Commentary

Robert Michels's skillful and incisive discussion of the history of the psychoanalytic case history exposes, in the manner in which we are accustomed to expect from him, the still chaotic state of our intentions when seeking to draw conclusions from what we tell each other we do.

Michels's review is exemplary and complements an earlier effort by Widlöcher (1994), alongside which I found it interesting to read—partly as an interesting comparison of French and American views of the issues. Widlöcher, interestingly, as he occupied Charcot's position at the Salpêtriére, focuses less on the issue of how far clinical reporting can “prove” anything and more on the issue of how clinicians can learn from experience. The difference in emphasis is relevant to Widlöcher's academic post. It has been argued that it was upon Freud's visit to Charcot at the Salpêtriére that he learned to value and interpret clinical data in a way rather at odds with the approach in the German-speaking, physiologically based medicine of the day (Schwartz 1999, p. 38). As Freud (1893) reported, Charcot said, “La theorie c'est bon, mais ça n'empeche pas d'exister. [Theory is good, but it doesn't prevent the existence of other things]” (p. 13). The tension between the difficulty of using “mere” clinical experience to advance knowledge and the necessity of doing so seems to me to live on and to be beautifully caught in Michels's address, as it is in current uncertainties about how psychoanalysts should establish confidence in their ideas. However, my main aim in this commentary is to take up four of the main points Michels has raised and to consider them a bit further.

Science. The first issue is the question of how we regard clinical data. Michels demonstrates the severe difficulty in drawing any secure conclusions from the clinical material we have accumulated so far in our discipline. He raises the question of what can constitute “scientific” data and how the various “selections” typically involved