Analyzing Talk: The Role of the Chairperson in a Teachers’ Meeting


This case study provides an overview of a conversation analysis study that examines how the management of talk is accomplished by a chairperson in a teachers’ meeting. The case describes each of the various steps of the conversation analysis study from deciding on an initial research focus to collecting, transcribing, and analyzing the data. It also offers practical advice for doing conversation analysis and suggestions for alternative conversation analysis research foci. Finally, post-reading exercises are provided to encourage the reader to conduct his or her own conversation analysis study.

Learning Outcomes

By the end of this case, students should be able to

  • Have a better understanding of conversation analysis
  • Understand ways to develop research foci
  • Understand ways to collect relevant data
  • Understand and carry out transcriptions
  • Analyze segments of authentic spoken discourse

Project Overview and Context

Conversation analysis (CA) is a method for investigating the underlying structure of naturally occurring talk or talk-in-interaction and how it is accomplished by its interlocutors. John Flowerdew (2013) states,

The aim … [of CA is] … to describe social interaction in terms of the actions it is used to perform, not from the outside, but from the inside, from the perspective of the user. (p. 117)

The conversation analyst works with data collected from real situated interactions rather than artificially created or manipulated scripts such as those created to exemplify a specific linguistic point or invented for the language classroom (Hutchby & Wooffitt, 1998; ten Have, 2007). The analyst creates highly detailed transcriptions of the naturally occurring data to represent more closely on the page the complexities of talk-in-interaction; in other words, talk as it is brought into being by its interlocutors. The process of making these transcriptions also aids the analyst in his or her familiarization with the particular data and overall analysis (Flowerdew, 2013; ten Have, 2007).

The analyst is interested in analyzing talk as it emerges on a turn-by-turn basis. In an ongoing sequential development of turns, each utterance in a sequence is “doubly contextual” (Heritage, 1989, p. 22). An utterance is seen as shaping the context of talk as it can only be sufficiently understood with reference to the preceding utterance; an utterance is also seen as renewing the context by forming an immediate context for the following talk. Interlocutors can achieve mutual comprehension via one party directing their utterance, the second part of a sequence, to the prior talk and the producer of the first utterance exercising the right to repair or continue the talk as a third action in the sequence. The analyst seeks to determine how interlocutors’ social actions and orientations to those actions are maintained and adjusted via the creation of this sequence of context-building next actions which moves the talk to its final destination while shaping its overall contextual structure. Thus, as George Psathas (1995) states, “the task of the analyst is … the discovery, description, and analysis of [the interlocutors’] produced orderliness” (p. 3).

I was first introduced to CA on my Master’s in Teaching English to Speakers of Other Languages (TESOL) course. This particular case study provides an overview of the various steps to a study I conducted using CA to examine how the management of talk is accomplished by the chairperson in a teachers’ meeting. This study focuses on how the role of the chairperson emerges and is oriented to by participants in the meeting and how the chairperson is able to control turn-taking and topic allocation to maintain the various interlocutors’ focus on the business at hand. The CA study (Boon, 2006) can be accessed on the English for Specific Purposes World website—

Research Design

Step 1: Deciding on an Initial Focus

Before starting to collect data, it is important to have a general research focus or “conversational phenomenon” (Flowerdew, 2013, p. 117) in mind. Ideas for research foci can be gained by doing prior reading in the CA field to determine what aspects of talk-in-interaction other researchers have analyzed. (For example, Wong and Waring, [2010] contains many ideas for possible second language teaching research foci.) Alternatively, research ideas may arise from particular incidents that occur in the classroom while teaching or by eavesdropping idly on other people’s conversations.

Originally, I had decided to focus my CA study on my interactions with a private student at a language school I taught at in Japan. I wanted to examine how closely our talk-in-interaction during “free conversation” lessons (lessons in which there were no specific language targets and topics to talk about emerged naturally between the two of us) paralleled that of talk between L1 interlocutors. Basically, I wanted to determine whether there was any pedagogic value to providing the student with these kinds of lessons as regards to more formal learning of the language.

Step 2: Collecting Data

Once an initial focus has been decided upon, researchers need to collect relevant naturally occurring data or may choose to use existing recorded data or data available in the public domain such as TV or radio broadcasts (ten Have, 2007). When collecting new data, it is important for researchers to obtain informed consent from the interlocutors for access to and permission to record their particular interactions. Researchers should also seek permission to be able to present on or publish any relevant research findings.

Consideration should also be made as to the most effective means of recording the data. These days, I use the application QuickVoice Recorder on my iPad for audio recordings as it works well and has an easy-to-use drag slider for rewinding and forwarding recorded data during the transcribing and analysis steps. Video cameras are also useful and have the added advantage of recording non-verbal interaction among interlocutors.

There are many problems that can occur while recording data; for example, the microphone may not pick up all of the audio, multiple speakers may make it difficult to determine what is being said at any given moment, batteries may run out, or a wrong button may be pressed. Therefore, it is useful to give any recording device a trial run, check the results, and make any necessary adjustments.

Another possible issue with CA data collection is the “observer’s paradox” (Richards, 2003, p. 108). Although researchers wish to collect authentic data, as interlocutors are aware they are being recorded, they may act artificially, different to their usual interactional behavior. This in turn may provide researchers with talk that is not representative of the everyday interactions of the particular social context being studied and may lead to unreliable data. However, the more often a video or voice recorder is placed within a particular environment, the more often interlocutors tend to forget they are being recorded and revert back to natural interactions.

In my initial CA study, I explained my research focus to my private student at the language school and asked his permission to record the lessons. After being granted permission, I used a standard cassette recorder and C90 cassette tape to make my first recording. My plan was to record lessons over a period of time to overcome any effects from the observer’s paradox and to capture authentic interactions between us.

Return to Step 1: Changing the Focus

When conducting research, problems may occur at any given stage of the process. In my case, at the end of the first lesson with my private student, he stated that he felt uncomfortable with being recorded and wished to cease his involvement with the CA study. As he was the only student at the language school that I was conducting free conversation lessons with, his decision in effect ended my particular research focus. I was unable to record the interactions and thus could not generate the necessary data. This meant that I had to return to Step 1 of the research process and come up with an alternative focus. Luckily, I remembered that I had an existing recording of naturally occurring interaction that I could utilize for a CA study. I had recorded a teacher’s meeting to assist my research for a different study (Boon, 2004). The meeting was organized by myself and my trainer to provide a professional development session for teachers at the particular institution and also to help me gain further insights from these teachers into effective instruction-giving in the language classroom (Boon, 2004). The meeting had comprised seven participants: three native English teachers, three Japanese teachers of English, and a chairperson (me).

My new focus was to determine from the recording of the meeting how a chairperson may control turn transition and allocation, topic allocation and maintenance, and maintain seven-party interlocutor concentration on a single stream of interaction rather than it breaking into a number of different separate conversations as what usually happens when a group of people get together in ordinary talk-in-interaction.

Although I had recorded the meeting with a C90 cassette tape and it had unfortunately run out before the end of the 2-hr meeting and the microphone had not picked up one of the interlocutors who had spoken quietly throughout the meeting, the data contained authentic talk-in-interaction, informed consent had already been agreed, and any distortion of the quality of interaction due to the observer’s paradox was negligible as interlocutors had interacted with the knowledge that the meeting was being recorded as data for my 2004 study and not for CA purposes. As I believed a subsequent analysis of the recording to determine how interlocutors oriented to and were accountable to locally established rules of interaction within the meeting offered little in the way of potential ethical harm to its participants, I continued with my CA study.

Step 3: Transcribing Data

Once researchers have collected their data, an important start of the analytic procedure is to transcribe it. Transcribing involves representing the sounds of a recording of natural interaction via the written format, thus allowing researchers and subsequent readers of the research a means of access to the “lived reality” (ten Have, 2007, p. 95) of what was said by the interlocutors at each turn-at-talk. Although time-consuming, researchers can begin to gain initial insights into the interaction being studied via repeated listening of the data while transcribing.

Unlike other research methods, CA involves undertaking a very detailed transcription of the data. A popular transcription system developed by Gail Jefferson (1979) includes noting the time, date, and place of the recording; line numbers; interlocutor names; words as spoken; sounds as uttered (e.g., D’ya wanna go?), silences or gaps; pauses; overlapping talk; latched speech; inaudible sounds; pace; stretches; stresses; and volume of utterance (ten Have, 2007). For example,

029] Mike:([Cough]) I use a combination of verbal and modelling e:::rrr (0.5) to say as I’m speaking I gesture with my hands what have you. WELL, obviously lower level more modeling a::n’ higher level less. °especially pretty much like dat°

030] Yasuko:(2.5) Yeah, same fo:r me as well like basically BUT FOR like maybe LO::W students for you guys maybe for us=

031] Chairperson:=Yeah.

032] Yasuko: will be a bit different but some lo::w lo:w low level students I will have to::: like use (2.0) well but we use the same thing, ahhhh, errrr, modelling >an’ ya know< the gestures and speaking slowly and give simple instructions.

033] Chairperson: Right (4.0) So its mainly verbal?

034] Yasuko: u:::hhhhhhhhhhh=

035] Erika:=I do modelling al[ot]

036] Yasuko: [Uhnnnn] (Boon, 2006)

In the excerpt from my 2006 CA study, double parentheses (Line 29) are used to provide any contextual information that occurs, yet is difficult to write phonetically. Here, the interlocutor, Mike, coughed before taking his turn at talk. Colons (Lines 29, 30, and 32) are used to depict elongated speech (e.g., e::rrr). Single parentheses that contain a number (Lines 29, 30, 32, and 33) signify a significant pause in talk (e.g., [0.5] represents a pause of half a second). Lexis that is underlined shows emphasis or a raise in volume by the interlocutor (Lines 29, 30, 32, and 33). Lexis placed between degree signs (e.g., °) (Line 29) shows that the words were spoken at a much quieter volume than the rest of the interlocutor’s talk. To maintain authenticity, lexis is also spelled as the interlocutor has uttered them. For example, in Line 29, Mike pronounces “that” as “dat.” Equals signs show that there is no discernable pause between interlocutors’ turn-taking. In Line 31, for example, the chairperson’s utterance of the back-channeling device, “yeah,” appears immediately between Yasuko’s utterance of “us” in Line 30 and “will” in Line 32. Lexis surrounded by arrows can show that the interlocutor’s speech has quickened (e.g., > <) or slowed (e.g., < >). For example, in Line 32, the filler, “an’ you know,” is spoken by Yasuko at a much quicker pace than the rest of her turn-at-talk. Finally, square brackets are used to show instances in which talk overlaps. For example, in Line 35, Yasuko responds to the current speaker’s talk before Erika has completed her turn.

Such detailed transcriptions are undertaken by researchers in CA to provide the reader with an accurate representation of the recording that is itself a mere copy of the spoken data that has naturally occurred (Hutchby & Wooffitt, 1998). For more information on Jefferson’s transcript symbols, please refer to the following links:

When conducting my 2006 CA study, to save time, I did a simplified transcription of the whole meeting (or the 90 min that had been recorded). This involved writing down the line numbers of each utterance, the name of the interlocutor who had uttered the words, and the words themselves. After data analysis, I was able to determine the parts of the data I wished to use in my research paper to illustrate the normative and deviant patterns of interaction within the teachers’ meeting. Once selected, I could return to the original recording, listen again, and conduct more detailed transcriptions of these parts in accordance with CA transcription principles.

Step 4: Exploring Data

After transcribing the data, researchers should begin to explore it via a process of “unmotivated looking” (ten Have, 2007, p. 121). By unmotivated looking, the conversation analyst examines the naturally occurring data with no (or at least few) preconceived analytical aims or ideas but rather to notice aspects of talk and the specific actions being accomplished by the interlocutors within the particular talk-in-interaction. In other words, the analyst builds a generic description that answers the question What is going on here, now? (Richards, 2003). ten Have (2007) suggests looking at turn-taking (and any interruptions to the turn-taking system), sequences, and repair while conducting an initial examination of the data.

In my 2006 CA study, I read through the simplified transcript of the data to obtain a greater understanding of what had been talked about and by whom in the teachers’ meeting. I then noted specific sequences in which the chairperson had successfully controlled turn-taking and topic allocation. I also highlighted instances in which disturbances to the established turn-taking system had occurred or digressions from the topic at hand had ensued.

Step 5: Analyzing Data

Once an initial exploration of the data has taken place, researchers should begin to focus more closely on specific sequences within the talk-in-interaction. Although it depends on the particular research focus one has chosen, possible interactional features to explore are

1. Turn-Taking

Ordinary talk is managed by interlocutor adherence to a turn-taking mechanism that is locally determined. The mechanism facilitates the smooth transition of turns between speakers to minimize gaps or overlaps, accommodates the allocation of the next speaker in a sequence of interaction and renders interlocutors accountable for modifying or disregarding the established system. Interlocutors are entitled to one turn at talk: a turn-constructional unit (TCU). At the end of a TCU, a transition-relevance place (TRP) comes into effect in which

1. The turn may be allocated to a next speaker by the current speaker.

If this does not occur …

2. The current speaker can relinquish the floor and a next speaker can select to the take the floor.

If 1 and 2 do not occur …

3. The current speaker may continue with a new turn at talk and the process is repeated at the next TRP. (Sacks, Schegloff, & Jefferson, 1974)

Thus, an interesting point of analysis for researchers can be interlocutor orientation to or departure from this locally managed mechanism.

2. Sequence organization

A conversation analyst may also examine how talk-in-interaction is sequentially organized. For instance, a greeting by an interlocutor (e.g., “How are things?”) sets up the expectation of a response by a second interlocutor (e.g., “Good, thanks”). In CA terms, these are adjacency pairs. The first-pair part of the utterance creates the contextual environment for the second-pair part of the sequence. In turn, the second-pair part is understood by its reference to the preceding utterance and also creates the context for the subsequent talk. Adjacency pairs may have a preferred or dispreferred second-pair part of the sequence; for example, acceptance (preferred) or rejection (dispreferred) of a request. It is usual that dispreferred responses are hedged by an interlocutor (e.g., “I’m really sorry, but …”) to soften them.

3. Repair

Repair is concerned with how interlocutors deal with trouble within the ongoing talk-in-interaction to maintain mutual understanding and coherence. Trouble may occur in talk-in-interaction due to interlocutor errors, slips, mishearings, requests for clarification, and so on. Repair may be signaled by “self-initiation” (e.g., “Where was that place again?”) or “other-initiation” (e.g., “Sorry, where d’ya say?”) Likewise, repair may be completed by “self-repair” (e.g., “Oh, I remember. It was Bournemouth”) or “other-repair” (e.g., “You mean Friday dontchuh”) (Flowerdew, 2013; ten Have, 2007). Repair sequences offer the analyst opportunities to examine how interlocutors locally manage and self-regulate talk-in-interaction to minimize potential breakdowns in communication.

4. Topic management

To partake in conversation, interlocutors need to be able to initiate, maintain, shift, and terminate topics to be talked about within the ongoing talk-in-interaction (Wong & Waring, 2010). Topic initiation may be signaled by elicitation (e.g., “What’s new?”), by announcing a “newsworthy” item (e.g., “I went to Thailand for my summer vacation”), or by commenting on the immediate environment (e.g., “Nice day, isn’t it?”). Topic shift may be signaled by the use of specific markers (e.g., “By the way …,” “That reminds me …”). Topic termination may be signaled by the introduction of pre-closing items that can indicate a desire to end the overall conversation (e.g., “Okay”). Thus, a conversation analyst may be interested in how topics are introduced, maintained, shifted, and terminated within specific interactions. Another question of interest may be related to which interlocutor is able to introduce and develop topics within the ongoing talk.

CA in Action

In my 2006 CA study, I focused on examining turn-taking procedures and topic management within the teachers’ meeting to determine how the participants helped establish and orient to the role of the chairperson by surrendering certain conversational rights that would be afforded to them during ordinary conversation.

First, I noticed that unlike ordinary conversation in which interlocutors have the right to select the next speaker after each TRP, participants in the meeting oriented to the chairperson as the official allocator of turns at talk. The chairperson had the sole right to initiate questions (first-part pair initiation moves) to introduce each item on the meeting agenda. By choosing not to select a specific speaker via an address form (name) after each question, the chairperson created an open floor environment in which any participant was free to respond. The movement of next turn operated on a rotation system between participants that constrained each member to contribute a response to each of the chairperson-initiated questions. If participants wished to contribute extra turns once their turn-at-talk had passed to a different speaker, they had to seek permission from the chairperson to do so (e.g., “Can I just say?”), thus orienting to the chairperson as turn coordinator.

The chairperson used back-channeling response devices (e.g., “Right, sure …,” “Yeah”) that were noticeably absent from other participants to help him exert further control over the turn-taking procedures. Back-channeling enabled the chairperson to acknowledge the response of each participant, quickly return the floor to the speaker, but also ensures that the chairperson is in a position to appropriate the floor at any available TRP. This was because the chairperson established the primary right to occupy the next slot after each question and answer sequence unless he forwent his right to do so. This enabled him to be omnipresent in the interaction and ensure turns returned to the chairperson so only he was able to initiate or allocate subsequent turns. For example,

111] Yasuko: So:::: but basically (1.0) It—I will say probably instructions were about (0.5) two minutes at most(?)

112] Chairperson: Right, sure=

113] Yasuko:=and if during the activity the student is going in the wrong way then I will point out an’ then get back to the way its supposed to be.

114] Chairperson: Right. so you—you’re checking=

115] Yasuko:=Uhnnn

116] Chairperson: all the way through a very long instruction process(?)

117] Yasuko: Uhnnn

118] Chairperson: Yeah. good (1.5) YEAH ITS A QUESTION THAT I::: I asked (1.0) because …

In this excerpt from my 2006 CA study, the chairperson’s use of the back-channeling device “right, sure” (Line 112) yields the floor and is understood by Yasuko as an opportunity to continue her talk. However, “right, so” (Line 114) is constructed as a pre-start to a longer TCU and allows the chairperson to take the floor, initiate a declarative question (Line 116), constrain the reply to an affirmative or negative response (Lines 115-117), and occupy the next slot in the sequence in which a third back-channeling device, “yeah” (Line 118), and its repetition, “yeah” (Line 118), is constructed to be able to shift topics and allow the meeting to move forward.

Second, in my CA analysis, I focused on topic management in the teachers’ meeting. In ordinary conversation, interlocutors have the freedom to initiate and develop topics within the talk-in-interaction. However, in the teachers’ meeting, the responsibility of the chairperson is to keep participants focused on the predetermined topic agenda. Thus, by the chairperson introducing each item on the agenda as a question, it formed the immediate context for second actions in a sequence. The preferred response of each participant therefore was to provide an answer that referred to the preceding question. In this way, the chairperson was able to control the topics that were talked about in the meeting.

Over a number of turns, however, topics can drift, continue for too long, or lapse prematurely. Thus, the role of the chairperson was to initiate further questions that steered participants back to the business at hand and to introduce new topics to be talked about to prevent certain issues dominating the meeting, avoid any potential lapses, and move the topics along to fit the allotted time.

In the teachers’ meeting, there are a number of deviant cases in which the chairperson’s role of turn allocator is undermined, back-channeling does not allow the chairperson to maintain the floor, or meeting participants are able to offer alternative items for discussion which are then taken up by others as a topic for discussion over a number of turns and lead to digressions in the meeting. The deviant cases are of particular interest as they reveal the various interactional strategies employed by the chairperson to regain control of the meeting. For further information on these deviant cases, please refer to Boon (2006).

Practical Lessons Learned

Although I was satisfied with my CA study at the time and was able to write the research up for publication, there are certain parts of the research process I would now do differently:

  • Be careful when recording. The original recording dates back to a training session/meeting held in the year 2000. At that time, I only had access to a very old cassette recorder, was only able to record 90 min of a 2-hr meeting due to using a C90 cassette, and the microphone of the cassette recorder failed to pick up one of the quieter meeting participants; 16 years later, technology has greatly advanced and the ready availability of quality video cameras and voice recorders or voice recorder applications makes recording a much easier activity. However, it is still easy with current devices to not place microphones in suitable positions, to run out of battery, to not have enough storage available, or quite simply to press the wrong button.
  • Demonstrate greater ethical awareness. Although I had sought permission to record the original training session/meeting from the participants with regards to capturing the content of the discussion to help with a different research project (Boon, 2004), I failed to inform the participants I would use the recorded data for CA analytical purposes. Although at the time I believed that my CA research would not harm participants in any possible way, in hindsight, I should have contacted them to check they agreed with me using the data for a new research focus. In addition, I should have anonymized participants’ first names within my publication to prevent any possibility of their identification by its readers.


Although somewhat time-consuming in respect to the highly detailed transcription procedure, CA offers language teachers a valuable means of engaging with and increasing awareness of the structure, organization, operations, and complexities of authentic spoken discourse. This in turn enhances our understanding of how to teach it to our students (Wong & Waring, 2010). After my 2006 CA study, I had a much deeper insight into the role of the chairperson in a meeting and how an effective chair exercises considerable control over the talk-in-interaction to ensure the smooth transition of topics and speakers, the concentration of multi-party participants on a single stream of talk, and that all items on the agenda are discussed before the end of the allotted time. Thus, it helped me to prepare for and to teach my business students with much greater confidence regarding how to interact in meetings in English.

I hope that reading this case study will have encouraged its readers to do more reading in the field of CA, to come up with their own research foci, and to conduct their own CA studies. It is via our small contributions to a cumulative collection of thick descriptions of naturally occurring talk-in-interaction that we help further our knowledge of the workings of authentic spoken discourse.

Exercises and Discussion Questions

  • Consult the references and recommended list of further readings and read more about Conversation Analysis (CA) methodology.
  • Do a Google search for “Jefferson transcription procedures” and find out about CA transcription symbols.
  • Record family and friends’ interactions (with permission).
  • Record your classes.
  • Listen to the recordings and pay attention to turn-taking organization, sequence organization, instances of repair, and topic management.
  • Make a transcription of a short fragment of recorded oral interaction. Use Jefferson’s transcription procedure.
  • Analyze the data. Think about what is going on in the data.
  • Develop a research focus for a CA study.
  • Collect data and conduct your CA study.
  • Share your research results with the teaching community by writing them up for publication or presenting them at a conference.


Boon, A. (2004). Instruction-giving: Effective procedural instructions. Modern English Teacher, 13(3), 4047.
Boon, A. (2006). Managing talk: The role of the chairperson in a teachers meeting. ESP World, 12(5), 1-18. Retrieved from
Flowerdew, J. (2013). Discourse in English language education. Abingdon, UK: Routledge.
Heritage, J. (1989). Current developments in conversation analysis. In D.Roger & P.Bull (Eds.), Conversation (pp. 2147). New York, NY: Multilingual Matters, Ltd.
Hutchby, I., & Wooffitt, R. (1998). Conversation analysis. Cambridge, UK: Polity Press.
Jefferson, G. (1979). A technique for inviting laughter and its subsequent acceptance declination. Retrieved from
Psathas, G. (1995). Conversation analysis: The study of talk-in-interaction. Thousand Oaks, CA: SAGE.
Richards, K. (2003). Qualitative inquiry in TESOL. Basingstoke, UK: Palgrave Macmillan.
Sacks, H., Schegloff, E., & Jefferson, G. (1974). A simplest systematics for the organization of turn-taking for conversation. Language, 50, 696735.
ten Have, P. (2007). Doing conversation analysis: A practical guide. London, England: SAGE.
Wong, J., & Waring, H. (2010). Conversation analysis and second language pedagogy. New York, NY: Routledge.
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles