DATE: October, 1991
EDITION: Volume V, Number 2
CONTENTS
-Uri Rueveni wrote that he and his wife Mira toured Japan and China this summer, with plans to conduct a workshop on network intervention at the Department of Psychology at the University of Hong Kong. He has recently been working on the issues of adult child caregivers for elderly parents. He developed a structure for children to interview their elderly parents about family history, and conducts a monthly support group for caregivers. Ross and Joan Speck, Uri and Mira, and Richard Bailey conducted a network intervention for a patient of Mark Sirkin. The patient was caught up in a folie-a-deux with an old friend who claimed to be the Messiah. The intervention appears to have been quite productive for Mark's patient and family. Mark and Uri have prepared a manuscript for an article on the intervention.
-Hans Maurer, a social worker from Switzerland, recently completed a series of visits to a number of sites around the world, including Montreal and Stockholm, to learn about current practice in network therapy. Gunnar Forsberg of Stockholm sent me a copy of Maurer's report of his study trip to Sweden. I have sent Hans Maurer a letter and included him on the NETLETTER mailing list. NETLETTER would welcome his correspondence describing his study visits. His report on the Swedish visit noted that there was little knowledge in Sweden of the Montreal team, which has been working for many years and is well known in France and French-speaking Switzerland [NETLETTER pleads guilty as well - we will check it out for the next issue]. In his encounters in Sweden (which included not only the group most frequently in correspondence with NETLETTER, but other Swedish groups and individuals as well), Maurer found his informants to be very enthusiastic about their network practice, claiming "failures close to none, results were always positive, completely aborted experiences were completely out of view." Comparing Swiss and Swedish experience, Maurer was looking for insight on the problem of how to get access to a network. Apparently, Swiss practitioners see the problem as the network interventionist initiating entry, while the Swedish experience is to wait for the client to approach his/her network members to convene a meeting. Other important lessons were the early use of a collective, rather than individual, approach in preparing the stage for network intervention, the utility of network intervention in crisis situations, the idea of network intervention as a last recourse after other methods have failed, the differentiation of the case worker from the network interventionist (with the latter entering from the outside of the system which includes the former), and the importance of bringing in "support members" so that every member of the client family has an informed, reliable, and supportive assistant in a meeting. Visiting network intervention projects at St. Lars Hospital in Lund, Langbro Hospital and St Goran Hospital in Stockholm, Maurer was impressed by the influence of the institutional systems' readiness to accept the network approach on the productivity of network intervention. Hans Maurer is planning to use the knowledge he has gained to train more Swiss practitioners in the network approach, and to develop research on the social networks of multiple-problem families in Geneva.
-Gunnar Forsberg visited Boston this summer, giving NETLETTER a chance to get an update on the Emergency Network Centre [Readers should note that this current project, described in NETLETTER IV,(2), 1990, is not the same as the earlier project, described in REVIEWS of this issue.]. The Centre was slated to meet with the Botkyrka Commune politicians and social welfare policymakers in September for its two-year review. There has been a strong positive response, with a large number of self- referrals. A crisis team, staffed by social workers and working with the police (particularly over the weekends), has been making referrals, as have the schools, allowing opportunities for preventive work. The Emergency Network Center provides supervision to the professionals on the crisis team. In contrast, only two of the social welfare districts in Botkyrka have been making referrals. The relatively small number of referrals from the social welfare office probably reflected the very high turnover in these offices, resulting in the loss of social workers who had had training in the social network perspective. Losses of funding loom. Seed money from the National Board of Social and Medical Health will run out. Funding from the Ministry of Health and Welfare ends January 1. Meanwhile, Commune social policy is changing its metaphor from that of the welfare state, in keeping with political ideological changes toward a more conservative direction in the country. The emerging metaphor is that of a business, with the ideal of a profitable social service organization getting fees from its referring organizations. Johan Klefbeck and his colleagues from the Emergency Network Centre continue to do training in social network therapy in and beyond Botkyrka, on an independent basis. Johan, the Centre Director, will be moving on next summer.
The REVIEWS section of this issue includes several reports on presentations at the 1990 annual conference of the American Psychological Association in Boston. I was hoping to encounter psychologists working on the concept of social network. I found none presenting research or practice on this topic. There were a number of presentations on social support, a concept which is related but not identical to social network. My experience has been that social network interventionist are more likely to cite the social support literature than social support experts are to cite the literature on social network analysis or intervention. A welcome exception is the collaborative work of social support practitioner and researcher Ben Gottlieb and social network interventionist Anne Coppard (Gottlieb & Coppard, 1987), reviewed in NETLETTER, II, (1), 1987, which described the use of network therapy to create support systems for the chronically mentally ill.
My experience at the conference was similar to my reading of the social support literature. I found myself chanting the mantra, "social networks are not support systems," as I heard people discussing social support in isolation from its social structural context, and without attention to complications from the fact that people in one's social context can be harmful rather than helpful, or sometimes both, and that the same people who are sources of support are sources of demands that can be draining. I was encouraged that a number of presenters spoke to the necessity of addressing social support systematically, as part of a complex social contextual field.
In the broad view, network therapy is one current in the stream of psychotherapeutic and social welfare work directed to addressing human needs by intervening into individuals' social contexts. Social support intervention is another, and both are relative newcomers compared to the fields of social work, pastoral ministry, and community healing practices from village and tribal traditions. Especially the given the difficulties that network intervention has encountered getting established in psychotherapy and social welfare practice, I find the relative isolation of network intervention and social support intervention from each other to be frustrating. I would be gratified if NETLETTER could play a role in facilitating the integration of the two fields, and welcome correspondence from readers on this topic.
Sarason's observations at the conference about the integration of knowledge about the interpersonal and intrapersonal dimensions of social support illustrate the significant contributions that the social support field can contribute to network therapy. Ben Gottlieb, in his panel contribution, was somewhat critical of the "psychologizing of social support." Yet, it appeared to me that a lot of valuable work has been done on the attributes of individuals and their influence on individuals' access to social support. Ross Parke's address provided an elegant illustration of the significance of individual attributes/capacities by describing some fascinating relationships between parental socialization practices and children's social competence and success. This type of research can provide the network therapist with more power in the coaching of individuals to improve, transform, or enlarge their personal networks.
I would also like to see the social support literature enriched by more work like Ben Gottlieb's and Anne Coppard's, which uses the network intervention paradigm to create positive changes in individuals' social support systems.
References
Gottlieb, B., & Coppard, A. (1987). Using social network therapy to create support systems for the chronically mentally disabled. Canadian Journal of Community Mental Health, 6, 117-131.
-Roberge, C. (1990). Single mothers' experiences of their social networks: A phenomenological study of separated women's relations. Unpublished doctoral dissertation, Massachusetts School of Professional Psychology, Dedham, Massachusetts.
[Editor's note: It was my privilege to serve as discussant in the colloquium on this dissertation presented by Carin Roberge.]
This is an exploratory, phenomenological study of the social network experience of separated or divorced women with children. Eight white women, who were middle-class before their separation, at least 28 years old, with children ages 3-12 were interviewed at least three years following their separation. As a phenomenological study, this research used a semi-structured interview designed to articulate the women's own account of their experience. The investigator looked for themes, and was not testing hypotheses. Subjects were asked about their experiences of their relationships, positive and negative aspects of relationships, and comparisons of relationships over time.
It is the nature of such studies that the findings are quite particular, and that it is not appropriate to treat the findings as facts which are generalizable to the population from which the sample is drawn. My own observations on the study and its findings should be taken with these qualifications in mind.
Several subjects spoke of the costs of asking for help. They reported that asking others for support often yielded relational stress as well as support, and that caring for others in their networks was often at the cost of worrying about others.
Subjects who complained about the burden of caring for children as blocking their access to social relationships seemed to be the least successful of the group at reconstructing their social worlds after separation. In reading the study, I began to develop a picture of what a "successful" social network reconstruction entailed for some of the subjects. The more successful subjects appeared to show a high level of conscious, articulate awareness of self-in-relation. They saw the changes in their social networks as very much the products of their own creative efforts, and were gratified by these achievements. They did not utilize their children for social support. Despite their pride of accomplishment, even these "successful" redesigners of their social worlds still reported that they were not receiving adequate social support.
Interviews of those women who were just at the three-year mark post-separation revealed that three years may not be enough time to get a picture of a stable equilibrium in the subjects' household and social network organization. At three years post-separation, some women appeared to be reporting the changes in their relational lives as works still very much in progress. Women who had moved out of their community after the separation described much more dramatic changes in their relational life than did women who remained in place after the separation.
Before reading the study, I had anticipated that subjects would describe major transformations in the relationships they had held in common with their ex-spouses. It has been my observation that separation challenges all the relationships held in common, with some people dropping away, some choosing sides, and a relatively few individuals maintaining separate ties with both members of a divorcing couple. None of the subjects mentioned this phenomenon, with the possible exception of one former victim of spousal abuse, who talked extensively about the opening up of her relational world once her husband was no longer in the picture to enforce her social isolation. Most subjects understood the changes in their social networks as driven primarily by the new role strains and economic strains of their separated parent status.
This study suggests to me an interesting hypothesis for future testing. Given the relative success of women who saw their social experience as the product of their own action, it may be that further study of divorced women will show that women who initiated the separation and divorce process are more likely to report successful social network reconstruction than women whose spouses initiated the breakup.
It is interesting to speculate on the influence of race and social class on the experience of the subjects. We might get quite different responses from poor or African-American women, who are socialized into a more collectivist network frame of reference.
The study also gave me a fresh perspective on the common wisdom (particularly among women) that men are far more likely than women to get involved in a new relationship soon after the breakup of the old one (unless they have already started the new relationship before breaking up the old one!). If one understands the "success stories" in this study to reflect women's greater competence in relational construction, reflecting the differences in gender socialization, then perhaps men find a new partner faster in order to get expert assistance in reconstructing their own social worlds.
**********************************************************************
-Wills, T. A. (Chair), Coyne, J. C., Cutrona, C. E., Rook, K.S., Sarason, I. G. (August, 1990). Social support: What does it mean, how does it work? Panel presented at the 98th Annual Convention of the American Psychological Association, Boston, MA.
[Editor's note: The panel discussion was organized around four questions. I paraphrase each question, and then summarize the discussion. Where my notes allow, I credit the comments of individual panelists.]
WHAT IS SOCIAL SUPPORT? Sarason elaborated the question as follows: Is social support better represented by social integration (i.e., the existence of social ties), or through the appraised availability of supportive functions? Is it a unidimensional construct, or is it best approached through measuring somewhat independent functions provided by social relationships? Is it best conceptualized as a property of the environment, or of the individual? (The latter represents the "personality model" of social support, which distinguishes people by the amount of social support they possess, often implying that the differences among individuals' amount of social support represents some individual characteristic, e.g., social competence.) Sarason recommended integrating current work on social support, which treats a broad range of interacting variables, i.e., variables in the interpersonal context and intrapersonal factors (including both the donor and the recipient of support). He urged that more studies utilize direct observations of social transactions. [My own impression of this field is that most studies rely on the subjects' own reports of social support.] Sarason stated that perceived support, based on subjective appraisal of self and others, proves to be a more "active ingredient" than received support, that is, a more objective reconstruction of supportive transactions.
Coyne joined Sarason's position about studying interaction of variables, asserting that it is "illusory to isolate variables." He noted that the concept of social support appears to have different meanings for different populations. Negative social reactions and interactions prove to be more powerful determinants of outcome than do presence or absence of social support. [Mantra: A network is not a support system.] Lack of support tends to correlate with more actively harmful social events. In most people's daily lives, support may not be an active concept; constraints and routines may prove to be more important. [Here, I would argue, is another argument for the social network concept, which includes but goes far beyond social support. Certainly, transactions with the personal network embody much of what constrains and routinizes our lives.]
Cutrona advocated for clarity and precision of research, calling for observational studies with behavioral coding. She spoke of her own work, which found that specific combinations of type of stressor and type of support behavior yielded subjects' reports of perceived supportiveness. For subjects to perceive social support in the context of situations involving controllable events, they needed problem-solving support. In the context of uncontrollable events, nurturing and soothing support produced a perception of supportiveness.
Rook spoke of his interest in social support as coming from his interest in the problem of loneliness. He stressed the importance of simple companionship, not just protection, in generating the experience of social support.
HOW DOES SOCIAL SUPPORT WORK? Cutrona amplified the question as follows: Is social support helpful exclusively (or primarily) under stressful conditions (the stress-buffering model), or does it operate irrespective of the stress level? What mechanisms are the most plausible? She noted the need for different strategies to study specific support (e.g., esteem support, yielding self-efficacy beliefs) and nonspecific support (e.g., sense of others' caring or of one's importance to others) She noted that specific and nonspecific mechanisms need to be studied differently. In studying social support, one looks at perceived, or expected, social support, or at received support, that is, the subject's retrospective tally of supportive acts from others. In the case of specific support mechanisms, it has been demonstrated that subjects, relying on recollection, cannot report received support accurately. There is therefore a need for observational studies to gather accurate data. In the case of nonspecific support, which is basically a concept of the subject's experience, "we already know how to measure fairly well." The subject goes through some kind of "averaging process," which [your editor infers] results in fairly reliable reports.
Sarason spoke of the importance of "self-acceptance" gained from social support in non-stressful situations. Rook noted the necessity of looking at the variables of support and control in tandem; he stated that these social phenomena co-occur, "contaminating the relational waters," and cannot be understood in isolation from each other. [This latter note of caution reminds me again of the network therapist's warning that one's social network is far more complicated than a simple "support system."]
IS SOCIAL SUPPORT BETTER REPRESENTED AS THE PRESENCE OF POSITIVES, OR AS THE ABSENCE OF NEGATIVES? Rook raised the question, how applicable are the effects of social support, as discovered through research on general populations, to people who are very isolated, or to highly vulnerable populations (e.g., caretakers for the ill or disabled, the elderly, etc.)? Rook stressed the importance of a primary, central relationship in providing social support. Coyne held out for the "primacy of the negative." Sarason claimed that people who report higher social support are better problem solvers and better able to anticipate the future than those who report lower social support.
WHERE DO WE GO FROM HERE? Time was running out, so there were the usual ritual, unanswered questions, i.e., what are the implications from research for intervention, what kinds of basic research are needed, and will help drive intervention design? The moderator provided a rousing ending by reporting a study showing that the degree of social integration was correlated with the prevention of AIDS among drug users.
**********************************************************************
Heller, K. (Chair), Thompson, M.G., Cohen-Silver, R., Gottlieb, B.H., Reich, J.W., Maton, K.I., Bembry, J.X., de Cuesta, A., & Caplan, R.D. (Discussant). (August, 1990). Understanding factors responsible for support intervention success and failure. Symposium presented at the 98th Annual Convention of the American Psychological Association, Boston, MA.
Mark Thompson's contribution was "Optimizing support interventions' effectiveness through a focus on transactional processes." His discussion of a support intervention with older adults was primarily in the form of a critique of existing work in this field. He spoke of "uniformity myths" in the literature, which assume homogeneity of problems in populations. He urged attention to the issues of existing ties versus new ties, and quantity versus quality of ties. Noting that most interventions try to add or graft new ties, he argued for enhancing existing, nonsubstitutable ties, e.g., strengthening support within family systems. He differentiated three subgroups of isolated elders. One group comprises individuals who have good-quality friend and/or family ties, but limited contact opportunity (for reasons of money, geography, health, etc.). Another group has lots of social contacts, but these are primarily negative. A third group has few contacts, and these are negative. He criticized many social-skills training interventions as making clients dependent recipients of aid, and advocated training for reciprocity and helping behavior. He suggested that social support interventions with older adults be framed as "role interventions," which enhance existing social connections, and prepare the client to establish new connections. Such interventions are best embedded in broader program goals.
Roxanne Cohen-Silver's contribution was "A study of successful and unsuccessful support providers following bereavement." This study examined both the central subject and members of that individual's support systems. [Because it was designed to look at least as well as most supportive individuals, it was closer to the network approach than many other social support studies.] Interested in the problem of loss of social support from one's network in response to stresssful events, Cohen-Silver looked at the problem of the discrepancy between the (often unrealistically high) expectations of the recipient of support and their intimates' assessment of the recipients' support needs (which are often inaccurate). Sixty widows and 108 significant others were interviewed. The others were drawn from the widows' identification of the three most and least helpful people in their lives. Questions probed perceptions of helpfulness, interdependence with the widow, intimacy with the widow, and distress over the widow's current situation. It was noteworthy that almost 50% of the people identified as important by the widows were described as not being very helpful. The groups identified as most and least helpful did not differ in pathology, life satisfaction, boundaries, perceptions of the widow's need or of their own helpfulness (everybody responded that they were "better than average" in helpfulness to the widow!). The groups did differ in self- report of frequency of offering support (by telephone and in person), in their assessment of intimacy, interdependence, love and affection, closeness and satisfaction with their relationship with the widow.
Benjamin Gottlieb's contribution was "Quandaries in translating support concepts to intervention." He found that research provides little guidance for conduct. The theoretical challenge for research is to understand the fit between the particular stressor, the characteristics of the social support, and the charactistics of the context in which the stressor and support are embedded. When one tries to take this model into practice, one finds fluid, changing situations which can defy one's efforts to specify the variables. Researchers' concepts of support may not fit people's experience, and hence appraisal, of support. There is always a danger that one's intervention may transform natural helpers into professionals. The study of naturally occurring relationships may not fit new ones, and most social support interventions focus on grafting new ties. Research has tended to study perceptions of social support, not actual transactions (Ben called this the "psychologizing of support."). Researchers' focus on recipients' judgments of social support has neglected the understanding of how the relational context shapes judgments, and interventions need to focus on relationships. Even within the context of a "psychologized" study of social support, little is known about the impact of individual personality on the experience and practice of support. Ben proposed that, since the culture of intervention programs differs so much from the natural ecology of supportive transactions, there should be more research which directly studies and compares intervention programs themselves, "studying the intentionally formed relationship."
John Reich's contribution was "Mental health interventions for older adults: Some unexpected results." He described an intervention study designed to assess the effects of efforts to improve personal control. Subjects were given four sessions of training designed to enhance perceived control and mastery. In the placebo condition, subjects were visited at home by a staff member who engaged in "friendly conversation." In the case of subjects whose baseline measure of locus of control identified them as "low internals" (i.e., they tended to locate the source of control over their lives' events outside of themselves), the placebo intervention improved their mental health! The training intervention proved to benefit the mental health of subjects who measured as "high internals" (i.e., they perceived the locus of control as within themselves). There was also an interesting interaction between the locus of control variable and the subjects' reports of the degree to which they were actually dependent on other people. "Low internals" who reported that they were dependent on others showed the strongest effect from the placebo intervention.
The contribution from Kenneth Maton, James Bembry, and Alfred de la Cuesta was "Social support nested within meaningful activity: Youth in community service." They described a "Magic Me" program which linked Baltimore middle-school children with elderly partners, and examined the impact of the program on depression, self-esteem, and perceived meaningful contributions. Noting that support is best understood in context, they asserted that supportive community interventions are most productive when they converge with meaningful activity. [I was struck with a parallel principle in network construction, i.e., that one is most likely to form new ties in the context of activities which are in themselves meaningful and important to the client.]
**********************************************************************
-Parke, Ross (August, 1990). Family-peer systems: In search of linking processes. Address presented at the 98th Annual Convention of the American Psychological Association, Boston, MA.
Noting that children are embedded in a variety of social contexts, with interpenetrating influence, Parke focussed on the influence of the parent/child relationship on peer relationships. Parents teach children social skills, they model and coach social interactions, and regulate the child's access to opportunities for social interaction with others. Parents are gatekeepers informally, in their choice of neighborhod to live in, in the homes the family visits, and in the overlap of the parents' and the child's social networks. As formal gatekeepers, parents control access to daycare, preschool, and membership in clubs, teams, and organizations. This parental intervention is more effective with younger children.
Parke reported on findings from observations of parent-child play, and assessment of the peer acceptance or rejection of the children observed. Parents of children rejected by peers were seen as generally less effective instructors. Children were more likely to be accepted by their peers when: a) The duration of play with parents was longer; b) Parents were more likely to use questions rather than directives in the play (the effect of suggestions on peer acceptance fell between questions and directives); c) Positive affect was displayed in the parent-child play; d) Parents demonstrated clarity of emotional expression.
Learning to regulate affect improves a child's acceptance by peers. Emotional understanding (deciphering and encoding emotional expression), which improves peer acceptance, is learned through emotional expressiveness in the family. Other studies have shown correlation between emotional decoding skills and sociometric status. Judges had more difficulty recognizing the emotional productions of rejected than of accepted children. The longer the duration of parent-child play, the greater the child's ability to recognize parental facial expressions. Judges were more able to recognize a child's emotional productions the longer the duration of parent-child play.
**********************************************************************
-Klefbeck, J., Bergerhed, E., Forsberg, G., Hultkrantz-Jeppson, A., & Marklund, K. (1987). Natverksarbete i multiproblemfamiljer - Forskningsprojekt i Botkyrka. Tumba: Botkyrka Kommun Kommungemansam Socialtjanst.
[Editor's note: Many thanks to Gunnar Forsberg for his painstaking translation into English of a substantial portion of the report of the Botkyrka Network Project, "Network work with multi- problem families." My summary review of his translation follows. Gunnar has promised to send along translation of more of the report later.]
Introducing the study, the authors noted that Sweden had become (in)famous in Europe for so readily placing children in foster care in response to perceived parental failure to care properly for their children. This international notoriety reflects an internal national debate, polarized between those who place primacy on "psychological parenting" which can provide adequately for the child's needs, and those who hold the position (based partially on psychoanalytic object relations theory) that the relationship with the biological parent is indispensable and irreplaceable. The authors advanced a network perspective, which emphasizes that children have supportive relationships with a wider set of individuals, and that the child's personal network can become even more important in cases where marital conflict, substance abuse, or emotional illness interfere with appropriate parental attention to children's needs. From a network perspective, the appropriate response to parental dysfunction is, first, to involve and support the child's personal network. If placement outside the family is unavoidable, placement within the network should be the first choice, so as to disrupt as few social ties as possible. In the event of institutional placement, network intervention can sustain social ties with the child's primary social environment. The network project emerged in Botkyrka Kommun, one of Sweden's poorest municipalities south of Stockholm, populated by Swedes from all over the country as well as immigrants from Europe and elsewhere. Social work and psychotherapy professionals in the Botkyrka region formed a study group in 1978 to learn and practice social network intervention. The five authors, members of the original study group, set up the network project in Alby, a section of North Botkyrka, in 1984, as a research and development project with the goal, "to investigate the consequences of a social worker working with the network perspective and applying network methods when handling all child welfare matters at Alby Social Welfare Office." Their network methods ranged from mapping to mobilizing networks through different types of network meetings around an individual or family in crisis. The project plan was accepted by the Swedish delegation for Social Research, and financially supported by the National Board of Health. Lennart Grossin of the University of Stockholm served as scientific consultant to the project. The fieldwork ran from September, 1984, through November, 1986, and covered 290 cases, including juvenile delinquency, parents' substance abuse, and child abuse, among other problems. Roughly half the cases involved juvenile delinquency, reflecting Swedish social policy that most delinquents under age 18 get their "social control" through the social bureau (social welfare office).
Cases of incest were directly referred for treatment to the psychiatric child guidance center [and thus, apparently, were not included in the demonstration project].
The original research proposal for the project was in the form of a straight experimental design, with all cases in Alby as the experimental group, and all cases in a comparable district serving as control. All cases in Alby were to be treated with a network approach, including network mapping, as close as possible to first contact. The effort to operationalize treatment with a network approach included intensified teaching of all social workers in Alby so that all would start with a common level of knowledge at the outset of the field study, mapping, and resource questions (e.g., Who in your network knows of the problem; What people in your network do you feel that you can turn to with the problem?). Practical and ethical issues emerged as the fieldwork began. Social workers believed that the mapping, which clients found quite stressful, came too early in the treatment process. Because so many of the cases involved issues of social control, with the government intervening into family life, workers and researchers recognized that ethical problems would emerge if the network approach were to be applied to every case handled by the Alby office. The project shifted from an experimental to an exploratory study, maintaining a strong emphasis on training social workers in the network approach, and supporting the use of resource questions and mapping whenever appropriate. The researchers shifted their stance from experimentation to participant observation, guiding the direction of their research on the basis of dialogue with the clients who were the subjects of the research.
The authors acknowledged several sources of theoretical and empirical knowledge guiding their project. They conceived the network as treatment system in the system-theoretical frame of reference developed by Bronfenbrenner (1979). The network is seen as the highest treatment level, including the microsystems of family, school, neighborhood, school, etc., and the mesosystem of contacts between the microsystems. Network intervention is often productive in opening up blocked channels of cooperation between agencies, i.e. activating the institutional mesosystem. To supplement their network maps, the authors utilized the geographical social support map developed by the Swedish sociologist Clarissa Kugelberg (1987). This map identifies the important support persons in the informant's life, and locates them as in the neighborhood, in another area of the municipality, another part of the country, or abroad. It appeared less anxiety-arousing than the network map, and was particularly useful for immigrants, for whom key supports were often far away. For a normative standard in evaluating network maps, they used Susan Phillips' (1981) study of network characteristics related to the wellbeing of normals. Briefly reviewing the standard network intervention literature, they made particular note of the Mount Tom emphasis on balance in assembling a network meeting (Halevy, et al., 1984).
Findings were reported on 61 cases, organized around a set of 8 questions:
Question One: How is the social network mobilized for the children and their parents involved in child welfare cases at the social bureau? The authors documented critical incidents and therapeutic interventions along a time line, starting with the opening of a case at the social welfare bureau. They found three types of patterns of activation or mobilization of the client networks: Initially high mobilization (roughly a quarter of the cases), initially low mobilization, building to a higher level later (roughly half), and low mobilization throughout the period of observation (roughly a quarter). Mobilizations tend to occur in "acute situations;" the authors found that network interventions were most productive at such times.
Cases with initially high mobilization could frequently be addressed with network meetings, ranging in size from less than ten to twenty participants. In about half the cases, the network meetings took place without an intervention team to lead them. These cases involved severe family crises, e.g., hospitalization of a parent, parental acute drug or alcohol abuse, death of a parent, intense child custody battles, or child abuse. With the encouragement of the social worker, it was often possible in these cases to involve the networks of the non-custodial parent (usually the father).
Cases with initially low mobilization began this way for a number of different reasons: a) The family voluntarily asked for assistance [and, presumably, was not in acute crisis]; b) the family wanted to avoid contact with the social bureau, and withdrew as soon as possible; c) neither the family nor social worker saw the problem as very serious at the time the case opened; and d) "uncomplicated" cases, e.g., shoplifting or truancy, which were handled in one or two interviews. Apparently in response to the initial mapping and resource questions, the clients (either at their initiative or with the active collaboration of the social worker) involved more people from different parts of their networks over the course of the intervention. This involvement could include participation in network meetings or in contacts with the clients outside of the social worker's presence. As an example, a noncustodial divorced father reported his concerns about his young children living with their mother, whom he described as abusing alcohol. Social workers met with each parent and gathered information from the pediatrician. The parents' conflict heated up, with acting out on both sides, until the father refused to return the children from a visit. The social workers had each parent draw maps of their children's networks, one for each parental household. The maps showed the parents how each each was cutting the children off from important social ties, as well as generating an inventory of the children's most important social relationships. The exercise produced a dramatic reduction of parental tension, and, at the close of the study, the parents were no longer battling over custody and had established a reasonably cooperative relationship.
Cases which sustained a consistently low level of network mobilization were, in many instances, the cases assigned to social workers who were critical of the network approach, preferring more traditional casework. Cases of juvenile delinquency often produced little network mobilization, reflecting the attitudes of family and social worker alike that, if the conduct problems did not escalate into more serious crime or substance abuse, there was not much to "stir up enough worry in the network." It was often difficult for social workers to gain access to families of children reported for delinquency. In general, social workers were unlikely to make extraordinary efforts to engage and work with families who did not make themselves accessible to the social bureau. Question Two: Which of the network methods (network mapping, step-by-step mobilization, small meetings, full-scale meetings, structured meetings) are applicable when treating what type of problem? This question turned out to be too ambitious for the data gathered. Efforts to divide cases into categories by symptoms created categories with too few members for analysis. The only category with a large enough set of members (parents' substance abuse) was not characterized by the use of a particular network intervention method. The authors then established categories based on the members of the family system who were the primary focus of the report to the child welfare office.
In cases where the primary focus of the report was the child, 90% of the cases were teenagers and the most frequently reported problems were substance abuse and/or criminality. Mapping was seldom used in this category. Typical were small network meetings, with the mesosystem of interaction between school and family most frequently addressed. Teenage gangs were often involved in meetings. Outcomes of these interventions tended to increase support between the child's parents, and to loosen gang ties. Cases in this category received the least amount of network intervention, perhaps reflecting families' and social workers' relatively high threshold for concern about delinquency. Within the service system, it is noteworthy that police reports on adolescent misconduct tended to arrive at the social bureau many weeks after cases came to police attention, leaving a cold trail and a family system less likely to be in crisis.
In cases with the primary focus on the parents, about half the reported problems involved alcohol abuse, with most of the rest of the problems evenly divided between psychiatric problems and custody conflicts. Many cases involved families with small children. Network mapping turned out to be a crucial intervention when dealing with disputes between divorced or separated parents, including custody disputes. As described in Question One, mapping was often used strategically to shift the focus from battles between the parents to their common interest in the child. Each parent would be asked to draw his/her perception of his/her perception of his/her child's social network. The parents would then sit down with the clinician and compare their maps, with pointed emphasis on what linkages remained intact for the child when s/he was with either parent. Absence of links between parental worlds was seen as problematic for the child, exacerbating loyalty conflicts. This would be particularly poignant for children of parents of different cultures/national origins. In about half the cases with a parent focus, network meetings were convened. The most common meeting was a small meeting, not led by a team, with two sectors of the network participating. Relatives were often strongly represented and quite active in these meetings, perhaps reflecting concerns in the extended family that the integrity of the family was threatened by the parents' symptoms. There tended to be more network intervention in these cases than in the child-focused cases, again perhaps reflecting more tolerance in the latter cases for the teenagers' delinquency.
Most of the cases involved reports focussing on both parents and children. Problems included family violence, child abuse, and psychiatric illness as the most frequently reported problems. The authors found these cases to most closely fit the label of "multiproblem family." Parents tended to show substance abuse, psychiatric problems, and violence; children tended to show psychiatric illness. The 25 families in this category generated 59 child welfare reports. Only six had children living with both parents; the rest lived in families with single parents or blended (step-) families. These families proved reluctant to produce network maps, perhaps reflecting their fear and distrust of the social bureau. When a family produced maps, there was more of a tendency for the members each to generate their own map, reflecting the support needs of individuals and their difficulties supporting each other. Given the distrust on the one hand, and the profound needs on the other, these cases either involved a great deal of network intervention or none at all. As a general summary of the findings on Question Two, the authors stated that the more problems cited in the original report and the social worker's assessment, the more network intervention had been done.
Questions Three and Four: Have the clients' networks changed during the treatment process, and, if so, how? The authors noted the limitations of their knowledge about actual changes, relying on social worker reports and questionnaires filled out by clients and social workers. They presumed that there were many positive and negative changes in client networks of which they were unaware. On the basis of what they did observe, they concluded that network changes did not occur when the client and social worker did not form a working alliance--because of client distrust, social worker's inability to find ways to reach the clients, or both. Four cases received second reports to the social bureau during the period of study. These cases all showed a low level of network change. The authors found "clear network change" in 87% of the cases, taking place mostly at the mesolevel, at which the interventions were primarily focused. The most common mesosystem level change was in the relationship between family and relatives. In about half the cases, there were changes in the family (microsystem) itself. In about a quarter of the cases, parents changed their work situation, created new social ties, or re- established old ones. In cases which yielded high levels of network change, the social worker had succeeded in establishing a flexible role within the client network, giving and gaining respect in ways which engaged and mobilized network members, who felt the worker understood their concerns. Different networks seemed to require different amounts of change for there to be impact on the presenting symptoms. In cases which showed the most network changes (six cases), there was symptom improvement in every case, with the symptom disappearing in five cases. In questionnaire responses, social workers and clients agreed that the interventions had resulted in network changes.
Questions Five and Six: How can the social network be mobilized to avoid placement of children? In the case of placement, how can the resources of the social network be used to minimize broken links between the child and the environment of origin (parents, siblings, relatives, friends, school, etc.)? There were a total of 18 cases in which placement would normally be chosen by the social bureau; 16 of them resulted in some form of placement outside of the home. (In the other two cases, conventional treatment intervention by the social worker or the child guidance clinic was sufficient to prevent placement.) The original 18 cases had smaller networks, with proportionally more broken linkages, than the other cases in the study. Seventy seven per cent of the 18 cases involved single-parent or blended families, in comparison to 54% of the total number of cases under study.
[Jodie Kliman notes an interesting twist on networks and placement, based on local press reports on problems of foster care with "crack babies" (children born to cocaine-abusing parents). When these children are placed within their parents' networks, they are more likely to be victimized by inappropriate, abusive contact with their drug-addicted parents. Some networks either participate in, or cannot protect children from, the toxicity of crack-addicted parents.]
Of the 16 placement cases, 7 had network meetings (4 with full-scale meetings with teams, 3 with one or more smaller, structured meetings). In 10 of the 16 cases, both children and their parents produced and discussed their network maps. In 11 cases, placements were arranged within the child's informal personal network. In 2 cases, the social network actually mobilized to make an institutional placement possible. For example, a 19 year-old drug addict was persuaded to participate in a drug treatment program by his formerly addicted brothers-in-law, who themselves had been helped by the program. In 3 cases, there was institutional placement with no success in mobilizing the network, either because of geographic splits in the network (e.g., a single African immigrant mother with kin overseas) or many breaks in important network linkages, accompanied by ongoing polarizations.
The authors made some comparisons between Alby social workers using the network approach, and social workers in other Botkyrka districts. In Alby, 86% of the placements were inside the primary personal networks, and 14% outside. In the rest of Botkyrka, 32% were placed inside the network, and 68 % outside. A systematic comparison was made of whom the social worker consulted before making a placement (family, relatives, friends and neighbors, school and daycare, government authorities). Compared to the other districts, Alby social workers were more likely to consult family and relatives, and less likely to consult authorities. The authors found that the Alby social workers who used the network approach had developed a "change of focus on their role." When interviewed, they stated that they saw their primary tasks as helping the child not to lose his/her "roots," and working with the relationships between the family of origin and the foster family. When a child was placed the resources of the natural network were always taken into account. The traditional foster care home tended to disappear. When placement in the network was impossible, the network was recruited to cooperate in the institutional placement. On occasion, network intervention resulted in a change of custody, with a disengaged father being re-enlisted and taking over the child's care. The authors stressed the importance of working through conflicts among family and relatives before placement within the network, in order to avoid disqualifying one or both parents, or taking sides in unresolved conflicts.
Question Seven - What were the changes in the social workers' experience resulting from the network perspective and network methods? Five cases were selected in which clients had both participated in network meetings and drawn their network maps. Students from the School of Social Work interviewed both the social workers and the clients in these cases. In addition, the authors interviewed social workers and clients during the project's pilot year.
Most of the social workers' responses were positive and enthusiastic about the network approach. They found the network intervention to be a "strong intervention," and emphasized the importance of establishing client confidence in the relationship with the social worker before embarking on mapping. The mapping process further improved the working relationship by emphasizing shared responsibility. The social worker's role as an authority tended to fade into the background, and the relationship between social worker and client became more equal. Social workers became less worried about child welfare cases with the network approach; they reported feeling "that everything doesn't depend on me." Social workers described client reactions to mapping as mostly positive, but found some clients to react with "puzzlement, fear, or suspicion." One client described to her social worker her pain at seeing in "black and white" that almost all her friends were alcoholic.
Social workers noted that, in some situations, the network approach can exacerbate conflicts with other authorities (school, police, etc.), which may expect speedy placement rather than cooperation with the informal placement. [Jodie Kliman and I noted similar problems of interagency strains with the community network approach (Trimble and Kliman, 1981).]
Social workers were particularly positive about the "process-oriented" network meetings in crisis situations. The authors were surprised to see how readily the social workers transcended their bureaucratic roles in such situations. Social workers reported that clients and others in their networks were generally satisfied with the meetings, "except for those who didn't get what they wanted."
On the topic of client resistance to the network approach, the social workers reported that they would respect clients' objections, and try to understand clients' hesitation or unwillingness. Asked how they would react if they, as clients, were asked to cooperate with network intervention, social workers imagined that they would be suspicious, and wondering about their own social worker's reasons for proposing network intervention. They imagined that confidence in their social worker would be very important. They felt that they would appreciate that their environment, and the people who were important to them, were taken into account.
Question Eight - What were the changes in the clients' experience as a result of the network perspective and network methods? Responses to the mapping were mostly positive, but the negative comments were striking. "A little ridiculous...too personal...it didn't matter to me, but I did it because it seemed important to my social worker." Positive comments noted the social worker's interest in the client's social situation, and in the question of how they would be able to get help and support. Most clients stated that the drawing of the map did not lead to new contacts or relationship changes; one said that she did get in touch with a friend whom she had not seen in a long time. Clients felt that they themselves already had a clear image of their networks before drawing their maps, but that mapping was helpful for their social worker's task.
Expectations prior to the network meetings were reported to have been positive (e.g., "to find out ways to make it better in the future") or negative (e.g., "My social worker proposed a meeting to be able to better check how bad my situation was."). All but one of the clients reported feeling exposed in the meetings ("everybody's eyes were directed at me...on the defendant's bench...mostly negative things came up). The one exception stated, "I did not take it so personally...Important things about my children were discussed." Clients who participated in more than one meeting were more positive about the later meetings. All five clients interviewed by the social work students said they thought that it had been a good experience to feel support from their closest friends and relatives at the meetings. Three clients felt that they had missed important people at the meetings. Two said they felt that there had been too many "unknown" authority persons present. One client complained that the meetings had tried to "push too much responsibility onto my relatives." Asked whether they would recommend that their friends, in similar situations, convene network meetings, clients responded that they would neither recommend nor advise against a network meeting. Clients did not report feeling that they had shared too much of their own personal thoughts and feelings at the meetings.
Reflecting on the clients' responses, the authors concluded that the clients' experiences were affected far more powerfully by the experience of the social bureau as a representative of the power of society intervening into the family in response to a child welfare report, than by the question of whether or not there was a network meeting. They found the study be particularly helpful in illuminating ethical questions about network intervention. They learned that the most productive stance toward crisis situation is to "take it seriously"--to avoid taking either a laissez-faire or a repressive stance.
References
Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA: Harvard University Press
Halevy, J., Hemley-van der Velden, E., Ruhf, L., & Schoenfeld, P. (1984). Process and strategy in network therapy. Family Process, 23, 521-523.
Kugelberg, C. (1987). Allt eller inget - Barn, omsorg forsvarsarbete. Stockholm: Carlssons forlag.
Phillips, S. (1981). Network characteristics related to the wellbeing of normals: A comparative base. Schizophrenia Bulletin,
Trimble, D., & Kliman, J. (1981) Community network therapy: Strenghtening the networks of chronic patients. International Journal of Family Psychiatry, 2, 269-289.
-David Duboff Participation in a self-help group as a network therapist.
[Editor's note: David Duboff, MSW, is a community organizer in Ann Arbor, Michigan, currently involved with a group addressing the permanent housing needs of the homeless mentally ill, and is a member of the Advisory Committee of the local community mental health center. David was a community organizer, mental health consumer, and leader of the Lynn Network Conference, a community organization which promoted network therapy in Lynn, Massachusetts, in the mid-1970's (The tale of the Lynn Network Conference, disguised as the "Ocean City" Network Conference, is recounted in Trimble, D., & Kliman, J. (1981). Community network therapy. International Journal of Family Psychiatry, 2, 269-289.). Following the examples set out in Ross Speck and Carolyn Attneave's pioneering Family Networks (1973), my network therapy practice in Lynn recruited people who participated in network therapy to become team members. David was both a team member for network assemblies and a member of the team which led meetings of "Survival Tribe," a therapy group which also served as a transitional network for its members. As a network therapy activist with the Network Conference, David participated in many discussions of the applications of the Speck and Attneave's (1973) spiral sequence model (retribalization, polarization, mobilization, depression, breakthrough, and exhaustion-elation) to community organization, community meetings, and other group phenomena. He contributes the following communication based on his use of the network therapist stance as a member and activist in self-help groups for mental health consumers.]
Therapy can be done in the context of self-help with the mentally ill. In order to do this, the therapist must be a mental health consumer and be a leader of one or more self-help groups. It is important that the therapist be a consumber so members can identify with him/her. This does not mean that the therapist does not play the role of an authority figure, like a traditional therapist. Authority derives from being in a leadership position and from being knowledgeable about mental health.
Of course, the nature of the interactions between therapist and members will be much less formal than in traditional therapy. The therapist has to adopt a flexible style that allows for mutual help. The people helped are not clients, but "members." The leader encourages people to help each other.
In order for this to work, the therapist must understand how to utilize networking in his/her work. A self-help group has network properties--multiple interconnections, stronger and weaker ties, and clusters (nexuses and plexuses). Lonely people need to be integrated into the group, which involves networking--people and small groups have to be introduced to the lonely person, and the leader has to work with the person individually. Sometimes, clusters become enmeshed and isolated from the whole group. These clusters have to be "loosened" through the integration of new members. Binds need to be loosened and bonds need to be strengthened--with the help of the whole group.
Different groups of consumers constitute a wider network. People know each other from after- care and hospital stays. These groups can be woven together through networking. For example, an anxiety support group and a group oriented towards socializing and providing moral/emotional support came to overlap over time (some of the same people in both groups). I played a role in orchestrating this weaving process. As the groups became interwoven, each group became strengthened with the introduction of new members. Social group members were able to do concentrated work on their anxiety-related problems, and anxiety group members were able to make new friends, attend holiday events, go on outings, etc.
The spiral model of group process described by Speck and Attneave (1973) in Family Networks works in a self-help group. I have seen it occur in one group on a continuous basis over a period of many years. The group may go through all the phases in the course of one meeting, or over several meetings, and, of course, the process is constantly occurring all the time as the group goes through phases based on whether members are more or less depressed, elated, anxious, or psychotic. One person's depression feeds into others'. Thus the entire group may enter a depression phase. As people struggle to break out of the depression, one member and then another succeeds. This can lead to a breakthrough for the entire group, which may occur at one meeting or over a long period of time. Having achieved a breakthrough, people are left exhausted or elated.
This whole process occurred, for example, before the group's annual holiday dinner. By the time of the dinner, people were ready for a retribalization, where they experienced the "network effect" of feeling good towards one another and having a group consciousness. Then, the process started over again. A polarization occurred against one member who had a lot of complaints. The process repeated itself, over months, as members worked to reintegrate this person into the group. The group had to incorporate some changes and established some new norms in the process.
Sometimes, a polarization can lead to the strenghtening of a norm. The group had always had a policy that no members get paid for work they do for the group. The President wanted to be paid $1 a day. This created a crisis. The group went through the whole spiral sequence very quickly in the process of deciding to stick with its norm. The President stepped down. A "revolutionary coalition" within the group had succeeded in staging a "coup." New people came into leadership positions.
For a number of years, I was active in the group, particularly during the four years I served as primary leader. Later, when I attended social work school, I stopped attending meetings, but continued to see several members on a regular basis. Thus, I kept in touch, and in a sense I worked through the people I saw to help the group resolve problems, particularly the depression that came over the group when I stopped attending meetings. My role was closer to that of a consultant, but I continued to serve as therapist as people came to me individually with their problems, and I continued to help loosen binds and strengthen bonds between people while not attending group meetings.
Thus, it can be seen that a therapist/member/leader can play a variety of roles. What it takes is flexibility, being conscious of the group process and what's best at a particular time, knowing when to "refer out" people to other therapists (particularly in a time when the group is very dense), and being aware of the use of self.
On one level, a self-help group is a particular kind of therapy group, where all the members, to varying degrees, serve as "therapists." One role of someone who has been trained in network therapy is to help other members understand group process and understand networks. I have shown some of my writing about therapy and networks to my mentally ill friends. As people become more conscious of what goes on in their groups, they can see the spiral sequence occurring and can help, intuitively, to push the group through stages, as I have done.
One group I worked with eventually dissolved. It became a network in which members saw whomever they wanted to, individually and in small groups. This is one of the aspects of networks -- groups exist when they serve a purpose, dissolve when they no longer serve a purpose for the particular members, and reemerge in new forms.
Reference
Speck, R., & Attneave, C. (1973). Family networks. New York: Pantheon.
-Thanks to Jodie Kliman, whose editing keeps NETLETTER intelligible and readable.
-Already planned for the next issue is a review of the English summary of Lennart Svedhem's (1991) Social networks and behaviour problems in school among children aged 11 to 13 years - A theoretical and empirical base for network therapy. There will also be a feature on current practice of network therapy, using the readers' summaries requested above.
-The Twelfth International Sunbelt Social Network Conference will be convened in San Diego, California, February 13-17, 1992. This is a gathering of social network analysts. Live Fyrand described her experience as a network therapist attending a Sunbelt Conference in NETLETTER, IV, (1), 1989. Network analysis is an important source of conceptual guidance and empirical information for network therapy; this year's conference looks particularly valuable because of Barry Wellman's workshop, Introduction to Network Analysis, which can provide the neophyte with a grounding in the basics of network analysis. The deadline for paper submission is November 15; send an abstract of no more than 200 words to the program coordinator, Phillip Bonacich, at Department of Sociology, University of California, Los Angeles, California. Telephone (213) 825-3017. To register, send a check for $40 payable to Sunbelt Social Network Conference, to Philip Bonacich at the above address.
-American Orthopsychiatric Association is struggling to remobilize itself.
They are convening their Annual Conference May 15-18, 1992, in New York City.
I plan to offer another panel for current practitioners to report on the status
of their projects. The deadline for program proposals is November 6; if you
are interested please call or fax Bertram J. Cohler at the University of Chicago,
telephone 312-753-3860, fax 312-955-9419.
USEFUL LINKS
This Site
was Originally Created and Maintained by Ransom D. Trimble 