Wednesday, 7 January 2015

Loughborough News: 8th Conversation Analysis Day


On December 15th 2014, I had the pleasure of attending the 8th International Conversation Analysis Day at Loughborough University. It is an annual event organized by the Department of Social Sciences, regularly at mid-December. The timing is perfect: it is a nice gathering before Christmas, and the end of the year gives the opportunity for an overview of previous year’s work. At the same time, it is useful to think about next year’s plans. Since I did not present a paper, I could fully concentrate on the potential application of the presented results in my work.

The entrance of the campus

I use Conversation Analysis (CA) in my work as a method because it helps to illuminate what people are actually doing when they are engaged in a conversation. Investigating language ideologies, I prefer this action-oriented approach for many reasons which are nicely summarized by two professors of the department which organized this meeting:
Rather than having memories, thoughts, attitudes, etc., that they carry around in their heads and produce on cue, people are shown to formulate or work up the nature of events, actions, and their own accountability, through ways of talking. These ‘ways of talking’ are constructive and action-oriented. They are constructive in the sense that they offer a particular version of things, rather than any other. They are action-oriented in the sense that any actual version of events, being a specific one produced on and for the occasion of its production, is always analyzably doing something (e.g. countering, complaining, praising, justifying), and not merely being dumped from memory into talk. (Edwards and Stokoe 2004: 500)
For me, these considerations are especially important: when analyzing my data, I always try to be aware of the co-constructedness of the accounts I work with. One of the consequences of this approach in the interpretation of data is that I do not write that ‘people think something’: people, rather, tell various things in various contexts, and doing so, they often recycle elements from previous conversations; they even quote or mimic others. This complexity makes CA work quite tricky, but, at the same time, exciting and enjoyable, too.

This one-day Loughborough conference gave a nice and inspiring insight into how CA can be used in various cultural, institutional and linguistic contexts. In the following paragraphs, I briefly highlight a couple of topics which are especially interesting and important for my work. Taking into considerations that the presented materials were in general confidential, I will provide constructed examples which are structurally similar to the presented ones, but not identical with those.

I liked the houses like this in the town.


The first presenter of the day, Ana Cristina Ostermann, who arrived from Brazil, presented recordings from a national healthcare survey which was carried out through telephone. As there are many phone call-based surveys (e.g. on public opinion), it is important to understand their mechanism better, especially if delicate situations emerge. For example, in the survey in question, respondents were asked about their sexual orientation. This question was originally roughly in the middle of the researchers' question list, but later it was replaced to the last place, because many people just simply hung up when asked about this. Since sometimes I also work with sensitive issues, it was also interesting for me how the question and the term heterosexual was interpreted by the participants. For example, gaps or discourse markers (dispreference markers) often marked that answerers did not feel themselves comfortable when encountering this question. It’s not so surprising if we consider that a question always does more than asking: there are always presuppositions included. For example, why is it important to ask such a question? Why is it formulated as a yes/no question? (In the case of yes/no questions, dominantly ‘yes’ is the preferred answer.) As for my own work, while analyzing interviews, I always try to make it explicit in the analysis what a question could imply. For example, when I ask teachers about what they consider as “the more important thing in teaching Hungarian language”, I presuppose at least two things: first, that something should be prioritized and second, that there is one single thing (see the singular ‘the most important’). In other cases, I formulate the question using a plural form: then, the expected answer is a list at least a couple of relevant topics, goals or activities. The fact that researchers take part in the co-construction of answers cannot be avoided: conversation is always a shared activity, even in institutional settings like a research interview. What is important is to be aware of this kind of complexity of conversations while analyzing data. 

Linguistic Landscape and fun: very 'informative' signs, close to the conference venue. 


Beyond surveys, there are many other settings when it is crucial to understand what is exactly happening in interaction. Such settings are, for example, healthcare encounters. It is important that doctors and patients understand each other very well in order to find the best diagnosis and treatment. The field of healthcare interaction seems to be a special sub-field of CA, and, considering its social impact, it is not surprising that many papers at this conference analyzed doctor-patient interaction.

Alison Pilnick's invited lecture showed how healthcare professionals and CA analysts can co-operate. She reflected on current trends in UK social healthcare which emphasize patient autonomy and patient-centered practices in joint decision-making (for example, in situations where it is needed to decide whether a test, an operation or a medication is necessary, and if yes, in what form). It is a basic argument that the patient should be involved into decision-making because decisions influence her or his own physical, psychical and emotional conditions. However, doctors regularly play a dominant role in decision-making processes. Is it a problem? Isn't it possible that this asymmetry can be perceived as beneficial by the patients, and that is one of the reasons why they collaborate in the maintenance of doctor-dominance? Alison Pilnick argued that instead of criticizing doctor-dominance as such, it would be more important to find the cases when doctor-dominance is clearly against the patients’ interests, and try to give feedback to practitioners in order to improve their interactional skills. These considerations, together with the ones in the following paragraph, are very fundamental in my work, too. I study school-related materials, and I usually encounter initiatives which aim at fostering pupil’s and teachers’ agency, and the implementation of pupil-centered learning environments. But what is beyond the recommendations, stated principles and reformed curricula? What is a ‘pupil-centered’ school like? My Marie Curie project targets, among others, these questions and I think there are a couple of issues in healthcare interaction which are useful to consider while studying school interaction. Primarily, I think that the analysis of power relations and agency in healthcare can give ideas for studies on education.

For example, questions like what ‘involvement’, ‘decision’ or ‘choice’ means, are both part of healthcare and educational investigations. At the CA Day, there were interesting examples which can boost our thinking about the answer. For example, Merran Toerinen’s paper investigated the formulation of healthcare practitioners’ questions, focusing on structures like ‘do you want X?’, ‘do you want me to X?’, ‘do you want somebody to X?’ and ‘it is Y enough to Z?’, for example, ‘do you want me to prescribe some pills for that?’, or ‘is it bad enough to start a special treatment?’. Toerinen emphasized that both basic question designs foreground the patient’s wishes and needs, but the agency relations are constructed differently. The first three variations of the same basic design highlight the person who recommends something, optionally including others (see ‘do you want somebody to…?’), while the second pulls the phenomenon to the center, and builds mainly on the patient’s perception.

Ruth Parry and her colleagues’ paper was based on excerpts from a conversation where the doctor, the patient and his partner were present, making the whole situation more complex as the person who accompanied the patient argued for or against the doctor’s suggestions. The analysis showed that the doctor consequently reduced the level of his authority: making the background of his suggestions explicit, he justified why he thought it would be better to do X rather than Y. That is, he did not use his authority as an unquestionable position from which he could predict anything to others. Doing so, he fulfilled the requirement of involving the patient into decision-making. However, it was also clear that in some cases the doctor did not accept the patient’s ‘no’ answers as valid, so he continued to argue for his preferences until they reached a compromise which was acceptable for all of the participants.

Clare Jackson and her colleagues’ presentation analyzed a very popular, award-winning television series, ’One Born Every Minute’. In this documentary series, the life of midwifes is followed. According to the webpage of the series,
One Born Every Minute provides helpful insights into the realities of giving birth, especially in a hospital environment. The intimate footage gives parents-to-be (or anyone thinking of having children) a unique inside view of what it is really like when life begins.
The authors of this paper highlighted that in this case, joint decision-making was promoted in different ways: from assertions through recommendations and suggestions until open questions, health care practitioners used various structures while interacting with midwifes and their partners. According to Clare Jackson and her colleagues, the linguistic structures which were involved can be positioned along a continuum where one extreme point is the health care practitioner’s point of view (e.g. ‘you should now go there’, that is, ‘I recommend to go there’), and the other extreme is when the decision is marked as exclusively the midwife’s business (e.g. ‘what do you think would be the best to do now?’).

This talk was in connection with a recurring challenge in my research: how to work with edited materials? CA scholars prefer unedited recordings because editing modifies, among others, temporal and spatial relations, and these have their important roles in interaction. But, at the other side, editing (manipulating) the recordings contribute to the watchers’ perception of the scenes, so  the investigation of edited materials tells a lot about how editors reconstruct certain situations, and what kind of implications can be found in those given reconstructions. So my answer to this challenge is that edited videos are worth analyzing, especially when the analysis focuses on ideologies. For example, in my Oulu paper, I presented how a teacher training video reconstructs and reinforces the teacher-centeredness of a lesson with various tools, including zooming to her, presenting her instructions and explanations in length, while cutting out the parts where students’ task solving activity would be in the focus, etc. The reconstructed scene was then a lesson where the students only formed the ‘background’ of the teacher’s work. If we consider that, according to the annotation of the video, this was intended to be a good example for future teachers, interesting implications can be found in connection with certain ideologies and priorities of Hungarian teacher training.


At the CA Day, I could listen to many other inspiring talks what I can’t describe here in detail. For example, Carrie Childs and Dave Walsh presented their study on self-disclosure and high grade assessment in the case of rapports. Alexandra Kent and Kobin Kendrick’s paper was about imperative directives and admonishments. Edward J. B. Holmes’s topic was the management of advice-resistance in ordinary talk.

There were two papers in which the focus was mainly on methodological and theoretical issues. For example, Elliot Hoey talked about the (not always so easy) distinction between pause, gap and lapse in interaction. Brian L. Due proposed a couple of examples which illuminated challenges in the analysis of the turn-taking system when Google Glass is in use. According to him, it is necessary to elaborate new methods for CA analysis, because although there are many studies on machine–machine or human–machine interaction, human–human interaction is underresearched in cases where (wearable) computers are involved.

Finally, it was a pleasure to listen to the closing invited talk by Paul Drew and Alexa Hepburn entitled ‘Absent Apologies’ in which they presented a lot of (very funny) examples. The cases showed how something can be co-constructed as transgression, how transgression is (not) admitted, and how apologies are (not) made in ordinary talk.

Greetings from the campus.


I really enjoyed the relaxed, informal atmosphere of the Conversation Analysis Day. Although I did not present a paper, I could discuss my work in progress with many of the presenters and the department members during the breaks, over lunch and dinner, and in the form of meetings on the following day. Dear All, thanks for this opportunity!


  1. Very interesting topic, thank you! I'm interested in your opinion on "nonviolent communication". Once I attended a (demonstration) occasion of a NVC-training (about children-parents relationship) and I wasn't convinced of what we tried to practice there, although maybe it's a useful method in the communication between the student and the teacher (or the children and the parent), perhaps it just requires a lot of time and practices...


  2. Dear Dia, thank you for your comment and the link. I was not familiar with the NVC approach, so now I could learn about that. What kind of practices did you learn at that demonstration? What made you think about it? I'm asking because I'm very interested in your experience (I haven't attended any NVC demos).

    From the point of view of analysis, I think that the thing which is the most important is how participants at an interaction co-construct what 'violence', 'aggression', etc. are, so it's hard to define these notions from outside. I think one of the strengths of CA is that it really helps you to concentrate on the actual interaction and find the answers for these questions in the given situation.