Self Psychology Psychoanalysis

Sorting Out the Maze of "Self Psychologies"

Crayton E. Rowe, Jr., M.S.W., B.C.D.-P

Crayton Rowe is the founder of the American Association for Psychoanalysis in Clinical Social Work. He is a founding member of The New York Institute for Psychoanalytic Self Psychology; and member of the International Psychoanalytical Association. He is a contributor to volumes in the Progress in Self Psychology series, the author of numerous psychoanalytic papers, and co-author of Empathic Attunement: The "Technique" of Psychoanalytic Self Psychology. He is in private practice in New York City.

This paper was presented at the 2002 Annual Conference of The New York Institute for Psychoanalytic Self Psychology. It may be quoted provided that the author, Crayton E. Rowe, Jr., M.S.W., B.C.D., and this website, www.selfpsychologypsychoanalysis.org, are cited as the source.

      Kohut (1984) was clear that his discoveries were an extension of analytic understanding rather than a deviation from traditional theory:

Self psychology does not advocate a change in the essence of analytic technique. The transferences are allowed to unfold and their analysis—the understanding of the transference reactions, their explanation in dynamic and genetic terms—occupies, now as before, the center of the analyst's attention [p. 208].

      He (1984) added: "Self psychology is at one with the technical principle that interpretation in general, and the interpretation of transferences in particular, is the major instrumentality of therapeutic psychoanalysis" (p. 210).

      Kohut (1977) arrived at his theoretical formulations through an empathic approach that he described as an "experience-near" mode of observation. It was a "commitment to a methodology of the observer's long-term empathic immersion in the psychological field—in particular, with regard to clinical phenomena, of his long-term, empathic immersion in the transference" (pp. xxi-xxii). This process is an intersubjective data gathering process, a continuous feedback loop, whereby the patient experiences the analyst's attunement. The analyst simultaneously experiences the patient's experience and so on.

      Within this intersubjective process, the analyst communicates the sense of what the patient is experiencing for the patient's validation. Kohut calls this long-term immersion process the understanding phase. It is simply capturing and communicating the nuances, blends, and intensities of what the patient is experiencing.

      As this long-term immersion process continues, the analyst and patient slowly acquire meaning of the evolving data in dynamic and genetic terms. Kohut calls this second phase the explaining or interpretation phase in which both analyst and patient mutually participate.

      As is well known, it was through this empathic process that, beginning with Ms. F., Kohut discovered the needs of his patients which he conceptualized as selfobject needs and selfobject transferences.

      Kohut's ability to suspend his theoretical convictions enabled him to perceive data more clearly from the patient's vantage point and, as a result, he was able to come to his conceptualization of "selfobject" as an impersonal experience of a function that is provided. His discovery was poignantly described (Kohut, 1971) in his work with Miss F.:

Thus I refused to entertain the possibility that I was not an object for the patient, not an amalgam with the patient's childhood loves and hatreds, but only, as I reluctantly came to see, an impersonal function, without significance except insofar as it related to the kingdom of her own remobilized narcissistic grandeur and exhibitionism [p. 288].

      Kohut's definition of selfobject as the experience of the functions provided was further clarified in a letter (Kohut, 1980) to a colleague: "Whether cognitively insufficiently separated from the self, as may be the case in infancy and in certain archaic merger states later in life, or, as is most often the case, in particular in the older child and in the adult, cognitively separated from it, the selfobject is defined by our experience of its function (p. 671; italics added).

      Kohut returned to his definition of selfobject as the experience of functions provided in his posthumously published work, How Does Analysis Cure? (Kohut, 1984). Here he considered the general meaning of selfobject to be "that dimension of our experience of another person that relates to the person's functions in shoring up our self " (pp. 49-50; italics added) and the specific meaning as relating to the earliest developmental experiences that are carried forward and into the present.

      In summary, Kohut consistently taught that the concepts of selfobject referred to the patient's experiences of functions provided within the analytic milieu. He was careful to distinguish that these theoretical formulations explained an individual's psychological and not physical reality.

      As the selfobject transferences unfold, the analyst must be able to differentiate their various forms in order to understand the specific developmental needs of the patient. Kohut (1984) stated:

The ability to differentiate between various types of selfobject transferences also gives us the opportunity to study in greater detail the developmental line characteristically associated with the archaic form of a particular self-selfobject relationship—from the archaic state that is revived at the beginning of the transference to the mature state which, as a result of the systematic and patiently pursued working-through process, may be attained at the end of successful analyses [p. 202].

      Kohut thought that much work remained to be done, especially in the area of exploring special needs of adolescents and the elderly as they face changing life circumstances. He (Kohut, 1984) acknowledged that varieties of selfobject transferences are yet to be discovered: "that there are still transferences—probably varieties of selfobject transferences—that have not yet been discovered and which, therefore, remain unanalyzed" (p. 209).

       Although Kohut left a legacy rich in directions for exploration, there has been an exodus of those who were to carry out these explorations. In fact, individuals and organizations representing self psychology have embraced disparate theoretical treatment modalities that are in obvious opposition to Kohut's foundational constructs that gave rise to the birth of self psychology and are the theoretical underpinnings that define self psychology. For example, at an international conference on the psychology of the self held in November 2001, there were six distinctly different theoretical perspectives featured: Boston Research Group on Change, Specificity Theory in Self Psychology, Relational Perspective of Stephen Mitchell, Self Psychology Forward Edge Perspective, Control Mastery Perspective, and Nonlinear Dynamic Systems Theory Perspective.

      Other obvious disparate theories embraced as self psychology that were not specifically featured at this conference are: intersubjective, optimal responsiveness, intimate attachment, motivational systems theory, and the dual transference model.

      In earlier articles (Rowe, 1994, 1996, 1999) I presented my concerns that Kohut's fundamental concepts were being redefined as an effort at integration in a growing number of theoretically disparate treatment models.

      These redefinitions have radically changed the meaning of Kohut's fundamental constructs and have pervaded the disparate treatment models that are embraced as streams of self psychology.

      Of greatest concern is the attempt to redefine the concept of selfobject and selfobject transference as object relations concepts. Reformulations were directed at shifting the meaning of selfobject as the experience of the function that is provided by the analyst to the experience of the analyst as the provider of the function. This object relations conceptualization of selfobject is in direct contradiction to Kohut's discovery of selfobject in his work with Ms. F. that he was an impersonal function and not an object.

      This reformulation assumes that the direction of the patient's experience necessarily is to the object/analyst, and by its assumption disallows an unrestricted direction-free attunement to the patient's experience that is necessary to perceive, understand, and allow for the development of the selfobject transferences.

      The reformulation of the concept of selfobject was clearly seen in Bacal's (1990) statement: "Self psychological theory has, in effect, focused on the experience of selfobject function, and has lost sight of the object that provides that function and of the importance of the relationship for the patient. Thus, I would add to Stolorow's caveat that the selfobject should be conceived of as a dimension of experiencing an object, that this experience is also embodied in a significant object" (p. 202). Bacal was referring to Stolorow's (1986) statement: "Thus, when we use the term 'selfobject,' we refer to an object experienced subjectively as serving selfobject functions" (p. 274).

      A redefinition of the meaning of selfobject and selfobject transferences is also apparent in motivational and intersubjective theory. For example, Lichtenberg, Lachmann, and Fossage (1992), 1996) evolved a theory of psychoanalytic technique that redefines the meaning of selfobject transferences as conceptualized by Kohut to that of selfobject experiences as aspects of what Lichtenberg (1989) has termed the attachment motivational system. Lichtenberg et al. (1992) clearly state that "mirroring, twinship, and idealization are central to the development of the attachment motivational system" (p. 136).

      The proposed theory of technique includes the concepts of motivational systems earlier formulated by Lichtenberg (1989):

  1. the need for psychic regulation of physiological requirements,
  2. the need for attachment and later affiliation,
  3. the need for exploration and assertion,
  4. the need to react aversively through antagonism or withdrawal (or both),
  5. and
  6. the need for sensual enjoyment and sexual excitement.

      Redefining the meaning of selfobject transferences as aspects of an attachment motivational system voids Kohut's meaning of selfobject and selfobject transferences as unique developmental needs that are the primary focus of analysis.

      From an intersubjective point of view, selfobject transferences are only one of many unconscious ways of organizing activity. For example, Trop (1994) states that they represent "one of a multiplicity of unconscious, automatic, and repetitive ways that patients organize their experience of the analyst. Selfobject transferences, in other words, are a class of invariant organizing principles" (p. 79).

      The patient comes to identify and understand the unconscious organizing principles as they emerge in the treatment. Successful psychoanalytic treatment is seen as facilitating the patient's establishment and consolidation of alternative organizing principles through new relational experiences with the analyst and through awareness (Stolorow and Atwood, 1992).

      Efforts have also been made to limit the focus on the selfobject transferences by including a dualistic treatment approach that focuses on object relations (Stolorow and Lachmann, 1981; Shane and Shane, 1988).

      The reformulation and redefinition of Kohut's fundamental constructs have necessarily led to a discounting of the importance of in-depth immersion into the patient's experience that led to Kohut's discovery of selfobjects. In short, there is no reason to pursue the patient's evolving unrequited selfobject needs if one does not believe they exist. In-depth focus on the patient's experience is pejoratively considered to be a "one-person psychology" in that the analyst's own subjectivity is not explored with the patient. This concept of treatment is taken primarily from relational theory (Mitchell, 1988; Aron, 1991) and is based on object relations theory, which does not recognize Kohut's discovery of selfobjects. It follows that those who hold to relational concepts would not comprehend the importance of the analyst's focus on the patient's experience to determine the patient's selfobject needs.

      Similarly, it is no surprise that Kohut's (1977) clinical observation of the spontaneous unfolding of selfobject transferences and the actualizing of inherent potentials or "designs" (p. 241) has been refuted (Stolorow, 1993, 1994). For example, Stolorow (1994) states: "Lastly, I share Gill's skepticism about the idea that the self possesses an innate nuclear program or inherent design awaiting a responsive milieu that will enable it to unfold" (p. 225).1 To accept Kohut's findings that one's potential can develop as the selfobject transferences develop would require an openness to the existence of unrequited selfobject needs and an awareness of the importance of in-depth attunement to allow for their discovery, understanding, and development in the treatment.

      Also, it is not surprising that as a result of the lack of recognition of the existence of the patient's evolving selfobject needs, the concept of optimal frustration, as being essential to understanding the depth of the patient's trauma surrounding the emergence of these selfobject needs, has been denied. Replacement concepts (e.g., "optimal responsiveness") have been developed that shift the analyst's focus away from the patient's experience to an object-related focus to the analyst as provider (Bacal, 1985; Terman, 1988).2

      In addition to the redefining and dismissal of Kohut's individual concepts, there is a parallel effort by a major self psychology organization to define self psychology within another theoretical model. This organization recently announced that it is planning "a series of programs to define self psychology within a relational psychoanalytic world."

      I am curious as to how theorists and major self-psychological organizations now seem to welcome the replacement of the fundamental constructs that have defined self psychology. This welcoming is a phenomenon unprecedented in the history of the development of other treatment models. Can you imagine those who support the Freudian drive model disowning Freud's concept of drives; or those who support ego psychology accepting a redefinition of Freud's and Anna Freud's concept of the ego; or object relations theorists disowning the concept of object representations? I, for one, cannot, as it is reasonable to think that those who believe in the validity of the fundamental constructs underlying a treatment model would wish to preserve and not disown their distinctions.

       Perhaps there is some recognition of the elimination of Kohut's concepts in that there is a minority voice within the maze of "self psychologies" that lays claim to the terms "mainstream," "traditional," or "classical" self psychology. Rather than preserving self psychology's fundamental constructs, however, these terms invite further confusion as their attempts at differentiation implicitly recognize other treatment models as being self psychology even though these models have dismissed the fundamental concepts that have defined self psychology.

       It is difficult to imagine these terms being applied to other treatment theories to differentiate differences from within their own knowledge base because that knowledge base discards its fundamental constructs. For example, "I am a traditional Freudian because, unlike other Freudians, I believe in drives." "I am a mainstream ego psychologist because unlike other ego psychologists, I believe in the structural concept of an ego." Or, "I am a classical object relations theorist because, unlike other object relations theorists, I believe in the concept of object representations."

      While such terms applied to these treatment models are obviously contradictory, unacceptable, and oxymoronic, not so when they are applied to self psychology.

      The attempts to redefine the contributions made by another theorist is unprecedented in any field. Do you know of any author blatantly claiming the title of another's work and then changing the storyline? I do not think that even copyright laws are prepared for this, as it is usual for those who claim the work of others to keep the storyline and change the title. Why change the storyline of Kohut's concept of selfobject and keep the title, selfobject? Since these redefinitions of Kohut's contribution of selfobject and selfobject transferences are new formulations, why not call them by other names?

      But how can this phenomenon of redefining Kohut's fundamental concepts be explained? Has Kohut made such a profound impact on psychoanalysis that theorists who have redefined his concepts fear to consider their redefinitions as being outside of self psychology? Is this why they retain the term selfobject—to keep an identity with self psychology? Or is it that the magnitude of Kohut's influence on psychoanalysis and the threat his theories pose to other theories require that his concepts be diminished? After all, Kohut's concepts of selfobject and selfobject transferences dramatically change how we listen to and understand our patients.

      One might quickly answer that Kohut poses no threat, that theorists have simply expanded on Kohut's ideas. Indeed, it is well known that worthwhile theories lead to the development of other theories. Certainly, Freudian theory led to the development of ego psychology and object relations theory.

       That theorists have simply expanded on Kohut's ideas may at least partially explain how self psychology has embraced disparate theories. If one accepts that a redefinition of a theory is a development of a theory, then it follows that one would believe that those modalities that have redefined Kohut's concepts belong within self psychology. It must be pointed out, however, that we cannot equate redefinition of the basic constructs of a theory with the development of that theory. Development of theory does not include its redefinition. Rather, it occurs when those who discard, preserve, or add to elements of a particular theory develop it without changing its fundamental meaning. Bergmann (1993) sees these groups as heretics, modifiers, and extenders. Heretics discard and fall outside of psychoanalysis; modifiers are considered to preserve concepts, and their contributions evolve from the concept; extenders expand psychoanalytic concepts without any modification of the concept.

      Goldberg (1984/1985) also sees three distinct groups in the development of psychoanalytic theory. He believes as Bergmann does that redefining fundamental concepts of a theory is not considered to be development of the theory. Goldberg classifies these groups as integrationists, compromisers, and anarchists. The integrationists attempt to bring divergent concepts together through translation. Compromisers believe there can be an inclusive umbrella theory. Anarchists emphasize the incompatibility of divergent theories.

      Even in the early development of Freudian theory, no one attempted to redefine Freud's fundamental constructs, in contrast to the redefinition of Kohut's fundamental contributions. For example, in the well-known heated battles that Freud (1914) had with Adler and Jung that led to two movements away from psychoanalysis, there was no thought of redefining Freud's concepts.

      In the former case, Freud was concerned that Adler was attempting to reify his concept of "organ inferiority" and "masculine protest" as a dominating force in character formation and neurosis. The battle was not over a redefinition of Freud's concept of character formation and neurosis.

       In the latter case, the question was not one of Jung's redefining Freud's fundamental contribution of neurosis. Rather, Freud was concerned that Jung was pushing the sexual factor into the background in the formation of neuroses.

      It should be emphasized that it is not my intention in this paper to evaluate the merits of these aforementioned disparate models of treatment. Instead, it is my concern that there has been an unprecedented redefinition of Kohut's fundamental findings and that the redefinition is are being accepted as self psychology.

      In conclusion, Kohut was clear that self psychology is an extension of psychoanalysis that embodies his fundamental contributions of selfobject and selfobject transferences. Treatment modalities that dismiss and/or redefine these fundamental contributions cannot be considered self psychology.


1.   Here, Stolorow agrees with Gill's (1994) objection to what he considers Kohut's positivist or objectivist view as contrasted with the social constructivist view that disavows consideration of innate proclivities or capacities that can be at least partially attained in treatment. The social constructivist view of treatment assumes that the psychic reality of the analyst and patient as well as the assessment of previous interactions is newly constructed in the present.

2.   See MacIsaac's (1996) excellent discussion of how Bacal's (1985) concept of optimal responsiveness shifts the analyst's focus away from the patient's experience.


Aron, L. (1991), The patient's experience of the analyst's subjectivity. Psychoanal. Dial. 1:29-51.

Bacal, H. (1985), Optimal responsiveness and the therapeutic process. Progress in Self Psychology, Vol.1, ed. A. Goldberg.  Hillsdale, NJ:  The Analytic Press, pp. 202-227.

——— (1990), Does an object relations theory exist in self psychology? Psychoanalytic Inq., 10:197-220.

Bergmann, M. (1993), Reflections on the history of psychoanalysis. J. Amer. Psychoanal. Assn., 41:929-955.

Freud, S. (1914), On the history of the psychoanalytic movement.  Standard Edition, 14:7-66. London: Hogarth Press, 1957.

Gill, M. (1994), Heinz Kohut's self psychology. A Decade of Progress: Progress in Self Psychology, Vol. 10, ed. A. Goldberg.  Hillsdale, NJ:  The Analytic Press, pp. 197-211.

Goldberg, A. (1984/1985), Translation between psychoanalytic theories.  The Annual of Psychoanalysis, 12/13:121-135. New York: International Universities Press.

Kohut, H. (1971), The Analysis of the Self. New York: International Universities Press.

——— (1977), The Restoration of the Self.  New York: International Universities Press.

——— (1980), Letter, July 24, 1980.  The Search for the Self, Vol. 4, ed. P. Ornstein. Madison, CT:  International Universities Press.

——— (1984), How Does Analysis Cure? ed. A. Goldberg & P. Stepansky.  Chicago: University of Chicago Press.

Lichtenberg, J. (1989), Psychoanalysis and Motivation. Hillsdale, NJ: The Analytic Press.

———, Lachmann, F.  &  Fosshage, J. (1992), Self and Motivational Systems: Toward a Theory of Psychoanalytic Technique. Hillsdale, NJ: The Analytic Press.

———, ———  & ——— (1996), The Clinical Exchange: Techniques Derived from Self and Motivational Systems. Hillsdale, NJ: The Analytic Press.

MacIsaac, D. (1996), Optimal frustration: An endangered concept. Basic Ideas Reconsidered: Progress in Self Psychology, Vol. 12, ed. A. Goldberg. Hillsdale, NJ: The Analytic Press, pp. 3-16.

Mitchell, S. (1988), Relational Concepts in Psychoanalysis. Cambridge, MA: Harvard University Press.

Rowe, C. (1994), Reformulations of the concept of selfobject: A misalliance of self psychology with object relations theory. A Decade of Progress: Progress in Self Psychology, Vol. 10, ed. A. Goldberg. Hillsdale, NJ: The Analytic Press, pp. 9-20.

——— (1996), Introduction: Notes on the integration, reformulation, and development of Kohut's contributions. Basic Ideas Reconsidered: Progress in Self Psychology, Vol. 12, ed. A. Goldberg. Hillsdale, NJ: The Analytic Press, pp. xiii-xxii.

——— (1999), The selfobject transferences reconsidered. Pluralism in Self Psychology: Progress in Self Psychology, Vol. 15, ed. A. Goldberg. Hillsdale, NJ: The Analytic Press, pp. 15-30.

Shane, M. & Shane, E. (1988), Pathways to integration: Adding to the self psychology model. Learning from Kohut: Progress in Self Psychology, Vol. 4, ed. A. Goldberg. Hillsdale, NJ: The Analytic Press, pp. 71-78.

Stolorow, R. (1986), On experiencing an object: A multidimensional perspective. Progress in Self Psychology, Vol. 2, ed. A. Goldberg. Hillsdale, NJ: The Analytic Press, pp. 273-279.

——— (1993), Thoughts on the nature and therapeutic action of psychoanalytic interpretation. The Widening Scope of Self Psychology: Progress in Self Psychology, Vol. 9, ed. A. Goldberg. Hillsdale, NJ: The Analytic Press, pp. 31-43.

——— (1994), Kohut, Gill, and the new psychoanalytic paradigm. A Decade of Progress: Progress in Self Psychology, Vol. 10, ed. A. Goldberg.  Hillsdale, NJ: The Analytic Press, pp. 221-226.

——— & Atwood, G. (1992), Contexts of Being: The Intersubjective Foundations of Psychological Life. Hillsdale, NJ: The Analytic Press.

——— & Lachmann, F. (1981), Two psychoanalyses or one? The Psychoanal. Rev., 68:307-319.

Terman, D. (1988), Optimum frustration: Structuralization and the therapeutic process. Learning from Kohut: Progress in Self Psyclf Psychology, Vol. 4, ed. A. Goldberg. Hillsdale, NJ: The Analytic Press, pp. 113-125.

Trop, J. (1994), Self psychology and intersubjective theory. The Intersubjective Perspective, ed. R. Stolorow, G. Atwood & B. Brandchaft. Northvale, NJ: Aronson, pp. 77-91.