THE MILIEU OF REAL RELATIONSHIPS IN RESIDENTIAL TREATMENT Michael S. Weiner, M.D. Freud writes that there are three impossible professions: Governing, Teaching, and Healing. Residential Treatment involves a complicated combination of all these. Little wonder then it is generally so poorly accomplished, so deservedly suspect a modality, often reserved for difficult patients who have failed in other forms of treatment--usually multihandicapped children or aggressively acting out adolescents. Yet, in our experience, competently conducted residential therapy is a powerful remedy and the treatment of choice for this group. [1] Early proponents of this form of treatment, August Aichhorn (1925), Bruno Bettleheim (1949), and Fritz Redl (1957), along with such less central figures as A. S. Neil (1960) and Lyward Burn (1956), have written convincingly of the therapeutic usefulness of the child's day- to-day experiences in a well-managed milieu. The descriptions of their pioneering work are quite similar. One is struck with the unitary quality of the settings they describe, consisting of a united and dedicated staff, loyal to a charismatic Director representing certain principles of treatment. There is little discussion of modalities of treatment or departmentalized services. Mostly their books report interactions of individual children and groups of children with each other and with adults. There are many descriptions of encounters between individual children and the Director, usually at unscheduled times in the milieu, precipitated by events that could not have been predicted but which become crucial therapeutic occasions. In general, the emphasis is on the development of (frequently transference) relationships and the use of these relationships and their vicissitudes for interpretation of intrapsychic conflict. In what I call classical milieu therapy, the essential feature is the focus on the evolution of relationships in the context of which children grow as opposed to the focus on various forms of treatment applied to the child. The gifted analysts who developed milieu therapy did not, however, have available the systematic theoretical understanding of borderline and narcissistic personality disorders which has evolved only more recently. Consequently, their form of psychoanalytic/interpretive/permissive milieu therapy had better results with individuals burdened with problems of development at the oedipal level than those children more frequently seen in residential treatment suffering from earlier occurring disturbances of object relations. Over the past years, there has been a trend away from permissiveness and interpretation in the milieu with increased attention to setting limits, behavior managing and modification, realistic consequences, and reality therapy. Some centers have abandoned the analytic point of view with varying degrees of lip service while others have preserved this approach in individual psychotherapy sessions with varying degrees of isolation from the life-space. Meanwhile, there have been significant advances in the psychoanalytic understanding of separation-individualization (Mahler, 1975), borderline personality organization (Kernberg, 1973), and narcissistic personality disorder (Kohut, 1971; Kernberg, 1973, 1976). Insights badly needed by the early analytic pioneers in milieu treatment of antisocial children are now available. However, the current trend away from using the milieu itself for the psychotherapeutic work of activating, understanding, and modifying problematic relationships prevents our using the increased understanding of character pathology with the many patients in residential treatment no suitable for psychotherapy. In an effort to preserve the values of classical milieu therapy, this paper attempts to elaborate its essential organizing focus on the therapeutic use of well-understood and controlled relationships in the life space. In doing so, we wish to exchange some of the approach to therapeutic control and use of relationships which derives from traditional psychoanalytic psychiatry for the approach of the existential school of psychiatry as understood by Havens (1973). I hope to show that the existential approach is more naturally suited to the conditions of the therapeutic milieu with its required levels of activity, affect, and behavior management, but still provides for activation and reorganization of self-object relationships distorting intrapsychic development. A related paper attempts to clarify with clinical illustrations the application of recently elucidated psychoanalytic insights to the conduct of milieu therapy. We will focus here, however, on the general organizing principles and process of milieu treatment. [2] The question most frequently asked of us by psychiatrists is, "What treatment modalities do you use?" Milieu therapy uses techniques and insights which have come from psychoanalysis, behavior modification, education, reality therapy, group work, and group therapy. However, milieu therapy uses none of these as a separately structure modality. Too often in residential treatment, these structures become resistances to the real relationships which we see as the main focus of our work with children. Psychoanalytically oriented individual therapy takes on this destructive function when a child wishing to share his thoughts and feelings is told, "Don't tell me now, save that for your therapist on Monday." Similarly, Behavior Modification can become a way of mechanically managing children from an emotional distance. Even what I call the "fast-footed child care worker technology" as usefully taught by Redl and Wineman (1957) in their book, The Aggressive Child, and by Trieschman (1969) in his book, The Other 23 Hours, can become a resistance to a real personal encounter. Instead, we insistently focus on the development of real relationship between children and adults as the essential medium of therapeutic work. This emphasis on real relationships as the main modality of treatment is likely here to raise even more persistent questions about the psychiatric foundations of milieu therapy. In his recent book, Approaches to the Mind, Leston Havens (1973) has described the four major psychiatric disciplines: The Objective Descriptive as represented by Kraepelin and Bleuler, the Psychoanalytic as represented by Freud, the Interpersonal as represented by Sullivan, and the Existential as represented by Jaspers, Minkowski, and Binswanger. Havens' book brilliantly clarifies the sources of data, basic assumptions, oversights, limitations, and particular usefulness of each school, advocating a more informed and selective eclecticism. Most clinical psychiatrists employ all of these approaches in their work, using whichever is most suited to the task at hand. Good residential treatment also requires use of all four of the psychiatric approaches, but the discipline of existential psychiatry is most suited to the main tasks of the milieu. I am most reluctant to associate myself in any way with the wild, flamboyant practices often linked to existential psychiatry, or with representations of this field other than Haven's restrained presentation defining the existential method in relationship to the other major psychiatric schools. Nor do I choose this approach out of personal preference. However, as described by Havens, existential psychiatry represents a highly disciplined approach to patients, which better than the other three major psychiatric schools, provides technical solutions to certain specific problems of milieu treatment and a systematic means of controlling the intense real relationships which necessarily comprise this form of treatment. This is not the appropriate place for even a brief summary of existential psychiatry, even if I could presume to improve on Havens' cautiously lucid explication of this complicated subject in his book and in his article (1974), "The Existential Use of the Self." Suffice it to say here that he discusses in detail the elements of the existential approach. These are: Being and Staying as ends themselves, the commitment to accepting appearances, translating and extending a shared grasp, "Keeping looking," the willingness to confront and change, all components of existential psychiatry which outline a clinical technique. Despite this inadequate recounting and my failure better to recapitulate Havens' well-crafted argument establishing this school as a disciplined psychiatric method, I would like to discuss here the particular usefulness of the existential approach in residential treatment, not to advocate existential approach in residential treatment, not to advocate existential psychiatry which has its own limitations: but in order to give an idea of the special problems of milieu treatment for which a systematic approach to the therapeutic use of real relationships is the solution. The five special problem areas of residential treatment for which existential psychiatry provides a disciplined attack are: ..First, the special countertransference problems involved in round-the- clock intimate and active work with child and adolescent patients. ..Second, the problem of managing and sharing the therapist's affect, access to which is especially important for child and adolescent patients in residence. ..Third, the problem of assigning realistic and therapeutic priorities to the meanings of behavior in the life space. ..Fourth, the problem of treating personality disorders involving vertical splitting. ..Fifth, the problem of the therapeutic relevance of love. In his book, Havens (1973, pp. 133-134) traces the development of existential psychiatry from Jaspers through Minkowski's crucial encounter with a depressed paranoid psychotic man whose private physician he had become and with whom he lived intimately for two months: Although my patient's case is relatively commonplace from the clinical point of clinical point of view, the circumstances in which I observed him were not. I have already said that I lived with him for two months. Thus, I had the possibility of observing him from day-to-day, not in the mental hospital or a sanitarium but in an ordinary environment. His manner of reacting to normal external stimuli, his ability to adapt to the demands of daily life, the variability of his symptoms and the particular annoyances are set out more clearly under such circumstances. This circumstance is accompanied by another. One cannot maintain a medical point of view 24 hours a day. We react to the patient as do other people around him. Compassion, gentleness, persuasion, impatience, and anger appear in turn. Thus, in the above circumstances, I not only observed the patient but had the possibility of projecting his psychic life on mine every moment. The psychoanalytic solution to these intense countertransference problems--the solution of distancing from the patient, of becoming and anonymous, and of reflecting inwardly is not available in residential treatment. Indeed, there are residential centers which attempt this so that the crucial professional treatment staff remove themselves from the milieu, leaving children and child care staff locked in painful relationships and further deprived by the feeling that the blessings of healing, understanding, and power exist outside the milieu of real relationships. In such centers, the area of distance between children and child care staff in the milieu on the one hand, and the professional treating (or perhaps more correctly, retreating) staff on the other becomes a fertile ground for development of transference and countertransference distortions which become so institutionalized and fixed as to resist any resolution. Worse, child care staff in such centers often model themselves after the professional staff and learn to keep "professional" distance from children. As a result, children often go through treatment without access to normal human relationships. Valuable as psychoanalytic insight is to our work, the psychoanalytic method is destructive to children when practiced in the life-space of the milieu. Individual psychotherapy may be useful adjunct to milieu treatment but liberal use of individual therapy in residential centers introduces, in addition to certain advantages, the danger of a withering of the milieu's therapeutic potency from neglect as staff wrongly assume that its therapeutic tasks will be accomplished in the scheduled individual sessions. Evolving from Minkowski's early attempt to live with a patient, into a precise technique of therapy existential psychiatry presents its own way out of these difficulties with a method more appropriate to the realities of residential treatment. The core method of existential psychiatry involves Being and Staying: that is, to join the patient where he is, to feel what he feels, to be with the patient. Any objectification of the patient, any distancing from him is a breach of existential technique into which countertransference distortions may enter. As Havens (1974, p. 10) points out, in the existential method, "the countertransference is not resolved in the private world of the therapist's thought, but, as with resolution of the transference, in the encounters of person to person." Consequently, in milieu treatment, the most valued work for staff is getting with kids, being with kids, both at the level of actual activity and in terms of emotional closeness. We are vigilant to spot when a child has "turned off" an adult who has been seeking to be with him. This observation stimulates more focused therapeutic work, aimed at bringing child and adult together. This constant concern over being turned off by a child, of wanting to stay with him and being unable to, is roughly equivalent in our work, to the role of the analysis of resistances as the organizing feature in analytic therapy. This focus of attention and activity also fits well with our belief that a child in group care, who is not special to anyone, carries great risk of distorted development. In adolescent psychiatry in general and residential treatment in particular, the open expression of affect by patients and therapists plays a more crucial role than in other therapies. As Havens contends, existential psychiatry more than any of the other approaches, gives the clearest technical guidelines for the control of the therapist's use of affect. In the existential method, the therapist is not permitted to express whatever affect he would like (as some wild existential analysts have interpreted) but instead strives to discipline himself through Being and Staying to feel only what the patient feels. When this is not possible, when the therapist wishes to stay with the patient but cannot under the present circumstances, then, and only then the therapist may express affect in the interest of Being and Staying: that is, in order to bring patient and therapist together. These two rules govern the use of affect in the existential method and usefully govern our reactions to adolescent patients in the exposed closeness of the milieu. Another reason for our selecting the existential approach as most suitable for residential treatment is its insistence on taking whatever the patient says and does at face value, as opposed to other psychiatric schools (the objective descriptive, psychoanalytic, and interpersonal) all of which Havens (1973) notes prize the search for hidden processes. Understanding the latent, symbolic, and covert meaning of behavior has of course enormous value in our work. However, the focus of these obscure meanings in a therapeutic setting wedded to a child's real living space can be very disorganizing. In many residential centers, there is extreme emphasis on understanding the symbolic meaning of behavior--all behavior is responded to as though symbolic and the child is deprived of the realistic feedback needed to shape behavior. Indeed, often his behavior is stripped of realistic impact, resulting in iatrogenically intensified regression and disorganization frequently confused with the child's illness. Many centers have dealt with this problem by reintroducing behavioral consequences designed to modify behavior (often rather mechanically applied by one group of staff) while another group of staff continues to "understand" the hidden meanings of behavior in symbolic terms. This leads to intolerable splitting up of the child's experience. We prefer to take what the child says and does at face value, which best allows us to provide at once both realistic behavioral consequences and to use the existential method to make behavior a meaningful part of the therapeutic encounter. The majority of children in residential treatment suffer from either established or incipient personality disorders, especially those of the antisocial, borderline, and narcissistic types whose dynamics are best described by Winnecott (1956), Kernberg (1967), and Kohut (1971). I cannot adequately discuss here their complex formulations or the background of their work in the contributions among others of Freud, Klein, Mahler, and Jacobsen. Simplistically speaking, they have extended Janet's early idea that mental illness results from disassociation of components of mental life because of insufficient psychic energy to a more precise understanding of the stressed within the structure of personality which lead to various patterns of splitting along natural horizontal and vertical cleavage lines. We see in may of our children various combinations of splitting off of elements of id, super-ego, and grandiose self--(with frequent splitting of good and bad libidinally vs. aggressively determined) self-object representations. Whole relationships in the context of a coherent milieu are particularly helpful for activating the split apart components and integrating them along with other elements of psychic structure into the maturing mainstream of a child's developing personality. The prevalent orthopsychiatric or what I call the "combination" approach to residential treatment which divided up the child's care and therapy between various services which separately meet his needs can be a disaster for this integrative work. At any rate, we do not seek, as does psychoanalysis, a therapeutic alliance only with the observing ego and other healthy portions of the patient's personality. Instead, we strive, as in existential psychiatry, to get with the patient, to reach and stay with all aspects of the child including the most hateful. We often form an alliance with the grandiose self or with the split off contents of the id. Thus, the therapeutic practice at times in our center of wholesale flattery, praising, affectionate holding, tickling, and visual and vocal admiration of our young patients and our otherwise promoting in them a warm glow of narcissistic contentment. The most important reason for our choice of the existential method as the main psychiatric approach in residential treatment is the most difficult to discuss. Children, even normal children, certainly children who have been given too little, do not thrive in group care without love. I believe that a significantly ill child in residential treatment who is not special to someone will not get better. (Indeed, a particularly unappealing child's failure to become special to someone has become one indication among others for regularly scheduled individual psychotherapy sessions as a parameter of our preferred treatment approach.) Of all the schools of psychiatry, only the existential deals systematically with what Havens calls "the educational component of love." He bases his discussion of this on Freud's classic description of the dynamics of resolution of the oedipal crisis. Havens (1973, pp. 301- 303) then concludes: The existential method is the principal attempt to harness and bring into therapeutics the educational component of love...This love's capacity to engender identifications. One of the technical requirements for this capacity to emerge is closeness, the being and staying. Another is the crisis which precipitated the need for the encounter, and then come the successive emotional crisis which being and staying necessitate. Identifications occur in a needful, emotional field: The personality must be in solution as it were in transit, then the therapist leaves pieces of himself behind in the patient. I suggest that leaving pieces of ourself behind in the patient is the chief mechanism of existential treatment. The method of existential psychiatry sets up the conditions necessary for the leaving behind. Actually, Havens' description greatly underestimates the educational component of love. We will later discuss Kohut's account of the beneficial effects of love in fostering cohesiveness of the self. His description of the effects of the mother's admiration of the infant on the development of narcissism, which I will elaborate subsequently, actually comes close in meaning to Havens' paraphrased quote of Buber's existential description of the central therapeutic act of "boldly swinging into the patient." In our work, we take great pains and some risks to favor the necessarily spontaneous evolution of real ties between children and adults. While the needs of the child take precedence, those of the adult must also be at play in a truly mutual relationship. On the dimension defined at one extreme by hurting and not caring about a child, loving and giving are only intermediate points along the way to the opposite pole of letting the child gratify an adult. In our experience no achievement so powerfully increases self esteem as the child's simultaneous awareness of being himself and pleasing a valued other. Consequently, while we agree with Shields' (1971) significant observation that the ability to cure antisocial children depends on the staff's capacity to bear pain, we place enjoying children at the top of our list of therapeutic activities. The central element of treatment is the child's capacity for growth. Elsewhere (Weiner, 1974) I have entitled this discussion, 'The Value of the Child Himself in the Stabilization of Self-Esteem.' We respect the child's positive potentials and his role in shaping the course of their unfolding which determines treatment. Indeed, our most important resource in a conviction of the intrinsic capacity for positive growth of each child who comes for care. This conviction, more than simply faith in the prospects of the new and often initially unappealing child, flows from an appreciation of the progress of the children already well along in treatment. Reminding ourselves of the undesirable features these children presented on admission as we enjoy their current successes has become an important technique of maintaining, often in the face of discouraging circumstances, our conviction about a particular child's as yet hidden virtues. But this is somewhat misleading; what is most important is not our faith in the capacities of all children, but our appreciation of the special value of a particular child which emerges from an actual encounter. Indeed, it is the child's ability to arouse our concern which makes us hopeful. We experience three distinct phases of treatment, described most simply as Initial, Middle, and Ending. These phases of treatment, defined more elaborately later, are actually manifestations of the normal evolution of relationships and group formation [3] and the related existential stages of Being, Staying, and Leaving Behind. In describing these stages of residential treatment, I have added to the purely existential issues consideration of transformations of narcissistic development as described by Kohut (1971) because this frequently failed aspect of growth is so important to most of our young patients. [4] The initial phase of treatment in our facility involves a complex series of interactions, variations in the evolution of which, more than anything else, specify the problems requiring attention. In our setting the initial phase of the treatment can be resolved into three separate elements which for certain reasons I define in terms of the staffs' therapeutic activity and countertransference (and only parenthetically in terms of the child's transference activity.) [5] The three stages of the initial phase of treatment are: 1) Our idealization of the child (while he experiences the honeymoon and gradual revealing of his grandiose self). 2) Our holding and staying (while the child tests the stability of treatment relationships). 3) The transformation of our idealization (while the child reveals his level of narcissistic development). For simplicity I will discuss these three elements as though they constituted three distinct stages of the initial phase of treatment. The first act of treatment "boldly swinging" into the child (to use Havens' (1974) recent paraphrasing of Buber (1957). As the child responds there occurs a process which I call the Idealization of the Child. This idealization is not based on the child's initial hiding parts of his real self from us, for we are mindful of the history, but on whatever he does show of himself, which together with his history, awakens our (of course narcissistically determined) fantasies of his idealized potential. Thus we invest our own self-esteem in the child. At this stage we are not so much interested in what we can get the child to do as we are in whatever he does. In this way our attitude toward the child parallels the mother's admiration of her baby. From this point, behavior modification enthusiasts can trace most of the rest of our treatment in a simple way. What matters is the gleam in our eye and the child's relating this to his own activity. The idealizing relationship in residential treatment is an active analogue of the analyst's more neutrally toned, yet still actively empathic, response to the mirror transference as described by Kohut (1971). Its goal is the therapeutic activation of the grandiose self. Some of the following comments by Kohut clarifying this process as it occurs in analysis are relevant to residential treatment: In this narrower sense of the term of mirror transference is the therapeutic reinstatement of that normal phase of the development of the grandiose self in which the gleam in the mother's eye, which mirrors the child's exhibitionistic display, and other forms of maternal participation in and response to the child's esteem and, by a gradually increasing selectivity of these responses, begin to channel it into realistic directions. (p. 116) The most significant relevant basic interactions between mother and child lie usually in the visual area: the child's bodily display is responded to by the gleam in the mother's eye... The acceptance of the child's body (especially oral and perioral region by tactile responses) leads under favorable circumstances to a basic equilibrium in the realm of the narcissistic cathexis of a cohesive body self. If the mother, however, recoils from the child's body (or cannot tolerate lending her own body to the child for his narcissistic enjoyment through the extension of his narcissistic cathexes to include the mother's body), then the visual interactions become hyper-cathected and, by looking at the mother and by being looked at by her, the child attempts not only to obtain the narcissistic gratifications that are in tune with the visual sensory modality but also strives to substitute for the failures that had occurred in the realm of physical (oral and tactile) contact or closeness. (Kohut, 1971, p. 117) We may thus conclude that the mother's exultant response to the total child (calling him by name as she enjoys his presence and activity) supports, at the appropriate phase, the development from autoerotism to narcissism--from the stage of the fragmented self to the stage of the cohesive self--i.e., the growth of the self experience as a physical and mental unit which has cohesiveness in space and continuity in time. (Kohut, 1971, p. 118) The mirror transference in the stricter sense of the word is, nevertheless, close to being a therapeutic reinstatement of a normal phase of development...and, in a correctly conducted analysis [or residential treatment] the mirror transference tends to become more and more akin to the normal developmental phase, that is, the sadistic elements diminish and the demands for affection and response take on the vigor, and approximate giving the pleasure, which is encountered in the corresponding phase-appropriate interactions between parent and child. (Kohut, 1971, p. 125) One may say that the experience of a unitary self, in consequence of a reliable narcissistic cathexis of the self image, is an important precondition for a cohesively functioning ego; that by contrast, the absence of such a cathexis tends to lead to disordered ego function; and, finally, that the narcissistic cathexes of a mirror transference may remedy the disturbance of the ego, i.e., improve ego functions via the intermediate step of supplying cohesiveness to the self. (Kohut, 1971, p. 132) Translated into terms of existential psychiatry, Kohut's account is a description of Being with the grandiose part of the self as it matures from infantile to childlike forms. In the second stage of the initial phase, the child tests the authenticity of our concern by showing some of the behavior which complicated his relationships in the past. We often welcome this revelation of a real part of the child and he may settle his inquiry feeling reassured. However, if the child has experienced significant past disappointment in relationships, he may desperately test our ability to withstand his hate with worrisome attempts to dissolve the treatment relationship. Often, for many youngsters whose life-problem has been repetitive failure of the nurturing environment, the treatment, Winnicott (1956) has stated, is simply to manage--to hold the child and the treatment setting together. In practice, this cannot be productively accomplished without love for the child, which enables us to see the valuable part of himself he reveals by his apparently negative behavior. The temper tantrum is a frequent way of testing the depth of the adult's valuing of the child and discovering whether adult and child can stay together. Its phenomenology and dynamics are crucial to the management of the second stage of treatment. The key to the resolution of this is not simply the child's convincing himself that he can trust the staff to control and tolerate him, but that he come to terms with his own guilt and an adult's anger along with an appreciation of himself being valued. Staff's enjoyment of the child, often aided by their ability to set firm limits when necessary, and more importantly, their acceptance of his "gestures of restitution" (Shields, 1971) which accompany the resolution of this stage, are essential elements of successful treatment. In this stage, the existential focus on staying, the willingness to confront and to change in order to stay, along with acknowledgement of patient's sacrifices to stay, are crucial aspects of treatment. [6] Having decided on the trustworthiness of the treatment relationships, the child and his workers focus on questions relating go the third stage of the initial phase of treatment. (These questions had actually already been at play in the second stage.) The child is no longer new, nor does he fear the absolute loss of the treatment relationship, and the staff has modified their idealization by an encounter with the real child of the second stage. Can the child sustain self-esteem in the face of the partial withdrawal of their admiration and a relative loss of relationship? At this stage both the child's realistic limitations and the realities of staff's level of energy and obligations to other children become more compelling features so that the relationship becomes more real. I call this third stage of the initial phase of treatment the "Transformation of Idealization." As the child relaxes from the fear and turmoil of the testing stage and the staff modify their idealization of him, we can assess the child's actual level of narcissistic development (Kohut, 1971) and his ability to maintain self-esteem in the face of interruptions and imperfections of the idealizing relationships. In this stage also there must be willingness to stay, to confront and to change in order to stay. The child makes sacrifices, and, because of the closeness, parts of the therapists are left behind in the patient. The vicissitudes of these relationships over time, the child's growing capacity to maintain esteem without them, his moving out to other areas of accomplishment, and his overcoming of developmental defects, compose the middle phase of treatment. Frequently, an improvement in appearance marks the transition from the initial to the middle phase. Anna Freud has said, children who are loved become more beautiful. Kohut (1971) comments that people with narcissistic personality disorders have poor circulation in skin and suffer from dryness, rashes, and feeling of coldness. We have noted that improvement in hair texture and complexion coincide with the improved appearance. Indeed, many of our previously deprived improving children manifest a pink glow suggestive of a reactive hyperemia. The maintenance and intermittent lapses of the idealing relationships permit the unfolding of the child's potential. As he indicates his interest in working on areas of past failure, frequently by resistance which compels our attention, the child is ready to use more specifically planned help. Often, before he attacks well-established and anxiety laden problems, he first chooses to explore some new areas of achievement. Sometimes this involves going back to learn skills appropriate to an earlier developmental phase. We frequently observe during this period that a youngster may appear more age-appropriately childlike. Esteem from accomplishments in more neutral areas support the child's eventual attack on more complicated and resistant problems. The development of skills, working on problems, and diversification of relationships are the activity of the middle stage. Its theme is the child's use of these accomplishments gradually to relocate within himself the value of himself in world which evolved in the special idealizing relationships of the initial phase of treatment. We frequently promote this process by changes of living group or staff in the middle to late stages of the middle phase. During the middle phase, children invite us to help them with specific problems which require quite individualized treatment planning. Frequently, however, the child engages in predominately inconspicuous and benign activity so that it is difficult to demonstrate active treatment except for evidence of his gaining self confidence and skill. We sometimes cannot write detailed treatment plans for our work in this often prolonged period and reluctantly resort to describing the process as "allowing the child to consolidate." Actually this is the phase of treatment during which the child most changes and becomes himself. His therapeutic progress during this time depends on the maintenance of the idealizing relationships--the gradual attention (and well-timed intensification) of which parallels his grown (and crises of narcissistic transformation)--but the larger aim of this work is the stabilization of self-esteem on the basis of real accomplishment. In order for the child to maintain treatment gains after leaving, it is crucial that expectations during this stage be set very high and that he be expected to master the difficulties of bureaucracy and of complicated and difficult social relationships. Turning to face the realistic problems of the family and gaining new perspective on the problematic relationships which led to the child's placement requires a relatively high level of narcissistic development. Consequently, this often occurs relatively late in the middle phase and signifies that successful termination may be possible. At this point the child can best participate in discharge and post discharge planning. The ending phase of treatment begins in the context of setting a termination date. Setting the date sometimes stimulates child and family to deal seriously for the first time with the problems of their relationship. We expect to see a brief return of old symptoms and other evidence of the child's and family's conflict about his return home. In dealing with these conflicts, reviews of the child's experiences, and working through the loss of leaving, comprise the main work of this phase. When the middle stage has been successful, that is, when the child's self-esteem has become detached from relationships with staff and rest on his real accomplishments--and especially when parents can acknowledge (admire) his achievements, we then see in the child much less pain and manifestation of loss at termination than is traditionally expected. In fact, we expect children in this stage to help and give to others including the adults who have been close to them. We view termination of successful treatment as a graduation, which it is indeed--emphasizing achievement rather than injury--stressing not so much what the child leaves behind, but what he takes away as his own. Separations are still painful. For a moment one might think that leaving behind so much of ourselves in patients would be depleting. This of course is not the case. Even the closeness involved is not exhausting. However, something important is lost whenever a child leaves, especially when treatment has been successful and the child goes as a whole person to a good environment. Havens (1974, p. 2) quotes Tarachow's (1962) representative attack on existential psychiatry from the psychoanalytic point of view: "If the therapist treats the patient as real he is using the patient to overcome his own sense of loneliness and sense of abandonment... The most basic temptation between two individuals is the urge to regress in the character of object relations and to dissolve boundaries and fuse." This idea of course ignores the forces present in healthy development which lead mother and child to move beyond symbiotic ties. The temptation is real, though, and is controlled throughout treatment by our and the child's awareness of the inevitable eventuality of his leaving as dictated by his developmental needs. His actual going, however, definitively disappoints again our narcissistically determined fantasies which we have invested in him. This experience at once purifies and diminishes the power of our investment in children--and that is a value and the limitation of our method. Footnotes: [1] Wintrob (1975) has discussed the differences between residential treatment, hospital treatment, and the considerations which would indicate choice of one or the other as the optimum fair of treatment for emotionally disturbed adolescents. [2] I must make clear here that I will discuss a particular approach within the field of residential treatment for which I would prefer to reserve the term milieu therapy despite its frequent misapplication to settings that are less than therapeutic and despite the fact that many proponents of milieu therapy might not favor my particular and sometimes personal representation. All I can claim to describe here is an approach to residential treatment enriched by contributions from similar work by many others, used by myself and co-workers to treat emotionally disturbed children and adolescents in a residential setting. [3] Garland (1965), Semrad (1951), and Mann (1951) have described similar phenomena from different points of view. Erickson's (1950) stages of individual development also apply: Trust (first stage of initial phase: idealization/honeymoon), autonomy (second stage of initial phase: testing), initiative (third stage of initial phase, transformation of idealization), industry (middle phase). Garland's (1965) model based on observation of therapeutic activity and summer camp groups is particularly useful: preaffiliation, power struggle, intimacy, differentiation, and separation. However, I prefer not to separate intimacy and differentiation, which make up our middle phase, because I wish to emphasize the role of the intimate relationships in facilitating differentiation. [4] Kohut proports to describe dynamics of the relatively rare narcissistic personality disorders. He does not describe as well as Kernberg (1973) the essential pathological personality structure of such individuals, but his clinical approach to narcissistic disorders and his theoretical descriptions of developmental transformations of normal narcissism are quite useful in our work with individuals manifesting some degree of narcissistic regression or heightening of narcissistic defenses. [5] Kohut's (1971) insights evolve from observation of certain kinds of transference phenomena--data available only in the psychoanalytic situation. Similar observation is not possible in residential treatment--partly because of our necessary activity in meeting the child's needs--and some (e.g., Dowling, 1973) would question the prudence of transferring Kohut's ideas. Consequently, I do not discuss the manifestations of narcissism in terms of the child's (latent idealizing transference) relationship to the staff, about which we know too little, but instead describe events in treatment from the point of view of the staff's relationship to the child which we can observe--focusing on the staff's idealizing activity as a mirror to the child's need. [6] These crucial techniques in treating children with antisocial behavior parallel important strategies suggested by Kernberg (1973) in therapy with adult narcissistic personality disorders: focusing on remnants of the patient's love for others, including the positive side of his sometimes disruptive attempts to maintain hope of being helped by the analyst, and systematic confrontation of the patient's behaviors which cause distance and meaninglessness in the relationship with the therapist with the goal of supplying self-object representations which can sustain self-esteem. I have discussed these issues with clinical illustrations in a related paper (Weiner, 1977). References: Aichhorn, A. (1925), Verwahrlostse Jugend. Vienna: Internationaler Psychoanalyteches Verlog Wayward Youth. New York Viking Press, New York, 1935. Meridian Books, 1955. Bettleheim, B. (1948), Truants From Life. Glencoe, Ill. Free Press. Bettleheim, B., and Sylvester, E. (1949), Milieu Therapy; Indication Buber, M. (1957), Elements of the Interhuman. Psychiatry, 20:105- 113. Burn, Michael. (1956), Mr. Lyward's Answer. London, England. Harnish Hamilton. Dowling, S. (1973), Unpublished memorandum to members of the American Association of Children's Residential Treatment Centers. Garland, J. A., Jones, H. E., and Kolodny, R. (1965), A Model for Stages of Development in Social Work Groups, In: Explorations in Group Works. ed. S. Bernstein. Boston: Boston University School of Social Work. Havens, Leston. (1973), Approaches to the Mind, Boston, Mass: Little Brown & Co. ----- (1974), The Existential Use of the Self. The American Journal of Psychiatry, 131:1:1-10. Kernberg, O. (1973), Borderline Personality Organization and Pathological Narcissism. New York: Aronson. Kohut, H. (1971), The Analysis of the Self. (Psychoanalytic Study of the child. Monogr. 4). New York: International Universities Press. Maier, Fritz. (1976), personal communication. Mahler, M., Pine, F., and Bergman, A. (1975), The Psychological Birth of the Human Infant. New York: Basic Books. Murphy, L. B. (1974), Growing Up in Garden Court. New York, N.Y.: Child Welfare League of America. Neil, A. S. (1960), Summerhill. New York, N.Y.: Hart Publishing Co. Redl, Fritz, and Wineman, David. (1957), The Aggressive Child. Glencoe, Ill.: Free Press. Shields, R. (1971), A Cure of Delinquents. New York: International Universities Press. Tarachow, S. (1962), Interpretation and Reality in Psychotherapy. Int. J. Psychoanal, 43:377-387. Trieschman, Albert, Whittaker, James, and Brendtro, Larry. (1969), The Other 23 Hours. Chicago, Ill.: Aldine Publishing Co. Weiner. M. S. (1975), Long Term Treatment of the Severely Disturbed Adolescent: Residential Treatment versus Hospitalizations. This Journal 14:436-450. Winnicott, W. W. (1956), The Antisocial Tendency. Collected Papers. London: Tavistock, 1958, pp. 306-315.