Evolution of human cooperation
Two papers this week help explain why humans cooperate, even with nonrelatives. Cooperation with relatives (activities that tend to decrease one's own reproductive success, while increasing that of others likely to share many of one's genes) is predicted by "selfish gene" theory, as formalized in Hamilton's rule. I've assumed that cooperation with nonrelatives is a beneficial side-effect of behavioral genes that evolved when most of our neighbors were relatives, as is still the case in parts of the Amazon and West Virginia. But other explanations have been proposed.
One hypothesis is that "human cooperation evolved as a result of high levels of lethal competition (i.e. warfare) between genetically differentiated groups." In other words, some groups of unrelated individuals happened, by chance, to have a higher fraction of individuals whose genes tended to increase within-group cooperation -- particularly, willingness to risk injury in battles with other groups -- and the overall frequency of those genes increased as victorious groups killed groups that happened to have a lower frequency of "cooperation genes." This process would tend to be undermined by within-group evolution (assuming selfish individuals tend to have more descendants) and by migration between groups. The latter could include abductions.
But are genetic differences between groups big enough for this "group selection" mechanism to work? In "Genetic differentiation and the evolution of cooperation in chimpanzees and humans", recently published in Proceedings of the Royal Society, Kevin Langergraber and colleagues compared genetic differences among competing aboriginal human groups with differences among competing chimpanzee groups. They found that genetic differences among chimpanzee troops were at least as great as differences among human groups. So, if humans are more cooperative than chimps -- budget deadlocks in the US Congress and my state legislature call this into question -- it's probably not because group selection is more effective in humans than in chimps, in increasing the frequency of genes favoring cooperation. The authors suggest that "both genetic and cultural differentiation between groups played a role in the evolution of altruistic cooperation."
What sorts of cultural differences among groups might be important? This week's second paper, recently published in PNAS by Sarah Mathew and Robert Boyd, claims that "Punishment sustains large-scale cooperation in prestate warfare." This paper addresses the issue mentioned briefly above, the problem of within-group increases in selfishness (if selfish individuals have more descendants) undermining increases in cooperation from between-group processes. In particular, they asked whether individuals that deserted during battles between groups were punished.
The researchers obtained data for 88 raids involving the Turkana tribe, and found that: 1) the chance of a man being killed is >1% for each raid he participates in, 2) desertion or acts of cowardice occur in at least 45% of raids, and 3) these acts often lead to severe beatings, after group discussion. Getting beaten could certainly cause an individual to change his behavior, but what effect, if any, do such sanctions have on the frequency of genes that affect willingness to take risks in battle? Do those who fight bravely end up with more wives and, more important, more descendants? Unfortunately, "distinguishing the effect of behavior during warfare from the effect of other factors that affect a person's value as a social or mating partner is beyond the scope of the present study."