_____________________________________________________________________________________________________________

Fam Proc 1:63-68, 1962

Family Therapy as a Research Arena

American Psychological Association meeting, New York, 1961. This paper is a product of the Family Therapy in Schizophrenia project of the Palo Alto Medical Research Foundation, and thus reflects the ideas and experience of the entire project staff. This work, directed by Gregory Bateson, has been supported by Mental Health Project Grant OM-324 from the National Institute of Mental Health, by the Veterans Administration Hospital, Palo Alto, and by the Mental Research Institute, Palo Alto Medical Research Foundation, Palo Alto, California.

No one who has attempted or even observed family therapymeaning conjoint treatment of a family groupwould think it inapt to call the family treatment situation an arena. There, with a therapist as actual observer, and also as representing a wider audience, the members present family dramas, spectacles, and often contests. But this as a research arena? Can it be, and should it be even if it can be? My own answer is "yes" to both questions, and I would like here to discuss 1) how the family therapy situation offers some special potentialities for important research and 2) why, at least at this stage of our knowledge, grasping these research opportunities seems quite consonant with holding basic therapeutic aims in interviewing family groups. This second point had best be dealt with first. Otherwise, doubts about the propriety of any deliberate mixing of research into therapy might block free and unbiased considerations of the research potentials of the family treatment situation. I see two main reasons why research and therapy are compatible in work with families. The first, to speak plainly, is that at present we know so little about family therapy, both theoretically and practically, that whatever we essay in treatment is bound in large measure to be only tentative and hopeful. We simply have not yet developed a well-defined concept and technique of treatment with certain known powers and limits, nor can this be approximated by relying on principles taken from the practice of psychotherapy with individual patients. Even if much more were known and agreed on in that field than is now the case, family therapy is too different from individual psychotherapy for any simple transfer of rules to be reliable. Thus, there is in this area no standard of "conservative treatment" from which exploratory operations would be a clear departure, properly requiring certain decisions and safeguards because extra risks would knowingly be run. Therefore, in practicing family treatment at present one has only the choice between being cautious in overall attitudewithout having adequate guides to what appropriate caution specifically would beor being more frankly exploratory. But to be exploratory is not necessarily to be irresponsible, and indeed the reverse may be the case here, as is true in many critical human and medical situations. Probably the most important fact we do know about families in treatmentat any rate the families of schizophrenics, my primary focus hereis that they are exceedingly resistant to any change in their basic patterns of interaction. This is true in spite of their sad condition as families and as individual family members, and in spite of the particular or recurrent family upheavals or frantic behavior they may often exhibit. The real problem in therapy is not that too much is likely to occur, but too little. Therefore, not only do we need research to improve and extend our knowledge, but we need to explore, with these families, for means of promoting enough real change to give alternative forms of family behavior and interaction a chance to exist; almost any change might be an improvement on the schizophrenic systematization. This is not to deny all need for caution, or all need to avoid provoking any really destructive behavior, but we should not absorb the families' own attitudes that any real changes or freedom of action would lead to catastrophes. Their operational, if not verbal, position is "For God's sake don't rock the boat,"or more accurately, "Don't change the way it's rocking now,"but why should we adopt the premises of confused and troubled families that come to us as authorities for help? We may now consider broadly what possibilities for exploratory research the family therapy arena offers by its very nature. The nature and results of such a consideration necessarily depend considerably on one's basic premises and conceptions about the nature of "psychopathology" or "mental illness" and its understanding. In my view, these terms themselves reflect a traditional emphasis on the individual in psychiatry and psychology. This individual-oriented point of view in our efforts to understand and deal with strange or pathological behavior has led to much concern with childhood and the remote past, with particular dramatic events, with fantasies, and with concepts of internal psychic organization. These concerns, however, mainly involve reports and inferences rather than observable events; they are not very helpful in relating behavior and interpersonal transactionseither symptoms and interpersonal influences, or therapeutic changes and patient-therapist interchanges. Accordingly, there has been a growing trend in psychiatry toward a more interpersonal approach, and family therapy represents a further development along this line. Family therapy, in fact, implies a point of view which may be explicitly
1

_____________________________________________________________________________________________________________

stated here: That which is most important for our understanding and handling of pathological or symptomatic behavior is the nature of actual current interaction between real individuals, especially family members. Such behavior and communicative interaction can be observed and studied directly, and should be the primary focus of scientific attention, with the past, subjective experience, and concepts of internal psychic mechanisms left in a secondary place. An important corrolary of this point of view is that ordinary human communication and its influence must be examined and considered seriously, in terms adequate to its complexity and importance. In particular, communication always and necessarily involves interrelated multiple messages by every participant, which are at different levels of communication. This may be obvious, as when one message is conveyed by words and the other by tone, movement, or by symptomatic behavior, but it is equally true when the words of a verbal message carry two meanings, at different levels, in relation to different possible contexts of reference. For example, a wife may tell her husband, "I want you to dominate me." Specifically, she says he should be in charge of their relationship; but at another level, by directing how he should behave, she is proposing that she be in charge. Since such related multiple messages are at different levels, there cannot be identity, nor direct contradiction; yet the nature of the relationship between them, and thus the influences exerted and effects produced by them jointly, is open to wide variation. There may be similarity and reinforcement, as when a mother says "Come here, dear" to her child in a warm voice, or with a beckoning movement. Or there may be incongruence and conflicting influence in many ways. Mother may say "Come here, dear" in an angry voice; she may say "Come here, dear" and look off to the side; she may ignore the child until he turns his attention and interest elsewhere, and then say "Come here, dear." These are only a few suggestive examples of kinds of incongruence; there are many others, some playing a quite positive part in the richness and subtlety of human communication. On this view, then, one first tries to relate confused or disturbed behavior to such conflicts among influential messages, and we note in practice that the husband whose wife says "I want you to dominate me" is likely to act confused, angry, or both, especially if he fails to notice and comment on this incongruency. In contrast, orthodox psychology and psychiatry has seemed based on an implicit model only slightly caricatured as "Messages are plain and simple, but some people receiving them are internally confused or divided." The difference in viewpoints is simple, and not absolute, but it makes a profound difference in one's whole approach to both individual and family problems. Given the general view point just outlined, it is clear that as an arena for research the family therapy situation fits its basic requirements as to the relevant data and conditions for collecting them. It is a situation in which individual communicative behavior, communicative reactions to this, and ultimately overall patterns of communication, influence, and interaction, are directly observable by the therapist. He does not have to rely only on reports of past events. Another way of viewing this is to point out that the therapist directly receives data on actual current family interaction that immediately includes more than one level of communicationthat is, as a simple but basic example, he sees how the family members interact while also hearing what they say about their interaction. In particular, this situation, continuing over a period of time, provides opportunities for repeated observation, and for the detection of repetitive patterns of interaction typical of certain families or kinds of families, despite differences in context or detail, in a way difficult to match. This special observation situation is particularly suited for investigations in three areas, which include some problems hard to investigate otherwise. First, in the course of family therapy it reasonably often occurs that the identified patient will start to behave in a noticeably more symptomatic way, briefly or for some while, than is usual in the sessions. Conceivably this might be avoidable if our therapeutic knowledge and skill were great enough, but as things stand, such behavioral changes do occur and provide an opportunity for direct investigation of any interpersonal causes and circumstances of schizophrenic behavior, especially if the preceding interaction is available on tape or film records. For example, in one family session a schizophrenic son, supported by the therapist, managed to say openly that he had meant a message to be critical of his mother. The father meanwhile kept denying that this was so, and the mother told the patient "I'll take all the hurt in the world if it will help you"; the patient then became confused, tearful, concretistic, paranoid, and claimed amnesia. Such incidents provide opportunities to investigate the contribution of all family members, and their system of interaction, to the occurrence of schizophrenic behavior, rather than focusing on the patient's "sickness" alone, or focusing on some one family member in a "Who is to blame?" fashion. It is obvious that there are many specific problems connected with such inquiriesfor example, how far back in the session record should the investigation start? And there are larger questions about the extent to which such occurrence of symptomatic behavior is similar or comparable to the schizophrenic break situation, or relevant for any ultimate etiology of schizophrenia. But any other inquiries into the etiology of schizophrenia or the psychotic break also involve similarly serious research problems. In particular, the lack of immediate data because the relevant events occurred in the past and were not directly observed by the investigator makes it seem appropriate first to investigate any observable onsets of symptomatic behavior as such, and to defer until later, when these are better understood, questions of possible wider or deeper relevance. One can not only examine such symptomatic outbreaks as responses to preceding family interaction, but also, viewing such behavior as itself communicative, one can observe its consequences for succeeding family behavior and interaction.

_____________________________________________________________________________________________________________

Thus, shortly after the criticism-parental reactions-symptomatic behavior sequence described above occurred, it was observable that the family had returned to its usual state, with mother mainly in charge, the father weakly platitudinous, and the son withdrawn. Combining these two approaches to symptomatic behavior thus leads towards the development of a circular view of family behavior. This begins the second type of inquiry for which the family therapy situation is adapted, the investigation of family homeostasis; that is, how the patterns of interaction in such familiesincluding the patient's schizophrenic behavior, but by no means limited to thisoperate to maintain the existing family system. This increasingly appears as the most important question in family therapy, or even for schizophrenia or psychopathology quite generally. Unless one is to fall back on some idea that people by fundamental nature are oriented toward disease, so that pathology is inherently self-sustaining, rather than "normality is normal," the central issue is not the question of the original root causes of schizophrenia. It is not even the question of what sort of present family interaction leads to schizophrenic symptoms in one member. Instead, the central issue concerns the nature of the maintenance of family systems involving a case of schizophrenia. What in the family interaction makes for the fundamental stability and persistence of these family systems that is so striking in the face of the general dissatisfaction and unhappiness of the members, their stated desires for change, and often the best efforts of a therapist? Here is something worth investigating, which in the family therapy situation can be inquired into directly and repeatedlyeven if still not readily understood. The third area for investigation obviously consists in the opportunity to observe directly what immediate or longer-term results in family behavior are produced by any given moves on the part of the therapist. Many important specific problems relate closely to this: how are moves of therapists countered by family behavior serving to maintain homeostasis? What upsets may occur in other family members if one improves? This is a question that also raises very serious technical and ethical problems about the uncontrolled effects, on related individuals not seen, of changes made by individual treatment. What sort of therapeutic moves are really effective? Here again, important practical difficulties and basic questions about interpretation and significance of any observed results are involved in such investigation: perhaps a significant change is initiated by some therapeutic move, but one cannot see the change clearly until later, so the connection is not noticed or uncertain; or if an immediate reaction is observable, is this of any real or permanent significance? Such questions cannot be answered in general now. But similar onesand others more difficultalso occur in any investigation into the nature or effects of therapy. In the family treatment situation, as regards therapeutic moves just as for the etiology and the maintenance of schizophrenia, at least we do have an arena where some pertinent direct observations can be made and followed up for some distance before it is necessary to plunge deeply into the realm of unobservables and into scientific-philosophical questions too broad to be resolved at present.