The Open Dialogue Approach to Acute Psychosis: Its Poetics and Micropolitics
JAAKKO SEIKKULA, Ph.D. MARY E. OLSON, Ph.D.*
In Finland., a network-based, language approach to psychiatric care has emerged, called "Open Dialogue." It draws on Ba-khtin's dialogical principles (Bakhtin, 1984) and is rooted in a Batesonian tradition. Two levels of analysis, the poetics and the micropolitics, are presented. The poetics include three principles: "tolerance of uncertainty," "dialogism," and "polyphony in social networks." A treatment meeting shows how these poetics operate to generate a therapeutic dialogue. The micro-politics are the larger institutional practices that support this way of working and are part of Finnish Need-Adapted Treatment. Recent research suggests that
* Both authors have equally contributed to the article and should be regarded as first authors.
Jaakko Seikkula, Ph.D., is Senior Assistant in the Department of Psychology at the University of Jyvaskyla and Professor at the Institute of Community Medicine at the University of Troms0, Norway. Send requests for reprints to Jaakko Seikkula, Department of Psychology, University of Jyvaskyla, P.O.Box 35, FIN 40351, Jyvaskyla. Email: seikkula@psyka.jyu.fi
Mary Olson, Ph.D., is on the faculty of Smith College School for Social Work and a research fellow at its Center for Innovative Practice. In the fall of 2001, she was Fulbright Professor to Finland in the Department of Psychology at the University of Jyvaskyla.
A Fulbright scholar award to Finland supported the research for and preparation of this article. Grateful acknowledgment goes to the entire staff of about 100 professionals at Keropudas Hospital and local outpatient clinics.
Open Dialogue has improved outcomes for young people in a variety of acute, severe psychiatric crises, such as psychosis, as compared to treatment-as-usual settings. In a nonrandomized, 2-year follow up of first-episode schizophrenia, hospitalization decreased to approximately 19 days; neuroleptic medication was needed in 35% of cases; 82% had no, or only mild, psychotic symptoms remaining; and only 23 % were on disability allowance.
Fam Proc 42:403-418, 2003
In Finland, a network-based, language approach to psychiatric care, termed Open Dialogue, has been pioneered at Keropudas Hospital in Western Lapland. One of the authors (JS) worked as a member of the original team. Other team members who have been writing about this approach include Jukka Aaltonen, Birgitta Alakare, Jyrki Keranen, and Kauko Haarakangas (Haarakangas, 1997; Keranen, 1992; Seikkula, Alakare, & Aaltonen, 2001a). Recent studies suggest that this model has improved the therapy of people suffering from first-episode psychosis by significantly reducing the incidence of hospitalization, the rate of recidivism, and the use of medication (Seikkula, Alakare, & Aaltonen, 2001b). This approach has gained widespread recognition   in  Northern   Europe where
403
Family Process, Vol. 42, No. 3, 2003 © FPI, Inc.
404 /
family process
Seikkula, together with Norwegian psychiatrist Tom Andersen, have fostered an international network of teams using open dialogue and reflective processes in acute-care settings in Russia, Latvia, Lithuania, Estonia, Sweden, Finland, and Norway. Far less recognized in the United States, this model is worthy of closer examination as a form of crisis intervention in the care of the most severe psychiatric problems.
Within a postmodern, social constructionist framework, Open Dialogue integrates different psychotherapeutic traditions into their origins and evolution. Within the family field, however, its starting point was Milan systemic therapy. Beginning with an overview of communication-based approaches to psychosis, the first part of this article will sketch broadly the theoretical and clinical evolution away from systemic family therapy sessions to network-based practice. It will then go on to specify the language practices of Open Dialogue and present an interview in order to look closely at what actually happens, moment by moment, in a treatment meeting. Finally, the last part will consider the institutional and training contexts in which this approach is embedded, and present the results of a study.
Our inquiry into the open dialogue approach draws on the two categories described by the community psychiatrist Marcelo Pakman (2000). He identifies the "poetics" and the "micropolitics" of therapy. The term "poetics" refers to the language and communication practices in face-to-face encounters (Hoffman, 2002; Olson, 1995). In Open Dialogue, we can locate three principles: "tolerance of uncertainty," "dialogism," and "polyphony in social networks" (Seikkula et al., 2001a). These terms echo and transform the original Milan team's principles of hypothesizing, circularity, and neutrality as the guidelines for the conductor of the session
(Selvini-Palazzoli, Boscolo, Cecchin, & Prata, 1980).
The micropolitics, or larger institutional practices, of Open Dialogue also can provide an important focus for examination and contrast. Most forms of family therapy have been office models with strategies for larger systems, while Open Dialogue is a communal practice orga nized in social networks.1 It is embedded in the larger transformation of public psychiatric services in Finland associated with a reform called "Need-Adapted Treatment" (Alanen, 1997, Alanen, Lehti-nen, Lehtinen, et al., 2000). As Pakman and others recognize, there is an urgent need in the U.S. for new, expanded models of dialogue that can address not only the poetics of the interview room but also the larger bureaucratic politics that can constrain and deaden them. As professionals struggle with these issues in a chaotic, procedure-driven, managed-care environment in America (Coffey, Olson, & Sessions, 2001), the experience in Finland can offer a clear alternative in its network care of the severest problems.
COMMUNICATION APPROACHES TO PSYCHOSIS
An interest in psychosis and schizophrenia was prominent in the early days of the family field. The research project of Gregory Bateson and his colleagues culminated in the landmark article on the double bind, (Bateson, Jackson, Haley, & Weakland, 1956). The treatment of psychotic patients and their families was, in fact, one of the significant starting points for family therapy.
Bateson's concept of double-bind communication came from a theoretical attempt to imagine the kind of context to which psychotic speech and behavior
1 The term "communal perspective" appears in the writing of Lynn Hoffman (2000), based on the idea of communal practice proposed by Tom Andersen.
SEIKKULA and OLSON
/ 405
would seem adaptive (Weakland, 1960). Subsequent writings by Bateson and his colleagues (1962) revised the original formulation of the theory:
The most useful way to phrase double bind description is not in terms of binder and a victim but in terms of people caught up in an ongoing system which produces conflicting definitions of the relationship and consequent subjective distress (p. 42).
Instead of looking solely at patterns of message exchange, Bateson (1962) shifted to emphasizing the larger system of relations that generates these paradoxes.
In the decades following the end of the Bateson project, other research endeavors with families and their psychotic children were undertaken. However, none of these turned out to be as significant in terms of the development of an identifiable therapeutic model for psychosis as the work of the Milan team (Hoffman, 1981). Their research became the next major clinical contribution that focused on the problem of psychosis by using a communication approach.
Indebted to the double-bind theory, the Milan team invented what they called the systemic model for families with severely disturbed, psychotic and anorectic children (Selvini-Palazzoli, Boscolo, Cecchin, & Prata, 1978). The Italians introduced the technique of the counterparadox to untangle paradoxical communication. For instance, they would offer the family a new logic in the form of a positive connotation or a new ordering of behavior in the form of a ritual (Boscolo, Cecchin, Hoffman, & Penn, 1987).
These ideas had a radical effect on the family field in both the United States and Europe. Yet, only in a few countries has the Milan model produced any kind of lasting influence on the psychotherapeutic treatment of psychotic patients. In the United States, there has been instead the
rise of the psychoeducational family approach, which comes from a different tradition, at least in terms of its stance toward family members. (C. Anderson, Hog-arty, & Reiss, 1980; Falloon, 1996; Falloon, Boyd, & McGill, 1984; Goldstein, 1996; McGorry, Edwards, Mihalopoulos, et al., 1996). The point of convergence between Open Dialogue and the psychoeducational program is the idea that neither the patient or family are seen as either the cause of the ps3'chosis or an object of treatment, but as "competent or potentially competent partners in the recovery process" (Gleeson, Jackson, Stavely, & Burnett, 1999, p. 390). There are many other important differences in the theoretical assumptions of Open Dialogue and the widely used psychoeducational models for treating psychosis. For a discussion of the contrast, please refer to Seikkula, Alakare, & Aaltonen (2001a).
Finally, Michael White (1995) has applied his narrative practice of externalizing the problem to psychosis. This method lessens the hostile voices of people with chronic symptoms by placing them outside the person rather than seeing them as a manifestation of inner experience. Similarly, Open Dialogue is oriented toward the outer, social dialogue but makes a more formal use of the network. Outside of Northern Europe, family therapy has not used a network focus in dealing with acute psychosis.
THE EMERGENCE OF OPEN DIALOGUE
From Family Therapy to Network Orientation
The team at Keropudas Hospital were using the Milan model when they began seeing families in the 1980s, and yet, this attempt to do systemic therapy in a public system soon encountered new and unforeseen practical dilemmas. The first impetus for the shift away from systemic family therapy had to do with the difficulty of
Fam. Proc, Vol. 42, Fall, 2003
406 /
FAMILY PROCESS
engaging families at Keropudas Hospital in family therapy. In the early 1980s, only a small number of patients and their families followed through with their referrals. Ideas and practices began to change in the search for solutions to such concrete problems.
While elegant in theory, there have been many similar problems reported about the Milan method in practice, especially when translating it into other cultural systems and other settings outside the private institute. There also have been many reports by practitioners of discomfort and difficulty in connecting with families when working this way (Andersen, 1992, 1995; Hoffman, 1992, 2002; Lannamann, 1998). These experiences seem to emanate from the nature of this highly expert and abstract model. Using the metaphor of "the game," the original Milan method tends to position the family as an object of therapeutic action, rather than as a partner in the therapeutic process. Another important set of critiques and revisions has arisen from feminist and social-justice theorists and therapists who objected to the systemic stance of neutrality in situations of abuse and violence (Goldner, Penn, Scheinberg, & Walker, 1990; MacKinnon & Miller, 1987).
In 1984, at Keropudas Hospital, it was the recognition of the negative effect of a distant and objectifying view of the family within the assessment procedure that led to a reorganization of the way admissions were handled at Keropudas hospital. The hospital staff began to organize a treatment meeting in advance of any kind of therapy. A further motivation for this change came from the fact that, since the hospital is in part a state psychiatric system, the issues of equity and access for all patients were central ones. In the Finnish practice, all patients must be accepted, not only those referred for-and willing to do-family therapy. Alanen and his team in Turku developed the original idea of
treatment meetings as a part of the need-adapted approach (Alanen, 1997).
Over time, this kind of meeting evolved into the main therapeutic forum itself. First, constructivist ideas, and then the idea of dialogism by Bakhtin (1984), Vo-loshinov (1996) and Vygotsky (1970) assisted in understanding the new phenomena arising in the new practice of organizing open meetings in contrast to family therapy sessions. Andersen's (1987, 1990, 1992) invention of the reflecting team and the Galveston group's collaborative language systems approach (H. Anderson & Goolishian, 1988) became significant clinical frameworks in the further development of what is now being called Open Dialogue.
The Organization of the Treatment Meeting
Dedicated to giving immediate help in a crisis, the basic format of the Open Dialogue is the treatment meeting, which occurs within twenty-four hours of the initial contact. It is organized by a mobile crisis team composed of outpatient and inpatient staff and takes place, if possible, at the family home. It brings together the person in acute distress with the team and all other important persons (i.e., relatives, friends, and other professionals) connected to the situation. The meeting takes place physically in an open forum as well, with everyone sitting in the same room, in a circle.
The responsibility for mobilizing the team and arranging for the meeting rests with the professional first contacted by the family. Those team members, who have taken the initiative to organize the meeting, take responsibility for conducting the dialogue. Either all the team members can participate in interviewing, or it can be decided beforehand that there will be a specific person asking the questions and facilitating a dialogue among the others in the room. The constellation
SEIKKULA and OLSON
/ 407
of the team varies according to the specific situation and the previous treatment history of the family, with all prior therapists invited to these meetings.
All decisions about ongoing therapy, medication, and hospitalization are discussed and made while everyone is present. There are no separate staff meetings for treatment planning. It is more advisable to focus on these treatment issues later in the meeting, after the family members have had a chance to express their concerns. The outcome of the meeting should be summarized at the end, especially the decisions that were made; if not, it should be stated that nothing was decided. The length of the meetings can vary, but a meeting of 1.5 hours often provides enough time.
Whether the patient is hospitalized or not, the same team remains involved and continues to meet with the person and the network-in some cases, over a short period of time, and in others, a much longer length—until the urgent situation and the symptoms dissolve. This idea of "psychological continuity," the sustained involvement of the same team over time, is critical to this approach. The team stays connected with the family until it is clear that people are out of danger.
The establishment of the treatment meeting altered the Milan-style practice of using a long interval between sessions and began to reveal the role of speech and language in the psychotic crisis. Crisis intervention generally was not part of the Milan model. In fact, the Italians viewed the report of a crisis as a "move" in the "family game," with the team's response strategically conceived to challenge it (Selvini-Palazzoli, et al., 1978). Thus, the departure in seeing families in crisis on a daily basis and working intensively with them is another important shift away from the Milan method.
Notwithstanding such differences, this first step toward Open Dialogue-the es-
tablishment of the treatment meeting-may have been prompted by the capacity to evolve that seems to be built into the Milan method itself. As Lynn Hoffman says, "Less like a set of procedures than a 'learning to learn' model," the systemic approach taught professionals to think re-flexively and transform their own premises and conduct in the face of impasses and difficulties (Boscolo et al., 1987, p. 28).
In the early eighties, Boscolo and Cec-chin became increasingly inspired by the work of cybernetic researchers von Foer-ster, Varela, and Maturana, who proposed the notion of a second-order cybernetic view. They emphasized that we cannot speak of a separate, observed system but only of an "observing system" that takes into account the lens of the observer. Thus, any encounter with a family is, in part, a creation of the ideas that professionals bring to their work. The seeds of this second-order shift were present at the end of the complex career of the original Milan team, notably in their article on circular questioning (Selvini-Palazzoli et al., 1980). The article emphasized the process of interviewing, rather than the characteristics of the family, and anticipated the linguistic turn the field would take by focusing on the conversational method rather than on the intervention.
The initial transformation in the Finnish team, which ushered in many other changes, was consistent with this second-order thinking and was started when the team altered its relationship with the family by treating everyone involved as a member of a partnership. The use of the cybernetic analogy has since dropped away, together with the metaphors of vision and observation. These were replaced, first by metaphors of voice and listening, and then by those related to sensing and touch (Hoffman, 2002). Above all, the idea persists that therapy
Fam. Proc, Vol. 42, Fall, 2003
408 /
FAMILY PROCESS
is conceived of as a process created jointly, with a deliberate emphasis on the spoken exchange and the circles of dialogue. Open Dialogue has retained other important ideas from the Bateson-Milan tradition, including the comunicational stance and the emphasis on affirmation, despite the dropping away of the positive connotation per se.
THE POETICS OF OPEN DIALOGUE
Tolerance of Uncertainty
The language practices of the treatment meeting in Open Dialogue have become quite distinct from those of any other form of network-centered therapy. As stated above, the foundation of the interview rests on the principles of "tolerance of uncertainty," "dialogism," and "polyphony." Here we will consider each independently, although they recursively work together. Tolerance of uncertainty is the counterpart to, in fact, the opposite of, the systemic use of hypothesizing or any other kind of assessment tool.
In practice, tolerance of uncertainty is constituted by frequent meetings and by the quality of the dialogue. It is important that meetings are held often enough, daily if necessary, that the family does not feel alone in the crisis. The team carefully monitors the scheduling of meetings and commonly includes the possibility of meeting daily for 10-12 days following the onset of a serious crisis.
Furthermore, uncertainty can be tolerated only if therapy is experienced as safe. Every severe crisis requires that the therapists and the family, for a period of time, manage the inherent ambiguities of the crisis situation, to which the dialogue, hopefully, provides Ariadne's thread. As part of setting these conditions, there is great attention paid to establishing a trustworthy therapeutic context, or "scene," so that the anxieties and fears stemming from the crisis can be mediated
and contained. Safety is established initially by hearing and responding to every person's voice and point of view, thus legitimizing each participant. If this kind of tolerance is constructed, there emerge more possibilities for the psychological (or what we might now call "dialogical") resources of the family and the patient, who thereby become agents who previously did not have a language to express their experience of difficult events.
As part of this approach, the question that a crisis poses, "What shall we do?" is kept open until the collective dialogue itself produces a response or dissolves the need for action. Immediate advice, rapid conclusions, and traditional interventions make it less likely that safety and trust will be established, or that a genuine resolution to a psychotic crisis will occur. Hypotheses are particularly avoided, because they can be silencing, and interfere with the possibility of finding a natural way to defuse the crisis (Andersen, 1990). The therapists therefore enter without a preliminary definition of the problem in the hope that the dialogue itself will bring forward new ideas and stories.
Finally, tolerance of uncertainty is different from, although reminiscent of, the "not-knowing" position proposed by Anderson and Goolishian (1992). The Galveston group defines a way of knowing where the client is the expert and the professional is the learner. The Finnish approach defines a way of being with others and with one's self that is a slightly different way of knowing. It is what Rilke (1984) meant when he wrote, "live your way into the answer" (p. 42).
Dialogism
Interwoven with enduring uncertainty is Bakhtin's (1984) idea of dialogue as the framework for communication among the team, the person, and the social network. In addition to constituting a network, this way of working engages in an effort to
SEIKKULA and OLSON
/ 409
reduce isolation by constituting a dialogue built around a communicative relationship with the patient and the persons involved with him or her. From a social constructionist perspective, psychosis is a temporary, radical, and terrifying alienation from shared, communicative practices: a "no-man's land" where unbearable experience has no words and, thus, the patient has no voice and no genuine agency (Holma, 1999; Seikkula, 2002). The therapeutic aim is to develop a common verbal language for the experiences that otherwise remain embodied within the person's psychotic speech and private, inner voices and hallucinatory signs.
The Bakhtinian idea of dialogue and its adaptation to the psychotic situation derive from a tradition that sees language and communication as primarily constitutive of social reality. Constructing words and establishing symbolic communication is a voice-making, identity-making, agen-tic activity occurring jointly "between people" (Gergen, 1999). The crisis becomes the opportunity to make and remake the fabric of stories, identities, and relationships that construct the self and a social world.
Open Dialogue thus translates Bakh-tin's concept of dialogism into a co-evolving process of listening and understanding. So described, it is consistent with what the French philosopher Jean-Francois Lyotard, leaning on Wittgenstein's concept of language games, calls the "game without an author," in contrast to the "game of speculation" of Western philosophy and debate. Lyotard describes the "game of audition" as a "game of the just" in which the "important thing is to listen," and when speaking, "one speaks as a listener" (Hoffman, 2000).
Seen this way, the idea of listening is more important in Open Dialogue than the process of interviewing. For this reason, the first questions in a treatment meeting are as open as possible to give
maximum opportunity for the family members and for the rest of the social network to be able to speak about whatever issues are most relevant to them at that moment. The team does not decide the themes in advance.
To generate dialogue from the very beginning, one of the tasks of the interview-er(s) is to "answer" what the patient or others have said. However, the answers usually take the form of further questions that are based on a previous utterance of the patient. From a Bakhtinian perspective, every spoken statement, or utterance, requires a reply. There is an aesthetic (a fitting together of utterance and reply) to the dialogue, that makes it "dia-logical," rather than "monological," which would be a speaker without a contributing listener (Volshinov, 1996).
In describing his term "heteroglossia," Bakhtin says that meaning is not fixed and intrinsic, although words carry traces and fragments of meanings from our diverse linguistic heritage. Since meaning occurs only in an ongoing exchange, the speaker and listener are intimately joined together in making sense of the psychotic episode. The therapeutic process requires creative participation in language that attends not only to what people say, but also to the existing feelings and sensuous responses that flow between them. Within the dialogical borderland where the person, the important others, and the professionals meet, a language for suffering may be born that can give the suffering a voice.
Polyphony
In Open Dialogue, there is no object—no structure or game-to be changed by therapy. Instead, there are multiple subjects, forming a polyphony of multiple voices. It was Anderson and Goolishian (1988) who first proposed the linguistic paradigm challenging the notion of a relational structure or communicational system ex-
Fam. Proc, Vol. 42, Fall, 2003
410 /
FAMILY PROCESS
isting within the family. White's (1995) post-structural approach takes a similar position.
The team no longer focuses on the family structure, but instead, on all the individuals involved. This means that the "system" is being created in every new dialogue, where the conversation itself constructs the reality, not the family "rules" or structure. Unlike the systemic approach, which focuses on intervening to change the system, the dialogic approach is designed to create a shared language that permits the meanings of the person's suffering to become more lucid within the immediate network.
As a result, Open Dialogue allows every person to enter the conversation in his or her own way. It is usual for the interviewer to begin with the person who asked for the meeting and then move on to other people, drawing out their concerns. Questions may be asked to assist in giving voice, such as "When did you become concerned about your son?". Most importantly, the interviewer pays meticulous attention to the communications of the person in distress, whose words and meanings form the focus of the dialogue. In contrast to the systemic use of circular questioning, the dialogical emphasis is on generating multiple expressions, with no attempt to uncover a particular truth
An important rule is that everyone present has the right to comment. The questions or reflections of the professionals should not interrupt the ongoing dialogue unless what they say fits in with the ongoing theme. They can comment either by asking another question related to the theme or by starting a reflecting dialogue about it with the other professionals (Andersen, 1995). An alternation between talking and listening in the reflecting process generates new opportunities for the patient and family to reconstrue their experience (Andersen, 1995; Seikkula, Aal-tonen, Alakare, et al., 1995).
Although influenced by the reflecting team idea, Open Dialogue is a less structured and more spontaneous kind of discussion. Reflections among the various professionals, who may have worked together in the same setting for years, occur in an impromptu manner, often during the most stressful or difficult moments. The reflections tend to promote a sense of emotional reassurance and help to create a story out of the person's psychotic communication.
When differences arise, the hope is to give all voices room to exist and thus encourage listening and exchange, rather than polarized, right-or-wrong thinking. This does not mean that everyone has to accept all points of view; people can disagree. Positive changes can take place simply from the airing of different perspectives in a safe climate. The goal is to generate joint understanding, rather than striving for consensus. Every effort is made to talk about any major issue concerning the patient or family only in their presence, including responses to the meeting itself. Therefore, there is minimal post-meeting review.
Thus, although rooted in the Milan tradition, Open Dialogue provides an important and well-developed example of the postmodern paradigm (Andersen, 1995; H. Anderson, 1997; Anderson & Goolishian, 1992; Hoffman, 2002; Penn, 2001). In accordance with Derrida (1971), there exists no "essence rigorously independent of that which transports it" (p. 229). In other words, there is no conception of truth or reality that can be known as separate from and outside of human expression. The therapeutic ingredient comes from the effect of dialogism on a social network as new words and stories enter the common discourse. To accomplish this, the language practices of the treatment meeting have the double purpose of holding people long enough (tolerance of uncertainty) so that the inexpressible can be given voice (dialogism) with the
SEIKKULA and OLSON
help of the important others in the network (polyphony).
THE STORY OF PEKKA AND MAIJA
The following dialogue is exceptional in the sense that the psychotic symptoms of the man, Pekka, disappeared in the course of this interview, and in the 7 years since this meeting, they have not recurred. This kind of result is not typical of the average case, where a psychotic crisis can be expected to last 2-3 years. However, this care is illustrative of the therapeutic process, in which words are jointly constructed for not-yet-spoken experiences. It is impossible to predict how long this will take. In some cases, as we see here, it can occur in the first meeting, while in most cases more conversations are needed.
A primary care physician met with Pekka, a 30-year-old married man who had worked in a hardware store. Pekka said that he was the victim of a systematic intrigue, and the men who were involved in this conspiracy were hunting for him. The physician contacted the admissions team at the psychiatric hospital, and a treatment meeting was set up. Present were Pekka, his wife Maija, the primary care doctor (D), a psychologist (Psych) and three nurses.2 The team met a tall and strong man with a much smaller wife. It was the wife who led them to the room, and they sat down next to each other. In the early phase of the meeting Pekka was speaking and Maija sat silent, but was looking at her husband, who, every now and then, looked back at his wife to see if she approved of his account.
When the team first attempted to interview Pekka, his speech was psychotic and incoherent, and it was impossible to understand him. For the first half-hour, the interview skipped from theme to theme, with no joint development of any topic.
2 Characteristics of the ease have been changed to make identification impossible.
/ 411
This situation finally changed when one of the nurses asked Pekka's wife about her concerns. This question initiated the beginnings of a dialogue where Pekka's psychotic speech started to shift.
Maija: Well, Pekka has been seeing things. He has been suspicious of everyone.
Pekka:   [Yes, and. .. .3
Maija: From my point of view, they all are a bit irritated with him.
Pekka: [. . .and I was saying that I will not. . .
Maija: And if one says something of the future. . ..
Pekka: [. . . .Yes, she is quite nervous, although. . . .
Maija: . . .the same kind of situation presented itself eight years ago.
Pekka:   [ It was quite a hassle
Maija:   He was even afraid of his father;
that his father would try to kill him.
Psych: How did it pass? Did you get any treatment?
Maija:   No, he has not had any treatment.
I do not even myself remember how it passed, perhaps it only tapered off. . .
Maija started to give a coherent description and offer details that made it possible for the team members to acquire some understanding of the situation. During this part of the interview, however, Maija and Pekka spoke simultaneously and thus they entered into the conversation polyphonically. The team did not try to structure this conversation by making each of them speak in turn. Instead, the professionals accepted this couple's style and the way they chose to engage in the conversation. After this initial exchange, Pekka started speaking more lucidly, in
3 [ - mark means the speech is being spoken simultaneously with the other speaker.
Fam. Proc, Vol. 42, Fall, 2003
412 /
FAMILY PROCESS
contrast to his earlier statements where the sentences and thoughts came out in a disorganized way. This glimpse of clarity signaled the beginning of a joint language for speaking about the situation.
Forty minutes into the meeting, Maija and Pekka began to describe events leading up to the onset of the psychosis. They painted in words a visual and moving picture of what happened, creating a narrative of experiences that previously had existed as scraps of speech without context. This shift took place within a conversation where the interviewer elicited a careful, slow-motion description of the events leading up to this crisis. When, as a result of the conversation, Pekka was able to put words to his experience, his psychotic expressions abated.
Maija and Pekka agreed that the point at which the psychotic symptoms began was on a Friday. The psychologist began to elicit more details about what happened on that Friday. Pekka explained that the holidays were coming, and because he was now out of work, he had no money for gifts. His former employer owed him bonus money that he should have paid Pekka. He was in emotional agony over the dilemma. Asking for the money might mean jeopardizing his friendship with his former employer, but not asking for it meant Pekka could not be a father to his family by buying gifts for them at Christmas. Despite his deep anxieties, Pekka decided to call his employer and ask for the bonus. When he did, his employer responded badly, accusing Pekka of blackmail. During this terrible conversation, there was, by chance, an electrical blackout in the area, and all the lights went out. Here is part of the description of the interchange:
Psych:   Yes, and Ray (the employer) said that you are blackmailing him?
Pekka:   Yes, arid . . .
D: And that was the end of the call?
Pekka: No, it was not at that point. It was when I said that "I am not blackmailing you, of course, but if, in any way, it could be possible, because there is a need for Christmas money."
Psych: Did he promise to do it, during the phone call?
Pekka: He said, "Yes, I will take a look at it." And at that moment the electricity went out. And it really was a terrible hassle. The computer, the electricity was fluttering. . . I felt that in some way he would make contact with me.
Psych:   Did that trouble you?
Pekka: Well, I was thinking that he really got startled. . .
Maija:   When the lights went out.
Pekka: I took it as a kind of sign that the blackmail was working . . .
At this point, events previously unsto-ried began to be told. It seemed as if Pekka was in a prison of conflicting injunctions about which he could not comment, nor could he escape the dilemma. He interpreted the terrifying coincidence of the electrical blackout within the frame of his entrapment. The team began to see that Pekka's paranoia was a culmination of many months of living in extreme tension, because he had no money. The team then encouraged the couple to continue to give more details of the sequence of events. As they did so, they assisted in deconstructing the psychosis further by talking about the emotions that overwhelmed Pekka during the onset of his symptoms. The interviewer had the impression that Pekka was re-living during the meeting the terror, which he may have felt when he initially started hallucinating. In order to give words to Pekka's emotional fears, the interviewer asked Pekka his first thought following the blackout:
SEIKKULA and OLSON
/ 413
Psych: It sounds like you were scared to death?
Pekka: Well, it wasn't that bad. But I was thinking that it would be better to leave the place. One never knew, when Ray could be so aggressive and so quick to argue, that how would you ever know what he would do. . .
Psych:    What was your first thought. . .
Pekka: . . . .that if he is coming. . . how in the world could you stop him if he is coming
D: He is coming to find you
Pekka:   Yes, he will come
D: Come and kill you, was that the
case?
Pekka: Well that is the, that is the. . .that is, of course, the worst thing that he could do. . .
To define Pekka's emotional experience, the interviewers used strong words: "He is coming to kill you, was that the case?" This statement from the team gave new, clear, and concrete expression to Pekka's fear in a way that he immediately accepted. The sense of safety and trust in the meeting and the connection between Pekka and the team was well established enough to allow the exchange to address Pekka's most dangerous fears. This interaction exemplifies the dialogical consequences of tolerating uncertainty.
At this point, the team reflects with each other. Reflections in Open Dialogue tend to occur when people are talking about the most terrifying elements of experience and are in danger of becoming disorganized. In the reflections, the team uses a stance of logical connotation. The term logical connotation is more apt than positive connotation, which places symptoms in the service of a beneficial premise or myth (Boscolo et al., 1987). Logical connotation describes how problematic experience or behavior makes sense in a par-
ticular context. Further, the reflecting conversation among the team members draws on dialogical principles. The focus here is not on creating an intervention but on creating a language for the couple's experience that reconstitutes voice and agency. It is the task of the team to search actively for new understandings of the problem. The themes developed by the team borrow from—and build on-the words used by Pekka and Maija.
Psych: If you could wait a moment, so that we can discuss among ourselves. What thoughts are we each having? What did this arouse in your mind?
D: Well, I, at least, started to think as
Pekka spoke that he is the kind of man who takes care of other people's concerns more than he does his own.
Pekka:   It is a little bit. . .
Psych:    More than himself?
D: Yes, more his neighbor's concerns
than his own.
Psych: That when Pekka asked for his end-of-year money from Ray, he started to worry how Ray would feel about it. . .
D: Yes.
Psych: He is more worried about what Ray thinks than the fact that this money belonged to him.
D: Yes, and I also started to think,
how difficult the situation was. . . . I am wondering if Pekka is the kind of man who finds it difficult to fight for his rights and go after what belongs to him [. . .]. I am also thinking if Pekka always describes things in such a detailed way, as he has done here. Or is this a sign of need and fear? Or does he want us to understand some issue in more detail? He explained so thoroughly what is difficult to understand, what was difficult to see.
Psych: Well, one could think that if one does not understand what took
Fam. Proc, Vol. 42, Fall, 2003
414 /
FAMILY PROCESS
place, it is a reason tor explaining very exactly what was happening. "What did it mean?" and "What made me think that. . .'?" In a way, the whole has disappeared and for that reason, one has to seek out the details to understand what it means.
D: And the things that are apparent
and the reasons that are given can also be a sign that the whole has disappeared, that one does not exactly know what things mean. . .
Psych: [referring to an earlier part of the meeting when Pekka explained that the TV was relaying private messages to himj Yes, it can no longer he possible to distinguish what is important and what is not. It is awful that one watches TV programs, having in mind that there are things that mean something only to me, although the programs are done somewhere in America.
D: ... many years ago.
In this dialogical way, the team members reflected on the incidents that Pekka and Maija had described. At the end of the meeting, the interviewer went back to some of the preliminary incidents to clarify if Pekka still had psychotic ideas about the electrical blackout and his ex-employer's reactions. The psychologist asked if Pekka thought that those things were coincidences, and he answered that he now thought that they were. The team agreed that if Pekka now thought that there were no magical powers affecting his relationship with his former employer, then he was no longer psychotic. In this shift, he seemed to show a new sense of personal agency, in contrast to being controlled by a destin3'.
In this example of an open dialogue, a language evolved to describe the terrible paradox that Pekka experienced in relation to his family and his employer. It is possible to view this psychotic situation from the perspective of double-bind theory and to notice how being able to name
and comment on this experience gives freedom from its captivity. Yet, the concept of the double bind has been abandoned, because it tends to suggest "an out-there reality" to be changed rather a "dialogical conversation" that can construct a path out of the psychotic world. From this point of view, the treatment meeting can be defined as a place where the words needed to talk about difficult things can be found within the back-and-forth movement of the conversational loom.
MICRO-POLITICS OF OPEN DIALOGUE
The effectiveness of the open dialogue approach is linked inextricably to its institutional and training contexts. Since 1984, at Keropudas Hospital, the approach has undergone systemic development, and the treatment meeting has been institutionalized as the standard admissions format. For the entire staff-including psychiatrists, psychologists, nurses, social workers-there is an ongoing, 3-year family therapy training program. These skills are taught democratically, under the assumption that any professionally trained person can acquire them. This democratic ethic in regard to the training is part of a larger ethic of participation and humility within the therapeutic culture of Keropudas.
In Western Lapland, a national healthcare program makes it possible for professionals to work in teams, in contrast to the U.S., where managed care's fee-for-service model has undermined them (Coffey et al., 2001). The team-based treatment meeting appears to have had a major beneficial effect on Keropudas hospital as a whole in that it mandates participation by inpatient staff in the community-based crisis teams, and that of outpatient staff in meetings on the wards. This kind of teamwork reduces the calcification of mental health perspectives as staff take on different positions within the hospital system.
SEIKKULA and OLSON
/ 415
Seen in this way, the open dialogue approach is not a model that is applied but a set of practices that are established throughout the hospital. As a result, there is integration with other forms of psychotherapy, especially individual therapy, but also traditional family therapy, art therapy, occupational therapy, and other kinds of rehabilitation services. The dialogical model organizes not only the treatment context but also the professional context. For this reason, the original team has not encountered the same short life-span of many other systemic teams whose host institutions have reacted to their presence with resistance and extrusion (Boscolo et al., 1987).
Despite all of these effective innovations, problems and failures in treatment still occur. This approach commits the team to work with the family during the failures and therefore to share the disappointments. The specific challenge for Open Dialogue, however, seems to be the administrative and practical problems of keeping teams together during the entire course of treatment, thus guaranteeing that there is psychological continuity for the person and network in crisis.
Statistics of Outcomes
A final observation about the micropoli-tics of Open Dialogue is the use of research and outcome data. It is critical in an environment that is dominated by the discourse of evidence-based practice to document outcomes. Open Dialogue is one of the most studied approaches to severe psychiatric crisis in Finland. Since 1988, there have been studies of treatment outcome and qualitative studies analyzing the development of the dialogue itself in the meeting (Haarakangas, 1997; Ker-anen, 1992; Seikkula, 1994; 2002; Seikkula et al., 2001a, b). Since this new approach was institutionalized, the incidence of new cases of schizophrenia in Western Lapland has declined (Aaltonen,
Seikkula, Alakare et al., 1997). Further, the appearance of new chronic schizophrenia patients at the psychiatric hospital has ceased (Tuori, 1994).
In an ongoing, quasi-experimental study of first-episode psychotic patients, Western Lapland was part of a Finnish national API (Integrated Treatment of Acute Psychosis) multicenter project, conducted by the Universities of Jyvaskyla and Turku together with STAKES (State Center for Development and Research in Social and Health Care) (Lehtinen, Aaltonen, Koffert, et al., 2000). The inclusion period for all non-affective psychotic patients (DSM-III-R) in the province was April 1992 through March 1997. As one of three research centers, Western Lapland had the task of starting treatment without beginning neuroleptic medication at the same time. This was compared to three other research centers, which used medication in a standard way, most often at the very beginning of the treatment. In Western Lapland, 58% of the patients studied were diagnosed with schizophrenia (Seikkula et al., 2001b).
In the comparison of the schizophrenia patients who participated in Open Dialogue versus those who had treatment as usual, the process of the treatment and the outcomes differed significantly. The Open Dialogue patients were hospitalized less frequently, and 35% of these patients required neuroleptic medication, in contrast to 100% of the patients in the comparison group. At the two-year follow up, 82% had no, or only mild non-visible psychotic symptoms compared to 50% in the comparison group. Patients in the Western Lapland site had better employment status, with 23% living on disability allowance compared to 57% in the comparison group. Relapses occurred in 24% of the Open Dialogue cases compared to 71% in the comparison group (Seikkula, Alakare, Aaltonen, et al., in press). A possible reason for these relatively good prog-
Fam. Proc, Vol. 42, Fall, 2003
416 /
FAMILY PROCESS
noses was the shortening of the duration of untreated psychosis to 3.6 months in Western Lapland, where the network-centered system has emphasized immediate attention to acute disturbances before they become hardened into chronics (Seikkula et al., 2001b).
In sum, it is important to see the open dialogue approach as the transformation of an entire psychiatric system, accompanied by administrative support, engagement with primary care physicians and psychiatrists, access to training, and ongoing outcome studies. The poetics of the interview are consistent with and reinforced by the micropolitics of the professional environment.
CONCLUSION
Gregory Bateson (1962) wrote that in relation to the double bind, "if this pathology can be warded off or resisted, the total experience may promote creativity" (p. 242). The open dialogue approach is a way of resisting the experience of "pathology." It builds instead a "transformative dialogue" within a social network (Gergen & MacNamee, 2000). While failure remains a daily occurrence when working with the severest psychiatric problems, the open dialogue approach offers new promise for many to find their way out of the labyrinth.
In many parts of America, the public mental health care system is in serious trouble. A recent report by the Surgeon General states that 80% of children and families who need mental health services do not receive appropriate mental health care (U.S. Public Health Service, 2000). Reports on children and adults "stuck" in hospitals have appeared in the national media, and there are lawsuits to remedy this situation in several states (Goldberg, 2001, July 9). Some managed care strategies have encouraged a decontextualized biological model that neither saves costs nor provides effective therapy.
At the same time, growing evidence suggests that community models such as the open dialogue approach can produce ethical and cost-efficient treatment. Inspired by Bakhtin's dialogical principles and other postmodern ideas, this way of working has humanized and improved the care of young people in acute severe crises, such as psychosis. The principles of Open Dialogue may be adapted to a variety of other severe difficulties. The idea of network therapy originally came from the United States but managed care has limited its applicability there. As we face the current crisis, perhaps it is useful to recall the "road not taken" and to take seriously the promise of the open dialogue approach.
REFERENCES
Aaltonen, J., Seikkula, J., Alakare, B., Haara-kangas, K., Keranen, J., & Sutela, M. (1997). Western Lapland project: A comprehensive family- and network-centered community psychiatric project. ISPS. Abstracts and lectures 12-16, October 1997. London.
Alanen, Y. (1997). Schizophrenia: Its origins and need-adapted treatment. London: Kar-nac Books.
Alanen, Y., Lehtinen, V., Lehtinen, K., Aaltonen, J., & Rakkolainen, V. (2000). The Finnish integrated model for early treat ment of schizophrenia and related psychosis. In. B. Martindale, A. Bateman, M. Crowe, & F. Margiseon (Eds.), Psychosis: Psychological approaches and their effectiveness (pp.235-265). London: Gaskell.
American Psychiatric Association. (1987). Diagnostic and statistical manual, of mental, disorders (3rd ed., rev.). Washington, DC: Author.
Andersen, T. (1987). The reflecting team: Dialogue and meta-dialogue in clinical work. Family Process, 26, 415-428.
Andersen, T. (1990). The reflecting team,: Dialogues and dialogues about dialogues. New York: Norton.
Andersen, T. (1992). Reflections on reflecting with families. In S. MacNamee & K. Gergen
SEIKKULA and OLSON
/ 417
(Eds.), Therapy as social construction (pp. 54-68). London: Sage.
Andersen, T. (1995). Reflecting processes: Acts of informing and forming. In S. Friedman (Ed.), The reflective team in action: Collaborative practice in family therapy (pp. 11 35). New York: Guilford Press.
Anderson, C, Hogarty, G., & Reiss, D. (1980). Family treatment of adult schizophrenic patients: A psycho-educational approach. Schizophrenia Bulletin, 6, 490—505.
Anderson, H. (1997). Conversation, language, and possibilities. New York: Basic Books.
Anderson, H., & Goolishian, H. A. (1988). Human systems as linguistic systems: Preliminary and evolving ideas about the implications for clinical theory. Family Process, 27, 371-393.
Anderson, H., & Goolishian, H. (1992). The client is the expert: A not-knowing approach to therapy. In S. MacNamee & K. Gergen (Eds.), Therapy as social construction (pp. 54-68). London: Sage.
Bakhtin, M. (1984). Problems of Dostojevskifs poetics. Theory and history of literature: Vol. 8. Manchester, UK: Manchester University Press.
Bateson, G. (1962). A note on the double bind. In C. Sluzki & D. Ransom (Eds), Double bind: The foundation of the communica-tional approach to the family (pp.39—42). New York: Grune & Stratton.
Bateson, G., Jackson, D., Haley, J., & Weak-land, J. (1956). Toward a theory of schizophrenia. In C. Sluzki & D. Ransom (Eds.), Double bind: The foundation of the commu-nicational approach to the family (pp.3-22). New York: Grune & Stratton.
Boscolo, L., Cecchin, G., Hoffman, L., & Penn, P. (1987). Milan systemic family therapy: Conversations in theory and practice. New York: Basic Books.
Coffey, E. P., Olson, M.E., & Sessions, P. (2001). The heart of the matter: An essay about the effects of managed care on family therapy with children. Family Process, 40, 385-399.
Derrida, J. (1971). White mythology: Metaphor in the text of philosophy. In A. Bass (Trans.), Margins of philosophy (pp. 207-271). Chicago: University of Chicago Press.
Falloon, I. (1996). Early detection and inter-
vention for initial episodes of schizophrenia. Schizophrenia Bulletin, 22, 271—283. Falloon, J., Boyd, J., & McGill, C. (1984). Family care of schizophrenia. New York: Guilford Press.
Gergen, K. (1999). An invitation of social construction. London: Sage.
Gergen, K., & MacNamee, S. (2000). From disordering discourse to transformative dialogue. In R. Neimeyer & J. Raskin (Eds.), Constructions of disorders (pp. 333—349). Washington DC: American Psychological Association.
Gleeson, J., Jackson, H., Stavely, H., & Burnett, P. (1999). Family intervention in early psychosis. In P. McGorry & H. Jackson (Eds.), The recognition and management of early psychosis (pp.380-415). Cambridge. Cambridge University Press.
Goldberg, C. (2001, July 9). Children trapped by mental illness. The New York Times, pp. Al, All.
Goldner, V., Penn, P., Scheinberg, M., & Walker, G. (1990). Love and violence: Gender paradoxes in volatile attachments. Family Process, 29, 343-364.
Goldstein, M. (1996). Psycho-education and family treatment related to the phase of a psychotic disorder. Clinical Psychopharma-cology, ii(Suppl. 18), 77-83.
Haarakangas K. (1997). The voices in treatment meeting. A dialogical analysis of the treatment meeting conversations in family-centred psychiatric treatment process in regard to the team activity. English Summary. Jyvaskyla Studies in Education, Psychology and Social Research, 130.
Hoffman, L. (1981). Foundations of family therapy. New York: Basic Books.
Hoffman, L. (1992). A reflexive stance for family therapy. In S. MacNamee & K. Gergen (Eds.), Therapy as social construction (pp.7— 24). London: Sage.
Hoffman, L. (2000). A communal perspective for relational therapies. In M.E. Olson (Ed.), Feminism, community, and communication (pp. 5-17). New York: Haworth Press.
Hoffman, L. (2002). Family therapy: An intimate history. New York: Norton.
Holma, J. (1999). The search for a narrative: Investigating acute psychosis and the Need-Adapted treatment model from the narra-
Fam. Proc, Vol. 42, Fall, 2003
418 /
FAMILY PROCESS
tive viewpoint. Jyvaskyla Studies in Education, Psychology and Social Research, 150.
Keranen, J. (1992). The choice between outpatient and inpatient treatment in a family centred psychiatric treatment system. English summary. Jyvaskyla. Studies in Education, Psychology and Social Research, 93.
Lannamann, J. W. (1998). Social construction and materiality: The limits of indeterminacy in therapeutic settings. Family Process, 37, 393-413.
Lehtinen, V., Aaltonen, J., Koffert, T., Rakko-loinen, V., & Syvalahti, E. (2000). Two-year outcome in first-episode psychosis treated according to an integrated model. Is immediate neuroleptisation always needed? European Psychiatry, 15, 312—320.
MacKinnon, L., & Miller, D. (1987). The new epistemology and the Milan approach: Feminist and sociopolitical considerations. Journal of Marital and Family Therapy, 13, 139 — 156.
McGorry, P., Edwards, J., Mihalopoulos,C, Harrigan, S., & Jackson, H. (1996). EPPIC: An evolving system of early detection and optimal management. Schizophrenia Bulletin, 22, 305-325.
Olson, M. E. (1995). Conversation and writing: A collaborative approach to bulimia. Journal of Feminist Family Therapy, 6(4), 21— 44.
Pakman, M. (2000). Disciplinary knowledge, postmodernism and globalization: A call for Donald Schoen's "reflective turn" for the mental health professions. Cybernetics and Human Knowing, 7, 105—126.
Penn, P. (2001). Chronic illness: Trauma, language, and writing: Breaking the silence. Family Process, 40, 33-52.
Rilke, R. M. (1984). Letters to a young poet (S. Mitchell, Trans.). New York: Random House.
Seikkula, J. (1994). When the boundary opens: Family and hospital in co-evolution. Journal of Family Therapy, 16, 401-414.
Seikkula, J. (2002). Open dialogues with good and poor outcomes for psychotic crisis: Examples from families with violence. Journal of Marital and Family Therapy, 28, 263-274.
Seikkula, J., Aaltonen, J., Alakare,B., Haara-kangas, K, Keranen, J., & Sutela, M. (1995).
Treating psychosis by means of open dialogue. In S. Friedman (Ed.), The reflective team in action (pp. 62—80). New York: Guilford Press.
Seikkula, J., Alakare, B., & Aaltonen, J. (2001a). Open dialogue in psychosis I: An introduction and case illustration. Journal of Constructivist Psychology, 14, 247-266.
Seikkula, J., Alakare, B., & Aaltonen, J. (2001b). Open dialogue in first-episode psychosis II: A comparison of good and poor outcome cases. Journal of Constructivist Psychology, 14, 267-284.
Seikkula, J., Alakare, B., Aaltonen, J., Holma, J., Rasinkangas, A., & Lehtinen V. (in press). Open dialogue approach: Treatment principles and preliminary results of a two-year follow up on first-episode schizophrenia. Ethical and Human Sciences and Services.
Selvini-Palazzoli, M., Boscolo, L., Cecchin, G., & Prata, G. (1978). Paradox and counter-paradox. 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.
Tuori, T. (1994). Skitsofrenian hoito kannat-taa. Raportti skitsofrenian, tutkfmuksen, hoidon ja kuntoutuksen valtakunnallisen kehittamisohjelman 10-vuotisarvioinnista. [Treatment of schizophrenia is effective]. Helsinki: Stakes raportteja 143.
U.S. Public Health Service. (2000). Report of the surgeon general's conference on children's mental health: Developing a national action agenda. Washington, DC: Author
Voloshinov, V. (1996). Marxism and the philosophy of language. Cambridge, MA: Harvard University Press.
Vygotsky, L. (1970). Thought and language. Boston, MA: MIT Press. .
Weakland, J. H. (1960). The "double bind" hypothesis of schizophrenia and three-party interaction. In C. Sluzki & D. Ransom (Eds.), Double bind: The foundation of the communicational approach to the family (pp.23-37). New York: Grune & Stratton.
White, M. (1995). Re-authoring lives: Interviews & essays. Adelaide, Australia: Dul-wich Centre Publications.