButthehopePeircespeaksofhererelatestotruth,nottoprooforverification;andthesubstanceofthehopesurelyrelatestotruthasordinarilyconceived.For,asIshalltrytoshowinthenextpages,theconfidencethatPeircespeaksofaspresupposedbythelogicalprincipleinquestionsmacksmoreofPeirce’sconfidenceinthesignificanceorSinnofdeclarativesentencesthatareproperlyanswerabletoexperienceorexperimentthanitcanofanyfaithofPeirce’sindeclarativesentencesortheirnegationsallhavingproofsorverifications.AtleastinDummett’ssense,Peirceisnotanantire-alist.InPeirce,thekeytohavingsignificanceorsense(andtograsp-ingsignificanceorsense)isnottheactualprospectofproof/disprooforverification/falsificationbutproperengagementwiththebusinessofinquiryandofreachingforverificationorfalsification.If(asIventuretothink)pragmaticismleavestruthjustasitwas,whatthenistherealpurportofpragmaticisminitsconnectionwithtruthandmeaning?Whatistheintendedimportofsuchdictaasthese?–Thereisnoconceptionsoloftyandelevatedthatitcannotbefullydefinedintermsoftheconceptionsofourhomely,instinctiveeverydaylife(MS313p.29,quotedinMisak1991:119)orthefamiliarfoundationalclaimConsiderwhateffectsthatmightconceivablyhavepracticalbearingsweconceivetheobjectofourconceptiontohave.Thenourconceptionoftheseeffectsisthewholeofourconceptionoftheobject.24(CP5.402,“HowToMakeOurIdeasClear,”1877)AccordingtothetranspositionofPeirce’sthoughtsthatIofferinthesereflections,therealpurportisrelativelysimple.EventhoughPeirceisarealistabouttruth,heisanoperationalistaboutmeaning.Thereisnospecificallypragmaticistconceptionoftruth,25butthereisapragmaticistconceptionofsense/significance/Sinn.APeirceanpragmatist,apragmaticist(asPeircewasledtosayinordertomakeroomforthedifferencesbetweenWilliamJamesandhimself),willscarcelythinkitworthsayingthatthereismoretorealitythancouldCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth117everbeputintohowevermanypropositions–that,truthbeingwhatitis,thereareallsortsoftruthsweshallneverformulateandnevercould.Hiskeyconcerniswiththepropositionsweshallourselvesarriveat,express,affirm,orbelieve.Itiswiththerealpurportormeaningofouractualutterances,andtheillusionswesoeasilyfallintoaboutwhatwecanmeanbywhatwesay.Thepragmatist’schiefcontributiontothesequestions,andthesourceofhiscritiqueof“vagabondideasthattrampthepublicroadswithoutanyhumanhabitation”(CP8.112,1900),lieswherethecitationlastgivenindi-catesthatitdoes.ItliesinPeirce’saccountofthegradesofclaritythatcanbeattainedinourunderstandingofthetermsthatenterintomeaningfulsentencesandinourgraspoftheconceptsthatenterintothepropositionsthatsuchsentencesexpress.Whereconcept-termsareconcerned,wehavethefirstgradeofclarity,accordingtoPeirce,ifwecanapplythetermtothingsinourexperience.Wehavethesecondgradewhenwecanproducethekindsofexplanationthatpassmusterasdictionarydefinitionsorthelike.Atthethirdgrade,ifwearetoattainthat,ourrecognitionalcapacitymusthavebeenelaboratedintoafurtherandbetterstateofpracticalreadiness,afullyoperationalstate,sotospeak,onethatengageswithinquiry,experience,secondness,guessing,retroduction...withthesethingsastheyareorcanbeinlife.Seeagaintherecentlydisplayedcita-tions.Thepracticalconceptionsmentionedinthesecondrelatetohabitsofaction.Theyalsoinvolvearatherspecificorientationto-wardpossibleoractualfutureexperience.(CompareCP8.194.)Itisatthethirdgrade(presupposingandnotsupersedingthefirstandthesecond)thatthegraspofsense/significance/Sinnofasymbolhastobemadecomplete.Thethirdgradeofclaritycanonlybeattainedif,independentlyofanyparticularperson’sefforts,thereawaitsonewhoseekstograspthemeaningofagiventermsomepubliclycompletedorcompletablemeaningforasufficientlydeterminedthinkertograsp.PeirceoffersnounitaryorfullanswertothequestionwhatitisthatcompletesthismeaningorsaturatestheSinnofthesymbol.(SeeMisak1991:12–35).Supposing,though,thatsomehowthissaturationhasbeenachievedorisintheprocessofbeingachieved,wemayexpectthepropositionexpressedbyasentencecomprisingsuchsymbolstode-pendontheSinnofitsconstituentparts.Thinkers’correspondinggraspofthepropositionanditstruth-condition,arrivedatthroughCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n118davidwigginstheirunderstandingofasentence’smodeofcomposition,mayormaynotputthemintoapositiontoverifyorfalsifythepropositionexpressedbythesentence.Wherethinkerscanverifyorfalsify,itneedstobenoaccidentthatthatisso.ButbeingplacedtoverifyorfalsifyisnotthegeneralformofthekindofreadinesswhichPeirceisconcernedwith.XIVSuchinoutlineisthesemanticoperationalismwhich,usinglan-guagemorecurrentthanthelanguageofPeirceansemeiotic,Ispec-ulativelyreconstructfortheauthorof“Fixation”atthetimeoftherereadingsandrewritingsrecordedinourfirstepigraph.Leavingin-tacttheordinaryidealoftruth,thepositioncounsels,nodoubt,thatoftenthebestwecanhopetoachieveisapproximationtotruth–andnothingwrongwiththat.If,ontheoneside,thereistruthitself(truthsthemselves),whichpropositionsseektotrackdown,thenontheothersidetherearepropositions,andthesensesofsentencesinuse.Theseareourartifacts.Assuch,theydependonusfortheircom-pletion.Theproperlysignificantsentence,bybeingthesentenceitisandhavingitssensedeterminedbywhateversenseswithwhichwecontriveforthesensesofitsconstituentstobesaturated,setsitselfagoalthatiteitherattainsordoesnotattain.Whetherthisgoalisattainedornotisinnowayuptous.Butwhatsentenceitisthathasbeenpropounded,withwhatsense,andengaginginwhatwaywithinquiry,thisisuptous.Itisahighlynontrivialachievementonourcollectivepartforasentencetosetitselfsuchagoal.Itisanachievementnotcontrivableatallunlesstheconditionsfortrutharecoordinatedwiththedemandsthatareplacedonaninquirertoavailhimselforherself,intherightway,ofexperience.Suchrealismandsuchoperationalismaremadeforoneanother.Onewordmore,aboutbivalence.SupposethatinacontextcthereisasentenceSeachofwhosecomponentshasasensethatislex-icallyandcontextuallyfullydeterminate,operationallycomplete,andreadyandwaitingincforthecomprehensionofanythinkerwhoisreadytoattaintothethirdgradeofclaritywithrespecttoitandreadytograsptheSinnofthesentencethatallthesecompo-nentsmakeup.SupposethatthepropositionexpressedbySincwillbetrueifandonlyif...,where“...”isafullydeterminatecondi-tion(determinateinc,evenifnotnecessarilyverballycompletelyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth119explicit).Ifso,thenthepropositionconveyedbySwillbesodeter-minatethatitisdeterminatewhatitisforthingstobeotherwisethantheyarewhen....Inthatcase,givensuchdeterminacyforS,nothingobstructsthefulldeterminacyofthesenseofthenegationofS.Forwhateverinccircumscribesthesenseof“...”willdeter-minethesenseof“thingsareotherwisethantheyarewhen....”SuchaformofstipulationseemsideallysuitedtotherulePeircegivesatCP4.492forhissystemofexistentialgraphs,totheeffectthattwoSEPS(signsofexclusion),“theoneenclosingtheother,butnothingoutsidethatother,canberemoved.”(SeeCP4.490–8,1903and4.572,1905)Thisisequivalenttothelawofdoublenegation.26Anantirealistmightdemur,butPeircehimselfentertainsnodoubtsatallaboutthepragmaticmeaningofaSEP-signthatconformstosucharule.XVInquiryconductedalongthelinesofPeirce’sfourthmethod,inherit-ingasitdoesthemeritsofvariouspredecessors,isaprocessthatgath-ersrationalstrength,wehaveclaimed,asitgathersforceandgathersforceasitgathersrationalstrength.Ontheproperunderstandingofthisprocess,wehavesaid,truthisconceivedasthepropertythatwecanhopetosteerourenquirytohomeupon;thebeliefsthatin-quiryfurnishestousarebeliefsthatitisrationalforus,howeverfallibly,topersistinuntilspecificgroundsfordoubtpresentthem-selves;andthemethodofinquirymakesroomforanyorallmodesofresearchorcriticism,whethercommonsensicalorscientific,thatpromisetoreachbeliefsbyroutesnotextraneoustothefactstheyareconcernedwith.Onewhoconductshimselfontheseprincipleswillbenomoreeagertodefine“rational”thanheistocircumscribelegitimatemethodsofexplorationanddiscovery;butsuchapersonwillsurelyinsistthatthemethodofinquiryisafullyrationalwayofarguingfromtheknowntotheunknown–thatitisaparagonofrationality.Suchanattitudewillappeartoconflictwithsomethingcom-monlyregardedasoneofthegreatinsightsofDavidHume.HumepointsoutinEnquiryConcerningHumanUnderstandingthatallreasoningsconcerningmattersoffactarefoundedintherelationsofcauseandeffect,andthefoundationofourunderstandingoftheseisexperience.Buthereheclaimstofindaproblem.HowmayICambridgeCompanionsOnline©CambridgeUniversityPress,2006\n120davidwigginsrationallyinferfrompastbread-eatings’havingnourishedmethatsimilareatingswillnourishme?Ifthereissuchaninferenceitisnotintuitive(knowablewithoutdemonstration);noryetisitdemon-strative.Whatisitthen?Itisexperimental,Humeimaginesyoursaying.Buttothisherepliesthatexperimentalreasoningsalreadypresupposethatthefu-turewillresemblethepast.Howthen,presupposingthis,cantheyshoworevensuggestthatthefuturewillresemblethepast?Humeinfersthatitisnotreasoningthatengagesustosupposethefuturewillresemblethepast.Itishabit,notreason.WhatoughtPeircetosay?Peircewouldbeginbyagreeingthatin-ferencefromtheknowntotheunknownisamatterofhabitandisnotdemonstrative.Buthabits,hewouldinsist,canbegoodorbad.Andgoodhabitscanexemplifyadistinctiveformofreasonableness.(SeeSectionII.)Afterall,weneedtoarguefromtheknowntotheun-known.Ifweneedto,thenitisreasonableforustodoso(intuitivelyrationalyoucansay,ifyouwish)anditisirrationalforusnottodoso–providedthatwedonotentrustourselvestoaparticularpolicythatthereisreasonforustoregardasreckless(asexposingourvitalneedstorisksthereisnonecessityforustoincur)orasill-calculatedtobringustobeliefsweshallacceptforreasonsnonextraneoustothefacts.IfHumewantstomakeapointabouthabit,lethimmakeitasapointabouttherelevanceofhabitstothescienceoflogic.Itisagoodpoint,andPeircewouldsecondit.(Cp.SectionII.)Butitisnoexcuseforanassaultonreasonassuch–unlessHume’saimistoputhimselfatthecenterofalong-runningcontroversy.Itiseasytoimaginethat,ifhewereallowedaresponse,HumewouldstillpressuponthequestionhowPeircecanarguenon-question-begginglyfrompastnourishingsbybreadtofuturenourish-ingsbybread,ifthispresupposesthegeneralclaimthatthefuturewillresemblethepast–whichissomethingyethardertoestablishthanfuturenourishingsbybread.TothisPeircewouldsurelyreply(hereanticipatingPopper)thatgoodargumentsfromtheknowntotheunknownhadbetternotpresupposethatthefuturewillresemblethepast.Foritisnoteventruethatitwill!Natureisnotregular....Itistruethatthespeciallawsandregularitiesareinnumerable;butnobodythinksoftheirregularities,whichareinfinitelymorefrequent.(W2,264)CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth121Moreover,whenwearguefrompastnourishingsbybreadtofuturenourishingsbybread,wearenot,accordingtoPeirce,simplyextrap-olatingapastregularity.Thatisnever,inhisview,avalidprocedure.IfthatwaswhatHumewasattacking,thenHumewasright,Peircewillsay,butfarshortoftheconclusionthatHumewasaimingfor.Whenweextrapolatearegularity,therehastobeanotherreasontodosobesidethefactthattheregularityhasheldsofar.Eveninthespecialcaseofthe“particularmethods”belovedoftheinductivere-liabilists,Peircewouldsay,itwouldbeutterlyinvalidtoarguefromthemerepastsuccessofamethodtoitsfuturesuccess.Withanymethod,therehastohavebeensomethingelsetocommendit.Andhereistheroleofabductivethought.(CompareSectionVII.)Letusdistinguishheretwocases.Thefirstisthatoftheordinarypersonwithanordinaryneednottostarve,whowantstoprolonglifeandneedssomedeterminateway,hereandnow,ofsortingthenour-ishingfromthenonnourishing.Anysuchsortingmusteitherdeployexistingcategorizationssuchas“bread”ordeployimprovementsuponthecategorizationsthepersonalreadyhas.Thereisnowhereelseforhimtoworkfrom.Insofaras“bread”isoneofthecatego-rizationsonwhichthepersonhabituallyreliesandonwhichheacts,heiscommittedtothinkthatthereissomethingaboutbread–asub-stancethathecanidentifywherenecessarywithsomeprecautionarycare–whichwouldexplainwhyitnourishes.Underinterrogationhewouldappeartobecommittedtothinkthereissomegeneralizationaboutbreadandnourishing(onehemaynotknowhowondemandtoformulateverycarefullyorarticulately)whichwouldnot,ifitweretested,befalsified.(CompareourdiscussionsofperceptionandmemoryatSectionVI.)Ifthequestionwereraisedwhy,oncefor-mulatedorreformulated,anysuchgeneralizationshouldbereliedupon,thepersonmightreplyfirstthatfaithinthisisamuchmorereasonablefaiththanfaithinthefuture’sresemblingthepast;andsecondthatsomesuchgeneralizationhastoberelieduponiflifeistogoon.Thereisnoalternative.Itwouldbeirrationalthennottoactonthebasisheisactingon.Criticizethatbasisandhewilllookforsomethingbetter,forsomethingthatisadequateforthematterinhand.Buttheonlypointofdepartureinthesearchforsomethingbetteristheplacewhereweare.Cp.PlatoPhaedo101D(adfin.);AristotleNicomacheanEthics1095b2.Asonemakesmoreandmoreexplicitthatwhichanordinaryper-sonmightsayindefenseofhishabitoftakingbreadtobenourishing,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n122davidwigginsoneconvergesonthecasewhereamoretheoreticalansweristobegiven.Thattheoreticalanswerisnot,accordingtothepragmatist,es-sentiallydifferent,onlymorediscursive.Itbeginsinthesameplace.Ifwearetodowhatwearenaturallycommittedtodoandargueinthiscasefromtheknowntotheunknown,thenwemustbeginbytryingtounderstandthethingthatisknown.Sothethingwemustunderstandbetterisbread.Theproblemofunderstandingorsinglingoutthisparticularkindisnot,however,oneweneedtosolveonitsown,orwithoutanyreferencetothestateofourinquiriesintootherempiricalquestions.Wecanonlyapproachitfromwhereweareatanygivenpointinourinquiry.Lookingatthingsfromwhereheis,theinquirernoticestheremarkablephenomenonthatsomehavenothingtoeatandstarveanddiewhileotherswhoeat,andeatbreadamongotherthings,sustaintheirlife.(Seethethirdepigraph.)Ifbreadnourishedthenitwouldbeamatterofcoursethatthosewhoateitsustainedlife.Soitseems,accordingtotheabductivehypoth-esis,thatbreadnourishes.Thisisageneralizationworthytotest;andintheinterimitisonetoliveby,pendinganyrefinementorrefutationthatitmaysuffer.HumeorhisfollowerswillnoticethatthePeirceanstrategyleanshereuponthefundamentalhypothesis.Sotheyareboundtoinquirewhatgroundsthefundamentalhypothesisitself.Onetemptingan-sweris:“Nothingholdsorissoorobtainsbutthatthereissomereasonwhyitisso.”ReadersofLeibnizwillrecognizethethought.27Itistruethattheclaimisquiteasgeneralastheclaimthatthefu-turewillresemblethepast,butitisafarbettercandidatetobetheregulativeassumptionofinquiry.Atleastitsuggestsnothingthatismanifestlyfalse.Stillbetter,itscarcelyneedstobethoughtofasanempiricalgeneralizationaboutreality.Itproposesratheracertainattitudetowardreality–anattitudethatitwouldbeunreasonableforusnottoshareinifwearetodothatwhichweshallperishbynotdoing.WhatthenistheconnectionbetweenSufficientReasonandthetwofoldprocedurethatPeircecommendstous?Supposeourmethod-ologicalstandisthatnothingholdsunlessthereisareasonwhyitshould.Thenwearecommittedtothinkthat,ifsomephenomenonCobtains,somethingmustbetruewhichexplainswhyCobtains.Butthenitmustbepossibleforustoarguebackwards,againstthecur-rentofdeductivesequence,andtoinferfromC’sobtainingwhateverCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth123bestexplainswhyCobtains.Butherewecomebacktoabduction,whichsuppliesselectedmaterialstoinduction.(SeeVIIend.)Ifstuffslikebreadnourish,theremustbesomethingorotherabouttheminvirtueofwhichtheydothat....Ofcourse“bread”maybethewrongbasisforanabductionandensuinggeneralization.Butthisisaques-tionthatwecanonlyattemptfromthemidstofalargebackground,alreadygiven,ofcollateralbeliefs,nonarbitrarysuspicions,conjec-tures,questions,andtherest.Thelabel“bread”isourprovisionalplace-holderforsomestufforotherthatmakesadifferencetolife’sbeingsustained.(Cp.CP4.234,1902)“Bread”providesuswiththematerialsforahypothesisthatcanbetested,qualified,reformulated,testedagain,andsoon.Inpracticeandsofar,somehypotheseshavestoodup.Whentheyfail,wewillstartrepairingthem.Itwouldbearbitrarytoproceedinanyotherwayandworsethanarbitrarynottoproceedinthisone.Ofthis,indeed,wecanbeintuitivelycertain.Noneofthisprovesthatbreadwillcontinuetonourish.SuchaproofwasnotwhatHumetookhimselftobeentitledtoaskfor.Whatheaskedwaswhatkindofreasonableinferenceitisthatgivestheconclusion(howeverfallible)thatbreadwillnourish.Theanswertohisquestionisthatitisafallibleextrapolation,whichweshouldbepracticallyirrationalnottoattempt,fromanabductivehypothesisthatweshouldhavebeenpracticallyirrationalnottotrytoformu-lateandtest,anabductivehypothesisarrivedatfromwhereverweactuallyare,andmadeinaccordancewiththebranchofthinkingthatthenineteenthcenturycalledlogic.Exceptinsofarasitsub-sumesthescienceofdeduction,itisnotthebusinessofsuchlogic,anditdoesnotneedtobeitsbusiness,tofurnishinfallibledirectionsbywhichtoarguefromtheknowntotheunknown28–onlydirec-tionsthatitwouldbeunreasonablenottoemploy.LetthosewhoareexpertintheclassificationofformsofreasonablenessnowclassifythevariouselementsofthisresponsetoHume’schallengeandletthemassignthemvariouslytotheintuitive,thedemonstrative,andtheexperimental.29notes1.SeeMisak(1991:ch.1,forinstance),whoalertsustotheconsistentlyandstrictlypragmaticsignificationofPeirce’suseofwordssuchas“mean.”CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n124davidwiggins2.SeeShort(2000:2,andn3).AnothercitationgivenbyShort:“Themethodofscienceisitselfascientificresult”(CP6.428,1893).3.Hookway(1985:52f).4.TheimportanceofthiscitationispointedoutbySkagestad(1981:141).5.ItisworthcomparingtheindispensabilityofthishypothesiswithoneoftheseveralrolesofLeibnizianSufficientReason.Foralltheseroles,seeWiggins(1996:117–32).6.HerethereisrichcollateralevidenceofPeirce’sintentions.Especially,perhaps,weshouldtakenoteofamanuscriptof1893–5thatCherylMisakdrawstoourattention:Asfortheexperienceundertheinfluenceofwhichbeliefsareformed,whatisit?Itisnothingbuttheforcefulelementinthecourseoflife.Whateveritis...inourhistorythatwearsoutourattemptstoresistit,thatisex-perience....Themaximthatweoughttobe“guided”byexperiencemeansthatwehadbettersubmitatoncetothattowhichwemustsubmitatlast.“Guided”isnottheword;“governed”shouldbesaid.MS408,p.147,1893–95,quotedinMisak(1991:83)7.Ananalogymaybehelpful.Augustinewrote“Dilige[deum]etquodvisfac.”Theexhortation“LoveGodanddowhatthouwilt”mayseemtobeutterlypermissive.Itseemssountilyoureflectthatsuchaninjunc-tionrequiresyoutodesirenothingGodwouldnotwishyoutodesire(ornothingyouthinkhewouldnotwishyoutodesire).Itdoesnotentailthatyoushoulddowhateveryouwill.(ThereisnodoubthowAugus-tine’sdoubledirectionistobeunderstood.Nodoubtitisaquestionwhetherour“imperativelogic”wouldneedgeneralmodificationlest“conjunctionelimination”destroythesenseofsuchdoublecommands.Betterthoughtoshowthedispensabilityofimperativelogic.)8.Insomethinghewrotebefore“TheFixationofBelief,”Peircehadal-readynotedthatthereisanimportantdifferencebetweenthesettle-mentofopinionwhichresultsfrominvestigationandeveryothersuchsettlement.Investigation“willnotfixoneanswertoaquestionaswellasanother,butonthecontraryittendstounsettleopinionsatfirst,tochangethemandtoconfirmacertainopinionwhichdependsonlyonthenatureofinvestigationitself”(CP7.317,1873).Bythetimesomeonehasreachedforthefourthmethod,hewillbefullypreparedforthis.9.Peircesometimestalkslikethis:“Abductioncommitsustonothing.Itmerelycausesahypothesistobesetdownuponourdocketofcasestobetried”(CP5.602,1903).Elsewhere,heistroubledbythefactthat“itiswellwithinboundstoreckonthatthereareabillionhypothesesthatafantasticbeingmightguesswouldaccountforagivenphenomenon.”SeePeirce(1929:269–83).Forhisresponsetothisdifficulty,seethesentencesfrom“Guessing”citedattheendoftheparagraphtowhichthepresentnoteattaches.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nReflectionsonInquiryandTruth12510.SeeMisak(1991:99).11.OnNicod,seeHempel(1945).12.SeeShort(2000:n9).13.HereoneiseagertoallowonPeirce’sbehalfforthefullforceofaremarkthathemadein1902–IowethereferencetoSkagestad(1981:39)–“Inreasoning,wehavethesingularphenomenonofaphysiologicalfunctionwhichisopentoapprovalanddisapproval”(CP2.152).14.Compare“[T]hetruthofthepuremathematicalpropositionisconsti-tutedbytheimpossibilityofeverfindingacaseinwhichitfails”(CP5.567,1901).Weshallsupersedeinthenextsectiontheapparentlyrela-tionalmodeofdiscourseadoptedinthesentencetowhichthefootnoteisannexed.15.UsefulcollationsofsourcesontruthwillbefoundinHaack(1997:91–107)(whichsetsoutsomewicked,curious,andinstructivecontrastsbetweenhertwosubjects)andinMigotti(1999).16.Thatistosaythatthefactthatpprescindsfromtheaspectsofrealitywithwhichitisnotconcerned.17.CompareP.F.Strawson(1965).18.SeeWiggins(2002).19.SeePutnam(1995b).20.Inordertoelucidateapredicate(withoutnecessarilydefiningitorgivingnecessaryandsufficientconditionsforitsapplication)onedeploysthepredicateandputstousetheconceptthatitintroducesinwaysthatexhibitthecharacteroftheconceptandrevealitsconnectionwithotherconceptsthatareestablished,coeval,orcollateralwithit,andalreadyintelligibleintheirownright.(Forthepedigreeofthetermelucidation,seeWittgenstein1921:3.263,4.026,4.112.)21.Forvariousattemptsofmyowntopursuethislineofinquiry,seethearticlecitedinnote18andthebibliographythere.Marks.Themarksofthe(firstlevel)concepthorsearethe(firstlevel)prop-ertiespossessedbyallthingsthatfallunderthefirstlevelconcepthorse.Thuswearriveatthemarksoftheconcepthorsebyasking,ofthingsthatsharethepropertyofbeingahorse,whatpropertiestheyhave.Theanswerinthiscasewillbethepropertiesofhavingahead,fourlegs,asolidhoof,aflowingmaneandtail,avoicethatisaneigh....Similarlythen,whatpropertiesdothingshavethatpossessthepropertyoftruth?22.SeeMisak(1995:121,125,127).23.NE4.xiii,undated,emphasisascitedinMisak(1991:157).24.Cp.“Wemustlooktotheupshotofourconceptsinorderrightlytoapprehendthem”(CP5.3,alreadycited).25.Aswesawinthefirstepigraph,Peircethinksthattruthisidentifiablebyreferencetotheconceptofinquiry.InSectionXII,Ihaveexploitedthatverythought.ButtruthisnotforthatreasonanepistemologicalCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n126davidwigginsorinquiry-basednotion.Itisamisunderstandingofthenatureofelu-cidationtosupposethattheconceptoftruthhadtobeepistemologicaljustbecauseoneelucidatoryroutetotruthwasthroughtheconceptofinquiry.Themethodofelucidationsneitherretracesapriorprocessofcontagionnoryetleavescontagioninitstracks.26.Inacomparableformulationdevisedintheserviceofadifferentconcep-tionofassertion,Dummettsaysthat“astatement,solongasitisnotambiguousorvague,dividesallpossiblestatesofaffairsintojusttwoclasses.Foragivenstateofaffairs,eitherthestatementisusedinsuchawaythatamanwhoasserteditbutenvisagedthatstateofaffairswouldbeheldtohavespokenmisleadingly,ortheassertionofthestatementwouldnotbetakenasexpressingthespeaker’sexclusionofthatpossi-bility”(Dummett,1959:149–50).ForDummettthisclaimispartofthebuild-upforaproofoftertiumnondatur.(“Ifastateofaffairsofthefirstkindobtains,thestatementisfalse;ifallactualstatesofaffairsareofthesecondkind,itistrue.Itisthusprimafaciesenselesstosayofanystatementthatinsuch-and-suchastateofaffairsitwouldbeneithertruenorfalse.”)OnepartofDummett’sdialecticalframeworkhereistheverifiabilityprincipletotheeffectthat“astatementcannotbetrueunlessitisinprinciplecapableofbeingknowntobetrue.”ForthatreasonDummettholdsthathisowndichotomyprinciplefallsshortofimplyingbivalence.IntheabsencefromPeirce’sthoughtaswehavereconstructeditofanysuchverifiabilityprinciple,itishardtoseewhatcouldpreventPeirce’ssemanticoperationalismfromdeliveringfullbi-valence.AnilluminatingcriticalcommentaryonDummett’sargumentwillbefoundinIanRumfitt(forthcoming).27.Leibnizputstheclaimtoatheologicaluse.Indeed,hesometimestriestoprovebyitsmeanstheexistenceofGod.ButSufficientReasonitselfisneithertheologicalnorteleologicalinitsoriginalpurport.Formoreonsomeofthesematters,seenote4.28.Oreventofurnishproceduresthat“will,ifpersistedinlongenough,assuredlycorrectanyerrorconcerningfutureexperienceintowhich[they]maytemporarilyleadus”(CP2.769,1905).Peircedoesmakesuchclaims,buttheyareinessentialtohiscontributiontothe“problemofinduction.”(OnthestatusofthesePeirceanclaims,seeMisak(1991:111,115).SeealsoSectionsI,XI.)29.PeircereferstoHumeratherinfrequently.ButseeCP6.500,19066.605,18915.505,1905.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nchristopherhookway5Truth,Reality,andConvergence1.introductionMostcontemporaryphilosopherswhoaresympathetictopragma-tismareanxioustodistancethemselvesfromPeirce’spragmaticclar-ificationoftheconceptoftruthas‘theopinionwhichisfatedtobeagreedtobyallwhoinvestigate’.ThisreluctanceisfoundbothinthosewhofollowDeweyandJamesingivinglittleroletoa‘realist’notionoftruthinaccountingforourcognitiveevaluationsandinthosewhodonottakethisdirection.Whydoesthesuggestionthattruthisamatteroffatedconvergenceofopinionseemsounpromis-ingtopragmatistsotherthenPeirce?Oneproblemconcerns‘buriedsecrets’or‘lostfacts’.Itseemsevidentthattherearemanytruthswhich,wearesure,wouldnotbediscovered,howeverlongweinquireintothem.Anotheristhatitseemstoyieldastrangeviewofwhatmakessomepropositionstrue:whetheritistruethataninchofrainfellonthemorningoftheBattleofHastingsdependsuponwhatevidencewillturnupinthefutureratherthanuponmeteorologicalconditionsinsouthernEnglandnearlyathousandyearsago.ItisalsochargedthatPeirce’stheoryobscuresaninsightwhichmanyfindcentraltowhatmakespragmatismattractive.James’spluralismsuggeststhatourpracticalconcernsandaestheticinterestshavearoleindeterminingwhetherasystemofbeliefsagreeswithreality;perhapstherearedifferentversionsofrealitywhichanswertodifferentpracticalconcernsandarenotincompetition.SincePeirceidentifiesthetruthwithwhatanyoneisfatedtobelieve,ifsheonlyinquiresforlongenough,itisnaturaltoconcludethathisaccountofrealitydependsuponiden-tifyingasinglefundamentalaimforinquiry,thatofcontributingto127CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n128christopherhookwaythegrowthoffinishedknowledge.Iftruthwerenotcharacterisedbyreferencetosuchageneralinterest,whyshouldweexpectallinquirerstodiscoverit?ThislineofthoughtledHilaryPutnamtosuggestthatPeirce’saccountoftruthcommitshimtowhatBernardWilliamshascalledtheabsoluteconceptionofreality.Thisisaviewof‘theworldasitisthereanywayindependentlyofourexperience’.Suchaviewwillabstractfromanythingthatbelongstoaspecificperspectiveandwillavoiddependenceuponfeaturesofourcognitiveapparatusthatarespecificallyhuman:itisaconceptionwhichis‘tothemaximumdegreeindependentofourperspectiveanditspeculiarities’.Thusitprovidesanaccountofrealitywhichomitssecondaryqualitiessuchascolours,whichomitsvalues,andwhichmakesnouseofsuchconceptsasthoseofartefactswhichanswertospecifichumaninterests.ForWilliams,thisviewofrealityexpressesanidealofob-jectivitywhichissoughtbythe(physical)sciences:anaccountofrealitywhichisnotrelativetoanyparticularperspective,andwhichisinprincipleavailabletoanycreaturethatinvestigatesthenatureofreality(1978,1985).Humans,extraterrestrials,androbotsmightallbeincludedinthisfatedconvergenceofopinion.1DescribingWilliams’sposition,Putnamwrites:‘anyconceivablespeciesofintelligentbeings(iftheyframehypothesescorrectly,per-formtheappropriateexperiments)can“converge”towardagreementonthelawsofidealphysics,inthefashionfirstenvisagedbyC.S.Peirce’(1992b:84).Elsewhere,hereferstothe‘Peirceanideaoftruth...asacoherentsystemofbeliefswhichwillultimatelybeac-ceptedbythewidestpossiblecommunityofinquirersasaresultofstrenuousinquiry’(1990:221);andtotheidea(sharedbyPeirceandmetaphysicalrealists)thatscientificinquirywillconvergeto‘oneidealtheory’(1994:353);to‘onecompleteandconsistenttheoryofeverything’(1990:223).Exploringhowfarthisistruewillhelpustounderstandthedevel-opmentofPeirce’sideasabouttruthandreality.Thepassageswhichappeartosuggestacommitmenttothe‘absoluteconception’comefrombefore1880.Eventhese,Ishallargue,neednotbeinterpretedascarryingthatcommitment.Laterwritings,whichemphasisethatsecondaryqualitiessuchascoloursarereal,andthatkindsofarte-facts(suchaslamps)arerealkinds,fitverypoorlywithit.ThereisevenanargumenturgingthatthecontributionofreligiousbeliefCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence129toourpersonalfulfilmentandtothesuccessofourscientificin-quiriesprovides‘evidence’whichreinforcesournatural(andoftenunacknowledged)beliefintherealityofGod.Itisalsostrikingthat,insomelaterpassages,theinsistencethatsufficientwell-conductedinquiryisfatedtotakeustothetruthisconsiderablyqualified.ItisimportantthatPeirce’spragmaticaccountoftruthwasfor-mulatedinordertoclarifytheconceptofreality:‘Theopinionwhichisfatedtobeultimatelyagreedtobyallwhoinvestigate,iswhatwemeanbythetrue,andtheobjectrepresentedinthatopinionisthereal’(W3,273,1878).Hisstrategywastoprovideanexplanationor‘clarification’oftruth,aconcepthethoughtofaslogical,andthenusethistoexplainwhatwemeanbyreality:therealistheobjectofatrueproposition.Theaccountwasrequiredtovindicatethever-baldefinitionofrealityas‘thatwhosecharactersareindependentofwhatanyonemaythinkthemtobe’(W3,271,1878);‘thatmodeofbeingbyvirtueofwhichtherealthingisasitis,irrespectivelyofwhatanymindordefinitecollectionofmindsmayrepresentittobe’(CP5.565,1901).Sowhatrecommendedtheviewoftruthascon-vergencewasthatitpromisedapragmaticclarificationofthisideaofrealitybeingindependentofthought.Explainingtheoriginsoftheconceptionofrealityin‘SomeConsequencesofFourIncapacities’(1868),Peircedescribedtherealas‘thatwhich,soonerorlater,infor-mationandreasoningwouldfinallyresultin,andwhichisthereforeindependentofthevagariesofmeandyou’(W2,239).After1880,hehadanewwaytoexplaintheconceptofreality:wedirectlyper-ceiveexternalthingsasexternal,and(by1903)whenIperceivearedbook,thatindependentlyexistingbookistheimmediateobjectofmyperception.2Afterconsideringhowweshouldunderstandtheabsoluteconcep-tionofrealityandtheconnectionsbetweentruthandafatedconver-genceofopinions(Section2),weturntoPeirce’swritings,arguingthatitisunclearwhethertheearlywritingsinvolvethecommit-menttotheabsoluteconceptionthatsomeofPeirce’scriticssuggest.Hiswritingsfromafter1880certainlydonotinvolvesuchacom-mitment.Moreover(Section4),theselaterwritingsdistinguishtheconceptsoftruthandreality,suggestingthatmetaphysicsismorein-dependentoflogicthanheinitiallysupposed.Thisreflectsanewwayofthinkingaboutreality(Section5).Finally(Section6)wenotethatitmaybeamistaketothinkoftheconvergencethesisasatraditionalCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n130christopherhookway‘theory’oftruthatall.ThishelpsustoseehowPeirceavoidstheproblemofburiedsecrets.2.truth,convergence,andtheabsoluteconceptionWhenapropositionistrue,‘anyonewhoinvestigates’isfatedtoarriveatbeliefinit.Now‘investigates’isatransitiveverb.Soanyonewhoinvestigateswhat?Whenthisisspelledout,avarietyofdistinctthesesemerge.(1)Ifapropositionistrue,thenanyonewhoinquires‘intothenatureofreality’(wellenoughandlongenough)isfatedtobelieveit.(2)Ifapropositionistrue,thenanyonewhoinvestigatessomequestiontowhichthatpropositionprovidestheanswerisfatedtobelieveit.The‘traditional’readingofPeirce,thereadingthatleadstotheclaimthatheiscommittedtoanabsoluteconceptionofreality,suggeststhatheaccepts(1):anytruthis,inprinciple,accessibletoanyin-quirer.(2)neednotleadtosuchaninterpretationofhisviews.Itiscompatiblewiththerecognitionthataparticularinquirermightbefatednevertoconfrontaquestiontowhichsometruepropositionprovidestheanswer,perhapsevenwiththeadmissionthatsomeinquirercouldneverevenunderstandsuchaquestion.Whereas(1)suggeststhatanyseriousinquirercouldeventuallyreachastablebe-liefinanytrueproposition,ifonlyshepursuedherinquiriesforlongenoughandwithsufficientdiligence,(2)leadsonlytoaconditionalclaim:ifaninquirerinvestigatesaquestiontowhichagivenproposi-tionprovidesthecorrectanswer,then,grantedthatherinquiriesareconductedwellenoughandcontinueforlongenough,sheisfatedtoarriveatbeliefinthisproposition.(2)iscompatiblewithrejec-tionoftheabsoluteconceptionofreality,foritiscompatiblewiththeviewthatourdifferentperspectivesarereflectedinthevaryingrangesofquestionsthatwecanunderstandortakeseriously.Itisthuscompatiblewithallowingthattherearetruthsthatsomein-quirerswouldneverdiscover,eventruthsthatsomeinquirerscouldneverdiscover,nomatterhowmuchefforttheyputintotheirinves-tigations.Andweshouldnotethatthiscouldariseintwoways.First,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence131itispossiblethattheywouldnever(orindeedcouldnever)consideranyquestionstowhichthosetruthsprovidetheanswers.Andsec-ond,eveniftheyconsideredsuchquestions,itispossiblethattheycouldneverinquireintothemeitherlongenoughorwellenough.Comparing(1)and(2)enablesustoseethatthetextsinwhichPeirceapplieshispragmatistprincipletotheclarificationoftrutharecom-patiblewitharangeofdifferentviewsofvaryingstrengths.And,ifwearriveataninterpretationthatfits(2)withoutentailing(1),theremaybenoreasontoanticipateanycloseconnectionbetweenPeirce’stheoryoftruthasconvergenceandtheabsoluteconceptionofreality.Theexampleofsecondaryqualitieshelpsustoseethedifferencebetweenthesepositions.Supposewegrantthatpossessionofcolourconcepts(green,blue,andthelike)isavailableonlytothosewithadistinctivekindofvisualapparatus,or,perhapsmoreplausibly,thatitisonlyavailabletothosewhoeitherpossesssuchvisualapparatusorcancommunicatewith(anddeferto)thosewhodoso.Colourcon-ceptsare‘response-dependent’andarethusonlyavailabletothosefamiliarwiththeappropriatekindofresponse–somethingisredonlyifitproducesappropriatesensationsin‘normal’observersin‘normal’circumstances.Unlesswethinkthatallinquirersmustpos-sessvisualapparatuslikeoursorthattheywillinevitablyencountercreaturesthatpossesssuchvisualapparatus,position(1)willentailthatcolourpropositionscannotbetrueandthattheirobjectsarenotreal.Position(2)hasnosuchentailment:itrequiresonlythatthosecapableofunderstandingquestionsaboutcoloursarecapableoffindingtheiranswers.Thistoomaybeaproblematicclaim–andindeedIshallquestionwhetherPeirceactuallyendorseditbelow.Butwhatisimportanthereisthat,unlike(1),itdoesnotinvolveacommitmenttotheabsoluteconceptionofreality.SoisPeircecommittedtotheabsoluteconceptionofreality?Therearepassageswhich,wemaysuppose,arehardtoreconcilewithaccepting(2)butrejecting(1).Forexample:Realityisindependentofthe‘vagariesofyouandme’(1868);itscharactersare‘independentofwhatwethinkthemtobe’(1877–1878).Dosuchclaimssuggestthatcolours,forexample,arenotpartofreality?Whether‘thingswhosecharactersareindependentofwhatwetakethemtobe’arethesameas‘thingswhosecharacterswouldbecontainedinaviewoftheworldwhich,toamaximumdegree,wasindependentofmyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n132christopherhookwayperspectiveanditspeculiarities’isveryuncertain.Ifnot,thenPeirce’sviewofrealitycanbedistinguishedfromtheabsolutecon-ception.Ofcourse,Peircedidnotdrawthesedistinctions,andhistextsoftenleaveitunclearwhatstancehewouldhavetakenhadhedoneso.However,ifwearetounderstandandevaluatehisviewsoftruthandreality,wemustkeepthesedifferencesinmindandcon-siderthedifferentpositionsthatarecompatiblewithwhathesays.3.somedevelopmentsinpeirce’sviewsInthissection,weexamineanumberofpassagesfromdifferentstagesinthedevelopmentofPeirce’sviewsinwhichhediscussestruth,reality,andtheconvergenceofopinion.Westartwithmate-rialfrombefore1880thatisoftentakentomanifestacommitmenttosomethingliketheabsoluteconception,butthatIshallarguedoesnotdoso.Wethenconsideranumberoflaterdiscussionsthatshowthat,after1880atleast,Peircehadnosuchcommitment.(a)1877–1878Putnam,andothercommentatorswhofindtheabsoluteconceptioninPeirce’swritingsontruth,refertotheclarificationoftruthandrealityin‘HowtoMakeOurIdeasClear’andthepassagewecitedfrom‘TheFixationofBelief’.Thefirstofthesepassagesemploysanexample:[All]thefollowersofsciencearefullypersuadedthattheprocessesofinves-tigation,ifonlypushedfarenough,willgiveonecertainsolutiontoeveryquestiontowhichtheycanbeapplied.Onemanmayinvestigatetheveloc-ityoflightbystudyingthetransitsofVenusandtheaberrationofthestars;anotherbytheoppositionsofMarsandtheeclipsesofJupiter’ssatellites;[etc.]...Theymayatfirstobtaindifferentresults,but,aseachperfectshismethodandhisprocesses,theresultwillmovesteadilytowardsadestinedcentre.Sowithallscientificresearch.Differentmindsmaysetoutwiththemostantagonisticviews,buttheprocessofinvestigationcarriesthembyaforceoutsidethemselvestooneandthesameconclusion.Thisactivityofthoughtbywhichwearecarried,notwherewewishbuttoaforeordainedgoal,isliketheoperationofdestiny.Nomodificationofthepointofviewtaken,noselectionofotherfactsforstudy,nonaturalbentofmindeven,canCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence133enablemantoescapethepredestinateopinion.Thisgreatlawisembodiedintheconceptionoftruthandreality.(W3,273)Althoughconsistentwithit,thisillustrationdoesnotrequirecon-vergencethesis(1).3Itrequiresonlythatthedestinedopinionwillbereachedbythosewhoseekananswertoaparticularquestion,inthiscasethosetryingtomeasurethevelocityoflight.Althoughtheexampleisnotworkedoutinanydetail,itisleftopenwhetherthosewhorejectthatscientificframeworkaredestinedtoinvestigatethequestionandreachthecorrectanswer.Foraphilosophersympathetictorealismandanxioustoexplainrealitybyreferencetoafatedconvergenceofopinion,twopossibil-itiesareparticularlydisturbing.Wemightallacceptpermanentlysomepropositionthatis,infact,false;andwemightreachnodes-tinedconvergenceatalluponsomematterwherethereisatruth.Realismsurelydemandsthatweallowforthesepossibilities;butPeirce’saccountoftruthandrealityappearstofindnoroomforthem.Immediatelyaftergivinghisclarificationoftruthandreal-ityin‘HowtoMakeOurIdeasClear’,Peirceconfrontsthisissue:‘Ourperversityandthatofothersmayindefinitelypostponetheset-tlementofopinion;itmightevenconceivablycauseanarbitrarypropositiontobeuniversallyacceptedaslongasthehumanraceshouldlast’(W3,274).However,heinsists,thisshowsonlythatwehavenotcarriedourinvestigations‘sufficientlyfar’:if,aftertheextinctionofourrace,anothershouldarisewithfacultiesanddis-positionforinvestigation,thetrueopinionmustbetheonewhichtheywouldultimatelycometo(presumablyunlesstheytoowere‘perverse’intheirinvestigations).Thirtyyearslater,in‘WhatPrag-matismIs’(1905),Peircereturnedtohisearlierclaimsabouttruthandreality.Afterexplainingthe‘destinedopinion’asonewhichis‘controlledbyarationalexperimentallogic’and‘doesnotdependuponanyaccidentalcircumstances’,heonceagainallowedthatthe‘perversityofthoughtofwholegenerationsmaycausepostponementoftheultimatefixation’(CP5.430).What‘perversity’meanshereisunclear:isthereaconnotationofcognitivefailureormalfunctiononthepartofinquirers?Thelaterreferenceto‘accidentalcircumstances’suggeststhat,perhapsthroughbadluck,relevantevidencemayescapeourattentionandweCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n134christopherhookwaymaythusfailtoreachthetruth.Whateverhehasinmind,reflectionononeofPeirce’sownexampleswillhelpustoseethathisapparentconfidencethat,ifonlywekeepinquiringforlongenough,theef-fectsofperversityandbadluckwillbeovercome,ismisplaced.Thedescriptionoftheconvergenceaboutthevelocityoflightempha-sisedthatmanyinquirers,usingdifferentmethods,wouldarriveatthesameresult.Theconvergencethesisrequiressomethingstronger:thattherewerenonewhowouldnothavereachedthecorrectanswerhadtheycontinuedforlongenough.If,asPeircesuggests,reachingthecorrectanswerdependsuponrevisingone’stechniquesandim-provingone’smethods,itwilldependuponthereliabilityandcom-pletenessofthebackgroundknowledgethatshapestheseimprove-ments.Ifwearesentdownthewrongpathbyfaultybackgroundknowledge,thereisnoguaranteethatweshallrecoverthestraightandnarrow,nomatterhowcarefulweareinourinquiries.4Asweshallnowsee,Peirce’sformulationsoftheconvergencethesisafter1880qualifyhisapparentconfidencethatweshallalwayseventuallyreachafatedconvergence.(b)ConvergenceandHopeIndiscussingsomeviewsofSchroeder’saboutthepresuppositionsofinquiryin1896,Peircebeganwithastraightforwardformulationoftheconvergencethesis:[As]toaninquirypresupposingthatthereissomeonetruth,whatcanthispossiblymeanexceptthatthereisonedestinedupshottoinquirywithref-erencetothequestioninhand–oneresultthatwhenreachedwillneverbeoverthrown.(CP432)ThepassagecontinuesUndoubtedly,wehopethatthis,orsomethingapproximatingtothis,isso,orweshouldnottroubleourselvestomakemuchinquiry.Butwedonotnecessarilyhavemuchconfidencethatitisso.Thissuggestsaconsiderableweakeningoftheclaimsfrom1877–1878.Transformingthecommitmenttoconvergenceintoahope,aregulativeideal,isapervasivefeatureofhislaterwritings.IndeedwhenPeirceplannedtorepublish‘HowtoMakeOurIdeasClear’in1903,heproposedtwochangestothepassagefrom(W3,273).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence135Thefirstsentencewasrevisedsothatitbegan‘...allthefollowersofscienceareanimatedbythecheerfulhope...’;andtheconclu-sionwas‘Thisgreathopeisembodiedintheconceptionoftruthandreality’(CP5.407;italicsadded).MurrayMurpheyhasemphasisedthatsometimebetween1880and1890,Peirce’searlierconstitutiveprinciplelinkingrealityandthedestinedfinalopinionwasweak-enedtoaregulativeonewhichheldthat‘inordertomakecertainthatagreementwillbepursued,itisnecessarytohopethatultimateagreementwillcome’(Murphey1961:301).5Second,notethatthishopeisfocusedon‘thequestioninhand’,suggestingconvergencethesis(2)ratherthan(1).Inthesamespirit,‘WhatPragmaticismIs’declaresthat:‘everymanofusvirtuallyas-sumesthat[theconvergencethesisistrue]inregardtoeachmatterthetruthofwhichheseriouslydiscusses’(italicsadded).‘Virtuallyassumes’maybeasimplealternativeto‘hopes’.Andthethesisisrestrictedtomattersthetruthofwhichweseriouslydiscuss:itisirrationaltoinquireintosomethingunlesswethinkthereisaseri-ouschancethatwearedestinedtoreachthecorrectanswerandcanescapebeingdiverteddownthewrongroadbyfaultybackgroundknowledgeorflawedtechniquesofinquiry.Solongaswearenotseriouslyconcernedtoinquireintosomematter,itseems,weneedneitherbelievenorevenhopethatfurtherinquiryintothematterwouldtakeustoafatedconvergence.Thus:Itisrationaltomakesomequestiontheobjectofaninquiryonlyifwecan(atleast)rationallyhopethatwewillreachasolutionthatwouldalsobereachedbyanyonewhoinquiredintothesamemanner(andwhoseinquirywasnothamperedbyperversityorbyunpropitious‘accidentalcircumstances’).ThisviewisalsomanifestinaresponsetoPaulCaruswho,toPeirce’sapparentsurprise,hadinterpretedPeirce’stalkoffatedconvergenceassuggestingthatourreachingthetruthwassomething‘inevitable’.Herespondedthatconvergencewas‘ahopethatsuchaconclusionmaybesubstantiallyreachedconcerningtheparticularquestionswithwhichourinquiriesarebusied’(CP6.610).AsPeirce’sphilosophydevelopedafter1878,hesooncametogivehisaccountoftrutharegulativestatus:wehopewewillconvergeonthetruthifweinquirelongenoughandwellenough.AndmanyofhisillustrationsandformulationssuggestthatthethesisshouldbeformulatedwithrespecttoparticularquestionsormattersforCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n136christopherhookwaydiscussion.Whenweclaimthatsomepropositionistrue,we(virtu-allyassume)thatanywell-conducted,sufficientlylonginquiryintothetruthvalueofthatpropositionwouldendupadmittingthatitwastrue.Andwhenweinquireintosomequestion,wehopethatthereisananswertoitwhichanynonperverseinvestigationwhichisunaffectedbyaccidentalcircumstanceswouldeventuallyaccept.Thus(2)isemployedratherthat(1);butitisinterpretedasareg-ulativeprincipleratherthanasa‘law’orasubstantivetruth.Thisweakerversioncanbeexpressedasfollows:(3)Itisrationalforsomeonetoassertthatp,ortoinquireintowhetherp,onlyifitisrationalforhertohopethatanyonewhoinquiredintowhetherp(longenoughandwellenough)wouldbefatedeventuallytoarriveatastablebeliefinp.(c)SecondaryQualitiesandRealityLetusnowreturntotheissuesaboutsecondaryqualitiesthatwereraisedinSection2.Sinceourconceptsofsecondaryqualitiesarerel-ativetoourhumansensoryapparatus,andourconceptsofartefactsarerelativetoourneeds,interestsandcapacities,theywouldhavenoplaceintheabsoluteconceptionoftheworld.IfPeirceiscommittedtothatconception,hemustdenytherealityofsuchconceptions.Inthatcase,itisrevealingthat,after1900,Peircechallengedthe‘vir-tualassumptionthatwhatisrelativetothoughtcannotbereal’(CP5.430,1905).Ontologicalmetaphysiciansusuallysaythat‘secondarysensations,’suchascolours,aredelusiveandfalse;butnotsothePragmaticist.Heinsiststhattherosereallyisred;forredis,bythemeaningoftheword,anappearance;andtosaythataJacqueminotroseisredmeans,andcanmean,nothingbutthatifsucharoseisputbeforeanormaleye,inthedaylight,itwilllookred.(CP8.194,1904;amuchfullerdiscussion,claimingthatcolouris‘external’andsothedenotationsofcolourtermsarereal,isfoundinCP6.327–8,from1909)Whetherhewouldhavedeniedtherealityofredinthe1870sisunclear,althoughtheinterpretationofhimasdefendingtheabsoluteconceptionofrealitywouldhaverequiredhimtodoso.InTheMinuteLogic(1902),Peirceexaminedthenotionofa‘real’,‘true’,or‘natural’class.Havingdefinedaclassas‘thetotalofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence137whateverobjectstheremaybeintheuniversewhichareofacer-taindescription’,heclaimsthataclassis‘natural’or‘real’whenitsmembers‘owetheirexistenceasmembersoftheclasstoacom-monfinalcause’.Hecontinues‘Inthecaseoflampsweknowwhatthatcauseis:thatinstinctwhichenablesustodistinguishhumanproductionsandtodivinetheirpurposeinformsuswithadegreeofcertaintywhichitwerefutiletohopeanyscienceshouldsurpass’(CP1.204).Theclassoflampsis‘real’becauseitcorrespondstoadistinc-tivehumanpurpose.(AlthoughPeircethenpointsoutthatbiologicalclassificationsarenotanswerabletopurposes,heinsiststhattheystillreflectfinalcausationofadifferentkind,andhemakesclearthathethinksthatclassesofartefactsare‘true’,‘real’,or‘natural’.)Thisexampleshowsthat,in1902,PeircesharedJames’sviewthatrealitycanbe‘relativetothought’ortohumaninterests,capacities,anddesires.Whethersomethingisred,oralamp,is‘external’and‘independentofwhatanyonethinksittobe’eveniftheconceptsinquestionaresensitivetoadistinctivehumanperspectiveandcouldnotbeunderstoodbyanyonewhowasunabletoenterthatperspec-tive.Itisalsousedtoillustratethepervasivenessofvaguenessinourthoughtaboutreality,somethingwhichrequiresfurtherqual-ificationoftheclaimaboutconvergence.Inviewofthevarietyofusestowhichweputlamps,theremaynotbeasinglepurposeordesirethatunifiestheclassoflamps.Instead,thatclassmaybeheldtogetherbyasetoflooselysimilar‘desires’whichcorrespondtothesedifferentuses.Lampscanbesubdividedintodifferentkindsthatcananswertomorespecificneedsandpurposes(CP1.205);andeventhesemorespecificdesireswillbevague,varyingalonganum-berofdimensions(CP1.206).Judgementisthusrequiredwhenweranklampsordecidewhethersomeobjectanswerstheintereststhatmakesuswantalamponaparticularoccasion;wemustbesensi-tivetothebalanceofpurposesandintereststhatarerelevantinthecontext(CP1.207).Withinrealornaturalclasses,‘objectsactuallywillclusteraboutcertainmiddlingqualities,somebeingremovedthisway,somethatway,andatgreaterandgreaterremovesfewerandfewerobjectswillbesodetermined’(CP1.207).‘Anditmaybequiteimpossibletodrawasharplineofdemarcationbetweentwoclasses,althoughtheyarerealandnaturalclassesinstrictesttruth’(CP1.208).OurinterpretationofPeirce’sconvergencethesismustallowforthefactthatinquirerswithdifferentinterestsmayarriveCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n138christopherhookwayatadifferentjudgementaboutwhethersomeobjectisalamp.Atbestwemusthopethatdifferentpeoplewillexercisejudgmentinharmo-niouswayswheneveritisimportantthattherebeagreement.6Somereaderswillwonderhowsomeonewhoexplainsrealityintermsofthefatedconvergenceofopinioncanendorsetheseclaims.Intheremainderofthechapter,IshallarguethattheyreflectachangeinPeirce’sphilosophicalposition,butitisachangeinhowhethoughtaboutreality,notachangeinhowhethoughtabouttruth.Itisaconsequenceofthisthattheaccountoftruthcomestohaveaverydifferentroleinhisphilosophy.MoreoverhisnewwayofthinkingaboutrealityembracesideasthatsomeofPeirce’scriticswouldidentifyaspragmatistinsightswhichPeircehimselffailedtoembrace.4.theconceptsoftruthandrealityAswehaveseen,Peirce’searlierformulationsoftheconvergencethesisallformpartofhissearchforanexplanationoftheconceptofreality.Theimmediatecorollaryoftheclaimthattruepropositionsarethosewearedestinedtobelieveisthatrealityistheobjectofthisfatedopinion.Thewaytocapturetheideathatrealityisexter-nal,thatitisindependentofwhatwethinkittobe,istoinsistthatanyinquirerwhoinvestigatedforlongenoughwouldbeboundtoacknowledgethetruthofanopinionthatwasproperlydescriptiveofreality.Realityisindependentofthoughtbecauseitisindependentofwhatanyindividualmaythinkatanyparticulartime(W2,467–70).Thisis,ifyoulike,hisconstitutiveaccountofreality:hesawitastheonlywaytomakesenseofhowrealityisindependentofouropinionswithoutsuccumbingtoscepticismbyturningthisin-dependentrealityintoanunknowablethinginitself.Inthissectionweconsidersomepassagesthatshowthattruthandrealitywerelessintimatelyconnectedafter1880.Thefirstistakenfromamanuscriptandthusmaynotrepre-sentalastingthemeinPeirce’sthought.Itconcernsthetensionsbetweentheconvergencethesisandfundamentallogicalprinciplessuchasthelawofbivalence.Unlessweaccept,implausibly,thateveryproposition(oritsnegation)willbetheobjectofafatedcon-vergence,Peirce’saccountoftruthshouldleadhimtoquestiontheprinciplethateachpropositioniseithertrueoffalse.HoweverCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence139considerthispassagewhichsuggestsawayofreconcilingaversionoftheconvergencethesiswithbivalence:[Every]propositioniseithertrueorfalse.Itisfalseifanypropositioncouldbelegitimatelydeducedfromit,withoutanyaidfromfalsepropositions,whichwouldconflictwithadirectperceptualjudgment,couldsuchbehad.Apropositionistrue,ifitisnotfalse.Hence,anentirelymeaninglessformofproposition,ifitbecalledaproposition,atall,istobeclassedalongwithtruepropositions.(EP2:284–5)Althoughthisiscloselyrelatedtothetruthasconvergencethesis,itisformulatedasadefinitionoffalsity:truthisthendefinedasanythingthatcannotberefuted.SincePeircewouldnotwanttoconcludethattheobjectof‘anentirelymeaninglessformofpropo-sition’isreal,hisflirtationwiththisstrategy–evenifonlybriefly–suggeststhatheisquestioningtheideathattheobjectofanytruepropositionisreal.Second,thedefinitionisflawed.Thequalifica-tionsintroducedinhisexplanationoffalsitysuggeststhathewasawarethatwidespreadignoranceorerrorinourbackgroundbeliefsmightproveapermanentobstacletotheconvergenceofopinion.Buttotrytorepairhisaccountbysayingthatourrefuting(orfailingtore-fute)somepropositionmustnotdependuponanyfalsepropositionsrendershisexplanationoffalsitycircular.ThedecisiveevidenceofhisabandoninghisanalysisofrealityastheobjectofatruepropositionisfoundinPeirce’sdefinitionoftruthforBaldwin’s‘DictionaryofPhilosophyandPsychology’in1902.Truthisstillexplainedintermsofconvergence:‘Truthisthatconcordanceofanabstractstatementwiththeideallimittowhichendlessinvestigationwouldtendtobringscientificbelief’(CP5.565).Realityisnowexplainedas‘thatmodeofbeingbyvirtueofwhichtherealthingisasitis,irrespectivelyofwhatanymindordefinitecollectionofmindsmayrepresentittobe’(CP5.565).Andhenowtakesseriouslythepossibilitythattheremightbetruthwherethereisnoreality,and,indeed,realitywherethereisnotruth.Howmightthereberealitywherethereisnotruth?Vaguenessmightofferoneexample.Thevagueproposition‘Xisbald’maybeneithertruenorfalseeveniftheunderlyingreality(thedistributionofhairsontheperson’shead)isfullydeterminate.Peirce’sownex-ample,tentativelyputforward,supposesthat‘ifinrespecttosomequestion–saythatoffreedomofthewill–nomatterhowlongtheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n140christopherhookwaydiscussiongoeson,nomatterhowscientificourmethodsmaybe-come,thereneverwillbeatimewhenwecanfullysatisfyourselveseitherthatthequestionhasnomeaning,orthatoneanswerortheotherexplainsthefacts,theninregardtothatquestion,therecer-tainlyisnotruth’(CP5.565).Peirceisexplicitthatthemetaphysicalquestionthenremainsopen.‘Evenifthemetaphysiciandecidesthatwherethereisnotruththereisnoreality,stillthedistinctionbe-tweenthecharacteroftruthandthecharacterofrealityisplainanddefinable.’Ifthemetaphysiciancanproduceadefensibleaccountofthemodeofbeingofthefreedomofthewill,onethatperhapsex-plainswhywecouldnotresolvethequestion,thenwemayadmitthatthereisarealitytowhichnotruthconforms.7Peircealsohasexamplesofhowtherecanbetruthwithoutreal-ity.Heconsidersa‘moralist’whodescribesanidealofasummumbonum,incircumstancesinwhich‘thedevelopmentofman’smoralnaturewillonlyleadtoafirmersatisfactionwiththedescribedideal’.Itmaybeaconsequenceofthisthat,withtime,anyonewhothinksaboutthematterlongenoughandcarefullyenoughwillbeledtosharethisviewofthegood(CP5.566).Thisappearstobeenoughtorenderpropositionsaboutthegoodandbadtrue.Theremaybeafatedconsensusontheacceptabilityofsuchclaims.Isthereanyrealitycorrespondingtosuchtruths?Thepassages,notedabove,inwhichPeircegrantsthatrealitymightbe‘relativetothought’,sug-gestthatthereis:thatrealitiesshouldbe‘relativeto’oursentimentaldispositionsandmoralnatureneeddisparagetheirrealitynomorethanthefactthatcoloursarerelativetooursensoryapparatusshouldleadustoquestiontheirreality.Iftheexampleisgenuinelytoof-feracasewherethereistruthbutnoreality,weneedtounderstandhowtheroleofthemoralistandthereinforcementofthemoralidealthroughtimesomehowmakesitsacceptancedependentuponwhatsomefinitegroupofindividualsthink.Perhapsitsgeneralacceptancedependsuponthecharismaticpowersofthemoralist:ifadifferentmoralleaderhasseizedthelimelight,adifferentmoralpropositionwouldhavebeentrue.Inthe1860sand1870s,Peirceaimedtoderivemetaphysicalcon-clusionsfromlogic.Henceanaccountoftruthasalogicalconcept–onewhichwasexplainedbyreferencetoitsroleininquiry–yieldedananalysisofreality,whichsomemightsupposetobeametaphysi-calconcept.Duringthe1880sand1890s,PeirceaddressedproblemsCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence141thatledhimtosearchforasystemofscientificmetaphysicsthat,interalia,woulddescribethedifferentmodesofbeingthatrealitiescouldhave.Peirce’sdiscussionofthefreedomofthewillexampleshowsthat,by1901,questionsofrealitycouldbeaddressedbymeta-physicianswhichwerenotsettledbyappealtoalogicalanalysisoftruth.Whenweregardapropositionastrue,wehopefor(andper-hapsexpect)convergenceuponit.Whensuchconvergenceoccurs,weshouldexpectanexplanationofit:whydowecometoagreeoncolours,onourmoraljudgements,onourbeliefsaboutthephysicalworld?Fordifferentkindsoftruths,differentkindsofexplanationmaybeappropriate–considerthedifferencesbetweenmoraltruths,mathematicaltruths,biologicaltruths,etc.The‘modesofbeing’ofmathematicalobjects,moralfacts,andscientificmatterswillbere-flectedinthekindsofexplanationsofconvergencethatwefindap-propriate.PeirceconsidersthepropositionthatCaesarcrossedtheRubicon.Itstruth‘consistsinthefactthatthefurtherwepushourarchaeologicalandotherstudies,themorestronglywillthatconclu-sionforceitselfonourmindsforever’.Anidealistmetaphysicianmayholdthatthecorrespondingrealityconsistsin–orisconstitutedby–thefacttheinquiryisfatedtotakethispath.Whereopinionwillnotconverge,realitywouldbeindeterminate.Butwemightalsoexplainconvergencebyshowinghowexperimentalinteractionwithanin-dependentlyexistingrealitywillsufficetoensurethatrealitymakesitselfmanifesttous.ThusevenifwedecidethatCaesar’scrossingtheRubiconissomethingreal,somethingindependentofthoughtintherequiredsense,thereisscopeforavarietyofmetaphysicalaccountsofwhatitsrealityconsistsin,ofwhatmakesthecorre-spondingpropositiontrue.Similarly,althoughitisuncontroversialthatmanystatementsofmathematicsaretrue,thesuppositionthatmathematicsdealsonlyinhypotheticalstructuresleavesroomforittobecontroversialwhetherarealistunderstandingofthesubjectmatterofmathematicsisdefensible(CP5.567).Andevenifareal-istaccountisaccepted,thereareopenquestionsaboutthemodesofbeingoftheobjectsofmathematicaltruths.Itisimportanttodistinguishsomeofthedifferentpointsbe-ingmadeinthissection.Issuesof‘realism’concernthemind-independenceofthefactorsthatdeterminewhetherparticularpropo-sitionsaretrueorfalse.Mathematicsismind-independentinatleastthisrespect,thatwhetheritistruethatfourplusfiveequalsnineCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n142christopherhookwaydoesnotdependuponwhetheranyoneactuallybelievesthatitistrue.Wewouldnotexplainitstruthasgroundedinthefactthatsomeparticularperson(orgroupofpeople)acceptsit.Ifasubjectmatterismind-independentinthisway,wecansaythatsatisfies‘basicrealism’.Thefirstpointisthatthefactthatapropositionsatisfiesbasicrealismleavesthemetaphysicalcharacter(the‘modeofbeing’)ofitsobjectsopen.Termssuchas‘lamp’collectthingstogetherthatembodyformsofthirdnessthataremind-dependent,whilenatu-ralkindtermscollectthingstogetherthatsatisfythesamemind-independentlaws.Physicalobjectsmanifestrealsecondness;‘ideal’or‘hypothetical’objectssuchasnumbersdonotmanifestrealsec-ondness.Perhapswecanthinkofbasicrealismasasortoflogicaldoctrineratherthanametaphysicalone.Andthenexamplessuchasthatofthelampshowjusthowvariousandcomplextheissuesaboutmind-independenceare.Certainlybasicrealismleavesopenmostofthedifferentoptionsthathavebeendiscussedbyphilosopherswhohavedebatedrealismaboutmathematicalobjects,externalthings,values,laws,causation,andthelike.Thesecondpointisthatonceweclarifytheconceptoftruthbysayingthatapropositionistrueifanyonewhoinquiredintoitwouldeventuallyarriveatastablebeliefinit,then,althoughtheapplicationoftheconceptoftruthtoapropositionmightguaranteebasicreal-ismaboutitsobjects,allthesefurthermetaphysicalquestionswillremainopen.Thismightbeexpressedbysayingthattheconceptoftruthisalogicalconceptandismetaphysicallyneutral.Sayingthisdoesnotnecessarilypreventourdefining‘thereal’as‘theobjectofatrueproposition’.Iftruthguaranteesbasicrealism,andordinarytalkof‘thereal’and‘reality’isconcernedwiththerequirementsofba-sicrealism(‘Therereallyareprimenumbersbetweentwoandten’),thentheapplicationoftheconceptofrealitywillbejustasmeta-physicallyneutralastheapplicationoftheconceptoftruth.Butthisclarificationoftheconceptofrealitywouldnotequipustounder-standandinquireintoimportantmetaphysicalquestionsabouttherealityofsecondnessandthirdness(forexample).Aricherunder-standingofmind-independenceisneededbeforewecaninvestigatewhetherthesethingsarereal.Butthereisathirdpoint.Peirce’sexampleofthemoralistsuggeststhattheapplicationoftheconceptoftruthdoesnotalwaysentailbasicrealism.Andifthatiscorrect,theconceptsoftruthandrealityCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence143comeapart;oncethisexampleisaccepted,thenwecannotsimplydefinetherealastheobjectofatrueopinion.Inthatcasetheconceptoftruthwillbeneutralinawaythattheconceptofrealityisnot:itdoesnotevenentailthetruthofbasicrealism.Andthatwouldmeanthatanewstartisneededonclarifyingtheconceptofreality.8Wenowfaceanimportantproblem.ForapragmatistlikePeirce,conceptsshouldbeexplainedbyreferencetoexperience:weex-plainwhatitisforsomethingtobehardbyshowinghowourac-tionsuponhardthingshavedifferentempiricalconsequencesfromsimilaractionsuponsoftthings.Theearlytheorythusheldthat,ifsomethingisreal,theninquiryintowhetheritisrealwillhaveanobservableconsequence,theconvergenceofopinion.Moreoverthisconvergencewillresultfromourregulatinginquiriesinthelightofexperience.Ifrealityandtrutharetocomeapart,orifrealityrequiresmoreclarificationthantheearlytheoryprovided,thenPeirceneedsanewwayoflinkingourthoughtaboutreal-itytoexperience.Whatisthis?Experienceisricherthatmanyphilosopherssupposeanditcontainsthematerialsforexplainingexternalityandindependenceofthought.95.experienceandrealityInhiswritingsfromthe1870s,Peirce’s‘realism’aboutlawsandabouttheexternalworlddependedcruciallyuponhisclaimthatproposi-tionsabouttheexternalworld,andpropositionsdescriptiveoflawsofnature,wouldformpartofafatedconvergenceofopinion.Ihavebeensuggestingthatinthefinalthreedecadesofhisphilosophicallife,hisrealismreceivedadifferentformulation.Externalthingsandlawgovernedchangesareobjectsof‘immediate’experience:ourex-periencemanifestssecondnessandthirdnessandthisisexplainedbyreferencetosecondnessandthirdnessinthebehaviourandinterac-tionsofexternalthings.Inturningtothephilosophyofperceptioninthinkingaboutrealism,heresemblesHilaryPutnamwho,inhis1994DeweyLectures,cametoinsistthatthetheoryofperceptionheldthekeytoproblemsaboutrealism(1999:13–14).Crucialdevelopmentsinthe1880sandaround1900involvein-sistingthatwehaveadirectperceptualawarenessofindependentexternalthings.Oneofthefirstofthesedevelopments,perhapsthemostimportant,isfoundinPeirce’s(unpublished)reviewofJosiahRoyce’sTheReligiousAspectofPhilosophy,whichhereadasCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n144christopherhookwaychallenginghisaccountoftruthandassuggestingthathecouldmakenosenseofthepossibilityoffalsebelief(seeHookway2000:Chapterfour).HeclaimedthatRoycerelieduponthemistakenviewthatwecanonlyrefertoanobjectaswhateverfitssomeindividualconceptordescription;andheproposedthatthemostfundamentalformofreferenceisindexical:WhenIsayImeanmydiscoursetoapplytotherealworld,theword“real”doesnotdescribewhatkindofworlditis:itonlyservestobringthemindofmyhearerbacktothatworldwhichheknowssowellbysight,hearing,andtouch,andofwhichthosesensationsarethemselvesindicesofthesamekind.Suchademonstrativesignisanecessaryappendagetoaproposition,toshowwhatworldofobjects,orasthelogicianssay,what“universeofdiscourse”ithasinview.(W4,250)Inhisreview,Peirceemphasizedthatthesamewastrueofordinaryperceptualjudgements:Ijudgethatthatisablackcomputer,thatthisisredbook.Andinmakingsuchjudgements,weareawareofsuchthingsasexternal.Inpickingthingsoutindexically,weareawareofthemasexistingindependentofourthoughtsaboutthem,assomethingexternalwithwhichweinteract.Thiswasveryexplicitinthelectureson“Pragmatism”thatweredeliveredinHarvardin1903.Describinghisexperienceofseeingayellowchairwithagreencushion,andhavingintroducedtheterm‘percept’forwhatisimmediatelypresenttothemindinperceptualexperience,heinsiststhatthechairisthepercept:‘ThechairIap-peartoseemakesnoprofessionsofanykind,essentiallyembodiesnointentionsofanykind,doesnotstandforanything.Itobtrudesuponmygaze;butnotasdeputyforanythingelse,nor“as”any-thing’(CP7.619,1903;forfurtherdiscussion,seeHookway1985:155ff).Inearlierwritingshehadsupposedthatthinkingofrealityasthecauseofourexperienceswoulddriveustowardsadmittingun-knowablethingsinthemselves.Ifthefactthatweperceiveexternalthingsasexternalismanifestwithinthecontentoftheexperienceitself,thisdangerdisappears.Moreoveritiseasytoseethatthisde-velopmentmakesiteasytoloosentheconnectionbetweentruthandreality:theremightbeafatedconvergenceuponmattersthatdonotmanifestexternalityorindependenceinthisway.Thepresenceofconvergenceuponanopiniondoesnotguaranteethatitsobjectswillbeexperiencedasexternalthingsthatinteractwithus.ItalsoleavesopenthepossibilitythattheremaybeotherwaysforrealthingstoCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence145beindependentofthought:thisrequires‘metaphysical’studyandcannotbegroundedsolelyinaccountsofthestructureofinquiryandthefatedconvergenceofopinion.AnotherdevelopmentinPeirce’sthoughtmayhavereinforcedhisrealism.AppealtothedefeasibleauthorityofcommonsensewasanenduringfeatureofPeirce’sthoughtand,after1905,hewashappytodescribehispositionas‘criticalcommon-sensism’.Commonsense,hetellsus,suggestsdirectrealismabouttheobjectsofperception(CP5.444,5.539),andcriticalcommonsensismendorsescommonsenseclaimsoncetheyhavebeenquestionedandrefinedintheinterestsofconstructingatheorythatmeetsourphilosophicalneeds.10Peirce’scriticalcommonsensismallowsforaqualifiedendorsementofoureverydayrealism.Aswenotedabove,similardevelopmentsoccurredinPeirce’sthoughtaboutlawsandgenerality.Inhisearlierwritings,theques-tionofrealismaboutnaturallawsisformulatedasaquestionaboutwhethersentencesformulatingsuchlawswillformpartofafatedconvergenceofopinionamongthosewhoinvestigate.Duringthe1880s,hedevelopedametaphysicalaccountoflawsasrealhabits,describingtheformstakenby‘thirdness’indifferentdisciplines(Hookway2000:Chapter6).Andbytheturnofthecentury,whenhesoughtaphenomenologicaldefenceofhissystemofcategories,heclaimedthat(external)‘mediation’andotherformsofnecessityweredirectlymanifestedinthecontinuouspatternsthatwerepresentinourexperience.Soourconceptofrealityreflectsboththepresenceofexternallawgovernedprocessesinexperience,andametaphysicalaccountofthemodesofbeingoflawsandexternalthings.Thischangeisinevitablyliberating.Ifexperiencerevealslaw-governedpatternsandchangesinthecoloursthatthingsdisplay(orinthebehaviourofartefacts),thereisnolongeranyneedtoques-tiontherealityofsecondaryqualitiesandpropertiessuchasthatofbeingalamp.Ifexperience(andourmetaphysicalstoryaboutreal-ity)revealcomplexpatternsofcontinuityandmediation,thereisnolongeranylogicalobstacletogivingaroletoourpracticalinterestsandaestheticsensibilitiesinidentifyingsomeofthesepatternsassalienttoourpurposesandinemployingparticularidealisationsindescribingthestructureofreality.Solongasthese‘reals’are‘ex-ternal’andarenotcreatedbyourbeliefsthattheyobtain,itisnoobstacletotheirrealitythattheyarerelativetoourperspectivesandtheirpeculiarities.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n146christopherhookwayInthe1860sand1870s,Peirce’sapproachtoquestionsaboutre-alitywastopickoutadistinctivesetofsignsorsentences,thosethatwearefatedtoacceptifweinquiresufficientlywell,andthentodefinerealityastheobjectsofthesesentences.This,hethought,wastheonlywaytogivepragmaticsensetotheconceptofreality.Bythe1880s,hewasalreadyembarkeduponajourneythatwouldofferhimanindependentempiricalhandleontheideaofarealitythatisexternaltousandindependentofourthought.Andthereisnoreasontothinkthatthiswayofarticulatinghisrealismwouldcommithimtoanyversionoftheabsoluteconceptionofreality.6.pragmatist‘theories’oftruthThesuggestionthatPeirce’stheoryoftruthhasmetaphysicalimpli-cationsasaresultofcommittinghimtoaversionoftheabsoluteconceptionofrealityhasbeenmadebyanumberofhisreaders.Thispaperhastriedtothrowdoubtontheseclaims.Thepassagesthatsup-portthediagnosisareallrelativelyearly,datingfromatimewhenPeircetiedtheconceptsoftruthandrealitytogethermuchmorecloselythanafter1880.Peirce’slaterwritingsabouttruthandreal-itygenerallyconflictwiththisinterpretationofhisviews.Moreovereventheearlywritingsthatappeartosupporttheseattributionsarecompatiblewithaweakerviewthatincorporatesadifferentideaofconvergence.Ishallconcludebyquestioninganothercommonas-sumptionaboutPeirce’swritingsontruth:thisisthatheoffersatheoreticalaccountofwhattruthconsistsin,arivaltoother‘consti-tutivetheoriesoftruth’suchasthecorrespondenceandcoherencetheories.11AsIhavearguedin“Truth,Rationality,andPragmatism”(Hookway,2000,Chapter2),pragmaticclarificationsofconceptsandpropositionsarebestseenasaccountsofthe(experiential)commit-mentsweincurwhenweassertorjudgethepropositioninquestion.TheaccountoftruthasconvergencereflectsthebeliefthatwhenIcommitmyselftothetruthofaproposition,Imustbeconfidentthat(oratleasthopethat)anydisagreementonthepartofotherscanbeputdowntolackofinformationontheirpart;possessionofmisleadinginformationontheirpart;cognitivefailingsontheirpart;perhapsdifferencesinthewaysinwhichwehaveresolvedtheper-vasivevaguenessofnaturallanguages,fixeduniversesofdiscourse,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence147orinterpretedceterisparibusclauses;andsoon.Similarly,whenIinvestigatesomequestion,Irelyupontheconfidenceorhopethatananswercanbefoundtowhichthissortofcommitmentisappro-priate.Peirceisnotofferinganaccountofwhatitisforapropositiontobetrue.Insteadheisclarifying:(1)Whatcommitmentsweincurwhenwetakeapropositiontobetrue.(2)Whatcommitmentsweincurwhenweseektruthinsomearea.Suchclarificationscanignorepropositionsthatwedonottaketobetrueandpropositionswhosetruthvalueisnotsomethingthatwecanmakeanobjectofseriousinquiry.And,aswehavenoticed,ourcommitmentmaybetonomorethanthereasonablenessofhopingthatinquirywillproduceconvergence.ConsiderthepropositionthatCaesarsneezedthreetimesonthemorningthathefirstcrossedtoEngland.Theprincipleofbivalencesuggestsithasatruth-value–althoughthevaguenessof‘morning’andperhapsevenof‘sneeze’mayleadthattobequalified.Thema-turePeirceshouldadmitthatthereisadeterminatereality:itfitspoorlywithourmetaphysicalviewsthattherewasagapinthehis-toryoftheuniverse;andwebelievethattherewererealeventswhichhadrealeffects,butwhichhavenotaffectedourcurrentcognitivestates.Butisthereany‘truth’?Wetakenocheckableriskswhenweassertthispropositionratherthanthepropositionthathesneezedfourtimes;weincurnocommitmentsthatmightrevealthatweweremistaken.Wecouldnotreasonablyhopethatinquirywouldmakesuchcommitmentspossible–although,ofcourse,evidencemightturnup.ItiseasytounderstandhowPeircemightconcludethatthereisnotruthhere.But,ifso,thisisbecausehisaccountoftruthhasadifferentsortofaimfrommoretraditionalkindsoftheories.12notes1.Notethatthisisaconceptionwhosevalue(andindeedwhoseintelligi-bility)Putnamrejects(1992b:Chapter5).2.Myuseoftermssuchas‘explanation’and‘clarification’todescribePeirce’saccountisintendedtoleaveopenthequestionofwhetheritCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n148christopherhookwayisintendedasatraditionalkindofphilosophicalanalysis.Inthefinalsectionofthispaper,Ishallsuggestthatpragmaticclarificationsshouldnotbeviewedasordinaryanalyses,asstatementsofthenecessaryandsufficientconditionsforapplyingproblematicconcepts.3.ThemoralsthatPeirceactuallydrawsfromthisexamplearethree.First,thatifapropositionistrue,differentinquirers,workingrelativelyinde-pendently,willcometoagreeuponit.Second,thatifitistrue,differentinvestigativetechniques,differentmethodsofinquiry,willnormallyarriveatthesameanswertothetargetquestion.Andthird,thatwhendifferentinquirersanddifferentmethodsortechniquesofinquiryap-pearnottoprovidethesameanswerstothequestion,thisisusuallyashort-termphenomenonthatwilldisappearonceourtechniqueshavebeenrefined.4.RobertBrandomhassuggestedinconversationthatPeirce’stheoryoftruthisrefutedbythefactthatignoranceanderrorcanpermanentlyblockthisfatedconvergence.5.MostscholarswouldagreewithNathanHouserthatPeirce’sreflectionsonRoyce’sReligiousAspectsofPhilosophyplayedanessentialroleinthattransition(seeW5,xlvi).6.Comparehere:‘Althoughitistruethat“Anypropositionyoupleaseonceyouhavedetermineditsidentity,iseithertrueorfalse”;yetsolongasitremainsindeterminateandsowithoutidentity,itneedneitherbetruethatanypropositionyoupleaseistrue,northatanypropositionyoupleaseisfalse’(CP5.448,1905)7.ThismayalreadyleadustoquestionwhetherPeirce’saccountisacon-stitutiveaccountoftruth.8.Iamawareofonlytheoneplaceinwhichsuchexamplesareusedtomakethispoint,sothereadershouldbewarethatitmaynotrepresentanenduringthemeinPeirce’sthought.Moreover,theexamplemaybeproblematic.Itseemstoinvolveacasewhere,evenifnotconsciously,convergenceinopinionissecuredthroughusingavariantonthemethodofauthority,amethodwhichPeircerejectedin‘TheFixationofBelief’.IfitcanbearguedthatPeirce’sclarificationoftruthidentifiesitwithbeliefonwhichtherewouldbefatedconvergenceamongthosewhousethemethodofscience,thenthismaynotbeanexampleofatrueproposition:theconvergenceinopinionissecuredinthewrongway.9.ThisprovidespartoftheexplanationofwhyPeircecametolinkhispragmatismtoJames’sradicalempiricism(Hookway1997).10.ThismightrepresentafurtherparallelwithPutnam’swritings:theDeweylecturesembraced‘naturalrealism’asastartingpointforphilo-sophicalreflection(Putnam1999).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTruth,Reality,andConvergence14911.See,forexample,Putnam’spaperonWilliamJames’stheoryoftruth(1997).12.ThischapterbuildsonareadingofPeirce’saccountoftruththatisde-fendedinChapters2–4ofTruth,Rationality,andPragmatism(Hook-way,2000)andin(Hookway,2002).IamgratefultoHilaryPutnamforveryhelpfulcommentswhensomeofthismaterialwaspresentedtoaconferenceonhisworkinMunsterin2000,toLeifWenarforhisadviceonanancestorofthispiece,toCherylMisakforanumberofhelpfulsuggestions,andtoDanielleBromwichforhelpingmetoidentifysomeconfusionsinanearlierversionofSection4.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\ncherylmisak6C.S.PeirceonVitalMatters11.introductionC.S.Peircearguedthatatruebeliefisthebeliefwewouldcometo,werewetoinquireasfaraswecouldonamatter.Atruebeliefisabeliefwhichcouldnotbeimprovedupon,abeliefwhichwouldforevermeetthechallengesofreasons,argument,andevidence.Peirceinitiallyputthisideainthefollowingunhelpfulway:atruebeliefisonewhichwouldbeagreeduponatthehypotheticalor‘fated’endofinquiry(SeeW3,273,1878).Itisthisformulationwhichisusuallyattackedbythosewhoseelittlevalueinthepragmatistviewoftruth.Butamuchbetterformulationisthis:atruebeliefisonewhichwouldwithstanddoubt,werewetoinquireasfaraswefruitfullycouldintothematter.Atruebeliefissuchthat,nomatterhowmuchfurtherweweretoinvestigateanddebate,itwouldnotbeoverturnedbyrecalcitrantexperienceandargument(CP5.569,1901,6.485,1908).Ihavearguedelsewhere(Misak2000:49f)thatthisformulation,unlikethefirst,isnotvulnerabletothestandardobjectionstothepragmatistaccountoftruth.Ihavealsoargued(Misak2000)thatthisformulationisveryfriendlytocognitivismaboutmorals–veryfriendlytotheideathatmoraljudgementsfallwithinthescopeoftruth,knowledge,andin-quiry.Ourethicalbeliefsmightwellaspiretotruth,asdoourbeliefsinscience,mathematics,anddiscourseaboutordinarymiddle-sizedobjects.AttimesPeirceappearswholeheartedlytoembracethecognitivistviewIwouldliketoattributetohim–hehappilyextendshisviewoftruthandinquirytomoraljudgements.Herearetwopassagesinwhichhiscognitivistintentionsareapparent.Inthefirst,aftersaying150CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters151thatatruebeliefisonethatwouldsurvivetherigoursofinquiry,hesaysthatbeliefsinethicscanbetrueorfalse.Inthesecond,hesug-geststhatmoraljudgementdrawsonexperience–experiencewhichisnotidenticaltothatwhichisfoundinscience,butexperiencenonetheless.(i)Butwhatelse,whenoneconsidersit,canour‘truth’everamountto,otherthanthewayinwhichpeoplewouldcometothinkifresearchwerecarriedsufficientlyfar?Thatwouldseemtobeallthatourtruthevercanbe.Sogoodmoralsisthekindofhumanbehaviorthatwouldcometobeapprovedifstudiesofrightbehaviorwerecarriedsufficientlyfar.Woulditnotbeagoodideatobeginatext-bookofethics...withthisdefinition:Ethicsisthetheoryofhowtodoasonewouldlikeifonehadconsideredsufficientlythequestionofwhatonewouldfindsatisfactory?(MS673,pp.12–13,1911)2(ii)Ethicsasapositivesciencemustrestonobservedfacts.Butitisquiteadifferentthingtomakeitrestonspecialscientificobser-vation...Theonlysolidfoundationforethicsliesinthosefactsofeverydaylifewhichnoskepticalphilosophereveryetreallycalledinquestion.(CP8.158,1901;seealso1.600,1903)But,asisusuallythecasewithinterpretingPeirce,mattersarenotquitesostraightforward.WhenIfirstsetoutapragmatistaccountofhowmoraljudgementmightbetruth-apt,IsaidthatPeircehimselfhadonlyunhelpfulthingstosayaboutethics(Misak2000:48).Hewasfrequentlykeentoinsistthatin‘vital’matters,whichincludesethicalmatters,3onemusteschewreasoninfavourofinstinct,forinvitalmatters,weneedtoreachadefiniteconclusionpromptly.Science,ontheotherhand,‘hasnothingatstakeonanytemporalventurebutisinpursuitofeternalverities...andlooksuponthispursuit,notastheworkofoneman’slife,butasthatofgenera-tionaftergeneration,indefinitely’(CP5.589,1898).Scienceisthusconcernedwithtruthandethicsisnot.Theflipsideofthepoint,hesuggests,isthat‘reallythewordbeliefisoutofplaceinthevo-cabularyofscience’(CP7.185,1901).Scienceconcernsitselfwitha‘formulareachedintheexistingstateofscientificprogress’–notwithabeliefuponwhichtoact.Peirceappearstoofferushereanextremekindofnoncogni-tivism,wheremattersofethicsdonotfallunderthescopeoftruth,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n152cherylmisakknowledge,andinquiry.Ethicsrequiresquicklyformedbeliefsuponwhichwecanactandhenceisamatterforgut-reaction.Thepreser-vationofthestatusquoseemsinevitable.Indeed,Peirceisclearthatthisview,whichheattimescalls‘sentimentalism’,‘impliesconser-vatism’(CP1.633,1898).Ethics,hesometimessays,isinfactnothingbutasortofcompositephotographoftheconscienceofthemembersofthecommunity.Inshort,itisnothingbutatraditionalstandard,accepted,verywisely,withoutradicalcriticism,butwithasillypretenceofcriticalexamination.(CP1.573,1905)Weseemtohaveheretheviewthatourcognitivistpractices–debatingandreasoningaboutmoralmatters,tryingtoimproveourviews,tryingtoweedoutmistakesandprejudices,etc.–arebasedonanerror,asMackie(1977)wouldsay.IshalltrytoresolvethistensioninPeirce’sworkbysettingoutthebackgroundtothecognitivistthoughtsexpressedinpassages(i)and(ii).MyargumentshallbethatifonelooksatPeirce’sepiste-mology,aboutwhichhedidnotwaver,onecanseethathisremarksaboutinstinctcanbefoldedintothecognitivistview.Thatis,theodd-soundingviewthatethicsmustgooninstinctis,onceweun-derstandtheplaceofinstinctinPeirce’sview,perfectlyconsistentwiththecognitivistview.AndweshallseethatsayingthatbeliefisoutofplaceinscienceisnotthebestwayPeircecouldhaveputhispoint.WhatPeircewasgettingatwhenhemadetheseoddre-marksisthatthescientistmustkeephiseyeonthefalliblenatureofbelief.42.truthandinquiryPeircefamouslyargued(in,forinstance,‘TheFixationofBelief’)thatinquirybeginswiththeirritationofdoubtandendswithastabledoubt-resistantbelief.Ifweweretohaveabeliefwhichwouldalwaysbeimmunetodoubt–whichwouldforeverfitwithexperienceandargument–thenthatbeliefwouldbetrue.Sincewecanneverknowwhenabeliefislikethat,ourbeliefsarefallible.Anyoneofthemmightbeshowntobefalse.Fallibilism,however,doesnotentailthatweoughttofollowDescartesandtrytobringintodoubtallbeliefsaboutwhicherrorisCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters153conceivable.Suchdoubtswouldbe,Peirceargued,‘paper’or‘tin’–notthegenuinearticle.Hesays:...thereisbutonestateofmindfromwhichyoucan‘setout’,namely,theverystateofmindinwhichyouactuallyfindyourselfatthetimeyoudo‘setout’–astateinwhichyouareladenwithanimmensemassofcognitionalreadyformed,ofwhichyoucannotdivestyourselfifyouwould...Doyoucallitdoubtingtowritedownonapieceofpaperthatyoudoubt?Ifso,doubthasnothingtodowithanyseriousbusiness.(CP5.416,1905)Ourbodyofbackgroundbeliefsissusceptibletodoubtonapiece-mealbasis,ifthatdoubtispromptedbysurprisingorrecalcitrantexperience.Wemustregardourbackgroundbeliefsastrue,untilsomesurprisingexperiencethrowsoneorsomegroupofthemintodoubt.Theinquirerisunderacompulsiontobelievejustwhathedoesbelieve...astimegoeson,theman’sbeliefusuallychangesinamannerwhichhecannotresist...thisforcewhichchangesaman’sbeliefinspiteofanyeffortofhismaybe,inallcases,calledagainofexperience.(MS1342,p.2,undated)SoonthePeirceanepistemology,aninquirerhasafallibleback-groundof‘commonsense’beliefwhichisnotinfactindoubt.Onlyagainstsuchabackgroundcanabeliefbeputintodoubtandanew,better,beliefbeadopted.Allourbeliefsarefalliblebuttheydonotcomeintodoubtallatonce.Thosewhichinquiryhasnotthrownintodoubtarestableandweshouldretainthemuntilareasontodoubtarises.Peircelinksthescientificmethodtothisepistemology.Itisthemethodwhichpayscloseattentiontothefactthatbeliefsfalltothesurpriseofrecalcitrantexperience.Inquiry‘isnotstandinguponthebedrockoffact.Itiswalkinguponabog,andcanonlysay,thisgroundseemstoholdforthepresent.HereIwillstaytillitbeginstogiveway’(CP5.589,1998).Acceptedhypothesesandtheories(‘es-tablishedtruths’)arestableandbelieveduntiltheyareupsetbyex-perience.Thescientificmethodisalsothemethodwhichleadstothetruth.Weaimatbeliefswhichwouldbeforeverstable–weaimatgettingthebestbeliefswecan.Wehaveinourvariousinquiriesanddeliber-ationsamultiplicityoflocalaims–empiricaladequacy,coherencewithotherbeliefs,simplicity,explanatorypower,gettingareliableCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n154cherylmisakguidetoaction,fruitfulnessforotherresearch,greaterunderstandingofothers,increasedmaturity,andthelike.Whenwesaythatweaimatthetruth,whatwemeanisthat,wereabeliefreallytosatisfyallofourlocalaimsininquiry,thenthatbeliefwouldbetrue.Thereisnothingoverandabovethefulfillmentofthoseaims,nothingmetaphysical,towhichweaspire.Truthisnotsometranscendental,mysticalthingwhichweaimatforitsownsake.Thisepistemologyanditsaccompanyingviewoftruthisentirelygeneral,despitethefactthatPeircecallsitthemethodofscience.Thatis,whatPeircecalls‘science’isextremelybroad.Anyinquirythataimsatgettingabeliefwhichwouldforeverstanduptoexpe-rienceandargumentabidesbythemethodofscience.WeshallseethatPeircethoughtthatmetaphysics(whenitiswell-conducted)andmathematicsarelegitimateaspirantstotruth.Andsoismoraldeliberation.Hethought,thatis,thatmetaphysics,mathematics,andmoralsmightsatisfyhispragmatistmaxim–themaximthatagenuinebeliefmustbelinkedtoexperience.Ihaveelucidatedelse-whereboththesemanticandtheepistemologicalargumentsinthisthought’sfavour5–inthenextsection,Iwillsimplygestureatthem.3.experience:mathematical,metaphysical,andmoralPeircethoughtthatthemotto‘Donotblockthewayofinquiry’‘deservestobeinscribeduponeverywallofthecityofphilosophy’(CP1.135,1899;seealso7.480,1898).Ahypothesiswhichhadnoconsequences,whichwasseveredfromexperience,wouldbeuselessininquiry.Itwouldbe,asWittgensteinputit,acoguponwhichnothingturned.Investigationintosuchhypothesesisboundtobebarrenandtodirectattentionawayfromworthwhilepursuits.Wecanaccepttheideathatabeliefmustberesponsivetoexpe-riencewithoutcommittingourselvestoanythingasstrongastheverificationismofthelogicalpositivists.ForPeircetakesthekindofexperientialconsequencesrequiredofvariousbeliefstobeverybroadindeed.Perceptionorexperienceisanythingthatisforceduponone.Itgoesfarbeyondwhatourears,eyes,nose,andskinreport:...anythingis,forthepurposesoflogic,tobeclassedunderthespeciesofperceptionwhereinapositivequalitativecontentisforceduponone’sCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters155acknowledgementwithoutanyreasonorpretensiontoreason.Therewillbeawidergenusofthingspartakingofthecharacterofperception,iftherebeanymatterofcognitionwhichexertsaforceuponus....(CP7.623,1903;seealso6.492,1896)Peircetakesanythingthatiscompelling,surprising,brute,orim-pingingtobeanexperience,regardlessofwhatcausesustofeelcompelledandregardlessofwhetherwecanidentifythesourceofthecompulsion:ThecourseoflifehasdevelopedcertaincompulsionsofthoughtwhichwespeakofcollectivelyasExperience.(CP8.101,1900)Experiencejustiswhateverpreventssomeonefrombelievingexactlywhathewantstobelieve–itiswhatkeepsusincheck(MS1342,undated;seealsoMS408,p.146,1893–1895).Peircearguesthattherearetwokindsofexperience–‘ideal’and‘real’.Thelatterissensoryexperienceandtheformerisexperienceinwhich...operationsupondiagrams,whetherexternalorimaginary,taketheplaceoftheexperimentsuponrealthingsthatoneperformsinchemicalandphys-icalresearch.(CP4.530,1905;seealso3.516,1896)Thissortofthoughtexperimentordiagrammaticexperimentorex-perimentintheimaginationis,Peirceargues,thecoreofmathe-maticalanddeductiveinquiry.‘Themathematician,likeeveryotherinquirer,beginsbyaconjecture,whichusuallyisthatacertaintrans-formationofhisicon[diagram]willleadhimto,ortowards,theendofhisinquiry.Hethenperformsthatexperiment....’6Hedrawssubsidiarylinesingeometryormakestransformationsinalgebraicformulaeandthenobservestheresults.Thoseresultsmightbesur-prising,andsincesurpriseistheforceofexperience,suchreasoningisanexperiment.Thissortofexperimentistrulyobservation,yetcertainlyinaverypeculiarsense;andnootherkindofobservationwouldatallanswerthepurposeofmathematics.(CP1.240,1902)Similarly,invaliddeductivereasoning,wearecompelledtoacceptaconclusion–thefactsstatedinthepremissescouldnotbe,ifthefactstatedintheconclusionwerenot.Theconclusionis,inthefirstinstance,irresistable.ItcomesuponthemindbeforeonecancontrolCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n156cherylmisakit.Onlylaterdowecriticallycomparetheconclusiontoournormsandideals.7Peircesometimesarticulateshispointaboutthebreadthofex-periencebysayingthateveryoneinhabitstwoworlds–theinnerandtheouter.Wereactwiththeouterworldthroughtheclashbe-tweenitandoursenses,andwereactwiththeinnerworld–theworldofmathematics,logic,andreasoning–byperformingthoughtexperiments.Inquiry,Peircesays,hastwobranches;oneisinquiryintoOutwardFactbyexperimentationandob-servation,andiscalledInductiveInvestigation;theotherisinquiryintoInnerTruthbyinwardexperimentationandobservationandiscalledMath-ematicalorDeductiveReasoning.8Thedistinctionbetweenthetwodifferentsortsofexperimentsisthattheresultsofdiagrammaticexperimentationexertsacompar-ativelyslightcompulsionuponusandwecanchangetheconstruc-tionofthosediagrams,whereastheouterworldisfullofirresistiblecompulsionsandishardtochange(CP5.474,1907,5.45,1903).Butnonetheless,‘theinnerworldhasitssurprisesforus,sometimes’(CP7.438,1893).Heintendstoleavethedifferencebetweenthetwosortsofexperiencevague:Wenaturallymakeallourdistinctionstooabsolute.Weareaccustomedtospeakofanexternaluniverseandaninnerworldofthought.Buttheyaremerelyvicinitieswithnorealboundarybetweenthem.(CP7.438,1893)PerhapsthecontrastbetweenthetwosortsofexperienceisbestmadebyPeirce’sdistinctionbetweenpracticalandtheoreticalbelief.Inthe1902manuscript‘Reason’sRules’(CP5.538–45),hesaysthatapracticalbeliefsuchas‘anthraciteisaconvenientfuel’willman-ifestitselfinadispositiontobehaveonthepartofthebeliever.Allthingsbeingequal,shewouldsometimesuseanthraciteweresheinneedofafuel.Inaddition,‘sensible’orempiricalconsequencescanbederivedfromthehypothesis.Forinstance,if(ceterisparibus)youweretolightit,itwouldburn.Ontheotherhand,a‘purelytheo-retical’beliefhastodonotwith‘habitsofdeliberateaction’orwithsensibleconsequences,butwith‘expectations’.Asexamplesofthe-oreticalhypothesesPeirceoffers‘thereisanimaginarycirclewhichistwicecutbyeveryrealcircle’and‘thediagonalofasquareisin-commensurablewithitsside’.Ofthelatter,hesaysthatalthoughitCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters157is‘difficulttoseewhatexperientialdifferencetherecanbebetweencommensurableandincommensurablemagnitudes’,therearenev-erthelessexpectations:...abeliefaboutthecommensurabilityofthediagonalrelatestowhatisexpectableforapersondealingwithfractions;althoughitmeansnothingatallinregardtowhatcouldbeexpectedinphysicalmeasurements....(CP5.539,1902)Thepragmaticmaximassertsthatifitisnottobe‘metaphysicaljargonandchatter’,abeliefmusthavealinkwithexperience–itmustissueinexpectationforpracticeortheory.Ifthereisanexpec-tation,thentheunexpectedcansurprisethebeliever.Thedifferencebetweenapracticalandatheoreticalbelief,saysPeirce,isthattheformerinvolvessensationthatis‘muscular’andthelatterinvolvessensationthatisnotmuscular(CP5.540,1902).ThusPeircethoughtthathypothesesinreligionmustissueinexpectations.In‘ANeglectedArgumentfortheRealityofGod’hesetshimselfthetaskofshowinghowthehypothesisofGod’srealitygivesrisetoexpectations.Ineachofthethreedraftsofthepaper,hebreaksoffinfrustration.Eachtimehebeginstotalkabout‘trac-ingoutafewconsequencesofthehypothesis’,heabruptlychangesthesubject.(See,forinstance,MS842,p.127.)Allhecancomeupwithisthatif‘Godisreal’weretruethenwewouldexpecttheretobeatendencytowards‘growthandhabit-taking’andwewouldexpectthatthingswouldbeharmoniousintheworld.9Attheendofthe1910‘Additament’tothepaper,heratherdisingenouslysays‘ThedoctrineoftheEnsnecessariumhasapragmaticistmeaning,althoughIwillnothereattempttosumupthewholeofitsmeaning’(MS844,lastpage;seealsoCP6.491,1910).ClearlymanywrinklesinPeirce’sbrandofpragmatismneedtobeworkedout.Nonetheless,itisclearthatwedonotneedtosay,withthelogicalpositivists,thatonlybeliefsinthephysicalsciencesmeetthepragmatiststandard.Willhypothesesaboutwhatisrightorwrong,orjustorunjust,meetthedemand–cantheybeshowntobesensitivetoexperiencesothattheyarecandidatesforbeliefandfortruth-values?Dotheysetupexpectationswhichcanbemetorunmet?WesawintheintroductionthatPeircesometimesveryclearlysaidthatmoraljudgementsarelinkedtoexperienceor‘ob-servedfacts’.Thesearethe‘observationsofeverydaylife’,observa-tionswhichdonotrequirespecialtrainingorequipment.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n158cherylmisakThosewhoarefamiliarwithPeirce’swritingswillimmediatelyseethatthesethoughtsareremarkablysimilartohisstatementsaboutmetaphysics.Metaphysics,hesays,isthoughttobein-scrutable‘becauseitsobjectsarenotopentoobservation’.Buttheblameforthe‘backwardstate’ofmetaphysicscannotbelaidthere,asmetaphysicsisindeedan‘observationalscience’(CP6.5,1898).It‘reallyrestsonobservations...andtheonlyreasonthatthisisnotuniversallyrecognizedisthatitrestsuponkindsofphenomenawithwhicheveryman’sexperienceissosaturatedthatheusuallypaysnoparticularattentiontothem’(CP6.2,1898).Observationsinthespecialsciencesrequirespecialinstruments,precautions,andskillbecausetheyareremotefromeverydaylife(CP1.242,1902).Otherphenomena,suchasthatwhichmetaphysicsstudies,are‘hardertosee,simplybecausetheysurroundusoneveryhand;weareimmersedinthemandhavenobackgroundagainstwhichtoviewthem’(CP6.562,1905,seealso1.134,1901).Theyarecommonplaceandbanal,buttheyareobservationsnonetheless.TherecertainlyisprimafaciereasontothinkthatPeirceisrightinthinkingthatthepracticeofmoraldeliberationisresponsivetoexperience,broadlyconstrued.Forwhenwedeliberateaboutwhatweoughttodo,wetakeourselvestobesensitivetoreasons,argu-ment,thoughtexperiments,andfirst-personexperience.Wetrytoputourselvesintheshoesofothers,tobroadenourhorizons,tolis-tentotheargumentsoftheotherside.Thatispartofwhatitistomakeamoraldecisionandpartofwhatitistotrytoliveamorallife.Itwouldn’tbeamorallife–itwouldnotbeengagedwiththecom-plexitiesofmoralrequirements–ifwesimplymadeourdecisionsabouthowtotreatothersbyfollowingaoracle,oranastrologer,orthetossofthedice.10Ofcoursewemustbepreparedforthepossibilitythat,asBernardWilliamsthinks,‘ethicalthoughthasnochanceofbeingeverythingitseems’(1985:135).Butthecommitmenttokeepingphilosophyintouchwithexperienceandpracticeissuchthatweshouldnotbetooquicktojumptothisconclusion.4.thetensionbetweenvitalmattersandscienceWenowhaveasketchofhowPeirce’sepistemologymightbefriendlytotheideathatmoraljudgementsarecandidatesfortruth.ButweCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters159haveyettoresolvethetensioninPeirce’swork.HowcanwehandlePeirce’sstatementsthatscience,butnotethics,goesonthehopethat‘thetruthmaybefound,ifnotbyanyoftheactualinquirers,yetultimatelybythosewhocomeafterthemandwhoshallmakeuseoftheirresults’(CP7.54,1902)?Howcanwehandlethethoughtthatinvitalmatters,wedonotaimatgettingtheanswerinthelongrun,butrather,wefollowinstinct,convention,andcommonsenseinordertogetananswerhereandnow?ThereisnousedenyingthatthedistinctionbetweenvitalandscientificmatterswasdeartoPeirce.Butitisfarlessdamagingforthecognitivistpositionthanitfirstappears.OnceweunderstandwhatPeircemeansby‘instinct’,‘experience’,and‘commonsense’andonceweunderstandtheirrolesinwhathecallsscientificinquiry,wecanseethatvitalmattersareindeedmattersforscientificinquiry.Peircebuildsinstinctintothescientificmethod–themethodofabduction,deduction,andinduction.Abductionisamatterofcom-ingupwith(Peircesometimessays‘guessingat’)anexplanationforasurprisingexperience.Onceabductionhasprovidedahypothesis,onededucesconsequencesfromitandtestsitbyinduction.Abduc-tionprovidessciencewithnewideasandthusscienceadvancesby‘thespontaneousconjecturesofinstinctivereason’(CP6.475,1908,5.604,1901).Thatis,whenasurprisingphenomenonneedsexpla-nation,instinctplaysacentralrole.Itprovidesthefalliblestartingpointsofthescientificmethod–thehypotheseswhoseconsequencesarethentestedbyinduction.Thatisonewayinwhichinstinct,ratherthanbeingsetagainstscienceorinquirywhichisaimedattruth,isapartofit.Anotherwayinwhichinstinctisapartofscienceisasfollows:whenonefactputsapersoninmindofanother,butrelated,fact,andonconsideringthetwotogether,hesaystohimself‘Hah!Thenthisthirdisafact’,...itisbyinstinctthathedrawstheinference.(MS682p.19,1913)Ifyoufeelthataninferenceiscorrect,thatfeelingisan‘instinct’andisverymuchlikethefeelingthatsomethingisred.Youhavenoreasontoacceptthejudgement–itjustcomesuponyou.Itisthusakindofexperienceorperception.Thisisnottosaythatthissortofinstinctisinfallible–itisasfallibleasallkindsofexperience(SeeCP1.404,1890).Peircealsomakesthiskindofpointbysayingthatinstinctisalignedtoourhabitsofreasoning:inPeirceanterminology,ourCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n160cherylmisaklogicautens.Reasoning‘istheprincipalofhumanintellectualin-stincts...reasoningpowerisrelatedtohumannatureverymuchasthewonderfulinstinctsofants,wasps,etc.arerelatedtotheirsev-eralnatures’(MS682p.8–9).Ourinstinctiveandhabitualcognitiveskills,asHookway(2000:255)putsit,guideourinquiries.Ofcourse,thesehabitscanbeflawed,butwenonethelessrelyuponthemuntiltheyareshowntobeflawed–untilwehaveevidencethattheyleadusastrayoruntilwecanexplainwhatiswrongwiththem.Ifwearetocontinuetoinquire,wemustassumethatourstockofhabitualcognitiveskillsisreliable.Peirceiscrystalclearthatsomething’sbeingsucharegulativeassumptionofinquirydoesnotmeanthatitistrue.Butsomething’sbeingaregulativeassumptionofinquirydoesentailthatweshouldbelieveitandthatweshouldconstructourphilosophyinsuchawayastomakeroomforitstruth.Andfinally,instinctis,forPeirce,alsoalignedwiththatwhichisnotdoubted–thatwhichformsour‘commonsense’fallibleback-groundbodyofbeliefs.Writinginanentirelygeneralwayaboutbeliefandinquiry,Peircesays‘...thepragmatistwillacceptwholesaletheentirebodyofgenuineinstinctivebeliefswithoutanyshadeofdoubt,tossingasidethetoydoubtsofthemetaphysicianasunworthyofamaturemind’.11ThatiswhatPeircemeanswhenhesaysthatinethicswemustgooninstinctoronthestatusquo.Evenifyouthink,generally,thattrustinginstinctis‘treacherousanddeceptive’,ifyoudon’tdoubtsomethingandhaveneverdoubtedit,youwillbelieveit.Thus:‘thatwhichinstinctabsolutelyrequireshimtobelieve,hemustandwillbelieveitwithhiswholeheart’.Ifsomethingseemsperfectlyevident,youcantryasyouwilltocriticiseit,butyouwilleventuallybeobligedtobelieveit.By‘commonsense’and‘instinct’,Peircemeans‘thoseideasandbeliefsthataman’ssituationabsolutelyforcesuponhim’.(CP1.129,1905)Instinctisthatwhichwehavenochoicebuttorelyupon.Instinctisjustwhatthewholeofpastexperiencehasputintoplace.InPeirce’sCambridgeLecturesof1898,oneoftheplacesinwhichtheproblematicgapbetweenscienceandvitalmattersismoststark,wehavethisthought:Wedonotsaythatsentimentisnevertobeinfluencedbyreason,northatundernocircumstanceswouldweadvocateradicalreforms.WeonlysaythatCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters161themanwhowouldallowhisreligiouslifetobewoundedbyanysuddenacceptanceofaphilosophyofreligionorwhowouldprecipitatelychangehiscodeofmorals,atthedictateofaphilosophyofethics,–whowould,letussay,hastilypracticeincest,–isamanwhomweshouldconsiderunwise.TheregnantsystemofsexualrulesisaninstinctiveorSentimentalinductionsummarizingtheexperienceofallourrace.(CP6.633/RLT:111)Parker(1998:50)notesthatwealsofindthesethoughtsinJames’swork.12In‘TheMoralPhilosopherandtheMoralLife’,Jamesex-tendsPeirce’sviewoftruthtoethics,saying‘...therecanbenofinaltruthinethicsanymorethaninphysics,untilthelastmanhashadhisexperienceandhissay’(1897a[1979]:184).Hearguesthatsoci-etymaybeseenasalong-runningexperimentaimedatidentifyingthebestkindofconduct.Itsconventionsthusdeserverespect.Ourbackgroundbeliefs,whileremainingfallible,capturetheexperienceofgenerations(1897a[1979]:206).Jamesthinksthat‘ethicalscienceisjustlikephysicalscience,andinsteadofbeingdeducibleallatoncefromabstractprinciples,mustsimplybideitstime,andbereadytoreviseitsconclusionsfromdaytoday’(1897a[1979]:208).PeirceisinagreementwithJameshere.Theethicaldeliberatormightbehesitanttoreviseherbeliefsandthishesitationcanbejustified.Butitisnotalwaysjustified:Likeanyotherfield,morethananyother[morality]needsimprovement,advance....Butmorality,doctrinaireconservatistthatitis,destroysitsownvitalitybyresistingchange,andpositivelyinsisting,Thisiseternallyright:Thatiseternallywrong.(CP2.198,1902)ForbothPeirceandJames,moraljudgementsareconnectedtoex-perienceinthewaythatallofourgenuinejudgementsare:‘justasreasoningspringsfromexperience,sothedevelopmentofsentimentarisesfromthesoul’sInwardandOutwardExperiences’(CP1.648,1898).Aswitheveryotherkindofexperience,‘[t]hatitisabstractlyandabsolutelyinfalliblewedonotpretend;butthatitispracticallyinfalliblefortheindividual–whichistheonlyclearsensetheword“infallibility”willbear–inthatheoughttoobeyitandnothisindividualreason,thatwedomaintain’(CP1.633,1898).Wehaveseenthat,forPeirce,thisholdsforanydomainofinquiry:wetakeourbodyofbackgroundbelieftobepracticallyinfallible,untilthecourseofexperienceweighsinagainstit.Instinctualandcommonsensebeliefsaresubjecttorevision,buttheyareheldfirmCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n162cherylmisakuntilexperiencepromptsthatrevision(CP5.444,1905).ThatisthePeirceIwanttofocusupon.Ethicsandscienceareinthesameboat,relyingondeeplyheld,butrevisable,backgroundbeliefsandhabits.Instincthasapositiveandessentialroletoplayinscienceandinmorality.13Enoughsaidabouthowrelianceoninstinctdoesnotdistinguishethicalmattersfromothermatters.Nowweneedtofocusontheotheraspectofthepoint–thecontentiousviewofscience,inwhichbeliefisoutofplace.Letuslookatthesourceofthetalkthat,whilethescientistcanwaitforananswerandthusdoesnotbelievehistheories,thedeliberatorinethicsneedsanimmediateanswerandthushasbeliefsonwhichtoact.OnesourceisPeirce’s1902appli-cationtotheCarnegieInstitute,pleadingforfundssothathecouldwritehisgrandworkonlogic.Therearemanydraftsofthisapplica-tioninthePeircePapersandsomeofthesedraftsshowveryclearlythatPeircedidnothaveasettledviewaboutthematterinquestion.Perhapshisdoubtsarebestexpressedonp.54ofsomeofthedrafts,wherewehavehimsayingthatthescientistisinabind–a‘doubleposition’:Asaunitofthescientificworld,withwhichheinsomemeasureidentifieshimself,hecanwaitfivecenturies,ifneedbe,beforehedecidesupontheacceptabilityofacertainhypothesis.Butasengagedintheinvestigationwhichitishisdutydiligentlytopursue,hemustbereadythenextmorningtogoonthathypothesisorrejectit...heoughttobeinadoublestateofmindaboutthehypothesis,atonceardentinhisbeliefthatsoitmustbe,andyetnotcommittinghimselffurtherthantodohisbesttotrytheexperiment.14Whatawonderfulstatementoftheproblem.Theinquirer(anyin-quirer)mustbereadytobelieveandtoactonthebelief,knowingfullwellthatitmightnotbetrue.Beliefisnotoutofplaceinscience–itisjusttemperedbyfallibilism.Thescientistmustbelieve,butbeconstantlyawarethatherbeliefmightbeoverturned.Thisisaper-fectstatementofthetrickypathonwhichthecriticalcommonsensephilosophermusttread.15Similarly,inthe1898‘DetachedIdeasonVitallyImportantTop-ics’,itturnsoutthatPeirce’sstatementthatthereisnobeliefinscienceisnotasalarmingasitfirstseems.Itisherethatwefindthenicemetaphorthatscienceiswalkinguponabog.Thereasonitcanonlysay‘thisgroundseemstoholdforthepresent.HereIwillstayCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters163tillitbeginstogiveway’isthatsciencealwaysstartswithanabduc-tiveinference.ForPeirce,theconclusionsofabductiveinferencesaren’ttobebelieved:theyaremereconjectures.ButAfterawhile,asScienceprogresses,itcomesuponuponmoresolidground.Itisnowentitledtoreflect:thisgroundhasheldalongtimewithoutshowingsignsofyielding.Imayhopethatitwillcontinuetoholdforagreatwhilelonger(CP5.589).Wecannowusethehypothesisorconjectureinpractice–wecanactonit.Foritnolongerrestsuponamereabduction.Ithasbeeninductivelysupported.Peircesays‘Inotherwordsthereisnowreasontobelieveinthetheory,forbeliefiswillingnesstoriskagreatdealuponaproposition’.Thescientist,however,willstillbeconcernedaboutwhetherinfactthetheorywillcontinuetosurvivethetrialsofinduction:hewillkeephiseyeonwhetherthehypothesiswillinthelongrunsurvivethetrialsofinquiry.Nonetheless:‘Wecalltheminscienceestablishedtruths,thatis,theyarepropositionsintowhichtheeconomyofendeavorprescribesthat,forthetimebeing,furtherinquiryshallcease’(CP5.589).AnothersourceofthecontentiousviewisPeirce’sCambridgeLec-turesof1898(RLT).Theselecturesarenotthebestplacefordiscern-ingPeirce’sconsideredviewaboutscienceandvitalmatters.HewasextremelyirritatedatJames,whohadcharitablysetupthelecturessothatPeircemightquiteliterallybeabletoputabitoffoodonhisplate.UponlearningthatPeirceintendedtoaddresstechnicalquestionsoflogic,Jamesaskedhimto‘beagoodboyandthinkamorepopularplanout’.Perhapshecouldratherspeakabout‘sepa-ratetopicsofavitallyimportantcharacter’.16Peirce,strugglingnodoubtwiththeshameofhavingtoberescuedbyJamesandhavingbeenshutoutofanacademicjobbyHarvard,poursscornontheHarvardphilosophersfortheirlackoftraininginlogicandsarcasti-callysaysthathewillindeedrestricthimselfto‘vitalmatters’.Thedraftsofthelecturesaremorescathingthanthelecturesactuallydelivered,showingthatPeircethoughtheoughttotrytoholdhisangerincheck,buthedoesnotaltogethermanageit.Itisinthiscontextthathemakestheextremeremarksaboutreasoningbeingoutofplacewithrespecttovitallyimportanttopics.Reasoning,hesneers,seemsnottobenecessaryforworldlysuccess.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n164cherylmisakTheseremarkssimplycannotbetakenseriouslyonceitisseenthatPeircewaswoundedaboutbeingtold,insuchanoffensiveway,toexcisethehardreasoningandlogicfromhislectures.AfterJameshaddied,Peirceclearlyfeltbadabouthavingbeensorude.Unwillingtogiveuponthepointthathisdearfriendhadan‘almostunexampledincapacityformathematicalthought’,henonethelesspromisestoendeavour‘tosubstituteaseriousandcourteous’tonefor‘thetoneIusedatHarvard’(CP6.182,1911).Andfinally,intheCambridgeLectures,wefindPeircesee-sawinginthesamebreathbetweentheideathatbeliefhasnoplaceinscienceandtheideathatitdoes.Firsthesays:Iwouldnotallowtosentimentorinstinctanyweightwhatsoeverintheoret-icalmatters,nottheslightest....True,wearedrivenoftentimesinsciencetotrythesuggestionsofinstinct;butweonlytrythem,wecomparethemwithexperience,weholdourselvesreadytothrowthemoverboardatamoment’snoticefromexperience.(CP1.634/RLT:111)Thisisthe‘nobeliefinscience’sideofthesee-saw.Wearenotreadytoactonbeliefinscience.Sciencemerelywritesinthelistofpremissesitproposestouse.Nothingisvitalforscience;nothingcanbe.Itsacceptedpropositions,therefore,arebutopin-ionsatmost;andthewholelistisprovisional.Thescientificmanisnotintheleastweddedtohisconclusions.Herisksnothinguponthem.Hestandsreadytoabandononeorallassoonasexperienceopposesthem.(CP1.635/RLT:112)Butinthenextbreath,Peircesaysthatsomeofthescientist’sconclusionsarecalled‘establishedtruths’–‘propositionstowhichnocompetentmantodaydemurs’(CP1.635/RLT:111).Establishedtruthsarethebackgroundbeliefswhichwetakeforgranted–thebeliefsagainstwhichwejudgenewhypotheses.Theyarewhatthecriticalcommonsensephilosopherfocussesupon.Peircedoesindeedthinkthatbeliefhasaplaceinscience.Ifwehavetochoose,ontheirownmerits,whichofPeirce’sop-posingviewsofsciencetoaccept,thechoiceiseasy.AsDuhem,Quine,andKuhnhavegonesofartoshowus,noscientifictheoryisoverthrowninaflashbyaloneexperience.Scientiststendtoin-sulatetheirtheoriesfromrogueexperiencesunlessthetheorycanbearsuchinsulationnomore.ItwouldbeanoddscientistindeedCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters165whoabandonedawell-supportedtheoryonthebasisofonecontra-dictoryexperiment.Peircejustmakesamistakehere–onethathasirritatedmanyacontemporarypragmatist.17Heattimesfailstoseehowthebackgroundtheoriesofscientistsareacceptedastrueuntilrecalcitrantexperienceoverwhelmsthem.Butofcourse,attimesheseesthisveryclearly.IfwedropPeirce’scontentiousthoughtaboutscience,wecandiscernacoherentandverysensibleposition.Inbothscientificandmoralmatters,wehavecherishedbeliefswhicharenonethe-lessresponsivetoexperience.Inethics,asinscience,weactonourexperience-drivenbackgroundbeliefs,whilerealisingthattheymightyetbeoverthrownbyfurtherexperience.5.amoreproductivetensionIhavearguedthatPeirce’sthoughtsaboutethicsandsciencecanbebroughtintoharmony–thatweoughttoseePeirceasholdingthatvitalmattersfallunderthescopeoftruth,knowledge,andwhathecallsthescientificmethodofinquiry.Asinscience,wehopeorassumethattherewillbeanupshottoourmoraldeliberations.ButthereisanothertensioninPeirce’sthoughtaboutscienceandethicswhichwewoulddowelltoleaveinplace.Foranycognitivistpositionwhichfailstoincorporatethetensionis,Ipropose,simpleminded.AtonepointPeircedistinguishesdisagreementinmoralmattersfromdisagreementabouttaste:‘Howeveritmaybeabouttaste,inregardtomorals,wecanseegroundforhopethatdebatewillul-timatelycauseonepartyorbothtomodifytheirsentimentsuptocompleteaccord’(CP2.151,1902).ThatisthecognitivistthoughtwhichIhavebeentryingtopreserveforPeirce.Buthethendeliverstheapparentlyanticognitivistthoughtthat‘Shoulditturnoutoth-erwise,whatcanbesaidexceptthatsomemenhaveoneaimandsomeanother?Itwouldbemonstrousforeitherpartytopronouncethemoraljudgmentsoftheothertobebad.Thatwouldimplyanappealtosomeothertribunal’(CP2.151,1902).Atfirstglance,thislookslikeastraightforwardcontradiction.ButPeircehereisrehearsinghisrathersubtlepositiononbivalence–apositionwhichisespeciallysuitedtomoraljudgement.Hethoughtthattheprincipleofbivalence–thateverystatementiseithertrueorfalse–isnotalawoflogic,butaregulativeassumptionofinquiry.IfitCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n166cherylmisakwerethecasethat,nomatterhowfarweweretopushourinquiries,therewouldbenoupshottothequestionathand,thenwemustsaythatthereisnotruthofthematteratstake.Thehopeorregulativeassumptionofinquiry–thatourinquirieswillhaveanupshot–wouldherenotbefulfilled.Sowefrequentlyhavehimsaying,inanentirelygeneralfashion:Itistruethatwecannotknowforcertainthatexperience,howeverlongandfull,everwouldbringallmentothesamewayofthinkingconcerningthesubjectofinquiry.Butthatistheonlyresultthatcansatisfyus,sothatwemustforevercontinueinthehopethatitwillcome,atlast.18Wemusthope,foranyquestionintowhichweinquire,thatbivalencewillhold.Andhereisthepointtowhichanycognitivistmustbealert:weexpectthatbivalencewillfailmoreofteninmoralinquirythaninchemistryandlessoftenthaninmattersoftaste.Aparticularlyhelpfultextregardingbivalenceandmoralsis‘TruthandFalsityandError’.HerePeirceconsidersthepossibilitythatforsomequestions,noanswerwouldbeforthcoming,nomatterhowlongthediscussionweretogoonandnomatterhowadvancedourmethodsofinquiryweretobecome.Perhapsthequestionofwhetherthereisfreewillislikethat:Theninregardtothatquestion,therecertainlyisnotruth.Butwhetherornottherewouldbeperhapsanyrealityisaquestionforthemetaphysi-cian...Evenifthemetaphysiciandecidesthatwherethereisnotruththereisnoreality,stillthedistinctionbetweenthecharacteroftruthandthecharacterofrealityisplainanddefinable.(CP5.565,1901)Afterdrawingthedistinctionbetweentruthandreality,Peirceverycarefullysaysthatitholdsnotjustforscience,butalsoforethics(CP5.566,1901)andformathematics(CP5.567,1901).Alloftheseinquiriesaimatthetruth:Nowthedifferentsciencesdealwithdifferentkindsoftruth;mathematicaltruthisonething,ethicaltruthisanother,theactuallyexistingstateoftheuniverseisathird;butallthosedifferentconceptionshaveincommonsomethingverymarkedandclear.Weallhopethatthedifferentscientificinquiriesinwhichweareseverallyengagedaregoingultimatelytoleadtosomedefinitelyestablishedconclusion,whichconclusionweendeavourtoanticipateinsomemeasure.AgreementwiththatultimatepropositionthatCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters167welookforwardto,–agreementwiththat,whateveritmayturnouttobe,isthescientifictruth.(CP7.187,1901)Therewillofcoursebedifferencesbetweenkindsofinquiry:themathematician,thechemist,andtheinquirerintowhatisthemorallyrightthingtodowillnotuseidenticalmethods.Norwilltheyfindthattheiraspirationshaveidenticalprospects.Norwilltheyallbetalkingaboutthesamesortofreality.AsHookwayputsit:Wemightagreethatmathematicalpropositions,ethicalpropositions,propo-sitionsfromthemoretheoreticalreachesofsciencecanallbeassessedastrueorfalse.Each,wemightsuppose,canbetestedor‘comparedwithre-ality’.Thismightinvolvelookingforaproof,consideringhowtheethicalpropositionwouldappealtosomeonewhotookupadistinctivedisinterestedviewpointonthings,ormakingexplanatoryinferencesaboutwhatbestsys-tematisesourothertheoreticalbeliefsandexperimentalresults.(Hookway2000:97)19Thatis,comparinghypotheseswith‘reality’isboundtotakediffer-entformsindifferentinquiries.AndPeirceisarealistaboutkindsofrealitythatarenotphysical.PeircegoesontomakethepointthatIhavebeenstressing:whatiscentralinthesevariousinquiriesisthesurpriseofexperience,againstabackgroundofstableexpectationorbelief.Hesays:‘Thusitisthatallknowledgebeginsbythediscoverythattherehasbeenanerroneousexpectation....Eachbranchofsciencebeginswithanewphenomenonwhichviolatesa[n]...expectation’.(CP7.188,1901)Letuslookatthemathematicalcase,aboutwhichPeirceisex-ceptionallyclear.In‘TruthandFalsityandError’,hesaysthat[t]hepuremathematiciandealsexclusivelywithhypotheses.Whetherornotthereisanycorrespondingrealthing,hedoesnotcare.Hishypothesesarecreaturesofhisownimagination;buthediscoversinthemrelationswhichsurprisehimsometimes.Ametaphysicianmayholdthatthisveryforcinguponthemathematician’sacceptanceofpropositionsforwhichhewasnotprepared,proves,orevenconstitutes,amodeofbeingindependentofthemathematician’sthought,andsoareality.Butwhetherthereisanyrealityornot,thetruthofthepuremathematicalpropositionisconstitutedbytheimpossibilityofeverfindingacaseinwhichitfails.(CP5.567,1901)CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n168cherylmisakPeircearguedthatmathematicsdoesnotanswertoaphysicalreal-ity–itisnotconcernedwithphysicalobjects,butwithpossibili-ties(CP4.234,1902,CP3.527,1896)ortheformsofrelations(CP4.530,1905).Peirceofcoursethinksthatrealitydoesgobeyondthephysical–generalsandpotentialities,forinstance,arereal.But,ashesays,thatisafurtherquestionforthemetaphysician–itgoesbeyondthebasicpragmatistelucidationoftruth.Whateveryourmetaphysicsmightbe,mathematicsaimsatthetruthinthesamewayotherkindsofinquiryaimatthetruth.Atruebeliefwouldbethebestbelief,werewetopursueourinquiriesasfarastheycouldfruitfullygo,andwhatmakesabeliefbestinmathematicsmightdifferfromwhatmakesabeliefbestinscience,orinmorals.‘TruthandFalsityandError’isthusawonderfultextforthecogni-tivistpragmatist.WehavePeircesayingthatabeliefcanbesensitivetoexperienceevenifthereisnounderlyingphysicalreality.Experi-ence,forPeirce,justisasurprise.Perhapsadomainofinquirywhichrestsonanunderlyingphysicalrealitywillhavemorestatementswhicharebivalent.Butsomekindsofinquiry,suchasmathematics,willbefullofbivalentstatementsandyettheyaresuchthatthereisnounderlyingphysicalreality.Moralityisanespeciallyinterestingdomainofinquirywithre-specttothesequestions.Inyetanothergrandproposedbookonlogic,wherePeirceisoutlininghisviewinasystematicway,hesayswhatsurelyanycognitivistmustaccept:moralityissomehowbothsub-jectiveandobjective.Moralityarisesfromhumanpredicamentsandhistory–it‘hasitsroots’in‘humannature’(CP2.156,1902).Yetitissuchthatweaimtogetarightanswer.Unliketaste,whichseemstobemostlysubjective,morals‘hasasubjectiveandanobjectiveside’(CP2.153,1902).Thereisacontinuumhere:‘taste,morality,ratio-nality,formatruesequenceinthisorder’,withtastebeing‘purelysubjective’.Peirce’selaborationofthisthoughttakesusfurtherintothesubtlemovesrequiredofmoralcognitivism.Itoftenhappens,hesays,thatamanseriouslyconsiderswhathisdutyisinacertaincase,butthendrawsamistakenconclusion:itis‘quitethereverseofthatwhichhewouldreachifcertainaspectsofthecasehadnotescapedhim’.Yetheisrighttodowhatseemstohimtobehisduty,despitethefactthatheismistaken.ItwouldbeveryoddtosuggestthatheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters169notfollow‘thedictatesofhisconscience’.Hemustactandhemustactonwhat,aftercarefulconsideration,hetakestobehisduty.Ifhe‘carriedhisself-discussionfurther’,hemighthavediscoveredthetruthaboutwhathisdutywas.Buthecannotbutdowhathethinksisright.Thereisarightanswertothequestion‘whatismyduty?’evenifIdonotreachit.‘Itseemsrighttome’comesapartfrom‘itisright’,asitmustinanyobjectivematter.Asinscience,wegoonthebestbeliefs,giventheavailableevidenceandargument,knowingthatfurtherevidenceandargumentmightproveusmistaken.ThenPeircecallsforacorrectioninmoralphilosophy:‘Itistruethatthemajorityofwritersonethicsinthepasthavemadetherootofmoralssubjective;butthatbestopinionisveryplainlymovingintheoppositedirection’(CP2.156,1902).Itcannotmovetoofarintheobjectivedirection,fortheonlygroundwehaveforourmoraljudgementsisfeeling:‘ouraversionforandhorrorof’incest‘issim-plyfelt’(CP2.171,1902).Thekindofexperiencerelevanttomoraljudgementislikelytobemuchlessuniformthanthekindofexpe-rienceonefindsrelevanttobeliefinlogicandinscience.Peircethuspresentsuswithanextremelysophisticatedcogni-tivism.Weoughttoexpectbivalencetofailmoreofteninethicsthaninphysicalscience,butmoraldeliberationnonethelessaimsatthetruth.Wecanseethattherealitytowhichethicaljudgementsfitisnotphysicalreality,yetethicaldeliberationisstillguidedbythesurpriseofexperience.Ethicsfallssomewhereinbetweenthehighlysubjectivedomainoftasteandthemuchmoreobjectivedo-mainofthephysicalsciences.WehavemovedveryfarindeedfromthethoughtthatPeircehadonlysillythingstosayaboutethics.6.thenormativesciencesandtheforceofexperienceOntheoutlineofPeirce’sviewwhichwehaveinhand,itwouldseemthatexperience,notreligion,notsomesetofphilosophicalprinciplesdeliveredtousbyutilitarianismorKantianism,isthelifebloodofmoraldeliberation.Thepragmatistcognitivistshould,byandlarge,stayawayfromatheoryofmoralityandtakeinsightfromwhereveritistobefound(SeeMisak2000:122f).ManyaPeircescholarwillbewonderingatthisjunctionwhyIhavebeensilentaboutwhatappearstobePeirce’stheoryofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n170cherylmisakmorality–hisoutlineofthenormativesciences.SoIwillconcludebyverybrieflylookingathowthefocusonexperiencefitswiththislateaspectofPeirce’sthought.Peircemakestheboldclaimthatlogicisdependentonethicsandethicsisdependentonaesthetics.Theclaim,onceunpacked,lookslessstartlingthanitdoesatfirst.Wedonothaveherethebizarrethesisthatlogicisbasedonmoralityandthatmoralityisbasedonart.Peircethoughtthatthenormativesciences(aesthetics,ethics,andlogic),althoughtheyhaveoftenbeenmistakenforpracticalsciences,arereallytheoretical.Theystudywhatoughttobe,notwhatwillbeorwhatis(CP1.281,1902,CP2.156,1902).Allthreepronouncesomethingsgoodandsomethingsbad.Allthreestudyformsofvoluntary,self-controlledconduct,aimedatanidealorend.Theysetoutruleswhichoughttobefollowedifouraimsaretobeattained.Theyinvestigatethe‘lawsoftherelationofPhenomenatoEnds,thatis,perhaps,toTruth,Right,andBeauty’(CP5.121,1903).Aestheticsaskswhatispossibletoadmireunconditionally.Thisisofcoursenothowweusuallythinkofaesthetics.Thescienceofaesthetics,Peircethinks,‘hasbeenhandicappedbythedefinitionofitasthetheoryofbeauty’(CP2.199,1902).Itislimitedtomat-tersoftasteonlyifweincludeundertheumbrellaoftaste‘formingatasteinbonnets...or...apreferencebetweenelectrocutionanddecapitation,orbetweensupportingone’sfamilybyagricultureorbyhighwayrobbery’(CP1.574,1905).Andinformingtheseprefer-ences,self-controlandcriticismisimportant.Onedoesn’tjustgoonone’swhims–onegoesonone’sconsideredanddisciplinedfeelings,experiences,andthoughts.Ethics,hesays,isbasedonaesthetics,aswecannotknowhowweshouldaimtobehaveorknowwhatispossibletoadoptasanulti-mateenduntilweknowwhatismostadmirable(CP5.130,1903).Andlogicisbasedonethics.Logicisnormative:itisthoughtwhichisunderself-control(CP1.606).‘Thinkingisakindofaction,andreasoningisakindofdeliberateaction;andtocallanargumentil-logical,orapropositionfalse,isaspecialkindofmoraljudgment’(CP8.191,1904).20SowhatdoesPeircethinkthatwecanadmireunconditionally?Hisratherunhelpfulansweris‘concretereasonableness’–reasonandthegrowthofreason(CP1.615,1903).ThatisthesummumCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters171bonumorthehighestgoodtowhichallofouractions,intentions,andprojectsmustanswer.WhenPeirceaskshowself-controlmanifestsitselfinethics,hesaysexactlywhatonewouldexpectofsomeonewhoholdsthatatruebeliefisonewhichbestaccountsforexperience,broadlyconceived.Hesaysthatwecriticiseourownconduct,comparethatconducttoastandard,askwhetherouractionsaccordwithourintentions,askwhetherwearesatisfiedordissatisfiedwithouractions,absorblessons,andreviewourideals(CP1.591–9,1903).‘Theexperienceoflifeiscontinuallycontributinginstancesmoreorlessilluminative’(CP1.599).WearethusquicklyreturnedtotheparallelbetweenPeirce’sviewofethicsandhisgeneralepistemology:Justasconductcontrolledbyethicalreasontendstowardfixingcertainhabitsofconduct,thenatureofwhich...doesnotdependuponanyacci-dentalcircumstances,andinthatsensemaybesaidtobedestined;so,thought,controlledbyarationalexperimentallogic,tendstothefixationofcertainopinions,equallydestined,thenatureofwhichwillbethesameintheend,howevertheperversityofthoughtofwholegenerationsmaycausethepostponementoftheultimatefixation.(CP5.430,1905)HereagainwehaveanexcellentstatementofPeirce’scognitivism.Whenweinquireabouthowweoughttocontrolourconduct,weare‘destined’toreachthetruth.Thatis,ourreachingapermanentlysettledbeliefdoesnotdependonaccident–wearedestinedtoreachthetruthinthesensethatexperienceandargumentwould,wehope,leadtoabeliefwhichwouldnotbeoverturned.ThisisathoughtwhichappearsagainandagaininPeirce’sepistemologyandtheoryoftruth.Wemaygobadlywrongforgenerations,butwehopethatthereisananswertothequestionathandwhichwouldfitwithalltheevidenceandargument,werewetohavesomuchofitthatfurtherinquirywouldbefruitless.Thebeautyoftheabovepassageisbothinmakingclearthemodestsenseof‘destined’andalsoindrawingtheexplicitparallelbetweenethicsandotherkindsofde-liberation.Inethicsweaimatgettingananswerwhichwouldsat-isfyouraimsininquiry:whichwouldforevermeetthechallengesofreasons,argument,andevidence.Peirce’sviewoftruthisindeedafriendofcognitivism,despitecertainofhisassertionsaboutvitalmatters.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n172cherylmisaknotes1.ThispaperhasbeenimprovedbycommentsfromThomasShort.2.Thismanuscript,titled‘ASketchofLogicalCritic’isagem,asitalsocontainsastatementofconsiderableimportanceforPeirce’sviewofbelief:‘Butitisoneoftheessentialsofbelief,withoutwhichitwouldnotbebelief,thatitbringspeaceofmind,oratleastrelieffromthestruggleofdoubt;sothatamancouldhardlybeconsideredsanewhoshouldwishthat,thoughthefactsshouldremainlamentable,heshouldbelievethemtobesuchashewouldwishthemtobe’(p.11;seealsothevariantMS675,p.8).SeeMisak(1991:59ff)andWiggins(1999)forthesignificanceofthisthought.3.AsChrisHookwaypointedouttome,avitalmatter,forPeirce,isanyurgentquestionaboutwhatweoughttodo.Thecategoryofthevitaliswiderthanthecategoryoftheethical.4.Theresolutionofferedhereis,inessence,theresolutionofferedinnote12ofChapter2ofMisak(1991).5.Misak(1995),especiallypp.59ff,97–127,152–62,171–8.6.MS328,p.43,1905.SeealsoCP3.363,1885;seealso4.233,1902,1.322,1903,5.162,1903,6.568,1905.7.MS453,loosesheets,1903.SeealsoCP2.96,1902,6.497,1906.8.MS408,pg.150,1893–1895;seealsoCP3.527,1896.9.CP6.490,MS842,p.16,MS843,unmarkedpage,105pagesfromtheendofthemanuscript.Hookway(2000:273f)suggeststhatPeircebackedofffromtheideathatonemustfindevidenceforGod’sreality,suggestingthatthemodelofabduction,deduction,andinductionisbeingstretchedtofitthereligiousquestion.IthinkthatPeircejustcouldn’tfindsuchevidenceandonlythendidhebackofffromtheideathathemustfindit.See(CP1.91,1896)fortheclaimthatwhetherprayersareefficaciousisaquestion‘opentoexperimentalinquiry’.10.SeeMisak(2000)forasustaineddiscussionoftheseissues.11.MS329p.12,1904.SeealsoCP5.445,1905andShort(2001)fortheideathatcommonsensebeliefsareofthenatureofinstincts.Hook-waypointstoasimilarconnectionbetweencommonsensebackgroundbeliefsandvitalmatters:‘Webeginwithafolkphysicsandafolkpsy-chologyandacommonsenseviewofmorality...’(2000:198,205).ItisimportanttoseethatoncethePeircescholarthinksofinstinctasfallingwithinPeirce’scriticalcommonsensism,mycentralthesisfol-lows.Thatis,thefactthatPeircethinksthatvitalmattersmustbedrivenbyinstinctdoesnotmakevitalmattersspecial.ForPeircetakesallofourinquiriestorelyonabackgroundof‘instinctive’undoubtedbelief–beliefuponwhichweact,untilexperiencepromptsustoreviseit.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nC.S.PeirceonVitalMatters17312.Itis,however,notclearwhether,intheend,Jamescanbedescribedasacognitivistaboutanykindofbelief.13.AsTrammell(1972)tellsus,Peircedidattimesinnocuouslyopposereasonandvitalconcerns.First,themoreemotionallycommittedoneistoavitalbelief,theharderitistoreasoninanunbiasedwayaboutit.Healsosometimessuggeststhatinductivereasoningrequiresakindofdetachmentfromimmediateconcerns,asitdependsonthenotionofalongrun.Andsometimeshesuggeststhathighlytheoreticalortechnicalscienceissuchthatinstinctislessreliablethanitisinlessabstractandlesstechnicalinquiry.Thesepointsareofcoursefinebyme–myargumentisthatPeircedidnotflat-outsetinstinctagainstreason,experience,andinquiry.Rather,heallowedaroleforinstinctinallkindsofreasoning.Instinctisnotoutofplaceinscienceanditisnotoutofplaceinmorals.Bothaimatthetruth,usingthemethodofexperienceandreasoning.(Hookway(2000:228),bytheway,notesthatthesharpdistinctionbetweentheoryandpracticewasconfinedtothe1880s.)14.MSL75,p.53–5ofthefirst88-pagevariant.15.SeeMisak(1991:50ff)foramoredetaileddescriptionofthispath.16.Trammell(1972)presentsanexcellentaccountofthisdispute.SeealsoHookway(2000:23f).17.See,forinstance,Levi(1983),(1984)andHookway(2000:210).18.MS1342,p.2ofvariants,undated.SeealsoMS408,p.147,1893–1895.19.IpreferthisstatementofHookway’sposition,ratherthanthefollow-ing:‘Sometruthscanbeunderstoodina“realist”manner,asdealingwithamind-independentreality,whileothersdealwithmatterswhosecharacterbearsmoremarksofourinterests,sentimentsorconstructiveactivities’(2000:77).Forthislatterwayofputtingthepointmakesitseemasiftherearedifferentkindsofreality,someofwhicharede-servingofthetitle‘realist’andothersnot.Thenthequestionmustbewhetherthosedowngradedformsofrealityoughttocountasreality.Inthesamevein,noticethat,despitePeirce’slanguage,therearenotdiffer-entkindsoftruth–eachkindofinquiryaimsatgettingananswerthatwillnotbeoverturnedbysubsequentexperience.Thatisthepragmatistelucidationoftruth.20.Peircealsoarguesthatlogicisbasedonethicsinthefollowingway:‘logicrequires,beforeallelse,thatnodeterminatefact,nothingwhichcanhappentoaman’sself,shouldbeofmoreconsequencetohimthananythingelse.Hewhowouldnotsacrificehisownsoultosavethewholeworld,isillogicalinallhisinferences,collectively’(W3,284,1878).Peirce’sclaimhereisthatinordertomakesenseofthenotionsoftruthandprobability,weneedtorefertoanextendedcommunityofinquirers.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n174cherylmisakWhenhesaysthat‘itisimpossibleforamantobelogicalunlessheadoptscertainhighmoralaims’,hisargumentis‘extremelysimple’.Itis:‘Allpositivereasoningdependsuponprobability.Allprobabilitydependsuponthesuppositionthatthereisa“longrun”.Butalongrunisanendlesscourseofexperience...’(MSL75p.13–14,variants,1902).Logicdoesn’trequirethatsomeonebecapableoftheheroismofself-sacrifice;itjustrequiresthatwerecognisethepossibilityofit.Itrequiresthatweseethatourinferencesarevalidiftheywouldbeacceptedbythehero.Onehasto‘referhisinferencestothatstandard’(W3,284).SeeMisak(1991:108ff)forasustaineddiscussion.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\ndouglasanderson7Peirce’sCommonSenseMarriageofReligionandScienceIn1905inalettertoF.C.S.Schiller,CharlesPeircerespondedtoSchiller’sattempttodefine“pragmatism”:“Iwouldletitgrowandthensayitiswhatacertaingroupofthinkerswhoseemtounder-standoneanotherthink,andthusmakeitthenameofanaturalclassintheNaturalHistoryfashion”(MSL390p.3).Wemightfol-lowPeirce’ssuggestioningivinganaccountofpragmaticphilosophyofreligion.Ifwedoso,wefindPeirce’sworkmarkingoutamiddlepositionamongtheworkoftheotherAmericanpragmatists.AtoneextremewefindJohnDeweywho,inACommonFaith,defendedaminimalnotionofreligiosityinwhich“God”stoodforthepowerofactualizinghumanideals.Deweyexplicitlyrejectedbothsuper-naturalismandthechurch.NextonthespectrumwefindWilliamJameswholikewisedownplayedtheimportanceofthechurch.How-ever,inTheVarietiesofReligiousExperienceandelsewhere,Jamesdescribedanddefendedtheimportanceoftraditionalkindsofin-dividualreligiousexperience.AttheotherextremepositionstandsJosiahRoyce,whombothSchillerandPeirceincludedamongthepragmatists(MSL390p.2).Royce’s“absolutepragmatism”initiallyprovidedphilosophicalargumentationindefenseofareligiousout-look;later,inTheProblemofChristianity,Royce,workingundertheinfluenceofPeirce,developedtheimportanceofthechurchasa“belovedcommunity.”Inthis“naturalclass”Peirce’sphilosophyofreligionstandssomewherebetweenthoseofJamesandRoyce.Inthisessay,IwouldliketomarkoutsomeofthedefiningfeaturesofthismediatingpositionandtoshowthatitisafittingpieceofPeirce’sphilosophicalarchitectonic.Peircedidnotwriteanysingletextthatheidentifiedashisphi-losophyofreligion.Butoverthecourseofhiscareerhegenerateda175CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n176douglasandersonbasicoutlookonreligionbyexploringtherelationshipbetweenwhathecalledthespiritofscienceandthespiritofreligion.Thespiritofreligion,whichforPeirceisdrivenbyinstinct,feeling,andcommonsense,aimsdirectlyatguidingtheconductoflife.Becauseofitsfo-cusonpractice,Peircebelieved,thespiritofreligiontendstobe–andneedstobe–somewhatconservative.Thespiritofscience,ontheotherhand,issuchthatscienceisopentochange.Whenscienceisproperlyunderstoodnotasastagnantbodyofbeliefsbutas“alivingandgrowingbodyoftruth,”weseeitsnaturalinclinationtofreedom,change,andliberality(CP6.428).AsPeircearguedin1898,thedifferencebetweenthetwospiritsisreasonablesinceinpracticewemustact,forthemostpart,onthebasisofourfundedexperienceandintheorywemustallowourselvestobeasflexibleandcreativeastheinformationwegatherallows.“Thusithappensquitenatu-rally,”Peircesaid,“thatthosewhoareanimatedwiththespiritofscienceareforhurryingforward,whilethosewhohavetheinterestsofreligionatheartareapttopressback”(CP6.430).Howevernaturalthisdifferencebetweenthetwokindsofinquiry,Peircebelieveditwasoftenpushedtoanextremebybothscientistsandreligionists.In1911hereassertedthat“notwospirits(tenden-cies)notdownrightconflictingcanwellbemoreopposedthanthespiritofscienceandthespiritofreligion”(MS851,p.1).Hewentontopointoutthatthisoftenleadstoananimositybetweenthetwo.Difficultiesarisewhenthenoveltiesofscienceencounterthenaturalconservatismofreligion.“Inthisway,”heargued,“scienceandreligionbecomeforcedintohostileattitudes”(CP6.431).Itiseasytoseethatthishasbeenacommonprobleminthedevelop-inghistoriesofscienceandreligion.Butweneedtonotethephrase“notdownrightconflicting”inthe1911passage.Althoughscienceandreligionplacetheiremphasesanddeveloptheirtendenciesindifferentdirections,theyarenot,asPeircesawit,“downrightcon-flicting.”Thus,thehistoricalconflictsareinprinciplemendable–theopposingspiritsshouldbeabletolivetogether.Howsucha“marriage”ofscienceandreligionmightbeeffectedwasathemeofPeirce’sthinkingbothearlyandlateinhiscareer.Peirceoftenchidedscientistsfornotbeingmoretolerantofreligiousbelief.Inhis1908essay“ANeglectedArgumentfortheRealityofGod,”heevenwentsofarastomaintainthatreligiousexperiencessuchasCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCommonSenseMarriageofReligionandScience177his“humbleargument”fortherealityofGodmightbeincipientsci-entificexperiences.The“humbleargument”wassimplyanaturalaffinityforbelievinginGodasthecreatororsustainerofthecos-mos.Assuch,Peircesuggested,it“isnothingbutaninstanceofthefirststageofallsuch[scientific]work,thestageofobservingfacts,orvariouslyrearrangingthem,andofponderingthemuntil,bytheirreactionswiththeresultsofpreviousscientificexperience,thereis‘evolved’(aswechemistswordit)anexplanatoryhypothesis”(CP6.488).Nevertheless,mostofPeirce’seffortsweredirectedtowardreconstructingourunderstandingofareligiouslife–wemightcallithispragmatizing,orpragmaticizing,ofreligion.originsandaimsofreligionAsdidWilliamJames,Peircedownplayedrationalorphilosophicalargumentationasasourceofreligiousbelief.ForPeirce,religiousbeliefwasmostoftenexemplaryofinstinctiveorcommonsensicalbelief.Inhis“NeglectedArgument”hemadethisexplicitinhisdis-cussionofbeliefintherealityofGod.Thereheargued,intryingtomakesenseofGalileo’sconceptionof“simplicity,”that“thesim-plerHypothesisinthesenseofthemorefacileandnatural”is“theonethatinstinctsuggests”(CP6.477).Thusthe“humble”beliefinGod’srealitywasaninitiallystronghypothesisbecauseitexcitedthe“peculiarconfidence”thatinstinctivebeliefscreate“inthehigh-estdegree”(CP6.477).Religiousbeliefisdirectlyexperientialandthereforebearsthestrengthofimmediacy;itiswhatPeirceoccasion-allycalled“practicallyindubitable.”Peirceoftenequatedinstinctiveandcommonsensebeliefswithfeelingandperception,furtheram-plifyingtheoriginarypowerthatreligiousexperiencedisplayed.1InhiseighthLowelllectureof1903,Peirceindicatedthatinstinctsug-gestedakindofdirectexperienceorperception:Ordinaryideasofperception,whichDescartesthoughtweremosthorriblyconfused,haveneverthelesssomethinginthemthatverynearlywarrantstheirtruth,ifitdoesnotquiteso.‘Seeingisbelieving,saysthemanofinstinct.’(CP5.593)In1896,Peircehadalreadydrawntheconnectionbetweenreligiousbeliefandperception,hintingatitsinstinctiveorcommonsensicalCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n178douglasandersonnature.Inoneofhismanyattacksonnominalismasunscientific,hebroughtperceptionintothereligiousarena:Wherewouldonefindsuchanidea,sayasthatofGod,comefrom,ifnotfromdirectexperience?Wouldyoumakeitaresultofsomekindofreasoning,goodorbad?....No:astoGod,openyoureyes–andyourheart,whichisalsoaperceptiveorgan–andyouseehim.[CP6.493]Inshort,philosophyisnottheoriginofmostreligiousbelief–suchbeliefisfelt,perceived,orexperienced.Thishasseveralimpli-cations.First,thepowerofinstinctivebeliefseemedtoPeirceexem-plifiedbyreligiousexperience,thusprovidingasolidbasisforthecommitmentsnecessarytotheconductoflife.Second,itmeantthatthecontentsofsuchbeliefswereinvariablyvague,general,andun-finished.Andfinally,thisindefinitenessmeantthatreligiousbeliefsrequiredphilosophy,science,ortheoreticalinquiryfortheirdevelop-ment–theirunfinishedstateneededexplicationandinterpretation.Thus,theinstinctive–perceptualoriginofreligiousbeliefmadeitusefulforitsprimaryaim–conductinglife.Butatthesametime,itmadeitsusceptibletomisuseandabuse,forwhichtheonlyantidotecouldbeanopennesstoongoingtheoreticalinquiryconcerningitsmeaning.The“mostdistinctivecharacteroftheCritical-Common-sensist,”Peirceremarked,“liesinhisinsistencethattheacriticallyindu-bitableisinvariablyvague”(CP5.446).Theinstinctivenessoftheideasofreligiousbeliefthusentailstheirindeterminatenessinvague-nessandgenerality.2Itisimportanttonotethatthisisastrengthofinstinctivebeliefs,asPeircesawit,notaweakness.Thevaguenessofthecommonsenseorinstinctivebeliefsallowsthemtoprovideadirectionorheadingforourconductwithoutforeclosingonthevari-etyofwaysinwhichthatdirectionmightbedeveloped.Inreligiousbeliefspecifically,itallowsforavarietyofreligiousexperiencesthataretetheredonlybyvague,workingconceptionsofGod,love,andthesummumbonum.Inshort,vagueandgeneralconceptionsaregoodenoughfortheconductoflife,thatis,formostofourpracticalconcerns.SuchaviewliesattheheartofPeirce’s“NeglectedArgu-ment.”Therethe“God”hypothesisisunderstood“asvagueyetastruesofarasitisdefinite”(CP6.466;seealsoCP6.494).Ina1905lettertoWilliamJames,Peircemadethepointinsimilarfashion:“Theidea[ofa“living”God]isavagueonebutisonlythemoreCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCommonSenseMarriageofReligionandScience179irresistibleforthat.Subtiledistinctionsareoutofplace;thetruthofcommonsenseisthatlittleaswecancomprehendtheauthorofallbeautyandpowerandthought,itisreallyimpossible,exceptbysophisticatingtheplaintruth,tothinkotherwisethanthatthereisalivingbeing”(MSL224:n.p.,7/26/05).Peircesawthedirectnessofreligiousexperienceasakindofknowledgebyacquaintance,thewaywemightknow,say,physicalsufferingthroughparticipationinacontestofendurance.Heseemedtoindicatethatthoughthedefinitenessofsuchideasislimited,wehaveaclosenesstothemthroughthisdirectacquaintancesuchthattheyareeasilyemployedinguidingeverydaypractices:“Nowordsaresowellunderstoodinoneway,yettheyareinvariablyvague;andofmanyofthemitistruethat,letthelogiciandohisbesttosubstitutepreciseequivalentsintheirplaces,still,thevernacularwordsalone,foralltheirvague-ness,answertheprincipalpurposes”(CP6.494).Thus,forreligiouspractice,vaguenessisavirtuenotavice;asPotterpointsout,“itisvaguenesswhichallowsournotionstobeaboutGod”(Potter1972:249).Thespiritofreligionistoconductourlivesundertheguidanceofinstinctive,commonsensebeliefsinsuchawayastoamelio-ratehumanexistence.Toestablishthisistheaimofthefirstofthethreenestedargumentsinthe“NeglectedArgument,”the“hum-bleargument”notedabove:“anynormalmanwhoconsidersthethreeUniversesinthelightofthehypothesisofGod’sReality,andpursuesthatlineofreflectioninscientificsinglenessofheart,willcometobestirredtothedepthsofhisnaturebythebeautyoftheideaandbyitsAugustpracticality,eventothepointofearnestlylov-ingandadoringhishypotheticalGod,andtothatofdesiringaboveallthingstoshapethewholeconductoflifeandallthespringsofactionintoconformitywiththathypothesis”(CP6.467).Thisseemsatallorderatfirstglance,butPeircebelievedhewassimplyreflect-inganexperiencethathadbeencommonamonghumanculturesandhistories.Acrosstimeandculturalspace,thereseemstobeageneralconsensusofthe“goodness”ofthosewhoactfromloveandcaringtowardtheinterestsofothers.AsKellyParkersuggests,forPeirce,those“whoareaffectedbythereligioussentimentwillbeat-tunedtothebenevolent,just,andwiseaspectsoftheworld,andwillmoldtheirlivessoastocontributetothesetendenciesinsocietyandintheorderofthings”(Parker1990:198).InsuchattunementthereisCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n180douglasandersonanaturalconservatism,ahumbleaffinityfortheacriticalinstinctiveandcommonsensebeliefs.However,thespiritofthisconservatismisdirectedtowardmoralactionandisnotantagonistictothespiritofscience.Itisinnocentratherthanignorant.Itis,asitwere,notyetawaketothespiritofscience.Notheorizing,interpreting,orprecisingoftheinstinctivebeliefsisyetatstake.Theoriginofthereligiouslifeis,fromPeirce’sangleofvision,adequatetotheaimofreligion’spracticaltask.OurvaguenotionsofGodandagape,orcherishinglove,aresufficientforustogetonwiththeprojectofbetteringhumanexistence.3Nevertheless,Peirceroutinelyworriedaboutthedangers–toscience,tomoral-ity,andtoreligionitself–whosegermslayatthisjunctureofreli-giousbeliefandpractice.Theconservatism,iftakentotheextreme,couldstandinthewayofanyquestfortruth.Thus,thewayoflifeguidedbyreligiousbeliefwasalwaysontrial;weareinfactalwayscarryingoutexperimentsonourbeliefsinthewayswelive.ThismarkstheclearlyJamesiansideofPeirce’spragmaticphilosophyofreligion:“Evenforthegreatestsaints,theactivemotiveswerenotsuchhopesandfears[ofheavenandhell],buttheprospectofleav-ingbehindthemfertileseedsofdesirablefruitshereonearth”(CP6.451).4Theconsequencesofouractionsare,inpartatleast,thepragmaticistictestofthetruthofourinstinctivebeliefs.Thebeliefstakeonalifeinhumanhistoryandremainopen,togrowth,de-velopment,andchange.Thus,forPeirce,the“reasonable”religiouspersonwillseethatthehypothesis[ofareal,lovingGod],irresistiblethoughitbetofirstintention,yetneedsProbation;andthatthoughaninfinitebeingisnottieddowntoanyconsistency,yetman,likeanyotheranimal,isgiftedwithpowerofunderstandingsufficientfortheconductoflife.Thisbringshim,fortestingthehypothesis,totakinghisstanduponPragmaticism,whichimpliesfaithincommonsenseandinstinct,thoughonlyastheyissuefromthecupelfurnaceofmeasuredcriticism.Inshort,hewillsaythattheN.A.[NeglectedArgument]istheFirstStageofascientificinquiry,resultinginahypothesisoftheveryhighestPlausibility,whoseultimatetestmustlieinitsvalueintheself-controlledgrowthofman’sconductoflife.(CP6.480)Peirce’sconceptionofthereligiouslifethussharesJames’sandDewey’smeliorism.ThedifferenceisthatPeircemakesamuchstrongercommitmenttotheregulativeidealsoftruthandgoodness.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCommonSenseMarriageofReligionandScience181Thiscommitmentrequiresanopennesstothegrowthanddevelop-mentofreligiousbeliefsaswithallbeliefs–religiousbeliefmustbeopentoreflectionandinquiry.ForPeirce,thismeantthatthespiritofreligionmustfindawaytomarryitselftothespiritofsci-ence.However,hebelievedthatanothermodeoftheorizingaboutreligiousbeliefstoodinthewayofthisproperunion.theologyandthetheorizingofreligiousbeliefAsagoodAristotelian,Peircedidsuggestthatitwasnaturalforper-sonstowanttotheorizeabouttheworldandabouttheirreligiousbe-liefs.Hemaintainedthatreligiousideas“areeasilydoubted”andweknowthatforhimdoubtwasthemainspringofinquiry.So,theques-tion,especiallygivenhiscriticalcommonsensism,wasnotwhetheronecaninquireintoreligiousideasorintootherideasthatmighthaveabearingonreligion,thequestionwashowoneistodothis.Theology,asPeirceunderstoodit,employsthewrongkindoftheo-rizingandleadstoaviciousconservatismthatthreatensbothscienceandreligion.Thefailureofreligions,fromapragmaticpointofview,hasbyandlargebeenafunctionoftheirengagingintheology.Peircebelievedthattheologymasqueradedasasciencewhileitwas,inessence,an-titheticaltothespiritofscience.Itindeedbroughtlogictobearonreligiousbeliefs,butitmistakenlytreatedreligionasaclosedde-ductivesystemofideas.Since“theologypretendstobeascience,”Peircearguedin1898,theologians“mustalsobejudgedasscientificmen”(CP6.3).Judgedinthisway,theologiansinvariablyandmis-erablyfailbecauseinsteadofseekingtruththroughopeninquiry,theytakeastheir“principalbusiness...tomakemenfeeltheenor-mityoftheslightestdeparturefromthemetaphysicstheyassumetobeconnectedwiththestandardfaith”(CP6.3).TheologyposesasanexplorationofthenaturesofGod,creation,andthecosmos.Butitisantiscientificjustinsofarasitistenaciousandauthorita-tive.Insteadofactuallyexploringthepossibilities,itbeginswithadogmaticplatformandseekstoinsulateitfromcriticism.Itisatbestanuncriticalcommonsensism.Indeed,theologyneednotevenbecommon-sensical,sincethetheologianmay–andoftendoes–arbitrarilyadoptanybeliefastheoriginofhisorherdeductivework.Moreoftenthannot,theologybecomesapracticalinstrumentCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n182douglasandersonfornarrowaims–aninstrumentthatisneitherscientificnortheoretical.Theology’smethod,ifitisamethod,istoexpressanddefendtena-ciouslyandauthoritativelysomespecificversionofreligiousideas.Todothis,ittriestospecifytheideassoparticularrulesandinter-pretationscanbenaileddown.Forexample,thevernacular“God”isreplacedwithanamedbeingorbeingswhoarehistoricallylocated,embodied,orotherwisedefinitelydescribed.Likewise,the“good”isreducedtoanarrowformulaofbehavior,asetofrulesthatcur-tailshumanvarietyandflexibilityindealingwithlifesituations.Inshort,theologiansproduceanddefendcreedsanddoctrines.Un-derthemethodoftheology“theChurchrequiressubscriptiontoaplatform–aCreed”(CP6.450).Bywayofthisversionoftheoriz-ingreligion,Peirceargued,wecannot“hopethatanybodyofpriestsshouldconsiderthemselvesmoreteachersofreligioningeneralthanoftheparticularsystemoftheologyadvocatedbytheirownparty”(CP6.427).TheologythusembodiesallthatPeirceresisted:tenacity,authority,closureofinquiry,andabsenceofgrowth.Ithasrepeat-edlyproveditselfadangertohumanity,and,asParkeraptlystates,theologiansare“tobechastisedasmuchformuddyingthewatersofreligionastheyareforobstructingthescientificspirit”(Parker1990:196);theologyasawayoftheorizingaboutreligiousbeliefisnotonlyunscientific,itisantagonistictoreligionitself.Thewatersofreligiongetmuddiedinatleastthreeways:(1)theologymakesreligionexclusionaryandantitheticaltotheprin-cipleoflove,(2)theologyoverlooksand/orrejectsreligiouspercep-tionandsentiment,and(3)theology’soverdeterminationofreligiousideasleadstoaninquiryitcannotsatisfy.Thislastwayalsomarksthepointatwhichtheologicalcredalismantagonizesthespiritofscience.Aswehavejustnoted,Peircebelievedtheology’sprimarytaskwastodemandadherencetoaspecificdoctrineandtoreject,usuallyinanarticulatefashion,anydeviationfromthisdoctrine:“religioustruthhavingbeenoncedefinedisnevertobealteredinthemostminuteparticular”(CP1.40).Theupshotisthattheologicalpracticeisfun-damentallyexclusionaryandestablishesthebasisforwarringamongdifferentreligiousoutlooks.BecausePeirceidentifiedGodwithloveandtookagapetobetheinstinctivebasisofareligiousattitude,hefoundtheexclusionarypracticetobefundamentallyirreligiousandCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCommonSenseMarriageofReligionandScience183immoral(seeCP6.441).Consequently,in“WhatIsChristianFaith?”heaskedthatwe“[d]iscountenanceasimmoralallmovementsthatexaggeratedifferences,orthatgotomakefellowshipdependonfor-mulasinventedtoexcludesomeChristiansfromcommunionwithothers”(CP6.445).Inshort,Peircewasradicallyantifundamentalist.Sincetheaimofreligionistoguidetheconductoflifemelioristically,theology,becauseitisessentiallyfundamentalist,isbothimmoralandirreligious.ThesecondproblemPeircefoundintheologywasitsobscuringofreligiousexperiencesandthereligioussentiment.Ontheonehand,becausetheologyisprimarilythedeductiveexplicationofasetofdogmaticpremises,thetheologiansneednothaveanydirectacquain-tancewithreligioussentimentorsensibility:“aman,”Peircesaid,“maybeanaccomplishedtheologianwithouteverhavingfeltthestirringofthespirit,buthecannotanswerthesimplequestionattheheadofthisarticle[WhatisChristianfaith?]exceptoutofhisownreligiousexperience”(CP6.435).Ontheotherhand,theology,inenforcingclosureonthecontentofbelief,oftenopenlychallengestheveryoriginofreligiousbelief.Thosewhodisciplethemselvestoatheologicaldoctrinebecomeautomatafollowingrulesratherthanbelieversinspiredtoproduceabetterlifethroughlove.Insteadoflookingtotheefficacyofloveindailylife,theologiansgenerateargumentationthatunnecessarilypoliticizesthereligiouslife:Then,afterareligionhasbecomeapublicaffair,quarrelsarise,tosettlewhichwatchwordsaredrawnup.Thisbusinessgetsintothehandsofthe-ologians:andtheideasoftheologiansalwaysappreciablydifferfromthoseoftheuniversalchurch.Theyswampreligioninfallaciousdisputations.Thus,thenaturaltendencyistothecontinualdrawingtighterandtighterofthenarrowingboundsofdoctrine,withlessandlessattentiontothelivingessenceofreligion,until,aftersomesymbolumquodcumquehasdeclaredthatthesalvationofeachindividualabsolutelyandalmostexclusivelyde-pendsuponhisentertainingacorrectmetaphysicsofthegodhead,thevitalsparkofinspirationbecomesfinallyquiteextinct.(CP6.438)Thus,theviciousconservatismoftheologymovesbeyondthework-ingconservatismofreligiousbeliefandworkstodestroythereligiouslifeoutofwhichitgrew.Thisviciousconservatismalsorevealstheology’sfinaldistortionofagenuinelyreligiouslife.Aswenotedearlier,forPeirce,religiousCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n184douglasandersonbeliefworkedeffectivelywithvagueideas–theideasgeneratedbyin-stinctorcommonsense.Theology’saim,however,isalwaystogiveconcretedetermination–Peircewouldcallitoverdetermination–tothevagueideas.But,asPotterpointsout,themoreoverdeterminedtheseideasbecome,themoredoubtabletheybecome:“Totheextentthatsomeparticularformulationofaninstinctivebeliefisdefinite(nonvague),thatformulationisopentodoubtanddemandscriticalreview”(Potter1972:225).ThisisthepointofPeirce’scriticalcom-monsensism.Thebeliefsmovefromtheroleofpracticalguidestothatofhypothesesinaninquiry.Theologyroutinelyfailstodrawthisdistinction.ForPeircethereisacleardivisionoflabor.Religioncaninitiallyeffectitspracticalworkwithitsvaguebeliefs.However,asnatu-ralcuriosityleadsustoinquireabouttheseideas,wemovefromreligiontoscience,frompracticetotheory.Thus,whentheology’snarrowdeterminationofreligiousbeliefscreatesdoubtandgeneratesinquiry,wemustturntothespiritandthemethodofscience.Buttheologyisunabletomakethetransition.Theheightofitsreasoningisexplicativededuction.Itcanonlyreturntoitsdogmaticdoctrineandauthoritativelyrepeatitsarguments.Itcirclesitswagonsandconservativelyholdsitsground.ForPeirce,however,agenuinelyreligiousoutlookwouldunder-standthelimitationsofitsaimandwouldknowtorelinquishtheresponsibilityofinquirytothescientificspirit.5Ifreligiouslifeistoamelioratetheworld,itmust,Peircebelieved,holdanabidingrespectfortruth.Suchrespectinvolvesanopennesstogrowth,todevelopment.Thus,asideasdevelopthroughthecommunityofin-quirers,theywillhaveagradualeffectonreligiousbeliefandsubse-quentlyonreligiouspractices:“Idonotsaythatphilosophicalsci-enceshouldnotultimatelyinfluencereligionandmorality;Ionlysaythatitshouldbeallowedtodosoonlywithsecularslownessandthemostconservativecaution”(CP1.620).APeirceanreligionmustactthroughitscommitmentsatthesametimethatitremainsopentoself-developmentandself-revision.PeirceputitthiswaywhenhewrotetoJamesin1897:“‘Faith,’inthesensethatonewilladhereconsistentlytoagivenlineofconduct,ishighlynecessaryinaffairs.Butifitmeansyouarenotgoingtobealertforindicationsthatthemomenthascometochangeyourtactics,Ithinkitisruinousinpractice”(MSL224,p.2).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCommonSenseMarriageofReligionandScience185Peircethusenvisionedreligioninareciprocaldependencewithscience;thetwomustengageinanongoingdialecticalrelationship.Anideathatiseffectiveasareligiousbelief,ifitistobetheorizedabout,mustturnitselfovertoscientificinquiry,tocriticism.Athe-ologicallydrivenreligionisunabletodothisandtherebybecomes,onPeirce’sterms,irreligious.Theologysimplyopensuptheneedforaninquirythatitcannotdeliver.RichardTrammellseesthispointmademanifestinPeirce’sdiscussionofmusementinthe“Ne-glectedArgument”:“Thisargumentshowsthatthesamecourseofmeditationwhich,forpracticalpurposes,producesalivingbeliefinGod,fromanotherpointofviewisthefirststageofatheoreticalin-quiry”(Trammell1972:19).ForPeirce,atheologicalreligionhasnoroutetogenuineinquiryandthereforemakesthespiritsofscienceandreligionfundamentallyantagonistic.Acriticallycommonsen-sistreligion,ontheotherhand,whileacknowledgingthedifferentspiritsofscienceandreligion,bringsthemintounionthroughamu-tualdependence.Consequently,asMichaelRaposapointsout,Peirce“thoughtitaltogetherreasonablethatcertainreligiousbeliefsshouldberevisedorevendiscardedasaresultofnewscientificdiscoveries”(Raposa1989:13).churchasacommunityofloveInexaminingPeirce’sthoroughrejectionoftheologyforitsirreli-giosityanditsantagonismtowardscience,wemightexpecthimtojoinJamesandDeweyinrejectingchurches,ororganizedreligions,ingeneral.JamesandDeweysharedPeirce’srejectionofcredalismandviciousconservatism.Jamesblamedachurch’sapparentneedtomanipulateandcontrolitsmembersforruiningthebeneficialeffectsofindividualreligiousexperience.Deweysimplyrejectedreligionsaltogether,arguingthattheyhadcausedtoomanyproblemshistor-icallywiththeirsupernaturalismstoberedeemed.Peirce,however,didnotfollowsuit.Instead,heturnedinthedirectionofRoyce,de-fendingthecentralimportanceofachurchforthereligiouslife’staskofamelioratinghumanexistenceinthisworld.Throughhispeculiar–andadmittedlytense–marriageofscienceandreligion,hearguedforthepossibilityofanontheologicalchurch.WenotedearlierthatforPeirceasforJamesreligionbeginsinper-sonalexperience.However,becausereligion’simportanceisfoundCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n186douglasandersoninguidingtheconductoflife,itimmediatelydisclosesthatithassocialconsequences.“Man’shighestdevelopments,”Peirceargued,“aresocial;andreligion,thoughitbeginsinaseminalindividualinspiration,onlycomestofullflowerinagreatchurchcoextensivewithcivilization.Thisistrueofeveryreligion,butsupereminentlysoofthereligionoflove”(CP6.493).Thechurchleadsusoutofour-selves,asagaperequires,andgeneratesconcernforothers.Itisthismovetowardselflessness–whichPeircealsotooktobeacentralfeatureofthespiritofscience–thatmakesourlives“social”andnotjustmechanicallyinteractive.“Theraisond’etreofachurch,”Peircebelieved,“istoconferuponmenalifebroaderthantheirnar-rowpersonalities,aliferootedintheverytruthofbeing”(CP6.451).Notonlydoesthechurchenablea“broaderlife,”italsoservesasavehicleforthesocialworkthattheprincipleofloverequires.Peirce’sdescriptionsofthechurch’sreasonforbeingarerootedinhisagapasm–hisbeliefthatloveisaneffectiveforceintheevo-lutionordevelopmentoftheuniverse.Whateveronethinksofitasacosmologicalspeculation,agapasmcanbeconvincingasanor-mativetheoryofhowhumanhistorycandevelopinameliorativefashion.6Itisagapasticlovethatovercomesself-interestandself-loveandturnstotheinterestsofothersandultimatelytotheinterestofthetruthofGod’scosmos.PeirceoftenadmittedtohavingcaughtadoseofSchellingianromanticismanditnodoubtrevealsitselfhere.Nevertheless,hisoutlookisreasonablyinlinewithhisrealism,hiscommitmenttoinquirytowardtruth,andhisbeliefinthepossibil-ityofagenuinecommunityofinquirers–hisdefenseofanagapasticchurchisnotmerelyahopefuladdendumtohisotherwork.Itisadevelopmentofhisowncriticalcommonsensism.Theprincipleofloveexpressedcommon-sensicallyintheGoldenRule,hemain-tained,“doesnot,ofcourse,say,Doeverythingpossibletogratifytheegoisticimpulsesofothers,butitsays,Sacrificeyourownper-fectiontotheperfectionmentofyourneighbor”(CP6.288).Thus,thechurch’sfunctionistodisseminatetheprincipleoflovetocombatthespecific“evils”generatedbyself-loveandself-seeking.Justasthecommunityofinquirerswasneededtomovescienceforward,thechurch,asabelovedcommunity,wasrequiredfortheameliorativeworkofreligion.Thechurchcannotbemerelyanin-stitutionfor“gettingtogether,”akindofsocialclub.“Areligiousorganization,”Peirceasserted,“isasomewhatidleaffairunlessitCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCommonSenseMarriageofReligionandScience187besworninasaregimentofthatgreatarmythattakeslifeinhand,withallitsdelights,ingrimmestfighttoputdowntheprincipleofself-seeking,andtomakesuretheprincipleofloveistriumphant”(CP6.448).Thetriumphsoughtcannotbeamatterofgettingpersonstosignon,intheologicalfashion,toacreed.Theprincipleoflovemustbedisseminatedthroughactionsthatarethemselvesgovernedbylove.This,again,isthepragmatictestofPeirce’sversionofreli-gionandofachurch.In“theMarriageofScienceandReligion”hemadethispointatlength:Butreligioncannotresideinitstotalityinasingleindividual.Likeeveryspeciesofreality,itisessentiallyasocial,apublicaffair.Itistheideaofawholechurch,weldingallitsmemberstogetherinoneorganic,systemicperceptionoftheGloryoftheHighest–anideahavingagrowthfromgen-erationtogenerationandclaimingasupremacyinthedeterminationofallconduct,privateandpublic.(CP6.429)TopursuethisRoyceandimensionofhisphilosophyofreligion,PeirceneededtoalterhischurchtomeettheconcernsexpressedbyJamesandDewey.Inbeingachurchoflove,hischurchmustbe“universal,”itmustconfidentlyturntheoryovertothespiritofscience,and,consequently,itmustbestrongenoughtolearnfromandgrowthroughitsownfailures.Inshort,Peirce’schurchwastobeadirectanswertothedangershelocatedintheologicalapproachestoreligion.Peirce’sagapasticchurchrequiresuniversalityinbothitsoriginanditsaim.Becauseitisgeneratedbyinstinctiveorcommonsensebeliefs,itisopentoandaccessiblebyeveryonefromthe“clodhop-per”tothescientist:“ithasalwaysseemedtomereasonabletosup-posethat,ifHe[aGodinwhomreligiouspeopleofallcreedsbelieve]reallyis,theremustbesomegoodreasonforbelievingso,otherwisethanonauthorityofsomekind,whichshouldappealtothelowliestmind...”(MS842,p.8–9).ThusPeirce’schurchisuniversallyopentoallwhowouldpayattentiontocommonsenseandinstinct.Thechurch’saimmustlikewisebeinclusive.Actionsundertheguidanceofagapemustreachouttotheinterestsofallpersonsthroughaconcernfortheir“perfectionment.”Personsarebroughttogethernotbyaforcedagreementtosomedoctrinebutbyagenuineacceptanceoftheirimportance.Lovemustbetransformative:“Love,recognizinggermsoflovelinessinthehateful,graduallywarmsitCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n188douglasandersonintolife,andmakesitlovely”(CP6.289).Atthepoliticallevel,Peirce’snontheologicalchurchmustnotactsoastodividebelievers,butmustseek“topatchupsuchpeaceasmightbewiththegreatreligiousworld”(CP6.447).Thus,Peirce’schurchnotonlyacceptseveryone,itattendstoeveryone.Thebidirectionaluniversalityoftheagapasticchurchplacesex-traordinarydemandsonit–demandsnotaltogetherunfamiliartothosewhowouldjoinPeirce’scommunityofinquirers.Athorough-goingselflessnessmustbecomeanordinaryhabit.Thechurchalsoneedsthestrengthtocontendwithitsownfinitudeandunfinishedstate.Itmust,ontheonehand,bewillingtoactthroughitscommit-mentstoitsvague,instinctivebeliefsoutofwhichitisgenerated.Atthesametime,itmustknowthatitsbeliefs–“workingcreeds”wemightcallthem–areopentocriticism,torefinement,andtogrowth.Theuniversalchurch’spracticesaretobetemperedbyanabidingrespectfortruth–anavenueofhumilitythatwasnotopentotheologicalchurches.Peircedemandedthathischurchbeconser-vativeinitspracticeandliberalinitstheory.Inthisway,itallowsitselftogrow,toreviseitselfinpragmaticfashion–aprocessthatfordifferentreasonsJamesandDeweydidnotthinkpossible.Insofarastheuniversalchurchcanlivewiththistension,itcankeepalivethereligiousattitudethatisitslife’sblood.Peircebelievedthechurchcouldlearnfromthespiritofscience“tobecomemoreandmoreperfect”insteadofsuffocatingitselfwithdoctrinesuntilthe“vitalsentimentthatgaveitbirthlosesgraduallyitspristinepurityandstrength”(CP6.430).InamuchmoresystematicwaythanEmer-sonandThoreau,PeircedefendedtheAmericantranscendentalists’desirefora“livingreligion.”ThelynchpintothesuccessofPeirce’suniversalchurchisthisfundamentalrespectfortruth.Achurchoflovemustalsobeachurchoftruth.Toactwithconcernforothersistoact,sofaraswecan,inlinewiththewaythingsare–fromareligiousperspective,afterall,itisGod’sworld,notours.Ablindandignorantloveislikelytofailtoachieveitspurpose.AndsinceforPeircetruthisalways“ontheway,”thechurchmustconfidentlyrespectchangeanddevelopmentthatresultfrominquiry.Inan1898manuscriptheentitled“ReligionandPolitics”Peircerecountedlearningthislessonfromamomentinpolitics.ADemocraticsenatorhadlefthispartybecauseitsplatformbecame“contrarytohisconvictions”(MS894p.1).ThenewspapersCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCommonSenseMarriageofReligionandScience189inNewYork“expressedcontempt”forthesenator’saction.Peirceagreedthatthesenatorhadacommitmenttohisparty,butarguedthathehadanevengreatercommitmenttothewholecommunityofcitizensandthereforemightberightinrejectinghisparty.Achurchis,Peircesuggested,analogoustotheparty–ithassomeplatformtowhichweowesomeallegiance.Nevertheless,“owewhatonemaytotheChurch,thetruthclaimspermanentallegiance”(MS894,p.2;CP6.450;seealsoCP6.426).Initsallegiancetothetruth,theuniversalchurchmustrecognizeitsneedforthespiritofscience.Peirce’schurch,asacommunityoflove,doesnotproducea“reli-gionofscience”inwhichonesimplyrejectsinstinctiveandcommonsensebeliefs.Thesebeliefsremainthe“bedrock”ofanyreasoning.Buthischurchmustbecriticallycommonsensistandmustaccepthispeculiarmarriageofscienceandreligion:Themanwhomreligiousexperiencemostdevoutlymovescanrecognizethestateofthecase.Whileadheringtotheessenceofreligion,andsofaraspossibletothechurch,whichisallbutessential,say,penessential,toit,hewillcastasidethatreligioustimiditythatisforeverpromptingthechurchtorecoilfromthepathsintowhichtheGovernorofhistoryisleadingthemindsofmen,acowardicethathasstoodthroughtheagesasthelandmarkandlimitofherlittlefaith,andwillgladlygoforward,surethattruthisnotsplitintotwowarringdoctrines,andthatanychangethatknowledgecanworkinhisfaithcanonlyaffectitsexpression,notthedeepmysteryexpressed.(CP6.432)ReligionandscienceworktogethertowardPeirce’ssummumbonum,thegrowthofconcretereasonableness–therealizingandactualizingofpurposesandmeaning(CP5.3,5.433).What,Peirceasked,“isman’sproperfunctionifitbenottoembodygeneralideasinart-creations,inutilities,andaboveallintheoreticalcognition?”(CP6.476).Scienceistobeneitherrejectednorromanticized;itsspiritistoworkinconcertwiththespiritofreligionasacriticalmea-sureofreligion’sinstinctandcommonsense.Togethertheygeneratethepossibilityofalivingreligionandalivingscience.conclusionPeirce’spragmatic–orpragmaticistic–philosophyofreligionholdstoatenuousmiddleground.AsforJamesandDewey,individualCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n190douglasandersonreligiousexperienceistheoriginofanyreligiosity.AndasforRoyce,acommunity–ultimatelyauniversalcommunity–ofreligiousbe-lieversisessentialtotheameliorativeconductthatreligiouspracticerequires.Tothispoint,Ihavesoughtprimarilytosketchadescrip-tionofPeirce’smediatingoutlook,nottodefendit.Itis,ofcourse,notwithoutitsdifficulties.Deweyanswillcontinuetowonderifthedangersofreligionsandchurchesdonotoutweighanypossi-blebenefits.SomewillwithChristopherHookwayraisequestionsconcerningPeirce’sclaimthatbeliefinGod,vagueorotherwise,isinstinctive(Hookway,2000:269–72).Likewise,therearedifficultiesofinternalconsistencylurkingwithinPeirce’soveralldescription.Nevertheless,Peirce’spragmaticphilosophyofreligionisbothprovocativeandsuggestive.ItisprovocativeinpartbecauseitisPeircewhodevelopedit.ManywhohavebeenattractedtoPeircebe-causeofhisworkinlogic,thephilosophyofscience,andsemeioticfindhisworkonreligiousissuesdisconcerting.Itwasthisprovoca-tiveness,Ithink,thatinpartinspiredThomasGoudge’swell-knownsuggestionthatthereweretwoincompatiblePeirces:areasonablescientificPeirceandalessreasonabletranscendentalistPeirce.WhatIhavetriedtoshowhereisthatPeirce’sphilosophyofreligionisnotfundamentallyopposedto,butworksinconcertwithhistheoryofinquiryandhispragmaticism.Thisisusefulifoneisinterestedintheviabilityofbothscienceandreligion.Alongtheselines,Peirce’sworkissuggestive.Onevirtueofhisoutlookisthatbothasareli-giousbelieverandasascientificinquirerhecanfaceallcriticismsdirectly;hecanlookdowntheroadtowardtheconsequencesofhisownconfrontationwithcriticismstoseehowhisownviewshavedevelopedorgrownandhowtheymightbefurtherrevised.Hisre-ligiosityisnotanongoingdeath-rattleandhisscienceisnotadog-maticnominalismterminatinginanequallydogmaticmaterialism.Inafirmbutarticulatemannerheclaimsaplagueonbothhouses:fundamentalistreligionanddogmaticscientism.Moreover,likemanyphilosophersofreligion,Peircewantstoseethebestofreligiousexperienceexpressitselfthroughauniversalchurch.Buthedoesnotuniversalizeinpatchworkfashionbytry-ingtomeldorsyncretizethevariousworldreligions.Instead,hethrowsreligionbackontoitsinstinctivebeliefsinalltheirvague-ness,askingthatwetakethemas“goodenoughfornow,”whileCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nCommonSenseMarriageofReligionandScience191wetogetherworkouttruthinthelongrun.Interestingly,Peirce’spragmaticandscientificreligionnotonlymakesroomforthein-corporationofnewscientifictruths,itopensustothegrowthanddevelopmentofspecificmoralinsightsunderthedirectionofourvagueandgeneralinstinctivebeliefs.Theprinciplesofloveandjus-ticeareopentoextensionbeyondwhateverhistoricalboundstheyatanytimerest.Peirce’sevolutionarylovecanunderwritehuman-ity’sgrowingresistancetoslavery,tosexism,toracism,andtosimpleabusewhereveritoccurs;itscentralaimistostrugglewithwhateverparticularevilsitencounters.Peirce’sanswertothehistoricististhatthevitalsentimentsofreligionaretrueinsofarastheyarevagueandgeneral.Whatishuman-madearethewaysinwhichweimplementthem,andtheseremainopentocriticismanddevelopment.Moraldevelopment–amelioratingconduct–isarealpossibilityinPeirce’spragmaticconceptionofreligion.Ifwearetotrytobringthevarietyofhumanexperiencestogether,notreductively,butastheyarelived,thenPeirce’scommonsensemarriageofscienceandreligionisinstructive.Themiddleisnotexcluded;wecanlivetowardtruthwithouthavingcertainty.Wecanactwithcommitmentwithoutbecomingtotalizing.TheseseemtoPeircetobethebestfeaturesofreligionandscienceinaworldinwhichtruthisonthehighway.Andheoffersusawayoflivingonelifewithbothinhand.notes1.ThedetailsoftherelationshipPeircetriedtoestablishbetweenper-ceptionandinstinctrequireanotherinquiry.Inthe1903pragmatismlecturesheraisedtheissueexplicitlyandsuggestedthepossibilityofperceiving“thirds.”Sofarasthisistrue,Peirceheldanearlierandevenmoreradical“radicalempiricism”thanthatofJames.Thiswouldhelpunderwritehisclaimsthat“God”isperceivableinsomefashionandisnodoubtlinkedtohisstrongScotisticrealism.However,italsoseemstocreateproblemsfordistinctionshemightwanttomakebetweenconceivingandperceiving.2.ForadetaileddiscussionoftheseandhowtheyarerelevanttoPeirce’sconceptionof“God,”seePotter1972.3.ItisimportanttonotethatPeirceoftenequated“God”and“love.”AgapeisGod’smodeofagency.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n192douglasanderson4.Indeed,Peircethoughtofhumanimmortalityinlightofhisrealism.Ourhabitsandpersonalitiesleavebehindrealeffects;itisbywayofthesethatwe“immortalize”ourselves.5.IshouldnotethatforPeircethisdoesnotmeanthatreligionmustsimplycapitulatetosomedogmaticallyheldsetof“scientific”views.Thespiritofscienceisinitsmethod,notinanyparticularhistoricalsetofbeliefs.6.Inhisessay“EvolutionaryLove,”Peircedescribesbothwaysofem-ployingagapasm.Initscosmologicalformitservesprimarilytomarkoutaworldinwhichneithersheercontingencynorsheerdeterminismreigns.Initssocialuse,theoneinwhichIamhereinterested,agapasmisofferedasamodelforhowhumanitymightworktowardsitsownperfection.SeeCP6.287–6.317.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nsandrarosenthal8Peirce’sPragmaticAccountofPerception:IssuesandImplicationsPeirce’sunderstandingofperceptioniscrucialinsituatinghisphi-losophywithinabroadrangeofissues.YetacursoryreadingofPeirceseemstoindicatethatwhathesaysaboutperceptionisbothincom-pleteandinconsistent,leadingbothtoanearlyneglectofhisaccountofperceptionandtowidelyvaryinginterpretationsofhisclaims,asinterestinthembegantogrow.ThefollowinganalysisofPeirce’sviewofperceptionwilltrytoresolvetheambiguitiesbybringingintofocusthesystematiccompletenessofPeirce’sunderstandingoftheprocessofperceivingandtheobjectofperception,atthesametimeshowingitsrelevanceforarangeofcontemporaryissues.Peirceholdsthatthescientificmethodistheonlygenuinemethodoffixingbelief,foritistheonlymethodbywhichbeliefsmustbetestedandcorrectedbywhatexperiencepresents(CP5.384).Andtheveryfirststageofscientificinquiryrequireshumancreativity.Peircecallstheprocessofcreativehypothesisformation‘abduction’todistinguishitfromtheinductiveprocessofdatacollection.HerejectstheclaimsofBritishEmpiricism,thatknowledgebeginswithfirstimpressionsofsense.Healsorejectstheclaims,suchasthatputforthbyDescartes,thatitbeginswithimmediatecognitionsorindubitableintuitions.Allknowledgebeginswithperception,butperceptionisnotthehavingofbrutegivens.Rather,thereisacreativeelementinperceptualawareness,aninterpretivecreativitybroughtbytheperceiver.ExactlyhowdeeplythisinterpretiveelementrunsinPeirce’sac-countofperceptionisopentoconfusionbecauseoftheseeminglycontradictorycharacterizationshegivesofthevarious“ingredients”inperceptualawareness.Thus,whileinrecentyearsagooddealofattentionisbeginningtobefocusedonPeirce’sunderstanding193CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n194sandrarosenthalofperceptualjudgmentsinrelationtothecontemporaryissueoffoundationsforknowledge,therearewidelyconflictingclaimsastowhetherheisafoundationalistoranantifoundationalist.TheseconflictingclaimstendtostemfromconflictingunderstandingsofPeirce’saccountofthepercept,theperceptualjudgment,andwhathecallsthepercipuum.Forexample,ChristopherHookwayholdsthatthepercipuumfusestheperceptandperceptualjudgmentintoasinglewholeandrepresentsPeirce’sattempttorejectfoundation-alism,thoughheneverattemptstoexplicatewhythisisso.1Ontheotherhand,DavidGruender,inhisdiscussionoftheinterrelationofobservationandtheoryinPeirce’sphilosophy,tendstointerpretwhatisgivenintheperceptalongfoundationalistlines.2Similarly,JeremiahMcCarthyarguesthatPeirceisledtoafoundationalistpo-sitionbecauseperceptualjudgmentsareimmunetodoubt.3CarlHausmanrecognizesdualmeaningsoftheperceptualjudgmentinPeirce’sphilosophy,butlinksthemtoaseeminglyunivocalmean-ingofthepercipuum.4TheensuinganalysiswillattempttobringintofocusPeirce’sunderstandingofthedualsensesofthepercept,theperceptualjudgment,andthepercipuuminthelogicofpercep-tualawareness,inordertoshowtheuniqueandfruitfulpositionheholds.Theperceptisthatsensoryelementwhichispresentedinper-ceptualawareness.Itinturninstigatestheformationofthepercep-tualjudgment,whichinvolvesacreativeinterpretationplacedupontheperceptorpresentedsensecontentor,inotherterms,anabduc-tionwhichyieldsahypothesisastowhatthecontentis(CP5.115).Peirceisnothereassertingthatwefirstobservetheperceptandthenproceedtointerpretitinajudgment.Norisheassertingthattheperceptandthecontentoftheperceptualjudgmentarephysi-cally,metaphysically,ornumericallydistinct.Hedoesnotholdthatwhatweareawareofissensedataofsomesortratherthanaphysi-calobject.Rather,theperceptasinterpretediswhatweimmediatelyperceiveandisthereality(CP5.568).ButPeircecharacterizesboththeperceptandtheperceptualjudgmentinquitedifferent,oftencontradictory,ways.Forexample,ontheonehandhecharacterizesperceptsasspecificallyindividual(CP7.633),asinsistentandforcingthemselvesuponus(CP6.541).Atthesametimeheholdsthatper-ceptsincorporategenerality(CP4.542).andinvolveprocesseswhichare“forallintentsandpurposesmental”(CP7.624).AndconflictingCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception195claimsaboundinPeirce’sdepictionsofperceptualjudgments.Heattimesclaimsthattheyareinfalliblytruestatementsaboutwhatisperceived(CP5.55),butalsostatesthatperceptualjudgmentsarefal-lible,forthereisnoinfallibletruth(CP5.544).Further,heattimesholdsthatperceptualjudgmentsyieldclaimsaboutthewaythingsappear(CP7.626),andatothertimesthattheyyieldclaimsaboutthewaythingsare(CP7.636n).Theclarificationofthewaytheseseeminglyconflictingstate-mentsrefertodifferentsensesoftheperceptandtheperceptualjudg-mentoperativeinperceptualawarenesscanbestproceedbyturningtothenatureofthepercipuum.ItwillbeseenthatPeirceusestheterm‘percipuum’intwodifferentsenses,awidesenseandanarrowsense.And,whenhisvariouscharacterizationsexemplifiedaboveareunderstoodinthelightofhisbroadandnarrowsensesofthe‘percipuum,’oroutcomeoftheperceptualjudgment,itcanbeseenthatPeirceusesboth‘percept’and‘perceptualjudgment’5inawideandnarrowsense,correspondingtothetwosenseofthe‘percipuum’:thepercipuuminitswidesenseasactuallyexperiencedintheflowofexperience(CP7.657;7.676),andinitsnarrowsenseisanana-lyticalorlogicalabstractionindicatinga“stopping”pointpriortoanticipatoryexpectationswithinthepercipuum.Theterm,percipuum,withintheperceptualsituationseemstohavebeenfirstusedbyPeirceinamanuscriptof1903(CP7.642–81).Hethereproposes“toconsidertheperceptasitisimmediatelyinterpretedintheperceptualjudgment,underthenameofthe‘per-cipuum’”(CP7.642).ThoughthetermisintroducedlateinPeirce’scareer,andmayseemtomanytoberepresentativeoftheunneces-saryobscurityoftenfoundinhiswritings,itwillinfacthelpbringintofocusthedistinctionstowardwhichheseemedtobegropingthroughouthiscareer,aswellasthenovelpositiontheseembody,inawaywhichwouldbeobscuredormisinterpretedbytheuseofmoretraditionalphilosophicalterms.Peircestatesthat“ThereisnoPercipuumsoabsoluteasnottobesubjecttopossibleerror”(CP7.676).However,Peirceishereus-ingtheterm“percipuum”initswidesense,asenseintendedinthecontexttoshowthattimeisnotcomposedofaseriesofdiscretein-stants.Ashetherestates,“Thepercipuumisnotanabsoluteevent”butratheroccursinaspanoftimewhichincludesmemoryandex-pectation(CP7.657).HerePeirceisconcernedwithemphasizingtheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n196sandrarosenthalcontinuityoftimeorthepassingtemporalspreadinwhichtheper-cipuumlookstoboththepastandthefuture.Yet,withinthewidesenseofthepercipuum,Peircemakesseveraldistinctionswhichareabstractionsforthepurposeofanalysis.Thepercipuuminitswidesense,asitactuallyoccursintheso-calledspeciouspresent,containsseveralanalyticaldistinctions,6oneofwhichisthepercipuuminitsnarrowsense(CP7.648).Theseareintendedtoindicatethattheverygraspofsensorycontentasrecognizedcontentwhichinturncanactivatehabitsofresponseinvolvesinterpretiveelements.Forthisgrasprequiresasynthesizedcriterionforgraspofpresentationasrepeatablecontent.Indeed,itistheinterpretiveprocessofgraspingtheindividualuniqueperceptasarepeatablecontent,as“thatwhichhasbeenseenbeforeandmaybeseenagain,”whichallowsfortheactivationofhabitsofanticipationinvolvedintheperceptualjudgmentinitswidesense.Thisprimitiveinterpretiveprocessinvolvestheformationofaper-ceptualjudgmentinthenarrowsense,yieldingapercipuumastheoutcomeofitsinterpretiveprocess.Itisthispercipuumwhichinturnbecomestheperceptfortheperceptualjudgmentinthewidesense,yieldingthepercipuuminthewidesense.Theonlytypeof“referencetofutureexperience”implicitinthepercipuuminitsnarrowsenseisthepossibilityoffuturepresentationsofgraspablecontentwhich,forpurposesofclarity,canperhapsbestbetermed“possibilityofrepetition”ratherthanpossibilityoffutureexperi-ence.Itprovides“samenessoftype”(MSp.740)but“containsnoassertionofafact”(MSp.740).Itisthatbywhichoneisabletograspacontentwhich,inbecomingarepetitionofpreviouscontent,caninstigateanticipationsor“activate”ahabit.Thispercipuumistheoutcomeoftheperceptualjudgmentinitsnarrowsenseandyields“repeatablecontent”whichservestoac-tivatehabitthough,asananalyticalstoppingpoint,itprovidesnoanticipationoffutureexperience.Theperceptualjudgmentinitsnarrowsenseistheprimitiveabductivehypothesisofapresentrepe-titionofpastexperientialcontent,andthecontentinfactbecomesarepetitionofpreviouslyexperiencedcontentsonlyastheperceptualjudgmentdoesassimilateittothosecontentsintheabductivepro-cessofrecognition.Or,asPeircenoteselsewhere,“Thepercipuumisarecognitionofthecharacterofwhatispast”(CP7.677).Theper-cipuumisgraspedbymeansofthecharacterofwhatispastand,asanCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception197analyticalabstractionforpurposesofanalysis,itcontainsnorefer-encetofutureexperience.AsPeircestates,“Whattwothingscanbemoredisparatethanamemoryandanexpectation?”(CP2.143).Yet,whiletheperceptualjudgmentinitsnarrowsensedoesnotincludeanticipationsoffuture-activities,theverycharacterofpastassim-ilationincorporatesassimilationofthatwhichhasbeenpartiallyconstitutedinaction.Eventhepercipuumastherepeatablecontentwhichactivateshabitisnottheproductofapurelypassiveassimi-lation.Humancreativityandactivityenterintoeverydimensionofperceptualawareness.Thusasananalyticalstoppingpointintheanalysisofpercep-tiononefindsarecognizedcontentorpercipuumwhichistotallydevoidofreferencetofutureexperienceyetwhich,asarecognizedcontent,isdependentuponandisinfacttheoutcomeofajudgment–theperceptualjudgmentinitsnarrowsense.Thus,the“sensingdi-mension,”asitentersthestructureofhumanawareness,isnotanabsolutegivenbutataken.Thecontentofthisanalyticalstoppingplaceisdifficulttoindicate,foreventhenarrowpercipuuminitspuritycanbeexpressedneitherinthelanguageofobjectivitynorinthelanguageofappearing.Itisaspontaneousqualitativeimmediacyinthat“itcannotbearticulatelythought,”foritlosesits“character-isticinnocence”intheveryattempt.Yetthiscontentisthereasananalyticalelementoftheperceptualsituation,servingasthebasisforourfullpredictivemeaningsassertedintheperceptualjudgmentinthewidesense.Theattempttoexplicitlygraspthispercipuumyieldsapprehen-sionofappearance.Butevenappearancesareapprehendedasappear-ancesofobjectivitiesandexpressedthroughthelanguageofobjec-tivitiesbecauseappearances,asgraspedthroughinterpretivehabits,reflect,intheirveryemergence,thestructuringsofobjectivities;theyreflectthestructuringsoftheveryanticipationswhichoneisat-temptingtowithholdinfocusingontheappearancequaappearance.Ordinarily,whenactionsoranticipationsarenotinhibitedbecauseofaquestionablesituation,weperceiveappearingobjects,notap-pearances.Whenaninterpretiveexpectationdoesnotwork,when“whatisthere”requiresreinterpretation,thenthefocusisturnedto“whatappears.”Appearancesarenotthebuildingblocksofpercep-tion,asheldbyphenomenalistpositionsofvarioussorts,butratheralevelbroughtaboutbyachangeoffocuswhenaproblemarises.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n198sandrarosenthalAnd,aswereinterprettheobjectivesituation,wenolonger“see”theappearanceofwhatwemistakenlytooktobethere,butanewappearingobject.Thus,appearancesthemselvesincorporatemean-ings,irreduciblemeaningswhichfunctiontoverifytheapplicationofobjectivemeaningsintheongoingcourseofexperience.Thefol-lowingfocusonPeirce’sclaimscanbestbeunderstoodintermsoftheapprehensionofappearancesastheclosestonecancomeintheongoingcourseofexperiencetothepercipuuminitsnarrowsenseastheabstractionofa“stoppingpoint”inthelogicalanalysisofperception.ThoughPeircespeaksofthepercipuumastheperceptimmedi-atelyinterpretedintheperceptualjudgment,heelsewherestatesthat“PerhapsImightbepermittedtoinventtheterm‘percipuum’toincludebothperceptandperceptualjudgment”(CP7.629),since“thedifferencesaresominuteandsounimportantlogicallythatitwillbeconvenienttoneglectthem”(CP7.629).Asheclarifies,“Theforcefulnessoftheperceptualjudgmentsfallsshortofthepureun-reasonablenessoftheperceptonlytothisextent,thatitdoesprofesstorepresentthepercept,whiletheperfectionofthepercept’ssurdityconsistsinitsnotsomuchasprofessinganything”(CP7.628).Thepercept,initssurdity,isinfalliblebecauseitdoesnotprofessany-thing.Andtheperceptualjudgmentisinfalliblebecause“tosaythattheperceptualjudgmentisaninfalliblesymptomofthecharacteroftheperceptmeansonlythatinsomeunaccountablemannerwefindourselvesimpotenttorefuseourassenttoitinthepresenceofthepercept,andthatthereisnoappealfromit”(CP7.628).Thustheperceptbyitselfprofessesnothing,whiletheperceptualjudgmentprofessesthepresenceoftheperceptasarecognizedcontent.Bothareinfalliblebecauseneitherprofessestheexistenceofanyobjectivefactortheanticipationofanypossibilitiesoffutureexperience.Fu-tureexperiencecannotshowtheperceptualjudgmentinitsnarrowsensetobeinerror,sinceitmakesnoreferencetofutureexperience.ThusPeirce,inreplyingtotheobjectionthataperceptualjudgmentisnotsoutterlybeyondallcontrolorcheckashesays,sinceitmayberevised,statesthatthe“perceptualjudgmentcanonlyrefertoasingleperceptwhichcanneverre-exist;andifIjudgethatitappearsredwhenitdidnotappearred,itmust,atleastbeacknowledgedthatitappearedtoappearred”(CP7.636).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception199Indistinguishingperceptandperceptualjudgment,Peirceob-servesthatperceptualjudgmentsareasunlikethepercept“astheprintedlettersinabook,whereaMadonnaofMurilloisdescribed,areunlikethepictureitself”(CP5.54).Thisexamplemayeasilyleadonetoviewtherelationbetweenperceptandperceptualjudgmentasanalogoustotherelationbetweennonlinguisticexperienceandlanguage.ThoughPeirce’sexampleisillchosen,heclearlyindicateselsewherethattheperceptualjudgmentis“ahighergradeoftheop-erationofperception”(CP7.634).ThedistinctionbetweentheperceptualjudgmentinitswideandnarrowsensescangainfurtherclarityifweturntoPeirce’sdiscussionofthefallibilityofmemory,whichisperhapsnotsoclearcutasitfirstappears.Nowletustakeuptheperceptualjudgment“Thiswaferlooksred.”Ittakessometimetowritethissentence,toutterit,oreventothinkit.Itmustrefertothestateoftheperceptatthetimethatit,thejudgment,begantobemade.Butthejudgmentdoesnotexistuntilitiscompletelymade.Itthusonlyreferstoamemoryofthepast;andallmemoryispossiblyfallibleandsubjecttocriticismandcontrol.Thejudgment,then,canonlymeanthatsofarasthecharacteroftheperceptcaneverbeascertained,itwillbeascertainedthatthewaferlookedred.(CP5.544;italicsadded)InjustwhatsensedoesPeircemeanthatmemoryissubjecttocrit-icismandcontrol?AsPeircenotesintheparagraphimmediatelyfollowingthequotationabove,“Perhapsthemattermaybestatedlessparadoxically”(CP5.545).AndPeirceproceedstodothisinadiscussionwhichconcludesthat“tosaythatabodyishard,orred,orheavy,orofagivenweight,orhasanyotherproperty,istosaythatitissubjecttolawandthereforeisastatementreferringtothefuture”(CP5.545).ItisevidentthatintheabovediscussionPeircehasswitchedfromtheterm“looks”totheterm“is”andthatthecharacteroftheperceptisascertainedbyreferencetothefuture;inthisway,then,bythetestoffutureexperience,memoryissubjecttocriticismandcontrol.However,inconfoundingthepercipuuminitswidesenseandapprehensionsofappearancesPeircedoesmakeanimportantpointconcerningappearances.Thoughheneverexplicitlydiscussesthepoint,itssignificanceistobefoundamongtheconfusedstatementsCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n200sandrarosenthalofthetwopassagescitedabove.“Allmemoryispossiblyfallible,”yet“sofarasthecharacteroftheperceptcaneverbeascertained,itwillbeascertainedthatthewaferlookedred”(CP5.544).IfonedoesnotgoontoPeirce’s“lessparadoxical”statementsoftheis-sue,thiscanbetakennotasaconfusedstatementconcerningfutureverificationbutratherasastatementconcerningthenatureofthecontentofthe“seeming”statement.If,asPeirceexplicitlystates,allmemoryisfallible,itisdifficulttoseehowthatwhichisindi-catedbytheseemingstatementcanprovidethebedrockofcertainty.Thereliabilityofmemorymustbequestionednotonlyinregardtowhatcanbepredicatedbasedonthepresentcontentbutinregardtotherecognitionofthepresentcontentitself.Memoryisinvolvedintheveryrecognitionofthatcontentwhichhasbeenseenbeforeandmaybeseenagain,agraspwhichallowsthecontenttobecomethebasisforpredictivemeaning.Thisbasis,then,isnotcertainbutrathersubjecttotheerrorofmemoryandincapableofprovidinganindubitablebedrockofempiricalknowledgeinanyfoundationalistsenseoftheterm.Whatisprovidedisnottheabsolutecertaintyoffoundationalistclaimsbut“pragmaticcertainty.”Theapprehensionofanappear-anceisindubitableinthesensethatitsfalsityisinconceivable.Itisbeyondconceivabledoubt,becausetodoubtitinthesensethatonethinksitmaybeprovenwrongissenseless;indeed,literallyso.Todoubtitistoputintoquestionsomethingforwhichthereisnotoolforgetting“behind”or“beneath”ittocompareitwithanythingmorefundamental.Forus,itmustitselfbethefinalcourtofappeal.Theapprehensionofanappearanceisnotcertainlytrueasopposedtopossiblyfalse.Itis“certain”inthesensethatneithertruthnorfal-sityisapplicabletoit.Theperceptualjudgmentinitsnarrowsensecannotevenbelabeledcertainlycorrectasopposedtopossiblyin-correct.Thereisnocorrectorincorrectrecognitioninvolvedatthislevel,forwhatthepercipuumisisdeterminedonlyinitsrecognitionandcanbedeterminedinnootherway.Itbecomesa“repetition”ofpreviouscontentsonlybybeingassimilatedtothosecontentsintheperceptualjudgment.Inrelationtomoretraditionalviews,thisconclusionissurelymoreparadoxicalthantheconclusionthattheperceptualjudgment,initswidesense,isfalliblebecauseitcanbeprovenwrongbyreferencetofutureexperience.PerhapsthenoveltyoftheformerCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception201conclusion,coupledwithhisownfailuretoclarifytheconceptualdistinctionstowardswhichhewasgroping,ledPeircesubtlytoswitchinhisattempttomakehispositionseem“lessparadoxical.”However,apartfromsuchspeculation,itdoesseemthatPeirce’sconfuseddiscussionstemsfromacarelessslippingbackandforthbetweentwoconceptswhichhelaterclearlydistinguished.Peirce’suseoftheperceptualjudgmentinitsnarrowsensehasbeendiscussedindetailbecauseitisthe“moreparadoxical”andlessemphasizedofthetwosenses.ThatthisissoisnodoubtduebothtoPeirce’sbriefandlateexpositionofadistinctioninlevelsofjudg-mentandtotheassumptionthatwhatistheoutcomeofajudgmentmustbecapableofbeingshowntobetrueorfalse.Bernsteinpointsoutthatiftheperceptualjudgmentcannotbetrueorfalseitisnotajudgment,7andagain,thatifthereisahypotheticalelementinvolvedineveryperceptualjudgment,theneveryperceptualjudgmentisfal-libleandsubjecttofuturetests.8Conversely,itisMcCarthy’saccep-tanceoftheclaimthatperceptualjudgmentsareimmunetodoubtwhichallowshimtoholdthatPeircemakesuseofanobservation–theorydistinctionwhichturnshimintoafoundationalist.9Theperceptualjudgmentinitsnarrowsensedoeshaveahypo-theticalelement,10forthejudgmentisahypothesisthatacontentis“thesameas”thatwhichhasbeenseenbeforeor“appearsas”(W1,471).However,theaboveanalysishasshownthattheperceptualjudgmentinitsnarrowsenseisnotfallibleandsubjecttofuturetests,foritmakesnoreferencestofutureexperience.AsPeircestates,itssurdityisalmostcomplete.Itcannotbecharacterizedastrueorfalsefor,asindicatedabove,wehavenomorefundamentalperceptualtoolbywhichtoassertitstruthorfalsity.Todenytheterm“judgment”tothatwhichcanbecharacterizedasneithertruenorfalseisonewaytoavoidthefrequentconfusionswhichpervadePeirce’sanaly-sisofperceptionbecauseofhisdualusesofthisterm,anditallowsonetofollowamoreconventionalterminologicalprocedure.Itisnot,however,toofferanobjectiontothatconceptwhichPeircein-tendsbytheterm.11Bycharacterizingthisprimitivesynthesisintermsofaperceptualjudgmentwhichyetcannotbetrueorfalse,hebringshomemoreforcefullytheradicalnoveltyofhisrejectionoffoundationalism.Whatis“given”atthemostfundamentallevelofperceptualawarenessisinfacta“taken,”anditincorporatesboththenatureofthetakingandthenatureofwhatistaken.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n202sandrarosenthalTheperceptualjudgmentinitswidesenseisindubitable,notinthesensethatthediscoveryofitsfalsityisinconceivable,foritstruthorfalsitymaybeascertainedbyfutureexperience,butratherinthesensethattherearenopositivegroundstostimulatedoubtpresentintheperceptualsituation.AsPeircehasstressedinhisrejectionofDescartes’universaldoubt,wecannotfeigndoubt(CP5.265).Unlesssomepositivegroundfordoubtisgivenintheperceptualsituation,perceptualjudgmentsandcertainvaguebeliefs(CP5.442)mustbetakenasindubitable,fortheycannotseriouslybedoubted–thoughtheyareeminentlyfallible,sincetheyaresubjecttothetestoffutureexperience.Thus,wearriveinabroadsenseatPeirce’s“fallibilism.”Thefallibilityoftheperceptualjudgmentinitswidesense,asitactuallyoccursinthepassageofthepresentandmakesaclaimaboutanobjectivestateofaffairs,liesinthefactthatitwillbere-jectedasfalseifitdoesnotfitwithfutureexperiencesanticipatedbyitsclaim.Theindubitabilityatthislevelentersinthesensethattheformationoftheperceptualjudgmentcannotbecontrolledandisbeyondlogicalcriticisminitsformation.Whilewecannotcriti-callycontrolthejudgment,however,wecancriticizeitsresultsandconclude,basedonfutureexperience,thatitisfalse.Underlyingtheverypossibilityofthesecommonsenseindubitableswhichmayturnouttobefalse,therehasbeenseentolieanindubitabilitytowhichneithertruthnorfalsityisapplicable,whichis“pragmaticallycer-tain.”DavidSavanaptlycharacterizesthistypeofdistinctionwhenhenotesthedifferencebetweenindubitableswhichcannotconceiv-ablybedoubtedandindubitableswhicharesoonlybecausethereisnopositivegroundtostimulatedoubt.12Underlyingtheverypos-sibilityofthesetestableperceptualjudgmentswhichmayturnouttobefalse,however,therehasbeenseentobeanindubitabilitytowhichneithertruthnorfalsityisapplicable,whichis“pragmaticallycertain.”Yet,thispragmaticcertaintyistheproductofinterpretiveactivity,itisabouta“taken”ratherthana“given.”Peirce’sstressoncertitude,then,farfromindicatinghisalliancewithfoundation-alism,infactleadstotheradical,novelnatureofhisrejectionofit.Ithasbeenseenthatperceptionisinfusedwithmeaningsstruc-turedbypossiblepurposiveactivity.Theroleofpurposiveactivityinthoughtandtheresultantappealtorelevanceandselectiveempha-siswhichmustultimatelybejustifiedbyworkabilityarekeytenetsCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception203ofPeirce’sunderstandingofthenatureofexperienceasexperimen-tal.And,interpretiveactivitybeginsatthemostprimordialleveloftheformationofrepeatablecontentwhichcanactivatehabitsofanticipation.Whatisyettobeexaminedisthewayinwhichthisunderstandingofperceptioninvolvestemporality.Peircemakesclearthatverificationinstancesarethemselvespos-sibleonlybecauseourperceptualmeaningscontainactivityandtem-poralreferenceintheveryheartofitsinternalstructure.AsPeircestresses,“Thereisnospanofpresenttimesoshortasnottocon-tain...somethingfortheconfirmationofwhichwearewaiting”(CP7.675).Butthis“peculiarelementofthepresent,thatitconfrontsuswithideaswhichitforcesuponus...issomethingwhichaccumu-latesinwholesoftimeanddissipatesthemoreminutelythecourseoftimeisscrutinized”(CP7.675).Or,asPeircesumsupthissenseoftemporalityattheheartofperceptualawareness,“ifwewishtoknowwhatthepercipuumofthecourseoftimeis,allwehavetodoisabstainfromsophisticatingit,anditwillbeplainenough...”(CP7.649).Thistemporalawarenesspresentinourcommonsenseperceptionsisalsothebasisforoursenseofthecontinuitywithinex-perience.Thusinaclaimparalleltotheaboveone,hestatesthat“Solongaswetrusttocommonsense,thepropertiesofatruecontinuumareamatterofcourse,”whilethroughourabstractelaborations“wefounderfromquagmireintoquicksand”(MS137,p.10).Hestressesthatthistemporallyrootedpercipuummakesnominalisticmaximsfutile.Asheelaborates,“Butitisremarkablethatincasewedonotacceptthepercipuum’sownaccountofitself...thenitwouldseemthatthereisnothingthatempiricaltruthcanmeanexceptaccor-dancewithwhatisgiveninthoseinstants,whichinthiscase,innowaytestifyconcerningoneanotherorinanywayrefertooneanother”(CP7.671).Thesametemporallyrooted,dispositionallyorganizedsenseofexpectationwhichisattheheartofperceptualclaimsisattheheart,also,oftheverypossibilityofpragmaticev-idenceandisrootedinourcommonsenseperceptualawarenessofthesenseoftemporalityandcontinuity.ThusPeircestatesofthepragmatist,“ThathewillhavenodifficultywithThirdnessisclearenoughbecausehewillholdthatconformityofactiontogeneralin-tentionsisasmuchgiveninperceptionasistheelementofactionitself,whichcannotreallybementallytornawayfromsuchgeneralpurposiveness”(CP5.212).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n204sandrarosenthalUntilthispointthemajoremphasishasbeenontheinterpretiveelementswhichenterintoourperceptionoftheworldinwhichwelive,evenattherudimentarylevelofwhatis“given”inexperience.ItmaybegintoseemthatPeirce,inrejectingfoundationalism,hasindeedsidedwiththeotheralternative.ButPeirce,withequalforce-fulness,rejectstheclaimsofantifoundationalism,forheholdsthatthereisa“hard”realitywithwhichweinteractandwhichprovidestheworkabilityforourinterpretations.ForPeirce,ourperceptualawarenessisadirectgrasp,thoughnotaspectatorgrasp,ofahardexternalreality.Tounderstandthisinteractionbetweenthecreativeandconstrainingforcesinvolvedinperceptualawareness,itwillbehelpfultoturntoPeirce’sunderstandingoftheworldthatisgraspedinourperceptualawareness.Peirceneverexplicitlyclarifieshisunderstandingof“therealworld,”thoughhereferstoitfrequentlythroughouthiswritings.Onethingthatisclearisthat“therealworld”fitsinadequatelywithintheconfinesofthelabels,‘realism,’‘idealism,’or‘phenome-nalism,’foritisadistinctivelypragmaticworld.Thefollowingdis-cussionproposestoshowthatPeirce,inrejectingtheroleofhumansasspectators,inunderstandingexperienceasaunityofinteractionbetweenhumansandthatfacticitywhichgivesitselfwithinexperi-ence,canholdatoncethattherealworldistheperceivedworld,thattherealworldhasanindependencefrommind,andyetthattheper-ceivedworldispartiallydependentuponthenoeticactandisthusrelativeinitsnaturetothemind.Thesupposedincompatibilityofthesethreecharacteristicsoftherelationofperceptiontotheworldstemsthefailuretoradicallyandonceandforallrejectthepresup-positionsofaspectatortheoryofknowledge.13Peirce’sabsoluteandradicalrejectionofthespectatortheoryofknowledgegivesriseto,andisinturnbroughtintoclearerlightby,hispragmaticconceptofworld.ThattherealworldistheperceivedworldisclearlyindicatedbyPeirceinseveralsuccinctpassages.Hestatesthat“Therealworldistheworldofsensibleexperience”(CP3.527);or,inotherterms,therealworldistheworldof“insistentgeneralizedpercepts”(CP8.148),whicharenotrepresentativeofanyunderlyingrealityotherthanthemselves14(CP2.143).Suchaworldisaconsistentsystemoffactsrigorouslyobeyingthelawsofnoncontradictionandexcludedmiddle,for,ashestates,“Dichotomyrulestheidealworld”(CPCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception2053.529),and“itispartoftheprocessofsensibleexperiencetolocateitsfactsintheworldofideas”(CP3.527).Suchagraspingofthesensibleworldintermsofasystemofideasisoftheveryessenceofthesensibleworld.AsPeircestresses,“ThisiswhatImeanbysayingthatthesensibleworldisbutafragmentoftheidealworld”(CP3.527).Further,thesystemofideasormeaningslimitsthefactswhichmayoccur“intheworld,”for,asPeircestates,“Weknowinadvanceofexperiencethatcertainthingsarenottrue,becauseweseetheyareimpossible...thereisnoroomfor(them)eveninthatidealworldofwhichtherealworldisbutafragment”(CP3.527).Thus,whatcanoccur“intheworld”mustconformtothepossibilitiesallowedforbytheworldofideasorthesystemofmeaningsintermsofwhichweapproachit.Thisclaimdoesnotleadtoaconventionalism,forPeircestressesthattherealworldisthatspecialpartoftheidealworldwhichsufficientexperiencewouldtendtocompelustoacknowledgeashavingabeingindependentofwhatwearbitrarily,orwillfully,create(CP3.527).So,forPeirce,dichotomyrulestherealworld,becauseitrulestheidealworldofwhichtherealworldisapart.Yet,hisviewofthenatureoftherealasindependentofthehumanmodeofgrasp-ingitindicatesthatsuchharddiscreteexactitudesdonotexist,forreality,accordingtoPeirce,isacontinuumwhich“swimsininde-terminacy”(CP1.171–2)becauseofitsindefiniterichness.Forthisreason,heholdsthattheprincipleofcontinuity,whichpervadestheindependentlyreal,is“fallibilismobjectified”(CP1.171).Furthertheindependentlyrealasacontinuumofeventsispreciselythattowhichneitherthelawofnoncontradictionnorthelawofexcludedmiddleisperfectlyapplicable.15Therelationbetweenthecontin-uumofqualitativeeventswhichconstitutesthecharacterofthemetaphysicallyrealindependentlyofthehumanmodeofgrasping,andthesystemoffactswhichconstitutestherealworld,findsitsanalogueinPeirce’sdistinctionbetweenanoccurrenceasaslaboftheuniverseinallitsinfinitedetail,andarealfactasthatwhichisextractedfromtheuniversebythepowerofthought(MS647,p.8).ThusPeirceholdsthattherealworldcanbecharacterized,also,astheworldofperceptualfacts,for“whatIcarrywithme”oftheper-cept“istheperceptualfacts”(CP2.141).HereliesthesignificanceofPeirce’sclaimthat“Nature,incon-nectionwithapicture,copy,ordiagramdoesnotnecessarilydenoteCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n206sandrarosenthalanobjectnotfashionedbyman,butmerelytheobjectrepresentedassomethingexistingapartfromtherepresentation”(CP3.420).Mill’sfailuretorecognizethismind-relatednessofworldlynature,accord-ingtoPeirce,ledhimastrayinhisanalysisofthe“uniformityofnature”(CP6.67).Peirceindicatestheabovepositionfromaslightlydifferentdirectioninhisclaimthat“Thereisnothingwhichisinitselfinthesenseofnotbeingrelativetothemind,thoughthingswhicharerelativetotheminddoubtlessare,apartfromthatrela-tion”(CP5.311).Orasheelaborates,a“this”isanobjectselectedbyasubjectfromthecontinuumofpossibility(MS942,p.16).Real-ityindependentofourthinkingexertsaninfluenceonourwaysofthinkingaboutit,butwhatfactsandobjectsitcontainsispartiallydependentupontheconceptualframeworkintermsofwhichper-ceivedobjectsandfactscanemergewithinthebackdropofaworld.Indeed,accordingtoPeirce“ExternalFact”canchangeinaccordancewiththewayhumanminds“feel,think,orsuffer”(MS642,p.16).Peirceoffersahelpfulclarificationabouthislimitedintentionsinhisnumerousstatementsconcerningtheindependenceofrealobjects,claimingthat,therealobjectcanbe“anobjectshapedbythink-ing...;butsofarasitisReal,itisnotmodifiedbythinkingaboutit(MS634,p.9).Theaboveanalysishasattemptedtoshowthattheperceivedworldisontologicallyonewithindependentrealityasaninfinitelyrichcontinuumofqualitativeevents.Yet,theperceivedworldisde-pendentuponthemeaningsystemwhichgraspsinawayinwhichrealityasindependentisnot.Theperceivedworld,thoughcon-crete,isnonethelessselectiveinthesensethataworld,asthecon-cretecontentdenotedbyasystemofmeanings,isawayinwhichtheconcretenessofrealitycanbedelineatedor“fixed.”Asystem,oncechosen,limitsthealternativespossiblewithinit,butalterna-tivesystemsmaybepossible.AsPeircenotes,“Trulynaturalclassesmay,andundoubtedlyoftendo,mergeintooneanotherinextricably”(CP1.209),andthusboundarylinesmustbeimposed,althoughtheclassesarenatural(MS427,p.40–1).Thecontinuityisthere;wherethe“cut”isimposedis,inpart,ourdecision.Liketheboundarylinesofnaturalclasses,the“boundarylines”thatconstituteourworldofperceptualexperiencemayhavebeendifferentlydrawn,givingrisetodifferentpossibilitieswithintheworld.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception207Whatappearswithinexperience,then,isalsotheappearanceoftheindependentlyreal;thereisnoontologicalgapbetweenappear-anceandreality.AsPeirceobserves,“Synechism...willnotadmitasharpsunderingofphenomenaandsubstrates.Thatwhichunderliesaphenomenonanddeterminesittherebyis,itself,inameasure,aphenomenon”(CP7.629).Further,itisatthesametime“tome”towhomitappearsandreflectsmyintentionallinkwiththeexternallyreal.ThusPeircecansaythat“Perhapsitmayreconcilethepsychol-ogisttotheadmissionofperceptualjudgmentsinvolvinggeneralitytobetoldthattheyareperceptualjudgmentsconcerningourownpurposes”(CP5.166).TheepistemicandontologicalunityofthesetwodimensionscanbeseenfromPeirce’spositionthatthoughthegeneralityofperceptualjudgmentsreflectsourownpurposes,yet“sincenocognitionofoursisabsolutelydeterminate,generalsmusthavearealexistence”(CP5.312).Peircefurtherindicatestheabovepositioninhiscrypticclaimthat“Theinkstandisarealthing.Ofcourseinbeingrealandexternal,itdoesnotintheleastceasetobeapurelypsychicalproduct,ageneralizedpercept”(CP8.261).ForPeirce,theseare“twosidesofthesameshield”(CP1.420).Or,asheeloquentlysummarizeshisposition,though“everythingwhichispresenttousisaphenomenalmanifestationofourselves,”this“doesnotpreventitsbeingaphenomenonofsomethingwithoutus,justasarainbowisatonceamanifestationbothofthesunandoftherain”(CP5.283).ForPeirce,then,perceptualfactsattheirverycoreemergeneitherfrommindalonenorfromthedynamicrealityoftheuniversealone,butratherfromtheinteractionofthetwowhichconstitutesexperience.Thisunificationundercutsthedi-chotomyoffoundationalismornonfoundationalismandalongwithit,thecloselyrelateddichotomiesofrealismorantirealismandob-jectivismorrelativismsinceeachofthesedichotomies,initsownway,representsthealternativesofanabsolutegroundingofknowl-edgeorskepticism.Thisinteractiveunityattheheartofexperienceclearsthewayforafullerunderstandingoftheverificationofourperceptualclaimsintheongoingcourseofexperience.TheextentoftheradicalconflictofinterpretationsconcerningPeirce’stheoryoftruthintheliteratureisperhapsbestcapturedinRobertAlmeder’sclaimthattheliteratureonPeircecontains“nofewerthanthirteendistinctinterpretationsofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n208sandrarosenthalPeirce’sviewsonthenatureoftruth.”16WithinPeircescholarship,theacceptanceofconvergenceandthefinalultimateopinionisnotdependentonone’sstanceintherealist–idealistcontroversy,thoughtheunderstandingofthenatureofthefinalultimateopinionasthattowardwhichinquiryonanysubjectwillconvergewillvaryaccord-ingtocamps.Thus,ashasbeenstatedfromthebackdropofacoher-encetheoryoftruth,thetruebedrockofpragmatismis“ultimatelytheentireframeworkofobjectivelogicandobjectiveidealism.”17Ontheotherhand,arealistinterpretationholdsthat“Theopinionreachedinthefinalopinion,unlikeopinionsreachedearlier,shallneverbeoverthrownalthoughthedegreetowhichthefinalopin-ioncorrespondstofactadmitsofindefinite,(butnotsubstantial)refinement.”18JohnE.SmithreferstoPeirce’stheoryoftruthasaconformtheory,whichhasadistinctadvantageovertheterm‘corre-spondence’inthatitcanavoidmuchofthehistoricalbaggageattach-ingtotheterm‘correspondence,’butitoperatesnonethelesswithinaframeworkofrealismandultimateconvergence.19Beforeexam-iningPeirce’stheoryoftruthinrelationtotraditionalalternativesofcorrespondenceorcoherence,itwillbehelpfulatthispointtoclarifythetypeofrealismwhichisintertwinedwiththecorrespon-dencetheoryoftruth,andtowhichPeirce’spragmaticpositionisopposed.ThiscanperhapsbestbeapproachedbyAlmeder’sespousalofPeirce’s“epistemologicalrealism.”HeproceedsbyshowingthatPeirceisnotaphenomenalistandnotanidealist,andthatPeirceoffersadefenseofbeliefintheexistenceofanexternalreality,are-ality,moreover,withwhichtheknowerisindirectcontact.20Withthesepointsthepresentinterpretationagrees.But,whatthisrealismalsoincludesforAlmeder,aswellasformostwhoaccepttherealistlabel,isthatthesenseinwhichtherealexternalworldweknow“isdependentonmindturnsouttobetriviallytrueandnecessaryforanyepistemologicalrealismwhereinitisanecessaryconditionthattheexternalworldbeknowable.”21Orassucharealismiselse-wherecharacterized,“Thereisaworldofobjectswhosepropertiesareneitherlogicallynorcausallydependentuponthenoeticactofanynumberoffiniteminds.”22PreciselywhatisdeniedbyPeirceisthe“furniturerealism”whichholdsthatthereisaworldofobjectsorfactswhosecharacterisinnowaydependentuponhumannoeticactivity.ItisthesegenerallyheldassumptionsassociatedwiththerealistlabelthataredeniedindenyingthatPeirceisarealist,for,asCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception209seenabove,theworldandtheobjectswithininarepartiallydepen-dentuponthenoeticactsoffiniteminds.Thus,whilePeircecannotbecalledanidealist23oraphenomenal-ist,neithercanhebeadequatelylabeledarealist.For,thoughPeirceholdsweareindirectcontactwithanexternal“brutelythere”real-itywhichlimitsourinterpretations,thusshowingheisnotacoher-encetheorist,yettherelationoftheknowertothisknownexternalrealitycannotbeunderstoodintermsofcorrespondence.And,al-thoughitmaywellbeanoversimplificationtosaythatcoherencetheoriesoftruthbelongtoidealismwhilecorrespondencetheoriesoftruthbelongtorealism,aninterpretationofPeirceasanepistemo-logicalrealistintheabovesenseindicatedbyAlmederandacceptedbymostothersusingthislabel,leadstotheviewthatatleasttheideallytrueandfinalopiniononanymatterwouldinvolvearela-tionofcorrespondence.24Tothequestion,whatalternativeremainswhenonerulesoutrealismaswellasidealismorphenomenalism,theansweris,thepragmaticalternative.Peirce’spragmatictheoryoftruthisultimatelyintertwinedwiththeentiregamutofhisuniquepragmaticepistemologyandmetaphysicswhichinterweaveinhispragmaticunderstandingoftheperceptualworld.BecauseforPeircetheherenessandnownessofeventsandtherealconnectionstheydisplayisindependentof,yetentersdirectlyintointeractionwith,ourconceptualizationsandthepossibilitiestheyallow,coherenceorconsistencyisnotasufficientcriterionforthetruthofempiricalassertions.Thereisanontologicaldimensiontowhatappearswithinexperiencewhichlimitsourinterpretationsintermsofworkability.25But,trueknowledge,evenideallytrueknowl-edge,couldnotbecorrespondence,forthenatureofourcreativelyinterpretive,interactivelinkwithreality,andthenatureofrealityasacontinuumwhich“swims”inindeterminacy,makestherela-tionofcorrespondenceliterallysenseless.Rather,inPeirce’swords,atruethoughtisonewhichanswers,whichleadstothoughtsinhar-monywithnature(MS934,p.24).Therelationof“answering”isultimatelytwo-directional.Realityanswersourquestions,andde-terminestheworkabilityofourinterpretations,butwhatanswersitgivesispartiallydependentonwhatquestionsweask,andwhatinterpretivemeaningsworkispartiallydependentuponwhichoneswebring.Truthisalwaysworldlytruth,forPeirceclaims“nothingelsethanaFactpossiblycanbea‘witness’or‘testimony’”(MS647,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n210sandrarosenthalp.26),andfacts,itwillberemembered,arealwaysrelativetotheframeworkofadiscriminatingmind.Yethestressesthatthewit-nessofafactisthereal,“sinceitistrulyinthatwhichoccurs”(MS647,p.9).Worldlytruthisthusperspectival,andotherperspectivesareal-wayspossible.Truthinvolvesconvergence,butconvergencewithinacommonworldwhichwehavepartiallymade,andcontinu-allyremakeinvariousofitsaspects.ThusPeirce,inspeakingoftruth,whetherscientific,moral,metaphysical,orcommonsense(CP5.565–8),statesthat“theperfecttruthofastatementrequiresthatitshouldinvolvetheconfessionthattheperfectdoctrinecanneitherbestatednorconceived”(CP5.565–8).Again,Peirceclaimsthatanessentialingredientoftruthincludesaconfessionofits“one-sidedness”(CP5.566).Thatthisisintendednotasafactuallimita-tiononpresentknowledgebutasatheoreticallimitationduetothenatureofknowledgeisfoundinPeirce’scomparisonoftheideallimitofconvergence,theidealofa“finalultimateopinion,”totheideallimitofpi.Itis“anideallimittowhichnonumericalexpressioncanbeperfectlytrue”(CP5.565).Itisanunattainableidealnotonlyinfactbutbytheverynatureofthatwhichsetstheideallimit(CP5.565).Thus,Peircecanpresentthefollowinghypotheticalsituation:“Supposeouropinionwithreferencetoagivenquestiontobequitesettled,sothatinquiry,nomatterhowfarpushed,hasnosur-prisesforusonthispoint.Thenwemaybesaidtohaveattainedperfectknowledgeaboutthatquestion.True,itisconceivablethatsomebodyelsewouldattaintoalikeperfectknowledgewhichshouldconflictwithours.Thisisconceivable”(MS409,p.112).Peircethengoesontosaythatthoughitistheoreticallypossibleitisnotpracti-callypossible“consideringthesocialnatureofman,”forwewould“comparenotes;andifweneverdocomparenotes,andnothirdpartytalkswithbothandmakesthecomparison,itisdifficulttoseewhatmeaningthereisinsayingwedisagree”(MS409,p.112).ThatPeirceisnotusingtheterm“perfectknowledge”inaloosecommonsensewaycanbeseenfromhisexplicitdistinctionbetweenitand“practicallyperfectbelief”(MS409,p.112).Thus,eventheidealofconvergencetoafinalultimateopinion,toperfectknowledge,isalwaysconvergencewithinanacceptedframe-workorperspective.And,therearealwaysotherandpossiblybetterwaysofcuttingintoreality,ofdelineatingthecontextwithinwhichCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception211convergencecanoccur.Thisisimpliedbytheverynatureofrealityasacontinuumwhichswimsinindeterminacybecauseofitsin-definiterichnesswhich“overflows”therigidiesofourinterpretivestructures.Thus,Peircestatesthatconvergencetowardonefinaltruthis“aregulativeprinciple,anintellectualhope,”andsucharuleofhopemustbefollowed,for“despairisinsanity”(CP1.405).Yet,evensucharuleofhope,the“cheerfulhope”whichanimatesthefollowersofscienceinvolves,inhiswords,“somethingapproximat-ing”only(CP3.432),forthe“indeterminate”natureofrealitymaymeanthatconcerning“theanswer,thatis,thefinalanswer...thereisnone”26(CP4.61).Theobjectswithinourworlddonotcopythein-dependentlyrealbutratheremergethroughourmodesofgraspingtheindependentlyreal.Nordothemodesofgraspingviawhichemergetheobjectivitieswithinourworldcopytheindependentlyrealbutrathertheyserveasconceptualtoolsfor“cuttingtheedges”oftheindependentlyrealcontinuumofeventswhich“swims”inindeter-minacy.Theideallytrueopinionwouldbethatopinionwhichwouldperfectlyworkinanticipatingpossibilitiesofexperience,andwouldworknotbecauseitadequatelycopied,butbecauseitadequately“cutinto”theindependentlyreal.Finally,theworldwithinwhichspecificmeaningsandbeliefsarise,andwithinwhichobjectsorfactsemergeforconsciousawareness,isnotacopyofanindependentre-ality,norisitidenticalwithanindependentrealityinitscharacterasindependent.Rather,suchaworldistheencompassingframeofreferenceorfieldofinterestoforganism–environmentinteraction,theultimatebackdropofrationalitywithinwhichemergingfactsaresituated.Wediscovertruthsaboutourworldonlybecausewehavefirstprescribedcontoursforourworld.Truebeliefsaretruebe-foretheyareactuallyverified,buttheverypossibilityofverificationemergesfromthebackdropofthetransformationoftheindefiniterichnessofrealityintoworldlyencounter.Truthistruthrelativetoacontextofinterpretation,notbecausetruthisrelativebutbecausewithoutaninterpretivecontexttheconceptoftruthismeaningless.WhatthisessayhasattemptedtoshowinsomedetailisthatforPeircetheperceptualfield,asitarisesinthecontextofhumanac-tivity,isanontologicallythick,resistingfieldofobjectswhichareessentiallyrelatedtotheinteractionalhorizonsofourworldandwhichallowfortheverystructureofthesensingwhichgivesaccesstothem.Initsemergence,theworldgroundsalllevelsofexperienceCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n212sandrarosenthalandknowledge,atthesametimegivingmeaningfulaccesstotheindependentrealityofthenaturaluniverse.Suchaviewundercutsalongtraditionofstandarddichotomieswhichareincreasinglyseentobeunsustainable.Notes1.Hookway(1985:166).2.Gruender(1983:181–287).3.McCarthy(1990:63–113).4.Hausman(1990:271).5.Hausman’sanalysisoftwomeaningsoftheperceptualjudgmentinPeirce’sphilosophydoesnotcorrespondwiththedistinctionbeingmadehere(Hausman1990:271–308).6.AllthesedistinctionsarediscussedbyRosenthal(1969)insomedetail,buttheothersneednotbeintroducedhere.7.Bernstein(1964:175).8.Bernstein(1964:173).9.McCarthy(1990:63–113).Hishighlyperceptiveanalysisgoesastrayherebecauseheremainstoogeneralinhisdiscussionofperceptualjudg-ments.10.ThusPeircestatesthatbothconceptionsandsensationsinvolvehypo-theticalinferences.11.Gruender(1983:281–7),inhisfoundationalistinterpretationofPeirce’sposition,viewstheinterrelationofobservationandtheoryinPeirce’sphilosophyintermsoftypesoflanguage,andseemstoplacetheinfil-trationofthetheoreticalintowhatisgivenatamoresophisticatedlevelthanisindicatedhere.Thus,hemaywellobjectbothtotheterminologyandtheconceptwhichitindicates.12.Savan(1965:40–1).13.ThispointisdiscussedinsomehistoricaldetailbyC.I.Lewis(1929:154)inrelationtothedevelopmentofhisownposition.AlthoughKantisconsideredthebeginningof“therejectionofthespectator,”hehimselfwasnotimmunetosomeofitspresuppositions.Thus,inacceptingthelattertwocharacteristics,herejectedthefirst.14.Ofcoursethereisasenseinwhichother“worlds”arereal.Forexample,theidealworldisarealidealworld.15.Peirceassertsthatthegeneralisthattowhichthelawoftheexcludedmiddledoesnotapply,whilethevagueisthattowhichtheprincipleofnoncontradictiondoesnotapply(CP5.448).Hethenexplicitlyidentifiescontinuitywithgenerality(MS137,pp.7–12).And,forPeirce,whateverCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sPragmaticAccountofPerception213isgeneralorcontinuousistosomedegreevague.SeeBuchler’s(1966:25)commentsonthispoint.Thus,neitherthelawofnoncontradictionnorthelawofexcludedmiddleisperfectlyapplicabletothecontinuous.16.Almeder(1985).17.Esposito(1980:229).18.Almeder(1980:52–57).19.Smith(1978:50ff.).20.Almeder(1975b:3–17).21.Almeder(1975b:14).Italicsnotintext.22.Almeder(1975b:9).Italicsnotintext.23.TheIdealismunderdiscussionhereisofcourseepistemologicalratherthanmetaphysicalidealism,thoughthetwoarecertainlynotunrelated.24.ThisispreciselytheconclusionreachedbyAlmeder(1975a).25.Sometypeofcoherencetheoryoftruthoperateswithintheframeworkofontologicalphenomenalismaswell.26.Thayer’s(1968:132)characterizationofPeirce’sconceptoftruthashav-ingthefunctionofKant’sregulativeideas“servingasaworkingstan-dardofcriticism”wouldapplyhere,butatamoreradicallevelthanthatintendedbyhischaracterization.SeealsoMurphey(1961:302).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nt.l.short9TheDevelopmentofPeirce’sTheoryofSignsEarlyandlate,Peirceanalyzedthoughtasaprocessofsign-productionandsign-interpretation.Buthisearlydoctrineof“thought-signs”wasdeeplyflawedand,therefore,hereviseditdras-tically,changingitsbasicprinciplesandgreatlyextendingitsscope.Ashedidthatinseveralstagesovermanyyears–ofteninlettersorunfinishedmanuscripts–whatwenowpossessislittlemorethanasequenceofcontradictions,aseriesofambitiousyetunfinishedsketchesofelaboratebutmutuallyincompatiblestructures.Andyet,Peirce’sworkonsignscannotbeignored;forhispragmatictheoryofinquiryandhissynechisticaccountofthemindareincompletewithoutit.ThisessaytracesthedevelopmentofPeirce’stheoryofsigns,orsemeiotic,culminatinginthemature,orpost-1906,versionofthattheory.Tospeakofthismaturetheoryatallistospeakhypo-thetically:ithastobeconstructedfromthesurvivingmanuscriptsofPeirce’slastyearsplusallthatisconsistentwiththemfromhisearlierwritings.1.theearlytheory:thedoctrineofthought-signs(1866–9)In1866,whenhehadjustturned27,Peircewrotethatare-presentationissomethingthatstandsforsomethingtosomeonewhosointerpretsit–moreprecisely,tothe“interpretant,”1whichthatpersonformsinresponsetothesignandwhichisasecondrepresen-tationofthesamething(W1,466).Already,then,wehavePeirce’sfundamental,unchangingconceptionofasign(or,inearlierdays,“representation”and,sometimes,“representamen”),asbeingoneofthreerelata–sign,object,interpretant–ofasingle,triadicrelation.214CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns215Thatmakesinterpretationessentialtosignhood.Significanceisnotadirectrelationofsigntoobject;instead,thesignificanceofasignisdeterminedbytheinterpretantwhichthatsignelicits.ButwehavealsothethesisthatcharacterizestheearlyperiodofPeirce’ssemei-oticthoughtandthatwasdramaticallyreversedinhislateryears:thatineverycasetheinterpretantofasignisanothersignofthesameobject.Peirce’searlyperiodculminatedinthe1868–9seriesofthreear-ticlesintheJournalofSpeculativePhilosophythatintroducedtheconceptofthe“thought-sign”(W2,193–272).Everythought,Peircethereinmaintained,interpretsaprecedingthoughtandisinterpretedinasubsequentthought(W2,225).Thethought-signsthatagiventhought,T,interpretsdetermineT’sreferent,orobject.Andthethought-signsinwhichTisinterpreteddetermineT’smeaning,orwhatitrepresentsaboutitsobject.Forexample,mystartledatten-tionisdrawntoashadowyshape:(A)“Somethingisinthatalley.”(B)“Maybeamugger!”(C)“I’dbettercrossthestreet.”Thethought,B,isnotaboutmuggersingeneral:itsobjectispresentedinthethought,A,ofwhichBisaninterpretation,viz.,thesomethingthatisinthealley.AndwhatBrepresentsaboutthatobjectisinterpretedinC,whereinamuggeristakentobeaformofdangerbestavoided.(Ob-viously,othersubsequentthoughtscandrawothermeaningsfrom“mugger.”)Ifthissamesortofanalysisappliestoeachthought,theneverythoughtisbothasignandaninterpretant.Hence,eachisbutamo-mentinaninfiniteregressusandinfiniteprogressusofthought-signs.Thatthoughtbeginsandendsintimeisaccountedforbyitsbeingacontinuum,packinganinfinityofinfinitesimalthoughtsintoafiniteflowofthought.2Amongmuchelse,thisentailsthatthereisnocognitionnotdeterminedbyapreviouscognition,hence,thatnoneisdetermineddirectlybyitsobject.Ifnocognitionisdeter-mineddirectlybyitsobject,thenthereisnointuitiveknowledge.Toestablishthattheoremanddeduceitsconsequenceswasthemainburdenofthe1868–9articles.But,apartfromitsexplicitlyanti-Cartesianintention,whatledPeirceintothisstrangedoctrine?WhyopposeDescartesbypositingacontinuumofthoughtsinterpretingthoughts?Peirce’sphilosophicalstudiesbeganwithKantandcenteredaroundtheKantianproblemofknowledge.AsagenerictermforCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n216t.l.shortanymentalcontent,KanthadadoptedWolff’sVorstellungbuthepro-vidednodiscussionordefinitionofVorstellungeningeneral(George,1982:31–4).PeircewasledfromKantbacktotheBritishempiricists;andalmostcertainlyhisterm“semeiotic”isatransliterationoftheGreekwordLockeintroduced,attheendofhis1690Essay,tonameanew“doctrineofsigns.”This,Lockeexplained,istobe“anothersortoflogic...thanwhatwehavebeenhithertoacquaintedwith,”bywhichtostudyideasas“signsthemindmakesuseoffortheunder-standingofthings”andwordsas“signsofideas”(Locke,1690[1965],v.2:309–10).This,apparently,iswherePeirce’ssemeioticanalysisofthoughtbegan.Itshouldbeemphasized,however,thatPeirce’sconceptofthoughtwasKantian,notLockean:thought,forPeirce,isalwaysconceptual,hence,generalincontent.Andhedevelopedthisideainadistinctlycontemporary(thoughalsoPlatonic)way,byidentifyingthoughtasinternalizeddiscourse.Thoughtisaspeciesofsemeioticbehavior,generallybutnotexclusivelyverbal,onaparwithspeechandwrit-ing;ourcapacitytothinkisdependentonourhavinglearnedalan-guage.Toconceiveofthoughtasasignthuspresupposesthatwordsandsentencesaresigns.Butwhatmakessomethingasign?Aristotle’splausibleviewwasthatwordssignifythoughtsbyconventionorcustomandtherebysignifythethingsofwhichthoughtsare“likenesses”(DeInt.1).Butifthoughtsthemselvesareinwords,thentheyarenotlikenesses.Ifthoughtsarewords,thenAristotle’saccount,thusmodified,wouldimplythatideassignifythingsbysignifyingideasofthings–andsoon,adinfinitum.Strangelyenough,Peirce’searlytheoryofsignsismuchlikethat.Thoughtsaresigns;signssignifythroughthought;therefore,thought-signssignifythroughotherthought-signs.WhatPeirceaddedtothiscombinationofKantandLockewithAristotlewastheideaofthecontinuum.Bymakingthetranslationofthoughtbythoughtproceedadinfinitum,butinacontinuousflow,heshiftedtheemphasisfromindividualthought-signstotheprocess–themovementofthought–itself:Itmaybeobjected,thatifnothoughthasanymeaning[i.e.,initself,apartfromitsbeinginterpreted],allthoughtiswithoutmeaning.Butthisisafallacysimilartosaying,that,ifinnooneofthesuccessivespaceswhichaCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns217bodyfillsthereisroomformotion,thereisnoroomformotionthroughoutthewhole.Atnooneinstantinmystateofmindistherecognitionorrepre-sentation,butintherelationofmystatesofmindatdifferentinstantsthereis.(W2,227)Inshort,sincemeaningcannotbelocatedinanythought-sign,itmustbefoundintheveryprocessbywhichonethoughtinterpretsanother.2.threeflawsfataltotheearlytheoryManycommentatorsassumethatPeirceneversubsequentlyaban-donedhis1868–9doctrineofthought-signs.However,thatdoctrinefacedatleastthreeproblems,anyoneofwhichwouldhavebeensufficientreasonforhimeventuallytohaveabandonedit.First,Peirce’sattempttoexplainhowthoughtssignifyfails.Hesupposedthatsignificancedependsoninterpretation,butthenex-plainedinterpretationasconsistinginsigns.Thus,theproblemofaccountingforsignificanceisnotsolvedbutismerelyhandedon,fromonesigntothenext.Nordoesitmatterthattheprocessofinter-pretationcontinuesadinfinitum.Thatmerelypostponesanansweradinfinitum.3Peirceanticipatedthisobjectioninthelongpassagequotedabove.However,thatresponsecontainsnoexplanationofhowthetranslationofonethought-signintoanotherproducessig-nificance.Merelyassertingthatitdoesisnotenough.Theassertionthatmeaningistobefoundinthemovementofthoughtratherthaninindividualthoughtsisunsupportedbyargument.And,asPeirce’sanalogytophysicalmotionisquestionableforthesamereasonthatthedoctrineitisintendedtosaveisquestionable,thatanalogyisofnohelpwhatsoever.Peirce’sfailuretoexplainhowthoughtssignifyisaccompaniedbyanotherfailure:toexplainhowsensecanbedistinguishedfromnon-sense.Thefactthat“Possibilitiespandertoprodigiousplentitude”canbetranslatedintootherlanguagesorintootherEnglishphrases,andthoseintostillothers,andsoon,adinfinitum,doesnotprovethatitsaysanything.Infact,itsaysnothing.Andsincethoughtcanbeformedinsuchwords,itfollowsthatthoughts,likespeech,canbenonsensical.ButonPeirce’stheory,thisnonsense,beingtrans-latable,isnotnonsense.Peirce’stheoryentailsthatcertainphrasesCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n218t.l.shortthatsignifynothingsignifysomething;and,therefore,histheoryismistaken.Second,Peirceassumedthateverythoughtmustactuallybeinter-preted,thatitsinterpretantmustbeactualandnotmerelypotential;forwhyelseholdthateachthoughtoccursinanactualbutinfiniteprogressionofthought-signs?Butsincetheinterpretantdeterminesasign’smeaning,itfollowsthatsignificanceisnothingmoreorlessthanthewayinwhichasignisactuallyinterpreted.Hence,erro-neousinterpretationisimpossible:significanceisentirelysubjec-tive.ThereisanalternativethatPeircedidnotclearlyarticulateatthattime,whichisthatsignificanceisapotentialityforaspecificsortofinterpretation–apotentialitygroundedonsomethingthatwouldjustifyinterpretantsofthattype.Ifsignificanceisgroundedinter-pretability,thenitispossibleforsomethingtobemisinterpreted–namely,whenaninterpretantisnotgroundedinthesign’sground.WhydidPeirce,inthisearlyperiod,identifysignificancewithbe-inginterpreted,ratherthanwithbeinginterpretable?Onemightaswellaskwhy,tenyearslater,hedeniedthatthereisanyrealdiffer-encebetweenahardandasoftthingbeforetheyareactuallyexposedtopressure.(Thelatteroccursfamouslyin“HowtoMakeOurIdeasClear,”atW3,266–7.)Inbothcases,itisevidentthatPeircethenlackedadefiniteconceptofpotentiality.Hisoccasionaluseofex-pressions,suchassomeinvolvingtheword“would,”thatseemtorepresentpotentialities,andevenhisoccasionalexplicitassertionsofpotentialities,isnoproofthathecouldaccommodatetheminhisphilosophy.ThusPeircecouldnotadmitthatadispositionalprop-ertylikehardnessconsistsinwhatwouldhappen,asdistinctfromwhatdoeshappen;andthesameappliestoasign’ssignificance.Third,Peirce’sinfiniteregressusofthought-signsisasunsatisfac-toryasishisinfiniteprogressus.Eachcognitionisdeterminedbyaprecedingcognitionthatestablishesitsobject.Butthatobjectisnottheprecedingcognitionitself;rather,itistheobjectofthepreced-ingcognition.Thustheobjectisnevertobefoundintheseriesofcognitionsatall,butstandsoutsidetheseries,approachedbutneverreachedaswetracethought-signsbacktoprecedingthought-signs.Forreasonswehavenoroomheretoexplore,Peirceinthe1860sviewedthisexternalobjectas,quaexternal,unknowable.But,onaprincipleannouncedinthesesameessays,thattheincognizableCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns219isinconceivable(W2,208),thismeansthattheobjectcannotbeadmittedasrealorevenaspossiblyreal.AsPeirceputit:Atanymomentweareinpossessionofcertaininformation,thatis,ofcog-nitionswhichhavebeenlogicallyderivedbyinductionandhypothesisfrompreviouscognitionswhicharelessgeneral...andsobacktoanidealfirst,whichisquitesingularandquiteoutofconsciousness.Thisidealfirstistheparticularthing-in-itself.Itdoesnotexistassuch.(W2,238,Peirce’semphasis)Peircewentontoarguethattheobjectofthoughtisreal,insofarasitisrepresentedincognition,andheconcludedthat“Thereisnothing,then,topreventourknowingoutwardthingsastheyreallyare”(W2,239).However,astheoutwardnessoftheoutwardthingisbyPeirce’sargumentunrealand,indeed,inconceivable,itisnotclearhowanyrealityoranyobjectofknowledgecouldbeidentifiedasan“outwardthing.”Noticethatexistenceexternaltothemindisnotallthatisdenied:individualexistenceisaswell.ThisfollowsfromtheprecedingplusPeirce’sKantianconceptionofcognitionasinvariablygeneral.For,bythatdoctrine,theindividualmustlieoutsidetheentireseriesofcognitions,and,thus,iftheexternalisunreal,soistheindivid-ual.Yet,fromhisearliestphilosophicalwritingsthroughhis1877–8paperson“TheFixationofBelief”and“HowtoMakeOurIdeasClear,”Peircenevergaveupthehypothesisthatthereisanexter-nal,individualcauseofsensation.Hemerelyattempted,invariousways,tocombinethatideawiththeideathattherealisasinquirytendseventuallytorepresentittobe.Thatnoneofthoseattemptswassuccessfulisattestedtobytheirnumber,andbythefactthat,beginningin1885,Peircetookaradicallydifferentapproach.3.thethirdflawcorrected(1885–1903)By1883,Peirceandhisstudent,O.H.Mitchell,haddiscoveredquantification(independentlyofFrege).Infirstorderpredicatelogic,quantifiersbindindividualvariables,andthelatterfunctionasrela-tivepronouns;together,quantifierandvariablerefertosomeortoalloftheindividualsinagivensetofindividuals.Thoughtheseformaldevelopmentsareneithernecessarynorsufficientforthepurpose,theyappearinfacttohaveledPeirce,by1885,torecognizeatypeofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n220t.l.shortsign–the“index”–thatisnotgeneralyetplaysanessentialpartincognition:...generalityisessentialtoreasoning....But[generalterms]alonedonotstatewhatisthesubjectofdiscourse;andthiscan,infact,notbedescribedingeneralterms;itcanonlybeindicated.Theactualworldcannotbedis-tinguishedfromaworldofimaginationbyanydescription.Hencetheneedofpronounsandindices....TheintroductionofindicesintothealgebraoflogicisthegreatestmeritofMr.Mitchell’ssystem.(W5,163–4)Peirceappliedthisargumentnotonlytodiscourseaboutparticularindividualsbuttogeneralstatementsaswell,andtostatementsinmathematicsandlogic(CP2.310–12,336–7,357,4.56–9,8.368n23).AsMurrayG.Murpheyremarks,Peirceusedtheterm“index”inthelate1860s,butnottorefertosignsdevoidofgeneralmeaning:Theindex“It”ofthe“NewList”isaconcept–namely,theconceptof“thepresent,ingeneral”...Theuseoftheterm“index”tomeanasignwhichrefersnottoaconceptbuttoanindividualdirectlydoesnotappearuntil1885...ItisatthispointthatthenotionofindividualitybecomesimportantforPeirce.(Murphey1961:299–300)Thus,by1885,Peircecanaccepttheindividualthing4asrealwith-outqualification,notmerelyideal,because,bythattime,hesawthatitsconceptualapprehension–descriptive,general–iscomple-mentedbyamoredirectformofexperience.“Theindex,”Peircewrote,“assertsnothing;itonlysays‘There!’”It“forciblydirects”oureyes,orotherwiseourattention(W5,163).Apointingfingeristheprimeexample.NoticethatPeircedidnotsaythatindicesoperateoutsideofcon-textscreatedbyothersigns;muchlessdidheassertthatindicesbythemselvesconveyknowledge.Heonlyassertedthatindicesarere-latedtotheirobjectsdirectly,unmediatedbygeneralconceptions,andthatitisbybeingconnectedwithindicesthatgeneralconcep-tionsareappliedtoindividualobjects.Peirce’sconceptionoftheindexrequirescarefulexplication.Theindex“signifiesitsobjectsolelybyvirtueofbeingreallycon-nectedwithit.Ofthisnatureareallnaturalsignsandphysicalsymp-toms”(W5,163).Therelationofindextoobjectissometimescausalorphysical;itwilloftenbecompulsive(asasharppokeinthebackCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns221callsone’sattentiontoitsasyetunidentifiedcause);butinanycaseitconsistsinaparticularorexistentialrelationbetweenindividuals.Butindicesarerarelypure.Instinct,custom,orconventiondrawsourattentiontotheextendedarmandrigidfingerandtellsustolookalongthelineitdefines;withoutthatinstinct,custom,orconven-tion,indexfingerswouldlessoftenberigid,forthentherecouldbenointention“topoint”(andtherigiditywouldnotbecalled“point-ing”).Nonetheless,thelinethatthefingerdefinesdependsonnoth-ingbutthefactthatitisextendedatthistimeinthisplaceinthisway.Instinctorconventioncallsattentiontothelinebutdoesnotmakethatline.Furthermore,indicesareinterpretedinlightofothersignswithwhichtheyoccur(“SeethatmanI’mpointingat?”)orinlightofbackgroundknowledgeaboutindicesoftheirtype.Higherthannor-malbodilytemperatureiscalled“fever”andistakentobeasymp-tomofinfectiononlybecauseweknowthatgenerallysuchacondi-tioniscausedbyinfection.Nonetheless,theindividualinstanceoffeverpicksoutanindividualinstanceofinfectionbyvirtueofbeingcausallyconnectedtoit,andthatcausalrelationisindependentofanyideaswemayhaveaboutit.Thuswecandistinguishtheindex-icalelementofsignificationwithintheseexamples,eventhoughitiseffectiveonlyincombinationwithother,moregeneral,elements.Peirce’sdiscoveryoftheindexicalsignenabledhimtorelinquishthethesisthateverycognitionmustbeprecededbyacognition,adinfinitum.Acognitioncombinesindicesandconcepts.Theindexpicksoutaparticularofanotherwisesignifiedtype,whichisthenmadethesubjectofapredicate(quantificationextendsthisanalysistogeneralstatements).Itfollowsthatiftheindexisdirectlycon-nectedtoitsobject,thensoisthecognition,throughtheindexitcontains.Thus,acognitiondoesnothavetobetheinterpretantofaprecedingcognitioninordertohaveanobject.Inhis1903Harvard“LecturesonPragmatism,”Peircedescribedthe“perceptualjudgment”as“thefirstjudgmentofapersonastowhatisbeforehissenses”(CP5.115,myemphasis).Thisisascleararejectionofhisearlierview–thatthereisnofirstcognitionbutthateverycognitioninterpretsaprecedingcognition–asonecouldwish.Butitdoesnotfollowthatsuchjudgmentsareintuitionsinthesenseofbeingentirelydeterminedbytheirobjectsand,thus,infallible;fortheunionofconceptwithindexisfalliblebecauseconjectural.OneCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n222t.l.shortcanmistakenlytakesomethingtobeanindexwhenitisnot,andonecanmistakenlytakeanindextobeofanobjectofatypeotherthanthetypeofthatobjecttowhichitisdyadicallyconnected.Inthesesamelecturesof1903(andelsewhere)Peircelikenedperceptualjudgmentstohypotheses,fromwhichtheydifferonlyinnotbeinginferential(CP5.181).Peirce’sfallibilism,hisanti-Cartesiandenialofintuitiveknowledge,waspreservedevenwhilethedoctrineofaninfiniteregressofjudgments,onwhichitwasoriginallybased,wasjettisoned.4.consequentgeneralizationanddeepeningofsigntheoryThediscoveryoftheindexicalsignhadtwofurtherimplicationsforthedevelopmentofPeirce’ssemeiotic.First,semeioticwasextendedbeyondthestudyofthoughtandlanguage.Foralthoughtheindexwasdiscoveredasplayinganessentialrolewithincognition,itisbyˆitsnature–asbeingcausalorotherwisenonconceptual–notlimitedtocognition.Naturalsigns,assmokeisoffire,thunderoflightning,feverofdisease,mustalsobeadmittedtosemeiotic’spurview.Furthermore,indicesareofcourseinterpretablebythought,butmoreessentiallybythatcomponentofthoughtwhichconsistsinanactofattention.Theroleofanindexistosetusinacertainˆdirectionwhenapplyinganassociatedidea.Aninterpretant,then,neednotalwaysbeanothersignofthesameobject.Butthiswasnotmadeexplicituntil1904,whenPeirceallowedinterpretantstobeactionsorfeelingsaswellasthought-signs(CP8.332).ThattripartiteclassificationwasbasedonPeirce’slistofthreecategoriesasrevisedinthe1880sand1890sonthebasisofthelogicofrelations,onwhichhehadbegunworkinthe1870s.Feelingsaremonadic,actionsaredyadic,andsigns,aswehaveseen,aretriadic.In1907,thesetypesofinterpretantwerenamedtheemotional,theenergetic,andthelogical(CP5.475–6),thoughatthattimealogicalinterpretantwasnolongerregardedasbeinginvariablyathought-sign(seeSection6).Butifsemeiotic’spurviewisextendedtononconceptualinter-pretants,thenwhynottononhumaninterpreters?Apersonpokedturnstolook,butsoalsoabrowsingdeer,startledbyanoise,raisesitsheadtolook;theseasoneddriver,seeingastopsign,stopswithoutthinking,butsoalsoabloodhound,nosetoground,followswithoutCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns223thinkingthespoorofitsquarry.Semeiotictherebybecameastudynotonlyofnaturalsignsbutalsoofnaturalprocessesofinterpre-tation.Andthatsuggestsawayinwhichthehumanmindmaybelocatedwithinnature,namely,asadevelopmentofmoreprimitivesemeioticcapacities.Thesecondimplicationofthediscoveryofindicesisthatitcom-pelsustorecognizearelationofsigntoobjectthatisdistinctfromsignification.Aswenotedearlier,Peirceconceivedofsignificanceasatriadicrelationinwhichaninterpretantmediatesbetweenthesignandtheobjectsignified.Butindicesarealsorelatedtotheirob-jectsdyadically.Itfollowsthatthetworelationsarenotthesame.Peircedrewthatconclusionin1885(W5,162–3)butdidnotatthattimesayhowthetworelationsarerelated.Itisobviousthatthesignrelationmustinsomewaybebasedonapriorrelationofthatwhichisasigntothatwhichisthatsign’sobject;butthenatureofthatdependencycouldnothavebeenformulatedbeforethedevelopmentofthematuresemeiotic,in1907.Still,Peirce’smostfamousdivisionofsigns,intoicons,indices,andsymbols(indifferentperiodsvariouslydesignated),turnsonthedifferentkindsofpriorrelationasignhastoitsobject.Evenin1866–7,whenthisdivisionwasdifferentlyconceived(W1,475,2:56),theexamplesPeircecitedanticipatedhislaterview,inwritingsof1885,1893,and1902–3(CP2.274–308)andc.1903(CP2.247–9).Hereagain,thebasisofclassificationisthelogicofrelations.Briefly,thepriorrelationofanicontoitsobjectismonadic,thatofanindexisdyadic,andthatofasymbolistriadic.Aniconisrelatedtoitsobjectmonadically,eitherbyresemblance,whereinsignandobjectbothhavethesameproperty,orbyexempli-fication,whereinitsobjectisapropertyitpossesses.Likeindices,iconsnormallyfunctioninacontextofothersignswhichdirectat-tentiontooneoranotherofathing’siconicaspects.Theindexwehavediscussed.Thesymbolaloneisbasedonatriadicandintrinsi-callysemeioticrelation,viz.,arulebywhichitistobeinterpreted.Notice,however,thatthisisdistinctfromthesign’sbeinginter-pretedinconformitytothatrule;thus,thedistinctionofthesign-relationfromthepriorrelationonwhichitisbasedismaintained.Mostcommentatorshaveassumedthataruleofinterpretationmustbeconventional,eventhoughPeirce,asearlyas1885,saidonlythatsymbols(thennamed“tokens”)areconventional“fortheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n224t.l.shortmostpart”(W5,162).Later,hespecificallysaidthattheinterpretinghabitmaybeeither“naturalorconventional”(CP2.307,1902)orthattheruleofinterpretationmaybeeitheraconvention,ahabit,oranaturaldisposition(CP8.335,1904).Thus,apeacock’smatingdisplaydoesnotobeyaconvention,andyetitmeanswhatitdoesonlybecauseofacorrespondinginstinct,onthepartofpeahens,sotointerpretit.Twoothertripartiteclassificationsofsigns,ca.1902–3,mayherebebrieflymentioned.Asigninitselfiseithermonadic–aqualityoffeeling,hence,aqualisign–dyadic–asingularobjectorevent,hence,asinsign–ortriadic–atypedefinedbyalawforformingitsreplicas,hence,alegisign(CP2.243–6).Aqualisigninitselfisamerepossibilityand,so,mustbeembodiedinasinsignactuallytofunctionasasign.Hence,somesinsignsaresuchbecauseofthequalisignstheyembody:theysignifynot,ornotonly,invirtueoftheparticularitiesoftheiroccurrence,asindices,butinvirtueoftheirqualities,asicons.Soalso,alegisignsignifiesonlythroughitsreplicas,whicharesinsignsofaspecialkind.Legisignsandtheirreplicasarebetterknownastypesandtokens,terminologywhichPeircealsointroduced(CP4.537,544).Legisignsaloneamongsignsexisttobeused,i.e.,replicated;andnothingreplicatesalegisignthatisnotformedforthatpurpose,i.e.,tosignifyaccordingtoaruleofinterpretation.Peircesometimesmaintainedthatsymbolscanonlybelegisignsandthattheirreplicasareindicesthatdirectattentiontothesymbolreplicated(CP2.249).Butnotalllegisignsaresymbolic.Pronounsareindexicallegisigns,sincetherulesforinterpretingtheirreplicasdirectonetothecircum-stancesinwhichtheyoccur(CP2.259).Andgeometricaldiagramsareiconiclegisigns,sincetherulefortheirinterpretationdirectsonetoattendtothespatialrelationstheyexemplify(CP2.258).Bycontrast,theobjectofasymbol’sreplicasisgivenbytheruleofinterpretationitself.Afinaltrichotomyofsignsformulatedin1903wasatfirstbasedonthecategoryofthesign’sobject,asapossibility,fact,orreason(CP2.243),butwasreformulatedin1908as“theNatureoftheInfluenceoftheSign,”asbeingeithermonadic,dyadic,ortriadic(CP8.373).In1903,thesesignsare,respectively,rhemes,dicentsignsordicisigns,andarguments,butin1906(CP4.538)and1908(CP8.373),theyaresemes,phemes,anddelomes.Inanycase,theyaregeneralizationsCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns225ofthefamiliarterm/proposition/argumenttrichotomy–ageneral-izationrequiredbythefactthatsignsarenolongertobelimitedtohumancognition.Theideaofthe1908reformulationisthis:asememerelypresentsitsobject;apheme,suchasanassertion,isforceful;whileadelome,suchasanargument,appealstotheinterpreter’sownreason.Inthecaseoflanguage,syntaxisaguidetothenatureofasign’sinfluence,butaguideonly;forincertaincontexts,amereword–“Yes!”–canconveyanassertionoranargument.55.thesecondflawcorrected(1896–1909)Sinceinterpretationisbasedonadistinctrelationofsigntoobject–onethatobtainswhetherornotaninterpretantisactuallyformed–itbecomespossibletodistinguishinterpretabilityfrominterpreta-tion.Fromthesolutiontothethirdofthethreeproblemswehavesurveyed,wecanthusderiveasolutiontothesecond.ButitisasolutionthatPeircecouldnothaveembracedbeforehedevelopedarealistconceptionofpotentiality(or,ashesometimessaid,realpossibility).Thishedidinhislateryears,whenheconstruedtherealityofalawinthestrongestpossibleterms,asa“would-be”thatisirreducibletoanyquantityofactualinstances(e.g.,atCP5.467,6.327,8.216–7,225),beginningin1896(CP1.420).6Butifalawinthissenseisreal,thensoaredispositionalproperties(whicharelawsabouthowoneoranotherindividualthingwouldbehaveundercer-tainconditions)andsoarepotentialities.In1905,Peirceexplicitlycorrectedhis1878remarkonhardnessandannouncedthat“itistherealityofsomepossibilitiesthatpragmaticismismostconcernedtoinsistupon”(CP5.453).ThusPeircewasable,after1896,toidentifysignificancenotwithactualbutwithpotentialinterpretation,orinterpretability,ashedidemphaticallyandrepeatedlybeginningin1902(e.g.,atCP1.542,2.242,274–5).Tobesure,hecontinuedtoinsistthateverysignmusthaveaninterpretant,butthatinterpretantisapotentiality.Thispointisembodiedinhislatedistinctionbetweenasign’s“imme-diate”interpretantanditsactual,or“dynamic”interpretants(seeSection9).Itis,then,itsbeinginterpretable,andnotitsbeinginterpreted,thatmakessomethingasign.Itfollowsthataninfiniteprogres-susofactualthought-signs,eachinterpretingthepreceding,isCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n226t.l.shortunnecessary.7Justasaparticularactofthinkingcanbeginwithabang,soalsoitcanendthatway:init,therecanbealastaswellasafirstthought.Thatisnottodenythateverythoughtremainsinter-pretablebyfurtherthought;inprinciple,thereisnoendtothinking.But,wheninfactitdoesend,thinking’srealityisnottherebycan-celled.6.therevolutionof1907:ultimatelogicalinterpretantsThethirdandsecondofourthreeproblemsweresolvedinconse-quenceofdevelopmentsof1885and1896,respectively,butthefirstnotuntil1907,whenPeircesuggested,thoughhedidnotfullyformu-late,anewaccountofhowsignssignify–onethatwasentirelydif-ferentfromhisearlieridentificationofsignificancewithanendlessprocessofsignsinterpretingsigns.Thefirststepinthisnewaccountwastorelinquishtheviewthateveryinterpretantofathought-signmustbeanotherthought-sign.Iftheinterpretantofasigncanbesomethingotherthanasign,andifthisholdsforsignsofeveryclass,thensignificanceneverdependsonaninfiniteseriesofinterpretants,evenifonlyinposse.In1902,Peircestillclaimedthatasignmustbesorelatedtoitsobject“astobecapableofdeterminingaThird,calleditsInterpre-tant,toassumethesametriadicrelation”tothatsameobject(CP2.274).Noticetheword“capable”:asignneednotactuallybein-terpreted.Yetaninterpretant,whenformed,musthavethe“sametriadicrelation”totheobjectashasthesignitinterprets:ergo,theinterpretantmustbeanothersignofthatobject.In1904,thethesisthatinterpretantsaresignswasreiterated(CP8.191,226n10),butalsointhatyearitsfirstmodificationmaybefound,where,inalettertoLadyWelby,Peircesuggestedthat“Takingasigninitsbroadestsense,itsinterpretantisnotnecessarilyasign,”sinceitmightbeanactionorfeeling(CP8.332).8This,aswehaveseen,becametheemo-tional/energetic/logicaltrichotomyofinterpretants.Butin1904–6,Peircecontinuedtoaffirmthe1868–9viewthatallthoughtsmustbeinterpretedinfurtherthought-signs.Thereasonisthatthemeaningofathoughtcannotbeexhaustedbyanynumberoffeelingsand/oractions,andthusathought’sinterpretantmustfallintothethirdcategoryofinterpretant.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns227In1907,Peircereversedhimselfbydrawingadistinctionwithinthecategoryoflogicalinterpretantsbetweenthosethataresignsandthosethatarenot.Thelatterhenamed“ultimatelogicalinterpre-tants”:...farfromholdingthatasigncanbethe“naked”,thatis,theultimatemeaningofasign,Iwasjustabouttoinsistthatitcannotbeso;andthegreatenigmathatleadsuptopragmatism,–atleasttomyformofthedoctrine,–is,“Whatcanthisnakedorultimatemeaningbe?”Idonotdenythataconcept,orgeneralmentalsign,maybealogicalinterpre-tant;only,itcannotbetheultimatelogicalinterpretant,preciselybecause,beingasign,ithasitselfalogicalinterpretant.9Thus,themeaningofathoughtmustbesomethingmore,orsome-thingother,thanitstranslationintofurtherthoughts.Whileathoughtmayalwaysbeinterpretedinfurtherthoughts,theremustalsobeanotherformofinterpretation,atleastpotentially.Theultimatelogicalinterpretant,asthereferencetopragmatismindicates,isthehabit:TosaythatIholdthattheimport,oradequateultimateinterpretation,ofaconceptiscontained,notinanydeedordeedsthatwilleverbedone,butinahabitofconduct,orgeneralmoraldeterminationofwhateverproceduretheremaycometobe,isnomorethantosaythatIamapragmaticist.(CP5.504,Peirce’semphasis)10Thehabitofconductinwhichthemeaningofaconceptistobefoundisthatwhichonewouldforminadoptingtheconceptorinapplyingittospecificsubjects.IfIbelievethatthestoveishot,thenIamdisposedtoacttowardthestoveaccordingly,avoidingitforsomepurposes,approachingitforothers.Andtopossesstheconceptofheatistobereadytoformsuchdispositions;itistoexpectthat,upontheobservationofcertaineffects,certainotherswilloccurorcanbemadetooccurifcertainconditionsaremet.Notoriously,itisnotpossibletospecifyallofthedispositionsthatonewhoadoptsaconceptmightform;butwhetherthatisaproblemforPeirce’sviewor,alternatively,correctlyindicatestheopennessofconceptstogrowthinlightofexperience,isanothermatter.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n228t.l.short7.pragmatismandsemeioticSincePeirce’spragmatictheoryofmeaningwaspublished,albeitnotunderthatrubric,inhis1878paper,“HowtoMakeOurIdeasClear,”itmaybewonderedwhyittook29moreyearsforthatdoctrinetolinkhandswithhissemeiotic.CouldPeircehavefailedforsolongtodiscernthecontradictionbetweenhis1878accountofmeaning,notforswornbutreaffirmedinlateryears,andhis1868–9doctrineofthought-signs?ItistruethatPeircedidnoworkonthetheoryofsignsfromthespringof1873tothewritingofhisaforementionedpaperof1885;butthatdoesnotexplainhisreiteratedassertions,from1902to1906,thatinterpretants,atleastthoseofthoughts,mustbesigns.Theanswerliesinthewayinwhichthepragmaticmaximof1878wasformulated:“Considerwhateffects,whichmightconceivablyhavepracticalbearings,weconceivetheobjectofourconceptiontohave.Thenourconceptionoftheseeffectsisthewholeofourconceptionoftheobject”(W3,266).AsPeircehimselfpointedoutin1906,theemphasisthroughoutisonconception,inorder,hesaid,“toavoidalldangerofbeingunderstoodasattemptingtoexplainaconceptbypercepts,images,schemata,oranythingbutconcepts”(CP5.402n3).Thus,whiletheroleofhabitininterpretingwordsorˆconceptsisimplicit,theinterpretantistakentobetheconceptofahabit.Thefundamentalrevolutionindoctrinethatoccurredin1907wastohaverecognizedthatitisthehabititself,andnotaconceptofit,thatistheinterpretant(moreprecisely,theultimatelogicalin-terpretant)ofaconcept.Verbalinterpretantsandverbaldefinitions,Peircethensaid,are“veryinferiortothelivingdefinitionthatgrowsupinthehabit.”11Again,Therealandlivinglogicalconclusionisthathabit;theverbalformulationmerelyexpressesit....Theconceptwhichisalogicalinterpretantisonlyimperfectlyso.Itpartakessomewhatofthenatureofaverbaldefinition,andisveryinferiortothelivingdefinitionthatgrowsupinthehabit.(CP5.491)Therevolutionof1907is,then,arevolutioninPeirce’spragmatismaswellasinhissemeiotic.Inboth,itisastepawayfromatooCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns229extremeintellectualism.1907isalsotheyearwhenPeircefirstdrewhispragmatismandhissemeiotictogetherintooneformulation.Unlikeactionsandfeelings,ahabithasthatcharacterofbeinggeneral,ofbeinginexhaustibleinconcreteinstances,thataconceptitselfhas.Itis,asPeircesaid,a“conditionalgeneralresolutiontoact”(CP5.402n3).Therefore,pragmatism–whichmaynowbeseen,reformulated,asthedoctrineoftheultimatelogicalinterpretant–isnotreductive:itdoesnotreducethegeneraltosomecongeriesofparticulars.Bythesametoken,pragmatismisnotanarrowlyprac-ticaldoctrine.Itdoesnotimplythatwethinkonlyinordertoact.Tothecontrary,inexperimentalscience,weactinordertotestourtheories,thus,toarriveattruethoughts.Andyet,bythisrevolutionof1907,webreakoutofthecir-cleofwords,ofwordsinterpretingwordsandthoughtsinterpret-ingthoughts.Thepragmaticdistinctionbetweenmeaningfulnessandmeaninglessnessbecomesthis:meaningfulspeechandthoughthaveultimatelogicalinterpretants,whilenonsensicalspeechandthought,thoughtheymayalwaysbetranslatedintofurtherthoughtsandwords,lackultimatelogicalinterpretants.Beinginterpretablebyhabitsofaction,meaningfulspeechengageswiththenonverbalworld:forexample,assertionsmaybeacteduponandtestedagainsttheconsequencesofthoseactions.Itshouldbenotedthatpragmatismpertains,Peircesaid,onlyto“intellectual”meaning(CP5.467,482).Peirceidentifiedmeaningingeneralwithinterpretantsingeneral(CP4.536,5.475).Therefore,emotionalandenergeticinterpretantscompriseotherdimensionsofmeaning,e.g.,emotiveandimperative,respectively.Tofeelangerwhenwronglychastised,shamewhenrightlychastised,correctlyinterpretschastisement;onedevoidofsuchfeelingsmaybesaidnottounderstandwhatisbeingsaidtohim(whetherduetolinguisticincompetenceortobeingasociopath).Theactionofonewhoobeysalegitimatecommandcorrectlyinterpretsthatcommand,eveniftheactionfollowsthecommandautomatically,withoutamediatingthought.Pragmatismthusbecamebutonepartofabroadertheoryoflinguisticmeaning.Andthat,inturn,waspartofabroadertheoryofsigns:e.g.,naturalsigns(aprey’sspoor,forexample)areinterpretedenergeticallybytheloweranimals,andthemeaningofaworkofartmaybegraspedpartlyorwhollyinanemotionalinterpretant.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n230t.l.short8.therevolutionof1907:interpretationasend-directedAcriterionfordistinguishingsignificantspeechfrominsignificantusesofwordsdoesnot,byitself,telluswhatsignifyingis.Merelytoissueinafeeling,anaction,orahabitofactiondoesnotmakesome-thingasign;suchresultshavemanycauses,fewofwhichwewouldidentifyassigns.Asecond,andfinal,stepPeircetookin1907wastodescribetheprocessofsign-interpretation,or,ashecalledit,semeio-sis,asbeingend-directed(CP5.472–3,484).12Thatinterpretationcanonlyoccurforthesakeofsomeendwasperhapsimplicitbefore,butnowitismadeexplicit.Ishallarguethattheend-directednessofse-meiosisaccountsforsignificance’sessentialfeatures.AsinPeirce’searlytheory,processiskey;buttheprocessthatmattersisoneinwhichevenasingleinterpretantisformed.Ifsemeiosisisend-directed,andifwearetousetheideaofse-meiosistoexplainconsciousthought,thenend-directednesscannotasageneralrulepresupposeconsciousness.Forthisreason,Peirce’smaturesemeioticrestsonthetheoryoffinalcausationthathefor-mulatedprimarilyintheyear1902(CP1.203–31,250,267–9,2.149,8.272),onthebasisofanalysesofphysicallawandtheoriesofevolu-tionmadefrom1891to1898(CP6.12–3,33,71–81,296–305,7.468–83,518–23)orearlier(seesuchhintsasthoseatW6,63andW3,244).Forthattheorymakesfinalcausation,ordirectednesstoanend,tobeindependentofconsciousdirection.ItisteleologicalinanAristoteliansense,asPeirce’sreferencestoAristotleindicate(CP1.211).Butitisalsobasedonandframedintermsofdevelopmentsinnineteenth-centurynaturalscience,specifically,statisticalme-chanicsandDarwin’stheory.ThisratherheterodoxinterpretationofDarwinismappearstohavebeensharedbyDarwinhimself(seeLennox1993andShort2002),whoalsoadoptedthelanguageoffinalcauses(introducedintheLatintranslationofAristotle).Thereisnoroomhereforafullexegesis,muchlessdefense,ofPeirce’sideaoffinalcausation(seeShort1981b,1983,1999:111–38).Butthegistofitisarguablythis:whenmechanicalevents,fortu-itouslyvaried,resultinatendencythatispracticallyirreversible,thattendencyistobeexplainedstatisticallyasduetovariationsofonetypebeingfavoredoverthoseofothertypes.InthecaseofDar-winiannaturalselection,differentialretentionofgeneticvariantsCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns231overmanygenerationsisexplainedbytheirdifferenttypesofeffect–someincreasingchancesofsuccessfulprocreationmorethanothers.Thiskindofexplanationisnotmechanistic,eveniftheparticularstepsoftheprocessareperfectlymechanical,becausethetendencyoftheprocessisexplainednotbymechanicalforces,whichareal-waysparticular,butbyselectionforgeneraltypesofoutcome(asIargueindetailinShort2002).13TypesofoutcomeasexplanatoryarewhatAristotle,translated,called“finalcauses.”Peirce’smaturesemeiotic,asearliernoted,mustbeconstructedfromfragments.Intheremainderofthissection,IfillinsomeoftheblanksPeirceleft.First,letusnotethatwhenanorganicfeature,X,isexplainedashavingbeenselectivelyretainedbecauseofatypeofeffect,E,thatitnormallyhas,thenXmaybesaidtoexistforthesakeofE(followingWimsatt1972,Wright1976).Inaword,EisX’spurpose.Apurposeisatypeforwhichtherehasbeenorcontin-uestobeselection;itmakesnodifferencewhethertheselectionisconscious.14Iftheheart’sseveralfeaturesaretheresultofvariationshavingbeenretainedbecausetheyfacilitatethepumpingofblood,thentheheartexistsbecauseitpumpsblood;ergo,pumpingbloodisitspurpose.Second,weshoulddistinguishexistingforapurposefromactingforapurpose:theheartfulfillsitspurposemechanically,notpurpose-fully.Whenactionispurposeful,variationandselectionareinternaltotheactionitself,atleastpotentially:purposefulactionissubjecttovariationwhenitfails,thesuccessfulvariants,ifany,beingthenselected.Thetypeselectedforistheaction’spurpose,withreferencetowhichwedefinesuccessandfailure.Selectionmayconsistinrepetitionofthesuccessfulvariant,orinmakingthatvarianthabit-ual(“learningfromexperience”),orincessationofactivity.Thus,ananimalthatmovesaboutrandomlyuntilitlocatesfoodisactingpur-posefully;butsoalso,rigidlyrepetitivebehavior,thoughitappearsmechanical,ispurposefulifitwouldbevariedwereitunsuccessful.Third,nothingpurposefuliswhollyarbitrary.Itmusthavesomewarrantorbasis,howeverfallible,orevenmistaken,thatrelatesittoitspurposeor,ifmistaken,thatseemstorelateittoitspurpose.Letustakeaprimitiveexample.Ablackbearscratchesaboutamongrottinglogsinsearch,wesay,ofgrubs;findinggrubsisitsaction’spurpose.Itsdispositionsotoactisbasedonapastcorrelation(whetherinthebear’sexperienceorinthecourseofursineevolution)betweenCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n232t.l.shorttheodorofrottingwoodandthepresenceofsucculent,protein-richgrubs.Anditsscratchingaboutinthisparticularlogisbased,inaddi-tion,onthepresenceofaparticularodorofthattype.Thecorrelationonthebasisofwhichananimalactsmaybeweak;iffoodishardtocomeby,evena1%correlationwouldjustifythisexpenditureofthebear’senergy.Hencefallibility:thebear’sbehaviormayfaileventhoughitisjustified.Mistakenjustificationissomethingelse.Amistakeoccurs,forexample,ifthepastcorrelationofodortogrubsnolongerobtains(perhapsallspeciesofinsectwhoselarvalformsinhabitrottingwoodhavebecomeextinct),orifthebear’solfactoryorgansaredefective,sothatitrespondstooneodorasifitwerean-other.Thistalk–ofsuccessandfailure,ofjustification,ofrealityversusappearance–isneitherfancifulnoranthropomorphic.End-directednessgroundsouruseofevaluativelanguageindescribingorganicfeaturesandanimalbehavior.Thesethreepointsyieldthis,perhapssurprising,corollary,thatanimalbehaviorcannotbefullydescribedandexplainedwithoutourusinggeneraltermsthatmay,intheindividualinstance,refertonothing.Thebear,wesay,issearchingforgrubsinthislogwhetherornotanygrubsareinfactthere.Normally(puttingasidenegative,modal,imperative,andsomeothercontexts),areferencetowhatdoesnotexistresultsinfalsestatement.“Sittingonalog”cannottrulydescribethebearifthelogdoesnotexist.But“Lookingforgrubs”cantrulydescribeeventhoseunhappybearsthatarefatedtoremainforevergrubless.Letusnamethisa“nonreferentialuse”ofageneralterm:itisausemadeinapositive,nonmodaldescrip-tionofsomething–somethingthatdoesexistandtowhichrefer-enceismade–thatinsomemannerimplicatesanobjectthatmightnotexist.Theprecedingaccountofpurposefulbehaviorshowshownonreferentialusagecanbemeaningful.Thebehaviorhastobede-scribedintermsoftypesthatareinvolvedinselectionandinbasesofselection,regardlessofwhetherthosetypesareinstantiatedintheindividualcase.Butnonreferentialusageisessential,also,toanyattempttosaywhatasignsignifies.Theobjectofasignmaybenototherwiseman-ifest(thefunctionofasignisrootedinthatfact),and,inthecaseofmanykindsofsign,theobjectofasignmightnotexistatall.15Toac-countforsignificanceistoaccountforthispeculiarsenseof“havinganobject,”whereintheobjectmustbespecifiednonreferentially.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns233Andthatiswhatourtheoryofpurposefulbehaviordoes.Indeed,thatpurposefulbehaviorisnotwhollyarbitraryisbutanotherwayofsayingthatitinterpretssigns.Toactpurposefullyistointerpretsomethingasasignofthatwhich,ifitobtains,willmakethatactionappropriatetoitsgoal.Thebear’sactionsinterpretacertainodorasasignofgrubs.Ouroverworkedbearshouldnotmisleadyou:thetheoryisnotrestrictedtotheprimitive.Peirce’spragmatismstated“aninsepara-bleconnectionbetweenrationalcognitionandrationalpurpose”(CP5.412,cf.402n3,428).Hismaturesemeioticrelatessignificanceingeneraltopurposeingeneral.Itis,inthatrespect,ageneralizationofpragmatism.Andneitherdoctrineassumesthathumanpurposesarenarrowlypractical.Forhistoricalreasons,thepeculiarproperty,“havinganobject,”isnamed“intentionality.”WhenFranzBrentanoretrievedthecon-ceptandthewordfromthemedievalScholastics,hemaintainedthatintentionalityisfundamentallyafeatureofmindandthatitiswhatdistinguishesthementalfromthephysical(1814[1973]:88–9).Bythephysical,hemeanttheworldstudiedinthenaturalsciences,andbythemental,hemeantprimarilyhumanconsciousnessasitisknownthroughauniqueformof“inner”perception(4andpassim).Muchlater,RoderickChisholmpointedoutthatsignificanceisaformofintentionalityandthattherearesignsoutsideofanyone’sconsciousness;but,adoptingBrentano’sview,hearguedthatthein-tentionalityofsuchsignsderivesfromthatoftheconsciousthoughtinwhichtheymaybeinterpreted(1952).ThatisinonewayclosetoPeirce’sview,sinceitidentifiessignificancewithinterpretability.YetPeircereversedthedependenceofintentionalityonconscious-ness.Hismaturesemeiotic,becauseitisteleological,accountsfortheintentionalityofsemeiosissansconsciousness;andthusitisabletoexplainthought’sintentionalityasduetothought’sbeingaspecialformofsemeiosis.Inrecentyears,somephilosophersofmindhavedeveloped,in-dependentlyofPeirce,asimilar“teleosemantics”(Millikan1984,Papineau1984,1987).Theirtheoryhasbeensubjectedtoastrongcri-tique(see,e.g.,Fodor1990:chap.3)whichPeirce’ssemeioticevades.Thechargeisthatananimal’sgoalcanbeidentifiedatdifferentlevelsofabstraction,and,thus,thattheobjectoftheallegedsignisambiguous(“thedisjunctionproblem”).Thebear’sgoal,forexample,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n234t.l.shortistofindgrubs,butmoregenerallyitistolocateprotein,andmoregenerallystill,itistosurvive.Istheobjectwhichtheodorsignifiesgrubs,orisitprotein,orisitameanstosurvival?Itwouldseemthattheobjectcanbeunambiguouslyidentifiedonlyifthereisaconsciousthoughtofit.Oursolutiontothisproblemisfoundinthecorrelation,orseemingcorrelation,thatjustifies,orwouldjustify,thebear’sresponse.Thatcorrelation,ifreal,isaninstanceofwhatweearlierreferredtoasthepriorrelationofasigntoitsobject.Thus,theobject,orsupposedobject,isexactlyasspecificasthatrelation,orseemingrelation,makesittobe.Theodoriscorrelated,reallyorapparently,onlywithgrubsofvariouswood-infestingspecies,andnotwithfishorothersourcesofprotein,andthereforethebear’sbehaviorinterpretstheodorasasignofgrubs,nothingmoreandnothingless.Letusconcludethisnecessarilyhastysketchbyhazardingasome-whatmoreformalstatementofPeirce’smaturesemeiotic.Itises-sentialthatwebeginwith“interprets...asbeingasign”andonlythendefine“sign”:Something(whetherfeeling,action,thought,etc.),R,interpretssomethingelse,X,asbeingasign,S,ofOifandonlyif(a)RisformedinresponsetoX,(b)Risgoal-directed,(c)RwillnotcontributetoitsgoalunlessOobtains,and(d)R’sgoal-directedresponsetoXhassomebasisinarelationorapparentrelationofXtoOorofthingsofX’stypeorapparenttypetothingsofO’stype.Inthisformula,oursymbol“O”standsintheplaceofexpressionsusednonreferentially.“Contribute”coversawiderangeofcases.Atoneextreme,O’sobtainingmakesRitselftobeitsgoal’sfulfillment.Attheoppositeextreme,Risanactofdesperationwhich,ifOob-tains,lowerstheprobabilityoffailureminutely.Inbetween,OmakesRfulfillaconditionofsuccess,butonlytosomedegreeofprobabil-itywhethergreatorsmall;andtheconditionmaybenecessaryandsufficient,oroneoftheseonly,orneither.Next,wecandefine“sign”intermsoftherebeinganunmistakenbasisforinterpretation:Xisasign,S,ofOifandonlyifthereisabasisonwhichapossiblecreature,havingapurposefordoingso,couldunmistakenlyinterpretitasbeingasign,howeverfallible,ofO.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns235Significanceisrelativetopurpose,butpossiblepurposesuffices.Thispointtendstomissed,sincehumancuriosityissuesinpurposestowhicheverypossiblekindofsignisrelevant.Notice,also,thatoneandthesamething,X,maybemanydifferentsigns,S,S’,...,relativetodifferentbasesofinterpretationandrelativetodifferentpossiblepurposes.Afoxyodorisasignofdangertotherabbitbutofdinnertothecougar;“E=mc2”isastatementofphysicallawtothephysicistbuttothegeneralpublicitisaniconofbraininess;tothosewhoworshipedit,anancientreligiousartifactwasanicon,evenquaembodiment,ofsuprahumanpowers,buttotheanthropologistitisanindexofsocialorganizationandculturaldevelopment.9.ramificationsTheprecedingmightbefaultedforfillingintheblankstooboldly.Initsdefense:itmakessenseoutofwhatlittlePeircesaidin1907;itcorrectsadeficiencyinhisearlytheoryofsigns;and,nowtobeshown,itisborneoutbysomefurtherdistinctions,ofconsiderableimportance,thatPeircedevelopedfrom1906to1909.Forthosedis-tinctionsarerequiredby,andwouldmakenosenseapartfrom,ateleologicalaccountofsemeiosis.In1904and1906,whenPeircefirsthintedatwhatwouldbe-cometheemotional/energetic/logicaltrichotomyofinterpretants,heproceeded,inthenextparagraph(CP8.333)orinthesamepara-graph(CP4.536),toclaim,asaseparatematter,thateverysignhasthreeinterpretants,which,in1906andagainin1909(SS.110–11),henamedtheimmediate,dynamic,andfinal.Clearly,then,thetwotrichotomiesareintendedtobedistinct.16Infact,theyarenotclassi-ficationsinthesamesense.Ratherthanbeinganontologicaltypingofinterpretants,theimmediate/dynamic/finaltrichotomyreferstostagesofsemeiosis.Atanyoneofthosestages,aninterpretantmaybeofvariousontologicaltypes,hence,eitheremotionalorenergeticorlogical.“MyImmediateInterpretant,”Peircewrotein1909,“isimpliedinthefactthateachsignmusthaveitspeculiarInterpretabilitybe-foreitgetsanyInterpreter”;again,“TheImmediateInterpretantisanabstraction,consistinginaPossibility”(SS.111).Inthesamepas-sage(cf.CP8.314–5),hedescribedthedynamicinterpretantasanactualeffectonaninterpreter,andthefinalinterpretantas“theoneCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n236t.l.shortInterpretativeresulttowhicheveryInterpreterisdestinedtocomeiftheSignissufficientlyconsidered,”oras“thattowardwhichtheactualtends.”Adynamicinterpretant,ifitisaninterpretantofasign,S,actualizesS’simmediateinterpretant;theoneisapossiblefeeling,action,orthoughtorhabit,whiletheothermakesthatpos-sibilityactual.Asthesamepossibilitycanbeactualizedindifferentways,mostsignscanhaveanynumberofdynamicinterpretants.Adynamicinterpretantmightrefineorsupplementtheimmediatein-terpretantitactualizes,orevennegateit,aswhenonethinksthataclaim(thereinunderstood)ismistakenordishonest.Aseriesofdy-namicinterpretants,informedbyotherobservations,mayapproachthefinalinterpretant,i.e.,theonethatwouldbeideallyadequatetoitspurpose.17Liketheimmediateinterpretant,thefinalinterpretantmayremainanunactualizedpossibility.Butasigncanbecorrectedonlyinrespecttosomerealitythatitpurportstorepresent.Eveniftheobjectsignifiedfailstoexist,theremustbeanimplicitreferencetosomethingthatdoesexist–thegrublesslog,themonsterlessloch,theidealrealmdevoidoffour-sidedtriangles–inrelationtowhichthefailedreferenceoc-curs.Hence,Peircedistinguishedbetweenasign’simmediateanddynamicobjects,thelatterbeingarealityandtheformerwhatthesignrepresentsthatrealitytobe.Forsomesigns,e.g.,apurelyquali-tativeicon,therewouldseemtobenodistinctionbetweendynamicandimmediateobjects.Wheresuchadistinctiondoesapply,asignmustsomehowindicateitsobjectinamannerthatenablesonetoidentifyitwiththeobjectsofothersigns;itisthroughtheagree-mentordisagreementofdiversesignsofthesamething(“collateralobservation,”inPeirce’sphrase)thatadiscrepancyofanimmedi-atefromadynamicobjectmaybediscerned(CP4.536,8.314,333,343,EP2:404–9).Asign’simmediateobjectandimmediateinterpre-tantaretwosidesofonecoin;soalsoitsdynamicobjectandfinalinterpretant.Peircewrote,“...mythreegradesofinterpretantwereworkedoutbyreasoningfromthedefinitionofasign”(SS.111).Buttherecouldbenofinalinterpretantifinterpretationwerenotend-directed.Ergo,thesegradespresupposeateleologicalaccountofsemeiosis.18Fur-thermore,therecouldnotbejustonefinalinterpretantpersign,ifsig-nificancewerenotcorrelativewithinterpretativepurpose,makingCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns237thesamethingdifferentsignsrelativetodifferentpurposes.Ergo,theideasofafinalinterpretantandadynamicobjectpresupposeananal-ysisofsignhoodliketheoneIhavesuggested.IdonotseeanotherwaytomakegoodsenseofPeirce’swritingsonsigns.ManywritershaveadoptedaversionofPeirce’searlytheoryasifitwerehisonlytheoryofsigns,fromwhichtheyhavederivedideasof“unlimitedsemiosis”(Eco1976:68–72)and“theindefinite-nessofreference”(Derrida1974:49).EvenifthosedoctrineswerenotcombinedwithanunPeirceaninsistencethatallsignificance,evenoficonsandindices,dependsonpurelyconventional“codes,”theywouldentailarelativismandirrealismthatareutterlyopposedtoPeirce’sownview.Inhismaturesemeiotic,theideasofafinalinterpretantanddynamicobject–evidentlyageneralizationofhisearlyidentificationoftruthasafinalfixationofbeliefandofre-alityastruth’sobject–extendthestructureofobjectivityfarbe-yondnaturalscience,factualassertion,and“intellectualmeaning,”tointerpretantsthatareemotionalorenergeticandtosignsthusin-terpretable.Insomecases,thefinalinterpretantwillnotbeatruetheorybut,rather,anappropriateactionorajustappreciation.Theimplicationsofthisforethicaltheoryandaesthetictheoryhaveyettobeexploited.notes1.WhydidPeircespeakofinterpretantsandnotofinterpretation?Heneversays.Isuggestthataninterpretationisaclassofinterpretantsthatare,insomesenseyettobedefined,equivalent.Aninterpretantisthusoneparticularbearerofaninterpretation.2.AnonstandardorLeibnizianconceptionofcontinuityisentailed;itisnosurprise,then,thatPeircesometimesentertainedsuchaconception.ButthisisoflittleimportanceforPeirce’stheoryofsigns,since,asweshallsee,heeventuallyabandonedthedoctrineofaninfiniteprogressusandregressusofthought-signs.3.ThisproblemhadbeennoticedbyGeorgeGentry,1952,andWilliamAlston,1956.SeealsoShort1981a.4.Isay“thing,”meaningindividualperson,horse,tree,becausePeircesometimesusedtheterm“individual”inanextremesensethatexcludesallspatiotemporalcontinuityand,hence,lawfulnessinitsmake-up.Inthetermsofhislatermetaphysics,individualsinthisextremesenseareCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n238t.l.shortactualbutnotreal;however,suchactualitiesarecomprisedwithinreal-ities,andthatdistinguishesPeirce’slaterfromhisearlierview,whereinindividualityisidealandneitheractualnorreal.5.Moreremainstobesaidaboutthesethreedivisionsofsignandtheirintersection(CP2.254–64),butnotinthisspace(seeShort1982).6.ThechangeindoctrineisadvertisedbyPeircehimselfinaMonistarticleof1897(CP3.527)referringtohisnominalisticviewofpossibilityonlyoneyearearlier(inarelatedMonistarticle,atCP3.442).MaxFischthereforegives1897asthedateofthechange(Fisch,1986:193–5,199n24).Buttheunpublishedmanuscript,“TheLogicofMathematics,”c.1896,anticipatesthatchange(CP1.420),and,asthesecondMonistarticlewaspublishedinJanuary,itmusthavebeenwrittenin1896.So1896seemsthemoreaccuratedate.7.Thisiscontradictedbyamuch-quotedpassageof1902,Peirce’sdefi-nitionof‘Sign’forBaldwin’sDictionary:“Anythingwhichdeterminessomethingelse(itsinterpretant)torefertoanobjecttowhichitselfrefers(itsobject)inthesameway,theinterpretantbecominginturnasign,andsoonadinfinitum....Iftheseriesofsuccessiveinterpre-tantscomestoanend,thesignistherebyrenderedimperfect,atleast”(CP2.303).1902is,however,thedateofpublication;thepassagewasprobablywrittenbeforethepassagesinPeirce’sunpublished“Syllabus,”c.1902,thatemphasize,rather,themerecapacityofasign(representa-men)todetermineaninterpretant:“...whilenoRepresentamenactu-allyfunctionsassuchuntilitactuallydeterminesanInterpretant,yetitbecomesaRepresentamenassoonasitisfullycapableofdoingthis;anditsRepresentativeQualityisnotnecessarilydependentuponitseveractuallydetermininganInterpretant...”(CP2.275).ThisisnottheonlyinstanceinwhichPeirce’smostclearanddefinitestatementofanextremepositionheraldsitsrejectionordrasticqualificationshortlyafter.8.Possibly,thefirstmentionofthismodificationisearlier,inthepassagedatedintheCollectedPapersc.1903,wheretheterm“representamen”isusedforthebroadestsenseof“sign”andPeircewrote,“ASignisarepresentamenofwhichsomeinterpretantisacognitionofamind”(CP2.242).9.BothofthesepassagesarefromMS318,asnumberedinRobin1967.ThisMSwaspublishedinpartinCP5.11–13,464–96,wherethedategivenisc.1906;butitisnowknowntobeof1907(seeRobin,1967:36).MS318,consistingoflabyrinthinemultipledraftsofanunfinishedarticle,hasneverbeenpublishedinitsentirety,despiteitsrichnessandimportance(thelargestportionpublishedsofarisEP2:398–433).ThefirstoftheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nTheDevelopmentofPeirce’sTheoryofSigns239twopassagesquotedisstillunpublished;thesecondmaybefoundinNE3/1:493–4.Asimilarpassage,fromavariantdraft,isinCP5.491;init,Peircesubstituted“final”for“ultimate,”whichhasengenderedsomeconfusion,since,asweshallsee,Peircesometimesusedtheterm“finalinterpretant”inaquitedifferentway.10.ThispassageisdatedintheCollectedPapersasc.1905.However,ifmyaccountofPeirce’sdevelopmentiscorrect,itmustbelaterthanthelastpossibledateatwhichPeircecouldhavealteredhis1906Monistarticle,“ProlegomenatoanApologyforPragmaticism,”fromwhichthequotationfromCP4.536istaken.AndsinceitissomuchakininwordingandthoughttoMS318,itswritingisprobablynearerto1907thanitisto1905.RichardRobindatesitasc.1905–8(1967:30).11.ThispassagefollowsthesecondofthetwofromMS318quotedearlier(seen.9).12.Thepassagehastobereadcarefully;seeShort1981a:204–7.13.Peirce,asiswellknown,alsopositedobjectivechance,i.e.,theoccur-renceofeventsnotconformingtomechanisticlaws(“tychism”).Ibe-lievehedidthisonlyinordertoaccountfortheevolutionoflawitself,andnottoaccountforirreversibletendenciestowardendsinthesystemsstatisticalmechanicsstudiesandinbiologicalevolution.However,ifIamwrongaboutthat,thentychismonlystrengthensthepresentpoint,thatend-directednessisirreducibletomechanisticlaw.14.Peircewrotethat“Apurposeismerelythatformoffinalcausewhichismostfamiliartoourexperience”(CP1.211),implyingthatinordi-naryusageapurposemustbesomethingofwhichoneisconscious.WrightandWimsattwouldagree,butIfindnothinginordinaryus-agethatjustifiesthatview.Weordinarilyspeakofthepurposeofanorganorotherorganicfeatureorinstinct,etc.,andwedosowithoutmakinganyassumptionofitshavingbeencreatedconsciously.Whatdoesseemtobeimpliedisthattheitemhasbeenselectedsomehowforatypeofeffect.HenceIadoptthemoreconvenientandfamiliarterm“purpose”inlieuof“finalcause.”YetthepointisPeirce’s:thatwhatmattersisselectionfortype,regardlessofhowthatselectionismade.15.Butifitsobjectdoesnotexist,orevenifitneednot,thenisthesignreallyoneofatriad?Peirce’sconceptionofthedynamicobjectresolvesthisproblem:seeSection9.16.Yetothercommentators,fromBuchler1939toLalor1997,havethoughtotherwise;fordiscussionoftheissue,seeFitzgerald1966:78n9,Short1981a:212–19,Liszka1990,1996:120–3.Short1996aisafulldefenseoftheviewpresentedhere.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n240t.l.short17.Whenourpurposeistoknowthetruth,thefinalinterpretantwillbesymbolic;hence,itwillhave,butitwillnotbe,anultimateinterpretant.18.In1906,Peirceconfessedthathisconceptionofthefinalinterpretantwas“notyetquitefreefrommist”(CP4.536).Norshouldithavebeenatthatdate,ifhissemeioticwasnotmadeexplicitlyteleologicalbefore1907.Infact,hisconceptionsoftheimmediateandthefinalinterpre-tantsvariedagooddeal,asdidhisterminology(cf.CP8.333of1904and8.343of1908).CambridgeCompanionsOnline©CambridgeUniversityPress,2006\npeterskagestad10Peirce’sSemeioticModeloftheMind1.introductionInthischapter,IshowhowPeirce’smodelofmindisgroundedinhissemeiotic,orgeneraldoctrineofsigns,agroundingmadepossiblebythelogicalpriority,inPeirce’sthought,oftheconceptofsignovertheconceptofmind.Ithencomparethismodelofmindwithsomemorerecentdoctrinesandtheories,andconcludewithsomecommentsonPeirce’srelevanceforcognitivescience,includingbothartificialintelligenceandhuman–computerinteraction.2.peirce’sdoctrineofsignsPeirce’sdoctrineofthoughtsignswasfirstintroducedinhisjustlyfa-mous1868articlesinTheJournalofSpeculativePhilosophyandlaterdevelopedingreaterdetailfrom1895untilPeirce’sdeathin1914.Inhis1868papersPeircespecificallytargetedDescartesandCartesian-ism,andarguedthatwehavenoabilitytothinkwithoutsigns.Thisargumentpresupposesapriorargumentthatallself-knowledgecanbeaccountedforasinferencesfromexternalfactsandthatthereisthusnoreasontopositanypowerofintrospection(CP5.247–9).Weneed,therefore,tolooktoexternalfactsforevidenceofourownthoughts,anditisthenanear-tautologytoconcludethattheonlythoughtssoevidencedareintheformofsigns:“Ifweseekthelightofexternalfacts,theonlycasesofthoughtwhichwecanfindareofthoughtinsigns”(CP5.251).Thisdoctrineisrepeatedin1909inthefollowingwords,whichrecallSocratesandAristotle:“Allthink-ingisdialogicinform....Consequently,allthinkingisconducted241CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n242peterskagestadinsignsthataremainlyofthesamegeneralstructureaswords;...”(CP6.338).BysayingthatallthinkingisdialogicPeirceishereplainlyim-plyingwhathemadeexplicitonnumerousotheroccasions,towit,thatallthinkingisinherentlycommunicationalandthusinherentlysocial.Evenmysolitaryruminationspresupposeacommunicationalstructureenabledbymymembershipinasociety.Wewillnote,how-ever,thatPeirceisnotsayingthatallthinkinginconductedinwords,onlythatitisconductedinsignsthatareofthesame“generalstruc-ture”aswords;thatis,heisnottakingthe“Orwellian”positionattackedinrecentyearsby,e.g.,StevenPinker,whichequateslan-guageandthought.Thoughtsarenotsimplywords,andPeirceneversupposedthattheyare.Infact,Peirce–likeanymathematician–thoughtnaturallanguageinadequateforthespecialpurposeofex-actreasoning,whichrequiresaspecialsymbolism,eitheranalge-braicoradiagrammaticone.Inkeepingwiththisview,hedevotedagreatpartofhislifetothedevelopmentofalgebraicanddiagram-maticnotationsforlogic,conceivingoftheuniversalandexisten-tialquantifiersindependentlyofFrege,anddesigninglogicdiagramswhichhavelatelyfoundfertileapplicationsinartificialintelligenceresearch.Thespecialsymbolscomprisingtheseandothernotationsare,however,ofthesame“generalstructure”aswords,andweturnnexttoanoverviewofwhatPeircetooktobethegeneralstructureofsignhood.In1868Peirceintroducestheconceptof“sign”asfollows:Nowasignhas,assuch,threereferences:first,itisasigntosomethoughtwhichinterpretsit;second,itisasignforsomeobjecttowhichinthatthoughtitisequivalent;third,itisasign,insomerespectorquality,whichbringsitintoconnectionwithitsobject.(CP5.283)Thesign,then,constructsorimposesanirreduciblytriadicrelationontheobjectofthesign,therespectorqualitythroughwhichthesignsignifies,andthethoughtthatconnectsthesignwithitsobject,athoughtwhichPeircelaterwastodubthe“interpretant”ofthesign.ThisirreducibilityoftriadicrelationsisoneofthecentralthemesinPeirce’sthought–athemealsolaterdevelopedbythenovelistandessayistWalkerPercy,whodescribedhimselfinthisrespectas“athiefofPeirce”(Samway,1995:130;Percy,1975:3–45).Asimpleex-ampleofatriadicrelationis‘JohngavethebooktoJim.’ThisrelationCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sSemeioticModeloftheMind243containsasitscomponentsthetwodyadicrelations‘Johngavethebook’and‘Jimreceivedthebook,’butthesetwodyadicrelationsdonottogetherexhaustthetriadicrelation‘JohngavethebooktoJim,’sincethesametwodyadicrelationswouldalsobepresentinthecasewhereJohngavethebooktoJane,whointurngaveittoJim.Similarly,whenaredlightsignifiestomethatIshouldstop,itmaybethecaseboththatIseetheredlightandthatIstop,buttheseconditionsmayholdwithouttheredlightsignifyinganything.Forinstance,Imayseetheredlight,notregisteritssignificance,butstopbecauseapedestrianiscrossingthestreetinfrontofmeorforsomeotherreason.Sothetwodyadicrelationsdonottogetherconstituteasignrelation.Generally,Peirceheld,thephysicalworldcanbecompletelydescribedintermsofdyadicrelationsandcom-poundsofdyadicrelations,whereasmentalphenomenacanonlybedescribedintermsoftriadicrelations.ThisconceptionofsignneedstobedistinguishedfromCharlesMorris’s(1946:7)later,behavioristdefinition,whichowesagreatdealtoPeirce,butwhichMorrisalsotookpainstodistinguishfromPeirce’sdefinition:Ifanything,A,controlsbehaviortowardsagoalinawaysimilarto(butnotnecessarilyidenticalwith)thewaysomethingelse,B,wouldcontrolbehaviorwithrespecttothatgoalinasituationinwhichitwereobserved,thenAisasign.Ashasbeennoted,e.g.,byGerardDeledalle(´2000:116–18),andbe-forethatbyMorrishimself(1946:288–9),thisisapsychologisticandbehavioristicdefinition,whichonbothcountsdivergesfromPeirce’scommunicationaldefinition.Deledalle(2000:116–118)alsocorrectlynotesthatMorrisadoptedPeirce’sterminology,butgavePeirce’stermshisownnewmeanings.Forinstance,Peirce’s“in-terpretant”(ofwhichmoreanon)isdefinedbyMorris(1946:17)notasasign,butasadispositiontorespond.WhatmaybeopentodebateiswhetherDeledalleisalsorightinaccusingMorrisofdis-solvingPeirce’striadicrelationsintopairsofdyads.Morris(1946:288)consideredthequestionandconcludedthattheintroductionofreinforcementintostimulus–responsepsychologymadepsychologyitselftriadicinthePeirceansense,aconclusionrejectedbyDeledalle(2000:118).IsuspectDeledalleisright,sinceanyirreduciblytriadicconceptofreinforcementwouldneedtobeintensionalinnature;anyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n244peterskagestadcausalconceptionofreinforcement,suchaswouldberequiredbyanyversionofbehaviorismfamiliartome,wouldbereadilyreducibletosequencesofdyads.Letusreturntotheinterpretant.Itisessentialtonotethattheinterpretantofthesignisnotidenticaltotheinterpreter;i.e.,theindividualmindinterpretingthesignisnotoneofthethreerefer-encesthatconstitutesignhood.Theinterpretantmay,andperhapsnormallydoes,ariseintheconsciousnessofanindividualsubject,butthesign-characterofthesignisnotaffectedbywhetherornotitdoes,andaneffectonanindividual’sconsciousnessisnopartofthedefinitionofa“sign.”Peirce’ssemeiotictriadisthusquitedifferentfromWalkerPercy’striadtoken–object–mind,exemplifiedbyHelenKeller’smind-openingrecognitionthatthetokensw–a–t–e–rbeingwrittenintoheronehandbyherteacherstoodforthewaterrunningoverherotherhand,thusconstitutingatriadamongthetokens,thewater,andHelen’smindwhichtookonetostandfortheother.PercyhimselfcertainlyexpressedhisawarenessofthisdifferenceinhiscorrespondencewithKennethKetner(Samway,1995:48).With-outsharingMorris’sbehaviorism,Percydoesseemtosharehispsy-chologisticconceptionofsignhood.ItistruethatPeircesometimesincludesareferencetotheindividualsubject,thesign-interpreter,asinthispassagefrom1897:Asign,orrepresentamen,issomethingwhichstandstosomebodyforsome-thinginsomerespectorcapacity.Itaddressessomebody,thatis,createsinthemindofthatpersonanequivalentsign,orperhapsamoredevelopedsign.ThatsignwhichitcreatesIcalltheinterpretantofthefirstsign.(CP2.228–9)HerePeircemayappeartodefine“sign”intermsofthetriad“something–somebody–somerespect,”i.e.,object–interpreter–ground.Butthisisonlyanappearance;Peircegoeson,intheverynextparagraph,tostatethattherepresentamenis“connectedwiththreethings,theground,theobject,andtheinterpretant,”nottheinterpreter.AndinhisthirdLowellLectureof1903,Peircedefines“representamen”withoutanyreferencetoaninterpreter:arepresentamenisasubjectofatriadicrelationtoasecond,calleditsobject,forathird,calleditsinterpretant,thistriadicrelationbeingsuchthattherepresentamendeterminesitsinterpretanttostandinthesametriadicrelationtothesameobjectforsomeinterpretant.(CP1.541)CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sSemeioticModeloftheMind245Signs,itmaybesaid,areofparticularinteresttoushumanbeingsbecausewedo,afterall,engageintheinterpretationofsigns,andbecausethisinterpretiveactivityiswhatconstitutesourthinking.Butourinterpretiveactivityisinnowayconstitutiveofthesignasasign,andthisiswhyPeircecanwithoutcircularitydefinethinking,andhencemind,intermsofsemeiosis,orsign-action,assignsareinthefirstinstancedefinedwithoutreferencetotheinterpretingmind.Inotherwords,apotentialinterpretiveactivityispresupposedbytheconceptofsignhood,butthesubjectofthatactivityisleftundefined,exceptinsofarasitispreciselythesubjectofaninterpretiveactiv-ity.Thisis,as,e.g.,Deledalle(2000:59–61,andpassim)hasnoted,amentalistic,notamaterialistic,conceptionofsignhood,butnotapsychologisticone.Theinterpretermight,butneednot,beanindi-vidual,andDeledalle(2000:33)approvinglyquotesthispassage,oneofmanyinwhichPeirceappearstolocatetheinterpretiveactivityinthecommunity,ratherthaninindividualminds:Meantime,weknowthatmanisnotwholeaslongasheissingle,thatheisessentiallyapossiblememberofsociety.Especially,oneman’sexperienceisnothing,ifitstandsalone.Ifheseeswhatotherscannot,wecallithal-lucination.Itisnot“my”experience,but“our”experiencethathastobethoughtof;andthis“us”hasindefinitepossibilities.(CP5.402)Communication,itmightbesaid,isnotsomethingwedowiththoughtsantecedentlyformedinsideourheads.Communication,toPeirce,isthecontextinwhichthoughtsareformed,andislogicallypriortothinkingprocessestakingplaceinindividualminds.Whatbringsthesignintoconnectionwithitsobjectissomema-terialqualityitpossesses,whichenablesittorepresentitsobjectinaparticularway.Apicturesignifiesbymeansofanassociationthatconnectsit,intheinterpreter’sbrain,withitsobject.Aweathervaneoratallysignifiessimplybybeingphysicallyconnectedwithitsob-jectbyachainofcausation.Words,finally,canfunctionassignsonlybecausetheyarecapableofbeingconnectedintosentencesbymeansofacopula(CP5.286).Later,Peircewastoclassifyallsignsasicons,whichsignifybyvirtueofresemblance,indices,whichsignifybyvirtueofaphysicalconnectionwiththeobject,andsymbols,whichsignifybyvirtueoftheexistenceofarulegoverningtheirinterpreta-tion(CP2.276–92).Now,theexactphysicalembodimentofsymbolsisofcourselargely,butnotentirely,amatterofconvention.TheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n246peterskagestadchoiceofRomanversusOldEnglishletteringmaybepurelycon-ventional,butthatiscertainlynotthecasewithone’spreferenceforArabicoverRomannumeralsforthepurposeofperforminglongdivision–apointalsomadebyPeirce;cf.Eisele(1979:11–34).TheimportantpointPeircemadein1868,andwaslatertodevelopindepth,isthattheexistenceofsymbolsdependscruciallyontheex-istenceofanotationthatiscapableofsymbolicinterpretation,andmoreoverthatourthinkingisfacilitatedorimpededbythespecificphysicalfeaturesofournotation.In“OnaNewListofCategories,”publishedin1867,Peirceintro-ducedhisfamoustrichotomyamongthethreechieftypesofsigns,orrepresentations,asPeircecalledthematthisstage:First.Thosewhoserelationtotheirobjectsisamerecommunityinsomequality,andtheserepresentationsmaybecalledlikenesses[later,icons].Second.Thosewhoserelationtotheirobjectsconsistsinacorrespon-denceinfact,andthesemaybetermedindicesorsigns.[Theterm‘signs’wassoontobeextendedtocoverallrepresentations.]Third.Thosethegroundofwhoserelationtotheirobjectsisanimputedcharacter,whicharethesameasgeneralsigns,andthesemaybetermedsymbols.(CP1.558)Forinstance,aweathervane,measlesspots,temperaturereadings,billowingsmoke,andexclamationsofpainareindices,apaintedportrait,anhistoricalnovel,oratheatricalperformanceisanicon,whileaphotographoratelevisednewsbroadcastisbothaniconandanindex.Theweathervaneindicatesthedirectionofthewindbyvirtueofhavingacausalconnectionofacertainkindtoitsobject,thewind.Apaintedportraitsignifiesapersonbyresemblingthatperson.Aphotographsignifiesbymeansofbothrelations.Anexam-pleofasymbol,finally,wouldbeaword,insofarasitsignifiesbymeansofapurelyconventionalrelationtoitsobject–unlike,e.g.,anonomatopoeia.Thethreetypesofsign,finally,standinahierarchi-calrelationship:symbolspresupposeiconsfortheirownexistence,whileiconsinturnpresupposeindices.Thetrichotomyoficon,index,andsymbol,wheresignsareclas-sifiedwithspecificreferencetotheirrelationtotheirobject,isonlyone–albeitarguablythemostbasicone–ofPeirce’svariousschemesforclassifyingsigns.Anothertrichotomydividessignsintoterms(laterrhemes,correspondingtopredicates),propositions(laterCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sSemeioticModeloftheMind247dicisigns),andarguments(laterlegisigns);argumentsareinturnde-ductive,abductive,orinductive,andsoforth.Peirce’smostfamousthree-dimensionalclassificationofsignsrecognizestenclassesofsigns.Wehavehereonlyattemptedanoverviewofthegeneralstruc-turesharedbywords,numerals,pictures,andothersigns.Finally,inPeirce’sviewasinJohnLocke’sbeforehim,thoughtsaresignsand,asweshallseenext,semeioticthusimpliesacompletephilosophyofmind,inwhichcognitionisthematizedasthedevelopmentofsigns,andnotasasuccessionofconsciousstatesofmind.3.thesemeioticmodelofthemindWritersasdiverseasthephilosopherJamesFetzer(1990:31–50)andthenovelistWalkerPercy(1975:3–45)havecreditedPeircewithde-velopingasemeioticmodelofthemind,i.e.,amodelinwhichsigninterpretationand,withit,intentionalityareessentialattributesofthemind.Wenowtakeacloserlookatwhatitmeansforthemindtobeessentiallyasigninterpreter.Byclaimingthatallthinkingissignification,i.e.,theproductionandinterpretationofsigns,Peircewasnotdenyingthepsychologicalfactthatthoughtscanbesubjectivelyexperiencedasinternalstatesofmind;whathewasclaimingisthatsuchexperiencesaremerelypsychologicalfacts:“[Every]thought,insofarasitisafeelingofapeculiarsort,issimplyanultimate,inexplicablefact”(CP5.289).Themeaningorintellectualvalueofathoughtliesinitspotentialforfurtherinterpretation,whetherbymymindorbysomeothermind;thatis,itliesinitssignhood.Knowledge,inPeirce’ssemei-oticdoctrine,consistslessinstatesofmind(“ultimate,inexplicablefacts”)thaninthepotentialityofexternalobjectstoinducecertainstatesofmind,andthispotentialitydependsonthespecificphysicalcharacteristicsofsaidexternalobjects.Consciousness,Peirceheld,isnotanessentialattributeofmind.Thus,inadiscussionofthena-tureofpsychologyfrom1902,Peircewrote,“Iholdthatpurpose,orrather,finalcausation,ofwhichpurposeistheconsciousmodifica-tion,istheessentialsubjectofpsychologists’ownstudies;andthatconsciousnessisaspecial,andnotauniversal,accompanimentofmind”(CP7.366).InanespeciallytopicalpassagePeircegoesontoemphasizethedependenceofourlanguagefacultyonexternaltoolsforlinguisticexpression:CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n248peterskagestadApsychologistcutsoutalobeofmybrain(nihilanimaleamealienumputo)andthen,whenIfindIcannotexpressmyself,hesays,‘Yousee,yourfacultyoflanguagewaslocalizedinthatlobe.’Nodoubtitwas;andso,ifhehadfilchedmyinkstand,IshouldnothavebeenabletocontinuemydiscussionuntilIhadgotanother.Yea,theverythoughtswouldnotcometome[emphasisadded].Somyfacultyofdiscussionisequallylocalizedinmyinkstand.(CP7.366)Havingdiscussedthispassageindetailelsewhere(Skagestad1999a),Iwillrestrictmycommentsonthisoccasion.ThecontextmakesclearthatonethingPeirceisdoinginthepassagequotedistoridiculetheideathatthefacultyofdiscussion,oranyothermentalfaculty,islocalizedinthebrainoranywhereelse.Heisnotsaying,nordoeshemean,thatthefacultyofdiscussionislocalizedintheinkstand.Thatitisnot,becomesclearwhenhegoesontosay,intheverynextsentence,“Itislocalizationinthesenseinwhichathingmaybeintwoplacesatonce”(CP7.366).Wemightnowbetemptedtodismissthereferencetotheinkstandasonlyajoke:localizationinthesenseinwhichathingcanbeintwoplacesatonceis,ofcourse,thesamethingasnolocalizationatall.So,itmightbeargued,whatPeirceisdoingisusingtheveryludicrousnessoftheideaofthemindbeinglocalizedinaninkstandasawayofhighlightingtheequalludicrousnessofsupposingthemindtobelocalizedinthebrain,oranywhereelse.This,however,willnotquitedo.Theinkstandexampleisajokeoncontemporarypsychologists;nodoubtaboutit.Butthosewhowouldarguethattheinkstandexampleisonlyajokeneedtoexplainpre-ciselyhowtheludicrousnessoftheinkstand-localizationhighlightstheludicrousnessofthebrain-localization.AndwhenPeircegoesontosay,“Onthetheorythatthedistinctionbetweenpsychicalandphysicalphenomenaisthedistinctionbetweenfinalandefficientcausation,itisplainenoughthattheinkstandandthebrain-lobehavethesamegeneralrelationtothefunctionsofthemind”(CP7.366),theanswerplainlyisthattheinkstandandthefrontallobestandeitherinidenticalorinsomerespectequivalentrelationstothefacultyofdiscussion.Thisrelationisnotliterallocalization,butrathersomethingthatmaybecalled“virtual”localization.Thisconcept,farfrombeingananachronism,wouldbequitecongenialtoPeirce,whotracedtheconceptofvirtualitytohisintellectualheroCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sSemeioticModeloftheMind249DunsScotus,andwhohimselfpennedthedefinitionof“virtual”inBaldwin’sDictionaryofPhilosophyandPsychology,towit:“Avir-tualX(whereXisacommonnoun)issomething,notanX,whichhastheefficiency(virtus)ofanX”(Baldwin,1902:vol.2,763,CP6.372).Similarly,the“localization”ofthemindintheinkstandorthebrainisnotlocalization,butsomethingwhichhastheefficiencyoflocal-ization,intheprecisesensethatthisiswhereyoulookforthemind.InPeirce’sviewyoudonotfindthemindinsidethebrain,anymorethanyoufindelectricityinsidecopperwires–ananalogyexplicitlycitedbyPeirce.Youfindthemindwherethereareinkstandsorothermeansofexpressingthoughts,paperorothervehiclesforpreservingandconveyingthoughts,andofcoursebrainscapable,throughtheintermediaryofeyesandhandsortheequivalent,ofinteractingwithexternaltoolsandmedia.Again,inPeirce’swordsfrom1902:Inmyopinionitismuchmoretruethatthethoughtsofalivingwriterareinanyprintedcopyofhisbookthantheyareinhisbrain.(CP6.364)Underlyingtheseutterancesisafullyarticulatedandwelldocu-mentedsemeioticmodelofthemind,inwhichcognitioniscon-struedasthedevelopmentofinternalorexternalsigns,andnotasasuccessionofconsciousstatesofmind.LikeJohnLockebeforehim,andwithexplicitindebtednesstoLocke,Peirceaffirmedthatthoughtsaresigns.Thoughtprocessesmayandoftendotakeplaceentirelytacitly,butwhatmakesthemthoughtprocessesisthesigncharacterofthethoughts,andthischaracterconsistslargelyintheirpotentialityforbeingexpressedandsobeingsusceptibletointerpre-tation.Peirceneverdeniedtheexistenceofconsciousness,andhedidnotdenythatwemayhaveintrospectiveknowledgeofourconsciousmentalstates,buthesimplydidnotregardcognitionasconsistingofsuchconsciousstates.Cognitionconsistsinthemanipulationofsignswhichmaybeexternallyembodied;aseachsigniswhatitisbyvirtueofitspossiblelaterinterpretations–i.e.,virtually–sotheminditselfisvirtual.In1902Peircewasexplicitlyarguingfortheclaimthattheessenceofmindisnotconsciousness,butpurpose,orfinalcausation.ThisviewofthemindhadalreadybeenarticulatedinsomeofPeirce’searlywritings,suchashisclassic1868articleswhich,aswesawearlier,alsointroducedPeirce’sdoctrineofsigns:CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n250peterskagestadFinally,nopresentactualthought(whichisamerefeeling)hasanymean-ing,anyintellectualvalue;forthisliesnotinwhatisactuallythought,butinwhatthisthoughtmaybeconnectedwithinrepresentationbysubse-quentthoughts;sothatthemeaningofathoughtisaltogethersomethingvirtual....Atnoinstantinmystateofmindistherecognitionorrepresen-tation,butintherelationofmystatesofmindatdifferentinstantsthereis.(CP5.289)Inalettertotheeditor,WilliamT.Harris,Peirceelaborated:Idonotsaythatweareignorantofourstatesofmind.WhatIsayisthatthemindisvirtual,notinaseriesofmoments,notcapableofexistingexceptinaspaceoftime–nothinginsofarasitisatanyonemoment.(CP8.284)StudentsofrecentphilosophyhavenodoubtbythispointfoundthemselvesremindedofLudwigWittgenstein.WithoutwishingtoventuredeeplyintotheverytrickyquestionoftherelationofPeirce’sthoughttothatofthelaterWittgenstein,IcannotforbeartonotethatPeirceismakingapointsimilartoonewhichWittgenstein(1958:152e)waslatertomakeasfollowsinthePhilosophicalInvestigations:Wesay“Iamexpectinghim”,whenwebelievethathewillcome,thoughhiscomingdoesnotoccupyourthoughts.(Here“Iamexpectinghim”wouldmean“Ishouldbesurprisedifhedidn’tcome”andthatwillnotbecalledthedescriptionofastateofmind.)AsIunderstandPeirce,hewouldagreewithWittgenstein’sparen-theticalcomment,butperhapsnotwiththesentencethatprecedesit.EvenwhenNN’sexpectedarrivaldoesoccupymythoughts,Peirce’sviewisthatwhatitmeansforhisarrivaltooccupymythoughtsismylatersurpriseshouldNNnotshowup,andnottheconsciousstate(“amerefeeling”)experiencedwhileexpectingNN;inotherwords,whatoccupiesmythoughtsisnotastateofmind.InPeirce’sview,thatis,thoughtsareinherentlydispositional,aviewechoingthatofDunsScotus(whomPeircegreatlyadmired)andinturnlaterechoedbyKarlPopper,whoheldknowledgetoresideexosomatically,inbooks,articles,andthelike,ratherthanintheconsciousexperienceoftheauthorsorreadersofsaidbooks,etc.Thatconsciousexperi-ence–inPopper’sviewasinPeirce’s–wasrendereddispensiblefortheanalysisofknowledgeorofmindbythatinherentlydispositionalnatureofknowledgewhichallowsknowledgetobeexosomaticallyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sSemeioticModeloftheMind251embodied.InPopper’swords,whichstrikinglyrecallPeirce’sdefini-tionof‘sign’assomethingwhichiscapableofbeinginterpreted:Itisitspossibilityorpotentialityofbeingunderstood,itsdispositionalchar-acterofbeingunderstoodorinterpreted,ormisunderstoodormisinterpreted,whichmakesathingabook.Andthispotentialityordispositionmayexistwithouteverbeingactualizedorrealized.(1972:116)Abook,then,regardedspecificallyasabook,isneitheraphysi-calnoramentalobject,butresidesinathirdrealmwhichPopper(1972:156)referstoastheworldof“objectivethoughtcontents”andwhichisdistinctfromboththephysicalandthementalworlds.Similarly,whenPeircerefusedtolocatethemindwithintherealmofconsciousness,hewasnotbeingamaterialist;hewasnotplac-ingthemindwithinthephysicalworld.Intheinkstandexample,torecall,thecontrastisnotdrawnbetweenconsciousnessandmate-rialobjects,butbetweenthebrainandtheinkstand–twophysicallocations–asthelocationofthefacultyofdiscussion.Andjustas,toPopper,abookwasnotapurelyphysicalobject,sotoPeircetheink-standwasnotapurelyphysicalobject,butwhat,inhis1901reviewofKarlPearson’sTheGrammarofScience,hetermeda“generalizedpercept”(CP8.144).WhilePopperdoesnot,tomyknowledge,commentonthespecificmaterialqualitiesthatmaycharacterizedifferentbooks,Peirce’sawarenessoftheimportanceofthesematerialqualitiesisrevealed,forinstance,inhispraisingCharlesBabbage(inanobituaryin1871)forpublishingavolumeoftablesoflogarithmswherehetriedfiftydifferentcolorsofpaper,andtenofink,beforesettlingonthecom-binationthatwaseasiesttoread,andthatthusmaximizedthecog-nitivevalueofthetables(W2,459).Peirce’spraiseofBabbageinthisrespectcouldofcoursebejustacasualobservation,ofnophilosoph-icalsignificance;however,alongthesamelineshealsocreditsBab-bagewithinventinganewnotationforkeepingtrackoftheintricateworkingsofhis“analyticalengine.”Later,inhisarticleon“LogicalMachines,”Peirce(1887:169)referstoalgebraas“thebestofallin-strumentsofthought,”andhegoesontocredititspower,aboveall,toapurelynotationalfeature,namelytheparenthesis.Theparen-thesis,Peirceobserves,couldinprinciplebedispensedwiththroughtheuseofplaceholderstorepresenttheresultsofintermediateop-erations–(a+b)c=dcouldbewrittena+b=t,ct=d–butatCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n252peterskagestadtheexpenseofanotationalclumsinessthatwouldmakealgebraalesspowerfulinstrument.AndthesearejustacoupleofexamplesofPeirce’snumerousreferencestoconcreteinstanceswherethespe-cificmaterialqualityofasignenablesittofunctionastheprecisekindofsignitis,thusinturnenablingtheprecisekindofreasoningitmakespossible.Peirce’sobservationsinthepreviousparagraphrecalltheSapir–Whorfhypothesisanditslateroffspring,theorality–literacythesis.IwanttomaketwocommentsontherelationofPeirce’ssemeioticdoctrineofmindtotheselaterhypotheses.Inthefirstplace,whilePeirceisclearlyembracingtheview,laterarticulatedbySapir,Whorf,andothers,thatlanguageisamediumofthought(althoughnottheonlyone,andnotnecessarilythemostimportantone),thepronouncementsquotedabovearequiteneutralwithrespecttotheSapir–Whorfhypothesisproper.Thatis,theviewthatthinkingwithapeniscognitivelydifferentfromthinkingwith-outone,doesnotimply(orpreclude)thatthinkinginonenaturallanguageiscognitivelydifferentfromthinkinginanyothernaturallanguage.Nor,asfarasIknow,hasPeirceevercommentedonthisquestiononewayoranother,orshownmuchinterestinnaturallanguage–asopposedtospecializedartificiallanguages–asamediumofthought.(Imust,however,admittonothavingreadev-erythingPeirceeverwrote.)Ontheotherhand,whatPeirceissayingclearlyisatoddswiththe“mentalese”modelwhichStevenPinker(1994:73–82)advocatesasanexplicitalternativetotheWhorfianhypothesis:ifthinkingweresimplyandsolelymentalprocessingcarriedoutbyaninternalTuringMachine,whoseresultsarethentranslatedintoEnglish,orwhatever,thenitwouldseemthatthepresenceorabsenceofwritingimplementscanmakenodifferencetothethinkingprocess.Inthesecondplace,IdotakePeircetobeembracingsomever-sionoftheorality–literacythesislatermadefamousbyEricHave-lock,WalterOng,MarshallMcLuhan,andothers,butitisimportanttospecifypreciselywhatkindofthesisisentailedorsuggestedbytheinkstandexample.Ifallthinkingissignaction,thenthinkingisatleastinpartanexosomaticprocess(althoughnotpurebehavior),andourthinkingprocesses–andtosomeextentperforcethecon-tentofourthinking–willvarywiththesignvehiclesavailabletous.WhilePeirce,aswehaveseen,isnotdenyingtheexistenceofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sSemeioticModeloftheMind253consciousness,somethingelseseemsheretobeeitherdeniedorrel-egatedtoaninsignificantstatus,namelytheoccurrenceofuncon-sciousthought.Peirce,fortherecord,heldthatunconsciousthought,whiledifferingfromconsciousthoughtonlybyitslesserdegreeofsusceptibilitytobeingcontrolled,constitutesfarandawaymostofourthinking(CP6.569;7.554,bothcitedinBrent,1998).However,heclearlydistinguishedsuchthinkingfromreasoning,properlyso-called(CP7.458–9).4.mindsandmachinesTherelevanceofPeirce’ssemeiotictocognitivesciencehasnotgoneunnoticed.Forinstance,inrecentyearsJamesFetzer(1990:34–50)hascontrastedPeirce’ssemeioticmodelofthemindwiththecur-rentlyinfluentialcomputationalmodelofthemind.AshasbeenpointedoutbyFetzer(1990:31–5),theNewell–Simonconceptofthemindasan“abstractandphysicalsymbolsystem,”sofundamentaltomostifnotallversionsofthecomputationalmodel,lacksthetriadicitycentraltoPeirce’sconceptofasign.Specifically,itlackstherecognitionthatasignisinherentlyasigntosomeone.IntheNewell–Simonmodel,symbolsaresymbolsbyvirtueofcertainfor-malfeaturestheydisplay,irrespectiveofwhetherornotanyorgan-ism(ormachine)actuallytakesthemtostandforanything.FetzerhascontrastedthismodelwithPeirce’ssemeioticmodelofthemind,inwhichthemindisessentiallyasignuser,andinwhichtherecog-nitionofsomethingasasymbol(oranyothersign)impliesthatitisasymboltosomesign-usingorganism(ormachine).Thethesisthatasignsystemcannotfirstbedefinedasapurelyabstract,for-malsystem,andthenafterwardsbegivenaninterpretation–andsoaninterpreter–hasofcoursebeenmostfamouslyarguedbyHilaryPutnam(1980:464–82),inwhatGeorgeLakoff(1987:229–38)hascalled“Putnam’sTheorem.”IthasbeenpointedoutbyEugeneFreemanthatPeirce’sdoctrineofsignsimpliesthatallreasoningisdiagrammatic,acorollaryrepeat-edlymadeexplicitbyPeircehimself,e.g.,inthispassagequotedbyFreeman(Freeman&Skolimowski,1974:477,CP3.393):“Forrea-soningconsistsintheobservationthatwherecertainrelationssubsistcertainothersarefound,anditaccordinglyrequirestheexhibitionoftherelationsreasonedwithinanicon.”ToavoidCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n254peterskagestadmisunderstandingitmustbestressedthatPeirceisnotsaying,hereoranywhereelse,thatallthinkingisdiagrammatic;bythewordreasoningPeircealwaysmeansspecificallytheprocessofdrawinginferences.Thevalueoftheiconicrepresentation,Peircerepeatedlyinsisted,layinthepossibilityitaffordedofperformingexperimentsonourthoughts,bychangingsomeelementsinthediagramandlit-erallyseeingnewrelationsappear.Inkeepingwiththisview,Peircedevelopedhissystemof“existen-tial”graphsasanewnotationforlogic.(ItisalsonotcoincidentalthatPeircewroteextensivelyontopologyandcartography.)IthasbeenclaimedbyMartinGardner(1982:54–8)thatthegraphswerenotin-tendedasapracticalimprovementoverexistingalgebraicnotations,butasameansoflayingbarethediagrammaticessenceofthoughtandlayingbeforethereader,inPeirce’swords,“amovingpictureofthought.”Butthismaybeafalsedichotomy,andGardnercertainlyexaggerateswhenhepostulatestheneedfor“agiganticeffortofprac-ticeandstudytomasterPeirce’sintricatetechniquetothepointofusefulness....”IthasbeenbrilliantlyshownbyKennethL.Ketner(1981:47–83)thatPeirce’sgraphicalnotationforpropositionallogicisnotsignificantlymoreintricateordifficulttolearnthanthealge-braicnotationnormallyusedinintroductorylogiccourses.AndJohnSowa(1984:149),workinginthefieldofartificialintelligence,hasgonefurtherandarguedthatPeirce’sgraphsforfirst-orderlogicareeasiertoreadandlearn,makepossibleshorterandsimplerproofs,andarethusofgreaterpracticalvaluetolinguisticsandartificialintelligencethanthestandardalgebraicnotation.Finally,JayDavidBolter(1991:85–106)hasrecognizedanddocumentedthefruitfulnessofPeirce’ssemeioticforinterpretinghuman–computerinteractioningeneral,andhypertextinparticular.Bethatasitmay,amovingpictureofthoughtcanalsobeex-hibitedbyamachine,asPeircemadeclearinhis1887articleonthelogicmachinesconstructedbyWilliamStanleyJevonsandAllanMarquand:Thesecretofallreasoningmachinesisverysimple.Itisthatwhateverrelationamongtheobjectsreasonedaboutisdestinedtobethehingeofaratiocination,thatsamegeneralrelationmustbecapableofbeingintroducedbetweencertainpartsofthemachine.(1887:168)Toillustrate,thesyllogism,‘IfAthenB;ifBthenC;therefore,ifAthenC,’canbeembodiedinamachinewherepushingtheleverACambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sSemeioticModeloftheMind255activatesthepistonB,whichinturnrotatesthewheelC,sothatpushingAineffectactivatesC.Machines,then,arecapableofrea-soning,inthesenseofdrawinginferences,butinPeirce’sviewthisisnotapeculiarityoflogicmachinesbutageneralcharacteristicofmachines,includinghereawidevarietyofapparatusnotgenerallythoughtofasmachines:“Accordingly,itisnofigureofspeechtosaythatthealembicsandcucurbitsofthechemistareinstrumentsofthought,orlogicalmachines”(1887:168).WithouthereventuringintospeculationsaboutwhatPeircemighthavesaidaboutmoderndigitalcomputers,weshallobservethathepointedtotwodifferencesbetweenpeopleandthelogicmachinesofhisday.First,machinesare“destituteofalloriginality,ofallinitiative.”Inamachine,Peircestresses,thisisagoodthing;itispreciselythemachine’slackoforiginalitythatmakesitpredictableandhenceuseful;aballoon,forinstance,haslimitedusefulnessbecauseithastoomuchinitiative:“Wenomorewantanoriginalmachinethanahouse-builderwouldwantanoriginaljourneyman,oranAmericanboardofcollegetrusteeswouldhireanoriginalprofessor”(1887:169).ThisispreciselytheviewofmachineslaterechoedbyPopper,whoreferredtothecomputerasa“glorifiedpencil,”aswellasbycom-puterpioneerJ.C.R.Licklider.ItwasalsoexpressedmorethanthirtyyearsbeforePeircewrotetheabove,byCharlesBabbage’sassociateLadyAda,CountessLovelace,reputedlyhistory’sfirstcomputerpro-grammer:“TheAnalyticalEnginehasnopretensionswhatevertooriginateanything.Itcandowhateverweknowhowtoorderittoperform”(Bernstein,1981:57).Second,Peirceobserves,alogicmachineislimitedbyitsdesign:“ithasbeencontrivedtodoacertainthing,anditcandonoth-ingelse”(1887:189).Now,Peirceisnotdenying,butexplicitlyad-mitting,thattherecouldbenondeterministicmachinesconstructedlikeJacquardlooms,incorporatingif–thenloops,andcapableofhan-dlingagreatvarietyofproblems(1887:170).Aproposofthis,Ketner(1988:50–1)hasspeculatedthatPeirce’spioneeringdistinctionbe-tweencorollarialandtheorematicreasoningconstitutesproofoftheexistenceofanondeterministicreasoningmachine–whichKetnercallsa“PeirceMachine”todistinguishitfromthedeterministicTuringmachine.Peirce’sinsightwasthatmathematicsitselfmakesuseof–andhastomakeuseof–reasoningwhichisnotpurelyme-chanical,butwhichessentiallyinvolvestheformationandtryingoutofhypotheses.TowhatextentthisreasoningcanberepresentedCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n256peterskagestadinamachine,andtowhatextentPeircethoughtitcould,arehow-everquestionsbeyondthescopeofthispaper.Whatwehaveseenisthat,inPeirce’sviewmachines,hardandsoft,rangingfrompensandinkstandstomathematicalnotationsandlogicdiagrams,playadecisiveandindispensableroleinourthinking.Andthisrecogni-tion,wehaveseen,isfirmlyrootedinPeirce’sbasicconceptionofthinkingassignaction,whichessentiallyinvolvesphysicalsignve-hicles,whosespecificmaterialqualitiescontributetomakingthemtheparticulartypesofsigntheyare,thusenablingustothinkthetypesofthoughtswedothink.Inthelatterrespect,asIhavearguedelsewhere(1993),Peirce’sse-meioticmodelofmindisrelevantnotonlytoartificialintelligence,butalsotothesocioculturallyfarmoreinfluentialresearchprogramknownas“intelligenceaugmentation,”theprogramwhichbroughtuswordprocessing,thepersonalcomputer,andtheInternet.Inthemanifesto-like“ConceptualFramework”paperfrom1962,whichspelledoutthegoalsandassumptionsoftheprogram,DouglasEn-gelbart(1962:9–11)arguedthatintelligenceshouldnotberegardedaslocatedintheindividualhumanmind,butratherinasystemcom-prisingthehumanbeing,his/herlanguage,artifactsforthinking–pencils,sliderules,computers,etc.–methodsofthought,andfinallytraininginthesaidmethodsandtheuseoftheartifacts.HereferredtothissystemastheHLAM-Tsystem,shortfor“HumanusingLan-guage,Artifacts,andMethod,inwhichheisTrained.”Withinthisframeworkweaugmenthumanintellectnotbymakingindividualhumanbeingssmarter,butbyprovidingthemwithaugmentationmeans–symbolsystems,tools,andmethods–sothattheresult-ingsystemwillbesmarterthantheunaidedhumanbeing.EngelbartwasstronglyinfluencedbyWhorf,butnotatallbyPeirce.YetinEngelbart’s“Framework”paperwehearadistinctechoofPeirce’scomparisonandcontrastbetweenthehumanmindandBabbage’sAnalyticalEngine.Liketheengine,thehumanbeingisalsolimitedbyitsdesign(1887:169):Theunaidedmindisalsolimitedinthisasinotherrespects;butthemindworkingwithapencilandplentyofpaperhasnosuchlimitation.Nolimitation,becausethemind,consideredasasignuser,doesnotresideinsideourheads,butintheexternalfieldofsign-action,whichiscapableofindefiniteaugmentationthroughhumaninventiveness.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nisaaclevi11BewareofSyllogism:StatisticalReasoningandConjecturingAccordingtoPeirce1.probabledeductionPeircewroteextensivelyondeduction,induction,andhypothesisbeginningwiththeHarvardLecturesof1865andLowellLecturesof1866.Theideasthatheexaminedinthoseearlydiscussionswerere-workedovernearlytwodecadesuntilthecomprehensivestatementofhisviewcontainedin“ATheoryofProbableInference”of1883thatwasincludedintheStudiesinLogic,bytheMembersoftheJohnsHopkinsUniversityandisreprintedinW4,408–450.ThisremarkablepaperdevelopedaversionoftheNeyman–Pearsonac-countofconfidenceintervalestimationthatincorporatedthemainelementsoftherationaleofferedforitsadoptionintheearly1930sandpresenteditasanaccountofinductiveinference.Inhisretrospectivereflectiononthequestionofinductionin1902(CP2.102),Peircerevealedsatisfactionwiththeviewsoninductionadvancedin1883andthisattitudeisconfirmedinotherremarksfromthatperiod.However,Peircedidexpressdissatisfactioncon-cerninghisnotionof“HypotheticInference.”AlthoughPeircecalleditHypotheticInferenceorHypothesisfrom1865to1883andlater,in1902,Peircereplacedtheterm“Hypothesis”with“Abduction.”InwhatIsaidabout“HypotheticInference”Iwasanexploreruponuntrod-denground.Icommitted,thoughIhalfcorrected,aslightpositiveerror,whichiseasilysetrightwithoutessentiallyalteringmyposition.Butmycapitalerrorwasanegativeone,innotperceivingthat,accordingtomyownprinciples,thereasoningwithwhichIwastheredealingcouldnotbetherea-soningbywhichweareledtoadoptahypothesis,althoughIallbutstatedasmuch.ButIwastoomuchtakenupinconsideringsyllogisticformsandthedoctrinelogicalextensionandcomprehension,bothofwhichImademore257CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n258isaaclevifundamentalthantheyreallyare.AslongasIheldthatopinion,myconcep-tionsofAbductionnecessarilyconfusedtwodifferentkindsofreasoning.(CP2.102)Peirce’sdescriptionofthesituationseemsreasonablyaccurate,asIhopetoshow.Peirce’swritingsonlogicintheperiodfrom1865to1870proposeanaccountofaformal“unpsychologisticlogic”that,unlikeFrege’slaterdiscussion,appliestoinductiveandhypotheticinferencesaswellastodeductiveinferences.1Theaccountofdeduction,induc-tion,andhypothesisPeirceofferedinthesethreeseriesoflecturesstartswithAristotle’stheoryofthecategoricalsyllogismasimprovedbyPeirce.FollowingAristotle,Peirceunderstoodinductiontoexhibittheformalstructureofatranspositionofoneofthepremisesandcon-clusionofavalidcategoricalsyllogism(the“explainingsyllogism”).InthecaseconsideredbyAristotlewheretheexplainingsyllogismisfigure1,moodAAAinBarbara,theconclusionoftheinductionisthemajorpremiss(the“rule”)ofthe“explaining”syllogismandthecon-clusion(the“result”)ofthatsyllogismreplacesthemajorpremiss.Theminorpremise(the“case”)oftheexplainingsyllogismremainsapremiseoftheinduction.Peirceclaimedthatthisformofargu-mentcharacterizesinductionby“simpleenumeration.”AfavoriteexampleofPeirce’sistheinferencefrominformationthataselectionofcloven-hoofedanimalsthatturnouttobeneat,swine,sheep,anddeerarealsoherbivoroustotheconclusionthatallcloven-hoofedanimalsareherbivorous.Theexplainingsyllogismis:Allcloven-hoofedanimalsareherbivorous(Rule);asampleofneat,swine,sheep,anddeerisselectedfromthecloven-hoofed.(Case).Hence,theneat,swine,sheep,anddeerselectedareherbivorous.(Result).Peirceclearlyunderstood,ofcourse,thattheminorpremiseofasyllogisminBarbaracouldbeauniversalaffirmative(A)proposi-tionsuchas“Allneat,swine,sheep,anddeerarecloven-hoofed”ratherthansingularstatementssuchas“thisneat,thissheep,etc.,arecloven-hoofed.”Buttheminorpremiseofthe“explainingsyllo-gism”isretainedasapremiseoftheinductionwhenRuleandResultofthesyllogismaretransposedtoyieldtheformofaninductivear-gument.AndthisminorpremiseorCaseisintendedtoconveytheinformationthatasampleSofindividualshavebeenselectedfromthepopulationcharacterizedbythemiddletermM.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism259AccordingtoPeirce’sreading,however,theCaseorminorpremiseoftheexplainingsyllogisminBarbarareportsmorethanthateachofagivensetofindividualsisamemberoftheclasscharacterizedbythemiddletermM.Asampleistakenfromtheclassrepresentedbythemiddleterm(inourexample,thecloven-hoofed).Thereportdescribesthespecimenstakenfromthepopulationcharacterizedbythemiddletermbytheminorterm.Inourexample,itisreportedthatthespecimensselectedfromthecloven-hoofedareparticularspecimensofneat,swine,sheep,anddeer.Moreover,inthesyllogismsinBarbaraeligibletobeexplainingsyllogismsforinductions,themethodofselectionfromthemiddletermMmustbesuchthatthespecimensareconsciouslyselectedonlyonthebasisofwhethertheyexhibitthecharacteristicsrepre-sentedbythemiddleterm(W1,264–5).Intheexample,theneat,theswine,thesheep,andthedeerarenotselectedbecausetheybelongtooneofthesefourspeciesbutsolelyonthebasisofwhethertheyarecloven-hoofed.ThisisthewayPeircecharacterizedthemethodofselectionin1865.HeusedessentiallythesamecharacterizationintheLowellLecturesof1866butcalledthemethodofselection“random.”Inthe1869“GroundsofValidityoftheLawsofLogic,”Peircewasassumingthatthelong-runrelativefrequencyorstatisti-calprobabilitywithwhichamemberofasetwillbeselectedgiventhatitisselectedatrandomisthesameforallmembersoftheset(W2,268).2Peirceexplicitlyacknowledgedthatapsychologicalorepistemicconstraintshouldbesatisfiedbythemethodofsampling.Whenwesaythatneatswinesheepanddeerareasampletakenatrandomofcloven-hoofedanimals,wedonotmeantosaythatthechoicedependeduponnootherconditionthanthatallshouldbecloven-hoofed;wecannotknowthatandthepresumptionistheotherwaysincethereisacertainlimitationofthatclassindicatedbyourhavingtakensofewinstances.Whatwemean,then,insayingthatneatswinesheepanddeeraretakenatrandomfromamongthecloven-hoofedanimals,isthatbeingcloven-hoofedwastheonlyconditionthatconsciouslyguidedusintheselectionoftheseanimals.(W1,433).Peirceadmittedthroughoutthathecouldnotgiveapurelyformaland,hence,logical(inthesenseofunpsychologicallogic)charac-terizationofthestrengthofinductivearguments.Inthisrespect,theydifferedfundamentallyfromdeductivearguments.ThereisnoCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n260isaaclevinotionofinductiveconsequencetocorrespondtodeductiveconse-quencethatbelongsproperlytoformallogic.Butinferences,canbeclassifiedasinductiveonthebasisofformalconsiderationsalone.Inthelate1860s,hisideawastoappealtothetransposi-tionsofpremisesandconclusionsofcategoricalsyllogismstopro-videthecharacterization.3ThisformalclassificationconstitutedthebasisforPeirce’sclaimthatinductionandhypothesiscouldbeobjectsofstudyundertherubricofanunpsychologisticformallogic.Peirceabhoredvacuumsinlogicalspace.Thereisobviouslyan-otherkindoftranspositionoftheexplainingsyllogisminBarbarawheretheminorpremiseorcaseandconclusionorresultofthatsyl-logismreplaceeachother.Peirceproposedtothinkoftheresultingargumentasinstantiatinghypotheticinference.Hypotheticinferences,likeinductiveinferences,canbeclassi-fiedbypurelyformaland,hence“unpsychological”criteria.Butthestrengthofsuchinferences,likethestrengthofinductiveinferences,takesintoaccountconsiderationsthatarenotpurelyformal.OneoftheproblemsinternaltoPeirce’sapproachisthatatleastfromthetimeoftheLowellLecturesof1866,hewishedtoregardin-ferencefromsamplefrequenciestopopulationfrequenciesor,moregenerally,tostatisticalhypothesesasparadigmaticofinductivein-ference.Thedifficultyisthatstatisticalclaimssuchas“Mostcloven-hoofedanimalsareherbivorous”or“90%ofcloven-hoofedanimalsareherbivorous”arenotcategoricalpropositionsandaredifficulttoconstrueascategoricalpropositionsforthepurposeofintegra-tionintosyllogisticargument.Butiftheycouldnotbesointegrated,theformsofinductiveinferenceswhoseconclusionsarestatisticalclaimscouldnotbeobtainedbytransposingthemajorpremisesandconclusionsofcategoricalsyllogisms.In1883,Peirceclearlyandexplicitlyrecognizedthatstatisticalclaimsthathetookoftentobemajorpremisesoftheexplainingsyl-logismsforinductionintheearlierpapersarenotcategoricalpropo-sitions.Heemphasizedthattheanalogybetweensyllogismsoftheform“AllMareP,SisM,thereforeSisP”andprobabledeductionsoftheform“TheproportionρoftheMsarePs;SisanM;thereforeitfollows,withprobabilityρ,thatSisP”andprobabledeductioniscer-tainlygenuineandimportant,therearefoursignificantdifferencesbetweenthesemodesofinference(W4,409–12)4:CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism261(1)Probabledeductionisrelatedtosyllogismasthequantitativebranchistothequalitativebranch.(2)Allthatisrequisiteforordinarysyllogismisthatthethreetermsofthesyllogismstandinsometransitiverelationsuchasinclusion,betterthan,etc.Probabledeductionpresup-posesquantitativeratios.“Forthattheremustbecountingandconsequentlyunitsmustexist,preservingtheiridentityandvariouslygroupedtogether.”(3)Insyllogism,theconclusionfollowsfromthepremisesac-cordingtoaformalrelationoflogicalconsequence.Inprob-abledeductionwheretheruleorsurrogatemajorpremiseassertsthattheproportionρoftheMsarePs,theinforma-tionthatSisanMdoesnotrendertheconclusionthatSisPevenprobable.ItisnecessarythatSbeselectedatran-domfromtheMs.AccordingtoPeirce,therequirementofrandomnesstakesintoaccountvarioussubjectivecircum-stancessuchasthemannerinwhichthepremisehasbeenobtained,theabsenceofcountervailingconsiderations;“inshort,goodfaithandhonestyareessentialtogoodlogicinprobablereasoning.”“InchoosingtheinstanceS,thegen-eralintentionshouldbetoselectanM,butbeyondthatthereshouldbenopreference;andtheactofchoiceshouldbesuchthatifitwererepeatedmanytimeswiththesameintention,theresultwouldbethatamongthetotalityofse-lectionsthedifferentsortsofMswouldoccurwiththesamerelativefrequenciesasinexperiencesinwhichvolitiondoesnotintermeddleatall.”Inotherwords,randomselectionis(a)selectionwithoutdeliberateorconsciousbiasand(b)selectionfromtheMsthatyieldsPswithprobability(accord-ingtothefrequencyinterpretationfavoredbyPeirce)equaltotheproportionρintheMs.Randomselection,however,isnotenough.“Acardbeingdrawnatrandomfromapicquetpack,thechanceisone-eighththatitisanace,ifwehavenootherknowledgeofit.Butafterwehavelookedatthecard,wecannolongerreasonthatway.”(4)Theconclusionofasyllogismisanecessaryconsequence.Theconclusionofaprobabledeductionisonlyprobable.Thatistosay,theruleormajorpremiseofsyllogismissuchthattheinferencefromCase(minorpremise)toResultCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n262isaaclevi(conclusion)istruth-preservinginallapplications.Inproba-blededuction,theinferencepreservestruthwitharelativefrequencyρ.Peircethenassertsitis“useful”thatweshouldhaveastrongerfeelingofconfidencethehigherthevalueofρ.AlthoughPeircewouldnevercallthisdegreeofcon-fidenceasubjectiveorbeliefprobability,itisclearinallhiswritingthatithasalltheearmarksofwhatmanycon-temporarieswouldcallasubjectiveprobabilityincludingthedispositiontotakerisks.Forthisreason,Iaminclinedtore-construePeirce’sviewasonethatadmitsthatinquirersmayassignsubjectivedegreesofprobabilisticbelieftohypothesesprovidedthosedegreesofbeliefcanbegroundedorjustifiedbyknowledgeofobjective,statistical,orfrequencyprobabil-ity.Hisobjectionistotakingnumericallydeterminatejudg-mentsofsubjectiveprobabilityseriouslyinscientificinquirywhentheylacksuchgrounding.PeirceextendedhisaccountofprobabledeductiontoinferencesfromsamplesofnMstotheproportionsofPsamongthenMs.Hefirstinvokedthebinomialdistributiontodeterminetheprobability(long-runrelativefrequency)ofn-foldsamplesexhibitinganygivenrelativefrequencyofPsamongnMs.TheWeakLawofLargeNumbersimpliesthatasnincreases,theprobabilityofobtainingarelativefrequencywhoseabsolutediffer-encefromρislessthansomespecificvaluedconvergesto1.Peircethenusedthenormalapproximationofthebinomialdistribution(supportedbytheCentralLimitTheorem)tosustainthefollowingargumenthecalled“StatisticalDeduction”:TheproportionroftheMsarePs(Rule).S,S,S,etc.areanumerousset,takenatran-domfromamongtheMs.Hence,probablyandapproximately,theproportionroftheSsarePs.Inkeepingwithhisaccountofprobablededuction,the“subjec-tive”conditionthatthesampleistakenatrandomshouldnotbepartoftheformofthisargument.Peirce’sconcerntodistilloutanunpsychologisticcomponentofthereasoninginvolvedprecludedthis.Similarly“probably”qualifiesthe“mode”oftheconclusionandisnotpartofitscontent.“Approximately”hereisintendedtosuggestthattheconclusionisnotthattherelativefrequencyisex-actlyrbutthatitfallsinsomemoreorlessvaguelyspecifiedinter-valaroundr.PeirceusedthenormalapproximationtothebinomialCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism263distributiontoillustratehowboundscanberoughlyspecifiedfortheintervaldependingonthesamplesizen.TheHarvardLecturesof1865donotexplicitlymentiondirectinferencefromstatisticalhypothesestotheoutcomeofsamplingexemplifiedbyprobableandbystatisticaldeductionortheinverseinductiveinferencefromtheoutcomeofsamplingtoastatisticalhypotheses.TheLowellLecturesintroduceinduction(inLectureIII)asinferenceofjustthiskind.However,indiscussionofinductionandhypothesisastranspositionsofPremiseandConclusionofSyl-logismsthereisnodiscussionofhowinductiontostatisticalhy-pothesisistoberepresentedassuchatransposition.YetPeircedidrecognizethemainpointsofdifferencebetweensyllogismandsta-tisticaldeductionmentionedaboveinthe1866Lectures.ApparentlyPeircewasthinkingin1866ofextendinghis1865accountofinduc-tion(andhypothesis)tostatisticalcasesbutdidnothaveanswersofthesortheofferslateronconcerningtherelationsbetweenproposi-tionsoftheform“r%ofMsarePs”and“AllMsarePs”andgrad-uallythoughtthesemattersthroughinthenextdecade.5Thus,thelecturesandpapersstartingin1866seemtobeattemptstoadjustPeirce’sengagementwiththeideathatinductionandhypothesisare,formallyspeaking,transpositionsofpremisesandconclusionsofcat-egoricalsyllogismstotranspositionsofprobableandstatisticalde-ductions.The1878and1883papersrepresenttheculminationofthiseffort.62.transposingpremisesandconclusionsofexplainingsyllogismsandapogogicinversionInthePriorAnalytics,Aristotlecharacterizedinductionastheprod-uctoftransposingthemajorpremiseandconclusionofacategoricalsyllogisminBarbara.TheminorpremiseisretainedandisspelledoutasassertingthattheindividualscoveredbytheminortermSaremembersofthesetcharacterizedbythemiddletermM.ButAris-totlealsoinsistedonpremisingtheconverseoftheminorpremise,positingthattheSsaretheonlyMs.Likemostcommentators,PeircetookthistomeanthatAristotle’sinductionisinductionbycompleteenumerationandisthus,strictlyspeaking,aspeciesofnecessaryordeductiveinference(W1,263).Peircerejectedthisattempttoreduceinductiontodeduction.Nei-therinductionnorhypothesisisreducibletonecessaryinference.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n264isaacleviOnecouldnotcometoknowthatthesampleexhauststhepopula-tionwithoutsomesortofinduction.TheinformationconveyedbythesampleofMsobserved–towit,theSs–cannotwarrantsuchajudgment.ThismuchPeirceexplicitlystatedin1865.YetPeircedidcon-strueinductionstobetranspositionsofthemajorpremiseandcon-clusionofacategoricalsyllogism.In1883,heexplicitlyabandonedthenotionthataninductiveinferenceisatranspositionofthemajorpremiseandconclusionofacategoricalsyllogism.ItisratherthetranspositionofthemajorpremiseorRuleofastatisticaldeductionandtheconclusionorResult.Thestatisticaldeductionhasthefollowingform:StatisticalDeductionRule:TheproportionroftheM’sareP’s.Case:s1,s2,...,snbelongtoanumeroussetSofobjectstakenatrandomfromamongtheM’s.Result:TheproportionroftheS’sareP’s(probablyandapproxi-mately).InductionisobtainedbytransposingtheRuleandResult.InductionResult:TheproportionroftheS’sareP’s.Case:s1,s2,...,snbelongtoanumeroussetSofobjectstakenatrandomfromamongtheM’s.Rule:TheproportionroftheM’sareP’s(probablyandapproxi-mately).PeircejustifiedthetranspositionthatyieldsInductionfromSta-tisticalDeductionbyarguingthattheprincipleofstatisticaldeduc-tionisthattwoproportions,thatofthePsamongtheMsandthatofthePsamongtheSs,areapproximatelyequal.Thatistosay,aslongasthemethodsofsamplingandtheavailableevidencewarranttheapplicabilityofthecalculusofprobability,theweaklawoflargenumbersandthecentrallimittheoreminsurethatthisprobableandapproximateequalityholds.Peircethenwrote:Ifthen,thisprinciple[ofstatisticalinduction]justifiesourinferringthevalueofthesecondproportionfromtheknownvalueofthefirst,itequallyjustifiesourinferringthefirstfromthatofthesecond,ifthefirstisunknownbutthesecondhasbeenobserved(W4,416)Thephrase“equallyjustifies”callsforcloserscrutiny,asPeircewellunderstood.Inthepassagecitedfrom1883,heclaimedthatCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism265hecouldspecifyconditionsunderwhichthisclaimcouldbemadegood.IntheHarvardLecturesof1865andLowellLecturesof1866,headoptedadifferentrationaleforclaimingthatthereisequaljustifi-cationforinferringpopulationfrequencyfromsamplefrequencyasthereisforinferringsamplefrequencyfrompopulationfrequency.Hethenbecamesilentontheissueuntil1878whenhecameoutwiththenewrationalization–oneheexpressedwithgreatclarityin1883.Butinspiteofanimportantchangeinhisviewsofinductionbetweenthelate1860sandthelate1870s,throughouttheentireperiodPeircethoughthecouldachievelegitimatelywhatAristotleattemptedtoachieveillegitimatelybytheillicitconversionoftheminorpremiseofacategoricalsyllogism.IbelievethatPeircealreadyendorsedmanyelementsofhisviewin1878and1883inLectureIVofthe1866LowellLectureswhencrit-icizingJ.S.Mill’scontentionthatinductiveinferenceis“grounded”onaPrincipleoftheUniformityofNature(W1,420–1).Constru-inghimthiswayprovidesawaytounderstandtheargumentoftheLowellLectureseventhoughthoselecturesoftenfailtobeexplicitaboutthevarioustransitionsintheargument.7Byusingcontrapositionandconversion,onecantransformvalidfirstfiguresyllogismsintovalidsecondandthirdfiguresyllogisms.ToobtainthesecondfigurethenegationoftheCaseofasyllogisminBarbarabecomestheconclusionandthenegationoftheResultapremise.Toobtainathirdfigure,thenegationoftheRulebe-comestheconclusionandthenegationoftheResultapremise.Peirceseemedtohavebeeninterestedinthese“apogogicinversions”be-causetheylooksimilarinformtohypotheticandinductiveinfer-ence.Butifthesyllogismsarepropercategoricalsyllogismsthisisnotstrictlyspeakingso.In1883,PeircesuggestedthatiftherulesaysthattheproportionofPsamongtheMsisρ,thenegationoftheruleisthatthesetofPsamongtheMsisoneoftherealvaluesbetween0and1otherthanρ.Usingthisandrelyingupontheweaklawoflargenumbers,Peirceshowedthattheapogogicinversionofastatis-ticaldeductionisaninduction.However,thestatisticaldeductionofwhichitisanapogogicinversionisnotthe“explaining”statisticaldeductionwhoseruleandresultaretransposedtoobtaintheinduc-tion.Andinthecourseofmakingthecaseforthisview,Peircewrotethatitisnecessaryifaninductionistohaveanyvalidityatallthattheexplanatorysyllogismshouldbeavalidstatisticaldeduction.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n266isaaclevi“Itsconclusionmustnotmerelyfollowfromthepremisesbutfol-lowfromthemupontheprincipleofprobability.Theinversionofordinarysyllogismdoesnotgiverisetoaninductionorhypothesis.”Thatisbecausetheinversionofanordinarysyllogismreliesonanillegitimateconversion(W4,424–7).Peirceexpressedasimilarviewin“Deduction,InductionandHy-pothesis”inthefamous1878seriespublishedPopularScientificMonthly(W3,328).In“OntheNaturalClassificationofArguments”of1867(W2,48)Peircepointedto“aresemblancebetweenthetranspositionofpropo-sitionsbywhichtheformsofprobableinferencearederivedandthecontrapositionbywhichtheindirectfiguresarederived.”Peircedoesnot,however,elaboratethepoint.Stillthisremarkdoesappeartobeanallusiontotheideacontainedinthe1878and1883papers.Con-fidenceinthisinterpretationisrenderedinsecurebythefactthatnowhereinthe1867essaydidPeircerequireexplicitlythattheex-plainingsyllogismforaninductionhaveastatisticalpremiseastheRule.Thatrequirementispresentinboth1878and1883.Nonethe-less,the1867essaydoessuggestthepossibilitythatinduction(andhypothesis)couldbeconstruedas“apogogicinversions”ofstatisticaldeductions.TheLowellLectureSeriesof1866doesnotevenrecognizetheanalogybetweentransposingtheRuleandtheConclusionofasyllo-gisminBarbaraandobtainingathirdfiguresyllogismbycontrapo-sitionfromthesamesyllogism.Yetthe1866lecturesdorecognizeinverseinferencefromsampletopopulationfrequenciesasparadig-maticofinductionandrecognizetheimportanceoftheprobableandapproximatesimilaritybetweensampleandpopulationinthisre-spect.Thus,thereisaprogressionfromthevirtualabsenceofstatis-ticalconsiderationsinPeirce’saccountofinduction(andhypothesis)in1865totheirreduciblystatisticalaccountofinductionofferedin1878and1883.3.frequentismandinsufficientreasonEvenin1865,PeirceinsistedthattheCaseorminorpremiseofanexplainingsyllogismreportsthattheSs(minortermoftheexplainingsyllogism)areaselectionfromtheclasscharacterizedbythemiddletermaccordingtoamethodthattookintoaccountasaprincipleofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism267selectionwhethertheitemisanMornotwithoutregardtootherconsiderations.Thismethodofselectioniscalledrandomselectionin1866andthereafter.AgainPeirceismoreexplicitinthelaterpapersthatselectinganitematrandomfromapopulationlicensestheassumptionthatthestatisticalprobabilityofselectingoneitemfromthepopulationisthesameasselectinganyotheritem.Thismethodofrandomselectionistheguarantorofthevalidityoftheprobabledeductiontotheoutcomeofasingleselectionandofthestatisticaldeductiontotheprobableandapproximateequalityofthesamplefrequencywiththepopulationfrequencyinastatisticaldeduction.IhavesuggestedthatPeirceheldthatthereareoccasionswhenonemaylegitimatelyjudgedegreesofbeliefconstruedasdisposi-tionstotakerisksofcertainkinds.Whentheinquirercanjustifytheresultofastatisticaldeductionbyappealtothestatisticalruleandthepremisethatthemethodofselectionisarandomone,thenceterisparibus,accordingadegreeofbeliefthattheresultwillbethetrueequaltothestatisticalprobabilityofsucharesulthappeningduetosuchselectionislegitimate.TheproblemthatPeircesoughttoaddressishowtoshowthattheinverseinferenceobtainedbytrans-posingRuleandResultthatisaninductionentitlesusequallytoinferthatapproximatelyr%ofthepopulationofMsarePsfromthedatathatinalargerandomsampleofMs,r%arePs.Successinthisrequiredthatnodegreeofbeliefbeassignedtotheconclusionunlessitcouldbederivedviaequationwithastatisticalprobabilitydeducedaccordingtoaproperstatisticaldeduction.Peircestoodopposedtooneapproachthatmanystudentsofinduc-tionincludingBayes,Laplace,andDeMorganhadfavored.Peirceen-thusiasticallyendorsedGeorgeBoole’scriticismoftheuseofinsuffi-cientreasonargumentstoformpriorsinordertoobtainposteriorsasearlyas1865andevenmoreexplicitlyin1866.Iftheproportionofredballsinthebagisunknown,itisstillthecasethatbystatisticalde-ductionbasedontheweaklawoflargenumbersandthecentrallimittheoremonecouldclaimthatnomatterwhatthepercentageρofredballsintheurnisthetrueone,therelativefrequencyofredballsselectedatrandomiswithgreatprobability(isalmostcertainly)ap-proximatelyequaltoρ.Ifviatheprincipleofinsufficientreason,oneassignsequalpriorprobabilitytoeachconjectureregardingthetruevalueofthepopulationfrequencyasBayes,Laplace,andDeMorganCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n268isaaclevirecommended,then,accordingtoBayes’Theorem,theposteriorprobabilityconditionalonevidencethattheproportionofredballsinthelargesampleselectedfromtheurnisrisgoingtobeashighasapproximatelyroftheballsintheurnfromwhichthesampleistakenaregoingtobered.AccordingtoPeirce,theprobabilityofahypothesisconditionalonevidenceisthelong-runrelativefrequencyofobtainingatrueconclu-sionfromtruedataininferencesofthatkind.TheinverseinferenceappealingtoinsufficientreasonwouldbeacceptabletoPeirceiftheequalpriorprobabilitydistributionoverhypothesesconcerningtheproportionofredballsinthebagcouldbeobtainedonthegroundthatthebagfromwhichtheballwasselectedwasinturnselectedatrandomfromabagofbagsinwhichallpossibleproportionswerepresentinequalproportions.Withoutsomesuchassumption,theappealtouniformpriorprobabilityandtoBayes’Theoremcannotyieldaposteriorconstruedinfrequentistterms.In1865and1866,PeirceexpressedadmirationforBoole’salgebraicwayofexpressingtheindeterminacyoftheposteriorprobabilityderivedfromBayes’Theoremwhennopriorisgiven(W1,238–9,404–5).Healsosharplycriticizedtheprincipleofinsufficientreasonasameansforreliev-ingtheindeterminacy(W1,401–3).Fromtheverybeginningofhiscareer,PeirceregisteredunwaveringoppositiontousinginsufficientreasonandBayes’TheoremtorationalizeinductionasLaplacehaddone.8In1878,Peircefamouslywrote:TherelativeprobabilityofthisorthatarrangementofNatureissomethingwhichweshouldhavearighttotalkaboutifuniverseswereasplentyasblackberries,ifwecouldputaquantityoftheminabag,shakethemup,drawoutasampleandexaminethemtoseewhatproportionofthemhadonearrangementandwhatproportionanother.Buteveninthatcase,ahigheruniversewouldcontainus,inregardtowhosearrangementstheconceptionofprobabilitycouldhavenoapplicability(W3,300–01).Peircereinforcedthisobjectiontotheuseofnumericallydetermi-natepriorcredalorbeliefprobabilitieswithoutgroundinginknowl-edgeofstatisticalprobabilitywithamorespecificobjectiontoin-sufficientreason.In1866,hecomplainedthatifoneisignorantwithrespecttothetruthofhand∼h,insufficientreasonassignshaproba-bilityof1/.However,onemaybethusignorantandalsoignorantcon-2cerningthetruthofh,h*,andh**,whereh*∨h**=∼h.InsufficientCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism269reasonrecommendsassigninghapriorprobabilityof1/3.Thisver-dictisinconsistentwiththeotherone.Butthereisnoprincipledwaytoresolvetheconflict(W1,402).4.uniformityofnaturePeircethoughtjudgmentsofnumericallydeterminatecredalproba-bilityoughttobegroundedinfullbeliefthatsomestatisticalproba-bilityistrue.Millmaintainedthatinductionsoughttobegroundedsomehowinappealtoamajorpremisethat,iftrue,wouldconverttheinductionintoacategoricalsyllogism.Ultimatelytherewouldbeahierarchyofsyllogismswherethefundamentalmajorpremisewouldbesomeprincipleofuniformityofnature.In1865,Peircecomplainedthatthisdemandforconvertinganinductionintoasyl-logismthreatenstoyieldaninfiniteregressofuniformityprinciples.In1866,Peirceaddedanadditionalmoreinterestingclusterofcomplaints.Thenotionofuniformityisunclear.AccordingtooneofMill’sformulations,theuniverse“issoconstituted,thatwhateveristrueinonecase,istrueinallcasesofacertaindescription”(SystemofLogic,v.1,Bk.3,Ch.3,§1).Peirceforcefullycalledintoquestiontheempiricalwarrantofthisclaim.Everystudentofphysicsknowsthatalawwhichisexactlyconformedtoinnaturewithoutinterferencefromotherlawsisalmostifnotquiteunknown.Everylawthatisdiscoveredthereforeisfoundafterafewyearsnottobeexact.Whatdowesay?Whythatitistrueinallcasesofacertaindescription;butwehaven’tfoundofwhatdescription....Thereisstillanothersenseinwhichwemightspeakoftheuniformityofnature.Ifweselectagoodmanyobjectsontheprinciplethattheyshallbelongtoacertainclassandthenfindthattheyallhavesomecommoncharacter,prettymuchthewholeclasswillgenerallybefoundtohavethatcharacter.Orifwetakeagoodmanyofthecharactersofathingatrandomandafterwardsfindathingwhichhasallthesecharacters,weshallgenerallyfindthatthesecondthingisprettynearthesameasthefirst.Itseemstomethatitisthispairoffactsratherthananyotherswhichareproperlyexpressedbysayingthatnatureisuniform.Weshallseethatitistheywhicharetheleadingprinciplesofscientificinference.(W1,420.)Peircedistinguishedbetweenformallawsthatwouldholdnomat-terwhatthe“stateofthings”mightbeandmateriallawsthatdonot.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n270isaacleviHearguedthattheuniformityprincipleheendorsedwillholdaslongasthereareanylawswhatsoever.Sincetheexistenceofsomelawsisapreconditionofknowledge,theuniformityhefavorsisaformalconditionofallknowledge.PeircethenwentontostatethatthepredictionoftheapproximateequalityoftheproportionofPsinthetotalpopulationofMsandtheproportionfoundinarandomsampleofMscorrespondstoaspecialuniformityintheworld.“Itisthetypeofalluniformityandallinduction.Statistics;thatisinduction”(W1,423).PeirceclearlythoughtthatadoptingthisnotionofUniformityofNaturedidnotadequatelyaccountforinductive–i.e.,statistical–inference,butmerelyrebaptisedthechallengetodoso.Thecrucialpointisthatbytheendofthisdiscussion,Peircehadexplicitlyacknowledgedevenin1866thattheconclusionofanin-ductiondidnothavetobeuniversalorapproximatelyuniversal.AnypropositionstatingthattheproportionofMsthatarePsisapprox-imatelyrcanbetheconclusionofaninduction.Incaseslikeesti-mationofastatisticalparametersuchastheproportionofredballsinthebag,allthatisrequiredisthattheproportionbeoneofthepossiblerelativefrequencies.9ThisisanimportantdeparturefromMill’sunderstandingofuniformity.5.nuisanceinformationTheprogressionintheemphasisPeirceplacedonstatisticalconsid-erationsinhisdiscussionofinductionin1878and1883pivotsonhisappealtotheWeakLawofLargeNumbersandtheCentralLimitTheoreminthecalculusofprobabilitiestoderivethevalidityofsta-tisticaldeductions.GiventherulethatstatesthattheproportionrofMsarePsandtheinformationthats1,s2,...,snareanumerousrandomsamplefromtheMs,statisticaldeductionwarrantsthecon-clusionthatprobablyandapproximatelytheproportionrofSsarePs.Thatistosay,probablythesamplewillberepresentativeofthepop-ulationwhentheproportionrofSsarePs.Moreover,forlargen,thisclaimwillholdforallproportionsbetween0and1inclusive.Conse-quently,theinquirerwhoknowsnothingabouttheoutcomeoftherandomsampletobetakenother,perhaps,thanthatitwillbetakencanusestatisticaldeductionandthecalculusofprobabilitiestoCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism271concludethattheoutcomeofrandomsamplingwillberepresen-tativewithveryhighprobabilityconcerningtheproportionofPsamongtheMs.Hence,theinquirerisentitledbeforesamplingtobenearlycertainthatheorshewillobtainsucharepresentativesample.OncethesampleistakenandtheproportionofPsamongtheSsascertained,theinquirermustreassesshissubjectiveprobabilitiesinthelightofthenewinformationavailable.Suppose,forexample,thataninquirerXintendstosampleatran-domfromcloven-hoofedanimals,takenoteofthepercentageofher-bivorousanimalsamongthem,andthenestimatethatthepercent-ageofcloven-hoofedanimalsthatareherbivorousisapproximatelyequaltothepercentageinthesample.Priortosampling,statisticaldeductionwillwarrantalmostcertainlythattheapproximateesti-mateofthepopulationfrequencywillbetrue.Theinquirerengagesinthesamplingandobtainsthefollowinginformation:(1)SistherandomsampleofMs.Specimensofneat,swine,sheep,deerhavebeenselected.(2)ThepercentageofPsamongtheSsisr.Cantheinquirerconcludeviastatisticaldeductionthatitisal-mostcertainthattheproportionofMsthatarePsisapproximatelyrasheorshecouldpriortoobtainingtheinformationspecifiedunder(1)and(2)?IbelievethatPeircegaveoneanswertothisquestionin1865and1866andanotheranswerin1878and1883.Peirceclearlyrecognized,asisevidentfromhisreviewofJohnVenn’sTheLogicofChancein1867,thatthedesiredinductivecon-clusioncannotbederivedviastatisticaldeductionfromtheinfor-mationcontainedin(1)and(2).HehadalreadyrejectedusingBayes’Theoremtosecurehighprobabilityforthatconclusion.Sohein-sistedthatnogradeofprobabilitygroundedinfrequenciescouldbeassignedtotheconclusion.Whatistheproblemwithcontinuingtojudgeithighlyprobablethatthesampleisrepresentative?Takefirsttheinformationaboutthesamplefrequencycontainedin(2).Giventheinformationthatthesamplefrequencyisr,theprobabilitythatthesampleisrepre-sentativeis1givenapopulationfrequencynearrandis0otherwise.WecannolongerarguethatitisalmostcertainthatthesampleisCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n272isaaclevirepresentativenomatterwhattheproportionofPsinthepopulationofMsis.Considernowtheinformationgivenin(1).Eveniftheinforma-tionaboutthesamplefrequencywereunknowntoX,knowingthattherandomsamplecontainedspecimensofneat,sheep,swine,deermight,asfarasXknew,berelevanttobeingherbivorousinthesensethatthefrequencyofbeingherbivorousamongthesekindsofanimalsisnottypicaloftheentirecloven-hoofedpopulation.Todoaproperstatisticaldeduction,Xwouldneedtohaveestablishedthatthein-formationthatthespecimensareofneat,sheep,swine,anddeerisirrelevantonthebasisofinformationaboutfrequencies.(ThisistheburdenofthecriticismofVenninPeirce’s1867review.)Typicallythiscouldnotbedoneeither.6.inductionandsemeioticPeirce’sinitialproposalforaddressingthesedifficultiescalledforanappealtoatheoryofsigns.The1865and1866papers,aswellastheseriesofpapersgiventotheAmericanAcademyofArtsandSciencesin1867,presentedthefirstversionsofwhatlaterbecamePeirce’ssemeiotic.Heintroducedtheseideasintothesepapers,Icontend,inordertoprovideabasisforconfrontingtheproblemofnuisanceinformationinthereferenceclassthatfacedhisaccountofinduction.SpacedoesnotpermitanextendeddiscussionofPeirce’searlytheoryofrepresentations.Butabriefsummaryisinorder.Therearethreetypesofrepresentations:signs(laterindices),copies(latericons),andsymbols.Neithersignsnorcopiesconveyinformation.Signshavedenotation,extension,orbreadthbutnoconnotation,comprehension,ordepth.Copieshavedepthbutnobreadth.Symbolshavedepthandbreadthandcarryinformation.Moreimportant,how-ever,isPeirce’sinsistencethatthedepthandbreadthofasymbolaredeterminedbytheinformationorbeliefstateoftheinquirer.Ac-cordingtoaninquirerX,the“informedbreadth”ofatermSisthesetofobjectsofwhichSmightbetrueasfarasXknows.Thein-formeddepthofSisthesetoftraitsattributableinthelightofwhatXknowstoanobjectofwhichSispredicated.Iftheinquirer’sstateofinformationisheldconstant,termMhasmoreinformedbreadththantermSifandonlyifithaslessinformeddepththanS.BothCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism273informedbreadthandinformeddepthcanincrease,however,iftheinquirer’sstateofinformationincreases.ObservethattheclaimthatSisarandomsamplefromMisthemi-norpremiseofthe“explainingsyllogism.”Thatistheinformationconveyedby(2).ThattheproportionofPsamongtheSsisrcon-veyedby(1)istheconclusionoftheexplainingsyllogismor,moreaccuratelyaccordingtothelaterwork,theprobablededuction.TheinductiveconclusionisobtainedbysubstitutingMforSinpremise(1).Tocometobelievethatapproximatelyr%ofMsarePsinvolvesanincreaseininformation.Peircecharacterizesthissitua-tioninthefollowingmanner.TheextensionorbreadthofMisatleastasgreatasthatofSbothrelativetotheinformationbeforeandaftertheshift(typicallyitisknowntobegreater).Thesubstitution,asaconsequence,incursariskoferrorthatmayactasadeterrenttotheinquirerX.Toovercomethisrisk,theinquirerneedsanincentive.ThatisprovidedbythefactthatthereplacementofSbyMin(1)increasestheinformeddepthofMduetotheadditionofthepredicatePtoitsinformeddepth.Andthisincreaseyieldsanincreaseininformationbyadoptingtheconclusionoftheinduction.Whethertheincreaseissufficienttojustifytheinductiondependsonthe“preferences”oftheinquirer.Itreflectstheextenttowhichheorsheispreparedtoriskerrortoobtainnewinformation.Thereisonesettinginwhichthetrade-offautomaticallyfavorsmakingthesubstitutionofMforS.ThatisinthecasewhereSisalistofthespecimens(theneat,swine,sheep,anddeer)selectedatrandomfromtheMs.AccordingtoPeirce,evenifthedescriptionsoftheindividualsselectedaresymbols,thedisjunctivetermS(isaneatoraswineorasheeporadeer)isnot,accordingtoPeirce,asymbolbecauseitdoesnotcharacterizeanycommonfeatureofthealternativespeciesand,hence,carriesnoinformation.Peircethenarguedthatthereisan“absolute”preferencefor“r%ofMsarePs”(whereMisasymbol)over“r%ofSsarePs.”AccordingtoPeirce’sview,aninquirerisnotentitledtoassessriskintermsofdegreesofbeliefunlessthedegreesofbeliefcanbegroundedinknowledgeofprobabilitiesasfrequencyinthelongrun.Peircedeniedthatthiswasfeasibleeveninthe1865HarvardLectures.Hence,theacknowledgedtensionbetweenriskoferrorandvalueofinformationgainedisnotrepresentableasaweightedCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n274isaacleviaverageofprobabilityoferrorandthevalueofinformation.Intheabsenceofaquantitativerepresentation,Peirceunderstoodthemat-tertobeoneofacomparative“preference.”Anditispossibleforoneinquirertorefusetodrawtheinductiveinferencewhereastheotherendorsesit.Inthecaseofinductionbysimpleenumeration,however,thepres-sureismuchstronger–orsoinsistedPeirce.Thenuisanceinforma-tioncontainedin(1)isnoinformationatall.ThedisjunctivesymbolShasnoinformeddepth.TheclaimthatSisasampleselectedatrandomfromtheMsisasinformativeastheclaimthatarandomsamplewasselectedfromtheMs.Hence,theclaimthatSisPisnomoreinformativethantheaddendumthatthesamplerandomlyselectedisincludedinthePs.PeirceseemedtothinkthataslongasthetermMhadmoreinformeddepththanS(accordingtothestateofinformationbeforereachingtheinductiveconclusion),theinquireroughttohaveamarkedpreferencefor“AllMareP”thenfor“AllSareP”and,hence,shouldtaketheinductiveleap.Ineffect,theinquirerisentitledtoignorethenuisanceinformationcontainedinSbecausethereisnone.Evenifoneendorsesthedistinctionbetweenrealpropertiesandmerepredicatesthatsomanyauthorsadvocatethesedaysandde-niesthatdisjunctionsofrealpropertiesarerealproperties,thestrongclaimsPeircemakesaboutabsolutepreferencedonotappeartofol-lowfromtheconcession.Peirceabandonedthisproposalsilently,butIbelievefairlyquickly,inthelate1860sandearly1870s.Heexplicitlyrejecteditattheturnofthetwentiethcentury.Andby1878hehadofferedanalternativeapproachtoreplacethismethodoffinessingthenuisanceinformation.InspiteofitsbriefdurationasPeirce’smethodofdissolvingtheproblemofnuisanceinformation,thisideaisimportanttotheunderstandingofPeirce’sthoughtfortworeasons:(a)itestablishesthatPeirce’ssemeioticwasinitiallyusedasameansofdissolvingsomedifficultiesinPeirce’saccountofinductionandhypothesisand(b)itcontributestoourunderstandingofthereasonswhyPeirceeventuallyabandonedthenotionofhypotheticinferenceasdistinctfrominductionandreplaceditbyaratherdifferentideaatthebegin-ningofthetwentiethcentury.Inaddition,itisworthnoticingthatin1865and1866,PeircetoyedwiththeideaofrationalizingampliativereasoningasaquesttoseekCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism275truthwhileshunningerrorinacontextwhereseekingtruthandavoidingerrormeansseekingmoreinformationwhileavoidingerrorandwherethisobjectiveisseenasacharacterizationofcommonfeaturesofproximategoalsofspecificinquiries.Beforeturningtothereasonsthatledtoreplacinghypotheticwithabductiveinference,itistimetoexplainhowPeircecametothinkofinductionin1878and1883.7.predesignationandself-correctionAninquirerXintendstosampleatrandomfromcloven-hoofedan-imals,takenoteofthepercentageofherbivorousanimalsamongthem,andthenestimatethatthepercentageofcloven-hoofedani-malsthatareherbivorousisapproximatelyequaltothepercentageinthesample.Priortosampling,statisticaldeductionwillwarrantalmostcer-tainlythatthepolicytheinquirerintendstoadoptformakinges-timatesofthepopulationfrequencywillbetrue.Atthatstage,forexample,theinquirerwillnotknowwhatthespecimensselectedatrandomfromthecloven-hoofedanimalsare.Retrospectively,how-ever,thestatisticaldeductionisunderminedbytheinformationthatthesampleselectedisasampleofSsandthattheproportionofPsinthesampleisr.Moreover,thosewhoseektouseBayes’Theoremtoderiveapos-teriorprobabilityonthedataforastatisticalconjectureorwhoseektousePeirce’searlyapproachtoexploittheinformationobtainedret-rospectivelyfromthedatatomakeajudgmentseemtobeinsomedifficulty.In1878and1883,Peirceproposedtothinkoftheinquirerasadopt-ingaprogramprospectivelyforusingthedatatomakestatisticalestimates.Theinquirerchoosesaprogramsuitedtohisorherob-jectivesandthensubsequentlyimplementsit.Implementationwillrequirecollectingdata;buttheinquireriscommittedtousingthedataasinputintotheprogramadoptedbeforehand.Beforeimplementingtheprogram,theinquirercanusestatisticaldeductiontoarguethatthepolicyofestimatingthepopulationfre-quencytobeapproximatelyequaltothesamplefrequencyisalmostcertainlytrue.Theinquireratthatpointlacksthesortofnuisanceinformationthatcouldinterferewithstatisticaldeduction.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n276isaacleviIfinsamplinganyclass,saytheM’s,wefirstdecidewhatthecharacterPisforwhichweproposetosamplethatclass,andalsohowmanyinstancesweproposetodraw,ourinferenceisreallymadebeforetheselatteraredrawn,thattheproportionofP’sinthewholeclassisprobablyaboutthesameasamongtheinstancesthataretobedrawn,andtheonlythingwehavetodoistodrawnthemandobservetheratio.ButsupposeweweretodrawourinferenceswithoutpredesignationofthecharacterP;thenwemightineverycasefindsomereconditecharacterinwhichtheseinstanceswouldagree.(W4,434)Theapproachexpressedhereisaclearformulationoftheapproachadvocatednearly50yearslaterbyJ.NeymanandE.S.PearsonandelaboratedbyA.Wald.Indeed,Peirce,liketheseauthors,requirednotonlythatthe“characters”whosefrequencyinthepopulationistobeestimatedbespecifiedinadvancebutalsothatthesizeofthesample(asthepassageclearlyindicates)bestipulatedbeforehand.Theinjunctionagainst“optionalstopping”isakeymethodologi-calmarkerdistinguishingadvocatesoftheNeyman–Pearson–WaldapproachtostatisticsfromadvocatesoftheBayesianapproachandtheircousins,advocatesofthelikelihoodview.Theinnovationisnotmerelythatcalculationsofprobabilityarerelativetotheinformationavailabletotheinquirerbeforetheex-perimentisinstitutedbutthataplanforimplementingaprogramfortakingdecisionsorreachingconclusionsisadoptedrelativetothepre-experimentinformation.Moreover,stepsaretakentosome-how“bind”theexperimentertofollowingthatprogramevenaftercollectingthedatanecessarytoimplementingtheprogram.Peircedidnotentirelyappreciatetheforceofhisowninvention.Hecontinuedtothinkofinductionas“inference”fromdatatakentobepremisestoaninductiveconclusionevenin1878and1883.Inferencecallsforusingthedatareportsaspremisesjudgedtrue,onthebasisofwhichtheconclusionisaddedtothebodyofbeliefs.Thedataarenotusedaspremises,accordingtothePeirce–Neyman–Pearson–Waldaccount.UsingdataaspremisesispreciselythesourceofthedifficultiesPeirce,Neyman,andPearsonallsoughttocircum-vent.Dataareusedas“input”intoaprogramthatdetermineswhatistobeaddedtoaninquirer’sbeliefs.Inmyjudgmentthereisaformofinferencefrompremisesrecog-nizableasinductive;butimplementationofthekindofpreplannedprogramforaddingnewinformationenvisagedbyPeircedoesnotCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism277fillthebill.Indeed,Peirce’sviewofinductionafter1878and1883andinallsubsequentwritingstandsopposedtoanyevaluationofhypothesesonthebasisofdataafterthedataarecollectedthatmaybecalledinferencefrompremisestoconclusion.10Nonetheless,Peircedidworryaboutretrospectiveassessmentofhypotheses.In1878,Peircecommentedonaninferencefromasam-pleofCretansallofwhomareliarsto“prettymuch”allCretansareliars:“whethertheremaybeanyspecialprobabilityinthat,Idonotknow”(W3,303).Ontheotherhand,incaseswheresamplesofMsareallPs,thelong-runfrequencyofclaimsofthetype“NearlyallMsareP”isverycloseto1.Peirce’sremarkin1883onstatisticaldeductionandinductionisbothrevealingandmoreinteresting:Thesetwoformsofinference,statisticaldeductionandinduction,plainlydependuponthesameprincipleofequalityofratios,sothattheirvalidityisthesame.Yet,thenatureoftheprobabilityinthetwocasesisverydifferent.Inthestatisticaldeduction,weknowthatamongthewholebodyofM’stheproportionofP’sisρ;wesay,then,thattheS’sbeingrandomdrawingsofM’sareprobablyP’sinaboutthesameproportion,–andthoughthismayhappennottobeso,yetatanyrate,oncontinuingthedrawingsufficiently,ourpredictionoftheratiowillbevindicatedatlast.Ontheotherhand,ininductionwesaythattheproportionρofthesamplebeingP’s,probablyhereisaboutthesameproportioninthewholelot;oratleast,ifthishappensnottobeso,thenoncontinuingthedrawingstheinferencewillbenotvindicatedasintheothercase,butmodifiedsoastobecometrue.Thededuction,then,isprobableinthissense,thatthoughitsconclusionmayinaparticularcasebefalsified,yetsimilarconclusions(withthesameratioρ)wouldgenerallyproveapproximatelytrue;whiletheinductionisprobableinthissense,thatthoughitmayhappentogiveafalseconclusion,yetinmostcasesinwhichthesamepreceptofinferencewasfollowed,adifferentandapproximatelytrueinference(withtherightvalueofρ)wouldbedrawn.(W4,416–7)AsIunderstandthispassage,inmakingtheestimatethatthepro-portionofMsthatarePsisρ,theinquirermayfailtoaddtheestimatetohisorherstockoffullbeliefs.Inthatcase,theinquirermightde-clarethatobtainingmoredataandmakingafreshestimate(usingthenewdataasinput)wouldeither“vindicate”thefirstestimateor“correct”it.Alternativelytheinquirermightactuallycometofullbeliefthatthefirstestimateistrue.TheinquirermightthenthinkCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n278isaaclevithatthereislittlepointinanadditionaltestbecausegiventhetruthofthefirstestimatetheprobabilityofvindicationbyrepeatperfor-mancewouldbeveryhighindeed.Nonetheless,theinquirershouldstilladmitthatifcountertofactthefirstestimatewerefalse,thenewestimatewoulduncoverthemistake.Asthepassagejustquotedindicates,Peircethoughtthatclaimingthatthe“conclusion”isprobablemeantsomethingdifferentthanitdidinthecaseofprobablededuction.Thus,hestoodindisagree-mentwithBayesiansandwithadvocatesofFisher’sfiducialargu-ment.Whattheclaimthattheconclusionisprobablemeansisthattheconclusionwasreachedbyahighlyreliableprocedurethatifreappliedwoulduncoveranymistakeintheestimate.Inductionisinthissenseself-correcting.PeircedidnotofferthisaccountofinductionashiscontributiontosolvingHume’sproblemofinduction.Hedidthinkthathisaccountavoidspresupposinganyprincipleofuniformityofnatureoruniver-salcausation.Evenso,theinquirerneedstoassumebeforehandnotonlythetruthofthestatisticalmodel(or,atleast,ofsomealterna-tiveverysimilartothatmodel)butalsothatthemethodofsamplingimplementstherequirementsofrandomnessnecessaryfortheac-quisitionofthekindofdatademandedbythemodel.Soinductionrelieson“mattersoffact”thatgowellbeyondthetestimonyofthesensesandtherecordsofourmemory.SincePeircedidnotdemandajustificationofcurrentbeliefsnotsubjecttoseriousdoubt,thiscircumstancedidnotappearaproblemtohim.Forthesamereason,PeircedidnotseektoavoidHume’sproblembypressingforvindi-cationratherthanvalidationinthesenseofReichenbach,Feigl,andSalmon.Hedidnotthinkthatinductionis“self-correcting”intheratherlaughablesensepromotedbyReichenbachaccordingtowhichtheinquirerinterestedinpositingalimitofrelativefrequencykeeps“correcting”hisorherestimateswiththeacquisitionofnewdataadinfinitum.ThereisoneseriousproblemwiththeaccountofPeirce’sviewofinductionIhavejustpresented.Ifdataareusedasinputandnotaspremises,theninductioncannolongerbeconsideredasinferencefrompremisestoconclusion.Tobesure,wesometimesspeakofdrawinginferencesfromsuppositionstakenastrueforthesakeoftheargument,butthepremisesofaninferenceareotherwiseclaimsjudgedtobetrueandusedasevidenceinjustifyingtheadditionoftheCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism279conclusiontothestockofbeliefs.Whendataareusedasinput,theyareneithersuppositionsnorpremisesjudgedtobetrue.SincePeircecontinuedin1883tothinkofinductiveinferenceasanargumentfrompremisestoconclusionexhibitingtheformofatranspositionofthemajorpremiseofastatisticaldeductionanditsconclusion,hisviewsseemtocontainelementsinsometension.IdoubtverymuchthatPeircesucceededincompletelyremovingthistensioninhisideas.Buthetookalargestepinthatdirectionwhenattheturnofthecenturyheexplicitlyabandonedtwokeyelementsofhisearlieraccountofinduction:(a)Theappealtosemeioticconsiderationsofinformedbreadth,depth,andinformation.(b)Theemphasisontakinginductionasatranspositionofpremisesandconclusionofasyllogism.Feature(a)hadalreadybeenabandonedsilentlyby1878.Feature(b)hadbeenexplicitlymodifiedin1878and1883.Attheturnofthecentury,Peircerealizedhehadtoretrenchstillmore.Hisreasonshadtodowithhismisclassificationofthetypesofinferencehecalled“hypothetic”asdistinctfrominductiveinference.8.induction,hypothesis,andabductionWhenPeirceintroducedhypotheticinferencein1865and1866,hecharacterizeditintermsofitsformandcontrastedthisformwiththeformofitsexplainingsyllogismandtheinductiveinferencethatsharedthesameexplainingsyllogism.Peirceattachedsomeimpor-tancetothispoint.Logicisunpsychologisticbecauseitinvestigatesforms.Inthecaseofdeductivelogic,validityisitselfapurelyfor-malmatter.Bywayofcontrast,inductiveandhypotheticvalidityarenotcharacterizableentirelyintermsofformalconsiderations.Buttheformsofinferenceeligibleforevaluationwithrespecttotheirstrengthasinductiveorhypotheticargumentscanbeidentified.By1883,Peircehadbrokenwiththeideathatinductionandhy-pothesisaretranspositionsofcategoricalexplainingsyllogisms.In-ductiveinferencesareobtainedbytransposingthemajorpremiseofastatisticaldeductionwiththeconclusionafterthefashional-readyexplained.Hypotheticinferenceisalsoobtainedbytransposi-tionsofpremisesandconclusionsofstatisticaldeductions.AndtheyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n280isaaclevidifferfrominductionsinthatthetranspositionisoftheminorpremisewiththeconclusion.Butthestatisticaldeductionwhosepremiseandconclusionaretransposedhasadifferentstructurefromtheexplainingstatisticaldeductionforinduction.Itisa“statisticaldeductionindepth.”Theideaseemstobethis.ThemajorpremiseofthestatisticaldeductionindepthassertsthatallMsareP1s,P2s,...,Pns.Shasanr-likenesstotheMs.Hence,probablyandapproximately,SresemblesanMinroutofthenre-spectsPi.Thereareseveralunexplainedfeaturesofthismodeofinference.Oneconcernstheinterpretationof“r-likeness.”Peircecouchedmat-tersthiswaytoguaranteethattheexplainingdeductionistrans-formedintoahypotheticinferencebytransposingtheminorpremiseandtheconclusion.Immediatelyafterpresentinghisidea,Peirceconcededthatinthe“extendedsense”hehadgiventothetermin-duction,theargument“issimplyaninductionrespectingqualitiesinsteadofrespectingthings.”ThePisare“arandomsampleofthecharactersofMandtheratiorofthembeingfoundtobelongtoS,thesameratioofallthecharactersareconcludedtobelongtoS”(W4,419).Thissuggestsanalternativeexplainingstatisticalde-ductionfortheso-calledhypotheticinference.ThemajorpremiseassertsthatobjectShasproportionrofthe(relevant)charactersofMs.ArandomsamplehasbeentakenofthecharactersofMs.Hence,probablyandapproximately,Shastheproportionrofthecharac-tersthusselected.Bytransposingthemajorpremiseandconclusionofthisprobablededuction,thesameinferenceinitiallyclassifiedashypotheticisexhibitedintheformofaninduction.Peircepreferredpresentingtheinferenceashypotheticbecauseoftheimpossibilityof“simplycountingqualitiesasindividualthingsarecounted.Charactershavetobeweighedratherthancounted”(W4,419).Butheclearlyhadappreciatedthattheinferencecouldbeclassifiedonewayortheother.And,inanycase,bothinductionandhypothesiswererecognizedbyPeircetobeampliativeratherthanexplicative.Apparentlytheawkwardnessofhissystemofclassificationeven-tuallyledPeircetoabandonthebasictemplateforclassifyingscien-tificinferencesthathehadusedfrom1865to1883whenhemadehismostimportantcontributiontothefoundationsofstatistics.BythebeginningofthetwentiethcenturyPeirceacknowledgednotonlyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism281thathisuseofhissemeiotictoassesstheacceptabilityofinductiveandhypotheticinferencewasoverblownbutalsothatsowashisuseofpermutationsofsyllogismstoclassifyampliativeinferencesintoinductiveandhypotheticinferences.Peircereplacedhisearlierclassificationwithanew,moreramifiedaccount.Adivisionbetweendeduction,induction,andabduction(orretroduction)replacedthetrichotomybetweendeduction,induc-tion,andhypothesis.Thechangewasmorethanterminologicalandcosmetic.Asalreadynoted,deductionincludednotonlycategoricalsyllogism,propositionallogic,thelogicofrelations,andquantifica-tionbutalsoprobableandstatisticaldeduction.Bothprobable(andstatistical)deductionareusedinscientificinquirytoexplicatetheconsequences(morecruciallythetestableconsequences)ofconjec-turesproposedasanswerstoaquestionorproblemunderinvestiga-tion.Abductionyieldstheconjectures.Inductioninvokesthedataobtainedbyobservationortestonthebasisofwhichconjecturedanswerstothegivenquestionareeliminated.AlthoughPeirceinsistedthroughoutthatthisclassificationisofdifferenttypesofinference,itseemstomebettertothinkofitasfocusedondifferenttasksthatareundertakeninproblem-solvinginquiry.Oneneedstoidentifypotentialanswerstothequestionunderinvestigation.Thatisthetaskofabduction.De-ductionfocusesonderiving(eitherapodicticallyorprobabilisti-cally)consequencesofconjecturesthatwereobtainedviaabduc-tion.Inductionevaluatesthestatusofconjecturesasaresultoftesting.Themodelisneitherthehypotheticodeductivemodelnorafalsi-ficationistmodel.AccordingtoPeirce,the“deductive”componentconcernsdeductionbroadlyconceivedtoincludestatisticaldeduc-tion.AndPeirceappreciated(asBraithwaitelaterdidbutPopperap-parentlydidnot)thattherejectionofastatisticalconjecturebecausethepredictionsderivedviastatisticaldeductionareobservedtobefalseisanampliativeinductiveinference.Inhislateryears,Peircecouldunderstandinductioninthiswaybecauseinductionincludedallampliativereasoning.InadraftofalettertoPaulCarusfrom1910,publishedasCP8.214–239,inductionsoconceivedincluded(a)quantitativeinduction,inductionfromasamplewhoserepresentativenessofthepopulationfromwhichitissampledissecuredbyrandomsampling;(b)qualitativeinduction,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n282isaacleviwhichPeirceadmittedheclassifiedashypotheticreasoningin1883;and(c)crudeinduction,whichappearstobeinductionbysimpleenumeration.11Throughoutthisdiscussiontheappealtotranspositionsofsyllo-gismsiscompletelyforgotten,asisthesemeiotic.Thereisnoef-forttoofferanaccountofthedifferencebetweendeduction,induc-tion,andabductionasadistinctionbetweeninferencesofdifferentforms.Peirce’searlyattemptstounderstandsomeaspectsofscien-tificreasoningotherthandeductivelogicaspartofformallogicareabandoned.Thedistinctionbetweendeduction,induction,andab-ductionfocusesonthedifferencesbetweenthetasksthataninquirerwillhavetoaddressinproblem-solvinginquiry.Consider,forexample,thetaskofidentifyingpotentialsolu-tionsoranswerstoagivenquestionorproblem.Thatisthetaskofabductivereasoning.Abductiveinferenceisfromthepuzzlingphenomenontoaconjecturedpotentialanswer.Onceapotentialanswerhasbeenproposed,onemightbringgeneralcriticalprinci-plestobearinascertainingwhethertheputativepotentialanswerisonethatisrelevanttothequestionunderstudy.Peircethoughtthattherearegeneralprinciplesforevaluatingsuchabductiveclaims.Thepragmaticorpragmaticistprincipleisthefundamentalprin-cipleofabduction.Itisnotageneralprinciplefordistinguishingtruth-value-bearingjudgmentsfromotherkindsofjudgments.Itdis-tinguishesbetweenpropositionsthatqualifyaspotentialanswerstotheproblemunderinvestigationandpropositionsthatdonotbyreferencetotheirtestableconsequences.Heretheconsequencesareconsequencesofthepotentialanswersgiventheinitialsettledbodyofbackgroundknowledgealreadytakenforgrantedasperfectlyfreefromdoubt(thoughvulnerabletofurtherscrutinyiftheresultsofsubsequentinquirywarrantraisingnewdoubts).Sincetheback-groundknowledgecanchange,theconditionsforbeingapotentialanswercanchange.Peirce’spragmaticprincipleisnotaverification-istsurrogatefortruth.Deduction(whichincludesbothnecessaryandprobablededuc-tion)isusedtoelaboratethetestableconsequencesoftheconjecturesformedviaabduction.Induction(whetherquantitative,qualitative,orcrude)contributestoconfirmingordisconfirmingtheseconjec-tures.Disconfirmingaconjecture,likeconfirmingit,is,ingeneral,CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism283anampliativeinductionbecause,forPeirce,conjectureswilltypi-callyinvolvesomesortofstatisticalelement.Thus,Peirce’smatureviewwas(a)stronglyopposedtothesortofprobabilismsowidelyendorsedbyphilosopherswhoendorseBayesianism,(b)supportiveoftheimportanceofformingtestableconjectures,asPopperinsisted,but(c)inoppositiontoboththeprob-abilistsandthePopperians,stronglyinfavorofampliativeinduction.Itisalsoclearthatbytheendofhiscareer,Peircehadjettisonedthenarrowantipsychologisticviewofthelogicofscientificinferencewithwhichhehadbegun.UnlikeCarnapandPopper,whosoughttokeeplogicfreeofcontext-sensitivenormativecomponents,Peircecametoacknowledgethecentralityofnormativityandcontexttologic.Thismustbesooncethedistinctionbetweenthethreekindsof“inference”isbasedonthedifferenceinthetasksand,hence,theaimsofthesekindsofreasoning.Ibelievethatthisheightenedsensitivitytocontextderivedfromthelessonshelearnedwhileat-temptingtoaddressstatisticalreasoning–atopicthatremainedatthecoreofhisphilosophicalinterestfromthebeginningofhiscareertoitsend.Earlyon(W1,LectureVI),Peircewarnedofthedangersoftoomuchrelianceonsyllogisticreasoning.ButPeircehimselfactuallysoughttobuildhisaccountofinductionandhypothesisontheformalrelationsbetweenthesekindsofreasoninganddeduction–especiallysyllogism.Hecametoappreciatefullythedangersofsuchrelianceonlyinthelastdecadeofhiscareer.Peirceshouldnotbefaultedforhisslownesstoseethedangersofflirtingwithtranspositionsofthesyllogism.Muchtwentiethcenturyandcontemporarythoughtoninductionandprobabilityre-mainstoocloselytiedtoparadigmscharacteristicofdeductiverea-soning.ToPeirce’scredit,heappreciatedtheseriousdifficultiescon-frontingthosewhoembracesomeformorotherofprobabilismandthosewhohavebeenbewitchedbythedeductivismandfalsification-ismofPopper.Hehadanunderstandingofthedistinctiveissuesthatthe“nonmonotonicity”ofinductivereasoninggenerates.Andhedidproposeanaccountof“inductivebehavior”thatanticipatedtheideasofNeymanandPearsonahalfcenturybeforetheypresentedthem.Hispragmaticprinciple,takenasthefundamentalprincipleofab-duction,isasophisticatedversionofPopper’sdemarcationcriterion.ButunlikePopper,whoseemedmysteriouslytothinkthatproducingCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n284isaaclevitestableconjecturesviaabductionmighthaveapointwithouttheinductivereasoningneededtodecidebetweenthem,Peirceunder-stoodthatthetestingofconjecturescalledformorethandeductivereasoning.Peircelearnedtobewaryofsyllogismsinawaythatnei-therCarnap,Popper,northeirepigonesevermanagedtodo.notes1.ThedocumentstowhichIreferincludetheHarvardLecturesof1865andLowellLecturesof1866(W1);“TheClassificationofArguments”and“UponLogicalComprehensionandExtension”of1867,thereviewofVenn’sLogicofChanceof1867,and“GroundsofValidityoftheLawsofLogic;FurtherConsequencesofFourIncapacities”of1869(W2).2.Ihavenodoubtthatheassumedthisin1865aswellalthoughthereisnotextualevidencetosupportthisclaim.ItseemsthatPeirceattheveryoutsetofhiscareerthoughtofinductiveinferenceinastatisticalsettingwhereprobabilityisunderstoodintermsoflong-runorlimitingrelativefrequency.However,hisappreciationofthesignificanceofstatisticalconsiderationsforhisviewofinductionandhypothesisastransposi-tionsofsyllogismsseemstohaveundergoneimportantchangesfromthattimethrough1883,asIshalltrytoexplain.3.Peircecontended(W1,265–6)thatotherformsofinductiveargumentcanbeobtainedbytransposingsyllogisticformsinthesecondandthirdfigurelegitimatedbycontraposition.(Suchtransformationsare“apogogicinversions.”BaconianinductioncombinesatranspositionofthemajorpremiseandconclusionofasyllogisminBarbara,suchastheinferenceto“Allcloven-hoofedareherbivorous,”withanotherinduc-tionobtainedfromasyllogisminCamestres(AEEFigure2),wheretheApropositionistheRuleinthesyllogisminBarbara,EisthenegationoftheResultofthesyllogisminBarbara,andtheconclusionisthedenialoftheCaseofthesyllogisminBarbara:AllCloven-hoofedareherbiv-orous.Asampleistakenfromnonherbivorousanimals).Specimensofrats,dogs,andapes.Sowehavethatrats,dogs,andapesarenotherbiv-orous.Anditisdiscoveredthatthesespecimensarenotcloven-hoofed.ThisexplainsthedatainaninductionwhosepremisesarethedenialsoftheCaseandResultinBarbaraandwhoseconclusionistheRule.PeircethenexploredFigure3AII,explainingsyllogismsinDatisi.Heclaimedtheirinversionscangive“indirect”supportfortheconclu-sionsoftheinductionswhoseexplainingsyllogismsareinBarbaraandCamestres.Inthefinalparagraphof“OntheNaturalClassificationofArgu-ments”of1867andmoreexplicitlyin“ATheoryofProbableInference”CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nBewareofSyllogism285of1883,Peircearguedthatinthecaseofprobablededuction,wheretheRuleorMajorPremiseisastatisticalassumptionratherthanacategor-icalone,thereislittleuseforthetranspositionsinthefigure3case.Theyallarevirtuallyreducibletothestatisticalanalogueoffigure1.4.“Deduction,InductionandHypothesis”of1878distinguishesbetween“necessary”deductionsinBarbaraand“probable”deductions“ofsim-ilarform”(W3,329).ThismayhavebeenPeirce’sattitudeinthe1866LowellLecturesandthediscussioninthepaper“TheClassificationofArguments”of1867.Itisveryexplicitlytheattitudein“GroundsofTheValidityoftheLawsofLogic”of1869(W2,p.267).In1883,Peirceretainedtheideathatstatisticalpremisesarerulesbutdistinguishedbe-tweenstatisticalrulesandcategoricalpropositionsasrules.Syllogismsarenecessarydeductions.Probabledeductionsarenot.Soprobableandstatisticaldeductionsarenotsyllogisms.Peirceretainedananalogybe-tweensyllogisminBarbaraandstatisticaldeduction;butasnotedabove,therearesomemarkedandimportantdifferences–manyofwhichherecognizedintheearlierworkbutwhosesignificancewasmoredifficulttoarticulatewithintheframeworkofhisearliertypology.5.Peircemayhavebeenstimulatedbyhisinvolvementwithhisfather,BenjaminPeirce,incompilingandanalyzingstatisticaldatapertainingtotheauthenticityofthesignatureofSylviaAnnHowlandtoanad-dendumtoherwillintheHowlandWillCase.BenjaminandCharlesPeircegavetestimonyin1867.HealsowroteareviewofVenn’sLogicofChancein1867that,asweshallsee,raisedaquestionaboutprobabledeductionthatposedissuesforinduction.Ishallarguethathemanagedtofinessethesedifficultiesonlyinthe1878and1883essays.Consider-ationofVenn’sideasmayalsohavecontributedtothedevelopmentofhisideasoninduction.6.Seefootnote4.7.Peirceexpressedacommon,mistakenprejudiceamongprobabilistswhenheinsistedthatprobabledeductionisthequantitativebranchofdeductionwhereassyllogismisthequalitativebranch.Byimplica-tion,thisimpliedasimilarrelationbetweeninductivetranspositionsofstatisticaldeductionsandtranspositionsofsyllogisms.In1883,Peirceexplicitlystatedthatwhenr=1or0,theinductionisan“ordinaryin-duction,”bywhichhemeantaninferencetoauniversalgeneralization.Intheearlierdiscussions,thisobservationwasnotmadebutIthinkPeircetookitforgranted.In1866,Peirceshiftedfromdiscussionofin-duction,illustratedbyinferencefromdataobtainedfromsamplingballsfromabagtoconclusionsabouttheproportionsofballsofagivencolorinthebag,toexamplesofinductionbytransposingcategoricalsyllo-gisms.ThereisonlytheslimmestofhintsastowhattheconnectionCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n286isaaclevibetweenthetwowaysofunderstandinginductiveinferencemightbe.YetPeircedidconsideraversionofthesamplingexamplewheresevenballsareselectedatrandomfromthebagandallarefoundtobeblackandtheinferenceisto“Almostallballsinthebagareblack.”Anditisclearthathethoughtthedifferencebetween“All”and“Almostall”tobeimportanteventhoughhedidnotexplainwhyhethoughtsoin1866ashedidin1883.Allofthismakessenseifweinterprethimasendorsingin1866theviewofthequantitativeandqualitativebranchesofdeduction.Intheearlyyearsofthetwentiethcentury,PeircebecameawareofBorel’sworkandbegantoacknowledgethatalthoughtosay“Allballsinthebagareblack”warrantsclaimingthatthestatisticalprobabilityofobtainingablackballonrandomselectionfromthebagis1,theconversedoesnothold.Howfullyheappreciatedtheramifica-tionsofthispointforhisviewsIcannotsay.8.Thissuggeststhatevenin1865Peircehopedtousehisaccountofin-ductionastranspositionofpremisesandconclusionsofanexplainingsyllogisminordertoaddressstatisticalinferencefromsampletopopula-tion.Inthatcase,theprogressionintheintroductionofstatisticalissuesIhavenotedinPeirce’spapersmayindicatetheextenttowhichhewasabletoaddressvariousproblemswithrelatingcategoricalsyllogismstowhathelatercalledstatisticaldeductionsand,hence,toprovidinganaccountofinductionandhypothesisastranspositionsofpremisesandconclusionsofdeductiveinferences.9.Iftherearenballsinthebag,thenumberofpossibleproportionsisn+1.Peircedidnotaddresstheissueofinfinitelymanyvalueswell;butforcountablyandevencontinuumlymanyvaluesitispossibletoaddressthematterinacoherentfashion.Peirceclearlymaintainedin1866thatinductiondoesnotpresupposethatnatureisuniformbutthatitmayinsteadexhibitstatisticalregularities.10.HeopposedargumentsforbalancingargumentsessentiallyrationalizedinBayesiantermsbothearlyandlate.11.Peircewasidentifyingspeciesofinductionthatarevariantsoftheclas-sificationinthedraftlettertoCarusin1901in“TheLogicofDrawingHistoryfromAncientDocuments.”BythenPeircewasusingtheterm“abduction”exclusivelyinthesenseofformingconjecturedanswerstoquestionsthatthencallfortesting(EP2:96–106).Hefurtherrefinedquantitativeinductiontomarkoffvariationswherethepopulationfromwhichsamplingisdoneisnotfinite.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nrandalldipert12Peirce’sDeductiveLogic:ItsDevelopment,Influence,andPhilosophicalSignificanceLogicmaywellbethesinglemostimportantkeytounderstand-ingPeirce’sthoughtandinfluence.ItwashisdeductivelogicthatbroughthimaninternationalreputationinhislifetimeandledtoconspicuousreferencestohisworkbyfiguressuchasPeano,Schroder,Russell,Venn,Jevons,andClifford.Peirce’shighest,and¨infactonly,academicpositionwasaslecturerinlogicatJohnsHop-kinsUniversity.Hehimselfsaidonnumerousoccasions–whenhewasn’temphasizinghisroleasaworkingscientistwiththeCoastSurvey,thatis–thathewasmainlyalogician.Hecalledhisexisten-tialgraphshischefd’oeuvre.1Logic,especiallythelogicofrelations,playedacentralroleinthedevelopmentofhisphilosophy.2HisthreeCategorieswerebasedon,andshowntobefundamentalby,thelogicofrelations.Thelogicofrelationsiscentraltohisanalysisofthefundamentaltriadicnotionofhissemeiotics,“signifiesto.”Hesawhistheoryofscientificmethodasjustlogic,broadlycon-strued.Andofpragmatismitself,heoftenrepeatedthatitwasnoth-ingmorethantheidealfixationofbelief,andthiswastheverygoaloflogic.Inthischapter,IwilladdressthreeaspectsofPeirce’slogic:asur-veyofPeirce’smajorideasandinfluence,asketchoftheirdevelop-mentfrom1868(thedateofhisfirstpublicationinlogic)untilhisdeathin1914,andfinally,severalphilosophicalandlogicalideasofenduringvalueinhiswork.Iwillrestrictmyselfheretowhatbothweandheunderstoodasdeductivelogic.Althoughhisdeductivelogicisself-contained,anddevelopedforthemostpartindependentofhistheoryofscientificmethod,whichincludedalsoinductionandabduction,itissometimesdifficulttoseparatedeductivelogicfromthismethod(andfromhisantilogicistconceptionofmathematics).287CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n288randalldipertIwillpayspecialattentiontothreethemes:tohisdefinitionsandconceptionsofthepurposeoflogic,towhatheregardedasthecoreprinciplesandconceptsofdeductivelogic,andtotherolethathistheoryofsigns(especiallydiagramsandiconicity)playsinthede-velopmentofhislogicalsystems.IwillarguethatthereareseveralsignificantchangesinwhatPeirceregardedasthekeyprinciplesoflogic.Atfirst,heregardedtheAristotelianfirstfiguremood,Bar-bara,asthecenterandeventhewholeoflogic.Inabriefperiodbetween1865and1867,hebecameconvincedthatBoole’s“alge-braic”approachtologicwasbasicallycorrect.By1870,andundertheinfluenceofDeMorgan,hemovedfromBoole’sownequation-formattoformulasusingasubsumptionoperator(–<)thathadtheimportantformal-algebraiccharacteristicsoftransitivityandanti-symmetryanddiscoveredaningeniousextensionofBooleanlogictorelations.Partofthisnewconceptioninvolvestheautonomoususeofasinglelogicalnotionforthecalculiofsubsumptioninthetheoryofclasses,forthematerialconditionalinpropositionallogic,andforwhatPeircetermed“illation”(“therefore”)–andtherebyatheoryofdeductioninthemannerofGentzen.Peirce’searliercon-ceptionimplicitlyplacesemphasisonwhatwewouldregardastheuniversalquantifier,withtheexistentialquantifierdefinedeitherseparatelyorderivatively(andoften,unsatisfactorily).Anumberofshiftsoccurinthelate1880sandespeciallyinthe1890swiththefirstattemptsatadiagrammaticlogic.Thebasiclogicalconstantandresultingcalculusbecameunequivocallypropositional,andthetran-sitivelogicalrelationisabandonedinfavorofdisjunction(togetherwithnegation).Theexistentialquantifieristreatedascoequalandinterdefinablewiththeuniversalquantifier.Thelaststepwastoshifttoconjunction(togetherwithnegation),acompletelydiagram-maticnotation,andtheprimacyoftheexistentialquantifierovertheuniversalquantifier.Iwillarguethatthisshiftactuallywastogiveupsomeofwhatwasadvantageousinhisearlierconceptionofthebasicprinciplesoflogic;IproposeextensionsandmodificationstotheexistentialgraphssothattheymightfullyfunctionasPeircehimselfintended.Hisnumerousandseeminglyvacillatingdefinitionsof‘logic,’inbroadersenses(thatembraceallidealizedthoughtandthoughtpro-cesses)andnarrowersenses(thatindicateprimarilydeductiveinfer-ence),presentamoreformidabledifficultytounderstandingPeirce.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sDeductiveLogic289Thisislikelywhyearlysecondaryliteraturehasnotfocusedonwhatisanembarrassmentofrichesand,atfirstblush,confusions.Iwillneverthelessmaintainthathismature,andIthinkveryfine,defini-tionoflogicis:Logicinthenarrowersenseisthatsciencewhichconcernsitselfprimarilywithdistinguishingreasoningsintogoodandbad,andwithdistinguishingprobablereasoningsintostrongandweakreasonings.Secondarily,logiccon-cernsitselfwithallthatitmuststudyinordertomakethosedistinctionsaboutreasoning,andwithnothingelse.(RLT:143)Since“goodandbad”arehereformal,qualitativeaspectsthat,ap-pliedstrictlytodeductivereasoning,amounttowhatwewouldterm“validity,”Peirce’sdefinitionofdeductivelogicisroughly:“Logicisthescienceofdistinguishing3validfrominvalidarguments,andthedevelopmentofallandonlythoseconceptsandtoolsthatareneces-sarytodoso.”Inspiteofnumerous,andapparentlyquitedifferent,statementsaboutboththebroadandnarrowsensesoflogic,IwillarguethatPeirceisactuallyperfectlyconsistentthroughouthislife.majorideasandinfluenceWithPutnam(1982)andDipert(1989),4aswellasintroductionsandnotestothevolumesoftheWritingsthathavesofarappeared,wearealmostinpossessionofaclearandcompleteoverviewofPeirce’slogicanditsplaceinthehistoryoflogic.Especiallyclearnowistheinfluenceofhispublishedworks,whichisfargreaterthanwasrecognizedasrecentlyas1980.However,afullevaluationofhisworkindeductivelogicmustawaittheassemblyandpublicationintheWritingsofwhatremainsofthemanuscriptsfortheproposedbook,TheGrandLogicof1893–4,TheMinuteLogicof1901–2,andthemanyfragmentsandpartialmanuscriptsusingthesystemofthe“existentialgraphs”intheperiodfrom1897untilhisdeathin1914.AnotherremaininglacunainPeircescholarshipisanappreciationandcarefulportrayaloftheworkofErnstSchroder,thedifferences¨betweenPeirceandSchroderonkeypoints,andacarefulassess-¨mentofSchroder’s(andthusPeirce’s)influenceondevelopmentsin¨thetwentiethcentury.5Inanutshell,Peirce’sfirstcontributionswere,first,minorimpro-vementstoBoole’slogic(1867–1868),togetherwithasophisticatedCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n290randalldipertunderstandingofAristoteliansyllogismonitsowntermsandasre-flectedthroughthisalgebra,thentheimportantandoriginaldevel-opmentofabroadlyBooleanlogicappliedtothelogicofrelations(1870–1894)andtransformedfromasystemofpureequations,andfinallyaburstofactivityinthelate1890sthatinterminglednorma-tivecomponentsfromhistheoriesofsigns(hissemeiotic),namely,iconicity,withhisearlier,moretraditionallynotatedlogicalsys-tem,whichcametoconstitutetheexistentialgraphs.Thislater,diagrammaticapproachtologicyieldedsophisticated,completelog-icaltheoriesthatlackstheBoolean,algebraicflavorofhispreviouswork.Infact,theyareconspicuouslyandadvantageouslyunlikeany-thingelseinthehistoryoflogic.Unlikeallearlierdiagrammaticlog-ics,Peirce’stechniquecouldexpressmultiplyquantifiedrelationalexpressions.Unlikehisearlier,symbolicworkinlogic,whichwewouldseeassomewhatcasualwithregardtorulesofformation,axiomatization,andevensemanticinterpretation,hisapproachtodiagramswasmuchmoreexacting.AnaccountoftheexistentialgraphsunfortunatelyremainedinunpublishedmanuscriptsuntilthepublicationofVolumeIV(“TheSimpleMathematics”)oftheCollectedPapersin1933.Sotheirin-fluenceonthecrucialfirstdecadesofthedevelopmentofmodernlogicalconceptsandnotationwasnil.Carefuldescriptionsandevalu-ationsoftheirimportancedidnotoccuruntiltheworkofJ.J.ZemanandDonRobertsinthelate1960sandearly1970s,6andwiththepub-licationofmoreofPeirce’sworkonthegraphsandonmathematicsinTheNewElementsofMathematicsin1976(NE).WithDeMorganinhisowncenturyanduntilAlonzoChurchandBensonMatesinthetwentiethcentury,Peircewasoneofveryfewlogiciansinrecentcenturiestoviewlogicasconnectedtoideasfrompreviouscenturies,especiallytoAristotle’sownandotheran-cientlogics,andtomedievallogic.Hisencyclopedic,historicist,andunmodernlybroadconceptionoflogic,aswellasthecomplicatedcurrentsoflatenineteenthcenturyphilosophyandmathematics,contributestomakinghismotivationsandthinkinginlogicdiffi-cultformodernreaderstoappraise.C.S.Peircewasthefirstpersoninthehistoryoflogictousequantifier-likevariable-bindingoperators(brieflyin1870,W2,392f,pre-datingFrege’sBegriffsschriftof1879);these“quantifiers”werestillmoreextensivelyusedinpublishedworksof1880andespeciallyCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sDeductiveLogic2911885.In1885,withoutanyknowledgeofFrege’spublicationofsixyearsearlier,PeircedeclarestheintroductionofquantifierstobetheresolutionofalongstandingprobleminBooleanlogic,namely,theproperexpressionofexistentialstatements,suchastheparticularcategoricalsofAristotelianlogic,“SomeF’sare(not)G’s.”Heex-pressedtheI-categoricalas“iFiGi,”usingsubscripted“indices”(here,i)forvariables,andthesymbols‘’and‘’forwhatwewrite,usingPeano’snotation,as‘∃’and‘∀.’7Hewas(againin1870)thefirstmajorfigureintheBooleantraditiontouseaspecialsymbolforclassinclusionorsubsumption(–<)asopposedtousingonlyequations.8Schroderfollowedhiminthis,butsubstitutinghissymbol¨
.InPeirce’searlyinterpretation,thissymbolalsostood,autonomously,forwhathetermed“illation”(the“therefore”:‘
’)andderivatively(until1885)forthematerialconditional.PeircerecognizedthatapropertyobservedbyDeMorgan(inva-lidityoftheSyllogismofTransposedQuantityininfinitedomains)wasalogicallyadequatecharacterizationofthedistinctionbetweenfiniteandinfiniteclasses,andhethusgaveadefinitionofaninfiniteclassthatisindependentof,anddistinctfrom,thebetterknowndef-initionsofDedekindandCantor.Interestedinwhatwewouldnowterm“computational”aspectsoflogic,hewasthefirstpersoninhistorytosuggestanelectricallogicalcircuitandhehadsomegrasp,startinginthe1880s,ofthenotionofeffectivecomputation.Hespec-ulatedaboutwhatbecameHilbert’sEntscheidungsproblem(decid-ability,1900)andwaslatersolvedbyChurch’sTheorem(1939)9–thatinfirstorderpredicatelogic,withquantifiersandrelations,therecanexistnoalgorithmfordistinguishingvalidfrominvalidar-guments(Dipert1984).JaakkoHintikkahasarguedthatPeircewasoneofthemostimportantfiguresinamodel-theoretictraditionthatfirstarosewithBoole,andallowedtheuniverseofdiscoursetobealtered,andthatthisallowedthemoderntheoryofsemanticsforlogicthatweseedevelopedinworksbyTarskiandGodel(Hintikka¨1988,Hintikka1997,andDipert1989).PeircecanbecreditedwiththerenewalofinterestintheLiarParadoxinthetwentiethcentury.10Whetherpartoflogicornot,Peircealsomadeimportantcontributionstothephilosophyofcol-lectiveentities,includingsets,andtothetheoryofnumbers,con-tinua,andinfinities.WithDedekind,hediscoveredatalmostthesametimeasPeanoaminimalandapparentlycompletesetofCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n292randalldipertpostulatesforthenaturalnumbers,nowknownasthePeanoPostulates.11Heanticipated,inanunpublishedbutfinishedpaper,“the”Shefferstrokefunction:thediscoveryoftwopropositionalcon-nectivessuchthateitherone,alone,issufficientfortheexpressionofthewholeofpropositionallogic.Hediscoveredvarioustechniquesforputtingquantifiedexpressionsintoacanonicalform,suchasaprenexnormalformthatseparatesanuninterruptedstringofquan-tifiersfromastrictlylogicalformulathatcontainsnoquantifiers(W5,178–183,1885).Heunderstood,proved,andutilized,especiallyinhisexistentialgraphs,thefactthatonecandefinetheuniversalquantifierintermsoftheexistential,orviceversa.PeircedidnotadvancetherigorousaxiomatizationoflogicinanywaysthatwerenotsimplycommonstandardsofrigorinnineteenthcenturymathematicsandpartoftheinheritancefromEuclidandearlyabstractalgebraists–althoughbythisseverecriterion,Fregemadenocontributionseither,andprogressinthisdimensionoflogicwasfirstpropelledbythepublicationofDavidHilbert’senor-mouslyinfluentialaxiomatizationofgeometryin1898,andthencetheAmericanpostulatetheorists(Scanlon,1991).ThegreatmajorityofPeirce’searlier,originallogicalwork,inwhatIwilltermhisAlgebraicPeriod,waspublishedinhislifetime.Thisperiodextendsfromhisfirstpublishedworkinlogicof1868throughtohispublished“OntheAlgebraofLogic:AContributiontothePhi-losophyofNotation”of1885.Inthisperiod,therearealreadyindi-cationsofa“semeioticturn”withinlogicthateventuallydevelopedintohisDiagrammaticPeriod.Theunpublishedbutextensivelyde-velopedGrandLogicof1893–4remainswithintheAlgebraicPeriod.ThetermsoflogicinthefirstpartoftheAlgebraicPerioduntil1885denoteclassesoroperationsinthemannerofBooleanlogic(evenwhentheyinvolverelations),andthesymbolismispredominantlynondiagrammaticandlinear.Thatis,notationconsistsoflinearse-quences(strings)ofsymbols.12Definitionsandinferentialrulesarealgebraicinstyle,informallymathematicalinpractice.Intheearlypartofthisperiod,1868–1885,propositionallogicistakentobejustaformofageneralcalculus.Inthelanguageofthetime,hypotheticallogicisaninterpretationofcategoricallogic,inwhichpropositionsdenotetheclassesofoccasionsortimeswhentheyaretrue(Dipert1981a).ExpandingonDeMorgan’sinsightandthecentralityofBarbaraintraditionallogic,PeirceholdsthroughoutthisperiodthatCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sDeductiveLogic293inference–allinference–dependsononealgebraicpropertyoflogicalconnectivesaboveallothers,namely,transitivity.13Inmyinterpretation,hispublishedarticle“TheLogicofRela-tives”(1897)isatransitionalworkbetweentheAlgebraicandtheDiagrammaticPeriods.Alreadyin1896therewasthegermofwhateventuallyconstitutedthesystemoftheExistentialGraphsthatheusedconstantlyfrom1900untilhisdeathin1914,namely,his“en-titativegraphs.”Itisnowusualtothinkofthislaterworkasmerelybeingan“iconic”treatmentofhisearlierdiscoveriesinthealgebraoflogic.Itisalsocommontothinkofthegraphsasintendedmerelyasatoolforthevisualrepresentationormanipulationoflogicalpropo-sitions,stillunderstoodmentallyaslinearlynotated.However,IwillarguethatitinsteadrepresentsalmostacompletebreakwithmostcentralmotifsinhisearlierworkintheAlgebraicPeriodandisbaseduponaphilosophyoflogicthatevenPeircehimselfdidnothavetheopportunityfullytodevelop.ThegreatesttragedyinPeirce’slifeasalogicianwashisfailuretofindapublisherforanybook-lengthworkonlogic,especiallyhisproposedwork,the“GrandLogic”of1892–3,withtheproposedti-tleofHowtoReason:ACritickofArguments.Thesourceofthetitle“GrandLogic”isnotknown(seedeWaal2000forthisandotherinformation),butespeciallysincethereisalaterbook-lengthprojectbyPeircefrom1902–3entitledtheMinuteLogic,wemayguessthatPeirceispossiblyalludingtotheLogicamagnaandLog-icaparvaofPaulofVenice,anauthorheknewwell.14LargesectionsoftheGrandLogicwerecomplete,butonlypartlypublishedinCP;anyportrayalofhislarge-scaleconceptionoflogicmustawaitarea-sonableorganizationofthesemanuscriptsandtheirpublication–promisedtooccupyallofVolume10oftheWritings.Peirce’slogicalworkundoubtedlyalsosufferedfromhislackofastableacademicposition,fromthelackofsophisticationandfollowingforcontem-porarydevelopmentsinlogicinAmerica,includinginmathematics,andfromthepaucityofinfluentialstudentshehadfromhisbriefperiodofteachingatJohnsHopkins(O.H.Mitchell,whodiedatanearlyage,andChristineLadd-Franklin;JohnDeweydidnotde-votehimselftodeductiveorsymboliclogic).OnemustalsocitethescatterednatureofPeirce’seffortsandinterests,acrossallmannerofphilosophical,mathematical,scientific,historical,linguistic,andevenliterarydisciplines,whileatthetimehewasmainlyemployedCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n294randalldipertasadatagathererbytheCoastandGeodeticSurveyandlater,af-terhisretirementtoMilfordin1890atage51,engagedinavarietyofmoney-makingschemes,andespeciallyinthelastdecadeoflife,combatinghispovertywhentheseschemesfailed.thedevelopmentofpeirce’slogicInthissection,Igoratherdeeplyintoselectedworks,butdonotattempttosurveyallofhisessaysandmanuscriptswiththesamedegreeofscrutiny.EarlyWorksThebasicideathatthereweresimilaritiesbetweensomefeaturesofalgebraandlogic,understoodintermsoftheAristoteliansyllogistic,wasobservedasearlyasthelateseventeenthcentury,andindepen-dentlydevelopedbyLeibnizandLambert,tonamejusttwo.Boole’sbasicideawasroughlythatcategoricalstatements,suchas“AllXsareYs,”couldbetreatedasanequation,xy=y,interpretedroughlyas“TheclassofXsisidenticaltotheclassofthingsthatarebothxandy.”(Inotherwords,x=x∩y.)This“Boolean”logichadop-erationsofmultiplication(roughly,classintersection)andaddition(classunion,butwithqualifications)andmultiplicativeandadditiveidentityelements1(theuniversalclass)and0(theemptyclass).Onedeviationinparticularfromstandardnumericalalgebramadeitoneofthefirstofthe“abstract”algebras,namelyxx=x.Peirce’searliestworkwaspartofaBooleanreformmovementintheperiodfrom1868(Peirce,Jevons)to1878(Schroder’s¨Opera-tionskreis)thatinterpretedBooleanaddition(classunion)inadif-ferentandmorewell-behavedway.Namely,forBoolehimself,ifclassesxandyhadanymembersincommon,‘x+y’wasundefined;theexpression‘x+y’wasonlymeaningfulifxandyweremutuallyexclusiveclasses.PeircehadreadBoole’s1854LawsofThoughtnolaterthan1865(age26).Alogicnotebookfrom1867survives,andinitwecanseesomeofthedevelopmentsleadinguptothepublica-tionsof1868and1870.In1867,PeircepresentedfivepaperstotheAmericanAcademyofArtsandSciences,andtheywerepublishedayearlater,1868,initsProceedings.Inthefirst,“OnanImprove-mentinBoole’sCalculusofLogic,”PeirceismetaphysicallyacuteCambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sDeductiveLogic295enoughfromtheoutsettodistinguishsymbolicallythe“logical”re-lationshipsandoperations,performedonclasses,fromtheirusualarithmeticalinterpretations,performedonnumbers.Thushewrites‘a=,b’asalogicalidentityholdingbetweenclasses,notnumericalequality;a+,bislogicaladdition,anda,bislogicalmultiplication.ThisaddressesoneofthecomplaintsinFrege(1880/1881:10)andFrege(1882),namelytheambiguoususeinBooleanlogicsofmathe-maticalandlogicalsymbols.Logicaladditionandmultiplicationareassociativeandcommutative;logicaladditionandmultiplicationarejointlydistributive,allasBooleanlogicisnowunderstood.Thein-verseoperationislesssatisfactorilydefined,namelyas,−a=1−,aorasthemaximalclasssuchthata,−a=,0.Infact,1and0areusedbefore0is“defined”as0=,x−,x=x−x.Peirceapparentlyintendstosaywiththisdoubleequationboththelogical0=,x−,xandthenumerical0=x−x.Thisisodd,however,since‘0’heredoesdoubledutyasthenullclassandasaminimalnumericalquantity.15Theremainingpartofthearticledealswithacalculusofprobabilities,withtheconnectingthreadbeingthat(27)ifa=,bthena=b.InCantorianterms,ifaandbareidenticalclasses,thenaandbareequinumerous(havethesamecardinality).Thecon-versedoesnothold.Observethat‘a’isproblematicallyambiguous,referringtoaclassinthefirstequationandtothenumericalsizeorcardinalityofthatclassinthesecond.Withthesetwonotations,firstforclassesbutwithaclass-union-likelogicaladdition,andthenforordinarynumericaloperations,theshifttoprobabilitiesrequiresthata+b=a+,b−,(a,b).16AsPeirceobserves(W2,212d),Boolehadensuredthatoperationsonprobabilitiessymbolicallyparalleledlog-icaloperationsbycreatinglogicaloperationsthatalreadyexcludedoverlappingcases.OnenotabledifficultyinBoole’s1854systemisinproperlyex-pressingtheIandOpropositionsofcategoricallogic:“SomeAisB”and“SomeAisnotB.”Itwouldbenatural,atleastinhindsight,toex-presstheI-categorical:a,b=,0(Peirce)andab=0(Boole).However,neitherBoolenorPeircewaswillingtoappealtoapropositionalnega-tionthatwouldallowthis;theirsystemsarestrictlyequational.17(InBoole’s1847bookthisismoresatisfactorilyexpressed.)CambridgeCompanionsOnline©CambridgeUniversityPress,2006\n296randalldipertPeirce’s1870article,“DescriptionofaNotationfortheLogicofRelatives,”isundoubtedlyoneofthemostimportantworksinthehistoryoflogic.Itisherethatageneralnotationformultiplyquanti-fiedrelations,andtechniquesformanipulatingthem,firstappear.A“relative”isaterm,inthesenseinwhichitisusedbyAristotelianlogicians.Thatis,therelationshipbetweenatermanda(one-place)propertyispreciselyduplicatedbytherelationshipbetweenarela-tiveandarelation.Aswewillsee,arelativeterm(‘relative’forshort)doesdoubleduty,semanticallyrepresentingacertainextensionorclass,namelythe“logicalsum”of(ordered)pairs(n-tuples)ofindi-viduals.Thisispreciselythemodernsemanticunderstandingoftheextensionofarelationofnplacesasasetofn-tuples.Butitalsoservesasanoperationonclasses.This1870paperisnottheveryfirstpublishedsymbolictreatmentofrelations:creditforthiswouldgotoLambert,orbetterknownandcruciallyinfluentialonPeirce,toDeMorgan’s“OntheSyllogismIV.”Thissaid,Peirce’s1870paperisremarkableforitssheerimagina-tiveness,butalsoforitsdisorderlypresentation.Inmanycases,thedevelopmentamountstoexperimentationwithvariousnotationsforrelations,andtothefollowingoutofalgebraicanalogies(suchaswithexponentiationandabinomialtheorem,somethingBooletooattempted,thoughnotforrelatives).SomeofthenotationalmethodsandanalogieswerenotevenusedagainbyPeircehimself.Neverthe-less,itisobviousthatthebasictechniquesallowedPeircetoex-pressquantifiedrelationalstatementsofenormouscomplexityand,inmanycases,toshowtheirequivalencetootherstatements.Forexample,whateverisloveroforservanttoawomanisthesameclassasthenonrelationallogicaladditionoftheloversofawomanandtheservantsofawoman:(l+,s)w=lw+,sw.Here,relationsareindi-catedbyitalicizedletters,andsimpleclassesbynonitalicizedletters.Observetoothat(asPeircenotes)juxtapositionindicatesanotionof“application”oftherelativetoaclass,andnotanysortofordinarylogicalmultiplication(intersectionofclasses).18Thisisthesenseinwhicharelativebehavesmorelikeafunctionoroperatorthanaclassorterm.19However,modernsettheoryfrequentlyequivocatesinthesamemanner(onceextended,thisisthekeyideaofCategoryThe-ory),andpredicateswerealsoconceivedas“propositionalfunctions”byFrege,Russell,andWhitehead.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sDeductiveLogic297Thereisnolistofaxioms,norevenanorderingofwhichtheoremsaremorefundamentalthanothers.Inthe“Conclusion,”Peircesaysthatcertainequationscanbetakenasaxioms,thenadds:“Buttheseaxiomsaremeresubstitutesfordefinitionsoftheuniversallogi-calrelations,andsofarasthesecanbedefined,allaxiomsmaybedispensedwith.Thefundamentalprinciplesofformallogicarenotproperlyaxioms,butdefinitionsanddivisions”(W2,4291870).Inotherwords,Peircetakesaslogic’stasktheconceptualanalysisofitsideas,ratherthantheorderingofitstruths.Withtheexceptionofhis1885essayandtheworkingpaperleadinguptoit,andhislaterexpositionofexistentialgraphs,hewasnevermuchconcernedwithaxiomsforlogic,oranyorderingoftheorems.Thiscausesforthemodernreaderanappearanceofcasualness,evensloppiness,thatis,however,moreduetoadifferenceintheconceptionandpurposeoflogic.Asearlyas1872–3,Peircehadsomeplantowriteabookonlogic.Thereexistsaseriesofmanuscripts,publishedasSections4–39ofVolumeIIIoftheWritings,thathavechapterheadingsandacomprehensiveandorganizedsubjectmatter,andthereexistsashortmanuscriptfromthesummerof1873entitled“NotesonLogicBook.”20SomeoftheseessayswereincorporatedintothePopularScienceseriesof1877–8.Thislastsectioncontainswritingonthecopula,relativeterms,andthesimplesyllogism.Itcontainslittlethatwasnotevidentinthepublished1870paper,butiswritteninafarmoreapproachablestyle.Ofsomeinterestarehisdiscussionsofthe“breadth”and“depth”ofterms,andhisviewthatalgebraisabranchoflogicandthatalgebraisnotexclusivelyamathematicalstudyofquantity.Thisessayismodifiedfromhis1867AmericanAcademyofScienceslecture,“UponLogicalComprehensionandExtension”(W2,70–86),andwasrecycledagain,withsomemodi-fications,intheproposedGrandLogicof1894(R-421;alsoR-725),wheremuchfrommedievalsemanticsisinterjectedorquoted.Al-thoughusuallyrefractedthroughtheinfluenceofmedievalseman-tics,thePort-RoyalLogic,Hamilton,andevenJ.S.Mill,Peircewasalwaysextremelyinterestedinexploringvarioussemanticdimen-sionsoftermsthatwentunderexpressionssuchasextension,inten-sion,denotation,connotation,comprehension,breadth,anddepth.ModernphilosophyoflanguagetendstostartwithFrege’sdistinctionCambridgeCompanionsOnline©CambridgeUniversityPress,2006\n298randalldipertbetweenreference(Bedeutung)andtheproblematicnotionofsense(Sinn).Muchaswithhishistoricallysophisticatedinvestigationsofvariouscollectivenotions,returningtoPeirce’sworksallowsustoseearicharrayofnewpossibilities,sometimesbrilliantlyexaminedbyPeirce,butthatweresuddenly,andsometimeswithlittleornoreason,abandonedaround1900.From1873through1878,PeircewasconsumedwithhisworkfortheCoastSurvey,and,laterinthatperiod,withhisonepublishedbook,PhotometricResearches(1878),aswellaswiththeprepara-tionofthe“IllustrationsoftheLogicofScience,”thatis,thePopu-larScienceMonthlyseriesof1877–8.Thisseriesbeginswithararewittyandtrenchantremark:“Fewpersonscaretostudylogic,be-causeeverybodyconceiveshimselftobeproficientenoughintheartofreasoningalready”(W3,242).Otherwise,however,thepe-riodfromthe“Notation...”of1870until1880wastheonlydecadeinPeirce’sadultlifewhenhedidnotdevotehimselfextensivelytologic.(Theotherlengthyperiodwithoutpublicationisfrom1885until1897,asPeircehimselfnotesatCP3.510.)Hisonestrictlylog-icalpublicationinthisperiod,anoteintheimportantBritishjournalMindin1876(W3,191–194),isremarkableforitsslightness,dealingwithcontrapositionandconversioninAristoteliansyllogisms,whencomplementation(non-Xs)oftermsispermitted.ThederivationofthemanyvalidmoodsthistechniqueallowsdoesshowPeircetobestillinthegripsoftheconvictionthattheBarbaramoodintheFirstFigureisofsingularimportance.ThisperiodendswithhisappointmentaslecturerinlogicatJohnsHopkinsUniversity,Baltimore,from1879to1884.JustasreadingDeMorgan’spaperonthelogicofrelationshadimmediatelyinspiredhisownpublishedworkonrelatives,itisclearthatreadingSchroder’s¨shortbookOperationskreisdesLogikkalkuls¨(1878)inspiredanef-forttowardamuchmoresystematicexpositionoflogicalprinciplesthanwasattemptedin1870.Thisisevidentintheunpublished“OntheAlgebraicPrinciplesofFormalLogic”(Fall,1879),withmanyreferencestoSchroder’spostulatesanddeductions;itbecamethe¨lengthyandrefined1880articleintheAmericanJournalofMath-ematics,“OntheAlgebraofLogic”(W4,163–209),inwhichtherearescantreferencestoSchroder,andtheexpositionisunfortunately¨lesssystematicthantheearliernotes.CambridgeCompanionsOnline©CambridgeUniversityPress,2006\nPeirce’sDeductiveLogic299Thedevelopmentofanalgebraoflogicinthepublished1880“OntheAlgebraofLogic”isfarmoreleisurelyandorganizedthanhadbeenthe“Notation...”oftenyearsearlier.Theinventivenesswithnotationismuchmoreconstrainedthanin1870,withonenotableexception.Peirceusesadashoverthesubsumptionsigntoindicateapropositionalnegation:‘s–

查看更多