Emotions in Moral Psychology
Content
TICS-734; No of Pages 6
Opinion
The role of emotion in moral psychology Bryce Huebner1, 3, Susan Dwyer2 and Marc Hauser3, 4, 5 1
Center for Cognitive Studies, Tufts University, Medford, MA 02155, USA Department of Philosophy, UMBC, Baltimore, MD 21250, USA 3 Department of Psychology, Harvard University, Cambridge, MA 02138, USA 4 Department Departm ent of Biolo Biological gical Anthropolo Anthropology, gy, Harvard Universit University, y, Cambri Cambridge, dge, MA 02138, USA 5 Department Departm ent of Evoluti Evolutionary onary Biology, Harvard University, University, Cambridge, MA 02138, USA 2
Re Rece cent nt work work in the the co cogn gnit itiv ive e an and d ne neur urob obio iolo logi gica call sciences scie nces indicates indicates an important important relat relationsh ionship ip betwe between en emotion and moral judgment. Based on this evidence, several researchers have argued that emotions are the source sou rce of our int intuit uitive ive mor moral al judgme judgments nts.. How Howeve ever, r, despite the richness of the correlational data between emotion and morality, we argue that the current neurological, behavioral, developmental and evolutionary evidence den ce is ins insuffi ufficie cient nt to dem demons onstra trate te tha thatt emo emotio tion n is
[3,4,7,8,11–14] 14].. In line with this research, we adopt the vernacular understanding of ‘emotion’ and merely appeal to the paradigm examples of emotional states that have been studied by such researchers: happiness, disgust, fear, contempt and guilt. The term ‘moral’ is more contentious. Yet, although there is debate over the scope of the term, there is a broad consensus that physically harming others and violating consideration considerationss of fair fairness ness are centr central al to the 18],, althou although gh the there re are clear clearly ly other other moral domain domain [15–18]
necess nec essary ary for ma makin king g moral moral jud judgme gments nts.. We sug sugges gestt instead, that the source of moral judgments lies in our causal-int causa l-intentio entional nal psychology psychology;; emot emotion ion often follo follows ws from these judgments, serving a primary role in motivating morally relevant action.
aspects of morality as well [19] well [19]..
Introduction The role of emotion in our moral psychology has long been the focus of philosophical dispute [1,2] [1,2].. Critically, it has often been asked: does emoti emotion on influen influence ce moral judgment judgment or mer merely ely motiva motivate te mor morall ally y rel releva evant nt act action ion?? Recent Recently ly neurologic neuro logical al and behav behaviora iorall data have been marshaled in favor of the claim that emotion is necessary for moral judgment [3 judgment [3 –7] 7],, if not both necessary and sufficient [7,8] sufficient [7,8].. We are not convin convinced ced that the evi eviden dence ce warran warrants ts such such conclusions. Although there is ample support for the claim that emotion has some role in moral thought and moral motiva mot ivatio tion, n, cur curren rentt eviden evidence ce is insuffi insufficie cient nt to explai explain n when or how emotion has a role in our moral judgments. We arg argue ue tha thatt the existi existing ng dat data a can cannot not estab establis lish h ‘the ‘the synchronic synch ronic claim’ that emoti emotion on parti partially ally or wholl wholly y consti consti-tutes our moral capacities, nor can it establish ‘the diachron hroniic cla claim im’’ th that at em emot otio ion n is nec neces essa sary ry for the the [9,10].. Our goal is developmen devel opmentt of our moral capacities capacities [9,10] to clarify the relationship between current empirical evidence and existing theories concerning the source of our moral judgments, concluding with some brief suggestions for future research. Before we begin, we pause to note that ‘emotion’ and ‘moral’ are contentious terms. We do not intend to enter into disputes over the meaning of these terms, and this is not an oversight on our part. The targets of our criticism have not typically explicated these terms or felt that it was nece necess ssar ary y to de define fine th them em to ma make ke th the e cl clai aim m th that at,, fo forr exam exampl ple, e, emot emotio ion n me medi diat ates es our our mora morall ju judg dgme ment ntss Corresponding author: author: Huebner Huebner,, B. ([email protected]). [email protected]).
Emotion has a role in our moral psychology It has long been recognized that violating moral norms is emotionally taxing [20]. [20]. Moreover, although our emotions do not alw always ays preven preventt us from from con contem templa platin ting g mor morall ally y reprehensi repre hensible ble acti actions, ons, feeli feelings ngs of guilt and shame typically typically compel us to stop short of immoral action. Indeed, psychopaths, who lack the capacity for empathy and guilt, often fai faill to inhibi inhibitt the their ir vio violen lentt ten tenden dencie ciess [9,21,22] [9,21,22].. Such Such observatio obser vations ns provi provide de the back background ground against which the role of emotion in moral psychology must be understood. However, they are also consistent with a variety of conce cept ptua uall lly y and and em empi piri rica call lly y di dist stin inct ct clai claims ms ab abou outt the the relationsh relat ionship ip betwe between en emoti emotion on and mora morall psycho psychology logy (Box ( Box
1), with at least four specifically implicated thus far: (i) Percei Perceived ved moral moral violat violation ionss often often evo evoke ke contem contempt, pt, [23,24].. shame, anger or disgust [23,24] (ii (ii)) Emotio Emotion n often often lea leads ds to morali moralizat zation ion [7,13,25] [7,13,25],, for for example, when disgust is cultivated in the service of the politically insidious goal of removing people from the realm of moral concern. [3,4,12,21] demonstrate that (iii) Neuroscien Neuroscientific tific studies studies [3,4,12,21] emotional structures are recruited in making moral judgments. (iv) Mora Morally lly relevant relevant action is often emotionally emotionally motivated, ate d, appear appearing ing early early in ont ontoge ogeny ny [12,26–30] and phylogeny [31]. [31]. First, Fir st, althou although emo emotio tion nmean ‘ac ‘accom compan panies ies’ ’ some some of responses our mor moral al judgments, thisgh does not that emotional ‘co ‘const nstitu itute’ te’ suc such h mor moral al judgmen judgments. ts. Sec Second ond,, althou although gh emotio emo tion n som someti etimes mes lea leads ds to morali moralizat zation ion,, this this is con consis sisten tentt with the hypothesis that emotion merely draws our attention to the morally salient features of our environment,
1364-6613/$ – see front matter 2008 Elsevier Ltd. All rights reserved. doi: doi:10.1016/j.tics.2008.09.006 10.1016/j.tics.2008.09.006 Available Available online xxxxxx
1
TICS-734; No of Pages 6
Opinion
Box 1. Five models of our moral psychology It is commonly agreed that moral psychology relies on a variety of computation compu tational al mecha mechanisms nisms.. Howev However, er, there is little agreement concerning the causal structure or temporal organization of these mechanisms. We suggest that the following five models delimit the most promising avenues for empirical enquiry (NB these are merely initial models; a complete story of our moral cognition is likely to be far more compl complicat icated) ed) ((Fig Figure ure I). The ‘Pu ‘Pure re Kantia Kantian’ n’ [61] model indicates that moral judgment is a rational, deliberative endeavor. According to the Pure Kantian, emotion generates reactive attitudes [62] on the basis basis of rat ration ional al app apprai raisal sals, s, but moral moral jud judgme gment nt is primarily a cons primarily conscious cious process of delib deliberate erate reflection. reflection. The ‘Pure [7,8],, by contrast, maintains that reason is the ‘slave of the Humean’ [7,8] Humean’ passions pass ions’, ’, that moral psychology psychology is esse essential ntially ly emoti emotive ve and that delibera deli berative tive mechanisms mechanisms are recruited recruited only to provide provide post-hoc post-hoc rationalizations of moral judgments. ‘Hybrid’ [3,4,10] models have also been developed on the basis of recent neurophysiological data, indicatin indi cating g that both emotional emotional and deli deliberat berative ive mecha mechanisms nisms are recruited in making moral judgments. Some advocates of this view [10] maintain that moral judgment always requires both emotion and reason; others [3,4,12] maintain that emotion has a important role only only in eva evalua luating ting persona personal-mo l-moral ral dil dilemma emmas s that require require indicate the physical phys ical inte interven rvention. tion. ‘Pure Rawl Rawlsian sians’ s’ [15,16,63] indicate existence of a distinctively moral faculty, operating independently of deliberative and emotional mechanisms. For the Pure Rawlsian, emotional mechanisms are recruited antecedent to moral judgment to translate moral judgments into morally relevant actions. Finally, the ‘Hybrid Rawlsian’ Rawlsian’ could indicate indicate that emotion is recruited in evaluating eval uating high-confli high-conflict ct personal personal dilem dilemmas mas and in resolving resolving ambiguous outputs from the moral faculty. However, the plausibility of thi this s mod model el has yet to be defend defended ed or subjec subjected ted to empiri empirical cal scrutiny.
Figure I. Five models of the moral mind. mind.
capturing attention and triggering distinctively moral cog [32–34] 34].. Third, existent neurological data are insufnition [32 nition ficient fici ently ly pre precis cise e to demons demonstra trate te the causal causal or tem tempor poral al rol role e of emotion in moral psychology. Finally, we take no issue wi with th th the e fo four urth th cl clai aim. m. In fa fact ct,, we not not on only ly ap appl plau aud d attempts to understand the role of emotion in motivating moral action but also suggest that this is probably its most important impor tant role [15] [15].. It is, howeve however, r, more more freque frequentl ntly y hyp hypoth oth-esized that emotion is both necessary and sufficient for the [8],, th that at maki making ng a mo mora rall possession posse ssion of mora morall conc concepts epts [8] judgment is nothing more, nor less, than being in a particular emo emotio tional nal sta state te and that that emotio emotional nal str struct ucture uress are recruited in making some [3,4,6,7,12,13] or even all [10] 2
Trends in Cognitive Sciences
Vol.xxx No.x Vol.xxx
moral judgments. Furthermore, building on a long philosophical sophic al tradi tradition tion [1,2] [1,2],, som some e cognit cognitive ive sci scient entist istss hav have e recently suggested that sympathetic concern is both ontogenetically and phylogenetically necessary for the development opm ent of kin kin-al -altru truism ism,, rec recipr iproca ocall altrui altruism sm and eve even n 39].. Unf Unfort ortuna unatel tely, y, we do genuine genuin e altru altruism ism [9,21,31,35–39] not believe the existent experimental paradigms are sufficient to test these hypotheses. To clarify what is theoretically at stake in this debate, consider consi der a parallel parallel set of concerns about ‘embo ‘embodied died cognition’. tion’. The strong strong emb embodi odied ed cog cognit nition ion the thesis sis holds holds tha thatt explicit motor representations partially constitute many concepts [40–42] 42].. The conce concept pt HAM HAMMER MER is tho though ughtt to consist of (i) features that typify hammers (e.g. a grasping shaft, a hard end) and (ii) the motor routines involved in goal-direc goaldirected ted grasping grasping and swing swinging. ing. Supporters Supporters of this hypothesis hypoth esis often appea appeall to neuroimagi neuroimaging ng data demonstrat demonstrat-ing that the word ‘hammer’ activates circuits classically associated with object categorization and circuits in the Howeve ever, r, activa activatio tion n in the primary prima ry motor cortex [43]. [43]. How motor cortex does not license the conclusion that the concept HAMMER includes motor routines. Given the poor temporal resolution of neuroimaging, it is just as likely that the concept HAMMER activates circuits dedicated to object categorization, which, in turn, activate circuits in the motor cortex. Distinguishing these hypotheses enables cognitive cognit ive scien scientists tists to target motor routines routines to see whether whether they are ‘part of our concepts’ or instead ‘stand in important causal relationships to them’. Analogously, we argue that cognitive scientists must target emotional mechanismss to see whethe ism whetherr the they y ‘const ‘constitu itute’ te’ mor moral al concep concepts ts or merely mer ely ‘st ‘stand and in an imp import ortant ant cau causal sal rel relati ations onship hip’’ to them. them. Behavioral data The central central behavi behaviora orall data data in fav favor or of the cla claim im tha thatt emotion is the source of moral judgment is the apparent modulation of moral judgment by disgust [7,44]. [7,44]. Participants responding to moral dilemmas at a dirty desk or when smelling a noxious odor, make more severe moral [6].. Highl Highly y susce susceptibl ptible e participartici judgments than controls [6] pants, hypnotically induced to experience ‘a brief pang of disgust’ when confronted with a neutral word, see moral transg tra nsgres ressio sions ns as mor more e mor morall ally y wro wrong ng in vig vignet nettes tes contai containning the hypnotically targeted word [44] word [44].. And, participants who watch a humorous clip from ‘Saturday Night Live’, as oppose opp osed d to a neutra neutrall con contro troll cli clip, p, report report fee feelin ling g mor more e positive mood and offer more utilitarian responses to the footbridge dilemma but not to the bystander dilemma [45] dilemma [45].. Yet, although these are fascinating effects, they are insufficient to demonstrate that emotion is necessary for moral judgment. As we see it, there are four difficulties with appeal app ealing ing to suc such h beh behavi aviora orall data data in est establ ablish ishing ing tha thatt emotions partially constitute moral judgments. Firs First, t, thes these e da data ta fa fail il to isol isolat ate e the the pr prec ecis ise e po poin intt at wh whic ich h emotion has a role in our moral psychology (see Box 2). 2). Studies [6,45] in which emotional stimuli are presented before the scenario is read, could modulate the inputs to the emotional system, enabling emotion to influence the interpretation of the scenario or the question. Or, emotion co coul uld d ac actt as a ga gain in on what what has has al alre read ady y be been en co conc ncei eive ved d as a moral infraction (thereby, increasing the severity of the
TICS-734; No of Pages 6
Opinion
Box 2. Where is the effect of emotion? Although various Although various expe experiment riments s demon demonstrate strate that strong emotional stimuli affect the response to moral scenarios, the exact point at which the effect occurs has not been critically critically exami examined. ned. On the basis of currently available data, the effect of emotion could occur (i) in the interpretation of the scenario, (ii) in the interpretation of the question, quest ion, (iii) in the production production of the moral judgment judgment or (iv) in reporting the judgment as a measurable response. The precise point at whic which h em emot otio ion n has has a role role is th the e mo most st impo import rtan antt issu issue e fo forr establ establish ishing ing the truth truth or fal falsit sity y of the emo emotio tional nal con consti stitue tuency ncy hypothesis (Figure (Figure I). I).
Trends in Cognitive Sciences
Vol.xxx No.x
resources. Existent data fail to address the plausible hypothesis that the apparent modulation of moral judgments by emo emotio tion n is an art artifa ifact ct of the redepl redeploym oyment ent of att attent ention ional al resources. With this in mind, we also note that although high socio-economic status (SES) participants treat harmless wrongdoings (e.g. cleaning a toilet with one’s national flag) as disgu disgusting, sting, whereas low-SES parti participa cipants nts trea treatt such suc h action actionss as univer universal sally ly wrong wrong and deserv deserving ing of pun punish ish-[13],, such data merely show that the class of uniment [13] versally wrong actions that are deserving of punishment has a permeable boundary. This is unsurprising. After all, differences in socialization, mediated by strong emotional cues, can ground ground conventiona conventionall judgme judgments nts about taboos that exhibit precisely these features. In summary, current behavioral data fail to license the claim that emotional emotional processes processes are the source of moral computation leading to judgment.
Figure I. Possible roles of emotion in moral cognition.
perceived wrong) antecedent to the operation of distinctively tivel y moral circuits circuits.. Moreover, Moreover, appeals to hypnoticall hypnotically y induced disgust [44] disgust [44] cannot cannot alleviate this worry. Because, here her e too too,, emo emotio tion n could could modify modify the inp inputs uts into into dis distin tincti ctivel vely y morall circu mora circuits its rather than modulating modulating the operation of
Neuroimaging data Neurobiological data have also been offered in support of the claim that emotional circuits are integral to evaluating morally moral ly salient salient stimu stimuli. li. For example, judgments judgments about morally salient claims (e.g. ‘The elderly are useless’) show increased activity in the frontal polar cortex (FPC) and medial frontal gyrus, when compared to judgments about
these moral circu circuits its themselves themselves.. Thus, although asking subjects to evaluate a moral question triggers the process of moral evaluation, the negative emotional state yields a more severe moral judgment because of an increased focus on th the e ‘ant ‘antec eced eden entl tly’ y’ mo mora rall lly y sa sali lien entt fe feat atur ures es of th the e scenario. Second, and on a related note, these data fail to demonstrate stra te that emoti emotion on tran transform sformss conven conventiona tionall judgme judgments nts into moral judgments. judgments. To demon demonstrat strate e that judgments can be shifted from the non-moral into the moral realm by inducing or amplifying an emotional response (a thesis suggested by Nichols [46] Nichols [46])) experiments analyzing the perceived moral relevance of a transgression would have to be used. Thus, all existent data fail to distinguish between three thr ee pla plausi usible ble hyp hypoth othese eses: s: (i) emotio emotion n acts acts as a gain gain antecedent to moral judgment, (ii) negative emotion alerts us to the moral salience of a situation and (iii) emotion triggers a genuine conceptual shift from merely conventional to moral. Third, there is little consistency in the scales used to meas me asur ure e th the e role role of em emot otio ion n in mora morall ps psyc ycho holo logy gy.. Some Some us use e a 7-point scale ranging from ‘perfectly OK’ to ‘extremely immoral’ [6] immoral’ [6],, others use a scale asking ‘how morally wrong’ an action is [44] is [44].. Yet others pose a dichotomous question about abo ut ‘ac ‘accep ceptab tabili ility’ ty’ or ‘permi ‘permissi ssibil bility ity’’ and som some e use a combin com binati ation on of the these se and oth other er scales scales within within the same same study.. This makes comparison study comparison across, across, and even within, [6] impossible. studies [6] studies Fourth, Fourt h, altho although ugh emoti emotion on yields ‘pra ‘practica ctical’ l’ judgme judgments, nts, it is unclear that this warrants treating emotion as constitutive tut ive of ‘moral ‘moral’’ judg judgmen ments. ts. Perhap Perhapss mor moral al cog cognit nition ion can be interf interfere ered d with with by int introd roduci ucing ng distra distracti cting ng emotio emotional nal stimuli [6] stimuli [6].. However, because disgust functions practically to help us avoid toxic toxic,, infec infectious tious or contaminat contaminating ing substances, stanc es, it could generate interrup interruptive tive judgments that coul could d co comp mpet ete e wi with th mora morall co cogn gnit itio ion n fo forr atte attent ntio iona nall
non-moral claims (e.g. ‘Telephones never ring’) [47] ring’) [47].. Moreover, morally salient stimu stimuli li evoke increase increased d functional functional connectivity between the left FPC, orbital frontal (OFC), anterior temporal, and anterior cingulate cortices (ACC) and limbic structures such as the thalamus, midbrain and Finally, ly, perso personal-m nal-moral oral dilem dilemmas mas basal foreb forebrain rain [48]. [48]. Final (e.g. footbridge dilemma) selectively recruit emotional circuits when compared to impersonal-moral dilemmas (e.g. bystander dilemma). On the the ba basi siss of thes these e and and si simi mila larr im imag agin ing g re resu sult ltss [3,4,12],, Greene [11] has proposed that moral judgment [3,4,12] requires requi res a prepotent prepotent emotional response (subserved (subserved by circuits in the medial frontal gyrus, the posterior cingulate gyrus gyr us and the angula angularr gyr gyrus) us) tha thatt dri drives ves moral moral disapp disapprov roval al and reflective utilitarian reasoning (implemented in the dorsolater dorso lateral al pref prefronta rontall corte cortex x [DLPFC]) [DLPFC]).. Althou Although gh these systemss typically system typically produce produce conve convergent rgent outpu outputs, ts, outpu outputs ts diverge diver ge in personal-m personal-moral oral dilemmas, dilemmas, generating generating confli conflict ct (evidenced by increased activity in the ACC) that must be resolv res olved ed by higher higher-co -cogni gnitiv tive e contro controll cir circui cuits ts in the ant anteri erior or DLPFC. DLPF C. Conve Converging rging data from neuroecono neuroeconomics mics indic indicate ate that unfair offers in an ‘ultimatum game’ elicit increased cortical activity in the anterior insula in addition to the DLPFC [49,50] DLPFC [49,50].. More broadly, emotional circuits in the insularr cortex insula cortex are diffe differentia rentially lly acti activated vated in the perce perception ption of inequity [50], [50], and it is hypothesized that activity in the insula ins ularr cortex cortex is indica indicativ tive e of the sen sensit sitivi ivity ty to nor normm violations implicated in deontological judgment. As interesting as these neurobiological data are, they only onl y show show ‘that ‘that som some e per percei ceived ved deonto deontolog logica icall violat violation ionss are associated with strong emotional responses’, a perspective that ‘few would doubt or deny’ [51]. [51]. More generally, the activity of emotional circuits provides only correlational data, showing that emotions emotions are associat associated ed with moral judgments. Such data (on their own) can never be used to infer infer cau causal sality ity,, and bec becaus ause e of the poor poor tem tempor poral al resolu resolutio tion n 3
TICS-734; No of Pages 6
Opinion
of neuroimaging, neuroimaging, cannot be used to assess when emotions have a rol have role e or whe whethe therr they they are const constitu itutiv tive e of mor moral al concepts. In summary, the mere activity of neural circuits classicall class ically y associated associated with emoti emotion on in proce processing ssing moral scenarios fails to distinguish between the claim that (i) emot em otio ions ns are are in inte tegr gral al to mo mora rall co comp mput utat atio ion n an and d (i (ii) i) emotions result from these computations. We, therefore, conclude that imaging data on their own are insufficient to determine whether emotions are activated before moral comput com putati ations ons,, dur during ing moral moral com comput putati ations ons or antece anteceden dentt to moral mor al com comput putati ations ons.. Future Future resear research ch com combin bining ing fun funcctional tional mag magnet netic ic resona resonance nce imagin imaging g (fM (fMRI) RI) wit with h eve event nt related potentials (ERPs) could help resolve this timing issue.
Trends in Cognitive Sciences
Vol.xxx No.x Vol.xxx
the flattening of negative emotion would yield more permissible moral judgments because of a failure to focus on the ‘an ‘antec tecede edentl ntly’ y’ mor morall ally y sal salien ientt featur features es of the sce scenar narios ios..
Neuropsychological data A much stronger test of the hypothesis that emotion is the source sou rce of mor moral al judg judgmen mentt is ind indica icated ted by stu studie diess of patien patients ts with adult-onset, bilateral damage to ventromedial prefrontal cortex (VMPC). VMPC patients exhibit: (i) a flattening ten ing of the their ir social social emo emotio tions, ns, indica indicated ted by beh behavi aviora orall and ph phys ysio iolo logi gica call me meas asur ures es;; (i (ii) i) an in inab abil ilit ity y to re rede depl ploy oy emotional representations representations previously associated with punishment and reward; (iii) an inability to anticipate future outcom out comes, es, pun punish ishmen ments ts and rewards rewards and (iv (iv)) a lac lack k of
The diachronic necessity of emotion? Correlational data from the mind sciences could be further reinforced by data demonstrating that emotion is necessary for the acquisition of moral concepts and the develexampl mple, e, opment of mora morall judgme judgment. nt. Hoff Hoffman man [56], [56], for exa appeals to the affectively laden tools used by parents to convey social rules and correct behavior in arguing that emotions are developmentally necessary for moral judgment. We disagree. Children are not given sufficient moral correction to account for the intricate structure of their moral psychology [57], [57], and although there is virtually no work on the morally relevant input the child receives, our sense sen se is that that mos mostt rule-b rule-base ased d correc correctio tion n is dir direct ected ed tow toward ardss conventional transgressions (‘Take your finger out of your nose!’) as opposed to moral ones (‘Don’t kill your brother.’). As in studies of child language acquisition, it is clear that we need need se seri riou ouss st stud udie iess of the the ac actu tual al inpu inputt the the ch chil ild d receives in the moral domain. An alternative ontogenetic hypothesis relies on the moral deficiencies deficiencies of psychopaths psychopaths.. Psych Psychopaths opaths engage in immoral acts, show a callous lack of concern for others
inhibitory control inhibitory control [10,52–54] 54].. Dama Damasio sio [10] argues argues that emotion is usually integral to cognition, and on this basis, hypothesizes that moral judgment is likely to rely on the emotional processes implemented in VMPC. Consistent Consis tent with this hypot hypothesis, hesis, frontotemp frontotemporal oral dementia (FTD), resulting from the deterioration of prefrontal and anterior temporal cortex, generates blunted emotion, disregard for others and a willingness to engage in moral transg tra nsgres ressio sions. ns. Mor Moreov eover, er, FTD pat patien ients ts show show a pro pro-nounced tendency to adopt the utilitarian alternative in personal moral dilemmas such as the footbridge case [55] case [55].. With these results results in mi mind, nd, Koenigs Koenigs,, Young and ccollea olleagues gues [53] examined examined the judgments of the previously mentioned [53] VMPC patients for moral and non-moral dilemmas. Each dilemm dilemma a was class classifie ified d as inv involv olving ing a per person sonal al or an impersonal violation [4] violation [4];; personal violations were further subdivided into low-conflict (characterized by short reaction tion tim times es and low varia variance nce in subjec subjectt jud judgme gments nts)) and high-conflict cases. VMPC patients were indistinguishable from controls, except in high-conflict dilemmas in which they were more likely to endorse the utilitarian outcome. Over Overal all, l, th thes ese e da data ta sh show ow th that at ev even en thou though gh VM VMPC PC patients patie nts experi experience ence a flattened flattened socio socio-emot -emotional ional profil profile, e, they nonetheless judge most moral dilemmas (i.e. impersonal moral dilemmas, low-conflict personal moral dilemmas and personal dilemmas in which ‘harm to another’ is pitted against a ‘benefit to self’) as do healthy controls. More Mo reov over er,, gi give ven n th the e limi limite ted d ra rang nge e of case casess on whic which h VMPC VMPC patients deviate, it is plausible that they fail to treat the morall mor ally y sal salien ientt featur features es of hig high-c h-confl onflict ict dil dilemm emmas as as morally salient. If our interpretation is correct, then moral cognit cog nition ion wou would ld yie yield ld devian deviantt out output putss as a res result ult of devian deviantt inputs, rather than as a result of a deficit in moral processing per se; sing parallels ls the interpret interpretation ation we offered offered se; this paralle earlie ear lierr for the hypnot hypnotism ism stu study dy [44] [44].. In this this case case,, th the e proces pro cesss of moral moral eva evalua luatio tion n wou would ld remain remain intact intact.. Howeve However, r,
and lac lack k the cap capaci acitie tiess for guilt, rem remors orse e and empathy empathy [9,21,32,58]. [9,21,32,58]. Bla Blair ir has cla claime imed d that that in nor normal mally ly develo developin ping g children, emotional circuits facilitate negative reinforcement for actions that generate distress cues. Psychopaths lack these emotional circuits, and Blair [9,21] Blair [9,21] argues argues from this fact to the claim that emotion is the developmental source of our moral concepts and that psychopathy is a developmental consequence of an early emotional deficit. Although psychopaths fail to distinguish moral from conventional transgressions, treating conventional violations as less permissible, more serious and as authorityindependent (i.e. as moral transgressions) [21] transgressions) [21],, such data do not not sp spea eak k to eith either er the so sour urce ce or the the co cont nten entt of a psychopath’s moral cognition. This pattern of response is equally well accounted for by a cold, calculated rationality, desi design gned ed to get get ‘o ‘off ff the the hook hook’’ and and to sa say y what what ot othe hers rs wa want nt to [9] himself himself has shown by using hear. Moreover, as Blair [9] age-matched psychopathic and non-psychopathic juvenile delinquents, even psychopathic juveniles draw the moralconventional distinction (although it is less pronounced in psychopathic than non- psychopathic juveniles), and make just as many references to welfare considerations as do non-psychopathic controls (although psychopathic juveniles were less likely to ascribe moral emotions to others). These data indicate a developmental trajectory for psychopathy, but contrary to what one would predict if emotion is developmentally necessary for acquiring moral concepts. Psycho Psy chopat pathic hic juve juvenil niles es app appare arentl ntly y lose lose the cap capaci acity ty to distinguish distin guish moral from conventio conventional nal viola violations tions over the course of development. Therefore, perhaps the deficiencies in the moral psychology of the psychopath are a developmental consequence of antisocial behavior, instead of the [58] argues, argues, a life filled with other way around. As Raine [58] antisocial behaviors could modify the moral psychology of an indiv individual idual,, allo allowing wing for the justificati justification on of immor immoral al behaviors and reducing cognitive dissonance. However, if
4
TICS-734; No of Pages 6
Opinion
this is true, the moral cognition of psych psychopath opathss is deviant as a result of deviant inputs rather than as a result of a deficiency in moral processing. A final piece of evidence for the developmental developme ntal necessity of emotion is indicated by the moral psychology of earlyonset ons et VM VMPC PC patien patients. ts. EarlyEarly-ons onset et les lesion ionss to VMP VMPC C [54,59] lea lead d to a phe phenot notypethat ypethat is unresp unrespons onsiveto iveto punish punish-ment, men t, lac lacks ks inh inhibi ibitor tory y contr control, ol, genera generates tes behavi behaviora orall deficiencies in moral and prudential domains and shows emotional deficits in guilt, remorse and empathy. Such patients pati ents also fail to acquire acquire moral conc concepts, epts, justifying justifying their behavior by appeal to the egocentric desire to avoid punishment [54,59] punishment [54,59].. However, such data only tell us that the acquisition acquisition of mora morall concepts concepts is downstrea downstream m from some soci social-e al-emoti motional onal mechanism mechanisms. s. To establis establish h that emotional processes are constitutive of moral cognition, wewouldneeda muc much h cl clear earerpictu erpicture re ofthe preci precise se defici deficits ts pr prese esent nt in ea early rly-o -onse nsett VMPC VMPC pat patien ients,testi ts,testing ng them them on the same battery of dilemmas used in the adult study. Given the rarity rarity of thi thiss disord disorder er,, it is unclea unclearr whethe whetherr the absence of moral cognition is a deficit in the acquisition of social rules produced by a lack of positive feedback, or even a result of deviant behavior that inhibits the maturation of moral cognition.
Box 3. Directions for future research
Future experiments should adopt a single scale for analyzing the effectt of emoti effec emotion on on moral judgme judgment, nt, or at least use different scales with one set of dilemmas to assess whether they generate dif differ ferent ent patter patterns ns of jud judgme gment. nt. For ex exampl ample, e, to exp explor lore e the potential poten tial ‘gain’ function function of emoti emotion on in shift shifting ing the seve severity rity of the moral judgment, we propose a 7-point 7-point scale scale ranging from obligatory (1) – permissible (4) – forbidden (7) [64] (7) [64].. To explore the potential driving force of emotion in labeling a situation as moral, we propose using a moralness scale that runs from conventional (1) to moral (7), each end point anchored by a concrete example. Experimental paradigms should be established to time-lock fMRI data data with with ERP data, data, and exp experi erimen mental tal paradi paradigms gms should should be developed to distinguish the effect of emotion on moral judgment from the effect of attentional redeployment and cognitive load on moral judgment.
Trends in Cognitive Sciences
Vol.xxx No.x
Conclusion We conclude that current evidence is insufficient to support the hypothesis that emotional processes mediate our intuitive moral judgments, or that our moral concepts are emotio emo tional nally ly con consti stitut tuted. ed. We sug sugges gestt instea instead d (for (for mor more e that our complete compl ete devel development opment,, see Refs [15,16,51,60]) [15,16,51,60]) that moral mor al judgme judgments nts are mediat mediated ed by a fast, fast, unc uncons onscio cious us process that operates over causal-intentional representation ti ons. s. The The mo most st im impo port rtan antt ro role le that that emot emotio ions ns migh mightt ha have ve is in motivating action. We offer these criticisms to sharpen future discussion and empirical enquiry. In addition, we hope hop e tha thatt future future stu studie diess wil willl be car carefu efull to dis disamb ambigu iguate ate the various ways in which emotional mechanisms might have a role in our moral cognition (Box (Box 3). 3). Do the social emotions push us into the moral domain, alerting us to the presence of morally salient properties? Do they modify the severity of our moral judgments? Do emotional structures inhibit our tendency to entertain thoughts that are contrary to moral mor al reason reasoning ing?? Do they they merely merely inh inhibi ibitt pra practi ctical cally ly irr irrati ationa onall act action ion?? We are opt optimi imisti sticc tha thatt more more carefu carefull research in the cognitive sciences will lead to more satisfying answers to these empirical and conceptual questions, and that the tools of cognitive neuroscience and neuropsychology will be at the forefront of future discussion. References 1 Hume, D. (1978 (1978)) A Treati Treatise se of Hum Human an Nat Nature ure,, Oxford University Press 2 Smith Smith,, A. (175 (1759) 9) The Theor Theory y of Mor Moral al Sentim Sentiment ents, s, Cambridge University Press 3 Greene, Greene, J.D. J.D. et al. (2001) An fMR fMRII investig investigati ation on of emo emotio tional nal al. (2001) engagement in moral judgment. Science judgment. Science 293, 293, 2105–2108 4 Greene, J.D. et al. (2004) The neural bases of cognitive conflict and controll in mora contro morall judgme judgment. nt. Neuron 44, 389–400 Neuron 44, 5 Haidt, Haidt, J. et al. (1997) Body, psyche, and culture culture:: the relat relationshi ionship p al. (1997) between disgust and morality. Psychol. Dev. Soc. J. 9, 107–131 6 Schnal Schnall, l, S. et S. et al. (200 (2008) 8) Disgustas embodied embodied moraljudgment. moraljudgment. Pers. Pers. Soc. Psychol. Bull. Bull. 34, 34, 1096–1109 7 Haidt, Haidt, J. (200 (2001) 1) The emotional dog and its rational tail. tail. Psychol. Psychol. Rev. 108, 814–834 8 Prinz, J. (2006) The emotional emotional basis of moral judgment. judgment. Philos. Philos. Exp. 9, Exp. 9, 29–43 9 Blair, Blair, R. (1997) (1997) Moral reason reasoning ing and the child child wit with h psyc psychop hopath athic ic tendencies. Pers. tendencies. Pers. Individ. Dif. 22, 731–739
Further experiments on patient populations (especially adult- and early-onset VMPC patients) should be carried out to examine the precise prec ise deficits that they exhibi exhibitt on moral tasks, including including the interaction between social emotions. Similarly, we should pursue experiments targeting the performance of psychopaths on moral dilemmas, instead of moral-conventional cases, to eliminate the use of pre previo viousl usly y rehear rehearsed sed ans answer wers s and a rat ration ional al des desire ire to cohere. A mor more e comple complete te accoun accountt of the com compon ponent ent pro proces cesses ses tha thatt underlie unde rlie moral judgment must be estab establish lished, ed, distingui distinguishing shing domain-specific from domain-general processes, and documenting how non-morally specific components interface to create new and potentially potentially morali morality-spec ty-specific ific representatio representational nal resourc resources es [15,16]. With [15,16]. With thi this s descri descripti ptive ve accou account nt in pla place, ce, it wil willl the then n be possible to map out the ontogenetic and phylogenetic patterns of change, and use neurobiological techniques such as transcranial magnetic stimulation to disrupt these components. Finally, we propose that emerging work on the performance of
10 Damasio, Damasio, A. (1994) (1994) Descartes Error, Harper Error, Harper Perennial 11 Greene, Greene, J.D. J.D. (2007) (2007) Why are VMPFC pati patients ents more utili utilitari tarian?: an?: a dual-process theory of moral judgment explains. Trends explains. Trends Cogn. Sci. 11, Sci. 11, 322–323 12 Greene Greene,, J.D. J.D. and Haidt Haidt,, J. (2002) (2002) How (and (and whe where) re) doe doess mor moral al judgment work? work? Trends Trends Cogn. Sci. 6, Sci. 6, 517–523 13 Haidt, J. (1993) Affe Affect, ct, culture, and morality, or is it wrong to eat your dog? J. Pers. Soc. Psychol. 65, dog? Psychol. 65, 613–628 14 Shweder, Shweder, R. and Haid Haidt, t, J. (2000) (2000) The cultural psychol psychology ogy of the emotions: emot ions: ancient and new. In Handbook Handbook of emotions emotions (Lewis, M. and Haviland, J., eds), pp. 397 –414, Guilford Press 15 Hauser, Hauser, M. (2006) Moral (2006) Moral Minds, Harper Minds, Harper Collins 16 Mikhail, J. (2007) Universal moral grammar. Trends Cogn. Sci. 11, 143–152 17 Joyce, Joyce, R. (2006) The (2006) The Evolution of Morality, MIT Press 18 de Waal, F. (1997) Good (1997) Good Natured, Natured, Harvard University Press 19 Haidt, Haidt, J. (200 (2007) 7) The new synthes synthesis is in moral moral psychology. psychology. Science 316, Science 316, 998–1002 20 Milgram, S. and Sabini, J. ((1978) 1978) On maintaining urban norms: a field
participants in virtual realityfor environments provides anemotion important and intriguing test ground examining the role of in moral motivation motivation as opposed opposed to moral judgmen judgmentt [65,66]. [65,66]. Such studies are only in their early days, however, they do indicate a promising promi sing and untapped untapped resea research rch environme environment, nt, espe especial cially ly one aimed at distinguishing judgment from action.
experiment in the subway. In experiment In Advances Advances in Environmental Psychology (Baum, A. et A. et al., al., eds), pp. 31–40, Erlbaum 21 Blair, R.J. (1995) A cogniti cognitive ve devel developme opmental ntal approach to mora morality: lity: investigating the psychopath. Cognition psychopath. Cognition 57, 57, 1–29 22 Nic Nichol hols, s, S. (2002 (2002)) Howpsychopa Howpsychopaths ths threat threaten en moral moral raitona raitonalism lism,, or is it irrational to be amoral. Monist amoral. Monist 85, 85, 285–303
5
TICS-734; No of Pages 6
Opinion
Trends in Cognitive Sciences
Vol.xxx No.x Vol.xxx
23 Rozin Rozin,, P. P. et et al. (1999) al. (1999) The moral-emotion triad hypothesis: a mapping between three moral emotions (contempt, anger, disgust) and three moralethics (community, (community, autonomy, autonomy, divinit divinity). y). J. J. Pers. Soc. Psychol. 76, 574–586 24 Rozin, P. (1997) Moralization and be becoming coming a vegetarian. vegetarian. Psychol. Psychol. Sci. 8, 67–73 25 Rozin Rozin,, P. P. et In Handbook of Emotions, et al. (1993) al. (1993) Disgust, In Handbook Emotions, pp. 575–594, Guilford Press 26 Gall Gallese,V. ese,V. (2003) (2003) The root rootss of emp empathy:the athy:the shared shared mani manifoldhypothes foldhypothesis is and the neural basis of intersubjectivity. intersubjectivity. Psychopathology 36, Psychopathology 36, 171–180 27 Hoffm Hoffman, an, M. (198 (1982) 2) Developmen Developmentt of proso prosocial cial motivation:empathy motivation:empathy and guilt. In Development of Prosocial Behavior (Eisenberg, Behavior (Eisenberg, N., ed.), pp.
46 Nichols, Nichols, S. (2004) Sentimental (2004) Sentimental Rules: On the Natural Foundation of Moral Judgment, Judgment, Oxford University Press 47 Moll, J. et al. (2002) (2002) The neural corre correlates lates of moral sensitivi sensitivity: ty: a functional functi onal magnetic resona resonance nce imaging imaging investig investigation ation of basic and moral emotions. J. emotions. J. Neurosci. 22, Neurosci. 22, 2730–2736 48 Moll, J. J. et al. (2003) al. (2003) Morals and the human brain: a working model. 14, 299–305 Neuroreport 14, Neuroreport 49 Sanfey, Sanfey, A.G. A.G. et (2003) 3) The neura neurall basis of economic economic decis decision-ma ion-making king et al. (200 in the ultimatum game. Science game. Science 300, 300, 1755–1758 50 Hsu, Hsu, M. et M. et al. (2008) al. (2008) The right and the good: distributive justice and neurall encod neura encoding ing of equity and efficiency. efficiency. Science 320, 1092–1095 Science 320, 51 Mikhai Mikhail, l, J. (2008)Moralcognit (2008)Moralcognitionand ionand Comput Computati ationa onall theory theory.. In Moral In Moral
281–313, Academic Press 28 Hoffman, M. (2000) Empathy (2000) Empathy and Moral Development: Implications for Caring and Justice, Justice, Cambridge University Press 29 Meltzoff, A.N. and Moore, M.K. (197 (1977) 7) Imitation of facial and manual gesturess by human neonates. Science gesture neonates. Science 198, 198, 75–78 30 Zahn-W Zahn-Waxler axler,, C. C. et et al. (1979) Child rearing and children’s prosocial initiations toward victims of distress. Child Dev. 50, Dev. 50, 319–330 31 Pres Presto ton, n, S. and and DeWa DeWaal al,, F. (2 (200 001) 1) Em Empa pathy thy:: its ultim ultimat ate e and and proximate bases. Behav. bases. Behav. Brain Sci. 25, 1–20 32 Bats Batson, on, C. et al. (1995) Immorality from empathy-induced altruism: whencompassio whencompa ssion n andjusticeconfli andjusticeconflict. ct. J. J. Pers.Soc. Psychol.68,1042 Psychol. 68,1042–1054 33 Pizarro, D. (2000) Nothing more than feelings?: the role of emotions emotions in moral judgment. judgment. J. Theory Soc. Behav. 30, Behav. 30, 355–375 34 Pizar Pizarro, ro, D. D. et al. (2006) The creativity of everyday moral reasoning: empathy, empat hy, disgust and moral persuasion. persuasion. In Creativity and Reason in (Kaufman,, J. and Baer, Baer, J., eds), pp. 81–98, Cognitive Development (Kaufman Cambridge University Press 35 Blair, R.J. and Cipolotti, L. (2000 (2000)) Impaired social response reversal: a case of ‘acquired sociopathy’. Brain sociopathy’. Brain 123, 123, 1122–1141 36 Decet Decety, y, J. J. et et al. (2002) al. (2002) A PET exploration of the neural mechanisms involved in reciprocal imitation. Neuroimage imitation. Neuroimage 15, 15, 265–272 37 Jackso Jackson, n, P.L. et al. (2006) Neural circuits involved in imitation and perspective-taking. Neuroimage perspective-taking. 31, 429–439 Neuroimage 31, 38 Flack, J. and deWaal, deWaal, F. (200 (2000) 0) ‘Any animal animal whatever’. whatever’. Darwini Darwinian an building blocks of morality in monkeys and apes. J. apes. J. Conscious. Stud. 7, 1–29 39 Sob Sober,E. er,E. andWilso andWilson, n, D D.. (19 (1998 98)) Unto Others, Others, Harvard University Press 40 Barsalou, Barsalou, L.W. et al (2003) Grounding Grounding conce conceptual ptual knowledge knowledge in al.. (2003) modality-specific systems. Trends systems. Trends Cogn. Sci. 7, Sci. 7, 84–91 41 Mahon, B.Z. and Caramazza, A. (2008) A critical look at the embodied cognition cogniti on hypothesis hypothesis and a new proposal for ground grounding ing conce conceptual ptual content. J. content. J. Physiol. (Paris) 102, 59–70 42 Lakoff, Lakoff, G. and Johnso Johnson, n, M. (19 (1999 99)) Philoso Philosophy phy in the Flesh: Flesh: The embodied Mind and its Challenge to Western Thought, Basic Thought, Basic Books 43 Pulvermu Pulvermu¨ller, F. (2005) (2005) Brain mechanisms linking language and action. Nat. Rev. Neurosci. 6, action. Neurosci. 6, 576–582 44 Wheatley Wheatley,, T. and Haidt, Haidt, J. (20 (2005) 05) Hypnotic Hypnotic disgus disgustt mak makes es moral moral
Psychology: The Neuroscience of Morality: Emotion, Disease, and PRESS S Development (Sinnott-Armstrong, W., ed.), pp. 81–92, MIT PRES Development 52 Damasio, Damasio, A.R. A.R. et et al. (1990) al. (1990) Face agnosia and the neural substrates of memory. Annu. memory. Annu. Rev. Neurosci. 13, Neurosci. 13, 89–109 53 Koenigs, Koenigs, M. et al. (2007) Damage to the prefrontal cortex increases utilitarian moral judgments. Nature judgments. Nature 446, 446, 908–911 54 Anderson, Anderson, S.W. S.W. et et al. (2006) Impairments of emotion and real-world comple com plex x behav behavior ior follow following ing childho childhoodod- or adu adultlt-ons onset et damag damage e to ventromedial prefrontal cortex. cortex. J. J. Int. Neuropsychol. Soc. 12, 224–235 55 Mendez, M.F. et al. (2005) (2005) An invest investigatio igation n of moral judgme judgment nt in frontotemporal dementia. Cogn. dementia. Cogn. Behav. Neurol. 18, Neurol. 18, 193–197 56 Hoffm Hoffman, an, M. (1983) (1983) Affect Affective ive and cognit cognitive ive proces processes ses in mor moral al interna inte rnaliz lizati ation: on: an inf inform ormati ation on proces processing sing app approa roach. ch. In Social Cognition and Social Development (Higgins, E. et al., al., eds), pp. 236– 274, Cambridge University Press 57 Mikhai Mikhail, l, J. (2008 (2008)) The pover poverty ty of the moral moral stimul stimulus. us. In Moral Psychology: The Evolution of Morality Morality (Sinnott-Armstrong, W., ed.), pp. 353–360, MIT Press 58 Raine, Raine, A. and Yang, Y. (2006) Neural found foundations ations to moral reasoning and antisocial behavior. behavior. Soc. Cogn. Affect. Neurosci. 1, Neurosci. 1, 203–213 59 Anderson, Anderson, S.W. S.W. et et al. (1999) al. (1999) Impairment of social and moral behavior related to early damage in human prefrontal cortex. Nat. Neurosci. 2, Neurosci. 2, 1032–1037 60 Dwyer, Dwyer, S. Mor Moral al dumbfo dumbfoundi unding ng andthe linguis linguistic tic analogy. analogy. Mind Mind Lang. (in press) 61 Kohlberg, L. (1969) Stage and sequence: the cognitive-developmental approach to socialization. In Moral In Moral Development and Behavior (Goslin, Behavior (Goslin, D., ed.), pp. 31–53, Holt, Reinhart, and Winston 62 Strawson, Strawson, P. (196 (1960) 0) Freedom and resent resentment. ment. Proceedin Proceedings gs of the 47, 1–25 British Academy Academy 47, 63 Dwyer, Dwyer, S. (199 (1999) 9) Moral competence competence.. In Philosophy and Linguistics (Murasugi, K. and Stainton, R., eds), pp. 169 –190, Westvew Press 64 Cushman, Cushman, F. F. et et al. (2006) al. (2006) The role of reasoning and intuition in moral judgments: testing three principles of harm. Psychol. Sci. 17, 1082– 1089 65 Slater, Slater, M. et al (2006) A virtual virtual reprise reprise of the stanle stanley y milgra milgram m al.. (2006) obedience experiments. PLoS experiments. PLoS One 1, One 1, e39
judgments more severe. Psychol. severe. Psychol. Sci. 16, Sci. 16, 780–784 45 Valdeso Valdesolo, lo, P. and DeSteno, D. (2006) (2006) Manipulatio Manipulations ns of emotional emotional context shape moral judgment. Psychol. Sci. 17, Sci. 17, 476–477
66 Kinga Kinga,, J. et J. et al. (2005 (2005)) Doi Doing ng the rig right ht thin thing: g: a com commonneura monneurall cir circuitfor cuitfor appropriate violent or compassionate behavior. Neuroimage behavior. Neuroimage 30, 30, 1069– 1076
6
