Paper Patients When Documents Stand in for Patients

This article analyses a seemingly mundane feature of a mental health centre for immigrants and refugees in Paris: the documents used by budding therapists undertaking their apprenticeships. Supervisors developed these documents in order to train therapist apprentices to learn the explanatory models of patients, identify the voice of patients, and incorporate medical anthropology into their therapeutic practice. The documents were central to the experiences of therapist apprentices because they occupied most of their time and were a substitute for supervision and patient contact. Documents disciplined the speech of therapist apprentices and focused their attention on specific aspects of patients’ histories. Therapist apprentices found these documents and documentary practices to be problematic because they reduced patients’ complex migration and medical histories to a series of tick boxes and short answers. These documents generated new forms of uncertainty among therapist apprentices about how to present clinical information about patients to their supervisors. This article is part of a larger study that considers mental health services for immigrants and refugees as communities of practice in which therapist apprentices learned to develop clinical and caring skills for vulnerable patient populations. By drawing on and contributing to scholarship on apprenticeship, uncertainty, documents, and bureaucracy, this article demonstrates how bureaucratic processes and documentary artefacts may generate unnecessary forms of uncertainty and hinder participation in communities of practice.


Introduction
In 2012, I spent the summer at a mental health clinic in the north of Paris for people with a variety of migration backgrounds. The centre provides therapy to people from all over the world and is staffed by psychiatrists and psychologists who are themselves multilingual, or who otherwise procure the services of professional interpreters. Health and social service professionals in a variety of public and private institutions refer patients to this centre because of its linguistic and transcultural expertise. Moreover, the centre is an important site of in-house training for budding therapists. These tend to be graduate students in clinical psychology but might also be medical students, residents in psychiatry, or health professionals who wish to learn to support patients who are grappling with mental disorder as well as instability in residency status, housing, employment, and integration in France.
When I first visited the centre, I often participated as an observer in assessment meetings in which one of the secretarial staff would read aloud referral documents that had been sent by external professionals to a group comprising the centre's senior clinicians and social worker. This group would collectively decide how to best chart a course of treatment for those who had been referred. I returned to the centre in 2014 for a longer stint of fieldwork, during which I learned that it was now the thérapeutes en formation [therapists-in-training, or TEFs], who presented referrals to the centre's clinicians by means of a double-sided presentation form. I also learned that the clinicians and social worker had introduced the form because they were not satisfied with the ways in which TEFs had initially presented patient referrals. What was particularly striking about this newer arrangement was how the presentation form was both a tool for guidance and a site of friction. It guided TEFs to identify the most vital elements of patients' medical and migration histories, but senior clinicians might castigate TEFs if they either relied on it too heavily or deviated too far from it. While forms, documents, and paperwork are a ubiquitousand thankless-feature of most organisations, I contend that these presentation forms are an important site of analysis because they dominated the TEFs' training and acted as a substitute for patient contact, and because they collapsed and encased the highly complex mental health and migration histories of patients into a series of tick boxes and short answers. I will argue that these forms, in the absence of clinical contact with patients, provoked uncertainty about how to understand the realities of patients, and created tensions between TEFs and their supervisors.
In the pages that follow, I describe my ethnographic analysis of apprenticeship and documents. I then introduce the centre and its form, and I situate this form within two rich bodies of scholarship: one on clinical socialisation that identifies how students-often of medicine-learn to think, speak, and carry themselves like professionals in their disciplines; and another on documents and bureaucracies, which illustrates how documents shape practices and create an image of the people they describe. In my analysis, I demonstrate how this form served to discipline the speech and thought of TEFs by encouraging them to identify the voice of patients and socialising them to speak like senior clinicians. My analysis also identifies how the form and the way it was implemented led to uncertainty when TEFs fluctuated between speculative and authoritative speaking about the realities of patients. The form caused TEFs to question the intentions and practices of senior clinicians because TEFs wished to learn about the complexities of patients' medical and migration trajectories rather than reduce these to a single document. By drawing on and contributing to literature on the anthropology of documentary artefacts and practices, I illustrate how TEFs used this particular kind of document to speculate about patients and the appropriate course of action. I contend that the notion of 'paper patients' allows us to analyse what it means for budding therapists to learn about migrant psychotherapy when paperwork substitutes for patient contact.

Ethnography of apprenticeship and documents
This article is part of an ethnography of therapeutic apprenticeship that analyses the experiences of budding psychologists and psychiatrists in four mental health settings in the greater Paris region for people with different migration backgrounds. I draw on seminal scholarship that characterises apprenticeship as legitimate peripheral participation in communities of practice (Lave 2011;Lave and Wenger 2009;Wenger 1998). Rather than focus on training, I use apprenticeship to understand how learning, identifying, and social membership are mutually constitutive. This article draws on ethnographic data I collected in one of the four mental health settings, as well as interviews with TEFs and their psychologist and psychiatrist supervisors. While not a budding clinician myself, in the course of my fieldwork I had the same role as TEFs in the centre: reading through referral documents, inscribing information from these documents onto the form (see Figure  1), presenting these forms to the supervisors, and occasionally observing patient consultations. The length of time TEFs spent at this centre varied from a few weeks to an academic year and their number varied. On those occasions when a patient came for an initial evaluation, TEFs needed to register in order to attend; typically, attendance at these evaluations might be limited to just a few TEFs. Initially, I shared the TEFs' concerns about the lack of opportunities to shadow clinicians' interactions with their patients and by the dominance of paperwork in their institutional lives. Gradually, however, I became fascinated by the ways in which these seemingly mundane documents served an important socialising and disciplining role for these budding therapists, and I chronicle these roles in what follows.

The work of paperwork among TEFs
TEFs wrote their name at the top of the presentation form and were responsible for presenting the information contained in the referral documents to the senior clinicians, who would then determine a course of therapeutic action for a patient. I will analyse these presentations later in this article. The first section of the form contained details regarding the professional and institution referring the patient, as well as biographical details about the patient. TEFs were also expected to indicate whether the patient was already undergoing some form of treatment. The remaining sections were meant to guide TEFs as they sorted through the different kinds of information contained in the referral documents. The boxes and prompts for short answers served as a system of organisation and classification of biographical and mental health information about the patient.
In her recent book, anthropologist Stéphanie Larchanché (2020) provides an ethnographic account of the same centre and its assessment meetings. Larchanché states that these are not just typical staff meetings, but rather represent an important site for the transmission of the centre's expertise to TEFs, who get to learn about complex referrals and directly confront clinical material (2020, 132). Larchanché also analyses the presentation form as a study sheet for TEFs that transforms their gaze so that they can understand the realities of patients; nevertheless, she quotes one of the centre's psychiatrists as a caution that these realities are not so easily organised as they are shown to be on the form (2020,. This article analyses how these forms stand in for patients and how therapist apprentices experience these forms as 'paper patients'. This article foregrounds the experiences of TEFs, whereas Larchanché's (2020) rich and compelling ethnography centres the perspectives of supervising clinicians. This distinction is important: as ethnographic accounts of clinical training and socialisation (e.g., Becker et al. [1961Becker et al. [ ] 2003Prentice 2013;Sinclair 2004;Wendland 2010) have long demonstrated, things look quite different from the relative vantage points of students and their supervisors. Taking the vantage point of TEFs, this article argues that these forms, in an absence of clinical contact with patients, provoke uncertainty about how to view patients' realities and generate tensions between TEFs and their supervisors.
By uncertainty, I refer specifically to the work of medical sociologist Renée Fox (1957Fox ( , 1980Fox ( , 2000, who identified three types of uncertainty which trainees face during clinical training: (1) the limits of one's own medical knowledge, (2) the limits of the field and what can be known, and (3) the inability to distinguish between these. Fox has detailed-as have the myriad of studies on clinical socialisation inspired by her seminal work-how clinical trainees go to great lengths to minimise the uncertainties they face. This body of scholarship has demonstrated how uncertainty can be a productive component of therapeutic apprenticeship as apprentices learn to master their craft. In the sections that follow, I will demonstrate how paperwork in this centre, and the presentation form in particular, generate a different kind of uncertainty to that which Fox observed. This uncertainty relates to how knowledge is translated-being about a patient but in the voice of a clinicianand then transposed onto a form like the one above.
To best understand how TEFs experience the work of paperwork in this centre, this article draws on a conceptual framework of apprenticeship which refers to legitimate peripheral participation within communities of practice (Lave 2011;Lave and Wenger 2009;Wenger 1998). Rather than simply spectate, apprentices participate and contribute to the activities of communities of practice. The community element of a community of practice requires apprentices to be gradually absorbed into the group and begin to identify with its more senior members. Indeed, apprenticeship is distinct from the transmission of knowledge since it involves the process by which one becomes a practitioner (Bryant 2008); regarding it as the latter risks the assumption of internalisation (Scheid 2002). As Volker Scheid's (2002, 164) ethnography of apprenticeship in traditional Chinese medicine has identified, attending to how one becomes a practitioner, rather than how knowledge is transmitted, allows us to consider learning as a more openended process that involves the interaction between identity, knowing, and social membership. In the present context, I consider how TEFs contributed to the work of these assessment meetings and how they became members of this community of practice. I also contend that TEFs' engagement with these forms is a unique space in which to examine the intersection between research on clinical apprenticeship and research on the function of paperwork. My analysis contributes to scholarship on apprenticeship since I demonstrate how paperwork about patients cannot be a substitute for immersion in practice with them.

Documentary multiplicity: Organisation and socialisation
Paperwork is a ubiquitous feature of just about any workplace, yet scholars in a variety of organisational contexts have identified that documents mediate practice rather than simply record it (Berg 1996;Berg and Bowker 1997;Brenneis 1994;Harper 1997;Hull 2012;Riles 2006). In an ethnographic account of the evaluation of grant proposals to several national agencies, Donald Brenneis (1994, 32) identified how participants in evaluation meetings self-discipline by incorporating standardised evaluation criteria into their recommendations, thereby anticipating a future audience. This perspective suggests that documents require a process of translation that renders information legible to other users, both current and future.
In a seminal review of documents and bureaucracy, Matthew Hull (2012, 253) suggests that documents are not merely instruments of bureaucracies but are instead constitutive of rules, ideologies, knowledge, practices, subjectivities, and the organisations themselves. Hull charts a significant shift from conceiving of documents in terms of their capacities for representation, to understanding them in terms of their form and the kinds of effects that are generated by their production and circulation. Documents also promote control by coordinating perspectives and activities, and they have a capacity for the construction of entities, subjects, and forms of sociality (2012,. Ethnographic research on documents in clinical settings has highlighted how documents serve multiple, simultaneous purposes, such as facilitating clinical evaluations, data collection, and clinical instruction (McKay 2012). Documents can be thought of as a 'paperwork technology' that stages interventions (Brodwin 2011, 190), and allows for different courses of action to take place (Street 2011). Documents used in assessment meetings reflect practices of representation and evaluation and are central to the processes of turning people into patients (Carr 2011, 53). Documents also serve an important socialisation purpose for medical students and interns, since it is through the proper use of documents that newcomers learn the cultural norms and mores and the communication patterns of subspecialties through which they rotate (Østerlund 2008, 216). My analysis contributes to scholarship on documents and paperwork in clinical settings-as well as more generally-since I demonstrate how the presentation form promotes coordination, discipline, and socialisation; however, I also demonstrate how documents can generate friction among those who use them, and uncertainty about how these documents translate information.
Forms are a specific genre of documents that are ordered and organised by section. Following the recommendations of Marie-Andrée Jacob (2007, citing Ben-Ari 1994), my analysis looks at the form and not just through it. Prior research has demonstrated how forms permit the highlighting of important information and thereby render other details less salient (Goodwin 1994;Heimer 2008;Mertz 2007). In this article, I suggest that these forms render other details invisible within the purview of TEFs since the forms train TEFs to think in institutionally sanctioned ways so as to recognise certain phenomena but not others. Even before the supervisors implemented the form, TEFs may have dwelt on certain details contained in the referral documents only to find that supervisors redirected, questioned, or even cut them off in order to divert their attention to certain details, thus sidelining others. Whereas the boundaries of knowing about a patient's case had previously been discursively produced through interactions with their supervisors, the form was now a textual way of establishing the same boundaries.

The process of identifying the patient voice
Beneath the sections about the referring professionals and institutions and the patients who were being referred, the TEFs encountered a series of tick boxes used to identify the presence or absence of various indicators of patients' conditions. These tick boxes were segmented into aspects of illness, sickness, and disease, reflecting the centre's adoption of Arthur Kleinman's (1981) theoretical framing of physical and mental suffering. This mental health centre was uniquecertainly in France, and perhaps even more broadly-in its anthropological orientation and its attention to the explanatory models of patients. Over the course of their apprenticeship, TEFs became well acquainted with Kleinman's scholarship and the terms illness, sickness, and disease. Sophie, 1 a graduate student in psychology visiting from Austria, commented on the structuring potential of these forms: I thought it was quite helpful because you could really structure yourself and really focus on the main information of the case because, sometimes if you have a dossier with a lot of information inside, you would get lost in the details.
Indeed, a dossier of referral information could be overwhelming for TEFs, and the form served to distil the information that senior clinicians valued most, while rendering other details invisible.
Senior clinicians instructed TEFs to use the sickness sections to identify the systemic factors that might cause or exacerbate mental anguish, as well as impede treatment. Homelessness, failed asylum applications, and experiences of racism or religious discrimination are the kinds of sickness factors that TEFs were to identify. TEFs used the illness section to identify the patient's voice, or the language that patients gave to their experiences of suffering. The sections on cultural representations, adult personality disorders, and defence mechanisms allowed TEFs to connect their graduate training in clinical psychology to the specific cases of patients.
Most TEFs, however, struggled with the form, or at least with the illness and disease sections, and found it difficult to identify the patient's voice in the referral documents which were, after all, written by the referring professionals and not the patients. Form-filling was undertaken in the absence of the patient or the referring clinician, raising questions as to how TEFs could learn what was at stake in the local worlds of patients by engaging with documents alone. Indeed, TEFs faced uncertainty due to the limits of their own knowledge concerning the realities of patients, as well as the limits of what was knowable about patients given their physical absence, and the reliance on paperwork as evidence instead. This represents a form of uncertainty distinct from what Fox (1957Fox ( , 1980Fox ( , 2000 observed, since the available knowledge about a patient's condition was constrained by what referring professionals shared.
Most of the information on the form was not easily reducible to boxes to be ticked or short blank spaces to be filled. When looking at forms TEFs had filled out, I often saw more blanks than completed spaces. Nathalie, a TEF who was studying to become a psychologist following an earlier career in advertising, commented that parts of the form were never used: It's too bad because you have the top part of the front page with the principal information, and then you have that part on the bottom that you never use … I found it was difficult to fill that out because I didn't think it was practical.
By looking at the form and not just through it, I contend that the blank spaces were just as significant as the spaces that were filled in. These blank spaces reflected TEFs' uncertainties about what could be known about a patient, given that they were only dealing with the paperwork sent by mental health and social service professionals. In other words, the task of finding the patient voice was constrained by the voice of these professionals. The blank spaces also reflected TEFs' uncertainties regarding their own knowledge about psychopathology, anthropology, and the social determinants of health.
TEFs also expressed confusion over the boxes on the form, such as the difference between 'social' and 'societal' determinants, and among 'magical-religious', 'spiritual', and 'traditional' values, and there was very little instruction offered by senior clinicians as to how to identify these details in referral correspondence. One TEF, Adrien, who was pursuing his studies in psychology after completing a PhD in chemistry, commented: Adrien suggested that TEFs lacked the 'detailed anthropological knowledge' needed to complete the form, and that there was an asymmetry of expertise since only the most senior clinicians might be able to disentangle this information. This asymmetry potentially posed an important challenge to the community of practice of this centre, since TEFs lacked the exposure to patients which would allow them to see as their supervisors saw. Rather, all they had to base their assessments on were the details in the referral documents. Adrien's comments reflect the dubious translation of anthropological knowledge into information that could be entered onto the form. Even if TEFs could parse the elements of illness, sickness, and disease in an abstract sense, they struggled to identify them in the information provided by the referring professionals. Sophie explained: I am not able to distinguish between them [elements of illness, sickness, and disease]. I mean, I see the distinction, but I don't really see the distinction when I read the referral information.
Taken together, Adrien and Sophie's comments suggest that TEFs were unable to translate anthropological knowledge but that their supervisors were. Perhaps this was because their supervisors had ongoing, regular patient contact, whereas the TEFs did not. What do these forms achieve if those tasked with using them find them indecipherable? Sociologist Carol Heimer (2008, 38-41) has identified how good forms structure attention when there is little uncertainty about what needs attention. Moreover, in a legal context, Rashmee Singh (2017, 512-513) has suggested that forms can distil complex activities into easily actionable tasks and, therefore, transfer expertise to administrators. In the present context, however, there was a lot of uncertainty about the material that needed attention. Moreover, as Adrien stated, the expertise seemed to lie with the head of the institution and was therefore not distributed to the others.
I assert that using forms to address complex matters where there was a great deal of uncertainty generated new forms of uncertainty for TEFs. I argue that these forms, in an absence of patient contact, invoked unnecessary uncertainty about how to view patients' realities and generated tensions between TEFs and supervisors. Supervisors implemented these forms to structure the attention of TEFs and streamline their information-gathering activities; however, TEFs had a tough time identifying how the information in the referral documents corresponded to the details requested on the form. Indeed, as Heimer (2008, 38) has stated, forms used in instances in which there is uncertainty about what needs attention, 'would be a maze of complicated skip patterns and subsections, most of which would be irrelevant to most of the users'. This was clearly the case as some of the TEFs suggested that they would simply not use certain sections of the forms or would not possess the requisite knowledge to be able to fill in certain details.
TEFs frequently lamented that they spent most of their time reading through referral documents and filling out these forms, especially since it came at the expense of shadowing the centre's clinicians. In fact, TEFs questioned the importance attached to the title of 'therapist-in-training', since the level of responsibility it actually conferred did not seem to live up to the title. One TEF, Aurelie, a student in psychology who grew up speaking Portuguese in addition to French, commented: I had the impression that 'therapist-in-training' gave us importance on paper.
Officially, we are therapists-in-training. It sounds nice. We are reassured with that title. But actually we did not do a whole lot of things from a therapeutic point of view. We were therapists-in-training, but we did not do a whole lot of therapy.
I asked one of the psychiatrists at the centre why they used the term 'TEF' rather than stagiaire [intern], a label much more frequently encountered in other clinical training settings in France. The psychiatrist, Fouad, who had studied medicine in Morocco before coming to France, stated that TEF was an intentional moniker: The word 'intern' in France has a very passive connotation. They are there to watch, to observe, and it is almost as if they are a bother. At most, they are given banal tasks like photocopying, but nobody looks after them. They are taken on for internships, but nobody looks after them like people. A therapistin-training is someone who is engaged in a relation of communication in an institution, has the capacity to speak with therapists, to learn one's job, and eventually participate in the care of patients. Larchanché (2020, 132) has discussed the intentionality of this title for therapist apprentices in the centre. I contend that by attending to the often-overlooked perspectives of therapist apprentices, we can see that TEFs take a critical stance on the title and the roles it appears to confer. Fouad's comments seem to closely resemble the distinction made by Lave and Wenger (2009) between observation and legitimate peripheral participation, the latter of which involves absorbing and being absorbed into the culture of practice. However, perhaps a key word in Fouad's statement is 'eventually', since TEFs universally commented on their lack of clinical exposure and supervision as being because they were only able to observe patient evaluations on an irregular basis. The ability to observe therapy sessions with patients required the permission of the clinician and the consent of the patient. Yet, as one of the psychologists described, clinicians did not often give this permission, or if they did do so did not extend this beyond the first appointment. This psychologist, Pamela, who trained in psychology in France after having emigrated from Canada, further commented: The institution cannot impose supervision … and it cannot compel the therapists to take TEFs into their consultations. So, all psychologists and psychiatrists are free to choose to have a TEF or not in their consultations. Very often, practitioners are not comfortable with that. They don't mind doing it for a first appointment, or an evaluation, but having a TEF in consultations over time can make a lot of practitioners feel uncomfortable because they believe it's not a typical way of doing things.
Pamela also explained that many TEFs felt deceived at not having sufficient clinical contact since, after all, they came to this centre with the intention of seeing patients because they were studying to become clinicians.

Learning to be good colleagues through the form
Fouad stated that one of the reasons for choosing the term TEF was so that these individuals would 'have the capacity to speak with therapists'. By reading through referral documents and filling out forms it would seem, however, that TEFs learned more about the professionals and institutions referring patients than about the patients themselves. While instructed to find the patient's voice in a referral document, TEFs instead found the professional's voice about a patient. So, instead of identifying the patient's voice, TEFs identified the voice of the professional, as well as the assumptions, values, and capacities of external professionals and institutions. Nathalie commented: The paper, it's a professional who wrote it, it's a letter, it's paper, but through the words you are able to perceive the counter transference of the professional … And I think it's really interesting and it's not what we see in our exercises in class. Generally, the clinical cases are neutral.
By engaging with the referral correspondence material, TEFs learned that seemingly neutral case studies that they might learn in their graduate classes were anything but when written by health, social service, and legal professionals who have their own biases and constraints. Nathalie continued: It's happened several times when I have remarked that, 'The professional, I think they are taking sides.' For example, there were letters written by psychologists, and it was really clear that they wanted to unload: 'The adolescent is getting on our nerves. We cannot take it any more.' They only gave the faults of the adolescent and I think that the psychologist is taking sides, they are not neutral, they are against the adolescent and want to get rid of the adolescent, it's not a neutral analysis.
Getting rid of patients may be a characteristic practice in a variety of health service contexts, leading to resentment when health professionals perceive their peers to be shifting their work onto others (Dodier and Camus 1997;Mizrahi 1985). Just as Patrick Castel (2005) has suggested that patients may serve as a resource for healthcare workers to learn about the practices of their peers, I contend that the referral documents in this context taught TEFs about the functions, resources, and intentions of referring institutions and professionals. Put simply, the form instructed TEFs how and how not to act towards their future patients and speak to their future colleagues.
Of course, referring professionals may be well-intentioned and may think the expertise of the centre's clinicians will truly benefit a patient. Adrien remarked: It's shocking, because sometimes you see letters and people say openly: 'We need someone who understands Africans, because this patient is African and we don't understand his problems.' Okay, the situation sounds funny, but if you think of it, it's not absurd. If you have a healthcare professional who is not able to help the patient, for one reason or another, I mean, maybe there is another reason but basically it doesn't work. And the person says: 'Okay, I've tried everything, and this person is African, I will send this person to a centre where they know, where they have a good handle on culture, cultures.' It's also oversimplification, there are a lot of stereotypes with this decision, but why not? They're trying alternatives.
Larchanché (2010, 335) has described how external professionals' referrals of West African patients to this centre reveal these professionals' generic or problematic notions of culture and cultural difference. I suggest that these referrals, whether problematic or not, have produced important pedagogical opportunities for TEFs, who have learned to read beneath the surface of referral documents, and for senior clinicians, who have attempted to discourage TEFs from using the same problematic language or notions of cultural difference that external professionals often espouse. In other words, by interacting with paper patients and not actual patients, TEFs learn to become good future colleagues more than they do good future clinicians.
In an edited volume on the ways in which documents represent artefacts of modern knowledge, Annelise Riles (2006, 7) has suggested that a document is simultaneously an ethnographic object, an analytical category, and a methodological orientation. By moving between the referral documents and the form, TEFs undertook a practice akin to ethnography whereby they explored the worlds of patients and referring professionals. In the words of Matthew Hull (2012), documents promote control because they coordinate perspectives and have the capacity to construct subjects. Donald Brenneis (1994) has emphasised how documents promote discipline among future users since they must anticipate future audiences. Returning to the present context, Riles (2006), Hull (2012), and Brenneis (1994) provide us with conceptual insights to understand how TEFs use referral materials and the forms to generate a kind of patient and professional, neither of which is neutral, as is often portrayed in textbooks. Instead, the documents generate a kind of patient whose migration history may have an effect on their psychopathology and a professional who might demonstrate problematic approaches to supporting these patients. These documents were instructive for TEFs, who were being trained by senior clinicians to discipline the ways in which they think and speak about immigrant patients pursuing psychotherapy so that they are legible to their future colleagues. The process of disciplining became especially apparent when TEFs presented patient cases to the senior clinicians during assessment meetings, described in the next section.

Learning to speak like a master clinician
Acclimatising to a community of practice necessarily requires apprentices to learn to speak like their supervisors. TEFs did not simply read referral materials and inscribe specific pieces of information onto forms; they used these forms to structure the clinical narrative of patients, which they presented to their supervisors and peers during assessment meetings. Since shadowing opportunities were rare, these assessment meetings represented the principal point of contact between TEFs and senior clinicians. TEFs therefore needed to appear prepared, professional, and competent in front of clinicians. TEFs used to read directly from the referral files during these meetings, often rambling for several minutes and touching on information senior clinicians considered tangential. These meetings only lasted one hour, during which several cases often had to be presented, so TEFs needed to summarise the details about referred patients' medical and social histories within two or three minutes. The form (Figure 1) not only served as a datagathering framework but also as a discursive itinerary so that TEFs would not deviate during their presentations. As described in other ethnographic accounts of case presentations in medical training (Anspach 1998;Atkinson 1995;Bosk 2003;Holmes and Ponte 2011;Light 1980;Prentice 2013;Schön 1983), presentations needed to be concise, to the point, and should illustrate the presenter's line of reasoning. These studies have also demonstrated that thorough and effective case presentations are an important step in carrying oneself as a future clinician.
My research presents a unique complement to these studies, which focus on the socialisation of medical students and residents, since I examine the language of socialisation among budding therapists. These studies have considered how future doctors take various steps to minimise uncertainty (Fox 1957(Fox , 1980(Fox , 2000, but the training at the centre presented a significant contrast: TEFs were encouraged to embrace uncertainty, speculate about the unknown, and ask questions without fear of asking dumb questions. Despite the seemingly open and flexible nature of this training environment, TEFs frequently commented that senior clinicians could be quite rigid in the ways they expected TEFs to speak when presenting the clinical narratives assembled using the form.
The form provided TEFs with a checklist for the kinds of information that senior clinicians valued. However, referral documents rarely contained this information and, as described above, TEFs had a hard time translating this information. As the sections below demonstrate, TEFs' presentations generally followed the organisation of the form. Just like in influential ethnographic accounts of case presentations, senior clinicians often interrupted the presentations to ask for additional information or to seek clarity. TEFs could be reprimanded if they deviated too much from the form, or if they relied too heavily on it. In the remainder of this article, I demonstrate how TEFs' uncertainties about the form became public, since the assessment meetings were a space where TEFs learned to discipline their speech through peer observation, affirmation by senior clinicians, and their fear of being castigated. In the next section, I walk through an example of one such presentation.

Presenting Ahmed
Let us consider a therapist apprentice's presentation of a 16-year-old adolescent we'll call Ahmed. 2 A nurse and social worker in the high school that Ahmed attended referred him to a psychologist at the centre. Adrien, the TEF who presented this referral, stated that Ahmed was originally from Bangladesh and had arrived in France after a ten-month journey through several countries, including Turkey, where he had spent seven months. The following extract is a reproduction of the exchanges between Adrien and Fouad, the psychiatrist: Adrien's presentation follows the order and vocabulary of the presentation form, and he used the terms illness, sickness, and disease. Learning to appropriately deploy these terms reflected an important component of acclimatising to a community of practice. This kind of vocabulary gives narrative structure to a patient's story (Good and DelVecchio Good 2000, 54). In the present context, these terms were meant to structure the ways that TEFs presented cases and, more generally, how they think about the life histories of patients.
Fouad's statement, 'Avoid the exotic and get back into the clinic', illustrated how he redirected Adrien's thinking so as to avoid making incorrect assumptions or over-interpreting a situation. Donald Schön (1983, 121) has identified how those who are at the beginning of their training in psychotherapy tend to leap to interpretations that are not based in the symptoms or the statements of the patients. Similarly, Kathryn Montgomery (2006, 123) has explained how students at the beginning of their clinical training tend to think of the most rare and exotic illnesses when they are confronted with the symptoms of a patient; the expression 'don't think of zebras', reflects the advice that supervisors give their students to avoid thinking of the most exotic animal when hearing the sound of galloping. In presenting Ahmed's case, Fouad told Adrien to avoid focusing on the unknown elements, and instead return to those that were more clinical and identifiable. The invocation 'avoid the exotic' could be considered as a request to avoid imprecise or potentially stigmatising information, and to focus on the elements that can eventually help in diagnosis. Fouad's instructions reflect an important point of tension and uncertainty. The TEFs were dealing with paper patients and were limited to the information contained in the referral documents. TEFs were encouraged to speculate but were also told to avoid exotic interpretations. The next section illustrates how supervisors invited TEFs to speculate and improvise in their speech about patients, while also correcting them when their speech deviated too far from norms. Indeed, this tension represents an important component of apprenticeship in communities of practice, where apprentices oscillate between improvisation and the necessity of adhering to authoritative standards.

Speculation and anticipating authority
During the presentations, supervisors questioned TEFs or encouraged them to try out and debate different ideas about referred patients' histories and diagnoses. In the example above, Fouad asked Adrien to speculate about Ahmed's condition. By asking TEFs to speculate, supervisors invited them to demonstrate their knowledge about the known and unknown elements of patients' histories and anticipate how this information, taken together, could be used to construct a clinical profile.
Although TEFs were asked to speculate there was a lack of explicit instruction regarding how to present information, and this was a significant source of uncertainty for TEFs. It was curious to hear how supervisors thought that TEFs learned to conduct these presentations. Pamela, one of the psychologists, explained how TEFs learned: I would say piece by piece … the reality is that there isn't always someone present that is going to be sure that the person who comes in knows from A to Z, everything, so the TEF is getting information, left and right, little by little.
TEFs described learning by observing and mimicking their peers, particularly those with more experience. For instance, Adrien said that they learned: By watching and listening. But, gradually, more precise instructions were given because the professionals were not always satisfied with how the files were presented. So gradually they started giving us clearer instructions like, 'we want this and this and this to be featured in the presentation'.
This form ensured that certain ways of thinking and speaking took precedence over others. Indeed, the supervisors had implemented the form as a checklist because TEFs' presentations did not adhere to the wishes of supervisors. In an absence of explicit instruction, apprentices often learn through a subtle, informal, and hidden curriculum (Hafferty and Franks 1994). While the form was an intentional and textual rendering of the kinds of information supervisors wanted, learning how to use the form properly required more informal observations of more experienced users.
In an account of how law students learned to think like lawyers, Elizabeth Mertz (2007, 77) described the multiple functions of speculation: to alert students that there is more taking place beneath the legal texts than is immediately obvious; to encourage students to be more aware of the strategic effects of proceeding in one way or another in their arguments; to initiate students into a particular genre of storytelling; and to further the opening up of legal readings to a wider array of cultural stories about why things happen. Mertz's analysis of speculation is helpful in the present context as it illuminates how TEFs were attempting to make sense of the scripts and forms of reasoning expected of them by their supervisors. Indeed, just as the first-year law students in Mertz's account were 'learning to unravel the cultural logics' of the legal profession, the TEFs in the centre were learning to anticipate the kind of authority they would need to perform as future therapists. During their presentations, TEFs were expected to be able to take the information contained on the forms and present it in a concise and coherent manner. Writing about nursing expertise, Sally Candlin (2002, 191) has suggested that establishing coherence requires individuals to make connections between sentences and sequences of sentences, and to determine how these connections fit with the overall framework of the activity. And as Cheryl Mattingly (1991Mattingly ( , 1000 has described, expert practice among clinicians involves their ability to put all the elements of a patient together in a narrative form. Drawing on the perspectives of Mertz (2007), Candlin (2002), and Mattingly (1991), I suggest that TEFs needed to read beyond the texts of the referral documents and anticipate the narrative structures their supervisors expected of them in order to speak with authority in front of their supervisors and peers. By using the form, TEFs were to first develop and then deploy the institutional vocabulary of the clinic. But I also contend that in the physical absence of patients TEFs found it challenging to implement this institutional framework since their presentations of patients were based on the information provided by the referring professionals. Their presentations and ability to speculate were hampered by the uncertainty of how to translate knowledge about patients written in the voice of professionals, and in the absence of patients.
Speculation was both freeing and constraining. It authorised TEFs to perform their knowledge, and it permitted supervisors to correct them, as described by Pamela, the psychologist: In asking the TEFs to speculate … it makes them think for themselves, it gives them the right to be, and … it's a way to readjust perhaps some people's thinking or beliefs or whatever the case may be. The TEFs need to be able to justify why they're saying that, but I also believe that if it doesn't fit with the framework of the institution, it will be known, the institution will try to correct that thinking.
Pamela's comments perfectly capture the tension between improvisation and authority. TEFs were encouraged to speak freely and 'think for themselves', yet they risked being 'corrected' when their speech did not adhere to the institutional vision and the norms of the clinic. Asking TEFs to stick to the facts contained in referral documents, but then to speculate-or essentially deviate from the factsrepresent contradictory imperatives. Instructions to not be too 'exotic' might generate uncertainty and confusion for TEFs, particularly since their presentations were informed by the material that was presented by referring professionals and not by the patients themselves.
Writing about improvisation and authority in clinical contexts, Laurence Kirmayer (1994) reminds us that authority permits improvisation but also constrains it. In the present context, by asking therapist apprentices to speculate, supervisors encouraged TEFs to improvise in their understandings of patient cases. However, supervisors simultaneously confined the possible range of improvisation to the kinds of thinking that were in line with the framework of the institution. Supervisors thus constructed what Charles Bosk (2003, 94-5) refers to as a 'binding definition of reality', where they set the limits about appropriate forms of improvised knowledge and challenge ways of thinking that might fall outside of these limits. Just as the form constrained the complexities of patient histories to tick boxes and short answers, the case presentations limited TEFs' speech to institutionally authorised forms of therapeutic talk. As the next section demonstrates, therapist apprentices found the tension between authority and improvisation to generate uncertainty about how to speak about patients. This tension also generated fear of being reprimanded by their supervisors.

Unlearning and the culture of fear
The interactions between TEFs and supervisors could at times be confrontational. TEFs including Aurelie, whose words appear below, commented that they were often anxious about presenting information in front of their peers and supervisors because of the fear of being disciplined or 'torn down': [One of the supervising psychiatrists] asks our opinions and we have a lot of liberty, but he freaked me out a bit [laughs] because I told myself that if I say something stupid, he is going to tear me down … he corrects people, and his way of doing it scares me. I find that it's not always friendly. But it's good for those who dare and who are not afraid to put themselves out there in a room full of ten people. It's better because they learn a lot faster. They think something, they propose a diagnosis, but then are told: "Oh no, it's not that, because there is this or that element, so it's not that at all." And the next time, they won't make that mistake. I think that because I don't dare as much, I retain less.
As Seth Holmes and Maya Ponte (2011) have illustrated, fear was often instilled in medical students in order to manage their language use and so that they would more easily be identified as a professional. While it might seem unsurprising that supervising clinicians in the present context would use similar kinds of instructional methods as in other clinical education contexts, what was paradoxical was how these supervisors thought these methods would encourage TEFs to unlearn the rigid ways of thinking they had acquired in their coursework and clinical training. While some TEFs could see value in the confrontational methods of their supervisors, others, like Sophie, pushed back against the notion that presentations needed to follow a rigid framework: For me, there is no wrong way of presenting; it's just a different way of viewing the issue. So sometimes I felt that they [supervising clinicians] were a bit too strict in their opinion.
Fouad, assuming the voice of a hypothetical TEF, justified the importance of making mistakes in while training and not later on as a professional: If I am training, that means that I am here to learn. I ask questions and I am not afraid to ask questions and I am not scared to make mistakes because I am in training. It's better that I pose a dumb question now, because that means that I will not ask it later.
That only some clinicians allowed TEFs into their consultations led TEFs to question whether the styles of practice of those who did not do so were aligned with the framework of the centre. Maria, who had trained in psychology in France after completing a doctorate in sociology in Spain, commented: I know that there were other therapists who I did not get to see work … that's why I think that the therapists at the centre, when they close their doors, everyone does what they want. It's not necessarily what they present to us.
This latter point represents an important concern of TEFs and an important challenge to the community of practice within this centre: how can apprentices become members of communities of practice if their supervisors do not let them fully participate in their practice?
The discrepancy between what therapists tell TEFs they do and what they actually do is a barrier to legitimate peripheral participation and the full immersion that apprenticeship entails.

Conclusion
In this article, I have argued that paperwork, in the absence of contact with patients, provoked uncertainty about how to view the realities of patients, and generated tensions between thérapeutes en formation [therapists-in-training, or TEFs] and their supervisors. I build on Fox's (1957Fox's ( , 1980Fox's ( , 2000 framework of uncertainty, which identified how uncertainty is an inevitable and generative component of clinical training. My analysis of 'paper patients' expands this framework by suggesting that a new form of uncertainty-which relates to translating a patient's lived experience through a series of documents-may not necessarily contribute to the development of therapist apprentices and may instead generate tension with their supervisors. By undertaking an ethnography of apprenticeship and documents, I analysed how budding therapists learned to use a presentation form and how this form became a socialising tool in the centre. This ethnographic approach foregrounded the experiences of TEFs within this centre's community of practice. Following the seminal work of Lave and Wenger (Lave 2011;Lave and Wenger 2009;Wenger 1998), this study has emphasised apprenticeship over training or knowledge transmission, since it considers how learning and membership within a community of practice were mutually constitutive.
This article contributes to scholarship on the coordinating and controlling function of documents, as well as scholarship on uncertainty and the socialisation and discipline of budding clinicians. Scholarship on documents-and the genre of forms in particular-has emphasised that documents may structure attention, distribute expertise, plan future action, coordinate perspectives, and promote discipline. This article highlights how many of these functions may be limited by the nature of the knowledge being considered and the nature of the form that is intended to capture this knowledge. My analysis demonstrates that when attempting to simplify complex knowledge about the human experience by transferring it to a form, much of this knowledge is lost in translation and the form can be unintelligible to its users.
Scholarship on socialisation and uncertainty has tended to foreground the experiences of medical students and junior doctors. My article expands this body of research by including the perspectives of budding 'psy-' professionals. Supervisors instructed TEFs to embrace the complexity of their patients' lives. TEFs undertook paperwork routines, which were meant to manage uncertainty by permitting them to identify the patient's voice. They were also designed with the purpose of enabling the TEFS to use a lens of medical anthropology in order to see how the mental health conditions of people may be impacted by structural violence. In the physical absence of patients, however, these paperwork routines generated uncertainty that was not necessarily overcome during TEFs' clinical apprenticeship.
By focusing on seemingly mundane tasks-the reading of referral documents, the inscription of patient information onto forms, and the presentation of this information to peers and supervisors-I contend that the work of paperwork and the presentations of TEFs revealed a great deal about the centre's community of practice, the uncertainty faced by TEFs, and the tension between improvisation and authority in therapeutic thought and speech. The structure and content of the form revealed how TEFs were meant to train their attention to prioritise certain patient details, while rendering others invisible. This had the effect of reducing complex information to boxes and short answers. Scholars (e.g., Heimer 2008; Singh 2017) who have examined forms in clinical and social service settings have demonstrated that forms should be used when there is little uncertainty about their use so as to ensure that expertise can be distributed from those with more experience to those with less. That was not the case here. TEFs questioned these forms and the kinds of work that they were supposed to perform. Their perspectives suggest that reducing complex life histories to tick boxes and short answers was antithetical to their supervisors' messages of embracing complexity and cultivating openness. By looking at the form, and not just through it, I contend that its blank spaces also revealed TEFs' uncertainties concerning their knowledge and what was knowable about the patient from the referral documents. These blank spaces reflected the problem of translating knowledge about patients since the referral documents were written in the voice of professionals and TEFs rarely encountered the patients they read about.
The centre and the form used Arthur Kleinman's (1981) framework of illness, sickness, and disease, and supervisors intended TEFs to use this framework to identify the explanatory models and the voice of patients. TEFs' difficulty with the form suggests that this complex anthropological knowledge was not easily translated to the restrictive, brief responses it sought to elicit. Since TEFs dealt mostly with paperwork rather than patients, they rarely had the opportunity to see how this anthropological knowledge may translate to actual lived experience. Over the course of their apprenticeship, TEFs might have observed at most a few initial patient evaluations per week. They thus lacked the ability to follow patients on a longer-term basis, to learn how this anthropological knowledge may inform clinicians so that they can provide better support to patients whose mental health conditions have been exacerbated by structural violence.
Supervisors instructed TEFs to find the patient's voice, but in the absence of physical patients TEFs found that the professional's voice extinguished that of the patient. Indeed, TEFs found it difficult to identify the patient's voice in their physical absence, and in having to deal with what was effectively a substitute-a paper patient. By learning to identify the voices of professionals, rather than patients, TEFs were learning how to become good colleagues who would not offload patients onto others or make problematic referrals.
While supervisors gave TEFs the liberty to speak freely and speculate, they also policed the speech of TEFs when it did not adhere to their guidelines-guidelines which they did not formally or explicitly outline. The case presentations during the assessment meetings, such as the one for Ahmed, reinforced the tension between the acceptance and minimisation of uncertainty. Supervisors wanted TEFs to unlearn the rigid ways of thinking about mental illness that are often acquired in clinical training and coursework. But despite encouraging TEFs to be creative as they speculated about patient cases, the supervising clinicians often used the same confrontational methods employed in the clinical training that they critiqued. While seemingly promoting a workplace culture of openness and exchange, the pedagogical method of supervisors instilled fear among TEFs.
The case presentation sessions were both freeing and constraining. By asking TEFs to speculate, supervisors encouraged them to deviate from the details contained in the referral documents. Supervisors also corrected TEFs when their interpretations veered too far from the centre's norms. In an absence of explicit instruction, TEFs expressed that they had learned to use the form and speak in front of their supervisors by observing their more experienced peers. As I have argued, in an absence of patients, the forms, and the kinds of speech that their use promoted, provoked uncertainty among TEFs concerning the realities of patients, and they generated tension between TEFs and their supervisors.
This article may seem critical of the supervisors in this centre. Clinical apprenticeship ethnographies that foreground the perspectives of apprentices often do come across as quite critical. My intention has been to centre the perspectives-and this includes the concerns-of the TEFs. As an anthropologist I admire the approach of identifying the voice and explanatory models of patients, and I find it refreshing that this centre has adopted a framework from medical anthropology. Based on this apprenticeship ethnography, I wish to offer a few modest recommendations to more closely incorporate TEFs into the centre's clinical work. First, to the extent possible, TEFs should be given more opportunities to shadow their supervisors, particularly their interactions with patients. This will allow TEFs to see how supervisors actually practise an explanatory model approach, rather than just hear it described. Second, given how much time TEFs spend with referral documents and this presentation form, supervisors should spend more time with TEFs looking at and discussing it. Conversations about the sections of the form, especially those that are often left blank and those that reduce complex information to tick boxes and short answers, might help render more explicit the intentions of supervisors.
Universität Siegen, and I am grateful for the invaluable input from the conference participants.

About the author
David Ansari is a lecturer in Sociocultural Anthropology in the Department of Anthropology at Washington University in St. Louis. His research and teaching interests centre on the intersections of migration and health, trauma and psychotherapy, health services and organisations, and clinical and therapeutic training. David's book Therapeutic Apprenticeship: Migration, Belonging, and Mental Health in France examines generational differences in perceptions regarding multiculturalism and diversity among psychotherapists in France. By attending to apprenticeship, David examines how newer generations of psychotherapists develop embodied caring and clinical skills under the guidance of supervising therapists. Therapeutic Apprenticeship also reveals generational differences between supervisors, many of whom came to France as immigrants, and their apprentices, many of whom were born in France with parents or grandparents who were born abroad. These generational differences represent evolutions in thinking about and enacting inclusivity, both in clinical contexts and beyond.