In science, no less than in other contexts, we often accept claims as true or well supported on grounds of authority. It is simply impossible to evaluate for oneself the evidence for all, or even most, of what scientists claim to know. In such cases, however, if our reasons for accepting a given hypothesis or theory are to be considered good scientific reasons, they must be based on good grounds for believing that the theory's acceptance by the authority in question was itself warranted in the light of good scientific methodology.

Many people accept Freudian psychoanalytic theory because Freud did or for the same sorts of reasons that Freud did. But were Freud's reasons good ones? I shall argue that they were not. From the point of view of at least one dominant current of thought in philosophy of science, the research methodology which Freud employed and on the basis of which he took his