Exploring classroom discourse

Exploring Classroom
Discourse
Routledge Introductions to Applied Linguistics is a series of intro-
ductory level textbooks covering the core topics in Applied Linguistics,
primarily designed for those entering postgraduate studies and
language professionals returning to academic study. The books take
an innovative ‘practice to theory’ approach, with a ‘back-to-front’
structure. This leads the reader from real-world problems and issues,
through a discussion of intervention and how to engage with these
concerns, before ﬁnally relating these practical issues to theoretical
foundations. Additional features include tasks with commentaries, a
glossary of key terms, and an annotated further reading section.
Classroom Discourse looks at the relationship between language,
interaction and learning. Providing a comprehensive account of current
perspectives on classroom discourse, the book promotes a fuller under-
standing of the interaction, which is central to effective teaching, and
introduces the concept of classroom interactional competence (CIC).
Walsh makes the case for a need not only to describe classroom
discourse, but to ensure that teachers and learners develop the kind
of interactional competence which will result in more engaged,
dynamic classrooms where learners are actively involved in the learning
process. This approach makes the book an invaluable resource for
language teachers and educators, as well as advanced undergraduates
and post-graduate/graduate students of language and education, and
language acquisition within the ﬁeld of applied linguistics.
Steve Walsh is a lecturer in the School of Education, Communication
and Language Sciences, Newcastle University, UK. He is the Editor of
Classroom Discourse, a Routledge journal.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Routledge Introductions to Applied Linguistics
Series editors:
Ronald Carter, Professor of Modern English Language
University of Nottingham, UK
Guy Cook, Professor of Language and Education
Open University, UK
Routledge Introductions to Applied Linguistics is a series of intro-
ductory level textbooks covering the core topics in Applied Linguistics,
primarily designed for those entering postgraduate studies and
language professionals returning to academic study. The books take
an innovative ‘practice to theory’ approach, with a ‘back-to-front’
structure. This leads the reader from real-world problems and issues,
through a discussion of intervention and how to engage with these
concerns, before ﬁnally relating these practical issues to theoretical
foundations. Additional features include tasks with commentaries, a
glossary of key terms, and an annotated further reading section.
Exploring English Language Teaching
Language in Action
Graham Hall
Exploring Classroom Discourse
Language in Action
Steve Walsh
‘The innovative approach devised by the series editors will make this
series very attractive to students, teacher educators, and even to a
general readership, wanting to explore and understand the ﬁeld of
applied linguistics. The volumes in this series take as their starting point
the everyday professional problems and issues that applied linguists
seek to illuminate. The volumes are authoritatively written, using an
engaging “back-to-front” structure that moves from practical interests
to the conceptual bases and theories that underpin applications of
practice.’
Anne Burns, Aston University, UK

Exploring
Classroom
Discourse
Language in Action
Steve Walsh
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

First published 2011
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
Simultaneously published in the USA and Canada
by Routledge
270 Madison Avenue, New York, NY 10016
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2011 Steve Walsh
The right of Steve Walsh to be identiﬁed as author of this work
has been asserted by him in accordance with sections 77 and 78
of the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including photocopying
and recording, or in any information storage or retrieval system,
without permission in writing from the publishers.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging in Publication Data
Walsh, Steve.
Exploring classroom discourse: language in action/Steve Walsh.
p. cm. — (Routledge Introductions to applied linguistics)
Includes bibliographical references and index.
1. Language and education. 2. Oral communication.
3. Discourse analysis. 4. Interaction analysis in education.
5. Second language acquisition. I. Title.
P40.8.W34 2011
370.1Ј4 – dc22 2010041937
ISBN: 978–0–415–57066–4 (hbk)
ISBN: 978–0–415–57067–1 (pbk)
ISBN: 978–0–203–82782–6 (ebk)
This edition published in the Taylor & Francis e-Library, 2011.
To purchase your own copy of this or any of Taylor & Francis or Routledge’s
collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.
ISBN 0-203-82782-1 Master e-book ISBN

Contents
Series editors’ introduction vii
1 Introduction 1
Introduction 1
The nature of classroom interaction 2
The IRF exchange structure 17
Challenges for teachers and learners 20
Summary 22
2 Classroom discourse and teaching 24
Introduction 24
Classrooms as contexts 24
Task-based language teaching 26
Interaction and teaching 31
Classroom discourse and teacher development 35
Summary 47
3 Classroom discourse and learning 48
Introduction 48
Language learning and interaction 48
Learning as a cognitive activity 52
Learning as a social process 62
Summary 66
4 Approaches to studying classroom discourse 67
Introduction 67
Recording and transcribing classroom interaction 67
Interaction analysis approaches 74
Discourse analysis approaches 81
Conversation analysis approaches 84
Summary 89
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

5 Alternative approaches to studying classroom
discourse 90
Introduction 90
Corpus linguistics 90
Combined approaches 98
Variable approaches 103
Summary 109
6 Enhancing understandings of classroom discourse 110
Introduction 110
The SETT framework: modes 111
The SETT framework: interactures 125
Using SETT 129
Summary 135
7 Classroom discourse as reflective practice 137
Introduction 137
The teacher as reflective practitioner 138
Developing reflective practices through SETT 146
Summary 157
8 Classroom interactional competence 158
Introduction 158
Interactional competence 159
Classroom interactional competence 165
Developing CIC in teachers and learners 177
Summary 180
9 Conclusions 182
Classroom discourse: looking back 182
Classroom discourse: future directions 186
Task commentaries 191
Appendix A 212
Appendix B 216
Glossary 217
Further reading 222
References 226
Index 235
Contentsvi

Series editors’ introduction
The Introducing Applied Linguistics series
This series provides clear, authoritative, up-to-date overviews of the
major areas of applied linguistics. The books are designed particularly
for students embarking on masters-level or teacher-education courses,
as well as students in the closing stages of undergraduate study. The
practical focus will make the books particularly useful and relevant
to those returning to academic study after a period of professional
practice, and also to those about to leave the academic world for the
challenges of language-related work. For students who have not pre-
viously studied applied linguistics, including those who are unfamiliar
with current academic study in English-speaking universities, the books
can act as one-step introductions. For those with more academic
experience, they can also provide a way of surveying, updating and
organizing existing knowledge.
The view of applied linguistics in this series follows a famous
definition of the field by Christopher Brumfit as:
The theoretical and empirical investigation of real-world problems
in which language is a central issue.
(Brumfit, 1995: 27)
In keeping with this broad problem-oriented view, the series will cover
a range of topics of relevance to a variety of language-related profes-
sions. While language teaching and learning rightly remain prominent
and will be the central preoccupation of many readers, our conception
of the discipline is by no means limited to these areas. Our view
is that while each reader of the series will have their own needs,
specialities and interests, there is also much to be gained from a
broader view of the discipline as a whole. We believe there is much
in common between all enquiries into language-related problems in
the real world, and much to be gained from a comparison of the
insights from one area of applied linguistics with another. Our hope
therefore is that readers and course designers will not choose only those
volumes relating to their own particular interests, but use this series
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

to construct a wider knowledge and understanding of the field, and
the many crossovers and resonances between its various areas. Thus
the topics to be covered are wide in range, embracing an exciting
mixture of established and new areas of applied linguistic enquiry.
The perspective on applied linguistics in this series
In line with this problem-oriented definition of the field, and to address
the concerns of readers who are interested in how academic study can
inform their own professional practice, each book follows a structure
in marked contrast to the usual movement from theory to practice. In
this series, this usual progression is presented back to front. The
argument moves from Problems, through Intervention, and only finally
to Theory. Thus each topic begins with a survey of everyday profes-
sional problems in the area under consideration, ones that the reader
is likely to have encountered. From there it proceeds to a discussion
of intervention and engagement with these problems. Only in a final
section (either of the chapter or the book as a whole) does the author
reflect upon the implications of this engagement for a general under-
standing of language, drawing out the theoretical implications. We
believe this to be a truly applied linguistics perspective, in line with
the definition given above, and one in which engagement with real-
world problems is the distinctive feature, and in which professional
practice can both inform and draw upon academic understanding.
Support to the reader
Although it is not the intention that the text should be in any way
activity-driven, the pedagogic process is supported by measured guid-
ance to the reader in the form of suggested activities and tasks that raise
questions, prompt reflection and seek to integrate theory and practice.
Each book also contains a helpful glossary of key terms.
The series complements and reflects the Routledge Handbook of
Applied Linguistics, edited by James Simpson, which conceives and
categorizes the scope of applied linguistics in a broadly similar way.
Ronald Carter
Guy Cook
Reference
Brumfit, C. J. (1995) ‘Teacher Professionalism and Research’, in G. Cook and
B. Seidlhofer (eds) Principle and Practice in Applied Linguistics. Oxford:
Oxford University Press, pp. 27–42.
Series editors’ introductionviii

Note
There is a section of commentaries on a number of the tasks, at the back of
the book from p. 235. The symbol in the margin indicates that there is
a commentary on that task.
TC
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Series editors’ introduction ix

1 Introduction
Introduction
The central thesis of this book is that language teachers can improve
their professional practice by developing a closer understanding of
classroom discourse and, in particular, by focusing on the complex
relationship between language, interaction and learning. The book
provides a comprehensive account of current perspectives on classroom
discourse, aiming to promote a fuller understanding of interaction,
which here is regarded as being central to effective teaching. While
classroom interaction has been the focus of attention for researchers
for more than ﬁfty years, the complex relationship between language,
interaction and learning is still only partially understood. The case is
made for a need not only to describe classroom discourse, but to ensure
that teachers and learners develop the kind of interactional competence
that will result in more engaged, dynamic classrooms where learners
are actively involved in the learning process. The concept of classroom
interactional competence (CIC) is introduced, described and problem-
atised as a means of developing closer understandings of how learning
and learning opportunity can be improved.
Rather than simply describing the discourse of second language
classrooms – an enterprise that has been underway for more than ﬁfty
years – the concern here is to promote understanding and facilitate
professional development. Chapter 1 focuses on the main features of
classroom discourse and considers how it is typically structured.
Chapters 2 and 3 look at the relevance of classroom discourse to
teaching and learning, while Chapters 4 and 5 evaluate different
approaches to studying classroom discourse. Chapters 6 and 7
introduce the SETT framework as a means of helping practitioners
evaluate their own use of language while teaching. Chapter 8 focuses
on the concept of classroom interactional competence. In Chapter 9
the main conclusions are presented.
This chapter considers some of the characteristics of second
language classroom interaction, offering an overview of the more
commonly found and widespread features that characterise classroom
interaction throughout the world. Using extracts from a range of
English language lessons, the aim is to offer a brief sketch of classroom
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

discourse features rather than a detailed description. Many of the
themes and issues raised here will be dealt with in more detail in the
rest of the book.
The chapter is divided into three sections. In the first section, I
present a description of the nature of classroom interaction in the
context of language teaching. The following section deals in more detail
with the underlying structure of classroom discourse, while in the final
section, I point to the future by highlighting some of the challenges
that face teachers and learners in relation to classroom interaction.
The nature of classroom interaction
When we reflect on classes that we have been in, either as teachers or
learners, we quickly realise that classroom communication is both
highly complex and central to all classroom activity. In the rapid flow
of classroom interaction, it is difficult to comprehend what is
happening. Not only is the interaction very fast and involves many
people, it has multiple foci; the language being used may be performing
several functions at the same time: seeking information, checking
learning, offering advice and so on.
Given its complexity and centrality to teaching and learning, it is
fair to say that any endeavour to improve teaching and learning should
begin by looking at classroom interaction. Everything that occurs in
the classroom requires the use of language. Like most human ‘joint
enterprise’ language underpins every action, every activity. It is through
language that ‘real world problems are solved’ (Brumfit 1995).
Crucially, in a classroom, it is through language in interaction that
we access new knowledge, acquire and develop new skills, identify
problems of understanding, deal with ‘breakdowns’ in the communica-
tion, establish and maintain relationships and so on. Language, quite
simply, lies at the heart of everything. This situation is further
complicated when we consider that in a language classroom, the
language being used is not only the means of acquiring new knowledge,
it is also the goal of study: ‘the vehicle and object of study’ (Long
1983: 67).
Introduction2
Personal reflection
Think about your own experiences as a learner or teacher. Why is
communication in the classroom so important? In what ways can
teachers make effective use of their language? What is the relation-
ship, if any, between the language used by teachers and learners and
the learning that occurs?

Yet despite its obvious importance, until recently, little time has
been given to helping teachers understand classroom interaction. While
researchers have gone to great lengths to describe the interactional
processes of the language classroom, few have used this knowledge
to help teachers improve their practices. Most teacher education
programmes devote a considerable amount of time to teaching methods
and to subject knowledge. Few, I suggest, devote nearly enough time
to developing understandings of interactional processes and the
relationship between the ways in which language is used to establish,
develop and promote understandings. Teachers and learners, arguably,
need to acquire what I call ‘Classroom Interactional Competence’ (CIC,
see Walsh 2006) if they are to work effectively together. That is,
teachers and learners must make use of a range of appropriate
interactional and linguistic resources in order to promote active,
engaged learning. Classroom interactional competence is discussed in
full in Chapter 8.
Let’s go back to the three questions you were asked to consider on
p. 2 and answer each in turn:
• Why is communication in the classroom so important?
• In what ways can teachers make effective use of their language?
• What is the relationship, if any, between the language used by
teachers and learners and the learning which occurs?
To some extent, the first question has been answered: communi-
cation in the classroom is so important because it underpins everything
that goes on in classrooms. It is central to teaching, to learning, to
managing groups of people and the learning process, and to organising
the various tasks and activities that make up classroom practices.
Communication refers to the ways in which language is used to
promote interaction; according to van Lier (1996), interaction is ‘the
most important thing on the curriculum’. If we are to become effective
as teachers, we need not only to understand classroom communication,
we need to improve it.
When we consider the second question, how can teachers make
effective use of their language, we must first define what we mean by
‘effective’. Given that the main concern of teachers is to promote
learning, effective, here, means language that promotes learning. There
are many ways in which teachers can influence learning through their
choice of language and their interactional decision-making. We come
back to this later in the chapter. For the time being, ‘effective’ simply
means using language that helps, rather than hinders, the learning
process (Walsh 2002).
Our third question is more difficult to answer. What is the
relationship between the language used by teachers and learners and
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 3

the learning that occurs? There is no clear and exact response to this
question. According to Ellis (1994), this relationship can be seen as a
strong one, where language use has a direct influence on the learning
that takes place; a weak one, where there is some link between the
language used and the learning that occurs; or a zero one, where there
is simply no relationship at all between the language used by teachers
and the learning that ensues. And of course, this relationship is difficult
to assess given that there are so many other factors that influence
learning. However, the very fact that this question is often asked does
suggest that there is a relationship between classroom language use
and learning. It is a question that underpins much of the discussion
in this book and one that we will be returning to.
In the remainder of this section, I offer an overview of the most
important features of second language classroom discourse. The discus-
sion that follows presents these features largely from the perspective
of the teacher who has the main responsibility for controlling the
interaction. Later in this chapter, we take a closer look at the language
used by teachers and learners (see below). Four features of classroom
discourse have been selected, largely because they typify much of the
interaction that takes place in classrooms and are prevalent in all parts
of the world:
• Control of the interaction.
• Speech modification.
• Elicitation.
• Repair.
Control of the interaction
One of the most striking features of any classroom is that the roles of
the participants (teacher and learners) are not equal, they are
asymmetrical. This is true of all classrooms: primary, secondary,
tertiary, monolingual, multilingual, with adult, teenager or very young
learners. It is also true of many other contexts in which institutional
discourse prevails and where roles are unequal: doctor/patient,
solicitor/client, shop assistant/customer, and so on. In each of these
settings, including classrooms, one party is in a position of power or
authority; that person has control of the patterns of communication
that occur and is able to direct and manage the interaction. In language
classrooms, teachers control patterns of communication by managing
both the topic of conversation and turn-taking, while students typically
take their cues from the teacher through whom they direct most of
their responses. Even in the most decentralised and learner-centred
classroom, teachers decide who speaks, when, to whom and for how
Introduction4

long. Teachers are able to interrupt when they like, take the floor,
hand over a turn, direct the discussion, switch topics. As Breen puts
it, it is the teacher who ‘orchestrates the interaction’ (1998: 119).
Learners, on other hand, do not enjoy the same level of control
of the patterns of communication, although there will certainly be
times when the roles of teacher and learners are more equal, allowing
more even turn-taking and greater participation by learners. For
much of the time, learners respond to the cues given by teachers: in
the form of a spoken response, an action (such as opening a book,
changing seats), or a change of focus (from a PowerPoint slide to
coursebook, for example, or from listening to the teacher to talking
to a classmate).
If we look now at some classroom data, we can see quite clearly
how teachers control the interaction. Look at extract 1.1 below, in
which a group of multilingual, intermediate adult learners are
discussing issues about law and order in their respective countries. In
line 1, we see how the teacher nominates a student (Erica), thereby
determining who may speak. Her question, ‘what happens if you
commit a crime’, both establishes the topic and provides a cue for
Erica, who must now reply to the question, which she does in line 3.
In line 5, we can see that the learner is experiencing some difficulty
and the teacher interrupts in line 6, indicated by = (a latched turn,
where one turn follows another without any pausing). Again, in line 6,
the teacher is controlling the interaction, seeking clarification and
correcting an error (‘what’s the verb’?). Not only does the teacher
control the topic, she controls the precise content of the learner’s
subsequent utterance in line 8, ‘they go to court’. Finally, in line 9,
the teacher brings L1’s contribution to an end ‘they can go home’,
controlling participation by inviting a response from another student
‘what happens in Brazil?’ Breen’s (1998) powerful metaphor of the
teacher orchestrating the interaction is in evidence throughout this
extract. Arguably, a teacher’s ability to ‘orchestrate the interaction’
in this way not only determines who may participate and when, it
influences opportunities for learning. It is also apparent when we look
at extract 1.1 that teachers have control over the amount of ‘space’
learners have in the interaction. For every contribution made by the
student, the teacher typically makes two: asking a question (in lines
1 and 6) and giving feedback (in lines 4 and 9). The consequence of
this is that teachers clearly talk more and occupy more of the inter-
actional space of the classroom. Learners’ opportunities to contribute
are largely controlled by the teacher. This three-part discourse
structure, comprising a teacher question, learner response and teacher
feedback is another feature of classroom discourse that exemplifies the
ways in which teachers control the interaction. We discuss this in some
detail under ‘Exchange structure’ – see page 17. (For a definition of
the transcription system please refer to Appendix B, p. 220.)
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 5

Extract 1.1
1 T: OK Erica could you explain something about law and
order in Japan what
2 happens if you commit a crime?
3 L1: almost same as Britain policeman come to take
somebody to police station
4 T: yes
5 L1: and prisoner questioned and if he is (5 seconds
unintelligible)=
6 T: =yes what’s the verb Eric Erica . . . if she or he yes
commits a crime they go
7 to
8 L1: they go to court yes but if they he they didn’t do that
they can go home
9 T: they can go home (. . .) very good indeed right what
happens in Brazil
To summarise, we have seen that teachers, through their unique
status in a classroom, and by the power and authority they have,
control both the content and procedure of a lesson, as well as
controlling participation.
Speech modification
One of the defining characteristics of all classroom discourse is
teachers’ modification of their spoken language. In some respects,
teachers’ use of a more restricted code is similar to the spoken language
of parents talking to young children. Typically, a teacher’s speech is
slower, louder, more deliberate, and makes greater use of pausing and
emphasis. Teachers also make a great deal of use of gestures and facial
expressions to help convey meaning. The modification strategies used
by teachers are not accidental; they are conscious and deliberate and
occur for a number of reasons. The first, and obvious one is that
learners must understand what a teacher is saying if they are going to
learn. It is highly unlikely that learners will progress if they do not
understand their teacher. A second reason is that, for much of the
time, teachers model language for their students. That is, they use
appropriate pronunciation, intonation, sentence and word stress, and
so on in order to give learners an opportunity to hear the sounds of
the target language. In many cases and in many parts of the world, a
teacher’s articulation of a second language may be the only exposure
to the language that learners actually receive. It is important, therefore,
that the L2 is modelled correctly and appropriately. A third reason
for speech modification is the fact that there is so much happening at
any one moment in a classroom that teachers need to ensure that the
class is following, that everyone understands and that learners don’t
Introduction6

‘get lost’ in the rapid flow of the discourse. In his 1998 paper, Michael
Breen talks about the need for learners to ‘navigate’ the discourse and
the fact that many learners do actually get lost from time to time.
It is the teacher’s responsibility to ensure that this does not happen
by making frequent use of repetition, by ‘echoing’ an individual
learner’s contribution for the benefit of the whole class, by seeking
clarification and so on.
An understanding of the ways in which second language teachers
modify their speech to learners is clearly important to gaining greater
insights into the interactional organisation of the second language
classroom and to helping teachers make better use of the strategies open
to them. What strategies do teachers use to modify their speech? We
can look at modified speech in two ways. On the one hand, teachers
employ a different range of linguistic resources to facilitate compre-
hension and assist the learning process. There are several features of
spoken classroom language that teachers normally modify in some way.
Perhaps the most obvious one is the use of simplified vocabulary and
the absence of more idiomatic or regional variations. Grammar too is
frequently simplified through the use of simpler and shorter utterances,
the use of a more limited range of tenses and fewer modal verbs.
Pronunciation is also often clearer, with slower articulations and wider
use of standard forms. Of course, teachers may ask themselves if they
are losing something of their identity by making all these changes. In
my experience, this does not happen to any great extent, although
teachers do employ their own particular ‘classroom idiolect’ (Walsh
2006): an individual way of talking that is normally based on their
personal conversational style. There are times when a teacher’s
classroom idiolect may help learners and there are other times when
learners simply do not understand because of the teacher’s idiolect.
On another level, teachers modify their interactional resources to
assist comprehension and help learners ‘navigate the discourse’. Most
notable is the use of transition markers to signal the beginnings and
endings of various activities or stages in a lesson. Words such as right,
ok, now, so, alright – typically discourse markers – perform a very
important function in signalling changes in the interaction or
organisation of learning. They function like punctuation marks on a
printed page: consider how difficult it would be to read a newspaper
without punctuation. The same applies in a classroom if teachers fail
to make appropriate use of transition markers. This important category
of discourse markers enables teachers to guide learners through the
discourse, hold their attention, announce a change in activity, signal
the beginning or end of a lesson stage. Crucially, they help a class
‘stay together’ and work in harmony.
In addition to the more obvious ways in which teachers modify
their speech discussed above, there are other more subtle strategies
that teachers use in order to clarify, check or confirm meanings. These
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 7

Introduction8
Task 1.1
Consider the extract below. To what extent does this teacher ‘guide’
learners through the discourse? As a learner in this class, in what
ways might you feel ‘lost’ and how might the teacher address this
problem?
66 T: =yes it’s the result of INtensive farming they call it (writes
on board) which is er (2) yeah
67 and this is for MAXimum proﬁt from erm meat so as a result
the animals suffer they
68 have very BAD conditions and very small erm they’re given
food to really to make
69 them big and fat and usually it’s unnatural and as you said
they HAVE to give them a
70 lot of anti-biotics because the conditions in which they’re
kept erm they have far more
71 disease than they would normally have so they give them
steroids to make them
72 stronger and of course this is now being passed through to
the HAMburger that you
73 eat is contaminated with er=
74 L: =sorry how do you spell anti- anti-biotics?
75 T: anti-biotics? anti-biotics yes? erm anti-biotics?
76 L: how to spell it?
77 T: oh how do you spell it right (writes on board) there’s er
I think I read a very shocking report
78 recently that nearly all for example chickens and beef now
pigs all all these that are reared
79 with intensive farming they’re ALL given anti-biotics as a
matter of course and of course the
80 public don’t hear this until quite a long time after we’ve been
eating it and this this is what
81 makes me angry quite a scandal really . . . sometimes when
I listen to these reports I think oh
82 perhaps I should be vegetarian and sometimes er you
wonder about the meat=
83 L4: =how the people who offer food on the street how can you
((2))
84 T: =er you can’t er check that they’re=
85 L4: =no no I mean what the name?
86 T: er oh well street vendors? you could say yes (writes on
board) er vendors from to sell yes?
TC

include confirmation checks, where teachers make sure they understand
learners; comprehension checks, ensuring that learners understand the
teacher; repetition; clarification requests, asking students for clarifica-
tion; reformulation, rephrasing a learner’s utterance; turn completion,
finishing a learner’s contribution; backtracking, returning to an earlier
part of a dialogue. These strategies operate at the level of interaction
rather than solo performance; they are used to ensure that the discourse
flows well and that the complex relationship between language use
and learning is maintained. Interestingly, perhaps, teachers rarely ask
learners to modify their speech, often relying instead on imposing their
own interpretation. Very often this results in teachers ‘filling in the
gaps’ and ‘smoothing over’ learner contributions, as a means of main-
taining the flow of a lesson or in order to create a flawless discourse.
Unfortunately, by so doing, learners may be denied crucial opportun-
ities for learning. Arguably, by seeking clarification and requesting
confirmation, by getting learners to reiterate their contributions, by
paraphrasing and extending learner contributions, in sum, by ‘shaping’
what learners say, teachers are greatly helping learners’ language
development.
An example of how this works is presented in extract 1.2 on p. 10.
Here, the teacher is working with a group of upper intermediate adult
learners and the focus is academic writing. By seeking clarification and
by negotiating meaning, the teacher helps the learners to express
themselves more fully and more clearly. Note how learner turns are
frequently longer and more complex than those of the teacher (122,
126). In the extract, this teacher works pretty hard to adopt a more
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 9
87 people selling things on the street whether it’s food or or
anything else we call it street
88 vendors yeah? so of course you can’t CHECK that that what
their hygiene is so well you’ve
89 no choice you take a RISK if you buy a hamburger from
someone who’s selling on the street
90 ((1)) I would think twice I think (laughs) be very very hungry
before I actually bought a
91 hamburger in the street alTHOUGH some of course are very
clean I wouldn’t say they’re all
92 unhygienic so yes just remember the words then er hygiene
hygienic unhygienic I I’m going
93 to CHECK the spellings in a minute I’m not sure they’re
completely correct

facilitative role, seeking clarification (121, 123, 129) and eliciting
from the learners descriptions of their writing strategies. Clarification
requests are extremely valuable in promoting opportunities for learning
since they ‘compel’ learners to reformulate their contribution, by
rephrasing or paraphrasing. There is clear evidence in this extract that
the teacher’s unwillingness to accept the learner’s first contribution (in
123, 125) promotes a longer and higher-quality contribution in 126.
Note too how the teacher shows confirmation and understanding (in
121, 123, 125, 129) through the backchannels ‘yes’ and ‘right’.
Backchannels are very important in all human interaction since they
tell the speaker that the listener has understood and is following what
is being said. They ‘oil the wheels’ of the interaction and ensure that
communication occurs. Consider how you feel during a telephone call
when there is silence at the other end of the line – you have no way
of knowing that you have been understood. The same is true in
classroom interactions.
Extract 1.2
121 T: =yes so tell me again what you mean by that?=
122 L: =the first is the introduction the second eh in this case
we have the ((3)) who you are to eh introduce yourself
a few words about yourself and where you live and
what I do [and]
123 T: [so ] . . . yes?=
124 L: =and then it’s the problem what happened . . .
125 T: yes=
126 L: =and you need to explain it and why you are writing
because probably you did something like you gave
the information to the police but it didn’t happen . . .
127 T: so can I ask you why did you write it in your head as
you said?=
128 L: =I don’t know it’s like a rule=
129 T: =right so it’s like a rule what do you mean? . . .
We have seen, then, that modified speech is a key element of class-
room interaction and one that can have profound effects on the
quantity and quality of the learning that takes place. Effective speech
modification ensures that learners feel safe and included and gives them
the confidence to participate in the interaction. It also minimises
breakdowns and misunderstandings and creates a sense of purposeful
dialogue in which a group of learners is engaged in a collective activity.
Although we do have some understandings of the role of speech
modification in teaching and learning, there are a number of questions
that remain unanswered:
Introduction10

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 11
• Do some types of speech modification create opportunities for
learning?
• Do some hinder opportunities for learning?
• How might teachers help learners modify their speech?
• How might teachers gain a closer understanding of the relationship
between speech modification and learner participation?
We shall return to these questions in Chapter 3.
Elicitation techniques
Elicitation techniques are the strategies used by (normally) teachers to
get learners to respond. Typically, elicitation entails asking questions.
Classroom discourse is dominated by question and answer routines,
with teachers asking most of the questions, while learners ask
correspondingly few questions. It is by asking questions that teachers
are able to control the discourse, especially given that they know the
answers to most of the questions they ask! Questions like these, where
teachers already know the answer (for example, ‘what’s the past tense
of go?’) are called display questions since they require learners to
display what they know. Classrooms are unique in that for most of
the questions that are asked, the answer is already known. Imagine if
you were to ask your friends or family questions to which you already
know the answer – they would find this very strange, abnormal even!
Yet in classrooms this practice is the norm. Display questions serve a
range of functions, including:
Task 1.2
Add as many reasons as possible to the following list of reasons
why teachers ask so many questions. Then suggest alternatives to
questions:
• To provide a model.
• To check comprehension.
• To test.
• To activate learners’ responses.
• To stimulate practice.
TC

• eliciting a response;
• checking understanding;
• guiding learners towards a particular response;
• promoting involvement;
• concept checking.
Essentially, the defining characteristic of display questions is to
check or evaluate: understanding, concepts, learning, previous learning
and so on. Responses tend to be short, simple, restricted, often
comprising one or two words. Rather than opening up space for
learning, they tend to close it down and result in a rather stereotypical,
almost mechanical type of interaction that is often exemplified in IRF1
sequences (see below).
Apart from display questions, teachers also ask genuine, more open-
ended questions, designed to promote discussion and debate, engage
learners and produce longer, more complex responses. These so-called
referential questions result in more ‘natural’ responses by learners,
often longer and more complicated, and resulting in a more conversa-
tional type of interaction. Referential questions often begin with a
wh- question such as who, why, what, etc. From a teaching and
learning perspective, the distinction between display and referential is
less important than the relationship between a teacher’s pedagogic goal
and choice of question. If the aim is to quickly check understanding
or establish what learners already know, display questions are perfectly
adequate. If, on the other hand, the aim is to promote discussion or
help learners improve oral fluency, then referential questions are more
appropriate. The extent to which a question produces a communicative
response is less important than the extent to which a question serves
its purpose at a particular point in a lesson. In short, the use of
appropriate questioning strategies requires an understanding of the
function of a question in relation to what is being taught.
Consider the two extracts following. In extract 1.3 on p. 13, the
teacher is working with a group of low-intermediate adult learners. The
class has recently read a story and here, the teacher is simply recapping.
It is immediately obvious that the turn-taking, participation and
contribution of each learner are all tightly controlled by the teacher’s
use of display questions. In lines 11, 13 and 15, the teacher simply
gets students to ‘display’ what they already know from what they have
read. The interaction is rapid and allows little space for full responses,
indicated by the latched turns (=: one turn follows another with no
pausing or silence). Learner responses are short, typically two or three
words and there is no space here for topic development (in lines 12,
14 and 16). We can surmise from this that the teacher’s goal was to
Introduction12

check understanding before moving on: her choice of display questions
here is entirely in tune with her teaching goal. The ensuing discourse
is ‘classic IRF’, with each teacher contribution serving to both evaluate
a learner response and move the discourse forward with another
prompt (again, in 11, 13, 15).
Extract 1.3
9 T: I’ll see if I have a (2) a photocopy (looks for papers)
right you can’t find it? look you
10 have this book and cos I’ve got another book here good
. . . so can you read question
11 2 Junya
12 L1: (reading from book) where was Sabina when this
happened?
13 T: right yes where was Sabina? (4) in unit ten where was
she?
14 L: er go out=
15 T: =she went out yes so first she was in the=
16 L: =kitchen=
17 T: =kitchen good and then what did she take with her?
18 T: L: =er drug=
19 T: =good she took the memory drug and she ran OUT
Compare extract 1.3 with extract 1.4 on p. 14. In 1.4, it is immediately
evident that learners have more interactional space and freedom in
both what they say and when they say it. This is a multilingual group
of advanced learners, preparing for a reading activity on the
supernatural. The teacher’s opening question is perceived as a genuine
one – he is seeking the opinions of the group. Note the two-second
pause (in line 50) and the relatively short responses by learners in 51
and 52. But it is the question why not in 53, accompanied by the seven
seconds of silence, which promotes the long learner turn in 54.
Seven seconds of silence is very unusual in most classrooms; typically,
the average wait time (the length of time that elapses between a
teacher’s question and learner response) is around one second. In line
54, and following seven seconds of silence, learner 3 produces an
elaborated response and works hard to express herself. While to us,
as outsiders, the meaning is not immediately apparent, the teacher seems
satisfied with her contribution and moves on to another student, Monica,
in 57. The teacher’s comments (in 53 and 57) are non-evaluative,
relating more to the content of the message than the language used to
express it. By being non-evaluative, and by asking genuine questions
and allowing pauses, the teacher succeeds in eliciting fuller, more
complex responses from the learners and in promoting a more engaged,
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 13

conversational type of interaction. His choice of questions is extremely
important to the resulting extended learner turns and produces a more
equal exchange, similar to casual conversation.
Extract 1.4
50 T: I agree do you do you believe in this kind of stuff? We
talked about UFOs and stuff yesterday (2)
51 L: no . . .
52 L: well maybe . . .
53 T: maybe no why not? (7)
54 L3: um I’m not a religious person and that’s the thing
I associate with religion and
55 believe in supernaturals and things like that and believe
in god’s will and that’s so far
56 from me so no=
57 T: I understand so and why maybe Monica? . . .
58 L4: well I’m also not connected with religion but maybe also
something exists but I
59 erm am rather sceptical but maybe people who have
experienced things maybe=
60 T: uh huh and what about you [do you]
Repair
Repair simply refers to the ways in which teachers deal with errors.
It includes direct and indirect error correction and the ways in which
teachers identify errors in the discourse. Clearly, there is a range of
types of error correction available to a teacher at any point in time.
As with all strategies, some will be more or less appropriate than others
at any given moment. The basic choices facing a teacher are:
• ignore the error completely;
• indicate that an error has been made and correct it;
• indicate that an error has been made and get the learner who made
it to correct it;
• indicate that an error has been made and get other learners to
correct it.
These choices correlate very closely to the work of conversation
analysts who recognise four types of error correction in naturally
occurring conversation: self-initiated self repair, self-initiated other
repair, other-initiated self repair, other initiated other repair (see Sacks
et al. 1974).
Introduction14

It is apparent when we look at classroom transcripts that error
correction occupies a considerable amount of teachers’ time. According
to van Lier, ‘apart from questioning, the activity that most characterises
language classrooms is correction of errors’ (1988: 276). He goes on
to suggest that there are essentially two conflicting views of error
correction: one that says we should avoid error correction at all costs
since it affects the flow of classroom communication, the other that
says we must correct all errors so that learners acquire a ‘proper’
standard. As teachers, we need to decide on the type and frequency
of error correction. Again, the strategies selected must be related to
the pedagogic goals of the moment. A highly controlled practice
activity requires more error correction than one where the focus is
oral fluency.
It is perhaps also true to say that, within the classroom, learners
do expect to have their errors corrected. While it may not be appro-
priate in more naturalistic settings for speakers to correct each others’
errors, in classrooms, this is both what learners want and expect. As
Seedhouse (1997: 571) puts it, ‘making linguistic errors and having
them corrected directly and overtly is not an embarrassing matter’.
Rather than deciding whether we should or should not correct errors,
teachers would do well to consider the appropriacy of a particular
strategy in relation to their intended goals. By adopting more con-
scious strategies and by understanding how a particular type of error
correction impacts on the discourse, teachers can do much to tailor
their error correction to the ‘moment’ and promote opportunities for
learning.
A sample of data is perhaps the best way to gain a closer
understanding of the need to tailor repair strategy to pedagogic goal.
Consider extract 1.5 on p. 16, in which the teacher is working with
a group of eight pre-intermediate adult learners. Her stated aim is ‘to
improve oral fluency’. The most striking feature of the interaction
is the overlapping speech (indicated [ ]). It is apparent from the data
that this teacher believes that repair is necessary; there are examples
of error correction in almost every teacher turn (277, 279, 281, etc.).
The student is really unable to express herself adequately owing to
the fact that the teacher interrupts so much in order to correct errors.
It is only in line 286 that she is really able to produce an extended
turn, presumably something the teacher wanted throughout given
her stated aim of improving oral fluency. While it is apparently this
teacher’s intention to help the learner by correcting errors, it is also
clear that over-correction is not very helpful. The flow of the exchange
is disrupted to the point that the learner is unable to clearly articulate
what she wants to say. While we are not claiming that this is a
deliberate strategy on the part of the teacher, this extract does underline
the need to match error correction strategy to the pedagogic goal
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 15

Introduction16
of the moment. In extract 1.6, we see a mismatch between the teacher’s
intended goal and the language used to realise it. The result is
interaction that is disjointed and lacking in coherence.
Extract 1.5
273 T: what about in Spain if you park your car illegally?
274 L4: . . . there are two possibilities
275 T: two [possibilities]
276 L4: [one] is er I park my car ((1)) and
277 T: yes . . . if I park . . . my car . . . illegally again Rosa
278 L4: (laughter) if I park my car [illegally]
279 T: [illegally]
280 L4: police stat policeman er give me give me
281 T: GIVES me
282 L4: gives me? a little small paper if er I can’t pay the money
283 T: it’s called a FINE remember a FINE yes?
284 L4: or if if my car
285 T: is parked
286 L4: is parked illegally . . . the policeman take my car and
. . . er . . . go to the
287 police station not police station it’s a big place where
where they have some
288 [cars] they
Extract 1.6
11 T: ok does anyone agree with his statement?
12 L: (2) erm I am agree=
13 T: =agree be careful with the verb to agree there you as
well Ensa that it’s WE
14 agree it’s not to be agree it’s to agREE Ok=
15 L: [oh I agree]
16 L: ((3))
17 T: I agree with you but not I AM agree with you the verb
is to agree ok so ((3)) to
Task 1.3
Look at extract 1.6 below. Comment on the teacher’s error correction
strategy. How appropriate is it here where the teacher is trying to
elicit student feelings and attitudes? What is the effect of the error
correction on the discourse?
TC

18 agree with (writing on board) is the preposition that
follows it I so it’s I agree
19 with you I disagree with you . . . ok em Silvie can you
em what were you going
20 to say?
21 L2: I agree with you because em when when we talk about
something em for
22 Example you saw a ((2)) on TV=
In this section, I have described some of the most important
features of classroom discourse and illustrated them using data
extracts. These features were teacher’s control of the discourse, speech
modification, elicitation and repair. I have tried to show how different
strategies are more or less appropriate according to the particular
pedagogic goal of the moment and the teacher’s understanding of local
context.
In the following section, I present a summary of the work on
spoken interaction in classrooms, whereby all interaction can be
described and analysed according to a three-part exchange structure.
The IRF exchange structure
One of the most important features of all classroom discourse is that
it follows a fairly typical and predictable structure, comprising three
parts: a teacher Initiation, a student Response, and a teacher Feedback,
commonly known as IRF, or IRE, Initiation, Response, Evaluation.
IRE is preferred by some writers and practitioners to reflect the fact
that, most of the time, teachers’ feedback is an evaluation of a student’s
contribution. Teachers are constantly assessing the correctness of an
utterance and giving feedback to learners.
This three-part structure was first put forward by Sinclair and
Coulthard in 1975 and is known as the IRF exchange structure. The
work of Sinclair and Coulthard has had a huge impact on our
understandings of the ways in which teachers and learners
communicate and has led to many advances in the field. IRF is also
known as a recitation script or tryadic structure (see the glossary).
Look at the example below:
Extract 1.7
1 Teacher: So, can you read question two, Junya. I
2 Junya: [Reading from book] Where was Sabina
when this happened? R
3 Teacher: Right, yes, where was Sabina. F
4 In Unit 10, where was she? I
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 17

5 Junya: Er, go out . . . R
6 Teacher: She went out, yes. F
(Walsh 2001)
In this extract, which is typical of all teacher–learner interaction
and occurs very frequently in classrooms all around the world, we can
see how the teacher opens the exchange and marks a new phase of
activity with the discourse marker ‘so’. This opening remark, or
initiation (I), leads to the question in line 1, which prompts the student
response [R] in line 2. In line 3, we see how the teacher offers feedback
(F) to what the learner has said (‘Right, yes’). Feedback is an important
feature of the three-part exchange since it allows learners to see
whether their response has been accepted or not. Frequently, feedback
entails some kind of evaluation, such as good, right, ok.
In line 3, the cycle begins again, with the next initiation (‘where
was Sabina?’), which is then clariﬁed in line 4 (‘in unit 10, where was
she?’). In 5, we see the learner’s grammatically incorrect response (‘er
go out’), followed in 6 by the teacher’s feedback and correction. This
second IRF sequence follows very logically from the ﬁrst and was
probably followed by a third. Based on this very brief extract, we can
make a number of observations about IRF, the most commonly
occurring discourse structure in any classroom:
• It enables us to understand the special nature of classroom
interaction.
• It enables us to understand why teachers talk so much more than
learners: for every utterance made by a learner (R), teachers typically
make two (I, F).
• It allows us to see how, if overused, classroom interaction can
become very mechanical, even monotonous. Teachers need to be
aware of this.
While the IRF sequence is both commonly found and appropriate
at certain times, there are other types of exchange that are more
desirable/useful to learning. We’ll come back to this point later.
Sinclair and Coulthard’s original work took place in L1 primary
classes. Based on recordings of teachers and pupils interacting in class,
they produced a hierarchical model for understanding classroom
discourse. They found that there were three basic kinds of exchange:
1 question-and-answer sequences;
2 pupils responding to teachers’ directions;
3 pupils listening to the teacher giving information.
Introduction18

While it is true to say that conversations outside the classroom
frequently have a three-part structure, speakers do not usually evaluate
one another’s performances. Just imagine how your friends or family
members would feel if you were to ‘evaluate’ their remarks all the
time!
Here’s an example of a typical ‘real-world’ exchange:
Extract 1.8
1 A: What’s the last day of the month? I
2 B: Friday. R
3 A: Friday. We’ll invoice you on Friday. F/I
4 B: That would be brilliant. R
5 A: And fax it over to you. I
6 B: Er, well I’ll come and get it. R
7 A: Okay. F
In extract 1.8, a business encounter, the interaction is opened by
A in line 1 with a question (an initiation). B’s response in line 2 is
then conﬁrmed by A in line 3 (feedback), followed by a second initia-
tion by A (‘we’ll invoice you on Friday’). Note how this second
initiation is not a question, but still requires some kind of a response,
which B gives in line 4. Note too how, in everyday communication,
the feedback move is optional. B’s response in line 4 is followed by
another initiation by A in line 5. (Although, it is also true to say that
feedback does not always occur in classrooms, it is far more prevalent
than in everyday exchanges outside the classroom. That is, most
responses by learners receive some kind of feedback from the teacher.)
Going back to extract l.8, we see how the exchange concludes with a
third tryadic exchange in lines 5–7 comprising an initiation by A (line
5), a response by B (line 6) and feedback by A (line 7). In everyday
settings then, even the most simple, ordinary encounter such as a
question and response often has three parts to it and not two as people
often think. It is also interesting to note that in the world outside the
classroom, responses and follow-ups are not usually reactions to test
questions (speaker A is not testing speaker B on what day it is, unlike
the teacher, above, who was testing the learners’ understanding), but
show that the speakers have understood one another, and are satisﬁed
with the way the interaction is progressing (Friday / that would be
brilliant / okay).
For language teachers, understanding the discourse of the classroom
itself is crucial, for we teach discourse through discourse with our
learners. This is another way of saying that in many parts of the world,
the main exposure to discourse in English that learners will have is
in the classroom itself, via the teacher. A number of studies have
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 19

compared the discourse of the classroom with ‘real’ communication
(e.g. Nunan 1987). But as van Lier says (1988: 267), ‘the classroom
is part of the real world, just as much as the airport, the interviewing
room, the chemical laboratory, the beach and so on.’
From this brief introduction to the exchange structure of classrooms,
we can make a number of important observations:
• All classroom discourse is goal-oriented. The responsibility for
establishing goals and ‘setting the agenda’ lies largely with the
teacher. Pedagogic goals and the language used to achieve them are
very closely related, even intertwined.
• The prime responsibility for what is said in the classroom lies with
the teacher. Teachers control the discourse through the special
power and authority they have, but also through their control of
the discourse. They control who may speak and when, for how long
and on what topic. They control turn-taking through the use of
IRF; not only do they initiate a response, they offer an evaluation
– further evidence of control.
• Learners take their cues from the teacher and rarely initiate a
response. Their role, one which they are socialised into from a very
early age, is to answer questions, respond to prompts and so on.
• The IRF sequence enables us to understand interaction in the
classroom, and comprehend its special nature. An awareness of IRF
enables us to consider how we might vary interaction more and
introduce alternative types of sequence (see Chapter 8).
• An understanding of the IRF sequence enables us to model spoken
language in the world outside the classroom, suggesting ways of
constructing dialogues for teaching, role-plays for practicing
conversation, etc.
Challenges for teachers and learners
From what we now know about classroom discourse, what are the
challenges that lie ahead for both teachers and learners? In this brief
overview, I posit a number of directions for future developments and
then return to these in subsequent chapters.
One of the most striking and noteworthy observations about
classroom discourse and language teaching is how little time is actually
spent making language teachers aware of its importance. Most teacher
education programmes, either pre- or in-service pay very little attention
to classroom interaction. Typically, teacher education programmes
offer some kind of subject-based preparation and training in classroom
methodology; a model comprising two strands which is used all over
Introduction20

the world. I would advocate a ‘third strand’ on teacher education
programmes that deals specifically with interaction in the classroom.
The aim is to sensitise language teachers to the centrality of interaction
to teaching and learning and provide them with the means of acquiring
close understandings of their local contexts. I suggest that classroom
processes will only improve once teachers have the means of
understanding local context and are able to improve it. Classroom
interaction lies at the heart of this.
A second and related challenge for teachers is the need to acquire
what I am calling Classroom Interactional Competence (CIC).
Although this is given a fuller treatment in Chapter 8, it is worth
introducing it briefly here. When we analyse classroom transcripts, it
is immediately obvious that levels of interactional competence vary
hugely from one context and from one teacher to another. Some
teachers, at some points in time, are very adept at managing interaction
in such a way that learning and learning opportunities are maximised.
Others use interactional strategies that ‘get in the way’ and that impede
opportunities for learning (Walsh 2002). Examples have been
presented throughout this chapter in the various data extracts we have
studied.
I define CIC as ‘teachers’ and learners’ ability to use interaction as
a tool for mediating and assisting learning’ (Walsh 2006: 130). The
assumption is that by first understanding and then extending CIC, there
will be greater opportunities for learning: enhanced CIC results in more
learning-oriented interactions. Teachers demonstrate CIC in a number
of ways. For example, ensuring that language use and pedagogic goals
are aligned is an important characteristic of CIC. As we have seen in
some of the extracts presented here, teachers’ use of language and their
goals must work together. Other features include the use of extended
wait time: allowing a reasonable time to elapse after asking a question
and not interrupting students all the time; extending learner responses
by careful management of the interaction and paraphrasing a learner’s
utterance, for example. Similarly, teachers need to be able to help
learners as and when needed by scaffolding a contribution, offering a
key piece of vocabulary or introducing a new phrase as and when
needed, for example. Achieving CIC will only happen if teachers are
able to understand interactional processes and make changes to the
ways in which they manage classroom interaction.
Other challenges facing teachers in the future is the need to gain a
fuller understanding of the relationship between classroom method-
ologies and classroom interaction. This will be dealt with more fully
in Chapter 2, but suffice to say here that a closer understanding of
how interactional features manifest themselves in, for example, task-
based learning, can only be of benefit to teachers and learners alike.
How, for example, does task-type affect interaction and what is the
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 21

consequence for learning? How might more effective management of
classroom interaction result in a more engaged, more dialogic type of
learning? And what do we know of the importance of interaction
during feedback following a task? This, according to many researchers
is the most important part of the task-based cycle and the one most
likely to lead to learning. There is much work to do in this area.
From a learner’s perspective, a number of challenges lie ahead.
Perhaps the biggest and most difficult one is the need to change the
interactional behaviours of learners so that they play a more equal role
in classroom discourse. When we consider the ways in which learners
are socialised into certain types of classroom behaviour, this is a huge
challenge. In most content-based subjects, learners answer questions,
respond to cues, follow the teacher’s initiative, avoid interrupting and
so on. And yet, in a language classroom, a very different set of
interactional traits is needed if learners are to play a more equal part
in the discourse. In language classrooms, we need learners to both ask
and answer questions, to interrupt where appropriate, to take the
initiative, seize the floor, hold a turn and so on. By following learnt
behaviours that are the product of many years of being socialised into
classroom rituals and practices, we may be facilitating the kind of
‘smooth’ discourse profile that prevails at the moment. But are we
helping to create interactions that result in learning? I suggest that we
need to encourage interactions that have a more ‘jagged’ profile in which
learners play a more central role in co-constructing meanings and in
ensuring that there are opportunities for negotiation, clarification and
the like. A jagged classroom interaction profile has more of the features
that would be found in naturally occurring interactions such as
everyday conversation, business encounters and the like. While not
denying that the language classroom is a social context in its own right,
many of its features are determined by the fact that control of the
communication lies with the teacher. In other contexts, roles are much
more equal, resulting in different interactional features. Turns are
longer, for example, and there are more frequent topic changes.
Overlaps and interruptions are more common, as are pauses. I am
suggesting that it is in this kind of interaction that learners have the
opportunity to acquire the kinds of linguistic and interactional
resources that will help them develop as learners. Teachers, while still
playing a more central role, would need a far more sophisticated
understanding of classroom discourse in order to be able to manage the
interaction. We’ll return to these ideas in Chapter 8.
Summary
In this chapter, I have provided a brief sketch of some of the main
features of L2 classroom discourse, presented under four main themes:
Introduction22

control of patterns of communication, teachers’ modified language,
elicitation and repair. These themes, or features of classroom discourse,
have been chosen because they are representative of the kinds of inter-
action that typically occur in second language (and other) classrooms.
One of the most commonly found structures – the IRF exchange –
was then introduced and exemplified. This ‘tryadic discourse’, first
postulated by Sinclair and Coulthard in 1975, is still by far the most
commonly occurring discourse structure to be found in classrooms all
over the world. Finally, we considered some of the challenges facing
both teachers and learners in language classrooms. In Chapter 2, we
build on some of these themes by looking at the relationship between
classroom discourse and teaching.
Note
1 IRF is explained in the next section. It means teacher initiation, learner
response, teacher feedback and is the basic unit of discourse in any
classroom.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Introduction 23

2 Classroom discourse
and teaching
Introduction
In this chapter, we look at the relationship between classroom inter-
action and teaching. The basic argument is that teachers can do much
to improve their professional practice and enhance learning by studying
their own interactions with students, a theme to which we shall return
later in the book (see Chapters 6, 7 and 8). The starting point here is
to consider how classroom contexts are created through interaction
and to then identify how teachers might gain a closer understanding
of specific features of the interaction. We then look at the place of
interaction in one specific teaching methodology, task-based language
teaching (TBLT), before looking at the ways in which teachers’ use of
linguistic and interactional resources can result in enhanced teaching.
In the final section of the chapter, we look at ways in which a focus
on classroom interaction can result in teacher development.
Classrooms as contexts
The word context has many different meanings and associations. A
quick search on the internet produces these:
1 The part of a text or statement that surrounds a particular word
or passage and determines its meaning.
2 The circumstances in which an event occurs; a setting.
3 The parts of a piece of writing, speech, etc., that precede and follow
a word or passage and contribute to its full meaning. Example:
‘It is unfair to quote out of context.’
4 The conditions and circumstances that are relevant to an event,
fact, etc.
(www.thefreedictionary.com/context)
From these definitions, we can see that the word context can be
used to describe both the background against which an event took

place and the language used in that event. Of course, there are many,
many ways of looking at context, but seeing contexts as a set of
circumstances accomplished through language will suit our purposes
here. Classrooms are unique social contexts in their own right. As we
saw in Chapter 1, classrooms can be characterised and described by
looking at a range of interactional features such as teacher elicitation
strategies, learner responses and teacher evaluations. Any of the
features identified in Chapter 1 mark the discourse as being typical of
a classroom. Most of us reading a transcript of a lesson would be able
to identify it as a piece of classroom discourse, and make specific
comments on what was happening in the discourse.
On the one hand, then, we can say that classrooms are easy to
identify and characterise as sites of social interaction. They have
interactional features that are easy to recognise and they follow certain
routines and procedures that set them apart from other social contexts
such as business meetings or dinner parties. And we are able to talk
about classroom contexts in different settings. For example, we know
that by studying transcripts, we can say with some certainty whether
the interactions took place in a primary or secondary classroom, an
L1 or L2 context, a lecture theatre, online, and so on. The reason that
we are able to do this is that the interaction reveals ‘what is really
happening’ in a classroom and enables us to make assumptions about
teaching and learning. So when we talk about the context of a lesson,
we can point to things like the age and level of the learners, whether
it’s a foreign or second language setting, who the teacher is, what
materials are being used, the amount of freedom the teacher has to
follow her own agenda, and so on. This is context at one level, and
it certainly provides us with useful information that may inform
practice or aid decisions concerning planning materials, writing a
curriculum, selecting assessment procedures and so on.
However, what we are interested in here is a different view of
context, one that is much more fine-grained and that helps us under-
stand the moment-by-moment decisions made by teachers ‘online’, as
they teach. Here, then, we are concerned to identify features of the
interaction that give a lesson or a teaching session its unique shape,
or ‘architecture’ (Heritage 1997). Only by looking in some detail at
the interactions that take place, I suggest, can we gain the kind of up
close and emic or insider understandings needed to inform good
practice. It is by looking at the ‘interactional architecture’ (c.f.
Seedhouse 2004) of a class, that we, as teachers, get some sense of
what is happening and, more importantly, why it is happening.
Detailed examinations of classroom discourse reveal how inter-
actants collectively co-construct meanings, how errors arise and are
repaired, how turns begin, end and are passed or seized. We can
identify specific features of the discourse that help us understand how
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Classroom discourse and teaching 25

teaching and learning are accomplished. Features like direct error
correction, wait-time, teacher echo, display questions and so on
provide vital clues as to the ways in which ‘space for learning’ is either
opened up or closed down. This kind of analysis can help us answer
questions such as:
• To what extent do teachers include or exclude learners from the
interaction?
• How are opportunities for learning created?
• Who holds the floor and for how long?
• What types of question are asked and how are they answered?
• How appropriate is the language to pedagogic goals?
Which types of discourse promote student engagement and
dialogue?
Perhaps more importantly, a fine-grained and detailed analysis of
micro-contexts offers us unique insights into what is being taught, how
and what learners are learning. By looking at the moment-by-moment
management of turns and topics we can see, in the interaction, what
is being learnt, what is not being learnt, what is the relationship
between what teachers teach and what learners learn. Our endeavour
is not simply to describe classroom interaction, it is to develop new
understandings and improve the ways in which we teach.
In this chapter, we turn our attention to the ways in which an
understanding of classroom discourse can help us enhance our under-
standing of teaching. We will look at how these understandings can
be enhanced in terms of whole methodologies (such as task-based
language teaching), in relation to the teaching of specific skills (such
as speaking or writing), and in response to specific strategies (such as
increasing wait-time, reducing teacher echo). Our focus is the moment-
by-moment interactions that collectively constitute a lesson, through
which contexts are co-constructed and in which meanings are
established. We begin with a consideration of what classroom discourse
can reveal about task-based language teaching.
Task-based language teaching
One of the most prevalent teaching methodologies to be found
throughout the world today is task-based language teaching (TBLT).
TBLT is derived from Communicative Language Teaching (CLT), in
particular the work of Hymes (1971) and Wilkins (1976). It aims to
bring ‘real-world’ contexts into the classroom, and it emphasises the
use of language for completing tasks rather than as a focus for study.
Classroom discourse and teaching26

Language is essentially incidental to task completion – the main focus
for learners should be the task. According to Willis (2001: 175), TBLT
sees language as ‘the driving force in language learning’.
One of the reasons for the popularity of TBLT is that interaction is
now widely regarded as being central to language acquisition because
it is through interaction that learners modify and develop their language
system – with or without teacher intervention. By completing a task with
others, especially oral communicative tasks, learners are able to identify
gaps in their own knowledge, notice connections between different
linguistic features, find ways of saying something even when they do not
have the most appropriate language, and so on. Essentially, task
completion puts learners in the kind of context in which they might find
themselves outside the classroom and obliges them to communicate
using whatever linguistic resources they have at their disposal.
Clearly, different tasks place different linguistic and cognitive
demands on learners, so teachers need to be able to predict what kind
of interaction a particular task will produce and what linguistic and inter-
actional resources learners will need to complete a task. Similarly, certain
tasks will generate more discussion and more interaction than others.
Consider task 2.1 below:
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Task 2.1
Look at the following two communicative activities, A and B. For each,
make a note of the kind of language that would be needed to complete
the task, and comment on levels of enjoyment and involvement that
each task might generate. Which task would you prefer and why?
What is the main difference between these tasks?
A
‘Girls and boys perform better and achieve greater success when they
attend single sex schools.’ To what extent do you agree or disagree
with this statement? Discuss in groups.
B
You have just crashed in a light aircraft on a remote, desert island.
You are one of three survivors. You must decide which of the following
items to use to help you stay alive. You may only choose five. Discuss
in groups and come up with a list and reasons for your choices:
• parachute;
• knife;
TC

According to Jane Willis (1996), one of the principal architects of
TBLT, there are three phases to this approach:
• pre-task;
• task;
• language focus.
Each phase will require different levels of involvement from both
teacher and students, different roles and different types of interaction.
Essentially, each task will pass through a number of stages, summarised
below:
Pre-task
• Lead-in. This will normally entail some kind of context setting
where the task and material are introduced. Here, learners’ previous
understanding of the topic, language or situation are activated (c.f.
schemata setting). The lead-in typically begins outside the material.
• Set-up. The task must be ‘set’, normally involving a clear instruction
and organisation of learning. How are students to work: in pairs,
groups, alone? Do they have some rehearsal time? How will
instructions be checked? Do students need a time limit?
Task
• Run the task. What is role of the teacher here? How to monitor the
task and make sure that learners are ‘on task’. How to deal with
student problems and offer support and guidance without being too
intrusive? How to maximise learners’ interactional space?
• box of matches;
• map;
• axe;
• bottle of whisky;
• ball of string;
• explosives;
• torch;
• magnifying glass.

• Close the task. How to bring the class back together in plenary?
How to deal with early or late ﬁnishers?
Language focus
• Feedback. How is feedback to be managed given that this is one
of the most important stages of TBLT in terms of learning?
How are learner contributions to be handled and supported or
scaffolded? How to ensure that learner contributions are maximised,
and optimised?
• Post-task. How to make the outcome of one task the focus of the
next? How to ‘build’ through the material?
Under TBLT, then, we can see that different types of interaction are
more or less appropriate according to which phase we’re at. If, for
example, the teacher’s goal is to set up a task, there will be minimal
learner involvement and teachers will simply ‘hold the ﬂoor’ and
normally give some kind of instruction that locates teaching in time and
space. While the aim of TBLT is to give learners greater opportunities
to play an active role in the learning process, it is both unrealistic and
inappropriate to discuss teacher talk in ‘quantitative terms’ as so often
happens. Teacher evaluations of their own interaction often refers to
the number of questions asked (‘I ask too many questions’), learner
involvement (‘they are a very passive class’) or simply the amount of
talk (‘I talk too much and my students say nothing’). As we have seen,
there are times when a lot of teacher talk is entirely appropriate and
necessary (setting up a task, for example), while there are others when
a much more ‘careful’ and restricted type of teacher talk is what is
needed (during the ‘in-task’ phase of TBLT, for example). In the next
section, we consider how teachers might make more or less use of
language in order to enhance the teaching (and learning) process.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Task 2.2
For each of the following extracts, and using the notes above, say
what stage of TBLT we are at, comment on the main focus of the
interaction and evaluate how effective the interaction is.
Extract 2.1
T: this morning we’re going to be reading about and talking about
er music and pop stars and famous people Last Christmas we
did a quiz do you remember and it was a picture quiz and some
TC

of the pictures were people from your countries. Erm let me see
could you work with Ben and could you three guys work together
(T organises groups). If you recognise any of those people who
they are and what you know about them and then what kind of
lifestyle do you think these people have.
Extract 2.2
253 L3: Ahhhh nah the one thing that happens when a person
dies ((2)) my mother used to
254 work with old people and when they died . . . the last
thing that went out was the
255 hearing ((4)) about this person =
256 T: =ahaaa
257 L3: so I mean even if you are unconscious or on drugs
or something I mean it’s probably
258 still perhaps can hear what’s happened (2)
259 L2: but it gets ((2))=
260 T: =aha (2)
261 LL: /but it gets/there are ((2))/=
262 L3: =I mean you have seen so many operation ((3)) and
so you can imagine and when
263 you are hearing the sounds of what happens I think
you can get a pretty clear picture
264 of what’s really going on there=
265 L: =yeah=
266 L: =and and . . .
267 L1: but eh and eh I don’t know about other people but
eh ((6)) I always have feeling
268 somebody watching watch watches me=
269 L4: =yeah=
270 L1: =somebody just follow me either a man or a woman
I don’t know if it’s a man I feel
271 really exciting if it’s a woman ((4)) I don’t know why
like I’m trying to do things better
272 like I’m eh . . . look like this . . . you feel! it . . . I don’t
know=
Extract 2.3
66 T: =yes it’s the result of INtensive farming they call it
(writes on board) which is er (2) yeah and
67 this is for MAXimum proﬁt from erm meat so as a
result the animals suffer they have very

Interaction and teaching
In this section, I present one extract of data (extract 2.4) as a means
of showing how teachers can create opportunities for learning through
their use of language and interactional resources (see Walsh 2002 for
a full version of this discussion). In the extract, the teacher facilitates
maximum learner involvement by constructing a micro-context in
which learners are maximally involved and in which learning
opportunities are clearly created. It is not my intention here to evaluate
the instructional skills of the teachers, merely to make the point that
much can be done to improve teaching when we pay close attention
to the way in which language is used.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
68 BAD conditions and very small erm they’re given
food to really to make them big and fat and
69 usually it’s unnatural and as you said they HAVE to
give them a lot of antibiotics because the
70 conditions in which they’re kept erm they have far
more disease than they would normally
71 have so they give them steroids to make them
stronger and of course this is now being
72 passed through to the HAMburger that you eat is
contaminated with er=
73 L: =sorry how do you spell anti- anti-biotics?
74 T: anti-biotics? anti-biotics yes? erm anti-biotics?
75 L: how to spell it?
76 T: oh how do you spell it right (writes on board) there’s
er I think I read a very shocking report
77 recently that nearly all for example chickens and beef
now pigs all all these that are reared
78 with intensive farming they’re ALL given anti-biotics
as a matter of course and of course the
79 public don’t hear this until quite a long time after
we’ve been eating it and this this is what
80 makes me angry quite a scandal really . . . sometimes
when I listen to these reports I think oh
81 perhaps I should be vegetarian and sometimes er you
wonder about the meat=
82 L4: =how the people who offer food on the street how
can you ((2))
83 T: =er you can’t er check that they’re=
83 L4: =no no I mean what the name?

In the extract that follows, there is clear evidence that the teacher,
by controlled use of language and by matching pedagogic and linguistic
goals, facilitates and promotes reformulation and clariﬁcation, leading
to greater involvement and precision of language on the part of the
learners.
Extract 2.4
In this extract with six pre-intermediate adult learners from Brazil,
Japan, Korea and Russia, the teacher’s stated aim is to provide
oral ﬂuency practice using material from Harrap’s Intermediate
Communication Games.
480 L4: the good news is my sister who live in Korea send eh
. . .
481 T: SENT=
482 L4: =sent sent credit card to me=
483 T: =ooh very good news . . .
484 L4: but bad news [is]
485 T: [the bad] news is . . .
486 L4: I don’t know password . . .
487 LL: /password/password/ (2)
488 L1: pin number=
489 T: =pin number . . .
490 L4: what?=
491 T: =pin number pin number=
492 LL: =/ahh pin number/pen number/=
493 T: =pin PIN not pen pin=
494 LL: =/pin/pin number/p-i-n/=
495 L1: =I always forgot my pin number
496 L: =ah pin number=
497 T: =I don’t know my pin number
498 L5: ((2)) she can phone you on mobile phone=
499 T: =she can . . .
500 L5: she can say [you]
501 T: [she can] . . .
502 L5: she can tell your pin number . . .
503 T: yeah she can [tell you your pin number ]
504 L5: [she can tell you] this pin number by phone . . .
505 L4: but I I can’t eh ring her because eh because eh the time
eh=
506 T: =the time difference?=
507 L4: =time difference=
508 L5: =you can count your time for example look what what’s
the difference time with your
509 country how many hours? (3)

510 L: eight hours=
511 L5: =eight hours ok you can phone early in the morning it
will be evening in your
512 country=
513 L4: =if I go to home if ((5)) if I call her Korea it’s eh [midnight]
514 L: [midnight] . . .
515 L5: ok you can phone in the morning ((3))=
516 L4: =yeah at [eight]
517 L5: [at eight] o’clock at nine o’clock you can call=
518 L6: =in Japan same=
519 T: =it’s the same eight hours?=
520 L6: =yeah=
521 L4: =I’m very busy=
522 L5: =what you are busy it’s eh just reason you [((5))]
523 L4: [((4))]=
524 T: =for god’s sake give him a break (laughter)
In the 44 lines of text that make up this extract, there are several
specific features of the teacher’s use of language that facilitate learner
involvement and create opportunities for learning:
Direct error correction
Maximum economy is used when correcting errors and the teacher opts
for a very open and direct approach to error correction, as preferred
by learners (Seedhouse 1997). This is far less time-consuming and
intrusive than the more ‘sensitive’ (and therefore time-consuming)
error correction strategies preferred by many teachers. Errors are
corrected quickly and directly (in 481, 485, 489, 493) and the discourse
is allowed to flow with minimal interruption by the teacher. While not
suggesting that all error correction should be direct and minimalist, there
is a certain logic in keeping error correction to a minimum in oral
fluency practice activities in order to reduce interruption and ‘maintain
the flow’. The teacher in extract 2.4 succeeds very well in achieving this
and the discourse is allowed to proceed with minimum interruption.
Content feedback
Many of the features of this extract mirror a naturally occurring
conversation and the teacher quite appropriately provides personal
reactions to comments made by learners: reacting to a comment made
(in 483) and making use of humour (in 524). Given that one of the
teacher’s stated aims is ‘to provide oral fluency practice’, her use of
conversational language is appropriate to her pedagogic purpose;
language use and pedagogic purpose coincide. The teacher’s use of
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

language strongly resembles utterances found in the ‘real world’ and
reinforces the aim of promoting oral fluency. Appropriate use of
conversational language creates an atmosphere that is conducive to
learning and is likely to promote learner involvement. Feedback on
the message rather than its form is also more conducive to genuine
communication and is more appropriate in the setting outlined here.
Checking for confirmation
Teachers who constantly seek clarification, check for confirmation and
who do not always accept the first contribution a student offers are more
likely to maximise learning potential than those who do not (Musumeci
1996). In the previous extract, the sole instance of the teacher checking
for confirmation (line 519) does serve to maintain the flow and keep
channels open. Here, a genuine question by the teacher not only facili-
tates a more interactive exchange, it ensures that learners are working
together and that misunderstandings are minimised.
Extended wait-time
One of the most striking features of the extract is its turn-taking
structure. As the discourse progresses, the teacher takes more and more
of a ‘back-seat’ and ‘hands over’ to the learners. In lines 507–523,
learners successfully manage the turn-taking and topic management
themselves with no intervention by the teacher. Extended wait-time,
the time allowed by teachers to answer a question (see, for example,
Nunan 1991) not only increases the number of learner responses, it
frequently results in more complex answers and leads to an increase
in learner/learner interaction. Again, this teacher confirms the import-
ance of maintaining harmony between language use and pedagogic
aim; the teacher’s use of language, consciously or subconsciously, is
very much in tune with her specific aim at this stage of the lesson.
Scaffolding
(See turns 489, 491, 497, 503.) Communication breakdown is a very
common feature of L2 classrooms. Often it occurs because learners do
not know a particular word or phrase or do not possess the appropriate
communicative strategies. To pre-empt breakdown, it is the role of the
teacher to intervene and feed in the missing language. Timing and
sensitivity to learner needs are of utmost importance and many teachers
intervene too often or too early. Scaffolding (Bruner 1990; Lantolf and
Thorne 2006) involves more than simply error correction. It is a skill
similar to the one possessed by many parents when helping their young
children struggling to find the right word at a given moment. It requires

the ability to listen actively and make economical use of language. The
examples in this extract illustrate this important practice very well:
latched modelling (in 489 and 491), where the teacher quickly models
the language needed at the end of a previous turn; alternative phrasing
(in 497), where she offers a different way of saying something;
prompting (in 503).
Of a total of forty-two turns (thirty made by learners, twelve by
the teacher), ten teacher contributions succeed (whether intentionally
or not) in engaging learners and in promoting longer, more complex
turns. Throughout much of the extract, there is clear evidence that
the teacher’s language use and pedagogic purpose are at one; that the
teacher’s stated goal of promoting oral fluency is consistent with her
use of language. Her verbal behaviour allows learners to play a full
and active role in the discourse, producing more complete, more
natural responses. Instead of ‘smoothing over’ the discourse and ‘filling
in the gaps’ by pre-empting learner responses, the teacher only
intervenes as and when necessary, giving language support, correcting
errors or adding a personal comment of her own.
As far as the learner contributions are concerned, it is evident from
this extract that learners and teacher are actively engaged in
constructing a piece of discourse which, in many respects, resembles
a conversation; an observation that again coincides with the teacher’s
pedagogic goal and reaffirms the need for teachers to be ‘in tune’ with
their aims and use of language as the lesson unfolds. Throughout this
piece, learners self-select (508–516), overlap (511/512, 514/515) and
latch (where one turn immediately follows another, as in turns
494/495, 507/508, 509/510); these are all features that are common
to naturally occurring conversation and add further weight to the
coincidence of language use and pedagogic purpose.
What becomes apparent from this extract is that we cannot talk
about the second language classroom context (singular); rather, we need
to talk about contexts (plural), in which teachers and learners jointly
construct the discourse. In extract 2.4, the discourse is constantly shift-
ing according to the changing agenda of the lesson and according
to different participation structures. Successful teaching entails using
linguistic and interactional resources that are ‘fit for purpose’, that enable
pedagogic goals and language use to coincide and that take account of
the changing nature of the discourse. Under this view of context, a
quantitative view of teacher talking time (TTT) is, to a large extent,
redundant. What is more important is the appropriacy of language used
in relation to the ‘context of the moment’ and task in hand.
Classroom discourse and teacher development
One of the most useful ways to help teachers develop and improve
their professional practices is to place classroom discourse at the centre
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

of the process. By helping teachers understand interactional processes
more fully and by getting teachers to study their own use of language
and its effects on learning, it is possible to greatly enhance microscopic
understandings of classroom processes, thereby improving the quality
of both teaching and learning.
A number of calls have been made over the years for teachers to
discover for themselves what kinds of interaction foster learning and
what the role of the teacher is in creating and managing that
interaction:
The challenge set before teachers is to recognise both the obvious
in their classrooms and the not so obvious within themselves and
their students, to understand how both of these dimensions shape
the dynamics of classroom communication, and to equip their
second language students with the competencies they need to get
the most out of their second language experiences.
(Johnson 1995: 169)
For teachers to implement pedagogical intentions effectively [. . .]
it is important [for them] to develop an understanding of the
interactional organisation of the L2 classroom.
(Seedhouse 1997: 574)
[. . .] the challenge for teachers of ESL students [is] to find out more
about the types of interaction that occur in their classrooms, and
to also reflect on teaching practice and curriculum implementation
which have the potential to facilitate second language development
in the classroom context.
(Glew 1998: 5)
Most of the studies on classroom interaction are descriptive rather
than pedagogic in nature (Thompson 1997: 99), concerned to further
understanding of the process of second language acquisition in the
formal context. While these studies have certainly increased under-
standing of the ways in which languages are learnt in the classroom
setting, they are less accessible to the very people they are intended to
influence: language teachers.
In the remainder of this chapter, we consider how language teachers
might improve their professional practices by focusing on classroom
discourse. In the discussion that follows, we evaluate how teachers’
use of their own classroom recordings can improve key aspects of their
performance. The discussion addresses four key areas for teacher
development that can be addressed by looking at classroom discourse
data:

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
1 Improving questioning strategies.
2 Making the discourse more communicative.
3 Improving interactive decision-making.
4 Dealing with reticence.
Improving questioning strategies
In Chapter 1, we saw how questioning occupies much of a language
teacher’s time. It is one of the most important strategies used in the
classroom to elicit response from learners. It is also, arguably, one of
the most difficult skills to master, even for the most experienced
teacher. In 1990, Keith Richards designed a ‘mini-course’ that set out
to improve teachers’ questioning skills through the use of film, self-
evaluation and micro-teaching. The quantitative results, summarised
below, showed a marked improvement in questioning strategies.
Questioning technique Before training After training
(frequency) (frequency)
Thought-provoking questions 37.3% 52.0%
Prompting questions 8.5% 13.9%
Repetition of teacher questions 13.7% 4.7%
Teacher answering own questions 4.6% 0.7%
This study clearly underlines the fact that teachers can be trained
to ask more appropriate (indeed, fewer) questions and hints at the
positive results – in terms of learner involvement – that such strategies
can bring. The results summarised by Richards indicate quite clearly
a change in the teachers’ questioning strategies and the increased
potential for learner involvement. Awareness raising, using data from
teachers’ own lessons, can have far-reaching effects on what students
learn and increase opportunities for meaningful participation.
In another study focusing on improving teachers’ use of questions,
Thompson (1997) highlights the need for language teaching profes-
sionals to ask appropriate questions and emphasises the complexity
attached to good questioning. In this project, Thompson set out to
help trainee teachers on an initial teacher education programme
to categorise question types as a way of helping them identify options
in their teaching. In other words, there was a concern to ensure that
choice of question reflected pedagogic goals. A three-way classifica-
tion of questions was devised, enabling questions to be analysed
according to:

(a) Form. Essentially, this is a grammatical categorisation: whether
the questions are closed or open, or, more traditionally, yes/no or
wh-. The distinction is important since closed questions are easier
to answer than open ones; asking a closed question may facilitate
involvement in the form of a short answer, which can then be
followed with an open wh- question to extend the learner’s
contribution.
(b) Content. The focus of the question may relate to personal facts,
outside facts or opinions. While the value of ‘personalising’
questions cannot be denied, since it often results in increased
learner investment in the discourse, Thompson argues that ‘per-
sonalisation does not necessarily mean that real communication
is taking place’ (1997: 101).
(c) Purpose. The distinction here is between questions that are for
display and questions that are for communication. The distinction
between referential and display questions has been stated elsewhere
(see Chapter 1). According to Thompson, the distinction between
display and referential questions is over-stated, the real issue being
the extent to which teachers behave as if they know the answer
to a question; genuine communication can be facilitated when
teachers act as if they do not know the answer to a question or
deliberately ask questions (about learners’ culture or professional
background) to which they do not know the answer.
Task 2.3
Comment on the questions used in the extract below. How effective
are the strategies used by this teacher? What evidence is there that
she is successful in handling questions?
28 T: Have you any idea what country life would have been
like?
29 L: Scary.
30 T: Why would it be scary?
31 L: Because it would be dark and there would be no cars or
anything.
32 L: There’d be no lights
33 T: But what’s it like in the country now?
34 L: It would have been hard work. There would have been big
tractors like till do the crops out of the ground.
35 T: Exactly.
TC

The main value of the framework lies in its potential to make
practitioners think about their choice of question in relation to what
they are doing in the classroom and, more importantly, about their
role in the interaction; a realisation, for example, that student contri-
butions do not always have to be evaluated (Thompson 1997: 105).
Arguably, the framework has even greater potential for raising
awareness when used in conjunction with teacher-generated data such
as audio recordings and/or lesson transcripts as against the reading
texts employed in the original study.
Making the discourse more communicative
For many years now, researchers and practitioners have been interested
in comparing the communication that takes place in classrooms with
that of the ‘outside’ or ‘real’ world. One of the main reasons for this
is the kind of concern expressed here by David Nunan more than
twenty years ago (1987: 144), ‘in communicative classrooms, inter-
actions may [. . .] not be very communicative after all’. The lack of
genuine communication in the contemporary second language
classroom, has been criticised more overtly by other researchers such
as Legutke and Thomas (1991: 8–9): ‘[. . .] very little is actually
communicated in the L2 classroom. The way it is structured does not
seem to stimulate the wish of learners to say something, nor does it
tap into what they might have to say.’
In a 2008 study, Scott Thornbury had teachers use their own lesson
transcripts as a means of raising awareness of the importance of
classroom interaction with small groups of trainee teachers. Trainees
were asked to record, transcribe and analyse one segment of a lesson,
identifying those features of their talk that were felt to be more or less
communicative and commenting on them in a written evaluation.
From their self-evaluations, the following features of communicative
classroom talk were identiﬁed:
• Referential questions, which require greater effort and depth of
processing on the part of the teacher, one possible reason for
language teachers’ preference for display questions over referential
questions.
• Content feedback, where the focus is on meaning, rather than
language form. This is quite rare in many second language
classrooms. Most of the feedback from teachers to students is form-
focused and deals with language-related issues.
• Wait-time. This is the amount of time a teacher waits after asking
a question before getting a response. Typically, it is very short, even
less than a second. Thornbury found that even slight increases in
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

wait-time result in an increase in the quantity and quality of learner
contributions and an increase in the number of learner questions.
• Student-initiated talk. Requests for clarification and confirmation
checks point to ownership of the discourse and suggest that learning,
through negotiated meaning, is taking place. When learners are
seeking clarification, asking questions or checking their own
understanding, they are more likely to be engaged with the learning
process.
The main conclusions from Thornbury’s study are that teachers
can be sensitised to making their classrooms more communicative
through a more appropriate use of language and interactional
resources. He also commented on the fact that getting students to work
in groups and pairs is not in itself ‘communicative’ and that learners
need to be helped and guided if they are to really engage with L2
classroom discourse. Later in this book, we’ll see how teachers can
foster student-centredness while performing a central role in managing
the interactions that take place. Simply ‘handing over’ to learners will
not, I suggest, in itself result in a student-centred, communicative
classroom.
In a second study aimed at helping teachers make more communi-
cative use of their language, Richard Cullen argues that there is a need
to analyse teachers’ use of language from a qualitative rather than
quantitative perspective. ‘Good’ teacher talk does not necessarily mean
‘little’ teacher talk; rather, effective teacher talk ‘facilitates learning
and promotes communicative interaction’ (1998: 179). Very often,
teachers are concerned to reduce the amount of teacher talk in their
classes, while, at the same time, increasing the amount of learner talk.
A quantitative approach to classroom talk is less helpful than a more
qualitative one in which talk is evaluated according to its purpose in
a lesson at a particular moment (see, for example, Seedhouse 2004;
Walsh 2006).
In the 1998 study, Cullen suggests that the communicativeness of
classroom discourse should be judged in accordance with what
constitutes ‘communicative’ in that context, as opposed to what is
communicative in other social contexts. Models of communication that
constantly compare the second language classroom with the ‘real
world’ are unrealistic and may be unattainable for many teachers; put
simply, the classroom has to be treated as a context in its own right.
Any attempt to judge the communicativeness of teacher (or learner)
talk should be made in relation to what is meant by ‘communicative’
in that context.
Four features of communicative teacher talk are identified –
referential questions, content feedback, speech modifications, negotia-
tion of meaning – in addition to features of non-communicative teacher

talk that include excessive use of display questions, form-focused
feedback, teacher echo, sequences of IRF discourse chains. Any analysis
of the communicative value of teacher talk should take account not
only of its ability to foster meaningful communication among learners,
but also of the extent to which it coincides with the teacher’s pedagogic
purpose. For example, the communicativeness of an instruction would
be measured in terms of how well that instruction was carried out or
the extent to which it permitted negotiation of meaning or clarification
by learners.
According to Cullen, there are three advantages to this approach.
First, the categories of teacher talk are based on what goes on in the
classroom, not what goes on in a another social context. Second, any
evaluation of the communicativeness of teacher talk is based on ‘what
it takes to be communicative in the context of the classroom itself’
(see above). Third, a model of communicative language teaching
recognising the importance of pedagogic functions is more likely to
be achieved by practising teachers than one that constantly compares
the L2 classroom with the outside world. By getting teachers to look
at the pedagogic value of their talk, the extent to which communication
and pedagogic purpose coincide, a far more realistic, and for many
practitioners, a far more easily attainable, model of classroom com-
munication can be constructed. The communicativeness of classroom
discourse has to be related to its context: the second language
classroom; only then can accurate and fair measures of its ‘worth’ be
made (Cullen 1998: 186).
Enhanced interactive decision-making
One of the key characteristics of effective teaching is good decision-
making. Put simply, good teachers make good interactive decisions;
decisions that are appropriate to the moment, that engage learners,
that seek out opportunities to teach, that facilitate learning and learning
opportunity.
Look at the extract on p. 42 in which a teacher working with a
group of intermediate adult learners is getting the class to describe
funny experiences from their school days. The extract begins with an
open question in 1, followed by a request for clarification in 2.
Following the clarification in 3, the learner gives a response in 4 that
is corrected and extended by the teacher in 5. His request for more
information (‘what happened, what made it funny’?) results in an
extended learner response (6), which is clarified by the teacher in 8
and re-articulated by L1 in 9. The teacher paraphrases this contribution
in 10 and this receives confirmation from L1 in 11.
What is interesting in this extract is the way in which the teacher
constantly clarifies, extends and summarises for the class. He does not
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

simply accept the first L1 contribution in 4, but asks for more details.
This process requires the teacher to make decisions that are interactive,
which enable learners to sustain a turn and allow the rest of the class
to follow the discourse. At any point in this interaction, the teacher
could have simply closed down the learner’s contribution and moved
on to another student. He chose not to and succeeded in engaging this
learner in an extended piece of spoken interaction.
Extract 2.5
1 T: what was the funniest thing that happened to you at
school (1) Tang?
2 L1: funniest thing?
3 T: the funniest
4 L1: the funniest thing I think out of school was go to picnic
5 T: go on a picnic? So what happened what made it funny?
6 L1: go to picnic we made playing or talking with the teacher
more closely because in the school
7 we have a line you know he the teacher and me the
student=
8 T: =so you say there was a gap or a wall between the
teacher and the students so when you=
9 L1: if you go out of the school you went together with more
(gestures ‘closer’ with hands)=
10 T: =so you had a closer relationship [outside the school]
11 L1: [yeah yeah]
Here’s what the same teacher said about this extract:
Basically he’s explaining that on a picnic there wasn’t this gap that
there is in a classroom – psychological gap – that’s what I’m drawing
out of him. There’s a lot of scaffolding being done by me in this
monitoring, besides it being managerial, there’s a lot of scaffolding
because I want to get it flowing, I want to encourage them, keep it
moving as it were. I’m clarifying to the class what he’s saying
because I know in an extended turn – a broken turn – and it’s not
exactly fluent and it’s not articulate – I try to re-interpret for the
benefit of the class so that they’re all coming with me at the same
time and they all understand the point being made by him.
The teacher’s comments suggest quite clearly that his interactive
decisions are both conscious and appropriate for this stage of the
interaction. He shows great awareness here of both this learner’s
needs and the needs of the class as a whole. He also makes very good
use of an appropriate metalanguage to discuss what he perceives to

be happening in the data (scaffolding, clarifying, extended turn, re-
interpret for the class, etc.). From these two extracts, we can see the
importance of developing interactive decision-making strategies in
order to promote learner engagement and to create opportunities for
learning.
In a 1996 study, Kathy Bailey addressed teachers’ decision-making
and looked at the ways in which teachers’ interactive decisions cause
them to depart from their lesson plan. Her concern was to develop
understandings of the decisions teachers took to depart from their
lesson plans. Essentially, the concern was to understand the ways in
which teachers move away from ‘ritualistic’ routines and practices
while teaching and adopt alternative practices at given moments in a
lesson. Participating teachers were simply asked to view lesson extracts
where they had departed from their lesson plan and justify the
interactional decisions taken. From their interview comments, Bailey
established the following six principles of interactive decision-making:
1 Serve the common good. Where several learners are perceived by
the teacher as having the same problem (raised by one student’s
question), there can be some justiﬁcation for dealing with the
problem (linguistic or otherwise) at length, resulting in a departure
from what was planned.
2 Teach to the moment. Lessons may be abandoned completely or
partially when a “window of opportunity” (page 28) results in the
teaching of something pertinent to the moment. (Here, the lesson-
plan was dropped completely in response to a question about the
American electoral college system.)
3 Further the lesson. Decisions pertaining to the organisation of
learning may result in changes to what was originally planned. For
example, a planned group-work activity may be rejected in favour
of completing the task individually.
4 Accommodate learning styles. Preferences by learners about the way
‘things get done’ in the classroom may result in departures from
what was planned.
5 Promote students’ involvement. Learners may be encouraged to
develop a contribution that is felt to be particularly valid or relevant
to a speciﬁc ‘teaching moment’.
6 Distribute the wealth: ensuring fair and even turn-taking by encour-
aging more reticent learners and ‘reining in’ the more vocal ones.
Much of the work done on initial and in-service teacher education
programmes focuses on the relationship between what was planned
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

and what happened in reality, or what teachers believe and what they
do in reality (Li and Walsh, forthcoming). Typically, teachers gain
credit for teaching to the plan and are criticised for deviating from
what was planned. Bailey’s study offers a very helpful alternative
approach and focuses on the decisions made while teaching. As I call
it, ‘online decision-making’ – pedagogic decisions made while teaching
– is at least as important as the decisions made before teaching or
while planning a lesson. Some may even say that online decision-
making, and the need for effective interactive decisions, are more
important than what goes on during planning. Good teaching is based
on far more than ‘planfulness’ (Leinhardt and Greeno 1986: 76). As
Bailey demonstrates quite clearly, skilled teaching is about the
interactive choices that are made online, about the decisions that are
taken in the fast flow of the lesson. Perhaps more importantly, good
interactive decision-making can be acquired by sensitising teachers to
the alternatives open to them, using the kinds of reflective practices
described by Bailey (ibid.).
Dealing with reticence
For many language teachers around the world, dealing with reticent
students is a very common problem and a dominant feature of
classroom life. Much learner discourse in larger classes is characterised
by single-word responses, whispered replies to teacher prompts, or
silence (see, for example, Lee and Ng 2009). Typically, there is very
little evidence of learner-initiated clarification requests, learner
questions or extended learner turns. Teacher talk may account for 80
per cent or more of the spoken interaction (c.f. Tsui 1998), with silence
and frequently repeated teacher questions predominant features of the
discourse. Many readers will identify very closely with this kind of
context.
Under such conditions, with minimal participation, it is unlikely
that much is being learnt at all. If learners are not engaged in the
discourse and participating actively, they have little opportunity to
try out (and later learn) new language, test their own hypotheses or
develop strategies for dealing with unknown language. For many
teachers, coping with silence is a major concern, often resulting in
teachers answering their own questions, a predominance of teacher
echo, an acceptance of one-word responses and a sense of frustration
and even failure.
In order to address this issue, Amy Tsui set up a study in 1996 in
Hong Kong in which teachers were required to make one video- or
audio-recording of a lesson, identify one problem in their classes, devise
strategies to overcome the problem, try them out and then evaluate
the whole process in a later lesson. The problem of learner reticence

manifested itself in a number of ways according to teacher perceptions.
These include: a lack of self-confidence and willingness to take risks;
a fear of mistakes and derision; a fear (on the part of teachers) of
silence; an uneven distribution of turns, with teacher bids going mainly
to the brighter learners; incomprehensible input – teachers’
contributions were too vague or too difficult to understand. The
practitioners in the study reported a number of strategies used to
overcome reticence in the language classroom:
• Lengthening wait-time. Silence was considered to be a waste of time
because, as one teacher commented, ‘time is too precious’. Earlier
studies confirm that teachers do not allow sufficient wait-time
owing to a need to cover the syllabus and a fear that silence
results in boredom and a loss of pace in the lesson (White and
Lightbown 1984: 236). Similar reasons were given by teachers in
the 1996 study. However, Tsui also reports that while silence is
‘not necessarily a bad thing’, excessive wait-time can increase learner
anxiety.
• Improving questioning strategies. Tsui notes that one of the
strategies identified by the teachers in the study was to improve their
questioning strategies by asking more referential questions and
fewer display questions. While an increase in referential questions
did not necessarily result in longer or better learner responses,
allowing planning time to answer those questions did; some teachers
commented on the value of getting students to write down their
answers before verbalising them.
• Accepting a variety of answers. Allowing a variety of responses and
emphasising that there was no one ‘right’ answer was found to be
of some value in initiating learner responses, with one teacher
allowing learners three choices after each question: answer, ask for
help or ask for more time.
• Making use of group work and peer support. Interaction in the
‘safety’ of a group before open class feedback was found to be
effective as a means of getting students to speak out since it allowed
rehearsal in a low risk, high gain situation.
• Providing content feedback. Content feedback was found to be
effective since learners knew that their errors would not be corrected.
Content feedback entails giving feedback on the message itself
rather than the words used to express the message.
In conclusion, Tsui suggests that a greater awareness of effective
anxiety-reducing strategies is needed if language learning is to be made
a more productive and enjoyable experience. Implicit in her remarks
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

is the suggestion that teachers need to be sensitised to strategies that
are effective in reducing anxiety, perhaps by reflecting on classroom-
recordings and having opportunities to discuss them.
To conclude this section and using some of the main themes
reviewed here, what do we now know about the role of classroom
discourse in improving teaching and learning? What are the main ways
in which an understanding of classroom discourse can be used in
second language teacher development?
The first thing to note is that any attempt to help teachers
understand classroom discourse is motivated, in the first instance, by
a desire to enhance learning. While the very term ‘teacher talk’
immediately conjures up images of teacher-centredness, most of the
studies reviewed here are instigated by a concern for the learners. The
use of appropriate questioning strategies is designed to increase learner
involvement, as is the concern to help teachers make more interactive
decisions – decisions that were exclusively guided by the learners.
Similarly, Tsui’s study (1996) was motivated by learner reticence.
Essentially then, learners are central to and instrumental in the process
of enhancing teachers’ awareness of language use in the L2 classroom.
A second theme that emerges from the discussion in this section is
the importance to teacher development of teacher-generated data.
Reflective practices are more easily accomplished when teachers analyse
their own data, using recordings from their own lessons. Perhaps more
importantly, we need to find ways of helping teachers analyse and
interpret data without having to transcribe everything. This is
something we return to in Chapter 6. For now, suffice it to say that
second language teacher development would benefit hugely from
having teachers collect, analyse and interpret data from their own
classes.
A third theme that emerges is the need to acquire an emic (insider)
perspective on classroom discourse and interactional processes. This
perspective is one that assumes that the use of teachers’ voices, actions
and stories is more conducive to gaining a true understanding of
interactional processes in the L2 classroom than the imposed
perspective of the researcher. An insider perspective is also easier to
obtain when teacher-participants are encouraged to verbalise their
experiences in some way. For example, in the studies reviewed in this
section, Bailey uses stimulated recall to initiate dialogues with
participating teachers; Tsui makes use of post-teaching interviews,
while Thornbury prefers a form of written self-evaluation. Each
procedure has its own merits and can be adapted according to local
conditions. Further, we can see that the complexities of classroom
interaction in the L2 context can best be understood when data are
collected using a variety of instruments: a multi-mode method. Apart
from enhancing reliability and validity, multi-mode methods of data

collection ensure that ‘all the angles are covered’, that the setting is
viewed from more than one position. From a teacher development
perspective, this means that any approach to helping teachers
understand classroom discourse should use more than one method:
for example, classroom recording with interview, lesson observation
plus focus group, etc. True understandings of the complexities of
classroom interaction can only be gained when we have as accurate
a picture as possible of what is really happening – this entails using
more than one type of data.
Summary
In this chapter, we have made a case for putting interaction at the
centre of effective teaching, using the argument that interaction lies at
the heart of effective classroom practice. By studying their own
interactions with students, either recorded or recorded and transcribed,
I suggest that teachers can do much to improve their professional
practice and enhance learning and learning opportunity. From an
understanding that contexts are not static and fixed, but dynamic and
fluid, we considered how teachers and learners co-create micro-
contexts through their interactions. It is in these micro-contexts that
learning occurs. The challenge for L2 teachers is to gain microscopic
understandings of their interactions with learners. By studying the
specific interactional strategies they use, teachers can promote more
active, engaged and dialogic learning environments. By placing
classroom interaction at the centre of their professional development,
language teachers are more likely, I suggest, to create micro-contexts
in which learning opportunities abound. In order to achieve this, I
would argue that teachers need to be equipped with appropriate tools
to study their own contexts. This is something to which we’ll return
throughout this book.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

3 Classroom discourse
and learning
Introduction
In Chapters 1 and 2, we established that L2 classrooms are as much
a social context as any other ‘real world’ context, such as a travel
agent’s, a dentist’s, an airport check-in counter. Rather than comparing
the L2 classroom to the ‘real world’ (which exists outside the class-
room), we should acknowledge that the L2 classroom is as much a
part of the real world as any other context. The point was made in
Chapter 2 that any language lesson should actually be seen as a series
of micro-contexts, which are jointly constructed by teachers and
students through their interactions. In all classroom contexts, linguistic
and interactional resources are the conduit through which opinions,
feelings, emotions, concerns are expressed and through which the
‘institutional business’ (i.e. language learning) gets done.
Our focus now shifts to an examination of the relationship between
interaction and learning, opening with a discussion of what we mean
by learning in an L2 classroom context. We consider why interaction
is regarded by many researchers and practitioners as lying at the heart
of learning and look at ways of uncovering the relationship between
interaction and learning. Much of the work that has been done to
explain the relationship between interaction and learning has
considered interaction from either a broadly cognitive position, in
which interaction somehow inﬂuences our ability to think and process
new knowledge, or from a more social perspective, in which learning
is viewed as a social process in which interaction plays a central role.
It is important to note that under both views of learning, interaction
plays a key and pivotal role.
Our discussion opens by considering the complex relationship
between language learning and interaction.
Language learning and interaction
Any discussion on learning a language in a formal context (a
classroom), as opposed to acquiring a language in an informal context
(outside the classroom), must begin with some deﬁnition of what is

meant by learning. We have all been language learners at some time
in our lives, many of us still are. Yet, what do we understand by
learning a second language? What does it mean ‘to know’ a language?
For most of us, learning means ‘acquiring’, ‘getting’, ‘having’,
‘possessing’ new skills or new knowledge. As learners, we say things
like ‘I can’t talk about what I did yet, I haven’t learnt the past tense’.
Eventually, once we ‘have’ the past tense, we are able to talk about
and describe past experiences. We often regard learning a language as
acquiring an ever-expanding repertoire of new skills and knowledge.
First, we ‘get’ the simple present tense, then we move on to the simple
past, present perfect and so on. We acquire one ‘building block’ (such
as ‘going to’ future) before we move on to the next. In many parts of
the world, languages are taught like this, in much the same way that
content-based subjects such as maths, history and science are taught.
We even talk about ‘delivering’ lectures, suggesting that knowledge
and skills can somehow be packaged and then offered to learners.
While it is easy to understand why this view of learning prevails,
it may be useful to consider an alternative. Rather than seeing learning
as ‘having’, we can also consider learning as ‘doing’ (see Larsen-
Freeman 2010). Under this view, learning is regarded as a process, an
activity, something we take part in, perform. Learning is regarded as
a dynamic, constantly shifting process in which participants collectively
construct meanings. Learning is not something we have or own, it is
something that we participate in – it entails encounters with others.
Learning is regarded much more as a social rather than a cognitive
process. Our actions, activities and interactions with others all work
together to determine what it is that we learn. Learning entails
completing a task, taking part in an activity, talking, discussing,
debating and arguing with others. It is this view of learning, as a
complex and dynamic process, which lies at the heart of contemporary
teaching methodologies such as TBLT (see Chapter 2).
It is not my suggestion here that learning is either having or doing;
it is almost certainly both (see Sfard 1998). What I am suggesting,
however, is that an understanding of learning in a formal, classroom
context must pay close attention to the ‘doing’ dimension since this
is something we can study, analyse and evaluate. We cannot look inside
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Classroom discourse and learning 49
Take a few minutes to think of your experiences as a language learner.
What does learning mean for you? Make a list, or write your own deﬁnition.
How have your ideas about learning changed over time?

Classroom discourse and learning50
the heads of our students and see what they are learning. We can look
at what they say, how they interact, how they use the L2 and so on;
this is where we can really begin to uncover some of the ﬁner nuances
of learning as a process. Under this view of learning, studying inter-
action, quite simply, is the same thing as studying learning (c.f. Ellis
2000; Pekarek Doehler 2010).
To summarise our discussion so far, we have established that:
• classrooms are social contexts just like any other context involving
human interaction;
• any lesson is made up of a series of micro contexts that are
established through the interactions that take place between
participants;
• it is through interaction that micro-contexts are created;
• learning is as much about what students do as what they have;
• interaction lies at the heart of learning.
Task 3.1
For extract 3.1 below, comment on what seems to be ‘happening’.
What do you think the teacher is trying to achieve here and what is
being learnt? How did you decide?
Extract 3.1
187 T: =what do we call I’m going to try and get the class
to tell you what this word
188 is that you’re looking for . . . er we talk about military
(claps hands) . . . military what?
189 L: ((1))=
190 T: =like ﬁght=
191 L: =kill=
192 T: =no not [kill]
193 L: [action] action=
194 T: =no ((2)) military?=
195 LL: =power=
196 T: =power think of another word military?
197 LL: ((3))force=
198 T: =so she believes in a FORCE for?
199 L: that guide our lives=
200 T: =that guides our lives=
TC

Interaction, from what we have said so far, is clearly very important
and something that we need to study. Why is this so? Why is interaction
seen as being so central to language learning? How can teachers and
learners gain a closer understanding of the interactional architecture
(Seedhouse 2004) of their classes? What impact might such an
understanding have on learning efﬁcacy? In the discussion that follows,
three reasons are presented for the centrality of interaction to learning.
According to van Lier (1996: 5), ‘interaction is the most important
element in the curriculum’, a position that coincides with that of Ellis
(2000: 209; original emphasis), ‘learning arises not through interaction,
but in interaction’. As such, interaction needs to be understood if we
are to promote learning. The discussion above supports both of these
claims: if we want to understand learning, we should begin by looking
at the interactions that take place in our classes. Further, there is now
increasing evidence to suggest that learner–learner interaction does not
have the impact on language learning that was once suggested (see,
for example, Foster 1998; Rampton 1999). This position suggests that
the teacher now has an important role to play in creating and managing
interaction that is ‘acquisition rich’ (Ellis 2000). The point is that even
in the most student-centred class, the teacher is instrumental to
managing the interaction (Johnson 1995); there is, then, a need to help
teachers acquire ‘microscopic understandings’ (van Lier 2000) of the
interactional organisation of the L2 classroom.
A second reason for studying interaction is that it allows practi-
tioners to make good interactive decisions. ‘Good teaching’ is as much
about the decisions teachers take while they teach as it is about the
planning that goes on before teaching (see Chapter 2). Good, effective
interactive decisions are the ones that inﬂuence learning, which create
space for learning (Walsh et al. 2010), which support learning rather
than impede it. Making appropriate choices in the interaction is not
something that can be easily mastered without some understanding of
the relationship between interaction and learning: a realisation, for
example, that asking yes/no questions all the time severely restricts
learners’ opportunity to offer extended turns and become engaged in
the interaction. Or that extended use of teacher echo (constant
repetitions) restricts the space that learners may have to interact.
Third, I suggest that if teachers are to enhance their understanding
of the interactional organisation of the L2 classroom, then the obvious
starting place is their own classes. For most teachers, professional
development entails improving the learning experience of their
students. In order to do this, I would argue, teachers should focus on
the interactions taking place in their own classes. For this to happen,
there is a need for more appropriate tools for investigating classroom
discourse, although a starting point for teacher development is for
teachers to simply record their classes and spend some time analysing
the interactions that took place. More will be said about this in
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Chapters 7 and 8. In the next section, the discussion offers an overview
of learning as a cognitive activity.
Learning as a cognitive activity
This section offers a necessarily brief overview of some of the more
important hypotheses that have attempted to explain the complex
relationship between interaction and learning. Three main hypotheses
are presented, all placing interaction at the centre of learning: Krashen’s
Input Hypothesis (1985), Long’s Interaction Hypothesis (1983, 1996),
Swain’s Output Hypothesis (1985, 2005).
These questions have been around for some time now and there is
still little agreement. Some researchers, for example, Cazden 1986,
argue that there is no direct relationship between the language used
in a language classroom and the language learnt; others, like Allwright,
suggest quite the opposite: ‘the processes of classroom interaction
determine what language learning opportunities become available to
be learned from’ (1984: 156).
According to Ellis (1990), it is helpful to distinguish between
interaction and formal instruction when assessing the role of classroom
communication in the process of SLA. Interaction is regarded as the
process by which samples of the target language are made available to
learners, while formal instruction is defined as the way in which the
teacher intervenes directly and provides samples of ‘specific linguistic
features for learning’ (1990: 93). For many teachers, providing samples
of language equates to modelling: the input that teachers provide and
that learners typically acquire in some form. According to Krashen
(1985), all input must be comprehensible if learning is to occur. His
input hypothesis is based on the assumption that learning will only occur
when the input learners receive – often in the form of a teacher’s
modelling – is at or slightly higher than the language level of the learners.
For modelling to be effective, it is clear that learners must be exposed
to an appropriate pronunciation, with correct word stress, intonation
and so on. Although the precise nature of a teacher’s input is still not
fully understood, it is clear that teachers do need to be sufficiently aware
of their use of language to both promote interaction and to allow
What is the relationship between the language used in the classroom and the
language learners acquire? Is there a direct, one-to-one relationship? What
can teachers do to maximise learning (i.e. acquisition) opportunities?

language acquisition to occur. Language that is too complex or too
simple will fail to engage learners in a level of activity that is at or slightly
beyond their level of competence. Language that is too ‘dense’, ‘wordy’,
or ‘idiosyncratic’ will have a similar effect.
And yet, despite the claims of Krashen, it is quite clear that
managing interaction entails far more than modifying input for
learners. Simplified input will not in itself result in SLA; comprehensible
input is ‘an insufficient condition for second language acquisition to
occur’ (Glew 1998: 1). Indeed, it is quite feasible that over-simplifica-
tion may have a counter effect on acquisition since negotiation will
become redundant (Musumeci 1996). Quality interaction, interaction
that is ‘acquisition rich’ (Ellis 1998: 145) has to be initiated, managed
and sustained by teachers through careful and knowing management
of the turn-taking sequences that occur in face-to-face communication.
Put differently, teachers and learners need to gain a comprehensive
understanding of the interactive processes that facilitate learning.
Interaction does not simply happen, nor is it a function of the teaching
methodology; interaction, in an acquisition rich classroom, is both
instigated and sustained by the teacher.
To illustrate this point, look at the ways in which interaction is
managed in extract 3.2 below. This teacher is working with a small
group of eight pre-intermediate adult learners in a multilingual context.
In the discourse, it is apparent that she uses a number of different
strategies to manage the interaction and ensure that the group is kept
together. In 285, for example, she acts as a model, correcting an error
and modelling the correct form, a task that is done quickly and
unobtrusively (indicated by the latched turns, marked =) in order not
to disrupt the flow of the interaction. In 287, she acts as ‘support’,
reinforcing L3’s contribution and making it available to the rest of
the class by writing it on the board. Finally, in turns 289, 291 and
293, her role is to inform, to act as a source of linguistic input in
response to the learner’s prompt in 288. This skilful management of
the discourse, entailing the adoption of different roles, ensures that
the class ‘stays together’, that learners are able to follow the lesson
and that learning opportunities are maximised.
Extract 3.2
284 L3: [the good] news is he boughted the new car=
285 T: =he bought a new car=
286 L3: =a new car a new car but bad news is (2) he crashed it
crashed crashed it his car=
287 T: =he crashed it yes the good news is (writes on board)
he bought a new car the bad news is he crashed it=
288 L3: =so if I want to say accident how to say?=
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

289 T: =he HAD an accident=
290 L3: =he had an accident=
291 T: =he had an accident or he crashed his car he crashed
his car=
292 L3: =what what is formal? (2)
293 T: ahh . . . about the same really he had an accident more
formal maybe . . .
294 L3: he had an accident (2)
Following on from the work of Krashen, Long proposed a slightly
different perspective on the role of interaction in learning. Long’s
interaction hypothesis (1983, 1996) argued that learning occurs when
students are negotiating meaning. Negotiation of meaning (also
‘negotiation for meaning’) was regarded by Long as being central to
second language acquisition. It is also the principle on which CLT
(communicative language teaching) is based. Essentially, learners
negotiate meaning when they are seeking or offering clarification and
confirmation. It is through this constant process of negotiating meaning
that second language learning occurs, according to Long, that is.
Long’s interaction hypothesis has received considerable scrutiny over
the years and has been found to be more or less in evidence according
to the different studies that have taken place. Early studies (see, for
example, Doughty and Pica 1986; Gass and Varonis 1985) found that
meanings are only negotiated when participants are all non-native
speakers of varying proficiencies and with different L1s. They also found
that the extent to which negotiation of meaning is determined is largely
dependent on who has access to the information: a learner seeking
information is more likely to negotiate meaning than a learner holding
it, for example. One of the main criticisms of Long’s hypothesis is that
the work was conducted under ‘laboratory conditions’; learners were
taken out of their regular classes and asked to perform certain tasks.
When we consider the extent to which negotiation for meaning occurs
in ‘normal’ classes, we find that it is considerably less than was first
claimed (see Foster 1998).
According to Long (1983, 1996), SLA is promoted when, through
communication breakdown, learners have to negotiate for meaning. By
asking for clarification and confirming comprehension – key features
of Long’s interaction hypothesis – it is argued that acquisition occurs.
Of course, in any conversation, meanings are negotiated and it can be
claimed that, to a large extent, the classroom is no different. The ‘social
process of negotiation of meaning’ (Pica 1997: 60) has been of consider-
able interest to researchers for many years. Negotiation enables learners
to provide each other with comprehensible input, to give and gain feed-
back on contributions and to modify and restructure utterances so that
meanings are made clear.

In his more recent version of the interaction hypothesis, Long
highlights the centrality of the more competent interlocutor in making
input comprehensible, in enhancing learner attention, and in
encouraging learner output. In Long’s own words (1996: 451–452):
[. . .] negotiation for meaning, and especially negotiation work that
triggers interactional adjustments by the NS or more competent
interlocutor, facilitates acquisition because it connects input, internal
learner capacities, particularly selective attention, and output in
productive ways.
There is clearly considerable emphasis here on the role of the
teacher, the more ‘competent interlocutor’, who is crucial to ensuring
that input is comprehensible and that learner output is ‘shaped’ in some
way so that it is ‘productive’. There seems to be a suggestion too in
the above citation that there is a need to reconsider the interaction
that occurs between teacher and learners, a departure from an earlier
version of the interaction hypothesis (Long: 1983), which also
addressed learner–learner interaction.
The teacher in extract 3.3 below is working with a group of
multilingual, advanced adult learners. He is preparing to introduce a
reading comprehension task on poltergeists and precedes the reading
with an open class discussion on ‘out of body’ experiences. It is clear
that the students in the extract have a high level of English and are
able to express themselves quite well, even when discussing complex
topics. Confusion arises in 146, however (indicated ((2)), as the student
is unable to recall the word ‘smoke’. The teacher’s clarification request
in 147 highlights the need for greater precision on the part of L1; his
request for clarification draws attention to the fact that there is a
problem that needs to be repaired. He interjects again in 151 and his
prompt eventually leads to another student offering the word smoke
in 155. Essentially, then, turns 146 to 158 entail negotiation for
meaning and arguably contribute to the extended and complex turn
produced by L1 at the end of the extract in 160. The teacher plays a
vital role in the discourse, highlighting the problem in the first instance,
offering support in the form of scaffolded input and finally guiding
the learner – with the help of a classmate – to her intended meaning.
Extract 3.3
144 L1: [I believe] in soul no I don’t believe in heaven certainly
not I believe that the universe eh we have place where
the souls are together and eh I remember ((5))we were
in sauna it was my grandmother
145 T: yes=
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

146 L1: =and something went wrong there and eh . . . the the
((2)) couldn’t go through the chimney
147 T: which couldn’t? . . . the the steam or . . . no?=
148 L1: =not steam eh blow? blow? not blow . . . eh something
from their fire they have eh=
149 T: =yeah=
150 L1: =a blow . . . not blow=
151 T: =a fire?=
152 L1: =when you make a fire=
153 T: =yes=
154 L1: =and eh what is=
155 L: =smoke?=
156 L1: =smoke smoke yes=
157 T: =uh huh=
158 L1: =smoke wouldn’t go through the chimney and the sauna
was erm warmed up?=
159 T: =yes=
160 L1: =warming up and it all stays in this room and the
room was full of this smoke and ((4)) and somebody
eh I don’t know eh probably a priest but I remember
erm how I saw my body on the floor lying down like
I was sleeping and my grandmother was pulling me
to the cold shower and then suddenly I like eh again
was con conscious?=
161 T: =yes=
We have already acknowledged that much of Long’s work is
based on studies that were conducted outside the classroom, under
‘laboratory conditions’. When we look at negotiated meaning in
the classroom, a different picture emerges, with many learners
negotiating only at word level, by repeating utterances, by remaining
silent, or by avoiding negotiation by ‘pretending’ to have under-
stood (see, for example, Tsui 1996; Foster 1998). In a 1996 study,
Musumeci found very little or no negotiation; indeed, quite the reverse
(1996: 314):
[. . .] teachers [. . .] speak more, more often, control the topic of
conversation, rarely ask questions for which they do not have
answers, and appear to understand absolutely everything the
students say, sometimes before they even say it!
The overall conclusion in Musumeci’s (1996) study – that there was
little or no evidence of sustained negotiation – is significant to the
current context; learners’ ability to formulate, re-formulate, clarify and

seek clarification are important indicators not only that language
acquisition has taken (or is taking) place but also that something is
being understood and eventually learnt. By ‘filling in the gaps’, teachers
may facilitate a coherent and flowing discourse, but they may also
deny learners opportunities to get to grips with target language forms
and identify potential problems in understanding. The present situation
indicates that negotiation is barely occurring at all in many L2
classrooms; significant changes in attitudes, expectations and the verbal
behaviour of both teachers and learners are necessary if language
acquisition and learning potential are to be optimised.
In terms of classroom communication, if negotiation for meaning is
to occupy a more central role in the learning process, there is a need
for a major shift in the ways in which understandings are accomplished.
This will entail greater emphasis on negotiated understanding, more
requests – by teachers and learners alike – for clarification, less accept-
ance by teachers of the ‘first response’ given by learners. In short,
discourse that may be less ‘easy on the ear’, but through which learning
opportunities are maximised and where problems and shortfalls (in
language acquisition) are more transparent.
Extract 3.4 offers an example of classroom discourse that is ‘less
easy on the ear’, but that allows identification of problems and
deficiencies in students’ language. Here, a group of advanced learners
is recalling memories of supernatural experiences. L1 is recalling an
incident from her childhood when she passed out in a sauna. Confusion
arises in 174: the teacher seeks clarification, asking L1 if the ‘person’
watching from above had a body. This is misunderstood in 175 as
the student thinks the teacher is alluding to her real body in the
sauna and continues recalling the experience in 177. The teacher
persists with requests for clarification in 178 and 180 and eventually
gets the response he was looking for in 181. While this whole
extract lasted less than one minute, it serves to illustrate a number of
valuable strategies, each of which, in some way, contributes to SLA.
First, there is a need on the part of teachers to really listen to a learner’s
response and evaluate its communicative potential. That is, is there
a message and is it unambiguous? Second, teachers would be well-
advised not to always accept a learner’s first contribution. Here,
for example, it would have been quite common practice for the
teacher to have accepted L1’s contribution in 175 and there would
have been no further negotiation for meaning. Third, teachers
should be prepared to persevere until they are satisfied that the intended
meaning has been conveyed. Furthermore, one reason these strategies
are valuable, not only from a SLA perspective, is that this is what
learners will encounter outside the classroom in their day-to-day
dealings with native-speakers.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Extract 3.4
172 T: =and did you see colours? was it like a colour [((2))]
173 L1: [it was] white it was white because I thought it was
smoke like it was like somebody smoking white and
green and [but]
174 T: [you you] saw from above did you see did you have a
body?=
175 L1: =I have body it was naked of course eh no [I did have
body]
176 T: [no I didn’t mean]
177 L1: I I (student wolf whistles) no I mean I watched me like
I was somewhere in the corner of the ceiling ((3)) and
I watched my body on the ﬂoor and my friend was
lying down . . .
178 T: Sure but but the YOU that was watching?=
179 L1: =yes I was watching all these=
180 T: =but did this you have a body? . . .
181 L1: no I don’t and I remember my grandmother was
shouting please please come in come back like and
my I remember my body was holding down and my
grandmother was trying to rise me up?=
182 T: =yes=
Furthering the work of both Krashen and Long, Swain (1985,
2005) proposed the output hypothesis as an alternative way of looking
at the role of interaction in L2 acquisition. Under this view, language
acquisition through interaction will only occur when learners have an
opportunity to speak; our focus should therefore be more on learner
output and less on the interaction or input which they receive. Bygate
(1988) reinforces the notion that speaking may be more important
than comprehensible input:
It is only when the learner is being required to piece together his
own utterances that he is being obliged to work out – and hence
learn – his own plans of verbal action, all the while evaluating his
output in the light of his meaning intention.
According to Swain (1985: 248–249), output is important because
it forces the learner to develop precise, coherent and appropriate
linguistic resources, ‘pushed language use’, and because it requires the
learner to pay close attention to syntax and test hypotheses. The
concept of ‘pushed output’ is central to this position; opportunities to
interact in a classroom, in themselves, may not be adequate. Through
the teacher’s negotiation of meaning, learners are helped to reﬁne their

contributions so that they can be understood (see extract 3.4). In other
words, learners have to pay attention not only to the form of an
utterance, but also to its function and degree of appropriacy at a given
point in an exchange.
In her most recent version of the output hypothesis, Swain (2005)
maintains that output enhances fluency and promotes ‘noticing’ by
allowing learners to identify gaps between what they want to say and
what they are able to say. Her position is that much can be learned about
classroom communication by looking at the dialogues that unfold
between teachers and learners and between learners and learners.
Adopting a more social view of learning, Swain stresses the dialogic
nature of language learning, arguing that an understanding of learning
processes can be enhanced by understanding dialogues more fully.
The kind of noticing Swain describes (see also Schmidt 1993) is
exemplified below in extract 3.5, in which a group of upper-
intermediate learners is preparing for a listening activity from New
First Certificate Masterclass on the topic of lying. Here, the give and
take of the dialogue is more conducive to gaining an understanding
of what is actually happening than the isolated utterances of teacher
or learners. In addition, the extract highlights the need for learners
(or teachers) to identify gaps in their interlanguage. When the teacher’s
prompt in 99 receives no immediate response, she is able to feed in
the missing vocabulary, white lie, in 101. By considering classroom
discourse as a form of dialogue, understandings about teaching and
learning processes can be greatly enhanced, and gaps in learners’
interlanguage identified.
Extract 3.5
94 T: =like for example what kind of situation might have been
necessary (4)
95 L: when you don’t want to explain what you have done and
you don’t want to have troubles
96 T: (2) right=
97 L: =to say nothing [and]
98 L: [when you] want to be polite[((1))]
99 T: [when] you want to be polite yes if you want to be
polite or if you don’t want to hurt somebody’s feelings
sometimes what do we call . . . we saw it last week a
little lie it’s not serious=
100 L: =oh yes . . .
101 T: =white lie good a little white lie so that may be to protect
somebody or to . . .
102 LL: =/white lie/=
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Interactive classrooms, where learners are engaged in task-based
learning, certainly promote learner independence. Yet there are
frequently problems with tasks that have little or no teacher super-
vision: students may not take the task seriously, make extensive use
of L1, withdraw from the task, dominate the discussion, or perform
poorly. Under these conditions, there is certainly a strong argument
that the teacher must play a more prominent role in the ensuing
dialogue.
An example of the way in which a teacher might play a more
prominent role without totally dominating the discourse is offered in
extract 3.6 below. Note how learner 3 holds the floor and controls
the topic of discussion. Apart from requests for clarification and
confirmation checks on the part of the teacher and other learners (in
39, 41, 45, 47), L3 has control of the topic and is able to produce
relatively long, coherent turns, culminating in the final, extended turn
in 50. This student succeeds in communicating her message – aided,
perhaps, by the fact that she has complete control of the topic.
Arguably, the opportunities for practice and for maximising SLA are
as great – if not greater – in this type of open-class discussion, where
the teacher plays a scaffolding role, than in more ‘private’ pair- and
group-work activities.
Extract 3.6
36 L3: =I see . . . in my city . . . one woman she has a baby =
37 T: =yes=
38 L3: =and she didn’t have any money for that and she stole
erm a tin of milk=
39 T: =a tin of=
40 L3: =of milk=
41 T: =milk=
42 L3: =milk for the children=
43 T: =yes
44 L: ((2))
45 T: she went to prison?
46 L3: she went to prison but at the same time time . . . how
do you say . . . the ((6))?
47 T: =sorry could you [repeat]?
48 LL: [prefect]
49 T: =the prefect [yes] the prefect
50 L3: [yes] he stole the money the city but no no problems no
problems but erm everybody knows he stole the
money and erm he’s got the money one day later he
says tells the I don’t want ((1)) I don’t want to go on
as a prefect (unintelligible)

The research that has been conducted on social interaction in the
formal, L2 context indicates a need among language teachers to
develop a less prominent, yet still inﬂuential, position (see, for example,
Walsh 2003). From this new position, rather than ‘taking a back seat’
or ‘handing over’ to learners, teachers would be empowered to
facilitate, monitor and evaluate student contributions, while paying
closer attention to the ways in which their (teachers’) language
contributes to the language learning process. From this position,
constant reﬁnements can be made to the quantity, quality and function
of teacher input in relation to desired learning outcomes.
This necessarily brief overview of some of the main theories that
have been proposed in order to explain the complex relationship
between interaction and language learning both underline its
importance and highlight an overall lack of agreement about its precise
function. It does seem clear that interaction contributes in some way
to language learning, but we still do not really know precisely how.
Ellis (1990) posits three roles for classroom interaction:
1 Strong: interaction in some way determines SLA.
2 Weak: interaction facilitates SLA.
3 Zero: interaction is not necessary for learning to take place.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Task 3.2
What do you think? Look at the extract below and consider the role
of interaction in L2 learning. To what extent do you think that learning
opportunities are created here? Do you think, based on what you have
read so far, that there is a relatively strong or a rather weak relationship
between interaction and learning?
273 T: what about in Spain if you park your car illegally?
274 L4: . . . there are two possibilities
275 T: two [possibilities]
276 L4: [one] is er I park my car ((1)) and
277 T: yes . . . if I park . . . my car . . . illegally again Rosa
278 L4: (laughter) if I park my car [illegally]
279 T: [illegally]
280 L4: police stat policeman er give me give me
281 T: GIVES me
282 L4: gives me? a little small paper if er I can’t pay the
money
TC

In this section, the discussion has centred on the role of interaction
in learning by considering three hypotheses: input, interaction and
output. The position taken is that while interaction is central to the
L2 teaching/learning process, the interactive processes that make up
classroom discourse is at present not adequately understood by either
teachers or learners. The discussion now turns to a view of learning
as a social, rather than a cognitive, process.
Learning as a social process
In the last fifteen or twenty years, there have been a number of
significant developments to challenge traditional and long-standing
views of both the nature of language and the nature of learning.
Perhaps the starting point for these developments was the seminal paper
by Firth and Wagner (1997 and revisited in 2007), which challenged
existing conceptualisations of learning. The Firth and Wagner paper
argues that learning should be seen as a social process and that
language should be viewed as a complex, dynamic system that is
locally managed by interactants in response to emerging communicative
needs. Under this view, learning can be traced in the moment-by-
moment co-construction of meanings and by using conversation
analysis (CA). The relatively new, emergent field known as CA-for-SLA
quite clearly views learning as participation (see above) and maintains
that we can measure and track learning through the interactions that
take place (Markee 2008).
This said, social views of learning have actually existed for much
longer. Socio-cultural theories of learning emphasise its social nature;
learners interact with the ‘expert’ adult teacher ‘in a context of social
interactions leading to understanding’ (Röhler and Cantlon 1996: 2).
Learning, under this perspective, entails dialogue, discussion and
debate as learners collectively and actively construct their own
understandings through interactions with others who may be more
experienced. This view of learning owes its origins to the influential
work of the Russian philosopher Lev Vygotsky (1978), whose theories
283 T: it’s called a FINE remember a FINE yes?
284 L4: or if if my car
285 T: is parked
286 L4: is parked illegally . . . the policeman take my car and
. . . er . . . go to the police station not police station
it’s a big place where where they have some [cars]
they

have been applied to language learning contexts by researchers such
as Lantolf (2000) and Lantolf and Thorne (2006).
By emphasising the social, dynamic and collaborative dimensions
of learning, proponents of socio-cultural theories of learning stress its
‘transactional’ nature: learning occurs in the first instance through
interaction with others, who are more experienced and in a position
to guide and support the actions of the novice. During this part of the
process, language is used as a ‘symbolic tool’ to clarify and make sense
of new knowledge, with learners relying heavily on discussions with
the ‘expert knower’. As new ideas and knowledge are internalised,
learners use language to comment on what they have learnt; spoken
interactions are used to both transmit and clarify new information and
then to reflect on and rationalise what has been learnt.
Learning and the Zone of Proximal Development (ZPD)
According to Lantolf (2000: 17), the ZPD should be regarded as ‘a
metaphor for observing and understanding how mediated means are
appropriated and internalised’. Lantolf goes on to offer his own
definition of the ZPD: ‘The collaborative construction of opportunities
[. . .] for individuals to develop their mental abilities’ (ibid.).
A number of key terms emerge from the work of Vygotsky and
Lantolf, including ‘collaboration’, ‘construction’, ‘opportunities’ and
‘development’. Other writers use a similar terminology: van Lier (2000:
252), for example, refers to opportunities for learning as ‘affordances’,
while Swain and Lapkin (1998: 320) talk about ‘occasions for
learning’. Ohta (2001: 9) talks about learners’ ‘. . .level of potential
development as determined through language produced collaboratively
with a teacher or peer’. As a construct in the present context, the value
of the ZPD lies in its potential for enabling consideration of the ‘give
and take’ in the teaching/learning process. The ‘collaborative
construction’ of opportunities for learning is examined through the
ways in which teachers and learners collectively construct meaning in
L2 classroom interaction.
In extract 3.7 on p. 64, for example, a group of upper-intermediate
students are discussing ways of regulating lives. The extract has
been selected to illustrate the importance of collaborative meaning-
making and the need to allow interactional space so that teachers and
learners can create opportunities for language acquisition. Note that
this is a very different stance to the one that has been advocated under
CLT methodologies; simply ‘handing over’ to learners and getting them
to perform in pairs and groups will not, under socio-cultural theory,
contribute much to language learning. Instead, the teacher plays a
focal role, guiding, clarifying, supporting and shaping contributions
so that learners have opportunities to reflect on and learn from the
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

unfolding interaction. Here, for example, in 398, 400, 402, the teacher
paraphrases and summarises L1’s previous contributions as a means
of offering support and enabling other students to follow the dialogue.
While the dialogue is mainly between the teacher and L1, this strategy
of summarising and checking and negotiating is important if all
class members are to understand and contribute to the discussion.
The rapid turn-taking in the extract (indicated by =) identifies it as
being almost conversational in nature, with one big difference: the
teacher participates in the dialogue but, more importantly, ensures that
messages are understood and refined for the other listeners, the other
students.
Extract 3.7
398 T: =so it’s eh . . . I . . . from a sceptical point of view what
you have is a way of regulating your life =
399 L1: =yes=
400 T: =and eh giving you direction =
401 L1: =yeah=
402 T: =and goals and meaning [and]
403 L4: [so] do you think ((6)) ?. . .
404 L1: no I think that eh for example that my argument is that
if I take alcohol I’m culpable I get sick=
405 T: =everybody does=
406 L1: =but I think for me it’s like a sign stop doing this=
407 T: =me too ((3))=
408 L1: =or take take ((2)) for yourself and you won’t feel sick
you’ll be like high?=
409 T: =yes=
410 L1: like yes ((4)) . . .
411 T: so it’s good for you as far as being good?=
412 L1: [yes I think so]
A key concept under this view of learning is that of ‘scaffolding’,
used to refer to the linguistic support given by a tutor to a learner
(Bruner 1990). Central to the notion are the important polar concepts
of challenge and support. Learners are led to an understanding of a
task by, on the one hand, a teacher’s provision of appropriate amounts
of challenge to maintain interest and involvement, and, on the other,
support to ensure understanding. Support typically involves segmenta-
tion and ritualisation so that learners have, in the first instance,
limited choice in how they go about a task that is broken down into
manageable component parts (Bruner, 1990: 29). Once a task has been
mastered, scaffolds are removed and the learner is left to reflect and
comment on the task.

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Clearly, the amount of scaffolded support given will depend very
much on the perceived evaluation by the ‘expert’ of what is needed
by the ‘novice’. In a classroom context, where so much is happening
at once, such fine judgments can be difficult to make. Deciding to inter-
vene or withdraw in the moment by moment construction of classroom
interaction requires great sensitivity and awareness on the part of the
teacher and inevitably teachers do not ‘get it right’ every time.
Consider how the word toe is scaffolded in the extract below
(in 261). A group of pre-intermediate learners is working with oral
fluency practice materials, but is unable to complete one task because
they do not know the word toe. The teacher’s demonstration (in 259),
modelling (in 261), reinforcement (in 263 and 265), and definition
(in 267) all serve to ensure that this piece of vocabulary is introduced,
used and remembered by learners.
Extract 3.8
259 T: =he dislocated his shoulder I don’t think he did I think
he did this come here come here come here I think
he did this (teacher stands on student’s foot)
260 L: AEERGH thank you my ((2)) (laughter)
261 T: the bad news is he what did he do? what did he do?
he (4) to step on someone’s toe . . .
262 L2: to step on someone’s =
263 T: =toe=
264 L2: =toe? this one up toe yes?=
265 T: (writes on board) he asked her to dance the bad news
is he stepped on her toes . . .
266 L2: this one toes=
267 T: =toes like fingers but on your feet=
268 L2: =ah (3)
There are a number of reasons why a social approach to learning
is attractive. In the first instance, the emphasis on the importance of
social interaction is in tune with what is currently considered to be
‘good practice’ in ELT methodology: an emphasis on discovery-based
learning through problem-solving; the use of task-based instruction
that emphasises ‘learning by doing’ (see Chapter 2); the centrality of
pair- and group-work not only to maximising interaction, but increas-
ingly to co-operative learning (Ng and Lee 1996) are all important features
of the contemporary EFL classroom. According to van Lier (1996),
social development can only become language acquisition when the
quality of the interaction is maximised. Collaboration with the teacher,
less able learners, more able learners and the individual’s own resources
can facilitate interaction that is both meaningful and productive.

Arguably, the quality of that interaction is very much dependent on
the teacher’s ability to manage complex interactional processes and
‘correctly’ interpret the learning environment.
Second, the process of ‘scaffolded instruction’ (Bruner 1983, 1990),
involves learners in taking risks; learning support is gradually
withdrawn as learners become more independent, solving problems
for themselves and gradually acquiring new knowledge and skills
about the L2 through a process of ‘dialogic inquiry’ (Wells 1999).
Central to the process is the support system offered by the tutor, the
extent to which scaffolds are left in place or withdrawn, the amount
of scaffolding given and the extent to which learners are made aware
of its value (Donato 1994).
Finally, under socio-cultural theories of learning, dialogue is crucial
in helping learners acquire new knowledge. For this to happen, learners
need interactional space and support to express their ideas or thoughts.
Opportunities for learning (i.e. language acquisition) are maximised
when new concepts and language can be both understood and
verbalised. However, the centrality of speech to learning has another,
more signiﬁcant dimension in that consciousness, considered by
Vygotsky as being central to learning, is developed through social
interaction. Learners become more aware, through participation in
social activity, of themselves as learners.
This section has presented a necessarily brief overview of the con-
tribution of socio-cultural theories of learning to an understanding of
second language acquisition in the formal context. The relevance
of Vygotskyan theories of learning stems from the importance attached
to the social, interpersonal dimension of learning, from the acknow-
ledgement that development can be assisted or ‘scaffolded’ and from
the mediating force of dialogue.
Summary
This chapter has reviewed a sample of the literature on class-based
second language acquisition. Starting with an explanation of the rela-
tionship between learning and interaction, the discussion then presented
a critical review of the place of interactionist theories of second
language acquisition. In the ﬁnal part of the chapter, a second strand
to the theoretical framework for language learning was presented in the
shape of socio-cultural theories of education and learning.
The main message of the chapter is that interaction in the second
language classroom is fundamental to language acquisition; that at
present, the interactional processes are only partially understood; that,
if interaction is to promote meaningful learning, it has to be mediated;
that the prime responsibility for creating interaction-centred learning
opportunities lies with the teacher.

4 Approaches to studying
classroom discourse
Introduction
This is the ﬁrst of three chapters that consider different approaches
to studying interaction in the L2 classroom. Any attempt to understand
classrooms must begin with description, as Kumaravadivelu observes:
What actually happens there [in the classroom] largely determines
the degree to which desired learning outcomes are realised. The task
of systematically observing, analyzing and understanding classroom
aims and events therefore becomes central to any serious educational
enterprise.
(Kumaravadivelu 1999: 454)
Essentially, any understanding of the ‘interactional architecture’
(Seedhouse 2004) of the second language classroom requires selection
and mastery of particular tools – it is these tools that are presented
here. In the sections that follow, a critique is presented of the principal
approaches that have been used to investigate classroom interaction:
interaction analysis, discourse analysis, and conversation analysis.
Chapter 5 presents alternative approaches to studying classroom
discourse, while Chapter 6 offers a framework that can be used by
researchers and practitioners to evaluate classroom interaction.
Any attempt to capture what really happens in classrooms usually
means making a recording, either audio or video, and then transcribing
that recording, either fully or partially. Before considering the principal
approaches that may be used to analyse classroom discourse, I will
discuss the issues involved in recording and transcribing classroom data.
Recording and transcribing classroom interaction
Classrooms are highly complex places where there is so much going
on at any one point in time that it would be very difﬁcult to capture
everything. Multiple interactions are the norm and multi-party talk
underpins every action, every activity, every moment. Not only are
there technical problems associated with recording what actually
happens, there are, more importantly, enormous issues associated
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

with transcribing spoken discourse as written text. In this section, I
briefly outline some of the difficulties associated with recording and
representing spoken interaction in classrooms.
Recording
The first decision that must be made is how to record classroom
interaction. Basically, there are four choices:
1 audio-recordings;
2 video-recordings;
3 observation;
4 narrative.
Audio recordings are, in many ways, the easiest means of capturing
spoken interaction in classrooms. Modern technology makes it very
easy to record the interaction, using digital recorders positioned
carefully around the room. More elaborate techniques might entail
the use of lapel microphones or multi-directional microphones that
are more sensitive and may result in higher quality recordings. The
main difficulty associated with audio-recordings is the presence of
background ‘noise’ – a constant presence that can make deciphering
very difficult. This can be overcome, to some extent, by ensuring that
recording devices are positioned in such a way that if something is
not clear on one recorder, it can be picked up on another one.
Video-recordings are a relatively straightforward means of recording
interaction in the classroom and have the added advantage of providing
a visual representation of what happened. To get the best recordings,
two cameras are usually needed: one at the front of the classroom
pointing to the back, one at the back pointing to the front. In this
way, it is possible to record all the interaction taking place, especially
if video cameras are supplemented by the use of digital recorders
positioned around the room. This system normally means that it is
possible to capture all actions and spoken interactions. The main
Approaches to studying classroom discourse68
Think of a recent class that you either taught or participated in as a student.
What can you remember about that class? What problems would you have in
recollecting ‘what happened’ in that class?
What are the main difficulties associated with ‘recording’ classroom
interaction? How would you deal with some of those difficulties?

disadvantage with the use of video is that it can be quite intrusive and
result in unusual or unexpected behaviour by students or teachers. The
other issue is how to transcribe gestures, body movements, actions,
etc. in the written representation of a lesson.
Observation is quick and easy to organise but may have other difﬁ-
culties. Observers may have to be trained in the use of an observation
schedule (see ‘interaction analysis’ below), and may not be able to
record every detail of the interaction. Observers may also be prone
to bias and record what they think they saw rather than what actually
happened. It is unlikely that two observers would record what they
saw in exactly the same way; there are then issues around the reliability
of the data and the extent to which the observation is a faithful record
of the class. One way of making observation more objective is to use
‘focused observation’, whereby the observer focuses on one detail in
the interaction such as teacher’s use of questions, oral feedback, learner
involvement, etc. This approach has the advantage of allowing the
observer to focus on only one element in the interaction and ignore
everything else.
A narrative approach to recording classroom observation entails
the use of an observer who writes a descriptive account of the lesson
as a narrative. The ‘story’ that is told of a particular lesson is then
used to make changes to practice or to investigate particular
phenomena in the interaction. Obviously, the main problem associated
with this approach is the difﬁculty of writing down every detail of
what occurs in a classroom. Even the best observers are sure to miss
key moments and cannot capture everything.
Whichever approach to recording is adopted, there are a number
of principles that should be followed:
• Ethical considerations. Written permission must be obtained from
all participants before any recording can begin. In the case of young
children, this may also involve getting the permission of parents.
In some contexts, the faces of students must not be visible in order
to protect their identity. Participants must be told what the
recording will be used for, how it will be used and how their
anonymity will be ensured. They must have the option of being
able to opt in or out of the recording and to leave at any time (see
BERA guidelines on research ethics by visiting: www.bera.ac.uk/
ethics-and-educational-research-2/).
• How much data? Depending on what the recordings are to be used
for, it is worth considering how much data will be needed. This
will depend both on the purpose of the recording and on the way
it is to be transcribed. It may, for example, be perfectly feasible to
use only a few hours of recording if the recordings are transcribed
in great detail (see below, transcription).
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Approaches to studying classroom discourse 69

• Sound quality. For audio and video-recordings, steps need to be
taken to ensure that the best quality recordings are made. There is
nothing more frustrating than working with recordings where only
part of the interaction can be heard. Things to think about include:
the choice of room, the use of carpet and curtains to help reduce
‘echo’, positioning of equipment, number of recorders, selective
recording where only some of the participants are included, and so
on.
• The role of the observer. The relation of the observer to the
participants needs to be given some thought: should they be known
or an ‘outsider’? In cases where the observer is also another teacher,
how might this affect the group dynamics? Where should the
observer sit and what part should they play in the lesson? What is
the potential for bias, or lack of objectivity?
Once recordings have been made, the next step is transcription.
Transcription
The main concern of transcription is to ‘represent reality’ as accurately
and faithfully as possible. There has been much discussion and
debate on this over past decades and there is still only partial agreement
as to the extent to which a written transcript can accurately repre-
sent a spoken encounter. In the same way that photography sets out
to provide a visual record of ‘reality’, a transcript offers a written
record of a spoken interaction. As we shall see, this is not without its
problems. Even photography has changed in the way it represents
reality: compare the photographs of thirty years ago with modern,
digitally enhanced photographs that provide a sharper, more detailed
representation of the same image. The same comparison may be said
of transcripts: a transcript may offer a rather bland summary of an
encounter or a highly detailed description that attempts to record every
detail.
Initial attempts to problematise transcription focused largely on the
technical side of representing in writing a spoken text. Clearly, there
are huge technical difficulties associated with actually hearing what
was said, eliminating background noise, including all the finer nuances
of spoken language, representing all of this in a written form. There
are technical difficulties too at the level of representation: do we opt
for broad or narrow transcripts, for example? A broad transcript
captures the essence of what was said, the words themselves or even
their intended meaning, but ignores the fine details such as a stressed
syllable, a pause, a rising intonation, overlapping speech. These tech-
nical difficulties remain as much a problem for transcribers as they
have always done.

However, and more importantly, there are factors to consider that
go beyond the purely technical side of transcription. Methodological
decisions made at the time a transcript is produced will greatly
influence our understandings of a particular discourse encounter such
as a second language class. The precise relationship between the
interaction that took place and the words and symbols used to
represent it is crucial and complex. That relationship involves the
transcriber making key decisions such as:
• Do I include all pauses and how do I represent them?
• How do I record particular gestures, facial expressions, body
movements, etc. – or do I simply ignore them?
• What is the ‘correct’ way to record emphatic speech? Do I even
need to record it?
• Do I organise the written text by turn at talk, as conversation
analysts do, or by linguistic utterance, showing breath groups and
intonation?
• Should intonation be included – if so how do I record it?
• Do I need to transcribe everything – or even anything?
The list goes on. The point I am making here is that important
decisions are taken at the time a transcript is produced: these decisions
will influence understandings of reality, our ability to interpret the data
and our potential to make changes to practice – if that is our intention.
By way of illustration, I am going to take two transcripts of the
same event and show how each offers a different interpretation of that
event. Both transcripts are taken from Bucholtz (2007) and relate to
a research project on nerdy American teenagers, conducted in 1998.
The transcripts below relate to an interview that the author conducted
with one teenager.
Extract 4.1
Fred: We’re always the nerds. We like it. We’re glad to be the nerds
and the squares. We don’t drink, we don’t do any drugs, we just get
naturally high, we do insane funny things. And we’re smart. We get
good grades.
In the first extract, the main details of the original interview have
been retained. As we read the extract, we get a good sense of Fred’s
feelings of being a nerd, what this means, why they like it and so on.
From an interviewer’s perspective, it seems to offer sufficient detail of
the interview, without burdening the reader with ‘unnecessary’ detail.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Essentially, extract 4.1 is a summary of the interview and is highly
typical of the kind of data used as ‘evidence’ in qualitative research
reporting interviews. While it certainly captures the essence of that
encounter, it omits all the detail of the interaction.
In contrast, consider extract 4.2 below, which offers a micro-
analytic rendition of the same event. Here, we see how the main ideas
represented in 4.1 were co-constructed by both speakers. Rather than
a triumphant boast about being a nerd, which is the impression given
in 4.1, the actual interaction demonstrates that the interviewer (Mary)
actually did quite a lot of work to elicit Fred’s intended meaning. The
details of this transcript reveal particular features that indicate that
this interaction was a joint enterprise and that meanings were
established together:
• Overlapping speech (in lines 1–2, 8–9, 16–17, etc., indicated [ ]).
• Acknowledgement tokens (in lines 22 and 27) that demonstrate
understanding by Mary.
• Pausing (in lines 20, 33 and 38) that may indicate interactional
trouble or pausing for thought.
• Latched turns (marked =) in lines 12–13, 30–31, 34–35), indicating
rapid speech or interruptions.
The overall impression we get as readers when we read extract 4.2
is that, rather than being boastful and triumphant, Fred is ‘feeling her
way’ in the discourse, expressing herself with some difﬁculty and
perhaps embarrassment. Mary, by contrast, prompts, guides and helps
the interlocutor to say what she really means – the hallmark of a good
interviewer.
Extract 4.2
1 Mary: [So ]
2 Fred: [We’re al]ways the nerds.
3 We like it.
4 Mary: You@’re the nerds?
5 Fred: We’re <creaky> {glad} to be the ner:ds,
6 a@nd the squa:res and,
7 Mary: Is that what
8 Fred: [we don’t– ]
9 Mary: [you say ] you are?
10 Fred: <[i?]> Well,
11 we don’t exactly s:–

12 We don’t always say it,=
13 =I say it. n@
14 Mary: @@[@! ]
15 Fred: [But- ]
16 Mary: @ You’re [[prou:d. ]]
17 Fred: [[you ]] know,
18 Mary: [@@ ]
19 Fred: [we don’t– ]
20 We just don’t (0.5) drink,
21 we d [on’t (.) <rapid> {d]o
22 Mary: [Mm. ]
23 Fred: any drugs,}
24 we don’t–
25 we just,
26 <smiling voice quality> {get ↑naturally high},
27 Mary: A[ha:. ]
28 Fred: [@: ]
29 Mary: [[So that makes you nerds?]]
30 Fred: [[We just do insane ]] funny things.=
31 =I don’t know,
32 maybe.
33 (0.6)
34 Mary: So:=
35 Fred: =And we’re smart.
36 We get <[e?]>-
37 good grades.
38 (1.3) Mostly.
As Bucholtz herself says (2007: 788):
It is obvious that my original transcript was not merely woefully
inadequate, but dangerously inaccurate in its representation of the
interaction. Fred’s comments are not the product of an auton-
omous, triumphant voice of nerd pride but are rather the result of
considerable co-construction (and obstruction) by me as the
researcher.
This illustration, I hope, serves to demonstrate the importance of
the decisions that are made at the transcription stage in terms of level
of detail and faithfulness of the written representation of the spoken
interaction. Consider how essential it is, then, in a classroom context,
to try to both capture and represent the interaction and to make a
case for the level of detail included in the transcript.
In the next section, I present the ﬁrst of three approaches to
analyzing spoken interaction.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Interaction analysis approaches
For many years, especially in the 1960s and 1970s, interaction analysis
(IA) was the most popular and widely used means of analysing classroom
interaction. According to researchers and practitioners at that time, one
of the most reliable, quantitative approaches to analysing interaction
was through the use of observation instruments, or coding systems,
to record what the observer thinks is happening at any given moment.
From these recordings and the ensuing statistical treatment, classroom
profiles can be established, which, it is argued, provide an objective and
‘scientific’ analysis of the interaction. According to Brown and Rodgers
Task 4.1
Whatever approach to transcription you adopt, you will need to
use some kind of transcription system. Consider the one below. How
appropriate is this for transcribing classroom discourse? Are any of
these features unnecessary? Would you include any other features
not mentioned here?
(Adapted from van Lier 1988; Johnson 1995)
Transcription system
T: - teacher
L: - learner (not identified)
L1, L2, etc.: - identified learner
LL: - several learners at once or the whole
class
S3, S4 - Student 3, 4, etc.
/ok/ok/ok/ - overlapping or simultaneous utterances
by more than one learner
[do you understand?]
[I see] - overlap between teacher and learner
= - turn continues, or one turn follows
another without any pause
. . . - pause of one second or less marked
by three periods
(4) - silence; length given in seconds
? - rising intonation – question or other
! - emphatic speech: falling intonation
((4)) - unintelligible 4 seconds a stretch of
unintelligible speech with the length
given in seconds
Paul, Peter, Mary - capitals are only used for proper nouns
T organises groups - editor’s comments (in bold type)
TC

(2002), over 200 different observation instruments now exist, while
Chaudron (1988) calculated that there were approximately twenty-six
systems available for analysing interaction in the L2 classroom.
The observation schedules used in IA have a number of features in
common:
• They use some system of ticking boxes, making marks, recording
what the observer sees, often at regular time intervals.
• They are considered reliable, enabling ease of comparison between
observers and generalisability of results (but see ‘limitations’ on
page 77).
• They make assumptions about the ways in which classroom discourse
progresses, often in a linear fashion that can be easily recorded.
• They are used extensively in teacher training, particularly for
developing competencies and raising awareness.
A review of the many different instruments that are now available
is beyond the scope of this chapter. Following Wallace (1998), in the
brief summary that follows, observation instruments are divided
according to whether they are system-based or ad hoc.
System-based approaches
By ‘system’ is meant that the instrument has a number of fixed
categories that have been pre-determined by extensive trialling in
different classroom contexts. There are several advantages to using a
fixed system: the system is ready-made – there is no need to design
one from scratch; because the system is well-known, there is no need
for validation; any system may be used in real-time or following a
recording; comparisons between one system and another are possible.
A brief summary of some of the better known schedules is now
presented, beginning with one of the earliest systems put forward by
Bellack et al. in 1966. Based on the interaction of fifteen teachers and
345 students, the instrument importantly identified a number of
pedagogical moves that could be categorised into common teaching
cycles. For example, the moves structure, solicit, respond and react
frequently occur together:
Extract 4.3
T: We’re going to look today at ways to improve
your writing STRUCTURE
T: Would you like to tell me one of the
mistakes that you made? SOLICIT
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

S: The type of the verb RESPOND
T: The verb, it means there’s a problem with
the verb REACT
(Walsh 2001)
Bellack et al.’s three-part exchange: solicit, respond, react – or as it
is now more commonly described: initiation, response, feedback (IR(F))
– is, even today, regarded as the very fabric of classroom interaction by
most practitioners (see Chapter 1). As such, it represents a significant
contribution to our understanding of the processes of classroom
interaction and reveals much about the ways in which teachers and
students communicate. Some researchers and practitioners argue that
an IR(F) exchange structure may indicate an overly teacher-centred
classroom, although this criticism has been countered by others. Kasper
(2001), for example, argues that learners can be more actively involved
in teacher-fronted classroom interaction when teachers offer more
participation rights in the conversation. Her work suggests ways in
which teachers can facilitate learner involvement in the interaction.
The work of Bellack et al. was further advanced by Flanders (1970),
whose FIAC system (Flanders Interaction Analysis Categories) assigned
classroom interaction to various categories of teacher and student talk:
Teacher talk
1 Accepts feelings.
2 Praises or encourages.
3 Accepts or uses ideas of pupils.
4 Asks questions.
5 Lectures.
6 Gives direction.
7 Criticises or uses authority.
Pupil talk
8 Pupil talk: response.
9 Pupil talk: initiation.
Silence
10 Period of silence or confusion.
Although this instrument does attempt to offer a finer grained
classification of classroom discourse, the categories are still rather broad

and it is questionable whether the instrument could adequately account
for the complex interactional organisation of the contemporary
classroom – content or second language – where teacher and learner
roles are, arguably, more equal and where student–student interaction
is commonplace. Again, this instrument also works on the assumption
that all classroom interaction progresses in a ‘neat and tidy’ manner –
this is clearly not the case.
In later years, instruments became more complex and included
more and more categories designed to capture the complexities of
classroom interaction. The COLT system (Communicative Orientation
to Language Teaching), for example, was proposed by Allen et al. in
1984 and comprised seventy-three categories. The main aim of COLT
was to enable the observer to make a connection between teaching
methodology and language use. The instrument is directly linked to
communicative methodology and considers how instructional differ-
ences impact on learning outcomes. It was devised in two parts.
Part A focuses on classroom organisation, tasks, materials and levels
of learner involvement, while Part B analyses learner and teacher
verbal interaction, considering such things as evidence of an informa-
tion gap, the existence of sustained speech, the quantity of display
versus referential questions. The COLT instrument is certainly one of
the most sophisticated devised to date vis-a-vis L2 classrooms and
makes use of a considerable range of both qualitative and quantitative
modes of analysis. A revised version of COLT was presented in 1995
by Spada and Frohlich.
While the most recent version of this instrument is certainly more
sophisticated and attempts to capture many of the features of class-
room interaction, it has its limitations, as the authors recognise:
If one is interested in undertaking a detailed discourse analysis
of the conversational interactions between teachers and students,
another method of coding and analyzing classroom data would be
more appropriate.
(Spada and Frohlich 1995: 10)
From this necessarily brief overview of three of the main instruments
used in system-based approaches to IA, it is fair to say that the
approach has a number of serious limitations:
• Patterns of interaction have to be matched to the categories pro-
vided; the results are, therefore, pre-determined. How do observers
deal with events or interactions that do not ‘ﬁt’ the categories?
• Coding systems emphasise the observer’s interpretation of events,
often at the exclusion of that of the participants. The perspective
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

is therefore etic (from the outside), rather than emic (from the
inside).
• No allowance is made for overlap; the categories for observation
are discrete and there is an underlying assumption that classroom
discourse proceeds in a sequential manner (T -> Ss -> T -> Ss and
so on). In fact, this is simply not the case: overlaps, interruptions,
back-channels, false starts, repetitions, hesitations are as common
in language classrooms as they are in naturally occurring conversa-
tion (Edwards and Westgate 1994).
• Observers may fail to agree on what they see: two observers may
‘see’ the same class in different ways and produce two different
accounts of the same interactions. This may have serious
consequences for validity and reliability.
• Coding systems fail to take adequate account of context. All
varieties of interaction are evaluated in the same way and from the
same perspective (Seedhouse 1996).
Task 4.2
Go back to the description of FIAC on page 76. Use the ten categories
given in this instrument to analyse the extract below. What problems
did you encounter? In what ways might this instrument be suitable
or unsuitable for L2 classrooms?
12 L: (2) erm I am agree with=
13 T: =agree be careful with the verb to agree there you
as well Ensa that it’s we! agree it’s not to
14 be agree it’s to agree! Ok=
15 L: [oh I agree]
15 L: ((3))
is to agree ok so ((3)) to agree with
17 (writing on board) is the preposition that follows it
I so it’s I agree with you I disagree with
18 you . . . ok em Silvie can you em what were you going
to say?
19 L2: I agree with you because em when when we talk
about something em for example you
20 saw a ((2)) on TV=
TC

As we have seen, it is unlikely that highly structured observation
instruments, such as the ones outlined in this section, can adequately
account for the complexities of most classrooms. In order to explain
the features of a particular context, it seems likely that less structured,
yet tailor-made instruments are more suitable. While still following a
broadly IA approach, these ad hoc observation instruments (Wallace
1998) are more sensitive to context and are normally designed with
a specific goal in mind. As such, they offer descriptions that are likely
to be both more accurate and more realistic.
‘Ad hoc’ approaches
In contrast to system-based interaction analysis, ad hoc approaches
offer the construction of a more flexible instrument, which may, for
example, be based on a specific classroom problem or area of interest.
The instrument may be designed as part of an action research project
in which practitioners wish to answer a particular question such as:
how can I improve the quality of the feedback I give my students?
Alternatively, the instrument might be designed through consultation
with either teachers or students working in a specific context.
Consider the following example, an extract from Walsh 2006 (see
Chapter 6):
F. Seeking clarification 1. Teacher asks a student to clarify
something the student has said.
2. Student asks teacher to clarify
something the teacher has said.
This instrument was used to help teachers gain closer under-
standings of their teacher talk. This is one category, from a total of
fourteen used in the original, focusing on the ways in which teachers
seek clarification (see Appendix A for the original).
Now look at the extract below and the commentary that follows.
In what ways does this instrument help us, as observers, to gain closer
understandings of the interaction?
Extract 4.4
121 T: =yes so tell me again what you mean by that?=
122 L:: =the first is the introduction the second eh in this case
we have the ((3)) who you are to eh
123 introduce yourself a few words about yourself and where
you live and what I do [and]
124 T: [so ] . . . yes?=
125 L: =and then it’s the problem what happened . . .
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

126 T: yes=
127 L: =and you need to explain it and why you are writing
because probably you did something like
128 you gave the information to the police but it didn’t
happen . . .
129 T: right=
130 L: =which is why it’s ((2))=
131 T: =so can I ask you why did you write it in your head as
you said?=
132 L: =I don’t know it’s like a rule=
133 T: =right so it’s like a rule what do you mean? . . .
In this extract, we can see several examples of the teacher seeking
clarification (in 121, 124, 131 and 133). Throughout the extract, the
teacher’s prompts stimulate the learner to offer an explanation: ‘it’s
like a rule’ (line 132). Although the actual explanation is not forth-
coming, even by the end of the extract, arguably, the teacher’s requests
for clarification do serve the purpose of ‘pushing’ this learner to say
more and offer an explanation. The teacher’s use of backchannels or
acknowledgement tokens (‘yes’ in line 126 and ‘right’ in line 129) also
tell the student that she has been understood, that she still has the
floor and that she should continue with her explanation.
The main advantage of this approach to IA is that it allows us, as
observers, to focus on specific details in the interaction that we can then
describe and attempt to explain. The whole process is much more from
the inside looking out and less from the outside looking in. In other
words, an ad hoc system is more likely to promote understanding and
generate explanations than the system-based approaches discussed
above. The fact that the instruments are designed in relation to a partic-
ular question or problem within a specific context makes the whole
research process more meaningful and realistic. Perhaps most important
of all is the fact that we are more likely to have confidence in the data.
We can summarise the main advantages of ad hoc approaches to
interaction analysis as follows:
• Ad hoc approaches to classroom observation give participants
ownership of the research design process and greater insights into
the issues under investigation.
• By focusing on the detail of the interaction, such approaches allow
practitioners to access and understand complex phenomena that
might otherwise take years of class experience to acquire.
• Ad hoc systems enable observers to focus on the microcosms of
interactions that might so easily be missed by the ‘broad brush’
descriptions provided by systems-based approaches.

Having given a brief critique of interaction analysis, I now turn to
discourse analysis approaches to analysing classroom interaction.
Discourse analysis approaches
Discourse analysis is the study of spoken or written texts. Its focus is
on words and utterances above the level of sentence and its main aim
is to look at the ways in which words and phrases function in context.
For example, the frequently used ‘Could you turn to page 36?’ might
be interpreted as a request under DA. Other utterances may be less
easy to classify. The phrase ‘the window’s open’, when uttered in a
classroom, might function as a request (please close it), an explanation
(that’s why it’s so cold), a drill (everyone repeat), a deﬁnition (as a way
of showing the meaning of open). According to Seedhouse (2004: 56),
‘the overwhelming majority of previous approaches to L2 classroom
interaction have implicitly or explicitly adopted what is fundamentally
a discourse analysis approach.’
Perhaps the earliest and most well-known proponents of DA are
Sinclair and Coulthard (1975) who, following a structural-functional
linguistic route to analysis, compiled a list of twenty-two speech acts
representing the verbal behaviours of both teachers and students
participating in primary classroom communication. The outcome is
the development of a descriptive system incorporating a discourse
hierarchy:
• LESSON
• TRANSACTION
• EXCHANGE
• MOVE
• ACT
Act is therefore the smallest discourse unit, while lesson is the
largest; acts are described in terms of their discourse function, as in
the two examples of speech acts below, evaluation and cue:
Extract 4.5
Act Function Realisation
Evaluation evaluates right so it’s like a rule what do you
mean?
Cue evokes bid yes so tell me again what you mean
by that
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Perhaps the most important contribution of Sinclair and Coulthard
to our understandings of classroom discourse is their realisation that
most classroom discourse follows an IR(F) structure. For every move
made by a student, teachers typically make two, as in:
I T: what’s the past tense of go?
F S: went
F/E T: went, excellent.
Thus, we can say that most teaching exchanges follow an IRF
structure, made up of three moves, with each move containing one or
more speech acts. Not only does this help to confirm the fact that
teachers do speak much more than students in most classrooms, it
also demonstrates how teachers often control the discourse. (For a full
discussion of the IRF exchange, please see Chapter 2.)
The main difficulty associated with allocating utterances to functions
is that any utterance can perform a range of functions (see the above
example). In a multi-party setting such as a classroom, where there
are so many things going on at the same time, deciding on a linguistic
function may be extremely problematic. According to Stubbs (1983),
it is almost impossible to say precisely what function is being performed
by a teacher (or learner) at any point in a lesson. As with most
functional analyses, inside or outside a classroom, an utterance can
have any number of functions depending on crucial contextual clues
such as who said it, to whom, how they said it, why they said it and
so on. Classification of classroom discourse in purely structural-
functional terms is consequently problematic. Levinson (1983) takes
up the same argument, stressing the fact that any one utterance can
perform a multitude of functions, especially in a classroom setting
where interaction patterns are so complex. The consequence of this is
that there is no way, under speech act theory, to account for gestures
or behavioural traits.
Task 4.3
Look at the extract below and analyse it according to language
functions. What difficulties did you encounter in deciding on the
function of each utterance? How well does IRF account for the overall
structure of this interaction?
TC

One of the main limitations of the Sinclair and Coulthard system
is that it was derived from data recorded in ‘traditional’ primary
school classrooms during the 1960s. In this context, and by the
authors’ own admission, it was possible to identify very clear status
and power relations between teachers and learners. Consequently, the
ensuing discourse had a very clear structure that was largely dominated
by question and answer routines. In the contemporary L2 classroom,
where there is, arguably, far more equality and partnership in the
teaching–learning process, it is doubtful whether the framework could
adequately describe the structure of classroom communication.
As Wu (1998: 529–530) says:
the cumulative effect of teacher–student interaction cannot be
accounted for within the [Sinclair and Coulthard 1975] framework.
It is clear that such hierarchical categorisation, though shedding
some light on the mechanism of teacher–student verbal exchanges,
is not enough to demonstrate its entire dynamic process.
To summarise this section, we can say that DA approaches are both
descriptive and prescriptive and attempt to categorise naturally
occurring patterns of interaction and account for them by reference to
a discourse hierarchy. The starting point is structural-functional
linguistics: classroom data are analysed according to their structural
patterning and function. For example, the interrogative structure
‘What time does this lesson end?’ could be interpreted as a request
for information, an admonishment, a prompt or cue. Any attempt to
analyse classroom data using a DA approach, therefore, involves some
simpliﬁcation and reduction. Matching utterances to categories may
be problematic owing to the issues of multi-functionality and the
absence of a direct relationship between form and function. In general,
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Extract 4.6
1 L: no that’s bad news=
2 L: =so it’s good news (laughter)
3 LL: /bad news/ ok / no no that’s good news/ . . .
4 L2: bad news . . .
6 L3: =ah good good news (2)
7 L1: no no that’s wrong you have to do bad news . . .
8 L2: yes it’s a bad news because [you]
9 L: [no but that’s] good news=
10 T: =that’s good news G N good news . . .

DA approaches fail to take account of the more subtle forces at work
such as role relations, context and sociolinguistic norms that have to
be obeyed. In short, a DA treatment fails to adequately account for the
dynamic nature of classroom interaction and the fact that it is socially
constructed by its participants. By the same token, DA approaches do
not adequately account for the range of contexts in operation in a lesson
and for the link between pedagogic purpose and language use.
In the ﬁnal section of this chapter, a critique is presented of a more
naturalistic approach to classroom discourse: conversation analysis.
Conversation analysis approaches
The origins of conversation analysis (CA) come from an interest in
the function of language as a means for social interaction (Sacks,
Schegloff and Jefferson 1974). CA is based on the premise that social
contexts are not static but are constantly being formed by the
participants through their use of language and the ways in which turn-
taking, openings and closures, sequencing of acts, and so on are locally
managed. Interaction is examined in relation to meaning and context;
the way in which actions are sequenced is central to the process.
In the words of Heritage:
In fact, CA embodies a theory which argues that sequences of
actions are a major part of what we mean by context, that the
meaning of an action is heavily shaped by the sequence of previous
actions from which it emerges, and that social context is a
dynamically created thing that is expressed in and through the
sequential organisation of interaction.
(1997: 162)
According to this view, interaction is context-shaped and context-
renewing; that is, one contribution is dependent on a previous one
and subsequent contributions create a new context for later actions.
Under this microscopic view of context, one person’s contribution is
inextricably linked to that of another person. Order in spoken
discourse is established through sequential organisation: the way in
which one utterance is connected to another.
To see how this works in a classroom, consider extract 4.7 below.
In the extract, every turn is uniquely linked to the previous and
following ones and there is a clear sense that this represents a coherent
piece of discourse that could have only occurred in a classroom
context. Turns 187–199 make up a piece of spoken text that is coherent
and logical because of the ways in which one turn is interdependent
on another. The whole extract is a good example of a piece of discourse
that is co-constructed by participants and that is shaped by a goal-
oriented activity: here, eliciting the word force.

Extract 4.7
187 T: =what do we call I’m going to try and get the class
to tell you what this word
188 is that you’re looking for . . . er we talk about military
(claps hands) . . . military what?
189 L: ((1))=
191 L: =kill=
192 T: =no not [kill]
193 L: [action] action=
195 LL: =power=
197 LL: ((3))force=
200 T: =that guides our lives=
Although the origins of CA lie in ordinary spoken interaction, its
relevance to institutional discourse – such as a classroom – cannot be
ignored. Here, the goals and actions of participants are closely linked
to, and to some extent constrained by, the institutional business. In a
language classroom, for example, most interactions are related to the
enterprise of learning a second language; turn and topic management,
sequential organisation and choice of lexis are all determined by that
enterprise and by the roles of interactants. Essentially, an institutional
discourse CA methodology takes as its starting point the centrality of
talk to many work tasks: quite simply, the majority of work-related
tasks are completed through what is essentially conversation, or ‘talk-
in-interaction’ (Drew and Heritage 1992: 3). In many work contexts,
we find an expert and non-specialist: doctor–patient, counsellor–client,
teacher–student. The interactions that take place in these contexts are
largely shaped by the roles performed by the interactants and by the
business of the moment. It comes as no surprise therefore that teachers
typically ask questions and students (usually) answer them. What CA
can do is to uncover something of the detail of these interactions by
looking at the ways in which contexts are co-created in relation to the
goal-oriented activity in which they are engaged (Heritage 1997: 163).
All institutions have an overriding goal or purpose that constrains both
the actions and interactional contributions of the participants according
to the business in hand, giving each institution a unique interactional
‘fingerprint’ (Heritage and Greatbatch 1991: 95–6). Thus, the inter-
actional patterning (or ‘fingerprint’) that is typical of, for example, a
travel agent’s will be different from that of a dentist’s, or newsagent’s.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

While the discourse of L2 classrooms does not and should not be
interpreted as fully resembling a conversation, there are some strong
parallels. Essentially, what takes place in an L2 classroom between
teachers and learners and learners and their peers can be described as
‘conversation’. It is, for the most part, two-way; it entails turn-taking,
turn-passing, turn-ceding and turn-seizing; it makes use of topic switches
and contains many of the features of ‘ordinary’ conversation such as false
starts, hesitations, errors, silence, back channelling and so on.
In the words of Edwards and Westgate (1994: 116):
The point is not that classroom talk ‘should’ resemble conversation,
since most of the time for practical purposes it cannot, but that
institutionalised talk [. . .] shows a heightened use of procedures
which have their ‘base’ in ordinary conversation and are more
clearly understood through comparison with it.
The relevance of a CA approach to the L2 classroom context is not
difficult to perceive. CA attempts to account for the practices at work
that enable participants in a conversation to make sense of the interaction
and contribute to it. There are clear parallels: classroom talk is made
up of many participants, and there have to be smooth transitions and
clearly defined expectations if meanings are to be made explicit. Possibly
the most significant role of CA is to interpret from the data rather than
impose pre-determined structural or functional categories.
For many researchers and practitioners working with classroom
data, CA offers a way of answering questions that are related to
teaching and learning. Rather than studying features of the interaction
per se – what ten Have (2007) calls ‘pure’ CA – the research interest
may be more closely linked to educational practices. This ‘applied’
form of CA may be used to provide insights that may inform the
actions of practitioners. For example, a CA analysis may reveal how
teachers create opportunities for learning in the interaction, or, indeed,
restrict them. As Richards (2006: 5) points out, CA can perform an
‘enabling rather than an enacting role’ in providing insights for
professional practice. Applied CA will seek to ‘translate’ the resulting
What interactional features make up the fingerprint of an L2 classroom?
Examples include questions and responses to questions.
How many others can you think of? How do these features compare to
those found in other social contexts, such as meetings, chat shows, job
interviews?

descriptions into terms relating to substantive issues in the relevant
field, such as ways of improving classroom practices.
The main features of a CA approach to analysing L2 classroom
interaction are summarised:
• The aim of CA is to account for the structural organisation of the
interaction as determined by the participants. There should be no
attempt to ‘fit’ the data to preconceived categories.
• The approach is strictly empirical. CA forces the researcher to focus
on the interaction patterns emerging from the data, rather than
relying on any preconceived notions which language practitioners
may bring to the data (Seedhouse 2004).
• The observer is seen as a ‘member’ of the interaction, trying to view
the experience through the eyes of the participants. The aim is to
offer an emic (insider) perspective.
• Context is dynamic and mutually constructed by the participants.
Contexts are therefore constantly changing as a lesson progresses
and according to local demands and constraints.
• Talk is seen, in an institutional setting, as being goal-oriented:
participants are striving towards some overall objective related to
the institution. In a language classroom, for example, the discourse
is influenced by the fact that all participants are focusing on some
pre-determined aim [SM1]– learning a second language.
• The analysis of the data is multi-layered. CA approaches emphasise
both context and the sequentiality of utterances. Because no one
utterance is categorised in isolation and because contributions are
examined in sequence, a CA methodology is much better-equipped
to interpret and account for the multi-layered structure of classroom
interaction.
An example may help to make some of these ideas clearer. Extract
4.8 below is taken from a class of adult learners who are preparing
to read a text on euthanasia. The teacher’s stated pedagogic goal for
this segment of the lesson was to improve oral fluency and there follows
a discussion on ways of prolonging life. In the extract, by looking at
longer extracts of data and using a CA approach, a number of
observations can be made.
In the extract, we can see that turn-taking between teacher and
students is fairly even – indeed, students have more interactional space
than the teacher, who only intervenes as and when necessary (in 115,
117 and 119). Turns follow one another quickly and there is little
pausing, as indicated by (=), showing latching: one turn follows another
without any pause. This may be significant since silence and pausing
are often signs of interactional trouble; there is, then, a sense that the
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

discourse flows freely and that participants have equal status and rights.
Participants make frequent use of acknowledgement tokens such as ugh
(line 111), yeah (113), oh (117), which serve the important function of
‘oiling the wheels of the interaction’ (McCarthy 2003) and letting
everyone know that they have been understood. There is no repair at
all in this extract, and although errors do occur, they do not impede
communication at all and are simply ignored. There is evidence too that
meanings are being negotiated (in 109–110, 112–113, 114–115,
116–117) – an indicator that something is being communicated.
Extract 4.8
109 T: [no ok alright] . . . so Jan you want to live forever?=
110 L3: =yeah if money can afford it I will freeze body =
111 L: =ugh . . .
112 L1: what are you going to do? . . . frozen frozen you body?=
113 L3: =yeah=
114 L1: =cryonics? . . .
115 T: =yeah it’s cry cry cryo[genics]
116 L1: [cryonics] cryogenics . . . no cryonics=
117 T: =oh is it? ok=
118 L1: =I think so I don’t know . . .
119 T: let me check it it might be in this one . . . (looks in
dictionary)
Even at a very basic level and focusing on a limited amount of data
(11 turns in extract 4.8), our understanding of ‘what happened’ can
be greatly enhanced under a CA methodology. Rather than allocating
utterances to preconceived categories, there is an attempt to ‘let the
data speak for themselves’, indicating how understandings are derived
in the goal-oriented activity of the moment. Minute adjustments in
the discourse can often be explained in relation to the moment-by-
moment co-construction of meaning as participants attend to some
pre-conceived agenda.
However, it cannot be denied that CA approaches do have a number
of limitations. These include:
• Selectivity. Some researchers criticise CA for being over selective.
Snatches of discourse and their ensuing commentaries may appear
to have been selected randomly with no attempt to evaluate their
significance to the discourse as a whole. Any selection of data may
appear contrived or idealised in order to illustrate a particular point
with little attempt to relate them to the exchange as a whole.
• Inability to generalise. A more serious criticism of CA approaches
is their inability to generalise findings because they focus on one

specific and very narrow context. While this may be true of many
qualitative research tools, it is particularly applicable to a CA
methodology owing to the centrality of context. That is not to say
that context-specific data are not valid or worthwhile; merely that
they cannot be extended to other contexts. This objection is
countered, however, if we acknowledge that the aim of classroom-
specific research is not so much to generalise as to promote
understanding and facilitate replication to another context. Class-
based, ethnomethodological research sets out to report trends,
patterns and tendencies rather than absolutes; ‘studies of classroom
interaction will clearly be extremely complex and tentative, and one
must take care not to draw hasty conclusions from superficially
identifiable interactional tokens’ (van Lier 1996: 143).
Summary
This chapter has provided a critical overview of some of the better-
known approaches to investigating L2 classroom interaction. Starting
with an account of the main issues that must be addressed when
recording and transcribing classroom interaction, I presented a critique
of the more traditional, and arguably more static, interaction and
discourse analysis traditions. In the final section conversation analysis
was offered as a potentially more powerful, though not unproblematic,
approach to recording classroom discourse.
In the next chapter, the focus shifts to alternative approaches to
analysing classroom discourse.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

5 Alternative approaches
to studying classroom
discourse
Introduction
Chapter 4 focused on some of the better-known approaches to
analysing classroom discourse in the form of interaction analysis,
discourse analysis and conversation analysis. In this chapter, we
consider other approaches that can be used to collect and analyse
classroom data and evaluate their relative merits or shortcomings. By
the end of this chapter you should have a pretty good idea about how
to choose an appropriate approach for analysing spoken interaction
in the classroom. In the next chapter, I introduce a framework for
evaluating classroom interaction that can be used to help teachers
improve their classroom practices.
Corpus linguistics
A corpus is a collection of texts that is stored electronically on a
computer or other form of electronic storage. These texts can be from
written sources such as books, magazines, junk mail, letters, advertise-
ments, business documents, literature, academic papers, emails and
internet pages. A corpus can also be a collection of spoken texts, such
as conversations, phone calls, speeches, TV chat shows, classrooms
and so on. A corpus, therefore, is a collection of real language that
people use in all types of situation. Because this language can be stored
Make a note of some of the difﬁculties we have discussed in analysing
classroom discourse. Based on what you now know, which approaches to
studying interaction seem appropriate to your context? How would you go
about studying interaction in one of your classes and how might you deal with
some of the problems you have listed?

easily on a computer, it can be searched quickly and easily using special
software. This makes it very suitable for research into how language
is really used. Corpus linguistics is the study of language in use, using
specialised software to carry out searches.
Using a corpus is not unlike using an internet search engine. It is
essentially a large database that you can use to find every occurrence
of a word or phrase. Just like an internet search, the ‘hits’ that result
from your search will, in a matter of seconds, pop up on your screen.
For instance, if we ask a corpus to find examples of the word
exactly, within seconds, the search software will give us a list that
looks like this:
These are called concordance lines. Concordance lines essentially
tell us how a word is used with other words, how it combines to create
particular meanings, which words it normally associates with and
so on. Essentially, concordance lines give us a lot of very useful
information about a particular word or words.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Alternative approaches to studying classroom discourse 91
train for a job and a lot of love. [p] This is exactly what the YMCA has and will continue to
full adjustable in order to support book sat exactly the right angle and weighed cords hold the
a free planning service and will work out exactly the amount of panelling you need to buy for
starters, there’s only one joke - which isn’t exactly original in the first place. It’s the one
wrong, that they’re without fault. They’re not exactly lying but they’re certainly not telling the
We opted for Shaker furniture, as it had exactly the clean lines we were looking for. A four-
if Moscow had not operated in every instance exactly as Washington might have wished. He did not
but Brand suddenly realised it summed up exactly his interest in Ella’s death. [p] As he
and the other two little girls were treated exactly alike. [o] We would be taken into the
close to the Celestial North Pole, but not exactly central. With the passage of centuries, the
I don’t remember.” But someone told you.” Not exactly. I just did know it was needful. I couldn’t
the ancient Britons, as I call them: they are exactly what that Roman (Suetonius or Agricola,
[p] Cos what happened all these girls did exactly the same thing. I was made quite clear that
There is no doubt they (the robbers) knew exactly what they were looking for.” He described
had liked it. Oh yes, I soothed, it was fun. Èxactly!” she beamed. À fun read, a woman’s book,
illustrious predecessors is play her part with exactly the right weight so that she appeals to
and arrange another abduction [p]. That was exactly what happened to Pam Green, whose two sons
TV Week [/h] [b] Ivan Waterman [/b] [p] WHAT exactly has Mariella Frostrup got? A funny surname?
to that word `fun again. Whatever it means exactly, manufacturers clearly feel that when it
And he finds out Ernesto is missing. That’s exactly the way it happened. Now, everybody tells me
see?” [p] I get you,” Stan said. Ì understand exactly what you mean.” And he waited with a child’s
in the baby boom year of 1947, Dewaere was exactly thirty-five years old when he died, a fact
crisis? What is the climax of the story? Why, exactly, does Sammy quit his job? Does anything lead
Source: www.collins.co.uk/Corpus/CorpusSearch.aspx

Alternative approaches to studying classroom discourse92
Apart from giving us concordance lines, corpus software can also
tell us how many times a word occurs in the whole corpus. However,
if this information is to be reliable and useful to us as teachers,
materials designers, or researchers, it has to come from a well-designed
corpus. When we construct a corpus, we must do so in a principled
way to ensure that it is representative, just as we would do if we were
conducting a survey using questionnaires, for example. So if we were
to build a corpus of classroom discourse, we’d need to think about
some or all of the following:
Example of a type of corpus Possible design considerations
Corpus of classroom interactions • What language levels?
• How many nationalities?
• How many age groups?
• What range of teachers: experienced,
novice, native speaker, non-native
speaker?
• How many classes?
• What types of classes?
Apart from the representativeness of a corpus, the other key
consideration we need to make is its size. Until quite recently, it was
Task 5.1
Go to the Collins Bank of English website at www.collins.co.uk/
Corpus/CorpusSearch.aspx . Use this free software to perform the
following operations for not really.
1 Under the corpus concordance sampler, type in not+really in the
‘type in your query’ box. Now tick the ‘British transcribed speech’
box.
2 Click ‘show concs’ (concordances). You’ll get a list of concordance
lines for not really.
3 Read through the concordance lines. What observations about
not really can you make? How might you use this with a group
of learners?
4 Consider how this approach could be used with a corpus of
classroom discourse. What might it reveal?
TC

assumed that the bigger a corpus the better it is. Some of the largest
corpora that are currently available contain over one billion words!
Such mega-corpora are usually held by English-language publishers
who make them available to their authors for materials design. Large
corpora are also used in making dictionaries. Here are some examples
of large publisher corpora:
• Collins, under the COBUILD project, has a corpus of 450 million
words of contemporary written and spoken English. Fifty-six
million words of this is available online as part of Wordbanks
Online English.
Source: www.collins.co.uk/Corpus/
CorpusSearch.aspx
• Cambridge University Press has a corpus of one billion words
which it makes available to its authors.
Source: www.cambridge.org/elt/corpus/cic.htm
• Longman has a written corpus of 100 million words, 5 million
words of American spoken English and 10 million words of
student writing.
Source: www.longman.com/dictionaries/corpus/
(O’Keeffe et al. 2007)
As a rule of thumb, a ‘large’ corpus means more than 5 million
words and ‘small’ usually means fewer than 5 million words. One of
the interesting developments in corpus linguistics in recent years, and
which is of relevance here, is that it is now becoming increasingly
common for researchers to use much smaller corpora of around
50,000–100,000 words. The reason for this is that these smaller
corpora are usually highly context-specific and applied to address a
particular problem. It is not difficult, for example, to see how a small
corpus of around 50,000 words of university small group talk could
be usefully applied to look at the ways in which local and international
students interact in seminars. Here, a small corpus has been used to
address a ‘real world problem’ in a very useful and illuminating way.
The reason this approach works is that all the data are recorded in a
single, homogeneous context and used in response to a particular
question or problem. The corpus is not simply descriptive, but pro-
active in relation to a particular issue.
The argument I am making then is that corpus linguistics, in the
present context, should focus more on what a corpus can do than what
a corpus is. (For definitions, the reader is referred to Biber et al. 1998;
Sinclair 1997). Corpus linguistics (CL), at least here, should be viewed
as a methodological tool that will help us investigate classroom
discourse. The range of tools that CL provides us with is discussed
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

briefly below, but essentially, CL will help us to automatically find
what we are looking for. CL will count frequencies and find key words
and it will allow us to find patterns that we could not otherwise find,
but it will not explain the dynamics of these interactions in the context
of the recordings.
Increasingly, CL is being applied to contexts and domains outside
of the study of language itself where the use of language is the focus of
empirical study in a given context. Such contexts include courtrooms
and forensic linguistics (Cotterill 2010), the workplace (Koester 2006),
the classroom and educational contexts (O’Keeffe and Farr 2003;
Walsh and O’Keeffe 2007), political discourse (Ädel 2010), advertising
and the media (O’Keeffe 2006), among other areas. In all of these cases,
CL is used as a tool and another approach is drawn on as a framework;
these include CA, discourse analysis and pragmatics (see below).
CL has at its disposal a range of techniques that can be used for
the study of language, all of which can be used to study classroom
discourse. These include:
Concordancing
As we saw in the example above, concordancing is a core tool in corpus
linguistics and it simply means using corpus software to find every
occurrence of a particular word or phrase. The search word or phrase
is often referred to as the node and concordance lines are usually
presented with the node word/phrase in the centre of the line with
seven or eight words presented at either side. These are known as Key-
Word-In-Context displays (or ‘KWIC’ concordances). Concordance
lines challenge us to read in an entirely new way, vertically, or even
from the centre outwards in both directions. They are usually scanned
vertically at first glance, that is, looked at up or down the central
pattern, along the line of the node word or phrase. Here are some
sample lines from a concordance of the word way using the Limerick
Corpus of Irish English (LCIE) taken from O’Keeffe et al. 2007:
Ireland is no different in a way then em what they were
of you anyhow? Now in a way ‘What Dreams may come’ it’s
in college in fact it’s a way of life and you find this
he present things in such a way that he would persuade
of life is to live in such a way that when you die your soul
he obviously lived a certain way of live and they wanted to
to deal with in a different way they couldn’t deal with it
stadium that’s the easiest way to describe it. There is a
find this the most effective way. Ok now today em you have as

there is no evidence either way. You can’t have evidence
from the top and works his way down. The theologian will
so it speaks and works its way up. The theologian starts
Our concordance analysis of this extract reveals the following:
• Way frequently co-occurs with the preposition in (in a way, in no
way, etc.).
• It is often used to form fixed or semi-fixed phrases (in a way, on
the way, a way of life, in such a way, etc.).
• Way is frequently followed by another preposition (to, down, up, etc.).
• It occurs more frequently in the middle of a clause than at the
beginning or end.
And so on. There are many other observations we could make and
just think how valuable this approach is when we use a larger extract,
restrict our search to three words either side of the node, or expand
it to seven words either side of the node. Concordance line analysis
is very productive in terms of looking at language patterning such as
idiomatic usages or collocation (for a detailed illustration, see O’Keeffe
et al. 2007).
Word frequency counts or word lists
Another common corpus technique that software can perform is the
extremely rapid calculation of word frequency lists (or word lists) for
any batch of texts. By running a word frequency list on a corpus, you
get a rank ordering of all the words in order of frequency. This
function facilitates enquiry across different corpora, different language
varieties and different contexts of use. Below, for example, are the
first ten words from three very different spoken corpora:
• Service encounters: a sub-corpus of the Limerick Corpus of Irish
English (LCIE) comprising shop encounters (8,500 words).
• Friends chatting: a sub-corpus of LCIE, consisting of female friends
chatting (40,000 words).
• Academic English: The Limerick-Belfast Corpus of Academic Spoken
English (LIBEL CASE, one million words of Academic English).
Even from the first ten most frequent words, we can see differing
patterns of use (any of these words can then be concordanced to look
at how they are functioning in the texts):
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

A B C
1 You I the
2 Of And And
3 Is The Of
4 Thanks To You
5 It Was To
6 I You A
7 Please It That
8 The Like In
9 Yeah That It
10 Now He is
Source: Walsh and O’Keeffe 2007.
From this very brief extract, we can make the following
observations:
• The shop (column A) and casual conversation (column B) results
show markers of interactivity typical of spoken English such as I,
you, yeah (as a response token), like, please, and thanks (see Carter
and McCarthy 2006).
• The first ten words of the academic corpus (C) lack the interactive
markers found in first two columns. The academic corpus has more
of the features of a written text such as a high frequency of articles,
a and the, use of the preposition of, use of that.
Again, there are many other observations that we could make from
this data. The point is that CL allows us to make important and reliable
observations from our data very quickly.
Key word analysis
This function allows us to identify the key words in one or more texts.
Key words, as detailed by Scott (2008), are those whose frequency is
unusually high in comparison with some norm. Key words are not
usually the most frequent words in a text (or collection of texts), rather
they are the more ‘unusually frequent’ (ibid). Software compares two
pre-existing word lists and one of these is assumed to be a large word
list that will act as a reference file or benchmark corpus. The other is
the word list based on the text(s) that you want to study. The larger
corpus will provide background data for reference comparison. For
example, we saw above that the is the most frequent word in the LIBEL
corpus of spoken academic English (see above). Scott (2008) notes the
key word facility provides a useful way of characterising a text or a
genre and has potential applications in the areas of forensic linguistics,

stylistics, content analysis and text retrieval. In the context of language
teaching, it can be used by teachers and materials writers to create
word lists, for example in Languages for Specific Purposes programmes
(e.g. English for Pilots).
Cluster analysis
The analysis of how language systematically clusters into combinations
of words or chunks (e.g. I mean, this that and the other, etc.) can give
insights into how we describe the vocabulary of a language (Wray
2000, 2002; McCarthy and Carter 2002). As O’Keeffe et al. (2007)
note, the way in which words cluster together into chunks has
implications for what we teach in our vocabulary lessons and how
learners approach the task of acquiring vocabulary and developing
fluency. As a corpus technique, the process of generating chunks or
cluster lists is similar to making single word frequency lists. Instead
of asking the computer to rank all of the single words in the corpus
in order of frequency, we can ask it to look for word combinations,
for example two-, three-, four-, five-, or six-word combinations. For
example, using Wordsmith Tools, here are the twelve most frequent
three-word combinations from the LIBEL corpus:
1 A lot of
2 And so on
3 I don’t know
4 At the moment
5 In other words
6 A number of
7 A couple of
8 A little bit
9 A bit of
10 As I said
11 First of all
12 As well as
(Walsh and O’Keeffe 2007)
What is interesting from this cluster analysis from a classroom
discourse perspective, is the frequency of words and phrases that are
used to ‘point’ in the discourse and to locate learning in time and place.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Phrases like as I said, first of all, in other words perform an important
signposting function and enable learners to find their way through the
discourse or ‘navigate’ their way in the interaction (Breen 1998). If
teachers do not signpost for their learners and highlight the direction
the discourse is going, it is a little like reading a newspaper without
punctuation; we very quickly get lost and confused. Effective teachers
use discourse markers naturally and almost without thinking.
In this section, I have outlined the main benefits of a CL approach
to the study of classroom discourse and highlighted its main advantages
in terms of speed and accessibility of data and reliability of findings.
In the next section, I consider how CL might combine with other
approaches, especially CA (Conversation Analysis).
Combined approaches
While CL offers a very useful means of analysing relatively large
corpora of spoken interactions, its focus is largely on individual and
collections of words. The perspective on the interaction that CL offers
is a distant one that portrays general trends and patterns rather than
a detailed, ‘up-close’ view of the interaction. To use Batstone’s (1994)
analogy of the view from an aeroplane as it comes into land, CL offers
us a view from around 10,000 metres. At this height, we can see the
landscape below and make out lakes, rivers and mountains. But we
cannot make out the detail. Our CL analysis allows us to make
Task 5.2
In the following extract, all the signpost words (discourse markers)
have been removed (marked XXXX in the text). Add in what you
consider to be the most suitable word or phrase. What functions are
these signpost words performing and why are they important here?
T: XXXX we’re going to be reading about and talking about
er music and pop stars and famous people Last XXXX
we did a quiz XXXX and it was a picture quiz and some
of the pictures were people from your countries. XXXX let
me see could you work with Ben and could you three guys
work together (T organises groups). If you recognise any
of those people who they are and what you know about
them and then what kind of lifestyle do you think these
people have. XXXX you can begin when you’re ready.
TC

generalisations about word frequencies and collocations, but it does
not permit us to get close to the interaction and to see how
interlocutors really communicate. To do this, we need to drop down
to around 200 metres.
At this height, we can see much more detail. We can see cars and
roads, we can make out the car park at the airport, we can even see
people. In order to get this level of detail, we need another approach
to our analysis. To do this, we need to combine CL with another
approach since CL is unable to account for some of the features of
spoken interaction that occur at the higher levels of utterance and turn.
In order to conduct a detailed analysis at this level of the discourse, I
propose that CL can usefully be combined with conversation analysis
(CA), an established and respected approach to providing detailed,
micro-analytic descriptions of spoken interaction. This combined
approach, using both CL and CA (henceforth, CLCA) cumulatively
gives a more ‘up-close’ description of spoken interactions in an
educational setting than that offered by using either one on its own.
From the analysis, we can gain powerful insights into the ways in which
interactants establish understandings and observe how words,
utterances and text combine in the co-construction of meaning.
As I have shown previously, the current trend in corpus-based
approaches to the study of discourse is towards smaller, more specialised
and context-speciﬁc corpora. Arguably, such corpora lend them-
selves to multiple layers of analysis and provide opportunities for
combining different methodologies. Previously, when it was ‘fashion-
able’ to only construct large corpora, the main focus was on lexical
patterning rather then discourse context because ‘large corpora
were lexically rich but contextually poor’ (Walsh et al. 2010). That
is, when researchers look at a lexical item in a 100 million word corpus,
they are detached from its context. However, when researchers record,
transcribe, annotate and build a small, contextualised spoken corpus,
a different landscape of the possibilities opens up in areas beyond
lexis to areas of use (especially issues of pragmatics, interaction and
discourse).
In what ways might it be desirable and useful to combine CL with
CA? As we have seen in Chapter 4, CA is essentially concerned with
turn-taking and explores all the associated features – such as topic, turn
management, turn preference, sequential organisation, repair – in
micro detail. At ﬁrst sight, it might appear that CL and CA have very
little in common. CL deals with large texts and pays little attention to
context, whereas CA looks at the microscopic details of shorter texts
that are highly contextualised. CA uses a qualitative analysis, CL’s is
quantitative. The main focus of CL is lexis: single words, combinations
of words, word clusters. CA, on the other hand, is more interested in
turns: sequential organisations, sequencing, shifts and so on.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

This said, in fact, both CA and CL have a number of features that
link them:
• Both use a corpus of empirical data.
• Both refer to baseline comparisons with other types of interactions
(canonical sequential order in the case of CA, reference corpora in
CL).
• CA offers an emic, close-up perspective, CL complements by
providing a bigger picture.
• Both start from the data and work outwards to construct context
(from turn order in CA, from patterns in CL).
• Both attach importance to lexis in creation of meanings (though
this is downplayed in CA).
• Word patterns (CL) often lead to consistent turn patterns (CA).
How might a combined CLCA approach be used in the study of
classroom discourse? The following example may help to illustrate this.
In a study on small group teaching in Irish universities, Walsh et al.
(2011) used a combined CLCA approach in order to provide ‘thick’
descriptions of spoken interaction. The authors argue that this com-
bined approach enables features of spoken discourse to be described
at both micro (word) and macro (text) levels. A CLCA methodology
reveals the relationships between interaction patterns and lexical
chunks. By looking at sequences of turns alongside, for example, dis-
course markers, new levels of understanding can be reached, the
authors claimed. As the authors themselves say:
This approach to analysis provides powerful insights into the
ways in which interactants establish understandings in educational
settings and, in particular, highlights the inter-dependency of words,
utterances and text in the co-construction of meaning.
(Walsh et al. 2011)
The following extract from the same study illustrates some of the
merits of a combined CLCA approach.
Extract 5.1
1 T: how are you getting on with your other ahh module (.)
2 ahh the the the ﬁlming one
3 S3: we’re ﬁlming a scene at the moment we’re editing and
4 it’s crazy

5 T: yeah you see it is crazy isn’t it (.) this week now
6 is going to be unbelievable
7 S3: it’s just music and we’re just putting it together
8 you see now (.) you know you’ve all the footage but
9 you’re there trying
10 S?: ( )
11 S3: we have so much footage and it’s just like you some
12 people have to accept that some of it
13 S?: ( )
14 T: yeah well it’s like essays isn’t it (.) I mean you
15 can’t write from the middle you know=
16 S4: =that was my idea so we can’t lose that and you’re
17 like going=
18 T: =who who’s the director?
19 S3: I wish I was the director
20 S2: in our in my group John C is
21 T: okay yeah you see that’s the thing like you know I
22 mean like really it does all come down to the
23 director and the people should respect that
24 immediately you know (.) that doesn’t happen that
25 often (.) you know what I mean it can get the roles
26 can get dispersed
(Walsh et al. 2011)
This extract is taken from a module on ‘film studies’. One of the
striking features that emerged from the quantitative analysis (CL) of
the corpus were the high frequencies of you know and you see (high-
lighted in extract 5.1). These discourse markers are used frequently in
classrooms in order to create ‘shared space’ where learning can occur.
According to Carter and McCarthy (2006), you see usually marks new
information while you know generally marks shared information. In
this study, there were significant differences in the frequencies of you
know and you see when compared with a corpus of everyday talk:
The interesting point here is that in the corpus of higher education
discourse, we find an exceptional number of you knows (marking
shared information) but we find more or less the same amount of you
sees (marking new information). The priority to build on and appeal
to shared knowledge and ‘shared space’ is central to both the pedagogic
and interactional process. In extract 5.1, the same result is borne out.
There are a large number of occurrences of both you see and you know,
but it is evident that you know is used much more frequently in order
to establish common ground, create shared space, demonstrate
empathy between tutor and students, and create a sense that ‘we’re
all in this together; I’m here to help you’.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

From this linguistic analysis of the extract (using CL), we now
turn to an analysis of the interaction (using CA). Our analysis of the
interaction in extract 5.1 reveals the following:
• Turns are evenly distributed and fairly symmetrical: it is not
immediately obvious who the tutor is.
• Students manage the turn-taking independently of the tutor in a
way that is similar to casual conversation.
• In response to the teacher’s opening turn, one student (S3) produces
an account of a group’s experiences of making a ﬁlm, including an
assessment of the situation (‘it’s crazy’), to which the tutor offers
a preferred (agreeing) response with the discourse marker ‘yeah’
and the repetition of the assessment, before building on this to
project what experiences will be like in the future.
In lines 11 and 12, S3 indicates that ‘some people’ may have
problems in accepting that material has to be cut, and in line 19, seems
to be expressing frustration either about the existing director, or the
lack of a director’s role in the group.
Throughout this extract, tutors take students’ experiences and
feelings and build on them, thus orienting to a pedagogic goal of
reinforcing appropriate behaviours and identities in the context of pro-
fessional practice. Throughout, the tutor plays a more or less equal
10,000
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
You see You know
LIBEL sub corp
LCIE
Figure 5.1 Comparison of ‘you see’ and ‘you know’ in LIBEL sub corpus and LCIE (normalised results)

role, listening attentively, as indicated by the use of acknowledgement
tokens (in lines 5, 14 and 21).
A lot of pragmatic work takes place in lines 21–26, where okay
marks a switch in orientation, and the tutor switches roles from
empathic listener to tutor. Here, there is a lot of interactional work
in order to change footing (Goffman 1981): okay yeah you see that’s
the thing like you know I mean like really. His stance after this preface
is that of teacher again, giving instruction and passing on new
knowledge. The interactional work is apparently needed in order to
change from equal interactant to tutor, to move from a position of
role symmetry to one of role asymmetry.
In the preceding analysis and discussion, I have tried to demonstrate
that CL and CA can be usefully combined in any study of classroom
discourse. Not only are they mutually beneficial, they actually offer each
other synergies and enable a deeper, richer level of analysis. The
approach is iterative: it requires a switch from CL to CA, back to CL
and then on to further CA. One approach informs the other, provides
direction and enables closer analysis. For example, the frequency
counts of you know and you see alone do not provide the whole picture.
Looking at how they actually operate in a piece of interaction gives a
much fuller ‘flavour’ of the relationship between linguistic forms and
interactional features. More importantly, this combined analysis gives
real insights into how people communicate in an educational setting.
Had this study used CL on its own, it would have achieved interesting
lists of high frequency items but it would not have brought us anywhere
near the depth of understanding offered by a CA framework. Similarly,
if the data had been viewed from a purely CA perspective, the analysis
would have identified some interesting interactional features but
overlooked the high frequencies of certain items and ignored how
specific lexical features operate in the interaction.
In the next section, the discussion turns to variable approaches to
analysing classroom interaction, which focus on promoting detailed
understandings of micro-contexts in a bid to understand the complex
relationship between language, interaction and learning.
Variable approaches
As teachers, we spend a lot of time talking about and even promoting
context. For example, for some time now language teachers have been
encouraged to always make sure that new language is presented ‘in
context’ rather than ‘in isolation’. We make frequent references to ‘real
world’ contexts, which apparently lie somewhere beyond the classroom
and in which language is used for ‘real’ communication. What’s more,
we contrast these ‘real’ contexts with the classroom context, which,
by inference, is not real and indeed may be viewed as being in some
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

way inferior to the real world. This is clearly not the case: the language
classroom is as much a ‘real’ social context as any other and should
be viewed as such.
According to Drew and Heritage (1992), much of the research on
L2 classroom interaction to date has adopted an approach whereby
context is viewed as something static, fixed and concrete. The majority
of studies have had one of two central goals, attempting to account
for either the nature of verbal exchanges, or the relationship between
SLA and interaction (Wu 1998). Whatever their focus, most studies
have referred to the L2 classroom context (singular), implying that
there exists such an entity and that it has fixed and describable features
that are common to all L2 contexts. One possible explanation for this
is the overriding concern of both practitioners and researchers to com-
pare L2 classroom interaction with ‘real’ communication. By following
this line of enquiry, many researchers have failed to acknowledge that
the classroom is as much a ‘real’ context as any other situation in which
people come together and interact. As van Lier (1988) says:
The classroom is in principle and in potential just as communicative
or uncommunicative as any other speech setting, no more, no less.
Nor should the ‘real world’ stop at the classroom door; the classroom
is part of the real world, just as much as the airport, the interviewing
room, the chemical laboratory, the beach and so on.
Task 5.3
Look at the following list of L2 classroom activities. Which ones do
you consider to be closer to ‘real-world’ communication and which
ones are more ‘artificial’? Place each on the continuum below and
then consider how you decided which one went where.
Artificial Real
1 Planning a composition with a group of students.
2 Doing a role play.
3 Discussing homework with students.
4 Modelling new language.
5 Watching a video and answering comprehension questions.
6 Taking part in a class debate.
TC

More recent work by Van Lier (2000) considers classroom inter-
action from a more variable perspective, arguing that the L2 classroom
environment is constantly shifting and that we need to focus more on
the different interactional processes at work in order to understand
that environment. He draws parallels with natural science, where
any changes to one species will affect another one. Other writers
have proposed that classroom interaction should be investigated from
a multi-layered perspective; a perspective where participants play a
crucial role in constructing the interaction and under which different
varieties of communication prevail as the lesson unfolds according to
particular pedagogic purposes (see, for example, Johnson 1995; Lantolf
2000; Seedhouse 2004; Tsui 1998).
The assumptions on which this variable view of context are based
are, first, that all L2 classroom discourse is goal-oriented; second, that
the prime responsibility for establishing and shaping the interaction
lies with the teacher (Johnson 1995); third, that pedagogic purpose
and language use are inextricably linked (Cullen 1998). L2 classroom
interaction is analysed according to the relationship between pedagogic
actions and the language used to achieve those actions. A variable
perspective offers, its proponents maintain, a more realistic interpreta-
tion of classroom discourse. Specifically, there is the recognition that
interaction patterns do and should vary according to the different
agendas and social relationships of the participants and according to
the linguistic and pedagogic purpose of the teacher. A blanket IR(F)
interpretation does not explain the finer variations that make up the
different contexts under which classrooms operate. Any analysis of
the interaction patterns in operation at a given point in a lesson must
take account of the fact that those patterns are dynamic and mutually
constructed by the participants, not static and pre-determined. The
notion of the single context is invalid and unworkable.
A number of writers have studied classroom interaction using a
variable approach. What follows is a brief summary of some of the
more recent studies.
van Lier (1988)
Van Lier identifies four types of L2 classroom interaction. The first,
which he calls ‘less topic-orientation, less activity-orientation’, is typical
of everyday conversation and therefore the least structured, allowing
the most freedom for self-expression. Type 2 (more topic-orientation,
less activity-orientation) is typical of the type of interaction that occurs
when information is provided in instructions or a lecture. The inter-
action is one-way and involves little in the way of an exchange of
ideas or opinions. His third type, more topic-orientation, more activity-
orientation occurs when information has to be exchanged following
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

specific and pre-determined lines, as in an interview, joke or story.
The final category, type 4 (less topic-orientation, more activity-
orientation), is typified by substitution drills, pair work and activities
that have very specific procedures.
While it is unlikely that van Lier’s classification is exhaustive and
capable of accounting for all types of interaction, it is certainly
representative of the typical patterns that occur. It also makes some
attempt to relate language use to activity; rather than proposing a
purely functional framework, van Lier’s scheme relates classroom
activity to type of language used.
Jarvis and Robinson (1997)
This study presents ‘a framework for the analysis of verbal interaction
between teacher and pupils in primary-level EFL lessons’ (1997: 212).
Considering the pedagogic functions of language, the researchers
identified a focus-build-summarise structure to classroom interaction,
based on six pedagogic functions, viz: (1) show acceptance of pupils’
utterances; (2) model language; (3) give clues; (4) elaborate and build
up the discourse; (5) clarify understandings; (6) disconfirm or reject.
The researchers examined teachers’ strategies for dealing with pupil
responses as a means of assessing the extent to which meanings were
aligned or formulated. This process, resulting in cognitive change, is
based on the Vygotskyan principle of appropriation (Vygotsky 1978,
1999), whereby children ‘appropriate’ new meanings through two-way
interaction with a more experienced interlocutor. According to Mercer
(1994), appropriation can be compared to a process of paraphrase
and recapping within the learner’s pedagogic framework. One of the
main findings of Jarvis and Robinson’s study is that teachers can
facilitate or hinder learning opportunities by using language that is or
is not pedagogically appropriate.
Kumaravadivelu (1999)
Kumaravadivelu conceptualises a framework for what he terms Critical
Classroom Discourse Analysis (CCDA) (1999: 453) aimed at ‘understand-
ing what actually transpires in the L2 classroom’ (ibid.). The framework
reflects the sociolinguistic, socio-cultural and sociopolitical dimensions
of classroom discourse. CCDA is socially constructed, politically
motivated and historically determined; the L2 classroom is viewed as a
constituent of a larger society that includes many forms of power,
domination and resistance (472). Understanding the interaction that
occurs requires an awareness of the voices, fears, anxieties and cultural
backgrounds that result in the commonly found mismatches between
‘intentions and interpretations of classroom aims and events’ (473).

Understanding classroom interaction, under the perspective advanced
by Kumaravadivelu calls for far more than an understanding of the
roles of input and acquisition in SLA; far more too than an awareness
of conversational conventions manifested in turn-taking routines.
Under CCDA, understanding the interaction of the second language
classroom requires an awareness of what participants bring to a class-
room, in terms of their beliefs, attitudes, knowledge, expectations,
conditioning and so on, as well as what actually goes on in the class-
room. Howard (2010) terms these external factors (what participants
bring) and internal factors (what actually happens). Clearly, any inter-
action is determined by a complex mix of external and internal factors.
Seedhouse (2004)
Using a CA methodology that focuses on turn-taking and sequence,
Seedhouse characterises four classroom contexts out of a total of six
he had identified earlier (Seedhouse 1996: 124–131):
(a) Form and accuracy contexts, where the focus is on linguistic form
and accuracy and the pedagogic purpose of the teacher is to elicit
from learners a string of forms for evaluation. Turn-taking and
sequence are tightly controlled by the teacher.
(b) Meaning and fluency contexts, where the teacher’s aim is to
maximise interaction within the classroom speech community
and maximise the learning potential of the classroom context. The
main focus is on fluency rather than accuracy and participants are
encouraged to express their personal feelings or emotions. Turn-
taking and topic management are less tightly structured, there is
more freedom for students to self-select and, in general, learners
have more interactional space.
(c) Task-oriented contexts, where learners communicate with each
other to complete a specific task using largely transactional
language.
(d) Procedural context, where the teacher’s aim is to ‘set something
up’, instruct or establish a procedure for work in progress.
Typically, there is no turn-taking at all and this context is
characterised by a single, long teacher turn and silence on the part
of the learners.
In an earlier study (1997), Seedhouse looks at the relationship
between pedagogy and interaction with regard to repair. How do
teachers organise repair? Specifically, what strategies do they use when
correcting oral errors and what strategies do learners expect them to
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

use? His finding, ‘that teachers perform a great deal of interactional
work to avoid performing direct and overt negative evaluation of
learner linguistic errors’ (1997: 563) indicates not only that teachers
tend to avoid overt error correction, but, perhaps more significantly,
that their choice of language and pedagogic purpose are in opposition.
That is, although the teacher’s intention is to correct errors (pedagogic
purpose), their choice of language seems to militate against this. While
learners accept that error correction is an essential part of the language
learning process, teachers seem to shy away from overt correction
because they believe it is in some way ‘face-threatening’ (Brown and
Levinson 1987). The stance adopted by teachers is largely influenced
by what would constitute an appropriate course of action outside the
classroom where overt correction might be considered less acceptable.
Yet in the language classroom, adult learners expect and indeed want
to be corrected. A teacher’s decision to correct errors in a less
‘threatening’ manner by carefully selecting language that avoids loss
of face may actually prevent or hinder repair from occurring.
Perhaps the most notable feature of each of the studies reviewed here
is the recognition that the L2 classroom is made up of a series of contexts
that are linked to the social, political, cultural and historical beliefs of
the participants (see, for example, Kumaravadivelu 1999). Contexts are
created by teachers and learners as they engage in face-to-face inter-
action and according to teachers’ pedagogic goals at a given moment.
Classroom interaction is therefore socially constructed by and for the
participants, leading some writers to suggest that we should think of
learning ‘as a process of becoming a member of a certain community
[necessitating] the ability to communicate in the language of this
community and act according to its particular norms’ (Sfard 1998: 6).
A variable approach to the study of L2 classroom contexts, by focusing
more on participation, enables greater understanding of ‘language
socialisation’ (Lantolf 2000: 156).
A second feature of the studies reviewed here is the emphasis they
place on the relationship between language in interaction and learning.
An understanding of the relationship between classroom communica-
tion and educational goals, the ways in that language use can facilitate
or hinder learning (Walsh 2002), has implications for teacher education
since it replaces ‘broad brush’ views of interaction with fine grained
paradigms that permit greater understanding of the interactional and
learning processes at work. By looking at longer stretches of discourse
and by considering the relationship between language use, pedagogic
goals and learning opportunities, it is possible to obtain a more
complete understanding of ‘what is happening’ in the discourse.
Third, in the studies reviewed in this section, there is an absence
of an agreed metalanguage for describing and accounting for L2
classroom micro-contexts. Seedhouse proposes four ‘contexts’ (2004:
102); Jarvis and Robinson six ‘pedagogic functions’ (1997: 212); van

Lier, four ‘types of interaction’ (1988: 156). This lack of an agreed
metalanguage makes the processes of comparison and generalisation
practically impossible, as the constructs used have different meanings.
Description and understanding of L2 classroom interaction is unlikely
to be advanced until an appropriate nomenclature is identified and
utilised by teachers and researchers alike. In Chapter 6, I present a
framework that offers, at least in part, a solution to this problem.
Summary
The aim of this chapter was to present alternative approaches for
studying classroom interaction, beginning with corpus linguistics (CL).
The main advantage of CL is that it offers rapid and reliable profiles
of classroom discourse and enables us to understand how linguistic
features ‘work’ in specific classroom contexts. For example, I com-
mented on the ways in which discourse markers (such as you know, you
see, right, ok, next, etc.) perform key functions in classrooms and, when
used properly, greatly assist the learning process. CL’s main
disadvantage is that it largely ignores the finer details of classroom
interaction and fails to recognise the ways in which meanings are jointly
achieved. A combined CLCA approach was proposed as one way of
overcoming such shortcomings, allowing a ‘multi-layered’ perspective
that offers a description of both linguistic and interactional features.
In the final section of this chapter, I presented a brief summary of
the work that has been done to promote variable approaches to
analysing classroom interaction. Here, there is a recognition that
classroom discourse cannot be viewed as being ‘all of a oneness’, where
interactional and linguistic features occur in a more or less fixed and
predictable way. Instead, variable approaches recognise that classroom
interaction proceeds in line with the pedagogic goals of the moment.
Micro-contexts are co-constructed in the interaction as participants
work towards clearly defined, and constantly shifting, goals. So, for
example, if the teacher’s goal is to set up a pair-work task, she may
use long teacher turns, frequent pausing, specific discourse markers
(such as first, next, then), and so on. If, on the other hand, her aim is
to elicit student opinions, there may be longer learner turns, more role
asymmetry, frequent overlaps, interruptions and false starts – in short,
the interaction will resemble more closely a casual conversation.
Recognising that classroom discourse is constantly shifting, that
goals are always changing and that language use and pedagogic goals
must work together are, I suggest, fundamental to teacher develop-
ment. One of the most effective means of developing as a teacher is
to gain closer understandings of classroom interaction: in Chapter 6,
I present a framework that is designed to help teachers improve the
quality of the interaction taking place in their classes as the first step
towards improving teaching and learning.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

6 Enhancing
understandings
of classroom
discourse
Introduction
In the last chapter, I presented an argument for adopting a variable
approach to classroom discourse, which emphasises the fact that
interaction and pedagogic goals are inextricably linked, that the
discourse is constantly changing according to the teacher’s agenda, and
that teaching and learning can be greatly improved by adopting a
variable approach. In this chapter, I present a framework (SETT: self-
evaluation of teacher talk) designed in collaboration with L2 teachers
and that aims to foster teacher development through classroom inter-
action. Essentially, my aim is to get teachers to think about classroom
interaction as a means of improving both teaching and learning. (For
a fuller discussion of this work, see Walsh 2006.)
Since its inception in 2006, SETT has been used in a number of
contexts, including: initial teacher education programmes (PGCE) for
English and Drama teachers (Walsh and Lowing 2008); INSET courses
for experienced teachers; a study evaluating the value of classroom
observation in the Middle East (Howard 2010); on CELTA
programmes around the world; a primary science classroom; various
secondary EFL contexts around the world; two university classroom
contexts, and an Irish medium secondary classroom. In short, the
framework has been used extensively to promote awareness and
understanding of the role of interaction in class-based learning and to
help teachers improve their practices.
The SETT framework comprises four classroom micro-contexts
(called modes) and fourteen interactional features (called interactures).
Classroom discourse is portrayed as a series of complex and inter-
related micro-contexts (modes), where meanings are co-constructed
by teachers and learners and where learning occurs through the ensuing
talk of teachers and learners. The next two sections offer a summary
of the framework, beginning with a description of each mode and then
characterising the different interactures.

The SETT framework: modes
The SETT framework is designed to help teachers both describe the
classroom interaction of their lessons and develop an understanding
of interactional processes as a way of becoming a ‘better’ teacher. The
starting point, based on the discussion in Chapter 5, is that the single, L2
classroom context does not exist; contexts are locally constructed by
participants through and in their interaction in the light of overall institu-
tional goals and immediate pedagogic objectives. The notion of ‘the L2
lesson context’ is too broad-brushed; ‘contexts are locally produced and
transformable at any moment’ (Drew and Heritage 1992: 19). In short:
• The L2 classroom context is made up of a series of micro-contexts,
linked to a range of ‘external’ factors (c.f. Howard 2010) such as
beliefs, attitudes, previous experience and so on.
• Contexts are created in and through the interaction, which is also
the way in which meanings are co-constructed (c.f. Lantolf and
Thorne 2006).
• The relationship between language use and pedagogic goals requires
closer understanding (c.f. Seedhouse 2004).
In other words, it is impossible to separate pedagogy and inter-
action, which come together through talk: pedagogic goals are
manifested in the talk-in-interaction. Using the term mode encompasses
the interrelatedness of language use and teaching purpose. A mode is
defined as ‘an L2 classroom microcontext which has a clearly defined
pedagogic goal and distinctive interactional features determined largely
by a teacher’s use of language’ (Walsh 2006: 62). The definition is
intended to emphasise the idea that interaction and classroom activity
are inextricably linked, and to acknowledge that as the focus of a lesson
changes, interaction patterns and pedagogic goals change too.
In the 2006 study, Walsh set out to design a framework that could
be used by teachers to evaluate and gain closer understandings of the
interactions taking place in their classes. The work was conducted in
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Enhancing understandings of classroom discourse 111
Based on what you now know about classroom discourse and your own
experience, how would you like to change your own use of language in the
classroom? How would you like to improve your teacher talk and improve the
quality of the interaction? What steps have you already taken to improve
classroom interaction and how effective were they?

a UK university’s English Language Centre and the participants were
the tutors and students of that centre. Recordings were made and
analysed using an applied conversation analysis methodology (ten
Have 2007). To summarise, the following procedures were followed:
• The total size of the corpus is approximately 100,000 words, or
twelve hours.
• Data were analysed using an applied CA methodology.
• Participants were then invited to study the transcripts in a series of
workshops in which the precise focus of each lesson was discussed
in relation to pedagogic goals.
• Interaction patterns were found to vary according to instructional
activity; for example, establishing procedures to complete an activity
resulted in a very different pattern of interaction to open-class
discussion.
• The different patterns manifested themselves in the turn-taking,
sequence of turns and topic management. Once a pattern had been
identified, the data were analysed for further examples of the same
pattern as is the ‘norm’ under conversation analysis.
(Psathas 1995: 52)
Following this procedure, four patterns, four micro-contexts were
identified. These were called modes: managerial mode, classroom
context mode, skills and systems mode, materials mode.
Each mode is made up of specific interactional features (such as
display questions, repair, content feedback) and particular pedagogic
goals. The four modes are included as being representative, rather
than comprehensive, and can be adapted to suit local contexts.
They are based on the analysis of frequently occurring classroom
discourse features and are designed to help teachers develop detailed
understandings of their own contexts.
Heritage and Greatbatch’s (1991) notion of ‘fingerprints’ is helpful
to the present discussion. In that study, the researchers identify a
number of socially constructed contexts in different institutional
settings that they term ‘fingerprints’ to differentiate interactional
organisations from one workplace to another. Thus, the ‘fingerprint’
of a doctor’s surgery will have a different exchange and participation
structure to that of a solicitor’s office. Following Heritage and
Greatbatch, under SETT, each L2 classroom mode has its own
distinctive fingerprint, comprising pedagogic and linguistic features.
Thus, the fingerprint of classroom context mode is markedly different
to that of managerial mode; both are different again from skills and
systems mode.
Enhancing understandings of classroom discourse112

The four modes, together with their interactional features and
typical pedagogic goals, are summarised in Table 6.1 above.
In the following sections, a description of each mode, together with
examples from the data, is presented. A full account of the transcription
conventions used appears in Appendix B. The numbers given in
brackets refer either to a whole extract, or a line in an extract; thus,
(6.1) refers to the whole extract, while (14) refers to line 14 in the
extract.
Managerial mode
Managerial mode accounts for what goes on in the organisation of
learning. Its prime pedagogic goal is to organise learning in time and
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Table 6.1 L2 Classroom modes
Mode Pedagogic goals Interactional features
Managerial To transmit information A single, extended teacher turn that
To organise the physical learning uses explanations and/or
environment instructions
To refer learners to materials The use of transitional markers
To introduce or conclude an activity The use of confirmation checks
To change from one mode of An absence of learner contributions
learning to another
Materials To provide language practice Predominance of IRF pattern
around a piece of material Extensive use of display questions
To elicit responses in relation to Form-focused feedback
the material Corrective repair
To check and display answers The use of scaffolding
To clarify when necessary
To evaluate contributions
Skills and To enable learners to produce The use of direct repair
systems correct forms The use of scaffolding
To enable learners to manipulate Extended teacher turns
the target language Display questions
To provide corrective feedback Teacher echo
To provide learners with practice Clarification requests
in sub-skills Form-focused feedback
To display correct answers
Classroom To enable learners to express Extended learner turns
context themselves clearly Short teacher turns
To establish a context Minimal repair
To promote oral fluency Content feedback
Referential questions
Scaffolding
Clarification requests
Source: Walsh 2006.

space and to set up or conclude classroom activities. It frequently
occurs at the beginning of lessons, as illustrated in extracts 6.1 and
6.2 below, characterised in the first instance by an extended teacher
turn of more than one clause and a complete absence of learner turns.
In each extract, the focus is on the ‘institutional business’ of the
moment, the core activity of the school, organisation or whatever. In
managerial mode, there are frequent repetitions, directives and
instructions. At the end of managerial mode, there is typically a
handing over to learners and a movement into another mode; in
extract 6.1, for example, the words ‘so Miguel’ act as a transition to
another mode, here, skills and systems. Note too the use of discourse
markers such as ok, so, now, right, etc., which help learners to follow
what is being said and give direction to the discourse.
Extract 6.1
1 T: Ok we’re going to look today at ways to improve your
writing and at ways that can be more effective for you
and if you look at the writing which I gave you back you
will see that I’ve marked any little mistakes and eh I’ve
also marked places where I think the writing is good and
I haven’t corrected your mistakes because the best way
in writing is for you to correct your mistakes so what I
have done I have put little circles and inside the circles
there is something that tells you what kind of mistake
it is so Miguel, would you like to tell me one of the
mistakes that you made (3)
Extract 6.2
1 T: now could you turn to page . . . 59 page 59 at the top
of the book (students find place in book)
When managerial mode occurs at the beginning of a lesson, the
teacher’s main concern is to ‘locate’ the learning temporally and
pedagogically (extract 6.1), or spatially (extract 6.2). Once learning
has been located, learners are invited to participate: so Miguel, would
you like to tell me one of the mistakes that you made. Locating
learning is an important first step in building a main context;
consequently, in many respects, managerial mode functions as a
support to the other three modes. We can say that it is an ‘enabling’
mode (McCarthy and Walsh 2003).
Although it is most commonly found at the beginning of a lesson,
managerial mode may occur post-activity or as a link between two
stages in a lesson, as indicated in extract 6.3 below, where the teacher’s

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
aim is to conclude an activity and move the lesson on. As in the
previous extracts, turn-taking is wholly managed by the teacher,
learners have no interactional space and the agenda, the pedagogic
goal of the moment, is firmly in the hands of the teacher. Once the
activity is concluded, the learners are organised into three groups and
the lesson moves from one type of learning (pair-work practice) to
another (open class checking in groups). Throughout, the teacher’s use
of language and pedagogic purpose are at one: the language used is
appropriate to the pedagogic goal of the moment.
Extract 6.3
5 T: all right okay can you stop then please where you are
. . . let’s take a couple of . . . examples for these and
. . . put them in the categories er . . . so there are three
groups all right this one at the front Sylvia’s group is A
just simply A B and you’re C (teacher indicates groups)
all right so . . . then B can you give me a word for ways
of looking (3) so Suzanna . . . yeah
The transition markers all right, okay, so signal the end of one
part of the lesson and alert learners to the fact that the lesson has
moved on, that pedagogic goals have been realigned with a shift in
focus to a new activity. These discourse markers are essential for
learners to follow the unravelling interaction and ‘navigate their way’
(Breen 1998) through the classroom discourse. They function like
punctuation marks in a written text, or intonation patterns in a spoken
text and are crucial to understanding. In cases where discourse markers
are not used, the boundaries between modes are difficult to detect and
learners may become confused as to what they are expected to do.
Most teachers have encountered comments by learners such as: ‘What
are we supposed to be doing?’; ‘Where are we?’; ‘What’s the task?’.
These, and similar questions, testify to the need for signposting and
the use of language that is related to the pedagogic goals of the
moment.
Task 6.1
Thinking about your own teaching, how does managerial mode
manifest itself in your classes? In what ways do you manage learning?
What difficulties do you encounter in organising and managing
learning and how might you overcome those difficulties?
TC

To summarise, managerial mode is characterised by one, long
teacher turn, the use of transition markers and an absence of learner
involvement. Its principal pedagogic purpose is the management of
learning, including setting up a task, summarising or providing
feedback on one particular stage of a lesson.
Materials mode
In this mode, pedagogic goals and language use centre on the materials
being used. All interaction typically evolves around a piece of material
such as text, tape, worksheet and so on. In most cases, the interaction
is tightly controlled and follows the IRF exchange structure. If we look
at extract 6.4 below, we see a clear example of materials mode.
Here, learners are completing a cloze exercise on sports vocabulary
and the teacher directs their contributions; the interactional organisa-
tion is almost entirely determined by the materials and managed by
the teacher. Teacher and learner turns are mirrored by the material:
the teacher elicits responses (in lines 81, 83, 85, 87, 90, 92, 95, 99,
104, 108) and learners respond (in lines 84, 86, 88, 91, 94, 96, 98,
etc.). The sequence is ‘classic IRF’, the most economical way to pro-
gress the interaction, with each teacher turn functioning as both an
evaluation of a learner’s contribution and initiation of another one.
There is only one turn (in line 103) that is not determined by pedagogic
goals, though it is related; unusually, it is a learner’s correction of the
teacher’s pronunciation. Very little interactional space or choice of
topic are afforded since the interaction is organised exclusively around
the material. Pedagogically, the focus can be interpreted as providing
vocabulary practice around a specific piece of material. Key items
of vocabulary are elicited, confirmed and displayed by the teacher
through echoes of a previous contribution.
Extract 6.4
81 T: now . . . see if you can find the words that are suitable
in in these phrases
82 (reading) in the world cup final of 1994 Brazil Italy 2 3
2 and in a shoot-out . . .
83 what words would you put in there? ((1))
84 L7: [beat]
85 T: [what] beat Italy 3 2 yeah in?
86 L7: in a penalty shoot-out
87 T: a what?
88 L7: in a penalty shoot-out
89 T: in a penalty shoot-out very good in a penalty shoot-out
. . . (reading) after

90 90 minutes THE?
91 LL: the goals goals goals (mispronounced)
92 T: [the match] was . . . what?
93 L: [match]
94 LL: nil nil
95 T: nil nil (reading) and it remained the same after 30
minutes OF (3)
96 L5: extra time
97 T: extra time very good Emerson (reading) but then Italy?
98 L5: lost (2)
99 T: but then Italy . . . what?
100 L5: lost=
101 T: =lost ok 3–2 in the penalty shoot-out after Venessi and
Bagio
102 (mispronounced) both missed
103 L: Bagio (correcting teacher’s pronunciation)
104 T: Bagio yes Spanish (reading) this was the fourth time
that Brazil had?
105 LL: =won=
106 T: =won . . .
107 LL: /won won/
108 T: the World Cup very good (5) and ((2)) what’s that word?
((5))
While turn-taking and topic choice in this mode are largely
determined by the material, there are varying degrees of association,
evidenced in extract 6.5. In this extract, a group of pre-intermediate
students is working on ﬂuency practice and while the interaction still
stems from the activity they are engaged in (indicated by the ‘on-task’
comments of the participants), learners are given far more interactional
space and manage the turn-taking themselves. The teacher is still
involved, but only intervenes when necessary to clarify (10, 16). The
IRF sequence no longer prevails and there is apparently more freedom
in topic choice. Closer analysis, however, reveals that this is in fact
not the case; while learners certainly have more freedom to self-select
or remain silent, contributions are made in response to the task. All
interactions here orient to the task and there is, in fact, little freedom
for learners to select topics and manage the interaction themselves.
Extract 6.5
1 L1: was shy so didn’t have a (1))=
2 L: so it’s good news (laughter)
3 LL: /bad news/ ok / no no that’s good news/ . . .
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

11 L2: ok so this one? (laughter)
12 LL: /oh/ yes that’s correct /yeah/ . . .
13 L1: so=
14 LL: /((3))/ he’s sick/ he’s/show me this one/=
15 L1: =no! it’s my card excuse me
16 T: so what’s up you have to say the bad news=
17 L2: =bad news because you can’t ski=
In materials mode, then, patterns of interaction evolve from
the material that largely determines who may speak, when and
what they may say; the interaction may or may not be managed
exclusively by the teacher. Though learners have varying degrees
of interactional space, depending on the nature of the activity, their
contributions are still bounded by the constraints imposed by the
task in hand.
Skills and systems mode
In skills and systems mode, pedagogic goals are closely related to
providing language practice in relation to a particular language system
(phonology, grammar, vocabulary, discourse) or language skill
(reading, listening, writing, speaking). Teaching objectives may also
relate to the development of speciﬁc learner strategies. Typically, the
interaction in this mode follows a lockstep organisation and the IRF
sequence frequently occurs. Turn-taking and topic selection are
determined by the target language and responsibility for managing the
turn-taking usually lies with the teacher. Pedagogic goals are oriented
towards accuracy rather than ﬂuency and the teacher’s concern is to
get learners to produce strings of accurate linguistic forms and
manipulate the target language. Direct repair and scaffolding have an
important role to play, as illustrated in the next extract.
In extract 6.6 a group of intermediate level students is practising
simple past forms. The teacher’s pedagogic goal, as evidenced in the
interaction, is to get the learners to produce patterns involving the use
of irregular simple past forms went and broke. The slight pause in
(218), (indicated [. . .]), provides the teacher with an opportunity to
scaffold the learner’s contribution in (219). Scaffolding involves the
‘feeding in’ of essential language as it is needed and plays an important

part in assisting learners to express themselves and acquire new
language. It is followed in (225) and (227) by direct repair, which is
also used in (233) and (235). Direct repair, involving a short, quick
correction, is a useful interactional strategy since it has minimal impact
on the exchange structure. Patterns of interaction are only slightly
disturbed and the ‘flow’ is maintained.
This combination of scaffolding and direct repair is found extens-
ively in skills and systems mode, enabling learners to attend to
specific features of their interlanguage while keeping the interaction
‘on track’, in line with the teacher’s pedagogic goals, the agenda
of the moment. Getting learners to ‘notice’ patterns (Schmidt 1990)
and identify relationships is a central goal in skills and systems
mode. Little attention is given to meaning, to communicative function;
the prime objective is to enable learners to understand and produce
target forms. It is widely acknowledged that form-focused instruc-
tion plays a significant part in the SLA process (c.f. Doughty and
Williams 1998; Ellis 2001); clearly, the teacher’s handling of learner
contributions through scaffolded instruction and repair are central
to that process.
Extract 6.6
218 L5: =the good news is he went to the went to [. . .]
219 T: he went to what do we call these things the shoes with
wheels=
220 L2: =ah skates=
221 L6: =roller skates=
222 T: =ROLLer skates roller skates so [he went
223 L5: [he went to=
224 L: =roller SKATing=
225 T: =SKATing=
226 L5: =he went to=
227 T: =not to just he went [roller skating he went roller skating
228 L5: [roller skating he went roller skating=
229 T: =lets hear it he went the good news [is
230 L5: [the good news is
he went skating . . .
231 T: good he went roller skating=
232 L5: =the bad news is he . . .was broken his leg=
233 T: =he? (2)
234 L5: he he has . . .
235 T: simple past . . .
236 L: he broke=
237 L5: =he broke he broke his leg=
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

In the next extract, 6.7, the teacher of an advanced group of learners
is trying to elicit the phrase military force. Of note here is the use of
display questions to guide learners to identify the missing word in (187,
189, 191, 195, 197, 199) and the very closely connected turn-taking
structure, evidenced by frequent overlaps ([ ) and latched turns (=).
The teacher’s intention is to elicit the target vocabulary military force,
which he eventually does (in 198). Asking display questions to guide
learners to the correct answer, or at least the one that the teacher is
looking for, is a notoriously complex skill, requiring deft management
of learner contributions to bring them closer to the answer being sought
(Johnson 1992). Frequently, ‘good answers’ are rejected simply because
they do not conform to the one the teacher is looking for, forcing
students to do the impossible and guess what’s inside the teacher’s
head (see also Lin 2000). In extract 6.7, learners are actively involved
in constructing meaning through the interaction, as opposed to being
passive recipients. This ‘dialogue’ is, arguably, more likely to result in
learning, since learners are engaged and having to think more carefully.
Extract 6.7
187 T: =what do we call I’m going to try and get the class to
tell you what this
188 word is that you’re looking for . . . er we talk about
military (claps hands) . . .
189 military what?
190 L: ((1))=
192 L: =kill=
193 T: =no not [kill
194 L: [action action=
196 LL: =power=
198 LL: ((3))force=
201 T: =that guides our lives
To summarise, in skills and systems mode, the focus is a speciﬁc
language system or sub-skill. Learning outcomes are typically achieved
through tightly controlled turn-taking and topic selection, determined
by the teacher. Learners respond to teacher prompts in an endeavour
to produce linguistically accurate strings of utterances. The interaction
is typically (though not exclusively) form-focused, characterised by
extended teacher turns, display questions and direct repair.

Classroom context mode
In classroom context mode, the management of turns and topics is
determined by the local context, ‘the communication potential of the
L2 classroom itself’ (van Lier 1988: 30). Opportunities for genuine
communication are frequent and the teacher plays a less prominent
role, taking more of a ‘back seat’ and allowing learners all the
interactional space they need. The principal role of the teacher is to
listen and support the interaction, which frequently takes on the
appearance of a naturally occurring conversation.
In extract 6.8 with a group of advanced learners, the teacher’s stated
aim is ‘to generate discussion prior to a cloze exercise on poltergeists’
and learners have been invited to share their experiences. The turn-
taking is almost entirely managed by the learners, with evidence of
competition for the floor and turn gaining, holding and passing, which
are typical features of natural conversation. Turns are also significantly
longer, contain more overlaps and latches and pausing is more
pronounced. In (261) for example, the two-second pause at the end
of learner 3’s turn is perceived by other learners as an invitation to
take up the discussion and two learners take a turn in (262) and (263),
before the original speaker (L3) regains the floor in (264). Topic shifts
are also managed by the learners (in 264, 269, 281), with the teacher
responding more as an equal participant, allowing the discourse to
develop within the topic frames selected by the learners. Note how
(in 280) the sub-topic of ‘neuroses’ is not developed and the original
speaker retakes the floor in (281), shifting to a new topic. The only
questions asked by the teacher are referential (in 276, 285, 287) and
extended learner turns dominate the sequence. Errors go unrepaired,
there are no evaluative comments and the only feedback given is
content-based, normally in the shape of a personal reaction.
Extract 6.8
256 L3: =ahh nah the one thing that happens when a person
dies ((2)) my mother
257 used to work with old people and when they died . . . the
last thing that went
258 out was the hearing ((4)) about this person=
259 T: =aha (2)
260 L3: so I mean even if you are unconscious or on drugs or
something I mean
261 it’s probably still perhaps can hear what’s happened (2)
262 L2: but it gets ((2))=
263 LL: /but it gets/there are ((2))/=
264 L3: =I mean you have seen so many operation ((3)) and so
you can imagine
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

265 and when you are hearing the sounds of what happens
I think you can get a
266 pretty clear picture of what’s really going on there=
267 L: =yeah=
268 L: =and and . . .
269 L1: but eh and eh I don’t know about other people but eh
((6)) I always have
270 feeling somebody watching watch watches me=
271 T: =yes=
272 L4: =YEAH=
273 L1: =somebody just follow me either a man or a woman
I don’t know if it’s a
274 man I feel really exciting if it’s a woman ((4)) I don’t know
why like I’m trying to
275 do things better like I’m eh . . . look like this . . . you FEEL
it . . . I don’t know=
276 T: =you think it’s a kind of spirit=
277 L1: =I think it’s just yeah somebody who lives inside us and
((3)) . . . visible area . . .
278 L4: I would say it’s just neurotic problems (laughter)
279 L1: what what . . .
280 L4: nothing nothing nothing . . .
281 L1: but have you seen city of angels=
282 L4: =no I haven’t=
283 L1: =with eh Meg Ryan and eh Nicholas Cage it’s a
wonderful story and I think
284 it’s true actually=
285 T: =and does this bother you=
286 L1: =what?=
287 T: =this feeling that you get does it bother you?=
288 L1: =it’s eh you know when I am alone I’m ok but if I feel
that somebody is
289 near I would be nervous=
290 T: =I would be very nervous . . .
The predominant interactional feature of extract 6.8 is the local
management of the speech exchange system; learners have considerable
freedom as to what to say and when. This process of ‘topicalisation’
(Slimani 1989), where learners select and develop a topic, is signiﬁcant
in maximising learning potential since ‘whatever is topicalised by the
learners rather than the teacher has a better chance of being claimed
to have been learnt’ (Ellis 1998: 159). In this extract, the academic
task structure and social participation structure (Johnson 1995) are
clearly more relaxed and opportunities for learning are increased. In
many classroom contexts, it is, however, not uncommon for teachers
to retain control of the interaction, interrupting students and taking

control of the topic. While this may be inevitable, there are times, as
in classroom context mode, when relinquishing control of turn-taking
and topic choice are fundamental interactional strategies. This is,
arguably, a skill that more teachers need to develop if more even, equal
participation structures are to be achieved.
The deﬁning characteristic of classroom context mode, then, is
interactional space: extended learner turns predominate as participants
co-construct the discourse. Teacher feedback shifts from form- to
content-focused and error correction is minimal. In short, the orienta-
tion is towards maintaining genuine communication rather than
displaying linguistic knowledge.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Task 6.2
Using the following extract:
1 Identify examples of each of the following modes (use Table 6.1
on page 113 to help you):
• Skills and systems
• Classroom context
• Materials
2 What difﬁculties did you encounter?
3 How might you use this type of analysis in your own teaching?
(In the extract, a group of intermediate level adult learners are
preparing to read a text called Pot Luck Suppers. Leslie is a character
in the coursebook and Yvette, Georgia and Haldoun are all students
in the class.)
29 T: =yes olive oil yes yes must be a good quality=
30 L: =and the other olive oil vegetable and made from
maize from corn=
31 T: =yes corn oil and yeah there’s many different
types of oil olive oil is the best and . . .
32 essential to Spanish omelette tortilla erm right so
if you were invited to one of Leslie’s
33 pot-luck suppers (referring to the book) erm right
what would you bring . . . along . . .
34 with you?
35 L: American or ((1))?
TC

36 T: yes this is this Leslie the American woman imagine
she [invited you to a pot-luck
37 supper
38 L: [a bottle of wine
39 T: a bottle of wine yes
40 L: a dish [of food
41 T: [what else? a dish of food yes what=
42 L: =dessert
43 T: yes possibly a dessert=
44 L: =or a starter
45 T: or a starter . . . yes so how would you know what
to bring?
46 LL: ((2)) the organiser give you a list ((1))
47 T: that’s right yes so it’s it’s quite well-organised you
can (laughs) imagine writing a list
48 oh yes I’ll ask Yvette she can bring a starter . . .
oh Georgia a nice salad Italians are
49 very good with salads and then maybe a main
course er Haldoun and perhaps er the
50 Japanse can do some more some more main
courses so it would be very er very
51 well-organised=
52 L: =usually((2)) like sometimes there are like 3 main
courses ((2)) like graduation in
53 sometimes we have parties and we have like er
roast chicken and another people
54 bring ((3)) for the main courses and we don’t have
nothing for nothing for dessert and
55 nothing for the starter [(laughs)]
56 T: [(laughs)] oh dear well Georgia perhaps when you
go back to Italy perhaps you can
57 organise one of these typical pot-luck suppers and
organise it well so you’ll have
58 plenty of desserts and plenty of starters (laughs)
but er do you think it’s a good idea?
59 L: sometimes, yeah
60 L: it’s nice because you don’t know what you’re
going to eat=
61 T: =it’s a surprise yes, yeah as long as you like
everything . . . I mean some people don’t
62 like certain things what adjective do we use for
people who don’t like that and hate
63 that and a lot of food they won’t eat . . . we call it
people are very . . . FUSsy fussy with

The SETT framework: interactures
As we have seen, the SETT framework uses four modes, each with
distinctive pedagogic goals and different interactional features. In this
section, the various interactional features are presented and discussed.
You may find that some of these are known to you, while others may
be quite new.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
64 their food (writes on board) right so fussy that’s
don’t like vegetables, never eat er
65 pasta (laughs)=
66 L1: =excuse me how to say if er you ((2)) for example
if you try some burger on corner
67 . . . on the the street and then you feel not very well
68 T: er . . . I’m not quite sure what you mean Yvette if
you?
69 L1: you can buy for example ((2)) just on the street
70 T: you mean street sellers . . . people selling food on
the street yes
Task 6.3
The SETT framework is made up of the fourteen interactional features
(interactures) below. In the table below, match the features with the
descriptions given. Two have been completed for you. Compare with
the original grid (see Appendix A).
Scaffolding Direct repair Referential questions
Content feedback Seeking clarification Display questions
Extended wait-time Teacher echo Extended teacher turn
Confirmation checks Teacher interruptions Turn completion
Extended learner turn Form-focused feedback
Feature of teacher talk Description
A. 1. Reformulation (rephrasing a learner’s
contribution).
2. Extension (extending a learner’s contribution).
3. Modelling (correcting a learner’s contribution).
TC

The interactures used in the SETT framework can be found in
varying degrees in any classroom. Some are more common than others
and occur with greater regularity throughout the discourse; display
questions, for example, are more widespread than referential, or
genuine, questions. Using a SETT analysis allows teachers to gain a
rapid profile of the kinds of features to be found in their classes and
assess their appropriacy. From Table 6.1, we can identify which
features, which interactures, belong to which mode. Content feedback,
for example, is more likely to be found in classroom context mode,
whereas skills and systems mode is more likely to contain examples
of direct repair, which, in turn, will be less visible in classroom context
mode. An interacture, then, can be defined as a particular interactional
feature that ‘belongs to’ or is typical of a mode: display questions are
interactures of materials mode, extended teacher turns of managerial
mode, for example. An interacture can be regarded as being more or
less appropriate, or mode convergent, at a given moment in a lesson
according to a teacher’s desired learning outcomes.
By making short recordings of your teaching and by listening to these
recordings, it is possible to increase your awareness of (a) which modes
are being used, and (b) which interactures appear in each mode. This
B. Correcting an error quickly and directly.
C. Content feedback Giving feedback to the message rather than the
words used.
D. Allowing sufficient time (several seconds) for
students to respond or formulate a response.
E. Referential questions Genuine questions to which the teacher does not
know the answer.
F. 1. Teacher asks a student to clarify something the
student has said.
2. Student asks teacher to clarify something the
teacher has said.
G. Making sure that teacher has correctly
understood learner’s contribution.
H. Learner turn of more than one sentence.
I. 1. Teacher repeats a previous utterance.
2. Teacher repeats a learner’s contribution.
J. Interrupting a learner’s contribution.
K. Teacher turn of more than one sentence.
L. Completing a learner’s contribution for the learner.
M. Asking questions to which teacher knows the
answer.
N. Giving feedback on the words used, not the
message.

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
combined analysis, first focusing on modes, then on interactures, will
give you a detailed profile of the interactions taking place in your classes
and permit you to make adjustments. For example, you may find that
in classroom context mode you only ask display questions, thereby
restricting the kinds of response open to learners. Or you may notice
that you constantly interrupt when eliciting. These are all important
features of your teaching that can be changed in order to improve
learning and learning potential for your students.
As I said at the beginning of this chapter, SETT has been used in
a range of contexts with different teachers, some experienced, some
less so. What follows is a brief account of teachers’ self-evaluations
where they comment on the interactures identified in their own
classes and evaluate their relative merits. In the extracts that follow,
there is clear evidence that interactures are interpreted by teachers as
being more ‘relevant’, ‘appropriate’ or ‘necessary’ according to mode.
Task 6.4
Using the extract below and the SETT grid, identify as many examples
of interactures as you can. Use the coding from Task 6.3: A, B, C,
etc. Were there any features that you did not find? What difficulties
did you experience?
129 L5: =I believe in trying new things and ((1)) ideas=
130 T: =er (1) you believe in being POSitive you mean?=
131 L5: =pardon?=
132 T: =do you believe in always being positive is that what
you mean?=
133 L5: =no . . . I believe to (3) to have a lot of achievement
134 T: (1) do you believe in what do you mean you you
should always take opportunities is
135 that what you mean no?=
136 L5: =no I want my life to be very (1)
137 T: happy?=
138 L5: =yeah and also I I do many things (1) many different
experiences=
139 T: =why don’t you say you just believe in experiencing
as many different things as you
140 want=
141 L5: =oh yeah=
142 T: =thats what I think you should say=
TC

An examination of extract 6.9, for example, reveals that some
interactures (content feedback) are more or less ‘relevant’ according
to the mode ‘mainly skills and systems’; in skills and systems mode,
feedback typically is form- rather than content-focused. Her comments
add weight to the argument that certain interactures are more appro-
priate in some contexts than others, that teacher talk has to reflect the
focus of the lesson, that mode convergence is necessary if learning
opportunities are to be maximised.
Extract 6.9
(Reading) ‘content feedback’ there was nothing because it wasn’t
relevant it wasn’t that type of lesson part of the lesson.
(Teacher 3)
Similarly, teacher 4 in extract 6.10 indicates that certain interactures
are more or less ‘necessary’ or ‘appropriate’ according to mode. While
it is certainly NOT the suggestion here that teachers should be trained
to ONLY use certain interactures in a certain mode, there is nonetheless
a perception that there are different degrees of appropriacy depending
on the mode in operation. In extract 6.9, for example, it is the teacher’s
belief that content feedback ‘belongs’ in a different mode, not skills
and systems.
Extract 6.10
I didn’t do any completing of turns, form-focused feedback didn’t fit
in really and there was no direct repair really because I suppose it
was fluency rather than accuracy and didn’t given anybody any
extended wait time because I didn’t think it was necessary but I would
have done it if they had needed it.
(Teacher 4)
Judgments concerning the relevance or appropriacy of a particular
interacture are made retrospectively in the light of desired teaching/
learning outcomes or in reaction to learner needs. It is not the
suggestion that such decisions are taken consciously ‘in the heat’ of a
teaching moment, merely that they can be usefully studied as part of
the reflective process of SETT. We will return to this in Chapter 7.
In extract 6.10, teacher 4 is in classroom context mode, clearly
aiming to elicit longer turns from the learners and give them more
interactional space – ‘it was fluency rather than accuracy’. Again, there
is a suggestion in his comments that certain interactures ‘fit in’ to
certain modes more than others and that teacher talk has to be adjusted

not only in line with desired learning outcomes, but also in response
to learner ‘needs’. In the data collected so far, it is evident that teachers
vary in their abilities to fine tune their language to promote learning
and make ‘good interactive decisions’ (Ellis 1998). Developing the kind
of understanding evidenced in the comments presented here concerning
the relationship between interactures and modes is perhaps one of the
ways in which second language teachers’ interactional choices can be
improved.
Using SETT
Thus far, the impression given is that L2 classroom interaction can be
classified very neatly into a finite number of modes, each with its own
particular set of pedagogic and interactional features. While there are
certainly examples of clearly delineated modes in the data, there are
also many instances where identifying modes is more problematic and
where features cannot be clearly shown. The framework, as already
stated, is intended to be representative rather than comprehensive. The
four modes depicted are quite clearly delineated by pedagogic goals
and interactional features; while there are some similarities, there
are also differences that make description possible. Yet the modes do
not claim to account for all features of classroom discourse, nor are
they sufficiently comprehensive to specify each and every pedagogic
goal. The first difficulty, then, is that the framework is incapable of
describing all aspects of classroom interaction. For example, inter-
actions that are not teacher-fronted, where learners work independently
of the teacher, are not described. Rather, the framework is concerned
to establish an understanding of the relationship between interaction
and learning; specifically, the interface between teaching objectives and
teacher talk. In essence, as a tool for teacher education, the frame-
work has to enable teachers to describe interaction relatively easily
and unambiguously. Yet tensions are inevitable. The discussion now
considers some of the difficulties that arise when using SETT. Because
classrooms are busy places and because the interactions that take place
are rapid, fast-changing and operating on many layers, description is
complex. The following features of the discourse are now discussed
using SETT:
(a) Mode switching: movements from one mode to another.
(b) Mode side sequences: brief shifts from main to secondary mode
and back.
(c) Mode divergence: where interactional features and pedagogic goals
do not coincide.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Mode switching
The overriding problem is one of demarcation; modes can be difficult
to distinguish and there are times when several modes seem to occur
simultaneously. The ‘neat and tidy’ examples presented in the earlier
part of this chapter do occur, but more often than not modes occur
in combination with other modes, rather than in isolation. Interactional
decisions are taken in the ‘here and now’ of a lesson, the moment by
moment sequence of planning and action, influenced by many factors,
including time constraints, teacher and learner agendas, the inter-
dependence of turns, unexpected occurrences. Lessons rarely progress
from A to Z; like conversations, deviations, topic-shifts, back-
channelling, repetitions, false-starts, overlaps all occur very regularly,
making description difficult to achieve. Add to this the fact that any
transcript is a reduced and idealised version of ‘what really happened’
and some measure of the problem can be gained.
Movements between modes, henceforth mode-switching, are very
common in the data and may be brought about either by teacher or
learner contributions, though teachers usually have prime respons-
ibility. The result is the same: the interaction becomes multi-layered
and more difficult to interpret and describe. In theory, any participant
in the discourse can say anything at any time and the ensuing
interaction takes sudden twists and turns in direction, which make
analysis difficult. In practice, this rarely occurs owing to the fact that
classroom interaction is goal-oriented. Understanding is gained by
considering the interrelatedness of the turn-taking, the fact that turns
do not occur in isolation, and by identifying pedagogic goals.
Previously, in Task 6.2, you analysed the following extract, 6.11. Here
the same extract is used to consider how movements between modes –
mode switching – are achieved. As we have already seen, we can
identify three modes here: materials mode; classroom context mode;
skills and systems mode. Mode switches are instigated by the teacher
on two occasions: in line 45, from materials to classroom context, and
in 47, from classroom context to skills and systems. These changes in
pedagogic focus are signalled in the language of the participants.
The extract begins in classroom context mode (lines 29–32) with
some open class discussion of types of oil used in cooking. The switch
to materials mode occurs in line 33, marked with the discourse marker
right and reference to the book. Lines 32 to 42 are all materials mode,
and the traditional IRF routine ensues; learners respond to the material
and have their answers confirmed in the teacher’s echoes in lines 39, 41,
43, 45. Turn-taking is tightly managed in relation to the material, little
interactional space is afforded and the role of the learners is to display
knowledge and understanding, which is evaluated by the teacher. The
second mode switch occurs in line 47, marked by laughter, which, here,

functions as a change of state token, allowing the discourse to change
focus, introduce a new topic and a new mode. At line 47, the interaction
moves from materials to classroom context mode, and is marked by the
mention of learners’ names. This ‘switch’ opens the door for a personal
contribution from a learner in line 52, who raises the sub-topic of
graduation dinners. Note here how the personalisation of the discourse
allows this learner to make an extended contribution about her
personal experiences.
The interaction is brought back to the main topic, ‘pot-luck suppers’
by the teacher’s question in line 58, preceded by the discourse marker
but. This transition takes the discourse back to materials mode in lines
58–62, when there is a very abrupt and unmarked switch to skills and
systems (‘what adjective do we use for people who don’t like that and
hate that and a lot of food they won’t eat. . . we call it people are very
. . . FUSsy’). This switch again invites a learner response in line 66,
which, as a vocabulary-related question, is consistent with skills and
systems mode. There is a request for clarification from the teacher in
line 68, followed by a self-initiated self-repair in line 69 and a
confirmation check in line 70.
It is important to note here that these switches occurred very rapidly
and over a short space of time. In order to ‘find their way’ in the
discourse, learners have to be alert to these extremely rapid switches
and learn to ‘read the signals’ so that interaction is maintained (see
Breen 1998); there are many instances in the data where such rapid
and often unmarked mode-switching does result in communication
breakdown that may or may not be repaired.
Extract 6.11
30 L: =and the other olive oil vegetable and made from maize
from corn=
31 T: =yes corn oil and yeah there’s many different types of
oil olive oil is the best and . . .
32 essential to Spanish omelette tortilla erm right so if you
were invited to one of Leslie’s
33 pot-luck suppers (referring to the book) erm right what
would you bring . . . along . . .
34 with you?
36 T: yes this is this Leslie the American woman imagine she
[invited you to a pot-luck
37 supper
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

42 L: =dessert
44 L: =or a starter
45 T: or a starter . . . yes so how would you know what to
bring?
47 T: that’s right yes so it’s it’s quite well-organised you can
(laughs) imagine writing a list
48 oh yes I’ll ask Yvette she can bring a starter . . . oh
Georgia a nice salad Italians are
49 very good with salads and then maybe a main course
er Haldoun and perhaps er the
50 Japanse can do some more some more main courses
so it would be very er very
51 well-organised=
52 L: =usually((2)) like sometimes there are like 3 main courses
((2)) like graduation in
53 sometimes we have parties and we have like er roast
chicken and another people
56 T: [(laughs)] oh dear well Georgia perhaps when you go
back to Italy perhaps you can
58 plenty of desserts and plenty of starters (laughs) but er
do you think it’s a good idea?
60 L: it’s nice because you don’t know what you’re going to
eat=
61 T: =it’s a surprise yes, yeah as long as you like everything
. . . I mean some people don’t
62 like certain things what adjective do we use for people
who don’t like that and hate
63 that and a lot of food they won’t eat. . . we call it people
are very . . . FUSsy fussy with
64 their food (writes on board) right so fussy that’s don’t
like vegetables, never eat er
65 pasta (laughs)=
66 L1: =excuse me how to say if er you ((2)) for example if you
try some burger on cormer

68 T: er . . . I’m not quite sure what you mean Yvette if you?
70 T: you mean street sellers . . . people selling food on the
street yes
Mode side sequences
A second and commonly found type of mixed mode occurs when there
is a brief departure from main to secondary mode and back to main
mode again. For example, the pattern may be classroom–context–
skills–and–systems–classroom–context, with classroom context the
main mode and skills and systems the secondary mode. Sequences like
this are henceforth referred to as mode side sequences. Side sequences
are a common feature of conversation and involve two speakers jointly
constructing and negotiating the dialogue, ‘feeling their way forward
together’ (Cook 1989: 54) and managing two topics and two exchange
structures. Equally, in the L2 classroom, participants progress
tentatively, each pursuing a particular agenda that is typically related
to that of the institution. Mode side sequences occur frequently, as
exempliﬁed the following extract.
In extract 6.12, a group of advanced learners is discussing life after
death. The main mode is classroom context, with a focus on personal
experiences. The mode side sequence occurs (lines 114 to 123), when
the learner’s turn (114) switches the mode to skills and systems. Once
the vocabulary question has been cleared up, the teacher brings the
discussion back to classroom context mode in (125). Note that the
impetus for a mode side sequence may come from either a learner or
the teacher, but that it is very often the teacher who brings the
discussion ‘back on track’, returning to the original mode. Here, there
is evidence of skills and systems mode in order to clarify a language-
related matter, but the departure is relatively brief and the main mode
is quickly re-established.
Extract 6.12
109 T: [no ok alright] . . . so Jan you want to live forever?=
110 L3: =yeah if money can afford it I will freeze body=
111 L: =ugh . . .
112 L1: what are you going to do? . . . frozen frozen you body?=
113 L3: =yeah=
114 L1: cyonics? . . .
115 T: =yeah it’s cry cry cryo[genics]
116 L1: [cryonics] cryogenics . . . no cryonics=
117 T: =oh is it? ok=
118 L1: =I think so I don’t know . . .
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

119 T: let me check it it might be in this one . . . (looks in
dictionary)
120 L: ((4)) . . .
121 T: cryogenics if you don’t freeze your body you freeze your
head isn’t that the way it is?=
122 L: =you can choose=
123 T: =oh really?=
124 L3: =it’s eh a ((2)) . . .
125 T: I see so if you don’t believe in religion=
126 L: =yeah=
In brief, then, we can say that mode side sequences occur when there
is a momentary shift from one (main) mode to another (secondary)
mode in response to a change in pedagogic goals. A number of such
mode side sequences have been identified, each following a similar
pattern:
• classroom–context–skills–and–systems–classroom–context
• materials–skills–and–systems–materials
• materials–classroom–context–materials
• managerial–skills–and–-systems–managerial
Both teacher and learners initiate mode side sequences, but the
responsibility for returning to the main mode lies with the teacher.
The extent to which the teacher is able to ‘keep on track’ and ensure
that learners do not become ‘lost’ is closely related to teachers’ ability
to move from one mode to another, adjusting language to the unfolding
text of the lesson. A mode side sequence is symbolised by an immediate
and obvious change in the talk-in-interaction, with different inter-
actional features and a different speech exchange system.
Task 6.5
Make a fifteen-minute recording of one of your classes, or, alterna-
tively, watch a video-clip of someone else teaching. Analyse the
discourse using SETT and identify which modes are used and why
and which interactures can be identified. Finally, in the light of what
you now know, how effective was the teaching in terms of its potential
for promoting learning?
TC

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
In this section, the framework for describing L2 classroom inter-
actional processes was presented and exemplified using extracts from
the data. The notions of mode switching and mode side sequences were
introduced, as were mode convergent and mode divergent teacher talk.
Summary
The descriptive framework presented in this chapter is designed to
enable teachers to access the interactional organisation of their classes.
The framework is intended as a means to an end rather than the end
itself, concerned to facilitate understanding, not code every interaction.
Consequently, it is representative, not comprehensive. Given the
uniqueness of the L2 classroom and the fact that every interaction is
locally produced, it is neither practicable nor realistic to propose that
an all-encompassing view of context can be derived, or that an
instrument is available that can accurately provide an emic perspective
of each interaction.
The framework relates pedagogic purpose to language use, enabling
teachers to identify ‘recurrent segmental patterns or structures’ (Drew
1994: 142) that can contribute to an understanding of what constitutes
appropriate teacher talk in a particular mode. This dynamic perspective
is intended to avoid the need for bland descriptive systems that adopt
an invariant view of L2 classroom interaction. By getting teachers to
relate their use of language to pedagogic goals and by examining
interactional features in each of the four modes, it is anticipated that
a greater depth of understanding can be gained in a relatively short
space of time.
In the data, four modes were identified and described according to
their pedagogic goals and interactional features. Managerial mode,
where the goal is the organisation of learning, features a single
extended teacher turn (usually an instruction or explanation) and an
absence of learner involvement. In materials mode, learning outcomes
are derived from materials-focused language practice: typically, the IRF
sequence dominates, making extensive use of display questions, form-
focused feedback and repair. Skills and systems mode typically follows
a similar interactional organisation to materials mode. However, turn-
taking and topic management may be less tightly controlled, and
pedagogic goals are not derived from materials, but from teacher and
learner agendas. In classroom context mode, on the other hand,
learners are allowed considerable interactional space; the focus is on
oral fluency, on the message rather than the forms used to convey it.
Each mode has its own characteristic fingerprint (Heritage and
Greatbatch 1991), specific interactional features that are related to
teaching objectives. While the characteristics identified in each mode
have a certain uniformity, there is also some degree of heterogeneity

(Seedhouse 1996) determined by the precise nature of the local context
and including factors such as the level of the students and the
methodology being used.
Modes are not static and invariant, but dynamic and changing.
There are movements from one mode to another, mode switching, and
between main and secondary modes (mode side sequences). Although
learners may initiate a switch, the responsibility for returning to the
main mode typically lies with the teacher. Switches from one mode to
another are marked by transition or boundary markers (right, now,
ok, etc.) with a corresponding adjustment in intonation and sentence
stress. A mode may last for one whole lesson or for much shorter
periods, with more frequent changes.
A teacher’s use of language may be mode convergent, where
pedagogic goals and language use are congruent, facilitating learning
opportunities, or mode divergent, where inconsistencies in pedagogic
goals and interactional features hinder opportunities for learning.
Occurrences of mode convergent or mode divergent verbal behaviour
may be purely ‘accidental’ in that teachers do not consciously plan to
coincide language use and teaching aims (however, see Chapter 7), or
more deliberate.

7 Classroom discourse as
reflective practice
Introduction
The aim of this chapter is to introduce readers to the idea that reflective
practice can be enhanced by making classroom discourse the main
focus of the reflection. We have already seen that interaction underpins
everything that takes place in a classroom. Here, I argue that there is
a compelling need to put classroom discourse at the centre of reflective
practice by advocating a more structured approach to the process of
reflection on practice. Why do I say this? The first, and perhaps most
important reason, is that the notion of reflective practice had become
a little tired and even dated. Without doubt, and as we shall see below,
the case for reflective practice is very strong; few would argue that if
we are to improve as teachers, we need to reflect on and change our
practices on a regular, ongoing basis. Practitioners and researchers alike
have been saying this for many years now. And yet, I argue, reflective
practice is only taking place in a limited way and with minimal impact
(see below). Perhaps reflective practice needs a makeover: using
classroom discourse as the focus might just be one way of doing this.
Second, and as we shall see below, few teachers are actually taught
how to ‘do’ reflective practice. Most teachers have experienced being
told to reflect on their practice, both on pre-and in-service teacher
education programmes. But very few, I suggest, are given the tools
and skills needed to do just that and to make reflection a part of a
teacher’s professional life.
Third, I am going to argue that in order to reflect on practice we
need to know exactly ‘what is happening’ when we are teaching. The
best way to understand teaching is to collect data and the best kind
of data is based on classroom interaction. By identifying a problem,
collecting data and then beginning the process of reflection and change,
I am very confident that reflective practice will not only occur on a
much larger scale, it will also have a greater impact on performance.
In the first part of the chapter, I present a critique of the notion of
teacher as reflective practitioner, while the second part deals with a
classroom discourse oriented approach to reflective practice.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

The teacher as reflective practitioner
The notion of reflective practice for teachers has been with us for
almost thirty years. First proposed by Schön (1983, 1996) as a means
of continuous professional development for educators and other
professionals, reflective practice is now regarded as a fundamental
principle of teacher education and teacher development. Essentially,
reflective practice entails thinking about a particular aspect of one’s
teaching and trying to change or improve it. The broadly-defined
process involves cycles of thoughtfully considering the effectiveness of
practice, taking steps to improve practice, and assessing the
effectiveness of those efforts, often in conjunction with other
professionals with complementary experience (Ferraro 2000). Hatton
and Smith (1995) described reflection as ‘a deliberate thinking about
action with a view to its improvement (34). It should be noted,
however, that, according to Sparks-Langer (1992), there does not
appear to be a single definition of reflective practice (henceforth, RP)
in the literature, a point that we’ll return to below.
Zeichner and Liston (1996) presented five features that they consider
to be key in reflective teachers. Reflective teachers are those who:
1 examine, frame and attempt to solve the dilemmas of classroom
practice;
2 are aware of and question the assumptions and values they bring
to teaching;
3 are attentive to the institutional and cultural contexts in which they
teach;
4 take part in curriculum development and school change efforts;
5 take responsibility for their own professional development.
The case for RP among a variety of practitioners has been advocated
for some time since Schön (see, for example, Eraut 1995; Edge 2001).
Many years ago, Dewey (1933) suggested that learning to teach must
include opportunities for new teachers to develop the capacity for
reflective action. More recently, in teacher education, it has been
acclaimed as a goal in many programmes (Hatton and Smith 1995).
Classroom discourse as reflective practice138
What does the term ‘reflective practice’ mean to you? Can you think of any
occasions where you have used reflective practice in your own teaching or as
a learner? What aspects of your teaching would you like to reflect on and how
might you do this?

Furthermore, Bartlett (1990) suggested that initial teacher training
involves more than skills and competencies, and that teachers need to
be equipped with the tools that will enable them to analyse their own
classroom practices and make adjustments to the teaching and learning
environment.
There is, then, plenty of evidence for the need for RP: it is,
essentially, seen as a ‘good thing’ and something that any structured
programme of professional practice should include. In initial teacher
education programmes, trainee teachers are told at a very early stage
that they must ‘reflect on their practice’; indeed, many written and
practical assignments include the ability to reflect critically on practice
as one of the main assessment criteria. What is less clear is the
precise meaning of RP and, more importantly, how to do it. Teacher
educators, I suggest, are quite content to tell their trainee teachers to
reflect on practice; few teach them how to do it. And yet what is needed
is for teachers in training to be taught how to reflect so that reflection
plus action becomes an integral element of their professional practice.
The only way that teachers can develop the skills needed to constantly
improve and refine their teaching is for them to become researchers
of their own practice. This is something that must be taught.
In the words of Russell (2005: 200), ‘The absence of any clear
agreement about what reflective practice is and how we recognise it
helps us to understand why it is not clear how to teach it.’ While
acknowledging that this is the current state of affairs, he then went
on to say that it ‘can and should be taught – explicitly, directly,
thoughtfully and patiently (204). Other writers have suggested how
RP might be taught. Alger (2006), for example, suggested that
reflection could be fostered through activities such as action research,
case studies, microteaching and reflective writing assignments, while
Park (2003) has identified reflective journals as potentially valuable
tools for developing reflective practitioners. He claimed that the
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Classroom discourse as reflective practice 139
Task 7.1
What aspects of classroom practice can we reflect on? Think of
three examples relating to teachers and three for learners. It’s probably
best to write each example as a question as I have done below. Then
decide HOW you would reflect and improve your practice in response
to each question.
Teachers Learners
How can I improve the questions How can I involve my students more
I ask? in my classes?
TC

journaling process did a number of things including stimulating critical
thinking, allowing teacher educators some insights into the minds of
the candidates, and helping the candidates to understand better the
process of learning.
Under the notion of second language teacher education, as distinct
from second language teacher training (Richards and Nunan 1990;
Wallace 1991), the focus of attention is on the need to enable teachers
to develop themselves. There is clearly more to teacher preparation
than skills training; teachers need to be equipped with the tools that
will enable them to find out about their own classrooms and make
adjustments (Bartlett 1990). In short, teachers must learn to change
their role from teacher to teacher-researcher, a logical extension of
what Wallace (1991: 8) terms ‘the applied science model’ of teacher
education (c.f. Schön 1983). This model of teacher education involves
a process of reflection in action whereby teachers and teachers in
training are involved in critical thought, questioning and re-appraising
their actions in the second language classroom. More recently, there
has been a call for the concept of reflection in action to be re-
interpreted as ‘reflection after action’ (Eraut 1995), since teachers
cannot reflect as they teach; instead, a cycle of action–reflection–further
action is preferred, with a slight distancing between each of the stages.
Under reflective practice, in addition to reflecting on present
classroom procedures, teachers adopt a retrospective stance and reflect
on past actions in an endeavour to increase their understanding of the
teaching/learning process (Wallace 1998). Competencies are acquired
through and by the participants who have an active role in their own
development, which in turn is based on two types of knowledge:
received knowledge, ‘the intellectual content of the profession’ (Wallace
1991: 14), including the specific knowledge (linguistic and pedagogic)
that language teachers need in order to perform their role; and
experiential knowledge, based on the experience gained in the
classroom and reflection on that experience. To use Wallace’s terms,
experiential knowledge is based on knowledge-in-action plus reflection
on that knowledge. Mann makes a similar distinction (2001: 58):
Received knowledge is the stuff of dictionaries and is more
verifiable. Experiential knowledge is not a matter of fact, but a
complex mix of feeling, thought and individual perspective.
Clearly, both types of knowledge are important, but it is the second
that is of most concern here since it rests on the assumption that
teachers can and should reflect on their practices and learn from them.
Central to the notion of experiential knowledge is collaborative
discussion, where thoughts and ideas about classroom practice are
articulated (Taylor 1985). To use Taylor’s term, articulation involves
the ‘shaping’ of ideas, the reformulation of concepts in a progression

towards enhanced understanding. Reflection on practice does not
occur in isolation, but in discussion with another practitioner, a form
of cooperative development, involving a ‘Speaker’ and an ‘Under-
stander’, who enhances professional understanding through dialogue
(c.f. Edge 1992).
Perhaps one of the best ways to promote reflective practice is for
teachers to engage in a process of action research. Action research rests
on the premise that teachers can and should investigate their own
classrooms (Cohen et al. 2007: 226). The starting point is the
identification of a problem; the process continues with data collection,
data analysis and finally outcomes are suggested. According to Carr
and Kemmis (1983: 220–221), action research can be defined as:
a form of self-reflective inquiry undertaken by participants in social
situations in order to improve the rationality and justice of their
own practices, their understanding of these practices, and the
situations in which the practices are carried out.
Action research involves identifying a specific classroom problem
and considering ways of solving the problem. The process of problem
posing is actually as important as problem solving and the process is
normally a collaborative one, involving discussion and dialogue with
a colleague or ‘critical friend’.
Cohen et al. (2007) identify eight stages in the process:
1 Identify the problem.
2 Develop a draft proposal.
3 Review what has already been written on the issue.
4 Restate the problem or formulate a hypothesis.
5 Select research procedures.
6 Choose evaluation procedures.
7 Collect data, analyse the data and provide feedback.
8 Interpret the data and make inferences.
The value and relevance of action research are self-evident: helping
teachers to focus on problems in their own classrooms and to identify
solutions is desirable from the position of both professional develop-
ment and learning. Perhaps the most important aspect of the action
research process is that it is data-led. Improvement in teaching per-
formance comes about through the collection and analysis of a small
amount of data. Here, ‘data’ are things like: recordings of a teaching
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

session; a set of test results; a completed observation schedule; minute
papers1
and so on. In short, collecting data means collecting evidence
that will help a teacher address a particular problem or issue. As
Johnson points out, this process of collecting and analysing data is
central to the professionalisation of teaching (1995: 29): ‘The more
research-driven knowledge teachers have, the better their teaching
performances will be.’
Under this view, ‘research-driven knowledge’ is data-driven and,
more importantly, based on data derived from a teacher’s own
classroom. According to Nunan (1989: 3), there are a number of
reasons why teachers might be interested in researching their own
classrooms:
• teachers have to justify educational innovations;
• teachers are constantly involved in intellectual and social change;
• becoming a teacher–researcher is a logical stage in the process of
professional self-development.
To this list is added another reason, that of ownership. I’d like to
suggest that ownership of the data is the single, most important factor
contributing to enhanced understanding and professional development.
Ownership of the data is, arguably, far more likely to result in a change
in teaching behaviour since one of the stages in getting teachers to
modify their teaching behaviour is for them to experience the change
(Thornbury 1996), a point echoed by Harmer (1999: 5), ‘It’s no good
talking about or imposing change [. . .], you have to let the changes
experience what you are talking about first.’ Because teacher-
researchers are both the producers and consumers of their research
(Kumaravadivelu 1999), since they own the data and are responsible
for effecting changes to classroom practices, the process is more private
and, arguably, less intimidating. The concern is to enhance
understanding of local context rather than generalise to a broader one.
Teachers engaged in action research are concerned to enhance their
understandings of the local context: a specific class, teacher or group
of students. As we have already seen, the process entails data collection,
analysis, reflection and discussion. Enhancing understandings of a
specific problem in a particular context rests on the following three
assumptions that will be dealt with in turn in the remainder of this
section:
• the research takes place in the classroom;
• teacher-researchers reflect and act on what they observe;
• understanding emerges through dialogue.

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Assumption 1: the research takes place in the
classroom
The kind of research we are describing here is small-scale, localised,
context-specific and private. This type of research contrasts starkly
to the kind that dominates second language teaching. According to
Nunan, much of the research that goes on in second language teaching
is deemed irrelevant, resulting in a ‘wedge between researcher and
practitioner’ (1996: 42). One of the reasons is that teachers’ voices are
often unheard, replaced by the researcher’s perspective; another is that
much research is still conducted under experimental conditions in class-
rooms created for research purposes (ibid.). This startling revelation
suggests an urgent need for research that is located in ‘ordinary’
classrooms, conducted by teachers for their own ends; understanding
and professional development can only be enhanced when the process
of inquiry is carried out in situ, in the teacher’s natural environment.
This was termed ‘ecological research’ by van Lier, drawing comparisons
with any natural environment in which the slightest change in one
sub-system will impact other systems (2000: 11):
Ecological research pays a great deal of attention to the smallest
detail of the interaction, since within these details maybe contained
the seeds of learning. The reflective teacher can learn to ‘read’ the
environment to notice such details.
The main attraction of this view of reflection in action is that teachers
work very closely with the data they collect in their own context, their
own ‘environment’, to use van Lier’s term. The understandings gained
Task 7.2
Consider each of the following scenarios. For each situation, say why
this might be a problem, then suggest what kind of data you would
collect to help you make improvements to your practice:
• Students refuse to speak English when working independently.
• Your instructions never seem to ‘work’ and are always mis-
understood.
• Your teaching is dominated by question–answer routines.
• Your students constantly do badly in the end of unit tests you set
them.
TC

from working with the detail of the data belong to teachers. Public
ownership does not come into question, neither does generalisability;
research is a process of inquiry, conducted by the teacher for the
teacher. The main advantage is that there is a unification of theory
and practice since the smallest details can be studied, changes
implemented and then evaluated (van Lier 2000). The main reason
for the potential for such microscopic analysis is the fact that the
research is located in a context that is both clearly defined and familiar
to the teacher-researcher.
Assumption 2: teacher-researchers reflect and act on
what they observe
The second assumption is very much in tune with the broad philosophy
of reflective practice: reflection results in a change in practice, based
on evidence in the form of data. Ideally, any change in practice should
benefit students. As mentioned previously, the first step in the process
is to identify a problem: the very act of ‘posing problems’ and coming
to understand them is, in itself, developmental. Problems may or may
not be solved; the real value lies in discussing options and considering
possibilities. Reflection and action result in a kind of ‘emergent
understanding’, an ongoing process of enhanced awareness.
For most second language teachers, this view of class-based research
might appear somewhat daunting – teachers are not automatically
equipped with classroom observation skills and may know even less
about how to process and analyse data. Not only are L2 teachers
normally too busy to take on additional duties such as data collection
and analysis, few have actually been trained in class-based research
techniques. The other issue is that many of the research methods
currently available have been designed for and by researchers, not by
teachers. There is a real paucity of research tools that can be used as
part of a reflective practice cycle and that do not require enormous
amounts of time and energy to master. My aim in the next section is
to demonstrate how classroom interaction data can be collected and
analysed quickly and easily and have an immediate impact on
understandings and teaching performance.
Assumption 3: understanding emerges through
dialogue
Action research has always been regarded as a collaborative process
involving dialogue (Kemmis and McTaggart 1992; Winter 1996;
Zuber-Skerritt 1996). Dialogue is a crucial part of the reflection–
action–further–action cycle, since it allows for clarification, questioning
and ultimately enhanced understanding. Conversation is the means by
which new ideas are expressed, doubts aired and concerns raised

(Wells 1999). It is through talk that we gain fuller understandings; I
can think of many occasions when a discussion with friends or
colleagues has helped me understand a new concept or a new idea
more fully. I am sure that readers will agree that a conversation is
often the best means of accessing new ideas and gaining closer
understandings of complex processes such as teaching.
From a theoretical perspective, socio-cultural views of learning are
very helpful here (see also Chapter 3). The main ‘message’ of this view
of learning is that all human development is underpinned by language,
often talk. Quite simply, if we wish to develop or improve in any aspect
of our lives, one of the first steps is to talk about it. New ideas and
new knowledge are ‘appropriated’; in other words, we gain ownership
of them. This process occurs through talk. Often, through a dialogue
with, possibly, a more experienced professional, we are ‘scaffolded’
or assisted in making sense and acquiring deeper insights of complex
phenomena. Consider this extract of interview data in which the
researcher (R) is talking to a teacher (T) about her use of teacher ‘echo’:
Extract 7.1
T: I was struck by how much echoing I did before and sometimes
there was a justification for it . . . but a LOT of the time . . . it
was just echo for the sake of echo so I was fairly consciously
trying NOT to echo this time.
R: And what effect did that (reduced echo) have on the interaction
patterns or the involvement of learners in the class, did it have
any effect that you noticed?
T: I think that it made them more confident perhaps in giving me
words because it was only going to come back to them if the
pronunciation WASn’t right rather than just getting ((1)) straight
back to them. When you’re eliciting vocabulary if they’re coming
out with the vocabulary and it’s adequate and it’s clear, there’s
no need for you to echo it back to the other students . . . you’re
wasting a lot of time by echoing stuff back.
(Teacher 1, interview 1)
Here we see very clearly the value of dialogue in promoting closer
understandings. The teacher is reflecting on her use of ‘echo’, the
repetition of student contributions – a common feature of classroom
discourse. Her realisation that echo can become a kind of habit (‘echo
for echo’s sake’) is probed by the researcher who asks about the effect
of echo on learner involvement. The teacher’s response is quite
revealing: she says that reduced echo makes learners more confident
and that a lot of echo is unnecessary. Arguably, this realisation may
not have occurred without an opportunity to discuss echo and reflect
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

on its effects. The researcher’s question allows her to think about her
language use and give reasons, possibly for the first time. It is this kind
of ‘light bulb moment’ that professional dialogue can create. Through
talk, new realisations and greater insights come about and get their
first airing. Dialogue can establish ‘proximal processes’, or contexts
that create opportunities for learning potential, which, according to
Bronfenbrenner and Ceci (1994: 578), ‘offer development growth in
terms of:
• differentiated perception and response;
• directing and controlling one’s own behaviour;
• coping successfully under stress;
• acquiring knowledge and skill;
• modifying and constructing one’s own physical, social and symbolic
environment.’
Again, using a socio-cultural construct, teachers are ‘scaffolded’
through their ‘Zones of Proximal Development’ (ZPD) to a higher
plane of understanding through the dialogues they have with other
professionals (van Lier 1996). Under this perspective, reflection and
action alone are insufficient means of allowing professional develop-
ment – scaffolded dialogues, where issues are clarified and new levels
of understanding attained, are central to reflective practice.
Developing reflective practices through SETT
In the last chapter, I presented the SETT framework, designed to enable
teachers to evaluate and analyse their interactions with students as a
means of improving classroom practice. The final task, Task 6.4, of
Chapter 6 invited you to make a short recording of your teaching and
analyse that recording using SETT.
In this section, I consider how the SETT framework might be used
as a tool to promote and foster reflective practice. The data used in this
section are taken from interviews with teachers who have used the
Task 7.3
Think of a time in your professional life when a conversation with a
friend or colleague helped you to understand something better. What
was the effect of the dialogue on your professional practice and what
changes did you make to your teaching, if any?
TC

framework to evaluate and improve their teaching. To summarise the
previous section in this chapter, the central argument here is that if
teachers are to become effective reflective practitioners, three conditions
are necessary:
1 Teachers need a tool, such as the SETT framework, to help focus
attention and direct reflections to specific features of their teaching.
2 Teachers need an appropriate metalanguage with which to discuss
their reflections. An appropriate terminology that is understood by
all parties is central to this.
3 There needs to be dialogue. Teachers engaged in reflective practices
should have an opportunity to discuss their reflections with a
colleague or other professional in order to make sense of their self-
observations.
The SETT framework (see Appendix A) was designed to help
teachers gain closer understandings of the complex relationship between
language, interaction and learning. It can be seen as a tool that promotes
and guides reflection and that, when accompanied by dialogue, will
result in changes in classroom practice that will, ultimately, benefit
learners. Similar frameworks have been advocated by other researchers
with an overall goal of making classrooms more dialogic and more
engaging for learners (see, for example, Mortimer and Scott (2003) and
Alexander (2008)). The SETT framework is intended to be used by
teachers evaluating their own language use: it therefore offers an emic,
insider perspective designed to facilitate reflection and action.
In the original study (2006), users of SETT made short (fifteen-
minute), ‘snapshot’ recordings of parts of their classes and analysed
these recordings using the framework. This entailed listening to the
tape twice: once to identify the four modes (see Table 6.1, Chapter 6),
then a second time to identify and make a note of specific interactures
such as teacher echo, direct repair, scaffolding, etc. This process was
repeated over a period of time in order to allow emergent under-
standings of classroom interaction to develop. From this dual analysis
(modes plus interactures), it is then possible to produce a kind of profile
of one’s teaching, taken from a purely interactional perspective. This
allows teachers to answer questions like these:
• Which modes seem to occur most frequently in my classes?
• How mode convergent (see Chapter 6) is my teaching?
• Are some interactional features (display questions, for example)
more common than others?
• How much interactional space do I give to my learners?
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

• Are there any particular features of the interaction in my classes
that I need to change or improve?
And so on. . .
Over time, and with repeated used, teachers gain in their classroom
interactional competence (see Chapter 8) and become much better
at promoting more dialogic, engaged interaction in which learners
really play a role. By completing the SETT reflective practice cycle in
Figure 7.1 on p. 149 and by engaging in a structured process of focused
reflection, action and further reflection, I suggest that improvements
to practice will occur naturally and relatively easily. Teachers become
genuine researchers of their own practice: collecting and analysing data,
reflecting and modifying teaching behaviour, discussing this process
with a colleague, collecting more data, and so on. The SETT reflective
practice cycle is summarised below; in the remainder of this section,
I present and analyse interview data from step 5 of the cycle as a means
of demonstrating this process in action. Pseudonyms have been used
throughout to protect the identities of the participating teachers.
In the first example, extract 7.2 below, the focus of the discussion is
‘wait time’, one of the interactional features that the teacher, Nick, had
chosen to discuss. For Nick, there is a growing realisation of the value
of wait-time in whole class open discussion (I just gave them whatever
time they needed). He comments on what actually happens following
a teacher prompt (they’re processing the question [. . .] and they HAVE
to literally look into their own minds and do they have an experience
which relates to the question) and makes the interesting observation that
for some students, this takes more time and they need to be given that
time (the wait-time is ALways more extensive for them).
It is apparent from Nick’s comments that this is the first time he
has been in a position to actually think about wait time as an important
phenomenon and one that teachers need to incorporate into their
teaching. Arguably, using the SETT procedure allowed him to analyse
this aspect of his teaching in far greater detail and make changes by
increasing wait-time where necessary. A simple change in practice
maybe, but it is one that is likely to have a huge impact on the kind
of interaction that occurs and a change that is likely to open up space
for learning. Note too in extract 7.2 how the interviewer (I) plays an
important role in guiding the discussion and in helping Nick to clarify
his own thinking around a particular issue.
Extract 7.2
T: I just found it was very enjoyable and the feedback, like
extended wait-time. Lots of GAPS here where you think there’s
nobody replying and then they suddenly come in

I: Was that conscious or was that just something . . .?
T: No I deliberately because I know that the far-easterners have
problems speaking and therefore I gave them I just gave them
whatever time they needed you know. In some cases they’re
processing the question and they’re processing the information
and they HAVE to literally look into their own minds and do they
have an experience that relates to the question. And this is the
case I think particularly with Roy with Yung rather and Jang
who are Korean I think the wait-time is ALways more extensive
for them.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
1. Record fifteen minutes of classroom interaction.
2. Analyse using SETT – identify modes.
3. Analyse using SETT – identify interactional features.
4. Evaluate the interaction and identify areas for attention or change.
5. Discuss with a colleague.
6. Make another recording and repeat steps 2–5.
Figure 7.1 SETT RP Cycle

In extract 7.3, we can see how the same teacher actually
incorporates extended wait time into his teaching. Here, we see Nick
trying to elicit a particular piece of vocabulary (shopping centre). Note
the three-second pauses between questions. This use of extended wait
time and the accompanying strategy of using a different question each
time leads to an extended learner turn in 3. Again, arguably, it is this
use of extended wait time that results in a longer response than the
one that might be expected by the teacher. The learner not only
provides the piece of vocabulary, she also comments on why people
go to shopping centres (it’s very convenient). The point I am making
here is that Nick is able not only to articulate and discuss a particular
aspect of his teaching (the use of extended wait time), he is able also
to demonstrate this practice while teaching.
Extract 7.3
1 T: what are the things that we have today with shopping?
what do we have today? (3)
2 what do we call these places today? (3) what’s the word
we’re talking about? (3)
3 L: shopping centres because you just go the shopping
centre you can buy everything its
4 very convenient
Under a dialogic approach to RP, a process of collaborative meaning-
making occurs, as evidenced in the next extract, 7.4. Here, teacher
and interviewer jointly construct understandings through their talk and
in relation to a particular tool. Notice how the interactants ‘feel their
way’ in the discourse, seeking and giving clarification and trying to
apply a new metalanguage to help them advance their understandings.
In line 1, for example, there is a request for clarification by this
teacher, Jill, relating to teacher echo. This is answered in lines 2–5, in
which the interviewer attempts to offer alternatives to echo (requests
for clarification and direct repair). In lines 7–9, the interviewer clarifies
the distinction between direct repair and teacher echo. He also makes
the point that teacher echo can have positive consequences (to allow
the whole class to HEAR it) or negative ones (it can be just a habit).
Compare the role of the interviewer in lines 2–5, where there is a lot
of hedging, frequent use of the modal could, and the use of softeners
(I would have thought) with his role in 7–9, where he is more directive,
more teacher-like we might say. In lines 7–9, there is a sense that the
interviewer is actually teaching the difference between direct repair
and teacher echo. This could be in response to Jill’s comment (I wasn’t
quite sure about that), or it could be stemming from a perceived need
to clarify the terminology.

In the second part of this extract (lines 10–21), there is a justification
by Jill for her use of echo, possibly in response to the interviewer’s
comment it can be just a habit. In lines 10–12, Jill is able to rationalise
her use of teacher echo (so the rest of the class can hear), before moving
on to explain that, from her completed SETT grid, some of her
contributions were examples of teacher echo, some examples of direct
repair. There is evidence here that this teacher is ‘sorting out’ in her own
mind not only the differences in meaning between the various inter-
actional features, but also her reasons for using them at specific points
in the lesson. The use of dialogue and the framework are central to this
process: she is essentially uncovering the interactional decisions taken
and describing – probably for the first time – her ‘online decision-
making’. We’ll come back to this below. Central to the process is the
potential, brought about by the presence of the interviewer, to air her
ideas and clarify interactional strategies taken ‘in the heat of the
moment’, while teaching. Note the assurance given by the interviewer
in lines 16–19, a kind of affirmation that Jill’s use of language and online
decision-making were appropriate for that stage in the lesson.
Extract 7.4
1 T: Yes yes, repeating what the students say. Is that teacher
echo?
2 I: It’s teacher echo but it could also be a kind of
clarification in a way or you could
3 actually even interpret it as direct repair, couldn’t you
because if they say ‘busy’
4 (mispronounced) or something instead of ‘busy’ and
you just say ‘busy’ that’s more
5 direct repair I would have thought than echo.
6 T: Yes I wasn’t quite sure about that.
7 I: If what you’re doing is correcting their pronunciation,
then that’s direct repair, if what
8 you’re doing is just regurgitating what they’ve said that’s
echo and again echo can be
9 to allow the whole class to HEAR it or it can be just a
habit.
10 T: Yep. I mean I do actually with this class I do use quite
a lot of teacher echo simply
11 because they do tend to speak very quietly and mumble
sometimes and it IS really so
12 the rest of the class can hear. This was pronunciation
so ‘busy’ (referring to SETT
13 grid) so that should have gone up to direct repair,
however, THESE were teacher
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

14 echoes ‘noisy expensive interesting’ because the
pronunciation was ok, but then I
15 repeated them for the reasons I mentioned they were
mumbled.
16 I: And I think everyone needs to HEAR a contribution.
I think that’s a GOOD use of
17 teacher echo and I think the sort of that we all probably
have the slightly negative one
18 is when somebody says something and you just repeat
it for no apparent reason.
19 I think that’s the one to be more aware of.
20 T: So again perhaps in the upper level classes when the
pronunciation is good and
21 clear, then it’s not necessary to repeat.
In the SETT RP data, many of the comments made by teachers were
about justifying their ‘online decision-making’. By ‘online’, I mean the
decisions that teachers make while teaching; a decision to cut an
activity short, or allow more time, or to get students to finish some
writing at home. Essentially, good teaching is about making the right
decision at a particular moment – I would go further and argue that good
decisions are often related to the ways in which interaction is managed
at a particular point in time. In extract 7.5 below, for example, the
teacher (Barbie) is talking about scaffolding, perhaps for the first time.
Her comments indicate that she is trying to both understand for herself
and explain to the interviewer how scaffolding occurs in practice (I’m
going to re-formulate it [. . .] I’m going to give them a lot of examples
so that’s all scaffolding isn’t it?). Barbie goes further and explains that
scaffolding occurs more in skills and systems mode because this is the
mode where the main focus is the language itself (it’s so focused on
language). What is actually happening in these, and other, extracts then
is that teachers are reflecting through the dialogues they have with the
interviewer, based on an earlier analysis of their own interactions with
students. I would suggest that this is a far more effective means of
promoting RP than simply asking people to reflect on their practice. Not
only are teachers able to discuss particular aspects of their teaching, they
are also able to give reasons for a particular strategy and make
observations about its appropriacy at a given moment.
Extract 7.5
I: Is scaffolding something you think you do more of in that type
of mode for example you’re in a skills and systems mode here.
Do you think it’s something that happens more in some modes
than others or is it maybe too difficult to say at this stage?

T: My first feeling would be yes because it’s so focused on
language that anything they give me that might not be correct
and not clear then I’m going to re-formulate it or anything they
don’t understand I’m going to give them a lot of examples so
that’s all scaffolding isn’t it?
(Teacher 3, interview 2)
As an alternative to simply talking about classroom interaction in
a semi-structured interview, it is also possible to get teachers to make
a video-recording of their teaching and discuss the recording. This
procedure, known as stimulated recall, has the immediate advantage
of allowing both interviewer and teacher to watch something together
and comment on it together. It is a very powerful research tool and
an excellent means of raising awareness about specific features of the
interaction. Stimulated recall was used in the SETT RP cycle as an
alternative to simply discussing a teacher’s performance. Teachers
were asked to make a video recording of part of a lesson, complete
the SETT framework and then discuss the playback with a colleague
or critical friend. The extracts that follow are taken from the stimulated
recall interviews. The classroom interaction is presented on the left
side of each extract; the teacher’s commentary on the right side.
In extract 7.6, the teacher, Joe, is commenting on an early part of a
lesson in which he is setting up a task. We can see from the interaction
that this is clearly managerial mode and the teacher’s goal is to locate
the activity in time and space (we’re going to be reading; last Christmas
we did a quiz). He is providing a context for the activity and helping
learners to understand what comes next, an instruction and
organisation of the groups. Joe comments on the mode (managerial) and
interactional feature (extended teacher turn). He then offers an
evaluation of the final part of the extract, the question, which, in Joe’s
own terms was rather ‘vague’ and ‘problematic’. Although we do not
know exactly what happened next, there is a suggestion by Joe that the
wording of his question created problems later on because it was too
vague. His comments highlight a problem that we are all only too aware
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Task 7.4
How might you use this approach (stimulated recall) to collect data
on your own teaching? What aspects of your classroom language
might you focus on using this approach and what problems do you
foresee in using it?
TC

of: the need for clarity when setting up tasks and giving instructions.
Joe’s reflections and self-evaluation, by pointing at specific features of
the interaction, enable a much finer grained analysis, both for us, as
outsiders, and for Joe commenting on his own teaching.
Extract 7.6
In the second stimulated recall extract, 7.7, we see a continuation
of Joe’s lesson in which he elicits information about the various pop-
stars. As Joe’s own comments reflect, he is asking genuine or referential
questions here in which there is a real exchange of information. As
we can see from the interaction, in many respects the extract resembles
casual conversation. Turn-taking is rapid (indicated by latching =) and
there are requests for clarification (e.g. from?) and acknowledgement
tokens (uh uh). It is not immediately evident who the teacher is: this
could be two friends chatting and roles are pretty symmetrical. Perhaps
one of the things Joe could have mentioned in his commentary is the
fact that there are opportunities here for extending learner contribu-
tions and for eliciting further information. We might even argue that
learning opportunities, in the form of speaking practice, have been
missed in this short extract. Nonetheless, what this and other extracts
show quite clearly is the relationship between classroom interaction
and teachers’ comments on it; this kind of data is both highly revealing
and highly suited to the promotion of reflective practice.
T: this morning we’re going
to be reading about and
talking about er music
and pop stars and famous
people. Last Christmas
we did a quiz do you
remember and it was a
picture quiz and some of
the pictures were people
from your countries. Erm
let me see could you work
with Ben and could you
three guys work together
(T organises groups). If
you recognise any of
those people who they
are and what you know
about them and then what
kind of lifestyle do you
think these people have?
So here I’m setting up the activity
so this is managerial and an
extended teacher turn. So this is
me setting up the activity, more
managerial.
It’s a very sort of vague question
which is problematic later on but
I thought this was . . . I want them
to speak to each other but it was
kind of problematic, maybe it was
too vague . . .

Extract 7.7
In the final stimulated recall extract 7.8, we see another teacher,
Mary, clarifying a piece of vocabulary, ‘discographics’. Our analysis
of the interaction reveals a number of interesting features:
• In 1, L1 comes up with an ‘invented’ piece of vocabulary,
‘discographics’, which is immediately met by surprise by Mary in 2.
• L1 tries to explain (in 3) and encounters some perturbation,
indicated by self-initiated self-repair and a four-second pause, which
Mary ignores, preferring to let L1 struggle a little longer.
• In 4, Mary interrupts L1 (indicated =) and seeks clarification,
offering an acknowledgement of L1’s previous contribution (‘like
you said’). Mary also scaffolds a more ‘precise’ term, offering ‘the
music business’ as a more appropriate phrase for ‘discographics’.
• In 5, it is apparent that L1 is not satisfied with this attempted
clarification, as indicated by her two questions, both suggesting
some doubt and confusion.
• Mary again interrupts (in 6), possibly preventing a fuller explanation
from L1 and possibly causing further confusion.
Mary’s own self-reflections on her data are equally interesting. Her
insights offer a detailed analysis of a repair strategy that may have
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
(Teacher stops the activity)
T: so you can help me
because I don’t know
some of these people er
who’s that who’s that
person in picture A?
L: Mei Chung
T: Mei Chung. What does
she do?
L: she’s a pop singer=
T: =she’s a a pop singer
. . . from?
L: from Taiwan=
T: =Taiwan is she famous in
lots of countries?
L: I think Taiwan and Hong
Kong=
T: =uh uh and in Malaysia
too?
L: yes
I’m asking referential questions
here because I didn’t really know
the answer to some of these. They
just told me. These are genuine
questions.

backﬁred and caused more confusion. She is able to rationalise the
whole process and take stock of the different courses of action taken,
and alternatives rejected (I was going to say it’s a false friend but I
decided not to because I thought that might confuse her). Mary is also
able to accept that she may have understood L1’s explanation and
that she possibly could have allowed more time, further evidence of
the interruption mentioned above. By her own admission, and as
evidenced in 5 (see above), there was some uncertainty about the
outcome of this repair being successfully achieved. There is doubt both
in Mary’s comments (there was still a doubt in my mind. . . .), and in
the questions asked by L1 (the music business? what is the name of
industry?).
Extract 7.8
(The teacher is eliciting vocab
items and collecting them on
the board. Learner 1 is trying
to explain a word)
1 L1: discographics=
2 T: =ooh what do you
mean?
3 L1: the people who
not the people the
(4) the business
about music
record series and=
4 T: =is this a word
you’re thinking of
in Basque or
Spanish in English
I don’t know this
word ‘disco-
graphics’ what I
would say is er
(writes on board)
like you said ‘the
music business’=
5 L1: =the music
business? what is
the name of of er
industry?=
6 T: =the music
industry as well it’s
actually better
I was going to say it’s a false friend
but I decided not to because I
thought that might confuse her . . .
maybe I misunderstood her now
when I look back at it . . . I under-
stood at the time that she meant that
this was a particular industry but
maybe she meant a business . . . but
I wasn’t prepared to spend a long
time on that because it didn’t seem
important even though there was still
a doubt in my mind . . .

It is clear, I believe, from these extracts, that stimulated recall is a
potentially very powerful approach that has much to offer reflective
practice, especially when the focus of the reflection is classroom
interaction. Even without the transcripts, much can be learned by
participants and it is a methodology that brings together very nicely
the various elements that I have argued are necessary for reflective
practice to work effectively, namely, a tool, some data and dialogue.
Stimulated recall is relatively easy to organise, is inexpensive and
unobtrusive; yet its potential for influencing professional development
is enormous.
Summary
In this chapter I have considered how reflective practice (RP) might
be realigned and improved if the focus of the reflection is classroom
interaction. By analysing extracts of classroom discourse and using
them for reflection, I have argued that teachers will learn how to reflect
on their practice in a structured and systematic way. Moreover, a
classroom discourse focused approach to RP allows the key elements
of the process to be addressed: first, a problem is identified; then data
are collected (here, in the form of classroom recordings); data are then
analysed using an appropriate tool and there is some reflection; finally,
new understandings and changes to practice emerge through dialogue
and discussion. The entire cycle is then repeated.
Using interview data, we have seen how emergent understandings
are collaboratively achieved; teacher and interviewer co-construct
meanings and uncover the detail of interactions in a bid to improve
practice. Taking this a little further, we have seen how stimulated recall
has much to offer RP in terms of its potential for providing a fine-
grained analysis of classroom discourse. The main point of this chapter
has been to demonstrate how teachers can become effective researchers
of their own practice and acquire some of the skills and tools that
they will be able to use in their professional lives. A classroom discourse
oriented approach to RP, I suggest, is more likely to result in
sustainable professional development and enable teachers to really gain
close understandings of the contexts in which they work.
Note
1 Minute papers are short, written evaluations by students on a teacher’s
teaching. They are quick to complete (hence ‘minute paper’) and give useful
feedback on specific teaching sessions.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

8 Classroom interactional
competence
Introduction
In this chapter, I present and develop the notion of classroom inter-
actional competence and consider how it can be characterised in
different contexts. Classroom Interactional Competence (CIC) is
defined as, ‘Teachers’ and learners’ ability to use interaction as a tool
for mediating and assisting learning’ (Walsh 2006: 132). It puts
interaction firmly at the centre of teaching and learning and argues
that by improving their CIC, both teachers and learners will
immediately improve learning and opportunities for learning. Building
on many of the ideas put forward in this book and elsewhere, I aim
to show how a better understanding of classroom discourse will have
a positive impact on learning, especially where learning is regarded as
a social activity that is strongly influenced by involvement, engagement
and participation; where learning is regarded as doing rather than
having (c.f. Sfard 1998).
The chapter falls into three sections. In the first, we look at the
work that has been completed on Interactional Competence, a con-
struct that has existed for more than twenty years and yet which
continues to attract a great deal of attention. In the second section, I
characterise classroom interactional competence (CIC), using data
extracts to examine the strategies open to both teachers and learners
to enhance interaction and improve opportunities for learning. In the
final section, we consider how teachers and learners might, in a very
practical sense, develop their own classroom interactional competence.
What do you understand by the term ‘interactional competence’? What is the
difference between, for example, interactional competence, communicative
competence and spoken fluency?

Interactional competence
It is apparent when studying spoken interaction that different speakers
have different levels of competence and varying abilities to express
their ideas and achieve understanding. This is true both in and outside
classrooms, of native and non-native speakers, regardless of their
language proficiency. Put simply, some people seem to be better able
to communicate than others, while some people seem to have difficulty
in conveying the most simple meanings. If we put this in the context
of the second language classroom, the situation becomes both more
complex and less understood. We’ll return to this in the next section.
Much of what happens in language classrooms, I suggest, is con-
cerned with individual performance rather than collective competence.
In other words, we, as teachers are constantly evaluating and assessing
our learners’ ability to produce accurate, fluent and appropriate
linguistic forms. Think about your own teaching and/or learning con-
text. Your focus of attention will almost certainly be on the ways in
which learners produce correct forms, articulate a particular point of
view, etc. Think too about the ways in which we test spoken language.
Again, the emphasis is on an individual’s ability to produce correct
utterances, rather than to negotiate meanings or clarify a point of view
or idea. Speaking tests focus heavily on accuracy, fluency, grammatical
structures, range of vocabulary and so on. They rarely consider how
effectively a candidate interacts or how well a candidate co-constructs
meanings with another interlocutor. In short, the focus of attention is
on individual performance rather than joint competence.
There may be many reasons for this position, not least of which is
the fact that a solo performance is easier to teach and easier to test
than a joint, collective one. To produce materials and devise tasks that
focus on interaction is far more difficult than to devise materials and
activities that train individual performance. Although contemporary
materials claim to adopt a task-based approach to teaching and
learning, they do not, I suggest, train learners to become better
interactants. All attention is directed towards the individual’s ability
to produce accurate, appropriate and fluent utterances.
In the ‘real world’, of course, effective communication rests on an
ability to interact with others and to collectively reach understandings.
Interactional competence, then, is what is needed in order to ‘survive’
most communicative encounters. Being accurate or fluent, in them-
selves, are, I suggest, insufficient. Speakers of an L2 must be able to
do far more than produce correct strings of utterances. They need to
be able to pay attention to the local context, to listen and show that
they have understood, to clarify meanings, to repair breakdowns and
so on. All of this requires extreme mental and interactional ability,
the kind of ability that will not, arguably, be trained by taking part
in pair-work tasks or group discussions.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

The notion of interactional competence was first coined by Kramsch
(1986: 370):
I propose (. . .) a push for interactional competence to give our
students a truly emancipating, rather than compensating foreign
language education.
What Kramsch seems to be saying here is that much foreign
language teaching adopts what Cook (2001) terms a ‘deficit’ model,
where second language speakers are perceived as being in some way
inferior to first language speakers, and where the performance of
second language speakers is somehow measured against that of first
language speakers. Indeed, in language testing contexts, many of you
will have come across descriptors or assessment criteria that use a
wording like ‘shows native-like fluency of the language’ in speaking
tests and so on. Kramsch, by contrast, argues that a focus on
interactional competence allows us to concentrate more on the ability
of learners to communicate intended meaning and to establish joint
understandings. Essentially, interactional competence is concerned
with what goes on between interactants and how that communication
is managed. Rather than fluency, we are concerned with what
McCarthy (2005) terms confluence: the act of making spoken language
fluent together with another speaker. Spoken confluence is highly
relevant to the present discussion since it highlights the ways in which
speakers attend to each other’s contributions and focus on collective
meaning-making. It is also a concept that lies at the heart of most
classroom communication, where interactants are engaged in a
constant process of trying to make sense of each other, negotiate
meanings, assist and query, support, clarify and so on. We might say
that, both inside and outside the classroom, being confluent is more
fundamental to effective communication than being fluent.
Since Kramsch’s 1986 paper, many researchers have struggled with
the notion of interactional competence without really coming to a
convincing and workable definition. More recent references emphasise
the fact that interactional competence is context-specific and concerned
with the ways in which interactants construct meanings together, as
opposed to looking at features of individual performance that lie at
the heart of communicative competence. For example, consider the
differences between the interactional resources needed in a context
where the emphasis is on a transaction, such as ordering a coffee, to
the interactional resources needed to take part in a conversation.
Clearly, in the first context, a basic knowledge of English will allow
you to order a coffee with minimal interactional competence. In the
second, however, and in most classroom contexts, much more
sophisticated interactional resources will be required if you are to
Classroom interactional competence160

successfully compete for the floor, gain and pass turns, attend to what
the speaker has said, interrupt, clarify and so on. We can see, from
these two examples, that interactional competence is highly context
specific and related very closely to speaker intent and to audience.
In an attempt to identify specific features of interactional competence,
Young (2003) points to a number of ‘interactional resources’ including
specific interactional strategies, such as turn-taking, topic management,
signalling boundaries and so on. Markee (2008) proposes three
components, each with its own set of features:
• language as a formal system (including grammar, vocabulary,
pronunciation);
• semiotic systems, including turn-taking, repair, sequence organisa-
tion;
• gaze and paralinguistic features.
As Markee (2008: 3) says, developing interactional competence in
a second language involves learners ‘co-construct[ing] with their
interlocutors locally enacted, progressively more accurate, fluent, and
complex interactional repertoires in the L2’. While I agree with the
first part of this definition and acknowledge, indeed emphasise, that
interactional competence is a joint enterprise, I have some issues with
Markee’s focus on accuracy, fluency and complexity as indicators of
interactional competence. I would go further and suggest that a person
who has a high level of interactional competence is not necessarily an
accurate speaker. As I hope to show in the next section, there are
alternative ways of characterising interactional competence.
Young (2008: 100) offers this definition of interactional competence:
Interactional competence is a relationship between participants’
employment of linguistic and interactional resources and the
contexts in which they are employed . . .
Here then, Young focuses on the relationship between ‘the linguistic
and interactional resources’ used by interactants in specific contexts.
At this stage, it might be helpful to look at some data in order to
clarify some of the ideas we have discussed so far on interactional
competence.
In extract 8.1 below, students are working independently of the
teacher in what might be termed an oral fluency practice activity. They
roll a dice and discuss a topic corresponding to the number shown on
the dice. If they have already discussed that topic, they simply select
another one. What is immediately obvious from the extract is the
amount of interactional work that students engage in to keep the
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

discussion moving and on track. Students have been asked to talk
about pictures of people that they are close to and in line 1 the topic
is launched with a question by L1 and an extended response plus
justification by L3 in lines 2–3. L2 shows empathy towards L3, making
the point that L3 is very busy and has little time to ‘play’ (sic) with
her boyfriend, a point taken up by L1, who says that they can at least
live together as a means of improving their relationship. Note how
the word search in line 6 is dealt with very quickly by L4 in line 7,
allowing the discussion to continue and avoiding a potential
breakdown. L4 extends her contribution with an interruption (and
know each other). The four-second pause in line 9 indicates an
opportunity for a change of speaker or change of topic and this is
proposed by L3. However, the change of topic does not occur
immediately as L2 interrupts in line 10 with an anecdotal story about
registration and the fact that the students doing the registration are
able to know who is single.
L2’s interruption prompts L3 to conclude her discussion about her
boyfriend in lines 14–19 (he’s not very handsome, he’s clever and serious
about his work). L1’s overlaps are in support of what L3 is saying and
offer some kind of affirmation or approval of the comments made by
L3. Note too how laughter is used as a way of offering approval and
affirmation throughout this extract. Finally, in line 20, L1 brings the
discussion back to the original topic switch proposed by L3 in line 9,
where there was no switch. Note the extended ten-second pause at the
end of the extract, suggesting that the students are considering another
topic and that there is now going to be a switch.
Task 8.1
Before you read any further, look at extract 8.1 and comment on these
students’ ability to manage and direct the interaction, unaided and
with no teacher involvement. How would you evaluate this in terms
of the students’ oral fluency? Are there any examples of what you
might call interactional competence?
Extract 8.1
1 L1: do you bring his photo with you
2 L3: eh . . . yeah but we have only a few photos because
we get together (.)only one year or
3 so=
4 L2: =and your work was very busy so you have no time
to play with him [<L03> Yeah]
TC

In terms of the interactional competence demonstrated in this
extract, we can make a number of observations about the interactional
resources employed and their impact on the overall ﬂow and coherence
of the discussion:
• Turn-taking. It is apparent that all four students manage the turn-
taking very well and are able to interrupt, hold and pass turns.
Interruptions occur, but naturally and in a supportive way. There
are no major breakdowns and the discussion ﬂows well.
• Repair. It is interesting to note that even though errors do occur, they
are largely ignored. (c.f. lines 1, 2, 4 etc., all containing errors). This
is what Firth (1996) refers to as the ‘let it pass’ principle; in many
business contexts where English is used as a lingua franca, inter-
actants largely ignore errors unless an error causes a problem for
understanding. In extract 8.1, the main repair comes in lines 6–8,
where the word ‘relationship’ is needed in order to clarify meaning.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
5 L1: but I suppose that you must (.)leave some enough
money to(.)live with your boyfriend
6 and in this way you can (.)improve the (.)eh (3 sec
unintelligible) how do you say=
7 L4: =relationship=
8 L1: =relationship yes [<L04> and know each other]=
9 L3: =yeah I think I now him very well now (laughs) (4) well
lets talk another topic=
10 L2: =I remember one thing when they choose register in
Coleraine and they organiser
11 know you [reg . . . register . . . register]
12 L: ˪ are single girl and they don’t know you have a
boyfriend
13 LL: ((laugh))
14 L3: I think eh that is is humerous ok he is not very
handsome and not very but I think he
15 is very clever ehm and he [<L01> a lot like you] no
(laughs) I think he is clever than
16 me and (laughs) (3) he do everything very . . . [seriously
17 L1: [seriously
18 L3: yeah (3 sec unintelligible) eh and eh in some eh . . .
in some degree . . . eh I . . . admire
19 him (laughs)=
20 L1: =thats a good thing [<L> Yeah] lets change another
topic (10)

• Overlaps and interruptions. Note how overlaps and interruptions
occur frequently (for example, in lines 4, 8, 12, 16, 20), but they
are supportive and designed to ensure that the interaction flows
smoothly. These overlaps and interruptions are examples of what
McCarthy (2003) refers to as good ‘listenership’: they signal to a
speaker that she has been understood, that the channels are open
and that the communication is working well. Essentially, they ‘oil
the wheels’ of the interaction and help to prevent trouble and
breakdowns from occurring. As a deliberate strategy, overlaps give
vital clues to speakers that they are being understood and that
something is being communicated.
• Topic management. One of the key indicators of the coherence of
a piece of spoken interaction is topic management and development.
In extract 8.1, we can see how the main topic of ‘relationships’ is
introduced, developed and discussed at length despite one attempt
to switch topic in line 9 – which is ignored until much later in line
20. Interactants are genuinely engaged with the topic and succeed
in maintaining it for some time and from a range of perspectives.
In short, we can say that this is a good example of coherent
discourse in which all participants are concerned to engage with
and develop a topic to the full.
To summarise this section, I present, in a table, the main features
of interactional competence and compare them with some of the ideas
proposed under what is referred to as communicative competence (c.f.
Hymes 1972; Canale and Swain 1980):
As a final word and pointing towards the future, it is important
to recognise the important work that is underway at the time of
writing. Research projects such as English Profile, led by Cambridge
University Press, working with a number of partner institutions (www.
englishprofile.org), now acknowledge that there is much work to be
done in terms of characterising interactional competence as the ‘fifth
skill’ (in addition to speaking, listening, reading and writing). It is
almost certain that the results of this and related research projects will
have enormous implications for language teaching, language testing
and materials design. In the next section, we therefore consider how
interactional competence might manifest itself in the classroom.
Task 8.2
Based on the discussion so far and what you may have read, how would
you define interactional competence as it relates to your own teaching
and learning context? What are the specific features of interactional
competence that you would like your learners (or yourself) to acquire?
TC

Classroom interactional competence
Turning now to a conceptualisation of classroom interactional com-
petence (CIC), defined here as ‘teachers’ and learners’ ability to use
interaction as a tool for mediating and assisting learning’ (Walsh 2006:
132), the starting point is to acknowledge the centrality of interaction
to teaching and learning. In the same way that we have seen that
interactants display and orient to learning through interactions that are
co-constructed, they also demonstrate differing abilities to jointly
create discourse that is conducive to learning. CIC focuses on the ways
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Table 8.1 interactional competence versus communicative competence
Interactional competence Communicative competence
Emphasises the ways in which The focus is on individual differences in
interactants co-construct meanings competence and the fact that one of the
and jointly establish understanding. aims of learning a language is to move to
the next level of competence.
Includes both interactional and Emphasises the knowledge and skills
linguistic resources, but places needed to use language in specific
more emphasis on the way the contexts as opposed to knowledge of
interaction is guided and managed language as an idealised system.
through turns-at-talk, overlaps,
acknowledgement tokens, pauses,
repair and so on.
Is highly context specific: the Context is everything: what we say is
interactional competence required in dependent on who we are talking to,
one context will not always transfer to where we are, why we are talking, what
another. Different interactional we have to say and when this takes place
resources will be needed in different (c.f. Hymes, 1972).
contexts
Largely rejects individual performance Emphasises individual performance and
in favour of collaborative enterprise. recognises that this can and will change.
Less concerned with accuracy and Accuracy, fluency and appropriacy lie at
fluency and more concerned with the heart of communicative competence
communication; this means that and are also the measures used to
speakers must pay close attention evaluate it.
to each others’ contributions and
help and support where necessary.
Places equal emphasis on attending Focuses more on individual speech
to the speaker as producing one’s production than on the listener and
own contribution; listening plays acknowledgement of what has been said.
as much a part in interactional
competence as speaking.

in which teachers’ and learners’ interactional decisions and subsequent
actions enhance learning and learning opportunity. From the extracts
of data below, we will consider how teachers and learners display CIC
and discuss the implications of this for gaining closer understandings
of the relationship between L2 learning and interaction. The
assumption is that by first understanding and then extending CIC, there
will be greater opportunities for learning: enhanced CIC results in more
learning-oriented interactions.
Given the context dependency of interactional competence, we are
attempting here to identify some of the features of classroom inter-
actional competence (CIC). How are meanings co-constructed in the
unfolding interaction? What do participants do to ensure that under-
standings are reached? How do they deal with repair and breakdown?
More importantly, how does CIC influence learning? In what ways
are interactants able to create, maintain and sustain ‘space for
learning’? Space for learning refers to the extent to which teachers and
learners provide interactional space that is appropriate for the specific
pedagogical goal of the moment. It does not simply mean ‘handing
over’ to learners to maximise opportunities for interaction. Rather,
creating space for learning acknowledges the need to adjust linguistic
and interactional patterns to the particular goal of the moment. Again,
the emphasis is on promoting interactions that are both appropriate
to a particular micro-context and to specific pedagogic goals.
In language assessment circles, I have already mentioned that it is
now widely predicted that interactional competence will become the
‘fifth skill’ (see previous section). Given that interlocutors display
varying degrees of competence in their joint construction of meanings,
I am suggesting here that teachers and learners also need to acquire
a fine-grained understanding of what constitutes classroom inter-
actional competence and how it might be achieved. Not only will such
an understanding result in more engaged and dynamic interactions in
classrooms, it will also enhance learning.
In the data, there are a number of ways in which CIC manifests
itself. First, and from a teacher’s perspective, a teacher who demon-
strates CIC uses language that is both convergent to the pedagogic
goal of the moment and that is appropriate to the learners. Essentially,
this entails an understanding of the interactional strategies that are
appropriate to teaching goals and that are adjusted in relation to the
co-construction of meaning and the unfolding agenda of a lesson.
This position assumes that pedagogic goals and the language used to
achieve them are inextricably intertwined and constantly being
re-adjusted (Walsh 2003; Seedhouse 2004). Any evidence of CIC
must therefore demonstrate that interlocutors are using discourse that
is both appropriate to specific pedagogic goals and to the agenda of
the moment.

In extract 8.2 below, a stimulated recall methodology has been used
(see Chapter 7). On the left side is the actual classroom interaction
and on the right the teacher’s commentary on it. I have selected this
extract because I believe it shows very clearly that this teacher’s
pedagogic goals and the interaction used to achieve them are at one,
they are working together – we can say they are ‘mode convergent’
(see Chapter 6). Essentially, his comments on the right indicate quite
clearly why certain interactional decisions were taken. For example,
in the interaction, we see evidence of the teacher constantly seeking
clarification, affirming and re-affirming and helping the learner to
articulate a full response (in 5, 7 and 9). Each of these responses is
designed, according to the teacher to ‘get it flowing’ and ‘to reinterpret
for the benefit of the class’. Not only is he helping the learner to
articulate his ideas more clearly, he is helping the rest of the class
to understand what is being said. From his comments and the evidence
provided in this extract, we can say quite clearly that pedagogic goals
and the language used to achieve them are at one and that this teacher
knows why he has made certain interactive decisions. We might say
that he is demonstrating a level of interactional awareness here: his
actions and his comments both highlight this.
Extract 8.2
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
1 T: what was the
funniest thing that
happened to you
at school (1)
Tang?
2 S1: funniest thing?
3 T: the funniest
4 S1: the funniest thing I
think out of school
was go to picnic
5 T: go on a picnic?
So what happened
what made it
funny?
6 S1: go to picnic we
made playing or
talking with the
teacher more
closely because in
the school we
have a line you
know he the
Basically he’s explaining that on a
picnic there wasn’t this gap that there
is in a classroom – psychological
gap – that’s what I’m drawing out
of him. There’s a lot of scaffolding
being done by me in this monitoring,
besides it being managerial, there’s
a lot of scaffolding because I want to
get it flowing, I want to encourage
them, keep it moving as it were.
I’m clarifying to the class what he’s
saying because I know in an ex-
tended turn – a broken turn – and it’s
not exactly fluent and it’s not articu-
late – I try to re-interpret for the
benefit of the class so that they’re all
coming with me at the same time
and they all understand the point
being made by him

A second feature of CIC is that it facilitates interactional space:
learners need space for learning to participate in the discourse, to
contribute to class conversations and to receive feedback on their
contributions. Interactional space is maximised through increased
wait-time, by resisting the temptation to ‘ﬁll silence’ (by reducing
teacher echo), by promoting extended learner turns and by allowing
planning time. By affording learners space, they are better able to
contribute to the process of co-constructing meanings – something that
lies at the very heart of learning through interaction. Note that this
does not necessarily mean simply ‘handing over’ to learners and getting
them to complete pair and group work tasks. While this may facilitate
practice opportunities and give learners a chance to work inde-
pendently, it will not, in itself, necessarily result in enhanced learning.
The same point has been made by others (c.f. Rampton 1999).
What is needed, I would suggest, is a re-think of the role of the
teacher so that interaction is more carefully understood, and so that
the teacher plays a more central role in shaping learner contributions.
Shaping involves taking a learner response and doing something with
it rather than simply accepting it. For example, a response may be
paraphrased, using slightly different vocabulary or grammatical struc-
tures; it may be summarised or extended in some way; a response may
require scaffolding so that learners are assisted in saying what they
really mean; it may be recast (c.f. Lyster 1998): ‘handed back’ to the
learner but with some small changes included. By shaping learner
teacher and me
the student=
7 T: =so you say there
was a gap or a
wall between the
teacher and the
students so when
you=
8 S1: if you go out of
the school you
went together with
more (gestures
‘closer’ with
hands)=
9 T: =so you had a
closer relationship
[outside the
school]
10 L1: [yeah yeah]

contributions and by helping learners to really articulate what they
mean, teachers are performing a more central role in the interaction,
while, at the same time, maintaining a student-centred, decentralised
approach to teaching.
Extract 8.3 is taken from a secondary class in China, where students
are talking about visits to museums. This teacher successfully creates
space for learning by using a number of key interactional strategies:
1 Extensive use of pausing throughout, some of these pauses are quite
extensive (c.f. lines 1, 3, 6 and 17, for example). Remember that
typically teachers wait for less than one second after asking a
question or eliciting a response. Pausing serves a range of functions:
• It creates ‘space’ in the interaction to allow learners to take a
turn-at-talk.
• It allows thinking or rehearsal time (c.f. Schmidt 1993) enabling
learners to formulate a response (see lines 44 and 46 where a
teacher pause is followed by a learner pause).
• It enables turn-taking to be slowed down, helping to make
learners feel more comfortable and less stressed.
• Increased wait time often results in fuller, more elaborated
responses, as in lines 40ff and 46ff.
2 A lack of repair. Students make some mistakes in this extract (line
34, word order; line 46, verb form ‘wasting of time’). These are
ignored since they do not impede communication and do not fulfil
the teacher’s pedagogic goal here: elicitation and sharing of personal
experiences. In this type of micro-context, error correction is not
seen as being necessary and the teacher disregards errors since they
are not of central concern.
3 Signposting in instructions. This teacher twice calls for a choral
response (in lines 23 and 28). She marks her instruction and
signposts that she wants the whole class to respond (‘together’). This
is a useful strategy in a multi-party conversation like a classroom
where calling out and ‘ragged’ choral repetitions are very common.
4 Extended learner turns (in lines 46–47, for example). The teacher
allows learners to complete a turn and make a full and elaborated
response. Often teachers interrupt and close down space when
learners are attempting to articulate something quite complicated.
Here, she does the opposite and allows the student space in the
interaction to make a full and useful contribution.
5 Seeking clarification (lines 34–37). The teacher is not entirely
satisfied with the first response and insists on the insertion of ‘often’
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

to make sure that this contribution is as accurate as possible. This
is a good example of a recast (see above).
In addition to creating space, in the same extract we can see how
space can be ‘closed down’. One of the main causes of this is where
teachers make excessive use of echo, repeating their own or students’
contributions, sometimes with no apparent reason or need. Note that
there are two types of echo:
• Teacher–learner echo: where a teacher repeats a learner’s utterance
for the beneﬁt of the class (lines 8, 11, 20 and 22 for example).
This is helpful and ensures that a class progresses together and that
everyone is ‘in the loop’. It is an inclusive strategy, which ensures
that the whole class comes along together and that there is
commonality of understanding (see extract 8.2 where the teacher
comments on the need to ensure that the whole class is ‘coming
with me at the same time’).
• Teacher–teacher echo: where a teacher simply repeats her own
utterance almost like a kind of habit (lines 4–5, 14–15, 32–33, 44).
This serves no real function, arguably, and may impede oppor-
tunities for learning since the teacher is taking up learners’ space
in the dialogue. It may be used as a kind of defence mechanism
since silence can be quite threatening.
Extract 8.3
1 T: class begins (3) good afternoon everyone
2 SS: good afternoon teacher
3 T: sit down please (3) so our topic today is museums(.)
4 talking about museums (.) have you ever been to
museums
5 (1)? Have you ever been to museums (2)? Yes of course.
6 And what ↑kind of museums have you been to (4)
NAME?
7 S: (unclear)
8 T: The national museum ↑yes thank you very much and
how
9 about you NAME?
10 S: (unclear)
11 T: History museum (.) thank you very much (.) so as you
12 mentioned just now (1) you have been to ((puts
powerpoint
13 slides of museums up))many kinds of museums (.) but
(.)do

14 you still remember ↑when did you go to those museums
for
15 the last time (2)? When did you go there for the last
16 time? For example when did you go to the national
museum
17 (.) the last time (4)? ((gets microphone from another
18 student)). Thank you
19 S: er maybe several month ago
20 T: several month ago thank you ok how about you?
21 S: I think several years ago
22 T: several years ago. Ok ((laughs)). Thank you very much
(3)
23 ok actually can you tell me together do you often go to
24 museums?
25 Ss: no
26 T: No so what you said is just the same as what I read in
27 the newspaper the other day(.) would you please read
the
28 title of this piece of news together ((points to
29 powerpoint slide))
30 Ss: ((reading aloud)) why are young people absent from
31 museums?
32 T: thank you (.) what does it mean? (.) NAME what does
the
33 title mean?
34 S: (.) why young people don’t go to museums
35 T: they don’t go to museums very?
36 S: often
37 T: very often thank you very much (.) and (.)so actually
38 you are young people ↑why don’t you go to museums
very
39 often(3)? NAME
40 S: (2) er because erm there’s nothing in the museums that
41 er attracts us and er even the mus- things in museums
are
42 usually very (.) old
43 T: old thank you very much ok so nothing can attract you
44 ((writes on board)) (3) what else? What else?
45 why don’t you go to museums very often? NAME
46 S: (3)I think going to museums is a wasting of time because
47 because I’m not interested in those old-fashioned things
48 T: ok thank you very much so you’re not in↑terested in it
49 it’s not interesting right? ((writes on board)) not
50 interesting (5) ok.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

So far, we have seen that two important features of CIC are the
convergence of language use and pedagogic goals, and the need for
interactional space. A third feature or strategy entails teachers being
able to shape learner contributions by scaffolding, paraphrasing,
reiterating and so on, a point that has already been mentioned.
Essentially, through shaping the discourse, a teacher is helping learners
to say what they mean by using the most appropriate language to do
so. The process of ‘shaping’ contributions occurs by seeking clari-
ﬁcation, scaffolding, modelling, or repairing learner input. In a
decentralised classroom in which learner-centredness is a priority,
these interactional strategies may be the only opportunities for teaching
and occur frequently during the feedback move (c.f. Cullen 1998).
Elsewhere (see, for example, Jarvis and Robinson 1997), the process
of taking a learner’s contribution and shaping it into something more
meaningful has been termed appropriation; a kind of paraphrasing
that serves the dual function of checking meaning and moving the
discourse forward.
We turn now to a further consideration of how these features of
CIC manifest themselves in classroom data. The extract below is taken
from an adult EFL class where the teacher is working with an upper-
intermediate group of learners who are preparing to do a listening
comprehension about places of interest. There are a number of features
in the extract that show evidence of CIC:
Extract 8.4
1 T: okay, have you have you ever visited any places
↑outside London?=
2 L1: =me I stay in (.) Portsmouth and er:: in Bournemouth
3 T: [where’ve you been?
4 L1: [in the south
5 T: [down (.) here? (pointing to map)
6 L1: yeah yeah
7 T: ↑why?
8 L1: er my girlfriend live here and (.) I like this student
9 place and all the people’s young and a lot (.) er go out
10 in the (.) evening its very [good
11 T: [right
12 T: anybody else? (4) Have you been anywhere Tury?
13 L2: Yes I have been in er (.) Edinbourg ((mispronounced)),
14 (())=
15 T: =so here here ((pointing to map))=
16 L2: =yes er Oxford (.) Brighton (.) many places (())=
17 T: =and which was your favourite?=

18 L2: =my favourite is London
19 T: (.) ↑why?
20 L2: because it’s a big city you can ﬁnd what what you [want
21 T: [mmhh
22 L2: and do you can go to the theatres (1) it’s a very (.)
23 cosmopolitan [city
24 L: [yes
25 L2: I like it very much=
26 T: =do you all (.) agree=
27 LL: =yes (laughter)
28 T: ((3)) laughter)
29 T: has anybody else been to another place outside
London?
30 L: no not outside inside
31 T: (.)mm? Martin? Anywhere?
32 L3: =no nowhere=
33 T: =would you like to go (.) [anywhere?
34 L3: [yes yes
35 T: [where?
36 L3: well Portsmouth I think it’s very (.) great=
37 T: =((laughter)) cos of the students [yes (.) yes
38 LL: [yes yes
39 L3: and there are sea too
40 T: Pedro?
41 L4: it’s a (.) young (.) place
42 T: mm anywhere else? (3) no well I’m going to talk to
43 you and give you some recommendations about where
you
44 can go in (.) England (.) yeah
(Carr 2006)
We can ascertain from this context (and from the lesson plan
accompanying these published materials) that the teacher’s main
concern is to elicit ideas and personal experiences from the learners.
The corresponding talk conﬁrms this in a number of ways. First, there
is no repair, despite the large number of errors throughout the extract
(see, for example, lines 2, 8, 13, 36, 39), the teacher chooses to ignore
them because error correction is not conducive to allowing learners
to have space to express themselves. Second, the questions she asks
are often followed with expansions such as ‘why’? (see for example,
7, 19), which result in correspondingly longer turns by learners (in 8,
and 20). Again, we would claim that both the teacher’s questioning
strategy and the longer learner turns are evidence of CIC since they
facilitate opportunities for both engaged interaction and learning
opportunity. Third, we note that there are several attempts to ‘open
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

the space’ and allow for wider participation of other learners. This
occurs, for example, in 12 (anybody else plus a four-second pause),
in 26 (do you all agree?), in 42 (anywhere else plus a three-second
pause). On each of these occasions, the teacher is attempting to include
other students in the interaction in a bid to elicit additional
contributions. Again, her use of language and pedagogic goals are
convergent, ensuring that learning opportunities are maximised.
Other features that show evidence of CIC include:
• the use of extended wait time, pauses of several seconds (in 12 and
42) that allow learners time to think, formulate and give a response.
Typically, teachers wait less than one second after asking a question
(see, for example, Budd Rowe 1986), leaving learners insufficient
time to respond.
• the use of requests for clarification (in 3, 5, 15) that serve to ensure
that understandings have been reached. Not only do such requests
give important feedback to the students, they allow the teacher to
ensure that the other students are included by clarifying for the
whole class.
• Minimal response tokens that tell the other speaker that under-
standings have been reached without interrupting the ‘flow’ of the
interaction (see, for example, 11 (right), 21 (mmhh). Again, the use
of such feedback is further evidence of convergence of pedagogic
goals and language use.
• Evidence of content feedback by the teacher who responds to the
message and not the linguistic forms used to articulate a particular
message. In extract 3 above, for example, the teacher responds in
an almost conversational way to almost all of the learners’ turns.
She offers no evaluation or repair of learner contributions, as would
be the ‘norm’ in many classroom contexts. Instead, she assumes an
almost symmetrical role in the discourse, evidenced by the rapid
pace of the interaction (note the overlapping speech in 3–5, 33–35,
and latched turns in 14–18 and 25–27).
In the same extract, there are a number of features of CIC that we
can highlight from a learner’s perspective. First, there is recognition
on the part of L1 that the appropriate reaction to a question is a
response, the second part of that adjacency pair, as evidenced in lines
2, 4, 6, 8. Not only does L1 answer the questions posed by the teacher,
he is able to recognise the precise type and amount of response needed,
ensuring that his contributions are both relevant and timely. He is also
sufficiently competent to appreciate that a question like ‘why’ in line 7
almost always requires an extended response, which he provides in 8.

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
His CIC is sufficiently advanced to appreciate that the teacher’s focus
here is on eliciting personal experiences – while his responses are
adequate and appropriate, they are certainly not accurate; yet this is
of little or no concern given the pedagogic focus of the moment. This
learner has correctly interpreted the teacher’s question as a request for
further information where accuracy is less important than the provision
of that information.
L1 also displays CIC in terms of his ability to manage turns, hold
the floor and hand over his turn at a particular point in the interaction.
He responds quickly to the teacher’s opening question, as indicated
by the latched turn in 2 and turn continuation in 4, indicated by the
overlapping speech. As well as being able to take as turn and hold the
floor, this learner (L1) also recognises key signals that mark a transition
relevance place – the teacher’s ‘right’ and accompanying overlap in
lines 9 and 10 signal to this learner that it is time to relinquish his
turn at talk and hand over to another learner. While it is the teacher
who ‘orchestrates the interaction’ (Breen 1998), nonetheless, L1 has
to be able to take cues, observe key signals and manage his own turn-
taking in line with what is required by the teacher. He must also
recognise that his own contributions are largely determined by the
teacher’s and by the specific pedagogic goals of the moment.
Extract 8.5
1 T: ok now <let’s start>(.) er today our topic is museum (1)
2 are you going to a museum in the coming summer
holiday?
3 NAME are you (4)
4 S: no I don’t
5 T: why?
6 S: because a museum is very boring if I have some spare
7 times er I would like to choose do something interesting
8 T: mmh so it’s boring to you right?
9 S: er yes
Task 8.3
Read the extract below, which is again taken from a senior high school
class in China where the topic for discussion is visiting a museum.
Based on what you now know about CIC, make a list of the features
of interactional competence in this extract for (a) the teacher and (b)
the learners.
TC

10 T: ok (2) erm how about NAME [how about you are you
going
11 S: [me either me neither
because I
12 think it’s boring too
13 T: why?
14 S: and er at first I don’t have the spare time and if I have
15 I will do something er in my hobbies
16 T: mmhh so you are not interested?
17 S: yes
18 T: thank you (2) well how would you (.) complete this
19 sentence((writing on board)) museum is a place where (.)
20 visitors can(.)how would you complete this sentence
erm
21 say NAME
22 S: look around [look around
23 T: [look around °thank you° (.) look around
how
24 about you NAME?
25 S: a museum is a place er where visitors can have fun
26 T: [have fun
27 S: [have fun
28 T: have fun by? by
29 S: have fun by getting knowledge (3) by getting new
30 ((mispronounced))knowledge
31 T: getting no knowledge (.)
32 S: (1) new new knowledge so- so- sorry
33 T: new kn- knowledge ok don’t be nervous (.) well yeah
well
34 you (.) learn some new knowledge in the museum mainly
35 by (.) looking or (.) listening or (.) reading (5) who’s
36 got the microphone? NAME (.)
37 S: (3) I think it’s by reading something
38 T: mm by reading some material ↑or how about photos?
39 S: er yeah read some photos and maybe some films
40 T: mmhh some films thank you or you look at some photos
41 right? Thank you (.) well it’s very clear that museums
42 are not so attractive to ↑you is that true (1) mmhh?
In this section, we have seen how CIC is portrayed in a number of
contexts. By ‘context’, I mean the physical, geographical and temporal
setting of the interaction in addition to the specific micro-context, or
mode, of the moment. As we have seen (both here and in Chapter 6),
the interactional and linguistic resources used by both teachers and
learners will vary considerably according to specific teaching and

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Task 8.4
In the following extract, identify any evidence of CIC from (a) the
interaction (on the left), (b) the teacher’s own commentary on the inter-
action (on the right). Based on your analysis, how might you change
your own classroom practice?
TC
learning goals at a particular point in time. One aspect of CIC is the
extent to which teachers match their use of language to their intended
goals. The point we are making here is that CIC is one aspect of
learning in formal contexts: teachers and learners, by making appro-
priate interactional choices through their online decision-making, both
facilitate the co-construction of meaning and display to each other their
understandings. CIC manifests itself through the ways in which inter-
actants create space for learning, make appropriate responses ‘in the
moment’, seek and offer clarification, demonstrate understandings,
afford opportunities for participation, negotiate meanings, and so on.
These interactional strategies help to maintain the flow of the discourse
and are central to effective classroom communication. They offer a
different but complementary view of learning through interaction to
that provided by a conversational analytic perspective that focuses
mainly on turn design, sequential organisation and repair.
In the next section, we consider how CIC can be developed in both
teachers and learners by looking at some practical applications of the
construct.
Developing CIC in teachers and learners
Elsewhere in this book (see, for example, Chapter 7), I have
commented on the need for appropriate tools, reflection and dialogue
in order to help teachers become more aware of the interactions taking
place in their local context. I have also made the point that in order
to enhance learning, we need to begin by gaining a closer under-
standing of context and that one of the best ways of doing that is to
gain a fuller understanding of interaction; specifically, the relationship
between interaction and learning.
How would you like to develop CIC in the context in which you teach? What
particular skills would you like to acquire or improve so that interaction is
enhanced? What might you do to help learners improve their CIC?

My aim in this section is to offer some specific strategies, activities,
tasks etc. designed to help both teachers and learners enhance their
CIC. Many of the suggestions that follow will entail you making
recordings or using transcripts and recordings of your, or a colleague’s
classes, or, alternatively using published materials. You may want to
re-read Chapter 3 that deals with recording and transcribing classroom
data.
Students are preparing to do
a board game and clarifying
vocabulary (fifteen minutes)
L: what does it mean singe,
singed my eyebrows?
T: singed, singed (writes on
board) (4) to singe means
really to burn but it always
has the sort of the
meaning to burn
something with hair=
L: =some people you know
too close to the fire so it
singed your eyebrows
burns your eyebrows
T: yeah, yeah
L: (12) bump into cupboard
door is it like hit?
T: yeah it’s like knock into
L: (20) fractured means like
twist
T: no fractured [means
broken]
L: [broken] (12)
What I liked about this is that they
were all asking questions. So for me,
that would point up the use of pro-
viding a structured type of activity;
although there’s a lot management
time setting it up, because they
know they’ve got that time, they’ll
play a far more active role in it. So
again, shutting up and letting them
get on with it.
Task 8.5
One of the ways in which teachers can improve their CIC is to
develop a metalanguage for talking about interaction in their teaching.
The following extract is taken from an interview with a teacher
who has just analysed a recording of a class. Comment on her use
TC
Extract 8.6

1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
of metalanguage: how well is it used and how does it help her to
explain and understand her own practice?
There were quite a lot of display questions which I think is
appropriate for low level classes, pre-intermediate, you tend to
use a lot of display questions so I had plenty of examples of those
like ‘How do you spell exciting?’ and then later I was asking
questions based on the text ‘Does she like Rome?’ of course
were questions I knew the answers to ‘What’s the adjective for
the noun pollution?’ There would be more referential questions
if it was more a discussion with a higher level group so that’s one
thing that came out. I used referential ‘polluted’? ‘Is your city in
China polluted?’ to Lee which enabled them to use a bit more
free sort of speech. But these were very limited the referential
questions to extend or reinforce the vocabulary.
Task 8.6
Make a 15–20 minute recording of part of a lesson. This can be a
class that you either teach or observe. If this is too difficult to organise,
watch a video of someone teaching . Focus on the teacher’s use of
language and identify examples of the following interactional features
(taken from SETT). How would you rate your CIC? Are their any
changes you would like to make to your practice based on this
analysis?
FEATURE DESCRIPTION
A. Scaffolding • Reformulation (rephrasing a learner’s contribution)
• Extension (extending a learner’s contribution)
• Modelling (providing an example for learner(s))
B. Direct repair Correcting an error quickly and directly.
C. Content feedback Giving feedback to the message rather than the
words used.
D. Extended wait- Allowing sufficient time (several seconds) for students
time to respond or formulate a response.
E. Referential Genuine questions to which the teacher does not
questions know the answer.
F. Seeking 1 Teacher asks a student to clarify something the
clarification student has said.
2 Student asks teacher to clarify something the
teacher has said.
TC

FEATURE DESCRIPTION
G. Extended learner Learner turn of more than one utterance.
turn
H. Teacher echo 1 Teacher repeats teacher’s previous utterance.
2 Teacher repeats a learner’s contribution.
I. Teacher Interrupting a learner’ contribution.
interruptions
J. Extended teacher Teacher turn of more than one utterance.
turn
K. Turn completion Completing a learner’s contribution for the learner.
L. Display questions Asking questions to which teacher knows the answer.
M. Form-focused Giving feedback on the words used, not the
feedback message.
N. Confirmation Confirming understanding of a student’s or teacher’s
checks contribution.
Summary
In this chapter, I have presented and characterised the construct
classroom interactional competence. Placing interaction at the centre
of learning, I have argued that in order to enhance learning and
learning opportunity, teachers should begin by developing their own
interactional competence. While I suggest that classroom interactional
competence is highly context specific (both in the general social/
geographic sense and in the more specific sense of ‘context of the
moment’), there are certain features of CIC that can be encouraged
and promoted in any setting. By adopting specific interactional
strategies, CIC can be greatly enhanced. These strategies include the
need for teachers to create space for learning, the importance of jointly
created understandings, the value of shaping learner contributions, the
need to engage and involve learners in dialogue, and so on.
Task 8.7
Make a recording of one of your classes or use a recording of a class
(published or personal). Based on what you now know about CIC,
evaluate the recording in terms of:
• the teacher’s use of ability to create ‘space’ for learning;
• evidence of learner engagement and involvement;
• specific strategies used by teacher or learners to promote under-
standing and establish shared meanings.
TC

Clearly, it is important for teachers to decide for themselves how
to improve their CIC. What I have attempted to do in this chapter is
to offer some thoughts on the various elements that make up CIC and
suggested how teachers might enhance their own understandings. Like
all professional development, there is no one ‘right way’ to improve.
However, understanding a speciﬁc context and developing skills
appropriate to that context are central to any endeavour towards
becoming a better teacher. Developing an understanding of classroom
interaction and improving the way that interaction is managed are, I
suggest, central to improving your teaching.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

9 Conclusions
In this, the ﬁnal chapter of the book, I try to bring together some of
the main themes that have been presented and discussed, and point
to future directions for the study of classroom discourse. For ease and
economy, the chapter is divided into two sections:
• Classroom discourse: looking back.
• Classroom discourse: future directions
Classroom discourse: looking back
At the beginning of this book, I gave a number of reasons for studying
classroom discourse. After more than ﬁfty years of research, it would
be fair to say that the reasons for studying interaction in second
language classrooms have not changed very much; what has changed,
is our approach and how we use the information we obtain. The
principal reason for studying classroom discourse is that it lies at the
heart of everything that takes place in classrooms; to restate van Lier,
it is ‘the most important thing on the curriculum’ (1996: 5). Under-
standing interaction helps us to understand teaching and learning and
improve the ways we, as teachers, go about our professional lives. Like
any human enterprise, we need to understand classrooms if we are to
improve them; analysing interactions is one of the most useful ways
of gaining insights into what goes on there.
Of course, one of the prime motives for looking at interaction so
closely is that it should help us to further our understanding of
learning. Views on this differ. For example, in 1994, Rod Ellis claimed
that L2 classroom interaction studies ‘have contributed little to our
understanding of how interaction affects acquisition’ (1994: 239). In
contrast, in 1998, Michael Breen made the following, opposing
statement:
Social relationships in the classroom orchestrate what is made
available for learning, how learning is done and what we achieve.

These relationships and the purposeful social action of teaching and
learning are directly realised through the discourse in which we
participate during lessons. [. . .] Furthermore, because the data
made available to learners in a classroom are a collective product
with which teachers and learners must interact actively as both
creators and interpreters, because what learners actually learn from
the classroom is socially rather than individually constructed, any
explanation of how language is learned must locate the process
within the discourse of language lessons.
(1998: 119)
Space does not permit a full examination of these polarised
perspectives, but sufﬁce it to say, at least for the time being, that one
reason for such differing points of view is that in order to understand
the role of interaction in learning, we need to understand what we
mean by learning. Our view of learning is conditioned very much by
our own experiences as learners, teachers and researchers. If we adopt,
for example, a largely cognitive view of learning, we may not see why
interaction is important at all (c.f. Ellis above). If, on the other hand,
we adopt a more social view of learning, where learning is seen as
being directly related to participation and involvement (c.f. Sfard
1998), then, clearly an understanding of interaction is both relevant
and important. (For a fuller discussion on this debate, see Seedhouse
et al. 2010). The position adopted here and throughout this book is
that learning is very much a social process; learners learn best, I would
suggest, when they are actively engaged in some kind of task, activity
or discussion. This position coincides very closely with the one
advocated under socio-cultural theories of learning and the work of
scholars such as Vygotsky and Lantolf.
In the context of understanding learning, much of the research on
classroom discourse has focused on describing the features that can
be found in classroom interaction. In the early chapters of this book,
we looked at the ways in which teachers modify their speech, ask
questions, correct errors and maintain overall control of the interaction.
A huge body of research now exists on repair (see, for example,
Seedhouse 2004) and questioning (see, for example, Walsh 2006),
perhaps the most commonly found features of classroom discourse in
any classroom, anywhere in the world. Similarly, stemming from the
work of Sinclair and Coulthard (1975), it would be fair to say that in
most classrooms the IRF/E exchange structure still prevails; teachers
typically ask questions, learners respond and have their responses
evaluated. This is how much classroom discourse is structured and this
is how interaction typically advances in any educational setting.
The ‘discovery’ of the IRF exchange structure by Sinclair and
Coulthard has been fundamental in fostering our understanding of the
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Conclusions 183

very fabric of classroom discourse. For example, we know that one
reason for the prevalence of IRF is that all classroom interaction is
goal-oriented and that the main responsibility for setting goals lies with
the teacher. By understanding the interconnectedness between
pedagogic goals and the language used to achieve them, much has been
done to unveil how learning takes place and how language, interaction
and learning are inextricably entwined (c.f. Seedhouse 2004). Put
simply, by developing understandings of interaction, we are advancing
our understandings of learning.
While IRF is the dominant type of discourse structure to be found
in classrooms, I am not suggesting that it is something we should be
advocating or encouraging; indeed, quite the reverse. By appreciating
that it does exist, we as teachers are in a position to ensure that we
do not fall into the trap of promoting a rather mechanical, rigid kind
of interaction in which there is little scope for creativity and
spontaneity. In earlier chapters, we have seen how it is quite feasible
to promote a different type of discourse in which learners play a fuller
role and in which there is far more evidence of engagement and
dialogue (c.f. Alexander 2008).
More recent studies of classroom discourse have adopted a variable
approach, in which the relationship between language and pedagogic
goals has been studied more closely. These studies provide us with an
alternative approach to describing classroom according to speciﬁc
micro-contexts or modes (see, for example, Seedhouse 2004; Walsh
2006). Rather than assuming that classroom discourse is all of a
oneness, a variable approach makes the important point that language
use and interaction vary according to a teacher’s agenda and what is
happening at a particular moment in a classroom. By understanding
how pedagogic goals and language use are interconnected, we obtain
a different perspective on classroom discourse, and one which is more
closely aligned to the decisions made by teachers and learners. Perhaps
most importantly, this perspective allows us to consider how the
discourse is advanced jointly; both teachers and learners play important
roles in collectively co-constructing meanings through their
interactions. An understanding of the give and take that occurs in the
interaction is fundamental to understanding learning (c.f. Pekarek
Doehler 2010).
Having made some mention of the what and why of classroom
discourse, I now turn to a consideration of the how. The early chapters
of this book and, indeed much of the research that has been conducted
to date, focused on the features and structure of classroom discourse
(the what) and offered reasons for studying it (the why). The ways in
which teachers and researchers have gained access to the complexities
of classroom interaction have changed over the past forty years or so.
Chapter 4 offered an overview of the main approaches to the study
Conclusions184

of classroom discourse, reviewing the work that has been conducted
from interaction analysis, discourse analysis and conversation analysis
perspectives. Each approach has its own merits and shortcomings and
should be adapted to local contexts. For example, it would be
unrealistic, in most situations, for teachers to provide detailed
transcripts of their classes using conversation analysis; equally, for most
classroom research, interaction analysis does not provide the kind of
detail needed to gain a close understanding of the interaction. One of
the central arguments of this book has been that teachers’ professional
development should be data led; if teachers are to improve their
practice, I suggest, they need to make recordings of their classes.
Clearly, however, it would be unrealistic to expect teachers to spend
hours transcribing and analysing their data. By using an appropriate
tool, such as the SETT framework, I have tried to offer alternative
approaches to collecting and analysing data, though it is also apparent
that there is still much to be done in helping teachers access usable
data (see below).
SETT, by aligning language use and pedagogic goals in a limited
number of modes, permits closer scrutiny of the discourse, identifying
interactional features that construct or obstruct opportunities for
learning (Walsh 2002, see also Nystrand 1997; Lin 2000; Nassaji and
Wells 2000). The relationship between learning and interaction is
understood from the teacher’s perspective in the light of intended
learning outcomes. Using SETT, the socially constructed discourse of
the language classroom can be scrutinised in teaching and learning
‘moments’; instances in the discourse where a particular interactional
decision appears to either facilitate or hinder learning opportunity. The
use, for example, of a display question in classroom context mode
might be examined; or the deliberate and careful use of verbal scaffolds
to facilitate understanding might result in enhanced understanding on
the part of the teacher. It is the framework that allows interactive
decisions to be related and compared to both teaching objectives and
learning opportunities. Essentially, the SETT framework is intended
to enable teachers to gain close, detailed understandings of their local
context. While it may need to be adapted to that context, it permits
a dynamic analysis of L2 classroom discourse by paying close attention
to intentions and events.
To summarise this section, I make the following observations on
the current position of classroom discourse research before turning to
a consideration of what the future holds:
1 Despite developments in teaching materials, understandings of
SLA, approaches to teaching and changes in teacher education pro-
grammes, the basic structure of classroom discourse has changed
very little; most classrooms are still dominated by IRF routines.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Conclusions 185

2 That said, when teaching goals and the language used to achieve
them are considered together, a different perspective emerges,
enabling more detailed, microscopic analysis of micro-contexts.
3 Different theories of learning impact greatly on our understanding
and perceptions of classroom interaction: the greater the emphasis
on learning as a social process, the greater the importance attached
to interaction.
4 In recent years, there has been greater emphasis on using our
knowledge about classroom discourse to train teachers and to
advance their professional self-development; however, this is only
happening on a relatively small scale.
5 Technological advances coupled with alternative approaches to
studying classroom discourse have resulted in the emergence of
different perspectives on the nature of classroom discourse and the
relationship between interactional and linguistic features.
Classroom discourse: future directions
So what does the future hold? At the beginning of this book, I posited
a number of challenges for teachers and learners in relation to under-
standing classroom discourse. One of these was to make an observation
that very little time is actually spent making language teachers aware
of the importance of classroom discourse. Although most second
language teacher education programmes offer some kind of subject-
based preparation and training in classroom methodology, there is very
little emphasis on promoting an understanding of classroom inter-
action. As I have tried to demonstrate in this book, an understanding
of interaction is crucial to effective teaching for a variety of reasons:
• Interaction is at the heart of learning.
• Understanding interaction is the ﬁrst step to improving awareness
of context.
• Reﬂective practice is more effective when the focus is classroom
interaction.
• Learners are more engaged and more involved in classes where
teachers understand how to make effective use of interaction.
• Teachers can research their own practices by focusing on inter-
action.
As mentioned in Chapter 1, it would not require too much
imagination to introduce a ‘third strand’ on initial teacher education
Conclusions186

programmes to supplement the two currently in use: language aware-
ness and teaching methodology. The third strand would be called
something like ‘interactional competence’ and would focus exclusively
on ways of helping teachers to understand classroom interaction as a
means of improving their own practice. Not only will this result in
more engaged, dialogic teaching, arguably it will make the whole
process of teaching and learning more enjoyable for all concerned.
A second, and related, direction for the future is the need to further
our understandings of interactional competence and classroom interactional
competence (see Chapter 8). As we have already seen, interactional
competence, the ‘fifth skill’, is set to become one of the major influences
on language teaching, testing and materials design. I suggest that both
interactional competence and CIC are highly context specific and that
teachers have a responsibility to gain understandings of what they mean
in their own context. It would be unrealistic to adopt a ‘one size fits all’
approach to this. Instead, there is a need to characterise interactional
competence for both teachers and learners in a range of contexts. This
will entail collecting and analysing classroom data and using
appropriate tools to analyse the data. Such an opportunity provides an
opening for teachers to become researchers of their own practice and
to enhance understandings of their own professional practice. In short,
it presents teachers with an occasion to engage in research that will
ultimately benefit themselves and their learners, as opposed to ‘having
research done to them’ by outside researchers.
A third challenge for the future is that we will need to develop new
understandings of classroom discourse as our understanding of what
we mean by ‘classroom’ changes. When we consider that many of us
now work in contexts in which technology plays a considerable role,
the traditional physical boundaries of the classroom no longer apply.
Instead, we find ourselves teaching and supporting learning by means
of virtual learning environments (VLEs), computer mediated communi-
cation (CMC), through the use of email, blogs, podcasts and so on.
Clearly, the impact of technology in the last twenty years has been
huge and is set to grow even further. These developments have to be
incorporated into our understandings of interaction and of the ways
in which interaction impacts on teaching and learning.
In the same way, we can see that our understanding of spoken
discourse in classrooms is also in a constant state of flux. This can be
explained in part by developments in technology that have enabled
us to capture and describe more effectively spoken interactions in
classroom settings. But this is only part of the story. Thanks to the
endeavours of researchers and practitioners and their efforts to
approach old problems in new ways by, for example, combining dif-
ferent research methodologies, asking different questions, introducing
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Conclusions 187

new theoretical perspectives, we are now better able to explain and
uncover the complexities of multi-party spoken interaction. Research
methods focusing on the micro-analysis of spoken classroom inter-
action, such as conversation analysis, have played a huge role in
promoting understandings. Similarly, researchers looking at multi-
modality and the interplay of the various features of any spoken
interaction – the words used, intonation and speech quality, gaze,
gestures, embodied actions, etc. – have contributed enormously to this
process.
If one of the prime motives for studying classroom interaction is to
advance our knowledge about learning, we should recognise that
views and attitudes towards learning are also changing. For a signifi-
cant number of researchers following what has come to be known as
a ‘CA-for-SLA’ agenda, learning can be studied ‘in the moment’ by
looking at participation, interaction, orientations and so on. Current
views of learning emphasising its social dimension lead me to believe
that a fourth future direction for classroom discourse research is to
advance understandings of learning as doing, participating, engaging,
etc. Learning is about far more than ‘having’, and while we cannot
ignore the cognitive dimensions of learning, these are much more
difficult to access. If we want to look for evidence of learning, we
should begin by focusing on the words and interactions of learners.
It is here that we will gain the most useful insights and be in a position
to influence the process of learning. This view of learning also resonates
with the kinds of skills that employers are looking for in their
employees: an ability to communicate effectively, to work as part of
a team, to demonstrate interpersonal and problem-solving skills, and
so on. Classroom discourse research that focuses on learning as
participation has much to offer not only in terms of education and
second language learning, but also in terms of informing workplace
practice and enhancing professional development.
On a more practical level, what does the future hold for classroom
practice? How might future classroom discourse research directions
inform teaching methodologies and help learners? If task-based
learning is as widespread as is claimed (see, for example, Ellis 2003),
what is the relationship between task-type and interaction and vice
versa? How might more effective management of classroom interaction
result in a more engaged, more dialogic type of learning? And what
do we know of the importance of interaction during feedback
following a task? This, according to many researchers (c.f. Willis
1996) is the most important part of the task-based cycle and the one
most likely to lead to learning. There is much work to do in this area.
In many parts of the world, classrooms are still dominated by the
IRF routine, with teachers doing 90 per cent of the talking and learners
playing a minimal role in the interaction. A future research direction
Conclusions188

in contexts such as this is for teachers and researchers to consider how
the IRF cycle might be broken, allowing for more engaged, dialogic
teaching and encouraging learners to play a more central role in the
interaction. While there are clearly cultural sensitivities around such
a proposal, I would suggest that changes might begin by looking more
closely at the interactions that take place and not by importing
expensive, Western teaching methodologies into contexts where they
are clearly not appropriate. Learning and learning opportunity can be
enhanced in any context in the world, but the starting point must be
to develop an understanding of that context. This can be best achieved,
I suggest, by encouraging practitioners to study classroom interaction.
From a learner’s perspective, a number of challenges lie ahead. In
earlier sections of the book, I talked about the importance of creating
‘space for learning’. I argued that the mains responsibility for doing
this lies with the teacher and that learners need to play a more equal
role in the discourse. When we consider the ways in which learners
are socialised into certain types of classroom behaviour, this is a huge
challenge. In most content-based subjects, learners answer questions,
respond to cues, follow the teacher’s initiative, avoid interrupting and
so on. And yet, in a language classroom, a very different set of
interactional traits is needed if learners are to play a more equal part
in the discourse. In language classrooms, we need learners to both ask
and answer questions, to interrupt where appropriate, to take the
initiative, seize the floor, hold a turn and so on.
By following learnt behaviours, which are the product of many years
of being socialised into classroom rituals and practices, we may be
facilitating the kind of ‘smooth’ discourse profile, which prevails at
the moment. But are we helping to create interactions that result in
learning? I suggest that we need to encourage interactions that have
a more ‘jagged’ profile in which learners play a more central role in
co-constructing meanings and in ensuring that there are opportunities
for negotiation, clarification and the like. A jagged classroom inter-
action profile has more of the features that would be found in naturally
occurring interactions such as everyday conversation, business
encounters and the like. While not denying that the language classroom
is a social context in its own right, many of its features are determined
by the fact that control of the communication lies with the teacher.
In other contexts, roles are much more equal, resulting in different
interactional features. Turns are longer, for example, and there are
more frequent topic changes. Overlaps and interruptions are more
common, as are pauses. I am suggesting that it is in this kind of inter-
action that learners have the opportunity to acquire the kinds of
linguistic and interactional resources that will help them develop as
learners. Teachers, while still playing a more central role, would need
a far more sophisticated understanding of classroom discourse in order
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Conclusions 189

to be able to manage the interaction, create learning opportunities and
shape learner contributions.
Finally, in terms of possible future directions in the methods used
for studying classroom discourse, we need to be clear about who the
user is. In particular, there is a need to make a distinction between
research tools for teachers and research tools for researchers. Up until
recently, I suggest, the same set of tools has been used by both teachers
and researchers, not always to good effect. As mentioned earlier in
this chapter, it is unrealistic to expect teachers to devote hours of their
time to recording, transcribing and analysing classroom data. What
they need is a tool that offers a ‘snapshot’ of their classes and that
allows them to develop understandings and acquire interactional
competence. For most teachers, their main concern is to grow in
professional expertise and make sure that learners have a rewarding
and enjoyable experience. Short recordings, analysed using some kind
of tool or framework, preferably developed by teachers for their own
use, is a much more realistic goal than wholesale transcription of a
full lesson. The aim should be to collect small amounts of data from
a sample of lessons and build up a proﬁle of a teacher’s classroom
interactional competence. From this, changes to practice can be made
and evaluated over time.
For researchers, future challenges will be concerned with collecting
data that offers greater insights into what really happens in classrooms.
Advances in technology are sure to play a part in this – for example,
we have already seen in Chapter 7 how corpus linguistics might be
combined with other methods to further understandings. Multiple
method research is likely to play an important part in helping
researchers uncover what is going on in language classrooms. In
addition to using technology, another aspect of this is to involve
participants more in the analysis. So many corpus-based studies do
not interview participants – this is surely essential if we are to maximise
our understanding of classroom interaction. We need to talk to teachers
and learners and get their perspective.
There are sure to be many other future directions that classroom
discourse research might take. What is clear is that there has to be a
continuation of the current trend of involving practitioners and
ensuring that they both beneﬁt from new advances in knowledge and
play a key role in producing that new knowledge.
Conclusions190

Task commentaries
Chapter 1
Task 1.1
This is an interesting piece of classroom discourse in many respects.
The most striking feature is the predominance of long teacher turns
and short learner contributions. In terms of speech modification, we
can observe that this teacher uses quite a complex range of vocabulary
for this level, pre-intermediate. It is unlikely that they would be able
to understand many of the words used here: steroids, contaminated,
as a matter of course, scandal, etc. A second observation is the way
in which this teacher moves from one topic to another: opening with
intensive farming, moving on to antibiotics, steroids, street vendors in
a very short period of time. Again, learners may find rapid topic
changes like this very difficult to follow unless they are marked in some
way. Note too that learners interrupt the teacher in lines 74 and 83
to ask language-related questions. In each case, there is some confusion
on the part of the teacher (in lines 75 and 84), followed by a further
clarification by the learner (in lines 76 and 85). Perhaps the most
interesting feature of this teacher’s classroom idiolect (her individual
way of talking in a classroom) is the fact that there is almost no
signposting for learners. Note how little use is made of discourse
markers like ok, right, so to mark the beginning and ends of transitions.
Note too how little evidence there is here of pausing, normally a
particular feature of modified speech and one that is essential for
learners trying to follow the discourse. This is especially important in
the final teacher turn, especially in lines 92–93 where the focus of the
discourse is language and the word hygienic. This change in focus is
not marked in any clear way, it is somehow embedded in the discussion
on street vendors and likely to cause great confusion for learners. In
response to the task questions, there is minimal guiding here by the
teacher. Learners are likely to feel lost and excluded owing to the high
level of vocabulary used, the lack of pausing, the absence of discourse
markers, the rapid topic changes, their perceived need to interrupt to
seek clarification.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Task 1.2
Other reasons for asking questions include:
• to get a piece of information from a student (a genuine question);
• to check concepts;
• to engage learners;
• to elicit full responses;
• to motivate learners.
While there are plenty of reasons for asking questions and this
is by far the most frequently used strategy by teachers for eliciting a
learner response, there are other strategies available, which might
be equally effective. These include the use of pictures as prompts
for a response, getting learners to ask questions across the class,
using single word cues, getting students to ask questions as well as
answer them, and so on. The point is that, although the teacher’s aim
is to elicit a response, it is not always necessary to ask questions to
do so.
Task 1.3
The first thing to note is the two-second pause at the beginning of the
learner’s response, possibly indicating that this learner is not sure how
to reply. Silence usually indicates some kind of interactional ‘trouble’
– here some doubt as to how to respond to the teacher’s question.
The latched turn in 12/13 (marked =) suggests that the teacher has
interrupted in order to correct the error * I am agree. This is followed
in 13–14 with an explanation by the teacher and an acknowledgement
(in 15) by the student. We are not sure what happens next (marked
((3)), three seconds of confusion), so it is difficult to evaluate the
extended explanation in lines 17–20. However, what we can say is
that this explanation is quite disruptive to the flow of the open type
of interaction, which is in evidence here. In other words, the teacher,
through her choice of repair strategy, has actually disturbed the
interaction and minimised learner involvement. A short, direct repair
might have been the answer here so as to avoid disruption and ensure
that learners’ involvement was maximised. Given that the teacher’s
aim is to elicit feelings, her choice of error correction strategy is
at cross-purposes with her pedagogic goal. A very quick correction
and return to the main focus of the lesson would have been more
appropriate.
Task commentaries192

Chapter 2
Task 2.1
The main difference between these two tasks is that A is much more
open and discursive than B, which has a more definitive end-point.
Essentially, the focus of A is the topic, whereas B is more closely
oriented to the task and to task achievement. To complete Task A,
students will need to know how to make suggestions, agree/disagree,
evaluate each others’ ideas and so on, whereas task B could actually
be achieved with very little language (student A: ‘parachute’; student
B: ‘ok’, etc.). Tasks that are more topic-oriented such as Task A,
normally require more preparation if they are to produce longer, fuller
student contributions. Depending on the level of the learners, teachers
may offer some structure to the task to ensure that it is achieved. Here
for example, students could be asked to individually list their own
arguments, compare in pairs, then fours, before coming up with one
common list and providing reasons for their choices to the class as a
whole. In terms of enjoyment, this will vary hugely from learner to
learner. In many ways, Task B is easier because it has a clear structure
and a definite end-point; learners know when they have finished the
task. On the other hand, Task A might appeal to an international group
of students because it allows a genuine exchange of opinions – students
always like to hear about their classmates!
In terms of interaction, Task A may have more breakdowns, more
long pauses, more errors and so on, while B may be complete with
only a cursory use of language. A will require more effective topic and
turn management, while B will pose a challenge in terms of the group
coming to some kind of agreement. To summarise, we can distinguish
between topics and tasks in the following way:
Topics
• good topics are ones which learners can identify with; depends on
their world knowledge;
• topics normally should be controversial (c.f. converging and
diverging topics);
• can be done as a formal debate;
• allow preparation time: language, task.
Tasks
• tasks are goal-oriented;
• tasks have a definite end-point (5 differences, 3 minutes to complete,
4 reasons);
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

• objectives should be non-linguistic;
• objectives should require interaction;
• visual stimulus may be helpful.
Task 2.2
Extract 1
This is clearly pre-task and set up. The teacher is locating learning in
time and place (‘this morning, last Christmas’, etc.) before organising
the groups and setting the task. While it seems quite clear, the
instruction is very short and is not checked in any way.
Extract 2
This is clearly the ‘task’ phase and learners are working quite
independently of the teacher in a free-ﬂowing discussion. The teacher
takes a ‘back seat’ and acts more like a member of the group, simply
showing understanding (256 and 260) and letting learners get on with
the task. Learners, for their part, manage all the turn-taking and topic
management on their own and for all intents and purposes, the
discussion resembles everyday conversation. This teacher succeeds
very well in using language that is highly appropriate to the stage of
the TBLT process and gives learners the space they need to complete
the task.
Extract 3
This is the language focus or post-task phase of the cycle, where
teachers typically get feedback following a task and offer their own
input. There are some problems in this extract: a misunderstanding in
74–76, a breakdown and repair in 82–84. Note too the excessively
long turn of the teacher and the fact that the topic seems to ‘meander’
somewhat. While the teacher seems concerned to engage learners in a
discussion, the students want to check their answers and learn new
vocabulary. There seems to be a mis-match between what the teacher’s
pedagogic goals are and what the learners perceive.
Task 2.3
This is a very good example of what is often referred to as ‘guiding
questions’. The teacher, by skilfully asking follow-on questions,
succeeds in helping learners to the answer she is looking for, which
eventually happens in line 34. This strategy requires great mental agility
on the part of the teacher who must listen carefully to what students
say. Note how it pays off in the extended learner turn in line 34. Her

first question is very open and is met with an inappropriate response, in
terms of what she is looking for. Rather than reject the response
‘scary’ in 29, she seeks clarification (in 30) before bringing the discourse
back on track with her question in 33. This teacher is very effective
in her questioning strategy and is ‘rewarded’ with a comprehensive
response in 34.
Chapter 3
Task 3.1
In extract 3.1, the teacher is trying to elicit military force. He uses a
number of devices to guide learners towards a correct response and
to ‘shape’ their contributions:
• In 188, he alerts the learners to the phrase he is looking for by first
clapping his hands and then saying military what?
• In 190, he provides the synonym fight as means of guiding learners
towards the correct response.
• In 192 and 194, he uses direct repair as a means of minimising
interruptions to the flow of the interaction and leaving the floor
open to learners.
• In 194 and 196, he uses a rising intonation to guide learners towards
the phrase military force.
• In 198, he emphasises the word force to show that this is the correct
response.
We can make the assumptions above (a) by analysing the interaction
carefully and paying close attention to the specific features of the
discourse and (b) by using our experience of classrooms and what
typically occurs when teachers elicit vocabulary. There is too the
methodological argument that encouraging learners to use intelligent
guessing is a useful strategy for helping them to both deal with
unknown lexical items and guess what might be the most appropriate
word – as in this case.
Task 3.2
In this extract, the teacher’s stated aim is ‘to promote oral fluency and
discussion’. While her choice of questions seems to suggest that this is
indeed her aim (what about in Spain if you park your car illegally?), she
manages the interaction in such a way that learners are unable to fully
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

articulate a response. Throughout this extract, the teacher interrupts and
corrects errors (in 275, 277, 279, 281, 283, 285), resulting in minimal
responses from learners. We can say that this teacher, by her use of
language and interactional strategies is actually obstructing learning and
learning opportunity. Her pedagogic goal (oral fluency) and the
language used to achieve that goal are actually in opposition. Learners,
in this micro-context, need interactional space and an opportunity to
express themselves; both are constantly denied here.
In terms of interaction and learning, the position adopted here is
that the relationship is a very strong one, especially when learning is
seen as participation. If learners are unable to participate, it is,
arguably, unlikely that learning can take place.
Chapter 4
Task 4.1
Choosing a transcription system is very much dependent on how
much detail you need to show in the data. If you are following a strict
CA approach, then every detail is important. If, on the other hand,
you are following a DA approach, you may find that less detail is
needed. The one below is appropriate in that it does show pauses
(though not in any great detail), overlapping speech and latched turns.
It also includes some prosodic features of spoken language such as
rising and falling intonation and emphatic speech. On the other hand,
it does not include details such as voice quality (such as ‘smiley voice’),
nor does it show in detail pauses of less than one second.
The guideline, then, has to be ‘fitness for purpose’. The transcription
system must show enough detail to be workable and to enable
interpretation of the data. On the other hand, if it has too much
detailed information it may be unusable by readers or other analysts;
it may also make the task of transcription so onerous that it becomes
self-defeating.
Task 4.2
The main difficulty associated with using FIAC is that we are trying
to apply language functions to classroom discourse as outsiders,
without really being sure what is happening at a particular point in
time. For example, using FIAC in this extract, line 11 may function
as (4) teacher asks a question or (6) teacher gives direction. Do we
classify line 13 as (6) gives direction or (7) criticises or uses authority?
We face similar difficulties when looking at student talk. While line 12
is clearly (8) pupil response, how do we classify the remaining learner
utterances? Is line 15 a response or is it an initiation?

The point is that all utterances can perform more than one function
and it is almost impossible to say, with conviction, what a particular
utterance is ‘doing’ at any given moment in classroom interaction. For
this reason, using FIAC is fraught with difficulty.
12 L: (2) erm I am agree with=
13 T: =agree be careful with the verb to agree there you as
well Ensa that it’s we! agree it’s not to
14 be agree it’s to agree! Ok=
15 L: [oh I agree]
15 L: ((3))
is to agree ok so ((3)) to agree with
17 (writing on board) is the preposition that follows it I so
it’s I agree with you I disagree with
18 you . . . ok em Silvie can you em what were you going
to say?
19 L2: I agree with you because em when when we talk about
something em for example you
20 saw a ((2)) on TV=
Task 4.3
Extract 4.6 (see below) has been included in order to illustrate
some of the limitations of a DA approach. The first obvious feature
is that the discourse is almost entirely managed by the students,
with only one teacher turn (in 10). The absence of IRF patterns of
discourse is also noticeable, as is the way in which turns are self-
selected, passed, seized or held. There are overlapping turns, where
several students speak at once (indicated by / /, or [ ]), and evidence
of meanings being negotiated by students independently of the
teacher, as in 4–5, 6–7. Interactional space is maximised, and there is
a sense that the floor has been ‘handed over’ to the learners, with the
teacher taking a ‘back seat’. In many respects, this piece of discourse
can be likened to casual conversation: turn-taking is rapid, roles are
symmetrical and participants have complete freedom as to when they
speak and when they remain silent. Such instances are not uncom-
mon in the contemporary L2 classroom, stemming perhaps from
learner-centredness and communicatively-oriented teaching method-
ologies. Were a strictly DA approach adopted, it is highly unlikely
that the unfolding discourse could be adequately captured, resulting
in idealisation of the data or misrepresentation of ‘what actually
happened’.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Extract 4.6
2 L: =so it’s good news (laughter)
3 LL: /bad news/ ok / no no that’s good news/. . .
Chapter 5
Task 5.1
Go to the Collins Bank of English website at www.collins.co.uk/
Corpus/CorpusSearch.aspx. Use this free software to perform the
following operations for not really.
1 Under the corpus concordance sampler, type in not+really in the
‘type in your query’ box. Now tick the ‘British transcribed speech’
box.
2 Click ‘show concs’ (concordances). You’ll get a list of concordance
lines for not really.
3 Read through the concordance lines. What observations about
not really can you make? How might you use this with a group
of learners?
4 Consider how this approach could be used with a corpus of
classroom discourse. What might it reveal?
Task 5.2
I have marked in possible options here, there may be others. In the
opening lines, the teacher locates the class temporally (today, last
class/last week). This is important so that learners have a point of
reference and can relate today’s session to a previous one. The teacher
then goes on to remind learners what happened in the last class (we
did a quiz ok/right/yes). Not only does this serve to remind learners
what happened last time, it also creates ‘shared space’, a sense that
we’re all in this together and we can now move forward with the work
we did last time. Shared space is crucial if learning is to take place

and most teachers work very hard to create such an environment (see
the discussion on shared space in Chapter 5).
The final two discourse markers (right/ok/now) mark the discourse
and function almost like punctuation in a written text. Their purpose
is to get learners’ attention and to draw attention to the instruction
that follows.
T: Today we’re going to be reading about and talking about er
music and pop stars and famous people Last class/week we
did a quiz right/ok/yes and it was a picture quiz and some of
the pictures were people from your countries. Right/ok/now let
me see could you work with Ben and could you three guys work
together (T organises groups). If you recognise any of those
people who they are and what you know about them and then
what kind of lifestyle do you think these people have. OK/now/
right you can begin when you’re ready.
Task 5.3
I have given a suggested answer to the task below. Note that decisions
concerning ‘real’ or ‘artificial’ may be based on how real or authentic
an activity is in class. For example, discussing homework is a very
‘real’ activity in class, as is planning a composition. Both activities
require language to be used for genuine communication. Other
activities, however, such as doing a role play, or answering compre-
hension questions are artificial in the sense that they are designed to
promote language practice, but not to encourage genuine communica-
tion. Clearly, most classes will include a combination of both ‘real’
and ‘artificial’ activities.
Artificial Real
2 4 5 6 1 3
7 Planning a composition with a group of students
8 Doing a role play
9 Discussing homework with students
10 Modelling new language
11 Watching a video and answering comprehension questions
12 Taking part in a class debate
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Chapter 6
Task 6.1
While there are no ‘right’ answers to this question, it is very likely
that managerial mode will be present in your teaching:
• when setting up a new activity;
• when concluding a task;
• when giving an instruction;
• when organising learners into pairs or groups;
• when monitoring a task, giving advice, etc.;
• when giving feedback after a task.
There may be other occasions. Essentially, the management of
learning occupies a great deal of time and it is something that most
of us, as teachers, are concerned to improve. Note too that, in many
contexts, the management of learners takes up a lot of class time. While
managerial mode does not specifically include things such as dealing
with difficult students, discipline and so on, these are clearly important
features of classroom life and cannot be ignored.
Some of the difficulties associated with managing learning include:
• giving clear instructions;
• setting and adhering to time limits;
• using signpost words appropriately;
• making a distinction between an ‘instructional voice’ and a ‘teaching
voice’ so that learners are able to understand instructions more
easily;
• moving learners in and out of group-work and plenary (open class)
activities.
There may be others specific to your own context. One of the best
ways of improving class management is to make short recordings of
stages in a lesson that involve close class management and analysing
them (see Chapter 6).
Task 6.2
This extract opens in skills and systems mode (lines 29–33). The
teacher is dealing with a language question concerning the difference

between olive oil and sunflower oil so this is skills and systems mode.
The slight hesitation and reference to the coursebook (erm in 33)
signals a transition to materials mode (lines 33–46). Note how the
turn-taking is tightly controlled, there are a lot of IRF sequences and
all the interaction evolves around the materials being used. There is
little scope for students’ personal opinions and the teacher’s role is to
get students to display their understanding.
In line 47, we see a transition to classroom context, signalled by
laughter and the reference to students’ names (Yvette, Georgia and
Haldoun). This has a striking effect on the interaction that ensues;
note the long learner turn (53–56) and the more personalised responses.
Students here have been given the floor and are free to comment on
their past experiences of pot luck suppers. Classroom context mode
continues until line 61 when the teacher switches modes, rather
suddenly, by asking a language focused question (fussy). Lines 61–70
are skills and systems mode and the focus is on language, with learners
seeking clarification and asking questions. There is some confusion
here (in 68), possibly caused by the rather rapid mode switch in 61.
The main difficulty with this kind of analysis is that we are outsiders
trying to work out what has happened in a particular context. This
problem does not arise when you use SETT to evaluate your own
teaching since you are more aware of the decisions taken and the local
context.
30 L: =and the other olive oil vegetable and made from maize
from corn=
31 T: =yes corn oil and yeah there’s many different types of
oil olive oil is the best and . . .
32 essential to Spanish omelette tortilla erm right so if you
were invited to one of Leslie’s
33 pot-luck suppers (referring to the book) erm right what
would you bring . . . along . . .
34 with you?
36 T: yes this is this Leslie the American woman imagine she
[invited you to a pot-luck
37 supper
42 L: =dessert
44 L: =or a starter
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

45 T: or a starter . . . yes so how would you know what to
bring?
47 T: that’s right yes so it’s it’s quite well-organised you can
(laughs) imagine writing a list
48 oh yes I’ll ask Yvette she can bring a starter . . . oh
Georgia a nice salad Italians are
49 very good with salads and then maybe a main course
er Haldoun and perhaps er the
50 Japanse can do some more some more main courses
so it would be very er very
51 well-organised=
52 L: =usually((2)) like sometimes there are like 3 main courses
((2)) like graduation in
53 sometimes we have parties and we have like er roast
chicken and another people
56 T: [(laughs)] oh dear well Georgia perhaps when you go
back to Italy perhaps you can
58 plenty of desserts and plenty of starters (laughs) but er
do you think it’s a good idea?
60 L: it’s nice because you don’t know what you’re going to
eat=
61 T: =it’s a surprise yes, yeah as long as you like everything
. . . I mean some people don’t
62 like certain things what adjective do we use for people
who don’t like that and hate
63 that and a lot of food they won’t eat. . . we call it people
are very . . . FUSsy fussy with
64 their food (writes on board) right so fussy that’s don’t
like vegetables, never eat er
65 pasta (laughs)=
66 L1: =excuse me how to say if er you ((2)) for example if you
try some burger on cormer
68 T: er . . . I’m not quite sure what you mean Yvette if you?
70 T: you mean street sellers . . . people selling food on the
street yes

Task 6.3
You should have the following if you have completed the grid correctly:
A. Scaffolding 1. Reformulation (rephrasing a learner’s contribution)
2. Extension (extending a learner’s contribution)
3. Modelling (providing an example for learner(s)
C. Content feedback Giving feedback to the message rather than the words
used.
D. Extended wait-time Allowing sufficient time (several seconds) for students to
respond or formulate a response.
E. Referential questions Genuine questions to which the teacher does not know
the answer.
F. Seeking clarification 1. Teacher asks a student to clarify something the
student has said.
2. Student asks teacher to clarify something the teacher
has said.
G. Extended learner turn Learner turn of more than one utterance.
H. Teacher echo 1. Teacher repeats teacher’s previous utterance.
I. Teacher interruptions Interrupting a learner’ contribution.
J. Extended teacher turn Teacher turn of more than one utterance.
M. Form-focused Giving feedback on the words used, not the message.
feedback
N. Confirmation checks Confirming understanding of a student’s or teacher’s
contribution.
Task 6.4
129 L5: =I believe in trying new things and ((1)) ideas=
130 T: =er (1) you believe in being POSitive you mean?= (F)
131 L5: =pardon?=
132 T: =do you believe in always being positive is that what
you mean?= (F)
133 L5: =no . . . I believe to (3) to have a lot of achievement
134 T: (1) do you believe in what do you mean you you should
always take opportunities is
135 that what you mean no?= (N)
136 L5: =no I want my life to be very (1)
137 T: happy?= (A)
138 L5: =yeah and also I I do many things (1) many different
experiences=
139 T: =why don’t you say you just believe in experiencing as
many different things as you
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

140 want= (A)
141 L5: =oh yeah=
142 T: =thats what I think you should say=
I have marked the most obvious features by using letters: each letter
is taken from the SETT grid in the previous task. So, in lines 130 and
132, the teacher is seeking clarification, and in line 134, this becomes
a confirmation check. I am interpreting 137 and 139 as examples of
scaffolding; the teacher is trying to help L5 articulate a response more
clearly by ‘feeding in’ the language required. Note that as we have
seen, ‘labelling’ utterances in classroom discourse in this way is
notoriously difficult and there may be other possible answers here. I
have used my knowledge of the context and of the data to present the
most plausible answers.
Task 6.5
This task has no commentary. The main rationale is to give you an
opportunity to try out the SETT framework for yourself and evaluate
its usefulness. It might be worth comparing your evaluation with that
of a colleague if this is possible.
Chapter 7
Task 7.1
There are literally hundreds of ways of completing this table. I have
offered just a few questions here. When devising reflective practice
questions, the main consideration should be their ‘answerability’.
While it may be possible to come up with lots of areas of focus, it is
important to recognise that only some questions can be answered,
while others may lie outside your control. Choose questions that give
a clear focus, which can be answered and which you have the capacity
to control.
Teachers Learners
How can I improve the questions I ask? How can I involve my students more in
What alternative ways of giving my classes?
feedback can I try? How can I encourage students to ask
How can I improve my instructions? questions?
I ask too many display questions; what What is the effect on interaction of
can I do about this? allowing ‘rehearsal time’?
What can I do to improve the ways in How can I get students to use English
which I elicit from my students? more actively in class?
How can I reduce my teacher echo? What techniques can I use to get
students to scaffold each other?

Task 7.2
Each of the following situations may cause problems, at least some of
the time. Obviously, my suggested solutions may have to be adapted
to your own context.
• Students refuse to speak English when working independently. This
is a ‘problem’ that teachers often comment on – ‘how can I get
them to speak English all the time?’ The solution might be simply:
don’t even try! It is probably more realistic to get students to use
English for short ‘bursts’ of activity and then allow them to use L1
again. Set time limits and challenge students to use English for five,
ten minutes or longer or just for the completion of a single task.
Get them to reflect on the problems they experienced; record them
during a task and let the class comment. And don’t forget, there is
nothing wrong with using L1 in class – it can be very useful indeed.
• Your instructions never seem to ‘work’ and are always misunder-
stood. This is something we all experience. Instructions are just
difficult to give and to be understood. Try the following:
• Break all instructions down. It should be possible to organise all
activities using 1, 2, 3.
• Check instructions by getting one student to summarise.
• Give part of an instruction and have students complete that
before moving on to the next part: instruction-action-instruction-
action, etc.
• Record your instructions and analyse them – where is it going
wrong and how might you improve?
• Your teaching is dominated by question–answer routines. This
needs a recording to be made or the help of an observer or ‘critical
friend’. You need to identify moments in your teaching when it
would be possible to elicit without asking a question. Use cues,
prompts, silence, pictures, etc. But most important, you need to be
more conscious about the use of questions and look at their purpose
in relation to what you want to achieve at a particular point in
time. Plan whole sections of a lesson where you ask no questions
and do encourage students to ask more questions so that you spend
more time answering their questions and less time asking your own.
• Your students constantly do badly in the end of unit tests you set
them. You need to analyse the tests and look at the teaching you
have done: are the tests fair? Do they include the content of what
you have been teaching and do they test in the way that you have
taught? For example there’s no point getting students to read long
texts if you want to test their grammar – choose an alternative.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Task 7.3
I believe that talking to colleagues is one of the most effective means
of improving your professional practice. Try it if you haven’t already
done so. Better still, get them to come and watch you teach having
first discussed a problem or issue and tried to implement a change.
The combination of observation plus dialogue is incredibly beneficial
and, in my experience, always gets results.
Task 7.4
Following on from the commentary on the previous task, the idea here
is to get you to appreciate the value of professional dialogue.
Stimulated recall is an excellent way of recording yourself teaching
and then discussing with a colleague. It works like this: make a
recording of one of your lessons (10–15 minutes is plenty, audio or
video). Find some time to watch or listen to the recording first on your
own and then with your colleague. Focus on one aspect of the
recording and discuss this, recording the conversation you have with
your colleague. Finally, use both recordings (of your teaching and the
discussion with a colleague) to try to reflect on, understand and make
some changes to your professional practice. The whole process is
designed to give you greater insights into your teaching and to allow
you to share problems and possible solutions with another
professional. It should take place in a friendly, supportive environment
and trust and mutual respect must be upheld at all times.
Chapter 8
Task 8.1
See the commentary in Chapter 8, pages 162–3.
Task 8.2
Like all of the tasks so far, the answer to this one will be very context-
specific and dependent on local conditions. If, for example, you work
mainly with adult learners who are highly motivated, your conception
of interactional competence will be very different to that of someone
working in a state school with very young learners. The following list
is neither exhaustive nor applicable to every context; it is merely
indicative of some of the things you may want to include in your
definition of IC in your context. Note that the features I have listed
below are different to those often used to characterise oral fluency;
here, we are interested in what goes on between speakers, not in solo
performance (see Chapter 8 for a fuller discussion on this).

Interactional competence: teachers Interactional competence: learners
Uses language appropriate to the goal Recognises the kind of response
of the moment. required: short or full response.
Listens and responds carefully to Is able and prepared to talk about
learners. personal experiences.
Avoids interrupting all the time. Helps other learners by scaffolding,
Allows learners to have ‘interactional suggesting, prompting, ignoring errors
space’. where appropriate.
Takes care to support learners by Is prepared to ask and answer questions.
scaffolding where necessary. Seeks opportunities to use English both
Allows learners to ‘struggle’ some of with teachers and classmates.
the time; avoids giving the answer. Is not afraid of making mistakes and will
Tries to help learners by ‘shaping’ take risks in open class discussions.
responses and not always accepting Respects the views of others and is
the first answer. prepared to be open-minded.
Avoids asking display questions Manages turn-taking well and is able to
all the time. ‘bring in’ other students in pair and
Is not afraid of silence and is group-work tasks.
prepared to use it to allow Is able to open and develop a range of
learners more ‘space’. topics.
Uses error correction that is Makes use of both linguistic and
appropriate to the moment. interactional resources to communicate
Creates a friendly, supportive effectively.
learning environment. Uses acknowledgement tokens (or
Avoids persistent use of IRF backchannels) – words like uh uh,
and is aware of its mmm, ok, right, yeah, etc. – to signal
consequence on interaction. understanding to a speaker.
Task 8.3
The following features of CIC can be identified in extract 8.5:
1 An absence of repair. Repair is again either totally lacking
(examples) or it is done in a very constructive way. For example,
the other-initiated, other-repair in line 31 ‘feels’ more like a request
for clarification than a direct error correction. This serves the
purpose of making students feel at ease and willing to take risks.
Again, in this micro-context where the prime focus is elicitation
and the sharing of personal experiences, repair has to be handled
in a sensitive way in order to maximise space for learners and for
learning. Note too the subtle error correction in line 40 (‘look at
some photos’). This is simply fed into the dialogue, a strategy that
serves to minimise interruption and maximise the ‘flow’ of the
discourse.
2 Seeking extensions. This teacher does not always accept the first
response she gets and tries to ‘push’ learners to say more (for
example, in lines 5 and 13 where she asks learners to justify their
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

response: ‘why’?). On both occasions she succeeds in getting an
extended response from learners who give reasons for a previous
contribution. (Note that the strategy of constantly asking ‘why’
might have the opposite effect and cause learners not to offer an
elaborated response. It should, therefore, be used sparingly).
3 Increased wait time – note the pausing:
• Line 4 (four seconds). This pause allows the student to formulate
a response and then go on to justify it.
• Lines 10 and 18 (both two seconds). Here the pause is used to
signpost the end of a particular stage in the lesson and mark a
transition to a new focus. This allows students to effectively
‘navigate the discourse’ (Breen 1998) and move on to a new stage
in the lesson. Pausing, then, has many different functions, in the
same way it has several functions in everyday conversation.
• Line 35 (five seconds). While this may not be deliberate, she is
looking for the microphone, this extended pause allows the
student to formulate a response. . .
4 Using ‘interactional pointers’. To help learners follow the discourse
and take part, this teacher uses specific spoken strategies that help
learners know what is expected of them and what their role is at
a specific point in time. These include: nominating a student (3, 10,
21, 24, 36); signpost words to mark transitions or a change in focus
(well (18, 41); locating learning in time and space (1). These
‘navigational aids’ help learners stay together and prevent them from
becoming lost in the discourse (c.f. Breen 1998).
5 Use of acknowledgement tokens (8, 16, 38, 40) is very important
in keeping channels open and ‘oiling the wheels of the interaction’
(McCarthy 2003). Arguably, listeners – here the teacher – play a
key and fundamental role in helping speakers to hold the floor and
stay focused.
Extract 8.5
1 T: ok now <let’s start>(.) er today our topic is museum (1)
2 are you going to a museum in the coming summer
holiday?
3 NAME are you (4)
4 S: no I don’t
5 T: why?
6 S: because a museum is very boring if I have some spare
7 times er I would like to choose do something interesting
8 T: mmh so it’s boring to you right?

9 S: er yes
10 T: ok (2) erm how about NAME [how about you are you
going
11 S: [me either me neither
because I
12 think it’s boring too
13 T: why?
14 S: and er at first I don’t have the spare time and if I have
15 I will do something er in my hobbies
16 T: mmhh so you are not interested?
17 S: yes
18 T: thank you (2) well how would you (.) complete this
19 sentence((writing on bb)) museum is a place where (.)
20 visitors can(.)how would you complete this sentence
erm
21 say NAME
22 S: look around [look around
23 T: [look around °thank you° (.) look around
how
24 about you NAME?
25 S: a museum is a place er where visitors can have fun
26 T: [have fun
27 S: [have fun
28 T: have fun by? by
29 S: have fun by getting knowledge (3) by getting new
30 ((mispronounced))knowledge
31 T: getting no knowledge (.)
32 S: (1) new new knowledge so- so- sorry
33 T: new kn- knowledge ok don’t be nervous (.) well yeah
well
34 you (.) learn some new knowledge in the museum mainly
35 by (.) looking or (.) listening or (.) reading (5) who’s
36 got the microphone? NAME (.)
37 S: (3) I think it’s by reading something
38 T: mm by reading some material ↑or how about photos?
39 S: er yeah read some photos and maybe some films
40 T: mmhh some films thank you or you look at some photos
41 right? Thank you (.) well it’s very clear that museums
42 are not so attractive to ↑you is that true (1) mmhh?
Task 8.4
1 Evidence of CIC in the interaction:
• Frequent and extended pauses.
• Learners asking questions.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

• Direct error correction (no fractured [means broken]).
• Extended learner turns.
• Interactional space given to learners by teacher.
2 Evidence of CIC in the teacher’s own commentary on the
interaction:
• Her observation that students are asking questions.
• Her observation that she needs to stay silent some of the time
(So again, shutting up and letting them get on with it).
• Her comments on the value of spending time setting up an
activity to maximise the learning potential of the task.
Extract 8.6
Task 8.5
This teacher uses a number of metalinguistic expressions to analyse
her own use of language. Her main focus is the questioning strategies
Students are preparing to do
a board game and clarifying
vocabulary (ﬁfteen minutes)
L: what does it mean singe,
singed my eyebrows?
T: singed, singed (writes on
board) (4) to singe means
really to burn but it always
has the sort of the
meaning to burn
something with hair=
L: =some people you know
too close to the ﬁre so it
singed your eyebrows
burns your eyebrows
T: yeah, yeah
L: (12) bump into cupboard
door is it like hit?
T: yeah it’s like knock into
L: (20) fractured means like
twist
T: no fractured [means
broken]
L: [broken] (12)
What I liked about this is that they
were all asking questions. So for me,
that would point up the use of pro-
viding a structured type of activity;
although there’s a lot management
time setting it up, because they
know they’ve got that time, they’ll
play a far more active role in it. So
again, shutting up and letting them
get on with it.

she used and the difference between display and referential questions.
What is interesting about her commentary is not so much that she
knows the difference, but that she appears to be very aware of why
she uses different types of questions. She comments on the fact that
display questions are more appropriate with lower levels and for
checking learning, whereas referential questions allow greater
‘freedom’ of expression.
Task 8.6
The main purpose of this task is to familiarise yourself with the
SETT framework and to evaluate your teaching using an instrument
that does not require wholesale transcription. Follow the procedure
given and Appendix A and repeat this 2–3 times. Each time you
should focus on one aspect of your teaching and try to make some
changes. If possible, get a colleague to observe you once you have
completed this procedure and identify examples of CIC in your
teaching (see Task 8.7).
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Appendix A
SETT: Self Evaluation of Teacher Talk
Procedure
• Make a 10-15 minute audio-recording from one of your lessons.
Try and choose a part of the lesson involving both you and your
learners. You don’t have to start at the beginning of the lesson;
choose any segment you like.
• As soon as possible after the lesson, listen to the tape. The purpose
of the first listening is to analyse the extract according to classroom
context or mode. As you listen the first time, decide which modes
are in operation. Choose from the following:
1 Skills and systems mode (main focus is on subject content, skills
or knowledge).
2 Managerial mode (main focus is on setting up an activity)
3 Classroom context mode (main focus is on eliciting feelings,
opinions, attitudes, etc.)
4 Materials mode (main focus is on the use of text, tape or other
materials).
• Listen to the tape a second time, using the SETT instrument. Write
down examples of the features you identify.
• If you’re not sure about a particular feature, use the SETT key
(attached) to help you.
• Evaluate your teacher talk in the light of your overall aim and modes
used. To what extent do you think that your use of language and
pedagogic purpose coincided? That is, how appropriate was your
use of language in this segment, bearing in mind your stated aims
and the modes operating.
• The final stage is a feedback interview with another colleague or
with me. Again, try to do this as soon as possible after the
evaluation. Please bring both the recording and SETT instrument
with you.

Lesson Cover Sheet
A. Lesson details:
Name:
Class:
Level:
Date:
Overall aim:
Age:
Materials:
B. Lesson modes identiﬁed
C. Self-evaluation of teacher talk
Evaluate your teacher talk in the light of your overall aim and modes used.
To what extent do you think that your use of language and pedagogic
purpose coincided? That is, how appropriate was your use of language in
this segment, bearing in mind your stated aims and the modes operating.
Continue on the next page if necessary.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Appendix A 213

A. Scaffolding 1. Reformulation (rephrasing a learner’s contribution)
2. Extension (extending a learner’s contribution)
3. Modelling (providing an example for learner(s)
C. Content feedback Giving feedback to the message rather than the words
used.
D. Extended wait-time Allowing sufficient time (several seconds) for students to
respond or formulate a response.
E. Referential questions Genuine questions to which the teacher does not know
the answer.
F. Seeking clarification 1. Teacher asks a student to clarify something the
student has said.
2. Student asks teacher to clarify something the teacher
has said.
G. Extended learner turn Learner turn of more than one utterance.
H. Teacher echo 1. Teacher repeats teacher’s previous utterance.
I. Teacher interruptions Interrupting a learner’ contribution.
J. Extended teacher turn Teacher turn of more than one utterance.
M. Form-focused Giving feedback on the words used, not the message.
feedback
N. Confirmation checks Confirming understanding of a student’s or teacher’s
contribution.
Feature of teacher talk Examples from your recording
A. Scaffolding
B. Direct repair
C. Content feedback
D. Extended wait-time
Appendix A214

E. Referential questions
F. Seeking clariﬁcation
H. Extended learner turn
I. Teacher echo
J. Teacher interruptions
K. Extended teacher turn
L. Turn completion
M. Display questions
N. Form-focused feedback
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Appendix A 215

Appendix B
Transcription system
The transcription system is adapted from van Lier (1988) and Johnson
(1995). Language has not been corrected and standard conventions
of punctuation are not used, the aim being to represent ‘warts and all’
the exchanges as they occurred in the classroom. Many parts of the
transcripts are marked unintelligible; it should be noted that the lessons
were recorded under normal classroom conditions with no specialist
equipment. Consequently, background noise, simultaneous speech and
other types of interference have, at times, rendered the recordings
unintelligible.
T: - teacher
L: - learner (not identiﬁed)
L1, L2, etc.: - identiﬁed learner
LL: - several learners at once or the whole class
/ok/ok/ok/ - overlapping or simultaneous utterances
by more than one learner
[do you understand?]
[I see] - overlap between teacher and learner
= - turn continues, or one turn follows
another without any pause.
. . . - pause of one second or less marked by three
periods.
(4) - silence; length given in seconds
? - rising intonation – question or other
! - emphatic speech: falling intonation
((4)) - unintelligible 4 seconds a stretch of
unintelligible speech with the length given in
seconds
Paul, Peter, Mary - capitals are only used for proper nouns
T organises groups - editor’s comments (in bold type)

Glossary
Acknowledgement token
a term used in Conversation Analysis (CA) to refer to words like ok, right, yeah,
sure, mm, uhuh. These words are used by listeners to acknowledge speaker
contributions and to help communication proceed smoothly.
Asymmetrical roles
in many professional or workplace settings, speaker roles are not equal. Consider,
for example, doctor/patient or teacher/student contexts: one speaker has more
knowledge and power than the other, thus affecting the way interaction progresses.
Backchannels
a term used by discourse analysts to refer to words like ok, right, yeah, sure. These
words are essentially discourse markers that refer back to a previous contribution
in spoken interaction.
Clariﬁcation request
where one speaker asks another for clariﬁcation.
Classroom idiolect
a teacher’s individual way of talking in class.
Classroom interactional competence (CIC)
the ways in which teachers and learners use language to mediate learning (see
Walsh, 2006).
Concept checking
teachers check key concepts, usually by asking questions, as a way of ensuring
that learners have understood a particular teaching point. Example: to check the
concept of past time:
T: Mary went to London yesterday. OK? So is she still there today?
Concordancing/concordance line
a term used in corpus linguistics to describe the way in which key words are
presented in context. A concordance line allows us to see how a particular word
is used with other words, how words combine in context.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111

Confirmation check
where speakers check that they have understood a previous utterance correctly.
Content feedback
feedback (normally from a teacher) on the message, not the language used to
express it.
Corpus
a corpus is a collection of texts that is stored on a computer. These texts may be
spoken or written, but must be ‘real’ and taken from natural contexts. A corpus
can be searched quickly and easily, making it very suitable for analysing language
in use.
Direct repair
where errors are corrected quickly and directly, with little explanation or
exemplification.
Discourse marker
words like sure, ok, right, now, so, which mark spoken discourse in some way. In
classrooms, these words often mark the beginning and end of an activity, or enable
movement from one stage of a lesson to another.
Display question
questions to which teachers already know the answer. Their function is to get
learners to ‘display’ what they know about something.
Echo
repeating (a) what a learner has said for the benefit of other learners, (b) what a
teacher has already said. Echo may help the flow of the interaction or it may ‘get
in the way’ if used excessively.
Elicitation
getting responses from students, normally, but not exclusively, by asking questions.
Extended learner turn
a learner contribution of several utterances. Most contributions are very short, one
or two words only. Extended learner turns give students greater opportunity to
express themselves and say what they really want to say.
Extended teacher turn
a teacher contribution of several utterances.
Extended wait-time
the time a teacher waits after asking a question or seeking a response. Typically,
this is very short, less than one second. Extended wait-time (of 2 seconds or more)
allows learners more time to think and prepare their contribution.
Form-focused feedback
teacher feedback that focuses on language rather than message.
Glossary218

Input hypothesis
first proposed by Stephen Krashen in 1981, who argued that second language
acquisition depends largely on students receiving comprehensible input.
Interaction hypothesis
first proposed by Michael Long in 1983 and based on Krashen’s theory. Long
proposed that learners acquire a second language when they negotiate meanings
through interaction with others. This was the hypothesis on which CLT was based.
Interactional competence
see Kramsch (1986). Considers the competence speakers need to communicate
effectively and emphasises what goes on between speakers rather than solo
performance.
Interacture
(Walsh 2006). Refers to the specific interactional features of classroom discourse
that may help or hinder communication. Includes things like wait time, confirmation
check, display question, etc.
IRF Exchange
based on the work of Sinclair and Coulthard (1975), who noted that most classroom
discourse follows a three-part structure:
Teacher Initiates: what’s the past tense of go?
Student Responds: went
Teacher Follows-up/Evaluates: went, excellent
The IRF exchange is still the basic unit of all classroom interaction and can be
found in almost any classroom, anywhere in the world.
Latching/latched turn
a term used in CA to describe the way in which one turn may follow another with
no pausing or silence.
Micro-context
refers to the ways in which interactants create contexts through their talk. Each
micro-context comprises specific pedagogic goals and the language used to
achieve them.
Mode
(Walsh 2006). A mode is a particular second language classroom micro-context,
which has specific and identifiable pedagogic goals and specific linguistic and
interactional features.
Mode convergence
when teachers use language that is appropriate to a particular micro-context.
Typically, mode convergence occurs when pedagogic goals and the language used
to achieve them are working together.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Glossary 219

Mode side sequence
a brief departure from one mode (for example, classroom context) to another (for
example, skills and systems) and then back again.
Mode switching
rapid movements from one mode to another.
Modelling
teachers frequently model new language by articulating a particular word, phrase
or structure with correct pronunciation, stress and intonation. This is important for
learners if they are to acquire the new language.
Negotiation of meaning
(see also, interaction hypothesis). Meanings are negotiated through interaction and
when speakers seek clarification or confirm intended meanings. According to Long,
negotiation of meaning is central to second language acquisition (SLA).
Noticing
(c.f. Schmidt 1993). The process of making connections, links etc. between one
word, phrase or structure and another. Noticing is considered by some to be a
crucial element in the process of SLA.
Online decision-making
the interactive decisions made by teachers while ‘online’, while teaching. ‘Good’
interactive decisions are those which promote learning and SLA.
Output hypothesis
first proposed by Merrill Swain (1985), who maintained that, for SLA to occur, the
focus of attention should be on the words used by learners, their output.
Overlaps
where one utterance occurs at the same time as another, a very common feature
of all spoken interaction.
Referential question
a genuine question, one to which a teacher does not know the answer (c.f. display
question).
Reflective practice
the process of reflecting on one’s teaching and making changes to practice.
Reformulation
refers to ways in which one speaker (normally the teacher) takes an utterance of
a learner and ‘reformulates’ it so that it is more accurate, more precise. (See also
the work of Lyster on recasts.)
Repair
a form of error correction.
Glossary220

Scaffolding
a term first coined by Bruner (1983) to denote the ways in which verbal scaffolds
are used to assist learners. In a classroom, this may entail teachers ‘feeding in’
specific words or structures as and when needed by learners. It has to be said
that learners also scaffold each other.
SETT
(self-evaluation of teacher talk, (Walsh 2006)). A framework designed to allow
teachers to gain a closer understanding of interactional processes in the classroom
as a means of improving their teaching (see Appendix A).
Shaping
refers to the ways in which teachers help learners to express themselves more
clearly by ‘shaping’ their contribution: rewording, adding additional language,
modifying, etc.
Stimulated recall
a research technique that can also be adapted for professional development. It
entails making a short recording of your teaching, listening or watching with a
colleague, discussing together and recording that discussion as a way of improving
your practice.
Teacher Talking Time (TTT)
the amount of time teachers spend talking while teaching, typically around 70–80
per cent of the total class-time.
Transition markers
typically, discourse markers such as right, now, ok, so, well used to mark transitions
or changes from one part of a lesson to another or from one activity to another.
Turn completion
used in CA to indicate where one speaker completes another’s turn.
Turn taking
all human interaction is underpinned by turn-taking. Understanding the process is
the domain of CA and entails looking at the ways in which turns are held, passed,
taken or given. An understanding of turns at talk offers unique insights into the
ways in which people interact and convey meaning.
ZPD
Zone of Proximal Development. A term first proposed by Vygotsky (1978) to refer
to the difference between learning alone or learning with a more experienced
‘knower’. As Lantolf (2000: 17) puts it: ‘[the ZPD is] the collaborative construction
of opportunities for individuals to develop their mental abilities.’ Key to this is
collaborative: we learn better, proponents of ZPD argue, when we are learning with
someone else.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Glossary 221

Further reading
The following is a suggested list of additional reading that you might
like to consult. Some of the references below appear in the main text,
others have been added. The brief comment after each reference is
intended for guidance only.
Books
Alexander, R.J. (2008) Towards Dialogic Teaching: Rethinking Classroom
Talk (4th edition), York: Dialogos.
Excellent introduction to the importance of classroom interaction and talk
to learning in a variety of contexts.
Bailey, K. and Nunan, D. (eds) (1996) Voices from the Language Classroom,
Cambridge: CUP.
This is a very useful collection of papers from a variety of ELT contexts
documenting the complexities of classroom life and classroom interaction.
Burns, A. (2009) Doing Action Research in English Language Teaching: A
Guide for Practitioners. London: Routledge.
This is a practical and comprehensive guide to conducting action research
projects.
Hall, G. (2011) Exploring ELT: Language in Action, London: Routledge.
Looks at the relationship between practice and theory when considering
some of the key issues that have emerged in English Language Teaching.
Johnson, K.E. (1995) Understanding Communication in Second Language
Classrooms, Cambridge: Cambridge University Press.
A very thorough overview of the various elements that need to be taken
into account in any study of classroom discourse.
Kramsch, C. (1993) Culture and Context in Language Teaching, Oxford: OUP.
An exploration of the links between cultural knowledge and second language
learning.
Lantolf, J.P. and Thorne, S. (2006) Sociocultural Theory and the Genesis of
Second Language Development, Oxford: Oxford University Press.
Looks at the relevance of socio-cultural theories of learning to second
language acquisition, adopting a largely social view of learning.

Mortimer, E. and Scott, P. (2003) Meaning Making in Secondary Science
Classrooms, Buckingham, UK: Oxford University Press.
Offers useful insights into the ways in which classroom interaction can be
studied as a means of facilitating learning and making teaching more effective.
Nunan, D. (1989) Understanding Language Classrooms, London: Prentice
Hall.
Though quite old now, this is still a very useful and thought-provoking
volume that considers alternative ways of approaching the study of language
classrooms.
Sargeant, P. (forthcoming) World Englishes: Language in Action, London:
Routledge.
An overview of the spread and current role of English and Englishes in the
world.
Seedhouse, P. (2004) The Interactional Architecture of the Second Language
Classroom: A Conversational Analysis Perspective, Oxford: Blackwell.
An excellent book that promotes the use of conversation analysis in the
study of classroom discourse.
ten Have, P. (2007) Doing Conversation Analysis: A Practical Guide, London:
Sage.
One of the best introductions to conversation analysis currently available.
Walsh, S. (2006) Investigating Classroom Discourse, London: Routledge.
This book offers a framework for studying classroom interaction as a means
of improving second language learning and teaching.
Young, R.F. (2008) Language and Interaction: An Advanced Resource Book,
London and New York: Routledge.
An excellent resource for studying language, social interaction and human
communication.
Young, R.F. (2009) Discursive Practice in Language Learning and Teaching,
Michigan: Wiley-Blackwell.
Provides a comprehensive account of discursive practices and social context
in second language acquisition.
Journal articles and chapters in books
The following is a list of some of the key papers that have been
published on issues relating to classroom discourse in the last thirty
years.
Allwright, R.L. (1984) ‘The importance of interaction in classroom language
learning’. Applied Linguistics, 5: 156–71.
One of the most inﬂuential papers on the relationship between classroom
interaction and learning.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Further reading 223

Breen, M.P. (1998) ‘Navigating the discourse: on what is learned in the
language classroom’, in W.A. Renandya and G.M. Jacobs (eds), Learners
and Language Learning. Anthology Series 39, Singapore: SEAMO Regional
Language Centre.
Looks at the ways in which learners ‘get lost’ in classroom interaction and
at the role of the teacher in preventing this from happening.
Fung, L. and Carter, R. (2007) Discourse markers and spoken English:
native and learner use in pedagogic settings, Applied Linguistics, 28(3),
410–439.
Provides a taxonomy of the most frequently found discourse markers used
in classroom settings and offers a description of their functions.
Howard, A. (2010) ‘Is there such a thing as a typical language lesson?’
Classroom Discourse, 1: 82–100.
Considers the effect of the observer on classroom discourse and tries to
uncover what really happens in ‘typical’ language classrooms.
Larsen-Freeman, D. (2010) ‘Having and doing: learning from a complexity
theory perspective’, in P. Seedhouse, S. Walsh, C. Jenks (eds), Reconceptual-
ising Learning in Applied Linguistics, London: Palgrave Macmillan.
Provides a stimulating discussion on polarised views of learning as either
‘having’ or ‘doing’, before turning to a critique of complexity theory as a
means of seeing language as a complex adaptive system.
Markee, N. (2008) ‘Toward a learning behavior tracking methodology for
CA-for-SLA’. Applied Linguistics, 29: 404–427.
Considers the role of conversation analysis in tracking learning in spoken
classroom interactions.
McCarthy, M. J. and Walsh, S. (2003) ‘Discourse’ in D. Nunan (ed) Practical
English Language Teaching, San Francisco: McGraw-Hill.
Offers a view of classroom interaction where teacher–learner discourse is
characterised by micro-contexts or discourse modes.
Pekarek Doehler, S. (2010) ‘Conceptual changes and methodological chal-
lenges: on language, learning and documenting learning in conversation
analytic SLA research’, in P. Seedhouse, S. Walsh, C. Jenks (eds), Recon-
ceptualising learning in applied linguistics. London: Palgrave-MacMillan.
Looks at the conceptions of language and learning proposed by proponents
of CA-for-SLA, whereby conversation analysis is used in the study of second
language acquisition.
Richards, K. (2006) ‘Being the teacher’: identity and classroom conversation
. . . A response to Sheen and O’Neill (2005), Applied Linguistics 27:
135–141.
Considers the ways in which teachers’ identities are co-constructed through
their interactions with learners and other teachers.
Further reading224

Seedhouse, P. (2005) ‘Conversation analysis and language learning’, Language
Teaching 38: 1–23.
A state of the art article on the role of conversation analysis in the study
of second language learning.
Sfard, A. (1998) ‘On two metaphors for learning and the dangers of choosing
just one’, Educational Researcher, 27: 4–13.
Proposes and evaluates two metaphors for learning: learning as having and
learning as doing.
Walsh, S., O’Keefe, A. and Morton, T. (2011) ‘Analyzing university spoken
interaction: a corpus linguistics/conversation analysis approach’, Inter-
national Journal of Corpus Linguistics [waiting for volume number].
Offers a methodology for studying classroom interaction that combines
highly qualitative conversation analysis methodologies with more quanti-
tative methods using corpus linguistics.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Further reading 225

References
Ädel, A. (2010) ‘How to use corpus linguistics in the study of political
discourse’, in A. O’Keeffe and M.J. McCarthy (eds), The Routledge
Handbook of Corpus Linguistics, London: Routledge, pp. 591–604.
Alexander, R.J. (2008) Towards Dialogic Teaching: Rethinking Classroom
Talk (4th edition), York: Dialogos.
Alger, C. (2006) ‘What went well, what didn’t go so well: growth of reflection
in pre service teachers’, Reflective Practice, 7 (3), 287–301.
Allen, J.P.B., Frohlich, M. and Spada, N. (1984) ‘The communicative
orientation of language teaching: an observation scheme’, n J. Handscombe,
R. Orem and B.P. Taylor (eds), On TESOL 83: the Question of Control,
Washington D.C.: TESOL.
Allwright, R.L. (1984) ‘The importance of interaction in classroom language
learning’, Applied Linguistics, 5: 156–71.
Bailey, K. and Nunan, D. (eds) (1996) Voices from the Language Classroom,
Cambridge: CUP.
Bartlett, L. (1990) ‘Teacher development through reflective teaching’, in
J.C. Richards and D. Nunan (eds), Second Language Teacher Education,
Cambridge: Cambridge University Press.
Batstone, R. (1994) Grammar, Oxford: Oxford University Press.
Bellack, A., Kliebard, H., Hyman, R. and Smith, F. (1966) The Language of
the Classroom, New York: Teachers College Press.
Biber, D., Conrad, S. and Reppen, R. (1998) Corpus Linguistics: Investigating
Language Structure and Use, Cambridge: Cambridge University Press.
Breen, M.P. (1998) ‘Navigating the discourse: on what is learned in the
language classroom’, in W.A. Renandya and G.M. Jacobs (eds), Learners
and Language Learning, Anthology Series 39, Singapore: SEAMO Regional
Language Centre.
Bronfenbrenner, U. and Ceci, S.J. (1994) ‘Nature-Nurture reconceptualised
in developmental perspective: a bioecological model’, Psychological Review,
101/4: 568–586.
Brown, J.D. and Rodgers, T. (2002) Doing Applied Linguistics Research,
Oxford: Oxford University Press.
Brown, P. and Levinson, S. (1987) Politeness: Some Universals in Language
Usage, Cambridge: Cambridge University Press.
Brumfit, C.J. (1995) ‘Teacher professionalism and research’, in G. Cook and
B. Seidlhofer (eds), Principle and Practice in Applied Linguistics, Oxford:
Oxford University Press, pp. 27–42.
Bruner, J. (1983) Child’s Talk, Oxford: Oxford University Press.

–––– (1990). ‘Vygotsky: A historical and conceptual perspective’, in L.C. Moll
(ed), Vygotsky and Education: Instructional Implications and Applications
of Sociohistorical Psychology, Cambridge: Cambridge University Press.
Bucholtz, M. (2007) ‘Variation in transcription’, Discourse Studies, 9/6:
784–808.
Budd Rowe, M. (1986) ‘Wait time: slowing down may be a way of speeding
up’, Journal of Teacher Education, 37: 43–50.
Bygate, M. (1988) Speaking, Oxford: Oxford University Press.
Canale, M. and Swain, M. (1980) ‘Theoretical bases of communicative
approaches to second language teaching and testing’, Applied Linguistics,
1: 1–27.
Carr, D. (ed) (2006) Teacher Training DVD Series (set of 15 DVDs). London:
International House.
Carr, W. and Kemmis, S. (1983) Becoming Critical: Knowing through Action
Research, Victoria, Australia: Deakin University Press.
Carter, R. and McCarthy, M.J. (2006) Cambridge Grammar of English. A
Comprehensive Guide. Cambridge: Cambridge University Press.
Cazden, C.B. (1986) ‘Classroom discourse’, in M.C. Wittrock (ed) Handbook
of Research on Teaching, New York: MacMillan.
Chaudron, C. (1988) Second Language Classrooms: Research on Teaching
and Learning, New York: Cambridge University Press.
Cohen, L., Manion, L. and Morrison, K. (2007) Research Methods in
Education, London: Routledge-Falmer.
Cook, G. (1989) Discourse, Oxford: Oxford University Press.
Cook, V. (2001) Second Language Learning and Language Teaching (3rd
edition), London: Arnold, co-published by Oxford University Press.
Cotterill, J. (2010) ‘How to use corpus linguistics in forensic linguistics’, in
A. O’Keeffe and M.J. McCarthy (eds), The Routledge Handbook of Corpus
Linguistics, London: Routledge, (pp. 578–590).
Cullen, R. (1998) ‘Teacher talk and the classroom context’, English Language
Teaching Journal, 52: 179–187.
Dewey, J. (1933) How we think: A restatement of the relation of reﬂective
thinking to the educative process, Boston: D.C. Heath and Company.
Donato, R. (1994) ‘Collective scaffolding in second language learning’, in
J.P. Lantolf and G. Appel (eds) Vygotskyan Approaches to Second Language
Research, Norwood, New Jersey: Ablex Publishing Corporation.
Doughty, C. and T. Pica (1986) ‘“Information-gap” tasks: Do they facilitate
second language acquisition?’ TESOL Quarterly, 20/2, 305–326.
Doughty and Williams (1998) Focus on Form in Classroom Second Language
Acquisition, Cambridge: Cambridge University Press.
Drew, P. (1994). ‘Conversation Analysis’, in R.E. Asher (ed) The Encyclo-
paedia of Language and Linguistics, Oxford: Pergamon.
Drew, P. and Heritage, J. (eds) (1992) Talk at Work: Interaction in Institu-
tional Settings, Cambridge: Cambridge University Press.
Edge, J. (1992) Co-operative Development. London: Longman.
–––– (2001) Action Research, Alexandria, VA: TESOL Inc.
Edwards, A. and Westgate, D. (1994) Investigating Classroom Talk, London:
Falmer.
Ellis, R. (1990) Instructed Second Language Acquisition, Oxford: Blackwell.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
References 227

–––– (1994) The Study of Second Language Acquisition, Oxford: Oxford
University Press.
–––– (1998) ‘Discourse control and the acquisition-rich classroom’, in
W.A. Renandya and G.M. Jacobs (eds) Learners and Language Learning,
Anthology Series 39, Singapore: SEAMO Regional Language Centre.
–––– (2000) ‘Task-based research and language pedagogy’, Language Teaching
Research, 49: 193–220.
–––– (2001) ‘Investigating form-focused instruction’, in R. Ellis (ed) Form
Focused Instruction and Second Language Learning, Malden, MA:
Blackwell.
––––(2003)Task-based Language Learning and Teaching, Oxford: Oxford
University Press.
Eraut, M. (1995) ‘Schön shock: a case for reframing-in-action?’ Teachers and
Teaching: Theory and Practice, 1: 9–22.
Ferraro, J.M. (2000) Reflective Practice and Professional Development, ERIC
Clearinghouse on Teaching and Teacher Education, Washington DC. ERIC
Identifier: ED449120.
Firth, A. (1996) ‘The discursive accomplishment of normality: On “lingua
franca” English and conversation analysis’, Journal of Pragmatics, 26:
237–259.
Firth, A. and Wagner, J. (1997) ‘On discourse, communication, and (some)
fundamental concepts in SLA research’, Modern Language Journal, 81:
285–300.
–––– (2007) ‘Second/foreign language learning as a social accomplishment:
elaborations on a reconceptualized SLA’, Modern Language Journal (Special
Focus Issue on: The impact of the ideas of Firth & Wagner on SLA), vol.
91: 798–817.
Flanders, N.A. (1970) Analyzing Teacher Behaviour, Reading, MA: Addison-
Wesley.
Foster, P. (1998) ‘A classroom perspective on the negotiation of meaning’,
Applied Linguistics, 19: 1–23.
Gass, S.M. and Varonis, E.M. (1985) ‘Non-native/non-native conversations:
a model for negotiation of meaning’, Applied Linguistics, 6: 71–90.
Glew, P. (1998) ‘Verbal interaction and English second language acquisition
in classroom contexts’, Issues in Educational Research, 8: 83–94.
Goffman, E. (1981) Forms of Talk. Oxford: Oxford University Press.
Green, J. (1983) ‘Exploring classroom discourse: linguistic perspectives on
teaching-learning processes’, Educational Psychologist, 18: 180–199.
Harmer, J. (1999) ‘Abide with me: change or decay in teacher behaviour?’
IATEFL Teacher Trainer and Teacher Development Special Interest Group
Newsletter, 2: 23–27.
Hatton, N. and Smith, D. (1995) ‘Reflection in teacher education: towards a
definition and Implementation’, Teaching and Teacher Education, 11 (1),
33–49.
ten Have, P. (2007) Doing Conversation Analysis: A Practical Guide, London:
Sage.
Heritage, J. (1997) ‘Conversational analysis and institutional talk: analyzing
data’, in D. Silverman (ed), Qualitative Research: Theory, Method and
Practice, London: Sage Publications.
References228

Heritage, J. and Greatbatch, D. (1991) ‘On the institutional character of
institutional talk: the case of news interviews’, in D. Boden and D.H.
Zimmerman (eds) Talk and Social Structure: Studies in Ethnomethodology
and Conversation Analysis, Berkeley: University of California Press.
Howard, A. (2010) ‘Is there such a thing as a typical language lesson?’
Classroom Discourse, 1: 82–100.
Hymes, D. (1971) On communicative competence, Philadelphia: University
of Pennsylvania Press.
––––(1972) ‘Models of the interaction of language and social life’, In
J. Gumperz and D. Hymes (eds), Directions in Sociolinguistics: The
Ethnography of Communication, New York: Holt, Rinehart and Winston,
pp. 35–71.
Jarvis, J. and Robinson, M. (1997) ‘Analysing educational discourse: an
exploratory study of teacher response and support to pupils’ learning’,
Applied Linguistics, 18: 212–228.
Johnson, K. (1992) ‘The instructional decisions of pre-service ESL teachers:
new directions for teacher preparation programmes’, in J. Flowerdew,
M. Brock and S. Hsia (eds) Perspectives on Second Language Teacher
Education, Hong Kong: City Polytechnic of Hong Kong.
–––– (1995) Understanding Communication in Second Language Classrooms,
Kasper, G. (2001) ‘Four perspectives on L2 pragmatic development’, Applied
Linguistics, 22: 502–530.
Kemmis, M. and McTaggart, R. (1992) The Action Research Planner (third
edition), Geelong, Victoria, Australia: Deakin University Press.
Koester, A. (2006) Investigating Workplace Discourse, London: Routledge.
Kramsch, C. (1986) ‘From language proﬁciency to interactional competence’,
The Modern Language Journal, 70 (4): 366–372.
Krashen, S. (1985) The Input Hypothesis, London: Longman.
Kumaravadivelu, B. (1999) ‘Critical classroom discourse analysis’, TESOL
Quarterly, 33: 453–484.
Lantolf, J.P. (2000) Sociocultural Theory and Second Language Learning,
Lantolf, J.P. and Thorne, S. (2006) Sociocultural Theory and the Genesis of
Second Language Development, Oxford: Oxford University Press
Larsen-Freeman, D. (2010) ‘Having and doing: learning from a complexity
theory perspective’, in P. Seedhouse, S. Walsh, C. Jenks (eds) Reconceptual-
ising Learning in Applied Linguistics, London: Palgrave Macmillan.
Lee, W. and Ng, S. (2009) ‘Reducing student reticence through teacher
interaction strategy’, ELTJ, 64(3): 302–13.
Legutke, M. and Thomas, H. (1991) Process and Experience in the Language
Classroom, Harlow: Longman.
Leinhardt, G. and Greeno, J.G. (1986) ‘The Cognitive Skill of Teaching’,
Journal of Educational Psychology, 78/2: 75–95.
Levinson, S. (1983) Pragmatics, Cambridge: Cambridge University Press.
Li, L. and Walsh, S. (forthcoming) ‘“Seeing is believing”: looking at EFL
teachers’ beliefs through classroom interaction’, Language Teaching
Research, 15/1.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
References 229

Lin, A. (2000) ‘Lively children trapped in an island of disadvantage: verbal
play of Cantonese working-class schoolboys in Hong Kong’, International
Journal of the Sociology of Language, 143: 63–83.
Long, M.H. (1983) ‘Native speaker/non-native speaker conversation and the
negotiation of meaning’, Applied Linguistics, 4: 126–141.
–––– (1996) ‘The role of the linguistic environment in second language
acquisition’, in W.C. Ritchie and T.K. Bhatia (eds) Handbook of Second
Language Acquisition, San Diego: Academic Press.
Lyster, R. (1998) ‘Recasts, repetition and ambiguity in L2 classroom discourse’,
Studies in Second Language Acquisition, 20: 51–81.
Markee, N. (2008) ‘Toward a learning behavior tracking methodology for
CA-for-SLA’, Applied Linguistics, 29: 404–427.
Mann, S.J. (2001) ‘From argument to articulation’, English Teaching
Professional, 20: 57–59.
McCarthy, M.J. (2003) ‘Talking back: “small” interactional response tokens
in everyday conversation’, Research on Language in Social Interaction, 36:
33–63.
–––– (2005) ‘Fluency and confluence: What fluent speakers do’, The Language
Teacher, 29(6): 26–28.
McCarthy, M.J. and Carter, R.A. (2002) ‘This that and the other: Multi-word
clusters in spoken English as visible patterns of interaction’, Teanga, 21,
30–52.
McCarthy, M. J. and Walsh, S. (2003) ‘Discourse’, in D. Nunan (ed), Practical
English Language Teaching, San Fransisco: McGraw-Hill.
Mercer, N. (1994) ‘Neo-Vygotskyan theory and classroom education’, in
B. Stierer and J. Maybin (eds), Language, Literacy and Learning in
Educational Practice, Clevedon, Avon: Multilingual Matters/Open
University.
Mortimer, E. and Scott, P. (2003) Meaning Making in Secondary Science
Classrooms, Buckingham, UK: Oxford University Press.
Musumeci, D. (1996) ‘Teacher-learner negotiation in content-based
instruction: communication at cross-purposes?’ Applied Linguistics, 17:
286–325.
Nassaji H. and Wells, G. (2000) ‘What’s the use of “triadic dialogue”?: An
investigation of teacher-student interaction’, Applied Linguistics, 21,
376–406.
Ng, M. and Lee, C. (1996) ‘What’s different about cooperative learning? And
its significance in social studies teaching’, Teaching and Learning, 17:
15–23.
Nunan, D. (1987) ‘Communicative language teaching: making it work’,
English Language Teaching Journal, 41: 136–145.
–––– (1989) Understanding Language Classrooms, London: Prentice Hall.
–––– (1991) Language Teaching Methodology, Hemel Hempstead: Prentice-
Hall.
–––– (1996) ‘Hidden voices: insiders’ perspectives on classroom interaction’,
in K.M. Bailey and D. Nunan (eds), Voices from the Language Classroom,
Nystrand, M. (1997) ‘Dialogic instruction: when recitation become conversa-
tion’, in M. Nystrand, A. Gamoran, R. Kachur and C. Prendergast (eds),
References230

Opening Dialogue: Understanding the Dynamics of Language Learning
and Teaching in the English Classroom, New York: Teachers College
Press.
Ohta, A.S. (2001) Second Language Acquisition Processes in the Classroom:
Learning Japanese, Mahwah, NJ: Erlbaum.
O’Keeffe, A. (2006) Investigating media discourse, London: Routledge.
O’Keeffe, A. and Farr, F. (2003) ‘Using Language Corpora in Language
Teacher Education: pedagogic, linguistic and cultural insights’, TESOL
Quarterly 37(3), 389–418.
O’Keeffe, A., McCarthy, M., Carter, R. (2007) From Corpus to Classroom,
Pekarek Doehler, S. (2010) ‘Conceptual changes and methodological
challenges: on language, learning and documenting learning in conversa-
tion analytic SLA research’, in P. Seedhouse, S. Walsh, C. Jenks (eds),
Reconceptualising learning in applied linguistics, London: Palgrave-
MacMillan.
Pica, T. (1997) ‘Second language teaching and research relationships: a North
American view’, Language Teaching Research, 1: 48–72.
Psathas, G. (1995) Conversation Analysis, Thousand Oaks: Sage.
Rampton, B. (1999) ‘Dichotomies, difference and ritual in second language
learning and teaching’, Applied Linguistics, 20: 316–340.
Richards, J.C. and Nunan, D (eds) (1990) Second Language Teacher
Education, Cambridge: Cambridge University Press.
Richards, K. (2006) ‘Being the Teacher’: Identity and Classroom Conversation
. . . A Response to Sheen and O’Neill (2005), Applied Linguistics 27: 135–141.
Röhler, L.R. and Cantlon, D.J. (1996) ‘Scaffolding: a powerful tool in social
constructivist classrooms’, Online. Available HTTP: http://edeb3.educ.msu.
edu./Literacy/papers/paperlr2.html (accessed 24 June, 2005).
Russell, T. (2005) Can reflective practice be taught? Reflective Practice, 6 (2),
199–204.
Sacks, H., Schegloff, E. and Jefferson, G. (1974) ‘A simplest systematics for
the organisation of turn-taking in conversation’, Language, 50: 696–735.
Schön, D.A. (1983) The Reflective Practitioner: How Professionals Think in
Action, New York: Basic Books.
–––– (1996) Educating the Reflective Practitioner Towards a New Design of
Teaching and Learning in the Professions, San Francisco, CA: Jossey-Bass.
Schmidt, R.W. (1990) ‘The role of consciousness in second language learning’,
Applied Linguistics, 11(2): 129–58.
Schmidt, R. (1993) ‘Awareness and second language acquisition’, Annual
Review of Applied Linguistics, 13: 206–226.
Scott, M. (2008) WordSmith Tools (Version 5), Liverpool: Lexical Analysis
Software.
Seedhouse, P. (1996) ‘Learning talk: a study of the interactional organization
of the L2 classroom from a CA institutional discourse perspective’,
unpublished thesis, University of York.
–––– (1997) ‘The case of the missing “no”: the relationship between pedagogy
and interaction’, Language Learning, 47: 547–583.
–––– (2004) The Interactional Architecture Of The Second Language
Classroom: A Conversational Analysis Perspective, Oxford: Blackwell.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
References 231

Seedhouse, P., Walsh, S. and Jenks (eds) (2010) Conceptualising ‘Learning’
in Applied Linguistics, London: Palgrave-MacMillan.
Sfard, A. (1998) ‘On two metaphors for learning and the dangers of choosing
just one’, Educational Researcher, 27: 4–13.
Sinclair, J. and Coulthard, M. (1975) Towards an Analysis of Discourse,
Slimani, A. (1989) ‘The role of topicalisation in classroom language learning’,
System, 17: 223–234.
Spada, N. and Frohlich, M. (1995) COLT Observation Scheme, Sydney,
Australia: Macquarie University, National Council for Educational Research
and Training.
Sparks-Langer, G.M. (1992) ‘In the eye of the beholder: Cognitive, critical,
and narrative approaches to teacher reﬂection’, in L. Valli (ed), Reﬂective
Teacher Education: Cases and Critiques, Albany, NY: State University of
New York Press, pp. 147–160.
Stubbs, M. (1983) Discourse Analysis: The Sociolinguistic Analysis of Natural
Language, Oxford: Blackwell.
Swain, M. (1985) ‘Communicative competence: some roles of comprehensible
input and comprehensible output in its development’, in S. Gass and
C. Madden (eds), Input in Second Language Acquisition, Rowley, Mass.:
Newbury House.
–––– (2005) ‘The output hypothesis: theory and research’, in E. Hinkel (ed),
Handbook on Research in Second Language Teaching and Learning,
Mahwah, NJ: Lawrence Erlbaum.
Swain, M. and Lapkin, S. (1998) ‘Interaction and second language learning:
two adolescent French immersion students working together’, The Modern
Language Journal, 83: 320–338.
Taylor, C. (1985) Human Agency and Language, Cambridge: Cambridge
University Press.
Thompson, G. (1997) ‘Training teachers to ask questions’, English Language
Thornbury, S. (1996) ‘Teachers research teacher talk’, English Language
Tsui, A.B.M. (1996) ‘Reticence and anxiety in second language learning’, in
K.M. Bailey and D. Nunan (eds), Voices from the Language Classroom,
–––– (1998) ‘The “unobservable” in classroom interaction’, The Language
Teacher, 22: 25–26.
van Lier, L. (1988) The Classroom and the Language Learner, London: Longman.
–––– (1996) Interaction in the Language Curriculum: awareness, autonomy
and authenticity, New York: Longman.
–––– (2000) ‘From input to affordance: social-interactive learning from an
ecological perspective’, in J.P. Lantolf (ed.) Sociocultural Theory and Second
Language Learning, Oxford: Oxford University Press.
Vygotsky, L.S. (1978) Mind in Society: the development of higher
psychological processes, Cambridge: Harvard University Press.
–––– (1986) Thought and Language (new Edition by Kozulin, A.), Cambridge:
MIT.
References232

–––– (1999) Collected Works Volume 6, R. Rieber and M. Hall (eds), New
York: Plenum Press.
Wallace, M. (1991) Training Foreign Language Teachers, Cambridge:
Cambridge University Press.
–––– (1998) Action Research for Language Teachers, Cambridge: Cambridge
University Press.
Walsh, S. (2001) ‘QTT vs TTT: never mind the quality, feel the width?’ The
IH Journal of Education and Development, 10: 11–16. .
–––– (2002) ‘Construction or obstruction: teacher talk and learner involvement
in the EFL classroom’, Language Teaching Research, 6: 3–23.
–––– (2003) ‘Developing interactional awareness in the second language
classroom’, Language Awareness, 12: 124–142.
–––– (2006) Investigating Classroom Discourse, London: Routledge
Walsh, S., and O’Keeffe, A. (2007) ‘Applying CA to a modes analysis of third-
level spoken academic discourse’, in H. Bowles and P. Seedhouse (eds),
Conversation Analysis and Languages for Speciﬁc Purposes, Bern: Peter
Lang.
Walsh, S. and Lowing, K. (2008) ‘Talking to learn or learning to talk: PGCE
students’ development of interactional competence’. Paper presented at the
British Educational Research Association Conference, Herriot Watt
University, Edinburgh.
Walsh, S., O’Keefe, A. and Morton, T. (2011) ‘Analyzing university spoken
interaction: a corpus linguistics/conversation analysis approach’, Inter-
national Journal of Corpus Linguistics [waiting for volume number].
Wells, G. (1999) Dialogic Inquiry: Towards a Sociocultural Practice and
Theory of Education, Cambridge: Cambridge University Press.
White, J. and Lightbown, P.M. (1984) ‘Asking and answering in ESL classes’,
The Canadian Modern Language Review, 40: 228–244.
Wilkins, D. (1976) Notional syllabuses, Oxford and New York: Oxford
University Press.
Willis, J. (1996) A Framework for Task-based Learning, London: Longman.
–––– (2001) ‘Task-based language learning’, in R. Carter and D. Nunan (eds)
The Cambridge Guide to Teaching English to Speakers of Other Languages,
Winter, R. (1996) ‘Some principles and procedures for the conduct of action
research’, in O. Zuber-Skerritt (ed.), New Directions in Action Research,
London: Falmer.
Wray, A. (2000) ‘Formulaic sequences in second language teaching: principle
and practice’, Applied Linguistics, 21 (4): 463–489.
–––– (2002) Formulaic Language and the Lexicon, Cambridge: Cambridge
University Press.
Wu, B. (1998) ‘Towards an understanding of the dynamic process of L2
classroom interaction’, System, 26: 525–540.
Young, R.F. (2003) ‘Learning to talk the talk and walk the walk: Interactional
competence in academic spoken English’, North Eastern Illinois University
Working Papers in Linguistics, 2: 26–44.
–––– (2008) Language and Interaction: An Advanced Resource Book, London:
Routledge.
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
References 233

Zeichner, K., and Liston, D. (1996) Reﬂective Teaching: An Introduction, New
Jersey: Lawrence Erlbaum Associates.
Zuber-Skerritt, O. (1996) ‘Emancipatory Action Research for Organisational
Change and Management Development’, in O. Zuber-Skerritt (ed), New
Directions in Action Research, London: Falmer.
References234

acknowledgement tokens 72, 80, 88,
154, 165, 211, 212, 221
action research 141, 144–6
Alexander, R. J. 191
Alger, C. 139
Allen, J. P. B., Froehlich, M. and
Spada, N. 77
Allwright, R. L. 52, 192
appropriation 106, 172
asymmetrical roles 4–6, 221
backchannels 10, 80, 130, 211, 221
Bailey, K. 43–4, 46
Bailey, K. and Nunan, D. 191
Bartlett, L. 138–9
Batstone, R. 98
Bellack, A., Kliebard, H., Hyman, R.
and Smith, F. 75–6
Breen, M. P. 4, 6, 182–3, 193, 212
Bronfenbrenner, U, and Ceci, S. J.
146
Brown, J. D. and Rodgers, T. 75
Brown, P. and Levinson, S. 108
Brumfit, C. vii, 2
Bruner, J. 66
Bucholtz, M. 71–3
Burns, A. 191
Bygate, M. 58
CA-for-SLA 62, 188
Carr, W. and Kemmis, S. 141
Carter, R. and McCarthy, M. J.
101
Cazden, C. B. 52
clarification requests 5, 7, 9–10, 22,
32, 34, 40, 41, 44, 54, 55, 57–8,
60, 79–80, 113, 125–6, 131,
150–151, 154, 155, 169–70, 174,
179, 195, 199, 205, 207, 208,
211, 218, 221
classroom, meaning of 187
classroom communication 2, 3, 15,
36, 52, 57, 59, 81, 83, 160, 177
classroom context mode 113, 120–3,
126, 130–4, 135–6, 185, 205
classroom idiolect 7, 195, 221
classroom interaction, and teacher
development and learning 35–47,
51–2, 140, 185
Classroom Interactional Competence
(CIC) 3, 21, 148, 211–12;
conceptualisation of 165–77;
context-specific 166, 176–7, 187;
defined 158, 165, 221; developing
177–80
classrooms as contexts 24–6, 40–1,
48, 104
cluster analysis 97–8
COBUILD project 93
coding systems 74
Cohen, L. Manion, L. and
Morrison, K. 141
collaborative decision-making
63–4
Collins Bank of English 93, 202
COLT system (Communicative
Orientation to Language
Teaching) 77–8
communicative competence 165
Communicative Language Teaching
(CLT) 26, 54
comprehension checks 9
concept checking 12, 221
concordancing/concordance lines 91,
92, 94–5, 202, 221
confirmation checks 7–8, 9, 34, 40,
54, 60, 113, 125–6, 131, 207,
208, 222
confluence 160
content feedback 33–4, 39, 40, 45,
112, 113, 125–6, 128, 174, 179,
207, 218, 222
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Index

context, in classrooms 24–5, 35, 36,
48, 65, 73, 86, 107–9, 111, 112,
121–3, 126–8, 130–1, 133–6,
185, 205, 216; promotion of
103–4
control of classroom interaction 4–6,
20, 183
conversation analysis (CA) 62, 84–9,
112, 188, 200; see also corpus
linguistics and conversation
analysis (CLCA)
Cook, G. 133
Cook, V. 160
corpus 222
corpus linguistics 90–8, 109, 190, 222
corpus linguistics and conversation
analysis (CLCA) 98–103
Critical Classroom Discourse
Analysis (CCDA) 106–7
Cullen, R. 40–1
data: collection of 67–74, 185, 190;
ownership of 142; recording
classroom interaction 67–74, 185;
teacher-generated 46; transcribing
classroom interaction 67–74, 185;
transition markers 7, 115, 116,
136, 225; transcription system 74,
200, 220
deficit model 160
departure from lesson plan 43–4
Dewey, J. 138
dialogue: and action research 141,
144–6; and learning 59–60, 62,
64, 66, 188; SETT reflective
practice cycle 148–53
direct error correction 26, 33
direct repair 113, 118–20, 125–6,
128, 147, 150, 151, 179, 196,
199, 207, 218, 222
discourse analysis 81–4, 200, 201
discourse markers 7, 18, 98, 101,
102, 109, 114, 115, 130, 131,
203, 222
display questions 11–13, 26, 38, 39,
41, 45, 112, 113, 120, 125–7,
135, 179, 180, 185, 207, 211,
218, 222
Drew, P. 135
Drew, P. and Heritage, J. 85, 104,
110
echo 222 see also teacher echo
ecological research 143–4
Edwards, A. and Westgate, D. 86
elicitation 4, 11–14, 17, 23, 25, 37,
72, 109, 113, 154, 156, 169, 196,
211, 216, 222
Ellis, R. 3, 51, 52, 53, 61, 122, 182
ELT methodology 65
English Profile 164
error correction 14–17, 108, 169,
183; see also direct error
correction 26, 33
evaluation 18, 20, 25, 183
experiential knowledge 140–1
extended learner turn 9, 13–14, 44,
109, 113, 114, 121, 123, 125–6,
150, 168, 169, 173, 180, 198,
205, 207, 214, 218, 222
extended teacher turn 113, 114,
120, 125–6, 135, 154, 180, 207,
218, 222
extended wait time 21, 34, 125–6,
150, 168, 169, 174, 179, 207,
212, 218, 222
feedback 18, 19, 29, 34, 172, 188
FIAC system (Flanders Interaction
Analysis Categories) 76–7, 78,
200–1
fingerprints 85–6, 112, 135
Firth, A. 163
Firth, A. and Wagner, J. 62
Flanders, G. 76
form and accuracy classroom
contexts 107–8
form-focused feedback 41, 113,
125–6, 135, 180, 207, 218, 222
form-focused instruction 119
Fung, L. and Carter, R. 193
Glew, P. 36, 53
Hall, G. 191
Hatton, N. and Smith, D. 138
Heritage, J. 25, 84
Heritage, J. and Greatbatch, D.
85–6, 112
Howard, A. 107, 193
Hymes, D. 26
input hypothesis 52–4, 223
interactional architecture 25
interactional competence 158–65, 165,
211, 223; context-specific 160–4,
165, 166, 187; see also Classroom
Interactional Competence
Index236

interactional space 5, 13, 63, 66, 87,
107, 115, 116, 118, 121, 123,
128, 130, 135, 166, 168, 172,
200, 201
interaction analysis (IA) 74–80,
200–1
interaction and language learning
48–52
interaction and teaching 31–5, 52,
65–6
interaction hypothesis 54–6, 223
interactive decision-making 41–4, 51
interactures 123–9, 147, 223
IRF exchange structure 12, 13,
17–20, 20, 23, 23 n1, 76, 82, 105,
116–18, 118, 130, 135, 183–4,
185, 188–9, 223
Jarvis, J. and Robinson, M.: six
pedagogic functions of classroom
interaction 106
Johnson, K. 36, 74, 142
Johnson, K. E. 191
Kasper, G. 76
key word analysis 96–7
KeyWord-In-Context displays 94
Kramsch, C. 160, 191
Krashen, S.: input hypothesis 52–4
Kumaravadivelu, B. 67; Critical
Classroom Discourse Analysis
(CCDA) 106–7
KWIC 94
language testing 160
Lantolf, J. P. 63, 108, 183
Lantolf, J. P. and Thorne, S. 63, 191
Larsen-Freeman, D. 193
latching/latched turns 5, 12, 35, 53,
72, 87, 120, 121, 154, 174, 175,
196, 200, 223
learner interruptions 195
learner-learner interaction 51, 77
learner questions 5–6
learning: as a cognitive activity
52–62; deﬁning 49–50; and
dialogue 59–60, 62, 64, 66, 188;
as doing 158, 188; Krashen's
input hypothesis 52–4, 223;
Long's interaction hypothesis
54–6, 223; as a social process
62–6, 145, 183; Swain's output
hypothesis 58–9, 224; task-based
60, 188, 197–8; and the Zone of
Proximal Development (ZPD)
63–6
Legutke, M. and Thomas, H. 39
Leinhardt, G. and Greeno, G. 44
'let it pass' principle 163
Levinson, S. 82
Longman 93
Long, M. H. 2; interaction
hypothesis 54–6, 223
managerial mode 113, 113–16, 126,
134, 135–6
Mann, S. J. 140
Markee, N. 161, 193
materials mode 113, 116–18, 130–3,
134, 135–6
McCarthy, M. J. 88, 160, 164
McCarthy, M. J. and Walsh, S. 193
meaning and ﬂuency classroom
contexts 107–8
micro-contexts 26, 31, 47, 48, 50,
103, 108, 109, 110, 111–12, 166,
176, 184, 186, 200, 211, 223
minute papers 157 n1
modes 111–13, 126–8, 147, 184,
223
mode convergence 126, 128, 136,
167, 223
mode divergence 129, 136
mode side sequences 129, 133–5,
136, 224
mode switching 129, 130–3, 136,
224
modelling 52, 53, 65, 104, 125–6,
172, 179, 203, 207, 218, 224
Mortimer, E. and Scott, P. 192
multi-party spoken interaction 188
Musumeci, D. 56
navigating the discourse 6, 7, 98,
115, 212
negotiation of meaning 40, 41,
54–8, 224; see also interaction
hypothesis
noticing 59, 119, 224
Nunan, D. 39, 142, 143, 192
observation 47, 68, 69–70, 74–5,
78, 79, 80, 110, 142, 144, 210
observation instruments 74–5
observation schedules 69, 75, 142
Ohta, A.S. 63
O'Keeffe, A., McCarthy, M. and
Carter, R. 94, 97
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Index 237

online decision-making 25, 44,
151–2, 177, 224
oral fluency 12, 15, 32–5, 59, 65,
87, 97, 107, 118, 128, 135,
159–60, 161, 162, 165, 199–200,
210
output hypothesis 58–9, 224
overlaps 15, 22, 35, 70, 72, 74, 78,
109, 120, 121, 130, 162, 163–4,
174, 175, 189, 200, 201, 224
Park, C. 139–40
Pekarek Doehler, S. 193
Pica, T. 54
procedural classroom contexts 2–17,
107–8
Psathas, G. 112
pushed language use 58
pushed output 58–9
question strategies 37–9, 45, 173, 183
questions, display 11–13, 26, 38, 39,
41, 45, 112, 113, 120, 125–7,
135, 179, 180, 185, 207, 211,
218, 222
questions, referential 12, 38, 39, 40,
45, 77, 113, 121, 125–6, 154–5,
179, 207, 215, 218, 224
received knowledge 140
recitation script 17
recording classroom interaction 68ff,
185
referential questions 12, 38, 39, 40,
45, 77, 113, 121, 125–6, 154–5,
179, 207, 215, 218, 224
reflection in action 140–1, 143
reflective journals 139–40
reflective practice 44, 46, 185, 186,
208, 224; and SETT 146–57;
teachers' role 138–46; see also
SETT reflective practice cycle
reflective teachers 138
reformulation 9, 32, 125, 140–1,
179, 207, 218, 224
repair 14–17, 107–8, 135, 163, 169,
172, 173, 183, 211, 224; error
correction 14–17, 108, 169, 183;
direct error correction 26, 33;
direct repair 113, 118–20, 125–6,
128, 147, 150, 151, 179, 196,
199, 207, 218, 222
repetition 7, 9
research, classroom-based 143–4
research tools, users of 190
response tokens 96, 174
reticence 44–6
Richards, K. 37, 86, 193
Röhler, L. R. and Cantlon, D. J. 62
Russell, T. 139
Sacks, H., Schegloff, E. and
Jefferson, G. 84
Sargeant, P. 192
scaffolding 21, 34–5, 42, 64–5, 66,
118–20, 125–6, 145, 146, 152,
168, 172, 179, 185, 207, 218,
225
Schön, D. A. 138
Seedhouse, P. 15, 36, 81, 192, 194;
four classroom contexts 107–8
SETT framework 135, 179–80, 185,
208, 225; contexts 110;
interactures 123–9, 147, 223;
mode convergence 126, 128, 136,
167, 223; mode divergence 129,
136; mode 111–13, 126–8, 147,
184; mode: classroom context
113, 120–3, 126, 130–4, 135–6,
185, 205; mode side sequences
129, 133–5, 136, 224; mode:
managerial 113–16, 126, 134,
135–6; mode: materials 113,
116–18, 130–3, 134, 135–6;
mode: skills and systems 113,
118–20, 126, 130–4, 135–6, 152;
mode switching 129, 130–3, 136,
224; practical exercise 216–19;
and reflective practice 146–57
SETT reflective practice cycle
148–57
Sfard, A. 108, 194
shaping learner contributions 9, 55,
63–4, 105, 121, 140, 168–9, 172,
180, 190, 199, 211, 225
shared knowledge 101
shared space 101
signposting in instructions 169
silence 13, 44–5, 76, 86, 87, 107,
168, 170, 196
Sinclair, J. and Coulthard, M. 17,
18, 23, 81–2, 83, 183–4
skills and systems mode 113,
118–20, 126, 130–4, 135–6, 152
Slimani, A. 122
space for learning 166, 168, 169–70,
189
Spada, N. and Fröhlich, M. 77
Index238

Sparks-Langer, G.M. 138
speech act theory 82
speech modiﬁcation 6–11, 9, 40, 183
spoken discourse, understanding of
187–8
stimulated recall 46, 153–7, 167–8,
210, 225
Stubbs, M. 82
student-initiated talk 40
Swain, M.: output hypothesis 58–9,
224
Swain, M. and Lapkin, S. 63
task-based cycle 21–2
task-based language teaching
(TBLT) 26–31, 49
task-based learning 60, 188, 197–8
Taylor, C. 140
teacher as reﬂective practitioner
138–46
teacher development 35–47, 51–2,
140, 185
teacher echo 7, 26, 41, 44, 51,
125–6, 145, 150–1, 168, 180,
207, 218, 222
teacher-generated data 46
teacher interruptions 4, 5, 15, 21,
22, 33, 72, 78, 109, 122, 125–6,
127, 155–6, 162–4, 169, 174,
180, 189, 196, 199, 200, 207,
211, 218
teacher-learner echo 170
teacher questions 5–6
teacher-researchers 144
teacher talking time (TTT) 35, 225
teacher-teacher echo 170
Ten Have, P. 86, 192
Thompson, G. 37–9
Thornbury, S. 39–40, 46
three-part exchange system 75–6
topicalisation 122
topics 164, 197
transcribing classroom interaction
67–74, 185
transcription 70–4, 113, 190, 200,
215, 220
transcription system 74, 200, 220
transition markers 7, 115, 116, 136,
225
tryadic structure 17, 19, 23
Tsui, A. 44–6
turn completion 9, 125–6, 180, 207,
218, 225
turn-taking 4–5, 12, 20, 34, 43, 53,
64, 86–8, 99, 102, 107, 112, 115,
117–18, 120, 123, 130, 154, 161,
163, 169, 198, 201, 205, 211, 225
van Lier, L. 3, 15, 20, 51, 65, 74,
89, 104, 121, 182, 220; ecological
research 143–4; four types of
classroom interaction 105–6
variable approaches to classroom
interaction 103–9, 184
vocabulary 7
Vygotsky, L. 62–3, 66, 106, 183
wait time 13, 26, 39–40, 45, 148
Wallace, M. 75, 140
Walsh, S. 21, 79, 111–12, 113, 158,
165, 192, 194
Walsh, S. and O'Keeffe, A. 97
Walsh, S. et al. 100–1
Walsh, S., O'Keeffe, A. and Morton,
T. 99
Wells, G. 66
Wilkins, D. 26
Willis, J. 27, 28
word frequency counts 95–6
word lists 95–6
Wu, B. 83
Young, R. 161
Young, R.F. 192
Zeichner, K. and Liston, D. 138
Zone of Proximal Development
(ZPD) 63–6, 146, 225
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
35
6
7
8
9
40111
1
2
3
4
45111
Index 239

Exploring classroom discourse

More Related Content

What's hot (20)

Viewers also liked (6)

Similar to Exploring classroom discourse (20)

Recently uploaded (20)

Exploring classroom discourse