David Sloan Wilson is an evolutionary biologist who studies a broad range of subjects relevant to humans in addition to nonhuman species. He is best known for his work on multilevel selection theory in which altruism and other prosocial behaviors evolve by benefiting whole groups, despite being selectively disadvantageous within groups. He has published in philosophy, psychology and anthropology journals in addition to his mainstream biological research. With philosopher Elliott Sober, he is author of Unto Others: The Evolution and Psychology of Unselfish Behavior (Harvard University Press, 1998).

The concept of altruism has had a turbulent history in evolutionary theory. Darwin had a clear conception of how altruistic behaviors can evolve (the first rise). Although altruists are less fit than nonaltruists within their own group, groups of altruists are more fit than groups of nonaltruists. Thus, altruism can be explained as a group-level adaptation that requires a corresponding process of group selection to evolve.

Group selection was rejected as an important evolutionary force in the 1960’s and was replaced by theories that seemed to explain altruism in individualistic terms, ultimately as the product of “selfish genes” (the first fall). The overarching metaphor of selfishness was hailed as a great advance but subsequent developments have shown it to be a massive wrong turn from which the field is only starting to recover. Multilevel selection (MLS) theory is back (the second rise), with two profound implications for evolutionary theory. First, the theories that were regarded as alternatives to group selection, such as kin selection, reciprocity, and selfish gene theory, are nothing of the sort. They assume the existence of groups and prosocial behaviors evolve by between-group selection, just as Darwin envisioned. Thus, MLS theory represents a synthesis and unification of previously discrepant theories. Second, not only does MLS theory include its so-called alternatives as special cases, but it goes further to explain other cases that previously were difficult to imagine. In particular, human groups emerge as important units of selection, despite the fact that their members are genetically unrelated and do not always base their behavior on expected return benefits.

Ironically, the same theory of multilevel selection that led to the first and second rise of altruism also leads to its second fall. Altruism is only one of several mechanisms that can evolve to increase the fitness of groups. Other mechanisms, which often appear more coercive and can be lumped under the term “social control”, are more efficient because they structure the activities of the group without great individual sacrifice. Thus, the case for groups as adaptive units is becoming increasingly secure while the case for altruism, as opposed to social control, remains tenuous. My talk will focus on this problem, which I regard as the current frontier of evolutionary research on altruism.

There are several reasons why altruism should remain a central concept in MLS theory and an essential feature of adaptive human groups (the third rise). First, it is very important to see social control as a form of altruism within the framework of MLS theory, even though the self-sacrificial component of the altruism is often very small. Second, high-cost altruism can evolve when altruists can segregate themselves sufficiently from nonaltruists. Plausible segregation mechanisms exist and have been demonstrated in social psychology experiments. Third, social control mechanisms can actually promote, rather than replace, strongly altruistic mechanisms. The prediction of symbiotic relationship between altruism and social control is empirically testable.