Medical teachers’ discursive positioning of doctors in relation to patients

Abstract Context An important part of a doctor’s identity is the social position he or she adopts relative to patients. Dialogic theory predicts that medical school discourses influence the positions students incorporate into their professional identities. As this may affect how students later exercise power in doctor‐patient relationships, we set out to examine how medical teachers position doctors in relation to patients. Methods Informed by Holland’s Figured Worlds theory, which draws important assumptions from Bakhtin, we chose dialogic research methodology to examine how educators’ language positions doctors and may influence students’ identity formation. We recruited a maximum variation sample of 10 teaching staff and used open prompts in individual semi‐structured interviews to elicit discourses of doctors’ social position. We used Sullivan’s dialogic methodology reflexively to identify informative speech acts (utterances) and to examine how the language used in these constructed doctors’ positions. Results Dominant discourses of Social Superiority, Technical Effectiveness, and Benevolence elevated doctors’ positions based on their social status, applied knowledge and trustworthiness, respectively. These positions were defended by predicating medical care on doctors’ mastery of treatments and their superior knowledge. A non‐dominant discourse of Distributed Power and Responsibility narrowed the positional gap by constructing doctors as empowering patients. Conclusions Whereas three conservative discourses upheld doctors’ elevated social position, a non‐dominant, transformative discourse distributed power. We suggest that doctors will form the best relationships with patients when they are aware of these discourses and know how to navigate them. In pursuit of effective and compassionate patient care, we commend critical pedagogy as a means of articulating non‐dominant discourses and increasing students’, educators’ and doctors’ awareness of how they learn the positions of doctors.

dents can respond individually to what they hear and see. They 'author' their own identities from the various positions offered to them by the many 'figures' they meet. This has important educational implications because it shows how the use of language can influence identity formation. 15 Critical pedagogy has recently been proposed as a means of examining, critically, how language enacts discourses of power, privilege and position, and how these influence professional identity. [16][17][18] The implementation of critical pedagogy in medical education, and thereby students' identity formation, could be promoted by a clearer understanding of how curriculum talk influences social positioning. [19][20][21] The purpose of this research was to help students develop identities that best serve the needs of patients. 3,7,[12][13][14] The research question was: How do medical teachers position doctors in relation to patients? We chose Figured Worlds theory and its companion qualitative methodology, dialogic analysis, to provide a robust methodological platform on which to answer the question.

| Ethics
The Joint Research Ethics Committee of the School of Medicine, Dentistry and Biomedical Sciences, Queen's University Belfast approved the project (approval no. 15.19).

| Setting
The research context was Queen's University Belfast (QUB), a research-intensive university, which had the only medical school in Northern Ireland, a geographically bounded region of the UK. Its 5-year undergraduate medical programme provides 2 years of clinical science education with some clinical contact followed by 3 years in clinical settings.

| Conceptual orientation
We chose a qualitative approach because this research set out to explore a complex social phenomenon. We took a critical realist position, which is well suited to exploring power relationships and the negotiation of identity. 22 This position does not seek to reveal a 'real' world in the usual sense, but it allows researchers to use the critical analysis of language as a means of understanding social practices, institutions, structures and relationships.
More specifically, the analysis was informed by Bakhtin's theory of dialogism, which is considered to have useful practical applications as well as being conceptually rich. 4,22 Dialogism takes the ontological position that people rely on others to develop a sense of who they are, but they also have a sense of their own creative potential.
Speech is central to Bakhtinian thinking because it mediates sociocultural influences, the formation of individual identity, and the exercise of power. 'Social speech' is in continuity with 'inner speech.' In both

Key Message
Encouraging students to be critical of teachers' language could encourage them to form sharing relationships rather than adopt elevated positions over patients. the social and individual contexts, people are engaged in multiple dialogues. The ability to author the many voices they hear gives people social agency and individual identity. By contrast with, for example, cognitive theories, dialogic theory regards speech as the exchange of lived ideas, including personal values and judgements, rather than of purely abstract concepts. The concept of 'addressivity' has an important place in Bakhtinian scholarship. According to this concept, people are in a perpetual state of addressing others and being addressed by others. Building on Bakhtin's work, Figured Worlds theory holds that, when we speak, we anticipate how the speech of others will author our identities and we author our own speech accordingly.
Dialogic methodology provides valuable possibilities for education research, which the distinction between 'small-d' discourse and There is no single 'correct' way of actioning these theoretical insights, just as there is no single 'correct' interpretation of any one speech act. We chose to use Sullivan's dialogic approach to guide this research. 22 Sullivan proposed that researchers should regard language as being imbued with speakers' intentions and desires, which are themselves entangled in social structures. Researchers using Sullivan's methodology, approach qualitative analysis with several core assumptions. One is that speech is 'language in action. ' 15 Another is that speech involves relations of power, 23 and another is that language involves unconscious desire. 22

| Research team
Two QUB medical students (SRB and a classmate), who had completed 2 years of the programme, a former internist and education researcher working in QUB (TD), and an Irish family practitioner and education researcher now working in Canada (MK) conducted the research. Two were men and two were women. Their cultures of origin (from Singapore, Switzerland, the Republic of Ireland and England), as well as their ages and professional positions, provided varying sociocultural perspectives.

| Design
This research explored how the language used by medical teachers in one-to-one, face to face interviews constructed doctors' positions in relation to patients. The theoretical assumption that justified this design was that the social speech of a medical curriculum influences medical students' inner speech. This influences students' identity development, which influences how students position themselves in relation to patients.

| Recruitment and participation
The researchers constructed a purposive sampling frame to ensure participants were scientists and clinicians of both sexes, who represented a range of ethnicities, career stages and disciplines. They included primary, secondary and tertiary care practitioners, all of whom contributed actively to the programme. Researchers invited medical teachers, by email, to contribute to an investigation into how medical teachers construct the position of doctors in relation to patients. They explained to prospective participants that the research was being conducted in order to help future medical students reflect critically on how they will position themselves in relation to patients after they qualify as doctors.

| Data collection
Having weighed up whether students, teachers, or neutral third parties should interview participants, the team chose to use students because their relationship with participants most closely replicated the teacher-

| Data analysis
Sullivan's dialogic approach requires researchers to arrive at their findings by achieving a balance between 'bureaucratic' and F I G U R E 1 Interview prompts 'charismatic' approaches. 22 A purely bureaucratic approach would vest authority in rules and procedures, systematic and exhaustive analysis of the transcripts, and researchers having an impersonal relationship with the data. A charismatic approach (for which no detailed guidelines exist) depends on researchers subjectively actualising the research procedures and responding reflexively to the data. We resolved possible tensions between these two extremes by conducting individual stages of data analysis meticulously, working reflexively as a team, and keeping an audit trail, when interpreting the data subjectively. We followed Sullivan's advice not to conduct hierarchical coding, but to extract the speech acts (utterances) that were best able to answer the research question and 'sound bites' that epitomised participants' different positions.
We scrutinised the text for speech acts that exemplified variation amongst participants, contradictions within and between their discourses, and texts that gave doctors more or less status and power in different ways. accompanied by a free text memo explaining the researcher's interpretation. Researchers copied blocks of text, rather than marking up words or phrases, to keep each piece of speech in its discursive context. They coupled these with analytical memos to make the process transparent to other team members. They elaborated the template and reapplied it to all speech acts. MK, who was naïve to the analysis up to that point, enhanced the team's collective reflexivity by giving her independent opinion on a random 20% sample of speech acts.

This analysis identified seven main themes: what doctors do;
what patients do; who doctors are; who patients are; changing times; social structures, and characteristics of the language. TD organised the data into these themes, which helped researchers identify four main discourses, described in the results section. To validate this provisional interpretation, the authors ran a workshop at an international conference, in which participants organised textual extracts into the four named discourses and gave critical feedback on the thematic structure. Then, the authors returned to the original audiorecordings to compare the interpretation against the data in their rawest state. Finally, they coded the data to the agreed themes using NVivo Version 12 analytic software (QSR International Pty Ltd, Doncaster, Vic, Australia) and wrote the narrative that follows using output from it.
Throughout these analytical stages (the details of which were recorded in a research log), researchers repeatedly re-read individual participants' speech acts, compared them against other participants' speech, and sought patterns within and between individual participants' responses.

| Reporting
Dialogic analysis places considerable importance on how findings are reported, which should be in a manner that is rich in quotations and textual features derived from participants' speech acts. The researchers' authorial voice should assemble participants' different voices into a coherent account of findings, highlighting and juxtaposing different positions. 22 Consistent with dialogic epistemology, the final account should not be regarded as an incontrovertible statement of how medical teachers position doctors, but as an interpretation that can generate interest and stimulate debate. It would make the results unreadable if all fragments of speech acts were attributed to individual participants so we have done this (using pseudonym and speech act number) only for longer (italicised) quotations.
We use the phraseology 'patients were …, patients did …' as shorthand for 'participants' discourses constructed patients as …' Table 1 gives some details of the 10 participants but not their specific academic disciplines because this would make individuals too easily identifiable. Participants positioned doctors in relation to patients within discourses of Social Superiority, Technical Effectiveness, Benevolence, and Distributed Status and Responsibility (abbreviated to Superiority, Effectiveness, Benevolence and Distribution, respectively), which we now illustrate. The following quotation exemplifies how, in the Superiority Discourse, social class gives doctors a high social position:

| Challenges to superiority, effectiveness and benevolence
There were many references to how doctors' position is becoming less elevated. One cause was redistribution of wealth. Another was wider availability of education. Another was travel. The main one was availability of knowledge:  positions and identities. 24 Students discursively assume the social position of doctors-to-be when they start interacting with patients. 24 They negotiate tensions between the imagined position of doctors, who have the power to 'make a difference', and the experienced position of students, who are powerless. 25 The positions that students' adopt and the positions they give patients and doctors differ between culturally different parts of the world. 26  To reduce any undue influence of the questions on the findings, we selected text that was not a direct response to interviewers' cues, and used non-directive prompting to probe participants' responses. The one exception was the final question, which gave doctors a position, but used a language construction that implicitly invited participants to agree or disagree with the suggestion made.

| Positions constructed by distribution
The implications of this work for medical education link with another emerging research topic, critical pedagogy. The purpose of this is to help students develop critical consciousness of doctors' and patients' social positions so that they develop the identity of compassionate, humanistic, socially conscious health professionals. 17,18,29,30 We offer our framing of doctors' positional relationships with patients as subject matter for critical pedagogy. We hope this will help students reflect on how doctors can use their position, and the power that comes with it, in virtuous ways rather than the vicious ways they may sometimes observe.

| CON CLUS IONS
We ask students, educators and doctors to reflect on how discourses of social superiority, technical effectiveness, benevolence and the distribution of power and responsibility, influence their relationships and other people's relationships with patients, so that we may all use our positions to best effect.

AUTH O R CO NTR I B UTI O N S
TD conceived the idea for this study, supervised the project, analysed the data, and wrote the article. SRB conducted the fieldwork, participated in the analysis, and contributed to the final manuscript. MK participated in the analysis and contributed to the final manuscript.

ACK N OWLED G EM ENTS
The authors acknowledge with thanks Clarence Ho's, Centre for Medical Education, Queen's University Belfast, Belfast, UK, contribution in conducting interviews and contributing to the analysis.

CO N FLI C T S O F I NTE R E S T
None.