Janet Schumacher Finell, New York

For always I am Caesar ... (Act I, Sc. II)
I am constant as the Northern Star,
Of whose true-fixed and resting quality
There is no fellow in the firmament...
That unassailable holds on his rank,
Unshaked of motion, and that I am he...
(Act III, Sc. I, Julius Caesar - William Shakespeare)

INTRODUCTION

Volumes of psychoanalytic writings focus on narcissistic pathology today. The topic is of central importance and attracts interest and lively controversy. Concurrent with the intense interest in narcissism is an ever-growing realization of the importance of countertransference in the analyst's work. Narcissistic reactions and countertransference part company at the point of considering the possibility that narcissism in the analyst may not always be a reaction to patient narcissism. It may in fact exist independent and quite separate from patient pathology.

Painful feelings of shame lower self-regard and affect self-esteem when the analyst is faced with defects in his idealized self-representation. The impossibility of processing denied and split-off feelings throws the burden of responsibility on to the analyst's personal analysis. If this fails to work through grandiosity, exhibitionism, aggression. and power, the potential for counter-transference over and above a specific reaction to a narcissistic transference in patients is enormous. The field affords many possibilities for gratification of narcissistic longings. Following Freud's spirit of shared humanity in regard to his adhering to the same model of neurosis in studying his own psyche and that of his patients, narcissistic issues should become a respectable area of concern among analysts.

RECOGNIZING NARCISSISM

The issue to which this paper is directed is: To what extent can we as analysts work through our narcissism in our personal analysis? If it has not been worked through, can we recognize its continued existence and accept it so that it does not interfere with our treatment cases? Can we seek out treatment for the express purpose of working through our own narcissism? In effect, can one see one's own narcissism? Can we select an analyst who will make an impact on our narcissism? Why is this area often neglected in the volumes of countertransference literature? Is this a blind spot. too sensitive to be dealt with in writing?

The honesty and self-disclosure evident in much of the countertransference literature (Tower, 1956; Racker, 1957) is sorely missed in this area. Is there unconscious collusion among analyststo avoid this issue except where it can be explained as 'narcissistic tension' (Kohut, 1977) which reflects 'residues' of narcissism in the analyst? The split between the patient and analyst--the narcissistic patient is hypothetically treated by a better-integrated, object-related analyst who, at the worst, struggles with residues of narcissism--may be a reaction formation against a commonly shared secret in the profession: that analysts, too, struggle with narcissism.

It is my hope that this paper will arouse questions in the reader's mind about these issues in him/herself. If it accomplishes that much, a lot will have been gained. One might ask, what role can study and reading psychoanalytic papers play? Since these experiences are filtered through the cognitive and intellectual ego, they are unlikely to have any impact on our introjects or our emotions.

In fact, they may intensify the intellectualizing, defensive ego functions rather than expand the experiencing ego's domain. Freud (1937) was pessimistic about activating dormant conflicts through analysis or through reading. He wrote:

The reader is 'stimulated' only by those passages which he feels apply to himself--that is, which concern conflicts that are active in him at the time. Everything else leaves him cold (p. 233).

Yet for some, reading and hearing papers can have a profound effect, activating barely thought-out concerns, bringing to consciousness preconscious issues, adding understanding to difficult treatment problems, or helping in the conceptualization of a clinical issue. However, where narcissistic defences predominate, material that can be challenging, anxiety-provoking, or otherwise disturbing will be attacked as inaccurate and wrong, and denigrated or rejected. Narcissistic defences are particularly impenetrable to cognitive appeals, thus making this area even more resistant to logical presentations than other characterological issues.

Given that the narcissistic individual is heavily invested in not processing his or her pathology and deeply defends and splits off troublesome feelings, narcissism in analysts finds many gratifications. Since the narcissistic individual has intense envy-aroused aggression and narcissistic rage, products of early frustration and disappointment (Kohut, 1971), or in another view, intense innate aggression (Kernberg, 1975; Rosenfeld, 1964), these feelings are not easily integrated in a self that splits off and projects unwanted feelings.

Personal analysis may not resolve the problem if the narcissistic individual discharges and splits off his/her aggression on to 'safe' targets. Grotstein (1981) writes about this phenomenon: 'if the analyst is held in a highly idealized position, then the negative transference may be split off and projected into objects on the outside' (p. 178). The search for safe targets may be gratified by the presence of colleagues of different theoretical positions. Institute politics provide an acceptable and convenient outlet for aggression.

Such politics also provide gratification for grandiosity, exhibitionism, and desires for power and control. Denying and splitting off aggression makes it unavailable in the transference. Defensively, split-off aggression fosters the treating or personal analyst's narcissistic feelings of goodness, self-idealization, or grandiosity, and promotes narcissistic over-estimation of the self. The treatment is purified of the sullying effects of the analysand's anger and contempt.

By ridding the self of unacceptable impulses, the analyst sees himself as loving, giving, and empathic toward his/her own patients while aggression is split off and re-directed against 'safe' targets - outsiders, non-patients, or those with different theoretical approaches or political alignments. The outsider may be viciously attacked to preserve the illusion of all-good self. These dynamics feed the analyst's narcissism and interfere with integration of aggressive aspects of the analyst's personality. Self-idealization or grandiosity promotes narcissistic over-estimation of the self. The analyst is then unable to process or work with the analysand's idealization of him/her, since to do so would be to explore underlying aggression. Aggressive aspects of the transference cannot be processed, since aggression is displaced to 'safe' extra-transferential objects.

Greenacre (1966) noted that the analyst's hostile drives may be directed to relatives as well as others in the analytic community. She wrote:

It is not only the narcissistic needs of the analyst but his failure to recognize his own hostile aggressive drives which seem to make trouble by dovetailing with the problems of these analysands. He may then too readily identify with his patient and accept the latter's splitting of transference with projection of negative transference onto others, especially accessory members of the patient's family or of the analytic community (pp. 210--11).

To return to the central theme of this paper:

How can one know if one is defending against narcissism, and, if so. what can be done about it? Narcissistic defences involve splitting, denial, and projection, and therefore make insight very difficult. The self in narcissism is struggling to maintain a'purified pleasure ego'. The self cannot integrate, process, or feel painful feelings such as hostility. envy, and contempt. These feelings are discharged, projected, split off, and induced in others in the dynamics ofnarcissism. Freud pioneered in self-analysis with his highly revealing

'The interpretation of dreams' (1900). His sharing his deepest conflicts should be a source of courage for all of us. Following Freud, can we face our grandiosity, sadism, splits and denials, projections and self-fragmentation? Can we deal with envy and aggression as possible reactions to our patients? Can we take an honest look at our own introjects and how these come alive in treatment situations?

Some individuals have moderate levels of narcissistic characteristics and splits that only become problematic when working with difficult patients. Broucek (1982) describes how shame and narcissism are related. Failure to live up to idealized and grandiose ideals arouses shameful feelings of failure. The painful affect of shame contributes to the difficulty in processing experiences that lower self-regard, while experiences that heighten self-regard may be used defensively to shore up the sense of self.

When the analyst is in training, narcissistic problems that are not picked up and worked through in personal analysis may be picked up by teachers, committees, and supervisors. Rangell (1982) reports a personal communication from Loewenstein noting'that the most unanalysable case was a successful narcissist' (p. 876). It is beyond the purview of this paper to address in depth the issue of how the analyst should resolve and work through narcissistic problems in the treatment. The treatment of narcissism is difficult and has been in the forefront of psychoanalytic literature. A variety of theories, approaches, and ways of thinking about this difficult problem are available. However, a short summary seems indicated.

Contemporary literature on narcissism includes primary writings (Kernberg, 1974, 1975, 1976. 1980; Gear et al., 1981; Grunberger, 1979; Kohut. 1971, 1977; Kohut & Wolf, 1978; Stolorow & Lachmann, 1980) as well as many reviews and critiques of narcissism and self psychologies (Blum, 1982; Cooper, 1981; Hanly, 1982; Hanly & Masson, 1976; Rangell, 1982; Robbins;1982; Saperstein & Gaines, 1978; Sprueill, 1974). The topic of narcissism has generated prolific writing well beyond the scope of this paper to cite. For purposes of this paper, the narcissistic character is defined as one in whom idealization and grandiosity, excess aggression and need for power and control reflect early deprivation and narcissistic injury. Splitting and projection are the major defences but may coexist along with higher level defences. Preoedipal and oedipal issues appear in the dynamics and it is my view that it is arbitrary to force dynamics into developmental or defensive explanations (Stolorow & Lachmann, 1980; Kernberg, 1974; Kohut, 1971). I propose a continuum model based on Freud's concept of the ego. Structure and neurotic conflict coexist with splitting and projection. Different situations activate different defensive reactions.

Following Kohut (1977), in some cases the analyst's narcissistic defences may be aroused in the treatment of narcissistic patients. while their higher level defences appear to predominate in work with better integrated patients. Both narcissistic reactions to specific patient dynamics, as well as pervasive narcissistic reactions as generalized modes of responding are addressed in this paper. The analyst's reactions to the patient's defences are likely to reflect the dynamic interaction between the two.

This paper is focusing not on the treatment of narcissism but on how one can determine whether or not the extent of narcissism in one's own personality isproblematical in terms of its potential interference in analytic work. It is, however, relevant to observe that when committees, supervisors, or colleagues pick up and alert the candidate to excess narcissism, it is unfortunate if this knowledge does not become part of the personal analysis. The major setting for resolving characterological narcissism is personal analysis. As noted, it is the responsibility of the analyst to treat the narcissism and to help the analysand arrive at manageable levels of sadomasochism, grandiosity and devaluation, exhibitionism and timidity, power, control, and dominance versus dependency, helplessness, and submissive dualities. Since the analysand will not be aware of the tendency to idealize and to feel positively toward the analyst, all the while splitting off and projecting devaluating and hostile feelings to others outside of the analytic situation, it is the personal analyst's responsibility to address the problem. In situations in which the personal analyst becomes theobject of hostility, contempt and devaluation, while the idealizing tendency is directed outside, the painfulness to the primary analyst brings it into the scope of analysis fairly quickly.

Conversely, idealization of the personal analyst with the analysand experiencing personal grandiosity through identification can be highly gratifying to both partners. The analysis has an overall positive aura and may continue interminably as long as nothing disturbs this folie a deux.

GRATIFICATION OF PSYCHOANALYSIS

A quotation from Rangell's (1982) recent article on the self confronts the fact that narcissism exists in the analyst as well as the patient. He writes:

Nor does narcissism constitutea sharp separator between the two participants in the psychoanalytic process. Problems in the countertransference can be traced to narcissistic injuries and defenses as frequently and regularly as in the transference (p. 876).

Kohut (1971) believes that the analyst's counter-transference difficulties 'in the analysis of narcissistic disorders are rooted in the analyst's own narcissism...' (p. 260). They 'do not differ from those which occur in the analysand' (p. 260) in terms of dynamics. Miller (1981) writes of the attraction of analytic work for those suffering early narcissistic injury.

Psychoanalytic work can afford immeasurable gratification to those eager to be loved and needed. Greenacre (1966) believed that narcissistic problems in the analyst parallel those of the patients so labelled. She cited over-idealization as a key issue in certain analysts with a strong need to be loved who 'find themselves enmeshed in too strong a positive transference in which it is difficult to decipher the negative elements' (p. 209). Kernberg (1975, p. 291) comments on the difficulties of analysts in training whose narcissistic resistances were not resolved in personal analysis. Some lose interest in analytic work and seek out different treatment modalities that offer immediate pseudo-intimacy rather than the deeper personal relationships of patient and analyst working together in depth. While this may be true in some cases, the analytic situation offers much gratification for analysts with intense needs to be loved,idealized, and to feel a sense of power and control over others. Analysts with such dynamics will tend to promote idealization, power, and control by taking a dominant position in relation to the analysand who is essentially submissive and masochistic in these dynamics. In these circumstances, analyst and patient collude and form a misalliance in the sense describedby Langs (1975). The narcissistic character structure of both is protected, and both receive a great deal of gratification that leaves the basic pathology untouched.

The specific constellation of transferencecountertransference addressed in this paper broadens that described by Kohut (1971). He discussed the analyst's inability to accept the patient's idealization because of the narcissistic tension aroused in the analyst. Wolf (1979) attributes difficulties in dealing with mirror or idealizing transferences as a reflection of the analyst's 'narcissistic vulnerability'. This type of reaction is illustrated with case material by Gedo (1975), presented later in this paper. In these situations, the analyst's unresolved narcissistic conflicts make it impossible for him to process and work with the patient's idealization.

Kohut (1971) describes also the reverse situation in which feelings of boredom, inattention, and emotional withdrawal occur in the analyst in reaction to the patient's non-'objectinstinctual' transference. He courageously cited:

specific hindrances in my own personality which stood in the way. There was a residual insistence, relatedto deep and old fixation points, on seeing myself in the narcissistic center of the stage... Thus I refused to entertain the possibility that I was not an object for the patient ... 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 (pp. 287--8).

Kohut's honesty about his own unresolved narcissistic issues is admirable and apparently led to his overcoming personal reactions which led to emotional withdrawal in work with patients who did not cathect him with object-instinctual libido. One should also recognize, however, that an enactment or projective identification is occurring in which the analyst is made to feel excluded and rejected the way the patient once felt in relation to significant others. Kohut's patient's behaviour may represent an identification with a narcissistic and rejecting parent. She excludes the analyst as she was excluded by her parents.

The present paper enlarges the scope of the study of narcissism in analysts by including those treatment situations in which idealization is fostered, accepted. and encouraged. In these cases, the analyst does not experience conscious narcissistic tensionsince the narcissistic gratification is ego-syntonic. As a result, there is no awareness of conflict either interpersonally between patient and analyst, or intra-psychically within the patient and analyst. Negative feelings are split off and directed to 'safe' objects.

IDEALIZATION AND GRANDIOSITY

This paper is specifically addressing idealization, grandiosity, and related issues. Kernberg and Kohut have been prominent in writing about these dynamics and treatment approach. Kernberg (1975) sees idealization as regressive and defensive against conflict and rage, while Kohut (1971) sees it as evidence of developmental arrest and missing structure. While Kohut sees it as a reflection of an early need to idealize parent figures, Kernberg sees it as a reaction formation against oral rage, loss, and disappointment. It should be noted that Kernberg's thinking follows Klein (1957), who believed that idealization and omnipotent longings were a reaction to early deprivation and losses, and were defensive against rage. Kernberg advocates interpretation toward the end of promoting ego development through integrating the split idealization and devaluation. A more reality-based ego would be reflected in a less harsh, punitive and grandiose superego. Kohut advocates an empathic approach in which structure is built through transmuting internalizations which involve 'a depersonalization of the introjected aspects of the image of the object, mainly in the form of a shift of emphasis from the total human context of the personality of the object to certain of its specific functions' (1971, p. 50).

Stolorow & Lachmann (1980), in an attempt to resolve differences in the understanding of idealization, offer case examples illustrating preoedipally based developmental idealization, and oedipal-conflict based idealization. Their explanation of idealization is based on the degree of self and object differentiation which appears in the transference De-differentiationof self and object in the transference is indicative of missing structure in contrast to an ability to maintain separate self and object images in the transference in patients with higher levels of development. I consider their explanation to be overly schematized as it is not based on a continuum model, following Freud's thinking on the ego.

The polarization and categorization that attributes narcissistic pathology to patients but not to analysts is paralleled by a theoretical schema that attributes narcissism to either developmental arrest or defensive regression from conflict. I believe that people are complex, and that narcissism reflects complex character configurations with elements of structural defect as well as regression from conflict. Moreover, I find the concept of missing structure problematic in that if something is missing, the void must be filled in some manner. Conceptually, it is hard to conceive of missing structure that is not filled with something--in psychoanalytic terms--either pathological or at least skewed structure.

McDougall's (1982) conception of narcissism not only deals with oedipal and pre-oedipal determinants, but takes issue with the popular stance of opposing narcissistic with neurotic disorders. Rather than categorize patients as narcissistic, neurotic, obsessional, masochistic or schizoid disorder, each individual 'mental mosaic' (1982, p. 375) needs to be understood. Since 'such understanding is necessarily filtered through our own intricate network of libidinal investments and narcissistic defenses' (p. 375), it is important not to bring our own biases into our clinical work. An over-emphasis on using a particular designation such as 'narcissistic personality disorder' might then, in my opinion, reflect projected narcissistic tendencies that can be defensively disavowed by finding them in one's patients rather than in oneself.

THE SELF IN NARCISSISM

It seems most helpful to me to conceptualize narcissism as a function of a self that has difficulty in containing painful states and feelings. Freud's (1920) concept of the purified pleasure ego can be applied to a state of the self in which unpleasantness is projected outward. In 'Beyond the pleasure principle' (1920), Freud described how internal impulses which could not be kept out of the infantile psyche through the protective shield or stimulus barrier were projected outward. He wrote:

There is a tendency to treat them [internal excitations] I as though they were acting, not from the inside, but from the outside, so that it may be possible to bring the shield against stimuli into operation as a means of defense against them. This is the origin of projection, which is destined to play such a large part in the causation of pathological processes (p. 29).

Today, the concept of projective identification. originated by Klein (1957), has come to be regarded as an important technical tool in working with narcissism (Rosenfeld, 1983; Segal, 1983). However, controversy exists about the concept and its clinical application (Finell, 1984).

The crux of the problem is that, by definition, it is impossible to discover independently that one is splitting since the unwanted feelings and experiences are pushed outside of the self and projected to others. McDougall (1980) clearly describes the clinical situation that exists when the analysand expels painful affects and attempts to induce them in the analyst. The nature of the process makes self-awareness impossible by definition unless outside help can be internalized. Since narcissistic defences eliminate the experience of conflict, and painful emotional states may be induced in the object, the interpretive work must address these dynamics. The elimination of feelings, the unawareness of pain, and its projection to the analyst need to become the focus of the transference analysis. This can occur only if the analyst himself/herself is not drawn into the patient's projections. Thus, if the analyst accepts the projection of grandiosity, the idealization is acted out. This type of unconscious interaction has been termed projective counter-identification (Grinberg, 1962).

CASE MATERIAL

I. Published case material affords an excellent way to studv controversial issues since the original source is available to all, in contrast to one's own case material which is inevitably subjective. Gedo's fascinating account of problems arising in the course of analytic work when a silent idealization was not analysed provides a good case example. In one of his cases he describes how the patient in the next to last session revealed a silent idealization:

the patient spoke of his satisfaction with the outcome of the analysis, listing his various adaptive gains, etc. He capped this recital of gratitude by stating, with matter-of-fact seriousness, that his analysis had probably been the best one ever performed. The analyst was completely taken aback by this revelation of a hitherto completely silent idealization; in the few moments remaining, he did interject that, from his vantage point such a view seemed to be unrealistic: the analysis had not been the best in his own experience.

The patient returned for the last session in great rage. He did not change his mind about ending the treatment, but he complained of severe gastric distress, which he connected with being deprived of his fantasy. He came back about six months later essentially to say that he was still enraged at the analyst, and in another year he returned for a last time to announce, with considerable venom, that he was planning another analysis--elsewhere (Gedo. 1975. pp. 491-492).

Gedo criticized the analyst who had treated this man for his focus on oedipal competition. He believes that the analyst failed to allow the idealization to develop within the transference. Interpretations were experienced as a traumatic reliving of the mother's attacks on the patient 'whenever he attempted to idealize his father: she may have required him to establish a pseudo-oedipal mode of relating to the latter' (p. 492). As a result, the patient's grandiosity which was projected on to the analyst via the idealization was not openly expressed until the penultimate session.

Gedo's criticism highlights the central point I am making--the patient's transference has become the exclusive focus of Gedo's critique. It is quite apparent that the patient is struggling with grandiosity and idealization. The analyst's narcissistic tension is revealed by his unempathic and hurtful comment that 'the analysis had not been the best in his own experience' (p. 492). Even if the analyst were aware but struggling with his own narcissism, he could have processed the patient's comment, realized consciously that he had erred seriously in not picking up the silent idealization, and commented that the patient's remark was most interesting; he would like to think about it, and would the patient be willing to extend the analysis to explore this issue? He could then explore whether he had been unconsciously colluding with the patient to maintain the idealization for his own narcissistic needs. Rather, the analyst's retort seemed hostile and defensive. His 'analyst-ideal-self' was disturbed when he realized that a powerful idealization was operating silently. The statement revealed a totally unanalysed segment of the patient's personality and longings. Had the analyst openly admitted to himself that he had failed but would try to correct his error through further work with the patient, and ideally further personal analysis of his own, one could conclude that the narcissistic tension aroused by this situation was at least potentially manageable and workable. In contrast, the analyst could not process his own narcissistic tension and could not respond either empathically or analytically. The collusive aspects of the situation were never processed by the analyst. who acted out his disappointment by attacking the patient. This account is fascinating not only because of the interesting case material, but because of Gedo's neglect of the role of the analyst's countertransference.

Gedo's article reveals just how difficult narcissistic problems are for all of usto deal with. While we can easily label our patients as suffering from pathological narcissism and a grandiose self, we attribute to ourselves a different level of pathology, calling our narcissism 'residues' which arise as a reaction to the patient's narcissism (Kohut. 1971; Kernberg, 1975; Stolorow & Lachmann, 1980). Why then, the split between patient and analyst? Essentially, the analyst strives to maintain what I have called the analyst-ideal-self which is seen as structured, conflict-based, and essentially oedipal except in so far as pre-oedipal issues are activated by patient pathology. Although Freud followed the medical model in his treatment approach, his self-revelations reveal that he saw himself as struggling with the same types of issues as his patients. His sense of humanity and egalitarian attitudes toward his patients were evident by his continual attention to his patients' comments and recommendations about his technique. In the enormous amounts of countertransference literature, the analyst's own narcissism is described as a residue, based on an assumption that the analyst is object-related and essentially free of narcissistic pathology. This attitude seems more self-protective than courageous, and suggests that hidden grandiosity makes it difficult for the analyst to confront his own narcissism openly and courageously.

II. Guntrip's (1975) article on his analysis with Fairbairn and Winnicott provides a fascinating account of the personal experiences of a leader in the field, in his analysis with two other major contributors to psychoanalysis. My impression is that the narcissistic issue was at the bottom of Guntrip's difficulties. He presented himself to Fairbairn at the age of 48, the age when, according to Kernberg (1975), narcissistic problems often begin to become painful. His presenting problem was feelings of illness and dizziness which he attributed to his amnesia for the traumatic death of his brother Percy, when he (Guntrip) was 31/2 years of age. Fairbairn was no longer in good health during Guntrip's analysis. Guntrip suggests that this may have been the cause of Fairbairn's overly intellectualized approach. which focused on oedipal material. Guntrip was only partially satisfied with the results, which helped him work through oedipal material. At the end of his analytic sessions, Guntrip and Fairbairn engaged in theoretical discussions during which Guntrip described Fair-bairn as coming alive, in contrast to his cold and distant manner during the sessions proper.

The analysis lasted for just over 1,000 sessions and Guntrip terminated when he realized Fairbairn was close to death. He feared his death would reactivate the traumatic death of his brother. He sought out Winnicott, with whom he had a much shorter analysis, just over 150 sessions over a period of 6 years. He found Winnicott's approach warmer, more congenial, and found him more comfortably human both in personality and in surroundings. Winnicott was able to help him ease up on the constant activity which defended against his fear of annihilation. He helped him get in touch with early traumatic experiences with his mother. Winnicott told Guntrip that he was good for him. Guntrip reports Winnicott's saying:

You too have a good breast. You've been able to give more than take. I'm good for you but you're good for me. Doing your analysis is almost the most reassuring thing that happens to me. The chap before you makes me feel I'm no good at all. You don't have to be good for me. I don't need it and can cope without it, but in fact you are good for me (p. 153).

Winnicott was enjoying Guntrip's role reversal, but knew it was problematic. When Winnicott suggested that the core of Guntrip's problems was in his 'primitive sadism, the baby's ruthlessness and cruelty, your aggression' (p. 153), Guntrip disagreed. He felt that Winnicott was resorting to Freudian and Kleinian instinct theory in regard to innate aggression. Throughout both analyses Guntrip kept notes of allsessions, which may have fostered his feelings of control of the analytic situation.

Winnicott's death ushered in dreams which Guntrip analysed himself, and which he believed led to his cure, at the age of 70. Not only did the loss of Winnicott reactivate the traumatic loss of Percy, but Winnicott's death made him a safe object for Guntrip to internalize since he could not now fail Guntrip in any way.

The core of Guntrip's problem was his experience of his mother as cold and indifferent to him, and a murderess to his brother. He had to take care of himself in order to survive. His omnipotence or infantile megalomania (Glatzer & Evans, 1977) was a defence against the fear of depending on his mother, and later on his two analysts in the transference. 'The only person who could fully analyse Guntrip was Guntrip' (Glatzer & Evans, 1977, p. 87). While Winnicott made more of an emotional impact on Guntrip than Fairbairn did, he could trust neither totally. In essence, he could only trust himself.

Glatzer & Evans (1977) suggest that both Fairbairn and Winnicott gave their highly gifted and fascinating patient unusual gratification which prevented his being able to experience aggression toward them. Guntrip engaged in theoretical discussions with Fairbairn and exchanged books and papers with Winnicott. Although Guntrip reported a number of acting out aggressive acts with Fairbairn, his hostile transference did not seem to have been worked through and integrated. Both analysts and Guntrip colluded in the sense of Langs (1975) so that each gratified the other's narcissism and avoided Guntrip's rage. Eigen (1981) notes that 'The atmosphere of mutual mirroring-liking contributed to muting anything unpleasant which could spoil the gratifying communion' (p. 107).

Guntrip's honest and moving account of his experiences gives us a chance to see at firsthand just how difficult narcissistic issues are to resolve. By remaining the ultimate authority, his omnipotence, a critical element in narcissism, was not worked through. The two analysts were devalued. Guntrip triumphed over them by curing himself. Relinquishing narcissistic control through trust in the analyst permits internalization and working through of the deepest conflicts. This did not seem possible in Guntrip's two analyses in so far as his account is unbiased. The resistances against permitting the dependency, trust and need for the analyst that is part of the working-through process in narcissism is demonstrated in this material. The ease with which the analyst can collude in these situations is also dramatically demonstrated.

III. A. In my own clinical experience I have attempted to pick up silent idealizations in my patients as well as avoid the over-gratification that subtly cuts off aggression. A recent experience confirmed my growing comfort with work on idealization. A patient stood up at the end of the session, and raved in a manically exaggerated manner about how wonderful psychoanalysis is and what a great analyst I am. She had got in touch with early and previously unavailable material about her parents and was extremely elated. Her manner seemed extreme and unrealistic to me and I suspected that other feelings might emerge which would provide important information on her hysterical problems as well as her grandiosity. I reflected her elation since she needed to idealize me and the process of analysis. However, when she plunged into depression I was able to help her connect her intense idealization of the analysis and the depression that often followed her elated moods. Her elation reflected moments of intense gratification in infancy which were followed by withdrawal and exclusion of her by her parents. Their exclusion created feelings of deep depression and loss which became a deeply-ingrained pattern. These dynamics reflected both her need for an idealized object and conflict over the rage that their exclusion aroused in her. In early years I would have felt very pleased and would have accepted the idealization as a valid compliment. I would have been quite disappointed and probably angry at the subsequent depression. I would have been unprepared for the shift in mood. At present I was able to process the idealization and work on it analytically and empathically.

B. In my early years as an analyst, patient grandiosity with devaluation of me left me feeling helpless and inept. It typically aroused feelings in me of not being needed, of being of no importance or value to the patient. Defensively, feelings of boredom protected me, but I was not able to use these countertransference feelings interpretively to enhance the patient's self-awareness and ego development. Recently, I dealt with counter-transference reactions while a narcissistic patient conducted self-analysis, rarely engaging me, but being proud and exhibitionistic about his discoveries. His grandiosity placed me in a devalued, unneeded role. I processed my countertransference feelings of boredom, rejection, and uselessness with the goal of bringing them into the analysis under the control of myego. The patient signalled his readiness by complaining that he felt he had gone as far as he could, and that I was too protective of him. He wanted me to be more challenging. The concept of projective identification made it clear that by inducing feelings of uselessness in me he was presenting a self that had been helpless in very traumatic situations. His protective distancing defended against dangerous fears of castration, reflecting pre-oedipal and oedipal anxieties. His transference behaviour had multiple meanings but the characterological issue predominating was intense narcissistic self-containment.

The time now seemed ripe for me to interpret that he had to keep me at a distance but wanted me to listen and admire him. He reacted with anxiety to this interpretation, and that night and over the next few weeks had a series of dreams involving violenceand dismemberment. These dreams represented fears of castration and annihilation by his mother, who had physically assaulted him in childhood, but who was periodically seductive and made him feel special in relation to his siblings and his father. She was capable of brutality, and on one occasion had tied him to a chair and threatened to cut off his legs. The father was uninterested and uninvolved with the patient, and provided no help against the mother.

In analysis, this patient's narcissistic armour provided protection against terrifying transference feelings. When he indicated that he was ready for analysis of his narcissistic transference, I interpreted his defensive need to keep me at a distance and keep control. Over time the patient's need to act as though he were conducting a'self-analysis', to bring in an inordinate number of defensive dreams and to report his inner developments in an intellectualized, controlled manner decreased. The analysis became emotionally alive as the patient came to experience his conflicts intrapsychically rather than expel and enact them by inducing feelings of uselessnes in me.

MUTUAL GRATIFICATION

Idealization, grandiosity, devaluation, and fear of dependency form an essential part of the narcissistic configuration. These feelings are painful and shame-inducing. As a result, they may go unanalysed in a treatment that may seem successful to both parties. The use of splitting and projection makes the feelings unavailable. Analysands in psychoanalytic training sometimes make use of teachers, supervisors, and others as targets for their aggression while idealizing the analyst, the analyst's analyst when known, or others analysed by the analyst. This material needs to be brought into the analysis with the aim of fostering a higher level of ego integration. Supervisors of analytic students have the responsibility of bringing to the supervisee's awareness excessive idealization or devaluation of the analyst, supervisor or teachers as well as the candidate's characterological grandiosity or timidity. The suggestion that this material be taken up in personal analysis should be made. If the personal analyst cannot work through the splits, only partial results will be achieved. The analysand's ability to deal with these issues with patients will be limited.

It is not unusual to observe dedicated and serious colleagues involved in personal interminable analysis in which idealization is not worked through. Clues that the personal analysis did not resolve these issues are idealization of a personal ongoing orterminated analysis and of the personal analyst, devaluation of others who follow a different approach or are not theoretical or political allies of the personal analyst, and excessively positive and loving feelings towards the personal analyst while hostility and devaluating feelings are split off and displaced to others. In these situations, the analysand maintains intense loyalty to the personal analyst and may not be able to seek out another analyst long after the analysis has been terminated.

The narcissistic personal analyst has a vested interest in maintaining the analysand's idealization, loving feelings, and dependency. The analysis can become interminable with the analysand quite fearful of leaving the analyst and yet secretly taking care of the analyst in the sense of Langs (1975) and Searles (1975). The analyst is sending out messages that he needs the patient. The patient's grandiosity is fed; he feels loved and needed and anxiety over abandonment is never experienced openly by the analysand. The analyst is the saviour, to use Greenacre's (1966) word, who will omnipotently rescue the analysand in a 'complete cure, approximating even a rebirth' (p. 210). The early deprivation and losses which caused the narcissistic problems are not worked through. The terror of separateness so poignantly described by McDougall (1980) is never experienced. Rather, gratification sustains the illusion of fusion and denies separateness and loss. A hostile symbiosis with intense dependency needs is hidden behind an externally loving, positive transference.

Failure of the analysand to keep up the idealization may arouse hostility and withdrawal from the personal analyst which may discourage further honest exploration of feelings. These dynamics can result in an interminable analysis with the patient becoming either stalemated or depressed and deteriorated. Patient and analyst collude to avoid working through loss, separation, and individuation. The analysand feels he cannot function separately. The personal analyst subtly fosters dependency and becomes, in a sense, the repository of the analysand's ego and its functions. The analyst may promote his own thinking while devaluing the analysand's contributions to the analysis. The analysand is encouraged to accept an essentially submissive and masochistic position in which the personal analyst's contributions are over-estimated. This position is damaging to the analysand's ego development.

If the personal analyst gives lip service to idealization in the treatment because of its popularity as a concept today, both parties may feel exonerated from any deep exploration of the issues. The transference hold is a powerful one. Outsiders, including friends and relatives, are unlikely to have much influence on these pathological symbiotic transference ties. The patient denies all awareness of the destructive attachment but will attack anyone who questions the treatment. If the personal analyst is deteriorating, the projective-introjective dynamics that prevail render the analysand exquisitely sensitive to the analyst's needs. If the primary analyst is in ill health, either physical and/or emotional, the analysand may show distressing signs of deterioration which he does not realize are a function of the transference. Usually these situations are ended through death, ill health, or unusual external exigencies that force the two apart.

Following Freud's recommendation that analysts return for analysis every five years, one can ask--with the same or a different analyst? How does this recommendation relate to those analyses in which we see our colleagues working with the same analyst for 20 years and over? Often the personal analyst seems successful and has a large and loyal following. Greenacre (1966) noted that when the analyst is well known and carries considerable prestige in the community from which the analysand comes, the treatment may get a:

kind of magic power. with improvement deriving from association with the analyst rather than from the analysand's own development. In such situations transference cures may result without fundamental change or growth in the patient (p. 211).

These dynamics promote an idealizing transference with resultant fantasies of the analyst's omnipotence. The analysand's grandiosity feels gratified by the liaison.

Rangell (1982) describes how personal charisma can generate a 'mystical union...by shock rather than by reason--in psychoanalysis no less than in social or political life or even art (p. 885). If the personal analyst has a following and is in a position to give the analysand narcissistic gratification such as power, control, and other external gratifications, dependency is fostered through reality gains as well as grandiose longings. The analyst's personality may be charisismatic, yet as Greenacre (1966) notes, he may appear 'disarmingly modest', but [have] a faith in the almost divine power of analysis' (p. 209). The personal analyst's grandiosity, faith and power appeal to the analysand whose grandiosity and power strivings receive vicarious gratification. For those with a hunger for an idealized object because of early or even later losses, the gratification of emotional and mystical appeal is beyond the capacity of the analysand to renounce. If the analysand is taking care of the personal analyst through a role reversal, he feels extremely powerful. In some cases, these dynamics hide an early loss on the part of both that motivates the clinging mutual dependency.

Emotional over-gratification and abandonment of neutrality also feed and trap the narcissistically deprived analysand who may want to let go but cannot. Such gratification is deceptive and may in fact be a reaction formation against the personal analyst's underlying narcissism and aggression. I believe that the warmest among us feel comfortable in the framework of neutrality, which serves as a restraint against over-gratifying the patient, and permits us to experience our loving impulses in fantasy, thereby making them available for ego-based, growth-producing therapeutic work. Empathic, emotional involvement and non-attacking interpretations are most effective tools in an analysis conducted from a neutral framework. Neutrality can be experienced as safe in thatit offers the maximum freedom for verbal expression of all impulses without the fear that the analyst will be seduced into responding.

Part of working effectively on these issues involves honest acknowledgement of a range of feelings. If one can process feelings of aggression in addition to loving ones toward one's patients, if one can acknowledge grandiosity, narcissism, envy, power and control strivings as part of the range of possible countertransference reactions that need to be processed consciously rather than denied and split off, one is on the way to being able to conduct an adequate analysis of one's patients.

SUMMARY

Narcissistic problems in analysts reflect the same dynamics as they do in patients. Splitting, projection, denial, and ego fragmentation defend against the experiencing of painful affects. Idealization and grandiosity, aggression, power, and exhibitionism and their opposites are the essential issues in narcissism. Psychoanalytic work can be extremely gratifying to the analyst's narcissism, leaving little incentive to resolve it personally or in one's patients. Politics and other extra-analytic situations can become the repository of split-off hostile and devaluing feelings while the personal analyst remains idealized.

Narcissism in analysts is relatively unexplored in the vast literature on countertransference, indicating that it arouses shameful and secretive reactions. Case material illustrated the points made. The overall conclusion is that the problem is a difficult one, particularly when the transference involves symbiotic features and mutual gratification. Working through narcissism should be central in the analyst's personal analysis in order to foster high levels of ego integration, self development, and analytic ability.

My thanks to my colleagues Emily Anne Gargiulo, Donald Whipple and Cecele Kraus.

REFERENCES

Blum, H. P. (1982). Theories of the self and psychoanalytic concepts: discussion. J. Amer. Psychoanal. Assn., 30: 959--978.

Broucek, F. J. (1982). Shame and its relationship to early narcissistic developments. Int. J. Psvchoanal.,

63: 369--378.

Cooper, A. M. (1981). Narcissism. In American Handbook of Psychiatry, vol. 7, ed. S. Arieti & H. Brodie, pp. 298--3 16.

Eigen, M. (1981). Guntrip's analysis with Winnicott:

a critique of Glatzer and Evans. Contemp.

Psychoanal., 17: 103--1 12.

Finell, J. 5. (1984). Projective identification; mystery and fragmentation. Current Issues in Psychoanalytic Practice, 1(4): 47--62.

Freud, S. (1900--1901). The interpretation of dreams. S.E. 4--5.

(1920). Beyond the pleasure principle. SE. 18.

(1937). Analysis terminable and interminable. S.E. 23.

Gear, M. C. et al. (1981). Working Through Narcissism--Treating Its Sadomasochistic Structure. New York: Aronson.

Gedo, J. E. (1975). Forms of idealization in the analytic transference. J. Amer. Psychoanal. Assn., 23: 485--505.

Glatzer, H. T. & Evans,W. N. (1977). On Guntrip's analysis with Fairbairn and Winnicott. Int. Psychoanal. Psychother., 6: 81--98.

Greenacre, P. (1966). Problems of over-idealization of the analyst and of analysis. In Emotional Growth. New York: Int. Univ. Press, 1971, pp. 743-761.

Grinberg, L. (1962). On a specific aspect of countertransference due to the patient's projective identification. Int. J. Psychoanal., 43: 436--440.

Grotstein, J. S. (1981). Splitting and Projective Identification. New York: Aronson.

Grunberger, B. (1979). Narcissism: Psychoanalytic Essays. New York: Int. Univ. Press.

Guntrip, H. (1975). My experience of analysis with Fairbairn and Winnicott. Int. Rev. Psychoanal., 2: 145-156.

Hanly, C. (1982). Narcissism, defence and the positive transference. Int. J. Psvchoanal., 63: 427-444.

_______ & Masson, J. (1976). A critical examination of the new narcissism. Int. J. Psvchoanal., 57: 49-66.

Kernberg, 0. F. (1974). Further contributions to the treatment of narcissistic personalities. Int. J. Psychoanat., 55: 215--240.

_______ (1975). Borderline Conditions and Pathological Narcissism. New York: Aronson.

______ (1976). Object-Relations Theory and Clinical Psychoanalysis. New York: Aronson.

(1980). Internal World and External Reality:

Object-Relations Theory Applied. New York: Aronson.

Klein, M. (1957). Envy and gratitude. In Envy and Gratitude and Other Works. 194 6--1963. New York: Delta Books. 1977, pp. 176--235.

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

_______ (1977). The Restoration of the Self. New York: Int. Univ. Press.

_______ & Wolf, E. S. (1978). The disorders of the self and their

treatment: an outline. Int. J.Psychoanal. 59: 413--425.

Langs, R. (1975). Therapeutic misalliances. Int. J. Psychoanal. Psychother., 4: 77--105.

McDougall, J. 980). Plea for a Measure of Abnormality.

New York: Int. Univ. Press.

_____(1982). The narcissistic economy and its relation to primitive sexuality. Contemp. Psychoanal., 18: 3 73--396.

Miller, A. (1981). Prisoners of Childhood. New York: Basic Books.

Racker, H. (1957). The meanings and uses of countertransference. Psychoanal. Q., 26: 303--357.

Rangell, L. (1982). The self in psychoanalytic theory. J. Amer. Psychoanal. Assn., 30: 863--891.

Robbins, M. (1982). Narcissistic personality as a symbiotic character disorder. Int. J. Psychoanal., 63: 457-473.

Rosenfeld, H. (1964). On the psychopathology of narcissism: a

clinical approach. In Psychotic States.London: Hogarth Press,

pp. 169--179.

_______(1983). Primitive object relations and mechanisms. Int. J. Psychoanal., 64: 261--267.

Saperstein, J. & Gaines, J. (1978). A commentary on the divergent views between Kernberg and Kohut on the theory and treatment of narcissistic personality disorders. Int. Rev. Psychoanal., 5:

413--423.

Searles, H. F. (1975). The patient as therapist to his analyst. In Tactics and Techniques in Psychoanalytic Therapy, vol. II: Countertranference, ed. P. Giovacchini. New York: Aronson.

Segal, H. (1983). Some clinical implications of Melanie Klein's work: emergence from narcissism. Int. J. Psychoanal., 64: 269--276.

Spruiell, V. (1974). Theories of the treatment of narcissistic personalities. J. Amer. Psychoanal. Assn., 22: 268--278.

Stolorow, R. D. & Lachmann, F. M. (1980).

Psychoanalysis of Developmental Arrests - Theory and

Treatment. New York: Int. Univ. Press.

Tower, L. E. (1956). Countertransference. J. Amer.

Psychoanal. Assn. 4: 224--255.

Wolf, E. S. (1979). Countertransference in disorders of the self. In Countertransference, ed. L. Epstein & A. H. Feiner. New York: Aronson.

Copyright (c) Janet Schumacher Finell

41 Fifth Avenue
New York
NY 10003
(M.S. received May 1984)