Criteria for Termination

In the course of considering the topic of this article, long-term analyses, I found as I began to contemplate any one aspect of the issue, the subject of psychoanalytic treatment unraveled before my eyes. Consider the manifest topic, long-term treatments; it immediately brings up the issues of termination, analyzability, and embedded in both these motifs is the question of the goal(s) of psychoanalysis. Once these questions are raised, I find myself plunged into a variety of other themes that seem impossible to exclude: the role of gratification in an analysis and the danger that Rose has pointed out, that is, analysis interminable and interminable “analysts” (Rose, 1974). Although all of these questions loom in front of us the subject that to me is surprisingly most compelling lies in the beginning of the analysis, more specifically in a process that I have termed analytic trust (Ellman, 1991). It is the way in which the analyst enters the patient's world that is a crucial determinant of how the analysand will respond to interpretative efforts, and eventually will be one of the most important determinants of how the treatment will end.

In this article, I first briefly highlight some conceptual difficulties that are present in the literature on termination. I begin by looking at Freud's ideas on when to end a analysis, and jump to Brenner, Novick, and Kohut. I then present two clinical vignettes that illustrate some of these difficulties, and end with a discussion of what I will call the cycle of analytic trust. I posit that it is the completion of this cycle that signals the end of the analysis. In the course of the article, I offer partial definitions of what I mean by analytic trust.

The course of psychoanalytic debate continues to be profoundly and penetratingly influenced by both Freud's concepts and psychoanalytic practices. It is my view that some of Freud's difficulties in conceptual consistency and in following his own dictates in th