Punishment and cooperation in nature
2012; Elsevier BV; Volume: 27; Issue: 5 Linguagem: Inglês
10.1016/j.tree.2011.12.004
ISSN1872-8383
AutoresNichola Raihani, Alex Thornton, Redouan Bshary,
Tópico(s)Evolutionary Psychology and Human Behavior
ResumoHumans use punishment to promote cooperation in laboratory experiments but evidence that punishment plays a similar role in non-human animals is comparatively rare. In this article, we examine why this may be the case by reviewing evidence from both laboratory experiments on humans and ecologically relevant studies on non-human animals. Generally, punishment appears to be most probable if players differ in strength or strategic options. Although these conditions are common in nature, punishment (unlike other forms of aggression) involves immediate payoff reductions to both punisher and target, with net benefits to punishers contingent on cheats behaving more cooperatively in future interactions. In many cases, aggression yielding immediate benefits may suffice to deter cheats and might explain the relative scarcity of punishment in nature. Humans use punishment to promote cooperation in laboratory experiments but evidence that punishment plays a similar role in non-human animals is comparatively rare. In this article, we examine why this may be the case by reviewing evidence from both laboratory experiments on humans and ecologically relevant studies on non-human animals. Generally, punishment appears to be most probable if players differ in strength or strategic options. Although these conditions are common in nature, punishment (unlike other forms of aggression) involves immediate payoff reductions to both punisher and target, with net benefits to punishers contingent on cheats behaving more cooperatively in future interactions. In many cases, aggression yielding immediate benefits may suffice to deter cheats and might explain the relative scarcity of punishment in nature. Punishment in nature: unresolved issuesIndividuals are often tempted to cheat in social interactions, thereby gaining a benefit at the expense of cooperative partners. To encourage partners to behave cooperatively, individuals might therefore use control mechanisms that render cooperative behaviour a more profitable option than cheating for the partner. One such mechanism is punishment (see Glossary) [1Clutton-Brock T.H. Parker G.A. Punishment in animal societies.Nature. 1995; 373: 209-216Crossref PubMed Scopus (712) Google Scholar]. Several laboratory studies have shown that punishment promotes cooperation among humans, typically using stylised laboratory games (e.g. [2Fehr E. Gächter S. Cooperation and punishment in public goods experiments.Am. Econ. Rev. 2000; 90: 980-994Crossref Scopus (2232) Google Scholar, 3Fehr E. Gächter S. Altruistic punishment in humans.Nature. 2002; 415: 137-140Crossref PubMed Scopus (3220) Google Scholar, 4Rockenbach B. Milinski M. The efficient interaction of indirect reciprocity and costly punishment.Nature. 2006; 444: 718-723Crossref PubMed Scopus (347) Google Scholar, 5Gächter S. et al.The long-run benefits of punishment.Science. 2008; 322: 1510Crossref PubMed Scopus (369) Google Scholar, 6Dreber A. et al.Winners don't punish.Nature. 2008; 452: 348-351Crossref PubMed Scopus (485) Google Scholar]). By comparison, only a handful of studies have shown that punishment promotes cooperation among non-human animals [7Bshary R. Grutter A.S. Asymmetric cheating opportunities and partner control in a cleaner fish mutualism.Anim. Behav. 2002; 63: 547-555Crossref Scopus (192) Google Scholar, 8Bshary R. Grutter A.S. Punishment and partner switching cause cooperative behaviour in a cleaning mutualism.Biol. Lett. 2005; 1: 396-399Crossref PubMed Scopus (168) Google Scholar, 9Bshary A. Bshary R. Self-serving punishment of a common enemy creates a public good in reef fishes.Curr. Biol. 2010; 20: 2032-2035Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar, 10Raihani N.J. et al.Punishers benefit from third-party punishment in fish.Science. 2010; 327: 171Crossref PubMed Scopus (100) Google Scholar, 11Raihani N.J. et al.Male cleaner wrasses adjust punishment of female partners according to the stakes.Proc. R. Soc. Lond. B. 2012; 279: 365-370Crossref PubMed Scopus (39) Google Scholar]. This relative paucity of evidence prompted arguments about why initial predictions that punishment should be common [1Clutton-Brock T.H. Parker G.A. Punishment in animal societies.Nature. 1995; 373: 209-216Crossref PubMed Scopus (712) Google Scholar] do not fit current data [12Brosnan S.F. et al.The interplay of cognition and cooperation.Philos. Trans. R. Soc. Lond. B. 2010; 365: 2699-2710Crossref PubMed Scopus (126) Google Scholar, 13Bshary R. Bronstein J.L. A general scheme to predict partner control mechanisms in pairwise cooperative interactions between unrelated individuals.Ethology. 2011; 117: 1-13Crossref Scopus (12) Google Scholar].Here, we critically assess empirical evidence for punishment in non-human species. We first outline how punishment can be distinguished from other forms of aggression that promote cooperative behaviour, such as coercion and sanctions, and then go on to discuss specific empirical examples of punishment. We end by discussing the conditions that are likely to favour punishment over alternative control mechanisms and whether these conditions are likely to be met in non-human species.What is (and what is not) punishment?Following the seminal paper by Clutton-Brock and Parker [1Clutton-Brock T.H. Parker G.A. Punishment in animal societies.Nature. 1995; 373: 209-216Crossref PubMed Scopus (712) Google Scholar], we assert that punishment occurs when an individual reduces its own current payoffs to harm a cheating partner. In doing so, the punisher reduces the payoffs of the cheat and thereby promotes cooperative behaviour from the cheat in subsequent interactions (Box 1). Thus, punishment is equivalent to 'negative reciprocity' [1Clutton-Brock T.H. Parker G.A. Punishment in animal societies.Nature. 1995; 373: 209-216Crossref PubMed Scopus (712) Google Scholar, 14Boyd R. et al.The evolution of altruistic punishment.Proc. Natl. Acad. Sci. U.S.A. 2003; 100: 3531-3535Crossref PubMed Scopus (1063) Google Scholar]. This functional definition is useful for studying punishment among non-human animals because punishment is not contingent on a capacity for mental state attribution and does not require the punisher to be aware of how its behaviour might influence that of the target [1Clutton-Brock T.H. Parker G.A. Punishment in animal societies.Nature. 1995; 373: 209-216Crossref PubMed Scopus (712) Google Scholar]. Punishers need not always be involved in the initial interaction with the cheat. For example, in 'policing' or 'third party punishment', a bystander observes a cheat and is willing to reduce its own current payoffs to reduce the payoff to the cheat. It is still largely unclear how punishers benefit from third-party punishment, however [14Boyd R. et al.The evolution of altruistic punishment.Proc. Natl. Acad. Sci. U.S.A. 2003; 100: 3531-3535Crossref PubMed Scopus (1063) Google Scholar, 15Gardner A. West S.A. Cooperation and punishment, especially in humans.Am. Nat. 2004; 164: 753-764Crossref Scopus (159) Google Scholar, 16Lehmann L. et al.Strong reciprocity or strong ferocity? A population genetic view of the evolution of altruistic punishment.Am. Nat. 2007; 170: 21-36Crossref PubMed Scopus (86) Google Scholar, 17Fowler J.H. Altruistic punishment and the origin of cooperation.Proc. Natl. Acad. Sci. U.S.A. 2005; 102: 7047-7049Crossref PubMed Scopus (306) Google Scholar, 18dos Santos M. et al.The evolution of punishment through reputation.Proc. R. Soc. Lond. B. 2011; 278: 371-377Crossref PubMed Scopus (88) Google Scholar].Box 1Payoffs associated with punishment and sanctionsFigure I shows categories of interactions among non-kin. The arrows indicate the initial instigator and recipient and the + or − signs represent fitness changes (as in [1Clutton-Brock T.H. Parker G.A. Punishment in animal societies.Nature. 1995; 373: 209-216Crossref PubMed Scopus (712) Google Scholar]).PunishmentIn punishment (Figure Ia), one individual cheats a partner, thereby increasing its immediate payoffs (++ relative to the payoff increase associated with cooperating, +) and imposing a fitness cost, −, on the partner. The partner then retaliates with a behaviour that reduces its immediate payoffs further, −, which imposes a fitness cost on the cheat. The fitness costs experienced by the cheat can be equal to (−) or greater than (−−) the payoff losses to the punisher of executing the punishment. In response to punishment, the cheat behaves more cooperatively in subsequent interactions with the punisher. Note that the cooperative interaction is mutually beneficial (in fitness terms) to both players compared to outside options (not interacting) but that, in the absence of punishment, the cheat could gain higher payoffs from exploiting a cooperative partner. Also note that some effects are on lifetime fitness (the effects of being cheated, of being punished and of mutual cooperation), whereas others are on immediate payoffs (the effects of cheating and of punishing). In interactions where cooperative behaviour is binary, the cheating is simply the opposite of cooperative behaviour. In interactions where cooperative behaviour can be a continuous investment, then cheating can be defined as any investment that is less than the population mean [55Bull J.J. Rice W.R. Distinguishing mechanisms for the evolution of cooperation.J. Theor. Biol. 1991; 149: 63-74Crossref PubMed Scopus (346) Google Scholar].SanctionsSanctions (Figure Ib) occur when one individual cheats a partner, thereby gaining a payoff increase relative to cooperating [56Herre E.A. et al.The evolution of mutualisms: exploring the paths between conflict and cooperation.Trends Ecol. Evol. 1999; 14: 49-53Abstract Full Text Full Text PDF PubMed Scopus (600) Google Scholar]. In response, the partner performs a self-serving behaviour that imposes costs on the cheat as a byproduct (2). This self-serving behaviour also serves to end the interaction. Thus, sanctions fit the concept of negative pseudo-reciprocity. We note that the term 'sanction' has also been used differently in the literature (e.g. [57Noë R. Despotic partner choice puts helpers under pressure?.Behav. Process. 2007; 76: 120-125Crossref PubMed Scopus (4) Google Scholar, 58Leimar O. Hammerstein P. Cooperation for direct fitness benefits.Philos. Trans. R. Soc. Lond. B. 2010; 365: 2619-2626Crossref PubMed Scopus (89) Google Scholar, 59Weyl E.G. et al.Economic contract theory tests models of mutualism.Proc. Natl. Acad. Sci. U.S.A. 2010; 107: 15712-15716Crossref PubMed Scopus (85) Google Scholar]). Under the concept of sanctions, there is no future to the interaction. Classic examples include the selective abortion by yucca trees of fruits that harbour too many seed-eating larvae of its pollinator, the yucca moth [60Pellmyr O. Huth C.J. Evolutionary stability of mutualism between yuccas and yucca moths.Nature. 1994; 372: 257-260Crossref Scopus (327) Google Scholar]; and the selective inhibition of nodule growth by leguminous plants in root parts in which rhizobia partner bacteria fail to fix significant amounts of nitrogen [61Kiers E.T. et al.Host sanctions and the legume–rhizobium mutualism.Nature. 2003; 425: 78-81Crossref PubMed Scopus (644) Google Scholar]. The permanent eviction of uncooperative individuals from a territory or group also fits the sanctions concept. For example, in coral-dwelling gobies (Paragobiodon xanthosomus; Figure II) dominant individuals sometimes evict similar-sized subordinates, because subordinates that grow too large can threaten the superior status of their dominant neighbour [62Wong M.Y.L. et al.The threat of punishment enforces peaceful cooperation and stabilizes queues in a coral-reef fish.Proc. R. Soc. Lond. B. 2007; 274: 1093-1099Crossref PubMed Scopus (117) Google Scholar]. Dominant gobies benefit from evicting overgrown competitors without the need for subordinates to behave more cooperatively (by reducing growth rate) in future. Indeed, evictees rarely return to their group and there is therefore little or no potential for them to cooperate more in response to eviction from dominants. Noë [63Noë R. Biological markets: partner choice as the driving force behind the evolution of cooperation.in: Noë R. Economics in Nature. Social Dilemmas, Mate Choice and Biological Markets. Cambridge University Press, 2001: 93-118Crossref Google Scholar] pointed out that sanctions often occur within a biological market in which individuals choose the best partner out of a possible range. Although the simple threat of terminating an interaction can be enough to promote cooperative behaviour [64Johnstone R.A. Bshary R. From parasitism to mutualism: partner control in asymmetric interactions.Ecol. Lett. 2002; 5: 634-639Crossref Scopus (91) Google Scholar], the additional opportunity of partner switching could enhance the effect [65Johnstone R.A. Bshary R. Mutualism, market effects and partner control.J. Evol. Biol. 2008; 21: 879-888Crossref PubMed Scopus (52) Google Scholar, 66Ferriere R. et al.Cheating and the evolutionary stability of mutualisms.Proc. R. Soc. Lond. B. 2002; 269: 773-780Crossref PubMed Scopus (209) Google Scholar, 67Foster K.R. Wenselaars T. A general model for the evolution of mutualisms.J. Evol. Biol. 2006; 19: 1283-1293Crossref PubMed Scopus (234) Google Scholar]. Indeed, switching to a different partner is an efficient way to select against cheating in marine cleaning mutualisms [68Bshary R. Schäffer D. Choosy reef fish select cleaner fish that provide high-quality service.Anim. Behav. 2002; 63: 557-564Crossref Scopus (165) Google Scholar].Figure IIA coral-dwelling goby. Dominant gobies sanction subordinates that breach a defined size threshold by evicting them from the group. Reproduced, with permission, from Joao Paulo Krajewski.View Large Image Figure ViewerDownload (PPT)As with punishment, some other control mechanisms also rely on responses to cheating that reduce the payoffs to cheats. However, unlike punishment, such responses do not necessarily reduce the current payoffs of the actor. Instead, several responses to cheating described in the literature are immediately self-serving and, hence, do not rely on future benefits arising from the increased cooperative behaviour of the target to be under positive selection. These examples do not fit the negative reciprocity concept but are instead cases of sanctions or 'negative pseudo-reciprocity' [19Bergmüller R. et al.Integrating cooperative breeding into theoretical concepts of cooperation.Behav. Process. 2007; 76: 61-72Crossref PubMed Scopus (180) Google Scholar, 20Bshary R. Bergmüller R. Distinguishing four fundamental approaches to the evolution of helping.J. Evol. Biol. 2008; 21: 405-420Crossref PubMed Scopus (111) Google Scholar] (Box 1).Coercion is another form of aggressive behaviour that can induce cooperative behaviour in the target. However, coercion differs from punishment because, from the point of view of the target, no interaction yields a higher payoff than interacting and cooperating with the aggressor. Thus, coercion occurs when targets of aggressive behaviour would do best to avoid interactions with the aggressor but are somehow prevented from exercising this higher paying outside option. Experiments on a coordination task in keas (Nestor notabilis) provide a good example [21Tebbich S. et al.Social manipulation causes cooperation in keas.Anim. Behav. 1996; 52: 1-10Crossref Scopus (56) Google Scholar]. In the experiment, one individual had to sit on a lever to lift a lid covering a food tray, thereby allowing another individual to feed. Under these conditions, dominant birds aggressively forced subordinate partners to sit on the lever without ever reciprocating. Forced copulations in animals also fit the concept of coercion [22Clutton-Brock T.H. Sexual coercion in animal societies.Anim. Behav. 1995; 49: 1345-1365Crossref Scopus (569) Google Scholar].Mixed evidence that punishment promotes cooperationThe effect of punishment on cooperation has been best studied in humans, typically using n-player public goods games (PGGs) under controlled laboratory settings [2Fehr E. Gächter S. Cooperation and punishment in public goods experiments.Am. Econ. Rev. 2000; 90: 980-994Crossref Scopus (2232) Google Scholar, 3Fehr E. Gächter S. Altruistic punishment in humans.Nature. 2002; 415: 137-140Crossref PubMed Scopus (3220) Google Scholar, 4Rockenbach B. Milinski M. The efficient interaction of indirect reciprocity and costly punishment.Nature. 2006; 444: 718-723Crossref PubMed Scopus (347) Google Scholar, 5Gächter S. et al.The long-run benefits of punishment.Science. 2008; 322: 1510Crossref PubMed Scopus (369) Google Scholar, 23Gächter S. Herrmann B. Reciprocity, culture and human cooperation: previous insights and a new cross-cultural experiment.Philos. Trans. R. Soc. Lond. B. 2009; 365: 2619-2626Google Scholar]. Players can punish free-riders by paying a small fee to impose a larger fine on the cheat. Although there is still some debate surrounding the evolutionary scenarios (Box 2), the majority of studies have shown that punishment promotes cooperation in n-player games (reviewed in [23Gächter S. Herrmann B. Reciprocity, culture and human cooperation: previous insights and a new cross-cultural experiment.Philos. Trans. R. Soc. Lond. B. 2009; 365: 2619-2626Google Scholar, 24Chaudhuri A. Sustaining cooperation in laboratory public goods experiments: a selective survey of the literature.Exp. Econ. 2010; 14: 47-83Crossref Scopus (688) Google Scholar, 25Balliet D. et al.Reward, punishment, and cooperation: a meta-analysis.Psychol. Bull. 2011; 137: 594-615Crossref PubMed Scopus (368) Google Scholar, 26Milinski M. Rockenbach B. On the interaction of the stick and the carrot in social dilemmas.J. Theor. Biol. 2011; https://doi.org/10.1016/J.JTBI.2011.03.014Crossref PubMed Google Scholar]), although considerable cross-cultural differences in the administration and effects of punishment exist [27Rand D.G. et al.Positive interactions promote public cooperation.Science. 2009; 325: 1272-1275Crossref PubMed Scopus (464) Google Scholar, 28Henrich J. et al.Costly punishment across human societies.Science. 2006; 312: 1767-1770Crossref PubMed Scopus (1004) Google Scholar]. Punishment might be less successful in two-player interactions. Dreber et al. [6Dreber A. et al.Winners don't punish.Nature. 2008; 452: 348-351Crossref PubMed Scopus (485) Google Scholar] used iterated two-player prisoner's dilemma games with and without a punishment option and found that, although punishment promoted cooperative behaviour, punishers achieved lower payoffs than did non-punishers. Instead, players that responded to cheats with reciprocal defection achieved the highest payoffs. In this experiment, punishers were disadvantaged by the relatively short time horizon of expected interactions with the current partner. Interactions lasted between one and nine rounds, which according to another recent experimental study [5Gächter S. et al.The long-run benefits of punishment.Science. 2008; 322: 1510Crossref PubMed Scopus (369) Google Scholar], is insufficient for punishers to recoup their initial investment in harming a cheating partner.Box 2The evolution of punishment in n-player gamesExperimental studies investigating the evolution of cooperation in n-player games have typically used n-player prisoner's dilemma (NPD) payoffs, rendering contributions altruistic [69Hamilton W.D. The genetical evolution of social behaviour. I.J. Theor. Biol. 1964; 7: 1-16Crossref PubMed Scopus (9410) Google Scholar, 70Hamilton W.D. The genetical evolution of social behaviour. II.J. Theor. Biol. 1964; 7: 17-52Crossref PubMed Scopus (2711) Google Scholar] and resulting in the tragedy of the commons [71Hardin G. The tragedy of the commons.Science. 1968; 1: 243-253Google Scholar]. Evidence indicates that humans willingly punish free-riders and that targets subsequently behave more cooperatively [36Cinyabugama M. et al.Cooperation under the threat of expulsion in a public goods experiment.J. Pub. Econ. 2005; 89: 1421-1435Crossref Scopus (162) Google Scholar, 37Jensen K. et al.Chimpanzees are vengeful but not spiteful.Proc. Natl. Acad. Sci. U.S.A. 2007; 104: 13046-13050Crossref PubMed Scopus (123) Google Scholar, 38Koyama N.F. et al.Interchange of grooming and agonistic support in chimpanzees.Int. J. Primat. 2006; 27: 1293-1309Crossref Scopus (58) Google Scholar, 39Hauser M.D. Costs of deception: cheaters are punished in rhesus monkeys (Macaca mulatta).Proc. Natl. Acad. Sci. U.S.A. 1992; 89: 12137-12139Crossref PubMed Scopus (123) Google Scholar, 40Jensen K. Punishment and spite: the dark side of cooperation.Philos. Trans. R. Soc. Lond. B. 2010; 365: 2635-2650Crossref PubMed Scopus (132) Google Scholar]. However, in one-shot games, this raises a second-order social dilemma as punishers invest in harming free-riders although the resulting benefit (of increased cooperation) is shared among punishers and non-punishers alike. Nevertheless, because punishment often promotes cooperative behaviour in one-shot games, evolutionary explanations for its emergence and stability have been proposed. For example, several authors [14Boyd R. et al.The evolution of altruistic punishment.Proc. Natl. Acad. Sci. U.S.A. 2003; 100: 3531-3535Crossref PubMed Scopus (1063) Google Scholar, 72Gintis H. et al.Explaining altruistic behavior in humans.Evol. Hum. Behav. 2003; 24: 153-172Abstract Full Text Full Text PDF Scopus (612) Google Scholar, 73Bowles S. Gintis H. The evolution of strong reciprocity: evolution in heterogeneous populations.Theor. Pop. Biol. 2004; 65: 17-28Crossref PubMed Scopus (448) Google Scholar] have suggested that punishment could spread through cultural group selection. Here, social learning facilitates the local spread of punitive behaviour, and demes with a high number of punishers outperform demes without punishment. Arguments over the importance of cultural group selection for the evolution of punishment have centred on two issues. First, Gardner and West [15Gardner A. West S.A. Cooperation and punishment, especially in humans.Am. Nat. 2004; 164: 753-764Crossref Scopus (159) Google Scholar] and Lehmann et al. [16Lehmann L. et al.Strong reciprocity or strong ferocity? A population genetic view of the evolution of altruistic punishment.Am. Nat. 2007; 170: 21-36Crossref PubMed Scopus (86) Google Scholar] pointed out that, because punishment is altruistic in these models, it relies on kin selection to spread in a population. Therefore, the logic of inclusive fitness theory still applies. Second, cultural group selection models struggle to explain how punishment becomes established when it is initially rare [15Gardner A. West S.A. Cooperation and punishment, especially in humans.Am. Nat. 2004; 164: 753-764Crossref Scopus (159) Google Scholar, 16Lehmann L. et al.Strong reciprocity or strong ferocity? A population genetic view of the evolution of altruistic punishment.Am. Nat. 2007; 170: 21-36Crossref PubMed Scopus (86) Google Scholar, 17Fowler J.H. Altruistic punishment and the origin of cooperation.Proc. Natl. Acad. Sci. U.S.A. 2005; 102: 7047-7049Crossref PubMed Scopus (306) Google Scholar]. Others have argued that punishment in one-shot games occurs because humans evolved in a social system in which repeated interactions are the norm and often take place in a communication network [74McGregor P.K. Animal Communication Networks. Cambridge University Press, 2005Crossref Scopus (282) Google Scholar]. Thus, humans are error-prone when confronted with anonymous one-shot interactions [18dos Santos M. et al.The evolution of punishment through reputation.Proc. R. Soc. Lond. B. 2011; 278: 371-377Crossref PubMed Scopus (88) Google Scholar, 75Trivers R. Mutual benefits at all levels of life.Science. 2004; 304: 964-965Crossref Google Scholar, 76Hammerstein P. Hagen E.H. The second wave of evolutionary economics in biology.Trends Ecol. Evol. 2006; 20: 604-609Abstract Full Text Full Text PDF Scopus (58) Google Scholar, 77Sigmund K. Punish or perish? Retaliation and collaboration among humans.Trends Ecol. Evol. 2007; 22: 593-600Abstract Full Text Full Text PDF PubMed Scopus (217) Google Scholar, 78Kummerli R. et al.Resistance to extreme strategies, rather than prosocial preferences, can explain human cooperation in public goods games.Proc. Natl. Acad. Sci. U.S.A. 2010; 107: 10125-10130Crossref PubMed Scopus (65) Google Scholar].More generally, explanations for the evolution of punishment might have been hindered by the use of inappropriate payoff matrices. Specifically, punishment is efficient when it reduces a free-rider's payoff below the population average; at this point, the target does best to contribute rather than free-ride. Further punishment is wasteful because it reduces group gains without increasing contributions from targets. Therefore, the net benefits of punishment are not a linear function of contributions as assumed in the NPD framework but are better described with a step function as assumed in the volunteer's dilemma game [79Dieckmann A. Volunteer's dilemma.J. Confl. Resol. 1985; 29: 605-610Crossref Scopus (288) Google Scholar, 80Raihani N.J. Bshary R. The evolution of punishment in n-player public goods games: a volunteer's dilemma.Evolution. 2011; 65: 2725-2728Crossref PubMed Scopus (42) Google Scholar]; Figure I). In non-linear public goods games, cooperators and punishers are expected to coexist in a stable mixed equilibrium [81Boza G. Szamado S. Beneficial laggards: multilevel selection, cooperative polymorphism and division of labour in threshold public goods games.BMC Evol. Biol. 2010; 10: 336-348Crossref PubMed Scopus (39) Google Scholar, 82Archetti M. Scheuring I. Coexistence of cooperation and defection in public goods games.Evolution. 2011; 65: 1140-1148Crossref PubMed Scopus (136) Google Scholar] and punishers can also invade when they are initially rare.Figure ISimple schematic of the costs and benefits of punishment according to the investment in punishment under the volunteer's dilemma framework. Although the costs increase linearly with increasing investment in punishment, the benefits (in terms of increased future cooperative behaviour from the punished individual) follow a step function. Any investment below the threshold yields no benefits, whereas further investment above the threshold yields no additional benefits. Reproduced, with permission, from [80Raihani N.J. Bshary R. The evolution of punishment in n-player public goods games: a volunteer's dilemma.Evolution. 2011; 65: 2725-2728Crossref PubMed Scopus (42) Google Scholar].View Large Image Figure ViewerDownload (PPT)In contrast to the large number of laboratory studies, there have been relatively few real-world studies of punishment and cooperation in humans. Notable exceptions have focused on hunter-gatherer societies and typically describe centralised punishment rather than peer punishment (see Glossary) [29Wiessner P. Norm enforcement among the Ju/'hoansi bushmen: a case of strong reciprocity?.Hum. Nat. 2005; 16: 115-145Crossref Scopus (146) Google Scholar, 30Rustagi D. et al.Conditional cooperation and costly monitoring explain success in forest commons management.Science. 2010; 330: 961-965Crossref PubMed Scopus (305) Google Scholar, 31Mathew S. Boyd R. Punishment sustains large-scale cooperation in prestate warfare.Proc. Natl. Acad. Sci. U.S.A. 2011; 108: 11375-11380Crossref PubMed Scopus (222) Google Scholar, 32Baumard N. Liénard P. Second- or third-party punishment? When self-interest hides behind apparent functional interventions.Proc. Natl. Acad. Sci. U.S.A. 2011; 108: E753Crossref PubMed Scopus (6) Google Scholar]. None of these studies have explicitly examined whether punishment causes an increase in the future cooperative behaviour of the target. Thus, these field studies do not help to elucidate the precise conditions that would favour punishment over alternative control mechanisms, such as terminating the interaction with a cheating partner, partner switching, or responding with reciprocal cheating (but see [33Hilbe C. Sigmund K. Incentives and opportunism: from the carrot to the stick.Proc. R. Soc. Lond. B. 2010; 277: 2427-2433Crossref PubMed Scopus (126) Google Scholar] for a theoretical approach). Similarly, very little work has addressed questions about the form that punishment is likely to take in reality and about the relative efficacy of different types of punishment. For example, rather than monetary fines, punishment can also take the form of physical aggression, verbal reprimands, negative gossip statements or ostracism [29Wiessner P. Norm enforcement among the Ju/'hoansi bushmen: a case of strong reciprocity?.Hum. Nat. 2005; 16: 115-145Crossref Scopus (146) Google Scholar, 34Masclet D. et al.Monetary and nonmonetary punishment in the voluntary contributions mechanism.Am. Econ. Rev. 2003; 93: 366-380Crossref Scopus (477) Google Scholar, 35Carpenter J.P. et al.Cooperation, trust, and social capital in Southeast Asian urban slums.J. Econ. Behav. Organ. 2004; 55: 533-551Crossref Scopus (139) Google Scholar, 36Cinyabugama M. et al.Cooperation under the threat of expulsion in a public goods experiment.J. Pub. Econ. 2005; 89: 1421-1435Crossref Scopus (162) Google Scholar]. These different types of punishment might impose variable costs on cheats and differentially affect their propensity to cooperate. More data on punishment in humans under real-world settings are clearly a research priority.In contrast to human laboratory studies, relatively few studies have demonstrated experimentally that non-human animals use punishment to promote cooperation. Perhaps surprisingly, evidence for punishment in closely related non-human primate s
Referência(s)