Key words: Teacher talk; Teaching exchange (IRE/F); International ESL; Membership Categorisation Devices

Abstract
In language classrooms turns of talk facilitate the meaning-making process as students and teachers collaboratively come to understand the discourse of knowledge they are co-constructing (Wells, 1999; Vygotsky, 1978) in their interactions together, teacher to student and student to student. Questions shape the essential teaching exchange IRE/F as a teacher initiates (I) the first move, a student responds (R) and the teacher again takes up a turn and evaluates (E) in the follow-up (F) move. As common and useful this exchange is for managing classroom behaviour, during the pivotal third turn in the essential teaching exchange (Young, 1992), there is potential for teachers to facilitate student talk when the teacher provides alternatives to a follow-up question (Dillon, 1985). When students talk in discussions there is potential for them to develop their communicative competence (Bachman, 1990; Canale, 1983; Canale and Swain, 1980). This case study of young adult English as Second Language (ESL) users in face-to-face interaction in a university preparatory study skills course indicates a limiting influence of teacher questioning on student talk in discussions. Rather than talk being generated by a teacher's questioning, alternatives to questions lead to increased length of turns in students' collaborative talk. Teacher plays a significant role in giving 'voice' to students whose role in discussion is limited by a less vocal membership category in the class. This study brings a discourse analysis focus to whole class discussion between teacher and international UNIPREP students in the higher education sector and provides a context for second language acquisition researchers, teachers and TESOL trainers.

Introduction
A pedagogy which creates an individual metaphor is a practice that affects fundamental aspects of teaching and learning at the interface, where teachers meet students in their common interactions in the classroom. Types of questions: open-closed and display-negotiation have been extensively analysed (Nassaji & Wells, 2000) to examine their impact on the content of interactions between teacher and student. Open questions such as "How do you cook rice in your country?" and closed questions as in "Is it 2 o'clock yet?" and display questions such as "What is the capital city of Pakistan?" and negotiation questions "What is the connection between culture and food?" are all used by teachers in exchanges that are a regular part of classroom life. Second language learning classes also make use of a range of question types. A teacher's cognitive challenge through a question of higher intellectual quality provides impetus to start classroom talk but it is not necessarily conducive to discussion.

Background
When a teacher poses a question for which there is a predetermined 'known' answer, the teacher occupies the role of 'primary knower' (Berry, 1981). The teacher poses a question and students are expected to provide a specific answer, the one the teacher had in mind. Display questions are typical of teacher-fronted lessons in which transmission of knowledge from teacher to student is the expected form of interaction. Students become adept at reducing the length of their answers to conform to the teacher's preferred composition of the answer. Display questions are therefore not conducive to discussion, when students are expected to express ideas and elaborate them. Use of open questions does change the teacher's role to one of 'secondary knower'; the teacher does not have control of the knowledge the student will provide. As students answer open questions, particularly of the negotiating kind, they have an opportunity to express their views, but even so their answers conform to the frame of the teacher's question.

Essential teaching exchange
The essential teaching exchange (Young, 1992) called triadic dialogue (Lemke, 1985) and known as IRE and IRF is the most common pattern of language interaction between teacher and students in a classroom. The exchange is well recognised as playing a key role in setting cognitive challenge for students and guiding direction of learning through co-construction of concepts (Wells, 1999). In each exchange:
"I" = initiation move (first turn) usually a question asked by the teacher;
"R" = response move (second turn) a reply made by a student in response to the question
"E" = evaluation move (third turn) of the student's response, also known as "F" = follow up of the student response, usually made by the teacher.

A teacher's third turn becomes problematic in discussion, when it includes a further question, even when the first question was an open one, as in this example:

The teacher's third turn (F) acknowledges the student's response (R) to her open question (I) by rephrasing the answer, but then includes another question, in this case "Any other ones?" Dillon (1985) maintained that questions foil and frustrate discussion. He suggested by way of contrast that alternatives to questions foster discussion and that as students maintain the floor during discussions, they attain a higher quality of language output than when they respond to a teacher's questions. This observation is particularly noticeable at the third turn in traditional IRE/F exchanges, so teachers have to consider alternatives to questioning if their students are to have opportunities to increase language output in a way that promotes discussion.

Alternatives to questioning
Alternatives at the third turn have been shown in Dillon's studies to elicit higher quality talk from students and to increase the length of their utterances in discussions. Criticism has been levelled at teachers' use of the IRE pattern claiming that triadic dialogue controls students' ideas and expression and limits the range of ways students can interact in a discussion in the classroom setting. From early sociolinguistic studies of the teacher's role in managing classroom interaction (McHoul, 1978; Mehan, 1979, and Cazden, 1988) transcribed texts have been used to analyse the essential teaching exchange (IRE). Generally research has been conducted in mainstream primary classrooms. This study brings those concepts into higher education sector among international students who are preparing for undergraduate and postgraduate studies where active participation in discussion rather than passive reception of information is a valued behaviour in the university tutorial setting.

The UNIPREP program provided coursework and face-to-face teaching in classrooms. Student talk was a valued component of academic skills students were developing. There were four courses of study, one of which was Studying at University (SAU). As part of the speaking component of English language development in SAU, students were involved in critical thinking. Topics included independent learning, motivation, democratic discussion and cultural responses when seeking and accepting help. Opportunities were created through discussions for students to develop communicative competence. The classroom offered a supportive environment for students to make propositions and have their peers comment, by adding and by modifying understanding from personal experience and from studied reading on relevant topics. The language they generated was output that provided a means of enhancing their linguistic competence. A discussion forum within UNIPREP coursework was selected as the context for the study. It offered natural opportunities for students to talk and for the teacher to provide reinforcement of discussion points and to manage who could take a turn and who might hold the floor next.

Discussion was considered a significant part of tutorial talk. By definition discussion is involved when people talk over a subject and if they investigate it by reasoning and argument. Other definitions include the concept of considering a question in open and usually informal debate, in addition to treating a topic formally in speech or writing. As students engage with a teacher's response to their statements, and with reactions from other students to the teacher's initiating move, they become involved in discussion. In a tutorial setting students are expected to make contributions that are focussed on a selected topic, rather than on a range of casual conversations that are more appropriate to a group of friends outside. In a discussion, talk involves gathering information and soliciting opinion thus providing opportunity for students to talk. It is also an invitation to participate in the cognitive exercise of comparing other students' view of the world.

In their multicultural UNIPREP classroom, there was scope for students to develop inter-cultural awareness while following discourse rules appropriate to the academic tutorial setting. Rules included turn taking, waiting for a transitionally relevant place to make a contribution, making an orientation to the topic being discussed and facilitating involvement of others in the group by allowing expression of personal views. During the pivotal third turn in the essential teaching exchange (Young, 1992), there was potential for teachers to facilitate student talk by providing an alternative to another follow-up question (Dillon, 1988).

Among alternatives to questioning Dillon (1994, pp.77-85) provided the following:

To the above five alternatives, back-channelling was included in the study. Back- channel signals included gesture, verbal signal and pause. Each signal allowed students to hold or take back their turns and continue expressing a view.

Method
A case study of adult English as Second Language (ESL) users in face-to-face interaction in a university preparatory study skills course was chosen to investigate the influence of a teacher's questions on student talk.

Selection of teacher and course content
The program coordinator on campus was also the course team leader of Studying at University (SAU). That course had discussion topics incorporated from week 5 in the 13-week program. As teacher of the study skills course amongst second language users of English, she recognised that students needed to be active learners across the four macroskills, and to have opportunities to develop their oral skills. One well-tried avenue to talk construction was discussion. Topics had been selected in the course materials that were relevant specifically to international students enrolled in a university preparation program. The teacher knew that the traditional IRE/F teaching exchange was characteristic of classroom talk and she was prepared to focus moves in the third turn to alternatives to a further question.

Selection of students and stage of the program
Students in two UNIPREP SAU classes participated during weeks 5, 6, and 7 of a 14-week program. They were familiar with routines of classroom talk and had experience with the expectation that they were to make contributions to discussion when topics were raised. Over a three-week period, when discussion topics were due for wider exposure, a ten-minute segment in each class was recorded on audiocassette. The teacher had selected the module for discussion from their class materials, namely personal stories of adjusting to study in an English-speaking environment.

Number of recorded sessions
Six audiotape recordings were made, two segments of talk from each class providing sufficient corpus for analysis and to capture talk on a range of discussion topics.
Students understood that they were to be participating in class as usual, in a natural way and to ignore the audio recorder and researcher-observer. They provided permissions of their willingness to participate in the recorded sessions.

Quasi-experimental action
For a ten-minute period in the discussion stage of the session, an audio recording was made of the teacher and students responding to the set topic. In whole-class interactions the teacher aimed to open up and maintain discussion in a natural way based on an opening topic question, such as "What were some of the stresses that Evelyn faced?"
Questions were posed and alternatives to questions offered by the teacher through the discussion period. For example, following a student statement, rather than closing the talk, the teacher rephrased the statement and paused, thereby encouraging a student to retain her turn.
In this example (M2) is a student and (T) is the teacher:

(M2): I said it's better for her to staying at home and do something++ instead of her studying++
T: So her family expected her to be a home person
In that case, the teacher provided a reflective statement as an alternative to another question. The teacher chose from the six alternatives to questioning as the choice of response at third turns in teaching exchanges and as prompts through interactions, (Dillon, 1994, pp.77-85).

Analysis of data
Language data were dealt with in the following way:
1. Transcriptions were made; teacher and student moves were identified.
2. Teacher's evaluative / follow-up moves were noted as a stimulus to student talk.
3. Questions and statement types were marked on the transcriptions.
4. Transcribed text was analysed for student responses following a teacher's initiation
5. Numbers of words uttered by students (both content and function words) in response to a teacher's question or statement were noted and tallied.
6. Comparison was made of student responses following teacher utterance types.
7. Role of membership categorisation devices, female 'voice' in discussion.

Results
Predominantly open negotiation types of questions were employed by the teacher. These had the effect of starting discussion when students did not initiate questions themselves. Open questions were expected in situations where the teacher was prompting personal views and did not have prior knowledge of the content of the students' talk. Display questions were minor occurrences in the data and usually had question tags attached, such as "that's what you said isn't it?" On occasions, statements with question tags "You don't like that, do you?" were treated as rhetorical questions, and therefore as not requiring a student answer. They were 'heard' as confirming responses and classified as declarative statements.

Transcriptions showed teacher and student talk in English, with false starts and fillers, content questions, students' answers, students' initiations, interruptions and extended talk with samples of discussion in written form available for closer analysis. Teacher's follow-up moves which demonstrated one of the six response types at the teacher's move were identified as demonstrating an influence on choice of students' moves and indicated length of their talk in the discussion mode. General notice was taken of the meaningful content of the students' talk. Word count of students' utterances demonstrated a difference between responses to teacher questions and to the other five types of prompts the teacher provided during discussion. Each type of response was analysed on a pie chart and some explanations offered as to the findings.

Both questioning by the teacher and alternatives to questioning yielded language production by students. Samples of the teacher initiation were provided from across the sample of transcriptions. When questions were asked, student responses tended to be short and undeveloped. Often the question had to be posed more than once. The following samples of talk indicate interactions for groups A and B, where T= teacher; H, M, F = Students
Effect of questions on discussion

The teacher posed questions often in the form of a tentative construction, using an auxiliary verb in the conditional form 'would' to soften the request. In English this structural form is preferred as it is thought to reduce the face-threatening act of asking a direct question. Teachers use polite request forms when asking students questions expecting them to provide an answer or proffer a view. Secondly, questions were not to be taken literally on all occasions. Students had to process the question and interpret the proposition as one requiring a pragmatic understanding of the questions as in the sample given. The teacher in this instance allowed whole class participation by asking for 'anyone else', which implied all people were invited to speak by adding to what the last speaker had said. The last speaker would feel inclined not to be the one to add more on hearing that statement. The proposition was not to be taken necessarily at face value; 'to add to that' can be explained as increasing content of what has already been said, or it can mean provide some other substantial content. Likewise the second question was not to be interpreted at a literal level. A reaction to the story they had heard and read together was implied.

Taking this example of a typical classroom question, a considerable level of interpretation or familiarity with English was required simply to determine the question. Then there was processing time to determine what and how to answer the teacher's questions in terms of the discussion theme. Simply put, questions were more difficult to interpret than alternatives to questions.

When display questions were posed by the teacher, minimal responses were likely. Students produced minimal answers with hesitant or little follow up. In the next sample following the minimal response 'Yes' by student H, the teacher proceeded to elaborate and develop a long turn, so discussion by students was foiled. (Dillon, 1993)

Effect of reflective statements on discussion

A reflective statement of a student comment was one in which the teacher stated her understanding of what the students had just said, giving her sense of it in an economical one sentence reflection. Reflection took the form of repetition or summary, characterising the student's utterance. Often the teacher would start the utterance with "So you're saying that…' and not change the intention of the speaker but make a reflective restatement. By rewording a student's statement in that way, the overall effect of clarification engaged the student in discussion and appeared to reduce confrontational effects of a question. In that sample of talk, student M extended expression of his view as a result of the teacher's reflective statement.

Reflecting on the discussion theme and reformulating a previous comment, the teacher engaged students and allowed them to expand expression of their ideas. Less imposing than a question was the teacher's reflective statement which immediately signalled to students that the previous student turn was valued, and considered worthy of personal reflection and retained as a discussion point. Generally students are used to teachers taking back the third turn, to acknowledge accuracy of a student response before posing the next question, often with little reflection on explicit or implied meaning of the student's previous response. So with an occasion to have another opportunity to talk following the teacher's endorsement of the previous response students were likely to continue, providing even further endorsement of the student's view in discussion.

Effect of declarative statement on discussion

In a declarative statement the teacher stated her 'pre-question' thought that came to mind as a result of what the student was saying. It is the thought which would trigger a follow-up question if the teacher were to ask the next question. It might not necessarily be the opposite of what was stated; it could be complementary to it, or simply informative of her thoughts, somewhat like the answer she would have given herself in response to her next question. The student speaker in such situations of hearing the teacher's declarative statement had the benefit of her thoughts on the matter. In the above sample, student M1 repeated the teacher's main idea, holding the turn as he formulated the content of his worldview in lines 9, 10 & 11.

Effect of speaker referral on discussion

The teacher stated a relationship between a current student's statement and a previous speaker's, referring one to another, offering potential for students to discuss a previous proposition.

Effect of statement of mind on discussion

Having heard a student statement, the teacher described what came into her mind. The student got to speak and respond to the teacher's true state of mind allowing discussion to develop. There was potential for that alternative to yield higher language production but the students' realisation of the ideas might have been different from the teacher's perspective.

Effect of statement of interest on discussion

The teacher stated an interest in hearing further about what the student had been saying. She showed a direct interest in the student's expressed view, and she wanted a definition or example, so interest was reflected in the statement she made to the student. Recognition of a viewpoint being well received by the teacher had a motivating effect on the student's engagement with discussion and it was evident as the student expanded his previous concept.

Effect of back-channel signal on discussion

When the teacher listened to students in discussion format, she provided verbal and non-verbal signals indicating that a speaker was being encouraged to continue. Non-verbal signals included a nod of the head, making eye contact, or other hand gesture. She acknowledged what was said by means of verbal signals, or a pause or fillers such as 'eh uhm' while looking intently at the speaker showing that she had no intention of interrupting. Pauses and attentive silences created a feeling of obligation by students to offer more language input to discussions. Back-channel signals (Hatch 1999) indicated that the student speaking could keep the turn and not be interrupted by the teacher although another student might have joined the discussion. The signals also indicated to students that they were on track. Given such assurance as in line 13 in this sample, student M expanded his views and provided a contrast in the discussion.

Back channel signals were used throughout the recorded segments of talk in discussion.

English language production
Production of language and length of student turn were higher in the alternatives than to direct questions, even of the open kind. Taken overall, on average students produced 15 words following a teacher's question. By contrast, utterances were longest from a teacher's statement of interest in the students' ideas in the discussion (36 words average). More questions were asked by the teacher than alternative forms of communication with students but those questions yielded less opportunity for students to talk, 10% on average.

Figure 1 showed quantity of English language production by students in Group A expressed as number of words in response to seven types of teacher verbal initiations. Statement of interest provided the alternative most likely to receive extended talk by students, followed by declarative statements, reflective statements, back channel-signals and statement of mind. The alternative of referring to another student yielded lowest count of number of utterances on this occasion, similar to the length of utterances from the teacher's questioning.

Figure 1: Length of Student Utterances in Group A (number of words)

Group A: Number of words Produced by Students in Response to Teacher's Stimulus Statement

Figure 2: Length of Student Utterances in Group B (number of words)

Group B: Number of Words Produced by Students in Response to Teacher's Stimulus statements

Discussion
Classroom communication exchanges between text and learner, teacher and students, students and students provided the learning context for discussion in tutorials. Teacher talk and talk generated by turns within the classroom discourse (Doughty & Williams, 1998; Edwards & Westgate, 1994; Dillon 1988, 1994) had an impact on the learning context and tended to foster discussion when the teacher was conscious of the roles of questioning and of alternatives to questioning.

Alternatives to questions provided opportunity for more language to be produced by students than direct questions. Although direct questions engaged students, the question often had to be repeated to gain an answer. When a response came, it was a brief answer without a clear development of the idea held in the question. It appeared that students were trying to second guess the teacher and provide a short accurate answer as a summary or non-elaborated point when the teacher posed a question.

Whereas questions tended to yield short answers, alternatives to questions more often produced longer responses which were picked up by other students and elaborated upon, extended, and exemplified. The IRF pattern of interaction did not preclude collaborative interaction between teachers and students, as previous research indicated. Students built on one another's contributions as Wells (1999, p.209) has also shown, "in a manner that advances the collective understanding of the topic under discussion". They brought into view elements from their cultural heritage that were not anticipated or produced when direct questions were posed at the third turn. As students they had to acquire tools that attuned them to the 'cultural logic' (Baker & Freebody, 1989) of the pervading teaching practice in an academic English tutorial. They had to perform student roles within parameters their teacher encouraged or allowed them to act out (McCarthy & O'Keeffe, 2004) while they could be seen also to conform to the quite narrow range of behaviour that their peers accepted in discussion.

Cultural influences on discussion in the diverse international group of students
There were some issues related to cultural expectations among the group of student participants. Of 40 students in the study, 30 were men and 10 were women; 70% were 25 years of age or younger. Over 50 % (23 men) were from China or countries in the Middle East (Saudi Arabia, United Arab Emirates, Oman, and Libya). Among the women, up to three in each class were Chinese; all other nationalities were represented by only one woman. Other countries represented included PNG, Solomon Islands, India, Brunei, Indonesia, Thailand, Ghana, Korea, and South Africa. Unless they were explicitly invited by the teacher to join the discussion, women contributed fewer turns of talk than men and fewer than might have been expected in an all female class.

Roles of female and male students in discussion
Women in this study were from paternalistic cultures and many were less inclined to initiate talk in English or to speak out in mixed multi-cultural company. Further, there may have been hierarchies of age and status that predisposed the female participants to turn taking rather than initiating a turn or interrupting others in conversation. Also they came from traditions in which reading and writing were academic pursuits more highly valued than the spoken word. They may also have been inclined to hold beliefs that the teacher should control discussion. They may have been acting out those beliefs, so they were hesitant and tended to wait for an invitation to contribute to the discussion.

For a section of the male cohort, having women in the class was a new experience. Men and women in their home countries were educated in separate institutions. This background experience may have caused them to be less inclined to acknowledge contributions from their female classmates or to hear them. Cultural background may have contributed to their dismissing the female viewpoint in the whole discussion. Further the men, particularly those from the Middle East appeared more confident than the women in speaking English. In order for the female voice to be heard, the teacher tended to reiterate the female speaker's statement by repetition or restating it with interest. These strategies gave 'voice' to the women and kept their ideas alive in the discussion rather than having them missed by quiet voices or dismissed as a male provided his ideas without acknowledging or following up on the previous female speaker.

Topic coherence
Sacks introduced the concept of membership categorisation device to provide a means of describing a category, for example gender which comprises one or more subordinate concepts of categories for example 'male' and 'female' and a set of rules which enables one to pair population members with a category (Coulthard, 1977, p.85).

Two further rules, the economy rule and the consistency rule. The economy rule states that if a member uses a single category from any device then he can be recognised to be doing adequate reference to person. The consistency rule states that once one member of a device has been used 'other categories of the same collection may be used to categorise further members of the population' (Coulthard, 1977, p.86). In the classroom, students who conformed to the teacher's prescription on a particular behaviour provided a category of how to contribute to a discussion. Sometimes the teacher made explicit how to identify in a specific category. Inevitably there were two categories of student, those who conformed and those who did not and they were identified by their behaviours as still within the same device 'this discussion'. A teacher's implied reference that the female Selin had offered a good idea for relief of stress (line 59) showed that she valued her response. The teacher gave an opportunity to develop the concept in lines 62, 64 and 67 as adjacency pairs teased out the male - female categories such that the 'voice heard' in this discussion was the female Selin. She claimed to use swimming as a means of relieving stress at university. The teacher ensured that her voice was heard.

The consistency rule then allows a corollary, the hearer's maxim: 'If two or more categories are used to categorise two or more members of some population and those categories can be 'heard' as categories from the same collection, then hear them that way. (Hester & Eglin, 1997). Feline was a female who used the sport of swimming as an example of a means of stress release and her view was not to be dismissed by a male voice claiming that she was off track in her answer. The two individuals mentioned together, Feline and Hashan were heard as being co-members of the device 'contributors to a discussion' and they were to be 'heard' as two equal participants. Many devices were duplicatively organised throughout the discussion.

The population consisted of a series of contributors and the talk was analysed from a view that the teacher aimed to give equal value to each speaker, and not one speaker was to be more valued than another. The participants in the population of the discussion consisted of those in the class. They were not unrelated devices all talking about different topics. They were related and bound by their category 'student' and device, the topic of discussion 'How to relieve stress at university'. Category-bound activities are those that are done by members of certain devices. 'Proposing relevant ideas' is an activity bound to the category 'participant' when it is a member of the 'discussion' device. Also there is an ordered relationship like student speaker, listeners, contributors, and teacher. A category-bound activity can be instanced to support or criticise. When a listener 'hears' the membership category the way the speaker intended, his response will be meaningful as part of that membership. However, if the listener has not 'heard' the sequence as the speaker intended, for example when the speaker provides a role that is unexpected or a controversial role for a female in the listener's view and the student listener has not identified the membership role, there is potential for communication breakdown as occurred in line 65 by Hashan.

As students with background languages other than English, they were using English in developing an understanding of cultural adjustment to Australian university tutorial setting. At the same time they were actually finding out what their own culture represented. That was a challenge which hitherto had not been properly noticed. Those who were culturally aware recognised that they were experiencing a process of finding a 'third place' (Crozet & LoBianco, 2003). A third place is the space where users of a language learn to manage personal reaction to content identified as typical of the target culture. As they learned they could be comfortably part of two cultures, they identified with behaviours common to both and they managed those that were distinct. They began to recognise they no longer clearly identified with one culture only; they had a third place to identify with.

Conclusion
There had been an expectation among students from particular education systems that the teacher was to provide all information in the classroom. Rather than it being a student's role only to listen, so that discussion responses might have been elicited more in keeping with the requirements for Australian university tutorial exchange, a climate of encouragement to speak developed. Statements of interest and reflection, referrals to previous speakers and use of non verbal support were made meaningful as students came to acknowledge a role in their own learning from contributions their colleagues made to the discussions.

This study has brought a discourse analysis focus to whole-class discussion between teacher and international UNIPREP students in tutorial sessions in the higher education sector. It has provided a context for second language acquisition researchers, TESOL trainers and teachers and shown potential as another site for imagining ESL Study Strategies pedagogy.

References
Baker, C. D., & Freebody, P. (1989). Talk around text: Constructions of textual and teacher authority in classroom discourse. In S. de Castell, A. Luke, & C. Luke (Eds.), Language, authority and criticism: Readings on the school textbook (pp. 263-83). London: Falmer Press.

Bachman, L. F. (1990). Fundamental considerations in language testing. Oxford: Oxford University Press.

Berry, M. (1981). Systemic linguistics and discourse analysis: A multi-layered approach to exchange structure. In M. Coulthard & M. Montgomery (Eds.), Studies in discourse analysis. (p. 126). London: Routledge & Keegan Paul.

Canale, M. (1983). From communicative competence to communicative language pedagogy. In J.C.Richards & R.W. Schmidt, (Eds.), Language and communication, New York: Longman.

Canale, M. & Swain, M. (1980). Theoretical bases of communicative approaches to second language teaching and testing. Applied Linguistics, 1(1), 1-47.

Cazden, C.B. (1988). Classroom discourse - the language of teaching and Learning. Portsmouth: Heinemann.

Coulthard, M. (1977). An introduction to discourse analysis. Essex: Longman.

Dillon, J.T. (1988). Questioning and teaching: A manual of practice. London: Croom Helm.

Dillon, J.T. (1994). Using discussion in classrooms. Buckingham: Open University Press.

Doughty, C. & J. Williams, J. (Eds.) (1998). Focus on form in classroom second language acquisition (pp.114-138). Cambridge: Cambridge University Press.

Edwards, A.D. & Westgate, D.P.G. (1994). Investigating classroom talk. London: The Falmer Press.

Education Queensland (2002). A Guide to Productive Pedagogies, Classroom reflection manual, Brisbane: Teaching and Learning Branch, Curriculum Implementation Unit.

Hatch, E. (1999). Discourse and language education. Cambridge: Cambridge University Press.

Hester, S. and Eglin, P. (1997). Membership categorization analysis: An introduction. In S. Hester, P. Eglin (Eds.), Culture in action: Studies in membership categorization analysis (pp. 1-23). Washington, DC: International Institute for Ethnomethodology and Conversation Analysis and University Press America.

Lemke, J.L. (1985). Using language in the classroom. Victoria: Deakin University.

Lo Bianco, J. & Crozet, C. (Eds.). (2003). Teaching invisible culture: classroom practice and theory. Melbourne, Vic.: Language Australia.

McCarthy, M. & O'Keeffe, A. (2004). Research in the teaching of speaking. Annual Review of Applied Linguistics, 24, 26-43.

McHoul, A. (1978). The organisation of turns at formal talk in the classroom. Language in Society, 27, p.25.

Mehan, H. (1985). The structure of classroom discourse. In A. Teun and A. Van Dijk, (Eds.), Handbook of discourse analysis, Vol.3, Discourse and dialogue. London: Academic Press.

Nassaji, H. & Wells, G. (2000). What's the use of 'triadic dialogue'?: An investigation of teacher-student interaction, Applied Linguistics, 21(3), 376-406.

Vygotsky, L. (1978). Mind in society. Cambridge, Mass: Harvard University Press.

Wells, G. (1999). Dialogic inquiry: Towards a sociocultural practice and theory of education. Cambridge: Cambridge University Press.

Young, R. (1992). Critical theory and classroom talk. Clevedon: Multilingual Matters.