As an analyst who identifies strongly with both self psychology and relational psychoanalysis, I feel I am in touch with many of the complex tensions that exist between these two psychoanalytic sensibilities. The relationship between them is difficult to pin down, in part because of the ongoing cross-fertilization, or as relationalists would say, the mutual influence, between the two perspectives. As part of a course I teach at the Chicago Institute on "Relational and Intersubjective Models," I have found it a useful exercise to try to specify the conceptualizations of the unconscious that underlie different models. We all recognize that such conceptualizations have been evolving, explicitly and implicitly, since Freud first conceived the topographical model. But in recent times, with a few exceptions (e.g., Stolorow and Atwood, 1992; Hirsch & Roth, 1995; Davies, 1996), we often have not been rigorous about specifying the assumptions about the unconscious underlying our own preferred theories. Such specificity is important because our models of therapeutic action are so intertwined with our understanding of the unconscious. As examples of the utility of this exercise, consider the following observations that I have found helpful in clarifying the relationship between self-psychological and relational theories, and even between different sub-theories within self psychology.

Over the past century, as our theories of therapeutic action have become more relational so have our models of the unconscious: the two go hand in hand. What began, in Freud's theories, as a model emphasizing interaction between largely endogenous intrapsychic entities and forces has, progressively, evolved into a model emphasizing internalization or appropriation of relational experience. Correspondingly, our theories of therapeutic action have evolved from an exclusive emphasis on interpretation (an intervention originally designed to impact the relationship between endogenous intrapsychic structures or forces) to a progressively greater emphasis on new relational experience, which presumably impacts old internalized relational templates.

It seems to me we have seen this same evolution occur within self psychology over the past 30 years. Kohut's theories of therapeutic action straddled the fence between interpretation and new relational experience. Likewise, the innovations he introduced in the understanding of the unconscious included new relational motivations (longings for selfobject experiences) and more "one-person" intrapsychic concepts such as structural deficits, and conflicts involving unmodulated, endogenously generated, narcissistic fantasies. The evolution of self psychological theory since Kohut can be understood as a progressive "interpersonalization" both of the unconscious and of the theory of therapeutic action. We now understand that the self, in pathology, is not simply understructuralized but is complexly structured by the internalized representations and procedural memories of early relational experience. Accordingly, our theories of therapeutic action have moved in the direction of the provision of new relational experience. It is not that interpretation no longer has a role, but its role is increasingly understood to be one of illuminating unconscious, relationally-derived "organizing principles" and affects, and of, in itself, providing a new relational experience.

Extending this kind of analysis to the dialogue between contemporary relational and self-psychological models, a fundamental difference can be seen in their conceptualizations of the unconscious. The relational unconscious (reflecting its roots in object relations theory) is conceived as a much more active, intrusive, controlling unconscious whose power to evoke both concordant and complementary identifications and to subject the analyst to co-created versions of the patient's internal world cannot be evaded. By contrast (reflecting its roots in American ego psychology), the contemporary self-psychological unconscious is more self-contained: it consists of "organizing principles," disowned affects, "negative selfobject" representations, "accommodations," and selfobject longings and fantasies, all of which control the patient's subjective experience but don't necessarily compel or enable him or her to control the experience of the analyst. Within this model, transference is experienced, but does not aggressively and inherently evoke complementary countertransference in the analyst. The core theories of therapeutic action differ accordingly. The self psychologist's empathic-introspective listening stance, and the interpretations arising from it, are sufficient to grasp, provide a new experience for, and transform the patient's more self-contained unconscious world; whereas the relationalist is always having to work his or her way out of co-created enactments, mutual projections and cross-identifications to provide the transforming needed new experience.

My own view is that these two seemingly irreconcilable views of the unconscious actually represent two constantly occurring aspects of unconscious experience, and that our models of analytic listening and therapeutic action need to take both aspects into account. Our listening stance must be able to hold in tension both a disciplined effort to empathically explore and grasp the patient's psychic reality and, at the same time, an equally empathic ongoing recognition of our own responses as we are affected by the patient. Similarly, our model of therapeutic action needs to hold in balance more empathic and interpretive interventions that seek to convey our understanding of the patient's subjectivity (conscious and unconscious), and more expressive interventions necessary to assert or reclaim our own subjectivity in the midst of reenactments.

This integrated understanding of the unconscious and therapeutic action is useful in advancing the ongoing debate in the literature on the merits of the concept of projective identification - a debate that seems to me to have obscured the fundamental questions that are at stake. If we pose the question less narrowly - that is, not just as a question about whether patients can project their feelings and impulses "into" therapists, as the original Kleinian model suggests - but rather as a question about the plausibility of the unconscious of one person intentionally and aggressively impacting the unconscious of another, it seems to me that the self psychological/intersubjectivist position is revealed to contain residual "one-person psychology" assumptions.

Stolorow and his associates (e.g., Stolorow, Orange, & Atwood, 1998) have been especially vocal and articulate opponents of projective identification. They have argued that belief in the concept amounts to constructing a "cordon sanitaire" around the analyst by making the patient responsible for the analyst's experience. I would argue that it is Stolorow et al. who are constructing a "cordon improbable" around both the patient and the analyst by failing to recognize the tendency of the unconscious of both parties to actively evoke both resonant and complementary responses in the other. Such a capacity is virtually axiomatic in all relational theories. It is also, I would argue, implicit in contemporary theories based in infancy research that emphasize the interaction of self and mutual regulatory processes (e.g., Sander, 1995; Beebe and Lachman, 2002) - theories which have found resonance among both relationally and self-psychologically-oriented analysts. I believe that what has been called projective identification could, at least in certain instances, be reconceptualized as a reenactment in the analytic relationship of early pathological mutual regulations (pathological, because their true aim was to regulate the parent's internal states, not the child's), and that analytic "cure" often involves the transformation of such pathological regulations into new mutual regulations that are truly facilitative of the patient's effective self regulation. Such transformations require the analyst to be empathically attuned both to the patient's effect on the analyst, and the analyst's effect on the patient.

References:

Beebe, B. & Lachmann, F.M. (2002), Infant research and adult treatment: Co-constructing interactions. Hillsdale, NJ: Analytic Press.

Davies, J.M. (1996), Linking the "pre-analytic" with the post classical: Integration, dissociation, and the multiplicity of unconscious process. Contemporary Psychoanalysis, 32, 553-576.

Hirsch, I. & Roth, J. (1995), Changing conceptions of the unconscious. Contemporary Psychoanalysis, 31, 263-276.

Sander, L. (1995), Identity and the experience of specificity of recognition: Commentary on Seligman and Shanok. Psychoanalytic Dialogues, 5, 579-592.

Stolorow, R.D. & Atwood, G.E. (1992), Three realms of the unconscious. In Contexts of Being: The Intersubjective Foundations of Psychological Life. Hillsdale, NJ: Analytic Press (29-40).

Stolorow, R.D., Orange, D.M., & Atwood, G.E. (1998), Projective identification begone! Commentary on paper by Susan H. Sands. Psychoanalytic Dialogues, 7, 719-725.