Economists and biologists have proposed a distinction between two mechanisms – “strong” and “weak” reciprocity – that may explain the evolution of human sociality. Weak reciprocity theorists emphasize the benefits of long-term cooperation and the use of low-cost strategies to deter free-riders. Strong reciprocity theorists, in contrast, claim that cooperation in social dilemma games can be sustained by costly punishment mechanisms, even in one-shot and finitely repeated games. To support this claim, they have generated a large body of evidence concerning the willingness of experimental subjects to punish uncooperative free-riders at a cost to themselves. In this article, I distinguish between a “narrow” and a “wide” reading of the experimental evidence. Under the narrow reading, punishment experiments are just useful devices to measure psychological propensities in controlled laboratory conditions. Under the wide reading, they replicate a mechanism that supports cooperation also in “real-world” situations outside the laboratory. I argue that the wide interpretation must be tested using a combination of laboratory data and evidence about cooperation “in the wild.” In spite of some often-repeated claims, there is no evidence that cooperation in the small egalitarian societies studied by anthropologists is enforced by means of costly punishment. Moreover, studies by economic and social historians show that social dilemmas in the wild are typically solved by institutions that coordinate punishment, reduce its cost, and extend the horizon of cooperation. The lack of field evidence for costly punishment suggests important constraints about what forms of cooperation can or cannot be sustained by means of decentralised policing.

We review evidence of the psychological and social costs associated with punishing. We propose that these psychological and social costs should be considered (in addition to material costs) when searching for evidence of costly punishment “in the wild.”

While admirable, Guala's discussion of reciprocity suffers from a confusion between proximate causes (psychological mechanisms triggering behaviour) and ultimate causes (evolved function of those psychological mechanisms). Because much work on “strong reciprocity” commits this error, I clarify the difference between proximate and ultimate causes of cooperation and punishment. I also caution against hasty rejections of “wide readings” of experimental evidence.

Strong reciprocity theorists claim that punishment has evolved to promote the good of the group and to deter cheating. By contrast, weak reciprocity suggests that punishment aims to restore justice (i.e., reciprocity) between the criminal and his victim. Experimental evidences as well as field observations suggest that humans punish criminals to restore fairness rather than to support group cooperation.

Guala points to a discrepancy between strong negative reciprocity observed in the lab and the way cooperation is sustained “in the wild.” This commentary suggests that in lab experiments, strong negative reciprocity is limited when uncertainty exists regarding the players' actions and the intentions. Thus, costly punishment is indeed a limited mechanism for sustaining cooperation in an uncertain environment.

Hunter-gatherer punishment involves costs and benefits to individuals and groups, but the costs do not necessarily fit with the assumptions made in models that consider punishment to be altruistic – which brings in the free-rider problem and the problem of second-order free-riders. In this commentary, I present foragers' capital punishment patterns ethnographically, in the interest of establishing whether such punishment is likely to be costly; and I suggest that in many cases abstentions from punishment that might be taken as defections by free-riders are actually caused by social-structural considerations rather than being an effect of free-rider genes. This presentation of data supplements the ethnographic analysis provided by Guala.

Experiments are not models of cooperation; instead, they demonstrate the presence of the ethical and other-regarding predispositions that often motivate cooperation and the punishment of free-riders. Experimental behavior predicts subjects' cooperation in the field. Ethnographic studies in small-scale societies without formal coercive institutions demonstrate that disciplining defectors is both essential to cooperation and often costly to the punisher.

The claims about (1) the lack of empirical support for a model of strong reciprocation and (2) the irrelevant empirical role of costly punishment to support cooperation in the field need qualifications. The interpretation of field evidence is not straightforward, and other-regarding preferences are also likely to play a role in the field.

Guala contests the validity of strong reciprocity as a key element in shaping social behavior by contrasting evidence from experimental games to that of natural and historic data. He suggests that in order to understand the evolution of social behavior researchers should focus on natural data and weak reciprocity. We disagree with Guala's proposal to shift the focus of the study from one extreme of the spectrum (strong reciprocity) to the other extreme (weak reciprocity). We argue that the study of the evolution of social behavior must be comparative in nature, and we point out experimental evidence that shows that social behavior is not cooperation determined by a set of fixed factors. We argue for a model that sees social behavior as a dynamic interaction of genetic and environmental factors.

Experimental studies on punishment have sometimes been over-interpreted not only for the reasons Guala lists, but also because of a frequent conflation of proximate and ultimate explanatory levels that Guala's review perpetuates. Moreover, for future analyses we may need a clearer classification of different kinds of punishment.

Guala argues that there is a mismatch between most laboratory experiments on costly punishment and behavior in the field. In the lab, experimental designs typically suppress retaliation. The same is true for most theoretical models of the co-evolution of costly punishment and cooperation, which a priori exclude the possibility of defectors punishing cooperators.

The spreading of reputational information about group members through gossip represents a widespread, efficient, and low-cost form of punishment. Research shows that negative arousal states motivate individuals to gossip about the transgressions of group members. By sharing information in this way groups are better able to promote cooperation and maintain social control and order.

It is argued that the generality of strong reciprocity theory (SRT) is limited by the existence of anonymous spontaneous cooperation, maintained in the absence of punishment, despite free-riding. We highlight how individual differences, status, sex, and the legitimacy of non-cooperation need to be examined to increase the internal and ecological validity of SRT experiments and, ultimately, SRT's external validity.

I argue that field (experimental) studies on (costly) peer punishment in social dilemmas face the problem that in equilibrium punishment will be rare and therefore may be hard to observe in the field. I also argue that the behavioral logic uncovered by lab experiments is not fundamentally different from the behavioral logic of cooperation in the field.

The standard theories of cooperation in humans, which depend on repeated interaction and reputation effects among self-regarding agents, are inadequate. Strong reciprocity, a predisposition to participate in costly cooperation and the punishment, fosters cooperation where self-regarding behaviors fail. The effectiveness of socially coordinated punishment depends on individual motivations to participate, which are based on strong reciprocity motives. The relative infrequency of high-cost punishment is a result of the ubiquity of strong reciprocity, not its absence.

We argue that standard experiments supporting the existence of “strong reciprocity” do not represent many cooperative situations outside the laboratory. More representative experiments that incorporate “earned” rather than “windfall” wealth also do not provide evidence for the impact of strong reciprocity on cooperation in contemporary real-life situations or in evolutionary history, supporting the main conclusions of the target article.

The target article misunderstands the research program it criticizes. The work of Boyd, Richerson, Fehr, Gintis, Bowles and their collaborators has long included the theoretical and empirical study of models both with and without diffuse costly punishment. In triaging the situation, we aim to (1) clarify the theoretical landscape, (2) highlight key points of agreement, and (3) suggest a more productive line of debate.

Guala appears to take social preferences for granted in his discussion of reciprocity experiments. While he does not overtly claim that social preferences are only by-products that arise in testing environments, he does assert that whatever they are – and how they evolved – they have little value in the real world. Experiments on animals suggest that social preferences may be unique to humans, supporting the idea that they might play a prominent role in our world.

Costly punishment's scarcity “in the wild” does not belie strong reciprocity theory as Guala claims. In the presence of strong reciprocators, strategic defectors will cooperate and sanctioning will not occur. Accordingly, natural field experiments are necessary to assess a “wide” reading of costly punishment experiments. One such field experiment exists, and it supports the hypothesis that costly punishment promotes cooperation.

The rarity of altruistic punishment in small-scale societies should not be interpreted as evidence that altruistic punishment is not an important determinant of cooperation in general. While it is essential to collect field data on altruistic punishment, this kind of data has limitations. Laboratory experiments can help shed light on the role of altruistic punishment “in the wild.”