top of page
Gordon, R.M. (1995) The Symbolic Nature of the Supervisory Relationship: Identification and Professional Growth, Issues In Psychoanalytic Psychology. 17, 2, 154-162.

The Symbolic Nature of the Supervisory Relationship:

Identification and Professional Growth

Robert M. Gordon, Ph.D., ABPP

Allentown, PA 18103

This article is based on my paper presented at the 18th Annual Scientific Conference, “Pitfalls of the Psychoanalytic Supervisory Process: Interpersonal and Self Psychological Perspectives”. It took place at the Washington Square Institute in New York on January 9, 1994.



Identification in supervision is compared to transference in analysis. The concept of identification with the supervisor is seen as important to professional growth. Positive identification is compared to the problems of idealized transference, disidentification and negative transference in supervision. It is difficult to manage the positive identification in supervision when the interventions of the supervisor are timed for the needs of the patient. This can be a narcissistic injury for the supervisee, and be confusing when teaching proper timing. The supervisor needs to be aware of the symbolic nature of the supervisory role to the supervisee’s unconscious and discuss the feelings within the relationship without doing therapy. The positive identification with the supervisor helps the therapist grow as a professional and enjoy the work.

Just as transference is important to analysis, so is identification important to supervision. In analysis, transference is interpreted for personal growth. In supervision, identification is part of the nonverbal process of learning a new role. It is often that someone explains how they do therapy, or arrange their practice, or set up their office in a particular fashion by stating that that was how it was done by their analyst or supervisor.

The symbolic role of the analyst serves as an opportunity to observe and rework the patient’s unconscious through mainly transference. The analyst will represent various aspects of the patient’s psychopathology. Solnit (1970) felt that the supervisory situation promotes an identification rather than a transference and regression. I see it more a matter of degree, with identifications and transferences occurring in both settings. The roles and tasks of the supervisor and supervisee limit the range of transference, and limit the degree of regression. The supervisor by being a reality based teacher of skills triggers the process of identification to facilitate learning. This promotes more identification than transference in supervision.

Example: A young therapist told his analyst that he sees him as the one who will heal him, but fears that the analyst will discover that he is unlovable and will reject him. In contrast he told his analyst that he sees his supervisor, as a wise teacher leading into the world of the profession. “He’s the parent to my teenage self, and you’re the parent to my child self. I don’t want him to heal me, I want him to show me the magic so I can be successful like him.” The analyst, asked if the same fears might not repeat themselves in that relationship. The analysand responded, “This relationship is more emotional, that relationship (the supervision) I only risk showing the therapist side of me.”

The above example illustrates that the analysand viewed the analyst from a position of more regression and transference than the supervisor who was more idealized and more an object of identification.

The identifications in supervision are important for the learning of tasks, skills and role behaviors. The transferences that do occur in supervision are discussed as an exception to the rule versus being the main focus as in analysis. The common exceptions would be when the supervisee’s transferences on to the supervisor are interfering with their working relationship, work with the supervisee’s patients, or the supervisee’s own analysis.

Example: A supervisee commented that her own analyst is too gratifying. Although the supervisor agreed, she remained silent. The supervisee stated that perhaps she can go into analysis with the supervisor. The supervisor discovered that the supervisee did not share these feelings with her analyst, fearing she would hurt his feelings. Sensing an idealized transference from the supervisee rather than just a positive identification, the supervisor encouraged the supervisee to bring up the issue in analysis. The supervisor then drew allusions to the supervisee’s own work, with her patients’ idealizations and devaluations. The supervisor stated that the supervisee may actually find supervision gratifying her transference wishes, due to the supportive nature of the relationship and getting answers to her questions. The supervisor explored whether the supervisee is imitating both her therapist and supervisor in being too supportive versus uncovering. The supervisee discussed a case where she was too supportive of her patient, rather than interpreting her patient’s idealizing transference. The supervisee then stated that she picked a therapist that wouldn’t confront her.

The above example is of an idealized transference rather than a positive identification. This transference is interpreted since it effects both her own analysis and work with patients. The genetics of it are not explored.

An idealized transference to the supervisor is not a form of learning, but rather can become a resistance to it. However, a positive identification with the supervisor is an important part of the learning process.

Compton’s (1985) review of the literature on identification distinguishes it from transference as being more adaptively selective, less tied to drives and less ambivalent, than transference. Identification is part of a learning process that can be unconscious or consciously employed. Identification can promote growth and adaptation.

On the contrary transference is a form of resistance, in which the past is not remembered. Fenichel (1945) defined it as “....... ... the patient misunderstands the present in terms of the past ... He ‘transfers’ the past attitude on to the present.” (p.29). With identification one wishes to be like the object; with transference there is a distortion of the object instead of remembering the past.

Generally, the supervisee at some level has the wish for the supervisor to idealize and admire, to protect, replenish, encourage and to teach the magics of the profession. If the analyst is the “mother”, mastering the world of emotions, then the supervisor is the “father”, leading into the world of work.

Baudry (1993) wrote that, “The role of the supervisor is analogous to that of the parent for a young child. In an earlier version of this paper, I had written ‘father’ instead of ‘parent’, but the term is both sexist and limiting.” I disagree on two of Baudry’s points. I see the supervisory role like that of a good parent figure not to a young child, but rather a young apprentice entering a guild. My use of the term “father” is not limited to (though most often associated with) a gender. I mean I as an unconscious role that acts as a catalyst to development. The symbolic father figure pulls the neophyte from the familiar environment, prepares and toughens the neophyte for the demands of the work world.

Support is only in service to the task. The mothering role, is for supporting and nurturing the self. Both roles are necessary for ego development. Supervision should have elements of both, but mainly more fathering than mothering.

The symbolic role of the analyst allows for an entry into the unconscious. The symbolic role of the supervisor allows for an entry into the profession. This latter role involves both more support and gratification, and more untimely confrontations than would be appropriate in the treatment milieu.

Of the three main components of psychoanalytic training: training analysis, didactic curriculum and supervision, Schlesinger (1990) concluded that, “... the main educational tool or procedure for learning to become an analyst is supervision.” This is because, according to Schlesinger, the supervisor must not only teach, but make an “educational diagnosis”. This “educational diagnosis” includes assessing whether an error is due to a “dumb spot” or a “blind spot” that is, due to ignorance or a product of neurosis, or both.

This dual role of teaching to overcome what Schlesinger calls “dumb spots” or ignorance, and interpreting “blind spots” that are about very personal issues, makes supervision the most complex of the three components of analytic training. It is also what distinguishes psychoanalytic supervision from the supervision in the non-analytic therapies, in that it includes discussion of the supervisee’s own issues. This will effect the identification with the supervisor. The supervisee may not see the supervisor as a good object, but as a punishing object. The supervisee may in turn take this identification into his or her own work with patients and become too critical with them.

In addition to the conflict about identification, there is also a conflict about timing and imitation. When trying to both teach and react to the supervisee’s errors, there is an implicit double message in supervision about timing. The supervisor must time the interventions to the supervisee based on the needs of the patient, not based on how ready the supervisee is to hear the intervention. While the supervisor is teaching about proper timing, quite the opposite is modeled in the supervision.

Example: A therapist presented a case of her patient who began to focus only on his symptoms, after a session where the therapist was overly confrontational with the patient. The therapist complained about the patient, much like the patient complained about his symptoms, without insight. The supervisor reminded the supervisee that in the previous session the supervisor had confronted her with her blind spot. After some denial, the supervisee admitted that she felt exposed and embarrassed.

The supervisor explained that it is confusing to teach good timing, and then confront a supervisee in a manner that would not occur in treating a patient. The supervisor explained that the timing was based on the needs of the patient, not the needs of the


The Supervisory Relationship 6

supervisee. It was explained that the patient worked on the manifest level of symptom complaints as a defense, after being confronted in a manner similar to what the supervisee experienced in supervision. The supervisor warned “don’t use my supervision timing, as therapy timing. My supervision timing is geared for therapists, not patients.”

The supervisor noted that the supervisee reacted in kind to her patient, being overly confrontational with her patient after she was confronted in supervision. The supervisor did not interpret the “parallel process” (when the supervisee enacts the symptoms and defenses of his or her patient in supervision, see Arlow 1963). The supervisor did not interpret the dynamics of the supervisee concerning the narcissistic injury, imitation of the critical object, or perhaps identification with the aggressor.

Rather, the supervisor explored the supervisee’s reactions and feelings and then clarified the confusing message about proper timing. The supervisor was also careful to reestablish the positive identification with the comment about the timing geared for therapists, not patients, to help the supervisee not feel like a patient, but as a resilient professional who can take the rigors of training.

As an aside, I have noticed that there is much more interest in the literature on “parallel process” than in identification in supervision. It seems that supervisors may find it easier to notice when their supervisees imitate their patients negative attributes than when they imitate their supervisors negative attributes.

Langs (1979) contrasts the usual teaching ambience with that of the supervision of psychotherapy. The usual teaching situation calls for a safe atmosphere for the openness of expression and support, gearing the level and rate of learning to the learning capacities of the student. However, Langs points out four qualifying factors with learning in supervision: 1) the situation is geared to the needs of not one but two, the patient and the supervisee; 2) both the teaching and learning about analysis and psychotherapy more than other learning situations occurs under the intense influence of the unconscious and personal factors often related to countertransference; 3) the supervision of psychotherapy often involves the exposure of very personal issues
in the supervisee, including his or her psychopathology, which lends a-special charge to the supervisory experience, and; 4) the learning of therapy can not simply unfold in a linear fashion from simple to complex, but rather the subject matter is complex and interrelated.

Langs adds that since the prime consideration is for the patient, not the supervisee, there will be times for a supervisory intervention based on patient need that will evoke conflict and discomfort for the supervisee. “The immediacy and frequent urgency of the therapeutic situation therefore exerts a major influence on the supervisory experience, and this consideration will sometimes necessitate teaching communications that might otherwise have been kept for a moment more propitious for the student.” (p.378).

What supervision does in effect, is have the supervisee offer for examination highly personal examples of work, hoping for admiration, in a frame much looser but similar to therapy. However, the supervisors reactions are geared for the good of the patient, with the defenses and the narcissism of the supervisee suffering a blow.

This is indeed a complex situation for the supervisor concerned with the needs of the patient and the needs of the supervisee. Just as transference needs to be protected in analysis, so does identification need to be protected in supervision.

A way to resolve this conflict is to maintain a balance of needs within the supervisory role. The supervisor needs be the symbolic good but demanding object. The supervisor can’t just be the father, but must at times be the mother to help the supervisee recover from narcissistic injuries common in supervision.

In analysis, the analyst must at some point interpret the idealization by the patient as a defense against aggression as in drive-conflict theory, or as an expression of a developmental task as in Kohutian theory. Either way, it is a road into the unconscious psychopathology. In supervision however, a positive identification is not for the sake of working through psychopathology, but to serve as good model to help the supervisee adjust, cope, master, and enjoy the profession.

In both therapy and supervision the interventions are sometimes seen as attacks which threaten self esteem. In both therapy and supervision, exploring the reaction to the interventions are important. The difference is that in analytic therapies, these reactions to the therapist's interventions are central to the analysis of the transference. In supervision, the supervisors understanding of the supervisory relationship is important, and that the main goal is to reestablish the supportive and optimal teaching aspects of the relationship. This need not be deeply explored with the supervisee, but must be managed by the supervisor. The relationship should be explored if it is negatively effecting the supervisee’s learning, analysis or work with patients. My main point is that, in supervision the supervisee needs to maintain the supervisor as a good model for
identification, not for the sake of personal growth, but for the sake of professional growth.

The supervisee’s identification is based on the supervisee’s transferences and perceptions of the supervisor. It is not based on an empathic understanding of the real person of the supervisor. In fact some distortion in the direction of idealization is probably necessary for identification. A supervisee may see a relaxed supervisor in session and wish to be easy going like her. When in fact, the supervisor may not be an easy going individual. Nevertheless, the real person of the supervisor needs to convey integrity and incorruptible ethics that are integral to a therapeutic relationship (Gordon, 1993).

Langs (1979) lists a hierarchy of indications for supervisory interventions, which emphasizes my point of the importance of the symbolic role that the supervisor plays for the supervisee, and when this relationship needs to be discussed in supervision. Although it is not Langs’ point that the supervisor is needed as a good model to the supervisee, he puts the relationship at the top of his hierarchy. For Langs the order of precedence is: “indications of countertransference in the supervisor or of a crisis in the supervision, signs of countertransference in the supervisee and of a crisis in the therapy, any major framework issue in either situation, and, last, the appearance of material in the therapeutic interaction that calls for teaching and discussion.” (p.364). This order, with it’s emphasis of dealing with what is going on in the supervisory relationship, seems to be a logical way out of the inherent contradictions in supervision,
and allows the positive identification to continue.

Example: A therapist trained in a non-analytic therapy model sought supervision in order to help fulfill her board requirements for supervision. She was also interested in learning more about psychoanalytic psychotherapy. Her pervious supervisory experience was purely a matter of learning techniques. She had no analytic therapy of her own. The supervisor set up the supervisory sessions with similar, but more relaxed ground rules than therapy. The supervisor started and ended on time. This was a problem for the therapist, who was frequently late for supervision, and often
started late and ran over time with her patients, The therapist readily grasped the psychoanalytic concepts and interventions but unconsciously discouraged her patients' work with transference. She-wanted to be the good object rescuer saving patients from their symptoms. She feared their unconscious material as she did her own. After a few months of supervision, the supervisee gave a reality justification for having to discontinue her supervision. The supervisor chose to interpret, since the supervisory relationship was now threatened with dissolution. The supervisor stated that patients will prematurely terminate their work with plausible reality demands and that it is to be expected as a defense against the uncovering of frightening unconscious material. The supervisor told the supervisee that it is extremely difficult to tolerate analytic supervision without benefit of analytic therapy.

The supervisor stated that the supervision like therapy, will stir up issues and resistances will build up if these issues are not interpreted. However, in supervision they are not resolved therapeutically, so it's understandable to unconsciously try to regulate the anxiety by finding some diplomatic way to leave supervision. The supervisee then said that she was in fact feeling too stirred up with her own issues in supervision. The supervisor suggested that the supervisee seek analytic therapy and continue with the supervision. The supervisor impressed the supervisee with the necessity for both in order to do deep work and to grow as a professional.

In this example the supervisor needed to interpret the supervisee’'s behavior since the working alliance was threatened by the supervisee's own defenses. The supervisor was very careful to discuss but yet protect the defenses of the supervisee and not delve into the genetics. The supervisor made supportive comments about the understandable stress of analytic supervision without therapy and offered a way to resolve it. The suggestion to seek analytic therapy was done in the context of professional growth, not psychopathology. This allows the supervisee to keep the supervisor as a good object of identification, and still be confronted.

If there exists the tendency to identify for the sake of learning, then there also exists the tendency to disidentify for the sake of resistance. The supervisees who need to disidentify, view the supervision as something to survive. Very little learning takes place, because they are often too grandiose to be taught. They often get into power struggles with their patients, and find the work stressful. Intense negative transferences on to the supervisor are more openly hostile and are indicative of even more psychopathology than disidentification. These situations must be interpreted in supervision, but can only be possibly resolved in analysis. If these issues cannot be resolved in treatment, then such individuals should be discouraged from being in the field.

Without a supervisor that served as an object of positive identification, the therapist lacks a model of an internalized professional guide and soother. Such a therapist, is more likely to feel the stress of the profession, and perhaps act out with patients, than one that had a positive supervisory experience with a good role model.

The process of identification in, supervision as an important nonverbal form of learning the professional role needs to be better understand and appreciated. The more the process of identification is understood, conflicts and contradictions in supervision can better be avoided or resolved.


Arlow, J.L. (1963). The supervisory situation. J. Amer. Psychoanal. Assn., 11, 576-594.

Compton, A. (1985). The concept of identification in the work of Freud, Ferenczi, and Abraham: a review and commentary. Psychoanalytic Quarterly, 54, 200-233.

Baudry, F.D. (1993). The personal dimension and management of the supervisory situation with a special note on the parallel process. Psychoanalytic Quarterly, LXII, 588-614.

Fenichel, 0. (1945). The Psychoanalytic Theory of Neurosis. New York: Norton.

Gordon, R.M. (1993). Ethics based on protection of the transference. Issues in Psychoanalytic Psychology, 15.2, 95-105.

Langs, R. (1979). The Supervisory Experience. Jason Aronson, New York.

Schlesinger, H.J. (1990). Supervision or training analysis: repetition or collaboration. 135-140. In Tradition and Innovation in Psychoanalytic Education. Clark Conference on Psychoanalytic Training for Psychologists. Edited by Murray Meisels and Ester R. Shapiro. Lawrence Erlbaum Assoc. Hillsdale, N.J.

Solnit, A.J.(1970). Learning from psychoanalytic supervision. Int. J. Psychoanal., 51, 359-415.

Back to Publications

bottom of page