September 25, 2002. I thank Auchincloss and Vaughan for clarification of their position. I concur with their assertion that serious errors may occur, with dire consequences for our field's credibility, when developmental assumptions are made solely on the basis of consulting room data from adult patients. This is where child analysis can make a significant contribution. Children in analysis demonstrate a remarkable capacity for introspection when interpretations follow and integrate the free associations of their fantasy play. This allows the child analyst to observe the inner world of the child as it emerges. Of course, child analysts must beware of the opposite error: that is, assumptions about later outcome made on the basis of developmental data alone.

I appreciate Auchincloss and Vaughan's making clear that we share a conviction about psychoanalytic theory building. Credible evolution of theory can occur only when psychoanalytic observations from the consulting room are integrated with research data generated in all relevant scientific domains.