Abstract

Duhem's discussion of physical theories as natural classifications is neither antithetical nor incidental to the main thrust of his philosophy of science. Contrary to what is often supposed, Duhem does not argue that theories are better thought of as economically organizing empirical laws than as providing information concerning the nature of the world. What he is primarily concerned with is the character and justification of the scientific method, not the logical status of theoretical entities. The crucial point to notice is that he took the principle of the autonomy of physics to be of paramount importance and he developed the conception of natural classification in opposition to accounts of physical theories that contravened it.