PREFERENCES NEITHER BEHAVIOURAL NOR ENTAL
Transcript of PREFERENCES NEITHER BEHAVIOURAL NOR ENTAL
DIPARTIMENTO DI ECONOMIA, MANAGEMENT E METODI QUANTITATIVI
Via Conservatorio 7 20122 Milano
tel. ++39 02 503 21501 (21522) - fax ++39 02 503 21450 (21505) http://www.economia.unimi.it
E Mail: [email protected]
PREFERENCES: NEITHER BEHAVIOURAL NOR MENTAL
FRANCESCO GUALA
Working Paper no. 5/2017
FEBRUARY 2017
FRANCESCO GUALA
Working Paper n. 2011-18
SETTEMBRE 2011
ARE PREFERENCES FOR REAL?
CHOICE THEORY, FOLK PSYCHOLOGY,
AND THE HARD CASE FOR COMMONSENSIBLE REALISM
FRANCESCO GUALA
Working Paper n. 2011-18
SETTEMBRE 2011
1
Preferences:NeitherBehaviouralnorMental*
FrancescoGuala
UniversityofMilan,Italy
Abstract:Recentdebatesonthenatureofpreferencesineconomicshavetypicallyassumedthattheyare
to be interpreted either as behavioural regularities or as mental states. In this paper I challenge this
dichotomyandarguethatneitherinterpretationisconsistentwithscientificpracticeinchoicetheoryand
behaviouraleconomics.Preferencesaredispositionswithamultiplyrealizablecausalbasis,whichexplains
whyeconomistsarereluctanttomakeacommitmentabouttheirinterpretation.
Wedon’tsay‘solubilityisacertainphysical-chemicalstructure’,butrather
thatthesolubilityofthosesubstancesthataresolubleisexplainedbytheir
possessionof a certainphysical-chemical structure. Similarly […] ifwe say
[…] that preferring A to B is ‘synthetically identical with’ possessing a
certain physical-chemical structure […] then we let ourselves in for what
seemstomeremarkableand insufficientlymotivatedextensionsofusage.
(Putnam1975:417)
1.Introduction
Somephilosophicalcontroversiesneverdie.Thedebateabouttheinterpretationofpreferencesandutility1
ineconomicsflaredupattheendofthenineteenthcentury;2inthe1940sand1950seconomistsseemed
to converge on a shared interpretation, but the consensuswas fragile:with the rise of behavioural and
experimentaleconomics,thecontroversyhasrestartedforcefullyduringthepastdecade.3Themainbone
* TheideasthatformthecoreofthispaperwerepresentedatseminarsheldattheuniversitiesofBayreuth,
Innsbruck,Helsinki,Paris1,andTurin.I’mgratefultoMikaelCozic,FranzDietrich,RobertoFumagalli,CyrilHedoin,ChristianList,CaterinaMarchionni,GianlucaPozzoni,BobSugden,JackVromen,PhilipWichardt,andmanyotherparticipantstotheaboveseminarsfortheirsuggestions,commentsandcritiques.IvanMoscatideservesspecialgratitude,forseveralconversationsthatmotivatedthewritingofthispaper.Allthemistakesare,ofcourse,myexclusiveresponsibility.
1 AlthoughIwillusetheterm‘preference’morefrequently,Iwilltakepreferencesandutilitytoberoughlysynonymous,inlinewiththestandardterminologyofcontemporarychoicetheorists(autilityfunctionisanindexofpreferences).
2 Forahistoricalreconstructionofsomekeyaspectsoftheoldcontroversy,seeLewin(1996),BruniandSugden(2007),Moscati(2013a,2013b,2015).
3 AnincompletelistofcontributionstotherecentdebateincludesHausman(2000,2008,2012),Mongin(2000a),Dowding(2002),Glimcheretal(2005),Ross(2005,2011,2014),Camerer(2008),GulandPesendorfer(2008),
2
ofcontentioniswhetherpreferencesandutilityshouldbegivenabehaviouralormentalinterpretation.Do
these theoretical constructs refer to overt behaviour, patterns of choice that are readily observable and
intersubjectively measurable? Or should we take them to describe psychological states, hidden in the
mindsofdecisionmakers?
These are generally assumed to be the only plausible interpretations. But the assumption is
dubious: inthenextfewpagesIwillmakeaproposalthatcutsacrossthebehavioural/mentaldichotomy
andtriestoreconciletheinterpretationofutilitywiththebestscientificpracticeineconomics.Iwillargue
that the two standard positions in this debate – ‘behaviourism’ and ‘mentalism’ – are to be rejected.
Preferences in the economic theory of choice are not mental states: they are dispositions that can be
realized in different ways depending on the circumstances of choice and on the characteristics of the
decision-maker.Iwillarguethatthisclaimdoesnotimplyacommitmenttoabehaviouristinterpretationof
preferences, however. On the contrary it is entirely compatible with the idea that psychology plays an
importantrole inchoicetheory.Butthisdoesnot implythatpreferencesorutilityaretobetransformed
intopsychologicalconstructs,inspiteofwhatsomeeconomistsadvocate(andothersdread).
Theargumentreliescruciallyonmultiplerealization,andonthe‘wide’applicabilityofchoicetheory
beyondthetraditionaldomainofhumandecisionmaking.Thisextensionrequiresthat‘preference’isgiven
anabstractinterpretation,alongthesamelinesasconceptslike‘force’inphysicsor‘fitness’inbiology.The
discoverythattherearedifferentkindsofforcespromptedthedevelopmentofdifferenttheoriessuchas
electromagnetism,thatnowcomplementthetraditionaltheoryofgravitation.Similarly,differenttheories
ofpreferencesoughttoexplainthepreferencesofdecisionunitswhoseinternalprocessesareunlikelyto
beexplicatedbytheprinciplesofhumanpsychology.
Thepaperisorganizedasfollows:sectiontwosummarizesclassicbehaviourismandexplainswhyit
isuntenable.Thissectionisrelativelyshortandunoriginal.ItwilloutlinewhatItaketobethefundamental
flawofbehaviourism,refertotherelevantliterature,andsavespaceforthediscussionofmentalisminthe
subsequent sections. Section three explains what role preferences play in economic explanations of
behaviour, in (broadly) dispositional terms. Section four shows that this interpretation is consistentwith
scientific practice in behavioural economics, focusing on the case of Prospect Theory. In section five I
illustrate how intransitive preferences can be multiply realized, using examples from human decision
makingandcommitteedeliberation.Sectionsixsummarizestheargumentandconcludesthemainbodyof
thepaper.Thelasttwosectionsareoptional,andwrittenmainlyforthesceptics.Sectionsevendefusesa
commonworry, namely that the proposed interpretation of preferencesmakes the economic theory of
choice void of any substantial content. Section eight criticizes the attempt to defendmentalism from a
functionalistperspective.
Hands(2009,2013,2014),Vromen(2010),Lehtinen(2011),Guala(2012),Fumagalli(2013),Clarke(2016),DietrichandList(2016),Okasha(2016),Angner(unpublished),Cozic(unpublished),NagatsuandPoder(unpublished).
3
2.Background:whypreferencescannotbejustbehaviour
Theinterpretationofpreferencesandutilitythatbecameorthodoxinthemiddleofthetwentiethcentury
isadirectdescendantofphilosophical‘positivism’.Positivism,tobesure,tookvariousforms:whilebland
positivistsmerelystressedthe importanceofempiricalevidencefortheproductionofknowledge,radical
ones advocated the elimination of all scientific terms that are directly unobservable or irreducible to
observable concepts. In physics, for example, operationalists tried to reduce themeaning of theoretical
statementstothesetofactionsthatscientistsperformwhentheymeasureaparameterinthelaboratory
(e.g. Bridgman1927).An electron according to this interpretation is not a small particlewith a negative
electriccharge,butthesetofoperationsweperformwhenweobserve,say,tracksinabubblechamber.In
psychology, behaviourists aimed at re-interpreting every psychological concept in terms of measurable
behaviour(e.g.WatsonandMcDougall1929).Termslike‘hunger’or‘pain’,shouldbereplacedbylawsthat
connectexternalphysicalstimuli(likefooddeprivation)withovertbehaviour(likefoodseeking).
Similar ideas played a prominent role in choice theory, where positivistic-minded economists
attempted(1)toreplace‘psychophysical’utility(pleasure)withanindexofpreferences;and(2)toreduce
preferencestoobservedchoice(behaviour).Thefirstpartoftheprogrammewasaccomplishedbyvarious
theorists(notablyPareto1906,andHicksandAllen1934)overthecourseofthirtyyears,andhasbecomea
pillarofmodernmicroeconomics.Thesecondpart,incontrast,provedtobemoreproblematicrightfrom
the beginning, and led to ambiguities that still hamper contemporary discussions on the status of
preferencesandutility.
ThebestknownapplicationofbehaviourismineconomicsisPaulSamuelson’stheoryof‘revealed
preferences’(Samuelson1938,1948,1950).Samuelson’sworkwaspartlymotivatedbyhisdissatisfaction
withHicksandAllen’spreference theory. Samuelsonpresentedhisproposal asa ‘direct attackuponthe
problem,droppingoffthelastvestigesoftheutilityanalysis’(Samuelson1938:62).WhereasinHicksand
Allen’s approach the concept of preference is primitive, Samuelson took choice as primitive.He showed
that the standard analysis of consumer’s behaviour (Neoclassical demand theory) only requires a simple
restriction(the‘weak’axiomofrevealedpreferences)onagents’choices:‘ifanindividualselectsbatchone
overbatchtwo,hedoesnotatthesametimeselecttwooverone’(Samuelson1938:65).Samuelsonthus
aimed at obtaining, without even mentioning utility or preferences, the same results that had been
achievedearlierbyPareto,HicksandAllen.4
Revealedpreference theorywas refinedanddeveloped invariousdirectionsduring the following
twodecades,5andSamuelson’s frameworkhasbecomeastandard tool inadvancedeconomic textbooks
4Althoughthe‘behaviourist’interpretationofSamuelson’sprojectismajoritarian,ithasitscritics.Mongin(2000a)for
examplearguesthatSamuelsonneverintendedthetheoryofrevealedpreferencestoreplaceordinalutilityanalysis.Foradetaileddiscussionoftherelationshipbetweenrevealedpreferenceandordinalutilitytheory,seeCozic(unpublished).
5 Seee.g.Samuelson(1948),Houthakker(1950),Afriat(1967).Hands(2013)proposesausefultaxonomyof
4
(e.g.Mas-Colelletal1995,Varian2005).Butinspiteofitsapparentsuccess,theoriginalpositivistproject
wasneverfullyachieved.Therearemanyreasonsforthis,6butthemostfundamentalonehastodowith
its narroworiginal scope: Samuelson’s revealedpreference approachwasmeant tobe a contribution to
ordinal utility theory, andhenceworkedwithin the domain of riskless decisions. The formalmapping of
preferencesontobehaviourwasattained inabelief-less context, inotherwords, taking forgranted that
consumershaveperfectinformationabouttheobjectsofchoice.Butifthelatterconditionisdropped,itis
easytodemonstratethattherecannotbeaone-to-onecorrespondencebetweenpreferenceandchoice.
Considerthefollowingtwocases:
(a)TonypreferstherestaurantPizzaVesuviotoPizzaBellaNapoli.HebelievesthatPizzaVesuviois
closedtonight.Therefore,hegoestoPizzaBellaNapoli.
(b)VincentpreferstherestaurantPizzaBellaNapolitoPizzaVesuvio.Hebelievesthatbothareopen.
Therefore,hegoestoPizzaBellaNapoli.
AlthoughTonyandVincent’schoicesareidentical,itwouldbeamistaketoconcludethatTonyand
Vincehavethesamepreferences.Thereasonisthat,inthestandardtheoryofeconomicchoice,behaviour
is determined both by preferences and by beliefs. As a consequence, the same choice (behaviour)may
resultfromdifferentpreferences,ifbeliefsalsodiffer.
Theproblemisubiquitous.TonymaydecidenottocontributetotheorganizationoftheChristmas
fairbecausehebelievesthatnooneelse intheneighbourhoodwillgiveanymoney,althoughheregrets
thatthefairwillnottakeplace.VinceincontrastdoesnotcontributebecausehehatesChristmasfairsand
wouldn’tcontributeevenifeveryoneelsedid.Bylookingattheirbehaviourwecannotdeterminewhether
thesituationthateachofthemisfacingisapublicgoodsgameornot,becausewedonotknowhowtheir
choicesdependontheirbeliefsaboutotherpeople’spreferencesandbeliefs.
Oneoption, of course, is to addbeliefs to the background conditions that allow themapping of
preferences onto choices. Such a move is eminently sensible and scientifically correct, but defies the
behaviourist goal: unless the beliefs are defined in behavioural terms, the revealed preference theorist
wouldmakeuseofthesortofpsychologicalconceptsthatsheintendedtoeliminateinthefirstplace.But
exactlythesameargumentholds,symmetrically,forbeliefs:itispossibletoderivebeliefsfrombehaviour
onlyifweknowthepreferencesofdecision-makers.Thebehaviourististrappedinamethodologicalcircle
thatdoesnothaveasatisfactorywayout.
ThispointhasbeenthoroughlymadebyAlexRosenberg (1993),DanHausman (2000,2012),and
others, so I will not devote much time to it here. Some economists have proposed sophisticated
interpretationsof choice theory thatattempt to circumvent theproblem in ingeniousways,butnoneof
differentversionsoftherevealedpreferenceproject.
6 ForacomprehensivereconstructionoftheevolutionofSamuelson’sposition,seeHands(2014).
5
them is able to recover the behaviourist programme without betraying its original goal.7 Without
informationalinputabout‘primitive’variables(i.e.,variablesthataredefinedinnon-behaviouralterms),it
isimpossibletotheorizeincoreareasofeconomicssuchasdecisionandgametheory.
Analternativestrategyistoabandonradicalbehaviourismandendorseaweakerinterpretationof
revealedpreferencetheory.Thisweakerbutmorereasonableinterpretationofpreferenceshasalwaysco-
habited with the radical behaviourist one: instead of claiming that preferences are behaviour, many
economists take Samuelson to mean that preferences aremanifested in behaviour. The latter view is
implied by Samuelson’s choice of terminology,8 and is consistent with a moderate empiricist attitude:
although preferences and choices are different things, choice is the empirical evidence that economists
traditionallyusetotesthypothesesderivedfrombeliefs,desires,andoptionsets.9
But endorsing themoderate interpretation shifts the problem one step beyond.What is choice
behaviour evidence of? If preferences are not patterns of choice, then what are they? What is the
alternative to a behaviourist interpretation?Does rejecting behaviourism imply thatwe should accept a
‘mentalistic’interpretationofpreferences?
3.Explainingwithpreferences
Most contemporary economists do not care much about behaviourism. Contemporary supporters of
revealedpreference theory aremore interested in raisingbarriers against the intrusionof psychology in
economics,thaninrevivinganout-datedphilosophicalprogramme.10Mostofthemhoweverstartfroma
common–andseeminglyplausible–presumption,namely, that the failureofbehaviourismwouldopen
thedoorstomentalismineconomics.Themainthesisofthispaperisthatthisworryisunjustified:thereis
roomforathirdposition,accordingtowhichpreferencesareneitherbehaviouralnormental.
Iwillarguelaterthatthethirdpositionisconsistentwithbestscientificpractice.Itis,inparticular,
consistentwiththewayinwhichtheconceptsofpreferenceandutilityareusedinbehaviouraleconomics
–ascientificprogrammethatisentirelydevotedtopromotetheintegrationofpsychologywitheconomics.
Iwillarguethatalthoughbehaviouraleconomicshasundoubtedly introducedpsychologicalconceptsand
mechanismsineconomicscience,ithasgivennoreasontointerpretutilityorpreferencesasmentalstates.
Preferences,asweshallsee,areinterpretedinbehaviouraleconomicsprettymuchinthesamewayasin
7 SeeHausman(2000,2012),Hands(2013)fordiscussion. 8 TheterminologywasintroducedinSamuelson(1948).Theexpression‘revealedpreference’doesnotevenoccurin
Samuelson’sfirst(1938)article. 9 Thisdistinctionbetweenastrongandweakformofrevealedpreferencetheoryisnowfirmlyentrenchedinthe
literature:seee.g.Hausman(2000,2012),Hands(2013),DietrichandList(2016),and,forbehaviourismmoregenerally,alsoMoore(2001).
10 GulandPesendorfer(2009)aretypicalinthisrespect.Anotherconcernistopreservethe(alleged)value-neutralityofeconomicscience;ontheinterpretationofpreferencesandthenormative/positiveusesofchoicetheory,seee.g.Hausman(2012),Hands(2013,section4)andOkasha(2016).Sen(1982)istheprecursorofmostcontemporarydiscussionsontheuseofpreferencesinnormativeeconomics.
6
traditional (rational) choice theory. As a preliminary step, therefore, it is necessary to explain how
preferencesaretypicallyusedbyeconomiststoexplainchoice.
Preferencesprovideinformationabouttherelativeattractivenessofdifferentstatesofaffairs.They
explain,forexample,whyTonyusuallydinesatPizzaVesuvioratherthanBellaNapoli(atVesuviotheysella
tastier, cheaperMargherita pizza), but also why he would change his habits if Vesuvio charged 5 extra
dollars for his favourite pizza. In general, preferences are explanatory relevant and help formulating
counterfactualclaimsaboutfutureorhypotheticalscenarios,whichmayinformthedecisionsofscientists
and policy-makers.11 At the same time, preference-based explanations are not trivial and competewith
alternativeexplanationsinthescientificarena.Tony’scocainehabitforexampleescapeschoice-theoretic
analysis,becauseitisinsensitivetovariationsinthecostsofconsumption.12
Evenwhentheyaregenuinelyexplanatory,however,preferencesdonotprovideinformationabout
manyinterestingquestions.Theydonottellus,forexample,how–throughwhichcausalmechanism– a
given price variationmay affect Tony’s behavioiur. They tell us that A (an agent) does B (engages in a
certainbehaviour)inC(asetofcircumstances),withoutsayinghowBandCarecausallyrelated.Adoptinga
venerableterminology,throughoutthepaperIwillassumethattosaythat‘AdoesBincircumstancesC’is
equivalenttosayingthatAhasadispositiontodoB.13Iwilltakethisasastipulativedefinition,ratherthan
an analysis of the everyday concept of ‘disposition’. The point is not to solve an ancient philosophical
disputeaboutadifficultmodalnotion,buttoborrowauseful terminologythathelpsarticulatethemain
features of preference-based explanations, and show that they share such featureswith other scientific
explanations.
The set of circumstances C is called the ‘trigger’ of B. For example, to say that crystals of NaCl
dissolveifyouputtheminwaterisequivalenttosaythatsaltissoluble,andwatertriggersthedissolution
ofsalt.Thankstomodernchemistryweknowquitealotaboutthisprocess:roughly,mixingNaClwithH2O
causesthedissociationofNaClintoions.Butknowledgeofdispositionsisinformativeandexplanatoryeven
ifwedonotknowthedetailsoftheunderlyingcausalprocesses.ThusifIsaythat‘thevaseisfragile’,Iam
suggestingthatweshouldhandleitwithcare.Butknowledgeofthecausalmechanismthatmayleadtothe
breakingofthevaseintolittlepiecesisnotnecessaryformanypragmaticpurposes.
Dispositionalviewsofpreferencesaretraditionallyassociatedwithbehaviourism.Thebehaviourist
is attracted to dispositions for epistemic reasons: because the causal basis is partly internal and
11 Onexplanatoryrelevanceandcounterfactualreasoningseee.g.Woodward(2008). 12Iamsimplifyinghere:strictlyspeakingthisistrueonlyifVince’sdemandcocaineiscompletelyinelastic,whichis
rarelytrueevenforharddrugaddicts(Hyman2009).Manycriticsoftraditionalchoicetheoryfocusoncasesthatareconceptuallyanalogoustothis:themaineffectofmoralnormsandcommitments,forexample(Sen1977),istomakespecificbehaviours(e.g.honestdealings)inelastictovariationsincostsandbenefits(e.g.bribes).Butagain,strictlyspeaking,suchcasesfalloutsidetherealmofchoicetheoryonlyincaseofcompleteinelasticity.
13 Idonotpretendtobeparticularlyoriginal:thereisanoldtraditioninphilosophy,goingbacktoGilbertRyle,thatidentifiesmentalstateswithdispositions.Sincepreferencesinchoicetheoryfulfilasimilarroleasdesiresinfolkpsychology,itisnotsurprisingtofinddispositionalaccountsofpreferencesinthecontemporaryliterature(e.g.Sugden1991:762;VanderbeekenandWeber2002;Dowding2002;Hausman2012;Cozic,unpublished).
7
unobservable, thebehaviouristdespairs thatwewill everknowanythingabout it. If ‘AdoesB inC’only
refers to observable events or states, the epistemicworry disappears.Unfortunately, however,wehave
seen that the connection between choice and preferences in choice theory is mediated by beliefs.
Preferences are ‘belief-dependentdispositions’ (Hausman2012:33), that is, the sortofdispositions that
willnothelpthebehaviouristproject,becauseunobservablestatesareincludedinC.Toavoidconfusion,
philosopherstendtouseacompletelydifferentlabel–‘functionalism’–fortheideathatmentalstateslike
desiresaredefinedpartlyintermsofcausalrelationswithothermentalstates(e.g.Block1980).ButsinceI
intendtoarguethatpreferencesarenotmentalstates,Iwillavoidthatterminologyhere.14
When it is used to explain, a disposition typically brackets the causal basis that connects C (the
trigger)withB(thebehaviour).Inasingleformula:
C[+causalbasis]→B.
Bracketing the causal basis has advantages and limitations. Starting from the latter, it is obvious
thatdispositionsareunfittoanswercertainexplanatoryquestions.15AsMolièrepointedout,itisuselessto
say that opium has virtuus dormitiva, if the issue is why opium makes you sleep. When the question
concerns the details of the causal process, dispositions do not provide useful information. However,
dispositionalexplanationshavetheadvantagethattheyallowtoexpressgeneralizationswhenthecausal
basisismultiplyrealizable.
Acausalbasisismultiplyrealizablewhenacertaintypeofeventtriggersthesameeffectacrossa
rangeofcircumstances,eventhoughthecausalmechanismsdifferfromcasetocase.Sinceeconomistsare
usually interested in the explanation of aggregate phenomena, dispositional explanations spare us the
trouble of giving extremely complicated and heterogeneous descriptions of the causes of behaviour
(Dowding2002,Ross2005).Butmultiplerealizationmaybetroublesomeeveninthecaseofindividuals:for
example,supposethatTonyoncefeltsickaftereatingpepperonipizza.Sincethen,hehasrefusedtoeatit
in three separate occasions: the first time, Tony recalled the effects of food poisoning and decided
instinctivelynottoeatpepperonipizza,followinghisgutfeeling.Thesecondtime,Tony’sstomachdidnot
react pre-emptively, but Tony deliberated not to eat pepperoni pizza because he remembered the food
poisoning.Finally,by the thirdoccasion,Tonyhaddevelopeda routine toavoidpepperonipizzawithout
eventhinkingaboutit.Inspiteofthedifferentcausalmechanisms,inallthesecasesitiscorrecttosaythat
Tonyhasadisposition(apreference)fornoteatingpepperonipizza.
Inordertosupportscientificgeneralizations,dispositionsmusthavesomestablefeatures.Andin
factmostdebatesinbehaviouraleconomicsconcernthepropertiesthatpreferenceshave.So,eventhough
knowledgeofdispositionsdoesnotrequireknowledgeoftheircausalbasis,understandingthecausalbasis
14Iwillreturntofunctionalismandmentalisminthelastsectionofthispaper.15Foradetaileddiscussionseee.g.Mumford(1998,chapter6).
8
mayhelptomodelthedispositionscorrectly.Themoreinformationwehaveaboutthecausalmechanisms
that connect C with B, the better we should be able to predict and explain the occurrence of B across
varioussituations.
Wecanthusdistinguishbetweentwoexplanatorytasks:atonelevel,economistsareinterestedin
providing explanations of behaviour; at another level,many (but not all) economists believe that giving
suchexplanationsrequiresthatwearealsoabletoexplainpreferences.Anotherwaytoputitistosaythat
preferences arepart of theexplanans (whatdoes theexplaining) at the former level, but constitute the
explananda(whatistobeexplained)atthelatter.Thuspreferencesexplainbehaviourinchoicetheory;but
psychologicalmechanismsexplainpreferencesinbehaviouraleconomics.Or,moreprecisely,psychological
mechanisms explainwhy certain preferences have the shape that they do have in themodels of choice
theory.
4.Forexample:ProspectTheory
The best way to illustrate the difference between explanations of behaviour and explanations of
preferences is to examine a specific model, and no model is better suited for this task than Prospect
Theory.ProspectTheorywasdevelopedbyDanielKahnemanandAmosTversky(1979;seealsoTversky&
Kahneman 1992) with the aim of systematizing a large body of experimental results produced by
psychologistsofdecision-makingduring the1960sand70s. The twomain featuresofProspect Theory–
whichdistinguishitfromtraditionalmodelsofrationalchoice–arethedistortionofsubjectiveprobabilities
(beliefs)andthereference-dependentutilityfunction.Sincepreferencesarethemaintopicofthispaper,I
willignorebeliefsandfocusonthelatteronly.
The classic shape of Kahneman and Tversky’s utility function is represented in Figure 1. The
intersectionofthehorizontalandverticalaxesdenotesthereferencepoint,thatis,thestateofaffairswith
respecttowhichtheagentevaluatesthepossibleoutcomesofherdecision. Inmanycasesthereference
point issimplythestatusquo.Thefirst importantideabehindProspectTheoryisthatthesameoutcome
maybeevaluateddifferentlybythesameindividualdependingonwhetheritisperceivedasagainoraloss
withrespecttothereferencepoint.Thesecondideaisthatutilitydecreasesinthedomainoflossesmore
steeplythanit increasesinthedomainofgains.The‘kink’ intheshapeoftheutilityfunctionreflectsthis
asymmetryintheevaluationofgainsandlosses.
9
Figure1:Reference-dependentutilityinProspectTheory.
Notice that so far there is no reason to interpret thepreferences summarized in Kahnemanand
Tversky’s utility function as psychological states. Although at times they seem to differentiate their
languageonpurpose–theyspeakof‘valuefunctions’ insteadof‘utilityfunctions’,forexample–inmost
respectsKahnemanandTverskydotheirbesttopresentProspectTheoryinthesameformatastraditional
choicetheories.16All thepsychologicalcontent lies in theexperimentalphenomena– like lossaversion–
thatKahnemanandTverskyrelyupontojustifytheunorthodoxshapeoftheirutilityfunction.Butthisdoes
not imply that the functionmustbe interpretedpsychologically. ThepreferencesofProspectTheoryare
notdifferentfromthoseoftraditionalchoicetheory–theydescribebehaviouraldispositions.Theprincipal
differenceisthatProspectTheoryusespsychologicalinformationtoattainamoreaccuraterepresentation
of preferences, whereas traditional choice theory tends to sacrifice realism in favour of normative
considerations(theagentsofthetraditionaltheoryarerationaldecision-makers,firstandforemost).17
WhatisthecausalbasisofpreferencesaccordingtoProspectTheory?‘Lossaversion’or‘reference
dependence’refertoexperimentaleffects,ratherthancausalmechanisms.So it isnotclearwhatkindof
psychologicalprocesses liebehindProspectTheory.Although the theory itself is remarkably silentabout
this aspect, the cognitive science literature offers many useful suggestions. Reference dependence for
examplemaybeexplainedbythefactthatourcognitivesystemisessentiallyadetectorofnovelties.18New
eventsconveypotentiallyimportantinformationregardingopportunitiesandrisks,towhichourcognitive
apparatusmustoftenreactquickly.Onecheapanddirtywaytodoitistoassesstheirpositiveornegative
valencewith respect to the status quo, so as to determine promptlywhether changing one’s behaviour
wouldimproveorworsenone’ssituation.
Loss aversion – the fact that the utility function is steeper in the domain of losses – may be
16Theoriginalpaperincludedanappendixwitharepresentationtheorem,forexample.SeealsoWakker(2010)for
anattempttointegrateProspectTheoryinthetraditionalconceptualapparatusofeconomictheory.17 Ontheimportanceofnormativeprinciplesinchoicetheory,seeSugden(1991),Guala(2000),Starmer(2000),
Okasha(2016). 18 Seee.g.Tiitinenetal.(1994),CorbettaandShulman(2002).
10
explainedby the fact that lossesandgainsareprocessed indifferentpartsof thebrain.19Losses tendto
triggernegativeemotionalreactions,andthe‘lossmodule’maybeover-sensitiveforevolutionaryreasons:
foracreaturelivinginadangerousenvironmentandalwaysonthebrinkofextinction,tryingtoavoidfatal
lossesismoresensible(infitnessterms)thanseekinguncertaingains.Apieceofevidenceinfavourofthis
hypothesis,forexample,isthatlossaversiongenerallyincreaseswhenthebudgetissmall(Camerer2005).
ProspectTheory,thus,isnotatheoryofpsychologicalpreferences–itisapsychologicaltheoryof
preferences.Anyre-interpretationofpreferencesinpsychologicalormentalterms–suchastherevivalof
hedonistic utility recently promoted by Kahneman and others (1997) – therefore must be based on
considerationsthatareindependentfromthosethatmotivatedthedevelopmentofProspectTheory.
5.Whenthecausalbasisvaries:intransitivepreferences
My rejection of the mentalistic interpretation of preferences is based on a distinction between two
different levels of explanation. And the distinction between the two levels is justified by multiple
realizability.Butmultiplerealizability isanempiricalmatter:whatifthecausalbasesofpreferenceswere
alwayspsychological,afterall?Tony’sdislikeofpepperonipizzawaspromptedinonecasebyanemotional
reaction,anothertimebydeliberatereasoning,andthethirdtimebyaningrainedhabitthat,presumably,
mustbeencodedinhisbrain.Inthiscase,theredoesnotseemtobeanyreasontodenythatpreferences
arepsychologicaldispositions.
Somecontemporaryattempts to identify ‘experiencedutility’or ‘utility in thebrain’areprobably
basedon this sortof reasoning.20 Ifnot in themind/brain,wherecouldutilitypossiblybe?Thequestion
seemsrhetorical ifwe interpretchoicetheorynarrowlyasa theoryofhumandecisionmaking,and ifwe
accept the internalist prejudice that pervades contemporary behavioural economics. But if these
assumptions are challenged,21 it is easy to see that (i) there may be different causal bases for the
preferencesofdifferentagents,and(ii)notallthebasesarenecessarilypsychological.
I will proceed again by way of an example. Experimental and behavioural economists have
demonstrated convincingly that preferences in some circumstances may be intransitive, and have
describedvariousmechanismsthatmayexplaintheseintransitivepatterns.Onemechanismwasidentified
a long time ago by Tversky and other choice theorists working on so-called ‘multi-attribute’ utility
functions.ImagineadecisionmakerfacingachoicebetweenthreeoptionslabelledX,Y,andZ.Eachoption
moreover is characterized by three attributes. The options and the attributesmay be anything that the
decision maker cares about, but to make the example more concrete let us suppose that the choice
concerns three typesof car characterisedbydifferentdegreesof comfort, speed, andprice.Now, letus
19 E.g.Blackfordetal.(2010) 20 Seee.g.Kahnemanetal.(1997),Glimcheretal.(2005),and–foracriticalperspective–Vromen(2010),Fumagalli
(2013). 21 Foraforcefulcritiqueofbothassumptions,seeRoss(2005).
11
supposethatthecars(X,Y,Z)arerankedoneachdimensionorcharacteristicasfollows:
1.Comfort:X>Y>Z
2.Speed:Z>X>Y
3.Price:Y>Z>X
The>symbolstandsfora(transitive)‘more…than’relation:forexample,alongthefirstdimension
carXismorecomfortablethanY,whichismorecomfortablethanZ.Now,sincenocarisbetterthanthe
others in all respects, the decision-maker faces the problem of weighing each characteristic and
aggregating the evaluations to form a single preference ranking. This task may be cognitively quite
demanding, so when presented with pairwise choices people tend to use a simple heuristic instead of
workingthroughtheaggregationprocedure.Theheuristicgoesasfollows:wheneveryouhavetomakea
decisionamongtwooptions,choosetheonethatbeatstheotheralongmostdimensions.Appliedtothis
particularcase,therulewouldgivethefollowingoutcomes:
XbeatsYincomfortandspeed,butlosesinprice.
YbeatsZincomfortandprice,butlosesinspeed.
ZbeatsXinspeedandprice,butlosesincomfort.
Using the simple heuristic thuswould generate intransitive preferences: X is preferred to Y, Y is
preferredtoZ,butZispreferredtoX.22
Themechanism behind preference formation, in this example, is clearly psychological: themain
reason to use the simple rule is to save cognitive effort. Each one of us can appreciate the lure of this
heuristicby introspection,andexperiments (e.g.Tversky1972)havedemonstrated thathumandecision-
makers have a tendency to engage in mono-dimensional evaluation before (or instead of) aggregating
across the attributes. A mentalistic interpretation of the causal basis of preferences in this case is
empiricallyandtheoreticallyjustified.
Itiseasyhowevertofindexamplesofintransitivepreferencesthatarenotbasedonapsychological
mechanismofthissort.Foreaseofcomparison, letusexamineacasethat is formallyverysimilartothe
previousone:auniversityhiringcommitteemustdecidehowtorankthreecandidates(DrSmith,DrBrown,
andDr Jones) for an academic position. The committee is constituted by threemembers,who rank the
candidatesasfollows:
1.Smith>Jones>Brown
22 Foraclassicexperimentalstudyofthisphenomenon,seeMay(1954).Otherexcellentdiscussionsofintransitivity
canbefoundinTversky(1969)andMongin(2000b).
12
2.Brown>Smith>Jones
3.Jones>Brown>Smith
Now, let us suppose that according to university rules the winner must be determined via a selection
processthatinvolvesasequenceofpairwisecomparisons,andthateachcomparisonmustberesolvedby
majorityvoting.Whenpresentedwithpairwiseoptions,thecommitteevotesasfollows:
SmithbeatsJones2-1
JonesbeatsBrown2-1
BrownbeatsSmith2-1
Thisisawell-knownexampleof‘Condorcetvoting’,aproblemthathasbeendiscussedextensivelyinchoice
theory.23 Condorcet voting is an interesting case of intransitivity generated by voting rules. A non-
psychologicalmechanismhereproducesanpeculiarpatternofpreferences.Althoughformallyanalogous,
themechanismthus is substantiallydifferent fromtheone thatexplains intransitivities inmulti-attribute
choice.The firstdifference is that thedecision-maker isanorganization rather thanan individualhuman
being.Thesecondoneisthatintransitivepreferencesaregeneratedbyaninstitutionalrule,ratherthanby
a heuristic.24 Any attempt to fix the problem, therefore, should better take this fact into account: the
committee does not suffer from bounded rationality, does not follow fast-and-frugal heuristics, and
transitivity cannot be restored by, say, changing theway inwhich the options are framed. An effective
interventionmusttargettheinstitutionalfeaturesoftheenvironment,forexampletherulesthatforcethe
committeetofollowamajority-votingprocedure.
Inthiskindofexternalistexplanation,thepreferencesofthecommitteearedeterminedtoalarge
extentbysomefeatureoftheenvironmentinwhichthecommitteemembershavetomakeadecision.The
causalbasisofpreferencesisnotalways,ornoteventypically,constitutedbypsychologicalprocesses.
6.Takingstock
The controversy on the interpretation of utility is one of the oldest debates in the foundations of
economics.Itsresilienceisduetoanunresolvedtensionbetween,ontheonehand,economists’aspiration
totheorizeatadifferentlevelfrompsychology,and,ontheother,thefactthatpsychologicalevidencehas
been an important source of information to build modern microeconomic theory. This has become
23 Seee.g.Farquharson(1969),ListandPettit(2011). 24 Thereare,ofcourse,psychologicalelementsinthecausalchainthatleadstothecommittee’spreferencestructure
(suchastheevaluationsoftheindividualcommitteemembers).Butsuchpsychologicalelementscannotexplaintheintransitivityofpreferences–not,atleast,withoutalotofhelpfromtheinstitutionalrules(seealsoRoss2005,2014).
13
particularlyevidentduring the last threedecades, since theriseofbehaviouraleconomicshas led to the
developmentofpsychologically-informedmodelsinmainstreameconomics.
Seeninthislight,theresistanceagainstmentalismmayseemallthemoreobsolete.Nonetheless,I
havearguedthatthenotionofpreferenceisusedinchoicetheory(includingbehaviouralchoicetheory)in
awaythatdoesnotcompelapsychologicalinterpretation.Thereasonisthatpreferencesaredispositional
termswithamultiplyrealizablecausalbasis.Becausetheyarebelief-dependentdispositions,behaviourism
fails.Butbecausetheyaremultiplyrealizable,mentalismdoesnothold.Insomecircumstances,especially
in the traditional domain of human decision-making, the causal basis is indeed mainly constituted by
psychologicalmechanisms.Butinothercasesitisnot.Thisexplainswhysomeeconomistsresistthenarrow
identificationofchoicetheorywiththepsychologyofdecision-making.
This is not an unscientific or a conservative position. Like ‘force’ in physics, ‘preference’ is an
abstract concept that must be filled with content in each specific domain of application. Like physical
forces,wehavegoodreasonstobelievethatpreferencesarerealexplanatoryvariables.Butwhattheyare,
exactly, issomethingthatscience–notphilosophy–will tell.Todiscoverhowmanykindsof forcethere
are,andhowtheywork,hasbeenan important task in thehistoryofphysics.Similarly, themechanisms
underlying preferences are potentially heterogeneous and new ones may be discovered with the
developmentofthisfieldofscience.
Thisbringsanendto themainargumentof thispaper.Withmultiple realizability, thecritiqueof
the behavioural/mental dichotomy is complete. Multiple realizability however presupposes a ‘wide’
interpretationofchoicetheory–asatheorythatisapplicabletothebehaviourofdecision-makersthatare
notnecessarilyhuman.Forthosewhodonotfindthispresuppositionpersuasive,therestofthepaper is
devotedtodispelafewworriesandobjections.InthenextsectionIaddressthefamiliarcomplaintthata
wide interpretation makes the theory of choice empty of explanatory content. In the final section, I
examinetheclaimthatbehavingconsistentlywiththeprinciplesofchoicetheoryisamarkofthemental–
andthusthatanyeconomicagentnecessarilyhasamind.
7.Firstcoda:thedomainofchoicetheory
There is no doubt that, historically, human decision-making has been the core domain of application of
microeconomics.Themodern theoryof rationalchoice isadirectdescendantofNeoclassicalconsumer’s
theory, and its extension to non-human behaviour is a relatively recent development. But today choice
theory can be and is routinely used to explain the behaviour of non-human animals, organizations, and
even artefacts. Onemayworry that these extensions of the theory compromise its explanatory power.
What is the purpose of attributing preferences and beliefs to robots, political parties, or fish? Political
partiesandrobotsmaybehave‘asif’theyhadpreferences,butsurelytheydonotreallywantanythingin
14
particular.Itisus,theobservers,whofinditusefultopostulatesuchtheoreticalconstructs;butthelatter
donotreallyrefertopreferencesinthegenuinesenseoftheterm.
Theproblemwiththisargument,ofcourse,isthatoneneedstoexplainwhat‘genuine’means.Ifby
‘genuine’ we mean a state generated by mental or psychological mechanisms such as those that are
implementedinhumanbrains,thencommitteesdonothavepreferencesinthissense.Butsincetheissue
thatwehavestarted from iswhetherpreferences (including thoseofhumandecision-makers) shouldbe
interpretedasmentalstates,clearlythisobjectionwouldbegthequestion.
Analternativewaytoproceedisthis:insteadofsearchingforhuman-likementalstates(whatever
theymaybe)innon-humandecision-makers,weshouldaskwhethertheattributionofpreferencestoany
(human or non-human) agent provides useful information about its behavioural dispositions. Are
preference-based explanations entirely ad hoc, or do they provide information that other kinds of
explanationfailtodeliver?Iftheywereadhoc,clearlythetheorywouldbevoidofsubstantialcontentand
oflittlescientificinterest.
But choice-theoretic explanations are not empty. They provide useful information about the
behaviour of creatures that are (1) consequence-driven, (2)motivated to pursuedifferent goals, and (3)
abletocomparethevalueofsuchgoals.25Anotherwaytoputitisthatchoicetheoryisvaluablyappliedto
creatures, artefacts, or organizations who possess a ‘control centre’ where information is processed,
consequences are evaluated, and trade-offs about costs and benefits aremade.How the control centre
works,orwhatitismadeof,isanissuethatmayinfluencewhichmodelofchoiceislikelytoworkbestin
eachspecificcase,butthatdoesnotprecludetheapplicationofchoicetheoryperse.26
Itmaybehelpfultoreviewafewcasesthatfalloutsidethedomainofthetheory,beginningwith
behaviour that is not consequence-driven (condition 1). When the doctor taps my knee with a little
hammer, my thigh muscle stretches and causes a jerk in the knee. Although the movement may be
described in choice-theoretic terms, such a descriptionwould bead hoc andwouldmischaracterize the
process. The knee-jerk reaction is a reflex: it does not result from a process of evaluation of the
consequencesofmyaction.
Multiplegoals(condition2):reflexbehaviourisnotalwayseasilydistinguishablefromothersimple
forms of decision-making. Russell’s famous chicken, who runs toward the feed until the day its neck is
wrung, isabletoevaluateonlyoneconsequenceof itsaction.Butempiricallytheredoesnotseemtobe
muchdifferencebetweenmono-consequentialistchoiceandreflexbehaviour:thechickeninaveryobvious
sense cannot help but run toward the farmer every time it sees him approaching, because it has been
programmed(orhaslearned)toevaluateonlythistinyaspectofthecomplexlifeofthefarm. 25ThereareclearanalogieswithRobbins’(1932)famousdefinitionofeconomicsas‘thesciencewhichstudieshuman
behaviourasarelationshipbetweenendsandscarcemeanswhichhavealternativeuses’:consequence-drivenbehaviourisbehaviourdirectedtowardends,andalternativeusesareanalogoustomultiplegoals.ForahistoricalperspectiveonRobbins’definition,seeBackhouseandMedema(2009).
26 Forasimilartakeontheimportanceofhavingacontrolcentre(or‘integratedworkspace’,inhisterminology)seeGodfrey-Smith(2013).
15
Weighing (condition3):a slightlymore intelligentanddiffidentcreaturewould take intoaccount
theprobabilitythatthefarmer isgoingtostring itsneck.Butthiswouldrequirethecapacitytocompare
differentoutcomes–suchasbeinghungryandbeingdead–togetherwiththeirrelativeprobabilities.The
choice-theoretic apparatus works best when we are dealing with a system that pursues different (and
potentiallyconflicting)goals,guidedbydifferentinformation.Thedecision-makermustbeabletoresolve
conflictsbetweenhergoalsbyweighingtheprosandconsofdifferentcoursesofaction.27
Most human behaviour does, of course, involve the evaluation ofmulti-attribute consequences.
Humans pursue a variety of goals simultaneously, and often face decisions that require the making of
trade-offs between different ‘goods’ and ‘bads’. A prominent humanway tomake such trade-offs is to
think about the consequences of the actions and to engage in a systematic process of comparative
evaluation. But conscious deliberation is by nomeans the only or themost important way tomake an
economicdecision.Whetherthecontrolcentreisinone’sbrain,one’scomputer,orinone’sstomachisof
secondaryimportance.28
That’swhythebehaviourofmanynon-humansystemscanbeexplainedinchoice-theoreticterms.
Weknowthatmiceandcrabshavepreferences–inthegenuinesenseoftheterm–becausewecanfind
their indifference points by varying costs and benefits and observing that their behaviour changes in a
systematicmanner. If laboratorymicearemadetowork toobtaindifferent typesofwater, forexample,
their behaviour fits a well-defined demand curve (Holm et al. 2007). If hermit crabs are given electric
shocks, theyaremore reluctant to leaveagood shell thanapoorqualityone (AppelandElwood2009).
Evenjellyfish,asweshallseeshortly,haveacontrolcentrethat“coordinatesrespondingwhenstimuliare
tendingtoelicitmorethanoneresponse”(Albert2001:474).
8.Secondcoda:functionalismandmentalism
Endorsing a wide interpretation of the domain of choice theory makes the multiple realizability thesis
plausible.Atthesametime,thewideinterpretationdoesnotmakethetheoryemptyortrivial.But,some
philosophersmay argue, a wide interpretationmay defeat the anti-behaviourist, anti-mentalist position
thatIamtryingtodefendinthispaper.Torealizewhyitmightbeso,itisusefultostartfromadistinction
betweentwofundamentalconceptionsofthemental:
It isnowacommonplace,at leastamongphilosophers, that therearetwofundamentallydifferent
conceptionsofmindandmentalstates.Oneisthatthemindisdistinguishedbyitsintrinsicnatureor
character:bywhatitisinitself.Itisdifferentfrommatterbecauseitismadeoutofdifferentstuff,as 27 Boththegoalsandtheweighingprocessesmustbestableenoughtoallowthepredictionofbehaviour.
28Seee.g.Zajonc(1980).Thosewhofailtoaccountforthemultiplecausalbasesofpreferencestypicallyendupdefendinganover-intellectualisedviewofpreferences.SeeforexampleHausman(2012),andAngner(unpublished)foracritique.
16
with Descartes’ res cogitans; or at least it has different intrinsic properties, such as being
phenomenallyconscious[…].
Theotherconceptionisthatmindisdistinguished,notbywhatitisintrinsicallybutbywhatitdoes
extrinsically:bytheroles itplays,thefunctions itdischarges.Onthisconception,tohaveamind is
simplytofunctioninacertainway.(ListandPettit2011:170-171)
Theconception thatmay threaten theanti-behaviouristandanti-mentalistposition is thesecond
one.Accordingtofunctionalists,
Mentalstatesare,at least inpart,statesthatplayacertainroleforanagent.Beliefs, forexample,
play the role of representing certain features of the world from the agent’s perspective, and
preferencesplaytheroleofmotivatingtheagent’sactions[…].Functionalism istheviewthatwhat
makes something amental state is simply that it plays the relevant role. (Dietrich and List 2016:
268)29
Thefunctionalistapproachisnotjustaphilosophicalcuriosity.Onthecontrary,itofteninformsthe
wayinwhichscientiststhinkandtalkaboutthemind.AccordingtoDavidAlbert,ascientistworkingatthe
Roscoe BayMarine Biology Laboratory in Vancouver, for example, some species of jellyfish display the
typical characteristics that we normally associate with the capacity to make choices. The behavioural
repertoireofMoonJellies(Aureliaorscyphozoa)
includeswimmingupinresponsetosomatosensorystimulation,swimmingdowninresponsetolow
salinity,diving inresponsetoturbulence,avoidingrockwalls, formingaggregations,andhorizontal
directional swimming. (…) The central nervous systemofAurelia sp. coordinatesmotor responses
with sensory feedback, maintains a response long after the eliciting stimulus has disappeared,
changes behaviour in response to sensory input from specialized receptors or from patterns of
sensoryinput,organizessomatosensoryinputinawaythatallowsstimulusinputfrommanypartsof
thebodytoelicitasimilarresponse,andcoordinatesrespondingwhenstimuliaretendingtoelicit
morethanoneresponse.(Albert2011:474)
Albert’sarticle is titled ‘What’son themindofa jellyfish?’. It suggests that themechanisms that
governtheresponsesofAureliatodifferentstimulijustifytheattributionofmentalstatestojellyfish.Butif 29DietrichandList(2016)donotendorsefunctionalism,butaweakerpositionaccordingtowhichplayinga
preference-roleismerelyindicative,ratherthanconstitutive,ofmentalstates.Themainreasonisthatmentalstatesmayhaveother,non-functionalcharacteristicsthatcanonlybeascertainedfromafirst-personalperspective.AstrongerfunctionalistpositionisdefendedbyListandPettit(2011),whotakeexclusivelyathird-personalperspectiveonthebehaviourofgroups.I’mgratefultoFranzDietrichandChristianListforclarifyingthispoint.
17
any creature that exhibits choice behaviour in this sense deserves to be attributed a mind, then every
preferenceisamentalstate–contrarytowhatIhavebeenarguingsofar.
Thisargumentisbasedonaprofoundequivocationofthesenseinwhich‘mentalism’isrelevantfor
economics.Asmentionedearlier,thedebateineconomicsconcernsmainlytherolethatpsychologicaldata
andpsychologicalexplanationshouldplayinchoicetheory.Itconcernsawayofdoingscience–ascientific
programme–ratherthanthephilosophicalproblemofdefiningthemental.Itisworthrememberingthat
functionalism originates from a scientific programme too, rather than from the need to resolve a
conceptualissue.Themaingoalofthisprogramme–‘ArtificialIntelligence’–wastounderstandcognition
by modelling the mind as a computational system. A cognitive system, according to this approach, is
essentiallyasoftwarethatcanbe implemented inmanydifferentways.Thehardwaredoesnotmatter–
therecanbemindsmadeofsiliconchips,vacuumtubes,orfleshandblood.As longas ‘something’plays
the right role in their control systems,non-humananimals, robots,plants,andalienscanbe legitimately
(scientifically)saidtohavementalstates.30
Farfrombeingmerelyterminological,thepointoffunctionalismwastoconvincephilosophersand
scientists that it is possible to understand complex cognition without ‘looking under the hood’. More
concretely,itpromisedtounderstandthemindevenwithoutanunderstandingofthewaythebrainworks.
Thesoftwarecanbestudiedindependentlyofitshardware.Thusfunctionalismwasmotivatedprimarilyby
amethodological,ratherthanbyanontologicalconcern.
Noticehowdifferentisthewayinwhichmentalismisemployedinthechoicetheorycontroversy.
Mentalistic interpretations of preferences have been brought to the fore by the rise of behavioural
economics.But themaingoalofbehavioural economistshasbeen topersuade social scientists that it is
importantto‘lookunderthehood’–thatis,tounderstandthepsychologicalandneuralbasisofeconomic
decision-making.Ifthereisamethodologicalcommitment,itisexactlytheoppositethanthefunctionalist’s
commitmentinAIandthephilosophyofmind.
Themainproblemwithafunctionalistdefenceofmentalismthenisthatitwouldbedisconnected
withanyspecificscientificprogramme. It isnotscientificallycontroversial thatsomethingdeterminesthe
choicesofaneconomicagent.Thescientificcontroversyconcernswhatthat‘thing’maybe,andaswehave
seen it is likelythat itdiffers fromcasetocase. Insomecase itmaybeaneuro-psychologicalprocess, in
another itmaybesomethingcompletelydifferent–aswhenwearedealingwithcommitteesor jellyfish.
Given thisdiversity, it seemsunlikely thata single scientific theorywillbeable toaccount for thecausal
basisofpreferencesacross thewholedomainofapplicationof choice theory. This isoneof the reasons
whyeconomistsprefertobenon-committalaboutthenatureofpreferences.Afunctionalisticmentalism
mayappealtophilosophers,butdoesnotmatchscientificpractice.
Anotherwaytoputitistosaythatthereisno‘scienceofthemental’inthefunctionalistsense.Or
rather:theonlyscienceofthementalwouldbechoicetheory.Butthentosaythatpreferencesaremental 30 Seee.g.theessaysonthemind-bodyproblemcollectedinPutnam(1975).
18
would be akin to saying that preferences are preferences – it would not make any difference in any
substantial,methodologicalsense.Behaviouraleconomicsincontrastisascientificprogrammecommitted
tostudythepropertiesandthecausesofhumanpreferences(amongotherthings).Itiscommittedtothe
claimthatsomepreferences (thoseofhumandecisionmakers)havecertainspecific features invirtueof
theunderlyingpsychologicalprocessesthatgovernchoicebehaviour.This isasubstantialhypothesisthat
makes a difference for the way in which economic science is and will be practised. Unfortunately, the
explanations that behavioural economics puts forward are unlikely to be applicable to jellyfish or
committees.Indifferentdomainsdifferentscientifictheoriesarelikelytoberelevant.Butthisissomething
thatascientificaccountofpreferencesmustbereadytocopewith.
When economists argue about ‘mindless’ or ‘mindful’ economics, in contrast, something very
important is at stake. The disagreement revolves around the use of psychology in economics. It has
important implications regarding the training of young economists, the theories and the methods they
shouldbe taught,andwhichoneswillhelpusmakeprogress in themodellingandexplanationofchoice
behaviour.
Economists do not argue about preferences as ‘mental state’ in the functionalist sense. This is
somethingaboutwhichthereisnodisagreement,becauseitdoesnotmakeanydifferenceforthewayin
whicheconomicscienceisdone.
References
Afriat,S.N.(1967)Theconstructionofutilityfunctionsfromexpendituredata.InternationalEconomicReview8:67–77.
Albert,D.J.(2011)What’sonthemindofajellyfish?NeuroscienceandBiobehavioralReviews35:474-482.
Angner,E.(unpublished)Whatarepreferences?
Elwood,R.W.&Appel,M.(2009)Painexperienceinhermitcrabs?AnimalBehaviour77:1243-1246.
Backhouse,R.E.andMedema,S.G.(2009)Definingeconomics:thelongroadtoacceptanceoftheRobbinsdefinition.Economica76:805-820.
Blackford,J.U.,Buckholtz,J.W.,Avery,S.N.,&Zald,D.H.(2010)Auniqueroleforthehumanamygdalainnoveltydetection.Neuroimage50:1188-1193.
Block,N.(1980)Troubleswithfunctionalism.InN.Block(ed.)ReadingsinPhilosophyofPsychology,Vol.1.Cambridge,Mass.:HarvardUniversityPress,261-325.
Bridgman,P.W.(1927)TheLogicofModernPhysics.NewYork:ArnoPress.
Bruni,L.&Sugden,R.(2007)Theroadnottaken:howpsychologywasremovedfromeconomics,andhowitmightbebroughtback.EconomicJournal117:146-173.
Camerer,C.(2005)Threecheers—psychological,theoretical,empirical—forlossaversion.JournalofMarketingResearch42:129-133.
19
Camerer,C.(2008)Thecaseformindfuleconomics.InA.Caplin&A.Schotter(eds.)TheFoundationsofPositiveandNormativeEconomics,NewYork:OxfordUniversityPress,pp.43-69.
Clarke,C.(2016)Preferencesandpositivistmethodologyineconomics.PhilosophyofScience83:192-212.
Corbetta,M.&Shulman,G.L.(2002)Controlofgoal-directedandstimulus-drivenattentioninthebrain.NatureReviewsNeuroscience3:201-215.
Cozic,M.(unpublished)Yetanotherattemptatclarifyingrevealedpreferencetheory.
Dietrich,F.&C.List(2016)Mentalismversusbehaviourismineconomics:aphilosophy-of-scienceperspective.Economics&Philosophy32:249-281.
Dowding,K.(2002)Revealedpreferenceandexternalreference.RationalityandSociety14:259-284.
Farquharson,R.(1969)TheoryofVoting.Oxford:Blackwell.
Fumagalli,R.(2013)Thefutilesearchfortrueutility.EconomicsandPhilosophy29:325-347.
Glimcher,P.W.,Dorris,M.C.&Bayer,H.M.(2005)Physiologicalutilitytheoryandtheneuroeconomicsofchoice.Games&EconomicBehavior52:213-256.
Godfrey-Smith,P.(2013)Cephaloidsandtheevolutionofthemind.PacificConservationBiology19:4-9.
Guala,F.(2000)Thelogicofnormativefalsification:rationalityandexperimentsindecisiontheory.JournalofEconomicMethodology7:59-93.
Guala,F.(2012)Arepreferencesforreal?Choicetheory,folkpsychology,andthehardcaseforcommonsensiblerealism.InA.LehtinenandP.Ylikoski(eds.)EconomicsforReal.London:Routledge,pp.137-155.
Gul,F.&Pesendorfer,W.(2008)Thecaseformindlesseconomics.inA.Caplin&A.Schotter(eds.)TheFoundationsofPositiveandNormativeEconomics,NewYork:OxfordUniversityPress,pp.3-39.
Hands,D.W.(2009)Economics,psychologyandthehistoryofconsumerchoicetheory.CambridgeJournalofEconomics34:633-648.
Hands,D.W.(2013)Foundationsofcontemporaryrevealedpreferencetheory.Erkenntnis78:1081-1108.
Hands,D.W.(2014)PaulSamuelsonandrevealedpreferencetheory.HistoryofPoliticalEconomy,46:85-116.
Hausman,D.M.(2000)Revealedpreference,belief,andgametheory.EconomicsandPhilosophy16:99–115.
Hausman,D.M.(2008)MindlessormindfulEconomics:amethodologicalevaluation.InA.Caplin&A.Schotter(eds.)TheFoundationsofPositiveandNormativeEconomics.NewYork:OxfordUniversityPress,pp.125-151.
Hausman,D.M.(2012)Preference,Value,ChoiceandWelfare.NewYork:CambridgeUniversityPress.
Hicks,J.&Allen,R.G.D.(1934)Areconsiderationofthetheoryofvalue.Economica1:52–76,196–219.
Holm,L.,Ritz,C.&Ladewig,J.(2007)Measuringanimalpreferences:shapeofdoubledemandcurvesandtheeffectofprocedureusedforvaryingworkloadsontheircross-point.AppliedAnimalBehaviourScience107,133-146.
Houthakker,H.(1950)Revealedpreferenceandtheutilityfunction.Economica17:159–74.
Hyman,G.M.(2009)Addiction:ADisorderofChoice.Princeton:PrincetonUniversityPress.
20
Koob,G.F.andLeMoal,M.(2006)NeurobiologyofAddiction.London:Elsevier.
Kahneman,D.&Tversky,A.(1979)Prospecttheory:ananalysisofdecisionunderrisk.Econometrica47:263-291.
Kahneman,D.,Wakker,P.P.&Sarin,R.(1997)BacktoBentham?Explorationsofexperiencedutility.QuarterlyJournalofEconomics112:375-405.
Lehtinen,A.(2011)Therevealedpreferenceinterpretationofpayoffsingametheory.HomoOeconomicus28:265-296.
Lewin,S.(1996)Economicsandpsychology:lessonsforourowndayfromtheearlytwentiethcentury.JournalofEconomicLiterature34:1293–1323.
List,C.&Pettit,P.(2011)GroupAgency.NewYork:OxfordUniversityPress.
Mas-Colell,A.,Whinston,M.D.,&Green,J.R.(1995)MicroeconomicTheory(Vol.1).NewYork:OxfordUniversityPress.
May,K.O.(1954)Intransitivity,utility,andtheaggregationofpreferencepatterns.Econometrica22:1-13.
Mongin,P.(2000a)Lespréférencesrévéléesetlaformationdelathéorieduconsommateur.Revueéconomique51:1125-1152.
Mongin,P.(2000b)Doesoptimizationimplyrationality?Synthese124:73-111.
Moore,J.(2001)Ondistinguishingmethodologicalfromradicalbehaviorism.EuropeanJournalofBehaviorAnalysis2:221-244.
Moscati,I.(2013a)Howcardinalutilityenteredeconomicanalysis,1909-1944.EuropeanJournaloftheHistoryofEconomicThought20:906-939.
Moscati,I.(2013b)WereJevons,MengerandWalrasreallycardinalists?Onthenotionofmeasurementinutilitytheory,psychology,mathematicsandotherdisciplines,1870-1910.HistoryofPoliticalEconomy45:373-414.
Moscati,I.(2015)Austriandebatesonutilitymeasurement,fromMengertoHayek.InR.Leeson(ed.)Hayek:ACollaborativeBiography,PartIV.NewYork:PalgraveMacmillan,pp.137-179.
Mumford,S.(1998)Dispositions.Oxford:OxfordUniversityPress.
Nagatsu,M.andPoder,K.(unpublished)Economists’conceptofchoice:anexperimentalstudy.
Okasha,S.(2016)Ontheinterpretationofdecisiontheory.EconomicsandPhilosophy32:409-433.
Pareto,V.(1906)Manualed’economiapolitica.Milano:SocietàEditriceLibraria.
Putnam,H.(1975)Mind,Language,andReality.Cambridge:CambridgeUniversityPress.
Robbins,L.C.(1932)AnEssayontheNatureandSignificanceofEconomicScience.London:Macmillan.Rosenberg,A.(1993)Economics:MathematicalPoliticsorScienceofDiminishingReturns?Chicago:UniversityofChicagoPress.
Ross,D.(2005)EconomicTheoryandCognitiveScience.Cambridge,Mass.:MITPress.
Ross,D.(2011)Estrangedparentsandaschizophrenicchild:choiceineconomics,psychologyandneuroeconomics.JournalofEconomicMethodology18:217-231.
Ross,D.(2014)Economicversuspsychologicalmodelsofboundedrationality.JournalofEconomicMethodology21:41-27.
21
Samuelson,P.A.(1938)Anoteonthepuretheoryofconsumer’sbehaviour.Economica5:61–71.
Samuelson,P.A.(1948)Consumptiontheoryintermsofrevealedpreferences.Economica15:243–53.
Samuelson,P.A.(1950)Theproblemofintegrabilityinutilitytheory.Economica17:355–83.
Sen,A.K.(1977)Rationalfools:acritiqueofthebehavioralfoundationsofeconomictheory.Philosophy&PublicAffairs6:317-344.
Sen,A.K.(1982)Choice,WelfareandMeasurement.Oxford:Blackwell.
Starmer,C.(2000)Developmentsinnon-expectedutilitytheory:thehuntforadescriptivetheoryofchoiceunderrisk.JournalofEconomicLiterature38:332-382.
Sugden,R.(1991)Rationalchoice:asurveyofcontributionsfromeconomicsandphilosophy.EconomicJournal101:751-785.
Tiitinen,H.,May,P.,Reinikainen,K.&Näätänen,R.(1994)Attentivenoveltydetectioninhumansisgovernedbypre-attentivesensorymemory.Nature372:90-92.
Tversky,A.(1969)Intransitivityofpreferences.PsychologicalReview76:31–48.
Tversky,A.(1972)Eliminationbyaspects.Atheoryofchoice.PsychologicalReview79:281–299.
Tversky,A.&Kahneman,D.(1992)Advancesinprospecttheory:cumulativerepresentationofuncertainty.JournalofRiskandUncertainty5:297-323.
Vanderbeeken,R.&Weber,E.(2002).Dispositionalexplanationsofbehavior.BehaviorandPhilosophy30:43-59.
Varian,H.(2005)IntermediateMicroeconomics.London:Norton.
Vromen,J.(2010)Onthesurprisingfindingthatexpectedutilityisliterallycomputedinthebrain.JournalofEconomicMethodology17:17-36.
Wakker,P.(2010)ProspectTheory:ForRiskandAmbiguity.Cambridge:CambridgeUniversityPress.
Watson,J.B.&McDougall,W.(1929)TheBattleofBehaviorism.NewYork:Norton.
Woodward,J.(2008)MakingThingsHappen.NewYork:OxfordUniversityPress.
Zajonc,R.B.(1980)Feelingandthinking:preferencesneednoinferences.AmericanPsychologist35:151-175.