Kangqiao Wang

There is an ongoing debate over the explanatory value of Hamilton’s rule. Among its various formulations, a general form derived from regression models—Hamilton’s rule–general (HRG)—has attracted particular attention because of its generality. While some authors, particularly Jonathan Birch, regard HRG as playing a distinctive unifying or organizing role in social evolution research, others have questioned its usefulness and explanatory significance. Drawing on the fact that different regression models give rise to different general forms of Hamilton’s rule, I introduce a model-selection perspective to the debate over explanatory value. From this perspective, proponents can be seen as defending a general preference for HRG, whereas critics highlight particular cases where HRG is unfit. Through concrete examples, I argue that a constructive way forward is to consider the model-selection problem: selecting the optimal general form of Hamilton’s rule relative to the given scenario. This problem can be operationalized using familiar statistical tools, and more importantly, it can help reveal the essence of the given scenario and enhance our understanding of social evolution in a different way.