Distributing Reflexivity through Co-laborative Ethnography

In ethnographic research and analysis, reflexivity is vital to achieving constant coordination between field and concept work. However, it has been conceptualized predominantly as an ethnographer’s individual mental capacity. In this article, we draw on ten years of experience in conducting research together with partners from social psychiatry and mental health care across different research projects. We unfold three modes of achieving reflexivity co-laboratively: contrasting and discussing disciplinary concepts in interdisciplinary working groups and feedback workshops; joint data interpretation and writing; and participating in political agenda setting. Engaging these modes reveals reflexivity as a distributed process able to strengthen the ethnographer’s interpretative authority, and also able to constantly push the conceptual boundaries of the participating disciplines and professions.


Introduction
Reflexivity is vital to ethnographic research. In fact, ethnography can very well be considered a research approach characterized by a practice of continuous movement between theory and empirical inquiry, a necessary to-ing and fro-ing between the immersive qualities of fieldwork and rather reclusive concept work. For ethnographic research practice, it is essential that fieldwork and concept work be both separate steps and dependent on each other. If the ethnographer lets her empirical material speak for itself without reflecting on its conceptual value, she becomes a travel writer; if she conducts her conceptual work without sufficient relation to the fieldwork, she becomes a philosopher; and if she conducts both steps at the same time, she weakens her professional authority, because she did not take sufficient time to reflect on the field experience. Fieldwork and concept work cannot be coeval, reflexivity keeps them apart and coordinates them at the same time.
This coordinating function of epistemic reflexivity has been discussed comprehensively, in terms of the role of the researcher in the field (Binder et al. 2013;Lindner 1981;Nader 2002;Rabinow 1977), in terms of the situatedness of concepts and knowledge production in the broader sense (Bourdieu and Wacquant 1992;Haraway 1988), and in terms of the writing of texts that expose the keys to one's own critical questioning (Knecht 2012). While the relevance of interaction for ethnographic research is obvious in all of these dimensions, reflexivity is essentially attributed to the researcher. This focus on the researcher as the epistemic subject in ethnographic and qualitative social science research more generally has moved subjectivity at the center of many debates about reflexivity (Mruck and Breuer 2003;Woolgar 1988). This concern reaches from subjectivity as a bias to be reduced through methodical control to subjectivity as the generative passage point to any interpretative process that needs to be made transparent and embraced. In both cases, the focus is on the cognizant subject as the source of epistemic gain or loss (Kuehner et al. 2016).
As anthropological ethnographers, we are firmly committed to considering subjectivity an opportunity for epistemic gain. We are also very aware of Michael Lynch's warning not to let reflexivity become an academic virtue, a source of privileged knowledge, or solely bound to the individual researching subject (Lynch 2000). As an ethnomethodologist, Lynch conceptualizes meaning-making practices as interactive practices. Drawing on the practice turn (Schatzki et al. 2001) ourselves, we adopt his approach of understanding reflexivity not just as an ethnographer's individual mental capacity, but also as an embodied process of making ethnographies accountable that necessarily involves the researcher as much as the participants in research fields. Building on the work of Kim Fortun (2001), George Marcus has recently as well advocated that, in contemporary anthropology, reflexivity encompasses "a discussion about the ethical and normative commitments of anthropological knowledge, and its accountabilities as such, in relation to specific publics, institutions, and global projects as ethnographers move recursively in their circuits of inquiry" (Marcus 2015, 92).
We advance this strand of discussion by presenting three cases of tying reflexivity to the ethnographic research process. Conducted over the last ten years, all three cases are rooted within our fieldwork in urban mental health care and with social psychiatry in Berlin, Germany, and include persistent efforts to develop a sustained collaboration between our ethnographic research group and several academic and clinical actors in social psychiatry. The efforts include shared research as well as hosting graduate and undergraduate projects and teaching formats. All three cases, employing what we have called "co-laborative ethnography" (Niewöhner 2016), share the common goal of contributing to ethnographic knowledge production through joint epistemic work with actors in the areas under study, and an analytical focus on concrete practices that define the norms for psychiatric care. 1 In addition, all three cases share a common starting point: the realization that individual epistemic reflexivity quickly reaches its limits in fields that are themselves marked by institutionalized reflexivity and that have ingested significant parts of the problematizations on which the ethnographer draws. In other words, we all share the anxiety of becoming ethnographers who return from the field with nothing to say that the "natives" do not already know and articulate themselves (Boyer 2015).
Of course, the productive interactivity between ethnographer and interlocutors in the field has been an issue since the late 19th century (Lassiter 2005). Yet only rather recently have more collaborative forms of ethnographic research come to be explored systematically. As we cannot do justice here to this growing literature that extends far beyond anthropology into several fields of qualitative social inquiry, we develop our argument by referencing the key domain of participatory research (Bergold and Thomas 2012), often called collaborative ethnography in anthropology (Lassiter 2005), to which we broadly assign our work as well. On the one hand, participatory approaches focus on marginalized groups in the broadest sense of the term. Researchers work with a group of people in order to reconstruct that group's knowledge and empower the group vis-à-vis an acute matter of (political) concern (Burawoy 2005;Lassiter 2008). On the other hand, participatory research aims at activating the immersive quality of fieldwork for ethnographic concept work. This collaborative ethnographic work takes increasingly experimental forms and is highly reflexive about ethnography as a method and its historical adaptations Fitzgerald and Callard 2015;Marcus 2010).
All these approaches epistemically embrace "strong reflexivity" (Harding 1993), which "appreciates the perspective of the researcher and her relationship to the field as a decisive source of data and interpretation" (Kuehner et al. 2016, 700), but also face and respond to a dilemma that has been associated with it: [T]he stronger the subjective account of the researcher gets, the more difficult it becomes to argue for the validity (not to mention any "truth") of her position. The dilemma is obvious: Giving up the authoritative position of the sovereign researcher and acknowledging her decentredness disavows any claim of interpretative authority regarding the subject matter of the research that goes beyond the pure self-reflection of the researcher. (Kuehner et al. 2016, 700) Essentially, we believe that collaborative approaches, as the one we present here, can address this dilemma by transforming questions of individual interpretative authority into questions of the shared production of and coping with epistemic contingency (Boyer 2015). However, in all participatory constellations, not being able to contribute meaningfully to the existing body of knowledge in the field remains a quandary of interpretative authority that continues to occupy social scientific knowledge production. In order to generate fresh insights, some social researchers return to their desks after a collaborative research phase with actors in the field to critically evaluate and deconstruct the collaboratively generated data individually against the backdrop of their own scientific discipline. The ultimate outcome of the project then amounts to a form of distanced critique, a means of knowledge generation considered by some as lacking legitimacy as it is unable to address adequately the research questions that those involved in prior collaboration consider vital (Latour 2004). At the other end of the spectrum, some researchers devote their social scientific enterprise entirely to the epistemic positions and problem definitions of their collaboration partners. This form of expanding existing knowledge has been critiqued too for contributing much to the field but little to the concept work within the researchers' academic community (Zuiderent-Jerak 2015). In anthropology, many consider the latter form of immersed collaboration a vital element of the discipline (e.g., Scheper-Hughes 1995). Since epistemic problems involved in this kind of collaborative project are less obvious in cases where researcher and field share the same motivations and (political) goals, most collaborative work occurs under such circumstances. In sociology and other social sciences, such collaborations are viewed with greater skepticism and debates on the importance of critical distance between fieldwork and analysis endure (cf. Hirschauer 2008). With regard to epistemic reflexivity, deconstructionist critique is, broadly speaking, founded on the researchers' individual reflexive capacities, while immersed collaboration conflates reflective practices with the epistemics of the collaboration partners.
In this article, we attempt to move on from this debate and treat practices of collaboration, reflexivity, and social science concept work in our ethnographic projects in close relation to each other. We show, first, that shared political or epistemic goals are neither mandatory for co-laborative ethnographic research nor for the practical achievement of reflexivity, rather the co-laborating partners need to commit themselves to a process of continued reviewing of professional knowledge practices. Second, we argue that colaboration as "temporary, non-teleological, joint epistemic work" (Niewöhner 2016, 2) contributes to the assemblage of reflexivity as a practice that is distributed across a set of places, people, and encounters. Research in highly reflexive fields then questions ethnographic knowledge production, but it does so in a generative sense. We encourage and support research in such fields and attempt to show that by distributing reflexivity through co-laboration across people and time, interpretative authority is gained, not lost.

Engaging in Discussion and Feedback
In our co-laborative research projects, establishing distributed reflexivity over time required continuous concerted encounters between co-laborating partners. Regular meetings created the basis for trust in the careful consideration of arguments and experiences presented in earlier project phases. In 2011, some of us (Bister, Klausner, and von Peter, together with our colleague, psychologist Manfred Zaumseil) initiated a regular academic working group of social psychiatrists/psychologists and anthropologists. Setting up this group was prompted by our ethnographic fieldwork experiences in mental health care institutions that showed us just how deeply both disciplines shared an interest in the fundamental social scientific perspectives on psychiatric classifications and knowledge manifest in the oeuvres of, for example, Goffman (1961) and Foucault ([1961Foucault ([ ] 1967. Consequently, we engaged in critical reading and interdisciplinary discussion of key concepts from psychology and anthropology, group members' publications and lectures, and interim results from our first co-laborative project, which focused on a practice-theoretically informed investigation of the classification of "chronicity" in the everyday life of mental health care institutions. 2 The more meetings we held, the more we realized that our exchange was gradually generating an increasing number of research questions that were productive for our respective disciplines, extending far beyond the reach of single research projects or perspectives. The working group aimed to clarify disciplinary perspectives, making them visible and debatable in the field. Since this was a prerequisite for the mutual generation of reflexivity, all participants were required to make both their theoretical assumptions and the logics of their epistemic and clinical practice explicit, and to engage in understanding cases, concepts and practices through the lens of the respective other disciplines. This proved to be particularly challenging for key concepts related to complex genealogies in both psychiatry/psychology and anthropology because, although these genealogies overlap at some points, they also diverge at others. Consequently, many intra-disciplinary foundations that were taken for granted demanded explication in this exchange. Over time, the working group developed a shared vocabulary that fostered joint learning and trusting debate. Examples of this are sessions in which the group debated concepts of "authenticity" and "experience" on the basis of anthropological and psychological readings and empirical material from the ethnographic research project on chronicity, such as observation protocols. Both notions have been widely discussed within each of the disciplinary histories, but in very different ways. In anthropology, for example, authenticity has been thoroughly deconstructed and is today largely understood as a performative category not imbued with an a priori exceptional normative force (Beck 2015;Handler 1986). In psychiatry, on the other hand, authenticity is an important attribution for engaged professional practices that marks a certain therapeutic stance (Hilgers 2018). In the colaborative meetings, we discussed whether authenticity and, in a similar vein, experience, could and should be considered as performative phenomena that are partially strategic and open to reflection, or whether they should be considered spontaneous expressions of a person true to a presumed essential core. It became clear that, for many of the participants, the tension between authenticity, experience, reciprocity, hierarchy, and normalization raised a number of vital disciplinary questions, some of which have since been pursued in individual research projects, such as the choreography of psychiatric practice (Klausner 2015), and the relationship of experience and expertise in psychiatric peer support (Schmid 2020). The participants from psychiatry and psychology debated how different therapeutic schools reveal different relationships to, and uses of, the concepts of authenticity and experience, and discussed those in relation to particular care practices such as psychiatric home treatment. Importantly, the practice of putting disciplinary concepts and assumptions to the test and also in confrontation with each other did not lead to a synthesis or agreed-upon interdisciplinary terminology. Rather, it afforded disciplinary reflexivity in the sense that both anthropology and psychiatry had to acknowledge the particularity, partiality, and limitation of their usage of the respective concepts. From this process, participants took away insights for their own disciplinary contexts as well as a better understanding of the other's epistemic practices, and published on these. The group had thus created a new epistemic space.
Over time, participants came to trust the symmetrical allocation of criticism across the disciplinary cultures represented and began searching for shared "matters of concern" (Latour 2004), for example, the perceived necessity to foster qualitative research in psychiatric care, to engage in long-term methodological reflections across the involved disciplines, and to establish an enduring network of and for researchers interested in such research endeavors. Despite this reciprocity, however, one asymmetry always remained: social psychiatry as a psychiatric practice has always represented a research field for social science research, whereas social scientific practice has not been a research interest for the participating psychiatrists. Consequently, psychiatrists were involved in co-laboration as scientists and as psychiatric-psychological specialists whose core task is the treatment of people with mental crises. Metaphorically, therefore, the psychiatrists constantly moved along a continuous gradient between the two poles "expert" and "academic partner" (not least qua attribution by the social scientists). While some colleagues from psychiatry and psychology particularly cherished this confrontation with social science perspectives and the critical evaluation of psychiatric work and expertise, it also proved challenging for the divergent expectations of the practical outcomes of the co-laboration, for example, how and where the results of the ethnographic research process and of joint discussions were to be published and which disciplinary audiences were to be addressed. Nevertheless, as we do not conceptualize co-laboration as a harmonious or harmonizing endeavor assimilating the different disciplinary concerns, but as a shared space from where the disciplines could potentially evolve their critical reflections beyond joint presentations and publications (von Peter et al. 2016;von Peter and Bieler 2017), this shared space allowed the members of the group to re-insert critical reflections into their own discipline (Bister 2018;Bister et al. 2016;Klausner 2015;Klausner et al. 2015;von Peter 2013).
Although the working group involved psychiatrists and psychologists, it still operated at a distance from the ethnographic field sites in which two of us (Bister and Klausner) carried out our ethnographies on mental health care practices at that time. This led us to think about how we could achieve "contextual reflexivity" as proposed by the anthropologist Jeannette Pols (2006). For Pols, care workers' engagement with ethnographic sequences that sketch and interpret concrete working practices can initiate reflection processes that represent constructive alternatives to conventional evaluation instruments that assess clinical everyday life according to best practice criteria (for a critique refer also to van Loon and Zuiderent-Jerak 2012). Accordingly, we initiated feedback workshops with employees in hospitals and community mental health care services with whom we conducted research. These workshops aimed at exchanging knowledge with the staff at their daily workplace and during working hours. Through practices of jointly reading observation protocols and putting the anthropologists' interpretations of them up for discussion, we provided the psychiatric teams with insights into ongoing ethnographic analyses opening these up for immediate re-examination. As we gave this feedback during our research visits, we continued to be available for further discussions and exchange beyond the feedback meetings. These feedback processes situated ethnographic analysis as critique within the research process and-most importantly-within the daily care settings, rather than imposing normative claims about good care from the outside. Situating critical accounts within the messiness of psychiatric care work demanded staying true to the continuous efforts and "everyday ethics" (Brodwin 2013) of caregivers. For us, the anthropologists, this demanded constant effort to translate our evolving analysis into accounts that spoke to the practitioners, requiring us to stay with the ambivalences of the field. Whenever we moved into a critical distance from the field to analytically deconstruct the field's discursive dynamics, we always returned to re-read this distanced critique through the field's own perspectives thus showing "face" to the field (Haraway 2008) and accepting a more-than-anthropological accountability for our analysis. Staff members confirmed that these feedback meetings during working hours provided them with a platform for a critical approach to the effects and implications of everyday treatment. The meeting on the ways in which daily accountability measures in community care co-define the governance of people who have been classified as mentally disabled provides an example here (cf. Bister 2020). This discussion added further detail to the weaving of our anthropological analysis and provided practitioners with new perspectives on previously little debated implications of their daily working practices. These meetings thus contributed to questioning the status quo and to revealing its contingency by multiplying perspectives.
The feedback workshops did not, however, result in direct interventions into existing daily routines. It was only in a subsequent ethnographic research project that a hospital team explicitly picked up our feedback method in order to question the effects of their own treatment practice using our ethnographic observation protocols. 3 Through the feedback workshops, we ourselves had received continuous feedback on the immediate relevance of our research for psychiatric practice, and the tensions between ethnographic analysis and its relevance and uptake in the field had become part of our empirical material.
Thus, by allowing interpretative authority to be distributed across a set of actors, we established a co-laborative space for epistemic reflexivity.

Generating shared analysis
In another ethnographic project, we learnt that reflexivity can be distributed by joint practices of coding and interpreting empirical data with the actors in the research field. Two of us (Schmid and von Peter) have tried this variant in a multi-professional discussion group on the topic of psychiatric peer support. While shared coding is an established practice in the social sciences (Bergold and Thomas 2012), this is usually shared between two researchers and not, as in this case, between researcher and the co-laborating partners (cf. Morrison et al. 2019).
In psychiatry, professional peer support (Genesungsbegleitung in German) is offered by people with experience with mental crises and diagnoses who now work as employees in psychiatric care institutions (Utschakowski 2015;Utschakowski et al. 2009). Some of these peer support workers have completed a specific training, the Experienced-Involvement Training (Utschakowski 2015;Utschakowski et al. 2009), and are supposed to use their subjective and personal experience and knowledge as a resource to support current patients in their recovery. In the first instance, we set up a multi-professional group based on the results of a previous project that looked at how day-patient care practices changed due to the introduction of peer support workers in day-care hospitals (Schmid 2013). We initiated discussion about some of the core hypotheses involved in a focus group consisting of eight participants: three peer support workers, a family expert, 4 two psychologists, a psychiatrist, and an anthropologist. This group met three times over a period of one year in two German cities. The discussions were recorded, transcribed, and then coded and interpreted in a joint process. Although two participants were not familiar with academic research processes, in the sense that they did not have an academic degree nor had they done research before, a thematic discussion unfolded about the role of peer support workers and their practical work tasks on psychiatric wards, a subject that we had not tackled before.
We attempted to adapt anthropological analytical practice by following, first, an open, and then, a focussed coding process in pairs (Emerson, Fretz, and Shaw 2011), each involving a person with lived experience of psychiatric wards and one without. In the first step, two teams ascribed thematic codes to sentences and paragraphs of the discussion transcript. In the second step, the whole group met again and discussed all codes, reconciling the different codes between the groups and deciding on overarching codes and their descriptions. In the third step, these codes were applied again by pairs of two to the exact same transcript, modifying a focused coding process. Each piece of material was thus repeatedly the object of coding by four teams in an open as well as focused coding process. Through this multiple coding approach, we sought to involve all group members and their different disciplinary perspectives (cf. Bergold and Thomas 2012). It was a time-consuming, intense and conflict-laden process that marked the beginning of a shared analysis, which was subsequently, after one and a half years of joint work, published in an article on the professional role of peer support workers in Germany (Heumann et al. 2018).
The shared practical goal of this group was to write an interdisciplinary paper on peer support work in Germany together with peer support workers. Consequently, while we initiated this shared research and work process across different disciplinary and professional perspectives, we also ensured that the peer support workers' questions were debated and answered. In the end, an elaborate negotiation process was initiated that encouraged participants to reflect and make explicit their perspective, core ethical and political values, as well as their use of specific concepts such as "empathy." Questions of labeling and wording surfaced similar to the solely academic working group described in the previous case. In addition, this mixed group also debated different objectives and prioritizations for the research process. While some participants saw this working group as part of a larger political agenda that is changing the status of peer support work in Germany, others were mainly concerned with the challenges in everyday life for peer support workers themselves and still others were interested in the possibility of developing analytical concepts for peer support work. Different ideas of analysis and "proper" empirical research sometimes clashed not only with other participants' anthropologically "undisciplined" analyses but also with research paradigms from medical academic contexts regarding, for example, questions of generalization from empirical material. Conversely, the practical impact of the work, which was the main objective for some participants, was not achieved through this publication of an academic paper.
Since the process involved both people with and without experience with qualitative empirical research material, the work of reading and analyzing the transcript of the discussion group came as a surprise to some. The question how to deal with the phenomenon of transcribed spoken language also raised questions. One participant, unable to identify with her parts of the discussion in the literal transcription, chose to drop out of the group. This co-analysis of the empirical material pointed to, and enhanced, differences between the participants at different levels, depending on experience with empirical material, status, and research objectives. In the end, the discussion group decided to write a multi-vocal paper that did not attempt to gloss over differences in perspectives, thus rendering the differences within the group productive. The joint publication emphasized the heterogeneity of the different perspectives rather than their synthesis while still pursuing the practical goal of a single paper.
The group co-laborated in a recursive process in which group members' differential experiences, assessments, and uncertainties regarding peer support work were described, compiled, and reflected. Curiously, the heterogeneity of the group as well as members' desires to learn something useful for their own respective work contexts, contributed crucially to the generative distribution of reflexivity. Distributed reflexivity thus multiplied possible interpretations of the empirical material and widened the conceptual repertoire of the group. To turn an anthropological phrase: This process provided multiple interpretations that proved to be more than one and less than many. In that way, it opened up possibilities for unexpected anthropological concept work.
The notion of experience provides an illuminating example of how anthropology gained from this process. The documentation of the multi-professional discussion group and the subsequent meetings as well as the resulting article informed the analysis of a concomitant ethnographic project, in which we explored how the peer support workers' former experiences of crisis were negotiated as a (counter) part of formally recognized expertise in daily mental health care practices (Schmid 2020). This ethnography shows that experience is treated as a fundamental category both in the field of anthropology (for example, in the form of field research experiences) and in the context of peer support work. Yet it is not well explored theoretically in either case. Looking at experience "at work" in the field of peer support work also provoked us to look back from a different perspective at experience at work in social and cultural anthropology: the field of psychiatric peer support work emphasizes the effects of an experience someone has "lived through" that plays out in psychiatric care situations. The emphasis on the effects and practical implications of experience shaped the reasoning of the ethnography.
Similar to the academic working group discussed in the previous section, the discussion group made the epistemic work of the research subjects visible. Additionally, this group engaged in a process of shared analysis and writing with academic and non-academic actors, aiming at a result in form of a paper. Hence, the time frame and the aim of this co-laboration were clearly defined, while that frame opened up the possibility of experimenting with usually exclusive academic working steps such as the coding of empirical data and writing an academic publication. In this example, epistemic reflexivity was generated as a "'sense-making' activity performed by people or inscribed into written documents" (van Loon 2015, 17), a context-dependent, distributed practice that enacted individual critical thinking as a relational phenomenon.

Participating in Political Agenda Setting
Our third and last case returns to the research project on chronicity in mental health care. This project revealed that an in-depth ethnographic study of the relation of mental distress and urban environments provided the opportunity to amplify co-laborative inquiries with scientists as well as with medical staff and social workers in psychiatry and beyond. Our aim therefore was to take this ethnography as an opportunity to produce knowledge to help build a better psychiatric care system (cf. Rose 2019) and to inform urban planning and social policy (cf. Söderström 2017) while generating vital theoretical and methodological insights for the social sciences (cf. Fitzgerald et al. 2016).
Trying to understand how neighborhoods affect mental health, one of us (Bieler) expanded extensive fieldwork with people with psychiatric diagnoses (including go-along interviews, Bieler and Klausner 2019a) and psychiatric staff by doing participant observation across public administration and a lobbying project of a social welfare organization targeting the inclusion of people with mental distress. Threats of eviction, the resulting homelessness of mental health care clients, and the lack of affordable housing for new clients in inner-city neighborhoods posed major problems for the psychiatric care system that demanded rapid political action (Bieler and Klausner 2019b).
The lobbying project under the umbrella of the welfare organization explicitly addressed these problems through alliance building with housing companies and social policy stakeholders-rather than simply demanding rights and resources-and by assessing the problems and needs of mental health care clients, social psychiatric services, and housing companies. By becoming a member of the lobbying project, we were able to experiment with balancing participation in the lobbying project's political agenda setting with contributions to anthropology. This approach provided access to otherwise confidential meetings with housing companies and political stakeholders and, simultaneously encouraged the distribution of reflexivity.
In order to involve the actors in continuous discussion on the potentials and limits of enhancing their mutual cooperation, in recurring presentations to the lobbying project's transdisciplinary advisory board, we tentatively reflected on observations conducted in the project (and in other field sites). This happened, for instance, by sharing an observation of a discussion between the project team, a federal politician, and a representative of private landlords. When the representative of private landlords asked for any specific features that could improve the living conditions of people with a psychiatric disorder, most project members spontaneously argued that stigmatization and negative stereotypes were the core problem resulting in exclusionary practices. However, one of the project members disagreed: According to him, focusing on stigmatization alone was not sufficient as inclusion was addressed as a moral issue, based on the assumption that people should share social psychiatry's ethics of care. Rather, he pressed to discuss what he called "hardware of inclusion," that is, material elements and infrastructures that could improve life for people with and without mental health problems, for instance finding solutions that would help avoid reciprocal nuisance in apartment buildings. This controversial comment depicted a critique of a romanticized notion of "community" that was held by many members in social psychiatry. Complementing these observations with insights from the goalongs with mental health care clients, we tried to provoke discussions on the importance of non-institutionalized urban spaces and actors that would not be captured with social network analysis or an analytical focus on social capital (cf. Blokland 2017). In this sense, this co-laborative approach resulted in staging a para-site, "risking interpretations together with the researcher about ideas fundamental to the political organization of their institutional contexts and functions, and how these ideas circulate, have effect, and change" (Deeb and Marcus 2011, 52).
Preparing these presentations was an important, tentative analytical step carried out during and alongside the ethnographic fieldwork process as topics crystallized and needed to be ethnographically grasped while not being fully "processed." This allowed for an evolving analysis over time. Currently, we analyze the empirical material by using the notion of the encounter (cf. Wilson 2017) in order to account for the manifold ways that people's lifelines are related beyond a dichotomous understanding of urban relationships as either "close" and reciprocal or absent, anonymous and isolating (Bieler forthcoming).
In addition, co-working in a stricter sense-co-writing two reports that featured recommendations for action, co-presenting the results at expert conferences, and co-designing the final expert conference of the projectallowed for more direct forms of mutual confrontation. Through this co-working, we were directly confronted with the ways in which the project team dealt with, legitimized and problematized practical obligations and occurring pressures, and were simultaneously able to confront the project team with (often) differing, anthropologically informed readings of the situation and construing complementing or alternative consequences. Being involved in the field, became part of the very same processes we analyzed (cf. Sánchez . Anthropological knowledge could thus be translated into feasible propositions: Whereas in previous publications broad definitions of supportive social relations in neighborhoods were generalized as enhancing recovery, for the projects' final report (featuring recommendations for political action) we suggested to include a section on the necessity of planning and developing urban spaces as a socio-material process. Finally, we argued that a "one size fits all" approach was insufficient, that positive social relations were not easily achievable by inclusion and needed to allow the possibility of social distancing and withdrawal and that, to create a dynamic planning process, situated ongoing transdisciplinary knowledge production was needed. The fact that these propositions were easily accepted by the advisory board of the welfare organization became an interesting empirical fact in itself: Besides a narrow, normative biopolitical promotion of mental health by advocating for the inclusion of people classified as "mentally ill" on a moral register, thereby naturalizing the concept of mental health, the projects' final report argues for situated, transdisciplinary socio-material processes of designing environments targeting the quality of urban coexistence-without a priori distinguishing between people with and without a psychiatric diagnosis and without neglecting material elements (cf. Hendrickx and Van Hoyweghen 2020).
To sum up, in this third case of distributing reflexivity, ethnographic analysis gradually emerged and evolved from attempts to solve the problems at hand by jointly problematizing the challenges and limits of mental health services due to current housing market pressures and seeking cooperative solutions with housing market actors and social policy stakeholders. The ethnographic analytical process thus took place tentatively in relation to and alongside fieldwork encounters and was not derived solely from reclusive analysis. Reflexivity was thus distributed through discussing specifically selected empirical material and preliminary analytical interpretations with the co-laboration partners within their everyday practices while simultaneously feeding back into and informing further analysis and ethnographic concept work.

Distributing Reflexivity
In this article, we have argued for understanding co-laboration as joint epistemic work that enables the distribution of reflexivity among different actors. Enacting reflexivity during and alongside the research process in recursive and ongoing confrontations between ethnographer and field as well as between different disciplinary thought styles offers a way of addressing dilemmas of strong reflexivity. Based on ten years of intensive co-laboration between anthropology and social psychiatry, we have illustrated different practices of distributing reflexivity: discussing disciplinary concepts in interdisciplinary working groups, feedbacking, and discussing preliminary ethnographic results, attempting joint data interpretation and writing and confronting the reflective practices of the field with anthropological problematizations. While these practices may differ in terms of timing within the research process, frequency of iteration, involved actors, addressed publics, and outcomes, they all share certain aspects. First, ethnographic analysis is a tentative generative process that arises from engagements with the respective field sites rather than from distance to it. Second, ethnographic analysis evolves as a recurring process that is continuously shaped by the attempt to make anthropological concept work meaningful to the reflexive attentions of the research partners. Third, the way the research subjects-understood as epistemic partners-self-critically problematize and practically deal with the challenges and limits of their professional activities of care and knowledge production are actively taken up in order to grasp and explore tensions with the evolving ethnographic analysis. Committing oneself to the mutual challenging of interpretations-including analytically informed ethnographic interpretations-is vital to the enhancement of analytical robustness, and adds to ethnographic authority.
Analysis thus gradually emerges from reflexivity distributed across people and across project stages and feeds into further reflecting practices. In this ongoing and long-term process, the epistemic practices of ethnographers and research subjects are continuously entangled but do not (necessarily) coalesce. This has two implications. First, co-laboration is most fruitfully practiced as a long-term process that extends beyond individual research projects. The research projects described earlier do not stand by themselves, they are mutually related and reinforce each other. They have by no means completed the development of co-laboration between anthropology and social psychiatry, rather the opposite is the case. We are continuously carrying and shaping our co-laboration further. By doing so, we take advantage of our long-term ethnographic presence in the field, constantly relating individual research projects to the co-laborative process. This raises different research questions and aims at communication with each other and with the field. Second, a successful distribution of reflexivity in joint epistemic work requires research fields that share an ethos of co-laboration: an ethical commitment to confront, discuss, and transform the challenges and contingencies of epistemic practices. While highly reflexive research fields might at first appear to be a potential "threat" to ethnographic knowledge production, we argue that reflexive fields encourage us to distribute reflexivity, taking ethnographic analyses into new and unforeseen directions while also experimentally advancing ethnography as a methodology.

Author's Note
Manuscript for the special issue of JCE entitled "Practices of Ethnographic Research," guest edited by Julian Hamann and Andrea Ploder

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Notes 1. Even though co-laboration is inherently an assembled practice between different actors and artifacts, in this article, we focus particularly on the implications of distributing reflexivity in ethnographic inquiries and less on the challenges for our co-laborating partners. 2. The project was titled "The Production of Chronicity in Mental Healthcare and Research in Berlin." Refer to funding details. 3. The project was titled "Mobilising Clinical Health Care: A Qualitative-Ethnographic Pilot Study about Health Care Practices in Psychiatric Home Treatment." Refer to funding details. 4. A "family expert" is a peer support worker who is not personally experienced with psychiatric diagnoses or psychiatric wards, but is experienced as a family member of someone who has been, or is in, psychiatric care.