Peter Wolff Responds

July 7, 1998. I would like to thank Howard Shevrin for his thoughtful and challenging comments. As he correctly points out, the underlying theoretical implications of my essay went considerably beyond what I was prepared to discuss at that time and in that context. To remedy the shortcoming, I will focus on the broader theoretical issues as succinctly as possible.

If I understand him correctly, there is a fundamental difference in our perceptions of what psychoanalysis is and is not. It seems to me that that difference boils down to the questions (1) whether psychoanalysis is or is not, a natural science “like any other natural science”; and (2) whether the validity of the “talking cure” does or does not depend on its natural science status. Had I assumed that psychoanalysis is a natural science, or that it has all the essential ingredients for becoming one, then a number of my arguments concerning the irrelevance of infant observations would be irrelevant; in that case, I would be inclined to agree with Shevrin that the fundamental psychoanalytic assumptions about the mind (the unconscious, repression) cannot be tested empirically “from the couch.”

However, I started from the assumption, implicit, that psychoanalysis is not a natural science (or for that matter any other kind of science); though the epistemological status of the psychoanalytic method may be ambiguous, I deem that ambiguity eminently worthy of serious study. Some commentators on the natural science status of psychoanalysis have recommended patience without premature closure, as it is still a very young science attempting to solve very complex issues of mental life. However, debates on the question whether psychoanalysis is natural science have gone on for over a hundred years, and there is no evidence that the debate has moved very far, if at all, from its original formulation. Nor do the results of almost fifty years of programmatic empirical research on various hypotheses derived from psychoanalytic