Disentangling Social Preferences from Group Selection

Abstract

Explaining human cooperation—and particularly the evolved proximate mechanisms that support it—is one of the most challenging enterprises of modern research, spanning both the natural and social sciences. A Cooperative Species is a serious and thoughtful attempt to give the most complete explanation available today, sophisticated both in the theory and in the empirical evidence gathered to support it. Bowles and Gintis, two well-known economists with wide knowledge in evolutionary biology and committed to the explanatory power of evolutionary models, belong to a group of researchers who in the last 10 years have developed a theory on the uniqueness of human cooperation, based on a strategy they call ‘‘strong reciprocity.’’ Strong reciprocity encompasses two main claims. First, the proximate (psychological) mechanism for human cooperation features social preferences—in particular altruistic motivations. This goes against the conventional wisdom in economics according to which human cooperation results from far-sighted self-interest and Smith’s invisible hand. Explanations based on self-interest founder in the face of incomplete contracts and market failures in interaction structures like the prisoner’s dilemma (PD) and the tragedy of the commons (pp. 5–6). Hardin (1968) proposed punishment as an alternative solution; but if punishers (P) are altruistically motivated (pp. 25–26, 31–32, 91–92, 165), self-interest does not explain cooperation. Social preferences like fairness as well as social emotions prompting cooperative conduct must be posited as a basic human endowment. The real puzzle is to explain how evolution shaped humans this way (p. 6). The other basic claim concerns evolutionary explanation. It follows directly from the view that fairness and moral emotions are biologically costly—altruistic in the biological sense of entailing a lifetime fitness sacrifice, posing a paradox for natural selection. Selection tends to favor nonaltruists (N); therefore, special evolutionary mechanisms are required to sustain behaviors that determine a loss in fitness. The authors solve this problem with a group selection (GS) model of altruistic cooperation. But in GS models the devil is in the details. So their solution comprises at least five presumably interrelated elements: one is clearly a novel and empirically relevant model of selective group extinction (Chap. 7), incorporating empirical estimates of intergroup warfare and of genetic differentiation between ancestral human groups. Given the low values for genetic differentiation (Chap. 6), the benefit-to-cost ratio of altruism must be high. Such high values are found in virtue of the selective extinction of groups through ancestral intergroup warfare, also estimated empirically (Chap. 6). Two further elements seem auxiliary ones, functioning to reduce the reproductive disadvantage of altruists compared to the N: an ancestral institution of food sharing supported in self-regarding preferences and conceived as a biologically selfish behavior (Chap. 7, p. 130); and assortment of altruists with altruists within groups, a phenomenon labeled ‘‘segmentation,’’ probably to highlight the fact that groups are further structured into subgroups that repeat at a lower level the positive assortment at the population level (Chap. A. Rosas (&) Department of Philosophy, Universidad Nacional de Colombia, Bogota, Colombia e-mail: arosasl@unal.edu.co