Coevolution of cooperation, causal cognition and mindreading

The evolution of cooperation between unrelated individuals has long been a puzzle in evolutionary biology. Formal models show that reciprocal altruism is approximately as stable as kin-based altruism when cooperators can assort. Why, then, is reciprocal altruism so rare? We suggest that the key lies in the difficulty of assortment based on underlying intentions: if individuals are able to reliably detect others' cooperative intent then cooperation is stable, but detecting intentions is notoriously difficult, especially when there are incentives to deceive. For this reason, we suggest, there is likely to be a coevolutionary relationship between human cooperativeness and our skills of social causal cognition: it is not a coincidence that we are both extraordinarily social, cooperating with non-kin to a degree not seen in other species, and extraordinarily good at inferring others' beliefs, intentions, and motivations, a skill sometimes known as mindreading. We discuss results of a recent study that provides evidence for this coevolutionary view of cooperation and social cognition.