Rethinking the Role of Regressive Experience in Psychoanalytic Supervision

In this article I shall try to make the case for the value of acknowledging and working with regressive phenomena that arise within the supervisory relationship. When I speak of regression I mean a range of relatively affectively intense and cognitively primitive experiences of supervisee or supervisor. These may include, for example, dreams about the therapy and the patient, or about the supervision and its participants; intense or conflictual feelings by the supervisee toward the patient (countertransference), or toward the supervisor, or by the supervisor toward the supervisee; any of the many enactments that are generally thought of as evidence of the “parallel process”; or having an experience of altered consciousness, such as a dissociative experience, or a somatic reaction in supervision. These experiences may be understood in terms of regression from more complex to more primitive modes of thinking and feeling, from more mature object relations to more primitive object relations, from more organized self states to more primitive self states. I also include among such phenomena the common supervisee experience of disorganization and “deskilling” (Sarnat, 1997b) that accompanies many student's efforts to give up old patterns in order to learn something new.

I believe that a variety of such experiences on the part of both supervisee and supervisor are relatively common, and occasional examples of all of these phenomena do appear in the literature. But only work with parallel-process phenomena and supervisee countertransference to patients are discussed in any depth in the supervision literature. I view this limitation in the supervision literature as a reflection of most psychoanalytically oriented supervisorsapos; discomfort with and lack of clarity about the best way to include these phenomena in the supervisory conversation. This discomfort leads to the neglect of these phenomena in supervision, resulting in both an impoverishment of the data that is