video

Chapter 3

GROUP COHESIVENESS

In this chapter I examine the properties of cohesiveness, the consider­ able evidence for group cohesiveness as a therapeutic factor, and the

various pathways through which it exerts its therapeutic influence. What is cohesiveness and how does it influence therapeutic outcome? The

short answer is that cohesiueness is the group therapy analogue to relation­ ship in individual therapy. First, keep in mind that a vast body of research on individual psychotherapy demonstrates that a good therapist-client rela­ tionship is essential for a positive outcome. Is it also true that a good ther­ apy relationship is essential in group therapy? Here again, the literature leaves little doubt that "relationship" is germane to positive outcome in group therapy. But relationship in group therapy is a far more complex con­ cept than relationship in individual therapy. After all, there are only two people in the individual therapy transaction, whereas a number of individ­ uals, generally six to ten, work together in group therapy. It is not enough to say that a good relationship is necessary for successful group therapy­ we must specify which relationship: The relationship between the client and the group therapist (or therapists if there are co-leaders)? Or between the group member and other members? Or perhaps even between the individ­ ual and the "group" taken as a whole?

Over the past forty years, a vast number of controlled studies of psy­ chotherapy outcome have demonstrated that the average person who re­ ceives psychotherapy is significantly improved and that the outcome from group therapy is virtually identical to that of individual therapy. 1 Further­ more there is evidence that certain clients may obtain greater benefit from group therapy than from other approaches, particularly clients dealing with stigma or social isolation and those seeking new coping skills.2

The evidence supporting the effectiveness of group psychotherapy is so compelling that it prompts us to direct our attention toward another

53

54 GROUP COHESIVENESS

question: What are the necessary conditions for effective psychotherapy? After all, not all psychotherapy is successful. In fact, there is evidence that treatment may be for better or for worse-although most therapists help their clients, some therapists make some clients worse.3 Why? What makes for successful therapy? Although many factors are involved, a proper therapeutic relationship is a sine qua non for effective therapy out­ come.4 Research evidence overwhelmingly supports the conclusion that successful therapy-indeed even successful drug therapy-is mediated by a relationship between therapist and client that is characterized by trust, warmth, empathic understanding, and acceptance.5 Although a positive therapeutic alliance is common to all effective treatments, it is not easily or routinely established. Extensive therapy research has focused on the na­ ture of the therapeutic alliance and the specific interventions required to achieve and maintain it. 6

Is the quality of the relationship related to the therapist's school of conviction? The evidence says, "No." Experienced and effective clinicians from different schools (Freudian, nondirective, experiential, gestalt, rela­ tional, interpersonal, cognitive-behavioral, psychodrama) resemble one another (and differ from nonexperts in their own school) in their concep­ tion of the ideal therapeutic relationship and in the relationship they themselves establish with their clients.7

Note that the engaged, cohesive therapeutic relationship is necessary in all psychotherapies, even the so-called mechanistic approaches-cognitive, behavioral, or systems-oriented forms of psychotherapy.8 A recent sec­ ondary analysis of a large comparative psychotherapy trial, the National Institute of Mental Health's (NIMH) Treatment of Depression Collabora­ tive Research Program, concluded that successful therapy, whether it was cognitive-behavioral therapy or interpersonal therapy, required "the pres­ ence of a positive attachment to a benevolent, supportive, and reassuring authority figure." 9 Research has shown that the client-therapist bond and the technical elements of cognitive therapy are synergistic: a strong and positive bond in itself disconfirms depressive beliefs and facilitates the work of modifying cognitive distortions. The absence of a positive bond renders technical interventions ineffective or even harmful.10

As noted, relationship plays an equally critical role in group psy­ chotherapy. But the group therapy analogue of the client-therapist rela­ tionship in individual therapy must be a broader concept, encompassing the individual's relationship to the group therapist, to the other group members, and to the group as a whole. t At the risk of courting semantic confusion, I refer to all of these relationships in the group with the term "group cohesiveness." Cohesiveness is a widely researched basic property of groups that has been explored in several hundred research articles. Un-

55 Group Cohesiveness

fortunately, there is little cohesion in the literature, which suffers from the use of different definitions, scales, subjects, and rater perspectives. 11

In general, however, there is agreement that groups differ from one an­ other in the amount of "groupness" present. Those with a greater sense of solidarity, or "we-ness," value the group more highly and will defend it against internal and external threats. Such groups have a higher rate of at­ tendance, participation, and mutual support and will defend the group standards much more than groups with less esprit de corps. Nonetheless it is difficult to formulate a precise definition. A recent comprehensive and thoughtful review concluded that cohesiveness "is like dignity: everyone can recognize it but apparently no one can describe it, much less measure it."12 The problem is that cohesiveness refers to overlapping dimensions. On the one hand, there is a group phenomenon-the total esprit de corps; on the other hand, there is the individual member cohesiveness (or, more strictly, the individual's attraction to the group}. 13

In this book, cohesiveness is broadly defined as the result of all the forces acting on all the members such that they remain in the group,14 or, more simply, the attractiveness of a group for its members. 15 Members of a cohesive group feel warmth and comfort in the group and a sense of be­ longingness; they value the group and feel in turn that they are valued, ac­ cepted, and supported by other members. 16t

Esprit de corps and individual cohesiveness are interdependent, and group cohesiveness is often computed simply by summing the individual members' level of attraction to the group. Newer methods of measuring group cohesiveness from raters' evaluations of group climate make for greater quantitative precision, but they do not negate the fact that group cohesiveness remains a function and a summation of the individual mem­ bers' sense of belongingness.17 Keep in mind that group members are dif­ ferentially attracted to the group and that cohesiveness is not fixed-once achieved, forever held-but instead fluctuates greatly during the course of the group. 18 Early cohesion and engagement is essential for the group to

encompass the more challenging work that comes later in the group's de­ velopment, as more conflict and discomfort emerges. 19 Recent research has also differentiated between the individual's sense of belonging and his or her appraisal of how well the entire group is working. It is not uncom­ mon for an individual to feel "that this group works well, but I'm not part of it."20 It is also possible for members (for example eating disorder clients) to value the interaction and bonding in the group yet be funda­ mentally opposed to the group goal.21

Before leaving the matter of definition, I must point out that group co­ hesiveness is not only a potent therapeutic force in its own right. It is a precondition for other therapeutic factors to function optimally. When, in

56 GROUP COHESIVENESS

individual therapy, we say that it is the relationship that heals, we do not mean that love or loving acceptance is enough; we mean that an ideal therapist-client relationship creates conditions in which the necessary risk taking, catharsis, and intrapersonal and interpersonal exploration may unfoid. It is the same for group therapy: cohesiveness is necessary for other group therapeutic factors to operate.

THE IMPORTANCE OF GROUP COHESIVENESS

Although I have discussed the therapeutic factors separately, they are, to a great degree, interdependent. Catharsis and universality, for example, are not complete processes. It is not the sheer process of ventilation that is important; it is not only the discovery that others have problems similar to one's own and the ensuing disconfirmation of one's wretched unique­ ness that are important. It is the affective sharing of one's inner world and then the acceptance by others that seem of paramount importance. To be accepted by others challenges the client's belief that he or she is basically repugnant, unacceptable, or unlovable. The need for belonging is innate in us all. Both affiliation within the group and attachment in the individ­ ual setting address this need. 22 Therapy groups generate a positive, self­ reinforcing loop: trust-self-disclosure-empathy-acceptance-trust.23 The group will accept an individual, provided that the individual adheres to the group's procedural norms, regardless of past life experiences, trans­ gressions, or social failings. Deviant lifestyles, history of prostitution, sexual perversion, heinous criminal offenses-all of these can be accepted by the therapy group, so long as norms of nonjudgmental acceptance and inclusiveness are established early in the group.

For the most part, the disturbed interpersonal skills of our clients have limited their opportunities for effective sharing and acceptance in inti­ mate relationships. Furthermore, some members are convinced that their abhorrent impulses and fantasies shamefully bar them from social inter­ action.t I have known many isolated clients for whom the group repre­ sented their only deeply human contact. After just a few sessions, they have a stronger sense of being at home in the group than anywhere else. Later, even years afterward, when most other recollections of the group have faded from memory, they may still remember the warm sense of be­ longing and acceptance.

As one successful client looking back over two and a half years of ther­ apy put it, "The most important thing in it was just having a group there, people that I could always talk to, that wouldn't walk out on me. There was so much caring and hating and loving in the group, and I was a part of it. I'm better now and have my own life, but it's sad to think that the group's not there anymore."

57 The Importance of Group Cohesiveness

Furthermore, group members sec that they are not just passive benefi­ ciaries of group cohesion, they also generate that cohesion, creating durable relationships-perhaps for the first time in their lives. One group member commented that he had always attributed his aloneness to some unidentified, intractable, repugnant character failing. It was only after he stopped missing meetings regularly because of his discouragement and sense of futility that he discovered the responsibility he exercised for his own aloneness: relationships do not inevitably wither-his had been doomed largely by his choice to neglect them.

Some individuals internalize the group: "It's as though the group is sit­ ting on my shoulder, watching me. I'm forever asking, 'What would the group say about this or that?"' Often therapeutic changes persist and are consolidated because, even years later, the members are disinclined to let the group down.24

Membership, acceptance, and approval in various groups are of the ut­ most importance in the individual's developmental sequence. The impor­ tance of belonging to childhood peer groups, adolescent cliques, sororities or fraternities, or the proper social "in" group can hardly be overestimated. Nothing seems to be of greater importance for the self-esteem and well­ being of the adolescent, for example, than to be included and accepted in some social group, and nothing is more devastating than exclusion.25

Most of our clients, however, have an impoverished group history; they have never been valuable and integral to a group. For these individuals, the sheer successful negotiation of a group experience may in itself be cura­ tive. Belonging in the group raises self-esteem and meets members' de­ pendency needs but in ways that also foster responsibility and autonomy, as each member contributes to the group's welfare and internalizes the at­ mosphere of a cohesive group.26

Thus, in a number of ways, members of a therapy group come to mean a great deal to one another. The therapy group, at first perceived as an ar­ tificial group that does not count, may in fact come to count very much. I have known groups whose members experience together severe depres­ sions, psychoses, marriage, divorce, abortions, suicide, career shifts, shar­ ing of innermost thoughts, and incest (sexual activity among the group members). I have seen a group physically carry one of its members to the hospital and seen many groups mourn the death of members. I have seen members of cancer support groups deliver eulogies at the funeral of a fallen group member. Relationships are often cemented by moving or haz­ ardous adventures. How many relationships in life are so richly layered?

Evidence Empirical evidence for the impact of group cohesiveness is not as extensive or as systematic as research documenting the importance of relationship

58 GROUP COHESIVENESS

in individual psychotherapy. Studying the effect of cohesiveness is more complex 27 because it involves research on variables closely related to co­ hesion such as group climate (the degree of engagement, avoidance, and conflict in the group) 28 and alliance (the member-therapist relationship). 29

The results of the research from all these perspectives, however, point to the same conclusion: relationship is at the heart of good therapy. This is no less important in the era of managed care and third-party oversight than it was in the past. In fact, the contemporary group therapist has an even larger responsibility to safeguard the therapeutic relationship from external intrusion and control. 30

I now turn to a survey of the relevant research on cohesion. (Readers who are less interested in research methodology may wish to proceed di­ rectly to the summary section.)

• In an early study of former group psychotherapy clients in which members' explanations of the therapeutic factors in their therapy were transcribed and categorized, investigators found that more than half considered mutual support the primary mode of help in group therapy. Clients who perceived their group as cohesive at­ tended more sessions, experienced more social contact with other members, and felt that the group had been therapeutic. Improved clients were significantly more likely to have felt accepted by the other members and to mention particular individuals when queried about their group experience.31

• In 1970, I reported a study in which successful group therapy clients were asked to look back over their experience and to rate, in order of effectiveness, the series of therapeutic factors I describe in this book.32 Since that time, a vast number of studies using analogous de­ signs have generated considerable data on clients' views of what as­ pects of group therapy have been most useful. I will examine these results in depth in the next chapter; for now, it is sufficient to note that there is a strong consensus that clients regard group cohesiveness as an extremely important determinant of successful group therapy.

• In a six-month study of two long-term therapy groups,33 observers rated the process of each group session by scoring each member on five variables: acceptance, activity, desensitivity, abreaction, and im­ provement. Weekly self-ratings were also obtained from each member. Both the research raters and group members considered "acceptance" to be the variable most strongly related to improvement.

• Similar conclusions were reached in a study of forty-seven clients in twelve psychotherapy groups. Members' self-perceived personality change correlated significantly with both their feelings of involve­ ment in the group and their assessment of total group cohesiveness.34

59 The Importance of Group Cohesiveness

• My colleagues and I evaluated the one-year outcome of all forty clients who had started therapy in five outpatient groups.35 Outcome was then correlated with variables measured in the first three months of therapy. Positive outcome in therapy significantly corre­ lated with only two predictor variables: group cohesiveness36 and general popularity-that is, clients who, early in the course of ther­ apy, were most attracted to the group (high cohesiveness) and who were rated as more popular by the other group members at the sixth and the twelfth weeks had a better therapy outcome at the fiftieth week. The popularity finding, which in this study correlated even more positively with outcome than did cohesiveness, is, as I shall dis­ cuss shortly, relevant to group cohesiveness and sheds light on the mechanism through which group cohesiveness mediates change.

• The same findings emerge in· more structured groups. A study of fifty-one clients who attended ten sessions of behavioral group ther­ apy demonstrated that "attraction to the group" correlated signifi­ cantly with improved self-esteem and inversely correlated with the group dropout rate.37

• The quality of intermember relationships has also been well docu­ mented as an essential ingredient in T-groups (also called sensitivity­ training, process, encounter, or experiential groups; see chapter 16). A rigorously designed study found a significant relationship between the quality of intermember relationships and outcome in a T-group of eleven subjects who met twice a week for a total of sixty-four hours.38 The members who entered into the most two-person mutu­ ally therapeutic relationships showed the most improvement during the course of the group. 39 Furthermore, the perceived relationship with the group leader was unrelated to the extent of change.

• My colleagues M. A. Lieberman, M. Miles, and I conducted a study of 210 subjects in eighteen encounter groups, encompassing ten ide­ ological schools (gestalt, transactional analysis, T-groups, Synanon, personal growth, Esalen, psychoanalytic, marathon, psychodrama, encounter tape).40 (See chapter 16 for a detailed discussion of this project.) Cohesiveness was assessed in several ways and correlated with outcome.41 The results indicated that attraction to the group is indeed a powerful determinant of outcome. All methods of deter­ mining cohesiveness demonstrated a positive correlation between cohesiveness and outcome. A member who experienced little sense of belongingness or attraction to the group, even measured early in the course of the sessions, was unlikely to benefit from the group and, in fact, was likely to have a negative outcome. Furthermore, the groups with the higher overall levels of cohesiveness had a signifi­ cantly better total outcome than groups with low cohesiveness.

60 GROUP COHESIVENESS

• Another large study (N = 393) of experiential trammg groups yielded a strong relationship between affiliativeness (a construct that overlaps considerably with cohesion) and outcome.42

• MacKenzie and Tschuschke, studying twenty clients in long-term in­ patient groups, differentiated members' personal "emotional relat­ edness to the group" from their appraisal of "group work" as a whole. The individual's personal sense of belonging correlated with future outcome, whereas the total group work scales did not.43

• S. Budman and his colleagues developed a scale to measure cohe­ siveness via observations by trained raters of videotaped group ses­ sions. They studied fifteen therapy groups and found greater reductions in psychiatric symptoms and improvement in self-esteem in the most cohesively functioning groups. Group cohesion that was evident early-within the first thirty minutes of each session-pre­ dicted better outcome.44

• A number of other studies have examined the role of the relationship between the client and the group leader in group therapy. Marziali and colleagues45 examined group cohesion and the client-group leader relationship in a thirty-session manualized interpersonal ther­ apy group of clients with borderline personality disorder. Cohesion and member-leader relationship correlated strongly, supporting Budman's findings,46 and both positively correlated with outcome. However, the member-group leader relationship measure was a more powerful predictor of outcome. The relationship between client and therapist may be particularly important for clients who have volatile interpersonal relationships and with whom the therapist serves an important containing function.

• In a study of a short-term structured cognitive-behavioral therapy group for social phobia47 the relationship with the therapist deep­ ened over the twelve weeks of treatment and correlated positively with outcome, but cohesion was static and not related to outcome. In this study the group was a setting for therapy and not an agent of therapy. lntermember bonds were not cultivated by the therapists, leading the authors to conclude that in highly structured groups, what matters most is the client-therapist collaboration around the therapy tasks.48

• A study of thirty-four clients with depression and social isolation treated in a twelve-session interactional problem-solving group re­ ported that clients who described experiencing warmth and positive regard from the group leader had better therapy outcomes. The op­ posite also held true. Negative therapy outcomes were associated with negative client-group leader relationships. This correlative

61 The Importance of Group Cohesiueness

study does not address cause and effect, however: Are clients better liked by their therapist because they do well in therapy, or does being well liked promote more well-being and effort?49

• Outcomes in brief intensive American Group Psychotherapy Associ­ ation Institute training groups were influenced by higher levels of en­ gagement.50 Positive outcomes may well be mediated by group engagement that fosters more interpersonal communication and self-disclosure.51

Summary

I have cited evidence that group members value deeply the acceptance and support they receiue from their therapy group. Self-perceived ther­ apy outcome is positively correlated with attraction to the group. Highly cohesive groups have a better overall outcome than groups with low esprit de corps. Both emotional connectedness and the experience of group effectiveness contribute to group cohesiveness. Individuals with positive outcomes have had more mutually satisf_,ving relationships with other members. Highly cohesive groups have greater levels of self­ disclosure. For some clients and some groups ( especially highly struc­ tured groups) the relationship with the leader may be the essential factor. A strong therapeutic relationship may not guarantee a positive outcome, but a poor therapeutic relationship will certainly not result in an effective treatment.

The presence of cohesion early in each session as well as in the early sessions of the group correlates with positive outcomes. It is critical that groups become cohesive and that leaders be alert to each member's per­ sonal experience of the group and address problems with cohesion quickly. Positive client outcome is also correlated with group popularity, a variable closely related to group support and acceptance. Although ther­ apeutic change is multidimensional, these findings taken together strongly support the contention that group cohesiveness is an important determi­ nant of positive therapeutic outcome.

In addition to this direct evidence, there is considerable indirect evi­ dence from research with other types of groups. A plethora of studies demonstrate that in laboratory task groups, high levels of group cohe­ siueness produce many results that may be considered intervening therapy outcome factors. For example, group cohesiveness results in better group attendance, greater participation of members, greater influenceability of members, and many other effects. I will consider these findings in detail shortly, as I discuss the mechanism by which cohesiveness fosters thera­ peutic change.

62 GROUP COHESIVENESS

MECHANISM OF ACTION

How do group acceptance, group support, and trust help troubled individ­ uals? Surely there must be more to it than simple support or acceptance; therapists learn early in their careers that love is not enough. Although the quality of the therapist-client relationship is crucial, the therapist must do more than simply relate warmly and honestly to the client.52 The thera­ peutic relationship creates favorable conditions for setting other processes in motion. What other processes? And how are they important?

Carl Rogers's deep insights into the therapeutic relationship are as rel­ evant today as they were nearly fifty years ago. Let us start our investiga­ tion by examining his views about the mode of action of the therapeutic relationship in individual therapy. In his most systematic description of the process of therapy, Rogers states that when the conditions of an ideal therapist-client relationship exist, the following characteristic process is set into motion:

1. The client is increasingly free in expressing his feelings. 2. He begins to test reality and to become more discriminatory in his

feelings and perceptions of his environment, his self, other persons, and his experiences.

3. He increasingly becomes aware of the incongruity between his expe­ riences and his concept of self.

4. He also becomes aware of feelings that have been previously denied or distorted in awareness.

5. His concept of self, which now includes previously distorted or de­ nied aspects, becomes more congruent with his experience.

6. He becomes increasingly able to experience, without threat, the therapist's unconditional positive regard and to feel an uncondi­ tional positive self-regard.

7. He increasingly experiences himself as the focus of evaluation of the nature and worth of an object or experience.

8. He reacts to experience less in terms of his perception of others' evaluation of him and more in terms of its effectiveness in enhanc­ ing his own development. 53

Central to Rogers's views is his formulation of an actualizing tendency, an inherent tendency in all life to expand and to develop itself-a view stretching back to early philosophic views and clearly enunciated a cen­ tury ago by Nietzsche.54 It is the therapist's task to function as a facilita­ tor and to create conditions favorable for self-expansion. The first task of the individual is self-exploration: the examination of feelings and experi­ ences previously denied awareness.

63 Mechanism of Action

This task is a ubiquitous stage in dynamic psychotherapy. Horney, for example, emphasized the individual's need for self-knowledge and self-re­ alization, stating that the task of the therapist is to remove obstacles in the path of these autonomous processes.55 Contemporary models recog­ nize the same principle. Clients often pursue therapy with a plan to dis­ confirm pathogenic beliefs that obstruct growth and development.56 In other words, there is a built-in inclination to growth and self-fulfillment in all individuals. The therapist does not have to inspirit cliems with these qualities (as if we could!). Instead, our task is to remove the obstacles that block the process of growth. And one way we do this is by creating an ideal therapeutic atmosphere in the therapy group. A strong bond be­ tween members not only directly disconfirms one's unworthiness, it also generates greater willingness among clients to self-disclose and take inter­ personal risks. These changes help deactivate old, negative beliefs about the self in relation to the world.57

There is experimental evidence that good rapport in individual therapy and its equivalent (cohesiveness) in group therapy encourage the client to participate in a process of reflection and personal exploration. For exam­ ple, Truax,58 studying forty-five hospitalized patients in three heteroge­ neous groups, demonstrated that participants in cohesive groups were significantly more inclined to engage in deep and extensive self-explo­ ration. 59 Other research demonstrates that high cohesion is closely related to high degrees of intimacy, risk taking, empathic listening, and feed­ back.60 The group members' recognition that their group is working well at the task of interpersonal learning produces greater cohesion in a posi­ tive and self-reinforcing loop. 61 Success with the group task strengthens the emotional bonds in the group.

Perhaps cohesion is vital because many of our clients have not had the benefit of ongoing solid peer acceptance in childhood. Therefore they find validation by other group members a new and vital experience. Further­ more, acceptance and understanding among members may carry greater power and meaning than acceptance by a therapist. Other group mem­ bers, after all, do not have to care, or understand. They're not paid for it; it's not their "job."62

The intimacy developed in a group may be seen as a counterforce in a technologically driven culture that, in all ways-socially, professionally, residentially, recreationally-inexorably dehumanizes relationships.63 In a world in which traditional boundaries that maintain relationships are increasingly permeable and transient, there is a greater need than ever for group belonging and group identity.64 The deeply felt human expe­ rience in the group may be of great value to the individual, Rogers be­ lieves. Even if it creates no visible carryover, no external change in behavior, group members may still experience a more human, richer part

64 GROUP COHESIVENESS

of themselves and have this as an internal reference point. This last point is worth emphasizing, for it is one of those gains of therapy--especially group therapy-that enrich one's interior life and yet may not, at least for a long period of time, have external behavioral manifestations and thus may elude measurement by researchers and consideration by man­ aged health care administrators, who determine how much and what type of therapy is indicated.

Group members' acceptance of self and acceptance of other members are interdependent; not only is self-acceptance basically dependent on ac­ ceptance by others, but acceptance of others is fully possible only after one can accept oneself. This principle is supported by both clinical wis­ dom and research.65 Members of a therapy group may experience consid­ erable self-contempt and contempt for others. A manifestation of this feeling may be seen in the client's initial refusal to join "a group of nuts" or reluctance to become closely involved with a group of pained individu­ als for fear of being sucked into a maelstrom of misery. A particularly evocative response to the prospect of group therapy was given by a man in his eighties when he was invited to join a group for depressed elderly men: it was useless, he said, to waste time watering a bunch of dead trees-his metaphor for the other men in his nursing home.66

In my experience, all individuals seeking assistance from a mental health professional have in common two paramount difficulties: (1) estab­ lishing and maintaining meaningful interpersonal relationships, and (2) maintaining a sense of personal worth (self-esteem). It is hard to discuss these two interdependent areas as separate entities, but since in the pre­ ceding chapter I dwelled more heavily on the establishment of interper­ sonal relationships, I shall now turn briefly to self-esteem.

Self-esteem and public esteem are highly interdependent.67 Self-esteem refers to an individual's evaluation of what he or she is really worth, and is indissolubly linked to that person's experiences in prior social relation­ ships. Recall Sullivan's statement: "The self may be said to be made up of reflected appraisals. "68 In other words, during early development, one's perceptions of the attitudes of others toward oneself come to determine how one regards and values oneself. The individual internalizes many of these perceptions and, if they are consistent and congruent, relies on these internalized evaluations for some stable measure of self-worth.

But, in addition to this internal reservoir of self-worth, people are, to a greater or lesser degree, always concerned and influenced by the current evaluations of others--especially the evaluation provided by the groups to which they belong. Social psychology research supports this clinical un­ derstanding: the groups and relationships in which we take part become incorporated in the self. 69 One's attachment to a group is multidimen­ sional. It is shaped both by the member's degree of confidence in his at-

65 Mechanism of Action

tractiveness to the group-am I a desirable member?-and the member's relative aspiration for affiliation-do I want to belong?

The influence of public esteem-that is, the group's evaluation-on an individual depends on several factors: how important the person feels the group to be; the frequency and specificity of the group's communications to the person about that public esteem; and the salience to the person of the traits in question. (Presumably, considering the honest and intense self-disclosure in therapy groups, the salience is very great indeed, since these traits are close to a person's core identity.) In other words, the more the group matters to the person, and the more that person subscribes to the group values, the more he or she will be inclined to value and agree with the group judgment.70 This last point has much clinical relevance. The more attracted an individual is to the group, the more he or she will respect the judgment of the group and will attend to and take seriously any discrepancy between public esteem and self-esteem. A discrepancy between the two will create a state of dissonance, which the individual will attempt to correct.

Let us suppose this discrepancy veers to the negative side-that is, the group's evaluation of the individual is less than the individual's self­ evaluation. How to resolve that discrepancy? One recourse is to deny or distort the group's evaluation. In a therapy group, this is not a positive de­ velopment, for a vicious circle is generated: the group, in the first place, evaluates the member poorly because he or she fails to participate in the group task (which in a therapy group consists of active exploration of one's self and one's relationships with others). Any increase in defensive­ ness and communicational problems will only further lower the group's esteem of that particular member. A common method used by members to resolve such a discrepancy is to devalue the group-emphasizing, for example, that the group is artificial or composed of disturbed individuals, and then comparing it unfavorably to some anchor group (for example, a social or occupational group) whose evaluation of the member is differ­ ent. Members who follow this sequence (for example, the group deviants described in chapter 8) usually drop out of the group.

Toward the end of a successful course of group therapy, one group member reviewed her early recollections of the group as follows: "For the longest time I told myself you were all nuts and your feedback to me about my defensiveness and inaccessibility was ridiculous. I wanted to quit-I've done that before many times, but I felt enough of a connec­ tion here to decide to stay. Once I made that choice I started to tell my­ self that you cannot all be wrong about me. That was the turning point in my therapy." This is an example of the therapeutic method of resolv­ ing the discrepancy for the individual: that is, to raise one's public es­ teem by changing those behaviors and attitudes that have been criticized

66 GROUP COHESIVENESS

by the group. This method is more likely if the individual is highly at­ tracted to the group and if the public esteem is not too much lower than the self-esteem.

But is the use of group pressure to change individual behavior or atti­ tudes a form of social engineering? Is it not mechanical? Does it not ne­ glect deeper levels of integration? Indeed, group therapy does employ behavioral principles; psychotherapy is, in all its variants, basically a form of learning. Even the most nondirective therapists use, at an unconscious level, operant conditioning techniques: they signal desirable conduct or attitudes to clients, whether explicitly or subtly.71

This process does not suggest that we assume an explicit behavioral, mechanistic view of the client, however. Aversive or operant conditioning of behavior and attitudes is, in my opinion, neither feasible nor effective when applied as an isolated technique. Although clients often report last­ ing improvement after some disabling complaint is remedied by behav­ ioral therapy techniques, close inspection of the process invariably reveals that important interpersonal relationships have been affected. Either the therapist-client relationship in the behavioral and cognitive therapies has been more meaningful than the therapist realized (and research evidence substantiates this) ,72 or some important changes, initiated by the sympto­ matic relief, have occurred in the client's social relationships that have served to reinforce and maintain the client's improvement. Again, as I have stressed before, all the therapeutic factors are intricately interdepen­ dent. Behavior and attitudinal change, regardless of origin, begets other changes. The group changes its evaluation of a member; the member feels more self-satisfied in the group and with the group itself; and the adaptive spiral described in the previous chapter is initiated.

A far more common occurrence in a psychotherapy group is a discrep­ ancy in the opposite direction: the group's evaluation of a member is higher than the member's self-evaluation. Once again, the member is placed in a state of dissonance and once again will attempt to resolve the discrepancy. What can a member in that position do? Perhaps the person will lower the public esteem by revealing personal inadequacies. However, in therapy groups, this behavior has the paradoxical effect of raising pub­ lic esteem-disclosure of inadequacies is a valued group norm and en­ hances acceptance by the group. Another possible scenario, desirable therapeutically, occurs when group members reexamine and alter their low level of self-esteem. An illustrative clinical vignette will flesh out this formulation:

• Marietta, a thirty-four-year-old housewife with an emotionally im­ poverished background, sought therapy because of anxiety and guilt stemming from a series of extramarital affairs. Her self-esteem was ex-

67 Mechanism of Action

ceedingly low; nothing escaped her self-excoriation: her physical ap­ pearance, her intelligence, her speech, her unimaginativeness, her func­ tioning as a mother and a wife. Although she receiued solace from her religious affiliation, it was a mixed blessing because she felt too un­ worthy to socialize with the church people in her community. She mar­ ried a man she considered repugnant but nonetheless a good man-certainly good enough for her. Only in her sexual affairs, partic­ ularly when she had them with several men at once, did she seem to come alive-to feel attractiue, desirable, and able to give something of herself that seemed of value to others. However, this behavior clashed with her religious convictions and resulted in considerable anxiety and further self-derogation.

Viewing the group as a social microcosm, the therapist soon noted characteristic trends in Marietta's group behauior. She spoke often of the guilt issuing from her sexual behavior, and for many hours the group struggled with all the titillating ramifications of her predica­ ment. At all other times in the group, however, she disengaged and of­ fered nothing. She related to the group as she did to her social environment. She could belong to it, but she could not really relate to the other people: the only thing of real interest she felt she could offer was her genitals.

Over time in the group she began to respond and to question others and to offer warmth, support, and feedback. She found other, nonsex­ ual, aspects of herself to disclose and spoke openly of a broad array of her life concerns. Soon she found herself increasingly valued by the other members. She gradually reexamined and euentually discon-fi,rmed her belief that she had little of value to offer. The discrepancy between her public esteem and her self-esteem widened (that is, the group ual­ ued her more than she regarded herself), and soon she was forced to en­ tertain a more realistic and positive view of herself Gradually, an adaptive spiral ensued: she began to establish meaningful nonsexual re­ lationships both in and out of the group and these, in turn, further en­ hanced her self-esteem.

The more therapy disconfirms the client's negative self-image through new relational experience, the more effective therapy will be.73

Self-Esteem, Public Esteem, and Therapeutic Change: Evidence

Group therapy research has not specifically investigated the relationship between public esteem and shifts in self-esteem. However, an interesting finding from a study of experiential groups (see chapter 16) was that members' self-esteem decreased when public esteem decreased.74 (Public esteem is measured by sociometric data, which involves asking members

68 GROUP COHESIVENESS

to rank-order one another on several variables.) Researchers also discov­ ered that the more a group member underestimated his or her public es­ teem, the more acceptable that member was to the other members. In other words, the ability to face one's deficiencies, or even to judge oneself a little harshly, increases one's public esteem. Humility, within limits, is far more adaptable than arrogance.

It is also interesting to consider data on group popularity, a variable closely related to public esteem. The group members considered most popular by other members after six and twelve weeks of therapy had sig­ nificantly better therapy outcomes than the other members at the end of one year.75 Thus, it seems that clients who have high public esteem early in the course of a group are destined to have a better therapy outcome.

What factors seem to be responsible for the attainment of popularity in therapy groups? Three variables, which did not themselves correlate with outcome, correlated significantly with popularity:

1. Previous self-disclosure.76

2. Interpersonal compatibility:77 individuals who (perhaps fortu­ itously) have interpersonal needs that happen to blend well with those of the other group members become popular in the group.

3. Other sociometric measures; group members who were often chosen as leisure companions and worked well with colleagues became pop­ ular in the group. A clinical study of the most popular and least pop­ ular members revealed that popular members tended to be young, well-educated, intelligent, and introspective. They filled the leader­ ship vacuum that occurs early in the group when the therapist de­ clines to assume the traditional leader role.78

The most unpopular group members were rigid, moralistic, nonintro­ spective, and least involved in the group task. Some were blatantly de­ viant, attacking the group and isolating themselves. Some schizoid members were frightened of the group process and remained peripheral. A study of sixty-six group therapy members concluded that the less pop­ ular members (that is, those viewed less positively by other members) were more inclined to drop out of the group.79

Social psychology researchers have also investigated the attributes that confer higher social status in social groups. The personality attribute of extraversion (measured by a personality questionnaire, the NEO-Pl) 80 is a very strong predictor of popularity.81 Extraversion connotes the traits of active and energetic social engagement, that is, a person who is upbeat and emotionally robust. Depue's neurobiological research82 suggests that such individuals invite others to approach them. The promise of the ex­ travert's welcome response rewards and reinforces engagement.

69 Mechanism of Action

The Lieberman, Yalom, and Miles encounter group study corroborated these conclusions.83 Sociometric data revealed that the members with the more positive outcomes were influential and engaged in behavior in close harmony with the encounter group values of risk taking, spontaneity, openness, self-disclosure, expressivity, group facilitation, and support. Ev­ idence has emerged from both clinical and social-psychological small­ group research demonstrating that the members who adhere most closely to group norms attain positions of popularity and influence. 84 Members who help the group achieve its tasks are awarded higher status. 85

To summarize: Members who are popular and influential in therapy groups have a higher likelihood of changing. They attain popularity and in­ fluence in the group by virtue of their active participation, self-disclosure, self-exploration, emotional expression, nondefensiveness, leadership, in­ terest in others, and support of the group.

It is important to note that the individual who adheres to the group norms not only is rewarded by increased public esteem within the group but also uses those same social skills to deal more effectively with inter­ personal problems outside the group. Thus, increased popularity in the group acts therapeutically in two ways: by augmenting self-esteem and by reinforcing adaptive social skills. The rich get richer. The challenge in group therapy is helping the poor get richer as well.

Group Cohesiveness and Group Attendance

Continuation in the group is obviously a necessary, though not a suffi­ cient, prerequisite for successful treatment. Several studies indicate that clients who terminate early in the course of group therapy receive little benefit. 86 In one study, over fifty clients who dropped out of long-term therapy groups within the first twelve meetings reported that they did so because of some stress encountered in the group. They were not satisfied with their therapy experience and they did not improve; indeed, many of these clients felt worse.87 Clients who remain in the group for at least sev­ eral months have a high likelihood (85 percent in one study) of profiting from therapy. 88

The greater a member's attraction to the group, the more inclined that person will be to stay in therapy groups as well as in encounter groups, laboratory groups (formed for some research purpose), and task groups (established to perform some designated task). 89 The Lieberman, Yalom, and Miles encounter group study discovered a high correlation between low cohesiveness and eventual dropping out from the group. 90 The dropouts had little sense of belongingness and left the group most often because they felt rejected, attacked, or unconnected.

The relationship between cohesiveness and maintenance of member­ ship has implications for the total group as well. Not only do the least

70 GROUP COHESIVENESS

cohesive members terminate membership and fail to benefit from therapy, but noncohesive groups with high member turnover prove to be less ther­ apeutic for the remaining members as well. Clients who drop out chal­ lenge the group's sense of worth and effectiveness.

Stability of membership is a necessary condition for effective short­ and long-term interactional group therapy. Although most therapy groups go through an early phase of instability during which some members drop out and replacements are added, the groups thereafter settle into a long, stable phase in which much of the solid work of therapy occurs. Some groups seem to enter this phase of stability early, and other groups never achieve it. Dropouts at times beget other dropouts, as other clients may terminate soon after the departure of a key member. In a group therapy follow-up study, clients often spontaneously underscored the importance of membership stability. 91

In chapter 15, I will discuss the issue of cohesiveness in groups led in clinical settings that preclude a stable long-term membership. For exam­ ple, drop-in crisis groups or groups on an acute inpatient ward rarely have consistent membership even for two consecutive meetings. In these clini­ cal situations, therapists must radically alter their perspectives on the life development of the group. I believe, for example, that the appropriate life span for the acute inpatient group is a single session. The therapist must strive to be efficient and to offer effective help to as many members as pos­ sible during each single session.

Brief therapy groups pay a particularly high price for poor attendance, and therapists must make special efforts to increase cohesiveness early in the life of the group. These strategies (including strong pregroup prepara­ tion, homogeneous composition, and structured interventions) 92 will be discussed in chapter 15.

Group Cohesiveness and the Expression of Hostility

It would be a mistake to equate cohesiveness with comfort. Although co­ hesive groups may show greater acceptance, intimacy, and understanding, there is evidence that they also permit greater development and expres­ sion of hostility and conflict. Cohesive groups have norms (that is, un­ written rules of behavior accepted by group members) that encourage open expression of disagreement or conflict alongside support. In fact, unless hostility can be openly expressed, persistent covert hostile attitudes may hamper the development of cohesiveness and effective interpersonal learning. Unexpressed hostility simply smolders within, only to seep out in many indirect ways, none of which facilitates the group therapeutic process. It is not easy to continue communicating honestly with someone you dislike or even hate. The temptation to avoid the other and to break off communication is very great; yet when channels of communication

71 Mechanism of Action

are closed, so are any hopes for conflict resolution and for personal growth.

This is as true on the megagroup-even the national-level as on the dyadic. The Robbers' Cave experiment, a famed research project con­ ducted long ago, in the infancy of group dynamics research," offers ex­ perimental evidence still relevant for contemporary clinical work. 93 A camp of well-adjusted eleven-year-old boys was divided at the outset into two groups that were placed in competition with each other in a series of contests. Soon each group developed considerable cohesiveness as well as a deep sense of hostility toward the other group. Any meaningful com­ munication between the two groups became impossible. If, for example, they were placed in physical proximity in the dining hall, the group boundaries remained impermeable. Intergroup communication consisted of taunts, insults, and spitballs.

How to restore meaningful communication between the members of the two groups? That was the quest of the researchers. Finally they hit upon a successful strategy. Intergroup hostility was relieved only when a sense of allegiance to a single large group could be created. The re­ searchers created some superordinate goals that disrupted the small group boundaries and forced all the boys to work together in a single large group. For example, a truck carrying food for an overnight hike stalled in a ditch and could be rescued only by the cooperative efforts of all the boys; a highly desirable movie could be rented only by the pooled contri­ butions of the entire camp; the water supply was cut off and could be re­ stored only by the cooperative efforts of all campers.

The drive to belong can create powerful feelings within groups. Mem­ bers with a strong adherence to what is inside the group may experience strong pressure to exclude and devalue who and what is outside the bounds of the group. 94 It is not uncommon for individuals to develop prejudice against groups to which they cannot belong. It is therefore not surprising that hostility often emerges against members of ethnic or racial groups to which entry for outsiders may be impossible. The implication

''Dynamic is a frequently used term in the vocabulary of psychotherapy and must be defined. It has a lay and a technical meaning. It derives from the Greek dunasthi, meaning "to have power or strength." In the lay sense, then, the word evokes energy or movement (a dynamic football player or orator), but in its technical sense it refers to the idea of "forces." In individual therapy, when we speak of a client's "psychodynamics," we are referring to the various forces in conflict

within the client that result in certain configurations of experienced feelings and behavior. In common usage since the advent of Freud, the assumption is made that some of the forces in conflict with one another exist at different levels of awareness-indeed, some of them are en­ tirely out of consciousness and, through the mechanism of repression, dwell in the dynamic un­ consciousness. In group work, dynamics refers to inferred, invisible constructs or group properties (for example, cohesiveness, group pressure, scapegoating, and subgrouping) that af­ fect the overall movements of the group.

72 GROUP COHESIVENESS

for international conflict is apparent: intergroup hostility may dissolve in the face of some urgently felt worldwide crisis that only supranational co­ operation can avert: atmospheric pollution or an international AIDS epi­ demic, for example. These principles also have implications for clinical work with small groups.

lntermember conflict during the course of group therapy must be con­ tained. Above all, communication must not be ruptured, and the adver­ saries must continue to work together in a meaningful way, to take responsibility for their statements, and to be willing to go beyond name­ calling. This is, of course, a major difference between therapy groups and social groups, in which conflicts often result in the permanent rupture of relationships. Clients' descriptions of critical incidents in therapy (see chapter 2) often involve an episode in which they expressed strong nega­ tive affect. In each instance, however, the client was able to weather the storm and to continue relating (often in a more gratifying manner) to the other member.

Underlying these events is the condition of cohesiveness. The group and the members must mean enough to each other to be willing to bear the discomfort of working through a conflict. Cohesive groups are, in a sense, like families with much internecine warfare but a powerful sense of loyalty.

Several studies demonstrate that cohesiveness is positively correlated with risk taking and intensive interaction.95 Thus, cohesiveness is not syn­ onymous with love or with a continuous stream of supportive, positive statements. Cohesive groups are groups that are able to embrace conflict and to derive constructive benefit from it. Obviously, in times of conflict, cohesiveness scales that emphasize warmth, comfort, and support will temporarily gyrate; thus, many researchers have reservations about view­ ing cohesiveness as a precise, stable, measurable, unidimensional variable and consider it instead as multidimensional.96

Once the group is able to deal constructively with conflict in the group, therapy is enhanced in many ways. I have already mentioned the importance of catharsis, of risk taking, of gradually exploring previ­ ously avoided or unknown parts of oneself and recognizing that the an­ ticipated dreaded catastrophe is chimerical. Many clients are desperately afraid of anger-their own and that of others. A highly cohesive group encourages members to tolerate the pain and hurt that interpersonal learning may produce.

But keep in mind that it is the early engagement that makes such suc­ cessful working-through later possible.97 The premature expression of ex­ cess hostility before group cohesion has been established is a leading cause of group fragmentation. It is important for clients to realize that

73 Mechanism of Action

their anger is not lethal. Both they and others can and do survive an ex­ pression of their impatience, irritability, and even outright rage. For some clients, it is also important to have the experience of weathering an at­ tack. In the process, they may become better acquainted with the reasons for their position and learn to withstand pressure from others. 98

Conflict may also enhance self-disclosure, as each opponent tends to

reveal more and more to clarify his or her position. As members are able to go beyond the mere statement of position, as they begin to understand the other's experiential world, past and present, and view the other's po­ sition from their own frame of reference, they may begin to understand that the other's point of view may be as appropriate for that person as their own is for themselves. The working through of extreme dislike or hatred of another person is an experience of great therapeutic power. A clinical illustration demonstrates many of these points (another example may be found in my novel The Schopenhauer Cure) .99

• Susan, a forty-six-year-old, very proper school principal, and Jean, a twenty-one-year-old high school dropout, became locked into a vicious struggle. Susan despised Jean because of her libertine lifestyle, and what she imagined to be her sloth and promiscuity. Jean was enraged by Susan's judgmentalism, her sanctimoniousness, her embittered spin­ sterhood, her closed posture to the world. Fortunately, both women were deeply committed members of the group. (Fortuitous circum­ stances played a part here. Jean had been a core member of the group for a year and then married and went abroad for three months. Just at that time Susan became a member and, during Jean's absence, became heavily involved in the group.)

Both had had considerable past difficulty in tolerating and express­ ing anger. Over a four-month period, they interacted heavily, at times in pitched battles. For example, Susan erupted indignantly when she found out that Jean was obtaining food stamps illegally; and Jean, learning of Susan's virginity, ventured the opinion that she was a cu­ riosity, a museum piece, a mid-Victorian relic.

Much good group work was done because Jean and Susan, despite their conflict, never broke off communication. They learned a great deal about each other and eventually realized the cruelty of their mu­ tual judgmentalism. Finally, they could both understand how much each meant for the other on both a personal and a symbolic level. Jean desperately wanted Susan's approual; Susan deeply envied Jean for the freedom she had never permitted herself In the working-through process, both fully experienced their rage; they encountered and then accepted preuiously unknown parts of themselues. Ultimately, they

74 GROUP COHESIVENESS

developed an empathic understanding and then an acceptance of each other. Neither could possibly have tolerated the extreme discomfort of the confiict were it not for the strong cohesion that, despite the pain, bound them to the group.

Not only are cohesive groups more able to express hostility among members but there is evidence that they are also more able to express hos­ tility toward the leader. 100 Regardless of the personal style or skill of group leaders, the therapy group will nonetheless come, often within the first dozen meetings, to experience some degree of hostility and resent­ ment toward them. (See chapter 11 for a full discussion of this issue.) Leaders do not fulfill members' fantasized expectations and, in the view of many members, do not care enough, do not direct enough, and do not offer immediate relief. If the group members suppress these feelings of disappointment or anger, several harmful consequences may ensue. They may attack a convenient scapegoat-another member or some institution like "psychiatry" or "doctors." They may experience a smoldering irrita­ tion within themselves or within the group as a whole. They may, in short, begin to establish norms discouraging open expression of feelings. The presence of such scapegoating may be a signal that aggression is being dis­ placed away from its more rightful source-often the therapist. 101 Leaders who challenge rather than collude with group scapegoating not only safe­ guard against an unfair attack, they also demonstrate their commitment to authenticity and responsibility in relationships.

The group that is able to express negative feelings toward the therapist almost invariably is strengthened by the experience. It is an excellent ex­ ercise in direct communication and provides an important learning expe­ rience-namely, that one may express hostility directly without some ensuing irreparable calamity. It is far preferable that the therapist, the true object of the anger, be confronted than for the anger to be displaced onto some other member in the group. Furthermore, the therapist, let us pray, is far better able than a scapegoated member to withstand confrontation. The entire process is self-reinforcing; a concerted attack on the leader that is handled in a nondefensive, nonretaliatory fashion serves to increase co­ hesiveness still further.

One cautionary note about cohesion: misguided ideas about cohesion may interfere with the group task. 102 Janis coined the term "groupthink" to describe the phenomenon of "deterioration of mental efficiency, reality testing, and moral judgment that results from group pressure." 103 Group pressure to conform and maintain consensus may create a groupthink en­ vironment. This is not an alliance-based cohesion that facilitates the growth of the group members; on the contrary, it is a misalliance based on

75 Summary

naive or regressive assumptions about belonging. Critical and analytic thought by the group members needs to be endorsed and encouraged by the group leader as an essential group norm. 104 Autocratic, closed and au­ thoritarian leaders discourage such thought. Their groups are more prone to resist uncertainty, to be less reflective, and to close down exploration prematurely. 105

Group Cohesiveness and Other Therapy-Relevant Variables

Research from both therapy and laboratory groups has demonstrated that group cohesiveness has a plethora of important consequences that have obvious relevance to the group therapeutic process. 106 It has been shown, for example, that the members of a cohesive group, in contrast to the members of a noncohesive group, will:

1. Try harder to influence other group members 107

2. Be more open to influence by the other members 108

3. Be more willing to listen to others 109 and more accepting of others110

4. Experience greater security and relief from tension in the group 111

5. Participate more readily in meetings112

6. Self-disclose more113

7. Protect the group norms and exert more pressure on individuals de­ viating from the norms114

8. Be less susceptible to disruption as a group when a member termi­ nates membership115

9. Experience greater ownership of the group therapy enterprise116

SUMMARY

By definition, cohesiveness refers to the attraction that members have for their group and for the other members. It is experienced at interpersonal, intrapersonal, and intragroup levels. The members of a cohesive group are accepting of one another, supportive, and inclined to form meaningful rela­ tionships in the group. Cohesiveness is a significant factor in successful group therapy outcome. In conditions of acceptance and understanding, members will be more inclined to express and explore themselves, to be­ come aware of and integrate hitherto unacceptable aspects of self, and to relate more deeply to others. Self-esteem is greatly influenced by the client's role in a cohesive group. The social behavior required for members to be es­ teemed by the group is socially adaptive to the individual out of the group.

In addition, highly cohesive groups are more stable groups, with better attendance and less turnover. Evidence was presented to indicate that this

76 GROUP COHESIVENESS

stability is vital to successful therapy: early termination precludes benefit for the involved client and impedes the progress of the rest of the group as well. Cohesiveness favors s~lf-disclosure, risk taking, and the constructive expression of conflict in the group-phenomena that facilitate successful therapy.

What we have yet to consider are the determinants of cohesiveness. What are the sources of high and low cohesiveness? What does the thera­ pist do to facilitate the development of a highly cohesive group? These important issues will be discussed in the chapters dealing with the group therapist's tasks and techniques.