Draftof15October2015

CHAPTER4:

CONTEXTUALISMANDSKEPTICISM:THEDEFEATOFTHEBOLDSKEPTIC

Contents1. ContextualismandtheProjectofSolvingtheAIPuzzle...................................................1

2. AGeneralPuzzleaboutSkepticalInclinationsWaxingandWaningvs.theAIPuzzleI’mTryingtoSolve.....................................................................................................................4 

3. UnspokenAI.......................................................................................................................................9

4. Actual,SpokenDisputesoverAI:I’mNotthatNiceContextualist!...........................10

5. AretheSkepticandHerMooreanOpponentBothMakingTrueClaimsasTheyArgue?:SomeDisputesAreGenuine!............................................................................................12 

6. AretheClaimsthatOurDisputantsAre‘Trying’toMakeBothTrue?:SomeDisputesAreDeeplyGenuine!..........................................................................................................15 

7. AretheClaimsthatOurDisputantsAre‘Trying’toMakeBothDeeplyImportant? 17 

8. Boldvs.TimidSkeptics...............................................................................................................18

9. Kornblith’sAttack:Full‐Bloodedvs.HighStandardsSkeptics...................................19

10. ThePhilosophicalInterestoftheHighStandardsSkeptic.......................................21

11. DoIRespondOnlytotheHighStandardsSkeptic?:TheHighStandardsSkeptic,theDeplorablePositionSkeptic,andtheBoldSkeptic.........................................27 

12. DoIRespondOnlytotheHighStandardsSkeptic?:TheSimultaneousDefeatofBoldSkepticsofBothKinds..............................................................................................................29 

13. MySupposedlyThoroughlyExternalistResponsetotheFull‐BloodedSkeptic 34 

14. IndexedAI....................................................................................................................................36

15. IrrelevanttoTraditionalEpistemologicalReflectiononSkepticism?.................40

16. IsMyContextualistPositionInexpressible?...................................................................44

17. TheFactivityProblem.............................................................................................................48

Appendix:DoIEvenKnowoAnyofThistoBeTrue?:SomeThoughtsaboutKnowledge,Belief,andAssertioninPhilosophicalSettingsandOtherKnowledgeDeserts.......................................................................................................................................................57 

References................................................................................................................................................69 

1

Draftof15October2015

CHAPTER4:

CONTEXTUALISMANDSKEPTICISM:THEDEFEATOFTHEBOLDSKEPTIC

1. ContextualismandtheProjectofSolvingtheAIPuzzle

Inthischapter,Iexplainanddefendthecontextualistnatureofmysolutiontothepuzzlethattheskepticalargument(AI)presentsuswith.Aclearviewofjustwhatthecontextualist—oratleastthiscontextualist—istryingtoachieve,andhowheistryingtoachieveit,iscrucialtomostoftheresponses,tightlytyingtogetherthe‘explain’andthe‘defend’aspectsofmyefforts.Iwillbespeakingformyselfhere,andtheremaybesignificantdifferencesinapproachamongdifferentcontextualistresponderstoskepticism,butIdothinkthatonissuespertainingtothecontextualistnatureofourresponsestoskepticism,Iamforthemostpartatleastroughlyalignedwithatleastmyfellow‘contextualistMooreans’whoweremostinfluentialonme:GailStine,StewartCohen,andDavidLewis.WhileIwillbetakingpositionsonspecificquestionsthatmyfellowcontextualistsmaynothaveanswered,andwhileIcan’tanddon’tpresumethattheywouldgivethesameanswers,IdothinkthatthemovesImakeforthemostpartfitwellwiththeirgeneralframeworks,andsoareanswerstheycouldwellgive.1

Iwillbedevotingalotofattentiontorespondingtoobjectionstocontextualistsolutionstoskepticism(andtomeinparticular)thathavebeenleveledinpublishedcriticisms,butwillalsobepresentingmydefensewithaneyeto

1ExceptionsIhaveparticularlyinmindinconnectionwiththat‘forthemostpart’:Insection13,I’mdealingwithanissuethatmaypertainspecificallytomeinvirtueofsomedetailsofmyaccountofknowledge,andsomaynotberelevanttoothercontextualistaccounts;andintheAppendix,ItakemyselftobesteppingquitefarfromissuesothercontextualistMooreanshaveaddressed,andthereforetobeworkingverymuchonmyown.

‐2‐

answeringmisconceptionsofthecontextualistapproachtoskepticismasoneislikelytoencounterthemonone’sphilosophicaltravels.Ofcourse,thesetwosourcesofresistanceoverlapaspopularlinesofcriticismfindtheirwayintoprint,or,Isuppose,aspublishedcriticismsfindtheirwayintothegeneralphilosophicalconsciousness.

Oneofthemostextreme,andmostcommon,worriesaboutcontextualistsolutionsisoneexpressed(perhapsmosteffectivelyandinfluentially)byErnestSosa:thattheyaresimplyirrelevanttotraditionalepistemologicalreflectiononskepticism.However,asthisworryconcernshowmyapproachrelatestootherphilosophicalapproaches,itisbesttoholdoffonrespondingtoitinfulluntillaterinthischapter,whenmyapproachhasbeenmorefullydeveloped.IwillproceedbypaintingapositivepictureoftheprojectthatIasacontextualistrespondertoskepticismamattempting,answeringthemostimportantobjectionsto,anddispellingthemostimportantmisconceptionsabout,contextualistresponsestoskepticismalongtheway.(Withoneexception:Oneofthemostimportantobjectionstocontextualistresponsestoskepticism,andmineinparticular,isthatleveledin(Schiffer1996),totheeffectthatourresponsesinvolveusinimplausiblyattributing‘semanticblindness’tospeakers.However,sinceSchiffer’sobjectionissotightlyboundupwithimportantobjectionstocontextualismitself,itwasanswered,inawayIamstillquitehappywith,invol.1,at(DeRose2009:174‐9).)

However,itisworthnotinghereattheoutsetthatthosewhohaveworrieslikeSosa’softenseemtoviewcontextualistresponsestoskepticismasengagedinaverydifferentinquiryfromtheonethatepistemologistsareusuallyortraditionallyinvolvedinwhentheyaddressskepticism.Contextualistand‘traditional’approachesareseenasbeingonoppositesidesofsomeimpressive‐soundingdivideorother.Thiscanbetiedupwithissuesofclassification,ofwhichsub‐areaofphilosophytheprojectsaresaidtobepartof:Contextualistsaredoingphilosophyoflanguage;traditionalists,epistemology.Or,relatedly,itcanbeseenassomekindofdifferenceinlevelsofsemanticascent:Traditionalistsareinvestigatingquestionsconcerningwhetherweknow;contextualists,questionsconcerningtheproperuseof‘know(s)’(e.g.,Sosa2000).

So,perhapsthefirstthingtobesaidincharacterizingthecontextualistapproachisthat,sofarasIcansee,that’sallwrong.Asageneralrule,whensuchadivideisposited,andthecontextualistissaidtobedoing‘notx,butonlyy’,whatisreallyhappeningisthatIamdoingxinawaythatinvolvesy.SoasIseethings,Iam

‐3‐

addressinganoldproblem—thetraditional,philosophicalproblemofskepticism,atleastasitisraisedbyonevenerableformofskepticalargument.AndthoughIamindeedbringinginsome(fairlylow‐tech)philosophyoflanguage,andsometalkaboutthetruth‐conditionsofsentencescontaining‘know(s)’,IamnotdoinganyofthatinsteadofaddressingthequestionofwhetherweknowthingsinlightofAI‐likeskepticalarguments.Rather,Iambringingintherelevantphilosophyoflanguageasawayofaddressingthesame(epistemological)issuethathasbeenaddressedbyotherapproaches.

Thisisallbestviewedinlightofthemethodologydiscussedacoupleofchaptersago.AswesawinChapter2(expandingonthemethodologicalremarksinSSP),IemployaMooreanmethodologicalapproachtosuchskepticalarguments,seeingtheproblemasoneofasetofclaims,eachofwhichcanseemveryplausible,butwhichseemmutuallyinconsistentwithoneanother.Sofar,ofcourse,thatdoesnotconstitutechangingproblems,butjusttakingasensiblemethodologicalapproachtoanoldproblem.GoingbeyondMoorehimself(sofarasIcansee),Itakethemostpromisingwayforwardinadjudicatingtheconflicttobeintheconstructionofasolutiontotheproblemwhichprovidesagood(‘damage‐controlling’)explanationforwhywegowronginourthinkingatthepointswherethesolutionchargesuswitherror.ButthisisstilljustsensibleMoore‐inspiredmethodology,evenifnotthemethodologyofMoorehimself,anddoesnotconstituteanybigchangeintheverynatureoftheproblembeingaddressed.

It’sherethatIbringincontextualism:It,togetherwithapartialaccountofwhatknowledgeis(a‘double‐safety’account,asI’vecometocallitsincethewritingofSSP,andthatwillbeexplainedmorefullyinChapter7)providethematerialsfortheexplanationIpropose.ButthesearejustthetoolsIusetodothejobofsolvingtheproblem;otherwould‐besolversofthisverysameproblemarewelcometouseothertools,andtonotusecontextualism,noranyphilosophyoflanguageatall,intheirattempts.IfIjudgesuchanalternativesolutiontobea(relativeand/orabsolute)failure,itwon’tbebecauseIinsistattheoutsetthatanygoodsolutionmustusetoolslikethosethatIuse,butbecauseIjudgethat,asitturnsout,thealternativejustdoesnotsucceedinsolvingthe(same,old)problem—theonethatItooamclaimingtosolve.

Asnotedinthefirsttwochapters,mycontextualistsolutiondoesdifferimportantlyfromthetypical‘straightforward’solutionsinthat,byofferingdifferentreadingsofthethreeclaimsconstitutiveofourpuzzle(AI’stwopremisesandthe

‐4‐

negationofitsconclusion),itdistributestheintuitivecostsamongmorethanoneofthethreeclaimsconstitutiveofourpuzzlebeforeitseekstoexplaintheproblemsaway:BoththefirstpremiseofAIandthenegationofAI’sconclusionaretrueononeofthesalientreadingsofit,falseontheother.ButIdon’tinsistthatishowthingsmustbedone:‘straightforward’(non‐contexualist)solutionsarewelcometochoosejustoneloserhere,andexplainawaywhywecanfinditsoplausible,thoughitis,accordingtothesolution,wrong.Indeed,amainpartofSSP(sects.15‐16)isarguingthatattemptstoprovidejustsuchdamage‐controllingexplanationsonbehalfofboldskepticism(whichchoosesthenegationofAI’sconclusionasthestraightforward,loneloser)turnoutnottosucceed.

2. AGeneralPuzzleaboutSkepticalInclinationsWaxingandWaningvs.theAIPuzzleI’mTryingtoSolve

Inthistellingstatement,TimothyWilliamsontouchesonanareaofconfusionaboutthenatureofcontextualistresponsestoskepticismthatitwillbeimportantforustoclearup:

Contextualism supplies a perfectly general strategy for resolving any apparent disagreement whatsoever. Since some disagreements are genuine, we should not always follow that strategy. The conflict of intuitions does not always disappear on further reflection. At least some intuitions are mistaken. Moreover, they are explicably, not blankly, mistaken. (2005: 694-5)

Ihopetheprevioussection,and,evenmoreso,theprevioustwochapters,makeclear(orrendersevenclearerwhatwasalreadyclearinSSP)thatIseemyselfverymuchinthegameofaddressingtheconflictofintuitionsthatatleastoneskepticalargumentpressesonus,andthatIrecognizeattemptsbynon‐contextualiststoexplainhowsomeoftheintuitionsinvolvedaresimplymistakenasaperfectlyfinewayinprincipletoaddresssuchconflicts.Whetherattemptsofsuchatypeactuallyworkoutintheendmustbejudgedonacase‐by‐casebasis.

TheimportantareaofpotentialconfusionIhaveinmindisonethatsurroundstheassumption,madeherebyWilliamson,thatcontextualistsolutionsdon’tmakedisputesoverskepticismouttobegenuinedisagreements.I’mnotsurewhatexactlyWilliamsonmeansby‘genuine’here,butoftenwhenit’ssaidthat

‐5‐

contextualistsolutionsrenderdisputesoverskepticismnon‐genuine,what’sbeingallegedisthatthecontextualistconstruesthejudgmentsbeingexpressedbythepartiestosuchadisputetobecompatiblewithoneanother(and,infact,tobothbetrue).Atanyrate,justwhenandinwhichwayscontextualismrendersdisagreementoverskepticismgenuineturnsouttobeasomewhatcomplicatedmatterthatIwillbeaddressinginthesectionsthatimmediatelyfollowthisone.MyfocuswillbeondisagreementsthatoccurbetweenAI‐wieldingphilosophicalskepticsandMoore‐likeresistersofskepticism,becausesuchdisagreementsaretheonesmostgermanetomyproject,andonesaboutwhichsomeofthemostseriousmisunderstandingsofthecontextualistapproacharise.

However,beforeaddressingsuchdisagreements,Ishouldnotethatthematerialaroundtheabove‐quotedpassageseemstoindicatethatWilliamsonalsohashiseyesonakindofdisagreement(oratleastadifference,ifonedoesn’tthinkitpropertouse‘disagreement’here)inintuitionsthatmightbehadbyasingleindividualatdifferenttimesandindifferentsettings:Inanepistemologyseminarinwhichskepticalworriesprevail,onemighthaveskepticalfeelings,inclinations,orintuitionstotheeffectthatshedoesnotknowthatO,whilethesamepersonmayquitefirmlyjudgeorintuitthatshedoesknowthatOinamoreordinarysetting.Thisgeneralphenomenon,theskepticalsideofwhichcanbedrivenbyavarietyofdifferentskepticalarguments,isfrequentlynoted,sometimescitingHume’sfamousobservationsabouthowskepticaldoubtsfeltinphilosophicalcontextsdissipateinothersettings.2Andsomemaythinkcontextualismissupportedbyitsabilitytoexplainhowsuchshiftsinintuitionsandleaningscanoccur,orthatcontextualiststhemselvesseektosupporttheirviewsbysuchconsiderations.Andwecontextualistshavenodoubtthoughtandwrittenthingstoencouragesuchanunderstandingofwhatwe’reupto(as,forinstance,insection2ofSSP).However,evengettingadecentgriponthephenomenatobeexplainedhere(justwhensuch

2Hume,ATreatiseofHumanNature,BookI,PartIV,section7(Hume1978:269):‘Mostfortunatelyithappens,thatsincereasonisincapableofdispellingtheseclouds,natureherselfsufficestothatpurpose,andcuresmeofthisphilosophicalmelancholyanddelirium,eitherbyrelaxingthisbentofmind,orbysomeavocation,andlivelyimpressionofmysenses,whichobliterateallthesechimeras.Idine,Iplayagameofback‐gammon,Iconverse,andammerrywithmyfriends;andwhenafterthreeorfourhour'samusement,Iwou'dreturntothesespeculations,theyappearsocold,andstrain'dandridiculous,thatIcannotfindinmyhearttoenterintothemanyfarther.’CrispinWrightprominentlyalludestothispassageat(Wright2005:240).

‐6‐

shiftsoccurandwhattheseshiftsarelike)turnsouttobequitetricky,Ithink,asthisallseemstoworkoutverydifferentlyfordifferentpeople.

Williamsonpointsoutthatanon‐contextualistexplanationofsuchshiftsispossible.Aftergivinganexampleofanotherquestionwhereouropinionmayshift,butwhereacontextualistexplanationwouldnotbeattractive,Williamsonverysensiblywrites:

Whether one knows (or one's belief is epistemically justified) also depends on the balance between many complex, unquantifiable, subtly interacting considerations. In many realistic cases, some weigh one way, others the other way. The concept of knowledge or of epistemic justification provides no algorithm for weighing all these factors against each other or integrating them into a final verdict. No wonder that the skilled sceptic can present the considerations that favour a negative verdict so vividly that they intuitively appear to outweigh the considerations on the other side. It does not follow that the sceptic is right, even in the context of the epistemology seminar; the case may not even be borderline. Nor does it follow that the sceptic is wrong, even in the everyday context. (2005: 695)

Infact,Williamsongoesontoarguethatanimportantfeatureofsuchshiftsfavoranon‐contextualistaccount:

As before, the intuitions that predominate in one context spill over to judgements about the truth-values of sentences as uttered in the other context. In the everyday context, it is intuitive that someone in the epistemology class who says 'Mary does not know that the train leaves at noon' is overestimating her epistemic difficulties. In the epistemology class, it is intuitive that someone in the everyday context who says 'Mary knows that the train leaves at noon' is underestimating her epistemic difficulties. Although such data are not decisive against contextualism, they tend to support the non-contextualist explanation. (2005: 695)

Fortherecord,IthinkWilliamson’sstatementaboutwhatthe‘data’arehereisoverlysimple:Ineithersetting,ifwehavetheground‐levelintuitionsWilliamsonassignstous,whenwethenturnourthoughtstothetruthvaluesofourclaimsmadeintheothersetting,whichisinrelevantwaysverydifferentfromthesettingfromwhichwearemakingthejudgmentabouttheclaim,Ibelievethatthingsgetverymurkytousveryquickly—anddifferentpeoplesayverydifferentthings.Further,Ithinkthatthemurkier,actualdataabouthowinonesettingweregardthejudgmentswemakeinverydifferentsettingswouldbemorefavorabletocontextualismplayinganimportantroleinexplainingwhat’sgoingonthanarethemuchcleaner‘data’thatWilliamsonseemstosuppose—thoughitislikelythatother

‐7‐

forces,havinglittletodowithcontextualism,arealsoimportantlyatworkhereaswell.

ButthemainpointIwanttomakeaboutthisissueisthatthereareimportantgroundsforacontextualistsolutiontoAI‐inspiredskepticismthatarenotreallydrivenbyitsabilitytoaccountforanyofthedetailsofhowskepticaldoubts,howevertheymightarise,waxandwaneintheirforceovertimeasonemovesfromonesettingtoanother,norbyitsabilitytoexplainhowonefeelsaboutthetruthvaluesofone’sownclaimstoknowthingsthatweremadeatdifferenttimesandinverydifferentcontexts,someskepticalandsomeeveryday.3Rather,Iamfocusedonaparticularintuitivelypowerfulskepticalargument,AI.Andevenwithrespecttothatargument,Iamnotevenparticularlydrivenbydetailsofhowpeoplereactwhentheskepticpresentsherargument.Irealizethatsome(thoughcertainlynotall)respondentsfindtheskepticalclaimthatisAI’sconclusiontempting,orevenmorethanjusttempting,ratherthanincredible,whileinthegripofsomethinglikeAI,andthatsuchskepticalfeelingscandissipatewithachangeofsettingsandmindset.ThoughI’veneverseenitdoneatallsuccessfully,andthoughitwouldbeaverycomplicatedproject,suchdetailscouldbeascertainedandthenbroughttobearonourpuzzle,givinggoodreasonsforacceptingonesolutionoveranother.SofarasIcansee,mysolutiontoAI,builtcruciallyoncontextualismandtworelatedbutimportantlydistinctwaysofevaluatingourbeliefs(intermsoftheirsensitivityvs.theirsafety),appearsaboutaswell‐positionedasanysolution,straightforwardorcontextualist,tobeabletosurvive,andevenexplain,themessydetailsofhowexactlydifferentpeoplereacttotheconflict(whentheyactuallyencounterit,andthenastheymoveoutoftheirencounter)comingtolight.ButIhaven’tbeenarguingformysolutionbasedonanysuchcarefullookatthoseextremelymessydetails,forthegoodreasonthatIamunabletoinanysatisfyingway.Indeed,evenifIweretolearnthatalmostnobody’s,or,ontheotherextreme,almosteverybody’s,overallresponsetobeingpresentedwithAI(atleastgivenasuitablycompelling

3Invol.1(DeRose2009),Idoaddressthematterofhowspeakersinhighvs.lowstandardscontextsevaluatethetruthvaluesofoneanother’sclaims(seeesp.pp.153‐4and160‐61),andwhatIsaythereIwouldalsobeinclinedtosayaboutcaseswherethesespeakersareconstruedasbeingthesamepersonatdifferenttimes.Mytreatmentofthematterthereisverymuchinkeepingwithmycurrentattitude.NotethatIamthererespondingtoanobjectiontocontextualism,ratherthanadvancingmyownpositiveargument,andthatmyresponsehasapronounced‘notsofast!’elementofclaimingthatthedatahere‘misbehave’inwaysthatdisabletheobjectiontocontextualismunderconsideration.

‐8‐

presentationofit)wastoacceptitsconclusion,ortobeatleaststronglyinclinedtodoso,thoseresultswouldnotmucheffectmythinkinginwaysIcannowsee,sincemysolutionisn’treallydrivenbysuchconsiderations.

Rather,whatIthinkweshouldfixoureyesonareafewfactsaboutourreactionstotheconstitutivepartsoftheAIpuzzle,takenindividually,thatdoseemtolerablyclear:Atleastwhentheclaimsareconsideredindividually,wedotendtofindtheclaimthatweknowthatOveryplausible,butwealsocanfindeachoftheskeptic’spremisesveryplausible.That’sverypuzzling,tosaytheleast,giventherelationamongtheseclaims.That’sourAIpuzzle,whichIhavebeenouttosolve.Evenif,quitecontrarytofact(insofarasI’vebeenabletodeterminetherelevantfacts),say,almosteveryonewereverystronglyinclinedtoacceptAI’sconclusionwhenpresentedwiththatskepticalargument,therewouldstillbethepuzzleofwhywearesostronglyinclinedtojudgethatwedoknowthatOinothersettings,eventhoughwearealsoinclinedtofindeachofAI’spremisestobequiteplausible.AndwhatIclaimtoexplainishowAI’spremisescanhavetheintuitiveappealthattheyhave,eventhoughitsconclusioncan(attheveryleast,inmostsettings)seemsoincredible.

Ofcourse,thoughIhavebeenfocusedononeparticularlystrongbasisforskepticism,contextualistsolutionsinimportantwaysanalogoustomysolutiontoAImayalsobeappropriatefordealingwithother,quitedifferent,skepticalarguments.Butifso,andifthesesolutionsreallyareinthesignificantwayanalogoustomine,theimportantquestionwillbewhethertheproposedcontextualistsolutionsaccomplishwhatIclaimtoaccomplishwithrespecttoAI:Canthecontextualistsemanticsof‘know(s)’beingproposed(supposingtheskepticalargumentinquestionconcernsknowledge),togetherwithsomethinglikea(nodoubtpartial)accountofwhatknowledgeis,wellexplain,andexplainbetterthanrivalaccounts(includingofcourseinvariantistaccounts),howthepremisesoftheparticularskepticalargumentbeingaddressedcanbeasplausibleastheyare,whileitsconclusionisasimplausibleasitis?Wecertainlyshouldnotbetraffickinginanythingthatcouldwithanyfairnessbecalledaone‐size‐fits‐all‘perfectlygeneralstrategyforresolvinganyapparentdisagreementwhatsoever.’4

4Thereissomethingofashallowrecipebywhichonemightusecontextualism(oralso,Isuppose,forthatmatter,bare,unsupportedclaimsoftwo‐wayambiguity)to,atleastbysomelowstandard,‘resolve’anyapparentdisagreementwhatsoever.HereisStanley’swayofputtingit(thoughheisdescribingthe

‐9‐

3. UnspokenAI

Itcanseemstrange,orevenwrongheaded,forthecontextualisttoappealtorulesofconversationforthesettingofepistemicstandardsintryingtosolvetheAIpuzzle,for,thoughareal,liveskepticcouldpresentAItoanotherpersonsheisactuallytalkingto,theconundrumAIconfrontsuswithseemsjustaspuzzlingwhenoneinsteadwrestleswithitinsolitude,withoutanyonesayinganythingatall.Howcouldaruleofconversationthenplaysuchapivotalroleinitssolution?Torevisitandexpanduponapointmadeinsection2ofSSP(thethird‘importantpoint’madeattheendofthatsection),thekeytotheanswertothatpointedquestionisthatourjudgmentastowhetherweknowsomething,evenwhenthatjudgmentisissuedsilentlyandwhileinsolitude,seemstobeinfluencedbyoursenseofwhetheritwouldberightorwrongtoclaimto‘know’it.Centraltomysolutionismycontentionthat,duetotheoperationoftheRuleofSensitivity,claimstoknowthatnot‐Handadmissionsthatonedoesnotknowthatnot‐Htendtoinvokehighepistemicstandardsthatone’sbeliefthatnot‐Hdoesnotsatisfy,whichproducesatendencyfortheclaimthatonedoes‘know’togofalseandtheadmissionthatonedoesnot‘know’tobetrue.Thattheclaimhassuchatendencytobackfire,andtheadmissiontoworkoutsowell,canmakeitseemtous,evenwhenwe’reconsideringthemattersilentlyandinsolitude,thatwereallydon’tknowthatnot‐H.

SomesuchmovealsoseemsneededtomakesenseofvariousbroadlyGriceanmaneuvers,includingsomeweshouldbekeentoendorse.So,forexample,insofarasonewastemptedtothinkthatitisnottruetosayofabrightredmailboxthatis

contextualist’shandlingofBank‐likecases,ratherthanitshandlingofskepticism):‘Supposewehavetwoclaims,Sandnot‐S.Italwaysopentosomeonetoresolvetheapparentconflictbymaintainingthatthetwooccurrencesof‘S’intheclaimsexpressdifferentpropositions,relativetotheirdifferingcontextsofuse’(2005:122).Buttheactualproposedcontextualistsolutionstoskepticismarenotjustinstancesofthatrecipe.Oratleastnomoresothantheactuallyproposedinvariantistapproachestotheproblemarejustinstancesoftheverygeneralinvariantist‘recipe’:‘Insistthatthetwooccurrencesof“S”intheclaimsalwaysexpressthesamepropositioninallcontexts,sothatthetwoclaimsaresimplyincompatible,andthendenyoneofthem.’That‘recipe’seemstobeonethat,forallwecanseeinadvance,couldbefollowedveryeffectivelyorextremelyineffectively(dependingcruciallyonhowgoodacaseismadefordenyingtheclaimthatisrejected).Well,ofcourse,thesamegoesforcontextualistsolutions.PerhapsWilliamsonwouldagreewithallthat.Still,itneedstobesaid.

‐10‐

onlytwofeetawayfromoneinbrightsunlight‘Itlooksredtome’,onsuchgroundsasthatitisnotdoubtfulenoughtomethattheobjectisredforthatsentencetobetrue(andIrealizethatinthesepost‐Griceantimes,‘insofar’maynotbeveryfaratallformanyofus),onewouldlikelybesotemptedevenwhenthinkingthroughtheissuesilentlyandinsolitudewhilestaringdirectlyatsuchamailbox,orimaginingdoingso,andnotonlywhenactuallyengagedinmakingtherelevantassertiontosomelistener.Thatpointdoesn’tnullifyGrice’sattempttoexplainawaythetemptation(Grice1961).Rather,Grice’smaneuversshouldbeunderstoodasdirectlyexplainingwhat’swrongwiththespokenclaim(intherelevantcircumstances),butthenalsoindirectlyexplainingwhywemaybetemptedbyafalseviewofthemeaningof‘seems’evenwhenthinkingtoourselves:Oursensethatitwouldbewrongtomaketheclaimintherelevantcircumstancescantemptustoward(wrongly)judgingthattheclaimwouldtherebefalse(whereonehasnodoubttheobjectreferredtoisred),evenwhenwemakethejudgmentsilently.

4. Actual,SpokenDisputesoverAI:I’mNotthatNiceContextualist!

ThoughIdon’tthinkthematterofwhichsideofvariousunenlightened(aswe’reabouttocallthem)debatesoverskepticismmayhave‘won’isveryimportant(seeDeRose2009:138‐9),somefolksmayhaveactuallytakenpartinsuchdisputes,andmaybeinterestedinwhetheritwastheyortheiropponentwhosecontentionswerecorrect.Butthereisanother,potentiallymoreimportant,reason,beyondthismatterofpersonalpride,thatsomehaveforbeinginterestedinwhatthecontextualisthastosayaboutsuchdisputes:Theythinkthatthecontextualist’srulingaboutthemisimplausible,andsocountsagainstthecontextualistapproachtoskepticism.

So,let’sconsidersuchadispute.5Imagineacharacterwecancallan‘undifferentiated’AIskeptic.Sadly,thepossibilityofcontextualismhasnotoccurred

5Inaddressingwhattosayaboutsuchadisputeoverskepticism,Itakemyselftobeaddressinganissuethatloomslargeinpopularconceptionsaboutwhatcontextualistsareupto.Butcriticismsofthecontextualistapproachtoskepticismthatstartwiththoughtsaboutwhatthecontextualistsaysaboutsuchdisputesareinthepublishedliterature.AsIpointoutinVolume1(DeRose2009:130‐131),aparticularlygoodexampleofthisis(Richard2004:215‐16).

‐11‐

toher,evenasapositionsomeoneelsemightaccept,soshejusttakesitforgrantedthatweeithersimplyknowthatOorwesimplydon’tknowthatO;6andonthebasisofAIsheholdsthatit’sthelatterofthosetwothatisthecase.

Nowwearrangeforourskeptictoenterintoaspokendispute.So,supposethisundifferentiatedskepticmeetsresistanceasshepresentsAI.ThepossibilityofcontextualismhasalsonotoccurredtotheundifferentiatedMoore‐likeresistersheencounters,whoalsopresumesthatweeithersimplyknowthatOorsimplydon’tknowthatO,butwhoinsistsinthefaceofourskeptic’sargumentthatit’stheformerofthosetwothatgetsthingsright.‘HowabsurditistosuggestthatIdonotknowit,butonlybelieveit,andthatperhapsitisnotthecase!’,heexclaims.7Backandforththesecombatantsargueinaheated,butas‐yet‘unenlightened’,disagreement—i.e.,adisagreementsadlyunenlightenedbythepossibilityofcontextualism.

Enterthenicecontextualist.Thischaracterisapparentlyhowcontextualistsareoftenimaginedtobe.Heisaveryaffirming,perhapsevennurturing,soul.Likeaparentwhoseekstointerveneonlyinneutral,balancedwaysinsquabblesamonghischildren,wheneverthisnicecontextualistencountersaphilosophicaldispute,oratleastonemeetingcertainconditions(perhapsonethatseemsdifficulttoacertaindegreetoadjudicate),heseeksto‘dissolve’itasbeingonlyanapparentdisagreement,insuchawaythatthepartiestothedisputeareallconstruedasmakingtrueclaimsastheyargue,and/or,relatedly,heseeks‘even‐handed’resolutionswherethepositionsofallthepartiestothedisputearerespectedandvalidatedto(atleastroughly)thesamedegree.8Wemightfurthersupposethatit

6Orshemayalsoaddthepossibilitythatit’ssimplyindeterminatewhetherweknow:thatclaimstotheeffectthatweknowthevariousOsareneithertruenorfalse.Recallthatonoursemi‐technicaluseof‘simply’,‘Wesimplyknoworwesimplydon’tknow’isn’tusedtoassertbivalenceaboutthesematters,buttoclaimthatthematterofwhatthecorrectanswerisdoesn’tvarywithcontextualshiftsinthemeaningof‘know(s)’.

7ThisexclamationisbasedonwhatMooresaidinhisfamous‘ProofofanExternalWorld’indefenseoftheclaimthatheknewthat‘Hereisonehandandhereisanother’:‘HowabsurditwouldbetosuggestthatIdidnotknowit,butonlybelievedit,andthatperhapsitwasnotthecase!’(Moore1959:146).

8Iaminspiredinmyuseof‘even‐handed’heretocharacterizehowmanyphilosophersseemtothinkofthemotivationsbehindcontextualistsolutionstoskepticismby(Wright2005:240‐242),butWrighthimselfusesthetermtodescribe,nothowwecontextualistsseektotreatthetwosidesofphilosophicaldisputesoverskepticism(philosophicalskepticvs.Moore‐likeresisterofskepticism),buthowweseektotreatskepticaldenialsofknowledgevs.commonsenseaffirmationsofknowledge(philosophicalskepticvs.ordinaryclaimsmadeinnon‐philosophicalsettings).Insoviewingus,Wrightiscertainlygettingclosertoour(oratleastmy)actualmotivationsthanisthecommonconceptionofcontextualistsIamusinghistermtocharacterize.

‐12‐

waslargelyinordertofacilitatesuchaniceresolutionofsuchoftennastydisputesoverskepticismthatournicecharacterwasledtoacceptcontextualisminthefirstplace.

Butthatain’tme,aswe’reabouttosee.Nordootheractualcontextualistsaboutknowledgeandskepticismtendtobelikethat,aswe’llalsonote.

5. AretheSkepticandHerMooreanOpponentBothMakingTrueClaimsasTheyArgue?:SomeDisputesAreGenuine!

Contextualismdoesopenupthepossibilityofrulingthatbothofourunenlighteneddisputantsaremakingtrueclaimsastheyargue.Butwhatdoactualcontextualistssayaboutsuchadispute?Probablythefirstthingtosayisthatwehaven’tcaredallthatmuchaboutthequestion—andwecertainlyaren’tfollowinganygeneralimperativetorulethatbothpartiestosuchadisputearemakingtrueclaims.Irejectalternativesolutionsbecausetheyturnoutnottowork,notoutofhandbecausetheyfailtodeliverthe‘nice’verdictthatbothpartiestothedisputearespeakingtruthfullyastheyargue.9

However,WrightseemstometostillbetakingustobemoreconcernedwiththefateoftheskepticthanIthinkwereallyare.AsIseeit,whenwecontextualistssetouttoaddressthepuzzlethattheskeptic’sargumentconfrontsuswith,wewereseekingtoadequatelyaccountforthepoweroftheskeptic’sargument,butweweremoreopenthanWrighttakesustobetosolutionsonwhichtheskepticturnedouttobejustmistaken.But,ofcourse,judgingmotivations—including,andperhapsespecially,one’sown—isdelicatebusiness!Still,Iespeciallydon’ttaketheportionofthefollowingpassagefromWrighttowhichI’maddingemphasistoaccuratelydescribewhatIsetouttoachieve:‘Aperfectlyeven‐handedtreatmentofscepticismandcommonsensewillallowtheirrespectiveconclusionstostandrelativetocontextandwillchargeneitherwithanycognitiveshortcomingunlessthechargeismatchedbyacorrespondingchargeagainsttheother.Sotheclaimsofcommonsensetoknowledgewillstand,albeitdefeasibly,nodoubt,modulocertainrelevantcontextualfactors.Andtheclaimsofscepticismtounavoidableignorancewillalsostand,modulocertainrelevantcontextualfactors’(Wright2005:241).

9Idoholditagainstboldskepticalsolutionsthattheyinvolveusin‘systematicfalsehood’inouruseof‘know(s)’(seeesp.section16ofSSP),butthisisgeneralsystematicfalsehoodinspeakers’useoftheterm,importantlyincludingusesIammuchmorehesitanttorulearefalse,ascomparedwithsomefalsehoodsissuingfromthemouthsofsomeMoore‐likedisputantsintheoddphilosophicaldisputeoverskepticism(seeDeRose2009:56‐59).AndIputthisforwardasthekindofproblemthatanysolutiontoagenuineconflictofintuitionsmustfaceandtrytoovercome.Irejecttheproposedboldskepticalsolutionsbecausetheir

‐13‐

Indeed,Icouldnotrejectthoseboldskepticalsolutionsforthatreason,because,likeotheractualcontextualists,Imyselfdon’toptforsucha‘nice’solutiononwhichbothpartiescomeoutmakingtrueclaims!However,IdidnotexplainthataspectofmyviewinSSPitself,butthereprovisionallyassumedaskeptic‐friendlyversionofcontextualismforeaseofexposition(seethesecond‘importantpoint’explainedinsection2ofSSP),waitinguntil‘SingleScoreboardSemantics’(‘SSS’)10tosayhowIwasinclinedtoaddressthoseissues.Soaquickreviewofthesituationiscalledforhere.

Whetheracontextualistwillrulethatourarguingpartiesarebothmakingtrueclaimsdependsonwhatnot‐so‐obviouscallhemakesonthegeneralissueofhowtohandleakindofdisputeinwhich:acontext‐sensitivetermiscruciallyinvolved;thedisagreeingpartiesaremakingclaimsthatatleastseemonthesurfacetobeinconflictwithoneanother;whethertheythinkaboutwhatthey’redoinginthesetermsornot,eacharguingpartyisexecutingconversationalmaneuversthattendtopushthecontentofthekeycontext‐sensitivetermtowardhavingavaluethatwouldmaketheirownclaimstrue;11andyettheyeachtakethemselvestobecontradictingoneanother,andmayevengiveexplicitindicationsoftheirintenttobecontradictingoneanother(‘No,you’rewrong.Sallyistall’).AcontextualistwhoseekstomakebothpartiestosuchdisputecomeoutspeakingtruthswilloptforsomethinglikewhatIcalla‘multiplescoreboards’viewonwhichthecontentofeachspeaker’sspokenclaimsmatchesthatspeaker’s‘personallyindicatedcontent’—thecontentthattheconversationalmaneuversthatspeakerismakingwouldtendtoputintoplace(DeRose2009:134‐5).Andmanyassumethatispreciselywhatthecontextualistshouldsayhere,aboutourimagineddisputeover

attemptstoovercometheirproblemturnoutnottoworkaswellastheircontextualistrivalinmakingsenseofthevariousintuitiveclaimsweare,oratleastshouldbe,tryingtoadjudicateamonginnavigatingthispuzzle.Importantlytoourcurrentconcern,Idefinitelydonotrejectthemoutofhandbecausetheyfailtodeliverthe‘nice’resultthatbothpartiestoourundifferentiateddisputearemakingtrueclaims.

10SSSwasfirstapaper(DeRose2004),thenChapter4of(DeRose2009).ThoughI’llbegivingpagereferencestothelatter,andthoughthelatterdoescontainimportantmaterialnotincludedintheearlierversion,everythingIuseandrefertoherewasalsointheearlierversion.

11Oneimportanttypeofmechanismfordeterminingthecontentofcontext‐sensitivetermsseemstobe‘rulesofaccommodation’onwhichthecontenttakesonvaluessoastomakewhatissaidtrue.Ifso,thenitistobeexpectedthattheclaimsofeachpartytosuchadisputewilltendtopushtheconversationalscoretowardvaluesthatmakesthatparty’sclaimstrue.

‐14‐

skepticism,orthatactualcontextualistsdotakethatoption,oreventhatitisadesiretoattaintheniceresultthatbothpartiesaremakingtrueclaimsthatmotivatescontextualisminthefirstplace.

Butthatmustnotbewhat’sreallymotivatingus,becausenoneofGailStine,StewartCohen,DavidLewis,ormyselfisinclinedtowardsuch‘multiplescoreboard’contextualistviews,butinsteadoptforviewsonwhichsuchadisputeisquiterealandit’simpossibleforbothunenlighteneddisputantstobespeakingtruthfully.12Onsuch‘singlescoreboard’views,the‘conversationalscore’fortheterminquestionisaffectedbytheconversationalmovesmadebybothpartiestotheargument,buttheresultofallthatmaneuveringisasingle‘score’thatgovernsthecontentofbothparties’useoftheterm,sothattheoneisdenyingjustwhattheotherisaffirming(2009:135‐6).Inmyowncase(andIsuspectinthecaseofStine,Cohen,andLewis),amainsourceofattractiontosucha‘singlescoreboard’approachisthatIverymuchsharethe(oftenstrong)senseourdisputantsarecontradictingoneanother,andtheapproachmakesgoodsenseofthatfeeling.Thatbasicapproach,though,leavesopenthequestionsofwhatthatsinglescoreisandsowhich,ifeither,ofthepartiesismakingtrueclaims.Thesearemattersthatwecontextualistshaven’tfoundveryurgent,but,asithappens,ontheparticularviewthatIleantoward,andonbothoftheviewsthatseemtobeCohen’sfinalists,neithertheskepticnorheropponentisspeakingtruthfullyastheyargue.Rather,bothpartiestothedisputearemakingclaimsthatareneithertruenorfalse(2009:58,144‐8).Lewistendstowardamoreskeptic‐friendlyversionofcontextualism,onwhichtheskepticisspeakingtruthfullyinsuchanargument,butLewisseemsnomoreanxiousfora‘makeeverybodyhappy’solutionthanCohenorIam,foronhisview,theskeptic’sopponentismakingfalseclaims.13AndthoughI‘provisionallyassumed’acertainpositiononthematterinSSPforeaseofexpositionandnotbecauseIthoughtitwascorrect,eventhatviewwasoneonwhichonlytheskepticismakingtrueclaimsinthetypeofdisputeinquestion.So,thoughcontextualismopensupthepossibilityfor

12Stine’srelevantpioneeringwritingsonthesematterswasquiteearly,and,duetohertragicearlydeath,shedidn’tfacetheseissuesveryexplicitly,somyclassificationofhermustbesomewhattentative,butIaminclinedtoreadher(inStine1976)astendingtoward(intheterminologyofSSS)asingle‐scoreboardviewofthe‘reasonableness’variety.

13See(DeRose2009:136‐8)fordiscussion.

‐15‐

ustodoso,noneofusseemstooptforaviewthatdeliversthe‘nice’resultthatbothpartiestosuchadisputearespeakingtruthfullyastheyargue.

Ofcourse,‘punishing’bothofthedisputants,asitwere,byrulingthatneitherofthemismakingtrueclaims,asCohenandIareinclinedtodo,introducesanelementof‘even‐handedness’toourtreatments,inthemannerofacertainkindofevenlyharshparent.ButitdoesclearlyshowCohen’sandmywillingnesstodepart,Isupposeinjustaboutasradicalawayaspossible,fromthekindof‘everybody’shappy’(inthe‘everybody’smakingtrueclaims’senseofthat)solutiontheachievementofwhichmotivatesthecontextualistoftheimaginationsofmany.

6. AretheClaimsthatOurDisputantsAre‘Trying’toMakeBothTrue?:SomeDisputesAreDeeplyGenuine!

Forthecontextualist(onewhofollowsthe‘basiccontextualiststrategy’ofsection2ofSSP),therearetruthsinthevicinityofwhatbothofthedisputantsinourunenlightenedargumentoverskepticismaresaying,andthesetruthscanbethoughtofaswhatthedisputantsare‘trying’toexpress,asitwere:WedoknowthatObyordinary,moderatestandardsforknowledgethattheMoorean’sresistantmaneuverswouldtendtokeepinplace(orputinplace,dependingondetailsofthestory),andwedonotknowthatObythestandardstheskeptic’smaneuvershaveatleastsometendencytoputinplace.Givencertaindetailsofhowthecontextualiststoryistold,oneorbothofourdisputantsmaybeblockingtheotherfrommakingthetrueclaimstheyareinthisway‘trying’tomake—i.e.,thetrueclaimstheywouldbemakingiftheirconversational‐score‐affectingmoveswereacceptedorwentunopposed(andoninvariantisttheories,wherethescoredoesn’tchange,wesaythatoneis‘trying’tosayjustwhatonedoessay).14Thisisthefeatureofmycontextualistsolutionthatthosewhotakemetobea‘nice’contextualistarelikelypickingupon,andtheymightrespondtowhatIsayintheprevioussectionasfollows:

14Here,Iemploy,whilebeingmoreexplicitaboutmyintendedmeaning,auseof‘trying’thatIusedbackin(DeRose1990:275).

‐16‐

OK, so you evidently don’t seek to make both parties of our dispute over skepticism out to be making true claims. But, still, you do insist on finding two true propositions, one of which can be expressed by the sentence ‘I know that I have hands’, and the other by ‘I don’t know that I have hands’, and though you don’t rule that our disputants are each succeeding in asserting one of these truths, you do make them out to be each ‘trying’ to express one of them, as it were, in the way explained above. So you really are, in a quite relevant way, seeking some kind of ‘no-fault’ dissolution to the disagreement, in which there is a truth that is at least in an important way standing behind each of the disputant’s positions. And it is a mistake to insist on such a solution to just any philosophical dispute. Some disputes are, as we may say, ‘deeply genuine’: the disputants cannot both be right, neither on the level of what they are actually claiming, nor on the level of what they are ‘trying’ to claim, as we have been using that phrase. And perhaps the dispute over skepticism is one of these deeply genuine ones.

Mysolutiondoesindeeddifferfrominvariantistsolutionsinthatitdoesnotmakethedisputesoverskepticisminquestionouttobedeeplygenuineinthisway.AndIsupposethatsomeeasy‐enough‐to‐imaginephilosophercouldstarttheirinquiryintoAI‐inspiredskepticismwithsomekindofsensethatthedisputehereisnot(inthepeculiarsensewearenowusing)deeplygenuine,andmightbestronglymotivatedtowardcontextualismandacontextualistsolutiontoskepticismpreciselybecausetheyvindicatethatsense.Inanextremecase,suchaninquirercouldbewelldescribedas‘insisting’onasolutionthatdeliversthatresult—andthatcouldbetheirinclinationtowardphilosophicalpuzzlesandconflictsgenerally,orperhapstowardthosethatmeetsomegeneralcondition,likebeingsufficientlydifficulttoresolve.

Butthatneednotbewhatmotivatescontextualism,andthat’snotwhatmotivatesme,sofarasIcantell.IrejectinvariantistsolutionstoAI,includingtheboldskeptic’ssolution,becausetheyturnoutnottoworkaswellinresolvingthepuzzleasdoesthecontextualistsolutionIdoaccept,notbecauseI‘insist’fromtheoutsetthatanysuccessfulsolutiontothisproblem—norcertainlytowardjustanysufficientlyknottyphilosophicalproblem—willhavetomakethedisputeinquestionouttobenotdeeplygenuine.15

15However,Idoherebyreiteratemyadmissionfromnote8aboutthedelicatenessofjudgingone’sownmotivations.

‐17‐

7. AretheClaimsthatOurDisputantsAre‘Trying’toMakeBothDeeplyImportant?

SomeunderstandcontextualistsolutionsasseekingtoprovidearesponsetotheAIdisputethatis‘even‐handed’inthatitacknowledgesthedeepimportanceofbothofthetruthsthatourarguingpartiesare‘trying’toexpress,likealecturingparent:‘Now,Abby,youaretryingtomakeaveryimportantpointhere.ButwhatBenistryingtosayisequallyimportant.Youtwoshouldjuststopallyourfighting,alloweachothertomakeyourimportantpoints,andrespecttheimportanceofwhattheotherisexpressing.’Anicecontextualistlikethis,happeninguponanargumentbetweenaskepticandaMoore‐likeresister,occurringindarkignoranceofthepossibilityofcontextualism,would,inbringingcontextualistenlightenmenttothisfight,counseleachdisputanttoperhapsappendsomekindofclarifyingphrasetotheirusesof‘know’,stoppresentingtheirclaimsasiftheywereatoddswithwhattheotherissaying,andachieveablessedharmonyinwhicheachismakingthetrueanddeeplyimportantclaimthathadbeenallalongbehindthevalueofthepositiontheyweretaking.

Butthebasiccontextualiststrategyisneutralabouttheattitudeoneshouldtaketowardthetruthsinquestion.Thoughitwouldnotbesurprisingtofindsomeoneattractedtocontextualistsolutionswhofindsthetruth‘behind’theskeptic’spositiontobeveryimportant(perhapssomedeep,orevenmenacing,factaboutthehumancondition),mysenseisthatweactualcontextualistsgenerallydon’tthinktheskepticaltruthinplayherereallyisverysignificant.Thatiscertainlytrueofme,aswewillseeinsection15,below.Buttheimportantpointtomakehereisthatcontextualistsneednotbesomotivated.Theconvictionthatthere’sboundtobesomethingimportanttolearnaboutknowledgefromapuzzleassharpasistheonepresentedbyAIneednotbedrivenbyanopinionthatthereissomedeeplyimportant(andperhapsmenacing)skepticaltruthsomehowstandingbehindtheskeptic’sstance.Thelessonstobelearnedmayinsteadhavetodowithwhatknowledgeis,andhowknowledgeclaimswork,insuchawayastododgemenacingskepticalclaims.

‐18‐

8. Boldvs.TimidSkeptics

Sofar,we’vebeenkeepingourAIskepticinthedark,unenlightenedbythethoughtofcontextualism.Oncethepossibilityofcontextualismisraised,ourAIskepticfacesachoice:Doesshetakeherargumenttoshowthatwedon’tknowevenbythestandardsforknowledgethatgovernourordinary,nonphilosophicalepistemictalk,oronlythatwedon’tknowbyunusuallyelevatedstandardsthatherargumentperhapshassometendencytoputintoplace?ThisisthechoicebetweenbeingaboldandbeingatimidAIskeptic,centraltothedivide‐and‐conquerstrategythatIpointtowhenIdistinguishbetweenthesetwoskepticsinsection2ofSSP.

OurquestiontotheAIskepticdoesnotpresupposecontextualism.Ourboldskepticcouldactuallybeacontextualistwhoacceptsvaryingstandardsforknowledge,butjustthinksthatAI’spremises,andsoitsconclusion,aretruebyordinarystandards—perhaps(butnotnecessarily)becauseshethinksthey’retruebyallallowablestandards.Butimportantly,shealsomaywellbeaninvariantistwhothinksAI’spremisesaretruebyordinarystandardsbecausethey’retruebythesinglesetofstandardsthateverprovidethetruth‐conditionsforclaimsto‘know’—whichordinarystandardsshemaythinkturnouttobemuchmoredemandingthanacursorylookatourratherbreezyknowledge‐ascribingbehaviormightleadonetothink.Orshecouldbeundecidedbetweentheabove:‘I’mnotsureaboutcontextualism,butinanycase,mypremisesaretruewhenevaluatedatthestandardsthatgovernourordinarytalk’.

Thetimidskepticacceptscontextualismandtakesherargumentonlytobeshowingthatwedon’tknowbytheelevatedstandardsforknowledgethatthepresentationofherargumenthassometendencytoputinplaceandnottobeshowingthatwedon’tmeetthestandardsthatprovidethetruth‐conditionsforordinaryusesof‘know(s)’.Shemayallowthatwedoknowbyordinarystandardsforknowledge,or,moreintriguingly,shemayleavethatquestionsomehowopen—perhapssuggestingsomethinglikethatitmustbeleftopenbecausethere’sjustnotellingwhetherweknowbyordinarystandards.

Thesetwoskepticsfaceverydifferentchallenges.Themainchallengetotheboldskepticconcernswhetherandwhy(orwhyintheworld,astheunsympatheticmayputit)weshouldacceptherboldskepticalclaims,andpressingthischallengeismyanti‐skepticalfocusinSSP.ThisistheskepticIseektodefeat.Thingsareofcourse

‐19‐

verydifferentwhenitcomestothetimidskeptic,forIagreewithherdefiningclaimthatwedon’tknowthingsbythehighstandardsthatthepresentationofAIhasatleastsometendencytoputintoplace.Hermainchallengeconcernstheskepticalinterestofherposition:‘Whethersuchatimidskepticalstanceisofanyinterestisatopicforanotherpaper’,IwriteinSSP,pointingto,butnotreallypressing,hermainproblem.16Aswewillseeinsection15,Iinfactdon’tthinkthetimidskeptic’strueclaimsareveryimportant,andsoIdon’ttakethe‘even‐handed’contextualistpositionthattheundifferentiatedskeptic(whohasnotyetconsideredthepossibilityofcontextualism)is‘trying’,asitwere,tomakeaveryimportantpointthatweshouldshowduereverenceto.

9. Kornblith’sAttack:Full‐Bloodedvs.HighStandardsSkeptics

HilaryKornblith’s‘TheContextualistEvasionofEpistemology’(Kornblith2000)providesaniceexampleofsomeofthekindsofmisunderstandingsthatthecontextualistapproachtoskepticismissubjectto—andithasbeenextremelyinfluential.17Itwillhopefullyfunctionasagoodtargetforpullingsomedoggedly

16That‘otherpaper’stilldoesnotexist.Itwould/willbeareworkingofChapter4(“Wittgenstein'sSuspicionandtheInsignificanceofPhilosophicalScepticism”)ofmydissertation(DeRose1990)—which,despitebeingperhapsmyfavoritechapterofmydissertation,istheonechapterthatIstillhavenotmadeintoanypublication.SomeoftheconclusionsIdraw—thoughnotthereasoningforthem—comeoutinsection15ofthischapter.

17ThoughthisdoesnotaffectthemaindisagreementsbetweenusthatI’mabouttoexplore,IshouldnotethatKornblithseemsalsotobeamongthosewhothinkofmeasthe‘nice’contextualistI’vebeendistinguishingmyselffromintheprevioussectionsofthischapter.Kornblith’spaperopenswiththis:

KeithDeRose’scontextualismoffersbothgoodnewsandbadnewsfortheskeptic.Thegoodnewsisthatwhentheskepticuttersthewords,“Idon’tknowanythingabouttheexternalworld,andneitherdoyou,”theskepticismakingatrueclaim.Thebadnewsisthatwhentheanti‐skepticuttersthewords,“Idoknowagreatdealabouttheexternalworldandsodoesalmosteveryoneelse,”theanti‐skepticismakingatrueclaimaswell.Putsomewhatdifferently:theclaimtheskepticmakesisirrelevanttotheclaimtheanti‐skepticmakes.(Kornblith2000:24)

‐20‐

persistentobjectionstocontextualistapproachestoskepticismoutbytheroot.Iwillrespondindetail(atspots,Isuppose,exhaustingdetail)toKornblith’sattack,butmostlyforthereasonthatdoingsoisagoodwayofclarifyingmycontextualistanswertoskepticisminawaythatshouldhelpinavoidingawiderangeofmisconstrualsofit.

Kornblithmincesnowords.Hispaperisafollow‐upto(Sosa2000)whereSosaexpressestheworrythatcontextualistsolutionstoskepticismaresimplyirrelevanttotraditionalepistemologicalreflectiononskepticism.ThisistheextremeandimportantworrythatImentionedatthebeginningofthischapterandwewilladdressitinsection15.But,aswe’llsee,Sosa,truetohisjudiciousnature,doesnotclaimthatcontextualistsolutionsareirrelevant;hejustprovidesgroundsforquestioningtheirrelevanceandthenwonderswhattherelevancemightbe.Kornblith,whotakesSosatobe‘extremelycharitableinhisdescriptionoftheepistemologicalimportofDeRose’scontextualism,’enthusiasticallytakesupthe‘lesscautious’stancethat‘DeRose’scontextualismisirrelevanttoepistemology’(Kornblith2000:25),andexplainswhyhethinksthatisso.

CentraltoKornblith’scaseisadivisionhedrawsbetweentwotypesofskeptics,whichislikemydistinctioninSSP(explainedinthesectionabove)inthatitsupposedlyfeaturesoneskepticwhoisbothmoreambitiousandinterestingthantheother:his‘Full‐Blooded’andhis‘HighStandards’skeptic.WewilldescribeKornblith’sskeptics,aswellashiseccentricVermonterwhomyou’reabouttomeet,inmoredetailsoon.ButnotefornowthatthebasicshapeofKornblith’scomplaintisthatIansweronlyhis‘HighStandardsSkeptic’,thesupposedlyuninterestingofhistwoskeptics:

Now my worry about DeRose’s contextualism can be put quite succinctly: DeRose responds only to the High Standards Skeptic, the skeptic who is an analog of my charmingly eccentric Vermonter and who acknowledges the importance and accuracy of substantive epistemological distinctions we wish to make. But this is a wholly trivial and uninteresting position. On the other hand, contextualism does nothing to address the Full-Blooded Skeptic, the skeptic who wishes to insist that all propositions about the external world are epistemologically on a par. But it is this latter skeptic who is making an historically important

ItseemsthatKornblithisimagininghisskepticandhisanti‐skepticasarguingwitheachother(basedontheskeptic’suseofthesecond‐personpronounandthentheanti‐skeptic’s‘Idoknow,’insteadofsimply‘Iknow’),inwhichcase,aswe’veseen,Idonotrulethatbothspeakersareassertingtruths.

‐21‐

and philosophically interesting claim. If skepticism is a position we need to worry about, it is Full-Blooded Skepticism, not High Standards Skepticism, which should concern us. The skeptical problem DeRose’s contextualism addresses is simply uninteresting from an epistemological point of view. (Kornblith 2000: 27)

WhileKornblith’shighstandardsskepticisperhapsrelativelyuninteresting,sheisalsodistinctfromthe‘boldskeptic’IexplicitlytakeasmymaintargetinSSP.AsKornblithdoesnotmentionmydistinctionbetweenthe‘bold’andthe‘timid’skeptic,muchlesshowhisdistinctionrelatestomine,IamunsurewhetherKornblithsimplymisunderstandsthenatureofmyannouncedtarget(thinkingheisdrawingthesamedistinctionasme,butjustusingdifferentlabels),orwhetherhe’sawareofthedifferencebetweenmy‘bold’andhis‘highstandards’skeptic,butthinksthat,despitemyintentions,myresponseinfactmissesitsintendedtargetandhitsonlythelessinterestingskeptichedescribes.WewillproceedinawaythatshouldanswerKornblith’sconcernsineithercase,andthatwillmakeuseofhisskeptics,aswellasanotherskepticthathisdiscussionwillpromptmetoaddtoourcastofcharacters,tomakeclearwheretheimportanceandinterestofcontextualistresponsestoskepticismlies.WewillstartbylookingattheconsiderablephilosophicalinterestIatleastthinkistobefoundeveninKornblith’shighstandardsskeptic;wewillthenseehowmyboldskepticdiffersfromthehighstandardsskepticinawaythataddsstillmoreinteresttoherposition;andthenwewillseehowmycontextualistresponsereallydoesanswertheboldskeptic.

10. ThePhilosophicalInterestoftheHighStandardsSkeptic

Kornblith’s‘Full‐Bloodedskeptic’isquitetruetohername(infact,Iwouldbetemptedtocallhersomethinglikethe‘wild‐eyed’skeptic),holdingthat‘wearenomorejustifiedinbelievingthatthereisanexternalworldthanthatthereisn’t’andthat‘Ihavenodegreeofjustificationwhateverformyclaimsabouttheexternalworld.None’(Kornblith2000:26).Bycontrast,hereishissupposedlyuninteresting‘HighStandards’skeptic:

If the skeptic agreed with us that there are differences in degree between the extent of justification which we have for various claims about the external world, but simply denied that we ever reach some very high standard required for knowledge, then skepticism would be a much less interesting claim. Indeed, imagine a skeptic who says the following: “Yes, I

‐22‐

agree that it is far more reasonable to believe that you are reading a paper in Oviedo than that you are standing in the middle of a road; the first is far better justified than the second. More than this, when it comes to deciding what to believe, there are significant differences in the degrees of justification for various propositions about the external world, and in a large class of cases, it is entirely unproblematic as to what one should believe. I simply deny that the level of justification one reaches in any of these cases is sufficient to call the resulting beliefs cases of knowledge.” This is not, of course, the skepticism of Descartes’ First Meditation; it is, instead, a much more modest and less exciting form of skepticism. Let us call this view High Standards Skepticism. (Kornblith 2000: 26)

Thoughit’snotabsolutelycleartome,ItakeitthatKornblith’shighstandardsskepticislikemyboldskepticinthatshemakesthe‘bold’(asIusetheterm)claimthatwearespeakingfalselywhenever,eveninordinaryconversation,weclaimto‘know’thingsabouttheexternalworld.ThatatleastishowI’minclinedtounderstandwhatKornblithmeanstobeclaimingwhenhesays‘Isimplydenythatthelevelofjustificationonereachesinanyofthesecasesissufficienttocalltheresultingbeliefscasesofknowledge.’Ifsheisnotinthatwaylikemyboldskeptic,herrelevancetomyproject,and,inparticular,whyKornbliththinkssheisthetypeofskepticthatmysolutionprovidesananswerto,becomescompletelyopaquetome.Andifthehigh‐standardsskepticdoesmakethisboldclaim,shestrikesmeasbeinginsomewaysI’llexplainquiteinteresting—andalsoinsomewaysbolderthantheDescartesoftheFirstMeditation,whomKornblithdoesfindquiteinteresting,but,Ithink,getsallwrong.18

18KornblithseemstobeinsinuatingthatDescartesisadvancingsomethinglikeKornblith’s‘full‐bloodedskepticism’intheFirstMeditation.HecertainlythinksthatDescartesisadvancingamorestridentskepticismthanmere‘highstandardsskepticism’.Now,DescartesisfrustratinglyinexplicitinMeditationOneabouttheforceoftheskepticismheintendstobeadvancing(inordertobattleinthelaterMeditations).ButIseenopositivebasisforthinkingDescartesintendstobeadvancinganythingnearlysowild‐eyedasKornblith’sfull‐bloodedskepticismanywhereinMeditationOne.Theforceoftheskepticismthatissupposedtobemotivatedbyconcernsaboutpossibledreaminginparagraphfiveisparticularlytoughtodiscern.Thekeystepsinthisargument(premises,conclusions)areleftlargelyinexplicit.Butoneclaimthatlookslikeastepintheargumentisexplicit:that‘therearenocertainindicationsbywhichwemayclearlydistinguishwakefulnessfromsleep’(Descartes1967:vol.1,146).Buildingonthat,oneguessesthatDescartesislikelysupposingsomepremisetotheeffectthatifthereareno‘certainindications’bywhichtomakethedistinction,andifperhapscertainotherconditionsaremet,thenanybeliefhemighthavetotheeffectthatheisawakeandnotdreaminglacksacertainstatus,andthenperhapssomeotherpremisetotheeffectthatiftheabovelackobtains,andperhapsifotherconditionsaremet,thenhisbeliefsinsuchordinarythingsasthatheisseatedbyafireplacealsolackacertainstatus—likelythesamestatus.Butwhatstatuscouldhehaveinmind?(Whatistheforceoftheintendedskepticism?)Ourbestclueseemstobeintheexplicitstep:Whatwould‘certainindications’(or‘suresigns,’byothertranslations)beneededfor?Thenaturalanswerwouldatleastseemto

‐23‐

Sometimeswejudgeaclaimtobe‘uninteresting’because,howeverimportantitwouldbeiftrue,wefindthechancethatitistruetobetooslight.However,Kornblith’schargeagainstthehighstandardsskepticseemstobethestrongeronethatevenifherclaimwereright,thatwouldstillnotbephilosophicallyimportantnews,asheseemstoindicateinthispassageinwhichhelikenshishighstandardsskeptictoaneccentricVermonterwhoispeculiarlydemandingaboutwhatconstitutes‘cold’weather:

What should we say to the High Standards Skeptic? On the one hand, I think it is quite natural to suggest that this skeptic has a deeply deviant view about the nature of knowledge, or at least, about how the term ‘knowledge’ should be used. But in practice, if confronted

be:inordertobecertainorsurethatoneisawake,andtheninordertobecertainorsureofthelikesofthatoneissittingbythefire,etc.IsupposeonecouldattributetoDescartespremisestotheeffectthat(ifperhapscertainconditionsaremet),oneneedscertainindicationsorsuresignsthatoneisawakeandnotdreamingtohaveanyjustificationwhatsoeverforthinkingthatoneisawake,andthentohaveanyjustificationwhatsoeverforthinkingthelikesofthatoneissittingbythefire.Butwhy?Whywouldcertainindicationsorsuresignsbeneededforthat?WouldwebedoingDescartesanyfavorsbyattributingsuchanargumenttohim?ThereareofcourseplentyofplacestoturntoforamoreseriouslookatDescartes’sdreamargument,butmyfavoriteis(Curley1978:46‐69),which(verysensibly,sofarasIcansee)rendersDescartes’sconclusionas:‘Noneofmybeliefsaboutordinary‐sizedobjectsinmyimmediatevicinityarecertain’(1978:52)—whichisaveryfarcryindeedfromtheskepticalconclusionthatnoneofthosebeliefshaveanydegreeofjustificationwhatsoever.Buthere,thoughIseenobasisforthinkingfull‐bloodedskepticismisbeingadvanced,andthoughIcertainlydon’ttakeDescartestobeadvancinganythingsowild,italsoseemsthatevensuchawildreadingcannotbesecurelyruledout.ButwhenwegettotheskepticismadvancedlaterintheFirstMeditation,seeminglywiderinscopebutevenmilderinforcethantheearlierdream‐inspiredskepticism,onebasedonskepticalpossibilitiesthatinMeditationThreeDescartescalls‘veryslight,andsotospeakmetaphysical’(Descartes1967:vol.1,159),theredoseemtobestrongpositivetextualgroundsforthinkingDescartesneverintendedtheskepticismbeingconsideredtobenearlyasstronginforceasfull‐bloodedskepticism—andalsogoodreasontothinkitisn’tevenmeanttobeasstridentashighstandardsskepticism,wherethatincludestheclaimthatordinaryclaimsto‘know’(orthecloseenoughLatinequivalent)gofalse.KeyhintstoDescartes’sintentionsaretobefoundinexplanationslikethe‘atheistgeometer’passageintheRepliestoObjections2,whereDescartesevaluatesthebeliefsofageometer,who,remaininganatheist,hasnotfollowedDescartes’stheisticescapefromtheskepticismestablishedinMeditationOne,andsowhopresumablyisstillsubjecttotheskepticismadvancedthere.Descartesexplains:‘Thatanatheistcanknowclearlythatthethreeanglesofatriangleareequaltotworightangles,Idonotdeny,Imerelyaffirmthat,ontheotherhand,suchknowledgeonhispartcannotconstitutetruescience,becausenoknowledgethatcanberendereddoubtfulshouldbecalledscience’(Descartes1967:vol.2,39).ImportanttoourcurrentconcernishowpositiveDescartes’sevaluationoftheatheist’sbeliefis,andhowweakinforceisskepticismthatDescartesthinkstheatheistissubjectto.Thisseemsweakerinforcethanevenhigh‐standardsskepticism.IhereuseatranslationofDescartesthatisparticularlyfriendlytothepointsIammaking,butonanyreasonabletranslation,thereseemslittleroomforthekindofunderstandingofDescartesKornblithispushinghere.See(DeRose1992)forfurtherdiscussionofhowIunderstandDescartesonthesematters.

‐24‐

with such a skeptic, it would probably be wise simply to capitulate. “Let us use the term ‘knowledge’ as you do,” I would say. Nothing much hangs on this. Since the skeptic agrees that we can make meaningful and important distinctions about how well justified we are in various claims, and agrees with us about which claims we should believe and act upon, nothing much turns on it. It is like dealing with the Vermonter who insists that he won’t say that it is cold outside unless it is at least 25 degrees below zero Farenheit. If he recognizes that there are important distinctions to be made in temperatures above minus 25, and that these distinctions have a bearing on how one should interact with the world, then the only difference between him and us is a bit of charming linguistic eccentricity. It would, however, be altogether different if this Vermonter thought that temperatures above minus 25 were all on a par, and had no differential physical effects. This latter character is more than a linguistic eccentric; he is making substantive claims about the world, claims which would be tremendously important if they were true. (Kornblith 2000)

This,andespeciallytheveryendoftheabovequotation,suggeststhatincontrasttowhatwemightcalla‘full‐bloodedVermonter’,Kornblithisjudgingthathis‘highstandardsVermonter’,and,byextension,hishighstandardsskeptic,aremakingclaimsthatwouldbeunimportanteveniftheyweretrue.(Afterall,ifKornblithwerethinkingintermsthatwouldmaketheincredibilityofaclaimsufficeforittobeuninteresting,surelyhisfull‐bloodedVermonter,eventhoughheismaking‘substantiveclaimsabouttheworld’,wouldbeamongtheveryfirstinlinetobewrittenoffasuninteresting!)

Butthatstrongerjudgmentseemssimplyincredibletome.PerhapsthiswillendupjustbeingapointonwhichKornblithandIareextremelyfarapartinourthinking(andinwaysInodoubtwon’tberesolvingheretothesatisfactionofthosewhoseethingsasKornblithdoes,thoughIwillsayabitaboutwhyIseethingsasIdo),butthisstrikesmeasimplausibleenoughtowarrantourlookingforanot‐so‐obviouswayofunderstandingKornbliththatmakeshispositionmoresensiblethanasurfacereadingwouldyield.AndhereIwouldnoteagainthatKornblith’spaperisaresponseto(Sosa2000),amainthemeofwhichisanattempttodisentangletheelementsofthecontextualistpositionthatconcernepistemologyfromtheaspectsimportanttothephilosophyoflanguage.PerhapsKornblithwouldagreethatifourclaimsto‘know’thingsabouttheexternalworldwerefalse,thatwouldindeedbenewsofconsiderablephilosophicalinterest,butthatinterestwouldbeconfinedtothephilosophyoflanguage,andthenewswouldnotbeimportanttoepistemology?Afterall,hedoeswritethingslike(nowwithemphasisadded):‘TheskepticalproblemDeRose’scontextualismaddressesissimplyuninterestingfromanepistemologicalpointofview.’Perhapshe’sthinkingthisnewsmightjustaffecthow

‐25‐

somepointsofepistemologyareexpressed,butwouldleavetheimportantepistemologyitselfunaffected,itsconsiderablephilosophicalimportancebeingconfinedtothephilosophyoflanguage?Orperhaps,onanotherreading,hethinksthatwhatphilosophicalimportancethereishereismostlyconfinedtothephilosophyoflanguage,buthealsothinksthatimportanceisnotconsiderableinanycase?

Atanyrate,thenewsthatallourclaimsto‘know’externalfacts,eventhosemadeinordinaryconversation,arefalsewouldcertainlybestartling—asindeedwouldbeanalogousnewsabout‘cold’.ButasIthinkofhowthisastonishingnewswouldaffectmyphilosophicalthinking,itsinitialimpactwouldcertainlyseemtobethehavocitwouldatleastthreatentowreakinthephilosophyoflanguage.Oneofthefirstthingstodoinresponsetosuchnewswouldbetotrytodeterminewhetheritwasduetosomeisolatedglitchaffecting‘know(s)’(and/or‘cold’),orwhethercommontermsofourlanguagemoregenerallyfailedinsucharadicalwayinallowingustomaketrueclaims(andhavingtruethoughts)abouttheworldbytheiruse.Butespeciallyasthethreattendedtowardthesecondofthesepossibilities,thenewswouldseemfarfromho‐hum.

AndIforoneamskepticalthattheresultingupheavalcouldintheendbeconfinedtothephilosophyoflanguage.ButthismaywellbebasedonwaysthatKornblithandIjustmovetothebeatofverydifferentphilosophicaldrums.Ibelievethatseparatingoutthefactorswhichaffectwhetherclaimsinvolvingphilosophicallyimportanttermsaretrueorfalsefromwhensuchclaimshaveotherpropertiesthatmightbeconfusedwithtruthandfalsityisanimportanttoolforunderstandingthemeaningsofthosekeyterms(andisatoolthatwouldbelargelydisabledforuseonatermonceahighstandardsskepticismaboutthattermwereaccepted),andthat,inturn,whileitiscertainlyveryfarfromallthatisinvolved,understandingthemeaningofthekeytermsinvolved—orperhapsmoreprecisely,avoidingradicalandprofoundmisunderstandingsofthemeaningsofthekeytermsinvolved—isimportanttoaddressingphilosophicalproblems.19Ifso,thentoechothethoughtsof

19Iamsometimesmentionedasacurrentphilosopher(suchcreatureswereapparentlymoreplentiful,oratleastmoreprominent,inearliertimes)whohas‘takenthelinguisticturn’inphilosophy—sometimesasaparticularlyclearexampleofsuch.I’mnotveryconfidentinmyunderstandingofthisdescription,butIsupposethatthesentencetowhichthisnoteattacheswouldbethekindofspottolookatindecidingwhetherthedescriptionfitsme.I’mguessingthatmyinclusionofthelimitingphrase‘whileitiscertainlyveryfarfromallthatisinvolved’mayharmmycandidacy.Butifthinkingthatunderstandingthemeaningofthekeyterms

‐26‐

myIntroductiontovolume1,wewouldbeboundtofallintoerroraboutknowledge,aswellasabout‘know(s)’,bydeprivingourselvesofsuchimportanttools.

Startinginthefollowingsection,IwillbedirectlyaddressingtheworriesthatseemmostbothersometoKornblithhimself.ButbeforeleavingthephilosophicalinterestofKornblith’shighstandardsskeptic,Ishouldnotetheimportantpointthat,whileI’vesofarbeenaddressinghowimportantthenewswouldbethatthisskeptic’sclaimswereright,thereisalsoanon‐hypotheticalinterestinthisskeptic’sposition,ifthatpositionincludesnotjustherincredibleclaim,butalsoargumentsforthatconclusion,insofarastheseargumentsarestrongandseemtopowerfullysupportherboldclaimthatwearealwaysspeakingfalselywhenweclaimto‘know’things.Ifthisskeptichasaformallyvalidargumenttotheconclusionthatwedon’treally‘know’eveninwhatseemstobeaparadigmaticcaseofexternalworldknowledge,andifthatargumentisdrivenbythepremisesthatcanseemboldlytrue,andespeciallyifwedoorareverytemptedtojudgethattheyareboldlytrue,thenthisissurelyphilosophicallyimportantnews,hopefullyexposingvariousgravemistakesregardingknowledgethatweareliabletomake.20

SimilarpointswouldholdforKornblith’shighstandardsVermonter.Howevershockingandimportantwouldbethenewsthathisclaimisright,whenthinkingnon‐hypothetically,I’minclinedtowritehimoffasaphilosophicallyunimportanteccentricifhe’sjustmakinghisclaimwithoutanyinterestingsupport.Butifwe’retalkingaboutahighstandardsVermonterwithapowerfulargumentforhisboldclaimthatwe’respeakingfalselywheneverwesayit’s‘cold’outsideandthetemperatureisaboveminus25degreesFahrenheit,thenthat’ssomeonethatI,asaphilosopher,wanttotalkto,evenifhispowerfulargumentisn’tnearlypowerfulenoughtorationallyeventemptustoaccepthisconclusion.Thoughwewillinthatcasehavearighttosupposefromtheoutsetthathisargumentmustgo(badly)wrongsomewhere,we’dstillbelikelytolearnsomethingimportanttryingtofigureoutwhereandhowitdoesgowrong.

involved,andespeciallyavoidingprofoundmisunderstandingsofthemeaningsofthekeytermsinvolved,canbeimportanttoaddressingphilosophicalproblemssufficesforonetocountashavingtakenthe‘turn’,thenIcertainlywanttopleadguilty.

20ThisisthevalueGrecoseesinthestudyofskepticalarguments;seen.16ofChapter2.

‐27‐

11. DoIRespondOnlytotheHighStandardsSkeptic?:TheHighStandardsSkeptic,theDeplorablePositionSkeptic,andtheBoldSkeptic

ChargeslikeKornblith’sclaimthat‘DeRoserespondsonlytotheHighStandardsSkeptic’havebeenmadebyothers,oftentakingtheformthatI,orthatthecontextualistrespondertoskepticism,failstorealizesomethingveryimportantabout(someof)themostimportantskepticalchallenges—namely,thattheskepticdoesnot(intheseimportantcases)thinkthatwefailtoknowbecausethestandardsforknowledgearesohigh,butratherbecausewefailtosatisfyevenordinaryorperhapsevenlowstandards.Thus,forexample,hereisRichardFeldmanalsolevelingjustthischarge:

[Contextualists] fail to come to grips with some important arguments for skepticism. This is because in at least some cases the leading idea behind skepticism is not that we fail to satisfy some extraordinarily high standards for knowledge but rather that, contrary to common belief, we typically don’t satisfy ordinary standards. This skeptical challenge gets remarkably little attention from the contextualists under discussion here, since they seem simply to assume that we do meet ordinary standards for knowledge. (Feldman 2001: 78)

And,handilycitingsomeofhispredecessors,KentBachhelpfullyputsthecharge(whichheendorses)inanutshell,impressivelycrystalizingtheconfusion:

Still, many philosophers have complained that contextualists do not really come to grips with the force and content of skeptical arguments (see Feldman, 1999, 2001; Klein, 2000; Kornblith, 2000; Sosa, 2000). The complaint is simple: the contextualist's attempt to marginalize skeptical arguments by restricting them to skepistemic contexts ignores the fact that skepticism denies that we have knowledge even by ordinary standards. The contextualist’s attempt to marginalize these arguments by restricting them to skeptical contexts ignores the fact that skepticism denies that we have knowledge even by ordinary standards. (Bach 2005: 68)21

ThisclearlyrepresentssomeprofoundmisunderstandingofwhatIatleastamupto—andI’mfairlyconfidentthiswouldbesoconcerningothercontextualists

21BredoJohnsenseemstohaveachargelikethisinmindaswell,here:‘[C]ontextualists(alongwithmanyothers)havebadlymisunderstood[theskeptic];accordingtotheskeptic,whataccountsformyfailuretoknowWisnotthatmybelieflackssomeextraordinarilyhighdegreeofepistemicstrength,butthatithasnoepistemicstrength’(Johnsen2001:394).

‐28‐

aswell.Formyintention,atleast,isclearlytodefeat,andnottoignore,theskepticwho‘deniesthatwehaveknowledgeevenbyordinarystandards’.

YetthereissomemerittoFeldman’sclaimthatcontextualists‘seemsimplytoassumethatwedomeetordinarystandardsforknowledge’:thatdoesseematleastafairlyaccurateobservationofhowactualcontextualists,includingmyself,engageskepticism.Sohowcouldcontextualistspossiblybeaimingtodefeataskepticwhosekeyclaimisonethatcontextualists‘simplyassume’iswrong?Icanseehowitcanlooklikecontextualistsmustbeignoringaskepticwhomakestheboldclaimthatwedon’tevenmeetordinarystandardsforknowledge.Andifcontextualistsareignoringthatskeptic,whatkindofskepticisleftforcontextualiststobereallyengaging?PerhapsjustsomethinglikeKornblith’shighstandardsskeptic?

WecanstartclearingupthismuddlebynotingtheslippagebetweenKornblith’sandmyschemesforclassifyingskeptics.IfKornblith’shighstandardsskepticsaysthatwe’realwaysspeakingfalselywhenweclaimto‘know’thingsabouttheexternalworld,whichisthecharacteristicclaimofmyboldskeptic,thenwhatistherelationbetweenherandmyboldskeptic?Answer:Kornblith’shighstandardsskepticisamorespecifictypeofcharacter:Sheisaboldskeptic,butthereareotherwaystobeaboldskeptic,andsomeofthemmightbemoreinterestingtoKornblith(andhisfellowcomplainers).

Kornblith’shighstandardsskepticthinksourclaimsto‘know’factsabouttheexternalworldarefalsebecause,whenitcomestosuchfacts,wenever‘reachsomeveryhighstandardrequiredforknowledge’.Butwhatofthedifferent,complementaryboldskeptic,pointedoutbyFeldman&Co.,whoalsothinkswenevermeetthetruthconditionsforordinaryclaimsto‘know’thingsabouttheexternalworld,butthinksthisnotbecausethosestandardsaresohigh,butratherbecauseshethinkswe’realwaysinsuchadeplorableepistemicpositionwithrespecttosuchexternalmattersthatwhenitcomestothemwedon’tmeetevenquitelowstandardsforknowledge?Callthisthe‘deplorablepositionskeptic’.22Thisistheskepticcontextualistssupposedlyignore,andshesoundslikeakindofskeptic

22Irealizethatit’seasytofeelthatoneislosingone’sgriponthedistinctionbetweenthesetwoskeptics(highstandardsvs.deplorableposition):Canthestandardsreallybesolowormoderateifwenever,oralmostnevermeetthem?ButIwillnotpressthatpotentialproblemformycritics,insteadshowingthat,insofaraswecanmakesenseofthedistinctionbetweenthesetwoskeptics,Ianswerbothofthem.

‐29‐

Kornblithatleastislikelytofindquiteinteresting.Sheismuchmorelikehisfull‐bloodedskeptic,whomKornbiththinksisofprofoundepistemologicalinterest,thanisthehighstandardsskeptic.Kornblithexplainshisfull‐bloodedskepticasonewhotrafficsinwild‐eyedclaimsabout(sometimescomparative)levelsofjustification—namely,thatwehavenoneofitatallforourexternalworldbeliefs,andnomoreforsomesuchclaimsthanforanyothers.ThedeplorablepositionskepticwejustmetrepresentsawaytofairlycloselyapproachKornblith’sfull‐bloodedskepticwhilestayingfocusedonknowledgeratherthanjustification,andsheseemstobeaskepticwhomKornblithwouldtaketobemakingsubstantiveepistemologicalclaims,andratherstrongonesatthat,ratherthanjustclaimsaboutlanguage,therebyperhapspiquinghisinterest.

So,howdoesthisdeplorablepositionskepticcomparewithmyboldskeptic?Shetooisaboldskeptic,everybitasmuchasisKornblith’shighstandardsskeptic.Thedeplorablepositionskepticandthehighstandardsskepticrepresenttwowaysofbeingaboldskeptic,andwhenIdefeattheboldskeptic,Imeantobedefeatingbothofthesecharacters.Myintentioncertainlyisn’tto‘respondonlytotheHighStandardsSkeptic.’

12. DoIRespondOnlytotheHighStandardsSkeptic?:TheSimultaneousDefeatofBoldSkepticsofBothKinds

Buthowmightsuchadefeatoftwosuchdifferentskepticsasthehighstandardsskepticandthedeplorablepositionskepticbeaccomplishedinonefellswoop?Andhowcanthecontextualist’stargetsforbattlingincludethesecondofthoseskepticsifthecontextualist‘simplyassumes’thatthisskeptic’skeyclaimiswrong?Areviewofhowthedefeatissupposedtobeaccomplishedwillanswerbothofthosequestions,startingwiththesecondofthem.

Recallfromthepreviouschaptersthenatureofthedefeatthatisbeingattempted,anditscontrastwiththeothermodeofcombattingskepticism.Thecontextualistisnottryingtorefuteanyoftheseskepticsbyderivingananti‐skepticalresultfromtheperhapsextremelymeagrestockofpremisesthatdonotbegthequestionagainsther.Ifthatwerethegoal,thenthewayinwhichthecontextualist‘simplyassumes’thatwemeetordinarystandardsforknowledge

‐30‐

indeedwouldshowthatdeplorablepositionskepticcouldnotlegitimatelybeamongthecontextualist’stargets.Butcontextualismhasneverbeenaboutthat–withrespecttoanyskepticaltargets.Itisinsteadbeingdeployedtowardtheverydifferent(andphilosophicallymoreimportant,Iclaim)goalofprovidingtheresolutionofthepuzzlethattheskepticalargumentpresentsuswiththatmakesthemostsenseofalltheintuitionsinvolved,whetherbyendorsingthemorbysuccessfullyexplainingthemaway,andtherebydefeatingrivalpositionsonthepuzzle,includingthatoftheboldskeptic.Inthatendeavor,itistobeexpectedthatvariousquestionswillbebeggedagainstrivals—justasitistobeexpectedthatthoserivals,includingthedeplorableconditionskeptic,willbegquestionsagainstthecontextualistMoorean.Wearesupposingthatadvocatesofthevariouspositionsonthepuzzlehaveprovidedwhateverpositiveargumentstheycanfortheintuitiveclaimsthattheyendorse,buthavehitrockbottom,findingnostill‐more‐powerfuldeeperclaimsonwhichtobuild,sowehavereachedthe(hopefully)intuitivestartingpointsofeachpositionthatwillbe‘simplyassumed’ratherthanarguedfor,andwearenowengagedindamage‐controlandexplanation,andfocusedontheaspectsofeachpositionthatarecounter‐intuitive.

Andrelativetothegoalcontextualistsarepursuing,boldskepticsofboththe‘highstandards’and‘deplorableposition’varietiesareextremelyvulnerabletosuchadefeat,duetofeaturesoftheirpositionthat,asboldskeptics,theyholdincommon.

Thetimidskepticdoesn’ttakeAItobeshowingthatwearewrongwhenwesayorthinkthatwe‘know’thatO.Hercharacteristicclaimisjustthatwedon’t‘know’bytheelevatedstandardsthatthepresentationofAIhasatendencytoputintoplace,whichseemsfarfromashockingclaim,especiallyafterweseeallthatcontextualismaboutknowledgeattributionshasgoingforit(asweseeinVolume1).Bycontrast,theboldskeptictakesAItobeshowingsomethingverysurprisingindeed,toputitmildly.‘Shocking’probablyisn’ttoostrongawordtodescribetheboldskeptic’spositionthatwehavealwaysbeenwrongwhen,inordinarysettings,wehavesaidorthoughtthatwe‘know’thisorthat.Andthisissowhetherourboldskepticisahighstandardsskepticoradeplorableconditionskeptic—orsomethingsomehowinbetweenthosetwo.

InlinewithwhatIurgeinsection10,thesurpriseofaboldskeptic’sposition,whichevervarietyofboldskepticsheis,portendswellfortheinterestofherposition.Shockingpositionsconcerningsuchthingsasknowledgeortruthare

‐31‐

boundtobeimportantnewsiftrue,andevennon‐hypothetically,powerfulargumentsforsuchshockingpositionsareboundtobephilosophicallyimportant.

Buttheshockingnatureoftheboldskeptic’sclaim,sohelpfultoherinterest,constitutesaveryseriouschallengetohersofarastherationalacceptabilityofherpositiongoes,forreasonswealsosawinChapter2.Sheisextremelyvulnerabletodefeat.Fornowshehastofacethequestionofwhyanyoneshouldrationallyfollowherreasoningtoitsboldskepticalconclusion,ratherthanacceptingsomeothersolutiontothepuzzlesheconfrontsuswith.ForwhatdoesthisboldAIskeptichavetostandon,otherthanhowplausibleherargument’spremisesare?AsI’veurged,theplausibilityofherpremisesisconsiderable—andmightevenbemorethanjustconsiderable.Buttheflip‐sideoftheshockingnatureofherconclusionisthatitsnegationwillalsohaveatleastaconsiderabledegreeofplausibility.AndsoourboldAIskeptics,ofboththe‘highstandards’and‘deplorableposition’varieties,facetheverychallengingquestionofwhyintheworlditwouldberationalforanyonetoacceptherstartlinglyboldconclusion,ratherthan,say,concludingwiththesubstantiveMoorean,that,atleastbythestandardsthatgovernourordinarythoughtandtalk,wereallydo‘know’thatnot‐H.ThisMooreanproposal,aswellasotheralternativestoboldskepticism,mayconstitutesomethingofanunexpectedintellectualsurpriseforsome,asourboldskepticsmightdoverywelltopointout,butmanymightquiterationallyfinditmuchlessincrediblethanthearrestingshockertheboldskepticsareaskingthemtoswallow.Inshort,aswehaveseen,whatwefacehereisapuzzle,andtheincrediblenatureofboldskepticismiswhatrendersit,inallitsvarieties,highlyvulnerabletodefeatbyothersolutionsthatmightcomealong.

So,whenacontextualistproposesasolutionthatheclaimsreallydoessolvetheproblem,byaccountingfortheplausibilityofalltheclaimsconstitutiveofthepuzzle(ratherthanjustembracingoneoftheinitiallydistastefulclaims),andarguesthathissolutionissuperiortothoseputforwardindefenseofboldskepticism,andclaimstohavetherebyrendereditrationallypreferabletoaccepthissolutiontothepuzzleoverthatoftheboldskeptic,itbetraysaprettybadmisunderstandingofwhatthecontextualistisattemptingtochargethathehasonlyrespondedtoonesub‐typeofboldskeptic.Thechargethatthecontextualistisnotansweringthedeplorableconditionskepticcouldmakesenseifthedeplorableconditionskepticweresomehowimmunetotheproblemsthecontextualistwaspressingagainstboldskepticism,and/orifthecomplainerwereputtingforwardaspromisingsome

‐32‐

proposedrealsolutiontoourpuzzlethatwasespeciallyavailabletothedeplorableconditiontypeofboldskeptic.Butthosemakingthechargedon’tseemtohaveanythinglikethatinmind.Or,atleastiftheydo,theyarenotveryforthcomingaboutwhytheythinktheproblemsbeingpressedagainstboldskepticismdon’taffectthedeplorableconditionsub‐typeoftheview,noraboutwhatsolutiontothepuzzle,especiallywell‐suitedtothedeplorableconditionvarietyofboldskepticism,theythinkispromisingandisbeingignored.

ItistruethatwheninSSPIexplorepossiblesolutionsavailabletotheboldskeptic,Ienduptryingoutasolutiontheadoptingofwhichwouldtendtomaketheboldskepticahighstandards,ratherthanadeplorablecondition,skeptic,andthismayhavecausedsomeoftheconfusionhere.23Butthat’snotbecausethedeplorableconditionskeptichasagoodsolutionuphersleevethatI’mawareofbutamchoosingtoignore.Rather,therelevantpartofSSP(sections15‐16)unfoldsasfollows.Inotetheproblemsthattheboldskepticfaces,and,thoughIdon’tbreakthingsdownintocases,theproblemsIpressaffectboldskepticsofbothvarieties.Ithennoteakindofmaneuverbywhichithasbeensuggestedthatboldskepticsmightdealwiththeirproblems,andactuallysolvetheAIpuzzle.Ithennoteacoupleofquitefundamentalproblemswiththatmaneuver,and,again,theproblemsIallegewouldn’tbeconfinedtotheuseofthemaneuverbyjustoneortheothersub‐typeofboldskeptic,butwouldaffectbothofthem.Ithennoteawaythataboldskepticmighttrytodealwithhernewproblems,anditisatthispointthatthepossibleescapeIconsiderontheboldskeptic’sbehalfwouldseemtopushhertowardbeingahighstandards,ratherthanadeplorablecondition,boldskeptic,Ithink.IthenarguethatmycontextualistsolutionprovesabettersolutionthantheoneIhavebeenledtoconsiderontheboldskeptic’sbehalf—andagain,theproblemhere,thoughit’satthispointbeingpressedagainstwhatlooksmostlikeahighstandardsskeptic,isonethatafflictsbothvarietiesofboldskeptic.So,atthecloseofthisstory,Iendupdealingwithaboldskepticwhoislookingmuchlikeahighstandardsskeptic.Butthatdoesn’tmeanthedeplorableconditionskepticisn’tbeingtargeted.Sheisstillamongtherivalswhoaresupposedtobedefeated.IhavesaidwhyIthinktheboldskeptic’ssolutionfails,andtheproblemsIbasethatconclusionon,upto

23Mysuspicion,however,isthattheconfusionwasinsteadlargelyproducedbythethoughtthatwe’vealreadyaddressedthatthecontextualistmustbeignoringthedeplorableconditionskepticbecausethecontextualist‘simplyassumes’thatsuchaskepticiswrong.

‐33‐

theend,areproblemsthatafflictthedeplorableconditionskepticinherattempt(ifsheevenmakestheattempt)tosolvetheAIproblemaswellasthehighstandardsskeptic.Thereasonthedeplorableconditionskepticgetsneglectedalittlebitattheendofthestoryisbecauseitisherfellowboldskeptic,andnother,whoseemstometohaveapotentiallypromisingmoveyettotry.

Ofcourse,itispossiblethattherearepromisingdamage‐controllingmovesavailabletotheboldskeptic,andperhapsmorespecificallytothedeplorableconditionskeptic,thatIhaven’theardofandjustcan’tsee.Ifthat’sthecomplaint,thenthewaytopressitwouldbetopresentthosemoves.AsIadmitinSSP,‘Likepracticallyanyclaimtohaveprovidedthebestexplanationofsomething,myclaimhereishostagetothepossiblefuturedevelopmentofabetterexplanationcomingalong.’ButifIammissingsomepromisingpossibilityforthedeplorableconditionskeptic,I’mnotdoingsoknowingly.Sheisatleastnotbeingintentionallyignored.Ifshehaspromisingmovesyettoplay,weneedtobetoldwhattheyare.AndsheisamongtherivalswhoaresupposedtobedefeatedinthewayIhaveexplained.ItistheverydifferenttimidskepticwhoissetasideandnottargetedinSSP.

Infact,somemightwellobjecttomyclaimtohavedefeatedtheboldskepticsofeithervarietyongroundsquiteunfriendlyindeedtothoseskeptics:thatboldskepticismwasalreadypre‐defeatedindependentofmyanti‐skepticalefforts—orthoseofanyotheranti‐skeptics,forthatmatter.Thesheerincredibilityofthepositionhasdefeatedit,faraboveanyepistemologist’spoorpowertoaddordetract.Thiswillbethestanceofthosewhomaketheir‘Mooreanchoice’quitedefinitivelyagainsttheboldskeptic,findinghersolutionfarinferiortosomeotheronoffer(e.g.,bare,bullet‐bitingMooreanism),andwhoaren’timpressedenoughbythepotentialofthekindsofdamage‐controlmaneuversavailabletoboldskepticismtobegintoevenbegintotemptthemto,ontheirbasis,makean‘enlightenedMooreanchoice’that’sanydifferentfrom,ormoredifficultforthemtomakethan,theiroriginal,unenlightenedchoice.Forthem,theboldskepticdidnotneedtobedefeatedbythedevelopmentofsomeothersolutionand/oranyargumentsabouttheviabilityofdamage‐controlmaneuvers.Still,theytaketheboldskeptictobedefeated,andtheymaystillbeinthemarketforanexplanationofhowthispuzzlearises,eventhoughthestancetheytaketowardboldskepticism,atleastinsofarasit’smotivatedbyAI,doesnotdependonanyunderstandingthatsuchanexplanationmightprovide.Ifonewithsuchstridentanti‐skepticalleaningsweretofindasolutionlikeminetobebothsuccessfulandthebestonoffer,Iimaginetheycould

‐34‐

seetheirwaycleartowardapplyingaperhapssomewhatextendeduseofournotionof‘defeat’towhatthe‘winning’solutiondoestorivalpositions,includingherethatoftheboldskeptic,evenifthosesolutionswerealreadydefeated(inastrictersense)beforethesolutioninquestionwasdeveloped.

13. MySupposedlyThoroughlyExternalistResponsetotheFull‐BloodedSkeptic

ThoughKornblithcomplainsaboutmyignoringhisfull‐bloodedskeptics,atonepointhesaysIhavearesponse,butclaimsthatwhatIresponseIhavehasnothingtodowithcontextualism:

DeRose does have a response to Full-Blooded Skepticism, but it does not involve his contextualism. DeRose, in addition to being a contextualist, is an externalist about justification. And DeRose’s treatment of what he calls the “strength of one’s epistemic position” is a familiar externalist account. On such a view, the skeptic is simply mistaken when he claims that the conditions for knowledge cannot ever be fulfilled; he is mistaken in thinking that in a situation of the sort we ordinarily take ourselves to be in when looking at a table (in standard conditions with properly functioning perceptual and cognitive equipment), we are no better justified in believing that we are looking at a table than that we are looking at a toaster. The externalist account of justification, or strength of epistemic position, is a familiar and controversial one. It may or may not be right. I myself am very sympathetic with externalism. But it is important to recognize that it is the externalist part of DeRose’s view which is doing the work in combatting Full-Blooded Skepticism. Contextualism does no work here. (2000: 27)

First,acoupleofsmallpoints.Strictly,beforethisbook,Ihavenot(atleastinprint)explicitlydealtwithKornblith’sfull‐bloodedskeptic,asthatcharacterisdistinctfrommyboldskeptic,andasIhaven’texplicitlyaddressedskepticismaimedatjustificationratherthanknowledge.(IwilladdressquestionsofjustificationinChapter7ofthisvolume.)ButitturnsoutthatI’maninternalistaboutepistemicjustification—notthatKornblithshouldhavebeenabletotellthat(thoughIdon’tknowwhyhethoughthecouldtellI’manexternalistaboutthat).However,aswe’venowseen,IdotakemyselftoberespondingtoaskepticnotthatfarfromKornblith’sfull‐bloodedskeptic:ourdeplorableconditionskeptic,who,asI’ve

‐35‐

noted,representsawaytocloselyapproachKornblith’sfull‐bloodedskepticwhilestayingfocusedonknowledgeratherthanjustification.

Andherewecometotheimportantpoint:Ihaven’trespondedtoanyskepticinthewayKornblithallegesabove.ThoughIstressthatitis‘roughandready’,IdouseanaccountofknowledgeinconstructingmyexplanationofhowtheAIpuzzleisgenerated,andtheninmyattempttosolvethatpuzzle.Andthisaccount,whichI’vesincecometocalla‘double‐safety’account,doesarticulatewhatcanwellbethoughtofasan‘externalist’conditionforknowledge.Hence,itwouldbeanexternalistaccountofknowledgeifit,together(onesupposes)withthestandardconditionsonknowledge(truebelief)weretakenasacompletetheoryofknowledge.Anditwouldcertainlymakefora(veryimplausibly,tomysensibilities)externalistaccountofjustification.IfIhadsimplyputforwardsuchanaccountofknowledgeorjustification,andthenrejectedsomeformofskepticismaboutknowledgeorjustificationbecausesuchskepticismwasinconsistentwithmytheoryofknowledgeorjustification,together,Isuppose,withmyviewofwhattheworldislike(tosecuretheresultthatworldswhereIgowrongwithrespecttoOreallyareallquitedistant),thenKornblith’sreactionwouldbeon‐target.Butthat’snotwhatI’vebeenupto.(That’stotakemetobeengagingwithskepticismmoreinthewaythatNozickdoes.24)Myresponsetoboldskepticshasnotbeen:‘Here’smycomplete(andthoroughlyexternalist)theoryofknowledge;andlook,givenmyviewofwhattheworldislike,wedomeeteveryconditionforknowingthatO,atleastbyordinarystandardsforknowledge,thatmytheoryposits;so,you’rewrong.’Myfocushasnotinthatwaybeenonexplaininghowweknow,butratheronwhyitcanseemthatwedon’t.Somyresponsetotheboldskepticsisbetterthoughtofasbeingalongtheselines:‘Onmyview,ofcourse,you’rewrong;asIamonyourview.ButIcanonmyviewexplain(away)thesourceofattractiontoyourview,likethis….Whatdoyougot?’It’sinthetask(crucialtotheprojectofdefeatingtheboldskeptics,aswe’reusingthatterm)ofexplaining(away)theintuitivepoweroftheskeptic’sargumentthatmyaccountofknowledgeisutilized.Itsrolethendoesnotrequireittobeacompleteorfinishedtheoryofknowledge(thankgoodness!),noreventobeclosetoacompletetheory,sothatoursatisfyingitwouldatleastprovidegoodreasonto

24Seeesp.thefirsttwosectionsofChapter7foramoresustainedexplanationofthedifferencebetweenwhatItherecharacterizeasNozick’spositiveexplanatoryapproachtoskepticismandthenegativeexplanatoryapproachIamattemptingintheworkthatKornblithiscritiquing.SeetherestofChapter7formyownattempttonowaddapositiveexplanationtomytreatmentofskepticism.

‐36‐

supposeweprobablydoknow,butjusttoprovidethematerialsfortheneededunderminingexplanation.Myuseofthedouble‐safetyaccountdoesn’tthencommitmetoaradicallyexternalistviewofknowledge,butjusttoanexternalistcomponentofknowledgebeingabletoplaythecrucialrolemysolutionassignstoitinaccountingfortheintuitivepoweroftheskeptic’sclaims.25

MovingnowtoresponddirectlytoKornblith’skeycharge:OnceweareclearonhowIrespondtoboldskeptics,wecanseehowwrongitistosupposethatit’ssomeexternalistaccountofknowledgethatprovidesmyresponsetoanyskepticsinawaythatdoesnotinvolvemycontextualism.Fortheaccount’sonlyroleinmyresponseistoprovideanexplanationofhowtheskeptic’sargumentcanmakeitseem(tosomeextent)thatwedon’tknow—toexplain(away)theintuitivepoweroftheskeptic’scase.AnditcannotplaythatroleinanythinglikethewayIemployitwithoutthecontextualism.It’smycontextualismandmy(partialandrough‐and‐ready,but,Iclaim,enoughtogetthejobdone)accountofknowledgeworkingtogetherthatprovidethecrucialexplanation.Neitherworksalone.

Ofcourse,onecouldwonderhowwelltherough‐and‐readyaccountofknowledgeIusecouldbeadvancedintoamorecompleteaccountofknowledgeandmadetofitwithaplausibleaccountofjustificationwhilestillallowingforthekindofexplanationIgiveinresponsetoskepticism.ThesequestionsaretakenupinChapter7,whereInowprovidethekindofpositiveanti‐skepticalexplanationofhowweknowwhatweclaimtoknowthatIhavebeenmisunderstoodasgivinginmypastwork(mainly,SSP).

14. IndexedAI

WenowturntoimportantcriticismsofthecontextualiststrategyleveledbySosa.WestartwithachallengebasedonamodifiedversionofAIthatSosaputsforward

25ThesupposedlyexternalistnatureofmytheoryofknowledgeisamainthemeofBredoJohnsen’saccountofmytreatmentofskepticismat(Johnsen2001:392‐4).ThoughIdonotheredirectlyrespondtoJohnsen,IhopethatmyresponsetoKornblithinthissectionmakesclearhowIwouldanswerJohnsen’srelatedconcerns.

‐37‐

(Sosa2004:50‐51).Usingthesubscript‘o’sothat‘knowso’designatesknowingbythestandardsforknowledgethattendtogovernordinary,non‐philosophicalusesoftheterm,theargumentisjustlikeAI,exceptalloccurrencesof‘know’areexplicitlyspecifiedasbeinggovernedbyo:

Indexed AI 1i. I don’t knowo that not-H. 2i. If I don’t knowo that not-H, then I don’t knowo that O. So, Ci. I don’t knowo that O.

Sosathinksthisargumentdoesaboutaswellastheoriginalinmotivatingsomeoftheskeptic’smostimportantworries,butisimmunetothecontextualist’sresponsebecauseinthisversionoftheargument,the‘contextualtermshavebeenrelativizedaway.’

ThecontextualistMoorean(likeme)willholdthatIndexedAIisunsound,andthatitgoeswrongatitsfirstpremise,whichissimplyfalse.Thechallengeforuswouldbetoaccountfortheplausibilityofthatpremise.Though,forthereasongiveninnote28,Idon’tthink1iisasintuitivelyattractiveas1,itisnonethelessquiteplausible,I’minclinedtoagree,sothisisarealchallenge.

Butitisnotreallyanewchallenge.Atleastforme,thiswasthecrucialpartofhowIviewedthechallengeIinitiallysetouttomeet,andtheabilityofthecontextualist’sresponsetoaccountforjustwhat’satissuehere—howAI’spremises,and,here,focusingonitsfirstpremise,canseemtobetruebyordinarystandardsforknowledge—groundedtheimportanceofthecontextualiststrategyinthefirstplace.AsIwroteinsection2ofSSP:

The contextualist strategy is important because AI initially seems to threaten the truth of our ordinary claims—it threatens to boldly show that we’ve been wrong all along in thinking and saying that we know this and that. For it doesn’t seem as if it’s just in some “high” or “philosophical” sense that Al’s premises are true: They seem true in the ordinary sense of ‘know’. In fact, one is initially tempted to say that there’s no good sense in which I know that I’m not a BIV or in which I can know I have hands if I don’t know that I’m not a BIV.

Onmyaccount,it’stheinsensitivityofanybeliefwemighthavethatnot‐Hthataccountsfortheplausibilityoftheskeptic’sdenialthatweknowthatnot‐H.ThoughIdonottakesensitivitytobearequirementforknowledge(whichbecomesimportantwhenAI’ssecondpremiseisaddressed),ontheskeptic‐friendlyversionofcontextualismIprovisionallyassumedinSSP,thisinsensitivitysecures,viathe

‐38‐

‘RuleofSensitivity’,theresultthatanyattempttoclaimto‘know’thatnot‐H(oranyattributionof‘knowledge’ofnot‐Htosomepersonotherthanthespeaker)willbefalse,andanyattempttoadmitthatonedoesnot‘know’thatnot‐H(orthatsomepersonotherthanthespeakerdoesnot‘know’thatnot‐H)willbetrue.Thisaccountoftheplausibilityofthepremisewouldnotleadustoexpectthatthepremisewillbeattractivetousonlywhenweareunderstandingittobeinvokingsomespecial,philosophicalinterpretationof‘know(s)’;afterall,itpredictsthatanyclaimto‘know’thatnot‐H,inanykindofconversationalsetting,willgofalse(andlikewise,anyadmissionthatonedoesn’t‘know’thatnot‐Hwillcomeouttrue).Ifwetaketheuseofordinarystandardstobenecessarytoacontext’scountingas‘ordinary’,thisaccountimpliesthatnocontextinwhichsuchaclaim(oradmission)ismadewillremainordinary.Still,itimpliesthatinanycontextinwhichtheclaimismade,nomatterhowmundane(atleastuptothepointwhentheclaimismade),makingtheclaimbackfires.Insofarasourjudgmentaboutthetruthvalueofaclaimisinfluencedbyoursenseofwhetheritwouldbetrueorfalsetoactuallymaketheclaim,thisaccountseemswell‐suitedforexplainingevenwhyitcanseem(atleasttosomeextent:theappearanceherehasdifferentstrengthsfordifferentpeople26)thatthereisnogoodsenseinwhichweknowthatnot‐H,sinceitpredictsthatthereisnocontextinwhichonecantruthfullymaketheclaim.

Theabove,though,isjustmyaccountundertheprovisionalassumptionofaskeptic‐friendlyversionofcontextualism.Onamoreserious(atleastbymylights)versionofcontextualism,itwon’tpredictthatjustanyclaimto‘know’thatnot‐Hwillbefalse.However,itwillstillimplythatanysuchclaimwillhaveaforceworkingagainstitstruth—aforcewhich,whileitmayonoccasionbeoverpoweredbysomeotherconversationalforce,wouldresultintheclaim’sfalsehoodwereitunopposed.Aclaimtoknownot‐H(whereone’sbeliefthatnot‐Hisinsensitive)willthenlikelyneverseemunproblematic.What’smore,onsomeoftheless‐skeptic‐friendlyviews,includingtheoneI’minclinedtoward(DeRose2009:144‐8),allitwouldtaketomakeaclaimto‘know’thatnot‐Htogonon‐true(evenifnotfalse)issomeskepticstatingthatyou‘don’tknow’thatnot‐H.Itwouldbeeasyforustowriteoffanysensewemighthavethatsuchaclaimhassometruth‐likevirtueasbeingillusory.Afterall,whatkindoftruthdisappearsuponmeetingverbalresistance?Ofcourse,wenowknowthatoncontextualism,nothingthatcan

26Seeesp.sections6‐8ofChapter2.

‐39‐

properlybecalled‘knowledge’appearsorvanishes,dependingonwhatkindofreceptionone’sclaimto‘know’meets(unlessthereceptionone’sclaimmeetsrocksone’sbeliefinthepropositioninquestion):Regardlessofcontext,oneknowsothatnot‐Handdoesnotknowhthatnot‐H.Itisonlytheabilitytoexpressatruepropositionbymeansofagivensentence,andnotthetruth‐valueofanygivenpropositionthatmightbesoexpressed,thatisreallyinjeopardyhere(DeRose2009:Chapter6).Still,priortoallthattheoryandclarification,wecanbeexpectedtolookaskanceatclaimswhosetruthrequiresacertainreception(oratleastthelackofahostilereception)byinterlocutors.27Sincemyaccounthasitthatthereisalwayssomeimportantforceworkingagainstthetruthofanyclaimto‘know’thatnot‐H(whereHiswell‐chosen),andsinceeffortstoresistthisforcewilllikelystrikeusasatleastsomewhatsuspicious,itisthereverseofsurprisingthatwewillquitegenerallyfindthetruthof‘Iknowthatnot‐H’tobeatleastproblematic.Thusitisnosurpriseonouraccountoftheplausibilityoftheskeptic’sfirstpremisethatwewouldhaveatleastsometendencytojudgethatthatpremiseistrueevenbyordinarystandardsforknowledge,orevenbyanygoodstandardsforknowledge.

Thus,thecontextualistsolutiontoAIcanalsoaccountforIndexedAI.Infact,itwasallalonginanimportantwayanattempttoaccountfortheformerinawaythatcouldalsoaccountforthepowerofthelatter.28

27ThissuspicioncanandIthinkdoesflyinbothdirections(atleastondifferentoccasions).ThosewithastrongsenseofthetruthoftheirordinaryclaimstoknowthatO(thosewithcertainstrong(substantively)Mooreantendencies)mayfindtheskeptic’sdenialsdeeplysuspiciousinlargepartbecausethosedenialscanbetrueonlyincontextsmeetingcertainconditions.

28Ofcourse,thisisn’ttosaythatthetwoargumentshavethesameamountofintuitivepower,generallyorspecificallyattheirfirstpremises,andIshouldnotethatIdon’tthinkthatresponderswillgenerallyfind1itobeasplausibleastheywouldhavefound1tobe.IndexedAI,alreadyatitsfirstpremise,willraisethepossibilityofcontextualisminawaythattheoriginalskepticalargumentdoesnot.OnepresentedwiththeoriginalversionofAImightnotconsiderthepossibilityofcontextualism,andsomightbestronglyinclinedtopresumethattheyeithersimplyknowthatnot‐H,orsimplydon’tknowthat.Sucharespondermaybemoreeasilycoaxedintoaccepting1than1i,butmaystillbetaking1tobeboldlytrue,inthatiftheyjudge2andthenCalsotobetrueinthesamespirit,theywilltakethistoimplythatthey’vebeenwrongallalonginthinkingandsayingthatthey‘know’thelikesofO.Bycontrast,ifsuchresponderispresentedinsteadwith1i,thiswilltendtoraisethepossibilityof(atleastsomethinglike)contextualismasaconsideration.Thismaymakethemsomewhatlessinclinedtoaccept1ithantheywouldbetoaccept1.

‐40‐

15. IrrelevanttoTraditionalEpistemologicalReflectiononSkepticism?

WenowturntoSosa’sworrythatcontextualistsolutionstoskepticism,includingmine,aresimplyirrelevanttotraditionalepistemologicalreflectiononskepticism.AsI’venoted,Sosadoesnotoutrightclaimthatthecontextualistsolutionisirrelevant,butisledtowonderwhattherelevanceis(Sosa2000:3‐4).Thisworryofcourseinvolvesnotjustwhatthecontextualististryingtoachieveandhow,butalsohowthatrelatesto‘traditional’approachestotheproblemofskepticism.Nowthatwe’vegottenclearonwhatthecontextualistresponsetoskepticismseekstoaccomplish,wearewellpositionedtoaddressthisworry.

Itturnsoutthatit’sonlyonaverynarrowconstrualofwhattraditionalreflectiononskepticismhasencompassedthatthisworryisrealized—animplausiblynarrowconstrual,ifyouaskme,butifitwereaccurate,thatwouldjustmeanthattraditionalreflectionwasn’tallthatvaluable.

Sosacompares

(a1) People often utter truths when they say “Somebody loves me.”

(a2) Does anybody love me?

with

(c1) People often utter truths when they say, “I know there are hands.”

(c2) Do people ever know that there are hands?,

adding:‘c2ispresentedasaquestionwemightposeinphilosophicalreflection,inaphilosophyjournalorconference.’Adistraughtpersonledtoaska2ishopingtobereassuredbytheanswerthatindeedsomebodydoeslovethemor,failingthat,islookingtohavetheirworstfearstothecontraryconfirmed;a1miserablyfailstoprovidetheneededcomfort.Inphilosophicalreflectiononskepticism,inphilosophyjournals,andinphilosophyconferences,manyhavebeenmovedtoask(thelikesof)c2.Whyisc1anymorerelevanttoc2,askedinaphilosophicalsetting,thana1istoa2?

Let’sforthetimebeingmakethe‘skeptic‐friendly’assumptionthatphilosophicaldiscussionsofskepticismaregovernedbyexceedinglyhighepistemic

‐41‐

standardsaccordingtowhichnobodyknowsthattherearehands.(Wecancallthese‘absolute’standards.)Asyounowknow,I’mintheendinclinedtorejectthatassumption,butmakingtheassumptionhelpstomakethequestionofrelevancemorepointed,andyoubynowknowhowtoconvertwhatwesaywhileworkingundersuchanassumptiontowhatIreallythink.Underthatassumption,whenc2isaskedinphilosophicaldiscussionsofskepticism,thetruthfulanswertoit,accordingtocontextualistsolutionstoskepticism,isthepotentiallydistressing‘no’.Thepressingquestion:Whyisc1anycomforthere—anymorethana1isinthedistressingsituationinwhichtheanswertoa2isthedistressing‘no’?

Acrucialdifferencebetweena1/a2andc1/c2isthattherelevant‘context‐sensitivity’ina1/a2—that‘me’refersoneachoccasionofusetothespeaker,andsoreferstodifferentpeopleasitisspokenbydifferentpeople—isclearenoughtoallthatnobody(well,exceptforthelikesofyoungchildren,whocanbecaughtupinvarious,oftenquitecute,‘paradoxes’here)islikelytobemisledintothinkingthata1impliesthattheanswertoa2ispositive.Bycontrast,ifthecontextualisttreatmentofskepticismiscorrect,itisonlycontroversiallyso,anditisfarfrombeingcleartoall.29Undercontextualistanalysis(andforthetimebeing,askeptic‐friendlyversionofit),theskepticraisesthestandardsforknowledge,andoursensitivitytoherstandards‐raisingmaneuversinclinesustoatleastsomeextenttogivenegativeanswerstoquestionsaboutwhetherthereisknowledgeinvariouscases.Atthesametime,however,wearesensitivetothefactthatit’susuallycorrecttosaythatpeopledoknowallsortsofthings—i.e.,tosomeextent,wearestillinclinedtorealizethatc1.Notrealizingthatthoseusualaffirmationsarecompatiblewithskepticaldenialsofknowledge,wenaturallygetconfused.

Sofar,thatallsoundsveryskeptical!(Thoughthisisinpartduetoourworkingundertheskeptic‐friendlyassumption.)Thecorrectanswertoc2,aswe’vebeenaskingitinphilosophicaldiscussions,is‘no’,andourtendencytothinkotherwiseisbasedonaconfusion—theconfusionofthinkingthatc1impliesthatc2shouldreceiveapositiveanswerinphilosophicalsettings.

29Ifyouthinkthecontroversyandlackofclarityherethemselvesfavorinvariantismovercontextualism,notethatit’salsoverymuchthecasethatifcontextualismandcontextualisttreatmentsofskepticismarewrong,theyareonlycontroversiallyso,andrecalltheargumentof(DeRose2009:159‐60,177‐79).

‐42‐

Butcrucially,theconfusion,ofcourse,pushesbothways.Notonlydoesc1pushustowardthinkingthatc2shouldbeansweredpositively,butwecanalsoeasilythinkthatanegativeanswertoc2threatenstoimplythatc1isreallyfalse.Andthat,Ithink,isalargepartofwhyanegativeanswertoc2(madeduringphilosophicaldiscussion)canseemsomenacingandsoimportant—orperhapsevenwhythatanswerseemssignificantlyimportantatall.Imean,really,doesn’tatleastmuchofthesensedthreatandinterestofskepticismcomefromthethoughtthatperhapswe’vebeenwrongallalonginthinkingandsayingthatwe‘know’variousthings?Ifitinsteadturnsoutthatalmostallourthoughtsandassertionstotheeffectthatwe‘know’variousthings—includingveryseriousthoughtsandassertionsmadeinveryserioussettingswhereappropriatelyhighstandards(butnotyetphilosophical,absolutestandards)forknowledgeholdsway—arereallycorrect,andespeciallyifitturnsoutthattheskepticneverevenmeanttobesuggestingotherwise,skepticismlosesmostofitssting,andinterest,atleastforme.30Thereinliesthecomfortofthecontextualistresponse.

Isthatsufficientcomfort?Howmuchskepticaldistressremainsunaddressed?Thatdependsonhowimportantitisthatweknowaccordingtothe‘absolute’standards.Inanextremecase,ifthehazardtoourabsoluteknowingistheonlythreatofskepticismthatonefindsatallimportant,onewillderivenocomfortatallfromthecontextualistsolution.Iofcoursedon’thavethatextremeviewandseethisverydifferently.IhaveelsewhereexplainedwhatIthinkitwouldtaketomeettheabsolutestandardsinquestion,andonmyproposal,itissimplyimpossibleforanyconceivablebeingtomeetthosestandardsforanybeliefitmighthold(DeRose1990:292‐309).Butasmyaccountofthatwillbecontroversial,Ishouldregisterherethat,quiteindependentlyofmyparticular,controversialaccountofwhatitwouldtaketomeettheabsolutestandards,formypart,oncetheskeptical

30ComparethisfromCohen,whoseattitudeIamhereatleastfairlycloselyfollowing:‘Whatistrulystartlingaboutskepticism,istheclaimthatallalong,inourdaytodaylives,whenwehaveclaimedtoknowthings,wehavebeenwrong—wehavebeenexpressingpropositionsthatareliterallyfalse.Iftheskeptic'spositionisinteresting,itisbecausehechallengesoureverydayknowledgeattributions.Thetheoryofrelevantalternativesaddressesthechallengebyarguingthatoureverydayknowledgeattributions—properlyinterpreted—arecorrect.Thepropositionsweactuallyexpressandhavebeenexpressingallalongareliterallytrue.Wedoknowrelativetothestandardsthatordinarilygovernthoseattributions.Nowitisofconsiderablylessinterest(althoughnotdevoidofinterest)thatthereissomeotherpropositioninvolvingstricterstandardsthatisfalse.Itisnotthecasethatweknowrelativetoskepticalstandards.Thisafallibilistreadilyconcedes’(Cohen1988:117).

‐43‐

strategyisseentohavenotendencytoshowthatanyofmyclaimsto‘know’—exceptthoseveryrareonesmadeinsettingsgovernedby‘absolute’standards—areinanywaywrong,andonceIstarttogetalookatwhatitwouldtaketo‘know’accordingtotheskeptic’sabsolutestandards,onanyreasonableaccountofthat,Ifindthedistresscausedbymyfailuretomeetthosestandardstobeveryslight,andsotospeakmetaphysical(toputthingsinproperlyCartesianterms)—perhapstobecomparedwiththe‘distress’producedbytherealizationthatI’mnotomnipotent.(Andifmyabove‐mentionedsuspicioniscorrect,itisperhapswellcomparedwiththedistresscausedbytherealizationthatI’mnot‘omnipotent’onaconstrualofthatattributeonwhichitisincoherenttosupposeanyconceivablebeingcouldpossessit).

c1doesnottellushowc2,interpretedaccordingtoabsolutestandards,shouldbeanswered—anymorethana1tellsonehowtoanswera2.Iftraditionalepistemologicalreflectiononskepticismwerestrictlylimitedtothenarrowquestionofhowtoanswerc2,whenit’sinterpretedintheabsolutewayinwhichitseemstobeusedonlyinquiterarediscussionsofskepticism,ifeventhere(wearestilltodecidewhetherourskeptic‐friendlyassumptionisright),thenSosa’sworrywouldbevalid:Thecontributionofthecontextualistsolutiontothatprojectwouldbetotellusthattheskeptic’sanswertotheloneimportantquestionoftraditionalinquiryaboutskepticismistherightoneandthatanytendencywemighthavetorejecttheskeptic’sanswerthat’sbasedonfactslikec1ismisguided.Ofcourse,thecontextualistsolutionmightverywellalso,inthatcase,leadustoconcludethattraditionalinquiryintoskepticismhasnotbeenconcernedwithaveryimportantquestion.Iwouldcertainlybeinclinedtojudgeso.

Butsurelytraditionalphilosophicalinquiryintoskepticismhasbeenconcernedmorethanthatunimportantquestion!Insofarasphilosophicalinquiryintoskepticismconcerns,notjustthequestionofhowc2,interpretedabsolutely,shouldbeanswered,butalsoconcernstheimportanceofthatquestion;insofarasitaddressestherelationbetweenphilosophicaldenialsof‘knowledge’andordinarythoughtabout‘knowledge’;and,crucially,insofarasphilosophicalinquiryintoskepticismseekstoaddressthetrulymenacingthoughtthatc1mightnotbecorrectafterall,thecontextualistsolutionisfarindeedfromirrelevant.Andinsofarasthere’sroomintraditionalphilosophicalinquiryintoskepticismforthatmenacingthoughttobeaddressedbyaninvestigationofwhethertheskeptic’sargumentrendersitreasonabletoacceptthatdisturbingthought,then,asIwouldhave

‐44‐

thoughtSSPhasmadeevidentallalong,butasIcertainlyhopethesepastthreechaptershavemadeclear,that’swhatmycontextualisttreatmentisallabout.Ithinkthatisjustthekindofquestiontraditionalapproachestoskepticismhavebeeninterestedin.ButifI’mwrongaboutthat,thenIsay:somuchtheworsefortraditionalinquiryintoskepticism.

16. IsMyContextualistPositionInexpressible?

Wenowmoveintoobjectionstheevaluationofwhichinvolvesthemechanicsofjustwhenandhowtheepistemicstandardsforknowledgeclaimsaremoved.Westartwithchargesthatmycontextualistpositiononskepticismisinexpressible.

Thisisformeaveryoldobjection,predatinganyactualwritingupofmyposition,eveninthedraftiestofdrafts,asitwasraisedbymyadvisorwhenIwasjusttalkingthroughmybasicapproachtoAIthatwasatthetimejuststartingtomakesensetome.Here’showIrecentlyrecalledtheeventonFacebook:

A quick recollection of working with my dissertation supervisor, Rogers Albritton:

When I was a grad student, and Albritton had bad problems with his sciatic nerve, I had to take over teaching his upper division undergrad class on skepticism. So I had to drive over to his condo a few times, ostensibly to talk about what to cover in class and how, but (as anyone who knew him could have predicted), it turned out just to be a discussion of the issues themselves, with basically zero attention to issues about how anything should be covered in class. He was laid up in bed, and I'd sit in a chair across the room. One time, the discussion went very long (which also cd have been very easily predicted), and, since it started with the bright afternoon sun shining through the window, with no lights turned on, it ended in total darkness, because we never bothered to turn on any lights. So, I ended up with the experience of his voice coming at me from out of the darkness, like the voice of God. (And if I had to cast someone as the voice of God, he’d be one of my first choices. Him or Orson Welles, depending on which personality traits of God’s were being highlighted by the role.) And I still remember just what he said to me at one point, out of the darkness, and how he said it, with the dramatic pause before the last word, which was delivered as only he could, dripping with playful mock disdain:

“So, according to you, there are these facts to the effect that, at least by ordinary standards, we know various skeptical hypotheses are false. But we can't state these facts, because the

‐45‐

very attempt to do so drives the standards up to where we no longer know. So if we can't speak them, what are these facts *doing*, just ... obtaining???!”

My response: “They’re facts. What’dya want them to do?”

Which got the biggest laugh from him I think I ever got.

TherewasofcoursesomethingoddaboutAlbritton’spresentationoftheallegedproblemthatonmyview‘wecan’tstatethesefacts’(andthisoddityhasalsocharacterizedthepresentationofmanywhohaveconfrontedmewiththisprobleminthesubsequentyears):Hehadlittleproblem(followingthewordingofmyownearlierattemptstodoso)statingjustwhataspectofmyviewwassupposedlyunstatable.AtleastifI’mrightabouthowphraseslike‘byordinarystandards’canwork,Icanemployalittle‘semanticascent’andexpresstherelevantpartofitinjustthewaythecriticjustdid(usually,followingmyverballead):‘Byordinarystandards,I(orwe)knowthatHisfalse.’(AndifI’mwrongabouthow‘byordinarystandards’works,wecouldcertainlyfindsomeother,thoughperhapsmoreugly,waytoengageinalittleascentandsaywhatneedstobesaidhere.31)Sotheobjectionshouldberefinedtothechargethatonmyview,onecannotexpressthefactsIallegebymeansofverysimplesentencesoftheform‘Iknowthatnot‐H’.

I’vesinceheardsuchinexpressibilityobjections,whetherrefinedornot,manytimes,andreaditafewtimes.32Here,forinstance,isWayneDavis:

31See(DeRose2009:215‐17)forexamplesof‘easylittlephrases’thatcanbeusedashandydevicesofsemanticascentandsomeexamplesofhowtheycanbesoused,butthenalsoforalittleguidanceonhowtoresorttomore‘tiresome’waysofachievingthateffectifyoudon’tgoformyuseofthehandylittlephrases.

32InadditiontoDavis,whoseobjectionweareabouttoencounter,see,e.g.,(Sosa1999:144‐5)and(Koethe2005:76).ThatIholdthatonecannevertruthfullyclaimknowledgeincontextsinwhichunderminingskepticalhypotheseshavebeenraised,orcannevertruthfullyclaimthatskepticalhypothesesarefalse,isoneofthemostcommonandfrustratingmisreadingsofmyview.Infairnesstothosewhosomisreadme,IshouldpointoutthatinSSP,Idonotactuallypresentthecasesoftruthfulattributionsof‘knowledge’toinsensitivebelieversthatIhaddevised,butonlymentioninafootnotethattherearesuchcases,referringreaderstomydissertation(DeRose1990)fortheactualcases.Whatthesecriticsweremissing,inadditiontothatnote,isthesecondofthree‘importantpoints’Imakeinsettingupmycontextualistsolutioninsection2ofSSP,whereIexplainthatIwillbeprovisionallyassuming,foreaseofexposition,a‘skeptic‐friendly’versionofcontextualism,onwhichtheraisingofskepticalhypothesesalwayssucceedsatinstallingtherelevantelevatedepistemicstandards,andthenexplainhowtoconvertmysolutiontolessskeptic‐friendlyversionsofcontextualism.BecauseIwasmakingthatworkingassumption,muchofSSPdoesreadasifIholdtheviewthatthesecriticsmisattributetome.Itwasin(DeRose2004)thatIstartedtoexplainhow‘skeptic‐

‐46‐

DeRose’s Rule of Sensitivity, however, produces a very curious result. It implies that while ‘‘S knows ~H’’ is true in everyday contexts, we cannot say or even think that S knows ~H in everyday contexts (DeRose, 1995, p. 39).* For saying or thinking that S knows ~H would raise epistemic standards by including some H worlds in the set of relevant worlds. Since ‘‘S believes ~H’’ would be true even in such worlds, ‘‘S knows ~H’’ would now be false. As far as I am concerned, the implication that it is impossible for us to truly assert an instance of ‘‘S knows P’’– even when it is true – is at least as abominable as the Nozickian conjunction.33

Thefirstsentenceoftheabove,readonitsown,wouldseemtobemakingthesimplechargeofinexpressibility—whichcouldthenbequashedbypointingoutourability,onmyview,totruthfullysayandthinkthatwe‘know’thatnot‐Hbymeansofsomewhatwordyclaimsthatinvolvealittlesemanticascent.Butthelastsentencequotedabovelevelstherefinedchargethatdoesnotallowforsuchanescape,andperhaps,inlightofthat,weshouldreadthefirstsentenceasalsoonlyexpressingtherefinedcharge,too.

Iamquiteuncertainaboutwhetherandwhytherefinedchargeshouldbetroubling,especiallyoncewedistinguishitfromthepotentiallymore‐troubling‐if‐correct,butclearlywrong,unrefinedcharge.Solongastherelevantpartofmyviewcanbeeasilyexpressedbymeansoffairlysimplesentencesthatinvolvealittleascent,woulditreallybetroublingifitcouldn’tbeexpressedbyevensimplersentences?Sometimes,thetruthofphilosophicallypuzzlingmattersisn’tmaximallysimple.Weshouldhavebeenthankfulthatthisoneisassimplyexpressibleasitis.

But,howevertroublingitwouldbeiftrue,eventherefinedchargeisfalse,inanycase.Thereasonitcanbeperhapssomewhatdifficulttotruthfullystate‘Iknowthatnot‐H’onmyviewisthatthereisaconversationalrule—my‘RuleofSensitivity’—bytheoperationofwhich,whensuchaclaimismade,theepistemicstandardstendtoberaisedtoalevelatwhichtheclaimisfalse.Butthat’sjustatendency,andtheoperationofthatrulecanofcourserunupagainst,andloseoutto,

(un)friendly’myowncontextualistthinkingreallyis.Andsuchmisunderstandingsofmyviewdoseemtohavebecomelessfrequentsincethen.

33(Davis2004:206).TheasteriskmarksthepointatwhichDavisattachesafootnotethatIwilldiscussinmynextfootnote,below.

‐47‐

otherconversationalforcesthatcanworktokeeptheepistemicstandardsmeetablylow.34

Andindeed,inmydissertation,IgavesomeexamplesofwhatIthought,andthink,tobetrueclaimsto‘know’thatskepticalhypothesesarefalse(despitethefactthatthe‘knower’s’beliefsthatthey’refalseareinsensitive)—casesthatappearinChapter6,section8ofthisbook.

And,aswewillseetowardtheendofthenextsectionbelow,thereisanotherwaytoseetoitthatone’ssimpleclaimsoftheform‘IknowthatP’aregovernedbymoderateepistemicstandards(onwhichonedoesknowthatnot‐H,onmyview):explicitlystatethatone’susesof‘know’aretobesounderstood—asIwillhaveoccasiontodomyselfintheAppendixtothisverychapter(seethesecondparagraphofthatAppendix).Havingdonethat,onecan,onmyviews,truthfullyasserttheverysimplesentence‘IknowthatI’mnotaBIV’.

Incaseithelps,Icanalsodojustthatrightnow,too,producinganinstanceofthetypeofclaiminquestionthatIatleastholdistrue(thoughIadmititstruth‐valueisfarfromobvious).Thefollowingclaimthatclosesthissectionshouldbeunderstoodasgovernedbyordinary,moderatestandardsforknowledge.

IknowthatI’mnotabraininavat.35

34Inthefootnotethat’sattachedtoquotationindentedaboveinthetextatthepointI’vemarkedwithanasterisk,Davisseemstocatchhismistake:‘DeRose’s(1995,p.39)mostofficialformulationoftherulesaysthatepistemicstandards‘‘tendtoberaised,ifneedbe’’soastorequiresensitivebelief.Butthenhistheorydoesnotcompletelyavoidabominableconjunctions,andneedstoexplainwhenshiftsdooccur’(Davis2004:278,n.6).But,ifDavisdoesindeedseethatIholdthatitispossibletotruthfullyassert‘Iknowthatnot‐H’,heforsomereasonretains,inhismaintext,thechargethatonmyviewthisisimpossible,soIamunsurewhetherhemeanstobewithdrawingthecharge.NordoesheconsidermyexamplesofwhenIclaimthisactuallyhappens(which,again(seethepreviousnote),infairnesstoDavis,arenotactuallypresentedinSSP,butonlytherereferredtoinanote,whichsendsthereaderto(DeRose1990)fortheactualcases),normyexplanations,givenalongwiththosecases,ofwhentheshiftsdoanddonotoccur.Asforwhyconjunctionscanbe‘abominable’(orcan‘clash’)eventhoughthetwoconjunctscaneachindividuallybegivenreadingsonwhichtheyaretrue,see(DeRose1998:70‐2)—thoughthisdoesn’tgoallthatfarbeyondthefairlyobviousobservationthatconjunctionsinvolvingcontext‐sensitivetermsoftendo‘clash’(giveanappearanceofinconsistency)whenthe‘conversationalscore’isheldconstantoverthewholeconjunction,eventhough,ifyouallowyourselftochangethescoremid‐sentence,youcouldassigntoeachconjunctareadingwhichwouldmakethewholeconjunctioncomeouttrue.ThisiswhythefeatureofmyviewIappealtoinsection10ofSSPtoexplaintheabominablenessofabominableconjunctions—namely,thattheyarefalsewhenevaluatedconsistentlyatanyepistemicstandard—doesdoagoodjobofexplainingwhatitseekstoexplain.

‐48‐

17. TheFactivityProblem

Usingthesubscript‘h’sothat‘knowsh’designatesknowingbythehighstandardsthatthepresentationofAIhassometendencytoputintoplace,and‘o’sothat‘knowso’designatesknowingbytheordinaryepistemicstandardsthatgoverneverydaytalk(andwe’renodoubtidealizinghere,butharmlesslyso,insupposingthereisasinglesuchstandardsodescribed),claimsalongthelinesofbothofthefollowingarecentraltocontextualisttreatmentsofskepticism,includingmine:

a. I knowo that I have hands

b. I don’t knowh that I have hands

Afterall,acentralaimofcontextualisttreatmentsistoprotectthelikesof(a),andappealstothelikesof(b)playakeyroleinourprotectionefforts.Butsinceknowingsomethingbyanystandard,includingo,requiresthetruthofthepropositionknown,itseemsthatsincewecontextualistsaccept(b),weshouldalsoacceptthelikesof

c. I don’t knowh that I knowo that I have hands.36

35AtleastasI’minclinedtoconstruetheposition,itmaybeopentotheclassicalinvariantist,andeventheskepticalclassicalinvariantist,toholdthatI’vejustassertedatruthinthesentencetowhichthisnoteattaches.AsIaminclinedtoconstruetheposition,itconcerns‘normal’claimsabout‘knowledge’,insomegoodsense,whereclaimswhosemeaningisexplicitlystipulatedasIdidinthetextarenotinthatway‘normal’,butarestipulated,technicaluses,verydifferentfromordinaryusesoftheterm.(Itisn’teasytoseehowmystipulationthatIamusing‘ordinary,moderatestandardsforknowledge’willbetreatedbytheviewsinquestion.Theskepticalinvariantistholdsthatordinaryclaimsaregovernedbyextremelyhighstandards,sotheycouldholdthatmystipulationinvokesthosestandards—inwhichcasemyclaimisfalse.Or,sofarasIcansee,theycouldthinkmystipulationinvokesthemoderatestandardsclaimsmustmeetbeforewewillinmaketheminordinarycircumstances,inwhichcasetheymaythinkmyclaimistrue.)Frommycontextualiststandpoint,bycontrast,theepistemicstandardsthatgovernknowledgeclaimsarealwaysmovingaround,andwhatI’mdoinginthetextisstipulatingthatmyuseof‘know’willhaveacertainvaluethatthetermcanhaveinquite,ordinary,non‐stipulateduse.

36(c)followsfrom(b),giventhefactivityofknowledgeoandanappropriateclosureprincipleforknowledgeh.[1.IknowhthatIknowothatIhavehands(Assumptionforreductio).2.IknowothatIhavehandsentailsthatIhavehands(bythefactivityofknowledgeo).So,3.IknowhthatIhavehands(by1,2,closureofknowledgeh).But,4.Idon’tknowhthatIhavehands(b).So,c.Idon’tknowhthatIknowothatIhavehands(by1‐4,reductio).]Ofcourse,theformulationofanappropriateclosureprincipleisamessydeal,sotheremaybesomequestionaboutthekeyinferenceto3,above.SoletmejustsayherethatIacceptmycommitmentto(c).(Infact,Ialsoacceptacommitmenttothis:Eventhosewhodon’tknowbyanystandardsthatknowingothatp

‐49‐

Andthisraisesacoupleofapparentproblemsforcontextualistresponderstoskepticismthatoftenflyunderthebannerofthe‘factivityproblem’—problemsofatypethat,tothebestofmyknowledge,werefirstraisedin(Williamson2001),buthavesincereceivedalotofdiscussion.37

Some(notWilliamson)whopress‘factivityproblems’present

d. I knowh that I knowo that I have hands,

orthelikesof(d)(theymaysubstituteforthe‘Ihavehands’part),which,ofcourse,directlycontradicts(c),aspartofthecontextualist’sposition.Idon’tthinkwecontextualistshaveeveractuallyclaimedtoknowhthatweknowotheverythingsweadmitthatwedon’tknowh.However,(a)isacentralclaimofours,andthatapparentlycanleadsometotakeitaspartofourpositionthatweknowhthat(a)istrue(seeBrendel2005:247).Anotherfactorthatmightbeinvolvedinleadingsomecriticstoascribe(d)tousisthatwecontextualistresponderstoskepticismwillseemtoassertthelikesof(a)incontextswhereskepticalscenarioshavebeenmadesalient,andsoseemtobecontextsthatwewilltaketobegovernedbystandardsh.Forreasonswewillsoonsee,thoseassertionsofourscanseemtobeactsbywhichwecontextualistsrepresentourselvesasknowinghthat(a)isso.Sometimesthosewhopresent‘factivityproblems’justtakeitforgrantedthatcontextualistrespondersacceptthelikesof(d),orwouldatleastclaimthatwecanknowhthatweknowothelikesofthatwehavehands.38Thus,whenthey(rightly)derivefromusacommitmenttothelikesof(c),theytakethemselvestohaveuncoveredaconflictinourposition.

However,(d)isnopartofmyposition—andIsuspectthesameistrueofothercontextualistresponderstoskepticism.Idenyit.Mypositionisthatwenotonlyfailtoknowhthatwehavehands,butalsofailtoknowhthatweknowothatwe

entailsthetruthofpcan’tknowhthattheyknowofactslikethattheyhavehands.Mydiscussionofapplyingskepticalargumentstothenon‐philosophicalinAppendixAofChapter5shouldgiveanindicationofhowIthinkthisallplaysout.)

37See,e.g.,(Williamson2001),(Williamson2005),(Brendel2005),(Wright2005),(Baumann2008),(Brendel2009),(Baumann2010),(Jäger2012),(Dinges2014).

38See,e.g.,(Baumann2008583),whichlistsatp.581otherpresentationsthatpressasimilarformoftheproblem.

‐50‐

havehands.(c)isnotsomeembarrassingcommitmentthatsnuckuponme,butissomethingI’velongaccepted—andhaveevenusedtomyownadvantage.39

Butdenying(d)isonething;havingtherighttodenyitisquiteanother.Imightbecommittedto(d),whetherIlikethiscommitmentornot.Insteadofthefactivityproblembeingonewhere(d)istakentobeapartofourpositionthatwe’dhappilyaccept,butthatturnsouttobeinconsistentwithsomethingwecontextualiststurnouttobeunhappilycommittedto(c),theproblemcaninsteadbepresentedasoneofourturningouttohavetwocommitments—to(c)andto(d)—thatareinconflictwithoneanother.Iacceptmycommitmentto(c),andsomyhopesofescapingtheproblemarepinnedonavoiding(d).Or,moregenerally,thefactivityproblemcanbeviewedasadifficultywecontextualistresponderstoskepticismhaveinstatingourposition,andparticularlyinstating(a),intheveryphilosophicalsettingsinwhichwe’relikelytodoso.

Ineithercase,thepotentialproblemcanbegeneratedby,oratleastsharpenedby,theKnowledgeAccountofAssertion(KAA),whichWilliamsonandIbothaccept.KAAcanbepackagedasaprincipleconcerningproperassertion,or(asWilliamsonlikestodealwithit)asarulegoverningthepracticeofassertion,or(asIoftenliketouseit)asaprincipleconcerningwhataspeakerrepresentsasbeingthecaseinassertingsomething:

KAA-PA: A speaker is well-enough positioned with respect to p to be able to properly assert that p if and only if she knows that p according to the standards for knowledge that are in place as she makes her assertion.

KAA-Rule: One must assert that p only if one knows that p according to the standards for knowledge that are in place as one makes one’s assertion.

KAA-Rep: In asserting that p, a speaker represents herself as knowing that p according to the standards for knowledge that are in place as she makes her assertion.

IhavepresentedtheseversionsofKAAinformswhicharerelativizedtothestandardsforknowledgeinplaceasaspeakermakesherassertion,acomplexity

39See(DeRose2009:207)forhowourfailuretoknowhthatweknowocanbeusedbythecontextualiststoescapeanadvanced‘nowyouknowit,nowyoudon’t’problem(the‘fortifiedfirst‐persondialogue’).AsIreportat(DeRose2009:209,n.18),IutilizedthisescapeintalksIgaveintheyears1996‐2001,butsincetherewererelated‘nowyouknowit,nowyoudon’t’problemsthatescapedthisescape,andhadtobehandledadifferentway,thisdidn’tcomeoutinprintuntil(DeRose2009).

‐51‐

thatwillbeunnecessarytoinvariantistslikeWilliamson,accordingtowhomasingleepistemicstandardgovernsallcontexts.RelativizingKAAinthiswayrendersitneutralwithrespecttothedebatebetweencontextualismandinvariantism:Theinvariantistwilljustsupposethatthestandardsinplacearealwaysthesameones.

Thepotentialproblemnowiseasytosee.GivenKAA,asserting(a)inacontextgovernedbyhwill(totaketheRepresentationformofKAA)involvemeinrepresentingitasbeingthecasethatIknowhthatIknowothatIhavehands—i.e.,willinvolvemeinrepresenting(d)asbeingthecase.Butthen,given(c),I’llberepresentingsomethingfalseasbeingthecase—and(totaketheotherformsofKAA)I’llbebreakingtheruleforassertionandmakinganimproperassertion.Alloftheserenderitconversationallynaughtytoassert(a)inacontextgovernedbyh,andcanbeseenasawaythatIgeneratesomekindofcommitmentto(d)byasserting(a)intheveryphilosophicalcontextsinwhichIamlikelytopresentmycontextualistsolutiontoskepticism,evenifIdon’texplicitlyassert(d).

Truthbetold,I’veneverseenmuchofaproblemhere.Thelessonforcontextualistresponderstoskepticismwouldjustseemtobe:Ifyouwanttoflat‐outassert(a)(withoutbeingnaughtyandwithoutgeneratingaconflictwithotheressentialelementsofthecontextualistsolutiontoskepticism),youshouldavoidallowinghtocometogovernthecontextsinwhichyouarespeaking.Butthat’snotreallyaspecialproblemforasserting(a):GivenKAA,ifonewantstonon‐naughtilyflat‐outassertmuchofanything(including,notoriously,anycontroversialskepticalclaims),oneshouldavoidallowinghtocomeintoforce,sinceweknowlittle‐to‐nothingbystandardh.

AndsuchavoidanceiswhatI,forone,atleasttrytoachieve.Idonotacquiesceintheskeptic’s‘attempts’toraisethestandardstoh.40WhentheskepticpresentsAI,Idon’t‘accommodate’herandstartspeakingasifIknowlittle‐to‐nothing.Iinsteadengageinverbalbehaviorwhichconstitutesresistingher‘efforts’,

40‘Attempts’isinscarequotesbecausetheskepticmaynotconstrueherselfastryingtodoanysuchthing:Shemaythinkhalwaysgovernseveryconversation,andthatsheisjusttryingtogetmetorealizethat.Buthoweversheviewswhatsheisdoing,bycontextualistanalysis,sheisexecutingmaneuverswhichhaveatendencytoraisetheepistemicstandards,andsoshecanbeviewed,inanextendedsense,as‘trying’or‘attempting’toraisethestandards.Seetherelateddiscussionaboveinsection5.

‐52‐

andatleasttryingto‘notlethergetawaywith’raisingtheepistemicstandardstoh.41

Now,therecouldbequestionsaboutwhetherIcansucceedinmyattemptstokeephfrombeinginstalled.Thatdependsonjustwhattherulesareforwhichepistemicstandardsgovernacontextwhenthespeakersinvolvedare‘trying’topushthosestandardsindifferentdirections.IcanvassedseveralcontextualistoptionsonthismatterinSSS(seenote9,above).Somepresentationsofthefactivityproblemclearlyassumevery‘skeptic‐friendly’versionsofcontextualism(bywhichImeanversionsonwhichtheraisingofepistemicstandardshappenseasily,oftenevenwhenthatmeansoverridingconversationalmaneuversthatwouldtendtokeepthestandardsfromsorising),thoughit’ssometimesdifficulttotell.

Butlet’sfirstseehowthisplaysoutonWilliamson’spresentationoftheproblem,which,interestingly,avoidsrelianceonKAAbyinstead(thoughthisiscloselyrelated)chargingthatthecontextualistrespondertoskepticismgetscommittedtoaMoore‐paradoxicalstatement:

Once sceptics have manipulated the context, in the epistemology seminar, contextualists are apt to console themselves with the thought that although most denials of “knowledge” in that context of scepticism are correct, in everyday contexts many assertions of “knowledge” are also correct. For example, although ‘I do not know that there is a whiteboard in the room’ expresses a truth as uttered in the seminar, 'He knows that there is a whiteboard in the room' expresses a truth as uttered simultaneously outside the seminar about the same person. But that thought underestimates the gravity of the situation in which the sceptic has put contextualists, on their own analysis. For since 'know' is a factive verb, the truth of 'He knows that there is a whiteboard in the room', as uttered outside the seminar, requires the truth of ‘There is a whiteboard in the room' as uttered outside the seminar and therefore as uttered inside the seminar too (since no context-dependence in the latter sentence is relevant here). Thus the consoling thought commits them to claims such as 'There is a whiteboard in the room', while their contextualism commits them to claims such as 'I do not know that there is a whiteboard in the room'. Consequently, in the seminar, they are committed to this:

(MK) There is a whiteboard in the room and I do not know that there is a whiteboard in the room.

41Imyselftendnottodosobyinsistingthat‘IknowthatIhavehands’,orthelike,butwhatIdodoesclearlyconstituteresistancetotheskeptic’sattemptedraisingofthestandards.

‐53‐

But (MK) is Moore-paradoxical: although (MK) could be true, it is somehow self-defeating to assert (MK). Thus contextualists are not entitled to the consoling thought. In the seminar, they should not say that 'He knows that there is a whiteboard in the room' expresses a truth outside the seminar . . . . At least in the seminar, contextualists are at the sceptic's mercy and cannot look outside for help. (Williamson 2005: 689)

Williamson’sassessmentsthat‘contextualistsareatthesceptic’smercy’andthatourcontextualism‘commits[us]toclaimssuchas’thesecond,negativehalfofhis(MK)maywellbebasedontheassumptionofaskeptic‐friendlyversionofcontextualism,thatwehaveseenisoptional,andisnotinfactmyview.Butit’snoteasytointerprettheclaimthatIam‘committed’tothelikesof‘Idon’tknowthatIhavehands’.Perhapstheideaisthattherestoftheabovepassageissupposedtobeunderstoodundertheassumption,setatthestartofthequotedmaterial,thattheskeptichassucceededininstallinghasthestandardthatwillgovernmyclaim,inwhichcase,ifIamgoingtosayeither‘IknowthatIhavehands’or‘Idon’tknowthatIhavehands’,thenithastobethelatterifIamtoavoidsayingsomethingfalse.ButifI,astheotherpartytotheconversation,haveanysayinwhethertheskepticsucceedsininstallingh,asIdoonmanynot‐so‐skeptic‐friendlyversionsofcontextualism,wecan’tjustassumethattheskepticsucceeds.Soperhapsthelessonweshouldtakefromallthisisthat,insofaraswesensethatit’spermissibletosticktoone’sgunsagainstaskeptic,wejustshouldn’toptforsuchaskeptic‐friendlyversionofcontextualism—andthenalsothatweshouldresistskeptics.Thiswouldconnectwiththemainlessontobedrawnofthe‘factivityproblem’ifwedoendorseKAA—namely,thewisepieceofstrategicadvicethat,assumingKAA,itisfoolishtocomplacentlylettheskepticdrivethestandardsuptohonthegroundsthat,havinggivenuphopeofbeingabletotruthfullysaythesimple‘IknowthatIhavehands’,youcanstillalwaysfallbackonbeingabletoflat‐outassertwithperfectproprietythelikesof(a).Thisisfoolish,givenKAA,because(a)’scommitmenttothetruthof‘Ihavehands’meansthat(a)tooisunassertableincontextsgovernedbyh.

ButbacktoWilliamson’s‘commitment’‐to‐MKversionoftheproblem:ThereareotherwaystounderstandthechargethatIam‘committed’tothelikesof‘Idon’tknowthatIhavehands’whentalkingtotheskeptic.Oneisjustasmalltweakawayfromtheinterpretationwetriedabove.WecantakeittobetheclaimthatifIamgoingtosayeither‘IknowthatIhavehands’or‘Idon’tknowthatIhavehands’,itwillhavetobethelatterifIamgoingtosaysomethingtrue(asopposedtojustavoidingsayingsomethingfalse).Inthisway,Iamcommittedto‘Idon’tknowthatIhavehands’whentalkingwiththeskepticonawiderrangeofversionsof

‐54‐

contextualism—including,interestingly,the‘gapview’whichImyselfleantoward(DeRose2009:144‐8),foronthatviewIwillsucceedinsayingsomethingtrueifIagreewiththeskepticthat‘Idon’tknowthatIhavehands’,butwillsaysomethingthatisneithertruenorfalseifIinsistthat‘IknowthatIhavehands’(inthefaceofastubbornskeptic).Wemaybetemptedtotakethisasareasontooptforaversionofcontextualismstilllessfriendlytotheskeptic,likethe‘vetopower’view(DeRose2009:140‐41),orperhapsa‘multiplescoreboards’view(DeRose2009:134‐5),onwhichIwillassertatruthifIinsistthat‘IknowthatIhavehands’,andsoonwhichIamnotinthewayunderconsideration‘committed’to‘Idon’tknowthatIhavehands’(andonwhich,inagoodsense,Iamnotso‘attheskeptic’smercy’).

ButIforonedon’tseethisasaverystrongmotivationformovingtosuchaversionofcontextualism,mostlyforreasonswe’llseeabitlater,butalsobecauseourcurrentunderstandingofthetermisaratherstrangewaytobe‘committed’tothesentence,‘Idon’tknowthatIhavehands’,for,amongotherproblems,itassumesImustasserteither‘IknowthatIhavehands’or‘Idon’tknowthatIhavehands’,andwhereisthatforcedchoicecomingfrom?Itdoesn’tseemthatIwouldhavetobeproblematicallyunforthcominginoptingtoassertneitherofthosetwothings.Whenpointedlyaskedwhether,accordingtome,IknowthatIamnotaBIV,Icouldforinstanceinsteadassert(a),atwhichpointI’dfeelI’veprettywelldescribedmypositionontheissueatquestion.Ofcourse,givenKAA,aswe’veseen,it’sproblematictoassert(a),too,inacontextgovernedbyh.ButthismayperhapsshowthatonereallydoesneedtorelyonKAAtodrivea‘factivityproblem’,andweshouldgiveuptryingtofoistsomeproblematic‘commitment’tothesentence‘Idon’tknowthatIhavehands’ontothecontextualistrespondertoskepticism.

ButperhapsthewayinwhichI’mcommittedto‘Idon’tknowthatIhavehands’issupposedtonotbeanythinglikethatI’minsomewayforcedtosayit,butjustthatI’mcommittedtowhatisassertedbeingtrueifI,orifsomecontextualistresponder,(orperhapseveniftheskeptic),weretoassertthatsentenceinacontextinwhichskepticismisbeingpressed(sinceI’dthenbegoingalongwithskeptic’sattempttoraisethestandards).Andthecontextualistwouldhavethatkindofcommitmenttothesentenceonawidevarietyofformsofcontextualism.(Notall:somewaysofworkingouta‘reasonablenessview’(DeRose2009:141‐2),forinstance,wouldallowustoavoideventhiskindof‘commitment’,aswouldcertainformsofcontextualismwhichplacea‘ceiling’onhowhighthestandardscango.)Butit’saveryoddcomplaintthatthecontextualistis‘committed’inthatsenseto

‐55‐

thesentence‘Idon’tknowthatIhavehands’,andthentotheMoore‐paradoxicalsentence,(MK),ofwhichitisapart.For,asWilliamsonpointsout,andisclearinanycase,Moore‐paradoxicalsentenceslike(MK)couldbetrue;that’sreallytheinterestingthingaboutthem.IfyouwanttoavoidmakingMoore‐paradoxicalpronouncements,then,contextualistrespondertoskepticismornot,you’regoingtohavetorefusetosaysomethingsthatwouldbetrueforyoutosay.Soit’snotmuchofacomplaintagainstsomeonethatthey’re‘committed’toaMoore‐paradoxinasensewhichrequiresforsucha‘commitment’toasentenceonlythatone’sviewsimplythatonewouldspeakatruthinassertingthatsentence.

AndiftheideaisjustthatI’mcommittedtothetruthof‘Idon’tknowthatIhands’,andthento(MK),whenthey’reevaluatedath(nevermindwhoisorcouldbesayinganyofthis),thenIhavenoproblemwithhavingsuchacommitmenttoaMoore‐paradoxicalstatement.SomeMoore‐paradoxicalstatementsaretrue,andI’mperfectlyhappytobecommittedto(MK),asevaluatedath,beingoneofthem.Infact,Iwouldn’thaveitanyotherway.

So,let’sstopworryingabout‘commitments’,andgobacktoKAA‐generatedfactivityproblems.Andwhatweseemtobeleftwithisapotentialproblem,generatedbyKAA(whichIaccept),forcontextualistresponderstoskepticismassertingourviewsincontextsgovernedbyh.Towhich,again,thesolutionwouldseemtobe:Don’tletyourassertionsbegovernedbyh—whichisverygoodadviceinanycase!Theremainingpotentialconcernwouldthenseemtobetheworrythattheveryphilosophicalcontextsinwhichcontextualistresponderspresenttheirsolutionscanappear,atleastonacertainviewofhowcontextualismoperates,tobeoneswhichareormaybeevenmustbegovernedbyh.

So,wearefacedwithacoupleof(closelyrelated)questions:Onthebestversionsofcontextualism,arecontextualists’assertionsofourviewtypicallygovernedbyh?And:Onthebestversionsofcontextualism,cantheskepticforceusintoacontextwhereourassertionsaregovernedbyh(arewe‘attheskeptic’smercy’)?Andlet’sjustsupposethatitwouldbeagoodobjectiontocontextualistresponsestoskepticismiftheanswerstothesequestionswereyes.

Fortheanswerstothesequestionsare:noandno.Andquiteclearlyso,thisoldcontextualistwouldsay.True,wecontextualiststypicallyfindourselvesassertingelementsofourresponsestoskepticismincontextsinwhichskepticalhypotheseshavebeenmentionedandarehangingintheconversationalair.But

‐56‐

thoughthosewhopressfactivityproblemsoftenassumethatthissufficesonourcontextualistviewstounfailinglyforcethestandardsuptoh,asI’vebeenstressinghere,that’sjustnotso.

ButwhydoIthinkouranswersaresoclearly‘no’?Well,sofar,I’vebeencitingthevariousoptionsforcontextualistsofthetypeIcanvassinSSS.And,yes,it’sveryfairtosaythatI(andothercontextualists,too)haven’tbeenverydefinitiveinwhichofthoseoptionswethinkisbest(thoughIandwedohaveourleanings).Andsomeoftheseoptions—includingtheoneIleantoward,aswesawabove—dohavepotentialfortheskeptictobeabletoconversationallypushusaroundinsomeways.

ButthechoiceamongtheoptionsIpresentinSSSthereoccursagainstthebackdropofacertainkindofdisputebetweenskepticsandthosewhoresistthem,disputesfallingintheparameterssetat(DeRose2009:131‐4).Andinpresentingourviews,contextualistsdon’thavetostaywithinthoseparameters.AndifyouareworriedthatIamofteninfactsubjecttofactivityproblemsinassertingmyviews,neverfear:Iinfactdon’tstaysoboxedin.Inparticular,inSSSIwasaddressingdisputeswherenopartytothediscussionoffersa‘contextualist,orproto‐contextualist,analysisofthedebatethatistranspiring’!AndrestassuredthatwhenIampresentingmycontextualisttreatmentofskepticism,Iofcoursedonot—atall—staywithinthatparameter.Andifitcomestothis,onceapartytoaconversationstartsexplicitlystatingwhatvaluetheiruseofacontext‐sensitivetermistohave,weareofcourseinawholedifferentsituation.Iftheotherpartyrefusestogoalongwithusingthevaluespecified,thenweareinaconversationalsituationwherea‘multiplescoreboards’treatmentseemsverymuchinorder,inwhichcasethecontextualistrespondertoskepticismisclearofthe‘factivityproblem’,forhisassertionsaren’tgovernedbyh(thoughtheskeptic’sassertions,governedastheyarebyh,arestillingravedanger).

So,yes,therecanbetoughandverytrickycallsforthecontextualist(orthenon‐contextualist,tryingtojudgewhicharethemostviableformsofcontextualism)tomakeconcerningthesemanticsofdisputestakingplace‘inthedark’aboutthepossibilityofcontextualism,wherethepartiestothedispute,thoughsointhedark,arestill‘trying’(inthesensewehavebeenusing)topushtheepistemicstandardsindifferentdirections.Maybe(thoughthisgoesagainstmyowninclinations)afairlyskeptic‐friendlytreatmentofsuchsituationsisinorder.Ifso,theninagoodsensetheunenlightenedMooreanisattheskeptic’smercy—and,givenKAA,athermercywithrespecttoherflat‐outassertionsgenerally,andnotjustwithrespecttoher

‐57‐

explicitclaimstoknowledge.Butthatwouldn’tmeanthat(oncontextualism,oronthemostviableversionsofit)thecontextualistwhorealizeswhatisgoingonisattheskeptic’smercy.Ifitreallycomestothis(andIdon’tthinkthisisreallynecessary),andyoudon’twantyourclaimstobegovernedbyh,justsayso.42Thatreallyshouldtakecareofmatters.

Appendix:DoIEvenKnowoAnyofThistoBeTrue?:SomeThoughtsaboutKnowledge,Belief,andAssertioninPhilosophicalSettingsandOtherKnowledgeDeserts

IntheprevioussectionIaddressedtheworrythatIfacesomespecial‘factivityproblem’inassertingthephilosophicalviewsI’mdefendinginthisbook.Theallegedspecialproblemwaslargelydrivenbymyadmission/commitmenttomylackofknowledgehofakeycomponentofmyview,andIspentsomespacediscussingtheconditionsunderwhichstandardshwouldgovernadiscussionofskepticism.IfearthatmydiscussionmayhavegiventheimpressionthatIthinkthatsolongasIkeepthestandardsforknowledgefromspinningoutofcontrol,I’mcompletelyinthecleartoassertmyphilosophicalviews,despitemyacceptanceofKAA,becauseIdotakemyselftoknowtheseviewstobetruebyordinaryormoderatestandardsforknowledge.

42AsalookatSSPwillreveal,Idon’tthereincludeanyexplicitinstructionsthatImeannottobegovernedbystandardslikeh.That’sbecauseIdon’tthinksuchpedantryisneededthere.PartofthereasonIsupposeisthatIdon’tconstruemyreaderasaskeptic,butjustassomeonewhoisinterestedintheissueofskepticism.Thus,thoughIamuptomyeyeballsinmentionsofradicalskepticalhypotheses,Idon’ttakemyownclaimstobetheregovernedbyh.ThatIamengagingincontextualistanalysis,andmakingitclearthatIdon’ttakemyselftoknowhmuchifanything,whileIdotakemyselftoknowmuchbyotherstandards,sufficetomakeitclearthatmyclaimsarenottobeunderstoodasgovernedbyh,Ihope.Whenthesituationdoescallforit(aswhenIamfacingarealskepticwhoispushingthestandardstowardh,andespeciallyifshe(orifsomenon‐skeptic)startsaccusingmeofhaving‘factivityproblems’,orifI’maskedwhetherItakemyclaimstobegovernedbyh),IwillsaythatIdon’ttakemytalktobegovernedbyh.Andifithelps,Iexplicitlydeclarethatrightnow.Thoseseemverybadstandardstouse!

‐58‐

Andhonestycompelsmetosaythat’sjustnotso.Idon’ttakemyselftoevenknowbyordinarystandardsthatmycontextualistsolutiontoskepticismisright.AndwhilethisgeneratesissuesconcerninghowIaminapositiontostatemyviews,theseturnouttobeverygeneralissuesinstatingphilosophicalviews—generalissuesthatIwillquicklyaddresshere.

Thisisimportanttoacknowledgenotonlytoavoidgivingafalseimpression,butalsotoansweroneformofwhatmaybeclassifiedasa‘factivityproblem.’Inpresentingwhatisoftenclassifiedasaversionofthatproblem,CrispinWrightdistinguishesthreecontexts:

Consider three epistemological contexts: that of quotidian common sense, Q; that of scepticism, S; and that of epistemological contextualism itself, C. (Wright 2005: 243)

Andinpressinghisparticularproblemwiththecontextualistsolution,WrightdoesnotassumeanypositiononwhetherCisdistinctfromS.Butwhatprobleminthevicinitycantherebeifwe’renotassumingthatC,thecontextinwhichthecontextualistpresentsherview,isS,thecontextinwhich,accordingtothecontextualist,wedon’tcountasknowingthethingstheskepticdenieswe‘know’?ForWright,the(oratleasta)problemisgeneratedbythis:HoweverCaligns(orfailstoalign)withtheothercontextsinquestion,ifthecontextualistadmitsthatshedoesnotknowthatherpositioniscorrectbythestandardsthatgovernC(thecontextinwhichshepresentsherview),sheistherebycommittedtohavingnopointtomakeonthesemattersinC:

Now the perspective of epistemological contextualism had presumably better be one from which the theorist can know both of the potentialities of quotidian common sense and of the limitations imposed by scepticism – can know both that common sense’s claims to knowledge are, in their proper context, perfectly good, and that scepticism’s denials of knowledge are perfectly good in theirs. For if contextualism cannot rationally profess that knowledge, it has no point to make! (Wright 2005: 243)

NowI’veadmittedthatIdon’tevenknowbyordinarystandards(Idon’tknowo)thatmycontextualistsolutionisright,andI’lladmittoWrightthatIdon’t‘knowC’thatmysolutionisright—Idon’tknowthatit’srightbythestandardsforknowledgethatgovernthecontextsinwhichIpresentit.43

43TheproblemsWrightismostkeentopressinvolvewhetherthecontextualistrespondertoskepticismcanmaintainsomeformof‘even‐handedness’that,speakingasonecontextualistrespondertoskepticism,Iwant

‐59‐

However,IinsistthatIdohaveapointtomake—indeed,severalofthem—eventhoughIdon’tknowthatI’mrightaboutthepointsI’mmaking,evenbystandardsoandc.

Perhapsthefirstthingtosay(andherewearegettingtowhyIthinkthisisageneralprobleminphilosophy)isthatIdonotatallfeelaloneinmypredicament:Ithinkphilosophersgenerallydon’tknow—byanygoodstandards—thepositionswetakeoncontroversialissues.44(And,yes,thattooissomethingIdon’ttakemyselftoknow,byanygoodstandards.)Ifthatrendersuspointless,weareintrouble!Infact,Ithinkthereissomethingtotheviewofphilosophersasspecialistsinaddressingquestionstowhichnobodyhasyetfiguredoutaknowledge‐producingwaytogetanswersto.Oneofourmostimportantspecialskillsisgeneratinganswerstosuchquestions,andgood(evenifnotknowledge‐producing)supportforthoseanswers.Ifanythinglikethatiswhatwearegoodfor,then‘Sticktothepointsyouknowtoberight’wouldbeabouttheworstadviceanyphilosophercouldfollowinherwork!

ThoughIholdthatwephilosopherstypicallydon’tknowthatourcontroversialpositionsinphilosophyarecorrect,evenbytheordinarystandardsforknowledgeatwhichwecountasknowinglotsofotherthings(andfortherestofthisAppendix,myusesof‘know’anditscognatesshouldallbeunderstoodasdesignatingknowledgeo.),itcanoftenfeeltousasifwedoknowthatweareright.Andthiscangiverisetodelusionsofknowledge.

nopartof.Ihopetheopeningofthischapter(thefirstsevensectionsofit)betterconveymyattitudeandapproachtotheproblemofskepticism.ButWright’sclaimaboutourneedto‘rationallyprofessknowledge’inordertohaveanypointtomakedoesalsoneedtobeaddressed.

IshouldnotethatthepassagefromWrightthatIdisplayaboveinthetextisdirectedspecificallyatwhathecalls‘subjectcontextualism’,whichisnot,intheterminologyIuse,aformof‘contextualism’atall.However,abitlater,Wrightbroadenshistargettoincludecontextualists(bymyuseofterms):‘Ihavepresentedtheargumentinawaythatrelateditdirectlytosubjectcontextualism.Butascriber‐contextualismfaresnobetter’(Wright2005:244).AndthoughWrightdoesn’tdevelophispointsasnicelyingoingafter‘ascribercontextualists’(whichiswhyIuseapassagefromtheearlierpartofhiscase),itseemsclearthathewouldtakethepointshemakesinthepassageIdisplayinthetexttoapplytosuchcontextualistsasmeaswell.

44AsGaryGuttingpointsout,whilephilosophicaltheorizingdoesnottypicallyresultinknowledgeofthecontroversialpositionswetakeontheissueswe’refocusedon,itdoesproduce,asaregularbyproduct,philosophicalknowledgeofsuchthingsasthattherearecertainimportantdistinctionstobedrawn(Gutting2009).

‐60‐

Truthbetold,Ithinkthatwedon’tevenreallybelievethatourcontroversialviewsinphilosophyarecorrect—thoughIshouldquicklyclarifythat,andparticularlythe‘really’Ithrowinthere.WhileIfeelquitecomfortableinjudgingthatwedonotknowtheitemsinquestion,whenitcomestobelief,Ithinkthatwearetypicallyinan‘in‐between’state,asEricSchwitzgebelputsitinhisveryhelpfulworkhere(Schwitzgebel2001;2002),inwhichitwouldn’tbetruetodescribeuseitheras‘believing’oras‘notbelieving’thesethings.Myuseof(thequiteelasticword)‘really’hereisintendedsothat‘reallybelieving’wouldbetobeinastateinwhichonecouldbetruthfullysaidtobelievetheiteminquestion.Inthissense,wedoreallybelieveallsortsofthings,including,Ithink(sinceI’mgenerallyquitegenerousinascribingbeliefsinlotsofother,non‐philosophicalcases),thingswearequiteunsureof.Example:IholdthatintheintendedsenseIdoreallybelievethatAbrahamLincolnwasbornin1809,thoughI’mveryunsureofthat.IseemtorememberthatbeingtheyeargivenbyGoldmaninanexampleinapaper(Goldman1967)thatIhaveinthepastreadandtaughtfrom.However,I’mnowquiteuncertainthat1809wastheyearusedintheexample,andsowhetheritisreallytheyearLincolnwasbornin.IfsomeoneweretoaskmewhatyearLincolnwasbornin,Iwouldn’tfeelinapositiontoflat‐outassertthatitwas1809,butwouldonlygiveahedgedanswer,e.g.,‘Ithinkitwas1809.’SoIcountusasintheintendedsensereallybelievingeventhingsthatweareveryunsureof.Yet,Istillthink(andhere,‘think’conveysphilosophicalacceptance:thistooissomethingIdon’treallybelieve),wedon’treallybelieveourcontroversialviewsinphilosophy—thoughneitherdowereallynotbelievethem.

Suchthoughtsseemfairlycommonamongphilosophersnowadays,butIwillillustratethekindofconsiderationsthatleadmetowardthembyusingtheexamplethatWilliamAlstonusedtomakesuchpoints(Alston1996:10‐11).So,I’manincompatibilistaboutfreewillanddeterminism.ThisisaviewIacceptandwilldefend,sometimespassionately(despitemylackofexpertise!),invarioussettingsofphilosophicaldiscussion.Thisisagoodexampleformetouseherebecause,notonlyisthisoutsideofmyareasofexpertise,butitisaviewonwhichthemajorityofphilosophers,andalsothemajorityofphilosopherswhohavestudiedtheissuemuchmorecloselythanme,seemtobelinedupagainstme.45Still,inthe

45BryanFranceshasdoneimportantworkingettingphilosopherstothinkaboutthechallengeoffacingdisagreementbyour‘epistemicsuperiors’,aswellasbyour‘epistemicpeers’.See,e.g.,(Frances2012).

‐61‐

philosophicalsettingsinwhichIsometimesfindmyselfcontendingforincompatibilism,InonethelessfeelstrangelyconfidentthatI’mright.ItindeedfeelstomeverymuchlikesomethingIknowtobethecase—andcertainlylikesomethingIbelievetobethecase.

Butthingswouldbeverydifferentifsomethingwerereallyridingonthematter—ifpracticalconsequencesweresomehowreallytiedtowhetherIwererightaboutthismatter.SupposeI’mupontheshipofsuper‐advancedaliens,whomIsomehow(andnevermindhow)knowtobetruthfulwhentheytellmethattheissueofwhetherfreeactioniscompatiblewithdeterminismisoneofthosephilosophicalquestionswehumanspuzzleoverthatactuallydoeshaveacorrectanswer;thatthey,thealiens,actuallyknowwhetherit’scompatibilismorincompatibilismthatiscorrect;andthattheywillgivemeachancetosavetheEarthandhumankindbygettingthequestionright:Igettogiveoneanswertothequestionwhethercompatibilismorincompatibilismistrue,andifIrefusetoanswerorgetitwrong,theywilldestroytheEarthandeveryonelivingthere,butifIgetitright,theywilldestroynothing,butwillreturnmetoEarthandthenpeacefullyleave.Or,tovarythecase,supposethatitisnotthefateoftheEarthandhumankindthatisatstake,butthatthealienswillinsteadgiveme1millionU.S.dollarsifIgivethecorrectanswer,butnothingifI’mwrong,beforereleasingmeandpeacefully

Inthe2009PhilPapersSurveys[http://philpapers.org/surveys/],theresultsfor‘Freewill:compatibilism,libertarianism,ornofreewill?’were:

Among‘Allrespondents’inthe‘Targetfaculty’:Acceptorleantoward:compatibilism 550/931(59.1%)Other 139/931(14.9%)Acceptorleantoward:libertarianism 128/931(13.7%)Acceptorleantoward:nofreewill 114/931(12.2%)

Amongspecialistsin‘Metaphysics’inthe‘Targetfaculty’:Acceptorleantoward:compatibilism 130/234(55.6%)Acceptorleantoward:libertarianism 50/234(21.4%)Acceptorleantoward:nofreewill 31/234(13.2%)Other 23/234(9.8%)

Amongspecialistsin‘PhilosophyofAction’inthe‘Targetfaculty’:Acceptorleantoward:compatibilism 23/43(53.5%)Acceptorleantoward:libertarianism 8/43(18.6%)Other 7/43(16.3%)Acceptorleantoward:nofreewill 5/43(11.6%)

‐62‐

leaving.Ineitherofthesecases,IwouldfeelverydifferentlyabouttheissuethanIdowhendiscussingthematterinaphilosophicalsetting.Andwhat’sreallyinterestingisthat,beyondtheeffectsonewouldlikelyexpecthighstakestohaveonthematter,Iwould(atleastuntilrecently)havebeenstronglyinclinedtogowiththeopinionofthemajorityofphilosophers,ratherthanmyownphilosophicalacceptanceofthematter,inthesecases.(ThisexamplehasbeensomewhatruinedformesinceIhavedippedabitmoredeeplyintothephilosophicalliteratureonfreeaction,andhavebecomemoreandmoreconvincedofhowlittlethecompatibilistpositionhasgoingforit.Thiscomplicatesmattersgreatly,asitisnowquiteimpossibleformetoconfidentlyguesswhatIwoulddointhesave‐the‐Earthcase,andmakesmenowthinkI’dgowithmyownincompatibilismwhenit’samilliondollarsthat’satstake.Butratherthandealwiththesecomplications,let’sjustfocusonmyearlierself,whowasnotsowellapprisedofthestateofthediscussion.)Iwouldhavebeenmoreinclinedtogowiththemajorityviewinthesave‐the‐Earthcasethaninthemoneycase,largely,Ithink,becauseofmydesire,outofpersonalpride,toenjoyhavingbeenrightallalong—whichdesiregetswipedoutasunimportantbythegreaterstakesinvolvedinthesave‐the‐Earthcase.(Yes,amilliondollarsisalotofmoney,but,perhapssadly,personalprideisaverystrongmotivatingfactorforme.)

Irealizethatit’squitediceytopredictwhatonewoulddoinsuchwildcircumstances,andIhavenowencountered(whenpresentingthesethoughtsatvariousplaces)lotsofinterestingguessesothershavemadeaboutwhattheywoulddointherelevantsituations,andinterestingopinionsaboutwhatitwouldberationaltodo.ButsolongasIwouldfeelatleastasignificanttemptationto‘flip’(togowithmajorityexpertopinion,ratherthanwithhowthingsseemtomepersonally),thisseemstobeinmarkedcontrasttohowrealbeliefs,eventhoseheldverytentatively,behaveundersuchstress.ConsideragainmyverytentativebeliefthatLincolnwasbornin1809.SinceI’msouncertainaboutthematter,IwillbequiteconservativeinwhatI’llstakeonthatbelief.Butinsituationsinwhichit’sclearthatIshouldgiveananswertothequestion(likewhensomethingbadwillhappenifIrefusetoanswerorgiveawronganswer,andwillbeavoidedonlyifIanswercorrectly;or,positively,ifsomethinggoodwillhappenonlyifIanswerandanswercorrectly),1809istheanswerI’mgiving,withnotemptationtooptforadifferentanswer.Ifyougreatlyraisethe(positiveornegative)stakesonme,youcanmakemefeelinvariouswaysveryunconfidentaboutwhatwillhappen.Whenthere’sahugenegativeresultonthetable,youcancausemegreatanxiety.Butyou

‐63‐

won’ttemptmetogowith1808or1810insteadof1809.Youmaymakemeaddthelikesof‘Heavenhelpme!’tomyanswer.Andyoumaymakemetryveryhardtosearchmymemorymorecarefully.ButinsofarasallIcancomeupwithisthistentativepushtowardtheanswer‘1809’,that’swhatI’mgoingwith,andthatisinitswayaneasy(evenifanxiety‐producing)callformetomake.I’mworried,butnotatalltemptedtoflip.Butinthecaseofmyincompatibilism(which,atleastinsomeheatedsettings,doesn’tfeelverytentativeatall),Iwillbeverytemptedtoflip,andgiveananswerdifferentfromwhatIgiveinphilosophicaldiscussion.

DosuchfactsabouthowIwouldhaveactedwhensomethingwasactuallyridingonwhetherIwasrightaboutthemattermeanthat(untilrecently)Ireallybelievedthatcompatibilismistrue?Ithinknot,butIdothinkthatthisandrelatedconsiderationsdopointtotheconclusionthatIdidn’treallybelievethatincompatibilismistrue—despitehowstronglyImighthavefeltthatIbelieveditwhenarguingaboutitinphilosophicalsettings.46

Oneofthesharpestcontrastsinmythinkingabouttheissuesintheimaginedhigh‐stakessituationsasopposedtothephilosophicalsettingsinwhichIactually

46CompareWilliamLycan,whomI’verecentlydiscoveredhasverysimilarthoughtsaboutbelief(thoughheseemssignificantlylesssubjecttoconsiderationsofpersonalpridethanIam!):

First,thereisaquestionofwhatphilosophersbelieve.Dowebelieveourownviewsinthesamesenseinwhichwebelieveordinarythings?HumeandMoorepointedoutthatthereisoftenaverystrikinggapbetweenourdoctrinesandourbehavior:ineverydaylife,epistemologicalskepticsmakeknowledgeclaims,idealistskickstones,antirealistsabouttimeconsulttheirwatches,andsoon.(Asimilarpointhasoftenbeenmadeaboutreligiousbelievers,manyofwhomdonotbehaveatallasifthecreedstheyprofessaretrue.)Itisimportanttoseethatthesearetypicallynotcasesofhypocrisyorotherinsincerity;whenweavowaphilosophicalthesisanddefendit,sometimespassionately,wemeanwhatwesay.Nonethelessthethesisdoesnotshowinournonverbalbehavior,oreveninouroff‐dutyverbalbehavior.

Also,supposeIhavebeenthuspassionatelydefendingoneofmyowncorephilosophicaldoctrines,sayfunctionalismregardingthemental.IhavesaidinprintthatfunctionalismisaviewthatIwouldkillordiefor.ButnowsupposeIencounteranoraclewhoknowsthetruthofthematter,orperhapsGodpartsthecloudsandtellsmethatinsixtysecondsHewillrevealwhetherornotfunctionalismistrue.Youinvitemetobet$1,000onfunctionalism.WouldItakethatbet?Ofcoursenot.Nor$10forthatmatter.Areyounuts?Thisisphilosophywearetalkingabout.Inreality,Ihavenoideawhetherfunctionalismistrue.(Lycan2013:115)

Gutting,withwhomLycanisinteractinghere,seemsinclinedtoagreewithLycanaboutthislackofbelief(Gutting2013:135).

‐64‐

thinkandtalkabouttheissueistheroleofexpertphilosophicalopinion.Itwouldweighveryheavilyinmydeliberationsintheimaginedhighstakesscenarios.Andimportantly,itseffectwouldnotbelimitedtomakingmemorehesitantaboutmyanswer,butitwouldhavethepowertoactuallyflipwhatanswerIgive.Inmythinkinginphilosophicalsettings,byverysharpcontrast,theexpertopinionofotherphilosopherscarriesatmosthardlyanyweightatall.

Andthisisforthegood,Ithink.It’sprobablygoodforphilosophicalprogress(whichIdobelievein,despitemyconvictionthatittypicallydoesnotleadtoknowledgeableanswerstothequestionswearefocusedon)thatinthesettingsinwhichwethinkaboutanddiscussphilosophywedonotletconsiderationslikecontraryopinionsbyexpertsmakeusgoallwishy‐washyaboutphilosophicalissues,orworsestill,quicklyflipfromwhatseemsrighttouspersonallytowhatseemstobethemajorityofexpertopinion,butinstead,inaway,sincerelyfeelconfidentaboutthepositionsthatseemrighttouspersonally,passionatelydefendthem,etc.

LetmebrieflyinterjecthowIthinkthisshouldeffecttherecentlyhotdiscussionoftheepistemologyofdisagreement.Manyofthecasesusedintheliteraturewhereitintuitivelyseemsthatoneshould,orthatit’satleastpermissibleto,‘sticktoone’sguns’,asopposedto‘conciliate’whenoneencounters‘peerdisagreement’overanissueconcernso‐called‘beliefs’inareasofcontroversy,likephilosophy.Isuspectthatthereasonit’salrighttosticktoone’sgunsinsuchexamplesisthatthesearen’trealbeliefstobeginwith.(And,relatedly,Ithinkthecredencewealreadyreallyassigntothepropositionsinquestionisnotnearlyashighasonewouldbeledtothinkbyhowconfidentweseemtobe,andinwaysfeelourselvestobe,aboutthoseitemswhenwe’reengagedinphilosophicaldiscussion.)It’salrighttosticktoone’sgunsinthesecasesonlybecausetheyareonlytoyguns,asitwere.47

Andinthesesettingsinwhichwearepassionatelydefendingthingswearenotevenclosetoknowingtobetrue,wefindourselves,amongotherthings,flat‐out

47Inthe(morejarring)casesofpeerdisagreementwhererealgunsareinplay,Ithinktherationalresponseistoreviseone’scredenceintheunderlyingmatterand/or(thoughmostcaseswillinvolveatleastsomeofboth)reviseone’sestimateofthereliabilityofone’sinterlocutor,inatleastroughly(andmaybemoreexactly)thewayproposedin(Worsnip2014).

‐65‐

assertingsuchthings,inflagrantviolationoftheknowledgenormforassertion.Thisisalsoprobablytothegood.Inphilosophyandother‘knowledgedeserts’,aswemightcallthem,wherewe’refocusedonquestionsthatnoneofusknowstheanswersto,itwouldbequiteadragtohavetobeconstantlyhedgingourassertions.So,oftenenough,itseems,wedon’t.Therearegreatdifferencesinpersonalstyleamongphilosophers,somehumblyhedgingtheirclaims,by,e.g.,throwinginparentheticalsofthelikesof‘Ithink’,whereothersmoreboldlyassertaway,hedgesbedamned.Butmostwillfindthemselvesassertingfromthehipinatleastsomephilosophicalsettings.(Inmyobservation,thisoftenhappensinatleastmoderatelyheatedphilosophicaldisputes,wheresomekindofbilateralescalationinprojectedconfidenceoftenseemstooccur.)Andwhenitdoeshappen,theresultingassertionsdon’tseemwrong—oratleast,notinthewaythatoneiswrongtoassert‘There’saservicestationroundthatcorner’toamotoristwhenoneisnowhereneartoknowingthattobeso.Inphilosophicaldiscussion,weseemtohavesomekindoflicensetoassertwithouthedgesthethingsweaccept,eventhoughwedon’tknowthemtobeso,andto,atleastinthatway,actasifweknowthingswedon’treallyknow.

I’mveryopentodifferentwaysofunderstandinghowthislicenseworks.Indeed,somemaywonderwhetherthespeechactsinquestionreallyoughttobeclassifiedasunhedgedassertions,asIhavedoneabove.Intheir(2014),whichisadefenseofcontextualismfromthefactivityproblem,butthen,asIamdoinghere,broadensouttoconsiderwhatphilosophersaregenerallydoingwhentheyputforwardtheircontroversialviews,MartinMontminyandWesSkolits(henceforth‘M&S’)construetheclaimsinquestionassomethingweakerthanassertions:

A key assumption of the argument generating the statability problem is that while in [the context in which she presents her theory], the contextualist asserts the content of her theory. But one may plausibly hold that the contextualist’s utterances have a slightly weaker assertoric force than assertions do. Consider the category of illocutionary acts called weak assertives, which includes conjectures, guesses and hypotheses. These illocutionary acts aim at truth, but their assertoric force is weaker than that of an assertion. On the current proposal, the force of the contextualist’s weak assertives would be somewhere in between the force of a conjecture and that of an assertion. Their illocutionary force would be comparable in strength to that of the weak assertives generated by a parenthetical use of ‘I think’. . . . This strikes us as a plausible description of what typical philosophers do when they defend their views, except that they tend to avoid stylistically frowned upon parentheticals. In a philosophical context, it is understood that many of the claims made are highly controversial and cannot be established decisively. There is thus an implicit understanding that speakers

‐66‐

do not represent themselves as knowing the content of every utterance they make. Utterances expressing controversial philosophical views are thus reasonably interpreted as having weaker assertoric forces than assertions do.

Weak assertives are governed not by a knowledge rule, but by a weaker epistemic requirement: a weak assertive that p is epistemically appropriate only if the speaker has some evidence for p. (Montminy & Skolits 2014: 327)

M&SareclearlybeingdrivenbythesamekindsofconsiderationsIam,andareseekingtoendupatroughlythesamedestination:somekindoflicenseintherelevantsettingsfortheunhedgedspeechactsinquestion.However,Iamnotmyselfinclinedtogettothisdestinationbydenyingthattheclaimsinquestionareassertions.ThismaybeinpartduetomyhavingasomewhatdifferentviewthanM&Sdoonsomeoftherelevantconversationalphenomena.Mostnotably,M&Swrite:

[I]t seems inappropriate to challenge an interlocutor who argues for, say, compatibilism about free will, moral cognitivism or epistemic contextualism, by asking, ‘Do you know that?’ or ‘How do you know?’ This provides further evidence for our claim that the contextualist is not asserting the content of her theory. (Montminy & Skolits 2014: 328)

AndhereIdisagree.Ihearsuchchallengesofteninthesettingsinquestion,andonthesurfacetheyseemtomenomoreconversationallyinappropriatethanaretheunhedgedstatementsthattheyareresponsesto.So,whileM&Sconclude,‘Thismeansthat...ourcontextualistdoesnotrepresentherselfasknowingthecontentofhertheory,’Iammoreinclinedtowardatreatmentonwhichwedoatsomelevelrepresentourselvesasknowingthecontroversialthingsweclaiminthephilosophicalsettingsinquestion,butourdoingsoisexcusable,atleasttosomeextent.

Inthisrespect,myapproachissomewhatlikeWilliamson’s(IsayasIshamelesslyseekto‘triangulate’),whoviews‘livelyseminardiscussion[s]’asexampleswheretheknowledgenormforassertionisoftenviolated,buttheviolationsarenotabigdeal,asheexpressesinthebelowpassage,whichalsocontainsanotherreason,beyondtheconsiderationdiscussedintheaboveparagraph,forwhyIdon’tjoinM&Sindenyingthattheclaimsinquestionareassertions:

To rule that the speakers are not making genuine assertions in such situations would be to trivialize the [knowledge] account [of assertion]. In natural languages, the default use of

‐67‐

declarative sentences is to make assertions, and the situations at issue are not special enough to cancel the default. Rather, the point is that the knowledge account does not imply that asserting p without knowing p is a terrible crime. We are often quite relaxed about breaches of the rules of the game which we are playing. (Williamson 2000: 258)

IshouldnotethatIdon’tnecessarilyjoinWilliamsoninchargingthatexceptingtheclaimsinquestionwould‘trivialize’theknowledgeaccountofassertion:Muchdependshereonwhatkindofcasecanbemadefortakingthemtobe‘specialenough’,asWilliamsonputsit.ButIdojoinhimintakingittobeanadvantageforatreatmentoftheseclaimsthatitnotrulethemtobenon‐assertions.Ifpointedchallengeslike‘Youdon’tknowthat!’,orifchallengingquestionslike‘Howdoyouknowthat?’,seemedasoutofplaceinresponsetotheunhedgedphilosophicalclaimsinquestionastheyaretohedgedclaims,thatwouldbethekindofconsiderationthatwouldtemptmetorulewithM&Sthattheseclaimsarenotassertions.ButIdon’tfindsuchchallengesconversationallyinappropriateinthatway,soIdoclassifytheclaimsasflat‐outassertions,asWilliamsondoes.

However,thoughtheclaimsinquestiondon’tseemspecialenoughtorulethemnon‐assertions,theredoesseemtobesomethingalittlespecialaboutthem–specialenoughtomakemethinkabitmoreisgoingonwiththemthanisexplainedbyjustWilliamson’sgeneral‘notallviolationsoftheruleareabigdeal’treatmentofthem.Toreturntothecontrastpresentedafewparagraphsabove,insettingsinwhichlittleisatstake,itmaynotbeahighconversationalcrimetoasserttoamotoristthat‘There’saservicestationroundthatcorner’,whenyouarenowhereneartoknowingthattobethecase.Andifyouarebeingattackedtoovigorouslyforthatconversationalmisdeed,therewouldbemerittoadefensealongthelinesof:‘Easeuponme!WhatIdidwasnotthathorrible.’Butwhatyoudidwasaconversationalmisdeed,and,Isense(thoughIadmitthisisallverydelicate),itisopentocriticisminaway(andnotjusttoanextent)thatoneisnotopentocriticismforleavingoutahedgewhenclaiminginaheatedphilosophicaldebatethat,say,freewillisincompatiblewithdeterminism.It’sthisthatmakesmesuspectsomekindofspeciallicenseorexcuseisatworkhere.Andtherelevantdifferencebetweenthesecasesseemstometobethattheclaimyoumaketothemotoristissomethingthat,fromherpointofview,youreallymighthaveknown—whichmakesitbadlymisleadingforyoutorepresentyourselfasknowingitbyflat‐outassertingit.Butthephilosophicalclaimisnotthekindofthinganyoneknows,sotherecanbesomespecialexcuseorlicenseforourengaging,asweseemtometodo,insomekindofpretensetoknowsuchcontroversialthingsinphilosophicalsettings,and

‐68‐

therebyflat‐outassertingthingsthatwearenowhereneartoknowingtobethecase.

Thispretenseofknowledgeseemslimitedinwhatitlicenses:Tome,anyway,itdoesnotseemrightinsuchsettingstosaythatyouknowthingsthatyoumerelyaccept.(Thismaybeinpartbecausetheclaimtoknowsomesuchp,becauseitislongerthanisjusttheflat‐outassertionofpitself,can’tbejustifiedbyhowconciseandhandyitis.)However,asI’venoted,itdoesseem,atleasttome,tolicenseflat‐outassertingthesecontroversialthingsthatweaccept,anditalsoseemstoshowupinthesignificanceofsaying‘Idon’tknow’,whichinsuchsettingsseemstoconveythatonedoesnoteven(chooseto)accepttheiteminquestion.So,itseemstomethatweslipintoamodeinwhichweincertainwaystalkasifweknowthecontroversialpositionsweaccept,evenwhenwedon’tknowthemtobetrue.Thisallowsustodispensewiththevarioushedgesthatwouldotherwiselittertoomanyofourclaimsina‘knowledgedesert’,andtoconciselyindicatewhetherweacceptaclaimbyflat‐outassertingitifwedo,andsaying‘Idon’tknow’ifwedon’t.

IfI’mrightthatthelatitudeweenjoyinthismodeofspeechdoesnotgosofarastolicensesayingthatweknowthingsthatwemerelyaccept,oneisputinanawkwardpositioninthatmodewhen,concerningsomeclaimthatonewantstocontinuetoaccept,oneisaskedthequestionsthat,unlikeM&S,Ifindappropriate:‘Doyouknowthat?’or‘Howdoyouknowthat?’.Foritiswrongtoanswerthatonedoesknow(ifI’mrightaboutourlicensebeinglimitedinthatway),butadmittingthatonedoesnotknow,whileengaginginthepretense,wronglyconveysthatonedoesnotevenaccepttheiteminquestion.Ihavefoundthebestoptioninsuchanawkwardsituationisto(perhapstemporarily)disengagefromthepretenseofknowingthesemattersand‘getreal’,byansweringalongthelinesof:‘Well,no,Idon’tknowit,ofcourse,butthatismyposition,andmyreasonsare...’Thisindicatesthatoneisexitingthemodeofspeechassociatedwiththepretenseofknowledgebecause‘Idon’tknow’,whichinthemodeindicatesthattheitemisn’tevenpartofthepositiononechoosestoaccept,isimmediatelyfollowedby‘butthatismyposition’.

‘Gettingreal’fromtimetotimeintheabovewayisprobablyagoodideainanycase,evenifoneisn’tputinanawkwarddialecticalpositionwhichpressuresonetodoso.Forifthisreallyissomekindofapretenseofknowledgethatweengagein,asIamsuggesting,it’sapretensethatoftengetsmistakenforreality,leadingphilosophers,andotherswhogetcarriedawaybyhowtheyfeelandtalk

‐69‐

whenoperatinginknowledgedeserts,into(sometimespathetic)delusionsofknowledge.

Needlesstosay,Idon’thavemythinkingonthesemattersverywellworkedout—andpresentingthepositionsofM&SandofWilliamsononeithersideofmineisnotjustdoneforthepurposeofclarifyingwhatmypositionisbycontrasttotheirs,butalsotopresentreaderswithsomeotherwaysthatIremainopento,bywhichtheseissuesmaybehandled.Thisallseemedworthbringinguphere,becauseitseemsthatworkingoutwhattosayaboutassertion(oratleastwhatappearstobeassertion)inphilosophicalsettingsandotherknowledgedesertswillalsoservetoaddressproblemsImightfaceinthespecialcaseofarticulatingmycontextualistsolutiontoskepticism.Inthemeantime,theworrythat,asacontextualistrespondertoskepticism,Ifacesomespecial‘factivity’problem,generatedbytheinvocationoffantasticallyhighepistemicstandards,isaddressedinsection17,above.

References

Alston,W.P.1996.'Belief,Acceptance,andReligiousFaith',InJ.Jordan,&D.

Howard‐Snyder(Eds.),Faith,Freedom,andRationality(pp.3‐27)Rowman&

LittlefieldPublishers.

Bach,K.2005.'TheEmperor'sNewKnows',InG.Preyer,&G.Peter(Eds.),

ContextualisminPhilosophy:Knowledge,Meaning,andTruth(pp.51‐89).Oxford:

OxfordUP.

Baumann,P.2008.'ContextualismandtheFactivityProblem',Philosophyand

PhenomenologicalResearch,76,580‐602.

‐70‐

‐‐‐‐‐‐2010.'FactivityandContextualism',Analysis,70,82‐89.

Brendel,E.2005.'WhyContextualistsCannotKnowtheyareRight:Self‐Refuting

ImplicationsofContextualism',ActaAnalytica,20,38‐55.

‐‐‐‐‐‐2009.'Contextualism,Relativism,andFacitivty:Analyzing'Knowledge'After

theNewLinguisticTurninEpistemology',InA.Hieke,&H.Leitgeb(Eds.),

Reduction,Abstraction,Analysis(pp.403‐416)OntosVerla.

Cohen,S.1988.'HowtobeaFallibilist',PhilosophicalPerspectives,2,91‐123.

Curley,E.M.1978.DescartesAgainsttheSkepticsHarvardUP.

Davis,W.A.2004.'AreKnowledgeClaimsIndexical?',Erkenntnis,61,257‐281.

DeRose,K.(1990).Knowledge,epistemicpossibility,andscepticism.(Unpublished

Ph.D.).UniversityofCalifornia,LosAngeles,UniversityMicrofilmsInternational.

‐‐‐‐‐‐1992.'Descartes,EpistemicPrinciples,EpistemicCircularity,andScientia',

PacificPhilosophicalQuarterly,73,220‐238.

‐‐‐‐‐‐1998.'Simplemights,IndicativePossibilities,andtheOpenFuture',

PhilosophicalQuarterly,48,67‐82.

‐‐‐‐‐‐2004.'SingleScoreboardSemantics',PhilosopicalStudies,119,1‐21.

‐‐‐‐‐‐2009.TheCaseforContextualism:Knowledge,Skepticism,andContext,Vol.1.

Oxford:OxfordUniversityPress.

‐71‐

Descartes,R.1967.ThePhilosophicalWorksofDescartes,Tr.E.S.Haldane,G.R.T.Ross;

1967ReprintCambridgeUP.

Dinges,A.2014.'EpistemicContextualismcanbeStatedProperly',Synthese,

Feldman,R.2001.'SkepticalProblems,ContextualistSolutions',Philosophical

Studies,103,61‐85.

Frances,B.2012.'DiscoveringDisagreeingEpistemicPeersandSuperiors',

InternationalJournalofPhilosophicalStudies,20,1‐21.

Goldman,A.I.1967.'ACausalTheoryofKnowing',JournalofPhilosophy,64,357‐

372.

Gutting,G.2009.WhatPhilosophersKnow:CaseStudiesinRecentAnalytic

PhilosopphyCambridgeUP.

‐‐‐‐‐‐2013.'RepliestoMargolis,Lycan,andHenderson',SouthernJournalof

Philosophy,51,133‐140.

Hume,D.1978.ATreatiseofHumanNatureOxfordUP.

Jäger,C.2012.'ContextualismandtheKnowledgeNormofAssertion',Analysis,72,

491‐498.

Johnsen,B.C.2001.'ContextualistSwords,SkepticalPlowshares',Philosophyand

PhenomenologicalResearch,62,385‐406.

‐72‐

Koethe,J.2005.Scepticism,Knowledge,andFormsofReasoning.Ithaca,NY:Cornell

UniversityPress.

Kornblith,H.2000.'TheContextualistEvasionofEpistemology',PhilosophicalIssues,

10,24‐32.

Montminy,M.,&Skolits,W.2014.'DefendingtheCoherenceofContextualism',

Episteme,11,319‐333.

Moore,G.E.1959.'ProofofanExternalWorld',PhilosophicalPapers(pp.127‐150).

London:GeorgeAllen&UnwinLtd.

Richard,M.2004.'ContextualismandRelativism',PhilosophicalStudies,119,215‐

242.

Schiffer,S.1996.'ContextualistSolutionstoScepticism',Proceedingsofthe

AristotelianSociety,96,317‐333.

Schwitzgebel,E.2001.'In‐betweenBelieving',PhilosophicalQuarterly,51,76‐82.

‐‐‐‐‐‐2002.'APhenomenal,DispositionalAccountofBelief',Noûs,36,249‐275.

Sosa,E.1999.'HowtoDefeatOppositiontoMoore',PhilosophicalPerspectives,13,

141‐153.

‐‐‐‐‐‐2000.'SkepticismandContextualism',PhilosophicalIssues,10,1‐18.

‐73‐

‐‐‐‐‐‐2004.'RelevantAlternatives,ContextualismIncluded',PhilosophicalStudies,

119,35‐65.

Stanley,J.2005.KnowledgeandPracticalInterests.Oxford:OxfordUniversityPress.

Stine,G.1976.'Skepticism,RelevantAlternatives,andDeductiveClosure',

PhilosopicalStudies,29,249‐261.

Williamson,T.2000.KnowledgeanditsLimits.Oxford:OxfordUniversityPress.

‐‐‐‐‐‐2001.'CommentsonMichaelWilliams'"Contextualism,Externalismand

EpistemicStandards"',PhilosophicalStudies,103,25‐33.

‐‐‐‐‐‐2005.'KnowledgeandScepticism',InF.Jackson,&M.Smith(Eds.),TheOxford

HandbookofContemporaryPhilosophy(pp.681‐700)OxfordUP.

Worsnip,A.2014.'DisagreementaboutDisagreement?whatDisagreementabout

Disagreement?',Philosophers'Imprint,14,1‐20.

Wright,C.2005.'ContextualismandScepticism:Even‐Handedness,Factivityand

SurreptitiouslyRaisingStandards',PhilosophicalQuarterly,55,236‐262.