Commentaries

When someone of the distinction of Peter Wolff addresses a body of research, all those who are intellectually committed to that field had better take notice. As a great admirer of Wolff's contributions over the years, it is a special pleasure for me to comment on

the arguments in his paper. His critique echoes that of other psychoanalysts (most particularly André Green, 1995, but see also Fajardo, 1993) and is therefore appropriately designated as a target article for this issue of the Journal.

The status of developmental propositions in the definition of psychoanalysis. When Sandler and Joffe (1969) summarized the core assumptions of the basic psychoanalytic model, they included the following: (1) psychic determinism; (2) the pleasure-unpleasure principle; (3) the nature of the biological organism; (4) the dynamic unconscious; and (5) genetic-developmental propositions. The last of these items is the idea that all behaviors are understandable as sequences of actions developing out of earlier (or even earliest infantile) events. It is undoubtedly the case that Freud's view throughout his life was that pathology recapitulated ontogeny (see, e.g., Freud, 1905). Fenichel (1945), for example, stated explicitly that psychoanalytic understanding of early development had substantially advanced as a byproduct of studies of the psychoses, particularly schizophrenia. It should not surprise us therefore that Freud recognized child observation as a legitimate epistemic strategy. His report on the “fort-da” game, as part of the developmental account of repetition compulsion, is perhaps his best-known example (Freud, 1920). For Freud, and almost all psychoanalysts who followed him, there existed a tacit assumption of an isomorphism between pathology and development that permits bidirectional causal inference from childhood to pathology and vice versa.

We are now all too aware of the weakness of the classical genetic-developmental proposition. Hartmann's caution (1950, 1955) that