Abstract:
In an insightful article published in this journal, Alan Musgrave (1981) has argued that once we distinguish between three kinds of assumptions, Milton Friedman's F-twist--according to which the truth of the assumptions of an economic theory is irrelevant to its acceptability--can be untwisted. It is shown that once we bring in more clarity on the formal identity of the different kinds of assumptions, Musgrave's contribution can be further refined: distinctions can be drawn between detectability and negligibility assumptions, and between domain and applicability assumptions; the suggested gap between as-if assumptions and other kinds can be removed; another type, that of joint negligibility assumption, can be introduced; the untwisting of the F-twist can be accepted in all cases except in the case of early-step (heuristic) assumptions; and the art of paraphrase is shown to constitute the basis for Musgrave's strategy and to have limits. Copyright 2000 by WWZ and Helbing & Lichtenhahn Verlag AG