Bulletin of the Menninger Clinic, 1997, 61:16-43.
Werbart, A. (1997). Separation, termination process and long-term outcome in psychotherapy with
severely disturbed patients. Bulletin of the Menninger Clinic, 61, 16–43. PMID:9066176
SEPARATION, TERMINATION PROCESS AND LONG-TERM OUTCOME
IN PSYCHOTHERAPY WITH SEVERELY DISTURBED PATIENTS1
Long-term consequences of working through of intratherapeutic separations were
studied in combined psychotherapy and milieu therapy with severely disturbed adult
patients. Seven of the 10 cases were consistent with the hypothesis about the
significance of therapeutic work with separation for the longitudinal outcome.
Inability to mourn was connected with less emotional and functional improvement.
Besides the therapist’s contribution in the middle phase, the patients’ different ways of
dealing with the trauma of termination were decisive.
KEY WORDS: Separation, termination, working through, mourning, therapeutic relationship,
”J’ai pu définir le psychisme comme le résultat de la relation entre deux
corps dont l’un est absent.”
André Green (1991). Méconnaissance de l’inconscient (science et
psychanalyse), p. 203.
”I cannot play with you,” the fox said. ”I am not tamed.” [...]
”What does that mean – ‘tame’?” [...]
”It means to establish ties. [...] if you tame me, then we shall need each
other. To me, you will be unique in all the world. To you, I shall be unique in
all the world [...] One only understands the things that one tames”, said the
fox. [...] ”If you want a friend, tame me...”
”What must I do, to tame you?” asked the little prince.
”You must be very patient,” replied the fox. ”First you will sit down at a
little distance from me – like that – in the grass. I shall look at you out of the
corner of my eye, and you will say nothing. Words are the source of
misunderstandings. But you will sit a little closer to me, every day ...”
The next day the little prince came back.
”It would have been better to come back at the same hour,” said the fox.
[...] ”One must observe the proper rites...”
Antoine de Saint-Exupéry (1943). The Little Prince, pp. 64ff.
The two opening quotations delimit a field between object loss and mutual attachment, an
area of particular difficulties in severely disturbed patients. It has often been claimed that the
therapist working with these patients represents the mother of the symbiotic relation, and a
1 This article may not exactly replicate the final version published in the Guilford Press journal. It is
not the copy of record.
”symbiotic crisis” in psychotherapy with psychotic patients was previously described by
Searles (1965; 1973), Feinsilver (1977; 1989), Pao (1979), Rinsley (1980), Goodrich (1984)
and Fritsch & Goodrich (1990). Stone (1961, passim) and Friedman (1969) had argued that
the therapist rather represents the mother of separation. Meltzer (1967) and Etchegoyen
(1991) reconsidered the psychoanalytic process in terms of identification and working through
of separation anxiety. From another tradition our thinking about the therapeutic relationship is
influenced by the work of Bowlby and Ainsworth on affectional bonds and object loss in and
beyond infancy (Bowlby, 1973; Ainsworth et al., 1978; for review of attachment theory see
Bretherton, 1991; Holmes, 1993; Goldberg et al., 1995). Scott (1987:60) recommended the
therapists ”review all the situations involved in therapy from the viewpoint of attachment and
separation [...] both currently and in the more distant past.” According to Searles (1986a; b)
intratherapeutic separations continually expose the borderline individual to the threat of loss
and difficulties with grieving. Rickman (1950) was probably the first to use separation
anxiety, awakened by intratherapeutic breaks, as a criterion of treatment progress. Some
research has been done on the effects of termination in psychoanalysis and psychotherapy
(Edelson, 1963; Firestein, 1978; Dewald, 1982; Gillman, 1982; Hartlaub et al., 1986; Kupers,
1988; Blanck & Blanck, 1988; Schachter, 1990; Becker, 1993; Tryon & Kane, 1995). In this
context, Pedder (1988) pointed out the inappropriateness of the term ”termination” with its
finite connotations. Janssen (1987:135; 173) describes the weaning from the psychoanalytic
therapy in the hospital setting in terms of separation, loss, mourning, and internalization of the
therapist, but he pays attention only to the short-term aspects of this process (up to 2 months
A previous investigation (Werbart, 1996), as well as clinical and theoretical evidence, led to
the hypothesis for the present study: Working through of intratherapeutic separations in long-
term treatment of severely disturbed patients, and of accompanying problems with
dependence and hostile feelings, would lead to longitudinal emotional and functional
development during, and after treatment. Lacunae in the therapeutic work on these areas and
retraumatization in connection with discontinuities in the therapeutic relationship would result
in persistent emotional and functional difficulties. According to the applied theoretical model,
the patient’s difficulties with separation, dependence, and hostility are gradually actualized
and enacted within the therapeutic relationship as well as in the universe of interpersonal
relations within the treatment environment. In this process, old patterns of relations to the
internal images of the primary objects, such as parents and siblings, are repeated in new
editions and in a different interpersonal context. This actualization and enactment can result in
a new traumatization, and contribute to tenacious holding on to distorted and regressive forms
of relating, but may also create possibilities for a working through and change. Repeated
experiences of the patient and therapist surviving separation, dependence and hostility are
decisive here. The termination process starts with the beginning of the treatment and
continues after the external interruption. The term ”separation” is used here in the second
sense as distinguished by Quinodoz (1991:26), applicable on the ”pre-Oedipal”, ”archaic” or
”narcissistic” level of object relations: both the absence and the presence of the object gives
rise to an unbearable painful perception of the other as non-ego. One of the principal aims of
treatment is to re-discover in oneself the capacity to experience – and survive – those feelings
which were banished from the emotional repertoire as a consequence of separation anxiety
and object loss. Quinodoz describes this process as the taming of solitude, a transition from a
wish to be an object of care to buoyance (in French: portance), a self-supporting capacity
making the patient independent of the object and capable of finding joy in ”flying with his
own wings” (Quinodoz, 1991:172).
The aim of this study is to investigate longitudinal consequences of more or less successful
working through of intratherapeutic separations in long-term treatment of severely disturbed
patients. The questions at issue are the following: (1) Is the work done with separation and
accompanying problems with dependence and hostile feelings related to the longitudinal
outcome, and, if so, how? (2) What factors in the patient and in the therapy contribute to
divergent patterns of interplay between the working through of intratherapeutic separations
and the longitudinal outcome? (3) Which obstacles to the working through of discontinuities
in the therapeutic encounter are characteristic for the existing setting? In order to answer these
questions a qualitative micro-analysis of repetitive patterns in consecutive cases was
performed. In closing, clinical implications for a wider spectrum of patients and settings are
discussed, as well as some relevant theoretical questions.
Treatment setting and subjects
The project was carried out at a Swedish therapeutic community for patients aged between 20
and 30 years of age with psychotic problems and/or severe personality disorders. The patients
were undergoing a combination of milieu and individual therapy. The treatment setting and
subjects have been described in detail in two previous papers (Werbart, 1995; 1996).
However, two factors in the design of the therapeutic environment should be mentioned here.
One was the rhythm of the staff’s presence during the weekdays, and their absence during nights
and week-ends. The other was the fusion of psychoanalytically oriented individual and milieu
therapy: The psychotherapists, known as the ”closest” in the domestic language of the
community, also participated in the day-to-day running of the institution, as well as in group
meetings. The therapists received qualified psychoanalytic-oriented supervision.
The target population comprised 10 in-patients consecutively admitted during the first four
years of operation of the institution, and who completed their treatment, 5 females and 5
males, with an average age of 25 years on admission, and an average stay of 47 months
(between 22 and 60 months). Four of these patients changed therapist during the course of
treatment, two of them once and two twice. Ten other patients, who discharged themselves
from the therapeutic community during this period after an average stay of 8 months (between
0.5 and 14 months), are subjects of a separate study (Tullhage & Werbart, 1995).
Case study method, research instruments, and data analysis
This longitudinal, prospective investigation was based on systematic case studies (Werbart,
1989). The research project started 2 years after admission of the first patients to the
therapeutic community. The first interviews, tests, and measurements were conducted 1
month (3 patients), 12 months (3 patients), and 24 months (3 patients) after admission, and
followed by data collection on each anniversary of admission, on discharge, at 2-year and 5-
year follow-ups. The patients are disguised and permission has been obtained for longer
Two forms of tape-recorded interviews were conducted with both patients and their therapists
during the treatment. The first was a modified version of the Chestnut Lodge Adolescent
Attachment Interview (Fritsch et al., 1987; Fritsch et al., 1992), and focused on development
of the patient’s feelings toward significant others, especially the therapist, in such areas as
Separation Difficulties, Emotional Dependency, Acceptance and Integration of Hostility,
Modulation of Affects, and the Core Problematic Area in the Therapeutic Relationship. The
second interview comprised the patient’s background and life history, coping with symptoms,
social functioning, education and work, friendship and love relations, contact with family of
origin, therapeutic relationship, the patient’s and therapist’s picture of one other, and
memories of critical episodes. This form of interview was also used at follow-ups with the
patients. Of the total of 185 interviews hitherto conducted with the 10 patients and their
therapists, 135 were analyzed (9 dyads early in the treatment, 9 dyads one year later, i.e. in the
middle phase, 10 dyads on discharge, 9 subjects at 2-year, and 8 subjects at 5-year follow-up).
Every interview was summarized by two or three independent judges according to a fixed
manual, differentiating between manifest content in the narratives, inferences and hypotheses.
These formulations were analyzed together at consensus meetings, and resulted in a consensus
summary of each case which, after having been referred to all the assessors led to a final
statement. The next step was to find gestalts in the material and to reconstruct the historical
development of repetitive patterns (Waelder, 1970).
All the cases studied were retrospectively classified according to the predominant patterns of
dealing with intratherapeutic separations during the treatment period. Pattern A, New
Solutions of Difficulties with Separation, Dependence, and Hostility, was defined as new
ways of handling separation, including the connected problems with dependence and hostile
feelings, both in the patient’s inner world and in the realm of interpersonal relationships,
accompanied by some working through of current strains. Pattern B, Retraumatization in
Connection with Separation, means unresolved separation crises, and persistent problems with
dependence and hostility left outside the therapeutic work. These difficulties were enacted in
the treatment situation and in connection with termination, leading to repetition of trauma on
account of both the patient’s and the therapist’s contributions. This classification was made
clinically on the basis of consensus summaries of Attachment Interviews.
The outcome was assessed by change in DSM III-R (American Psychiatric Association, 1987)
diagnosis between admission, discharge, and follow-ups, consumption of psychiatric care
(average of in-patient days) and psychopharmacological agents (average of chlorpromazine
equivalent per day) 2 years before the treatment, 2 years after discharge, and during the 5th
year of follow-up. Furthermore, the Global Assessment Scale (GAS; Endicott et al., 1976)
and the Strauss-Carpenter Outcome Scale (Strauss & Carpenter, 1972) were used at
admission, at the 2-year and 5-year follow-up.
Longitudinal outcome patterns were studied for the 10 cases at the 2-year, and for 8 of them
likewise at the 5-year, follow-up. The outcome was classified into three categories: (1)
Longitudinal Emotional and Functional Improvement, including both intrapsychic, and outer,
adaptive changes; (2) Functional Improvement without Inner Changes; and (3) Persistent
Emotional and Functional Difficulties through the whole follow-up period. The condition for
assessment of functional improvement was less severe DSM III-R diagnosis, reduction in
consumption of either psychiatric care or psychopharmacological agents, or both, increased
GAS scores and a level above 60 (GAS scores up to 60 are usually treated as clinical disease,
and above 70 as normal functioning), and/or increased Strauss-Carpenter scores with a level
above 10 (scores up to 10 are treated here as clinical disturbance), at least at the 2-year
follow-up. The clinical classification of emotional changes was based on consenus summaries
of follow-up interviews in such areas as dynamic hypotheses concerning repetition, enactment
and change in the inner world, and persistence versus modification of imagos of primary
objects. For every case significant factors after discharge were reported.
Intratherapeutic separations and longitudinal outcome
Two patterns of reactions to discontinuities in the therapeutic relationship were identified
(Appendix 1). Five cases were assigned to pattern (A), New Solutions of Difficulties with
Separation, Dependence and Hostility. The other five patients were classified as pattern (B),
Retraumatization in Connection with Separation. The categorizing of the longitudinal
outcome (Appendix 2) resulted in five cases being assessed as (1) Longitudinal Emotional
and Functional Improvement, two cases as (2) Functional Improvement without Inner
Changes; and three cases as (3) Persistent Emotional and Functional Difficulties. The
interplay between the working through of intratherapeutic separations and the longitudinal
outcome is presented in Table 1.
(1) Emotional and
Alice; Cecilia; Frida; Henry
(3) Persistent Emotional
(A) New Solutions
Table 1. Working through of intratherapeutic separations, dependence and hostility, and
longitudinal outcome patterns (courses inconsistent with the hypothetical pattern are
Emil Diana; Gustav; Jenny
The patients Alice, Cecilia, Frida, and Henry (pattern A1), as well as Diana, Gustav, and
Jenny (pattern B3) are grouped in a way that is consistent with the hypothesis about the
significance of therapeutic work with separation for the longitudinal outcome. For these
patients the working through of separation difficulties and the testing of new solutions in the
therapeutic environment were accompanied by positive development, while retraumatization
in connection with intratherapeutic separations was connected with persistent difficulties after
the treatment period. However, factors others than work with separation could be decisive for
these cases. The remaining three patients, Bruno (B1), Klas (A2), and Emil (B2) are more or
less inconsistent with the expected pattern, and are treated here as ”critical cases”. In the
following discussion, each cell in Table 1 is examined and compared with regard to admission
diagnosis, initial patterns, middle phase, reactions to termination, and significant factors in the
Hypothetical positive sequence (A1): Four patients showed signs of New Solutions of
Difficulties with Separation, Dependence, and Hostility, followed by longitudinal emotional
and functional improvement. The only exception was Alice’s increased consumption of
psychiatric care and reduced Strauss-Carpenter score during the 5th year of follow-up. Initial
diagnoses in these patients were, respectively, schizophrenia in remission; depressive disorder
and borderline personality; schizoaffective disorder in remission and borderline personality;
obsessive-compulsive disorder and borderline personality. In the early phase of the
therapeutic relationship these patients showed a wide spectrum of reactions to separation:
rejection of the therapeutic relationship and denial of the patient role, followed by attempts at
maintaining a pretended high level of functioning; denial of unmanageable strong reactions to
separation, followed by seclusion, shielding and autistic encapsulation (Tustin, 1990); strong
and chaotic reactions to separation, followed by rapid regression; attempt at self-sufficiency
and manifest reactions to separation with paranoia, catatonia and autistic encapsulation.
In the middle phase varied examples of new solutions could be found: acceptance of
dependence and recognition of anger in the therapeutic relationships, limited to intellectual
handling of strong feelings; capacity to invest in the relationship, to discriminate affects and
tolerate anger in the presence of therapist; some tolerance for conflictual dependence needs
and depressive feelings; diminished dread of outrageous feelings and the taking on of
responsibility for one’s own symptoms. However, limitations in the therapeutic working
through of, above all, negative transference were found in all these cases. Object loss in
connection with changes of therapist during the treatment period was not worked through in
either of the two cases where this arose. Discharge from the therapeutic community was a
hard experience for two patients; another was able to avoid the loss by establishing a private
relationship with her therapist; for yet another patient it was a relief to dispense with the
emotionally provocative climate of the therapeutic community. Significant factors after
discharge were psychotherapy outside the therapeutic community during, and years after the
stay, respectively, a real relationship with a former therapist, and one patient’s conscious
effort to manage his life, test new solutions, and sort out old unresolved interpersonal
Hypothetical negative sequence (B3): For three patients severe limitations in the work done
with separation, dependence, and hostility, and repeated retraumatization in connection with
separation was followed by persistent emotional and functional difficulties for a long time
after the treatment. Two patients had lower scores on the Strauss-Carpenter Outcome Scale at
the 2-year follow-up, and their consumption of psychiatric care and of
psychopharmacological agents increased during the follow-up period compared to the 2 years
before the stay at the therapeutic community. Even though Diana showed some improvement
at the 2-year follow-up her emotional and functional difficulties became more and more
severe. These patients were the only ones with unchanged psychiatric diagnoses, and their
initial diagnoses were: psychotic disorder and borderline personality; undifferentiated
schizophrenia; generalized anxiety disorder, phobia and dependent personality. Early
reactions to separations varied from strong and chaotic reactions, followed by recent and rapid
regression, or attempts at self-sufficiency and inability to express regret, in spite of visible
reactions, to absence of affective reactions, even if separation was experienced as the death of
the object and a threat to survival.
Multiple examples of retraumatization in connection with separation and object loss could be
found in all cases in the middle phase. In one case the patient’s deep regression was
encouraged by the therapist, although the actualized sexual and aggressive feelings were
neglected by both parties, and a persistent pattern of sado-masochistic relating and strong
mutual ambivalence was established. One patient showed a significant functional
improvement, encouraged by the therapist, who proffered himself as ”better” parental figure,
but neglected the thereby actualized aspects of the relationship to the primary objects.
Another patient never relinquished her wish for ”a new, better mammy” and her clinging to
exchangeable but indispensable objects. The two changes of therapist were never worked
through, and she reacted with her first psychotic breakdowns.
Termination was a catastrophic experience for all these patients. In two cases the shared
fantasy was that separation was impossible, and discharge was precipitated by the therapists
leaving the therapeutic community. These patients reacted with repeated breakdowns, self-
destructivity and durable psychoses. The third patient tried to satisfy her outspoken object
hunger through lingering manic and paranoid delusions. All three patients actively recreated
separation traumas during the follow-up period. One patient twice initiated new
psychotherapeutic contacts, but was rejected after short periods of time by the therapists.
Another ”flew from life” to addiction, well aware that the price for him would be new
psychoses. The third patient disrupted therapy for a long period of time, and acted out her
delusion that she could find new real parental figures.
Comparison between positive and negative courses: Neither diagnosis nor Strauss-Carpenter
Scale ratings at admission differentiated between the positive and negative courses: initial
heavy psychiatric diagnoses and low Strauss-Carpenter scores (up to 10) seem to be equally
distributed across the two patterns. Low initial GAS scores (up to 60) were noted in all cases
of Retraumatization and Persistent Emotional and Functional Difficulties (B3), compared to
one case (out of four) of New Solutions and Emotional and Functional Improvement (A1).
However, GAS scores on a one-month basis during the treatment period showed considerable
fluctuation in all patients. No systematic differences in reactions to separation could be found
between the two contrasting patterns in the initial phase of treatment. Early strong and chaotic
reactions and rapid regression were, for example, typical for Frida (A1) and Diana (B3), and
denial of affects and attempt at self-sufficiency for both Alice, Cecilia, Henry (A1), and
Differences between these two patterns became apparent in the middle phase. The enactment,
in transference and in action, of conflicts and relations to primary objects led to new solutions
in some areas for patients following the hypothetical positive sequence, while the negative
courses were characterized by traumatizing repetition of pathogenic patterns. Termination was
traumatic for half of the patients in the first group and for all in the second. Some pleasure in
separation, the shouldering of own responsibility, and independence could be found in three
cases of positive course, but in none of the cases of negative course. Increasing dissociation
from the treatment period as well as contempt and devaluation of the therapist could be
observed in two patients representing the positive, and one the negative course. Good
longitudinal outcome was associated with the patients’ active attempts at ”cure” after
discharge, such as psychotherapy, real relationship with the former therapist, or determination
to survive and change ones own life. In all cases of negative course the patients actively
recreated separation traumas during the follow-up period, and none of them were in regular
psychotherapy outside the therapeutic community. On the whole, some working through of
intratherapeutic separations in the middle and terminal phase, actively promoted by the
therapist, as well as the patient’s own attempts at new solutions of separation difficulties years
after termination, seems to be decisive for the longitudinal inner and outer development. On
the other hand, retraumatization in connection with intratherapeutic separations and severe
limitations in working through of difficulties with separation, dependence and hostility,
together with the patient’s active recreation of separation traumas in the posttreatment phase,
seems to be decisive for persistent emotional and functional difficulties.
Critical cases (A2, B1, B2): Three cases were inconsistent with the hypothesis: one of new
solutions of separation difficulties not accompanied by longitudinal emotional improvement,
and two of retraumatization in connection with intratherapeutic separations, although at least
functional improvement in the posttreatment phase was exhibited.
Klas (A2) is a case of new solutions, although with functional improvement without inner
changes. His initial diagnosis was dysthymic disorder and avoidant personality. Early in the
treatment Klas was passive, shy, withdrawn, and hesitant about becoming attached to anyone.
In the middle phase of treatment there emerged a clear pattern of transference and
countertransference: Klas’ defensiveness and absence of emotional reactions awoke a desire
in the therapist to ”force his way through”, and Klas responded with displaced peevishness.
The maintenance of this concealed pattern appeared to be gratifying to both parties. Klas’
spectacular progress had been a result of feats accomplished by will-power and conscious,
cognitive strategies of using ”counter-thoughts” in order to do what he most feared. The warm
relationship with his active and practically oriented therapist helped Klas to accept his
dependency need, to get rid of his paralyzing passivity, and to tolerate his anger. However, his
inner conflicts were not solved, but rather handled in a opposite, counterphobic way. An
obstacle to working through, at discharge, was the unspoken agreement between patient and
therapist not to address all the libidinal and aggressive aspects of the therapeutic relationship,
and Klas reacted with time-limited self-sufficiency and the feeling of being forsaken. In the
posttreatment phase Klas continued his struggle with passivity by means of counterphobic
solutions. The therapeutic contact continued 3 more years, and Klas became able to feel
sorrow and to mourn his father, whose death has been the triggering factor in Klas’ collapse
some years before treatment.
Bruno (B1) displayed longitudinal emotional and functional improvement, despite limitations
in the work done with separation, dependence, hostility, and unsolved transference. His
diagnosis at admission was dysthymic disorder, social phobia, avoidant and paranoid
personality. Bruno’s initial denial of unmanageable strong reactions to separation, his
seclusion and shielding was broken by his first female therapist’s investment in ”symbiotic
attachment”. Bruno enacted with her his real or fantasized Oedipal triumph, and he elaborated
rigid control rituals at separation and reunion, followed by increased physical violence
towards nonliving objects. The therapeutic relationship was considered by others to be
increasingly destructive, and Bruno was allotted a new male therapist. This measure was
never worked through. The ”therapeutic acting out”, together with both therapists’ tendency
to infantilize the patient’s more adult expressions of sexuality and aggression, ensnared Bruno
in tenaciously levelling accusations of ”maltreatment”, which made it possible for him to
cling to the treatment situation, and at the same time to turn away in anger and
disappointment. The termination became an enduring power struggle. Two factors appear to
have contributed to the apparent inner and outer improvement, which had taken place first
after discharge: Bruno ”dumped” his guilt and negative self-image at the therapeutic
community which he thereafter needed to avoid, and he worked through some of his core
problems in regular psychotherapy after discharge. At follow-ups he was of the opinion that
he had needed half of the time he had spent in psychotherapy to deal with the previous
destructive treatment experience.
In the case of Emil (B2), serious limitations in the work done with separation, dependence,
and hostility, as well as severe traumatization at termination were followed by functional
improvement. His initial diagnosis was schizophreniform disorder and unspecified personality
disorder. Emil’s denial of a dependence relationship with the therapist, and his increasingly
aggressive attempts to be a ”helper to the helpers” were substituted after 1 year by psychotic
reactions to separation and clinging to the therapist and his body. Emil denied the therapist’s
role as a professional, at the same time as the therapist proffered himself as ”better” parental
figure. In an attempt to be ”only good” Emil eradicated every direct expression of anger. His
therapist, who strongly identified himself with Emil’s picture of primary objects, fantasized
about Emil’s violence and berserk fury, but neglected aggression in their relationship and
noticed only Emil’s longing. Emil reacted to termination with delusional canceling of what
occurred, denial that he ever had been in treatment, and the psychotic conviction that he and
the therapist were friends for life, at last being able to live together. This delusional realization
was followed by a psychotic breakdown. He had previously reacted in the same way to his
mother’s death. In the ascetic religious group Emil returned to, some time after discharge, he
could find social and spiritual community, vindication of his withdrawal from the hostile
world, and release from sexuality and aggression. Up to the 2-year follow-up he was free from
psychosis and showed a considerable functional improvement, not paralleled by inner
changes. In his opinion the treatment period gave him nothing and he needed two years of his
life to put himself together after his stay there.
Comparison between cases consistent and inconsistent with the hypothetical sequence:
Despite new solutions, in one case the longitudinal outcome was limited to functional
improvement without inner changes. Despite retraumatization, positive longitudinal outcome
was found in two cases, one representing both emotional and functional improvement, and
one functional improvement without inner changes. The three ”critical cases” had the highest
scores on the Strauss-Carpenter Outcome Scale at follow-ups and showed marked
improvement in all functional variables. A conscious determination, nourished by the
therapist, to live a better life and to force outer changes was of crucial importance for the
positive development after treatment in the case of Klas, although the unsolved pattern of
transference and countertransference hindered his emotional development. Despite the
catastrophe at termination, psychotherapy and religious salvation, respectively, contributed to
Bruno’s and Emil’s increased level of functioning. Neither diagnosis, Strauss-Carpenter Scale
ratings at admission, nor the therapists’ contributions seem to differentiate between cases
consistent and inconsistent with the hypothesis. Countertransferential difficulties and
unsolved transference were found in all cases of good outcome. In all cases of
retraumatization the therapist’s contribution was evident. Examples of a lasting therapeutic
contact after discharge, as well as of radical dissociation from the treatment period and
enduring resentment could be found both in cases consistent with the hypothesis and in the
”critical cases”. The difference between these two courses seems to be determined by factors
outside the boundaries of the therapeutic community, especially the patients’ different ways of
dealing with termination trauma.
Termination process and obstacles to the working through of intratherapeutic separations
The termination process started with the beginning of the treatment and the obvious question:
”Why should I attach to the therapist if I have to leave him/her and the treatment setting?”
The first separation confronting the patient was admission, which entailed a loss of the
previous environment and adaptation, and confrontation with a strange and often threatening
therapeutic culture with high demands of relating and emotional involvement. The new
emotional attachment and the budding of dependence awakened conflicting feelings and
strong anxiety. The patients relived new editions of previous strains and either reacted with
hostility or denied this affect. Early actualization of these problems was decisive for
interruption of treatment in half of the dropout group (Tullhage & Werbart, 1995). For
patients who remained in treatment, transition to the middle phase was, in half of the cases,
marked by acute psychotic crises or self-destructivity, and, in others, by severe regression to
an infantile dependence and clinging, sometimes nourished by the therapist. Repeated critical
episodes provided the possibility of working through and testing new solutions, but also the
risk of new traumas and the petrifaction of old patterns.
In 7 of the 10 cases, termination showed itself to be an acute crisis which patients could not
cope with on their own. These patients experienced discharge as a loss of part of their own or
the mother’s body, or as a threat to their own ego-boundaries. They reacted with panic,
anxiety, intolerable rage, psychotic disorganization, self-destructivity, and protracted inner
and outer difficulties. The pattern of patients reversing the situation of separation could be
observed in four patients, all of whom left the therapeutic community some weeks before the
agreed discharge date. Only three of the patients showed signs of some pleasure in separation,
the taking on of own responsibility, and independence. The termination phase was a hard
experience also for the therapists, evoking fantasies of never getting rid of the patient.
Exaggerated expectations of treatment results, and idealization of one’s own therapeutic
culture can lead to staff dissatisfaction and unconscious, hostile feelings towards patients who
do not become as well as expected. This can enmesh the patient and staff in destructive
relationship patterns which quickly take over at termination. Some of the therapists acted out
their countertransference and their own separation difficulties by prolonging the treatment, by
abrupt termination or by concluding their employment.
Shortcomings in the work done with separation, dependence and hostility left apparent traces
in the posttreatment phase. Solutions applicable in the therapeutic environment could not
automatically be employed in new conditions. For three patients the process in the
posttreatment phase was determined by the loss of a real and indispensable supportive object,
leading to chronic deterioration in emotional and social functioning. Previous mutual
idealization of the therapeutic relationship could turn into mutual contempt and devaluation.
Dissociating themselves from the treatment period, and displacing their own guilt feelings and
own bad or weak parts to the treatment setting, could bring relief to certain patients, both
those doing well and those with increasing difficulties after termination. During the follow-up
period half of the patients regarded their stay at the therapeutic community as a negative and
Neither the patients nor the therapists were prepared for the termination. A recurring pattern
was that the patient displayed extreme anger or was incapable of managing it in connection
with termination. He or she felt themselves to be misunderstood or not seen by the therapist,
whilst the therapist idealized the treatment, patient, and therapeutic relationship. Consensus
formulations of the Core Problematic Area in the Therapeutic Relationship showed that all
therapists had marked difficulties in seeing and accepting their own and the patients’ negative
feelings. Common to the therapists was a tendency to overestimate the working alliance and
idealize the results of collaboration. Irrespective of the therapists’ contribution to their
patients’ progress, the therapists were surprisingly blind to the patient’s difficulties with
separation, dependence, and, above all, hostility, and to expressions of these difficulties in
transference through the whole treatment period. In all cases the greatest limitation was
unsolved patterns of transference and countertransference, sometimes leading to acting out.
Both parties’ avoidance of painful and negative feelings made the inevitable mourning
The working through of separation difficulties and the testing of new solutions during the
treatment period were associated with longitudinal emotional and functional improvement in
four of the ten cases studied. In three other cases retraumatization in connection with
discontinuities in the therapeutic encounter was followed by persistent emotional and
functional difficulties. In all, seven cases were consistent, and three inconsistent with the
hypothesis about the significance of therapeutic work with separation for longitudinal
outcome. The difference between a positive and negative course was constituted by the
therapist’s contribution to working through or to retraumatization in the middle and terminal
phase and by the patient’s contribution in the posttreatment phase. In all cases of positive
course, the patients actively attempted to use new solutions for separation difficulties after
termination, and in all cases of negative course the patients recreated separation traumas.
Positive development after discharge was reinforced by continuing in regular psychotherapy,
forming real relationship or seeking religious salvation. The difference between courses
consistent and inconsistent with the hypothesis was determined by factors outside the
treatment setting after discharge, and especially the patient s’ different ways of dealing with
the trauma of termination during the posttreatment phase.
None of the courses studied were fully compatible with the ideal sequence of actualization
and enactment of problems with separation, dependence, and hostility, accompanied by
survival and working through of such experiences in the therapeutic relationship, followed by
mourning in the terminal and posttreatment phase. The problems of separation, both current
and past, as well as hostile feelings, hate, destructivity, and the enactment of these feelings in
the therapeutic relationship and on the interpersonal stage of the therapeutic community were
neglected by the therapists. This often led to a breakdown of the therapeutic relationship,
traumatic and unnecessarily painful terminations, and to recurring problems for the patients in
the posttreatment phase of reorientation and adjustment. If no work is done with anger,
disappointment, and destructivity, then the patient is also unable to experience the joy in
separating, being more mature, gaining freedom, and becoming more independent.
Termination is then merely a loss. Consequently, the patient either clings tight, turns away in
rage and disappointment, denies the separation altogether, or relapses into his or her earlier,
These problems do not have to do with termination; they primarily concern the confusion as
to what should happen in the middle phase, that is, after attachment has been established. At
the therapeutic community attention was narrowly directed towards the patient’s attaching,
opening up, and relaxing his or her guard and defenses, only to be then separated from the
therapist and the community. According to the therapists’ view, this process was one of
”thawing out” the patient, forging an attachment, and then separating. A recurrent experience
reported by the patients was that of being seduced into a warm, close, and trusting relationship
only to be then betrayed, abandoned, and left alone out in the cold. The study of the first 10
patients in the dropout group (Tullhage & Werbart, 1995) demonstrated that the therapists’
orientation towards emotional attachment resulted for all patients in strong, unmanageable
and contradictory affects concerning dependence, aggression and hate. At the same time, the
therapists could not cope with the awakened feelings and reactions. This pattern was common
both for patients who completed and those who interrupted their treatment.
Alternative explanations to the results obtained in this study could focus on the nature of the
patient’s psychiatric problems or pretreatment status and functioning. McGlashan
(McGlashan, 1984; 1986a; 1986b; McGlashan & Heinssen, 1988) showed that the variance in
longitudinal outcome was related to diagnostic categories of schizophrenia, schizoaffective
disorder, borderline personality disorder, and unipolar affective disorder. Strauss and Carpenter
(1977) demonstrated the predictive power of variables such as employment, social contacts,
psychotic symptoms, and duration of hospitalization. In the present study neither initial
diagnosis nor Strauss-Carpenter Scale ratings at admission differentiated between the two
patterns of dealing with intratherapeutic separations, or between cases of good and poor
outcome. A further investigation could include a close examination of the patient’s
characteristic reactions to previous separations and object loss before the debut of manifest
psychiatric problems. The results of this study were obtained by micro-analysis of a limited
number of cases in a specific setting and need to be tested on new cases in different settings.
The study showed that the working through of separation, dependence, and hostility during
and after the treatment is decisive not only for the vicissitudes of the therapeutic process but
also for the long-term course. One obstacle to this work was ideas of a ”new-object
relationship” (Bibring, 1937), ”a new beginning” (Balint, 1952), and a ”corrective emotional
experience” (Alexander & French, 1946; Alexander, 1961), fostered in this all-inclusive
therapeutic environment. The therapists’ shared ”private theory of pathogenesis and cure”
(Abend, 1979; Arlow, 1981; 1986), based on the model of compensation for symbiotic deficit
by means of emotional attachment to new and better ”parents”, along with separation-
individuation, appears not to reflect the real course of events. When the metaphor of
regression to an infantile stage, growth, and maturation is taken literally the result is a total
”maternal environment”. The emphasis on attachment and dependency actualizes in the
patient archaic relationship patterns and split-off images of the parental figures, and arouses
hard-to-handle, negative, hostile feelings that are omitted from the therapeutic working
Denial, split, and projection of hostile feelings, which are concomitants of emotional
dependence and of separations, are the main pathogenic defenses in patients with psychosis
and/or severe personality disorders (see Kernberg, 1991; 1992; 1993) – and the main
countertransferential obstacle for therapists (Winnicott, 1947/1975; Boyer, 1994; Feinsilver,
1994). These difficulties do not appear to be unique to the studied therapists and the treatment
modality. However, the therapeutic task has been hindered by the fusion of the role of
psychotherapist and milieu therapist, making both roles difficult for the staff to handle,
especially over long treatment periods. The everyday presence of the therapists at the
therapeutic community contributed to their avoidance of the patients’ central problems with
separation, dependency, and hostile feelings. A prerequisite for systematic work with negative
transference is clear boundaries marking psychotherapy as something apart from the everyday
life of the treatment institution. When both parties are involved in the same ”drama”, and also
share everyday life, then it can be difficult to discern the different roles they adopt – and in
which play. There is a risk that the therapeutic relationship loses its character of ”as-if”,
”make-believe” or mutual play. When the enactment of the archaic relations to primary
objects is stimulated but not attended to, the therapeutic relationship runs the risk of becoming
a new, traumatic experience in the patient’s struggle for an independent existence. The
greatest obstacle is a combination of the patient’s emotional needs being awakened and
monopolized by the treatment institution whilst at the same time the termination of the stay,
or of the therapeutic contact, is definitive (Werbart, 1995).
We have to relinquish the idea of the ”total” therapeutic environment, a place and time for the
decisive and complete cure. Even a positive outcome at termination does not guarantee the
durability of the effects of treatment (Werbart, 1993). Long stays in a therapeutic community
characterized by intensive relating within the sheltered area, a fusion of individual and milieu
therapy, and a lack of ways of working outside culminate in dramatic terminations and
difficulties in managing the period after the stay. The often forgotten phase subsequent to
completion of formal treatment is of central importance: The patient’s new experiences and
solutions must be transferred to a life-context outside the therapeutic culture. The therapeutic
community can merely initiate something, never complete or finish it. If the stay at the
therapeutic community is not to engender a new form of hospitalization, continued care
dependence or a repetition of the patient’s earlier failures, we have to learn to see the
treatment as inevitably fragmentary.
Some theoretical aspects
Why is the working through of difficulties with separation, dependence, and hostility of such
central importance to the longitudinal outcome? The nature of intrapsychic change is often
conceptualized in terms of mourning and internalization, both connected with separation and
object loss. Garcia Badaracco (1986:142) describes the first aspect of change as de-
identification from ”maddening” internal objects, Modell (1991:233) – as ”unfreezing” of
conflict situations, and Laplanche (1992:173) as a ”new version or translation [...] preceded
by the painful work of Lösung, of detachment from the former version”. The second aspect of
change – internalization – presupposes both gratification and frustration, or, in terms of Blatt
and Behrends (1987), both repetition and disruption of the mother-infant unity. Quinodoz
(1991:131) points out two main difficulties on the way to psychic change. The one has to do
with the analyst, and the other with the patient. An inevitable part of the ”taming of solitude”
is the patient’s hate toward the analyst, who is perceived as responsible for the pain the patient
is experiencing. The main obstacle to this work is the therapist’s countertransference
resistance to accepting the hate and destructiveness of the patient, as well as the therapist’s
In the studied therapeutic community the ”widening” of the therapeutic framework was
connected with ideas about compensating the patient for a supposed deficit in ”good-enough
mothering”. An unconscious correlate to this ”private theory of cure” is the therapist’s wish to
provide the patient with that which he or she had never received, resulting in the therapist’s
hate of the patient who hinders this wish fulfillment. This is paralleled by the patient’s wish of
being one with the object and never separating, a common mechanism behind the ”negative
therapeutic reaction”. With an incisive formulation one may postulate that a precondition of
therapeutic success is that the patient’s wish never is satisfied, and that the patient realizes and
accepts this. Psychic change presupposes access to both love and hate, and the ability to
sustain psychic pain. The shared avoidance of hostile feelings and painful experiences by both
the patient and the therapist results in pathological forms of mourning, and in mutual
idealization, followed by devaluation and contempt after termination. Hostile feelings, left
outside the therapeutic work, corrupt the therapeutic relationship. When the dependence
relationship is characterized by a ”deficit” in positive, satisfying and relatively conflict-free
involvement although with a ”surplus” of ”experienced incompatibility” and aggressive
charge, the result is not maturational process or therapeutic change, but pathogenic
identification. Such a ”negative” of the mechanism described by Blatt and Behrends (1987) is
active in infantile trauma, long-term sequelae of extreme man-made trauma (Werbart &
Lindbom-Jakobson, 1993), and in the potentially new, traumatic consequences of
transference, noted by Thomä and Kächele (1988:9). The present study demonstrates that the
analogy between the therapeutic cure and the work of mourning has not only a metaphoric but
also a literal, and specific meaning. It may be hypothesized that the actualization and
enactment of relations to primary objects in the therapeutic setting is a precondition of
working through, although the failure in repetition of old patterns, followed by the work of
mourning of lost internal objects, is necessary for therapeutic success.
The hypothesis about the significance of therapeutic work with separation for longitudinal
outcome was confirmed in 7 of the 10 studied cases. The working through of these difficulties
and new solutions of crises in the middle phase, while not assuring a trauma-free termination,
had consequences for the course after discharge. Unsolved difficulties with separation,
dependence, and hostility in the middle phase resulted in traumatic and extremely painful
terminations. Besides the therapist’s contribution in the middle phase, the patients’ different
ways of dealing with the trauma of termination were decisive for the variance in the
longitudinal outcome, as demonstrated by 3 ”critical cases”, inconsistent with the hypothesis.
The often postulated linear pattern of symbiotic block followed by resolution and separation-
individuation was not confirmed in this study. The greatest obstacle to the therapeutic task
was a combination of the patient’s emotional needs being awakened and monopolized by the
treatment setting while, at the same time, the problems of separation, dependence, and
hostility were neglected in the therapeutic work. Inability to mourn was connected with less
emotional and functional improvement. Limitations in the work of mourning of lost internal
objects during the treatment period had severe longitudinal consequences after termination.
So the little prince tamed the fox. And when the hour of his departure
drew near –
”Ah,” said the fox, ”I shall cry.”
”It’s your own fault,” said the little prince. ”I never wished you any sort
of harm; but you wanted me to tame you ...”
”Yes that is so,” said the fox.
”But now you are going to cry!” said the little prince.
”Yes that is so,” said the fox.
”Then it has done you no good at all!”
”It has done me good,” said the fox ”because of the colour of the wheat
”Good-bye,” said the fox. ”And now here is my secret, a very simple
secret: It is only with the heart that one can see rightly; what is essential is
invisible to the eye [...] Men have forgotten this truth [...] But you must not
forget it. You become responsible, forever, for what you have tamed.”
Antoine de Saint-Exupéry (1946). The Little Prince, pp. 66ff.
This study was supported by grants from the Swedish Council for Social Research, Stockholm
County Council, and the Federation of the County Councils. The author gratefully
acknowledges the support and useful comments given by Professor Johan Cullberg and
Professor Carl-Otto Jonsson, as well as the work done with consensus summaries by Ann-
Sofie Bárány and Björn Sahlberg.
Licensed psychologist Werbart is a psychoanalyst in private practice and a researcher at the
Psychosocial Research Centre, Stockholm County Council, and at the Stockholm University,
Department of Pedagogy.
APPENDIX 1. DSM III-R diagnoses at admission and patterns of changes in dealing with
separation, dependence, and hostility during the treatment period
Pattern A: New Solutions of Difficulties with Separation, Dependence, and Hostility
Alice (schizophrenia in remission)
Early*: Rejection of the therapeutic relationship; denial of the patient role; attempts at maintaining a pretended
high level of functioning
Middle phase*: From psychotic breakdowns to acceptance of dependence; recognition of anger and intellectual
handling of strong feelings
Termination: Pleasure in separation; relapse to precocious self-sufficiency
Posttreatment: Mutual idealization turns into contempt and devaluation; relapse to psychosis
Cecilia (depressive disorder; borderline personality)
Early: Denial of unmanageable strong reactions to separation; seclusion, shielding, and autistic encapsulation
Middle phase: Inability to handle separations; capacity in the presence of a real object to discriminate affects and
recognize and tolerate hostile feelings without injuring herself or attempting suicide
Termination: Realization of wish for real object by mutual acting out
Posttreatment: Real relationship with the former therapist
Frida (schizoaffective disorder in remission; borderline personality)
Early: Early strong and chaotic reactions to separation; rapid regression
Middle phase: From reversible psychotic dilution and acts of self-damage to capacity to endure conflictual
dependence need and depressive feelings of grief, sorrow and regret
Termination: Some pleasure in separation, followed by severe regression, panic, decomposition, and persistent
Posttreatment: Increasing dissociation from the treatment period and steady progress in psychotherapy
Henry (obsessive-compulsive disorder; borderline personality)
Early: Attempt at self-sufficiency and manifest reactions to separation with paranoia, catatonia and autistic
Middle phase*: Active denial, reversing of the situation, and some grandiosity
Termination: Pleasure in separation; taking on responsibility and independence; idealization and denial of
Posttreatment: Internalization of the therapist’s functions; gradually more realistic picture of the therapist;
delusions transformed into painful inner conflicts
Klas (dysthymic disorder; avoidant personality)
Early: Absence of affects and manifest reactions to separation; nobody being permitted to become significant;
Middle phase: Manageable reactions to separation and acknowledgment of the therapist’s significance
Termination: Time-limited self-sufficiency and feeling of being forsaken
Posttreatment: Ongoing mutual idealization and an unworked-through pattern of transference and
countertransference; continuous outer adjustment
Pattern B: Retraumatization in Connection with Separation
Bruno (dysthymic disorder, social phobia; avoidant and paranoid personality)
Early: Denial of unmanageable strong reactions to separation; seclusion and shielding
Middle phase*: Rigid control over the therapist, unmanageable deadlock and new trauma
Termination: Clinging to the treatment situation and turning away in anger and disappointment
Posttreatment: Projection of guilt and of bad and weak parts of the self onto the treatment setting, and
dissociation from the treatment period
Diana (psychotic disorder; borderline personality)
Early: Strong and chaotic reactions to separation; diluted and borderline psychotic states
Middle phase: Severe regression to an infantile dependence and clinging; separation as loss of libidinal object
and of parts of the self; unmanageable deadlock
Termination: Repeated breakdowns, annihilation anxiety and severe regression; protracted difficulties
Posttreatment: Mutual admiration turns into anger and hate; new trauma; stuck in an archaic universe of
relations to primary objects
Emil (schizophreniform disorder; personality disorder)
Early: Rejection of the therapeutic relationship; denial of the patient role; attempts at maintaining a pretended
high level of functioning
Middle phase: Relapse to psychotic solutions and severe regression to infantile dependence and clinging; denial
of the therapist’s importance as a therapist
Termination: Total denial of separation; psychotic acting out and delusional realization of the wish for real
Posttreatment: Projection of guilt and of bad and weak parts of the self onto the treatment setting and
dissociation from the treatment period
Gustav (schizophrenia, undifferentiated)
Early: Attempt at self-sufficiency and inability to express regret; manifest reactions to separation
Middle phase: From relapse to psychosis to contact with own suppressed anger
Termination: Incapability of reacting before the discharge and self-inflicted breakdown at termination
Posttreatment: Retraumatization, ”flight from life”, addiction and durable psychoses; stuck in an archaic
universe of relations to primary objects
Jenny (generalized anxiety disorder, phobia; dependent personality)
Early*: Absence of affective reactions; separation as death and loss of an object vital for survival
Middle phase*: From psychotic solutions to some contact with suppressed anger; inability to deal with
separation; clinging to the therapist
Termination: Clinging to the treatment setting, followed by manic defenses, psychotic break down and self-
Posttreatment: Retraumatization; stuck in an archaic universe of relations to primary objects
* = change of therapist
APPENDIX 2. Longitudinal outcome patterns: emotional and functional development at the 2- and 5-year follow-up
Patient Significant factors
after discharge at admission and at
DSM III-R diagnosis Consumption of
2 years before/ 2
years after/ 5th year
cal agents 2 years
before/ 2 years after/
5th year after (mg
Scale (GAS) at
admission and at
(1 – 100)
at admission and at
(0 – 16)
P a t te r n 1 : L ong i tu d in al Em o ti ona l an d Fu nc t io na l I mp ro v e me n t (i n t ra p sy ch i c and a dap t iv e c han ge s )
psychotherapy less severe 46; 33; 101
psychotherapy change to no diagnosis 14; 0; 0
real relationship change to no diagnosis 129; 28; 0
psychotherapy less severe 130; 16; 0
conscious effort by
P a t te r n 2 : Functional Improvement without Inner Changes
religious salvation less severe 201; 40; ?
conscious effort by
P a t te r n 3 : P e r s i s te nt E mo t ion a l a nd F unc t i ona l D i f f ic u lt i e s
75; 48; 33
0; 0; 0
83; 0; 0
172; 73; 5
159; 0; 0
65; 80; 70
45; 70; 75
50; 60; 70
62; 70; 80
65; 65; 70
9; 12; 8
8; 15; 15
10; 11; 13
7; 13; 14
9; 12; 15 less severe 110; 0; 0
256; 9; ?
0; 0; 0
52; 70; ?
50; 75; 80
7; 14; ?
9; 15; 15 change to no diagnosis 40; 0; 0
unchanged 165; 10; 73 128; 115; 72 35; 60; 45 6; 12; 7
unchanged 61; 114; 21 102; 339; 330 35; 40; 60 8; 7; 9
unchanged 45; 134; ? 33; 54; ? 58; 50; ? 8; 5; ?
Abend, S. (1979). Unconscious fantasy and theories of cure. Journal of the American
Psychoanalytic Association, 27, 579-596.
Ainsworth, M.D.S., Blehar, M.C., Waters, E., & Wall, S. (1978). Patterns of attachment: A
psychological study of the strange situation. Hillsdale, NJ: Lawrence Erlbaum
Alexander, F. (1961). Unexplored areas in psychoanalytic theory and treatment. In The scope
of psychoanalysis: Papers of Franz Alexander (pp. 319-335). New York: Basic Books.
Alexander, F., & French, T.M. (1946). Psychoanalytic therapy. New York: Roland Press.
American Psychiatric Association (1987). Diagnostic and statistical manual of mental
disorders, (3rd ed. rev.). Washington, DC: APA.
Arlow, J.A. (1981). Theories of pathogenesis. Psychoanalytic Quarterly, 50, 488-514.
Arlow, J.A. (1986). The relation of theories of pathogenesis to psychoanalytic therapy. In
A.D. Richards & M.S. Willick (Eds.), Psychoanalysis: The science of mental conflict (pp.
49-63). Hillsdale, NJ: Analytic Press.
Balint, M. (1952). New beginning and the paranoid and depressive syndromes. International
Journal of Psycho-Analysis, 33, 214-224.
Becker, T.C. (Reporter). (1993). Panel: The difference between termination in psychotherapy
and psychoanalysis. Journal of the American Psychoanalytic Association, 41, 765-773.
Bibring, E. (1937). On the theory of the results of psychoanalysis. International Journal of
Psycho-Analysis, 18, 170-189.
Blanck G., & Blanck, R. (1988). Ego psychology and termination in psychoanalysis and
psychotherapy. Journal of the American Psychoanalytic Association, 36, 961-984.
Blatt, S.J., & Behrends; R.S. (1987). Internalization, separation-individuation, and the nature
of therapeutic action. International Journal of Psycho-Analysis, 68, 279-297.
Bowlby, J. (1973). Attachment and loss, vol. 2, Separation: Anxiety and anger. London: The
Boyer, L.B. (June, 1994). Analyzing through the countertransference. Paper presented at the
XIth International Symposium for the Psychotherapy of Schizophrenia, Washington, DC.
Bretherton, I. (1991). The roots and growing points of attachment theory. In C. Murray
Parkes, J. Stevenson-Hinde, & P. Marris (Eds.), Attachment across the life cycle (pp. 9-
32). London and New York: Routledge.
Dewald, P.A. (1982). The clinical importance of the termination phase. Psychoanalytic
Inquiry, 2, 441-462.
Edelson, M. (1963). The termination of intensive psychotherapy. Springfield, Ill.: Charles C
Endicott, J., Spitzer, R.L., Fleiss, J.L., & Cohen, J. (1976). The Global Assessment Scale: A
procedure for measuring overall severity of psychiatric disturbance. Archives of General
Psychiatry, 33, 766-771.
Etchegoyen, R.H. (1991). The fundamentals of psychoanalytic technique. London: Karnac
Feinsilver, D.B. (1977). Symbiotic block with psychotic patients. International Journal of
Psychoanalytic Psychotherapy, 6, 132-145.
Feinsilver, D.B. (1989). Transitional play with regressed schizophrenic patients. In M.G.
Fromm, B.L. Smith (Eds.), The facilitating environment: Clinical applications of
Winnicott’s theory (pp. 205-237). Madison, CT: International Universities Press.
Feinsilver, D. (June, 1994). Psychotherapy and the comprehensive treatment of schizophrenia:
A bio-psycho-social model. Opening address at the XIth International Symposium for the
Psychotherapy of Schizophrenia, Washington, DC.
Firestein, S.K. (1978). Termination in psychoanalysis. New York: International Universities
Friedman, L. (1969) The therapeutic alliance. International Journal of Psycho-Analysis, 50,
Fritsch, R.C., de Marneffe, C., & Goodrich, W. (1987). Chestnut Lodge Attachment Scale.
Rockville, Md. Unpublished manuscript.
Fritsch, R.C., & Goodrich, W. (1990). Adolescent inpatient attachment as treatment process.
Adolescent Psychiatry, 17, 246-263.
Fritsch, R.C., Goodrich, W., & de Marneffe, C. (1992). Concurrent validity of an Adolescent
Attachment Scale. International Journal of Adolescent Psychotherapy, 1, 298-316.
Garcia Badaracco, J.E. (1986). Identification and its vicissitudes in the psychoses: The
importance of the concept of the ‘maddening object’. International Journal of Psycho-
Analysis, 67, 133-146.
Gillman, R.D. (1982). The termination phase in psychoanalytic practice: A survey of 48
completed cases. Psychoanalytic Inquiry, 2, 463-472.
Goldberg, S., Muir, R., & Kerr, J. (Eds.), (1995). Attachment theory: Social, developmental,
and clinical perspectives. Hillsdale, NJ: Analytic Press.
Goodrich, W. (1984). Symbiosis and individuation: The integrative process for residential
treatment of adolescents. Current Issues in Psychoanalytic Practice, 1, 23-45.
Green, A. (1991). Méconnaissance de l’inconscient (science et psychanalyse) [Unawareness
of the unconscious (science and psychoanalysis)]. In R. Dorey, C. Castoriadis, E.
Enriquez, R. Thom, J. Ménéchal, W.-H. Fridman, G. Berquez, & A. Green,
L’inconscient et la science (pp. 140-220). Paris: Dunod.
Hartlaub, G., Martin, G., & Rhine, M. (1986). Recontact with the analyst following
termination: A survey of seventy-one cases. Journal of the American Psychoanalytic
Association, 34, 895-910.
Holmes, J. (1993). Attachment theory: A biological basis for psychotherapy? British Journal
of Psychiatry, 163, 430-438.
Janssen, P. (1987). Psychoanalytic therapy in the hospital setting. London & New York 1994:
Kernberg, O.F. (1991). The psychopathology of hatred. Journal of the American
Psychoanalytic Association, 39 (Suppl), 209-238.
Kernberg, O.F. (1992). Aggression in personality disorders and perversions. New Haven &
London: Yale University Press.
Kernberg, O.F. (April, 1993). Plenary discussion. European Psychoanalytical Federation
Conference, Vienna, 1993.
Kupers, T.A. (1988). Ending therapy: The meaning of termination. New York: New York
Laplanche, J. (1992). Seduction, translation, drives. London: Psychoanalytic Forum &
Institute of Contemporary Arts.
McGlashan, T.H. (1984). Long-term outcome of schizophrenia and the affective disorders.
Archives of General Psychiatry, 41, 586-601.
McGlashan, T.H. (1986a). Long-term outcome of borderline personalities. Archives of
General Psychiatry, 43, 20-30.
McGlashan, T.H. (1986b). Long-term follow-up perspectives: Schizotypal personality
disorder. Archives of General Psychiatry, 43, 329-334.
McGlashan, T.H., & Heinssen, R.K. (1988). Hospital discharge status and long-term outcome
for patients with schizophrenia. schizoaffective disorder, borderline personality disorder
and unipolar affective disorder. Archives of General Psychiatry, 45, 363-368.
Meltzer, D. (1967). The psycho-analytic process. Perthshire, Scotland: Clunie Press.
Modell, A.H. (1991). Changing psychic structure through treatment: preconditions for the
resolution of the transference. In T. Shapiro (Ed.), The concept of structure in
psychoanalysis (pp. 225-239). Madison, CT 1991: International Universities Press.
Pao, P.-N.(1979). Schizophrenic disorders: Theory and treatment from a psychodynamic
point of view. New York: International Universities Press.
Pedder, J.R. (1988). Termination reconsidered. International Journal of Psycho-Analysis, 69,
Quinodoz, J.-M. (1991). The timing of solitude: Separation anxiety in psychoanalysis. London
& New York 1993: Routledge.
Rickman, J. (1950). On the criteria for the termination of an analysis. International Journal of
Psycho-Analysis, 31, 200-201.
Rinsley, D. (1980). Treatment of the severely disturbed adolescent. New York: Jason
Saint-Exupéry, A. de (1943). The little prince. London 1945: Heinemann.
Schachter, J. (1990). Post-termination patient-analyst contact. International Journal of
Psycho-Analysis, 71, 475-486.
Scott, J.P. (1987). The emotional basis of attachment and separation. In J.L. Sacksteder, D.P.
Schwartz, & Y. Akabane (Eds.), Attachment and the therapeutic process: Essays in
honor of Otto Allen Will (pp. 43-62). Madison, Connecticut: International Universities
Searles, H.T. (1965). Collected papers on schizophrenia and related subjects. London:
Searles, H.T. (1973). Concerning therapeutic symbiosis. In Countertransference and related
subjects: Selected papers (pp. 172-191). New York: International Universities Press.
Searles, H.T. (1986a). Separation and loss in psychoanalytic therapy with borderline patients:
further remarks. In My work with borderline patients (pp. 287-326). Northvale, NJ:
Searles, H.T. (1986b). Some aspects of separation and loss in psychoanalytic therapy with
borderline patients. In My work with borderline patients (pp. 327-357). Northvale, NJ:
Stone, L. (1961). The psychoanalytic situation. Madison, Connecticut: International
Strauss, J.S., & Carpenter, W.T. (1972). The prediction of outcome in schizophrenia, I.
Characteristics of outcome. Archives of General Psychiatry, 27, 739-746.
Strauss, J.S., & Carpenter, W.T. (1977). The prediction of outcome in schizophrenia, III.
Five-years outcome and its predictors. Archives of General Psychiatry, 34, 159-163.
Thomä, H., & Kächele, H. (1988). Psychoanalytic practice: Vol. 2, Clinical studies. Berlin &
Heidelberg 1992: Springer-Verlag.
Tryon, G.S., & Kane, A.S. (1995). Client involvement, working alliance, and type of therapy
termination. Psychotherapy Research, 5, 189-198.
Tullhage, E., & Werbart, A. (1995). Drop-out in psychotherapeutic environment for severely
disturbed patients. Stockholm: Unpublished manuscript.
Tustin, F. (1990). The protective shell in children and adults. London & New York: Karnac
Waelder, R. (1970). Observation, historical reconstruction, and experiment: An
epistemological study. In Psychoanalysis: Observation, theory, application (pp. 635-
675). New York 1976: International University Press.
Werbart, A. (1989). Psychotherapy research between process and effect: The need of new
methodological approaches. Acta Psychiatrica Scandinavica, 79, 511-522.
Werbart, A. (1993). Psykoterapeutiska mekanismer vid kombinerad miljö- och
individualterapi med psykotiska och personlighetsstörda patienter [Psychotherapeutic
27 Download full-text
mechanisms in combined milieu and individual therapy with psychotic and
personality-disturbed patients]. (Final progress report). Swedish Council for Social
Research, Psychosocial Research Unit, Report 1993:1.
Werbart, A. (1995). The future of psychotherapeutic environments – learning from experience
at a Swedish therapeutic community. Psychoanalytic Psychotherapy, 9, 269-288.
Werbart, A. (1996). The vicissitudes of therapeutic collaboration in long-term treatment of
severely disurbed patients. (Unpublished).
Werbart, A., & Lindbom-Jakobson, M. (1993). The ”living dead” – survivors of torture and
psychosis. Psychoanalytic Psychotherapy, 7, 163-179.
Winnicott D.W. (1975). Hate in the countertransference. In Through paediatrics to
psychoanalysis (pp. 194-203). London: The Hogarth Press. (Original work published in