Abstract

Duhem is commonly held to have founded his view of history of science as continuous on the ‘metaphsical assertion’ of natural classification. With the help of a strict distinction between formal and material characterization of natural classification I try to show that this imputation is problematic, if not simply incorrect. My analysis opens alternative perspectives on Duhem's talk of continuity, the ideal form of theories, and the rôle of ‘bon sens’; moreover it emphasizes some aspects of Duhem's realism that play an important part in his philosophy of science