Mind Soc (2010) 9:171–192 DOI 10.1007/s11299-010-0079-9
Has punishment played a role in the evolution of cooperation? A critical review Nicolas Baumard
Received: 26 March 2010 / Accepted: 21 September 2010 / Published online: 1 October 2010 Springer-Verlag 2010
Abstract In the past decade, experiments on altruistic punishment have played a central role in the study of the evolution of cooperation. By showing that people are ready to incur a cost to punish cheaters and that punishment help to stabilise cooperation, these experiments have greatly contributed to the rise of group selection theory. However, despite its experimental robustness, it is not clear whether altruistic punishment really exists. Here, I review the anthropological literature and show that hunter-gatherers rarely punish cheaters. Instead, they avoid dealing with them and switch to other partners. I suggest that these data are better explained by individual selection, and in particular by partner choice models, in which individuals are in competition to be recruited by cooperative partners. I discuss two apparent problems for partner choice theories: large-scale cooperation and punishments in economic games. I suggest that rather than favouring group selection theory, these two phenomena provide evidence in favour of individual selection: (1) people produce large-scale cooperation through institutions in which punishment is not altruistic but rewarded on an individual basis; (2) punishment in experimental games can be explained without altruism and is indeed often better explained by individual interests. Keywords Punishment Group selection Partner choice Hunter-gatherers Institutions Behavioural experiments
N. Baumard Institute of Cognitive and Evolutionary Anthropology, University of Oxford, Oxford, UK N. Baumard (&) Institut Jean-Nicod, CNRS-EHESS-ENS, Paris, France e-mail:
[email protected]
123
172
N. Baumard
1 Introduction In 2002, Fehr and Ga¨chter (2002) published a study on altruistic punishment in humans. They showed that, in an experimental game, participants punish defectors in a cooperative activity although the punishment is costly for them and yields no material gain. They also showed that cooperation flourishes if punishment is possible, and breaks down if it is ruled out. This was not the first time that such a pattern was observed (see Yamagishi 1986 for similar results). However, for the first time, the existence of punishment was linked to evolutionary theories of cooperation. Fehr and Ga¨chter explained that their experimental results were good evidence in favour of group selection (Gintis et al. 2003; Sober and Wilson 1998). First, they demonstrated that people are willing to punish others altruistically—that is, at a cost to themselves and for no benefits. This was taken as evidence that humans are able to sacrifice their individual welfare to increase the welfare of the group, and that cooperation has evolved not for the benefit of individuals but for the benefit of the group. Second, they demonstrated that punishment is a powerful tool to sustain cooperation. At the time when Fehr and Ga¨chter published their results, group selection was considered unlikely since it was easy for defectors to exploit the contribution of altruistic individuals. However, if defectors are punished by altruistic individuals, they can no longer benefit from their defection, and sacrificing oneself for the group becomes evolutionary advantageous.1 This was just what Fehr and Ga¨chter’s experiment suggested. Since then, punishment has become a central theme both among theoreticians (Boyd et al. 2003, 2010; Kurzban and DeScioli 2009; Sripada 2005) and experimentalists (Barclay 2006; de Quervain et al. 2004; Henrich et al. 2006, 2010; Price 2005). However, despite this surge of interest in punishment, the empirical evidence remains weak. Supporters of group selection and students of punishment in behavioural games have assumed that evidence of punishment in the laboratory means evidence of punishment in real life. But is this really the case? In this article, I contend that there is no empirical evidence that altruistic punishment has played a role in the evolution of human cooperation. It is actually a well-established fact in anthropology of law that in small scale societies wrongdoers are not punished by altruistic individuals willing to defend the interests of the group. When someone harms or steals someone else, most of the time nothing is done to punish the culprit. If the wrongdoing is very serious and threatens the safety of the victim, she may retaliate in order to preserve her reputation or deter future aggression. In most cases, however, people simply stop wasting their time interacting with immoral individuals. Thus, rather than supporting the above-cited experimental results, ethnographic data seem to contradict them and to invalidate the theory of the evolution of cooperation by altruistic punishment. I will thus suggest that ethnographic data are better explained by individual selection, and in particular by partner choice theory 1
This theory raises the problem of second-order punishment (some individuals may free-ride the cost of punishment by contributing to the common good but refusing to punish those who do not cooperate). Boyd et al. have proposed a solution (Boyd et al. 2003). We will see in Sect. 2.2 that this solution is empirically implausible.
123
Role of punishment in the evolution of cooperation
173
(Bshary and Noe¨ 2003; Bull and Rice 1991; Noe¨ et al. 1991; Roberts 1998). Individual selection is often reduced to partner fidelity, as typified in the iterated Prisoner’s Dilemma game (Axelrod 1984; Axelrod and Hamilton 1981; Trivers 1971). Individuals who at any time fail to cooperate with their partner can be penalised by those same partners in subsequent trials. These models have been shown to be limited in their ability to explain cooperation outside dyadic relationships (Boyd and Richerson 1988). In the last two decades, a new class of individual selection models, based on partner choice, has emerged. In partnerfidelity models, guarding against cheating is the central issue. By contrast, in partner choice models, the emphasis is less on guarding against cheating and more on competition between potential cooperative partners. Partner control and partner choice models imply different evolutionary dynamics of cooperation. In partner control models, an individual sanctions an uncooperative partner by aborting the relationship, thereby losing the advantages of cooperation in the process. In partner choice models, an individual reacts to an uncooperative partner by starting a new cooperative relationship with another partner. Thus, there may be no cost in sanctioning uncooperative partners. Indeed, switching to a better partner is in the interest of individuals. It is worth emphasising that partner choice models do not require any kind of altruistic punishment. First, switching from one individual to another, presumably more cooperative individual is not altruistic but clearly beneficial to the agent. Second, leaving one’s partner because he is cheating is not a punishment since it does not aim to inflict a cost on him. The cost inflicted on the cheater (through the loss of his partner) is just a by-product of the fact that his previous partner prefers to interact with someone else. To sum up, partner choice works solely on the basis of individual interests. In this theory, cooperative dispositions have evolved not to help the group, but rather to benefit the individuals themselves. The starting point of these theories is that, in the ancestral environment, individuals were in competition to be recruited in cooperative ventures. They had an interest in finding the best cooperative partners and, conversely, they needed to be chosen by the best cooperative partners. This competition is likely to lead to selection for cooperative dispositions. Indeed, individuals who harm and steal from others, cheat and take more than their fair share of the benefit, will be avoided in the future and will progressively loose the opportunity to enjoy the benefits of cooperation. We shall see that this is what we observe in hunter-gatherer societies. As Kaplan and Gurven (2001) put it, cooperation may emerge from the fact that people in hunter-gatherer societies ‘‘vote with [their] feet.’’ (on this point, see Aktipis 2004). In this article, I shall study the ethnographic literature on punishment among hunter-gatherers. I conclude that it is unlikely that punishment has played a role in the evolution of human cooperation. On the contrary, the stability of cooperative behaviours seems to rely on the fact that individuals can switch from one partner to another. In the discussion, I shall argue that there is further evidence that cooperation in groups can be stabilised without punishment. I shall also discuss two apparent problems for partner choice theories: large-scale cooperation and punishments in economic games. I will suggest that rather than casting doubt on
123
174
N. Baumard
individual selection and supporting group selection, these two phenomena provide evidence in favour of individual selection. We shall see that (1) people produce large-scale cooperation through institutions in which punishment is not altruistic but rewarded on an individual basis and that (2) punishment in experimental games can be explained without altruism and is indeed often better explained by individual considerations. To sum up, in an explanation of the evolution of cooperation, the use of punishment seems both uncertain—there is no conclusive evidence in favour of its existence—and unnecessary—cooperation seems to flourish very well in its absence.
2 Altruistic punishment among hunter-gatherers 2.1 Individuals do not punish cheaters, they just avoid them The idea that punishment plays a central role in stateless societies by sustaining cooperation and suppressing cheating is quite old. In a seminal work, Durkheim wrote that ‘‘In primitive societies (…) law is wholly penal (…)’’ (Durkheim 1893, p. 76). He contended that although restorative law has increasingly displaced penal law in modern societies, the latter appeared first. However, Durkheim was writing a century ago, at a time when there were no studies of hunter-gatherer societies. Durkheim had to rely on descriptions of ancient civilisations such as early Greece, Egypt, China and India. He knew that such societies had highly penal systems of law, and he reasoned that if ancient civilisation had so much penal law, earlier and simpler societies would have even more. But he was wrong. These civilisations, albeit quite old, were already based on intensive agriculture and produced huge surpluses that financed penal institutions such as courts, prisons, etc. (Ostrom 1990). They differ radically from the small-scale societies that were soon to be studied by the first ethnographers. In these societies, punishment is marginal, as RadcliffeBrown reports in one of the first studies on a hunter-gatherer society (Andaman, India): ‘‘There was no such thing as the punishment of crime by society. If one person injured another it was left to the injured one to seek vengeance if he wished and if he dared. There were probably always some who would say with the criminal, their attachment to him overcoming their disapproval of his actions. The only painful result of anti-social actions was the loss of the esteem of others. (…) it was in most instances sufficient to prevent such actions.’’ (Radcliffe-Brown 1922) Modern studies of hunter-gatherers confirm Radcliffe-Brown’s statement. ‘‘The Hazda have no procedure for arbitration. They have to solve their disputes for themselves almost always without any intervention by third parties. In fact, most disputes are resolved by self-segregation and attract hardly any attention’’ (Woodburn 1982, p. 252). Similarly, among the Chenchu of Southern India, Furer-Hameindorf notes that:
123
Role of punishment in the evolution of cooperation
175
‘‘There is no institutionalized system of arbitration and no provision for any coercive action in more serious cases such as the abduction of wife. Under these circumstances, it is usual for one of the parties to leave the group, and the rivals will then avoid each others’’. (Fu¨rer-Haimendorf 1967, p. 20) Such societies are stateless—without chiefs or any political authority—and the handling of most, if not all conflict is radically decentralised among the adversaries. They do not even have judges or arbitrators who intervene authoritatively in the conflict of their fellows (Black 2000). The only such third party among the Mbuti (Pygmee) of Zaire is the so-called ‘camp-clown,’ who dampens conflict by drawing attention to himself and making everyone laugh (Turnbull 1965, pp. 182–183). In other societies, onlookers may simply beg the parties to stop, step between them, or pull them apart when they quarrel or threaten to fight (Knauft 1987; Lee 1979). But most adversaries handle their own problems with no help from anyone. Although hunter-gatherers sometimes use violence against wrongdoers, it is typically inflicted by the aggrieved party. Among Inuits for instance, a male relative to the victim traditionally takes vengeance against a killer (Balikci and American Museum of Natural History 1970; Hoebel 1954). As we have seen in Sect. 2, these behaviours can be easily explained without punishment. The aggrieved party seeks only to defend the reputation of his kin and prevent future aggressions. There is thus no evidence that punishment plays a central role in societies without penal institutions. On the contrary, evidence favours partner selection, as cheating results in a switch to other partners, thereby creating a market for cooperative partners. Woodburn notes: Units are highly unstable, with individuals constantly joining and breaking away, and it is so easy to move away that one of the parties to the dispute is likely to decide to so very soon, often without acknowledging that the dispute exists (Woodburn 1982, p. 252). Although both parties lose their partners, in the long run, this partner-switch process is costly to cheaters: The Adamanese cannot afford to arouse permanent resentment or contempt of his own local group, for this group is the land owning unit, within the territory of which he is entitled to hunt and collect. He can join another land-owning group only if its members are willing to accept him, and a notoriously troublesome and lawless man is unlikely to be welcome to another group. Furer-Hameindorf makes the same statement: ‘‘Spatial mobility and the ‘settling of disputes by avoidance’’ allows a man to escape from social situations made intolerable by his egoistic or aggressive behaviour, but the number of times he can resort to such a way out is strictly limited. There are usually two or three alternative groups he may join, and a man notorious for anti-social behaviour or a difficult temperament may find no group willing to accept him for any length of time. Unlike the member of an advanced society, a Chenchu cannot have casual and superficial relations with a large number of persons, who may be somewhat indifferent to his conduct
123
176
N. Baumard
in situations other than a particular and limited form of interaction. He has either to be admitted into the web of extremely close and multi-sided relations of a small local group or be virtually excluded from any social interaction. Hence the sanctions of public opinion and the resultant approval or disapproval are normally sufficient to coerce individuals into conformity and only strong emotions, such as the passionate attachment of lovers living in an incestuous union, will induce Chenchus to run the risk of social exclusion.’’ These observations validate partner choice models: individuals do not punish wrongdoers, they rather avoid them. The above-cited studies are focused on highly immoral behaviours (murder, theft, incest, etc.). But the same observations can be made for other kinds of uncooperative behaviours. In a review of the literature on the food exchanges of hunter-gatherers, Gurven (2004) shows that people choose their partners on the basis of their willingness to share, and that they may exclude some partners because their level of cooperation is too low: ‘‘For example, one Pilaga family temporarily left the village in response to giving twice as frequently as it was receiving food from another family, consistent with their common complaint, ‘‘I have given something to him but he has not given to me’’ (Henry 1951, p. 199). Although Mamainde food distributions appear egalitarian (quantities given to each family is inversely proportional to number of families present), Aspelin (1979) notes several cases where one unproductive family with a precarious position in the village was frequently excluded from receiving shares. Altman (1987, p. 147) describes a collusion between two Gunwinggu family clusters to share less food with a third cluster who was ‘‘not producing enough.’’ This sanction induced higher production and sharing by the third cluster, wherein the other two family clusters resumed normal relations. Among the Washo, a ‘‘person who would not share with others of the same household, or who was generally stingy would not be included in the networks of sharing and would be ‘talked out’ of his household’’ (Price 1975, p. 16). Baksh and Johnson (1990) relate a similar anecdote in which a household that ‘‘did not like to work cooperatively, or participate in communal undertakings’’ was driven out of the village. An unproductive family ‘‘quickly gets pressure to contribute its own share’’ among the Agta, where social ostracism ultimately forces them to relocate’’ (Griffin 1982, p. 20). Similarly, in a review of political anthropology, Boehm (1999) shows that people leave their leaders when they find their behaviours immoral and in contradiction with their own interests. Losing one’s partner is the ultimate threat among hunter-gathers. Most of the time, however, people do not need to go that far—simply to signal one’s disagreement toward uncooperative members is enough to motivate them to cooperate more. Indeed, as Boehm (1999) notes: ‘‘If a forager is seriously ostracized (in the sense that no one will speak to him, literally, or cooperate with him economically), his very means of livelihood may be threatened’’ (p. 59). This
123
Role of punishment in the evolution of cooperation
177
explains why ethnographic reports suggest fairly smooth interactions and also why criticism, ridicule and public opinion are so powerful (Boehm 1999; Cashdan 1980; Turnbull 1965). People have to take into account others’ opinions if they do not want to be left alone to fend for themselves one day. Briggs’s report of a socially inept woman named Niqi among the Inuit is a good illustration of this point (Briggs 1970 here summarised by Boehm): Her [Niqi’s] emotionally unruly family was kept at distance by the rest of the band even though the members were not wholly excluded from group social life and sharing of food. Niqi herself was known for regularly becoming irritated in ways that adult Inuit normally manage to mask; in compensation she smiled much too broadly (Briggs 1970: 218). One problem, then, was that her personal style was too volatile for a people who were averse to exhibitions of strong emotion, and of anger in particular. In addition, Niqi’s small family tended not to share food as much as other families in the band, and stayed socially separate in a number of ways. Band members were resentful of her family as would-be free-riders who were deemed stingy and jealous, and also as people who did not keep their emotions under control. Briggs (1970: 214–223) details the quite grumbling and factional hostility that developed out of these tensions, and notes that very direct hints were made that Niqi should participate in the camp women’s tasks, such as cooking fish heads for everybody. At length, these well-masked but hostile signals drew Niqi away from her own family’s subsistence activities, if only briefly, to make a contribution to group’s subsistence. But she quickly went back to her former ways, and the social distancing of her family continued. The tensions rose and fell, and Niqi’s social status vacillated between grudging acceptance and semiostracism. The rest of the group distanced Niqi’s family mainly because of Niqi’s bad temler, stinginess, and failure to help, and the family tended to camp away from the core group. (p. 56) This example is a good illustration of the fact that disapproval and criticism precede exclusion. Even the socially inept Niqi understood that, although in the end she was unable to reform herself and was excluded by others who preferred not to interact with her. This example also shows that partner choice need not be an all or nothing issue. Individuals can cooperate to various degrees, and adjust their level of cooperation based on their trust in their partners. Finally, the only clear cases in which individuals inflict a cost on a wrongdoer seem to be cases of self-defence, when the victim takes revenge by attacking the wrongdoer, and when the group collectively punishes a very dangerous individual that has become a public enemy. However, these kinds of behaviour cannot be considered as altruistic. In the case of revenge, the victim has a clear interest in defending herself in order to win a reputation as someone who does not tolerate aggression (Hoebel 1954). This behaviour is not specific to humans and can be observed in many species (Clutton-Brock and Parker 1995). It is the old instinct of possession and self-defence. Here, ‘retaliation’ would be a better term than
123
178
N. Baumard
‘punishment’, since the victim’s behaviour has often nothing to do with cooperation (for instance, it does not aim to deter cheating in the group), aiming only to defend the victim’s reputation and interests. The killing of public enemies can be explained in a similar way. Indeed, it is clearly in the direct interests of the people who carry out the murder, since the dangerous individuals threaten everyone’s safety. Moreover, the assassination is usually organised, everyone assaulting the individuals at the same time or everyone (including the family of the individual) agreeing not to take revenge on the killers. Since murders are mainly prevented by the threat of revenge by kin, killing a public enemy in these conditions is not an altruistic act and is not very costly (and indeed, the killer is sometimes an old man or a young boy) (Boehm 1999; Clastres 1972; Hoebel 1954; Lee 1979; Rouland 1976).2 Although these defensive behaviours may have played a role against the evolution of aggressive behaviours, it is not clear that they have an impact on everyday cooperation. To conclude, studies of hunter-gatherers present at once evidence against the group selection account of punishment (individuals do not punish altruistically and cooperation is not enforced by punishment) and evidence in favour of the individual selection (partner choice is sufficient to create an incentive to cooperate with others). Few studies, however, have systematically examined punishment in a hunter-gatherer society. In the following section, I will review in detail the only study that does so (Wiessner 2005). This is all the more important given that this study is often quoted as showing that hunter-gatherers punish others, and is therefore used as evidence in favour of group selection (e.g. Bernhard et al. 2006; Gintis 2009). On the contrary, we shall see that this study can count as evidence against the group selection account of punishment. 2.2 A case study: punishment among the Ju/’hoansi Polly Wiessner analysed data from 308 conversations among the Ju/’hoansi (!Kung) Bushmen in the Kalahari desert. Wiessner finds the same pattern as previous ethnographers: people do not punish free-riders. First, because individuals are supposed to defend themselves. ‘‘all people as autonomous individuals are expected to stand up for their rights in dyadic relationships’’. (p. 135). But mostly because uncooperative actions are not punished. Cheaters are just avoided. ‘‘minimal energy was wasted on incorrigible free-riders; unless they were disruptive to community, they were accepted for whatever they and their families did contribute.’’ (p. 135) ‘‘In view of arguments that punishment by coalitions is largely a means to curb free-riders (Price et al. 2002), it is surprising what a small percent of 2
Among the Inuits, as a double safeguard against blood revenge on the executioner, close kinsmen may themselves be called upon to carry out the community will (Hoebel, 1954).
123
Role of punishment in the evolution of cooperation
179
punishment (14% of all cases) was aimed at able-bodied free-riders who regularly failed to produce enough to share widely with the community or tried to consume more than their share on a daily basis. Only 5 of 32 cases involved criticism for low work effort even though there were capable individuals in all camps who produced far less than others. People did not bother with those who were regularly lazy, stingy, or greedy. Rather, the penalty for low production or stinginess was not verbal punishment but low social regard, fewer marriage opportunities, or fewer exchange partnerships (Wiessner 2002b).’’ (p. 124) Like other hunter-gatherers, the Ju/’hoansi do not seem to punish others. Instead, they simply avoid cheaters and lazy individuals. Wiessner analysed a variety of behaviours that could be described as ‘punishments’. These punishments were divided into four categories: (a) put-downs through pantomime, joking, or mocking, (b) mild criticism and complaint, (c) harsh criticism or complaint, and (d) criticism plus violent acts. Of the 171 cases of punishment recorded, 22% involved mocking, joking, or pantomime; 41%, outright complaint or criticism; 35%, harsh criticism; and 2%, harsh criticism and violence. We will see that the analysis of these punishments confirms the idea that cooperation is not enforced by punishment but rather by partner choice. First, joking, mocking and criticism (categories a, b and c) seem to mainly serve to communicate people’s disagreement to others and to warn them about the future of their relationships. The second and third categories of punishment involved outright criticism, usually when the target person was present or within earshot. In the evenings it was not unusual to hear one person register a loud complaint that shattered the still night air, reaching all hearths. Most complaints represent a rupture in relations. Rejection was expressed in the response of the transgressor: ‘‘Jusi wesi !xau mi’’ ‘‘Everybody rejects (or refuses) me.’’ People who delivered harsh complaints were unlikely to share or give other forms of assistance to the target until their anger was quelled. (p. 122) In line with the idea that people express their disagreement to their partner, most criticisms rarely go further: The most common outcome of Ju/’hoansi verbal punishment was that the message was heard by the target without visible response or apology. (p. 132) It worth noting that people first use joking, mocking and pantomime to express their concern. They communicate their opinion indirectly so that the person who is blamed can save her face. Contrary to a group selection account of punishment, people do not want to harm others or to be too aggressive. They just want to convey a message. Second, when criticisms go further, the typical punishment does not consist in harming the criminal. Instead, it consists in stopping interactions with him. In other words, it means switching to another partner, which is precisely what partner choice theory predicts. Wiessner gives six examples of serious punishment. They all take
123
180
N. Baumard
the form of partner switching. Since we have few examples of major ‘punishments’, it is worth examining some of them in detail. In 1974 the two most important instances of coalition building were directed against strong individuals who exhibited reclusive behavior by going off to summer territories, withdrawing from broader circles of cooperation, and indicating that visitors were not welcome. Genealogies were recited repeatedly to establish equal or greater rights to the summer territories and to rally co-landholders. (…) Tension remained high when both groups returned to/Kae/kae for the dry season and persisted for several years thereafter. Sharing declined, but no further sanctions were exerted. (p. 132) Another major case of coalition building in 1974 involved ostracism of a woman who had several passing relationships with Bantu men. Ju/’hoan men are threatened by women who consort with Bantu men because external unions deplete their pool of possible marriage partners. The woman in question left the area after harsh and repeated criticism and subsequently died. (p. 133) Another case of ostracism that had occurred in the previous year was discussed in 1974. This involved one of the best hunters in the area, who was believed to have committed incest. Proof of the incest was further confirmed in the eyes of the Ju/’hoansi when his daughter died rather suddenly from a lump that appeared on her back. After extensive slander, the offender went into voluntary exile in the nearby hills with his family. Close kin continued to visit and reside with him. By 1996 he was settled at/Kae/kae, where he had become a respected member of the community. (p. 133) The second case involved the expulsion of a group of families who had recently moved to Xamsa village. At first things went very well, but when one member of the group killed a large animal and gave meat to only one person in the host village, negative feelings mounted. One incident led to the next and a coalition aimed at extraditing the temporary residents was formed. After several months of tension, the group moved on, established their own village some 9 km away, and resumed good relations. Such fission and fusion in Ju/’hoan camps is not uncommon. The third case involved the expulsion of a family from Xamsa in which the wife was drunken and promiscuous and the children unruly. People felt threatened by the wife’s sexual behavior with Bantu, her bad parenting, and her wild kids. The village formed a coalition against the family, jointly refused to share with them, and even asked me not to give them anything. (p. 133) In all these cases, the so-called ‘punishment’ could be better described as cessation of a relationship or partner switching. Although the last category—violent punishments (d)—may suggest the existence of altruistic punishments. However, in all the cases, the victim was involved in the action. The involvement of the victim suggests that reputation management and
123
Role of punishment in the evolution of cooperation
181
self-defence may be responsible for these behaviours. This is consistent with Wiessner’s statement that people are expected to stand up for their rights in dyadic relationships. In other words, the last category of the so-called ‘punishment’ may be better described as retaliation. Wiessner finally turn to the cost of punishment. It appears to be relatively low. Of the 124 cases of punishment considered, 11 (8%) had some negative repercussions for the initiators and coalition members. And again, in line with the partner choice, these costs take the form of the cessation of relationships. The three cases with the most severe consequences, severed relations between groups, involved land issues detailed above in which the stakes were high for all members of the punishing coalition, particularly in one case where the transgressors included members who were known to be hotheaded and to have killed in the past. (p. 135) In these cases, it is not clear to what extent the cessation of relationship is costly. Indeed, in the long term, it may be better to leave an uncooperative partner for the prospect of finding a better one. Two further observations run against the idea that cooperation among huntergatherers is enforced by punishment. First, group selection theorists have proposed the existence of second-order punishment—that is, punishment of those who do not punish (those who free ride on punishment) (Boyd and Richerson 2002; Fehr and Fischbacher 2003; Henrich and Boyd 2001; Sober and Wilson 1998). This hypothesis was not confirmed here; not a single case of second-order punishment occurred during the conversations Wiessner reports in her analysis. Second, whereas ability to mediate is highly valued among the Ju/’hoansi, as it is in most societies, willingness to punish was regarded with mixed feelings. On the one hand, those who punished frequently or harshly were considered to be tchi nlai (angry, sharp, or biting thing) and were frequently told to be quiet and stop causing trouble; on the other, camp members were often relieved to have someone else bring tense issues into the open. Twenty-seven percent of high-status individuals did punish very frequently, but there is little to indicate that punishing is an important source of status in comparison with hunting, healing, or xaro exchange (Wiessner 2002b). Consistent with the ideal that leaders should be mediators, not punishers, nlore kxaosi, the most respected figures in camps and recognized senior owners of the land, refrained from joining punishing coalitions in 28 out of 45 (62%) conversations in which they were present. (p. 138) In other words, the most successful individuals refrain from being too aggressive. They choose the best partners and do not waste their energy on cheaters. Wiessner’s therefore study confirms previous ethnographic reports and illustrates the fact that punishment plays a marginal role in the maintenance of cooperation, while partner choice is the leading force in the evolution of cooperation.
123
182
N. Baumard
3 Discussion 3.1 Partner choice: an alternative way to promote cooperation Throughout this review of the anthropological literature, we have seen that the existence of altruistic punishment is quite uncertain, and that cooperation can be sustained solely on the basis of individuals’ interest in cooperating with the best partners. The possibility of partner choice is confirmed by the evidence we have that hunter-gatherers constantly exchange information about others’ behaviour (Cashdan 1980; Wiessner 2005), and that they accurately distinguish good cooperators from bad cooperators (Gurven 2004; Tooby et al. 2006). People talk about theft, stinginess and failure to cooperate, but also about generosity, cooperativeness and honesty. As Boehm writes: ‘‘In effect, the band keeps a dossier on every individual, noting positive and negative points.’’ (p. 73) These observations are corroborated by a range of experiments which demonstrate that people tend to select the most cooperative individuals and that the least cooperative individuals are gradually left out of cooperative exchanges (Barclay 2004, 2006; Barclay and Willer 2007; Chiang 2010; Coricelli et al. 2004; Ehrhart and Keser 1999; Hardy and Van Vugt 2006; Page et al. 2005; Sheldon et al. 2000; Sylwester and Roberts 2010). Further studies show that people are perfectly able to detect the cooperative tendencies of their partners. They rely on cues such as their partners’ apparent intentions (Brosig 2002; Frank et al. 1993), the costs of their actions (Ohtsubo and Watanabe 2008), or the spontaneity of their behaviour (Verplaetse et al. 2007). They also actively seek these types of information and are willing to incur costs to get it (Kurzban and DeScioli 2008). These results are strengthened by recent experimental games who suggest that when participants have more options (for instance, when they can retaliate against those who have punished them), punishment and second-order punishment do not necessarily lead to more cooperation because those who have been punished retaliate against punishers (Cinyabuguma et al. 2004; Denant-Boemont et al. 2007; Dreber et al. 2008; Herrmann et al. 2008; Nikiforakis 2008; Rand et al. 2009). One recent experiment nicely illustrates the validity of partner-choice mechanism (Pradel et al. 2008). One hundred twenty-two students of six secondary school classes played an anonymous dictator game, which functioned as a measure of cooperation. Afterwards and unannounced, the students had to estimate their classmates’ decisions and did so better than chance. Sociometry revealed that the accuracy of predictions depended on social closeness. Friends and disliked classmates were judged more accurately than liked classmates or those met with indifference. Moreover, more cooperative participants were friends with more moral persons than were less moral person. The results confirm the existence of the two prerequisites for the evolution of cooperation through partner choice: the predictability of moral behaviour and the association of more cooperative individuals. In the competition of cooperative partners, the most cooperative individuals end up interacting with each other and enjoy bigger common good. This conclusion is not only valid for hunter-gatherer societies. Indeed, throughout much of human history, there was no penal system whatsoever, and cooperation was
123
Role of punishment in the evolution of cooperation
183
sustained only by partner-choice and self-defence (Hoebel 1954). In tribal societies like the Nuer for instance, ‘‘self-help, with some backing of public opinion, is the main sanction’’ (Evans-Pritchard 1940). Even in large-scale societies in which there is a police force and a penal system, many cooperative activities still work without any punishment. For instance, the study on European and Maghribi medieval traders (Greif 1993), Jewish New York jewellers (Bernstein 1992), and Chinese middlemen in South Asia (Landa 1981), has shown that highly valuable goods and services can be negotiated and exchanged peacefully, without any punishment. All exchanges are based on trust, and anyone who cheats once will lose everyone’s trust. Nobody will take the risk of exchanging with him, and he will have to leave the business. Similarly, on a smaller scale, many economic associations rely on partner choice to sustain cooperation. For instance, in a review on rotating credit associations,3 Ardener (1964) noted that ‘‘the member who defaults in one association (…) may not be accepted as a member of any other.’’ In some communities rotating credit institutions have become so rooted in the economic and social system that exclusion would be a serious form of deprivation. As citizens of modern states, we are used to thinking that crimes are only prevented by the penal system. However, the penal system only monitors major wrongdoings (violence, harassment, theft) and most of our daily interactions with our partner, our friends or our colleagues work smoothly without punishments. This process has been extensively studied by sociologists seeking to understand why some groups (firms, churches, states, clans, etc.) show more cooperation than others (see Hechter 1987 for a general theory; see Hechter and Kanazawa 1993 for a comparison of japanese and american societies). Although their analysis was based on rational choice models, these works are perfectly compatible with evolutionary theories. Indeed, they show that cooperation depends on the effectiveness of the monitoring of group members’ cooperation (through public production, public sharing, collective rituals, etc.) and on group members’ dependence on the benefits offered by the group: The force of community sanctions ultimately flows from the dependence of members. The most serious sanction of all in intentional community –as it is in tribal society- is expulsion but the severity of this punishment depends entirely on the values of alternatives outside group boundaries. Most successful intentional communities do attempt to limit their members’ group affiliations. Often members are required to surrender their goods on entry (Kanter 1972: 82), and once having thus ‘‘invested’’ in membership, exiting group is more costly. Not all investment need to material. Members invest their time and energy in the intentional community. They invest in a certain kind of on-thejob training: mastering the community jargon, learning the rules, abiding by status distinctions. Rarely do the investments have value outside the group. (p. 161)
3
These associations are widespread in countries where access to credit is difficult. They are made up of individuals who agree to make regular contribution to a fund that is given, in whole or in part, to each contributor in rotation.
123
184
N. Baumard
The importance of dependence explains why cooperation is more successful in ethnic minorities (Landa 1981 #141; Hechter 1987 #175): since members of minority face high costs in leaving their group, they are extremely dependent and cannot afford being rejected by the members of their group. 3.2 How do we explain large-scale cooperation? Although partner choice can be an efficient way for cooperation to emerge, it cannot explain the whole of human cooperation. It may explain why individuals in smallscale societies manage to trust each other and to cooperate in the long term, but it does not explain how larger societies produce public goods such as infrastructures, schools, or social security. In an influential work on common pool resource institutions (irrigation systems, fisheries, pastures, etc.), Ostrom (1990) concludes that common pool resources cannot be managed relying only on the motivations deployed in the ancestral environment. Even in repeated settings where reputation is important and where individuals share the norms of keeping agreements, reputation and shared norms are insufficient by themselves to produce stable cooperative behaviour over the long run. If they had been sufficient, appropriators could have avoided investing resources in monitoring and sanctioning activities. In all of the longenduring cases, however, active investments in monitoring and sanctioning activities are quite apparent. (p. 93) Producing public goods and enforcing cooperation on a large scale requires more than reputation and a willingness to cooperate (see also Hardin 1968; Hechter 1987; Olson 1965). At first sight, the existence of large-scale cooperation looks like a problem for individual selection theories, and although partner-choice models may be able to explain cooperation outside dyadic interactions (contrary to standard partner fidelity models), group selection may be a better theory to explain large-scale cooperation. This conclusion is all the more tempting given that large-scale cooperation usually relies on punishments and sanctions. And indeed, supporters of group selection often use it to highlight the limits of individual selection and the need for altruistic punishment (Gintis et al. 2005). However, we shall that far from confirming group selection, the analysis of large-scale cooperation confirms the premise of individual selection. Indeed, in large-scale cooperative institutions, punishment is organised in such a way that members of institutions have a personal interest in punishing. Thus, although large-scale institutions do rely on punishments, institutional punishments do not require any altruism from the institutions’ members. Why is this so? First, monitoring others is often costless, as Ostrom notes: The cost of monitoring is low in most CPRs [Common Pool Resources] as a result of the rule in use. Irrigation rotating systems, for instance, usually place the two actors most concerned with cheating in direct contact with one another. The irrigator who nears the end of a rotation turn would like to extend the time of his turn (and thus the amount of water obtained). The next irrigator
123
Role of punishment in the evolution of cooperation
185
in the rotation system waits nearby for him, and would even like him to start early. The presence of the first irrigator deters the second from an early start, the presence of the second irrigators deters the first from a late ending. Neither has to invest in additional resources monitoring activities. Monitoring is a byproduct of their own strong motivations to use their water turns to the fullest extent. (p. 95) Second, sanctioning is profitable: Private benefits are allocated to those who monitor. When internal monitoring is accomplished as part of a specialized position accountable to the other appropriators, several mechanisms increase the rewards for doing a good job or exposing slackards to the risk of losing their position. In the Spanish huertas, a portion of the fines is kept by the guards; the Japanese detective also keep the sake they collect from infractors. All of the formal guards positions are accountable to the appropriators, and thus the monitors can be fired easily if discovered slaking off. Because the appropriators keep monitoring the guards, as well as each others, some redundancy is built into the monitoring and sanctioning system. (p. 96) Thus, far from being a confirmation of group selection and of the role of altruistic punishment in the evolution of cooperation, the analysis of large-scale cooperation contradicts group selection (see also Hechter 1987 for a similar analysis). If people were altruistic punishers, institutions would not need incentives to motivate people to punish. But this is exactly what we observe: people punish others because they have an incentive to do so! This point is even clearer in modern states in which policemen, judges, detectives, etc., all receive wages paid by citizens via their taxes. Note that this does not mean that individuals are self-interested maximisers as in classic rational choice analysis. Although most of the analyses of common goods have been conducted in the framework of rational choice, they are fully compatible with an evolutionary approach according to which humans are moral animals. Indeed, these studies have often emphasised the willingness of members of groups holding common pool resources to comply with social norms and to find fair agreements on the distribution of benefits. What these studies point out is that the moral motivation is not enough to sustain large-scale institutions. For instance, some people may find the institutional arrangements unfair and think that they have a legitimate right to cheat and take their fair share of the benefits. Others may notice this cheating and cheat in their turn. Sooner or later the institution will collapse. This is why institutions need more than just good will. They need stable membership, monitoring incentives, graduated fines, exclusion of outsiders, and conflict resolution. It is worth emphasising this point, since group selection supporters often use the limits of rational choice analyses to claim that individuals are ‘strong reciprocators’. The evidence that individuals are not totally selfish does not mean that their cooperative disposition has evolved through group selection. It could also have evolved by individual selection, notably through partner choice. Thus, evidence of cooperative dispositions is not evidence of group selection. Empirical studies of
123
186
N. Baumard
institutions are therefore compatible with any kind of evolutionary theory of cooperation, and indeed, as we have seen in this section, their results seem to fit better with individual selection than with group selection.4 3.3 How do we explain punishment in economic games? If altruistic punishment is not necessary for cooperation to evolve, why then do participants punish others in experimental games? Isn’t Fehr and Ga¨chter’s experiment a counterargument against the theory of the evolution of cooperation by partner choice? Not necessarily. There are indeed a variety of alternative explanations based on individual selection. First, consider second-party punishment: this behaviour can easily be explained in terms of partner choice. In the ultimatum game, when the responder refuses to take an unfair share of the pie (and then loses his money), his behaviour can be interpreted as a cessation of relationship with the proposer (although he incurs a cost in the experimental game, in the ancestral environment, this behaviour would mean switching to a better partner). Participants do not punish (they do not aim to sustain cooperation), they just defend their interests by refusing to be cheated (on second party punishment in the perspective of individual selection, see also Kurzban and DeScioli 2009) .5 Now consider third-party punishment. First, note that third-party punishments are smaller than second-party punishments (Fehr and Fischbacher 2004). Second, thirdparty punishment (unlike second party punishment; see Bolton and Zwick 1995) diminishes when such acts are carefully kept anonymous (Kurzban et al. 2006). Third, people may punish out of simple spitefulness. In public good games, for instance, some participants punish both high and low contributors in order to increase their own relative pay-off (Cinyabuguma et al. 2004; Falk et al. 2005; Saijo and Nakamura 1995). Fourth, people may punish for strategic reasons, thinking that, in a repeated PGG with only four participants, it is advantageous to threaten low cooperators (but see Falk et al. 2005). Fifth, participants’ behaviour in experimental games can be explained by methodological factors. Participants receive some 4
Group selection supporters also claim that group selection is necessary to explain the emergence of large-scale institutions (and not only their stabilization through punishment). Good institutions would flourish and out-compete bad institutions. However, common pool resources are just the cases that contradict this point. Indeed, as Ostrom shows, each institution with its rules, its symbols, its participants, is precisely adapted to its environment. It cannot travel easily and replace another institution. Moreover, empirical studies clearly show that institutions do not evolve by mutation and selection but rather by adjustment and conflict resolution. When appropriators design at least some of their own rules (design principle 3), they can learn from experience to craft enforceable rather than unenforceable rules. This means paying attention to the costs of monitoring and enforcing, as well as the benefits that accrue to those who monitor and enforce the rules. (p. 96).
5
The same analysis can be used for the Prisoners’ Dilemma (Dreber et al. 2008; Wu et al. 2009). In each round, participants chose between cooperating, defecting, or punishing the other participant. Here punishment is not altruistic, since individuals are directly involved in the interactions and may benefit considerably from threatening their partners. By punishing, they show their partner that they are not going to allow themselves to be exploited in the following rounds. Again, this behaviour may have been adaptive in the ancestral environment, where it was not uncommon to interact again and again with the same individuals.
123
Role of punishment in the evolution of cooperation
187
money and have the chance to use it to punish. They may just punish to use their money and not go back home without having done anything in the lab. The other possibility is that experimental games may contain demand characteristics. Participants may figure out that the experiment is about punishment and want to please the experimenter by punishing. So far, I have invoked only selfish motives to explain third-party punishment. However, in partner choice theories, people also have moral motives to punish others. To understand why, we need to study the dynamics of the competition for cooperative partners in more detail. In the ancestral environment, individuals were in competition to be recruited for the most fruitful ventures, and it was vital to share the benefits of cooperation in a mutually advantageous manner. If individuals took a bigger share of the benefits, their partners would leave them for better partners. If they took a smaller share, they would be exploited by their partners, who would receive more than what they had contributed to produce. This competition to attract cooperative partners is thus likely to have led to selection for a ‘sense of fairness’, a cognitive device that binds individuals to respect others’ interests (goods, services, investments, etc.) in cooperative interactions (Baumard 2008, 2010a; Baumard et al. 2010). Human society is a vast cooperative venture in which individuals try to share the cost and benefits of cooperation in a fair way. This is why they think that they ought to contribute to cooperative actions if they receive some benefits, and also why the distribution of benefits is often proportional to contributions to the common good. This is the best way to share the costs and benefits in a fair manner. This is also why individuals have duties to help others in case of difficulty, as well as why these duties are limited because mutual aid has to be equally shared. Individuals cannot ask more than what is compatible with others’ interests, and do not give more than what is compatible with their own interests. This instinct of fairness may explain why people may want to punish wrongdoers. Indeed, in the same way that people give resources to others who are entitled to them (e.g. because these others have contributed to the production of the resources, or because they had previously given some resources), people should take some resources from others who are not entitled to them (e.g. because they had stolen them or taken more than what they had contributed to produce). This is the reason why people vote and protest to punish financial firms guilty of misbehaviours or why they give money to NGO such as Amnesty International or Oxfam. This is also the reason why people support just revenge in state less society. That is the best they can do in a society where there is no institutions to easily punish criminals. In line with this theory, economic games suggest that punishment in behavioural games is partly motivated by fairness and not by the desire to help the group. For instance, participants are ready to pay a cost to reduce high earners’ incomes (and increase low earners’ incomes) even when there is no contribution phase and thus no cheaters to punish. They just want to implement the fairest distribution of resources (Dawes et al. 2007; Johnson et al. 2009). Participants are also willing to punish participants who contribute and punish too much because, arguably, these participants require too much from others and are thus unfair with low contributors (Baumard 2010b; Herrmann et al. 2008). Finally, participants punish unfair
123
188
N. Baumard
distributions in the dictator game even when these distributions are better from the group’s point of view (Leibbrandt and Lo´pez-Pe´rez 2008). Although this duty to restore fairness may explain why people punish wrongdoers, it is quite limited and may not play an important role in the evolution of cooperation. Indeed, we not have to sacrifice ourselves to protect the rights of others, just as our duty to rescue others or to prevent crimes is limited (we are not obliged to risk our lives to save strangers or to be killed to prevent a robbery—this would be great, even heroic, but surpasses the requirements of duty). Most people consider that it is good but beyond duty to risk one’s life to support political dissidents in foreign countries for example. Since punishing others is often very costly (except in economic games where it costs only a couple of euros and where there is no threat of retaliation), we should not observe many of the costly punishments examined by evolutionary scientists (e.g. physically attacking the wrongdoer). Indeed, they may generally be beyond the call of duty. The only cases where costly punishments are expected to be common are when punishment coincides with retaliation (i.e. the victim is also the punisher and has a direct interest in punishing) or when some institution has an interest in punishment and the power to carry its cost (e.g. the state has an interest in preventing crime and can pay policemen and judge to implement its policy). As we can see, altruistic punishment is not the only explanation for third-party punishment. There are multiple motives which might explain why participants punish in experimental games: reputation management, spitefulness, manipulation, fairness, desire to play with the money or to conform to the experimenter’s perceived objective. Insofar as humans do not seem to punish a lot in everyday life, it seems more parsimonious to consider that punishment in experiment games is not clear evidence in favour of altruistic punishment.
4 Conclusion Altruistic punishment has been at the centre of evolutionary theories of cooperation. However, in the light of the anthropological literature, it seems unlikely that it has played a role in the evolution of human cooperation. Anthropological data fit better with the theory according to which cooperation has evolved by partner choice, through competition to find the best cooperative partners. This theory is compatible with the existence of large-scale cooperation, which has been shown to be sustained through institutional rewards and punishment, and with the existence of punishment in experimental settings, which may be the product of many different individual motivations. Of course, this demonstration does not totally rule out the possibility that cooperation has evolved through altruistic punishment. It could be the case that anthropological archives are too uncertain or that our analyses of institutions or experiments are too rudimentary. Is there thus another way to decide between altruistic punishment and partner choice? I think so, as the two theories make different predictions regarding punishment. Group selection theory holds that punishment aims to promote the welfare of the group (Boyd et al. 2003; Fehr and
123
Role of punishment in the evolution of cooperation
189
Ga¨chter 2002; Henrich and Boyd 2001). This implies that punishment should be calibrated to help the group and deter crime. Here, group selection parallels the utilitarian doctrine of punishment, which contends that punishment should be used to deter crimes and maximise the welfare of society (Polinsky and Shavell 2000; Posner 1983). The utilitarian theory of punishment holds, for instance, that the detection rate of a given crime and the publicity associated with a given conviction are relevant factors in assigning punishments. If a crime is difficult to detect, the punishment for that crime ought to be made more severe in order to counterbalance the temptation created by the low risk of getting caught. Likewise, if a conviction is likely to get a lot of publicity, a law enforcement system interested in deterrence should take advantage of this circumstance by ‘‘making an example’’ of the convict with a particularly severe punishment, thus getting a maximum of deterrence for its punishment. By contrast, the theory of fairness developed in Sect. 3.3 predicts a restorative logic. A crime creates an unfair relationship between the criminal and her victim, and people have the intuition that something should be done to restore the balance of interests—either by harming the criminal or by compensating the victim. In intuitive terms, someone is punished because she ‘‘deserves’’ to be punished. I have argued elsewhere (Baumard 2010b) that empirical data are more consistent with the theory of fairness. Indeed, experimental studies, relying on a variety of methodologies, suggest that when people punish harmdoers, they generally respond to factors relevant to a retributive theory of punishment (magnitude of harm, moral intentions) and ignore factors relevant to the group selection theory (likelihood of detection, publicity, likelihood of repeat offending) (Baron et al. 1993; Baron and Ritov 2008; Carlsmith et al. 2002; Darley et al. 2000; Glaeser and Sacerdote 2000; McFatter 1982; Roberts and Gebotys 1989; Sunstein et al. 2000). In line with these results, field observations have extensively demonstrated that, in keeping with this prediction, the level of compensation in stateless societies is directly proportional to the harm done to the victim: for example, the wrongdoer owes more to the victim if he has killed a family member or eloped with a wife than if he has stolen animals or destroyed crops (Hoebel 1954; Malinowski 1926/1985). To conclude, punishment does not seem to be a group adaptation. It follows the logic of fairness rather than the interests of the group.
References Aktipis C (2004) Know when to walk away: contingent movement and the evolution of cooperation. J Theor Biol 231(2):249–260 Ardener S (1964) The comparative study of rotating credit associations. J Anthropol Inst Great Britain Irel 94(2):201–229 Axelrod R (1984) The evolution of cooperation. Basic Books, New York Axelrod R, Hamilton W (1981) The evolution of cooperation. Science 211(4489):1390–1396 Balikci A, American Museum of Natural History (1970) The Netsilik Eskimo, 1st edn. Natural History Press, Garden City Barclay P (2004) Trustworthiness and competitive altruism can also solve the ‘‘tragedy of the commons’’. Evol Hum Behav 25(4):209–220
123
190
N. Baumard
Barclay P (2006) Reputational benefits for altruistic punishment. Evol Hum Behav 27:325–344 Barclay P, Willer R (2007) Partner choice creates competitive altruism in humans. Proc Biol Sci 274(1610):749–753 Baron J, Ritov I (2008) The role of probability of detection in judgments of punishment. Unpublished manuscript Baron J, Gowda R, Kunreuther H (1993) Attitudes toward managing hazardous waste: what should be cleaned up and who should pay for it? Risk Anal 13(2):183–192 Baumard N (2008) A naturalist and mutualist theory of morality. Ecole Normale Supe´rieure, Paris Baumard N (2010a) L’origine de la morale Une histoire naturelle du bien et du mal. Odile Jacob, Paris Baumard N (2010b) Punishment is a group adaptation. Mind Soc. doi:10.1007/s11299-010-0080-3 Baumard N, Boyer P, Sperber D (2010) Evolution of fairness: cultural variability. Science 329(5990):388 Bernhard H, Fischbacher U, Fehr E (2006) Parochial altruism in humans. Nature 442(7105):912–915 Bernstein L (1992) Opting out of the legal system: extralegal contractual relations in the diamond industry. J Legal Stud 21(1):115–157 Black D (2000) On the origin of morality. In: Katz L (ed) Evolutionary origins of morality: crossdisciplinary perspectives (pp. xvi, 352 p.). Imprint Academic, Thorverton Boehm C (1999) Hierarchy in the forest: the evolution of egalitarian behavior. Harvard University Press, Cambridge Bolton G, Zwick R (1995) Anonymity versus punishment in ultimatum bargaining. Games Econ Behav 10(1):95–121 Boyd R, Richerson P (1988) The evolution of reciprocity in sizable groups. Theor Biol 132(3):337–356 Boyd R, Richerson P (2002) Group beneficial norms can spread rapidly in a structured population. J Theor Biol 215(3):287–296 Boyd R, Gintis H, Bowles S, Richerson P (2003) The evolution of altruistic punishment. Proc Natl Acad Sci USA 100(6):3531–3535 Boyd R, Gintis H, Bowles S (2010) Coordinated punishment of defectors sustains cooperation and can proliferate when rare. Science 328(5978):617 Briggs JL (1970) Never in anger: portrait of an Eskimo family. Harvard University Press, Cambridge Brosig J (2002) Identifying cooperative behavior: some experimental results in a prisoner’s dilemma game. J Econ Behav Organ 47(3):275–290 Bshary R, Noe¨ R (2003) The ubiquitous influence of partner choice on the dynamics of cleaner fish–client reef fish interactions. In: Hammmerstein P (ed) Genetic and cultural evolution of cooperation. MIT, Cambridge, pp 167–184 Bull J, Rice W (1991) Distinguishing mechanisms for the evolution of co-operation. J Theor Biol 149(1):63 Carlsmith K, Darley J, Robinson P (2002) Why do we punish? Deterrence and just deserts as motives for punishment. J Pers Soc Psychol 83(2):284–299 Cashdan E (1980) Egalitarianism among hunters and gatherers. Am Anthropol 82(1):116–120 Chiang Y (2010) Self-interested partner selection can lead to the emergence of fairness. Evol Hum Behav 31(4):265–270 Cinyabuguma M, Page T, Putterman L (2004) On perverse and second-order punishment in public goods experiments with decentralized sanctioning. Working paper, Department of Economics, Brown University Clastres P (1972) Chronique des Indiens Guayaki; ce que savent les Ache´, chasseurs nomades du Paraguay. Plon, Paris Clutton-Brock T, Parker G (1995) Punishment in animal societies. Nature 373(6511):209–216 Coricelli G, Fehr D, Fellner G (2004) Partner selection in public goods experiments. J Conflict Resolut 48(3):356–378 Darley J, Carlsmith K, Robinson P (2000) Incapacitation and just deserts as motives for punishment. Law Hum Behav 24(6):659–683 Dawes CT, Fowler JH, Johnson T, McElreath R, Smirnov O (2007) Egalitarian motives in humans. Nature 446(7137):794–796 de Quervain DJ, Fischbacher U, Treyer V, Schellhammer M, Schnyder U, Buck A et al (2004) The neural basis of altruistic punishment. Science 305(5688):1254–1258 Denant-Boemont L, Masclet D, Noussair C (2007) Punishment, counterpunishment and sanction enforcement in a social dilemma experiment. J Econ Theory 33(1):145–167 Dreber A, Rand D, Fudenberg D, Nowak M (2008) Winners don’t punish. Nature 452(7185):348–351 Durkheim E´ (1893) De la division du travail social. F. Alcan, Paris
123
Role of punishment in the evolution of cooperation
191
Ehrhart K-M, Keser C (1999) Mobility and cooperation: on the run. Se´rie scientifique (CIRANO);99s-24 Evans-Pritchard EE (1940) The Nuer, a description of the modes of livelihood and political institutions of a Nilotic people. The Clarendon, Oxford Falk A, Fehr E, Fischbacher U (2005) Driving forces behind informal sanctions. Econometrica 73:2017–2030 Fehr E, Fischbacher U (2003) The nature of human altruism. Nature 425(6960):785–791 Fehr E, Fischbacher U (2004) Social norms and human cooperation. Trends Cogn Sci 8(4):185–190 Fehr E, Ga¨chter S (2002) Altruistic punishment in humans. Nature 415:137–140 Frank R, Gilovich T, Regan D (1993) The evolution of one-shot cooperation: an experiment. Ethol Sociobiol 14:247–250 Fu¨rer-Haimendorf Cv (1967) Morals and merit: a study of values and social controls in South Asian societies. Weidenfeld & Nicolson, London Gintis H (2009) The bounds of reason: game theory and the unification of the behavioral sciences. Princeton University Press, Princeton Gintis H, Bowles S, Boyd R, Fehr E (2003) Explaining altruistic behavior in humans. Evol Hum Behav 24(3):153–172 Gintis H, Bowles S, Boyd R, Fehr E (2005) Moral sentiments and material interests: origins, evidence, and consequences. In: Moral sentiments and material interests: the foundations of cooperation in economic life, pp 3–39 Glaeser EL, Sacerdote B (2000) The determinants of punishment: deterrence, incapacitation and vengeance. SSRN Greif A (1993) Contract enforceability and economic institutions in early trade: the maghribi traders’ coalition. Am Econ Rev 83(3):525–548 Gurven M (2004) To give and to give not: the behavioral ecology of human food transfers. Behav Brain Sci 27:543–559 Hardin G (1968) The tragedy of the commons. Science 162:1243–1248 Hardy CL, Van Vugt M (2006) Nice guys finish first: the competitive altruism hypothesis. Pers Soc Psychol Bull 32(10):1402 Hechter M (1987) Principles of group solidarity. University of California Press, Berkeley Hechter M, Kanazawa S (1993) Group solidarity and social order in Japan. J Theor Polit 5(4):455–493 Henrich J, Boyd R (2001) Why people punish defectors: weak conformist transmission can stabilize costly enforcement of norms in cooperative dilemmas. J Theor Biol 208:79–89 Henrich J, McElreath R, Barr A, Ensminger J, Barrett C, Bolyanatz A et al (2006) Costly punishment across human societies. Science 312(5781):1767–1770 Henrich J, Ensminger J, McElreath R, Barr A, Barrett C, Bolyanatz A et al (2010) Markets, religion, community size, and the evolution of fairness and punishment. Science 327(5972):1480 Herrmann B, Ga¨chter S, Tho¨ni C (2008) Antisocial punishment across societies. Science 319: 1362 Hoebel EA (1954) The law of primitive man: a study in comparative legal dynamics. Harvard University Press, Cambridge Johnson T, Dawes C, Fowler J, McElreath R, Smirnov O (2009) The role of egalitarian motives in altruistic punishment. Econ Lett 102(3):192–194 Kaplan H, Gurven M (2001) The natural history of human food sharing and cooperation: a review and a new multi-individual approach to the negotiation of norms paper presented at the conference on the structure and evolution of strong reciprocity, Santa Fe Institute Knauft B (1987) Reconsidering violence in simple human societies: homicide among the Gebusi of New Guinea. Curr Anthropol 28(4):457–500 Kurzban R, DeScioli P (2008) Reciprocity in groups: information-seeking in a public goods game. Eur J Soc Psychol 38(1):139 Kurzban R, DeScioli P (2009) Adaptationist punishment in humans. Context Kurzban R, DeScioli P, O’Brien E (2006) Audience effects on moralistic punishment. Working paper Landa JT (1981) A theory of the ethnically homogeneous middleman group: an institutional alternative to contract law. J Legal Stud 10(2):349–362 Lee RB (1979) The !Kung San: men, women, and work in a foraging society. Cambridge University Press, Cambridge Leibbrandt A, Lo´pez-Pe´rez R (2008) The envious punisher. Working paper, University of Zurich Malinowski B (1926/1985) Crime and custom in savage society. Rowman & Littlefield, New York McFatter RM (1982) Purposes of punishment: effects of utilities of criminal sanctions on perceived appropriateness. J Appl Psychol 67(3):255
123
192
N. Baumard
Nikiforakis N (2008) Punishment and counter-punishment in public good games: can we really govern ourselves? J Public Econ 92(1–2):91–112 Noe¨ R, van Schaik C, Van Hooff J (1991) The market effect: an explanation for pay-off asymmetries among collaborating animals. Ethology 87(1–2):97–118 Ohtsubo Y, Watanabe E (2008) Do sincere apologies need to be costly? Test of a costly signaling model of apology. Evol Hum Behav 30(2):114–123 Olson M (1965) The logic of collective action; public goods and the theory of groups. Harvard University Press, Cambridge Ostrom E (1990) Governing the commons: the evolution of institutions for collective action. Cambridge University Press, Cambridge Page T, Putterman L, Unel B (2005) Voluntary association in public goods experiments: reciprocity, mimicry and efficiency. Econ J 115(506):1032–1053 Polinsky AM, Shavell S (2000) The economic theory of public enforcement of law. J Econ Lit 38(1):45–76 Posner R (1983) The economics of justice. Harvard University Press, Cambridge Pradel J, Euler HA, Fetchenhauer D (2008) Spotting altruistic dictator game players and mingling with them: the elective assortation of classmates. Evol Hum Behav 30(2):103–113 Price ME (2005) Punitive sentiment among the Shuar and in industrialized societies: cross-cultural similarities. Evol Hum Behav 26:279–287 Radcliffe-Brown AR (1922) The Andaman islanders: a study in social anthropology. The University press, Cambridge Rand D, Dreber A, Ellingson T, Fudenberg D, Nowak M (2009) Positive interactions promote public cooperation. Science 325(5945):1272–1275 Roberts G (1998) Competitive altruism: from reciprocity to the handicap principle. Proc R Soc Lond B 265:427–431 Roberts J, Gebotys R (1989) The purposes of sentencing: public support for competing aims. Behav Sci Law 7(3):387–402 Rouland N (1976) La Sanction du droit chez certaines populations esquimaudes: proble`mes the´oriques et modes d’intervention de la communaute´ dans les proce´dures de re`glement des conflit. Centre d’Etude Arctique, Paris Saijo T, Nakamura H (1995) The ‘‘spite’’ dilemma in voluntary contribution mechanism experiments. J Conflict Resolut 39(3):535–560 Sheldon KM, Sheldon MS, Osbaldiston R (2000) Prosocial values and group assortation. Hum Nat 11(4):387–404 Sober E, Wilson D (1998) Unto others: the evolution and psychology of unselfish behavior. Harvard University Press, Cambridge Sripada C (2005) Punishment and the strategic structure of moral systems. Biol Philos 20(4):767–789 Sunstein C, Schkade D, Kahneman D (2000) Do people want optimal deterrence? J Legal Stud 29(1):237–253 Sylwester K, RobertsG (2010) Cooperators benefit through reputation-based partner choice in economic games. Biol Lett 6:659–662 Tooby J, Cosmides L, Price ME (2006) Cognitive adaptations for n-person exchange: the evolutionary roots of organizational behavior. Manag Decis Econ 27:103–129 Trivers R (1971) Evolution of reciprocal altruism. Q Rev Biol 46:35–57 Turnbull CM (1965) The Mbuti pygmies: an ethnographic survey. American Museum of Natural History, New York Verplaetse J, Vanneste S, Braeckman J (2007) You can judge a book by its cover: the sequel. A kernel of truth in predictive cheating detection. Evol Hum Behav 28(4):260–271 Wiessner P (2005) Norm enforcement among the Ju/’hoansi Bushmen: a case of strong reciprocity? Hum Nat 16(2):115–145 Woodburn J (1982) Egalitarian societies. Man 17(3):431–451 Wu J, Zhang B, Zhou Z, He Q, Zheng X, Cressman R et al (2009) Costly punishment does not always increase cooperation. Proc Natl Acad Sci USA 106(41):17448–17451 Yamagishi T (1986) The provision of a sanctioning system as a public good. J Pers Soc Psychol 51(1):110–116
123
