Bowen Family Systems Theory and Practice: Illustration and Critique
Jenny Brown *
* Private practice, 2 Oswald St Mosman NSW 2088; djkbrown @ Ozemail.com.au
This paper will give an overview of Murray Bowen's theory of family systems. It will describe the model's development and outline its core clinical components. The practice of therapy will be described as well as recent developments within the model. Some key criticisms will be raised, followed by a case example which highlights the therapeutic focus of Bowen's approach.
Murray Bowen's family systems theory (shortened to `Bowen theory 'from 1974) was one of the first comprehensive theories of family systems functioning (Bowen, 1966, 1978, Kerr and Bowen, 1988). While it has received sporadic attention in Australia and New Zealand, it con- tinues to be a central influence in the practice of family therapy in North America. It is possible that some local family therapists have been influenced by many of Bowen's ideas without the connection being articulated. For example, the writing of Guerin (1976, 1987), Carter and McGoldrick (1980, 1988), Lerner (1986, 1988, 1990, 1993) and Schnarch (1991, 1997) all have Bowenian Theory at the heart of their conceptualisations.
There is a pervasive view amongst many proponents of Bowen's work that his theory needs to be experienced rather than taught (Kerr, 1991). While this may be applic- able if one can be immersed in the milieu of a Bowenian training institute, such an option, to my knowledge, is not available in this country. Bowen's own writings have also been charged with being tedious and difficult to read (Carter, 1991). Hence it seems pertinent to present this influential theory in an accessible format.
DEVELOPMENT OF THE MODEL Murray
Bowen was born in 1913 in Tennessee and died in 1990. He trained as a psychiatrist and originally prac- tised within the psychoanalytic model. At the Menninger Clinic in the late 1940s, he had started to involve mothers in the investigation and treatment of schizo- phrenic patients. His devotion to his own psychoanalytic training was set aside after his move to the National Institute of Mental Health in 1954, as he began to shift from an individual focus to an appreciation of the di- mensions of families as systems. At the NIMH, Bowen began to include more family members in his research and psychotherapy with schizophrenic patients. In 1959 he moved to Georgetown University and established the Georgetown Family Centre (where he was director until his death). It was here that his developing theory was extended to less severe emotional problems. Be- tween 1959 and 1962 he undertook detailed research into families across several generations. Rather than develop- ing a theory about pathology, Bowen focused on what he saw as the common patterns of all `human emotional systems'. With such a focus on the qualitative similarities of all families, Bowen was known to say frequently, `There is a little schizophrenia in all of us' (Kerr and Bowen, 1988).
In 1966, Bowen published the first `orderly presenta- tion 'of his developing ideas (Bowen, 1978: xiii). Around the same time he used his concepts to guide his inter- vention in a minor emotional crisis in his own extended family, an intervention which he describes as a spectacu- lar breakthrough for him in theory and practice (Bowen, 1972 in Bowen, 1978). In 1967, he surprised a national family therapy conference by talking about his own family experience, rather than presenting the anticipated formal paper. Bowen proceeded to encourage students to work on triangles and intergenerational patterns in their own families of origin rather than undertaking individual psychotherapy. From this generation of trainees have come the current leaders of Bowenian Therapy, such as Michael Kerr at the Georgetown Family Centre, Philip Guerin at the Centre for Family Learning, Betty Carter at the Family Institute of Westchester, and Monica McGoldrick at the Family Institute of New Jersey.
While the core concepts of Bowen's theory have changed little over two decades, there have been significant expansions: the focus on life cycle stages (Carter and McGoldrick, 1980, 1988) and the incorporation of a feminist lens (Carter, Walters, Papp, Silverstein, 1988; Lerner, 1983; Bograd, 1987).
Bowen's focus was on patterns that develop in families in order to defuse anxiety. A key generator of anxiety in families is the perception of either too much closeness or too great a distance in a relationship. The degree of anxiety in any one family will be determined by the current levels of external stress and the sensitivities to particular themes that have been transmitted down the generations. If family members do not have the capacity to think through their responses to relationship dilemmas, but rather react anxiously to perceived emotional de- mands, a state of chronic anxiety or reactivity may be set in place.
The main goal of Bowenian therapy is to reduce chronic anxiety by 1) facilitating awareness of how the emotional system functions; and 2) increasing levels of differentiation, where the focus is on making changes for the self rather than on trying to change others. Eight interlocking concepts make up Bowen's theory. This paper will give an overview of seven of these. The eighth attempts to link his theory to the evolution of society, and has little relevance to the practice of his therapy. However, Wylie (1991) points out in her bio- graphical piece following Bowen's death that this interest in evolutionary process distinguishes Bowen from other family therapy pioneers. Bowen viewed himself as a scientist, with the lofty aim of developing a theory that accounted for the entire range of human behaviour and its origins.
1. Emotional Fusion and Differentiation of Self
`Fusion 'or` lack of differentiation' is where individual choices are set aside in the service of achieving harmony within the system. Fusion can be expressed either as a sense of intense responsibility for another's reactions, or by emotional `cutoff 'from the tension within a rela- tionship (Kerr and Bowen, 1988; Herz Brown, 1991). Bowen's research led him to suggest that varying degrees of fusion are discernible in all families. `Differentiation ', by contrast, is described as the capacity of the individual to function autonomously by making self directed choices, while remaining emotionally connected to the intensity of a significant relationship system (Kerr and Bowen, 1988). Bowen's notion of fusion has a different focus to Minuchin's concept of enmeshment, which is based on a lack of boundary between sub-systems (Minuchin, 1974). The structural terms `enmeshment ' and `disengagement 'are in fact the twin polarities of Bowen's `fusion '. Fusion describes each person's reactions within a relationship, rather than the overall structure of family relationships. Hence, anxiously cutting off the relationship is as much a sign of fusion as intense sub- missiveness. A person in a fused relationship reacts im- mediately (as if with a reflex, knee jerk response) to the perceived demands of another person, without being able to think through the choices or talk over relation- ship matters directly with the other person. Energy is invested in taking things personally (ensuring the emotional comfort of another), or in distancing oneself (Ensuring one's own). The greater a family's tendency to fuse, the less flexibility it will have in adapting to stress.
Bowen developed the idea of a `differentiation of self scale 'to assist in teaching this concept. He points out that this was not designed as an actual instrument for assigning people to particular levels (Kerr and Bowen, 1988: 97 ^ 98). Bowen maintains that the speculative nature of estimating a level of differentiation is com- pounded by factors such as stress levels, individual dif- ferences in reactivity to different stressors, and the degree of contact individuals have with their extended family. At one end of the scale, hypothetical `complete differentiation 'is said to exist in a person who has resolved their emotional attachment to their family (ie shifted out of their roles in relationship triangles) and can therefore function as an individual within the family group. Bowen did acknowledge that this was a lifelong process and that `total 'differentiation is not possible to attain.
Bowen described triangles as the smallest stable relation- ship unit (Kerr and Bowen, 1988: 135). The process of triangling is central to his theory. (Some people use the term `triangulation ', deriving from Minuchin (1974: 102), but Bowen always spoke of `triangling '.) Triangling is said to occur when the inevitable anxiety in a dyad is relieved by involving a vulnerable third party who either takes sides or provides a detour for the anxiety (Lerner, 1988; James, 1989; Guerin, Fogarty, Fay and Kautto, 1996). An example of this pattern would be when Person A in a marriage begins feeling uncomfortable with too much closeness to Person B. S / he may begin withdraw- ing, perhaps to another activity such as work (the third point of the triangle). Person B then pursues Person A, which results in increased withdrawal to the initial triangled-in person or activity. Person B then feels neglected and seeks out an ally who will sympathise with his / her sense of exclusion.This in turn leads to Person A feeling like the odd one out and moving anxiously closer to Person B. Under stress, the triangling process feeds on itself and interlocking triangles are formed through- out the system. This can spill over into the wider com- munity, when family members find allies, or enemies to unite against, such as doctors, teachers and therapists.
Under calm conditions it is difficult to identify tri- angles but they emerge clearly under stress. Triangles are linked closely with Bowen's concept of differenti- ation, in that the greater the degree of fusion in a rela- tionship, the more heightened is the pull to preserve emotional stability by forming a triangle. Bowen did not suggest that the process of triangling was necessarily dysfunctional, but the concept is a useful way of grasp- ing the notion that the original tension gets acted out elsewhere. Triangling can become problematic when a third party's involvement distracts the members of a dyad from resolving their relationship impasse. If a third party is drawn in, the focus shifts to criticising or worry- ing about the new outsider, which in turn prevents the original complainants from resolving their tension. According to Bowen, triangles tend to repeat themselves across generations. When one member of a relationship triangle departs or dies, another person can be drawn into the same role (eg `villain ',` rescuer', `victim ',` black sheep ', `martyr'). For example, in my own family of origin I found myself moving into the role of peace- maker after the death of my mother, who had mediated the tension between my father and brother. This on- going triangle served to detour the anxiety that had been played out between fathers and sons in the family over the generations.
3. Nuclear Family Emotional System
In positing the `nuclear family emotional system ', Bowen focuses on the impact of `undifferentiation 'on the emotional functioning of a single generation family. He asserts that relationship fusion, which leads to tri- angling, is the fuel for symptom formation which is mani- fested in one of three categories. These are: (A) couple conflict. (B) illness in a spouse. (C) projection of a problem onto one or more children.
A. Couple Conflict
The single generation unit usually starts with a dyad--a couple who, according to Bowen, will be at approxi- mately equal levels of differentiation (ie both have the same degree of need to be validated through the rela- tionship). Bowen believed that permission to disagree is one of the most important contracts between indi- viduals in an intimate relationship (Kerr and Bowen, 1988: 188). In a fused relationship, partners interpret the emotional state of the other as their responsibility, and the other's stated disagreement as a personal affront to them. A typical pattern in such emotionally intense relationships is a cycle of closeness followed by conflict to create distance, which in turn is followed by the couple making up and resuming the intense closeness. This pattern is a `conflictual cocoon '(Kerr and Bowen, 1988: 192), where anxiety is bound within the conflict cycle without spilling over to involve children. Bowen suggested the following three ways in which couple con- flict can be functional for a fused relationship, in which `Each person is attempting to become more whole through the other '(Lederer and Lewis, 1991). (1) Con- flict can provide a strong sense of emotional contact with the important other. (2) Conflict can justify people's maintaining a comfortable distance from each other without feeling guilty about it. (3) Conflict can allow one person to project anxieties they have about them- selves onto the other, thereby preserving their positive view of self (Kerr and Bowen, 1988: 192).
B. Symptoms in a Spouse
In a fused relationship, where each partner looks to the other's qualities to fit his / her learned manner of re- lating to significant others, a pattern of reciprocity can be set in motion that pushes each spouse's role to oppo- site extremes. Drawing from his analytic background, Bowen described this fusion as `the reciprocal side of each spouse's transference '(Kerr and Bowen, 1988: 170). For example, what may start as an overly responsible spouse feeling compatible with a more dependent partner, can escalate to an increasingly controlling spouse with the other giving up any sense of contributing to the relationship. Both are equally undifferentiated in that they are defining themselves according to the reactions of the other; however the spouse who makes the most adjustments in the self in order to preserve relation- ship harmony is said by Bowen to be prone to develop- ing symptoms. The person who gets polarised in the underfunctioning position is most vulnerable to symp- toms of helplessness such as depression, substance abuse and chronic pain. The overfunctioning person might also be the one to develop symptoms, as s / he becomes overburdened by attempts to make things `right 'for others.
C. Symptoms in a Child
The third symptom of fusion in a family is when a child develops behavioural or emotional problems.This comes under Bowen's fourth theoretical concept, the Family Projection Process.
4. Family Projection Process
In the previous two categories the couple relationship is the focus of anxiety without it significantly impacting on the functioning of the next generation. By contrast, the family projection process describes how children develop symptoms when they get caught up in the previous generation's anxiety about relationships.
The child with the least emotional separation from his / her parents is said to be the most vulnerable to de- veloping symptoms. Bowen describes this as occurring when a child responds anxiously to the tension in the parents' relationship, which in turn is mistaken for a problem in the child. A detouring triangle is thus set in motion, as attention and protectiveness are shifted to the child. Within this cycle of reciprocal anxiety, a child becomes more demanding or more impaired. An example would be when an illness in a child distracts one parent from the pursuit of closeness in the marriage. As tension in the marriage is relieved, both spouses become in- vested in treating their child's condition, which may in turn become chronic or psychosomatic.
As in all of Bowen's constructs, `intergenerational pro- jection 'is said to occur in all families in varying degrees. Many intergenerational influences may determine which child becomes the focus of family anxiety and at what stage of the life cycle this occurs. The impact of crises and their timing also influences the vulnerability of certain children. Bowen viewed traumatic events as significant in highlighting the family processes rather than as actually `causing 'them.
5. Emotional Cutoff
Bowen describes `emotional cutoff 'as the way people manage the intensity of fusion between the generations. A `cutoff 'can be achieved through physical distance or through forms of emotional withdrawal. Bowen distin- guishes between `breaking away 'from the family and `Growing away 'from the family.` Growing away' is viewed as part of differentiation--adult family members follow independent goals while also recognising that they are part of their family system. A `cutoff 'is more like an es- cape; people `decide 'to be completely different to their family of origin. While immediate pressure might be re- lieved by cutoff, patterns of reactivity in intense relation- ships remain unchanged and versions of the past, or its mirror image, are repeated. Bowen proposes that: If one does not see himself as part of the system, his only options are either to get others to change or to withdraw.
If one sees himself as part of the system, he has a new option: to stay in contact with others and change self (Kerr and Bowen, 1988: 272 ^ 273).
`Cutoffs'are not always dramatic rifts. An example of a covert emotional cutoff would be one family member maintaining an anxious silence in the face of another's anger. The pull to restore harmony overwhelms the ability to stay in contact with the issue that has been raised.
A central hypothesis of Bowen's theory is that the more people maintain emotional contact with the pre- vious generation, the less reactive they will be in current relationships. Conversely, when there are emotional cut- offs, the current family group can experience intense emotional pressure without effective escape valves. This family tension is like `walking on eggshells', as issues which remain unresolved from the cutoff are carefully avoided. Triangling provides a detour, as family mem- bers enlist the support of others for their own position in relation to the cutoff.
6. Multigenerational Transmission Process
This concept of Bowen's theory describes how patterns, themes and positions (roles) in a triangle are passed down from generation to generation through the projection from parent to child which was described earlier. The impact will be different for each child de- pending on the degree of triangling they have with their parents.
Bowen's focus on at least three generations of a family when dealing with a presenting symptom is certainly a trademark of his theory. The attention to family patterns over time is not just an evaluative tool, but an interven- tion that helps family members get sufficient distance from their current struggle with symptoms to see how they might change their own part in the transmission of anxiety over the generations. As Monica McGoldrick (1995: 20) writes in applying Bowenian concepts: By learning about your family and its history and getting to know what made family members tick, how they related, and where they got stuck, you can consider your own role, not simply as victim or reactor to your experiences but as an active player in interactions that repeat themselves.
7. Sibling Positions
Employing WalterToman's (1976) sibling profiles, Bowen considered that sibling position could provide useful information in understanding the roles individuals tend to take in relationships. For example, Toman's profiles describe eldest children as more likely to take on respon- sibility and leadership, with younger siblings more com- fortable being dependent and allowing others to make decisions. Middle children are described as having more flexibility to shift between responsibility and depen- dence and `only 'children are seen as being responsible, and having greater access to the adult world. Bowen noted that these generalised traits are not universally applicable and that it is possible for a younger sibling to become the `functional eldest '. Bowen was especially interested in which sibling position in a family is most vulnerable to triangling with parents. It may be that a parent identifies strongly with a child in the same sibling position as their own, or that a previous cross generational triangle (eg an eldest child aligned with a grandparent against a parent) may be repeated. If one sibling in the previous generation suffered a serious illness or died, it is more likely that the child of the present generation in the same sibling position will be viewed as more vulnerable and therefore more likely to detour tensions from the parental dyad.
Helping the client understand and think beyond the limitations of their own sibling position and role is a goal of Bowenian family of origin work. Clients are encouraged to consider how assumptions about relation- ships are fuelled by their sibling role experience. As with other aspects of Bowen's theory, the impact of gender and ethnicity on sibling role is not considered. For example, there is no exploration of how a family's ethni- city influences which birth order position and which gender is more valued, or how the gender of any sibling position tends to influence whether the role is primarily relational (female), or task oriented (male).
THE MODEL IN CLINICAL PRACTICE
Bowen's is not a technique focused model which incor- porates specific descriptions of how to structure therapy sessions. The goal of therapy is to assist family members towards greater levels of differentiation, where there is less blaming, decreased reactivity and increased respon- sibility for self in the emotional system. Perhaps the most distinctive aspects of Bowen's therapy are his emphasis on the therapist's own family of origin work, the central role of the therapist in directing conversation and his minimal focus on children in the process of therapy.
Bowen views therapy in three broad stages. Stage one aims to reduce clients' anxiety about the symptom by encouraging them to learn how the symptom is part of their pattern of relating. Stage two focuses adult clients on `self 'issues so as to increase their levels of differenti- ation. Clients are helped to resist the pull of what Bowen termed the `togetherness force 'in the family (Bowen, 1971 in Bowen, 1978: 218). In the latter phases of therapy, adult clients are coached in differentiating themselves from their family of origin, the assumption being that gains in differentiation will automatically flow over into decreased anxiety and greater self-responsibility within the nuclear family system.
The Role of the Therapist
The role of the therapist is to connect with a family without becoming emotionally reactive. Emphasis is given to the therapist maintaining a `differentiated 'stance. This means that the therapist is not drawn into an over re- sponsible ^ under responsible reciprocity in attempts to be helpful. A therapist position of calm and interested investigation is important, so that the family begins to learn about itself as an emotional system. Bowen in- structs therapists to move out of a healing or helping position, where families passively wait for a cure, `To getting the family into position to accept responsi- bility for its own change '(Bowen, 1971 in Bowen, 1978: 246).
Bowen warns of the problems of therapists losing sight of their part in the system of interactions, where they may be inducted into a mediating role in a triangle with the family. Hence there is a high priority given to understanding and making changes within the therapist's own family of origin. In training, the emphasis is on the trainees' level of differentiation, and not on therapeutic technique. The therapist's resolution of family of origin issues is reflected in the:
... Ability to be in emotional contact with a difficult, emo- tionally charged problem and not feel compelled to preach about what others should do, not rush in to fix the problem and not pretend to be detached by emotionally insulating oneself (Kerr and Bowen, 1988: 108).
The therapist is active in directing the therapeutic con- versation. Enactments are halted so as to prevent the escalation of clients' anxiety. Clients are asked to talk directly to the therapist so that other family members can `listen and `` really hear'', without reacting emotion- ally, for the first time in their lives together '(Bowen, 1971 in Bowen, 1978: 248). Bowen himself would avoid couple interaction in the room and concentrate on inter- viewing one spouse in the presence of the other. Bowen clearly avoided asking for emotional responses, which he saw as less likely to lead to differentiation of self, prefer- ring mostly to ask for `thoughts ',` reactions' and `im- pressions' (Bowen, 1971, in Bowen, 1978: 226). He called this activity `externalizing the thinking of each client in the presence of the other '(Bowen, 1975 in Bowen, 1978: 314).
Children in Bowen's Therapy A surprising feature of Bowen's family therapy is his tendency to minimise the involvement of children. While Bowen might include children in the beginning stage of therapy, he would soon dismiss them, focusing on the adults as the most influential members of a family system (Bowen, 1975 in Bowen, 1978: 298). Excluding a child from therapy responsibility is viewed as a de- triangling manoeuvre. When parents cannot use the child as a `triangle person 'for issues between them, and the therapist resists taking the replacement role in the triangle, parents can begin differentiating their respec- tive selves from one other.
The beginning sessions in Bowenian therapy focus on information gathering in order to form ideas about the family's emotional processes, which concurrently pro- vides information to family members about the present- ing problem in its systemic context. The presenting problem is tracked through the history of the nuclear family and into the extended family system. A multi- generational genogram is a useful tool for recording this information (McGoldrick and Gerson, 1985; Kerr and Bowen, 1988: 306 ^ 313). The therapist looks for clues about the emotional process of the particular family, including: patterns of regulating closeness and distance, how anxiety is dealt with in the system, what triangles get activated, the degree of adaptivity to changes and stressful events, and any signs of emotional `cutoff '. Information collected is acknowledged to be extremely subjective, especially when extended family are dis- cussed; but stories about past generations are viewed as useful clues to the roles people occupy in triangles and the tensions that remain unresolved from their families of origin. If for example, a member of the extended family is described as `the rebel ', the therapist explores what events gave rise to this label, who else has occupied this role across the generations and how triangles formed around family crises involving `rebellion '. Calming family members' anxiety in the early stages of therapy might involve helping them to make connections between the development of symptoms and potent themes in a family's history. Another aim will be to loosen the central triangle that has formed around, and maintains, the presenting problem. Teaching clients about systems concepts as they operate in their own family is part of therapy at this stage. This does not mean attempting to convince people to do things differently but to encour- age family members to see beyond their biases so that it is possible for them to consider each person's part in the family patterns.
Questions that Encourage Differentiation
The therapist asks questions that assume that the adult client can be responsible for his / her reactiveness to the other. An example would be, `How do you understand the way you seem to take your child's acting out so personally? 'In response to such questions, family members are encouraged to take an `I 'position where they speak about how they view the problem, without attacking, or defending against, another family member (Bowen, 1971a in Bowen, 1978: 252; Goodnow and Lim, 1997). Clients are taught to make personal statements about their thoughts and feelings in order to facilitate a greater sense of responsibility in a relationship. For example, an accusatory statement such as, `You are so selfish to cause this much worry for your parents! ', is shifted to, `I am really concerned that this might affect your school grades '. The parent is encouraged to `own' their worries, rather than to project their anxieties through blaming statements. Developing such a `self- focus' is said to be crucial in lowering anxiety and enabling `person to person 'relationships where each family member can think about the part they play in problematic interactions.
Creating a Multigenerational Lens
Bowen's multigenerational model goes beyond the view that the past influences the present, to the view that patterns of relating in the past continue in the present family system (Herz Brown, 1991). Hence the therapist uses questions to encourage clients to think about the connection between their present problem and the ways previous generations have dealt with similar relationship issues. For example, if the onset of a symptom followed a death in the family, the therapist asks about how grief has been dealt with in previous generations. Questions seek to uncover family belief systems as well as the way relationships have shifted in response to loss. Tracking symptoms and exploring related themes over at least three generations makes it more difficult for individuals to blame one another for individual deficiencies. As therapist and family members see how patterns repeat over generations, it is possible to identify the `automatic ' reactions of family members towards each other:
The ability to act on the basis of more awareness of rela- tionship process (not blaming self or others, but seeing the part each plays) can, if done repeatedly in important rela- tionships, lead to some reduction in emotional reactivity and chronic anxiety (Kerr and Bowen, 1988: 132).
This is probably the central technique in Bowenian therapy. The client is first helped to recognise both the subtle and the more obvious ways that they are `Triangled 'by others, and the ways in which they attempt to triangle others in their turn. The therapist uses ques- tions to facilitate the family members' awareness of their roles in family triangles. Simple open ended tracking questions, using what Herz Brown (1991) terms the four `Ws' (who, what, when and where) help clients to become `detectives' in their own interpersonal systems. It is often very difficult for family members to identify the triangles they participate in, and the sometimes covert ways in which they detour anxiety. An example would be a client who was struggling to understand her negativity towards her father. When questioning included her mother's role in these emotions, the client began to see that her view of her father was influenced by her position in a triangle. As her mother's ally in this triangle, she viewed her father as the inadequate husband who left her mother feeling needy. Once triangles have been identified, family members are helped to plan ways of communicating a neutral position to others, leaving the dyad to communicate directly with each other.The goal is for a family member to find a less reactive position in the face of the other's anxiety. This will require different stances in different systems, ranging from refusing to discuss the deficiencies of another behind his/her back, to reversing one's usual reaction in a triangle. For example, when the predictable pattern in the family system is to keep distance between those who haven't been able to work out their problems, the therapist helps a family member to plan strategies that shift their usual role in maintaining the avoidance. The family member might encourage more involvement between the conflictual twosome, or change the subject when invited to discuss the conflict. Reversal is a key de- triangling technique.When for example a family member A complains about how uncaring another person is, person C reverses the predictable sympathetic response, substituting a casual comment about how considerate person B seems for not putting demands on A's time and energy. Unlike a strategic intervention, the goal of any detriangling stance is not to change the other's rela- tionship but to express one's neutrality about it. A calm and thoughtful neutral stance prevents one from anxiously reacting to the tension of another relationship by `taking sides'.
Coaching: Family Therapy with an Individual
Another distinguishing feature of Bowen's model is its validity in working with a single adult. The term `coach- ing' describes the work of the therapist giving input and support for adult clients who are attempting to develop greater differentiation in their families of origin. Clients should feel in charge of their own change efforts, with the therapist acting as a consultant. Bowen thought that a person's efforts to be more differentiated would be more productive when the focus shifted away from the intensity of the nuclear family to the previous generation. The emphasis is on self-directed efforts to detriangle from family of origin patterns. An individual's efforts can modify a triangle, which in turn ripples through to change in the whole extended family.
Bowen described `coaching' as `family psychotherapy with one family member' (Bowen, 1971 in Bowen, 1978: 233). This therapy takes on the flavour of teaching, as clients learn about the predicable patterns of triangles. The therapist supports their efforts in returning to their families to observe and learn about these patterns. Clients practise controlling their emotional reactivity in their family and report their struggles and progress in following sessions. During family of origin coaching, clients use letters, telephone calls, visits and research about previous generations to gain a systemic perspec- tive on their family's emotional processes and a sense of their own inheritance of these patterns. The therapist prepares clients for the anxiety they will encounter if they shift from their customary roles in their families of origin. Any such changes will inevitably disturb the predictable balance of family patterns and therefore heighten anxiety and resistance. Change is viewed as a three step process where (a) one takes a new position, (b) family members react and (c) the new stance is main- tained in the face of pressure to revert to the original position (Herz Brown, 1991). Bowen (1978) emphasised that it is what happens in step `c' that really determines whether change occurs.
Bowen's model has been adopted and developed by many prominent therapists. Rather than attempt to summarise all of these developments, I shall focus on the applications of the model by Betty Carter and Monica McGoldrick which have influenced the practice of the Family Institute of Westchester in New York and the Family Institute of New Jersey.
Since the early 1980s, the work of Carter, McGoldrick and their colleagues has expanded Bowen's framework to include attention to the family life cycle (Carter and McGoldrick, 1980, 1988.) As well as the `vertical' flow of anxiety through the generations, Carter included an assessment of `horizontal' stress as families move through various stages of the life cycle.Vertical and hori- zontal patterns converge, as multigenerational tensions impact on the ways that life cycle tasks and disruptions are negotiated. The stress of life cycle changes affects the choice of family of origin issues focused upon in the current generation. Using a life cycle perspective, symptom development is viewed in the context of an unresolved adjustment to a life cycle task.
Acknowledging the significance of gender, race, ethnicity and class on a family's progression through life cycle stages was an important development in family assess- ment (eg McGoldrick, Pearce and Giordano, 1982; Carter et al., 1988; McGoldrick, Anderson and Walsh, 1988; Herz Brown, 1991). This much broader focus provides what Carter has called a `multi-contextual lens'. These variables are part of the context of the family's `horizontal' story and underlie the potent themes of a family's multigenerational legacy. Patterns of gender across the generations are viewed as powerfully contri- buting to the roles that people occupy in the family emotional system. The inclusion of gender sensitivity in a Bowenian framework means that the therapist helps clients to look not only at patterns of relating over the generations but also to critique the roles they occupy in relationships. Such a focus is not confined to the family system's gender expectations but includes questions that look for connections to socially defined gender roles. Betty Carter, in developing her work from the women's project (Carter et al., 1988), has outlined how Bowen's key concepts (fusion, differentiation and triangles) need to be viewed differently from a feminist position. Gender roles will determine the way men and women express fusion, with women socialised to be dependent and ap- proval seeking and men socialised to withdraw and emo- tionally `cut off'. Carter asserts that the concept of fusion can easily be misused to pathologise the `over-involved female' while overlooking the distant male. With a `gender sensitive lens', a Bowenian therapist validates rather than pathologises the relational concerns of women and explores ways that men can take responsi- bilities in this sphere. The distancing of a male will be seen not only as a symptom of lack of differentiation but also as a socially prescribed reaction.
Likewise, the nature of a relationship triangle is influ- enced by gender related behaviour. Carter illustrates the different ways a therapist might view a triangle with and without the feminist lens. The triangle of a husband in a distant position, with his wife and mother in conflict, would be viewed by a feminist Bowenian therapist as `a case of two women bumping into each other as each tries to carry out her family responsibilities in the face of the man's withdrawal' (Carter et al., 1988). Inter- ventions will respect the women's roles and dilemmas and focus on how the husband can choose to be more involved in both significant relationships. Without such a lens, the detriangling strategy would typically be to have the husband set more boundaries with his mother --which has the effect of preserving the gendered stereotype of the `possessive' mother in law.
The therapist is challenged to recognise that no inter- vention is free from societal constructs in regard to gender and power (including race, ethnicity, class and sexual orientation) so that `every intervention will have a different and special meaning for each sex' (Carter et al., 1988). Thus therapists expand their questioning to ask about the relational impact of each spouse's income and ethnicity. The organisation of child care and housework is also explored. Therapists are encour- aged to challenge men's excuses that work prevents family involvement and women's expectations about financial support (Carter, 1996). An awareness of the impact of therapists' own value system on their therapy is also stressed (Carter, 1992).
For Bowenian therapists in the nineties, the core of Bowen's theory of symptom development and change re- mains unaltered. What has been added is attention to how wider socio-political issues of power and hierarchy are played out as couple or family problems. A broad range of systemic techniques such as restorying and circular questioning can readily be incorporated into the model (Carter and McGoldrick, 1988).
CRITIQUE OF BOWEN'S MODEL
Bowen's model of family therapy is perhaps most distinc- tive for its depth of evaluation beyond symptoms in the present. Its focus on emotional processes over the gen- erations and on individuals' differentiation within their systemic context offers family therapists a multi-level view that has usually been reserved for psychodynamic therapies. Bowen's model pays attention to the emotional interaction of therapists and their clients and expects that the process of therapy must in some way be applied to the therapists' own lives, so that they are able to remain meta to the client family system.
A number of Bowenian therapists acknowledge that the wider focus of Bowen's model can be a drawback in that that many clients want only to address symptom relief in the nuclear family (Young, 1991). For the Bowenian therapist, symptom reduction is seen only as the ground work from which families can proceed less anxiously towards working on detriangling and im- proved levels of differentiation. Herein lies a clear danger of discrepancies in client and therapist goals.
While Bowenian therapy has been embraced by some leading feminist therapists, such as Betty Carter and Harriet Goldhor Lerner, it has also received its share of criticism from a feminist perspective. Deborah Leup- nitz (1988) points out that Bowen, along with other male family therapy pioneers, has paid rather too much atten- tion to the mother's contribution to symptom develop- ment in the child. Some support for this can be found by scanning the index to Kerr and Bowen (1988), where `fathers'do not warrant a category yet `mothers'are refer- enced in relation to families of schizophrenics, levels of differentiation in the child, and their role in triangles (Kerr and Bowen, 1988: 395). [The index to Bowen's own collected papers, Family Therapy in Clinical Practice, however, includes one reference to `fathers' and none to `mothers': Eds.] A perceived over-investment by a mother in her child is seen as a sign of undifferentiation.
Unlike the current feminist therapists who use the Bowenian model, Murray Bowen (along with many of his Georgetown colleagues) failed to contextualise maternal behaviour. Patriarchal assumptions about male ^female roles and family organisation are not acknowledged or critiqued, which leaves women vulnerable to having their socially prescribed roles pathologised. Women are readily labelled as `over concerned', and their active, rela- tional role in families too easily labelled as `fused' and `undifferentiated'. There is no questioning of societal norms that can be seen to `[school] females into undiffer- entiation by teaching them always to put others' needs first' (Leupnitz, 1988: 43).
The women's project in family therapy asserts that a model such as Bowen's pressures the woman to `back off' while placating and courting the distant male (Carter et al., 1988). Carter asserts that this is not only biased against women but disrespectful of men since the model assumes men's limitations in terms of emotional engage- ment in therapy and family relationships. An ongoing challenge for feminist Bowenian therapists is to recon- struct a therapy language of intimacy and attachment that is not misused to imply dysfunction (Bograd, 1987; Carter et al., 1988).
Another criticism that flows from the biases of Bowen's `male defined' terminology, is that his is a therapy lacking in attention to feelings (Luepnitz, 1988). It is asserted that Bowen's therapy focuses on being rational and objective in relation to emotional processes, which rele- gates to a low priority the expression of emotions in therapy. My own experience of this model, with its invi- tation to explore the `tapestry' of one's family across the generations, is that it is an emotionally intense therapy. While Bowen may emphasise the goal of helping the client learn about their family's emotional processes, in practice it is the experience of the emotions, embedded in family of origin relationships that is a key motivator for the client to undertake family of origin work. I recall Betty Carter, in asking a man about his relationship with his own father, tapping deeply into emotions that moti- vated him to make changes in his ways of relating.