Introduction
Empirical ethics has become an accepted approach in bioethics.Footnote 1 Yet theoretical and methodological debates over the eventual combination of empirical research and normative analysis are still ongoing. One core issue relates to the interaction between empirical research and normative analysis. Some want to keep this interaction linear; others argue for a cyclical process.Footnote 2 A second core issue concerns the justification of outcomes. Whereas some say normative analyses should be guided by bioethical theories and principles, others use coherency of judgments or shared credibility as justification. A third core issue is the role of the ethicist.Footnote 3 Some place the ethicist in the role of an expert and stranger to the practice under consideration; others regard the ethicist more like a participant or a critical friend.
We have developed an approach to empirical ethics that can be called dialogical practice.Footnote 4 Central to our approach is a cyclic way of working, in which empirical data collection and normative-ethical analysis are interrelated and located in dialogues with and between participants in a practice. Our dialogical approach combines philosophy (hermeneutic ethics) and social science (responsive evaluation) and is structured but flexible in terms of specific contexts and research questions. The purpose of this article is to present this dialogical approach and to show how it can be made effective within a rapidly changing practice. This is illustrated by a case example: the development of a forced detoxification program in psychiatry. We start with a theoretical section describing our approach as a sample of contextual empirical ethics. This is followed by a presentation of the case example and a description of how our approach was applied.
Two Approaches to Contextual Empirical Ethics
We regard the experience of people as a source of moral knowledge and the starting point for ethics. Musschenga divides into two kinds empirical ethics approaches that take seriously the factual convictions of people and their factual argumentation practices.Footnote 5 He labels them “contextual empirical ethics”; the first approach is grounded in the framework of reflexive equilibrium inspired by Rawls,Footnote 6 and the second, in that of epistemic contextualism.
Reflexive equilibrium approaches start with the moral intuitions of the well-considered judgments of participants in a practice. According to Rawls a reflexive equilibrium is reached when judgments and principles match each other. Judgments can, however, match different sets of principles. In order to reach a broad reflexive equilibrium, then, the relative strength and weakness of these different sets must be determined by examining the relevant background theories. Empirical ethicists use this approach not only to describe the moral system of a person or a group but also to assess the coherence of well-informed judgments with principles and background theories. The cycle between empirical work and normative-ethical analysis is (often) gone through many times, ideally until empirical saturation and rational coherence are reached. The reflexive equilibrium approach is particularly useful in theory building. The ethicist has an expert role within the ethical analysis process, guided by the reflexive equilibrium criteria of coherency. Issues of feasibility and implementation are not included in this type of empirical ethics.
In the kind of empirical ethics that Musschenga places within the framework of epistemic contextualism,Footnote 7 the concept of practice is even more central. Although research is directed toward the factual convictions of participants in a practice, the final aim is the reconstruction of the “morality of concrete practices,” which can be described as “internal morality,” in contrast to “external morality.”Footnote 8 Musschenga states that internal morality is a normative notion. The context (or practice) within which one has justified beliefs has a particular moral outlook, which one can only appreciate through a process of education. The internal morality of medical practice not only is formed by inside practitioners but also can be induced by outside influences. For example, cost-effectiveness considerations have now become part of medicine’s internal morality.
Can epistemic contextualist approaches go beyond the description and reconstruction of a practice and be relevant in solving practical moral problems? Musschenga gives an affirmative answer to this question. The ethicist, after having determined first the relevant context and then the basic beliefs it contains, has to establish if a policy can be induced from it. This requires not only a description of the factual convictions of people in a practice but also interpretive research of the social meaning of goods, practices and institutions.Footnote 9 The last step requires, in addition to inside knowledge of a practice, argumentative skills. Musschenga visualizes the role of the ethicist in epistemic contextualist approaches as a well-informed, philosophically trained participant in the further development of the morality of a certain practice. The ethicist can even develop into a “connected critic”Footnote 10 who delivers immanent criticism based on the internal morality of the practice or the morality of the broader community in which the practice is embedded.
Our empirical ethics approach is an example of this latter development. We regard the experience of people as a source of moral knowledge and the starting point for ethics. The moral beliefs, intuitions, and reasoning of people in a practice contain normative-ethical knowledge grounded in the morality of that practice. These can be used to draw normative conclusions relevant for that practice. In this process, the ethicist is not an outside expert who judges a practice but someone who is (at least temporarily) part of that practice and supports its further development. The ethicist has a specific role to play by fostering dialogical learning processes, leading to new normative insights and actions.
Dialogical Empirical Ethics
Our dialogical approach to empirical ethics is based on a specific tradition in philosophy, hermeneutic ethics,Footnote 11 and a social science approach, responsive evaluation.Footnote 12 In hermeneutic ethics, dialogue is seen as a vehicle for moral learning and the development of normative conclusions.Footnote 13 Responsive evaluation aims at understanding moral issues in practice through the exchange of concrete, detailed experiences and perspectives in dialogue.Footnote 14
The ethicist starts and moderates a dialogical learning process with and between people involved (stakeholders) in a practice.Footnote 15 By exploring and articulating the various, sometimes conflicting perspectives, she or he assists stakeholders to understand these multiple perspectives, and through mutual dialogue these may evolve into new perspectives and broader solutions to the problems at hand. The ethicist contributes to the dialogue insofar as she or he becomes involved in a problem on the basis of his or her own moral experience. His or her role is essentially twofold: the moderator of and (not more than) a participant in a dialogical learning process. The ultimate goal of this dialogue is in Gadamerian terms a “fusion of horizons”Footnote 16 reaching a shared understanding. In line with the tradition of responsive evaluation, justification of the outcomes is sought in the shared credibility of the stakeholders within a practice and the fairness and authenticity of the process. Fairness designates the genuine possibility of each participant being included and having a sayFootnote 17 in the process. Authenticity designates the enhanced personal and mutual understanding of participants and the enhanced capacity to act and change the practice in line with the shared understandings.Footnote 18 This dialogical approach can be used for both theory building and practice transformation.
We use the methodology of responsive-interactive evaluationFootnote 19 to set up a structured dialogical learning process in which issues are addressed and developed in interaction with involved stakeholders.Footnote 20 Our methodological heuristics asks for a cyclic, iterative way of working: findings from one respondent, stakeholder group, and research phase form the input for the next respondent, stakeholder group, and research phase, in a continuous process of interpretation and analysis of data. Furthermore, stakeholder groups are first consulted separately so they can develop their own stance and voice, and then they mutually discuss and integrate their perspectives. These heuristics translate into some typical research steps (or phases): (1) creating social conditions in the setting, (2) eliciting issues from different stakeholders separately, (3) deepening issues in homogeneous groups, and (4) integrating issues in heterogenic groups. Usually research methods such as in-depth interviews and focus groups are used, but creative techniques like paste-ups, drawing, photo elicitation, and narrative workshops are also considered helpful, especially for articulating inner worlds and silent voices.Footnote 21 Through this cyclic process, an integrative hermeneutic learning process within a practice is facilitated. In some cases, this requires intensive interaction and cooperation, as we show in the next case.
A Case Example: Forced Detoxification in Psychiatry
Background
In a closed ward of a Dutch mental hospital, staff had serious concerns about patients who were sectioned under the mental health act and whose use of drugs impeded diagnosis and treatment. To the staff’s frustration, the only course open to them was to control these patients. In 2007, a new program—forced detoxification—began to be developed. The program aimed to provide a diagnosis and formulate a treatment plan for each patient and thus to revive a ward in chaos. Patients were placed in an isolated detoxification section of the ward for three months.
The Dutch legal conditions for coercive interventions in psychiatry (the BOPZ ActFootnote 22) are based on respect for autonomy understood as self-determination. Coercive interventions can only be based on the danger criterion: serious danger for the patient or others. In the ward no such serious danger was indicated, resulting in a situation in which patients were largely ineligible for treatment and lacked every prospect of improvement. Furthermore, the situation disrupted the ward and demotivated the staff. This caused a dilemma for the management. Should they continue to respect the right to self-determination, or should they intervene despite the apparent lack of a legal basis? The recent shift in drug usage, from mostly cannabis to amphetamines, reinforced this dilemma. Patients stayed in the ward for an unacceptably long time (one to three years), and the change in drug use caused a significant increase in aggressive incidents. The ward became progressively more control oriented. Checks for drug possession, deception, suspicion, mistrust, conflict, and aggression were daily events. Staff described their jobs as “playing policeman” and “offering a hotel service.” Treatment decisions were reduced to medication adjustment and granting permission to go outside.
The hospital management realized that the enforced detoxification program put them on slippery legal ice but were also convinced of the urgent need to do something with this stagnated patient group. Should legal complaints arise, they were prepared for a juridical test. The patients’ parents supported the project.
Aims and Approach
The detoxification approach followed a paradoxical strategy: it used coercion to stimulate the patients to become more active and to reverse the ward climate of aggression, control, and coercion to foster more livable working relationships. The program was set up as a development project, as hardly any knowledge and examples were available, and was intended to anticipate events. It had to be developed and evaluated along the way, depending on emerging issues and experiences. The ward was split into two separate wings. One became the isolated detoxification ward. We were asked to support the detoxification project with empirical ethical research. The research objective was to illuminate expected or unexpected treatment effects, to indicate contributive and obstructive elements in the process, and to support the treatment policy and implementation process from all stakeholder perspectives. Special attention to changes in the relations between staff and patients was required.
Results
The project started with the doors being shut between the wings and a sniffer dog searching the ward for drugs. A new therapeutic milieu was initiated, in which rewarding good (rather than penalizing bad) behavior was essential. The aim of this approach was to stimulate patients positively and to change staff attitudes toward patients. Legal complaints and aggression by patients, which were feared beforehand, did not occur. Patients complained a lot but were undoubtedly more relaxed and approachable. To everyone’s surprise, a social atmosphere developed between patients. They began to help and support one another; such interactions were new. Although (given the history of coercion) their distrust never disappeared, patients gradually became friendlier with the staff. Patients’ motivation and confidence grew as they felt new possibilities were appearing. Visiting parents were happy “to be able to hold normal conversations again.” Staff were pleased about the unprecedented peace on the ward and the new social interaction. In these first months the trust in patients and in the detoxification approach grew rapidly.
Patients could now be diagnosed. Two patients showed serious psychotic symptoms even without drugs, but the five other patients improved with appropriate medication adjustments. These patients embraced the possibility of treatment. After they were transferred back to the regular wing of the ward after three months, a remarkable group dynamic occurred among them—they were taking pride in no longer using drugs. One eventually started using soft drugs again, persevered, and was placed in a care residence where controlled use was permitted. Three others went to their parental homes, the fifth to a rehabilitation house. Although there were concerns over the two seriously ill patients, in general there was a strong feeling of success and confidence in the new approach. Five stagnated patients, previously considered as candidates for the long-stay ward, were now on a rehabilitation course outside the clinic, and solitary confinement also decreased by 75 percent.
The team faced new challenges when the project became dependent on external placements. The patient population became more diverse, and this resulted in more disappointing clinical outcomes. Unlike the first patients (whose positive group dynamics guided them through), more patients quickly relapsed in the follow-up after transition from the detoxification ward. Among several patients a negative group dynamic of drug use and dealing arose. This took its toll on the staff and on the treatment confidence gained and also affected patients who were still in the detoxification ward. Unlike the first patients, they started to dread the transition. Nine months after the start of the project, a synchronous motivation and confidence decline occurred in both staff and patients. To curb this crisis, in the eleventh month the four most difficult patients were discharged. In a series of multidisciplinary reflection meetings, those involved concluded that further treatment refinement was required, and a continuation plan was drawn up. It was believed that more individualized treatment and support structures were needed, while at the same time more could be done to make use of group dynamics. The staff decided to make treatment plans the overall guiding principle in the ward routines, to make relapse prevention even more central, and to achieve an increase in team spirit. The first experiences were positive, and the patient success rate grew again.
The primary project goals proved feasible: to get a grip on the chaos and to keep patients drug-free to establish a clean diagnosis of their psychiatric condition. Thus opportunities emerged for the secondary goals: to make use of the patients’ drug-free state to awaken their internal motivation and to make them liable for treatment again. Many patients proved more responsive to treatment after detoxification. The relapse of several later patients evoked discussion on the feasibility of the secondary goals. The treatment adjustments made these aims viable again. The most remarkable and unexpected effects concerned the social interaction among patients, such as the support and trust they gave one another. Both patients and parents praised their improved relationships and said, “this should have happened much sooner.” The successful patients spoke of an increased feeling of independence, identity, and self-esteem. Although initially indicated for long-stay hospitalization, they were, much to their surprise, capable of functioning outside the clinic. Others were placed in better-suited environments. Some very ill patients controlled their drug use better. Aggression on the ward dropped, and the use of solitary confinements sharply declined. In terms of team attitude and interaction, a turnaround was apparent “from control to treatment and supportive care.”
Dialogical Empirical Ethics within a Rapidly Changing Practice
After this outline of the background, aims, and results of the project, we now look at the research process and the methods used. The research objective was to describe and formatively support the development of the detoxification project. Our contribution was to systemize learning processes and create feedback loops and dialogue among stakeholders, thus supporting practice improvements. We identified clients, their families, the psychiatrist, nursing staff, social workers, and managers as stakeholder groups. To be able to monitor the fast developments and changes that were expected, and to build rapport with the client group, the researcher (first author) was placed on the ward for three full days a week. The other two authors acted as peer debriefers to discuss methodological decisions and stimulate reflexivity. The researcher had a key to move freely in both wings of the ward, had documentary access, and could attend all regular and most ad hoc meetings.
On the day the detoxification project started and seven patients were placed in the detoxification ward, the researcher arrived, was introduced, and started the research. He began creating social conditions for research and building rapport with stakeholder groups. Besides participant observations and informal conversations, he conducted a series of in-depth interviews. After two months a multiperspective description was made of the project’s origination, goals, first experiences, and early emerging issues, and these were discussed in focus groups with the team and the project management. This procedure (interviews/transcription, analysis, feedback, and discussion in focus groups) was repeated three times during the project. Halfway through, we wrote a report describing program development and implementation, and attitudinal changes. This had a primarily performative aim, conceptualizing and structuring these topics to systematize and foster learning processes. These were discussed in focus groups with the nurses, the treatment team, and both groups combined. Toward the end, this process was repeated, aiming at the integration and embedding of findings; see Table 1.
Table 1. Project Activities
At the beginning, we used regular interview and focus group research procedures. These were very well suited to family members and institute managers outside the ward. With stakeholders on the ward, however, a more challenging process evolved. We knew that establishing rapport with this patient group would be a challenge, but working through the issues and supporting the required changes within the nursing and therapeutic teams also proved to be complicated. We first explain how we approached the patients and then how we dealt with staff.
The patients were mainly recently detoxified young men in their twenties suffering from schizophrenia, with a history of coercion in the institution. Most patients initially kept their distance from the researcher and tested his trustworthiness. Although he was explicit about the reason for being there, he decided to wait for them to show an interest in sharing their experiences for research purposes (for instance, in an interview). He spent the first few months participating in their daily activities, such as playing table tennis, drinking coffee, and smoking cigarettes. He related to them as friends: he chatted about their daily experiences, or what he had been doing during the weekend, and shared jokes with them. Humor appeared to be excellent for easing the atmosphere and creating a feeling of mutual respect and equality. After a while some patients expressed an interest in research participation. It soon became clear, however, that their limited concentration span and the formality of the interview setting prevented them from talking openly. We therefore adjusted our methods and continued the usual conversations and interactions that had developed in the course of time. Some intensive relationships developed. The researcher kept a logbook of these on-the-fly interviews, analyzed and summarized recurring themes and issues, and shared these with the patients to verify interpretations and obtain permission to use them in the research process and dialogue with other stakeholders. The issues included “undergoing coercion,” “boredom,” and “absence of fresh air” and were fed back to professionals.
Although the first period of the detoxification project showed very successful patient results, the alterations the staff had to make were vast and wide ranging. This period was hectic and intense. The detoxification program and the new therapeutic milieu had to be developed while patients were already on the detoxification wing of the ward. The many views on what should be done entailed fierce discussions among the staff. They went through many rapid and meandering developments, challenging their routines and requiring communication and attitudinal changes. Meanwhile this patient group was keen to take advantage of staff uncertainty. Three months into the project the tensions escalated into serious conflicts between the nurses and the psychiatrist. This dominated the atmosphere and threatened the process we had been asked to support. We adopted a mediating role, resolving the immediate team crisis, and supported the implementation of communication procedures to exclude miscommunications and to secure nurses’ input in the decisionmaking processes. We have elaborated on this mediating process elsewhere.Footnote 23 Here we focus on our role in helping the staff to change their moral practice. These goals were expressed in the motto “From control to treatment,” which was adopted by a team that had almost forgotten how that should be done. It required a rethinking and gradual reshaping of moral routines on both a group and an individual level. Therefore more reflection and dialogue were necessary than could be realized by regular interview and focus group procedures. As part of our assignment to support the project development, we searched for ways to canalize these processes and make the required changes happen. Because we were embedded in their daily routines and adjusted to the various stakeholders’ communication styles, we developed more on-the-fly techniques. We illustrate in the following how we fostered the dialogical learning processes in this way.
We had interviews, focus groups, and many intermediate encounters with the team manager, psychiatrist, social worker, and (later on) psychologist. The center of dialogue within the therapeutic team was however their weekly patient presentations, in which they evaluated patient progress and made treatment decisions. Difficult issues raised concerned the nature of patients’ motivation (“Wouldn’t they say anything to leave the clinic?”) and accurate assessment of their capacities. The researcher was present in these weekly meetings during the whole project, initially to observe and describe the project, but gradually taking a more active role. Many implicit normative issues played a role in the patient evaluations. To make these explicit, the researcher started to ask questions. Usually, however, questions and conversations stimulated by the researcher merged into the meeting, intensifying normative reflection in the deliberations. An example was whether or not patients could be trusted, given the risk of their slipping back into control habits. Another intervention was drawing attention to unheard voices. We would bring in the issues we heard from patients, or focus attention on the perspective of a person in the group if we felt his or her perspective was not being noticed. In this way, we introduced the key characteristic of good dialogue (inclusion, openness, and engagement) into these central weekly meetings. Our role was Socratic, supporting normative reflection and analysis by posing questions and drawing attention to implicit values and norms, and including all involved perspectives.
The most intensive dialogical processes emerged with the nurses, as the project affected their work radically. The required change in the balance between keeping control and giving supportive care required a redefinition of the internal ethics of the team. Many nurses thrived on what they called “getting back to our profession again” and were passionate to rediscover what “good nursing” meant to them. Real changes to the therapeutic environment were a complex multifaceted issue including different normative and attitudinal challenges. Because he was on the ward daily and attended most of their daily transfer meetings, nurses started to treat the researcher as a dialogical partner. Apart from these meetings (in which he took a Socratic role), nurses saw him individually to discuss their issues. These issues concerned concretizing the new therapeutic milieu, reshaping their attitude toward and routines concerning clients, regaining their pivotal role in treatment processes, regaining their informative role in decision procedures, and so on. Many of these issues were canalized and addressed in task-orientated work groups supported by the researcher. Furthermore, he set up a work group with the senior nurses, to monitor and evaluate all these processes. Some nurses, who were cynical about patient possibilities, clung to the old control morality; others, who were more optimistic, emphasized “real contact” with and counseling of patients. The ward changes deeply affected their individual moral commitment, and nurses used the researcher as a counseling partner. Some could not handle the required changes and left the ward.
The Role of Interpretation, Conceptualization, and Ethical Theory
We continued our usual dialogical research structure of interviews, homogeneous focus groups, and heterogeneous dialogue groups, including transcription, analysis, and member checking of our interpretations. Yet the formative research aim, rapid developments, and wide-ranging issues, including the transformation of the internal morality of the ward, demanded an intensified and integrated dialogical hermeneutical process. Radicalizing Gadamer’s idea of truth before method,Footnote 24 we chose to concentrate on the transformative process. Letting the flux of the hermeneutical process guide our methods, we gradually merged our dialogical activities into regular meetings, daily routines, and informal meetings. In addition to interviews and focus groups, the researcher used ethnographic methods like participant observations and Socratic conversations. He operated as a permanently available hermeneutical worker on the ward. This was gratefully acknowledged on the ward, as all participants felt a need to reflect and talk more extensively about the issues and developments there.
Despite the flexibility of methods, we did not abandon methodological rigor. Our basic heuristics consist of a cyclic, iterative way of working: findings from one respondent, stakeholder group, and research phase form the input for the next, in a continuous process of interpreting and analyzing data. Furthermore, it requires us first to consult stakeholder groups separately, to enable them to develop their own stance and voice, and then mutually to discuss and integrate their perspectives. These heuristics were easily incorporated and safeguarded in this integrated way of working. More complex was keeping the technique of member checking intact. Normally, data from interviews or focus groups are transcribed and analyzed and given to respondents to check for possible misinterpretations or wrong conceptualizations, thus reducing bias and heightening contextual credibility. In the dialogical dynamics we performed in this quickly changing context, we used informal member checks to validate findings. Besides asking questions and drawing attention to implicit normative issues or blind spots, the researcher checked his interpretations in dialogical questioning. If he noticed that his conceptualizations or interpretations were picked up and used more structurally by practitioners, he started to question these to test their credibility.
To some extent, this is indeed a messy process.Footnote 25 Hermeneutic learning processes are often quite dynamic. In our view the ethical researcher is a participant in such mutual hermeneutic processes, not a normative expert. She or he moderates a dialogical process in which participants play the main role in all parts of the cyclical process of empirical data collection, interpretation, and ethical analysis. The researcher is not normatively neutral in this process. She or he performs interpretative conceptualizations of data, makes connections with theoretical issues and debates, and consequently brings in elements of external moralities and understandings. Yet these are not treated as given normative frameworks. As epistemic contextualists, we placed the justification of outcomes in the specific context in which they were used, not in any outside position, preordained principles, theoretical framework, or judgment of an ethical expert. Moral justification is not found by transcending the context but is reached in the internal dialogical process, implying credibility for participants in the context and the fairness and authenticity of the process. Theoretical insights may have a value if they are recognized by the participants. In an embedded approach, this asks for rigid self-reflection, a critical but at the same time humble attitude, and creative communication skills on the part of the ethical researcher.
Conclusion
Our empirical ethics approach works within the framework of epistemic contexualism. We aim at reconstructing and fostering the transformation of the internal morality of a practice through a dialogue with all participants and perspectives involved. The researcher acts as a participant in a hermeneutic process of moral development: changing moral understanding and meanings through dialogical interaction. Our approach focuses on transformation of practice, but this also enables us to develop and test theories (always in touch with practice). The case presented in this article is extraordinary in that it concerned a complex and rapidly changing practice. To be able to monitor and support the fast changes, the empirical researcher had to be present as an embedded participant. This entailed an adjustment of the regular repertoire of in-depth interviews and focus groups to the specific characteristics of the practice, weaving dialogue naturally into daily routines, gatherings, work groups, and informal encounters on the ward. We used techniques from the field of anthropology and action research, such as participant observations and keeping field notes, and used work groups. In terms of dialogue, a methodical shift took place toward dialogues on the fly.
Adapting our interactive approach to this practice, we were able to foster reflection and moral learning both in the stakeholder groups and with individual participants. Patients became more responsible and constructive, interested in building up their lives. Nurses changed their role from controller to coach of and co-worker with patients. Psychiatrists became more aware of the need to involve nurses and patients in decisions about treatment and care. These learning processes are not simple and linear. Many intensive and often meandering interactions are needed to keep on going and determine the right track. In these processes, the researcher acts as a catalyst, asking questions, stimulating normative reflection, and making people attentive to silent voices. The researcher has to be careful and reflective, constantly checking whether his or her interventions are helpful and conducive to the hermeneutical process of moral learning. She or he definitely has an influential role, not as an external ethical expert but as a person who is present and involved, asking questions and suggesting the need for further deliberation in a constantly changing practice striving for improvement.
