Sociologists have often looked to the natural and physical sciences for theoretical and methodological models. Marx was interested in developing social laws analogous to those in physics, and Durkheim based many of his core theoretical ideas on an analogy between social systems and biological organisms. Among classical sociologists, the strongest dissenting voice came from Weber, who argued that the social sciences required distinct theories and methods since they were studying intentional action.

Like many metatheoretical issues, this debate between the founders of the discipline has been replayed often in contemporary sociology, yet there are still strong disagreements about the extent to which social scientists can learn useful lessons from the natural and physical sciences. One recent attempt to use the natural sciences as a model for the social sciences comes from historical sociology, where several prominent scholars have argued that evolutionary biology could provide useful guidelines for understanding the relationship between history and theory in explanation.

This paper has two main goals. First, we want to outline the structure of explanation in evolutionary biology, and in the process clear up some misconceptions in the sociological literature. These misconceptions arise from two sources: (1) historical sociologists have derived their views of evolutionary biology almost exclusively from one source, Stephen J. Gould, who does not represent the mainstream of the discipline; and (2) many historical sociologists have misread Gould's work, and have incorrectly depicted him as opposed to the use of general theory and traditional scientific methods.

After outlining the way explanation in evolutionary biology works, we return to the question of whether we can learn lessons from them that will improve historical explanations in sociology. We argue that there are useful lessons, concerning both what theory can do (focusing on the contrast between explanation and prediction, and the importance of constructing abstract typologies) and on the ways in which historical effects can be modeled (concentrating on what evolutionary biologists call 'developmental constraints' and 'genetic drift').