Abstract

Hamilton introduced two conceptions of social fitness, which he termed neighbour-modulated fitness and inclusive fitness, and he argued that the two concepts are equivalent. This argument relies on two assumptions—actor’s control and additivity—that can be justified as approximations under ‘_δ_-weak selection’, which is selection on tiny differences between the mutant and the wild type. The assumption of _δ_-weak selection stems from a methodological stance that takes cumulative adaptation to be the explanatory target of social evolution research, together with an empirical commitment to a gradualist picture of how cumulative adaptation arises. In a process of gradual, cumulative adaptation, short-term change can be calculated using either fitness concept, but inclusive fitness has a special role to play as the criterion for improvement and the standard for optimality.