Paolo Migone

The concept of "Expressed Emotion" (EE) represents one of the most important discoveries in the field of psychosocial treatment of schizophrenia over the last decades. According to a recent review (Kavanagh, 1992, p. 616), it "may prove to be the most significant treatment breakthrough in schizophrenia since the discovery of neuroleptic medication." The studies that later led to the refinement and validation of this concept, and to its utilization for preventing schizophrenic relapses, were initiated in England by Brown and his collaborators (1958, 1962, etc.) in the late fifties. At the time, the cultural atmosphere was characterized both by the newly discovered phenothiazines treatment for schizophrenia, and by the English social psychiatry movement, which, with its trend towards deinstitutionalization, was gradually challenging the traditional vision of mental illness and its treatment. The discharge of many patients from psychiatric hospitals and their management in the community brought new problems to the attention of mental health professionals and patients' families. The importance of EE studies derives from the possibility, offered by these studies, to test, through a rigorous methodology, a psychosocial approach to the treatment of schizophrenia.

It is interesting to note that this trend is somehow at odds with the current supremacy of biological models of understanding and treatment for this disorder. Although a psychoanalytic or "insight-oriented" psychotherapy of schizophrenia has been shown to be ineffective in some recent studies (Gunderson et al., 1984; McGlashan, 1984; M. Stone, 1986), EE studies have shown that a structured form of family psychotherapy does indeed influence the course of schizophrenia, thus opening the way to the refinement of these psychological interventions and to their understanding in a more general conceptual framework.

The aim of this article is to explore the psychotherapeutic techniques derived from EE studies, trying to use a psychoanalytic framework. Indeed, in these last three decades the authors who have investigated the EE concept have remained, with few exceptions, at a descriptive level, without trying to conceptualize why EE influences the course of the disorder. Most authors rely on the concept of vulnerability, namely on the stress-diathesis model of schizophrenia (Zubin & Spring, 1977). In this framework, EE is conceived as a stressor that can provoke relapse by increasing patients' arousal beyond an optimal level. However, little is known about the mechanism of relapse, or precisely why the emotions related to high EE supposedly trigger arousal and relapse of schizophrenia.

Relying on previous papers (Migone, 1988a, 1991, 1993), I shall try to apply the psychoanalytic concept of projective identification to the EE model. By doing so, I hope to contribute to a broader effort aimed to fill the gap between psychiatric and psychoanalytic research in this area. Moreover, I shall try to see whether some of the clinical experiences and conceptualizations put forward by psychoanalytically oriented psychotherapysts belonging to the last generations have something to say to psychiatric clinicians and researchers. Projective identification, in particular, has gained increasing attention as a useful conceptual tool in psychotherapy, and some authors have emphasized its usefulness also in general psychiatry (e.g., see Jureidini, 1990, for its use in the clinical interview, in psychiatric wards, and in the work with families). However, its use in the field of EE has not yet been explored.

In their initial research, Brown et al. (1958) found that discharged psychiatric patients tended to have a different outcome according to their different living arrangements (living with parents, spouse, in hostels, etc.). Patients who were living alone or with siblings were having a relapse rate of 17%, which increased to 32% for those who were living with their parents, and to 50% for those who were living with their spouses. In a second study, Brown et al. (1962) found that a fundamental variable related to a poor outcome was the level of "expressed emotion" in relatives. They divided 128 families of psychiatric patients in two groups, at "high and low emotional involvement" depending on the evaluation of the key relative on scales of hostility and expressed emotion. At the end of one year of follow up, 76% of patients from the high involvement group showed a worsening of symptoms and social behavior, with 56% of readmissions, as compared to only 28% of the patients living in families at low emotional involvement showing worsening of symptoms, and 21% being readmitted. In this study, expressed emotion was still evaluated without any standardized assessment instrument.

In summary, Brown et al. (1962) found that the most important variable related to poor outcome was the level of expressed emotions shown by family members: the higher the level of emotion and hostility in the family, the higher the likelihood for the patient to have a relapse within the next year.

In 1966, Rutter & Brown developed a semi-structured interview, called Camberwell Family Interview (CFI), to measure family characteristics and to improve study methodology. The EE index was formulated in 1972 by Brown et al., and included five scales: "Hostility," "Emotional Over-Involvement" (EOI), "Critical Remarks," "Warmth," and "Dissatisfaction." Of these, only the first four scales have shown to be correlated to outcome. For this reason, Vaughn & Leff (1976b) deleted the "Dissatisfaction" scale and replaced it with the "Positive Remarks" scale. They also modified the CFI, making it shorter and more manageable; the final version included the following five scales:

1) "Criticism": measured as frequency of critical comments, inferred both by the actual content of phrases (expressions of disapproval, resentment or rejection, and devaluation), and by specific vocal characteristics (rate, volume, and tone of speech).

2) "Hostility": this measure reflects a more global or generalized criticism and/or rejection of the patient. It is measured on a three point scale: it can be present either as a generalization of criticism, or as an overall rejection, or both.

3) "Emotional Over-Involvement": it is measured on the basis of both the past behavior referred to in the interview (exaggerated emotional response shown in the past; excessive attitude of devotion or self-sacrifice; overprotective behavior) as well as the actual behavior shown during the interview (excessive emotional involvement, emotional behaviors such as worries and sadness when relatives talk about the patient; dramatization); it is measured on a five point scale, ranging from "absent" to "marked."

4) "Warmth": measured on a five point scale, based mostly on voice tone, spontaneity, sympathy, empathy, interest in and concern for the person.

5) "Positive Remarks": expressions of praise and approval of the patient's personality, measured as frequency of positive comments about the patient.

As mentioned earlier, only the first four variables have been shown to be associated with differences in the course of schizophrenia; the first three have become part of the EE index because of their more direct correlation with the course of the disorder. However, recently "Warmth" has also been regarded as possibly associated with a better course, especially in low EE, and it has been suggested that the low relapse rate in some patients who live in high EE families might be explained by this variable. "Positive Remarks," on the contrary, appears to have no influence on patients' course.

Since "Hostility" is often associated with high "Criticism," it can be stated that "Emotional Over-Involvement" and "Criticism" are the main components of EE. "Criticism" has been found very common (30% to 70%) and equally distributed in spouse and parents, while "Emotional Over-Involvement" is less common (8% to 30%) and equally distributed in daughters and sons, but is more frequent in mothers than fathers (Leff, 1991).

To sum up, EE may be considered as an indicator of the emotional "temperature" of the family, a sign of the intensity of the emotional level of a given relative in a given time period. As Brown et al. (1972, p. 246) initially wrote,

The threshold levels of "high EE" have been empirically determined on the basis of their ability to predict schizophrenia relapses in the nine months after discharge. Several studies have been carried out that focus on the relationship between EE level at hospital admission and relapses. Brown et al. (1972) first observed that 58% of patients living in high EE environments relapsed, as compared to 12% of patients living in low EE environments. These results have been replicated in several studies (Vaughn & Leff, 1976a, 1976b; Vaughn et al., 1984; Leff & Vaughn, 1985; Moline et al., 1985; Jenkins et al., 1986; Nuechterlein et al., 1986; Tarrier et al., 1989; etc.), showing that high EE in the patient's family is associated with a three-to-fourfold greater risk of relapse in the 9 or 12 months following discharge, as compared to relapse risk in patients living in low EE.

Three factors tend to interact with EE (they also, in turn, interact with each other):

1) Medication: drugs protect patients living in low EE families from stressful life events, and help in protecting patients living in high EE families, but are less effective in presence of both high EE and life events. Some studies (Hogarty et al., 1988) show that low dosage neuroleptics (5-10 mg. of Fluphenazine Decanoate every two weeks) are more effective if used flexibly, according to family EE.

2) Time of Face-to-Face Interaction with Relatives at High EE: if this time exceeds 35 hours per week, the likelihood of relapses markedly increases (one of the goals of some psychosocial approaches based on the EE studies is to reduce weekly hours of face-to-face interaction).

3) Life Events: life events (e.g., death of a relative, major changes in work or life situation, etc.) precipitate relapse, and have an additive effect with EE. In high EE patients, life events are not necessary to provoke relapse, because high EE is in itself a sufficient stressor (Birley & Brown, 1970; Leff et al., 1973; Leff & Vaughn, 1980). Falloon et al. (1984) believe that the higher EE, the smaller the life event is needed to provoke relapse.

However, high EE is not specific to relatives of schizophrenic patients; it has been found also in families of patients with depression, bipolar disorder, anorexia nervosa, and obesity (Vaughn & Leff, 1976b; Hooley et al., 1986; Miklowitz et al., 1988; Hodes & Le Grange, 1993; etc.), and in families of patients suffering from non-psychiatric conditions, such as epilepsy, diabetes, Parkinson's disease, ulcerative colitis, and Crohn's disease. Finally, the EE construct has also been validated in transcultural studies.

The EE index has shown satisfactory validity in psychophisiological studies and in studies on direct interaction.

In the first kind of studies, Skin Conductance (SC) response has been found higher in patients who expect to interact with a high EE relative, or who are in the same room with him, rather than in a room with a low EE relative (Turpin et al., 1988; Tarrier, 1989).

The studies on direct interaction have employed the Affective Style (AS) index, which is quite similar to the EE index and is based on three scales, notably criticism (harsh or benevolent), intrusiveness, and induction of guilt feelings. Both EE and AS scores are related and predict relapses (Valone et al., 1983; Miklowitz et al., 1984). High EE relatives, particularly those with high "Criticism," are more critical than low EE relatives, and tend to be more intrusive and to induce more guilt feelings, particularly those at high "Emotional Over-Involvement." Furthermore, EE is strictly related with Communication Deviance (CD) score, measuring parents' inability to establish and maintain a shared focus of attention. According to some longitudinal studies by Goldstein (1987), high CD, negative AS, and high EE precede the onset of schizophrenic symptoms and increase the likelihood of their appearance up to four times.

Concerning the type of interaction between patient and his relatives, Leff (1988), among others (Miklowitz et al., 1989; Strachan et al., 1989; Birchwood & Tarrier, 1992; etc.), stated that linear models are rather simplistic, whereas circular models are more appropriate to understand EE. In other words, the association between high EE in the family and an increased relapse rate does not necessarily imply that high EE is the cause of the patient's relapse; it could also represent the family's response to a particularly difficult or disturbed patient, who later may relapse because of the difficult course of his own disorder. Furthermore, a vicious circle may develop between relatives' EE and patient's symptoms: the former may represent an attempt to cope, while the patient may worsen because of relatives' negative EE. At any rate, since family EE and patient characteristics influence each other, according to General System Theory, by modifying one of the two elements of the system, it should be possible to modify the other element as well.

Several studies have been realized in recent years to test this hypothesis, and it has, in fact, been shown that by modifying EE level in the family of a schizophrenic patient, it is possible to lower the relapse rate. Different approaches to family therapy have been used, some of them (such as the one developed in North America by Ian Falloon) being more behavioral than others (such as the approach developed in England by Julian Leff). The family therapy approach derived by EE studies is generally called "psychoeducational" (Anderson et al., 1986), because it involves, first of all, an education of the family about the nature of the mental illness suffered by the schizophrenic member, about the course of the disorder, etc. In this way the family members are reassured and, believing that the patient suffers from an illness biological in nature, they feel less guilty and stop putting pressure on the patient or asking him to change his behavior, which is presumably what he is not able to do.

Interestingly, one author (Greenley, 1986) has tried to reconceptualize the EE factor in the framework of a wider theory of "social control"; he has postulated that the major components of EE ("Criticism" and "Emotional Over-Involvement") provide support for viewing it as a measure of "High Intensity Interpersonal (HII) Social Control." This mechanism is employed by the family because of the fears and anxieties raised by the fact of having a schizophrenic member, particularly when the relatives do not view the patient as ill. In fact, relatives' EE tends to be higher when negative symptoms, rather than positive symptoms, are present, in other words, when the relatives do not see florid and unequivocal symptoms of this disorder. In this context, High Intensity Interpersonal Social Control is considered to be a stressor for the schizophrenic patient, who is not well equipped to cope with the relatives' attempts to change his behavior (which probably is the only adaptive solution available to him).

Greenley, using the data from the pioneering 1972 study by Brown, Birley & Wing, was able to show the usefulness and construct validity of this model. In particular, he demonstrated the two following hypothesis: (1) families who express more fears and anxieties about their schizophrenic member will more likely exhibit high EE; (2) families who do not view their schizophrenic member as mentally ill will show increased EE at higher level of anxieties and fears, and, conversely, families who view their schizophrenic member as mentally ill will not show increased EE at higher level of anxieties and fear.

As I said before, I will try to go beyond the discussion by Greenley and to see if psychoanalytic concepts, such as projective identification, will help us to explain how EE works.

The term was originally formulated by Melanie Klein in 1946, when she spoke of the projection of a part of the subject onto the object, with whom the subject remains identified, so that he exerts a "control" on him (or from "inside" of him). When M. Klein had this clinical intuition and used this term for the first time, she was somehow dissatisfied with it; subsequently the concept was refined and further explored by many analysts (who, incidentally, were working mostly with schizophrenic patients), to the point that this concept became of central importance in Kleinian thought. Currently, it is widely used also by non Kleinian analysts, and many authors have shown its usefulness also for patients with less severe forms of psychopathology. It can be regarded as a "bridge concept" between classical and interpersonal psychoanalysis, and a useful theoretical tool for understanding family dynamics (Zinner & Shapiro, 1972). An in-depth discussion of this concept does imply an adequate consideration of wider metapsychological problems (for example the use of metaphors), and of the relationship between this concept and other previous conceptualizations (such as the concept of transference and countertransference). For this reason, only a brief mention of its principal clinical aspects will be made here, while I refer to other papers for a longer discussion of the underlying theoretical problems (Migone, 1988b, 1989a, 1995 ch. 7). Since the concept of Projective identification today is used in different ways, and may encompass various clinical phenomena (Sandler, 1988), I will use the schema suggested, among others, by Ogden (1979, 1982).

Ogden divides the clinical phenomena of projective identification in three phases, which somehow overlap one another: (1) "projection," (2) "interpersonal pressure," and (3) "reinternalization." (These three phases are not to be confused with the three historical phases of the development of the concept of projective identification described by Sandler [1988], even if there are some similarities.) The three phases will be briefly described.

(1) First phase: "projection". It is assumed that the person who uses projective identification has first of all an unconscious need to get rid of a part of himself and to project it onto someone else. Various reasons may explain why a person needs to project. Kleinian authors use a metapsychological (although concrete) jargon, and speak of bad parts of the self that could be considered dangerous for the self, or, vice-versa, good parts that could be in danger of being destroyed by the bad parts of the self. These, then, need to be put into someone else in order to be protected and kept safe. These explanations are scarcely testable, although the concept of projection is commonly accepted as an explanation of some clinical and social phenomena, such as scapegoating, for example. At any rate, such metaphors may prove to be useful in understanding complex clinical situations and overcoming some especially difficult moments encountered in the course of a therapy. What is important to know here is that this first phase alone does not yet involve the use of the term projective identification; it involves simply the concept of projection.

Projection can be conceived of as an intrapsychic phenomenon, not necessarily affecting the object (the other person), who may be unaware of being the target of projective identification. In this case, projection may have defensive functions for the subject, who does not need to concretely modify or "control" the object. For a more in-depth discussion of the concept of projection, with its metapsychological and clinical aspects, I refer to the classical discussions by authors of the "Freudian" tradition, on the one hand (Freud, 1885, 1896, 1911, 1915, 1921, etc.; A. Freud, 1936; etc.), and of the "Kleinian" tradition, on the other (M. Klein, 1930, 1931; Isaacs, 1948; Segal, 1973; etc.). The most interesting aspect of projective identification, for our present discussion, concerns the second phase.

(2) Second phase: "interpersonal pressure". This phase is the one most directly related to EE concept and Greenley's concept of "High Intensity Interpersonal Social Control." There are two main differences between simple projection and projective identification. First, in the latter situation the person must be involved in an actual interpersonal relationship with another person, and not simply in a fantasized one. (In a sense, using EE terminology, we might say that the length of time spent in "face-to-face" interactions between the patient and his relatives may increase the intensity of this phenomenon.). Second, the person shows an interpersonal pressure, or control, to make sure that the other behaves in a manner consistent with the feeling that he has projected on him. Using Greenley's words, we might say that the person exerts a "High Intensity Interpersonal Social Control" because he needs the other to change his behavior. Supposedly, if the other does not change his behavior, he comes to represent a threat for the projector; therefore the projector continuously needs, in subtle or open ways, to exert various kinds of pressure to ensure that the person who received the projection really is the person that the projector wanted him to be. In a way, we can also conceive projective identification as a projection that has not been completely successful, so that the projector needs to exert pressure on the object to reassure himself of the success of this defensive operation. It is for this reason that some authors (e.g., Kernberg, 1987) conceive projective identification as a primitive defense mechanism (present mostly in borderline and psychotics), while projection is considered to be a more mature and successful defense (present mostly in neurotics). However, other authors (e.g., Meissner, 1988) reject the distinction between projection and projective identification made on the basis that the latter induces the object of the projection to respond, claiming that "complementary pulls" are always at work in all projections occurring within an interpersonal context. Whether we accept this equation between projection and projective identification or not, the presence of an interpersonal context for phenomena such as "complementary pulls" must occur. In this article I prefer to use the term projective identification, rather than simply projection, because it is the one that has been the most widely employed by those authors who have studied these complex interpersonal phenomena.

The clinical phenomena considered as examples of this interaction usually are intimate or close relationships, such as the mother-child relationship or the patient-analyst relationship. In all these situations there is some form of dependency of one person upon another, in some cases because of physical or psychological needs of survival, similar to those relationships studied in the framework of the "double bind" concept (Bateson et al., 1956). Typically, a therapist who "receives" a projective identification from a patient may develop a new set of feelings, and only during later self-scrutiny come to understand that they, so to speak, "belonged" to the patient. Furthermore, since this process is unconscious and can be very subtle, sometimes the therapist, in a way, may "become" someone else. Interpersonal transmission of affect is well known in psychotherapy (as well as out of psychotherapy), and this mutuality of emotional response was already illustrated long ago by the simultaneous psychophysiological recordings of patient and therapist (Greenblatt, 1959). For this phenomenon, Redl (1966) suggested the term "contagion" of mood. Luborsky (1984, pp. 137-139), discussing this issue, says that there seems to be evidence that neither extremely field-dependent therapists (who may be more apt to get caught up by such contagion) nor highly field-independent therapists (who may be too uninvolved and unable to form a warm relationship with the patient) are helpful. He argues that moderately field-dependent therapists may be more effective, and that alertness to the existence of this phenomenon helps to preserve therapist's equanimity.

We may recall that Paula Heimann (1950), in her pioneering work on countertransference, said that the feelings of the therapist are a "creation" of the patient. Actually, long before Heimann's contribution, it was Helene Deutsch who clearly anticipated the enlarged view of countertransference in a paper, published in 1927, on "Occult processes occurring during psychoanalysis." Indeed, some authors have linked the concept of projective identification to phenomena such as telepathy, folie à deux, Jung's concept of "psychic infection," "Devil's possession," "evil eye," suggestion, hypnosis, and the like, all phenomena in which there is a threat to personal identity and autonomy (Bilu, 1988; Bolko & Merini, 1988, 1991; etc.).

The interest around the concept of projective identification started with many analysts when they were working with extremely difficult or regressed psychotic patients and felt very uncomfortable with them, overly "controlled" or under pressure, or experienced a new set of feelings that were difficult to index as "countertransference." For them, it was more clinically useful (and possibly reassuring) to believe that in those moments they were not reexperiencing an old (and not well analyzed) aspect of themselves, but that a new and disturbing feeling was simply "put into them" by these very sick patients. This, in my opinion, is the origin of this concept, and it might explain the increased need to change the old terminology "transference/countertransference" into the new one, "projective identification/projective counteridentification" (a term coined by Grinberg in 1957). Furthermore, studies focusing on the concept of projective identification, which ran in parallel to those concerned with a new "enlarged" view of countertransference that started in the fifties (Heimann, 1950; etc), produced many interesting clinical intuitions, such as the use of the analyst's own feelings to know and understand the patient's unconscious, the concept of "evocation of a proxy" (Wangh, 1962), "externalization" (Brodey, 1965), role "actualization" and "role-responsiveness" in the transference (Sandler, 1976), the use of the patient's feelings on the part of the analyst to know and understand (or even "supervise" or "interpret") himself (Searles, 1975; Langs, 1978; Hoffman, 1983), etc.

(3) Third phase: "reinternalization". This phase is related to treatment: if it does not occur, there is no change in the patient's psychological functioning and consequently he still needs to use projective identification. In this phase, through the therapeutic interaction (and in a complex and not fully understood way) the patient "reinternalizes" the part that was projected into the therapist, because he is ready now to keep it inside of himself. According to the various metaphorical explanations that have been given to this phenomenon, the therapist must "digest" or "metabolize" this dangerous part, and make it ready to be later "digested" by the patient. In other words, during the therapeutic interaction this projected part is transformed by the therapist and made more manageable by the patient. This process (with the concepts of "container," "contained," etc.) was first described by Bion (1962, 1963), who understood it in a concrete way, and conceived that the child can reinternalize the bad parts previously projected into the mother after she has transformed them with her thought activity, i.e., her "reverie." (We can see here also an echo of the concepts of "holding" and "good enough mother" of Winnicott, 1958.)

According to a non-metapsychological explanation, and using a learning model, the patient, in the course of the interaction with his therapist who (maybe for months) tolerates the anxieties and fears that have been projected into himself, learns how the analyst does it, learns new skills or adaptive behaviors useful to cope with emotional stressors. For example, the therapist may show the patient, often through his own behavior rather than through verbal interpretation, that it is indeed possible to tolerate stressful feelings (anxieties, fears, depression, persecutory ideas, suicidal thoughts, etc.) and to survive. Winnicott (1958), among others, said that an important therapeutic experience consists in the therapist's surviving the patient's destructive projections and provocations. The therapist may talk about these feelings, and at times eventually even look at them with the instrument of irony. The improvement may occur also because the fears or anxieties are explained or interpreted - they are changed into something less dangerous or stressful. Many authors, instead of emphasizing the cognitive change through interpretation, underline that most often the improvement occurs because the therapist simply does not "discharge" again onto the patient the projected feeling, and keeps it inside of himself. In fact, interpretation as such may often be experienced by the patient as a discharge, and it is the emotional containment on the therapist's side (silence, nondefensive attitude, etc.) that breaks the vicious circle and teaches the patient that it is indeed possible not to use projective identification as the only way of functioning. (Incidentally, it is interesting to note that, even in these recent investigations, many authors prefer to rely on concepts, such as "discharge," that remind us of the old metapsychological terminology of Freudian drive theory).

Returning now to the interaction patterns in high EE families, we know that the relatives who increase the likelihood of relapse in the schizophrenic member show "Criticism" and "Emotional Over-Involvement," all characteristics described by Greenley (1986) as a form of controlling behavior. But why should these family members need to be so controlling? Presumably, they believe the patient is not ill, and interpret his behavior as voluntary, as purposely inappropriate, etc., so that they feel offended in their self-esteem and discouraged about their expectations. On the contrary, if they are told by the caring staff that the patient is biologically ill, and they are educated about schizophrenia, they can stop exerting pressure on him and can learn to better accept him, hence decreasing the likelihood of relapses.

What can the concept of projective identification contribute to a greater understanding of EE? Generally speaking, trying to describe the EE phenomenon with the metapsychological terminology of projective identification, it can be assumed that projective identification goes in both directions, and reinforces itself circularly and endlessly, unless it is interrupted with a psychotherapeutic or "psychoeducational" intervention. The patient may project disturbing feelings onto his relatives, who are unable or unprepared to contain them, so that they give them back again to the patient with another projection (what has been called projective counteridentification by Grinberg, 1957, 1962, 1979). With their EE, they put pressure on the patient to take them back, a process that the patient is unable to complete. Conversely, relatives may project onto the patients their own disturbing feelings (such as persecutory and guilt feelings about the illness, fears about schizophrenia, anger, frustration, etc.). "Negative symptoms" are especially prone to misattribution by the relatives, and function as a target of projective identification, since they are not easily understandable and less obviously related to schizophrenia than are hallucinations or delusions (Leff & Vaughn, 1985; Kavanagh, 1992, p. 608). Incidentally, "Emotional Over-Involvement," which represents an important component of EE, is seen by Leff (1991) as a form of "overidentification" or "symbiosis" (p. 213) of the relatives towards the patient, and as a direct expression of their sense of guilt (p. 215). Indeed, it is very likely that the sense of guilt is the most important unwanted and unacceptable part of themselves that the relatives unconsciously project onto the patient, who may feel overwhelmed and much less equipped than they to handle persecutory ideation. Some typical features of high EE relatives seem to endorse the idea that they are struggling with a system of projections, rather than a series of practical management problems. Typically, for example, high EE relatives seem to be constantly struggling to "control the uncontrollable" (Hooley, 1985; Brewin et al., 1991), as if they are unconsciously afraid of some implications of the patient's behavior. They often refuse to acknowledge the legitimacy of the illness, blaming the patient for his disturbance (Leff & Vaughn, 1985). Furthermore, high EE relatives are critical of the patient's character, while low EE relatives comment negatively on behavior alone, registering a fundamental acceptance of the patient as a person. High EE relatives also seem unable to complete a process of mourning the loss of the idealized individual - what the patient might have been - while low EE relatives are sad about this loss but appear to have had more progress in the mourning process and are better able to acknowledge the loss as permanent (McCarthy, 1993). Thus, low EE relatives supposedly have less sense of guilt and no need to project it onto the patient.

The concept of projective identification may be helpful, both to therapists and relatives, as a complement to psychoeducation, because it touches upon the fantasy level and can help the relatives to understand (or "interpret") some aspects of their distress. Actually, we might say that psychoeducation itself is a form of interpretation of the cause of mental illness, and it can be usefully complemented by other types of explanations and interpretations related to the psychoanalytic concept of projective identification. These types of "interpretations" can help the relatives to better understand what is going on in the family. The EE concept rests on a rather general descriptive level, and necessarily implies a lack of specificity in the therapeutic intervention, while an understanding of the specific projective identifications that are behind the high EE of a given relative (whether these projections involve sense or guilt or other feelings) may allow more accurate, and possibly more effective psychoeducational (i.e., interpretive) interventions.

More specifically, the process described in the third phase of projective identification according to Ogden ("reinternalization") may be the same process that occurs during the psychoeducational treatment. In a way, the therapist teaches the relatives to behave like those psychoanalysts who are familiar with the concept of projective identification. In other words, the therapist teaches the relatives to contain the disturbing feeling they receive from the patient (or from the wrong idea they have of the patient), and therefore to lower their EE, breaking the vicious circle of anxieties and fears, so that the patient feels relieved. In the same way that the psychoanalyst is pable to feel less anxious or less distressed, thanks to the knowledge of projective identification, the relatives of the schizophrenic patient are able to diminish their pressures and criticisms and lower their EE, thanks to the knowledge of schizophrenia (which is "interpreted" to them), they learn, for example, that schizophrenia has a strong biological component, that the patient is not totally responsible for his behavior, that they are not guilty, that they should change their expectations in order to prevent frustrations, and so forth.

We have spoken about relatives' EE, but what about therapists' or staff workers' EE? Some interesting recent studies have focused on the EE level in the staff working with long-term adult mental patients (Moore, Ball, & Kuipers, 1992; Moore & Kuipers, 1992). A wide range of EE level was found in staff members; high EE in the staff has not been found related to identified stressors in the workplace (such as staff shortage, poorly defined roles, etc.) or with staff workers' general health, but was more dependent on the specific clinical attributes of the patient, i.e., the presence of nuclear symptoms of schizophrenia. This suggests that EE does not represent a stable, enduring trait-like quality, but a characteristic of the state of ongoing transactions between two individuals. In these studies, the staff members' approach did seem to have positive impact on patients' behavior; for example, the staff's approach rated as low-EE facilitated a positive self-expression in the patient, as shown by a higher score for patient self-affirmation.

In summary, it is possible to say that the psychotherapeutic work with schizophrenic patients over the last generation, which has led also to the development of the concept of projective identification, may be of interest in the research on psychosocial treatment of schizophrenia based on the EE model. It seems that different investigations, based on different theoretical assumptions, have come up rather similar clinical conclusions. While the psychoeducational approach is simply based on a different interpretation of the illness, interpretation is the main intervention on which psychoanalytic therapy itself is based. Patients are supposed to change after they acquire new insights into their own symptoms and are educated about them. After all, Freud himself (1905, 1917) once defined psychoanalysis as a form of "after-education" (Nacherziehung).

One of the limitations related to the concept of projective identification (as well as to other psychoanalytic concepts) is that it is too overinclusive and elastic (and not easily "falsifiable," in Popper's [1934, 1957] sense). Nevertheless, its popularity among psychotherapists might express the need for concepts of this kind in understanding family interactions and the transmission of emotions between people.

I would like to make two further comments on this complex and interesting aspect of psychotherapy of schizophrenia, and discuss some current developments of psychoanalytic theory and technique, in order to underline some similarities in different psychotherapeutic models.

The first comment is related to the well-known theory of psychotherapy of the "San Francisco Psychotherapy Research Group" led by Weiss and Sampson (Weiss, Sampson & the Mount Zion Psychotherapy Research Group, 1986; Weiss, 1993). According to this model, called "Control-Mastery Theory," the patient improves only if the therapist passes a "test" that the patient unconsciously asks him to pass. In the transference process, the patient repeats past relationships, and may induce the therapist to behave like previous transference figures. This theory, however, does not give emphasis to repetition compulsion or drive discharge, but highlights the patient's unconscious plan to test the therapist, in the patient's hope that the therapist will behave differently than was expected. An interesting aspect of this model is that it represents an autonomous theory of therapy, with few metapsychological concepts (Eagle, 1984, ch. 9; for a discussion of Control-Mastery Theory from the point of view of cognitive science, see Migone & Liotti, 1998). There are various kinds of tests, and some of them might be very similar to the tests that have to be passed by therapists when they are the target of a patient's projective identification. For example, as Ogden (1982, pp. 83-84) has also observed, if a patient realizes that his therapist is capable, after all, to "survive" or not feel too angry or depressed, despite the many attacks, criticisms, and depressive ideas the patient unconsciously throws onto him, he may feel relieved, and may start to believe that it is indeed possible to tolerate very disturbing feelings without being destroyed by them, or he may learn some adaptive skills from his therapist by unconsciously identifying with him, and so on.

The second and final comment is more critical, and is not directly related to EE but to the changing nature of psychoanalytic technique. As we know, psychoanalytic technique has changed considerably over the last decades. Some changes are related to the increased importance given today to the concept of projective identification. It seems that a certain use of this concept in psychoanalysis has increased the crisis of interpretation. This tendency started in psychoanalysis in the mid sixties, when the hermeneutic critique (Ricoeur, 1965; Habermas, 1968; etc.) led to relativize the concept of "true interpretation" as main curative factor, and emphasized the importance of other curative factors related to the therapeutic interaction in the "here and now" (construction of new meanings, of narratives, etc.). This concept was later elaborated by many psychoanalytic authors (Spence, 1982; etc.). Therefore, while the concept of interpretation was gradually becoming a "weak concept" in psychoanalysis, new concepts were replacing it as key ones in the psychoanalytic model (Galli, 1988a, 1988b, 1990; Migone, 1989b, 1989c, 1995). One of these concepts is that of "setting" or "therapeutic environment," with its functions of "container" and "contained" developed by Bion (1962, 1963, 1967), which rely heavily on the concept of projective identification. These theoretical developments are also related to the increased degree of severity of patients undergoing psychoanalytic treatment, especially during the second half of this century.

In saying this, I want simply to point out that, generally speaking, there are two ways in which the concept of projective identification is used clinically. Some authors still rely on the central tool of interpretation; they use this concept to better understand the dynamics of the interaction, and to interpret these dynamics to the patient, i.e., to make him aware of his own responsibilities (for example, interpreting projection). This could be seen as a traditional technique, in which the concept of projective counteridentification overlaps the "enlarged" use of countertransference, and in which the analyst, through his interpretations, tries to distinguish between his nontransference responses to the patient and his countertransference in the "narrow" sense.

On the other hand, other authors (for example, Ogden, 1982, p. 20, and some followers of Bion), explicitly emphasize that interpretation is not needed to bring about therapeutic change. Rather, it is through the interaction that the patient improves. As explained earlier, some patients may perceive interpretations as a way to discharge those emotions they had tried to project. These patients may learn more from the therapist's ability to tolerate frustrations, lack of understanding, tensions, anxieties, confusion, ambiguities, etc., than from his attempts to interpret or to explain. In this way, therapy often becomes a silent interaction, a silent containment, a work of patience, and a working through of our own feelings.

It is apparent here that this issue is also related to the EE concept. Is low EE beneficial because it is a consequence of new meanings of the illness given to the family (through "interpretation" or psychoeducation), or because it simply implies a breaking of the vicious circle of negative emotions, a "corrective emotional experience," a sort of "cure without understanding"? It is interesting to note that the type of family therapy derived from EE studies is often called "psychosocial treatment," not "psychotherapy," and there is an ongoing discussion of the correct definition of the psychological treatments of schizophrenia. A possible definition of "psychotherapy" (which includes psychoanalysis) is the ambitious attempt to modify the individual, his psychic structure, while "psychosocial" treatments or interventions could mean primarily the modification of the patient's environment, such as his family, since it would be too frustrating to expect a change in the schizophrenic patient's behavior in the short run, only with psychotherapy. Schizophrenic patients, like other patients with psychoses and severe personality disorders, have an alloplastic, rather than autoplastic, way of functioning, i.e., their functioning relies heavily on environment. Of course, with psychosocial treatment as well the aim in the long run is to modify the patient (for example, to prevent relapse), but most often the patient's improvement is not stable since it is due to the environmental changes that are constantly needed.

We can say that this distinction between psychotherapy and psychosocial therapy has a long history in psychoanalysis, where the concept of insight was used originally to give stability to a patient's change, to make him autonomous from the environment. Antinomy in the usefulness of insight versus emotional experience, remembering versus experiencing, specific versus general (or nonspecific) therapeutic factors, has always been present in the history of psychoanalysis. It dates back to Ferenczi & Rank (1924), was raised again by Alexander et al. (1946), and today is more alive than ever, as is well shown by the popularity of Object Relation Theory. Although I am not questioning the therapeutic efficacy of corrective therapeutic experiences, it seems to me that there is a risk here that without acknowledging it, we may simply slip back to the old European anthropo-phenomenological approach (Jaspers, Binswanger, etc.). It should not be forgotten that psychoanalysis was, in fact, developed by Freud to add something new to these approaches and to differentiate from them.

Incidentally, a similar observation can be made concerning the concept of Self, which today has become very popular. Current Self Psychology deemphasizes the usefulness of interpreting the repressed content, i.e., the work on Ego-Id conflicts, in favor of work on more experience-near feelings, i.e., using the instrument of empathy.

The phenomenologists, too, were not relying on the concept of the unconscious and were working with psychotic patients. Interestingly, certain quotations by some of the best known phenomenological psychiatrists (such as Binswanger) seem to be taken from contemporary psychoanalysts working with the concept of projective identification. Binswanger used to say that an important and specific aspect of the attitude of the therapist, when he meets the patient, is to feel inside of himself the feelings that the patient evokes in him, to let these vibrations stay within himself, to experience what they evoke in him and, perhaps, to talk about them with the patient.

I would like to conclude with an exhortation: while we build "new" theories of psychotherapy, we should not forget the experience of these colleagues of one century ago.

Alexander F., French T.M. et al. (1946). Psychoanalytic Therapy: Principles and Applications. New York: Ronald Press. Internet edition: http://www.psychomedia.it/pm/modther/probpsiter/alexan-2.htm.

Anderson C.M. et al. (1986). Schizophrenia and the Family. A Practictioner's Guide to Psychoeducation and Management. New York: Guilford Press.

Bateson G., Jackson D.D., Haley J., Weakland J. (1956). Towards a theory of schizophrenia. Behavioral Science, 1, 4: 251-264 (trad. it.: Verso una teoria della schizofrenia. In: Cancrini L., a cura di, Verso una teoria della schizofrenia. Torino: Boringhieri, 1977, pp. 75-100).

Bilu Y. (1987). "Possession" and mechanisms of interiorization and exteriorization. In Sandler J., editor, Projection, Identification, Projective Identification. Madison, CT: Int. Univ. Press, 1988, ch. 11.

Bion W.R. (1962). Learning from Experience. London: Heinemann (trad. it.: Aprendere dall'esperienza. Roma: Armando, 1983).

Bion W.R. (1963). Elements of Psychoanalysis. London: Heinemann (trad. it.: Gli elementi della psicoanalisi. Roma: Armando, 1981).

Bion W.R. (1967). Second Thoughts: Selected Papers on Psychoanalysis. London: Heineman.

Bion W.R. (1967). Second Thoughts. New York: Aronson.

Birchwood M. & Tarrier N., editors (1992). Innovations in the Psychological Management of Schizophrenia. New York: Wiley.

Birley J.L.T. & Brown G. (1970). Crises and life changes preceding the onset or relapse of acute schizophrenia: clinical aspects. British Journal of Psychiatry, 116: 327-333.

Bolko M. & Merini A. (1988). Introduction to the Italian translation of Sandler J., 1988 (Proiezione, identificazione, identificazione proiettiva. Torino: Bollati Boringhieri, 1988, pp. 9-15).

Bolko M. & Merini A. (1991). Osservazioni sulla identificazione proiettiva: "Through the looking glass". Psicoterapia e Scienze Umane, XXV, 4: 19-33.

Brewin C.R., McCarthy B., Duda K., Vaughn C.E. et al. (1991). Attributions and expressed emotion in the relatives of patients with schizophrenia. Journal of Abnormal Psychology, 100: 546-554.

Brodey W. (1965). On the dynamics of narcissism: I. Externalization and early ego development. Psychoanal. Study Child, 20: 165-193.

Brown G., Carstairs G.M., Topping G. (1958). Post-hospital adjustment of chronic mental patients. The Lancet, Sept. 27, 1958, pp. 685-689.

Brown G.W. et al. (1962). Influence of family life on the course of schizophrenic illness. Brit. J. Prev. Soc. Med., 16: 55.

Brown G.W., Birley J.L.T., Wing J.K. (1972). Influence of family life on the course of schizophrenic disorders. Br. J. Psychiatry, 121: 241-258.

Deutsch H. (1926). Okkulte Vorgänge während der Psychoanalyse. Imago, 12: 418-433. Engl. transl.: Occult processes occurring during psychoanalysis. In: G. Devereux, editor, Psychoanalysis and the Occult. New York: International Universities Press, 1953; also in: Helene Deutsch, The Therapeutic Process, the Self, and Female Psychology: Collected Psychoanalytic Papers (History of Ideas Series. Edited by Paul Roazen). New Brunswick, NJ: Transaction Publishers, 1992.

Eagle M.N. (1984). Recent Developments in Psychoanalysis. A Critical Evaluation. New York: McGraw-Hill (reprint: Cambridge, MA: Harvard Univ. Press, 1987). Internet edition of chapter 12: "Developmental defect versus dynamic conflict": http://www.psychiatryonline.it/ital/docum8-i.htm  (trad. it.: La psicoanalisi contemporanea. Bari: Laterza, 1988. Edizione su Internet del cap. 11, ed. or. cap. 12: "Carenze di sviluppo e conflitto dinamico": http://www.publinet.it/pol/ital/documig8.htm).

Falloon I.R.H., Boyd J.L., McGill C.W. (1984). Family Care of Schizophrenia. A Problem Solving Approach to the Treatment of Mental Illness. New York: Guilford.

Ferenczi S. & Rank O. (1924). Entzvicklungsziele der Psychoanalyse. Wien: Internationaler Psychoanalytischer Verlag (English transl. by Caroline Newton: The Development of Psychoanalysis. New York: Nervous and Mental Disease Publishing Co., 1925; trad. it.: Prospettive di sviluppo della psicoanalisi. In: Fondamenti di psicoanalisi, Vol. III. Rimini: Guaraldi, 1973).

Freud A. (1936). The Ego and the Mechanisms of Defense. New York: International Universities Press, 1946.

Freud S. (1895 [1892-97]). Extracts from the Fliess papers. Draft H: Paranoia. SE, 1.

Freud S. (1896). Further remarks on the neuro-psychoses of defence. SE, 3.

Freud S. (1905). On psychotherapy. SE, 7.

Freud S. (1911). Psychoanalytic notes on an autobiographical account of a case of paranoia (dementia paranoides). SE, 12.

Freud S. (1915). A case of paranoia running counter to the psychoanalytic theory of the disease. SE, 14.

Freud S. (1917). Introductory lectures on psychoanalysis (1916-17). Part III. General theory of the neuroses (1917). Lecture XXVIII: Analytic therapy. SE, 15.

Freud S. (1921). Some neurotic mechanisms in jealousy, paranoia and homosexuality. SE, 18.

Galli P.F. (1988a). Le ragioni della clinica. Psicoterapia e Scienze Umane, XXII, 3: 3-8.

Galli P.F. (1988b). Clinical theory and psychotherapeutic practice between dogmas and stereotypes. Per la salute mentale/For Mental Health, 1, 2/3: 113-119.

Galli P.F. (1990). "Psychoanalysis as the story of a crisis". Presented at the Rapaport-Klein Study Group Annual Meeting, Austen Riggs Center, Stockbridge, Massachusetts, June 10, 1990. Internet edition: http://www.psychomedia.it/rapaport-klein/galli90.htm  (Italian version: Le psicoanalisi e la crisi della psicoanalisi. Ricerche di Psicologia, 1990, XIV, 4: 39-58 [also in: Canestrari & Ricci Bitti, 1993, pp. 173-188; also in: Galli P.F., La persona e la tecnica. Appunti sulla pratica clinica e la costruzione della teoria psicoanalitica. Milano: Il Ruolo Terapeutico, 1996, pp. 57-75; new edition: Milano: Franco Angeli, 2002]. German version: Die Psychoanalysen und die Krise der Psychoanalyse. In: Martin Kuster, editor, Entfernte Wahreit. Tübingen: Diskord, 1992, pp. 146-165. English version: Crisis of psychoanalysis? From the scientistic solution to the semantic perspective. Italian Journal of Psychiatry and Behavioral Sciences, 1994, IV, 1: 23-28).

Goldstein M.J. (1987). The UCLA high-risk project. Schizophrenia Bull., 13: 505-514.

Greenblatt M. (1959). Discussion of papers by Saslow and Matarazzo and Lacey. In E. Rubinstein & M. Parloff, editors, Research in Psychotherapy. Washington, DC: American Psychological Association, 1959, pp. 209-220.

Greenley J.L. (1986). Social control and expressed emotion. J. Nerv. Ment. Dis., 174: 24-30.

Grinberg L. (1957). Perturbaciones en la interpretacion motivadas por la contraidentificacion proyectiva. Revista de Psicoanalisis, 14: 23-30.

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

Grinberg L. (1979). Countertransference and projective counteridentification. Cont. Psychoanal., 15, 2.

Grotstein J. (1981). Splitting and Projective Identification. New York: Aronson (trad. it.: Scissione e identificazione proiettiva. Roma: Astrolabio, 1983).

Gunderson J.G., Frank A.F., Katz H.M., Vannicelli M.L., Frosch J.P., Knapp P.H. (1984). Effects of psychotherapy in schizophrenia: II. Comparative outcome of two forms of treatment. Schizophrenia Bull., 10, 4: 564-598.

Habermas J. (1968). Erkenntnis und Interesse. Frankfurt: Suhrkamp (trad. ingl.: Knowledge and Human Interest. Boston: Beacon Press, 1971; trad. it.: Conoscenza e interesse. Bari: Laterza, 1970).

Heimann P. (1950). On countertransference. Int. J. Psycho-Anal., 31: 81-84. Also in R. Langs (editor), Classics in Psychoanalytic Technique. New York: Aronson, 1981, pp. 139-142 (trad. it.: Sul controtransfert. In: C. Alberella & M. Donadio, a cura di, Il controtransfert. Napoli: Liguori, 1986, pp. 81-86).

Heimann P. (1950). On countertransference. International Journal of Psycho-Analysis, 31: 81-84. Also in: Langs R., editor, Classics in Psychoanalytic Technique. New York: Aronson, 1981, pp. 139-142.

Hodes M. & Le Grange (1993). Expressed Emotion in the investigation of eating disorders: a review. Int. J. Eating Disorders, 13, 3: 279-288.

Hoffman I.Z. (1983). The patient as interpreter of the analyst's experience. Contemp. Psychoanal., 19: 389-422 (trad. it.: Il paziente come interprete dell'esperienza dell'analista. Psicoterapia e Scienze Umane, 1995, XXIX, 1: 5-39).

Hogarty G.E. et al. (1988). Environmental/personal indicators in the course of schizophrenia research group. Dose of fluphenazine, familiar expressed emotion and outcome in schizophrenia: results of a two year controlled study. Archives of General Psychiatry, 45: 797-805.

Hooley J.M. (1985). Expressed Emotion: a review of the critical literature. Clinical Psychology Review, 5: 119-139.

Hooley J.M. et al. (1986). Levels of Expressed Emotion and relapse in depressed patients. British Journal of Psychiatry, 148: 642-647.

Isaacs S. (1948). The nature and function of phantasy. Int. J. Psycho-Anal., 29: 73-97.

Jureidini J. (1990). Projective identification in general psychiatry. British Journal of Psychiatry, 157: 656-660.

Jenkins J.H. et al. (1986). Expressed emotion, maintenance pharmacotherapy, and schizophrenic relapse among Mexican-American. Psychopharmacology Bulletin, 22: 621-627.

Kavanagh D.J. (1992). Recent develppments in Expressed Emotion and schizophrenia. Br. J. Psichiatry, 160: 601-620.

Kernberg O.F. (1987). Projection and projective identification: developmental and clinical aspects. J. Am. Psychoanal. Ass., 4: 795-819. Also in Sandler J., editor, 1988, ch. 7.

Klein M. (1930). The importance of symbol formation in the development of the Ego. In The Writings of Melanie Klein (editors: R.E. Money-Kyrle et al.). London: Hogarth Press, 1975.

Klein M. (1931). A contribution to the theory of intellectual inhibition. In The Writings of Melanie Klein (editors: R.E. Money-Kyrle et al.). London: Hogarth Press, 1975.

Klein M. (1946). Notes on some schizoid mechanisms. Int. J. Psycho-Anal., 27: 99-110. Also in The Writings of Melanie Klein (editors: R.E. Money-Kyrle et al.). London: Hogarth Press, 1975 (una versione del 1952 è stata tradotta negli Scritti, 1921-1950. Torino: Boringhieri, 1978, pp. 409-434).

Langs R. (1978). A model of supervision: the patient as unconscious supervisor. In Technique in Transition. New York. Aronson, 1978, pp. 587-625.

Leff J. (1988). La famiglia del paziente schizofrenico: un punto di forza per il terapeuta. In: Migone P., Martini G., Volterra V. (editors), New Trends in Schizophrenia. Vol. I, pp. 330-353. S. Maria a Vico (Caserta): Fondazione Centro Praxis.

Leff J.P. (1991). Achieving emotional separation between high EE relatives and schizophrenic patients (Presented at the Xth International Symposium for the Psychotherapy of Schizophrenia. Stockholm, August 11-15, 1991). In: A. Werbart & J. Cullberg, editors, Psychotherapy of Schizophrenia: Facilitating and Obstructive Factors. Oslo: Scandinavian University Press, 1992, pp. 212-224.

Leff J.P. et al. (1973). Life events and maintenance therapy in schizophrenic relapse. Brit. J. Psychiatry, 123: 659.

Leff J.P. & Vaughn C.E. (1980). The interaction of life events and relatives expressed emotion in schizophrenia and depressive neurosis. Brit. J. Psychiatry, 136: 146.

Leff J.P. & Vaughn C.E. (1985). Expressed Emotion in Families: Its Significance for Mental Illness. New York: Guilford.

Luborsky L. (1984). Principles of Psychoanalytic Psychotherapy. A Manual for Supportive-Expressive Treatment. New York: Basic Books (trad. it.: Princìpi di psicoterapia psicoanalitica. Manuale per il trattamento supportivo-espressivo. Torino: Bollati Boringhieri, 1989).

McCarthy B. (1993). Personal communication.

McGlashan T.H. (1984). The Chesnut Lodge Follow Up Study, II: Long term outcome of schizophrenia and the affective disorders. Arch. Gen. Psychiatry, 41: 586-601.

Malin A. & Grotstein J. (1966). Projective identification in the therapeutic process. Int. J. Psychoanal., 47: 26-31.

Meissner W.W. (1988). Projection and projective identification. In Sandler J., editor, 1988, chapter 2.

Migone P. (1988a). "Emotività' Espressa" (EE): un'interpretazione psicoanalitica basata sul concetto di Identificazione Proiettiva. In: Migone P., Martini G., Volterra V. (editors), New Trends in Schizophrenia. Vol. III, pp. 279-281. S. Maria a Vico (Caserta): Fondazione Centro Praxis.

Migone P. (1988b). La identificazione proiettiva. Il Ruolo Terapeutico, 49: 13-21. Internet edition: http://www.psychomedia.it/pm/modther/probpsiter/ruoloter/rt49ip88.htm.

Migone P. (1989a). Discussione critica del concetto psicoanalitico di identificazione proiettiva secondo Ogden. Prospettive psicoanalitiche nel lavoro istituzionale, 7, 2: 159-177.

Migone P. (1989b). The evolution of the psychoanalytic concept of "true interpretation" and the scientific status of psychoanalysis one century later. Presented at the VIII^th International Forum of Psychoanalysis "Psychoanalysis one century later" (Rio de Janeiro, october 9-14, 1989).

Migone P. (1989c). Il rapporto tra psicoanalisi e scienza: dal neopositivismo a Grünbaum. Rivista di Psichiatria, 24, 3: 153-160 (see the web site: http://www.psychomedia.it/pm/modther/probpsiter/ruoloter/rt50-89.htm).

Migone P. (1991). "Expressed Emotions and Projective Identification: a psychoanalytic hypothesis on the EE factor in the treatment of schizophrenia. A bridge between psychiatric and psychoanalytic concepts". Paper presented at the Xth International Symposium for the Psychotherapy of Schizophrenia (Stockholm, August 11-15, 1991).

Migone P. (1993). Identificazione proiettiva: una interpretazione psicoanalitica della Emotività Espressa? Schizofrenia, V, 1/2: 29-41.

Migone P. (1995). Terapia psicoanalitica. Milano: Franco Angeli (A review on the Internet: http://www.psychomedia.it/pm-revs/books/migone1a.htm).

Migone P. (2004). La identificazione proiettiva. Bollettino d'informazione dell'Ordine degli Psicologi della Regione Emilia-Romagna, IX, 4: 23-31

Migone P. & Liotti G. (1998). Psychoanalysis and cognitive-evolutionary psychology: an attempt at integration. International Journal of Psychoanalysis, 79, 6: 1071-1095. The paper, the Internet discussion, and the review by Paul Williams can be downloaded from the web site http://ijpa.org/archives1.htm. (A shorter version of this article, authored by Paolo Migone, was presented as an invited paper at the panel "Psychic reality and pathogenic beliefs: the patients' theories about themselves and the relational world", at the 39th Congress of the International Psychoanalytic Association, San Francisco, July 30-August 4, 1995).

Migone P., Martini G., Volterra V., editors (1988). New Trends in Schizophrenia (Proceedings of the International Meeting New Trends in Schizophrenia. Bologna, Italy, April 14-17, 1988). S. Maria a Vico (Caserta): Fondazione Centro Praxis (4 volumes).

Miklowitz D.J. et al. (1984). Interactional correlates of expressed emotion in the families of schizophrenics. Brit. J. of Psychiatry, 144: 482-487.

Miklowitz D.J. et al. (1988). Family factors and course of bipolar affective disorders. Arch. Gen. Psychiatry, 45: 225-231.

Miklowitz D.J. et al. (1989). Is Expressed Emotion an index of a transactional process? I. Parents' Affective Style. Family Process, 28: 153-167.

Moline R.A. et al. (1985). Family expressed emotion and relapse in schizophrenia in 24 urban American patients. Am. J. Psychiatry, 142: 1078-1081.

Moore E., Ball R.A., Kuipers L. (1992). Expressed Emotion in staff working with the long-term adult mentally ill. Br. J. Psychiatry, 161: 802-808.

Moore E. & Kuipers L. (1992). Behavioural correlates of expressed emotion in staff-patient interaction. Social Psychiatry Psychiatr. Epidemiol., 27: 298-303.

Nuechterlein K.H. et al. (1986). Expressed emotion, fixed dose fluphenazine decanoate maintenance, and relapse in recent-onset schizophrenia. Psychopharmacology Bulletin, 22: 633-639.

Ogden T. (1979). On Projective Identification. Int. J. Psychoanal., 60: 357-373. Also in Ogden, 1984, chapter 2, pp. 11-37.

Ogden T.H. (1982). Projective Identification and Psychotherapeutic Technique. New York: Aronson (trad. it.: Identificazione proiettiva e tecnica psicoanalitica. Roma: Astrolabio, 1994. Trad. it. del cap. 6: L'identificazione proiettiva e il ricovero psichiatrico. Prospettive psicoanalitiche nel lavoro istituzionale, 1984, 2: 200-214).

Popper K.R. (1934). The Logic of Scientific Discovery. New York: Basic Books, 1959 (trad. it.: Logica della scoperta scientifica. Torino: Einaudi, 1970).

Popper K.R. (1957). Conjectures and Refutations. New York: Basic Books, 1962 (trad. it.: Congetture e confutazioni. Bologna: Il Mulino, 1972).

Redl F. (1966). When We Deal with Children. New York: Free Press.

Ricoeur P. (1965). De l'interprétation. Essai sur Freud. Paris: Seuil (trad. it.: Della interpretazione. Saggio su Freud. Milano: Il Saggiatore, 1967; trad. ingl.: Freud and Philosophy. New Haven, CT: Yale University Press, 1970).

Rutter M. & Brown G.W. (1966). The reliability and validity of measures of family life and relationships in families containing a psychiatric patient. Social Psychiatry, 1, 1: 38-53.

Sandler J. (1976). Countertransference and role responsiveness. Int. Rev. Psychoanal., 1976, 3: 43-47. Internet edition: http://ijpa.org/jsjan00.htm (trad. it.: Controtransfert e risonanza di ruolo. In: C. Alberella & M. Donadio, a cura di, Il controtransfert. Napoli: Liguori, 1986, pp. 189-197).

Sandler J., editor (1988). Projection, Identification, Projective Identification. Madison, CT: Int. Univ. Press (trad. it.: Proiezione, identificazione, identificazione proiettiva. Torino: Bollati Boringhieri, 1988).

Segal H. (1964). Introduction to the Work of Melanie Klein. London: Heineman (trad. it.: Introduzione all'opera di Melanie Klein. Firenze: Martinelli, 1968).

Searles H.F. (1975). The patient as therapist to his analyst. In Giovacchini P.L., editor, Tactics and techniques in psychoanalytic therapy. Vol. II: Countertransference. New York: Aronson, 1975. Anche in: Searles H.F., Countertransference and Related Subjects. New York: Int. Univ. Press, 1979 (trad. it.: Il paziente come terapeuta del suo analista. In: Searles H.F., Il controtransfert. Torino: Bollati Boringhieri, 1994, cap. 18, pp. 280-336).

Spence D.P. (1982). Narrative Truth and Historical Truth. Meaning and Interpretation in Psychoanalysis. New York: Norton (trad. it.: Verità storica e verità narrativa. Firenze: Martinelli, 1987).

Stone M.H. (1986). Exploratory psychotherapy in schizophrenia-spectrum patients. Bull. Menninger Clinic, 50: 287-306.

Strachan A.M. et al. (1989). Is Expressed Emotion an index of a transactional process? II. Patient's Coping Style. Family Process, 28: 169-181.

Tarrier N. (1989). Electrodermal activity, expressed emotion, and outcome in schizophrenia. Br. J. Psychiatry, 155 (suppl. 5): 51-56.

Tarrier N. et al. (1989). The community management of schizophrenia: a controlled trial of a behavioural intervention with families to reduce relapse. British Journal of Psychiatry, 154: 625-628.

Turpin G., Tarrier N., Sturgeon D. (1988). Social psychophisiology and the study of biopsychosocial models of schizophrenia. In H. Wagner, editor, Social Psychophysiology: Theory and Clinical Applications. Chichester: Wiley.

Valone K. et al. (1983). Parent expressed emotion and affective style in an adolescent at risk for schizophrenia spectrum disorders. Journal of Abnormal Psychology, 92: 399-407.

Vaughn C.E. & Leff J.P. (1976a). The influence of family and social factors on the course of psychiatric illness: a comparison of schizophrenic and depressed neurotic patients. Br. J. Psychiatry, 129: 125-137.

Vaughn C.E. & Leff J.P. (1976b). The measurement of expressed emotion in the families of psychiatric patients. Br. J. Social Clin. Psychology, 15: 157-165.

Vaughn C.E. et al. (1984). Family factors in schizophrenic relapse: replication in California of the British research on expressed emotion. Arch. Gen. Psychiatry, 41: 1169-1177.

Wangh M. (1962). The "evocation of a proxy": a psychological manoeuvre, its use as a defense, its purposes and genesis. Psychoanal. Study Child, 17: 451-472.

Weiss J. (1993). How Psychotherapy Works. Process and Technique. New York: Guilford (trad. it.: Come funziona la psicoterapia. Presentazione di Paolo Migone e Giovani Liotti. Torino: Bollati Boringhieri, 1999).

Weiss J., Sampson H. & the Mount Zion Psychotherapy Research Group (1986). The Psychoanalytic Process: Theory, Clinical Observation, and Empirical Research. New York: Guilford (trad. it. del cap. 1: Introduzione al lavoro del "San Francisco Psychotherapy Research Group". Psicoterapia e Scienze Umane, 1993, XXVII, 2: 47-65).

Winnicott D.W. (1958). Collected Papers (1931-1956). Through Pediatrics to Psycho-Analysis. New York: Basic Books (trad. it.: Dalla pediatria alla psicoanalisi. Scritti scelti. Firenze: Martinelli, 1975).

Zinner J. & Shapiro R. (1972). Projective identification as a mode of perception and behaviour in families of adolescents. Int. J. Psychoanal., 53: 523-530.

Zubin J. & Spring B. (1977). Vulnerability: a new view of schizophrenia. Journal of Abnormal Psychology, 86: 103-126.

[Note: A version of this paper appeared in Contemporary Psychoanalysis, 1995, 31, 4: 617-640. Earlier versions  were presented at the International Meeting New Trends in Schizophrenia, Bologna, Italy, April 15-17, 1988, and at the X International Symposium for the Psychotherapy of Schizophrenia, Stockholm, Sweden, August 11-15, 1991. The author thanks Giovanni de Girolamo, M.D., Morris N. Eagle, Ph.D., Brigid MacCarthy, Ph.D., and Matthew Hodes, M.D., who provided important comments and suggestions, and Daniela Cavallo, Ph.D., who helped in completing the review of the literature on Expressed Emotion.]

Paolo Migone, M.D.

Via Palestro, 14

43123 Parma

Italy

Tel./Fax +(39) 0521-960595

E-Mail <migone@unipr.it>

Web Editor: Paolo Migone. For suggestions or corrections, please e-mail to: migone@unipr.it