Introduction
JessicaBrownandMonaSimion
1.1.Introduction
Thenotionofdefeathasbeencentraltoepistemology,practicalreasoning,and ethics.Withinepistemology,itisstandardlyassumedthatasubjectwho knowsthat p,orjustifiablybelievesthat p canlosethisknowledgeorjustified beliefbyacquiringaso-called ‘defeater’,whetherevidencethatnot-p,evidence thattheprocesswhichproducedherbeliefisunreliable,orevidencethatshe haslikelymisevaluatedherevidence.Withinethicsandpracticalreasoning,it iswidelyacceptedthatasubjectmayinitiallyhaveareasontodosomething, althoughthisreasonislaterdefeatedbyheracquisitionoffurtherinformation.
Thenotionofdefeathasbeencentraltoawiderangeofdifferentphilosophicaldebates,including,butnotlimitedto:
(1) Thenatureofjustificationandknowledge:sinceknowledgeand justificationaretakenbymanytobedefeasible,theextenttowhich oneaccountoranotherofthenatureofknowledge/justificationisable toaccountfor/accommodatedefeatconstitutesanimportantground forassessingitstheoreticalcredentials(seee.g.Sudduth2018).
(2) Internalismversusexternalism:severalepistemologistsworrythat epistemicexternalismhasahardtimeaccommodatingpsychological defeat;atthesametime,conversely,ifjustificationsuperveneson mentalstatesalone,asperinternalism,itseemsmysteriousthatit couldbedefeatedbynormativedefeaterslyingoutsideofthecognizer’ s ken(seee.g.Pappas2017).
(3) Epistemicnormsandreasons:foraccommodatingthephenomenonof defeat,debatesonepistemicnormsandepistemicreasonsoweus,ata minimum,anaccountofepistemicnormativeoverriding,aswellasan accountofreasonsagainstbelief(seee.g.Simion2020)
(4) Evolutionarydebunkingarguments:thenatureofdefeathas repercussionsforthedebateaboutthemetaethicalimplicationsof
evolutionaryexplanationsofmorality.Itisarguedthatsome,ifnotall, humanmoralbeliefsaretheproductofevolutionaryforces:wewere selectedforhavingusefulratherthantruemoralbeliefs.Learningabout theevolutionaryoriginofourmoralbeliefsundercutstheirprimafacie justification,orsothechallengegoes(seee.g.Silva2016).
(5) Evidenceandhigher-orderevidence:sinceevidenceiswidelytakento bedefeasible,aplausibleaccountofevidenceshouldcomewitha correspondingplausibleaccountofdefeat.Forinstance,oneimportant desideratumforanysuchaccountisthatitexplainsthedefeatingpower ofhigher-orderevidence,namelyofevidencethatone’ s first-order beliefsaretheoutputofa flawedprocess(seee.g.Brown2018).
(6) Closureandtransmission:onepopularsolutiontoallegedfailuresof closureprinciplesforknowledgeandtransmissionprinciplesforwarrantisknownas ‘thedefeatsolution’:roughly,thethoughtgoes,closure andtransmissionholdprimafacie,andtheintuitionsoffailureareto beexplainedintermsofpsychologicaldefeat.Thissolution,ofcourse, hangsontheassumptionthatthereissuchathingtobeginwith:i.e. thatpsychologicaldefeatisagenuineepistemiccategory(seee.g.Pryor 2004).
(7) Disagreement:onewaytocharacterizethedebatebetweenconciliatory andsteadfastviewsofdisagreementisascentredaroundthequestion: canthetestimonyofone’speercarrydefeatingpower?Steadfast-ism answers ‘ no ’,conciliationismanswers ‘ yes ’.Thecorrectaccountofthe natureofdefeatcanhelpsettletheissue(seee.g.FrancesandMatheson 2019).
(8) Reductionismversusanti-reductionismabouttestimony:saythata suspectformurderStellsyouthatshedidnotdoit.Accordingtoboth ofthemainviewsintheepistemologyoftestimony,youarenot justifiedtobelieveS.Accordingtoreductionism,that’sbecauseyou alwaysneedpositive,non-testimonialreasonstobelievewhatyouare beingtold.Incontrast,accordingtoanti-reductionism,youareprima faciejustifiedtobelieveS,butyourjustificationisdefeated.Thecorrect natureofthenatureofdefeatwilllikelygoalongwayinthedirectionof settlingtheissue(seee.g.Green2008).
Itisusefultocategorizecontributionstothedefeatliteratureasfallingunder oneofthefollowingtwobroadcategories:(i)thenatureandextentofdefeat and(ii)kindsofdefeaters.Inwhatfollows,wewill firstbrieflyrunthroughthe stateoftheartthuscategorized(Sections1.2and1.3).Last,wegivean
overviewofthevolume’schaptersandexplainhowtheybuildonthestateof theart(Section1.4).
1.2.TheNatureandExtentofDefeat
1.2.1DefeatersasReasons
The firstandnowconsideredtheclassicviewonthenatureofdefeatin epistemologyisduetoPollock(1986).Accordingtothisview, D isadefeater of E’ssupportfor p for S ifandonlyif(i) E isareasontobelieve p for S,and (ii) E&D isnotareasontobelieve p for S (henceforth Pollock’sview).
Theaccounthasalotgoingforit; first,itpromisestocutacrossnormative domains,invirtueofbeingframedintermsofreasons;afterall,epistemologistshardlyenjoyexclusivityonreasons.WithPollock’sviewinplay,itiseasy toseehowwecouldgeneralizeittocoverdifferenttargets(e.g.actions)and types(e.g.moral,prudential)ofnormativity.Second,Pollock’sviewmakes goodontheintuitivethoughtthatdefeatersareactualizersofthepossibilityof apositivenormativestatustobeoverriddenorundercut;whattheviewsays, inanutshell,isthatdefeatersarethekindofthingsthatrenderapermissible beliefimpermissible.
Morerecently,though,Pollock’sviewhascomeunderheavyattack.First,as Chandler(2013)pointsout,Pollock’sviewisproblematicinthatitcannot accommodatewhatcanbecalled justifyingdefeaters.Asthenamesuggests,a justifyingdefeater JD of E’ssupportfor p for S doesdoubleduty:itprevents E frombeingareasontobelieve p for S but,atthesametime,itgives S anew reasontobelieve p.So, JD isadefeaterfor S,butitdoesnotfulfilclause(ii)of Pollock’sview,as E&JD stillisareasonfor S tobelieve p.
Second,theviewhasbeenfoundwantingonpriorplausibilityinvirtueof theveryfactthatit’sstatedintermsofreasons.Pollock’sview fitssnuglywith evidentialismaboutjustification:theepistemicstatusofabeliefisdetermined bythereasonsforandagainstbelieving.Assuch,onPollock’sview,thereisno defeatwithoutreasons.Withtheincreasedpopularityofexternalist,processbasedaccountsofjustification,however,conceivingofdefeatinterms ofreasonsisnotveryhelpful.Onereasonforthisisthat,onstandard formsofreliabilism(processreliabilism,virtuereliabilism,properfunctionalism),reasonsareneithernecessarynorsufficientforjustification.If reasonsdon’thaveanyjustificatorypower,however,itismysteriouswhy theymighthavedefeatingpowers.Furthermore,notabledefendersof
reliabilismdenythatreasonsaresubstantiveepistemicnormativecategoriesto beingwith(e.g.Lyons2009;Kornblith2015).
1.2.2DefeatersasReliableProcesses
Reliabilisttheoriesofjustificationhavebeenextremelypopularinthelast threedecadesandcomeinavarietyofforms,thegistoftheviewisthatabelief isjustifiedifandonlyifformedviaa(normally)reliableprocedure,orability. Reliabilismisatheoryofprimafaciejustification.Assuch,inlinewith normativetheoriesingeneral,itneedsatheoryofdefeatinordertohold water.ThestandardreliabilistaccountofdefeatcomesfromAlvinGoldman:
Thealternativereliableprocessaccount(ARP): S’sbeliefisdefeatedifthere arereliable(orconditionallyreliable)belief-formingprocessesavailableto S suchthat,if S hadusedthoseprocessesinadditiontotheprocessactuallyused, S wouldn’thaveheldthebeliefinquestion(Goldman1979).
OnecanseehowARPisanelegantreliabilisttranslationofthePollockian thoughtthatdefeatisthekindofnormativeentitythat,whentakeninconjunction withtheextantepistemicsupportfortherelevantbelief,failstorenderitjustified.
BobBeddor(2015)isthe locusclassicus forcriticismofARP;ifBeddoris right,ARPisbothtooweakandtoostrong.AgainstARP’ssufficiencydirection,Beddoroffersthefollowingcase:
ThinkingaboutUnger: Harryseesatreeinfrontofhimat t.Consequently,he comestobelievethepropositionTREE: 〈Thereisatreeinfrontofme〉 at t.Now, HarryhappenstobeverygoodatformingbeliefsaboutwhatPeterUnger’s1975 time-slicewouldadviseonetobelieveinanysituation.Callthiscognitiveprocess hisUngerPredictor[...].What’smore,[...]wheneveritoccurstoHarrythat Ungerwouldadvisehim(Harry)tosuspendjudgementabout p,thiscausesHarry to[...]suspendjudgementabout p.So,ifHarryhadusedhisUngerPredictor,he wouldhavecometo[...]suspendjudgementregardingTREE.
WhatthiscasesshowsisthatARPistooweak:contraARP,formybelief that p tobedefeated,itisnotenoughthatIwouldchangemymindabout p in acounterfactualworldduetoemployingsomereliableprocess.Whatthis objectionidentifiesisthatARPisnormativelytooweak:justbecauseIwould changemymindinworld W,itdoesnotfollowthatI should changemymind inworld W:defeatisanormativenotion.
Here’sBeddor’scaseagainstARP’snecessitydirection:
Jobopening: MashatellsClarencethatherdepartmentwillhaveajob openinginthefall.ClarencebelievesMasha;assumingthatMashaisusually reliable,Clarence’sbeliefcountsasprimafaciejustified.Sometimelater, ClarencespeakswiththeheadofMasha’sdepartment,Victor,whoinforms himthatthejobsearchwascancelledduetobudgetconstraints.Nowsuppose thatClarenceharboursadeep-seatedhatredofVictorthatcauseshimto disbelieveeverythingthatVictorsays,what’smore,noamountofrational reflectionwouldridClarenceofthisinveteratedistrust.Consequently,he continuestobelievethattherewillbeajobopeninginthefall.
ThiscaseshowsthatARPisalsotoostrong:justbecause,inallcounterfactualworlds,Iwouldirrationallyandstubbornlyholdontomybelief,itdoes notfollowthatI should doso.Onceagain,ARPisnotnormativeenoughtodo thejobitissupposedtodo.
1.2.3DefeatScepticism
Wehaveseenthat,withinepistemology, itisstandardlyassumedthatasubject whoknowsthat p,orjustifiablybelievesthat p,canlosethisknowledgeorjustified beliefbyacquiringadefeater.Withinethicsandpracticalreasoning,itiswidely acceptedthatasubjectmayinitiallyhaveareasontodosomethingalthoughthis reasonislaterdefeatedbytheiracquisitionoffurtherinformation.
However,thetraditionalconceptionofdefeathasrecentlycomeunderattack. Somehavearguedthatthenotionofdefeatisproblematicallymotivated (Lasonen-Aarnio2010,2012,2014;HawthorneandSrinivasan2013). Accordingtoasecondstrandofattack,thenotionofrevisionarydefeatis inconsistentwithimportantdesiderataonanaccountofknowledgeorjustificationsuchasnaturalism,externalism,evidentialism,orarule-basedaccount (Greco2010;Lasonen-Aarnio2010,2012,forthcoming,Beddor2015;BakerHytchandBenton2015;Weatherson,2019).Afurtherworryconcernshowthe defeatofaproposition’sstatusasevidenceiscompatiblewithaBayesianframework(e.g.Weisberg2009;Pryor2013;Greco2017).Furthermore,thosewho adoptasteadfastviewonpeerdisagreementcanbeunderstoodasdenyingthe defeatingeffectofevidencethatapeerdisagreeswithone(e.g.Kelly2005).
Inthelightoftheseattacks,somehavesuggestedthatweshoulddiagnose theintuitionthatthereissomethingwrongwithsomeonewhocontinuesto
believethat p inthefaceofdefeatingevidencenotbythesuggestionthatthe relevantbelieflosesitsstatusasjustifiedorasconstitutingknowledgebut insteadinsomeotherway,e.g.bysomevarietyoferrortheory.
AccordingtoWilliamson(2009:315),forinstance,whenweintuitthat one ’sknowledgegetsdefeated,weare,infact,confusingbetweenwhetherone knowsandwhetherit’sprobableonone’sevidencethatoneknows.Another popularsuggestionisthatcontinuingtobelievethat p insuchcircumstancesis toexhibitanepistemicdispositionlikelytoleadtotroubleinthelongrun (Lasonen-Aarnio2010;HawthorneandSrinivasan2013).Accordingtothis view,incasesofallegeddefeat,ourintuitionsfailtodistinguishbetween impermissibilityandmereblameworthinessgeneratedbyirresponsible behaviour.
Againstdefeatscepticism,inrecentwork,oneofus(Brown2018:Ch.5)has arguedthatitinvolvesanunnoticedandunacceptablyhighcost:oncewe explicitlydistinguishbetweencontributory(i.e.synchronic)andrevisionary (i.e.diachronic)defeat,itishardtoargueagainstrevisionarydefeat.Thatis becauseitisdeeplyimplausibletodenythecontributorynotionofdefeat i.e. roughly,evidenceagainst p thatispartoftheinitialbodyofevidence.For,to dosoiseffectivelytodenythatwhenevidenceaffectsjustification,itisone’ s overallevidencethatmattersandnotmerelyapartofit.But,onceweaccept thecontributorynotionofdefeat,itishardtodenytherevisionarygiventhe plausibleepistemicsymmetrybetweencasesofcontributoryandrevisionary defeat.
1.2.4Higher-OrderDefeat
Higher-orderevidenceisevidenceaboutwhatevidenceonepossessesorwhat conclusionsone’sevidencesupports.Oneinterestingquestioninvolving higher-orderevidenceis:howshouldourbeliefsrespondtoourbeliefsabout ourbeliefswhenwedohavehigher-orderevidence(Feldman2005; Christensen2010;Kelly2005,2010).Arelatedissueinrecentepistemology isexplainingthedefeatingpowerofhigher-orderevidence,namelyofevidence thatone’ s first-orderbeliefsaretheoutputofa flawedprocess(seeforinstance Christensen2007,2010;Kelly2010;Lasonen-Aarnio2014).Suppose,for illustration,thatyouareapilotwhocalculateswhetheryouhaveenough fueltoreachtheclosestairport.Uponcalculating,youreachtheconclusion thatyouhavemorethanenoughfueltogettoanairport fiftymilesfurther thantheoneinyourinitialplan.Supposeyouthenglanceatthealtimeterto
seethatyou’reat10,500feetandrememberthathypoxiaisariskataltitudesof 10,000feetandhigher.Younowhaveevidencethatyoumighthavehypoxia, andthusyouhaveevidencethatyoumighthavemiscalculated.Areyounow justifiedinbelievingthatyoucangettothemoredistantairport?Areyou justifiedinbelievingthatyourevidencesupportsthatclaim?(Christensen 2010).
Answering ‘ yes ’ tobothquestionsisintuitivelyextremelyproblematic:it licensesanextremelyformofdogmatism.Answering ‘ no ’ tothe firstquestion and ‘ yes ’ totheseconddoesn’tseemverypromisingeither:evenifyoucan’t actuallybringyourselftobelieve F,beingjustifiedinbelievingyourevidence supports F primafaciejustifiesyouinbelieving F.Answering ‘ no ’ toboth questionsisthetraditionalwaytogoforchampionsofdefeat,andprobably themostwidelyspreadviewintheliterature.Forone,thisistheanswerthat suggestsitselfonboththemainviewsonthenatureofdefeat(thereasonsbasedandthealternativereliableprocess-basedaccounts).
Onehighlydebatedviewintheliteraturecombinesdefeatscepticismwith answering ‘ yes ’ tothe firstsectionand ‘ no ’ tothesecondquestion.Onthis view,then,yoursecond-orderevidencefailstodefeatyour first-orderjustification:youarestilljustifiedtoholdyour first-orderbeliefthatyou’llreachthe secondairport,butyouarenolongerjustifiedinbelievingthatyourevidence supportsyourclaim.Thisviewhasbecomeknownasthe ‘level-splittingview’ (Lasonen-Aarnio2010,2014;Wedgwood2011;Coates2012).Thus,onthis ‘level-splittingview’,thepilotisjustifiedtobelievethatthey’regoingtoreach thesecondairportandthattheirevidencedoesnotsupportthisconclusion (henceforthRationalAkrasia).
Feldman(2005:110–11)arguesthatitisimpossibleforthisbelieftobeboth trueandreasonablesincethesecondconjunctunderminesthereasonableness ofthe firstconjunct(seealsoe.g.Elga2007;Christensen2007;2010;Horovitz 2014;Titelbaum2015).Furthermore,thelevel-splittingviewhasalsobeen arguedtolicenceproblematictheoreticalandpracticalreasoning(e.g.Brown 2018;Horovitz2014).Forinstance,ifRationalAkrasiaobtains,ifthepilot believesthatshehasenoughfueltoreachthesecondairport,thenitseemsthat shecanexploittheclaimthatshehasenoughfueltoreachthesecondairport inherpracticalreasoning.Forexample,shemightreasonthatsinceshehas enoughfueltoreachthesecondairport,sheshouldnotbothertolandonthe firstairport.Nonetheless,whileignoringthe firstairportonthegroundsthat shehasenoughfueltoreachthesecondone,sheoughttoalsoadmitthatit’ s unlikely onherevidencethatshe’llreachthesecondairport!Butactinginthis wayseemsentirelyunreasonable.
1.3.1ByMechanism
Howdoesdefeatwork?Pollock’sdistinctionbetweenrebutting(alsoknownas overriding)andundercutting(alsoknownasundermining)remainsthe classicanswertothisquestion.
AccordingtoPollock(1986),therearetwowaysinwhichdefeatersact: roughly,theyeitherspeakagainstthecontentofone’sbelief,oragainstthe credentialsofitsformationprocedure.Toseehowthisworks,saythatmy friendMarytellsmethatTRAIN:Thetrainleavesateightsharp,andthereby IcometojustifiablybelievethatTRAIN.Now,saythatmyothergoodfriend, Alice,sharplydisagrees:accordingtoAlice,thetrainleavesat8.30.The thoughtisthat,afterreceivingAlice’stestimony,Iamnolonger(fully) justifiedinbelievingTRAIN,invirtueofacquiringevidenceagainstTRAIN beingthecase.Inthiscase,Alice’stestimonyconstitutesitselfinarebutting defeaterformybelievingTRAIN.
Incontrast,undercuttingdefeatersspeakagainstthecredentialsofthe sourceofmybelief.Considerforinstance,analternativescenarioinwhich AlicetellsmethatMaryisacompulsiveliarabouttrains’ schedules.After receivinghertestimony,Iamnolonger(fully)justifiedinbelievingTRAIN basedonMary’sassertion.
OnPollock’sview,arebuttingdefeater,then,forabeliefthat p of S isa reasonfor S foreitherbelievingnon-p orforbelievingsomeproposition q, where q isincompatiblewith p.Inthat,onPollock’sview,rebuttingdefeaters workbyoverridingthereasonsinfavourofbelievingthat p.Incontrast, undercuttingdefeatersdon’tspeakagainst p,butratheragainstthecredentials ofthebeliefformationprocedureemployedinformingthebeliefthat p.An undercutting defeaterforabeliefthat p of S isareasonfor S attacksthe connectionbetween S’sgroundforbelieving p and p.
OneworrythatarisesaboutPollock’sdistinctioniswhetheritisplausible thatitisanexclusiveone,astheaccountseemstosuggest.Pryor(2013),for instance,givesanexamplemeanttoillustratethatarebuttingdefeatercan,in virtueofitsrebuttingnormativepower,alsoconstituteitselfinanundercuttingdefeater:toseethis,suppose S’sjustificationforbelieving p fromevidence E isrebutted,as S acquiresnewevidence E*thatmorestronglysupports q, whichisincompatiblewith p.Whenthishappens,Pryorargues, S isleftwith thequestionofwhy p isfalseevenif E istrue.Pryor’ssuggestionisthatthebest answeravailableto S isthat,inthepresentcontext, E isnotareliableindicator
ofthetruthof P.Hence E*alsoactsasanundercuttingdefeater,sinceit supplies S withsomeevidenceagainstthereliabilityof E insupporting p.
Incontrast,Sturgeon(2014)arguesthatundercuttingandrebuttingdefeatersexemplifytwoessentiallydifferentkindsofdefeat.AccordingtoSturgeon, forinstance,incontrasttorebuttingdefeat,undercuttingdefeatisnotfully independent:itactsintandemwiththesubject’sbeliefthatthesourceunder attackistheonethatisactuallyresponsibleforgeneratingthebeliefin question.
1.3.2ByNormativeStatus
Animportantquestionaddressedintheliteratureondefeatconcernsthe normativestatusofdefeaters.Thenormativestatusquestionconcernsthe issueofwhetherdefeatersneedtobejustifiedthemselves.Lackey’s(1999) distinctionbetweendoxastic(akamerelypsychological)andnormativedefeatersiscentraltothisdebate.InLackey’sview,doxasticdefeatersaremerely psychologicaldefeaters:theyarebeliefsof S,notnecessarilytrueorjustified, thatspeakinfavouroftheclaimthat S’sbeliefthat p isfalseorisbasedonan unreliablesource.Normativedefeatersare(good)reasonsforentertaining doxasticdefeaters.AccordingtoLackey,then,theanswertothenormative statusquestionis ‘ no ’:defeatersneednotbejustifiedthemselves.
ManyepistemologistsagreewithLackey,anditisfairtosaythatLackey’ s distinction whichitselfentailsthatdefeatersneednothavepositiveepistemic status hasbecomeaclassicintheepistemologicalliteratureofthelastyears. AccordingtoAlvinPlantinga(2000:364–65),forinstance,irrationaland unwarrantedbeliefscandefeatbeliefsthathaveimpeccableepistemicstatus. SupposeIbelievethatI’mmadeof flesh,blood,andbone.Ithencometo believe,duetosomecognitivedisorder,thatmyheadismadeofblownglass. AccordingtoPlantinga,giventhatIcometoholdthissecondbeliefInowhave adefeaterforthepriorbelief,evenifthedefeaterwasformedbywayof cognitivemalfunction.AccordingtoPlantinga,thiscaseisonewhereIhave a ‘rationalitydefeater’:giventhatIacquirethesecondbelief,itisnolonger rationalformetoholdthe firstone.
ContraPlantinga,Alston(2002)arguesthatonlybeliefswithpositive epistemicstatuscandefeatbeliefsthathavepositiveepistemicstatus,and, furthermoreabelief D candefeatbelief A onlyif D hasgreaterwarrant than A.Oneofus(Simion2020)hasalsoarguedthattheexistenceof merelypsychologicaldefeatimpliestheexistenceofmerelypsychological
justification,i.e.itimpliesjustificationinternalism.Ifthisisso,externalists aboutjustificationshouldbeespeciallyworriedaboutthecategoryofmere psychologicaldefeat.
1.3.3ByPsychologicalStatus
Anotherimportantquestionfortaxonomizingdefeatconcernswhetherdefeatersneedtoenjoypsychologicalstatusornot,or,moreclearlyput,whether onlymentalstates(e.g.beliefs)candodefeatingwork,or,tothecontrary,facts outsideourskullcanalsohavedefeatinginfluenceoverthenormativestatusof ourbeliefs.
Itisfairtosaythatitiswidelyacknowledgedthat,inthecaseofknowledge defeaters,theanswertothepsychologicalstatusquestionis ‘ no ’:factsoutthere intheworldcanactuponourknowledgeablebeliefsandtherebylowertheir epistemicstatus.TheclassiccasetoillustratethispointisGilbertHarman’ s (1973:143–44)assassinationcase.Supposethatapoliticalleaderhasbeen assassinated.Areporterwhoisawitnesstotheassassinationdictatesdetailsof theeventtohisnewsagencysothatthestorymaybeincludedintheday’ s final editionofthepaper.Jillpicksupthepaperandreadsthestoryandbelieves thatthepoliticalleaderhasbeenassassinated.However,beforeJillpicksupthe newspaperandreadsthestory,loyaliststothepoliticalleaderdeclareon nationwidetelevisionthatthebulletactuallystruckandkilledsomeonein thepoliticalleader’sentourage.Jillreadsthetruestoryinthepaperbutmisses thefalsereportontelevision.ManypeoplebelievewithHarmanthatJill doesn’tknowthatthepoliticalleaderhasbeenassassinated.Thereare,however,severaldifferentviewsonthemarketastowhatexplainsthisdatum. AccordingtoSwinburne(2001)andPollock(1986),theexplanationissocial: thefactthatJilldoesnothaveknowledgeistheconsequenceoftherebeinga trueproposition(suggestiveofadefectinjustification)thatiswidelybelieved inJill’ssociety.Othersthinkthattheexplanationlieswiththeeasyavailability oftherelevantdefeatinginformation(Sudduth2018).
Wehaveseenthatepistemologiesthatincorporatedoxasticdefeaterstypicallytakethemtodefeatjustification(Alston1989:238–9;Bergmann2006: 155–6)orsomespeciesofrationality(Plantinga2000:357–66).Whenitcomes tonon-doxasticdefeaters,however,thingslookslightlydifferent:whilenondoxasticdefeatersarewidelytakentohaveanimpactonknowledge,whether merefactsoutthereintheworldcandefeatjustificationisamorecontroversialissue.Theepistemicinternalistwillhavetoanswer ‘ no ’:afterall,by
internalism’slights,justificationsupervenesexclusivelyonfactsinternaltothe cognizer’smind.Epistemicexternalists,however,canaffordtohavemixed viewsabouttheissue.SandyGoldberg(2016),forinstance,arguesthat evidenceonedoesnothave,butthatone should havehad,candefeatone’ s justification.Hethinksthiscomesaboutinvirtueofsocialfacts,i.e.invirtueof thereasonableexpectationsofothersgroundedinnormsinternaltoasocial practice.AccordingtoLackey(2008),too,merenormative(non-doxastic) defeatcandefeatjustificationincaseswherethebelieverhasnegligently ignoredimportantcounterevidence,invirtueofepistemicresponsibility constitutingacrucialcomponentofepistemicjustification.
1.4.SummaryoftheVolume
Thisvolumebringstogetherrecentworktore-examinetheverynotionof defeat,anditsplaceinepistemology,andinnormativitytheoryatlarge.
InChapter2,TheNormativityofKnowledgeandtheScopeandSources ofDefeat,SanfordC.Goldberginvestigatesthenatureandscopeofdefeat withanaimtoargueinsupportofthecategoryofnormativedefeat,aswell asinfavouroftherebeingasocialdimensiontoknowledge.Tothiseffect, thechapterappealstoapriorgraspofthenormativityofknowledgeitself itsroleinentitlingasubjecttoconfidenceandinauthorizingothersto believeonthestrengthofone’sepistemicstanding toshedlightonthe natureandscopeofdefeat.Goldberg’sstrategyistofocusoncasesinwhich anotherwiseepistemicallywell-positionedsubjectfailstoenjoythesenormativestandings,andtoarguethatthebestexplanationisthepresenceof normativedefeaters.
Chapter3,TheStructureofDefeat:Pollock’sEvidentialism,Lackey’ s Framework,andProspectsforReliabilism,byPeterGrahamandJackLyons, investigatesthestructureofdefeat.Ithastwomainaims;the firstismostly critical:itarguesthatseveralclassicaldistinctionsthattheliteratureondefeat endorsesareproblematic.Inparticular,accordingtoGrahamandLyons,the traditionalcategorizationofdefeatersasdoxasticandnormativeismistaken: first,unjustifiedbeliefscan’tdefeat,thereforetherearenosuchthingsas merelydoxasticdefeaters;second,thereasonwhyignoredevidencecandefeat isdifferentfromtherationaletraditionallytakentosupporttheexistenceof normativedefeat.Thesecondaimofthechapteristodevelopanovel, reliabilism-friendlyviewofthenatureofdefeat.Onthisview,havinga defeaterforabeliefthat p isamatterofeitherhavingwarranttobelieve
not-p,orelsehavingwarranttobelievethattheirwarrantforbelievingthat p areinadequate(wherewarrantisunderstoodinnon-evidentialistterms).
Chapter4,LosingKnowledgebyThinkingaboutThinking,byJennifer Nagel,putsforthanoveldefenseofinfallibilisminepistemologyagainstthe threatcomingfromthephenomenonofdefeat.Defeatcasesareoftentakento showthateventhemostsecurelybasedjudegmentcanberationallyunderminedbymisleadingevidence.JenniferNagelarguesthatdefeatcasesreally involvenotanexposureofweaknessinthebasisofajudgement,butashiftin thatbasis.Forexample,whenthreateningdoubtsareraisedaboutwhether conditionsarefavourableforperception,oneshiftsfromabasisofunreflective perceptualjudgementtoabasisofconsciousinference.Inthesecases,thebasis ofone’sknowledgeislost,ratherthanrationallyundermined.
Chapter5,DispositionalEvaluationsandDefeat,byMariaLasonen-Aarnio, arguesthatwhatexplainstheintuitionofimpermissibilityinputativecasesof defeatresidesnotintheimpermissibilityofthetargetbeliefsthemselves,but ratherinthebeliever’scriticizabilityformanifestingbadepistemicdispositionstobelieve.Subjectswhoretaintheirbeliefsinthefaceofhigher-order evidencethatthoseverybeliefsareoutputsof flawedcognitiveprocessesareat leastveryoftencriticizable.This,however,onLasonen-Aarnio’saccount,is notbecausesuchhigher-orderevidencedefeatsvariousepistemicstatusessuch asjustificationandknowledge.Instead,shearguesthattheymanifestdispositionsthatarebadrelativetoarangeofcandidateepistemicsuccessessuchas truebeliefandknowledge.Inparticular,beingdisposedtoonlygiveupbelief inresponsetohigher-orderevidence whenthatevidenceisnotmisleading wouldrequiresubjectstohavedispositionsthatdiscriminatebetweencasesin whichtheiroriginalcognitiveprocessesis fine,andcasesinwhichtheymerely seemedtobe fine.But,accordingtoLasonen-Aarnio,suchdispositionsarenot normallyhumanlyfeasible.
InChapter6,Suspension,Higher-OrderEvidence,andDefeat,Lordand Sylvanfocusontheepistemiceffectofhigher-orderevidence.Thechapter makestwomainclaims:anegativeoneandapositiveone.Theirnegative claimisthatextremeviewsabouttheissue claimingeitherthathigher-order evidencehastrumpingeffects,orthatithasnonewhatsoever aremistaken. Thepositiveviewinturnhastwoparts.The firstpartdefendstheideathat higher-orderevidenceprovidesdirectreasonsforsuspendingjudgmentthat typicallyleaveevidentialsupportrelationsonthe firstorderintact:insteadof destroyingtheserelations,thereasonsforsuspensiondefeatorcompetewith theepistemicreasonsforbeliefgeneratedbytheserelations.Secondly,the LordandSylvanframeworkpurportstoexplainhowthisdefeatispossibleby
showinghowthesedistinctivereasonsforsuspensionofjudgment flowfrom theconstitutionofsuspensionofjudgment.
InChapter7,ReasonsforReliabilism,BobBeddormotivatesanddevelops asynthesisbetweentwoleadingapproachestojustificationwhichare typicallydevelopedinisolationfromeachother;the firstonecomesfrom thereliabilisttradition,whichmaintainsthatabeliefisjustifiedprovidedthat itisreliablyformed,whilethesecondonecomesfromthe ‘ReasonsFirst’ tradition,whichclaimsthatabeliefisjustifiedprovidedthatitisbasedon reasonsthatsupportit.OntheviewproposedbyBeddor,justificationis understoodintermsofanagent’sreasonsforbelief,whichareinturnanalysed alongreliabilistlines:anagent’sreasonsforbeliefarethestatesthatserveas inputstotheirreliableprocesses.Beddorarguesthatthissynthesisallowseach traditiontoprofitfromtheother’sexplanatoryresources.Inparticular,it enablesreliabiliststoexplainepistemicdefeatwithoutabandoningtheirnaturalisticambitions.
Chapter8,Knowledge,Action,andDefeasibilitybyCarlottaPavese,reviews somemotivationsfora ‘knowledge-centredpsychology’—apsychologywhere knowledgeenterscentrestageinanexplanationofintentionalaction, it outlinesanovelargumentfortheclaimthatknowledgeisrequiredfor intentionalaction,anddiscussessomeofitsconsequences,inparticularfor thedebateonthedefeasibilityofknow-how.AsPaveseargues,aknowledgecentredpsychologymotivatestheintellectualistviewthatknow-howisa speciesofknow-that.Initsmoreextremeform,thisviewiscommittedtoan epistemologicallysubstantialclaim i.e.thattheepistemicprofileofknowhowisthesameasthatofpropositionalknowledge.Ifthatiscorrect,one corollaryofintellectualismisthatthedefeasibilityofknow-howpatternswith thatofknowledge.Againstrecentchallenges,Pavesearguesthatthispredictionisburnedout,forknow-howandknowledgeareindeeddefeatedexactly whenone’sabilitytointentionallyactisdefeated.
MatthewMcGrath’sChapter9,UndercuttingDefeat:WhenitHappensand SomeImplicationsforEpistemology,investigatesrecentscepticismaboutthe Pollockianoperativecommonalitybetweenrebuttingandundercuttingdefeat. Inparticular,thechapterlooksintotheplausibilityoftheclaimthatthe mechanismofundercuttingdefeat,incontrasttothatofrebuttingdefeat, occursonlyinconjunctionwithcertainhigher-ordercontributions,i.e.with beliefsaboutthebasisonwhichonedoesorwouldbelieve.McGrathargues that,inthecaseofdefeatofinferentialjustification,undercuttingdefeatisa genuinephenomenonandtakesroughlytheshapePollocksuggests,not needinghelpfromhigher-orderbeliefsorjustifications.However,according
toMcGrath,fornoninferentialjustification,thePollockianaccountisin trouble.Thisdifference,itisargued,hasimportantimplicationsforepistemology:forone,whatseemstofollowisthatthereislessnoninferential perceptualortestimonialjustificationthaniscommonlythought.
Chapter10,DefeatersasIndicatorsofIgnorance,byClaytonLittlejohnand JulienDutantdevelopsanovelaccountofthenatureofrationalitydefeat. Accordingtotheirview,defeatersareindicatorsofignorance,evidencethat we ’renotinapositiontoknowsometargetproposition.Whentheevidence thatwe’renotinapositiontoknowissufficientlystrongandtheprobability thatwecanknowistoolow,itisnotrationaltobelieve.LittlejohnandDutant arguethattheiraccountretainsallthevirtuesofthemorefamiliarapproaches thatcharacterizedefeatintermsofitsconnectiontoreasonstobelieve ortoconfirmationbutprovidesabetterapproachtohigher-orderdefeat. Furthermore,itisargued,theviewprovidesaunifiednormativeframework, onethatgivesaunifyingexplanationofthetoxicityofdifferentdefeatersthat isgroundedinaframeworkthateitherrecognizesknowledgeasthenormof belieforidentifiesknowledgeasthefundamentalepistemicgoodthatfull beliefcanrealize.
InChapter11,CompetingReasons,JustinSnedegarconsidersdifferent waysthatreasons,bearingonouroptions,cancompetewithoneanotherto determinetheoverallnormativestatusofthoseoptions.Twokeyclaims defendedinthechapterare(i)thatthetheoryofthiscompetitionmustinclude adistinctroleforreasons against,inadditiontoreasons for,and(ii)thatthe theorymustallowfor comparative verdictsaboutwhichoptionsaremore stronglysupportedthanothers,ratherthansimplywhichoptionsarerequired orpermitted.Snedegarrejectsasimpleandfamiliarbalancingaccountofthe competition,aswellasanaccountthatunderstandsthecompetitioninterms ofgivingandansweringcriticismsoftheoptions,andheintroducesanew accountthatincorporatesadistinctroleforreasonsagainst.
InChapter12,PerceptualReasonsandDefeat,MarkSchroederfocuseson thedefeasibilityofperceptualevidence.Ifsomethinglooksredtoyou,itis reasonabletobelievethatitisred,butifyourealizeyouarewearingrosetintedglasses,itmaynotbereasonableatalltobelievethis,unlessyouhave someindependentsourceofevidence.Schroedercomparesseveralmodelsfor howtounderstandthisphenomenon.Thesemodelsdifferintheiranswersto twoquestions:whatevidencewegetabouttheexternalworldthroughperception,andwhatourhavingthatevidenceconsistsin.Accordingto Schroeder,somemodelshavetheadvantageof fittingseamlesslyintogeneral accountsofnon-monotonicinferencebutcarrywiththemacommitmenttoa
restrictedspaceofpossibleoptionsingeneralepistemology.Theirchallengeis toextractanadequatetreatmentofobjectivedefeatfromtheireleganttreatmentofsubjectivedefeat.Othermodelsleadtocommitmentsaboutthe differencesinwhatexplainsreasonablebeliefingoodandbadcases.Their challengeistoextractanadequatetreatmentofsubjectivedefeatfromtheir eleganttreatmentofobjectivedefeat.AccordingtoSchroeder,hisfavourite, non-factivecontentmodelsharesneitheroftheseproblems,anditoffers paralleleleganttreatmentsofbothobjectiveandsubjectivedefeat.
References
Alston,W.(1989). EpistemicJustification.Ithaca:CornellUniversityPress. Alston,W.(2002).Plantinga,naturalism,anddefeat.In:JamesBeilby(ed.) NaturalismDefeated?EssaysonPlantinga’sEvolutionaryArgumentagainst Naturalism. Ithaca:CornellUniversityPress:176–203.
Baker-Hytch,M.andBenton,M.(2015).Defeatismdefeated. Philosophical Perspectives,29(1):40–66.
Beddor,B.(2015).Processreliabilism’stroubleswithdefeat. ThePhilosophical Quarterly,65(259):145–59.
Bergmann,M.(2006). JustificationwithoutAwareness.NewYork:Oxford UniversityPress.
Brown,J.(2018). Fallibilism:EvidenceandKnowledge.Oxford:OxfordUniversity Press.
Chandler,J.(2013).Defeatreconsidered. Analysis,73(1):49–51.
Christensen,D.(2007).Epistemologyofdisagreement:Thegoodnews. PhilosophicalReview,116(2):187–217.
Christensen,D.(2010).Higher-orderevidence. PhilosophyandPhenomenological Research ,81(1):185–215.
Coates,A.(2012).Rationalepistemicakrasia. AmericanPhilosophicalQuarterly, 1(2):113–24.
Elga,A.(2007).Reflectionanddisagreement. Noûs,41(3):478–502. Feldman,R.(2005).Respectingtheevidence. PhilosophicalPerspectives,19: 95–119.
Frances,B.andMatheson,J.(2019).Disagreement.In:EdwardN.Zalta(ed.) The StanfordEncyclopediaofPhilosophy (Winter2019Edition).Stanford:Stanford UniversityPress.https://plato.stanford.edu/archives/win2019/entries/disagree ment/.
Goldberg,S.(2016).Ontheepistemicsignificanceofevidenceyoushouldhave had. Episteme ,13(4):449–70.
Goldman,A.(1979).Whatisjustifiedbelief?In:G.S.Pappas(ed.) Justification andKnowledge.Dordrecht:Reidel:1–25;reprintedinA.I.Goldman(2012). ReliabilismandContemporaryEpistemology.NewYork:OxfordUniversity Press:29–49.
Greco,D.(2017).Cognitivemobilehomes. Mind,126(501):93–121. Greco,J.(2010). KnowledgeasanAchievement.Cambridge:CambridgeUniversity Press.
Green,C.R.(2008). TheEpistemologyofTestimony. InternetEncyclopediaof Philosophy.http://www.iep.utm.edu/ep-testi/.
Harman,G.(1973). Thought.Princeton,NJ:PrincetonUniversityPress.
Hawthorne,J.andSrinivasan,A.(2013).Disagreementwithouttransparency: Somebleakthoughts.In:D.ChristensenandJ.Lackey(eds.) The EpistemologyofDisagreement:NewEssays .Oxford:OxfordUniversityPress: 9–30.
Horowitz,S.(2014).EpistemicAkrasia. Nous,48(4):718–44.
Kelly,T.(2005).Theepistemicsignificanceofdisagreement.In:J.Hawthorneand T.S.Gendler(eds.) OxfordStudiesinEpistemology. Vol.1.Oxford:Oxford UniversityPress:167–96.
Kelly,T.(2010).Peerdisagreementandhigher-orderevidence.In:R.Feldmanand T.A.Warfield(eds.) Disagreement. Oxford:OxfordUniversityPress:111–74.
Kornblith,H.(2015).Theroleofreasonsinepistemology. Episteme ,12(2): 225–39.
Lackey,J.(1999).Testimonialknowledgeandtransmission. ThePhilosophical Quarterly,49:471–90.
Lackey,J.(2008). LearningfromWords.Oxford:OxfordUniversityPress.
Lasonen-Aarnio,M.(2010).Unreasonableknowledge. PhilosophicalPerspectives, 24:1–21.
Lasonen-Aarnio,M.(2012).Higher-orderevidenceandthelimitsofdefeat. PhilosophyandPhenomenologicalResearch ,LXXXVIII:314–45.
Lasonen-Aarnio,M.(2014).Thedogmatismpuzzle. AustralasianJournalof Philosophy,92(3):417–32.
Lasonen-Aarnio,M.(2020).Enkrasiaorevidentialism. PhilosophicalStudies,177: 597–632.
Lyons,J.(2009). PerceptionandBasicBeliefs:Zombies,Modules,andtheProblem oftheExternalWorld.Oxford:OxfordUniversityPress.
Pappas,G.(2017).Internalistvs.externalistconceptionsofepistemicjustification. In:EdwardN.Zalta(ed.) TheStanfordEncyclopediaofPhilosophy (Fall2017 edn).https://plato.stanford.edu/archives/fall2017/entries/justep-intext/. Plantinga,A.(2000). WarrantedChristianBelief.NewYork:OxfordUniversity Press.
Pollock,J.(1986). ContemporaryTheoriesofKnowledge.Savage,MD:Rowman andLittlefield.
Pryor,J.(2004).What’swrongwithMoore’sargument. PhilosophicalIssues,14: 349–78.
Pryor,J.(2013).Problemsforcredulism.In:C.Tucker(ed.) Seemingsand Justification.Oxford:OxfordUniversityPress:89–131.
Simion,M.(2020). EpistemicNormsandEpistemicFunctions.UnpublishedBook Manuscript.
Sturgeon,S.(2014).Pollockondefeasiblereasons. PhilosophicalStudies,169(1): 105–18.
Sudduth,M.(2018). DefeatersinEpistemology.InternetEncyclopaediaof Philosophy.https://www.iep.utm.edu/ep-defea/#SH1b. Swinburne,R.(2001). EpistemicJustification.Oxford:ClarendonPress.
Titelbaum,M.(2015).Rationality’ s fixedpoint(or:indefenceofrightreason).In: S.T.GendlerandJ.Hawthorne(eds.), OxfordStudiesinEpistemology. Vol.5. Oxford:OxfordUniversityPress:253–94.
Weatherson,B.(2019). NormativeExternalism.Oxford:OxfordUniversityPress. Wedgwood,R.(2011).Justifiedinference. Synthese,189(2):1–23. Weisberg,J.(2009).Commutativityorholism?Adilemmaforconditionalizers. BritishJournalforthePhilosophyofScience,60:793–812. Williamson,T.(2009).Replies.In:P.GreenoughandD.Pritchard(eds.), WilliamsononKnowledge.Oxford:OxfordUniversityPress:279–384.
