Imperfect Reasons and - Arizona State dportmor/Imperfect_Reasons_and_Rational... · Imperfect Reasons…

  • Published on
    28-Aug-2018

  • View
    212

  • Download
    0

Embed Size (px)

Transcript

<ul><li><p>ForthcominginNos</p><p>ImperfectReasonsandRationalOptions</p><p>DOUGLASW.PORTMORE</p><p>ArizonaStateUniversity</p><p>Agentsoften face a choiceofwhat todo.And it seems that, inmost of these choicesituations, the relevant reasons do not require performing some particular act, butinsteadpermitperforminganyofnumerousactalternatives.Thisisknownasthebasicbelief.Below,Iarguethatthebestexplanationforthebasicbeliefisnotthattherelevantreasons are incommensurable (Raz) or that their justifying strength exceeds therequiring strength of opposing reasons (Gert), but that they are imperfect reasonsreasonsthatdonotsupportperforminganyparticularact,butinsteadsupportchoosinganyofthenumerousalternativesthatwouldeachachievethesameworthyend.Intheprocess,Idevelopanddefendanoveltheoryofobjectiverationality,arguingthat it issuperiortoitstwomostnotablerivals.</p><p>Agents often face a choice ofwhat todo.And it seems that, inmost of thesechoicesituations,therelevantreasonsdonotrequireperformingsomeparticularact,but insteadpermitperforminganyofnumerousactalternatives.1 Itseems,for instance, that I could now, in accordance with reason, do any of thefollowing:watchTV,readanovel,practicethepiano,volunteerforOxfam,workonthispaper,playwithmydaughter,orprepareformynextlecture.FollowingJosephRaz, Iwillcall this thebasicbelief.Razcalls it thebasicbelief,because itseems tobesufficientlywellentrenched inourcommonsense thinking thatweshouldgiveitcredenceunlessitcanbeshowntobeincoherentorinconsistentwithsomeofourrightlyentrenchedviews (1999,100).Ashesees it, then, thebasicbeliefisoneofourstartingpoints,andourtaskistoexplainitratherthantodefendit.However,evenjustexplainingitisnotaneasytask,foritseemsthatthere couldbe such rationaloptionsonly if therewereexactlyequal reason toperformeachoftheoptionalactalternatives,andyetitisdifficulttobelievethatsuch is thecase.2For instance, it isdifficult tobelieve that Ihave justasmuchreasontowatchTVastoeithervolunteerforOxfamorworkonthispaper.Afterall,volunteeringforOxfamseemsvastlysuperiortowatchingTVintermsoftheamountofimpersonalgoodthatitwoulddo,andworkingonthispaperseemsvastlysuperiortowatchingTVintermsoftheamountofgooditwoulddomeassumethatImnotatthismomenttootiredtoworkproductivelyonthispaper.3Moreover, the fact that, in many of these choice situations, the relevant actalternativesremainrationallyoptionalevenaftertherehasbeenaslightincrease</p></li><li><p>2</p><p>in the number and/or strength of the reasons that favor just one of thealternativesshowsthattheiroptionalstatuscouldnothavebeendue,inthefirstplace,toaperfectbalanceofreasons(Gert2008,14).Forinstance,itseemsthatitwould stillbe rationallypermissible forme to continue toworkon thispapereveniftherewereaslightincreaseinthestrengthand/ornumberofthereasonsthat favor justoneof theotheroptions,aswhere, say,Oxfam institutesanewpolicy ofgiving eachvolunteer adelicious cookie. So thepuzzle is to explainhow, inmost choice situations, there couldbe somany rationallyoptional actalternatives if, as seems to be the case, there is not exactly equal reason toperformeachofthem.</p><p>Iwillarguethatthispuzzlecanbesolvedbytakingabroaderviewofthingsand that, oncewe take this broader view,we find that there is in fact equalreason to do each of the more broadly conceived alternatives.4 As I willdemonstrate, the relevant alternatives are not, as weve been supposing, theparticularactalternativesavailabletotheagentatagivenmoment,butratherthevariouscoursesofactionthattheagentcouldperformovertheremainderofherlife.</p><p>1.ImperfectReasonsandRationalOptions</p><p>Often times,whatanagenthasmostreason todo isdeterminedbyfactsaboutwhat shehasmost reason to achieve.But such facts often fail to support anyspecific alternative, for there is oftenmore than oneway to achieve the sameresult.To illustrate,consider the following threeexamples.First, the fact that Ineedtogettotheairportprovidesmewithareasontotakeanyofthefollowingequallyattractivemeans togetting there:abus,a taxi,ora trainassume thatthese are all equally attractive given their comparative cost, comfort, andconvenience.5Second,thefactboththatIhaveapaperthatneedstobefinishedin amonths time and that Iwillmost likely finish itby then if andonly if Iimplementapolicyofspendingtwohoursadayworkingonitprovidesmewitha reason to spend two hours (any two hours) each day over the nextmonthworkingonit.And,third,thefactthatIneedtospendoneofthenexttwodaysgradingexamsandtheotherpaintingthebackyardfenceinordertomeetcertainimportantdeadlinesprovidesmewitha reasoneither (1) toplanon spendingtomorrowgradingexamsandthenextdaypaintingthebackyardfenceor(2)toplan on spending tomorrow painting the backyard fence and the next daygradingexams.</p><p>These three sorts of reasons arewhat I call imperfect reasons, for they areanalogous to imperfectduties in that theyallow for significant leeway inhowone chooses to complywith them.6 Just as the imperfect duty of beneficence</p></li><li><p> 3</p><p>requiresonlythatonebebeneficenttoacertainextentbut leaves itup toonesdiscretion towhomandonwhichoccasions tobebeneficent, imperfectreasonsspeakinfavorofachievingsomeworthyendbutleaveituptoonesdiscretionwhichoftheequallyattractivemeanstoachievingthatendtotake.</p><p>ImperfectreasonsariseinwhatJoshuaGertcallsmultipleoptioncases,casesinwhich there ismore than one equally attractivemeans to achieving the sameworthyend(2003,10).Inmultipleoptioncases,itisrationallypermissibletotakeanyoftheequallyattractivemeanstoachievingthatend,fortherelevantreasonsare imperfectreasons,whichdonotsupportperforminganyparticularact,butinstead supportperforming any of the acts thatwould each achieve the sameworthyend.</p><p>Thattherearemultipleoptioncaseswheretherelevantreasonsareimperfectreasonsis,Itakeit,uncontroversial.ButiftheexistenceofmultipleoptioncasesisasuncontroversialasIclaim,wemightwonderwhyphilosopherssuchasRaz(1999) andGert (2003) have held thatwe cannot account for the basic beliefsimplybyappealingtothem.Theansweristhattheyeachbelievethatmultipleoptioncasesareinsufficientlynumeroustoaccountforthebasicbelief.</p><p>Todemonstratethis,RazcitesacaseinwhichawomannamedMaryhastheopportunity to seeapowerfulperformanceofagoodplayather local theatertonight but decides to stay home instead (1999, 99). Razs view is that bothalternativesgoingtothetheaterandstayinghomearerationallypermissible.Andalthoughheadmits thatmany reasonsare imperfect reasons (orwhathecallsreasonsthatarenottimespecific),hedeniesthatsuchreasonsallowustoaccountforthebasicbelief.Hesays:</p><p>Inmanycasesthereasonsfordoingonethingoranotherarenottimespecific:thesamereasons and the same opportunity to conform to them will apply on a number,sometimesanindefinitenumber,ofoccasions.Thisdoesnot,however,explainthebasicbelief.Quiteapartfromthefactthatdelayisnotcostless,thebasicbeliefappliestotimespecificreasonsaswell.Mary,inourexample,doesnothavetogototheplayevenonthe last evening of its run. Shemay still just not feel like it and do something elseinstead.(1999,100)</p><p>Likewise, Gert believes that the set of cases in which there are rational</p><p>options is significantly broader than the set of multipleoption cases. Gertbelieves, for instance, that it is rationally permissible both to sacrifice $200 toprevent forty children from suffering from seriousmalnutrition for fortydaysandtorefusetodoso,choosinginsteadtospendthatmoneyononeself(2003,89).Asheseesit,thisisacasewherethereisarationaloption,buthedeniesthatitisamultipleoptioncase.Gertsays:</p></li><li><p>4</p><p>Multipleoption casesdepend cruciallyon the fact that the classof rationally justifiedoptionsinanygivencasecanplausiblybeseenaswaysofdoingthesamething.Intheexamples thatwereused tomotivate the justifying/requiringdistinction,on theotherhand,thepairsofjustifiedoptionswerealwaysofthefollowingform.</p><p>(a) Makeasacrificeforanaltruisticreason.(b) Donotmakethesacrifice.</p><p>Suchpairscannotplausiblybeconstruedasalternatewaysofpursuingthesameendoractingonthesamereasons[italicizedlettersaremine].(2003,13)</p><p>This, I will argue, ismistaken. I will show that options a and b can be</p><p>construedastwoalternativewaysofpursuingthesameend.AsGertadmits,itiscrucialtogetthelevelofdescriptionright,forwhetheritstruethatIamdoingwhatIamrationallyrequiredtobedoingdependsonhowwedescribewhatitisthatIamdoing(2003,1314).Toillustrate,consideragainthecasewhereImustgettotheairport,andletusassumethatIhappentobetakingataxitogetthere.IfweaskwhetherImrationallyrequiredtobedoingwhatIamdoing,theanswerwilldependonthelevelofdescription.If,ontheonehand,wedescribewhatIamdoingastakingataxitotheairport,theanswerwillbeNo,becausetakingthebus isanequallyattractivemeansofgetting to theairportand Iam rationallypermittedtodothatinstead.If,ontheotherhand,wedescribewhatIamdoing,moregenerally,asgoing to theairport, theanswerwillbeYes,because Iamrationallyrequiredtogototheairport(orsoweresupposing).So,atonelevelofdescription,IamrationallyrequiredtobedoingwhatIamdoing,and,atanother,Imnot.</p><p>IthinkthatthesamecanbesaidofGertsexample(and,aswellsee,ofRazsexampleaswell).Atamoregenerallevelofdescription,therearetwocoursesofaction that Imight takeover time,one that includesmyactingselfinterestedlynow and altruistically later and another that includesmy acting altruisticallynowandselfinterestedlylater,anditisplausibletoconstruethesetwocoursesofactionas twowaysof trying toachieve thesame thing:areasonablychoiceworthyfuture.7Ifwethinkthatanyreasonablychoiceworthyfuturewillcontainbothaltruisticactsandselfinterestedacts,thenitdoesntmatterwhichofthemIperformnowandwhichof them Iperform later so longas, in the end, Illbeperformingasufficientnumberofeach.</p><p>InRazs case,we can say something similar. Ifweacknowledge thatMaryconceivesofheragencyasbeingextendedovertime,thenweshouldthinkthat,attherelevantlevelofdescription,thepertinentchoiceisnotbetweenseeingthisplay on its last night and staying home, but between two courses of action,containingbothacertainamountofrelaxationandacertainamountofculturalenrichment.SoeventhoughthisisagoodplayanditsMaryslastopportunity</p></li><li><p> 5</p><p>toseeit,thiswillnotbeMaryslastopportunitytodosomethingasentertainingand culturally enriching as seeing this play. And surely she isnt rationallyrequired to take advantage of every such opportunity.On the broader view,then,whateachofus, includingMary,shouldbe trying todo is tocompleteareasonablychoiceworthyfuture,andanyreasonablychoiceworthyfuturewill,formostofus, includebothsomerelaxationandsomeculturalenrichment.Itsimportanttonote,then,thatifMarystayshometonightandrelaxes,therewillbeotheropportunitiesforhertosimilarlyenrichherself,andthat,ifshegoestotheplay, therewill be other opportunities for her to get the relaxation that sheperiodicallyneeds.Whatsimportant,then,isnotthatshetakesadvantageofanyspecific opportunity for relaxation or cultural enrichment, but that she takesadvantageofbothsortsofopportunitiessufficientlyoftenoverthecourseofherfuture.</p><p>Sowhileadmitting that therearemultipleoptioncases,RazandGertclaimthatsuchcasesareinsufficientlynumeroustoaccountforthebasicbelief,andsotheyresorttomorecontentiousclaimsintryingtoaccountforthebasicbelief:inRazscase, to theclaim that therearewidespread incommensurabilitiesamongreasons, and, in Gerts case, to the claim that reasons have two separabledimensions of strength (i.e., requiring strength and justifying strength). I,however, will argue that multipleoption cases are sufficiently numerous toaccountforthebasicbeliefand,thus,thatweneedntappealtosuchcontentiousclaimstoaccountforthebasicbelief.MyargumentwillbethatRazandGertfailto appreciate the fact thatmost typical choice situations aremultipleoptioncases, because they fail to view our choices from the appropriate level ofdescription: that of choosing between various future courses of action. To beclear, I am not going to argue that the contentious claims that Raz andGertappeal to in accounting for thebasicbelief are falseor even that they arenotneededtoaccountforsomethingelse.Iwillargueonlythattheyarenotneededtoaccountforthebasicbeliefand,thus,thatwecanaccountforthebasicbeliefwithonlythetheoreticalapparatusthatallsidesagreeon:imperfectreasonsandmultipleoptioncases.Also,IshouldwarnthereaderthatIwillnotbedefendingthebasicbelief.AsRazpointedout,our task is toexplain it,and thatswhat Iaimtodo,andassimplyaspossible.</p><p>Inthenextsection,Ipresentmyowntheoryofobjectiverationality,atheoryaccording towhichmost choice situations aremultipleoption cases. Then, insubsequentsections,Idefendthetheory.</p><p>2.TheFutureCourseofActionTheoryofObjectiveRationality</p></li><li><p>6</p><p>BeforeIstatemytheory,letmebeclearonwhatitisatheoryabout.Itis,asthename suggests, a theory aboutwhat it is objectively rational to do.Objectiverationality is about what agents ought to do in the most fundamental andunqualifiedsense.Soitsnotaboutwhattheymorallyoughttodo,norisitaboutwhattheyprudentiallyoughttodo.Instead,itisaboutwhattheyoughttodo,allthingsconsidered.Thus,tosaythatanact isobjectively irrational istosaythattheactionabsolutelyshouldnotbeperformed(Gert2004,137).Somesuggestthat we should understand objective rationality in terms of advisability:objectively irrationalactionsarethosethatnofullyinformedandwellmeaningadviserwouldrecommendperformingontheassumptionthattheagentwouldtake the advice.8Moreprecisely, though,we shouldunderstand thenotion ofobjectiverationalityintermsofreasons:anactisobjectivelyirrationalifandonlyif there are decisive reasons for the agent to refrain from performing it.9 Theobjectiverationalstatusofanactis,then,purelyafunctionofthereasonsforandagainstperformingitasopposedtoitsalternatives,regardlessofwhetherornottheagentknows,oriseveninapositiontoknow,whattheyare.</p><p>By contrast, the subjective rational status of an act depends, not onwhatreasons there are, but on what reasons the agent takes there to be or,alternatively, on the practicalmental functioning of the agent.On the formerview,anact issubjectively irrational ifandonly if theagenthasbeliefswhosetruthwouldgiveherdecisivereasons torefrain fromperforming it.Andwhentheagenthasinconsistentbeliefs,theactwillbesubjectivelyrationalrelativetosome beliefs, but subjectively irrational relative to others (Parfit 2008).On thelatterview,anactissubjectivelyirrationalifandonlyifitindicatessomefailurein thepracticalmental functioningof theagent (Gert2004,160). Iwont takeastand onwhich, if either, account of subjective rationality is correct. Formypurposes, it is importantonly thatwehavea clearunderstandingofobjectiverationality,ofwhichInowofferatheory.</p><p>Atanygivenmoment,ti,priortodeath,therearevariouswaysasubject,S,might liveout therestofher life.ForeachpossiblewayofcompletingSs life,there isawholeseriesofsuccessiveactions thatSwillperform ifandonly ifScompletesherlifeinexactlythisway.Eachofthesepossiblewaysofactingoverthe remainder of Ss life iswhat I call a future course of actionan FCA forshortavailable to S at ti.10 The theory of objective rationality that I endorseholdsthatwhetherornotitisrationallypermissible(i.e.,objectivelyrational)foran agent to perform a given act is a function ofwhether or not any of theavailable FCAs that include her performing this act are themselves rationallypermissible.11CallthistheFutureCourseofActionTheoryof...</p></li></ul>