No one denies that contemporary human beings cooperate extensively with non-kin. This social behavior sets us apart from even our closest primate relatives, who tend to only display strong cooperative behaviors with kin. But explaining this difference is no easy task: modern culture exerts such a strong influence on our behavior that it is easy to imagine that our propensity to help those to whom we are not related might be a relatively new human innovation. Were our ancestors nepotistic, favoring kin while trying to compete with non-kin? If so, is there a mismatch between our evolved tendencies and the current cultural expectation of cooperation with unrelated acquaintances? Have we been tricked by culture to mistakenly provide mutual aid to our fellow humans?
These are the questions that frame a recent study published in the journal Science. In their article entitled “Co-Residence Patterns in Hunter-Gatherer Societies Show Unique Human Social Structure”, a research team led by Kim R. Hill and Robert S. Walker showed that contemporary hunter-gatherer tribes lack the social structuring necessary for strong kin selection to effect cooperative behavior (Hill et al. 2011). Because of extensive local migration of both males and females, local groups tend to be composed predominantly of non-kin or distant relatives, suggesting that cooperative behaviors had to evolve by mechanisms other than kin selection. Interestingly, the dispersal that breaks down the potential for kin selection within the local group also prevents strict genetically-based group selection from producing cooperation either: in order to understand why these tribes cooperate, one has to appeal to some combination of cultural group selection and genetic group selection at the larger “tribal” level. What is clear is that our current tendency in industrialized societies to cooperate with non-kin is not at odds with the tendencies of our hunter-gatherer contemporaries, suggesting that modern culture is not necessary to foster cooperation with unrelated acquaintances.
This is a really nicely-composed paper and represents some interesting findings that need to be considered, especially by those in the “inclusive fitness can explain all cooperation” camp. But there is a bit of an “Achille’s Heel” here, and that is the source of the data used to understand relatedness. From what I can tell, the database used to conduct this study (the Group Compositions in Band Societies Database) is based on genealogies collected through census. If this is true, we lack a critical piece of information: how these people are actually related. While it is safe to assume some concordance between reported relatedness and actual relatedness, the problem with this study is that it assumes strong concordance. If people are not in fact related genetically in a way that parallels their perceived relationships, this data could be misleading us. I worry that we think that we are studying biological relationships when in fact we are really studying social relationships.
This concern relates to one of the key conclusions of the paper, which validates the theory proposed by Bernard Chapais that monogamous pair-bonding was the key to fostering early human cooperation. I have not read Chapais’ book Primeval Kinship, but the basic argument is that with pair-bonding comes known relatedness and once you know that your relatives have all dispersed into local groups you are more likely to cooperate with those groups than raid and/or kill them. The supposed support that this recent study provides for Chapais’ theory is given particular prominence in an article by Nicholas Wade in the New York Times entitled “New View of How Humans Moved Away From Apes”, and I found it disturbing how uncritically this link was accepted. Why is pair-bonding necessary for this sort of cooperation? And why didn’t pair-bonding and the kin recognition it allowed provide for more local nepotism rather than less? What’s the connection between pair-bonding and the real finding of this study, which is that humans have a unique pattern of dispersal? Isn’t it possible that a multimale/multifemale “promiscuous” mating system could have produced the same kind of cooperation with non-kin? Why assume that pair-bonding is at all causative?
Clearly this study provokes a lot of questions. Actual genetic relatedness data might begin to get at the underlying significance of the clear finding here, which is that our hunter-gatherer ancestors probably could not have relied on kinship to build cooperation.Articles, Cooperation, Cultural Evolution, Group Selection, Human Evolution, Kin Selection, Mismatch theory