Discussion

It is the nature of work that identifies itself at the juncture of two different areas that have not yet articulated their similarities and differences (to be controversial). It is also the nature of such work to possess enormous potential for moving along one or both areas, and both are the case, I believe, with Professor Fonagy's work. In the brief remarks with which I respond to Professor Fonagy's paper, I will attempt to highlight some of the points of controversy and potential, while steering my remarks toward the spirit of this important series organized by Dr. Baudry, concerning the role and value of longitudinal research for psychoanalysis. In doing so I hope to lay out why longitudinal research is necessary and useful.

Prediction and control, one is taught early on, are the goals of science. But unusual difficulties arise with respect to the study of peoples' lives, as opposed to inanimate forces. One can then take resort in the human sciences as an alternative to the natural sciences, a direction I do not advocate. It is preferable to work as best one can within the constraints imposed by any single method for uncovering antecedent-consequent relationships, with an appreciation of some inherent limitations. In deriving antecedent-consequent relationships, prediction must be distinguished from postdiction, with whose inductions it is often confused. We know about the post hoc, ergo propter hoc logical fallacy, in which everything is seen as clear and inevitable when seen in retrospect, but ill-defined and shadowy when seen in prospect. Thus, one thinks, of course an adult analysand who has suffered a severe trauma during the second year of life becomes