Primates and Philosophers_How Morality Evolved

by Frans de Waal

  Biologists prefer bottom-up over top-down accounts, even though there is definitely room for the latter. Once higher order processes have come into existence, they modify processes at the base. The central nervous system is a good example of top-down processing, as in the control the prefrontal cortex exerts over memory. The prefrontal cortex is not the seat of memory, but can “order” memory retrieval (Tomita et al. 1999). In the same way, culture and language shape expressions of empathy. The distinction between “being the origin of” and “shaping” is a fundamental one, though, and I will argue here that empathy is the original, pre-linguistic form of inter-individual linkage that only secondarily has come under the influence of language and culture.

  Bottom-up accounts are the opposite of Big Bang theories. They assume continuity between past and present, child and adult, human and animal, even between humans and the most primitive mammals. We may assume that empathy first evolved in the context of parental care, which is obligatory in mammals (Eibl-Eibesfeldt 1974 [1971]; MacLean 1985). Signaling their state through smiling and crying, human infants urge their caregiver to pay attention and move into action (Bowlby 1958). The same applies to other primates. The survival value of these interactions is obvious. For example, a female chimpanzee lost a succession of infants despite intense positive interest because she was deaf and did not correct positional problems (such as sitting on the infant, or holding it upside-down) in response to its distress calls (de Waal 1998 [1982]).

  For a human characteristic, such as empathy, that is so pervasive, develops so early in life (e.g., Hoffman 1975; Zahn-Waxler and Radke-Yarrow 1990), and shows such important neural and physiological correlates (e.g., Adolphs et al. 1994; Rimm-Kaufman & Kagan 1996; Decety and Chaminade 2003) as well as a genetic substrate (Plomin et al. 1993), it would be strange indeed if no evolutionary continuity existed with other mammals. The possibility of empathy and sympathy in other animals has been largely ignored, however. This is partly due to an excessive fear of anthropomorphism, which has stifled research into animal emotions (Panksepp 1998; de Waal 1999, appendix A), and partly to the one-sided portrayal by biologists of the natural world as a place of combat rather than social connectedness.

  What Is Empathy?

  Social animals need to coordinate action and movement, collectively respond to danger, communicate about food and water, and assist those in need. Responsiveness to the behavioral states of conspecifics ranges from a flock of birds taking off all at once because one among them is startled by a predator to a mother ape who returns to a whimpering youngster to help it from one tree to the next by draping her body as a bridge between the two. The first is a reflex-like transmission of fear that may not involve any understanding of what triggered the initial reaction, but that is undoubtedly adaptive. The bird that fails to take off at the same instant as the rest of the flock may be lunch. The selection pressure on paying attention to others must have been enormous. The mother-ape example is more discriminating, involving anxiety at hearing one’s offspring whimper, assessment of the reason for its distress, and an attempt to ameliorate the situation.

  There exists ample evidence of one primate coming to another’s aid in a fight, putting an arm around a previous victim of attack, or other emotional responses to the distress of others (to be reviewed below). In fact, almost all communication among nonhuman primates is thought to be emotionally mediated. We are familiar with the prominent role of emotions in human facial expressions (Ekman 1982), but when it comes to monkeys and apes—which have a homologous array of expressions (van Hooff 1967)—emotions seem equally important.

  When the emotional state of one individual induces a matching or closely related state in another, we speak of “emotional contagion” (Hatfield et al. 1993). Even if such contagion is undoubtedly a basic phenomenon, there is more to it than simply one individual being affected by the state of another: the two individuals often engage in direct interaction. Thus, a rejected youngster may throw a screaming tantrum at its mother’s feet, or a preferred associate may approach a food possessor to beg by means of sympathy-inducing facial expressions, vocalizations, and hand gestures. In other words, emotional and motivational states often manifest themselves in behavior specifically directed at a partner. The emotional effect on the other is not a by-product, therefore, but actively sought.

  With increasing differentiation between self and other, and an increasing appreciation of the precise circumstances underlying the emotional states of others, emotional contagion develops into empathy. Empathy encompasses—and could not possibly have arisen without—emotional contagion, but it goes beyond it in that it places filters between the other’s and one’s own state. In humans, it is around the age of two that we begin to add these cognitive layers (Eisenberg and Strayer 1987).

  Two mechanisms related to empathy are sympathy and personal distress, which in their social consequences are each other’s opposites. Sympathy is defined as “an affective response that consists of feelings of sorrow or concern for a distressed or needy other (rather than the same emotion as the other person). Sympathy is believed to involve an other-oriented, altruistic motivation” (Eisenberg 2000: 677). Personal distress, on the other hand, makes the affected party selfishly seek to alleviate its own distress, which is similar to what it has perceived in the object. Personal distress is therefore not concerned with the situation of the empathy-inducing other (Batson 1990). A striking primate example is given by de Waal (1996: 46): the screams of a severely punished or rejected infant rhesus monkey will often cause other infants to approach, embrace, mount, or even pile on top of the victim. Thus, the distress of one infant seems to spread to its peers, which then seek contact to soothe their own arousal. Inasmuch as personal distress lacks cognitive evaluation and behavioral complementarity, it does not reach beyond the level of emotional contagion.

  That most modern textbooks on animal cognition (e.g., Shettleworth 1998) fail to index empathy or sympathy does not mean that these capacities are not an essential part of animal lives; it only means that they are being overlooked by a science traditionally focused on individual rather than inter-individual capacities. Tool use and numerical competence, for instance, are seen as hallmarks of intelligence, whereas appropriately dealing with others is not. It is obvious, however, that survival often depends on how animals fare within their group, both in a cooperative sense (e.g., concerted action, information transfer) and in a competitive sense (e.g., dominance strategies, deception). It is in the social domain, therefore, that one expects the highest cognitive achievements. Selection must have favored mechanisms to evaluate the emotional states of others and quickly respond to them. Empathy is precisely such a mechanism.

  In human behavior, there exists a tight relation between empathy and sympathy, and their expression in psychological altruism (e.g., Hornblow 1980; Hoffman 1982; Batson et al. 1987; Eisenberg and Strayer 1987; Wispe 1991). It is reasonable to assume that the altruistic and caring responses of other animals, especially mammals, rest on similar mechanisms. When Zahn-Waxler visited homes to find out how children respond to family members instructed to feign sadness (sobbing), pain (crying), or distress (choking), she discovered that children a little over one year of age already comfort others. Since expressions of sympathy emerge at an early age in virtually every member of our species, they are as natural as the first step. An unplanned sidebar to this study, however, was that household pets appeared as worried as the children by the “distress” of family members. They hovered over them or put their heads in their laps (Zahn-Waxler et al. 1984).

  Rooted in attachment and what Harlow termed the “affectional system” (Harlow and Harlow 1965), responses to the emotions of others are commonplace in social animals. Thus, behavioral and physiological data suggest emotional contagion in a variety of species (reviewed in Preston and de Waal 2002b, and de Waal 2003). An interesting literature that appeared in the 1950s and “60s by experimental psychologists placed the words “empathy” and “sympathy�
� between quotation marks. In those days, talk of animal emotions was taboo. In a paper provocatively entitled “Emotional Reactions of Rats to the Pain of Others,” Church (1959) established that rats that had learned to press a lever to obtain food would stop doing so if their response was paired with the delivery of an electric shock to a visible neighboring rat. Even though this inhibition habituated rapidly, it suggested something aversive about the pain reactions of others. Perhaps such reactions arouse negative emotions in rats that witness them.

  Monkeys show a stronger inhibition than rats. The most compelling evidence for the strength of empathy in monkeys came from Wechkin et al. (1964) and Masserman et al. (1964). They found that rhesus monkeys refuse to pull a chain that delivers food to themselves if doing so shocks a companion. One monkey stopped pulling for five days, and another one for twelve days after witnessing shock delivery to a companion. These monkeys were literally starving themselves to avoid inflicting pain upon another. Such sacrifice relates to the tight social system and emotional linkage among these macaques, as supported by the finding that the inhibition to hurt another was more pronounced between familiar than unfamiliar individuals (Masserman et al. 1964).

  Although these early studies suggest that, by behaving in certain ways, animals try to alleviate or prevent distress in others, it remains unclear if spontaneous responses to distressed conspecifics are explained by (a) aversion to distress signals of others, (b) personal distress generated through emotional contagion, or (c) true helping motivations. Work on nonhuman primates has furnished further information. Some of this evidence is qualitative, but quantitative data on empathic reactions exists as well.

  Anecdotes of “Changing Places in Fancy”

  Striking depictions of primate empathy and altruism can be found in Yerkes (1925), Ladygina-Kohts (2002 [1935]), Goodall (1990), and de Waal (1998 [1982], 1996, 1997a). Primate empathy is such a rich area that O’Connell (1995) was able to conduct a content analysis of thousands of qualitative reports. She concluded that responses to the distress of another seem considerably more complex in apes than monkeys. To give just one example of the strength of the ape’s empathic response, Ladygina-Kohts wrote about her young chimpanzee, Joni, that the best way to get him off the roof of her house (much better than any reward or threat of punishment) was by arousing his sympathy:

  If I pretend to be crying, close my eyes and weep, Joni immediately stops his plays or any other activities, quickly runs over to me, all excited and shagged, from the most remote places in the house, such as the roof or the ceiling of his cage, from where I could not drive him down despite my persistent calls and entreaties. He hastily runs around me, as if looking for the offender; looking at my face, he tenderly takes my chin in his palm, lightly touches my face with his finger, as though trying to understand what is happening, and turns around, clenching his toes into firm fists. (Ladygina-Kohts, 2002 [1935]: 121)

  De Waal (1996, 1997a) has suggested that apart from emotional connectedness, apes have an appreciation of the other’s situation and a degree of perspective-taking (appendix B). So, the main difference between monkeys and apes is not in empathy per se, but in the cognitive overlays, which allow apes to adopt the other’s viewpoint. One striking report in this regard concerns a bonobo female empathizing with a bird at Twycross Zoo, in England:

  One day, Kuni captured a starling. Out of fear that she might molest the stunned bird, which appeared undamaged, the keeper urged the ape to let it go…. Kuni picked up the starling with one hand and climbed to the highest point of the highest tree where she wrapped her legs around the trunk so that she had both hands free to hold the bird. She then carefully unfolded its wings and spread them wide open, one wing in each hand, before throwing the bird as hard she could towards the barrier of the enclosure. Unfortunately, it fell short and landed onto the bank of the moat where Kuni guarded it for a long time against a curious juvenile. (de Waal, 1997a, p. 156)

  What Kuni did would obviously have been inappropriate towards a member of her own species. Having seen birds in flight many times, she seemed to have a notion of what would be good for a bird, thus offering us an anthropoid version of the empathic capacity so enduringly described by Adam Smith (1937 [1759]: 10) as “changing places in fancy with the sufferer.” Perhaps the most striking example of this capacity is a chimpanzee who, as in the original Theory of Mind (ToM) experiments of Premack and Woodruff (1978), seemed to understand the intentions of another and provided specific assistance:

  During one winter at the Arnhem Zoo, after cleaning the hall and before releasing the chimps, the keepers hosed out all rubber tires in the enclosure and hung them one by one on a horizontal log extending from the climbing frame. One day, Krom was interested in a tire in which water had stayed behind. Unfortunately, this particular tire was at the end of the row, with six or more heavy tires hanging in front of it. Krom pulled and pulled at the one she wanted but couldn’t remove it from the log. She pushed the tire backward, but there it hit the climbing frame and couldn’t be removed either. Krom worked in vain on this problem for over ten minutes, ignored by everyone, except Jakie, a seven-year-old Krom had taken care of as a juvenile.

  Immediately after Krom gave up and walked away, Jakie approached the scene. Without hesitation he pushed the tires one by one off the log, beginning with the front one, followed by the second in the row, and so on, as any sensible chimp would. When he reached the last tire, he carefully removed it so that no water was lost, carrying it straight to his aunt, placing it upright in front of her. Krom accepted his present without any special acknowledgment, and was already scooping up water with her hand when Jakie left. (Adapted from de Waal 1996)

  That Jakie assisted his aunt is not so unusual. What is special is that he correctly guessed what Krom was after. He grasped his auntie’s goals. Such so-called “targeted helping” is typical of apes, but rare or absent in most other animals. It is defined as altruistic behavior tailored to the specific needs of the other even in novel situations, such as the highly publicized case of Binti Jua, a female gorilla who rescued a human child at the Brookfield Zoo in Chicago (de Waal, 1996, 1999). A recent experiment demonstrated targeted helping in young chimpanzees (Warneken and Tomasello 2006).

  It is important to stress the incredible strength of the ape’s helping response, which makes these animals take great risks on behalf of others. Whereas in a recent debate about the origins of morality, Kagan (2000) considered it obvious that a chimpanzee would never jump into a cold lake to save another, it may help to quote Goodall (1990: 213) on this issue:

  In some zoos, chimpanzees are kept on man-made islands, surrounded by water-filed moats…. Chimpanzees cannot swim and, unless they are rescued, will drown if they fall into deep water. Despite this, individuals have sometimes made heroic efforts to save companions from drowning— and were sometimes successful. One adult male lost his life as he tried to rescue a small infant whose incompetent mother had allowed it to fall into the water.

  The only other animals with a similar array of helping responses are dolphins and elephants. This evidence, too, is largely descriptive (dolphins: Caldwell and Caldwell 1966; Connor and Norris 1982; elephants: Moss 1988; Payne 1998), yet here again it is hard to accept as coincidental that scientists who have watched these animals for any length of time have numerous such stories, whereas scientists who have watched other animals have few, if any.

  Consolation Behavior

  This difference between monkey and ape empathy has been confirmed by systematic studies of a behavior known as “consolation,” first documented by de Waal and van Roosmalen (1979). Consolation is defined as reassurance by an uninvolved bystander to one of the combatants in a preceding aggressive incident. For example, a third party goes over to the loser of a fight and gently puts an arm around his or her shoulders (figure 2). Consolation is not to be confused with reconciliation between former opponents, which seems mostly motivated by self-interest, such as the imperative to restore a disturbe
d social relationship (de Waal 2000). The advantage of consolation for the actor remains wholly unclear. The actor could probably walk away from the scene without any negative consequences.

  Figure 2 A typical instance of consolation in chimpanzees in which a juvenile puts an arm around a screaming adult male who has just been defeated in a fight with his rival. Photograph by the author.

  Information on chimpanzee consolation is well quantified. De Waal and van Roosmalen (1979) based their conclusions on an analysis of hundreds of postconflict observations, and a replication by de Waal and Aureli (1996) included an even larger sample in which the authors sought to test two relatively simple predictions. If third-party contacts indeed serve to alleviate the distress of conflict participants, these contacts should be directed more at recipients of aggression than at aggressors, and more at recipients of intense rather than mild aggression. Comparing third-party contact rates with baseline levels, the investigators found support for both predictions (figure 3).

  Figure 3 The rate at which third parties contact victims of aggression in chimpanzees, comparing recipients of serious and mild aggression. Especially in the first few minutes after the incident, recipients of serious aggression receive more contacts than baseline. After de Waal and Aureli (1996).