Strengthening strong reciprocity.

Gintis (2000) presents a model of the evolutionary emergence of strong reciprocity that may in part explain a high level of sociality in human groups despite a low level of relatedness among members. A strong reciprocator is predisposed to cooperate with others and punish defectors at a cost to himself. Gintis (2000) argues that those groups that have strong reciprocators as members might have an evolutionary advantage under conditions of high social instability, where cooperation based on self-regarding reciprocity, socalled reciprocal altruism (Trivers, 1971), will collapse. However, both the argument that leads to Eqn (8), which provides the condition for self-interested agents cooperating, and the derivation that leads to Eqn (16), which provides the condition for an increase in the fraction of strong reciprocators in the population, are inaccurate. Moreover, several important papers have noted that Gintis (2000) assumes without justification that strong reciprocators are obligate cooperators, and have even speculated that strong reciprocity cannot be evolutionarily stable unless ∗Corresponding author. Tel.: 1-413-586-7756; fax: 1-775-402-4921. E-mail addresses: [email protected] (K. Deng), [email protected] (H. Gintis), [email protected] (T. Chu).