Introduction
1.
Theessaysinthisvolumediscussancientepistemology,especiallyinPlato,Aristotle, andthePyrrhonianskeptics.Onecentralthemeiscognitiveconditionsandtheir contents.¹Inparticular,howare epistêmê, doxa, dogma,andthecorresponding verbs,tobeunderstood?Is epistêmê knowledgeasitisconceivedofnowadays? Are doxa and dogma beliefasitisconceivedofnowadays?Considerationofsuch questionsshouldnotbelimitedtotheGreekwordsjustmentioned.SinceGreekhas arichcognitivevocabulary,weshouldalsoconsiderotherGreekwordsthatare sometimestranslatedas ‘knowledge,’ suchas eidenai and gignôskein,aswellasother Greekwordsthataresometimesthoughttoindicatebelief,suchas hupolêpsis and oiesthai.Weshouldalsoconsiderwhatknowledgeandbeliefare.
The firstworkinthePlatoniccorpusthatengageswithsomeoftheseissuesis the Apology,whichIdiscussinChapter2.Socratesisoftenthoughttohavesaid thattheonlythingheknowsisthatheknowsnothing.Ifhesaysthis,andif ‘know’ isusedunivocally,heseemstocontradicthimself.Forifhe knows thatheknows nothing,hedoesn’tknowanything,contrarytohisclaimtoknowsomething(viz., thatheknowsnothing).Ontheotherhand,ifhedoesn’tknowanything,thenhe can ’tknowthatfact.Forifhedid,hewouldknowsomethingafterall.Yettheview thatSocratessaysthatheknowsthatheknowsnothing,thoughbynomeans universal,²iswidespread.³InChapter2,Iaskwhether,inthe Apology,Socrates says,orimplies,thatheknowsthatheknowsnothinginawaythatinvolvesselfcontradiction.Ialsoconsidertherelatedbutdifferentquestionofwhetherhesays thatheknowsnothing,aclaimthat,unliketheclaimtoknowthatoneknows nothing,isnotself-contradictory.⁴
¹Cognitiveconditionsincludestatesofknowingandbelieving,aswhenJohnknows,orbelieves, that2+2=4.Cognitivecontentsincludewhatisbelievedandknown,e.g.that2+2=4.
²Seee.g.C.C.W.Taylor, Socrates:AVeryShortIntroduction (Oxford:OxfordUniversityPress, 2nded.,2019),44–50.
³Foronerecentexample,seeS.Bakewell, HowtoLive (London:ChattoandWindus,2010),124.For another,seeH.Lagerlund, SkepticisminPhilosophy (NewYork:Routledge,2020),whosaysthat ‘Socratesfamouslyproclaimedthatheknowsthathedoesnotknowanything’ (2;cf.26).Icitefurther examplesinChapter2.
⁴ Onemightarguethatitimpliesacontradiction,sinceitisanassertion;and,accordingtooneview, ifoneassertsthatp,onetakesoneselftoknowthatp.Butthatviewofassertioniscontroversial;and IsuggestinChapter2thatPlatodoesnotacceptit.
EssaysinAncientEpistemology .GailFine,OxfordUniversityPress(2021).©GailFine. DOI:10.1093/oso/9780198746768.003.0001
At Ap.21b4–5,Socratessays:
ForIamawareofbeingwiseinnothing,greatorsmall. Egôgardêoutemegaoutesmikronsunoidaemautô(i)sophosôn.
Thispassageisn’texplicitlyself-contradictory,foritusestwodifferentcognitive phrases, sunoidaemautô(i) and sophosôn. However,thepassagewouldbeimplicitlyself-contradictoryifthesephrasesweresynonymous,orifto suneidenai somethingimpliedbeing sophos withrespecttoit.⁵
However,toknowsomethingwouldnotusuallybethoughtsufficientforbeing wisewithrespecttoit.⁶ Itwouldnormallybethoughttobemoredifficulttoattain wisdomthantoattainjustanyoldknowledge.IfSocratesclaimstoknowthatheis notwise,thatdoesnotimplyacontradictionunlesstheknowledgehetakes himselftohaveiswisdom.Butthe Apology doesnotsupportthatview.Indeed, itisnotevenclearthat suneidenai isbeingusedforknowledge:thewordcanbe usedforbeingawareofsomethinginawaythatfallsshortofknowingit.In Chapter2,IarguethatSocratesissayingeitherthatheknowsthatheisnotwise (whereknowingsomethingfallsshortofbeingwisewithrespecttoit);orelsethat heisawarethatheisnotwise(wherebeingawareofsomethingfallsshortof knowingit).Neitherclaiminvolvesorimpliesacontradiction.⁷
Onemightarguethatif,in Ap.21b4–5,Socratesclaimstolackallwisdom,he contradictstheoracle’sclaimthatnooneiswiserthanheis(21a)andthatheis wisest(21b).Fortheoracledoesn’tlie(21b).OnesolutionistosaythatSocrates disclaimswisdomonlybeforehearingwhattheoraclesaid;afterhearingwhatit said,hedecidesthatheiswiseafterall.Hedoesn’tcontradicthimself,holding bothpandnotpatthesametime;rather,uponreflectionhechangeshismind.⁸
Idefendadifferentsolution,accordingtowhich,evenafterreflectingonwhat theoraclesaid,Socratescontinuestobelievethathelacksallwisdom.Thisis consistentwithwhattheoraclesays.ForSocratescanbewiserthanotherswithout anyofthembeingwise,justasonepersoncanbewealthierthananotherwithout
⁵ ThisishowRichardKrautunderstandsthepassage.Hetranslates ‘sunoida’ as ‘know,’ andsaysthat ‘youcannotknowthatyouarenotwiseeveninasmallway;fortoknowsomethingistohaveasmall amountofwisdom’ : SocratesandtheState (Princeton:PrincetonUniversityPress,1984),272n.44.
⁶ In Rep.428c–429a,Platotakes sophia tobejustonekindof epistêmê.However,in Meno 96e–100c heusesavarietyofcognitivetermsinterchangeably,including sophia and epistêmê. Tht.145d7–e6 explicitlyidentifies epistêmê and sophia.Despitethisterminologicalvariation,Platooftenseemsto recognizedifferentkinds,orlevels,of epistêmê.
⁷ Eitherway,ontheinterpretationI’vesuggested21b4–5usestwodifferentcognitivetermsfortwo differentcognitiveconditions.GregoryVlastos,bycontrast,arguesthatSocratesnotonlytakesavariety ofcognitivetermstobesynonymousbutalsotakeseachofthemtobeambiguousasbetweentwo sensesof ‘know,’ whichVlastoscalls ‘certainknowledge’ and ‘elencticknowledge.’ Seehis ‘Socrates’ DisavowalofKnowledge,’ PhilosophicalQuarterly 35(1985),1–31.ReprintedinG.Fine(ed.),PlatoI, 64–92,at82–4,91.SomecommentatorsagreewithVlastos’sbasicpointbutfavordifferentsensesof ‘knowledge’ thanhedoes.
⁸ Forthisview,seeKraut, SocratesandtheState,271.
eitherofthembeingwealthy.It’sjustthatSocratescomesclosertobeingwisethan othersdo,justasonepersoncanbeclosertobeingwealthythanothersare. Similarly,Socratescanbewisestbybeingtheonewhoisclosesttobeingwise,just asthehealthiestpersonintheroomneednotbehealthy.It’struethat,after reflectingontheoracle,heclaimstohavehumanwisdom(23ab);buthedoesn’t thinkthatitisgenuinewisdomorevengenuineknowledge.Foritconsistsjustin thefactthat,whenhedoesn’tknowthatp,neitherdoeshethinkheknowsthatp: hediffersfromothersinnothavinganyfalsepretensestoknowledge.Thoughthis iscompatiblewithhisbeingwise,itdoesn’timplythatheiswise.Anditremains hisconsideredview,evenafterconsideringtheoracle’spronouncement,thathe lacksallwisdom.
AlthoughSocratesconsistentlyclaimstolackallwisdom,heclaimstoknow somethings.Forexample,at29bheclaimstoknow(eidenai)amoraltruth.This doesn’tcontradicteitherhisclaimtolackallwisdomortheviewthathuman wisdomisn’tgenuinewisdom.Fortheknowledgehesayshehasin29bisnotpart ofhishumanwisdom.Norishavingitsufficientforhimtobegenuinelywise;itis knowledgeofalessdemandingsort.
IndecidinghowbesttounderstandSocrates’svariousclaimsabouthiscognitivecondition,itisimportanttobearinmindthat,asIhavementioned,Greekhas manycognitiveterms;anditisnotalwaysclearhowtounderstandtheminagiven context.Forexample,wehaveseenthatthereisdisputeaboutwhetherSocrates’ s claim,in21b,to suneidenai somethingimpliesthathetakeshimselftobewise withrespecttoit,ortoknowitinawaythatfallsshortofbeingwisewithrespect toit,ortobeawareofitinawaythatfallsshortofhavingknowledge.Similar questionsariseaboutothercognitiveterms.Forexample,at23bSocratessaysthat he egnôken thathiswisdomisworthless.To gignôskein somethingcanbetoknow it.Butthetermcanalsobeusedforagraspthatfallsshortofknowledge.Howcan wedecidewhichwayitisusedhere?Further,therearedisputesaboutwhat knowledgeis.How,then,doweknowwhenitisappropriatetotranslateagiven Greekwordas ‘knowledge’?
Hereitishelpfultodistinguishthe concept ofknowledgefromparticular conceptions ofit.⁹ Theconceptofknowledgeprovidesanabstractaccountof whatknowledgeis;particularconceptions fillinthisaccountinmoredeterminate ways.Twopeoplecanagreeabouttheconceptofsomethingwhiledisagreeing abouttheright,orbest,conceptionofit.Thedistinctionbetweenaconceptanda
⁹ ThisistheterminologyIuseinChapter5,whereIalsoprovideafullerexplanationofthe distinctionthanIdohere.IdrawthesamedistinctioninChapter2,thoughindifferentterms,speaking insteadoftheconceptofknowledgeandcriteriaforapplyingit.InChapter12,IprovidewhatIcalla workingaccountofsubjectivity;thisroughlycorrespondstoaconceptofit,thoughitislooserand moredisjunctivethanaretheconceptsIsuggestofknowledgeandbelief.InChapter12,Isometimes use ‘concept’ whereIwouldnowuse ‘conception.’ InChapter14,Idistinguishacoreconceptionfrom anoutershell;theseroughlycorrespondto,respectively,aconceptandaconception,asIexplainthose notionshere.
conceptionisanalogoustothedistinctionbetweenajobdescriptionandthe candidatesforit.Thejobdescriptionspecifieswhatcriteriathesuccessfulcandidateshouldsatisfy.Thecandidatespurporttosatisfythem.Someinfactdon’tdo so.Butmorethanonemightdoso;andoneofthemmightsatisfythedescription bestofall.Forexample,ajobdescriptionmightsaythatthesuccessfulcandidate shouldbeskilledatteachingatavarietyoflevelsandshouldalsobeagood administrator.Somecandidatesmightsatisfyoneofthesecriteriabutnotthe other;ifso,theydon’tmakethegrade.Morethanonecandidatemightsatisfyboth criteria,butsomemightdosobetterthanothers.Thejobdescriptionisanalogous toaconcept;thecandidatesareanalogoustoconceptions.
Itaketheconceptofknowledgetobethatofatruth-entailingcognitivecondition thatisappropriatelycognitivelysuperiortomeretruebelief.¹⁰ Particularconceptionsofknowledgesayhowknowledgeisappropriatelycognitivelysuperiortomere truebelief:theysaywhatittakestogetoverthatthreshold.Ifto suneidenai somethingisn’ttruthentailing,orifisn’tappropriatelycognitivelysuperiorto meretruebelief,itisn’tknowledge.Itdoesn’tsatisfythejobdescription.Ifagiven cognitivetermisusedforaconditionthatisbothtruthentailingandalsoappropriatelycognitivelysuperiortomeretruebelief,itisusedforknowledge:itsatisfies thejobdescription,thoughweshouldthenaskhowwellitdoesso.Isthecandidate minimally,orwell,qualified?
Evenwhenagivencognitivetermcan,inagivencontext,beproperlytranslated as ‘knowledge,’ itmightbemisleadingtodosoifitobscuresthefactthat,inthat context,morethanoneGreektermisused,allofwhichcan,insomecontexts, properlybetranslatedas ‘knowledge.’ Further,aswehaveseen,someGreek terms e.g. suneidenai and gignôskein canbeusedbothforknowledgeandfor somethinglessthanthat.Wehavetopaycarefulattentiontothecontextto determinehowthetermsareused.Evenifagiventermisalwaysusedfor knowledge,itmightindicatedifferentlevelsofknowledgeindifferentcontexts. Forexample,twocognitivestatesmightbothcountasgenuineknowledge(not justapproximationstoknowledge),yetonemightbedeeperthantheotherby havingabetterexplanationofwhywhatoneknowsistrue,orabetterjustification forone’sbeliefthatsomethingisso.¹¹Further,differentdialoguesmighthave differentconceptionsofknowledge.
EvenifeachofthepassagesIhavediscussedsofarisinternallyconsistent,and eveniftheyarealsoallcompatiblewithoneanother,otherpassagesmightbe
¹⁰ Thisconceptdoesnotimplythatknowledgeiseitheraspeciesorimplicationofbelief,thoughit iscompatiblewithbothofthoseoptions.
¹¹Tosaythattherearedifferentlevelsofknowledgeisnottosaythat ‘knowledge’ hasdifferent senses.Ifsomethingisalevelofknowledge,itisgenuineknowledge:itisacognitiveconditionthatis appropriatelycognitivelysuperiortomeretruebelief.Buttherecanbedifferentwaysofbeing cognitivelysuperiortomeretruebelief.Idiscusslevelsofknowledge,aswellasofjustificationand explanation,below;seealsoChapter2.
moredifficulttoaccommodate.However,inChapter2,Iarguethatnowherein the Apology doesSocratesclaim,orimply,thatheknowsthatheknowsnothingin awaythatinvolvescontradiction.Everypassagethatmightseemtodosoismore plausiblyreadinadifferentway,eitherasclaimingthatheisaware(inawaythat fallsshortofknowing)thatheisn’twise;orasknowing,inalow-levelway,thathe lacksknowledgeofahigher-levelsort.NotonlydoesSocratesnotclaimtoknow thatheknowsnothinginawaythatisself-contradictory;butneitherdoeshe claimtoknownothing.Tobesure,wehaveseenthat,at21b,heclaimstolackall wisdom.But,aswehavealsoseen,lackingallwisdomdoesnotimplylackingall knowledge.And,aswehaveseen,elsewhereinthe Apology heclaimstoknow somethings thoughwithoutimplyingthatheiswisewithrespecttothem.
Ifmyaccountiscorrect,the Apology implicitlydistinguisheslevelsofknowledge.ForSocratestakeshimselftohavesomeknowledgethatfallsshortof wisdom:hehaslow-levelknowledgebutlackshigh-levelknowledge.Healso implicitlydistinguishesknowledgefrombelief.If,forexample,in21b suneidenai isusedforawarenessthatfallsshortofknowledge,itisarguablyusedformere belief,thatis,forbeliefthatfallsshortofknowledge.(However, suneidenai doesn’t mean ‘merebelief. ’ Aswehaveseen,itcanbeusedinotherways.Wehavetorely onthecontexttotellushowitisusedonagivenoccasion.)Further,at23ab,after consideringwhattheoraclesays,Socrateseventuallyarrivesatasettledandwellreasonedviewaboutwhatitmeantandofwhat,inthelightofthat,hisown cognitiveconditionis.Hehasacquiredwhathetakestobeawell-justified truebelief,thoughnotonethathethinksisjustifiedwellenoughtocountas knowledge.¹²However,the Apology doesn’texplicitlysaywhatknowledge, wisdom,belief,ortruebeliefare.Platoismoreexplicitaboutsomeofthese cognitiveconditionsinthe Meno,whichIdiscussinChapter3.¹³
2.
Inthe Meno,Platoexplicitlydistinguishes epistêmê fromtrueorcorrectbelief (alêthes/orthêdoxa).(At97,healsonotessomesimilaritiesbetweenthem:both implytruth;bothimplybelief;andtheyareequallyreliableguidestoactionona givenoccasion.)Healsodefines epistêmê astruebeliefthatistieddownwith reasoningabouttheexplanation(aitiaslogismos)(98a3–8).Thatis,onehas epistêmê thatpifandonlyifonehasthebeliefthatp,pistrue,andonecan explainwhypistrue.¹⁴
¹²Thisisnottosaythatherejectstheviewthatknowledgeisjustifiedtruebelief.Hemightthink thatknowledgeisjustifiedtruebelief,wherethejustificationhastobeofacertainsort;notanyold justificationwilldo.Idiscussthisviewfurtherbelow.
¹³Chapter2alsodiscussessomeissuesthatarerelevanttoChapter3;seeesp.n.6andn.7.
¹⁴ Forreasonsforthinkingthisisatleastanoutlinedefinitionof epistêmê,seeChapter3,sect.4.
Inthepreviousparagraph,Iused ‘belief ’ totranslate doxa.However,the Meno doesn’tprovideanaccountof doxa;andthereisdisputeaboutwhether doxa,as Platoconceivesofit,countsasbeliefasitisstandardlyunderstoodnowadays.Tobe sure,noteveryonenowadayshastheexactsameviewofbelief.ButonwhatItaketo bethestandardview,whichIshallassumehere,forAtobelievethatpisforAto takeptobetrue(withtheaimofitsbeingtrue:takingittobetrueforthesakeof exploringahypothesisdoesn’tcount).¹⁵ InChapter3,Iassumeratherthanargue that doxa,asPlatousestheterminthe Meno,means ‘belief ’ inthissense;hence Isimplytranslateitas ‘belief.’ ButinChapter5,Iarguethat doxa,asitisusedinthe Phaedo,isbeliefinthissense.InChapters9and10,IbrieflydiscussAristotleon doxa andbelief.InChapter11,IaskwhetherSextususes dogma forbelief.
Onereasontothinkthat doxa,asitisdiscussedinthe Meno,isbeliefinthe senseoftakingtobetrueisthat98adefines epistêmê asaspeciesof doxa.¹⁶ However,thedialoguealsosometimestakes doxa tobemerebelief:thatis,belief thatfallsshortofknowledge.Forexample,in97e–98atrue doxai arecomparedto Daedalus’ statueswhich,unlike epistêmê,areunstable.When doxa ismerebelief, epistêmê isnotaspeciesof,anddoesn’timply, doxa;similarly,knowledgeisnota speciesof,anddoesn’timply,merebelief.ThoughPlatouses doxa bothforgeneric takingtobetrueandformerebelief,hedoesn’tusethetermintwodifferent
¹
⁵ Forthisviewofbelief,seee.g.S.Blackburn, sv ‘belief,’ TheOxfordDictionaryofPhilosophy, (Oxford:OxfordUniversityPress,1996);E.Schwitzgebel, ‘Belief,’ inZalta,E.(ed.), Stanford EncyclopediaofPhilosophy,(Summer2015Edition);B.Williams, ‘DecidingtoBelieve,’ inhis ProblemsoftheSelf (Cambridge:CambridgeUniversityPress,1973);R.Stalnaker, Inquiry (Cambridge:MITPress,1984);D.Velleman, ‘ThePossibilityofPracticalReason,’ Ethics 106(1996), 694–726at706–7;andhis ‘OntheAimofBelief,’ inhis ThePossibilityofPracticalReason (Oxford: OxfordUniversityPress,2000),244–81.Thoughtheviewthatbeliefistakingtobetrueiswidespread, noteveryoneunderstandstakingtobetrueinexactlythesameway.IdiscussthisinChapter11,where IalsoexplainhowIunderstandthenotion.Foradifferentviewofbelief,onwhichitneednotinvolve takingtobetrue,seeM.Frede, ‘TheSkeptic’sBeliefs,’ inhis EssaysinAncientPhilosophy (Minneapolis: UniversityofMinnesotaPress,1987),ch.10;andhis ‘TwoKindsofAssentandthePossibilityof Knowledge,’ alsoinhis Essays,ch.11.IdiscusshisalternativeinChapter11. ¹⁶ Thisissometimesdoubted.Forexample,in ‘ThreePlatonistInterpretationsofthe Theaetetus, ’ in C.GillandM.M.McCabe(eds.), FormandArgumentinLatePlato (Oxford:ClarendonPress,1996), ch.6,at93,D.SedleynotesthatPlatosaysthattruebeliefsbecome(gignontai)knowledge(98a6),which mightmeanthattheythenceasetobetruebeliefsjustas,whenachildbecomesanadult,sheceasesto beachild.(ThoughSedleymentionsthisinterpretation,hedoesnotcommithimselftoit.Contrast J.MossandW.Schwab, ‘TheBirthofBelief,’ JournaloftheHistoryofPhilosophy 57(2019),1–32,at9.) But gignesthai canbeusednon-temporally,asitisine.g. Meno 82d2.Ifitisusednon-temporallyin 98a6,thepointisthatitturnsoutthatknowledgeistruebeliefthatisappropriatelybound:thatisthe conclusionoftheirreasoning.If,however, gignesthai isusedtemporally,thepointmightjustbethat truebeliefbecomesknowledgefromnothavingbeenknowledge;itneednotindicatethatthetruebelief ceasestobeatruebelief,thoughofcourseitceasestobea mere truebelief.
Onemightsaythatratherthantakingknowledgetobeaspeciesoftruebelief,Platomeansjustthat truebeliefisacomponentofknowledge.WhenwesaythatwaterisH₂0,wedon’tmeanthatwaterisa speciesofhydrogen;wemeanthathydrogenisacomponentofwater.Butevenonthisview,itremains thecasethatthesortoftruebeliefthatisinvolvedinthedefinitionofknowledgeisgenerictakingtobe true;meretruebeliefisnotatissue.Ishallcontinuetospeakofknowledgebeingaspeciesoftruebelief; buttheviewthattruebeliefisacomponentofknowledgewoulddoformypurposes,aswouldtheview thatknowledgeimpliestruebelief.
senses.¹⁷ Rather,heusesittomeantakingsomethingtobetrue,butsometimesa givencontextmakesitclearthatonlysomebeliefsareatissue.
Let’snowturnto epistêmê.Itissometimesthoughtthat,indefining epistêmê in 98a,Platoisdefiningknowledgeasjustifiedtruebelief.¹⁸ Iasklaterwhether epistêmê isknowledge.Butlet’sask firstwhetherhedefines epistêmê asjustified truebelief.Hisdefinitionmentions part ofthejustifiedtruebeliefaccountof knowledge:itsaysthat epistêmê isaspeciesoftruebelief.Butishavingan aitias logismos havingajustification?AccordingtoMylesBurnyeat,itisnot: ‘the Meno’ s leadingconditiononknowledge, aitiaslogismos (98a)isGreekforworkingoutthe explanationofsomething, not forassemblingajustificationforbelievingit,which theslavealreadyhasatastagewhenPlatodenieshehasknowledge(85c).’¹⁹ Burnyeatmakestwopoints.Oneisthat aitiaslogismos doesn’tmean ‘justification.’
TheotheristhatPlatodoesn’ttakeknowledge(epistêmê)tobejustifiedtruebelief.
IagreewithBurnyeat’ s firstpoint.That,however,leavesopenthepossibility thatPlatothinksthathavingthesortofjustificationthat’sneededforone’sbeliefs tocountas epistêmê consistsinhavingan aitiaslogismos;and,inChapter3, IarguethatthatisPlato’sview.
AstoBurnyeat’ssecondpoint,it’struethat,by85c,theslavehas some justificationforhisbeliefsbutissaidtolack epistêmê.Butthisshowsonlythat Platodoesn’tthinkit’ssufficientforhaving epistêmê thatonehaveatruebelief thatisjustifiedtoanyolddegreeorinanyoldway.Rather,hethinksonehas epistêmê whenonehasajustifiedtruebelief,wherethejustificationinvolvesbeing abletoexplainwhywhatonebelievesistrue.Notalljustificationisonapar;some isbetterthanothers.Intaking epistêmê torequireexplanation,Platoisn’tbypassingjustification;heisspecifyingthesortofjustificationthat’sneededfor epistêmê.²⁰
Wecannowaskwhether epistêmê isknowledge.Onceagain,thedistinction betweentheconceptofknowledgeandparticularconceptionsofitprovesuseful.²¹
¹⁷ Pace MossandSchwab, ‘TheBirthofBelief,’ 8.Insayingthatthe Meno uses doxa bothforthe genusandformerebelief,IretractaclaimmadeinChapter3n.32.
¹
⁸ Seee.g.D.M.Armstrong, Belief,Truth,andKnowledge (Cambridge:CambridgeUniversityPress, 1973),137.
¹⁹ M.F.Burnyeat, ‘SocratesandtheJury:ParadoxesinPlato’sDistinctionbetweenKnowledgeand TrueBelief,’ ProceedingsoftheAristotelianSociety,suppl.54(1980),173–91,at187.SeealsoM.F. Burnyeat, ‘Introduction’ to TheTheaetetusofPlato (Indianapolis:Hackett,1990),271.
²⁰ Fortheviewthatitisaddingacertainsortofjustification notjustanyoldjustification totrue beliefthatyieldsknowledge,seee.g.J.Cargile, ‘OnNearKnowledge,’ Analysis 31(1971),145–52,at 145–6;andR.Fogelin, PyrrhonianReflectionsonKnowledgeandJustification (NewYork:Oxford UniversityPress,1994),ch.1.Inspeakingofaddingsomethingtotruebelief,Idonotmeantoendorse whatissometimescalledtheadditivemodelofknowledge,forwhich,seeChapter3n.39.
²¹OrsoIthink.Butnoteveryone findsthedistinctionbetweenconceptandconceptionhelpful. Indeed,in ‘ExplanationintheEpistemologyofthe Meno, ’ OxfordStudiesinAncientPhilosophy 48 (2015),1–36,W.Schwabisskepticalaboutwhetherthereevenisaconceptofknowledge(2);hencehe focusesinsteadoncomparingSocrates’saccountof epistêmê inthe Meno withcontemporary(2)(or standardcontemporary:25)accountsofknowledge.HeconcedesthatSocrates’saccountof epistêmê in the Meno islikesomecontemporaryaccountsofknowledge;buthearguesthatsuchaccountsaren’t
AccordingtotheconceptofknowledgeIhavesuggested,knowledgeisatruthentailingcognitiveconditionthatisappropriatelycognitivelysuperiortomere truebelief. Epistêmê ,asPlatodefinesit,countsasknowledgewhenknowledgeisso described.Foritimpliestruthandisalsocognitivelysuperiortomeretruebelief, sinceitalsoinvolvesjustification,wheretherelevantjustificationconsistsin explainingwhywhatoneknowsistrue.Tobesure, epistêmê doesn’t have tobe justifiedtruebelieftobeknowledge;that’sjustone,controversial,conceptionof knowledge.²²Butitisaconceptionofknowledge:it fitsthejobdescriptionevenif, intheend,onefavorsadifferentcandidate.
Platotakes epistêmê tobeknowledgeassuch,notjustonetypeofknowledge:he thinksthatitistheonlycandidatethatsatisfiesthejobdescription.Forhethinks thatanycognitiveconditionthatfallsshortof epistêmê isatbestmeretruebelief.²³ Hence,inhisview,onecanhave epistêmê,andsoknowledge,thatpistrueonlyif onecanexplainwhypistrue.Bycontrast,incontemporaryepistemologyitis oftenassumedthatonecanknowthatpistruewithoutknowingwhyitistrue. Thisindicatesadifferenceinconceptions,butnotintheconcept,ofknowledge. Weshouldnotbesurprisedthattherearedifferentviewsofhowknowledgeis appropriatelycognitivelysuperiortomeretruebelief.
WehaveseenthatPlatothinksthattoknowthatpistrue,oneneedstobeable toexplainwhyitistrue.Ononeview,hethinksthat,todothis,onemusthavea deepsynopticgraspofasortfewcanhave.Analternativeisthat,thoughhe requiresthissortofexplanationinsomecases(inparticular,forknowledgeof forms),hedoesn’t(inthe Meno)requireitforall epistêmê.Forexample,hesays thatonecanknowwhoMenois(71b)andhowtogettoLarissa(97a).Todoso,an explanationisneeded;butitneednotbeonethatinvolvesadeep,synopticgrasp thatveryfewpeoplecanhave.Nordoesitrequireagraspofforms:thoughforms areexplanatory,therearealsoexplanationsthatdonotinvokeforms.Hencewe neednot(asissometimesdone)dismisstheseexamplesasmereanalogies;they areliteralexamplesofthingsonecanknow.
Inrequiringexplanationforallknowledge,Platomakesknowledgemore difficulttoacquirethansome thoughnotall donowadays.Butinallowing
‘standard.’ If epistêmê,asPlatodescribesit,islikesomecontemporaryaccountsofknowledge,that’ s sufficient(thoughnotnecessary)foritsbeingknowledge.
²²Vlastos,forexample,saysthathardlyanyonenowadaysacceptsthejustified-true-beliefaccount ofknowledge: ‘Socrates’ DisavowalofKnowledge,’ n.25.
²³Nordoesthe Meno countenanceacognitiveconditionthatishigherlevelthan epistêmê (whichis consistentwiththedialoguecountenancingdifferentlevelsof epistêmê,asitseemstodo).Ofcourse, Platodidn’thavethe dedicto beliefthat epistêmê isknowledge:hedidn’tknowEnglish.Mypointisthat givenwhathesaysaboutvariouscognitiveconditions,itisreasonabletodescribehisviewastheview thattheonlyknowledgethereisis epistêmê,andthatanyothercognitiveconditionisatbestmeretrue belief.
levelsofexplanation,andsoofknowledge,hedoesn’tmake all knowledge as difficulttoacquireasheissometimesthoughttodo.²⁴
ItissometimesthoughtthatthefactthatPlatorequiresexplanationfor epistêmê showsthat epistêmê isunderstanding.Anditissometimesthoughtthat understandingiseitherjustone(especiallyhigh-level)typeofknowledgeorelseit isdifferentfromknowledgealtogether.Howwedecideaboutthisdependsnot onlyonhowwethinkPlatoconceivesof epistêmê,butalsoonhowweconceiveof knowledgeandunderstanding.WhitneySchwab,forexample,arguesthat epistêmê isunderstandingbutnotknowledge,whereasLindsayJudsontakes understandingtobeatypeofknowledge.Theyagreethat epistêmê isunderstanding;buttheydisagreeaboutwhetherunderstandingisatypeofknowledge.²⁵ If onethinksthathavinganexplanationofwhywhatoneknowsistrueissufficient forunderstanding,then epistêmê isunderstanding.Platowouldinthiscasehold theviewthatunderstandingistheonlysortofknowledgethereis.However,aswe haveseen,hethinkstherearelevelsofexplanation,andsoofunderstanding.Some explanationsareeasiertocomebythanothersare;andsosome epistêmê iseasier tocomebythanlinkingittounderstandingmightsuggest.
Platoisoftenthoughttoholdtheso-calledTwoWorldsTheory(TW),accordingtooneversionofwhichtherecanbeknowledgeonlyofforms,andbeliefonly aboutsensibles.²⁶ Whatevermaybetrueelsewhere,the Meno rejectsthisview.For, aswehaveseen,thedialogueallowsknowledgeofsensibles:wecanknowwho Menois(71b),andhowtogettoLarissa(97a).Italsoallowsbeliefaboutforms.For Socratessaysvariousthingsabout,forexample,theformofvirtue e.g.thatitis oneandthethesameinallcases(72b–d) buthenotonlydisclaimsknowledgeof whatitisbutalsothinksonemustknowwhatsomethingisinordertoknow anythingatallaboutit(71).Itisreasonabletothinkthatheisexpressinghisbeliefs aboutwhatitis.Herebeliefismerebelief;and,ofcourse,onecan’t,att1,both knowandmerelybelievethatp.However,onecan firstmerelybelievethatpand thencometoknowthatp.²⁷ Further,aswehavealsoseen,Platodefinesknowledge
²⁴ FortheviewthatPlatoallowslevelsofexplanation,andsolevelsofknowledge,seealso R.Wedgwood, ‘Plato’sTheoryofKnowledge,’ inD.Brink,S.Meyer,andC.Shields(eds.), Virtue, Happiness,Knowledge:ThemesfromtheWorkofGailFineandTerenceIrwin (Oxford:Oxford UniversityPress,2018),33–56.Ifknowledgerequiresexplanation,differentlevelsofexplanationare sufficientfordifferentlevelsofknowledge.However,evenifonedoesn’tthinkknowledgerequires explanation,onecouldstillpositlevelsofknowledgeby,forexample,distinguishinglevelsofjustification:amongjustificationsthataresufficientforturningtruebeliefintoknowledge,somemightbe higherlevelthanothersby,forexample,beingmoresynoptic.
²
⁵ Schwab, ‘Explanation’;L.Judson, ‘The Meno, ’ inG.Fine(ed.), TheOxfordHandbookofPlato (NewYork:OxfordUniversityPress,2019;2nded.),161–81.
²⁶ ForfurtherdiscussionofTW,seeChapters4and10.DiscussionofTWtypicallyfocuseson sensiblesandforms,andthatwillbemyfocushere.However,Platoalsocountenancesothersortsof entities e.g.soulsandgod.
²
⁷ HenceSedley’ssuggestionabout98a(seen.16above)doesnotsupportTW,sinceitleavesopen thepossibilitythattherecanbeknowledgeandbeliefaboutthesameobjects;itrulesoutjustone person ’shavingbothknowledgeandtruebeliefaboutsomethingatoneandthesametime.
asaspeciesofbelief,wherebeliefisgenerictakingtobetrue.Ifknowledgeisa speciesofbelief,then,wheneveroneknowssomething,onetherebybelievesit. HencePlatorejectsTWbothwhenbeliefismerebeliefandwhenitisthegenus.
3.
WehaveseenthatPlatorejectsTWinthe Meno.Butitisoftenthoughtthathe acceptsitinsuchmiddledialoguesasthe Phaedo and Republic.²⁸ Ihaveargued elsewhere,however,that Republic 5leavesopenthepossibilitythatonecan havebeliefsaboutformsaswellasknowledgeofsensibles.Ihavealsoargued elsewherethat,inthe Republic,Socratessaysthathehasmerebeliefaboutthe formofthegood(506c)andthatthephilosopherwhoreturnstothecavewillhave knowledgeofthethingsthere,thatis,ofsensibles(520c).²⁹
Butwhataboutthe Phaedo,whichisgenerallythoughttohavebeenwritten afterthe Meno butbeforethe Republic?InChapter4,Iarguethatthe Phaedo also rejectsTW.Forexample,Socratesdiscussesa doxa thatphilosophershave;their doxa countenancesforms.Andindiscussingthetheoryofrecollection,hecountenances epistêmê ofe.g.one’sfriendandhiscloak,andofalyre.Onemight dismisstheseremarksaslooseor,inthecaseof epistêmê,asmereanalogies.Butit isnotjustthathementions doxa aboutformsand epistêmê ofsensibles.Itisalso that,asweshallsee,hismoretheoreticalremarksmakeitreasonabletotakehis examplesatfacevalue.
Onemightarguethat,thoughthe Phaedo countenances doxa aboutforms and epistêmê aboutsensibles,itdoesnotcountenance beliefs aboutformsor knowledge ofsensibles,since doxa isnotbelief(atleast,notasweunderstandit nowadays),and epistêmê isnotknowledge(atleast,notasweunderstand itnowadays).IaddressthisissueinChapter5.Indoingso,Iagaindistinguish theconceptsofknowledgeandbelieffromconceptionsofthem.Ihavesuggested thattheconceptofbeliefistakingsomethingtobetrue,withtheaimofitsbeing true.IthasbeenarguedthatPlato,inatleastsomedialogues,doesnotuse doxa for generictakingtobetrue,butjustformerebeliefor,in Republic 10butperhaps alsoelsewhere,soastoincludewhatwewouldcallnon-doxasticappearances,
In ‘TheBirthofBelief,’ onereasonMossandSchwabgivefortheirviewthat doxa isnotbelief(which theytaketobegenerictakingtobetrue)isthatithasitsownspecialobjects,whereasgenerictakingto betruedoesnot;inparticular,theyarguethatPlatodoesnotcountenance doxa aboutintelligibles.In the Meno,however, doxa doesn’thaveitsownspecialobjects.InChapter4,Iarguethatthesameistrue ofthe Phaedo,andIhavearguedelsewherethatthesameistrueofthe Republic
²⁸ Fortheviewthatthe Phaedo holdsTW,see,forexample,D.Gallop, Plato:Phaedo (Oxford: ClarendonPress,1975).Theviewthatthe Republic doessoiswidespread.
²⁹ Fine, ‘KnowledgeandFine,Beliefin Republic V’ and ‘KnowledgeandBeliefin Republic V–VII,’ bothreprintedinmy PlatoonKnowledgeandForms (Oxford:ClarendonPress,2003),chs.3and4, respectively.
aswhentheoarappearsbentinwatertome,butIdonottakeittobebent.Ithas alsobeenarguedthat,insomedialogues,Platotakes doxa tobenarrowerthanwe takebelieftobe,insofarasherestrictsittoclaimsthatareacceptedasaresultof activereflection.³⁰
Whatevermaybetrueelsewhere,theseviewsdonot fitthe Phaedo.Noneof the doxai mentionedinthe Phaedo isanon-doxasticappearance.Andthough someofthe doxai itmentionsarecarefullyconsidered,notallofthemare.For example,thephilosophers’ doxa thatisdescribedat66b1–67b5iscarefully considered.Butthebonesandsinews ’ doxa thatSocratesshould flee(99a2)is notcarefullyconsidered.Nordoesthe Phaedo use doxa tomean ‘merebelief.’ In somecases doxa isusedneutrally,withoutitsbeingimpliedthatthe doxa isamere belief.Forexample,whenCebessaysthathewantstohearSocrates’ s doxa about thesoul(70b9),hedoesnotmeantoimplythatSocrateshasamerebeliefhere (thoughSocrateswouldnodoubtdisclaimknowledgehere).Hewantstoknow whatSocratesthinks,whathetakestobetrue.Butevenif doxa werealwaysused withtheconversationalimplicaturethatthe doxa atissueisamerebelief,that wouldn’timplythat doxameans ‘merebelief ’:weneedtodistinguishthe meaning ofthetermfromits extension or reference inagivencontext.
Let’snowturnto epistêmê.AccordingtoLloydGerson,theterm ‘epistêmê’ in the Phaedo ‘appearstobeusedsolooselythatitisvirtuallyequivalentto “cognition. ”’ Forexampleitisusedforobjectsofsense-perceptionaswellasfor forms(cf.73c8,d3,74b2).Aself-consciousrestrictionintheuseofthetermfor epistêmê suchaswe findinthe Republic isnotinevidencehere.’³¹Aswehaveseen, somecommentatorsthink epistêmê isn’tknowledgebecauseitismoredemanding thanknowledgeis.Gerson,however,thinksthat,inthe Phaedo , epistêmê istoo extensivetocountasknowledge.Oneofhisreasonsisthatthedialogueallows epistêmê ofsensibles.Ihaveargued,however,thatPlatoallows epistêmê of sensibles(wherethatcountsasknowledge)bothinthe Meno,whichprecedes the Phaedo,andinthe Republic,whichfollowsit.Soweshouldnotbesurprisedif healsodoessointhe Phaedo .
³⁰ Aswehaveseen,in ‘TheBirthofBelief ’,MossandSchwabarguethat,atleastinsomedialogues, Platouses doxa notforgenerictakingtobetruebut,atbest,formerebelief,andsomorenarrowlythan weuse ‘belief.’ Theyalsomention,butdonotdiscuss,theviewthatPlatosometimesuses doxa soasto includewhatwewouldcallnon-doxasticappearances.In ‘Plato’sAppearance-AssentAccountof Belief,’ ProceedingsoftheAristotelianSociety 114(2014),213–38,J.Mossarguesthat,atleastin Rep 10,Platouses doxa soastoincludenon-doxasticappearances.(Cf.H.Lorenz, TheBruteWithin: AppetitiveDesireinPlatoandAristotle (Oxford:OxfordUniversityPress,2006).)Shealsoarguesthat insomelaterdialoguesPlatorestricts doxa toclaimsthatareacceptedasaresultofactivereflection, andsomorenarrowlythanweuse ‘belief.’ Forthislatterview,seealsoS.Broadie, ‘TheKnowledge Unacknowledgedinthe Theaetetus, ’ OxfordStudiesinAncientPhilosophy 51(2016),87–117,though shesometimesqualifiesherviewwith ‘typically’ (e.g.92,96).
³¹L.Gerson, KnowingPersons:AStudyinPlato (Oxford:OxfordUniversityPress,2003),66.See alsohis ‘TheRecollectionArgumentRevisited,’ inM.McPherran(ed.), Recognition, Remembranceand Reality, Apeiron suppl.vol.32(1999),1–15.
Ialsoprovideanaccountofhow epistêmê isconceivedinthe Phaedo onwhich itisknowledge:thatis,ittreats epistêmê asatruth-entailingcognitivecondition thatisappropriatelycognitivelysuperiortomeretruebelief.For epistêmê,asitis conceivedinthe Phaedo ,requiresatruth-entailing logos (e.g.76b)aswellasa certaindegreeoffamiliarity;neitheroftheseisrequiredfor doxa.
Itisnotclearwhetherthesortof logos thatthe Phaedo requiresfor epistêmê mustbeexplanatory.Ifitmustbe,thenall epistêmê,inthe Phaedo asinthe Meno, requiresexplanation;but,alsoasinthe Meno,therearelevelsofexplanation,and notallexplanationrequiresagraspofforms.If,however,the logos needn’talways beexplanatory,then,inthe Phaedo,onecansometimeshave epistêmê that somethingissoevenifonecan’texplainwhyitisso.Nonetheless,all epistêmê goesbeyondmeretruebelief,evenifnotallofitrequiresexplanation.Onthis reading,the Phaedo differsfromthe Meno which,aswehaveseen,requires explanationfor epistêmê
InChapter6,Icontinuemyexplorationofthe Phaedo’sepistemology,focusing onthetheoryofrecollection.Plato’sargumentfortheclaimthatwhatwethinkof aslearningisreallyjustrecollectingwhatweonceknewhasbeenmuchmaligned. Forexample,ithasbeenarguedthatitisinvalid,thatitiscircular,andthatitbegs thequestion.IdefendPlato’sargumentagainstthesecharges:whichisnottosay thattheargumentissound.
Thetheoryofrecollectionisgenerallythoughttopositinnateknowledge.Ihave arguedelsewherethatthe Meno,whichalsopositsthe,ora,theoryofrecollection, doesnotcountenanceinnateknowledge.³²InChapter6ofthepresentvolume, Iarguethatthe Phaedo doesnotdosoeither.Sofarfromdoingso,itrejectsits existence.Indeed,Plato’sargumentforrecollectiondependsonrejectingit.
InChapter7,Idiscussafurtheraspectofthe Phaedo’sepistemology,comparingitsaccountofperceptionwithacelebratedpassageonperceptionin Theaetetus 184–6.It’sdebatedwhether,inthelatterpassage,Platotakesperceptiontobepropositional,³³ortofallbelowthatthreshold.Idefendthelatterview. Someofthosewhofavorthatviewof Theaetetus 184–6thinkthatthemiddle dialogues,bycontrast,takeperceptiontobepropositional.Idefendadifferent view.Althoughthe Phaedo agreeswith Theaetetus 184–6atanabstractlevel, insofarasbothdialoguesarguethatperceptionisnotknowledge,theygive
³²SeeG.Fine, ThePossibilityofInquiry:Meno’sParadoxfromSocratestoSextus (Oxford:Oxford UniversityPress,2014),ch.5.Onpp.172–3,Iarguethatthe Phaedo alsorejectsinnateknowledge;but mydiscussionisverybrief.InChapter6ofthepresentvolume,Iexpandit.
³³Inaskingwhetherperceptionispropositional,Iamaskingjustwhethereverycaseofperceptionis acaseofperceivingthatsomethingisso.Iamnotassuminganyparticularanalysisofwhat propositionsare.Onemightarguethatevenifperceptionisn’tpropositional,itcouldbeconceptual. However,onmyviewand,Ithink,onPlato’s,onehastheconceptofFifandonlyifonegrasps propositions;onehastheconceptofcat,forexample,ifandonlyifonegraspsthatcatsarethusandso. SoifPlatotakesperceptiontobenon-propositional,healsotakesittobenon-conceptual.However, onecanhavetheconceptofF,andsograspsomepropositionsaboutF,withoutbeingabletoanswer the ‘WhatisF?’ questionaboutF.
differentreasons.In Tht.184–6,thepointisthatknowledgeispropositional, whereasperceptionisnot.Henceperceptionfallsshortnotonlyofknowledgebut alsoofbelief;forbelief,likeknowledge,ispropositional.Inthe Phaedo,thepointis thatperceptioncan’tgraspanytruths aboutforms,butwisdom(phronêsis) requiressuchagrasp.Thatviewiscompatiblewithperception’sbeingpropositional,foritleavesopenthepossibilitythatperceptiongrasps other propositions e.g.aboutsensibles.Butwhatthe Phaedo saysisalsocompatiblewiththeviewthat perceptionisnotpropositional.Unlike Theaetetus 184–6,the Phaedo takesnostand onthisissue.Perhapsthat’sbecauseitsconcernslieelsewhere:inshowingthatwe needtoengageinnon-perceptualreasoningifwearetoacquirewisdom,whichis justonesortofknowledge,anespeciallyhigh-levelsort.Forthispurpose,itdoesn’t matterwhetherperceptionispropositional.Ofcourse,perceptual reasoning reasoningbasedonperception ispropositional.Butthe Phaedo distinguishes perceptualreasoningfromperception.Althoughtheformerispropositional,it doesnotfollowthatthelatteris;nordoesitfollowthatitisnot.
4.
InChapter8,IturnfromPlatoto[Plato].HereIexploreanunjustlyneglected spuriousdialogue,the Sisyphus.Socratesbeginsbyproposinganaccountof deliberation.Sisyphuschallengesthisaccountandsuggestsadifferentone, accordingtowhichdeliberationisatypeofinquiry.Socratesproceedstochallenge thatview.Healsochallengestheviewthattherecouldbesuchathingas deliberation,ifdeliberationrequiresknowledge;forthesortofknowledgeit wouldrequireisunattainable.If,however,deliberationdoesnotrequireknowledge,deliberationismereguesswork.Butdeliberationisnotmereguesswork.So deliberationisimpossiblewhetherornotitinvolvesknowledge.
ThisParadoxofDeliberationneatlyparallelsMeno’sParadox(or,asitis sometimescalled,theParadoxofInquiry)asitisdiscussedinPlato’ s Meno.³⁴ AccordingtotheParadoxofDeliberation,therearejusttwooptions:deliberation eitherinvolvesknowledgeoritismereguesswork;eitherway,deliberationis impossible.Thesetwooptionsdonotseemtobeexhaustive.What,forexample, abouttruebelief?Sometruebeliefsareneitherknowledgenormereguesswork. Perhapsifwehaveandrelyonthem,wecandeliberate.AccordingtoMeno’ s Paradox,oneeitherknowsordoesnotknowwhatoneisinquiringinto;either way,inquiryisimpossible.Now,whereasknowledgeandguesswork(theoptions offeredinthe Sisyphus)arenotexhaustiveoptions,knowingandnotknowingare exhaustive.Orsoitmightseem.However,itissometimesthoughtthatMeno’ s
³⁴ Seeesp.80d–e.IdiscussMeno’sParadoxindetailin ThePossibilityofInquiry.
Paradoxtakesknowledgetobecompleteknowledge,andnotknowingtobe completeignorance;andthesearenotexhaustiveoptions.Ifso,wecanreplyto Meno’sParadoxbynotingthattherearecognitivestatesthatfallshortofcomplete knowledgebutaremorerobustthanignorance:forexample,truebeliefand knowledgethatislessthancomplete.Ifonehasandreliesononeofthem,one caninquire.ButitisnotclearthatMeno’sParadoxtakesknowledgetobe completeknowledge,andnotknowingtobecompleteignorance.Itmightinstead takeknowingandnotknowingtobeexhaustiveoptions.Inthiscasewecanreply thattherearedifferentwaysofknowingandofnotknowing,andsomeofthem allowinquiry.Forexample,whetheroneistotallyignorantorhasmeretruebelief, onelacksknowledge.Althoughonecan’tinquireifoneistotallyignorant,onecan inquiresolongasonehasandreliesonrelevanttruebeliefs.Thestructuralparallel betweentheParadoxofDeliberationandMeno’sParadoxshouldleadusto considerwhethertheycanbedisarmedinparallelways.
Aristotleagreeswiththe Sisyphus thatdeliberationisatypeofinquiry;andhe thinksdeliberationispossible.³⁵ So,justashethinksthatMeno’sParadoxcanbe answered,³⁶ hethinksthattheParadoxofDeliberationcanbeanswered.Itis unclearwhenthe Sisyphus waswritten;butitisreasonabletothinkitwaswritten beforethe NicomacheanEthics.³⁷ ItdoesnotfollowthatAristotleisreplyingtoit; nonetheless,someoftheconnectionsarequiteinteresting.Wheneverthedialogue waswritten,itisafascinatingbridgebetweenthe Meno andAristotle’sethicsand epistemology.
5.
InPartII,IconsiderAristotle’sepistemology,focusingonissuesexploredinPart I.InChapter9,Iaskhowheconceivesof epistêmê inthe PosteriorAnalytics (APo.) Inparticular,isitknowledgeand,ifso,isitknowledgeassuchorjustakindof knowledge?Inconsideringthisquestion,IcompareAristotle’saccountof epistêmê inthe PosteriorAnalytics withPlato’saccountofitinthe Meno.Aswehaveseen, in Meno 98aPlatosaysthat epistêmê istruebeliefboundwithan aitiaslogismos. Thatis,Ahas epistêmê thatpjustincasepistrue,Abelievesthatp,andAcan explainwhypistrue.Thisisadefinitionof epistêmê assuch,notofjustone kindof epistêmê.Itisalsoadefinitionofknowledgeassuch.Thatis,itistheonly
³⁵ See EN 1142a.FordiscussionofAristotle’sviewthatdeliberationisatypeofinquiry,see K.Nielsen, ‘DeliberationasInquiry,’ PhilosophicalReview 120(2011),383–421.
³⁶ IdiscussAristotle’sreplytoMeno’sParadoxin ThePossibilityofInquiry,ch.6.
³⁷ C.W.Müller, DieKurzdialogderAppendixPlatonica (Munich:WilhelmFink,1975),94–104,and D.S.Hutchinson, ‘Introductionto Sisyphus, ’ inJ.M.CooperandD.S.Hutchinson(eds.), Plato: CompleteWorks (Indianapolis:Hackett,1997),1707,plausiblysuggestthatthe Sisyphus waswrittenin themid-fourthcentury;seeChapter8.
truth-entailingcognitiveconditionPlatocountenancesthatgoesbeyond,andis appropriatelycognitivelysuperiorto,meretruebelief.Henceheimposesmore demandingconditionsonknowledgethaniscommonnowadays,sinceitisnot usuallythoughtnowadaysthat,toknowthatpistrue,onemustbeabletoexplain whypistrue.(However,aswehavealsoseen,itisnotclearthathetakesall explanationstobedifficulttocomeby.)Thismaynotbethemostfamiliar conceptionofknowledgenowadays.Bethatasitmay,itisaconceptionof knowledge:it fitsthejobdescription.
ButwhataboutAristotle?In APo.1.2,hesays:³⁸
Wethinkwe epistasthai somethingwithoutqualification(haplôs)(andnotinthe sophistical,incidentalway)wheneverwethinkwe ginôskein theexplanation (aitia)becauseofwhichthethingis, ginôskein thatitistheexplanationofthat thing,and ginôskein thatitcannotbeotherwise.
Wemayparaphrasethisasfollows:A epistataihaplôs x=df.A ginôskei y,whichis theexplanationofx;A ginôskei thatyistheexplanationofx;andA ginôskei thatx isnecessary.³⁹
haplôs (withoutqualification, simpliciter)mightindicatethatAristotleisdefiningjustanidealkindof epistêmê,onethatleavesroomforothersortsof epistêmê thatfallshortoftheideal.⁴⁰ Alternatively, haplôs mightindicatethatheisdefining epistêmê assuch:ifsomethingfallsshortof epistêmê sodefined,itisn’t epistêmê, period,but,atbest,itapproximatestoit.⁴¹Sometimes,butnotalways,beingan idealFandbeinganFdiverge.Forexample,alinethatisn’tperfectly,ideally straight,isn’treallystraight;itjustapproximatestobeingstraight.Bycontrast,an idealparenthasallthevirtuesthatarepropertobeingaparent.Butonecanbea
³⁸ Thereisdisputeaboutwhether1.2offersadefinition,orsomethinglessthanthat.Ishallassume, withBarnes,thatAristotleisofferingadefinition.SeeJ.Barnes, Aristotle:PosteriorAnalytics (Oxford: ClarendonPress,1975(1stedition=B1),96;1994(2ndedition=B2)),90.Aristotledefines,not epistêmê,but epistasthai:heusestheverb,notthenoun.Althoughit’sbeenarguedthattheverbismore specializedthanthenoun,Idon’tthinkthataffectsanythingIsayhere;andIshallmovefreely betweenthem.
³⁹ Ileavetoonesidethequestionofwhatnecessityamountstohere.Foroneaccount,see M.Peramatzis, ‘AristotleonKnowledgeandBelief: PosteriorAnalytics 1.33,’ inG.Salmieri(ed.), KnowingandComingtoKnow:EssaysonAristotle’sEpistemology (Pittsburgh:PittsburghUniversity Press,forthcoming).ThankstoPeramatzisforlettingmecitethispaperinadvanceofitspublication.
⁴⁰ ThissometimesseemstobeBurnyeat’sview:seeM.F.Burnyeat, ‘AristotleonUnderstanding Knowledge,’ inE.Berti(ed.), AristotleonScience:ThePosteriorAnalytics, ProceedingsoftheVIII SymposiumAristotelicum (Padua:Antenore,1981),97–139,ate.g.100–1.(IdiscussBurnyeatmore fullyinChapter9;seealsoChapter2,n.6andn.7.)SeealsoR.Pasnau, AfterCertainty (Oxford:Oxford UniversityPress,2017),3–8,142–5.
⁴¹ThismaybeBarnes’sview:seeB1,90,149;cf.B2,89–92.InChapter9,Ialsoconsiderathird interpretationof haplôs,whichIleavetoonesidehere. Nous issometimestakentobecognitively superiorto epistêmê asdefinedin1.2.Aristotlevacillates,atleastterminologically,astowhether nous is atypeof epistêmê.71b6–7seemstoleaveopenthepossibilitythatitis.Butaccordingto2.19itisn’t.In EN 6.7Aristotlesaysthat sophia is nous plus epistêmê.Ishallleavethisissuetoonesidehere,sincemy focusison epistêmê andcognitiveconditionsthatfallshortofit.
