Linguistic-ethnographic perspectives on the operating theatre
The starting point for analysing the materials we had collected was that people use linguistic and a range of other communicative resources, such as gesture, gaze and body posture, to construct social realities. In order to gain insight into these social realities I needed to analyse in detail how people talk and use their body in situated interaction. That's why I audio- and video-recorded interactions around the operating table, rather than making fieldnotes alone: one cannot record the details of interaction as it unfolds in on-the-spot-note-taking. A number of 'sensitising concepts' or theoretical pointers guided my initial explorations of these data.
The first pointer drew my attention to the range of activity types inside the operating theatre. I noticed that while staff in operating theatres frequently engaged in conversations their main focus is on what Goffman (1981) calls 'coordinated task activity'. In this activity speech only becomes 'critical when something doesn't go as expected' (Goffman 1983: 4): surgeons and nurses often quietly coordinate their collaborative work through embodied action, without using speech all the time. It also struck me that bodily arrangements in surgical activity often run contra to conventional social expectations in relation to interpersonal distance, bodily comportment and gaze in conversations (Bezemer 2014).
The second pointer drew my attention to the sequential ordering of actions and spoken utterances during operations. Embodied actions, including gestures made with hands and with instruments, can prompt or be prompted by other actions, or they can prompt or be prompted by a spoken utterance. For instance, a surgeon ceasing to tie knots and holding the end of the stitch in fixed position prompted an assistant to cut the stitch (Bezemer et al. 2011a); and so did a surgeon saying, 'yeah you can cut that'. Mondada (2011) calls such sequences 'paired actions'; indeed they are organised in the same way that, say, an 'adjacency pair' of a question and an answer is. I also considered the affordances of speech and different kinds of actions. For instance, surgeons frequently make references to highly specific points in space; for example when one surgeon suggests to another where he thinks the other should cut. These references can only be achieved in gesture: speech does not allow them to be as spatially precise as one can be with the tip of an instrument (Bezemer et al., 2014).
The third pointer drew my attention to the 'gains and losses' involved in representing the social world inside the operating theatre in writing for specific audiences. For instance, I looked at the 'preference cards' of surgeons, that is, notes detailing for each consultant and for each of the procedures they perform the instruments that they want to have at their disposal. These cards distribute the work of surgeons and nurses in particular ways, and they help understand some of the observed interactions between them (Bezemer et al. 2011b).
While these perspectives draw on a significant body of literature they are underrepresented in medical journals. When we started the research only a handful of studies on learning and communication in the operating theatre had adopted linguistic and/or ethnographic perspectives, and all of them except one (Moore et al. 2010) were published in social science journals (see Weldon et al., 2013 for a literature review). The same applies to my methodological orientation, which is recognised in a range of social science journals yet rarely adopted in medical journals. I engaged with small excerpts from a purposive data sample, rather than coding much larger, representative data samples as other publications on learning and communication published in medical journals typically do (see Blom et al. 2007). My aim was to seek generalisable and falsifiable explanations. In the words of Burawoy, I sought to 'extract the general from the unique' (Burawoy 2009: 21).
Alex did her own analysis of the video-recordings. The audience for her work was the world of surgery. She knew that to be recognised by this world she was also going to have to publish in medical journals. Hence, she drew on a framework of theories cited in medical journals, such as Ericsson (2004); and coded larger data samples including the video-recordings of teaching and learning at the operating table and the audio-recordings of interviews surgical trainers and trainees (see Cope 2013). Notwithstanding our theoretical and methodological differences our partnership continued to be of mutual benefit. In the periods that we each worked on the data we met up on a weekly basis, giving detailed feedback on each other's emerging analyses and ideas. These interactions shaped my analysis in important ways. Rather than going back to the surgeons I studied, who were practically unavailable for post-observational conversations, Alex taught me about inter-professional dynamics in the operating theatre, and specific biographical details of staff. In return, I helped her where I could. For instance, we had lengthy discussions about the categories she defined as part of her coding schemes, and I was one of a number of 'raters' who coded a subsample of the data to test the 'inter-rater reliability' of her coding scheme.
I was well aware of the lack of familiarity among surgeons with linguistic ethnographic perspectives. However, I did aim to share my insights with them and other 'non-clinical' academic communities. The first papers I wrote drew directly on the observations I have outlined above and appeared in journals such as Symbolic Interaction and Applied Linguistics Review. These papers fed into guest lectures I gave for Imperial's Master's programme in Surgical Education and workshops I ran in national and international conferences attended by surgeons, such as the Ottawa Conference, a major conference on education and assessment in medicine. Attendees at these sessions expressed great interest in what I had to say about learning and communication in the operating theatre. Yet their comments on evaluation forms suggest that while some found my analyses useful, others were less convinced. They recognised what I rendered visible in detailed transcripts and ethnographic accounts, but they were not always clear as to how these analyses could be used to help improve surgical care. 'So what?', they asked, for instance, when I pointed out that teaching medical students in the operating theatre is often organised in sequences of Initiation, Response and Feedback.
Meanwhile, I moved on from the ethnographic ambition to 'understand', in my case, learning and communication in the operating theatre, to now also contribute, to 'make a difference', however small. Through my residency I had become familiar with the discourse of 'improvement' in health care (all my colleagues in the department were talking about it) and noted the frequent use of terms such as 'patient safety', 'human factors', 'situation awareness', 'non-technical skills', and 'decision making'. I realised that to make 'impact' I had to develop ways to connect my insights in, say, the sequential organisation of talk, with some of those categories. The two case studies I present in the following section illustrate how I did that.