Working With the Split Transference

Modern Psychoanalysis, Vol. 3, No. 2, 1978 (P. 217 – 232)

In attempting to resolve the resistance of the patient to freely experience and verbalize all thoughts and feelings, the author recognized that resolution may depend on the patient forming alternative transference alliances. This may be described as splitting the transference.

The technique of allowing for a therapeutic split transference finds its roots in the theoretical concept of splitting.

When splitting occurs within the transference relationship, it is seen as resistance to allowing all thoughts and feelings to develop towards the single object of the analyst. Siding with or joining a preoedipal resistance is sometimes an effective means of resolving the resistance. This is done when the psychoanalyst agrees that a relationship with a single therapeutic agent simply may not be desirable for the patient. If the intensity of the relationship is too threatening, some patients may defend against developing a transference or may feel the need for two therapeutic agents into good and bad mothers, or split one analyst in separate, unconnected parts.

Psychoanalysts often work with patients who feel that they need to maintain pre-existing relationships with therapeutic agents, or who need to find additional therapeutic agents. Often, the additional therapeutic agent is another analyst (see Ormont and Strean, 1978, for an explication of splitting the transference between group and individual analysts), but institutions, organized groups, other healing practitioners – even religion – may be used as well.

The literature suggests that both cancer (Brody, 1977-78; J. Unger to be published; LeShan, 1977; Spotnitz and Meadow, 1976) and alcoholism (R. Unger, 1978) are diseases with pre-oedipal origins in which highly charged feelings are directed toward or against the self or outside objects. In these diseases, permitting or encouraging the patient to use additional therapeutic agents, aids in maintaining systemic equilibrium.

The preoedipal origins of these diseases is revealed in developmental patterns which are seen to be typical for the patients. Studies (LeShan, 1978) show that victims of cancer often have a history of inadequate mothering, where the mother was experienced as cold, unavailable, or inaccessible. When this pattern of loss of the loved object is recreated in later life, the trauma can serve as the catalyst for the development of the disease. The obsessive orality in alcoholism, is also evident. The alcoholic is seen as an overly dependent person, defending against the recognition of his dependency needs by making his primary love object an inanimate bottle. The addiction to the bottle represents the alcoholic’s attempt to recapture the early pleasurable experience of feeding from the mother (Spotnitz and Meadow, 1976). The disturbance in the early mother/child relationship may be seen in the behavior of the alcoholic, which reveals fixation on the early oral level. The pattern of drinking (or, the impulse to drink in non-drinking alcoholics who have learned to curb acting on their impulses) reflects repetitive and incessant seeking for an experience which will finally gratify the need for satiation.

The intense oral dependency, which is sometimes combined with an oral-sadism, makes these patients demanding and difficult to treat. Other therapeutic agents typically used in addition to the analyst are the medical doctor, for cancer, and Alcoholics Anonymous (AA) for the alcoholic.


In the treatment of cancer the use of medical therapeutic agents is usually crucial. The cancer patient has a life-threatening disease which does not allow the psychoanalyst the time to move along comfortably analyzing the patient’s resistances and waiting for the development of the transference neurosis. The cancer patient working with a psychoanalyst and an oncologist will develop a complex set of transference reactions.

The transference relationship a cancer patient forms with his physician has its own peculiar characteristics because of the nature of the treatment. All such treatment – surgery, chemotherapy, and radiation – involves insult and strain to the body. Administering chemotherapy and radiation requires the art of balancing one poison against another, in hopes that the poison of the drug or X-ray does not prove more destructive to the body than the poison of the disease. The resilience of the body may make a one-shot surgical intervention less damaging than long term treatments, but the body pays for submitting to a radical technique.

The physiological damage incurred through orthodox medical treatment of cancer is well understood. Where cure is effected, it is because the natural resources of the body have been strong enough to throw off the toxins from both the treatment and the disease.
The evidence that some people are cured by poisons, while others are not, suggests a resistance mechanism similar to that found in psychoanalysis. In our psychoanalytic work we generally think of resistance as a mental operation, whose resolution is accompanied by a change in attitudes or beliefs. Success in treating somatic symptoms in preoedipal disorders suggests however, that body resistances are amenable to psychoanalytic technique.

Psychoanalysis resolves the resistance through the transference relationship. In establishing a relationship with the analyst, old conflicts and fixations can be reactivated and resolved, and the patient can proceed to new developmental levels. So, too, in the medical treatment of cancer, resolution of the resistance that has become organically manifest, may depend on the relationship the patient forms with the physician. The relationship is especially problematic: the physician is administering toxic medication to the patient, and, while the patient may willingly agree to the medication, his body responds as though it has been poisoned. Typical side-effects of cancer-fighting medical treatments are nausea, dizziness, weakness, loss of hair, disturbances of the gastro-intestinal system, and interference with the immunological response.

Despite the fact that the physician fills the patient with toxic material, the patient is not helped to discharge the toxicity. The patient generally keeps from conscious knowledge any awareness of being fed poisons. He acknowledges only gratitude for the physician’s persistence in treatment and professional skill in applying medically accepted techniques. The patient does not acknowledge feelings of aggression towards the object – the physician – who treats him with toxic substances. If these feelings remain unconscious, the relationship to the physician may, in time, become as toxic to the patient as the treatment itself.
Standard medical treatment of cancer, then, creates a situation which is potentially psychologically harmful to the patient. Awareness is dual. The conscious agreement to ingest a poison violated the body’s awareness. The cancer patient who is able to tolerate conflicting data integrates the separate awarenesses. The cancer patient who needs to maintain a good/bad object split is unable to integrate the fact that a good doctor is giving bad medicine. Failure to address inadequate integration may result in the patient assuming a posture of pseudo-integration.

The recognition of the potential destructive effect of inadequate discharge of toxic material has led modern analysts to formulate techniques for the release of frustration-aggression, and for eventual immunization from the toxic material. (Spotnitz, 1976).
With the cancer patient who is in analytic and medical treatment, there is a ready-made split transference. The poisonous substances that the physician provides makes him an easy target on which to project the image of bad-object. Whole analytic sessions may be spent with the patient reviling the physician. It may be that understanding the patient experiences from the analyst makes the relationship with the bad-object tolerable.


B is married and has four children. Her husband is an emotionally distant man, and the only emotional closeness within the family is between B and her youngest daughter, E.

B completed a classical analysis when she was first married. The analysis focused mainly on B’s understanding her feelings for her mother. Through treatment, B was able to produce early memories of being afraid that her mother hated her and wanted to kill her. The analysis produced no change, however, in her habitual way of dealing with this painful relationship. She continued to remove herself emotionally from her mother, and their infrequent communication was unpleasant. During the course of treatment, B developed an intense relationship with an older woman who served as a substitute good mother. Although this relationship was more gratifying than her primary relationship with her mother, there were still frustrations. The friend was no always available and did not seem to love B as much as B love her. B’s excessive dependency, however, prevented her from experiencing feelings of aggression. Instead of being angry about the friend’s inaccessibility, B became overly available and overly giving.

Years later at the age of 50, B was diagnosed with breast cancer. Treatment consisted of surgical removal of the breast. Shortly after B’s recovery from the operation, E, the youngest daughter, graduated from college and returned home to live. Her return was especially problematic for the mother/daughter relationship. B became too much concerned about her daughter’s whereabouts; E resented the intrusion after four years of living away from home. Both B and E began analysis began analysis. Since B’s former analyst had moved, she began treatment with Dr. S, a different analyst. Treatment terminated almost immediately, as a result of a confrontation with the analyst. B told the analyst that she didn’t get the feeling that the analyst liked her. The analyst replied, “Why do I have to like you to treat you?” B experienced this intervention as a rejection. She compared the coldness and “business-like manner” of her new analyst with the warmth and love of her former analyst. She did not want to be in treatment with an analyst who did not love her, and terminated.

Eight years later an X-ray revealed that B’s cancer has metastasized. This recurrence coincided with an emotional trauma B was undergoing in relation to her mother-substitute friend. The friend had recently remarried and was now entirely unavailable to B. B and E repaired their relationship, and the daughter took an active part in decisions about treatment approaches to the cancer. They agreed on a three-pronged approach: B re-entered analysis with Dr. S, followed a strict dietary program designed for cancer patients, and took a hormone inhibitor prescribed by her physicians.

Before long, conflict arose over the different treatment modes. B had been a patient of her physicians for 30 years and had confidence in them. She felt she would not die as long as they treated her. She sought their permission and approval for that part of her nutritional support program having to do with the use of laetrile (an apricot pit extract said to have anti-carcinogenic properties). None of her physicians endorsed it: two of them considered it useless; the third, who supplied the hormone medication, refused to treat her if she took laetrile.

B continued all three treatment modes. She learned to lie to her physicians, presenting herself as a dutiful patient. Several years after initiating treatment, she became convinced of he harmful effects of X-rays, which were being used for diagnostic purposes, and was even able to conveniently “forget” to take the recommended bi-annual X-rays. During this time, favorite topics of conversation between B and her daughter centered around how little physicians know about nutrition, how rigid physicians are, and how insensitive to the psychological effects of disease.

The analysis dealt with B’s feeling that she should do what others wanted her to do. A chronic pattern of depressed “house sitting” was reversed. B began lessons in tap dancing, acting, and tennis. A turning point occurred when her fear of being a bad child was activated in the transference. B had trouble paying for treatment on time. The analyst insisted on timely payment, and threatened to terminate treatment if this condition were not fulfilled. In the ensuing sessions, B was able to verbalize her intense feelings of rejection and fear of the psychoanalyst’s anger. She recognized that her own behavior aroused anxiety and fears of rejection. To date, there are no signs of the spread of the cancer. Moreover, X-rays taken before B discontinued that procedure, showed actual regeneration of previously deteriorated bone tissue.


B’s shifting attributions of good and bad to the same objects, a splitting, enabled her to maintain her longstanding relationships with medical doctors cast as all good objects. Her conviction that no harm would come to her as long as she was in treatment with them suggests primitive idealization (Kernberg, 1975).

The classical analysis was unsuccessful in bringing these feelings into the transference. Thus the defense against experiencing aggressive feelings towards external objects while maintaining the relationship remained a primary defense. Either she experienced feelings of hurt and fled the relationship, as with her mother and Dr. S, or she denied the feelings and the denial allowed her to continue the relationship, as with her mother-figure friend. The recurrence of cancer led to dramatic shifts in attribution of good and bad qualities to objects. Her friend’s marriage reawakened feelings of abandonment and worthlessness originally felt in relation to her mother. Rather than risk destruction of the object through her aggressive feelings, B resorted to the regressive narcissistic defense of self-destruction (depression and, speculatively, cancer).

E’s willingness to serve as a therapeutic agent replaced the love object lost when her friend married. B experienced feelings of closeness, the feeling of being taken care of, and the feeling that another person was willing to expend considerable time and energy meeting her needs. With the arousal of these feelings, the shift of attribution from bad to goo as regards E became easy. Objects associated with E, Dr. S, and also proponents of non-toxic cancer treatments became good too. The medical doctors who refused to approve the treatment plan E had suggested became the all bad objects, and B distanced herself from them by lying.
B had functioned all her life in a pseudo-cooperative way, while harboring unconscious hostility towards those whom she saw as restricting her freedom. It was only when a split transference was available that aggression could be discharge in psychoanalytic sessions. When B was able to transfer the bad-object images onto her therapeutic agents, she felt justified in rebelling against their wishes and admonitions. At first she rebelled against Dr. S by terminating treatment. Later however, when the impulse to rebel against her physicians was aroused, her feelings of dependency on them for medication made termination impossible.
Good and bad objects, then, were clearly separated. Resolution of the split occurred when B experienced negative feelings for both Dr. S and the medical doctors, while simultaneously experiencing feelings of dependency. The dependency prevented her from acting on her impulse to flee from the bad object. B became able to tolerate her aggressive feelings towards the love object, and aggressive feelings directed towards her from the loved object, and still maintain the relationship. Good and bad in herself and in the loved object became acceptable.


Some patients come into analysis with a pre-existing loyalty to a religious or philosophical system. That system may be embodied in an actual institution like the Catholic Church or may be an abstract idea like belief in life-after-death. Private patients who have the option of choosing their analysts will often choose one who seems to share their philosophical views. Clinic patients, on the other hand, have no choice, and the analyst may have difficulty in the development of a narcissistic transference.

The issue of boundary definition becomes especially important if the transference is shared with a philosophical or religious system. When the boundary definition takes the form of a commitment to an outside system, the analyst carefully monitors his communications about that system. The patient who proudly boasts of his identification with a system may be saying, as well, that he values his uniqueness – the identity which belonging to this group affords him. Here, the analyst may plead ignorance of the philosophy and ideals of the system, and may even ask to be educated. Or, the patient who complains that his therapist doesn’t know enough about his system of beliefs may be saying also that he wishes his therapist were more like him. The analyst gratifies this request by making communications that reflect awareness and sensitivity to the issues of the patient’s particular brand of beliefs. One patient developed a feeling of kinship for the analyst, and even wondered if the analyst was a secret member of the same organization.

Where the analyst has not yet understood the ideal closeness/distance needs of the patient, the patient may resort to fight or flight. This mechanism arises out of a fear of dissolution, and is a defensive maneuver to reassert boundary definitions. Fight and flight are attempts to establish separateness: fight through contradiction, and flight through distance. Fight creates strong boundary definition and is ego-enhancing. Flight, however, leaves the situation essentially unchanged.


Mr. and Mrs. P are clinic patients who have been seen as a couple for three years. Mr. P is an abstinent alcoholic, active in AA; Mrs. P is active in Alanon, the organization for spouses of alcoholics. Both are Catholic.

During the first year of treatment, no overt transference communications were made. Mr. P usually remained silent in the sessions; when addressed by his wife or the analyst, he answered defensively or attackingly. Mrs. P talked incessantly, seemingly unconcerned about whether anyone was listening. Finally, in one session, Mr. P asked the analyst how much she knew about alcoholism. She pleaded ignorance and confessed that they were the only alcoholics she had worked with. She offered them the only information she had about alcoholism – that the well-known bio-chemist, Roger Williams, felt that alcoholism reflected a condition of hypoglycemia, and that the only way to get an alcoholic to stop drinking was to correct the hypoglycemia. Mr. P’s response was his first active attempt at communication. He said that the hypoglycemia idea was pure bunk, and was appalled at most people’s ignorance about alcoholism. For several sessions he patiently explained to the analyst the AA concept of the disease. He defined the 12 steps of AA, and described the complex network of communication that typifies the “hot line” service of AA.

After some time, the analyst mentioned that she was thinking a great deal about the use of adjunct institutions in analytic treatment and asked if they could help her understand it better. The analyst knew they would be favorably disposed to this because they had repeatedly emphasized that neither of them would be in treatment without AA. Mr. P brought in pamphlets and articles on alcoholism. He recommended that the analyst attend AA or Alanon meetings.

Shortly after that, Mr. and Mrs. P told the analyst that they were going to be speakers at an Alanon meeting. They asked her to come to hear them.

Mr. P had become more talkative in sessions, and a working alliance was established with the analyst. As the relationship between Mr. P and the analyst became increasingly more cooperative, Mrs. P began to withdraw. She came to one session praising the analyst for helping Mr. P and said that she had decided that he should continue the analysis individually, but that she would get treatment elsewhere. The next several sessions were spent exploring the negative transference with Mrs. P. Part of the material that came out was that the analyst’s religion bothered her. She felt that spiritual counseling by another Catholic would be more beneficial. Mrs. P agreed to stay in treatment until all of them decided that her leaving was appropriate.

Mr. and Mrs. P began bringing up sexual material for discussion, Mrs. P confessed that she hadn’t enjoyed sex since the birth of her second child. To avoid sex, she made sure that her husband was asleep before she went to bed. Exploration revealed that she felt guilty about having had a partial hysterectomy shortly after the birth of her second child. She had been assured by her doctor that the operation was necessary. She sought a second opinion, which concurred with the first. She discussed the matter with her priest who advised her that she was not practicing a form of birth control because the operation was a medical necessity. She was unclear about the origin of the guilt because her priest had determined the act of having the operation to be free of sin.

The analyst pointed out that Mrs. P had indeed sinned. Her sin was a sin of thought, rather than behavior. Even though the surgery was medically justified, Mrs. P had welcomed the procedure for its birth control function. She did not want more children, and the operation gratified that wish. The analyst observed that according to the New Testament, one was accountable for thoughts as well as behavior. Mrs. P had indeed sinned. The analyst advised her to confess to her priest.

The analyst continued asking interventions that reflected familiarity with Catholicism and with living a spiritual life. Mrs. P began to feel a spiritual kinship with the analyst. She reaffirmed her commitment to stay in treatment, and said that she felt the analyst was a spiritual messenger sent to her by God.


Both Mr. and Mrs. P were involved with two institutional adjuncts to therapy. Mr. P needed to see that the analyst would accept his affiliation with AA; Mrs. P needed to see that the analyst would accept her affiliation with the Church. Both institutions had provided comfort to the patients; their identities were wrapped up in the sense of belonging they felt from being members of groups of people like themselves. They needed to know that therapy would support this identity rather than take it away.
Mr. P’s response to the threat of loss of individuation was to fight. He wanted to stay in the therapy relationship and convince the analyst of the merits of his philosophical system. The analyst allowed herself to be convinced, then to question, to seek more information and to question again. Mrs. P’s response to the treatment was to feel alienation and desire flight. She felt that she should find an analyst who shared her spiritual values. When the analyst personified these values for Mrs. P, she was able to see the analyst narcissistically.

When the couple maintained both affiliations, the feelings around the adjunct institutions were brought into and absorbed by the transference. Thus the narcissistic transference developed and therapeutic progress was possible.

Each patient comes into treatment with pre-existing loyalties and affiliations. New ones also are formed during treatment. It is important to understand the significance to patients of maintaining loyalties outside the therapeutic dyad. Outside loyalties may be negatively charged, as in the case of Mrs. B, or positively charged, as in the case of Mr. and Mrs. P. Resolution of the splitting of the transference does not necessarily mean that the patient give up ideologies or membership in outside institutions. Rather, resolution of the split transference requires that all of the feelings experienced towards the adjunct system be experienced, as well, within the context of the transference relationship.


Patients who come into treatment without outside affiliations – or who through treatment are able to establish outside connections – usually manifest some level of object relatedness. The severely regressed patient, however, presents a clinical picture of someone without attachments to objects separate from himself. Nevertheless such a person will be able to form a narcissistic transference. Here the patient creates an internal circle, encompassing himself and the analyst, in which the analyst is experienced as part of himself.

Where the analyst embodies different roles for the patient, there may be extreme polarization and vacillation in perception between one role and the other. The narcissistic transference is maintained because the patient is not able to validate in reality his fantasized ascribed roles.


R, twenty-five years old, has been in treatment for two years. He has profoundly low self-regard, defended against by marked boasting about his professional and sexual accomplishments. He is sporadically employed and lives at home with his parents.
When R first came into treatment, he was spending hundreds of dollars on prostitutes and charging their fees to his Mastercharge account. He had run up a bill he had no way of paying since was, at the time, unemployed. In treatment, he told tale after tale of visits to prostitutes, trying to impress the analyst with his sexual exploits. He even claimed to have had an orgasm during an analytic session. Therapeutic communications in this phase of treatment were neutral and minimally stimulating. As the transference developed and R began to have feelings for the analyst, his visits to the prostitutes were examined as a treatment destructive resistance. He was asked whether the money he was spending would interfere with his ability to continue treatment. An alliance was formed with that part of him that wanted to exert impulse control. He asked the therapist to hold his credit card so that he could not use it.

The first year of treatment was spent with the patient making requests of the analyst. Would she have dinner with him at the Playboy Club; could he spend the night at her house if he found himself in the city late at night; could he live with her, ride around the city with her, rescue her if her building burned down? When these requests were explored in terms of how they would help his analysis, the patient gave rationales: “You would get to see me in social situation,” “You would see me more often,” “It would show me that you care.” These fantasies were neither encouraged, nor discouraged, but, rather, neutrally explored.

Analyzing rather than gratifying his demands proved to be extremely frustrating for R. He ordered the therapist to meet his demands. When asked what gave him the idea that he could order her, he replied: “You have to do what I’m telling you. I’m paying you for that.” He was advised that he was paying for analysis, not for a slave. His response to even minimal frustration was to threaten to leave treatment. For more than a year, he left every session saying that he would not return.

The content of most sessions dealt with transference issues: “Tell me you care.” “Do you trust me?”, or requests for gratification of erotic fantasies. The sparse historical material had to do with feelings of worthlessness, sexual feelings, and fantasies. There was a confusion of sexual identity. The patient mistakenly referred to himself in the female gender.

In the second year of treatment, the patient asked the analyst if he could refer to her as his wife. He began writing letters to her addressed “To my wife.” In the sessions, he alternately referred to her as his wife, or his therapist. When asked about this once, he answered: “They’re two different people. As my wife, you are successful and I like it. As my therapist, I feel jealous, maybe even hatred. As your wife* [*Here the reference to himself as the wife is an example of gender confusion], I can talk better, think better, do everything better. As husband and wife, I see the relationship as one.”

As the husband/ wife fantasy developed, R experienced the relationship as less frustrating, and was able to verbalize his feelings. Threats to terminate became infrequent. Instead he said, “I feel like shoving my fist down your throat.” He explained that as a husband he could love the therapist and would not leave her, even if he was angry. He asked innocently if he would be divorced if he felt angry. The fantasized marriage also allowed him to explore his wishes in relation to women. He expressed the desire to be dominated and controlled and referred to his experiences with prostitutes, where he had acted out sado-masochistic dominance fantasies.


R’s rationale for choosing the analyst was the first indication of narcissistic transference. He had video-taped a conference in which she had participated and he said that she must be an exhibitionist, like himself, to have spoken on stage to so many people. It has been noted that in narcissistic transference, seemingly inconsequential items, such as the analyst using the same kind of pen as the patient, take on momentous and symbolic meaning**. [**Personal, communication, Marie Coleman Nelson.] For this patient there were similarities between his relationships with prostitutes and with his analyst: they were both women and they were both being paid for their services. These similarities were sufficient to cause him to identify the analyst with prostitutes. When the patient informed the analyst that he had reached orgasm in a session, he indicated that the analyst and the prostitutes were synonymous for him, and revealed his belief that the rules were the same for both situations.

R’s self-esteem was so low that he assumed that the only woman he could get to spend time with him was a woman he paid. Conflict then arose: he was acculturated enough to know that most men do not pay to get wives; yet he was not well enough integrated to have the feeling or the capacity to be able to get a wife (woman) without buying her. The resolution of the conflict took place on a fantasy level — that is, through the idea of the therapist as his wife.

Initially the verbal contact with the analyst, whom the patient saw as attractive and intelligent, was sufficient gratification. The patient seemed to get pleasure out of the mere fact of the analyst speaking to him. He experienced a sense of accomplishment that she would accept him as a patient and speak to him during sessions. When he became accustomed to this level of gratification, he sought new ways of being transferentially satisfied. The fantasy of the analyst as his wife reduced his feeling of inadequacy in the world and increased his feeling of adequacy in the transference relationship.

The splitting of the transference object into wife and therapist probably prevented further regression. A more direct confrontation with his feelings of worthlessness and self-hatred would have rendered the patient, who was only tenuously functional, dysfunctional. Maintaining the reality perception of the analyst as therapist allowed the therapy to proceed; creating the fantasy perception of the analyst as wife allowed the patient gratification enough so that he could more comfortably explore his feelings and fears. The fantasized husband/ wife relationship also enabled him to experience more fully the commitment to the relationship. As a husband, he understood that it was role-appropriate to stay in the relationship and to tolerate and resolve the negative feelings; as a patient, he permitted himself the idea of acting out whatever he felt.

The gender confusion created even further fractionalization of the transference. The frequent gender slips indicated that he experienced the analyst as therapist, as wife, and also as husband.

The narcissistic transference gives us important information about the history of object relations. It gives us information about the patient’s self-identifications. The triple narcissistic split-transference evidenced by this patient allowed him to show the analyst a composite picture of his three critical points of conflict. Resolution of the triple split-transference would lead to integration of the self-identifications.