Tomyparents
Preface
Philosophyishard.Ethicsishard;epistemologyishard;decisiontheoryishard;logic ishard.Allthepartsofphilosophyarehard,butthosefouraregoingtobe particularlyrelevanttothestoryI’mtellinghere.Theymatterbecausetheyareall evaluative.Someonewhoviolatesethicalprinciplesisimmoral;someonewhoviolatesepistemologicalprinciplesisirrational;someonewhoviolatestheprinciplesof decisiontheoryisimprudent;someonewhoviolateslogicalprinciplesisillogical. Andtosaythatsomeoneisimmoral,irrational,imprudent,orillogicalistonegativelyevaluatethem.
Butitiseasytofeeluneasywiththissetoffacts.Ifitissohardto figureoutthe truthinthese fields,whyshouldwenegativelyevaluatesomeoneforfailingto conformtothesehardto findstandards?Doesn ’tfairnessrequirethatweonly judgepeoplebystandardstheycanknowabout?I’mgoingtoarguethisisnot right thattoevaluatesomeoneisnecessarilytoimposeastandardonthem,and theymaynotevenknowwhatthestandardis.Indeed,theymaynothaveanyreason tobelievethetruthaboutwhatthestandardis,andinextremecasesmayhavegood reasontoendorseafalsestandard.
Thispositionisuncomfortable,sinceitiseasytofeeltheunfairnessofholding someonetoastandardthattheydonotaccept,andcouldnotreasonablyaccept. Manyphilosophersthinkthatweshouldeithersupplementorreplacetheseexternal standardswithinternalstandards.An ‘internalstandard’ hereisonethattheperson beingevaluatedeitheraccepts,orhasgoodreasontoaccept.Tosupplementthe externalstandardsistosaythattherearetwowaystoevaluatepeople.Itisgoodto liveuptothecorrectstandardsinethics,epistemology,anddecisiontheory,and badtoviolatethem.Butitisalso,saythesupplementers,goodtoliveuptoone’ sown standards,andbadtoviolatethem.Thereplacerssaythatconformitytoone’ s ownstandardsismoreimportantthanconformitytoexternalstandards;insome deepsense(atleastsomeof)theheroesofethics,epistemology,anddecisiontheory arepeoplewhoabidebytheirownstandards.
Iamgoingtopresstwoproblemsagainstthiskindofview.Theproblemsaremost pressingforthereplacers,buttheyunderminethepositionofthesupplementerstoo. The firstproblemisthatthiskindofviewhasproblemswithfanaticsandideologues.Everyideologuewhothoughtthattheyhad figuredouttheonetruewaythings mustbedoneandreactedviolentlyagainstthosewhodidn’tagreeweredoingwellby theirownlights.It’snotgood,inanyway,tobethatkindofideologue.Weshouldn’t lookbackattheReignofTerrorandsay, “Well,atleastRobespierreandSaint-Just werelivinginaccordancewiththeirownvalues.” Aimingto fittheworldtoone’ sown valuesisadangerousgame;it’sonlyworthplayingifyou’vegotthevaluesright.When wefocusourattentiononideologueswhohavegoneofftherails,theideathatitis unfairtoholdpeopletoastandardtheycan’tseefeelslikesomethingthat’saproblem intheorybutnotinpractice.
Thesecondproblemwiththeinternalviewisthatitleadstoanastyregress.Itis,to besure,hardtotellwhatthetruevaluesare.Butchoosingsomevaluesdoesnotend ourproblems.Moralityishardevenonceyou’vesettledonamoraltheory.Thisisa pointfamiliarfrom,forexample,Sartre’sdiscussionoftheyoungmantornbetween dutytohismotherandhiscountry.
Whatcouldhelphimmakethatchoice?TheChristiandoctrine?No.TheChristiandoctrine tellsuswemustbecharitable,loveourneighbour,sacrificeourselvesforothers,choosethe “ narrowway, ” etcetera.Butwhatisthenarrowway?Whomshouldwelovelikeabrother the soliderorthemother?...Whocandecidethat apriori?Noone.Nocodeofethicsonrecord answersthatquestion.(Sartre2007,31)
Wecanevaluatetheyoungmanbyhisownlightsandstillbeinawayunfairtohim. PerhapsitturnsoutthatthetrulyChristianthingtodoisto fightNazis,butthe youngmanconcludes(reasonablybutfalsely)thatitistohelphismother.Andhe doesthat.Ifwearemovedbytheunfairnessofholdinghimtoastandardhedoesnot endorse,weshouldalso finditunfairtoholdhimtoaconsequenceofhisown standardthathedoesn’trecognize.Butnowwhatisleftoftheinternalstandard?It mustbethatitisgoodtodonotwhatisbestbyone’sownlights,butwhatonethinks isbestbyone’sownlights.Butperhapsonecouldevenbewrongabout that. (I’lldiscussanexampleofthisinchapter1.)Andtheinternalviewcollapsesinto theviewthatweshouldevaluatepeoplebywhattheythinktheythinktheythink...their ownviewssupport.
Thisisallabsurd,anditmakestheproblemwithfanaticsandideologueseven worse.Perhapswecouldarguethatsomeideologuestakeactionsthatareincompatiblewithwhattheysaytheirvaluesare.Buttheydonotactagainstwhattheythink theirownvaluesrequire.
Perhapswecanmotivatetheimportanceoftheinternalpointofviewnotby thinkingaboutfairness,butbyfocusingonananalogywithrecklessagents.IfI firea cannondownFifthAvenueatpeakhour,Idosomethingmorallyhorribleevenif miraculouslyIdon’thitanyone.Myactioniswrongbecauseitisreckless.Perhapsif Idosomethingthatisprobablymorallywrong,Iammorallyrecklessinjustthesame way.Andthat’strueevenifmyactionturnsoutnottobewrong.Sowhatmattersis notjustwhatisrightandwrong,butprobabilitiesofrightnessandwrongness.Ithink thiskindofreasoningfailstoo,andthereareimportantasymmetriesbetween physicalrisk(asisinvolvedin firingcannonsdownbusystreets)andmoralrisk. I’llspendchapters3and4outliningtheseasymmetries,andwhytheytellagainstthe ideathatthereisadistinctivewrongofmoralrecklessness.
The firsthalfofthebookdiscussesthesignificanceoftheinternalpointofviewin ethics.AsI’veindicated,Idon’tthinktheinternalpointofviewisparticularly important,thoughI’llspendabitoftimetowardstheendofPartIlookingat somemorelimited,andhencemoreplausible,claimsforitsusefulness.Thesecond partofthebookturnstoepistemology,andtotheideathatonecannotreasonably havebeliefsthatonebelieves(orshouldbelieve)tobeunreasonable.
Again,theissueturnsonhowimportantisconformitytoone’sownstandards.The mostcommonphilosophicalviewaroundhereisakindofsupplementingview,nota replacingview.Itisimportant,sayseveralphilosophers,tohavebeliefsthatareboth
actuallyreasonableandalsoreasonablebyone’sownlights.AndI’mgoingtopush backagainstthat.OnereasoncomesfromworkbyTimothyWilliamson.What’ s reasonabletobelieveturnsonempiricalfactsaboutone’ssituation.Sincewedon’t haveGod-likeperfectaccesstoourownempiricalsituation,wemightnotrealize whatisreasonabletodoinourownsituationjustbecausewedon’tknowprecisely whatsituationwearein.Insuchcases,itseemsweshouldreacttothesituationwe areactuallyin,nottoourbestguessaboutwhatsituationthatis.
TherewillbetwoprimarythemesofPartIIofthebook.Oneechoesthe firstpart ofthebook.Sometimeswecannotknowwhatitwouldbetobereasonableby ourownlights.Soaddingarequirementthatreasonablepeoplearedoingwell bytheirownlightsthreatenstotriggeraviciousregress.I’mgoingtoarguethat thisthreatisrealized.Theotherthemeisthatthephenomenathatphilosophershave thoughtcouldonlybeexplainedbyaddinganinternalconstraintontobeliefcanbe adequatelyexplainedbyamorecarefulattentiontothenatureofevidence,andwhat ittakesforonetohaveevidenceandforthatevidencetosupportabelief.I’llargue thatsuchexplanationsarepreferabletoexplanationsintermsofinternalconstraints (suchasonlybelievewhatyoubelieveisreasonabletobelieve).Thisisinpartbecause theyavoidregressandimplausibleknowledgeaboutone’sownsituation;inpart becausetheyonlycommitustothingsweareindependentlycommittedto;andin partbecausetheyexplainamuchbroaderrangeofcasesthanareexplainedbythe allegedinternalconstraints.
IhavemorepeopletothankforhelpwiththisbookthanIcouldpossiblylisthere. I’mnotevensureatwhichpointoftimeIshouldstartthethanks.Twenty-oddyears agoasagraduatestudentatMonashIwasn’tworkingon this project.Butthepicture thatpervadesthisbook,thatinphilosophyeverythingiscontestableandthereareno safestoppingpoints,owesalottotheamountoftimeIspentasagraduatestudent thinkingabout,andbeingtaughtabout,heterodoxapproachestologicandto decisiontheory.
MostofthebestfeedbackI’vereceivedonthevariouspartsofthebookhascome fromgraduatestudents.Someofthesecondpartofthebookisbasedonan epistemologyseminarItaughtatRutgers.ItaughtagraduateseminaratMichigan offanearlydraftofthebookmanuscript.AndI’vetaughtseveralmini-coursesat StAndrews,andpresentedatevenmoreworkshopsandsymposiathere,offpartsof thebook.IneverycasethefeedbackIreceivedfromcolleaguesand,evenmore frequently,graduatestudents,changedthebookforthebetter.
PartsofthebookarebasedonpresentationsatororganizedbytheUniversityof Aberdeen,UniversityofOxford,UniversityofVienna,UniversityofKonstanz, UniversityofZurich,UniversityofGraz,MassachusettsInstituteofTechnology, PrincetonUniversity,OhioStateUniversity,UniversityofSydney,Australian NationalUniversity,andUniversityofMelbourne.I’vepresentedpartsofitatthe BellinghamSummerPhilosophyConference,theNightofPhilosophyinNewYork City,andtheAustralasianAssociationofPhilosophyannualconference.AndI’ ve discusseditwiththeCorridorReadingGroupinNewYork,andtheEthicsLunch groupinAnnArbor.I’mverygratefulforallthefeedbackIgotatthosepresentations. Aswellasallthoseaudiences,I’dliketoparticularlythankDerekBall,Jessica Brown,SarahBuss,HermanCappelen,RuthChang,StewartCohen,JoshDever,
TomDonaldson,AndyEgan,ClaireField,KatherineHawley,ScottHershowitz, TorfinnHuvenes,JonathanJenkinsIchikawa,JimJoyce,ZoeJohnsonKing,Maria Lasonen-Aarnio,BenLevinstein,JuliaMarkovits,MatthewMcGrath,SarahMoss,Jill North,CarolinePerry,QuentinPharr,LewisRoss,AndrewSepielli,JoeShin,Holly Smith,MartinSmith,andEliaZardiniforparticularlyvaluablefeedback.(AndI’ m alreadydreading findingoutwhoIshouldhaveincludedonthislistbutdidn’t.)
RalphWedgwoodreadthewholemanuscriptandprovidedcommentsthatimproved itininnumerableways.Thankstohim,andtoPeterMomtchiloffformakingsuchan astutechoiceofreaderforthemanuscript.
InPartIofthebook,especiallysections2.4,2.6,3.4,3.9and6.1,Idrawonmaterial frommypaper, “RunningRisksMorally”.ItisreprintedbypermissionfromSpringer Nature, PhilosophicalStudies,Volume167,Issue1,pp.141–63,doi:10.1007/s11098013-0227-2,Copyright©2013,SpringerScience+BusinessMediaDordrecht.Many oftheargumentsinchapter11,andinsection12.1, firstappearedinmy “Disagreements,PhilosophicalandOtherwise,” in TheEpistemologyofDisagreement:New Essays,editedbyDavidChristensenandJenniferLackey,Copyright©2013,and isreproducedbypermissionofOxfordUniversityPress:https://global.oup.com/ academic/product/the-epistemology-of-disagreement-9780199698370.
EliseWoodardcompiledtheindexforthebook.Andshecorrectedmanyerrorsin thebookinthecourseofjustdoingtheindex.ThanksalsoOxfordUniversityPress forcapturingmanymore.I’msureseveralremain,andthoseareallmyfault.
TheidiosyncraticworkflowIusedforwritingthiswouldhavebeenimpossible withoutFletcherPenney’sMultimarkdown(boththelanguageandtheComposer software)andJohnMacFarlane’sPandoc,andI’mverygratefultobothofthemfor buildingsuchvaluabletools.MuchofthebookwasdraftedunderthedomeintheLa TrobeReadingRoomattheStateLibraryofVictoria,andI’msogratefulthatVictoria hasmaintainedthatspace,andthatbuilding.
Earlyinthedevelopmentofthisbookproject,Iwashonouredtobecomethe first MarshallM.WeinbergProfessorofPhilosophyattheUniversityofMichigan,Ann Arbor.WithoutthesupportMarshallhasprovidedtomyresearch,andtothe researchprojectoftheUniversityofMichiganmorebroadly,thisprojectwould havebeenunimaginable.Myinaugurallecturewas “RunningRisksMorally,” mostof whichappearsinonewayoranotherinpartIofthebook.The firstdraftofthebook waswrittenwhileonasabbaticalfundedthroughtheWeinbergProfessorship.But beyondthat,thevibrantintellectualcommunityhereatMichiganreliesineverso manywaysonMarshall’ssupport.Icouldn’ttellyouhowmuchthisbookrelieson feedbackfromgraduatestudentswhohavereceivedWeinbergfellowships,orwho cametoMichiganinpartbecauseoftheWeinbergCenterforCognitiveScience. Whilethisisbynomeansaworkofcognitivescience,itisinfluencedinmanyways bywhatI’velearnedfromcognitivescientiststalkingattheWeinbergCenter.And IreallycannotthankMarshallenoughforhissupportforMichigan,andforits research.
Finally,I’dliketothankIshaniMaitraandNyayaMaitraWeathersonfor,well, everything.Ishanididn’tjusttalkthroughallthethingsinthisbookwithme,and improveditinsomanyways,butshealsotalkedthroughallthethingsIcutfromthe book.Andsheimprovedthoseportionstoo.
1.Introduction1
1.1ToThineOwnSelfBeTrue1
1.2FourQuestions2
1.3NormativeExternalismDefined8 1.4Guidance9
1.5Symmetry10
1.6Regress13
1.7TwoRecentDebates17
1.8ElizabethandDescartes18
1.9WhyCallThisExternalism?20
1.10PlanofBook22
PartI.Ethics
2.AllAboutInternalism27
2.1SomeDistinctions27
2.2TwoWaysofMaximizingExpectedGoodness31
2.3VarietiesofInternalism32
2.4AnInitialConstraint34
2.5MotivationOne:Guidance36
2.6MotivationTwo:Recklessness37
2.7MotivationThree:Symmetry39
3.AgainstSymmetry41
3.1GuiltandShame41
3.2JacksonCases42
3.3Motivation45
3.4WelfareandMotivation48
3.5Motivation,Virtues,andVices50
3.6TheWeakMotivationPrinciple(WMP)52
3.7TheStrongMotivationPrinciple(SMP)57
3.8MotivationThroughThickandThin61
3.9Moller’sExample65
4.ADilemmaforInternalism68
4.1SixFormsofInternalism68
4.2TwoDifficultCases70
4.3InadvertentVirtueandMisguidedConscience72
4.4EthicsandEpistemology74
4.5RationalityandSymmetry78
4.6Conclusion80
5.BlameandMoralIgnorance84
5.1DoesMoralIgnoranceExcuse?84
5.2WhyBelieveMIE?86
5.3ChapterPlan87
5.4BlameandDesire88
5.5Blame,Agents,andTime90
5.6ActinginIgnoranceIsNoExcuse91
5.7AgainstCounterfactualInterpretationsofActingfromIgnorance92
5.8AgainstMotivationalInterpretationsofActingfromIgnorance94
5.9AdoptingaDecisionProcedureandActingonIt97
5.10CalhounonBlameandBlameworthiness98
5.11MoralMistakesandMoralStrangers103
5.12TwoApproachestoBlame106
6.DoubleStandards109
6.1Hypocrites109
6.2ValueComparisons112
6.3TheExternalist’sCommitments114
PartII.Epistemology
7.Level-CrossingPrinciples119
7.1First-OrderandSecond-OrderEpistemology119
7.2ChangeEvidentialism120
7.3MotivationsforLevel-Crossing121
7.4ThePlanfortheRestoftheBook123
7.5Evidence,Rationality,andWisdom124
7.6Evidence,Thought,andMathematics125
8.Higher-OrderEvidence130
8.1VarietiesofHigher-OrderExamples130
8.2DiagnosesandAlternatives133
8.3TirednessandAbduction135
8.4ExplainingAllFourCases138
8.5AgainstBracketing143
9.Circles,EpistemicandBenign146
9.1NormativeExternalismandCircularity146
9.2Inference,Implication,andTransmission148
9.3Liberalism,Defeaters,andCircles150
9.4PyrrhonianSkepticismandNormativeExternalism156
9.5EasyKnowledge157
9.6What’sWrongwithEasyKnowledge?160
9.7Coda:Testing165
Introduction
1.1ToThineOwnSelfBeTrue
Earlyin Hamlet,LaertesdepartsElsinoreforParis.Ashepreparestogohisfather, LordPolonius,offershimsomepaternaladvice.Hetellshimtotalklessandsmile more.Hetellshimtospendallhismoneyonclothes,sincethat’showtheyrollin Paris.Hetellshimtoneitheraborrowernoralenderbe,thoughthelatteris presumablyredundantifhe’stakentheadvicetodate.Andheconcludeswiththis advice,destinedtoadornhighschoolyearbooksforcenturiestocome.
Thisaboveall:tothineownselfbetrue,Anditmustfollow,asthenighttheday,Thoucanst notthenbefalsetoanyman.
Itisn’tcompletelyclearwhatPoloniusmeanswhenheadvisesLaertestobetrueto himself,butitisplausiblethathemeanssomethinglikethis:
Followyourownprinciples!
Orperhapssomethinglikethis:
Dowhatyouthinkisright!
AndunliketherestoftheadvicePoloniusgives,manyphilosophershavefollowed himinthinkingthisisaverygoodidea.
Theprimaryaimofthisbookistoargueagainstthisidea.Followingone’ sown principles,ordoingwhatonethinksisright,arenotingeneralverygoodideasatall. Iwillcall normativeinternalism theviewthatweshouldbeguidedbynormsthatare internaltoourownminds,inthesensethatourbeliefs,andour(normativeevidence) isinternaltoourminds.AndIwillopposethatview,arguingfor normative externalism.
Normativeexternalismistheviewthatthemostimportantstandardsforevaluatingactions,mentalstates,andagentsaretypicallyexternaltotheactor,believer,or agentbeingevaluated.Itcanbeappropriatetoholdsomeonetoamoral,oran epistemic,standardthattheydonotendorse,oreventhattheycouldnotbe reasonablyexpectedtoendorse.Ifonehasbadstandards,thereneedbenothing wronginviolatingthem,andthereisnothinggoodaboutupholdingthem.
Thatlastparagraphmadealotofdistinctclaims,anditisworthspendingsome timeteasingthemapart.Butbeforewegettoodeepintheweeds,Iwanttohaveon thetabletheguidingprincipleofthebook.Beingtruetoyourself,inthesenseof conformingtotheprinciplesonehas,oreventotheprinciplesonehasreasonto
have,isjustnotthatimportant.Whatisimportantisdoingtherightthing,beinga goodperson,andhavingrationalbeliefs.Ifonehasmisguidedviewsabouttheright, thegood,andtherational,thenthereisnothinggoodaboutconformingtothose misguidedviews.Andthismatters,becausemanypeoplehaveviewsabouttheright, thegood,andtherational,thatareverymisguidedindeed.
1.2FourQuestions
1.2.1Actions,agents,oradvice
Ifonesays,withPolonius,thatitisgoodtoconformtoone’sownprinciples,thereare anumberofdistinctthingsonecouldbemeaning.
Onecouldbemakingaclaimaboutparticular actions.(Oraboutparticularbeliefs, butwe’llfocusonactionsforthenextfewparagraphs.)Soonecouldbesayingthat actionsthatconformtotheactor’sprinciplesaretherebyinsomesenserightorgood, andthosethatviolatetheactor’sprinciplesareinsomesensewrongorbad.
Alternatively,onecouldbemakingaclaimabout agents.Soonecouldbesaying thatpeoplewho(typically)conformtheiractionstotheirprinciplesareinsomesense good(orlessbad)people,andthosewhoviolatetheirownprinciplesareinsome sensebad.
Oralternativelyagain,onecouldbemakingaclaimabout advice.Onecouldbe sayingthatwhetherornottheclaimsintheprevioustwoparagraphsarestrictly correct,itisexcellenttoadvisepeopletoactaccordingtotheirprinciples.Thereare plentyofcaseswhereadvisingpeopletodotheoptimalthingisbad,especiallyif aimingfortheoptimalresultislikelytoleadtocatastrophe.Sothisviewaboutadvice isinprincipledistinctfromtheviewsaboutactionsandagents.
TheformofexternalismIwilldefendisopposedtotheviewsinallofthelastthree paragraphs.Butitismoststronglyopposedtotheviewaboutactions,andleast stronglyopposedtotheviewaboutadvice.Indeed,Iwon’thavealottosayabout advicethroughoutthebook;excepttonoteoccasionallywhenintuitionsaboutadvice seemtobegettingusedillegitimatelytojustifyconclusionsaboutactions.ButIdon’t meantoimplythattheviewshavetostandorfalltogether.Aviewthatisexternalist aboutactions itthinksitdoesn’tmakeanydifferencetothecorrectevaluationofan actionwhethertheactorendorseditornot butinternalistaboutagents itthinks thereissomethinggoodaboutpeoplewhosticktotheirprinciplesandbadabout thosewhodonot iscertainlyworthconsidering.Butitisn’tmyview;Imeanto opposeallthreeprecisificationsofwhatPoloniussays.
1.2.2Aboveall?
PoloniusdoesnotjustsayLaertesshouldbetruetohimself.Hesaysthisissomething ‘aboveall.’ Thissuggeststhatheiselevating Dowhatyouthinkisright toacentral place,makingitmoreimportantthanprincipleslike Respectotherpeople,or Make theworldbetter,oreven Dotherightthing
TheexternalistviewIsupporttakescompletelytheoppositetack.Theprinciple Do whatyouthinkisright isofnoimportanceatall.
Butthereisalargemiddlegroundposition.Thisiseasiesttoseeifweassumethe debateisaboutagents,notactionsoradvice,soI’llpresentitforagents.Butit shouldn’tbetoohardtoseehowtogeneralizetheidea.
Wecouldholdthatdoingwhatonethinksisrightisoneofthevirtues,something thatcontributestoapersonbeingagoodperson.Orwemightthinkthatfailingtodo whatonethinksisrightisavice,somethingthatcontributestoapersonbeingabad person.Andwemightthinkoneorother(orboth)ofthosethingswithoutthinking themparticularlyimportantvirtuesorvices.Onecouldcoherentlyholdthatthereis avirtueinholdingtoone’sprinciples,evenifonethinksthatothervirtuestodowith honesty,courage,respect,andthelikearemoreimportant.Andonecouldcoherently holdthatdoingwhatonethinksiswrongisavice,eveninthecasewhereonehas falseenoughviewsabout first-ordermoralquestionsthatdoingwhatonethinksit rightwouldmanifestevenmoreseriousvices.
Indeed,onemightthinkthatordinaryEnglishgoesalongwiththis.Wedotalk somewhatadmiringlyaboutpeoplewhoareprincipledorresolute,andsomewhat disdainfullyaboutpeoplewhoarehypocritical.¹
I’mgoingtoclassifythiskindofview,theonethatsaysthatdoingwhatonethinks isrightisimportanttocharacter,butnotofmaximalimportance,asamoderate internalistview.Andmyexternalismwillbeopposedtoit,likeitisopposedtothe viewthatbeingprincipled,andavoidinghypocrisy,arethemostimportantvirtues. Thepossibilityofsuchamoderateinternalistviewisimportant,becauseotherwise wemightthinktheargumentagainstinternalismwouldbetooeasy.Historyisfullof fanaticswhoconvincedthemselvesthattheyweredoingtherightthingwhilecausing immenseharm.Itishardtobelievethattheoneprincipletheydidconformto, Follow yourownprinciples,isthemostimportantprincipleofall.Butperhaps,justperhaps, theirresolutenessisinasmallwayavirtue.Atleast,aphilosophicalviewthatsays thatitisavirtue,albeitoneoffsetbymountainsofvice,isnotabsurd.
1.2.3Ethics,epistemology,andmore
I’vebeeninterpretingPolonius’sdictumasbeingprimarilyaboutethicssofar.But viewslikehisareavailableinmanyotherareasofphilosophy.I’llmentionthreemore here,the firstofwhichwillbeamajorfocusofthisbook.
Beliefissubjecttoevaluationonanumberoffronts.Beliefsaretrueorfalse,but thathardlyexhauststheirvirtuesorvices.Sometruebeliefsarebadinvirtueofbeing luckyguesses,orleapstounwarrantedconclusions.Somefalsebeliefsaretheresultof sensiblyfollowingtheevidencewhereitleads,andjustbeingunluckilymisledinto error.Soaswellasevaluatingabelieffortruth,wecanevaluateitforresponsiveness totheevidence.I’mgoingtoargue,somewhatindirectly,thatabeliefisrationaljust incaseitisresponsivetotheevidenceinthisway.²
¹Thoughtobeclear,Idon’tthinktheEnglishwords ‘principled’ and ‘resolute’ actuallypickoutthe so-calledvirtueofupholdingone’sownprinciples.FollowingRichardHolton(1999),Ithinkthosewords pickoutdiachronicpropertiesofaperson.Theyapplytoapersoninpartduetothatperson’sconstancy overtimeinsomerespect.Followingone’sprinciplesisn’tlikethis;itisapurelysynchronicaffair.
²Thoughgettingclearonjustwhatthislastsentencecommitsmetowillrequiresayingmoreabout whatevidenceis.Fornow,itwon’tdomuchharmtoequateevidencewithbasicknowledge.Aproposition
Butifthat’swhatrationalityis,thensubjectscanalsohavebeliefsaboutthe rationalityoftheirownbeliefs.Andwecanaskwhethersubjectsaredoingwellat believingbytheirownlights.Tobelievesomethingjustistobelieveitistrue,soifour onlystandardforbeliefistruth,theneveryonewillbelievewellbytheirownlights. Butitispossibletobelievesomething,andevenrationallybelieveit,whilebelieving thatthatverybeliefisirrational.Or,atleast,soI’llargue.
Isthisabadthing?Shouldwemarksomeonedownforbelievinginawaythatthey taketobeirrational?I’mgoingtoarguethatweshouldnot.It’sgoodtobelievetruths. It’sgoodtobelieveinaccordwithone’sevidence.Andthat’sasfarasweshouldgo. It’snotgoodtobelieveinaccordwithwhatonebelievestheevidencesupports,unless onetherebyendsupwithabeliefthatisgoodforsomeotherreason.Andit’snotbad tobelievesomethingthatonebelievesisnotsupportedbyone’sevidence,unlessone endsupwithabeliefthatisbadforsomeotherreason.
Justasintheethicscase,wecanseparateoutanumberofdistinctquestionshere. Assumeyouthinkthereissomethingphilosophicallyimportantaboutbeliefsthatare irrationalbythelightsofthebelieverthemselves.Youcouldsaythatthisisabadmakingfeatureofthebeliefitself,orabad-makingfeatureofthebeliever,or,perhaps thatitisbadtoadvisepeopletohavebeliefsthatareirrationalbytheirownlights.That is,wecanreplicatetheact,agent,oradvicedistinctioninsideepistemology,thoughthe ‘acts’ arereallythestatesofholdingparticularbeliefs.Andifyoudothinkthesebeliefs, orbelievers,arebadinsomeway,thereisafurtherquestionabouthowmuchbadness isinvolved.Isbelievinginawaythatonethinksisirrationalasbadasnotfollowingthe (first-order)evidence,ormorebad,orlessbad?(Orisbadnessthewrongconcepttobe usinghere?)
Wewillseedifferentphilosophicalviewsthattakedifferentstandsonthese questionsthroughoutPartIIofthebook.I’mgoingtodefendafairlysimple,and fairlyextreme,position.Itisn’tabad-makingfeature,inanyway,ofabeliefthatthe believerthinksitisirrational,norisitabad-makingfeatureofbelieversthatthey havebeliefstheythinkareirrational.Itisn’tevenabadhabittoroutinelyhavebeliefs thatonethinksareirrational;thoughI’mgoingtobealittlemoretentativein defendingthatlastconclusion.Thegeneralprinciplethroughoutistomotivateand defendapicturewherewhatmattersisconformitytotheactualrules betheyrules ofactionorrulesofbelief ratherthanconformitytowhatonetakes(oreven rationallytakes)therulestobe.
Thedisputesofthelastfewparagraphshaveallbeenoverepistemology,fairly narrowlyconstrued.Buttherearesomeotherdisputesthatwemighthavetoo,where thedifferencebetweenconformitytoexternalrulesandconformitytoone’sversion oftherulesmatters.I’mnotgoingtosaymuchaboutthenexttwodisputes,butthey arehelpfultohaveonthetable.
Somelivesgobetterthanothers.Whenweactforthesakeofothers,whenweact benevolently,weaimtoimprovethelivesofothers.Callsomeone’ s welfare that p ispartofthesubject’sevidenceifthesubjectknows p,anddoesn’tknow p becausesheinferreditfrom somethingelse.
quantityweimprovewhenweactbenevolently.³Philosophersdisagreealotabout whatwelfareis,sosomeofthemarewrong.AndthoughI’mnotgoingtoarguefor this,itseemstomethatthedisagreeingpartieseachhavesuchgoodargumentsthat atleastsomeofthephilosopherswhoarewrongareneverthelessrationalinholding thepositiontheydo.Sothatimpliesthatarationalpersoncouldhaveachoice betweentwoactions,oneofwhichactuallyproducesmorewelfare,andtheotherof whichproducesmorewelfareaccordingtothetheoryofwelfarethey(rationally) hold.Assumingthepersonwantstoactbenevolently,or,iftheactisdirectedtotheir owngood,theywanttoactprudentially,isthereanythinggoodaboutdoingthething thatproducesmorewelfareaccordingtothetheoryofwelfaretheyhold?My position,thoughI’mnotgoingtoargueforthisinthisbook,isthatthereisnot. Whatmattersforbenevolentorprudentialactionishowwellone’sactdoesaccordingtothecorrecttheoryofwelfare.Itdoesn’tmakeanactionbenevolent,orprudent, iftheactionisgoodaccordingtoamistakentheoryofwelfare.That’strueevenifthe theoryofwelfareisone’sown,orevenifitistheonethatisrationalforonetohold.If one ’stheoryofwelfareisapurelyhedonisticexperientialtheoryofwelfare,thenyou mightthinkyouareimprovingthewelfareofothersbyforce-feedingthemhappy pills.Butifthattheoryofwelfareisfalse,andwelfareinvolvespreferencesatisfaction, orautonomy,thensuchanactionwillnotbebenevolent,norwillitberationalto performonbenevolentgrounds.
Wecanmakethesamekindofdistinctionwithindecisiontheory.Let’sassumefor nowthatapersonhasrationalbeliefs,andwhentheylackbelieftheyassignarational probabilitytoeachuncertainoutcome,andtheyvaluetherightthings.Thereisstilla questionabouthowtheyshouldactinthefaceofuncertainty.Unlikethequestions aboutethics,epistemology,orwelfare,thereisanorthodoxanswerhere.Theyshould maximizeexpectedutility.Thatis,foreachact,theyshouldmultiplytheprobability ofeachoutcomegiventhatact,bythe(presumablynumerical)valueofthat outcome–actpair,andadduptheresultingproductstogetanexpectedvalueof theact.Thentheyshouldchoosetheactwiththehighestexpectedvalue.Butwhile thisistheorthodoxviewofdecisiontheory,therearedissentersfromit.⁴ Thebest recentstatementofdissentisinabook-lengthtreatmentbyLaraBuchak(2013).And someonewhohasreadBuchak’sbookcanthinkthatherviewistrue,or,perhaps, thinkthatthereissomeprobabilitythatitistrueandsomeprobabilitythatthe orthodoxyistrue.
Sonowwecanaskthesamekindofquestionaboutconformitytothecorrectrules versusconformitytotherulesonethinksarecorrect.⁵ Assumethatsomeonedoesnot havethecorrectbeliefsabouthowtorationallymakedecisions.Andassumethatthey performanactwhichisnotrational,accordingtothetruedecisiontheory,butis
³Therearealotofdifferentthingsthatpeoplecallwelfareinthephilosophicalliterature.I’mtakingthe ideaoftyingitdefinitionallytobenevolentactionfromSimonKeller(2009).
⁴ I’msuppressingdisputeswithinorthodoxyabouthowjusttoformulatetheview,thoughthose disputeswouldalsosufficetogetthekindofexampleIwantgoing.
⁵ Ifthemoraltheoriesonegivescredencetorejectexpectedvaluemaximization,thentherewillbeeven morecomplicationsattheintersectionofethicsanddecisiontheory.IttayNissan-Rozen(2015)hasareally nicecaseshowingthecomplicationsthatarisefortheinternalistwhenmoraltheoriesdonotassume orthodoxdecisiontheory.
rationalaccordingtothedecisiontheorytheyaccept.Istheresomethinggoodabout thatdecision,andwouldtherehavebeensomethingbadaboutthemdoingthething thatcorrecttheoryrecommended?Mypositionisthatthereisnot.Therational decisionsaretheonesrecommendedbycorrectdecisiontheory.Thereisnothingto besaidforconformingtoone’sownpreferreddecisiontheory,ifthattheoryisfalse.
1.2.4Actualorrational
SofarI’vefocusedonthedistinctionbetweenprinciplesthatareexternaltotheagent, andprinciplesthatareinternaltotheagentinthesenseofbeingbelievedbythe agent,orbeingtheagent’sownprinciples.WhenIcallmyviewexternalist,itisto indicatethatIthinkitistheexternalprinciplesthatmatter.Butthereisanother categoryofprinciplesthatIhaven’tfocusedon,andwhichareinsomesenseinternal. Thesearetheprinciplesthattheagentshould,rationally,accept.
Nowifwesaythattheagentshouldrationallyacceptallandonlythetrue principles,thentherewon’tbeadistinctionbetween Followthetrueprinciples and Followtheprinciplesitisrationaltoaccept.Butlet’sworkfornowwiththeassumptionthatthereisadifferencehere;thatjustlikewithanythingelse,agentscanbe rationallymisledaboutthenatureofethics,epistemology,welfare,anddecision theory. ⁶ Thenthereisanotherpossibility;thatagentsshouldfollowtheprinciples thattheyhavemostreasontobelievearetrue.
Thisgivesanotherwayfortheinternalisttorespondtotheproblemofhistorical monsters.Let’sthinkaboutoneparticularcase,onethatI’llreturntooccasionallyin thebook:MaximilienRobespierre.⁷ Whateverelseonecansayabouthim,noone seriouslydoubtsthatRobespierrealwaysdidwhathethoughtwasright.⁸ Butdoing whathethoughtwasrightinvolvedsettingofftheReignofTerror,andexecutingever somanypeopleonincredibly flimsypretexts.Wecan’treallysaythattheprinciplehe didwellby, Dowhatyouthinkisright,isonethatshouldbevaluedaboveall.We mentionedabovethatwecouldreasonablysayitisagood-makingfeatureof Robespierrethathewasprincipled,evenifitisoutweighedbyhowabhorrenthis
⁶ JuliaMarkovits(2014)arguesthatagentshaverationalreasontoacceptthefundamentalmoraltruths. MichaelTitelbaum(2015)arguesthatagentshaverationalreasontoacceptthefundamentalepistemologicaltruths.I’massumingfornowthatbothofthesepositionsarefalse,becauseitgivesmyopponents moreroomtomoveiftheyarefalse.ClaireField(forthcoming)respondstoTitelbaum’sarguments.Note herethatwhenIsaythatanagentcanberationallymisledaboutmoralityandepistemology,Iamnot claimingthattheycanrationallyhaveoutrightfalsebeliefsaboutmoralityandepistemology.Ijustmean thatrationalityisconsistentwithhavingsomethingotherthancompletecertaintyintheclaimsthatare actuallytrue.
⁷ TherearemorehistoricalsourcesonRobespierrethanwouldberemotelypossibletolist.Thethings IsayherearelargelydrawnfromrecentworkbyPeterMcPhee(2012),RuthScurr(2006),andespecially MarisaLinton(2013).ThestudyoftheCommitteeofPublicSafetybyR.R.Palmer(1941)ishelpfulfor seeingRobespierreincontext,andespeciallyseeinghimalongsidemenwithevenmoreextremecharacteristicsthanhis.
⁸ Mostrevolutionaryleadersareeitherpower-hungryorbloodthirsty.ButRobespierregenuinelyseems tohavebeenneitherofthose,exceptperhapsattheveryveryend.Linton(2013,97–9)isparticularlyclear onthispoint.
setofprinciplesturnedouttobe.Buttheinteresthereisinwhetherwecan findsome internalistprinciplethatcanbesaidtobetrue ‘aboveall’ inhiscase. ⁹
Robespierrehadamplereasontobelievethathehadendeduponthewrongtrack. Hewasn’tbrainwashedintobelievingthattheTerrorwasmorallyjustifiable;the reasonsforitwereclearlypresenttohim.TheresultsoftheTerrorweren’tplaying outinsomedistantland,orintheholdofaslaveship,theywererightinfrontofhim. Andheknewalotofmoralandpoliticaltheory.Hewaswelleducatedintheclassics. HereadMontesquieu.Heread,andadored,Rousseau.Hesatthroughhoursupon hoursofdebateeverydayabouttheefficacyandmoralityofgovernmentactions, bothbeforeandduringhisreign.Evenifonethinks,asIdo,thatsometimesthe reasonsfortheimmoralityofanactionarehiddenfromtheactor,thatcanhardlybe saidtobetrueinRobespierre’ scase.
SoIthinkwecanreasonablysayinRobespierre ’scasethatheviolatedtherule Followtheprinciplesitisrationaltoaccept.Andthatruleisaninternalrule,insome sense.Ifwetakeittobetheprimaryrule,thenwewon’tjudgepeoplebystandards thatarehiddenfromthem.Wemayjudgethembystandardstheydon’taccept,but onlywhentheyhavereasontoacceptthestandards.SoI’lltreatitasanother internalistapproach,thoughverydifferentfromtheapproachthatsaysitismost importantforpeopletofollowtheirownprinciples.
Sowehavetwoverydifferentkindsofinternalistapproachestoethics,epistemology, welfare,anddecisiontheory.Onesaysthatitis(most)importantthatpeoplefollow theirownprinciples.Theothersaysthatitis(most)importantthatpeoplefollowthe principlestheyhaverationalreasontoaccept.The first,initsstrongestform,says absurdthingsaboutthecaseoffanatics.AsI’llargueatlengthinwhatfollows,italso leadstonastyregresses.Theseconddoesnothavetheseproblems.Butitisveryhardto motivate.Wewillspendsometimeonthereasonsphilosophershavehadforwanting viewslikePolonius’s.Allofthese,I’llargue,pushtowardstheideathatthemost importantthingisthatpeoplefollowtheprinciplestheyactuallyaccept.Noneofthem, whenconsideredcarefully,giveusareasontopreferprinciplestheactororbelieverhas reasontoaccepttotheprinciplesthatareactuallytrue.Retreatingfrom Followyour ownprinciples to Followtheprinciplesitisrationaltoaccept letstheinternalistavoid harshcaseslikeRobespierre,butatthecostofabandoningtheinterestingreasonsthey havefortheirview.
1.2.5Somecaveats
I’vespokenfreelyinthissectionaboutthetruemoralprinciples.Thatwayof speakingpresupposesthattherearemoraltruths.Imeantobeusingthephrase ‘moraltruths’ inasnon-committingassenseasispossible.Idon’tmeantosaythat themoraltruthsaremind-independent.Ifitistruethatmurderiswronginvirtueof ourdisapprovalofmurder,itisstilltruethatmurderiswrong,andthat’senoughfor
⁹ Onethingthatwon’trescueintuitionsaboutthecaseistosaythat Dowhatyouthinkisright is importantonlyiftheagentis ‘procedurallyrational.’ Robespierreusedtherightmethodstoformmoral beliefs:hereadwidely,talkedtolotsofpeople,andreflectedonwhatheheardandsaw.Hejustgotthings catastrophicallywrong.GideonRosen(2003;2004)placesalotofemphasisonproceduralrationalityin defendingaformofinternalism,thoughhisaimisverymuchnottotrackintuitionsaboutparticularcases.
currentpurposes.NordoImeantoinsistthatthemoraltruthsareinvariantacross spaceandtime.Therearehardquestionsabouthowweshouldevaluateactorsfrom differenttimesandplacesifaformofmoralrelativismistrue.Butthosequestionsare largelyorthogonaltotheonesI’minterestedin.
Iamineffectassumingawayaverystrongformofmoralrelativism,onethat makesmoraltruthrelativetothemoralprinciplesoftheactorbeingevaluated.But that’snotaplausibleformofmoralrelativism.Ifmoralrelativismistrue,thenwhat moralityisrelativetoismuchmoreinclusivethanasingleperson;itissomethinglike aculture,orapractice.Andthereisadifferencebetweenwhatapersonaccepts,and whatistrueintheircultureorpractice.
Asbrieflynotedabove,I’malsoassumingthatthereisadifferencebetweenwhatis trueandwhatitisrationaltoaccept.AllIreallyneedhereisthatitcanberationalto belessthanfullycertaininsomemoralandepistemictruths.I’mnotgoingto assume,forexample,thatonecanrationallybelievemoralorepistemicfalsehoods. I’vespokenaboveasifthatispossible,butthatwasaconvenientsimpli fication. What’sgoingtoreallymatterisjusttheexistenceofagapbetweenwhat’strueand what’sreasonabletobelieve,andthatgapcanariseevenifallthethingsthatare reasonabletobelievearetrue.
Finally,youmayhavenoticedthatweendedupalongwayfromanythingthat couldbeplausiblyattributedtoLordPolonius.WhenhetellsLaertestobetrueto himself,I’mprettysurehe’snotsayinganythingaboutwhetherLaertesshouldhave beliefsthatarerationalbythestandardsthatLaertesshouldrationallyaccept.Yet whetherLaertes(oranyoneelse)shouldhavesuchbeliefsisoneofthequestionswe endedupbeinginterestedin.ThegoodLord’sroleinthisplaywasjusttointroduce thedistinctionbetweenfollowingone’sownprinciplesandfollowingthetrue principles.Withthatdistinctiononstage,wecanletPoloniusexitthescene.
1.3NormativeExternalismDefined
Normativeexternalismistheviewthatthemostimportantevaluationsofactionsand actors,andofbeliefsandbelievers,areindependentbothoftheactororbeliever’ s beliefaboutthevalueoftheiractionorbelief,andoftheevidencetheactororbeliever hasaboutthevalueoftheiractionorbelief.Theaimofthisbookistodefend normativeexternalism,andexplorewhyitisphilosophicallyimportant.
Itistemptingtostrengthenthiskindofnormativeexternalismfurther,andsay thatwhatoneshoulddoandbelieveiscompletelyindependentofwhatonebelieves oneshoulddoandbelieve.Butthisstrongindependenceclaimcan’tberight.(I’ m gratefulheretoDerekBall.)Ifonethinksthatoneshouldmurderone’sneighbours, thenoneoughttogetprofessionalhelp.Sometimesnormativebeliefschangethe normativesignificanceofotheractions.SotheexternalistclaimI’mdefendingisa littleweakerthanthisgeneralindependenceclaim.Itallowsthatanormativebelief B maychangethenormativestatusofactionsandbeliefsthatarenotpartofthe contentof B.ButtheexternalismI’mdefendingisstillgoingtobestrongenoughto allowalotofcritics.
Thestrongestkindofnormativeinternalismsaysthatthevalueofactionsand beliefsistightlytiedtothebeliefsthatactorsandbelievershaveabouttheirown
actionsandbeliefs.Itsaysthatthemostimportantmoralpreceptistodowhatyou thinkisright,andthemostimportantepistemologicalpreceptistobelievewhat youthinktheevidencesupports.Thestrongversionofinternalismisnotapopular position.Butithasanimportantroletoplayinthenarrativehere.That’sbecause therearemanyinteresting,andpopular,moderateversionsofinternalism.Yetonce welookatthemotivationsforthosemoderateversions,we’llseethattheyreallyare argumentsforthestrongest,andleastplausible,version.
Wecangeneratethosemoderateformsofnormativeinternalismbylookingatthe fourquestionsfromtheprevioussection.Someinternalistssaythatinternalismis truejustforactors(orbelievers),notforactions(orbeliefs).Somesaythatinternalist principlesarepartofthemoral(orepistemological)truth,notprinciplestoputabove all.Somesaythatinternalistprinciplesapplytojustoneofethicsorepistemology, notboth.Andsomesaythatwhatmattersisnotconformitytotheprinciplesone actuallyholds,butconformitytotheprinciplesonehasevidencefor.Andanswers tothesequestionscanbemixedandmatchedindefinitelytoproducevarietiesof internalisttheses.Here,forexample,arethreeprinciplesthatarebothwidely believed,andwhichyoucangetbymixingandmatchinganswerstothefour questions.
• Itisavicetofrequentlydothingsonebelievesarewrong,evenifthoseactions areactuallyright.
• Wrongactionsareblameless,andhencedonotreflectbadlyontheactorwho performsthem,ifthatactorbelievestheactionisright,andhasgoodreasonfor thatbelief.
• Abeliefisirrationalifthebelieverhasgoodevidencethatthebeliefisnot supportedbytheirevidence,evenifthat ‘higher-order’ evidenceismisleading.
AndI’mgoingtoarguethatthebestargumentsforthosepositionsovergeneralize; theyareequallygoodasargumentsfortheimplausiblestrongversionofinternalism. Sotheyarenogood.
Partoftheargumentherewillbepiecemeal:showingforaparticularinternalist thesisthattherearenogoodargumentsforitbutfortheargumentsthatleadall thewaytothestrongestformofinternalism.AndIcan’thopetodothatforallthe possiblethesesyoucouldgetbymixingandmatchinganswerstothefourquestions. ButIcanhopetomakethestrongformofexternalismmoreplausible,bothby showinghowithandlessomedifficultcases,andbyshowingthatthemostgeneral argumentsagainstitdonotwork.
1.4Guidance
ToillustratethekindofstorylineIsketchedintheprevioussection,let’sconsiderone popularargumentagainstexternalism.Theexternalistsaysthatpeopleshoulddothe rightthing,whateverthatis,whetherornottheyknowthattherightthingisinfact right.Itisoftenobjectedthatthisisnotparticularlyhelpfulguidance,andmorality shouldbemoreguidingthanthis.WeseeversionsofthisobjectionmadebyTed Lockhart(2000,8–9),MichaelSmith(2006,143),AndrewSepielli(2009,8),William MacAskill(2014,7)andbyHilaryGreavesandTobyOrd(2017).Theseauthors
differbetweenthemselvesaboutbothwhynormsthatarenotguidingarebad,some sayingtheyareunfair,othersthattheyareunhelpful,andaboutwhatconclusionwe shoulddrawfromthisfact.Buttheyagreethereissomethingbadabout Dotheright thing invirtueofitnotbeingguiding,andthinkweneedsomethingmoreinternalist.
Butifyouthink Dotherightthing isnotguiding,andweneednormsthatare guidinginjustthatsense,someverystrongconclusionsfollow.Afterall,ifnonguidingrulesarebad,thentheyshouldn’tbeanypartofourmoraltheory.Soitisn’t justthatweshouldtakehypocrisytobeonevicealongsidecowardice,dishonesty, andsoon,buttobetheonlyvice.Afterall,ifthereareothervicesatall,thenmorality asawholemaynotbeguiding.Nowwhois Dotherightthing notguidingto? Presumablytopeoplewholackfullmoralknowledge.Butsomeofthesepeoplewon’t havefullepistemologicalknowledgeeither.Sobythestandardthat Dotherightthing isnotguiding,principleslike Dowhatevertheevidencebestsuggestsisright,or Do whatevermaximizesexpectedrightness won ’tbeguidingeither.Ifwecan’texpect peopletoknowwhat’sright,wecan’treallyexpectthemtoknowwhat’sprobably righteither.
Sotakingguidancetobeaconstraintinthiswaypushesustoaversionof internalismthatreliesonactualbeliefsaboutrightness,notbeliefstheevidence supports,andreliesonaversionthattakesconformitytoone’sownvaluestobe ‘aboveall.’ Butifwedothat,wecan’tsayeitheroftheplausiblethingsIsuggested variousmoderateinternalistscouldsayaboutRobespierre.Thetwosuggestionswere tosaythatconformitytoone’sownvalueismerelyonevirtueamongmany,andthat goodpeopleshouldconformnottotheiractualprinciples,buttotheprinciplestheir evidencesupports.Ifwetakeguidancetobeaconstraint,thenbothwaysoutare blocked.Robespierrefailedbysomeveryimportantstandards,buthecouldn’tbe guided(inwhateversensetheinternalistmeans)bythosestandards.
We’llseethisstorylineafewtimesinwhatfollows.Theexternalistviewseemsto havesomeunattractivefeatures.Butwhenwespelloutjustwhatthefeaturesare, we ’llseetheyaresharedbyallbutsomeveryimplausibletheories.Thiswon’tjust holdinethics.TheepistemologicalpictureI’mgoingtodrawallowsforkindsof reasoningthatappearontheirfacetobeunacceptablycircular.Butwhenwetryto sayjustwhatthiskindofcircularitycomesto,we’llseethatblockingitwouldprovide enoughresourcestogroundanargumentforPyrrhonianskepticism.
1.5Symmetry
Ingeneral,one’sevidenceisrelevanttowhatoneshoulddo.Thenormativeexternalistdeniesanaturalgeneralizationofthislittleplatitude.Althoughevidenceabout mattersoffactisrelevanttowhatoneshoulddo,evidenceaboutthenormative,about thenatureofmoralityandrational,isnot.Evidenceaboutwhethertoturnleftor rightisrelevanttorationaldecision-making,evidenceaboutwhatiswrongorrightis irrelevant.Orsosaystheexternalist.
Thislookslikeanargumentagainstexternalism:itdeniesaveryplausiblesymmetryprinciple.Theprinciplesaysthatweshouldtreatallkindsofuncertainty,and allkindsofevidence,thesame.I’mgoingtospendmuchofthe firsthalfofthisbook
arguingagainstthesymmetryprinciple,butfornowlet’squicklysetupwhywe mightthinkthereisapuzzlehere.
We’llstartbythinkingthroughanexampleofwhereevidenceisrelevantto mundaneaction.Aperson,we’llcallhimBaba,islookingforhiscarkeys.Hecan rememberleavingtheminthedrawerthismorning,andhasnoreasontothinkthey willhavemoved.Sothenaturalthingtodoistolookinthedrawer.Ifhedoesthis, however,hewillbesadlydisappointed,forhistwo-year-olddaughterhasmovedthe carkeysintothecookiejar.
ThingswouldgobestforBabaifhelookedinthecookiejar;thatwayhewould find hiscarkeys.Butthatwouldbeaveryoddthingforhimtodo.Itwouldbeirrational tolookthere.Itwouldn’tmakeanysense.Ifhewalkeddownthesteps,walked straighttothecookiejar,andlookedinitforhiscarkeys,itwouldshockany onlookersbecauseitwouldmakenosense.Itusedtobethoughtthatitwouldnot shockhistwo-year-olddaughter,sincechildrenthatyounghadnosensethat differentpeoplehavedifferentviewsontheworld.Butthisisn’ttrue;wellbefore agetwochildrenknowthatevidencepredictsaction,andaresurprisedbyactionsthat don’tmakesensegivenaperson’sevidence(He,Bolz,andBaillargeon2011).Thisis becausefromaveryyoungage,humansexpectotherhumanstoactrationally(Scott andBaillargeon2013).
Inthisexample,Babahasawell-foundedbutfalsebeliefaboutamatteroffact: wherethecarkeysare.Let’scomparethistoacasewherethefalsebeliefsconcern normativematters.Theexampleisgoingtobemorethanalittleviolent,thoughafter thistheexampleswillusuallybemoremundane.Andtheexamplewill,inmy opinion,involvethreedifferentnormativemistakes.
Gwennegisataconference,andisintroducedtoanewperson. “Hi,” hesays, “I’mGwenneg,” andextendshishandtoshakethestranger’shand.Thestrangerreplies, “Nicetomeetyou,but youshouldn’tshakemyhand.IhavediseaseD,andyoucan’tbetoocarefulaboutinfections.” AtthispointGwennegpullsouthisgunandshootsthestrangerdead.
Nowlet’sstipulatethatGwenneghasthefollowingbeliefs,the firstofwhichisabout amatteroffact,andthenextthreeareaboutnormativematters.
First,GwennegknowsthatdiseaseDissocontagious,andsobadforhumansboth intermsofwhatitdoestoitsvictims’ qualityandquantityoflife,thatthesudden deathofapersonwiththediseasewill,onaverage,increasethenumberofqualityadjusted-life-years(QALYs)ofthecommunity.¹⁰ Thatis,althoughthesuddendeath ofthepersonwiththediseaseobviouslydecreasestheirQALYsremaining,tozeroin fact,thedeathreduceseveryoneelse’sriskofcatchingthediseasesomuchthatit increasestheremainingQALYsinthecommunitybyamorethanoffsettingamount.
Second,Gwennegbelievesinastrongversionofthe ‘straightrule.’ Thestraight rulesaysthatgiventheknowledgethatxpercentoftheFsareGs,otherthingsequalit isreasonabletohavecredencethatthisparticularFisaG.Justabouteveryone believesinsomeversionofthestraightrule,andjustabouteveryonethinksthatit
¹⁰ QALYsaredescribedinMcKieetal.(1998),whogoontodefendsomephilosophicaltheses concerningthemthatI’mabouttoassigntoGwenneg.
