Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Systemically and emotionally-focused marital therapies : a comparative outcome study Goldman, Audrey A. 1987-10-18

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
831-UBC_1988_A2 G64.pdf [ 11.92MB ]
Metadata
JSON: 831-1.0053663.json
JSON-LD: 831-1.0053663-ld.json
RDF/XML (Pretty): 831-1.0053663-rdf.xml
RDF/JSON: 831-1.0053663-rdf.json
Turtle: 831-1.0053663-turtle.txt
N-Triples: 831-1.0053663-rdf-ntriples.txt
Original Record: 831-1.0053663-source.json
Full Text
831-1.0053663-fulltext.txt
Citation
831-1.0053663.ris

Full Text

SYSTEMICALLY AND EMOTIONALLY-FOCUSED MARITAL THERAPIES: A COMPARATIVE OUTCOME STUDY by AUDREY A. GOLDMAN B.Ed., University of British Columbia, 1960 M.Ed., University of British Columbia, 1978 A DISSERTATION SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF EDUCATION in THE FACULTY OF GRADUATE STUDIES Department of Counselling Psychology We accept this Dissertation as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA January, 1987 ® Audrey A. Goldman, 1987 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia 1956 Main Mall Vancouver, Canada Department V6T 1Y3 DE-6(3/81) ii Abstract The present study compares the effectiveness of two interventions in the treatment of marital discord: a sequentially integrated systemic intervention focusing on reframing interactional patterns and an emotionally focused intervention, focusing on accessing emotional experiences underlying interaction patterns. Forty-two couples seeking therapy were randomly assigned to one of these treatments or to a wait-list control group. Each treatment was administered in ten weekly sessions by seven experienced therapists. Through an implementation check it was determined that the treatments had been implemented in accordance with the treatment manuals. Tests of equivalence showed that the groups were equivalent on pre-test levels, demographic variables and in the strength of their working alliance with their therapists. Post-test results indicated that both treatment groups made significant gains compared to untreated controls on measures of marital adjustment, conflict resolution, target complaint reduction and goal attainment, but that neither treatment group gained significantly more than the other. At follow-up, a further 16 weeks later, post-test levels on all measures were maintained by the sequentially integrated systemic group. The emotionally-focused group maintained levels on conflict resolution, and continued to achieve target complaint improvement but slipped back on pre-test goals and relapsed on marital adjustment. Results suggest that two very different treatments, one of which had not been tested before, are both effective in helping couples alleviate marital distress, but that the sequentially integrated systemic therapy is more self-sustaining. iii TABLE OF CONTENTS Abstract ii Acknowledgements x CHAPTER I: Introduction 1 Background 1 The Problem 6 CHAPTER II: Literature Review 12 Outcome Research in Marital TherapyConceptual Models for Experiential and Systemic Marital Therapies 17 Experiential Theory 1The Process of Change in Experiential Therapj' 22 Family Systems Theory 26 Communication Theories 30 The Process of Change in Communication Theory 35 Minuchin and Structural Family Therapy 41 Milan Therapy .44 The Process of Change in Milan Therapy 47 Ethical Considerations in Communication and Milan Therapy 51 Issues in Integration of Models 53 Incompatibility of ModelsToward an Integration at the Theoretical Level 56 Sequential Models of Integration .....65 Shared Systems Theory Concepts 68 A Sequential Model for Integration at the Level of Practice 81 iv Rules for the Choice of Techniques in SIS 85 Outcome Research in Strategic Family Therapy ; 93 CHAPTER III: The Methodology of the Study 98 Design of the Study 101 The Treatments 2 Emotionally-focused Marital Therapy 10Sequentially Integrated Systemic Marital Therapy 105 Research Procedures 107 Client Population and Sampling 108 Therapist Selection and Training 9 Checking of Clinical Procedures: Implementation Check 110 Timing of Interventions Ill Instruments Used IlOutcome Instruments Ill Therapy Process Measures 116 Subject Variables 117 Data Analysis 118 HypothesesData Analysis Procedures 119 CHAPTER IV: Results and Data Analysis 121 Introduction 12Section I: Testing Assumptions of Group Equivalence 122 Therapist Effects 125 Implementation Check 6 Section II: Analysis of Treatment Effects 128 9 Statistical Background 12V Differences Between the Groups 130 Pre-test 13Post-test 14Follow-up 0 Differences Between Times 141 Pre-test vs. Post-test 3 Post-test vs. Follow-up 144 Group by Time Interactions 6 Analyses of Interaction Effects Between Sex and Treatment over Time .... 150 Section III: Descriptive Data 157 Descriptive StatisticsProportion of Treated Couples Who Improved 158 Deterioration 160 Structured Interviews 1 Posttest Questionnaires 16Follow-up Questionnaires 3 Summary of Results 166 CHAPTER V: Discussion of Results 168 Summary 16In What Ways Were the Treatments Effective? 171 1. Outcome Measures 172. Standardized Interviews 2 Additional Considerations in Evaluating Treatment 174 Differences in Treatment Effects at Follow-up 177 Conclusions 190 Limitations 1 VI Future Research 192 References 8 Appendix A: Sequentially Integrated Systemic Couples' Therapy Manual 211 Appendix B: Emotionally Focused Couples' Therapy Manual 246 Appendix C: Implementation Checklist 26LIST OF TABLES Table I: Means, standard deviations and ANOVA F values: for demographic variables across groups 124 Table II: Summary ANOVA: Couples Therapy Alliance Scale (Experimental groups) 125 Table III: Couples Therapy Alliance Scale: Means (M), standard deviations (SD) and ANOVA F-values 125 Table IV: Table of means (M) and standard deviations (SD): Dependent measures 131 Table V: Summary MANOVAs: for dependent measures 133 Table VI: Summary ANOVAs: for dependent measures 134 Table VII: Comparison of groups on individual occasions: Significance probability of F-tests 137 Table VIII: Comparison of times for individual groups: Significance probabilities of F-tests 142 Table IX: Dependent F-tests: Significance of individual interactions ...147 Table X: Summary MANOVAs: For DAS, CRS, TC and GAS: Interaction effects between sex and treatment over time 152 Table XI: Percentages of couples who improved based on the DAS 159 viii LIST OF FIGURES Figure 1: DAS at pre, post and follow-up 153 Figure 2: CRS at pre, post and follow-up 4 Figure 3: TC at post and follow-up 155 Figure 4: GAS at post and follow-up 6 IX Acknowledgements I would like to thank the members of my committee, John Friesen, Malcolm Greig, Sharon Kahn and Harold Ratzlaff for their expertise and patient guidance with tasks pertaining to theory and methodology, organization and editing. I am especially grateful to my chairperson, Leslie Greenberg, whose wisdom, vision and insight inspired this research project and whose unflagging support enabled me to finish. I would also like to thank the following: my children-Alan, Rhonda and Linda—for their loyalty and devotion; my colleagues, for their help, notably Claire Winstone for her dedication throughout the time it took for this dissertation to evolve. Finally, I would like to express personal appreciation to Fred Swartz for his caring, his willingness to listen and advise and for his excellent drafting skills, which were invaluable to this document. 1 CHAPTER I: Introduction Background Marriage is one of the primary vehicles for the satisfaction of intimacy needs. It is generally acknowledged that the capacity for intimate relationships is vital for the growth and development of human beings. The self is felt to be most meaningful when it is part of a larger context, in relationship with something outside or beyond itself (Campbell, 1980). L'Abate (1977) defines intimacy as the sharing of fears or hurt feelings with someone who can be trusted (usually a mate). These fears maj' be related to unresolved grief, past frustrations and failures, inadequacies, loneliness or poor self esteem. Dicks (1967) and other theorists argue that marriage is the ultimate adult equivalent to the positive attachment of the parent-child relationship. Marriage is the human situation in which the personality structure and the capacity for sustained, meaningful and satisfying human relations are most fully challenged. The marital relationship is a crucial relationship in many peoples' lives. Adulthood and couple relationships provide the opportunity for dependency gratification; the chance to have one's wishes and needs respected and to be the most important person to an important other (Wile, 1981). Bowlb3T (1969) maintains that attachment behaviour in adults is an essential feature of our humanness, "to judge attachment behavior in adults as inappropriate is to overlook the vital role that it plays in the life of man from the cradle to the grave" (Bowlby, 1969, p. 208). There is now evidence to suggest that positive close relationships seem to help "inoculate" those involved against the stresses of life. In analyzing life 2 histories, Lowenthal and Haven (1968) noted that the people who were the happiest and healthiest in later years were those who were or had been involved in one or more close personal relationships. Rosow (1967) suggests that the depth of intimacy experienced with others is key to adaptation over the lifespan, while the findings of Lowenthal and Weiss (1976) indicate that men and women find the motivation to live autonomous, satisfying lives only through the presence of one or more mutually supportive, intimate dyadic relationships. Given the evidence of the importance of intimate relationships in avoiding stress and engendering health, it is not surprising to note that marital problems constitute a significant proportion of all mental and emotional disorders and have wide-reaching ramifications. Problems often extend beyond the unhappiness experienced in the married state (L'Abate, 1983). The negative consequences of marital disruption for children have been well documented (Wallerstein and Kelley, 1975). These observations lend credence to the notion of the centrality of intimate and nurturing dyadic relationships. If we agree with the generally accepted premise that marriage or sustained couple relationships provide the best context for intimacy, it then becomes imperative for us not only to understand ways of facilitating its development, but to gain a deeper understanding of ways to prevent marriage breakdown and/or to help repair and revitalize such relationships when they falter. As with all human relationships, one of the key determinants in marriage breakdown seems to be unresolved interpersonal conflict. Studies have shown that couples' conflict behaviours are related to their marital distress (Koren, Carlton & Shaw, 1980). Interpersonal conflict, itself, seems to be an inevitable part of all human relationships (Deutsch, 1969). Individuals attain and maintain a sense of 3 internal emotional comfort by being in control of their environments, including their important relationships. This need for control inevitably produces a struggle regardless of the quality of the relationship (Guerin, 1982). Gurman (1978) states that it is generally agreed by theorists of varying orientations that marital conflict is always the result of attempts to define the relationship between husband and wife. Conflict, then, does not necessarily need to be pathological or destructive. Rather, it can be seen as productive. When productive, it prevents stagnation and is a medium through which problems can be aired and solutions arrived at; it is the root of personal and social change. Conflict helps to establish group and personal identities. If aimed at a resolution of tensions between antagonists, conflict is likely to have stabilizing and integrative functions for the relationship (Deutsch, 1969). Koren et al. (1980) suggest that both mutual satisfaction with outcomes and the attainment of objective resolutions are crucial aspects of conflict resolution and this is confirmed in the literature (Birchler & Webb, 1977; Birchler, Weiss & Vincent, 1975; Glick & Gross, 1975). It seems apparent that much current marital conflict is not productive insofar as tensions are not resolved and integrative functions for the relationship are, therefore, not achieved. In Canada and the U.S., one marriage in three currently ends in divorce (Adams & Nagnur, 1981; L'Abate, 1983). Marital satisfaction appears to be related in part to the number of specific unresolved problems between partners (Birchler & Webb, 1977). Guerin (1982) has noted that severe marital conflict is related to high levels of emotional arousal and anxiety, an unstable emotional climate and low feelings of warmth, calm and safety. Marital conflicts have been defined as dysfunctional when they cause psychological or 4 physical injury, decrease interpersonal trust and fail to generate constructive changes in subsequent interaction between partners (Feldman, 1982). Conflict between marital partners is often associated with dysfunctional interactional cycles which have evolved systematically. Communications theorists (Watzlawick, Beavin & Jackson, 1967) refer to repetitive series of exchanges between partners which reflect a rule about the relationship. Jackson (1965a) was the first to emphasize the importance of the notion of such rules in defining the nature of the relationship (Steinglass, 1978). In a typical cycle, each partner has a specific role. For example, the wife may engage in nagging criticism while the husband engages in passive withdrawal. In explaining their frustrations, the couple sees the sequence in a linear, causal fashion. The husband states that his withdrawal is a defence against her nagging while she sees this as distortion and notes that she is critical because of his passivity (Steinglass, 1978). From an interactional or systemic perspective, however, the cycle involves circular causality: when she nags, he withdraws and his withdrawal leads to her nagging which in turn leads to his withdrawal and so the cycle has no beginning or end. Negative interaction cycles in couples have been observed to evolve repeatedly when dealing with issues concerning closeness and distance, dominance and submission. Once a negative interaction cycle is in place, it is self-perpetuating: it seems to become its own reason for existence (Greenberg & Johnson, 1984). Wile (1981) has delineated a number of couples' patterns as well: pursuer/distancer, demanding/withdrawn, mutual accusation/mutual withdrawal, attack/attack. The high incidence of unresolved conflict and marital distress can be seen as antagonistic to the development of intimacy and leads to a search for the most effective method of helping to repair marital relationships. In the last decade, 5 marital therapy has evolved into one of the most significant methods of intervention (L'Abate, 1983). Intervention at the level of the marital relationship has a substantial impact at other levels and if only one level could be approached, marital therapy would seem to have the greatest potential for effecting change in both the individual and the family (Lewis, Beavers, Gosset & Austin Phillips, 1976). In a summary of the results of 77 studies of marital and family therapy, Gurman and Kniskern (1981) reported that marital therapies seemed to produce positive change in almost two-thirds of clients, whereas individual therapies for marital problems showed improvement in only 48 per cent of clients. Although it is suggested in the literature that marital therapy as a whole is effective, there is not enough empirical data representing all orientations to allow meaningful comparisons among differing types of approaches. Thus it has not been clear which general orientations or specific techniques are effective in treating marital problems (L'Abate, 1983). Despite Jacobson and Weiss's (1978) evidence of the effectiveness of a behavioural approach and cumulative data which suggest a 64 per cent rate of superiority over control conditions, Gurman (1978) has concluded that "behavioral, as compared with nonbehavioral marital therapy studies, yield few results to indicate clearly the superiority of behavior therapy over no treatment or other treatments" (L1 Abate, 1983, p. 280). Because there has been substantial homogeneity of interventions across different theoretical models, the development of such comparative studies has been impeded. For example, Sager (1981), a psychodynamic practitioner, and Weiss (1978), a behavioural practitioner, both use interventions such as behavioural contracting. Some models have also become more encompassing. O'Leary and Turkewitz (1978) point out, for example, that behaviour therapy has tended to 6 become ambiguous as a result of the cognitive trend in psychology. In a recent significant outcome study Johnson and Greenberg (1985) clearly differentiated and delineated a cognitive behavioural approach from an experiential approach, finding the experiential treatment superior to the cognitive behavioural treatment. It appears, therefore, that it may be possible to show differential treatment effects. The Problem As stated above, marital therapy in general has been shown to be effective and indeed superior to other forms of therapy in treating marital problems. However, very little empirical data exists to suggest what specific techniques within the general rubric of marital therapy are most effective. Previous marital therapy research has been focused predominantly on behavioural problem solving therapies (Jacobson, 1978b, 1979; Jacobson & Weiss, 1978) and has not included comparative outcome studies involving other methods. While behavioural marital therapy has shown some impressive gains with mildly or moderately distressed couples, there is less persuasive evidence of the power of these methods with severely disturbed couples (Gurman & Kniskern, 1981). Moreover, the authors cite frequent methodological inadequacies and question the relevance of many of the outcome measures for clinical practice (Gurman & Kniskern, 1978). Marital therapy has been criticized as an ill-defined field that lacks an adequate theoretical base (Ackerman, 1972; Bowen, 1978; Haley, 1971; L'Abate, 1983; Olson, 1970). Although theory has developed considerably in the last ten years, there is still no widely shared language nor a comprehensive theoretical foundation (Vines, 1979). Although specific comprehensive theories have been 7 developed to explain marital distress and dysfunction, they tend to stress specific unitary components of behaviour rather than more global or integrated components of dyadic relationships. L'Abate (1983) notes, "Major theories stress altering feelings, altering rationality or altering behavior with no theory incorporating all three areas of human functioning. Each major theoretical orientation presupposes that changing one area of functioning (feeling, thoughts or behavior) will produce changes in the other two areas" (p. 226). While the notion of a theory which integrates all three areas of human functioning has enormous appeal, it would seem crucial that research be generated in order to test the kinds of therapies that would attend to the individual dimensions of feelings, thoughts and behaviours. The groundwork for such an undertaking, however, has not yet been adequately laid, as evidenced by the paucity of research supporting each theoretical model and the lack of empirical data to allow meaningful comparisons among the differing approaches. Nor would it seem possible to do more than speculate about global or integrated components of dyadic relationships until such time as marital distress and dysfunction are specified and operationalized, for the purposes of investigating the potency of varying techniques in effecting change. Based on the close link established between marital dissatisfaction and conflict patterns, then, one of the main goals, if not the sine qua non of marital therapy, is to facilitate the resolution of conflict patterns between spouses, specifically the escalating type of conflict pattern. An important research question which follows from this addresses the issue of the effects of different interventions on couples' conflict resolution. Two promising broad approaches to marital therapj', neither of which have had much research investigation, are "Emotionally-Focused" marital 8 therapy, based on experiential therapies, and "Sequentially Integrated Systemic" marital therapy, based on systemic therapies. From an intuitive perspective, both therapies are appealing. Clinicians experienced in either or both of them have attested to positive outcomes resulting from their use. The Emotionally Focused Marital therapy has been empirically tested and found to be effective (Johnson & Greenberg, 1985) but the Sequentially Integrated Systemic Marital therapy has not been empirically tested. In order to increase understanding of both therapies and to determine whether present claims as to their effectiveness are supported by empirical evidence, further sytematic investigation is required. The aim of the present study is to conduct such an investigation in order to compare the effects of an emotionally-focused marital therapy with the effects of a sequentially integrated systemic marital therapy. A brief description and delineation of the emotionally-focused therapj' (EF treatment) and of the sequentially integrated systemic therapy (SIS treatment) is presented below. A more detailed conceptual model for these approaches is presented in Chapter II and a summary description, including a set of interventions specific to each of the two treatments, is presented in Chapter III. The manual for the SIS treatment can be found in Appendix A and the manual for the EF treatment can be found in Appendix B. The delineation of the EF and SIS approaches demonstrates that the two are not homogeneous and allows them to be compared for outcome assessment purposes. The cornerstone of the emotionally-focused marital therapy is emotional experiencing. The role of emotional experience in therapeutic change and in the resolution of interpersonal conflict and/or marital conflict or change has been controversial and has evoked a variety of responses from different theorists. Some 9 have seen it as possibly detrimental or irrelevant (Jacobson & Margolin, 1979; Stuart, 1969), whereas others have seen it as essential to therapeutic change (Greenberg & Johnson, 1984; Greenberg & Safran, 1984a). Of those who see emotional experiencing as essential to change in therapy, Greenberg and Safran believe that newly generated emotional experience not currently in awareness can provide an important motivating force to implement new behaviours. In experiential marital therapy the goal is for partners to access and acknowledge-in both themselves and the other--previously unexpressed feelings underlying their reactive behaviour patterns. This process can lead to a change in both the way partners are perceived by each other and how they communicate with each other. For example, if one partner sees pain and sadness expressed by one who has previously been seen as demanding, this will most often lead to positive change and can evoke feelings of comfort and support instead of self-protectiveness against attack (Greenberg & Johnson, 1984). In contrast to the Emotionally-focused marital therapy, the emphasis in the Sequentially Integrated Systemic marital therapy is on changing the conceptual frame of reference or meaning attributed to the couple's negative interactional cycle and thereby modifying patterns of communication (Steinglass, 1978). For the systemic therapist, emotional experiencing plays only a very small part, if any, in the process of change. In the SIS therapy the awareness and expression of feelings ma}' or may not be a consequence of the therapeutic process but they are not seen as at all necessary to or catalytic in the process. Wile (1981) notes that whereas traditional causal theory is linear, "the generating concept of system is circular causality" (p. 27). Each partner's behaviour is seen as a reaction or adjustment to the behaviour of the other. (One partner withdraws because the 10 second nags, while the second nags because the first withdraws). Reframing of the marital interaction and the spouses' positions in this interaction becomes the catalyst for change. Watzlawick (1976) has defined reframing: "to change the conceptual or emotional setting in relation to which a situation is experienced and place it into another frame that fits the 'facts' of the same concrete situation equally well or better and thereby change its entire meaning" (p. 122). Therapy based on a systemic view is quite pragmatic and treatment contracts negotiated with clients usually hinge on presenting complaints, which for the purposes of establishing goals are defined in very concrete terms. Based on this, therap}' is present-centered and the therapist is primarily concerned with the interaction. Interpretation is not used to foster either genetic or interactional insight; instead, the negative interactional cycle is reframed in order to give it new meaning and is then prescribed with the intent of creating recalibration or change in the system (Gurman, 1981a). The present study is an investigation and comparison of the effects of an emotionally-focused treatment, a sequentially integrated systemic treatment and a wait-list control condition on marital discord. The study is a comparative outcome study to evaluate the efficacy and differential effects of these two marital treatments and a wait-list control condition in helping couples to improve marital adjustment, resolve conflict cycles, reduce target complaints and reach specified relationship goals. The dependent measures are the Dyadic Adjustment Scale, the Conflict Resolution Scale, Target Complaints Reduction and Goal Attainment Scaling. The units of measurement are the couples' combined mean scores on the four measures. 11 Before conducting the study it was expected that the two treatment groups would not differ from each other on the dependent measures but that both would show significantly more improvement than the control group at post-test and at four month follow-up. Secondty, it was expected that the couples exposed to SIS therapy or EF therapy would show significant differences on the dependent measures from pre-test to post-test and pre-test to follow-up but that couples in the control condition would not differ. To address the second expectation, each of the groups was analyzed separately, over time. Finally, it was expected that there would be group by time interaction effects between the SIS and EF groups over occasions but not between either of those groups and the control group on the dependent measures. 12 CHAPTER II: Literature Review The following review focuses on the outcome literature in marital therapy. The emphasis is placed on existing studies specific to experiential and systemic marital therapies, with a brief review on behavioural marital therapy and research. Conceptual models that form the underpinnings for the emotionally-focused and the sequentially integrated systemic marital therapies are presented. Outcome Research in Marital Therapy A significant increase in research on the outcomes of family and marital therapies has characterized the last ten years. This is reflected in the contrast between a review of outcome studies in family therapy by Wells, Dilkes and Trivelli (1972) where only 13 relevant reports could be identified and a review where 500 reports were examined with total N approaching 5,000 (Gurman & Kniskern, 1981). In their review, Gurman and Kniskern state that evidence suggests that both behavioural and nOn-behavioural treatments are effective beyond chance and concluded that, in the marital therapy reviewed, 65 per cent of cases improved (Gurman & Kniskern, 1981). In single group studies, conjoint marital therapy is found to be superior to alternative treatments in 70 per cent of comparisons and inferior in only 5 per cent (L'Abate, 1983). Of Gurman and Kniskern's (1978, 1981) overall conclusions on marital therapy studies, those that have relevance here are: 1. Couples benefit most from treatment when both partners are involved in therapy, especially when seen conjointly. 13 2. Individual psychotherapy for the treatment of marital problems has a poor record of positive outcomes and a strikingly high rate of negative outcomes. 3. Short term therapies (8 - 12 sessions) seem to be at least as effective as treatments of longer duration. 4. Therapist relationship skills have major impact on the outcome of marital family treatment regardless of the "school" orientation of the clinician. 5. Behavioural marital therapy offers insufficient research to justify the training of therapists in this approach alone. 6. The only interventions which have received consistent positive empirical support as facilitating outcomes of marital therapy, regardless of the style of such therapies, are those that increase couples' communication skills (Birchler 1979, Gurman & Kniskern, 1977; Jacobson, 1978b 1979). Further empirical evidence of the efficacy of communication skills training comes from Azrin, Naster and Jones (1973) who used a type of behavioural contracting and communication training approach and Azrin, Besalel, Bechtel, Michalicek, Mancera, Carroll, Shuford and Cox (1980) who added more extensive communications and modified procedural and methodological flaws of the previous study, including a larger sample and a control group. In the latter study, "reciprocity counselling," which provided communication training and instruction, behavioural contracting and instruction in mutual reinforcement was found superior to a control condition which only encouraged discussion of problems. According to Gurman and Kniskern (1981), behavioural marriage therapy seems to be about as effective for minimally to moderately distressed couples as non-behavioural methods (Jacobson, 1978a, 1979), though it should be noted that behavioural and non-behavioural studies often use very different outcome criteria. 14 Like psychodynamic marital therapy, behavioural marital therapy has not shown much empirical support in the treatment of severely distressed marriages (Gurman & Kniskern, 1981). L'Abate (1983) has cogently pointed out that although current behavioural research claims valid and often excellent results, it must be noted that behavioural research attempts to measure complex relationships by assessing only a few discrete concrete variables in the marital relationship (i.e. exhibition of specific behaviours). Widespread methodological problems include using instrumentation with little or no established validity and reliability and using self-report data without acknowledging the possibility of improvement reported for any reasons other than real gains (i.e. justifying treatment or pleasing the therapist). Until recently, comparative studies in marital therapy had been limited to comparisons of interventions comprised within single treatment models. O'Leary and Turkewitz (1981), for example, conducted a comparative outcome study of behavioural contracting (similar to the "good faith" contracts of Weiss, Hops and Patterson (1973) and communication training (modelling, feedback, role playing). Although there were no overall differences between the two treatment groups, treatment over no-treatment was effective. Margolin and Weiss (1978), in a similar comparison of two types of behavioural treatment with a non-specific intervention (control), found behavioural communications skills training with a cognitive restructuring component more effective than behavioural communication skills training only. A few years ago, Gurman and Kniskern (1981) stated that a conclusive assessment could not be made of the general comparative efficacy of behavioural vs. other marital and famity treatment methods and that such studies were nearly 15 non-existent. By 1984, only one clinical (non-analogue) study had compared behavioural interventions with another form of marital therapy. Liberman, Levine, Wheeler, Sanders and Wallace (1976) compared behavioural marriage therapy including communication skills training and contingency contracting with "interactional therapy" which included description and discussion of feelings. Although both groups improved, there was no difference between groups on self-report measures, but the behavioural group did show improvement on problem solving methods. Methodological problems (lack of a control group, non-random assignment and small therapist samples), however, hampered the generalizability of conclusions based on this study. A significant contribution has now been made by Johnson and Greenberg (1985) who have conducted an outcome study in which a cognitive behavioural intervention which taught problem solving skills and an experiential intervention which focused on emotional experiences were compared with a control group. Results showed that both treatment groups made significant gains over untreated controls on measures of goal attainment, marital adjustment, intimacy levels and target complaints reduction. More striking were the effects of the experiential treatment, which were superior to those of the cognitive behavioural treatment on marital adjustment, intimacy, target complaints and goal attainment levels. At follow-up, marital adjustment scores in the experiential group were still significantly higher than in the cognitive behavioural group. Their study is particular^' important because of the use of a control group, random sampling procedures, a sufficient sample size (45 couples), pre-post test measures, therapist implementation checks and the provision of treatment manuals and training for each treatment condition. Replication of the findings on emotional therapy would seem valuable and 16 informative. The fact that therapists of psychodynamic, analytic and systemic theoretical orientations alike have employed a large number of behavioural techniques such as contingency contracting and communication skills training in their marital therapy has undoubtedly been a contributing factor in the dearth of comparative outcome studies across models. With regard to experiential approaches, the scarcity of research may be partly attributable to a previous sketchiness in the elaboration of the theoretical underpinnings of experiential mechanisms of change as well as the lack of clear identification and description of the sequences of interventions that are associated with them. A description and delineation of an emotionally-focused treatment which makes possible a meaningful comparison between a behavioural and an experiential approach has been provided by Greenberg and Johnson (1984). It seems, then, that the logical and illuminating next step is to compare such an experiential approach with a systemic approach to marital therapy. This necessitates specification of a sequentially integrated systemic treatment. The treatment suggested here is delineated in Greenberg and Goldman's manual, which can be found in Appendix A. The experiential treatment suggested here is the emotionally-focused treatment, as delineated by Greenberg and Johnson (1984), which can be found in Appendix B. Outcome literature in marital therapy, particularly that pertaining to comparative studies across models, has been discussed. Before reviewing any studies specific to the experiential or systemic treatments, it is important to describe the conceptual models that form the underpinnings for these two approaches. 17 Conceptual Models for Experiential  and Systemic Marital Therapies Experiential Theory Experiential therapy is an outgrowth of humanistic-existential theory (Greenberg & Johnson, 1984). Although developed as an individual therapy, Gestalt therapy, one of the major experiential therapies, addresses the issue of organism/environment interactions (Greenberg, 1982; Perls, 1973). In this wholistic field theory, the organism is understood as it exists in relation to and in interaction with its environment . . . taking in what it needs and is available from the environment. . . . Rather than a dualistic idea of a need arising in the organism as an entity separate from the environment, a need . . . (including psychological contact needs) . . . is regarded as a psychological event that comes into being at the contact boundary between the organism and the environment in which it arises. Needs which result in this fashion are regarded as determining human behaviour. Behaviour, therefore, is a function of the total field. (Greenberg, 1982, p. 2) Gestalt theories, therefore, lend themselves to an integration with systemic perspectives in which context is seen as an important determinant of behaviour. In their work with individuals, however, Gestalt therapists have focused more on the individual's awareness and not attended to the importance of the environmental context. With this emphasis on awareness, an individual's current organization of the world has been the focus of therapy. Individuals are regarded as having inherent tendencies to survive and grow and to develop their capabilities in ways which will serve to maintain or enhance them. All behaviour, then, stems from the individual's quest to actualize himself or herself. As such, it can be assumed that an individual's behaviour at any given moment is his or her optimal 18 means of actualizing himself or herself in the currently perceived environment. Blocks to awareness and experiencing have been seen as central to individual problems. Such blocks may result from various contingencies. For example, blocks occurring as a result of "unfinished business" affect the individual when needs have not been fully satisfied and are out of awareness. Although such needs recede into the background, they still press for closure and so can interfere with an individual's ability to respond to new situations. Restrictions of awareness, avoidance and disowning aspects of current experience (Greenberg, 1982; Greenberg & Johnson, 1984) can also result in blocks. For Perls, the "self is the totality of the sensing, acting organism, and not a static structure. Part of the work of therapy involves expanding the "I-boundary" and allowing for a broader range of awareness. The I-boundary is therefore a description of the functioning of the dynamic, shifting nature of the ego or what Rogers referred to as "self-concept" (Greenberg, 1982). Blocks to awareness and experiencing or interests of the forming self that are not allowed are regarded as alien to the ego and are perceived as threats to the individual's integrity. To combat such perceived threats, the individual alters his or her ego functioning so that the need gratifying sequence is interrupted. These disturbances occur without awareness and hence distort the individual's apprehension of realitj' and allow him or her to avoid direct and immediate contact with the "here and now." Two of these "alterations of the boundary" which are relevant to couples' therapy are "projection" and "introjection." These boundary alterations are generally accepted as two of the major processes by which ego functions are altered and are considered pathological when they are maintained outside the individual's awareness: 19 1) Projection. In this process, the individual attributes disowned aspects of the self to others. The aspects of the self that are contacted are alienated and regarded as foreign to the self. Particular emphasis is paid to attributions in which the individual does not distort reality seriously but shows his or her over-concern in perceptual selectivity of certain phenomena. This often occurs when judgements and evaluations of the individual are imagined to be occurring in the environment and results in over-concern with what other people think. 2) Introjection. In this process, aspects of the environment are identified with as if they were aspects of the self. The excitations and interests of the self are not contacted. The introjects are contacted as if they were the self and this results in conflict (Greenberg, 1982). This concept of introjection can be related to a parallel concept in Object Relations theory, where introjection refers to the build-up of organized clusters of memory traces regarding significant others and the images of self in relationship to those persons. This process results in a unique distortion of "self and "other" perceptions: neither are perceived as whole beings with ambivalent qualities; ambivalence is defended against by the mechanism of splitting where positive and negative values are kept apart (Segraves, 1982). This leads to the notion of projective identification wherein internal conflicts are translated into more concrete modes. In Gestalt terms, the related processes are projection and introjection or a combination of both, which refer to the projection of inner objects onto the other. As applied to couples' counselling, the phenomenon of projective identification can be seen in operation when a client consistently misperceives or exaggerates some aspect of the spouse's (or therapist's) personality that is denied in himself or herself. The goal of therapy is the 20 re-experiencing of good and bad parts, thus allowing the client to reintegrate dissociated parts of the self (Segraves, 1982). Within the experiential framework which stresses individual awareness, Satir (1967) has added a critical focus on the importance of congruent communication and closeness in relationships (Greenberg & Johnson, 1984). The essence of Satir's position is that there is a reciprocal interrelationship between communication difficulties and individual self-esteem. She believes that healthy interpersonal relationships require those involved to have a sense of individuality and relatedness (Segraves, 1982). She notes that individuals need to learn to discriminate among internal feelings, images and introjects, and external reality. Although the emphasis on communication and interrelatedness places her approach in a communication and systems framework, the emphasis on affect also places her approach in an experiential framework (Greenberg & Johnson, 1984). Gurman (1981a), Wile (1981) and Greenberg and Johnson (1984) all suggest integrative theories of marital therapy which draw upon aspects of intrapsychic as well as interactional theory. In support of this, Gurman (1981a) states: It is likely that treatment approaches which systematically consider and attempt to produce change on multiple levels of psychological experience will facilitate the development of interventions that are more flexible and responsive to (clients) and will . . . lead to more positive and enduring clinical outcomes. (p. 422) In an experiential marital therapy an overriding goal is to achieve change in each partner as well as in the marital interaction. According to Gurman, the behaviour of each partner is not always under interactional control. While one partner's symptoms often have communication meaning for the relationship and often reflect a disturbance in the relationship, they do not necessarily have functional value. One partner's symptoms may reflect both his or her disturbance and the relationship 21 disturbance. Contrarj' to many notions in the family therapy field, while change in one partner necessarily changes the marital system, system change is not always required in order for change to occur in one or the other partner (Gurman, 1981a). An experiential approach encourages development of a shared perspective and mutual caring between partners and, in addition, places a priority on the fulfillment of individual needs. Wile (1981) sees psychological symptoms as emanating from the deprivation of needs that individuals are currently experiencing. The result is a lack of satisfaction and control that would be necessary to make conditions liveable. In this view, adulthood is seen as providing an opportunity for dependency gratification, the chance to have one's wishes and needs respected and the opportunity to be the most important person to an important other. Adults as well as children seem to share an awkwardness in expressing what are recognized as universal human needs. This view is in contradistinction to a depth analytic view which traces psychological symptoms to infantile impulses and developmental defects where the focus is on regressive gratification and secondary gain, (i.e. exploitive control over others). The major proposition of this view is that it is not individuals' feelings and wants that cause problems in marriage but rather their disowning or not allowing of these feelings that leads to ineffective communication and escalating interactional cycles (Greenberg & Johnson, 1984). For example, an angry feeling expressed in a straightforward manner often has a clarifying effect; but when it is suppressed and seen as unacceptable or pathological it takes on an infantile quality (Wile, 1981). Greenberg and Johnson note that some major needs in couples are for closeness, contact/comfort and intimacy. Intrapsychic fears of closeness and interactional patterns which prevent closeness are, then, goals for change. 22 The Process of Change in Experiential Therapy. In the experiential model of couples' therapy, change occurs in each partner as well as in the marital interaction. This suggests that the process occurs within the individual as well as within the context of the relationship. Change or growth within the individual involves developing a broader range of experiential awarenesses, including an awareness of unmet needs for closeness and intimacy and the legitimization of feelings of vulnerability or deprivation (Greenberg & Safran, 1984a; Greenberg & Johnson, 1984; Wile, 1981). The mechanism for such interpersonal change suggested by the author is emotional experiencing, "events in therapy in which newly formed . . . emotions are brought into awareness and expressed" (Greenberg & Safran, 1984, p. 560). These authors note the need for a perspective which recognizes that emotional experiencing is implicated at all levels of behaviour and that emotional experiencing is a crucial facet of individual therapeutic change. The authors suggest a model which assumes that emotional experience is as much a function of information processing which takes place at preconceptual, expressive motor and schematic memory levels as it is a function of conceptual cognition. The notion of emotional experience as a complex integrative process implies that purely cognitive change at a conceptual level of information processing will not necessarily produce change at a feeling level in an individual. Acknowledging aspects of experience and certain primary emotions not currently in awareness provides individuals with adaptive, affective responses which aid problem solving. Therapists, however, often tend to construe affect as a dependent variable in human functioning and regard emotion as disruptive to the therapeutic process. Mahoney (1980) has noted that cognitive therapists tend to view feelings too narrowly and to overlook the importance of unconscious processes. He recommends 23 that cognitive therapy needs to address these issues in order to provide a more adequate understanding of human adaptation. Zajonc's (1980) argument for the primacy of affect has implication for such therapy. Rachman (1981) posits three independent systems, including affective and cognitive and suggests that therapeutic focus should be guided by the dysfunctional system(s). Rice (1974) asserts that reflection of feeling is potentially one of the most active and powerful tools available to a therapist and calls upon the evocative function of the therapist whereb}' the client is encouraged to experience and recreate emotions surrounding a problematic event. Those advocating change through emotional experiencing assert that newly generated emotional experience can provide an important motivating force to implement new adaptive behaviours; the absence of this new emotional experience may result in a lack of motivation to implement such new behaviours. The notion of emotional experience as a complex integrative process implies that purely cognitive change at a conceptual level of information processing will not necessarily produce change at a feeling level in the individual (Greenberg & Safran, 1984b). Greenberg and Safran (1984b) have noted a number of intrapsychic change processes in which affect plays an important role. Two of these are relevant to couples' therapy. The first is the process of acknowledging previously unacknowledged biologically adaptive primary emotions which aid problem solving. The second is the modification of state dependent cognitions which emerge for therapeutic consideration only when the individual is in the aroused affective state in which the response of interest previously occurred. "These affective change processes occur in each individual during the process of successful (experiential) couples' therapy" (Greenberg & Johnson, 1984). Arousal of currently experienced feelings, i.e. anger, sadness, can provide access to such 'state dependent learnings.' 24 The mechanism of change is the learning of new responses in the domain of "hot cognitions" (Greenberg & Safran, 1984b). This is more likely to take place if the individual is in the state in which the response that needs to be modified originally occurred. Accessing these hot cognitions can be particularly important in clarifying couples' interactions. Often clients are not able to recall momentary construals leading to key behaviours when the problem is being discussed coolly in therapy. Helping couples recreate the situation and relive the emotions in therapy often makes the cognitions governing these behaviours more available for clarification and modification (Greenberg & Johnson, 1984). Interventions range from using vivid language to reflection of feeling and heightening of non-verbal expression. Once this state has been evoked, experiential procedures from Gestalt therapy (Perls, Hefferline & Goodman, 1951) or client centered therapy, such as empathic reflection and evocative responding (Rice & Greenberg, 1984), could be used to encourage emotional experiencing. Adults often cry in association with making the profound behaviour altering cognitions that may result (Greenberg & Safran, 1984b). The goal in experiential marital therapy is to incorporate such inner experiential changes into the relationship. Accessing of adaptive primary emotion provides self-defining information and motivation which enhances problem solving. Such therapeutic changes will have an impact on the couple relationship. At least five processes of change in the emotionally-focused treatment of couples have been identified by Greenberg and Johnson (1984). 1. An individual perceives himself or herself differently by bringing into focus awareness experiences not previously dominant in this person's view of self, for example, "I see and accept my vulnerability." 2. The spouse, upon witnessing the partner's new affective expressions, perceives the partner in a new way; for example, "I see your need for caring and contact rather than your hostility." 25 3. The individual's personal reorganization leads to different behaviour in the interaction of the spouses; for example, "I now ask you for reassurance from a position of vulnerability." 4. The spouse's new perceptions of the partner leads to different responses; for example, "I comfort you rather than withdraw." 5. As a function of their partner's new behaviours, the individuals come to see themselves in a new way; for example, "Since I can fulfill your needs, I see myself as valuable and necessary to you." Extending the focus from the present interactional patterns to the intrapsychic concerns of one or both partners is designed to enable them to become more aware of currently experienced, unmet needs, to explore these awarenesses on an emotionally meaningful level and thereby expand "I boundaries" and initiate further individual growth. The self is redefined in the relationship for both partners. Different aspects of the self are accepted and expression of these leads to changes in the interactional sequences. This process is instrumental in enabling each partner to appreciate the other's position and accept certain behaviours in the other that were previously unacceptable (Greenberg & Johnson, 1984; Wile, 1981). Because of the high demand for disclosure, the process is conducive to building intimacy and emotional bonds, both of which are key aspects of marital satisfaction (Tolstedt & Stokes, 1983). The process of inner change is also important in helping each partner to deal with "boundary alterations" that may have resulted from "unfinished business" or perceived threats to his or her integrity. These restrictions have important consequences for dyadic interaction. Disturbances which have occurred out of awareness make it likely that individuals will see their spouses in terms of past experiences or relationships instead of in the current context. As individuals become aware of these projections and introjections and/or the process of projective identification, they may be able to re-experience and integrate dissociated parts of 26 themselves (Segraves, 1982) and develop new awarenesses of their partners. Sharing these experiential processes can also further help to invoke the sympathy and understanding of their partners, leading to change and resolution of conflict in the relationship. Family Systems Theory There are varying "schools" or orientations within the broad framework of family and marital systems theory. Different models of practice have evolved from these theoretical schools. The family systems theories and the models of practice which have contributed to the development of the sequentially integrated systemic marital therapy tested here are briefly described. A description of models is then followed by a discussion of the issues surrounding integration. Such an integration has been supported by some therapists and renounced by others. The arguments for and against integration are summarized. Following this is a proposal and elaboration of an integration of these different theories at the level of practice. It has been pointed out b}' Steinglass (1978) that "system theory, as it has been applied to marriage and marital discord is at the moment more a series of loosely connected concepts than an integrated theory of marriage" (p. 300). Olson (1970) and more recently, L'Abate (1983) have concluded that marital therapy does not have a firm theoretical base. For Steinglass (1978), the essence of a systems approach (from which marital systems theory has emanated) is defined as "attention to organization, to the relationship between parts, to concentration on patterned rather than linear relationships and to a consideration of events in the context in which they are occurring rather than in isolation from their environmental context" 27 (p. 304). The principle of "wholeness" seems crucial to an understanding of the development of marital systems therap}'. The main principle of General Systems theory is that a collection of elements, once combined, produces an entity that is greater than the additive sum of each of the parts. The premise is that a system cannot be adequately understood or explained once it has been broken down into its separate parts. It also proposes that any single element within a system cannot be thought of as acting independent!}' (Steinglass, 1978). Marital systems are conceived as similar to cybernetic systems, in which the generating concept of causality is circular, rather than linear (Bateson, 1971; Jackson, 1967) and complex interlocking feedback mechanisms and behaviour patterns repeat themselves in sequence. An individual's symptoms serve as homeostatic mechanisms which regulate the couple's transactions (Gurman, 1981a). Wile (1981) notes that the concept of homeostasis is an organizing principle in family systems theory. In relation to marital therapy, the couple, having achieved equilibrium, is seen as resisting or counteracting forces that threaten this equilibrium. The author states that traditional causal theory is linear and that "the notion of circular causality changes the picture. Each partner's behaviour is now viewed as a reaction or adjustment to the behaviour of the other. One partner withdraws because the second nags, while the second nags because the first withdraws" (p. 28). Weeks and L'Abate (1982) note that "the behavior of a system is the product of a complex series of transactions" (p. 25). The symptoms of partners are seen as both system-maintained and system-maintaining and all individual problems are seen as an outgrowth of marital-family disturbance. Marital conflict 28 is viewed as a result of interaction; intrapsychic (especially unconscious) forces are considered irrelevant. The symptoms of husband and wife are assumed always to have interpersonal meaning and to function as communicative acts, so that a symptomatic individual cannot be expected to change unless his or her family system changes (Gurman, 1981a). Based on this, Gurman and Kniskern (1978) have shown that treating an individual in a vacuum for a marital problem may produce deterioration in the marriage. As suggested earlier, the experiential theorist's notion that emotional experiencing is pivotal to the process of change in marital therapy is not shared by system-oriented marital therapists, who show some similarity to cognitive therapists (Beck, 1976; Ellis, 1973; Meichenbaum, 1977) in their belief that awareness and expression of feelings may or may not be a consequence of the therapeutic process. Such processes are not seen as catalytic or necessary at all for change in therapy. Because there are varying schools or orientations within the broad framework of family and marital systems theory, the major schools and their accompanying theories of change are elaborated separately here. This is followed by an integration at the level of practice which suggests a broad system-oriented approach to change and approach to intervention. Although family and marital systems theories have incorporated some of the broad principles of general systems theories, it is incorrect to suggest that all systems therapies are based mainty on general systems theory. Communications theory, a systems model of family and marital interaction, is one theory that is more directly tied to and derived from general systems theory (Gurman, 1978). The author notes, "there have evolved a number of family systems theories, with a wide range of adherence to and deviation from the language and logic of general 29 systems theory" (p. 507). Several influential family systems theories had emerged in the few years prior to 1978, among those of note: the strategic therapy of Haley (1963, 1976), the communications-oriented (or interactional) therapy of the Palo Alto group (Fisch, Weakland & Segal, 1982; Sluzki, 1978; Watzlawick, 1976; Watzlawick, Beavin & Jackson, 1967; Watzlawick, Weakland & Fisch, 1974; Weakland, 1976; Weakland, Fisch, Watzlawick & Bodin, 1974), the structural family therapy of Minuchin (Liebman, Minuchin, Baker & Rosman, 1976; Minuchin, 1974; Minuchin, Baker, Rosman, Liebman, Millman & Todd, 1975), and the eclectic therapy of Satir (Satir, 1967). Since that time, publications about Milan therapy and the work of Selvini-Palazzoli, Boscolo, Cecchin, Prata (1978, 1980) have had a significant impact on the field of family systems theory. Of these theories, emphasis is placed here on an elaboration of those which are drawn from in the conceptualization of the Sequentially Integrated Systemic therapy used in the present study. Both the Milan group's theories and communications-oriented theories are included because they have had considerable influence upon the development of the techniques and interventions that are associated with the integrated systemic therapy. Haley is included under the heading of Communication Theories as his paradigm could be seen as having its genesis in the communications model even though he is also associated with structural therapy and has become concerned with larger sequences of behaviours more closely associated with structural models. Satir is not discussed here because her work seems more relevant to experiential therapy and has been elaborated earlier (see Experiential Models). Minuchin is summarized briefly; his structural work has relevance for integrated systemic marital therapj' even though his focus and the publication of work has been directed away from the marital dyad and 30 more toward whole family systems. Communication Theories. The multi-disciplinary group at the Mental Research Institute (also known as the Palo Alto group) has for the past 25 years been developing concepts of marriage and family interaction based on communication theory (Steinglass, 1978). Communication theory is a systems model of marital interaction and therapy more directly tied to general systems theory than others (Gurman, 1978). The central focus in communication theory is away from thoughts and feelings (internal processes) of the individual, which are seen as a distraction, toward the far more useful data regarding communicational "input" and "output." Speculations about the fantasy life, motivation or structural organization of the mind are seen as merely confusing and even destructive if the goal is to describe and understand the rules that govern human behaviour. This focus applies when considering marital disorders (Steinglass, 1978). Again, inner processes are disregarded; this is in sharp contrast to a psychoanalytic or intrapsychic approach. Fisch, Weakland and Segal (1982) note, "Although one could always speculate about possible broader implications ... we are not . . . attempting to present a comprehensive theory of human nature, of human existence or 'the mind' but only to state ... a theory as close as possible to practice" (p. 6), and later, "a focus on communication and interaction within the family leads to much more emphasis on actual behaviors, what is observably going on in the present, rather than on the past, the internal and the inferential" (p. 8). This is unlike Bowen with his emphasis on development and family of origin. The M.R.I, group points out that viewing behaviour in its immediate context represents a shift in epistemology from a search for linear cause and effect chains to a cybernetic or systems view or "the understanding and explanation of any . . . behavior in terms of its place in a 31 wider, ongoing, organized system of behavior involving feedback and reciprocal reinforcement throughout" (Fisch et al., 1982, p. 9). This shift also implies less of an emphasis on individual deficits than has been espoused in psychoanalytic approaches. Haley (1980) supports this group's systemic and interpersonal view of behaviour; when he refers to ideas that have handicapped therapists, he notes that whereas the therapist with an organic theory thinks of the schizophrenic as a defective person, a therapist with a systemic theory realizes that it is the "social function of a young psychotic to fail" (p. 12). Fisch et al. (1982) stress that the Palo Alto group's most basic view, "a meta-view to which all the rest are subsidiary" (p. 10) in regard to the nature of truth or reality is that there are not views that. are more or less real or true, but that "some views may be more useful or effective than others in accomplishing one's chosen end, but this is a pragmatic criterion, not one of 'reality'" (p. 11). Watzlawick et al. (1974) notes that there are two separate orders of reality. The first deals with physical properties of objects and our perception of them, while the second is based on the attribution of meaning and value to these objects. The authors state that most human problems involve not only the second order reality and following from this there are not 'true' underlying problems but rather that problems depend upon how individuals view things. From a theoretical perspective, communication that occurs is seen as divided into the following categories: syntax, semantics and pragmatics (Watzlawick et al., 1967). Syntax refers to the ways in which information is transmitted; in marital communication, aspects of who-to-whom speech, percentage of speaking time for each partner, conciseness of speech and "ratio of information to noise" are taken into consideration. Semantics refers to the meaning of the communicational act and how 32 it is understood by the receiver. Of interest in marital communication is clarity versus confusion, existence of private, shared communicational systems of "code" words or gestures. The pragmatics of communication refers to the behavioural effects, i.e. in marriage, is a message acknowledged or invalidated? Does it serve a mutually supportive function for partners or is it a source of conflict? Communication theorists also distinguish between "report" and "command" aspects of a particular communicational act. The distinction is between the conveying of information (report) as opposed to those aspects of communication that address the relationship between partners. Thus it is expected that two partners in a marriage must not only send information but define the nature of their relationship to each other. One can imagine two or more messages given by marital partners to each other which have conflicting "command" aspects. Steinglass (1978) cites an example of a wife who recites the events in her day, seemingly accepting the fact that her husband is reading the newspaper as she is doing so. This represents one command aspect; that her recitation is trivial and it is acceptable for her husband to read the newspaper at that time. However, if after 20 minutes, the wife states accusingly, "you haven't heard a thing I said," then an alternative command aspect has been issued; that she is perhaps unwilling to accept a response from her husband which might reinforce his tendency to treat her verbal communication as trivial. If a series of similar exchanges occurs regularly in this marriage, then it might be concluded that a "describable" pattern of communication exists (Steinglass, 1978). Jackson (1965b) believes that such a pattern of communication reflects a rule about the nature of a marital relationship. He stresses the importance of rules in defining the nature of a marital relationship. In the early years of 33 marriage two people bargain to work out the rules that will govern the nature of their relationship. Jackson refers to the "marital quid pro quo" as the initial bargain that is struck between husband and wife. If the marriage contract based on such bargains is flexible, the couple might do well but if the agreement is too rigidly defined the couple might be at risk from the stresses that require change in the communicational patterns (Steinglass, 1978). According to Watzlawick et al. (1967), communication cannot be considered pathological if it occurs in a single act; thus this group is concerned with the sequencing and patterning of communication. The study of patterns of communication centers around the syntactical and semantic qualities of the patterns, and the degree of clarity and confusion as well as the pragmatics. Is the message accepted, rejected or disqualified? An individual may invalidate his or her own or the other individual's communication (Steinglass, 1978). The second main concept related to pathological communication is that of paradoxical communication. Such communication is one that moves in two opposite and internally inconsistent directions at the same time. Although such communication can be considered a regular occurrence, when marital communication patterns take on paradoxical features at critical times, i.e. in crisis, then the situation becomes pathological. A type of paradoxical communication that has been noted as implicated in schizophrenia is the "double bind" (Sluzki & Ransom, 1976). This refers to a communicational situation in which two logically inconsistent messages are simultaneously communicated and there is a third message, an injunction against commenting about this inconsistency. With regard to marital choice, those who are strong proponents of communication theorj' would not likety acknowledge the relevance of unconscious need 34 fulfillment, as in psychoanalytic or dynamic schools or repetition of childhood experience, as in Bowen Theory (Gurman, 1978). Consistent with a social learning view, communication theorists would agree that marital conflict does not follow from existing psychopathology in individuals who marry, but that such conflict is a result of their interactions. Like behaviourists, they do not see the genesis of conflict as being as important as the current organization of the interaction between spouses (Gurman, 1978). Haley (1963), a communication theorist, sees relationships and marriage conflict as largely a process of the struggle for power and control. Conflict is not primarily seen as an attempt to define a relationship but is likely to occur when couples communicate at multiple levels, i.e. "report" and "command." When this happens, paradoxical communication occurs. Since change requires movement to the next higher level of logical type (Watzlawick et al., 1974) and interpersonal systems do not usually have rules that allow for changing their own rules, partners are often unable to resolve their conflicts. Gurman (1978) speculates about the complementarity of the theoretical position of the communicationists, Bowenites and object relations theorists. Klugman (1975a) has suggested overlap, in that "fused" individuals (Bowen, 1976) are synonymous with paradoxical communicators. "Symptomatic behavior may be viewed as a system of communication, one designed to support the delusion of fusion" (p. 4). The purpose of such a style of relating is revealed to be one of protecting (defending) the individual from the anxiety that he or she associates with differentiation of self (Gurman, 1978). Dicks (1967) notes that "what object-relations theory would call unconscious collusion is described by Haley in phenomenological and behavioral terms" (p. 121). Rausch, Barry, Hertel and Swain (1974) argue that the view one holds of one's self and of others, together with the 35 needs and affects characterizing the relationships between the images, set contraints upon the ways in which interpersonal messages will be received. A poorly "differentiated" individual with, by definition, rigid perceptual capacities will have fewer options for processing information. In communication theory "the restriction in variety of any part of the system limits the system ... \and] will limit the ability of the system to adapt to new circumstances" (p. 25). Here the restriction in a part of the system is within the individual. Gurman (1978), in agreement with Klugman (1975a) notes that the mechanism of "collusion" (a phenomenon in object relations theory) requires paradoxical communication for its maintenance and also produces paradoxical communication. With regard to this notion, Rausch et al. (1974) note, "a collusive joint pattern of avoidance seems particularly likely to foreclose the possibility of \a] shift to a metacommunicative level" (p. 80). Whereas communicationists conceptualize interaction on the dyadic level, Bowenites and object relations theorists conceptualize it on the monadic level (Gurman, 1978). The Process of Change in Communication Theory. Haley's (1963) central focus in marital conflict is on the power struggle between spouses, the modification of rules on boundaries and the management of power and authority (Sluzki, 1983). Bateson (1979) sees the use of these terms as reflecting a linear epistemology rather than a cybernetic world view. The Palo Alto group focuses on the inadvertent, problem-maintaining solutions that people engage in. One of the key principles that the MRI group adheres to in working with clients is that it is something in the clients' "attempted solutions," the ways they try to alter a problem, that contribute most to the problem's maintenance or escalation (Fisch et al., 1982). Problems are maintained inadvertently and persistence in inappropriate handling of difficulties does not necessarily require defects in family organization (as 36 in structural theory) or mental deficits in the participants (as in psychoanalytic theory). Conflict occurs when spouses do not differentiate between a "difficulty" and a "problem" or when "the solution becomes the problem" (Watzlawick et al., 1974). Watzlawick et al. (1974) note that it is the meaning attributed to a situation and therefore its consequences, but not its concrete facts that is problematic. Conflict between spouses is the result of attributional deficiency or cognitive misconceptions. "This distinction between facts and premises about the facts is crucial for an understanding of change (Watzlawick et al., 1974, p. 54). Gurman (1978) points out that the premises of these authors seem analogous to the "irrational assumptions and categorical imperatives ("should," "ought" and "must") that form the core of the rational emotive model of psychopathology (Ellis, 1973) and with the "implicit assumptions," "arbitrary influences" etc. of Beck's (1976) cognitive approach. Gurman (1978) notes, however, that the therapeutic interventions that follow from these models, i.e. developing awareness of and modifying thoughts and belief structures, are very different from those of the Palo Alto group. If, as the communication theorists state, problem formation and maintenance are seen as part of a vicious circle process in which well-intended solution-behaviours maintain the problem, then alteration of these behaviours should interrupt the cycle and initiate resolution of the problem (also referred to as the "symptom"). The therapist's primary aim, then, need not be resolution of all difficulties but merely the initiation of a reversal. In order to do this, not only must the therapist get a clear view of the problem behaviors and the behavior that functions to maintain it; he (or she) must also consider what the most strategic change in the 'solutions' might be and take steps to instigate these changes-in the face of the client's considerable commitments to 37 maintaining them. (Fisch, Weakland & Segal, 1982, p. 19) The communication therapist's view that the problem is the attempted solution and that problems are maintained inadvertently through negative communicational cycles, has led to the notion that "awareness" of intrapsychic or historical factors is irrelevant to the therapeutic process. Haley (1976) comments, " . . . the main goal of therapy is to get people to behave differently and so to have different subjective experiences" (p. 49). Gurman (1978) adds, The basic assumption of communicationists in this regard is that when change has been achieved in one domain of experience, it radiates to all other domains, from the behavioral-interactional to the subjective. This assumption also appears in analogous form in the standard contention by systems-oriented therapists that a therapeutic focus on process is superior to a focus on content. This assumption is explained by reference to the general systems notion of equifinality, that is, that no matter where one begins, the conclusion will be the same. (p. 529) This outlook is manifested in the systems-oriented therapist's lack of attention to historical events. Foley (1974) articulates this, "Whether the subject is money, sex, children, or in-laws, the pattern will be the same . . . the clinician need only get some idea of the couple's interaction in a given area to understand how they relate" (pp. 42-43). Gurman (1978) has pointed out that these views are similar to those of behaviourists who have regarded feelings as epiphenomena of overt behaviour change but in the context of marital therapy, no longer assume such a correspondence. Research is cited by Strupp and Hadley (1977) that has demonstrated that feelings of well-being and reliably counted behavioural changes do not show consistently high positive correlation. Consistent with this view, Sluzki (1978) writes: ... a system-oriented therapist will take into consideration effects rather than intentions . . . the effects of behavior will be carefully noted, while, on the contrary, no inference will be made about the motivations of the participants. Even further, issues about the 38 motivation or intention will be considered irrelevant to the understanding of interpersonal processes. This choice must not be taken as a blunt denial of the existence of motivation, intentions or human volition in general. It only happens that those inferences do not add any relevant information for purposes of conceptualizing and/or treating marital disorders from a systems perspective. (p. 367) System change is seen as necessary for change to occur, both in the couple interaction and within the individual (Gurman, 1981a). Here the treatment focus is on the "symptoms" or reactive interactional cycles of behavior that have developed between partners. Negative interactional patterns between husband and wife are assumed always to have interpersonal meaning or to function as communicative acts so that one partner cannot change unless the system changes. A sustained, present-centered focus on the couple interaction is, therefore, provided by the therapist (Wile, 1981). The therapist works with the couple to change these negative cycles. The mechanisms for change in a communications approach to marital and family therapy stem from a paradox paradigm and, in general, involve reframing or positively connoting the couple or family's symptoms and/or system, followed by making interventions centering around a suggestion of "no change," i.e. prescribing the symptom or negative interactional pattern. Weeks and L'Abate (1982) state that, based on a paradox paradigm, symptoms need to be understood dialectically, in terms of opposition. The major implication of dialectics for labeling is that the therapist must change the emphasis from finding or inventing pathology in clients to focusing on strengths. Like Haley (1980), these authors note that ascribing negative labels tends to perpetuate behaviour through the self-fulfilling prophecy. The label(s) also tend to generalize to people in overall ways and influence how clients behave, as well as others' perceptions of them. In order to focus on the positive aspects of symptoms, the therapist must understand how they have been 39 adaptive for clients, i.e. speculate as to what function they have been serving in the family or marital system. Once the positive function of the symptom has been identified it can be seen as a vehicle of change even though its function has also been one of precluding change in the system (Weeks & L'Abate, 1982). Reframing and relabeling, then, are seen as ways of facilitating positive therapeutic outcomes. Watzlawick (1976) has defined reframing as: "to change the conceptual or emotional setting in relation to which a situation is experienced and place it into another frame that fits the 'facts' of the same concrete situation equally well or better and thereby change its entire meaning" (p. 122). He notes that effective reframing consists of a successful change of the second order frame of reference or reality and is based on the communicationist idea that there is not some "true" underlying problem but that the problem lies in how individuals view things. Reframing is a broader term which refers to a change in meaning attributed to the situation, while relabeling refers to changing the label attached to the individual or problem without changing the frame of reference and is subsumed under reframing (Weeks & L'Abate, 1982). A positive label applied to a disturbing behaviour implies to individuals that they have been given both permission to have that behaviour and an expectation of positive change. Watzlawick (1976) stresses that successful reframing must be communicated in a "language" that is congenial and acceptable to the client's way of conceptualizing his or her world or second-order reality. The therapist's ability to "adopt" the client's view of reality is considered very useful in this approach. In this way "resistance not only ceases to be an obstacle but becomes the royal road to therapeutic change" (p. 123). The author conceptualizes paradoxes (therapeutic double-binds) as the next step in making reframing acceptable or even compelling to a client. 40 Watzlawick, Beavin and Jackson (1967) have defined the therapeutic double-bind as the mirror image of a pathogenic bind. Within the psychotherapeutic situation, an injunction is given that creates "paradox" because the client is told to change by remaining unchanged. This puts the client into an untenable situation. If he or she complies, he or she no longer "can't help it." The symptom is then under the client's control, which is the purpose of therapy. If the client resists the injunction, he or she can do so only by not behaving symptomatically, which is also the purpose of the therapy. Weeks and L'Abate (1982) point out that reframing has the same purpose as paradoxical injunctions; the only difference is that the former is implicit and the latter is explicit. Whereas a paradoxical directive may involve telling partners, for example, quite specifically to do more of the same or remain unchanged ("prescribing the symptom"), reframing carries the same message but it is implicit. This is true because a so-called "undesirable" behaviour, given a positive label, seems to be desirable. A classification scheme for paradoxical interventions has been developed by Rohrbaugh, Tennen, Press and White (1981) and sheds some light on the process. They refer to "prescribing," "restraining" and "positioning" operations. When prescribing, the therapist encourages or instructs the client to engage in the specific behaviour to be changed. In restraining, the therapist discourages change and may even deny that change is possible. Restraining embodies the message, "you probably shouldn't change" (p. 456). The therapist may tell the client to "go slow" or discuss with him or her the possible dangers of improvement. Paradoxical positioning is exemplified in the work of the Palo Alto group. Here the therapist attempts to shift a problematic "position"-usually an assertion that an individual . . . is making about himself or his problem-by accepting and exaggerating that 41 position. This intervention is used when the person's position is assessed as being maintained by a complementary or opposite response by others. The authors cite an example of a "depressed" person who has been encouraged to "cheer up." Because this can have the opposite of the intended effect, the therapist positions himself or herself by agreeing with and exaggerating the client's view of the problem, perhaps saying, "Considering your situation, I am surprised you are not more depressed than you appear to be" (p. 463). The authors recognize the potential that this strategy has for making things worse with certain more suggestible individuals (e.g. hysterics) and emphasize the importance of careful assessment. They also stress that taking such positions is indicated more when clients seem resistant. Rohrbaugh et al. (1981) distinguish between "compliance-based" and "defiance-based" strategies. It is noted that some paradoxical interventions are effective because the client attempts to comply with the therapist's directives (compliance-based), while others work because the client rebels against the therapist's intervention. Compliance-based interventions are more likely to work with symptoms such as obsessions, anxiety attacks and somatic complaints which are maintained to some extent by the client's attempts to be rid of them. Defiance-based strategies are based on the expectation that the client will react to a suggestion or directive. Haley (1976) suggests that this type of intervention is meant to influence the client to change by rebelling. Minuchin and Structural Family Therapy. Minuchin's (1974) model is based on three major assumptions: 1) the individual operates within a social context and it is this context that defines the constraints within which individual behaviour exists. Thus, there is a never-ending interaction between the individual and his or 42 her environment which mutually affects and influences each other. 2) This social context can be seen as having a structure. 3) Some structures are good and some bad (Steinglass, 1978). Minuchin (1974) believes that it is possible to understand the structure of a marriage if three specific dimensions are observed: 1) Organizational characteristics (membership, systems and subsystems, and boundaries). The marital dyad is seen as a subsystem within the family which is composed of two individual subsystems. Borrowing from pure systems theory, Minuchin (1974) states that the boundaries of a particular system are "the rules defining who participates and how" (p. 53). The function of boundaries is to protect the differentiation of the system. In order to grow, marriages must have clear boundaries. These insure that partners are clearly defined as a separate system to be protected from interference from competing subsystems such as in-laws or children. However, boundaries must not be so rigid as to prevent interaction between the marital subsystem and the outside world, or between husband and wife. In evaluating a marriage's level of function, Minuchin identifies three general types of boundaries: a) disengaged, b) enmeshed, and c) clear boundaries which he proposes exist on a continuum. Although Minuchin states that these do not refer to a qualitative difference between functional and dysfunctional kinds of boundaries but instead to a transactional style or preference for a certain type of interaction, Steinglass (1978) notes that his theory implies that clarity of boundaries makes it easier for a marriage to thrive. 2) Patterning of transactions. This dimension is somewhat similar to homeostasis in the MRI model, but it is more diffuse in the data base it uses. 43 The focus of the MRI model is on specific communicational acts whereas Minuchin pays particular attention to the relationship between context and behaviour. "Transactions are not merely communicational acts between transmitters and receivers; they also include intricate interrelationships between environmental contexts and individual behavior" (Steinglass, 1978, p. 328). In this model, therefore, patterned transactions are conceptualized more in spatial than temporal terms and it is the relationship between different variables in space, rather than the sequential order of their occurrence, that becomes critical in making judgements. These sequences of interactions between husband and wife are evaluated to determine the juxtaposition of different functional roles within transactions. The relationships, first between context and behaviour and then between husband and wife, suggest the transactional structure within the marriage. Like Sluzki (1978), Minuchin (1974) does not deny that motivation occurs but indicates that it is not necessary to describe behaviour in motivational terms. 3) Response to stress. Although there are four potential sources of stress on the family (interaction between individuals and extra-familial forces, interaction between the family and extra-familial forces, developmental transition and idiosyncratic sources), an underlying family structure can be identified in the common patterns of adaptation to stress that emerge. One type of structure can lead to adjustment and the other to rigidity; however, the emphasis is on parts that fit together in a particular organizational pattern (Steinglass, 1978). Steinglass (1978) notes that many of Minuchin's concepts are somewhat vague. Minuchin's emphasis on structural concepts leads to a style of therapy which deals largely with the "here and now." In summary, the author notes, "The emphasis is on maps in which structural variables are represented in a 44 spatial dimension only; past history, although interesting, does not . . . have a logical or consistent role in conceptualization of normality or pathology in this model. The emphasis on stress and adaptation to stress is a logical extension of the conceptual reliance on structure" (p. 330). Milan Therapy. Between 1965 and 1967, Selvini-Palazzoli became increasingly disillusioned with the psychoanalytic method and started to become involved with whole families. She was sooned joined by Boscolo, Cecchin and Prata and thus began the Milan group, which applied a variation of the "behind the mirror" team approach to family therapy. Although they have concentrated on whole families rather than the marital dyad, many of the principles of Milan therapy are adaptable and central to the marital systems therapy under study here. Like the Palo Alto group, their epistemology is based on systems theory, information theory and cybernetics (Bateson, 1972). Selvini-Palazzoli et al. (1978) emphasize circular rather than linear causality, which is considered more "useful" in clinical work with families. Within a circular framework it is assumed that individuals and objects have or "show" characteristics in relation to the contrasting characteristics of others (one person is old only because another is considered younger). A difference is always considered a relationship between whatever is being compared and such relationships are always reciprocal or circular. If one member of a family is defined as attacking, then by definition this implies that others are good. A circular orientation allows implicit information to become more explicit and offers alternative points of view. In contrast, a linear orientation is restrictive and tends to mask important information. Similar to the Palo Alto group, Milan therapists believe that rather than being limited to the constraints of content, intention and ongoing behaviour, clinicians who use this approach are able 45 to think more broadly in terms of content, reciprocity, effects of behaviour or beliefs, connections between behaviours and patterns that form self-perpetuating loops (Selvini-Palazzoli et al., 1978). Weeks and L'Abate (1982) note that the "logic" used by a paradoxical therapist is circular rather than linear but that our language basically consists of linear and causal, subject-predicate sequences, e.g. If —, then This makes thinking and articulating about circular processes difficult and the Milan group (1978) refers to this as the "tyranny of linguistic conditioning" (Bateson, 1972). The team has been able to maintain circular assumptions by substituting the verb "to be" in their conceptualizing about families. To say that a husband "shows" depression implies a systemic perspective, that the behaviour may be related to the interaction and allows evaluation of his behaviour in relationship to his wife. However, saying that he is depressed is to imply attribution or internal causes. Parallel to the Palo Alto group, the Milan team is interested in the message conveyed by the symptom rather than the symptom itself (Selvini-Palazzoli et al., 1978). The Milan team's assumptions about the nature of truth are also circular or pragmatic, as are those of the Palo Alto group. The Milan group points out that the statement that is most "true" at any particular moment is that which is most useful. When assessing the meaning of behaviour, there are only different "punctuations" or points of view, no certainties. They use the word hypothesis rather than truth in order to convey this. For the clinician, statements are considered to be pragmatically true when they are useful in facilitating constructive change with clients. The usefulness is judged on the basis of feedback, i.e. the "truth" of the therapist's statement is based on the responses of the family and how useful it is in helping the family to make constructive changes. For example, 46 referring to a person in a marital dyad as "depressed" may not only not be "useful" but may also be harmful and an obstacle to change; it may in fact allow one or both partners to attribute the problem to the "depressed" spouse and so prevent change. In order to facilitate constructive change it will be most useful for the therapist to understand and point out how the symptom (depression) has been useful or adaptive for both husband and wife. Thus, as in communications approach, the therapist focuses on the clients' strengths and the positive rather than negative aspects of the problem. Accepting most of the premises of general systems theory, Selvini-Palazzoli et al. (1978) see the family as a "self-regulating system which controls itself according to rules formed over . . . time through a process of trial and error" (p. 4). They adhere to notions that the whole is greater than the sum of its parts, change in one part will affect every other, and the whole manifests a totality with tendencies to homeostasis, equilibrium, transformation and equifinality. More than some other family therapists, however, they stress the capacity of family systems to change on their own. While systems therapists in Western philosophical tradition tend to see the stable organization of components as central to the systemic perspective (morphogenesis and morphostasis), the Milan team are more similar to Eastern philosophies, in that they tend to see systems as representing an ongoing process of everchanging interconnectedness (continuous (homeostatic) fluctuation and discontinuous transformation) (Tomm, 1982). These differences are implicated at practical levels. For example, some structural and strategic therapists look for patterns of interaction that reveal entrenched patterns of organization and try to realign the system in more adaptive ways, i.e. through defining clearer boundaries, breaking up cross-generational alliance, restoring hierarchies, etc. The Milan therapist, however, 47 while assuming that the family system only appears to be stable and that the rest of the system is changing around its "stuckpoints," attempts to identify the point(s) at which the system seems to be stuck. He or she then intervenes to introduce new connections or a new time factor at these points so that the system may be freed to continue to change spontaneously (Selvini-Palazzoli et al., 1978). Cronen, Johnson and Lannamann (1982) describe the stuck point in a system as that point at which particular ideas, meanings, or beliefs are connected and locked into a paradoxical tangle or loop. The idea of the therapist intervening briefly, and not to change the whole system is, however, parallel to Fisch et al.'s (1982) characterization of the therapist's primarj' aim which is to initiate a reversal that will interrupt the problem-maintaining behaviour cycle, rather than attempt to change the whole system. Milan therapists believe that time is required for new information to reverberate through the system, for reciprocal feedback between members and for the impact of the intervention to be assessed. Thus monthly, rather than weekly, sessions are common (Selvini-Palazzoli et al., 1980). When families don't seem to be changing the therapist may "broaden" the "field of enquiry," often bringing in significant others, perhaps, as Bowen does, members of the family of origin. The Process of Change in Milan Therapy. The main catalyst for change in Milan therapy seems to be the intervention which occurs after the interview with the family. This intervention represents the conclusions of the team, which are delivered by the therapist as a message from the team. Team members, who have been behind the mirror during the session, note family reactions to the message. The conclusions of the team are based on a brainstorming of team 48 members following the session, where the emphasis is on elaborating a systemic understanding, i.e. What effect is the submissive behaviour of mother having on father and the children? The message may be in the form of an opinion, a prescription for "no change," a declaration of impotence, ritual prescription, etc. (Selvini-Palazzoli et al., 1978). An important technique which corresponds to the reframing process in other systemic therapies is positive connotation (Selvini-Palazzoli et al., 1978). It is believed that when the behaviours of the family members (including symptomatic behaviour) are connected in a circular fashion and connoted positively as beneficial for the family they are more easily accepted. An important purpose of positive connotation is to allow new information to gain entry into the family system. Positive connotation also serves to legitimize the prescription of "no change" that follows. This parallels the sequence of reframing negative interactional cycles and then prescribing the symptom (no change) which is followed by the Palo Alto group. If a particular pattern of behaviour is construed as being a good thing, then it follows that it should continue. Kraemer (1982) has pointed out that systemic therapy is a "fiercely disciplined attempt to view the family in the same manner as Spinoza urges us to view Nature, that is, as a system that, when understood clearly, must be accepted as it is" (p. 354). In Milan therapy, an additional phrase like "for the time being" is always added to a "no change" prescription (Tomm, 1982). This phrase is very important since it implies that future patterns of behaviour may be different and also introduces an element of time. What is being positively connoted is the family's homeostatic tendencies. The connections made in the intervention suggest alternative solutions and possibilities for change. "As a result, the therapeutic paradox is not binding but is implicitly open" (Tomm, 49 1982, p. 26). The therapist delivers the message briefly, thus introducing new information to the system, and thus avoids further interaction and leaves the family to make sense of the message. Selvini-Palazzoli et al. (1978) do not assume that they know the solution to the family's problem or that they are able to tell members how to be free of problems. The premise is that the family system has transforming potential and that the intervention will suggest change at some deeper level of meaning (comparable to Watzlawick's (1976) notion of second order change) as opposed to the concrete or content level. When a task is prescribed, it is often not important whether or not it is carried out. It is the reaction of the family members to the instructions and the implicit ideas that are the change agents. The impact of a ritual derives from its implicit challenge to a covert ritual or myth in the system. It is considered possible that a second catalyst of change in Milan therapy could be the interview process itself and Selvini-Palazzoli et al. (1980) have conjectured that if the interview were conducted "properly," a formal end-of-session intervention or message might not be required. The information elicited through the style of interviewing might be enough to induce a transformation. However, it may be that the new information and connections generated in the interview need to be "consolidated" by the intervention. The style of questioning used in the interview is called "circular questioning." It is based on the principle of circularity as elaborated earlier. Circular questioning refers to a style of questioning about differences in relationships that is designed to elicit information about circular processes. Three basic types of difference questions are asked: a) differences between family members on an issue or characteristic, b) 50 differences between relationships and c) differences between individuals or relationships over time, i.e. "Were Father and Mary closer or more distant after grandmother died?" The kind of information sought through such questions is guided by the hypotheses of the team. Questions framed in such a manner tend not to engender client resistance, as do more direct questions. For example, a question such as, "Is Mother closer to Jim or Nancy?" would seem to allow those involved to be less concerned with the intent of the question or about giving a desired impression than a question such as, "Are Mother and Jim close?" In addition to the principle of circularity, the two key principles of hypothesizing and neutrality are considered essential by the Milan group (Selvini-Palazzol et al., 1980). 1) Hypothesizing: the team conceptualizes about the family based on their observations as well as the family's responses to the interventions. They are viewed not as true or false but as "useful" or "not useful." Hypotheses are related to the connections that are postulated between events, behaviours, meanings, etc. and track relational patterns. In general, they address the question, "Why is this family coming for therapy at this time?" Hypotheses are not stated because the system "requires" that members not be aware of the most relevant connections and also because the suggestion would be met with denials and resistance would be strengthened. (This is similar to the Palo Alto group's ideas that awareness is not relevant or even therapeutically advisable.) Behaviour sequences are tracked, enabling the therapist to build his or her own inferences rather than focusing on thoughts or feelings which are more inferential from the clients' perspectives. 2) Neutrality is based on the epistemological assumption that a system can be best understood through the examination of each individual's perception of the differences. In order to obtain 51 information, the therapist must be perceived as a neutral figure, not taking sides with any members and resisting pressure to become involved in the family's linear view of the problem. He or she must give each member equal time, never agreeing with one and not another. Questions should be constructed in such a way as to avoid implying a moral or ideological stance. If they are not sure where the therapist stands on an issue, family members are more likety to give more accurate information rather than to offer responses that they think will make them look good to the therapist. The Milan group notes that, in essence, the basic paradox presented to the therapist by the family is, "Here is the individual who has a problem and needs to change-but our family is fine and we expect to remain the same." Systems theory however postulates that change in one part requires some complementary change in the whole—the therapist's task, therefore, is to counter the family's paradox. His or her response may not necessarily be paradoxical but often is. For example, an injunction not to change is paradoxical in that the content message (lower level or first order logical type) occurs in a social context (higher level or second order logical type) that specifies that the therapist's function in the community is to help people to change, thus "Paradox and Counter Paradox," (Selvini-Palazzoli et al., 1978). Ethical Considerations in Communication and Milan Therapy. There has been some criticism of paradoxical strategies on ethical grounds; it has been viewed by some who consider it deliberately manipulative as less desirable than interventions which are designed to influence clients more directly. Rohrbaugh et al. (1981) assert that paradoxical strategies seem to be least applicable in situations of crisis or extreme instability—at those times, clients are usually more amenable to 52 direct influence attempts and can best be helped by the therapist offering structure, taking control or in other ways stabilizing the situation. The authors seem to assume that if clients are more vulnerable they will not be resistant or that if they intend to comply it will follow that they will change. The idea of the therapist working to stabilize such situations is an appealing and common sense proposition but questions pertaining to how and in what sequence this can best be done are raised. Kraemer (1982), in suggesting similarities between the philosophical models of Spinoza and systemic therapists offers speculations that seem applicable. Kraemer notes that much is made of the paradoxical and playful elements in the prescription and that "some fear that it is mischievous and unethical to instruct people to continue their disturbed behaviour, because that seems to require further suffering which should, they imply, be unnecessary now that help is at hand . . ." (p. 355) but concludes that "the principal task is neither to be paradoxical nor playful, but to discover and accept the family as it is, paying particular attention to the function of the symptom. To propose that it can be maintained for the time being is simply the most effective way of communicating that attitude. The therapists temporarily abandon any desire to change the family" (p. 355). Kraemer states that he is suggesting that the ethical propositions of Spinoza are a first step in an ethics of systemic therapy and that the philosopher has argued that distress is diminished by seeing that it is "natural and necessary" and that loss could not be avoided. He adds: By reframing a loss as inevitable and a pitiable state as natural and necessary, Spinoza has touched upon one of the central disciplines of systemic therapy, which is always to be attentive to the necessities in the system as it presents itself. This is paradoxical only because the family is expecting to be told that their problem is not necessary and can therefore be removed. The family in therapy must always be 53 seen as a natural . . . system, however disordered or perverse. Accept anything less and the therapist will lose sight of the function of the symptom and will begin to think it is "unnatural" . . . Then he or she will be paralyzed by it just as the family is. (p. 355) Kraemer notes that Bateson (1979) expresses similar views, "A sort of freedom that comes from recognizing what is necessarily so. After that is recognized, comes a knowledge of how to act" (p. 219). This seems to reflect a deep respect for natural systems. Without this, the systemic view and particularly positive connotation would only imply "weak and indiscriminate approval" of anything the therapist happens to notice. "On the contrary," Kraemer suggests, "the idea is to notice the specific and essentially secret contribution that the presenting problems make to the family's survival" (p. 356). Issues in Integration of Models A number of family therapists have espoused some form of integration of models, some at a theoretical level, but most often at a practical level. As one approach to integration, some therapists have advocated a shift during the course of therapy from the interventions suggested by one model to those suggested by another. Other therapists have argued against integration, or even a mixing of models, either at the level of theory and/or practice. Incompatibility of Models Some have stated directly or implied that the models are incompatible. Rohrbaugh (1984), who is opposed to integration, asserts that the hazards-for clients, therapists and systemic therapy theorists-outweigh the benefits and points 54 out that while the models may share certain premises, they present irreconcilable views on the problem that therapy seeks to change. At the practical level he believes that the mixing or shifting of models changes and dilutes therapeutic focus, with the resulting risk being longer treatment. According to Colapinto (1984), true integration can be distinguished from eclectic attempts to adhere to two incompatible models, the latter being indefensible because it requires that therapists shift basic assumptions as well as techniques. Starting from the level of theory, Rohrbaugh (1984) points out that different therapists see either different systems as significant and "things in the bushes" or problems as significant and this, necessarily, implies differences in "where" or "how" to intervene. According to the author, "Several of the formulations find pathology in how relationships are organized; others do not" (p. 30). While some therapists (Madanes, 1981) (Milan) assume that the symptom is functional; another (MRI) does not. One requires a great deal of hypothesizing (Milan); another eschews inference entirely (MRI). One construes context very narrowly (MRI); another places virtually no limit on what is relevant (Milan). Rohrbaugh concludes that these therapies do not fit together theoretically and that since theory determines what we do, it should be taken seriously. According to Rohrbaugh (1984), at a practical level, the mixing or shifting of models in addition to diluting therapeutic focus, makes the task of providing clear and consistent action or intervention much more difficult. Of those therapists who are generally opposed to integration, some have proposed a "fit" between the type of therapy and the type and social context of the family. McKinnon, Parry and Black (1984) conclude that "the different and often incompatible positions that the three strategic approaches (MRI, Haley/Madanes & Milan) take on major theoretical and methodological issues appear related to the 55 evolution of each approach in different social contexts and with different problems and family types" (p. 21). The authors note, for example, that the MRI approach is likely to best address neurotic conditions of adults in middle to upper class families and interpersonal conflict, particularly between parent and adolescent. The directive approach of Haley, which focuses on inadequate hierarchies and coalitions, is seen as adapted to "centrifugal" families (those with diffuse external boundaries, corresponding to Minuchin's (1974) "disengaged" category) which exhibit behaviour disorders. The metaphoric approach of Madanes appears adapted to centripetal families (more rigid boundaries, tight knit or "enmeshed" according to Minuchin's category), where dysfunctional behaviour is seen to emanate from the rigidity of repressive rules or the functional nature of the symptom. An approach that examines the family's epistemology (Milan) is perceived as well adapted to tightly knit externally closed systems (centripetal) in that it addresses the underlying false assumptions and (through neutrality) escapes the inevitable disqualification of the therapist. A more tentative stance toward "integration" is taken by Coyne (1984), but he warns that attempts to shift back and forth or borrow fully from one approach while working within another in a single case, will likely lead to problems with continuity and coherence. The author notes that the style of information gathering and intervention of each model sets a context for therapy and "the choice of a particular intervention should take into account not only its immediate impact but its influence on the therapeutic context" (p. 27). He concludes that, overall, it is probably easier for the therapist to shift from a one-down (strategic) position to a one-up (structural) position and that the decision to take a more structural approach after having worked strategically may be difficult to reverse. 56 As an alternative to an integrative approach, de Shazer (1984) offers the notion of fit, contending that both the structural and strategic therapist have been successful in developing techniques which "fit." Fit is used here in an analogous fashion to that of the capacity of a key to fit a lock and refers to the capacity of the therapist to understand, accept and incorporate the family's frame of reference and terms, and to provide a course of therapy which creates positive change. The author believes that structural and strategic therapists each do effective therapy when following their own models and that having these two separate, well-defined models affords a better depth of understanding and perspective from which to evaluate situations. He also notes that as full integration might leave the therapist with "just one eye and, therefore, a lack of depth perception," (p. 36) he advises against attempting it. Toward an Integration at the Theoretical Level In evaluating the "current move toward 'eclecticism,'" Fraser (1982) examines the relationship between structural and strategic schools in a conceptual and practical framework. While he rejects the notion of integrating or intermingling "these two often divergent views" (p. 18) at the level of therapy, the author departs from the views of the authors cited above in that he finds some basis for integration at the theoretical level. With reference to theory, he notes that both structural and strategic schools see families as rule-governed systems which operate with tolerance limits for departure from those rules. Fraser points out that both perspectives outline nearly identical reasons for system dysfunction: "Action is not needed but is taken by the system and 2) action is needed by the system but is 57 not taken" (p. 14). According to Watzlawick et al. (1974), a third option is when action is needed but taken on the wrong level. Fraser (1982) also points out that both positions acknowledge the operation of "negative feedback loops" at a basic level of family functioning. A first principle in family functioning is the maintenance of homeostasis through the initiation of negative feedback loops when a family member activity exceeds tolerable limits. A "negative feedback loop" is defined as one in which system action is directed toward the goal of reducing deviation from a given set of tolerance limits. The regulation is the critical element of the negative feedback loop. He cites the following as an example, In a family system when a small child straj's too far from the safety of its porch, we may see a watchful parent call or go to retrieve the child. Once the child is within the "safe" limits of the yard and porch again, the parent may again return to their previous activity. Dysfunction may be viewed when this process continues after the child has reached adolescence. The dysfunction is ... a product ... of the rigid adherence to overfy restrictive tolerance limits. The adolescent is being given messages to return and stay within the home when his developmental task is to differentiate and leave. What is dysfunctional is that the tolerance limits haven't changed to match the evolution of the child-in-the-family. . . . The punctuation of the cycle . . . may be altered to begin with any given element. It is cyclical (pp. 15-16). As stated, both structural and strategic schools acknowledge the operation of these negative feedback loops at a basic level of family function. However, a difference becomes apparent as the two address the nature of system dysfunction. The structural school sees dysfunction as a rigid homeostatic transaction which must be broken, i.e. too much enmeshment or disengagement (Minuchin, 1974); in other words, either rigid or overly permeable boundaries. The example given above of the adolescent subjected to too much control is one of overly rigid limits. Conversely, the strategic school sees the dysfunction as emanating from repetitive cycles of first-order changes (changes of intensity, duration, etc. within a given set 58 of rules or premises) that are viewed as "positive feedback cycles." Fraser notes, "a positive feedback loop process is the 'other side of the system coin' from the negative loop, homeostatic process. Where negative loops decrease deviation, positive loops are deviation-amplifying processes ... a repetitive solution which exacerbates rather than solves the dilemma" (p. 16). The example cited is of a marriage where one partner is judged not to be assertive enough. One spouse may ask for more independent and assertive action from the other, yet when the other responds, it appears only to be more dependent compliance with the dissatisfied spouse's requests. If more independence is now demanded, the other spouse is in a "be spontaneous" paradox where he or she can only be independent by rejecting the other's demands and remaining dependent. If he or she attempts independent action at the other's request he or is is only again demonstrating more compliance and dependence. The only way for the cycle to be reversed adequately is for the original solution to be retracted. Once the other spouse is not directed to be independent and assertive he or she can then at last exercise truly independent activity. A vicious cycle may therefore be replaced by a virtuous one, and this new, now positive deviation may then be amplified, (p. 16) The central point of the strategic view is that the only way that the cycle can be adequately reversed is for the original solution to be retracted. The key difference between strategic and structural is that, whereas the structural therapist's task is to break a rigid, negative loop, homeostatic structure, the strategic therapist sets out to interrupt a positive loop, vicious cycle. The strategic school sees change as occurring through the creation of beneficent or virtuous cycles. As stated, the strategic therapist's view of system dysfunction is based upon repetitive cycles which are conceptualized as "escalating positive feedback cycles." Thus a cyclical model of dysfunction is developed which refers to a repetitive solution which exacerbates rather than solves a problem. Resolution occurs through less rather than more of the same solution or a change in system rules, as in the example above, where the spouse is not directed to be independent and assertive. According 59 to Fraser, "the therapist produces change by creating a logical leap, or rule change within the system, such that actions based upon the new rules will create successive deviation amplifications in a virtuous cycle of problem resolution" (p. 17, 1982). A difference that emerges in the two schools as a result of this discrepancy in how dysfunction and in turn, change, is perceived. This is highlighted by Fraser when he notes that, whereas the negative loop level (seen by the structural therapist as central to change) focuses upon shorter time spans involving discrete trials of homeostatic transactions, the positive loop level (seen by the strategic therapist as central to change) takes in longer time spans, which include multiple instances of the negative loop level transactions. At the positive loop level, the homeostatic adjustments lose their sharp definition and the focus is more on the outcome of transactions between family members. Consequently, the focus of the two schools is different—particularly with regard to "time sequence" and "level of analysis." Whereas the structural therapist will tend to focus more on short time periods in the life of the family, the strategic therapist will look at transactions between members over longer time periods. The structural school tends to focus upon the whole system as it is operating in the present. The therapist "maps" the system and tries to break up what are perceived as rigid, dysfunctional structures through altering subsystem boundaries; the goal is to create new transactional patterns. (A father and daughter would be an example of a subsystem within a family system and the therapist seeking to alter these boundaries may try to break up a coalition between father and daughter.) Fraser states, "the Structural view's emphasis upon structural transactions around a homeostatic base may be a consequence of their intense and immediate engagement 60 with as much of the system as possible to determine, on a here-and-now basis, the nature of family rules. At this close distance, structure stands out, as does the homeostatic level process" (p. 18). Thus, the emphasis is on a time-limited, intense analysis of family transactions, which draws attention to the importance of role relationships and striving for homeostatic balance. The Strategic school, however, following an often retrospective analysis of transactions reported by one or a few system members, studies behaviour over a longer time period. In this process, the Strategic therapist loses the definition provided by the structural therapist's focus on discrete role structures. The strategic therapist does not see the transaction with an entire family system or its subsystems as a critical part of the process. Perhaps meeting with only one or a few members, the therapist focuses on those most heavily engaged in repetitive attempts at a solution to the problem, as targets for intervention. Fraser notes, "the strategic perspective's emphasis upon overall rule change and positive feedback processes may be a consequence of their belief that system information may be collected in retrospect from one or two members. Accurate or not, such data are based upon an analysis of the outcome of many separate trials, and at this level, the structural characteristics lose some salience, as do the negative loop transactions. The result is an analysis of the problem seen in "positive loop terms" (p. 18). Despite these discrepancies between the two schools in perception of dysfunction as well as focus of intervention with regard to time sequences and level of analysis, Fraser speculates that a model for integration can be found in Bateson's (1979) "alternating evolutionary ladder" across time (p. 17). The negative loop levels are seen as discrete iterations each time. The positive loop level 61 connects these negative loop levels by acting upon the outcome of each to modify the next one." (p. 17). Fraser (1982) concludes that the process is the same and that the major difference between these two schools is the time sequence and level of analysis sampled by each school. Pointing out that although each leads to a different set of therapist actions with different subsets of individuals, he argues that similar results might be expected from each process. Fraser's (1982) basis for a rejection of integration at the practical level relates to 1) the notion of the self-fulfilling prophecy and 2) the principle of second order change. Regarding the self-fulfilling prophecy, Fraser believes that the therapist's initial choice of interventions "shape and are shaped by the family system" (p. 19). Fraser supports Minuchin (1974) in his assertion that: "whatever approach (the therapist) chooses in attempting to restructure the family will have an impact upon the family's response to him (or her). It will open certain pathways of intervention and close others" (p. 97). Regarding second order change, Fraser likens a therapeutic shift from structural to strategic approaches to a major change of therapeutic rules for both the therapist and family. In responding to Stanton's (1981c) suggestion that a therapist can shift from a structural to a strategic approach according to the situation's demands, Fraser has emphasized the potential difficulty of this shift. To support this, he points to the difficulty that social systems have in making such shifts, stating that it is impossible to employ both methods simultaneously and extremely difficult to alter them sequentially. With reference to such a sequential process, he translates Stanton's approach into "do what is logical, direct, and easy first; then if it doesn't work, do something crazy or different"; and further on, "this view of shifting between approaches even at a point of failure . . . may be 62 more difficult than is implied . . . this attempt at integration appears not to be that at all, but instead to be a statement that basic change in a therapist's rules at a point of failure ma}' yield creative problem solving" (Fraser, 1982, p. 20). The suggestion by both Fraser (1984) and Minuchin (1984) that the initial choice of intervention will have a lasting impact on the family and is shaped and will be shaped by the family system seems important. However, as suggested in the Sequentially Integrated Systemic approach outlined in this study, choosing other pathways and sequentially altering interventions following the initial choice of intervention seems appropriate. The model for this is similar to Sluzki's (1983). Sluzki notes that the practice guided by the three models-structural, strategic and Milan-also utilizes therapeutic interventions from other models. While he acknowledges that each of these models provides conceptual rationale to a specific set of interventions, Sluzki questions whether these types of interventions are mutually exclusive. He states, If we accept the notion that process, structure and world views are nonexclusive, dialectically related levels of analysis of interpersonal phenomena, that is not the case. In fact, each systemic change can be discussed in terms of interactional, structural and world view parameters. Even the specific sets of therapeutic interventions that clearly derive from one of the models can be analyzed from the angle of the others. . . . Even further, in many cases one given therapeutic intervention can be said to represent two intermediary models. For instance, what process-oriented (strategic) therapists define as repunctuation, world view centered (Milan) therapists will call reframing. (p. 474) As Sluzki concludes, the realization that the three intermediary models are rooted in a common paradigmatic frame expands the repertoire of conceptual and technical tools of the familj' therapists and empowers them with choices between a wide range of mutually potentiating family variables, hypotheses and interventions. Implicit in this view is the notion of "digging where the ground is soft," through 63 the assessment of family patterns, finding the marker for change, and subsequently applying the right intervention at the right time. As Sluzki asserts, none of these interventions violate the base paradigm. Thus, different change processes might be appropriate at different times in therapy and in different therapeutic contexts. It is from this perspective that the notion of the macro-sequence of interventions arises. Specifically, in reference to the systemic marital therapy used in this study, the steps are seen as circularly related, rather than rigidly sequenced. For example, in initiating therapy with couples, the therapist might begin with the restructuring intervention with some, while with others, the choice might be positive connotation of their negative interactional cycle or restraining. The choice would depend on the therapist's and the team's assessment of the couple's interactional patterns and their decision as to what approach would work best within this context. Referring to Fraser's description of Stanton's approach as following the dictum of doing what is "logical" first and then, if it doesn't work, doing something "crazy," the construction of structural as "logical" and strategic as "crazy" or "different" seems inappropriate. As Kraemer (1982), concludes, the principal task of a strategic approach is "neither to be paradoxical nor playful, but to discover and accept the family as it is, paying particular attention to the function of the symptom. To propose that it can be maintained for the time-being is simply the most effective way of communicating that attitude" (p. 355). To support this, Kraemer cites Spinoza's notion that the pain arising from the loss of any good is mitigated as soon as the man who lost it perceives that it could not have been preserved in any way. Kraemer goes on to state that By reframing a loss as inevitable and a pitiable state as "natural and necessary," Spinoza has touched upon one of the central 64 disciplines of systemic therapy, which is always to be attentive to the necessities in the system as it presents itself. This is paradoxical only because the family is expecting to be told that their problem is not necessary and can therefore be removed. The family in therapy must always be seen as a natural . . . system, however disordered or perverse. Accept anything less and the therapist will lose sight of the function of the symptom and will begin to think it is "unnatural" . . . then he or she will be paralyzed by it just as the family is (p. 355). Rather than construing that the change in approaches occurs at a point of failure, this author agrees with Papp (1983) who states, "If the symptom is being used as a secret weapon in a covert battle or has become embedded in a repetitious style of interaction, attempts to alleviate it will most likely be undermined. The therapist is then placed in a paradoxical position and the focus is on the consequences of upsetting this--of being asked to eliminate a symptom that the family has an investment in keeping but cannot acknowledge openly. In such cases an indirect or paradoxical approach that focuses on the consequences of upsetting the investment is most expedient" (pp. 29-30). The emphasis in this study is upon the family's placing the therapist in this position and also upon drawing attention to the consequences of upsetting the investment. Although moving from a structural to a strategic approach does imply some change in rules, the Sequentially Integrated Systemic approach outlined in this study suggests that departure from the initial approach makes sense if the couple continues to be stuck in negative interactional patterns, particularly if the therapy takes a U-turn and the therapist "reframes" the problem or negative cycle, thus providing a sensible rationale for such a rule change. By acknowledging that he or she has been mistaken, the therapist implicitly validates the wisdom of the couple's system and accepts the paradoxical position in which the couple has placed him or her. In this way the therapist is free to focus on the consequences to the couple 65 of eliminating a symptom that the couple has an investment in keeping but cannot openly acknowledge. Sequential Models of Integration As an alternative to integration per se, some family therapists have approached cases initially from either a structural or strategic stance and then switched mid-stream to another model. One such therapist, Stanton (1981c), proposes moving from a structural to a strategic approach and provides general rules or guidelines for such a shift. With regard to marital therapy, initially the therapist deals with the couple through a primarily structural approach: joining, accommodating, testing boundaries, restructuring, unbalancing and increasing intensity. Then, under the following conditions, the therapist switches to a mainly strategic approach when structural techniques are not succeeding: when the therapist finds the family "resisting" or no change is occurring or he or she is unsure of what is happening with the couple at the level of content or is unclear about where to go in treatment. Stanton (1981c) adds that when prior information at the outset leads the therapist to believe that the couple or family would not be responsive to structural techniques (i.e. the family may have a "dysfunctional" member and such families seem not to have been as responsive to conventional structural therapies), the therapist may choose to adopt a strategic approach. Andolfi (1980), who often works with families with severely dysfunctional members, proposes switching models during therapy. However, he begins instead with a strategic approach and works toward the structural. The implicit premise 66 for this approach seems to be that members of such families are not able to "individuate" or behave more "separately" and a paradoxical position, which is more homeostatic than that of the family, provides an opportunity for them to begin to do so. Once this is achieved, they are considered "individuated" enough to respond to structural methods. Miller (1984) presents another model for shifting interventions suggested by different approaches during the course of therapy. Her approach is reminiscent of that of McKinnon, Parry and Black (1984) with respect to matching therapy to the type of family and particular problem. However, it has been included here because it is a more dynamic and flexible approach than that of McKinnon et al., in that it allows for the mixing of models in a serial fashion, according to the perceived multiple and changing needs of different members at different times. Miller considers that a mixing of models is potentially very useful, particularly in cases where both therapeutic neutrality and non-neutrality are seen as necessary and complementary, i.e. when one member of the system is seeking therapeutic support to balance lack of support from the family or societal system. By initially using a directive, non-neutral stance in such a case, Miller notes that, . . . therapeutic non-neutrality in relation to this situation may include a temporary alliance with this family member, a more directive style of therapy and a goal-directed approach to outcome; in order to balance these aspects of non-neutrality, it is seen as critical to incorporate a more systemic, neutral stance in terms of interviewing, assessment and development of hypotheses. Without certain areas of neutrality, i.e. connecting all present members of the system as well as extended family and historical information, this case would have floundered in the absence of adequate information and perhaps failed due to an impoverishment of information to feed back into the system. (Miller, 1984, p. 61) In her work with a family with an electively mute daughter who refused to talk, Roberts (1984) also shifted models, in this instance from a Haley strategic to 67 a Milan approach. Her case example effectively demonstrates the use of two different models with one family. Roberts notes that beginning with the Haley model was effective in connecting with the "isolated, therapy-reactive family" and in instituting behavioural programs with them and speculates that beginning with the Milan model would not have kept the family in treatment. However, beginning with a more directive model and switching to Milan therapy when there was "no change" opened up new response patterns for the family and therapist and allowed "more flexibilitj' to work indirectly with the family (with "change" a more open-ended issue) as well as an emphasis on the evolutionary historical context that seemed to fit the family's multi-generational experience" (p. 52). The author concludes that defining a model to be used in treatment and then switching models offers two areas of strength that are not as clearly available when models are integrated: first, family therapy models have different strengths and limitations that may meet a family's needs better at a particular point in time and, second, the process of switching can provide a significant "news of a difference" for both the family and team. In answer to the question of when to shift, Roberts proposes broader guidelines than Stanton, who bases his decision to shift on the family's "resistance" to change. Roberts provides the following indicators: 1) when the therapist's own affective response suggests a dead end (i.e. anger, sense of futility, etc.), or 2) feedback from the family and sessions suggests that they are stuck in redundant patterns, and 3) "when the model's construction of reality seems to be limiting information that needs to be taken from the family and their social network" (p. 53). 68 Shared Systems Theory Concepts In contrast to those generally opposed to integration of family therapy models, as stated earlier, a number of therapists have proposed integrative models, most often at the level of practice. Some, Sluzki (1983) being a notable example, take the stand that structural, strategic and systemic views are subviews of the overall general system theory view on families. Implicit in Sluzki1 s approach is the idea that integration could occur through reference to shared general systems theory concepts even though the approaches differ through their respective focus on family process, structure and world views. He suggests that "models that share a systemic root are those that focus primarily on process, primarily on structure and primarily on world views" (p. 470). Process models correspond to communication-oriented therapies, structure corresponds to structural family therapies, (i.e. Minuchin), and world view corresponds to Milan therapy. Halej^'s notions span both process and structure. McKinnon (1983) notes, "although Haley's early work tended toward description of process he later emphasizes the form or structure these descriptions implied" (p. 427). As an example of this she points out that Haley asserts that double bind communication occurs in organizations in which hierarchy is incongruent, with reference to structural aspects. Sluzki (1983) goes on to summarize the models. He points out that the model emphasizing process states that symptoms, conflicts and problems are anchored in larger recursive interpersonal loops or patterns, i.e. are pieces of interactional sequences that tend to perpetuate themselves. Therapeutic interventions are focused on the recursive loops that contain the problem. Behaviour strategies for disrupting these specific patterns are symptom prescription or prescription of 69 non-symptomatic behaviour belonging to the sequence. The disruption of the necessary nature of the sequence frees the symptomatic or non-symptomatic behaviours that were captured in the cycle-as symptoms disappear, the family recovers a set of alternatives that were lost when the behaviours came to be. Notions crucial to this view are pattern, punctuation of the sequence of events and family rules. Sluzki comments that patterns may be more elusive to detect when symptoms are stable and thus other behaviours in the interactional pattern are also not fluctuating. At such times, it may be possible to infer the rules that regulate interpersonal processes to which symptoms are anchored "by means of activating another intermediary construct, the one centered on structure (p. 471). An analog}7 is drawn wherein process is to structure as verb is to noun. Therapists guided by a structural model explore and map boundaries (rules of participation) and hierarchies (rules of power). Sluzki (1983) asserts that "the modification of rules on boundaries and on the management of power and authority in a family has a profound impact on a variety of substantive transactions, including the disruption of those interactional patterns that contain and maintain symptoms" (p. 471). Even if the initial intent is to focus on interactional patterns, when this is not fruitful, it will be possible to focus on structural variables. Whereas interactional information emerges from difference, structural mapping is based on invariance. The world-views-oriented model is predicated on the notion that each of us carries a set of belief structures that organize our behaviour on the basis of assumptions, attributions, convictions, etc. and that "all communicative acts (discourse and actions alike) provide direct access to the world views of the actors, as the world view organizes the interface between the individual and his (her) environment" 70 (p. 472). Reality is anchored for the family through key words, symbols and histories that condense prescriptions of behaviours, agreements about punctuations, boundaries and interpersonal rules in general. Family members are believed to share common ideologies or ways of organizing realities. "Recursively the performance of any interactional pattern evokes the underlying world view" (p. 472). Family history and mythology can be described as an agreement about the order and meaning attributed to the events of common experiences—rendered memorable because they represent agreements about present reality, contracts about values, goals, etc. Symptomatic behaviours tend to be quickly incorporated as part of the family's reality and their activities quickly define rules about roles. Symptoms, as well as the complementary behaviours of nonsymptomatic members, serve as markers that evoke family agreements about views and values. Based on this model, the therapist's interventions will be directed toward selectively changing the organization of the specific fragments of the family's reality that provide ideological support to those symptomatic interactional patterns in order to "jolt the pattern and dislodge the symptoms" (p. 473). This is often done through positive connotation of either previously labeled negative behaviours (or behaviours that have been positively labeled in order to alter polarizations in the system and break patterns of punctuation) or proposing alternative equally plausible organizations of past or present reality that make retention of the symptomatic behaviour unnecessary. In summary, Sluzki's (1983) "inescapable conclusion" is that symptomatic behaviours can be said to be anchored in circular interactional patterns, viewed as reminders of structural traits which recursively contribute to maintain them and by their part in world views that in turn provide the ideology that supports them. 71 Process and structure are seen as a dialectic pair, whereas the construction of reality, connected with the other two refers to a different logical level. "However, level of analysis allows the description of a recursive loop that accounts for the maintenance of a symptomatic . . . behavior" (p. 474). According to Sluzki, recognizing that the three intermediary models are rooted in a common paradigmatic frame expands the repertoire of conceptual and technical tools of family therapists and empowers them with choices between a wide range of equivalent potentiating family variables, hypotheses and interventions. Although Fraser (1982) was previously opposed to integration at the practical level, in a later article, Fraser (1984) allows that further analysis may lead to at least one mode of practical integration. He argues that practical integration is possible if a model of general systems theory within the "sociocultural domain of systems" is adopted rather than an organismic level view of systems theory. In summarizing organismic level premises, Buckley (1967) suggests that 1) system structure is primary and all functions are a consequence of structure 2) overall system purpose is state maintenance and stability which is maintained by a negative feedback process called homeostasis 3) change is a nuisance and only based upon the need to interrupt one relatively stable state to move to another 4) evolution is slow and gradual and exists within a narrow range. Duncan and Fraser (1983) liken the differences in the premises of organismic and sociocultural levels to the differences between structural and strategic views, in that order. Premises of a sociocultural model suggest that 1) process is primary and structure is a description of process; 2) purpose does not exist within the system but is attributed to it by an observer; the system tends to grow on the principle of "what works"; 3) the system is open and requires variability in order 72 to maintain viability; 4) major processes of the system are continual movement to greater complexity, flexibility and differentiation; 5) transactions between members involve an exchange of information; 6) change is an ongoing process; 7) evolution is at the heart of the system; whereas many changes are gradual, major evolutionary changes often occur in rapid, discontinuous jumps. Duncan (1984) describes Buckley's (1967) system as a hierarchical classification scheme. He points out that the organismic/homeostatic view is a lower level biological view from which the construct of "function" emanates and the sociocultural view is of a higher level "process adaptive" type. This is elaborated more fully in the following section (see Duncan, 1984). Fraser (1984) points out that Buckley's analysis of levels implies that "if our domain of study is social rather than biological ss'stems, the process level assumptions should be used" (p. 47). He also points out that Maruyama (1963) and Speer (1970) both cogently argue against a view of social systems as structural^' bound to a homeostatic balance, describing deviation amplifying processes (as in the "positive loop level" described earlier) as the main factor in the evolution of social systems and that Hoffman (1971) has taken a similar position. In underlining some of the problems of using organismic level assumptions, Fraser observes that An organismically structural view applied to social systems can be seen to fall into the category of teleological explanation. In other words, a family member's action tends to be explained in terms of its function in relation to a supposed family structure, and the explanation of the family structure is in terms of its functions, (p. 50) As an example of such a teleological explanation, he cites a situation where spouses are experiencing lack of contact and are described as members of a disengaged or diffused marital dyad while the definition of a disengaged marital dyad might be in 73 terms of lack of contact between partners. This view suggests that actions and structures have an ultimate purpose in the internal balance state of the system. A process view of social systems, instead, "tends to refer to efficient causes, with little reference to inherent purpose, either in the system or in actions" (p. 50). It is based on the notion that a social sj^stem is in a constant process of flux involving elaboration and differentiation of interaction in response to internal and external variations. As an example, the process view might explain the blaming of a wife in terms of the ongoing positive feedback cycle of a spouse's angry outbursts being met by either more and more withdrawal by the husband followed by greater angry demands by the wife and more withdrawal or the reverse. Fraser notes, "the phenomenon is explained by reference to current feedback cycles with no reference to necessary past causes or future oriented purposes or functions. The cycle is an efficient explanation of itself and needs no reference to ultimate purposes such as structure maintenance" (p. 51). Thus in a process view, purpose is a construct of an observer and is not inherent in the system itself; there is no grand design of which all actions can be explained as being in the service. The author concludes that "a sociocultural process level set of assumptions appears appropriate, and preferable to lower level system assumptions, for use in guiding system based practice" (pp. 53-54). Because the system is believed to be in a constant process of change rather than state maintenance, judgements of the desirability or undesirability of interactions are made in terms of virtuous or vicious cycles as opposed to rigid, homeostatic transactions. In other words, if a wife's anger leads to her husband's withdrawal and the cycle escalates, with negative repercussions for both partners, this is seen as a vicious cycle; whereas, if the 74 anger leads to greater closeness, it would be seen as virtuous. This is in contrast to a homeostatic/organismic view which would tend to see an anger cycle between spouses as a manifestation of an undesirable, perhaps overly-enmeshed marital dyad. In a process view, a problem is seen as an escalating deviation amplification rather than a departure from change minimization. Finally, "the system is not expected to respond to dissonance or variation with an inherent reaction to reject or minimize the difference in a process resistive to change. The changing of process implies introducing dissonance or new variation to initiate a new cycle of deviation amplification, assimilation and accommodation, rather than breaking a homeostatic, relatively stable state, to create a new relatively stable state" (Fraser, 1984, p. 54). In support of his argument for practical integration, Fraser suggests that if a therapist chooses to adopt the above process level premises, he or she will likely see a system engaged in "repeated escalating solution patterns" around a variation in the system. Members chosen for therapy focus might be those that are "accessible, influential, or frustrated by the unsuccessful solution pattern" (p. 55). According to Fraser, the goal is to introduce dissonance to interrupt the vicious problem cycle and to initiate possible virtuous cycle resolution. Clients' premises, metaphors and patterns as well as those metaphors and actions of other therapy views could be adopted and used. For example, although an intervention might look like "enactment" and the therapeutic metaphor might include a description of the "function of symptoms" ("enactment" is usually associated with structural therapy and a "functional view of symptoms" is associated with an organismic level approach), the intent might be to interrupt the vicious cycle through behaviour or premise change (based on a process level premise) rather than to break homeostatic 75 structures and eliminate symptom function (based on organismic level premises). Duncan (1984) proposes a model for integrating two views of systems: one which uses the construct of "function" with one that does not. As stated earlier (Duncan & Fraser, 1983; Fraser, 1984) the difference in premises between organismic and sociocultural levels is parallel to the difference between structural and strategic approaches, respectively. Duncan (1984) traces the notion of functionality to both psychoanalytic and behavioural perspectives. As reflected in family therapy, the psychoanalytic view holds that symptoms serve a systemic function, e.g. diverting conflict, stabilizing a marriage, maintaining homeostasis, while a view similar to the behavioural holds that symptoms function to provide interpersonal gains such as for power and leverage in relationships. With regard to the notion of "function" in family systems theory, Duncan and Fraser (1983) argue that this construct emerges from an organismic/biological (or organismic/homeostatic) level view of systems rather than a process adaptive or sociocultural level view (Buckley, 1967). As Fraser (1984) notes, the organismic/homeostatic level system is lower hierarchically and is characterized by energy exchange, structure, function and morphostasis. The sociocultural level system is higher in the hierarchy and characterized by information exchange, ongoing process, fluid structure and morphogenesis. To explain more fully, Fraser (1984) draws from Buckley (1967), who argues that while the construct of function is descriptive of lower level organismic systems, it is inadequate for sociocultural systems. In an organismic/homeostatic system, the function of a given physiological structure can be determined by its future consequences for the organism because the evolution and development of the structure itself is well understood (e.g. natural selection). Also function can be ascertained because the same structure can be seen to perform the same function in other systems. However, in a 76 sociocultural system, there is no specific structure that performs a stability function within a well-defined limit that is normal for every system (Buckley, 1967). Buckley argues that there is not enough information and knowledge to determine the adequacy of a behaviour to fulfill a systemic function because of the lack of understanding of how an element and the transactions around the element have developed, i.e. there is no process like natural selection that is understood for complex social systems (Duncan, 1984, pp. 61-62). Duncan (1984) then speculates that the construct of function, which is still largely relied upon in family therapy, could be seen as a major differentiating point that bears on integration. A biological view of family systems which takes "functionality" into account is based on a contradictory premise to a sociocultural view and, therefore, could be seen as precluding integration. Duncan does not follow this dictum, however, and, instead, asserts that function can be applied in selected instances and its use can be defended, particularly if a higher level sociocultural view of systems is adopted. In this way, some form of integration may be possible. He rejects the idea that the notion of functionality of symptoms is appropriate with all clients in all situations; such a view would seem to assume the veracity of an organismic/homeostatic view of family systems and also might limit therapeutic freedom. He states, An alternative perspective, from a process/adaptive level of systems, allows for increased flexibility by enabling a therapist to utilize a lower level systemic construct ("function") if it fits the presentation of the client/system. Beginning with the premises of ongoing process, morphogenesis and nonfunctionality (characteristics associated with a higher level sociocultural view) frees the therapist to construct his/her intervention to match the clients' world view and historical presentation" (p. 62). There are two main situations when the construct of function is seen by Duncan as appropriate: 1) when it fits the scenario presented by the client and facilitates therapeutic change and 2) when such an interpretation is new and introduces useful dissonance into the family system. In the first instance, the 77 therapist is encouraged to try out tentatively particular interventions which imply functionality in order to determine the usefulness or applicability of such a construct in the given situation. An example (which is briefly summarized here) is presented of a psychotic female client, diagnosed as depressive, who described her life as one of sacrifices for her family and of abusive treatment and ridicule from her sisters. She expressed hostility toward her sisters for continually using her to "take the fall for them." As an intervention, Cindy was told that she was fulfilling a valuable function in her family; she stabilized her family with her illness and protected her sisters from conflict through her continual sacrificial behaviour of taking the fall for them. The therapist expressed his admiration for her loyalty to her sisters. An important part of such an approach is therapist flexibility: it is the therapist who must be flexible to fit the client's presentation rather than the client who must fit the therapist's orientation. In this example, the notion of function emerged from the world view of the client and an intervention was designed that matched the client's presentation of the problem, with the intention of facilitating change. Duncan points out that, although the intervention is built upon constructs involving a functional view of symptoms and an organismic/homeostatic model of systems, it could also be seen as consistent with a higher level sociocultural model of systems in that it was designed to interdict the problem maintaining vicious cycle that surrounded the symptom. In concluding, he notes, The premises of the process/adaptive system model . . . enable maximum therapeutic flexibility and are therefore probably the most useful set of constructs from which to operate. From this higher level perspective, i.e. of sociocultural systemic process rather than biological systemic process, a therapist is free to utilize any lower level construct or any therapeutic language (e.g. dynamic, behavioural, etc.) from other views that match what the client presents. (Duncan, 1984, p. 64). 78 The author then states, "In the therapeutic system, there is a reality that emerges from the transactional and historical presentation of the client. . . . one specific perspective of reality, such as the strategic, systemic and structural views that utilize the construct of function will restrict therapeutic freedom by fitting every client into that reality . . ." (p. 64). This is similar to Rice and Greenberg's (1984) approach in their emphasis on client reality and operations and events in therapy which call for an "understanding which transcends the particular orientation in which the operations were identified and studied" (p. 17) and the importance of using interventions that are best suited to particular client contexts. Finally, Duncan (1984) summarizes his paradigm for integration: "A more pragmatic position may be to . . . select the reality or set of theoretical constructs such that (it) matches the transactional and historical template of the client while basing one's overall treatment goal upon the overriding process adaptive system" (p. 64). Liddle (1984) has attempted to step back and view "integration" in a broader context. He sees it as conceptually and practically important and vitalizing. Like Sluzki (1983), he believes that integration can counter the unproductive, competitive, territorial struggles in a field. "That is, understanding the points of convergence and divergence among approaches can serve to illustrate a field's interconnectedness" (p. 66). This model of integration is presented as an alternative to dichotomization or an "either/or" thinking modality; the movement toward it challenges therapists to construct practical treatment models that are applicable across a wide range of clinical situations. Liddle makes it clear that he is not advocating "undisciplined, unsystematic eclecticism," and stresses that it makes sense to think of the treatment models as different pieces of a systemic whole but not to mold these into one grand therapy design. In practice, Liddle notes that models do 79 not have to be compatible in order to be integrated but need further work in delineating the dimensions on which they can and should be compatible so that integration can be facilitated. Liddle concludes that further rhetoric about whether or not to integrate is passe and calls for model construction or theory building. The essence of the author's position is that a structural/strategic therapy can be constructed so as to be useful at both conceptual and operational levels and that the models can be defined as having complementary theories of change. He advocates a synergistic, dialectical process as opposed to the wholesale combination of two models in back-to-back fashion. In support of this, he invokes the systems notion of nonsummativity that holds that a system is more than the sum of its parts and yields "a complexity for which the elements, considered separately, could never account" (Watzlawick, Beavin & Jackson, 1967, p. 125). Liddle (1984) suggests that complex premises and organizing principles of change supercede a preoccupation with integrative models and therapists' overfocusing on technique. In structural/strategic therapy, change occurs with support and challenge to previous realities and the accessing of new alternatives. Change is not always a straightforward, continuous process; there are patterns of behaviour that would be judged as resistances or relapses. The Ericksonian tradition in strategic therapy uses these resistant processes in the "utilization" technique-here the therapist uses whatever the client presents as a focus for therapy and accepts, reframes and accentuates a tendency toward no-change or return to a previous way of being. In structural/strategic therap}', this procedure becomes one of the overarching principles, along with the structural principle of support and challenge and allows the therapist to, at times, accept, reframe and even accentuate a tendency to "no-change" or return to a previous way of being. 80 Structural views of change have been identified as "continuous" and strategic as "discontinuous" in nature. Criticizing this "either/or" approach, Liddle states that a structural/strategic theory of change can embody both sets of premises and are held together in the therapist's imagination; neither is more correct nor primary than the other. "Each is the other's counterpart, they comprise the yin and yang of change" (p. 70). In support of this, he invokes the work of Prigogene and Stengers (1984) on how fluctuations in organisms and being "shaken-up" can give rise to growth, forms of new complexity, reorganization and a higher order (Ddssey, 1982). In clinical application, therapy could be said to proceed through the resolution of contradictions which then leads to new contradictions. This process can be thought of as "oscillation between thesis and antithesis to the synthesis transcending this dichotomy" (Watzlawick, Weakland & Fisch, 1974, p. 91). Although therapy is not seen as techniqueless, techniques are thought of as multi-directional, interactional processes. The therapeutic focus is upon change. Guidelines for model construction are observed to be minimal but Liddle (1984) cites some impressive models (Feldman, 1979; Feldman & Pinsof, 1982; Gurman, 1981a; Jacobson & Margolin, 1979; Pinsof, 1983). In concluding, Liddle (1984) finds an example of the needed direction in research paradigms in Rice and Greenberg (1984). In their analysis of client performance patterns and underlying mechanisms of change from a variety of theoretical orientations, these authors are seen to have identified classes of change phenomena or observable client/therapist interactions that seem to enhance change. Liddle states, "Although these researchers are working in the area of verifying individual change processes, their conceptualization and methodology could easily 81 serve as a model for the family therapy field. Work of this genre should orient us in the right direction~the linking of our theoretical formulations to an empirical practice base" (p. 77). Different change processes might thus be appropriate at different times in therapy and in different therapeutic contexts. Integration will then involve the applying of appropriate interventions at appropriate times to evoke appropriate change processes, provided that the interventions and processes do not violate some higher level premises or the shared systems which form the base paradigm. A Sequential Model for Integration at the Level of Practice In the present study integration is not attempted at the level of theory. Although Sluzki (1983) perceives structural, strategic and systemic views as subviews of the overall system theory view of families, he only implies that integration could occur and does not provide an integrated model, per se. An integrated model is also not provided by either Duncan (1984), who asserts that integration may be possible if a higher level, sociocultural view of systems is adopted, or Fraser (1982), who speculates that a model for integration can be found in Bateson's (1979) "alternating evolutionary ladder" across time. Presently there does not seem to exist a sufficiently evolved paradigm for a theoretical integration, nor is the development of such a paradigm within the scope of this study. For the purposes of this study, a practical treatment model is presented. Similar to Liddle's (1984), the focus is on organizing principles of change rather than on integrative theoretical models. The Sequentially Integrated Systemic treatment is an attempt to provide a practical treatment model which can be used 82 with a wide range of clients of varying backgrounds and levels of distress or conflict. The model is based upon the premises of both structural and strategic approaches. From the Structural it draws upon the principle of support and challenge which allows the therapist to at times accept, reframe and even accentuate a tendency to "no change" or return to a previous way of being. From the Strategic it draws upon the notion that the therapist uses resistant processes as a focus for therapy and reframes and accentuates a tendency toward no change or return to a previous way of being. The Sequentially Integrated Systemic approach outlined in the present study seems closest to the sequential models of integration described earlier (p. 65, this paper). While some advocates of a sequential model refer to it as "shifting models" (Stanton, 1981c) or "mixing models" (Miller, 1984), the procedure in common seems to be a technically eclectic one. It follows Rice and Greenberg's (1984) model of applying appropriate interventions at appropriate times to evoke appropriate change processes, provided that the interventions and processes do not violate higher level premises or the shared systems that form the base paradigm. Another model which forms the underpinning of the Sequentially Integrated Systemic approach is one that Papp (1983) developed within the context of the Brief Therapy Project at the Ackerman Institute. During that project, Papp explored the use of paradoxical and strategic interventions with families and couples. Her model seems to be technically eclectic as well. It incorporates something of the notion of "fit" in classifying and tailoring interventions according to the different functions of symptoms in different family situations. Noting that some symptoms are less vital to family equilibrium, she states, 83 If the symptom is primarily a response to a crisis or a transitory event it is not necessary for the therapist to become preoccupied with the consequences of change as the family will in all probability quickly absorb them. In such cases a direct approach in which the therapist merely defines the problem and advises the family what to do about it is appropriate. On the other hand, if the symptom is being used as a secret weapon in a covert battle or has become embedded in a repetitious style of interaction, attempts to alleviate it will most likely be undermined. The therapist is then placed in a paradoxical position of being asked to eliminate a symptom that the family has an investment in keeping but cannot acknowledge openly. In such cases, an indirect or paradoxical approach that focuses on the consequences of upsetting this investment is most expedient, (pp. 29-30). This notion of the family presenting a paradox to the therapist is a cornerstone of Milan therapy and forms a thesis of Selvini-Palazzoli et al.'s Paradox and Counterparadox (1978). In answering the question of how the therapist can determine what function the symptom serves, Papp (1983) concludes that it is sometimes not possible to do so before intervening, as only the feedback, i.e. family response to the intervention, provides this information. This seems to provide a rationale for a somewhat pragmatic and technically eclectic approach, one which authors such as Fraser (1982) and Rohrbaugh (1984) find haphazard and detrimental to therapeutic focus. Specifically at the level of practice, Papp appears comfortable with alternating paradoxical and direct interventions. In defence of this she asserts, "Certain families indicate from the beginning that logical interventions will prove futile; for example, families in which bizarre transactions take place, or in which a high degree of anxiety, defensiveness, denial, guilt, or anger prevents the family from 'hearing' the therapist. In order for the therapist's suggestions to make sense, the interventions must address themselves to the premises under which the family is operating. If the family is operating from some powerful hidden belief or carrying 84 out an injunction involving some tradition from the past, a common sense approach is unlikely to be effective" (p. 31). It is at such times that Papp uses paradoxical messages that address these beliefs or injunctions. She and her group alternate direct and paradoxical interventions to test the family's readiness to change. They use the paradoxical interventions to define continually the covert transactions that are hindering it. As developed for this study, the five main interventions in the Sequentially Integrated Systemic marital therapy represent an attempt to integrate structural (Minuchin) and strategic (MRI) approaches at the level of practice, in order to provide a technically eclectic therapy much the same as that of Papp (1983) and Stanton (1981c). Although the techniques of Milan therapy (Selvini-Palazzoli et al., 1978) influenced the development of Steps 4 (a) and (b) ("Positive Connotation" and "Prescribing the Symptom"), those two steps most closely parallel their counterparts in MRI. The goal for such therapy is that it be responsive to the needs of clients and/or the requirements of the therapeutic situation. Five of the seven steps which form the core of the sequentially integrated systemic treatment are presented in this chapter. In addition, each of these five steps is discussed with relation to the approach it developed from: strategic, structural, Milan. The first two steps-1) Defining the Issue Presented and 2) Identifying the Negative Interactional Cycle-are not discussed here because they are of a preliminary, problem definition nature and, so, similar to the beginning two steps in the Emotionally-focused marital therapy. As noted in Chapter III, the SIS therapy is designed to include a team of trained therapists who are positioned behind the mirror during therapy sessions, and who consult with the therapist assigned to each couple regarding the reframes and 85 prescriptions. Rules for the Choice of Techniques in SIS The SIS therapy does not include all of the techniques of the systemic therapies on which it is based. The SIS therapy is instead designed to parallel the structural/strategic marital therapy of Stanton (1981c) and the paradoxical and strategic therapy developed by Papp (1983). The authors of the SIS manual base their choice of the particular techniques that comprise the SIS therapy on their own clinical experience. Like the techniques in Papp's and Stanton's approaches, the techniques selected from Structural and Strategic and those influenced by Milan seem to fit together in a coherent manner. The techniques selected also seem to be the most effective in creating change in couples. The last five steps are: 3) Restructuring The purpose of the restructuring step is to try to intervene directly in order to encourage the couple to change their interactional cycle before trying strategic or indirect interventions. It is based on the structural therapy of Minuchin (1974) and parallels Stanton's (1981c) approach. The interventions in this step, which are drawn from the structural approach of both of these therapists, include joining, testing boundaries, enactment, unbalancing and increasing intensity. The interventions rest on a view of change as a straightforward, continuous process. Treatment is aimed at restructuring the system, i.e. establishing or loosening boundaries, as appropriate for a particular couple. If a couple is "enmeshed" 86 (overinvolved with each other), for example, then the therapist will suggest ways of communicating and interacting that will lead to the establishment of "boundaries" between them. Problems are believed to result from a rigid dysfunctional structure; their usefulness will disappear with system transformation (Stanton, 1981c). Minuchin (1974) defines the goal of structural therapy as inducing a more adequate family organization which will maximize growth potential in each of its members. He notes that people will change when alternative possibilities are presented which make sense to them and when alternative transactional patterns have been tried out. At that time, new relationships appear which themselves become self-reinforcing. The whole idea behind restructuring is to shift supports around, recognizing that people will not move to the unknown in a situation of danger. The healing potential of the relationship is assumed and supports are provided to facilitate movement. If restructuring is successful, therapy continues along these lines. The main focus of the Restructuring step in the SIS couples' therapy is on changing the partners' interactions by getting them to talk and behave differently to each other in and between sessions. As a result of clinical observations, it is concluded that such a direct technique is sometimes successful in enabling partners to change the way that they communicate with each other. If partners are able to change through these direct interventions, this likely demonstrates that their difficulties are primarily a response to a crisis or a transitory event rather than ingrained, or part of a hidden belief, premise or covert transaction and, therefore, amenable to such direct interventions (Papp, 1983). If Restructuring is successful, then therapy continues along this line. 87 4) Reframing the Interaction. If, as Stanton (1981c) suggests, prior information leads the therapist to believe that the couple would not be responsive to direct interventions and/or the therapist and team find that the family demonstrates resistance (more specifically, as Papp (1983) has suggested, the symptom or cycle seems to be used in a covert battle or seems to have become embedded in a repetitious style of interaction), resistance will likely be manifested through an adherence to old patterns and an absence of change. At such times it seems that the therapist is placed in the paradoxical position of being asked to eliminate a symptom that the couple has an investment in keeping but cannot acknowledge openly. To deal with this situation it is necessary, Papp states, to adopt "an indirect or paradoxical approach that focuses on the consequences of upsetting this investment" (Papp, 1983, p. 30). It is for this reason that, when the therapist and team perceive that the couple continue to be stuck in negative interactional patterns, the therapy takes a U-turn and continues along more strategic lines, in which indirect attempts to change are applied. The U-turn is a reversal of therapeutic position and allows the therapist and team to note that they must have been mistaken in their perception of the problem because obviously the "fight" is serving an important function and it would be premature to change this too quickly. This is based on the systemic assumption of homeostasis that pathological systems adhere to solutions which maintain the system as well as the notion that symptoms are highly adaptive for the couple in helping them to maintain a steady state in their system. Here, as in the following steps, the perception of change is of a less direct process (corresponds to the notion of a positive feedback loop) which takes into account or 88 utilizes client patterns of behaviour which could be judged as resistances. By taking a U-turn, the therapist and team align with the couple and position themselves on the side of "no change." a. Positive Connotation: After conferring with the team, the therapist begins the reframing step by positively connoting the couple's negative cycle. This is an important technique based on Milan therapy, corresponding to the reframing process in other systemic therapies, notably MRI Strategic therapy. It is derived from the belief that when the (symptomatic) behaviours of family members are connected in a circular fashion and connoted positively as beneficial for the family, they are more easily accepted. An important goal of positive connotation is to allow new information or alternative world views to gain entry into the family system and to address in a respectful way the fears the couple has about change as well as recognizing and acknowledging the functional aspects of the symptomatic cycle. This leads to the defining of the relationship between partners in a positive way. Positive connotation is an important step in legitimizing the prescription of "no change" which is to follow. This is paralleled by the sequence of reframing negative interactional cycles and then prescribing the symptom (no change) which is followed by the MRI group. Using positive labels or relabeling of the couple's negative cycle is in keeping with premises of paradoxical psychotherapy espoused by Weeks and L'Abate (1982), Haley (1980) and the MRI group (Watzlawick, 1976; Watzlawick et al., 1974). In order to focus on the positive aspects of symptoms, the therapist must understand how they have been adaptive for clients. Once the positive function of the symptom has been identified, it can be used as a vehicle of change. This intervention is seen, then, as the first step in facilitating positive therapeutic 89 outcome. b. Prescribing the Symptom: As noted above, Positive Connotation is an important step in validating the partners and their relationship and acknowledging the positive function of the fight. It also leads the way for symptom prescription. This can be traced to the Milan notion that if a particular pattern of behaviour is construed as being a good thing then it follows that it should continue. After positively connoting the negative cycle the therapist prescribes or tells the couple to continue to fight. Here, the therapist reiterates the important function of the fight and tells partners to continue it at home. This is an example of the therapeutic double-bind (the mirror image of a pathogenic bind) which was developed by Watzlawick, Beavin and Jackson (1967), forerunners of the MRI school, where an injunction is given that creates "paradox" because the clients are told to change by remaining unchanged, putting them into an untenable situation. If they comply, they no longer "can't help it" and the symptom is under their control (the purpose of therapy); if they resist, they can only do so by not behaving symptomatically, which is also the purpose of therapy. Thus, the implicit message of the reframe or positive connotation is made explicit in the paradoxical directive to continue fighting. This intervention is consistent with Haley's (1976) notion that the goal of therapy is to encourage people to behave differently and thus have different subjective experiences. Awareness, while not necessarily irrelevant to the therapeutic process, would be more likely to follow rather than precede the behaviour change. Sluzki (1978) observes that in prescribing a symptom to a symptomatic member in the presence of the mate, the therapist tries to shatter the pattern that perpetuates it. This occurs in two ways: 90 i) When the patient is told "fake the symptom, and fake it well," the other member of the dyad is implicitly being told that the symptomatic behaviour to which he or she may be exposed may be false, therefore inhibiting "spontaneous" responses that in turn may reinforce and perpetuate the symptom; and ii) it subtly increases the consensus about the patient's control over the symptom, and decreases the chances of his or her claiming spontaneity. It induces the notion that if a person can produce a symptom, he or she may also be able to reduce it. As in Milan therapy, the additional phrase, "for the time being" ("for now") is added to the "no change" prescription. Thus, although the couple's homeostatic tendencies are being positively connoted, the phrase "for now" suggests alternative solutions and possibilities for change; the therapeutic paradox, rather than being binding, is implicitly open. The team supports the therapist in positively connoting and prescribing the symptom. There are times when some members of the team support the reframe and symptom prescription while others express reservations about it, in order to identify with client reluctance or resistance. 91 5) Restraining By discouraging change or even denying that it is possible, the therapist continues to align himself or herself with a position of "no change." The restraining step is based on a belief which is common to Strategic (MRI and Haley) that it is risky for the therapist to be explicitly aligned with system tendencies toward change and instead needs to be seen by the client as accepting the status quo. In this way it could also be seen as consistent with the Systemic (Milan) belief in therapeutic neutrality. Partners in a system habitually tend toward stabilizing the system, by following patterns already familiar to them, and thus strengthening the status quo and sj'mptoms. The strategy of restraint portrays the therapist as uncommitted to changing the clients, particularly to having them change quickly. The Go Slow step is applicable when there is a slight shift toward improvement or clients show signs of making some moves toward resolution. Clients are instructed not to do anything further-particularly nothing specific. Instructions are general and vague and most of this intervention is comprised of offering believable rationales for "going slow." The Dangers of Improvement step can be seen as part of the restraining step and is an extension of "go slow" but also carries the message "you probably shouldn't change." It is useful when clients, perhaps even though showing some tendency toward change, are not able to get out of their vicious cycles or can be used when change is not occurring. The therapist mobilizes change by benevolently suggesting that change is something to be feared. Eliciting disadvantages may help partners to see alternatives or that improvement is not "ideal," thereby feeling less 92 compelled to put pressure on themselves to change and this may help them relax. Partners would then have altered their attempted solution of trying too hard which Strategic therapists see as central to the problem and there would likely be a lessening of the problem as a result. 6) Consolidating the Frame While introducing a reframe or presenting it on one occasion sets the process of change in motion, it does not seem sufficient to create lasting change, particularly with more distressed couples, where the negative cycle has been highly adaptive over time and become a pervasive pattern. In order to make a more lasting impact, the frame needs to be consolidated. In consolidating, the therapist in essence reinforces and generalizes the reframe. Consolidating the frame seems to be a crucial step in the process of SIS therapy; it is described in further detail in the SIS manual (Appendix A). Although therapists of a systemic orientation may adhere to such an approach, the notion of consolidation as presented here has not been reported in the literature on systems theory or therapy. However, "Consolidating Change" as a stage of therapy has been identified by Friesen (1985) and Erickson (in Omer, 1982). 7) Prescribing a Relapse In the sense that the therapist aligns with "no change," this step can be seen as consistent with MRI and Haley Strategic therapies. In portraying the therapist as uncommitted to change, this step can be seen as consistent with the Milan concept of therapist neutrality. 8) Positioning Positioning is a concept that is linked to the MRI Strategic school and Haley. In positioning, the therapist attempts to shift a problematic "position"-usually an assertion that one or the other partner is making about himself or herself and his or her problem~by accepting and exaggerating that position. Positioning is important in working with reactive or resistant clients. A therapist can discourage resistance by responding to clients in ways that are not opposition to their position. If, for example, one or both partners express pessimism about the treatment, the therapist must position himself or herself in such a way as to accept and exaggerate this. (For further discussion on Positioning, or any one of these steps, please see The Sequentially Integrated Systemic Manual, Appendix A.) Outcome Research in Strategic Family Therapy Although there has been a significant increase in research on outcomes of family and marital therapies in general in the last ten years, outcome studies specifically comparing marital therapy across models is sparse. There is also evidence of a paucity of outcome research relevant to marriage using systemic or strategic therapies (Gurman & Kniskern, 1981; Stanton, 1981a). However, seven 94 research studies that investigate treatment outcome using strategic-oriented family therapy have been cited by Stanton (1981a) who has suggested that, based on Gurman and Kniskern's (1978) scale of family research design quality, these studies utilized superior research designs in comparison with the average for studies of other family therapy approaches. On the basis of his review, Stanton (1981a) has concluded that, depending on the kind of patient population, strategic orientation to family therapy either shows substantially better results or considerable promise when compared to standard forms of treatment. These studies as well as one relevant study cited by Weeks and L'Abate (1982) are briefly reviewed here. Langsley, Fairburn and De Young (1968) investigated family crisis therapy using techniques similar to MRI Brief Therapy Center and Haley (i.e. brief problem-focused therapy, emphasis on the present rather than the past, small changes as the goal, attention to family hierarchy, reframing, positive interpretation, giving firm directives and concrete tasks for homework). Half were randomly assigned to family crisis therapy without hospitalization and half to standard treatment. Results from an 18 month follow-up showed that family crisis therapy cut in half the number of days patients subsequently spent in the hospital. However, methodological problems somewhat mitigated the impressiveness of the results (Stanton, 1981a). Alexander and Parsons (1973) compared a behaviourally oriented, crisis centered family therapy based on strategic techniques and systems theory derived from Haley (1963) and Watzlawick et al. (1967) with client-centered, eclectic dynamic approaches and a no treatment control group in treating delinquency. Recidivism was cut in half in the systems treatment group, suggesting the superiority of this treatment. A three year follow-up by Klein, Alexander and 95 Parsons (1977) showed the incidence of problems in siblings to be significantly lower for the family system treatment. Olson et al. (1980) states that Alexander and Parson's approach is usually characterized as "behavioural" but has elements of strategic therapy and uses a form of reframing referred to as relabelling. Gurman and Kniskern (1981) note that Alexander and Parsons shifted from a strictly behavioural approach to a more systemic one when the behavioural was unsuccessful. As with other similar studies (Garrigan & Bambrick, 1977) the authors do not specify the set of strategies used. Weakland, Fisch, Watzlawick and Bodin (1974) did short-term follow-ups on 19 per cent of their cases. The problems represented a broad range of individual, work and family problems. Clients were not seen for more than ten sessions and an average number of sessions was seven. Seventy-two per cent of the clients were either successful (40%) or significantly improved (32%) as measured by the Goal Attainment Scale. The lack of control or comparison groups limits conclusions that can be drawn (Stanton, 1981a). A structural/strategic approach to family therapy with drug addicts was investigated by Stanton and Todd (1979). Follow-up results after six months suggested that family therapy treatment conditions were superior to other conditions. Clinical findings of relevance here indicate that with married addicts, the relationship with family of origin had to be dealt with first before focusing on the marriage, if treatment was to succeed. With regard to pre-post treatment changes in family interaction patterns, Stanton noted that this is the first study to provide experimental support of the theory that system change is necessary for symptom change. 96 Garrigan and Bambrick (1975, 1977, 1979) conducted a six year research project investigating outcomes of Zuk's (1966) "go-between" therapy for families with disturbed children. Follow-up of the first study suggested the treatment group showed more improvement in the identified patient's perception of family adjustment, despite teacher judgement of little difference in classroom behaviour. The second study suggested the treatment group showed significantly more improvement in the identified patient's behaviour in class and at home. This study was notable in that it used one of the few measures of marital dyadic function available in the literature. Results suggest that family therapy enabled these couples to reestablish communication within the marriage on more meaningful and facilitative levels of relating. Treatment was not as effective in the third study. Gurman and Kniskern (1978) rate the design of the first two studies as good and very good, respectively. Wagner, Weeks and L'Abate (1980) conducted an experiment with 56 married couples, all non-clinical volunteers, to study the effectiveness of direct, straightforward and paradoxical letters. Results suggest that all three treatment groups made significant improvement in marital functioning as compared with the control group, and the paradoxical group did not differ significantly from the other two experimental groups. Because of certain design flaws and because the couples did not initially present with specific problems or show resistance, the authors concluded that this may not have been a good test and that paradoxical techniques with such couples may be unnecessary and/or inappropriate. The authors assert that this and Weakland et al.'s (1974) study are the only empirical investigations of paradoxical psychotherapy other than several studies by Frankl. This study seems to be the only such study that focuses solely on the marital dyad. 97 It is difficult to draw overall conclusions from the above outcome research that are relevant to marital therapy. Several studies did show that family therapy embodying characteristics of strategic therapy was effective in helping clients achieve goals, improve communication, and reduce incidences of drug addiction, hospitalization and delinquency. However, these studies were often hampered by methodological problems such as lack of control or comparison groups or failure to delineate treatment approaches carefully. Two of these studies which are claimed to be the only empirical investigations to focus on paradoxical psychotherapy other than Frankl's, suffer from design flaws. The results of Wagner et al.'s (1980) study cannot be generalized to a more distressed population because the couples participating in the study were not distressed at the outset. In the present study, important design and methodological issues are addressed in an attempt to prevent such problems and produce a study that is empirically sound and generalizable. For this reason, two treatment groups and a control group are used and two marital therapy approaches carefully delineated and implemented according to treatment manuals. A complete description of the design and methodology of this studj' can be found in Chapter III. 98 CHAPTER III: The Methodology of the Study This chapter includes information about the methodology and data analysis used in this study. There is a brief discussion of methodological issues in the evaluation of marital therapy, followed by an outline of the design of the study, including the client population and sampling, therapist selection and training, discussion of instrumentation and methods of data analysis. The following criteria for evaluating the adequacy of outcome studies in marital amd family therapy which have been formulated by Gurman and Kniskern (1978) are considered applicable to this study: 1. Controlled assignment to treatment conditions: random assignment, matching of total groups or matching in pairs. 2. Pre-post measurement of change. 3. No contamination of major independent variables, including therapist experience levels, number of therapists per treatment condition and relevant therapeutic competence (e.g. a behavioural therapist using emotionally focused therapy for the first time provides a poor test of the power of an emotionally focused method). 4. Appropriate statistical analysis. 5. Follow-up: three months or more would be most desirable. 6. Treatment equally valued, particularly by therapists participating in the study-tremendous biases are often introduced when this criterion is not met. 7. Treatment carried out as described or expected. 8. Multiple change indices and multiple vantage points used in assessing outcomes. 99 9. Outcome not limited to change in the "identified patient," i.e. either one or the other partner. 10. Data on other concurrent treatment and its equivalence across groups. 11. Equal treatment length in comparative studies. 12. Outcome assessment allowing for both positive and negative change. The above issues were addressed in conceptualizing this study and provision was made for ensuring their implementation. Gelder (1978) and others emphasize that treatment operations must be specified and implementation checked and that there is no point in studying a treatment which cannot be reproduced. Luborsky, Woody, McLellan, O'Brien and Rosenwerg (1982) support Gurman and Kniskern's (1981) emphasis on the importance of explicating the independent variable with their design of a treatment manual for therapists which specifies the nature and sequence of therapist interventions and allows a check on whether the therapist adheres to the treatment model. O'Leary and Turkewitz (1978) note that treatment manuals should be made available and that without such manuals, it is almost impossible to draw conclusions from treatment studies. Core issues surrounding subject and therapist selection and assignment, the need for multidimensional measures taken from different perspectives and appropriate follow-up are similar to those found in the psychotherapy literature, in general. O'Leary and Turkewitz (1978) discuss the need for a clear description of the sample population in detail, to make subject exclusion criteria clear as well as the problems of "therapist effects." They note the need for a minimum of 3-4 therapists per experimental condition in order to allow for appropriate generalizations regarding treatment and the study of whether a treatment program can be 100 successfully implemented by therapists of varying styles. The authors also state that it is important to note that crossing therapists with treatments has clear advantages and disadvantages. One advantage noted is that differential treatment outcomes can be more readily attributed to the treatments rather than the skill of the therapists. A serious disadvantage, however, is the possibility of systematic bias if all or most of the therapists involved have a theoretical orientation that favours one of the treatments. They also stress the need for therapy implementation checks. There is general consensus that control groups are necessary. While the logical power of controlled treatment studies is generally acknowledged, Gurman and Kniskern (1978) point out that there is no such thing as a true control group; supposedly "untreated" patients very often get themselves treated (Bergin and Lambert, 1971). They also note that it seems almost impossible to have ultimate control over the matching of treatment and control groups with regard to the severity of the presenting problem. Finally, the ethics of either a placebo-control or wait-list control group are problematic. One procedure, known as "treatment on demand" (Gurman & Kniskern, 1981) presents a viable alternative. In the present study, where the researcher is concerned with linking systemic and emotionally focused interventions to specific outcomes, it is considered necessary to have either a placebo control group or a wait-list control group. Other measurement issues which have been considered are the difficulties generated by the demand characteristics of marital interaction observations and self-report indices (Johnson & Bolstadt, 1973; Mitchell, 1973). This study is designed to allow for replicability through treatment manuals, with Subjects and therapists assigned according to the conditions specified above and 101 multidimensional measures, a control group and implementation checks used. Design of the Study The present study is a comparative outcome study to investigate the differential effects of the two treatment levels of the independent variable, specifically, sequentially integrated systemic marital therapy and emotionally-focused marital therapy, and a control condition on the dependent measures of marital adjustment, conflict resolution, target complaint reduction and specified relationship goals. The design takes the following form: R o. Ti o2 o3 R o, T2 02 o3 R o, T3 o2 Treatment 1 is the systemic treatment, Treatment 2 is the emotionally-focused treatment and the Control is a wait-list control. There are 14 couples in each treatment as well as in the control condition. Summaries of the two treatments and the essential characteristics which differentiate them follow. (Therapists are nested under treatment and treated as a fixed factor in analysis). Manuals for the sequentially integrated systemic (systemic) treatment and the emotionally-focused (experiential) treatment can be found in Appendix A and Appendix B. 102 The Treatments  Emotionally-focused Marital Therapy The emotionally-focused marital therapy used in this study is derived largely from experiential therapies and somewhat from systemic therapies, and is directed toward present affective experience in an interactional context. Change is seen to occur in each partner as well as in the marital interaction. Because of the assumption that change occurs within the individual as well as within the context of the relationship, the therapy is directed toward interpersonal relationship changes as well as change within each partner. Partners are regarded as active perceivers who construct meanings and organize perceptions and responses on the basis of current emotional states, disowning aspects of present and past experience because of blocks to awareness or because of perceived requirements of the present relationship. The therapist attempts to induce changes within each partner and link these changes to the relationship. Initial interventions (usually within the first two to three sessions) include delineating conflict issues and attempted solutions, helping each partner to establish his or her position, and identifying and exploring negative interactional patterns. Partners are then encouraged to develop awarenesses of unmet needs for closeness and intimacy and legitimize feelings of vulnerability or deprivation. The mechanism for such change is emotional experiencing, wherein underlying feelings or newly formed emotions are brought into awareness and expressed. (It is difficult to say specifically when each intervention will occur because the steps of the treatment are cyclical and tend to occur in an iterative fashion, however the above-mentioned 103 processes will generally take place during the middle phase of therapy.) The therapist helps partners to recreate situations and relive emotions in order to make the cognitions governing behaviours available for clarification and modification. Interventions may also be directed toward helping partners to experience and integrate dissociated parts, i.e. when partners see spouses in terms of past experiences or relationships instead of the current context. The therapist then helps partners to incorporate such changes into the relationship. The above processes lead to both partners redefining the self in the relationship. Different aspects of the self are accepted and expression of these leads to changes . in the interactional sequences. Partners are able to appreciate the other's position and accept previously unacceptable behaviours in their spouses. There is a high demand for disclosure, conducive to the building of intimacy and emotional bonds which are critical to marital satisfaction. (This would often occur during the last few sessions.) Therapeutic techniques include methods of Gestalt therapy (Perls et al., 1951), empathic reflection, evocative responding (Rice, 1974) and reframing the problem and responses in terms of underlying feelings. The treatment manual for training includes the following steps: 1. Define issue as presented. 2. Identifj' negative interactional cycle. 3. Facilitate clients in accessing and accepting previously unacknowledged emotions underlying the cycle. 4. Redefine the problem cycle in terms of these new emotions and the clients' interacting sensitivities. 5. Encourage identification with previously unacknowledged aspects of experience by enactment of redefined cycle. 6. Facilitate acceptance of partners' positions. 104 7. Encourage clients to state needs and wants arising from their new emotional synthesis. 8. Facilitate new solutions. 9. Help clients to integrate new perspectives of the self and the other, solidify new relationship positions and ways of achieving intimacy. Establishment of a therapeutic alliance is a priority as much of the treatment implementation and effectiveness is contingent upon it. A therapeutic alliance is defined here as developing an emotional bond between counsellor and client. It is believed that such an alliance is developed through empathy, and the sharing of emotional experiences as well as sharing of common goals. Validation-support and legitimization of partners and their concerns and positions-is part of developing an alliance. The success of the therapy is considered contingent on a balance in the validation of both partners and the development of a therapeutic alliance and shared perspective. In summary, then, the essential characteristics which differentiate the emotionally-focused marital treatment suggested here from other treatments is the focus on the present affective experience in an interactional context. It is considered crucial for partners to access underlying feelings and incorporate resulting changes in their experiencing into the relationship. The assumption is that when different aspects of the self are accepted by both partners, this will lead to deeper level relationship changes, underlying levels of intimacy, emotional closeness and in changes in conflict cycles. The success of the therapy is considered contingent on the validation of both partners and the development of a therapeutic alliance. 105 Sequentially Integrated Systemic Marital Therapy The sequentially integrated systemic marital therapy used in this study is derived from family and marital systems therapy and is based on an integration of process, structure and world views-orientations. Therapy is directed primarily toward change at the interactional level. Changing repetitive, self-perpetuating negative interactional cycles is believed to lead to second order change in the system: change not only in behaviours but in rules governing interactions, in process or relationships. The basic assumption is that when change is achieved in the interactional domain of experience it will spread to other domains-internal and experiential. Because marital conflict is not seen to follow from previous psychopathology in partners and because awareness of internal processes is seen as irrelevant to the therapeutic process, therapy is directed toward present interactions and changing present symptomatic behaviours which are believed to be anchored in circular interactional patterns, which are recursively maintained in structural traits. The individual and shared world views of the partners provide the ideology that supports such structures and interactional patterns. Problem behaviours and negative interactional cycles are seen as serving a purpose or function in the marital system and change needs to occur around points at which the system seems to be stuck. Problem formation and maintenance are a vicious circle process in which attempted solutions to problems are inadvertently created and maintained. The therapist's primary aim is not to attempt to resolve all difficulties but instead to initiate a reversal in the repetitive negative communicational or interactional cycles and to change the frame of reference or meaning attributed to the situation. The mechanisms for change in sequentially integrated systemic 106 marital therapy stem from structural as well as paradox paradigms. The attempt is first to establish each partner's position and to get a clear delineation of the problem behaviours and behaviour that serves to maintain the symptom or problem. This is followed by attempts to restructure the couple's interaction. If this is not successful, interventions include reframing or positively connoting the symptom (negative interactional cycle) by changing its conceptual or emotional meaning and focusing on its positive rather than negative functions in the marital system. This is followed by prescribing the symptom and restraining, where the therapist questions possibilities for change and cautions partners to "go slow." The treatment manual for training includes the following steps: 1. Defining the issue presented. 2. Identifying the negative interactional cycle. 3. Restructuring. 4. Reframing the problem a. Positive connotation b. Prescribing the symptom 5. Restraining a. Go slow b. Dangers of improvement 6. Consolidating the frame. 7. Prescribing a relapse. The therapist's "positioning" in relation to partners is considered critical in creating a context for change. In positioning, the therapist attempts to shift a problematic position of partners-usually an assertion that one or the other spouse is making about self or problem~by accepting or exaggerating that position. By not pushing for change and responding to clients in ways that do not oppose their positions, the therapist is helped to discourage reactivity or resistance and thus to 107 create a context for change. In summary, then, the essential characteristics which differentiate the sequentially integrated systemic marital treatment suggested here from other treatments is its exclusive focus on the current interactions between partners, along with its reframing of negative interactional patterns, prescribing of the symptom and use of restraining tactics. The therapist responds to clients in such a way as not to oppose their positions or perceptions of reality. This kind of response is believed to be a critical variable in the change process. Again it is difficult to say specifically when each intervention occurs because of the cyclical nature of the steps; however, defining issues, identifying negative cycles and restructuring tend to occur in the first third of therapy, followed by reframing and symptom prescription during the middle third. Restraining tactics and consolidation of frames occur repetitively and prescribing a relapse occurs after a behavioural change or shift. Research Procedures In order to qualify for the study one or both partners were required to score within the predetermined limits on a measure of marital satisfaction. In addition, it was necessary for couples to meet the criteria listed below under "Client Population and Sampling." Pre-measures were then taken (Target Complaints, Goal Attainment Scaling and Conflict Resolution Scale). During the induction session and in the course of completing measures, specific conflict cycles were discussed and delineated as the focus of therapy. Subjects were then randomly assigned to treatment and therapist. The couples assigned to the wait-list group had access to 108 immediate referral to outside therapy if deemed necessary by them, as well as a course of therapy following administration of the two treatment groups. Couples in the treatment groups received ten one-hour weekly therapy sessions in the Education Clinic at U.B.C. Sessions were audio- and video-taped. Client Population and Sampling The subjects for this study were recruited from newspaper, radio and T.V. advertisements. Respondents who met the criteria below were offered ten free sessions designed to help them break recurrent conflict cycles. Before counselling, clients had an orientation interview and were screened according to the following criteria: 1. Clients must have lived together for a minimum of 18 consecutive months. 2. Clients must not have immediate plans for divorce or separation. 3. Clients must not have received psychiatric treatment/hospitalization within the two year period previous to treatment. 4. Clients must not have characterological problems or be suffering from clinical depression or any psychopathology such as schizophrenia. 5. Clients must not be addicted to alcohol or drugs, according to self-report. 6. Clients must not have engaged in incidences of physical abuse with each other, according to self-report. 7. One client in each dyad must fall into the "distressed" range on the Dyadic Adjustment Scale (Spanier, 1976), that is, one must score below 95 but not below 60 and the other between 60 and 120. 8. Clients must consent to research procedures, i.e. completing test forms, 109 video-taping and audio-taping. Demographic data on clients was collected and includes age, length of time together, number of children, education level of partners, occupation, previous therapy experience, previous marriage and income level in order that this population can be described in detail. After screening, appropriate clients attended the Education Clinic at the University of British Columbia for counselling. Therapist Selection and Training There were 14 therapists in this study, seven conducting each treatment. All of the therapists had at least a Master's degree in counselling psychology, clinical psychology or social work and some were doctoral candidates in the counselling psychology program. Each therapist had a minimum of three years experience in general counselling in addition to at least two years' experience in couples' counselling. None of the therapists in either group had more than four years' experience in counselling couples. In addition, all of the therapists in the EF group were committed to that orientation, having had at least 24 hours of supervised training in experiential therapy beyond the minimal education and training requirements cited above and all of the therapists in the SIS group were committed to that orientation, having had no less than 30 hours of training in structural/strategic therapy, beyond the minimal education and training requirements cited above. If they met the minimal criteria, therapists were matched between the two groups with regard to number of years of general counselling, number of years of couples' counselling and number of hours of specialized training in either the experiential or the structural/strategic orientation, respectively. 110 Therapists were trained to implement the therapy manuals for this study by Dr. Leslie Greenberg in a group setting (twelve hours of training for each group). Group supervision of therapists by Dr. Greenberg and the researcher took place once weekly. There was a team of three similarly trained therapists behind the mirror during the SIS therapy sessions, who viewed videotapes of the sessions and consulted with the couple's primary therapist regarding reframes and prescriptions. Checking of Clinical Procedures: Implementation Check In order for external observers to provide a rigorous check of clinical procedures, implementation checks were made following the therapy sessions. These implementation checks were made by trained independent raters (three hours of training) who were blind to the treatment conditions they were observing, and who rated segments of video and audio tapes. The raters had a minimum of two years experience in counselling. An Implementation Checklist (Appendix C) enabled the raters to determine if the interventions stipulated in the treatment manuals had, in fact, occurred. A total of 93 out of the 280 sessions or approximately three of each couple's series of ten sessions were chosen at random and observed by the raters. Two segments of ten minutes each were taken from the middle and final third of these 93 sessions. In this way each couple was observed for a total of 60 minutes of their therapy. (Refer to p. 113 in the next section, entitled Instruments Used, for a description of the Implementation Checklist.) During the weekly supervision meetings of the Emotionally Focused therapists, tapes were reviewed in order to ascertain that the treatment was being implemented according to the EF manual. The presence of the team of trained Ill therapists behind the mirror during the Sequentially Integrated Systemic therapy served a parallel function by ascertaining that the treatment was being implemented according to the SIS manual. Timing of Interventions The steps of both treatments in this study, as outlined in the manuals, tend to be iterative in nature. There was, therefore, no strict time sequence of interventions corresponding perfectly to sessions, and, if necessary, earlier steps were repeated in later sessions. In general, however, in both groups, the first three treatment steps were accomplished during the first three sessions of therapy. There was some correspondence between Steps 4, 5 and 6 in the Emotionally Focused therap}', with the next four sessions of therapy and Steps 4 and 5 in the Sequentially Integrated Systemic therapy corresponding somewhat to Sessions 4, 5, 6 and 7 of that therapy. Steps 7, 8 and 9 in the Emotionalty Focused therapy, like Steps 6 and 7 in the Sequentially Integrated Systemic therapy, were seen as resolution steps and generally occurred during the last three sessions of either therapy, respectively. Instruments Used Outcome Instruments 1. The Dyadic Adjustment Scale (DAS) (Spanier, 1976). This scale was used for screening and as a general measure of outcome at post-test and at follow-up. The 112 DAS is comprised of 32 items arranged into four subscales measuring dyadic satisfaction (10 items), consensus (13 items), cohesion (5 items) and affectional expression (4 items), and is at present the preferred instrument for the assessment of marital adjustment in relation to reliability (.96 Cronbach's Coefficient Alpha) and validity. Validity data gave a correlation between the DAS and the Locke Wallace Marital Adjustment Scale (1959) of .86. Distress level as measured on this instrument correlated with satisfaction concerning conflict outcomes and objective conflict resolution in Koren et al. (1980). Spanier points out that the scale can be considered to be a measure of the adjustment of the dyad as a functioning group rather than a measure of individual adjustment to the relationship. The scale has a theoretical range of 0 to 151. The mean total score in the norming sample for married and divorced couples was 114.8 (S.D. 17.8) and 70.7 respectively. The reliability of the subscales is Consensus .90, Satisfaction .94, Cohesion .86 and Affectional Expression .73. The majority of items involve a five or six-point Likert-type scale defining the amount of agreement of the frequency of an event. A rating measure for global happiness and for commitment is included in the Satisfaction subscale. 2. Target Complaints (TC) (Battle, Imber, Hoehn-Saric, Stone, Nash & Frank, 1966). This measure was filled out by each partner in conjunction with their therapist during the initial interview given at post-test and at four month follow-up. This measure is recommended in Waskow and Parloff (1975) as a core battery instrument for use in psychotherapy outcome research. It is comprised of 3 five-point scales on which the client is asked to rate the amount of change on three different complaints. (In this study clients will be asked to rate the amount 113 of change on three complaints related to the main conflict in the relationship.) Numerical values can be assigned to each rating point. The client's score on the instrument then becomes a mean value consisting of the sum of the ratings for all target complaints divided by the number of complaints rated. Battle et al. (1966) provide evidence as to the validity of this measure; it shows significant correlations with four other outcome measures. In particular, the main complaints derived from a target complaint interview are congruent with complaints obtained in intensive psychiatric interviews. The authors state that, as an outcome measure, Target Complaints is informative, makes good clinical sense and responds differentially to experimental manipulation. With regard to reliability or consistency of clients' initial definitions of problems, clients' rankings of problems between pre-post psychiatric interviews shows a correlation of .68. Jacobson, Folette and Elwood (1984) suggest that measures which tap couples' presenting problems most directly-as Target Complaints and Goal Attainment Scaling—are preferred instruments in assessing marital therapy. 3. Goal Attainment Scaling (GAS) (Kiresuk & Sherman, 1968). This measure was filled out by each client in the assessment interview as well as at post-test and at four month follow-up. The clients were asked to set specific behavioural goals in relation to their main presenting concern and the attainment of these goals was evaluated using this procedure. In this study, clients were asked to focus upon the main goal in relation to their marital issues and to define five levels of attainment of that goal: "expected or most likely results," "somewhat better than expected results," "much better than expected results," "somewhat less than expected results," and "much worse than 114 expected results." For each level the client was asked to list three observable and quantifiable behaviours (with emphasis on specifying the frequency, i.e. "I would like my husband to listen to me and give me feedback at least three times a day") as well as an affective indicator, i.e. how would they feel or want to feel if this happened? Clients were then asked to indicate their level of attainment at pre-treatment and at post-treatment and at follow-up were again asked to indicate their level of attainment. Although it is possible that the level of attainment could be lower than the initial level before treatment, the treatment goal is for the level of attainment to improve. A spouse might describe the "somewhat better than expected" level of goal attainment as 1) being able to reach consensus on decisions 75 per cent of the time, 2) being able to discuss openly issues when consensus was not reached and as a result, 3) only having one escalating conflict cycle a month. This could lead to the spouse feeling more content and secure in the relationship. The GAS scores are based on a standard score system (T scores) having a mean of 5 and a standard deviation of 1. The range for goal outcomes is -2 for "much worse than expected" to +2 for "much better than expected." Outcome data can be grouped for analysis without losing the import of individual client goals. A standard score may be generated for each client to evaluate his or her position before and after therapy. Essentially the GAS has the following characteristics: 1) a set of statements of goals for an individual; 2) a system of weights for those goals; 3) a set of expected outcomes for these goals ranging from "most unfavourable" to "most favourable"' 4) a follow-up scoring of these outcomes; and 5) a score summarizing the outcome across all goals. 115 In the assessment interview clients formulated their goals and rated their present level of attainment independently of each other. At termination and follow-up, the level of attainment was measured again. 4. Conflict Resolution Scale (CRS) Subscale of Enriching and Nurturing Relationship Issues, Communication and Happiness (ENRICH) (Fournier, Olson & Druckman, 1983). This measure was completed by each client in the assessment interview as well as at post-test and at four month follow-up. The ten items were specifically developed to identify interpersonal processes that become problematic for many couples. To determine construct validity, the relationship between new measures and existing measures that are consistent with theoretically derived hypotheses relevant to the construct were assessed. The CRS is significantly correlated with the Locke-Wallace Marital Adjustment Scale. The Alpha coefficient for the CRS (Enrich) is .75 and Test-retest reliability is .90. All items are answered on a five point Likert-type scale: 1) Strongly agree, 2) Moderately agree, 3) Neither agree nor disagree, 4) Moderately disagree and 5) Strongly disagree. Raw scores on the CRS are converted into percentile scores so that each individual can be compared to national (U.S.) norms. Individual percentile scores are calculated for both male and female partners. One methodological feature of the CRS is the assessment of social desirability (modified Marital Conventionalization Scale, Edmonds, 1967) and the subsequent correction of individual percentile scores. The individual revised scores adjust each category percentile score according to 1) each individual's relative amount of "idealistic distortion" 2) the empirical relationship between each scale. The couple scores provide a summary of the convergent or divergent opinions that couples have 116 about their relationships. The couple scores were designed to tap the four main dimensions of 1) differences or disagreements in partner responses, 2) potentially negative agreements in partner responses, 3) indecisive responses, and 4) similar responses or agreements that appear to be positive for the relationship. Therapy Process Measures 5. The Couples' Therapy Alliance Scale (AS) (Pinsof & Catherall, 1983). This instrument was completed by each client after the third therapy session as a measure of the client's view of the therapeutic relationship. The measure has three parts: agreement between client and therapist on therapeutic goals, the existence of personal bonds between therapist and client and the development of tasks that are perceived by the client as relevant to his or her concerns. These components are identified by Bordin (1979) as key to the concept of the therapeutic relationship. These are also viewed in relationship to the self, the other partner and the couple's relationship as a whole. The measure is comprised of 28 items: (11 relate to self, 11 to other partner and 6 to the couple relationship), which the subject responds to on a Likert-type seven point scale. This instrument is intended to control for the general or relationship factors which have been shown to be important in predicting therapeutic outcomes. The task dimension which measures engagement in the tasks of therapy would also appear to be particularly relevant in this study to show if clients are equivalently engaged in both types of therapy and perceive them as equally relevant. 6. Therapist Intervention Report. Therapists in this study were required to complete a checklist at the end of each session. Therapists were asked to check off the interventions they used in the session and to estimate and report the 117 frequencjr of use. This checklist is comprised of the same categories of interventions as the Implementation Checklist. 7. Implementation Checklist. The Implementation Checklist is comprised of 25 coding categories, five of which are either general categories or problem definition categories descriptive of interventions common to both therapies. The remaining 20 categories taken from the treatment manuals are made up of ten categories which describe interventions typical of the Sequentially Integrated Systemic therapy and ten which describe interventions typical of the Emotionally Focused therapy. An intervention is defined as a "complete therapist statement," of which the beginning and end are noted by the raters to ensure that they are both evaluating the same units. (Unknown to the raters the odd numbered items on the checklist refer to EF interventions and the even-numbered items refer to SIS interventions.) (See Appendix C for copy.) Subject Variables 8. Demographic Questionnaire. This questionnaire addresses the following: number of years spent together as a cohabiting couple, number of children living in the home, previous marriages or marital therapj' of spouses, approximate amount of family income, age of spouses as well as educational level and presenting occupations. The purpose of this questionnaire is to provide an accurate description of the sample population. 118 Data Analysis Hypotheses The present study is a comparative outcome study to investigate the differential effects of two treatments, an emotionally-focused marital therapy and a sequentially integrated systemic marital therapy and a control condition on the dependent measures of increased marital adjustment, conflict resolution, specified relationship goals and target complaint reduction. The hypothesis is that a sequentially integrated systemic treatment and an emotionally-focused treatment have differential effects on the dependent measures and that these two treatments are more effective than the control condition on these variables as measured at post-test and at follow-up. Because there is little evidence to suggest which treatment might be superior, the alternatives to the null hypotheses are stated as bi-directional. The first null hypothesis relates to the question of how the groups differ from one another at post-test and at follow-up. It is stated as follows: 1) HQ - It is hypothesized that couples exposed to sequentially integrated systemic (SIS) therapy, emotionally-focused (EF) therapy and the control condition will not be different from one another on the four dependent measures of marital adjustment: the Dyadic Adjustment Scale (DAS), the Conflict Resolution Scale (CRS), Target Complaints Reduction (TC) and the Goal Attainment Scale (GAS) at post-test and at follow-up. This hypothesis is applied to the four dependent variables, both separately and jointly. The dependent variables are the couples' combined mean scores on the 119 DAS, CRS, TC and GAS. The second null hypothesis relates to the question of how the groups change over time. It is stated as follows: 2) H/p - It is hypothesized that couples exposed to SIS therapy, EF therapy and the Control condition will not differ on the four dependent measures over time. Each of these groups is analyzed separately over time. The two particular time comparisons of interest are 1) pre-test vs. post-test for the three groups and the two depedent measures, DAS and CRS and 2) post-test vs. follow-up for the SIS and EF groups on all four dependent measures, DAS, CRS, TC and GAS. The dependent variables are the couples' combined mean scores on the above measures. The third null hypothesis relates to the question of how the group differences change over time, i.e. the patterns of interactions. It is stated as follows: 3) HQXT - It is hypothesized that there will be no group by time interaction effects between the SIS and EF groups over occasions. This hypothesis could be restated: the group differences (possibly null) will remain constant over time. The two particular time comparisons of interest are 1) pre vs. post for the SIS and EF groups on the DAS and CRS, and 2) post vs. follow-up for the SIS and EF groups on DAS, CRS, TC and GAS. The dependent variables are the couples' combined mean scores on these four measures. Data Analysis Procedures The data analysis in the main experiment for this study was conducted using a repeated measures design with three treatment groups, three occasions and four dependent measures. Both multivariate and univariate analyses were 120 performed. The analyses sought to answer the questions generated by the three hypotheses in order to determine how the groups differed from one another, how they changed over time and how the group differences changed over time. All three treatment groups were included at post-test but the control group was not included in the four month follow-up. A statistical analysis was also conducted to check for the presence of sex effects. The main analysis was preceded by testing of assumptions of group equivalence with respect to demographic data, pre-test scores and the strength of the alliance between therapist and couples. An implementation check was conducted to determine whether or not the treatments were implemented according to the manuals. An analysis of variance was performed to examine for the presence of a therapist effect with couples nested in therapists nested in treatments; Therapists were treated as a fixed factor in the design. 121 CHAPTER IV: Results and Data Analysis Introduction In this chapter the details of the data analysis are reported. There are three main sections: the first pertains to assumptions of equivalence and validation of test instruments and the last to descriptive data. The essence of this chapter, however, is the second section, which is comprised of the analysis of data for the main experiment. The main experiment was basically a repeated measures design with three treatment groups (Sequentially Integrated Systemic, Emotionally Focused and Control) and three occasions (pre-test, post-test and follow-up). The dependent measures considered were the Dyadic Adjustment Scale (DAS), the Conflict Resolution Scale (CRS), the Target Complaints Instrument (TC) and the Goal Attainment Scale (GAS). Both multivariate and univariate analyses were performed. The analyses were complicated by two factors: 1) no follow-up of the control group was conducted and 2) pre- to post-test change scores were not available for the TC and GAS because they are post-therapy reports. As stated in Chapter 3 in connection with the three hypotheses, the three main questions that these analyses sought to answer were: 1. How did the groups differ from one another? 2. How did the groups change over time? 3. How did the group differences change over time, i.e. what were the patterns of interactions? 122 In answer to the above questions, the exact details with appropriate qualifications are given in the body of this Chapter. In broad outline, the answers were that 1) there were not any group differences at pre-test; 2) at post-test both SIS and EF groups showed increased scores over the controls; 3) at follow-up, the SIS group maintained this increase but the EF group did not maintain it for all of the dependent measures. Although the basic unit of measurement was the couple's score (the combined score of the male and female within the dyad, divided by 2) it was also important to test for the possibility of differential effects of treatment for males and females. Results showed there were no significant effects; these results are elaborated upon in Section III after the section on follow-up. In general, tests were performed at the 5% level of statistical significance. For presentation purposes, the significance probability, p, is given. If p <.05, then the result was considered statistically significant. Tables of means and standard deviations, summary MANOVAs, ANOVAs and significance probabilities of t-tests, denoting comparisons of groups on occasions, comparisons of times for individual groups and significances of individual interactions are included in this Chapter. Section I: Testing Assumptions of Group Equivalence Before investigating the effectiveness of the two treatments it was crucial to determine whether or not assumptions of group equivalence had been met, i.e. to examine the pre-treatment scores of the three groups on the dependent measures in order to rule out the possibility that there were significant initial differences which would confound the results. Drawing meaningful and logical conclusions about the 123 effectiveness of the two therapy treatments would be contingent upon such evaluation. Testing the assumptions of group equivalence was comprised of 1) comparing the two treatment groups and the control group with respect to the demographic data and pre-test scores; 2) comparing the strength of the alliance between the two treatment groups and therapists. It was also necessary to test for the presence of therapist effects and to conduct an implementation check to determine whether or not the treatments were implemented according to the manuals. A multivariate analysis of variance was conducted on 11 demographic variables (group by demographic variable: number of years together, number of children from current marriage, previous therapy, duration of previous therapy, if any, family income, age and level of education of each spouse). In the multivariate analysis there were no significant differences found on these separate factors when tested at the .05 level of significance, corrected by the Bonferroni Procedure (Miller, 1966). Details of the multivariate analysis may be found in Table I. The scores of the two treatment groups on the Couples Therapy Alliance Scale (Pinsof & Catherall, 1983), taken after the third session, were analyzed in order to test the assumption that there were not any significant group differences in the quality of alliances between therapists and their couples. If a significant difference were shown it would present a confounding factor in the interpretation of treatment effects. The maximum possible score on the CTAS scale is 196. The Sequential^ Integrated Systemic couples' mean score was 151.36 (SD = 14.53) and the Emotionally-Focused couples' mean score was 157.82 (SD = 14.79). The results of the analysis of the couples alliance scores are presented in Table II. 124 Table I: Means, standard deviations and ANOVA F values: for demographic variables across groups Variables (N = 42) Groups SIS EF C F P M/SD M/SD M/SD df(2,39) No. of yrs. together 8.29 12.64 12.85 1.10 .342 5.44 10.85 10.30 No. of children 1.14 1.57 1.36 .48 .623 1.17 1.22 1.08 Previous therapy 1.79 1.43 1.64 1.96 .154 .426 .514 .497 Length of therapy (years) .357 .857 .786 .97 .389 .744 1.027 1.25 Combined Income 4.07 3.57 3.93 .36 .702 1.54 1.55 1.73 Male Age 36.71 41.71 41.64 1.52 .232 5.59 10.80 8.92 Previous Marriage (Male) 1.79 1.57 1.64 .72 .492 .426 .514 .497 Male Education Level 4.07 4.64 4.36 .53 .590 1.44 1.39 1.55 Female Age 35.21 37.71 39.43 .86 .430 4.64 10.80 8.96 Previous Marriage (Female) 1.71 1.86 1.79 .40 .672 .469 .363 .426 Female Education Level 4.86 4.07 4.03 2.65 .084 .949 1.14 1.18 125 Table II: Summary ANOVA: Couples Therapy Alliance Scale (Experimental groups) N = 28 Source of Variation SS df MS F P Between 292.509 1 292.509 1.361 .254 Within 5587.51 26 214.905 — — No significant difference was found in this analysis or in an analysis of male and female alliance scores considered separately (Table III), or in an analysis of the individual subtests of the alliance scale. Table III: Couples Therapy Alliance Scale: Means (M), standard deviations (SD) and ANOVA F-values Groups Variables SIS EF F P M/SD M/SD df(l,26) Males 151.29 156.57 .6941 .412 15.63 17.87 Females 151.43 159.07 1.198 .284 17.06 19.79 Therapist Effects Since the effects of therapists are confounded with the treatment groups, it was necessary to test for the presence of therapist effects along with the 126 assumptions of group equivalence. An analysis of variance was performed to examine for the presence of a therapist effect, using a Repeated Measures design with couples nested in therapists nested in treatments. Results showed there were not any significant differential therapist effects on the four outcome measures, DAS, CRS, TC and GAS. The multivariate F statistic using pre, post data was F(28,22) = .57, p = .92. The multivariate F statistic using post, follow-up data was F(56,13) = .85, p = .676. Univariate tests were conducted; they did not indicate differential therapist effects on any of the four outcome measures, DAS, CRS, Target Complaints and Goal Attainment on any occasion (corrected by the Bonferroni procedure X = .05/2 at pre, post and pre, post, follow-up and X = .05/4 at post, follow-up). Based on these results, it was concluded that there were no therapist effects and therapists were therefore not included as a factor in subsequent analyses. Implementation Check The last step before analyzing treatment data was an implementation check to determine if the treatments were implemented according to the treatment manuals. This implementation check was conducted by two trained, independent raters, blind to the treatment conditions they were observing, who rated segments from video and audiotapes of the therapists. A total of 93 out of the 280 sessions or approximately three of each couple's series of ten sessions were chosen at random and observed by the raters. Two segments of ten minutes each were taken from the middle and final third of these 93 sessions. In this way, each couple was observed for a total of 60 minutes of their therapy. As noted in 127 Chapter III, the Implementation Checklist is comprised of 25 coding categories, five of which were either general categories or problem definition categories descriptive of interventions common to both therapies. The remaining 20 categories were made up of ten categories which described interventions typical of the Sequentially Integrated Systemic therapy and ten which described interventions typical of the Emotionally-Focused therapy. An intervention was defined as "a complete therapist statement," of which the beginning and end were noted by the raters to ensure that they were both evaluating the same units. A total of 2,268 interventions were evaluated by the raters. Of the 2,268 interventions coded, 64 or 2.8% were coded so as to be inappropriate to the treatment condition being observed. Of these inappropriately coded, 42 occurred in the Emotionally Focused treatment condition and 22 in the Sequentially Integrated Systemic. The more cognitive, structured and clearly specified nature of the Sequentially Integrated Systemic treatment tended to make it more clearly identifiable. Inter-rater reliability was calculated on 624 interventions taken from 25 randomly chosen sessions (26.9% of the total sessions observed.) The two raters agreed on 584 of the interventions rated (94% agreement). Inter-rater reliability was then calculated, using Cohen's (1960) statistic Kappa. This statistic is a conservative estimate of agreements, corrected for the proportion of agreement to be expected by chance alone. The 40 disagreements that occurred between the raters were comprised of 15 cross-treatment (n=15) disagreements and 25 cross-intervention (n = 25) disagreements. The Kappa statistic for inter-rater agreement regarding treatment was computed at .94. The Kappa statistic for inter-rater agreement regarding interventions was computed at .95. These results, as well as the 128 individual ratings, suggest that the treatments were implemented according to the manuals and that both the overall treatments and the interventions within treatments can be reliably differentiated. Section II: Analysis of Treatment Effects Statistical Background The basic design of the experiment was a repeated measures design with three treatment groups (Control, Sequentially Integrated Systemic and Emotionally-Focused) and 3 occasions (pre, post, follow-up), with the basic unit of measurement being the couple's average score on four dependent measures (DAS, CRS, TC and GAS). As there was no measurement of Controls at follow-up, nor pre, post change scores on the TC and GAS, the basic experiment was re-analysed with the following variations: 1. As a two group (SIS, EF) by three times repeated measures experiment on DAS, CRS 2. As a three group (SIS, EF, Control) by two times (pre, post) repeated measures experiment on DAS, CRS 3. As a two group (SIS, EF) by two times (post, follow-up) repeated measures experiment on TC, GAS. In the repeated measures framework, the basic effects tested are groups, time by group and time. More formally, the null hypotheses, are: 1. HQ - there is no difference between the groups (averaged over time). 2. H-p - there is no difference between times (averaged over groups). 129 3. HGxT " there is no group by time interaction. This group by time null hypothesis HQXT can be restated as: the (possibly null) group differences remain constant over time. This hypothesis makes no statement about the magnitude of group differences and they could possibly be null. To isolate where differences occurred, several re-analyses were done. First, the basic MANOVAs were reanalyzed as univariate experiments. Second, the experiment was analysed, using one time only (this tests the hypothesis HG for each time-there will be no averaging over time). Third, the experiment was analysed using one group only (this tests the hypothesis Hx for that group-there will be no averaging over groups). Finally, one group and one time were dropped (this tests HQXT m the remaining two by two design). In order to combine results from these procedures, which produce non-independent tests, Bonferroni's procedure was used along with other post hoc comparisons. Briefly, the Bonferroni procedure is a method of splitting up a hypothesis into a pre-determined number (n) of sub-hypotheses. The significance probability, p, of a test of each of the sub-hypotheses is computed and compared with X = a/n. If p < X, the sub-hypothesis is significant at level a. For example, if a multivariate hypothesis with four dependent variables were split up into four univariate components, then n = 4. There is one other statistical point of note. This concerns the power of the tests. Since this is basically a repeated measures design, the power of the test of HQ is quite dependent on the "within group between couple" variability, as the group to group comparisons are basically comparisons of different groups of couples. In contrast, the tests of H-p and HQXT are independent of this variability and because this variability is appreciable in this experiment, the test of HQ has 130 appreciably less power. Differences Between the Groups The first part of the analysis of treatment effects addressed the question of how the groups differed from one another (See HQ at pre, post and follow-up). The data for this part of the experiment were analyzed using MANOVAs, then univariate tests, followed by t-tests to compare the groups on individual occasions. Means and standard deviations are shown in Table IV, the MANOVAs are shown in Table V and the ANOVAs in Table VI. Briefly, results suggest (with qualifications) that there were no differences between the groups on the DAS or CRS at pre-test. However, at post-test there was a difference between the two treatment groups, SIS and EF, and the control group on the four dependent measures DAS, CRS, TC & GAS. The two treatment groups were not significantly different from each other but were significantly higher than the controls at post-test (Table VI). At follow-up, there were no significant differences between the SIS and EF groups on any of the four dependent measures. (Table VII). This last statement is strongly qualified and will be discussed in the section pertaining to "Differences Between Times." The differences were not statistically significant but real and will be demonstrated in that analysis. In the remainder of this section the results obtained above are examined in greater detail. Pre-test. This part of the data analysis began with an examination of pre-treatment Dyadic Adjustment Scale (DAS) and Conflict Resolution Scale (CRS) scores to determine whether there were significant initial differences between the three experimental groups on these two dependent measures. Target Complaint 131 Reduction (TC) and Goal Attainment Scale (GAS) were not included because all groups start at the same pre-test score level on these two measures. Table IV: Table of means (M) and standard deviations (SD): Dependent measures DAS Pretest Posttest Follow-up SIS EF C M SD M SD M SD 83.86 8.81 86.32 8.25 82.50 7.11 96.75 13.12 100.14 14.24 80.86 9.93 101.0 8.64 92.05 14.7 CRS Pretest Posttest Follow-up SIS EF C M SD M SD M SD 28.89 2.93 26.07 2.71 27.54 4.13 33.29 4.35 32.14 3.91 28.03 3.31 33.23 4.69 31.68 5.18 Table IV: Table of means (M) and standard deviations (SD):  Dependent measures (cont'd) TC Posttest Follow-up SIS EF C M SD M SD M SD 3.72 .684 4.02 .700 2.46 .709 3.69 .73 3.46 1.05 GAS Posttest Follow-up SIS EF C M SD M SD M SD 5.25 .893 5.71 .699 4.00 .554 5.50 .921 5.00 1.02 SIS = Sequentially Integrated Sj^stemic Group EF = Emotionalry-Focused Group C = Control Group 133 Table V: Summary MANOVAs: for dependent measures Pre, post, follow-up for DAS, CRS N = 22 Source Wilks Approx. F Approx. df p Lambda Groups .82944 1.95 2,19 .169 Time .19642 17.39 4,17 .000 Time x Group .53436 3.70 4,17 .024 Pre, post for DAS, CRS N = 42 Source Wilks Approx. F Approx. df p Lambda Groups .63136 4.91 4,76 .001 Time .53159 16.74 2,38 .000 Time x Group .63858 4.78 4,76 .002 Post, follow-up for DAS, CRS, TC & GAS N = 22 Source Wilks Approx. F Approx. df Lambda Groups .83738 .83 4,17 .527 Time .37459 7.10 4,17 .001 Time x Group .49230 4.38 4,17 .013 134 Table VI: Summary ANOVAs: for dependent measures  Summary ANOVAs For DAS Pre, Post, Follow-up N = 22 (SIS, EF) Source of Variation SS df MS F P Groups 44.988 1 44.988 .24 .642 Subj. (error) 4028.688 20 201.434 Time 2627.454 2 1313.727 25.04 .001 Time x Group 440.602 2 220.301 4.20 .022 Error (B) 2098.938 40 52.473 Pre, Post N = 42 (SIS, EF, C) Source of Variation SS df MS Groups Subj. (error) Time Time x Group Error (B) 2020.212 6676.375 1466.665 1053.008 2029.563 2 39 1 2 39 1010.106 171.189 1466.665 526.504 52.040 5.90 28.18 10.12 .006 .001 .001 Post, Follow-up N = 22 (SIS, EF) Source of Variation SS df MS Groups Subj. (error) Time Time x Group Error '(B) 141.840 3919.188 426.594 316.422 892.250 1 20 1 1 20 141.840 195.659 426.594 316.422 44.612 .73 9.56 7.09 .405 .006 .015 135 Summary ANOVAs for CRS Pre, Post, Follow-up N = 22 (SIS, EF) Source SS df MS Groups Subj. (error) Time Time x Group Error (B) 86.367 420.352 582.957 4.555 444.832 1 20 2 2 40 86.367 21.018 291.479 2.277 11.121 4.11 26.21 .21 .056 .001 .816 Pre, Post N = 42 (SIS, EF, C) Source of Variation SS df MS Groups Subj. (error) Time Time x Group Error (B) 154.820 565.813 280.492 114.352 451.250 2 39 1 2 39 77.410 14.508 280.492 57.176 11.571 5.336 24.242 4.942 .009 .001 .012 Post, Follow-up N = 22 (SIS, EF) Source of Variation SS df MS Groups Subj. (error) Time Time x Group Error (B) 48.082 460.891 14.223 3.266 251.523 1 20 1 1 20 48.082 23.045 14.223 3.266 12.576 2.09 1.13 .26 .164 .300 .616 136 Summary ANOVAs for TC Post, Follow-up N = 22 (SIS, EF) Source of Variation SS df MS F P Groups 0.0 1 0.0 0.0 .999 Subj. (error) 20.023 20 1.001 Time 2.876 1 2.876 15.912 .001 Time x Group .580 1 .580 3.027 .088 Error (B) 3.615 20 .181 Summary ANOVAs for GAS Post, Follow-up N = 22 (SIS, EF) Source of Variation SS df MS F P Groups .278 1 .278 .274 .606 Subj. (error) 20.318 20 1.016 Time 1.278 1 1.278 4.808 .040 Time x Group 1.278 1 1.278 4.807 .040 Error (B) 5.318 20 .266 137 Table VII: Comparison of groups on individual occasions:  Significance probability of F-tests  For DAS Time Between Groups Pre Post Fol. C vs. SIS .658 .001 C vs. EF .201 .000 SIS vs. EF .452 .518 .097 C vs. SIS vs. EF .455 .000 For CRS Time Between Groups Pre Post Fol. C vs. SIS .325 .001 C vs. EF .277 .006 SIS vs. EF .014 .472 .472 C vs. SIS vs. EF .092 .002 138 For TC Time Between Groups Posttest Follow-up t C vs. SIS .000 C vs. EF .000 ... SIS vs. EF .262 .556 For GAS Time Between Groups Posttest Follow-up t C vs. SIS .000 — C vs. EF .000 ... SIS vs. EF .138 .243 139 Multivariate analyses of variance of these pre-test measures did not yield statistical^ significant differences (Wilks A approximate F(4,76) = 1.60, p = .183). Univariate analyses of variance did not show significant differences between the three groups either on DAS or CRS, corrected by the Bonferroni procedure (X = .05/2). t-tests comparing all possible pairs on occasions indicated there were no significant differences between pairs of groups (treatment vs. treatment or treatment vs. control) on the pre-test (Table VII), with the exception of a comparison between the Sequentially Integrated Systemic (SIS) and the Emotionally-Focused (EF) scores on the CRS. Because differences between the three groups on the CRS were not indicated on either the multivariate analysis or the univariate analysis of the CRS, this difference between SIS and EF could be legitmately regarded as a spurious random occurrence; however, to be on the safe side, more attention should be paid in the repeated measures analysis to the "interaction" terms, as related to the CRS. The pretest mean for all of the couples on the DAS was calculated at 84.2 (SD = 8.05). The mean for the SIS group was 83.9 (SD = 8.81), for the EF group it was 86.3 (SD = 8.25) and for the control group it was 82.5 (SD = 7.11) (Table IV). Scores for all three groups ranged from 59 to 116. The mean for all of the couples on the CRS was calculated at 27.5 (SD = 3.43). The mean for the SIS group was 28.9 (SD = 2.92), for the EF group it was 26.1 (SD = 2.7) and for the control group 27.5 (SD = 4.12) (Table IV). Scores for all three groups ranged from 14 to 45. The multivariate homogeneity of variance assumption was checked and significant differences were not found: Box's M = 3.70791, Chi-Square = 3.433 with 6df, p = .753 (approx.) 140 Post-test. In order to test differences among the three treatment groups at post-test a multivariate anaylsis of variance (MANOVA) was conducted between the groups over the four treatment outcome variables: the DAS, CRS, TC and GAS. This analysis suggested that there were statistically significant differences among the three groups: Wilks A approx. F(8,72) = 5.80, p = .000. Univariate analyses of variance showed significant differences among groups on all four dependent measures (DAS, CRS, TC and GAS). In order to avoid the problem of an escalating Type I error rate, the Bonferroni procedure was used to calculate the critical significance level for each univariate test (X = .05/4 = .0125). Using Tukey's procedure (Glass & Stanley, 1970) post hoc comparisons were conducted to find whether mean differences shown were between treatment groups and controls or between the two treatment groups or both. These comparisons yielded the following results: 1. Both treatment groups were significantly higher than controls on the four outcome variables: DAS, CRS, TC and GAS. 2. Although both treatment groups scored significantly higher than the controls, there were not significantly different treatment effects noted between the SIS and the EF groups. These results were confirmed by t-tests which made all possible comparisons between pairs at post-test corrected by the Bonferroni procedure (Table VII). Follow-up. Sixteen weeks after post-test, couples in the two treatment groups completed follow-up measures, which included the four dependent measures, DAS, CRS, TC and GAS. Eleven of the 14 couples in each treatment group completed this follow-up data. Two couples in each group did not complete follow-up because they were separated and one couple in each group did not return 141 the forms. Again at follow-up, the focus was on the question of whether or not the SIS and EF groups would be significantly different from one another. To test the differences among the SIS and EF treatment groups at follow-up, a multivariate analysis of variance (MANOVA) was conducted between the groups over the four treatment outcome variables: DAS, CRS, TC and GAS. This analysis suggested that there were not statistically significant differences between the two groups: Wilks A approx. F(4, 17) = .79, p = .547. Univariate tests suggested there were not significant differences between the two groups on any of these four measures, corrected by the Bonferroni procedure (X = .05/4). These results were confirmed by t-tests which compared the SIS and EF groups at follow-up (Table VI). It is important to note that these group comparisons at follow-up were made with tests that are lacking in statistical power and, strictly speaking, they "fail to show a difference" rather than "show no difference." Differences Between Times The second part of the analysis of treatment effects addressed the question of how the groups changed over time, pre vs. post, post vs. follow-up and pre vs. post vs. follow-up (see Kf, Chapters III, IV). This part of the experiment was analyzed using repeated measured MANOVAs, then univariate tests, followed by dependent t-tests which compared times for individual groups. These results are to be found in Tables V, VI and VIII. Briefly, with regard to pre-test vs. post-test, the results suggest that both the SIS and EF treatment groups made significant gains on the two outcome 142 Table VIII: Comparison of times for individual groups: For DAS Significance probabilities of Group F-tests Time SIS EF C 1 vs. 2 1 vs. 3 2 vs. 3 0.000 0.000 0.688 0.002 0.429 0.007 0.497 1, 2 & 3 .001 .002 For CRS Group SIS EF C 1 vs. 2 1 vs. 3 2 vs. 3 0.012 0.005 0.349 0.000 0.005 0.655 .641 1, 2 & 3 .002 .001 For TC Group Time SIS EF 2 vs. 3 0.133 0.003 For GAS Group Time SIS EF 2 vs. 3 1.00 0.016 143 measures DAS and CRS, whereas the controls did not change significantly over this time. With regard to post-test vs. follow-up, post-test levels were maintained by the SIS group on all four dependent measures, DAS, CRS, TC & GAS in contrast to the EF group, which maintained post-test levels on the CRS but not on the DAS, TC and GAS. In the remainder of this section, the significant results obtained above, are examined in greater detail. Pre-test vs. Post-test. For this part of the data analysis a MANOVA was conducted, which analyzed changes in the three groups respectively over the time between the pre- and post-treatment on the two treatment outcome variables DAS and CRS. (Because the TC and GAS do not yield pre, post change scores, it was not possible to include them in this test. Instead, in analyzing whether post-test changes occurred and were maintained on TC and GAS, it was necessary to examine the post, follow-up progression of the treatment groups on the TC and GAS. It was also relevant to examine how the two treatment groups differed from each another on TC and GAS at follow-up as outlined in the previous section, which compared differences between groups on occasions.) This analysis suggested that there were statistically significant changes in the two groups between pre- and post-treatment: Wilks A approx. F(2,38) = 16.74, p = .000 (Table V). One-way analyses of variance showed significant changes in the two treatment groups from pre- to post- on both the DAS and CRS, (corrected by the Bonferroni procedure) (X = .05/2) (see Table V, pre, post). The MANOVA and ANOVAs were followed by dependent t-tests (corrected by the Bonferroni procedure) which indicated that there were significant differences in both treatment groups on the DAS as well as the CRS between pre- and post-treatment but that there were no significant differences between the pre- and post-treatment period for the control group on either of these dependent measures (Table VIII). These comparisons suggested the following result: Both treatment groups made significant gains from pre- to post-test on the two outcome measures, DAS and CRS, whereas the controls did not change significantly. Post-test vs. Follow-up. The focus at follow-up was on the question of whether the post-test levels attained by the SIS and EF treatment groups on the four dependent measures, would be maintained at follow-up. To determine whether the post-test levels of the two treatment groups were maintained at four-month follow-up, a repeated measures multivariate analysis of variance-MANOVA-was conducted, which analyzed changes in the SIS and EF groups, respectively, over the time between post-test and follow-up, on the four outcome variables. This analysis suggested that there were statistically significant changes within the two groups between post-test and follow-up, Wilks A approx. F(4,17) = 7.10, p = .001, (Table V). In addition to the post-test, follow-up repeated measures MANOVA, an examination of the repeated measures MANOVA which takes into account pre, post, follow-up changes in the groups confirms statistically significant differences within the groups. Wilks Lambda approx. F(4,17) = 17.39, p = .000 (Table V). Repeated measures two-way analyses of variance showed significant differences within the groups from post- to follow-up on the DAS, TC & GAS but did not show significant differences from post- to follow-up on the CRS. (Table VI). To pinpoint these differences, these ANOVAs were followed by dependent t-tests (corrected by the Bonferroni procedure, X = .05/4) which indicated: 1. There were not significant differences within the SIS group between post and 145 follow-up on any of the four dependent measures, DAS, CRS, TC and GAS (Table VIII). 2. There were significant differences in the EF group between post- and follow-up on the dependent measures, DAS, TC and GAS but not on CRS (Table VIII). Inspection of the means (Table IV) indicated a reduction in scores on the EF group. In association with pre-test, post-test findings, these comparisons suggest the following results: 1. Both treatment groups made significant gains from pre- to post-test on the two outcome measures DAS and CRS. Levels on both DAS and CRS were maintained by the SIS group at follow-up. 2. Whereas the post-test levels attained by the SIS group on the two dependent measures DAS and CRS were maintained at follow-up, the EF group maintained post-test levels on the CRS but did not maintain post-test levels on the DAS. 3. The SIS group maintained levels on Target Complaints Reduction and Goal Attainment from post- to follow-up. 4. Whereas the post-test levels attained by the SIS group on Target Complaints reduction and Goal Attainment were maintained at follow-up, the EF group did not maintain these levels on either TC or GAS from post-test to follow-up. 146 Group by Time Interactions The last part of the analysis of treatment effects for the main experiment addressed the question of how the group differences changed over time; specifically what were the patterns of interactions? (See HQXT, Chapter 3,4). This part of the experiment was analyzed with Repeated Measures MANOVAs, then univariate tests, followed by dependent t-tests which indicated the significance of individual interactions. These results are to be found in Tables V, VI and IX. Briefly the pattern on pre vs. post, post vs. follow-up and pre vs. post vs. follow-up generally confirms those from the previous section. (In the previous section, it was suggested that both groups made significant gains at post-test, and that while the SIS group maintained levels on all four measures, at follow-up, levels were maintained by the EF group on CRS but not on DAS, TC and GAS.) There is one exception regarding Target Complaints in the test for interactions which does not confirm previous results; this exception and other results are discussed in greater detail in the remainder of this section. To answer the question of whether or not there were group by occasion interactions between pre-test and post-test a MANOVA was conducted which included the two treatment groups and the two outcome variables DAS and CRS. Results suggested statistically significant differences: Wilks A, approx. F(4, 76) = 4.78, p = .002 (Table V). Univariate analyses of variance showed significant interaction effects from pre- to post- on both the DAS and CRS, corrected by the Bonferroni procedure (X = .05/2) (Table VI). These analyses were followed by t-tests to check the significance of individual interactions. These t-tests (corrected by the Bonferroni procedure X = .05/3) suggested significant interaction effects between the 147 Table IX: Dependent F-tests: Significance of individual interactions For DAS Time Between Groups 1 vs. 2 1 vs. 3 2 vs. 3 1 vs. 2 vs. 3 C vs. SIS C vs. EF SIS vs. EF .001 .001 .823 .034 .015 .022 For CRS Time Between Groups 1 vs. 2 1 vs. 3 2 vs. 3 1 vs. 2 vs. 3 C vs. SIS C vs. EF SIS vs. EF .044 .002 .401 .615 .616 .816 For TC Time Between Groups 2 vs. 3 SIS vs. EF 0.088 For GAS Time Between Groups 2 vs. 3 SIS vs. EF 0.040 148 control group and each of the treatment groups on both the DAS and CRS dependent outcome measures but did not show significant interaction effects between the SIS and EF groups on either of these dependent measures (Table VII). These comparisons added strength to the previous conclusion that both treatment groups made significant gains from pre-test to post-test on the two outcome measures DAS and CRS, whereas the controls did not change significantly on these measures over these times. Group by occasion interactions between post-test and follow-up were examined by a repeated measures MANOVA which included the two treatment groups and the four outcome measures, DAS, CRS, TC and GAS. Confirming previous tests, the analyses suggest statistically significant interactions for the post-test, follow-up period: Wilks A approx. F(4, 17) = 4.38, p = .013 (Table V). Examination of the repeated measures MANOVA which included the two outcome measures DAS and CRS and analyzed pre, post, follow-up as well as post, follow-up interactions suggested there were statistically significant interactions: Wilks A approx. F(4, 17) = 3.70, p = .024 (Table V). To pinpoint these differences, dependent F-tests were conducted (please see Table IX) (corrected by the Bonferroni procedure) which indicated 1. there were significant interaction effects between pre-test and follow-up for the SIS and EF treatment groups on the DAS outcome measure but not on the CRS. 2. there were significant interaction effects between post-test and follow-up for the two outcome measures DAS and GAS but not on the outcome measures CRS & TC. 149 These findings confirm the above-mentioned comparisons of post-test with follow-up and point to an interesting discrepancy between the two treatment groups in maintenance of gains made on the outcome measures DAS and GAS. However, in analyzing the interactions it was noted that, although the results pertaining to the DAS and GAS are similar to those in the previous section, this is not so for the Target Complaints. With reference to TC, the p value for SIS vs. EF for the post vs. follow-up period is .088 (see Table IX), thus failing to reach significance (X = .05). These results on the interaction test are somewhat different from the results on tests of differences over time and suggest that levels attained by the EF group on Target Complaints were maintained relative to the SIS group. Overall, however, there is an indication of some slipping back by the EF group on Target Complaints. Further implications are discussed in Chapter V. There is some apparent contradiction in the results in Section II compared with findings in Section I. The resolution lies in the nature of statistical testing, where the outcomes are, in fact, "proven differences" and "no proven differences" (rather than "differences" or "no differences"). In addition, it should be noted that previous findings which suggested no significant differences between treatment groups at follow-up were based on a test which compares groups with one another and, as noted earlier under "Statistical Background", this is not as powerful a test as the interaction tests used in this experiment. Given the statistically more powerful test, which examines changes within each group over time, and taking into account the factor of less variability within groups, important differences in treatment effects emerge at follow-up. (A statistically more powerful test is one that is more able to detect differences of a given magnitude. Here the differences are of such a magnitude that the more powerful test can detect them but the less powerful one 150 cannot. If, for example, the power of this experiment were to have been increased by having a larger sample, it is a fairly safe conjecture that at least one of the less powerful comparisons of the two groups would have shown significant differences at follow-up.) Implications of these results are discussed further in Chapter V. Analyses of Interaction Effects Between Sex and Treatment over Time In the previous analyses, the unit of measurement was the couples score; this did not, therefore, test for the possibility of differential effects of treatment of males and females. The analyses were not considered complete without investigating whether there were significant differences between male and female responses to the two treatments. Since the composition of each couple was uniformly heterosexual, it is only the changes in sex effects that are of interest. Therefore, the most appropriate question seemed to be: What were the interaction effects between sex and treatment over time? If the most statistically powerful test available was applied to answer this question and ruled out interaction effects between sex and treatment over the three occasions, it could then be concluded that there were not sufficient significant differential effects between males and females to warrant futher investigation. To test for these interaction effects between sex and treatment over time, a repeated measures multivariate analysis of variance was conducted which included the two treatment groups (SIS and EF) and the four outcome measures over time, specifically the two outcome measures, DAS and CRS, at pre, post, follow-up and at pre, post, and the four outcome measures, DAS, CRS, TC and GAS at post, 151 follow-up. (Please see Table X.) The pre, post analysis on the DAS and CRS suggests that there were no statistically significant second order interaction effects for sex and treatments across occasions: Wilks A approx. F(4,154) = .76, p = .555, (Table X). Univariate tests indicated there were not significant sex effects on either the DAS or the CRS, corrected by the Bonferroni procedure (X = .05/2). The post, follow-up analyses using male and female scores on the four outcome measures DAS, CRS, TC and GAS suggests that there were no significant interaction effects among sex and the two treatments from post-test to follow-up: Wilks A approx. (4,37) = .52, p = .722 (Table X). Univariate tests indicated there were not significant sex effects on any of these four measures, corrected by the Bonferroni procedure (X = .05/4). An examination of the repeated measures MANOVA and the univariate tests which analyzed pre, post, follow-up interactions using male and female scores on DAS and CRS confirms that there were not any statistically significant interaction effects among sex, treatments and occasions from pre-test to follow-up: Wilks A approx. F(4,37) = .23, p = .918. See Table X for a complete delineation of the Repeated Measures MANOVA for Sex Effects. These results suggest that there were not differential effects for the Sequentially Integrated Systemic treatment and the Emotionally Focused treatment on males and females over time. 152 Table X: Summary MANOVAs: For DAS, CRS, TC and GAS: Interaction effects between sex and treatment over time For DAS and CRS Pre, Post, Follow-up N = 22 Source Wilks A Approx. F Approx. df P Sex .94744 1.08 2,39 .340 Groups by Sex .99708 .06 2,39 .945 Time by Sex .92201 .78 4,37 .544 Time by Group by Sex .97538 .23 4,37 .918 Pre, Post N = 42 Source Wilks A Approx. F Approx. df P Sex .96447 1.42 2,77 .248 Group by Sex .98923 .21 4,154 .933 Time by Sex .99904 .04 2,77 .964 Time by Group by Sex .96179 .76 4,154 .555 For DAS, CRS, TC and GAS Post, Follow-up N = 22 Source Wilks A Approx. F Approx. df P Sex .94367 .55 4,37 .699 Group by Sex .98823 .11 4,37 .978 Time by Sex .94572 .53 4,37 .714 Time by Group by Sex .94681 .52 4,37 .722 153 Figure 1: DAS at pre, post and follow-up M E A N S C 0 R E S 100 95 90 85 80 • / / 4- / / / / V Sequentially Integrated systemic -l Emotionally focused _L Control 0 PRE-TEST POST-TEST FOLLOW-UP 154 Figure 2: CRS at pre, post and follow-up 33 32 311 30 M E A N S C 0 R E S 29 28 27 26 V Sequentially Integrated systemic Emotionally focused — 1 Control L s L PRE-TEST POST-TEST FOLLOW-UP 155 Figure 3: TC at post and follow-up M E A N S C 0 R E S 4.5 4.0 3.5 3.0 2.5 2.0 < > Sequentially Integrated systemic Emotionally focused Control X < POST- TEST FOLLOW-UP 156 Figure 4: GAS at post and follow-up M E A N S C 0 R E S 5.8 5.5 5.0 4.5 4.0 Sequentially Integrated systemic Emotionally focused .—— Control X 0 L POST-TEST FOLLOW-UP 157 Section III: Descriptive Data This section is comprised of 1) descriptive statistics which address issues of deterioration and proportion of treated couples that improved 2) a synthesis of information gathered from couples in response to informal standardized interviews conducted at post-test and at four month follow-up. Descriptive Statistics In reference to methodological issues in marital therapy outcome studies, authors (Gurman & Kniskern, 1978; Jacobson, 1984) have asserted that although the majority of researchers in outcome studies base their inferences regarding treatment efficac}r on statistical comparisons between treatment and control groups, these group differences in mean performance provide limited information for clinicians attempting to evaluate particular kinds of therapy. Such results do not answer questions about the response to treatment or the amount of change shown by particular couples comprising the sample. Alternatively, these authors have suggested that outcome studies should provide additional information which reveals more about the nature of response patterns of couples within the sample. Jacobson, Follette and Elwood (1984) have suggested inclusion of information concerning deterioration, the proportion of treated couples who improved and, finally, the couples' own evaluation of their response to treatment. (The latter could be based on standardized interviews at post-test and at follow-up.) Such information was gathered regarding this study and it is considered essential to the discussion of the process of therapy which is presented in Chapter V. 158 Proportion of Treated Couples Who Improved. Another way of thinking about the data pertains to the proportion of treated couples who improve. Jacobson (1984) suggests as a criterion for improvement that a couple change from pre-test to post-test by at least 1.96SE. In this study, the couples' scores on the DAS, which is considered to be the most comprehensive and psychometrically established instrument for measuring marital adjustment, were used to calculate such improvement. The criterion for improvement was calculated as a rise in score points of 4.00 based on the formula of Jacobson et al. (1984) which takes into account the reliability of the measure (.96 for the DAS) and the standard deviation of change scores (in this study the standard deviation of change scores was averaged over the three groups, SIS, EF and Controls and calculated at 10.2). Using this criterion, for post-test scores it was found that twelve of the 14 SIS couples (86%), and ten of the 14 EF couples (71%) showed improvement. At follow-up, however, 11 of the remaining 11 SIS couples (100%) and five of the remaining 11 EF couples (45%) showed improvement. The data are displayed in Table XI. Since the above figure of 4.00 score points seems small as a criterion, a more stringent category was established. For this category, called "marked improvement," the reliability of the DAS measure was not taken into account and the SE was calculated as 10.2, which, as mentioned above, is the standard deviation of change scores averaged over the three groups in this sample. The criterion for "marked improvement," then, was a rise in total score points of 19.99 or 20 points (1.96 x 10.2). Using this category, then, four of the SIS couples (29%) and six of the EF couples (43%) showed marked improvement at post-test. At four month follow-up two of the remaining 11 SIS couples (18%) and two of the remaining EF couples (18%) showed marked improvement. 159 Thus, while approximately a third of all of the treated couples in both groups showed marked improvement at post-test, a very high percentage of couples in both groups showed moderate improvement at post-test. However, the distribution across treatments with regard to moderate improvement changed substantially from post-test to four month follow-up, with a substantially higher percentage of the SIS group continuing to show moderate improvement. These findings seem consistent with previous findings using statistical comparisons and are discussed in Chapter V. The data are displayed in Table XL Table XI: Percentages of couples who improved based on the DAS Posttest N = 28 Group n Improved Marked Unimproved Improvement SIS 14 86% 29% 14% EF 14 71% 43% 29% Follow-up N = 22 Group n Improved Marked Unimproved Improvement SIS TT" 100% 18% 0 EF 11 45% 18% 55% 160 Deterioration. In discussing deterioration, Jacobson, Folette and Elwood (1984) stated that in research in behavioural marital therapy approximately 5% of treated couples actually deteriorate (this would be predicated on the assumption that scores were normally distributed). In this study, the four dependent outcome variables all allowed for the measurement of deterioration but again the DAS, the most psychometrically sound, was used to calculate deterioration. If the 5% estimate is applied here, using DAS scores, then as many as two couples in this study would show scores deteriorating to the point of statistical significance. In fact, none of the treated couples' scores decreased to this point, either at post-test or at follow-up, using the previous criterion of 1.96SE (19.99), where SE is equal to the standard deviation of the difference in change scores for the three groups (10.2 in this sample). Thus, none of the couples in either group showed statistically significant deterioration at post-test or at follow-up. Although none of the couples showed significant deterioration, there was some decrease in DAS scores from pre-test to post-test; two of the SIS couples and one of the EF couples' scores dipped below pre-test levels. At follow-up, however, none of the remaining 11 SIS couples' scores dipped below pre-test levels but four of the EF couples' scores did. These findings seem consistent with previous findings using statistical comparisons in that follow-up levels on the DAS were not maintained by the EF group and are discussed in Chapter V. 161 Structured Interviews Posttest Questionnaires. In response to the Post-Treatment Standardized interview, most of the couples noted that their relationship had improved as a result of therapy. Couples in both groups seemed to define relationship changes in the following general ways: 1. Becoming more aware of their own and their partner's thoughts and feelings and being more responsive to these and able to express views more openly to one another. 2. Being able to talk with partners more calmly about important relationship issues 3. Becoming more aware of relationship dynamics 4. Feeling an increase in the level of trust and sense of safety in the relationship and more supported by their partners. The couples in the SIS group referred more to becoming better able to express ideas and views and/or to discuss issues (79%), whereas EF couples were more likely to characterize relationship changes in terms of expressing feelings or emotions (71%). In response to a specific question of what was most helpful in therapy, individuals from both groups referred to therapist objectivity and the safe environment created (75-79 percent in each group) but, of these, members in the SIS group characterized their therapists more as "neutral" whereas the EF members described their therapists as "empathic" and "accepting." Seventy-five to 80 percent of all couples reported improvement in communication patterns and about 60 percent in each treatment group alluded to being able to confront one another calmly about 162 issues. Sixty percent in each group alluded to probing questions by the therapist or memorable experiences in particular sessions. For example, one SIS husband cited the therapist question, "When did you stop thinking that your opinions counted?," as an influence in change. An EF husband remembered a session when his wife expressed her fear of not being in control, noting the important impact that this had made upon him and the relationship. Others in the EF group spoke about "dissolving anger," "understanding and accepting one another's differences," "hearing and feeling my partner's pain," and "emotional release." They commented on becoming more aware of their partners' vulnerabilities and sensitivities and feeling more responded to by their partners. Individuals in the SIS group spoke frequently of the team and noted their contributions, i.e. support, neutrality, consensus, non-judgemental feedback, provocative messages. Several SIS members spoke of the benefits of rebelling against the team and against the team's advice to fight deliberately or to "go slow." Three SIS couples noted that because they knew that others besides the therapist were involved, they felt more confident about the therapeutic suggestions. One SIS wife noted that the message from the team made her realize she didn't feel equal to her husband; an SIS husband reported that the message made him see how he and his wife were caught up in a "vicious cycle." One SIS couple referred to the therapy (the restructuring step) as "allowing more clear communication." One SIS wife reported that the therapist telling them to "go slow" and later predicting a relapse led them to see the purpose of their pattern and added, "it was knowing I had more control and wasn't powerless to change." With regard to attributions of change, the groups differed. In response to an open-ended question, members of the EF group spontaneously attributed change 163 to becoming aware of their own or partner's feelings that had been buried beneath their reactions (65 percent referred to this.) The SIS group members spontaneously pinpointed change as occurring as the result of messages from the team (75% referred to the effect of the team messages and of these, half referred specifically to the positive connotation of their fight by the therapist or team). Individuals were asked to classify improvement in their relationship with regard to changes in thoughts and beliefs about their relationship, changes in experiencing and expression of feelings to partners, and changes in doing things differently with partners and negotiating more. When compared with each other, the treatment groups each placed about equal emphasis on changes in the domains of thoughts and beliefs about their relationships and equal emphasis on doing things differently with partners and negotiating more. However, as might be expected from the nature of the treatment, the EF members rated changes in feelings and emotions and expression of these considerably higher than did the SIS group. (75% of EF members rated changes in feelings high, whereas only 32% of SIS members rated them high.) Within the SIS group itself, couples gave a higher weighting to changes in thoughts and beliefs than to feelings, whereas within the EF group itself, ratings of changes in thoughts and beliefs were about equal to feelings and emotions. A final question of whether more sessions were needed elicited similar responses in both groups. Half of the couples in each group said they would have liked one or two more sessions. Follow-up Questionnaires. At follow-up, four months after the termination of treatment, the focus of the standardized questionnaire was on the remaining 22 couples' (11 in each group) perceptions of how therapj' was helpful to them and in 164 what ways their relationship had changed. Even though there was a greater emphasis on emotional experiencing in the emotionally focused therapy, the majority of individuals in both groups (83% in the SIS group and 80% in the EF group) noted that therapy had helped them by enabling them to express feelings to their partners and to understand their partners' feelings. An interesting but perhaps predictable difference was the EF group's greater emphasis on the importance of feeling understood and accepted by their therapists (82% of the EF group as opposed to only 58% of the SIS group rated this as very helpful in therapy.) With regard to specific perceived changes resulting from therapy, (i.e. behaviour, feelings, self-perception and perception of partners) about half of the couples in each group rated their behaviour as changing substantially. A fairly large number of individuals in both groups noted definite changes in perception of their partners resulting from therapy (67% in the SIS group and 64% in the EF group). However, in regard to changes in self-perception resulting from therapy, only 55% of the EF individuals noted definite changes in self-perception whereas 75% of individuals in the SIS group noted definite changes in self-perception. When asked to rank order seven change processes in therapy, individuals in the SIS group gave the most weight to seeing their partners in a new way (62% ranked this as their most important change in therapy and 21% ranked it as the second most important change). The SIS group also ranked "changing expectations of partners" and "expressing feelings to partner" as either first or second in importance and equally important as each another. In the EF group, the highest emphasis was placed either on "expressing feelings to partner" (45% of individuals ranked this highest) and "seeing partners in a new way" (45% ranked this highest). The third priority was on "feeling accepted and supported by partners" (32% ranked 165 this either first or second highest). Thus at follow-up, individuals in both groups placed emphasis on seeing partners in a new way, and considered this highly important in their change process. However, those in the EF group placed greater emphasis on expressing feelings to partners and feeling accepted and supported by partners as priorities in change whereas SIS members placed emphasis on changing expectations of their partners as well as on expressing feelings to partners. These responses with regard to attributions of change may seem to contradict post-test responses; indeed, there does seem to be greater emphasis by the SIS group at follow-up than at post-test on expression of feelings and less emphasis on positive connotations (regarding the important function of the fight) by the team. However, this contradiction may be more apparent than real: at post-test the SIS group's pinpointing of change as occurring as the result of team messages was elicited in response to an open-ended question about what led to change in therapy. In contrast, at follow-up the same group's pinpointing of change as occurring through seeing their partners in a new way, changing expectations of partners and/or expressing feelings to partners was elicited through an item requiring rank-ordering, where options for responses were provided. While eight of the 11 EF couples spontaneously said they would have liked a monthly check-up to report to their therapist, receive feedback, and continue to work on relationship issues, this was true for only three of the SIS couples. 166 Summary of Results Preliminary tests of equivalence and validations were performed and presented prior to the analysis of the main experiment. These tests showed no significant differences between the three groups on demographic variables and no significant differences between the SIS and EF groups on the quality of alliance between therapists and couples. Tests for the presence of therapist effects suggested there were not any significant differential therapist effects on the four dependent outcome measures DAS, CRS, TC and GAS. An implementation check was conducted; results suggested that the treatments were implemented according to the manuals that both the overall treatments and the interventions within treatments can be reliably differentiated. Examination of repeated measures MANOVA and univariate tests using individual male and female scores on the DAS and CRS confirms there were no statistically significant interaction effects among sex, treatments and occasions (from pre-test to follow-up), ruling out the possibility of differential effects for the SIS and EF treatment groups on males and females. The more crucial tests were for the pre-test equivalence of groups on the real dependent measures used in the main analysis. These showed that there were no statistically significant differences among the three groups on the DAS or CRS at pre-test. These results also addressed the first question in the main analysis: how did the groups differ from one another at pre, post and follow-up. With regard to post-test, the analyses show that the SIS and EF groups, while not significantly different from each other, were significantlj' higher than the controls on the four dependent measures, DAS, CRS, TC and GAS. At follow-up, on these tests, which are statistically less 167 powerful, there were no significant treatment effects indicated between the SIS and EF groups on the same four measures. With regard to the second question about differences in groups over time, results of the statistically more powerful analysis of pre-test vs. post-test suggest that both the SIS and EF treatment groups made significant gains on the two outcome measures DAS and CRS whereas the Controls were not significantly changed. With regard to post-test vs. follow-up, post-test levels were maintained by the SIS group on all four dependent measures in contrast to the EF group, which maintained post-test levels on the CRS but not on the DAS, TC and GAS. The analysis of the interactions confirms the pattern from pre- to post- noted above. On their DAS and CRS scores, the treatment groups both made gains over this time compared with the controls, but neither gained significantly more than the other. Between post- and follow-up the EF scores slipped relative to the SIS scores on DAS and GAS, but not on CRS and only marginally on the TC. 168 CHAPTER V: Discussion of Results Summary The purpose of this study was to outline and describe a sequentially integrated systemic treatment for marital dissatisfaction or conflict and to compare its effectiveness with an emotionally focused treatment, which had previously been delineated and had received some empirical validation (Johnson & Greenberg, 1985). The 42 couples who participated in this study were recruited through a newspaper article and randomly assigned to one of the two treatment groups or the wait-list control group. Each couple in the treatment groups received ten one-hour sessions of conjoint marital therapy administered according to the two treatment manuals. Through an implementation check, it was determined that the treatments had been implemented in accordance with the manuals and were sufficiently disparate so as to be reliably differentiated. Preliminary sests of equivalence prior to the main analysis showed that there were no significant differences among the three groups on demographic variables. There were also no significant differences between the two treatment groups on the quality of couples' alliance with their therapist after initial therapy sessions. The 14 therapists (seven administering each treatment) were balanced with regard to educational background and experience and there was no evidence to suggest the presence of differential therapist effects. Examination of the repeated measures MANOVA which tested for sex effects ruled out the possibility of differential effects of sex for the two treatments. The main analysis is related to the three research hypotheses (HQ, HT and HC-xT) which were generated by the three research questions. The multivariate 169 analysis which compared the groups with one another on the measures of marital adjustment, conflict resolution target complaints improvement and goal attainment (HQ) shows that the SIS and EF groups, while not significantly different from each other, are significantly higher than the controls at post-test. These results suggest that the couples in the two treatment groups improved on these four measures from pre-test to post-test, whereas the controls did not. Results of a statistically more powerful repeated measures multivariate test that was used to examine differences within groups over time (Hf) also indicate that both the SIS and EF treatment groups made significant gains at post-test on the two repeated measures of marital adjustment and conflict resolution. The analysis of the interactions OHGxT) confirms the pre to post pattern noted above: On their DAS and CRS scores, both treatment groups made gains over this period compared with the controls, but neither gained significantly more than the other. Thus there is sufficient evidence to indicate rejection of the null hypothesis and support the conclusion that both treatments were effective. At follow-up, there were no significant differences in treatment effects suggested when the two treatment groups were compared with each other. However, this test (HQ) was the least powerful of the three tests used. The second test examined differences within groups over time and the third was an analysis of interaction effects. Results from these more powerful tests generally suggest that whereas post-test levels were maintained by the SIS group on the measures of marital adjustment, conflict resolution, target complaints and goal attainment, the EF group maintained post-test levels on conflict resolution but did not maintain levels on a measure of marital adjustment and slipped back on target complaint improvement and goal attainment. Although this evidence suggests that 170 the EF group did to some extent slip back on Target Complaints and Goal Attainment at four month follow-up, EF couples did continue to achieve target complaint improvement and reach pre-test goals. Before this study, the SIS treatment had not been empirically tested-nor had any form of structural/strategic marital therapy. The emotionally-focused marital therapy had been tested before (Johnson & Greenberg, 1984). To the extent that the EF therapy was found to be effective (particularly at post-test) in the present study, the findings replicate those of Johnson and Greenberg's (1984) study which was seminal in establishing the effectiveness of EF therapy. In that study the EF therapy was compared with behavioural problem-solving (PS) marital therapy; couples in both groups improved but the EF treatment was found to be considerably more effective than the PS treatment. There were two main differences between Johnson and Greenberg's study and the present one: 1) Johnson and Greenberg's sample was comprised of "moderately distressed" couples whereas these couples were considered to be more severely distressed, 2) gains made by EF couples in that study were maintained at two month follow-up whereas EF couples in this study did not maintain some of their gains over a longer four month follow-up period. Discussion regarding the categories of distressed couples and the implications of these differences are presented later in this chapter. Overall, the results of this study at post-test suggest that two very different treatments stemming from two very different theoretical frameworks are both effective in helping couples alleviate marital distress and/or resolve conflict patterns, but that the Sequentially Integrated Systemic therapy is more self-sustaining. One treatment, EF therapy, focuses on encouraging the expression and understanding of affective, inner experiences in partners. The other, SIS therapy, focuses exclusively 171 on current marital interactions, reframing patterns of behaviour and prescribing symptoms. Although both treatments seem to lead to positive changes in marital satisfaction and conflict resolution and enable couples to reduce target complaints and reach goals, the SIS therapy seems to have more lasting effects. Neither of these treatments is comprised of the skill training interventions associated with behavioural marital therap}', the results of this study seem to support Johnson and Greenberg's (1985) conclusion that distressed couples do not necessarily require problem solving and skill training in order to resolve differences or change interactional patterns. In What Ways Were the Treatments Effective? Before considering the implications of the difference in treatment effects at follow-up, the discussion will centre on 1) clarification of how or in what specific ways the treatments were effective at post-test and 2) elaboration of ways of evaluating treatment effectiveness other than tests of statistical significance. 1. Outcome Measures. Gains made on the Dyadic Adjustment Scale suggest that the couples' level of marital adjustment improved during the course of both SIS and EF therapies. Spanier (1976) has defined marital adjustment as an "ever-changing" process rather than an unchanging state, comprised of dyadic satisfaction, consensus, cohesion and affectional expression. Gains made on the Conflict Resolution Scale reflect improvement in interpersonal processes which had been problematic for couples. Items are related to the subject of conflict, itself. Couples were asked to rate themselves on items, for example, "I sometimes feel that our arguments go on and on and never get resolved." Gains made on Target 172 Complaints reflect improvement on three different complaints related to the couple's main conflict in their relationship. For example, one wife identified friction over matters of organization and finance as her prevailing target for change. One husband wanted to reduce the number of their fights per week. Gains made on Goal Attainment reflect specific behavioural goals in relation to the couple's main presenting concern. For example, one husband stated as a goal, "getting organized to take care of all the details in everyday life so that there would be fewer tensions in the relationship." Specific measurable results which show that the goals had been achieved were then elicited from clients. As an example of part of the process, with the above-mentioned husband, the expected results were elicited as "coming home from work and having a congenial evening with no discussion of any mess that I had left in the morning." 2. Standardized Interviews. It is interesting to note that responses to post-test questionnaires, like the data from the outcome measures, reflect similarities in the impact of the two treatments. Despite the focus on reframing interactions and symptom prescription in the SIS treatment as opposed to the focus on inner experiencing and expression of feelings in the EF treatment, at post-test couples in both groups shared common perceptions about the dynamics of their change process. For example, an equal number in both groups indicated that the treatments had made an impact on their awareness of their relationship dynamics and communication patterns. An equal number also experienced an increase in the level of trust and sense of safety in their relationships, felt more supported, more able to talk calmly about important issues, do things differently with partners and negotiate differently with one another. These latter effects, which seem to reflect greater overall relationship comfort and mutual support, might be seen as 173 predictable outgrowths of successful treatment based on either systemic (SIS) or experiential (EF) approaches. It could be argued that the above-noted similarities in responses related to interactional and communication patterns might be expected partly because of the overlap in the first two steps of both treatments, in which partners are helped to define issues and interactional patterns. However, what is somewhat surprising is that an equal number of couples in both groups emphasized changes in their ability to talk about their thoughts and beliefs about their relationship and also stressed that they had become more aware of their partner's thoughts as well as feelings. It seems that, although the focus in the SIS treatment is more on thoughts and beliefs and in the EF treatment, more on emotional experiencing, the changes in emotional experiencing in the EF therapy may lead to clients changing their thoughts and beliefs as well and, conversely, changes in the meaning attached to situations which are involved in the reframing process in SIS therapy may lead to clients changing their feelings. Responses to post-test questionnaires suggest that although couples shared many common perceptions, they were also affected by the two treatments in different ways. When commenting spontaneously on the effects of therapy, EF couples referred to emotional responses such as "dissolving anger," feeling an "emotional release" or becoming more aware of their partner's sensitivities and vulnerabilities. Typical informal comments by SIS couples included frequent references to the team—their expertise and neutrality as well as the provocative messages of "go slow" and "don't change." SIS couples seemed to feel that such messages led to their new, often contrary responses and to changes in their interaction patterns. The awareness of the presence of the team seemed to make couples feel, by their own admission, more confident about the therapeutic 174 suggestions. SIS members also characterized their therapists as neutral, whereas EF members tended to refer to their therapists as empathic and accepting. The tendency of EF couples to see their therapists as supportive may have induced confidence in therapeutic suggestions and compensated for the confidence in the team that was expressed by the SIS couples. With regard to their own attributions of causes of change, the groups also differed. Again, members of the SIS group invoked the team, pinpointing change as occurring through team messages and the team's positive connotation of the important function that their fights had in their relationships. As might be expected, EF members attributed change to their becoming aware of their own and partners' underlying feelings buried beneath their interactive cycles. Finally, EF members gave considerably higher ratings than SIS members to changes in feelings and emotions. These differences in members' responses to post-test questionnaires suggest that the groups did experience substantial qualitatively different responses to the two treatments. Additional Considerations in Evaluating Treatment Evaluation of the treatment effectiveness in this study is largely based on statistical analysis of the group differences in mean performance, magnitude of change in the two treatment groups over time and interaction patterns. Because this statistical analysis does not address questions about the nature of individual couples' response patterns, additional descriptive information is provided about the percentage of couples who improved and the number of couples who deteriorated. This descriptive information seems consistent with the statistical information and 175 supports the overall conclusion that both treatments were effective and that the integrated Systemic therapy had more lasting effects. Another consideration in evaluating treatment outcome which has not yet been discussed is the clinically significant changes made by couples during therapy. Jacobson, Folette and Elwood (1984) believe that one important concern in marital therapy studies is the number of couples that leave therapy with a non-distressed rather than distressed relationship. They operationalize a non-distressed relationship as "one in which the spouses score within the non-distressed range on one or more pre-selected measures of marital functioning" (p. 117). These authors believe that criteria such as this, which are related to couples' level of satisfaction derived from their relationship, are universally relevant. Supporting this position, Revenstorf, Hahlweg, Schindler and Kunert (1984) suggest that in research in marital therapy literature, couples whose relationships remain in the distressed level of functioning are often described as treatment failures, even though they have shown a significant amount or magnitude of positive change. By the above criterion, the 67 percent of all treatment couples in this study who finished therapy with relationships rated in the non-distressed range would be considered successful but the 33 percent who finished in the moderate to severely distressed range would be considered treatment failures. (The criteria for severely, moderately distressed and non-distressed were based on Spanier's (1976) calculations of the total mean DAS sample scores for married (X= 114.8, SD=17.8) and divorced (X = 70.7, SD = 23.8) couples. In this study, those couples who scored above 100 on the DAS were judged to be in the "non-distressed" range, those below 100 and above 85 to be in the "moderately distressed" range and those below 85 to be "severely distressed.") To consider those 33% as treatment failures 176 would seem to overlook the fact that they had made substantial gains, showing either significant or marked improvement during the course of therapy. Similarly, considering those who finished therapy with relationships in the moderately distressed range as treatment failures would not account for the fact that all treated couples had been exposed to only brief, time-limited therapies in which some change rather than a "cure" was the goal. Accepting the premises of Jacobson et al. (1984) and Revenstorf et al. (1984) and viewing clients whose relationships are in the distressed range at post-test as treatment failures seems extreme and serves to exclude other important criteria of change. Three other criteria which seem relevant to treatment outcome are: 1) where couples were functioning at the outset, i.e. non-distressed, severely distressed, 2) percentage of couples who improved significantly, and 3) the duration and frequency of the treatment program itself. In developing a broader perspective of viewing treatment outcome, then, it is agreed that it does seem important to acknowledge the relevance of descriptive categories such as distressed vs. non-distressed, percentage of couples improved and deterioration. However, it seems equally important not to lose sight of valid empirical considerations such as comparisons of group means, magnitude of change over time and interaction effects, all of which lend themselves to drawing meaningful inferences about the value of therapy treatment, in what specific ways it works and when and for whom it works best. 177 Differences in Treatment Effects at Follow-up Statistical tests show that while post-test levels were maintained by the SIS group on all four outcome measures, the EF group did not maintain levels on three of those four measures. As stated, the EF group did maintain their post-test changes in conflict resolution but relapsed toward pre-test levels on the measure of marital adjustment (the DAS), and slipped back on target complaint improvement and goal attainment. Although the more powerful statistical tests suggest that the EF group slipped back on TC and GAS at follow-up., actual follow-up data showed that EF couples continued to achieve target complaints improvement and reach pre-test goals. It seems clear from observing the actual mean scores of both groups on TC and GAS at follow-up (Table IV) that the EF group was functioning in a comparable range to the SIS group and did not demonstrate clinically significant differences. Examination of the mean scores of EF couples on TC at follow-up shows that they ranged from "slightly improved" to "somewhat improved" on their target complaints that were elicited before therapy and examination of EF couples' mean scores on GAS at follow-up shows that they still achieved "expected results" on the goals set out before therapy. At follow-up, SIS couples' scores on TC were slightly higher than those of EF couples but also between "slightly improved" to "somewhat improved" and on GAS, SIS couples were slightly higher but the overall average was between "expected results" and "better than expected results." Indeed, from a clinical perspective, the conclusion that the two groups were functioning at similar levels on target complaint improvement and goal attainment is supported by the first statistical test, which compares the treatment groups at follow-up and shows that the groups were not significantly different from 178 each other. Descriptive data supports the finding that EF couples did not maintain their improvement as much as SIS couples at follow-up. Data pertaining to the percentage of couples who improved shows a contrast between post-test and follow-up: at post-test, 86 percent of the SIS couples and 71 percent of the EF couples showed significant improvement whereas at follow-up one hundred percent of the SIS couples showed significant improvement but only 45 percent of the EF couples continued to show significant improvement. At follow-up only 45 percent of SIS couples had slipped back into the distressed range but a substantially higher number (73%) of EF couples were in that range. Although these couples do not necessarily represent treatment failures, there are a sufficient percentage of EF couples who returned to distressed levels to warrant further investigation. Indeed the implications of the fact that a large percentage of couples were distressed when they began therapy, combined with the fact that more of the EF couples reverted to distressed levels in contrast to the SIS couples, forms the core of the following discussion. Results of the present study suggest that while both SIS and EF therapies are effective, the Sequentially Integrated Systemic therapy has a more lasting or self-sustaining effect. In order to make a definitive statement which also addresses the issue of generalizability, it seems critical to consider the relevant factors in the couple population that comprised the sample of the study. One such variable, which seems a key consideration in evaluating outcome in this study, is the level of marital functioning of the sample. Couples treated were drawn from a population with a mean on the DAS of 84.2 and their relationships ranged from moderately to severely distressed. At the outset all of the couples' relationships 179 ranged from moderately to severely distressed (scores between 96 and 71.5 on the DAS). Fifty-seven percent of all treated couples had relationships classified as severely distressed at pre-test, barely above Spanier's (1976) criteria for divorced couples (X = 70.7). A more definitive conclusion would be, then, that the effects of the sequentially integrated systemic marital therapy are more lasting or self-sustaining than those of the emotionally-focused therapy with couples whose relationships are more severely distressed. There may seem to be some inconsistency between the above findings and Johnson and Greenberg1 s (1985) findings in that their EF group did not relapse at follow-up. However, it should be noted that Johnson and Greenberg's follow-up period was shorter, being of only two months duration, and couples in their study were considerably less distressed at the outset than those in the present study (the authors identified their couples' population as having moderately distressed relationships, with a mean on the DAS of 92.1, as opposed to the present population of couples with more severely distressed relationships, with a mean of 84.2). Moreover, EF couples in the present study did achieve higher levels, in the non-distressed range, at post-test. Before making inferences about the difference between groups in treatment effects at follow-up, it seems relevant to provide a brief description of the characteristics of a "distressed relationship." The categories of "severely distressed," "moderately distressed" and "non-distressed" have been operationally defined according to score limits on the DAS; however, a definition of a "distressed relationship" has not been provided. Gottman, Markman and Notarius (1977) generally found that "distressed clinical couples" are likely to be more negative toward each other and more likely to reciprocate negative affect than non-clinical couples. Although couples 180 with either distressed or non-distressed relationships in their study engaged in problem solving and "mind reading," couples in distressed relationships were more likely to deliver their problem solving or mind reading statements with negative affect. The authors refer to the proclivity of each partner in a distressed relationship to talk about himself or herself and his or her position, "in distressed couples communication is more likely to be characterized by a "summarizing self syndrome rather than summarizing the spouse or summarizing both positions" (p. 469). Guerin (1982) characterizes severe marital conflict by "intense projection, a high degree of mutual blaming and total inability to maintain self-focus on the part of either spouse" (p. 15). By not maintaining a self-focus Guerin seems to be referring to the notion of each person not taking at least half responsibility for his or her own contribution to negative interactions or fights; the opposite would be the ability to see the other's point of view much of the time. Guerin adds, "In these cases the clinician spends all or most of his or her time trying to keep the instantaneous reactivity in the relationship under some degree of control" (p. 15). L'Abate (1983) believes that closeness-distance issues may be a crucial determinant of distressed relationships. He describes a "depressed" relationship as one in which partners are inflexible because of a dependent mode; neither has achieved an appropriate sense of mastery and as a consequence both are excessively vulnerable to loss because they tend to increase their dependency during stress. Hinchcliffe, Hooper and Roberts (1978) notes that such a pattern is counterproductive because it does not allow the couple to support each other successfully in threatening situations. L'Abate (1983) concludes that spouses who are dependent on each other to provide specific complementary role functions show greater distress, whereas egalitarian marriages may provide the best conditions for satisfactory interaction. 181 Couples in distressed relationships have been described as inclined to reciprocal negative affect, including anger and blaming. A key premise of the emotionally-focused therapy is that underneath negative affect and anger there is a great deal of unresolved and unexpressed pain and fear of further hurt; it is believed that the expression and sharing of these hurtful feelings-of grief, vulnerability, inadequacy, loneliness and/or poor self-esteem—leads to the development of emotional intimacy (L'Abate, 1977). For this reason, the development of intimac}7 is a central goal of the Emotionally-Focused therapy. The therapeutic push is for partners to become more intimate by becoming aware of those unresolved, unexpressed feelings of hurt and fear beneath the anger. The therapist's role in the EF marital treatment is to acknowledge the presence of these hurt feelings and indicate their powerful effects on the marital relationship. The hope and belief is that by experiencing intimacy in the relationship the couple will also resolve relationship issues and achieve greater marital satisfaction. The EF therapy is designed to help partners access unacknowledged feelings underlying problematic reactions and to share and accept each other's feelings. It is felt that through this process the couple can attain intimacy. The process following the preliminary steps involving identification of issues and the negative cycle, begins with: Step 3: Access and accept unacknowledged feelings underlying problematic interactions, and continues through Steps 4, 5, 6 and 7; Step 4: Redefine the problem in terms of newly synthesized emotional experiences; 5: Identify with disowned aspects of experience in the redefined cycle; 6: Accept partner's position and 7: Express and clarify needs and wants. (For a more elaborated version of each step, please see the EF manual, Appendix B). 182 Implicitly and by definition, couples who had accomplished the goals of Steps 6 and 7 would no longer be distressed and would be highly resolved with regard to relationship difficulties. If, for example, they were able to communicate successfully about their newly experienced emotional responses and accept one another's positions (Step 6) they would not be inclined to demonstrate distressed characteristics, such as very negative affect or a "summarizing self syndrome. If partners had attained Step 7 and were clarifying thoughts and feelings to each other and asking spouses to meet their needs and wants they would not likely be engaged in mutual blaming or manifesting an inability to maintain a self-focus, as Guerin's (1982) severely distressed couples. The last two steps, 8 and 9, include the emergence and establishment of new solutions and integration of new positions by the couple. Involvement in these processes presupposes that couples would no longer be functioning in the distressed range. Although the steps tend to be cyclical in application, successful accomplishment of the goals of steps 8 and 9 would depend on partners' completing the previous steps of achieving intimacy and resolution of relationship issues. Although the data suggests that the EF group had improved significantly at post-test, data at four-month follow-up suggests that couples in the EF group relapsed to some extent, particularly on the DAS. At post-test, 71 percent of EF couples were significantlj' improved and the same percentage were functioning in a non-distressed range; at follow-up, only 45 percent of EF couples showed significant improvement and 73 percent were again functioning in a "distressed" range on the DAS. These follow-up results suggest that exposure of couples whose relationships were more severely distressed to a brief, time-limited course of emotionally focused therapy did not lead to lasting changes, particularly in overall level of marital 183 adjustment. One interpretation which seems most cogent, particularly in light of Johnson and Greenberg's (1985) results, where EF couples had not relapsed at two month follow-up, is that a brief, emotionally-focused therapy administered to couples whose relationships were more severely distressed did not create sufficient intimacy or strong enough relationship resolution to engender lasting change. It is likely that the fact that couples in the present study had more severely distressed relationships at the outset made the attainment of emotional intimacy and relationship resolution more difficult. Following what is known about distressed couples, these couples were probably engaged in more mutual blaming and demonstrating more negative affect toward one another than less distressed couples. For this reason, it is probable that their negative interactional patterns had become reactivated after therapy. There was likely a greater inability on the part of partners to maintain the self-focus necessary to access underlying feelings. The greater tendency of partners to "summarize self rather than summarizing spouses undoubtedly limited their ability to acknowledge one another's vulnerabilities and consider the other's point of view. Given these circumstances couples would be less able to make lasting changes, particularly with only brief therapy. It would seem particularly difficult for couples whose relationships were severely distressed to reach the level of intimacy and resolution that would be necessary if they were to continue negotiation processes on their own, create new solutions and integrate new perspectives. From an affectively-based perspective, then, the expression of needs and wants, the development of emotional intimacy and the resolution of relationship issues are considered central to the change process and the power of change to hold in marital therapy. Without intimacy and resolution, the next and final steps 184 of creating new solutions and integrating new perspectives does not seem possible and change, if it has occurred, is not likely to last. The participation in the present study of couples with more severely distressed relationships who were exposed to EF therapy of only brief duration seems to be a contributing factor in the relapse of the EF group at follow-up. A further possible interpretation of the EF group's relapse on the DAS lies in the nature of the experiential therapy which is more evocative than the systemic therapy. This may have led couples to idealize their relationships and may have created higher, even somewhat unrealistic expectations of therapy. These idealistic expectations may have made EF couples more susceptible to feelings of disillusionment during the period immediately following therapy and this phenomenon could have been reflected in their performance on a measure of overall marital satisfaction at four month follow-up. It is interesting to note that EF couples' scores on the Conflict Resolution Scale at follow-up do not suggest a relapse as do their scores on the other measures, particularly the Dyadic Adjustment Scale. In searching for possible reasons, it was noted that rather than being focused on substantive issues or issues of affection, cohesion and intimacy (as the DAS), the ten items on the CRS are focused specifically on conflict processes or the "fight," itself, i.e. "Sometimes we have serious disputes over unimportant issues." Exposure to EF therapy may have made couples more aware of their conflict behaviours and led to a concomitant desire to stop fighting and negotiate more. However, the discrepancy between EF couples' performance on CRS and the other three measures at follow-up suggests that while they had improved with regard to conflict behaviours, they had not necessarily improved on more global measures of marital satisfaction (DAS). To 185 some extent they did seem to have reached their targets and goals. It seems that specific conflict resolution or a reduction in fighting behaviour and even some target complaint improvement and attainment of specific relationship goals can occur without overall improvement in marital adjustment, which may depend on other factors such as attainment of intimacy or resolution of more encompassing relationship issues. It seems that although EF therapy did not seem to create sufficient emotional intimacy and resolution of relationship issues to be self-perpetuating with couples in more severely distressed relationships, particularly in maintaining marital satisfaction, this was not true for the SIS therapy, which had a more sustained effect with the same population. Data have been presented which suggest that exposure of more severely distressed couples to a brief, time-limited course of Sequentially Integrated Systemic therapy led to more lasting changes, particularly in levels of marital adjustment. A brief review of the premises of SIS therapy provides a basis for discussion of possible qualities which may have contributed to its lasting effect. Sequentially Integrated Systemic marital therapy is directed primarily toward change at the interactional level. One premise is that changing repetitive, self-perpetuating interactional cycles between partners will lead to second order change, not only in behaviours but in the rules governing interactions and in the process of the relationship. Because marital conflict is not seen to follow from previous psychopathology in partners and awareness of internal processes is seen as unnecessary for change, the sequentially integrated systemic therapy, unlike the emotionally-focused therapy, is not directly aimed at helping partners to access and acknowledge underlying feelings, express needs and v/ants to one another, attain 186 intimacy or even to resolve relationship difficulties, within the context of the therapy itself. The basic assumption of the sequentially integrated systemic therapy is that when change is achieved in the interactional domain of experience it will spontaneously spread to other domains-including the internal and experiential. Thus therapy is instead directed toward initiating a reversal in repetitive negative communication patterns or interactional cycles and to changing the frame of reference or meaning attributed to these phenomena. In the sense that the therapist's primary aim is to initiate this reversal but not to attempt to get partners to resolve relationship difficulties during treatment, as in EF therapy, the immediate goals of SIS therapy could be said to be less extensive. One of the mechanisms for change in SIS therapy is believed to be the reframing of the symptom (negative interactional cycle) by changing its conceptual or emotional meaning. The therapist's positioning, i.e. accepting the clients' frame of reference and restraint, is designed to discourage resistance and create a context for change. The process (following the preliminary steps of defining issues and identifying the negative cycle) begins with Step 3: Restructuring, and continues through Steps 4, 5, 6 and 7; Step 4: Reframing the problem; Step 5: Restraining; Step 6: Consolidating the frame and Step 7: Prescribing a relapse. (For a more elaborated version of each step, please see the Sequentially Integrated Systemic Marital Therapy Manual, Appendix A). In searching for an explanation as to why the SIS therapy was more self-sustaining than the EF therapy, differences were noted between the two treatments which may have interacted differentially with couples in more severely distressed relationships. In considering the particular characteristics of the couples in distressed relationships and the way that SIS therapy may work in a more 187 persistent fashion with this group, it was postulated that such couples are caught up in more entrenched fight cycles than other couples and probably view their relationships more negatively, tending to become deeply discouraged and perhaps despairing of possibilities for change. Perhaps because these fight cycles become so entrenched and negative affect (resentment, anger) so strong with these couples, a brief period of EF therapy may not be sufficient to enable them to access underlying feelings, express newly synthesized needs and wants or acknowledge these to their partners and achieve enough intimacy to create lasting relationship change. Because of the intense negative feelings and consequent negative conceptualization of their relationship, an approach which offers distressed couples an alternative, positive way of thinking about their interaction may be necessary for enduring change. The SIS therapy seems highly dependent on partners' changing their perceptions about their relationship and making more positive meaning attributions about their negative interactional patterns. Once the interaction changes, it seems self-sustaining. The content of the therapist's message, which embodies the reframe and prescribes the symptom seems to provide a conceptual framework through which couples can view their relationship and a guide for making new "rules" about the relationship. As noted earlier, SIS couples indicated at post-test that they were strongly influenced by the team's provocative messages, expertise and neutrality. The fact that the message represents the consensus of the team, a highly respected group, seems to lend credence and legitimacy. Therapist neutrality, the position of attributing positive meaning or function to negative interaction cycles, and restraining and consolidation of the frame seem to set a process in motion whereby couples see the positive function of their fighting or negative patterns and, based on this overall view, are able to be more 188 self-accepting as well as accepting of partners. This level of acceptance may pave the way for more realistic expectations of themselves, their relationship and goals for the therapy. Couples are likely provided with an ongoing positive framework for viewing their relationship, one which is much needed in such a distressed and conflictual situation. Such a framework probably engenders second order change in which change occurs in the rules governing the interaction rather than in the interaction itself. Once the rules governing interactions have changed, there is no issue of maintaining this by "good practice" as in the learning theory view. Instead, the interaction has been reframed and this makes options for new actions or behaviours available. Systemic and interactional therapists (Selvini-Palazzoli et al., 1978; Watzlawick et al., 1974) have hypothesized that brief therapy works by initiating a reversal in negative interactional or communicational patterns and that this reversal is enough to lead to enduring change, even with distressed clients. The results of the present study seem to attest to this. The fact that the SIS therapy was brief did not seem to interefere with its ongoing effectiveness with severely distressed couples. Perhaps through its application, a self-sustaining process of change was set in motion. To determine whether, in fact, follow-up data supported any of these speculations, couples' responses were reviewed from the follow-up questionnaires. It is interesting to note that even though the therapies are believed to work in different domains, couples in both groups shared similar perceptions about how their relationships had changed (83 percent of SIS couples and 80 percent of EF couples noted that therapy had helped by enabling them to express feelings and understand partners and half of the couples in both groups rated their behaviour as changing 189 substantially). It seems that even though the nature of the therapies is very different, there are some parallels in effects. More important, being able to express and understand feelings is not necessarily, in itself, the marker for lasting change. Couples' responses in regard to self-perception seem relevant to the discussion regarding the more lasting effect of the SIS therapy. More SIS couples noted changes in self-perception and more SIS couples ranked changes in expectations of partners as highly important whereas EF couples did not. First, it seems as if spouses' changes in self-perception may be linked in some way with changes in conceptualizing their relationship interactional patterns and in turn in forming more realistic expectations of self and partners, and so may render lasting relationship change more attainable. Second, such changes in self-perception may reflect a more positive attitude or feeling about self which may generalize to the relationship. Third, changes in self-perception may lead to a greater ability to "self-focus," which maj7 lead to greater relationship resolution. (Self-focus has been seen as a marker of less distressed partners (Guerin, 1982)). Finally, the observation that a much greater number of EF couples (eight out of 11 as opposed to three out of 11 SIS couples) expressed a desire for monthly check-ups lends weight to the notion that SIS therapy is more self-sustaining and leads to second order change, in which the need for practice or "boosters" is not necessary as in EF therapy, where changes need to be consolidated or maintained by continued work. 190 Conclusions The main findings of this study are that two separate, reliably differentiated treatments for marital therapy stemming from different theoretical frameworks were effective in increasing marital adjustment, resolving conflict, improving target complaints and attaining relationship goals with a population of couples whose relationships were in the moderately to severely distressed range. Although the therapies were both found to be effective at post-test, the SIS therapy, which had not been empirically tested, demonstrated more lasting effects at follow-up. The EF therapy is directed toward helping partners to access, express and accept feelings underlying problematic interactions in order to create emotional intimacy and relationship resolution. The SIS therapy is directed toward changing the frame of reference or meaning attributed to negative interactional patterns in order to create second order change—in the rules governing the interactions and the process of the relationship. These results suggest that an experiential, affectively based treatment and an interactional systemic treatment are both viable approaches for creating change in marital therapy, but that the latter treatment is more self-sustaining. The success of the EF treatment attests to the significance of emotional experience in modifying the nature of intimate relationships. The success of the SIS treatment attests to the significance of reframing of negative interactions in the creation of lasting change in intimate relationships. Although the conclusion of this study is that two different treatments were both effective in alleviating marital distress, gains made by the two treatment groups at post-test were not consistently maintained at follow-up. Post-test levels were maintained at follow-up by the SIS group but there was some relapsing by 191 the EF group, particularly in the domain of marital adjustment. The EF group did, however, continue to improve on target complaints and reach acceptable levels regarding relationship goals. A further, more refined conclusion is that, with reference to couples who are functioning at a more distressed level, the effects of the sequentially integrated systemic therapy are more lasting or self-sustaining than those of the emotionally-focused therapy. Limitations In this study, the integration of the structural and systemic therapies took place only at the level of practice. Although two authors of note, Duncan (1984) and Fraser (1982), have suggested models for integration at the level of theory, neither sufficiently articulates such a model. Further development of a theoretically integrated model was not within the scope of the study. Because there was no theoretical integration, the implications for the SIS treatment and the findings of this study may be contested. Possible inconsistencies between interventions in the Sequential^ Integrated Systemic treatment could be seen as a potential problem, particularly if it is believed that theory determines what is done and that divergences in theory could lead to a discontinuity between interventions. An attempt was made, however, to prevent such discontinuity by developing a sequentially integrated systemic model (the SIS therapy) at the level of practice, along the lines of Sluzki (1983) and Papp (1983). Such a model, while rooted in a common paradigmatic frame, allows for a flexibility in therapeutic techniques, drawn from the intermediary models of the structural, strategic and Milan schools and tailored to the requirements of the therapeutic situation. Further speculation 192 about an integration at the theoretical level would seem fruitful for future analysis and discussion. Future Research Some theoretical questions suggested by this study are related to which specific aspects of each treatment lead to change and the processes involved in such change. These issues are complex and are not within the scope of this study. The question of how emotional or systemic processes lead to change should be addressed by research which focuses specifically on the process of change. The discrepancy between EF couples' performance on a measure of conflict resolution and a measure of marital satisfaction raises another interesting theoretical question about the relationship of specific conflict resolution or reduction of fighting behaviours and overall improvement in marital adjustment and whether the one necessarily leads to the other. In this study the length of therapy was ten sessions. This might be extended in order to give distressed couples more time to resolve seriously dysfunctional conflicts. For example, twelve to fourteen sessions might be provided with such couples. This study involved only four month follow-up. Future research could benefit by a longer term follow-up period, i.e. a one year follow-up study. At the same time, it would be desirable to include the control group in both short and longer term follow-ups, in order to provide a clear baseline. Finally, larger samples would allow researchers to detect smaller differences. As an example, in this study there was inconsistency between results when the groups were compared with one another on each occasion (HQ) and when each group was tested over 193 time (HT)- Had the power of the experiment been increased by having a larger sample, it is probable that at least one of the powerful comparisons of the treatment groups would have shown significant differences at follow-up, thereby increasing the consistency between tests in the study. At the level of clinical practice the study raises questions about how to help therapists by clarifying for them points in clients' change processes, and more specifically, articulating the particular appropriate interventions which correspond to these identified points. How best to train therapists to implement emotionally focused interventions or deliver sequentially integrated systemic messages is another question. In practical application the EF and SIS therapies seem to have strong complementarity. A combination of experiential and systemic approaches is an alternative which is rich with possibilities. Couples often have complex problems which are not fully addressed by one therapy to the exclusion of the other. Using SIS and EF interventions in some sequential or integrated fashion would seem optimal in order to deal with the complexity of marital problems. Repetitive fight cycles which have become so entrenched that couples are resistant to an affective, experiential approach are often a major problem. Sequentially Integrated Systemic interventions would be useful for reframing those cycles in order to interrupt and alleviate them. Similarly, an accompanying problem for couples is their inability to access underlying emotions and develop intimacy. The Emotionally Focused approach would provide therapeutic opportunities for partners to experience and express such emotions. One issue which has not been addressed is the impact of the team in the integrated systemic therapy; both at theoretical and practical levels. From a 194 theoretical perspective it could be speculated that couples exposed to SIS therapy derive the benefit of the concentrated, collective efforts of trained professionals, a so-called "think-tank," a group which devotes considerable time and effort toward discussing the couple's relationship and problem solving in order to facilitate change. It could be speculated that, in some hypothetical sense, couples receive more than the designated weekly hour of therapy and this could be a possible influence in making the SIS therapy more lasting or self-sustaining. From a practical perspective, SIS marital therapy is less feasible to implement-from a "person-power" as well as financial standpoint. For this reason, there may be realistic cost-effectiveness considerations in instituting such a program, either in research or in everyday practice. Considering the high cost of SIS therapy, research could be undertaken into more efficient methods of delivery, i.e. reducing the number of therapists involved in actual sessions. Finally, the issue of generalizability must be addressed. The first question addresses the effect of therapist factors on generalizations regarding treatment. In this study therapists were treated as a fixed factor. This means that therapists were not randomly assigned to either of the treatment groups but instead different therapists executed the two treatments. EF therapists had education and specific training in experiential therapies and SIS therapists had education and specific training in systemic therapies. The fact that therapists in each of the treatments were committed to that treatment could be seen as advantageous in that they would demonstrate greater enthusiasm for and knowledge of the treatment. This tends to prevent the bias which is believed to occur when a therapist's theoretical orientation is in conflict with the treatment. However, treating therapists as a fixed factor could also be seen to limit generalizability. When therapists are 195 treated as a fixed factor, treatment effects may be confounded with differential therapist skills. Random assignment of therapists or crossing therapists with treatments, where the same group of therapists implement both treatments, is seen by some researchers as a way of preventing confounds between treatment effects and differential therapist skills (O'Leary & Turkewitz, 1978). If it is true that random assignment ensures that all of the therapist population is equally represented in each treatment group, it can then be assumed that using that method would make differential treatment outcomes more readily attributable to the treatments rather than to the differential characteristics or skills of the therapists. However, O'Leary and Turkewitz (1978) have pointed out that random assignment of therapists can have a serious disadvantage. This would occur if more of the therapists happen to have a theoretical orientation that favours one of the treatments to the exclusion of the other; systematic bias would then be introduced. For example, as applied to the present study, if a preponderance of therapists from a structural/strategic background were to have conducted both treatment programs, a bias would have been introduced b}' greater enthusiasm for the SIS interventions, greater knowledge of the specific SIS procedures and more clinical experience in treating clients using SIS methods. Because it was decided to treat therapists as a fixed factor in this study—different therapists executed the two treatments—measures were taken to try to prevent the possible consequences of confounding of treatment effects with differential therapist skills. One such measure which was suggested by O'Leary and Turkewitz (1978) was to use large numbers of therapists. The use of seven therapists in the SIS treatment group and seven therapists in the EF treatment 196 group exceeded those researchers' recommended minimum of three-four per group. Another measure taken to avoid this confound was the provision of the extensive training programs and close supervision for therapists in each treatment group. Treating therapists as a fixed factor could also limit generalizability in that therapists who prefer SIS therapy may be seen to represent a population with characteristics that are different than the population represented by those who prefer EF therapy. Specifically, bias could occur if something in the SIS therapist's background and training causes him or her to function differently than the EF therapist. While this notion has merit, it could be alternatively concluded that both SIS and EF therapists in this study were representative of the same general population of therapists. This conclusion is based upon the equivalence of education, training and experience that existed between both groups as well as the close matching of the two groups with regard to number of years of general counselling, couples' counselling, and number of hours of specialized training in either an experiential or systemic approach. However, in order to gain more insight and further address questions of generalizability based on therapist representativeness, it is recommended that this study be replicated, using random assignment of therapists or crossing of therapists with treatments. Finally, it is important to consider generalizability with regard to the population of couples participating in this study. As noted, this study was conducted using brief therapy with a population of couples in rather severely distressed relationships and, therefore, inferences about the impact of the treatments and the issue of their lasting effect may be limited to such a population. Continued outcome research with less severely distressed couples, or a longer treatment program with more distressed couples, or a combination of approaches 197 may help to elucidate further the strengths and limitations of these two approaches, as well as their inherent potential for change. 198 References Ackerman, N. (1972). The growing edge of family therapy. In C. Sager & H. S. Kaplan (Eds.), Progress in group and family therapy. New York: Br unner/Mazel. Adams, O. B., & Nagnur, O. N. (1981). Marriage, divorce and mentality. A life table analysis for Canada: 1975-77. Ottawa: Statistics Canada. Alexander, J. F., & Parsons, B. V. (1973). Short term behavioral intervention with delinquent families: Input on family process and recidivism. Journal of Abnormal Psychology, 81, 219-225. Andolfi, M. (1979). Family therapy: An interactional approach. New York: Plenum Press. Andolfi, M. (1980). Prescribing the families own dysfunctional rules as a therapeutic stratagy. Journal of Marital and Family Therapy, 6, 29-36. Azrin, N. H., Besalel, V. A., Bechtel, R., Michalicek, A., Mancera, M., Carroll, D., Shuford, D., & Cox, J. (1980). Comparison of reciprocity and discussion-type counselling for marital problems. American Journal of Family Therapy, 8(4), 21-28. Azrin, N. H., Naster, B. J., & Jones, R. (1973). Reciprocity counseling: A rapid learning based procedure for marital counseling. Behavior Research and Therapy, 11, 365-382. Bateson, G. (1971). A systems approach. International Journal of Psychiatry, 9, 242-244. Bateson, G. (1972). Steps to an ecology of mind. New York: Ballantine Books. Bateson, G. (1979). Mind and nature. New York: Dutton. Battle., C, Imber, S., Hoehn-Saric, R., Stone, A. R., Nash, E. R., & Frank, J. D. (1966). Target complaints as criteria of improvement. American Journal of Psychotherapy, 20, 184-192. Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: International Universities Press. Bergin, A. E., & Lambert, M. J. (1971). The evaluation of therapeutic outcomes. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of psychotherapy and behavior change. New York: John Wiley & Sons. 199 Birchler, G. R. (1979). Communication skills in married couples. In A. S. Bellack & M. Hersen (Eds.), Research and practice in social skills training. New York: Plenum Press. Birchler, G. R., & Webb, L. J. (1977). Discriminating interaction behaviors in happy and unhappy marriages. Journal of Consulting and Clinical Psychology, 45, 494-495. Birchler, G. R., Weiss, R. L., & Vincent, J. P. (1975). Multimethod analysis of social reinforcement exchange between maritally distressed and non-distressed spouse and stranger dyads. Journal of Personality and Social Psychology, 31, 349-360. Bordin, E. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Therapy, Research and Practice, 16(3), 222-229. Bowen, M. (1976). Family therapy and family group therapy. In D. H. L. Olson (Ed.), Treating relationships. Lake Mills, IA: Graphic Press. Bowen, M. (1978). Family therapy in clinical practice. New York: Jason Aronson. Bowlby, J. (1969). Attachment and loss. New York: Basic Books. Buckle}', W. (1967). Sociology and modern systems theory. Englewood Cliffs, NJ: Prentice-Hall. Campbell, S. M. (1980). The couples' journey. U.S.A.: Impact Publishers. Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37-46. Colapinto, J. (1984). On model integration and model integrity. Journal of Strategic and Systemic Therapies, 4(3), 38-42. Coyne, J. (1984). Introducing structural interventions into strategic therapy: A caution. Journal of Strategic and Systemic Therapies, 3(3), 28-32. Cronen, V., Johnson, K., & Lannamann, J. (1982). Paradoxes, double binds, and reflexive loops: An alternative theoretical perspective. Family Process, 1, 91-112. Deutsch, M. (1969). Conflicts, productive and destructive. Journal of Social Issues, 25, 7-43. Dicks, H. V. (1967). Marital tensions. New York: Basic Books, Inc. Dossey, L. (1982). Space, time and medicine. Boulder, CA: Shambala Publications. 200 Duncan, B. L. (1984). Adopting the construct of functionality when it facilitates system change: A method of selective integration. Journal of Strategic and Systemic Therapies, 4(3), 60-65. Duncan, B. L., & Fraser, J. S. (1983). Systems of systems: Integration and disintegration among system views of family therapy. Paper presented at American Association for Marriage and Family Therapy, Washington, D.C. Edmonds, V. H. (1967). Marital conventionalization: Definition and measurement. Journal of Marriage and the Family, 29, 681-688. Efron, D. (Ed.) (1984). Special issue on integration - case studies: Mixing and switching models in clinical practice. Journal of Strategic and Systemic Therapies, 3(4). Ellis, A. (1973). Humanistic psychotherapy: The rational emotive approach. New York: The Julian Press. Feldman, L. B. (1979). Marital conflict and marital intimacy: An integrative psychodynamic behavioral-systemic model. Family Process, 18, 69-78. Feldman, L. B. (1982). Dysfunctional marital conflict: An integrative interpersonal-intrapsychic model. Journal of Marital and Family Therapy, 8, 417-428. Feldman, L. B., & Pinsof, W. M. (1982). Problem maintenance in family systems: An integrative model. Journal of Marital and Family Therapy, 8, 295-308. Fisch, R., Weakland, L., & Segal, L. (1982). Tactics of change. San Francisco: Jossey-Bass. Foley, V. (1974). An introduction to family therapy. New York: Grune & Stratton. Fournier, D. G., Olson, P. H., & Druckman, J. M. (1983). Assessing marital and premarital relationships. In E. E. Filsinger (Ed.), Marriage and family assessment. Beverley Hills, CA: Sage Publishing. Frankl, V. (1960). Paradoxical intention: A logotherapeutic technique. American Journal of Psychotherapy, 14, 520-535. Fraser, J. S. (1982). Structural and strategic family therapy: A basis for marriage, or grounds for divorce? Journal of Marital and Family Therapy, 8(5), 13-22. Fraser, J. S. (1984). Process level integration: Corrective vision for a binocular view. Journal of Strategic and Systemic Therapies, 4(3), 43-57. Friesen, J. (1985). Structural-strategic marriage and family therapy. New York, 201 London: Gartner Press, Inc. Garrigan, J., & Bambrick, A. (1975). Short term marital therapy with emotionally disturbed children. Journal of Marriage and Family Counselling, 1, 379-385. Garrigan, J., & Bambrick, A. (1977). Family therapy for disturbed children: Some experimental results in special education. Journal of Marriage and Family Counselling, 3, 83-93. Garrigan, J., & Bambrick, A. (1979). New findings in research on go-between process. International Journal of Family Therapy, 2(1), 76-85. Gelder, M. G. (1978). Some priorities for research in psychotherapy. Paper presented at the Society for psychotherapy research, Toronto. Glass, E., & Stanley, J. (1970). Statistical methods in education and psychology. Englewood Cliffs, NJ: Prentice-Hall. Glick, B., & Gross, S. J. (1975). Marital interaction and marital conflict: A critical evaluation of current research strategies. Journal of Marriage and the Family, 8, 505-512. Gottman, J., Markman, H., & Notarius, C. (1977). The topography of marital conflict: A sequential analysis of verbal and non-verbal behavior. Journal of Marriage and the Family, 8, 461-477. Greenberg, L. S. (1982). The relationship in Gestalt therapy. In M. Lambert (Ed.), Psychotherapy and patient relationships. Homewood, IL: Dow Jones-Irwin. Greenberg, L. S., & Johnson, S. M. (1984). Emotionally focused couples therapy: An integrated affective systemic approach. In N. S. Jacobson & A. S. Gurman (Eds.), The clinical handbook of marital therapy. New York: Guilford Press. Greenberg, L. S., & Safran, J. D. (1981). Encoding and cognitive therapy: Changing what clients attend to. Psychotherapy Theory Research and Practice, 8, 163-169. Greenberg, L. S., & Safran, J. D. (1984a). Integrating affect and cognition: A perspective on the process of therapeutic change. Cognitive Therapy and Research, 8(6), 559-578. Greenberg, L. S., & Safran, J. D. (1984b). Hot cognitions: Emotion coming in from the cold. Cognitive Therapy and Research, 8(6), 591-598. Guerin, P. J. (1982). The stages of marital conflict. The Family, 10(1), 15-26. Gurman, A. S. (1978). Contemporary marital therapies: A critique and 202 comparative analysis of psychoanalytic, behavioral and systems theory approaches. In T. J. Paolino & B. S. McCrady (Eds.), Marriage and marital therapy. New York: Brunner/Mazel. Gurman, A. S. (1981a). Integrative marital therapy. Toward the development of an interpersonal approach. In S. H. Budman (Ed.), Forms of brief therapy. New York: Guilford Press. Gurman, A. S. (1981b). Creating a therapeutic alliance in marital therapy. The American Journal of Family Therapy, 9(3), 84-87. Gurman, A. S., & Kniskern, D. P. (1977). Enriching research on marital enrichment programs. Journal of Marriage and Family Counselling, 3, 3-11. Gurman, A. S., & Kniskern, D. P. (1978). Research on marital and family therapy: Progress, perspective and prospect. In S. L. Garfield & A. E. Bergin (Eds.), Handbook of psychotherapy and behavior change. New York: John Wiley & Sons. Gurman, A. S., & Kniskern, D. P. (1981). Family therapy outcome research: Knowns and unknowns. In A. S. Gurman & D. P. Kniskern (Eds.), Handbook of family therapy. New York: Brunner/Mazel. Haley, J. (1963). Strategies of psychotherapy. New York: Grune Stratton. Haley, J. (1971). A review of the family therapy field. In J. Haley (Ed.), Changing families. New York: Grune & Stratton. Haley, J. (1973). Uncommon therapy: The psychiatric techniques of Milton H. Erickson. New York: Ballantine Books. Haley, J. (1976). Problem solving therapy. San Francisco: Jossey-Bass. Haley, J. (1980). Leaving home. New York: McGraw Hill. Haley, J. (1984). Ordeal therapy. San Francisco: Jossey-Bass. Hinchcliffe, M. K., Hooper, D., & Roberts, F. J. (1978). The melancholy marriage. New York: Wiley and Sons. Hoffman, L. (1971). Deviation-amplifying processes in natural groups. In J. Haley (Ed.), Changing families. New York: Grune & Stratton. Jackson, D. D. (1965a). The study of the family. Family Process, 4, 1-20. Jackson, D. D. (1965b). Marital quid pro quo: Family rules. Archives of General Psychiatry, 12, 589-594. Jackson, D. D. (1967). The individual and the larger context. Family Process, 6, 139-154. 203 Jackson, D. (1968). Therapy, communication and change. Human Communication, volume 2. Science and Behavior Books, Inc. Jacobson, N. S. (1978a). A review of the research on the effectiveness of marital therapy. In T. J. Paolino & B. S. McGrady (Eds.), Marriage and Marital Therapy. New York: Brunner/Mazel. Jacobson, N. S. (1978b). Specific and nonspecific factors in the effectiveness of a behavioral approach to the treatment of marital discord. Journal of Consulting and Clinical Psychology, 46, 442-452. Jacobson, N. S. (1979). Increasing positive behavior in severely distressed marital relationships: The effects of problem solving training. Behavior Therapy, 10, 311-326. Jacobson, N. S. (1984). A component analysis of behavioral marital therapy: The relative effectiveness of behavior exchange and communication problem solving training. Journal of Consulting and Clinical Psychology, 52(2), 295-305. Jacobson, N. S., Folette, W. C, & Elwood, R. C. (1984). Outcome research on behavioral marital therapy: A methodological and conceptual reappraisal. In K. Hahlweg & N. S. Jacobson (Eds.), Marital interaction: Analysis and modification. New York: Guilford Press. Jacobson, N. S., & Margolin, G. (1979). Marital therapy: Strategies based on social learning and behavior exchange principles. New York: Brunner/Mazel. Jacobson, N. S., & Weiss, R. L. (1978). Behavior marriage therapy III. The contents of Gurman et al. may be hazardous to our health. Family Process, 17, 149-164. Johnson, S. M., & Bolstadt, O. D. (1973). Methodological issues in naturalistic observation: Some problems and solutions for field research. In L. A. Hamerlynck, L. C. Handy & E. J. Mash (Eds.), Behavior change: Methodology, concepts and practice. Champaign, IL: Research Press. Johnson, S. M., & Greenberg, L. S. (1985). Differential effects of experiential and problem-solving interventions in resolving marital conflict. Journal of Consulting and Clinical Psychology, 53(2), 175-184. Keller, J. F. , & Elliott, S. S. (1982). Reframing in marital therapy: From deficit to self-sacrifice as focus. Family Therapy, 9(2), 119-126. Kiresuk, T. J., & Sherman, R. E. (1968). Goal attainment scaling: General method for evaluating comprehensive community mental health programs. Community Mental Health Journal, 38, 443-453. Klein, N. C, Alexander, J. F., & Parsons, B. W. (1977). Impact of family systems intervention on recidivism and sibling delinquency: A model of primary prevention and problem evaluation. Journal of Consulting and 204 Clinical Psychology, 45, 469-474. Klugman, J. (1975a). Towards a unified theory of family therapy. Unpublished manuscript, Yale University. Klugman, J. (1975b). Towards a unified technique of family therapy: A meta-strategy of psychotherapy. Unpublished manuscript, Yale University. Koren, P. E. (1978). Marital conflict interaction, relationships among objective and subjective measures. Unpublished doctoral dissertation, University of Utah. Koren, P. E., Carlton, K., & Shaw, D. (1980). Marital conflict: Relationship among behaviors, outcomes and distress. Journal of Consulting and Clinical Psychology, 48(4), 460-468. Kraemer, S. (1982). A note on Spinoza's contribution to systemic therapy. Family Process, 21, 353-357. L'Abate, L. (1977). Intimacy is sharing hurt feelings. Journal of Marriage and Family Counseling, 3, 13-16. L'Abate, L. (1983). Handbook of marital interventions. New York: Grune & Stratton. Langsley, D., Fairburn, R., & De Young, C. (1968). Adolescence and family crisis. Canadian Psychiatric Association Journal, 13, 125-133. Leventhal, J. (1979). A perceptual motor processing model of emotion. In P. Pliner, K. R. Blankstein & I. M. Spigel (Eds.), Advances in the study of communication and affect, Vol. 5. Perception of emotions in self and others. New York: Plenum Press. Lewis, J. M., Beavers, W. R., Gossett., J. T., & Austin Phillips, V. (1976). No single thread: Psychological health in family systems. New York: Brunner/Mazel. Liberman, R. P., Levine, J., Wheeler, E., Sanders, N., & Wallace, L. J. (1976). Marital therapy in groups: A comparative evaluation of behavioral and interactional formats. Acta Psychiatrica Scandinavica, Supplement 266, 1-34. Liddle, H. (1984). Toward a dialectical-conceptual-coevolutionary translation of structural-strategic family therapy. Journal of Strategic and Systemic Therapies, 4(3), 66-80. Liebman, R., Minuchin, S., Baker, L., & Rosman, B. L. (1976). The role of the family in treatment of chronic asthma. In P. Guerin (Ed.), Family therapy. New York: Gardner Press. Lowenthal, M. F., & Haven, S. W. (1968). Interaction and adaptation: Intimacy as a critical variable. American Sociological Review, 33, 20-30. 205 Lowenthal, M. F., & Weiss, L. (1976). Intimacy and crises in adulthood. The Counselling Psychologist, 6, 88-94. Luborsky, L., Woody, C. E., McLellan, T., O'Brien, C. P., & Rosenwerg, J. (1982). Can independent judges recognize different psychotherapies? Journal of Consulting and Clinical Psychology, 50(1), 49-62. McKinnon, L. (1983). Contrasting strategic and Milan therapies. Family Process, 22, 425-441. McKinnon, L., Parry, A., & Black, M. (1984). Strategies of family therapy: The relationship to styles of family functioning. Journal of Strategic and Systemic Therapies, 4(3), 6-22. Madanes, C. (1981). Strategic family therapy. San Francisco: Jossey-Bass. o Mahoney, M. J. (1980). Psychotherapy and the structure of personal revolutions. In M. J. Mahoney (Ed.), Psychotherapy process: Current issues and future directions. New York: Plenum Press. Margolin, G., & Weiss., R. L. (1978). Comparative evaluation of therapeutic components associated with behavioral marital treatment. Journal of Consulting and Clinical Psychology, 46, 1476-1486. Maruyama, J. (1963). The second cybernetics: Deviation-amplifying mutual causal processes. American Scientist, 51, 161-179. Meichenbaum, D. (1977). Cognitive behavior modification. New York: Plenum Press. Meissner, W. W. (1978). The conceptualization of marriage and family dynamics from a psychoanalytic perspective. In T. J. Paolino & B. S. McCrady (Eds.), Marriage and marital therapy. New York: Brunner/Mazel. Miller, D. (1984). Therapeutic neutrality as a primary distinction between models in system-oriented family therapy. Journal of Strategic and Systemic Therapies, 3(4), 55-62. Miller, R. G. (1966). Simultaneous statistical inference. U.S.A.: McGraw Hill Inc. Minuchin, S. (1974). Families and family therapy. Cambridge, MA: Harvard University Press. Minuchin, S., Baker, L., Rosman, B., Liebman, R., Millman, L., & Todd, T. A. (1975). A conceptual model of psychosomatic illness in children. Archives of General Psychiatry, 32, 1031-1038. Mitchell, K. M. (1973). Effective therapist interpersonal skills: The search goes on. Invited address, Michigan State University. 206 Nadelson, C. C. (1978). Marital therapy from a psychoanalytic perspective. In T. J. Paolino & B. S. McCradj' (Eds.), Marriage and marital therapy. New York: Brunner/Mazel. Olson, D. H. (Ed.) (1970). Marital and family therapy: Integrative review and critique. Journal of Marriage and the Family, 32, 501-538. Olson, D., Russel, C, & Sprenkle, D. (1980). Marital and family therapy: A decade review. Journal of Marriage and the Family, 42(4), 973-992. O'Leary, K. D., & Turkewitz, H. (1978). Methodological errors in marital and child treatment research. Journal of Consulting and Clinical Psychology, 46, 747-758. O'Leary, D. K., & Turkewitz, H. (1981). A comparative outcome study of behavioral marital therapy and communication therap}'. Journal of Marital and Family Therapy, 7, 159-169. Omer, H. (1982). The macrodynamics of Ericksonian therapy. Journal of Strategic and Systemic Therapies, 5, 34-44. Papp, P. (1983). The Process of Change. New York: Guilford Press. Penn, P. (1982). Circular questioning. Family Process, 21, 267-280. Perls, F. (1973). The Gestalt approach and eyewitness to therapy. Ben Lomond, CA: Science and Behavior Books. Perls, F., Hefferline, R., & Goodman, P. (1951). Gestalt therapy. New York: Julian Press. Pinsof, N. (1983). Integrative problem-centered therapy: Toward a synthesis of family and individual psychotherapies. Journal of Marital and Family Therapy, 9, 19-35. Pinsof, W., & Catherall, D. (1983). The couples therapy alliance scale manual. Chicago, IL: The Chicago Center for Family Studies, Northwestern University. Plutchik, R. (1980). Emotion: A psychorevolutionary synthesis. New York: Harper & Row. Prigogene, I., & Stengers, I. (1984). Order out of chaos. New York: Barton Books. Rachman, S. (1981). The primacy of affect: Some theoretical implications. Behavior Research and Therapy, 19, 279-290. Rausch, H. L., Barry, W. A., Hertel, R. K., Swain, M. A. (1974). Communication conflict and marriage. San Francisco: Jossey-Bass. 207 Revenstorf, D., Hahlweg, K., Schindler, L., & Kunert, H. (1984). The use of time series analysis in marriage counseling. In K. Hahlweg & N. S. Jacobson (Eds.), Marital interaction: Analysis and modification. New York: Guilford Press. Rice, L. (1974). The evocative function of the therapist. In D. Wexler & L. Rice (Eds.), Innovations in client centered therapy. New York: John Wiley & Sons. Rice, L. N., & Greenberg, L. S. (1984). Patterns of Change. New York: Guilford Press. Roberts, J. (1984). Switching models: Family and team choice points and reactions as we moved from the Haley strategic model to the Milan model. Journal of Strategic and Systemic Therapies, 3(4), 40-53. Rohrbaugh, M. (1984). The strategic systems therapies: Misgivings about mixing the models. Journal of Strategic and Systemic Therapies, 3(3), 28-32. Rohrbaugh, M., Tennen, H., Press, S., & White, L. (1981). Compliance, defiance, and therapeutic paradox: Guidelines for strategic use of paradoxical interventions. American Journal of Orthopsychiatry, 51(3), 454-467. Rosow, I. (1967). Social integration of the aged. Glencoe, NY: Free Press. Sager, C. J. (1981). Couples therapy and marriage contracts. In A. S. Gurman & D. P. Kniskern (Eds.), Handbook of family therapy. New York: Brunner/Mazel. Satir, V. (1967). Conjoint family therapy. Palo Alto, CA: Science & Behavior Books. Segraves, R. T. (1982). Marital therapy: A combined psychodynamic-behavioral approach. New York: Plenum Press. Selvini-Palazzoli, M., Boscolo, L., Cecchin, G., & Prata, G. (1978). Paradox and counterparadox. New York: Jason Aronson. Selvini-Palazzoli, M., Boscolo, L., Cecchin, G., & Prata, G. (1980). Hypothesizing-circularity-neutrality: Three guidelines for the conductor of the session. Family Process, 19, 3-12. de Shazer, S. (1984). Fit. Journal of Strategic and Systemic Therapies, 3(3), 34-37. Skynner, A. C. (1976). One flesh, separate persons: Principles of family and marital psychotherapy. London: Constable & Co. Sluzki, C. E. (1978). Marital therapy from a systems theory perspective. In T. Paolino & B. McCradj' (Eds.), Marriage and marital therapy. New York: 208 Brunner/Mazel. Sluzki, C. E. (1983). Process, structure and world views: Toward an integrated view of systemic models in family therapy. Family Process, 22, 469-476. Sluzki, C. E., & Ransom, D. C. (Eds.). (1976). Double bind: The foundation of the communicational approach to the family. New York: Grune & Stratton. Smith, M. L., & Glass, G. V. (1977). Meta-analysis of psychotherapy outcome studies. American Psychologist, 32, 752-760. Spanier, G. (1976). Measuring dyadic adjustment. Journal of Marriage and the Family, 38, 15-28. Speer, D. C. (1970). Family systems: Morphostasis and morphogenesis, or "Is homeostasis enough?" Family Process, 9, 259-278. Stanton, M. D. (1981a). Strategic approaches to family therapy. In A. S. Gurman & D. P. Kniskern (Eds.), Handbook of family therapy. New York: Brunner/Mazel. Stanton, M. D. (1981b). An integrated structural/strategic approach to family therapy. Journal of Marital and Family Therapy, 7(4), 427-439. Stanton, M. D. (1981c). Marital therapy from a structural/strategic viewpoint. In G. P. Sholeval (Ed.), Marriage is a family affair: A textbook of marriage and marital therapy. Jamaica, NY: S. P. Medical and Scientific Books. Stanton, M. D., & Todd., T. C. (1979). Structural family therapy with drug addicts. In E. Kaufman & P. Kaufman (Eds.), The family therapy of drug and alcohol abuse. New York: Gardner Press. Stanton, M. D., & Todd., T. C. (1981). The family therapy of drug addiction. New York: Guilford Press. Steinglass, P. (1978). The conceptualization of marriage from a systems theory perspective. In T. J. Paolino & B. S. McCrody (Eds.), Marriage and marital therapy. New York: Brunner/Mazel, pp. 298-367. Strupp, H. H., & Hadley, S. W. (1977). A tripartite model of mental health and therapeutic outcomes. American Psychologist, 32, 187-196. Stuart, R. B. (1969). Operant interpersonal treatment for marital discord. Journal of Consulting and Clinical Psychology, 33, 675-682. Tolstedt, B. E., & Stokes, J. P. (1983). Relation of verbal, affective and physical intimacy to marital satisfaction. Journal of Counselling Psychology, 30, 573-580. Tomm, K. (1982). The Milan approach to family therapy: A tentative report. 209 Unpublished manuscript. Vines, N. (1979). Adult unfolding and marital conflict. Journal of Marital and Family Therapy, 15, 5-14. Wagner, V., Weeks, G., & L'Abate, L. (1980). Enrichment and written messages with couples. American Journal of Family Therapy, 8(3), 36-44. Wallerstein, J. S., & Kelley, J. B. (1975). The effects of parental divorce: Experiences of the preschool child. Journal of the American Academy of Child Psychiatry, 14, 600-616. Waskow, I. E., & Parloff, M. B. (1975). Psychotherapy change measures. Washington, D.C.: National Institute of Mental Health. Watzlawick, P. (1976). The psychotherapeutic techniques of "reframing." In J. L. Claghorn (Ed.), Successful Psychotherapy. New York: Brunner/Mazel. Watzlawick, P., Beavin, J. H., & Jackson, D. D. (1967). Pragmatics of human communication: A study of interactional patterns, pathologies and paradoxes. New York: W. W. Norton. Watzlawick, P., Weakland, J. H., & Fisch, R. (1974). Change: Principles of problem formation and problem resolution. New York: W. W. Norton. Weakland, J. H. (1976). Communication theory and clinical change. In P. Guerin (Ed.), Family therapy: Theory and practice. New York: Gardner Press. Weakland, J. H., Fisch, R., Watzlawick, P., & Bodin, A. (1974). Brief therapy: Focused problem resolution. Family Process, 13(2), 141-168. Weeks, G., & L'Abate, L. (1982). Paradoxical psychotherapy: Theory and practice with individuals, groups and families. New York: Brunner/Mazel. Weiss, R. L. (1978). The conceptualization of marriage from a behavioral perspective. In T. J. Paolino & B. S. McCrady (Eds.), Marriage and marital therapy. New York: Bruner/Mazel. Weiss, R. L. (1980). Strategic behavioral marital therapy. In J. P. Vincent (Ed.), Advances in family intervention: Assessment and theory. Greenwich, CT: JAI Press. Weiss, R. L., Hops, H., & Patterson, G. R. (1973). A framework for conceptualizing marital conflict: A technology for altering it. Some data for evaluating it. In L. A. Hammerlynck, L. C. Hardy & E. J. Mash (Eds.), Behavior change: Methodology, concepts and practice. The 4th Banff Conference on behavior modification. Champaign, IL: Research Press. Wells, R. A., & Dezin, A. E. (1978). The results of family therapy revisited: 210 The nonbehavioral methods. Family Process, 2 7(3), 251-274. Wells, R. A., Dilkes, T., & Trivelli, N. (1972). The results of family therapy: A critical review of the literature. Family Process, 7, 189-207. Wile, D. E. (1981). Couples therapy: A non-traditional approach. New York: John Wiley & Sons. Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35, 151-175. Zuk, G. H. (1966). The go-between process in family therapy. Family Process, 5, 162-178. 211 Appendix A: Sequentially Integrated Systemic Couples' Therapy Manual 212 THERAPY MANUAL: SEQUENTIALLY INTEGRATED SYSTEMIC COUPLES' THERAPY FORMAT: THERAPY STEPS THERAPIST ACTIVITIES DEFINITION OF TERMS © Greenberg and Goldman INTEGRATED SYSTEMIC TREATMENT FRAMEWORK L. Greenberg and A. Goldman I. Therapy Steps Step 1. Defining the Issue Presented Therapist Activities Step 2. Identifying the Negative Interactional Cycle Homework Therapist Activities Step 3. Restructuring Homework Therapist Activities Step 4. Reframing the Problem a. Positive Connotation b. Prescribing the Symptom Therapist Activities Step 5. Restraining a. Go Slow Homework Therapist Activities b. Dangers of Improvement Homework Therapist Activities 214 Step 6. Consolidating the Frame Step 7. Prescribing a Relapse (Note: Therapist Activities and Homework are not treated separately under the last two steps but incorporated within them.) II. Definition of Terms Reframing Restraining Tactics (basis for) Positioning Compliance and Defiance-Based Strategies Note: the therapy is conducted with the help of a Team of trained therapists who remain behind the mirror during the sessions. 215 THERAPIST'S MANUAL, SEQUENTIALLY INTEGRATED SYSTEMIC COUPLE'S THERAPY The Interventions in this treatment tend to occur in a circular fashion within sessions rather than a linear sequence across sessions. Therefore, this manual focuses upon the steps of the process rather than attempt a session by session account. The steps in the process, key interventions, examples of these, general therapist activities accompanying each step and homework assignments follow below. Step 1. Defining the Issue Presented The purpose of this step is to establish each partner's view of the problem and how partners perceive their own and their spouse's role in it. Partners are encouraged to make a full statement of their positions and how they perceive their own and their partner's roles in the problem. Partners are asked to state their goals for the therapy. Part of the value of this is that it usually points to the couple's central issue. If clients have difficulty setting goals, ask what the minimal change would be. Therapist Activities Establishing an alliance or "joining." Direct questions and probes. Summarizing and integrating information. The following interventions are helpful in the alliance building process as well as in establishing goals and a focus for the therapy: 216 1. Assess each partner's position and form an alliance with each by affirming their position. 2. Join with each partner by talking in their language, recognizing their concerns, etc. The following interventions are helpful in establishing communicational premises in order to neutralize or make evident certain rules of the couple (Sluzki, 1982): 1. Each partner should be asked/told to speak in the first person singular, "I" instead of "we." This is an attempt to discourage "mind reading: and/or "collective agreement" between partners, either or both of which usually become apparent in the first interviews. By saying "speak for yourself the roles and rules of the couple and their way of negotiating with the therapist—seeing themselves and each other—are challenged. (This intervention is seen as a "testing ground" for the couple's way of negotiating with the therapist about the premises of therapy, as well as their ability to make shifts and changes). Example: Client: One of the things that always makes us angry is when her mother starts to interfere in our business. Therapist: I'd like you to tell me just from your own point of view right now-whenever your wife' s mother gets involved in your affairs, you get angry? Client: But we both get angry. Therapist: I understand, but just now I'd like you each to be the expert on yourself and just speak about it for yourself and let B speak for herself. You were saying, when your mother-in-law 217 . . . ? 2. If partners make impersonal statements about personal matters (value judgements, opinions), then tell them to make personal statements. Example: Client: Husbands who are supportive don't try to question and criticize their wives when dinner isn't ready on time. Therapist: So you get upset when Frank complains that dinner isn't ready on time? 3. If one partner engages in "mind reading" or makes an "unsubstantiated reference" to the other's subjective state, then ask the other partner what it is that he/she perceives, thereby differentiating perceptions from inferences. Example: Client: Sarah thinks I'm an erratic driver and always clams up and scowls at me when we're in heavy traffic. Therapist: I understand that Sarah seems nervous when you're driving in heavy traffic—and you believe that she thinks you're an erratic driver. Can you check that out with Sarah? (or) Sarah, how do you perceive Bill's driving? Step 2. Identifying the Negative Interactional Cycle The purpose of this step is to identify sequences of problematic reactions as the couples describe or enact them. Couples are asked to describe the problem and the negative interactional cycles are identified by the therapist. These cycles may occur spontaneously in the session or the couple may be asked to engage in an enactment of the problem. As they occur, patterns in the cycles become apparent. 218 For example, the cycle with one couple was identified at this stage as: When A begins to push and take charge, B responds by asserting her rights and accusing A. A reacts by being more insistent and controlling while B continues to assert and blame. A finally gives up and withdraws and then B withdraws as well. In such cycles, each of the partner's solutions to the problem intensifies the problem. Homework Instruct partners to observe or deliberately have an argument like this at least once during the week using the rationale that it would be helpful to be able to identify what happens in their arguments and to report this back the following session. Therapist Activities Assessing the components of the cycle through questioning, exploring, and clarifying each partner's reactions to the other. To clarify the cycle and positions, the therapist might say, "What did you do then?" and then to the other partner, "When your partner does this, what do you do?" Mind reading, collective agreement, making impersonal statements about personal matters are approached as in Step 1: A set of specific questions that are helpful in gathering information about the problem are given below. These are divided into questions about: 1. The problem 2. Attempted solution 219 3. Goals The Problem a. What is the problem that brought you here? b. How is it a problem for you? c. How has this problem interfered with your daily life? d. What consequences has it had for you? e. Can you give me an example? f. Is there some particular problem that stands out most? g- What is the most important difficulty for you at this particular time? h. What made you decide to come to therapy at this particular time? i. Why was this step taken now? Attempted Solutions a. How have you been attempting to handle or resolve this problem? b. Everybodj' tries as best they can to deal with their problems: I would like to know what things you have been trying to solve the problem even though they may not have worked as well as you wished they would? c. Are there other people who have helped you to deal with this problem? Goals a. At a minimum, what would you hope to see happen or be different, as a result of coming to therapy? b. What at the very least, would you like to see happen as a result of therapy? 220 c. What would be a significant indicator of change or proof that something has happened? d. What would be a sign of a positive step? e. I am going to ask you the same question. What for you would be an indication that you had taken a positive step? f. Let us assume that you have resolved or achieved this step. What new problem might possibly appear, hypothetically? g. Who else will be affected if you changed? h. What else might change? Step 3. Restructuring The purpose of this step is to try to intervene directly in order to get the couple to change their interactional cycle before attempting strategic or indirect interventions. If the direct restructuring interventions are successful, the therapy continues along these lines. The couple is first asked to enact a problem and to attempt to come to a mutually satisfactory resolution. This enactment generally reveals the dysfunctional positions and patterns of communication that constitute the negative interactional cycle. As one partner begins to speak to the other, the therapist intervenes to block any of the communication that will lead to the old negative cycle and supports those that are in line with the needed pattern of communication. The restructuring intervention involves supporting each member of the couple in the same way and then directly suggesting that they change the manner in which they talk to each other in order to resolve the negative cycle. For example, the therapist could say to the pursuer, "I see that you are very good at expressing 221 your side and this has been helpful to you but if you are to resolve this conflict it will be necessary to change the way that you talk to your partner so that she will hear you. I want you to talk to her in a way that she will be able to hear." In general, it is important to ask the partner who is the pursuer or the more active person in the interaction to speak in a different way to the other. The different responses that are needed are guided by the assessment of the dysfunctional pattern. This may be to move out of a dominant position or to move closer by being more self disclosing. Example: _T: (to blamer) John, if you want to get your wife to listen to you, you need to talk to her in a way that she can understand or, John, how could you talk to your wife so that she can understand what you are saying? John: Wives who are critical and nagging are impossible to live with. TT: So you get upset when Sharon criticizes you . . . John: She always looks at me so contemptuously-as if she disapproves of everything I do-(to Sharon) I can't believe you do that when all I wanted in the first place was your respect! (Wife responds in angry fashion and cycle begins.) jT: John, I understand you feel criticized by your wife, but how could you speak to Sharon so that she could hear what you are really saying? If the husband is able to speak to her in a positive fashion, telling her what is desired without "putting her down" and she responds in an angry fashion, then the therapist intervenes with her in a similar fashion to restructure the 222 interaction: T: Sharon, I understand that you feel angry with John when you feel blamed but I wonder if you can respond to what John has asked. Homework If the couple is able to change the negative interactional cycle in the session, following this intervention, then they are instructed to talk to each other in this new fashion at home at least two or three times during the week. The specific manner in which each person should speak is clearly identified. If they are not able to change, the therapist might repeat the following: "I understand that the fight has been useful to you but if you are to resolve it, it will be necessary to change the way that you talk to each other"; adding, "When this occurs at home, I'd like you both to think about how you could speak to each other and respond to each other so that each of you could hear and understand what the other is saying. John, I understand that you feel criticized by Sharon, but how could you speak to her so that she could hear you? . . . and Sharon, I would like you to listen to what he says and respond in such a way that you two could resolve this issue." Therapist Activities 1. Clarifying 2. Questioning 223 3. Suggesting enactments 4. Giving directives designed to make interactions more positive. 5. Applying communication skills principles a. getting partners to phrase communication as requests rather than commands b. getting partners to state what specific positive action is desired rather than what is not desired c. getting partners to refer to the future rather than the past-past actions can't be reversed *Note: If the couple returns, having successfully completed the homework assignment and reversed the negative interactional cycle, then direct interventions are continued and the therapy is moving toward termination. They are asked to communicate in the session in the new way, this is reinforced and the homework assignment is repeated. Step 4. Reframing the Problem If the couple continues to be stuck in negative interactional patterns, the therapy takes a U turn and continues along more strategic lines in which indirect attempts at change are implemented. This involves reframing the problem or "negative cycle" as positive. The overall purpose of reframing is to reframe or restate a situation so that it is perceived in a new way, the meaning attributed to the situation being changed. The reframe used here, at the same time as attributing a new meaning to the negative cycle, suggests "no change." 224 Initially, when clients report that they have been unsuccessful in their homework attempts at improved communication and conflict resolution, the therapist makes a U turn by commenting, "We must have been mistaken, obviously the 'fight' is serving an important function and it would be premature to change it too quickly." This response is based on the assumption that pathological systems adhere to solutions which maintain the system, that symptoms are highly adaptive for the couple in helping to maintain a steady state in their system and that it is important for the therapist to align himself/herself with the couple. The reframe is achieved in two steps: first, positive connotation and second, prescription of the cycle. Positive Connotation The therapist begins the reframing step by positively connoting the couple's negative cycle. The goal in doing this is to address in a respectful way the fears the couple has about change as well as recognizing and acknowledging the functional aspects of the symptomatic cycle, partly as a step toward defining the relationship between the partners in a positive way. Positive connotation is an important step in providing a rationale for the paradoxical directive of "no change" which is to follow. Example: With a controlling/blaming interaction, the therapist would start by complimenting the couple and linking the problematic interaction to the survival of the couple system: _T: I think you're both very perceptive and sensitive to each other and respond to each other in ways that maintain the relationship as it is. . . . You two are so closely connected, so 225 sensitive to each other and have worked out a balance that's just right for you. Even though it may be painful or unpleasant at times, it still seems important in keeping you two in balance ... I see that this pattern of disagreement is part of the struggle between you to maintain the relationship and to define yourselves and this fight has been very important . . . The therapist might use phrases such as: "I see how important you are to each other, otherwise you wouldn't be so concerned." Or, "You two are very deeply involved; only with such involvement would people care so much to fight about it." In an intimacy struggle, angry feelings or interactions are redefined as 'caring.' The intensity of the fight can be defined as 'passion,' 'deep involvement,' 'mysterious and powerful connection.' In a competitive or dominance struggle, the anger is defined as "assertion of rights," "need for self definition," "being a complete person," etc. While the whole system should be included in the positive connotation, it is often helpful to direct the connotation to the partner who is the more active in the cycle, who is considered by the couple to be "causal" in the cycle, i.e. blamer, victimizer, etc. The goal is to define this person as having control over the interactional symptom at a certain level and being motivated by good intentions such as love and protection, those that have positive value and are helpful to the other member. However, a certain shared responsibility of A and B over the cycle is always implied. With a couple in which the "blamer" complained about what 226 she perceived as her husband's criticism (his part in the cycle was seen as "controlling"), the therapist might say: "As I've said, I think you're both very perceptive and sensitive and that is a measure of the caring between you. (To the blamer, A) A, you're particularly sensitive to how B responds to you-and that sensitivity is very important and helpful to B-because it's your way of letting B know how much you care about him and B, your attentiveness and forcefulness is part of how you show your caring and express your closeness. If you didn't care, you wouldn't invest so much energy in telling her what you think about her." With regard to a positive connotation of a withdrawal, the therapist might say, "When you both withdraw, it demonstrates your sensitivity to each other and is helpful in that you each protect each other from further pain that might occur if you were to continue to confront each other." Thus, the couple's relationship is positively connoted-defined in a positive way by the therapist's describing them as caring and sensitive to each other. In general, blaming can be redefined as 'sensitivity' and control as 'responsiveness.' Anger is positively connoted as 'forcefulness' and as 'motivated by caring and a desire for closeness.' Withdrawal is connoted as 'being helpful' and 'having a positive effect.' Some other examples of positive labels that can be used are: passive — the ability to accept things as they are submissive — seeking authority and direction to find oneself insensitive — protecting oneself from hurt seductive ~ wanting to attract other people and be liked oversensitive ~ tuned in to other people, very alive and aware 227 controlling -- structuring one's environment oppositional - searching for one's own way of doing things crying -- ability to express emotion, especially hurt. In the positive connotation, the therapist attempts to capture the unique dynamic which is operating in the relationship to stabilize it and maintain it at its present "safe" level. An attempt is made to identify the positive aspect of each person's idiosyncratic role in maintaining the system as it is. Prescribing the Symptom As noted, positive connotation is an important step in validating the partners and their relationship and acknowledging the positive function of the fight. It is also crucial in providing a rationale for the paradoxical directive, of continuation of the cycle, or "no change," which is to follow. After positively connoting or reframing the negative cycle, the therapist prescribes or tells the couple to continue the fight. Here, the therapist reiterates the important function of the fight and tells the couple to do this at home. The therapist, for example, might say to a couple in a predominantly blaming/withdrawing interaction, "A, I think it's important that you continue to show your caring for B by going after him, badgering him occasionally so that he feels loved. B, you should continue to accept things as they are by being silent and occasionally, when you get lonely, respond to A by doing things that you know will give her cause to make contact with you such as leaving things lying around. It is also important to, at times, disagree with her to help her have a chance to assert herself in response." 228 When working with an explicit symptom such as a phobia or compulsion, prescribing the symptom to the symptomatic member in the presence of a spouse is designed to change the pattern that perpetuates the symptom in two ways: (1) When the client is told to "fake the symptom, fake it well," it is implied to the partner that the symptomatic behaviour to which he/she may be exposed may be false, therefore inhibiting "spontaneous" responses that in turn may reinforce and perpetuate the symptom. (2) It subtly increases the consensus about the client's control over the symptom and decreases the chances of their claiming spontaneitj7. It induces the notion that if a person can produce a symptom, he/she may also be able to reduce it. Therapist Activities 1. Going "one down" by admitting having been mistaken about the importance of the fight. 2. Adopting a position of "no change." 3. Acknowledging the importance of the fight. 4. 'Capturing' any positive aspects of the couple's negative cycle and positively connoting. 5. Assessing the function that the negative interactional cycle (fight) is serving in the couple's relationship. 6. Prescribing the interaction in terms of its functions. 7. Using the hypothesis about the function of the fight in framing a rationale about the fight. 229 8. Linking the problem to the couple and their relationship, i.e. "I see this as part of the struggle between the two of you, to define yourselves, which is an important struggle in any relationship. 9. Linking the problem to the couple system but directing the message (positive connotation) through the partner who is the most symptomatic or amenable, i.e. blamer, by complimenting that partner. 10. As Sluzki (1978) has suggested, if A is defined as "blamer," "victimizer" or IP, etc., it is helpful to reduce physical distance with A or mirror A's body position. 11. If A and B concur in defining A as "victim," for example, and B as "victimizer," then find a way of reversing the roles/labels and state the reversal forcefully. 12. If A or B expresses or attributes to the other feelings that have negative connotations in our culture (i.e. to have those feelings is to be mad, bad, sick . . . ), then relabel or reframe that feeling using a positive connotation. Step 5. Restraining The purpose of the restraining step is for the therapist, by discouraging or even denying that change is possible, to continue to align him/herself with a position of "no change." Restraining is based on the notion that it is risky for 230 the therapist to be explicitly aligned with system tendencies toward change and, rather, needs to be seen by the client as accepting the status quo. Partners in a system habitually tend toward stabilizing the system, following patterns already familiar to them, thus strengthening the status quo and the symptoms. The strategy of restraint portrays the therapist as uncommitted to changing the clients, particularly to having them change quickly. This can have the effect of freeing clients to be more receptive to the therapist and cooperative with the suggestions that the therapist gives. And by removing the sense of urgency to change, clients are helped to relax their problem maintaining efforts. Rohrbaugh et al. (1981) distinguish between "soft" restraining: "you probably shouldn't change" exemplified by interventions such as "go slow" and "dangers of improvement," and "hard" restraining: "you probably can't change," where the therapist mobilizes change by benevolently suggesting that change may not be feasible. With "hard restraining" the therapist may even suggest that further struggles to overcome the problems will only frustrate the client and make matters worse, and that the most reasonable goal of therapy at this point is for the client to learn to accept and live with the problem which, after all, does serve an important function. The restraining step (soft restraining) is applicable when there is a slight shift to improvement. The clients, after being given a specific task, have returned, reporting definite and welcome improvement. (Although this step appears at this point in our sequence, restraining could also be used early in the sessions with clients whose main attempted solution is "trying too hard" or who press the therapist for action while they remain passive or uncooperative.) 231 Go Slow This intervention embodies the message "you probably shouldn't change." When the clients show signs of making some moves toward resolution, they are instructed not to do anything further-particularly nothing specific. Instructions are general and vague and most of this intervention is comprised of offering believable rationales for "going slow." For example: _T: "This week, it would be important not to do anything too quickly to bring about any further resolution." Examples of generalized rationales for "go slow" would be: 1. "Change, even for the better, requires major adjustment." 2. "One needs to determine a step at a time how much change would be optimal--rather than too much." 3. "You might be better off with 50% improvement rather than 75%." Examples of Therapist Response to Clients who Report Improvement in Response to  a Specific Intervention: "It's good to hear that you've done some things this week and that you feel more accepting and less angry toward each other, but our basic feeling is that you're moving too fast, and the important thing is you need to slow down and this is the big trap for someone with your kind of problem—going too fast . . . (goes on, elaborating on this theme). (Clients protest) 232 T: The team just phoned--our supervisor said if it's absolutely essential that you do some of those things, you'll just have to, but to do them to an absolute minimum . . . etc." Homework Clients are not instructed to do anything in particular. Instructions here involve restatement of the restraining tactics to "go slow, take it easy as change that is too fast is dangerous~not lasting." Therapist Activities 1. Avoids showing optimism or encouragement about the possibility of change. 2. Acknowledges the good news about change but seems worried and explains that, welcome as the change may be, it's too fast and is making the therapist uneasy. 3. Encourages the couple to hold back on further improvement, at least until the next appointment. Dangers of Improvement This can be seen as part of the Restraint Step and an extension of the "go slow," but also carries the message "you probably shouldn't change." It is useful when clients, even though showing some tendency toward change are not able to get out of their vicious cycles. The therapist mobilizes change by benevolently suggesting change is something to be feared. Eliciting disadvantages helps the partners to see that improvement is not "ideal," thereby feeling less compelled to put pressure on themselves to change and this may help them relax. Partners have then altered their attempted solution of trying too hard and there would likely 233 be a lessening of the problem as a result. As with "go slow," when the partners show some slight shift toward improvement, "restrain"~this time by asking if they are aware of what the dangers are that are inherent in resolving their problem. It is important to phrase the question in such a way as to imply that there are dangers rather than asking if there are any dangers, which gives them an "out" or opportunity to decline. Alternatively, if clients respond that they realty want to resolve their problems and that there could not possibty be any dangers, the therapist would then generate some drawbacks that exist for each of the partners. It is important to generate at least one credible drawback for each in order to legitimize the position that there are, not just could be, dangers to improvement. Ity suggesting dangers of improvement, the therapist is able to: 1. Legitimize partner's blocks to change. 2. Increase motivation, i.e., "If you stopped being angry and critical with your spouse, he might feel neglected and lonely, or if you do not withdraw you might get into even more arguments." 3. Use this tactic to reframe partner's non-compliance with assignments. For example, T\ I understand that you must have had good reasons for not doing the homework last week-I think in a way there's an important message in this, that maybe at some level you're not aware of. Perhaps—let me ask—can you see any dangers in improvement?" (By linking the question about the dangers 234 of improvement to the clients' failure to do the homework, the therapist is suggesting limits to his/her responsibility for the couple changing and placing more responsibility on the couple.) The following is an example of a therapist's response when a couple is showing some signs of improvement but has not done the homework tasks or when the therapist believes that the improvement may not last and anticipates further problems: "Although I'm aware of your enthusiastic participation in this program and your outward desire to change, I was also aware my supervisor pointed this out to me-of the skillful way that you've challenged this program and~we wondered if there is something about resolving your issue, at a level you're not aware of—that might present a problem—uh, that perhaps there's some unknown danger to you in resolving your conflict. The fact is from what you've both told me, that you've tried in many ways to solve this problem—(perhaps lists ways or alludes to previous therapy) we've all agreed-there are just too many times, you need to think deeply about this. What, by resolving this problem, would be likely to happen-what would be the impact? It would be best not to limit your thinking by being too concrete-or practical-perhaps you could just go with your imagination—so we don't eliminate any possibilities." Homework The homework is similar to that for "Go slow" in that the restraining tactic of "go slow" and listing the dangers of improvement b}' both clients and therapist are invoked at the end of the session. 235 When the issue is one of partners not completing previous homework tasks, the therapist would repeat the message, "As I said, I understand that you must have had good reasons for not doing the assignment last week-I think in a way, there's some underlying meaning in this." And then review the dangers, "And as we discussed, change might be dangerous to your relationship in the following ways . . . " (list ways). Therapist Activities: (The steps for "go slow" are applicable). In addition: 1. Positively connotes clients' blocks to change. 2. Frames the question about "dangers" using "What would be . . ti rather can "can either of you think of any dangers . 3. In order to encourage them to comply with further assignments, reframes lack of compliance regarding homework by linking it to "dangers of improvement. it 4. Benignly presents an attitude of resignation. 5. Challenges client's belief that problem symptom is directly changeable and thereby sets stage for client to prove the therapist wrong by changing. Step 6. Consolidating the Frame By suggesting a reframe, positively connoting, prescribing the symptom and using restraining tactics, the therapist adds something novel to the couple's 236 situation, capturing an unacknowledged level of reality in the system in such a way that the clients are intrigued. Through this process of reframing and restraining, the therapist has portrayed himself/herself as uncommitted to change and implicitly accepted the couple's immediate reality. The couple's resistance has begun to diminish, and the therapist has begun to form an alliance with them. However, while introducing a reframe or presenting it on one occasion sets the above processes in motion, it does not seem sufficient to create lasting change, particularly with more distressed couples, where the negative cycle has been highly adaptive over time and become a pervasive pattern. In order to make a more lasting impact, the frame needs to be consolidated. To do this, the therapist must reinforce and generalize it. Much like a lawyer building a case, the reframes must be restated, and in all following interventions the "supporting evidence" must be related to the basic reframe and position presented by the therapist. Repeating and linking the reframe to various couple interactions and situations will serve to "anchor the frame"; i.e., hold it in place, secure it as a ship is secured to the ocean floor. Anchoring the frame lends further credence to the therapist's position of aligning with the couple and further supports the notion that the negative cycle has an important function in the couple's system. Step 7. Prescribing a Relapse Prescribing a relapse is applicable toward the end of therapy if improvement has occurred. As with Restraining, the purpose of this step is for the therapist to align himself/herself more with a position of "no change" or, at the very most, a position of being uncommitted to changing clients, particularly to having them 237 change quickly. Again, it is seen as risky for therapists to be seen as explicitly aligned with system tendencies toward change and instead, more effective for them to be seen as supportive of stability and no change. Prescribing a relapse also serves the purpose of suggesting to clients that they have control over symptoms—if they can produce them, they may also reduce them. The therapist is placing the client in a therapeutic bind: If the symptomatic cycle does occur again, the therapist predicted it, so it is under his/her control. If it does not occur, it is under the client's control. Prescribing a relapse can be seen as an extension of Restraining and as part of the therapist task of "Consolidating the Frame." When either or both members report a decrease in the conflict or intensity of symptoms, the therapist expresses surprise as well as vague concern and recommends a slight relapse. In prescribing the relapse, it is preferable to have the negative cycle enacted in the session and then prescribe it for the future as something to be engaged in when useful. The intent here is to set a frame far beyond therapy so that the fighting that will occur will be covered by the reframe. The enactment can be very powerful for it really does put people in the position of being in control of this "out of control" cycle. A frame for the relapse is needed, either "you have needed this in the past" or "you will need it in the future" or paying homage or respect to an important aspect of their relationship. These are seen as ways that they have learned that do work for them, and they are advised not to throw them all out the window, as they may occasionally be useful. Example: 238 Therapist: I know that you're both saying there's been a change--and in one way I suppose that's good-but I'm always worried when things happen too quickly--and I wonder--what effect will this have?-I mean, I do believe that this fight between you serves as a function and now, if that isn't occurring, what could happen?—what might the disadvantages be?-As we said, A, if you stopped criticizing B, he might feel neglected and lonely or he might not withdraw as he usually does and you two might get into even more or worse arguments. . . . In any event, I would like to suggest that next week you go back to your old style and have a fight-at least twice. B, perhaps you could help A to get angry and critical again-Can you think of any way you can do that?—I'm really worried that if you're not getting angry and criticizing B, A, then he'll take over and do it instead. Client A: But we don't want to fight any more-that's why we came here-why should we go back? Client B: Yes-why would we want to fight again? Therapist: Yes, I understand that this must seem confusing—but I really must caution you. As I said, the fight serves an important purpose. Without it . . . (gives disadvantages) and I do think it's really important for both of you to do this. 239 At this point, the therapist could get the couple to enact their cycle in the session if at all possible and then prescribe it as homework.) For example, Therapist: A, can you think of something you like to criticize B about? Can you tell him about that right now? Definition of Terms  Reframing Reframing has been defined earlier as restating a situation or problem so that it is perceived in a new way such that it makes new behaviours possible; the meaning attributed to the situation is changed. Watzlawick, Weakland and Fisch (1974) have defined reframing: "to change the conceptual and/or emotional setting in relation to which a situation is experienced and place it into another frame that fits the 'facts' of the same concrete situation equally well or better and thereby changes its entire meaning." Watzlawick et al. note that there are two separate orders of reality: (1) first order reality deals with physical properties of objects and our perception, (2) second-order reality, based on the attribution of meaning and value to these objects. The overwhelming majority of human problems involve only the second-order reality. Within this realm, there are no objective criteria as to what is real-the meaning or value attributed to an objective situation or even the nature of a relationship is not objective. Effective reframing consists of a successful change of this second-order frame of reference and is based on the idea that there is not some 'true' underlying problem but that the problem lies in how people view things. Successful reframing must be communicated to clients in a language that is congenial and, therefore, acceptable to clients' ways of conceptualizing their worlds, 240 their second-order realities, in order that they can accept an interpretation of "reality" that is different from their own. The therapist must learn to communicate with clients in their own language. The ability to adopt clients' views of reality is critical. In addition to using the client's own language, the use of paradox is another way of making reframing acceptable or compelling to the client; i.e., telling a client to do more of the same: "I want you to complain to your wife for at least one hour each day next week." Reframing and relabelling are, then, important catalysts for change. "Reframing" is a broader term which refers to a change in meaning attributed to a problem or situation, while "relabelling" refers to changing the label attached to a person or problem by changing the frame of reference and is subsumed under reframing. A positive label given to some disturbing behaviour gives individuals a sense of having permission to have that behaviour and expectation of positive change. A sense of support imparted by the therapist in such an intervention is an important component for creating this expectational set. Through Reframing, we have noted, the meaning attributed to the situation is changed. In a couple's therapy context, the frame of reference is often changed in the sense of moving from the individual to the dyadic or systemic level. (The implication is that both members have joint responsibility for the problem and its solution.) Reframing and relabelling are contingent upon the therapist having an accurate perception of the client's internal frame of reference, with emotions and meanings that are involved within that frame. Reframing and relabelling are powerful therapeutic tools which can build upon this internal frame of reference and yet surpass it by serving as catalytic agents for change. In other words, the 241 therapist necessarily empathizes or understands the internal frame, the feelings and cognitions surrounding these, but intentionally reframes or relabels, attaching different labels or changes the conceptual or emotional viewpoint in relation to which a situation is experienced in order to generate change in the clients, making more response alternatives open to them. Restraining Tactics (basis for) When a living system deviates toward either the polarity of change or stability, processes are activated that pull the system in the direction of the opposite tendency, thereby keeping an equilibrium. Total stability leads to rigidity and total change leads to dissolution and both lead to death of the system. The balance between both tendencies is crucial both for stability and change within the system. If the therapist wants to favour change in a rigid system, he/she treads that fine line. That can be done by defending stability while freeing the system from the restrictive rules that prevent members from working out changes. Positioning In positioning, the therapist attempts to shift a problematic "position"~usually an assertion that one or the other partner is making about himself/herself and his/her problem-by accepting and exaggerating that position. Positioning manoeuvres are defiance-based interventions through which the therapist preempts or outdoes the clients' or couple's problematic assertions about themselves/their problems. One use of this intervention in this therapy is when one person's position is assessed to be maintained by a complementary or opposite position by the person's spouse. For example, when a client's pessimism is reinforced by an optimistic or encouraging response from her mate, the therapist may "outdo" the client's pessimism by defining the situation as even more dismal 242 than the client had originally held it to be. Positioning is important in working with reactive or resistant clients. A therapist can discourage resistance by responding to clients in ways that are not in opposition to their position. If, for example, one or both partners express pessimism about the treatment, the therapist must position himself/herself in such a way as to accept and exaggerate this. Example: Client: My wife and I have been to several different counsellors. Nothing has helped. They all told us that we were incredibly "stuck." We agree and I'm sure there's nothing you could say or do. Our problem is impossible. It involves something that happened eight years ago, and nothing will ever change! Therapist: I would tend to agree with you-sounds pretty bad~you've tried many things—undoubtedly worked with the best people. I understand how discouraged you must feel. You've tried so much and nothing has worked. It would be foolish to be optimistic. I think skepticism about the program is much healthier and I'm pleased that you're expressing that skepticism to me. I would encourage you to do that whenever you're feeling that way. After all, whatever happened is so long ago and so serious that it's probably difficult to imagine changing that ..." 243 In this way, then, the therapist avoids creating unnecessary client resistance by: 1. Accepting the clients' statements. 2. Recognizing the values they represent. 3. Not making inflammatory or non-credible statements. In the previous example, the therapist has validated the client(s) by implicitly supporting their belief that their problems are complex and not amenable to an easy solution. In order to ensure cooperation regarding interventions, they should be presented to clients in a way that is consistent with the position clients are conveying. For instance, if a wife has decided that there is some behaviour in her husband that is causing the problem, the therapist should avoid presenting a point of view contrary to the client's, even if the therapist does not agree. Instead the therapist should align himself/herself with the client's position and work within the client's frame of reference, taking this into account when framing an intervention. Example: Partner A, Susan: Frank doesn't regard my needs as important and doesn't care about me. He's very cold and uncaring. One of these days I'm going to walk out. I try to talk to him about our problem, suggesting a walk together, and he says he doesn't have anything to say and retires to the den to watch sports on the 'boob tube.' Actually his father is much the same way; he always falls asleep in the corner after dinner and never 244 shows Frank's mother any affection. It runs in the family. Frank: Mmmm, yes, my father is that way, it is difficult for both of us to communicate and I don't know what I can do about it. I guess it's ingrained and I find it impossible to show affection. (The therapist has observed that, although Susan tries to communicate her needs to her husband, her messages are often unclear, abstract, and vague. The therapist has also noted that Susan often criticizes Frank and that the couple's negative cycle might change if Susan were to avoid criticizing and be more explicit in what she wanted. However, the therapist is also aware that both partners, especially Susan, attribute Frank's difficulty behaviour to a personality deficit and that Susan has defined herself as a victim. In order to ensure the couple's cooperation, it will therefore be important for the therapist to position herself somehow in such a way as to show acceptance of their formulation while at the same time framing an intervention that will allow them to change their position and the interaction.) Therapist: I'm impressed by the way you two are in basic agreement about this issue even though you have your differences. I understand that Frank seems to be following in his father's footsteps and that it's a basic flaw but I just wondered-it seems to me that this is his way of showing his caring and concern for you and his desire to protect the relationship. For after all, by not talking and retiring to the den, he is protecting both of you from the pain of dealing with your difficult problems and helping you two stay together. Frank, 245 I'm impressed by your willingness to take on the role of the villain in the interests of your relationship~in spite of how painful this might be for you, and I understand that it is motivated by your caring for Susan and your desire to nurture the relationship. Susan, given that Frank has some sort of deficit in communicating and is showing protection by not talking, how could you talk to him and let him know how important it is to you to talk together and solve your problems? Compliance and Defiance-Based Strategies Compliance-based interventions are those in which change occurs when clients tr}7 to comply with a therapeutic intervention or prescription. Either the clients will find it impossible to comply with the prescription or the prescription will create an aversive or punishing situation for the clients. Defiance-based interventions are those which are used with the expectation that the clients will rebel or react against a suggestion or directive from the therapist. This type of intervention is meant to influence clients to change by rebelling. Change occurs when the client rebels against the therapist's influence attempt. (Positioning tactics and Restraining strategies are also defiance-based and utilize client resistance.) Copy only with permission of authors. Appendix B: Emotionally Focused Couples' Therapy Manual 247 THERAPY MANUAL: EMOTIONALLY FOCUSED COUPLES' THERAPY FORMAT: FRAMEWORK ASSESSMENT THERAPY STEPS AND THERAPIST ACTIVITIES TERMINATION OPERATIONAL DEFINITIONS THERAPIST INTERVENTIONS, DESCRIPTIONS © Greenberg and Johnson 248 EMOTIONALLY FOCUSED TREATMENT FRAMEWORK L. Greenberg and S. Johnson 1. Define issues as presented. 2. Identify negative interactional cycle. 3. Facilitate clients in accessing and accepting previously unacknowledged emotions underlying the cycle. 4. Redefine the problem cycle in terms of these new emotions and the client's interacting sensitivities. 5. Encourage identification with previously unacknowledged aspects of experience by enactment of redefined cycle. 6. Facilitate acceptance of partner's position. 7. Encourage clients to state needs and wants arising from their new emotional synthesis. 8. Facilitate new solutions. 9. Help clients to integrate new perspectives of the self and the other, solidify new relationship positions and ways of achieving intimacy. 249 Session 1. Assessment Therapist Tasks 1. Delineate conflict issues more precisely and attempted solutions. Identify themes in core struggle. 2. Discuss each partner's perception of the problem. Observable behaviours are noted but the focus is upon how each partner sees the self and the other in this relationship and the stances or positions each takes in the relationship. Identify sequences of problematic reactions as the couples narrate or enact them. How do this couple connect, maintain distance, attempt to influence and protect themselves against each other and the therapist? Allow a 10 minute discussion of the presenting problem for research purposes. Key events are noted. The strengths of the relationship when it is functioning well are assessed. The couples' expectations of the relationship are explored. Norms as to power/control, dependence/independence, and closeness/distance are noted. The therapist also considers the developmental stage of the relationship and the level of commitment. 5. Enquire about the family of origin and life history of the partners. 3. Note and explore patterns in the process of interaction. 4. Enquire regarding the history of the relationship. 250 Note partners' views of male and female roles and the norms mentioned above. Hypothesize vulnerabilities and sources of anxiety stemming from life experiences which may be reflected in the present relationship. How do interaction patterns impact the individual's self-concept and self-esteem? 6. Present Treatment Rationale. The therapist frames problems in terms of deprivation, unmet needs, and interacting sensitivities in the relationship. The problem is framed in terms of stuck emotional chain reactions which have become automatic and which both partners have participated in building and now are imprisoned by. PROCESS NOTE: The goal of the therapist throughout the session is to establish a working alliance, to create rapport and trust with both partners and give them hope for positive outcomes. Since this is an information gathering and diagnostic session much more of the interaction will be therapist-client in nature than in the following sessions where client-client interaction will increase. The therapist by his/her behaviour also creates expectations for the process of the sessions, for example by encouraging clients to speak for themselves not for the other and discouraging disruptive interruptions. Typical therapist activities. Empathic Responding. Direct Questions and Probes as to issues, interaction patterns and intrapersonal anxieties. Observe/Hj'pothesize regarding the central struggle in the relationship. Framing of problem in terms of treatment perspective. 251 Steps of Treatment This therap}' tends to occur in a circular rather than a linear sequence therefore this manual will focus upon the steps of the process rather than attempt a session by session account. The steps in the process and the key interventions follow below. These steps are elaborations of the framework stated above. 1. Define issue as presented. Define problems as seen by the clients. Establish each person's view of the problem, and how they perceive their own and their partner's role in the problem. Establish shared goals. Each person is encouraged to make a full and complete statement of their position. Therapist Activities: Direct questions and probes; Empathic responding; Summarizing and integrating information; Validate opposing reality claims and positions and each partner's need to be right and innocent of blame. 2. Identify negative interaction cycles. An example of such a cycle might be "when you demand attention he withdraws by leaving the room. You become more upset as he refuses to talk to you. You finally give up and also withdraw. Finally after a day or so he initiates superficial contact." In such cycles each of the partners' solutions to the problem intensifies the problem for the other. The therapist explores behaviours, feelings and perceptions involved in the cycle in order to clarify each partner's position in the "dance." Behaviour towards the partner is linked to underlying feelings. Such cycles may be talked about 252 and reconstructed or they may occur in the therapy session where the therapist identifies and comments upon them as they happen. For example, the therapist comments "I notice that when you start to express your views on this topic your partner asks you why you see things that way, and then you seem to get confused and start to explain . . . etc. Negative messages such as blaming the partner are explored in terms of underlying needs. The framing of behaviour in terms of an interactional cycle fosters a perspective of mutual responsibility. The partners are encouraged to develop their position more fully and their positions are validated. Therapist Activities: The therapist identifies and connects elements in the cycle by means of questioning, exploring, clarifying and interpreting each partner's perceptions, feelings and reactions to the other. Negative alienating reactions occurring in the session are pointed out and discussed, for example, mind reading of the other partner or making negative dispositional attributions. Blaming behaviour is not ruled out as unhelpful but used by the therapist to search for the feeling underlying specific accusations. It is developed further rather than challenged as unacceptable. The therapist uses open ended explorations and only interprets if clients are unable to discover their own experience. Examples: a) To clarify cycle and positions the therapist says: What did you do then? or When your partner does — what do you do? You criticize Jack for never holding you and for being cold to you, 253 when he does this how do you feel? b) To draw attention to interactional patterns the therapist says: It seems that when your partner talks you interrupt—I'm wondering what is happening for you, what is it that you are experiencing when you do this? 3. Access and accept unacknowledged feelings underlying problematic interactions. Emotional responses at the periphery of awareness are attended to and linked to heightened self-perceptions. Particular attention is paid to vulnerabilities, fears and unexpressed resentments. Significant events arousing strong emotion are at times reconstructed, or enacted in the session and are focused upon to reveal underlying emotion. Clients are thus exposed to aspects of self and the other not previously acknowledged. This is to be distinguished from the ventilation of superficial and/or defensive reactions; it is a new synthesis of emotional experience. An example of such a superficial reaction would be an angry reaction expressed with no awareness of a sense of threat or underlying fear. These reactions are explored for the underlying experience of fear. Therapist Activities: Evocative responding (see the end of the manual for a detailed description of the modified form of this intervention). This intervention involves focused reflection, probing and interpretation by the therapist. The therapist may attempt to supply missing feelings, or supply sentences for a client to finish. The therapist may also attend to bodily sensations the client is experiencing and to non-verbal behaviour in general. Images and metaphors may also be created to heighten and clarify emotional responses. The focus is upon looking at inner 254 experience and the owning of such experience. This experience is then validated by the therapist. There is a continuing focus on the emotional experience occurring in the present. 4. The problem is redefined in terms of newly synthesized emotional experiences. The problem is now construed in terms of adult unmet needs and sense of deprivation and alienation. Interacting sensitivities are explored and interpreted and individual experience is translated into the meaning carried for the other spouse and the relationship. Such interpretations integrate the clients' affective, cognitive and behavioural experiences. Fears and coping reactions are validated and related back to the responses taught in the family of origin and to key self images. The current need for these responses is explored. New perspectives on the relationship and the partners' behaviour created by the new emotional synthesis are now integrated. For example, a blaming response may be seen as an expression of a need for love or a withdrawal seen as a fear response instead of as an attempt to punish or hurt. Attempts are made to capture these new feelings as they are occurring in interactions during the session. The clients are encouraged to interact with each other in the sessions and to share their underlying feelings as they emerge in the session in reaction to their partners. There is a strong focus on what is occurring in the present between the partners. These feelings are explored fully, both in terms of their personal meanings and their meaning to the partner. Therapist Activities: The impact on the relationship of the personal vulnerabilities 255 explored in Step 3 are now clarified. The therapist interprets elements in the interactional sequence in terms of underlying needs and fears which stem from interacting sensitivities. For example, Jim is vigilant regarding actions of Jill's that he perceives as rejecting and responds by bullying; Jill is sensitive to bullying and responds by rejecting. This cycle prevents contact and the meeting of the partners' needs in the relationship. Evocative responding may also be used as well as interpretations of issues and defensive reactions in terms of family of origin schemata. A present centered focus is maintained and partners are regularly asked what they feel right now in response to what their partners just said. 5. Identifying with disowned aspects of experience in the redefined cycle. The cycle, redefined in terms of underlying emotional experience and needs,. is enacted deliberately in order for the partners to become more aware of their underlying needs and to gain a sense of control of these automatic responses. The clients are instructed to become more fully "who they are" by engaging deliberately in their part of the "cycle rather than trying not to engage in this behaviour. For example, the withdrawer and the pursuer are both encouraged to experience more fully their underlying feelings and needs which were previously disowned. Aspects of experience such as the withdrawer's fear of being overwhelmed and need to protect and the blamer's feelings of being unloved and need for support are fully discussed and then prescribed. Each person is asked to identify with disowned aspects of their experience, to develop their position fully and to engage deliberately in some of the behaviours associated with their previously disowned feelings and needs. This is an intrapsychically oriented intervention focusing on enacting disowned parts rather than enacting the negative 256 interactional cycle as occurs in some paradoxical interventions. Distancing partners, for example, may be asked either in the session or for homework to protect themselves deliberately or practice putting up a wall as a way of becoming more aware of and gaining control over this sometimes problematic aspect of their own behaviour. Pursuers are asked to engage deliberately in support seeking behaviours and to become aware of their need to be nurtured and the feelings associated with this. If one partner feels too dependent or feels anxiety about being intimate, he or she is asked to identify with the dependent or fearful aspect of their experience rather than to deny these parts or try to disown them. Both partners are reassured at this point that even though it might seem strange or be difficult to act in a manner that they construe as problematic (such as dependent or afraid) that these are the feelings they are actually feeling and that this is only being more congruent. It is emphasized that it is important in resolving marital conflict to take responsibility for one's feelings and that accepting these feelings and deliberately behaving in ways associated with these feelings will give them more control and choice of these feelings and behaviours. Once partners have identified with disowned aspects of their experience it is possible to integrate these aspects both intrapsychically and also into the relationship. Identifying with disowned aspects of experience is worked on in the session and given as homework and people are asked to do it deliberately if possible, or to "go with" their experience when they begin to feel their previously disowned experience rather than fighting against that aspect of themselves. 257 Therapist Activities: Suggests people identify fully with previously disowned aspects of their experience. The therapist conveys an ultimate acceptance of each person's position, feelings and needs by suggesting that people do what they are doing rather than try not to. Although there is a "prescription" of certain behaviours and experiences, the focus is on having people do what they do with full awareness and responsibility (in order to make previously automatic responses deliberate) rather than to prescribe a paradox to gain therapeutic control of the interactional cycle. 6. Acceptance of partner's position. The focus is now upon the communication to the spouse of the newly experienced emotional responses, and the partner's acceptance of these responses. The therapist facilitates acceptance of the other's needs on the part of each spouse, primarily by tracking interactions and blocking or exploring non-accepting responses. The therapist helps the couple construct the conversation they might have had if they had been in touch with and able to report their feelings and vulnerabilities. The phobic avoidance of the expression of vulnerability in the relationship is usually confronted in this process. This session is not directed towards the teaching of the specific skill of empathic listening but toward helping partners reveal new aspects of themselves to their mates and facilitating a new intimacy and contact between the partners. Blocks to one partner's ability to hear and accept the other's experience are examined and interpreted in terms of that partner's view of self, past learning in family of origin and catastrophic fears. The therapist facilitates acceptance of self and others in contrast to the usual pattern of reciprocal disqualification which occurs in distressed 258 relationships. Therapist Activities: Evocative responding; interpretation and labelling to clarify relationship events; drawing attention to the nature of responses to the partner and the impact of these responses, and suggesting alternatives. Example: a) I feel alone (experience of abandonment and helplessness integrated in previous steps) because you never show yourself, your feelings; never really show me how you feel. b) I don't show you my feelings, well I suppose I don't, I'm afraid to show you, when I have I get attacked. Therapist: (to A). How can B show you his feelings in a way that you can hear them? 7. Expression of needs and wants. The emotional synthesis of the issue in terms of intra-individual and interpersonal experience leads to a clarification of needs and wants in the relationship. One partner can now directly ask for what he or she wants or needs from the other, and the implications of these desires for the individuals and the relationship can be examined. Key attitudes underlying the positions each partner has taken in the relationship begin to be explored. Therapist Activities: Focus interaction upon the expression of needs and wants. Clarify and interpret such needs if necessary. 8. New solutions. 259 The statement of needs and wants, accessed, integrated and accepted by the spouse, leads to the creation of new alternative solutions to the couple's struggle and the presenting problem which is symptomatic of this struggle. The therapist clarifies and explores aspects of the solution with the couple and again helps them to confront blocks to positive responding. The therapist also draws attention to and highlights new positive patterns of interaction. New solutions constitute a redefinition of the relationship, for example, a relationship may become one in which one person can state needs and the other can give support rather than a relationship in which one has to coerce and bully the other into seeing and responding in a certain way. New solutions are assessed in terms of the needs of both partners and their general feasibility and if possible enacted in the session. Therapist Activities: Clarify and explore new solutions, for example, how can a partner help his spouse trust him and feel safe in the relationship? Perhaps he can do this by engaging in activities that he knows reassure her that she is important to him. This sense of safety will then enable her to respond to him in ways that he finds satisfying. 9. Integrate new perspectives. The therapist helps the couple develop a shared perspective, a detailed picture of the relationship, and engage in metacommunication as to the past and present nature of the relationship. The therapist clarifies new positions and positive sequences of emotional response and the new interactional cycles. The past relationship positions taken by the partners and the negative cycle are discussed. New goals for future relationship development as well as new 260 ways of creating and maintaining intimacy are discussed. Therapist Activities: Summarizing. Termination issues. PROCESS NOTE: These nine steps tend to be cyclical; the therapist may circle back to previous steps if necessary, or begin the cycle of steps focusing upon some new aspect of the couples' core struggle. In the sessions the partners continue to expand their awareness of their positions in the relationship and the needs and fears underlying these positions. As positions, interaction patterns and key underlying emotional responses become clearer the couple's manner of interacting becomes less reactive and automatic, alternative behaviours, feelings and thoughts are experienced and experimented with. The partners develop a shared perspective of the relationship and begin to "woo" each other back into intimacy. Since previously unaccepted aspects of the self have been accessed, validated, expressed and integrated into the relationship anxiety reducing defences are less and less evident. As therapy continues, ideally the therapist does less and the partners interact more and more, helping each other through the therapeutic process. Termination Session: This session, like assessment, will always follow a certain format. The treatment process will be reviewed, new interaction patterns highlighted, and the present state of the relationship in terms of goodwill, trust, open contact, closeness and positive affect assessed and summarized. The original presenting problem is discussed and post treatment measures completed. 261 OPERATIONAL DEFINITIONS NEED: Awareness of an urgent lack of nurturance, safety, or basic relatedness necessary for survival and a sense of well-being. Boszormenyi-Nagi suggests that the other is the object, the "ground" for an individual's identity delineation and security needs and labels this "ontic dependence." INTERACTIONAL CYCLES: Sequences of behaviours where the response of one partner becomes the automatic stimulus for an automatic reaction in the other, e.g. I nag because you ignore me, no I ignore you because you nag. Such cycles are alienating and usually spiral into more intense conflict. INTERACTING SENSITIVITIES: The strategies designed to cope with the special sensitivity or vulnerability of one partner which elicits the special vulnerability of the other resulting in an alienating emotional chain reaction. The issue to which the partner is sensitive often has historic significance. This term then refers to the sensitivity which underlies core feeling reactions which lead to negative interactional cycles. POSITION: A point of view, perspective or orientation in a relationship. A view of the self in relation to the other which creates a set of expectations which guide perceptions, feelings and behaviour. Positions tend to become rigid and polarized in a context of threat to self-esteem or well-being. CONTACT: To meet or come together, to touch, to connect or experience reciprocal 262 openness, allowing the other to impact you. To communicate openly on an intense personal level. To touch-to permit part of the body/self to come in contact with so as to feel. INTERPRET: To clarify meaning by connecting or relating one element in a situation to another, for example, by connecting relationship behaviours to intrapsychic perceptions of the self. It is also a process of imposing meaning upon events, or creating a new frame of reference. Can be more or less confrontive. CLARIFY: To make the implicit explicit-deals with what is just beyond awareness. Symbolize more completely. Can be a mild form of interpretation. Therapist Intervention. Evocative Responding This intervention consists of probes or statements which attempt to clarify and heighten the clients' emotional experience in therapy and make the automatic a focus of conscious awareness. The elements of experience focused upon are: Stimulus (cue and appraisal) Arousal Response The therapist's focus depends upon the process of therapy: SITUATION 1. If Stimulus, Arousal and Response (S.A.R.) are all clearly experienced in awareness, that is if the stimulus is clear and differentiated, arousal 263 is present and acknowledged, and response is expressed with ownership and inner awareness then the therapist pushes for more differentiated inner awareness and a clearer expression of experience and needs. Thus spouses are exposed to aspects of themselves and each other that is beyond awareness. EXAMPLES: Client -When you yell at me like that I feel hot and shaky, I feel afraid and I just have to get away, so I leave the room. Therapist -How do you experience the fear, as shakiness? Client -When you look concerned and sit close to me like that I feel very uncomfortable. I feel smothered, hemmed in, so I turn away, close off and ignore you till you go away. Therapist -Smothered, you feel like you don't have room to breathe. That's scary, you feel anxious? (Client nods), what will happen if you don't turn away? Client -She will expect me to be a certain way, warm, and I can't feel a certain way. I know that I'm not the husband she wanted. SITUATION 2. If Arousal is missing, the therapist heightens using images, probes and interpretations. EXAMPLE: Therapist -264 Is that painful for you? Client -Yes, very. Therapist -It almost sounds like you're in a cave and shouting help, help, and all you feel you get is the echo of your own voice. * * * * Client -I have to deal with it, not burden him with my jealousy, he's struggling too. Therapist -Sounds like you want to hold his hand and help him while he makes love to his lover. SITUATION 3. If the Stimulus is not clear, specific and alive the therapist focuses upon cues and the meaning of the cues to the individual so as to differentiate the impact of a particular stimulus. EXAMPLE: Therapist -What is it that sparks off your cynicism and makes it hard for you to listen to him? Client -He's so condescending, I get hostile. Therapist -265 What about the way he does this gets to you? Client -He is so logical, never lets go, and that look on his face of I know better. Therapist -He seems cold and superior. SITUATION 4. If the Response is unclear the therapist differentiates the Stimulus or helps the couple enact the sequence so that they may unfold and dismantle their interactions in terms of emotion, cognition and behaviour. EXAMPLE: Therapist -So what happens when Pat tells you that she doesn't want to make love, and turns away? Client -Nothing, I accept it, might ask her why. Therapist -I'm wondering if you don't feel hurt or feel that need to get back at her? Appendix C: Implementation Checklist 267 COUPLES PROBLEM SOLVING PROJECT IMPLEMENTATION CHECKLIST Couple No. Session No. Rater Instructions to Raters: Place one check mark on the rating form under an intervention each time that intervention is noted. An intervention is defined as a therapist statement. INTERVENTION CHECKLIST A. Problem Definition 1. The problem is defined/redefined in terms of the emotions underlying the positions taken in the relationship and the sense of deprivation experienced by the couple. 2. The problem is defined in terms of goals for therapy, stated in specific behavioural terms. 3. The therapist clarifies and elaborates the basic positions taken by the partners in the relationship. 4. The therapist gathers data on attempted solutions to the stated problem. B. Management of Attacking Behaviours 5. The therapist validates or develops the positions implied by negative behaviours, i.e. blaming and interprets such behaviour in terms of underlying needs and feelings. 6. The therapist positively connotes negative behaviours, i.e. blaming, in 268 terms of the function of the behaviour in the cycle. C. Directing the Process of Therapy 7. The therapist probes for and heightens emotional experience, especially fears and vulnerabilities, clarifying emotional triggers and responses and focusing upon inner awareness. 8. Restructuring a) The therapist redefines one or both spouses' positions in a way that emphasizes the complementarity of their positions and punctuates the interaction differently. Redefinition is used to direct partners to change their perceptions or behaviours, i.e: Partner A identifies B as the source of difficulty and complains about B's distancing, retiring behaviours, unavailability when she needs support. The therapist responds by asking, "So when did you give up trying to be close?" or "How do you let him get away with being so distant?," thus placing the emphasis or punctuation differently and directing partners to see their cycle differently. The implicit suggestion is that Partner A might want to reconsider and perhaps try to get what she wants from B. b) The therapist "unbalances" by forming a temporary coalition with the partner who is "one down." The therapist tells partner A to push B for what she wants, i.e. (To partner A): T: I think it's important that you stand up more for yourself. 269 It's important to tell him that this is what you want! A: Fred, I'd like you to attend the Sunday evening dinners with my family. B: I'd like to, Jane, but I can't because your family drives me to distraction. T: (unbalancing and throwing weight behind A) I think it's important for you to tell him again. I don't think you should let him sweet-talk or persuade you. 9. The therapist tries to help clients discover emotional meanings, and differentiate and elaborate upon these. 10. The therapist reframes the couple's negative cycle in positive terms, emphasizing the protective function of each person's position. 11. The therapist encourages partners to be aware of and identify with the feelings underlying their positions in the cycle. 12. The therapist uses information provided by the partners as supporting evidence for the reframe. D. Dealing with Blocks 13. The therapist clarifies the interacting sensitivities underlying behaviour and the meaning of individual emotional experience is interpreted in terms of its effects on the partner and the relationship. 14. The therapist acknowledges improvement but restrains the couple by suggesting that partners "go slow" and that too much change too quickly might be. dangerous. 15. The therapist helps partners explore blocks to accepting the other's 270 positions in terms of underlying feelings, self-concept or experiences in family of origin. 16. The therapist elicits and speculates on the dangers of improvement or change. Information from family of origin may be used to elaborate on changes. E. Problem Solving 17. The therapist facilitates partners in accessing underlying emotional needs and wants and expressing them to their partner. 18. The therapist prescribes the negative interactional cycle. 19. The therapist helps clients to crystallize their new views of their partners and explore new feeling responses. 20. The therapist uses information provided by the clients as supporting evidence for a reframe. 21. The therapist clarifies with partners a shared perspective on the relationship. Metacommunication is facilitated. 22. The therapist expresses concern at reports of change and recommends a relapse. Additional Categories 23. Information gathering. 24. Refocus on topic. 25. Not codable (e.g. therapist assigning homework). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0053663/manifest

Comment

Related Items