1

RELIGIOUSDISAGREEMENT

HelenDeCruz

This is a draft for a book that has appearedwith Cambridge University Press. Pleaserefertothefinalpublishedversion:

https://www.cambridge.org/core/elements/religious-disagreement/18A9B767E64FA47F97CD8E9814233210

Contents

1Howshouldwerespondtoreligiousdisagreement?

1.1Thesignificanceofreligiousdisagreement

1.2Conciliationismandsteadfastness

1.3Isreligiousdisagreementphilosophicallyintractable?

1.4Isreligiousdisagreementinsensitivetoevidence?

1.5Privateevidenceandreligiousdisagreement

2Irrelevantinfluencesandreligiousdisagreement

2.1Youonlybelievethatbecause

2.2Howirrelevantinfluencesaffectreligiousbeliefs

2.3Thesafetyobjection

2.4Thearbitrarinessobjection

2.5Makingreligiousdiversityworkinthephilosophyofreligion

3Conversionanddisagreementwithformerselves

3.1ReligiousconversionandtheIndependencePrinciple

3.2Conversionasatransformativeexperience

3.3Disagreementwithone’sformerself

3.4Disagreementwitharecentlyconvertedpeer

2

3.5Rationalargumentandconversion

3.6Thoughtfuldisagreement

4What(ifanything)canweinferfromcommonconsent?

4.1Whatcommonconsentmightmean

4.2TheoriginalConsensusGentium:frominnatenesstotruth

4.3Commonconsent,self-trust,andevidence

4.4Synergyandtheepistemicsignificanceofconsent

5Religiousexpertiseanddisagreement

5.1Theimportanceofreligiousexpertiseindisagreement

5.2Whatisareligiousexpert?

5.3Twomodelstosolvethenovice/expertproblem

5.4Theexpertasteacher

5.5Decidingwhototrust:Maimonidesonthesages

6Whyphilosophymattersforreligiousdisagreement

References

3

Abstract

This book examines what we can learn from religious disagreement, focusing ondisagreement with possible selves and former selves, the epistemic significance ofreligiousagreement,theproblemofdisagreementsbetweenreligiousexperts,andthesignificance of philosophy of religion. I will show how religious beliefs of othersconstitute significant higher-order evidence. At the same time, we should notnecessarily become agnostic about all religious matters, because our cognitivebackgroundcolors thewayweevaluateevidence.Thisallowsus tomaintain religiousbeliefsinmanycases,whileneverthelesstakingthereligiousbeliefsofothersseriously.

4

Chapter1:Howshouldwerespondtoreligiousdisagreement?

1.1Thesignificanceofreligiousdisagreement

In themidst of thiswar ofwords and tumult of opinions, I often said tomyself:Whatistobedone?Whoofallthesepartiesareright;or,aretheyallwrongtogether?Ifanyoneofthemberight,whichisit,andhowshallIknowit?(Smith,1902,4)

Joseph Smith grew up in a religiously diverse community in New York (Palmyra andManchester) during the Second Great Awakening (about 1790-1850s). He came intocontact with many religious (mainly Christian) denominations, including Methodism,Presbyterianism,Baptism,butalsowithfolkreligiousmagic,whichwaspracticedinhisfamily. The open disagreement between these religious groups,whichwere trying towinconverts,troubledhim.Whomshouldhetrust?

Presbyterians were most decided against the Baptists and Methodists,andusedallthepowersofbothreasonandsophistrytoprovetheirerrors,or, at least, tomake thepeople think theywere in error.On theotherhand, theBaptists andMethodists in their turnwereequally zealous inendeavoringtoestablishtheirowntenetsanddisproveallothers.(Smith,1902,3–4)

The fifteen-year-old Smith solved this conundrum by retreating into the woods, andaskingGodforguidance.Thiseventuallyledtoaseriesofvisions,whichinturnledhimto establish a new religiousmovement, the Church of Jesus Christ of the Latter DaySaints(alsoknownasMormonism).Putinepistemologicalterms,Smith’ssolutiontotheproblemofreligiousdisagreementwastotrytoseekadditionalevidence,inhiscase,intheformofrevelation1.Thisexampleillustratesthatreligiousdisagreementconstitutessomeformofevidence.Itisapeculiarformofevidence,inthatitdoesnotdirectlybearon the truthor falsityof religiousbeliefs,but ratheronusasepistemicagents.Whenconfrontedwithconflictingviewpoints,wesometimestrytogathermore information,as inSmith’scase.Disagreement is thusaformofhigher-orderevidence.Higher-orderevidencehasafewpeculiarfeatures.Forexample,itsvalueseemstobedependentonwho is evaluating it. If Kabita disagrees with Dan about the epistemic credentials ofBuddhism,thenKabita’sbeliefsarehigher-orderevidenceforDan,butnotforKabita.ItwouldindeedbeabitpeculiarifKabitasaid,“Iamaverythoughtfulsortofperson,andIamaBuddhist.MybeliefmustconstitutesomeevidenceforBuddhism!”Shewouldbehubristic,tosaytheleast.ButitisnotatallunusualifDantookKabita’sbeliefassomeformofevidence,thinkingalongthefollowinglines,“IknowKabitaisathoughtfulsortof person. If she is a Buddhist, maybe Buddhist beliefs—about enlightenment,1This account of Joseph Smith’s “first vision” is canonical amongMormons, but it issomewhatidealized.Foramorenuancedaccount,seeTaves(2016).

5

reincarnationandthelike—arenotasoutlandishasIthoughttheywere.Maybethereissomething I am missing.” This asymmetry merely demonstrates that disagreementconstitutes evidence that is relative to the agent, not that it would be irrelevant(Christensen,2010;Matheson2009).

Joseph Smith’s case also illustrates another feature of religious disagreement:recognizingitsepistemicsignificancehasapractical,real-worldimpactonthereligiousbeliefswehold.Onceweseethatpeoplewhoarejustasthoughtfulandwell-informedaswearecometoverydifferentreligiousviewpoints,wecannolongergoontakingourown religious views for granted. Nicholas Wolterstorff (1996) draws a distinctionbetween analytic and regulative epistemology. Analytic epistemology, according toRoberts and Wood (2007, 20), aims to produce “theories of knowledge, rationality,warrant,justification,andsoforth,andproceedsbyattemptingtodefinetheseterms.”Bycontrast,regulativeepistemologyisamorepracticallyorientedwayofthinkingaboutthese concepts; it tries to provide guidance for how to shape our doxastic practices(Wolterstorff borrows the term “doxastic practice” from Alston [1991]). A doxasticpractice is a system of habits bywhichwe form our beliefs. Regulative epistemologyproposesdoxasticpracticesthathelpustoacquirebeliefsthatareresponsiblyformed.They can, for instance,beaimedatobtainingasmany truebeliefs asweareable,orthey canbemore risk-averse andhelp us to avoidmakingmistakes. As James (1902)alreadynoted,thereissometimesatensionbetweenthesetwodesiredstatesofaffairs(believing true things, and avoiding believing false things), so a risk-seeking personmightbemoreinclinedtobelievewhatisnotcertain,whilearisk-aversepersonwouldavoidit2.Thusadoxasticpracticeneedstospecifyfirstwhatepistemicutilitywewouldliketoobtain,forexample,obtainingtruebeliefs,avoidingfalsebeliefs,oravoidingfalsebeliefsofspecifickinds.Oncespecified, itcanhelpusobtaintheseutilities.Regulativeepistemologiesareoftenborneoutofaconcreteneed,whichisprecipitatedbyasocialandintellectualcrisis(Wolterstorff,1996).InthecaseofDescartesandLocke,thiswasthe unraveling of the medieval Christian consensus in the seventeenth century. Toprovideasimplifiedpictureofwhathappened,at theendofMiddleAges thegeneralconsensusonmoralandfactualmattersweakenedasaresultofseveralfactors.Theseincluded the increasing recognition that there is a wide diversity of religious beliefsacross the world, due to increasing contact with foreign cultures as a result ofcolonialismandtrade.Reportsofreligiousbeliefsinothercultureswereoftensecond-hand,notsystematicallycollected,anddistorted.Nevertheless,theyprovidedevidencethat religious beliefs varied considerably, and thatmonotheismwas not universal. As

2See Pettigrew (2016) for a recent formal argument that vindicates James and thatshows that it is rationally permissible for epistemic risk-seekers to go significantlybeyond the evidence, and believe something for which they can never haveincontrovertibleevidence,suchastheexistenceoftheexternalworld,ofotherminds,orofGod.

6

Hume (1757, 2) summarized it, “no two nations, and scarce any twomen, have everagreedpreciselyinthesamesentiments.”

As we will see in chapter 4, observations like these weakened the argument fromcommonconsentfortheisticbelief,theargumentthattheismmustbetruebecauseitisuniversal. Added to thiswas the growth of experimental science,which showed thatmanyreligiousclaims,suchasabouttheageoftheearthortheoriginofspecies,werefalse.Furtherepistemicshiftsoccurredwiththeendoflogicalpositivisminthemiddleofthetwentiethcentury.Logicalpositivismsoughtnewepistemiccertaintybyappealtoempirically verifiable statements. With its downfall, it became clear that scientificfindings could not take on the role that formerly religious dogmas had played in theMiddle Ages. Today, wemay be experiencing another epistemic crisis, the increasingpolarization and tribalization of beliefs, exacerbated by political echo chambers. Forexample, a study by Gauchat (2012) shows that scientific beliefs have becomeincreasinglypoliticallypolarizedintheUnitedStatesfromthe1970sonward.Giventhatourbeliefsareincreasinglypolarized,whatdoxasticpracticesshouldweadopt?Theaimofthissmallmonographistoberegulative,ratherthananalytic,eventhoughitwillusetoolsofanalyticepistemology.Iwillnothereattempttomakeacomprehensivesurveyofreligiousdisagreementinallitsdifferentforms.Rather,Iwillexaminewhatpracticalconclusionswecandrawinthefaceofparticularformsofreligiousdisagreement.

1.2Conciliationismandsteadfastness

Let’s for the moment assume that disagreement about religion has some evidentialvalue (I will further in this chapter respond to some objections to this claim). Socialepistemologistshavebeendebatinghowweshouldrespondtothisevidence.Takethisexample,adaptedfromClaytonLittlejohn(2013):

Complacent atheist:Clayton is a complacent atheist: he strongly believesthere is plenty of evidence against the existence of God. However, he isalsoawareof the fact that thereare severalphilosopherswhobelieve inGod. Many of these have thought carefully about the matter, and areexperts in epistemology, metaphysics, and other relevant philosophicalsubdisciplines.

Therearetwobroadlinesofresponseopentothecomplacentatheist.Thefirstoptionfalls under the umbrella of conciliationism. The conciliatory position3holds that weshould revise our opinions, or become less confident of them, in the face ofdisagreement with someone we consider to be an epistemic peer about the subjectmatter.SoifClaytonbelievesthat,say,Linda(atheistphilosopher)isjustasthoughtfuland epistemically virtuous as he is, and that she has access to the same body ofevidence,heshouldrevisehisbeliefs.Maybeheshouldsuspendjudgmentontheissue3SeeChristensen(2011)foranin-depthexplanationofthisterminology.

7

entirelyandbecomeanagnostic,asFeldman(2007)recommends.Ifhedoesnotwishtogo this far, he should at least become less complacent in his atheism, say,move hiscredence4that atheism is true from .9 to .7 (depending on whether he believes heshouldlendequalweighttoLinda’sviews).

Asecondoptionistoremainsteadfast,andnotchangeone’scredencesatall.Thereareseveralmotivationsforremainingsteadfast,whichmayapplyinareligiouscontext.Forexample, Wedgwood (2007) points out the epistemic asymmetry between my own(religious)experiencesandevidence,bothofwhichguidemedirectly,versus thoseofothers,whichcanonlyguidemeindirectly.Thisasymmetryexplainswhyavividreligiousexperience can have strong evidential force forme, but not for the person I tell myreligiousexperienceto.Indeed,thereisanimpressivecollectionofreligiousexperiencesinJames(1902),andmorerecently, intheAlisterHardyReligiousExperienceResearchCentre,which has collected over 6,000 reports since 1969. These experiences have aspecific phenomenology, e.g., “The experience was unbelievably beautiful, and I willneverforgetthequalityofthatbrightwhitelight.Itwasawesome.”Buttheirevidentialforceishardtoconveytothirdparties.

Note that conciliationists do not always change their views. For one thing,conciliationists have not given up their belief that conciliationism is right in spite ofencountering many epistemologists who disagree with them. Under somecircumstances it is reasonable to stick to youroriginal beliefs, for instance,when it ismore likely that the other party has made a mistake. But if you do not have anyindependentreasontothinkthatyourinterlocutor,withwhomyoudisagree,hasmadeamistake,conciliationismdoesrequiresignificantbeliefrevision.Epistemologistshaveproposed several principles that would separate these two ways of responding todisagreement.Oneoftheseistheindependenceprinciple:

Independence: In evaluating the epistemic credentials of another’sexpressed belief about p, in order to determine how (or whether) tomodifymyownbeliefaboutp,Ishoulddosoinawaythatdoesn’trelyonthereasoningbehindmyinitialbeliefaboutp(Christensen,2011,2)5.

This principle can help reasoners guard against blatant circular reasoning (“Well, ofcourse,sinceatheismistrue,Lindamustbewrong”)andencouragesepistemichumility.Itmaps out plausible courses of action inmany cases of peer disagreement, such asChristensen’s (2007) classicmentalmath case. Inmentalmath, two restaurant goerssplitthebillandendupwithdifferentcalculationsofhowmuchtheyowe,afteraddingthetip.Itseemscommendabletoloweryourcredenceinyouroriginalbelief,say,that

4Anagent’scredenceinapropositionthatpmeasuresherdegreeofconfidenceinp.5We will look at another proposed key principle that separates conciliationism andsteadfastness,namelyuniqueness,inchapter2.

8

you each owe 23 dollars, in the face of the other person who has come up with adifferent amount, say, 26 dollars. However, sometimes a disagreement does notconstituteevidenceagainstone’sownbelief,butagainsttheviewthattheotherpersonisone’sepistemicpeer.JenniferLackey(2010)imaginesthefollowingsituation,termedelementarymath: IfindoutthatmyfriendHarrythinksthat2+2=5.Thisshouldnotlead me to revise my belief that 2 + 2 = 4, but rather lower my opinion of Harry’sarithmeticalcapacities.Claytoncould, inasimilarvein,concludethatLindaiswoefullymisled about the question of God’s existence, even if she is in general an excellentphilosopher. Examples like these indicate that our intuitions aboutwhat to do in thefaceofdisagreementwilldivergedependingonwhatthedisagreementisabout.Andaswewillsee,thecausesofthedisagreementarealsorelevant.

This bookwill examinedifferent formsof religious disagreement (or agreement), andwhat we can learn from them. It is written in a broadly conciliationist spirit: I amworking from the assumption that religious disagreement does provide higher-orderevidencetoone’s religiousbeliefs,andthat it should impactone’sbeliefs. In thenextchapters I will look at disagreement with possible selves, with former selves, theepistemic significance of agreement about religion, the problem of religious expertdisagreement, and conclude by outlining the significance of philosophy of religion inreligious disagreement. In each of these scenarios, I will show how conciliationismprovidestherightresponse,andhow—atthesametime—itdoesnotmeanweshouldnecessarilybecomeagnosticaboutall religiousmatters.The reasonweshouldnotbeagnostic is that our own cognitive background constraints and colors the way weevaluate evidence. This allows us to maintain religious beliefs in many cases, whilenevertheless taking the religious beliefs of other people seriously, and often alsorevisingourbeliefsinthelightofthem.

Chapter2examineswhat itmeanstobe indisagreementwithpossibleselves:what ifyou had been born and raised in Afghanistan,where 99 percent of the population isMuslim,or inPapuaNewGuinea,where99percent isChristian? Inall likelihood,youwouldhave endedupholding themajority belief. Should thisworry you? Iwill arguethatitshouldnot,butthattheroleofirrelevantinfluencesstillposesaproblematthemacrolevel,specificallyinconstrainingtherangeofviablehypothesesinthephilosophyofreligion.Chapter3looksatdisagreementwithyourformerself:ifyouconvertedtoadifferentreligiousaffiliation,canyoubeconfidentthatyourpresentbeliefismorelikelytoberight?Iwillarguethatbecausereligiousconversionisepistemicallyandpersonallytransformative,youcannotassumethatthis is thecase.Thebestwaytoevaluatethebeliefs of a convert (including yourself) is to engage in reasoned debate. Augustine’sarguments in De utilitate credendi (On the usefulness of belief) will illustrate thisapproach. Chapter 4 looks at the flipside of disagreement, namely the epistemicsignificance of agreement, in particular agreement about the existence of thesupernatural. I will examine the argument from common consent, its merits andproblems. Chapter 5 will analyze how we ought to respond to disagreement amongreligious experts. It looks at models of expertise and the proper response to expert

9

disagreement.Iproposeanewmodelofexpertise,theexpert-as-teacher,incorporatingadvice offered byMaimonides in hisGuide of the Perplexed. Chapter 6 concludes byshowingthatphilosophicalreflectioncanplayaconstructiveroleinreligiousdebate.

I will now consider three arguments against conciliationism in the face of religiousdisagreement.Thefirstisthatreligiousdisagreementistoomessyandcomplextobeofphilosophical interest. The second is that religious beliefs are insensitive to evidence,and therefore, cannot be revised in the light of higher-order evidence, rendering thediscussionmoot.Thethird is thatrelevantevidence inreligiousdisagreement,suchasreligiousexperience,isprivateandcannotbesharedbetweenparties.

1.3Isreligiousdisagreementphilosophicallyintractable?

Clear-cutcaseslikementalmathelicitconciliatoryintuitions:IfIhavenoreasontothinkthatIambetteratmentalarithmetic,itwouldseemprudenttobelessconfidentwhenmy epistemic peer and I come up with different numbers. But what about religion,politics, and all those other messy cases where we frequently find ourselves indisagreement? Maybe the concept of epistemic peer is not useful in such cases, asAdamElga(2007)andothershaveargued.SupposethebeliefweareinterestedinistheexistenceofGod,asconceptualizedintheAbrahamictraditions.Beliefinthisbeingissotied upwith our other beliefs, including political andmoral beliefs, that it is hard toassess to what extent the other person is an epistemic peer. Kelli Potter (2013) hasargued that it in many cases difficult to gauge whether a religious disagreement isgenuinelyadisagreement.

Given the messiness of religious disagreement, one can see why the philosophy ofdisagreement—in spite of a clear and continued interest in the topic—tends to useclear-cutexamplessuchasrestaurantbillsandsimplevisualperception,insteadofreal-world religious cases. However, excluding messy cases from epistemologicalconsideration would leave us none the wiser about the rational status of beliefs wegenuinelycareabout,suchasinpolitics,philosophy,religion,andmorality.Wecannotusetoyexamplestoreasonourwayintothemorecomplexcases,inpartbecausethesetoyexamplesalreadyelicitdiffering intuitions (comparementalmathwithelementarymath).

Arguably, themost interesting cases of disagreement occur when parties come withdifferent sets of background beliefs. In some of these cases, the parties concernedconsider their interlocutors to be peers, even though they do not know if the otherpersonhasexactly thesameevidenceor isequallyvirtuous.Canpeople insuchcasesstillbecalledepistemicpeers?Itdependsonone’snotionofepistemicpeerhood.Theterm “epistemic peer” was originally coined by Gary Gutting (1982) who describedepistemicpeersintermsofintellectualvirtues.AishaandBenjaminareepistemicpeersiftheyaresimilarinattentiveness,thoroughness,andothervirtues.Althoughthisistheoldestdefinitionofepistemicpeerhood,anditisnotoftenusedintherecentliterature,

10

my survey on religious disagreement among academic philosophers (De Cruz, 2017)revealsthatitisstillpopular.Sixtypercentofsurveyedphilosophersfavoredadefinitionofepistemicpeersas similar in intellectual virtues.Subsequentdefinitions focusedoncognitiveequality(Lackey,2010),whereAishaandBenjaminareepistemicpeersiftheyare similar in their cognitive capacities and limitations, and on evidential equality(Christensen, 2007), where they are epistemic peers if they have access to the sameevidence for the domain under consideration. Sameness of evidence is a difficultcriteriontomeet.Evenpeoplewhoarecloselymatchedintrainingandexpertise,suchasdissentingphilosophersof religion,willnothaveaccess tothesameevidence (e.g.,theywill have read different papers, gone to different graduate schools). Even peerswhohaveaccesstothesameevidencemaynothaveassesseditcorrectly:perhapstheydisagreefundamentallyonwhichtheoreticalvirtuestouse intheirdiscussion,suchassimplicity,fruitfulness,generality,andcoherencewithbackgroundknowledge(Douven,2010). Different weightings of such virtues could lead to divergent appreciations ofnaturaltheologicalarguments,suchasthecosmologicalargument.

Suppose thatwedonotknowwhetherpartieshave thesameevidence,arecognitiveequals, or are equally virtuous, does this make their disagreement epistemicallyirrelevant?Thisdoesnotseemtobethecase.Evenifone’sinterlocutorisanepistemicinferior, suchasanundergraduatestudentversusaprofessor, thedisagreementdoesconstitutesome(albeitweak)evidence.Thereare,ofcourse,manycaseswherewedonotneed toheedourepistemic inferiors (e.g., ifmy five-year-oldand I comeupwithdifferentnumbersinamentalmathproblem,Idonotneedtorevisemyconfidencethatmycalculationisright).Butinmanysituations,wesimplydonotknowifapersonisinasgoodanepistemicpositionasweare(King,2012).Thisisnotjustinmessyreligiousdisagreementsbuteveninmoreclear-cutcasessuchasmentalmath:AishamaybelievethatsheandBenjaminareequallygoodatmentalarithmetic,butinrealityBenjaminissignificantlyweaker.

Lackey (2010) favors the concept of ordinary disagreement. In a case of ordinarydisagreement, Aisha and Benjamin consider themselves to be epistemic peers on thetopicpriortotheirdisagreement,andtheycometorealizethattheydisagree. Insuchsituations,whilewedonotknowwhetherthetwopartiesareevidentiallyorcognitivelyequal,themerefactofdisagreementconstitutes(defeasible)evidence.Attheveryleast,thedisagreementshouldleadustoinquirefurtherintotheother’sposition,bylookingat the reasons they might have for holding it. For the purposes of this book, I willunderstand epistemic peer disagreement as ordinary disagreement, unless otherwisespecified.

1.4Isreligiousdisagreementinsensitivetoevidence?

A secondworry for thephilosophicaldiscussionofpeerdisagreement is that religiousbeliefsmightnotbesensitivetoevidenceinthesamewayasordinarybeliefsare.When

11

weargue about religion, it is not uncommon to hear appeals to personal satisfactionandmeaning.Whenreligiousbelieverstrytowinconverts,theywillsaythingssuchas“Havinga relationshipwith Jesusbringsme joy!” rather than“Herearesomereasonswhy I think the existence of God is more likely than God’s non-existence”. Neil VanLeeuwen(2014)hasarguedthatreligiouscredencesarelargelyinsensitivetoevidence.ToVanLeeuwen,thebelief“Godiswatchingme”iscognitivelydistinctfromthebelief“Thepolice arewatchingme”. The latter beliefwould be vulnerable to evidence; theformerwouldnot.Religiousbeliefsarevulnerabletospecialauthority,bypeoplewhoarerespectedintheirreligiouscommunityandwhofulfillaspecialrolethere.Althoughthisclaimthatreligiouscredencesareinsensitivetoevidenceisadescriptiveclaim,notanormativeone, it has repercussions for theepistemologyof religiousdisagreement.Howcouldweusereligiousdisagreementashigher-orderevidenceifitweregenuinelythecasethatourreligiousbeliefswerepsychologicallyinvulnerabletoevidence?

Thereissubstantialevidencethatreligiousbeliefsareprocessedinapeculiarway.Forexample, Larissa Heiphetz and colleagues (2013) examined how adults and childrenreasonaboutbeliefs(boththeirownandthoseofotheragents).Theyfoundthatyoungchildren(agedfiveandolder)alreadydrawadistinctionbetweenfact-basedbeliefs(e.g.,the sizeof germs) andopinion-basedbeliefs (e.g.,which color is theprettiest).Whenasked whether two agents who disagreed about an opinion (e.g., whether broccolitastes nice), children and adults tended to think that both agents could be right. Forfactual beliefs, they thought only one agent could be right. Religious beliefs fellsomewhere in between, with adults responding at chance level about whether bothagents could be right. Andrew Shtulman (2013) found that undergraduates aremorelikely to refer to authorities when justifying their belief in the existence of religiousentities(e.g.,angels,God,souls),comparedtotheirbelief intheexistenceofscientificentities(e.g.,fluoride,electrons,genes).

However,religiousbeliefsarenotuniqueinthisway.Politicallypolarizedbeliefssuchasbeliefs about climate change and evolutionary theory in the United States show thesamepatternofresistancetoevidence,apatternthatmightbeexplainedbyprocessingfluency (Levy, 2017). Although religious beliefs are intimately tied to factors such aspersonal identity and meaning, they still are to an important extent about factualmatters (the same holds for political beliefs). This is why attempts such as Gould’s(2001)non-overlappingmagisteria,whichaimstoneatlyseparatethedomainofscienceasthedomainofstatementsoffact,andthedomainofreligionasthedomainofoughtstatements, fail. It is inpracticeoftennotpossibletoseparatethefactualclaimsfromnormative or preference claims in religious statements. If religion did not make anystatementoffact,butmadeonlyclaimsaboutvalueandethics,theseclaimscouldnotbe justified using purported facts. For example, one could not argue that one shouldlove one’s neighbor because it pleases the Creator, because that is a (purported)statementoffact(Godispleasedbyneighborlylove)(Worrall,2004).

12

1.5Privateevidenceandreligiousdisagreement

Peter van Inwagen (1996) has argued for the steadfast view by appealing to privateevidence.We frequently have some (incommunicable) insight or experience that wemight suppose the other person lacks. This can act as a symmetry breaker:whenwehavegoodreasonstothinkwehaveinsighttheotherpartylacks,thereisnoreasontomoveourbeliefsintheirdirection.Inacasethathasbecomesomethingofaclassicinthe epistemology of disagreement, van Inwagen expresses his puzzlement that DavidLewis,aphilosopherheadmires,disagreesfundamentallywithhimaboutwhetherfreewillanddeterminismarecompatible—vanInwagenthinkstheyarenot;Lewisthoughttheyare.Tobreakthesymmetry,hearguesthathehassomesortofspecialinsightthatLewis,forallhisperspicacity,lacks.

But how can I take these positions? I don’t know. That is itself aphilosophical question, and I have no firm opinion about its correctanswer.IsupposemybestguessisthatIenjoysomesortofphilosophicalinsight ... that, for all hismerits, is somehowdenied to Lewis. And thiswouldhavetobeaninsightthatisincommunicable—atleastIdon’tknowhowtocommunicate it—,or Ihavedoneall I cantocommunicate it toLewis,andhehasunderstoodperfectlyeverythingIhavesaid,andhehasnot come to sharemy conclusions. Butmaybemybest guess iswrong.(vanInwagen,1996,138)

This exampledemonstrateshowadopting the steadfast view canerode thenotionofepistemicpeer:ifvanInwagenbelieveshisallegedepistemicpeertolacksomeinsighthepossesses,hedoesnotreallyconsiderhimapeer(atleastnotaboutthequestionoffree will), but sees himself as in a superior position. What are we to make of suchprivateevidence? In the religiousdomain, theobviouscandidate for incommunicable,unshareableprivate evidence is religious experience. But atheistsmay alsohavenon-propositional,non-inferentialevidencefortheirposition.Theoccurrenceofevilssuchasthe suffering of innocent children may give the atheist an experience of God’s non-existence(Gellmann,1992).

Religious experiences are common, but less common than religious beliefs. A surveyamong ordinary believers by the Pew Forum indicate that 59 percent of Americansregularlyhavereligiousexperiences6,whichmakethemlesscommonthanthenumberof Americans who believe in God (around 90 percent in the same survey), or thanpeople who consider themselves members of a religious denomination (over 70percent).AnthropologicalresearchbyTanyaLuhrmannamongevangelicalChristiansoftheVineyardcommunityindicatesthatreligiousexperiencesaredependentonpractice.EvangelicalslearntodistinguishtheirownthoughtsfromGodwhospeakstothem.Thisis,asLuhrmann(2012b,39)putsit,“askilltheymustmaster”.Theymasterit,gradually,using a variety of techniques such as individual and collective prayer, and more

6http://www.pewforum.org/2015/11/03/chapter-2-religious-practices-and-experiences/

13

imaginativeexercisessuchasimagininghavingadatenightwithGod.Thisgivesrisetoapeculiarproblem,whichLuhrmanndubsan“epistemologicaldoubleregister” thatshefrequentlyencounteredamongtheparishionersshestudied(Luhrmann,2012a,380).Inspite of the vivacity of their religious experiences, practitioners often remain unsurewhether a particular religious experience is really of God, or stems from theirimagination. As one congregant put it prosaically, “sometimeswhenwe think it’s thespiritmoving,it’sjustourburritofromlunch”(Luhrmann,2006,149).Atthesametime,they feel they are certain that God exists. This certainty is not a result of any givenexperience of God, which is highly ambiguous for them. The Roman Catholic mysticTeresaofÁvila(1577[1921])broachedasimilarproblem.Religiousexperiencescanbevery vivid, but their memory and vividness quickly fade and leave the personexperiencingthemunsurewhethertheexperiencewasafigmentoftheir imagination.(Teresaalsodiscussedthepossibilitythattheymightbeofdiabolicalorigin—thiswouldofcoursestillbolsterbelief inasupernatural realm,butnotnecessarily inanall-goodGod.)

Thereareotherproblemswithusing religiousexperienceasa symmetrybreaker.Thediversity of religious experiences precludes a straightforward interpretation along thefollowinglines:NobuyoisaShintōpriestess.Sheworks inanurbanShintōshrine,andregularlyexperiences thepresenceofkami, spiritswhoenter the shrineandwhoareworshipped. She concludes from this that the spiritual phenomena of Shintō aregenuine,andthatkamiexist.Forthis,shemightusesomethinglikeSwinburne’s(2004)credulityprinciple: if itseemsepistemicallythatkamiarepresent,thenprobablykamiarepresent.Otherpeoplehavereligiousexperiences thatprovidepurportedevidenceof other religious claims, not easily compatible with Shintō (such as Abrahamicmonotheism).Giventhatevenreligiousexperienceisnotafirmsymmetrybreaker,thestandingof incommunicable insights isalsodoubtful. Inanycase, itworksbothways:Lewismayhave incommunicable insights about freewill that are somehowdenied tovan Inwagen. If you have no special reasons to believe that you are less likely to bewrong, and the other seems an epistemic peer in other relevant respects, privateevidencecannotbreakthesymmetry,andsoitcannotjustifythesteadfastposition.Insum, itmay be difficult in practice to knowwhether the peoplewe find ourselves inreligious disagreement with are peers. But even if they are not peers, their beliefsconstitutesomeformofhigher-orderevidenceforus.

Chapter2:Irrelevantinfluencesandreligiousdisagreement

2.1Youonlybelievethatbecause...

ThemedievaltheologianAbūḤāmidMuḥammadibnMuḥammadal-Ghazālī(ca.1058–1111) considered the role of irrelevant influences on religious beliefs. Al-Ghazālī wasborninthePersiantownofTabaran,inthedistrictofTus(northwesternIran),wherehe

14

receivedatraditionalIslamiceducation.Asayoungman,hemovedtoNishapur,wherehestudiedunderal-Juwaynī,aprominentAshʿariteteacher.AshʿarismwasanorthodoxSunni theological school that was mindful of scripture without being literalist. ItopposedtheMuʿtazila,anotherphilosophicaltheologicalschoolthatprizedreasonasasource of knowledge, and that was heavily influenced by ancient Greek—particularlyAristotelian—philosophy.

Al-GhazālīwasverycriticalofMuslimphilosopherswhodrewupontheancientGreeks,collectively termed falāsifa (the philosophers); his main beef was with Ibn Sīnā(Avicenna) and his arguments directly challenged Ibn Sīnā’s claims. In particular, al-Ghazālī was skeptical of the philosophers’ self-professed reliance on reason andargument.Hearguedthatthephilosopherswereguiltyoftaqlīd7,uncriticallyacceptingthe views of Aristotle and other ancient philosophers. Like Christians and Jews, thephilosophershadthemisfortuneofbeing“born intoanun-Islamicatmosphere [ghayrdīnal-Islām],andtheirancestorshadpursuednobetterways”(al-Ghazālī,11thcentury[1963a], 1–2). Orthodox Muslims, by contrast, received divine revelation that wasproperly transmitted to them through the Qurʾān and ḥadīth (Griffel, 2017). So, al-Ghazālīreckonedhehadthecorrectreligiousbeliefs,comparedtoChristians,Jews,andadherentstoancientphilosophy. InhisspiritualautobiographyDeliverancefromError,he expressed more doubts about having the right religious views. He observed thatchildren of Jews, Christians, or Zoroastrians tend to almost exclusively follow theirparents’religion.

... as I drew near the age of adolescence the bonds of mere authority(taqlīd)ceasedtoholdmeandinheritedbeliefslosttheirgripuponme,forIsawthatChristianyouthsalwaysgrewuptobeChristians,Jewishyouthstobe JewsandMuslimyouths tobeMuslims. I heard, too, theTraditionrelated of the Prophet of God [ḥadīth] according to which he said:“Everyonewhoisbornisbornwithasoundnature[fiṭrah];itishisparentswhomakehimaJeworaChristianoraMagian[Zoroastrian]”.(al-Ghazālī,ca.1100[1952],21).

This consideration of religious diversity was the start of a skeptical meditation, notunlikeDescartes’Meditations. If religiousbeliefs,whichare sucha largepartofone’sviews, are the result of the accident of one’s birth, how can we know anything? Al-Ghazālī resolved to put his religious faith on firmer epistemological grounds. Heconsidered, and rejected, senseperceptionas theprime sourceof knowledge, as it issometimes wrong: it was generally known in al-Ghazālī’s time that although the Sunappearstobesmallerthanacoin,itislargerthantheEarth.Evenourintellectualbeliefscanbewrong,asweoftenholdmistakenandunfoundedbeliefswhenwearedreaming.

7Taqlīd denotes theuncritical acceptanceof testimonyor authorityof onepersonbyanother;itusuallyhasanegativeconnotation.

15

Andgiven that ourpresent earthly life is often compared to adream-like statewhencomparedtowhatwewillknowintheafterlife,howcanweknowthatourintellectualbeliefsareright?Inbrief,al-Ghazālīcannotbeconfidentthathesuccessfullycastofftheinfluences of his upbringing. How could he know that his attacks against thephilosophers were justified, given that his own beliefs, like theirs, were heavilyinfluencedbycircumstancesbeyondhiscontrol?(Wewillconsiderhisownresponsestothisquestionin2.4.)

Thisworryisafamiliarone:youonlybelievethetenetsofagivenreligionorideology,because youwere raised as aMuslim, a Christian, or an atheist. Irrelevant influencesraiseaskepticalchallengethat is intimatelytiedtotheepistemologyofdisagreement.GiventhatmyreligiousbeliefsaretoalargeextentaresultoftheaccidentofwhereandwhenIwasborn,howcanIbejustifiedinholdingthem,especiallyonceIbecomeawareofreligiousdiversity?Thisquestionwillbethefocusofthischapter.Theepistemologicaldebate about such irrelevant influences focuses on uniqueness and permissivism.Uniquenessholdsthat,foragivenpropositionp,“thereisjustonerationallypermissibledoxasticattitudeonecantake,givenaparticularbodyofevidence”(White,2014,312).Permissivism, by contrast, allows for some leeway: there ismore than one rationallypermissibledoxasticattitudewecantakegiventheevidence(Schoenfield,2014).Inthischapter,Iwillexaminewhatirrelevantinfluencesareintheshapingandmaintainingofreligiousbeliefs,andwhethertheyraiseepistemologicalworriesforreligiousbelievers.Iwill argue that while we cannot escape these influences, we can let religiousdisagreement work to our epistemic advantage, both in philosophy of religion andamongreligiousbelieversmoregenerally.

2.2Howirrelevantinfluencesaffectreligiousbeliefs

Whenisaninfluencethatshapesourbeliefsirrelevant?KatiaVavova(inpress)proposesthefollowinggeneraldefinitionofirrelevantinfluences:

An irrelevant influence (factor) formewith respect tomybelief thatp isonethat(a)has influencedmybeliefthatpand(b)doesnotbearonthetruthofp.

Being raised in a particular religious community shapes one’s religious beliefs to animportantextent.Peopleinrelativelyreligiousnationsacquiremoreorthodoxreligiousbeliefscomparedtopeoplefrommoresecularnationswhoaresimilartotheminotherrespects,suchasgender,age,andeducation.Thisisinpartbecausewelearnreligiousbeliefs fromanearlyageon,butalsobecause inpredominantly religiousnations, thepoolofpotentialpartners,friends,orco-workerscontainsmorereligiouspeople.Peoplehavea tendency to adopt thebeliefsof thosearound them; this conformistbias alsoplaysa role in shaping religiousbeliefs.Buteven inmore secularnations,wearenotinsulated from irrelevant influences. In such countries, the beliefs of parents play animportant role in shaping their children’sbeliefs,whereas inmore religiouscountries,parentalinfluenceonthereligiosityoftheiroffspringisweaker,andinfluencefromthe

16

surroundingcommunityisstronger(KelleyandDeGraaf,1997).Thesecomplexpatternsindicatethatwhereandwhenoneisbornhavealargeinfluenceonthereligiousbeliefsonehas.

This is also true for philosophers of religion. In an open-question survey I conductedwith139philosophersof religion (DeCruz, inpress), I found thatnearlyhalf of them(43.9 percent)mentioned religious upbringing or education as a reason forwhy theybecame interested in philosophy of religion. Here is a representative answer to thequestion“Canyoutellsomethingaboutthefactorsthatcontributedtoyourspecializinginphilosophyofreligion?”byamaleassociateprofessoratasmallliberalartscollege:

I was raised Catholic and have a strong respect for that tradition. Thisrespecthasledmetobeinterestedinothertraditionsaswell.Atthesametime,asIhavestudiedphilosophy,Ihavebeenintriguedbyargumentsforatheism.Myparentsweredeeplyreligiousandintellectuallyengagedwiththeirfaith;thishassurelyhadaninfluenceonme.

Education,too,seemstohaveplayedanimportantroleinthereligiousbeliefsofmanyphilosophersofreligionwhotookpartinthissurvey.Inparticular,introductorycoursesinphilosophyhaveled,insomecases,tothelossofreligiousbelief.Hereisaresponsebyafemalefullprofessorataresearch-intensiveuniversity:

When I was a child I was a very committed believer and participant inChristianity. I gradually lost my faith, and the finishing element was asectiononphilosophyof religionwhen I tookan introductoryphilosophycourseinmyfirstyearatuniversity.Theshockwashugeand(believeitornot),Iwassomewhatsuicidal:IfeltInolongerhadanymeaninginmylife.Ithink,eversincethen,Ihavebeentryingtounderstandwhathappenedtome,andwonderingwhetherIreallyneededtoabandonmyfaith.

Ifshehadgonetoadifferentuniversity,presumablyonewithamorereligion-friendlyintroductoryphilosophycourse,shewouldperhapsnothavelostherfaith.ThisechoesaworrybyJerryCohen(2000)thattheplaceyoudecidetostudyatwouldhavea largeinfluence on your subsequent philosophical beliefs. Cohen chose to study at Oxford,ratherthanHarvard,forhisgraduatedegree.Atthetime,graduatestudentsatOxfordtendedtoaccepttheanalytic/syntheticdistinction,whereasthoseatHarvardtendedtorejectit.Thiswasnocoincidence,andnottheresultofpurelyindependentreasoningonthepartofthesegraduatestudents,buttheresultofwhattheyweretaught:

Ibelieve, rather, that ineachcasestudentswereespecially impressedbythe reasons respectively for and against believing in the distinction,becauseineachcasethereasonscamewithalltheaddedpersuasivenessofpersonalpresentation,personal relationship,andso forth.So, in somesenseof“because,”andinsomesenseof“Oxford,”IthinkIcansaythatIbelieve in the analytic/synthetic distinction because I studied at Oxford.

17

Andthatisdisturbing.ForthefactthatIstudiedatOxfordisnoreasonforthinkingthatthedistinctionissound.(Cohen,2000,18)

Similarly, in some sense of “because”, and in some sense of “Muslim education andupbringing”, al-Ghazālī was predisposed to think the philosophers’ arguments wereincoherent. In his Incoherence of the Philosophers, he rejected twenty of theirarguments,forexample,thattheycannotdemonstratethattheworldispre-eternal(i.e.,hadanoriginoutsideoftime).Thefalāsifabelievedthattheworld, likeGod,waspre-eternal, because it would be impossible for an eternal being to generate somethingtemporal.Al-Ghazālīarguedagainstthisclaim:Godisnotacause,likeothercauses,andwecanimaginethattheworldwascreatedatagiventime.Ifwecanimaginethat,itispossible, and an omnipotent being could have accomplished this (Griffel, 2017). Al-Ghazālī’s arguments are sophisticated—he used philosophical tools such as possibleworlds(whathecalled“alternativeworlds”)andmodallogic,whichwerederivedfromGreek philosophy. But if he had been raised in what he called an “un-Islamicatmosphere”andhadnotbeenfamiliarwithAshʿarism,hewouldlikely,attheveryleast,havebeenfriendliertowardsthosesameargumentshesoardentlyopposed.Thisseems,on the faceof it, troubling forphilosophersof religion, and foradherents to religiousbeliefs(includingatheists)ingeneral.

What gives irrelevant influences their sting? Identifying this has been a subject ofcontinued debate, especially on the question of evolutionary and cultural factors inshaping our moral beliefs. Below, I will outline two ways to spell out the worry:irrelevant influences make our beliefs unsafe (the safety objection), and irrelevantinfluencesmakeourbeliefs subject toanunacceptablearbitrariness (thearbitrarinessobjection).

2.3Thesafetyobjection

Perhapsweshouldbeworriedaboutirrelevantinfluencesonourbeliefs,becausetheymakeourbeliefsunsafe.IfIhadbeenborninanotherculture,oranothertime,Iwouldhavehaddifferentreligiousbeliefs.JohnStuartMillformulatedtheworryasfollows:

And the world, to each individual, means the part of it with which hecomesincontact;hisparty,hissect,hischurch,hisclassofsociety;...anditnever troubles him that mere accident has decided which of thesenumerousworlds is theobjectofhis reliance, and that the samecauses,whichmakehimaChurchmaninLondon,wouldhavemadehimaBuddhistoraConfucianinPekin.(Mill,1859,35)

Millmakestworelatedobservations,namelythatreligiousbeliefsofdifferenttimesandculturescontradicteachother,andthatbeliefsthatweregenerallyheldinpastcultures,orthatarebelievedbymostpeopleindifferentcultures,arewidelyseenaserroneous,evenabsurd,bypeoplewhoarepartofourcontemporaryculture.He formulates the

18

pessimistic induction thatbeliefswidelyheld todaywillbe rejected in the future.TheonlywayIcanthinkthatIamholdingtruebeliefsistoassumethatIamlucky,becauseit is easy to conceive that I would have ended upwith very different beliefs. TomasBogardus(2013,384)spellsoutthissafetyobjectionasapremise-conclusionargument:

1. If you had been born and raised elsewhere, else when, and formed religiousbeliefsusingthesamemethodyouactuallyused,then,byyourownlights,youeasilymighthavebelievedfalsely.

2. Therefore,yourreligiousbeliefswerenotformedsafely.

3. Therefore,yourreligiousbeliefsdon’tcountasgenuineknowledge.

One problem with this argument is that safety is a controversial requirement forknowledge. Agents sometimes seem to possess knowledge, even though theiracquisitionofitwasaresultofirrelevantinfluences.Let’slookatareal-worldexample,whereknowledge isunsafebut stillplausiblyknowledge.Belief inevolutionary theoryand in human-induced climate change is highly polarized in the United States.Whilescientists almost universally accept these beliefs, the general public is divided aboutthem.Themajorityofconservativesdisbelieveboth,andthemajorityofliberalsacceptboth.Morescientificallyliterateconservativesare,ifanything,evenmorelikelytorejectthescientificconsensus.Levy (inpress)arguesthat thispolarization isduetothewaypeoplegaugetestimony.Becauseweoftencannotcheckthecontentoftestimony,weneedtorelyoncuesthatsignalwhetherthe informant istellingthetruth.Thesecuescanbedividedintotwocategories:competence(toguardagainstlearningfrompeoplewhomakemistakes)andbenevolence(toguardagainstlearningfrompeoplewhowanttodeceiveus)(e.g.,Lane,Wellman,&Gelman,2013).Conservativesandliberalsusethesameheuristics,butbecausetheanti-sciencesentimentsinconservativemilieusaresorampant since the 1970s, conservatives, while accepting scientists as competent,perceivethemaslessbenevolentthanliberalsdo.Thus,liberalsareepistemicallyluckierthan conservatives about scientific matters. Because liberalism is science-friendly (atleast in theUS, andat least about these two topics), their “chainsofdeference tracebacktotherelevantscientificexperts,andthereforetoproperlyconstitutedcollectivedeliberation.” Conservatives are not so lucky: their chains of deference “end in‘merchants of doubt’ ... or maverick scientists” (Levy, in press). This sociologicalphenomenonmakesAmericanliberals’scientificbeliefsunsafe:iftheyhadbeenbornina conservative family or environment, theywould likely not have ended upwith thecorrectscientificbeliefs,buttheirbeliefsareplausiblystillknowledge.

Lookingatal-Ghazālī throughaMillian lens,heargued thatalthoughChristians, Jews,andMuslimsacquiretheirknowledgeinthesameway(throughdeferencetotestimony),only Muslims have access to proper, undistorted, divine revelation. If he had goodreasonstobelievethattheothersarelesslikelytoberight,thenitisnotaproblemthathewasluckytobebornintothecorrectreligiousatmosphere.Whileal-Ghazālīwasableto refute the views of the philosophers, using their own methods, and could thus

19

discreditthem,thisdoesnotautomaticallyvindicatehisownbeliefs.InDeliverancefromError,hearguedthatmysticalperceptionthroughSufipracticewasabletoprovidesuchvindication. But as we have seen in section 1.5, and will explore further on in thischapter,religiousexperienceisalsocoloredbythecommunitythatthemysticisbornin.Ifyouhavereasonstobelieveyouareright,thentheluckofbeingborninacommunitythatholdsthecorrectbeliefsdoesnotprecludeknowledge.

2.4Thearbitrarinessobjection

Another objection against irrelevant influences is that they seem tomake our beliefsarbitrary, inawaythat reduces (orperhapseveneliminates) their justification.This isthearbitrarinessobjectionagainstpermissivism(Simpson,inpress).Arbitrarinessposesa problem for both conciliationists and steadfasters. Here is the problem forconciliationists:accordingtopermissivism, therecanbe leeway inhowwerespondtoevidenceas longasweareusinggoodepistemic standards (Schoenfield,2014).But ifthat is the case, why should I revisemy religious beliefswhen Imeet someonewhoholdsadifferentsetofbeliefs?Aslongaswearebothusinggoodepistemicstandards,there isnopressure to revise.Yet,conciliationistshold thatbeliefsofepistemicpeersconstitutehigher-orderevidenceinthelightofwhichweshouldre-evaluateourbeliefs.Thusitwouldseemthatyoucannotbebothaconciliationistandapermissivist.

Butsteadfastersalsofaceaproblemiftheywishtoendorsepermissivism.Toseehow,imaginethatitistruethattherearemanyrationalwaystorespondtotheevidenceforwhetherornotGodexists.ThetotalevidencerelevantforGod’sexistence(E)contains,for example, the apparent fine-tuning of cosmological constants in a way that is lifepermitting, and human-induced and natural suffering. Suppose that Anjelica says, “Ithink there aremany rationalways to respond toE, but as far as I’m concerned, theevidence supports the Anselmianmonotheism, and I hold this beliefwith a credenceof.7.”HowcanAnjelicarationallyendorseherownbeliefaswellasallowingthatotherbeliefsarerational,e.g.,Miguel’s,whoisanatheist,andonlyholdsa.1credenceintheexistenceofanygod?8

Roger White (2014) likens cases where irrelevant influences play a large role in ourbeliefswithhaving to ingestabelief-togglingpillwhich randomlywould causeone tobelievenot-pwhereonebelievedpbefore(orswappingthecredencethatpof.7withoneof.1).Believingthatpwithacredenceof.7isthuspurelyaccidental;Imighthavebelieved otherwise. Is it rational to hold fast to this belief? The onlyway inwhich itmightbe,istoassumethatI’vebeenluckyiningestingtherightpillorgrowingupintheright community (White, 2014), and that assumption seems unwarranted. This8 The question of whether permissivists can also be conciliationists is a topic ofcontinuedinterestanddiscussion(seee.g.,Ballantyne&Coffman,2012).

20

motivatesuniqueness:thereisonlyonewaytorespondrationallytoabodyofevidence.The arbitrariness objection thus raises a challenge to any kind of religious belief. Itwould seem that suspension of belief is the rational response in the face of religiousdiversity, as we cannot know we were lucky in holding the right beliefs. Even amotivated agnosticism cannot be maintained, because this position would also bepresumably be the result of irrelevant influences. This is not a desirable epistemicsituation.

One way for the permissivist to respond is to argue that religious beliefs are notanalogous to pill-popping cases. The pill-popping example presents an isolated belief,whereasreligiousbeliefsarecloselyinterwovenwithotherbeliefsandaretheresultofmany different irrelevant influences (rather than just a single factor), e.g., family,education, friends, books one happens to read. The belief that God exists is oftenrelated tomany other beliefs, such as attitudes to abortion, gay rights, and politicalorientation.Thus,whereandwhenyouarebornisnotlikeingestingapillthatcanleaveyouwitharightormistakenbelief. It issomethingthatshapesyourwholeattitudetotheworld,includinghowyourespondtoevidence.

Thepermissivistcouldalsorespondthatwehavedifferentbackgroundconditionsthatmakeitbothrationalforustorespondtotheevidenceindifferentways,andtoseetheotherperson’sbeliefsasevidence.Severalauthorshavearguedalong these lines.Forexample, according to Robert Simpson (in press) one could be a permissivist about agivenquestionifthefollowingconditionsapplytomydissentingepistemicpeerandme:

1. Ourdifferentviewsareduetoushavingdifferentcognitiveabilitiesandapplyingdifferentepistemicstandards.

2. Theuseofourepistemicstandardsisoptimallytruth-conduciveforus,giventhecognitiveabilitiesthatwehavetoworkwithintheapplicationofourrespectiveepistemicstandards.

By cognitive abilities, Simpson (in press)means the abilities thatweput intopracticewhenapplyingepistemicstandards.Byepistemicstandards,hemeansmethodsagentsuse to assess the evidence they have gathered. For example, two detectives givedissenting assessments of the evidenceof a crime in a court of law.Onedetective isverygoodatgaugingwhethertestimonyisinsincere(ahumanliedetector),theotherisnotasgoodinthisbutsheisexcellentatholisticallyassessingthetotalevidenceatthesceneofacrime.Itdoesnotseemunreasonablefortheformertoplacemoreweightontestimony, and for the latter to place more weight on the total evidence. Suchconstraints are still quite stringent, but they would allow for permissivist attitudestoward religious and other beliefs. Because our background environment has such alargeinfluenceonthereligiousbeliefswehold,wewillendupwithdifferentcognitiveabilities and different epistemic standards with respect to religious beliefs. Take tworeligious experts who disagree about the nature of God. One is trained in religiouspractices fromanearlyageon,especiallypractices tocultivateaweandwonderas inJudaism (Wettstein, 2012), and accords much weight on experiential evidence. His

21

colleagueisabiblicalscholarwhohasadeepunderstandingofscripturaltexts,butsheisnottrainedinsuchpractices.ItdoesnotseemunreasonableforhertoaccordmoreweighttotheevidenceofwrittensourcesthantotheoccasionalexperientialevidenceofGod’s existence that shehas. Thebestwe cando isworking from theseepistemicstandardsandusingthesetoolstocometowell-grounded,justifiedbeliefs.

Canaresponsealongtheselinesworkforal-Ghazālī?Hisviewswereprofoundlyshapedby his own upbringing and education, but the philosophical tools he acquired whilestudyingunderal-Juwaynīdidallowhimto formulatenovelobjectionsagainstMuslimphilosopherswhoratheruncriticallyacceptedAristotelianargumentsfortheeternityofthe universe and other claims. It should also be noted thatwhile al-Ghazālī’swork isstrongly influenced by Ashʿarism, he does differ from this school in subtle respects(Frank, 1994). Take his notion of taqlīd, and its subtle differences with how theAshʿarites dealt with it. Like Ashʿarite authors, al-Ghazālī saw a tension betweenuncritical acceptance of the testimony people hold in esteem (taqlīd) and religiousknowledge. The Ashʿarites believed that testimony, by itself, could not lead toknowledge.Theyconsideredintuition,perception, inference(inductivereasoning),andformalreasoning(deductivereasoning)assourcesofknowledge.Bycontrast,accordingtoal-Ghazālī taqlīdcould leadone tohold the rightbeliefs. Inhis Incoherenceof thePhilosophers, he argued that accepting testimony from proper sources, such as theprophets (as recorded in the Qurʾān and ḥadīth) could lead to knowledge, but thatknowledgeobtainedthroughcriticalreflectionissuperior(Griffel,2005).

The charge against the falāsifa can thus be refined: al-Ghazālī saw their uncriticalacceptanceofancientGreekphilosophyasaproblem,becausetheythemselvesprizedtheirreasoning,cleverness,andwit.However,iftheyweretogiveuptheirtaqlīd,theywouldcometorealizethatthesetransmittedbeliefsarewrong,andwouldcometothetrueviewsoforthodoxIslam.Thus,acomplexpictureemergeswhereal-Ghazālīseesitasincumbentuponpeoplewhoareabletocriticallyreflecttoevaluatetheirbeliefs,andseewhethertheyholduptoscrutiny. Ifweapplythistophilosophersofreligionortoanyreligiousbelievers,itisinevitablethattheyareshapedbythereligioustraditionsoftheir birth, but that does not prevent them from reflecting upon their beliefs. Theyshoulddo this, especially if an interlocutor (perhaps coming fromadifferent religioustradition) offers them reason to doubt their views. Indeed, itwould be hubristic andintellectuallydishonestnottodoso.Butif,onreflection,theiroriginalbeliefsholdup,theirjustificationincreases.

Thepeculiar resultof thisapproach is that itworksregardlessof thecircumstances inwhichyouwerebroughtup.Thetotalbodyofevidenceforanyreligiousposition(e.g.,the existence of God) is fiendishly complex and large. Most people will only have apartial grasp of E. But if they do the best they can, they can have justified religiousbeliefsregardlessoftheperspectivetheycomefrom.Inthispicture,aphilosopheroranordinarybelieverwhocriticallyandhonestlyexaminesarguments for theexistenceofGodcanobtainahighdegreeofjustification,ascanacontemplativewhoseeksoutand

22

cultivates theirmind and habits in order to gainmeaningful religious experiences, assomeonelikeTeresaofÁvilaandal-Ghazālīrecommended.

2.5Makingreligiousdiversityworkinphilosophyofreligion

One may ask whether the strategy al-Ghazālī recommended for reflective thinkersworks. Belief polarization is a well-attested collective phenomenon whereby peopletendtodiverge,ratherthanconvergeinthefaceofdisagreement.Severalbiasescausethis phenomenon, including confirmation bias, which leads one to evaluate morepositively evidence in line with beliefs one already holds. Even highly-educatedreflective thinkers are unaware of their own biases. For example, participantsconsistently think that they are less susceptible tobiases than the averageAmerican,theirclassmatesinaseminar,ortheirfellowtravellers(Pronin,Lin,&Ross,2002).Giventhebiasblindspot,thereisnowaythatal-Ghazālī(ca.1100[1952],21)canconfidentlystatethat“thebondsofmereauthority(taqlīd)ceasedtoholdmeandinheritedbeliefslosttheirgripuponme,”becauseitishardtoknowhowbiasedone’sownreasoningis.Moreover, in Deliverance, al-Ghazālī became more pessimistic about reasoning as asourceofreligiousknowledgeandarguedthatmysticalexperience(throughpracticingSufism)wouldprovideknowledge.However,mysticalexperiencetendstovarywiththereligioustraditionsinwhichtheyoccur,soit isnotclearthatknowledgebasedonthiswould escape the arbitrariness objection. It is no coincidence that al-Ghazālī hadmysticalexperiencesthatsupportIslam,whileTeresaofÁvilahadexperiencesthatwereinlinewithCatholicism.

If individualdebiasingisdifficultandonlyhasmodestsuccess,onewaytoaddresstheproblemofirrelevantinfluencesistoletreligiousdisagreementworkproductively.Wecan do this by actively encouraging or cultivating religious diversity, especially in areflective forum such as philosophy of religion. This strategy has been proposed inphilosophy of science (Kitcher, 1990; Longino, 1991). These authors have differentmotivations for why encouraging diversity of opinion and disagreement may bepracticallyrational.Theirrecognitionofscientificpracticeasacollectiveendeavorformsacommonelement.PhilipKitcher(1990)seesamismatchbetweenwhatisrationalforanindividualscientisttobelievecomparedtowhatisrationalforscientistsasagrouptopursue. Sometimes it is conducive to scientific progress that individual scientists holdunpopularormaverickviews,evenifthesearenotaswellsupportedbytheevidenceasthemainstream view, for example plate tectonicswas aminority view in geology foroverhalfacenturyuntil itwasvindicatedbygeomagnetism,biogeography,andotherobservationsinthe1950sand60s.Agentswhoaredrivenbymotivesthatarenotpurelytruth-orienteddobetterthanagentswhoareonlyinterestedinthetruth.Theoptimalepistemicsituationisoftenonewhereagentscombineasearchfortruthwithlesspuremotives, suchas aquest for fame,or trying toobtain a result quickly.Applied to thephilosophyofreligion,someauthorshaveworriedthatphilosophersofreligionhaveanagenda, i.e., they have less puremotives. Theywant to proselytize, and are perhaps

23

morebiasedthanphilosophersinotherdisciplines(Draper&Nichols,2013).ButthisisnotaprobleminKitcher’sapproach:itisfineifphilosophersofreligionhaveanagenda,aslongastheyarenotalldrivenbythesamemotives.

HelenLongino(1991)hasarguedthatevenifscientiststrytoremainunbiased,theywillstill exhibit biases of their research program, education, gender, race, and class.Contextualvalues,whicharenotdirectedat the truth,will inevitablyplaya role.Onewaytosolvethisproblemisthroughstandpointepistemology—apowerlesssubjectcanhave privileged knowledge because of her social situation (for example, she can seehurdles or solutions invisible to others). But this will not suffice to eliminate biases.HenceLonginorecommendstheactivecultivationofdiversepointsofview,asharingofintellectual authority, which ultimately—through productive dialogue—facilitatestransformativecriticism.Applied tophilosophyof religion, thiswouldmeantheactivecultivation of viewpoints that are not mainstream. Given the preponderance ofChristianity in the discipline, it could benefit from the cultivation of Muslim, Jewish,Mormon,Wiccan,Shinto,andotherpointsofview.Moreover,giventhatphilosophyofreligionisstilloverwhelminglymaleandwhite,theparticipationofmarginalizedvoicessuchaswomen,genderqueerindividuals,andnon-whites,wouldnotonlyallowforthesharingof intellectualauthority,butalsofor insightsonsuchtopicsastheatonement,freewill,orthenatureofGod,thatwillbecoloredbyintersectionalcharacteristics.

Wecannowseehow irrelevant influences canplaybothaproductiveandanegativeroleinthephilosophyofreligion.Wecannotovercomeourbiasesthatarearesultofalifetime of socialization and experiences—even ifwe try our very best, like al-GhazālīandDescartesdid.Butwith theparticularcognitivemindsetswehave,wecandoourbest to arrive at true beliefs. This allows for a permissivist picture of religiousdisagreement.Atpresent,academicphilosophyofreligionispredominantlywhite,male,and Christian. This poses constraints on the ideas that can be generated, leading theproblemofunconceivedalternatives (posedbyKyle Stanford [2006] as aproblem forscientificrealism).Stanfordobservesthatscientificcommunitieshaverepeatedlyfailedto come up with reasonable alternatives to dominant scientific theories, althougherstwhile unconceived alternatives are later confirmed by evidence and adopted byscientificcommunities.Thesamecanbearguedforphilosophyofreligion,whichleaveslargeswathesofconceptualspaceunexplored.

Evenifweassumethatatleastsomeoftoday’sreligioustraditionscapturemeaningfulparts of religious reality, thoughtful dialogue would be to our advantage. Take JohnHick’s (1988) argument for religious pluralism: the world is religiously ambiguous, itdoesnotclearly favoronereligiousworldviewoveranother,and itcanbe interpretedreligiouslyornon-religiously.HickseesparticularreligiousconceptssuchasAllāh,Viṣṇu,or the triuneGod, as phenomenal interpretations of a noumenal religious reality. Heuses theparableof theblindmenand theelephant,whereeachblindperson feels adifferentpartoftheelephant(e.g.,thesmoothtusks,theflexibletrunk,thesturdylegs)andjudgewhattheyarefeeling.Giventhatourunderstandingofthereligiousdomainisshaped to a large extent by irrelevant influences, we will have a particular religious

24

outlook.Butthisdoesnotmeanthattheopinionsofothersholdnoweight.Ifweacceptthatthereligiousdomainisindeedcomplexandambiguous—withoutperhapsgoingasfar as to endorseHick’s Kantian distinction betweennoumenal and phenomenal—wecan learn something from the religiousbeliefsofothers.While theaccidentofwhereandwhenwearebornposes serious constraintsonourown religiousbeliefs,wecanlearnfromtheviewpointsofthosewhoaredifferentlysituated.JeromeGellman(1997)suggestsadifferentwayoflookingatthis.ApparentlycontradictoryexperiencesofGodpointtoaninexhaustibleplenitudeonGod’spart,forexample,Godisbothpersonalandimpersonal,hencehe,she, they, itareexperiencedbysomeaspersonal (e.g., theisticreligions)andbyothersasimpersonal(e.g.,Daoism).

To conclude, while the accident of where and when we are born leads to thearbitrarinessobjection,thisdoesnotnecessarilymeanwehavetogiveupanyarticulatepositiononreligiousmatters.Asamatteroffact,wecanusereligiousdiversitytoourepistemicadvantagegiventhatintellectualdiversityhelpscommunitiesoftruthseekersto obtain beliefs that aremore truthful andmore justified.Given the complexity andambiguityofthereligiousdomain,suchdiverseperspectivessheddifferentlightsonthequestionofGod’sexistenceandrelatedmatters.

Chapter3:Conversionanddisagreementwithformerselves9

3.1Religiousconversionandtheindependenceprinciple

MiguelandCatherinearebothatheistphilosophersandlong-timefriendsfromgraduateschool.MiguelrespectsCatherineandbelievessheishisepistemicpeer.Bothofthemhavebeenraisedinaliberalenvironmentwhereatheismisquitewidespread.Buttheyhavealsocarefullythoughtabouttheism.Neitherofthemisaphilosopherofreligion,but they know the arguments for and against theism and they both agree that theargumentsagainstoutweightheargumentsfor.Asaresult,botharequitecomplacentatheists:theythinkthereisnoseriouscasetobemadefortheism.

MiguelmeetsCatherineataphilosophyconference—it’sbeenaboutayearsincetheylastmet.Hehasbeenlookingforwardtoseeinghisoldfriend.Astheycatchupoveracupofcoffee,itbecomesclearthatCatherineisnowatheist.

“But why?,” Miguel asks, “Didn’t you think all that theist philosophy is dubiousmotivatedreasoning?”

“Ichangedmymind,”Catherinereplies.

9This chapter is adapted from “Religious conversion, transformative experience, anddisagreement”,PhilosophiaChristi(inpress).

25

This situation appears to be a case of peer disagreement. Miguel and Catherine are(presumably)equallythoughtfulinthedomaininquestion,andtheyhaveaccesstothesameevidence.Whiletheirepistemicsituationisnot identical,theywillstillbesimilarenoughtotakeeachothertobeepistemicpeers,asituationthatoccursfrequently ineverydaylife.Aswehaveseeninchapter1,epistemologistshavedebatedwhetheroneshould conciliate, that is, revise one’s opinion that p, when facedwith a disagreeingpeerwhoholdsthatnot-p,orwhetheronecanremainsteadfast.

Oneprinciplethathasbeenproposedtoseparatethesetwotypesofresponses istheindependence principle. Recall, this principle says that in evaluating the epistemiccredentialsofsomeoneelse’sbeliefaboutp Ishouldnotrelyonmyreasoningbehindmy initialbeliefaboutpwhendeterminingwhetherandhowto revisemyownbeliefaboutp. There are counterexamples to the independence principle,where one has ahighjustifiedbeliefthatp,andwhereitdoesseemthatonecanrelyonone’soriginalreasoning. If someone trulybelieved that2 +2=5, thiswoulddemonstrate that thispersonis,inalllikelihood,notasclear-headedasyou.

Would a religious conversion be more analogous to mental math or more likeelementarymath?Manypeoplehavearesponsealongtheselines:“Idon’tknowwhathappenedtoJosh.Hewassuchareasonableguy,butthenhejoinedScientology.Ithinkhe’s gone off the cliff”, or, “I used to respect Amy, but now she is no longer anEvangelical Christian, I cannot respect her anymore.” In such cases, people treat theconversionasasituationwhereone’sepistemicpeerhasacquiredsuchanoutlandishsetofbeliefsthattheybegintodoubtthepeerhood,akintoelementarymath.Inothercases,where thenewbelief is at least a liveoption (in the Jamesian sense), the casemay appear more like mental math (e.g., “I’m not Jewish, but it does look like aninterestingreligionwitharichtheology.AmymightberightforallIknow,orImightberight.”)Thereisafurtherproblemfortreatingconversioncasesaspeerdisagreements:religious conversions tend tobe transformative.Thismakes ithard toassesswhethertheperson is still one’s epistemic peer, orwhat theproper response should be.As aresult of this transformation, the beliefs of a dissenting peer, arising from a religiousconversion,donothavestraightforwardevidentialvalue.

3.2Conversionasatransformativeexperience

Someexperiencestransformus,bothinwhoweareandwhatweknow.L.A.Paul(2014)has termed such experiences “transformative experiences”. They are personallytransformative,inthattheychangewhoyouareasaperson,andtheyareepistemicallytransformative, in that they give you new information by virtue of a particularexperience thatcanonlybeobtainedbyhaving thatexperience.Paulhasarguedthatwe cannot rationally make decisions that will transform us because there is anasymmetrybetweenwhowearenow,andwhatweknownow,andwhowewillbethen,andwhatwewillknowthen.Wecannotmakeacomparisonbetweenbeforeandafter

26

to calculate the expected utility, for example, adopting a child may be a wonderfulexperienceformyfriend,butImaycometoregretit.Yet,weoftenmakelife-changingdecisions:choosingasubjectof study,gettingmarried,or joininga religious tradition.Notalltransformativeexperiencesaretheresultofconsciousdecisions.Becominglong-termillordisabledisasituationfewpeoplechoose,yetittransformswhattheyknow(e.g., what it is like to live in the face of imminent death), and who they are (e.g.,someonewho became blind). The transformative nature of disability is illustrated byhowdisabledandnon-disabledpeoplethinkabouttheirqualityof life:sightedpeopletend to think being blind is a lot worse than blind people think it is, including blindpeoplewhowerepreviouslysighted(Carel,Kidd,&Pettigrew,2016).

Isreligiousconversiontransformative?Saul/Paul’sconversiontoChristianityontheroadto Damascus suggests that a single transformative event can lie at the basis ofconversion.AsrecordedinActs9(inathird-personnarrative)andthePaulineepistlessuch as Galatians 1 (in a first-person narrative), Saul of Tarsus was a Jew whopersecuted the early Christian community. On his way to Damascus, where he wasextending his mission to have Christians imprisoned, he had a dramatic religiousexperience:hesawablindinglight,andheardthevoiceoftherisenChrist,whoaskedhimwhyhewaspersecutinghim.Saulthengotbaptized,andstartedmissionaryworkfor the nascent Christian church. This account of conversion in terms of a singledramatic experience shaped subsequent discussions of what conversion is like, forinstance,intheworkofearlypsychologistssuchasJames(1902).Subsequentresearchintoreligiousconversionssuggestsamoregradualpattern.Conversionsarenotsingleexperiences,butratherthecumulativeeffectofmanysmallerexperiencesanddecisions.Evenamongpeoplewhoself-identifyasborn-againChristians,themajorityexperienceagradual,ratherthanasudden,conversiontoEvangelicalProtestantism(Dixon,Lowery,& Jones, 1992). Nevertheless, religious conversions transform a person’s system ofbeliefs(thusareepistemicallytransformative),andchangetheirpersonality.Thisisnotso much a change in personality traits, as it is in goals, feelings, attitudes, and lifemeaning(Paloutzian,Richardson,&Rambo,1999).

Giventhataperson,post-conversion,isbothepistemicallyandpersonallytransformed,itbecomesdifficulttogaugewhetheraformerepistemicpeerisstillanepistemicpeer.Forexample,JonathanEdwardshasanextensiverecordofthedeepeningofhisfaithinseveralphases:

... my sense of divine things gradually increased, and becamemore andmore lively, and hadmore of that inward sweetness. The appearance ofeverythingwasaltered:thereseemedtobe,asitwere,acalm,sweetcast,or appearanceofdivineglory, in almosteverything.God’sexcellency,hiswisdom,hispurity and love, seemed toappear ineverything; in the sun,moon,andstars,inthecloudsandbluesky,inthegrass,flowers,trees,inthewaterandallnature;whichusedgreatlytofixmymind(Edwards,1821,xxv).

27

HowcanweknowthatEdwards,post-conversion,isinanepistemicallybetterpositionthanbefore?HowcanEdwardshimselfknowthatthisisthecase?

3.3Disagreementwithone’sformerself

Conversion thus gives rise to two forms of disagreement—disagreement with one’sformer self, and disagreement with friends and family. Let’s look at someone likeCatherine,who is now indisagreementwithher former self. Shenowbelieves thingsshewouldhavefoundimplausiblejustayearago.Shouldsheaccordmorecredencetohernewbeliefs?Ifconversionweresolelytheresultofacareful,rationaldeliberation,itwouldseemproperthattheconvertismoreconfidentaboutherbeliefspost-conversion.However,conversionoccursasaresultofseveralfactors.Theseincludethedesiretobeof the same religious tradition as one’s family and friends, or the desire for self-improvement and for having a religion that ismore in linewith one’s self-image. Forexample,inaqualitativestudyofFrenchyoungadultswhoconvertedtoIslam(Lakhdar,Vinsonneau,Apter,&Mullet,2007),severalparticipantsmentionedthattheybelievedregular prayer and observing food taboos and the Ramadan would help them to bemorepersonallydisciplined.Suchconcernsmaybepractically rational,butdonotsaymuchaboutthetruthorfalsityofthebeliefstheyadopted.

The influential Lofland–Stark model of religious conversion (Lofland & Stark, 1965)outlinestwobasicconditionsthatpromptreligiousconversion:first,peoplemustfeelanacute need or tensionwithin their religious problem-solving perspective, and second,theyformaffectivebondswithmembersofthenewreligion/affiliationtheyconvertto,whichfacilitatesthesteptothenewreligion/affiliation.Thismodelhasattractedmuchattentioninthesocialstudyofreligion.Morerecentsociologicalaccountsofconversion(e.g.Kox,Meeus,&Hart,1991)identifytwotypesofattractionofreligiousmovementstonewconverts: ideological,throughofferinganewperspectiveonlife,andsocial,byprovidingasatisfactorysocialnetwork.

Thus, conversion seems tobe to a largeextent the result of irrelevant causal factors,suchasthepresenceofsocialgroupsthatonemightfeelathomewithandseekcloseraffiliationwith.However,one’soriginalreligiousbeliefs(orlackthereof)arealsolargelytheresultof irrelevant influences.Aswehaveseen inthepreviouschapter, there isacontinueddebateonwhether irrelevant influencescanactasdefeaterstoourbeliefs.This debate often centers on the question of whether our response to the evidenceallowsforsomelatitude,aspermissivistspropose,orwhetheratotalbodyofevidenceonlyallowsforonerationalattitude,asproponentsofuniquenesshold.

If we assume that sociologists are right and that most conversions are caused by amixture of social and personal factors, few ofwhich are relevant to the truth of thebeliefs, how should a evaluate her past self’s beliefs? Vavova (in press) recommendsthat inordertoevaluatetheepistemicsignificanceof irrelevant influences,welookatwhatcausestheseinfluences.Iftheinfluencesgive“goodindependentreasontothink

28

that you are mistaken with respect to p, you must revise your confidence in paccordingly—insofarasyoucan.”Forinstance,ifabeliefistheresultofbrainwashingorwishful thinking, this does seem to present substantial higher-order evidence thatundermines the belief. Extreme cases of conversion that involve indoctrination orcoerciveforcewouldthusbecastinanegativelight,asindoctrinationandcoercionareingeneralpoorbelief-formingmechanisms.Butformanyothercases,whichinvolvethetypical mixture of social factors and personal motivations, it is not so clear whethertheseirrelevantinfluencesprovidetheconvertwithgoodindependentreasonsthatsheismistakenwithrespecttohernewbeliefs.Thefactorsunderlyingconversioncasesdonot seem to bemore epistemically vicious or benign than factors underlying originalreligiousbeliefformation(e.g.,parentalreligiousaffiliation).

Areligiousconverthasonepieceofhigher-orderevidencethatsomeonewhodoesnotconvertdoesnotpossess,namelyfirst-personalexperienceofchangingone’sreligiousbeliefsinadeepandsignificantway:theconvertknowsthatherreligiousbeliefscanbechanged10. This might lead converts to become more accepting and open to thereligious views of others—after all they had different religious beliefs prior to theirconversion. If Catherine nowholds beliefs she found implausible just a year ago, thistells her something meaningful not just about her current religious beliefs, but alsoabout the fragility and revisability of religious beliefs in general. Thus it would seemrational for a convert to remain open to the views of dissenting epistemic peers. Bycontrast,Migueldoesnothavetherelevantexperiencesthatpresumablywereatleastin part the basis of Catherine’s conversion. Given that he has not, he should remainopen to the possibility that Catherine has relevant evidence that he lacks. Thus,conciliationism seems a rational response in the light of the conversion of a formerepistemicpeer.Iwilllookinmoredetailatthisquestioninthenextsection.

3.4Disagreementwitharecentlyconvertedepistemicpeer

Iwillnowlookatthequestionofwhat(ifany)epistemicconclusionsMiguelcandrawfrom Catherine’s conversion, in the absence of further information (let’s assumeCatherinehadtodashtotheairport,soMiguelnevergetstohearherreasons).Shouldit lower his confidence in his complacent atheism? One could argue that Catherine’sconversiondoesnotprovidehimwithanynewinformation.Miguelispresumablyawareof the distribution of opinions about theism. He knows that most of the worldpopulation (about 85-90 percent) are theists, most of whom are monotheists(Zuckerman, 2007), and that a substantial number of academics (albeit less than 50

10This is in addition to any first-order evidence that prompted the conversion, e.g.,Paul’sreligiousexperienceontheroadtoDamascus.Notethatnotallconversioncasesinvolvesuchfirst-orderevidence;somemaybepurelytheresultofsocialfactorsanddonot offer any first-order evidence, but then the convert still has the second-orderevidenceofhavingconverted.

29

percent)areatheists(Gross&Simmons,2009).Hemightalsoknowthat73percentofacademicphilosophersareatheists (Bourget&Chalmers,2014)11.SinceMiguelsharesmostofhisbackgroundbeliefswithacademicphilosophers,hewouldprobablyconsiderthemasepistemicpeers.Miguelcouldmaintainhiscredencesaccordingly.Theredoesnotseemtobeanyspecial informationgainedbyanold friendhavingconvertedtoabeliefhedoesnotshare.

However,Miguelhasacquiredanewpieceofinformation:someonewithwhomhehadmany background beliefs in common, and whom he has always respected as hisepistemicpeer,changedhermindonthequestionoftheism.IfheconsidersCatherineto be similar to him in many respects (e.g., shared graduate school experience,philosophers they both admire), they are similar in relevant background knowledge.ThusitwouldseemepistemicallyprudentforMigueltoatleastfollowupwithCatherine(e.g.,asimplee-mailsayingsomethinglike,“IhavetoconfessIwasabitsurprisedthatyou’reatheistnow.Couldyoutellmemoreaboutit?”).Afterall,shemayhavereasonsthathehasnotproperlyconsidered.

Thereisanotherreasonwhytheconversionofafriendwhowas(atleastpreviously)anepistemicpeerissignificant:wetendtoattachmoreweighttothetestimonyofthosewhoareclosetousthantothetestimonyofstrangers.Maybethis isbecausewecangaugetheepistemiccredentialsoffamiliarindividualsbetter.Butfamiliarityalonedoesnotexplain theextraweightweaccord to thosenearanddear tous.Afterall, if thatwere the case we would put more stock in people who are former friends, or infrenemies,andthisdoesnotseemtobethecase(indeed,wetendtobemorecautiouswhenitcomestobothcategories).Beingfriendsengendersepistemicpartiality(Stroud,2006).Toddlersalreadyshowatendencytovaluethetestimonyofpeopleclosetothemmore thanthe testimonyofstrangers (Harris&Corriveau,2011).Thisheuristicmakessenseinthecontextofepistemicvigilance:sincewenotonlyneedtosortoutaccuratefrominaccuratetestifiers,butalsothosewhoaretruthfulfromdeceitful,itmakessensetotrustpeoplewhoarewelldisposedtowardus.Thuspeoplehaveatendencytoplaceselectivetrustinpeopletheyseeasbenevolent:peoplewithwhomwehaveamutuallytrustingrelationshiparelesslikelytodeceiveus(Sperberetal.,2010).

Trustinfriendsmayalsobevaluablebeyondpurelyepistemicreasons,justlikeself-trustis intrinsically valuable (Pasnau, 2015). Intellectual theorizing should not be seen aspurely dissociated from our emotional lives: sometimes it makes sense to have ourbeliefs accord with those of our friends. This indeed often happens in the case ofconversion,wherepeopleconvert to the religious traditionof friendsor relatives.Forsuch reasons,Miguel should bemore diligent in following up Catherine’s reasons forconverting. And perhaps likewise, Catherine should be diligent in following up withMiguel to explain her reasons for converting. I will now look at a case study ofconversion that illustrateswhy reasoned debate is the proper response to a friend’s,andone’sown,conversion.11butseeDeCruz(2017),where50.2%ofsurveyedphilosopherswereatheists.

30

3.5Rationalargumentandconversion

TheAfrican theologianAugustineofHippo (354–430)converted fromManichaeismtoChristianity. In both religious traditions, he was an apologist, a teacher, and anevangelist.His letterDeUtilitateCredendi (Ontheusefulnessofbelief) isaddressedtoHonoratus,astudentfriend.AugustineconvertedhisfriendtoManichaeismwithmuchdifficulty,andinthislettersethimselftheunenviabletasktotrytoconvincehisfriendtoconverttoChristianity.ButwhyshouldHonoratus,whomhepreviouslypersuadedofthemeritsofManichaeism,nowtrustAugustine?

For Augustine to achieve his purposes, Honoratus would have to bepersuaded of Augustine’s trustworthiness while discounting Augustine’smuchearlierevangelisticcampaignsasaManichee.HonoratuswouldalsohavetograntthatAugustinehimselfhadnotbeendeceivedasecondtimeasheappearstohavebeenmisledthefirsttimewhenhewaspersuadedbytheManichees.(Asiedu,2001,128)

The letter is revealing in that Augustine does not dwell on his own conversionexperience,ashedoeselsewhere,e.g., inhisConfessions(4thcenturyCE[1961]),asasource of knowledge. Rather, the letter focuses on biblical hermeneutics (thediscrepanciesbetween theOldandNewTestament,whichManicheansoftenpointedoutasreasonstorejectChristianity)andonhigh-levelepistemicprinciplessuchastrustandcredulity.The letterexamines the reasonsManicheansgave for rejectingCatholicdoctrines,andarguesthatthesereasonsweremistaken.Inthisway,AugustinearguesthatHonoratus’ (andhis formerself’s) reasons foracceptingtheManichaeandoctrineweremistaken:

Well, they harangued at great length and with great vigor against theerrorsofsimplepeople,whichIhavesincelearnedisextremelyeasyforanyonetodowho ismoderatelyeducated;and if they taughtusanyoftheirowndoctrineswethoughtwemustmaintainitbecausenothingelseoccurredtoustosetourmindsatrest.(Augustine,5thcentury[1953],i,2,292)

The passage in viii, 20 recounts Augustine’s own faith journey (what born-againChristianswouldcalltheir“testimony”).Hefirstrevealsthathewasalready“inastateofseriousdoubt”aboutManichaeismwhenhelastpartedfromhisfriend,andthathisdoubt grew evenmore after he saw the underwhelming performance by the famousManicheanFaustus.“Youremember,hiscomingtoexplainallourdifficultieswasheldouttousasagiftfromheaven.Well,Irecognizedthathewasnobetterthantheothersofthesect,exceptforacertaineloquencehehad”(Augustine,5thcentury[1953],viii,20, 306). The account of his own conversion anticipates the later fuller testimony inConfessions,especiallybooks5and6.WhileAugustinedoesnotusehisownconversionas a source of evidence to Honoratus, he nevertheless presents it as a model that

31

Honoratuscanemulate.Forexample,hedrawscloseparallelsbetweenhisownformerdoubts (now happily resolved), and Honoratus’ present doubts. This echoes theconciliationist position in theepistemologyof disagreement.Recall, the conciliationistpositionholdsthatbeliefsofotherswerespectasourpeersprovidesignificantevidence(inpartbecauseourpeersmayhavereasonsorargumentswemaynothaveconsideredproperly),andcallforbeliefrevision.

Theletter(particularlysectionsvii–xi)alsodevelopsanintricatephilosophyoftestimony,where Augustine asserts that all knowledge must begin in trust of those who haveproperauthority, ratherthan inreason.Hegivestheexampleof thetrustweplace inourparents:

How will children serve their parents and love them with mutualdutifulness iftheydonotbelievethattheyaretheirparents.Thatcannotbeknownbyreason.Whothefatheris,isbelievedontheauthorityofthemother,andastothemother,midwives,nurses,slaveshavetobebelieved,for the mother can deceive, being herself deceived by having her sonstolenandanotherputinhisplace.(Augustine,5thcentury[1953],xii,26,313)

To Augustine, the chief problem with Manicheans is that they do not recognize theimportanceoftrustintestimonyofthosewhohaveproperepistemicauthority.Insteadtheyholdupamirageofhowweshouldacquirebeliefs:“theypromisetogivetothosewhomtheyattractareasonevenfortheirmostobscuredoctrines”(ix,21).Thus,theyare not being intellectually honest with their adherents and converts, because it issimplynotpossibletogivereasonsforobscuredoctrines,withoutresortingtotrust inauthoritative testifiers12. I will return to the topic of epistemic authority in the nextchapter.

In De Utilitate Credendi Augustine sees testimony merely as a basis of belief. In hisConfessionshealsoregardsitasabasisforknowledge(King&Ballantyne,2009).Yet,intheformerheclearlyseestestimonyasapreconditionforknowledge,asmanypeoplewould lack the relevant reasoning skills to work out religious truths for themselves.Augustine draws an illuminating parallel between friendship and trust in testimony.Against those who do not think that testimony can be a useful source of belief,Augustineargues

Idonotseehowanyonewhoacceptsthatastruecaneverhaveafriend.For if to believe anything is base, either it is base to believe a friend, orwithout such belief I cannot see how anyone can go on speaking aboutfriendship.(Augustine,5thcentury[1953],23,309)

12Thisholdsalso for thosewhostudythesciences.Atsomepoint,oneneedsto trustthose with authoritative knowledge, as it is impossible to experimentally verifyeverythingforoneself.

32

Whileitisnotexplicit,itisclearthatAugustinehereisappealingtohisfriendtoreadhisargumentswithmorecharityandpatience thanhewould read thearguments fromaCatholic writer he does not know, with the charity one owes a friend. The tonethroughout the letter underlines this point, with Augustine frequently expressingunderstandingatwhatmustbeasurprisetohisfriend,forexample,“Youareamazed,Iamsure.ForIcannotpretendthatIwasnotformerlyofaverydifferentopinion”(vi,13,301);hefrequentlyanticipatesobjectionsHonoratusmighthave,suchas,“Butyouwillprobablyasktobegivenaplausiblereasonwhy, inbeingtaught,youmustbeginwithfaithandnotratherwithreason”(ix,22,308).

Overall,DeUtilitateCredendishowsagoodmodelofpeerdisagreementinthefaceofconversion. Augustine’s aims in the letter aremodest. As he points out in his closingparagraphs, he does not refute the Manichean doctrines or defend Catholic ones.Rather, he aims to show that the arguments that Manicheans have offered againstCatholic doctrines (namely, alleged discrepancies between the Old and NewTestaments) are not as strong as he previously believed. He also argues thatManichaeismdoes not live up to its own standards of belief based on reason, ratherthantestimony—astandardthatAugustinealsodemonstratestobeunattainable.Alongthe way, Augustine can also demonstrate to himself (as he also did at length in hisConfessions)thathe isnowinabetterepistemicpositionthanhispre-conversionself.Even though Augustine may not be justified in thinking he now holds the correctdoctrines, his new beliefs are based on more sophisticated grounds, whereas hispreviousreasonsforrejectingCatholicdoctrinesprovedtobeinadequate.

3.6Thoughtfuldisagreement

Thecaseofreligiousconversionpresentsaseriesofdifficultiesfortraditionalaccountsofepistemicpeerdisagreement,becauseconversionisatransformativeexperience:itisdifficult todecidewhetheraconvert is inabetterepistemicpositionpost-conversion.Conversion is rarely the outcome of a rational decision process, but the result ofirrelevant influences such as personal tension and the religious views of friends. ThiswasalmostcertainlythecaseforAugustine,aswecansee inpassagesthroughoutDeUtilitateCredendi,forexample,

WhenIdepartedfromyouacrosstheseaIwasalreadyinastateofseriousdoubt;whatwas I to hold;whatwas I to give up? Indeedmy hesitationgrewgreaterdaybyday from the time that I heard the famous Faustus.(Augustine,5thcentury[1953],viii,20,306)

Most casesof conversionarenot causedby forces thataremorepernicious than theonesthatgaverisetotheoriginalsetofbeliefs(e.g.,beliefsofone’sparents).Giventhatbothself-trustandtrustinfriendsisvaluableforepistemicandnon-epistemicreasons,itdoesnotseemwrongtoaccordprimafacieweighttoafriend’snewsetofbeliefs ifone saw this friend as an epistemic peer prior to conversion. Rational argument, as

33

illustratedbyAugustine’sDeUtilitateCredendi,canplayacrucialroleinevaluatingpeerdisagreementcausedbyconversion.Itnotonlyhelpsone’sfriendstocriticallyevaluatetheirbeliefs,butalsoallowstheconverttocriticallyreflectonherownbeliefs.

Chapter4:What(ifanything)canweinferfromcommonconsent?

4.1Whatcommonconsentmightmean

Muchwork in social epistemologyhas focusedondisagreement. There is comparablylittleworkontheepistemicsignificanceofagreement.Inavarietyofeverydaysituations,findingyourselfinagreementthatpwilloftenjustifiablyincreaseyourcredencethatp.If IworkoutatrickymathematicalequationandIfindthatmyepistemicpeerhasthesame result, I justifiably becomemore confident that I am right. I can become evenmore convinced I am right if a greatmany people (of variousmathematical abilities)arrive at the same result, independent fromme. The same intuitions are elicited bymore complex cases, for example, if two medical doctors converge upon the samediagnosis, it would seem right for them to accord more confidence to it. This isespeciallythecaseiftheycametotheirdiagnosesindependently(Goldman,2001).Buteveniftheydidhaveacommonsource,forinstance,theymightendorsethediagnosisof another medical doctor who is highly respected in her field, agreement can haveepistemicsignificance. (Iwill saymoreontheroleof source-independence lateron inthechapter.)

Agreementhassomeepistemicsignificance,buthowmuchweightshouldweattachtoit? As we saw in chapter 2, philosophers from various traditions have expressedreservations about uncritically accepting the consent of the religious community onehappens tobelong to.Toal-Ghazālī, as tomanyothermedievalMuslimphilosophers,taqlīd(mindlesstrustintestimony)isinferiortoreasoning.Itmaybeacceptableforthelaypersonbutnotforphilosophers.Millagrees:

toathinkertheargumentfromotherpeople’sopinionshaslittleweight.Itis but second-hand evidence; andmerely admonishes us to look out forandweighthereasonsonwhichthisconvictionofmankindorofwisemenwasfounded.(Mill,1874,156)

However,upuntil theseventeenthcentury, theologiansregardedcommonconsentassubstantial evidence for the existence of God. The argument from common consent,alsoknownastheconsensusgentium,establishesthetruthoftheismthroughitsnearuniversality.Rollins(2015,84)givesthefollowingformulationoftheargument:

P1:BeliefinGodis(nearly)universal.

P2:Foranygivenpropositionp,ifbeliefinpis(nearly)universal,pmustbetrue.

34

P3:So,ifbeliefinGod(s)is(nearly)universal,God(s)mustexist.

∴Godmustexist.

Statedbaldlyinthisform,theargumentisnotveryimpressive.Examplesofbeliefsthathadnear-universalconsentandthatturnedouttobefalseare legion, includingaboutcometsbringingdiseaseandmisfortune,orabouttheintellectualcapacitiesofwomen.Yet,consensusgentiumwasamongthemostpopularnaturaltheologicalargumentsinthe EarlyModern period, with proponents such as John Calvin, Pierre Gassendi, andJohn Wilkins. But it fell out of favor in the eighteenth century, and today is hardlymentionedinmajoranthologiesandreviewsofnaturaltheologicalarguments.Inpart,itlost its appeal because of increasing doubts about the reliability of the cognitivemechanismsthatwerethoughttobeatthebasisoftheisticbelief,aswillbeshowninthe next section. But to some extent, it fell out of fashion because of an increasingepistemic individualism, exemplified by authors such as Descartes and Locke, whofocusedontheindividualthinkerandhisappealtoreasonandargument.Aswesawinchapter1,thisepistemicindividualismwasaconsequenceofanepistemiccrisis—alossof confidence in themedieval religious,moral,andscientific consensus.However, thesocial nature of belief formation has experienced something of a revival in socialepistemology. Social epistemologists, like traditional epistemologists, focus on howagentsformbeliefsandacquireknowledge.Butunliketraditionalepistemologists,theysee knowledgeproduction andbelief formation as a primarily social enterprise. Sincesocialmechanismsplayan important role in theacquisitionof religiousbeliefs (aswesawinchapter3),questionsabouttheirreasonablenessshouldtakeintoaccountwhatothersbelieveandhowwedealwithsociallyacquiredknowledge.

It is useful to re-evaluate the common consent argument in the light of socialepistemology.Thischapterwillreviewthreeversionsoftheargumentandgaugetheirstrengths and weaknesses: an early and representative version by John Calvin,Zagzebski’s argument from self-trust, and Kelly’s argument that common consent isevidence.Ithenproposeafourthversion,makinguseofnewlydevelopedtoolsinsocialepistemology,whichformalizetheintuitionthatifyoufindyourself inagreementwithothers,youcan(undersuitableconditions)attachhighercredencetoyourbeliefs.Onesurprising result of applying these new tools is that diversity of religious opinion canstrengthen the argument from common consent. But overall, I will argue, commonconsentonlyhasweakevidentialvalue.

4.2Theoriginalargumentfromcommonconsent:Frominnatenesstotruth

Most traditional versions of the argument from common consent proceeded in twoinferential steps: from common consent to innateness, and from innateness to truth.Thereasonwhybelief inGod isnear-universal is thatGod implantedthebelief inourminds. In thisway,Godfigures in theexplanation forcommonconsent.The followingformulationbyCalvinexemplifiesthistwo-stepprocess:

35

There is within the human mind, and indeed by natural instinct, anawarenessofdivinity. [...]Godhimselfhas implanted inallmenacertainunderstandingofhisdivinemajesty.[...][T]hereisnonationsobarbarous,no people so savage, that they have not a deep-seated conviction thatthereisaGod[...]Fromthisweconcludethatitisnotadoctrinethatmustbe first learned inschool,butoneofwhicheachofus ismaster fromhismother’swombandwhichnature itselfpermitsnoonetoforget. (Calvin,1559[1960],book1,chapter3,43–46)

Calvinmakesempiricalclaimsabouttheprevalenceofreligiousbeliefsacrosscultures,whichwouldreceivescrutiny inthefollowingcenturies,aswesternersgot increasinglyin contactwithpeople fromother religious traditionsdue toexploration, colonialism,andtrade.Theawarenessofdivinity(sensusdivinitatis)isahypothesisthatexplainsthewidespreadbelief inGod. The consensus gentium lost its philosophical appeal in partbecause this two-step inference became increasingly problematic (Reid, 2015): itwasunclearwhetherbeliefinGodwasreallyinnate,especially,asHume(1757)andothersobserved, given that there were many societies in which belief in God was absent.Religiousbeliefmaybewidespread,buttheismisnot,andevenreligioussentimentsarenotcross-culturallyuniversal.AsHumeobserved

The belief of invisible, intelligent power has been very generally diffusedoverthehumanrace,inallplacesandinallages;butithasneitherperhapsbeen so universal as to admit of no exceptions, nor has it been, in anydegree, uniform in the ideaswhich it has suggested. Some nations havebeendiscovered,whoentertainedno sentiments of Religion, if travellersandhistoriansmaybe credited; andno twonations, and scarce any twomen,haveeveragreedpreciselyinthesamesentiments.Itwouldappear,therefore, that thispreconceptionspringsnot fromanoriginal instinctorprimary impression of nature, such as gives rise to self-love, affectionbetween the sexes, love of progeny, gratitude, resentment; since everyinstinctofthiskindhasbeenfoundabsolutelyuniversal inallnationsandages, and has always a precise determinate object, which it inflexiblypursues.(Hume,1757,introduction)

Moreover, even ifwe grant that religious beliefs arewidespread, this does not showthat theism is true, because the mechanisms that lie at the basis of religious beliefformationmightnotbetruth-conducive.Hume(1757)arguedthataninabilitytoexplainandcontrolfutureeventswasatthebasisofpolytheism,whichheconsideredtobetheearliest form of religious belief. Humans anthropomorphized the environment, whichgave them a sense of control—they could now cajole and beseech the gods. Sincewishfulthinkingandfeararenotgoodbelief-formingmechanisms,commonconsenttoGod’sexistencewasvulnerabletoarebuttingdefeater.Arebuttingdefeatertopgivesusreasonforbelievingnot-p,whereasanundercuttingdefeatertopgivesonereasontosupposethatone’sgroundforbelievingp isnotsufficientlyindicativeofthetruthofp(Pollock,1987).Forexample,ifIhaveevidencethatafriendcouldnothaveknownthat

36

p,Ihaveanundercuttingdefeaterforhertestimony.ButifIhaveevidencethatshewaslyingtomethatp,Ihavearebuttingdefeaterformybeliefthatp,formedonthebasisof her (false) testimony. If religious beliefs were the result of unreliable cognitivemechanisms, we would have fresh reasons to doubt the veracity of these beliefs.Hume’sargumentsagainstconsensusgentiumhingeondetailsofhowreligiousbeliefsareacquiredandontheirdistributionacrosscultures.Hedidnothavegoodfirst-handevidence about the latter, but used unreliable reports of travelers and historians. Athorough assessment of the common consent argument would thus require anextensive review of the cross-cultural distribution of religious beliefs and theiracquisition. This is beyond the scope of this book. I will here focus on the sensusdivinitatisclaimandevidencefromcognitivesciencethatisrelevantforit.

Thediscussionabouttheismanditscognitiveoriginshasseenarevivalinthepastfewdecades.According toAlvinPlantinga (2000),belief inGodcanbeproperlybasic, justlike our belief in otherminds, an externalworld, or the past. Classic foundationalismholds that beliefs can only be properly basic if they fall into one of the followingcategories: they are incorrigible, evident to the senses, or self-evident. Any beliefwehold,inordertobejustified,musteitherbeproperlybasic,orderivefrombeliefsthatare properly basic. Plantinga argued that many beliefs that we hold without explicitreasoning, sensoryevidence,or argumentareproperlybasic. These include thebeliefthat otherminds exist, that the past exists, and that there is an external world.Wecannotknowforcertainthatotherpeoplehaveminds,asinternalmentalstatesarenotevidenttooursenses,andtheyarenotself-evidenteither(asisshownbythefactthatmostnon-humananimals,veryyoungchildren,andpeoplewithautismlacktheabilitytoattributebeliefstoothers).Yetbeliefinothermindsisproperlybasic.IfGodinstilledin human beings a sensus divinitatis, and if it is properly functioning, in a congenialenvironment,thebeliefscouldbeproperlybasicinthewayPlantingasuggests.

KellyClarkandJustinBarrett (2010)havearguedthat thecognitivescienceof religionprovidessupportfortheideathattherewouldbeasensusdivinitatis.TheydistinguishbetweenCalvin’sversionofthesensusdivinitatis,whichisaninnatebeliefinGod,andPlantinga’s version which conceptualizes the sensus divinitatis as dispositional: it iselicited by particular experiences, such as of awe-inspiring natural beauty ormorallyrelevantsituations.ClarkandBarrettidentifytwocomponentstothesensusdivinitatis:theoryofmindandagencydetection,alsoknownastheHyperactiveAgencyDetectionDevice(HADD).

Theory of mind is our ability to explain the behavior of other agents by attributingdesires,beliefs,andothermentalstatestothem.Thankstotheoryofmind,weareabletounderstandactionsthatotherwisewouldbepuzzling,suchassomeoneactingonthebasis of a false belief, searching for an object that is no longerwhere they originallyplacedit.Four-year-oldchildrenalreadydifferentiatebetweenthementalstatesofGodandthoseof limitedagents: theypredict thatGod,butnottheirmother,wouldknowthe contents of a closed box (e.g., Barrett, Richert, & Driesenga, 2001). There is alsotentative evidence linking theory of mind directly to religious belief: in one study

37

(Norenzayan,Gervais,&Trzesniewski, 2012)beingon the autistic spectrumpredicteddecreasedbeliefinGod,andmentalizingdeficits—areducedabilityintheoryofmind—mediatedthisrelationship.

HADD is an alleged mechanism (or set of mental mechanisms) that allows us todistinguish agents in our environment. The anthropologist Stuart Guthrie (1993)proposed that humans are prone to detect agents in their environment because itmakesevolutionarysensetodoso,asthecostsoffalsenegatives(failingtodetectanagent) far outstrip the costs of false positives (mistakenly detecting an agent). Forexample,hikerstendtomistakebouldersforbearsbutnotbearsforboulders,becausenot spotting bears in time could have potentially dire consequences, whilemisinterpreting boulders just gives them a little jolt. Barrett (2004) and others havefurtherdevelopedthistheory,arguingthatagencydetectionishyperactive, i.e.,pronetomakefalsepositives.Thesearearesultoftheexpectedpayoffsoffalsenegativesandfalse positives. Given the low cost of a false positive, and the high cost of a falsenegative,we can expectHADD tomisfire frequently. Thiswould give rise to religiousbeliefs, such as a creature I saw in the forest, which are subject to further culturaldevelopment,suchasforestspirits.

ClarkandBarrett(2010)arguethatsensusdivinitatisconsiststheoryofmindandHADD,andtheycall this thegod-faculty. IthascomponentsbothofCalvin’snativistaccount,which sees religious beliefs as diffuse and underdetermined, and of Plantinga’sdispositionalaccount,asreligiousbeliefscanbeelicitedbyparticularexperiences.Oneproblemwith thisproposal is thatHADDand theoryofmindarenot theonlymentalmechanisms that are involved in religiousbelief. For example, promiscuous teleology,the tendency of young children (and adults) to attribute purpose and design to theworldaroundthem,mightcontributetotheformationofreligiousbelief(e.g.,Kelemen,2004).Thetendencyofhumans,frominfancyonward,tooverattributecausationandtoprefer agents as causes could also lie at the basis of theism and could underliearguments such as the kalām cosmological argument (De Cruz and De Smedt, 2015).One could further expand the god-faculty until all relevant contributing cognitivefacultiesaresubsumedunderit.

Asecondmoreseriousproblemisthatresearchershavenotfoundconsistentevidencefor a link between HADD, theory of mind, and religious belief. Several authors haveattemptedtofindacausalrelationshipbetweenthepropensitytooverattributeagencyandreligiousbelief,buttheyhavefailedtodoso(seevanLeeuwen&vanElk,inpress,for an overview). The role of theory of mind in religious belief formation remainscontentious too, for example, contrary to Norenzayan et al. (2012), Lindeman,Svedholm-Häkkinen,andLipsanen(2015)foundnocorrelationbetweentheoryofmindand religious belief, and only a modest correlation between teleology and religiousbelief.EvenifHADDplaysaroleinreligiousbeliefformation(forinstance,researchersmay not have implemented the proper research design to test for this), it is unclearwhetherthiswouldvindicatethe ideaofasensusdivinitatis.ForHADDissupposedtomisfirefrequently,elicitingmanyfalsepositives,inadditiontodetectingtheagentswho

38

are reallypresent. In theabsenceofenough informationonhowHADDoperatesandwhatrole (ifany) itplays in the formationof religiousbelief,anyattempttoestablishthepresenceofasensusdivinitatisbasedamongothersonHADDispremature.

It is possible that the failure to find clear causal relationships between cognitivemechanismsandreligiosityindicatesalackoftherightempiricalmeasures.AsitstandscurrentempiricalevidencedoesnotsupportCalvin’stwo-stepconsensusgentiumfrominnate religious belief to the truth of theism (the second could not be establishedthrough science in any case, given itsmethodological naturalism). Even if therewerestrong evidence for such a causal path, Calvin’s argument would be vulnerable to aproblemofunderdetermination.Religiousbeliefsarecross-culturallydiverse.Scholarsinthecognitive scienceof religiondonot think thatmonotheism is thedefault religiousposition. Rather, the proposed cognitivemechanisms that lie at the basis of religiousbelief would support a wide variety of religious beliefs. Calvin and Plantinga have aresponsetotheproblemofreligiousdiversity:thenoeticeffectsofsin.DuetotheFall,humancognitivecapacitiesaremarredandideasaboutGodhavebecomedistorted.AsIhave argued (De Cruz & De Smedt, 2013), this solution presents its own problems:withinthecognitivescienceofreligion,thereisnonaturalisticparalleloftheFalltothenaturalizedsensusdivinitatis.

Thisdoesnotindicatethatanycommonconsentargumentbasedontheinnatenessofreligious beliefs is unsound. However, the lack of empirical support for a sensusdivinitatisandthelackofanaturalisticaccountofthenoeticeffectsofsinmakeithardtonaturalizethecommonconsentargument.Onecouldstillarguethatthereisasensusdivinitatis and that God has implanted it to instill religious belief in people acrosscultures,nottakingonboardtheculturalbaggageof theFall.Thisargumentdoesnotstand or fall with empirical evidence. But note that traditional defenders of theconsensusgentium,suchasCalvin,makeuseofempiricalclaimstobolstertheviewthatGodhasimplantedasenseofthedivineinallpeopleacrosscultures.Inthelightofthis,Iwillnowexaminecommonconsentargumentsthat lookpurelyat thedistributionofbeliefs,whileremainingagnosticabouttheirgenealogy.

4.3Commonconsent,self-trust,andevidence

Linda Zagzebski (2011) proposes a common consent argument that appeals to thenotionofself-trust.AccordingtoZagzebski,humanshaveanaturaldesirefortruth,andtheybelievethatthisdesirecanbesatisfiedusingtheirfacultiesandcognitiveprocesses,including,perception,memory,andreasoning.Asaresultofthis,wehaveabasic,non-reflectiveandinescapableself-trust,atrustinourowncognitivefacultiestoobtainthetruth. For example, Iwant to knowwhat theweather is like (desire for truth); I lookoutsideandseemanycloudsinthesky.Itrustmyfaculties,andbelievethatitiscloudy(self-trust).Inaddition,wealsooftentryreflectivelytogetatthetruth;shetermsthisreflectivequality“epistemicconscientiousness”.Itcomesindegrees.Justalittleofitisa

39

naturalextensionofourpre-reflectiveself-trust.Butsomepeopletryespeciallyhardtoget tothetruth,usingmetacognitiveskills tosee if theyaredoingwell (e.g.,checkingthe result they obtained twice, and asking themselves such questions as “Have Ioverlooked evidence?”, “Am I in an echo chamber?”, “Can I use another method tocheckmyresults”).Whilethereare limitstoself-reflection,Zagzebskibelievesthatwecanmakeareasonableassessmentofhowepistemicallyconscientiousweare.

Onceweaccordtrust toourselves,Zagzebskibelieves,weareobligatedtoaccordthesametrusttothefacultiesofothers,because“[m]anyotherpeopleappeartometobejustasconscientiousasIamwhenIamasconscientiousasIcanbe”(Zagzebski,2011,29). If a largenumberofpeoplehaveagivenbelief, suchas thatBelgiumbecameanindependentcountryin1830wehavesomereasontobelievethatthisistrue,becausepeoplehaveanaturaldesirefortruthandweshouldtrustthembecausewehaveself-trust. Similarly, given that millions of people believe in the existence of God, I havereasontotrustthem.Thistrustisdefeasible.Forexample,itmightturnoutthatpeoplewhobelieveinGodarenotasconscientiousintheirbeliefsasatheistsoragnostics.Butprimafacie,thetrustoughttobeaccorded.ElizabethFricker(2014)thinksthismoveisproblematic, because it assumes that others are epistemically like us, which is anempiricalmatterandnotsomethingwecandecideapriori.Moreover,trustingothersisnotunavoidableinthesamewayastrustingourselvesis,usingthefollowinganalogy:

Suppose the only way across a chasm, to escape a deadly predator, isoverarotten-lookingbridge.SoIhavenochoicebuttorelyonittotakemyweight, taking a leap of epistemic faith in doing so. This in nowaymeans that I amguiltyof culpable inconsistency, if (having survivedmyordeal!)Irefusetocrossothersimilar,rotten-lookingbridges,whenIamnotconstrainedbycompellingpracticalmotivestodoso.(Fricker,2014,197)

Even if thecognitivedispositionsthatgeneratereligiousbeliefsaresimilar formeandotherpeople, itdoesnotfollowthat Ishouldextendmyself-trusttoothers.Whatwewould be observing is that many people are crossing the rotten bridge rather thanawaiting their fate with the predator. This says something about their relativeassessmentofbridgesandpredators,perhapsthattheyhaveabetterchanceofsurvivalcrossing the bridge than staying in the vicinity of the deadly (presumably hungry)predator.Oncewetakethislineofreasoning,consensusgentiumisnolongerbasedonself-trust,butaformofevidence.

Thomas Kelly (2011) has formulated (in the same edited volume as Zagzebski) aconsensus gentium argument that regards religious belief as a form of evidence. ForKelly(2011,142),theunderlyingreasoningbehindanyargumentfromcommonconsentisanargumenttothebestexplanation:“Iamjustifiedinconcludingthatpisthecaseonthebasisofthefactthatpisthedominantopinioninthegrouponlyifthetruthofpispartofthebestexplanationofthefactthatpisthedominantopinioninthegroup.”Forreligiousbelief,heconsidersthefollowingdatum.

40

Datum: A strong supermajority [i.e., more than 60%] of the world’spopulationbelievesthatGodexists.

IfthebestexplanationforthewidespreadbeliefinGodisGod’sexistence,thenwecaninfer that God exists based on the datum. Kelly considers several objections. Forexample,thedatummightbefalsebecause“God”doesnothavethesamereferent.Butthe strongest argument he finds is that the datum is insignificant because thesupermajority was not produced by independent convergence. In many—indeedprobably—in most cases of common consent, majority opinion is not producedindependently. The fact that the majority of Belgians believe Belgium became anindependentcountryin1830isduetothemlearningthisatschool.However,thetruthofthisfactispartoftheexplanationthatBelgians(iftheyrememberwhattheylearnedatschool)generallybelieveit.ThebestexplanationforwhytheschoolcurriculaincludethisisthatBelgiumindeedbecameanindependentcountryin1830,andthusinspiteofa lack of independent convergence, common consent here satisfies Kelly’s necessaryconditionoffiguringinthebestexplanation.

IsthisthecaseforbeliefinGod?Thisisunclear.Let’stakeKelly’s(2011)examplewherethereisnootherevidenceforagivenbelief.SupposeIseethebinsstandingoutsideofmyneighborsonadaythatthebincollectionnormallydoesnotoccur,Imayreasonablyconcludemyneighborsbelievethebincollectionwillhappentoday.Themostplausibleexplanationforthisisnotthatmyneighborshavesuddenlybecomedelusional,butthattheyhave informationthat, forsomeorotherreason, isnotavailabletome. Itseemsrationaltometoputmybinout.IfwelookatcommonconsenttobeliefinGod,itisaresultofacomplexmixofcognitive,sociological,andsociologicalfactors.Althoughthismixismorecomplexthanthecausaloriginsofthebeliefinthebincollectionday,onecouldarguethattheexistenceofGodisareasonableexplanationforwhythebeliefissowidespread.ButitisunclearwhetherananalogousargumentcouldbemadeforGod.

4.4Synergyandtheepistemicsignificanceofconsent

Under some circumstances, peer agreement that p does not only heighten yourcredence thatpbutalso raise thecredencehigher thaneitherpeers’ initial credence.Take, forexample, twodetectivesor twomedicaldoctorswhoconsidereachotherasepistemicpeersandcometothesamesuspectordiagnosis.Onehasacredenceof.95,and the other .97. Upon learning that my epistemic peer has reached .95, while Ihave.97,itseemsplausibleformenottosplitthedifference(lowermycredenceto.96),butbecomeevenmoreconfident.KennyEaswaranetal.(2016)formalizethisintuitionin theirconceptof synergy.Synergyallows twoagents toupdate theircredences toahighercredencethantheirinitialcredences,usingtheUpcorule:

41

Here,wehavepeersP,QandRwhohavepriorcredencesp,qandr.Uponlearningeachother’scredencesaboutA, theyshouldupdate their credencesaccording to theUpcorule. For example, suppose Catherine has a .8 credence that God exists, and sheencountersAnita,whohasa.7credencethatGodexists,andCatherineconsidersAnitaherepistemicpeeronthismatter.Ifsheweretosplitthedifference,shewouldhavetodowngradehercredence toGod’sexistence to .75.Butwith theUpco rule,Catherinecanupgradehercredenceupto.9,asfollows:

Thus,ifyouencountermanyepistemicpeerswhohaveahighcredencethatGodexists,you can increase your credence. It becomes relevant to see how belief in God andcredences inbelief inGodaredistributed,andwhoone’sepistemicpeersmightbe. Ifwe take “God” tomean theGodof theAbrahamicmonotheisms,worldwidebelief inGodwasatabout55percent in201013.This fallsshortofcommonconsentbut isstillverywidespread.IfthedefinitionismoreencompassingtoalsoallowformonotheisticstrandsofHinduismandothertraditions,thebeliefgoesuptoabout70percent.About89percentoftheUSpopulationbelievesinGod(slightlydownfromthepreviousyears)andabout63percentbelieveinGodwithcertainty,withafurther20percentwhoarefairlycertainthatGodexists14.“Certain”and“fairlycertain”arenotprecisecredences,buttheyare,wecanassume,quitehigh,significantlyover.5.Thus,areligiousbelieverwhofindsherselfsurroundedbyreligiousbelieverscanupdatehercredencetoahigherone.

However, religious belief is significantly lower among academics, and even more soamongphilosophers,aswesawinchapter3.Thepercentageofatheistsandagnosticsineliteuniversities in theUS isabout60percent (Ecklund&Scheitle,2007),but if smallliberal arts colleges, teaching-intensive state schools and community colleges areincluded, thepercentageof facultymemberswhobelieve inGodorahigherpower isabout 75percent (Gross& Simmons, 2009).Amonga sampleof philosophers (mainly

13http://www.pewforum.org/2012/12/18/global-religious-landscape-exec/.

14http://www.pewforum.org/religious-landscape-study/belief-in-god/.

42

drawnfromeliteuniversitiesintheUS,theUK,andCanada)thereare72.8percentwhodisbelieve inGod (Bourget&Chalmers,2004).Butmostphilosophersdonotconsiderreligious questions in their academic work; only a small minority (philosophers ofreligion) do. The majority of philosophers of religion (over 70 percent), unlikephilosophersingeneral,leantowardstheism(Bourget&Chalmers,2004;DeCruz,2017).From these numbers it is difficult to saywhat the common consentwould be that isrelevanttoaphilosopherwhoconsiderstheexistenceofGod(probably,mostreadersofthisbookareinthisposition).

Whetherthere iscommonconsentamongone’sepistemicpeersthusdependsonthegrouponeconsidersrelevantforthisquestion:theworldpopulation,theUSpopulation(assuming one lives in the US, the percentage of theists is lower in other westerncountries),academics,philosophers,orphilosophersofreligion.Thenotionofepistemicpeerhoodisrestrictedtodomain,soanepistemicpeerisinthiscaseapeeraboutthequestionofGod’sexistence.It isunclearwhetheracademicsorphilosopherswouldbeinabetterepistemicpositioninrespecttothisquestion.Throughoutthisbook,Iusethenotionofordinarypeerdisagreement(i.e.,disagreementwithpeopleweconsidertobeourepistemicpeersinday-to-daylife),ratherthanusingaparticularnotionofpeerhood(e.g., evidential, cognitive). But even with this loose notion it is hard to find a non-question beggingway of determiningwhomight be our epistemic peers aboutGod’sexistence.Differentreligioustraditionsemploydifferentstandardsofwhatwouldcountasrelevantevidenceandwhatwouldputone inabetterepistemicposition, includingmysticalexperience,scripture,theologicaltraining,aproperly-workingsensusdivinitatis,ortherightkindoffaithfulattitude.GiventhatpeopleacrosstheworldareinterestedinthequestionofGod’sexistence,andthatitisnotclearwhowouldcountasanepistemicpeer,itseemsprudenttousetheglobalpopulationastherelevantreferenceclass.

Epistemologiststakeitthatconsensusthatpstrengthensone’sjustificationforbelievingthat p, and that disagreement that p, especially with an epistemic peer or superior,should decrease one’s justification. However, there are several cases where diversitycan increaseourtrust intheconsentthatremains.Takedisagreement inthescientificdomain.Whenscientificexpertsdisagree, forexample,about theprecisemechanismsinvolved in natural selection, this is often taken as evidence that these theories areincorrect.But,asDellsén(inpress)hasargued,dissentbetweenscientistscanindicatethat their conclusionsarenotpurely a resultof sociological factors, suchas scientistswhounthinkinglyfollowpopularopinionoronescientist’sauthority.Frequentdissentinfact bolsters our confidence in scientific ideas forwhich there iswidespread consent,suchasclimatechangeandevolutionthroughnaturalselection,eventhoughscientistsmaydifferintheiropiniononhowquicklyclimateischangingorhownaturalselectionoperates. Barnett (in press) argues that when we evaluate the evidential value ofcommonconsent,weneedtoconsideritcouldbeexpectedthattheagentswouldreachthesameconclusion.Hetakesthecaseoftwostudentswhoreachthesameansweronalogicexam,butitturnsoutthatonestudenthasuncriticallycopiedtheother.Here,itseems straightforward that agreement between these two students lends no extra

43

weighttothesolutionbeingthecorrectone.However,supposeonestudentiscopyingtheother,butshecheckshisresultswhiledoingso.Shedoesnotcorrectthembecauseshethinksthatheisright,andsotheyendupwiththesameresult.Inthissecondcase,theagreementdoeslendmoreweighttothesolution.

Barnett’s rule canexplainwhy common consent in echo chambers, such as Facebookreactions to news, is of very limited value. Suppose Saray, a liberal philosopher, isengaged with her friends on Facebook on whether ending net neutrality15would beoverallworseformostInternetusers.Thetopicofnetneutralityisonewithwhichfewpeoplehavefamiliarity,andevenexpertscannotforeseetheramificationsofendingit(moreover, this is a politically charged topic with liberals pro net neutrality, andconservativesmore split about the issue). Saray believes that ending net neutrality isgoing to be bad and accords a high credence (.8) to this. But because she mainlyinteractswithotherprogressives,it isunsurprisingthattheytoothinkitwillbebadtoendit.SupposeherfriendsNaomiandBenjaminalsothinkendingnetneutralitywillbebad,withcredences .7and .9.UsingUpco, theywouldarriveatanupdatedcredenceof.988.(Sincetheresultingcredenceofeachagentishigherthantheirinitialcredences,this isan instanceofsynergy.)However,giventhattheyare inanechochamber,withlikeminded people it is unsurprising that they will agree. Common consent onlyincreasesmycredenceifIshouldnotexpect,inadvance,toreachthesameconclusionasmypeers.ThatwouldsuggestSarayshouldnotuseUpcotoupdateherbeliefinnetneutrality.Bycontrast,ifSaray’sconservativefriend,Calum,alsosaysthatscrappingnetneutralitywillbeterrible,sheshouldbemoreconfidentitwillbebad,becausehedoesnotshareherliberalviewpoints.

Applyingthis toreligiousbeliefs, religiousdiversitycancountertheviewthatreligiousbeliefs are purely the result of following authority. Indeed, new religiousmovementssuchastheChurchofJesusChristoftheLatterDaySaints,whichJosephSmithfoundedfollowing thereligiousdiversityhesaw inhiscommunity inNewYorkareevidenceofindependent reflectionon thequestionofGod’s existenceandotherquestionsaboutultimatereality.Thefactthatsomanypeopleacrosscultureshituponsimilarreligiousbeliefs(e.g.,insuper-powerfulandsuper-knowingagents,withwhomtheycaninteractbyperformingrituals) isanintriguingconvergence.ThelackofanempiricalbasisforasensusdivinitatismakesitdifficulttodefendaCalvin-stylecommonconsentargument,whichmovesfromcommonconsenttoinnateness,andfrominnatenesstotruth.Butitmakesthemoredirectcommonconsentargumentmoreplausible. Insofarasreligiousbeliefs across cultures have a common core of belief in supernatural agents, andwecannotexpectinadvancetoseesuchconsent,thebeliefhassomeprimafacieevidentialvalue.

15NetneutralityistheprinciplethatInternetserviceprovidersmusttreatalldataontheInternetthesame.

44

Chapter5:Religiousexpertiseanddisagreement

5.1Theimportanceofreligiousexpertiseindisagreement

Our everyday lives are governed by an extensive division of labor. We depend onexperts,includingengineers,plumbers,medicaldoctors,philosophers,mathematicians,physicists, priests, and car mechanics. We not only defer to experts, but we oftenoutsourcequestionswehaveaboutvariouspracticalandtheoreticalproblemstothem.Thisdividebetweenexpertsandpeoplewhodependon them iswidening, given thatexpert knowledge is increasingly becoming specialized and extensive. As a result, itbecomes increasingly harder for non-specialists to evaluate the claims expertsmake.This lack of transparency becomes a problemwhen experts disagree. In the religiousdomain, experts disagree about fundamental questions such as whether God exists,whether God became incarnate, whether there is divine revelation, and how this ismanifested. They alsodisagree aboutminutiae such aswhether the Trinity shouldbeunderstood through a social or Latinmodel, orwhether therewill be freewill in theafterlife.(Iwilldiscussinmoredetailwhatreligiousexpertsmightbefurtheroninthischapter,butforthetimebeing,youcanthinkaboutsuchpeopleaspriests,theologians,philosophersofreligion).

At first sight, a novice confronted with such disagreements has few resources toevaluatewhoisrightandwhomtotrust.Thisgivesrisetothefollowingpuzzle:howcanlaypeopleknowwhichexpertisright,andwhichexperttotrust,iftheycannotevaluatethe truth ofwhat experts are saying? This is what Alvin Goldman (2001) termed thenovice/expertproblem: theproblemofhowanovice,whilebeinganovice, canmakejustified judgments about the credibility of rivaling experts. Next to the practicalproblemofchoosingtotrustparticularexperts,expertdisagreementinadomainDalsogivesrisetodeeperepistemologicalworriesaboutD,namelywhetherthedisagreementshouldlowerourconfidenceinstatementsmadebyexpertsinDorevenwhetherthereareanyexperts inD.Religion isafieldrifewithexpertdisagreement,bothwithinandespeciallyacrossdifferentdenominationsandfaiths.Thenovice/expertproblemisalsoimportantbecausedisagreementscanoftenbereducedtoexpertversions.Supposeyouhaveanatheist(Abby)andatheist(Theo)whoarenotexperts,debatingthepossibilityoftheism.

Abby: I can’t believe youbuy intoall that theist humbug. Seriously, ifGodexistedhewouldpreventbadthingsfromhappening.

Theo:ButGodmayhavegoodreasonstoletbadthingshappen.Perhaps,itservessomegreatergood.

Abby:Okay,thatsoundsplausiblesometimes.Butwhywouldheallow,say,mysistertodieofcancer?Andifshehadtodie,whydid ittakemonthsandcausesomuchpain?WhydidheallowmyfathertoabusemewhenIwasachild?

45

Theo:Look,Idon’tpretendtoknowtheanswertoallofthat,andI’msorryaboutwhathappenedtoyouandyoursister.Idon’twanttominimizewhatyouwentthrough.ButGod’sreasonsmayjustbeinscrutabletous.

Conversationslikethesetakeplaceineverydaycontextsofreligiousdisagreementandoftenarereminiscentofmoresophisticatedphilosophicaldiscussions. Inthisexample,Abby’s formulationofanevidentialargument fromevil remindsoneofRowe’s (1979)classic example of a fawn that dies in a forest whose suffering does not seem topromote any greater good, and Theo ultimately appeals to skeptical theism16. Thischapterfocusesonthenovice/expertproblemofexpertdisagreementinphilosophyofreligion. My primary aim is to provide a model of expertise that can help solve thenovice/expertproblem.Inthenextsection,Iaskwhomightqualifyasareligiousexpert,drawingondifferenttheoriesofexpertiseoutlinedintheepistemologicalandcognitivescience literature. I then look at twomodels of deference to experts, the expert-as-authority model (Zagzebski, 2012) and the expert-as-advisor model (Lackey, 2018).While I argue that the lattermodelbetter in the faceofdisagreement, I show that itdoes not provide a clear heuristic of what to do in the face of disagreement amongexperts.Toremedythis,Iforwardanewmodel,theexpert-as-teacher,whichprovidesausefulsetofheuristics thatnovicescanusewhendealingwithexpertdisagreement. IexaminetherecommendationsofthemedievaltheologianandphilosopherMaimonidesin Guide of the perplexed (12th century [1963a]) on how to deal with conflictingopinionsofJewishreligiousexpertsasawaytofleshoutthismodel.

5.2Whocountsasareligiousexpert?

Across cultures, religiousexperts are common. In societieswith littledivisionof labor(beyondsexualdivisionoflabor),includingsmall-scalehunter-gathererandhorticulturalcultures, religious experts are among the few experts that a community can support.Theyareshamans,healers,andpriests,calledupontohelpmembersofthecommunitydeal with a variety of practical and epistemic problems. Consider baganga (singular:nganga),religiousexpertsinCentralAfricawhocallsuponancestorspiritsonbehalfoftheir clients (Janzen,1978). Inorder todo this, thengangacreates special containerscalledminkisi (singularnkisi) to summonand focus thepowerof ancestor spirits. Theancestorsarecalledtohelpinavarietyofsituationssuchascuringillnessorinfertility,andtoprotectagainstmisfortune.Angangaisarespectedindividualwhosestatusandearning increases ifheorsheisperceivedaseffective.Themakingofminkisi,ofteninthe shape of bowls and anthropomorphic figures, is a tricky practice that requiresextensive knowledge of symbolism: figurines are covered with nutshells or seashells,

16SeealsoLevinstein(2015,AppendixB),whohasoutlinedaformalmodeltoshowhowdisagreementsbetweenagentsareisomorphictotheirhypotheticalexpertversions.

46

(metaphors forwombs),mirrors as symbols for the boundary between the ancestorsandtheliving,teethandnailstosignifyaspirit’sforcefulness,andparticularplantswithvariousmeanings.Bagangaareconsideredexperts intheircommunitiesbothbecauseof theirextensiveknowledgeof thespiritworldand theirability tosummonspirits tohelphumanswithproblemstheyfaceineverydaylife.

Thisexampleindicatesthatreligiousexpertiseis,atleastinpart,asocialphenomenon.Thepracticalelementofbagangaexpertiseiscrucial:theirreputationistoalargeextentdetermined by the degree to which they can help clients by engaging in successfulreligious practice. Their theoretical knowledge, for instance, about whichminkisi aremostappropriateunderspecificcircumstances(e.g.,theuseofnutshellsandseashellsinminkisitocombatinfertility)isanintegralpartofthis.

Therearedifferentnotionsofexpertiseintheliteratureproposedbyphilosophersandcognitive scientists. I will briefly review them and then examine their suitability forthinking about religious experts. One influential notion of expertise is the veritisticnotion, which sees an expert as someone possessing a large store of knowledge (atminimum,truebeliefsarrivedatinajustified,non-Gettierizedway),comparedtonon-experts.Forexample,Goldmanconceptualizesanexpertasfollows:

Anexpert(inthestrongsense)indomainDissomeonewhopossessesanextensivefundofknowledge(truebelief)andasetofskillsormethodsforaptandsuccessfuldeploymentofthisknowledgetonewquestions inthedomain.(Goldman,2001,92)

For many domains of expertise, we do not know whether what the experts have isknowledge.This isnotonlythecaseforreligion,butalsoforotherdomainswhereweassume there are genuine experts, such as philosophy. Philosophers do not haveindependentmeasurestoassesswhethertheirintuitionsinsuchdomainsasethicsandepistemologyarethecorrectones,andneedtorelyonconsensus in theircommunity(Cummins, 1998).Given that consensus correlateswith the strengthof intuitions, butnotwith their correctness (Nagel,2012), thismakes ithard forphilosophers toassesswhethertheyhaveknowledgeortruebelief.Theevidentialnotionofexpertisefacesasimilarproblem.Accordingtothisnotion,

S is an expert with respect to domain D if and only if S possessessubstantially more and/or better evidence concerning propositions in Dthanmostpeopleintherelevantcomparisonclass.(Goldman,inpress)

Unfortunately,what counts as evidence is not clear-cut. For instance, in the religiousdomain there is disagreement aboutwhether private religious experiences constitutemore relevant evidence (this view is held in more experientially-based, mysticaltraditions such as Sufism in Islam or Pentecostalism in Christianity), or whether anextensiveknowledgeofdoctrineandscripturaltextswouldmakeoneanexpert.Thereseems tobeno straightforwardway toadjudicatewhat countsasevidenceandwhatshouldhavemoreevidentialweight.Thedifficultiesofestablishingexpertiseintermsof

47

knowledge or evidence have led some authors to focus more on the social role ofexpertise, and to cash it out in terms of skills and practice. For example, Collins andEvansarguethat

contributoryexpertiseenablesthosewhohaveacquire[anexpertskill]tocontribute to the domain to which the expertise pertains: contributoryexperts have the ability to do things within the domain of expertise.(Collins&Evans,2008,24)

Beingabletocontributetoafield(notmerelyunderstand itandgraspthetheoreticalelementsofit)indeedcapturesanessentialelementofexpertise.Butthisdefinitionstillleavesoutthesocialnatureofexpertise,andtheabilitytohelpothers.Goldman’sCAPdefinitionofexpertiseattemptstoincorporatethis:

S isanexpert indomainD ifandonly ifShasthecapacitytohelpothers(especially laypersons) solve a variety of problems in D or execute anassortment of tasks inD which the latter would not be able to solve orexecute on their own. S can provide such help by imparting to thelayperson(orotherclient)his/herdistinctiveknowledgeorskills.(Goldman,inpress)

Thesenotionsofexpertisehelpustodefinewhomightbeanexpert,buttheydonotidentifyexpertsamongmembersofourcommunity,nordotheyofferclearguidelinesfortheextenttowhichweshoulddefertoexperts,especiallyiftheirtestimonyconflicts(the novice/expert problem mentioned earlier). I will look at two solutions to thenovice/expertprobleminthenextsection.Eachprovidesconcreterecommendationsonwhat a novice should do when faced with an expert in Dwhen she wants to learnsomethinginD.ThemaindifferencebetweentheseapproachesisthatZagzebski(2012)recommendssuspendingone’sownjudgment,andthatLackey(2018)doesnot.

5.3Twomodelstosolvethenovice/expertproblem

LindaZagzebski (2012)puts forwardadefenseofepistemicauthority,atopicthatsheacknowledgesdoesnotsiteasilywiththecontemporaryemphasisonindividualismandfree choice. Political philosophers have attempted to justify political authority, as itseems inescapable that we are subject to it. But what about epistemic authority?Althoughitseemslessinevitable,wedofindourselvesfrequentlyinthesituationwherewehave to relyonexperts (e.g.,makingadiagnosisof illness,predictinghowclimatewillchange).Zagzebskiarguesthatdeferencetoepistemicauthoritycanberational. Itdoes not require being able to believe on someone else’s command, which iscontroversial (although, see Peels, 2015 for a recent defense that we can choose tobelieve).Itmerelyrequirespreemption.Preemptionistheprocessbywhichreasonsforbelieving something a personmight have are replaced by other reasons, in this case,

48

reasonsbasedonwhatanauthoritysaysordoes.Thepreemptionthesis,drawnfromJosephRaz,goesasfollows:

Preemption:Thefactthattheauthorityhasabeliefpisareasonformetobelievepthatreplacesmyotherreasonsrelevanttobelievingpandisnotsimplyaddedtothem.(Zagzebski,2012,107)

On the faceof it,bracketingone’sownreasons in favorofanepistemicauthority (anexpert) seems likeabadprocedure to formbeliefs.What reasonswouldwehave fordoingso?Zagzebskibelievesthatthereisastraightforwardjustification:supposeIamanoviceinD,andE isanexpertinD,andIamwonderingaboutformingtheD-relevantbeliefthatp.GiventhatEknowsmoreaboutDthanIdo,sheismorelikelytoberightaboutpthanIamifIweretotrytofigureoutwhetherpistrueonmyown.Toargueforpreemption,Zagzebskiforwardsthetrackrecordargument.Supposethatyouareinanexperimentwherea light flashesgreen80percentof the time,andred20percentofthetime.Youneedtopredictwhenthelightflashesgreeninordertoobtainareward.Given that the change in color occurs at random, the rational and utility-maximizingstrategyistoalwayspredictagreenlight.Althoughratsandpigeonsmaximize,humanstendtoperformsub-optimallybyonlypredictinggreenin80percentofcases,endingupwithfewercorrectguesses,about68percent17.Analogously,ifEismorelikelytogetitrightthanIamasanovice,IdobetterifItrustherinallmattersconcerningD.Thatiswhy Zagzebski thinks that we should treat experts as authorities (expert-as-authoritymodel).

While thisargumentmightwork if I know justoneexpert (orhave reasons to thinkapersonisanexpert)inD,thingsbecomeabittrickierwhenthereareseveralexpertsinD, especially if they disagree. As Lackey (2018) cautions, if there aremultiple expertswhodisagreefundamentallyaboutanumberofissues,youhadbetterpickagoodone.This caution seems especiallywarranted in the religious domain,where following thewrong religious expert potentially not only has detrimental epistemic effects (leadingonetohavewrongbeliefsabout thereligiousdomain),butalsopotentiallybadmoraleffects (leading one to hold morally unpalatable beliefs), or perhaps even ill salvificeffects.Theworrydoesnotonlyariseinthecaseofdisagreement,butalsointheidealcase Zagzebski has in mind where there is only one viable epistemic authority I canfollow. If I have to replace any reasons for belief inD with those ofE,what reasonswouldpossiblyleadmetoconsiderEanexpertinthefirstplace?MaybeIshouldtrustEbecausemanyotherstellmethatshe isanexpert.Atsomepoint, itbecomesdifficultand implausibletobracketawayone’sownreasonsforbeliefeveninadomainone isnot an expert in. As Lackey (2018) cautions, such deference “provides all of theresourcesforrenderingrationalthebeliefsofparadigmaticallyirrationalcommunities,”includingYoungEarthcreationists,FlatEarthers,orwhitesupremacists.

17Unfortunately,whilemanyauthorsmentionthisexperiment,Iwasunabletotraceitssource.

49

Given theproblems that the expert-as-authoritymodel faces, Lackey recommends analternative, the expert-as-advisormodel. She starts out from several caseswhereweuseexpertadvice,andcometoourownconsideredviewsasaresult.Expertwitnessesinatrialcanprovidemembersofthejurywithcompellingreasonsforwhypistrue,butthe jury will take these reasons into consideration together with other pieces ofevidencetoreachtheirverdict.Similarly,ethicsconsultantsathospitalscanadviseonwhether a givenprocedure is in thebest interest of thepatient andother interestedparties,butthedecisionto,say,refuseorfollowagiventreatmentisuptothemedicaldoctors,patients,andsometimestheirfamilies(asinthedecisiontoturnoffaventilatorthatkeepsapatientalive).Insuchcases,wedonothavetoreplaceourreasonswiththebeliefsoftheexpert.Togeneralize:

An expert that is an advisor does not give authoritative testimony orpreemptive reasons for belief; rather, her testimony provides evidenceforbelievingagivenpropositionand,inthisway,offersguidance.(Lackey,2018,238)

Anexpertopinion is regardedasa relevantpieceofevidence,next tootherpiecesofevidence.Lackey(2018)motivatesthissolutiontothenovice/expertproblembyappealtothepuzzleofisolatedsecond-handknowledge(firstdescribedinLackey,2011).Thisoccurs when someone knows that p through expert testimony, and lacks any otherrelevant information. Itwould seemthat it is improper toassert thatp in suchcases,even though one knows thatp (whichwould be a counterexample to the sufficiencycondition of the knowledge norm of assertion, which says that knowing that p is asufficientconditionforassertingthatp).Oneofthecasessheusestoelicittheintuitionthat flat-out assertionsof expertsdonot give sufficient reasons for accepting a givenbeliefthatpisDOCTOR.

Matilda is an oncologist at a teaching hospital who has been diagnosingandtreatingvariouskindsofcancerforthepastfifteenyears.Oneofherpatients,Derek,was recently referred toher office becausehehas beenexperiencing intense abdominal pain for a couple of weeks. MatildarequestedanultrasoundandMRI, but the resultsof the tests arrivedonherdayoff;consequently,alltherelevantdatawerereviewedbyNancy,acompetentmedicalstudentinoncologytrainingatherhospital.BeingabletoconferforonlyaverybriefperiodoftimepriortoDerek’sappointmenttoday, Nancy communicated to Matilda simply that her diagnosis ispancreaticcancer,withoutofferinganyofthedetailsofthetestresultsorthereasonsunderlyingherconclusion.Shortlythereafter,MatildahadherappointmentwithDerek,whereshetrulyassertstohimpurelyonthebasisofNancy’sreliabletestimony,“Iamverysorrytotellyouthis,butyouhavepancreaticcancer.”(Lackey,2011,34–35)

Note that in this scenario,Nancy isa student—neitheranexpert,norapeer (Benton,2016).Ifthecaseismodifiedwherethepersonreviewingtheevidenceonbehalfofthe

50

doctor is a fellow oncologist, or even a team of oncologists, or a student nearingcompletionofhertraining,theassertionseemsless improper.Maybethe intuitions intheDOCTORcasealsohave todowith the fact that thedoctor cannotexplain to thepatient why the evidence points to the pancreatic diagnosis. As Benton (2016, 506)argues, it is unclear whether “we expect that experts always have an obligation toexplaintoanon-expertwhatisbehindtheformationoftheiropinion.”Wedoseemtoexpectofexperttestimony,morethanofordinarytestimony,thatanexpertcanexplainthe reasons behind her testimony. Lackey (2018) argues that cases of second-handisolated knowledge favor her expert-as-advisor model over Zagzebski’s expert-as-authoritymodel.Expertsneedtofunctionasadvisorstonon-experts,andinordertodosotheyneedtobeabletoexplainorbackuptheirclaims.

However,thepuzzleofisolatedsecond-handknowledgedoesnotexclusivelyfavortheexpert-as-advisor model. It is compatible with a number of other solutions to thenovice/expertproblem,suchastheoneIwilloutlineinthenextsection,theexpert-as-teacher.Theexpert-as-advisormodeldoesnotsolve thenovice/expertproblemwhenwearefacedwithdisagreeingexperts.Thequestionofwhichexpertalaypersonoughtto trustwhen experts disagree is now replacedwith the question ofwhich advisor alaypersonshouldturnto,incaseseveraladvisorsdisagree.Moreover,whiletheadvisormodelprovidesmoreresourcestodismissthebeliefsofirrationalcommunitiessuchaswhitesupremacistsandYoungEarthCreationistsasirrational,thereisnoclearmandateagainstfollowingtheadviceofsuchcommunities.

Tosumup,theexpert-as-advisormodelisnotconstrainedenoughinprovidingguidanceforwhichexpert to trust. Italso fails tocapture thatsometimesweought to treatanexpertasanauthority.Intheadvisormodel,wearefreetochoosewhetherweshouldfollowtheethicsconsultant,believetheexpertwitness,orheedthefitnesscoach.Butifamedicaldoctorrecommendsanantibioticstreatmentforsepsis,itwouldbeill-advisedtodismisshiscounselinfavorofthatofaherbaldoctor.Moregenerally,iflisteningtogenuineexpertsbecomesoptional (as theexpert-as-advisormodel suggests), itwouldbe acceptable for policymakers to ignore experts, for instance, on climate change. Ifexpertspredictthatclimatechangewilllikelyhavedevastatingconsequences,itwouldbeprudenttoheedthem.Aplausiblemodelofexperttestimonywouldneedtosatisfyatleastthefollowingtwodesiderata:ontheonehand,itisnotagoodideatobracketone’sownreasons.Ontheotherhand,expertsdohavegenuineauthoritythatrequiressomeformofdeference.

5.4Theexpertasteacher

Conceptualizingexpertsasteachersprovidesanalternativetothepreviousmodels.Aswehave seen, themain shortcomingof the formermodel is that it asks laypeople tocordonofftheirownreasoninginfavoroftheexpert’sopinion.Thelattermodelisnotconstrainedenough:insomesituationswereallyshouldtreatanexpertasanauthority.

51

The novice/expert situation is often a novice/teacher situation. Cross-culturally,teachingisoneofthemainwaysinwhichexpertiseistransmitted.Thisdoesnotalwaysin the formofexplicit,verbal instruction. Itcanalsoconsistofa teacherdirecting theattentionofthelearnertoarelevantpartofthetaskathand,orprovidingnegativeorpositive feedback on a task the learner has performed (Legare, 2017), for example, apotterpointingoutthecorrectspeedwithwhichthepotter’swheelturns,andhelpingtoadjusttheapprentice’sclayonit.

Ellen Fridland (in press) has argued that teaching is crucial for complex, cumulativeculture (unique to humans) because it allows for innovation. Thanks to teaching, alearner knows what parts of a culturally transmitted skill are relevant, how they fittogether, and how the skill might be improved. By contrast, pure imitation is moreconservative.Childrentendtooverimitate,i.e.,toimitatefeaturesofaprocessthatdonotcontributetothesuccessfuloutcomeofthatprocess(Nielsen,Mushin,Tomaselli,&Whiten, 2014). This observation is cross-culturally robust and has been replicatedseveral times. Adults are not immune to it either, copying causally irrelevant actionswhentheyimitatethesolutiontoopenapuzzlebox(Whitenetal.,2016).

Whywouldwelosetimeandenergycopyingirrelevantelementsofanaction?Partofitmay simply be caution. Many cultural skills are opaque, and it is often unclear to anovicewhichpartsoftheactionmatter.Suppose,forexample,thataplantisonlyediblewhen it is extensively treated: it has to be dried, pounded, and boiled beforeconsumption.Withoutadetailedtoxicologicalanalysis,wecannotknowwhichpartsofthisprocessmakes theplant lose its toxicity. It is better tobe safe than sorry and tocopy all the actions, rather than hasten the process and risk being poisoned. Evenamong experts, overimitationmight occur for this reason.AsNick Shea (2009) notes,scientistswilloftenfollowtheminutiaeofaprotocol,e.g.,aparticulardoseofsolvent(10mlinsteadof,say,25ml),becauseofthetimeandresourcesinvolvedincarryingoutthe experiment. By contrast, if one were trained in that lab one would know therationaleforusingparticularsolventsandcouldexperimentwiththem.Theadvantageofteachingisthatitfreeslearnersfromslavishlycopyingmodelsbecausetheycometounderstandtherationaleforcertainsteps.Thisallowsthemnotonlytobecompetentintheirdomain,butalsotocontributetoit,anaspectofexpertisethatisemphasizedbothinCollinsandEvans’(2008)notionofcontributoryexpertise,andinGoldman’s(inpress)CAP definition of expertise. Once you know which parts of the plant processing aretheretodetoxifyit,andwhicharemerelytheretomaketheproducttastebetter,youcanskipsomestepswithoutendangeringyourhealth.Butunlikeimitation,thisrequiresexplicit, verbal instruction. This can provide understanding, which is much harder toobtainthroughimitativelearningalone.

The notion of teacher as expert captures something cross-culturally stable about thetransmissionof expertise in human cultures: humanshavebeen reliant on experts totransmit knowledgeand thisoftenoccurs ina teaching situation. Thisdoesnotmeanthateveryexemplarofthenovice/expertproblemisanovice/teachersituation.Rather,it isusefulwhenone isanovice inD, facedwithexperts inD, to regard themasone

52

wouldateacher:teachershavesomeauthority,asisthecaseintheexpert-as-authoritymodel,but theydonot require totaldeferenceor screeningoffone’sownreasons.Alearner issensible ifshelooksforreasonsforwhyateacheroffersthetestimonytheygive.Acceptingwhatateachersayswithoutlookingforthesereasonswouldbetaqlīd,i.e.,uncriticalacceptanceoftestimony(seechapter2).

Someepistemologists(e.g.,Pritchard,2008)havearguedthatpropositionalknowledge,evencausalpropositionalknowledge,isdistinctfromunderstanding.AllisonHills(2009)proposesthatunderstandingcannotbetransmittedtothesameextentaspropositionalknowledge. She gives the specific example of moral understanding, and argues thathavingunderstandingofamoralpropositionprequiresmorethanjustknowingthatp.Specifically, one should be able to explain why p in one’s own words, follow anexplanationofwhypgivenbysomeoneelse,drawfromqtheconclusionthatp,andsoon. Although testimony does not always transmit understanding immediately anddirectly,teaching, ifdoneproperlywithawillinglearner,results inthetransmissionofinsight (this is a high bar to clear, as one requires both a good teacher and awillinglearner).

The model of expert-as-teacher is applicable in many domains. In science, it isexemplified by the research-intensive university, where research helps to informteaching.Inthereligiousdomain,itisespeciallyapt,asmanyreligiousexpertsexplicitlyhave the title “teacher”. For example, in Hinduism, Buddhism, Jainism, and Sikhism,havegurus (teachers),who canhelp laypeople reach religious insight. In Judaism, theterm “rabbi” also means teacher. Rabbis are not priests—they do not have specialauthority toperform rituals,but theyhavea substantial knowledgeof the Jewish lawanditstraditionwhichgivesthemauthoritytotransmitit.Thetitle“teacher”doesnotmean that these experts always literally teach, but it says something about theircontributory expertise: they are able to contribute to laypeople’s knowledge byimpartingtheirdistinctiveknowledgeorskills.

5.5Decidingwhomtotrust:MaimonidesontheTalmudicsages

The model of expert-as-teacher provides some general guidelines on how novicesshould evaluate teachers. It strikes a balance between themaximal deference of theexpert-as-authority model and between the pick-and-choose attitude that may beengendered by the expert-as-advisor model. It indicates that we should see expertsprimafacieasauthorities,butnotscreenoffourownreasoning.Maimonidesprovidesaninsightintohowthismodelmightwork.

TheJewishmedievalphilosopherMosesMaimonides(RabbiMōšehbēn-Maymōn,alsoknown by the acronym Rambam, 1138–1204) was born in Cordoba, which had aflourishingJewishcommunityandcultureunderIslamicrule,withreligiousdiscussionsamongChristians,Muslims,andJews,aswellasan increasing influenceofAristotelianphilosophy.WhenCordobawastakenoverbytheAlmohadCaliphatein1148,Christians

53

andJews losttheirprotectedstatusandwereofferedthechoicebetweenconversion,death,or,exile.Maimonides’familychoseexileandsettledinFez,Morocco,whichwasalso a multicultural andmultireligious city. HisGuide of the perplexed (completed in1190)wasostensiblywrittenforhisstudent,JosephbenJudah,aJewishphysicianandpoet,who couldnot integratephilosophywith religious teachings. Theguide is a richresource of philosophical thought, including on the existence and attributes of God(suchashis incorporeality),negativetheology,andextensivecritiquesofChristianandIslamicphilosophicaltheology(Seeskin,2017).

Several parts of theGuidedealwith the Talmudic sages, or Rabbis (with a capital R).ThesewereTalmudicscholarsofthefirstfivecenturiesCE,whosecorpusofwritingsliesat the basis of the Halakhah, the Jewish law. Traditional Jewish practice is moreinfluencedby therabbinic interpretationsof theTorahthanby itsoriginal text,hencethe importance of these writings. There was, of course, a distinction between thewrittenTorahandtheRabbis’understandingofit.ThisdistinctionwasformalizedinthenotionofadualTorah,consistingof theoriginalwrittenTorah,and its interpretation,theoralTorah,which formspartof theTalmud.TheHalakhahwasnecessarybecausethePentateuchitselfisoftenbriefandlackingindetail.TheRabbishelpedtoturnthoseterse formulations into concrete directives for practicing Jews. For example, Leviticus23:42 states that for the Feastof Tabernacles (sukkot), “You shall dwell inbooths forsevendays.AllnativeIsraelitesshalldwellinbooths”.Nodetailsareprovidedaboutthedimensions and materials of these booths, about whether it applies to everyone orwhetherthereareexemptionsforthosewithdisabilities,orwhetherdwellingrequiressometransferofone’shousehold itemsto thebooths.Therewereseveral techniquesfor interpretation, including analogical reasoning (look for similar cases where thePentateuch does provide more detail), exegetical arguments, logical reasoning, andtransmittedtraditions.Usingdifferenttechniquesledtodifferentinterpretations,whichiswhytheTalmudcontainssomanydisagreements(Berger,1998).

Throughout his works, Maimonides shows tremendous respect for the Talmud.Nevertheless,theTalmudicauthorsdisagreedwitheachotheronseveral fundamentalquestions, such as the timing of the Coming of theMessiah.Moreover,Maimonidesdisagreedwith themonparticular topics suchas the validityof astrology.He soundlyand categorically rejected astrology, deeming it a formof pseudo-expertise. Hismainreasons for this rejection are that astrology, contrary to astronomy, fails on scientificgrounds,andthatitascribestocelestialbodiesdivineornearlydivinepowers,whichiscontrarytoJudaism’scommitmenttotheunityandsovereigntyofGod.Healsosawitasanimpedimenttofreewill(Langermann,2000).

Theendorsementofastrologyby thesagesposedadilemmaforMaimonides.On theonehand, hewants to defer to their epistemic authority; on the other, hewishes torejectastrology.Hissolutiontothedilemmaisnottocordonoffone’sownreasonsandviewswhenconsideringtheviewsofthesages:

54

IknowthatyoumaysearchandfindsayingsofsomeindividualsagesintheTalmud and Midrashim whose words appear to maintain that at themomentofaman’sbirth,thestarswillcausesuchandsuchtohappentohim ... it is not proper to abandonmatters of reason that have alreadybeenverifiedbyproofs,shakelooseofthem,anddependonthewordsofasingleoneofthesagesfromwhompossiblythematterwashidden...Amanshouldnevercasthisreasonbehindhim,fortheeyesaresetinfront,notinback.(Maimonides,12thcentury[1972],472)

If a sage makes a statement that contradicts what has been verified by proofs(Maimonides’ prime source of knowledge, reason), then hemay bemistaken, or wemisinterpretwhathewrites,orhiswritingshouldbe interpreted in thecontextofhistimeandplace.MaimonidesmakesthismoreexplicitinhisGuide,whereheencouragesthereadertothinkaboutthehistoricalcontextinwhichthesageswrote:

Do not ask me to show that everything they [the Sages] have saidconcerningastronomicalmattersconformstothewaythingsreallyare.Foratthattimemathematicswereimperfect.Theydidnotspeakaboutthisastransmittersofthedictaoftheprophets,butratherbecauseinthosetimestheyweremen of knowledge in those fields or because they had heardthese dicta from the men of knowledge who lived in those times.(Maimonides,12thcentury[1963b],III,15,459)

Maimonides encourages the reader to see the sages as experts in religious matters,interpreting what prophets have said and the Jewish law, but this does not makeexperts in all domains, for example, astronomy or astrology. To Maimonides eveninexperiencedstudentsarenotcompletelypowerless in the faceofdissentingexpertsand can use their own reasoning to evaluate the conflicting testimony of religiousexperts.Theycanusetheirownreasonstodiscardimplausibleclaims,for instance,bythinking about cultural influence/limitations of past experts, resisting halo effects, orunderstandingreasonsforwhytheexpertsmightbedisagreeing.Inthischapter,Ihavearguedthatwecanproductivelythinkofexpertsasteachers.Expertshelpustoobtainunderstandingofaconceptualspacewetrytolearnmoreabout.Inthereligiousdomainitisnotuncommontothinkofreligiousexpertsasteachers(e.g.,gurusandrabbis).Theadvantageofthismodelisthatexpertsgetdeferenceduetotheirepistemicauthority,butthatwecanstillappraisecriticallywhattheyhavetosay.Maimonides’approachtotheTalmudicsagesillustrateshowwecantreatreligiousexperts:withduerespect,butwithoutscreeningoffourownreasons.

Chapter6:Whyphilosophymatterstoreligiousdisagreement

55

Regulative epistemology aims to shape our doxastic practices, to help us obtaindesirablebeliefoutcomes.Throughout thisbook, I’vebeenagnosticaboutwhat theseoutcomes could be. Epistemologists have disagreed about what the fundamentalepistemicgood(s)couldbe:amongothers,theyhaveproposedtrue(accurate)beliefs,knowledge, justified beliefs,warranted beliefs, and understanding. For some authors,suchasal-Ghazālī,weshouldaimfornothinglessthanreligiousknowledge.Al-Ghazālīwaspessimisticabouttradition,senseperception,andreasonassourcesofknowledge,andhebelievedthatmysticalexperiencewastheonlysecurewaytolearnaboutGod.However,mostpeopleacceptawiderangeofsourcesforjustificationandknowledge.These include perception, memory, reason, and testimony. All of these channels arevulnerabletoskepticalworries.

Religious disagreement elicits skeptical worries about whether we can have anyreligiousknowledgeatall,andwhetherwecantrust(any)religiousexperts.Sincemostofusliveinreligiouslydiversecommunities,weareconfrontedwithsuchdisagreementson a daily basis. Philosophical argumentation and analysis can make distinctivecontributions in the public spherewhen religious disagreement occurs. These includearguments in philosophy of religion and analysis of religious disagreements by socialepistemologists. Ideally, philosophical reflection should be more thanmere post hocrationalizationsofbeliefswealreadyhold; itshouldgenuinely improveourbeliefsandthe way we form them. Some authors, such as Regina Rini (2017), have expressedskepticism about our ability to alter our individual epistemic practices. She gives theexampleof fakenews,wherepeoplereasonablytrustthetestimonyofthosewhoaresimilartothemandthusaretrickedintosharingandendorsingfabricatednewsstories,oftenof ahighlypartisannature.Others, suchas JonathanHaidt (2001)propose thatreasoning(philosophicalorotherwise) is largelyposthocrationalization.Wereasontoaffirm beliefs we already hold. This pessimistic picture even seems to hold up forphilosophers. In a series of papers, Eric Schwitzgebel and colleagues (see e.g.,Schwitzgebel, 2014, for an overview) investigated whether ethicists behave moremorally praiseworthy compared to other philosophers.Moral self-improvement is animportantmotivationforethics(e.g.,virtueethics). Ifethicscouldhelpus leadagood(ethical)life,weshouldexpectethiciststobehavemoreethicallythanotherpeople.Butthisisnotthecase.Onawiderangeofmeasures,suchassigningupasanorgandonor,eating meat, not stealing library books, clearing away trash after talks, ethicists aresimilar to other philosophers or the population at large. This seems to indicate thatwhatethicistsaredoingisposthocreasoningratherthanshapingtheirethicalbeliefsinanywaythathasanimpactonwhattheyaredoing.

Nevertheless,smallalterationsinindividualreasoningpracticescanpotentiallyimprovedebates in the public sphere. There is an increasing body of empirical researchindicating that our reasoning is intimately connected to argumentation, and thatreasoningisinherentlysocial(Mercier&Sperber,2017).Debatesabouthighlypolarizedand loaded topics tend tobeof poorquality, but thequality of the argumentation isimproved when conducted in small face-to-face groups with the soft prodding of

56

moderators. For example, Luskin, O’Flynn, Fishkin, and Russell (2014) let parents ofschool-aged children in a community in Northern Ireland debate about the future oflocalschools.ManyideasaboutthistopicarepolarizedbetweenRomanCatholicsandProtestants, who, until recently were engaged in violent conflict. Nevertheless, thedebatesproved fruitful andconstructive,andcitizens left thedebateswith consensusonsometopics.HugoMercierandDanSperber(2017,chapter17)discusstheexampleof the early abolitionistmovement as a case where debate about amorally chargedtopic (slave ownership) resulted in a dramatic shift in policy and attitudes. Theabolitionistswon their casebypointingout inconsistencies in theiropponents’ views.For example, those who were pro-slavery argued that slaves were well treated (andcertainlybetteroffthantheywouldhavebeeninAfrica),becausetherewaseconomicincentive to do so. By contrast, William Wilberforce, an MP who lobbied for theabolitionist cause, argued that it was more profitable for slavers to crowd slavestogether while being transported, so the argument that economic incentive leads tobettertreatmentdidnotalwaysholdup.ThisandotherargumentsultimatelyconvincedthemajorityofMPs to gradually abolish slave trade.Reasoningalsoproves fruitful inchangingpeople’smindsaboutpolarizedtopicssuchasclimatechange,withAmericanliberalsmore likelytoacceptscientificclaimsaboutclimatechangeashuman-inducedcompared to conservatives.Michael Ranney andDav Clark (2016) showed thatwhenparticipants see short instructional videos that explain the mechanisms of climatechange,theiracceptanceofthescientificclimatechangeconsensuswasincreased,bothinliberalsandconservatives.

Theseexamplesindicatethatargumentationworks,andthatdebatescanbefruitfulinchangingattitudes,eventhosethatarehighlyentrenched.Philosophersofreligionoftenprovideexpertversionsofordinaryreligiousdisagreements,suchasabouttheproblemofevilortheepistemicsignificanceofreligiousrevelation(seechapter5).Philosophershave developed elaborate versions of such everyday debates, and have carefullyconsideredcounterargumentsanddefenses.Thus,whilephilosophersofreligionarenotinaprivilegedepistemicpositionabout,say,theexistenceofGod,theyareexperts inarticulating reasons and arguments about God’s existence and other religious topicsexplicit.Thus,abetter take-upofphilosophicalargumentsabout religion in thepublicsphere can help us improve discussions among laypeople. An emerging literature inexperimental philosophy indicates that the intuitions of philosophers in a variety offields, includingethicsandepistemology, arequite similar to thoseof laypeople (e.g.,Schwitzgebel&Cushman,2015).Whilewecaninterpretthispessimistically,namelythatthereisnosuchthingasphilosophicalexpertise,thefactthatthereisnothinginherentlyphilosophical about philosophy is also a good thing. In particular, the claim thatphilosophersdrawonthesamesortsofreasoningandresourcesasordinarypeoplealsomeans that philosophical knowledge does not require any special defense orjustification (Gutting,2009).Thismeans thatwecanevaluatephilosophicalclaims thewayweevaluateotherclaimsorarguments,bylookingattheevidence,thesoundnessof theargument, and soon.Wesaw that irrelevant causal factors suchasupbringingandeducationplay a role in the religiousbeliefswe for themostpart enduphaving

57

(chapter 2), and that they alsoplay a role in conversions (chapter 3). Philosophersofreligionarenot immuneto irrelevant influences.But this isnotaproblem;wecan letdiversity work. Given the wide diversity of religious beliefs across cultures, it isimportantthatphilosophersofreligionengagewithawiderangeoftraditions,notjustthosefoundinChristianityinwesterncultures.Suchargumentscanthenbeusedinthepublic sphere to reason about religious matters. Intellectual diversity and thoughtfuldebatehelpcommunitiesobtainmorejustifiedbeliefs.

Acknowledgements

Many thanks to Johan De Smedt, Hugh Burling, Kevin Schilbrack, Neil Van Leeuwen,MatthewBenton,KateKirkpatrick,LiamKofiBright,JonathanIchikawaJenkins,fortheircommentstoanearlierversionofthismanuscript.

References

Al-Ghazālī, A. H.M. (11th century [1963]). The incoherence of the philosophers (S. A.Kamali,Trans.).Lahore:PakistanPhilosophicalCongress.

Al-Ghazālī,A.H.M.(ca.1100[1952]).Deliverancefromerror.InW.M.Watt(Trans.),Thepracticeandfaithofal-Ghazālī(pp.1–359).London:Allen&Unwin.

Alston,W.P. (1991).PerceivingGod.Theepistemologyof religiousexperience. Ithaca,NY:CornellUniversityPress.

Asiedu,F.B.A.(2001).ThelimitsofAugustine’sPersonalAuthority:TheHermeneuticsofTrust inDeutilitate credendi. InK.Paffenroth&K. L.Hughes (Eds.),Augustineandliberaleducation(pp.124–145).Lanham:Rowman&Littlefield.

Augustine. (4th century CE [1961]). Confessions (R. S. Pine-Coffin, Trans.). London:Penguin.

Augustine.(5thcentury[1953]).Theusefulnessofbelief(Deutilitatecredendi).InJ.H.S.Burleigh (Ed.), Augustine: Earlier writings (pp. 284–323). Philadelphia: WestminsterPress.

Ballantyne, N., & Coffman, E. (2012). Conciliationism and uniqueness. AustralasianJournalofPhilosophy,90,657–670.

Barnett, Z. (in press). Belief dependence: How do the numbers count? PhilosophicalStudies,1–23.

Barrett,H.C.(2004).DesignversusdescentinShuarchildren’sreasoningaboutanimals.JournalofCognitionandCulture,4,25–50.

58

Barrett, J. L.,Richert,R.A.,&Driesenga,A. (2001).God’sbeliefsversusmother’s:Thedevelopmentofnonhumanagentconcepts.ChildDevelopment,72,50–65.

Benton, M. A. (2016). Expert opinion and second-hand knowledge. Philosophy andPhenomenologicalResearch,92,492–508.

Berger,M.(1998).Rabbinicauthority.Oxford:OxfordUniversityPress.

Bogardus, T. (2013). The problem of contingency for religious belief. Faith andPhilosophy,30,371–392.

Bourget, D., & Chalmers, D. J. (2014). What do philosophers believe? PhilosophicalStudies,170,465–500.

Calvin, J. (1559 [1960]). Institutes of the Christian religion (F. L. Battles, Trans.).Philadelphia:WestminsterPress.

Carel, H., Kidd, I. J., & Pettigrew, R. (2016). Illness as transformative experience. TheLancet,388(10050),1152–1153.

Christensen, D. (2007). Epistemology of disagreement: The good news. PhilosophicalReview,116(2),187–217.

Christensen, D. (2010). Higher-order evidence. Philosophy and PhenomenologicalResearch,81(1),185–215.

Christensen, D. (2011). Disagreement, question-begging, and epistemic self-criticism.Philosopher’sImprint,11(6),1–22.

Clark,K.J.,&Barrett,J.L.(2010).Reformedepistemologyandthecognitivescienceofreligion.FaithandPhilosophy,27,174–189.

Cohen,G.A.(2000).Ifyou’reanegalitarian,howcomeyou’resorich?Cambridge,MA:HarvardUniversityPress.

Collins,H.,&Evans,R.(2008).Rethinkingexpertise.ChicagoandLondon:UniversityofChicagoPress.

Cummins, R. (1998). Reflectionson reflective equilibrium. InW.Ramsey&M.DePaul(Eds.),Theroleofintuitioninphilosophy(pp.113–127).NewYork:Rowman&Littlefield.

DeCruz,H. (inpress).Religiousbeliefandphilosophicalviews:aqualitativestudy.ResPhilosophica.

De Cruz, H. (2017). Religious disagreement: An empirical study among academicphilosophers.Episteme,14,71–87.

DeCruz,H.,&DeSmedt,J.(2015).Anaturalhistoryofnaturaltheology.Thecognitivescienceoftheologyandphilosophyofreligion.

59

DeCruz,H.,&DeSmedt, J. (2013).Reformedandevolutionaryepistemologyand thenoeticeffectsofsin.InternationalJournalforPhilosophyofReligion,74(1),49-66.

Dellsén,F.(inpress).Whenexpertdisagreementsupportstheconsensus.AustralasianJournalofPhilosophy,1–15.

Dixon, R. D., Lowery, R. C., & Jones, L. P. (1992). The fact and form of born-againreligiousconversionsandsociopoliticalconservatism.ReviewofReligiousResearch,34,117–131.

Douven, I. (2010).Simulatingpeerdisagreements.Studies InHistoryandPhilosophyofScienceA,41,148–157.

Draper, P.,&Nichols, R. (2013).Diagnosing cognitive biases in philosophy of religion.TheMonist,96,420–444.

Easwaran, K., Fenton-Glynn, L., Hitchcock, C.,&Velasco, J. D.(2016).Updating on thecredences of others: Disagreement, agreement, and synergy. Philosophers’ Imprint,16(11),1–39.

Ecklund,E.H.,&Scheitle,C.P.(2007).Religionamongacademicscientists:Distinctions,disciplines,anddemographics.SocialProblems,54(2),289–307.

Edwards,J.(1821).Atreatiseconcerningreligiousaffections,inthreeparts.Philadelphia:JamesCrissy.

Elga,A.(2007).Reflectionanddisagreement.Noûs,41,478–502.

Feldman, R. (2007). Reasonable religious disagreements. In L. Anthony (Ed.),Philosopherswithoutgods(pp.194–214).Oxford:OxfordUniversityPress.

Frank, R. M. (1994). Al-Ghazālī and the Ash‘arite school. Durham and London: DukeUniversityPress.

Fricker,L.(2014).Epistemictrustinoneselfandothers—Anargumentfromanalogy?InL. Callahan& T.O’Connor (Eds.),Religious faith and intellectual virtue (pp. 174–203).Oxford:OxfordUniversityPress.

Fridland, E. (in press). Do as I say and as I do: Imitation, pedagogy and cumulativeculture.MindandLanguage.

Gauchat,G.(2012).Politicizationofscienceinthepublicsphere:AstudyofpublictrustintheUnitedStates,1974to2010.AmericanSociologicalReview,77,167–187.

Gellman, J. (1997).Experience ofGod and the rationality of theistic belief. Ithaca andLondon:CornellUniversityPress.

60

Gellmann,J.(1992).Anewlookattheproblemofevil.FaithandPhilosophy,9(2),210–216.

Goldman, A. I. (2001). Experts: Which ones should you trust? Philosophy andPhenomenologicalResearch,63,85–110.

Goldman,A.I.(inpress).Expertise.Topoi.

Gould, S. J. (2001). Nonoverlappingmagisteria. In R. Pennock (Ed.), Intelligent designcreationism and its critics. Philosophical, theological, and scientific perspectives (pp.737–749).Cambridge,MA:MITPress.

Griffel,F.(2005).Taqlīdofthephilosophers:al-Ghazālī’sinitialaccusationinhistahāfut.In S. Günther (Ed.), Ideas, images, and methods of portrayal: Insights into classicalArabicliteratureandIslam.LeidenandBoston:Brill.

Griffel, F. (2017).Al-Ghazālī’s Incoherenceof thephilosophers. InK. El-Rouayheb&S.Schmidtke (Eds.), The Oxford handbook of Islamic philosophy (pp. 191–200). Oxford:OxfordUniversityPress.

Gross, N., & Simmons, S. (2009). The religiosity of American college and universityprofessors.SociologyofReligion,70,101–129.

Guthrie,S.E.(1993).Facesintheclouds.Anewtheoryofreligion.NewYork&Oxford:OxfordUniversityPress.

Gutting,G.(1982).Religiousbeliefandreligiousskepticism.NotreDame,IN:UniversityofNotreDamePress.

Gutting,G.(2009).Whatphilosophersknow:Casestudiesinrecentanalyticphilosophy.Cambridge:CambridgeUniversityPress.

Haidt,J.(2001).Theemotionaldoganditsrationaltail:Asocialintuitionistapproachtomoraljudgment.PsychologicalReview,108,814–834.

Harris, P. L., & Corriveau, K. H. (2011). Young children’s selective trust in informants.PhilosophicalTransactionsoftheRoyalSocietyB,366,1179–1187.

Heiphetz, L., Spelke, E. S., Harris, P. L., & Banaji, M. R. (2013). The development ofreasoningaboutbeliefs:Fact,preference,and ideology. JournalofExperimentalSocialPsychology,49,559–565.

Hick, J. (1988). God and the universe of faiths. Essays in the philosophy of religion.London:Macmillan.

Hills,A.(2009).Moraltestimonyandmoralepistemology.Ethics,120,94–127.

61

Hume, D. (1757). The natural history of religion. In Four dissertations (pp. 1–117).London:A.Millar.

James,W. (1902).Thevarietiesof religiousexperience.Astudy inhumannature.NewYork:Longmans,Green,andCo.

Janzen,J.(1978).ThequestfortherapyinLowerZaire.Berkeley:UniversityofCaliforniaPress.

Kelemen, D. (2004). Are children “intuitive theists”? Reasoning about purpose anddesigninnature.PsychologicalScience,15,295–301.

Kelley,J.,&DeGraaf,N.D.(1997).Nationalcontext,parentalsocialization,andreligiousbelief:Resultsfrom15nations.AmericanSociologicalReview,62,639–659.

Kelly,T.(2011).Consensusgentium:Reflectionsonthe‘commonconsent’argumentfortheexistenceofGod.InK.J.Clark&R.J.VanArragon(Eds.),Evidenceandreligiousbelief(pp.135–156).Oxford:OxfordUniversityPress.

King,N.L. (2012).Disagreement:What’s theproblem?Oragoodpeer ishardto find.PhilosophyandPhenomenologicalResearch,85,249–272.

King, P., & Ballantyne, N. (2009). Augustine on testimony. Canadian Journal ofPhilosophy,39(2),195–214.

Kitcher,P.(1990).Thedivisionofcognitivelabor.JournalofPhilosophy,87,5–22.

Kox,W.,Meeus,W.,&Hart,H.(1991).Religiousconversionofadolescents:TestingtheLoflandandStarkmodelofreligiousconversion.SociologicalAnalysis,52(3),227–240.

Lackey,J.(2010).Whatshouldwedowhenwedisagree?InT.Gendler&J.Hawthorne(Eds.),Oxfordstudies inepistemology (Vol.3,pp.274–293).Oxford:OxfordUniversityPress.

Lackey, J. (2011). Assertion and isolated second-hand knowledge. In J. Brown & H.Cappelen (Eds.), Assertion: New philosophical essays (pp. 251–275). Oxford: OxfordUniversityPress.

Lackey, J. (2018). Experts and peer disagreement. In M. Benton, J. Hawthorne, & D.Rabinowitz (Eds.),Knowledge, belief, and God: New insights in religious epistemology(pp.228–245).Oxford:OxfordUniversityPress.

Lakhdar, M., Vinsonneau, G., Apter, M. J., & Mullet, E. (2007). Conversion to IslamamongFrenchadolescentsandadults:Asystematicinventoryofmotives.InternationalJournalforthePsychologyofReligion,17(1),1–15.

62

Lane,J.D.,Wellman,H.M.,&Gelman,S.A.(2013).Informants’traitsweighheavilyinyoungchildren’strustintestimonyandintheirepistemicinferences.ChildDevelopment,84,1253–1268.

Langermann, Y. T. (2000).Maimonides’ repudiation of astrology. In R. S. Cohen& H.Levine(Eds.),AMaimonidesreader(pp.131–157).Dordrecht:Springer.

Legare, C. H. (2017). Cumulative cultural learning: Development and diversity.ProceedingsoftheNationalAcademyofSciences,114,7877–7883.

Levinstein,B.A.(2015).Withallduerespect:Themacro-epistemologyofdisagreement.Philosopher’sImprint,15(13),1–20.

Levy, N. (2017). Religious beliefs are factual beliefs: Content does not correlate withcontextsensitivity.Cognition,161,109–116.

Levy,N.(inpress).Duedeferencetodenialism:Explainingordinarypeople’srejectionofestablishedscientificfindings.Synthese,1–15.

Lindeman,M.,Svedholm-Häkkinen,A.M.,&Lipsanen,J.(2015).Ontologicalconfusionsbut not mentalizing abilities predict religious belief, paranormal belief, and belief insupernaturalpurpose.Cognition,134,63–76.

Littlejohn, C. (2013). Disagreement and defeat. In D. E.Machuca (Ed.),Disagreementandskepticism(pp.169–192).LondonandNewYork:Routledge.

Lofland, J., & Stark, R. (1965). Becoming a world-saver: A theory of conversion to adeviantperspective.AmericanSociologicalReview,30,862–875.

Longino, H. E. (1991). Multiplying subjects and the diffusion of power. Journal ofPhilosophy,88,666–674.

Luhrmann, T. M. (2006). The art of hearing God: Absorption, dissociation, andcontemporaryAmericanspirituality.Spiritus:AJournalofChristianSpirituality,5,133–157.

Luhrmann,T.M.(2012a).AhyperrealGodandmodernbelief.CurrentAnthropology,53,371–395.

Luhrmann, T. M. (2012b). When God talks back. Understanding the AmericanEvangelicalrelationshipwithGod.NewYork:Vintage.

Luskin, R. C., O’Flynn, I., Fishkin, J. S., & Russell, D. (2014). Deliberating across deepdivides.PoliticalStudies,62(1),116–135.

Maimonides,M. (12th century [1963a]). The guide of the perplexed (Vol. 1; S. Pines,Trans.).Chicago:UniversityofChicagoPress.

63

Maimonides,M. (12th century [1963b]). The guide of the perplexed (Vol. 2; S. Pines,Trans.).Chicago:UniversityofChicagoPress.

Maimonides, M. (12th century [1972]). Letter on astrology. In I. Twersky (Ed.), AMaimonidesreader(pp.463–473).Springfield,NJ:BehrmanHouse.

Matheson, J. (2009). Conciliatory views of disagreement and higher-order evidence.Episteme,6(3),269–279.

Mercier, H., & Sperber, D. (2017). The enigma of reason. Cambridge, MA.: HarvardUniversityPress.

Mill,J.S.(1859).Onliberty.London:JohnW.ParkerandSon.

Mill,J.S.(1874).Threeessaysonreligion.London:HenryHoltandCo.

Nagel, J. (2012). Intuitions and experiments: A defense of the case method inepistemology.PhilosophyandPhenomenologicalResearch,85,495–527.

Nielsen,M.,Mushin, I., Tomaselli, K.,&Whiten, A. (2014).Where culture takes hold:“Overimitation” and its flexible deployment in Western, Aboriginal, and Bushmenchildren.ChildDevelopment,85,2169–2184.

Norenzayan, A., Gervais, W. M., & Trzesniewski, K. H. (2012). Mentalizing deficitsconstrainbeliefinapersonalgod.PLoSONE,7,e36880.

Paloutzian, R. F., Richardson, J. T., & Rambo, L. R. (1999). Religious conversion andpersonalitychange.Journalofpersonality,67(6),1047–1079.

Pasnau,R.(2015).Disagreementandthevalueofself-trust.PhilosophicalStudies,172(9),2315–2339.

Paul,L.A.(2014).Transformativeexperience.Oxford:OxfordUniversityPress.

Peels,R.(2015).Believingatwillispossible.AustralasianJournalofPhilosophy,93,524–541.

Pettigrew, R. (2016). Jamesian epistemology formalised: an explication of ‘thewill tobelieve’.Episteme,13,253–268.

Plantinga,A.(2000).WarrantedChristianbelief.NewYork:OxfordUniversityPress.

Pollock,J.L.(1987).Defeasiblereasoning.CognitiveScience,11,481–518.

Potter,K.(2013).Religiousdisagreement:internalandexternal.InternationalJournalforPhilosophyofReligion,74,21–31.

Pritchard,D. (2008).Knowing theanswer,understanding, andepistemic value.GrazerPhilosophischeStudien,77(1),325–339.

64

Pronin,E., Lin,D.Y.,&Ross, L. (2002). Thebiasblind spot:Perceptionsofbias in selfversusothers.PersonalityandSocialPsychologyBulletin,28(3),369–381.

Ranney, M. A., & Clark, D. (2016). Climate change conceptual change: Scientificinformationcantransformattitudes.TopicsinCognitiveScience,8(1),49–75.

Reid, J. (2015).Thecommonconsentargument fromHerbert toHume. Journalof theHistoryofPhilosophy,53,401–433.

Rini,R.(2017).Fakenewsandpartisanepistemology.KennedyInstituteofEthicsJournal,27(2),E43–E64.

Roberts, R. C., & Wood, W. J. (2007). Intellectual virtues. An essay in regulativeepistemology.Oxford:OxfordUniversityPress.

Rollins,J.(2015).Beliefsandtestimonyassocialevidence:Epistemicegoism,epistemicuniversalism,andcommonconsentarguments.PhilosophyCompass,10,78–90.

Rowe, W. L. (1979). The problem of evil and some varieties of atheism. AmericanPhilosophicalQuarterly,16,335–341.

Schoenfield,M.(2014).Permissiontobelieve:Whypermissivismistrueandwhatittellsusaboutirrelevantinfluencesonbelief.Noûs,48,193–218.

Schwitzgebel,E.(2014).Themoralbehaviorofethicistsandtheroleofthephilosopher.InC.Luetge,H.Rusch&M.Uhl(Eds.),Experimentalethics:towardanempiricalmoralphilosophy(pp.59-64).Dordrecht:Springer.

Schwitzgebel,E.,&Cushman,F.(2015).Philosophers’biasedjudgmentspersistdespitetraining,expertiseandreflection.Cognition,141,127–137.

Seeskin, K. (2017). Maimonides. Stanford Encyclopedia of Philosophy, retrievedSeptember 21, 2017, fromhttps://plato.stanford.edu/archives/spr2017/entries/maimonides/

Shea, N. (2009). Imitation as an inheritance system.Philosophical Transactions of theRoyalSocietyofLondonB:BiologicalSciences,364,2429–2443.

Shtulman,A.(2013).Epistemicsimilaritiesbetweenstudents’scientificandsupernaturalbeliefs.JournalofEducationalPsychology,105,199–212.

Simpson,R.M.(inpress).Permissivismandthearbitrarinessobjection.Episteme.

Smith,J.(1902).HistoryoftheChurchofJesusChristoftheLatter-DaySaints.PeriodI:HistoryofJosephSmith,theprophet,byhimself.Volume1.SaltLakeCity:DeseretNews.

Sperber,D., Clément, F., Heintz, C.,Mascaro,O.,Mercier,H.,Origgi, G.,&Wilson,D.(2010).Epistemicvigilance.MindandLanguage,25,359–393.

65

Stanford, K. (2006). Exceeding our grasp: Science, history, and the problem ofunconceivedalternatives.NewYork:OxfordUniversityPress.

Stroud,S.(2006).Epistemicpartialityinfriendship.Ethics,116,498–524.

Swinburne,R.(2004).TheexistenceofGod(6thed.).Oxford:ClarendonPress.

Taves,A.(2016).Revelatoryevents.Threecasestudiesoftheemergenceofnewspiritualpaths.Princeton,NJ:PrincetonUniversityPress.

Teresa of Ávila. (1577 [1921]). The interior castle or the mansions.http://www.ccel.org/ccel/teresa/castle2.pdf.

vanInwagen,P.(1996).Itiswrong,everywhere,always,foranyone,tobelieveanythinguponinsufficientevidence. InJ.Jordan&D.Howard-Snyder(Eds.),Faith,freedomandrationality(pp.137–154).Savage,Maryland:RowmanandLittlefield.

VanLeeuwen,N.&vanElk(inpress).Seekingthesupernatural:Theinteractivereligiousexperiencemodel.Religion,BrainandBehavior.

VanLeeuwen,N. (2014).Religious credence isnot factualbelief.Cognition,133, 698–715.

Vavova,K.(inpress).Irrelevantinfluences.PhilosophyandPhenomenologicalResearch.

Wedgwood,R.(2007).Thenatureofnormativity.Oxford:OxfordUniversityPress.

Wettstein, H. (2012). The significance of religious experience, New York: OxfordUniversityPress.

White,R.(2014).Evidencecannotbepermissive.InM.Steup,J.Turri,&E.Sosa(Eds.),Contemporarydebatesinepistemology(pp.312–323).Oxford:OxfordUniversityPress.

Whiten, A., Allan, G., Devlin, S., Kseib, N., Raw, N., & McGuigan, N. (2016). Sociallearninginthereal-world:‘over-imitation’occursinbothchildrenandadultsunawareofparticipationinanexperimentandindependentlyofsocialinteraction.PloSOne,11(7),e0159920.

Wolterstorff, N. (1996). John Locke and the ethics of belief. Cambridge: CambridgeUniversityPress.

Worrall,J.(2004).ScienceDiscreditsReligion.InM.L.Peterson,&R.J.VanArragon(eds.),ContemporaryDebatesinPhilosophyofReligion(pp.59-72).Malden,MA:Blackwell.

Zagzebski,L. (2011).Epistemicself-trustandtheConsensusGentiumargument. InK.J.Clark&R.J.VanArragon(Eds.),Evidenceandreligiousbelief(pp.22–36).Oxford:OxfordUniversityPress.

66

Zagzebski,L. (2012).Epistemicauthority:Atheoryoftrust,authority,andautonomyinbelief.Oxford:OxfordUniversityPress.

Zuckerman,P.(2007).Atheism.Contemporarynumbersandpatterns.InM.Martin(Ed.),The Cambridge companion to atheism (pp. 47–65). Cambridge: Cambridge UniversityPress.