DOI: 10.1126/science.1220999, 874 (2012);336 Science

Frans B. M. de WaalThe Antiquity of Empathy

This copy is for your personal, non-commercial use only.

clicking here.colleagues, clients, or customers by , you can order high-quality copies for yourIf you wish to distribute this article to others

  here.following the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

  ): April 17, 2013 www.sciencemag.org (this information is current as of

The following resources related to this article are available online at

http://www.sciencemag.org/content/336/6083/874.full.htmlversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/content/336/6083/874.full.html#relatedfound at:

can berelated to this article A list of selected additional articles on the Science Web sites

http://www.sciencemag.org/content/336/6083/874.full.html#ref-list-1, 13 of which can be accessed free:cites 23 articlesThis article

http://www.sciencemag.org/cgi/collection/psychologyPsychology

subject collections:This article appears in the following

registered trademark of AAAS. is aScience2012 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

Apr

il 17

, 201

3w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

14. G. A. Bonanno, M. Westphal, A. D. Mancini, Annu. Rev.Clin. Psychol. 7, 511 (2011).

15. A. B. Adler, P. D. Bliese, D. McGurk, C. W. Hoge,C. A. Castro, J. Consult. Clin. Psychol. 77, 928 (2009).

16. R. J. McNally, R. A. Bryant, A. Ehlers, Psychol. Sci. PublicInterest 4, 45 (2003).

17. K. Mulligan et al., J. Consult. Clin. Psychol. (2012).18. J. S. Goldstein, Winning the War on War: The Decline

of Armed Conflict Worldwide (Dutton, New York,2011).

19. S. Pinker, The Better Angels of Our Nature: Why ViolenceHas Declined (Viking, New York, 2011).

Acknowledgments: I thank N. Breslau, D. M. Wegner,and S. Wessely for helpful comments on previous draftsof this article.

10.1126/science.1222069

PERSPECTIVE

Frans B. M. de Waal

The view of humans as violent war-prone apes is poorly supported by archaeological evidence and onlypartly supported by the behavior of our closest primate relatives, chimpanzees and bonobos. Whereas thefirst species is marked by xenophobia, the second is relatively peaceful and highly empathic in bothbehavior and brain organization. Animal empathy is best regarded as a multilayered phenomenon,built around motor mirroring and shared neural representations at basal levels, that develops into moreadvanced cognitive perspective-taking in large-brained species. As indicated by both observational andexperimental studies on our closest relatives, empathy may be the main motivator of prosocial behavior.

After the devastations of World War II,humans were routinely depicted as “killerapes”—in contrast to the real apes, which

were regarded as pacifists. Books by KonradLorenz, theAustrian ethologist, andRobertArdrey,an American journalist, contributed to the ideathat a hallmark of humanity is aggression. Untilwell into the 1980s, this remained the dominanttheme of biological approaches to human behav-ior. This literature is now recognized as one-sidedbecause it overlooked our species' capacity forcooperation, empathy, and prosocial behavior.

Species-typical tendencies normally comewith built-in rewards. Nature has ensured thatwe find fulfillment in eating, sex, nursing, andsocializing, all of which are necessary for survivaland reproduction. If there were truly a geneticbasis to our participation in lethal combat, weshould willingly engage in it. Yet soldiers report adeep revulsion to killing and shoot at the enemyonly under pressure (1). After these experiences,they often end up with substantial psychologicaldamage. Far from being a recent phenomenon,hauntingmemories of combat were already knownto the ancient Greeks, such as Sophocles, whodescribed Ajax's “divine madness,” now knownas posttraumatic stress disorder (PTSD).

Even though evidence for individual murdergoes back hundreds of thousands of years, com-parable signs of warfare (such as graveyards withweapons embedded in a large number of skele-tons) are lacking from before the Agricultural Rev-olution [about 12,000 years ago (2)]. This is not toimply that war was absent before then, but it doesmean that the common assumption that our ances-tors waged perpetual wars and knew peace only at“precarious interludes” (as Winston Churchill sur-

mised) lacks solid archaeological backing. Duringmost of our prehistory, we were nomadic hunter-gatherers, whose cultures are nowadays not partic-ularly known for warfare (3). They do occasionallyraid, ambush, and kill their neighbors (4), but moreoften trade with them, intermarry, and permit travelthrough their territories. Hunter-gatherers illustratea robust potential for peace and cooperation.

Going back farther in time, we end up withArdipithecus ramidus, a 4.4-million-year-old hom-inin that has been described as relatively peaceful,owing to its reduced canine teeth as compared tothose of the chimpanzee (Pan troglodytes) (5), whocan be lethally violent during territorial encountersbetween communities. However, the conclusiondrawn from Ardipithecus' dentition that our an-cestors were less war-prone than the apes is notrigorous unless the bonobo (P. paniscus), whichalso has relatively small canines (Fig. 1), is in-cluded. Despite being as closely related to us aschimpanzees, the behavior of bonobos fails to sup-port traditional violence-based scenarios of humanevolution. Deadly aggression among bonobos hasthus far not been observed, neither in captivity norin the wild, and xenophobia is only weakly devel-oped. Bonobos sometimes mingle across territorialborders, where they engage in sex, grooming, andplay. They are known as the “make love, not war”primates for solving dominance issues throughsexual activity (6). Indeed, it has been suggestedthat these apes “may approach more closely to thecommon ancestor of chimpanzees and man thandoes any living chimpanzee” (7).

In addition, developments in psychology, neu-roscience, behavioral economics, and animal be-havior have begun to question the view, dominantuntil a decade ago, that animal life, and by ex-tension human nature, is based on unmitigatedcompetition. In primatology, the countermove-ment started with research into the survival valueof friendships (8) and conflict resolution (9).

After the discovery that chimpanzees often kissand embrace shortly after a fight within theirgroup, numerous studies have documented “rec-onciliations” in nonhuman primates. Method-ologies comparing postconflict observations withbaseline data to determine how species membersbehave in the presence versus absence of previ-ous antagonism show that primates are generallyattracted to former opponents, seeking friendlycontact especially if they otherwise enjoy a mu-tually beneficial relationship. Relationship valueappears to drive post-conflict repair (10). Thebehavioral expression of reconciliation varies,but its general effect is a rapid return to preex-isting levels of tolerance and affiliation. Thisreunion process has been reported for macaques,gorillas, golden monkeys, capuchins, and manyother primates, but also for nonprimates, such aswolves, dolphins, and hyenas. Reconciliation is acommon social mechanism that would be super-fluous if social life were ruled entirely by dom-inance and competition.

The level of cooperation among nonhumanprimates tends to be underappreciated. In order toset it apart from human cooperation with non-relatives, aid among primates is sometimes as-cribed largely to kinship (11). This claim has notheld up, however, on the basis of DNA extractedfrom chimpanzee feces in the wild. Males with-out genetic ties make up the majority of mutuallysupportive partnerships (12). The same seems toapply to bonobos. Female bonobos maintain aclose social network that allows them to collect-ively dominate the majority of males despitethe fact that females are also the migratory sex,which means that they are largely unrelated with-in each community (6). Both of our closest primaterelatives are marked, therefore, by high levelsof nonkin cooperation, probably explained bywell-developed reciprocity.

Expressions of empathy are common in apesand resemble those of our own species. In childresearch, for example, a family member is typ-ically instructed to feign distress or pain, uponwhich touching, stroking, and close-up eye-contactby the child is interpreted as a sign of sympatheticconcern. In chimpanzees, bystanders at a fight goover to the loser and put an arm around his or hershoulders or provide other calming contact (Fig. 2).Data from several thousand postconflict observa-tions in chimpanzees indicate that consolationreduces the recipient's arousal and follows thesame sex difference as reported for sympatheticconcern in children, with female apes providingcomfort more often than males (13). Bonobos ex-press the same tendency sociosexually by means

Living Links, Yerkes National Primate Research Center; andDepartment of Psychology, Emory University, Atlanta, GA30322, USA. E-mail: dewaal@emory.edu

18 MAY 2012 VOL 336 SCIENCE www.sciencemag.org874

on

Apr

il 17

, 201

3w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

of genital contacts. A comparison of chimpanzeeand bonobo brains supports the general view ofbonobos as more empathic (6). This species hasmore gray matter in brain regions involved in theperception of distress, including the right dorsalamygdala and right anterior insula, and a betterdeveloped circuitry for inhibiting aggression (14).

Human empathy has been described as takingthe perspective of another or imagining oneself inanother's position. Psychologists commonly applythis cognitively demanding explanation of empathyeven if the immediacy of the response hints atsimpler processes. If we see a child fall and scrapeits knee, we flinch, and exclaim “ouch!” as if whathappened to the child happened at the same instantto ourselves. Another way of looking at empathy isas a multilayered phenomenon that starts withautomatic state-matching based on motor mimicry

and shared neural representations(15). We should not be surprised,therefore, by unconscious empa-thy, such as when human studyparticipants mimic observed fa-cial expressions and report cor-responding emotions even thoughthe expressions were presented toobriefly for conscious perception(16). In this view,whichmaymaponto nested neural processing (17),cognitive perspective-taking isa secondary development builtaround more elementary mecha-nisms, such as state-matching andemotional contagion (Fig. 3).

The evolution of empathy isthought to go back to mamma-

lian maternal care.Whether amouse oranelephant, amotherneeds to be exquis-itely in tune with in-dications of hunger,danger, or discomfort in her young.Sensitivity to emotional signals con-fers clear adaptive value. This hypo-thetical origin of empathy wouldexplain the observed sex differencesas well as the stimulating effect ofoxytocin (18). That empathy is rootedin bodily connections between indi-viduals is reflected in pain contagionin mice (19) and yawn contagionin apes and humans (20). Mirrorneurons are often mentioned in thiscontext, even though their preciserole remains a point of speculation.The fact that these neurons werediscovered not in humans, but inmonkeys, supports the idea of evolu-tionary continuity. Social animalsneed to coordinate travel, communi-cate about danger, and assist groupmates in need. Bodily synchroniza-tion and sensitivity to the emotionalstates of others ranges from rapidspreading of alarm through an entiregroup to a mother ape returning to awhimpering youngster to help itfrom one tree to the next by drapingher body between the two. The firstis a reflex-like transmission of fear,whereas the mother ape is more dis-criminating because she needs toassess the reason for her offspring'sdistress in order to ameliorate itssituation.

The idea that empathy translatesinto altruism and helping is widelyassumed for humans and has alsobeen proposed for other mammals(21). Reports of spontaneous assist-

ance among primates are abundant and are alsoavailable for elephants and cetaceans. For ex-ample, a female chimpanzee may react to thescreams of her closest associate by defending heragainst an aggressive male, thus taking great riskon her behalf. Such coalitions are among themost systematically studied forms of coopera-tion in primatology (22). One advantage of anempathy-based explanation is its ability to ex-plain “unrepaid” altruism, such as that showntoward nonreciprocating nonkin. This type ofbehavior is well illustrated by the adoption oforphans by wild male chimpanzees, who maydevote years of costly care to unrelated juveniles(23). Although empathy, such as between a moth-er and offspring or between cooperation partners,is likely to be adaptive, not each and every ap-plication of this capacity needs to be for it toretain overall adaptive value.

Increasingly, the importance of mammalianprosocial tendencies is backed by experimentsthat range from demonstrating that rats givepriority to the liberation of a trapped companionover eating chocolate (24) to those showing thatapes are prepared to assist others even in theabsence of incentives, go out of their way to giveothers access to food, or choose shared benefitsover selfish ones (25, 26). Studies have also dem-onstrated more complex expressions of empathyamong apes, such as “targeted helping” (assistancebased on an appreciation of the other's specificneeds), both in their spontaneous behavior (21)andduringcontrolledexperiments (27).This increasedknowledge suggests that nearly the full spectrumof empathy-based altruism may be representedamong nonhuman primates, including the cognitiveperspective-taking that marks human altruism.

Empathy automatically produces a stake in an-other's welfare; that is, the behavior comes with

Fig. 1. Two bipedal bonobos, an adult female (left) and ado-lescent male (right), show the species' relatively long legs, whichmakes it anatomically more similar to early hominins than to thechimpanzee, with its longer arms and shorter legs (photographby Frans de Waal). Ardipithecus ramidus may be the closest com-parison to the bonobo in terms of its overall body proportions,grasping feet, and reduced canine teeth. Ardipithecus is thought tohave been relatively peaceful, and bonobos are likewise marked byhigh sensitivity to others and low levels of violence.

Fig. 2. Contact comfort is critically important in the lives of apes,such as here between two chimpanzees watching a disturbance intheir group (photograph by Frans de Waal). Apes go out of theirway to console distressed parties, showing the same sex differencein this tendency as humans.

www.sciencemag.org SCIENCE VOL 336 18 MAY 2012 875

SPECIALSECTION

on

Apr

il 17

, 201

3w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

an intrinsic reward, known as the “warm glow”effect. Humans report feeling good when they dogood and show activation of reward-related brain

areas (28). It will be important to determinewhether the same self-reward system extends toother primates.We do know from studies on rodents,apes, and humans that empathy is biased towardthe ingroup. For example, while watching the yawnsof videotaped conspecifics, chimpanzees frequently

joined the yawns of their own group membersbut not those of unfamiliar individuals (23). Thisingroup bias makes sense from an evolutionary

perspective, because it is with themembers of one's own group thatapes cooperate. At the same time,however, it poses a profound chal-lenge for the modern human world,which seeks to integrate a multitudeof groups, ethnicities, and nations.The flip side of the ingroup bias inempathy is lack of empathy for theoutgroup, as is typical of xenophobia.

Nevertheless, empathy may beour only hope to deal with theseissues. We know that it can be ac-tivated by outsiders, even by mem-bers of a different species, such aswhen we empathize with a strandedwhale and move it back into theocean. This is not an outcome forwhich empathy evolved, yet oncein existence, capacities are oftenemancipated from their evolution-ary origin. If it weren't for empathywith all life forms, including ene-my lives, soldiers would have noreluctance to kill nor would they re-turn from the battlefield with PTSD.Although it is true that empathy hastrouble reaching beyond the in-group, it is an automated responsethat does not allow itself to be fullysuppressed by rationalizations and

political indoctrination. This is another lessonfromWorld War II, with examples such as OskarSchindler and the guardians of Anne Frank. Tobetter understand the power of empathy requiresinvestigation of its neurological basis as well asits evolutionary antiquity.

References1. D. Grossman, On Killing: The Psychological Cost of

Learning to Kill in War and Society (Back Bay Books,New York, 1995).

2. I. J. N. Thorpe, World Archaeol. 35, 145 (2003).3. D. P. Fry, The Human Potential for Peace (Oxford Univ.

Press, New York, 2006).4. R. W. Wrangham, L. Glowacki, Hum. Nat. 23, 5 (2012).5. C. O. Lovejoy, Science 326, 74 (2009).6. F. B. M. de Waal, Bonobo: The Forgotten Ape (Univ. of

California Press, Berkeley, CA, 1997).7. H. J. Coolidge, Am. J. Phys. Anthropol. 18, 1 (1933).8. B. B. Smuts, Sex and Friendship in Baboons (Harvard Univ.

Press, Cambridge, MA, 1999) (originally published in 1985).9. F. B. M. de Waal, A. van Roosmalen, Behav. Ecol. Sociobiol.

5, 55 (1979).10. F. B. M. de Waal, Science 289, 586 (2000).11. R. Boyd, Science 314, 1555 (2006).12. K. E. Langergraber, J. C. Mitani, L. Vigilant, Proc. Natl.

Acad. Sci. U.S.A. 104, 7786 (2007).13. T. Romero, M. A. Castellanos, F. B. M. de Waal,

Proc. Natl. Acad. Sci. U.S.A. 107, 12110 (2010).14. J. K. Rilling et al., Soc. Cogn. Affect. Neurosci. 7, 369 (2012).15. S. D. Preston, F. B. M. de Waal, Behav. Brain Sci. 25, 1,

discussion 20 (2002).16. U. Dimberg, M. Thunberg, K. Elmehed, Psychol. Sci. 11,

86 (2000).17. J. Panksepp, Science 334, 1358 (2011).18. P. J. Zak, A. A. Stanton, S. Ahmadi, PLoS ONE 2, e1128 (2007).19. D. J. Langford et al., Science 312, 1967 (2006).20. M. W. Campbell, F. B. M. de Waal, PLoS ONE 6, e18283

(2011).21. F. B. M. de Waal, Annu. Rev. Psychol. 59, 279 (2008).22. A. H. Harcourt, F. B. M. de Waal, Coalitions and Alliances in

Humans and Other Animals (Oxford Univ. Press, Oxford, 1992).23. C. Boesch, C. Bolé, N. Eckhardt, H. Boesch, PLoS ONE 5,

e8901 (2010).24. I. Ben-Ami Bartal, J. Decety, P. Mason, Science 334,

1427 (2011).25. F. Warneken, B. Hare, A. P. Melis, D. Hanus, M. Tomasello,

PLoS Biol. 5, e184 (2007).26. V. Horner, J. D. Carter, M. Suchak, F. B. M. de Waal,

Proc. Natl. Acad. Sci. U.S.A. 108, 13847 (2011).27. S. Yamamoto, T. Humle, M. Tanaka, Proc. Natl. Acad.

Sci. U.S.A. 109, 3588 (2012).28. W. T. Harbaugh, U. Mayr, D. R. Burghart, Science 316,

1622 (2007).

10.1126/science.1220999

PERSPECTIVE

Samuel Bowles

The origins of such varied features of contemporary life as the national state and the desire touphold generous and civic social norms are to be found in a combination of conflict between groupsand attenuation of both inequalities and conflicts within groups. In contrast to the adoption of abetter tool or a more productive crop, which can be adopted by a single individual, a new institutionworks only if most people adopt it. This explains why collective action against those benefittingfrom the status quo at the expense of others, as well as conflict between groups governed bydifferent norms and institutions, figures so prominently in our capacity to adapt to changingcircumstances and to harness new knowledge for human benefit.

Conflict has a bad name, one that it richlydeserves for the suffering, tragedy, andwaste of human and material resources

that it brings about. But conflict—both violentand civil, both within and between societies—hasalso been a midwife for humanity’s most

cherished values and institutions: among themdemocracy, the rule of law, and a propensity tohelp others and to abhor injustice.

I will make the case that it was warfare thatculled Europe’s once-motley collection of gov-ernments to produce the modern national state,which, as a result of subsequent conflicts withinnations, would become liberal and eventuallydemocratic. This occurred because, not contentto free ride on the sacrifices of others, peoplewere willing to take mortal risks in pursuit ofdemocratic and liberal values. And this, if I amright, is itself a result of millennia of conflict be-tween groups of ancestral humans where, CharlesDarwin wrote, the groups with large numbers of“courageous, sympathetic and faithful members,who were always ready to...aid and defend each

Perspective-takingtargeted helping

Sympathetic concernconsolation

State-matchingemotional contagion

Fig. 3. The Russian doll model of multilayered empathy. The doll'sinner core consists of the perception-action mechanism (PAM)that underlies state-matching and emotional contagion (15). Builtaround this hard-wired socioaffective basis, the doll's outer layersinclude sympathetic concern and targeted helping. The complexityof empathy grows with increasing perspective-taking capacities,which depend on prefrontal neural functioning, yet remainfundamentally connected to the PAM. A few large-brainedspecies show all of the doll's layers, but most show only theinner ones.

Santa Fe Institute, 1399 Hyde Park Road, Santa Fe, NM87501, USA; and University of Siena, Siena, Italy. E-mail:samuel.bowles@gmail.com

18 MAY 2012 VOL 336 SCIENCE www.sciencemag.org876

on

Apr

il 17

, 201

3w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from