The term ‘family therapy’ covers a variety of approaches. At one extreme it is a method drawn from one or more of a range of theoretically based schools that seeks to help an individual patient who presents with a clinical syndrome. At the other extreme family therapy is a way of thinking about psychotherapy in general; the intervention may involve the individual alone, the nuclear family, or an extended network, but the focus is the relationships between people. According to this view psychopathology reflects recurring, problematic interactional patterns among family members and between the family, and possibly, other social institutions, and may include doctors and helping agencies. Midway between these two positions is one that views the family as acting potentially as a resource or as a liability for an identified patient; different interventions are thus needed to enhance the positive effects of family relationships as compared with those that seek to minimize or negate their noxious effects. As we will elaborate in this chapter such a range of interventions makes it tricky to define and research family therapy.
The family has long been recognized as a fundamental unit of social organization in the lives of human beings. Regardless of the specific pattern of family life, the foundational narratives, myths, legends, and folklore of all cultures emphasize the power of family relations to mould the character of the individual and serve as an exemplar of the moral and political order of society.
In the past 150 years new academic disciplines, among them anthropology, sociology, and social history, have devoted much attention to the diverse forms of family structure and function found in different cultures at various historical periods. Constrained perhaps by Western medicine's focus on the individual patient, psychiatry has been tardy in formulating a view of the family other than as a source of genetically transmitted diseases, hence the emphasis on inquiring about the prevalence of mental illness among relatives.
Scattered through Freud's writings are interesting comments about marital and family relationships and their possible roles in both individual normal development and psychopathology (Sander, 1978). Freud's description of unconscious processes such as introjection, projection, and identification explained how an individual's experiences could be transmitted across the generations in a family. Freud's successors elaborated on his formulations, e.g., in 1921 J. C. Flugel published the first detailed psychoanalytic account of family relationships (Flugel, 1921).
Strongly influenced by the work in the UK of Anna Freud, Melanie Klein, and Donald Winnicott, the child guidance movement devised a model of one therapist working with the disturbed child and another with the parents, most often the mother on her own. The two clinicians collaborated in order to recognize how the mother's anxieties distorted her perception and handling of her child, which compounded the child's own developmental anxieties. This work, however, was conducted by psychiatric social workers and only a minority of psychiatrists.
Things took a different turn in the US. There, Ackerman (1958), who coined the term ‘family therapy’ in the 1950s, had introduced the idea of working with the nuclear family of a disturbed child using psychodynamic methods. An interest in working with the family, including two or more generations, arose concurrently in several psychiatric centers. Most of the pioneers of so-called ‘transgenerational family therapy’ were analysts who used many of the concepts of object relations theory that they recast into their own conceptual language.
Thus, Murray Bowen (1971) in his work with psychotic children found that their capacity to differentiate themselves emotionally from their families (especially from mother) while still retaining a sense of age-appropriate emotional belonging was impaired by the legacy of unresolved losses, trauma, and other upheavals in the lives of parental and grandparental generations. Bowen also devised the genogram, a schematic depiction of family structure, with a particular notation for significant family events; this forms a standard part of contemporary family assessment in clinical practice (see pp. 60–2 section on assessment).
Boszormenyi-Nagy and Spark (1984) in their contextual therapy also addressed this transgenerational theme by describing how family relationships between generations and between adults in a marriage were organized around a ledger of entitlements and obligations; this conferred on each person a sense of justice about their position. This, in turn, reflected the experience in childhood of neglect or sacrifices made on a person's behalf for which redress was sought in adult life.
Bowen had also introduced the principles of systems theory into his work with families. A system may be defined as a set of inter-related elements that function as a unity in a particular environment. General systems theory (GST) was propounded in the 1940s by the German biologist, Ludwig von Bertalanffy (1968); he outlined the principles by which any system (inanimate, animate, or ideational) can be described. Key concepts of GST are hierarchy, the emergence of new properties in the transition from one level of organization to another, and formulations derived from thermodynamics, which describe the exchange of energy between the system and its environment. A family may be considered a partially open system that interacts with its biological and sociocultural environments.
Working with delinquent youth in New York, Salvador Minuchin and his colleagues recognized the relevance of systems thinking to their interventions. The youngsters often came from economically impoverished, emotionally deprived families, headed by a demoralized single parent (most often mother) who alternated between excessive discipline and helpless delegation of family responsibilities to a child or to her own disapproving parent. Such families were understandably mistrustful of words and beyond the reach of conventional ‘talking’ therapies. Minuchin's emergent structural family therapy came to deploy a series of action-oriented techniques and powerful verbal metaphors that enable the therapist to ‘join’ the family, and to reestablish an appropriate hierarchy and generational boundaries between the various subsystems (marital, parent/child, siblings).
Later, treating so-called ‘psychosomatic families’ where the presenting problem was a child suffering from anorexia nervosa, unstable diabetes, or asthma, Minuchin's team noted that unlike the chaotic, leaderless disengaged ‘delinquent families’ these, while middle-class, intact, and articulate, often were enmeshed. Their members avoided overt expressions of dissent or challenge to ostensible family unity. Typically, marital conflict was detoured through the symptomatic child, resulting in maladaptive coalitions between parent and child, or between grandparent and child, the inclusion of third parties (e.g., a helping agency) into family life. All this led to a loss of appropriate boundaries. Because words were used to avoid change in these well-educated families, Minuchin and Fishman (1981) again looked to actional strategies to challenge their unspoken fears of conflict and change.
Jay Haley's (1976) Strategic Therapy combined aspects of Minuchin's model with ideas of the psychotherapist, Milton Erickson; his hypnotherapy techniques had skillfully exploited the notion that a covert message lurks behind overt communication that defines the power relationship between people. This applies to a patient's ties with his family and their professional helpers.
Another important series of theoretical developments took place in Palo Alto, California, where a group of clinicians gathered around the anthropologist Gregory Bateson (1972) in the 1950s. In his field work, Bateson had noted two relational patterns:
Symmetrical, in which each participant's behavior induces the other to do more of what they were already doing as equals. Power struggles in a marriage or between parents and an adolescent, arguments over compliance with medication or family conflict preceding psychotic relapse or an alcoholic binge exemplify such symmetrical escalation; and
Complementary, in which participants arrange themselves such that, for example, one is dominant and the other subordinate. The doctor–patient relationship or the parent–child relationship often is of this type, while a pattern of rigid complementarity characterizes the marriages of many patients suffering from chronic anxiety states, agoraphobia, and chronic dysthymia.
The ability to switch from complementary to symmetrical patterns and vice versa, and to alternate between dominant/subordinate and co-equal positions at different times and on various matters are skills that the Bateson approach teaches. It views psychopathology as the product of people getting stuck in once relevant but now dysfunctional modes of relating and problem solving.
Bateson's group also noted that implicit in communication were tacit, nonverbal ‘metacommunications’ that defined the relationship between the participants. Contradiction or incongruence between these two levels when each message carried great persuasive, moral, or coercive force to the recipient formed part of what they labeled a ‘double-bind’. When combined with a tertiary level injunction that forbade escape from the field of communication, this double-bind was proposed as a possible basis for schizophrenic thinking (Bateson et al., 1956, 1962).
All these aforementioned system-oriented approaches assume that the family is a system observed by the therapist. However, therapists are not value neutral. As described, in some models they take an active role in advocating and orchestrating specific changes in accordance with a preconceived model of family functioning. Yet these models ignore therapists’ biases as well as the relevance of their relationships with families. This probably reflected the determination of certain American family therapists to distance themselves from psychoanalytic theory, and also led them to neglect the family's history, how it altered during the life cycle, and the relevance of past notable events.
In response to these criticisms there was a move away from the here-and-now, problem-focused approach that had characterized most behavioral and communicational views of psychopathology. The Milan school (Selvini-Palazzoli et al., 1980; see pp. 62–4 in section on course of therapy), whose founders were all psychoanalysts, developed circular questioning, a radically new method of interviewing families. Furthermore, observers behind a one-way screen formulated hypotheses about the family-plus-therapist system and its relevance to the clinical process.
A Norwegian group (Andersen, 1991) developed the ‘reflecting team dialogue’ in which, following a therapy session, the family could observe the therapists discussing their problem, possible causes, and unresolved factors, which might have led them to seek certain solutions they had persevered with despite obvious lack of success, while neglecting alternative solutions.
Family therapists also began to consider that families might be constrained from experimenting with new solutions to difficulties because of the way they had interpreted their past experiences or internalized the explanatory narratives of their family, the expert's, or society at large.
This led to a shift from regarding the family as a social system defined by its organization (i.e., roles and structures) to a linguistic system. According to this view the narrative a family relates about their lives is a linguistic construction that organizes past experience and relationships, and their significance, in particular ways. Other narratives are excluded from consideration. When a family with an ill member talks to health professionals, conversations are inevitably about pathology (a problem-saturated description). The participants ignore times when the problem was absent or minimal, or when they successfully confined it to manageable proportions. A different story might be told if they were to examine the context and relationships that might have led, or could still lead, to better outcomes.
A number of narrative, social constructionist, or solution-focused approaches (the terms are essentially interchangeable) make use of these concepts (De Shazer, 1985; Anderson and Goolishian, 1988; White and Epston, 1990). Philosophically, they align themselves with postmodernism, a movement that challenges the idea that there is a basic truth or grand explanatory theory known only by experts.
Many criticisms of the above systems approaches to family therapy have been leveled. These include:
disregard of the subjective and intersubjective experiences of family members;
neglect of the family's history;
denial of unconscious motives that influence individuals in a relationship;
although people are reciprocally connected in a family system the power they exert on one another is not equal (this is highlighted particularly in the problem of violence against women and in various types of child abuse);
inequality and other forms of injustice based on societal attitudes towards differences in gender, ethnicity, class, and the like, are uncritically accepted as ‘givens’;
minimizing the role of therapeutic relationship, including attitudes family members develop toward the therapist and her feelings towards each of them and to the family as a whole.
This critique has led to an interest in integrating systems-oriented and psychoanalytic concepts, particularly those derived from object relations theory. Attempts at a general level are those of Flaskas and Perlesz (1996), Braverman (1995), and Cooklin (1979), and the feminist perspective (Luepnitz, 2002); specific disorders such as schizophrenia (Ciompi, 1988), and anorexia nervosa (Dare, 1997) have also been targeted. One variant of integration is John Byng-Hall's (1995) masterful synthesis of attachment theory, systems thinking, and a narrative approach.
A further criticism of systems-oriented therapies is their minimizing the impact of material reality such as physical handicap or biological forces in the cause of mental illness, and sociopolitical phenomena such as unemployment, racism, and poverty. These are obviously not merely the result of social constructions or linguistic games. The distress they inflict are real in the extreme.
The ‘psychoeducational’ approach, ‘family crisis intervention’, and ‘family-sensitive practice’ have evolved in the context of the burden that schizophrenia places on the family and the potential for responses of members to influence the course of the illness. This has paved the way for a series of interventions:
educating the family about what is known regarding the nature, causes, course, and treatment of schizophrenia;
providing the family with opportunities to discuss their difficulties in caring for the patient and to devise appropriate strategies;
clarifying conflict in the family not only about the illness but also about other issues;
regularly evaluating the impact of the illness on the family as individual members and collectively;
helping to resolve other conflicts not specifically related to the illness, but which may be aggravated by the demands of caring for a chronically ill person.
This type of work may be carried out with several families meeting together. Whatever the case, promising results have been achieved in reducing relapses and frequency of hospital admission (McFarlane et al., 1995).
The limitations of psychoeducational programs for psychiatric disorders and vulnerability to relapse after a psychotic episode have been shown to reflect the emotional climate of the family. It is noteworthy that these potentially disruptive patterns of interaction often are not detectable by the clinician who interviews the patient alone rather than observing him in the context of a family interview (Thompson et al., 2000). Furthermore, the difficulties therapists encounter working with such families vary at different phases of treatment.
While the conventional view claims critical comments are significantly correlated with relapse, it also appears that, at least in some patients with bipolar disorder, a comparative excess of genuinely positive and supportive comments by family members may also be associated with relapse.
Family crisis intervention, initially devised for families with a schizophrenic relative, but since applied to other clinical states, operates on the premise that deterioration in mental state or a request by the family to hospitalize a member may well reflect a change in a previously stable pattern of family interaction. Convening an urgent meeting with patient, spouse, and other key family members, even in a hospital emergency center, is associated with a reduced rate of admission.
While integrating some concepts from systems, postmodern, and psychoeducational approaches, cognitive-behavioral therapy emphasizes the importance of identifying and directly modifying dysfunctional ideas and behavioral patterns of family members. When families are in distress they frequently perceive each other's reactions (behavioral and emotional) in a distorted way, which may in turn elicit counterproductive reactions. Persistent deleterious cycles are set up in which family members continually misperceive and/or misinterpret one another and react accordingly. Therapy aims to help family members correct their selective negative biases, negative attributions of one another, negative predictions, dysfunctional assumptions, and unrealistic standards. Cognitive-behavioral therapists work to help family members increase positive behavioral changes, engage in pleasurable activities, and improve communication and problem-solving skills (Epstein and Schlesinger, 2003).
Notwithstanding the application of these various approaches in adult psychiatry for at least three decades, indications remain ill defined compared with other forms of psychotherapy. Moreover, controversy has dogged the subject. This is not altogether surprising. Pioneering family therapists acted perhaps with a touch of hubris when claiming that their innovative approaches were suited to most clinical conditions. Ambitiousness rode high. With the passage of time, a more balanced view evolved that encompasses the notion that a systemic context is advantageous in assessing and treating any psychiatric problem, although it is not axiomatic that family therapy will be the treatment of choice (or even indicated).
We should bear in mind that family therapy is a mode of psychological treatment, not a unitary approach with one central purpose. One only has to note the diversity of theoretical models we discussed earlier, with their corresponding variegated techniques. Attempts to link indications to specific models have proved ill advised and contributed little to the field overall.
It has also become clear that conventional diagnoses as listed in DSM-IV or ICD-10 do not serve well as a source to map out indications for family work. DSM-IV has a minimal section, the so-called V diagnoses, covering ‘relational problems’, which are not elaborated upon at all (American Psychiatric Association, 1994). All we are told is that the problem in relating can involve a couple, a parent–child dyad, siblings, or ‘not otherwise specified’. ICD-10 ignores the relational area entirely.
In mapping out indications, we need to avoid the complicating factor of blurring assessment and therapy. A patient's family may be recruited in order to gain more knowledge about his diagnosis and subsequent treatment. This does not necessarily lead to family therapy. Indeed, it may point to marital or to long-term supportive therapy. Thus, we need to distinguish between an assessment family interview and family therapy per se.
Finally, a typology of family psychopathology that might allow the diagnostician to differentiate one pattern of dysfunction from another and identify appropriate interventions accordingly is elusive. Here, empirical evidence is inconclusive and clinical consensus lacking. An inherent hurdle is determining which dimensions of family functioning are central to creating a family typology (Bloch et al., 1994). Communication, adaptability, boundaries between members and subgroups, and conflict are a few of the contenders proffered (we offer our own classification below).
It does not help that there are no clear associations between conventional psychiatric diagnoses and family type. Efforts to establish links, such as an anorexia nervosa family (Minuchin et al., 1978) or a psychosomatic family (Clarkin et al., 1979) have not been fruitful. Similarly, work in the area of the family and schizophrenia (e.g., Bateson et al., 1956 and Bowen, 1978) have not yielded durable results. Instead, research supports the view that no particular type of family dysfunction differentiates between specific types of mental illness (as designated on Axis 1 of DSM IV). Rather, having a mentally-ill family member acts as a general stressor on the family that may lead to impaired functioning across a range of family-related activities (Epstein and Schlesinger, 2003). Consistent with the systemic view, such illness-induced family dysfunction may aggravate the course of the illness or complicate its management.
What follows is our attempt to distill past clinical and theoretical contributions, particularly the work of Walrond-Skinner (1978) and Clarkin et al. (1979). There are many ways to cut the pie; resultant categories are not mutually exclusive entirely given the considerable overlap in clinical practice; and a particular family may require family therapy based on more than one indication. We also must stress that family dysfunction is obvious in certain clinical situations but more covert in others, and often concealed by a specific member's clinical presentation. Six categories emerge:
The clinical problem manifests in explicitly family terms; the therapist readily notes family dysfunction. For example, a marital conflict dominates, with repercussions for the rest of the family or tension between parents and an adolescent child dislocates family life with everyone ensnared in the conflict. In these sorts of situations, the family is the target of intervention by dint of its obvious dysfunctional pattern and family therapy the treatment of choice.
The family, nuclear or extended, has experienced a life event, stressful or disruptive in type, which has led to dysfunction or is on the verge of doing so. These events are either predictable or accidental and include, for instance, accidental or suicidal death, financial embarrassment, serious physical illness, the unexpected departure of a child from the home, and so forth. In all these circumstances, any family equilibrium that previously prevailed has been disturbed; the ensuing state becomes associated with family dysfunction and/or the development of symptoms in one or more members. In some instances, family efforts to rectify the situation inadvertently aggravate it.
Continuing, demanding circumstances in a family are of such a magnitude as to lead to maladaptive adjustment. The family's resources may be stretched to the hilt, external sources of support may be scanty. Enduring physical illness, persistent or recurrent psychiatric illness, and the presence in the family of a frail elderly member are typical examples.
An identified patient may become symptomatic in the context of a poorly functioning family. Symptoms are an expression of that dysfunction. Depression in a mother or an eating problem in a daughter or alcohol misuse in a father, on family assessment, is adjudged to reflect underlying family difficulties.
A family member is diagnosed with a specific condition such as schizophrenia, agoraphobia, obsessive-compulsive disorder, or depression; the complicating factors are the adverse reverberations in the family stemming from that diagnosis. For example, the schizophrenic son taxes his parental caregivers in ways that exceed their ‘problem-solving’ capacity; an agoraphobic woman insists on the constant company of her husband in activities of daily living; a recurrently depressed mother comes to rely on the support of her eldest daughter. In these circumstances, family members begin to respond maladaptively in relation to the diagnosed relative and this paves the way for a deterioration of his condition, manifest as chronicity or a relapsing course.
Thoroughly disorganized families, buffeted by a myriad of problems, are viewed as the principal target of help, even though one member, for instance, abuses drugs, another is prone to violence, and a third exhibits antisocial behavior. Regarding the family as the core dysfunctional unit is the relevant rationale rather than foci on each member's problems individually.
We reiterate that family therapy may be a treatment of choice in all these categories, but not necessarily the only one. Thus, in helping a disturbed family struggling to deal with a schizophrenic son, supportive therapy and medication for the patient is likely to be as important as any family treatment. Similarly, an indication for family therapy does not negate the possibility of another psychological approach being used for one or more members. For instance, an 18-year-old adolescent striving to separate and individuate may benefit from individual therapy following family treatment (or in parallel with it) while the parents may require a separate program to focus on their marital relationship.
These are more straightforward than indications; they are self-evident and therefore mentioned briefly.
The family is unavailable because of geographical dispersion or death.
There is no shared motivation for change. One or more members wish to participate but their chance of benefiting from a family approach are likely to be less than if committing themselves to individual therapy. (We need to distinguish here between poor motivation and ambivalence; in the latter, the assessor teases out factors that underlie it and may encourage the family's engagement.)
The level of family disturbance is so severe or long-standing or both that a family approach seems futile. For example, a family that has fought bitterly for years is unlikely to engage in the constructive purpose of exploring their patterns of functioning.
Family equilibrium is so precarious that the inevitable turbulence (Goldenberg and Goldenberg, 1996) arising from family therapy is likely to lead to decompensation of one or more members, e.g., a sexually abused adult may do better in individual therapy than by confronting the abusing relative.
A member with a psychiatric condition is too incapacitated to withstand the demands of family therapy. The person in the midst of a psychotic episode or someone overwhelmed by severe melancholia is too affected by the illness to engage in family work.
An identified patient acknowledges family factors in the evolution of his problem but seeks the privacy of individual therapy to explore it, at least initially. For example, a university student struggling to achieve a coherent sense of identity may benefit more from her own pursuit of self-understanding. Such an approach does not negate an attempt to understand the contribution of family factors to the problem.
Family assessment, an extension of conventional individual psychiatric assessment, adds a broader context to the final formulation. Built up over a series of interviews, the range and pace of the inquiry depends on the features of the case. Its four phases are: history from the patient, a provisional formulation concerning the relevance of family issues, an interview with one or more members, and a revised formulation.
In some cases, it is clear from the outset that the problem resides in the family as a group (see indications); in this context, the phases below are obviously superfluous.
The most effective way to obtain a family history is by constructing a family tree. This provides not only representation of structure but additional information is obtained about important events and a range of family features. Scrutiny of the tree also becomes a source of noteworthy issues warranting exploration and, eventually, of clinical hypotheses.
Personal details are recorded for each member such as age, dates of birth and death, occupation, education, and illness, as are critical events (e.g., migration, crucial relational changes, major losses, and achievements), and the quality of relationships.
An erudite discussion of the family tree—its construction, interpretation, and clinical uses—is presented by McGoldrick and Gerson (1985).
Useful guidelines are to work from the presenting problem to the broader context, from the current situation to its historical origins and evolution, from ‘facts’ to inferences, and from nonthreatening to more sensitive themes.
Commonly, questions are preceded by a statement such as: ‘In order to better understand your problems I need to know something of your background and your current situation’. This is enriched by questions that refer to interactional patterns: ‘Who knows about the problem? How does each of them see it? Has anyone else in the family had similar problems? Who have you found most helpful, and least helpful thus far? What do they think needs to be done’. Attitudes of members can thus be explored and light shed on the clinical picture.
Questions aimed at understanding the current context include: ‘What has been happening recently in the family? Have there been any changes (for example, births, deaths, illness, losses). Has your relationship with other members changed? Have relationships within the family altered?’
At this point a broader inquiry flows logically—in terms of members to be considered, and in the time span of the family's history. Information about parents’ siblings and their families, grandparents, and a spouse's family may be pertinent. Other significant figures, which may include caregivers and professionals, should not be forgotten.
Apart from information about the extended family's structure, questions about the family's response to major events can be posed: for example, ‘How did the family react when grandmother died? Who took it the hardest? How did migration affect your parents?’
Relationships should be explored at all levels covering those between patient and other members and between those members themselves. Conflicted ties are illuminating. Understanding the ‘roles’ adopted by members is also useful, for example, ‘Who tends to take care of others? Who needs most care? Who tends to be the most sensitive to what is going on in the family?’
Asking direct questions about members is informative but a superior strategy is to seek the patient's views about their beliefs and feelings and to look for differences between members; for example: ‘What worries your mother most about your problem? What worries your father most?’ Several lines of inquiry may reveal differences:
Pursuing sequential interactions: ‘What does your father do when you say your depressions are dreadful? How does your mother respond when your father advises you to pull up your socks? How do you react when she contradicts your father?’
‘Ranking’ responses: ‘Everyone is worried that you may harm yourself. Who worries most? Who is most likely to do something when you talk about suicide?’
Looking for changes in relating since the problem: ‘Does your husband spend more or less time with you since your difficulties began? Has he become closer or more distant from your daughter?’
Hypothetical questions dealing with imagined situations: ‘How do you think your relationship with your wife will change if you don't improve? Who would be most likely to notice that you were getting better?’
Triadic questions help to gain information about relationships that go beyond pairs; for example: ‘How do you see your relationship with your mother? How does your father see that relationship? How would your mother react to what you have told me if she were here today?’
Two questions about the family arise following the above interview: (1) How does the family typically function, and (2) Do any family features pertain to the patient's problems?
A schema to organize ideas about family functioning builds from simple to complex observations: structure, changes, relationships, interaction, and the way in which the family works as a whole.
The family tree will reveal the many family structures possible—single parented, divorced, blended, remarried, sibships with large age discrepancies, adoptees; unusual configurations invite conjecture about inherent difficulties.
Data will be obtained about significant family changes and events. Timing of predictable transitions such as births, departures from home, marriages, and deaths is pertinent. Have external events coincided with these transitions? (times at which the family may be more vulnerable). How have demands placed on the family by such changes been met?
Relationships refer to how members interact with one another. This is typically in terms of degree of closeness and emotional quality (e.g., warm, tense, rivalrous, hostile). Major conflicts may be noted as may overly intense relationships.
Particular interactional patterns may become apparent. These go beyond pairs. Triadic relationships are more revealing about how a family functions. A third person is often integral to defining the relationship between another pair. A conflict for instance may be rerouted through the third person, preventing any direct resolution. A child may act in coalition with one parent against the other or with a grandparent against a parent.
At a higher level of abstraction, the clinician notes how the family works as a whole. Particular patterns, possibly a series of triads, may emerge, which may have recurred across generations. For example, mothers and eldest sons have fused relationships, with fathers excluded, while daughters and mothers-in-law are in conflict.
Idiosyncratic shared beliefs may be discerned that explain much of the way the family does things. ‘Rules’ governing members’ behavior towards one another or to the outside world may flow from these beliefs. For example, a family may hold that ‘you can only trust your own family; the outside world is always hostile,’ they may therefore avoid conflict at any cost, and prohibit seeking external support.
Evidence of family difficulties may be found at each of these five levels. If they are, the question arises whether these relate or not to the identified patient's problems.
Links between family functioning and the patient's problems take various forms, but the following categories cover most clinical situations. More than one will often apply: the family as reactive, the family as a resource, and the family in problem maintenance.
The patient's illness, or its exacerbation, may have occurred at a time of family upheaval. The precipitant for the upheaval may have been the illness itself. An escalating combination of the two may pertain. The illness may have occurred in the face of family stress; it pressurizes the family all the more, and this in turn exacerbates the illness.
The family may be well placed to assist in treatment. This may be as straightforward as supervising medication, ensuring clinic attendance, and detecting early signs of relapse or providing a home environment that promotes recovery and its maintenance. The family may also call on friends and agencies, professional or voluntary, to offer support.
Interactions revolving around the patient's illness may act to maintain it in one of three chief ways. First, the illness itself becomes a way of ‘solving’ a family problem, the best that can be achieved. For example, anorexia nervosa in a teenager due to attend a distant university may lead to her abandoning this plan as she feels unable to care for herself. Were she to leave, parental conflict would become more exposed and her mother, with whom the patient is in coalition against her father, would find herself unsupported. The illness therefore keeps the patient at home and enmeshed in the parental relationship, and also provides a focus for shared concerns and an ostensible sense of unity.
Secondly, maintenance of the illness does not solve a family problem but may have done so in the past. An interactional pattern persists even though it lacks utility. In the previous example, the father's mother died 9 months later. His wife subsequently expressed feelings of closeness, feelings not experienced by him for years; their relationship gradually improved. Both parents, however, continued to treat their daughter as incapable of achieving autonomy, reinforcing her own uncertainty about coping independently if she were to recover.
Thirdly, persistence of illness reflects a perception by the family of themselves and their problems, to which they are bound by the persuasive power of the narrative that they have shaped for themselves; the narrative may have stemmed from the helping professionals’ explanatory schemas.
The clinician will by now have made an initial assessment of the patient's problems and of the family context. An interview with one or more informants, usually family members, is the next step. Several purposes are served: to corroborate the story, to fill in gaps, to determine influences impinging on the patient, and to recruit others to help. A family meeting is most effective in order to accomplish these goals.
Problems may arise in trying to implement the session. The patient may resist family members being interviewed for all sorts of reasons, e.g., symptoms have been kept secret, the patient regards it as unfair to burden others, he is ashamed of seeing a psychiatrist, he is fearful the family will be blamed or he is suspicious of them. These concerns need ventilating, particularly if the family is pivotal and treatment will be enhanced by their involvement. The patient will agree in most cases. Where the health or safety of a patient or others is threatened, refusal may be overridden on ethical grounds. Otherwise, refusal must be respected. The question of a family session can be raised later after a more trusting relationship has been cemented.
Who should be seen depends on the purpose of the interview; generally, all those living in the household and likely to be affected by the identified patient's illness should participate. Of course, some family members may be living elsewhere but are very much involved. The more family factors pertain, the more desirable the attendance by all members. The patient's views should be sought as he will provide insight into who he considers are key people.
The clinician will have garnered substantial information by the time the family is seen. He should reflect on any biases that may have crept into his thinking about the family, and how the situation might influence them to draw him into alliances. This may well happen when conflict prevails. The clinician strives to act neutrally, his sole interest that of ‘helping in the situation.’ A nonjudgmental stance is paramount.
Introductions are made, names and preferred modes of address clarified. The clinician then explains the meeting's purpose. The details may well influence future participation. Everyone is then invited to share their views about the nature and effects of problems they have encountered.
The clinician may have an idea about how the identified patient's problems relate to family function and can test it out by probing questions and observing of interactions. This idea is typically kept to himself as it is unhelpful to present a hypothesis prematurely. Instead, he seeks details about everyday events and infers patterns thereafter. For example, rather than focusing on ‘closeness’, he enquires about time spent together by the family, whether intimate experiences are shared, who helps with family tasks, and so on.
Triadic relationships can be scrutinized both through questioning (What does A do when B says this to C?) and observation (What does A do when B and C reveal tensions?). The scope for circular questioning is enhanced if several members participate. A third person may be asked to comment on what two others convey to each other when a particular event occurs. This approach of not asking predictable questions to which the family may by now have stereotypical responses often challenges them to think about their relationships in a fresh way.
Information is elicited that elaborates the family tree. Observations are made concerning family structure and functioning, e.g., who makes decisions, who controls others and in what areas, the quality of specific relationships, conflict, alliances, how clearly people communicate and how they approach problems. The discussion then extends to all spheres of family life: beliefs, traditions, rules, and values.
Throughout the interview the clinician affirms the experiences of all members by not only attending to concerns, but also acknowledging strengths and their efforts to tackle their difficulties.
The interview ends with a summary of what has emerged. The clinician may ask to continue the assessment on a second occasion or may recommend family therapy at this point. If the latter, he then explains its aim and rationale. Arrangements are set for a follow-up session, purportedly the launch of the family therapy per se, but in essence a continuation of the ‘work’ in progress.
As more information becomes available at each of the aforementioned levels, the initial formulation can be revised as necessary. The five observational levels of structure, transitions, relationships, patterns of interaction, and global family functioning are reexamined in terms of the family as reactive, resourceful, or problem maintaining. Appropriate interventions can be planned, at least for a follow-up session. We are now ready to turn to the course of typical family therapy.
With the phase of assessment concluded and a family approach agreed upon, therapy begins. We should recall, however, that a family may be referred as a group from the outset on the premise that the problem is inherently a family-based one. In this case, the initial stage incorporates assessment and this is made explicit.
Given the plethora of ‘schools’ of family therapy, as described earlier, it would be laborious to map out the course of treatment associated with each of them. Instead, we will focus on the approach pioneered by the Milan group (Selvini-Palazzoli et al., 1980) but we should stress that it has undergone much elaboration and refinement over 25 years. Our account tends to highlight the original features. First, we need to comment briefly on the roles the therapist may assume.
Beels and Ferber (1969) who were among the first observers to consider various roles for family therapists, divided them into ‘conductors’ and ‘reactors’; the differentiation remains useful as it transcends schools. The therapist as conductor is represented in the work of practitioners such as Satir, Bowen, and Minuchin. Virginia Satir (1967) is a good illustration. With her emphasis on communication, she espoused the notion that the family therapist is a teacher who shares her expertise in optimal communication by setting goals and the direction of treatment. In her case, she guided the family to adopt a new form of language in order to resolve problems in communication that she saw as the root of their troubles. Additionally, the therapist instills confidence, promotes hope for change and makes them feel comfortable in the process.
In Satir and fellow conductors, the therapist is an explicit authority, who intervenes actively in implementing change.
The therapist as reactor plays a different role by resonating with, and responding to, what the family manifests. Therapists in the psychoanalytic tradition belong to this group as do what Beels and Ferber label system purists. Typically, the therapist shares observations about patterns of relating that emerge during the sessions. We will illustrate this aspect when describing the Milan approach (Selvini-Palazzoli et al., 1980). We have selected it arbitrarily as we cannot possibly give accounts of every school.
With assessment complete, the therapist (sometimes a pair) meets with the family. With her preparatory knowledge, she shapes a hypothesis about the nature of the family's dysfunction. As a reactor, he has the opportunity, on observing patterns in vivo, to confirm her ideas. Such patterns usually emerge from the start making the therapist's job correspondingly easier. Apart from hypothesis testing, another task in this session is to engage the family fully so that they will be motivated to reattend. We could interpolate a dictum here: a primary aim of the first session is to facilitate a second session. A key element in encouraging engagement is for the therapist to promote a sense of curiosity in members so that they raise questions about themselves and the family as a group (Cecchin, 1987).
The chief strategy used is circular questioning, which we touched on in the assessment section (Tomm, 1987). Although it is easy to imagine doing, it is tricky to do well. The main purpose is to address the family's issues indirectly; this avoids pressurizing particular members and perhaps provoking their resistance. For example, the therapist asks questions of an adolescent about how his parents get on with each other; or a mother about how her husband relates to the eldest son; or a grandmother about which grandchild is closest to the parents; and so forth. This mode of inquiry generates illuminating data about individual members and about the family as a group. In this phase, it helps to clarify the hypothesis, to engage participants and affords the therapist greater facility to remain neutral and thus avoid forging alliances with an individual or subgroup. Because the system and not the identified patient is the target of change, the therapist is wary of showing bias. (This does not preclude transient alliances adopted for strategic purposes; these, however, need to be limited in time and distributed throughout the system.)
The therapist and family ‘work’ together for an hour or so on the basis of promoting curiosity, circular questioning, and neutrality. A number of options then follow. If the therapist is part of a team, her colleagues will have been observing the proceedings through a one-way screen. The family's consent, of course, will have been obtained previously. During a break the team—observers and therapist(s)—systematically pool impressions (Selvini, 1991). This is invariably a rich exchange as team members often note something others may have missed. As a result of these deliberations, a consensus about family functioning evolves. Conclusions are drawn and converted into ‘messages’. The therapist returns to the family briefly to convey them. This is akin to the Delphic Oracle. The actual messages and their oracular quality comprise a potent intervention but not necessarily more cogent than interventions in the form of circular questions made earlier. Indeed, the advent of the narrative school has brought with it a de-emphasis on the ‘therapist's message’ on the premise that ‘truth’ is a shared construction.
The messages, usually between one and three, are given crisply and with maximal clarity. ‘Homework’ may be assigned and another session planned (unless termination was set for this point). Messages have several purposes including the promotion of intersessional ‘work’. Three or 4 weeks is commonly set aside between meetings, and for good reason. During this time, the family, armed with new ideas, will tackle them in their day to day lives. It is not critical how they do so but important that they do so. To get back to the point about curiosity, and as Cecchin (1987) has argued, the family's interest in their own functioning should have been so aroused that they will be motivated to continue looking at themselves between sessions.
One of the authors (see Allman et al., 1992) has conducted research on the nature of the message that led to devising a classification. Messages are divisible into three broad groups: supportive, hypothesis related, and prescriptive. In the first, the message has a reassuring, encouraging, or otherwise supportive quality but it is not related to the hypothesis. A complimentary message might be that ‘The team were impressed by how open you all were in the session’ and a reassuring one that ‘This is like a new start for the family; there are bound to be uncertainties’.
Hypothesis-related messages refer to the hypothesis worked out by the therapeutic team, and may assume diverse forms. It may be stated directly, e.g., ‘Susan has assumed the role of therapist for her parents and sister in order to prevent the family's disintegration.’ There may be reference to change such as ‘The team can see John taking responsibility to look after himself; John and his father's improved relationship has allowed this to occur’. The family may be offered options, an outline of possible choices related to the hypothesis, e.g., ‘The family could risk being more open or you could continue to keep things to yourselves’. Paradoxical messages are a means to communicate a hypothesis that invites the family to revisit a feature of their functioning so that the family's difficulties are positively promoted and explicitly encouraged, e.g., ‘The team sense that your problem is working for the good of your marriage; sticking with your illness can save the marriage’. The paradox may also be split in that the family are told about a divergence of opinion in the team (Papp, 1980). For instance, the family may be informed that some team members believe it too risky for them to communicate openly, whereas others suggest this can be done safely.
Through a prescriptive message the family is given a task directly. This may or may not be related to the hypothesis. For example, the family is urged to meet on their own before the next session in order to explore what inhibits a member from relating closely to the others.
Whatever the form of message, the therapist attempts to de-emphasize the pathological status of the identified patient and to apply what the Milan group refers to as positive connotation. The latter, a brilliant innovation, rests on the premise that all behavior is purposeful, and that the purpose can be construed positively. An adolescent's ‘symptom of open grieving’ is reframed as serving the family by sparing them the anguish of grief. This quality of message calls for creative thinking and flies in the face of the customary view of symptoms as evidence of psychopathology. Again, curiosity enters the picture as the family hears this positive communication concerning an issue that they have hitherto regarded as negative and abnormal.
The above process continues during succeeding meetings and attention is paid to what occurs in the family between sessions. Duration of therapy depends on how entrenched the family dysfunction is rather than on the status of an identified patient's problems. Thus, systemic change is aimed for and the family encouraged to consider a substitute mode of functioning that is feasible and safe. In practice, sessions range in numbers from one to a dozen. If progress has not been achieved by about session 8, it is likely that alternate ways of helping the family and/or the identified patient are called for.
Termination is less problematic than in individual or group therapy. The reason is obvious. The family has come as a living group and will continue to be one after the therapist bows out. In most approaches, even when the therapist is a prominent conductor, the family's own intrinsic resources are highlighted so that these can be drawn on and exploited further upon the therapist's exit. Determining the endpoint is usually straightforward in that there is a shared sense that the work has been accomplished. A hypothesis (or set of) has been introduced, tested, and confirmed. The family system has been carefully examined in order that impediments are recognized and understood and better modes of functioning devised and implemented. The family does not have to leave functioning optimally. Instead, termination occurs when there is agreement that the family is equipped with new options and feels confident to try them out over the long term.
As alluded to earlier, this may be determined alongside a judgment that an identified patient (or other member occasionally) requires another therapy in his or her own right. A clear example is an adolescent who has felt unable to separate and individuate. While family work has explored the system that blocked ‘graduation’ to adult psychological status, the sense prevails that he could benefit from individual or group therapy by building on changes already achieved. In another example, the parents may conclude, with the therapist's support, that they have an agenda that is not pertinent to their children and therefore best handled in couple therapy.
Where assessment has been carried out diligently and motivation for change sustained, treatment proceeds smoothly. This is not to negate a possible crisis buffeting the group. But rather than being derailed, the family is encouraged to regard the crisis as a challenge with which to grapple.
Family treatment does not always succeed. Indeed, deterioration may take place, albeit in a small proportion of cases (Gurman and Kniskern, 1978). What are common difficulties encountered? The nonengaging family is problematic in that while evidence points to the need for family intervention, members cannot participate, usually because they resist letting go ‘the devil they know’. In another variation, engagement of particular members may fail. This is particularly so in the case of fathers who tend to see the target of therapy as the identified patient rather than the family as a whole.
Missed appointments may punctuate therapy, often linked to turbulent experiences between sessions or apprehension about what a forthcoming session may reveal. Like any psychotherapy, dropout is possible. On occasion, this is reasonable inasmuch as the indication for family therapy was misconstrued. In other circumstances, dropout is tantamount to failure and may derive from such factors as therapist ineptitude, unearthing of family conflict that they cannot tolerate, and inappropriate selection of a family approach based on faulty assessment.
We have referred to the possible occurrence of a family crisis. Given that the family continues as a living group during treatment, they are exposed to all manner of vicissitudes, and these may disrupt the therapeutic work. For example, an overdose by the identified patient, abrupt marital separation, or a psychiatric admission may take its toll and serve to jeopardize treatment.
In discussing the ending of the treatment, we commented on outcome. Obviously, not all families benefit. The family's dysfunction may be so intractable as to be impervious to change, hypotheses may be ‘off the mark’, the family may lack adequate psychological sophistication, members may retreat in the face of change because of insecurity, and so forth.
Occasionally, dependency becomes a problem as the family senses a greater security when relying on the therapist. The latter may inadvertently foster dependency by assuming a role of authoritativeness that impedes a growing partnership. The family's own resources are then not given expression.
Finally, a family subgroup may harbor a secret that threatens the principle of open communication between members. The therapist may be inveigled into this group, although he stressed at the onset that keeping secrets is not conducive to the therapeutic process. For example, a call to the therapist from a spouse that she is having an affair that she will not disclose to her husband or children imposes a burden on both therapist and the family work.
Astute judgment is required in these situations. No ready-made prescriptions are available but instead a keen awareness in the therapist that difficulties are possible even in a highly motivated and well selected family. The general principle, however, is to prevent their evolution if at all possible or to recognize them early and ‘nip them in the bud’.
In appraising the contemporary state of adult family therapy research, the choice is to see the glass as either half full or half empty. We opt for the more optimistic scenario. We need to remind ourselves that adult psychiatry family therapy is a toddler, dating only from the 1970s. During this time, immense strides have been made, particularly in the development of theoretical concepts. Pioneers in the field were chiefly therapists, working with families and tantalized by the nature of the process rather than its effectiveness. In hindsight, this makes sense. Models were completely lacking, the how to conduct treatment crying out for creative ideas. As can be seen in the theoretical part of the chapter, these have emerged bounteously, and continue to do so. The result is a rich array of therapeutic approaches, including several comprehensive theoretical contributions (Gurman and Kniskern, 1991). The growth has occurred at a dizzy pace with the inevitable consequence of overload. How can we make sense of the competing offerings? Is integration needed in order to forestall fragmentation of the field? Have we reached the point to reflect on what the terrain looks like? Are we now better placed to carry out outcome studies and to evaluate relative effectiveness? Tough questions and the research pathway is obstructed by many hurdles.
Observers of family therapy research, among them Gurman et al. (1986) and Bednar et al. (1988) have sought to clarify evolutionary themes and options for further work. Notwithstanding this collective endeavor, we have still not reached the enviable position say of an integrated model such as cognitive-behavioral therapy that, by dint of its relatively integrated status, has been systemically investigated, both its process and outcome, so that we are building up knowledge about how cognitive-behavioral therapy works and for what types of patients.
A complicating aspect of family therapy research is to define components of the approach, namely the therapist assembling a natural group, of varying composition, in which a dominant goal is to alter its functioning. This is altogether a more daunting matter compared with the relatively straightforward task of examining the effectiveness of say a well described treatment given to a single patient presenting with a well defined depressive syndrome.
Even if we were able to design solid outcome studies, we would be left with the conundrum of what constitutes the desired outcome and how to measure it. We can illustrate this by citing the conclusions of Asen and his colleagues (1991) in their investigation of 18 London families. Fundamental differences among the researchers emerged when handling the data. The team had decided to apply a multidimensional set of measures to assess change and at individual, dyadic, and family levels. At follow-up they noticed changes at the first two levels but not in the family as a group. The latter involved ratings of, inter alia, communication, boundaries, alliances, adaptability, and competence. The researchers were refreshingly candid in sharing their doubts about how to deal with the findings. Several contradictory interpretations were offered, e.g.: an absence of change in family functioning; the measure of that functioning nonreactive to treatment as it was a trait measure; and an inappropriate model of family therapy applied in the first place. Asen et al. concluded that the ‘assumptive worlds’ of therapists and researchers were being approved rather than the families themselves, a conclusion that makes good sense and an issue continuing to ensnare researchers. (These ethically related dimensions are discussed by Bloch et al., 1994, in The family in clinical psychiatry.)
A research team in Oxford (Bloch et al., 1991) encountered similar difficulties in their evaluation of 50 consecutive families treated in an adult family therapy clinic. Whereas two-thirds of the patients were judged to be improved at termination, only half the families were rated as functioning better or much better. Again, like the Asen team, the investigators were left with questions as how to determine what had actually been achieved.
A methodologically simpler way to wrestle with the issue is to focus solely on the identified patient's progress. Hafner et al.'s (1990) work exemplifies this choice—a case-controlled evaluation of family therapy in an inpatient setting with subsequent hospital admission data applied as the chief change criterion. Satisfactory as this study is in terms of design, the omission of a family system outcome measure leaves us hankering for more information about the group's functioning following the intervention.
With these tricky matters in mind, let us consider what research in the adult family therapy field needs to sort out. The diffuse question of whether family therapy works or not in this setting is of limited utility, and is reminiscent of the sterile debate that typified psychotherapy outcome research in the wake of Hans Eysenck's throwing down the gauntlet in 1952 (Alexander et al., 1994). While subsequent meta-analyses demonstrated that psychological interventions overall exerted useful effects across a range of conditions, the field was still open to the criticism that efficacy of a specific therapeutic approach for a particular clinical state remained unanswered. The NIMH collaborative study on the treatment of depression was an advance. Family therapy should not repeat the same error and so squander opportunities and time. Instead of posing the futile question of whether family therapy works in adult psychiatry, we should instead ascertain whether a specific approach, whose character is well identified and adherence by therapists to it confirmed, is useful for both the identified patient, with a specific presentation, and the family's functioning, again well defined.
Research has begun to fulfill these desiderata. Many studies exploring interventions in families containing a schizophrenic member have described principles of treatment, the rationale upon which it is based, aspects of the process, and outcome measures in the patient and (in some cases) the family (see, for example, Falloon et al., 1986). Helpful reviews can be found in Dixon and Lehman (1995) and Mueser and Bellack (1995). Although not as advanced as developments in schizophrenia, research conducted in the area of affective disorders has been innovative, and should pave the way for formal outcome studies (see Weber et al., 1988; Keitner, 1990).
The Maudsley study on anorexia and bulimia nervosa aptly illustrates how outcome research can contribute to the clinician (Russell et al., 1987). In a well controlled study, 80 patients were randomized to either family therapy or ‘routine individual supportive therapy’, following their discharge from an inpatient weight/restoration program. Treatment of the family involved an average 10 sessions, and individual treatment 15 sessions, spaced out over a 1-year period.
Family therapy focused on engaging the family and providing them with information about the eating disorder and the effects of starvation. Parental anxiety was acknowledged and efforts made to help parents take control of their daughter's diet. In parallel with improved physical status, therapy turned progressively to typical adolescent issues of separation and individuation and how these might be accomplished. A structural approach was applied, with systemic and strategic measures incorporated when progress slowed down.
Family therapy of a specific type can be applied to the family as a group in the light of system dysfunction. Thus, while the above research concerning particular psychiatric states, and involving an identified patient, is necessary for progress, this does not preclude outcome studies where the family is the principal target of change. We illustrate this with a particular form of family grief therapy developed in the Center of Palliative Care in the University of Melbourne (Kissane et al., 1998). The model was derived from empirical research on the outcome of family grieving in an oncology setting. A 13-month follow-up yielded five family clusters of which two were distinctly dysfunctional, two functional, and an intermediate group at risk of maladaptive grieving. Three dimensions of family functioning were critical: cohesion, managing conflict, and expressiveness. The investigators then devised a model highlighting the goals of promoting cohesiveness, expressiveness, and optimal management of conflict. A corresponding screening instrument was applied to identify dysfunctional families.
Fifteen therapists were trained to use the emergent treatment guidelines and to work under close supervision in order to ensure that they adhered to the model. The randomized controlled trial (RCT) showed clearly the model's suitability and feasibility. Treatment began prior to the death of a terminally ill parent and extended into the bereavement period. Outcome measures included individual psychosocial morbidity and adaptation and the family's functioning. The model and its practical application are described in detail in Family focused grief therapy (Kissane and Bloch, 2002). The findings of the RCT are at the time of writing being subjected to statistical analyses.
This necessarily schematic account of research developments on family therapy in adult psychiatry suggests likely future trends. We can best summarize what research should strive for as: ‘Specificity is of the essence’.
While postmodernist foundations of narrative therapies might suggest that they are less amenable to traditional research of the sort we have described, this has not proven to be entirely so. A group of researchers in London studied the accounts by family members of their experiences caring for an acutely psychotic relative, and discerned two patterns of narrative. In one that was described as having meaning, members’ stories depicted themes of reparation and restitution and integrated the illness into ongoing family life. In the other, described as frozen or chaotic narratives, members viewed the illness as a series of random events (affinity with Byng-Hall's model—Byng-Hall, 1995; Stern et al., 1999). The clinical implications of these two patterns and their relation to empirical studies of relapse prevention await elucidation.
Psychoeducational interventions for children whose parents suffer from a major affective disorder have been modified to pay attention to the children's narratives of their experiences. Initial findings indicate the possibility of improving the children's resilience and coping with their parents’ illness (Focht and Beardslee, 1996). This research approach is promising in terms of its preventative potential and could be extrapolated to the adult sphere.
From a few charismatic figures practicing idiosyncratic, innovative methods of family therapy, the field has developed into an immense, skillfully marketed enterprise in many countries, particularly the US, with hundreds of books, scores of training courses, several dozen journals, and a year-round program of local, national, and international conferences and workshops (Liddle, 1991).
Formal training may occur in one of three contexts (Goldenberg, I. and Goldenberg, H., 1996).
University-based, degree-granting programs view family therapy as a distinct profession, with its own corpus of knowledge, and offer diploma, masters, PhD, and postdoctoral training.
Free-standing institutes also tend to see family therapy as a distinct discipline and provide part-time training, usually of shorter duration than most university-based programs. A prerequisite for entry in most of these is that the candidate has completed basic training in one of the health professions.
Within university-affiliated hospitals and clinics that provide professional training in psychiatry, psychology, social work, and occupational therapy, many provide a brief course in the theory and practice of family therapy as part of general professional training.
Although there is a vast spectrum of training experiences to which students are exposed, most programs include:
Live supervision of clinical work with the supervisor (and often other students) observing the trainee and family from behind a one-way screen. Some clinicians consider the one-way screen to be dehumanizing and too objectifying of the family as well as adding to the trainee's performance anxiety. They advocate instead a model of co-therapy (trainee and supervisor), often with other students sitting in the interview room in full view of the family.
Video recording of the trainee's work, which is then reviewed by her in the presence of supervisor and fellow students is widely used. Tapes of particular models conducted by eminent therapists are also popular.
Whether training requires familiarity with concepts and techniques of a variety of schools or whether it is preferable to develop expertise in only one school remains debatable. Free-standing institutes tend to be run by therapists of a particular school, so that after a generally cursory overview of the field training is restricted to a specific model. This is even more likely when the program is part of general education in psychiatry, psychiatric nursing, psychology, and social work.
Diversity of schools and training reflects an uncertainty as to whether family therapy is a distinct profession, a method of conceptualizing psychopathology, or a set of therapeutic methods to add to the armamentarium of the mental health professional. This issue is further compounded by the aforementioned trend toward integrating psychodynamic, attachment, systems, feminist, and narrative approaches.