Chimps don't just get mad, they get even

Humans are the most cooperative species on the planet, and the most punitive. This is no coincidence. When I promise to take a friend to the airport to catch an early-morning flight, I have to overcome the temptation to sleep an extra couple of hours. My motivation to follow through on my promise is influenced by my awareness of the consequences: my friend will miss her plane, and I will feel guilty. My friend will probably be angry with me, and she may refuse to help me the next time I ask. Worse yet, she may tell other people what I did, and they may share her outrage and anger. The threat of these kinds of sanctions helps to sustain cooperation within dyads and larger groups (1–4). There is now considerable interest in the evolution of cooperation and punishment in human societies, and there have been a number of efforts to explore the phylogenetic origins of cooperative motives in other primates (4–8). Virtually all of the work on other primates has focused on the willingness to provide benefits to conspecifics. In this issue, Jensen et al. (9) turn the tables and examine chimpanzees' propensity to impose sanctions on familiar group members who commit transgressions.

To study chimpanzees' propensity for punitive behavior, Jensen et al. (9) devised an ingenious experimental protocol in which one chimpanzee was given the opportunity to respond to the loss or inaccessibility of valued food items by pulling a rope that caused a platform to collapse and the food to fall out of reach. This setup allowed the researchers to examine how chimpanzees responded when food was inaccessible or taken away from them, how they responded to disparities in outcomes between themselves and others, and their sensitivity to the role others played in their losses.

In the first experiment, Jensen et al. (9) compared the chimpanzees' responses to variation in the accessibility and desirability of rewards. Sensibly enough, chimpanzees were more likely to pull the rope and dump the items on the floor when they were given access to inedible pieces of plastic and bamboo than when they were given access to tasty items. Similarly, they were more likely to collapse the table when food was inaccessible to them than when it was placed within reach. So far, there is nothing particularly social about the chimpanzees' responses: when they are unable to obtain access to food, they are likely to pull the rope and collapse the table. The key finding from this experiment is that the chimpanzees were as likely to collapse the table when another chimpanzee had access to the food as when no one had access to the food. Thus, chimpanzees do not seem to retaliate against other chimpanzees simply for getting lucky.

In a second experiment, the researchers explored the chimpanzees' sensitivity to the nature of losses that they suffered. In this experiment, food was placed on a sliding platform that could be moved out of the actor's reach. As before, the actor could pull a rope that

Chimpanzees are predisposed to impose sanctions on those that harm them.

caused the platform to collapse and the food to fall. In one condition, a human experimenter moved the platform away from the actor and slid it to within reach of another chimpanzee. In another condition, the experimenter did the same thing, but there was no other chimpanzee present to receive the food. In the last condition, a chimpanzee in the opposite cage was able to pull the platform away from the actor and gain access to the food. Not surprisingly, the chimpanzees were more likely to collapse the table when they lost food than when they were left alone to eat in peace. However, the chimpanzees were significantly more likely to respond punitively when they were victimized by other chimpanzees than when they were the victims of the experimenter's whims.

The Upside of Anger

Jensen et al. (9) speculated that the chimpanzees' behavior might also be influenced by their emotional state. It is easy to tell when a chimp gets mad: they become aroused and perform conspicuous displays and noisy tantrums. The researchers found that the chimpanzees were more likely to collapse the table when they were aroused than when they were calm. Moreover, they were most likely to get upset when they were the victims of theft by other chimpanzees. The authors suggest that the chimpanzees' punitive behavior was motivated by anger. Apparently, chimpanzees do not just get even, they get mad.

The results of these experiments suggest that chimpanzees are predisposed to impose sanctions on those that harm them. Such reprisals are costly because angry chimps incur the physiological costs of arousal and the energetic costs of displays, and run some risk of being injured if their threats escalate into physical confrontations. However, if punitive action deters others from taking their food or stealing their mates, then the direct benefits derived from punitive sanctions may outweigh the costs. This kind of punitive behavior will readily evolve (10). The anticipation of such reprisals may lead to what has been characterized as respect for possession in other species of primates.

It is not clear what role the threat of punitive sanctions plays in stabilizing cooperation in chimpanzees. Chimpanzees collectively patrol the boundaries of their territories, form coalitions, and jointly mate-guard females (11). They also participate in simple forms of exchange. For example, after one chimpanzee is groomed by another, he is more likely to permit his former grooming partner to share his food (12). However, we have no evidence that the failure to reciprocate in these kinds of interactions generates reprisals like those observed by Jensen et al. (9). Of course, this could be because the threat of sanctions is so effective in maintaining cooperation that reprisals are rarely needed. However, it could be that chimpanzees are unwilling to perform costly acts of retaliation in more naturalistic situations.

In this set of experiments, punitive action was cheap and easy. All the chimps had to do was pull a rope, and food fell to the floor. Pulling the rope required little effort, and they did not lose any food themselves. What would they have done if it was more costly to retaliate? Would they have collapsed the table if they lost part of their own food? Would you? Humans have a pretty strong taste for revenge, even when it costs them dearly. In the ultimatum game, which has now been played by thousands of subjects in dozens of societies (13, 14), one player is given a sum of money and can decide how much to keep and how much to allocate to a second player. The second player then decides whether to accept or reject the offer. If the offer is accepted, each player takes home the amount that they were allocated; if the offer is rejected, both players go home with nothing. Rejections provide a way for the second player to retaliate against the first player for unsatisfactory offers. Although the offers are anonymous and the players are strangers, most people around the world reject low offers, even when the stakes are high.

Low offers in the ultimatum game provoke anger because they are perceived to be unfair. People display a strong preference for equitable outcomes (14), and there has been a lively discussion about whether other primates share such preferences (15–17). The results of this experiment add to a growing body of evidence that suggests that other primates are not particularly averse to inequitable outcomes. Here, the chimpanzees were just as likely to collapse the table when the experimenter delivered food to an empty cage as they were when the experimenter delivered food to another individual. They seemed to be motivated by anger, not by envy, with the ultimate aim of deterrence.

Prosocial Punishment

These experiments demonstrate that chimpanzees are willing to retaliate

Low offers in the ultimatum game provoke anger because they are perceived to be unfair.

against those who transgress against them. However, humans take punitive action one step further. We are willing to impose punishment on those who transgress against others, even if we sustain no harm ourselves. For example, people are willing to spend their own money to punish those who make unfair offers in the ultimatum game (18). In real life, whistle-blowers risk their careers to call attention to corporate wrong-doing, even when they are not harmed directly. This kind of “altruistic punishment” is thought to play an important role in the evolution and maintenance of cooperation in large groups (1). Punishment of infractions by third parties is not limited to humans; policing plays an important role in the maintenance of cooperation in the highly specialized cooperative societies of social insects (19). Chimpanzees are clearly willing to retaliate against chimpanzees who steal from them, but would they also police and impose costly punishment on chimpanzees who steal from others?

The fact that we can now ask and devise ways to answer these kinds of questions is really quite remarkable. We used to rely on intuition, schooled by natural history and informed by theory, to divine the motives and preferences that underlie behavior in other animal species. Now, we have ways to test these intuitions systematically in the field and laboratory. Both positive and negative evidence can be illuminating because “it is not enough to review the highlights of succorant behavior, it is equally important to consider the absence of such behavior when it might have been expected” (20). The elegant experiments conducted by Jensen et al. (9) provide compelling proof of the fruitfulness of this approach.