Epilogue

One ineluctable conclusionemerges clearly from this study: an analyst's theoretical orientation has a marked impact on the way he thinks about patients and the way he works with them. This conclusion raises an equally striking question: how can clinicians who think and behave so differently get equally good analytic results? Before addressing these points let me consider some of the problems inherent in the use of the methodology from which they have arisen.

The private nature of psychoanalytic data has always been the subject of scientific frustration and criticism, most recently by Grünbaum (1984). Spence (1986) clearly demonstrated the inadequacy of most psychoanalytic reports for explanatory purposes, as contrasted with their interpretive use, and made an eloquent plea for the development of more adequate approaches to psychoanalytic reporting. Audio- and videorecording, while a step in the right direction, runs aground on the volume of data that must be explored, a volume unmanageable for many purposes.

The method of data collection used in this study, a brief case presentation followed by process notes taken by the treating analyst at the time of the session, is the backbone of psychoanalytic supervision and has been used since the dawn of psychoanalytic history. It is, of course, essentially different from data obtained via supervision in that the material being studied is presented in written form rather than in person verbally by the analyst. This has the virtue of making the data more publicly available, but at the price of omitting much of the nonverbal affective contents of the analyst's report and rendering impossible closer investigation through questioning about specific aspects of the analytic exchange. Like all methods of presenting psychoanalytic material, this one has its problems. They