In 1964, William Hamilton presented a mechanism for the evolution of altruism, which was perceived by its main promoters as an alternative to explanations based on “group selection,” invoking advantage to the population to account for the evolution of such traits. Less than ten years later, Hamilton used the framework developed by George Price to model the evolution of an altruistic gene in a structured population, a result that has been interpreted as a spectacular conversion to group selection. This paper revisits the modeling research on altruism and the considerable semantic ambiguities concerning the levels of selection in the late 1960s and early 1970s, by studying in close detail the reflections and exchanges among Hamilton, Price, Robert Trivers, and Ilan Eshel. The challenge in this research was not simply to find and model robust mechanisms for the evolution of altruism, but to interpret their properties in unambiguous terms that could be accepted by other researchers. The continuing debate over the levels of selection results from the tension between the properties of the models and the words used in interpreting them.

This content is only available via PDF.