Collegiality , Therapy and Mediation — The Contribution of Experts in Swedish Mental Health Law

Independent experts serve a vital role in how the human rights of patients are protected in mental health law. This article investigates the contribution of court-appointed psychiatrists (APs) in civil commitment court hearings. Analysis is based on 12 court hearings that were audiotaped. Supplementary informal interviews with participants were also conducted. Data were analysed through a combination of rhetoric analysis and discourse analysis. Analysis of the hearings reveals that APs do not fulfil their function to critically investigate treating psychiatrists’ (CPs) recommendations that patients meet commitment criteria. They typically do not ask any questions from CPs, and the few questions that are asked do not cast light on the legal issues at stake. To further understand the role of APs, their communication has been analyzed in terms of four interpretative repertoires: collegial, disclosing, therapeutic and mediating. In conclusion, the human rights of patients subjected to involuntary commitment might be at risk when therapeutic concerns are built into the process. The specific Swedish model where APs deliver their own assessment about whether commitment criteria are met may be counterproductive. This argument possibly extends to the role of medical members in mental health tribunals in the United Kingdom, Australia and New Zealand.


Introduction
People suffering from mental disorders have always been a powerless and vulnerable group.Throughout history, they have been stigmatized in local communities and subjected to exclusion processes.In modern societies, they face problems in labor and housing markets, and are challenged with discrimination due to negative attitudes and prejudice [1].This means that human rights issues are particularly pertinent in relation to mental health.In recent years, there has been an increased emphasis on positive rights for people with mental disorder.This is expressed in the UN Convention on the rights of persons with disabilities, which promotes rights regarding inclusion and independent living, health, standard of living, as well as participation in education, employment and cultural life [2].
Historically, there has been more interest in the negative rights of people diagnosed with mental illness focusing the protections against coercive interventions.Today, most democracies in the world have legislation that regulates coercive interventions.Such laws typically specify criteria for admission and also the forms for mandatory review in court-like proceedings.However, having legislation and procedures in place does not guarantee that human rights are actually realized.Starting in the 1950s, a large number of studies indicated severe deficiencies in the legal protections of patients [3][4][5][6][7][8].There was often vast geographical variation in commitment rates that signified inconsistent application of commitment laws by clinicians.Such inconsistencies were not corrected by the legal procedures that were designed to strengthen the legal rights of patients.Courts, tribunals and similar legal bodies were reluctant to genuinely challenge psychiatrists' medico-legal assessments, and the term "rubber stamp procedures" was coined to capture how legal decisions appeared to be nothing more than a ritual confirmation of clinical decisions.Despite some early research, very few recent studies have investigated legal monitoring from a human or legal rights perspective.One influential trend is the therapeutic jurisprudence movement [9,10], which has shifted the attention towards a broader discussion of the function of such procedures, particularly focusing their potential therapeutic values.All in all, we know little today about how well current court-like procedures actually protect the rights for patients under compulsory psychiatric care.
Legal decisions regarding compulsory psychiatric care are difficult because they relate to a subject matter about which lawyers have little expertise.To critically evaluate the assessments put forth by a treating psychiatrist arguing for a coercive intervention, most legal systems therefore have mechanisms that provide courts with independent psychiatric expertise.In many Commonwealth states, the legal body that makes the decision includes a psychiatrist (a medical member) [11][12][13].In Sweden, the court instead appoints an independent psychiatrist to evaluate the case and provide an opinion in addition to that of the treating psychiatrist [6].In contrast, in some states of the United States there is no built-in measure to guarantee psychiatric expertise and the selection of experts is under the discretion of the court [14].
Independent experts serve a vital role in how the human rights of patients are protected in mental health law.This article investigates the contribution of court-appointed psychiatrists (APs) in Swedish court hearings concerning compulsory psychiatric care.Previous research into similar legal proceedings has largely relied on data derived from either court documents or accounts from participants.This study instead draws on naturalistic data in form of audio-recordings from actual hearings.Court discourse will be investigated by focusing argumentation and also the linguistic components of expert participation.

Expert Role and Expert Discourse
The role of experts in legal proceedings can be viewed in light of a broader societal development of 'expertification'.In an information society, there is an increased dependency on specialized knowledge in a wide variety of contexts.Expertification is the process of the division of intellectual labour, where experts of various kinds gain increased impact in a growing number of social practices [15,16].In mass media, news stories lend credibility through a narrative format in which experts are invited to comment on current affairs.Within the political domain, increasing complexity shifts power from politicians to bureaucrats and experts, a phenomenon that is sometimes conceived as a serious democratic problem.Similarly, law can be viewed as the expertification of conflict resolution in society.The notion of "juridification" addresses the increasing impact of law on societal matters in general and represents the general trend of expertification [17].
Broadly speaking, the role of experts in mental health law can be placed in a wider context about who will be assigned expert status in various social contexts, how experts enact their roles, how they maintain authority and how their statements are interpreted and used by the recipients who originally called for the expert statement.The type of hearings we explore incorporate three layers of expertification.First, clinical experts make initial decisions to commit certain individuals to compulsory care.These clinical decisions are then subjected to the scrutiny of legal expertise.Thirdly, courts appoint independent experts (psychiatrists) to assist in scrutinizing the original clinical assessments.Hence, we stand before an intricate intersection of experts with different roles that all have a bearing on the interpretation of legal criteria, and, ultimately, the realization of human rights [18].
As an institution, a court serves the function to resolve social conflicts and attains its legitimacy from a special form of discursive practice.The purpose of court discourse is to solicit different viewpoints and identify objective facts that are relevant to the legal issue at hand.In mental health law, the court will face the problem to penetrate arguments that originates from a clinical context.The treating psychiatrists, typically arguing for the necessity of coercive intervention, will use terminology and draw on professional observational skills that are difficult for the court to critically examine.
Another challenge for the court regards the uncertainties inherent in psychiatric assessments.A substantial part of the authority ascribed to experts in general is derived from scientific knowledge [19].Rogers [20] has pointed to the difficulties in using psychiatric knowledge in legal decisions making.The most pertinent problem in this respect regards how psychiatric assessments rarely meet the standards of accuracy and validity that are expected in legal contexts.These problems remain despite the development of elaborate diagnostic systems like the DSM [21] and ICD [22].Considerations regarding coercive interventions often pertain to prognostic assessments that are uncertain, even when they are made by experienced professionals [23,24].To make things even more problematic in the context under investigation in this article, unlike many other medical interventions, there is no firm evidence base to lean on.There are no randomly controlled studies that investigate the efficacy of inpatient commitment.The few such studies that have investigated outpatient commitment have shown negative results [25,26].
To understand the role of court-appointed psychiatrists, it may be helpful to broaden the perspective to the functions of experts in society in large.Sundqvist [27] defines the expert as a person with special knowledge who in a political context is called for as an advisor or problem-solver.This applies to a broad variety of cases: experts like the military commander commenting current war events in news broadcasts; the bank employee advising a customer into financial decisions; the scientist assisting in public policy decisions; an IT helpdesk; or those experts who appear in legal proceedings.These experts share a similar claim of possessing particular kinds of knowledge.Their knowledge gains authority from different sources: research, education and/or personal experience, typically acquired through professional practice.The expert is needed when someone is confronted with a problem or has to make a decision where complex knowledge is relevant.It is significant that the expert role does not include responsibility for the consequences of expert statements.Typically, the expert does not take an active part in direct decision-making.This means that they cannot be held accountable in dealings where it is difficult to make predictions, such as psychiatric prognosis.
Generally speaking, the participation of psychiatric experts in legal proceedings serves two functions: to assist in understanding the problem at hand, and to provide legitimacy to the measures that are taken.Given the uneven playground between patients and psychiatrist, and how legal decision-makers have limited competence in assessing psychiatric arguments, the most important role of the independent psychiatric expert in court hearings about coercive intervention is to assist the court by providing a critical examination of the arguments and facts presented by the treating psychiatrist.
The provision of legitimacy occurs on a symbolic level, but it is also accomplished in discourse.The power that accompanies the expert role is discursive in its nature.In essence, the accomplishment of the expert is to enter into, facilitate and improve an on-going discourse.The role of the expert is reproduced in language acts and it is through the skilful application of linguistic tools that the expert creates credibility for her assessment.

Mental Health Law in Sweden
In Sweden, the rights of patients under compulsory care are protected by the 1991 Compulsory Psychiatric Care Act (LPT).Compulsory commitment requires that three criteria are met: (1) the patient has a severe mental disorder; (2) the patient opposes treatment or is unable to consent to treatment; and (3) the patient has an indispensable need of care (danger to others is relevant to this criterion).Since 2008, it is possible to put in place compulsory community care when patients are discharged from in-hospital compulsory care.
In terms of procedure the human rights of patients are protected with an independent review by a psychiatrist of the initial decision to commit.Compulsory care must be approved by an administrative court within four weeks of admission.The court can then prolong compulsory care after another four months, and thereafter every sixth months.Patients also have options to appeal for additional court hearings.
Swedish administrative courts consist of a professional judge and three lay judges that come to a joint decision.In court, a chief psychiatrist (CP)-typically the treating psychiatrist-argues for the continuation of compulsory care.The patient is usually present and represented by a lawyer.It is mandatory for the court to appoint an independent psychiatrist (AP) unless it is obviously unnecessary.The law does not specify the function or the procedural role of the AP, other than stating that s/he can ask questions to the patient and the treating psychiatrist.Although not required by law, APs are expected to present a statement about whether the admission criteria are satisfied.Such statements are delivered as a routine feature of all LPT hearings.Part of the rationale of having an AP is that the very knowledge of being under the scrutiny of another psychiatrist would encourage the treating psychiatrist to be more careful in considering the case [28].APs are mandated by The National Board of Health and Welfare and should demonstrate excellence within the discipline of psychiatry, having extensive clinical experience.
The question about who is qualified to fulfil the role as expert in compulsory psychiatric care hearings is not self-evident and groups other than psychiatrists have made unsuccessful claims for that authority.The Swedish Psychological Association views psychological competence as important, while the service user organization the Swedish National Association for Social and Mental Care has argued that courts should appoint people with personal experience of mental illness as expert advocates for patients [28].
The Swedish approach to supplementing the court with medical expertise can be seen as an attempt to find a middle ground compared to solutions elsewhere.In Sweden, APs deliver opinions on the legal issues that the courts determine, but without being part of the court.Contrary to the United States, experts are picked from an authorised roster, which means that the court itself does not assess the qualifications or relevance of the expert.APs are not subjected to cross-examination.
In Swedish administrative law, hearings are complimentary to the written documentation that is submitted by the CP prior to the hearing [28,29].In LPT cases, the CP has submitted a written statement supplemented with the form that was completed at the time the initial decision was made to commit.Only rarely does the patient submit anything in writing to the court.In administrative law, the obligation to make sure that all relevant facts are presented to the court rests with the court itself.In recent years, however, the parties are becoming more active in illuminating as many aspects of the case as possible.
Several studies have reached the finding that APs rarely differ with the opinion of treating psychiatrists-the rate of agreement between APs and CPs being 99 percent, 91 percent and 99 percent [30][31][32].Given the ethical and legal complexity of compulsory care and the lack of guidance from empirical research, the strong tendency of agreement calls into question whether APs actually realise the expectations of critically examining the assessments of treating psychiatrists.
As is the case in most industrialized democracies, apparently reasonable admission criteria and legal procedures are in place that appear to protect the human rights of mental health patients.This article examines the extent to which those protections are realized in practice and how court-appointed experts contribute to protecting patient rights.

Methods
The analysis in this article is based on twelve LPT court hearings, with three different APs.The hearings were chosen randomly and originate from one county administrative court, at different occasions and with three different judges.Complimentary data from fourteen hearings before the same court has been used from a prior study conducted by the first author [6].The average length of the 26 hearings was 24 minutes.Similar tribunal hearings in an English study were 75 min [12].Each hearing was audio-recorded and all written court documents in the cases have been gathered.Nine supplementary semi-structured interviews were carried out with participants in the hearings: judges, lay judges, official parties/chief psychiatrists, citizen parties, attorneys and appointed psychiatrists [33]. 1 These were carried out in conjunction with hearings and respondents were asked to reflect on participants' roles in the hearings.The data were collected in 2000-2001.Since then, some minor revisions have been made in administrative and mental health law: most notably that the number of administrative courts have been reduced, and formal regulation of outpatient commitment has replaced the previous practice of granting patients long-term leave from inpatient commitment.However these changes have not likely affected the role of appointed psychiatrists.Accordingly, our findings here are relevant to today's practice.
We have investigated the discourse with a combination of rhetoric analysis and discourse analysis.Classic rhetoric analysis build on Aristotle's three means for a speaker to persuade an audience [34].One method is to draw upon logical arguments (logos), but of equal importance is what is expressed through the character of the speaker (ethos) and the emotional appeal of what s/he says (pathos).Our analysis has taken its starting point in the logos of each hearing: What topics are highlighted and which ones are excluded?What facts are brought forth?Are there concealed arguments?
For the next stage, we have applied discourse analysis to capture participants' active use of language, including what is communicated and how it is conveyed.The discourse analysis also casts light on aspects of ethos and pathos and is suitable for questions like: How do APs reproduce their status as experts?What references are made to other experts?Who are they addressing?To what extent are questions moralizing?More specifically, the contributions of APs have been analysed in terms of interpretative repertoires-culturally variable ways of discussing and evaluating certain actions or events that constitute a cultural common sense [35].Wetherell and Potter [36] argues that interpretative repertoires are seldom questioned-or even noticed-within a given institution or culture.Gilbert and Mulkay [37] have demonstrated how scientists promote their own research by skilful use of "empiristic repertoires".In a previous study, we identified four repertoires applied by attorneys in LPT hearings: defending, voice-giving, therapeutic and legalistic [38].Interpretive repertoires are created and reproduced in interaction between all parties involved in a hearing.Thus, the APs are not the sole creators of the repertoires they apply. 2We expect that such repertoires would be construed in relation to other repertoires that exist in legal and medical discourses in general.
The primary source for the analysis has been transcriptions from the proceedings. 3Auditory aspects such as significant uses of tone of voice or laughter have been noted in some cases.In the transcriptions, we have taken care to capture the sense of spoken language, in particular with regards to expert vs. vernacular language and how utterances refer to previous talk.In a sense, this is unfair to the participants in these hearings, as their talk may come out as disorganized and strange to a reader who is not accustomed to reading detailed transcripts of how people actually talk. 4  1 Informed consent has been obtained from all patients and interviewees.To ensure anonymity, names of persons and places mentioned in the text have been altered.At the point in time when our grant proposal was accepted, the Swedish Council for Working Life and Social Research made its own ethical review of projects before they were accepted.
2 Note that depending on the power dynamics within a particular setting, some participants have more powers to influence the construction of repertoires than others.For a discussion of how a similar notion-that of "medical voice"-is constructed in interaction, see [39].All authors observed hearings and conducted interviews.Author Jacobsson had the main responsibility for transcribing data at the first stage of the analysis, while primarily authors Jacobsson and Sjöström have been active in refining and validating the interpretation of repertoires and excerpts.Analysis was carried out on the 12 hearings collected for this In the next section, we provide a general understanding of how these hearings are carried out by paying particular attention the case of one patient (Mr. Anders Svensson).Mr. Svensson's case was chosen partly because the professionals involved saw it as particularly difficult to assess, which would put the AP's contribution in the spotlight.The judge pointed out in an interview that the AP's performance had a special impact in this case.This hearing also lasted four times longer than the average hearing.All this gives us reason to expect that arguments would be more detailed, explicit and transparent than normal.Possible deficiencies in clarity and logic would therefore be likely to be found elsewhere as well.To provide a broader picture, the analysis will also consider other cases.The examples we have chosen in the text represent core features of the court discourse identified in this study.

How a Hearing Is Conducted
Anders Svensson's court hearing took place in a conference room at the clinic where he was committed.It was one of about a dozen undertaken by the court on that day.The court travelled between 3-4 different hospitals in the metropolitan area.Typically, one attorney would be ordained to assist all clients whose cases are heard on the same day.
This hearing was occasioned through an application of Mr. Svensson's chief psychiatrist for an extension of ongoing compulsory care.The court schedules 30 min for each case, including the court's deliberation and public announcement of its decision.Since hearings normally are announced only two or three days prior to the date, there is little time for participants to make preparations.In most cases, the AP would have read the documents submitted to the court prior to the hearing and, at times, also studied case records at the hospital.
Most patients appear in court without anyone else present than their attorney.However, Mr. Svensson was also accompanied by his mother and a community support worker.In the conference room, the judge was positioned at the short end of a large table, with the clerk/secretary and three lay judges next to him.The attorney (public counsel) and his client were positioned at the opposite short end of the table.Anders Svensson's mother and the community support worker occupied the seats next to them, but on the long side of the table.Opposite from them, next to each other, were the AP and the CP.
LPT hearings are typically conducted according to relatively distinct phases and the tone of the conversation is usually quite informal.After a brief introduction, this hearing began with the attorney presenting arguments on behalf of the patient, whose appeal occasioned the hearing.Both Mr. Svensson himself and his mother made remarks about details.Next, the CP briefly commented on the arguments of the attorney.The judge followed up with a couple of questions on the same theme.In this discussion, both Anders Svensson and his mother contributed with further comments.
This phase was ended when the judge initiated the phase in which the AP questioned the parties.The AP began addressing Anders Svensson and his mother, before proceeding to the CP.In a second round of further questioning, the AP altered between the parties.Judges rarely interfere in this phase but may add some questions for clarification.At the conclusion of this phase, an audio-recorder was turned on and the AP delivered her formal statement, declaring that the legal criteria for commitment were satisfied.
The judge then asked if anyone would like to comment the AP's statement, which probed the CP to add an argument about the criterion relating to severe mental disorder.The judge asked the CP if that would be her final statement before the patient's attorney was invited to present her closing argument.
project.We could also confirm the existence of the repertoires presented here in the 14 hearings from the previous project, without finding any additional repertoires.
After about an hour, the hearing was concluded, and after a short deliberation, the parties were called in again to hear the judge announcing that Mr. Svensson's appeal had been denied and that compulsory treatment thus would continue.
To provide a picture of how much different actors talk during hearings, we have counted the number of words spoken in our material (excluding the phase where the decision is announced).The APs contributed 20 percent of what was said, judges nine percent, CPs 21 percent, patients (present in 10 hearings) 25 percent, attorneys 22 percent, family members of the patient (present in four hearings) four percent and others (present in one hearing) one percent. 5

Analysis
The first analytical step was to understand the logos, or logical structure, of court discourse.However, despite one of the authors being a law professor and one having extensive experience from research into compulsory psychiatric care, we struggled to follow the logos of the discussions (cf.[38,40]).Participants typically focused on the patient and his/her situation, but in a fashion that covered a broad range of aspects without connecting them to legal criteria.The APs did not differ from the other participants in this respect.Judges generally took a passive approach, they rarely requested clarification, nor did they interrupt discussions that were obviously irrelevant to the legal matter at hand.Nevertheless, apart from a few patients and attorneys, most participants described the hearings as meaningful in interviews.
An important component of the AP's task is to conduct a critical inquiry into the logos of facts and assessments presented by the CP.Surprisingly, we found that in three of 12 hearings, the AP did not ask any questions to the CP (nine of 26 if hearings from Sjöström's previous study [6] are included).Without asking questions, the experts will hardly provide court members with a deeper understanding-nor identify possible flaws-of the arguments from the CP.
In those cases where APs actually asked CPs questions, two general observations are vital: Firstly: APs largely addressed general issues relating to the patient and her circumstances.Issues regarding the mental condition and social situation of the patient dominated the APs' questions to the extent that there was little room to critically examine the arguments from the CP.
Secondly: not once did an AP ask questions that genuinely challenged the argument of the CP.APs did hardly ever inquire into the degree of certainty in assessments or the reliability of sources of information, nor did they probe into details about considerations of diagnosis, need of care or insight.Even if occasional questions were posed regarding matters such as the need of care criterion, APs typically did not follow up on such questions to make sure that replies were relevant and lucid.Similarly, APs rarely questioned facts or assessments in the documents presented to the court.The descriptions of symptoms and circumstances presented by CPs in the hearings appeared quite fragmented and unclear, especially in a context of ascertaining legally valid information.Similar observations have been made by Holstein [14] in the United States, and Perkins [12] in the United Kingdom.Doctors rarely put examples and descriptions of the patient's situation into a context that allowed listeners-lay or expert-to form their own opinion about what is claimed (cf.[41,42]).
The following excerpt from Mr. Tore Waller's hearing illustrates the lack of coherence in medical evidence.Mr. Waller was diagnosed with schizophrenia and had been treated with compulsory care on several previous occasions.At the time of this hearing, he had been on long term leave from the hospital while still being committed under LPT.A critical issue in this hearing was Mr. Waller's attitude towards medication and the risk for relapse if he would discontinue his current medication.Faced with questions from APs, CPs typically took a quite independent approach and often strayed quite far away from the topic.In Excerpt 1, it is next to impossible to understand how what the CP then says relates to the specific question how the discontinuations of medication would lead to relapse, even if starting with a "yes" seems to indicate that he is addressing the question.Rather than addressing the question as asked, the CP elaborates on side effects and the patient's failure to pay his rent.Nevertheless, the AP provides back-channelling as if what the CP says is relevant to his first question ("yes", "mm", "mm, mm", "yes, precisely").The AP leaves the initial question unresolved when proceeding to a new question about supervised housing and then about the staffing situation.
The next question concerns whether Mr. Waller has ever been able to manage reasonably without medication.The CP replies with an unequivocal "no" without specifying a time range.The AP probes into the matter asking if the relapses are "quick" and if they involve florid psychotic symptoms.
The CP confirms briefly and then proceeds to raise concerns about Mr. Waller's insight and that he used to bet on horses.Excerpt 1 represents the entire phase where the AP asks questions to the CP.The questioning has not contributed much new information, but appears to have reinforced the impression that a discontinuation of Mr. Waller's medication would inevitably lead to relapse into a severe psychotic condition.
Even though APs did not particularly contribute to provide clarity to the testimony of CPs in their questioning of the CPs, nor exhibit a critical investigation, participants did largely appreciate the contribution of the APs.
Most of the factual ground for a court's decision was gathered from documents produced in non-legal contexts by persons not present during the hearing.It was difficult for the APs as well as the courts to question the validity of such material.Hence, APs had to rely on second-hand information that has gained status as "reliable enough".One strategy for APs to demonstrate a critical stance towards such second-hand sources was to ask questions about isolated details from hospital records.In doing so, they demonstrated knowledge of the case but without seriously challenging the comprehensive arguments put forward by the CPs.
We will now proceed by discussing four interpretative repertoires (collegiate, disclosing, therapeutic and mediating) that we have identified. 6Two traits of how the repertoires operate will be highlighted.Firstly, we will argue that they counteract a truly critical inquiry of evidence.This raises the question of how the contributions of APs are meaningful to participants.Therefore, we will also address how the interpretive repertoires that are applied serve to reproduce the AP's status as an expert.

Collegial Repertoire
The collegial repertoire was identified in all nine hearings where APs asked questions to the CP.It expresses a kinship between the AP and one of the parties in court, namely the treating psychiatrists whose assessment the AP is assigned to critically investigate.The two doctors share a similar expert status, have similar education and are members of the same profession.It is quite likely that the AP has a history of previous professional or social relations with the CP, and if not, future encounters may be envisioned.Thus, apart from sharing medical interpretative frames, the two doctors are likely to share a sense of concord.In our interviews, one CP described how she viewed the AP as "strongly supportive", while an AP acknowledged that "the greatest risk, I suppose, is to get into an alliance with the chief psychiatrists".In the hearings, we have identified a collegial interpretive repertoire that expressed the interdependence of APs and CPs.
The collegial repertoire permeated much of the APs' involvement in LPT hearings and took several quite different expressions.The reluctance to raise truly inquisitive questions discussed above is one dimension of the collegial repertoire.Another aspect is how APs provided supportive back-channelling to the CPs, and also how the two doctors discussed patients in the third person.In Excerpt 1 the AP addresses the CP with her first name, something we interpret as an expression of closeness and concord.APs' questioning of CPs often took the character of a collaborative effort to present arguments about the patient meeting with commitment criteria.For example, in Anders Svensson's hearing, the AP reinforced the CP's assessment that Mr. Svensson was psychotic by putting forth additional information from the hospital files.
Another aspect of the collegiate repertoire was found in the hearing of Muhammed Jahri, where the AP and the CP engaged in whispering conversations at several occasions during the hearing.When the judge poses a question concerning uncertainty in Mr. Jahri's diagnosis, the CP replies using the pronoun "we".Thus, she indicates that this assessment is produced jointly by herself and the AP.
Here, the AP is included in statements from the official authority responsible for the decision to initiate a coercive intervention.In this case, it is particularly troublesome that the issue of uncertainties in diagnosis is not discussed openly, allowing the patient and his attorney to raise counter arguments and the court to have a better factual basis for its decision.

Disclosing Repertoire
The disclosing repertoire was identified in ten of the eleven hearings where the patients was present.It appeared when APs interviewed patients and an impression was created that the AP was able to disclose signs of mental illness that the lay people in court might have missed.It can be understood in the context of an expectation among clinicians and members of the court that patients were capable of "composing themselves" for the short duration of a court hearing [6,14].
Like with CPs, APs often refrained from following up patients' responses to their questions.However, exchanges with patients were often left unresolved in a way that made the patient's message appear strange.APs were also inclined to frame their questions in a manner that made discussions incoherent.To any, listener, the conversation in Excerpt 2 would come out as strange.Let us consider how the AP contributes to this.First of all, it is difficult to follow a logic in the different questions she asks, whether we try to see them as follow-ups from Mr. Svensson's replies or as a set of questions that were planned beforehand.Furthermore, the questions in themselves are quite unclear and it is understandable that Mr. Svensson struggles to make sense of them and respond appropriately.To the extent that the questions are designed to reveal the content of possible delusions on behalf of Mr. Svensson, the conversation here does not result in clarifying what they might be.It is also noteworthy that the questions are often quite leading, often implying symptoms of mental illness.Moreover, we note that for most of the issues raised in the excerpt, there is no resolution.The AP does not follow up to clarify, even though what Mr. Svensson says often appears out of context and extremely difficult to understand.Crucially, listeners do not learn about the time-frame of the delusions or bizarre ideas that Mr. Svensson is portrayed having, how consistently they recur or if he holds them at the time of the hearing.Other features that contribute to incoherence is how the AP at a few times quite forcefully interrupts Mr. Svensson with challenges ("I beg your pardon!" and "what did you say?").
Everyday conversations are managed through strong norms about interaction, where one essential "politeness strategy" is the mutual collaboration of interlocutors [43].Given the norms that exists for everyday conversation, Mr. Svensson is put in a difficult position to respond to the questions asked by the AP.Throughout the conversation, he comes out as hesitant in his replies, but it is clear that he makes an active effort to cooperate in having a conversation.For example, this is apparent in how he begins several of his utterances with the word "yes", which in this context cannot be taken as a straightforward expression of agreement to the content of what the AP said.The incoherence of the questions in themselves also makes it difficult for Mr. Svensson to provide answers that come out as distinct and rational.However, given the collaborative norms for conversation, he is nevertheless likely to try to say something and attempt meaningful responses even to questions he does not understand.
From a logos point of view, the conversations in Excerpt 2 lead by the AP do not help the court to gain deeper understanding into the patient's private circumstances or his current mental condition.Rather, the effect seems to be that the patient's contributions to the conversation indicate mental illness.Holstein [14] observed similar patterns when investigating mental health proceedings in the United States, where participants talked about a skill to allow patients to "hang themselves".Similarly, one of the CPs we interviewed emphasized an important ability of APs: "you can tell that she's really great and that she can talk to the patient in a totally, like, right way."This manner of presenting the patient as crazy may be crucial to the court's decision [6,14].
The disclosing repertoire also serves to demonstrate the expertise of the AP, how she possesses a special communicative skill with which she can assist the court in revealing hidden signs of mental illness.In part, this is possible given the institutional expectations that are inherent in patient-doctor conversations.In medical contexts, such encounters have a distinct institutionalized form.Patients in general are socialized to cooperate with their doctor [44].The therapeutic relation rests on assumptions of trust, that takes another flavour when the psychiatric assessment becomes relevant in legal contexts [20,45].From the patient's perspective, information that is provided in a context of therapeutic trust, can be used "against her" in a legal proceeding.It seems as if some of this generalized trust in doctors extends to APs in court hearings, which in turn facilitates the APs' use of the disclosing repertoire.

Therapeutic Repertoire
Therapeutic goals and the relevance of on-going and proposed treatment are relevant to the legal decision-making.However, at times therapeutic concerns appeared to overshadow and transgress whatever relevance therapeutic discussions had for the legal decision at hand.This is exemplified in Excerpt 3, which follows after the AP has asked questions to the CP regarding Anders Svensson's future treatment.The CP relates plans about bringing in a psychologist and also about a treatment home specialized on the particular kind of problems that Mr. Svensson has.Again, the AP's contribution cannot be understood as being primarily guided by the logos of finding whether legal criteria are satisfied or not.Rather, the AP is trying to motivate the patient to comply with the treatment that has been suggested by the treating psychiatrist: "wouldn't that be good?","You see one advantage with this . . ." Possibly, this conversation could be construed as relevant to the third criterion in § 3 LPT, whether the patient opposes to treatment or lacks ability to form an opinion.However, no such a connection is made by anyone elsewhere in the hearing, in the AP statement or in the court's written motivation of the decision.
Another example of the therapeutic repertoire is found in an excerpt from another patient-Eva Göransson.As Excerpt 4 begins, the AP leans over the table, looks the patient into the eyes while changing her voice to a more personal tone: With a personal tone of voice, the AP signals a status as helpful doctor, at the same time downplaying the legal purpose of this conversation.She uses a diminutive form ("staying") for the recommendation that the patient should remain at a locked hospital ward and alludes to the four months that the court would decide on as "only for a short period".When Ms Göransson raises an objection about the side effects of her medication, the AP suggests that "it's perhaps possible to make a compromise", but without holding any powers to ensure that this would occur.It seems here that the AP's communication here serves to persuade the patient to accept further treatment and comply to treatment, rather than to assist in clarifying the legal issue at hand.
The professionals in court sometimes acknowledged that the APs did take a therapeutic stance.Unprompted, a judge touched on this issue in an interview: "I don't really approve of the way they [APs] sometimes start a therapeutic conversation during the hearing."However, nowhere in the hearings did a judge interfere to redirect the discussion towards legally relevant issues when APs engaged in a therapeutic repertoire.
One function of the therapeutic repertoire-that appeared in all eleven hearings in which the patient was present-may be that it alleviates some of the anxieties caused by a coercive intervention.In interviews and informal conversations, psychiatrists pointed out how patients suffer from taking part in the hearings.(Interestingly, this was not reflected in frequent complaints from patients about possible stresses of the hearings cf.[6]).Although it appears humane, there is an obvious risk that attempts to ease the stress that patients might experience during hearings also veil the legal role of the AP.The similarities in form to clinical doctor-patient conversations may also lure patients to provide information that otherwise would have been withheld.In this sense, the therapeutic repertoire shares a problematic affinity with the disclosing repertoire.

Mediating Repertoire
At the time of a LPT hearing, the CP and the patient are involved in a clinical relationship.Sometimes, the court becomes a stage for a therapeutic dispute over treatment content, goals and plans that transcends the legal issue that the court is assigned to resolve.In interviews, some CPs reported that they saw the hearings as opportunities to have franker discussion with patients.Peay [5] has reached similar findings when interviewing psychiatrists in the United Kingdom.The adversarial organization of courtroom conversation, the fact that there were other, impartial persons present and the patient having the support of an attorney would likely contribute to this function of the hearings.
The mediating repertoire appeared in all but one of the eleven hearings where the patient was present and serves a function when hearings become an arena for negotiation of clinical issues.In Excerpt 4 above, the final line of the AP: "So I heard but I thought that it's perhaps possible to make a compromise" is one such example.Excerpt 5 illustrates how the mediating repertoire unfolds in an extended context.

AP
The fact that he is here under LPT gives the clinic a certain responsibility.The particular manner in which the AP poses a question here, creates a context where the clinic's legal obligation to provide treatment for every citizen in need, is construed as a matter of discretion.The fundamental right that his doctor takes responsibility for treatment is here reconstructed as a generous offer for Mr. Svensson.The CP is accordingly represented as responsible, generous and compromising.Mr. Svensson's hesitant-and perhaps puzzled-response, again abiding to basic conversational rules of politeness, are not followed up by the AP, and we are left with an impression that some sort of concord exists between Mr. Svensson and his CP.
The mediating repertoire was also a means for APs to position themselves as neutral bystanders in the conflict between the CPs and patients.The APs offered the parties with opportunities to comment on what had been proposed from the other side.One example of this is when Anders Svensson's mother alerted the court that a psychological test indicated that her son in fact was not psychotic.The AP then turned to the CP to solicit a brief comment.The factual matter was not resolved, but through asking a question, the AP had demonstrated her impartiality.
Through the application of a mediating repertoire, the fundamental conflict between the legal parties was toned down.The mediating repertoire rarely had bearing on the legal assessment of the case.However, similar to the therapeutic repertoire, it may have served to create a sense of patient participation.By extending the width of the APs' expertise, the mediating repertoire also helped reproducing their status as experts.

Concluding Discussion: Human Rights, Expertise and Legitimacy
The human right protection from being arbitrarily committed to compulsory psychiatric care relies on expertise.This article has investigated how court-appointed psychiatrists assist Swedish administrative courts in critically examining the recommendations of treating psychiatrists to commit patients.The results give rise to some concern.First of all, APs only occasionally asked questions to treating psychiatrists.Questioning that was directed towards CPs tended to be carried out in a collegial repertoire in which the APs allied with the CPs to solicit reasons in support of the latter's application to the court.In contrast, patients (when present) were always subjected to questioning that was often performed in a disclosing repertoire where patients were encouraged to provide responses that affirmed expectations that they were mentally ill.Moreover, the two other repertoires-the therapeutic and the mediating-shifted attention away from the key task of assessing whether commitment criteria were met.This created confusion about the APs' role in the hearings to the effect that patients were enticed to collaborate in a disclosing repertoire that was harmful to what they sought to achieve in the hearings.The credibility of the problematic traits identified here is reinforced by the fact that they featured in Anders Svensson's hearing, where the discussions were exceptionally extended and participants attested to the importance of the AP's contribution.The overall findings here helps understanding the strong agreement in assessments between CPs and APs [30][31][32].
These findings raise three questions: how do APs manage to reproduce their expert status, why is it that they fail to efficiently assist courts with expert knowledge, and what changes can be envisioned to improve practice?
At least two features appeared vital to the reproduction of APs' expertise in the court hearings.First, the expert status was accentuated through linguistic means in the application of interpretive repertoires.The best example is perhaps the disclosing repertoire, in which the APs demonstrated aptitude in apparently making patients reveal disclosed symptoms of illness.The AP's professional competence was further emphasised in the detailed questioning concerning treatment (often undertaken in therapeutic and mediating repertoires), although such discussions had limited relevance to legal criteria.
The mediating and therapeutic repertoires contributed to the reproduction of expertise in another way.On a rhetorical level, experts emphasised their pathos and ethos through these repertoires.The APs appeared as emphatic persons who cared about patients.Through these repertoires, APs were also able to advice CPs on clinical matters.Being useful in these extra-legal aspects appeared to have contributed to the sense that the APs served a meaningful role in the hearings despite their limited input in substance.
The APs' failure to adequately challenge the assessments of CPs does reflect difficulties that are inherent in their task both on a theoretical and on a practical level.The fact that similar findings are at hand with regards to the attorneys of patients [40] suggest that the problem cannot be reduced to collegiality.One challenge for APs is the relatively limited time they have to prepare before hearings.The lack of a solid research base is a challenge that perhaps explains the tendency towards speculation that Beck [15] finds typical for another kind of risk discourse, namely the one involving environmental issues.In fact, we have not found a single reference to research in the twelve hearings analyzed.The reluctance to challenge recommendations of CPs can also be seen as part of a more general trait of mental health hearings.Hearings are held in administrative courts, with the explicit rationale to expose patients to less stressful experiences than envisioned in the more adversarial general courts.This is in line with the ideas of therapeutic jurisprudence (TJ) that are becoming increasingly popular in the United States, Australia and New Zeeland [46,47].It seems values favoured in TJ such as accommodating the expected stress of patients and providing a sense of participation and being listened to have been catered for when APs have applied particularly the therapeutic and mediating repertoires.However, these efforts might have drawn the attention away from the human rights issues at stake, particularly given the very restricted time provided for the hearings.On a more general level, these problems can be associated with the longstanding tradition of paternalism in mental health care [6,48,49].
Finally, then, how can expert contribution be arranged and performed in mental health hearings in order to better accommodate human rights?Our results provide reasons to raise caution over moving to far in building therapeutic concerns into the legal process.It may be more respectful to patients to maximise the transparency about what is at stake in a commitment hearing and thus promote sharper discussions about how legal criteria are met.With regards to the Swedish situation, we conclude that the format where APs are assigned to present their own opinion of whether commitment criteria are met may be counterproductive.The scarce resource of the AP would come to better use if s/he were allowed to focus on scrutinising the assessment of the CP, and act like an independent advisor to the legal decision-makers.This argument can possibly be extended to jurisdictions where expertise is built into decision-making bodies, like is often the case with medical members in mental health tribunals in the United Kingdom, Australia and New Zealand.

3
Transcriptions from audiotapes are somewhat adapted to written standards.[Square brackets] indicate clarifying contextual information inserted by the authors, CAPITAL LETTERS indicate loud speech, italics indicate stressed utterances, and hyphens (-) indicate that there is an immediate overlap between utterances.

Excerpt 4 :
Court hearing regarding Eva Göransson AP Do you know what I think?Personally, I'm going to say what I think and I'll be saying that you should stay.I really think also that you should take the medicine and you know this is only for a short period.It is important that when you leave you still have your apartment.PATIENT Yes it is.AP Sorry?PATIENT I'd like to take Trilafon but not in this form, I have considerable side effects.AP So I heard but I thought that it's perhaps possible to make a compromise.PATIENT Yes.
PATIENT Yes that's when I asked if you could touch someone's breasts and those sort of things, I don't think in that way anymore.AP No, but it's also the case that sometimes you must think about things secretly because they are telling you "you can't' think like that", is it . . .Because I mean, it's also about what you can deliberately think, what else could you think about deliberately . . .PATIENT Yes that it . . .that it, yes . . .yes that in principle you shouldn't eat and stuff and not eat, what you want, what you'd think is most fun.AP WHAT DID YOU SAY? [loudly] PATIENT Well, for example that plants can't feel things and stuff like that.AP WHAT DID YOU SAY? [loudly] PATIENT That plants can't feel things, stuff like that.