ANRV331-PS59-11 ARI 4 November 2007 20:27 Putting the Altruism Back into Altruism: The Evolution of Empathy Frans B.M. de Waal Living Links, Yerkes National Primate Research Center, and Psychology Department, Emory University, Atlanta, Georgia 30322; email: [email protected] Annu. Rev. Psychol. 2008. 59:279–300 Key Words First published online as a Review in Advance on perception-action, perspective-taking, prosocial behavior, June 5, 2007 cooperation The Annual Review of Psychology is online at http://psych.annualreviews.org Abstract This article’s doi: Evolutionary theory postulates that altruistic behavior evolved for 10.1146/annurev.psych.59.103006.093625 by EMORY UNIVERSITY on 01/23/08. For personal use only. the return-benefits it bears the performer. For return-benefits to play Copyright c 2008 by Annual Reviews. a motivational role, however, they need to be experienced by the or- All rights reserved ganism. Motivational analyses should restrict themselves, therefore, 0066-4308/08/0203-0279$20.00 to the altruistic impulse and its knowable consequences. Empathy Annu. Rev. Psychol. 2008.59:279-300. Downloaded from arjournals.annualreviews.org is an ideal candidate mechanism to underlie so-called directed al- truism, i.e., altruism in response to another’s pain, need, or distress. Evidence is accumulating that this mechanism is phylogenetically an- cient, probably as old as mammals and birds. Perception of the emo- tional state of another automatically activates shared representations causing a matching emotional state in the observer. With increasing cognition, state-matching evolved into more complex forms, includ- ing concern for the other and perspective-taking. Empathy-induced altruism derives its strength from the emotional stake it offers the self in the other’s welfare. The dynamics of the empathy mechanism agree with predictions from kin selection and reciprocal altruism theory. 279 ANRV331-PS59-11 ARI 4 November 2007 20:27 “You begin with the effect of behavior on ac- Contents tors and recipients; you deal with the problem of internal motivation, which is a secondary INTRODUCTION................. 280 problem, afterward. ...[I]f you start with mo- ORIGIN OF EMPATHY............ 282 tivation, you have given up the evolutionary LEVELS OF EMPATHY ........... 282 analysis at the outset.” Emotional Contagion............. 282 This is a perfectly legitimate strategy that Sympathetic Concern............. 283 has yielded profound insights into the evo- Empathic Perspective-Taking ..... 285 lution of altruism (e.g., Dugatkin 2006). Un- UNDERLYING MECHANISMS . 286 fortunately, however, these insights have not Perception Action Mechanism .... 286 come with a new terminology: Evolutionary Russian Doll Model .............. 287 biology persists in using motivational terms. FROM EMPATHY TO Thus, an action is called “selfish” regard- ALTRUISM ..................... 288 less of whether or not the actor deliberately Emotional Contagion............. 288 seeks benefits for itself. Similarly, an action is Sympathetic Concern............. 289 called “altruistic” if it benefits a recipient at Empathic Perspective-Taking ..... 289 a cost to the actor regardless of whether or EMPATHY AS EVOLVED not the actor intended to benefit the other. PROXIMATE MECHANISM OF The prototypical altruist is a honeybee that DIRECTED ALTRUISM ........ 291 stings an intruder—sacrificing her life to pro- CONCLUSION .................... 292 tect the hive—even though her motivation is more likely aggressive than benign. This us- age of the terms “selfish” and “altruistic” of- Sympathy ...cannot, in any sense, be tentimes conflicts with their vernacular mean- regarded as a selfish principle. ing (Sober & Wilson 1998). Smith (1759, p. 317) The hijacking of motivational terminol- Empathy may be uniquely well suited for ogy by evolutionary biologists has been un- Altruism bridging the gap between egoism and altru- helpful for communication about motivation (biological ism, since it has the property of transforming definition): per se. The way to clear up the confusion is behavior that another person’s misfortune into one’s own to do what Trivers did when he decided that increases the feeling of distress. evolutionary analyses require that effects be recipient’s fitness at a Hoffman (1981a, p. 133) by EMORY UNIVERSITY on 01/23/08. For personal use only. considered separate from motivation. Con- cost to the versely, motivational analyses require us to performers keep motivation separate from evolutionary Ultimate cause or INTRODUCTION considerations. It is not for nothing that biol- Annu. Rev. Psychol. 2008.59:279-300. Downloaded from arjournals.annualreviews.org goal: the benefits an organism or its close Discussions of altruistic behavior tend to suf- ogists hammer on the distinction between ul- kin derive from a fer from a lack of distinction between function timate and proximate (Mayr 1961, Tinbergen behavior, hence the and motivation. This is due to the contrasting 1963). The ultimate cause refers to why a probable reason why emphasis of biologists and psychologists, with behavior evolved over thousands of gener- the behavior was the former focusing on what a particular be- ations, which depends on its fitness conse- favored by natural selection havior is good for, and the latter on how it quences. The proximate cause, on the other comes about. hand, refers to the immediate situation that Proximate cause: situation that Evolutionary explanations are built around triggers behavior, and the role of learning, triggers behavior and the principle that all that natural selection can physiology, and neural processes—typically the mechanism work with are the effects of behavior—not the the domain of psychologists. (psychological, motivation behind it. This means there is only Proximate and ultimate viewpoints do in- neural, physiological) one logical starting point for evolutionary ac- form each other, yet are not to be con- that enables it counts, as explained by Trivers (2002, p. 6): flated. For example, primate cooperation is 280 de Waal ANRV331-PS59-11 ARI 4 November 2007 20:27 promoted by social tolerance. Through its ef- trinsically rewarding qualities in that it of- fect on food-sharing, tolerance evens out pay- fers the actor an emotional stake in the re- off distributions (de Waal & Davis 2003, Melis cipient’s well-being, i.e., if helping the other Directed altruism: et al. 2006). Tolerance likely is a proximate ameliorates the helper’s internal state (see helping or mechanism that evolved to serve the ultimate Empathy as Evolved Proximate Mechanism, comforting behavior goal of cooperation, which is to yield benefits below). Extrinsic rewards, on the other hand, directed at an for all contributors. are less likely to play a role. By definition, al- individual in need, Cooperation and altruistic behavior are truism carries an initial cost, and positive con- pain, or distress thought to have evolved to help family mem- sequences occur only after a significant time Intentional bers and those inclined to return the favor interval (e.g., the recipient reciprocates) or altruism: the altruist deliberately (Hamilton 1964, Trivers 1971). Regardless of not at all (e.g., care for dependent kin), making seeks to benefit whether this is the whole explanation or not for rather poor learning conditions. either the other (see Sober & DS Wilson 1998, EO Wilson Intentionally selfish altruism would re- (intentionally 2005), the point is that ultimate accounts quire the actor to explicitly expect others to altruistic altruism) or stress return-benefits, i.e., positive conse- return the favor. Despite the lack of evidence itself (intentionally selfish altruism) quences for the performer and/or its kin. Inas- for such expectations in animals, they are of- much as these benefits may be quite delayed, ten assumed. The common claim that humans Empathy-based altruism: help and however, it is unclear what motivational role, are the only truly altruistic species, since all care born from if any, they play. This becomes clear if we con- that animals care about are return-benefits empathy with sider more closely what drives directed altru- (e.g., Dawkins 1976, Fehr & Fischbacher another ism, i.e., altruistic behavior aimed at others in 2003, Kagan 2000, Silk et al. 2005), miscon- Empathy: the need, pain, or distress. There are three ways strues reciprocity as a motivation. It assumes capacity to (a)be in which directed altruism may come about: that animals engage in reciprocal exchange affected by and share with a full appreciation of how it will ulti- the emotional state 1. Altruistic impulse. Spontaneous, disin- of another, (b) assess terested helping and caring in reaction mately benefit them. Helpful acts for imme- the reasons for the to begging or distress signals or the sight diate self-gain are indeed common (Dugatkin other’s state, and ( ) identify with the of another in pain or need. 1997), but the return-benefits of altruistic be- c havior typically remain beyond the animal’s other, adopting his 2. Learned altruism. Helping as a condi- cognitive horizon, i.e., occur so distantly in or her perspective. tioned response reinforced by positive This definition time that the organism is unlikely to con- outcomes for the actor. extends beyond what nect them with the original act. This ap- exists in many 3. Intentional altruism. Help based on the by EMORY UNIVERSITY on 01/23/08. For personal use only. plies to most reciprocal altruism in the animal animals, but the term prediction of behavioral effects. One kingdom.