「文藝と思想」第 79 号 2015 年 2 月 （41） ～ （63） 頁
The ‘Monologic’ of Student-Student
Interactions in EFL Classes
in Japanese Universities
Paul STONE
AbstractIn a detailed study of student-student classroom interaction, Hauser (2009) describes group discussions in a Japanese university EFL class as being fairly monologic. By this he means that the students in his study rarely engaged with one another’s ideas to develop an argument, offer support to one another, or chal-lenge one another. In these interactions, students took it in turns to be the ‘primary speaker’ who delivered their opinion to the rest of the group, while other group members responded with minimal backchannels (e.g. non-lexical tokens such as “un” or head nods) but did not make full claims to turns-at-talk. In this paper I intend to develop and broaden Hauser’s discussion of the monologic of student interactions through a detailed analysis of a range of interactional data col-lected in a Japanese university classroom. I do so by employing a multimodal approach to interaction analysis, as well as ethnographic data collection methods that seek to involve the participants in the research process.
Introduction
Student-student interactions are seen as an important part of the learning process in many approaches to education. Dialogic teaching, constructivism, stu-dent-centred learning and many other approaches and theories all stress the
importance of interactions amongst students to improve education. Within L2 education approaches such as the 1980s communicative approach, task-based lan-guage learning and the concept of Interactive Competence have similarly empha-sised the important role that communication plays in learning and in the classroom. Given this focus on interaction between learners it seems important that we have studies that investigate this interaction in some detail.
There have been some concerns from EFL and SLA researchers (e.g. Firth and Wagner, 2007; Hellermann, 2008) that not enough research has been undertaken into actual examples of classroom interaction. However, over recent years a number of studies utilizing a conversation analysis (CA) framework have been describing and analyzing classroom talk (e.g. Seedhouse, 2004 and Hellermann, 2008). Many of these CA studies have focussed on language. However, language is only ever one mode of communication among many (Norris, 2004, 2011) and there has been much recent interest in investigating communication from a multimodal perspective (e.g. Kress, 2010).
The research reported on here adopts a multimodal perspective to communica-tion and applies that perspective to a detailed analysis of classroom interaccommunica-tion, while also following CA’s concern with the sequential ordering of talk. The particu-lar focus of this paper is one particuparticu-lar type of interaction within the classroom: group discussions between learners. Group discussions are intended to engage learners in meaningful language use and negotiation of meaning, providing a space for learners to think together and develop ideas. As well as providing learners with opportunities to use the L2 together, discussions should be unbounded spaces where ideas can develop, not according to some predetermined trajectory but according to the countless possibilities created by the meeting of multiple voices (Wegerif, 2013).
One problem with group discussions in classrooms across the world, both L1 and L2, is that learners often do not fully engage with one another’s ideas as they want to avoid disagreement and maintain group harmony (Wegerif, 2013). In a CA study of Japanese university students’ English language group discussions, Hauser (2009) shows how the interactions he observed operated according to a kind of monologic, where each student took one turn as ‘primary speaker’ to
deliver their opinion to the rest of the group. The rest of the group would receive the primary speaker’s turn using only minimal backchannels (e.g. using non-lexi-cal sounds such as “mm” or through embodied actions such as nods) and without developing ideas or attempting to reach any kind of consensus.
In performing this study I observed many of the phenomena described by Hauser in his chapter. Here I intend to add to those observations with my own, developed in a detailed micro-analysis of 30 hours of video data recorded in my classroom. I intend to develop and broaden Hauser’s concept of the monologic of the ‘primary speaker’ and how it affects turn-taking practices in classroom interac-tions. I also intend to move beyond a descriptive position to consider why students might orient to the ‘primary speaker’ identity and the implications that this has for L2 classroom education.
Methodology: CA, MDA and MIA
Like many other studies into L2 classroom interaction I make use of tools and methods from CA. This includes a focus on the turn-by-turn construction of talk and the use of notions such as adjacency pairs and repair to help understand the organization of that talk (see Markee (2000) for a discussion of CA methodology).
CA is an important theoretical and methodological framework that has been used by a number of researchers to investigate the mechanisms of everyday talk, as well as classroom interaction (e.g. Seedhouse, 2004), and its narrow focus is exemplary at uncovering the structural patterns of talk-in-interaction. However, as an approach designed to investigate talk-in-interaction, CA places spoken lan-guage at the centre of analysis and investigates other modes of communication through a methodological framework structured by the mechanisms of talk. And with its focus on describing the mechanisms of the conversational turn-taking system and the rules for their use, a ‘pure’ CA does not look up from the micro-context in which talk occurs to consider the broader social micro-context.
In the study from which this analysis is taken I am not only interested in describing spoken language, but aim to analyze the whole multimodal ensemble of which talk is only ever one part of. I am also not only interested in the mechanics
of interaction, but also in investigating how social actors perform educational prac-tices through interactions that also produce social relationships and identities that position the social actors in complex social and cultural networks (and affect condi-tions for learning). I therefore do not take a ‘pure’ CA approach and look beyond the immediate sequential context of the language used to take a more social approach to my analysis, making use of Scollon’s (2001) mediated discourse theory and Norris’ (2004, 2011) development of this in her multimodal interaction analy-sis. That is, I do not limit my analysis to what is said, but also investigate the use of modes other than language (e.g. gesture, gaze, and so on) as well as other aspects of the interaction, such as the relationships of participants, that would usually not be analyzed by CA researchers.
Mediated discourse analysis (MDA) is a wide-ranging theoretical perspective that incorporates elements of a CA approach, but is more socially engaged than CA. It seeks to address questions about the complexities of social practices by explicating the links between social action and discourse. One of the major motiva-tions behind MDA is Scollon’s attempt to develop a theoretical remedy for dis-course analysis that operates without reference to social actions on one hand (e.g. CA), and social analysis that operates without reference to discourse on the other (e.g. Bourdieu’s (1977) practice theory). That is, Scollon was concerned that much analysis of discourse failed to consider the wider social context, whereas much social analysis neglected to analyse actual language use. A mediated discourse perspective sees social practice and discursive (linguistic) practice as mutually constitutive – one being accomplished through the other. Analysis seeks to describe how the discursive practice (i.e. the ‘talk’) mediates the social practice (such as purchasing a cup of coffee) and how the social practice mediates the talk. Social actors are seen as being positioned by chains of actions that occur at the intersection of what Scollon calls the interaction order (the social roles and rela-tionships in a situation), discourses, and the historical body (ways of being in the world) of the social actor (Scollon and Scollon, 2004).
The principles that underlie MDA stress that discourse should principally be seen as consisting of actions (rather than systems of representation, such as lan-guage). Social actions are defined as the social actor at the moment of performing
an action with a communicative mode (such as gaze, gesture, language, etc). That is, the action as a unit of analysis includes both the social actor and the mode within it (Scollon actually uses the term mediational means rather than mode, following Wertsch’s (1998) sociocultural theory). So the social action of saying “hello” requires a social actor (i.e. a person) to perform it, and also a system of representa-tion (in this case language) to make it meaningful. Thus defined, the social acrepresenta-tion includes the histories of the social actor and the mode within it and, from an MDA perspective, all action is performed according to the histories of the actor and the mode. That is, the social actor’s history of social experience will affect how they act in the present, and all communication depends on shared systems of represen-tation, such as a shared language, that derive their meanings from their histories of use. So a shared language and a shared social history (for example, growing up in a similar social environment) help two social actors to communicate with one another.
MDA provides the theoretical perspective from which I have undertaken this project. In terms of analysing the data I have employed the methodological frame-work of multimodal interaction analysis (MIA), which as mentioned above is a development of MDA by one of Scollon’s students (Norris, 2004, 2011). This methodological framework is designed to investigate all of the actions taken in real-time, face-to-face social interactions, without placing spoken language at the centre of analysis. MIA takes some of the central theoretical concerns of MDA and combines them with visual research methods to take a more multimodal approach to analysis. Concepts and tools from MIA are useful in allowing us to analyse how participants produce actions on multiple levels and across multiple communicative modes.
Both MDA and MIA suggest that ethnographic data collection methods are important in opening up a close interactional study to become a broader, socially-informed study. As such, I also utilize more interpretive, ethnographic research procedures to inform my analysis of classroom interactions. There are different understandings of what ethnography is, but of particular importance to this paper is a type of ethnographic data that Scollon and Scollon (2001) call observer’s
observer returning their observations to the group under study. This allows par-ticipants in the study to comment on the findings of the analysis – a process that can be difficult for both the participant (who may not like to hear descriptions of their own behaviour) and the researcher (who may find that their own interpreta-tions of the data are wide of the mark). This presents opportunities to both confirm findings and find starting points for deeper understandings. One important way in which Norris presented her analysis to the participants in her study was to utilize
playback methodology. This involved showing the participants the data pieces that
had led her to a certain analysis and asking for their interpretation before giving her own. This allowed the participants to become involved in the research process, providing the researcher with useful insights, and to also benefit from the researcher’s analysis of their interactions. As discussed below, I utilized playback methodology in this study.
So, in this project I take CA’s concern with the sequential ordering of talk, but combine this with the broader social and multimodal approaches of MDA and MIA. This allows me to perform a micro-analysis of the mechanics of interaction, while also considering what is going on from a social perspective.
A note on transcription
As will be described in the following section, the project centred on transcrip-tions that I made of video and audio data. Transcription is a process of transforma-tion that involves reducing complex phenomena for the purpose of analysis (Flewitt, 2006) and a multimodal transcript must transform both visual and audio aspects of an interaction into a printable format. Transcription is also theory, reflecting the aims of the research and directing the research findings (Ochs, 1979). My interest here is in the multimodal performance of classroom interac-tions. The multimodal transcription process described in the following section makes it much easier to represent posture, for example, than a transcription process that relies on written descriptions. However, it is important to remember that it is only ever a representation of what happened, as is the video and audio data itself. So the transcriptions used in this project are representations of
representations that reflect the interests of myself as a researcher. There are many other ways that the data might be transcribed that would allow for different focuses.
Transcription systems must also attend to two separate fields in that they must accurately represent the structure of the events for analysis, but must also be as clear as possible for presentation to an audience (Goodwin, 2001). As can be seen below, to present the data in the simplest format for the reader here I have opted not to include images in the transcripts presented in this paper. This has the result that this paper places spoken language at the centre of the analysis. It is, however, important to bear in mind that the analysis that I performed did not do so (as it took a multimodal approach) and that the extracts below are only one representation of the data collected for this study.
Method and Data
I first collected approximately 30 hours of video data of group interactions in a first-year university classroom over one 15-week semester (by way of compari-son, Hauser’s study was based on an analysis of 38 minutes of interaction). The class was an English communication class that met once a week and consisted of 15 students and their teacher. The recordings are of groups of learners performing English discussion tasks and activities in pairs or small groups, rather than teacher-learner interactions. To ensure that the audio was captured clearly I placed audio recorders (recording high quality digital files) on each desk and after class I used simple video editing software to combine the audio files with the video data.
I then made detailed multimodal transcripts (Norris, 2004) of these interactions, which also served as the initial analysis. The purpose of the transcripts is to describe the sequences of actions taken by participants in the interactions. Multimodal transcription is complicated, involving a number of steps and methods, and the multimodal transcription methodology described by Norris involves making multiple transcripts for any one interaction. She suggests first separating the communicative modes, as far as this is analytically possible, and producing a
separate transcript for each mode (e.g. one transcript for spoken language, another for gesture, and so on). Transcripts for modes that can be seen in the video data (such as gesture) involve a different approach to those that cannot be seen (such as spoken language).
To make transcripts for modes that can be visually represented by the video data, the researcher takes screenshots from the video to illustrate the start and the end of the performance of each social action (where necessary a third screenshot is captured during the middle of the performance of the action). For example, the researcher must find the onset of the performance of a gesture and take a screen-shot at this moment. The researcher must then do the same for the end of the performance of the gesture. The third screenshot would capture the peak point of the gesture. Once screenshots have been made for each action performed they can be arranged chronologically to create a transcript of all of the actions performed in that mode. For social actions taken in modes that cannot easily be represented in a screenshot (such as language or music) text is used. Following Norris, the basic unit of analysis for spoken language is the intonation unit. The appendix to this article contains a note of the conventions employed for the transcription of spoken language in this paper.
Once these separate transcripts are produced, they are then combined and the final transcript displays only the most important aspects of the interaction (these relate to the focus of the project and are decided by the researcher and sometimes the participants in discussion with the researcher). So each transcript may consist of images, transcriptions of spoken language or written descriptions.
Once the multimodal transcripts/initial analysis was complete I went through my initial observations to look for patterns and points of interest to investigate in more detail. In this way the project took an inductive approach as findings emerged from the data. After making analyses I then took my observations to students and other relevant parties (e.g. other teachers) and engaged in video-playback ses-sions with them. These interactions with the participants formed an important part of the research process. These sessions allowed me to discover how the learners themselves interpreted the data and so gave the learners a voice in the research project. After speaking with learners I would revisit the data and analyse
it again making use of their interpretations to provide new ways of seeing what was happening. So, as well as being inductive, the project also took an iterative approach as I moved back and forth between the data, my notes and learners’ interpretations. Following this process I was able to identify themes in the data and reach conclusions about what was going on.
Analysis
The project, which is ongoing, is a large one and has provided a wide range of themes for analysis and discussion. In this paper I focus on only one phenomenon. As discussed in the introduction, Hauser (2009) has described group discussions in EFL university classes in Japan as being somewhat monologic in nature, domi-nated by the ‘primary speaker’ identity, and it is this primary speakership that I focus on here.
It is worth noting that Hauser does not use the term identity in his paper and that it may be possible to use the term ‘role’ instead. I choose to use the term identity here as I see each action that a social actor performs as also being a per-formance of identity, so that identity can be seen as fluid and performed in real-time. In an interaction we perform certain identities, some of which may be more stable and may be carried beyond the interaction (e.g. male, female, Japanese, etc) and some of which may be less stable (e.g. superior, equal, friendly, etc). I would argue that performing an ‘English teacher’ identity (for example) may at times involve performing identities in interactions such as ‘counsellor’, ‘listener’, ‘lan-guage model’, and so on. For the students in Hauser’s study (and in this study) being an ‘English student’ means at times being a ‘primary speaker’ (and being able to perform this identity successfully).
Hauser shows in his analysis how each student in a group takes one turn as the primary speaker, during which time they deliver their opinion to the rest of the group. Other students support this with non-committal backchannels that make no claim for a turn-at-talk. Once each student has had their turn as the primary speaker they do not need to take another turn. There is no attempt to reach a con-sensus as a group, or defend or forge a position. This can also be observed in the
data analysed in this study.
Excerpt 1
Excerpt 1 is taken from the beginning of a recording of a group of three students
discussing the role of women in society. The discussion question was given to these students by the teacher at the start of the activity on a piece of paper. We can see that each student is primary speaker once, and that primary speakership is passed (almost literally) around the table. S1 is primary speaker from lines 1-9, S2 is primary speaker from lines 10-18, and S3 is primary speaker from lines 19-31. While a student is primary speaker the other students perform a supporting role. For example, in lines 4-5 both S2 and S3 nod and say the non-lexical receipt token “un” simultaneously while S1 is the primary speaker. They do so during a slight pause in S1’s talk. We can see similar actions in lines 14, 21, 22 and 28 (in lines 14
and 28 these backchannels occur not only in a slight pause, but also following a gaze by the primary speaker at one of the other students). These actions demon-strate that the other students are listening and give the primary speaker permis-sion to keep speaking. They do not comment on anything the primary speaker is saying, nor do they make a claim to a full turn-at-talk.
There are clear boundaries where primary speakership is passed around the table. In line 6, S1 gives a somewhat emphatic nod at a point where her turn as primary speaker could be understood as complete (it is at the end of a complete utterance delivered with falling intonation). This nod is a claim to have finished speaking. In lines 7-8, both S2 and S3 nod in receipt of this, but neither of them subsequently attempts to claim primary speakership. This is another feature noted by Hauser; students are reluctant to claim primary speakership and must often be nominated. In line 9, S1 nominates S2 to take the next turn as primary speaker. In line 18, at the end of her turn as primary speaker, S2 similarly nominates S3 to be the next primary speaker. In line 29, S3 claps and adds the coda “I think so” to her turn to make a claim to have finished. This claim is received by S1 and S2 with minimal receipt tokens in lines 30-31.
The primary speaker identity explicated in Excerpt 1 could be seen in interac-tions where students were involved in discussions that were intended to resemble debates (i.e. one topic was discussed by a group of students who were supposed to give their opinions and respond to the opinions of others). In fact, the perceived formal nature of the debates was one reason that students adopted the turn-taking practices described. Other interaction patterns, related to the primary speakership pattern (but not described in Hauser’s chapter) are also evident in the group inter-actions analysed in this study.
I call one such pattern the ‘primary questioner-answerer’ pattern, and it is really another manifestation of the same interaction principles that drive the ‘primary speaker’ interactions. Just as the primary speaker is a dominant interaction iden-tity that one student may assume for an extended period in an interaction and that every student must assume at some point, in other group discussions we can see a ‘primary questioner’ identity that serves a similar function. The types of interac-tions in which the ‘primary questioner’ identity could be seen included
information-gap types (where different students have different information from one another, for example when discussing the relevant merits of two or three dif-ferent things that not every group member has all of the information about), “getting to know you” types of discussions, or discussions where a list of ques-tions of some type were in some way pre-determined. All of these interacques-tions required students to ask one another questions to obtain some kind of information.
In Excerpts 2 and 3 I will demonstrate the ‘primary questioner’ identity. In the discussion from which Excerpt 2 is taken, which occurred at the beginning of the semester, the students were asked to “get to know” one another in groups of three. Whereas the teacher had envisaged a somewhat conversational style (i.e. with equal rights of participation and an open turn-taking structure) to these brief discussions, the students adopted a more formal, structured approach. They began, through a series of pointing gestures and gazes, by establishing who would ask questions to who. Once the order of turns had been decided they proceeded to speak to one another. The students in my data quite often nonverbally (i.e. using gestures and gaze) negotiate the order in which they will pass around the primary speaker/questioner identity prior to the actual start of the discussion, which is also something that Hauser describes in his study. The following excerpt occurs about two minutes into the interaction from which it was taken.
Excerpt 2
directed at (although this has already been determined). She then proceeds to ask S2 a question, which S2 responds to (although we can see from the repetition of words, pauses and use of a Japanese filler (“eh to”) that she has some difficulty in producing her turn). There is no elaboration on this and no response by S1. Instead, S2 immediately gazes toward S3 in line 3, thereby nominating her as the recipient of her question, and asks her a variant of the same question that S1 asked in line 1. S3 responds in line 4 (like S2 she uses the Japanese token “eh to”). There is a two-second pause after S3 responds, but again there is no attempt to develop the topic and there is no response to this by S2. Instead, S3 gazes at S1 and in line 5 she asks her the same question.
After S1 has answered the question the participants have completed one round of question-answers, where they have all asked and answered the same question. In line 6, without any further discussion of their hometowns, S1 proceeds to ask S2 another question, gazing at her as she does so. This question is then passed around the table, as it were, with each student having a turn as both questioner and answerer at some point.
Gaze plays an important role here is helping to structure the turn-taking. While the two participants currently fulfilling the ‘questioner’ and ‘answerer’ identities gaze at one another, the third participant has no active part to play in the talk and most often gazes at the desk. Often this participant is an ‘overhearer’ (listening to what the other participants are saying), but also quite often appears to be paying little attention to the ongoing interaction.
While being an ‘overhearer’ participants often assume a less upright postural position, usually leant over the desk in front of them. At moments when a student begins speaking (and so shift from being an ‘overhearer’ to being a ‘speaker’) they frequently change posture. This postural shift usually starts with the student leaning over the table and gazing downwards (the ‘overhearer’ role just described) and then moving into a more upright position with gaze directed at the questioner ready to answer the question directed at them.
Figure 1 (which consists of screenshots that have been taken from the video and subsequently processed to protect the identity of the participants) explicates this postural shift. In the upper screenshot, S3 (who is sat on the left of the image) is
gazing at the desk, as is S1 (who is sat in the centre of the image). S1 is here being an ‘overhearer’. S2 (on the right of the image) is asking S3 a question while also gazing at S3. In the lower screenshot, S3 is answering the question that S2 addressed to her. We can see that S3 is now sitting in a more upright position when compared with the upper screenshot (the postural shift is indicated by the arrow). That is, as S3 gives her answer to the question and therefore becomes the
S2 asking question to S3
S3 answering question
‘answerer’ she also performs a postural shift.
In Excerpt 2 we can see that each student has one turn as ‘questioner’ and one turn as ‘answerer’ and that the question (and the ‘questioner-answer’ identities) are passed around the table. Each student has a clear identity at any one moment in the interaction: ‘questioner’, ‘answerer’ or ‘overhearer’.
Excerpt 3 is taken from a discussion later in the semester. This discussion
involved four students talking about emotions (e.g. things that make them happy, scared, and so on) and came at the end of a unit in the course textbook that focusses on phrases that describe emotions. The students had been asked by the teacher to sit together and to ask one another questions about their emotions, using some of the vocabulary that they had been practicing in the textbook.
Excerpt 3
In line 1, S1 asks the other three students a question. However, the order in which the other students should respond to this question has not been predeter-mined (as it was in Excerpt 2). S4 responds nonverbally with a nod in line 2 and then S2 responds in line 3. S3 has not responded and, orienting to the principle that all of the participants must respond to the question, in line 4 S1 gazes towards S3
to nominate her to speak. S3 gazes back at S1 in line 5 but does not respond. This is a source of trouble for the interaction and S1 attempts to resolve the problem by restating the question (she places the words “amusement park” at the beginning of the utterance, placing some emphasis on them, and also places extra stress on the word “amusement”). This demonstrates that S1 believes the problem to be that S3 did not hear or understand that the question is about amusement parks. S3 repeats the words “amusement park” with rising intonation (demonstrating some problem in understanding), pauses for two seconds, and then offers her answer. Her response suggests that she has not fully understood the question as she pauses before she answers and then talks about the “park” in her answer (rather than the “amusement park” she was asked about). Although S3 has clearly had some problem answering the question this is not taken up by any of the other par-ticipants (although they all subsequently claimed that they thought S3 had prob-lems understanding, they did not wish to perform any face-threatening actions). In line 9, S1 nods and says “finish”.
In lines 1-9, S1 has been the ‘primary questioner’ and each other participant has been the ‘answerer’ once. S3 had some problems understanding the question, but still attempted to perform the role of ‘answerer’ after S1 had nominated her to speak and rephrased the question. We can see that the participants are orienting to the expectation that every participant, except the ‘questioner’, must answer the question. In line 9, S1 makes a claim to have completed her turn as the ‘primary questioner’.
I use the term ‘primary questioner’, rather then simply ‘questioner’, as the instructions for the activity did not require the students to take it in turns to ask the others questions in this way. The activity specified that students should ask each other about their feelings, but did not explain how this should be done. The students could have adopted a more free-flowing approach to the interaction with any student asking any other student a question at any time (and then using the answers as a starting point for more discussion). However, the students here decide that one student at a time should ask all of the others a question and that the other students should each have one opportunity to answer that question. Once each student has delivered their answer they have fulfilled their obligations
until either a new question is asked or they are required to become ‘primary ques-tioner’ themselves.
S2 is the ‘primary questioner’ from lines 10-19, selecting who should respond to her question with the use of gaze. As well as asking the question and selecting the ‘answerer’, we can see in lines 10, 12, 15, 17 and 19 that she also provides some receipt tokens as the ‘answerers’ are speaking. The interaction is between her and whoever is answering the question at the time, rather than between the group as a whole. (Or it may be better to conceive of the ‘primary questioner’ and the ‘answerer’ as the focus of the interaction and the other students as peripheral par-ticipants). Again, answers are not elaborated on. When a participant has given the bare minimum response to a question that participant has performed their contri-bution to the interaction and no longer needs to speak.
Discussion
We can see in these three excerpts that rather than the interaction being dia-logic, with students engaging with one another’s voices, offering support or chal-lenging one another, they simply take turns at delivering their responses to the question being asked. It is for this reason that Hauser describes the group discus-sions in his chapter as monologic. Students who are not performing one of the ‘primary’ identities in the interaction either receive the speaker’s utterances with non-committal tokens, such as nods, or else background the interaction in their consciousness. So, ideas are not developed as a group and the performance of turns as ‘primary speaker’ and ‘answerer’ resembles something more of a presen-tation to a relatively passive audience. The focus is on each student having the opportunity to speak unchallenged, rather than meaningfully engaging with one another.
There are some benefits of this for classroom interaction. It provides a compara-tively pressure-free space for students to speak in English, as other students rarely challenge the speaker or rush them when they are having problems. For example, S3 in Excerpt 1 has some problems producing her turn but is given time and space by S1 and S2 to say what she wants to. Similarly in Excerpts 2 and 3 we
can see how quite rigid ‘questioner-answer’ interaction patterns constrain the talk, but also afford students a safe, predictable space in which to speak. This also ensures that all students speak at some point (as the students in this study orient to the principle that every ‘answering’student must speak at least once in response to each question), as can be seen in lines 1-8 of Excerpt 3. So one benefit of ‘primary speaker’ and ‘primary questioner-answerer’ identities is that they allow each student in a group to have a turn-at-talk.
However, I wish here to move from a descriptive stance to a more critical stance. The participants in the interactions analyzed in this paper can be seen to adopt somewhat rigid turn-taking patterns that are related to the performance of certain interactional identities (e.g. ‘primary speaker’ and ‘primary questioner’). These identities both afford and constrain the ways in which the participants may interact with one another and reveal something of the expectations that these participants have about classroom interaction and their own role within it. At times, these expectations and identities may limit students’ attempts to achieve the aims of the lessons and prevent the students from fully participating in the interaction.
For example, in Excerpt 2 we can see how the participants at times foreground and at other times background the ongoing interaction, and this is linked to the performance of the ‘overhearer’ and ‘questioner-answerer’ identities. At moments when the participants move between performing these different identities we can see them restructure their consciousness as their focus changes. Norris (2004, 2011) discusses how awareness and consciousness can be observed phenomenally in interactions and describes how changes in a participant’s awareness are indi-cated by bodily movements. She also shows how such changes are often linked to the performance of different identities. The postural shifts described in this article (and explicated in Figure 1) show a refocusing of the participants’ awareness as they foreground a new action.
As a student is performing the ‘overhearer’ identity the ongoing interaction is quite often backgrounded in their consciousness. For example, S1 in Figure 1 is backgrounding the interaction while she focuses on the action of preparing for her upcoming turn-at-talk. This is possible because the turn-taking in these
interactions is often highly predictable and the next question is often known in advance (e.g. in the interaction from which Excerpt 2 was taken S3 and S1 knew in advance what question S2 and S3 would ask them). Participants confirmed in inter-views what could be seen in the video data: i.e. they often use the moments when they are not required to speak to prepare their upcoming utterances. So, during these moments the participants are often more focused on preparation for an upcoming turn and are backgrounding the ongoing interaction.
When the participants move from the ‘overhearer’ identity and assume a speak-ing role, as S3 does in the bottom screenshot in Figure 1, they re-focus their awareness to foreground the ongoing interaction. This re-focusing is indicated by postural shifts that the participants usually perform as they begin to speak. This is important, because it demonstrates that the students are not continually focused on the ongoing interaction with one another, but are moving between the perfor-mance of more active speaking identities and more passive non-speaking identities that structure the way the interaction is performed (it is almost as if some of the students are being “switched on” and awaking from a dormant state as they sit upright at the start of their turn-at-talk). So, the participants who are not assuming one of the speaking identities are quite often not fully engaged in the interaction and are not required to contribute full turns at talk (this is not to say that they pay no attention to the ongoing interaction, but rather that it is not foregrounded in their consciousness). Also, the speaking identities are assumed only for a certain amount of time and then relinquished, and once relinquished there is no need to speak again. This means that participants who are not performing a ‘primary speaker’ identity can background the interaction and contribute little to it (except for minimal backchannels, such as nods or nonlexical tokens like “un”) so that the interactional space is comparatively closed and the talk cannot develop fully.
When not performing one of the speaking identities the students’ participation is peripheral to the interaction and, as has already been discussed, the ongoing interaction is backgrounded in their consciousness. While students at times need opportunities to carefully plan and prepare what they would like to say in English, it is also desirable that they engage in more spontaneous interactions in which they are continually foregrounding the ongoing talk, rather than shifting their
focus between more or less passive/active identities (switching themselves “on” and “off”). This might allow for greater and less predictable interaction from all members of the group. That is, it would be desirable for students to not adopt the more passive ‘overhearer’ identity once they have finished a turn-at-talk, but to remain focused on contributing to the talk and engaging with one another (this is not to suggest that all participants talk at once, but that they perform more active roles as engaged listeners who might speak at any moment).
The students in this study are orienting to a particular understanding of what it means to be a student taking part in classroom talk. This is what we might call a language display model of classroom interaction, where students are focused on accurately producing certain utterances (to display their ability to produce these utterances). There is little focus on the meaning of what others say and the talk rarely develops beyond basic exchanges of information and subsequent receipt tokens (e.g. “What music do you like?”, “I like J-pop”, “I see”). In this way the interactions are not dialogic, but have a more monologic character.
This is not how things always work outside of the classroom and I would argue that this kind of interaction does not in itself adequately prepare students for the kinds of group interactions that they may face in the future. I would also argue that it deprives them of important opportunities to engage in spontaneous talk that are not often available outside of the classroom (this being an EFL context). Also, from a dialogic teaching perspective (e.g. Wegerif, 2013) this is not an effective form of classroom interaction for promoting learning. The ‘primary speaker’ identity pro-motes a turn-taking structure that mediates and constrains the practice of taking part in a group discussion and limits the opening of dialogic space between the participants.
It may be beneficial, therefore, in group discussions of this kind, to introduce interventions or to teach students interaction strategies that help open up the dialogic space (and therefore promote learning). For example, in debate activities other voices that offer different perspectives or that explicitly challenge the stu-dents’ opinions could be introduced into the interaction, forcing the students to defend their opinions in more detail or at least engage with other voices. Or it could be possible to discuss the findings of research such as Hauser’s, or indeed
this study, with students and to encourage them to take different approaches to turn-taking in their discussions with one another (being careful to discuss why they might want to take these different approaches).
The ‘Student’ Identity
As mentioned above, I would argue that one major reason for the primary speak-ership and associated turn-taking practices described in Hauser and this paper is the students’ understanding of what it means to be a ‘student’. This understanding is developed and stored in the habitus (Bourdieu, 1977) through experiences of being a student at school. There, students are often called upon to speak to the class or the teacher in order to display some kind of knowledge, but are rarely asked to engage in interactions with other students’ voices. They are therefore unfamiliar with more dialogic turn-taking practices that encourage the develop-ment of ideas with others in interaction, and are more used to presenting pre-formed utterances to others at the teacher’s request. For many first-year university students, being a student means giving your opinion when the teacher asks you to and being quiet when others speak, but not necessarily challenging others or developing ideas through talk. That is to say, being a student is, quite often, being monologic.
So, in the interactions analysed in this study the participants were performing a version of the student identity learned through interactions in high school class-rooms. These interactions were often shaped by the physical environment of the classroom and by the role of the teacher. Students were often sat in rows of desks all facing the teacher, who controlled interactions by nominating speakers to display some kind of knowledge. In the EFL classes under study the participants were in a different kind of physical space (small groups sat facing one another at desks) and were interacting without the teacher’s guidance. The experience of classroom communication learned through high school education provides the participants with principles from which to organize the interactions in this unfamil-iar EFL classroom environment (a number of the participants in the study expressed surprise about the way interaction was organized in these classrooms). However, they still oriented to the principle, learned at high school, that for
students classroom speaking is a kind of display (whether this is a display of knowledge, an opinion, or a skill such as language ability). They did not for the most part use language to create meanings together, to negotiate ideas or engage with one another’s viewpoints. One reason for the lack of elaboration seen in
Excerpt 2 (for example) was the students’ reluctance to adopt a conversational
style, as they saw this as inappropriate to the ‘student’ identity that they were performing. They believed that they should speak in a more formal style that they believed was more appropriate for a formal educational environment.
In closing, it is worth noting that the ‘primary speaker’ and ‘primary questioner-answerer’ identities do not dominate all student-student interactions in the data analysed. In fact, students perform a wide range of interactional identities. However, the ‘primary speaker’ identity is frequently evident and does form one of the major interactional identities in this classroom. As discussed above, it most often appears in debate-type activities or in discussions of pre-determined ques-tions, especially in groups of three or more students. It is focussed on when stu-dents adopt what we might call a more institutional focus to their classroom practice (that is, when the students assumed they were performing some kind of formal, institutional activity, such as a debate). However, there are other moments in the data, which are beyond the scope of this paper to discuss in detail, where students take a more personal or relational focus and the interactions are, accord-ingly, structured somewhat differently. These moments became more common as the semester progressed and included the ‘boundary’ talk that occurred in between activities.
References
Bourdieu, P. (1977). Outline of a Theory of Practice. Cambridge: Cambridge University Press. Firth, A. and Wagner, J. (2007). ‘Second/foreign language learning as a social accomplishment:
elaborations on a reconceptualized SLA’. The Modern Langage Journal 91, pp. 800-819. Flewitt, R. (2006). ‘Using video to investigate preschool classroom interaction: education
research assumptions and methodological practices.’ Visual Communication 5, pp. 25-50. Goodwin, C. (2001). ‘Processes of seeing visual analysis: an ethnomethodological approach.’ In
Leeuwen, T. V. and Jewitt, C. (Eds.) Handbook of Visual Analysis. London: Sage.
H. and Kasper, G. (eds.) Talk-in-Interaction: Multilingual Perspectives. Hawaii: University of Hawaii, pp. 216-244.
Hellermann, J. (2008). Social Actions for Classroom Language Learning. Clevedon, UK: Multilingual Matters.
Kress, G. (2010). Multimodality: A Social Semiotic Approach to Contemporary Communication. Abingdon, Oxon: Routledge.
Markee, N. P. P. (2000). Conversation Analysis. Mahwah, NJ: Erblaum. Norris, S. (2004). Analyzing Multimodal Interaction. London: Routledge.
Norris, S. (2011). Identity in (Inter)action: Introducing Multimodal (Inter)action Analysis. Berlin: De Gruyter Mouton.
Ochs, E. (1979). ‘Transcription as theory.’ In Ochs, E. and Schieffeling, B. B. (Eds.) Developmental
Pragmatics. New York: Academic Press, pp. 43-72.
Schegloff, E.A. (2000). Overlapping talk and the organization of turn-taking for conversation.
Language and Society 29, pp. 1–63.
Scollon, R. (2001). Mediated Discourse: The Nexus of Practice. Abingdon, Oxon: Routledge. Scollon, R. and Scollon, S. W. (2001). Intercultural Communication (2nd edition). Oxford:
Blackwell.
Scollon, R. and Scollon, S. W. (2004). Nexus Analysis: Discourse and the Emerging Internet. London: Routledge.
Seedhouse, P. (2004). ‘The Interactional Architecture of the Language Classroom: A Conversation Analysis Perspective’. Language Learning 54 (supplement 1), pp. 1-300. Wegerif, R. (2013). Dialogic: Education for the Internet Age. Abingdon, Oxon: Routledge. Wertsch, J. V. (1998). Mind as Action. New York: Oxford University Press.
Appendix Transcription notation (adapted from Schegloff, 2000) [ ] Overlapping or simultaneous talk
(1.5) Numbers is parenthesis indicate silence represented in tenths of a second (.) A dot in parenthesis indicates a silence of less than 0.5 seconds
? Rising intonation :: A stretched sound
WORdy Capital letters indicate increased volume
>< The talk in between these symbols is markedly faster <> The talk in between these symbols is markedly slower
