If one is interested in explaining both the social dynamics and the genetics of human punishment, the everyday behaviors of Late-Pleistocene type foragers are of special interest, even though their ethnographic description is neither complete nor even consistent. There exist some remarkable and strong social central tendencies among the 150-plus documented foraging societies that qualify as “Late-Pleistocene appropriate” (LPA), for these people cluster in mobile egalitarian bands, form moral communities, condemn and punish predatory behaviors like bullying and cheating, and actively favor altruistic cooperation (Boehm Reference Boehm2000; Reference Boehm2008). Two important social goals are: (1) keeping political life egalitarian and (2) promotion of cooperation; and their methods range from shaming, to ostracism, to capital punishment.

In terms of assessing punishment's costs and possible second-order free-rider problems, capital punishment is of special interest because deviants are likely to resist being killed; furthermore, they may be avenged. My coded data on 50 LPA societies (see Boehm, in press) reveal patterns that complement Guala's analysis (see Table 1).

Table 1. Capital Punishment in 50 LPA Foraging Societies

At present there is an Inuit and Australian Aborigine bias in this sample, which may be skewing the data somewhat in favor of sorcery. More generally, the data will be inherently “incomplete” due to reticence, because colonial administrations punish indigenously legitimate executions as murder (see Lee Reference Lee1979); but, with that caveat, the main targeted deviant pattern involves forceful personalities that go against the egalitarian grain, while (with much smaller numbers) devious predators and sexual malefactors seem to come in second. I would suggest that over the millennia all LPA foragers have been executing serious deviants (see also, Otterbein Reference Otterbein1986) on a rare but significant basis, and that the main culprits have been would-be dominators such as sorcerers or serial killers.

Mobile foragers live in groups averaging 20–25 persons, which are composed largely of nonrelatives or distant relatives (Hill et al. Reference Hill, Walker, Božičević, Eder, Headland, Hewlett, Hurtado, Marlowe, Wiessner and Wood2011) but often contain pairs of siblings. There is sometimes the possibility that a male may avenge a close kinsman's death even if the victim was a major social deviant (e.g., Boehm Reference Boehm2011; van den Steenhoven Reference van den Steenhoven1962), so potentially the risks in using capital punishment were high unless the problems of predictable resistance and possible retaliation could be coped with.

In another analysis that concentrates on LPA foragers' methods of social control, in a smaller sample of 10 societies I discovered that in six of them capital punishment was done by delegating a close kinsman to kill the culprit by ambush (see also Woodburn Reference Woodburn1982), while in six (mainly overlapping) societies in the same sample ethnographers reported merely that “the group” killed the culprit.

Often it is impossible to tell whether these group actions involved collective killing or delegation, but out of 22 Bushman homicide cases there is one well-described account of an assassination attempt that turns collective; first a man hits a serial killer with poisoned arrows when he is awake, then the latter wounds a woman and kills her husband, and finally the entire group attacks him as he is dying of the poison (Lee Reference Lee1979). The impression is of a group agreement that the man must go, but of great inefficiency. In a tribal example of a highly efficient communal execution, it is clear that the point of everybody participating is to avoid retaliation by making it impossible to determine who actually killed the culprit (Boehm 1986).

Risks are greatly reduced when a band delegates a close male kinsman of the target to do the job: first, because he will ambush the deviant in his sleep; and second, because otherwise predictable lethal retaliation will be set aside because of the kin tie. But a remaining evolutionary paradox is that the delegated executioner faces costs (the slight risk of the target fighting back, and the definite loss of a close male kinsman), while the rest of the band (non-kin and distant kin) can be seen as free-riders who benefit substantially but pay no costs.

However, there is another way setting up the genetic cost/benefit analysis. The executioner who pays such costs is merely caught in a structural position, in which he becomes the chosen executioner because he is close kin, whereas the free-rider roles of those who abstain are also determined by social position. Thus, free-rider genes are not at issue because the free-riding is determined situationally.

In this light, we may reconsider the ethnographically well-described Mbuti case Guala cites from Turnbull (1961). Cephu cheats on a meat-acquisition system which is designed to bring in a fair share of game for all the participating families; and, collectively, most of the band actively shames him in ways that are humiliating while Cephu's loyal followers stand aside – but do not actively back him. This too is situational, because they are kin. It is worth noting that the sanctioning goes beyond shaming when one band member threatens the arrogant Cephu with ejection from the band; but he is taking little risk because the people backing him are in a state of moral outrage.

I emphasize that the several families associated closely with Cephu likely would be conventionally modeled as free-riding defectors because they stand aside; and also, that in fact this is not a matter of opportunistic free-rider genes in action. It is simply a situational matter, and over the millennia such stepping aside has had nothing to do with genes. In such contexts, the free-rider problem does not apply.

Guala has opened up some interesting questions, and has used ethnographic data in doing so. Perhaps these further ethnographic nuances may serve as useful food for thought, for scholars who use experiments with students (or non-LPA nonliterates) to try to understand human nature.