SYSTEMIC FUNCTIONAL LINGUISTICS

This user-friendly student guide is the essential resource for all those engaged in studying
systemic functional linguistics (SFL).
Assuming no prior knowledge, this guide is divided into nine chapters which can be read
independently of one another and used for purposes of reference. The reading section maps
out and mediates the key SFL literature. The application guides show how SFL has been and
can be applied to various domains, from translation to healthcare communication. The term
guides demystify the core terminology and the vocabulary guides aid readers in dealing with
the most commonly used terms in text analysis.
Systemic Functional Linguistics is an invaluable guidebook for all those studying functional
grammar and SFL within linguistics, applied linguistics and related courses.

Christian M.I.M. Matthiessen is currently Professor at University of Complutense Madrid;

Distinguished Professor, Department of Linguistics, University of International Business and
Economics and Distinguished Professor, School of Foreign Languages, Hunan University. He
has authored and co-authored over ten books and 160 book chapters and journal articles.

Kazuhiro Teruya is the author of a number of book chapters and journal articles including
several systemic functional books: two volumes of description of Japanese, a co-authored book on
key terms in SFL, an introduction to SFL in Japanese (edited and co-authored) and an edited series
of Matthiessen’s collected works.
SYSTEMIC FUNCTIONAL
LINGUISTICS
A Complete Guide

Christian M.I.M. Matthiessen

and Kazuhiro Teruya
 Designed cover image: “The Conversation”, by Martin Emond. By courtesy of
Christian M.I.M. Matthiessen
First published 2024
by Routledge
4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
605 Third Avenue, New York, NY 10158
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2024 Christian M.I.M. Matthiessen and Kazuhiro Teruya
The right of Christian M.I.M. Matthiessen and Kazuhiro Teruya to be identified as authors of this
work has been asserted 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.
Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
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
Names: Matthiessen, Christian M. I. M., author. | Teruya, Kazuhiro, author.
Title: The Routledge guide to systemic functional linguistics : terms, resources and applications /
Christian M.I.M. Matthiessen, Kazuhiro Teruya.
Other titles: Systemic functional linguistics
Description: First edition. | Abingdon, Oxon ; New York, NY : Routledge, 2023. |
Includes bibliographical references and index.
Identifiers: LCCN 2022042355 | ISBN 9781138938274 (hardback) |
ISBN 9781138938281 (paperback) | ISBN 9781315675718 (ebook)
Subjects: LCSH: Functionalism (Linguistics)
Classification: LCC P147 .M39 2022 | DDC 401/.430183—dc23/eng/20221004
LC record available at https://lccn.loc.gov/2022042355

ISBN: 978-1-138-93827-4 (hbk)

ISBN: 978-1-138-93828-1 (pbk)
ISBN: 978-1-3156-7571-8 (ebk)
DOI: 10.4324/9781315675718

Typeset in Bembo
by Apex CoVantage, LLC
 CONTENTS

Preface xi
Acknowledgements xiv

1 Introduction to the notion of a guide to SFL 1

1.1 The notion of a guide 1
1.2 SFL’s holistic approach to semiotic systems 2
1.3 The “architecture” of language in context according to SFL 7
1.3.1 The organization of the systemic functional metalanguage 10
1.3.2 The ordered typology of systems operating in different
phenomenal realms 13
1.3.3 The global semiotic dimensions organizing language in context 17
1.3.4 The local semiotic dimensions organizing stratal subsystems 20
1.3.5 Theory and description and analysis 22
1.4 Heuristics, matrices, delicacy (of focus), complementarities, helixes 23
1.5 Start with phenomena – not with the study of phenomena 31
1.6 The context of reading and using SFL 34
1.7 The organization of the guide to SFL 34
Notes 35

2 Reading guides 37
2.1 Pre-systemic literature and influences (up to around 1960) 37
2.2 Overviews of literature on systemic functional linguistics 39
2.2.1 Introductions to SFL 39
2.2.2 Companions, handbooks and anthologies 43
2.2.3 Term guides 46
2.3 Descriptions of the lexicogrammar of English 47
2.4 Variation in systemic functional linguistics: “architecture”
and terminology 50
v
 Contents

2.5 SFL and other disciplines 53

2.6 SFL and other (contemporary) approaches 58
2.6.1 Close dialogue partners 59
2.6.1.1 The Birmingham School 59
2.6.1.2 Social semiotics 61
2.6.1.3 Critical Linguistics, CDA 62
2.6.2 Fellow travellers in linguistics at a distance 73
2.6.2.1 Range and groupings of fellow travellers 73
2.6.2.2 West-Coast Functionalism 75
2.6.2.3 Corpus linguistics 76
2.6.2.4 Beyond fellow travellers 78
2.6.3 In relation to stratal subsystems 78
2.6.3.1 Phonology 79
2.6.3.2 Lexicogrammar 80
2.6.3.3 Semantics 87
2.6.4 Genre and register studies 90
2.7 Systematic literature reviews, research synthesis and meta-analysis 91
Notes 97

3 Resource guides 99
3.1 Organized according to the hierarchy of stratification 99
3.1.1 The notion of stratification 99
3.1.2 Accounts of context 102
3.1.2.1 Pre-SFL accounts of context 102
3.1.2.2 Into SFL 104
3.1.2.3 Functional diversification: field, tenor and mode 107
3.1.2.4 Context extended along the cline of instantiation 113
3.1.3 Accounts of semantics 115
3.1.4 Accounts of lexicogrammar 125
3.1.5 Accounts of phonology and phonetics (graphology and
graphetics) 131
3.1.5.1 Phonology 131
3.1.5.2 Phonetics 135
3.1.5.3 Graphology and graphetics 137
3.2 Organized according to the cline of instantiation 138
3.2.1 The stratification-instantiation matrix 139
3.2.2 Codal (variation) studies 142
3.2.3 Register (variation) studies 144
3.2.4 Semogenesis in relation to the cline of instantiation 152
3.3 Organized according to the spectrum of metafunction 155
3.3.1 Content plane resources distributed metafunctionally 157
3.3.1.1 Textual systems 157
3.3.1.2 Interpersonal systems 160
3.3.1.3 Ideational systems 162

vi
 Contents

3.3.1.4 Semantics and lexicogrammar 165

3.4 Accounts of semiotic systems other than language
(multimodal studies) 167
3.5 Organized according to axiality (the hierarchy of axis) 172
Notes 175

4 Application guides 179

4.1 Domains of research and application 179
4.2 General methodological considerations 181
4.2.1 Phenomena and the study of phenomena 181
4.2.2 The study of phenomena: stages in the research process 182
4.2.3 Research methods 184
4.2.4 Authentic: data collection – points of view: sites 195
4.3 System description 196
4.3.1 The task of describing linguistic (sub-) systems 196
4.3.2 Description as upward move along the cline of instantiation 199
4.3.3 Descriptive guidance: theoretical and typological 203
4.3.4 From sets of instances to paradigms to systems 207
4.3.5 The use of texts from different registers 208
4.3.6 A note on “corpus based” vs. “corpus driven” 221
4.4 Text analysis 222
4.4.1 Examples in the SFL literature that can serve as models 222
4.4.2 Text and options in types of text analysis 222
4.4.3 Stage of description: pre-systemic vs. systemic 229
4.4.4 Role of text: text as specimen vs. text as artefact 229
4.4.5 Means of analysis: manual vs. automated 230
4.4.6 State of text as data: raw text vs. annotated text 231
4.4.7 Phase of text: text as product vs. text as process 232
4.4.8 Multisemiosis: mono vs. multi (analysis of
multisemiotic texts (“MDA”)) 233
4.4.9 Analysis and interpretation 236
4.5 Language development, language and education 238
4.5.1 Learning how to mean: phases 239
4.5.2 The beginning of educational linguistics in Britain 242
4.5.3 Knowledge in education 246
4.5.4 Options in educational linguistics 248
4.6 Translation studies 253
4.6.1 Phases of development of SFL-informed studies of
translation 253
4.6.2 Overview of SFL-informed studies of translation 254
4.6.3 Research methodology adopted in SFL-informed studies of
translation 256
4.7 Institution of medicine (healthcare) 258

vii
 Contents

4.7.1 Clinical linguistics and healthcare communication studies 258

4.7.2 Clinical linguistics: language as clinical phenomenon 258
4.7.3 Healthcare communication studies: language as healthcare
resource 262
4.8 Summary 271
Notes 274

5 Guide to alternatives, questions, issues and debates 275

5.1 Choices among metalinguistic options 275
5.2 Theoretical issues 278
5.2.1 The theoretical architecture of language in context 279
5.2.2 Context: stratified and/or extended along
the cline of instantiation 281
5.2.2.1 Terminology 281
5.2.2.2 Context: interpretation in terms of the hierarchy of
stratification and/or the cline of instantiation 283
5.2.2.3 The nature of ideology 290
5.2.3 Content plane: the stratal status of system networks 291
5.2.4 Content plane: the number of metafunctions 293
5.2.5 Content plane: rank 296
5.2.6 Exploration of the relationship between individual meaners
and their speech fellowships 297
5.3 Descriptive issues 298
5.3.1 Interpersonal 299
5.3.1.1 Interpersonal systems: negotiation 299
5.3.1.2 Interpersonal systems – appraisal: stratal location within
content plane 301
5.3.1.3 Interpersonal systems – subject: the role of subject in the
description of different languages 302
5.3.2 Textual systems – theme: the boundary between Theme and
Rheme 304
5.3.3 Experiential systems – TRANSITIVITY (including process type) 305
5.3.3.1 The development of Halliday’s description of the system
of TRANSITIVITY 305
5.3.3.2 The system of PROCESS TYPE in particular 308
5.3.4 Logical 313
Notes 314

6 SFL term guides (metalanguage) 316

6.1 SFL terminology 317
6.2 “Tricky” terms 321
6.2.1 Theoretical terms 321
6.2.1.1 Overview the “tricky” theoretical terms 321
6.2.1.2 “Function” 321
viii
 Contents

6.2.1.3 “System” 329

6.2.1.4 “Meaning” 331
6.2.2 Descriptive terms 332
6.2.2.1 “Text” and “discourse” 333
6.2.2.2 “Clause” and “sentence” 334
Notes 335

7 Analysis guide: English lexicogrammar 337

7.1 Lexicogrammatical analysis of text 337
7.1.1 Workbooks and models 337
7.1.2 Annotation, presentation and tools 340
7.1.2.1 Boundary annotation (chunking) 340
7.1.2.2 Analysis representation 342
7.1.2.3 Tools, standards 343
7.1.3 The choice of what to include in the analysis 344
7.1.4 Interpretation of analysis 346
7.2 Clause analysis guides 348
7.2.1 Trinocular view of the clause 348
7.2.2 Constituent parts of clauses 350
7.2.2.1 The view “from below” 350
7.2.2.2 Types of probe 351
7.2.3 Correspondences across metafunctional
function structures of the clause 356
7.2.4 Thematic analysis 360
7.2.5 Modal analysis of the clause 361
7.2.5.1 Modal analysis: the mood element 364
7.2.5.2 Modal structure of the clause in relation to thematic
and transitivity structure; maximal modal structure 367
7.2.5.3 Freedom: free vs. bound clauses 368
7.2.5.4 Subject 373
7.2.6 Transitivity analysis of the clause 375
7.3 Vocabulary guides 381
7.3.1 Introduction 381
7.3.2 Verbs 384
7.3.3 Adverbs and conjunctions 386
7.3.4 Prepositions 424
7.3.5 Continuatives 429
Notes 429

8 Conclusion 430
Note 433

9 Appendix 434
9.1 Online sources 434

ix
 Contents

9.1.1 Websites 434

9.1.2 Presentations, YouTube channels 435
9.1.3 Tools 435
9.2 Conventions 436
9.2.1 System networks 436
9.2.2 Realization 441
9.2.3 Structural representation 442
9.2.4 Semantic structure 442
9.3 Transcription and annotation 442
9.4 Summary of SFL overviews 444
9.5 Summary of descriptions of the lexicogrammar of English 444
9.6 Key descriptions of different languages 444
9.7 Displays of text analysis 444
9.7.1 Chunking of texts: unit boundaries 450
9.7.2 Phonology: prosodic analysis 450
9.7.3 Lexicogrammar 452
9.7.3.1 Cohesive analysis 452
9.7.3.2 Univariate analysis 452
9.7.3.3 Multivariate analysis 454
9.7.4 Tabular analysis displays 457

References 458
Index 519

x
 PREFACE

Our Systemic Functional Linguistics is oriented towards a wide variety of readers in terms of both
field (e.g. regions of language in context and areas of application) and tenor (e.g. in terms of
expertise, ranging from university students to active scholars). It is certainly the kind of book I
would have found useful when I began my studies of linguistics (I think!).
When I first came across systemic functional linguistics (SFL) soon after I had started my
undergraduate studies at Lund University in Sweden in the mid-1970s, I was thrilled – since I
was able to view it against the background of other linguistics theories and approaches around
that time, and it offered fundamentally new insights into language as a rich resource for mak-
ing meaning. I found it hard to get hold of relevant readings, and I hadn’t even come across the
name “systemic functional linguistics”, so when I wrote an undergraduate thesis on the topic,
I called it Hallidayan Linguistics. Assuming that one could get hold of all the relevant publica-
tions informed by SFL in those days, it would in principle have been possible to read and digest
all of them, and this remained true into the 1980s (cf. Halliday and Martin, 1981, Readings in
systemic linguistics).
However, by some time in the 1990s, it was no longer possible to read all the contributions
informed by SFL, partly because there were so many of them, partly because they now cov-
ered a very wide range of areas, and partly because they were presented in a range of different
languages nobody would be likely to know all of. In response to the growing number of SFL
publications and as a resource for the SFL community, Michael Halliday and I set out to compile
a systemic functional bibliography in the early 1990s, and it was copied and further developed
by other systemic functional linguists. To keep up with the developments, I created a FileMaker
database of systemic functional publications. It now contains thousands of publications, but I’m
well aware that it is far from complete. In fact, thanks to the collective worldwide engagement
with SFL in many different languages producing a valuable flow of publications, we would need
a team – and possibly crowdsourcing – to enable us to update such a database. Even this would
not be enough; we would need computational algorithms of the kind employed by e.g. Google
Scholar and Academia. (At the same time, I am keenly aware that, as in so many other traditions,
a great many insights exist in the oral culture, having been generated in countless conversations
but not documented in writing.)
This expansion – or even explosion – of contributions within, or informed by, SFL is quite
remarkable, perhaps especially if we consider how few linguistic theories and frameworks

xi
 Preface

formulated in the 1960s are still around (cf. Matthiessen and Yousefi, 2022). SFL clearly did not
have built-in obsolescence, and while some more recent paradigms based on new technology
or conceptualizations have arguably suffered from “Kuhn loss” (cf. Post, 1971: 229, fn. 38), SFL
embodies a robust development with continuity. Using SFL, researchers keep adding descrip-
tions of “new” languages, researchers continue to develop areas of application established in the
1960s while at the same time adding new ones, and researchers open up dialogues with leading
experts in other disciplines.
This is indeed a thriving global community of scholars, researchers and students, engaging with
SFL in many contexts and in many languages. Naturally there will be different views, different
preferences for ways of theorizing and describing areas; but, remarkably, SFL has continued to
operate collaboratively, as a growing diverse thriving community. There have been no “linguis-
tic wars” (cf. Harris, 1993). While there have certainly been healthy and productive difference
opinions, and a few scholars have left the SFL community over the years to pursue different goals
and engage in other dialogues, there have on the whole been no acrimonious disagreements – no
“linguistic wars” as I just put it. This is in large part due to Michael Halliday’s deep understanding
of how science and disciplines prosper productively, and to his service as an admired role model
showing that mutual support is healthier and more productive than aggressive competition; but
it is also due to the collective sense of the fundamental importance of the goals of what Halliday
came to call “appliable linguistics”.
As I noted earlier, it is no longer possible to read and engage with the incredibly rich and
voluminous range of contributions informed by SFL across an expanding spectrum of areas
of theory, description, analysis and application. And there are a number of fairly recent edited
handbooks, encyclopaedias and thematic edited volumes that include important contributions
with expertise in different particular areas. Our guide is thus a different kind of contribution,
reflecting our attempt to provide an integrated account presenting one overview, even if flawed.
Here Bertie Russell’s (1947) comments introducing his magnificent history of Western Philoso-
phy are immediately relevant:

Apology is due to the specialists on various schools and individual philosophers. With
the possible exception of Leibniz, every philosopher of whom I treat is better known
to some others than to me. If, however, books covering a wide field are to be writ-
ten at all, it is inevitable, since we are not immortal, that those who write such books
should spend less time on any one part than can be spent by a man who concentrates
on a single author or a brief period. Some, whose scholarly austerity is unbending, will
conclude that books covering a wide field should not be written at all, or, if written,
should consist of monographs by a multitude of authors. There is, however, something
lost when many authors co-operate. If there is any unity in the movement of history,
if there is any intimate relation between what goes before and what comes later, it is
necessary, for setting this forth, that earlier and later periods should be synthesized in
a single mind. The student of Rousseau may have difficulty in doing justice to his
connection with the Sparta of Plato and Plutarch; the historian of Sparta may not be
prophetically conscious of Hobbes and Fichte and Lenin. To bring out such relations
is one of the purposes of this book, and it is a purpose which only a wide survey can
fulfil.

In other words, as in SFL in general, complementarities (Halliday, 2008) are absolutely cen-
tral: to understand and appreciate the amazingly wide range of contributions, both theoretical
and applied, within SFL, we need complementary accounts. In our account here, we have tried

xii
 Preface

to foreground the kind of overview that is (hopefully!) coherent and consistent because it is
produced by only two voices, in this way complementing recent edited overviews by groups of
specialists. One obvious challenge is that systemic functional linguists around the world keep
contributing new findings, publishing them in various languages. To keep up, we need to rely
on collective initiatives and search engines.
Halliday used to say that he was probably the last generation of linguists who could be
generalists – for example, in terms of the hierarchy of stratification, contributing to contexts,
semantics, lexicogrammar and phonology, as he did. (In Halliday’s case, also generalist in terms
of his notion of appliable linguistics.) I have tried to be a generalist myself because I find so many
aspects of language fascinating and because the holistic view is very helpful, a way of counteract-
ing what David Bohm (1979) called the “fragmentation of knowledge” and a basis for systems
thinking. However, precisely because I have tried and keep trying, I have a very strong sense of
what Halliday meant!
But, fortunately for me, this guide to SFL has been a collaborative project. Kazuhiro Teruya
and I have worked on all aspects of the book together. Among other benefits, this has meant
that while we have focussed on systemic functional linguistic publications in English, we have
together been able to range over and check work in several other languages, with Kazu adding
Japanese and Chinese to our collective multilingual meaning potential. This addition has been
helpful as we have tried to discern the outline of SFL since quite a few significant contributions
have appeared in Chinese and Japanese – one being the recent book on SFL in Japanese he has
just published (Teruya, 2022), with chapters written by him in Japanese, translated by him into
Japanese (by Michael Halliday, Heidi Byrnes, John Bateman and myself) – the whole manuscript
having been shaped and edited by him. Like the preparation of the present book, this turned out
to be an enormous task, one he had to interleave with this book.
We hope that our Systemic Functional Linguistics will be a useful resource, a real guide to SFL.
We have designed it so that it provides an index into the literature, including general introduc-
tions, handbooks and encyclopaedias.
Christian M.I.M. Matthiessen

References
Bohm, David. 1979. Wholeness and the implicate order. London: Routledge and Kegan Paul.
Halliday, M.A.K. 2008. Complementarities in language (Halliday Centre Series in Appliable Linguistics).
Beijing: The Commercial Press.
Halliday, M.A.K and J.R. Martin (eds.). 1981. Readings in systemic linguistics. London: Batsford.
Harris, Randy Allen. 1993. The linguistics wars. Oxford: Oxford University Press.
Matthiessen, Christian M.I.M. and Moslem Yousefi. 2022. Systemic functional linguistics as a resource in
appliable linguistics: New applications. Language, Text, Context, 4(1): 114–145, January.
Post, Heinz R. 1971. Correspondence, invariance and heuristics. Studies in History and Philosophy of Science
2: 213–255.
Russell, Bertrand. 1947. History of western philosophy. London: George Allen and Unwin.
Teruya, Kazuhiro (ed.). 2022. Imi ga yoku wakaru yooni naru tame no gengogaku: Taikei gengogaku eno shootai
(Linguistics for a better understanding of meaning: An introduction to systemic functional linguistics).
Tokyo: Kuroshio Publishers.

xiii
 ACKNOWLEDGEMENTS

The authors would like to express gratitude for permissions to reproduce material in the
book as noted:
Figure 2.4 from © M.A.K. Halliday 1978. Language as social semiotic: The social
interpretation of language and meaning. Figure 1. London: Edward Arnold; Figure 5.9 from
© M.A.K. Halliday 1978. Language as social semiotic: The social interpretation of language
and meaning. Table 3. London: Edward Arnold; Figure 3.2 from © M.A.K. Halliday 2003.
Introduction: On the architecture of human language. In M.A.K. Halliday, On language and
linguistics. p. 13. Volume 3 of the collected works of M.A.K. Halliday, edited by Jonathan J.
Webster. Continuum, an imprint of Bloomsbury Publishing Plc.; Figure 3.5 from © M.A.K.
Halliday 2002. Computing meanings: Some reflections on past experience and present pros-
pects. Figure 7. In Guowen Huang and Zongyan Wang (eds.), Discourse and language func-
tions. Shanghai: Foreign Language Teaching and Research Press; Table 5.7 from © M.A.K.
Halliday 1985. Dimensions of discourse analysis: Grammar. p. 45. In Teun A. van Dijk (ed.),
Handbook of discourse analysis. Volume 2. New York: Academic Press, an imprint of Else-
vier; Figure 3.13 from © M.A.K. Halliday 1997. Linguistics as metaphor. Figure 2. In Kristin
Davidse, Dirk Noel and Anne-Marie Simon-Vandenbergen (eds.), Reconnecting language:
Morphology and syntax in functional perspectives. Published by John Benjamins Publishing
Company; Table 4.4 from © M.A.K. Halliday 1998. On the grammar of pain. Table 12.1.
Functions of Language 5.1. Published by John Benjamins Publishing Company; Figure 5.1
from © M.A.K. Halliday, 1961. Categories of the theory of grammar. p. 244. Word 17.3.
Routledge, an imprint of Taylor & Francis Group. Reproduced by permission of Taylor &
Francis Group. [This is still pending]; Figure 2.7 from © Norman Fairclough 2010. Critical
discourse analysis: The critical study of language. p. 133. London and New York: Rout-
ledge, an imprint of Taylor & Francis Group. Reproduced with permission of the Licensor
through PLSclear; Table 4.6 from © Robert E. Longacre 1974. Narrative vs other discourse
genres. Chart 1. In Ruth Brend (ed.) Advances in tagmemics. Amsterdam: North-Holland, an
imprint of Elsevier; Figure 5.2 from © Robin Fawcett 2013. Choice and choosing in systemic-
functional grammar. Figure 6.2. In Lise Fontaine, Tom Bartlett and Gerard O’Grady (eds.),
Systemic functional linguistics: Exploring choice. Cambridge: Cambridge University Press.
Reproduced with permission of the Licensor through PLSclear; Figure 5.7 from © Robin
Fawcett 2000. A theory of syntax for systemic functional linguistics. Figure 4. Amsterdam.

xiv
 Acknowledgements

Published by John Benjamins Publishing Company; Figure 5.3a from © J.R. Martin 1992.
English text: System and structure. Amsterdam. Published by John Benjamins Publishing
Company; Figure 5.5 from © J.R. Martin 2010. Semantic variation: Modelling realisation,
instantiation and individuation in social semiosis. Figure 1.9. In Monika Bednarek and J.R.
Martin (eds.), New discourse on language: Functional perspectives on multimodality, identity,
and affiliation. Continuum, an imprint of Bloomsbury Publishing Plc.; Figure 5.6 from ©
J.R. Martin 2010. Semantic variation: Modelling realisation, instantiation and individuation
in social semiosis. Figure 1.12. In Monika Bednarek and J.R. Martin (eds.), New discourse
on language: Functional perspectives on multimodality, identity, and affiliation. Continuum,
an imprint of Bloomsbury Publishing Plc.; Figure 9.5 from © M.A.K. Halliday 1985. Dimen-
sions of discourse analysis: Grammar. p. 32. In Teun A. van Dijk (ed.), Handbook of discourse
analysis. Volume 2. New York: Academic Press, an imprint of Elsevier; Figure 9.7 from ©
M.A.K. Halliday and Ruqaiya Hasan 1976. Text 1 on p. 340. Cohesion in English. Informa
UK Limited. Reproduced by permission of Taylor & Francis Group.

xv
 1
INTRODUCTION TO THE
NOTION OF A GUIDE TO SFL

1.1 The notion of a guide

This book is intended to serve as a gateway for anyone interested in engaging with language and
other semiotic systems drawing on systemic functional linguistics (SFL) by reading the litera-
ture, doing discourse analysis, developing descriptions of languages, contributing to the general
theory – or undertaking any other kind of work where a map of the systemic functional semiotic
territory will be helpful.
Since the 1960s, systemic functional linguistics has developed into a powerful resource
for engaging with language and other semiotic systems operating in context – an ecological
approach to semiotic systems, i.e. systems of meaning. It has expanded considerably in coverage
in a number of ways:

• the theory has expanded in coverage to include all the semiotic dimensions that have been
identified in the organization of semiotic systems – in terms of the hierarchy of stratifica-
tion, the spectrum of metafunction, the cline of instantiation and so on;
• descriptions have also expanded in coverage – of English, but also of an increasing number
of other languages since Halliday’s early work on Chinese and the descriptions of languages
spoken in Africa in the 1960s (Yoruba, Mbembe, Nzema);
• the range of applications has also increased, quite dramatically; in the 1960s, systemic
functional linguists contributed to linguistic theory, descriptions of particular languages,
problems in education and in translation, and since then many other areas have been added,
including computational modelling, administration, the media, healthcare and clinical
applications and the law;
• the community of contributors has also expanded significantly from groups of linguists in
Britain to groups around the world – thus also leading to an expansion of the range of lan-
guages in which systemic functional work is constructed and presented, now including – in
addition to English – French, German, Danish, Norwegian, Swedish, Spanish, Portuguese,
Italian, Indonesian, Vietnamese, Chinese and Japanese.

As SFL has expanded in these ways, the need for publications giving students, intermediate and
advanced users, and active contributors access to the growing body of ideas and literature has

1 DOI: 10.4324/9781315675718-1
 Introduction to the notion of a guide to SFL

increased. Thus scholars have produced introductions, workbooks, handbooks, companions,

glossaries and similar resources; but there is as yet no “guide to SFL”. This is a significant gap
we hope to fill with this book.
But what is a guide? How does it differ from an introduction, a manual or a handbook?
Searching the web reveals wordings such as a guide to the good life, a guide to recognizing your
saints, a guide to the thesis-writing that is a guide to life, a guide to watching the N.F.L. playoffs, a guide
to apprenticeship, a guide to the orders of trilobites and a guide to Berlin. These may suggest a sense
of what a guide is or does: given a map of concrete or abstract space, it helps readers navigate
through this space to reach certain destinations – certain objectives. This is different from an
introduction, where the emphasis is arguably on sketching the first outlines of the map of the
territory to be explored (but cf. Feez, 2010, Montessori and early childhood: A guide for students).
It is perhaps more like a manual or a handbook, and one of the book titles a web-search finds is
indeed The social justice advocate’s handbook: A guide to gender. However, in the particular domain
of SFL, there is now a “handbook”, which is a different kind of contribution from our guide (see
Chapter 2). Perhaps our guide is like Using archives: A guide to effective research. Indeed, we could
imagine calling the book Using SFL: A guide to effective research and application and we would have
a good model: Butt et al.’s (1995) Using functional grammar: An explorer’s guide, although that book
is more of an introduction to “functional grammar”, specifically in the context of education.
Still, the title of this popular book makes the important connection between exploring and
guide. Another good model is Taylor’s (1998) guide for “Italian/English translators”, which is
both theoretical and practical, as a good guide needs to be. He provides background to models
and descriptions of both language and context, with SFL as a central source, and offers guidance
in translation between Italian and English, in both directions, importantly providing examples
drawn from a number of registers.
Our guide to SFL thus complements introductions, workbooks, handbooks and companions,
while at the same time drawing on them and referencing them. Our objectives are:

• to offer maps of SFL – its “architecture” of language and other semiotic systems, its
literature;
• to give users of SFL navigational help – pointers needed in studying SFL and putting it
to work in research and application.

Our guide is thus designed to enable users to get as much mileage as possible out of SFL as a
resource for engaging with language and also other semiotic systems.

1.2 SFL’s holistic approach to semiotic systems

SFL can be located on a map showing developments of linguistics, language sciences, applied
linguistics, applied language sciences and semiotics since the 1960s, and we will return to this
task; but it is helpful to take a step back so that we can see more general trends in the develop-
ment of modern science, and relate SFL to one of these trends, so that it becomes possible to
identify useful similarities across a number of disciplinary approaches.
Modern science emerged in the western part of Eurasia around five centuries ago, and it has
come to be dominated by one general kind of approach to the study of phenomena of different
kinds – the approach that has come to be known as Cartesian Analysis (see e.g. Capra, 1996).
This approach is to dissect a phenomenon into component parts, drilling down until compo-
nents that are small enough to be subjected to scientific analysis have been reached. In linguistics,
Cartesian Analysis has manifested itself in various ways – arguably, in modular architectures of

2
 Introduction to the notion of a guide to SFL

language characteristic of various generative models, but undoubtedly in the focus on constitu-
ent structure since the 1930s. It represents an aspect of what Halliday (1977a) has called the
image of language as rule (cf. also Martin, 2017: 128), and Seuren (1998) has characterized
as formalism – both interpreting these related concepts as one strand in the development of
Western Linguistics.
Cartesian Analysis contrasts with another approach to the study of phenomena of different
kinds, one which Capra (1996) calls a holistic worldview, or in a related sense, an ecological
view. This involves systems thinking rather than componential analysis, and researchers tend to
try to develop unified theories to transcend the fragmentation of knowledge that results from
the increasing specialization encouraged by Cartesian Analysis (as diagnosed by Bohm, 1980; for
an example of an attempt to overcome the problems in psychology through the development of
a unified theory, see Henriques, 2003, 2011). It is characteristic of ecological studies, but also
of the trans-disciplinary approach of general systems theories – a recent manifestation of which
is the study of complex adaptive systems promoted by the Santa Fe Institute.1 In linguistics, the
holistic approach has manifested itself in various ways – for example, in relational conceptions
and models of language, where linguistic categories are defined by the relations they enter into
(as in Stratificational Linguistics and systemic functional linguistics), in approaches where con-
text is taken into account together with all the subsystems of language, in approaches where
language is interpreted ecologically – so functionally, both in terms of its internal organization
and in terms of its adaptation to context. The holistic approach represents an aspect of what
Halliday has called the image of language as resource, and Seuren (1998) has characterized as
ecologism.
As linguistics had gradually come to be dominated by the highly specialized orientation that
emerged with generative linguistics by the mid-1970s, Halliday (1975c/2003a: 74–75) com-
mented on the two possible ways for linguistics to make a unique contribution, suggesting that
specialization was not the only strategy:

To be highly specialized is in its way a kind of defence, a means of protecting one’s

identity. We are constantly being reminded of how many others feed on language: phi-
losophers, psychologists, rhetoricians, speech pathologists, communications and media
experts, and many more besides. Yet there are significant aspects of language that none
of these groups takes account of; so linguists have tended to retreat and to consolidate
the terrain that is out of others’ reach.
...
It must be said at the outset that there have always been linguists whom this image did
not fit. But it is true that during the 1960s the majority, in the United States at least, did
adopt a highly specialized work style. Faced with a choice between two ways of being
unique, one that of seeing language in the round, from all angles (where disciplines for
which language is an instrument, not an object, see only one or two angles), the second
that of seeing what lies at the core of language (to which the other disciplines do not pen-
etrate), linguists chose the second, in this way marking out their own area of specialization
and using the discipline ‘linguistics’ rather than the object ‘language’ to characterize their
domain. Linguistics became the study of linguistics rather than the study of language.
Associated with this specialization was a determination that linguistics should be
useless, that it was a theory without applications.

He makes related but complementary comments in an interview conducted by Heloise Collins

and Geof Thompson [GT] in 1998 (in Martin, 2013b: 164):

3
 Introduction to the notion of a guide to SFL

MAKH: Yes; if what you had in mind with ‘isolating’ was the mainstream tradition from Bloom-
field via Chomsky in North America, with its insistence on autonomous syntax, with
the way that they took language as a thing in itself, rather than as some element in a
wider social system and process, then I think that’s fair enough.
GT: In the isolating tradition, socio-linguistics and pragmatics, for example, become things
you can push aside if you’re not interested in them, whereas, within the systemic-
functional approach, you can’t.
MAKH: That’s absolutely it. In a sense, the only reason why that tradition created sociolin-
guistics and pragmatics was because these weren’t in the theory of language in the first
place, where they should have been. And I always said that we didn’t need a concept
of sociolinguistics, because our concept of linguistics always was ‘socio’. And similarly
with pragmatics: to me this has always been simply the instantial end of the semantics.
We don’t need a separate discipline. Another dimension of the isolation, of course, is
the isolation between system and text. If you’re focusing on the system, the text is just
data, which has no place in the theory. Then when somebody does want to come and
study the text, they do it under a totally different disciplinary banner and both sides lose.

(On the conception of the domain covered by “sociolinguistics”, see also Hasan, 2009b, espe-
cially Chapter 1, ‘Wanted: a theory for integrated sociolinguistics’.) It is clear that SFL is located
within the general strand of linguistics where language is viewed holistically as a resource and is
theorized ecologically as a system of systems, and that SFL scholars have continuously worked
towards a unified theory (see e.g. Halliday, 1993b; Halliday and Matthiessen, 2006; Matthies-
sen, forthcoming a). But it is also worth noting that SFL embodies something approaching a
synthesis. For example, while language is not viewed as rule and it is not represented by some
type of rule system, Halliday and other systemic functional linguists have developed forms of
theoretical representations that are more explicit and formalized than what is often found in
“functional linguistics” in general (certainly in the West-Coast Functionalist tradition) – one
key example being system networks with realization statements, which have been modelled and
implemented in computational language systems. Thus if we were to review Seuren’s (1998)
discussion of strengths and weaknesses with formalism and ecologism, we would find that SFL
has avoided weaknesses characteristic of ecologism – and also of formalism.
The holistic, unified theory of language in context is a resource for developing compre-
hensive descriptions of particular languages in their contexts. Thus the theory is an account
of language as a human system – as a higher-order semiotic system, but descriptions are con-
cerned with particular languages (see Sections 1.3.5 and 4.3). The balance between theory and
language is such that while the theory is designed to empower descriptivists, it should not force
them to impose categories posited originally in the description of Latin, English, Chinese or any
other “major” language that has been very influential in descriptive work: descriptive categories
such as Theme, Subject, Actor or tense, aspect must be empirically grounded in text evidence
from the language under description. Just as the general theory of language should be holistic,
descriptions of particular languages should be comprehensive. Descriptions should be thick
descriptions in Ryle’s (1971: Chapters 36 and 37) sense, taken up by Geertz (1973: Chapter 1):
this notion is discussed with reference to Subject in Halliday and Matthiessen (2014: 145–150).
Comprehensive descriptions make it possible to identify fractal patterns such as the dispersal of
the categories of expansion and projection throughout the content system of English (Halliday
and Matthiessen, 2006; Halliday and Matthiessen, 2014: Chapter 10; Matthiessen, 1995a). Thus
like holistic theory, comprehensive descriptions support systems thinking, and can be developed
by means of the strategies of systems thinking.

4
 Introduction to the notion of a guide to SFL

At the same time, holistic theory and comprehensive descriptions ensure the appliability of
SFL. While Halliday didn’t introduce the term appliable linguistics until the early 2000s
(e.g. Halliday, 2002b, 2008a, 2015), creating a theory and descriptions that could serve as
resources for applications designed to solve a wide range of problems in communities had been
part of the agenda since the 1950s, and developments in SFL can often be linked to efforts to
solve problems.2 One important aspect of the goal of creating an appliable kind of linguistics is
the sense of social accountability (e.g. Halliday, 1984d). The readiness to apply the resources of
SFL to problems in education, healthcare, law and various other institutions differentiates it from
a number of other currents in “functional linguistics”. See further Sections 2.4, 2.6.1.3 and 4.1.
Since SFL is holistic in nature, interpreting phenomena ecologically and striving towards
unified theories and descriptions, the first step is to locate language (and other semiotic) systems
within an ordered typology of systems operating in different phenomenal realms – physical,
biological, social and semiotic systems. This is done to ensure (i) that language is understood not
only semiotically, but also socially, biologically and physically; (ii) that the properties ascribed to
language are located systematically within the relevant order of system; and (iii) that the condi-
tions are optimal for dialogue with workers within other disciplines concerned with language in
one of its guises as a lower-order system.
Halliday (2003c: 2) describes this ordered typology as follows:

A language is a system of meaning – a semiotic system. Here, as in all my writing,

“semiotic” means ‘having to do with meaning (semiosis)’; so a system of meaning
is one by which meaning is created and meanings are exchanged. Human beings use
numerous semiotic systems, some simple and others very complex, some rather clearly
defined and others notably fuzzy. A language is almost certainly the most complicated
semiotic system we have; it is also a very fuzzy one, both in the sense that its own limits
are unclear and in the sense that its internal organization is full of indeterminacy.
What other kinds of system are there? I shall assume there are three: physical, biologi-
cal and social. One way to think of these is as forming an ascending order of complexity.
A physical system is just that: a physical system. A biological system, on the other hand,
is not just that; it is a physical system (or an assembly of physical systems) having an
additional feature, let us say “life”. A social system, in turn, is an assembly of biological
systems (life forms) having a further additional feature – which we might call “value”:
it is what defines membership; so, an assembly of life forms with a membership hierar-
chy. So a social system is a system of a third order of complexity, because it is social and
biological and physical. We could then think of a semiotic system as being of a fourth
order of complexity, being semiotic and social and biological and physical: meaning is
socially constructed, biologically activated and exchanged through physical channels.

Other, more detailed, accounts can be found in Halliday (1996a), Halliday and Matthiessen
(2006), Cléirigh (1998), Matthiessen (2007b, 2009c, forthcoming a). We have summarized the
properties of the four orders of system in Table 1.1, based on Matthiessen (forthcoming a),
where the properties are presented and discussed and references are made to comparable propos-
als for ordered typologies of systems and unified theories coming from other disciplines. Physi-
cal and biological systems are both material in nature; they can be observed and measured in
material terms. In contrast, social and semiotic systems are both immaterial in nature; Halliday
(2003c, 2005b) uses “semiotic” as a cover term for them, thus making the contrast between mat-
ter and meaning. More precisely, he suggests that systems of all orders are in a sense mixtures of
matter and meaning, with matter dominating in physical systems and meaning in semiotic ones.

5
 Table 1.1 Properties of systems of increasing complexity operating in different phenomenal realms (adapted from Matthiessen, forthcoming a)

Property Material Immaterial

Physical systems Biological systems Social systems Semiotic systems

Primary Higher-order
genesis emergence Big Bang, c. 13.8 beginning, c. 3.5 billion multiple times of multiple times of c. 200–250 K years ago,
billion years ago years ago emergence emergence in Homo sapiens
sapiens (AMHs)

Introduction to the notion of a guide to SFL

mode of cosmogenesis + evolution
cosmogenesis
phenomenal realm (domain) cosmos planet earth certain biological certain social groups groups of Anatomically
populations Modern Humans
(AMHs)
distinguishing characteristic extension in + life (individuation, + value (social order) + meaning + metafunctional modes
6

characteristic property space-time self-replication) of meaning

characteristic relational role networks stratification: + “form” between
form of content and outer substance strata:
organization expression lexicogrammar and
phonology
instantiation potential event potential action potential behaviour potential meaning potential
instance event (biological) act behaviour (social act) act of meaning (semiotic act)
individuation individual – organism (“doer”, “agent”) person (aggregate of meaner (aggregate of semiotic roles; “speaker/
social roles or personae; listener”
“social agent”)
collective – (biological) population social groups < society meaning groups < speech fellowship (“speech
community”)
 Introduction to the notion of a guide to SFL

We shall not discuss the ordered typology of systems further here – we have indicated sources of
fuller discussion earlier; but we would like to emphasize again that it is an important resource as we
view language and other semiotic systems from diferent angles (trinocularly, as it were) and as we
engage in dialogues with scholars coming from disciplines other than linguistic or semiotic ones.
Based on the ordered typology of systems sketched earlier, the next step is to strive towards
a holistic account of language at its own systemic order – towards a holistic account of language
as a semiotic system and of other semiotic systems, a unified theory.

1.3 The “architecture” of language in context according to SFL

To achieve a holistic account of language – or rather of language in context, Halliday and
other systemic functional linguists have proposed an “architecture” that is based on semiotic
dimensions in terms of which different relations can be defined (cf. Halliday, 2002a, 2003c;
Hasan, 2009a, 2015, 2019; Martin, 1992a, 1999a, 2006, 2010, 2013c; Halliday and Matthiessen,
1999/2006; Matthiessen, 2007a, forthcoming a).3 The dimension that differentiates semiotic
systems as 4th-order systems from social systems as 3rd-order systems is the hierarchy of strati-
fication, and this dimension is characterized by the relation of realization (cf. Butt, 2008, on
realizational systems).
The hierarchy of stratification is one of the semiotic dimensions that organizes language in
context globally; it orders all of language in context into a number of stratal subsystems, ranging
from context through the content strata of language to the lowest of the expression strata, where
language interfaces with the expressive resources of the body – phonetics (or graphetics, or sign).
The other global dimensions of organization are the cline of instantiation and the spectrum of
metafunction. The cline of instantiation extends language in context in “phases” from poten-
tial to instance, with intermediate patterns: the cline is a continuum from the meaning potential
of language embedded in the cultural potential of a community to acts of meaning unfolding as
texts in their contexts of situation: see Figure 1.1.
The spectrum of metafunction disperses language in context into different modes of
meaning – field/ideational, tenor/interpersonal, and mode/textual. This spectrum is mani-
fested throughout context and the content plane of language, but it also penetrates into the
expression plane (in particular the highest units of the expression plane). These global dimen-
sions are introduced in Section 1.3.3, and discussed in more detail in Section 3.1.
Each of the global dimensions “slices” the overall resources of language in context in a par-
ticular way, and this can be seen when we combine them intro matrices – in particular, the
stratification-instantiation matrix and the function-stratification matrix: see Chapter 3. The hier-
archy of stratification organizes language in context into an ordered series of stratal subsystems:

context (see Section 3.1.1)

language
content plane
semantics (see Section 3.1.2)
lexicogrammar (see Section 3.1.3)
expression plane
phonology (graphology, sign1) (see Section 3.1.4)
phonetics (graphetics, sign2) (see Section 3.1.4)

Each of these stratal subsystems is organized internally according to the same architectural plan
in terms of two local dimensions – the hierarchy of rank and the hierarchy of axis. Since the

7
 Introduction to the notion of a guide to SFL

Figure 1.1 The cline of instantiation extending through context and language

8
 Introduction to the notion of a guide to SFL

plan is the same for each stratal subsystem, we can say that rank and axiality are fractal principles
of organization; they are local dimensions that are manifested in diferent stratal environments. In
systemic functional accounts, the hierarchy of rank is usually given priority over axiality in terms of
defining the domain of operation of systems (see e.g. Matthiessen, forthcoming a, for an overview
and references to various explorations, in particular in computational systemic functional model-
ling): each stratal subsystem is as it were stratified internally by the hierarchy of rank, but here the
local principle is one of composition; for example, in the description of the lexicogrammar of
many languages, it makes sense to posit a rank scale of clause > group > word > morpheme, i.e.
a clause consists of groups, a group consists of words, and a word consists of morphemes. (In many
languages, the rank between clause and word involves both phrases and groups.)
Each unit of the hierarchy of rank of a stratal subsystem is the point of entry for axis, or more
specifically, each primary class of each unit is the point of entry of the paradigmatic or systemic
axis of the dimension of axiality. Thus, the lexicogrammatical stratal subsystem of a language
is organized internally into a compositional hierarchy of units – a rank scale, say clause >
group > word > morpheme; and clause is the point of entry for the organization along the
paradigmatic axis, represented by means of system networks. This is what makes it possible to
construct lexicogrammatical function-rank matrices showing the overall distribution of lexico-
grammatical systems such as theme, mood and transitivity in terms of rank and metafunction,
as shown and discussed later in reference to Figures 1.7 and 3.9.
We can now relate the local semiotic dimensions to the global ones, and locate the global ones
within the ordered typology of systems, as shown in Figure 1.2.4 The figure includes one more
principle that is important to take account of: what we say about the “architecture” of language in
context is construed in terms of the systemic functional metalanguage (just as alternative accounts
are construed in terms of the metalanguages of other traditions). But metalanguage is itself orga-
nized as a semiotic system; crucially in the current context, it is stratified (and like language, it is
“embedded” in context – metacontext, i.e. the context of doing linguistics or more generally of
doing semiotics), as discussed by Halliday and Matthiessen, Matthiessen (2007a, forthcoming a); cf.
also Teich (1999a). The highest stratum of metalanguage is the stratum of theory, but theoretical
specifications are realized by lower-order specifications. For example, the theory of paradigmatic
organization is realized by the theoretical form of representation known as system networks.
The conception of the organization of language in context goes back to Halliday’s (1961)
pioneering presentation of the theory of grammar – presented in Berry’s (1975/7) two volumes
of introduction to what came to be known as scale-&-category theory. This early account has
been modified and expanded; and some of the terms have been replaced – an important one
being Firth’s “exponence”, which was replaced by “realization”, taken from Lamb’s (e.g. 1966)
stratificational theory (see Halliday, 1992c).
According to the current version we have just sketched, there are three kinds of semiotic
dimension – hierarchy, cline and spectrum; these are set out in Table 1.2.

Table 1.2 Semiotic dimensions – global and local

Type of dimension Characteristic Relation Domain of operation

Global Local
hierarchy ordered series realization stratification rank, axiality
cline continuum “specialization”5 instantiation delicacy (paradigmatic)
metafunction dispersal unification metafunction –

9
 Introduction to the notion of a guide to SFL

Systemic functional linguists have sometimes used the terms for the diferent types of dimension
somewhat diferently, e.g. referring to the “hierarchy” of instantiation, but it is important to
recognize that hierarchies and clines have very diferent properties.

1.3.1 The organization of the systemic functional metalanguage

The resource for construing theories of language in context and for enacting applications is the
systemic functional metalanguage (shown as the foundation in Figure 1.2). This resource is, like
language, a semiotic system, and it has the general properties of higher-order semiotic systems.
Importantly in the present context, it is stratified into an ordered series of strata. These may
vary somewhat depending on the domain in which one operates with the systemic functional
metalanguage, but let us present the organization that was developed in the first articulation of
the stratification of the systemic functional metalanguage in computational linguistic research.
In the context of computational linguistic research, it turned out to be helpful to model
the systemic functional metalanguage as stratified into four strata extended from the stratum of
theory to that of implementation:

• the highest stratum of the metalanguage – the stratum of systemic functional theory.
Given the nature of SFL, this stratum is holistic in orientation, covering as many aspects as
currently possible of language in context as semiotic systems in an ordered typology of sys-
tems (semiotic > social > biological > physical) and manifested in comprehensive descrip-
tions of particular languages in their particular contexts. Thanks to the holistic nature of
the theory and the comprehensive nature of descriptions, the metalanguage is appliable
(e.g. Halliday, 2015: 97); and the separation of the metalanguage into a series of strata also
increases its appliability.
• the theoretical stratum is realized by the stratum below – the stratum of theoretical
representation, which of course centrally involves system networks with realization state-
ments. System networks are informed by the theory of paradigmatic organization and its
relation (through realization) to syntagmatic organization. They stand in a “natural” rela-
tionship to that theory (just as lexicogrammar stands in a natural relationship to seman-
tics), but they do not cover the full extent of the theory of paradigmatic organization.
For example, paradigmatic continua cannot be represented by systems in system networks
since their terms are discrete (which is why researchers have explored the use of topolo-
gies, e.g. Lemke, 1987; Martin and Matthiessen, 1991, and of fuzzy set theory, e.g. Mat-
thiessen, 1995c); and they may fail to capture patterns of paradigmatic agnation that, as it
were, bridge less delicate distinction in a system network. At the same time, the stratum
of theoretical representation is not explicit enough for a variety of applications, including
computational ones, so it needs to be re-represented or realized by a more explicit type of
representation.
• the stratum of theoretical representation is realized by the stratum of computational
representation, i.e. by a form of representation that is explicit enough that it can sup-
port computational implementation. Explicitness can be interpreted in an “engineering”
sense: can the representation support implementation, which in practice means “has it
been implemented”? In computational linguistics, researchers have developed a number
of computational representations that can be suitable as realizations of system networks
with realization statements, and a number of them have been tried over the years, includ-
ing production rules (e.g. Patten, 1988) and typed feature structures (e.g. Bateman and
Momma, 1991). Such computational representations have typically been used for a range

10
 Introduction to the notion of a guide to SFL
11

Figure 1.2 The architecture of language in context according to SFL, starting with metalanguage for modelling semiotic phenomena
 Introduction to the notion of a guide to SFL

of tasks; they have not been designed only to represent system networks. They are more
explicit than system networks in the sense that they can be, or have already been, imple-
mented in some programming language; but there is a trade-off here: just like the theoreti-
cal representation of system networks does not cover all of the systemic functional theory
of paradigmatic organization, there are aspects of system networks that have not yet been
realized in the form of computational representations, notably recursive logical systems and
systemic marking conventions (cf. Henrici, 1965; Matthiessen, 1988b; Kasper, 1988; Bate-
man, 1989; Teich, 1995, 1999a).
• the stratum of computational representation is realized, or implemented, one level
down by means of some programming language, such as LISP, Prolog, C++, Java or
Python. Programming languages are used for many tasks, although they may have been
designed for a certain range of tasks, as when LISP was developed to serve tasks in AI (as
opposed to say numerical tasks of other programming languages at the time); so they are
even further removed from the theoretical stratum than is the stratum of computational
representation.

In terms of disciplinary expertise, the four metalinguistic strata implicate a team of researchers
with complementary areas of expertise: linguists are responsible for theory and theoretical repre-
sentation, computational linguists are responsible for computational representation, and software
engineers are responsible for implementation. In such a team, computational linguists can serve
as mediators, dialoguing with both linguists and software engineers (cf. Patrick, 2008).
As we go down the hierarchy of the stratification of metalanguage, new theoretical challenges
may present themselves. Prominently, computational linguistic projects will put pressure on the
stratum of systemic functional theory to develop theoretical accounts of processes, crucially pro-
cesses of instantiation (cf. Matthiessen and Bateman, 1991)6. While there has been a good deal
of exploration of various processes, perhaps in particular logogenesis, much needs to be done
in terms of theorizing linguistic (and more generally, semiotic) processes. This is hardly surpris-
ing; like linguists in other frameworks and traditions, systemic functional linguists have tended
to focus on the description of systems (potential) and the analysis of texts (instance) rather than
on the processes along the cline of instantiation that relate the two. In contrast, computational
linguists have worked on linguistic processes since the 1960s.
The stratification of the systemic functional metalanguage sketched earlier has been “tuned”
for use in computational contexts of application, but there are other kinds of application, like
the implementation of accounts in educational practice. This highlights the need to theorize the
meta-context in which metalanguage operates in addition to the metalanguage itself. In a way,
this is a motif that goes back to Halliday’s (1964a) paper called “Syntax and the consumer”. It
turns out that the contextual parameters already established for accounts of the context of lan-
guage are also relevant, i.e. field, tenor and mode.
Thus the meta-context in which we operate the metalanguage can be characterized func-
tionally as follows (see also Section 1.6):

Field is “what’s going on” in the context of doing linguistics, or more generally of doing
semiotics: (i) the different fields of activity involved in analysing, describing, comparing
and contrasting and theorizing semiotic phenomena and in applying the results to address
a wide range of problems that arise in the institutions that make up a community; and also
(ii) the different fields of experience brought into purvey, and even into existence, through
these activities – different domains of “knowledge”, e.g. expert knowledge concerning
pedagogy in education, concerning human anatomy and physiology in healthcare.

12
 Introduction to the notion of a guide to SFL

Tenor is concerned with who are taking part in various fields of activity, the tenor of
relationship among these participants as interactants engaging with one another in dif-
ferent role relationships. These role relationships centrally include institutional roles, e.g.
academic colleague to academic colleague, teacher to student, student to student, aca-
demic researcher to clients outside the academic community (professionals, members of
the general public and so on). Here we can also capture and represent citation networks
based on who refers to whom. But alongside institutional role relationships, there are
other important ones based on status (power), familiarity (contact) and, of course, socio-
metric roles and also value systems. In terms of value systems, the emphasis in SFL on
appliability in general and social accountability in particular plays a significant role.
Mode is concerned with the part the metalanguage plays in the metalinguistic context.
Significantly, it includes the rhetorical model, covering all the rhetorical uses we put our
metalanguage to, e.g. observational, experimental, analytical, theory expanding, descrip-
tive, informative, argumentative, didactic, promotional. In addition, it includes the channel –
the different academic specialized elaborations of oral and graphic, increasingly enhanced
by new technologies and involving mixtures and blends of in person and virtual “modes”;
and also the division of labour among the different metalinguistic strands that we have at
our disposal, including systems of theoretical and computational representation.

While these contextual parameters are, in principle, independently variable, there are naturally
certain combinations that are more common, having involved more academics and practitioners,
or more highly valued. For example, in SFL, descriptions with comprehensive coverage pre-
sented to inform users are highly valued, but in Chomskyan generative linguistics, argumenta-
tive texts dealing with alternatives within certain areas of theory have been more highly valued.

1.3.2 The ordered typology of systems operating in different

phenomenal realms
As Figure 1.2 shows, the theory of systemic functional linguistics construes a holistic model of systems
of all kinds, arranged in an ordered typology of increasing complexity, ranging from physical systems
to semiotic systems (e.g. Halliday, 1996a; Halliday and Matthiessen, 1999/2006; Matthiessen, 2007b,
forthcoming a). This ordering is analytical, but it can also be hypothesized that it corresponds to the
sequence of emergence in cosmogenesis,7 a term used and elucidated by Layzer (1990):

1st-order systems: physical systems – they were the first to emerge with the Big Bang
on the order of 13.5 billion years ago. These systems involve energy and matter extended
in space-time and are subject to a range of forces, the laws of nature. They range in scale
from the microscopic realm of the quantum world to the whole cosmos. As cosmos
has expanded and cooled, going through different epochs, it has increased in complex-
ity over time, e.g. through the gradual differentiation of different forces from an initial
unitary force, gravity being the first to split from other forces. However, while physical
systems change through cosmogenesis, they are not subject to evolution in a technical
sense; evolution is a mode of cosmogenesis that emerged with 2nd-order systems.
2nd-order systems: biological systems = physical systems + life – they emerged under
very specific physical conditions on our planet around 3.5 billion years ago. “Life” is
shorthand for a new order of organization out of complex chemicals involving individu-
ation, self-replication, memory in the form of genetic code and evolution as a new mode
of cosmogenesis, with natural selection.

13
 Introduction to the notion of a guide to SFL

3rd-order systems: social systems = biological systems + value (or social order) –
they must have emerged on many occasions in different biological populations. “Value”
is shorthand for individuals taking on different roles in role networks within different
groups engaging in different kinds of behaviour – i.e. undertaking different kinds of
labour in their roles, with the potential for division of labour and also of social hierarchy.
An individual is now a “person”, an assemblage of social roles or personae, and biological
acts are now also social behaviour. Role networks and forms of behaviour come together
in the organizations of institutions such as the family, friendship, education, healthcare
and science; and aggregates of such institutions form societies.
4th-order systems: semiotic systems = social systems + meaning – like social systems
they must have emerged on many occasions in different social groups. They have the
properties of social systems – centrally, individuals as persons interacting in different roles
in different social roles (personae), but they have the added property that distinguishes
them as semiotic rather than purely social – the property of meaning. Consequently,
semiotically, persons are also meaners, and interaction involves the exchange of mean-
ings with other meaners in meaning groups. Meaning groups form speech fellowships,
to use Firth’s term, or speech communities. To be able to carry or even create meaning,
semiotic systems must be stratified into two planes, content and expression.8 This
minimal stratification is characteristic of simple semiotic systems – primary semiotic
systems; but more powerful semiotic systems, higher-order semiotic systems, are further
stratified within both planes – in the case of language, the content plane is stratified into
semantics and lexicogrammar and the expression plane into phonology and phonetics
(or graphology and graphetics, or their analogues in sign languages). Both primary and
higher-order semiotic systems operate within context, which is a higher-order content
plane: see Figure 1.3. Primary semiotic systems are micro-functional: different micro-
functional meaning potentials correspond one-to-one to different contexts. In contrast,
higher-order semiotic systems are meta-functional: their meaning potentials embody
simultaneous metafunctional modes of meaning and are freed from particular types of
context. This contrast is illuminated in systemic functional studies of ontogenesis (cf.
Sections 3.2.4 and 4.5.1).

The four orders of system can be grouped into two macro-orders, viz. material systems, com-
prising physical and biological systems, and immaterial systems, comprising social and semiotic
systems. This diferentiation is explored by Halliday (2005b), who provides a subtle account of
the ordered typology in terms of the relative mixture of meaning, now in the generalized sense
of “order”, and matter.
Coming from different disciplines, a number of scholars have also proposed ordered typolo-
gies. We have already referred to Layzer (1990), whose account is grounded in astrophysics.
These accounts tend to lump social and semiotic systems together under a heading such as
“culture”. This brings out the fact that they are “immaterial”; they are concerned with social
and semiotic order. However, such accounts fail to reveal the fundamental distinction between
“value” and “meaning” that is needed to explore and theorize semiotic systems. Thus while
social systems are networked, in the sense of role networks, they are not stratified into content
and expression – an shown schematically in Figure 1.3.
In the move up this ordered typology, each higher-order system emerges out of the system
next below (not only analytically but also in the course of cosmogenesis); it is manifested in the
organization of that order of system next below and also inherits its lower-order properties of
organization. Thus semiotic systems are also social, so also biological and also physical: meaning

14
 Introduction to the notion of a guide to SFL
15

Figure 1.3 The distinction between social systems and semiotic ones: semiotic systems are social systems + meaning, so they are stratified into content and expression
within context
 Introduction to the notion of a guide to SFL

is enacted by members of social groups, it is embodied in biological organisms and manifested

physically, e.g. as sound waves. In addition, once a new order of system has emerged, it will
co-evolve with systems of the lower orders (except for physical systems, which do not evolve).
For example, Deacon (1992, 1997) has hypothesized that language and the human brain have
evolved together; this co-evolution would serve to explain the extraordinary complexity of the
human brain. Similarly, Dunbar (e.g. 1996) has suggested that there is a correlation between the
size of social groups and the emergence of language.
As a 4th-order system, language is thus also social, biological and physical; and this is very
evident in branches of linguistics – or linguistics and phonetics, if the latter is treated as a separate
discipline, as it sometimes is. So while language is a semiotic system, it is also manifested within
systems of a lower order. Consequently it can be studied physically, biologically and socially as
well as semiotically within a range of different sub-disciplines of linguistics and other disciplines,
as illustrated in Figure 1.4.
These different disciplines and sub-disciplines cover different approaches to conception and mode
development of theory, of nature of data and methodology (cf. Chapter 4) and they operate with
distinct dialogic interfaces to other disciplines. Not surprisingly, the lower the order of manifesta-
tion of language, and of other semiotic systems, that is in focus, the closer the study is likely to be to
paradigms developed within “hard” sciences – which essentially means sciences devoted to the study
of material systems. In this respect, phonetics can be regarded as the leading sub-discipline within
linguistics (and may be treated institutionally with a given university as a separate department).
This simply reflects the general principle that the breakthroughs to modern science have
followed the degree of complexity inherent in the ordered typology (e.g. Halliday, 1996a; Hal-
liday and Matthiessen, 1999/2006): first physical sciences with Kepler, Galileo and Newton,
then biological sciences with Darwin (and Wallace), later supplemented by Mendel (although
it took quite a while for his contribution to be recognized), then social sciences with Marx,
Weber and Durkheim, and then semiotic sciences. . . . But here we have to pause: while Sau-
ssure made a valiant effort to lay the foundation of a science of semiotic systems, he failed for

Figure 1.4 Orders of systems operating in different phenomenal realms and studies of these realms in
different disciplines and sub-disciplines

16
 Introduction to the notion of a guide to SFL

various interesting reasons, a key one being the creation of the dichotomy between langue and
parole, which created a distance between the domain of “data” and the domain of “theory” (cf.
Halliday, 1991d).
While Figure 1.4 represents each type of study as focussed on one systemic order, this is just
the main focus; a number of disciplines and sub-disciplines are concerned with the relationship
between at least two orders of manifestation of language (and other semiotic systems). For exam-
ple, sociolinguistics is concerned with the relationship between language and society, although
the theory of language often remains fairly unelaborated – which is why Hasan (e.g. 2009b) has
proposed a reconceptualization of sociolinguistics informed by systemic functional theory. Other
significant contributions include Steiner’s (1991) theoretical integration of the Leont’ev activity
theory into an SFL-based framework for analysis of text in context and van Leeuwen’s (e.g. 1996)
account of social actors. These contributions make it possible to attend to and analyse social behav-
iour in contexts of situation where systemic systems are minimally active, as in the traditional cate-
gory of “language in action”, or even absent, while still bringing out the way in which context as a
higher-order semiotic system serves to coordinate not only semiotic processes but also social ones.
Such relationships between different systemic orders are also sites where new areas of study
tend to emerge (see already Thibault, 2004) – one recent example being neurosemiotics (García
and Ibáñez, forthcoming) and another new development being studies of translation investigat-
ing correlations between semiotic discourse patterns and biological acts observed through eye-
tracking and key-stroke logging, as in Alves et al. (2010). Other systemic-functionally informed
examples of the value of taking the biological embodiment of semiotic systems into account are
Asp’s (2013) and Gil’s (2013) accounts of systemic choice in the light of neuroscience.

1.3.3 The global semiotic dimensions organizing language in context

The four orders of system sketched in Figure 1.4 share certain organizational properties – prop-
erties that are candidates for inclusion in a general systems theory. For example, systems of all
orders are extended along a cline of instantiation from potential to instance – hence Halliday’s
(e.g. 1991d) well-known analogy between meteorological systems and language (which can be
generalized to other semiotic systems as well): see Table 1.3. This says that the linguistic system
is to text as the climate is to weather. We observe and analyse phenomena at the instance pole of
the cline of instantiation, compiling representative samples of instances; and we theorize them
at the potential pole.
There are other properties that characterize the organization within systems of all kinds – like
the compositional hierarchies that emerge when systems begin to increase in complexity, as in the
move from protolanguage to language during ontogenesis. Such properties have been investigated
in various manifestations of “system science”, including General System Theory, articulated and
envisaged by Bertalanffy (1968) and Boulding (1956), and developments related to that tradition
(e.g. Skyttner, 1996, 2001), the study of complexity – of complex adaptive systems (e.g. Gell-
Mann, 1994), and also the emergent discipline of network science (Barabási, 2016). There are

Table 1.3 The cline of instantiation as a property of systems in general – as illustrated by the analogy
between the linguistic system and the climate

Potential Subpotential/instance type Instance

meteorological phenomena climate sub-climate/weather pattern weather

linguistic phenomena linguistic system register/text type text

17
 Introduction to the notion of a guide to SFL

general properties in addition to composition that have already been shown to be relevant to the
workings of language such as Steels’ (1998) notion of level formation.
Language is extended along the cline of instantiation as one phenomenon; the linguistic sys-
tem, texts and intermediate patterns are not different phenomena but simply reflect the different
phases of instantiation that we can observe, even if only indirectly, analyse, describe and theorize.
There are patterns that are remembered and stored as “templates” – as partially or completely
instantiated systemic patterns (noted in various linguistic traditions under a number of headings,
including “phraseology”, “idiomatic expressions”, “formulaic language” and “constructions”;
Halliday (1992b/2003b: 367) comments on such patterns:

we should be reminded that the part played by the system in the history of any par-
ticular act of meaning is variable. Many sentences are stored readymade; they are more
or less formulaic for the speaker and for the listener, and any given instance of their
occurrence harks back to previous instances of the same wording rather than being
engendered afresh by choosing within the system. Examples from English are:
Take it or leave it!
I couldn’t believe my ears.
It’s a contradiction in terms.
He doesn’t know what he’s talking about. Not by any stretch of the imagination.
You can’t change human nature.
You must keep your eye on the ball.
etc.
Of course these items are still located within the system and derive their meaning
ultimately by semantic proportionality in the usual way; but their immediate source is
the trace of earlier occurrences.

The cline of instantiation is one of the global dimensions organizing language in context; lan-
guage and context are both extended from the potential pole of the cline to the instance pole,
as shown in Figure 1.5.
Context is extended from the context of culture at the potential pole to contexts of situa-
tion at the instance pole, with the intermediate region of cultural domains (or institutions) and
situation types. Approached from the context of culture, this region can be characterized as
subsystems of cultural potential – cultural domains, or institutions; and approached from con-
texts of situation, it can be characterized as recurrent situations that are similar enough to form
situation types.
Similarly, language is extended from the linguistic system at the potential pole, which func-
tions in the context of culture, to texts at the instance pole; and here there is also an extended
intermediate region. Approached from the linguistic system, this region can be characterized as
subsystems of language – registers, or functional varieties of language functioning in different
cultural domains; and approached from texts, it can be characterized as recurrent texts that are
similar enough to form text types.
The vertical axis in Figure 1.5 represents another global dimension of organization, viz. the
hierarchy of stratification. This dimension appeared with the emergence of semiotic systems in
the ordered typology of system; it is an organizational property that distinguishes them from
social systems and enables semiotic systems to carry or even create meaning. The relationship
between language and culture is thus interpreted stratally as one of realization – an insight that
was made explicit by Halliday (1978): he interpreted context as a semiotic system, i.e. as a system

18
 Introduction to the notion of a guide to SFL

Figure 1.5 The extension of context and language along the cline of instantiation, taken from Halliday
(1991d)

meaning.9 Drawing on Hjelmslev (1943), Martin (1992a) theorizes context as a connotative

semiotic system, one that has another or other semiotic systems as its expression plane, in con-
trast with denotative semiotic systems such as language, which have their own expression planes.
Language is further stratified internally into the two kinds of plane just mentioned, the con-
tent plane and the expression plane; and each of these is further stratified: the content plane is
stratified into semantics and lexicogrammar, and the expression plane is stratified into phonol-
ogy and phonetics (or graphology and graphetics in spoken language, or their analogues in sign
languages of deaf communities).
There is one more global dimension, i.e. one more dimension that organizes all of language
in context. This is the dimension of functional diversification. This is a principle of organiza-
tion that permeates both context and language, more specifically the content plane of language
(and to some extent also the prosodic zone of phonology, or arguably graphology). This func-
tional organization of language is intrinsic to it, as explained by Martin (1991); in other words,
its distinct from, though ultimately related to, the extrinsic uses of language, as shown by Hal-
liday and Hasan (1985). The functional diversification of both context and language “guides”
the realizational relationship between them. Halliday (1994a) characterizes the relationship suc-
cinctly as follows:

systemic theory gives prominence to discourse, or ‘text’; not – or not only – as evi-
dence for the system, but valued, rather, as constitutive of the culture.10 The mecha-
nism proposed for this constitutive power of discourse has been referred to as the
‘metafunctional hookup’: the hypothesis that (a) social contexts are organic – dynamic
configurations of three components, called ‘field’, ‘tenor’, and ‘mode’: respectively,
the nature of the social activity, the relations among the interactants, and the status
accorded to the language (what is going on, who are taking part, and what they are
doing with their discourse); and (b) there is a relationship between these and the meta-
functions such that these components are construed, respectively, as experiential, as

19
 Introduction to the notion of a guide to SFL

Figure 1.6 The resonance between context and language based on the functional diversification of both –
specified at the instance pole of the cline of instantiation

interpersonal, and as textual meanings. Register, or functional variation in language,

is then interpreted as systemic variation in the relative prominence (the probability of
being taken up) of different options within these semantic components.

The relationship between the “organic components” of context and the metafunctions extends
along the cline of instantiation from the potential pole to the instance pole, and, as mentioned,
it extends stratally from contents down through the content plane of language, even involving
prosodic phonology. The natural stratal relationship between context and language – semantics
in the first instance, and between semantics and lexicogrammar within the content plane is thus
based on the resonance across the modes of meaning that are functionally diferentiated. The
relationship is summarized diagrammatically by Halliday and Hasan (1985) with reference to the
instance pole of the cline of instantiation: see Figure 1.6.
Figure 1.6 specifies the relation between context and language at the instance pole of the
cline of instantiation – i.e. between context of situation and text; but both field, tenor and mode
and the metafunctions are extended throughout the cline of instantiation, from the potential
pole to the instance pole,11 as illustrated in Figure 3.4.
The three dimensions of the hierarchy of stratification, the cline of instantiation and the
spectrum of functional diversification together determine the global organization of language in
context. This “space” can be represented diagrammatically in different ways; one way is to map
it out by means of a three-dimensional matrix, as in Figure 1.7.
Using this matrix, we can also locate the domains of the local dimensions organizing each
stratum, as illustrated for lexicogrammar in Figure 1.7. This local organization is represented
by another kind of matrix, a function-rank matrix, introduced by Halliday as he had added the
theory of metafunction to SFL (e.g. Halliday, 1970a). Let us now discuss the local dimensions
of organization.

1.3.4 The local semiotic dimensions organizing stratal subsystems

The local dimensions of organization are the hierarchies of axiality and of rank (the rank scale).
They organize all stratal systems, and in this respect they are fractal, i.e. principles of organiza-
tion that are manifested in the different stratal environments defined by the global hierarchy of
stratification.

20
 Introduction to the notion of a guide to SFL

Figure 1.7 The intersection of the hierarchy of stratification, the cline of instantiation and the spectrum
of functional diversification represented as a three-dimensional matrix

The hierarchy of axis is the same wherever it is manifested: the paradigmatic axis is primary,
and patterns along the syntagmatic axis are “derived” from it by means of realization statements.
The primacy of the paradigmatic axis is unique among linguistic theories, and was introduced
by Halliday in the 1960s (argued for by Halliday, 1966a, and illustrated in descriptions in e.g.
Halliday, 1967, 1967/8, 1969). Paradigmatic organization is represented by means of system
networks located at the second stratum of the stratified metalanguage (see Section 1.3.1), the
stratum of the stratum of theoretical representation (for algebraic and graphic conventions, see
Section 9.2), and within a given stratum, these system networks are distributed according to
the hierarchy of rank of the stratum, and primary classes within each rank (e.g. at group rank:
nominal group, verbal group, adverbial group).
The hierarchy of rank, the compositional hierarchy known as the rank scale, varies from one
stratum to another within a given language, and it is also subject to variation across languages.
For example, the phonological rank scale of English is tone group > foot > syllable > phoneme,
but that of Mandarin is tone group > foot > syllable, as described by Halliday (1992d). By now,
there are quite a range of descriptions of the rank scales of different languages at the strata of
lexicogrammar and phonology, but a great deal of work remains to be done on the description
of composition within the semantic strata in terms of rank or other forms of composition (for
issues relating to rank-based descriptions, see Section 5.2.5).
Just as we can intersect global dimensions to create maps of the overall resources of lan-
guage in context in the form of matrices, as indicated in Figure 1.7, we can intersect the global

21
 Introduction to the notion of a guide to SFL

Table 1.4 The division of phonological labour in terms of the phonological rank scale in English (from
Matthiessen, 2021a)

Rank Prosodic ~ Articulatory

Pitch Rhythm Salience Articulation

tone group TONE, TONICITY,

TONALITY
foot FOOT COMPOSITION,
ICTUS STATE
syllable SALIENCE articulatory postures
(“phonotactics”)
phoneme PLACE, MANNER &c

dimension of metafunction with the local dimension of rank to form function-rank matrices for
the two strata of the content plane, semantics and lexicogrammar. Such matrices are focussed
on the paradigmatic axis, and serve as overviews of the systems of a given stratum distributed
according to metafunction and rank, as illustrated for the lexicogrammar of English in Table 3.9
and Table 3.10.
Such matrices for phonological systems can be formed by intersecting rank scale with the dis-
tinction between prosodic patterns (pitch and rhythm) and articulatory patterns (manner of articu-
lation) so as to provide an overview of the distribution of phonological systems of a language. The
phonological matrix for English is outlined in Table 1.4. Here the system of salience is interpreted
as involving both prosodic and articulatory features, since it relates to the function of a syllable in
a foot (as Ictus or Remiss) and to the articulatory quality of the syllable, notably the vowel quality
(which is characteristic of English as a foot-timed language). In general, in many languages, the
syllable serves as the phonological gateway between prosodic and articulatory phonology.
Semantic, lexicogrammatical and phonological matrices constitute “maps” of the system of
each stratum, and enable us to see the linguistic division of labour across ranks, which is valuable
in the description of a particular language and indispensable in the comparison of two or more
languages.

1.3.5 Theory and description and analysis

In engaging with linguistic and other semiotic phenomena construed by the systemic functional
metalanguage (cf. Figure 1.2), systemic functional linguists have always been careful to draw a
clear distinction between the general theory of language and descriptions of particu-
lar languages – continuing the Firthian tradition (e.g. Firth, 1957c, 1968), and drawing on
Hjelmslev’s insights in Glossematics. In his “personal perspective” written for the first volume of
his collected works, On Grammar, Halliday (2002d: 12–13) writes:

I am often asked about my views on “linguistic universals”. The answer is that I fol-
low Hjelmslev and Firth in distinguishing theoretical from descriptive categories. The
theoretical categories, and their interrelations, construe an abstract model of language
(and other semiotic systems); they are interlocking and mutually defining. The theory
that is constituted in this way is continually evolving as it is brought to bear on solv-
ing problems of a research or practical nature. (No very clear line is drawn between

22
 Introduction to the notion of a guide to SFL

“(theoretical) linguistics” and “applied linguistics” – except institutionally where, for

example, an education authority will give teachers release time and professional credit
for a degree course called “applied linguistics” but not for one called “linguistics”.)
Descriptive categories are categories set up in the description of particular lan-
guages. When people ask about “universals”, they usually mean descriptive categories
that are assumed to be found in all languages. The problem is that there is no mecha-
nism for deciding how much alike descriptive categories from different languages have
to be before they are said to be “the same thing”. There is a method, based on the
(theoretical) category of system, for matching up descriptive categories across lan-
guages but they are not claimed to be universal, and no grand hypothesis stands or falls
by their “universality”. The unity of human language, and its relation to the human
brain, is proclaimed by the multifaceted architecture of the theory.

The distinction between theory and description has been clear throughout the history of SFL,
even during its pre-systemic-functional scale-&-category phase (e.g. Halliday, 1961: Section 1.11).
Thus Bamgboṣe (1966) deals with theory in Part 2 of his book (grammar, rank scale, structure,
delicacy, system, exponence and so on) before he sets out his description of Yoruba, which is
essentially a descriptive move down the rank scale; and in the final section he provides “analysis
of samples from the text”. Theoretical and descriptive terms are distinguished clearly in Mat-
thiessen, Teruya and Lam (2010). While the universalist stance of Chomskyan linguistics domi-
nated for a few decades in the second half of the 20th century, based on actual descriptions of a
growing number of languages, there is now a growing recognition of the “myth of universals”,
cogently argued by Evans and Levinson (2009), which can be related to what Firth (1957c)
called the “universalist fallacy”.

1.4 Heuristics, matrices, delicacy (of focus), complementarities, helixes

SFL is, as we have emphasized, a holistic approach to language. This holistic approach taken – and
developed – in SFL is reflected in many ways, ways that we will have reason to discuss through-
out this guide; they include:

• the development of a theory of language in context modelled as a semiotic space of

multiple intersecting semiotic dimensions, each of which is the domain of operation of
a particular kind of relation (e.g. dimension: stratification; relation: realization); two (or
more) semiotic dimensions may be intersected to define matrices providing overviews of
the resources filling the semiotic space, including:
• the stratification-instantiation matrix: the intersection of the hierarchy of stratifi-
cation and the cline of instantiation to show the stratified resources of language in con-
text extended along the cline of instantiation from the potential pole to the instance
pole, as shown in Figure 3.5;
• the function-stratification matrix: the intersection of the hierarchy of stratifica-
tion and the spectrum of metafunctions to show the stratified resources of language
in contexts dispersed across the spectrum of modes of meaning, as shown in Table
3.16;
• the function-rank matrix: the intersection of the hierarchy of rank (rank scale) to
show the ranked resources of semantics or lexicogrammar (or context) dispersed across
the spectrum of modes of meaning, as shown below in Table 3.9 and Table 3.10;

23
 Introduction to the notion of a guide to SFL

• the articulation of two related methodological strategies based on the intersecting

dimensions defining the semiotic space:
• shunting: moving along any of the dimensions to ensure that the phenomena under
investigation are observed, analysed or described not only from one position but from
different positions (introduced in Halliday, 1961), e.g. shunting along the cline of
instantiation to shift from instance observation to system modelling, shunting along the
cline of delicacy to shift from a lexical approach to a grammatical one;
• trinocular perspectivizing: like shunting, moving along a dimension, but specifi-
cally one involving an ordering of viewpoints so that the phenomenon being investi-
gated can be observed, analysed or described “from above”, “from roundabout” and
“from below” (as shown in Halliday, 1978; e.g. viewing the grammatical system of
mood along the hierarchy of stratification “from above” in terms of speech function
within semantics (and by another step up, in terms of tenor within context), “from
below” in terms of tone within phonology, and “from roundabout” in terms of its own
organization as a grammatical system of the clause together with other clausal systems
such as polarity and deicticity);
• proportionalities: proportionalities are inherently systemic but can be stated in terms
of syntagmatic patterns; the basis is simply saying that a is related to b in the same way
as p is related to q, or a : b :: p : q for short. They are thus used to identify systemic
contrasts or relationships of agnation. See further later on proportionalities.
• reactances: this term originated with B.L. Whorf, reflecting his background in
chemical engineering. In addition to being detectable by explicit markings, gram-
matical categories may be characterized by the way in which the grammar “reacts”
to them. For example, in English, perceptive mental clauses are characterized among
other things by the fact that the simple present and the potentiality type of modulation
(realized by can) are very closely agnate – more so than in the general case; for example,
in the perceptive mental clause And now, from the window (which is not my typical window)
I can see a kid trying to hit a baseball in the park, the verbal group can see serving as Process
is close to see – not synonymous, of course, being a close agnate, which distinguishes
‘perceptive’ clauses from ‘mental’ clauses of other types of sensing (contrast I want the
new edition and I can want the new edition, which are quite distinct) and also from ‘behav-
ioural’ clauses (contrast I am watching a kid trying to hit . . . and I can watch a kid trying to
hit . . ., which are also quite distinct).
• the search for complementarities in theorizing and description (e.g. Halliday, 1998b,
2008b): viewing a phenomenon or the account of the phenomenon from different vantage
points that appear to engender conflicting views but transforming these conflicting views
into complementary perspectives (e.g. the grammatical and lexical views of patterns of
wording, the transitive and ergative views of experiential configurations);
• the progression in conceptualization, presentation, curriculum design in terms of helical
movement (rather than linear or cyclical movement): moving upwards towards increasing
information and insight, covering the same territory by revisiting particular areas at a higher
order of information and insight.

The examples we have listed here are not intended to be self-explanatory at this point, but we
have included them to indicate the general orientation of SFL in its approach to, and engage-
ment with, language and other semiotic systems. They are important in particular because sys-
temic functional linguists have always striven towards holistic – and unified – theory and towards

24
 Introduction to the notion of a guide to SFL

comprehensive descriptions, so they need strategies to cope with the daunting complexity of
language in context. We will come back to the items on the preceding list, but let us provide a
couple of examples of the use of proportionalities, of complementarities and of helical moves.
(1) Proportionalities are inherent in the system of language, being a property of its paradig-
matic or systemic organization, so the use of proportionalities is an important methodological
strategy inherent in the notion of system and the significance given to paradigmatic organiza-
tion. It reflects the conception of language as a vast network of relationships rather than as an
inventory of “things”, as shown by this early statement by Halliday (1964a/2003b: 39):

transformational grammars of English recognize a passive transformation relating such

pairs of sentences as the man eats cake and cake is eaten by the man. The analogue in
a “scale-and-category” grammar (to use a name by which the version of a system-
structure grammar that my colleagues and I have been working with has come to be
known) would be a system at clause rank whose terms are active and passive: as for
example the question transformation is paralleled by a clause system whose terms are
affirmative (transformational grammar’s “declarative”) and interrogative. In fact no sys-
tem of voice at clause rank is introduced into our present description of English. We
could say that this is because it does not represent our interpretation of the facts. But
the question is: what “facts” are being interpreted? The system implies proportionality:
given a system whose terms are a, b, c, then the set of their exponents12 a1 a2 a3 . . ., b1
b2 b3 . . ., c1 c2 c3 . . . are proportionally related: a1 : b1 : c1 :: a2 : b2 : c2 and so on. This
holds good, it seems to us, to a reasonable extent (such that the simplicity of the general
statement is not outweighed by the complexity of further statements that are required to
qualify it) of affirmative and interrogative in the clause, and of active and passive in the
verbal group where the description does recognize a system of voice; but not of active
and passive in the clause. In other words, John was invited by Mary, this house was built by
my grandfather, the driver was injured by flying glass, John’s been dismissed from his job and it
was announced that the committee had resigned are not explained as all standing in the same
relation to a set of their active counterparts. Such a relationship is shown, but indirectly
(as the product of a number of systemic relations) and not always by the same route.

That is, the use of proportionalities makes it possible to reason about paradigmatic relationships
without reifying the elements being related; thus proportionalities depend on, and reveal, para-
digmatic values (valeurs). There are many examples in the SFL literature of the use of proportion-
alities in the development of descriptions; one descriptive example from Halliday (1997/2003b:
249) will sufce here:

how do we establish systematic patterns of predictable meaning relationship? Structural

proportionalities provide evidence, but the underlying proportions are systemic, and
located within metafunctionally defined regions of the grammar – which enable us to
set up proportionalities such as these, from the region of modality in English:
I think [they’re away] : I don’t think [they’re here] ::
it’s possible [they’re away] : it’s not certain [they’re here] ::
they must [be away] : they can’t [be here]

Since the use of proportionalities is not all that common (nowadays) outside SFL, it may be
helpful to cite a number of examples: see Table 1.5. There are additional examples in the

25
 Introduction to the notion of a guide to SFL

Table 1.5 Examples of proportionalities in the systemic functional literature

Area Examples Reference

transitivity: we’re selling cosmetics : cosmetics are selling :: Halliday

transitive/ergative we’re running Jones for chairman : Jones is running for (2003b: 60–61)
interpretations chairman
the clothes were washed : the clothes washed ::
the door was opened : the door opened ::
her hair’s being grown long : her hair’s growing long ::
the horse shouldn’t be jumped: the horse shouldn’t jump
modality: explicit/ I think [they’re away] : I don’t think [they’re here] :: Halliday
implicit orientation it’s possible [they’re away] : it’s not certain [they’re here] :: (2003b: 249)
they must [be away] : they can’t [be here]
phonology (1) within the half-close vowel: Halliday
ui : iu :: un : ing ::: [both w-y : y-w] (2006: 308)
ei : ou :: en : eng [both a-y : a-w]
(2) and between all of these and those with open vowel:
uai : iao :: uan : uang ::: [both w-y : y-w]
ai : ao :: an : ang [both a-y : a-w]
(3) oral only:
ie : e : uo :: [y-a : a-a : w-a]
ia : a : ua [y-a : a-a : w-a]
(4) nasal only:
in : ong :: [y-y : w-w]
ian : uang [y-y : w-w]

systemically oriented steps into systemic functional theory: Matthiessen and Halliday (2009),
Martin (2013a), and Martin et al. (in press).
(2) Complementarities are recognized in SFL both as inherent in language and therefore as
a strategy for theorizing language and describing particular languages. Halliday (2008b: Section
5.2) characterizes complementarity as follows:

Complementarity means having things both ways – that you eat your cake and have
it, as I put it earlier. At its simplest, this implies that you fit two (or more) matching
parts together to make a pattern, like the complementary colours that, when mixed
together, make white.
...
The three complementarities in language that I have been talking about are also of
three distinct kinds. As I described it, grammar and lexis form a complementarity of
focus; system and text form a complementarity of angle; speaking and writing form a
complementarity of state. We could gloss these further by identifying the dimension
along which they are counterposed.

(1) Grammar and lexis are opposed in delicacy: meaning as generalities or as particulars.
(2) System and text are opposed in instantiation: meaning as (collections of) instances or
as potential.
(3) Speaking and writing are opposed in (manner of) realization: meaning as happening
or as thing.

26
 Introduction to the notion of a guide to SFL

But all have a critical property in common: they are all strategies for extending the
semogenic power of language – its meaning potential as a semiotic system.

In addition to the three general complementarities that Halliday (2008b) deals with – comple-
mentarities that are as it were inherent in the architecture of language – we can also recognize
complementarities that we find in the description of particular languages or in the comparison
of such descriptions: see Table 1.6.
We will just give two examples. (i) The lexicogrammar of English has evolved two comple-
mentary models of processes of consciousness, either emanating from the senser or as impinging
on him or her. They are characterized succinctly by Halliday (1996a/2002d: 400):

Complementarities are found in those regions of (typically experiential) semantic space

where some domain of experience is construed in two mutually contradictory ways.
An obvious example in English is in the grammar of mental processes, where there
is a regular complementarity between the “like” type (I like it; cf. notice, enjoy, believe,
fear, admire, forget, resent . . .) and the “please” type (it pleases me; cf. strike, delight, con-
vince, frighten, impress, escape, annoy . . .). The feature of complementarities is that two
conflicting proportionalities are set up, the implication being that this is a complex
domain of experience which can be construed in different ways: here, in a process of
consciousness the conscious being is on the one hand ‘doing’, with some phenom-
enon defining the scope of the deed, and on the other hand ‘being done to’ with the
phenomenon functioning as the doer. All languages (presumably) embody comple-
mentarities; but not always in the same regions of semantic space (note for example

Table 1.6 Examples of lexicogrammatical systemic complementarities

Metafunction Systemic domain Systems References

ideational experiential/ construing “augmentation”: Matthiessen (2004a)

logical circumstantiation/logico-
semantic type
transitivity construing participation: transitive Halliday (1967/1968); Davidse
model/ergative model (1992a, 1996b); Halliday
and Matthiessen (2014:
Section 5.8)
transitivity construing sensing: mental clauses: Halliday (1996a/2002d: 400)
emanating/impinging

transitivity construing pain: as process, quality, Halliday (1998a)

thing
temporality construing process time: tense/ Matthiessen (2004a)
aspect
interpersonal assessing modality: modalization/ Matthiessen (2004a)
information evidentiality
textual discursive flow theme/information Halliday (1967/1968); cf. also
in relation to texts: Halliday
(1981); Martin (1993b);
Halliday and Greaves (2008);
Matthiessen (1992)

27
 Introduction to the notion of a guide to SFL

the striking complementarity of tense and aspect in Russian). One favourite domain is
causation and agency, often manifested in the complementarity of transitive and erga-
tive construals.

(ii) The lexicogrammars of languages around the world have evolved at least two complementary
models for construing processes unfolding in time – the systems of tense and of aspect. We can
see them as complementary when we compare languages using essentially one or the other, as
in the case of English and Chinese; but there are also languages that embody both, like Russian,
Hindi and (Modern Standard) Arabic. In tense systems, the unfolding of the process in time is
modelled in terms of its location relative to the ‘now’ of speaking, or some reference time inter-
mediate between the ‘now’ of speaking and the time of the process – commonly, past, present or
future relative to ‘now’ or another reference time. In aspect systems, the unfolding of the process
in time is modelled in terms of its boundedness in time – bounded, grammatically ‘perfective’,
or unbounded, grammatically ‘imperfective’. The two models are compared diagrammatically
in Figure 1.8.
(3) To illustrate the deployment of a helical move rather than a linear of cyclical one
in the interpretation of phenomena being explored, let us again draw on Halliday for an
illustration of the deployment, taking Halliday’s (1991a, 1996b) helical interpretation of

Figure 1.8 The complementarity of systems of tense and of aspects as models of process time

28
 Introduction to the notion of a guide to SFL

“literacy” as an example (for a recent well-rounded contribution to the engagement with

literacy informed by SFL, see de Silva Joyce and Feez, 2016; for literacy and pedagogy,
see e.g. Martin and Rose, 2005). After taking a number of helical steps in his interpreta-
tion of literacy, Halliday (1996b) provides a summary helical diagram, reproduced here as
Figure 1.9 (for “writing systems”, see Section 3.1.5.3), and describes the move as follows
(1996b/2007b: 123–124):

I have tried to trace a course through what Graff called the labyrinth of literacy, while
interpreting literacy in linguistic terms. The route has led through a number of stages,
which could be summarized as follows:

1 The written medium

engaging with the material environment to produce abstract symbolic objects called
“writing”.
2 Writing systems
mapping these symbols on to elements of language and constructing them into written
text.
3 Written language
construing meaning through lexicogrammar in written text: lexical density, nominal-
ization, grammatical metaphor.
4 The written world
construing experience through semantics in written language: the world objectified as
the basis of systematic knowledge.
5 The technology of literacy: (1) revisited
from books to computers: refining the medium, realigning writing and speech, tech-
nologizing discourse.
6 The frontiers of literacy: (2) revisited
from writing to other systems of visual semiotic: expanding the potential for meaning.
7 The contexts of literacy: (3) revisited
from text to context: locating written language in its socio-cultural environment.
8 The ideology of literacy: (4) revisited
from the construction of experience to participation in the social-semiotic process.

This suggests a kind of helical progression, as set out in Figure 6.1 [reproduced here as
Figure 1.9].

His (1991a) version is slightly diferent, and also worth quoting:

I have tried to trace a course through the linguistics of literacy. It has run something
like this:

1 Written medium: graphological level, material interface with the body, leading to:
2 Writing systems: relation to other levels, phonology and lexicogrammar, leading to:
3 Written systems: how meaning is construed in the lexicogrammar of writing, leading to:
4 The written world: the discourse-semantics of writing as construction of experience;
leading to:
5 The technology of literacy: reconstituting (1) as technological processes, leading to:
6 The frontiers of literacy: reconstituting (2) as visual semiotic systems, leading to:

29
 Introduction to the notion of a guide to SFL

7 The contexts of literacy: reconstituting (3) as the construction of meaning by text-in-

context, leading to:
8 The ideology of literacy: reconstituting (4) as the construction of society.

This suggests a kind of helical progression, as follows:

Figure 1.9 Halliday’s helical interpretation of “literacy”, ascending from the written medium to the
ideology of literacy

As noted, systemic functional linguists need to face the potentially overwhelming complex-
ity of language in context, as they develop both holistic theory of language in context and
comprehensive descriptions of particular languages in their contexts. While many linguists
have followed “mainstream” science in relying on Cartesian Analysis, drilling down to ever
smaller domains until they become manageable in terms of the theory at their disposal, this
runs counter to holistic thinking and systemic functional linguists have engaged in “systems
thinking”. Systems thinking includes the strategies listed earlier. These strategies help sys-
temic functional linguists pursue a holistic approach to language in context, instead of the
generally favoured approach associated with Cartesian Analysis (recommended in Descartes’s
Meditations).
Trying to explain the difference to Bill Mann and Christian Matthiessen over lunch in
Marina del Rey in 1980, Michael Halliday drew contrasting figures reproduced in Figure 1.10
on a paper placemat. On another occasion, he told Matthiessen that the best piece of advice he
had ever given a PhD student was to get a large sheet of paper to accommodate all the informa-
tion he needed to engage with. Another useful piece of advice was this: once one has drawn
the overview map, it may be helpful to exercise selective inattention in order not to get over-
whelmed by all potentially relevant details all at once.
Halliday was realistic about what can be achieved. Noting the enormous effort involved in
developing a description of a “new” language as one’s PhD project, he said that once a student
has successfully completed his or her PhD he would say something along the lines of “Well
done! Now set your description aside, and start again ignoring descriptions of English (or what-
ever major language might have influenced the description)”. In other words, the task cannot be
achieved in one go; one must adopt a kind of helical movement.

30
 Introduction to the notion of a guide to SFL

Figure 1.10 Holistic approach vs. Cartesian Analysis

1.5 Start with phenomena – not with the study of phenomena

Students of linguistics or applied linguistics are likely to have the starting point of their way into
linguistics determined by the programme – the introductory core course. And while research-
ers have a choice as to how to move in, in principle, this is often pre-selected by expectations
deriving from research committees, panels and funding agencies. In general, Dylan Thomas’s
approach adopted for Under Milkwood seems insightful:

To begin at the beginning:

It is spring, moonless night in the small town, starless and bible-black, the cob-
blestreets silent and the hunched, courters’-and-rabbits’ wood limping invisible down
to the sloeblack, slow, black, crowblack, fishingboat-bobbing sea. The houses are blind
as moles (though moles see fine tonight in the snouting, velvet dingles) or blind as
Captain Cat there in the muffled middle by the pump and the town clock, the shops
in mourning, the Welfare Hall in widows’ weeds. And all the people of the lulled and
dumbfound town are sleeping now.

That is, “to begin at the beginning” means that we are well advised to adopt a thematic
approach – or a macro-thematic one (cf. Martin, 1993b), which we can interpret as one that
provides us with an orientation to our course of study or research plan. But what constitutes
material for an insightful orientation? The short answer for much of the history of linguistics
since the early 1960s has been: current linguistic theory. But we can imagine another

31
 Introduction to the notion of a guide to SFL

answer – viz. language in context, i.e. the realm of phenomena that linguistic theories are
designed to engage with rather than the study of this realm of phenomena.
Thus Halliday used to say that the first year of a linguistics programme should not be about
linguistics but about language: since while language is pervasive in their daily lives, students are
not aware of it, or only of certain exposed aspects, and it is just part of the background, so they
need to learn to attend to language, i.e. to notice and observe it systematically – they need to
develop a feel for it (see Matthiessen, 2022). One method is to learn how to keep a discourse
diary, logging the uses of language and their context throughout a day or a week, as illustrated
in Figure 2.10 (e.g. Ure and Ellis, 1977; Halliday, 1985f; and cf. also Halliday, 1984c). Another
method is focussed on the expression plane; students need to learn not only to listen attentively
to spoken language but also to manipulate it practising what we might call phonetic yoga (cf.
Matthiessen, 2022). For example, they can start by pronouncing one of the vowels of a language
they are familiar with, then keeping everything else constant, they can alternate between spread
and rounded lips. This will begin to give them a sense of articulatory systemic contrasts. Then
they can keep the lips constant, but change the position of the tongue, either “horizontally”
or “vertically”. Once they have developed a conscious analytical sense of producing familiar
sounds, including the correlation between articulatory choices and auditory distinctions, they
can begin to experiment with sounds from languages they are not familiar with.
When we engage with language, learning to observe it and reflect on it, various properties
built into the theory of language in context in SFL are likely to stand out. We’ll mention two
central and directly related ones – that language appears to us as events, not as static construc-
tions and that language is a resource for making meaning, not a rule system. Already in the
article that’s often cited as the starting point for the scale-&-category phase of the development
of SFL, Halliday (1961/2002d) emphasizes the eventive nature of language:

The data to be accounted for are observed language events, observed as spoken or as
codified in writing, any corpus of which, when used as material for linguistic descrip-
tion, is a “text”.13
(p. 38)

The unit being the category of pattern-carrier, what is the nature of the patterns it
carries? In terms once again of language as activity, and therefore in linear progression,
the patterns take the form of the repetition of like events. Likeness, at whatever degree
of abstraction, is of course a cline, ranging from “having everything in common” to
“having nothing in common”.
(p. 46)

(The eventive nature of observed language can of course be highlighted by using the term “lan-
guaging”, which Halliday, 1985c/2003b: 192, comments on as follows: “the noun language is of
course a grammatical metaphor – Peter Doughty preferred to talk about languaging, to show that
our object of study is more process than entity”.)
The other central property that emerges when we engage with language is that it is a
resource – more specifically, a resource for making meaning (what Halliday, e.g. 1973, has called
a “meaning potential”). Halliday (1977a/2003b: 94–95) points out that this is the experience of
language that young children have and grow up with before they enter school:

By the time he is two years old, a child has a considerable awareness of the nature and
functions of language. When he starts to talk, he is not only using language; he is also

32
 Introduction to the notion of a guide to SFL

beginning to talk about it. He is constructing a folk linguistics, in which (i) saying, and
(ii) naming-meaning, denote different aspects of the same symbolic act. And language
functions for him both in reflection and in action: as a way of thinking about the world
(including about himself), structuring his experience and expressing his own personal-
ity, and as a way of acting on the world, organizing the behaviour of others and getting
them to provide the goods-and-services he wants. . . .
Soon, however, the child will go to school; and once he is there, his ideas about
language will be superseded by the folk linguistics of the classroom, with its categories
and classes, its rules and regulations, its do’s and, above all, its don’ts. Here a fundamen-
tal ideological change takes place in the child’s image of language – and, through this,
in his image of reality. Up till now, language has been seen as a resource, a potential
for thinking and doing; he has talked about it in verbs, verbs like call and mean, say and
tell, and rhyme. From now on, language will be not a set of resources but a set of rules.
And the rules are categorical – they operate on things; so he must talk about language
in nouns, like word and sentence, and noun and verb, and letter.
...
So we have enshrined in our folk linguistics these two views, one of language as
resource, the other of language as rule. The two co-exist; but since one is a product
of our primary socialization, and belongs to the reality that is learnt at our mother’s
knee, while the other is part of a secondary reality and belongs to the realm of
organized knowledge, they impinge on each other scarcely at all. But in our prevail-
ing ideology, the dominant model is that of language as rule (our schools teach the
formal grammar of logic, not the functional grammar of rhetoric); and it is only
when we come across the writings of those with a different vision of language,
like Malinowski, Hjelmslev and Whorf, or alternatively when we make a deliberate
effort to change the prevailing image, as some teachers and educators are trying to
do, that the notion of language as resource surfaces from our unconscious and we
begin to build on the insights that we possess by virtue of this simple fact, observed
from the moment of birth (if not before), but so easily forgotten by the philosophers
of language, that people talk to each other.

In this context, it is worth noting examples of how one can move into a linguistic paper: many
of J.R. Martin’s contributions are grounded in a relevant text in its context, ensuring what J.R.
Firth called the renewal of connection. For example, Catford (1969: 200) writes: “Ulti-
mately, ‘abstract’ linguistics gets its ‘justification’ when ‘the linguist’ ‘finally proves his theory by
a renewal of connection with the processes and patterns of life’ and ‘experience’.”
By starting with the linguistic – and more generally semiotic – phenomena rather than
with the study of such phenomena, researchers can avoid reifying phenomena according to
particular traditions. But this can obviously be hard in the institution of academia: certain
constructs become popular or fashionable – like universals, impoverished input; ideology,
identity, gender; and scholars and students feel tempted or obligated to address them, taking
them more or less for granted. Here academic communities may become divided, as in the
tension between the stance taken in Conversation Analysis, where categories can only be
posited in the close examination of particular conversations if the interactants can be shown
to orient to them, and Critical Discourse Analysis, where categories such as power, gender
and class are often assumed. Methodologically, studies informed by SFL should foreground
the need for dialogic shunting, in the attempt to arrive at positions that synthesize thesis and
antithesis positions.

33
 Introduction to the notion of a guide to SFL

1.6 The context of reading and using SFL

This guide is organized into eight chapters (see the next section). They are all designed so that
they can be read independently of one another, and used for purposes of reference. How they
are accessed will naturally depend on the contexts in which users of the guide operate. There
are probably innumerable different contexts of use, but we can identify a number of likely uses:

• users starting on a course of study of SFL in order to use its resources in research and applica-
tions. Here Chapter 1 will provide an overview sketch, and Section 2.2 can then be consulted
for introductory and reference literature. The choice of other parts of the book to focus on
next will depend on the particular areas of interest: different areas are detailed in Chapter 3 – for
example, contributions to semantics in Section 3.1.2; and guides to different areas of applica-
tion, including ones “internal” to linguistics (“immanent”, in Hjelmslev’s sense), are provided
in Chapter 4. The first two sections, 4.1. and 4.2, are relevant to applications in general.
• users who have background in lexicogrammatical or discourse analysis informed by SFL
and who want to continue to use SFL in conducting analysis (rather than description) may
want to start with the first two sections of Chapter 4, Sections 4.1 and 4.2, and then move
on to Section 4.4 on text (discourse) analysis while using Chapter 7 as a reference resource.
• users who have experience with SFL in text analysis (including lexicogrammatical text
analysis) and would like to initiate a project describing a language in terms of SFL may want
to review the first two sections of Chapter 4, Sections 4.1 and 4.2, and then move on to
Section 4.3 on the application of SFL in the development of language systems.
• users coming to SFL from another member of the family of functional text-based approaches
to language may find it helpful to start with Section 2.6 on SFL and other contemporary
functional text-based approaches, and then continue according to their interest, say in text
analysis of language-system description.
• users interested in some area of application other than, or in addition to, text analysis and
system description, can turn Section 4.6 on translation, Section 4.7 on healthcare commu-
nication or Section 4.8 on other areas.
• users intending to criticize SFL from the standpoint of some other linguistic framework will
do well to review Chapter 1 to ensure that they have a sense of the formative concerns of
SFL and theoretical framework, since history is full of examples of scholars misunderstand-
ing metalanguages that are quite different from their own, and then move on to Section 2.2
for pointers to SFL introductory and reference literature. Chapter 5 may then be helpful to
consult since it deals with alternatives, questions, issues and debates within SFL.

1.7 The organization of the guide to SFL

This guide to SFL is organized into a number of guides to particular areas or activities:

Chapter 2, Reading guides: This chapter is designed to help readers find areas of the
SFL literature relevant to their interest. It identifies key publications that can be consulted
in the course of exploring some particular area of interest, including online publications
and resources such as corpus tools.
Chapter 3, Resource guides: This chapter surveys the resources of language in con-
text, including semiotic systems other than language, providing an “index” into these
resources with pointers to the relevant literature.

34
 Introduction to the notion of a guide to SFL

Chapter 4, Application guides: This chapter is concerned with SFL in use – with how
to go about deploying what SFL offers in terms of general methodology, and with the
use of SFL in particular areas of application (language description, text [discourse] analy-
sis, education, translation, healthcare).
Chapter 5, Issues guides: This chapter is concerned with debates in SFL, identifying
theoretical and descriptive issues that have emerged since the 1960s. These issues range
from ones that are of a fundamental theoretical nature, like the number of strata in terms of
which language is interpreted, to particular descriptive issues, like the question of where to
draw the boundary between ‘material’ and ‘relational’ clauses in the description of English.
Chapter 6, Term guides: This chapter provides a brief overview of terminology in SFL
as part of the general “metalanguage”, and identifies publications particularly focussed
on systemic functional terms in English and other languages used in SFL publications.
Chapter 7, Analysis guides: This chapter is designed as a complement to workbooks in
SFL dealing with contextual, semantic and lexicogrammatical analysis of texts; it includes
vocabulary guides compiled based on known challenges in the analysis of English text in
the identification and classification of grammatical and lexical items.
Chapter 8, Conclusion: This chapter takes a step back to locate our guide to SFL in rela-
tion to SFL and the SFL community in general, and identifies gaps that need to be filled
for example by undertaking new research projects.

While there are obviously interdependencies and cross-references, the chapters have been writ-
ten so that they can be consulted independently of one another. After the chapters, we included
the Appendix, a list of material that is not included in the main chapters of the book but which
are crucially relevant to some, most or all of the chapters.

Notes
1 See: www.santafe.edu
2 Halliday (2015: 97) comments: “At a very general level, I have always thought of linguistic theory as
something to be applied, to real problems either in research or in some domain of practice; eventually I
came up with the term ‘appliable’ linguistics to encapsulate this preoccupation with a theory as a mode
of action that is based on understanding.”
3 Some strands within SFL over the years have pursued different kinds of architecture, and some of them
have turned into distinct lines of development, an important trajectory being that initiated by Richard
Hudson around 1970, when he set out to provide a non-transformational alternative to Chomsky’s
transformational grammar: first Hudson’s (1971) “generative systemic grammar” and then his (1976)
“Daughter Dependency Grammar” or DDG. While DDG still had strong links to systemic functional
grammar (SFG; cf. Hudson, 1987), he then moved on to develop “Word Grammar” (Hudson, 1984,
2010), which in one sense was a move further away from SFG but in another sense marked a return to
a more Hallidayan approach to language. Being influenced by both Hudson and Halliday, Robin Faw-
cett set out to develop a generative systemic functional grammar – an early theoretical statement being
Fawcett (1973a, 1973b). He developed this into a full-fledged model together with a team at Cardiff
University, and this is often known as the “Cardiff Grammar”, which has led to many publications over
the years (see e.g. Schulz and Fontaine, 2019). This theory is arguably more comparable to computa-
tional linguistics architectural models than to the Hallidayan multi-dimensional relational model. See
further Chapter 5.
4 The global and local dimensions represented in the figure as having as their domain semiotic systems
are specific to semiotic systems. However, as organizational principles a number of them can be shown
to apply to the “architecture” of systems, or at least more complex systems, of all orders (cf. Section
1.4). This can be illustrated by Halliday’s (e.g. 1990a) characterization of the cline of instantiation in
his account of language in context by analogy to a physical system, more specifically meteorology: the
cline from climate to weather with intermediate patterns.

35
 Introduction to the notion of a guide to SFL

5 There are two modes of “specialization”: instantiation in the case of the cline of instantiation and
elaboration or refinement in the case of the cline of delicacy.
6 For recent overviews of SFL and Natural Language Processing/Computational Linguistics, see
O’Donnell (2017) and Bateman et al. (2019). Sketching the history of interaction going back to the
1950s, they review successes and challenges, and highlight the significance of the emergence of statisti-
cal NLP, identifying new opportunities. However, they do not focus on the task for SFL to theorize
processes, i.e. to interpret the modelling of processes in computational linguistic systems in terms of
(expanded) systemic functional theory.
7 Here we use “cosmogenesis” to refer to the genesis of systems of all orders; it has also been used more
specifically to denote the genesis of physical systems, supplemented by other types of genesis like bio-
genesis and anthropogenesis.
8 These two planes are reflected in the Stoic-Saussurean theory of the sign as embodying two “halves”,
signified (content plane) and signifier (expression plane); but semiotic systems are systems of meaning,
not simply collections of signs – as is brought out clearly in Halliday’s (1975a) account of the phases of
ontogenesis.
9 Halliday’s interpretation of context as a semiotic system is a very important theoretical advance, and
it was further illuminated by Martin’s (e.g. 1992a) interpretation of it as a denotative semiotic sys-
tem, drawing on Hjelmslev’s insights. Context has often been referred to as “social context” in SFL.
Although this term makes very good sense, it is important to note that given the ordered typology of
systems, it follows automatically that context and any other semiotic constructs are also social; ana-
lytically, it is fundamentally important to distinguish between social and semiotic characterizations of
context, situations and similar notions. Semiotic characterizations obviously are additive in the sense
that they add “meaning” to the social characterizations.
10 This is of course what lies behind the notion of a “canon” of texts for a given combination of culture
and languages.
11 In her insightful review of systemic functional work on field, tenor and mode, Bowcher (2017: 391)
introduces these three functional parameters of context as follows: “Field, Tenor and Mode together
form a contextual framework for describing the context of situation – that is, they represent the param-
eters across which the relevant features of a situation in which language is involved may be grouped.”
While the notion of “relevant features” relates to context of situation at the instance pole of the cline of
instantiation in particular, the three contextual parameters extend along the cline of instantiation from
the context of culture via the intermediate patterns of institution/situation type to context of situation
at the instance pole.
12 In current terms, adopted later in the 1960s, their “realizations” (cf. Halliday, 1992c, on the Firthian
term “exponence”).
13 In a footnote on “text”, referring to Firth’s work, Halliday (1961/2002d: 73) writes: “Here “text”
refers to the event under description, whether it appears as corpus (textual description), example
(exemplificatory) or terminal string (transformative – generative).”

36
 2
READING GUIDES

This chapter is designed to provide an index into the SFL literature, ranging from general intro-
ductions to particular areas to research monographs and key articles. In the next chapter, we
will focus on a few areas of application, discussing methods and examples of previous studies,
but here we will present a general-purpose map of the systemic functional literature. This map
is designed among other things to enable readers to go beyond introductory works in order to
locate original sources and also in order to find more detailed, research-oriented accounts.
The map is designed fairly much along the lines of a traditional survey of the literature in
different areas. At the end of the chapter (Section 2.7), we discuss an approach to surveying
past research that is more powerful and informative – the systematic literature review involving
research synthesis and often also meta-analysis of past results.

2.1 Pre-systemic literature and influences (up to around 1960)

SFL grew out of the Firthian tradition – variously known as Firthian linguistics, the London School
or, in reference to its approach to phonology, prosodic analysis, and Firth had been influenced by
the Polish-born British anthropologist Bronislaw Malinowski (e.g. Firth, 1957a; Honeybone, 2005;
Butt, 2001, 2019), who has continued to be a source in the development of SFL; but as Halliday
developed what became SFL, he, together with colleagues, drew on other linguistic traditions as
well, including US American anthropological linguistics and the continental European functional-
ism of the Prague School work and Hjelmslev’s interpretation of European structuralism. And Hal-
liday had received his early training in linguistics in China, including his training in linguistic field
work (see Peng, 2015). These various strands of scholarship relevant to the development leading up
to SFL – and also of course to later work within SFL – are summarized in Table 2.1, and Halliday
(1985c) provides a helpful sketch of sources about a quarter century into the development of SFL.
We have identified scholars and the traditions they initiated (roughly from the period 1920–1960),
key contributions within these traditions and helpful secondary sources – overviews in the history
of linguistics and interpretations of the contributions, including discussions within SFL.
Overviews of the pre-systemic literature, and of connections to related traditions include:

• Malinowski: extrinsic functionalism, contextualism – Steiner (1983); Hasan (1985a); Hal-

liday (1985d); Butt and Wegener (2007); the volume of contributions on Malinowski by

37 DOI: 10.4324/9781315675718-2
 Table 2.1 Relevant pre-systemic literature

Scholar and tradition Key original contributions Helpful overviews and interpretations

Malinowski and his Malinowski (1923, 1935, 1944) pre-SFL: R. Firth (1957b) [a critical tribute to Malinowski from an anthropological point of view, but with
functionalism and a contribution by J.R. Firth, 1957b];
contextualism in from the point of view of sociology and anthropology: Turner and Maryanski (1979) [an overview of
anthropology functionalism in anthropology and sociology, summarizing Malinowski’s functionalism and tracking its
influence, including on the work by Talcott Parsons];
SFL: Hasan (1985a) [on the complementarity of Malinowski and Saussure in a theory of meaning]; Butt
and Wegener (2007) [on context in Malinowski, and the 19th c. work by Wegener]; Steiner (1983)
[a general history of contextualism in British linguistics]; Steiner (2005a, 2015) [on the importance of
Malinowski in functional translation studies]
Firth, the London Firth (1957a, 1968) scholars in the Firthian-Hallidayan tradition: Catford (1969) [an informed overview of Firthian linguistics
School and as a whole]; de Beaugrande (1991) [an overview of Firth’s contribution to linguistics in the context of
prosodic analysis other leading linguists of the 20th century, including M.A.K. Halliday, with detailed references to Firth’s

Reading guides
work]; Monaghan (1979) [Neo-Firthian linguistics]; Steiner (1983) [British contextualism]; Bazell et al.
(1966) [contributions drawing and extending Firth’s work]; Hill (1966) [a very accessible account of
38

prosodic analysis, with examples from Turkish]; Butt (2001, 2019) [Firth in relation to Halliday and the
development of SFL, and in Butt, 2019, the intellectual context in which he produced his work];
scholars outside the tradition: Dinneen (1967: Ch: 10) [contextual theory and prosodic analysis]; de
Beaugrande (1991) [on Firth’s work, with extensive quotes from Firth’s publications]; Rébori (2002);
Thomas (2011: 178–184) [entry on Firth as one of 50 key thinkers about language and linguistics]; Senis
(2016a, 2016b) [on the significance of Firth’s prosodic analysis alternative to phoneme theory]
Mathesius, the Mathesius (1911, 1928, 1975); SFL scholars: Halliday (1974a) [Prague’s FSP in an SFL interpretation]; Davidse (1986) [SFL and the Prague
Prague School Trubetzkoy (1939); Vachek School] and cf. Hajičová (2019);
(1960, 1964); Garvin (1964); Outside SFL: Jindrich (1995) [“the embedding of the Prague Circle into the broader cultural ambience of
Daneš (1964, 1974a, 1974b); the 1920s and the 1930s” (p. viii)]; Nekula (1999)
Firbas (1992)
US American Sapir (1921); Whorf (1956) SFL scholars: Halliday (1985a, and elsewhere in grammatical descriptions) [extensive use of Whorf ’s notion of
Anthropological reactances, and his conception of cryptogrammar]; Martin (1988) [Whorfian interpretation of the construal
Linguistics of core aspects of Philippine culture in Tagalog in terms of resonant features of the language – syndromes,
(Boas – Sapir – or “conspiracies”];
Whorf) outside SFL: Lucy (1992); Lee (1996)
 Reading guides

the anthropologist Raymond Firth (1957d) is also a valuable source, including one chap-
ter by J.R. Firth (Firth, 1957b). For non-SFL discussions of Malinowski as a precursor
to pragmatics, see e.g. Senft (2005); Korta (2008), and as a pioneer in functionalism in
anthropology and sociology, see Turner and Maryanski (1979).
• Firth and the London School: system-&-structure theory, prosodic analysis – Robins
(1961); Bazell et al. (1966); Dinneen (1967); Catford (1969); Monaghan (1979); Kachru
(1980); de Beaugrande (1991); Rébori (2002); Honeybone (2005); Plug (2008); Thomas
(2011); and in relations to SFL: Steiner (1983); Henderson (1987); Butt (2001, 2019).
For recent non-SFL discussions of J.R. Firth, see also e.g. Senis (2016a, 2016b) and Léon
(2007).
• The Prague School: Jindrich (1995); in relation to SFL: Davidse (1986); Hajičová (2019);
on “functional sentence perspective” and textual organization in particular, cf. also Halliday
(1974a), Firbas (1992: 99).
• The Chinese context: Zhang et al. (2005), Peng (2015), Halliday (2017).
• Marxism: Halliday (2015); Webster (2019: 49–51).

Naturally, many of these connections have kept being renewed since the pre-systemic period. Butt
(2019) helps readers access Firth by sketching the academic context in which Firth was working; read-
ing Firth in reference to those of contemporaries he wrote for can guard against misunderstandings.

2.2 Overviews of literature on systemic functional linguistics

Since the 1960s, scholars have produced an extensive literature on SFL in general and on particu-
lar areas of activities. For a long time, there was only one introduction to SFL, Margaret Berry’s
two volumes introducing systemic linguistics (Berry, 1975/7). They covered the systemic version
of scale-&-category theory, but not yet the functional addition – i.e. the theory of the metafunc-
tional organization of language and its resonance with the functional organization of context
(Halliday, 1978) – and not the early work on text in context that began to emerge only in the
late 1970s (notably Hasan, 1978, 1979). Two anthologies of papers appeared in this period, Hal-
liday (1976a) and Halliday and Martin (1981), making available a range of contributions that had
appeared in various places or which had not previously been published. While these collections
helped making systemic functional work more accessible, there were still important contributions
that only existed in mimeo, notably Hudson’s (1964) scale-&-category description of the Beja,
Huddleston et al.’s (1968) study of scientific English (but cf. Huddleston, 1971), El-Menoufy’s
(1969) description of English intonation, Barnwell’s (1969) description of Mbembe and Mock’s
(1969) description of Nzema. Of the descriptions of languages other than English that was pub-
lished, Bamgbos e’s (1966) description of the grammar of Yoruba seems to be unique.
˙
The current situation is very different. While there is still no comprehensive overview of SFL
in its manifestation as a general theory of language (and other semiotic systems) with descriptive
examples from a wide range of languages1 – the closest still being Caffarel, Martin and Matthies-
sen (2004), there are (1) introductions geared towards discourse analysis and (2) a number of
edited volumes designed to provide general coverage of SFL – including handbooks, compan-
ions, term guides, thematic edited volumes and edited volumes with more various contributions.

2.2.1 Introductions to SFL

There is, as yet, no comprehensive introduction to SFL as a kind of general linguistics; such
an introduction would cover the general theory of language in context, introduce descriptions

39
 Reading guides

of different languages (including comparison and typology), and give a sense of a range of
applications (for this, we have to turn to edited, multi-authored companions, handbooks and
overviews: see Section 2.2.2). Instead, introductions tend to focus on a particular stratum (often
lexicogrammar) taking account of one particular language, as shown in Table 2.2. There are also
introductions to linguistics informed by the tradition that SFL grew out of – Robins (1964),2
and by SFL itself, a valuable contribution being McCabe (2017). She covers various aspects of
linguistics, including insights from SFL wherever appropriate, and concludes with a chapter on
“fields of linguistics” with contributions by specialists in different areas.
The most common language is, not surprisingly, English, but there is now a growing number
of introductions using a language other than English as the way in, including Chinese (Hu, Zhu
and Zhang, 1989), Spanish (Ghio and Fernández, 2005), Danish (Andersen, Petersen and Sme-
degaard, 2001), Swedish (Holmberg and Karlsson, 2006) and Japanese (Teruya, 2004, 2022b.).
Since Halliday published the first edition of his Introduction to Functional Grammar in 1985
(Halliday, 1985a, 1994c; Halliday and Matthiessen, 2004, 2014), other more introductory books
to the grammar have appeared: Thompson’s (1996, 2004, 2014) Introducing Functional Grammar,
Bloor and Bloor’s (1995, 2004) Functional analysis of English: a Hallidayan approach; Fontaine’s
(2013) Analysing English grammar, Banks’ (2019a) Systemic functional grammar of English and a
workbook has been published: Martin, Matthiessen and Painter (2010). As new introductions
appeared, an ecosystem emerged around IFG and then kept changing with new contributions,
so IFG could expand its coverage and move in the direction of a reference account of English.
These introductions to the IFG account are geared towards applications of the account in differ-
ent contexts, the analysis of English text being a central concern. However, another introduction
to the grammar, Matthiessen and Halliday (2009), was designed as a first step into the theory of
grammar in SFL with examples from Chinese, English and Japanese. The “second step” is taken
by Martin (2013a), and the “third step” by Martin et al. (in press).
The various introductions written to be more introductory than Halliday’s IFG have dif-
ferent orientations. Thompson’s (2014) Introducing Functional Grammar is perhaps the closest to
an introduction to IFG, but it is arguably geared in particular towards text analysis. Butt et al.
(2000) is oriented towards teachers, and Eggins (2004) is an introduction focussed on English
within the “Sydney School” version of SFL developed by J.R. Martin and his colleagues and
students (for this historical background, see e.g. Martin, 1999a; Rose and Martin, 2012; Tann,
2017).
Bloor and Bloor (2013) characterize their introduction as follows:

This third edition is an indispensable introduction to systemic functional linguis-

tics, which can be used independently or in preparation for M.A.K. Halliday and
C.M.I.M. Matthiessen’s Introduction to Functional Grammar. The book is an ideal
text for students of linguistics, applied linguistics and grammar – those new to the
field, or who have a background in traditional grammar, as well as teachers of English
language.

Table 2.2 does not include short chapter-length introductions in handbooks, but there are
a number of helpful overviews of SFL or particular areas of SFL that can serve as the first
way in:

• in handbooks and encyclopaedias on general linguistics: Halliday (1994a); Caffarel (2010);

• in companions to English language studies: Coffin and O’Halloran (2010), who shed light
on “describing English” by drawing both on SFL and corpus linguistics;

40
 Table 2.2 Introductions to different aspects of SFL presented in English – according to strata (see also Table 9.3)

Context Semantics Lexicogrammar Phonology Text analysis

Halliday’s IFG Chapters 2–10 (Chapter 1) (Chapter 9)

Thompson’s (2014) IFG
Butt et al.’s (2000) UFG Chapter 8 Chapters 2–7 Chapter 9
Bloor and Bloor (1995, 2004, 2013) Chapters 3–10 Chapters 5, 11
Fontaine (2013) Chapters 1–8 Chapter 9

Reading guides
Martin, Matthiessen and Painter (1997, Chapters 1–6 Chapter 7
2010)
41

Halliday and Hasan (1976) textual: system of cohesion cohesion

Tench (1990) whole book: intonation
and rhythm
Halliday and Greaves (2008) whole book: intonation
and rhythm
Eggins (1994, 2004) Chapters 3–4 Chapters 5–10 Chapter 11
Eggins and Slade (1997, 2005) interpersonal interpersonal systems – √
systems
Salkie (1995) largely introduction to
cohesion
Martin and Rose (2003, 2007) √ [“discourse √
semantics”]
Martin and Rose (2008) √ [“genre relations”]
 Reading guides

• in handbooks and encyclopaedias on discourse analysis: Halliday (1985b) on lexico-

grammatical discourse analysis, Eggins and Martin (1997) on genre in discourse analysis,
Rose and Martin (2012) on genre in the Sydney School, Schleppegrell (2012) on SFL
in general as a framework in discourse analysis, Kress (2012) on multimodal discourse
analysis, Lemke (2012) on multimodal discourse analysis, partly in autobiographical
mode on discovery and development of approaches, Martin (2015a) on cohesion and
texture;
• in encyclopaedias of communication sciences and disorders: Armstrong (2005, 2016).

Beyond the introductions, there are some volumes that ofer more advanced accounts, e.g.
Martin (1992a) on “English text” – focussing on semantics in context, Matthiessen (1995a)
on English lexicogrammar organized around system networks, with typological outlooks,
Halliday and Matthiessen (1999/2006) on ideational meaning. But, in general, accounts
going beyond the introductions tabulated in Table 2.2 are most likely to be found in the
companions, handbooks and other edited overviews presented under Section 2.2.2 (see
Table 9.3).
The introductions surveyed in Table 2.2 are sorted according to the strata they focus on or
include major introductions to, but the last column of the table is concerned with text analysis –
that is, a focus defined in terms of the cline of instantiation (see Figure 1.1) rather than the hierar-
chy of stratification. While text is in the first instance a semantic unit, as emphasized by Halliday
(1981) (defined by reference to context as “language functioning in context”, e.g. Halliday and
Hasan, 1976), text analysis can be undertaken at any stratum, and there are introductory accounts
ranging from phonology to semantics:

• phonological text analysis (prosodic, i.e. intonation and rhythm): Halliday and Greaves
(2008);
• multi-semiotic systems and “texts”: face to face interaction – speech and gesture: both
Muntigl (2004a) and Martinec (2004) are short pioneering contributions that also serve
as introductions to the nature and description of gesture and its relation to speech, and
Martinec (2004) proposes system networks for gestures operating within the experi-
ential metafunction; Martin and Zappavigna (2019) provide an overview of “paralan-
guage” in general (for filmic representations of face to face interaction, see later in this
list);
• multi-semiotic systems and “texts”: visual expression: Kress and van Leeuwen (2001) offer
an early non-technical overview of multi-semiosis, using the stratification of semiotic
systems into content (which they stratify into “discourse” and “design”) and expression
(which they stratify into “production” and “distribution”) as a major organizing principle.
Drawing on insights gained from computational modelling (e.g. Delin and Bateman, 2002),
Bateman (2008a) introduces a much more technical and explicit framework for analysing
multisemiotic “documents”, differentiating different “layers” (cf. the notion of strata) and
emphasizing the significance of variation according to genre (this framework is presented as
a course by Bateman, 2014); and this approach is now included and expanded in Bateman,
Wildfeuer and Hiippala (2017);
• multi-semiotic systems and “texts”: film: in addition to providing a “toolkit” for transcrib-
ing and analysis of printed pages and web pages, Baldry and Thibault (2006) also extend
this toolkit to films; Bateman and Schmidt (2012) offer a systemic functional framework for
film analysis that is informed by and grounded in film studies, e.g. positioning their work

42
 Reading guides

in relation to the pioneering efforts by Christian Metz (e.g. Metz, 1966), discussing e.g.
his famous construct of the grande syntagmatique and complementing it with their grande
paradigmatique.

The work by Bateman and his extended team promises to be the way forward for systematic
multimodal discourse analysis, taking us beyond the early exploratory work. Their framework is
guiding an increasing number of multi-semiotic research projects.

2.2.2 Companions, handbooks and anthologies

While thematic SFL volumes were published in the last decade or so of the 20th century
(e.g. Ghadessy, 1988, 1993a, 1999; Christie and Martin, 1997; and four volumes of Fest-
schrift for Michael Halliday), in the 21st century (cf. Halliday, 2013a), thematic volumes
have now been supplemented by a number of companions and handbooks. These are
summarized together with other reference sources in Table 9.3 in the Appendix, and set
out here:

• Hasan, Matthiessen and Webster (2005, 2007), Continuing Discourse on Language (cf. Hasan,
1985e): two volumes covering a wide range of areas of research within SFL by specialists
presented in 30 chapters, published on the occasion of M.A.K. Halliday’s 80th birthday.
While now a decade and a half old, it is very comprehensive and includes many leading
experts.
• Halliday and Webster (2009), Continuum Companion to Systemic Functional Linguistics: designed
as a companion to SFL with general chapters on the framework (a general introduction;
methods, techniques and problems; ideas and new directions; corpus-based research; dis-
course studies; key terms) and on different areas of research (language development; language
and other primates; computational linguistics; clinical applications; multimodality; context;
literary stylistics). This companion to SFL is now being revised and will appear in a new edi-
tion in the next couple years.
• Fontaine, Bartlett and O’Grady (2013), Systemic Functional Linguistics: exploring choice: a the-
matic volume focussed on a central principle of paradigmatic organization in SFL, viz.
choice, with contributions by a large number of systemic functional scholars and also by
Sydney Lamb on “Systemic networks, relational networks and choice”. The volume con-
sists of five parts: Part I – Choice: theory and debate; Part II – Cognitive and neurolinguistic
views on choice; Part III – Linguistic constraints on choice; Part IV – Cultural and contex-
tual constraints on choice; Part V – Interpreting choice.
• Webster (2015a), The Bloomsbury Companion to M.A.K. Halliday: designed as a companion
to M.A.K. Halliday, and thus to his linguistics – SFL, organized into four parts (Halliday’s
life; Halliday: the making of a mind; Halliday: ideas about language; directions and devel-
opments from Halliday). It includes one chapter by Halliday himself on influences from
Marxist theory.
• Bartlett and O’Grady (2017), The Routledge Handbook of Systemic Functional Linguistics. This
handbook covers a wide range of topics within SFL in 41 chapters. It is organized into five
parts: Part I – A theoretical overview; Part II – At clause rank; Part III – Below the clause;
Part IV – Above the clause; Part V – SFL in practice: an appliable theory.
• Thompson et al. (2019), The Cambridge Handbook of Systemic Functional Linguistics. This
handbook covers a number areas of SFL in 16 chapters, but it is a more selective handbook

43
 Reading guides

than Bartlett and O’Grady (2017). It consists of two parts: Part I – SFL: the model (Firth
and the origins of systemic functional linguistics, key terms in the SFL model, semantics,
the clause, the rooms of the house, context and register, intonation, continuing issues
in SFL, the Cardiff Model of functional syntax, SFL in context) and Part II – Discourse
analysis within SFL (models of discourse in systemic functional linguistics, cohesion and
conjunction, semantic networks, discourse semantics, appraisal, SFL and diachronic stud-
ies). It leaves out a number of areas, including multimodal studies – descriptions of various
languages, language comparison and typology, translation studies; phonology (apart from
prosodic phonology) and morphology; phylogenesis and ontogenesis; institutional linguis-
tics in SFL (e.g. education, healthcare, language and the law); SFL and systems thinking.

Certain areas within SFL have been the focus of handbooks and thematic edited volumes, which
often contain contributions from vantage points other than that of SFL; for example:

• Educational linguistics: Halliday (2016); Hasan and Martin (1989); Christie and
Martin (1997); Christie and Martin (2007); Rose and Martin (2012); Christie (1985,
1990, 2002, 2005, 2012); Fine (1988b); Christie and Derewianka (2008); Schleppegrell
and Colombi (2002); Schleppegrell (2004); Byrnes (2006); Ortega and Byrnes (2008);
Byrnes, Maxim and Norris (2010); Byrnes (2019); Mickan (2019); language and other
semiotic systems in education cf. McCabe, O’Donnell and Whittaker (2007): many con-
tributions now, but a pioneering one is Mohan (1986) – cf. also Dalton-Puffer (2007).
Here we can also include developmental studies: Halliday (1975a, 2003a); Painter (1984,
1999, 2017); Torr (1997, 2015); Williams (2019), which are linked to studies of school
discourse by McCabe (2021).
• Multimodal (multisemiotic) studies, including multimodal discourse analysis (MDA):
van Leeuwen and Jewitt (2001); Ventola, Charles and Martin Kaltenbacher (2004);
O’Halloran (2004); Royce and Bowcher (2006); Ventola and Moya-Guijarro (2009); Jones
and Ventola (2008); Jewitt (2009, 2014); Dreyfus, Hood and Stenglin (2011); Starc, Jones
and Maiorani (2015); Moya-Guijarro and Ventola (2022).
• Translation studies: Steiner and Yallop (2001); Munday and Zhang (2017); there is now
a new volume introducing translation studies informed by SFL: Wang and Ma (2021), and
a volume of interviews they have conducted with systemic functional linguists involved in
translation research, Wang and Ma (2020). Kim et al. (2021) provide an overview of “the
interaction between SFL and translation studies”, with examples of a number of major lan-
guages from around the world.

There is now a large number of edited volumes that are not strongly thematic, often arising
out of congresses and conferences, but which contain varied and valuable contributions; they
include Hasan and Fries (1995); Banks (2004); Wu, Matthiessen and Herke (2008); Fang and
Wu (2009); Neumann et al. (2017); Baklouti and Fontaine (2018); and Brisk and Schleppe-
grell (2021).
There are a number of papers and edited volumes with contributions covering particular
geographic regions, and journals providing venues for scholars from particular regions:

• China: Annual Review of Functional Linguistics; Peng (2015); Webster and Peng (2017); cf.
also Huang (2011) on “challenges of developing SFL in China”;
• South-East Asia: Rajandran and Manan (2019);

44
 Reading guides

• Japan: JASFL Journal;3

• Brazil: Revista Letras; Barbara and Cabral (2013); Vasconcellos (2009).

Helpful overviews of diferent aspects of SFL can also be found in Festschrifts dedicated to:

• M.A.K. Halliday: (Steele and Threadgold, 1987, with contributions by a wide range of
scholars, also outside SFL; and the three volume series of meaning and choice in language:
Hasan and Martin, 1989 [educational linguistics], Berry et al., 1996 [mostly lexicogram-
matical accounts], Fries and Gregory, 1996 [discourse in society]); and see also Hajičová
(2019);
• Ruqaiya Hasan: Bowcher and Liang (2016);
• Michael Gregory: de Villiers and Stainton (2001);
• Erich Steiner: Kunz et al. (2014);
• Theo van Leeuwen: Zhao et al. (2018);
• Robin Fawcett: Tucker et al. (2020);
• J.R. Martin: Zappavigna and Dreyfus (2020).

Interviews with systemic functional scholars often provide a helpful complementary view of
their contributions to their own writings:

• M.A.K. Halliday has been interviewed on many occasions by scholars from different back-
grounds, probably starting with Halliday (1974b), where he was interviewed by Herman
Parret, who also interviewed other leading linguists of the time. Many interviews with
Halliday have been collected by Martin (2013b).
• Ruqaiya Hasan: In her own words: an interview with Ruqaiya Hasan, Butt and Liang (2015), an
interview which traces Hasan’s career and thinking and where among other topics Hasan
talks about the complementarity of Bernstein and Vygotsky.
• “Key figures” contributing to social semiotics (Theo van Leeuwen, Gunther Kress, Jay
Lemke, J.R. Martin, Christian M.I.M. Matthiessen): Andersen et al. (2015).
• Gunther Kress (see Li, 2020):
www.tandfonline.com/doi/abs/10.1080/01596300500199908;
https://cgscholar.com/community/community_profiles/new-learning/community_
updates/94214;
www.learntechlib.org/p/103359/
• Bev Derewianka: In conversation with Professor Beverly Derewianka www.youtube.com/watch?
v=4L51RpQQH5Y&t=3s
• Bev Derewianka and Sally Humphrey: interviewed by Robyn Cox, Teaching knowledge
about language www.youtube.com/watch?v=fXtguIyc7Eg
• Fran Christie: interviewed by Robyn Cox, Writing as a necessary dimension of language and
literacy education www.youtube.com/watch?v=LTHaO23qrKY
• Christian M.I.M. Matthiessen: Matthiessen (2000), Matthiessen et al. (2022); Martin and
Matthiessen (2011) interviewed at TESOL: http://blog.tesol.org/interview-with-professors-
james-martin-and-christian-matthiessen/

There is also a book of interviews with scholars who have made contributions to the study of
translation informed by SFL: Wang and Ma (2022).

45
 Reading guides

2.2.3 Term guides

Systemic functional terminology naturally overlaps to a great extent with terminology in lin-
guistics general, but there are differences worth noting:

• there are terms that are distinguished in SFL but not in linguistics in general, notably
“group” and “phrase”. SFL follows Bloomfield (1933) in making a distinction between
“endocentric” and “exocentric” constructions: groups are endocentric constructions –
expansions of word complexes, whereas phrases are exocentric constructions – like clauses,
and in fact often interpretable as “shrunken” clauses;
• there are terms from other linguistic traditions that have been borrowed in SFL but which
are now not widely used outside SFL, like “agnation” and “agnate” from Gleason (1965);
“stratum” and “stratification” from Lamb’s Stratificational Linguistics; “reticulum” (as in
Martin, 1992a) from the Hartford Stratificationalists; “connotative” vs. “denotative” semi-
otic” from Hjelmslev (1943);
• there are terms that are fairly restricted to SFL – certainly terms originating in SFL like
“system” and “system network”, “rank”, “rankshift”, “metafunction”, “grammatical met-
aphor”, “delicacy”, “cline of instantiation”, “logogenesis”, the “phases of ontogenesis”,
“protolanguage” in the sense of the first phase of ontogenesis. These include quite a few
descriptive terms, although some of them are now used outside SFL as well, like “polar-
ity”: many terms in the description of various systems such as transitivity (Medium, Range,
Scope, Senser, Phenomenon, Sayer, Verbiage, Carrier, Attribute, Token, Value, etc.),
modality (modalization, usuality, modulation, readiness, etc.), tone (tone 1, tone 2, etc.),
logico-semantic type (projection, expansion; elaboration, extension, enhancement); and
the interpersonal system of appraisal;
• there are terms that are used in SFL and also in a number of other traditions, but sometimes
in somewhat different senses, like “register”, “genre”, “taxis”, “hypotaxis”, “parataxis”,
“ergative”, “equative”, “Goal” (taken from Bloomfield, 1933, roughly comparable to “patient”
outside SFL in the case grammar tradition).

Since systemic functional linguists strive for holistic theory of language and comprehensive
descriptions of particular languages as an expression of the complexity of language, there
has always been a pressure to find appropriate terms. On the whole, Halliday and other sys-
temic functional linguists have tended to adapt or re-purpose existing terms – in contrast with
Hjelmslev (e.g. 1943), who chose to introduce a great number of new terms as part of his Glos-
sematics. Interesting issues also arise in the translation or coining of terms in languages other
than English. For example, in Japanese, scholars have either tended to go with the prevalent
approach to using Sinitic resources for technical terms or they have followed the approach
adopted in Okudian functional linguistic of using native Japanese terms on the grounds that
they are more accessible to users outside linguistics, importantly in the context of education
(Teruya, 2022a).
Readers of SFL may thus need help with terms, and term guides have been published since
around 1980, either in the form of full-length books – two of them, 30 years apart: de Joia and
Stenton (1980) and Matthiessen, Teruya and Lam (2010), which has been translated into Chi-
nese by Huang and Chen (2016) – or as appendices to introductions or edited volumes: Halliday
and Martin (1981) [list of technical terms at the end of the collection of papers]; Matthiessen
(1995a); Matthiessen and Halliday (2009); Halliday (2009a); Ghio and Fernández (2005) [Span-
ish glossary based on Matthiessen and Halliday, 2009]; Martin (2013: Appendix 1). Andersen,

46
 Reading guides

Petersen and Smedegaard (2001: 338–341) provide a Danish-English list of terms in SFL, and
Holmberg and Karlsson (2006: 210–218) include a table of terms in Swedish, English, Danish
and Norwegian.
In Matthiessen, Teruya and Lam (2010: 1–44), we discuss the nature of technical and sci-
entific terms in SFL, and show what aspects of the overall resources of SFL such terms repre-
sent, noting that terms obviously only lexicalize certain meanings within the total resources.
The terms, which are all taken from accounts of SFL in English, are listed alphabetically, as in
a dictionary, but we emphasize that they form a network – one that would be revealed in a
thesaurus view of the terms. This term guide has now also been translated into Chinese and
published as Matthiessen, Teruya and Lam (2016). Halliday (2009b) offers 13 terminological
displays that show how terms are related to one another, as illustrated by his entry for “theory”:
see Figure 2.1. Halliday (1984a) discusses a number of the challenges encountered with lin-
guistic terminology. One central challenge is that key categories such as Subject are ineffable
(1984a/2002d: 306–307):

But that which makes the category of Subject learnable is also that which ensures that
it will be ineffable. How can we generalize, in a single definition, or even in an article
or a book, the whole of the shared experience of Subjecthood of the adult speech
community – or even that of one novitiate member of it?
If a language had been a designed system, matters would have been different.
Designed systems are designed so as to be effable; in fact, effability is a necessary
condition of design. You cannot design unless the principles can be made explicit.
But a language is an evolved system; and evolved systems rest on principles that are
ineffable – because they do not correspond to any consciously accessible categorization
of our experience. Only the relatively trivial meanings of a natural language are likely
to be reducible to (meta-)words. Fundamental semantic concepts, like those underly-
ing Subject, or Theme, Actor, New, definite, present, finite, mass, habitual, locative,
are, in an entirely positive way, ineffable.

While we cannot overcome the inefability of such concepts, we can manage it by locating them
in relational networks, as in Figure 2.1 or more generally in a thesaurus, so that we can explore
them in terms of their valeur in the overall system.

2.3 Descriptions of the lexicogrammar of English

After developing a pre-systemic functional description of Chinese (e.g. Halliday, 1956b, 1959),
Halliday turned to the task of describing the lexicogrammar of English. The results of his long-
term project of describing it were, of course, synthesized in his Introduction to Functional Grammar
(IFG), starting with the first edition in the mid-1980s (Halliday, 1985a). The foundational stud-
ies from the 1960s appear in Halliday (2005a), together with more recent investigations couched
in terms of the current IFG framework – including his seminal investigation of “the grammar
of pain”.
Building on Halliday’s IFG description of English, many systemic functional linguists have
extended the coverage in various ways (for more introductory versions, see Table 2.2; for an
introduction providing a bridge from “formal” grammar [really traditional grammar] to func-
tional grammar, see Coffin, Donohue and North, 2009). For an overview of developments
within the IFG framework as of the mid-noughts, see Matthiessen (2007a). The most compre-
hensive extension to date is Matthiessen (1995a), where the description of the lexicogrammar

47
 Reading guides

Figure 2.1 Display showing lexico-semantic relations between “theory” and related terms in SFL, adapted
with changes from Halliday (2009b: 230)

of English is organized in terms of system networks.4 At the same time, there have been many
extensions to the description of different areas of the lexicogrammar of English, including:

• Textual: Ghadessy (1995) is an edited volume dealing with thematic organization in six
parts (I Emergence and development of the concept of Theme; II Practical implications of
thematic development [including translation]; III Thematic development in academic and
non-academic texts; IV Contribution of thematic development to stylistics; V The company
the Theme keeps; VI Thematic development and textual accountability); Hasan and Fries
(1995) is an edited volume with contributions to textual and interpersonal descriptions in
a number of languages, and it includes one chapter on theme in Old English (Cummings,
1995) and two on Theme in relation to text organization (Fries, 1995). Forey and Sampson
(2017) review descriptions of textual systems, focussing on the system of theme, also with
reference to semantic patterns. Their overview is complemented by Clarke’s (2017) discus-
sion of the system of cohesion. Halliday and Greaves (2008) provide an updated version of
the description of the system of information as part of their description of English intona-
tion (see also Smith and Greaves, 2015).
• Interpersonal: Hunston and Thompson (1999) contains contributions by linguists from
different traditions, including a chapter by Thompson and Zhou (1999) on certain inter-
personal Adjuncts in text, and an early characterization of the system of appraisal by Martin

48
 Reading guides

(2000). Halliday and Greaves (2008) describe the interpersonal significance of the system
of tone in English. While interpreted as a semantic system by Martin and White (2005),
appraisal is also relevant to the extension of the description of the interpersonal lexico-
grammar of English, with implications for the account of interpersonal lexis (see also e.g.
Hood and Martin, 2007). Andersen (2017) reviews descriptions of interpersonal systems,
focussing on speech function and mood in particular.
• Experiential: Davidse (1999) explores various aspects of the system of transitivity,
including the complementarity of the ergative and transitivity models of transitivity and the
nature of relational clauses. She has also built a bridge to Langacker’s Cognitive Grammar,
and work by her and students and colleagues drawing on both frameworks has continued
to illuminate ideational lexicogrammar (e.g. Vandelanotte, 2009 on projection). Contribu-
tions involving the description of experiential systems and their use in analysis are included
in Martin and Veel (1998), Simon-Vandenbergen, Taverniers and Ravelli (2003), Jones and
Ventola (2008). The description of the experiential lexicogrammar has also been extended
towards lexical delicacy: Tucker (1998) on adjectives and Neale (2002, 2006); Matthiessen
(2014c) on the system of process type. Davidse (2019) reviews descriptions of the experi-
ential systems of the lexicogrammar of English, also with reference to some work on other
languages.
• Logical: An early account of the clause complex in scientific English by Hudson can be
found in Huddleston et al. (1968). Nesbitt and Plum (1988) illuminate the logical systems
of clause complexing in probabilistic terms (discussed by Halliday, 1991b), followed up by
Matthiessen (2002a, 2006). The relationship between the grammar of clause complexing
and the semantic organization of texts as logico-semantic or rhetorical relational complexes
is theorized and investigated empirically by Matthiessen and Thompson (1988), Matthies-
sen (1995a, 2002a), and Stuart-Smith (2001) and Matthiessen and Teruya (2015c) report
on studies of the lexicogrammatical realization of relations in such semantic complexes that
include clause complexes. Butt and Webster (2017) offer an introduction to the logical
resources of grammar.

For a summary of descriptions of English, see Table 9.2 (Descriptions of the lexicogrammar of
English in the Appendix); cf. also Table 3.15.
Soon after Halliday published his accounts of various areas of the lexicogrammar of
English in the late 1960s (Halliday, 1967/8, 1970d), other systemic functional scholars
building on his work began to introduce variant descriptions – early ones being Hudson
(1971) and Fawcett (1974–76). Drawing on his early work in the 1960s and early 1970s,
Richard Hudson went on to develop frameworks that were different in various respects, first
his Daughter Dependency Grammar (Hudson, 1976, 1987) and then his Word Grammar
(first presented in Hudson, 19845), both of which used English as the language of descrip-
tion. Robin Fawcett has stayed closer to Halliday’s SFL, in terms of both theory and the
description of English, but together with colleagues and students at Cardiff University, he
has continued to contribute to the systemic functional description of English, developing
a distinctive description of English that clearly resembles Halliday’s but also departs from it
in various ways (see e.g. Huang, 2017; Neale, 2017). The work on the extension of gram-
mar in delicacy to describe lexis in the Cardiff tradition pursued by Gordon Tucker (e.g.
Tucker, 1998) is relevant to all systemic functional accounts of lexicogrammar. This also
includes the challenge of dealing with collocation, idioms and “phraseology”: Tucker (2005,
2007, 2016).

49
 Reading guides

2.4 Variation in systemic functional linguistics: “architecture”

and terminology
In pursuing the development of SFL as a kind of appliable linguistics (Halliday, 2002b, 2008a),
Halliday has consistently worked towards – and emphasized the value of – a powerful theory of
language. For example, in his Foreword to de Joia and Stenton (1980: viii), he writes:

the resources of language are extraordinarily rich, and the ways in which things can
be related to each other are of an intricacy that we have hardly yet begun to conceive
of. The theory should not restrict the kinds of interpretative statements we can make;
it needs to be rich enough to allow for all kinds of elaborations and extensions –
especially extensions in delicacy, given that what we know about language today is
only a fragment of what there is to be found out – without having to be patched up
and mended all the time. For this reason my conceptual apparatus has always been
extravagant rather than parsimonious. Since I was drawing no conclusions from the
form of the theory, beyond the obvious one that language must be such-and-such if
such-and-such a theory can account for it, there was no virtue in restricting its scope.

Being “extravagant”, SFL has considerable scope for a wide range of applications; but this also
means that it has a built-in potential for variation, both dialectal and registerial metalinguistic
variation (cf. Chapter 5). SFL has been developed to serve as a flexi theory (cf. Matthiessen,
2010), one where diferent dimensions can be played of against one another as linguists explore
how to interpret and model some particular domain of phenomena. For example, should the
relationship between clause and text be interpreted and modelled as a compositional one in
terms of a rank-scale or as a realizational one in terms of stratification? The general theory allows
for both options, but precisely because they fall within the compass of the theoretical potential,
it was possible for SF linguists to arrive at the insight that the relationship is interpreted most
productively as one of realization rather than one of composition (contrast from and around
Tagmemics, e.g. Longacre, 1979, on the paragraph as a grammatical unit), as Halliday and Hasan
(1976) and Halliday (1981) make very clear.
Being a flexi theory, SFL has inspired and accommodated a good deal of theoretical variation
over the decades. By analogy with variation in language, we can explore this variation in the SF
metalanguage as being either dialectal variation (i.e. variation according to user) or registerial
variation (i.e. variation according to use). For example, when Richard Hudson attempted to
answer Chomsky’s questions about language in terms of SFL, he was using SFL in a different
way and he produced what we might characterize as a registerial variant, viz. Hudson (1971);
and still pursuing this goal of producing non-transformational answers to Chomsky’s questions,
he developed his framework of Daughter Dependency Grammar (Hudson, 1976), where he also
(as the name suggests) drew on the tradition of dependency grammar going back to Tesnière
(1959, which had also been taken up in computational linguistics in the first half of the 1960s
by David Hayes (and of course further back to late Medieval grammarians [see e.g. Covington,
1984]). From the point of view of the SF metalanguage in general, this was a restricted register.
At the same time, there has also been variation of the other kind – dialectal variation, or at
least that is how this variation has often been presented, viz. the contrast between the so-called
Sydney Grammar, which represents a continuous development since Halliday’s (e.g. 1961) early
work in the 1960s, and the so-called Cardiff Grammar, which was pioneered by Robin P. Faw-
cett in the 1970s (important early publications being Fawcett, 1973a, 1974–1976, 1980, 2000,
2008, 2013, 2017).6 However, at the same time, these variants also differ registerially. While they

50
 Reading guides

have been put to a number of similar uses – notably in computational linguistics – their ranges
of uses differ significantly: the “Sydney Grammar” has been deployed in a very wide range of
contexts and research and application. The educational work is, of course, very well-known,
but there is a growing number of other areas, including translation studies and other types of
multilingual studies, workplace studies, forensic linguistics, clinical linguistics, healthcare com-
munication studies, critical/positive discourse analysis, multimodal studies. And the “Sydney
Grammar” has been used as a theoretical and typological descriptive resource in the description
of a far greater number of languages.
The differences between the “Sydney” and “Cardiff ” traditions are not, of course, limited
to the theory of grammar and the descriptions of the grammars of languages around the world.
While they are based on the same approach to axis, with system networks playing a central role
in the organization of subsystems (including the agreement on lexis as the most delicate gram-
mar and the probabilistic nature of systems), they differ theoretically in terms of the basic orga-
nizational principles – in terms of their “architectures”, including the key aspects of stratification
and metafunction.
The architecture of Halliday’s SFL is entirely relational in nature; the overall organization of
language (and other semiotic systems) in context – and in relations to systems of other kinds – is
theorized in terms of a number of semiotic dimensions, each of which is the domain of some
specific semiotic relation such as instantiation, realization or composition. The architecture of
Fawcett’s SFL is modular rather than relational in nature; the overall organization is specified in
terms of interacting modules – just as in computer software architectures (and thus as in genera-
tive linguistic architectures). It is of the same kind as the architectures we tend to find in accounts
of computational linguistic systems (cf. Matthiessen and Bateman, 1991).
In terms of their descriptions of the lexicogrammar of English, the “Sydney Grammar” and
the “Cardiff Grammar” are close in various ways – not surprisingly, since Fawcett (e.g. 1980,
1987, 2000, 2008) and his team drew on Halliday’s pioneering descriptions from the 1960s
onwards, while both revising and extending them. However, there are areas where Fawcett’s
account is closer to traditional and/or formal accounts, notably in the area of clause combin-
ing. At the same time, there are significant descriptive contributions that can be rich sources for
future developments within both traditions – important examples of Cardiff contributions being
Tucker (1998, 2005, 2007, 2016) and Neale (2002, 2006).
Returning to variation within SFL in general, let us summarize this variation diachronically
focussing on approaches to grammar (SFG) departing from Halliday’s work in the 1960s: see
Figure 2.2. Important developments to note in the context of variation within SFL are:

• Since his early work on what turned into SFG in the 1960s, Halliday has continued this
development over the decades with his students and colleagues (cf. Matthiessen, 2007a).
• In the late 1960s and 1970s, Richard Hudson drew on this work, but attempted to use it
to answer Chomsky’s questions about language, in a sense culminating in his proposal for
a non-transformational alternative to Chomsky’s (Extended) Standard Theory – Hudson’s
(1976) Daughter Dependency Grammar. Moving on, he produced a textbook on sociolin-
guistics, and this provided him with a context for returning to Halliday’s questions about
and engagement with language while retaining the insights he had derived from depen-
dency grammar, and the outcome of this engagement was his Word Grammar7 (Hudson,
1984; and for a fairly recent account, Hudson, 2021).
• In the late 1960s and early 1970s, Robin Fawcett, returning from Africa as a teacher to
the UK to pursue a PhD, was influenced by Halliday’s work on SFG, but also by Richard
Hudson’s revisions and new versions, and while in certain crucially important respects he

51
 Reading guides

Figure 2.2 Systemic functional grammar since the 1960s – offshoots and interconnections

stayed closer to Halliday than to Hudson, he launched an extended research programme in

the 1970s that led to the development of the “Cardiff Grammar”, a programme he has pur-
sued productively together with colleagues and research students. Like Halliday’s research
programme, Fawcett’s has remained productive, and is still continuing.
• Independently of the development of variants of SFG associated with Hudson and Fawcett,
in the 1970s, Martin Kay set out to develop a computational version of SFG8 (like Halliday,
he had been involved in Margaret Masterman’s machine translation group in the 1950s, some
time after Halliday, 1956a, had put forward the use of a “mechanical thesaurus” in what came
to be known as machine translation). The result of his effort was Functional Unification
Grammar (Kay, 1979). One of his key objectives was to develop a functional framework for
representing grammatical resources declaratively that could be used in both computational
generation and analysis (i.e. procedurally different uses of the declarative representation of
the grammatical resources). Kay was employed at Xerox PARC, Palo Alto, and a junior col-
league of his, Robert Kaplan, was influenced by Kay’s ideas, and incorporated some of them
in the joint development with Joan Bresnan of Lexical Functional Grammar.9

The chart in Figure 2.2 is necessarily a simplified temporal map of developments relating to
SFG. For example, it can be amended to include Lamb’s Stratificational Linguistics – although
the interaction between SFL and Stratificational Linguistics is not limited to the interpretation
of grammar (but cf. Lockwood’s, 2003, account, which draws on Halliday’s description of the
grammar of English), and also to include the interactions between SFG and Ron Langacker’s
Cognitive Grammar initiated by Kristin Davidse (cf. also e.g. Vandelanotte, 2009).

52
 Reading guides

But the chart still brings out an important fact about the development of SFL internally and
in relation to other traditions: it is arguably in the area of grammar where the greatest number of
influences have occurred within general linguistics (whether acknowledged or not).

2.5 SFL and other disciplines

Throughout the development of SFL, SF linguists have maintained dialogue with a range of
other disciplines. Sometimes, these dialogues have meant taking and adopting ideas from past
work, as in the case of Malinowski’s contributions to anthropology and Whorf ’s contributions to
anthropological linguistics, but often the dialogues have been true exchanges of ideas and find-
ings, as in the case of the extended dialogue between Bernstein and SF linguists in the develop-
ment of educational linguistics.
In the early days of the 1950s and 1960s, such dialogues were pursued as a natural part of the
development of SFL; since then interdisciplinarity has become widely recognized as a way to
move forward – it has even become fashionable, but in view of this development it is important
to note that the orientation towards and engagement with other disciplines has been part of SFL
from the beginning. At the same time, it is important to note the value that Halliday assigns to
transdisciplinarity. Thus Halliday (1990b/2003b: 169) writes:

At the same time, however, it is noticeable that with all the varied activities that go
under the name of applied linguistics we have still not really achieved a transdisci-
plinary perspective. I say “transdisciplinary” rather than “inter-” or “multi-disciplinary”
because the latter terms seem to me to imply that one still retains the disciplines as the
locus of intellectual activity, while building bridges between them, or assembling them
into a collection; whereas the real alternative is to supersede them, creating new forms
of activity which are thematic rather than disciplinary in their orientation.

The interaction between SFL and other disciplines has been initiated both within SFL and
within other disciplines; in the latter case, this has typically occurred when researchers in other
disciplines found that they needed accounts of language compatible with their own views.
Part of the success these dialogues have had is due to the permeability that characterizes SFL
(e.g. Halliday, 1985c). Thus there have been conferences and publication devoted to inter-
disciplinary dialogue and the crossing of boundaries, typically reflected in their titles, e.g. Martin
and Wodak (2003); Young and Harrison (2004); Thompson and Hunston (2006); Christie and
Martin (2007); Benson and Greaves (2009); Neumann et al. (2017).
Examples of dialogic interfaces with other disciplines include the following:

• Computational linguistics: Teich (2009), Bateman et al. (2019), and see Patrick (2008)
for an example of a project involving linguists, computational linguists and software
engineers.
• Fuzzy theory: Kobayashi (1995); Kobayashi and Sugeno (1994); Matthiessen (1995c).
Starting around 1990, there was an extended period of dialogue between fuzzy theory
and SFL, initiated by Professor Michio Sugeno when he was at the Tokyo Institute of
Technology, and then continued at his lab at the Brain Science Division of the RIKEN
Institute 2000–2005, and beyond. The dialogue included centrally how to represent sys-
tems and other aspects of systemic functional theory by means of fuzzy set theory, but
it went much further, including Michio Sugeno’s notion of computing based on natural
language.

53
 Reading guides

• Sociology: Bernstein (1973); Christie and Martin (2007); Maton and Doran (2017). Start-
ing in the 1960s, there has been an extended and still ongoing dialogue with Basil Ber-
nstein’s sociology (e.g. Atkinson, 1985). It began in the systemic functional engagement
with and interpretation of his theory of codes and cultural transmission (e.g. Hasan, 1973;
Halliday, 1994d), and then it expanded to other areas of his scholarship where he assigned
a central role to language, centrally including the discursive construction of knowledge –
variable across disciplines both in terms of the discourses involved and the nature of the
knowledge structures.
• Education: Christie (1985); Halliday (2007a); Rose and Martin (2012); Williams and
French (2016). The dialogue with scholars and practitioners in education goes back to the
1960s. It provides a powerful model of how to nurture such dialogues thanks to a number of
educators who have been aware of the centrality of language in learning and gone on to do
MAs and PhDs in SFL, creating the field of educational linguistics. Educational linguistics
have nurtured and expanded the dialogue, undertaking education-based research, produc-
ing material for the classroom and designing and implementing approaches, curricula and
programmes. Byrnes, Maxim and Norris (2010) present all aspects of a very successful ini-
tiative at Georgetown University.
• Anthropology: Apart from the “virtual” dialogue with Malinowski, carried on in SFL,
long after Firth’s actual dialogues (e.g. Hasan, 1985a; Halliday and Hasan, 1985; Butt and
Wegener, 2007; Hasan, 2009a), another interesting dialogue has been with Benjamin
Colby. He became interested in SFL because he wanted to explore the possibility of infer-
ring culture from culturally significant text (cf. Colby, 1966; Colby and Colby, 1981),10 and
invited Michael Halliday as a visiting scholar for a few months in the first half of 1980 to the
University of California, Irvine, where Halliday worked on the specification of a systemic
functional grammar for computational modelling with Mark James (see Halliday, 2005a),
which later became the foundation of the Nigel text generation grammar at the Informa-
tion Sciences Institute (e.g. Matthiessen and Bateman, 1991). While it wasn’t realized at
the time, Colby’s vision of the use of SFL for such a project is still relevant. Elsewhere,
researchers have used computational tools to discern cultural trends, as in Michel et al.’s
(2011) “Culturomics” and Google Trends.
• Film theory: While there have been a number of systemic functional studies of film, the
dialogue with film theory, including the pioneering work by Christian Metz (e.g. 1966),
was really initiated by Bateman and Schmidt (2012) (cf. also Bateman, 2007); subsequent
publications have built on the foundational framework they have provided the community
with, e.g. Tseng (2013); Tseng and Bateman (2012); Wildfeuer (2014); Wildfeuer and
Bateman (2017); Bateman, Wildfeuer and Hiippala (2017).
• Primatology: There has been a very productive dialogue between Sue Savage-Rumbaugh
and her group of primatologists within the human-ape culture at the Language Research
Center at Georgia State University (e.g. Dubreuil and Savage-Rumbaugh, 2019) and sys-
temic functional linguists, resulting in a number of publications: e.g. Benson et al. (2002,
2004); Benson and Greaves (2009); Benson and Thibault (2009). One conclusion based on
the systemic functional analysis is that “Kanzi’s considerable language abilities have been
underestimated” (Benson et al., 2002).
• Neuroscience: There has been a long line of SFL-informed contributions to clinical lin-
guistics (e.g. Halliday, 2005d; Armstrong, 2009; Asp and de Villiers, 2010, 2019; Ferguson,
Spencer and Armstrong, 2017), but the dialogue with neuroscience in the context of the-
ory and basic research has so far been more exploratory, although in particular thanks to the
research by García and his group it seems to be expanding: Halliday (1995); Melrose (2006);

54
 Reading guides

García and Ibáñez (2017); Trevisan and García (2019); cf. also Williams (2005a) – and the
potential of the “mediation” through Stratificational Linguistics: Lamb (2013); García, Sul-
livan and Tsiang (2017).
• Network science (in the sense of Barabási, 2016): While SFL has developed an approach
that resonates with network science as it has emerged in the last couple of decades, there
is probably as yet no publication exploring possibilities for dialogue, but the potential is
considerable.

It is of course possible to engage with other disciplines indirectly at another level: systemic
functional linguists have sometimes developed accounts of phenomena that have traditionally
been within the domain of other disciplines. One interesting example is Halliday’s (1993b)
exploration of a language-based theory of learning. Learning theories have typically been within
the domains of psychology and education, so here Halliday “intrudes” on their territory; but
he does so because the “mainstream” approaches had not, in his view, taken account of the
role of language as a central resource in learning. His contribution is a step in the direction of
a language-based theory of learning, and he identifies 21 features that he thinks “are critical
to a language-based theory of learning”. His exploration was received positively; Wells (1994)
pointed out that it complements Vygotsky’s (e.g. 1962) work as a psychologist in the 1930s (cf.
Wertsch, 1985, who played an important role in drawing the attention of “Western” scholars to
Vygotsky’s pioneering research).
While it is not clear that Halliday’s step towards a language-based theory of learning has been
taken up by the “mainstream”, it has led to further research, applications and publications.11
Byrnes (2006) brought Halliday and Vygotsky together in a stimulating volume with contribu-
tions from scholars from different traditions (see also Schleppegrell, 2004: Chapter 2); and in
a recent journal article, Han and Kellog (2019) argue that metaphors introduced in English
discussion of Vygotsky within constructivist and cognitivist interpretations matter (including
Bruner’s, 1986, constructivism, scaffolding12 and narrative) and can lead us away from Vygotsky’s
intended model. Pointing out that Vygotsky was not a cognitivist, they suggest how Halliday’s
work complements Vygotsky’s, and how his notion of construal (e.g. in relation to how children
construe their experience of the world as meaning) can help us transcend constructivist and cog-
nitivist readings of Vygotsky. Shrestha (2020) brings together Dynamic Assessment informed by
Vygotskyan theory and systemic functional linguistics, addressing the key question of (pp. 2–3)
“how DA combined with SFL, as powerful and systematic pedagogic and assessment tools, can
be applied to academic writing assessment and subsequently promote students’ academic writing
and conceptual development over a period of time”.
In the case of Halliday (1993c), the “intrusion” into an area typically dealt with by other
disciplines was necessary because he dealt with phenomena that they had tended to ignore.
And this is of course often the case because language is so pervasive in human life and plays a
central role in many institutions like education, administration, the media, the law, healthcare,
hospitality, diplomacy and politics that may be felt to be the territory of other disciplines. Sys-
temic functional linguists often work in such institutional settings, and in fact they have usually
worked as members of teams with practitioners from the institutions where the work is being
done or with researchers from other disciplines concerned with those institutions – for example,
in education, in healthcare and in workplace studies.
Not surprisingly, dialogue and collaboration across disciplines have been a theme in SFL and
have been examined in a number of SFL conferences and workshops. Examples that have been
documented include education (e.g. Rose and Martin, 2012), healthcare (e.g. Slade et al., 2015)
and computational modelling and AI (e.g. Teich, 2009, Bateman et al., 2019; Patrick, 2008). It

55
 Reading guides

Figure 2.3 The engagement with language in context across disciplines in relation to the stratal organization
of language

seems that in such cases, success depends on the extent to which members of projects can draw
on a (meta-) theoretical model that maps out the territory and brings out the way in which the
different disciplinary areas of expertise complement one another.
It may be helpful to relate the engagement with language in general across different disci-
plines to the stratal organization of language in context (cf. Section 3.1): see Figure 2.3. This fig-
ure suggests that the inner strata of language in context – the “form” strata of lexicogrammar and
phonology (or graphology) – are least likely to be investigated in disciplines outside linguistics in
general, exceptions being computational approaches. Moving to the outer strata of language –
the “substance” strata of semantics and phonetics (or graphetics) – and to context, we are more
likely to find engagement with language in disciplines other than linguistics, and they are of
course often studied in “hyphenated” branches of linguistics such as sociolinguistics and articula-
tory phonetics, each with its own dialogic interface to other disciplines. In addition, we can also
note studies of language in context at lower systemic orders of manifestation – social, biological
and physical.
Expanding on Figure 2.3 to go beyond stratification as a point of reference,13 we can refer to
Halliday’s (1978: 11) map of linguistic studies in relation to other work in other disciplines: see
Figure 2.4. Halliday identifies the inner strata of language, adding the relevant view of language:
“language as system”. He also includes a number of other views: “language as art”, “language
as behaviour” and “language as knowledge”. While he drew this map over 40 years ago, the
general outlines are still highly relevant, and his comments are helpful in exploring theoretical
territories, potential for dialogue and opportunities for application (Halliday, 1978: 11–12):

A diagrammatic representation of the nature of Linguistic studies and their relation to

other fields of scholarship will serve as a point of reference for the subsequent discus-
sion (figure 2.1 [reproduced here as Figure 2.4]). The diagram shows the domain of

56
 Reading guides

literary
studies
archaeology
and prehistory
philosophy
language
history of as art
language logic and
language mathematics
families
linguistic
change communications
engineering
geography language
as system
idiolect
substance: phonic/ phonetics: physics
acoustic
language varieties: graphics articulatory
dialect form: grammar and perceptual
register (diatypic) vocabulary
semantics: physiology
situational conceptual

language language language

as behaviour as knowledge pathology
aphasia etc.
socialization: internalization:
function language language production and
and interaction typology universals understanding

(socio-
linguistics) (psycho-
linguistics)

sociology
psychology
social
anthroplogy
culture human
biology

Figure 2.4 The nature of linguistic studies and their relation to other fields of scholarship

language study – of linguistics, to give it its subject title – by a broken line; everything
within that line is an aspect or a branch of linguistic studies.
In the centre is a triangle, shown by a solid line, which marks off what is the
central area of language study, that of language as a system. One way of saying
what is meant by ‘central’ here is that if a student is taking linguistics as a university
subject he will have to cover this area as a compulsory part of his course, whatever
other aspects he may choose to take up. There are then certain projections from
the triangle, representing special sub-disciplines within this central area: phonetics,
historical linguistics and dialectology – the last of these best thought of in broader
terms, as the study of language varieties. These sometimes get excluded from the
central region, but probably most linguists would agree in placing them within
it; if one could give a three-dimensional representation they would not look like
excrescences.
Then, outside this triangle, are the principal perspectives on language that take us
beyond a consideration solely of language as a system, and, in so doing, impinge on

57
 Reading guides

other disciplines. Any study of language involves some attention to other disciplines;
one cannot draw a boundary round the subject and insulate it from others. The
question is whether the aims go beyond the elucidation of language itself; and once
one goes outside the central area, one is inquiring not only into language but into
language in relation to something else. The diagram summarizes these wider fields
under the three headings, ‘language as knowledge’, ‘language as behaviour’, ‘language
as art’.
The last of these takes us into the realm of literature, which is all too often treated
as if it was something insulated from and even opposed to language: ‘we concentrate
mainly on literature here – we don’t do much on language’ as if ‘concentrating on lit-
erature’ made it possible to ignore the fact that literature is made of language. Similarly
the undergraduate is invited to ‘choose between lang. and lit.’. In fact the distinction
that is being implied is a perfectly meaningful one between two different emphases or
orientations, one in which the centre of attention is the linguistic system and the other
having a focus elsewhere; but it is wrongly named, and therefore, perhaps, liable to be
misinterpreted. One can hardly take literature seriously without taking language seri-
ously; but language here is being looked at from a special point of view.
The other two headings derive from the distinction we have just been drawing
between the intra-organism perspective, language as knowledge, and the inter-organism
perspective, language as behaviour. These both lead us outward from language as a
system, the former into the region of psychological studies, the latter into sociology
and related fields. So in putting language into the context of ‘language and social man’,
we are taking up one of the options that are open for the relating of language study to
other fields of inquiry. This, broadly, is the sociolinguistic option; and the new subject
of sociolinguistics that has come into prominence lately is a recognition of the fact that
language and society – or, as we prefer to think of it – language and social man – is a
unified conception, and needs to be understood and investigated as a whole. Neither
of these exists without the other: there can be no social man without language, and
no language without social man. To recognize this is no mere academic exercise; the
whole theory and practice of education depends on it, and it is no exaggeration to
suggest that much of our failure in recent years – the failure of the schools to come to
grips with social pollution – can be traced to a lack of insight into the nature of the
relationships between language and society: specifically of the processes, which are
very largely linguistic processes, whereby a human organism turns into a social being.

2.6 SFL and other (contemporary) approaches

Throughout its history, SFL has naturally been developing in the environment of other
approaches to language in linguistics. The relationship to such other approaches has, naturally
enough, been very varied, changing over the decades. In a number of cases, SFL has been in a
continuous dialogue with other approaches, as in the case of Lamb’s (e.g. 1966) Stratificational
Linguistics and later versions of it (García, Sullivan and Tsiang, 2017); in other cases, develop-
ments have proceeded in parallel, with little or no interaction.
This is really a continuum, but it may be helpful to distinguish between two poles – one of
high degree of interaction and another of low degree of interaction. (For a particular perspective
on “the role of systemic functional linguistics as a linguistic theory in the twenty-first century”,
see Bateman, 2017.)

58
 Reading guides

2.6.1 Close dialogue partners

At the pole of high degree of interaction, we can identify both approaches that in some sense
derive from SFL and ones that are parallel to it, but have involved a good deal of dialogic
exchanges:

• Younger Sibling tradition:

• Birmingham School: discourse analysis
• (Partial) off-shoots from SFL:
• Social semiotics
• Critical Linguistics, Critical Discourse Analysis (CDA)
• Parallel linguistic traditions:
• The Prague School
• Stratificational Linguistics (also known as cognitive linguistics), now Relational Net-
work Theory
• Tagmemic Linguistics

Later we will discuss in some more detail the Birmingham School, social semiotics and the criti-
cal orientation in Critical Linguistics and CDA.
One linguistic tradition running parallel with SFL that has been a particularly important dia-
logue partner is Lamb’s Stratificational Linguistics (e.g. Lamb, 1966; Lockwood, 1972; Makkai
and Lockwood, 1973), later transformed into cognitive linguistics (e.g. Lamb, 1999) and now
developed into Relational Network Theory (e.g. García, Sullivan and Tsiang, 2017). There have
been two strands of influence here. One is directly associated with the dialogue between Michael
Halliday and Sydney Lamb that started in the 1960s (see e.g. Lamb, 2013; Halliday, 1974b, and
other interviews with Halliday in Martin, 2013; and Halliday’s foreword to García, Sullivan and
Tsang, 2017) and the other with the Hartford Stratificationalists J.R. Martin learned from as a
student in Canada in the 1970s (see e.g. Martin, 2014), including Harold Gleason (e.g. Gleason,
1965) and Waldemar Gutwinski (e.g. Gutwinski, 1976) and their use of reticula in text analysis,
adapted for SFL by Martin (1992a). This latter influence is documented by Martin (2014) in his
“autobiographical essay”, where he also discusses the importance of the network representation
in Stratificational Linguistics in general and the value of being taught Michael Gregory’s version
of scale-&-category theory at York University.

2.6.1.1 The Birmingham School

The Birmingham School originated with the work by Michael Halliday’s junior colleague John
Sinclair (1933–2007). They had worked together fairly closely for part of the 1960s while they
were both at Edinburgh University. Halliday (2008b: 37–38) sketches the background:

Sinclair and I discussed collocation at some length. We set up a project for studying
it, by collecting a corpus of tape-recorded English conversation; and we made certain
predictions regarding the outcome – what we would find out. I guessed that colloca-
tion would turn out to be a purely lexical phenomenon: that is, that the tendency for
lexical items to collocate would be independent of their grammatical environment –

59
 Reading guides

unaffected by unit boundaries and structural relationships. Sinclair took the oppo-
site view, expecting collocational patterns to be dependent on relationships within
the grammar. But we also discovered our own personal complementarity: although
we could both play both parts, I thought more like a grammarian, while Sinclair
thought more like a lexicologist. So when I left Edinburgh for London in 1963,
Sinclair took over the project and ran with it. He took it with him to Birmingham
in 1965, reconceptualized and reorganized it, and published the results in English
Lexical Studies.
(Sinclair, Jones and Daley, 2004; cf. Halliday, 1966b; Sinclair, 1966)

The concerns that have been particularly prominent in the work by Sinclair, his collaborator
Malcolm Coulthard and other scholars who have come in since the earliest days (e.g. Michael
Hoey, David Brazil, Amy Tsui, Susan Hunston, Michael Toolan) fall within the research and
application activities of SFL. The Birmingham School is oriented towards appliability, which
has been evident in various areas, including education – the classic being Sinclair and Coulthard
(1975), and law, where Malcolm Coulthard has been one of the pioneers in developing forensic
linguistics (e.g. Coulthard, Johnson and Wright, 2017). Over the years, there have been various
collaborative engagements involving Birmingham and SFL scholars, and Birmingham contribu-
tions and SFL papers have often appeared in the same publications:

• Corpus-based studies. While early corpus-based studies in SFL had been focussed on
grammar and intonation (e.g. Halliday, 1959; Huddleston et al., 1968; El-Menoufy, 1969),
the Birmingham work was initially focussed on lexis, with important contributions to the
study of collocation (e.g. Sinclair, 1991), but then it also extended towards grammar, in a
sense taking up Firth’s notion of colligation (an approach that might be characterized as the
lexicographer’s dream, e.g. Hunston and Francis, 2000) and applied in translation studies by
Mona Baker (e.g. Baker, 1995).
• Discourse analysis (including dialogue analysis), which began with the pioneering work
on classroom discourse by Coulthard and Sinclair (Sinclair and Coulthard, 1975), drawing
on Halliday’s (1961) scale-&-category model, was then taken further under the leadership
of Malcolm Coulthard (e.g. Tsui, 1986, Coulthard, 1992, on spoken discourse, Coulthard,
1994, on written discourse; and Coulthard, 1985, which is an overview of approaches
to discourse analysis). Discourse analysis in both SFL and the Birmingham School have
included contributions in a critical vein, as in the work by Carmen Caldas Coulthard (e.g.
Caldas-Coulthard, 2003; Caldas-Coulthard and Coulthard, 1995; Caldas-Coulthard and
van Leeuwen, 2002; Caldas-Coulthard and Moon, 2010).
• Forensic linguistics has received a major boost from the Birmingham School. In particu-
lar, Malcolm Coulthard has also been one of the pioneers in the development of forensic
linguistics, including corpus and discourse analysis as a central part of the approach (e.g.
Coulthard and Johnson, 2010; Coulthard, Johnson and Wright, 2017).
• Supporting the work on spoken discourse analysis, Birmingham scholars have contributed
to the description of intonation in English – in particular David Brazil, well known for
extending the account of intonation to discourse (e.g. Brazil, 1975, 1997), and his project is
taken further by O’Grady (2010); for some discussion of his work in relation to descriptions
in SFL, see Tench (1990).
• Analysis of evaluative discourse and the language of appraisal has been a longstanding area
of research in the Birmingham tradition, especially in the work (inspired) by Susan Hunston
(e.g. Hunston and Sinclair, 1999; Hunston, 2010; Hunston and Su, 2017).

60
 Reading guides

Since the Birmingham School has taken the corpus as a central construct from the start, and
began by focussing on lexis, it is natural that scholars have kept drawing on J.R. Firth’s work (as
SFL scholars have done), including centrally his notions of collocation, colligation and prosody
(see e.g. Tognini-Bonelli, 2001, in particular Chapter 8). Here it is interesting to note that the
representation of corpora and the tools for analysing corpora have invited lexical studies in the
first instance and created a lexicographic approach to lexis. This important methodological point
is discussed by Halliday (2002c), and what might be called the Birmingham approach to the rep-
resentation of corpora without annotation (following Sinclair’s, 1992, call to “trust the text”) is
contrasted with what might be called the Lancaster approach involving annotations (which can
of course be set aside) by McEnery and Hardie (2012).
While the Birmingham School and SFL constitute separate strands, they have remained on
speaking terms and had many productive exchanges over the years. Through their editing work,
Susan Hunston and Geoff Thompson have contributed significantly to the dialogue between
the Birmingham School and SFL – in particular, Hunston and Thompson (1999) on evaluation
and appraisal (a volume where Martin’s, 2000, seminal systemic functional account of appraisal
resources in English appeared), further corpus-based work on “evaluation” by Hunston (1993,
2011) and Hunston and Su (2017); and, more generally, Hunston and Thompson (2006) on
corpus linguistic research and SFL.
The range of research and application activities in SFL goes far beyond the list of Birming-
ham School, of course; for example, in SFL there has been much more work in what we call
multilingual studies, in computational modelling, in multimodal studies. But there is a produc-
tive complementarity; for example, in terms of institutional settings, systemic functional lin-
guists have pioneered educational linguistics in the institution of education while Birmingham
scholars have pioneered forensic linguistics in the institution of law.

2.6.1.2 Social semiotics

As far as SFL is concerned, social semiotics began with Halliday’s (1978) framing of language
as a social semiotic system – and, importantly, of the study of social semiotic systems (“the
social interpretation of language and meaning”) as an alternative to the framing provided by the
increasingly dominant macro-discipline of cognitive science (for further details, see also Mat-
thiessen, 2017). This was later taken up, perhaps in particular by Halliday and Matthiessen, and
developed as part of an ordered typology of systems operating in different phenomenal realms
(see Section 1.3.2):

• 1st-order systems – physical systems

• 2nd-order systems – biological systems: physical systems + life
• 3rd-order systems – social systems: biological systems + value (social order)
• 4th-order systems – semiotic systems: social systems + meaning

This ordered typology makes it possible to characterize social semiotic systems quite explicitly
(and to compare and contrast them with bio-semiotic systems, as in Halliday and Matthiessen,
2006): see Halliday (1996a), Halliday and Matthiessen (1999/2006), Halliday (2005b), Matthies-
sen (2007b, forthcoming a).
Halliday’s (1978) book is a collection of papers from the 1970s all concerned with dif-
ferent aspects of language as a social semiotic but also with other social semiotic systems. It
complemented his earlier collection, Explorations in the Functions of Language (Halliday, 1973),
and it contextualized his work on ontogenesis, Learning How to Mean (Halliday, 1975a), where

61
 Reading guides

protolanguage is interpreted as a (primary) social semiotic system. In 1986, Gunther Kress,

Ruqaiya Hasan and J.R. Martin asked Halliday about the origins of his engagement with semi-
otics (Martin, 2013b: Ch. 6):

I was in Urbino for two or three summers in the early seventies, late sixties. That was
when I first interacted with semiotics in the continental sense. It seemed to me that the
general concept embodied in semiotics was a very valuable one because it enables me
to say: “Here is a context within which to study language”. Partly it’s simply saying:
“OK. We can look at language as one among a number of semiotic systems.”

Halliday’s conception of social semiotics as a theme in the study of higher-order systems became
very influential, and was taken up and developed in publications by a growing number of schol-
ars in the 1980s: see Figure 2.5 and Figure 2.6. Towards the end of the decade, Hodge and Kress
(1988) published a book titled Social Semiotics, which they noted continued a strand of work they
had been involve in since the 1970s and which had led to the publication of Language as Ideology
(Kress and Hodge, 1979). In their acknowledgements, they write:

Of the many writers who have influenced us who are mentioned in the references,
we would like to single out M.A.K. Halliday, not only for his texts such as Language as
Social Semiotic, but for his inspiring example as a researcher, teacher and explorer of the
social functions of language.

In his own contributions to social semiotics, Halliday ranged quite widely, from context to
lexicogrammar. One crucial contribution was his interpretation of context as a system of mean-
ing, as a semiotic system, set forth in Halliday (1978); this is clearly a considerable advance over
a conceptualization of context only in social terms (cf. Argyle, Furnham and Graham, 1981).
But his contributions to the study of the role lexicogrammar in language as a social semiotic are
also noteworthy, also in view of the fact that as social semiotics developed partly alongside SFL,
scholars tended to focus on areas other than lexicogrammar, often without the grammarian’s
commitment to explicit descriptions. Examples of Halliday’s contributions include: Halliday
(1991d), Halliday (1992b), Halliday (1998b) and Halliday (2001b/2006).
Andersen et al. (2015) provide an overview of the development of social semiotics through
interviews with scholars they regard as key contributors (Christian Matthiessen, Jim Martin,
Gunther Kress and Theo van Leeuwen) and a discussion of “central themes”. For biographi-
cal information about Gunther Kress and his many contributions to the development of social
semiotics, see Li (2020); for Theo van Leeuwen’s varied contributions (e.g. van Leeuwen, 1996),
see the recent Festschrift, Zhao et al. (2018).

2.6.1.3 Critical Linguistics, CDA

As shown in Figure 2.5, the development of Critical Linguistics (CL) drew on SFL in the 1970s,
and CL in turn also contributed to the emergence of Critical Discourse Analysis (CDA) in the
1980s. Martin and Rose (2007: 314) emphasize the connections between SFL and CDA:

Critical discourse analysis (hereafter CDA) has always had close links with SFL reach-
ing back to its foundation in the work of Fowler et al. (e.g. 1979) on critical linguistics
at East Anglia in the 1970s. Halliday’s conception of linguistics as an ideologically com-
mitted form of social action has been one important factor in this dialogue; and SFL’s

62
 Reading guides

Figure 2.5 Intellectual strands in the development of social semiotics (see Matthiessen, 2017)

63
 Reading guides
64

Figure 2.6 Social semiotics and neighbouring fields of activity, including multimodality (multimodal studies), Critical Linguistics and Critical Discourse Analysis
 Reading guides

relatively rich semantic orientation to text in context across languages and modalities
has meant that CDA has regularly visited the theory in search of tools for analysis
where close systematic readings of texts are required. Martin (2000) reviews various
connections from the perspective of SFL; and Chouliariki [Chouliaraki, CMIMM and
KT] and Fairclough (1999) look back from CDA; Young and Harrison (2004) bring
together work across the two traditions.

The critical stance manifested the development of Critical Linguistics in the 1970s by the
“East Anglia group” (e.g. Fowler et al., 1979; Kress and Hodge, 1979) and later of Critical
Discourse Analysis (e.g. Fairclough, 1989, 2003, 2010) can be related to the commitment to
social accountability (e.g. Halliday, 1975c, 1984d; Matthiessen, 2012) that has been part of
the development of SFL as an appliable linguistics from the start, as evidenced by applications in
education already in the 1960s. Here it is worth pausing to consider what Halliday (1975c) said
about the need for social accountability in linguistics in the mid-1970s, since this reflected the
background in linguistics against which subsequent critical movements arose; commenting on
the stance in “mainstream” theoretical linguistics as it had emerged in the 1960s against applica-
tions in areas such as language teaching, he writes (1975c/2003a: 75):

In matters such as these, theoretical linguists refused to admit the social accountability
of their subject, and withdrew their expertise from activities that could have been ben-
eficial to large numbers of people. I am not saying that all linguists held aloof. Nor am
I implying, in an excess of linguistic paternalism, that miraculous solutions lay around
the corner.

Later in the same paper, he continues (1975c/2003a: 89):

In short: linguistics has not yet faced up to the question of its social accountability.
Social accountability is a complex notion which cannot be taken in from one angle
alone. It is not defined as satisfying some abstract or symbolic entity such as a board
of trustees, the business community or the taxpayer. Nor is it the same thing as sat-
isfying our own individual consciences, which is a purely private luxury. There is an
ideological component to it, which consists at least in part in eliminating some of the
artificial disciplinary boundaries that we have inherited and continued to strengthen.
At this level, many of the arguments lead back to this same point: that there are strong
boundaries between academic disciplines, which hamper intellectual development,
and induce both overspecialization and underapplication.

Significantly, he relates social accountability to the value of application (as opposed to “underap-
plication”) and the need to overcome disciplinary boundaries (“overspecialization”) – later rec-
ognized terminologically as aspects of his notion of appliable linguistics (e.g. Halliday, 2002b,
2008a). Halliday has on many occasions emphasized that SFL embodies both theory and action;
it is not only reflection but also praxis, or a mode of action, manifested in positive interventions
in various applications (cf. Kilpert, 2003), e.g.:

[Halliday, 1985c/2003b: 197]: What is perhaps a unifying factor among these who
work within this framework is a strong sense of the social accountability of linguistics
and of linguists. Systemic theory is designed not so much to prove things as to do
things. It is a form of praxis. I have often emphasized that language, both in its nature

65
 Reading guides

and in its ontogenetic development, clearly reveals a dual function; it is at once, and
inseparably, a means of action and a means of reflection. Linguistics, as metalanguage,
has to serve the same twofold purpose. Systemic theory is explicitly constructed both
for thinking with and for acting with.
[Halliday, 2001a/2003b: 272–274] linguistics was (like language itself) a mode of
action, a way of intervening in social and political processes; and this has remained as a
significant motif of work in systemic functional linguistics. . . .
Linguistics who engage in professional activities of this kind are taking an interven-
tionist view of their science, one where the scientist is an active participant in the social
process and defines the concept of “doing science” in these inclusive terms.

As appliable linguistics, SFL thus embodies both theory and application; and both have been
informed by social theorists coming from anthropology and sociology, including Bronislaw
Malinowski, Basil Bernstein, Mary Douglas and Karl Marx. There is a long tradition of studies
in SFL addressing social accountability by undertaking critical studies of both system and text
concerned with shedding light on ideology, both in a field-based sense of worldview and a
tenor-based sense of power and solidarity; for example: Halliday (1971) on the power of tran-
sitivity to construe diferent world views as deployed by William Golding; Halliday (1976b) on
the diferent models of teaching construed by the grammar of English; Hasan (1986) on the
ontogenesis of ideology; Martin (1986) on text-based analysis of ideology; Halliday (1990b) on
the perils of the commonsense world view embodied in everyday grammar; Matthiessen (1993c)
on the folk model of the mind taken over into cognitive science; Martin (2002) on the discourse
of reconciliation as a positive model; Butt, Lukin and Matthiessen (2004) on the role of gram-
mar in the incitement to war; Veloso (2006) on both the reflection and construction of ideology
in super hero comic books after 9–11 – and further Veloso (2012), Veloso and Bateman (2013);
Martin et al. (2013) on restorative justice; Matthiessen (2015f) on the subliminal construal of a
patriarchal world view in Noah’s Ark; Lukin (2019) on ideology and discourses of war (see also
Lukin, 2017). These and other comparable studies use diferent aspects of the “full machinery”
of SFL. There is also a “critical turn” in multimodal studies, e.g. Djonov and Zhao (2013),
O’Halloran et al. (2016; O’Halloran, Tan and Marissa, 2015).
Drawing on SFL as it had developed by the mid-1970s or so, Roger Fowler, Gunther Kress
and other scholars associated with the “East Anglia” group developed a strand that came to be
down as “Critical Linguistics”, and produced studies that were informed by a critical stance,
relating linguistic selections in texts to control and ideology: Fowler et al. (1979); Kress and
Hodge (1979); Fowler (1991).14 For example, Trew (1979) shows the ideological loading in
news reports through transitivity analysis, e.g. suppression of agency involving the police in
killings; this theme of systematic ideological bias in the news was later taken up by Lukin (e.g.
2019). Fowler’s (1995) study of “the language of George Orwell” is also an important contribu-
tion in this context, since Orwell was a “critical” writer, both of non-fiction and of fiction, and
Fowler uses SFL in his study.
A decade after Fowler et al.’s (1979) demonstration of the value of Critical Linguistics, another
major “critical” contribution appeared – Norman Fairclough’s (1989) book Language and Power,
which is treated as the foundation of Critical Discourse Analysis. At this point, he characterized
the approach as “Critical Language Studies” (p. 5):

The approach to language which will be adopted here will be called critical language
study, or CLS for short. Critical is used in the special sense of aiming to show up
connections which may be hidden from people – such as the connections between

66
 Reading guides

language, power and ideology referred to above. CLS analyses social interactions in a
way which focuses upon their linguistic elements, and which sets out to show up their
generally hidden determinants in the system of social relationships; as well as hidden
effects they have upon that system.

This is very resonant with one of the many aims of SFL at the time in general (as in Hasan,
1986; Martin, 1986) and with Critical Linguistics in particular. At the beginning of the book,
Fairclough states that he had “written [the book] for two main purposes” (p. 1):

The first is more theoretical: to help correct the widespread underestimation of the
significance of language in the production, maintenance, and change of social relations
of power. The second is more practical: to help increase consciousness of how language
contributes to the domination of some people by others, because consciousness is the
first step towards emancipation.

This is again very much aligned with the conception of SFL noted above as a resource for both
reflection and action. One diference with respect to action would arguably be that systemic
functional linguists, working with professionals in diferent institutions, would extend the prac-
tical activity from consciousness raising to positive interventions, as in educational linguistics: the
work by J.R. Martin and other members of the so-called Sydney School set out an educational
programme from the start that included the aim of giving students access to genres of power,
both in the sense of being able to analyse genres used for the purpose of control and in the
sense of being able to master the production of powerful genres (e.g. Rose and Martin, 2012;
Drury and Jones, 2010; Dreyfus et al., 2016; Humphrey et al., 2010). The same applies to other
systemic functional linguists engaging with institutions of education. Importantly, educational
eforts have involved empowering students with the resources of grammatics, as in the work by
Geof Williams and Ruth French (Williams, 2005b; Williams and French, 2016), the approaches
adopted with bilingual learners in relation to social equity, as in Harman (2018), and with
the tools of multimodal analysis to help them engage in “critical thinking”, as in O’Halloran,
Marissa and Sabine Tan (2015).
Fairclough has presented his conception of discourse in various publications over the years,
and of the methods to be used in a critical approach to the analysis; let us quote from one of his
influential, oft-cited books, Fairclough (1992: 97–98):

The approach I have adopted is based upon a three-dimensional conception of dis-

course, and correspondingly a three-dimensional method of discourse analysis. Dis-
course, and any specific instance of discursive practice, is seen as simultaneously (i) a
language text, spoken or written, (ii) discourse practice (text production and text
interpretation), (iii) sociocultural practice. Furthermore, a piece of discourse is embed-
ded within sociocultural practice at a number of levels; in the immediate situation, in
the wider institution or organization, and at a societal level; for example, one can read
an interaction between marital partners in terms of their particular relationship, rela-
tionships between partners within the family as an institution, or gender relationships
in the larger society. The method of discourse analysis includes linguistic description of
the language text, interpretation of the relationship between the (productive and inter-
pretative) discursive processes and the text, and explanation of the relationship between
the discursive processes and the social processes. A special feature of the approach is
that the link between sociocultural practice and text is mediated by discourse practice;

67
 Reading guides

how a text is produced or interpreted, in the sense of what discursive practices and
conventions are drawn from what order(s) of discourse and how they are articulated
together, depends upon the nature of the sociocultural practice which the discourse is a
part of (including the relationship to existing hegemonies); the nature of the discourse
practice of text production shapes the text, and leaves ‘traces’ in surface features of the
text; and the nature of the discourse practice of text interpretation determines how
the surface features of a text will be interpreted. On page 98 there is a diagrammatic
representation of this approach [Reproduced here as Figure 2.7].

Fairclough’s “three-dimensional conception of discourse” can be interpreted as a selection of

some aspects of the systemic functional architecture of language in context within which text (or
discourse) can be located: see Figure 2.8. Naturally, the mapping proposed in our representation
can be discussed, and other links explored, but it is quite clear that whereas Fairclough operates
with one dimension in the systemic functional sense of the semiotic dimensions of the archi-
tecture of language in context, SFL operates with three: (i) the cline of instantiation extended
between potential and instance, (ii) the hierarchy of stratification of language in context, and (iii)
the ordered typology of systems operating in diferent phenomenal realms. Thus in the analysis
of any given text, the analyst can – and should – shunt along all three dimensions. At the same
time, of course, in an SFL engagement with text, the spectrum of functional diversification
needs to be evoked and considered: the three parameters of context – field, tenor and mode,
and the metafunctions – ideational (experiential, logical), interpersonal and textual. While the
concerns of CDA are located within tenor since they have to do with power, control, domi-
nance, discrimination and inequality, they relate in interesting ways also to field and mode, both
of which are likely to be relevant to the analysis.

Figure 2.7 Fairclough’s (2010: 133) “diagrammatic representation” of “three-dimensional conception of

discourse” and “three-dimensional method of discourse analysis”

68
 Reading guides

Figure 2.8 Possible mapping of Fairclough’s “three-dimensional conception of discourse” onto part of
the systemic functional theory of language in context extended along the cline of instantiation
and ordered internally in terms of the hierarchy of stratification and externally in terms of the
ordered typology of systems operating in different phenomenal realms

In Language and Power, Fairclough (1989: 110–111) lists “ten main questions (and some sub-
questions) which can be asked of a text” – notably “of a text” rather than of “discursive prac-
tices” or “social practices”, which are involved in separate interpretation and explanation (see
Figure 2.7); these questions have been formulated to be accessible to readers with an interest in
language and power but with no specialist background – they are not designed with linguists in
mind. The main questions are as follows:

A Vocabulary
1 What experiential values do words have?
2 What relational values do words have?

69
 Reading guides

3 What expressive values do words have?

4 What metaphors are used?
B Grammar
5 What experiential values do grammatical features have?
6 What relational values do grammatical features have?
7 What expressive values to grammatical features have?
8 How are (simple) sentences linked together?
C Textual structures
9 What interactional conventions are used?
10 What larger-scale structures does the text have?

They can all be interpreted systemic-functionally, and located within a matrix: Table 2.3. This
matrix is organized in terms of the spectrum of functional diversification and the hierarchy of
stratification – two of the global semiotic dimensions in the systemic functional “architecture” of
language in context. This makes it possible to see where Fairclough’s ten questions are located;
they are concerned mainly with lexicogrammar – eight of the ten questions, with only two
about semantics, and none (in this list) about context and none about the expression plane, i.e.
phonology or graphology. However, some of the lexicogrammatical questions have implications
for context. Seven of the questions are concerned with the experiential mode of the ideational
metafunction and the interpersonal metafunction; there are no questions about textual patterns
and only one about the logical mode of the ideational metafunction. It is also worth noting that
within lexicogrammar, ranks are not diferentiated, as can be seen when we compare it with
a full-fledged function-rank matrix such as the ones reproduced as Table 3.7 (from Halliday,
1970a) and Table 3.8 (from Halliday and Matthiessen, 2014: 87).
Halliday (2001a/2003b: 279) summarizes the developments we have sketched so far as
follows:

In the 1970s Roger Fowler introduced the concept of “critical linguistics”, using
grammatics to reveal the political or other ideological constructs that were latent in
the wording of a text; and this has been followed up in an impressive body of work in
“critical discourse analysis” (Fairclough, 1992). (Not all comparable efforts at critical
analysis are equally successful or convincing; some writers barely engage with the text
at all, and give little linguistic evidence for the interpretation that is being proposed.)

Fairclough has continued to be very productive, addressing a variety of serious problems through
critical analysis; Language and Power has now appeared in a third edition, and he has authored
and co-authored many articles and books. As the collection of contributions included in Young
and Harrison (2004) shows, the interaction between CDA and SFL has been productive (cf. also
Bloor and Bloor, 2007).
Meanwhile, while Fairclough was very explicit about the relationship to SFL, other strands
have been added to CDA as a broad programme. Here a meeting in Amsterdam in 1991 was
important in bringing together a group of scholars with related concerns, described by Wodak
and Meyer (2009) as follows:

The CDA as a network of scholars emerged in the early 1990s, following a small
symposium in Amsterdam, in January 1991. Through the support of the University of

70
 Table 2.3 Systemic functional interpretation in terms of a function-stratification matrix of Fairclough’s (1989: 110–111) ten questions to ask of a text

Stratum Ideational: logical Ideational: experiential Interpersonal Textual Other

context [field] [tenor] [mode]

semantics 9. What interactional 10. What larger-scale structures
conventions are used? does the text have?

Reading guides
lexicogrammar grammar 8. How are (simple) 5. What experiential 6. What relational values do
71

sentences linked values do grammatical features have?

together? grammatical 7. What expressive values to
features have? grammatical features have?
lexis 1. What experiential 2. What relational values do
values do words words have?
have? 3. What expressive values do
words have?
4. What metaphors are used?
phonology/graphology
 Reading guides

Amsterdam, Teun van Dijk, Norman Fairclough, Gunther Kress, Theo van Leeuwen
and Ruth Wodak spent two days together, and had the wonderful opportunity to
discuss theories and methods of Discourse Analysis, specifically CDA. The meeting
made it possible to confront with each other the very distinct and different approaches,
which have, of course, changed significantly since 1991 but remain relevant, in many
respects.

Gunther Kress, who was part of the East Anglia Group, and Theo van Leeuwen have, of course,
a long history of engagement with and contributions to SFL, to social semiotics and have pio-
neered the framework for “reading images” (Kress and van Leeuwen, 1996, 2006; cf. Taylor,
2017). Adding to the published resources on CDA, van Leeuwen (2008) has compiled his work
on CDA as a book; drawing on both Foucault and Halliday and also Bernstein’s concept on
reconceptualization, he consolidates a decade and a half of work in the area. He covers a range
of topics, including “visual racism”, and applies his own framework of social action and social
actors.
Coming from traditions other than SFL, van Dijk and Wodak have added other concerns
and working concepts to CDA: van Dijk has argued for a cognitive component,15 and Wodak
has added a historical dimension, with further emphasis on empirical evidence for critical inter-
pretations. This emphasis has been manifested in the adoption of corpus linguistic tools and
techniques, a seminal study being Baker et al. (2008) (see also www.monabaker.org) – cf. also
Caldas-Coulthard and Moon (2010), an expanded version of a contribution from 1999. Work
by, and inspired by, Wodak has been very productive (e.g. Wodak and Meyer, 2009; Weiss and
Wodak, 2003) and is resonant with SFL (as illustrated by Martin and Wodak, 2003), also in the
use of corpus evidence. Wodak (2015) turns much needed critical attention to the current status
of universities.
As noted earlier, SFL has been “primed” from the start to engage in research and applications
related to social accountability, broadly speaking, including critical ones. The critical approach
has been supplemented within SFL by J.R. Martin in the shape of PDA, Positive Discourse
Analysis. Drawing initially on the discourse of reconciliation from South Africa, Martin (e.g.
2004a) has worked on establishing what can serve as positive models in communities embodying
solidarity. This brings out a certain duality in tenor relations, foreshadowed in Brown and Gil-
man’s (1960) classic study of interactant pronouns in terms of power and solidarity.
Critical studies – Critical Linguistics, Critical Discourse Analysis and Critical Discourse
Studies – can be related to the notion of social accountability, as can PDA. In turn, social
accountability is an aspect of appliable linguistics – part of its value system or axiology. Appliable
linguistics covers different linguistic activities, i.e. not only text or discourse analysis, but also
language description, language comparison and typology, theory and of course all the myriad of
applications drawing on appliable linguistics as a resource. However, since many research activi-
ties and applications depend centrally on discourse analysis, it is helpful to articulate the nature
of appliable linguistics in the domain of discourse analysis. One initial approximation is Mat-
thiessen’s (2014a) chapter introducing and characterizing Appliable Discourse Analysis, or ADA.
If we extend appliable linguistics along the cline of instantiation, we can locate the differ-
ent phases of description and analysis, from Appliable System Description (ASD) to Appliable
Discourse Analysis (ADA): Figure 2.9. Critical stances constitute one of many possible orienta-
tions within appliable linguistics, if we understand “critical” in the sense of being concerned
with (exposing) power, control, domination and other unequal relationships within the tenor
parameter of context. Importantly, just like appliable linguistics in general, such critical stances
are not, or should not, be confined to discourse analysis. There are many circumstances where

72
 Reading guides

Figure 2.9 Phases of appliable linguistics along the cline of instantiation

they need to be applied to regions higher up the cline of instantiation. Thus Halliday’s (1990b)
paper “New ways of meaning: a challenge to applied linguistics” involves a critical reading of
lexicogrammatical systems at the potential pole of the cline of instantiation, and it has helped
energize the development of ecolinguistics (Webster, 2008).16

2.6.2 Fellow travellers in linguistics at a distance

2.6.2.1 Range and groupings of fellow travellers

At the pole of low degree of interaction, we can identify developments often initiated as reac-
tions against Chomsky’s generative linguistics in the US after the 1960s or even after the 1970s,
i.e. well after the foundation of SFL. These include:

• Functionalism: functionalism has been an important motif in 20th and 21st century lin-
guistics, and it goes back much further in the history of linguistics, of course (see e.g.
Halliday, 1977a; Seuren, 1998; and for a comparison of their two accounts, Matthiessen,
forthcoming a). Functional currents whose participants have engaged in extended dialogue
with SFL are mentioned earlier; here we can note continental European functionalism
other than the Prague School, West-Coast Functionalism and Japanese Functionalism as
shaped by Okuda:17
• French and Dutch Functionalism (for the Prague School, see earlier): Caffarel (2010)
includes references to both French and Dutch Functionalism, and Fries (2001) discusses
SFL as a possible close relative to French Functionalism (in the André Martinet tradi-
tion). Halliday refers to the work by Claude Hagège in various publications, and Hagège
was invited as a plenary speaker to ISFC in Ghent in 1994. François (2018) presents his
interpretation of certain aspects of SFL in relation to functional(ist) theories of language.

73
 Reading guides

• West-Coast Functionalism: There are many introductions to West-Coast Functional-

ism,18 which is an open and variable linguistic territory, going back to the 1970s. There are
some publications bringing this type of functionalism together with ideas from SFL, includ-
ing Thompson (1985, 1990), Matthiessen and Thompson (1988), Matthiessen (2002a).
There are also studies presenting findings that are more in tune with Halliday’s description
than more traditional views, e.g. the interpretation of “complement clauses” as “objects”;
Thompson (2002) is closer to Halliday’s projection analysis than to more mainstream
accounts.
• Japanese Functionalism: There are various Japanese Functionalism works since the
1960s. They are included in Teruya (1998, 2007). Japanese Functionalism has emerged
from the Linguistics Research Group (Gengogaku Kenkyukai) led by Yasuo Okuda
(Okuda, 1985) since the late 1950s. The coverage of the domain of their linguis-
tic enquiry extends from lexicology (Miyajima, 1994), morphology (Suzuki, 1972),
grammar (Gengogaku Kenkyukai (ed.), 1983, Takahashi, 2005) to language teaching
(Okuda and Kokubun’ichi (eds.), 1974). There are some similarities between Japanese
Functionalism and SFL, for example, the notion of clause (bun) and Okuda’s notion
of collocation (rengo). The former constitutes the highest ranking unit of the lexico-
grammatical units in both theoretical and descriptive traditions, and the latter is com-
parable to the ‘effective’ experiential clause structure in SFL.
• Cognitive linguistics: Like functionalism or functional linguistics, cognitive linguistics
covers a number of different strands, including the different manifestations characteristic of
the US East Coast – Ray Jackendoff, and the West Coast – George Lakoff at UC Berkeley
and Ron Langacker at UC San Diego. Kristin Davidse and her group have drawn on both
SFL and Langacker’s Cognitive Grammar, bringing out interesting connections, compat-
ibilities and complementarities in the process, an early example being Davidse (1996b) and
a more recent example being Vandelanotte (2009) – but there are many more contribu-
tions. In Lakoff ’s variety of cognitive linguistics (e.g. Lakoff, 1987), a central phenomenon
of study has been metaphor (Lakoff and Johnson, 1980, and many publications since then).
It is called “conceptual metaphor”, but it is really primarily experiential lexical metaphor –
central properties of which had already been identified by Whorf (1956) in the 1940s. The
complementarity between “conceptual metaphor” and systemic functional work on (ide-
ational) grammatical metaphor is explored by Halliday and Matthiessen (2006). Gonzálvez-
García and Butler (2006, 2014) map out an extended functional-cognitive space, analysing
contributions in terms of a number of parameters (see also later in this section), and Butler
(2019) has contributed a chapter drawing on this extensive metalinguistic research project
to the Cambridge SFL encyclopaedia (Thompson et al., 2019).
• Construction Grammar: Like Lakoffian cognitive linguistics, Construction Grammar
originated at UC Berkeley, the leading scholar being Charles Fillmore. He captured
insights from his earlier work on case grammar (Fillmore, 1968, 1977), and then worked
with colleagues to develop Construction Grammar and the closely related resource
FrameNet e.g. Goldberg (1995, 2006); Hoffman and Trousdale (2013); Boas and Sag
(2010). By some degrees of separation, there is a British analogue to Construction Gram-
mar, deriving from the Birmingham tradition; this is Hunston and Francis’s (2000) (lexi-
cal) pattern grammar. From a systemic functional point of view, Construction Grammar
and pattern grammar have some key similarities. One is that they are syntagmatically
oriented; they start with syntagmatic constructions or patterns. Another is that they focus
on the mid-region along the cline of delicacy between grammar and lexis – a region that

74
 Reading guides

can be approached productively from either pole of the cline of delicacy. This is also the
region where many constructions studied in Construction Grammar can be located (cf.
Figure 2.12).

There are, naturally, other fellow travellers. Here we can just mention the growing work in
“probabilistic linguistics”, in a sense picking up a thread of interest from the 1950s maintained
by Halliday (1959) and a number of other SFL studies since then (e.g. Halliday, 1991b, 1991c,
1993c, and, for an overview, Matthiessen, 2015a) – with a number of contributions brought
together almost two decades ago by Bod, Hay and Jannedy (2003). Research in this area has
been fuelled by the increase in the size of corpora and advanced in corpus linguistics (see later)
and, importantly, the development of statistical NLP (a foundational text being Manning and
Schütze, 1999). Also worth noting are the strands outside SFL since the 1990s relating the
conception of language to general theories of complex adaptive systems, including the focus
on emergent complexity (cf. Beckner et al., 2009), e.g. Steels (1998) – coming from compu-
tational considerations and with contributions to Fluid Construction Grammar (e.g. Steels and
Beuls, 2019), and Larsen-Freeman and Cameron (2008) – as a contribution to issues in applied
linguistics.
A number of these approaches are committed to empirical evidence from authentic texts and are
often brought together under the heading of “usage-based theories”. These include those functional
and cognitive approaches basing descriptions on texts, and on systematic samples of texts in the form
of corpora. Studies of particular areas framed in terms of SFL are very likely to refer to and drawn on
contributions from other functional frameworks and also cognitive ones, a representative example
being Degand’s (1996) account of causation in Dutch and French and also Fine (1988a).
Functional and cognitive approaches in general have been compared and contrasted system-
atically by Gonzálvez-García and Butler (2006, 2014). Their 2014 book is a massive undertak-
ing, comparing and contrasting 16 models, based on a questionnaire with 58 items or features
grouped into six major groups (pp. 148–153). The questionnaires were sent out to respondents
for each model, and also to proponents of the Minimalist Program (Chapter 3). The responses
form the primary data for their analysis or map of the “functional-cognitive” space, informed
by three sophisticated statistical techniques (Chapter 4). Drawing on their primary data, they
devote six chapters to the characterization and comparison of the 16 models, with ample refer-
ences to the relevant literature; for each group of questions, one can locate SFL in relation to
the other models they survey.

2.6.2.2 West-Coast Functionalism

West-Coast Functionalism can serve as a cover term for work in the US to provide functional
accounts and explanations as alternatives to the Chomskyan tradition that had become increas-
ingly dominant in the 1960 and the 1970s. A number of the pioneers in this kind of functional-
ism had done their PhDs in the Chomskyan climate of the 1960s and began to react against it.
They began to develop a functional approach in the second half of the 1970s, with important
contributions from the first phase including Givón (1979a, 1979b, 1982, 1983), Hopper and
Thompson (1980, 1982), Hopper (1982). Some of them referred to the Prague School and to
Tagmemic Linguistics; references to SFL were less common.
These studies set out to study and explain grammar by referring to patterns in discourse; they
were concerned with grammar and discourse mainly for the sake of grammar. Early examples
include the examination of tense and aspect systems typically based on narrative discourse (e.g.
Hopper, 1982). Later, conversation became an important focus, thanks largely to interaction

75
 Reading guides

between Sandy Thompson and a number of her PhD students in linguistics and Emanuel Sche-
gloff and the Conversation Analysis framework in sociology, one important publication being
Ochs, Schegloff and Thompson (1996). This strand of research has combined with an interest
in the emergence of grammar in discourse, as in Bybee and Hopper (2001), adding to the work
on grammaticalization (e.g. Hopper and Traugott, 1993).

2.6.2.3 Corpus linguistics

“Corpus linguistics” tends to be used as a term in somewhat different ways, ranging from a sense
of a distinct school within linguistics with its own paradigm (or perhaps paradigms, since there
is a great deal of variation in terms of (meta-)theoretical assumptions, even between, say, the
Lancaster and Birmingham varieties, e.g. their positions on corpus annotation), with dedicated
journals and conferences, to a sense of a toolkit used and developed within a variety of schools
in linguistics, including SFL. As a toolkit, corpus linguistics is closely related to work within
NLP (Natural Language Processing), which tends to have more of a computational base and
more powerful forms of representation and analysis (including machine learning techniques). In
SFL, corpus linguistics has arguably been a toolkit from the start – thus not a separate branch or
school; so given this situation, it does not make sense to say that research is based on SFL and
corpus linguistics since corpus linguistics is simply part of the methodology used in SFL from
that start together with e.g. manual text analysis.
The use of text as a basis for descriptions of linguistic systems has a long history; many descrip-
tive grammars included collections of texts from which examples were drawn. Early systematic
efforts to compile corpora are usually taken to be The Survey of English Usage, led by Randolph
Quirk, and The Brown Corpus, initiated by Freeman Twaddell. Halliday (1994a/2003b: 440) cites
Randolph Quirk and his group (including Jan Svartvik), and “more recently” Sinclair, as influ-
ences with respect to “the theory and practice of corpus linguistics”. But there were other ear-
lier precursors; Charles C. Fries had, for example, recorded phone conversations and collected
informal letters (e.g. Fries, 1985, and other contributions to Fries and Fries, 1985). McEnery
and Hardie (2013) provide a general history of corpus linguistics, and Wu (2009) and Sharoff
(2017) discuss the use of corpora in SFL.
When Halliday undertook research in the 1950s that would feed into the development of SFL,
there was no separate field of “corpus linguistics”, but there were early efforts that used texts as
data in the way that has become common in corpus linguistics, e.g. using corpora as a source of
quantitative information, as in Ellegård’s (1953) pioneering study of changes in the use of do as an
auxiliary in a period of around 250 years in different grammatical environments defined by mood
type and polarity. Halliday’s PhD thesis from the mid-1950s, published as Halliday (1959), treated
the Mandarin text of The Secret History of the Mongols as a corpus, and provided a number of counts
of grammatical features. His work on the description of the English system of intonation began in
the late 1950s and was based on a small corpus of spoken English (Halliday, 1963a, 1963b, 1967).
In the 1960s, systemic functional research continued to be based on corpora, for example
Huddleston et al.’s (1968) study of scientific English, El-Menoufy’s (1969) description of Eng-
lish intonation, and descriptions of the grammars of a number of languages other than English
(Barnwell, 1969; Mock, 1969). The explicitly corpus-based studies proved impossible to get
published because publishers were sensitive to Chomsky’s critique of such studies (cf. McEnery
and Hardie, 2013), so unfortunately they have often been overlooked even after corpus-based
work became acceptable to publishers.19 Halliday’s (1975a) investigation of early child language
development, begun in 1969, was also corpus-based, as were subsequent studies with the same
focus (including Painter, 1984; Torr, 1997).

76
 Reading guides

Since these relatively early uses of corpora in SFL, corpora have continued to play an impor-
tant role, one that has grown in significance with the growth of corpora in size and the increas-
ing sophistication of corpus tools. In terms of research methods, corpus-based research has
proceeded alongside “discourse” analysis and “ethnographic methods”. Systemic functional
researchers have used corpora and tools used by the community of linguists involved in corpus
research in general (e.g. small corpora in the Brown and LOB tradition, including ICE cor-
pora, and the London-Lund corpus of spoken British English, the Penn Treebank, and larger
corpora, including the Bank of English, the British National Corpus, COCA and tools such as
Word Smith, Sketch Engine, W-Matrix, ELAN), and they have also compiled their own cor-
pora (including the ones mentioned earlier, the Cardiff child language corpus, the mother-child
corpus compiled by Hasan and her team, the UTS-Macquarie OZ Talk Corpus, the EDCOM
corpus, Halliday’s and Banks’ corpora of scientific English) and developed their own tools,
notably Mick O’Donnell’s (2012) UAM Corpus tool20 and Wu’s (2000, 2009) SysConc and,
for multimodal analysis, O’Halloran’s Multimodal Analysis tool.21 These tools are, naturally, also
possible to use in research that is not specifically based on SFL.
Corpora have been compiled and investigated for a wide range of purposes in SFL, and a
number of examples are listed in Table 2.4. Some of the studies listed in the table have involved

Table 2.4 Examples of use of corpora in SFL

Area of use Qual/quant Nature of corpus References

language description basically qual Barnwell (1969), Mock (1969)

basically qual corpus of selection of Kumar (2009), Mwinlaaru
registers (2016)
quant corpus of ethnographic Nesbitt and Plum (1988)
interviews with dog
breeders
quant part of Bank of English Halliday and James (1993)
multilingual Sharoff (2006)
quantitative: text archive of range of Matthiessen (2006)
systemic profiles registers
translation studies quant textbook used for Ghadessy and Gao (2001)
teaching translation
quant selection of registers from Teich (2003); Hansen-Schirra,
different languages: the Neumann and Steiner
CroCo corpus (2012)
qual and quant verbal art: selections of Wang (2017); Zhang (2020a);
poetry, drama, novels Wang and Ma (2020);
Ma and Wang (2021);
Matthiessen, Arús and
Teruya (2021)
language qual longitudinal: mother Halliday (1975a, 1984c); Painter
development tongue (1984, 1999); Torr (1997);
Torr and Simpson (2003)
qual and quant longitudinal: ESL Xuan (2014) [student writing]
(Continued )

77
 Reading guides

Table 2.4 (Continued)

Area of use Qual/quant Nature of corpus References

clinical linguistics narrative therapy: “six con- Muntigl (2004b)

joint sessions involving
a narrative therapist
with one couple”
healthcare qual consultations in Slade et al. (2015)
communication emergency
departments – the
EDCOM corpus
qual and quant multimodal corpus of Zhang Peija (2018)
public health posters in
NY and HK
register studies qual and quant verbal art: TV drama Law (2017)
qual and quant news, BNC Lukin (2019)

the compilation of new corpora, while other investigations have been based on existing corpora
accessible to the research community at large.

2.6.2.4 Beyond fellow travellers

Sometimes or even often, the term “mainstream linguistics” is used, although it can be prob-
lematized. To the extent that there is such a current in linguistics, neither SFL nor its fellow
travellers represent the “mainstream”. Arguably, Chomskyan linguistics (or however we choose
to call it) does constitute the mainstream, or at least it has up until fairly recently – if we measure
mainstream in terms of dominance and institutional power. In relation to the Chomskyan “min-
imalist programme”, Asp (2017) comments on some aspects of this manifestation of Chomskyan
programme with respect to “systems and functions”.

2.6.3 In relation to stratal subsystems

In the preceding subsections, we have considered SFL in relation to other traditions taken as a
whole. Here we will consider briefly particular stratal subsystems of language. The degree to
which there have been dialogues, one-way influences and cross-fertilization varies consider-
ably from one stratal subsystem to another. To oversimplify, before we get into more detail: the
greatest interaction has taken place within lexicogrammar, and the least within semantics; and in
the case of phonology, the prosodic analysis part of the foundation of systemic phonology was
acknowledged to some degree in autosegmental phonology once generative phonologists had
“rediscovered” prosodic analysis (cf. Henderson, 1987).
The variation in interaction across the strata is partly related to where work outside SFL
was done in disciplinary terms: in work on lexicogrammar, computational modelling turned
out to be a meeting ground; in work on semantics, the grounding of a good deal of semantics
outside SFL in logic and philosophy (formal, model-theoretic semantics) increased the distance
considerably (cf. Halliday, 1977a); but in contrast, work in phonology tended to be a linguistics-
internal undertaking.
78
 Reading guides

When we track developments in different linguistic traditions throughout the 20th cen-
tury and into the 21st century, there are many synchronicities. For example, phonologi-
cal studies tended to blossom before grammatical studies, and these in turn came before
semantic studies and explorations of text of discourse. But there were always precursors and
outliers, of course, including Firth’s “proto-pragmatics” (cf. Levinson, 1983: xii) and Har-
ris’s (e.g. 1952) interest in discourse analysis.
Before discussing the stratal subsystems of language in relation to approaches other than SFL,
we can highlight key differences: see Table 2.5. We will then start with phonology, moving
upwards via lexicogrammar to semantics.

2.6.3.1 Phonology
Phonology was, of course, a focus in linguistics on both sides of the Atlantic from the 1930s
onwards. The US tradition was essentially concerned with articulatory phonology and was based
on the phoneme (e.g. Swadesh, 1934; hence the term “phonemics”), with Kenneth Pike’s work
providing an important complement outside the Bloomfieldian–Post-Bloomfieldian tradition
(e.g. Pike, 1948). The continental European tradition included centrally the work on phonology
within the Prague School, which was also largely oriented towards the phoneme, but it fore-
grounded the paradigmatic axis of organization rather than the syntagmatic one, as is brought
out in Nikolai Trubetzkoy’s (1939) posthumous foundational study. His compatriot Roman
Jakobson continued the Prague School work on phonology, but in the late 1940s he transformed
paradigmatic parameter values into syntagmatic “components” of phonemes (Jakobson, 1949).
As he moved to the US East Coast, fleeing Nazi-occupied Europe, he added this version of
Prague School phonology to the Bloomfieldian tradition (e.g. Jakobson et al., 1951), and it was
this version that was incorporated into generative phonology, as in Chomsky and Halle (1968).

Table 2.5 Systemic functional accounts of semantics, lexicogrammar and phonology in comparison with
other approaches

Stratum Key distinctive SFL properties in comparison with other approaches

semantics • systemic organization

• metafunctional coverage of all modes of meaning
• extension along the cline of instantiation from semantic system to acts of meaning
in text
• orientation upwards towards context and outwards towards semiotic systems other
than language
• scale from text to elements of meaning
lexicogrammar • systemic organization
• metafunctional organization of systems
• rank-based constituency, with syntax and morphology as unified domains within
grammar
• extension in delicacy from grammar to lexis, with the two forming a unified
resource of wording
phonology • systemic organization
• rank-based organization encompassing both prosodic and articulator phonology
• prosodic patterns based on naturally occurring connected speech (i.e. rhythm based
on isolated citation forms)
• both functional and natural: functional as resource for realizing wordings as prosodic
and articulatory patterns, natural in relation to phonetics

79
 Reading guides

Thanks to J.R. Firth, the British phonological tradition during this period was quite different,
being based on prosodies rather than on phonemes (e.g. Firth, 1948).
Developing early accounts of systemic phonology in the 1960s, Halliday and others worked
with Firth’s prosodic analysis, but an important innovation was his systemicization of both pro-
sodic and articulatory features, which meant that in some sense all phonological features could
be treated as prosodies freed from the kinds of structural constraints imposed by phonemics, and
they could be organized systemically or paradigmatically. The significance of the systemic ori-
entation in the interpretation of phonology and descriptions of particular phonological systems
was first shown by Halliday in his account of prosodic phonology in English (Halliday, 1963a,
1963b, 1967), and then used in the description of phonological units of all ranks by Barnwell
(1969) in her description of Mbembe phonology. While prosodies were echoed in the tiers in
autosegmental phonology, systemic phonology remains the only theory of phonology based on
paradigmatic organization as the primary mode of organization (see further, Matthiessen, 2021b:
Chapter 6).

2.6.3.2 Lexicogrammar
As in the case of studies of phonology, there are geographically distinct traditions or schools in
the study of grammar emerging from the first half of the 20th century, and many of them are the
same: US American structuralism, later partly transformed into generativism; US West-Coast
Functionalism, developing as a reaction against the formal approach of generative linguistics;
Prague School Functionalism; and Firthian system-structure theory transformed by Halliday
and his collaborators into systemic functional linguistics – and systemic functional grammar in
reference to the grammatical subsystem of language.
However, while the move in generative phonology towards “prosodies” in Firthian and sys-
temic terms seems to have a kind of rediscovery of insights from Firth’s prosodic analysis, the
influence of Halliday’s systemic functional grammar on generative grammars from the 1970s and
1980s has been more direct, although not always highlighted or at least acknowledged by these
generative linguists, making it sometimes difficult to trace the flow of ideas.
Some key lines of development of theories of grammar since the 1960s and interactions
between them are shown in Figure 2.7. The systemic functional lines of development are shown
roughly in the middle, between US American structuralist and generative developments and
continental European developments within the Prague School.
The systemic functional lines draw most directly on Firth’s system structure theory; the
first explicitly theoretical step in the direction of what was to become the systemic functional
theory of grammar (“SFG”) was taken by Halliday (1961), who introduced what came to be
called scale-&-category theory (introduced in detail by Berry, 1975/1977). In a sense, Halliday
was filling in an area that Firth had barely touched – more specifically, the area of grammar
between Firth’s work on context (and to some extent semantics) and his work on phonol-
ogy (accompanied by phonetic studies). This was an essential step in the move towards a more
holistic theory, and accounts that could support text analysis (see Halliday, 1964a). As the 1960s
unfolded, Halliday moved towards a more clearly systemic grammar (Halliday, 1966a), i.e. one
where the paradigmatic axis was given priority, with syntagmatic patterns derived from it by
means of realization statements, and also towards a more explicitly functional grammar, more
specifically a theory of grammar incorporating his emerging theory of metafunction (e.g. Hal-
liday, 1967/1968, 1969, 1970a, 1970d). In the 1960s, there were some significant interactions
between Halliday and Lamb (see e.g. Lamb, 2013; and Halliday’s foreword to García, Sulli-
van and Tsiang, 2017), who was developing his Stratificational Linguistics (e.g. Lamb, 1966;

80
 Reading guides

Lockwood, 1972), not shown in the diagram, and also between Halliday and Prague School
linguists, in particular with respect to the textual metafunction (e.g. Halliday, 1974a). (Halliday
had developed his insights into the complementarity of the systems of theme and information
in his work on Chinese in the 1950s, before he became familiar with the Prague School work.
As Fries, 1981, shows, Halliday kept these two systems distinct, separating Theme and Given,
unlike Mathesius in his pioneering contributions to the Prague School account.)
By the 1970s, the theory of grammar had emerged as SFG – both systemic and functional.
This was the decade when variants of SFG appeared – those developed by Hudson (1971) and
by Fawcett (1973a). Hudson’s work gradually took him away from SFG, in a couple of steps:
Hudson’s (1976) Daughter Dependency Grammar and then his (e.g. 1984) Word Grammar (cf.
Hudson, 1987). But Fawcett remained within SFL, and together with a team in Cardiff devel-
oped what has come to be known as the “Cardiff Grammar” (e.g. Fawcett, 2008; Schulz and
Fontaine, 2019).
During the same period, Halliday’s SFG influenced approaches to grammar outside SFL.
Martin Kay set out to develop a computational version of SFG, and this led him to the cre-
ation of FUG, Functional Unification Grammar (e.g. Kay, 1979). Kay was at Xerox Palo Alto
Research Center at the time, and his work influenced a junior colleague and co-researcher, Ron
Kaplan, who went on to develop Lexical Functional Grammar together with Joan Bresnan at
Stanford University (where Halliday visited in 1973 and again in 1980 on a sabbatical). (Both
Halliday and Kay had worked on Margaret Masterman’s machine translation project, at different
times in the 1950s.)

Figure 2.10 Systemic functional grammar and other approaches to grammar

81
 Reading guides

Kay’s (e.g. 1979) work on FUG led to a more general development of grammars broadly based
on his notion of unification in the (computational) representation of grammatical descriptions.
We can recognize a family of unification-based grammars, where descriptive fragments of
the grammar of a language are unified to create more complete descriptions – see Figure 2.11;
these grammars can all be contrasted with the Chomskyan line of development, originally
conceived of as “transformational grammar”. The criterion for membership in the family is
the nature of the representational system rather than higher-level theory (cf. Figure 1.2), and
according to this criterion we can recognize FUG and SFG as members, but also generative
approaches developed in the late 1970s and 1980s, in particular LFG (Lexical Functional Gram-
mar; e.g. Kaplan and Bresnan (1982); Bresnan et al., 2016) and HPSG (Head-Driven Phrase
Structure Grammar; e.g. Pollard and Sag, 1993; Boas and Sag, 2010) and, more specifically
within computational linguistics, Aravind Joshi’s TAG (Tree Adjoining Grammar; e.g. Joshi and
Schabes, 1997). Another approach to the characterization of (part of) this family is Winograd’s
(1983) notion of feature and function grammars. During the 1980s, GPSG (Generalized Phrase
Structure Grammar; e.g. Gazdar, Pullum and Sag, 1985) was also a generative alternative to the
Chomskyan line of development. It has disappeared from the scene now, but it had one interest-
ing notion from the point of view of SFG. Transformations were reinterpreted as “meta-rules”
(e.g. the relationship between active and passive clauses); they were meta in relation syntagmatic
patterns, so can be thought of as a syntagmatic step towards paradigmatic relations.

Figure 2.11 SFL in relation to other approaches (to grammar) in terms of the theoretical and representational
levels of metalanguage

82
 Reading guides

If we move up one stratum or level within metalanguage from representation to theory,

we can recognize other alignments; in particular, here we can identify a family of functional
grammars based on theoretical principles such as the functional organization of grammar,
the relationship between grammar and discourse and the some version of the recognition
that the relationship between grammar and semantics is natural rather than conventional (or
“arbitrary”). This family includes the Prague School and West-Coast Functionalism. With the
development of interest in discourse, Dik’s (e.g. 1978) Functional Grammar (FG) has arguably
moved closer to this higher-level notion of functionalism – Functional Discourse Grammar
(e.g. Hengeveld and Mackenzie, 2008), whereas Role and Reference Grammar (RRG) still
seems more concerned with functional representations than with high-level functional theory.
Butler (2005) compares FG, RRG and SFG in terms of their treatment of information focus.22
Going beyond the grammatical frameworks mentioned earlier, Butler and Gonzálvez-García
(2014) and Gonzálvez-García and Butler (2006) report on a multidimensional analysis of con-
tributions to what they call the “functional-cognitive” space (see also Section 2.6.3.3).
The approaches to grammar mentioned earlier and included in Figures 2.10 and 2.11 came
into focus in the 1980s, often with roots in the 1970s. There are certain themes that would
be helpful to include, like the continued interest in and computational implementations of
dependency grammar. Key conceptions can be traced back to the Middle Ages (Covington,
1984), but in more recent times, Tesnière’s (1959) posthumous book has played a significant
role and Hays (1964) helped draw attention to the dependency conception of syntagmatic rela-
tions in computational linguistics. As already noted earlier, Hudson (1976) combined systemic
organization with dependency structure, and dependency grammar has played a significant role
in computational linguistics, in particular in the development of dependency parsers such as
the Stanford Parser. In terms of systemic functional theory, dependency can be interpreted as a
hypotactic interpretation of configurational patterns as well as hypotactic ones.

Figure 2.12 Constructions located along the cline of delicacy

83
 Reading guides

Dependency grammars are thus focussed on the syntagmatic axis in the first instance, offer-
ing an alternative to constituency-based theories of syntagmatic patterns. Another development
not yet covered here is focussed on the relationship between grammar and lexis – theorized in
terms of the cline of delicacy in SFL, as shown in Figure 2.12 (e.g. Halliday, 1961; Hasan, 1987;
Tucker, 1996b, 1998, 2002, 2005, 2007; Matthiessen, 1991a, 2014c; Neale, 2002, 2006). This
is what we might call the family of Construction Grammars. Fillmore (1988: 36) charac-
terizes constructions as follows: “By grammatical construction we mean any syntactic pattern
which is assigned one or more conventional functions in a language, together with whatever
is linguistically conventionalized about its contribution to the meaning or the use of structures
containing it.” This is usually discussed in terms of “form-meaning pairings”, also with reference
to Saussure’s characterization of the sign as a combination of signifier and signified.
Since the early work on constructions by Fillmore and his colleagues at UCB (e.g. Fillmore,
1988), which was informed by his earlier research into deep case (Fillmore, 1968) and frames
(Fillmore, 1977) and which is related to the development of FrameNet (e.g. Ruppenhofer et
al., 2016),23 different members of the family of Construction Grammars have been developed.
Goldberg (2006: 213–214), one of the key scholars in the UC Berkeley tradition, identifies four
approaches aligned with “the constructionist approach”: Unification Construction Grammar
(e.g. Fillmore, 1999), Cognitive Grammar (e.g. Langacker, 2008), Radical Construction Gram-
mar (e.g. Lakoff, 1987; Goldberg, 1995) and Cognitive Construction Grammar (e.g. Goldberg,
1995). We can also add Sign-Based Construction Grammar (which also draws on HPSG; Boas
and Sag, 2010) and Fluid Construction Grammar (e.g. Steels and Beuls, 2019); for variation
within the family of Construction Grammars, see further Hoffman and Trousdale (2013).
The members of this family share several properties (summarized by Goldberg, 2003: 219,
as seven tenets, formulated partly to bring out the contrast with “the mainstream generative
approach”), but at the same time they also differ in various ways, for example in how they
relate grammar to semantics and how explicit they are. Crucial among the shared properties
is the conceptualization of the continuity between grammar and lexis. In their introduction
to the Oxford Handbook of Construction Grammar, Hoffman and Trousdale (2013: 1) write:
“Instead of assuming a clear-cut division of lexicon and syntax, Construction Grammarians . . .
consider all constructions to be part of a lexicon-syntax continuum (a ‘construction’, Fillmore,
1988; see also Jurafsky, 1992; Goldberg, 2003: 223).” They give as examples a single word
(apple), an idiom construction (X take Y for granted), a comparative construction (John is taller
than you) and a resultative construction (She rocks the baby to sleep). Constructions such as idioms,
comparative constructions and resultative ones are located along the cline of delicacy interme-
diate between general grammatical systems and delicate lexical ones, as indicated in Figure 2.1.
(For an examination of possible systemic functional accounts of phraseology, see Tucker, 2002,
2007, also with references to relevant contributions from the Birmingham School.)
Langacker (2005) compares three members of the family of Construction Grammars: his own
Cognitive Grammar, Radical Construction Grammar and “just plain Construction Grammar”,
taking Goldberg (1995) as “by default as a comprehensive and readily accessible statement”. He
sheds light on their similarities and differences. Among the differences, he elucidates one under
the heading of “grammatical form” that is significant in the overall architecture of language and
with the respects to the conception of grammar (i.e., lexicogrammar). Interpreted theoretically
in terms of SFL, the difference has to do with the stratal conception of lexicogrammar. In Lan-
gacker’s Cognitive Grammar, grammar is embodied in the relationship between semantic and
phonological structures whereas in the other two approaches, there is a distinct “level” of gram-
matical form. In Langacker’s (2005: 104–105) formulation:

84
 Reading guides

In Cognitive Grammar, as is clearly spelled out in all the published formulations, the
form in a form-meaning pairing is specifically phonological structure. . . . crucially,
it does not include what might be called grammatical form. In both Construction
Grammar and Radical Construction grammar, the form part of a form-meaning pair-
ing does include grammatical form. Thus Goldberg (1995: 51) speaks of “a pairing
between a semantic level and a syntactic level of grammatical functions”. More explic-
itly, Croft (2001: 62) says that a construction is symbolic by virtue of being “a pairing
of a morphosyntactic structure with a semantic structure”.
The point is not a trivial one or just a matter of terminological preference. The
difference goes straight to the heart of two foundational issues: the nature of gram-
mar and its relationship to meaning. In this respect, the vision of Cognitive Gram-
mar is clear, consistent, and a radical departure from previous traditions. Grammar
is symbolic in nature, where symbolic structures reside in the pairings of semantic
and phonological structures. On this view, grammar (or grammatical form) does not
symbolize semantic structure, but rather incorporates it, as one of its two poles. If
grammar is wholly reducible to assemblies of symbolic structures, it is incoherent
to say that some aspect of grammar functions as the symbolizing element in such
assemblies.

This diference is clearly stratal in nature (as it would be if viewed in terms of Stratificational
Linguistics). From a systemic functional point of view, it is interesting to ask how the organi-
zation of language into stratal planes, the content and expression planes, is handled in difer-
ent versions of Construction Grammar; how the natural relationship between semantics and
lexicogrammar is conceptualized; and how the distinction between congruent and incongruent
(i.e. metaphoric) relations semantics and lexicogrammar are theorized and modelled. (While
various discussions of Construction Grammars refer to the Saussurean conception of the sign,
SFL relates to the further development by Hjelmslev, critically involving the stratification of the
content and expression planes.)
While Hunston and Francis’s (2000) pattern grammar is not usually included in discussions
of Construction Grammars, it is also worth noting here. It is “usage based”, and it foregrounds
patterns that become “visible” when corpora are investigated with corpus tools. Such patterns
are syntagmatic in the first instance and in terms of trinocular vision, they thus represent a view
of lexicogrammar from below in terms of finding patterns but interpret them from above. Dis-
cussing the complementarity of grammar and lexis, Halliday (2008b: 71) provides a view of their
valuable contribution:

They say that their pattern grammar approach “avoids a distinction between lexis and
grammar” (p. 36). I would perhaps say that it puts the distinction into perspective,
rather than avoiding it; they make considerable use of grammatical categories (see
for example their introductory sentence of “ergative verbs” on p. 183), just as they
use the concept of word or lexical item throughout the book. What they are doing
is medium delicacy grammar, setting up grammatical classes “on the basis of their
behaviour” and giving them names (e.g. “shell nouns”, p. 185). Their perspective
typically foregrounds the view “from above”, so that “they (shell nouns) are divided
into two groups on the semantic grounds suggested above”; this in fact corresponds
to the distinction between locution and idea, both of which are then distinguished
from fact-like nouns. The results are entirely in harmony with the categorization,

85
 Reading guides

derived from grammar, of the three types of projection: reports, ideas and facts (Hal-
liday, 1985a: 227–251).

We began our brief comments on the family of Construction Grammars by referring to the
cline of delicacy within lexicogrammar, noting that members of this family share Halliday’s
(1961) insight that grammar and lexis are note separate “modules” but form a continuum (see
also Section 3.1.4). Typical “constructions” are located somewhere within the regions between
the grammatical and lexical zones of lexicogrammar. In SFL, they need to be interpreted and
placed not only syntagmatically but also paradigmatically; this is precisely how their location
along the cline of delicacy can be determined (cf. Tucker, 2007). There is another semiotic
dimension relevant to the interpretation of their nature, viz. the cline of instantiation. Lexico-
grammatical patterns discussed under headings such as phraseology, idioms, formulaic language
and clichés obviously fall within the overall lexicogrammatical system of a language, and more
generally within its meaning potential, and they can be characterized at some degree of delicacy,
but they are also located along the cline of instantiation, the limiting case being pre-instantiated
patterns. Halliday (1992b/2003b: 367–368, emphasis in original) makes this point in reference
to “readymade” wording:

Many sentences are stored readymade; they are more or less formulaic for the speaker
and for the listener, and any given instance of their occurrence harks back to the previ-
ous instances of the same wording rather than being engendered afresh by choosing
within the system. Examples from English are:
Take it or leave it!
I couldn’t believe my ears.
It’s a contradiction in terms.
He doesn’t know what he’s talking about.
Not by any stretch of the imagination.
You can’t change human nature.
You must keep your eye on the ball.
etc.
Of course these items are still located within the system and derive their meaning
ultimately by semantic proportionality in the usual way; but their immediate source
is the trace of earlier occurrences. (This is sometimes seen as a form of intertextual
effect; but it is actually rather different. Intertextual potential is strongest where the act
of meaning exploits the full resources of the system; it tends precisely to be weakened
where the act of meaning is locked into a formula, since this inhibits the search for
other semiotic input.)

Halliday’s point that such items “are still located within the system” is shown by the fact that
there is usually a set of related systemic variants. For example, when we search the Corpus of
Contemporary American English (COCA), we find variants such as:

Only by the greatest stretch of the imagination did the section . . .

by no stretch of the imagination should it be . . .
by what possible stretch of the imagination does the square . . .
a flexible smartphone isn’t exactly a stretch of the imagination
it’s not too much of a stretch of the imagination . . .

86
 Reading guides

it’s not a stretch of the imagination at all . . .

. . . is not such a stretch of the imagination
there was nothing that was in the wildest stretch of the imagination the least bit “murky” . . .
this is a crazy stretch of the imagination . . .
. . . sounded like an unnecessary stretch of the imagination . . .
by any stretch of the sociological imagination, the recipe for marginalization . . .
my job isn’t exciting or highly paid by any stretch of the American imagination

SFG is diferent from all the other approaches to grammar in a number of respects; two central
ones are:

• Theoretically, it is the only approach that is paradigmatically-based – hence “systemic”;

and from this follow a number of other properties, including
• the continuum from grammar to lexis;
• the metafunctional organization;
• the continuum from system to text (i.e. from the potential pole of the cline of instan-
tiation to the instance pole).
• Metatheoretically, SFG has been designed to be appliable in nature from the start (cf. Sec-
tion 4.1) – articulated by Halliday (1964a) under the heading of “syntax and the consumer” –
and its range of actual application is almost certainly vastly more extensive than any of the
other approaches mentioned here.

It is also important to note that the approach to grammar in SFL always also involves the other
stratum of the content plane, i.e. semantics. In this respect, the functional approach means that
grammar is interpreted as natural grammar, in the sense that grammar is shown to stand in a
natural relationship to semantics. This also underpins the theory of lexicogrammatical meta-
phor, where incongruent relations of realization between semantics and lexicogrammar are taken
account of and described. The traditional focus was on lexical metaphor, and Halliday comple-
mented this by giving attention to grammatical metaphor (e.g. Halliday, 1985a: Chapter 10,
1988a; Halliday and Matthiessen, 1999/2006: Chapter 6; Simon-Vandenbergen, Taverniers and
Ravelli, 2003). But the metaphor extends from grammar to lexis, so it is properly theorized as
lexicogrammatical metaphor. (Within cognitive linguistics, lexical metaphor has been studied
under the heading of “conceptual metaphor”, which includes examples of the kind Whorf,
1956, had highlighted.) See further Section 3.3.1.4.

2.6.3.3 Semantics
While lexicogrammar is one of the two inner strata of language, constituting the form strata
together with phonology (graphology) (cf. Figure 2.3), and has been dealt with largely within
linguistics – but also within computational linguistics, semantics is one of the two outer strata,
constituting the substance strata together with phonetics (graphetics). Since it is an outer stra-
tum, it has come into view within a wider spectrum of disciplines, including not only linguistics
and but also philosophy and logic, ethnography, rhetoric, neurolinguistics and neurosemiotics,
and cognitive science as a macro-discipline (which covers both cognitive psychology and AI):
see Figure 2.13 (taken from Halliday and Matthiessen, 1999/2006).
Broadly, as shown in Figure 2.13, we can identify three traditions in the approach to meaning
in linguistics and other disciplines:

87
 Reading guides

Figure 2.13 Approaches to meaning in different traditions

• functional approaches to meaning: these have been developed in anthropologically/

ethnographically oriented traditions such as SFL, the Ethnography of Speaking and Tag-
memic Linguistics. Here text tends to be taken to be the primary domain of meaning –
characterized as “language functioning in context” in SFL (e.g. Halliday and Hasan, 1976).
We find a number of studies that are close enough to SFL that it is possible to draw on them in
SFL, including Grimes (1975), Longacre (1996), Longacre and Hwang (2012); Saville-Troike
(1982). Here we can also add contributions from the Prague School and from the Birming-
ham School. The orientation in contributions broadly characterizable as functional semantics

88
 Reading guides

is toward rhetoric rather than towards logic, so contributions to rhetoric such as Kinneavy
(1971) can also serve as sources of insight in the further development of systemic functional
semantics.
• formal approaches to meaning: these have originated outside linguistics in philosophy
and logic. In particular, there is a tradition of model-theoretic formal semantics, pioneered
by Richard Montague (who was influenced by Alfred Tarski’s work on semantics for for-
mal logic in the 1930s, e.g. Tarski, 1936) in a series of publications collected by Thomason
(1974). This work has been continued in philosophy and logic, explored in computational
linguistics, but also been taken up in linguistics by linguists, including Barbara Hall Partee,
Emmon Bach and Edward Keenan (see e.g. Lappin and Fox, 2015; Aloni and Dekker,
2016). Formal semantics tends to be focussed on clauses or sentences rather than on texts,
but it has been extended by Hans Kamp in his “Discourse Representation Theory” (e.g.
Kamp and Reyle, 1993). The interaction between model-theoretic approaches to meaning
and SFL has been minimal, one area where they may come into contact being computa-
tional modelling (cf. Bateman and Rondhuis, 1997).
• cognitive approaches to meaning: these range from ones that are aligned with formal
linguistics in the Chomskyan sense, in particular the work by Ray Jackendoff (e.g. 1972,
1983, 1988, 2012), to ones that are part of cognitive linguistics of the West Coast variety
in the US, associated in particular with George Lakoff and his group at UC Berkeley (e.g.
Lakoff and Johnson, 1980, 1999; Lakoff, 1987), but also with Ron Langacker at US San
Diego (e.g. Langacker, 1990, 2008, 2009, 2013).

The closest relationship between the systemic functional approach to meaning and other
approaches is to functional approaches to meaning and to Lakof’s version of cognitive linguis-
tics (cf. Halliday and Matthiessen, 1999/2006), both in its concern with metaphor and in its
concern with the role of language in the realm of politics; but Kristin Davidse has pioneered an
approach that draws both on Langacker’s cognitive linguistics and functional linguistics (see e.g.
Vandelanotte, 2009; Davidse and Breban, 2019). Butler (2019) locates SFL in “what has been
called functional-cognitive space, a multidimensional space based on a wide range of proper-
ties, in which various functionally and/or cognitively and/or constructionist to language can be
plotted” (see Gonzálvez-García and Butler, 2006; Butler and Gonzálvez-García, 2014).
Against the background on these brief notes on functional, formal and cognitive approaches
to meaning, we can ask what sets SFL apart in its approach to meaning – what is characteristic
of systemic functional semantics? Theoretically, it is distinct in a number of ways, including
centrally:

• semiotic (in ordered typology of systems): semantics is construed as a stratal subsystem

within language as a higher-order semiotic, located in an ordered typology of systems (Sec-
tion 1.3.2), so as part of language semantics is socially enacted and biologically embodied
and physically manifested. This means, among other things, that the modelling of meaning
in semantics incorporates “sociological” considerations (cf. Halliday, 1972) – not only neu-
rological (or cognitive) ones.
• stratal – content plane: in systemic functional linguistics, semantics is theorized as the
higher (and outer) of the two content strata, standing in a natural relationship to the lower
(and inner) stratum, lexicogrammar, and serving as an interlevel or interface to contexts and
other systems that lie beyond language (see Figure 2.3).
• systemic: like all other systems of language in context, semantics is modelled paradigmatically –
it is organized systemically as a resource for making meaning, with systemic terms being

89
 Reading guides

realized structurally along the syntagmatic axis. This means, among other things, that
meanings are interpreted in terms of their value or valeur in their systems, and that choices
in meaning are given an important place in the overall account (cf. Tsui, 1989). Thus both
valeur and signification are crucial aspects of the modelling of meaning in SFL (cf. Hasan,
1985a).
• textual: semantics extends from whole texts, defined by reference “upwards” as language
functioning in context (Halliday and Hasan, 1976), to the smallest elements of meaning; its
compositional scope is not restricted by the grammatical rank scale of units – it extends far
beyond the domains of meaning realized by clauses and clause complexes, which is why a
number of systemic functional linguists have followed Martin (e.g. 1992a) in referring to
semantics as “discourse semantics”.
• metafunctional: while many traditions distribute meaning across two “modules”, seman-
tics and pragmatics, where semantics covers ideational meaning and pragmatics covers inter-
personal and textual meaning (speech acts, reference, information structure and so on), in
contrast, in SFL, semantics covers the full metafunctional spectrum of meaning, which has
important consequences for the overall account of meaning (cf. Halliday and Matthiessen,
1999/2006).
• multimodal: since semantics is modelled stratally as an interlevel, it is also an interface to
meaning to denotative semiotic systems other than language, including the ones that have
already been taken account of and described in SFL (see Section 3.4).

The interface to the modelling of meaning in systems other than language is very clear in the
research by John Bateman and his colleagues on the description of the ideational semantics of
space working together with models of space needed by robots moving around and perceiving
space (see e.g. Bateman et al., 2010). Their work provides essential guiding principles for the
further development of semantics as a true interface to other semiotic systems, both social semi-
otic ones and bio-semiotic ones (cf. Matthiessen, 2015e). The interface nature is now also being
illuminated in neurolinguistic work undertaken by Adolfo García, Augustín Ibáñez and other
members of their group, in in their work with explicit references to SFL. García and Ibáñez
(2017) interpret neurolinguistic studies in the light of the semantics of ‘material’ clauses (seman-
tically, figures of doing-&-happening), and put forward a proposal for a revised version of the
stratified SFL model of language in context that has an explicit place for sensorimotor systems.

2.6.4 Genre and register studies

Functional variation, i.e. variation according to context of use, has had a central place on
the systemic functional research agenda since the 1960s (e.g. Halliday, McIntosh and Strev-
ens, 1964; Ure, 1971; Ure and Ellis, 1977), developed as a generalization of Firth’s notion
of restricted languages.24 It been called register variation or diatypic variation, and it’s
been contrasted with dialectal variation and codal variation (e.g. Hasan, 1973, 1989, 2009b;
Halliday, 1978, 1994d) – see Section 3.2. While different in conception, genre has also been
used as a way of conceptualizing functional varieties of language within SFL – viewing such
functional varieties from the point of view of context, as in the “Genre Model” developed by
J.R. Martin (e.g. 1992a) and his colleagues over the years, originally typically in educational
linguistics (see e.g. Rose and Martin, 2012; Drury and Jones, 2010; Dreyfus et al., 2016;
Humphrey et al., 2010; Mahboob et al., 2010). Within SFL, different perspectives on func-
tional variation have thus been developed, and we compare them in Section 5.2.1 (cf. also
Section 6.2 on “tricky terms”).

90
 Reading guides

At the same time, different disciplinary traditions have come to study the phenomenon of
functional variation in context using different but ultimately related labels, in particular “genre”,
“register” and “text type” (cf. also Biber and Conrad, 2009, for a non-SFL overview).

• Genre studies have been undertaken in different disciplines; they include not only linguis-
tics (in SFL and also in Tagmemics or the tradition close to Tagmemics represented by
Longacre, 1974, 1996; see also Longacre and Hwang, 2012) but also anthropology, rheto-
ric, studies of literature or verbal art – and by extension also studies of art created in other
semiotic systems (e.g. discussions of genres in the history of painting) and in composite (or
“multimodal”) semiotic systems (e.g. discussions of genres in film studies).
• Register studies have tended to be undertaken in linguistics, first essentially within SFL,
going back to Halliday, McIntosh and Strevens (1964), Ure (1971), Hasan (1973) and Halliday
(1978), but since the late 1980s and more especially the 1990s register studies (Leckie-Tarry,
1995) have also been taken up by or embedded within other linguistic traditions –
Biber and Finegan (1994) being central in establishing a North-American tradition along
the lines formulated by Biber (1994).
• Text typology studies tend to be associated with continental European contributions start-
ing in the 1970s, Werlich’s (1975) model being an interesting and influential contribution;
text typology has also been taken up in translation studies, again ones where continen-
tal Europe figures prominently (e.g. Snell-Hornby, 1995; Nord, 2005). (For a more gen-
eral discussion of linguistic variation, or “heterogeneity”, as a challenge to translators, see
Steiner, 2004, 2005b.)

There are, naturally, various diferences among these approaches, but arguably the most funda-
mental one is that the SFL account of functional variation is part of a holistic view of language
in context that has led to the development of a unified theory. Functional variation – whether
explored under the heading of register (diatypic variation) or genre – is defined and character-
ized not as a stand-alone account but rather as part of the total theory in terms of the relations
this variation enters into, which are centrally the cline of instantiation and the hierarchy of strati-
fication, both of which are needed to locate the zone of variation: see Section 3.2.
Different approaches to the study of functional variation in language are charted in Figure
2.14 (see Matthiessen, 2019, for further details). The chart is not exhaustive, leaving out e.g.
work on sub-languages in machine translation (e.g. Kittredge, 1987; Kittredge and Lehberger,
1982; Grishman and Kittredge, 1986) and on text typologies in translation studies, but it gives
an indication of the “neighbourhood” in which systemic functional accounts of functional varia-
tion are developed and used under the headings of “register” or “genre”. There is now a tra-
dition of recognizing and comparing three approaches to “genre”, particularly in the context
of education, including academic writing; this tradition goes back to Hyon’s (1996) valuable
comparison of the Swalesian genre tradition (beginning with Swales, 1990), genre in “New
Rhetoric” (e.g. Miller, 1984) and in the “Genre Model” developed in SFL by J.R. Martin and
his team. It has been further discussed in various places, including Swales (2009) and Malavska
(2016). However, it is important to note that this framing of work on genre provides quite a
limited view, leaving out relevant work on functional variation in other traditions.

2.7 Systematic literature reviews, research synthesis and meta-analysis

There are different ways of reviewing past research as it has been “reported” in the literature. On
the one hand, scholars may choose different fields of activity: they may report on the literature,

91
 Reading guides
92

Figure 2.14 Explorations of functional variation in context in different linguistic and semiotic traditions
 Reading guides

Figure 2.15 Choices in “traditional” literature reviews

simply inventorying past work or recounting it to construe the history of a field of research; or
they may explore the literature in terms of contributions made, evaluating them and arguing
about them by supporting them or refuting them. On the other hand, scholars may focus on
different domains of research and scholarship – empirical outcomes of analysis or description,
theoretical categories and issues, or methodology. At the same time, we can also note that the
literature review may have the status of a stand-alone article or section in an article, a book or
a research proposal, or it may be spliced into, or integrated with, another text. The different
possibilities are set out in idealized schematic form in Figure 2.15. Of course, there are other
important choices reviewers have to make, e.g. in terms of field considerations – coverage (rang-
ing from comprehensive or exhaustive [and exhausting!] to selective according to certain criteria
such as representativeness or impact in the field) and tenor considerations – expertise of audience
(ranging from general public to expect scholars).
The best places to find examples of different types of literature review are, of course, in jour-
nal articles, textbooks and – importantly – handbooks and encyclopaedias. We have inventoried
a number of these earlier in the chapter, organized according to different academic concerns,
and we also provide tabular overviews in the Appendix (Table 9.3, Table 9.4 and Table 9.5).
The choices specified in Figure 2.15 probably cover most literature reviews that have appeared
in SFL – although that is of course a claim that needs to be tested in a large-scale literature
review! However scholars in a number of disciplines have transcended this traditional approach
to surveying and reviewing the literature, developing more systematic approaches to literature
reviews encompassing research synthesis and, where possible, meta-analysis (see Cooper and
Valentine, 2009; Cooper, 2017). Unlike our own literature review here and most SFL literature
reviews (but see examples of exceptions later), research synthesis involves searches of bibliographic

93
 Reading guides

databases according to explicit criteria for inclusion and exclusion of contributions to the litera-
ture and also clear consistent schemes for coding or classifying contributions that are included,
often with additional “meta-data” such as region of scholarship and language of presentation.
The terminology is fairly fluid here, as Cooper and Hedges (2009) note:

The terms research synthesis, research review, and systematic review are often used interchangeably
in the social science literature, though they sometimes connote subtly different meanings.
Regrettably, there is no consensus about whether these differences are really meaningful.
Therefore, we will use the term research synthesis most frequently throughout this book.
The reason for this is simple. In addition to its use in the context of research synthesis, the
term research review is used as well to describe the activities of evaluating the quality of
research. For example, a journal editor will obtain research reviews when deciding whether
to publish a manuscript. Because research syntheses often include this type of evaluative
review of research, using the term research synthesis avoids confusion. The term systematic
review is less often used in the context of research evaluation, though the confusion is still
there, and also it is a term less familiar to social scientists than is research synthesis.

In our discussion of literature reviews, we will use the term systematic review to indicate
that the criteria for the review have been made explicit – crucially the criteria for including and
excluding past research reported in the literature. Such systematic reviews are the prelude to
research synthesis; here we, as reviewers, take a step back to determine how distinct research
contributions complement one another (obviously allowing for overlap and even contradiction)
and can be brought together to produce a more comprehensive account. If the results that are
reported by diferent contributions are comparable, we should be able to compare and contrast
them. When we do so, we treat the results that have been reported as “data”, but as 2nd-order
data (since these data are reported data), and we can undertake a meta-analysis of these data.
This is of course easier if the reported data are quantitative in nature, as is often the case in clini-
cal research studies or sociological questionnaire-based studies, but using quantitative data is not
a necessary condition.
Drawing on work in sociology, Norris and Ortega (2006) have introduced research synthesis
into applied linguistics. Ortega (2015) characterizes it succinctly as follows:

Research synthesis refers to a continuum of techniques and research procedures that

have been developed by social scientists with the aim of reviewing literature system-
atically. Simply put, the methodology produces contemporary literature reviews that
differ from traditional literature reviews in taking an empirical perspective on the task
of reviewing. Syntheses (also known as systematic reviews in some fields) investigate
and evaluate past findings in a systematic fashion, always explicating the methodology
followed in the review so as to enable replication by other reviewers. The approach
began in the late 1970s in the fields of psychology, education and medicine and its
use has been widespread across the social sciences since the late 1980s. In applied lin-
guistics, the methodology was introduced in the mid 1990s and has witnessed a rapid
development particularly since the late 2000s. . . .
Meta-analysis is probably the best-known technique systematically synthesizing
quantitative research. However, meta-analyses are a restricted form of synthesis that
cannot always be conducted. For one, synthesists can engage in meta-analysis only
when the body of research to be synthesized is clearly quantitative, that is, experimen-
tal, quasi-experimental or correlational. In addition, the methodology of meta-analysis
is driven by questions about the magnitude of an effect or causal relationship and
94
 Reading guides

involves mathematical ways of summarizing past findings that demand the availability
of a large number of studies.

In SFL, de Silva Joyce and Feez (2016: 276–278) discuss the use of meta-analysis in the context
of researching literacy:

Meta-analysis is a process of reviewing the results of multiple research studies in a

particular field to look for contrasts, similarities, patterns or disagreements. The aim
of meta-analysis is to look for recurring evidence of statistical significance in order to
draw generalisable conclusions.

They ofer Hattie (2009) as an example. In SFL, the most likely applications of meta-analysis
would probably be in studies based on the analysis of large corpora and of widely distributed
questionnaires (although the latter method is not very common in SFL since it involves relying
on what people say they do, say and think rather than on what they actually do, say and think –
projected information rather than non-projected, straight, information). For meta-analysis to
be a possible further step beyond 1st-order analysis, researchers would have to develop com-
mon standards to be used in recording and reporting findings, ensuring that they are more
widely shareable than is currently the case. Naturally, the use of particular tools such as Mick
O’Donnell’s UAM Tools or ELAN will help researchers collaborate and share results (cf. Sec-
tions 2.6.2.3, 7.1.2 and 9.1.3).
Thanks in large part to the efforts by Winfred Xuan, who learnt research synthesis from
John Norris and Lourdes Ortega at Georgetown University, and his collaborators, there is now
a growing body of syntheses of research in SFL: Mwinlaaru and Xuan (2016) on systemic func-
tional typology and description of particular languages, Xuan and Chen (2019) on grammatical
metaphor, Xuan and Chen (2020) on projection, and Chen, Xuan and Yu (2022) on translation
studies. Another recent contribution is Schwartz and Hamman-Ortiz’s (2020) “comprehen-
sive review of the empirical scholarship exploring the intersection of SFL, elementary writing
pedagogy, teacher education, and Els”, where they set out the methodology clearly, detailing
“eligibility criteria”, the “review process” and “coding and analysis procedures”, and then report
on their findings.
The research synthesis of systemic functional work on grammatical metaphor will serve as a
good model. Xuan and Chen (2019) summarize their study as follows:

Adopting the methodology of research synthesis, the present study takes stock of pri-
mary research studies on grammatical metaphor (GM) within systemic-functional
linguistics and maps the research landscape since Halliday proposed this concept in
1985. Based on the data from 118 primary studies, we first report on the publica-
tion and methodological features of these studies. We then conduct content analysis,
which is based on five domains of interest, viz. language description, register variation,
education, language comparison, and translation. Major findings are: (i) The type of
GM in these studies is dominated by ideational GM; less attention has been given to
interpersonal GM; (ii) Textual GM has emerged as a heated issue recently. (iii) While
education represents the most studied field, there remains considerable research space
in other domains. Finally, some theoretical issues and implications for future directions
are highlighted.

The procedure they follow can be represented diagrammatically as a pseudo-flow chart: see
Figure 2.16.
95
 Reading guides

Figure 2.16 Diagrammatic representation of Xuan and Chen’s (2019) research synthesis of SFL work on
grammatical metaphor

96
 Reading guides

Naturally, the design of any particular research synthesis project will depend on the research
questions that are formulated. However, the steps that need to be taken are very likely to be
quite similar to those taken by Xuan and Chen (2019), as are the databases consulted and the
information generated by the steps leading to the synthesis of the findings. The synthesis is of
course a result in its own right – a contribution to the systemic functional “knowledge base”
of collective contributions, and this knowledge base can serve as a resource for various activi-
ties including not only bibliographic searchers but also formulation of new research projects to
fill research gaps identified through the research synthesis: see Section 4.2 on general research
methodology.

Notes
1 Matthiessen is preparing what might be characterized as a multilingual IFG.
2 Robins (1964) includes a section on “M.A.K. Halliday: systemic grammar” (Section 7.4.3) as part of his
presentation of “Earlier post-‘structuralist’ theories” – theories that “developed more or less contempo-
raneously with Chomskyan transformational-generative theory”. In addition to systemic grammar, he
introduces Tagmemics and Stratificational linguistics.
3 See www.jasfl.jp
4 The first two editions of IFG (Halliday, 1985a, 1994c) only included examples of system networks; but
in the subsequent editions, system networks were added for all major areas of the grammar of the clause
and of groups/phrases (Halliday and Matthiessen, 2004, 2014). However, these networks were not used
as navigational tools in the way they are in Matthiessen (1995a).
5 For more recent information about Word Grammar (apart from various publications), see: http://
dickhudson.com/word-grammar/
6 We can also note Michael Gregory’s “Communication Linguistics”, developed originally at York Uni-
versity in Toronto (e.g. Gregory, 1985, 2002; Malcolm, 1985; Stainton and de Villiers, 2001) – a
resource for Asp and de Villiers (2010).
7 Personal communication with Richard Hudson (Matthiessen), at UCL in the late 1970s.
8 Personal communication with Martin Kay (Matthiessen).
9 It seems to be difficult to find references to the role of SFG in the development of LFG in the published
literature, but Joan Bresnan noted the influence of SFG on the development of LFG during the Q-&-A
session after her talk in AILA in Tokyo in 2000.
10 Referred to by Halliday in Thibault’s (1987: 5) interview with him as “culturally pregnant texts such as
religious narrative”.
11 Williams (2015: 327) observes: “It is a paper that has often been cited as highly innovative but, so far,
it has not been extensively utilized to produce research.” So here is a great research opportunity still
waiting for the community.
12 The scaffolding metaphor, introduced by Bruner (1975), was taken up also by systemic functional
scholars in valuable contributions to educational linguistics such as Gibbons (2002).
13 It is interesting to draw analogous maps based on other semiotic dimensions. For example, using the
spectrum of metafunction as a frame of reference, we could draw a map of approaches to interpersonal
meaning within different disciplines, including psychology: Osgood (e.g. 1964), Ekman (e.g. 1992),
Ekman, Friesen and Tomkins (1971); sociology: Triandis (1977), Turner (2000); and of course the
growing industry of sentiment analysis in computational approaches.
14 The critical stance adopted is a manifestation of the critical turn in a number of disciplines, informed by
the Frankfurt School and the work by Jürgen Habermas in particular (cf. Figure 2.5). Among linguists
taking a professional interest in activism and initiatives with the potential of improving the human
condition, Halliday and Chomsky represent two interestingly different positions. Chomsky is of course
well known for his activism, but it is treated as completely distinct from his theoretical linguistics; but
Halliday has been pursuing an appliable kind of linguistics since the 1950s.
15 Van Dijk (e.g. 2008) has criticized the systemic functional conception of context, arguing that it needs
to be more cognitive; for example, he writes (p. 28): “I shall show that the SF approach to context is
misguided, and needs to be abandoned.” O’Grady (2019) reviews his criticism as part of his overview
of SFL and CDA. In addition, we can note that the account offered by Halliday and Matthiessen
(1999/2006) provides a semiotic interpretation of the phenomena that are often viewed as cognitive, an

97
 Reading guides

interpretation where semiotic systems are located in an ordered typology of systems and thus embodied
in biological systems. However, van Dijk does not refer to this work in his book, nor to other works
that could have shed light on his understanding of SFL at the time, e.g. Halliday (1993a, 1995, 2005b),
Hasan (1992c, 2004), Painter (1999).
16 The Wikipedia entry on “Ecolinguistics” includes the following observation (1 June 2020): “Michael Hal-
liday’s 1990 speech New ways of Meaning: the challenge to applied linguistics is often credited as a work which
provided the stimulus for some linguists to consider the ecological context and consequences of language.
Among other things, the challenge that Halliday put forward was to make linguistics relevant to overarching
contemporary issues, particularly the widespread destruction of ecosystems. The main example Halliday
gave was that of ‘economic growth’, describing how ‘countless texts repeated daily all around the world
contain a simple message: growth is good. Many is better than few, more is better than less, big is better
than small, grow is better than shrink’, which leads to environmentally destructive consequences.”
17 This list is certainly not exhaustive. Because functionalism has tended to be pushed to the margins since
the 1960s, when Chomskyan formalist linguistics became dominant, there were important linguistic
voices that were essentially not heard, especially scholars from scholars in the “developing world”.
18 The coastline is the US one, and there are indeed important centers along the West Coast – probably
in particular, in Santa Barbara, California, and Eugene, Oregon; but the “space” is actually more
of an abstract intellectual one, and leading contributors have also come from other parts of the US
(importantly including Paul Hopper, Barba Fox and Cecilia Ford) and indeed other parts of the world.
There have been fairly few references to SFL for a variety of complex reasons – arguably more to
the Prague School and Tagmemics. One reason is that the pioneers were trained in the 1960s when
Chomskyan linguistics was becoming increasingly dominant, and they began to rebel against it in the
1970s, sometimes constructing functional positions without the benefits of the European functional
traditions – though the Prague School was given serious attention. Another reason turned out to be
that unlike Halliday and SFL, West-Coast functional linguists were not really concerned with the
development of an appliable kind of linguistics; they were more focussed on discourse-based answers
to questions about grammar (rather than on discourse in its own right, at least initially). Yet another
reason was that an influential group of West-Coast Functionalist scholars began to work with Conver-
sation Analysis, and Emanuel Schegloff, in the mid-1980s; and they accepted the CA focus on micro-
analysis of conversation – viewed from below as social interaction (rather than as an ongoing exchange
of meanings), as can be seen in the important influential publication brought out by Ochs, Schegloff
and Thompson (1996). The partnership with CA brought with it, to a certain extent, the same kind of
anti-system stance as can be found not only in CA but also in some varieties of corpus linguistics – thus
again very different from SFL, where Halliday had theorized the relationship between system and text
in terms of the cline of instantiation. (In this respect, Givón’s version of West-Coast Functionalism is
arguably substantially different.)
19 They are not covered by Sharoff (2017).
20 See: www.corpustool.com
21 See: http://multimodal-analysis.com/products/index.html
22 Butler (2005: 591) writes: “The approach to information structuring taken in SFG derives ultimately
from Prague School ideas, but with one particularly important difference.” Actually, Halliday had
worked out the outline of his approach working on Chinese before he became familiar with the Prague
School work. The important difference is that Halliday separates theme and information, already in his
early work on Chinese. For the difference between the two approaches, see also Fries (1981).
23 See: https://framenet.icsi.berkeley.edu/fndrupal/
24 Restricted languages are extreme cases of registers, being almost closed – in contrast with casual con-
versation, which is towards the most open end of the cline between closed and open registers. It is not
the case that “register” is a term simply replacing J.R. Firth’s term “restricted language”, as has been
claimed on occasion. The term “register” was suggested by Jean Ure with reference to Reid (1956): see
e.g. Halliday (2008b: 114).

98
 3
RESOURCE GUIDES

Descriptions of semiotic resources – contextual resources and the resources of languages and
other denotative semiotic systems – are, not surprisingly, spread throughout the SFL literature.
Here we will bring these descriptions together in overviews; we will include both overviews and
summaries and representative examples (exemplars). By way of general overview of the resources
identified by SFL, we include Table 3.1. It identified key publications according to the dimen-
sions of the systemic functional architecture of language in context.
There are many ways in which we could have organized our overview of guides to accounts
of different resources – e.g. metafunctionally or stratally. Here we will choose the hierarchy of
stratification as the primary organizing principle, starting with context. In this connection, it is
important to note that accounts of the lexicogrammatical stratum now exist for a reasonably wide
range of languages, so there is considerable awareness of typological variation, but accounts of
context and semantics still tend to be drawn primarily from English material. When we turn to
phonology, moving from the content plane to the expression plane, it is of course not possible to
engage in implicit “transfer comparison”; linguists cannot simply take over descriptions originally
grounded in English, as can happen with descriptions of English context and English semantics
(cf. the explicit inclusion of “English” in Martin’s, 1992a, English Text: System and Structure).

3.1 Organized according to the hierarchy of stratification

In Section 1.3, we sketched the overall architecture of language in context according to SFL
(noting that we will identify various alternatives in Chapter 5). We emphasized that the archi-
tecture is based on a number of semiotic dimensions that are the domains of different relations
such as realization and instantiation. This dimensional-relational organization of language in
context is visualized in Figure 1.2, and we will use the global dimensions – i.e. the hierarchy of
stratification, the cline of instantiation and the spectrum of metafunction – to present our guides
to systemic functional resources, both theoretical and descriptive ones.

3.1.1 The notion of stratification

Approaches to the study of language tend to reflect the stratification of language – or here
more appropriately – of language in context, although linguistic theories do not agree on the

99 DOI: 10.4324/9781315675718-3
 Resource guides

Table 3.1 Overviews of the “architecture” of language in context (excluding treatments of particular stratal
subsystems: see this chapter)

Focus Publications

overall overall “architecture”: Halliday (1994a, 1996a, 2002a, 2003b); Halliday and
dimensions dimensions (vectors) of Matthiessen (2014: Ch. 1); Matthiessen (2007a,
organization forthcoming a);
computational modelling: Matthiessen and Bateman
(1991)
the content systems of language Martin (1992a); Eggins (2004)
in context grounded in a
description of English
particular stratification and instantiation Halliday (1991b, 1994a, 1994b)
dimensions stratification, instantiation and Martin (2008, 2010)
individuation
stratification, delicacy, Berry (2017)
realisation and rank
instantiation Matthiessen and Bateman (1991); Matthiessen (1993b)
metafunction Halliday (1978, 1979a); Martin (1991, 1996c)
metafunction and rank Halliday (1970d, 1978); Halliday and Matthiessen
(2014: Ch. 2)
axis: systemic organization Halliday (1966a, 1969); Martin and Matthiessen (1991);
Hasan (1996b); Hasan et al. (2007); O’Grady,
Bartlett and Fontaine (2013); Martin (2013a)
axis: structural organization Halliday (1966a, 1979b); Matthiessen (1988b); Martin
(1996c); Bateman (2008b)
rank Halliday (1961, 1966b, 1967); Huddleston (1965)

identification of subsystems. From the point of view of SFL, there has been a tendency in lin-
guistics to confuse stratification and rank (cf. Halliday and Hasan, 1976; Halliday, 1981) and not
to treat stratification and metafunction as independent dimensions. Nevertheless, it makes sense
to start with stratification as the dimension upon which we base our survey of the resources of
language in context. Having dealt with the stratal subsystems, we will then complement the
overview by turning to other semiotic dimensions. Within SFL, scholars have explored the
nature of stratification itself, e.g. interpreting stratification in terms of Lemke’s (1984) notion
of metaredundancy (e.g. Halliday, 1992c; Martin, 2010; Taverniers, 2019) and highlighting
the “robustness of realizational systems” (Butt, 2008) and there are some theoretically different
proposals for the stratal organization of language in context; we discuss these in Section 5.2.
Here we can simply note that while stratification and inter-stratal realization are fundamental
to the understanding of semiotic systems, they still require a great deal of further theoretical
work, and also attention in terms of explicit modelling. (The latter point is crucial, but does not
become apparent in manual text analysis.) While diagrams representing theoretical models are
only visualizations, not explicit discursive accounts of models, they help make explicit certain
aspects of such models and bring out salient aspects of them. Therefore, we have included three
visual representations of the stratification of language in context:

• Figure 3.1 is a version of the widely used co-tangential circle representation of language in
context, where each circle represents a stratum and the relative size of each stratal subsystem
100
 Resource guides

Figure 3.1 The stratal organization of language in context into the content plane (semantics and lexic-
ogrammar) and the expression plane (phonology and phonetics, in the case of spoken language)

is shown by the size of the circle. Language and context are clearly distinguished and within
language, the content plane and the expression plane are also distinguished. (It is important
to note that language is stratified into four strata, the expression plane being stratified like
the content plane into two strata. It is thus a quadristratal system, not a tristratal one.)
• Figure 3.2 is Halliday’s (e.g. 2003c: 13) stratal diagram designed to show (1) the increase in
stratification from proto-language to language (cf. also Butt’s, 2015: 48, elaborated some-
what different version of this diagram), (2) the distinction within both linguistic planes
between form strata (lexicogrammar, phonology) and substance strata (semantics, phonet-
ics), and (3) the nature of the substance strata as interfaces to the environments of language,
the eco-social environment in the case of semantics and the bodily environment in the case
of phonetics (cf. insights presented by Thibault, 2004).
• Figure 3.3 is a version of the stratal diagram presented by Matthiessen (forthcoming b)
designed to bring out the distinction between form strata (internal to language) and sub-
stance strata (interface strata), the distinction between the content plane and the expres-
sion plane, the difference between natural inter-stratal realization and conventional (or
“arbitrary”) inter-stratal realization, and the distinct principles of organization within the
content plane and the expression plane.

These diferent visual representations are complementary since they are designed to make
somewhat diferent facets of the stratal theory of language in context visually accessible. Addi-
tional graphic representations can be found in Halliday (1973), where he brings out diferences
between uses of language and intrinsic metafunctional organization; Halliday (1985f), where he
presents a stratal diagram designed to show diferences among writing systems (redesigned here
as Figure 3.9); e.g. Martin (1992a, 2010), where he ofers another pictorial representation to
elucidate the stratification of context and the nature of context as a connotative semiotic system.
101
 Resource guides

Figure 3.2 Stratification in protolanguage and in language, from Halliday (2003c: 13)

Halliday and Matthiessen (1999/2006) propose a version of Figure 3.1 designed to show that
ying/yang theory is in fact a theory of signs and so of stratification. Taverniers (2019) uses ver-
sions of Figure 3.1 to explore that nature of semantics in SFL.
We will now “scan” the resources of language in context from above, starting with context,
and then proceeding downwards from semantics to lexicogrammar within the content plane of
language, and then to phonology, phonetics within the expression plane and finally to graphol-
ogy and graphetics as alternative strata to phonology and phonetics within the expression plane.

3.1.2 Accounts of context

3.1.2.1 Pre-SFL accounts of context

Context has been part of SFL theory and descriptions from the start, having been “imported”
from Malinowski’s anthropological theory involving context of situation (e.g. Malinowski, 1923:
306 ff.) and context of culture (e.g. Malinowski, 1935: 17, 21–22, 51) into linguistics by Firth.
Malinowski’s account of context was data-based or even data-driven, growing out of his field-
work.1 He needed to develop a framework for dealing with – analysing and interpreting – texts
and other activities in Trobriand society, and he also needed such a framework to enable him to
translate texts into English to make them accessible (cf. Steiner, 2005b).
Firth focussed on Malinowski’s context of situation, proposing a schema for the analysis con-
texts of situation,2 introduced and described by Firth (1957c/1968: 177) in his discussion of “an
analytic dispersion of the statement of meaning at a series of levels” as follows:

No hard and fast lines can be drawn at present to form a strict classification for contexts
of situation. Some might prefer to characterize situations by attempting a description

102
 Resource guides

Figure 3.3 Stratification of the content and expression planes of language, with differentiation between
internal strata (“form”) and interface strata (“substance”), from Matthiessen (forthcoming b)

of speech and language functions with reference to their effective observable results,
and perhaps also with reference to a linguistically centred social analysis.
The technical language necessary for the description of contexts of situation is not
developed, nor is there any agreed method of classification. At this level there are
great possibilities of research and experiment. It will be maintained here that linguistic
analysis states the interrelations of elements of structure and sets up systems of ‘terms’
or ‘units’ and end-points of mutually determined interior relations. Such interior rela-
tions are set up in the context of situation between the following constituents:

1 The participants: persons, personalities and relevant features of these,

The verbal action of the participants.
The non-verbal action of the participants.
2 The relevant objects and non-verbal and non-personal events.
3 The effect of the verbal action.

No linguist has yet set up exhaustive systems of contexts of situation such that they
could be considered mutually determined in function or meaning. There is some
approximation to this, however, in Malinowski’s [1935] Coral gardens and their magic,
and here and there in special studies of contexts of personal address and reference, and

103
 Resource guides

of well-defined technological activities such as fishing or weaving or making war, and

of rituals of various kinds.

Firth was concerned with the context of situation of a text. In general, as Halliday has pointed
out, he tended not to venture beyond the generalized actual towards an overall system such as
the context of culture. (To deal with the daunting task of describing a culture, Malinowski,
1944, had suggested institutions as the way in.)

3.1.2.2 Into SFL

Firth’s construct was then re-interpreted by Halliday and colleagues in the 1960s as a level of
analysis embodying three parameters, field, tenor and mode3 (e.g. Halliday, McIntosh and Strev-
ens, 1964; Gregory, 1967).
The next step taken by Halliday (e.g. 1978) was to interpret context as a system of meaning,
a semiotic system. As a semiotic system, it was treated as being stratally “above” language and
other comparable semiotic systems serving as their semiotic environment. The relationship was
characterized clearly in Hjelmslevian terms by Martin (1992a): context is a connotative semi-
otic system realized through denotative semiotic systems including language, paralanguage,
kinesics, proxemics, pictorial systems – a number of which have since become the focus of
research in SFL (see Section 3.4). Halliday (1978) suggested that context and language resonated
with one another in functionally differentiated terms: field ~ ideational systems, tenor ~ inter-
personal systems, mode ~ textual systems.
The next step was to interpret the relationship between Malinowski’s context of situation and
context of culture theoretically in terms of the cline of instantiation; Halliday (1990a) inter-
preted them as being located at the outer poles of the cline of instantiation – context of culture
at the potential pole and context of situation at the instance pole, locating institutions and situa-
tion types between them. This depended on the theoretical dimension of the cline of instantia-
tion, which had already been explored as far back as Halliday (1973) in this characterization of
language in terms of potential and instance (as in can mean and means). This 1990 clarification in
terms of the cline of instantiation was indeed an important step. Malinowski (e.g. 1935: 51–52)
himself had not been very clear about the relationship,4 although there is an indication that con-
text of culture is a distillation of contexts of situation:

In the course of our analysis it has become increasingly clear that the contextual definition
of each utterance is of the greatest importance for the understanding of it, and that this
contextual reference must be two-fold. In the first place, an utterance belongs to a special
context of culture, i.e. it refers to a definite subject-matter. Each of the sayings, phrases
and narratives which I have here adduced belongs definitely to a certain division of our
Supplement and each such division corresponds to an aspect of Trobriand gardening.
But side by side with this context of culture or context of reference, as it might also
be called, we have another context: the situation in which the words have been uttered.
A phrase, a saying or a few sentences concerning famine may be found in a narrative,
or in a magical formula, or in a proverbial saying. But they also may occur during a
famine, forming an integral part of some of those essential transactions wherein human
beings co-operate in order to help one another. The whole character of such words is
different when they are uttered in earnest, or as a joke, or in a narrative of the distant
past. The words need not be idle in any of the cases. We have shown the function of
narrative. Even a joke about a serious subject may do its part in begetting a traditional

104
 Resource guides

attitude, an attitude which in the long run might prove of considerable significance in
tribal life, and this is the most important result of an utterance from the point of view
of a scientific theory of meaning.

But he predicted the path of development taken within SFL; Malinowski (1944: 5) wrote:

Culture, as the widest context of human behavior, is as important to the psychologist

as to the social student, to the historian as to the linguist. I submit that the linguistics
of the future, especially as regards the science of meaning, will become the study of
language in the context of culture.

Halliday (1990a) appeared at the end of a very active period of exploring the organization
of context by J. R. Martin and his group, at that stage mostly in the intellectual environ-
ment of educational linguistics. Dealing with educational tasks, their approach had been to
stratify context instead of extending it as one stratum along the cline of instantiation; and
this became part of the “Genre Model” in Martin’s “Sydney School”,5 documented in detail
after a very productive decade by Martin (1992a: Chapter 7). There have also been other
suggestions for further work on context that difer in one way or another from the overview
presented in this section; see e.g. Bartlett (2017). We will discuss the options taken in SFL
in the theorization of context in Section 5.2.1; they are reflected in overviews of contexts,
including:

• Ghadessy (1999): an edited volume with chapters covering different aspects of context,
including dynamic modelling.
• Butt and Wegener (2007): an overview of context in SFL up through the mid-noughts,
going back to Malinowski’s field-based framework, Firth’s adaptation of his insights, and
interpretation by Mitchell’s description of service encounters in Cyrenaica, and the founda-
tional work in SFL in the 1960s (Halliday, Ellis, Ure). After this brief overview, they adopt
a historical perspective: “we set out by investigating the early proposals concerning context,
with some interpretive excursions into the nineteenth century.” As part of this historical
review, they set out details of “Wegener’s typology of context”.
• Hasan (2009a) is “concerned primarily with an examination of the category of context in
Halliday’s systemic functional linguistics”. She traces the development of this category back
to Malinowski and Firth, and locates context within the overall architecture of language,
based on Halliday (1991d), and summarizes a number of her own contributions, like her
notions of material situational setting, relevant context and contextual configuration (the
setting of systemic values in terms of field, tenor and mode), the relationship between con-
text and text structure – and her systemic description of field, tenor and mode. She also
refers to Saussure, Weinreich and Labov.
• Bowcher (2010) “outlines the history and theoretical development of the concept of context
of situation in Systemic Functional Linguistics”. She begins by identifying different uses of
the term “context”, summarizes the work in the “early years” and comments on context
in linguistics also outside SFL, including Hymes’ Ethnography of Speaking (e.g. Hymes,
1967). Against this background, she explores the integration of context into systemic func-
tional theory, including the “context-metafunction hook-up hypothesis” (Hasan’s term)
and the location of context in terms of instantiation and realization. In addition, she reviews
certain systemic functional descriptions by means of system networks of the field parameter
within context. See further Bowcher (2017, 2019).

105
 Resource guides

• Wegener (2011) presents a PhD thesis entirely devoted to context: it “theorises context
by setting out from the historical genesis”, sets forth “a study of modelling context”,
and “applies a model of context”, in a particular institutional setting, viz. “emergency
care for the purpose of understanding communication within the hospital environ-
ment”. In addition to these substantial contributions, the thesis includes four appendi-
ces that among other things provide an overview of systemic functional descriptions of
context networks developed by Ruqaiya Hasan and David Butt and publications relat-
ing to different aspects of context by Wegener and co-authors. Among the conclusions,
Wegener suggests that there are difficulties with system network representations of
contexts, and as a way forward she points to “A person centred approach to contextual
descriptions” since this approach “appears to offer a useful way to incorporate cogni-
tive aspects of meaning making yet it remains to be seen what impact this will have
on modelling and theory and how best these individual differences can be represented
for different audiences”. This could be related to the work on computational models
informed by SFL, both the “Penman tradition”, with its various offshoots, and the
work by Robin Fawcett and his Cardiff team – and importantly also Bateman’s (1985)
PhD thesis on “utterance in context”, and his theoretical emphasis on “intersubjective
achievement in discourse”.
• Bowcher (2017) reviews accounts of field, tenor and mode as “a conceptual framework for
describing the context of situation – that is, they represent the parameters across which the
relevant features of a situation in which language is involved may be grouped”.
• Tann (2017) and Gardner (2017) complement the contributions by Hasan, Butt, Bow-
cher and others who focus the interpretation, theory and description of context along the
Halliday-Hasan trajectory; they review contributions by Martin and his group within what
they have called the “Sydney School”.
• Tann (2017) provides an up-to-date account of the approach to context taken by Mar-
tin (e.g. 1997, 2010) and his group, often under the heading of the “Sydney School”
(e.g. Rose and Martin, 2012).
• Gardner (2017) presents genre as a contextual stratum within the “Sydney School”,
indicating the reasons given for differentiating genre and specifications of field, tenor
and mode, called “register” in the model.
• Bartlett (2017) traces the development of accounts of contexts within SFL, criticizing
certain aspects of Halliday’s account, claiming that: “Halliday’s classical Marxist orienta-
tion entailed an intellectual programme to correlate variations in linguistic behaviour with
variations in social structure (specifically, class and situation type), which has resulted in
an overly deterministic/congruent modelling of the relationship between contextual and
linguistic features, and which underplays divergence, diversity and the unique socialisation
of the individual” and sketching his own proposal – “a scalar supervenience conception
of context” (see also Section 5.2.2.2). Along the way, he relates the discussion not only to
Malinowski and Firth, but also to fellow travellers – Bernstein, within sociology, and to
scholars outside this tradition, like Vygotsky.

Meanwhile, we will continue to deal with the model developed by Halliday since the 1960s:
Halliday, McIntosh and Strevens (1964), Halliday (1978, 1990a, 2002a), Hasan (2009a), Mat-
thiessen (2015c).

106
 Resource guides

3.1.2.3 Functional diversification: field, tenor and mode

The parameterization of context into the three systemic parameters of field, tenor and
mode has remained quite stable since the 1960s. The term “style” was used in Halliday, McIn-
tosh and Strevens (1964) but was replaced by “tenor”, taken from Spencer and Gregory (1964),
but the basic parametric framework has continued to serve as a foundation for further accounts
of context (for some variation, see Table 5.2). In this respect, work on context is similar to SFL
in general: SFL has developed in an evolutionary way, not in a revolutionary fashion; it has
developed by expanding the territory for which it is responsible, not by working over essentially
the same territory with a succession of accounts, each new version replacing the previous one.
Thus while many descriptions of field, tenor and mode have been added, in particular with a
focus on “English” context, the theoretical framework has arguably stood the test of time, also
supporting descriptions of contexts in which languages other than English operate (as in Rose,
2001; Thomson, Sano and de Silva Joyce, 2017) and applications in various institutional settings.
The theory has not, of course, remained static; it has been expanded along different lines, e.g.
in work explicitly relating the characterization of tenor to work on role networks in sociology.
Interestingly, the parametric account of context was in place in SFL before Halliday devel-
oped his theory of linguistic metafunctions.6 The interpretation of context in terms of field,
tenor and mode came first, and it constituted an ecological approach to context – one that was
sensitive to the extrinsic functions or uses of language (cf. Halliday, 1985d; Martin, 1991). Then
Halliday realized that the theory of metafunctions that he had developed independently based on
this observations of the clusterings of systems in his systemic description of the lexicogrammar
of English (starting with the “Bloomington grammar”) enabled him to observe and articulate
resonances or correspondences between the contextual parameters of field, tenor and mode
and the metafunctions – ideational, interpersonal and textual. He presented this “hook-up”, as
Hasan has called it (cf. Hasan, 1995b),7 systematically in Halliday (1978).
But let us focus here on the descriptions put forward in the SFL literature of contexts in
terms of field, tenor and mode. We will begin with Halliday’s (1978) general characterization
of field, tenor and mode – formulated as what we might call pre-systemic glosses since they had
not yet been turned into system networks (Table 3.2). Then we will turn to what we might call
discursive glosses characterizing contexts of situation of particular texts (Table 3.3), and then say
something about attempts at developing systemic descriptions of field, tenor and mode at, or
close to, the potential pole of the cline of instantiation8 (Table 3.4).
Halliday (1978: 142) introduces his characterization of field, tenor and mode by reference to
the (context of) situation of a text, interpreting it semiotically, i.e. as made of meanings:

We shall be able to show something of how the text is related to the situation if we can
specify what aspects of the context of situation ‘rule’ each of these kinds of semantic
option. In other words, for each component of meaning, what are the situational fac-
tors by which it is activated?

The question then becomes one of characterizing the context of situation in appropriate terms,
in terms which will reveal the systematic relationship between language and the environment.
This involves some form of theoretical construction that relates the situation simultaneously to
the text, to the linguistic system, and to the social system. For this purpose we interpret the situ-
ation as a semiotic structure; it is an instance of the meanings that make up the social system.9

107
 Resource guides

Actually it is a class of instances, since what we characterize will be a situation type rather than
a particular situation considered as unique.
(In Halliday, 1990a, situation and situation type are located relative to one another along the
cline of instantiation as instance and instance type; and by another along the cline of instan-
tiation, situation types are related to institutions, which in turn are related to the context of
culture. See Section 3.1.2.4.) Halliday (1978: 142–143) then goes on to say that the situation
“consists of ” (i) field – “the social action”, (ii) tenor – “the role structure”, and (iii) mode –
“the symbolic organization”. His characterizations of these three parameters are quoted in the
first row of Table 3.2.
After providing the characterizations of field, tenor and mode in the first row of Table
3.2, Halliday elaborates on these three parameters (pp. 143–145) and makes a distinction of

Table 3.2 Halliday’s (1978) general pre-systemic characterization of field, tenor and mode

FIELD TENOR MODE Reference

the social action: that the role structure: the cluster the symbolic Halliday
which is ‘going on’, and of socially meaningful organization: the (1978:
has recognizable meaning participant relationships, particular status that 142–143)
in the social system; both permanent attributes is assigned to the text
typically a complex of of the participants and role within the situation; its
acts in some ordered relationships that are specific function in relation to
configuration, and in to the situation, including the social action and the
which the text is playing the speech roles, those that role structure, including
some part, and including come into being through the channel or medium,
‘subject-matter’ as one the exchange of verbal and the rhetorical mode.
special aspect; meanings;
[social action:] play, verbal [power:] equal, authority, . . . –
art, entertainment, [expertise:] expert-novice, GENERAL
instruction, disseminating [familiarity:] intimate, SYSTEM
(information), unknown, . . . [1] first-order
transmitting [institutional role:] parent- (“social”)
(knowledge), exploring child, mate-mate,
(knowledge), manageress-sales person,
maintenance of teacher-pupil, pastor-flock,
institution, regulation individual-society, story
of social interaction, teller-audience, stand-up
commercial transaction, comedian-audience, seller-
professional consultation, buyer, doctor-patient, . . .
[‘subject matter’:] [speech-role:] [medium:] written (to be [2] second-
theme & thesis (‘plot’) – questioner-answerer, read privately, to be order (also
domains of experience questioner-suggester, read aloud, . . .), spoken “semiotic”)
brought into existence advisor-advisee, . . . (scripted, unscripted, . . .);
through language [turn-raking:] dialogic,
monologic
[rhetorical mode:] essayistic,
narrative, exploratory,
didactic, explanatory,
persuasive, exhortatory,
polemic, . . .

108
 Resource guides

fundamental importance. This is the distinction between 1st-order and 2nd-order. First-order
categories exist independently of language and may be manifested directly in the social system,
but second-order ones are “defined by the linguistic system” – and, we might add, by the
systems of other denotative semiotics. First-order categories are set out in the second row of the
table, and second-order categories in the third row of the table. While field and tenor have both
first- and second-order categories, mode exists only as a second-order parameter, as Halliday
(1978: 145) explains:

All the categories under this third heading are second-order categories, in that they are
defined by reference to language and depend for their existence on the prior phenom-
enon of text. It is in this sense that the textual component in the semantic system was
said to have an ‘enabling’ function vis-à-vis the other two: it is only through the encod-
ing of semiotic interaction as text that the ideational and interpersonal components of
meaning can become operational in an environment.

To get a sense of how the characterizations of field, tenor and mode can be used in the
analysis of the contexts of situation of particular texts, we can review a number of examples:
see Table 3.3. While not systemic in the sense of being based on systemic descriptions of the
cultural potential (“context of culture”) of a community, these characterizations are often very
perceptive and insightful, and a good source of inspiration for researchers trying to produce
descriptions of contexts of situation. For example, one can scan them to identify and classify
the kinds of contextual properties that contextual descriptivists have been alert to, one can
look for contexts of situation that are close to one under one’s own investigation, and one
can draw on them to try to move towards general systemic descriptions of the field, tenor and
mode potentials of a culture (while remembering Malinowski’s, 1944, advice about institu-
tions as the true isolates of culture).
So far we have given general characterizations of field, tenor and mode (Table 3.2) and
discursive glosses specifying particular contexts of situation discussed in the SFL literature
(Table 3.3). However, one of the long-term goals in SFL is, naturally, to develop systemic
descriptions of field, tenor and mode at the potential pole of the cline of instantiation – in
other words, to describe the cultural potential of a community. This is indeed a daunting
task – one that US American anthropologists and anthropological linguists in a sense also
targeted in the 1950s (e.g. Hall and Trager, 1953, who sketched categories for the analysis
and description of culture). Given the overwhelming nature of the task, Malinowski (1944)
recommended moving in through institutions; he writes (p. 54):

No element, “trait,” custom, or idea is defined or can be defined except by placing it

within its relevant and real institutional setting. We are thus insisting that such insti-
tutional analysis is not only possible but indispensable. It is maintained here that the
institution is the real isolate of cultural analysis. It is also maintained that any other type
of discussion or demonstration in terms of isolated traits or trait complexes, other than
those which would follow the institutional integration, must be incorrect.

This is indeed a viable research programme for the development of descriptions of contexts of
culture. However, some systemic functional scholars have proposed system networks for field,
tenor and mode, as shown in Table 3.4. Wegener (2011: Appendix 1, 223–241) provides an
overview of the descriptive proposals put forward by Hasan in a number of publications and,
building on her work, by Butt (2004). Bowcher (2014) discusses issues that arise in the work

109
 Table 3.3 Examples of discursive descriptions of the contexts of situations of texts and of the situation types of registers

Field Tenor Mode Reference

(a) Verbal art: entertainment through (a) Writer and readers: writer adopting role Text as ‘self-sufficient’, as only form of Halliday (1978: 146) –
story-telling as recounter: specifically as humourist social action by which ‘situation’ is Thurber’s The lover and his
(b) (i) Theme: human prejudice (partly projected through subsidiary role defined. lass
(‘they’re different, so hate them!’). as moralist), and assigning complementary Written medium: to be read silently as
Projected through: role to audience. private act.
(ii) Thesis (‘plot’): fictitious (b) Mate and mate: animal pair as projection Light essay; original (newly-created) text
interaction of animals: male/female of husband and wife; each adopting own projected onto traditional fable genre,
pairs of hippopotamuses, parrots. (complementary) role as reinforcer of structured as narrative-with-dialogue,
shared attitudes. with ‘moral’ as culminative element.
Child at play: manipulating movable Small child and parent interacting: Spoken, alternately monologue and Halliday (1978: 115) –
objects (wheeled vehicles) with related child determining course of action, dialogue, task-oriented; pragmatic, (i) Father and child playing
fixtures, assisted by adult; concurrently (i) announcing own intentions, referring to processes and objects of
associating (i) similar past events, (ii) (ii) controlling actions of parent; situation, (ii) relating to and furthering

Resource guides
similar absent objects; also evaluating concurrently sharing and seeking child’s own actions, (iii) demanding
110

objects in terms of each other and of corroboration of own experience with other objects; interposed with narrative
processes. parent. and exploratory elements.
Instruction: the instruction of a novice – Equal and intimate: three young adult males; Spoken: unrehearsed Halliday (1978: 226)
in a board game (e.g. Monopoly) acquainted Didactic and explanatory, with undertone
with equipment present – but with hierarchy in the situation of non-seriousness
– for the purpose of enabling him to (2 experts, 1 novice) – with feedback: question-and-answer,
participate – leading to superior-inferior role correction of error
relationship
A general, imaginary problem of verbal Adult and three children: adult (neither Informal spontaneous speech. Dialogue: Halliday (1985b: 52–53)
behaviour: how to refer to a baby teacher nor parent) interviewer, informal; question – and – answer exchanges. “Dimensions of Discourse
whose sex is unknown, without children self-conscious but relaxed. Exploratory; hypothetical and logical analysis” – discussion
offending against the parents, the baby Speech roles: adult questioning, children in orientation, moving towards (partly among 1 adult and 3
(later in life), or the language suggesting. humorous) resolution. nine-year-olds
Maintenance of institutionalized system Authority (in both senses, i.e. person holding Written to be read aloud; public act Halliday and Hasan (1980:
of beliefs; religion (Christianity), and authority, and specialist) to the audience; (mass media: radio); monologue; text is 40) – Radio talk by
the members’ attitude towards it; audience unseen and unknown (like whole of relevant activity Bishop Woolwich
semi-technical readership), but relationship institutionalized Lecture; persuasive, with rational argument
(pastor to flock)
 Verbal regulation of social interaction ‘Member’ (individual) addressing ‘collective’ Written to be filed (i.e. to form part of Halliday and Hasan (1980:
through sanctions of the legal (society) using formula prescribed by documentary records): text gives status 13) – Transfer of whole
system: codification of exchange of collective for purpose in hand (as social act) to non-verbal transaction; (freehold or leasehold)
property (‘deed of transfer’), including text is formulaic (i.e. general, with
certification that transaction falls within provision for relating to specific instances)
particular class of transaction defined Performative (i.e. text constitutes, or
by value of commodity exchanged ‘realizes’, act in question)
(a) General. Retail selling in department Manageress and new salesgirl; a complex Natural, spontaneous speech. One- Halliday (1985a: 370) – The
store: silver department. Task: selling status relationship embodying (a) senior- sided dialogue (monologue with silver text
silver ware. junior, (b) expert-novice, (c) teacher- acknowledgements), Part 1, expository:
(b) Specific. Instruction to new member. apprentice; with a fourth, personal exposition – doubt – explanation. Part
Task: teaching how to sell silverware. relationship at a metaphorical level, 2, exhortatory: injunction – doubt –
Means of achievement: [premise 1] (d) mother-daughter. illustration and reassurance.
virtues of silver, [premise 2] customers’

Resource guides
appreciation thereof, [action]
encouragement of this appreciation.
111

extending, exploring or transmitting addressed to specialists, to learners or to phonic or graphic channel, most Halliday (1988a) – register:
knowledge in the physical, biological or laymen from within the same group (e.g. congruent (e.g. formal ‘written scientific English [“a
social sciences specialist to specialist) or across groups (e.g. language’ with graphic channel) or less semiotic space within
lecturer to students) so (e.g. formal with phonic channel), which there is a great deal
and with variation in rhetorical of variability”]
function – expository, hortatory,
polemic, imaginative, and so on
Keeping people informed: disseminating Expert to laity, but with only moderate Written: print: prose format. Halliday and Matthiessen
information by mass media: print: distance. Rhetorical mode: informative; (1999/2006) “Construing
newspaper: daily. Expert standpoint: impersonal, with semi-technical. experience through
State of natural environment: natural acknowledgment of uncertainty; attitude Accompanied by other visual modes: meaning”
phenomena: weather – present to subject-matter neutral. maps, with special (technical) symbols. – register: weather
conditions and prognostications for Audience: interested general public, plus some Day-to-day continuity, with date and forecasting
immediate future (0–2 days), specifying groups of special concern (whose actions place of origin specified.
locality (varying focus, from district to may be affected by the information).
world, with emphasis on ‘region’). Monologue.
(Continued)
 Table 3.3 (Continued)

Field Tenor Mode Reference

Special features: (1) explanations of

events; (2) warnings of disaster;
(3) information for special interest
groups: those engaged in agriculture
and gardening, boating and fishing,
aviation, sport.
professional consultation: medical, client: patient-applicant and agent for channel: aural – visual contact: telephone Halliday and Hasan
application for appointment . . . consultant: receptionist; maximum social conversation; (1980: 23) – Making an
distance . . . spoken medium . . . appointment
economic transaction: purchase of retail agents of transaction: salesman-customer; channel: aural + visual contact; Halliday and Hasan (1980:
goods: perishable food . . . social distance: near maximum . . . spoken medium 78) – Service encounter

Resource guides
object activity (second order field) a hierarchy of social relations in the the technologies associated with a radio Bowcher (1999: 168) –
which largely constitutes the subject commentary box, near maximal social commentary, process-sharing between ABC rugby league radio
112

matter distance between the commentator and the prespecified participants, play-by-play commentary
other speaking participants, maximal social commentary and ‘colour’ commentary
distance between speaking participants creating a type of monologue presented
and the listeners, non-negotiability of to a mass of unseen listeners
the number and type of participants and
the extent of participation, along with
little individual freedom to enter and exit
the commentary by those who verbally
participate in the social process
 Resource guides

Table 3.4 Systemic descriptions of field, tenor and mode in terms of system networks (note in addition,
Butt, 2004, based on Hasan’s description and referred to in the SFL literature but only available
in manuscript form)

Field Tenor Mode Reference

transmission: oral (doing)/ status: visual: none/one-way/ Martin (1992a: Ch. 7)

written (studying) equal/ two-way and
oral: domestic/specialised unequal aural: none/one-way/
[recreational/trades] contact: two-way
written: administration/ involved/ none & none: public/
exploration distant private and
affect: visually/aurally objectified
marked field-structured
[positive/ [abstraction:
negative]/ accompanying/
not marked constituting]/
genre-structured
projection: projected/
unprojected
material action: present/ Hasan (1999a) – field
non-present [deferred/ and mode descriptions
absent] explored by Bowcher
verbal action: ancillary/ (2014, 2017) and viewed
constitutive [practical/ in terms of multimodal
conceptual] analysis by Bowcher
sphere of action: (2007). Wegener (2011:
specialised/quotidian Appendix 1, 223–241)
[institutional/individuated] presents a summary
iteration: stop/go of contextual system
[independent/aligned/ networks by Hasan
integrated] (1999a) and Butt (2004).

towards descriptions of field and mode, focussing on the role of language, ancillary, constitutive
and material action (see also Bowcher, 2017, 2019, and cf. Bartlet, 2019).10 Matthiessen (2015b)
presents a fairly detailed description of field of activity, drawing on Jean Ure’s unpublished work;
this description has also been presented in various other publications, including Matthiessen
(2009a), Matthiessen, Teruya and Lam (2010) and Matthiessen and Teruya (2015a), and it is
summarized in Halliday and Matthiessen (2014).

3.1.2.4 Context extended along the cline of instantiation

As shown in Figure 1.5, the relationship between context and language extends along the cline
of instantiation from the potential pole to the instance pole, i.e. from the language system, or
meaning potential, in the context of culture to the acts of meaning making up a text in its con-
text of situation (see Figure 3.11 in Section 3.2.1). This is brought out when stratification and
instantiation are intersected in a stratification-instantiation matrix, as shown in Figure 3.5.
Naturally, the extension of context along the cline of instantiation applies to all three functional
parameters; field, tenor and mode are extended from the context of culture at the potential pole
of the cline to all the contexts of situation at the instance pole.

113
 Resource guides

The relationship between context and semantics is highlighted by Halliday’s (1975c/2003b:

79–80) discussion of speech act theory:

When the social context has been idealized out of the picture, a theory of speech acts
provides a means of putting it back again. It celebrates the linguists’ rediscovery that not
only do people talk – they talk to each other.
The study of speech acts, which as Searle remarks is important in the philosophy of
language, starts from the speaker as an isolate, performing a set of acts. These include,
among others, illocutionary acts – questioning, asserting, predicting, promising and
the like; and illocutionary acts can be expressed in rules. In Searle’s (1965) words,
The hypothesis . . . is that the semantics of a language can be regarded as a series of
systems of constitutive rules [i.e. rules that constitute (and also regulate) an activity
the existence of which is logically dependent on the rules] and that illocutionary
acts are performed in accordance with these sets of constitutive rules.
The meaning of a linguistic act thus comes within the scope of philosophical enquiry.
But a language is not a system of linguistic acts; it is a system of meanings that defines
(among other things) the potential for linguistic acts. The choice of a linguistic act –
the speaker’s adoption, assignment, and acceptance (or rejection) of speech roles – is
constrained by the context, and the meaning of the choice is determined by the con-
text. Consider a typical middle-class mother and child exchange: Are you going to put
those away when you’ve finished with them? – Yes. – Promise? – Yes. The second of these
yesses is at one level of interpretation a promise, a concept which (in Searle’s now classic
demonstration) can be explained by reference to conditions of three types: preparatory
conditions, the sincerity condition and the essential condition. But its significance as
an event depends on the social context: on modes of interaction in the family, socially
accepted patterns of parental control, and so forth – and hence on the social system,
Malinowski’s “context of culture”. To describe the potential from which this utterance
derives its meaning, we should need to specify such things as (sub-culture) professional
middle class, (socializing agency) family, (role relationship) mother-child, (situation
type) regulatory, (orientation) object-oriented; and to interpret it as, at one level, a
move in a child’s strategy for coping with a parent’s strategy of control. (How this
may be done can be seen from the work of Bernstein and Turner, from one view-
point, and of Sacks and Schegloff from another.) We shall not want to say that the
child’s utterance is ‘insincere’; but nor shall we want to interpret it as ‘one speech act
conveying another’, which introduces an artificial distinction between a speech act
and its use, as if to say “As an idealized structure, this is a promise; when it is instanti-
ated, by being located in a social context, it functions as something else”.

We have quoted Halliday at some length because his account illustrates a key point about meaning
in context in relation to the cline of instantiation – showing how an instantial act of meaning must
be understood against the background of contextual settings further up the cline of instantiation
towards the potential pole, and because it provides information relevant to later developments
of the interpersonal semantics of speech function and negotiation in relation to the contextual
parameter of tenor, e.g. Halliday (1984b), Martin (1992a), Hasan (1996a), Hasan et al. (2007),
and it is also relevant to later exploration of appraisal, afliation and the like (e.g. Knight, 2010a,
2010b). Meanings are diversified and even multiplied according to the vast array of options inher-
ent in context as system11 – and importantly as a systemic assembly of sub-cultural systems. One

114
 Resource guides

line of exploration that reflects this insight is the development of semantic descriptions of subsys-
tems of the overall meaning potential that are resonant with particular situation types somewhere
midway along the cline of instantiation, as in Halliday’s (1972) example of maternal regulatory
semantics.
In Halliday’s example, the choices in meaning made by mother and child are instantial and
they are made in an instantial context, the context of situation of the dialogue between them.
But they make them against the background of linguistic and cultural organization higher up
the cline of instantiation, towards the linguistic meaning potential and the context of culture or
cultural meaning potential. As far as culture is concerned, it is not treated as uniform or homog-
enous; Halliday points out that the example illustrates a particular sub-culture, that of “profes-
sional middle class”; and the other settings of contextual parameters also point to particular
values within the overall cultural system, as a system of systems.
As already noted, the functional contextual parameters of field, tenor and mode extend along
the cline of instantiation from potential to instance. At the potential pole, they constitute simul-
taneous regions within the cultural potential of a society, exemplified in Table 3.4, all the diver-
sity characteristic of any given society. At the instance pole, they are reflected in the “glosses”
on the contexts of situation of particular texts, listed in Table 3.3. While these may appear to be
disconnected, a central part of the long-term project of developing descriptions of the contexts
of culture of different societies (cf. Rose, 2001; Thomson, Sano and de Silva Joyce, 2017) is
precisely to develop systemic descriptions of field, tenor and mode within the cultural potential
of societies and to relate them to selections within their systems in the characterizations of the
contexts of culture of particular texts.
This is a truly daunting task – and it will help to use institutions as domains within a culture
as gateways, as suggested by Malinowski (1944), indicated in the quote in Section 3.1.2.3, but
let us give a hint of an indication of what is involved by referring to Halliday’s (1984a/2002d:
313–322) example of a dialogic passage between father and son:

Nigel at 5;4 (from Halliday (1984c))

NIGEL: Shall I tell you why the North Star stays still?
FATHER: Yes, do.
NIGEL: Because that’s where the magnet is, and it gets attracted by the earth; but the other stars
don’t so they move around.

He provides a fairly comprehensive semantic, lexicogrammatical and prosodic analysis of this

passage, and in addition he “glosses” the context of situation of the text in terms of field and
tenor. But he also characterizes the field and the tenor parameters in terms of the context of
culture – the cultural system at the potential pole of the cline of instantiation: see Figure 3.4.
Here it is clear that the “glosses” used to characterize the field and tenor selections of the con-
text of situation are ultimately derived from the specification of the field and tenor parameters
of the context of culture. For example, within tenor, the interaction between father and son
instantiates the parent-child role dyad within the institution of the family, more specifically the
subtype of family characteristic of “class: middle class: intellectual”. This highlights the fact that
many – even most in today’s world – contexts of culture are highly diversified and heterogenous.

3.1.3 Accounts of semantics

In SFL, semantics is theorized, like all aspects of language in context, in terms of the relation-
ships it enters into and serves as the domain of, as is brought out by Taverniers (2019). It is

115
 Resource guides

Figure 3.4 The context of situation of the “North Star” text and its relationship to the context of culture

Figure 3.5 Halliday’s (2002a) stratification-instantiation matrix

116
 Resource guides

theorized as the higher of the two content strata, the stratum of meaning (in a narrow sense [cf.
Table 6.2; Taverniers, 2019]; see also Section 6.2.1.4 on distinct but related senses of “mean-
ing”). It is realized by the stratum of lexicogrammar, the stratum of wording, and it serves as
an interface to what lies beyond language – context, of course, and also the content of other
semiotic systems operating alongside language in context (including bio-semiotic systems, i.e.
sensorimotor systems). The stratal location of semantics in relation to language and context and
other semiotic systems is diagrammed in Figure 3.6. This figure shows that:

• semantics forms the content plane of language together with lexicogrammar. The two
content strata, content1 and content2, are called semantics and lexicogrammar, but they are
both concerned with content, and derived semogenetically from a split of the content plane
into these two strata, as Halliday (1975a, 2003a) has shown in his account of ontogenesis.
This split augments the potential for patterns of agnation within the content plane, includ-
ing ones involving metaphor (see the next item), as shown for the semantics and grammar of
dialogue by Halliday (1984b). The relationship between semantics and lexicogrammar is the
general stratal one of realization, as illustrated by e.g. Halliday (1972, 1984b), Hasan (1996b).
In computational work, this relationship has also been modelled by means of the “chooser-
&-inquiry” framework (e.g. Matthiessen, 1988a, 1990; Matthiessen and Bateman, 1991);
• the stratal boundary between semantics and lexicogrammar is a natural one rather than a
conventional (or “arbitrary”) one, represented by the dotted circle in Figure 3.6 (unlike the
boundary between lexicogrammar and graphology); and this makes possible the emergence

Figure 3.6 The stratal location of semantics in relation to other linguistic strata, context and other semiotic
systems

117
 Resource guides

of metaphor, expanding on the congruent pattern of realization between semantics and

lexicogrammar. Metaphor extends throughout lexicogrammar, from grammar to lexis; in
addition to the traditional notion of lexical metaphor (also known as “conceptual meta-
phor”, in the strand of cognitive linguistics drawing on Lakoff and Johnson, 1980), gram-
matical metaphor has been recognized and studied (e.g. Halliday, 1988a, 1998c; Halliday
and Martin, 1993; Derewianka, 1995, 2003; Halliday and Matthiessen, 1999/2006: Chap-
ter 6; Simon-Vandenbergen, Taverniers and Ravelli, 2003);
• of the two content strata, lexicogrammar is internal to language and is realized by the
internal stratum of the expression plane, viz. phonology (or graphology), and semantics is
the interface between language and what lies outside language, providing the strategies for
transforming this into meaning in language (e.g. Halliday, 1973);
• what lies outside language is on the one hand context as a connotative semiotic system,
i.e. one that is parasitic upon other semiotic systems in that it uses them as their expression
plane, and on the other hand denotative semiotic systems other than language (both
content in other social semiotic systems and content in biosemiotics systems);
• through semantics, language thus “resonates” with its semiotic environment – i.e. with
context, and this resonance is differentiated functionally according to the functional param-
eters of context (field, tenor and mode) and the metafunctions of language (ideational,
interpersonal and textual). This resonance is covered by the general notion of inter-stratal
realization (shown by diagonal arrows in Figure 3.6). Since it is functional in orientation,
semantics must be seen “from above”, from the point of view of contexts, and this means
among other things that the basic unit of semantics is language functioning in context – i.e.
text (e.g. Halliday and Hasan, 1976; Halliday, 1981);
• semantics also interfaces with meaning in other denotative semiotic systems including
systems that may accompany language such as gaze, facial expression and gesture in face to
face interaction, images of various kinds in printed documents, but also systems that often
operate on their own such as dance, music, architecture and sculpture. Meanings in these
other semiotic systems can be construed by the semantic systems, but certain meanings in
the semantic system can also be “translated” into other semiotic systems (cf. Mohan, 1986);
• at the same time, semantics also interfaces with bio-semiotic systems (Halliday and Mat-
thiessen, 1999/2006); it construes meanings created by our perceptual systems and it enacts
linguistic meanings as motor programmes. For example, when we see a car moving to the
right around a corner, we can construe it as the driver turned right at intersection, and when
we are told turn left at the intersection, we can enact this command as a motor programme
appropriate to our means of transportation.

Publications dealing with the diferent aspects of the stratal location of semantics just enumer-
ated are specified in Table 3.5. The publications included in the table are not exhaustive, but
even if more publications had been included it would still remain the case that the stratal rela-
tions between semantics and other systems need a great deal more work, including descriptive
accounts such as Hasan’s (1984a) investigation of the semantic realization of the Placement ele-
ment in the structure of the narrative context of nursery tales.
These observations about semantics in SFL derive from its stratal location, as shown in Figure
3.6. At the same time, like language in context in general, semantics is extended along the cline
of instantiation (shown in Figures 3.1–3.5):

• At the potential pole of the cline of instantiation, semantics is “manifested” as the meaning
potential of language – its semantic system (see Figure 3.10).

118
 Resource guides

Table 3.5 Works dealing with the different aspects of the stratal location of semantics

Stratum Language Other denotative semiotic systems

context relationship between context and

semantics:
realization of Placement in nursery tale:
Hasan (1984a);
general tendencies (according to functional
diversification): Martin (1992a)
semantics internal organization of semantic relationship to other denotative
system: semiotic systems:
see function-rank matrix in Table 3.6 Halliday and Matthiessen (1999/2006);
Bateman et al. (2010);
Matthiessen (2015e);
(implied in) accounts of multimodality:
Martinec (2005);
Kress (2010, 2012); Bateman, Wildfeuer
and Hiippala (2017); and to biosemiotics
systems: García and Ibáñez (2017)
lexicogrammar relationship between semantics and –
lexicogrammar:
realization in message semantics: Hasan
(1996b), Hasan et al. (2007)
realization in register-specific semantic
systems: Halliday (1973), Patten (1988)

Table 3.6 Accounts of semantics according to the cline of instantiation

Potential Subpotential/instance type Instance

semantic system: register/text type: text:

see function-rank matrix register-specific semantic systems: text as semantic choice:
in Table 3.7 sociosemantic semantics, regulatory semantics: Halliday (1977b); Hood
Halliday (1972), and in computational and Forey (2008)
modelling: Patten (1988)
coding orientation: Hasan (1989, 2009b)

• At the instance pole of the cline of instantiation, it is manifested as acts of meaning that
make up texts unfolding in time.
• At the region intermediate between potential and instance, it is manifested as subsystemic
patterns characterized as registers (from the point of view of the potential pole) or as text
types (from the point of view of the instance pole).

Examples of studies dealing with semantics within diferent regions of the cline of instantiation
are given in Table 3.6.
Halliday (1972) developed the notion of a sociosemantic system network in the context
of the work led by Basil Bernstein, also involving Geoffrey Turner (cf. Turner, 1973, 1987) and

119
 Resource guides

Bernard Mohan. He emphasized the nature of semantics as an interlevel – an interface between the
rest of language and what lies outside language (as far as the content plane of language is concerned).
To demonstrate this approach, he viewed semantics “from above”, from the point of view of a situ-
ation type within context, choosing as an example of a regulatory context, more specifically one
where a mother needs to exercise semantic control over a young boy’s behaviour, warning or threat-
ening him. Thus his interpretation of semantics was as a strategic resource – as options in meaning
for getting something done. Curiously, while other systemic functional linguists have taken note of
this pioneering work (see e.g. Hasan et al., 2007: 704–707), there has been little further develop-
ment of other semantic descriptions of registers as strategies operating in particular situation types.
As the stratum of meaning, semantics is metafunctionally diversified into different modes
of meaning, each with its own characteristic mode of expression (e.g. Halliday, 1979a, 1981;
Martin, 1992a, 1996a; Halliday and Matthiessen, 1999/2006; Taverniers, 2019):

• ideational: construing our experience of the world inside and around us as meaning, mod-
elled either as configurations (experiential) or as series (logical);
• interpersonal: enacting our roles and relations as meaning;
• textual: engendering ideational and interpersonal meanings as a flow of information (dis-
course) in context.

This metafunctional spectrum of meanings can be intersected with the internal compositional
hierarchy of semantics, or semantic rank scale, to define a semantic function-rank matrix, as
shown in Table 3.7. In the matrix, we have also listed key publications dealing with the par-
ticular region of the semantic system viewed in terms of metafunction and rank. (This list of
accounts of semantics is supplemented by works focussed on semantic system networks orga-
nized chronologically: see Table 3.8.)

Table 3.7 Semantic function-rank matrix with references to a selection of publications presenting descri-
ptions of English semantics

Rank – domain General Ideational Interpersonal Textual

Logical Experiential

terms used in Ideation base Interaction base Text base

Halliday and
Matthiessen’s
(1999/2006)
account of the
meaning base
text Martin (1992a); logico-semantic macro-proposal: macro/
Martin and organization: Martin (1992b); hyper-
Rose (2003) Matthiessen appraisal: Martin Theme
[written (forthcoming b) and White ^ New:
language] (2005) Martin
(1993b)
sequence ideation base:
Halliday and
Matthiessen
(1999/2006:
Ch. 3)

120
 Resource guides

Rank – domain General Ideational Interpersonal Textual

Logical Experiential

figure; move ideation base: move: speech

[exchange]; Halliday and function:
message Matthiessen Halliday (1984a);
(1999/2006: Hasan (1985d,
Ch. 4) 1996b)16
exchange: Berry
(1981); Martin
(1992a); Ventola
(1987); Eggins
and Slade
(1997/2005)
element ideation base:
Halliday and
Matthiessen
(1999/2006:
Ch. 5)

As can be seen from the function-rank matrix in Table 3.7, semantics is organized internally
according to a compositional hierarchy, or semantic rank scale (for issues relating to the
semantic rank scale, see Section 5.2.5). Semantics extends from texts (and even macro-texts)
to the smallest elements of meaning; and it covers all metafunctional modes of meaning, viz.
ideational (experiential and logical), interpersonal and textual. In this respect, the systemic func-
tional theory of semantics contrasts with many other conceptions – ones where semantics is
limited in scale to propositions or predications and in mode of meaning to ideational meaning
and other aspects of meaning are “exported” to pragmatics.
The cells in Table 3.7 define the domains of different semantic system networks. That is, the
internal organization of semantics is stated in terms of the paradigmatic axis, with patterns
along the syntagmatic axis derived by means of realization statements, and paradigmatic orga-
nization is represented by means of system networks. Systemic functional descriptions of the
resources of semantics in terms of semantic networks began to be undertaken and presented
and published in the 1970s – either approached in its own terms as a meaning potential (e.g.
Halliday, 1972 [cf. Turner, 1987], on sociosemantic networks, mentioned earlier, and Halliday,
1984b, on the system of speech function) or approached from below in terms of lexicogram-
matical systems with implications for semantics “beyond the clause”: Halliday and Hasan’s
(1976) description of the textual lexicogrammatical system of cohesion, which was then com-
plemented by a semantic interpretation by Martin (1992a) – see also Taverniers (2019: 80).
In their survey of “semantic networks”, Hasan et al. (2007) track their development through
the decades. We have summarized their chronology in Table 3.5. The account given by Hasan
et al. (2007) is updated by Fung and Low (2019), who also include an overview of applications
of Hasan’s message semantic system networks (pp. 348–350).
In the tradition of semantic networks, paradigmatic organization is, of course, represented by
means of system networks. But there are interesting questions about the possible need to supple-
ment semantic system network (1) in order to provide more information about paradigmatic
patterns and (2) in order to support semantic reasoning.
121
 Table 3.8 Hasan et al.’s (2007) chronology of the development of semantic networks in SFL, with additional works not covered by them

Period Development of semantics Pages in Hasan Additional work on semantic Other contributions to semantics
et al. (2007) networks

1960s Hasan (1968): options in statement 700–701

1970s register specific semantic networks (regulatory): 702–707
Halliday (1972)
Turner (1973, 1987)
1980s non-register specific semantic networks 708–709 register-specific semantic the chooser-&-inquiry framework as an interface
(1) “paradigmatic semantics and syntagmatic networks: between lexicogrammar and higher-order
syntax”: Fawcett (1980) Patten (1988): register-specific sources of meaning: Matthiessen (1988a, 1990);
semantic networks as compiled Matthiessen and Bateman (1991)
solutions to recurrent
contextual problems
(2) “language exhaustive semantic networks”: 709–715 speech functional semantic “upper modelling”: ideational meanings networked

Resource guides
Hasan’s semantic network (Hasan, 1985d, 1989, network: Halliday (1984a); – for text generation: Matthiessen (1987b),
1992a, 1992b; Cloran, 1994; Williams, 1995) exchange: Berry (1981) Bateman (1990)
122

1990s Projects drawing on Hasan’s semantic network: 715– (discourse) semantic networks topology as a way of representing semantic
Maley and Fahey (1991); Williams (1995, (including semantic correlates agnation: Martin and Matthiessen (1991)
1999, 2001); Hall (2004) of cohesion): Martin (1992a)
rhetorical unit: Cloran (1994) ideational semantic network:
Halliday and Matthiessen
(1999/2006)
2000s Interpersonal semantic networks tending appraisal systems: Martin (2000); continuing work on grammatical metaphor;
towards lexical realizations Martin and White (2005), work on “spatial ontology” within ideational
Oteíza (2017); for discussion, semantics (drawing on Halliday and Matthiessen,
see also Halliday (2008b) 1999/2006), interfacing linguistic semantics
with ontologies needed in robotics modelling:
Bateman et al. (2010)
2010s Continued applications of Hasan’s semantic
network to various domains: see Fung and
Low (2019)
Adaptation of Hasan’s semantic network to
(questions in) Cantonese: Fung (2015); Fung
Ka Chun (2018)
 Resource guides

(1) The theory of paradigmatic organization of the semantic system includes systems of
options with “attached probabilities”, options related to one another as poles on a cline, options
on different systemic paths showing patterns of agnation and other properties that are hard
or impossible to capture by means of system networks. A number of these challenges emerge
already in the use of lexicogrammatical system networks (cf. Matthiessen, 1988b; Bateman,
1988, 1989; Teich, 1995, 1999a), and they are magnified as we apply representation by means
of system networks to the paradigmatic organization of the semantic system. At least two related
paths of exploration have been suggested.
One is to interpret systems in terms of fuzzy logic, taking terms in systems as names of
fuzzy sets, as sketched in Matthiessen (1995c) based on the research in Michio Sugeno’s Lab at
the Tokyo Institute of Technology, in particular by Ichiro Kobayashi. Another is to interpret
semantic space in terms of topology as a complement to the typological interpretation embod-
ied in system networks, as suggested by Martin and Matthiessen (1991), inspired by Jay Lemke’s
(1987) work on genre topology (see further Martin, 2003, on a topology of genres used in his-
tory; Martin and Rose, 2008, on genre topology in general). These paths of exploration remain
possibilities that could be taken up more generally in sustained research, but they have been, and
continue to be, metaphors that are helpful in conceptualizing semantic organization. Within
language, a number of issues similar to those that emerge in descriptions of the paradigmatic
organization of semantics also emerge in phonetics – not surprisingly, since both are interface
strata relating to what lies outside language (cf. Section 3.1.5.2).
(2) Since systemic functional scholars have developed a semiotic view of phenomena that are
treated as cognitive in cognitive science, foregrounding the central role of language (e.g. Halli-
day, 1978; Halliday and Matthiessen, 1999/2006), processes that are routinely taken to be cogni-
tive must also be illuminated in a semiotic perspective as semiotic processes, or more specifically
semantic processes. Such processes include reasoning quite centrally – cf. Hasan (1992b) on the
role of semantic sequences in the “rationality in everyday talk”, so reasoning – in all its varieties,
including deduction and inference – must be a semantic process and as such it must be supported
by the representation of the resources of semantics (cf. Halliday and Matthiessen, 1999/2006:
Chapter 3; Chapter 10, Section 10.2).
The need for accounts of semantic reasoning arises in many contexts of research and appli-
cation, including multilingual studies; for example, it is necessary to provide models of the
multilingual meaning potentials translators and interpreters operate with, showing how they can
reason about knotty translation problems like recreating the representation of process time in a
language with an aspect system by means of the resources in a language with a tense system (see
e.g. Matthiessen, 2018b).
But are semantic system networks up to the task of representing semantic resources and
processes of reasoning? The short answer would have to be “no”. If they are re-represented by
means of some variety of knowledge representation designed to support reasoning, the situation
will change (see e.g. Bateman, 1992), but the question is if the properties of semantic system
networks that have been valuable in manual text analysis (as in Hasan, 1989; Cloran, 1994; Wil-
liams, 1995; and the longer list of studies discussed by Fung and Low, 2019) can be combined in
semantic representation with the properties needed to support semantic reasoning.
Having sketched accounts of semantics in SFL, we can round off by noting the difference
between the SFL approach to meaning in terms of semantics and accounts of meaning in various
other framework where meaning is dealt with by semantics and pragmatics. In SFL, semantics
covers all modes of meaning – not just those traditionally recognized in philosophy and logic
in terms of reference or denotation, and it extends from the semantic system at the potential
pole of the cline of instantiation to the instance pole. In the “mainstream”, meaning has often

123
 Resource guides

been restricted to ideational meaning, and even to ideational meaning within the system, and
other aspects of meaning have been “exported” to pragmatics – including aspects that would
be handled in SFL by the relationship between context and semantics such as politeness phe-
nomena, which have to do with the relationship between the contextual parameter of tenor
and interpersonal semantic systems. The “mainstream” chunking of meaning into two separate
modules, semantics and pragmatics, is represented diagrammatically in Figure 3.7. The diagram
in the figure also brings out the fact that there is also a tendency to export the treatment of
“instantial meanings” like the reasoning in a particular context of situation in the processing of
text to pragmatics.
There have been many formulations of the distinction between “semantics” and “pragmat-
ics”, particularly as “pragmatics” began to take off as an institutionalized field of study, and there
is now a very extensive literature. However, Morris (1946: 219), who introduced the term
“pragmatics” in 1938 (Allwood, 1981; Gazdar, 1979), is still an excellent point of reference:

The following definitions retain the essential features of the prevailing classification,
while freeing it from certain restrictions and ambiguities: pragmatics is that portion of
semiotic which deals with the origin, uses, and effects of signs within the behaviour
in which they occur; semantics deals with the signification of signs in all modes of
signifying; syntactics deals with combinations of signs without regard for their specific
signification or their relation to the behaviour in which they occur.

Figure 3.7 Semantics in SFL ranges across all modes of meaning and extends along the cline of instantiation
from potential to instance, but in “mainstream” approaches, semantics is often restricted to
ideational meaning (often restricted in rank to the semantics of predications or propositions)
while other manifestations of meaning are “exported” to pragmatics

124
 Resource guides

The definition of “syntactics” is, of course, quite diferent from the systemic functional con-
ception of lexicogrammar as a resource permeated by the diferent metafunctional modes of
meaning. The definition of “semantics” invites the traditional conception oriented towards
ideational meaning, with “signification”, although Morris had revised his earlier definition
(Morris, 1938: 84), viz. “semantics as the study of ‘the relations of signs to the objects to
which the signs are applicable’”. In principle, “signification” could be extended to include
at least the interpersonal mode of meaning as well as the ideational one (cf. Hasan, 1985a,
on signification in relation to valeur). Similarly, while “pragmatics” has come to be confined
to the interpersonal and textual modes of meaning, uses and efects are equally relevant to
the ideational mode of meaning in systemic functional theory. Use is not part of a separate
component or module of meaning; it is simply inherent in the relationship between meaning
potential and acts of meaning – the instantiation of the meaning potential as text unfolding
in contexts of situation, in both the direction of generation (production) and the direction
of analysis (comprehension). And the “interpreter”, which was part of Morris’s (1938: 84)
definition of pragmatics – “the relation of signs to interpreter” – is also equally relevant to the
ideational mode of meaning, viz. the meaner as a person construing their experience of the
world around them and inside them.12

3.1.4 Accounts of lexicogrammar

In SFL, lexicogrammar is theorized, like all aspects of language in context, in terms of the
relationships it enters into and serves as the domain of. It is theorized as the lower of the two
content strata, content2 or the stratum of wording; this means that it stands in a natural relation-
ship to the higher of the two content strata, semantics (content1): patterns of wording are, as it
were, semantically “transparent”, and they are motivated by the different metafunctional modes
of meaning. Thus since semantics and lexicogrammar are both part of the content plane of lan-
guage, they are both organized metafunctionally into the spectrum of metafunctions: ideational
(experiential and logical), interpersonal and textual.
Lexicogrammar realizes semantics (and thus indirectly context), and it is in turn realized by
phonology (graphology, or sign in sign languages of deaf communities). In other words, it is
realized by the higher of the strata within the expression plane of language, the lower one being
the expression plane’s interface to what lies beyond language (the bodily resources for producing
and perceiving sounds, in the case of phonetics).
Unlike the natural stratal boundary between semantics and lexicogrammar, the one between
lexicogrammar and phonology (graphology, sign) is largely conventional – “arbitrary” in Eng-
lish translations of Saussure’s line of arbitrariness. This does not mean that there are no natural
patterns of realization; there is a higher degree of naturalness within the domain of prosodic
phonology – in particular, intonation – than within articulatory phonology and this extends
to “paralinguistic” features such as voice quality, but even within the articulatory domain,
there are some motivated patterns of realization (sound symbolism or phonaesthesia, including
onomatopoeia).
In contrast with “mainstream” accounts of the system of wording of language, SFL treats
them as a unified system – the system of lexicogrammar. This means that:

• the grammar of the word, “morphology”, and the grammar above the word, “syntax”, are
not fragmented into two separate modules but are instead unified as a resource of word-
ing (stated already by Halliday, 1961). Interestingly, linguists outside the SFL tradition have

125
 Resource guides

increasingly come to adopt the term “morphosyntax”, bringing the two domains along the
rank scale together;
• grammar and lexis are not separated into two different modules, syntax and morphology
on the one hand, and the lexicon on the other, but they are instead treated as zones along
a continuum – the cline of delicacy, ranging from the most general (grammar) to the most
delicate (lexis) with a region in-between that has been explored outside SFL in Construc-
tion Grammar and in lexical pattern grammar (cf. Gonzálvez-García and Butler, 2006,
2014; Butler, 2019).

Like the meaning potential of semantics, the wording potential of lexicogrammar is theorized
paradigmatically in the first instance (Halliday, 1966c). The paradigmatic organization is typi-
cally represented by means of system networks, including both networks of systems and realiza-
tion statements (see Sections 9.2.1 and 9.2.2 and Matthiessen, 2021b). This is brought out very
clearly in Matthiessen (1995a), which was written as a complement to Halliday’s Introduction to
Functional Grammar (IFG): unlike IFG, Matthiessen’s Lexicogrammatical Cartography (LexCart) is
organized around system networks, which constitute the map in this cartographic project.
These system networks have units of different (primary) classes as their point of entry, their
systems cover the metafunctional spectrum, and they extend in delicacy from the grammatical
zone of lexicogrammar towards the lexical zone. The systems that constitute the resources of the
lexicogrammar of a language can be viewed by means of a function-rank matrix (for English,
e.g. Halliday, 1970d, 1976a, 1978; Halliday and Matthiessen, 2004, 2014; for Chinese, Halliday
and McDonald, 2004; Japanese, Teruya, 1998, 2004, 2007). Halliday’s (1970a/2005a) version
of the function-rank matrix is reproduced here as Table 3.9 and the function-rank matrix in
Halliday and Matthiessen (2014: 87) as Table 3.10. The latter was designed as an index into the
presentation of the lexicogrammar of English, so can be used here to look up systems of interest.
The overall systemic-functional map of English has remained fairly stable over the years, but
there have been descriptive developments and some revisions and re-interpretations, including
ones that can be detected by comparing the two matrices in Table 3.9 and Table 3.10:

• The system of polarity was taken to be logical in 1970, but has been re-interpreted as
interpersonal at clause rank (in the description of English).
• In the verbal group, “catenation” has been given a comprehensive description in terms of
verbal group complexing, starting with Halliday (1985a).

But many of the changes in Halliday’s description of English since the 1960s have to do with
elaboration in delicacy and adjustments within particular systems (such as the system of process
type within transitivity, see Matthiessen, 2018a). This reflects the holistic approach to lan-
guage in general and to the development of descriptions of particular languages.
Halliday’s IFG and also Matthiessen’s LexCart have helped to stimulate the development
of a growing number of descriptions of the lexicogrammars of languages other than English –
LexCart provides some guidance in the form of “typological outlooks” for the major system,
later supplemented by Matthiessen (2004a). But there are now additional sources of guidance,
including descriptions of a growing range of languages that can be taken as models (eight of
which are represented in Caffarel, Martin and Matthiessen, 2004). For the task of developing
systemic functional descriptions of particular languages, see Section 4.3.
To provide a sense of the steady development of lexicogrammatical descriptions of a grow-
ing number of languages, we have adapted and updated a table from Teruya and Matthies-
sen (2015): see Table 3.11. (There are additional recent contributions in Martin, Doran and

126
 Table 3.9 Halliday’s (1970a/2005a: 169) function-rank matrix

Rank Class Metafunction

Ideational Interpersonal Textual

Logical (complexes) Logical Experiential Structural Cohesive

clause (major) taxis and logico- polarity transitivity mood theme cohesion
semantic type types of process, types of speech types of message (‘above the
participants, function (identity as text sentence’:
circumstances modality relation) non-structural
(identity clauses) (the WH-function) (identification, relations)
(things, facts and predication, reference;
reports) reference, substitution
substitution) and ellipsis;
conjunction;

Resource guides
group verbal secondary tense tense person voice
catenation (verb classes) (‘marked’ (‘contrastive’ lexical cohesion
127

options) options)
nominal classification modification attitude deixis
sub-modification epithet function attitudinal modifiers determiners
enumeration intensifiers ‘phoric’ elements
(noun classes) (qualifiers)
(adjective classes) (definite article)
adverbial (incl. narrowing ‘minor processes’ comment conjunction
prepositional) sub-modification verb prepositional (classes of comment (classes of discourse
relations (classes adjunct) adjunct)
of circumstantial
adjunct)
word (incl. (verbals, compounding lexical ‘content’ lexical ‘register’ collocation
lexical nominals, derivation (taxonomic (expressive words) (collocational
item) adverbials) organization of (stylistic organization organization of
vocabulary) of vocabulary) vocabulary)
information unit tone information
intonation systems distribution and focus
 Table 3.10 Halliday and Matthiessen’s (2014: 87) function-rank matrix, designed as an index to chapters and sections of the book

Stratum Rank Class Logical Experiential Interpersonal Textual

lexicogrammar clause taxis and logico- - transitivity mood [Ch. 4] theme [Ch. 3]
semantic type [Ch. 5]
info. unit [Ch. 7] - key [Ch. 4] information [Ch. 3]
group/ nominal modification thing type, nominal mood, determination
[Ch. 8]
phrase [§ 6.2] classification, person,
[Ch. 6] epithesis, assessment
qualification
verbal [§ 6.3] tense event type, aspect polarity, modality contrast, voice
adverbial modification circumstance type comment type conjunction type

Resource guides
[§6.4]
128

prepositional - minor transitivity minor mood

phrase
[§ 6.5]
word derivation denotation connotation
morpheme
phonology tone group tone sequence; tone [Ch. 4] tonicity [Ch. 3]
tone concord
[Ch. 7]
complexes simplexes
 Table 3.11 Seminal and comprehensive descriptions of different languages in (proto) systemic functional terms (excluding translation studies)

1950s 1960s 1970s 1980s 1990s 2000s 2010s

Chinese [Mandarin] Halliday A series of MA Halliday (1992d, Halliday and

(1956b, thesis supervised 1993d) McDonald (2004);
1959) by MAKH Halliday (2006);
at Sydney Li (2007)
University
Chinese [Cantonese] Tam (2004) Fung (2015, 2018)
Vietnamese Hoang (1997, Thai (2004)
2012); Thai
(1998)
Tibetan Wang (2020)
Mongolian Zhang (2020a, 2020b)
Thai Patpong (2005)

Resource guides
Japanese Teruya (1998) Teruya (2004, 2007) Ochi (2013)
Korean Park (2013)
129

Tagalog Martin (2004b)

Telugu Prakasam (2004)
Bajjika Kumar (2009,
2012); Kashyap
(forthcoming)
Western Desert Rose (2001, 2004)
Gooniyandi McGregor (1990)
Arabic (Modern El-Rabbat (1978) Williams (1989) Bardi (2008)
Standard) [Egyptian]
Beja Hudson (1973)
Yoruba Bamgboṣe
(1966)
Mbembe Barnwell
(1969)
Nzema Mock (1969)
(Continued)
 Table 3.11 (Continued)

1950s 1960s 1970s 1980s 1990s 2000s 2010s

Akan Matthiessen
(1987a, 1987b)
Dagaare Mwinlaaru (2017)
Oko Akerejola (2005)
English Halliday [see Halliday [see Halliday (1985a) Martin (1992a);
Halliday, Halliday, Matthiessen
2005a] 2005a]; Halliday (1995a)
and Hasan
(1976)
French Huddleston Caffarel (2004,

Resource guides
and Uren 2006b)
(1969)
130

Spanish Lavid, Arús and

Zamorano-
Mansilla (2010)
German Steiner and Teich
(2004)
Danish Andersen, Petersen
and Smedegaard
(2001)
Swedish Holmberg and
Karlsson (2006)
Finnish Shore (1992)
 Resource guides

Figueredo, 2020; and in Martin, Quiroz and Figueredo, 2021, on interpersonal grammar.
Thematic issues of WORD with systemic functional accounts of the nominal groups in dif-
ferent languages are being prepared.) This table is of course not exhaustive. We have included
PhD theses only where they provide comprehensive overviews of the stratal subsystem of a
language (typically the clause grammar) that have not subsequently been published in book
form. For certain languages there is now a fairly extensive literature, including a good deal of
work presented in languages other than English (e.g. the literature on Chinese in Chinese).
Particular areas of a number of languages have now been described – e.g. the nominal group in
Indonesian, reference in Weri, the system of theme in Korean, various aspects of Portuguese,
but since these do not provide general overviews, we have not included them in the table.
There are various areas where partial descriptions of a language will be included because they
are needed to carry out a particular task, as in translation studies (e.g. Kim, 2007a, on the system
of theme in Korean; Macdonald, 2019, also on Korean; cf. also Kim et al., 2021) or comparison
and contrast in support of second/foreign language education.
The information in Table 3.11 reflects the steady progress in the description of different
languages, but it gives no indication of what lexicogrammatical systems are in focus in the dif-
ferent contributions, which are – as we put it, comprehensive or seminal. To supplement the
information in this table, we provide another table where we use the function-rank matrix to
identify overviews of particular lexicogrammatical systems: see Table 3.12 (adapted from Mat-
thiessen, 2007a, and updated). Since the sketches of the clause grammars of eight languages are
identified in Table 3.11, we haven’t repeated them here, but it is clearly the case that the clause
grammars have been covered most extensively in the systemic functional descriptions of different
languages (cf. Figure 4.11). Table 9.4 in Section 9.5 of the Appendix presents an overview of
coverage in a number of descriptions of the lexicogrammar of English categorized in terms of
rank, metafunction and system.

3.1.5 Accounts of phonology and phonetics (graphology and graphetics)

3.1.5.1 Phonology
Moving downwards to the expression plane of language, we will start with phonology. The
development of systemic functional phonology from Firthian prosodic analysis is documented
by e.g. Henderson (1987), Tench (1992b), Halliday (2000) (and cf. also Halliday, 1970c, in rela-
tion to Halliday, 1992d).
In SFL, phonology is conceived of as a resource – a resource for making sound in the service
of realizing wordings, and, by another stratal step, meanings. (For SFL and other approaches to
phonology, see Section 2.6.3.1.) The phonology of language is thus its sounding potential,
and like the meaning potential and the wording potential it is modelled paradigmatically and
represented by means of system networks. The “architecture” of phonology according to SFL is
outlined by Matthiessen (2021a); for a theoretically deep exploration, see Cléirigh (1998). Apart
from the description of English phonology found in a number of publications (see later in this
section), the longest sustained descriptive project is Prakasam’s (e.g. 1972, 1976, 1977, 1979,
1982b, 1987) work on the phonology of Telugu, also with theoretical contributions based on
his descriptive insights.
Phonology is the higher of the two expression plane strata of language, the lower being pho-
netics (see Figure 3.1 and Figure 3.2). Like lexicogrammar, phonology is internal to language – a
stratum of form, whereas phonetics is like semantics, it is an interlevel or interface to what lies
beyond language, viz. the human bodily potential for producing and perceiving sounds. The

131
 Table 3.12 Lexicogrammatical function-rank matrix with references to a selection of publications

Rank General Class General Logical Experiential Interpersonal Textual

Complex Simplex Structural Cohesive
clause English: transitivity mood &c. theme conjunction
Matthiessen various: English: English: Fries ellipsis &
(2002a) Matthiessen Andersen (1981, 1995); substitution
(2018a) (2017) Ghadessy
English: Davidse Japanese: (1995);
(1999, 2017); Teruya Matthiessen
Matthiessen (2017) (1995b); Berry
(1999, 2014c); (1996); Forey
Tucker (2016) and Sampson
[Cardiff (2017); Huang
Grammar]; (2017) [Cardiff

Resource guides
Banks (2016); Grammar]
132

Neale (2017) Spanish: Arús

[Cardiff (2017)
Grammar]
phrase prepositional
group McDonald nominal English:
(2017); Fontaine
Fontaine (2017) and
and Schöntal cf. Fries
(2019) (1970);
Chinese: Li
(2017)
Indonesian:
Made (1988)
adjectival English: Tucker
(2017)
[Cardiff
Grammar]
 verbal various: Quiroz tense
French: English:
Caffarel Matthiessen
(2017) (1983, 1996)
French:
Caffarel
(1992)
word Matthiessen
(2015d): word
grammar
in SFL –
morphology
morpheme Halliday
(2008b:
58–61):
some

Resource guides
meanings
133

construed by
derivation in
English
[cutting
across
rank]
strata: Taverniers
grammatical (2017)
metaphor
delicacy: Hasan (1985c, Hasan (1985c, Oteíza
lexis 1987); Tucker 1987); Neale (2017) [but
(1998); (2002, 2006); interpreted
Matthiessen Matthiessen as located
(1991a); Cross (1995a; at semantic
(1992) 2014c) stratum]
general Butt and Davidse (1999) Matthiessen Clarke (2017)
Webster (1992)
(2017)
 Resource guides

stratal boundary between phonology and lexicogrammar is that between the expression plane
and the content plane of language and it is largely conventional (or “arbitrary”) rather than natu-
ral (see earlier, with references to principled exceptions). In contrast, the relationship between
phonology and phonetics is internal to the expression plane and it is natural, like the relation-
ship between lexicogrammar and semantics. (Outside SFL, scholars have in fact used terms like
“natural phonology” as a reaction against the very abstract generative phonology of the 1960s,
in particular Chomsky and Halle, 1968.)
In SFL, phonology is thus theorized as both functional and natural – functional in that it is a
resource for realizing wordings as soundings, and natural in that it stands in a natural relationship
to phonetics. When phonology is theorized along these lines, two separate though overlapping
domains emerge quite clearly, with the syllable as a gateway between the two:

• Prosodic phonology: the phonological resources of higher-ranking phonological units

such as the tone group and the foot in the description of English (e.g. Halliday, 1963a,
1963b, 1967, 1970b, 1985e; El-Menoufy, 1969, 1988; Tench, 1990, 1996; Martinec, 1995;
Halliday and Greaves, 2008; Greaves, 2007; O’Grady, 2017) – with excursions into voice
quality (e.g. Wan, 2009, 2010), and song and rap (e.g. Caldwell, 2010, 2014a, 2014b),13 and
work on speech synthesis (Teich, Watson and Pereira, 2000). Halliday (1970b) is designed
as a coursebook for students learning English intonation and rhythm, with accompanying
recordings. Halliday and Greaves (2008) is geared more towards an appliable description of
intonation in English, with many examples from different dialectal varieties of English and
spoken samples on an accompanying CD.
• Articulatory phonology: the phonological resources of syllables and (where applicable)
phonemes: Tench (1992a: Part I). Matthiessen (1987c) provides an outline of Akan pho-
nology with a focus on phonological words, syllables and phonemes; the model account in
this area is Halliday’s (1992d) description of syllables (or syllable finals) in Mandarin (more
specifically the Peking dialect); phoneme: Tench (2017).

Both these phonological domains are addressed in the contributions to two edited volumes on
systemic functional phonology, Tench (1992a) and Bowcher and Smith (2014).14
Complementing these two books, Tench (2011) provides a manual for transcribing spoken
English. His approach is not specifically systemic-functional, but on the one hand, he helps
“novice” students become observant – and enables them to transcribe spoken English, which is
very much a systemic functional approach of actually engaging with language – and on the other
hand he provides useful comparisons of systems of transcription, in particular a table comparing
Halliday with ToBi and a couple of other approaches to the annotation of intonation patterns
(p. 133).
Most systemic functional contributions to phonology have been concerned with prosodic
phonology and have been focussed on English and, while Tench (1992a) includes descriptions of
a number of different languages, Bowcher and Smith (2014) deals specifically with English;15 this
is an understandable extension of the long-term project since the 1960s of developing descrip-
tions of English to support text analysis (cf. Halliday, 1964a), including in clinical context (e.g.
Fine, 1991; Ovadia and Fine, 1995; de Villiers et al., 2007; Asp and de Villiers, 2010: descrip-
tion: 35–43, and applications throughout their book) and it resonates with the extension into
paralanguage and song-and-rap (e.g. Caldwell, 2007, 2008; and in the Firthian tradition, Robins
and McLeod, 1956a; Abercrombie, 1968). In contrast, most descriptions of articulatory phonol-
ogy in SFL have been concerned largely with languages other than English, including the work
in the 1960s by Bamgboṣe (1966) on Yoruba, Barnwell (1969) on Mbembe and Mock (1969)

134
 Resource guides

on Nzema. Tench (1992a) covers descriptions of Gooniyandi, Telugu, Algerian Arabic, Welsh
and Mandarin, and beyond articulatory phonology on Swahili and Yoruba.

3.1.5.2 Phonetics
As noted at the beginning of this section, phonology is the higher of the two expression plane
strata, phonetics being the lower one; and just as the two content strata, lexicogrammar and
semantics, phonology and phonetics stand in a natural relationship to one another. Thus the
phonological system constitutes a “semioticization” of the human bodily potential for producing
and perceiving sounds. Halliday (2003c: 14–15, emphasis in original) characterizes the relation-
ship between the two as follows in semogenic terms:

What I have been talking about here is the stratification of the “content” facet of the
original sign. Simultaneously, an analogous stratification took place in the “expression”
facet. Sound displaced gesture as the primary modality, and this likewise split into two:
an abstract organizational space (phonology) where sound is systemized to meet up
with the lexicogrammar, and speech sound as articulatory and auditory processes tak-
ing place in the human body (phonetics).

Like Firth’s prosodic analysis approach to phonology, systemic functional phonology is abstract
in the sense that it is organized to serve as a resource for realizing patterns of wording, and
captures the valeur of terms in phonological systems, but at the same time, again like prosodic
analysis, it is grounded in phonetic facts. This is very evident in the work by British phoneti-
cians drawing on Firth’s work – David Abercrombie (e.g. 1967) and Ian Catford (e.g. 1977) in
particular – who have been important in the development of systemic phonology, as is evident
for example in Halliday’s comments in various places on English as having a foot-timed system
of rhythm. In fact, Catford has contributed to phonetics in general – his Fundamental Problems
in Phonetics (1977) can be treated as a blueprint for further development of phonetics along sys-
temic functional lines – and at the same time, he has contributed to systemic phonology (e.g.
Catford, 1985). In his introduction to his 1977 book on phonetics, Catford (1977: 1) writes:

It was my original intention to entitle this book Anthropophonics: the Phonetic Categori-
zation of Human Sounds. I was dissuaded from this course of action on the grounds that
the use of this unusual term might be suggestive of an eccentricity that would discour-
age potential readers. The term anthropophonics has, however, a most respectable history
of use in the works on phonetics, going back to the great Polish-Russian linguist Jan
Baudouin de Courtenay. . . .
By anthropophonics, Jan Baudouin de Courtenay meant the study of what is com-
mon to all mankind in the sphere of vocal sound production: the study of the total
sound-producing potential of man.

So potential is central to the nature of phonetics as a resource. And another British phoneti-
cian, Peter Ladefoged, who like Catford had moved from Abercrombie’s Edinburgh to the US,
also contributed to what we can think of as systemic phonetics through his work on features in
phonology and phonetics (e.g. Ladefoged, 1988, 2004). For some further discussion, see Mat-
thiessen (2021b: Chapter 6).
Systemic functional linguists working on phonology have also undertaken phonetic studies
to support their phonological descriptions – an approach that continues the Firthian tradition of

135
 Resource guides

phonetically grounded phonology. This has probably been applied in particular to the descrip-
tion of intonation. While tone contours are well known to be acoustically composite (pitch,
loudness, length), pitch traces are still a valuable part in the study of intonation. Watt (1992)
reports on “an instrumental analysis of English nuclear tones”, presenting instrumental analysis
for the single and compound tones identified and described by Halliday (1967, 1970b); see also
Watt (1994) for further details. Since Watt undertook his instrumental analysis, phonetic tools
have become increasingly accessible to working phonologists (for an overview from the 1990s,
see Ladefoged, 1997, who was an enthusiastic instrumental fieldworker). Bill Greaves had used
SIL’s helpful tools, but then switched to Praat, which has become a great standard tool for
researchers working on intonation (and of course on other areas of phonology and phonetics).
Halliday and Greaves (2008) use Praat extensively in their description of the system of intonation
in English, and this approach has also been followed by other systemic functional linguists (see
e.g. Bowcher and Debashish, 2019).
Just like semantics, phonetics is an interlevel – an interface to what lies beyond language.
What lies beyond language includes the human bodily potential for producing and perceiv-
ing sound, and by another step in the ordered typology of systems, the manifestation of sound
physically as sound waves (and of course engineering solutions in speech synthesis, speech analy-
sis). However, it also includes expressive resources that border those of phonetics, as suggested
in Figure 3.8. Thus in terms of semiotic systems that operate alongside spoken language, it
includes voice quality and other uses of the voice in sound production but it also includes visual
manifestations of phonetic sound production, the most exposed ones being facial expressions, in

Figure 3.8 Phonology and phonetics in relation to other human systems on the expression plane

136
 Resource guides

particular the shape of the lips, which makes “lip reading” (or “speech reading”) possible. More
generally, the face, including the mouth, is a key expressive resource for interpersonal meanings,
or meanings that can be interpreted as interpersonal in terms of the metafunctional organization
of language (for the face as an interpersonal expressive resource in sign language, see Johnston,
1992a, 1992b). In terms of “bio-semiotic” systems (Halliday and Matthiessen, 2006), phonet-
ics interfaces with sensorimotor systems – a multimodal or heteromodal connection (which of
course has implications for the content plane; cf. in neuroscience e.g. Fernandino et al., 2016).

3.1.5.3 Graphology and graphetics

In written language, the expression plane is organized into graphology and graphetics instead of
the two expression strata of spoken language, phonology and graphetics. Neither stratum seems
to have received much attention in SFL (but cf. Albrow, 1972), the main contribution being
Halliday’s (1985f) typology of writing systems, adapted here as Figure 3.9.16 He differentiates
writing systems according to their relationship to other strata, primarily either phonology –
alphabets and syllabaries – or lexicogrammar – logographic writing systems, or characteries. For
some discussion, see also Chapter 1 of Halliday’s IFG.
Writing systems have different calligraphic potentials, the Chinese charactery and the Arabic
script being two outstanding examples of scripts transcending writing and shading into drawing
and art more generally. This is one area where graphology and graphetics may receive greater
attention in SFL in the future thanks to the development of multimodal studies. Another area is
the connection between writing and layout; work on layout originating in computational model-
ling is relevant here: Bateman et al. (2001) – in (manual) analysis: Bateman (2008a), Bateman et al.
(2016). Conceived of in the context of computational modelling, Sefton (1990) offers a specifica-
tion of the incorporation of “punctuation and typography” as part of a text generation system.
Taking a step back, we can also view graphology and graphetics as part of Halliday’s (1996b)
holistic picture of literacy referred to as an example of helical exploration in Section 1.4 (see

Figure 3.9 Adapted representation of Halliday’s (1985f: 43) typology of writing systems in relation to the
hierarchy of stratification

137
 Resource guides

Figure 1.9). Here “writing systems” are part of the picture, but in work on literacy in SFL the
focus has tended to be on other parts of the total picture (cf. Rose and Martin, 2012).
For the illumination of reading from the vantage points of different disciplines based on a
shared understanding of reading, a “single perspective”, see Goodman, Fries and Strauss (2016).
In their preface, they set out the goals for their important book:

Some researchers have devoted their careers to what they call a simple view of reading.
They think that if they can find a few tests that correlate well with reading comprehen-
sion that by teaching to those tests children will be taught to read.
In this book we’ll show that there is nothing simple about reading. Reading is one
aspect of language that is really quite complex. Fortunately, language, including writ-
ten language, is something we humans are really good at learning.
The purpose of this book is to explain not only what reading really is, but why com-
mon sense makes it seem to be something quite different from that reality. In other words,
we will explain this grand illusion. And, as we shall see, unraveling the secrets of the grand
illusion of reading will teach us about far more than reading itself. We will learn about
how remarkable human language is. It will help us understand how the brain uses lan-
guage, visually, to navigate the world. It will teach us about what it means to be human.

Their book is, then, also an illustration of the insights to be gained by seeing language holistically
and by interpreting in in the ordered typology of systems.

3.2 Organized according to the cline of instantiation

While stratification is a hierarchy organized into discrete orders, like rank, instantiation is a cline,
like delicacy; it is a continuum extending from the potential pole of the cline to the instance
pole with intermediate patterns in-between. The instance pole is where we can observe semi-
otic patterns unfolding in time; here we sample texts in contexts unfolding over relatively short
periods of time, say from less than one minute to several hours. Such instances accumulate over
longer periods of time forming patterns that we recognize as recurrent types, or even as systems
at the potential pole. But, as Halliday has pointed out on many occasions (e.g. Halliday, 1991d,
1994d, 2002a), the theory does not differentiate text, intermediate patterns and systems along
the cline as different phenomena (in the way that langue and parole or competence and perfor-
mance have been postulated in much linguistic theorizing), only as different “phases” of one
and the same phenomenon; so in terms of the study of this phenomenon, we use the cline of
instantiation to locate ourselves at different viewpoints – close up, we observe instances unfold-
ing in time, and at a far greater distance, we observe systems evolving through time. Of course,
we infer generalizations about systems evolving based on samples of instances: the instance pole
of the cline of instantiation is where the observer samples data, while the potential pole is where
we locate our generalizations in the shape of theories and descriptions.
The cline of instantiation is represented diagrammatically in Figure 3.10. This figure also
shows how we can use the cline of instantiation to locate data and theory at the two poles of
the cline, relating them by means of moves along the cline: deduction, induction and abduc-
tion. Here deduction means going from a description of the system of a language to texts at the
instance pole of the cline, trying to find examples supporting the description. Induction means
moving in the other direction, assembling generalizations of patterns found in texts. These
two directions can be related to Tognini-Bonelli’s (2001) distinction between corpus-based and
corpus-driven research; but this distinction leaves out a third move, or combination of moves,

138
 Resource guides

Figure 3.10 The cline of instantiation extended between the potential pole and the instance pole – and
the locus of data and theory, and of three investigative moves: deduction, induction and
abduction

which is crucially important. This is abduction, the kind of dialectic move between descriptions
of the system and text instances practiced in text analysis – typically manual, but not necessarily
so. See further Section 4.3.6.

3.2.1 The stratification-instantiation matrix

Like the hierarchy of stratification, the cline of instantiation is a global semiotic dimension, and
it intersects with the hierarchy of stratification to form a two-dimensional semiotic space. The
semiotic resources of language in context are “dispersed” in this space, and they can be surveyed
by means of a stratification-instantiation matrix, as shown by Halliday (2002a): see Figure
3.11. (This matrix can be extended downwards in stratification to cover the expression plane of
language as well – phonology and phonetics, graphology and graphetics, and their analogues in
sign languages such as Auslan.)

139
 Resource guides
140

Figure 3.11 The stratification-instantiation matrix as presented by Halliday (2002a)

 Resource guides

The resources at the potential pole of the cline of instantiation – the system of culture or “con-
text of culture”, the semantic system or meaning potential, and the lexicogrammatical system or
wording potential – have been surveyed in terms of their locations in the hierarchy of stratifica-
tion in Section 3.1, as has the phonological system or sounding potential.
The development of descriptions of the systems at the potential pole of the cline of instantia-
tion will be discussed in Section 4.3, and the use of such descriptions in the analysis of instances,
i.e. of texts in their contexts of situation, will be explored in Section 4.4. Here we will focus
on the resources in between the potential and instance poles – the resources that are explored
in terms of systemic variation in instantiation, viz. codal (variation) studies and in register
(variation) studies.
In principle, we should also include a section on dialect (variation) studies, but while dialectal
variation is very important in societies and has been placed on theoretical maps in SFL (going
back to Halliday, McIntosh and Strevens, 1964; Gregory, 1967; Hasan, 1973; Halliday, 1978;
Gregory and Carroll, 1978), there has been relatively little descriptive work on dialects as semiotic
resources17 (see also Benson and Greaves, 1984; Lemke, 1995, for relevant discussion of dialects).
The three different kinds of variation are “plotted” in terms of the cline of instantiation and
the hierarchy of stratification in Figure 3.12. In terms of the cline of instantiation, they can be
located within different regions of the cline. Dialectal variation is close to the potential pole,
as evidenced by the tendency for dialects to drift apart gradually during migrations of speakers,
as documented for a number of language families. Codal variation is fairly close to the poten-
tial pole since it involves whole sub-cultures or classes within a community (cf. Halliday, 1994d),
while register variation can be located somewhere midway between the potential pole and the
instance pole. Approached from the instance pole, it can be characterized in terms of text types.
In terms of the hierarchy of stratification, there are two properties to take into consideration,
viz. the location of the variation (phonology, lexicogrammar or semantics) and the possible exis-
tence of a higher-stratal constant. Registerial and codal variation are both modes of semantic
variation (although Hasan, e.g. 2009b, has used the term “semantic variation” for codal variation
in the first instance). However, they differ with respect to a higher-stratal constant. In the case of
registerial variation, there is no higher-stratal constant: this is semantic variation correlating with

Figure 3.12 Kinds of variation in relation to the cline of instantiation and the hierarchy of stratification –
dialectal, codal and registerial variation

141
 Resource guides

contextual variation; registers constitute different modes of meaning in different contexts, which
is why Halliday (e.g. 1978) has characterized registers as the meanings at risk in particular
contexts. In the case of codal variation, there is a higher-stratal constant: codes are different ways
of meaning within comparable contexts, like the contexts critical to socialization identified by
Bernstein and studied linguistically by Hasan (1989, 2009b). In contrast with both codal variation
and registerial variation, dialectal variation is “lower-level” variation – phonological varia-
tion and to some extent morphological and lexical variation within lexicogrammar; and it has a
constant within semantics, which is reflected in the characterization of dialects as different ways
of saying the same thing (e.g. Halliday, McIntosh and Strevens, 1964; Halliday, 1978).

3.2.2 Codal (variation) studies

In her classic characterization of code, register and social dialect, Hasan (1973: 228) offers the
following way into the exploration of codes, having discussed dialects:

Two outstanding differences between the two [dialects and codes] immediately draw
attention to themselves: while the extra-linguistic factor(s) correlating with social
dialect are incidental, those correlating with code are said to be causal; if the rela-
tionship between the two in the former case is simply that of co-occurrence, the
relationship between the two in the case of code is that of logical dependence, which
presupposes co-occurrence. Second, while social dialect is defined by reference to its
distinctive formal properties, the code is defined by reference to its semantic proper-
ties, thus involving the consideration of the formal levels only. That is to say, it can be
argued that the semantic properties of the codes can be predicted from the elements
of social structure which, in fact, give rise to them. This raises the concept of code
to a more general level than that of language variety; indeed, there are advantages in
regarding the restricted and the elaborated codes as codes of behaviour, where the
word behaviour covers both verbal and non-verbal behaviour.

This provides an important systemic functional take on Bernstein’s (e.g. 1971) theory of codes.
Significantly, it was clear already at this stage in the development of the theory that the locus
of variation is within the semantic stratum, which is the vertical axis in Figure 3.12. Halliday
(1994d) tracks the successive refinements in the characterization of codes in Bernstein’s theory
and documents the history of dialogue between Bernstein and systemic functional linguists. He
then characterizes codal variation in comparison with the other two kinds of variation, dialect
variation and register variation (1994d/2007: 242–243):

Let us return for a moment to code, register and social dialect. What distinguishes
register variation from dialect variation is that, prototypically, dialects are different ways
of saying the same things, whereas registers are ways of saying different things. In other
words, dialectal variation is phonological or lexicogrammatical, with a higher-level
constant in the semantics (same meanings, different wordings and sounds). With regis-
ter variation there is no higher-level constant: the meanings are different because they
represent different spheres of activity. With code, on the other hand, there is semantic
variation: the meanings are different – but there is a higher-level constant, namely the
context of situation (which in systemic theory is modelled as another level in the stratal
organization of language). Thus for example if the situation is one of a mother explain-
ing a particular principle of behaviour, semantic variation refers to the fact that, in the

142
 Resource guides

same context and with the same semiotic function, different options in meaning may
be taken up. Such variation might of course be random; but if it is systematic, such that
it distinguishes statistically between one group within the population and another, this
constitutes a difference of code.

He raises the question of where codes are located along the cline of instantiation – the horizontal
axis in Figure 3.12, and problematizes the answer (1994d/2007b: 236–237):

The difficulty we have with the concept of “code” in Bernstein’s model, as I have sug-
gested elsewhere (Halliday, 1992d: 71–73, 88), is that – very much as with the concept
of global warming – we need to position ourselves at some midpoint along the scale. In
order to be able to do this, we have to view “code” simultaneously from both perspec-
tives, seeing it both as variation in the system (in the semiotic climate, or meaning poten-
tial) and as different patterns in the text (in the semiotic weather, the way this potential
is instantiated). The regularities that Bernstein is observing and accounting for lie just at
this intermediate depth.

Halliday (1997/2003b: 259–261) problematizes the location of codal variation along the cline of
instantiation (the horizontal axis in Figure 3.12):

where do we model the codes on the cline of instantiation?

We can throw some light on this question by looking into the other type of semanti-
cally based variation, that of register. It might be worth noting here that the distinction
between the language of commonsense knowledge and the language of educational and
technical knowledge, as discussed by Jim Martin and myself in our book Writing Sci-
ence (1993), is also a difference of code, though one that is a little closer to variation of
the register kind than are the original Bernstein codes. I have suggested elsewhere the
reasons why I think that code is the most difficult kind of variation to model: because
it lies near enough to the “system” end of the cline to be hard to typify instantially.
Figure 3.2 [reproduced here as Figure 3.12] makes this point in diagrammatic form.
System and instance form a complementarity such that we can examine variation, as a
phenomenon having to be located intermediately between the two, from either end. If
we consider register variation first: viewing from the “instance” end, we can recognize
a text type as a collection of similar instances. But when we shift perspective and see it
as systemic variation, each of these text types appears as a register, a kind of subsystem
which redounds with the properties of the context in terms of field, tenor and mode.
Likewise, we can look at code variation from the perspective of the instance. It is
extremely difficult to do this, because it means recognizing different semantic styles
(different “ways of meaning”, in Hasan’s terms) within one and the same language –
which may appear from above, as different ways of “meaning the same thing” (e.g. of
grounding the reasons for a given judgment, or of interpreting a given complex class
of experiences), but end up by creating different models of reality. The critical step
here is to identify that which various instances have in common, such that the shared
features constitute a code. . . .
Now that we have a more comprehensive account of the grammar, Hasan’s work on
mother – child interaction together with her theoretical modelling of semantic varia-
tion have been able to provide both a powerful demonstration of the principle of code
variation and a further confirmation of the validity of Bernstein’s original findings.

143
 Resource guides

An early report of Hasan’s findings based on her large project was Hasan (1989), and this and
other contributions are collected in Hasan (2009b). For Bernstein’s code theory in the context
of the “critical turn”, see O’Grady (2019: 463–464). For Karl Maton’s framework of “Legiti-
mation Code Theory” building on Bernstein’s code theory in relation to SFL in the work by
J.R. Martin and his group, see Maton and Doran (2017).

3.2.3 Register (variation) studies

As can be seen in Figure 3.11, Figure 3.12 and Figure 3.13, registers are located within the
region intermediate between the potential pole of language and the instance pole. They can
be interpreted as functional varieties of the overall meaning potential of language – as meaning
subpotentials operating in particular institutional settings.18 They are complemented by views
of functional varieties positioned at the instance pole of the cline of variation – observations of
patterns of meaning in texts that are sufficiently alike to be recognized as text types.
In undertaking register studies, researchers may focus on very general registers such as sci-
entific language (cf. Halliday’s, 1988a, characterization of “the language of physical science”)
or on, say, types of text emerging thanks to channels of communication enabled by new tech-
nology.19 (All registers have of course originally emerged as patterns of meaning in text, and
all registers are constantly undergoing change in progress – although they may become highly
stylized and codified.)
As the location of registers midway along the cline of instantiation indicates, the system of
a language can be interpreted and described as an assembly of registers (e.g. Halliday, 1978;
Matthiessen, 1993a). This is in fact one way in which linguistic systems achieve the property
of metastability; as they evolve (see Section 3.2.4), they keep adapting to their contextual
conditions, and this process of adaptation includes the emergence of new registers but also the
disappearance of registers that are no longer useful to the speakers of a community (although
they may of course be preserved as ritualized varieties). Thus unlike languages, many registers
have “lifespans”, and their evolution can be tracked if historical corpora are available. Halliday
(1988a) provides a model for the study of the evolution of registers with his account of the his-
tory of scientific English. Since then there have been other systemic functional studies of the
evolution of particular registers, e.g. Nanri’s (1993) study of the evolution of news reports of
(attempted) assassinations and Bank’s (2008) further investigation of the evolution of scientific
language: for more examples, see Table 3.15 under phylogenesis.
Turning from the location of registers along the cline of instantiation to the hierarchy of
stratification, we can note, referring to Figure 3.12 and Figure 3.13, that the locus of register
variation is the semantic stratum, and since the relationship between semantics and lexicogram-
mar is a natural one, lexicogrammar is also implicated in register variation, which means (among
other things) that registers can also be investigated by means of lexicogrammatical analysis. But
register variation is semantic variation in the first instance. There is no higher-stratal constant;
registers vary with the variation in the values of the field, tenor and mode parameters of context.
These values are located mid-region along the cline of instantiation, in the region character-
ized as subsystem/instance type in Figure 3.11. Contextually, this is the region of institutions/
situation types.20 This is why Halliday (e.g. 1978) has characterized registers as meanings at
risk in context.
In view of the characterization of registers by reference to context as “meanings at risk” in
context, it makes sense to approach them “from above”, from the point of view of context, just
as we define text by reference to context (Halliday and Hasan, 1976; Halliday, 1981). Thus in
SFL registers are usually identified, analysed, characterized and classified based on the contextual

144
 Resource guides

Figure 3.13 The location of codal variation along the cline of instantiation, from Halliday (1997/2003b: 259)

Figure 3.14 The locus of registerial cartography in terms of the cline of instantiation and the hierarchy
of stratification

parameters of field, tenor and mode (see Section 3.1.2.3), as can be seen from the contributions
to Ghadessy (1988, 1993a) and from the examples of systemic functional register studies given
in Table 3.14.
One strand in this research is the long-term research programme known as registerial car-
tography; see e.g. Matthiessen (2015b, 2015c, 2019, 2021c). This cartographic effort involves
characterizing registers trinocularly, i.e. contextually, semantically and lexicogrammatically: see
Figure 3.14. Here registers are identified “from above” by reference to settings of field, tenor
and mode values. The contextual structures realizing different settings are also specified, and
they are further characterized semantically, for example by global to local textual patterns and by

145
 Resource guides

logico-semantic complexing based on the Rhetorical Structure Theory style of analysis (Mat-
thiessen, forthcoming b), and the semantic patterns are in turn examined in terms of their
lexicogrammatical realizations.
It is possible to fix a certain setting of field, tenor and mode values within context, and then
attempt to describe a register-specific semantic system that is tailored or adapted specifically to
the combination of field, tenor and mode values. This approach to register is illustrated by Halli-
day’s (1972) notion of sociosemantic system networks, i.e. semantic networks that represent
particular registers (see also the immediately following).
In frameworks where the analysis of texts is based on computational tools providing
automatic analysis, the approach to register analysis is likely to be “from below”, from
lexicogrammar – and within lexicogrammar, fairly low-ranking features are likely to be
involved, as in Biber’s (e.g. 1988, 1994, 1995) Multidimensional (“MD”) analysis (for
comments informed by SFL, see Ghadessy, 2003). Naturally, these approaches from differ-
ent starting points are ultimately complementary. Sophisticated computational techniques
have also been used in SFL, an important model being Teich et al. (2016: 1668):

We analyze the linguistic evolution of selected scientific disciplines over a 30-year time
span (1970s to 2000s). Our focus is on four highly specialized disciplines at the bound-
aries of computer science that emerged during that time: computational linguistics,
bioinformatics, digital construction, and microelectronics. Our analysis is driven by the
question whether these disciplines develop a distinctive language use – both individu-
ally and collectively – over the given time period. The data set is the English Scientific
Text Corpus (SCITEX), which includes texts from the 1970s/1980s and early 2000s.
Our theoretical basis is register theory. In terms of methods, we combine corpus-based
methods of feature extraction (various aggregated features [part-of-speech based],
n-grams, lexico-grammatical patterns) and automatic text classification. The results of
our research are directly relevant to the study of linguistic variation and languages for
specific purposes (LSP) and have implications for various natural language processing
(NLP) tasks, for example, authorship attribution, text mining, or training NLP tools.

In addition to the diferent stratal approaches to registers, they can, at the same time, be approached
both qualitatively and quantitatively. Quantitatively, they can be theorized and described as the
resetting of the probabilities of the general probabilities of the semantic system (and by another
step, by a stratal descent, of the lexicogrammatical system). These registerial probabilities can
be induced from relative frequencies identified in the analysis of texts sampled as representative
of one register or another. Qualitatively, registers can be theorized and described as semantic
systems adapted to particular types of context – institutional settings or situation types, as in Hal-
liday’s (1972) example of maternal regulatory semantics. There have been fairly few examples
of this approach to register characterization, but Patten (1988) has used it in a text generation
system, pointing out its strength based on AI problem solving work: registers can be viewed as
compilations of solutions to recurrent problems, avoiding the work of solving the problems by
searching the overall semantic space every time.
There are a number of overviews of register theory and of register studies in SFL, includ-
ing Ghadessy (1988, 1993a), Lukin et al. (2008), Matthiessen (1993a, 2015b, 2019) and Moore
(2017). Systemic functional studies of particular registers are listed in Table 3.13; for accounts
of register theory, see Table 9.3. While Table 3.13 is hopefully representative, it is of course not
exhaustive.21 Studies have appeared in many different places, and also in different languages; new
studies are being added all the time now also thanks to the journal Register Studies, which was

146
 Table 3.13 Examples of register studies in SFL

Register (macro and micro) Mode Focus of study References

General Particular varieties Medium Other semiotic

overviews theory of register variation (diatypic Ghadessy (1993a);

variation) and of analytic and Matthiessen (1993a);
descriptive work on register Lukin et al. (2008);
Matthiessen (2015b,
2015c, 2019); Moore
(2017, 2020)
Journal: Register Studies
conversation (casual) chat; chunks (gossip, spoken – interpersonal systems: semantics Eggins and Slade (1997,
opinion, anecdote (exchange, speech function, appraisal) 2005)
. . .) and lexicogrammar;
genres within (chunks of) casual

Resource guides
conversation
147

service encounters face-to-face spoken Ventola (1987)

telephonic spoken tenor-interpersonal slice through strata Matthiessen et al. (2005)
in analysis of telephonic service
encounters ordering pizza
spoken negotiation of relationships in call centre Hood (2010)
dialogue
spoken voice quality and appraisal in call centre Wan (2010)
dialogue
customer service spoken and written appraisal analysis of written and spoken Hui (2010)
complaints customer service complaints
conversation spoken – based on Sinclair and Coulthard (1975) Tsui (1986)
rank scale (acts, moves, exchanges,
sequences and transactions)
utterances interpreted as acts, and
characterized in terms of interpersonal
system (essentially speech function)
(Continued)
 Table 3.13 (Continued)

Register (macro and micro) Mode Focus of study References

General Particular varieties Medium Other semiotic

classroom discourse spoken – modelled on Halliday’s (1961) scale-&- Sinclair and Coulthard
category theory (1975)
ranks of lessons at level of “discourse” (acts,
moves, exchanges, transactions, lessons);
structure of units at different ranks
spoken exchanges between teachers and students Gibbons (2007)
of English as a second language,
drawing on Halliday & Vygotsky
classroom teacher- spoken gaze, gesture, “This study explores the multimodal and Amundrud (2017)
student consultations spatial linguistic contours of the individual

Resource guides
position feedback consultation classroom
148

curriculum genre (Amundrud, 2017),

which is comprised of five stages:
Opening, Conferring, Advice, Scoring,
and Closing. . . . Through the systemic-
functional multimodal discourse analysis
(SF-MDA) of classroom discourse . . .
these consultations were examined for
the use of spatial position, gaze, gesture,
and language.”
academic discourse academic lecture spoken characterization of an academic lecture Malavska (2016)
drawing on SFL and other genre
frameworks
scientific discourse written – Huddleston et al. (1968);
Banks (1991)
written – evolution of the language of physical Halliday (1988a)
science (English) over 500 years
 written – evolution of scientific discourse in the Banks (2005)
late 17th century
written – disciplinarity: variation across disciplines Teich et al. (2016)
written visual semiotics Lemke (1998)
popular science written – investigation of popular science involving Fuller (1995, 1998)
comparison of David Suzuki and Jay
Gould
popular science books and articles, Teich (2003)
textbooks in English and German with
a focus on cross-linguistic variation;
analysis by means of corpus tools
school science “scientific literacy for participation”: Knain (2015)
analysis of examples from a school
textbook, e.g. dealing with energy

Resource guides
transfer, and students’ interaction
around equipment, a scientific
149

instrument and a computer, also

image analysis; chapters providing
background for conducting linguistic
and pictorial analysis
student writing across disciplines in Nesi and Gardner (2012)
higher education
discourse of taxonomic reports, . . . written Wignell, Martin and Eggins
geography (1987)
social science written Wignell (1997, 1998, 2007a,
2007b);
written, web-based multimedia Zhao (2011)
mathematical (general overview) written symbolism, O’Halloran (2005)
discourse visual images
historical discourse historical recounts, written – Eggins, Wignell and Martin
accounts, (1993); Coffin (1997,
explanations, 2006); Rajandran (2012)
expositions
(Continued)
 Table 3.13 (Continued)

Register (macro and micro) Mode Focus of study References

General Particular varieties Medium Other semiotic

media discourse (general overview) written Iedema, Feez and White

(1994); Thomson and
White (2008); Fowler
(1991); Lukin (2010)
news discourse written – evolution of “hard news”, articles about Nanri (1993);
(attempted) assassinations
written images Caple (2008); Bednarek and

Resource guides
Caple (2012)
sports news written construing sports stars as celebrities Dreyfus and Jones (2010)
150

sports commentary spoken – play-by-play sports commentary (rugby), Bowcher (1999, 2001, 2003,
with intonation analysis 2004)
advice columns written Thibault (1988)
media interviews spoken Bell and van Leeuwen (1994)
“social media” written (typically) emojis, images, Twitter and social media; appraisal Zappavigna (2011)
clips analysis
business magazines written discursive representation of companies Rajandran (2014)
corporate reports CEO statements written Rajandran and Taib (2014)
business discourse written Ghadessy (1993b)
administrative (general overview) written Iedema (1995, 1996, 1997)
discourse
legal discourse agreements written the language of privacy in Japanese and Chik (2018)
English
 verbal art stories, poetry written – Hasan (1985b)
poetry written Butt (1984); Hasan (1985b;
1988); Halliday (1988b);
Lukin (2003); Lukin and
Webster (2005)
drama written (to be Birmingham School approach to dialogue Burton (1980)
performed) in Pinter plays
(written to be) the semiotics of creativity in MD House Law (2017)
performed
narrative O’Toole (1982);Toolan
(1989, 2009)
chronicle Thibault (1991)

Resource guides
graphic novels, comics written pictorial (images) ideological analysis; multimodal analysis Veloso (2006); Veloso,
Bateman and Lau
151

(forthcoming); Bateman
et al. (2016)
song (written to be) sung music folk ballad: music and lyrics Steiner (1988); O’Toole
or rapped (2015)
pop song Cranny-Francis and Martin
(1991)
rap, including the phonology/phonetics Caldwell (2007, 2008, 2010,
of the rap voice 2014a, 2014b)
 Resource guides

launched in 2019 in order to provide a dedicated channel of publication for studies drawing on
different traditions (cf. the launch issue, which includes contributions from a number of scholars
who have engaged with register addressing the same questions from the editors of the journal).

3.2.4 Semogenesis in relation to the cline of instantiation

As we have seen, the cline of instantiation enables us to locate different kinds of variation in
language (together with the hierarchy of stratification). In addition, it also helps us interpret
different phases of semogenesis, the process of creating meaning within different timeframes.
Through this general notion of semogenesis, it is possible to bring together phenomena that
have been studied in quite distinct sub-disciplines within non-SFL linguistics or even other
disciplines. Let us quote the characterization of semogenesis within different timeframes from
Halliday and Matthiessen (1999/2006: 18–19):

Meanings do not ‘exist’ before the wordings that realize them. They are formed out
of the impact between our consciousness and its environment. We need, therefore, a
further guiding principle in the form of some model of the processes by which mean-
ing, and particular meanings, are created; let us call these semogenic processes. Since
these processes take place through time, we need to identify the time frames, of which
there are (at least) three. [footnote: There may be more, but these are the ones that
matter here.]

(i) First, there is the evolution of human language (and of particular languages as
manifestations of this). Known histories represent a small fraction of the total time
scale of this evolution, perhaps 0.1%; they become relevant only where particular
aspects of this evolutionary change have taken place very recently, e.g. the evolu-
tion of scientific discourse. This is the phylogenetic time frame.
(ii) Secondly, there is the development of the individual speaker (speaking subject).
A speaker’s history may – like that of the biological individual – recapitulate some
of the evolutionary progression along epigenetic lines. But the individual experi-
ence is one of growth, not evolution, and follows the typical cycle of growth,
maturation and decay. This is the ontogenetic time frame.
(iii) Thirdly, there is the unfolding of the act of meaning itself: the instantial construc-
tion of meaning in the form of a text. This is a stochastic process in which
the potential for creating meaning is continually modified in the light of what
has gone before; certain options are restricted or disfavoured, while others are
improbable or opened up. We refer to this as the logogenetic time frame, using
logo(s) in its original sense of ‘discourse’.

These are the three major processes of semohistory, by which meanings are continu-
ally created, transmitted, recreated, extended and changed. Each one provides the envi-
ronment within which the ‘next’ takes place, in the order in which we have presented
them; and, conversely, each one provides the material out of which the previous one is
constructed: see Figure 3.1–3.6 [reproduced here as Figure 3.15].
As the upward pointing arrow suggests, the individual’s (transfinite) meaning poten-
tial is constructed out of (finite) instances of text; the (transfinite) meaning potential of
the species is constructed out of (finite) instances of individual ‘meaners’. Following
the downward arrow, the system of the language (the meaning potential of the species)

152
 Resource guides

Figure 3.15 Processes of creating meaning located along the cline of instantiation – three semo-histories
(Halliday and Matthiessen, 1999/2006: 18)

provides the environment in which the individual’s meaning emerges; the meaning
potential of the individual provides the environment within which the meaning of the
text emerges.

This framing of semogenesis in terms of the cline of instantiation enables us to relate logogenesis
to ontogenesis, and ontogenesis to phylogenesis; it also enables us to compare and contrast the
diferent patterns in the creation of meaning, such as the expansion of the meaning potential by
the gradual disassociation of systemic variables, as described by Halliday in various places (e.g.
Halliday, 1991b).
Systemic functional studies dealing on the three semo-histories are exemplified in Table 3.14,
together with any names used in “mainstream” linguistics. There is perhaps no comparable con-
ception of an area of study of semo-history outside SFL, but some strands of semo-history have
of course long been pursued under widely recognized labels such as historical linguistics. The
term “language acquisition” has become used widely since the 1960s, but Halliday and other
systemic functional linguists working in the area have preferred to avoid it because of undesir-
able implications of the metaphor of acquisition instead discussing it in terms of ontogenesis or
language development – learning how to mean. The problems with the conception of language
development as “language acquisition” have also been pointed out by experts outside SFL,
notably by Larsen-Freeman (2011), who gives examples of how scholars and students have been
misled by this metaphor.
The references in Table 3.14 include both particular studies and surveys of studies in SFL.
The different strands within research dealing with semo-history have not been given equal
attention. Broadly speaking, there is a complementarity between SFL research and research
outside SFL (Williams and Lukin, 2004):

• In terms of phylogenesis, historical linguistics has of course a long tradition, established

as part of the motif of romanticism and evolution in the 19th century, and there are a vast

153
 Table 3.14 Examples of studies of the three semo-histories in SFL

Timeframe Mode of genesis Focus Name studies in References

“mainstream” linguistics

phylogenesis evolution deep time language evolution Matthiessen (2004b); Rose (2005); Halliday (2010)
up to around 8,000 years historical linguistics [evolution of registers:]
before now scientific discourse: Halliday (1988a, 2010); Banks (2005, 2008);
O’Halloran (2005)
news reports: Nanri (1993); Urbach (2013)
advertisements: Starc (2010, 2015)
[lexicogrammatical features:] ‘existential’ clauses with there:
Martínez-Insua (2013)
[overview:] Banks (2019b)

Resource guides
grammaticalization studies [on the systemic conceptualization of grammaticalization:]
Matthiessen (1995a, 2004a), Halliday (2008a);
154

[case study:] Martínez-Insua (2013); Mwinlaaru and Yap (2017)

ontogenesis growth lifetime: early childhood language acquisition [case studies:] Halliday (1975a; 1984c), Painter (1984, 1999,
2003), Torr (1997, 1998); Torr and Simpson (2003);
Derewianka (1995, 2003)
[overviews:] Painter, Derewianka and Torr (2007), Torr (2015),
Painter (2017), Williams (2019)
[language-based theory of learning:] Halliday (1993b)
lifetime: school years Christie and Derewianka (2008); Christie (2012)
lifetime: academic career Guerra Lyons (2021)
logogenesis unfolding text time: logogenetic (not really an area of Matthiessen (1993b); Martin (1999b); Körner (2000);
patterns study, unless one turns Matthiessen (2002b); Zappavigna (2011); Caldwell and
to computational Zappavigna (2011); Lam (2015) with reference to context;
linguistics) cf. also explorations of the complementarity of synoptic and
dynamic views of texts: Martin (1985a); Ventola (1987)
text time: process of [in computational modelling:] Matthiessen and Bateman (1991);
unfolding O’Donnell (1990, 1999); O’Donnell and Sefton (1995)
 Resource guides

number of studies outside SFL (for SFL in relation to diachronic studies in general, see
Banks, 2019b; for deep history – the evolution of language, see Matthiessen, 2004b). Here
one special contribution within SFL has been the study of the evolution of particular reg-
isters, pioneered by Halliday (1988a) in his study of the evolution of scientific English, and
taken further e.g. by Banks (2008) on scientific language and Nanri (1993) and Urbach
(2013) on the evolution of news reports. The potential for systemic functional contributions
to the study of language change is considerable, also in view of the probabilistic conception
of language: language evolution can be studied in terms of changing probabilities – see later.
• In terms of ontogenesis, there have certainly been a vast number of studies of “language
acquisition” since the 1960s, but at least the early ones tended to be focussed on “form”
rather than meaning, taking the transition into the mother tongue as the starting point
and thus overlooking the early phase or phases of children learning how to mean, and
they have tended to be cross-sectional rather than longitudinal (cf. Ortega and Byrnes,
2008, on longitudinal L2 studies); so SFL contributions have been unique in construing
ontogenesis as “learning how to mean” (Halliday, 1975a, 2003a) with evidence from lon-
gitudinal case studies, going back in children’s lives to their protolinguistic phase and even
pre-protolinguistic activities (for overviews, see Painter, Derewianka and Torr, 2007; Torr,
2015; Painter, 2017; Williams, 2019; Thwaite, 2019).
• In terms of logogenesis, there is not much in the way of theoretical models and empirical
studies outside SFL in linguistics (with certain interesting exceptions like Kempson, Meyer-
Viol and Gabbay, 2001) – contributions tending to come from computational linguistics,
e.g. parsing algorithms and text planning, although they are not theorized linguistically. In
SFL, there have been a number of contributions to the conception and study of logogen-
esis drawing on, or inspired by, computational SFL, e.g. Matthiessen and Bateman (1991);
O’Donnell (1990, 1999, 2013); Matthiessen (1993b, 2002b); Zappavigna (2011); Caldwell
and Zappavigna (2011).

Beyond the actual contributions made in SFL so far to the study of semo-history, it is important
to consider the potential for further studies created by systemic functional theory and the nature
of descriptions of particular languages. Within the phylogenetic timeframe, we can interpret
Ellegård (1953) as a preview of what’s possible within a systemic functional version of histori-
cal linguistics. He studied the emergence of do as a finite operator, “periphrastic do” providing
“do support”, over a period of roughly 250 years, showing how its presence is determined by a
combination of features from the systems of mood type and polarity and how changes can be
described probabilistically as gradual changes in relative frequencies in texts from late Middle
English to Modern English around 1700: see Figure 3.16. (Ellegård’s work was carried out long
before Labov’s introduction of the notion of change in progress.)
Ellegård’s (1953) pioneering study is entirely consistent with Halliday’s theoretical model of
change involving the disassociation of associate variables and changes in probabilities, collected
in Halliday (2005c).

3.3 Organized according to the spectrum of metafunction

In addition to the hierarchy of stratification and the cline of instantiation, there is one more
global semiotic dimension – the spectrum of metafunction, or functional diversification: see
Figure 3.17. The spectrum of metafunction is extended from the meaning potential of language
to texts as acts of meaning. In terms of the hierarchy of stratification, functional diversification
permeates context and the content plane of language in the first instance, not the expression

155
 Resource guides
156

Figure 3.16 Phylogenesis as gradual change in systemic probabilities derived from relative frequencies in a historical corpus, based on Ellegård’s (1953) study
 Resource guides

Figure 3.17 Metafunctional spectrum in relation to the hierarchy of stratification and the cline of instantiation,
permeating context and the content plane of language and extending from potential to
instance

plane; but within the expression plane, it does penetrate into the higher ranks of phonology,
“prosodic phonology”, and the different functional modes of expression (see Halliday, 1979a)
can be recognized throughout the expression plane.

3.3.1 Content plane resources distributed metafunctionally

Adopting the spectrum of metafunction as a “projection” system, we can survey the resources
of the content plane metafunction by metafunction, cutting across the stratal boundary between
semantics and lexicogrammar: see Table 3.15. The publications listed in the table provide fairly
general overviews; there are, of course, numerous studies involving the analysis of texts using the
different metafunctional systems.
The description of systems within the different metafunctions started in the 1960s with
Halliday’s work on English (e.g. Halliday, 1967/8, 1970a), but there were also some accounts
of other languages (see Table 9.5). Once the first edition of Halliday’s Introduction to Functional
Grammar had appeared in 1985, supported by his guide to lexicogrammatical text analysis (Hal-
liday, 1985a, 1957), the descriptions of the different metafunctional resources of the grammar
were generally available to be used in text analysis and to form the basis for further descriptions
of English and to serve as a model for systemic functional descriptions of other languages.

3.3.1.1 Textual systems

Textual systems provide the resources for enabling the creation of ideational and interper-
sonal meanings as a flow of information (or “flow of discourse”) in context. Since this flow is

157
 Table 3.15 Resources according to the spectrum of metafunction – key semantic and lexicogrammatical systems, with relevant systems of intonation also specified

Metafunction Study/system References

textual cohesion as lexicogrammatical resource: Halliday and Hasan (1976); Halliday and Matthiessen (2014: Chapter 9)
as semantic resource: Martin (1992a) – identification: Martin and Rose (2007: Chapter 5)
overview and review: Clarke (2017)
periodicity Martin and Rose (2007: Chapter 6)
theme Halliday and Matthiessen (2014: Chapter 3); Matthiessen (1992, 1995b); Ghadessy (1995); Forey and
Sampson (2017); theme in the Cardiff Grammar: Huang (2017); theme in Spanish: Arús (2017)
information Halliday and Matthiessen (2014: Chapter 3)
interpersonal negotiation and mood Halliday (1984b); Martin (1992a: Chapter 2); Hasan (1996b); Thibault (1995); Halliday and
Matthiessen (2014: Chapter 4); Andersen (2017); mood in Japanese: Teruya (2017); Martin and
Rose (2007: Chapter 7)

Resource guides
modality Halliday (1970a); Halliday and Matthiessen (2014: Sections 4.5 and 10.2); Butler (1982, 1988);
Verstraete (2001)
158

appraisal overviews: Martin and White (2005); Martin and Rose (2007: Chapter 2); White (2015); Oteíza
(2017); Hood (2019); further work on affect: Benítez-Castro and Hidalgo-Tenorio (2019)
Hunston and Thompson (1999) [an edited volume with both SFL and non-SFL contributions];
Hyland and Guinda (2012) [an edited volume with both SFL and non-SFL contributions]
affiliation Knight (2010a, 2010b); Zappavigna (2012, 2014)
tone Halliday and Greaves (2008)
ideational experiential ideation base Halliday and Matthiessen (1999/2006)
ideation Martin (1992a: Chapter 5); Martin and Rose (2007: Chapter 3)
transitivity Halliday and Matthiessen (2014: Chapter 5, Section 10.5); Davidse (1999, 2017); transitivity in the
Cardiff Grammar: Neale (2017)
logical logical organization of text Mann, Matthiessen and Thompson (1992); Matthiessen (forthcoming b);
Martin (1992a: Chapter 4); Martin and Rose (2007: Chapter 4)
clause complexing: Matthiessen (2002a); Butt and Webster (2017)
taxis and logico-semantic type
 Resource guides

characterized by peaks of prominence and troughs of non-prominence of different textual sta-

tuses, the metaphor of “flow of information” could alternatively be replaced by a related one –
“wave of information” (see Halliday, 1985g; Matthiessen, 1992). Textual statuses are assigned to
ideational and interpersonal meanings in text in context by speakers to guide their listeners in
processing these meanings (and also to help themselves in producing texts), as shown by Mat-
thiessen and Bateman (1991), Matthiessen (1992) and Halliday and Matthiessen (1999/2006).
Textual systems cover resources within the content plane that have been characterized outside
SFL in terms of functional sentence perspective, communicative dynamism, topicality, informa-
tion structure (borrowed from Halliday), pragmatics (as in Dik’s, 1978, functional grammar –
excluding the interpersonal) and reference.
Viewed from the point of view of lexicogrammar, textual systems are of two kinds, as shown
in the function-rank matrix in Table 3.9:

• textual systems realized by textual structures: the systems of theme (clause) and of informa-
tion (information unit);
• textual systems realized by cohesive devices creating cohesive links: the systems of cohesion
(grammatical: conjunction, reference, ellipsis-&-substitution; lexical: lexical cohe-
sion), shown in the right-most column in the table.

Viewed “from above”, from the point of view of semantics, textual systems are, arguably, all
concerned with progression in text – the progression of successive textual statuses of difer-
ent kinds, creating patterns of periodicity (see Martin, 1992a, 1996a; Martin and Rose, 2007:
Chapter 6). Semantically, the system that corresponds to the cohesive system of conjunction
has been interpreted as logical – either modelled as Martin’s semantic system of conjunction
specifying explicit and implicit “logical connections” in text (Martin, 1992a; Martin and Rose,
2007: Chapter 4) or modelled as logico-semantic, or rhetorical-relational, complexing in text
by means of Rhetorical Structure Theory (e.g. Mann, Matthiessen and Thompson, 1992; Mat-
thiessen, 1995a) and its systemicized version (Matthiessen, forthcoming b).
The textual system of theme has been given a great deal of attention since the early 1980s,
stimulated by Fries’s (1981) pioneering study of the relationship between the selection of Theme
clause by clause and the “method of development” of text – followed up in a number of pub-
lications, e.g. Fries (1994, 1995, 2002); see e.g. Ghadessy (1995), Hasan and Fries (1995). This
work also included discussions about how far Theme extends into the clause, as in Berry (1996).
Forey and Sampson (2017) give an overview of work on Theme in SFL, mainly descriptions
of English, also noting accounts of thematic organization “above the clause” (pp. 134–137) (cf.
Thompson, 2007).
The textual system of information assigns textual statuses of a structural nature just like the
system of theme, but its domain is the information unit rather than the clause; so the systems
of theme and information are independently variable, with the usual unmarked and marked
combinations characteristic of simultaneous systems. The system of information was originally
described by Halliday (1967, 1967/8), and is part of the description presented in his Introduction
to Functional Grammar. It is treated in some detail in Halliday and Greaves (2008), and also in
some other systemic functional descriptions of intonation. Halliday’s term “information struc-
ture” has been adopted outside SFL, so there are a number of publications that may be consulted,
e.g. Schwabe and Winkler (2007).
The textual systems collectively known as cohesion began to be described already in the
1960s by Ruqaiya Hasan and Michael Halliday, and were presented in their classic book Cohe-
sion in English (Halliday and Hasan, 1976), which has been a resource in text-based research

159
 Resource guides

in a number of areas. This work was supplemented by Hasan’s (1984b) account of cohesive
harmony, which captures interaction among lexico-referential cohesive chains. In Halliday
and Hasan (1985), Hasan offers a typology of cohesive systems that also supplements Halliday and
Hasan (1976).

3.3.1.2 Interpersonal systems

Interpersonal systems provide interactants with the resources for enacting their roles and rela-
tions as meaning, including their value systems, empowering them to negotiate and calibrate
the tenor of their relationship. They cover meanings that have been characterized outside SFL
in terms of speech act, connotation, stance, evaluation, sentiment and also (in relation to tenor)
politeness and impoliteness, rapport and empathy.
Interpersonal systems have also been given considerable attention within both lexicogram-
mar and semantics, as in Halliday’s (1984b) account of the nature and ontogenesis of dialogue,
involving both the semantic system of speech function and the lexicogrammatical system of
mood, supported by the description of tone within prosodic phonology (e.g. Halliday, 1967,
and Halliday and Greaves, 2008). All three are brought together in Halliday’s IFG (e.g. Halliday
and Matthiessen, 2014: Chapter 4). Related grammatical systems have also been described, with
particular focus on the system of modality (e.g. Halliday, 1970a; Halliday and Matthiessen,
2014: Section 4.5 and Section 10.3; Butler, 1982, 1988).
In the late 1980s, research started on what was to become a project describing interpersonal
lexis. Poynton (1990, 1996) described the interpersonal resources of the nominal group (with
careful attention to the tenor parameter within context), and this was one source in the devel-
opment of the system of appraisal: Martin (2000), Martin and White (2005), Martin and Rose
(2007: Chapter 2); White (2015), Oteíza (2017), and Hood (2019), who also includes references
to work on languages other than English (pp. 401–402). The system of appraisal covers subsys-
tems realized primarily through lexis – attitude in particular (judgement, affect and apprecia-
tion), but also subsystems realized primarily through grammar – graduation (force and focus)
and engagement.22 Benítez-Castro and Hidalgo-Tenorio (2019) propose a revised description
of the attitudinal region of affect.
The system of appraisal has been treated as an interpersonal semantic system (thus located
at the intersection of semantics and interpersonal in Table 3.16). In Halliday and Matthies-
sen (2004, 2014), modality, mood evaluations, propositional and speech functional comments,
and attitude are all interpreted lexicogrammatically as different modes of interpersonal assess-
ment, modality being the most highly grammaticalized subsystem in English. In his discussion
of the complementarity of grammar and lexis, Halliday (2008b: 48-) discusses interpersonal
lexicogrammar:

So the lexicogrammar adopts two contrasting perspectives for construing all this com-
plexity. The one is specific and openended; hence flexible, but low in information:
this is the lexical perspective, good for seeing phenomena as particular. The other is
general and systemic: hence high in information, but creating closure: this is the gram-
matical perspective, good for seeing phenomena as generality. The two perspectives are
complementary; any phenomenon can be images of the whole.
Interpersonally, the same two perspectives come into play. Some interpersonal
meanings are highly generalized, like the enactment of dialogic roles (speech func-
tion); Hasan (1992 [1992a]) and Hasan and Cloran (1990) present semantic networks
for interpersonal systems of speech function (questions and commands) showing

160
 Table 3.16 Function-stratification matrix

Stratum Ideational Interpersonal Textual

Logical Experiential

context field tenor mode

semantics logical organization of figuration (Halliday negotiation, including progression; periodicity (Martin,
text: logico-semantic and Matthiessen, speech function 1992a; Martin and Rose, 2007)
complexing Mann, 1999/2006); (Halliday, 1984b;
Matthiessen and ideation (Martin, Berry, 1981; Martin,

Resource guides
Thompson, 1992; 1992a; Martin and 1992a; Martin and
Matthiessen, 1995a, Rose, 2007) Rose, 2007)
161

forthcoming b); appraisal (Martin and

conjunction (Martin, White, 2005; Oteíza,
1992a; Martin and 2017)
Rose, 2007)
lexicogrammar taxis and logico-semantic transitivity (Halliday mood; modal assessment theme and information (Halliday cohesion (Halliday and
type (Halliday and and Matthiessen, (including modality) and Matthiessen, 2014: Ch. 3) Hasan, 1976; Halliday
Matthiessen, 2014: 2014: Ch. 5) (Halliday and and Matthiessen, 2014:
Ch. 7) Matthiessen, 2014: Ch. 9)
Ch. 4)
phonology (tone sequence and – tone (Halliday and tonicity (Halliday and Greaves,
concord) (Halliday and Greaves, 2008: 2008: Sections 3.4.2, 3.4.3, 5.1)
Greaves, 2008: Sections 3.4.1, 5.2)
Section 5.3)
 Resource guides

their realizations in the grammar. With options in the way something is evaluated
(“I approve/I disapprove”), or contended (“I agree/I disagree”), the borderline
between grammar and lexis is shaded over; systems of appraisal, as described by Mar-
tin and White (2005), represent more delicate (more highly differentiated) options
within the general region of evaluation. . . . There are then mixed systems where the
two perspectives intersect; for example the system of modality, in which the various
degrees of probability and usuality have multiple realizations, including those where
items which elsewhere function lexically are organized into systemic sets – wordings
like certainly, perhaps, I think, I’m convinced and many others. . . .
Thus in the interpersonal domain, the organization of meaning into two regions,
the lexical and the grammatical, is less polarized; there is not such a clear demarcation
between the general and the particular in the management of human relationships.
The two contrasting perspectives are still distinct; but it becomes more apparent that
the difference between them is one of depth of focus, not one of discontinuity in the
phenomena themselves.

Halliday’s point in the last paragraph can be seen clearly when we interpret both modality and
appraisal (in particular, attitude) as manifestations of the system of modal assessment. The
system of modal assessment is extended from grammar to lexis, with modality being more
highly grammaticalized and attitude being more highly lexicalized. When viewed in this way,
modality and attitude can be seen to shade into one another within certain types of modal assess-
ment (cf. Matthiessen, 2007c).

3.3.1.3 Ideational systems

Ideational systems provide speakers with the resources for construing their experience of the
world around and inside them as meaning. While interpersonal systems are systems for action – or
rather, interaction, ideational systems are thus systems for reflection, for construing theory. The
ideational metafunction provides two modes of construing experience – the logical mode,
where experience is construed serially as chains (as in the clause complex), and the experien-
tial mode, where it is construed organically as configurations (as in the transitivity structure of
the clause). These two ways of modelling experience complement one another, and languages
mix them in different ways (cf. Halliday and Matthiessen, 1999/2006: Chapter 7; Matthiessen,
2004a). In fact, there is also registerial variation within languages; speech tends to deploy the
logical model where writing tends to deploy the experiential one, as shown by Halliday (1985f,
1988d) and Matthiessen (2002a). The tendency in writing is strengthened by ideational gram-
matical metaphor (see Table 3.17).
While there are early indications of the description of the logical mode of the ideational
metafunction (e.g. Halliday, 1965, 1977b; Hudson in Huddleston et al., 1968) before the
extended description in Chapter 7 of Halliday’s IFG, most early work focussed on the experi-
ential clause system of transitivity (for an overview of successive accounts, see Table 5.5). Hal-
liday (1967/8) outlined the description, including an early version of the system of process type
and the interpretation of the system of transitivity in terms of the complementary transitive
and ergative perspectives, and his analysis of passages from a novel by Golding, The Inheritors, also
advanced the understanding of the system as a resource for construing (angles on) experience
(Halliday, 1971). Halliday (1976a) includes a short paper on the system of process type, and by
the time the first edition of Halliday’s IFG was published in 1985, the description of the system
of transitivity in English was in place, and this version has become a reference account both

162
 Resource guides

Table 3.17 Lexicogrammatical metaphor differentiated in terms of metafunction – ideational, interpersonal,

and in terms of delicacy – grammatical zone, lexical zone

Metafunction Grammatical zone Lexical zone

ideational ideational grammatical metaphor – e.g. ideational lexical metaphor – e.g.

a sequence realized as if it were a figure kinds of consciousness realized as
by a clause rather than congruently by a if they were states or levels within
clause complex vertical space (to be high/low; elated;
higher consciousness)
Perhaps the most exposed, easily observable This is the traditional notion of
type is nominalization (involving metaphor, now also known as
transcategorization), which has of “conceptual metaphor”.
course been recognized traditionally
and discussed widely (also in the classic
transformational literature, where it
played a role in the move towards more
lexicalist approaches).
interpersonal interpersonal grammatical metaphor – interpersonal lexical metaphor –
e.g. a command realized as if it were a e.g. a personal endearment realized
question by an interrogative clause instead as if the person were a cute animal
of congruently by an imperative clause (possum) or sweet food item (cream
puff) or by an interpersonal flip to a
negative connotation (bastard)
Metaphors of mood have been widely
discussed under the heading of “indirect
speech acts”.

for descriptive elaborations and alternatives, and for use in text analysis: see Davidse (2017),
Matthiessen (2018a). Halliday’s description of the system of transitivity has been used to illu-
minate particular areas of experience – notably Halliday’s (1998a) description of “the grammar
of pain”, and also Kashyap and Matthiessen’s investigation of the construal of our experience of
space (Matthiessen and Kashyap, 2014; Kashyap and Matthiessen, 2017, 2019). Davidse (2017:
87–89) summarizes some of the descriptions of transitivity in languages other than English, and
Matthiessen (2004a) sets for generalizations about transitivity systems in languages around the
world (cf. also Teruya and Matthiessen, 2015; Kashyap, 2019).
Halliday’s IFG account has been extended and illuminated by Davidse’s (1999) studies of
the nature of the complementarity of the transitive and ergative models of transitivity (Davidse,
1992a) and of existential clauses (Davidse, 1992c), the intricacies of (identifying) relational
clauses (Davidse, 1992b, 1996a), the construal of possession (Davidse, 1996b). For further inves-
tigation of identifying relational clauses, based on evidence from a registerially varied corpus, see
Zhang Ruihua (2016). Matthiessen (1991b) explores relational clauses as part of the “grammar
of semiosis”, and he reports on quantitative studies of selections in transitivity in Matthiessen
(1999, 2007c [2006]). Matthiessen (2014c) outlines a long-term project classifying and inter-
preting Levin’s (1993) verb classes in terms of Halliday’s description of the system of transitiv-
ity; see also Lemmens (1988) for the description of particular verb classes.
In the development of the description of the “Cardiff Grammar” framework, Robin Fawcett
and his colleagues have drawn on Halliday’s description of transitivity, but proposed alternatives
at certain points, e.g. the location of the borderline between ‘material’ and ‘relational’ clauses

163
 Resource guides

(Fawcett, 1987); see Neale (2017) and Schulz and Fontaine (2019: 249–251) for a summary of
the Cardiff approach to transitivity in English, and Halliday and Matthiessen (1999/2006: 504–
505) for a discussion of Fawcett’s (1987) description. They have documented criteria for dis-
tinguishing different transitivity types, presented in Fawcett (2011), Neale (2017). Neale (2002,
2006) has extended the system of process type in delicacy, drawing on the Birmingham work
on the COBUILD verb dictionary.
In the work on systems within the ideational metafunction, Kristin Davidse has suggested a
bridge between SFL and Ron Langacker’s (e.g. 2008, 2009, 2013) Cognitive Grammar, and a
number of studies of a “functional-cognitive” nature have now appeared, e.g. Laffut (2006) on
three-participant clauses, Lemmens (1988) on causative constructions, drawing also on Levin
(1993), and on projection, Vandelanotte (2009).
The semantic system corresponding to the lexicogrammatical system of transitivity has
been discussed in a number of the publications on transitivity mentioned earlier. It has also been
sketched under the name of figuration as part of the ideation based proposed by Halliday and
Matthiessen (1999/2006) as a model of ideational semantics. The ideation base makes it pos-
sible to interpret metaphors of transitivity in terms of the relationship between semantics and
lexicogrammar.
Turning now to the logical mode of the ideational metafunction (cf. Ellis, 1987a; Taylor
Torsello, 1996), we can note again the early contributions that previewed aspects of Halliday’s
description of clause complexing, group formation and also word structure. Halliday (1965)
differentiates multivariate structure and univariate structure; as later publications make clear
(e.g. Halliday, 1978, 1979b), multivariate structure covers textual, interpersonal and experiential
structure, whereas univariate structure is engendered by the logical mode of the ideational meta-
function. He gives examples of univariate structures in the formation of clause complexes, groups
and words (in particular, patterns of derivation). Hudson, in Huddleston et al. (1968), gives an
account of the clause complex in scientific discourse, noting distinctions between parataxis and
hypotaxis, and specifying which types of hypotactically dependent clauses can be “highlighted”
by means of theme predication. And Hudson (1971) describes “complex sentences” as part of a
step towards a non-transformational generative grammar.
Halliday (1977b) reveals aspects of his account of clause complexing in the analysis of a text.
The fully developed version appears in Chapter 7 of An Introduction to Functional Grammar (Hal-
liday, 1985a), complemented by his description of group and phrase complexes in Chapter 7A.
His account of the clause complex is illustrated by a number of examples in the analysis of texts,
including the dialogue in Halliday (1985b), the poem by Tennyson in Halliday (1988b) and the
fund raising letter in Halliday (1992a). It is further illuminated by his studies contrasting proto-
typically spoken and written language: Halliday (1985f, 1987a) and also his essay on language
and the order of nature (Halliday, 1987b).
Butt and Webster (2017) provide an introduction to the logical mode of the ideational meta-
function and to logical systems operating in complexing and groups, but they do not give a gen-
eral overview of descriptions of logical systems in SFL. Here we can note further contributions
concerned with key aspects of the logical mode of the ideational metafunction:

• probabilistic studies: Nesbitt and Plum (1988), discussed by Halliday (1991b); Matthiessen
(2002a, 2007c [2006]);
• clause complexing in relation to rhetorical-relational (logico-semantic) complexing in text: Mat-
thiessen and Thompson (1988), Matthiessen (1995a, 2002a), Matthiessen and Teruya (2015b);
• studies of projection (in various language): Vandelanotte (2009); Teruya (2010); Arús-Hita
et al. (2018);

164
 Resource guides

• clause complexing in aphasic discourse: Armstrong (1992);

• the complementarity of the interpersonal and logical modes of organization in exchanges
in dialogue: Ventola (1988), Eggins (1990);
• work on languages other than English: Ouyang (1986) on Chinese; Martin (1996c) on
Tagalog; Teruya (2006, 2007) on Japanese; and in translation studies: Choi (2013).

3.3.1.4 Semantics and lexicogrammar

Since the relationship between semantics and lexicogrammar as the two content plane strata of
language is a natural one (see e.g. Figure 3.6), systems listed in Table 3.15 within the different
metafunctions are both semantically and lexicogrammatically relevant and in areas where the
description has advanced far enough, they are typically paired, as in the case of speech function
and mood, and of figuration and transitivity. This “pairing” occurs where semantic domains
correspond to grammatical ones – which is determined by the upper bound of the grammatical
rank scale. The upper bound is the clause, and clauses linked togico-semantically into “chains”
of clauses, i.e. tactic combinations of clauses into clause complexes. Thus clauses are paired in
the semantics with messages, moves and figures, and clause complexes with sequences. There
are no grammatical domains above that of the clause and the clause complexes, but semantics
extends compositionally beyond messages, moves, figures and sequences all the way up to whole
texts, and even macro-texts (cf. Martin, 1992b; Muntigl, 2006), which can be interpreted as text
complexes. The relationship between the compositional scales in semantics and lexicogrammar
is depicted in Figure 3.18.

Figure 3.18 Composition in semantics in relation to composition in lexicogrammar – partly overlapping

rank scales

165
 Resource guides

From the point of the metafunctional organization of the content plane of language, one funda-
mental question is whether semantics is like lexicogrammar in having one rank scale that unifies
the distinct metafunctional patterns. That is, within lexicogrammar it can be argued that the
rank scale is oriented towards composition of the experiential kind – i.e. constituency – but it
can still serve as a framework for textual and interpersonal patterns. However, the organization
within semantics is “looser”, at least above the semantic domain realized by the grammatical
domain of the clause (and clause complex), and the different metafunctions may need to be
represented by different compositional scales:

• logical: text “composed of ” a logico-semantic (rhetorical-relational) complex;

• textual: text “composed of ” waves of textual statuses – in particular, of thematicity and of
newsworthiness, creating a kind of undulating progression or periodicity;
• interpersonal: text “composed of ” a succession of exchanges (with monologic texts as the
limiting case, consisting of mostly initiating moves) and of prosodies of varying extent;
• experiential: text “composed of ” chunks according to the nature of the experience being
construed, e.g. episodes in stories and recounts, methods in procedures, and “taxa” in taxo-
nomic reports.

The complementarity of these metafunctional compositional scales in semantics needs to be

explored in further text-based research. On the one hand, it is possible that these diferent
metafunctional compositional scales all operate simultaneously; on the other hand, it is possible
that texts vary in orientation, certain types being oriented towards the ideational (and thus field
within context) and others towards the interpersonal (and thus tenor within context): see Hal-
liday (2001b). In addition, we may need to allow for considerable variation across registers in
the compositional hierarchy or hierarchies of semantics. This issue is discussed in Section 5.2.6.
In the metafunctional organization of the content plane, we can note fractal patterns: the
same metafunctional principles of organization are manifested within both lexicogrammar and
semantics, and within different domains along the compositional scales. Halliday’s (1981) dem-
onstration of the analogy in organization between texts and clauses paved the way for the rec-
ognition of fractals within the content system of language, and Matthiessen has suggested that,
in addition to being like a clause, a text is also like a clause complex (e.g. Matthiessen, 1995a,
2002a, forthcoming b). Martin (1996a) shows that the metafunctional modes of expression first
identified within lexicogrammar by Halliday (1979b) also operated within text, and Halliday and
Matthiessen (1999/2006) discuss fractals within the ideational metafunction.
The matrix in Table 3.15 foregrounds metafunctional origin of systems over stratal location,
but it is also helpful to view the systems originating within the different metafunctions according
to their stratal location in a function-stratification matrix. Such a matrix is a global version
of the function-rank matrix used to provide a map of the system of a content stratum (as in Table
3.7 for semantics and Table 3.9 for lexicogrammar), and it typically collapses the local rank scale,
or represents systems only from the highest rank: see Table 3.16.
By differentiating the systems operating within the different metafunctions, we can also bring
out incongruent patterns of realization alongside congruent ones; we can provide a comprehen-
sive account of metaphor. Metaphor is based on the existence of incongruent patterns of realiza-
tion between semantics and lexicogrammar coming into existence alongside congruent ones, and
they involve de-coupling and recoupling of realizational relationships – what Martin (1997) calls
stratal tension (see further, Ravelli, 2003). Naturally, it can only occur between semantic and
lexicogrammatical domains that are “paired” through realization, i.e. its upper bound is that of
the highest-ranking domain of lexicogrammar – that of the clause and its tactic combinations into

166
 Resource guides

clause complexes. (This still leaves open the possibility of metaphors emerging between context
and semantics, of course.) In this respect, all metaphor within the content plane is lexicogram-
matical metaphor. Metafunctionally, metaphor operates within the ideational metafunction
and within the interpersonal one: ideational metaphor and interpersonal metaphor.23
And it is manifested both within the grammatical zone of lexicogrammar (e.g. Halliday, 1985a:
Chapter 10; 1988a, 1998c; Halliday and Martin, 1993; Halliday and Matthiessen, 1999/2006:
Chapter 6; Simon-Vandenbergen, Taverniers and Ravelli, 2003) and the lexical one (e.g. Simon-
Vandenbergen, 2003), giving us four regions of lexicogrammatical metaphor: see Table 3.17.
Simon-Vandenbergen (2003) reports on a study of lexical metaphors used in construing pro-
cesses of saying, and shows that they typically involve interpersonal connotations. So ideational
lexical metaphor may have interpersonal as well as experiential and textual consequences, and
the same is true of ideational grammatical metaphor: for example, if a figure is realized incon-
gruently as if it were a thing by a nominal group, instead of congruently by a clause, it will no
longer be mapped onto a proposition or a proposal, so it loses its status of arguability (cf. Hoey’s,
1999, study of Chomsky’s rhetoric).

3.4 Accounts of semiotic systems other than language

(multimodal studies)
One of the important characteristics of SFL is its engagement with semiotic systems in general,
which is still unusual among linguistic theories; Halliday’s (1978) characterization of language
as a social semiotic system among other such systems paved the way for the study of semiotic
systems other than language (for more of the background to the conception of social semiotics,
including influence form continental European semiotics, see interviews with Halliday in Mar-
tin, 2013b). While the contributions from 1990s and in particular from the last two decades have
become well known not only in SFL but also in other traditions concerned with “multimodal-
ity” (see terminology later), the explorations go much further back, in fact to the 1960s. These
early explorations are important because they demonstrated the potential of SFL to expand its
coverage of the phenomenal realm of semiotic system.
In a way, we can even cite Halliday’s pioneering work on the description of intonation
in English, begun in the late 1950s (Halliday, 1963a, 1963b, 1967), since on the one hand it
represented an aspect of language often treated as marginal by other linguistic theories at the
time or even left out of the picture (with important exceptions such as Pike, 1945) since it did
not fit most theoretical constructs based on syntagmatic patterns and constituency, and on the
other hand it provided a model for handling semiotic resources that are prosodic in nature (cf.
Matthiessen, 2009a). With his pioneering account of protolanguage, begun in the late 1960s,
Halliday (1975a, 2003a) showed how to approach a semiotic system quite unlike “adult” lan-
guage in fundamental ways – human protolanguage. This was again important in its own right,
but also provided a model for exploring other non-linguistic semiotic systems. That is, while
protolanguage is, of course, the first phase in the ontogenesis of post-infancy language (see Sec-
tion 4.5.1), it is also a distinct semiotic system in its own right. (Here we can note that it can
be treated as a model for researchers dealing with semiotic systems other than language even
though they have tended to rely primarily on Halliday’s account of adult language.)
But there was another interesting and significant contribution in the 1960s: Winograd’s
(1968/1981) systemic description of tonal harmony. It is important among other things because
it demonstrates the power of system networks to deal with semiotic systems of a different kind
than language. It may not be very visible today – e.g. it is not referred to by van Leeuwen
(1999) – but from the point of view of modelling semiotic systems it is still completely relevant

167
 Resource guides

(cf. Steiner, 1988). Indeed, outside SFL, its pioneering status has been recognized; discussing
work on automatic harmonic analysis, López (2017: 13) writes:

Researchers in the field mostly mention the approach from Terry Winograd, in 1968,
as the pioneer work in the task of automatic harmonic analysis. This work is not only
important because it is the first and pioneering work in computing a harmonic analy-
sis, but also because it linked the computational techniques used in natural language
processing to music.

Winograd’s (1968/1981: 269) review of the properties of his systemic description (“grammar”)
of tonal harmonies is worth quoting because it can guide systemic work on music and on other
semiotic systems and highlights the value of the systemic part of the account:

1 The structure resulting from realizations is not necessarily a one-dimensional

string but can have several independent dimensions (time, pitch, loudness).
2 At any rank in the grammar a number of different systems can operate indepen-
dently. Thus at note rank, we would have systems of pitch, duration, articulation,
and possibly others.
3 The resulting structure is not segmented in a single way but may have several con-
stituent structures, even in the same dimension. This could account for the analysis of
music into phrase groupings, melodic groupings, harmonic groupings, and so forth.
4 The syntax is based on the use of systems, containing limited sets of mutually
exclusive features. Systems of this sort abound in music, with systems of pitches,
durations, chord types, inversions, key, time signature, and numerous others.

These four points are of fundamental importance in the interpretation, modelling and descrip-
tion of any semiotic system. (In his work on musical harmony, Winograd is, not surprisingly
given his background in computer science and AI, a good deal more explicit in the use of system
networks than has tended to be the case in “multimodal studies” in SFL – but that is definitely
a (systemic) feature, not a bug.)
These explorations can be drawn on also in current work on interpreting, modelling and
describing semiotic systems other than language. They illustrate some of the reasons why SFL is
well positioned to make advances in the study of semiotic systems in general, of particular semi-
otic systems other than language, and of semiotic systems operating together to make meaning
in context in a complementary fashion. But before we continue, let us list some aspects of SFL
that have been and continue to be helpful in making advances:

• the interpretation of semiotic systems as systems for creating, or at least carrying, meaning
in context (rather than the more limiting conception of such systems as sign systems);
• the axial orientation towards the paradigmatic orientation of organization of semiotic sys-
tems as primary – the systemic organization as primary, freeing up accounts from being restricted
by “rules” for defining potentially very different syntagmatic modes of expression (which had
proved crucial in the interpretation of intonation and of protolanguage referred to earlier), while
at the same time including specifications of how paradigmatic patterns are realized by syntag-
matic ones (as in Winograd, 1968/1981), which may reflect different modes of expression;
• the differentiation of meaning into different modes of meaning organized microfunc-
tionally, macrofunctionally or metafunctionally, which may tend towards distinct modes of
expression;

168
 Resource guides

• the recognition of variation in instantiation along the cline of instantiation, opening up

the possibility of interpreting a given semiotic system at different points along the cline – e.g.
as a general semiotic system at the potential pole or as a registerially tailored one somewhere
midway along the cline of instantiation;
• the interpretation of meaners as persons with personal meaning potentials accrued and
adapted to the roles they play in different groups within different institutions.

These points are explicit or implicit in Halliday’s (1978) argument for interpreting language as a
social semiotic system. Indeed, studies drawing on SFL in the late 1980s and in the 1990s (and
beyond) tended to frame their work in terms of social semiotics.
Now, if we set out to explore semiotic systems other than language and language work-
ing together with other semiotic systems, where would be an ideal starting point? Arguably,
we should start with dialogue – with face-to-face exchanges of meanings; this would be both
phylogenetically and ontogenetically motivated. (As SFL studies of early language development
have shown, protolanguage is multimodal from the start, drawing on both vocalization and
gesture within their expression plane.) This is in fact where a group of anthropologists (Ray
Birdwhistell, Edward Hall) and anthropological linguists (George Trager, also Henry Smith
and Charles Hockett) began their pioneering work in the 1950s; they explored the body as a
resource in communication, analysing and describing what we might call somatic semiotic
systems (i.e. semiotic systems with some facet of the body as their expression plane): their
foundational studies are listed in Table 3.4.
There was an interest in detailed micro-analysis of spoken text in face-to-face interaction,
and another important study from this period is Pittenger, Hockett and Danehy (1960). They
provide a detailed analysis of the audio of the first five minutes of a psychiatric consultation
between a therapist and a patient, to shed both psychiatric and anthropological light on it (cf.
also Schwartz, 1962). This is in fact the kind of study one could imagine being undertaken in
SFL together with colleagues from other disciplines – an example of what can be expected of
Appliable Discourse Analysis. Another project from the same period in the 1950s, also involving
a mixture of disciplinary experts, was concerned with the analysis of members of a family one of
whom was a psychiatric patient. After a long delay it was written up as a manual – McQuown
et al. (1971: 2):

This manual constitutes a general introduction to the theory of micro-analysis of inter-

views with a focus on overt behavior, to the individual systems of analysis of the speech
and body-motion of participants in such interviews, to the techniques of manipulating
taped and filmed materials in order to facilitate such analysis, and to the theoretical
frames suitable for the interpretation of the materials and for their use in psychothera-
peutic and other practical applications.

In 1962, a crucial conference on “Kinesics and Paralanguage” was held at Bloomington, Indiana.
In her report on this conference, Bateson (1963: 3) writes:

A great part of the discussion centered on the wider usefulness of the linguistic model
in handling paralinguistic and kinesic data, and the difficulty of isolating units for
analysis outside of language proper, and identifying the level on which they function.
However, a developed theory of communications, or semiotics, would have to include
the exchange of communications through touch, smell, and taste, as well as through
sight and hearing.

169
 Resource guides

In SFL, apart from the pioneering work on intonation and protolanguage, the first sustained
lines of research into semiotic systems other than language were initiated in the 1980s (see
Table 3.19), but unlike the US American work begun in the 1950s (see Table 3.18), SFL
scholars focussed initially on exosomatic semiotic systems using visual semiotic expression

Table 3.18 The study of kinesics, paralanguage and proxemics pioneered by US American anthropologists
and anthropological linguists in the 1950s and 1960s

Somatic semiotic system Domain References

kinesics Birdwhistell (1952: 3): “The following represents an Birdwhistell (1952, 1970)
attempt to review certain methodological aspects
of the study of body motion as related to the non-
verbal aspects of inter-personal communication.
The term kinesics has been chosen to cover the
multilevel approach (physical, physiological,
psychological, and cultural) to such phenomena.”
Poyatos (1993: 132): “Given all the discrepancies
we find in the literature, kinesics can be
defined as: the conscious or unconscious
psychomuscularly-based body movements and
intervening or resulting still positions, either
learned or somatogenic, of visual, visual-acoustic
and tactile or kinesthetic perception, that,
whether isolated or combined with the linguistic
and paralinguistic structures and with other
somatic and objectual behavioral systems, possess
intended or unintended communicative value.”
paralanguage Trager (1961): “One of the main categories is that Trager (1958, 1961)
of the voice qualities. These involve pitch range,
pitch control, rhythm control, tempo and other
phenomena. . . .
Another set of paralanguage categories is that of the
vocalizations distinguished as vocal characterizers,
vocal qualifiers, and vocal segregates.
The vocal characterizers of laughing and crying
probably occur everywhere. . . .
The vocal qualifiers seem to involve only intensity,
pitch height, and extent (or duration). . . .
The vocal segregates, noises of the kind usually
indicated in writing as uh, uhunh, sh, tsk, and
the like, are so much like language sounds that
their description and identification is the most
difficult of all.”
proxemics Hall (1966: 1): “The central theme of this book is Hall (1959, 1966)
social and personal space and man’s perception
of it. Proxemics is the term I have coined for the
interrelated observations and theories of man’s use
of space as a specialized elaboration of culture.”

170
 Resource guides

Table 3.19 Accounts of semiotic systems other than language or of language together with other semiotic
systems

Focus Type References

overviews “classical” painting, sculpture, O’Toole (1994, 2011)

architecture
images Kress and van Leeuwen (1990, 1996,
2006)
publications including handbooks van Leeuwen and Jewitt (2001); Jewitt
SFL contributions (2009, 2014)
but also other
frameworks
journals Visual Semiotics
GeM model multimodal page Bateman (2008a, 2014); Hiippala (2016);
Zhang Peija (2018)
general explicit Bateman, Wildfeuer and Hiippala (2017);
multimodal model Martinec (2005)
general various semiotic Ventola, Charles and Kaltenbacher
collections systems, and (2004); O’Halloran (2004); Royce
multimodality and Bowcher (2006); Amano (2006);
Jones and Ventola (2008); Ventola and
Moya-Guijarro (2009, 2021); Bowcher
(2012)
study of visualization of cf. Doran (2016)
phenomena description
visualization of Matthiessen (2002b); Zappavigna (2011);
analysis Caldwell and Zappavigna (2011)
phenomena paralanguage: gesture, Martinec (2004); Hood (2011);
facial expression Zappavigna et al. (2010); Martin and
Zappavigna (2019)
voice quality Wan (2009, 2010)
comics and graphic Bateman, Wildfeuer and Hiippala (2017);
novels Veloso and Bateman (2013); Bateman
and Veloso (2013); Veloso, Bateman
and Lau (forthcoming); Veloso (2015);
Bateman et al. (2016)
picture books Painter, Martin and Unsworth (2014)
multimodal public Zhang Peija (2018)
health posters
mathematical O’Halloran (2005); Doran (2016)
discourse, including
symbols and
diagrams
print and digital Wong (2019)
media
film Baldry and Thibault (2006); Bateman
and Schmidt (2012); Wildfeuer (2014);
Wildfeuer and Bateman (2017)
171
 Resource guides

planes, the key early ones concerned with descriptions to support analysis being: painting,
sculpture, architecture (“displayed art”) – O’Toole (1994, 2015), and “images”, including
photographs, drawings, paintings – Kress and van Leeuwen (1990, 1996). These foundational
accounts were general in orientation, not tuned to particular registers, but Lemke (1998)
focusses on “scientific text”, and this register (or genre) oriented approach was also adopted
by other researchers (see the following). The work by Kress and van Leeuwen was very much
also part of the development of social semiotics, to which they were central contributors (cf.
Figure 2.6).
As researchers developed descriptions of semiotic systems other than language (Table 3.19)
and of semiotic systems operating together in context (“multimodality”), they have provided
overviews of the semiotic resources of particular systems comparable to Halliday’s (1970a, 1973,
1976a, 1978) original function-rank matrix mapping the lexicogrammatical resources of Eng-
lish. In particular, O’Toole (1994, 2011) pioneered function-rank matrices for painting (1994:
24), sculpture (1994: 36) and architecture (1994: 86). Such overviews are helpful not only as
synoptic charts to readers but also as guides to researchers, as exemplified e.g. by O’Halloran et
al. (2016); see Table 3.20.

3.5 Organized according to axiality (the hierarchy of axis)

Axiality is ordering of organization into the paradigmatic and syntagmatic axes (for references,
see Table 3.1). Since Halliday (1966c), the paradigmatic axis has been treated as primary (see
further e.g,. Matthiessen and Halliday, 2009; Halliday, 2013b; Martin, 2013a, 2015b; Matthies-
sen, 1995a, 2015g). This is the axis that brings out the nature of language as a resource for mak-
ing meaning – language as choice in meaning (Halliday, 2013b), and it is represented by means
of system networks (see Section 9.2.1 for conventions). Syntagmatic patterns are derived from
the systemic organization by means of realization statements (see Section 9.2.2) associated with
systemic terms (options or features in system networks).
Since the conception of language as a resource for making meaning is embodied in paradig-
matic organization, we can conceive of the overall organization of language paradigmatically –
in principle, as a vast system network made up of a number of systems such as speech function
(semantics), mood (lexicogrammar) and tone (phonology). These systems are located at the
potential pole of the cline of instantiation, as shown in the stratification-instantiation matrix in
Figure 3.5. They are dispersed stratally and also metafunctionally (within the content plane), and
can be located by means of various matrices (cf. Figure 1.7):

Stratal matrices:
Content plane: function-rank matrices:
lexicogrammar: Table 3.9, Table 3.10, Table 3.12
semantics: Table 3.7
Expression plane: phonology: Table 1.4
Function-stratification matrix: Table 3.16 (Table 2.3)

Thus maps of language as a resource are systemic in the first instance; the regions of semiotic
space that they display are systems whose points of origin can be defined in terms of stratifica-
tion, metafunction and (within a given stratum) rank (including primary class within ranked
units). In this way, it is possible to produce completely comprehensive overviews of particular
languages. Of course, the systems identified and named in matrices are specified at primary or at
least at low delicacy, so within lexicogrammar, they are grammatical systems rather than lexical

172
 Table 3.20 Function-rank matrices for language and pictorial semiotics adapted from O’Halloran et al. (2016: 10–11) with changes in the characterization of the
linguistic resources

Metafunction Stratum Compositional unit System Description

Language
logical content: semantics text rhetorical relations (logico-semantic system for developing texts and logico-semantic complexes
complexing)
content: clause logico-semantic type and taxis system for linking clauses into clause complexes
lexicogrammar
experiential content: semantics figure [configuration] figuration system for construing a quantum of change as a figure
content: clause transitivity . . . congruently realized by the transitivity system of the
lexicogrammar clause
interpersonal content: semantics move speech function system for enacting a quantum of interaction as a move in a
dialogic exchange

Resource guides
content: clause mood . . . congruently realized by the mood system of the clause
173

lexicogrammar
textual content: semantics text progression (of textual statuses system for managing the flow of meanings as text unfolding
[thematicity, news, identifiability, in context – a succession of messages
continuity and contrast])
content: clause; information unit theme; information . . . congruently realized by the theme system of the clause
lexicogrammar and the information system of the information unit
pictorial
semiotics
experiential content work narrative theme; representation; nature of the scene
setting
episode processes; participant roles; and visual happenings, actions and relations
circumstance roles
figure [as part of posture; dress characterization of the participants
episode]
(Continued)
 Table 3.20 (Continued)

Metafunction Stratum Compositional unit System Description

interpersonal work angle; camera distance; lighting visual effects
episode proportion in relation to the whole happenings, actions and relations with respect to the whole
image; focus; perspective image
figure [as part of gaze – visual address direction of participants’ gazes as internal to image or
episode] external to viewer
textual work compositional vectors; framing the organization of the parts as a whole, with the visual
marking (e.g. framing) of certain parts
episode relative placement of episode; framing position of the happenings, actions and relations in relation
to the whole image, and the visual marking of certain
aspects

Resource guides
figure [as part of relative placement of the figure position of figures in relation to happenings, actions or
episode] within the episode; arrangement; relations, and the visual marking of certain aspects of
174

framing those figures

 Resource guides

ones; but since lexicogrammar is a continuum from grammar to lexis ordered in delicacy, lexical
systems represent elaborations in delicacy of grammatical ones identified in matrices.
Since the resources of a language are vast, system networks are also vast, and they take up a
great deal of space to display graphically, even at fairly low degree of delicacy, as is evident from
Matthiessen’s (1995a) systemic map of the lexicogrammar of English. One way of giving an indi-
cation of the systemic organization of language while at the same time managing the size of the
display is to represent systems only by their names, without the details of their entry conditions,
terms and associated realization statement. Such an index to the systems of the clause described
in Halliday and Matthiessen (2014) is presented in Matthiessen (2021b).
Viewing the semantic, lexicogrammatical and phonological resources of a language by means
of system networks is helpful in the development of the description of the language, in the use
of the description in text analysis, and also in various applications such as the use of system
networks as a diagnostic tool in second/foreign language teaching (e.g. Gibbons and Markwick-
Smith, 1992; Xuan and Matthiessen, forthcoming). However, there are also uses of the descrip-
tion of a language where we need to access the description “from below” in the first instance,
i.e. from the vantage point of syntagmatic patterns. Naturally, this is the case in analysis such as
lexicogrammatical parsing. Here we can compile syntagmatic views for easy access to syntag-
matic patterns (see Kasper, 1988; O’Donnell, 1994; Matthiessen and Nesbitt, 1996; O’Donnell
and Bateman, 2005: Section 3.8; Bateman and O’Donnell, 2015). Such a view can be thought
of as a dictionary view of the lexicogrammatical resources, complementing the canonical the-
saurus view represented by system networks.

Notes
1 Butt and Wegener (2007) discuss Philipp Wegener’s functionalism and contextualism in relation
to Malinowski, but it is not clear to what extent Malinowski’s own conception of context, which
grew out of his fieldwork in the 1910s and efforts to analyse and present texts that he had collected,
was influenced by Wegener’s. Malinowski (1923: 297–298) includes his name only in a general list
of linguists: “Indeed, for some time already, we have had, side by side with the Arts of Language,
attempts at posing and solving various purely theoretical problems of linguistic form and mean-
ing, approached mainly from the psychological point of view. It is enough to mention the names
of W. von Humboldt, Lazarus and Steinthal, Whitney, Max Muller, Misteli, Sweet, Wundt, Paul,
Finck, Rozwadowski, Wegener, Oertel, Marty, Jespersen and others, to show that the Science of
Language is neither new nor unimportant. In all their works, besides problems of formal grammar,
we find attempts at an analysis of the mental processes which are concerned in Meaning. But our
knowledge of Psychology and of psychological methods advances, and within the last years has made
very rapid progress indeed. The other modern Humanistic Sciences, in the first place Sociology
and Anthropology, by giving us a deeper understanding of human nature and culture, bring their
share to the common problem.” Malinowski (1935) develops his account of context further, but it
is still ethnographic in orientation, and he does not mention Wegener. However, Nerlich (1990:
117) writes, “inspired by Wegener, Malinowski had introduced the expression ‘context of situa-
tion’ into English linguistics, which was fruitfully exploited by the so-called London School.” And
Firth (1957b: 94–95) makes the connection to Wegener’s work: “Among the linguists mentioned
in the Supplement, the leading German comparatists are missing but W. von Humboldt, Sweet, and
Jespersen are there, and notably Wegener (1885), to whom Malinowski owed his early notions of
the Situation, Wegener was one of the first to propound what he called the Situationstheorie.” Firth
then goes on to give a brief summary of Wegener’s Situationstheorie, noting that Gardiner referred
to and drew on this work. In his review of Nerlich (1990), Malkiel (1991) characterizes Philipp
Wegener as “a sort of precursor and intellectual hero, a proto- or palaeo-pragmatist” (cf. Levin-
son, 1983: xii, who refers to “Malinowski, Firth and other ‘protopragmaticists’” and he continues
“Pragmatics prior to 1957, it could be argued, was practised (if in an informal way) without being
preached”). Indeed, Wegener’s (1885: 21–27) introduction of his different kinds of situation, “die

175
 Resource guides

Situation der Anschauung”, “die Situation der Erinnerung” and “die Situation des Bewusstseins”,
reads like an armchair-based pragmatics account of a speaker’s mental processes rather than like an
ethnographic approach to context of the kind Malinowski developed. Naturally, Firth (1957b: 103)
distances himself from the mentalist conception of “situation”, commenting after characterizing
Wegener’s account: “In some respects, this analysis has links with my own point of view though I
do not require his trinity of situations, nor do I wish to introduce a reference to retentiveness nor to
consciousness of self or of personal identity.”
2 The well-known framework for describing “social setting”, or (in our systemic functional terms)
context of situation, introduced later by Hymes (1967) is comparable to Firth’s construct (cf. Hal-
liday, 1985d).
3 Capitalization conventions vary in the spelling of field, tenor and mode. If they are treated as the
parameters of context, it makes sense to spell them with initial lower case (cf. ideational, interper-
sonal and textual). If they are used as names of system networks, then it makes sense to spell them
with all capitals, or small, capitals: field, tenor and mode (cf. transitivity, mood and theme).
Sometimes, they are spelt with initial capital, but this does not follow the general convention of
using initial cap to indicate structural functions (cf. Actor, Subject and Theme), and is thus best
avoided.
4 Martin and Rose (2008: 9) claim that “Malinowski interpreted the social contexts of interaction as
stratified into two levels – ‘context of situation’ and ‘context of culture’, and considered that a text
(which he called an ‘utterance’) could only be understood in relation to both these levels.” They
represent this diagrammatically as their Figure 1.2 on p. 10. (There is figure in Butt et al., 2000: 7,
which is similar in that it represents context of situation below context of culture as “extralinguis-
tic levels”.). But it is far from clear that what Malinowski says about the relationship between the
two means that he interpreted them as two stratified levels. The interpretation of the relationship
as one of extension along the cline of instantiation arguably makes more sense. However, since
Malinowski did not operate with either dimension, the issue is really one of how best to make
sense of the complementarity of his “context of situation” and “context of culture” – and from
the point of view of SFL, the cline of instantiation is an excellent candidate, as argued by Halliday
(1990a).
5 In the context of SFL, the “Sydney School” has come to refer to the work by J.R. Martin and his
group (e.g. Rose and Martin, 2012), but there is some variation: Davidse (2017: 89) contrasts the
“Sydney school”, in a more general sense of SFL linguists working in Sydney, with the “Cardiff
school”, referring to the work by Robin Fawcett, Gordon Tucker and their colleagues and students.
This contrast is usually referred to as “Sydney grammar” vs. “Cardiff grammar”.
6 Halliday, personal communication. This can also be inferred from the sequence of publications in
the 1960s and 1970s.
7 The relationship is a challenging one to interpret, theorize and model. Different lexical metaphors
will bring to the fore distinct views of it, but something along the lines of “resonance” would be
helpful since it does not give the impression that there are independent “things” to be “hooked up”
but rather patterns of distinct orders resonating or harmonizing – a concept Matthiessen (1990) had
used [Matthiessen, Christian. 1990. Metafunctional complementarity and resonance in syntagmatic
organization. MS. Dep. of Linguistics, University of Sydney.] and discussed with Hasan. See Hasan
(2014).
8 The contextual parameters of field, tenor and mode are sometimes presented as if they were (only)
properties of the context of situation, but this is misleading. They are extended throughout the
cline of instantiation, from the context of culture to the context of situation, as shown clearly
in various places, including Halliday (1984a: Appendix; 2002a): see Figure 3.11 earlier in this
volume.
9 Let us comment on Halliday’s formulation here. In view of the articulation of context being extended
along the cline of instantiation (e.g. Halliday, 1990a) and the later development of the ordered typol-
ogy of systems operating in different phenomenal realms (semiotic > social > biological > physical,
e.g. Halliday and Matthiessen, 1999/2006: Ch. 13), it would be better to say “it is an instance of the
meanings that make up the context of culture”.
10 Bowcher (2014: 176) writes: “Recently there has been a renewed interest in modelling the
concept of ‘context of situation’ within Systemic Functional Linguistic theory (SFL). This inter-
est has focused on developing a framework of descriptive features and on the viability of using

176
 Resource guides

system networks as the basis for this framework.” However, she follows Halliday (1991d) in
locating context of situation and the instantial pole of the cline of instantiation, with context
of culture at the potential pole. So while unfolding instances create instantial systems, systemic
descriptions of field, tenor and mode must be located towards the potential pole of the cline of
instantiation. She notes (p. 179): “The further along the continuum one moves from system to
instance, the greater the degree of descriptive specificity emerges. However, context of situation
can also be seen as a potential in that as one moves along the cline of instantiation back towards
the context of culture, we can categorise different types of contexts of situation at grosser and
grosser degrees of description, with very generalised categories encompassing a range of specific
but related contexts of situation.” This could be misinterpreted: the cline of instantiation is not
a matter of more specific vs. grosser “degrees of description”, but of more instantial vs. more
systemic. Descriptions of field, tenor and mode systems are thus more likely to be located within
the range along the cline of instantiation between context of culture and situation type – but not
at context of situation.
11 An early demonstration of the “multiplication” of interpersonal meanings in relation to different
tenor settings comes from outside SFL – Brown and Gilman’s (1960) classic study of personal pro-
nouns and their meanings in relation to power and solidarity.
12 Hasan’s “message semantics”.
13 An interesting and valuable pre-systemic prosodic study is worth noting here: the analysis of music
and text in Yurok songs by Robins and McLeod (1956a).
14 It is also important to note (1) that the domains are permeable in the sense that while there are pro-
totypically prosodic and prototypically articulatory phenomena, we need to recognize that certain
features may be treated prosodically or articulatorily in a given language, and (2) that phenomena
that have tended to be treated phonemically in particular in the US American structuralist tradition
and its early generative successor are often better interpreted prosodically, even if the domain of
the prosody is the syllable. In fact, the syllable is a “gateway” between the prosodic and articulatory
domains of phonology.
15 Since a companion volume on other languages has been envisaged.
16 An insightful BA honours thesis concerned with text generation should also be mentioned: Sefton
(1990).
17 Researchers describing languages have, of course, been explicit about their choice of dialect, as in
the case of Barnwell’s (1969) description of Mbembe. For a systemic functional discussion of the
translation of dialects, see Chen (2019).
18 The term “register” has been used in different sense in the work on the “genre model” by J.R. Mar-
tin (e.g. 1992a) and his group. See also Section 6.2 later in this volume.
19 But note that “new technologies” are typically associated with channel within the mode parameter
of context, often allowing for a variety of field and tenor values. For example, Twitter is used in
many ways, including verbal brawling, promotion and advice; see Zappavigna (2012). For “genres
on the web”, see Mehler, Sharoff and Santini (2010), and J.R. Martin’s caution in the foreword (ibid:
vii): “What if genres cannot be robustly characterised on the basis of just a few easily computable
formal features? What if a flat approach to contextual variables and representational features simpli-
fies research to the point where it is hard to see how the texts considered could have evolved as
realisations of the genres members of our culture use to live? Would we be wise to complement flat
computationally based quantitative analysis with thick manual qualitative description and see where
the two trajectories lead us? And do we need to balance commercially driven research with ideologi-
cally committed initiatives (who for example will benefit from the genre informed search engines
inspiring so many of the papers herein)?”
20 This is where registers are located along the cline of instantiation; they are functional varieties
operating in institutions. Sometime scholars relate them to “context of situation” but this involves
a slippage, a kind of knight’s move: texts operate in contexts of situation, but registers operate in
institutions (which can be approached from the instance pole of the cline of instantiation in terms of
situation types).
21 Analyses of texts interpretable as “verbal art” (cf. Hasan, 1985b, 2007) are often to be found under
the heading of “stylistics” or “style”, e.g. Spencer and Gregory (1964), Hasan (1964, 1975), Prakasam
(1982a), Cummings and Simmons (1983), Birch and O’Toole (1988), Adejare (1992), Lukin and
Butt (2009).

177
 Resource guides

22 It is important to distinguish between the interpersonal enactment of feelings, emotions and atti-
tudes and the experiential construal of such sensations (e.g. Matthiessen, 2007c), but the interper-
sonal metafunction may co-opt experiential resources.
23 The possibility of textual metaphor has also been explored in the systemic functional literature, but
examples that have been cited are of a different nature from ideational and interpersonal metaphor.
Critically, they do not involve the “stratal tension” involved in the “as if ” nature of metaphor, i.e.
meaning a realized by wording q, as if it were meaning b.

178
 4
APPLICATION GUIDES

4.1 Domains of research and application

In this chapter, we will provide guides to domains where SFL is being used extensively. It is of
course used in domains other than the ones we are able to cover in detail, and new domains keep
emerging, like neurosemiotics (e.g. García and Ibáñez, 2017; Matthiessen, forthcoming e). This
is due to the nature of SFL as appliable linguistics – it has been designed from the beginning as
a kind of linguistics that has the potential to be applied to address questions and problems that
arise not only in academia but in different institutions in communities throughout human soci-
eties, as outlined by Halliday (2002b, 2008a), including ones that have been examined critically
(Matthiessen, 2012); and appliable linguistics includes Appliable Discourse Analysis (Mat-
thiessen, 2014a). There are now a number of publications that include contributions addressing
the appliability of SFL, e.g. Mahboob and Knight (2010) and Huang, Chang and Dai (2010).
Trevisan and García (2019) identify the use of SFL in experimental stimulus design as a new
“appliable horizon”.
Appliability can be viewed in terms of the needs of potential users. Using the title “linguistics
and the consumer”, which echoes Halliday’s (1964a) “syntax and the consumer”, Martin (1998)
provides a succinct account of the challenges encountered in efforts to make a positive difference
in education – both professional and political challenges, and the experiences he documents can
serve as a guide in many areas other than education as well.
The guides in this chapter are designed to help with research and application in these domains.
Thus while Chapter 2 is concerned with the “content” of SFL, this chapter also foregrounds
methodology – methods of research and application that have been developed and deployed in the
domains that we will discuss, and also methods that could be adopted and adapted in future work.
We begin by sketching general methodological considerations (Section 4.2), and we then
turn five domains as illustrations; the domains that we will cover are:

• system description (language description, comparison and typology) (Section 4.3);

• text analysis (or discourse analysis, including corpus analysis) (Section 4.4);
• language development, language and education (educational linguistics) (Section 4.5);
• translation studies (Section 4.6);
• institution of medicine: clinical linguistics, healthcare communication studies (Section 4.7).

179 DOI: 10.4324/9781315675718-4

 Application guides

After the general methodological considerations in Section 4.2, we will start with a domain
that is very general in nature – the development of a description of a part of the system of
a language or of the whole system (Section 4.3). Here we will thus focus on the potential
pole of the cline of instantiation, as represented diagrammatically in Figure 1.1 earlier in
this volume. Next we move along the cline of instantiation to the instance pole to focus
on the analysis of text using existing descriptions (Section 4.4). In terms of instantiation,
text analysis is more “specific” than system description, but like system description it is still
an activity of a general kind that can be applied to address a wide range of problems in dif-
ferent institutional settings (e.g. Martin, 2009; Martin and Rose, 2007; Eggins and Slade,
2005; Matthiessen, 2014a), uses in healthcare communication studies (e.g. Iedema, 2007;
Slade et al., 2008, 2015) and translation studies being a couple of several growing areas (e.g.
Munday and Zhang, 2017; Kim et al., 2021; Wang and Ma, 2021, 2022; Zhang, 2020).
Text or discourse analysis involving two more semiotic systems, often called MDA (multi-
modal discourse analysis), or more specifically, SF-MDA, has been developed and applied
to an increasing number of contexts since the mid-noughts (e.g. Bateman, Wildfeuer and
Hiippala, 2017; O’Halloran, Tan and Wignell, 2019).
After discussing system description and text analysis, we turn to research tasks and applica-
tions that can be located within one or other of the institutions that make up a culture (cf.
Malinowski, 1944; Matthiessen, 2014a, 2021c). Given the nature of SFL as a kind of “appliable
linguistics”, this chapter could be expanded to include many other areas that can be sited in
particular institutions, e.g. language (and other semiotic systems) in institutions of the law, of
government, administration and management, of entertainment, and urgent concerns that are
manifested across institutions, e.g. the linguistics of recreation and entertainment, the linguistics
of Mother Gaia (ecolinguistics), the linguistics of social transformation and empowerment, and
(on a more depressing note), the linguistics of violence. Here we focus on the following three
domains of studies:

Educational linguistics is concerned with the role of language and of other semi-
otic systems in processes of learning and teaching in institutions of education (Section
4.5) – and also, to some extent, in the institution of the family, both before the start
of formal education and during formal schooling (e.g. Painter, 1999; Christie and
Unsworth, 2005; Rose and Martin, 2012; McCabe, 2021).
Translation studies deal with the recreation of meaning in the processes of
translation and interpreting (Section 4.6; e.g. Steiner, 2005a, 2015, 2019). They
can be related to the comparison of language systems (e.g. Matthiessen, Teruya and
Wu, 2008) – comparison located at the instance pole of the cline of instantiation
(cf. Figure 1.1) – but they also have clear institutional locations relating to the pro-
fessions of translators and interpreters (e.g. in translation agencies) and institutions
within which they provide professional services (e.g. international agencies such as
the UN, service providers such as hospitals).
Finally, healthcare communication studies focus on communication in institu-
tional settings such as hospitals and clinics (Section 4.7; e.g. Slade et al., 2015), and the
same is true of studies dealing with other professional activities (e.g. Forey and Lock-
wood, 2010). In healthcare communication studies, researchers are concerned with
language and other semiotic systems as resources in the performance of healthcare, for
example in medical consultations between patients and healthcare practitioners but
also among healthcare practitioners, as when a patient is “handed over” from one team
to another, and among patients and their friends and family members.

180
 Application guides

One area of institutional application with potentially far-reaching consequences for communi-
ties that has not been a central focus in the history of SFL is language planning, although
Halliday (1975b/2007b: 216–218) includes it in the agenda for sociolinguistic research:

Language planning has two aspects to it, the linguistic and the sociolinguistic. The first
is concerned with the development of language as a system (i.e. with language itself),
the second with the development of language as an institution (i.e. with the relation of
the language to its speakers).

Recently, language planning has been given attention in SFL by Yang and Wang (2017). They
examine language policy in the PRC, but they also ofer a framework by addressing a number of
general research questions. They ask how SFL can be applied to language policy, how this appli-
cation can bring out features of language policy, and, drawing on the new approach they sketch,
how problems that emerge during the implementation of language policies can be addressed.

4.2 General methodological considerations

4.2.1 Phenomena and the study of phenomena

In SFL, researchers have tried to maintain a clear conceptual and terminological distinction
between the phenomena being studied and the study of those phenomena. As Halliday (1984a)
points out, there is a pervasive terminological ambiguity in English, with the same terms being
used for both; some examples are given in Table 4.1. For instance, grammar is used to refer to
both the phenomenon of the subsystem of wording of a language, as in the grammar of Nahuatl,
Nahuatl grammar, and the study of that subsystem, as in systemic functional grammar, transformational
grammar. This can cause confusion, and to provide conceptual and terminological clarity, Hal-
liday (e.g. 1996a) has introduced grammatics as a term for the study of the phenomenon of gram-
mar; here grammatics is to grammar as linguistics is to language.
Systemic functional theory is thus part of the study of linguistic, or more generally
semiotic, phenomena (“systemic functional semiotic theory”) – as are the methods based
on the theory, i.e. systemic functional methodology. Semiotic phenomena are viewed
through the “lens” of the theory, as shown in Figure 4.1. The theory guides the way we (as
researchers) observe, sample, analyse and describe these phenomena. These are not passive
ways of reflecting the phenomena as we study them; rather we project them into semiotic
existence, as it were. A helpful analogy is Albrecht Dürer’s use of a window or grid to
help him draw a “phenomenon” in perspective (his drawing machine): see Figure 4.2. As
observers, analysts, interpreters and so on, we engage with phenomena according to the
“lens” or “grid” that we choose to use. Consequently, even observation is not – and cannot

Table 4.1 Terms for phenomena and the study of phenomena

Phenomena Study of phenomena

semantics (of Western Desert, Chinese, English) (functional, formal, model theoretic) semantics
(English, Japanese, Dagaare) grammar (functional, formal, generative, structural) grammar
(English, Japanese, Dagaare) phonology (functional, phonemic, prosodic, generative)
phonology

181
 Application guides

Figure 4.1 The fundamental distinction between phenomena and the study of phenomena

Figure 4.2 Albrecht Dürer’s representation of a perspective drawing using a “perspective machine”1

be – theory-neutral (see e.g. Matthiessen and Nesbitt, 1996), and decisions regarding the
representation of text in a corpus are highly theoretical ones, different representations fore-
grounding different aspects of text in context (cf. Halliday, 2002c).

4.2.2 The study of phenomena: stages in the research process

Viewing the study of phenomena as a process, we can identify a number of phases. We can begin
by positing an idealized representation in terms of linear time, from past via present to future:

182
 Application guides

Figure 4.3 The research bridge – given present research questions, objectives, goals and problems, we imagine
(envisage) research methods that will enable us to produce the answers, meet the objectives,
achieve the goals and arrive at the solutions

• past: knowledge (findings, results) produced by research already undertaken – surveyed in

literature reviews or enhanced literature reviews, viz. research syntheses, in order to take
stock, locate the research frontier and identify unchartered areas (research gaps);
• present: research gaps forming the basis of research questions (and problems), in turn lead-
ing to research objectives;
• present to future: formulation of research plans of designs specifying research methods
that will enable us to move from questions to answers, from objectives to the meeting of
objectives, from problems to solutions;
• future: answers to research questions (new findings, results), further research needed.

In order to be able to move from present research questions to future answers to these questions
in a research project, we need to imagine (envisage) the research method or range of methods
that will take us to the answers; this is like constructing a bridge from the present to the future:
see Figure 4.3. Aspects of the process of undertaking a doctoral research project are documented
and discussed by Burns, Kim and Matthiessen (2009).
An example of an account of stages in an SFL-informed research project is given in Table 4.2.
(For an overview of the method of systematic review of research, see Figure 2.16.) Since SFL is
a type of appliable linguistics, it supports action research (“AR”). In an overview of the use
of action research in applied linguistics, Burns (2015) characterizes it as follows:

AR is the superordinate term for a set of approaches to research which, at the same
time, systematically investigate a given social situation and promote democratic change
and collaborative participation. Participatory action research (PAR), learning, participant
inquiry, practitioner inquiry and critical action research (CAR), action cooperative inquiry are
all terms broadly underpinned by the assumptions and approaches embodied in AR.

183
 Application guides

Table 4.2 Example of stages in an SFL research project

Nature of Stages References

project

text analysis, Stage 1: data collection – largely automated by means of Python scripts O’Halloran
multimodal Stage 2: multimodal analysis – manual but supported by Multimodal et al. (2016)
Analysis Image software
Stage 3: data mining – automated: machine learning is used to
“develop data mining algorithms for analysis of the large dataset
of multimodal texts with the images from Dabiq”, also drawing on
Wikipedia classification of human socio-cultural life
Stage 4: information visualization: interactive visualization is to
“display the relations between the multimodal texts according to
various dimensions”

The common features they share are to: (a) undertake research to bring about positive
change and improvement in the participants’ social situation; (b) generate theoretical as
well as practical knowledge about the situation; (c) enhance collegiality, collaboration
and involvement of participants who are actors in the situation and most likely to be
affected by changes; and (d) establish an attitudinal stance of continual change, self-
development and growth.

Action research is, naturally, well suited to the needs in educational linguistics (cf. Burns, 1999,
2010). In SFL, educational linguistics research projects are often designed to include positive
intervention as part of the research and the results of the interventions then feed back into
further research, as in the work on genre (e.g. Rose and Martin, 2012) and grammatics (e.g.
Williams, 2005a; French, 2010; Williams and French, 2016). But there are other institutional
domains where phases of research and research-based intervention follow one another, health-
care communication being an increasingly active area of study (e.g. Slade et al., 2015).
In general, we can interpret action research as another instance of the helical or spiral move-
ment Halliday has used in SFL, as discussed in Chapter 1 (cf. Figure 1.2).
Let us now consider the research methods that have been used in SFL – or that could be used
in future research – to take us from research questions to answers.

4.2.3 Research methods

Research methods in SFL have always been closely aligned with the theory – with the systemic
functional “architecture” of language in context. They have been informed by early work, even
pre-systemic functional work, by Halliday and others in the 1950s, continuing the Firthian
tradition of focussing on text in context (e.g. Robins and McCleod, 1956a,b; Mitchell, 1957;
Halliday, 1959) but, in Halliday’s case, also informed by his fieldwork experience in China in the
late 1940s. Fieldwork has, of course, been significant also because of Malinowski’s pioneering
work as a participant observer in the Trobriand Islands in the 1910s, introducing and establishing
modern fieldwork in anthropology (e.g. Kuper, 1983; Young, 2004). Since Firthian linguistics
included work on both the expression plane – the Firthian work on phonetics and phonology
(prosodic phonology) – and the content plane, the repertoire of research methods had to span
both, so it was varied from the start; this coverage has been extended in SFL because it has
emphasized holistic theory and comprehensive descriptions.

184
 Application guides

Naturally, new research methods have been added as new technologies enabled new
approaches – as in the case of corpus studies, including NLP techniques (e.g. Halliday and James,
1993; Teich et al., 2016; O’Halloran et al., 2016), of translation studies involving eye-tracking
(e.g. Alves et al., 2010; Serbina, Niemietz and Neumann, 2015), of neurolinguistic studies involv-
ing brain imaging techniques (e.g. García and Ibáñez, 2017), but also as new approaches were
developed to carry out research in institutional settings, perhaps in particular action research in
educational settings mentioned earlier (cf. Burns, 2015). As the need arises, established research
methods have been added to the systemic functional research repertoire, as in the case of Hal-
liday’s (1975a) use of the diary method in his longitudinal case study of one child learning how
to mean. More recently, systemic functional linguists have adopted and adapted research synthe-
sis, systematic reviews and meta-analysis from applied linguistics in general (e.g. Mwinlaaru and
Xuan, 2016; Xuan and Chen, 2019; Chen, Xuan and Yu, 2022; de Silva Joyce and Feez, 2016).
Since the 1950s, naturally occurring text in context has been a key source of data in SFL,
and the analysis of text has been a key research method (cf. Halliday, 1964a), both to shed light
on particular texts as artefacts and to develop systemic descriptions based on texts as specimens
(cf. Halliday and Matthiessen, 2014), as in Halliday’s work on Chinese and later development
of the description of the system of intonation in English based on a corpus of spoken text (e.g.
Halliday, 1967); from the start, the analysis was both qualitative and quantitative. The earliest
example is Halliday’s (1959) study of the Language of the Chinese Secret History of the Mongols (the
published version of his PhD thesis). This is a corpus-based study consisting of one text. Halliday
analysed it manually to arrive at qualitative descriptive generalizations about the grammar, but he
also counted items so that he could produce quantitative profiles of terms in a number of systems.
Before he undertook this study, he had learned the general methodology of field work in
survey of dialects of Cantonese in the Pearl River Delta when he worked as a research assistant
for Professor Wang Li in 1949/1950; but while he was able to generate an important data set
using elicitation techniques, and planned to use this as his source for a PhD, he was told that
nobody could supervise this kind of research, so he switched to the “Chinese Secret History of
the Mongols”, and he was not able to write up the earlier study (cf. Halliday, 2006).
While research in SFL based on authentic (naturally occurring) texts has been dominant in
SFL – treated both individually and as samples in corpora, researchers have also used the meth-
odology of field work in the description of a growing range of languages. However, let us take a
step back, and survey research methods more systematically. Very helpful discussions of how to
conduct research in a number of areas in SFL can also be found in Unsworth (2000) and, with
a focus on literacy research, de Silva Joyce and Feez (2016).
Since research design involves informed choice among different research methods, we will rep-
resent research methods as a resource available to researchers – a research potential that can be
modelled as a system network: see Figure 4.4. When researchers plan or propose or embark on
a new research project, they need to select options from the systems in this system network. In
any given project, they may, of course, make more than one pass through the system network, for
example selecting both ‘qualitative’ and ‘quantitative’, but possibly for different sample sizes of the
texts (or other data sets) they are examining. (Thus it often makes sense to analyse a complete corpus
using computational tools to arrive at a ‘quantitative’ analysis, and to complement this with a manual
analysis of a small sample of the corpus to develop a ‘qualitative’ interpretation of the patterns.)
According to our description of research methods, there are a number of simultaneous sys-
tems of choice open to researchers (each of which could, of course, be elaborated in delicacy).

QUAL/QUANT – qualitative vs. quantitative: we either describe our data qualitatively

in discursive terms (as in manual text analysis, e.g. Halliday, 2002e; Birch and O’Toole,

185
 Application guides

Figure 4.4 Linguistic research methods represented as a system network to show choices available to
researchers (note that all the systems can serve as the basis of “mixed methods”)

1988) or we measure our data quantitatively in numerical terms (as in corpus analy-
sis, e.g. Halliday, 2005c; Teich, 2003; Thompson and Hunston, 2006; Hansen-Schirra,
Neumann and Steiner, 2012; Teich et al., 2016; Kunz et al., 2021). SFL studies have
used both methods, though qualitative methods have dominated, in large part because
of the difficulty of automating higher-level analysis of large samples of text. One of
the large-scale quantitative projects is Hasan’s (1989, 2009b) investigation of codal, or

186
 Application guides

semantic, variation in mother-child dialogues using principal components analysis.

Another study that can serve as a helpful model is Teich et al. (2016), which uses a
data-mining approach to register studies (outside SFL, cf. the rediscovery of discovery
procedures implemented by means of unsupervised machine learning, e.g. Lee, 2018).

Halliday (1994d/2007b: 241–242) offers an accessible characterization of principal com-

ponents analysis, used by Ruqaiya Hasan in her research on codal variation:

Hasan analysed the results in terms of “principal components”: each principal com-
ponent is a cluster of features which is identified by the program as accounting for
a high percentage of variation in the data. The essential property of this analysis is
that neither of the variables, social or semantic, is taken as given: neither the sex/class
matrix of the original sample, nor the systemic relation among the semantic features
in the network, is incorporated into the input. Thus the program could show up quite
different patterns of socio-semantic variation, having nothing to do with sex or class
and revealing no obvious underlying semantic motifs.

As noted, SFL studies have usually been ‘qualitative’ rather than quantitative, but thanks to techno-
logical advances in the compilation and analysis of data, it has become possible for researchers to
use quantitative methods to a larger extent (as in linguistics in general). Importantly, the quantita-
tive considerations had of course been included in the theory from the start by Halliday (1959),
and quantitative studies have certainly increased significantly in the last couple of decades.
The combination of quantitative and qualitative studies has often been referred to as “mixed
methods” (e.g. Denzin and Lincoln, 2018), although this term has also been used for other
combinations (as in Slade et al., 2015). From a systemic functional point of view, qualitative and
quantitative are complementary rather than “mixed” since they can be derived from the overall
architecture of language in context. Complementary methods can involve different selections
of options in research methods in the system in Figure 4.1, not only ‘quantitative’ and ‘qualita-
tive’ but also e.g. ‘natural’ and ‘experimental’, ‘authentic’ and ‘elicited’.
(2) NATURALNESS – natural (non-experimental) vs. experimental: we either use natu-
rally occurring phenomena as data or we generate phenomena as data through experimental
set ups. SFL studies have been largely natural or non-experimental since language and other
semiotic systems tend to behave very differently under experimental conditions, the work on
early language development being an important case in point (e.g. Halliday, 2003a; Painter,
1984, 1999; Torr, 1997). Thus Halliday (2003a: 183) observes, “These are concepts deriving
from the stage of concrete operations which Piaget associates typically with the age range 7 to
11, although his age assessments tend to be a bit late because they are based on experimental
rather than natural behaviour.” For related discussion of the goals of “ecological validity” vs.
“experimental control”, see Trevisan and García (2019).
As noted earlier, SFL studies are almost always ‘non-experimental’ – i.e. naturalistic – in
nature. This has a theoretical basis in the insight that in the study of semiotic systems, experi-
mental methods are likely to perturb the phenomena under investigation and thus to skew the
findings, as has been noted by Halliday (e.g. 2004) in relation to studies of ontogenesis (which
contrast with experimental methods used by psychologists like Piaget and many linguists investi-
gating how children learn how to mean but conceived of as “first language acquisition”). Painter
(2000) discusses the nature of data in studies of language development. However, experimental
or experimentally created conditions may have a role to play, as in the study of linguistic (and,
more generally, semiotic) processes as in TPR (translation process research), where Alves et al.

187
 Application guides

(2010) provide an interesting model combining more psycholinguistic methods with SFL. And
outside SFL, certain research based on experimental conditions has turned out to resonate with
longstanding insights in SFL derived from natural situation – importantly the work by Chris-
tiane von Stutterheim and her team investigating how speakers of different languages perceive
and construe video clips showing motion through space. (Another consideration in relation to
the possibility of using experimental methods in SFL is this: the very richness and complexity
of the SFL model is invaluable in the analysis and interpretation of authentic data but it can make
the task of setting up an experiment quite challenging.)
(3) AUTHENTICITY – authentic vs. elicited: we can use ‘authentic’ phenomena that occur
without our intervention as researchers or we can intervene and ‘elicit’ data using some form of
elicitation strategy – fieldwork elicitation based on translation or some other stimulus or con-
struction of examples when linguists speak the language they are investigating vs. questioning,
by means of informal interviews or surveys, either structured interviews or questionnaires, as in
the early work in Bernstein’s research group.2 SFL researchers have tended to draw on authentic
data, but in certain areas researchers have also used elicitation techniques.3
‘Authentic’ (naturally occurring) data may involve taking field notes in the course of observ-
ing texts unfolding in real time (as in Slade et al.’s, 2015: 13–24, tracking of medical consulta-
tions in emergency departments in large hospitals) or recording texts unfolding in real time or
using an existing record (as in most corpus studies). Recording texts in real time has become
increasingly easy thanks to advances in technology but also more constrained in terms of ethi-
cal and legal considerations. (Trained field workers can certainly transcribe spoken texts in real
time, as Halliday, 1975a, did in his case study of a young child learning how to mean, but most
researchers would now rely on video/audio recording.)
Even though the use of ‘authentic’ data has been favoured in SFL research, elicitation can
have an important role to play, as in Plum’s (1988) elicitation of stories – stories that would
probably not otherwise have occurred but which provided important data, and in studies using
interviews or semi-structured interviews as data, as in Slade et al. (2015: 14), where interviews
with patients and healthcare practitioners complemented the recordings of their medical con-
sultations. It is very important to note that what people think and say that they do and say may
actually be very different from what they actually do and say, so while a great deal of research in
sociology and also in certain areas of applied linguistics is based primarily on questionnaires and
other forms of surveys, from the point of view of SFL, this should really be complemented by
extensive analysis of ‘authentic data’ in the form of texts in context.
Another research context where elicitation may be needed is the collection of traditional
texts like folk tales in communities where only a few speakers are left and the texts no longer
occur spontaneously, as illustrated by Abbi’s (2021) unique anthology of texts from the Anda-
man islands.
At the same time, fieldwork elicitation or construction of examples can be essential in lan-
guage description, enabling linguists to “probe paradigms”. Even if descriptivists have access to
a large corpus of text when they develop their descriptions of linguistic systems, they will still
need to probe systemic combinations that do not occur in the corpus. For instance, if we are
describing the intersection of the systems of mood type and polarity in a language, it may turn
out that we find many examples in the corpus we are working with of both ‘positive’ and ‘nega-
tive’ in combination with ‘declarative’, and also enough examples of ‘elemental interrogatives’
(“wh-interrogatives”) but none for ‘polar interrogatives’. So then we need to elicit examples
to check if the combination is possible, and, importantly, what its interpersonal meaning is in
dialogic exchanges. The systemic functional literature of language description contains many
examples of such elicitation in the service of paradigm probing.

188
 Application guides

(4) ORDER OF DATA – primary data vs. secondary data: we can engage with semiotic
phenomena themselves, as primary data (however we arrive at them), or we can investigate stud-
ies of semiotic phenomena, as secondary data.
As has usually been the case in linguistics and more generally in the humanities, SFL has
typically drawn on ‘primary’ data – usually text in context (including “multimodal discourse”),
but also to some extent on data from surveys and questionnaires (which can of course also be
regarded as “texts” albeit ones instantiating very specialized registers). There are areas where
linguists tend to use ‘secondary’ data – accounts of ‘primary’ data rather than the original source;
for example, this may happen in language typology (although typologists consulting reference
grammars of different languages and other comparable accounts may of course analyse examples
given in those sources as ‘primary’ data).
However, in the last few years, there has been a more general move towards the use of ‘sec-
ondary’ data in studies involving research synthesis, typically coupled with systematic litera-
ture reviews (see Section 2.7 earlier, with references to SFL examples) instead of the traditional
type of literature review and with meta-analysis. In particular, Winfred Xuan has drawn on
the work by Lourdes Ortega and John Norris in applied linguistics. They introduced research
synthesis in applied linguistics (e.g. Norris and Ortega, 2006), drawing on work in other dis-
ciplines; research synthesis is a very common strategy in medicine. Together with colleagues,
Xuan has examined work in SFL on language typology (Mwinlaaru and Xuan, 2016), gram-
matical metaphor (Xuan and Chen, 2019) and translation studies (Chen, Xuan and Yu, 2022).
De Silva Joyce and Feez (2016: 94–95) discuss the use of systematic reviews in literacy research.
To support meta-analysis, systemic functional linguists can move towards agreed standards for
representing the results of analysis. This is not an easy task, but the pay-off in terms of collective
achievements could be immense.
(5) AUTOMATION – manual analysis vs. automated analysis: we can analyse our data
(whether natural or experimental, authentic or elicited) manually or, using computational tools,
we can analyse these data automatically. There is a clear trade-off here (see also Matthiessen,
2014a): manual analysis tends to operate with small samples of texts, but it can be quite high-
level in terms of rank and stratification (high-ranking grammatical analysis, and high-stratal
semantic and contextual analysis), whereas automated analysis tends to operate with large sam-
ples of texts (corpora – if these samples are principled and systematic) but, due to current
constraints on automation, it must be fairly low-level (based on graphological patterns and
low-ranking grammar).
Since SFL studies often involve high-level analysis, they have often had to rest on ‘manual’
analysis, and books published to support the study of language in practical terms have tended to be
geared towards text analysis at one stratum or another (e.g. Martin and White, 2005; Martin and
Rose, 2007; Martin, Matthiessen and Painter, 1997, 2010). But corpus linguistic techniques have
also been included since the 1960s and as computational technology has improved, offering more
storage capacity and more sophisticated tools for analysis (tools for analysing text, but also statistical
tools), systemic functional linguists have, naturally, been able to use ‘automated’ analysis.
One relatively early systemic study where the method used is set out explicitly is Halliday and
James (1993); they investigate the probability of the terms of a small number of grammatical sys-
tems, and discuss how they dealt with sampling, ambiguity and other issues that arise in a “raw”
corpus (i.e. one that has not been annotated), in their case a part of the Birmingham COBUILD
Corpus. More recent systemic functional studies using automated analysis where the approach is
set out explicitly include Teich (2003), Hansen-Schirra, Neumann and Steiner (2012), Teich et al.
(2016), Kunz et al. (2021). Methodological issues are also discussed and illustrated by contributors
to Hunston and Thompson (2006) and by Wu (2009). See further Section 4.4 later in this chapter.

189
 Application guides

(Since corpora have been used in SFL since the beginning, and they have been analysed using cor-
pus tools, the notion that SFL and corpus linguistics are distinct and bridges must be constructed
between them is fundamentally misguided. This would only make sense if the focus is on “corpus
linguistics” as an institutionalized species of linguistics – not if it is used as a methodological toolkit.)
(6) DIRECTION – analysis vs. synthesis: we can study language and other semiotic systems
by analysing instances of these systems, i.e. texts, but alternatively we can also study them by
synthesizing instances – typically by means of computer models, as when speech synthesis is
undertaken as a way of studying the properties of speech.
Like other linguistic traditions, SFL has tended to work with ‘analysis’ as a method of inves-
tigating phenomena rather than ‘synthesis’. Thus virtually all the extensive contributions to the
study of texts (discourses) have been based on text analysis (discourse analysis), whether manual
or automated. But, in some projects, researchers have used the other direction of investigation –
viz. synthesis, an early exploratory study being Fawcett (1973a): this is the body of work on
text generation by computer (Matthiessen and Bateman, 1991; O’Donnell and Bateman, 2005;
Teich, 2009; Bateman and O’Donnel, 2015; Bateman et al., 2019).
We have summarized key contributions to the options in research methods shown in Figure
4.4 in Table 4.3. These key contributions discuss the methods and/or provide illustrations of
their use. For example, Hasan (1989) illustrates the use of the statistical method of principal
components analysis in determining systemic values that tend to go together in differentiating
dyads between mothers and young children in terms of social class and the sex of the child, while
Wu (2009) and Sharoff (2017) provide general overviews of research methods involving corpora.
In principle, all combinations “predicted” by the system network of research methods in
Figure 4.4 are possible although they have not been equally common in SFL research. Each
time we traverse the network of options in research methods, we can make different sets of
selections of options. By traversing the network more than once, we adopt “mixed” or com-
plementary methods. Let us just exemplify combinations of options in the two top systems
in the system network: see Figure 4.5.

Figure 4.5 The intersection of two research method systems – qualitative vs. quantitative, and naturalistic
(non-experimental) vs. experimental
190
 Application guides

Table 4.3 Examples of systemic functional studies using the different research methods set out in Figure 4.4

System Term 1 Term 2

qual/quant qualitative quantitative

[Most SFL studies] Hasan (1989, 2009b): principal
components analysis based on
semantic system network
Halliday (1959), Nesbitt and Plum (1988),
Halliday and James (1993), Matthiessen
(1999, 2002a, 2006, 2015a): relative
text frequencies as a base for
probabilistic interpretation of systems
naturalness naturalistic (non-experimental) experimental
(experimentation) [Most SFL studies] Examples include [Few experimental SFL studies] Alves
the naturalistic study of child et al. (2010), Serbina, Niemietz and
language development using the Neumann (2015): the use of keystroke
“diary” method. logging and eye-tracking in the
investigation of translation as process
authenticity authentic (naturally occurring) elicited
observing Plum (1988): elicitation of stories
Slade et al. (2015: 13–14): field notes Slade et al. (2015: 14): semi-structured
recording; record interviews
Halliday (1975a): real-time transcription
Huddleston et al. (1968): corpus-based
register description
Halliday (1967), El-Menoufy (1969):
corpus-based description of intonation
order of data primary secondary
[Most SFL studies] Research synthesis: Mwinlaaru and
Xuan (2016): language typology;
Xuan and Chen (2019): grammatical
metaphor; Xuan and Chen (2020):
projection; Chen, Xuan and Yu
(2022): translation
automation manual analysis automated analysis
Halliday (1985b), Martin, Matthiessen Halliday and James (1993): automated
and Painter (1997, 2010): corpus analysis
lexicogrammatical text analysis; Wu (2009), Sharoff (2017): corpus-
Martin and Rose (2007): semantic text based research
analysis (written); Teich et al. (2016): data mining with
Eggins and Slade (2005): register features in study of discourses
lexicogrammatical and semantic text from different disciplines
analysis O’Halloran et al. (2016): data mining
Halliday and Greaves (2008): prosodic combined with multimodal discourse
text analysis analysis in study of terrorist discourse
direction analysis synthesis
[Most SFL studies] Matthiessen and Bateman (1991):
generation of text by computer;
O’Donnell and Bateman (2005)

191
 Application guides

As noted earlier, in complex research projects, we are likely to need to choose more than one
of the research methods set out in Figure 4.4. Combinations of research methods have been
discussed in the general literature on research under the headings of multimethod research
(including “mixed methods”) and of triangulation (cf. Fine, 1985).4

• In multimethod research, researchers will choose more than one option from the same
system, for example, ‘qualitative’ and ‘quantitative’, or ‘authentic’ and ‘elicited’.
• In triangulation, researchers will undertake multimethod research, but the emphasis is on
supporting the same analysis using different methods (as the navigational and survey meta-
phor suggests).

From the point of view of SFL, the most important point is that research methods may
complement one another, bringing out diferent aspects of the phenomena we are studying;
so “mixed methods” might be better conceived of as complementary methods (on the
methodological importance of complementarity, cf. Halliday, 2008b). This is illustrated by
O’Halloran et al. (2016); they used a combination of multimodal discourse analysis (MDA),
data mining and information visualisation to investigate text-image relationships in “the Eng-
lish language version of Dabiq, ISIS’s ofcial online internet magazine”. They describe their
methodology as follows (pp. 6–7):

The mixed methods approach involves qualitative, social semiotic analyses of the text
and image relations in Dabiq which are transformed into quantitative data using mul-
timodal annotation software. These quantitative results inform machine learning tech-
niques that are applied for data mining of text and image relations in large multimodal
data sets which contain the same or edited versions of the images from Dabiq. The
outcomes are a multilevel contextual model with empirical techniques for mapping
discourse patterns in large databases of multimodal texts. The approach complements
and extends existing research in terrorism studies which apply discourse analytical
perspectives, mixed methods methodologies, quantitative computational approaches,
and sociological and cultural studies approaches to the study of violent extremist pro-
paganda materials.

Their study thus shows very clearly the value of combining complementary research methods.
Complex research projects may, of course, be needed under various conditions, but they
are very likely to be needed in the study of institutions of various kinds such as the highly
elaborated institutions of modern nation states – e.g. educational institutions, legal institu-
tions, healthcare institutions and financial institutions, but also institutions that have most
likely always been part of human history in some version or another such as the institu-
tions of family and friendship. Thus in the investigation of communication in emergency
departments in hospitals, the researchers recorded large samples of interactions between
patients and healthcare practitioners and analysed them, primarily manually, but we also
conducted interviews and focus groups and we used questionnaires (see Slade et al., 2015;
Pun et al., 2017).
Let us outline a very schematic design involving complementary research methods that might
be used in researching an institution such as a school or a hospital: see Figure 4.6. It is helpful
to differentiate a number of orders of abstraction in the research, starting with the phenomena
under investigation and moving to the study of these phenomena:

192
 Application guides

• The lowest order of abstraction is what we can observe that people do and say – their
actions and activities, social (interactive) behaviour and exchanges of meaning. We can
observe these phenomena by recording them or by reporting them in field notes.
• Next comes people’s projection of this lowest order – what they believe and what they say
that they do and say. If they talk about this in naturally occurring context such as tea breaks
at a workplace, we can of course record their talk, but we can also elicit their accounts of
what they believe and say about what they do and say by means of elicitation techniques
such as (ethnographic) interviews, questionnaires and focus groups.
• The observations we make through recordings, field notes or elicitation techniques can
be interpreted as being of the third order in our proposed hierarchy: they are grounded
in the phenomena we study, but they are part of the study of these phenomena – thus
at one remove from the phenomena themselves (and consequently subject to errors of
observation).
• The output of the observations – however they are represented – can then be the basis of
different modes of analysis, including text (discourse) analysis, whether manual or based on
the techniques of corpus approaches, and “content analysis” in the sense of the term as it
has tended to be used in social sciences (and humanities) outside linguistics.
• While the last stage is, in principle, the last in a given research study, we need to include
the possibility of more abstract studies designed to analyse the results of individual studies.
At this 5th order of abstraction, we undertake analysis of lower-order analyses, typically
produced by a number of separate studies; in other words, we undertake meta-analysis
(cf. Section 2.7).

As we move up the hierarchy of abstraction, we obviously move further away from the first-
order phenomenal realm – what people say and do – but the study of such phenomena is still
empirically grounded in observable data. Obviously, if we rely on what people believe and say
that they say and do, we are at one remove from what they say and do precisely because we rely
on their projections; it is important to be aware of the fact that what people believe and say that
they say and do may be very diferent from what they actually say and do – particularly, with
respect to their beliefs about what they say, as Halliday has noted in various places, e.g. in Hal-
liday (2002d: 325–326).
As we move into the study of phenomena, issues naturally arise concerning the reliability
and precision of our observations, and then of our analysis of the products of our observations.
Therefore, researchers have developed various notions designed to help us assess the validity
of research findings, including inter-rater reliability, inter-coder reliability, consistency and
conformity, and statisticians have designed techniques for assessing findings. (There are many
books on the use of statistics in linguistics, including in corpus studies, applied linguistics and
language assessment, a number of them using the R programming language; here we will men-
tion just Brezina, 2018.)
In SFL, a few studies have investigated the consistency of analysis across analysts, in particular in
doing process type analysis, an area of description known for indeterminacy and alternative accounts
(cf. Table 5.1): O’Donnell, Zappavigna and Whitelaw (2008); Gwilliams and Fontaine (2015).
Both studies found a fairly high degree of inconsistency across analysts and offer insight-
ful explanations that shed light both on the area of the grammar itself and on criteria used
by analysts. But they both relied on volunteers invited through SFL mailing lists and on their
self-assessment of expertise. Gwilliams and Fontaine (2015) included only those participants
who characterized themselves as “advanced users of SFL”, and O’Donnell, Zappavigna and

193
 Application guides

Figure 4.6 Example of combination of methods in relation to different orders of phenomena and study of
phenomena

Whitelaw (2008) excluded “coders” who “had an individualistic approach to process types”.
In other words, neither study used an assessment of the degree of expertise of the participants
in the studies independent of the participants’ own assessment. In our own experience, even
“advanced users of SFL” may not have been trained to the point where they have mastered all
the considerations involved in the description of the system of transitivity in general and of
process type in particular – especially, the more cryptogrammatical aspects, including reac-
tances; and they may not have a very clear assessment of their own level of expertise. However,
this is of course precisely one key reason why assessing the quality of the different steps in
Figure 4.6 is an important research methodological concern and why studies of the kind just
discussed are needed.
The SFL community will continue to make advances in research methodology. Meanwhile,
for the individual analysts or teams of analysts, it is highly advisable:

• to ensure consistency in the analysis of comparable cases;

• to keep track of alternative analyses;
• to document considerations (criteria) involved in the choice among alternative analyses.

In Sections 4.3 through 4.7, we will refer to guides, handbooks and other contributions to
research methodology in particular areas of research. However, even if a given publication is
focussed on a particular area, it will contain guidelines that are of more general value. For
example, Ferguson and Armstrong’s (2009) book is worth consulting in fields of research other
than its target, viz. communication disorders. It is also very helpful to identify studies that can
serve as models, and then use their research designs and guides. The ability to replicate a study
based on the documentation provided of the research design and methodology is a good measure
of its usefulness as a model.

194
 Application guides

4.2.4 Authentic: data collection – points of view: sites

Before we turn to particular areas of research, let us elaborate further on how to collect authentic
data. One way of ensuring authenticity is to pay close attention to context, and to select a con-
sistent point of view that can be specified in terms of some aspect of context, siting the point of
collection – e.g. an institutional site (e.g. at the post office, in the classroom, at the dinner table –
conceived of in contextual terms in the first instance rather than in material terms) or a personal
site (tracking a person within an institution or across institutions, e.g. a patient journey through a
hospital, a student’s learner path, a person’s interaction in different roles in successive institutions
during a day or other period of time).
The sampling of text for the purpose compiling corpora that can serve as a general resource
for a variety of research projects has undergone a significant change from the early corpora of the
1960s (like Brown and LOB and London-Lund corpora of English) to more contemporary cor-
pora of texts sampled in reference to the different parameters of context (the early corpora were
oriented mainly towards mode). This is what Gu (e.g. 2002) calls “situated discourse”; in systemic
functional terms, such corpora are contextualized and have been designed in reference to some
combination of field, tenor and mode values so that their contextual properties are “controlled”,
and texts can be selected accordingly (see also Section 4.3.5). One helpful example outside SFL
but quite compatible with SFL is the CANCODE corpus (www.nottingham.ac.uk/research/
groups/cral/projects/cancode.aspx) and another is the BAWE corpus of student writing in
higher education, which is informed by SFL (www.coventry.ac.uk/research/research-directories/
current-projects/2015/british-academic-written-english-corpus-bawe/).
There is clearly a tension between different goals for a given corpus; in particular, the
goal of compiling a corpus of “situated discourse” may point in one direction, but the goal
of creating a “big data” corpus may point in another direction (contrast the orientation in
Conversation Analysis, CA, originating in sociology with that of the corpus development
within the Birmingham School). O’Halloran, Gautam and Jin (2021) note that in multi-
modal analysis, it is now possible to compile “big data” multimodal corpora, and thus also to
deploy “big data analytics”. The same applies to language corpora, of course. So it is impor-
tant to be aware of the trade-offs.
In SFL, the concern with corpora of “situated discourse” has been on the agenda for a long
time, as in the early work on scientific English (Huddleston et al., 1968; Halliday, 1988a) and later
contributions such as Banks (2008). One important version of this approach is the discourse or
language diary. Here the site of collection is personal, tracking a person (or group of persons)
as they “language” their way through a certain period of time, e.g. day, taking on different roles in
different groups operating within different institutions. As part of an account of register, Ure and
Ellis (1977: 228) provide an example of such a language diary, tracking Kofi M. during a day, taking
note of time, language events and language used as follows.

Language diary

Name: Kofi M. First Language: Fanti

Date: July 1969
Place(s): Near
Winneba
Office reference no. Working day/Free day
___________ (Cross out which does not apply)

195
 Application guides

Time Language events Language used

6:15 a.m. Asked personal boy to go for my newspapers. Akan

6:30 a.m. Read newspapers, discussed important headlines with wife. Akan (mixed)
7:00 a.m. Listened to the news. English
7:15 a.m. Had breakfast with wife and children; discussed teachers, teachers New Akan (mixed)
Salary Scales.
7:25 a.m. Left home for school with children. Advised them to avoid soiling their Akan
clothes.
7:35 a.m. Discussed with headteacher expansion to the school block to stop the shift Akan (mixed)
system.
8:15 a.m. Led pupils at prayers. English
8:30 a.m. Started the day’s work with Religious Instruction. English
10:00 a.m. Trained the school football team. English
10:30 a.m. Explained the bad behaviour of a pupil to his father who called at the school. Akan (mixed)
11:00 a.m. Taught school some hymns. English
12:00 p.m. Marked class exercise books. English
2:30 p.m. Explained to a co-worker how to go about the class magazine. Akan (mixed)
3:50 p.m. Left school for home with headteacher. Discussed the difficulty of getting Akan (mixed)
school grants for next year.
4:30 p.m. Sowed some seeds in my garden with personal boy. Akan
5:30 p.m. Asked personal boy to fetch water for my bath. Akan
6:00 p.m. Left for choir practise. English
7:40 p.m. Started preparing teaching notes for next week. English
8:00 p.m. Listened to bulletin of home news. English
8:30 p.m. Continued with notes preparation. English
10:00 p.m. Wished wife goodnight and went to bed. English

It would of course be possible, even desirable, to add other contextual features, such as insti-
tutional site and field, tenor and mode ranges of values, but they can often be inferred from
the descriptors of the language events. (Compare also the use of frames, scenarios, script and
other contextual conceptions from AI in Asp and de Villiers’s, 2010, examples of analyses of
disordered discourse.) Discourse diaries have been central in longitudinal case-based studies of
language development in SFL, beginning with Halliday (1975a), whose diary is presented in
Halliday (1984c). But there are many other possible uses (cf. Halliday, 1985f); they can be used
in longitudinal studies in various institutional settings, as in the study of learner paths in educa-
tion or patient journeys in healthcare. Guerra Lyons (2021) presents a pioneering study of the
development of the academic writing by a selection of linguists over decades, addressing the
methodological issues of sampling and modes of analysis.

4.3 System description

4.3.1 The task of describing linguistic (sub-) systems

Describing the system of a language is a challenging but obviously fascinating task. It is chal-
lenging because it means that the focus is on the potential pole of the cline of instantiation, so
196
 Application guides

the person undertaking the task is, in principle, responsible for covering the whole system of the
language under description – its entire meaning potential (including its wording potential and
its sounding potential). Of course, this task can be divided into sub-tasks based on the semiotic
dimensions of the systemic functional architecture of language – as long as the sub-tasks are
undertaken ecologically:

stratification: we can zoom in on one of the strata of language instead of taking on all of
them. This is a reasonable way of managing the complexity of the descriptive task, provided
that the stratal neighbourhood is taken into account, i.e. that the stratum is approached eco-
logically, not in isolation. Within a given stratum, intra-stratally, we can zoom in further to
focus on one rank, again adopting an ecological approach. Here it is important to choose the
rank in such a way that it is both inter-stratally and intra-stratally informative. For semantics,
this points to the text, for lexicogrammar to the clause and for phonology to the highest
prosodic unit, like the tone group. Thus the clause is both the gateway to the semantics of
the text and the domain within which lower-ranking units of lexicogrammar operate.
metafunction: alternatively, we can select a metafunctional slice through the system, focus-
sing on both semantics and lexicogrammar, or concentrating on one or the other while
always referring to the higher- or lower-stratal correlates. For example, we can explore
speech function (semantics) and mood (lexicogrammar) together, shunting between
them; or we can explore one, say the mood system of the clause while constantly relating
it to the semantics of speech function. Working with Mongolian, Zhang (2020) provides
a helpful model of how to develop a description of interpersonal systems, taking not only
semantics and lexicogrammar but also context (tenor) into consideration.

Intersecting stratification and metafunction, we can locate the patterns we are concerned with
within a stratification-function matrix, or, by another step internal to a given stratum,
within a function-rank matrix. By locating the patterns within such a matrix, we ensure that
the contours of the overall system have been specified, helping us identify interactions to take
into account, e.g. between theme and transitivity or between theme and mood.
However we select our focus of descriptive attention, the cline of delicacy will also be meth-
odologically very helpful: we start with low delicacy while trying to maintain a reasonably broad
coverage, and then against the picture that emerges, we can increase the delicacy. (Somewhere
along the way, we will meet the region of intermediate delicacy characterized by the kinds of
pattern investigated by Construction Grammarians.) This will ensure that the overall systemic
map has already been sketched. In other words, we develop the systemic description by gradual
approximation.
There is one more semiotic dimension that is methodologically very helpful, the cline of
instantiation:

instantiation: we can shunt downwards along the cline of instantiation, focussing on part
of the overall system – a register (in the sense of functional variety of language operating
in a context characterized by certain field, tenor and mode values) or set of registers.
Thus in her description of the lexicogrammar of Thai, Patpong (2005) first compiled
a corpus of narratives (told by traditional story-tellers), used it as the empirical base of
her lexicogrammatical description, and then tested the description against samples from
other registers. This is a natural, well-motivated way of managing the complexity of the
descriptive task. (Compare explorations of the system of semantics focussed on particu-
lar registers, and the semantic strategies used in those registers to address the goals of the
context in which they operate: Halliday, 1972; Turner, 1987; Patten, 1988.)
197
 Application guides

As we make decisions about where to locate our descriptive eforts, we can thus be guided by
the semiotic dimensions of systemic functional theory, and we can reason about trade-ofs as we
try to ensure that our descriptive eforts are manageable. For example, if we want to produce a
description that is extended in delicacy, it may make sense to move down the cline of instantia-
tion to a particular register.
There is considerable cumulative experience in SFL with language description. From the start,
SFL researchers have been working on the description of the systems of different languages. Hal-
liday’s pioneering work started with Chinese in the 1950s (collected in Halliday, 2006), and then
extended to English in the 1960s. During the 1960s, researchers turned to other languages – French,
Yoruba, Nzema and Mbembe; and the work on a wider range of languages took off in the 1990s
(see Table 3.11 earlier in this chapter). This was also the period when Longacre (1964) published his
Grammar Discovery Procedures: A Field Manual, a book informed by Tagmemic Linguistics that is still
worth consulting, especially in view of the fact that there is a need in linguistics in general, not only
in SFL, for materials guiding linguists in the development of descriptions. (US American structur-
alists had long been concerned with procedures for developing descriptions of languages ensuring
the validity of the descriptions proposed. Chomsky, 1957, called these “discovery procedures” and
argued for “evaluation procedures” instead. This was surely one reason not enough attention was
paid in linguistics in general to procedures or methods for developing descriptions, but “discovery
procedures” can now be revisited in the context of advances of machine learning techniques and
neural network representation. Given an extensive corpus, such techniques can “discover” aspects
of the system of the language under investigation – pace Chomsky, 1957.)
There are many reasons for producing descriptions of languages, and systemic functional
descriptions are designed to be appliable – so maximally relevant to the needs of communities
of speakers in different institutional settings (cf. Halliday, 1964a, on “syntax and the consumer”,
and Martin, 1998, on “linguistics and the consumer”, with an orientation towards education).
In the case of “major” languages, they tend to be descriptions of a range of descriptions avail-
able, even fairly comprehensive “reference” descriptions. But, unfortunately, most languages still
spoken around the world have not been given rich comprehensive descriptions. In the case of
endangered languages, the task of producing such descriptions is particularly urgent, and there
is considerable awareness in linguistics of the need for language maintenance, revival or even
documentation of languages bound to disappear (e.g. Nettle and Romaine, 2000; Harrison,
2007; Evans, 2010). Linguists emphasize the sustainability of communities of speakers and what
they lose if they lose their language – and what is lost to humanity in general as part of our
collective heritage. What is lost tends to be framed in ideational terms – the loss of knowledge,
including the knowledge of the environment – but we should add that it is equally important to
consider the loss in interpersonal terms – the loss of modes of interaction, of ways of enacting
tenor relationships (cf. Diamond, 2013).
Developing a description of the linguistic system of a particular language is a daunting but
exciting task. As noted earlier, the system of a language is located at the potential pole of the
cline of instantiation, as shown in Figure 1.1, so when we set about describing the system of
a language, we have to take into account a reasonable registerial range of instances. Naturally,
this is most important when we describe the content plane systems of a language, the semantic
system and the lexicogrammatical system, since registerial variation is primarily variation in
content – it is semantic variation in the first instance (“meanings at risk”), but it is also lexico-
grammatical variation (thus, wordings at risk), since the stratal relation between them is natural
rather than conventional (arbitrary).
Since the 1960s, descriptions of linguistic systems in SFL have tended to focus on the lexi-
cogrammatical subsystem, at least as a way into an account of a given language. As exemplified

198
 Application guides

in Table 4.4 (to be discussed later), researchers have typically drawn on sets of texts, compiled
their own corpora of texts or used existing corpora (the method of using ‘authentic’ texts in
Figure 4.4); but they have also sometimes used elicitation, although elicitation has not been used
very often since researchers have tended to describe languages they themselves speak fluently.
In such cases, linguists have instead deduced examples from the language they speak as a way
of illustrating or probing a category. This is the method of constructing examples – of eliciting
examples from oneself, and often needs to be used in probing paradigms, for example to deter-
mine whether two systems are simultaneous or ordered in delicacy.

4.3.2 Description as upward move along the cline of instantiation

Describing the system of a language means in the first instance moving along the cline of instan-
tiation from the instance pole to or at least towards the potential pole: we meet instantial patterns
in texts in their contexts of situation, and we look for regularity of occurrence, taking regulari-
ties as the basis for positing instance types, and then we gradually generalize the instance type,
positing systems as we begin to approximate the description of general patterns at the potential
pole of the cline of instantiation. The descriptive move is visualized in Figure 4.7. (As Figure
4.10b later shows, in the development of a description of the lexicogrammar of a language, one
is likely to move through a number of cycles of generalization and interpretation as systems.)
The descriptive move thus involves generalization or induction step by step. There is a rela-
tionship between the size of the sample of texts at the instance pole of the cline of instantiation
and the distance it is empirically justified to move up along the cline of instantiation. The closer
our target is to the potential pole, the larger the sample that we need to compile if we want to
produce a reasonably reliable description, as shown in Figure 4.8.
Figure 4.8 includes two limiting cases.
(1) We may focus on a single text as an artefact of interest in its own right – probably
because it is given high value in the culture or a sub-culture, like a sacred text, a literary text or

Figure 4.7 Description of linguistic systems as an upward move along the cline of instantiation from
observation of instantial patterns in texts in their contexts of situation via regularities in
patterning interpreted as instance types to instance types interpreted as systems

199
 Application guides

Figure 4.8 The relationship between the descriptive target along the cline of instantiation and the size of
the sample of texts (corpus) needed for a reliable description

a legal text such as The Universal Declaration of Human Rights. This is most likely to involve an
act of analysis rather than one of description (thus presupposing an existing description), and the
sample is simply the artefact under investigation – unless other similar texts are brought in to
enable us to investigate the text in a comparative fashion.
(2) We may focus on the overall system of a language located at the potential pole of the cline
of instantiation, in which case we will need a very large corpus, one that will merit the label
of reference corpus. How large is difficult to say, but if we look at the trend since the 1960s,
when a large corpus was around 1 million words, to the present, when reference corpora are on
the order of hundreds of millions of words, we can safely say that such a corpus would need to
be on the order of hundreds of millions of words. However, the sample size is obviously not the
only consideration; importantly, the registerial composition of the corpus is quite crucial: see
Section 4.3.4. Naturally, additional considerations come into play if we aim to include systemic
probabilities in the description. The general principle is that we need to estimate what volume
of text we need to get a sufficient number of instances of a given feature. Thus if the feature
is very delicate, we need more text; similarly, if it is highly marked, we also need more text,
the general principle being that we need sufficient instances of marked features. Clearly, more
research needs to be done to provide researchers with further guidance, including meta-analytic
studies surveying the size and nature of corpora used in language description.
As an example of using a reference corpus, taken from outside SFL, we can note the first
“corpus-based” reference grammar of English – the Longman grammar by Biber et al. (1999),

200
 Application guides

which was designed to complement Quirk et al. (1985), which had of course already drawn on
SEU (the Survey of English Usage). The corpus they used “contains over 40 million words of
text representing six major register categories” (p. 4), which they suggest provides “a sound basis
for reliable analyses of grammatical patterns” (p. 24). In comparison, when Halliday and Mat-
thiessen revised the second edition of Halliday’s IFG (Halliday, 1994c), they subscribed to a 60
million word sub-corpus of the COBUILD Corpus, consisting of spoken and written texts from
various registers. While reasonably large, this corpus turned out to be too small to find enough
examples of certain “constructions” that they investigated. (We will return to this example later,
discussing the choice of “register categories”.)
(3) Intermediate between the two poles, there are descriptions that target registers, ranging
from “macro-registers” such as scientific English (cf. Huddleston et al., 1968; Halliday, 1988a) to
“micro-registers” or fairly restricted registers such as weather forecasts and stock market reports.
Such restricted registers came into focus in work in machine translation in the late 1970s and
early 1980s when researchers realized that such systems would be more successful in generating
automatic translations if they tailored their systems to such registers (or “sub-languages” as they
called them; e.g. Kittredge, 1987). Here it seemed that a sample of text of around 15,000 words
was sufficient in order to do the descriptive tuning for such a register. As a variant of a purely
register-focussed description, the descriptivist can use a particular register as a way into the
language under description. Thus Patpong (2005) began her systemic functional description of
Thai by compiling a corpus of traditional stories, developed her description based on patterns in
this corpus and then checked it against a corpus of texts from a wider range of registers. This is
a contextually motivated way of making the process of description more manageable. (Compare
the use of narratives by non-SFL descriptivists, e.g. focussing on tense-aspect systems in different
languages, as in Hopper, 1982.)
A key challenge in language description is, as noted, to decide on the size and composition of
the corpus upon which the description is based. Caruso et al. (2014) review approaches to the
determination of corpus size, identifying lemma-based ones and text-based ones. Approaches
based on lemmas, or lexical items, are of course oriented towards lexical studies (cf. Halliday,
2002b, on the bias in the representation of corpora towards lexical studies), and they need to
be supplemented by considerations of grammatical and also semantic features. Here text-based
approaches could be promising, but Caruso et al. (2014: 136) conclude their review saying, “no
practical and rigorous method is yet available as de facto standard in corpus linguistics to define
the size of a corpus in terms of number of texts”. They then go on to propose “a statistical
method estimate minimum corpus size” and test it on a particular register (“domain”), “part of
a corpus currently being compiled for the construction of a tourism thesaurus”.
To recapitulate: systemic functional description is, in the first instance, a move along the cline
of instantiation from the instance pole towards or up to and including the potential pole, and the
“material” for the development of systemic functional description of systems is thus located at
the instance pole – texts in their contexts of situation. The move involves gradual generalization
or induction from instantial patterns via regularities that can be identified as texts are analysed
and classified as instance types to systemic interpretations of instance types as they accumulate.
In principle, we can imagine that in the future the process of description can be fully auto-
mated through machine learning techniques (which we could think of as revisiting the dis-
covery procedures of US American structuralist linguistics; cf. Longacre’s, 1964, field guide).
Although “low-level” patterns can be identified (even with fairly simple concordance tools –
e.g. n-grams), the task of developing systemic descriptions remains a largely manual undertak-
ing, and it probably will continue to remain one for the foreseeable future as long as we aim
high, as systemic functional linguists do. Aiming high means targeting systemic and functional

201
 Application guides

descriptions, throughout the lexicogrammatical rank scale, and also throughout the semantic
compositional hierarchy – and covering all metafunctions. This is a tall order for automated
analysis (as is illustrated by the challenge of developing systemic functional parsers).
Rounding off the discussion of the use of corpora as the material for inducing the system in
the process of description, let us note that there is also, of course, a place for a butterfly collec-
tion approach to the development of systemic descriptions. It is always helpful to keep looking
for examples of interest to one area of description or another, monitoring texts one meets as
part of daily life, e.g. in consuming news or attending to casual conversations. Since large corpus
samples of casual conversation are still more restricted than samples of written text or even of
other varieties of spoken texts, examples occurring in spontaneous spoken discourse can be of
special significance – and at the frontier of the system evolving, as Halliday (2002c) notes. For
example, relative clauses with resumptive personal pronouns that are co-referential with the rela-
tive item occur in casual speech, but not in formal writing, as in Well, something is going to happen
to me that I don’t know what it’s going to be (from a talk by Alan Watts). Both the butterfly collection
approach and the investigation of large corpora are quite likely to lead to examples that even
trained linguists would not deem grammatical when they (try to) construct them.
We have suggested that the development of description involves induction, i.e. generaliza-
tion of instantial patterns found in a corpus to distil more general patterns in the description
targeting the potential pole of the cline of instantiation. But this one-way move along the cline
of instantiation towards the potential will at some point need to be complemented by system-
based text analysis and paradigm probing:

• System-based text analysis: When we start developing a new description by distilling

generalizations about instantial patterns identified in texts in context, there is as yet no
description. But as the new description begins to emerge, it can be tested against additional
data; i.e. we can use it in systemic text analysis, checking whether the description holds up
in the analysis of additional texts, or needs to be extended or revised. This may be done
by starting the description with texts from one register and then expanding the registerial
sample to undertake systemic text analysis. In principle, this is just like working with a
fairly established description such as Halliday’s (1985a) original “IFG” description of Eng-
lish, applying it in the analysis of large volumes of text, and then extending or revising the
analysis. In fact, this has been going on for decades with the IFG description. This process
can be interpreted as one of abduction (a form of reasoning due to C.S. Peirce, but viewed
here as it has been applied in computational linguistics and AI, e.g. Hobbs et al., 1993).
• Paradigm-probing: When we develop descriptions based on generalizations from a cor-
pus of texts, we can represent these generalizations as paradigms, starting with “working
paradigms” (see Section 4.3.4) and moving towards systems. One of the recurrent questions
will be whether two or more systems are simultaneous or ordered in delicacy. For example,
if we establish one system of polarity, positive/negative, and one system of mood type,
say an early version such as imperative/declarative/polar interrogative/elemental interroga-
tive, we may find that while ‘positive’ versions of the different mood types are common,
‘negative’ ones are much less common – and may even be absent in the corpus for one or
more combinations, e.g. ‘negative’ and ‘polar interrogative’. If we hypothesize the polarity
and mood type are simultaneous systems and define a two-dimensional paradigm, how do
we deal with the gap in this paradigm, the absence of ‘negative’ and ‘polar interrogative’?
One obvious approach is of course to expand the corpus, but if we still cannot find any
examples, what is the next step? This is where probing the paradigm is a useful descriptive
strategy: we try to construct examples that are negative polar interrogatives, and then we

202
 Application guides

check them with native users of the language. If they accept them, then it is likely that they
are possible, only quite marked, perhaps with a special meaning. This would be the case
for English: the clause Wasn’t I real before? enacts a biased expectation as to the value of the
polarity (‘I thought I was real before.’). (The construction of such examples may of course
involve elicitation, which may be based on translation of an example from a language shared
by the linguist and the language consultant.) However, if we probe the same paradigm for
‘negative’ and exclamative declaratives, we would in fact find that they do not occur, so we
would have to describe this gap in our provisional, working paradigm systemically.

In establishing and probing systemic paradigms, we need to take account not only of overt
patterns but also covert ones. This means probing for reactances to reveal cryptotypes, as dis-
cussed for research methods in work on transitivity by Davidse (2017: 90–93) for transitivity and
Quiroz (2020) for process types within transitivity.

4.3.3 Descriptive guidance: theoretical and typological

The descriptive move along the cline of instantiation from instance towards potential is induc-
tive in nature. However, this move can be guided in manual analysis by sources of information
other than the sample of texts within the corpus used for the description, and the same applies
to paradigm probing. The central sources of information are:

• The general systemic functional theory of language (in context) provides theoretical guid-
ance. The theory provides descriptivists with a general template – the “architecture” of
language, prompting them to look for systemic patterns as well as structural ones; semantic
agnations as well as grammatical ones; function structures as well as syntagms; textual, inter-
personal and logical modes of expression and meaning as well as experiential ones; prosodic
patterns as well as articulatory ones. In other words, the general theory invites descriptivists
to look far beyond what would be included in mainstream descriptions – reference grammars
and the like (i.e. the general theory makes more of the language “visible” to them than what
is usually included in reference grammars); and it guides them by giving them a sense of what
a multidimensional comprehensive description of a language needs to include.
• At the same time, the general theory does not impose particular descriptive categories such
as Subject, Agent, Theme, (present) tense, (nominative) case and (plural) number on the
description of a “new” language. Language-specific categories are not built into the general
theory – precisely to avoid the mistake of foisting Latin on all other languages as traditional
(often missionary) linguists tended to do or English as generative linguists have tended to
do (cf. English-centric notions like “non-configurational”, “pro-drop” and “free (vs. fixed)
word order”). (This insight goes back to Firthian linguistics; cf. the points made outside
SFL by Evans and Levinson, 2009, about “the myth of language universals”.) Instead, more
particular guidance is derived from generalizations across descriptions of particular lan-
guages – descriptions that are empirically grounded in texts in context. Such generaliza-
tions constitute typological guidance. For example, if in the course of the description
of a language, it turns out that the verbal group tends to come at the end of the clause, in
its experiential function as Process and its interpersonal function as Predicator, it is highly
likely that the language ends the clause with an “interpersonal finale”: as speakers are about
to hand over to their listeners, they indicate to them ‘this is what I’m doing in my interac-
tion with you (the speech role I adopt), and what I expect you to do by way of response
(the speech role I assign to you)’.

203
 Application guides

So descriptivists should be guided by the general theory of language and by typological general-
izations as they compile their samples of texts (corpora) and analyse them in an attempt to arrive
at systemic generalizations. Aren’t they allowed to “cheat” by looking at model descriptions of
English – e.g. Halliday’s description of the lexicogrammar of English (Halliday and Matthiessen,
2014; Matthiessen, 1995a, with typological pointers) and Martin’s description of the semantics
of English (Martin, 1992a; Martin and Rose, 2007)? No – well . . . good model descriptions
can actually be very helpful, but the crucial point is that they should not be restricted to English.
To stimulate their imagination of what is linguistically possible and to expand their linguistic
horizons, researchers setting out on the description of a “new” language can and should draw
inspiration from the description of a variety of other languages. In Cafarel, Martin and Mat-
thiessen (2004), they can consult the description of eight languages, and quite a few languages
have been added to the descriptive treasure trove since then: see Teruya and Matthiessen (2015).
One special case of using the description of a given language as a “model” as one develops
a description of a “new” language is transfer comparison. This method was originally char-
acterized by Halliday (1960/1966: 39) as follows, and has also been discussed by Teich (1999a):

In the comparison of languages we may take advantage of the fact that . . . there are always
several different ways of describing the same linguistic phenomenon; it is thus possible to
adopt the description of one language to that of another. The aim of this ‘transfer com-
parison’ is to draw attention to the resemblances between the two languages.

(In a way there is an analogue to transfer comparison in translation; in work largely compatible
with SFL but not informed by SFL, Mann and Weber, 1979, refer to this strategy as dialect
adaptation.) The diferent sources of information that can guide the development of a

Figure 4.9 Sources of information informing development of the description of a “new” language

204
 Application guides

description of a “new” based on a registerially informed sample of texts in contexts are sum-
marized in Figure 4.9.
The general goal systemic functional linguists set for themselves as they embark on descrip-
tions of particular languages is to produce comprehensive descriptions of each language they
engage with, doing justice to each language in its own terms. There are fundamental theoreti-
cal reasons for this goal: given the nature of language as a complex adaptive system, the only
way in which it is possible to reveal its fundamental nature and to reason about it in terms of
systems thinking is to work with comprehensive descriptions, e.g. by identifying fractal patterns
manifested throughout the lexicogrammar of a given language. At the same time, since systemic
functional accounts are designed to be appliable – to have the power to be applied to a wide
range of community needs – they must be resources, and only comprehensive descriptions can
serve as appliable resources.
Consequently, systemic functional descriptions tend to be developed by means of gradual
approximation. First the delicacy of focus is fairly low so that it will be possible to achieve com-
prehensive coverage at a fairly low degree of delicacy, as reflected in a function-rank matrix of a lan-
guage where all the systems of primary delicacy are identified and labelled. Once this overall sketch
of the linguistic resources has been produced (like a map of some territory of the scale of one to a
million), it becomes possible to increase the delicacy of the description gradually. As the description
of a particular area is increased (e.g. the system of transitivity, or the system of logico-semantic type),
its systemic environment has already been mapped out; it is not being described in isolation. This
increase in delicacy is repeated, as shown diagrammatically in Figure 4.10a.
We can call this approach to description the holistic approach of gradual approxima-
tion. It is the kind of approach that cartographers would adopt. It contrasts fundamentally with
the approach of Cartesian Analysis (cf. Capra, 1996, who argues in favour of the holistic
approach over Cartesian Analysis). This approach has dominated Chomskyan generative linguis-
tics. Its starting point is a set of theory-internal questions about language derived from Western
Philosophy, more specifically from the debate in epistemology between rationalism and empiri-
cism. Adopting the position of rationalism, generative linguists within this Chomskyan strand
have drilled down to areas likely to yield results of theoretical interest given their stance and
of manageable extent. This approach is shown in the lower half of Figure 4.10a. (For an early
characterization of the “failure of generative grammar” because of its lack of comprehensive
descriptions, see Gross, 1979 – drawing insights from Zellig Harris, Maurice Gross had expertise
in both mathematical linguistics and corpus-based research. It is, however, important to note
that there are linguists emerging out of the generative tradition who can be said to have departed
from Chomsky’s Cartesian Analysis, importantly including McCawley, 1998, and those genera-
tive linguists committed to the development of parsing grammars, where patterns encountered
in texts rather than only theoretically elevated phenomena must be handled.)
In the holistic approach, the process of gradual approximation proceeds in phases. Here the
expansion of systemic coverage of a language increases, phase by phase, and being grounded in
empirical evidence from text, this will correlate with the gradual accumulation of texts (sampled
from the instance pole of the cline of instantiation, within different registers: see Section 4.3.4).
The correlation between the expansion of systemic descriptive coverage and the expansion of texts
within different registers that have been analyses is represented diagrammatically in Figure 4.10b.
When systemic functional linguists work with the approach of gradual approximation
(Figure 4.10a, b), the first descriptive approximation is typically a sketch of the major gram-
matical systems of the lexicogrammar at clause rank, as shown in Figure 4.11. This makes
excellent descriptive sense since the grammar of the clause in a language is the gateway of
lexicogrammar to the semantics of the text. This first descriptive approximation will serve

205
 Application guides

(a)

(b)

Figure 4.10a, b Development of systemic description as a step-wise probe of expansion of systemic

coverage and accumulation of texts analysed in terms of the emergent description

206
 Application guides

as the foundation for further description that on the one hand extends the description in
delicacy towards lexis and the other hand “drills down” into the grammar of lower-ranking
units – for which the clause serves as the environment. This is thus an ecological approach
to the description of the lexicogrammatical system of a language. Naturally, the character-
istics of the lexicogrammar of a particular language may make it necessary to pay attention
to some domain or domains other than that of the clause. For example, in developing a
description of Kalam, one would need to consider the grammar of clause complexing in
relation to the grammar of the clause from the start, and in developing a description of
Mapuche, one would need to consider the grammar of the verb at word rank in relation to
the grammar of the clause from the start.
We have mentioned theoretical guidance and typological guidance in the development of
new systemic descriptions. Since we have focussed on the development of the lexicogrammati-
cal stratal subsystem of languages, we should emphasize that the guidance is very likely to come
“from above”, i.e. in terms of the hierarchy stratification from the vantage point of semantics
and, by another stratal step, the vantage point of context. The importance of approach in the
description and comparison of the lexicogrammars of different languages “from above”, i.e.
from the point of view of semantics, was first highlighted and articulated by Martin (1983) in
his article on the identification and tracking of references in English, Tagalog and Kate. The
value of moving in “from above” has been demonstrated recently by Zhang (2020a, 2020b) in
his description of the interpersonal resources of Khorchin Mongolian.

4.3.4 From sets of instances to paradigms to systems

The move from recurrent instantial patterns observed and identified in texts to systemic descrip-
tions involves intermediate stems. An important one is the development of working para-
digms that can form the foundation for the further development of systems in system networks.
As he explores the grammar of pain, Halliday (1998a/2005a: 320–321) discusses organizing

Figure 4.11 First descriptive approximation in terms of coverage (rank and delicacy)

207
 Application guides

examples of expressions of pain into a working paradigm as a step towards the development of
a systemic description:

In section 7, I will organize these into a working paradigm – that is, into a pre-systemic
form which makes it possible to compare and contrast them systematically in terms of
their grammatical features. People do not speak in paradigms (an obvious point, but
one that is easily forgotten when you are investigating language!); but they are a neces-
sary working resource for grammarians, whose task is to bring out the potential that
lies behind what people say. The purpose of this particular paradigmatic display is to
examine expressions of pain from the standpoint of transitivity. What type of process is
‘pain’? What part does the ‘pain’ element play in the total configuration by which the
process is construed?

The working paradigm that Halliday presents is reproduced here as Table 4.4. Halliday (1998a/
2005a: 321) describes it as follows:

I shall first set out the paradigm in summary form (cf. Table 12.1), including (a) the
‘pain’ expression; (b) a brief analysis of the expression, showing (i) the type of process,
(ii) the relevant structural functions; and (c) an agnate expression from some other
semantic domain. In those cases where the pattern of intonation is not the one that
is typical for English (that is, where the tonic prominence is not in its “unmarked”
place, on the final lexicalized element), the element that carries the prominence is
shown in bold. The ‘pain’ expression, in each case, is a typical form of wording,
representative of one grammatical construction whereby ‘pain’ is commonly talked
about in ordinary dialogue. The analysis suggests the transitivity category to which
it approximates most closely – its primary address in this region of the grammar; in
some instances a secondary interpretation has been offered, enclosed between square
brackets. The agnate expression is one that shares the same primary grammatical
features.

4.3.5 The use of texts from different registers

The selection of texts serving as the data in the development of systemic descriptions is stra-
tegically very important; it can make a significant difference in terms of the investigation of
different areas within the overall lexicogrammatical system of a language. Examples of texts
that have been used in the development of descriptions of the lexicogrammars of different
languages are given in Table 4.5. All the descriptions listed in the table are based on reasonably
extensive evidence from text, and a number of researchers have sampled naturally occurring
texts in order to compile their own corpora – in particular, when no corpora had already been
compiled previously.
What are the characteristics needed for a corpus chosen or compiled in order to support the
development of lexicogrammatical descriptions? Such corpora differ from the first generation
of corpora compiled for English (e.g. Brown, LOB, ACE and the ICE corpora) and a number
of other languages in that they should ideally be composed of whole situated texts (in Gu’s, e.g.
2002, sense of “situated discourse”), as noted in Section 4.2.4 earlier in this chapter:

• The corpus should be composed of whole texts operating in known contexts so that they
have integrity as semantic units and so that it is possible to draw on contextual and semantic

208
 Table 4.4 Example of a working paradigm: Halliday’s (1998a) paradigm of pain expressions

(a) ‘Pain’ expression (b.i) Type of process (b.ii) Structural functions (c) Agnate expression

A Pain as process 1 my knee hurts/aches relational: attributive body part = Carrier the ground slopes
Attribute/Process the paint sticks (‘is sticky’)
2 my knee’s hurting/ material: middle body part = Medium my nose is bleeding
aching [existential: occurring] [body part = Setting] [the roof ’s leaking]
3 I hurt/ache (here) relational: attributive person = Carrier I sympathize
[behavioural] Attribute/Process [I grieve, I worry]
[person = Behaver]
4 I’m hurting/aching material: middle [existential: person = Medium I’m falling
(here) occurring] [behavioural] [person = Setting] [I’m peeling]
[person = Behaver] [I’m trembling]

Application guides
5 it hurts/aches (here) existential: existing impersonal Setting it echoes
6 it’s hurting/aching existential: occurring impersonal Setting it’s raining
209

(here)
7 my knee’s hurting me mental: impacting body part = Phenomenon/Agent the heat’s bothering me
person = Senser
8 you’re hurting me material: effective person = Goal; you’re pushing me
other (person/object) = Actor
9 I’ve hurt my knee material: effective person = Actor; I’ve broken my glasses
body part = Goal
10 I’ve hurt myself material: effective/reflexive person = Actor; I’ve ruined myself
(on the knee) body part = Location
11 that hurts relational: attributive/ Attribute/Process; that dirties (‘causes things to be dirty’)
agentive other (object/process) = Agent

B Pain as quality 12 my throat feels sore relational: attributive ‘pain’ = Attribute; the meat smells bad
body part = Carrier
13 my throat’s feeling sore (as no. 12) (as no. 12) your face is looking thin

(Continued)
 Table 4.4 (Continued)

(a) ‘Pain’ expression (b.i) Type of process (b.ii) Structural functions (c) Agnate expression

14 I feel sore (here) (as no. 12) ‘pain’ = Attribute; I feel sad
person = Carrier
15 I’m feeling sore (here) (as no. 12) (as no. 14) I’m feeling sad
16 it’s sore (here) existential: existing impersonal Setting it’s cold outside
17 the wound is painful relational: attributive ‘pain’ = Attribute; other the climate’s healthy
(object/process) = Carrier
C Pain as thing 18 I’ve got a headache/a pain relational: attributive/possessive ‘pain’ = Attribute (possessed); I’ve got a chest cold/a hole in my
in my neck person = Carrier pocket
19 that gives me a headache relational: attribute/possessive/ (as no. 18) that gives me a thought

Application guides
agentive other (object/process) = Agent
20 that’s giving me a headache material: effective: benefactive ‘pain’ = Goal; that’s giving me a lot of help
210

person = Beneficiary;
other = Actor
21 do you feel any pain? mental: scope-defining ‘pain’ = Phenomenon; do you see any smoke?
person = Senser;
22 my pain is bad (today) relational: attributive ‘pain’ = Carrier; my hopes are high
quality = Attribute
23 are you in (great) pain? relational: attributive/ ‘pain’ = (circumstantial) Attribute; are you in great suspense?
circumstantial person = Carrier
24 the pain suggests (that) . . . verbal ‘pain’ = Sayer the damp suggests that . . .
 Table 4.5 Examples of data sets of texts used in different systemic functional descriptions of lexicogrammars

Language Reference Descriptive focus Authentic text Elicited text

Japanese Teruya (1998, 2007) grammar: clause, group, text sampled from different registers; examples constructed by linguist as
word existing large corpus of Japanese native speaker
Chinese (Mandarin) Li (2003) grammar: clause translation of an English novel into Chinese
Thai Patpong (2005) grammar: clause texts of the register of traditional folk tales,

Application guides
supplemented by smaller samples from
other registers
211

Akan Matthiessen (1987a) grammar: clause, group, texts of the register of traditional folk tales examples elicited through translation
word to probe systemic possibilities
(paradigm probing)
Dagaare Mwinlaaru (2017) grammar: clause See Table 4.8 examples constructed by linguist as
native speaker
Bajjika Kumar (2009), grammar: clause text sampled from different registers
Kashyap (2012)
Khorchin Mongolian Zhang (2020a) interpersonal systems: institutions: (1) family: daily conversations;
semantics and grammar (2) school: conversations between
colleagues; (3) government agency:
government officials’ visit to local peasant
 Application guides

patterns in the development of the lexicogrammatical description – obviously in the devel-

opment of the description of thematic systems, but also more generally.
• The texts should be sampled from a reasonably wide range of registers representing dif-
ferent field, tenor and mode values – extending beyond stories and conversations to include
texts from other registers that figure prominently in the language under description, a cen-
tral consideration being what wordings are at risk in different registers.
• The texts should ideally be represented in their original form as video or audio recordings
or as print documents with contextual notes if possible in addition to being transcribed,
reformatted and chunked in preparation for different modes of analysis.

The size of the corpus or corpora supporting the description is naturally also an important
consideration. Nowadays standard corpora available for major languages with long descriptive
traditions (and good sources of funding) tend to be tens of millions of words, or even hundreds
of millions of words (in their corpus-based description of the grammar of English, Biber et
al., 1999, used a corpus of 40 million words), and even these mega-corpora may not be large
enough to find a reasonable number of instances of highly marked features (which may therefore
have to be studied by means of elicited or constructed examples). Of course, from a grammati-
cal point of view, it would be better to have measures in terms of clauses and clause complexes
rather than words, but computational tools for automatic chunking into grammatical units are
not likely to produce very reliable results (varying of course with the language under descrip-
tion). When no standard corpora are available, researchers need to sample texts themselves from
diferent registers and compile their own corpora, and they will necessarily have to start on a
much smaller scale.
As we noted earlier, the corpus should be composed of texts from a reasonably wide range
of registers, and here researchers can draw on the long tradition in SFL of characterizations of
the contextual parameters of field, tenor and mode, and register variation according to different
field, tenor and mode values (e.g. Halliday, 1978).
If we increase the delicacy in the contextual specification of the environment in which
the registers we sample enough, we arrive at a point where we can provide some guidance
as to what kinds of text are likely to be a good source for the initial or early investigation
of a lexicogrammatical or a semantic system. Such guidance has to do with what mean-
ings and wordings are likely to figure prominently in different registers – the meanings and
wordings “at risk” in a given register. This kind of guidance is in a sense inherent in the
text typology proposed by Longacre (1974) within the broadly Tagmemic tradition and
elaborated in later publications (Longacre, 1996; Longacre and Hwang, 2012): Longacre
specifies lexicogrammatical features characteristic of the different text types or genres in his
typology, as illustrated in Table 4.6. And similar insights led to the frequent use of narra-
tives in West-Coast functionalist discourse-based studies of tense-aspect systems in various
languages, beginning in the second half of the 1970s (e.g. Hopper, 1982), and to the later
discovery of conversation as a central source in the study of various grammatical features (cf.
Ochs, Schegloff and Thompson, 1996). In fact, conversation played an important role in
Hopper and Thompson’s (2001) review of their transitivity hypothesis (Hopper and Thomp-
son, 1980, 1982), which had originally related grammar and discourse in general terms, not
in a register-sensitive way.
Thanks to the work in SFL on register variation since the 1960s, we know that the “mean-
ings at risk” vary across registers with the values of the contextual parameters of field, tenor and
mode; and since the relationship between semantics (meaning) and lexicogrammar (wording) is
a natural one, we can – by another stratal step – also consider the “wordings at risk”. A number

212
 Application guides

Table 4.6 Longacre’s (1974: 358) examples of text types (genres) with lexicogrammatical characteristics

− Prescriptive + Prescriptive

+ succession narrative (many types) procedural (many types; includes

(chronological) instructional)
1. 1/3 persons 1. pseudo (non-specific person)
2. Actor oriented 2. Goal oriented
3. accomplished time encodes as past or 3. projected time encodes as past,
present present or future
4. chronological linkage 4. chronological linkage
drama
1. multiple ½ person
2. accomplished time as concurrent
3. dialogue paragraphs without quotation
formulas
− succession expository (several types, includes descriptive) hortatory (sermons, pep talks, etc.)
(chronological) 1. any person (usually 3rd) 1. 2nd person
2. subject matter oriented 2. addressee oriented
3. time not focal 3. commands, suggestions encode as
4. logical linkage imperatives of ‘softened’ commands
4. logical linkage

of the meanings and wordings at risk are reviewed systematically by Martin (1992a) for English,
and we can expect similar cross-stratal patterns for other languages.
So let us ask what registers will serve as good starting points for the investigation of dif-
ferent lexicogrammatical systems, using the contextual parameters of field, tenor and mode.
We will start with field of activity within field, and then add some comments on tenor and
mode. As we will see, it is possible to arrive at the same area of “wordings at risk” starting
from field or tenor or mode. This is because while these three contextual parameters are, in
principle, independently variable, there are favoured combinations and combinations that
do not seem to occur at a given period of time (gaps – cf. Lemke, 1995, on disjunctions).
For example, if we want to study the resources for enacting assessments (evaluations) in a
given language – e.g. the interpersonal resources of modality, evidentiality or appraisal (e.g.
Martin and White, 2005, on the English resources) – we might approach through field of
activity – sharing of (personal) opinions, exploring of (public) values; or through tenor –
sociometric roles (affect), valuation (of domains of experience); or through mode – orienta-
tion towards tenor. This is because systemic functional accounts tend to be (deliberately!)
over-determined.
(i) Field: field of activity. As a heuristic, we can use the description of field of activity
that we have developed over a number of years, and documented in e.g. Matthiessen, Teruya
and Lam (2010), Matthiessen and Teruya (2015a), Matthiessen (2015b). Drawing on Matthies-
sen (2015b), we have summarized characterizations of our classification of field of activity, taking
two steps in delicacy in Table 4.7, where we have also included references to relevant accounts
in Martin and Rose (2008) and in Eggins and Slade (2005). (Importantly, the descriptions devel-
oped by Martin and his colleagues within their “Genre Model” do of course extend well beyond
the largely educational registers covered in Martin and Rose, 2008, but their book is a good way
into the expanding body of work within that model.)

213
 Table 4.7 Field of activity – primary and secondary types, with references to key publications within the SFL “Genre Model”

Field of activity Key accounts in the sfl literature based on the “genre
model”

Primary type Characterization Secondary type Characterization Written: Spoken:

Martin and Rose (2008): Eggins and Slade (2005)

expounding expounding our experience explaining explaining phenomena by specifying (Chapter 4 Reports
of classes of phenomena how they happen (temporal and Explanations)
according to a general sequence of events) or how explanations
theory (ranging they happen (causal sequence of
from commonsense events, contributing factors)
folk theories to categorizing categorizing (documenting) (Chapter 4 Reports and
uncommonsense phenomena, by specifying Explanations) reports

Application guides
scientific theories) their properties, location in
taxonomies and parts
214

reporting reporting on our experience chronicling chronicling the flow of particular (Chapter 3 Histories)
of particular phenomena events, linearly (as in historical recounts, biographies
(as opposed to general recounts, logbooks) – verbal (Chapter 5 Procedures and
classes of phenomena) timelines, or cyclically (as in Procedural Recounts)
modern news reports) procedural recounts
surveying surveying particular places –
verbal maps (as in topographic
reports, e.g. in guidebooks)
inventorying inventorying particular entities (as
in catalogues)
recreating recreating our experience of [narrating, (Chapter 2 Stories) stories:
the world imaginatively, dramatizing] narratives
that is, creating imaginary
worlds having some direct
or tenuous relation to the
world of our daily lives –
through narration and/or
dramatization
 sharing sharing our personal lives, [experiences, values] (Chapter 2 Stories) stories: chat; opinion, teasing,
prototypically in private, anecdotes, exempla gossip
thereby establishing,
maintaining and
negotiation personal
relationships – sharing
our personal experiences
and/or sharing our
personal values
doing doing social activities, collaborating by members of one group
prototypically engaging collaborating with one another,
in interactive social using language to coordinate
behaviour, thereby their activities
collectively achieving directing by one person directing the other
some task members of a group
recommending recommending people promoting by inducing them (promotion:

Application guides
to undertake some recommendation for the benefit
activity, thus very likely of the speaker, as in consultations)
215

foreshadowing a ‘doing’ advising by advising them (recommendation

context for the benefit of the addressee, as
in advertisements)
enabling enabling people to instructing by instructing them in how to (Chapter 5 Procedures and
undertake some undertake the activity; or procedural recounts)
activity, thus very likely procedures
foreshadowing a ‘doing’ regulating by regulating (controlling) their (Chapter 5 Procedures and
context behaviour procedural recounts)
protocols <or: embedded
in procedures>
exploring exploring our communal arguing by arguing about positions and ideas (Chapter 3 Histories)
values and positions, expositions, discussions
prototypically in public reviewing by reviewing a commodity
(in contrast with (goods-&-services)
‘sharing’)
rallying by rallying people around an idea or
a cause (as in a political speech,
by members of the establishment
or by protesters)
 Application guides

Once we have elaborated the specification of field of activity to secondary delicacy, we can begin
to identify correlates in the grammar – the “wordings at risk”. For example, once we have difer-
entiated ‘reporting’ into diferent kinds according to the domain of phenomena being reported
on, we can begin to identify grammatical areas likely to be “at risk”: reporting on events – i.e.
chronicling events, will very likely involve the grammar of time, including the grammar of pro-
cess time (modelled as tense and/or aspect in languages) and the grammar of location and extent
in time, but reporting on places – i.e. surveying territories – will likely involve the grammar of
space, including static processes and participants, and circumstances.5
To give an indication of how different fields of activity are likely to put different areas
of the grammar of a language “at risk”, we have constructed the display in Figure 4.12. For
example, texts operating in contexts of expounding knowledge about phenomena in general
(rather than particular instances) – i.e. building field in Martin’s (e.g. 1993b) sense (see “field
of experience” below) – are good for investigating the grammar of generalizations about the

Figure 4.12 Fields of activity and areas of grammar likely to be “at risk”

216
 Application guides

world as a resource for construing scientific or folk theories; in contexts expounding knowl-
edge by categorizing phenomena in general, texts such as descriptive and taxonomic reports
are good for investigating the grammar of taxonomy – experientially, relational clauses and
nominal groups; interpersonally, gnomic statements; and textually, theme selection involving
lexical cohesive relations. (For example, in investigating theme in Japanese, we found that
in narrative texts operating in recreating contexts, the famous postposition wa only occurs
occasionally, but in taxonomic reports, it is very frequent because such texts involve constant
shift from one topical theme to another as the taxonomy is being construed clause by clause
in the text.)
Within the field parameter of context, there is one more sub-parameter in addition to field of
activity; this is field of experience – the experiential domain or “subject matter” construed by
language and other denotative semiotic systems in a context. It covers all the domains of experi-
ence that fall within the cosmology of a community and would be covered in an encyclopaedia
and reflected in the lexical organization described in a proper thesaurus, but it also covers the
different models that a community of speakers operate with – always a folk model, or common-
sense model, and possibly also scientific, or uncommonsense, models (or educational versions
of them. There are good examples in the systemic functional educational linguistics literature,
e.g. Wignell, Martin and Eggins, 1989; Painter, 1999; and cf. also Halliday and Matthiessen,
2006: Chapter 14). Thus both folk models and scientific models can be expounded, but folk
models may be construed through proverbs. While field of experience and field of activity are in
principle independently variable, certain combinations are very likely, e.g. taxonomic organiza-
tion of experience and expounding: categorizing and procedural sequential organization and
enabling: instructing.
To investigate the ideational resources of language in terms of texts operating in different
fields of experience, we can draw on the strand within anthropological linguistics (including
the ethnography of speaking) known as ethnosemantics (or ethnoscience; e.g. Kronenfeld,
1996: 47–53), which originated with studies in the 1950s and used “componential analysis”
taken from Roman Jakobson’s version of Prague School phonology (Goodenough, 1956). Eth-
nosemantics is characterized by Saville-Troike (2002: 103) as being “concerned primarily with
discovering how experience is categorized by eliciting terms in the informants’ language at
various levels of abstraction and analyzing their semantic organization, usually in the form of a
taxonomy or componential analysis”. This has included work on folk taxonomies of plants, of
diseases, of rhetorical modes (genres) – aspects of the lifeworlds of the speakers of a community.
In ethnosemantics, the method of elicitation is often used to obtain lexical items. However, in
a systemic functional descriptive project, it would make sense in the first place to sample texts
from different experiential domains, and to analyse them not only to identify lexical items but
more generally to take stock of lexicogrammatical patterns – patterns that realize the ideational
semantic construal of the domain.
(ii) Tenor. While we focussed on one of the two major parameters of field, viz. field of
activity, we need to consider all the basic parameters of tenor: institutional role, power (status)
role, familiarity (contact) and sociometric role (affect) are important in sampling texts from dif-
ferent registers as material for the description of interpersonal systems, and it is also helpful to
consider the valuation of the field of discourse (domain of experience).
Institutional role should be controlled for in the identification of contexts in which texts
are sampled, since the varieties of a given language spoken by people serving in different roles
may be important in a given community, reflecting the division of labour according to institu-
tional roles, including occupational roles (cf. notion like teacher talk, e.g. Thwaite and Rival-
land, 2009; as an important example, Kealley, 2007, presents a study of the linguistic resources

217
 Application guides

used by nurses, and there are of course highly specialized roles that may be of particular signifi-
cance to a community because of their association with the law or ritual).
Power (status) role and familiarity (contact) often combine to define the socio-semiotic
space vertically and horizontally within which interactants operate, enacting their locations in
this space – calibrating, confirming and potentially also contesting the space they have been allo-
cated. So they should be taken into consideration in the sampling of texts and the compilation
of the corpus to be used as the empirical based of the descriptive project. This is of course very
clear in those languages where power and familiarity combinations have been “grammatical-
ized”, as in the system of “speech level” in Korean, which is simultaneous in the grammar of the
Korean clause with the system of mood (e.g. S.E. Martin, 1992). Similarly, considerations apply
to Japanese, which has been described comprehensively by Teruya (2007), Javanese and various
other languages of East and South-East Asia. Of course, an early well-known study drawing
attention to the significance of the combination of power (vertical) and solidarity (horizontal)
is Brown and Gilman (1960), who described pronominal systems in reference to (in our terms)
these tenor parameters. And the ever-expanding literature on politeness (and now also impo-
liteness) provides many examples of how languages provide interpersonal resources enabling
interactants to negotiate the tenor of the relationship between them. In very traditional tribal
communities, this socio-semiotic space may not be very greatly expanded; they will be fairly
egalitarian communities where everybody essentially knows everybody else (e.g. Harman, 1999;
Johnson and Earle, 2000; Lucassen, 2021), strangerhood thus being rare. However, systems of
kindship may play quite a central role.
The tenor parameter of sociometric role is significant in the selection of texts needed to
explore the affective resources of language, the resources for enacting the “charge” between
speaker and addressee – neutral or charged, and if charged, positive or negative. So settings
of this parameter are relevant to the investigation of the linguistic resources of interpersonal
assessment – as in the description of the system of attitude in particular within the appraisal
resources of a language, along the lines set out by Martin and White (2005). These resources
cover interpersonal connotation within lexis, purely interpersonal items but also areas of gram-
mar. Such resources are also sensitive to another aspect of tenor, viz. valuation.
While sociometric role concerns the enactment of the charge among interactants, valuation is
focussed on the evaluation of areas within the field of discourse, again either neutral or loaded, pos-
itive or negative. This may relate to other tenor considerations – constraints on topic areas within
fields of experience in relation to the role relationships, taboo being an extreme example. For
example, very personal experiences may be associated with field of activity of sharing and tenor
relationships of intimacy (familiarity), equal status (power) and close friendship (institutional role).
Therefore, considerations of field and tenor combinations can be significant in the search for
contexts to sample texts in. There are many examples of what we can interpret as profiles of
fields of activity for different tenor roles in the literature. To cite just one social-psychological
study, based on questionnaires, Henderson and Argyle (1985) show that work colleagues in dif-
ferent tenor relationships (social friends, friends at work, workmates, disliked colleagues) have
different profiles of fields of activity they are likely to engage in (e.g. helping each other with
work, discussing work, chatting casually, discussing personal life). Awareness of such possible
variation is obviously important in the development of corpora of text in context, of “situated
discourse”, as a basis for language description. The mappings between tenor and field are also
central more generally in a community to the division of labour and the formation of socio-
semiotic hierarchies.
Varying the settings of tenor variables is, as suggested earlier, a central method to be used
in exploring interpersonal systems. This is illustrated systematically and powerfully by Zhang

218
 Application guides

(2020a, 2020b) in his context-calibrated study of interpersonal systems in Mongolian. There is

one further related reason for sampling texts operating in contexts with distinct settings of the
same parameters: the meanings enacted by one and the same interpersonal system may be mul-
tiplied or “flipped” in different tenor environments. The classic example of the tenor multiplica-
tion of interpersonal meanings is Brown and Gilman’s (1960) study of “pronouns of power and
solidarity”. The same lexicogrammatical pronominal system may vary in meaning depending
on the settings of the parameters of power (status) and familiarity (solidarity); for example, the
choice of an “informal” addressee pronoun over a “formal” one may enact solidarity if the inter-
actants are of equal status, but distance if they are of unequal status, either domination or respect.
Because of the need to negotiate tenor relations, such interpersonal systems can become quite
systemically elaborate (cf. the reference to speech level in Korean), a kind of elaboration that
may involve interpersonal lexicogrammatical metaphor. The interpersonal “flip” we referred
to earlier often occurs in systems of interpersonal amplitude, where for example negative items
may be co-opted as boosters. Another kind of flip is syntagmatic in nature: for English, Poynton
(1990, 1996) shows that familiarity and sociometric role may be realized between interactants
either by shortening their given names – the shortening as it were symbolizing the reduction in
distance, or by extending them by means of interpersonal affixes of affection – thus extending
the prosody of affection. Similarly, such flips occur with irony and sarcasm, and will be interest-
ing to document in the language under description.
(iii) Mode. Mode includes medium (spoken or written language), channel (oral or visual, or
complex combinations – many languages now operating in contexts enabled by new technolo-
gies of the channel), turn (dialogic or monologic), rhetorical mode (didactic, explanatory, narra-
tive, persuasive, exhortatory, polemic, diverting and so on); and also, in our account, orientation
towards field or towards tenor in terms of contextual goals, division of labour among semiotic
systems and between semiotic systems (processes of meaning) and social systems (processes of
behaving).
The full range of possible combinations is not taken up in every language during a particular
period of history, and this will obviously influence the identification of contexts in which to
sample texts belonging to different registers. In human history, the evolution of the written
mode is quite recent, going back around 5,000 years or so in certain places, where writing
evolved out of drawing initially essentially in inventorying contexts for book keeping – and the
technology of the channel needed for writing. Thus for most of our history, human cultures
have been oral, and many languages exist mainly or only in the spoken mode – especially those
with few speakers, including severely endangered ones. Even in languages with a written mode,
the distribution of registers may be such that there are gaps – gaps filled by other languages in
multilingual societies, e.g. in formal education, in written news reporting, in the legal system.
So this is also important to take into account in the compilation of a registerially rich and varied
corpus of the language under description.
In terms of division of labour among semiotic and social processes, there is also considerable
variation. In face-to-face interaction, language and gesture combinations may have evolved
to divide up the semiotic labour in different ways for different languages (e.g. Lantolf, 2010,
on English and Spanish in the construal of motion lexicogrammatically and gesturally). In the
didactic rhetorical mode, communities will vary with respect to the extent to which youngsters
are taught more or less silently by demonstration of social practices and also semiotically.
In compiling a corpus of “situated discourse” for a language under description, we thus have
many factors to take into consideration. The early corpora from the 1960s tended to be char-
acterized in terms of mostly mode categories, like written and then spoken when this became
technically possible, channel (e.g. print, face-to-face), turn (monologic, dialogic) – with the

219
 Application guides

addition of traditional “genre” categories like fiction and news. However, even a consideration
of all the different aspects of mode requires more dimensions of classification, and in order to
compile a corpus of truly “situated discourse”, we need to involve field and tenor as well.
The contextual parameters of field, tenor and mode can thus be used to guide the sampling
and selection of texts as we compile a corpus to support the description of (aspects of) the sys-
tem of a given language. The choice of contextual parameters and the ranges of settings within
each parameter will depend on the focus and scope of the description. For example, if we are
concerned with interpersonal systems, it makes sense to concentrate on settings of tenor values
in particular. As an illustration of how such contextual considerations can inform the sampling
and selection of texts from different registers, we have adapted the overview tables of the cor-
pora compiled by Mwinlaaru (2017) in his description of Dagaare and by Zhang (2020a) in his
description of Khorchin Mongolian as Tables 4.8 and 4.9.
In referring to the different wordings at risk, we have used “semanticky” terms in the first
instance – “the grammar of assessment”, “the grammar of opinion”, “the grammar of inter-
personal control” and so on – to foreground the point that semantics is the interface between
context (in this case, field of activity) and grammar, and also to ensure that we do not jump to
language-specific expectations too soon. The principle behind this strategy was articulated early
on in a crucial contribution on the textual semantic task of “participant identification” in text
by J.R. Martin that drew on both the insights from the Hartford Stratificationalists and the SFL
modelling of language in context: Martin (1983), and for a recent powerful example focussed on
tenor and interpersonal systems in semantics and lexicogrammar, see Zhang (2020a). His con-
textual characterization of the registerially varied texts of his corpus is adapted as Table 4.9 here.
Returning briefly to our mention of the reference corpus used by Biber et al. (1999) in
Section 4.3.2, we can note that while their corpus was compiled based on “six major register
categories”, the corpora compiled for systemic functional descriptions of languages should be

Table 4.8 Adaptation of Mwinlaaru’s (2017: Table 2.2) showing field (field of activity) and mode values of
the archive of text he compiled for his description of the lexicogrammar of Dagaare

Field of activity Mode

Written Spoken

Dialogue Monologue Dialogue Monologue

doing A tribute for a installation of queen

University VC mother
enabling school textbook
recommending concert advertisement
exploring arguments in conversations;
radio panel discussions;
expounding
reporting biblical narratives in conversations and farmers’ reports at
narratives meetings; radio interviews; agricultural workshop;
radio announcements
recreating short stories unscripted play; religious film folktales
sharing conversations meetings; Face- speeches
book posts and comments

220
 Application guides

Table 4.9 Adaptation of Zhang’s (2020a: Table 1.1.) “contextual characterization” in terms of field (field
of discourse), tenor and mode of the data used for his description of interpersonal systems in
Khorchin Mongolian

Institution Field: field of discourse Tenor Mode

family mostly domestic; some [institutional roles:] nuclear a mixture of constituting

specialized family and extended family; field and accompanying
[familiarity:] close contact field, but mostly
[status roles:] equal status constituting field; mostly
school mostly specialized and [institutional roles:] co-workers/ dialogue
technical; some close friends
domestic [familiarity:] close/distant contact
[status roles:] equal status
government mostly specialised; [institutional roles:] government
agency some technical and officials and peasants
domestic [familiarity:] distant contact
[status roles:] unequal

based on considerably more delicate differentiation of registers because it is only when we reach
higher degrees of delicacy that we begin to see the significant effect of “wordings at risk”. This
greater context-based differentiation is illustrated by the corpora characterized in Table 4.8 and
4.9. These examples taken from the research projects concerned with the development of sys-
temic functional descriptions of Dagaare and of Khorchin Mongolian are also closer to Gu’s (e.g.
2002) notion of a corpus of “situated discourse”.

4.3.6 A note on “corpus based” vs. “corpus driven”

Let us return to the relationship between samples of texts at the instance pole of the cline of
instantiation and the development of descriptions of the system of a language at the potential
pole of the cline. As part of the development of corpus linguistic methodology, Tognini-Bonelli
(2001) characterized the well-known and often-used contrast between corpus-based accounts
and corpus-driven ones as follows (pp. 10–11; cf. also Biber, 2010):

Chapter Four considers what can be referred to as the ‘corpus-based’ as a methodol-

ogy that uses corpus evidence mainly as a repository of examples to expound, test or
exemplify given theoretical statements. It is argued that, in this context, the potential
of corpus evidence is not exploited fully because, in order not to threaten dramatically
some existing theoretical positions, the richness of language usage is in many ways
sacrificed and is not allowed to shape the descriptive and theoretical statements that
should ideally account for it.
Chapter Five defines, in contrast with the ‘corpus-based’ approach, the ‘corpus-driven’
approach where the corpus is used beyond the selection of examples to support or quan-
tify a pre-existing theoretical category. Here the theoretical statement can only be for-
mulated in the presence of corpus evidence and is fully accountable to it. This approach,
it is argued, brings about a qualitative change in the description of language and shakes
some major assumptions underlying traditional linguistics. One of these, discussed in this
chapter, is the distinction between ‘lemma’ and ‘inflected forms’ which are usually taken
as equivalent in terms of meaning and function.
221
 Application guides

What Tognini-Bonelli calls “theoretical statements” and “theoretical category” would usually
correspond to descriptive statements and categories in SFL (although certain categories that
she suggests emerge from the corpus-driven approach are theoretical in nature, in fact echoing
theoretical categories proposed by Firth: collocation, colligation and prosody).
The corpus-based approach can be interpreted as a deductive move along the cline of
instantiation, from potential to instance, as represented diagrammatically in Figure 3.9, and the
corpus-driven approach can be interpreted as an inductive move along the cline of instantiation,
from instance to potential. The most far-reaching examples of the inductive move would argu-
ably be techniques involving some kind of machine learning, as in grammar induction, although
at present “corpus-driven” seems to be limited in corpus linguistics by the fairly low-level
computational tools currently being used. However, the distinction between the corpus-based
or deductive approach and the corpus-driven or inductive approach needs to be supplemented
by the abductive approach. This approach involves repeated moves between potential and
instance: descriptions of systems at the potential pole of the cline are used in the analysis of texts
and the instance pole, and the results of the analysis fed back into the description, expanding
it in some way so that it is being improved in the course of abduction. This characterizes text
analysis – one of the major activities in SFL.

4.4 Text analysis

While the description of the system of a language – or a stratal subsystem of a language – is
located at the potential pole of the cline of instantiation, text analysis is located at the other pole
of the cline, the instance pole (cf. Figure 1.1 earlier in this volume). Thus it is less daunting than
system description because even if the text being analysed is a very long one or a macro-text like
a semester-long course in a subject or a textbook it still instantiates only a fraction of the overall
system, and, crucially, it typically draws on an existing description of the system of the language
that the text or texts instantiate – although pre-systemic text analysis is an important strategy (i.e.
the analysis of text before a description of the system has been developed, typically in the service
of the development of such a description6).

4.4.1 Examples in the SFL literature that can serve as models

The SFL literature has a number of examples of texts analysed and then interpreted based on the
analysis, ranging from introductions to the task of the analysis of texts of different kinds (e.g. Halliday,
1985b; Hasan, 1985b) to studies that provide good models of how to go about text analysis (e.g. Hal-
liday, 1992a, 1994b). Some of these examples are listed in Table 4.10 with notes of features that are
methodologically helpful. (Compare also the examples of use of corpora in SFL listed in Table 2.4
earlier.) Martin and Rose (2007) devote a chapter to “tackling a text” (Chapter 8) semantically, pro-
viding illustrations of text analysis; and Martin, Matthiessen and Painter (2010) discuss and illustrate
text analysis in the final chapter of their functional grammar workbook (Chapter 7). Halliday and
Greaves (2008) demonstrate and illustrate prosodic analysis of text. Providing historical background,
Hodge (2017) considers systemic functional “discourse analysis” in relation to CDA concerns; and
He (2019) takes stock of systemic functional text analysis, also noting the potential for future work.

4.4.2 Text and options in types of text analysis

Text analysis has turned out to be one of the major activities in SFL – and also in other func-
tional linguistic frameworks (though not in all). It is alternatively labelled “discourse analysis”,

222
 Application guides

Table 4.10 Examples of studies involving text analysis with notes on exemplification of methodology

Study Kind of text Methodological model

Halliday and Hasan (1976) (various) cohesive analysis; examples of texts annotated for
various types of cohesion
Mishler (1984) medical interviews cohesive analysis
Halliday (1985b) dialogue between adult the different aspects of the analysis of (spoken) text:
and children transcription, prosodic annotation (pp. 31–32);
metafunctional clause analysis (pp. 32–46) and
group/phrase analysis (pp. 45–49);
cohesive analysis (pp. 49–51);
grammatical metaphor (pp. 51–52);
contextual “gloss” (pp. 52–53)
Halliday (1992a) fundraising letter lexicogrammatical analysis
Martin (1992b) the ZPG letter

Mann, Matthiessen and semantic analysis: Rhetorical Structure Theory

Thompson (1992) (RST) analysis
Halliday (1994b) thesis examination prosodic transcription (pp. 200–206)
deliberation lexicogrammatical analysis: tabulation of
metafunctional clause analysis (pp. 207–229):
taxis; theme, mood and modality, transitivity
Martin (1992a); Martin (various) semantic analysis: the systems described
and Rose (2007)
Cloran, Stuart-Smith and advertisement comparison of: message semantic analysis, RST
Young (2007) analysis and phasal analysis
Halliday and Greaves (2008) (various) prosodic analysis

and discourse analysis outside SFL is often treated as a field distinct from and parallel to linguis-
tics, with contributions from a range of other disciplines (see e.g. van Dijk, 1997a, 1997b; de
Beaugrande, 1997). However, from a systemic functional point of view, “text” and “discourse”
are not distinct phenomena, but rather one and the same phenomenon, perhaps seen in some-
what different perspectives (e.g. Halliday, 2011): for discussion, see also Section 6.2.2.1.
In some traditions outside SFL, “discourse” seems to be used to refer roughly to the combina-
tion of text (semantics) and situation, or context of situation, (context). Text and situation will, of
course, occur together as closely aligned processes or products. Thus Halliday (1978: 145) notes:

The patterns of determination that we find between the context of situation and the
text are a general characteristic of the whole complex that is formed by a text and its
environment.

If it seems useful, we can call this whole complex “discourse”, but it is important to note that
text and context of situation are stratally distinct and subject to separate lines of analysis, and,
importantly, “texts” engendered by semiotic systems other than language may operate within
the same context of situation alongside the (verbal) text. We can expect to find similar semiotic
complexes across any stratal border that is natural rather than conventional – not only between
context and semantics but also between semantics and lexicogrammar and between phonology
and phonetics (and graphology and graphetics).

223
 Application guides

As we have noted, text is located at the instance pole of the cline of instantiation, but it is
also extended stratally: see Figure 4.13. In the first instance, text is (the product of) a process
of meaning – the highest-ranking unit of the stratum of semantics.7 So it can be characterized
as language functioning in context (cf. Halliday and Hasan, 1976), taking the stratal location
and the view from above, from the point of view of context as criterial. But text as meaning is
realized as text as wording, and text as wording is, in turn, realized as text as sounding or text
as writing. It is important to recognize this realizational chain since many computational tools
only enable us to analyse text as writing (or, in the case tools like Praat, acoustically as sounding);
most higher-stratal analysis still has to be manual in nature.
Let us begin by specifying a simple system network representing options available to text ana-
lysts: see Figure 4.14. The system network is represented as a number of simultaneous systems,
each with two alternative options. The options are characterized in Table 4.11 together with
references to the SFL literature.
We will discuss the systems set out in Table 4.11 in some detail in the following subsections,
but first we will comment on one fundamental trade-off in text analysis that in principle affects
all kinds of analysis. This is the trade-off between the degree of comprehensiveness of the
analysis and the volume of text analysed: see Figure 4.15. Comprehensiveness is represented by
the vertical axis and volume by the horizontal axis. Assuming a fixed quantum of work avail-
able, the relationship is an inverse one: the more comprehensive the analysis, the smaller the
volume of text, and the less comprehensive the analysis – highly selective analysis, the larger

Figure 4.13 The location of text in the overall theory of language in context in terms of (i) the hierarchy
of stratification and (ii) the cline of instantiation

224
 Application guides

Figure 4.14 Choices open to text analysts

225
 Table 4.11 Choices open to text analysts and examples of relevant studies

System Term 1 Term 2

stage (of description) Pre-systemic Systemic

Text analysis undertaken typically to support the Text analysis referencing a systemic description, e.g. Halliday (1985b)
development of systemic descriptions – thus
illustrated in accounts of different languages
role (of text) Text as specimen Text as artefact
Text analysis undertaken to further the description Text analysis undertaken to illuminate a text treated as an object in its own right,
of a system, e.g. Halliday and James (1993), typically a highly valued text, e.g. text as verbal art: Halliday (1971), Hasan (1985b);
Halliday (1998a) but also other texts given value, including as exemplars (e.g. Halliday, 1992a, 1994b;
Martin, 1992b)
means (of analysis) Manual Automated

Application guides
Text is analysed manually, with patterns being Text is analysed automatically using various computational tools (ranging in power and
matched to descriptions of contextual and sophistication from those traditionally used in corpus linguistics to those used in NLP
226

linguistic systems. Computational tools may and information technology), typically focussed on “low-level” patterns. These include
be used to record the manual analysis, e.g. concordancing tools, parsing, data mining techniques (e.g. Wu, 2009; Teich et al., 2016;
O’Donnell’s (2012) UAM Corpus Tools. O’Halloran et al., 2016; Bateman et al., 2019), and for multi-semiotic (multimodal) studies
key-stroke logging (used in translation studies) and eye tracking (used in both translation
studies and film studies), e.g. Alves et al. (2010), Tseng, Laubrock and Bateman (2021).
state (of text as data) Raw text Annotated text
Most text analysis projects in SFL have probably Teich (2009: 115–124)
operated with “raw” texts but there’s a shift Annotated “multimodal” texts: Hiippala (2014); Zhang Peija (2018)
towards value-added texts, i.e. annotated texts.
phase of text Text as product Text as process (logogenesis)
Most instances of systemic functional text analysis Analysis of text as unfolding through time: Matthiessen (1995a, 2002b); Zappavigna
have focussed on texts as product. (2011); Caldwell and Zappavigna (2011); Gregory’s phasal analysis (e.g. Gregory,
1984, 2002; Malcolm, 2010)
multi-semiosis Text alone (monomodal) Text + other semiotic “texts” (multimodal) [MDA]
Many examples in the systemic functional Painter, Martin and Unsworth (2014); Bateman, Wildfeuer and Hiippala (2017);
literature; for models, see e.g. Halliday (2002e) O’Halloran et al. (2016); see also Section 3.2.
 Application guides

Figure 4.15 Methodological trade-offs in text analysis – comprehensiveness of analysis and volume of texts
analysed

the volume of text that can be analysed. Comprehensiveness encompasses both stratal and
metafunctional coverage.
Assuming texts are chosen as specimens of something more general along the cline of instan-
tiation in the direction of the potential pole, like registers (text types) or the overall system of
language, then we can and must be clear about minimal and maximal reliability and potential
for analytical bias. If the analysis is highly selective and the volume of text analysed is small, then
the analysis will be minimally reliable and subject to analytical bias. If the analysis is exhaustive
relative to a comprehensive description and the volume of text is large, then the analysis will be
maximally reliable and it should not be subject to much analytical bias. If the analysis is quanti-
tative in nature, then there are of course various general statistical tests to determine reliability
and significance, like chi-squared test, and software packages supporting quantitative analysis
like SPSS. (For discussion of corpus size relative to the purpose of investigation, see also Caruso
et al., 2014. For a discussion of “big data” in multimodal analysis, see O’Halloran, Gautam and
Jin, 2021.)
There is another related trade-off – between the degree of automation of the analysis and
the volume of texts that can be analysed: see Figure 4.16. The potential for automation depends
on the “level” of analysis, level being first the stratum of analysis (graphology or phonology,
lexicogrammar, semantics, context) and second the rank within a given stratum. The potential
for automation decreases as we move upwards: automation of analysis of graphological patterns
is straightforward; various tools work at this “level”, like frequency counters (e.g. visualized by
means of word clouds), concordancers. At the next stratum, some degree of automation is also
possible: word class taggers (or “part of speech” [POS] taggers) are widely available for different
languages, and there are now also parsers for general use such as the Stanford Parser (e.g. Teich,
2009; Bateman et al., 2019: 564). However, such parsers do not provide rich systemic func-
tional lexicogrammatical analyses – the kind of lexicogrammatical needed to support a stratal
ascent into semantic analysis; here we still have to rely on manual analysis (although researchers

227
 Application guides

have conducted interesting experiments with second-order parsing of standard parser output to
approximate systemic functional analysis). For discussions of the challenges involved in full-scale
systemic functional parsing, see e.g. O’Donnell and Bateman (2005), Teich (2009), Bateman and
O’Donnell (2015), and Bateman et al. (2019: 565–570).
As we move further upwards, to semantics and then to context, the chances of automated
analysis diminish further. There are systems for automated sentiment analysis (e.g. Faulkner,
2014: Chapter 2), but they are often based on look-up of orthographic words in dictionaries, so
they can only provide rough sketches of appraisal analysis. However, more sophisticated forms
of automated analysis are possible, as Faulkner (2014) shows in his automated analysis of positive
vs. negative stance in student essays based on classification models of text and clause. Similarly,
“semantic taggers” such as UCREL (https://ucrel.lancs.ac.uk/usas/) still tend to be limited in
their abilities although the push for “meta-data” as a way of managing the ever-expanding flow
of text is growing in information technology and related fields.
Taking the volume of texts analysed into account, we can note that manual analysis will
restrict us to small samples of text – probably no larger than 100,000 words unless ample fund-
ing is available to employ a team of experienced analysts or some form of crowdsourcing can be
devised. Automated analysis will allow us to process volumes of any size, but we are restricted in
the “level” of analysis to low levels.
One way of dealing with these constraints is to develop an approach that combines auto-
mated analysis of large volumes of text at a low level and manual analysis of small samples taken
from such large samples, as Halliday (1998a) does in his study of the grammar of pain and
O’Halloran et al. (2016) do in their investigation of violent extremist text.

Figure 4.16 Methodological trade-offs in text analysis – automation of analysis and volume of texts
analysed

228
 Application guides

4.4.3 Stage of description: pre-systemic vs. systemic

The great bulk of systemic functional text analysis has been systemic in the sense that it has relied
on existing systemic functional descriptions of systems at different strata of different languages.
However, as part of the development of the description of a language that has not previously
been described in systemic functional terms, it is necessary to undertake what we might call
pre-systemic analysis – analysis designed to produce evidence for generalizations about dif-
ferent systems.
In systemic text analysis, analysts use existing descriptions of any or all of the strata of
language in context (or of other semiotic systems), and match patterns in the text to options
in the relevant systems, often using (fragments of) function structures as a step in the systemic
analysis. For English, there is now a reasonably complete set of books that can be used as guides
informing the analysis, including:

• context: Martin and Rose (2008) on analysis of context;

• semantics: Martin and Rose (2007) on analysis of written text in context;
• lexicogrammar: Martin, Matthiessen and Painter (2010), designed as a workbook for Hal-
liday’s IFG;
• phonology: Halliday and Greaves (2008) on prosodic analysis, i.e. intonation and rhythm.

Informed by the account presented by Martin (1992a), Eggins (2004) provides an introduction
to all the strata of English above phonology/graphology, as do Eggins and Slade (2005) for the
analysis of English casual conversation. Reference sources that can be consulted in lexicogram-
matical analysis are summarized in Table 9.4.
Systemic analysis of text shows actual choices made in text, but always against the background
of choices that could have been made. This evocation of “shadow texts” as possible variants of the
text being analysed can be a good basis for the interpretation of results of the analysis (cf. Mat-
thiessen, 2001; Johansson, 2007), bringing out consistent selections and syndromes of features, as
illustrated by a number of the examples of lexicogrammatical text analysis in Table 4.10 earlier.
Systemic text analysis can also be used to record the choices made as a text unfolds in time –
text segment by text segment. Here text is viewed as process rather than as product: see Section
4.4.7 on the phase of text being analysed.
In pre-systemic text analysis, systemic functional linguists often sample texts in the com-
munity of speakers of the language they are targeting, taking care to select texts from a variety of
registers, including ones likely to give early insights into one metafunction or another, as illus-
trated for Isaac Mwinlaaru’s description of the lexicogrammatical system of Dagaare in Table 4.8
and Zhang Dongbing’s description of the interpersonal systems of Khorchin Mongolian in Table
4.9. For example, to begin the exploration of the textual organization of clauses, one should
select texts from registers characterized by distinct methods of development, e.g. chronological
vs. taxonomic. For further discussion of the use of text analysis in the development of systemic
functional descriptions, see Section 4.3.

4.4.4 Role of text: text as specimen vs. text as artefact

In text analysis, texts may be sampled to serve either of two distinct roles: they may be of inter-
ests in their own rights as artefacts given value in a culture or sub-culture, or they may simply
serve as evidence for some generalization being sampled as specimens. We can characterize this
distinction in terms of the cline of instantiation.

229
 Application guides

If texts are treated as artefacts, we stay at the instance pole of the cline, and illuminate the
particular texts because they are of inherent interest. This is the case with texts that serve as
verbal art – as literature (cf. Hasan, 1985b, forthcoming; Birch and O’Toole, 1988; Lukin and
Webster, 2005; Miller and Turci, 2007; Lukin and Butt, 2009; Webster, 2015b). But many other
kinds of texts may similarly be of interest in their own right, e.g. political speeches, historically
significant texts like Darwin’s writings, certain laws such as the Hong Kong Basic Law and
declarations such as the Declaration of Human Rights. (Note that the term “artefact” has also
been used in the linguistic and semiotic literature to denote an instance of a semiotic system,
typically other than language, i.e. the analogue of a text instantiating the meaning potential of
that semiotic system.)
If texts are treated as specimens, we are interested in patterns higher up the cline of instanti-
ation, beyond the texts that we have sampled – say, moving up to a particular register (e.g. Bow-
cher, 2001), or further up to illuminate codes (cf. Hasan, 2009b), or all the way to the linguistic
system (cf. Figure 4.16). So in sampling texts at the instance pole of the cline of instantiation,
we have to ask what kinds of generalization the text sample is expected to support, i.e. where
along the cline of instantiation we want to locate our generalizations. The higher up the cline
we want to move, the larger the text sample must be. But of course it’s not just a matter of size;
we must consider the registerial composition of the sample as well,8 ensuring that the registerial
composition is commensurate with the claims we want to make about our description in terms
of its location along the cline of instantiation.

4.4.5 Means of analysis: manual vs. automated

When systemic functional text analysis started in the 1960s, the only feasible option open to
analysts was manual text analysis, and many projects involving text analysis are still based on
manual text analysis. However, tools enabling automated text analysis have become increasingly
sophisticated and powerful, as recent studies such as Teich et al. (2016) demonstrate. Still, there
is an upper limit to automation, as indicated in Figure 4.17 and explained earlier in connection
to the figure. In their study undertaken about three decades ago using the tools available in the
Birmingham School of corpus linguistics, Halliday and James (1993) use automated analysis of
a large sample of text, a subset of the Birmingham corpus (the “Bank of English”), in order to
infer grammatical probabilities based on relative text frequencies. They discuss the challenges
they faced, e.g. due to ambiguous patterns, and methods that they used to deal with them (like
manually examining a manageable number of instances).
For analysts, it is essential to be aware of the trade-off between manual and automated analy-
sis, and to make the choice in an informed way, rather than simply follow the beaten path. There
has been a tendency in linguistics outside SFL to forge ahead with corpus linguistic methods
without taking a step back to ask how much of the total picture they can actually reveal. Often
results are reported in quantitative terms, without giving readers a feel for the texts that have
been analysed, also leaving their contexts of situation underspecified. A good check is to ask
how far any given investigation has moved in the direction of semantics, and, by another step,
context.
The potential tension between large-scale automated analysis and small-scale manual analysis
is illustrated by the contrast between corpus linguistics and Conversation Analysis. However,
SFL provides the framework needed for a judicious selection of mixed methods in this area – or
better: complementary methods. Many systemic functional linguists have in fact adopted
complementary methods in their research based on text analysis (for multimodal analysis, cf.
O’Halloran et al., 2016); one example is provided by Halliday (1998a) in his investigation of the

230
 Application guides

“grammar of pain”, where he both analyses a medical consultation manually and applies corpus
tools to a large general corpus. In his plenary talk at ISFC 2002 hosted by Liverpool Univer-
sity, Michael Hoey – not a systemic functional linguist, but a highly valued fellow traveller –
addressed the complementarity in terms of his personal experience. He noted that he had started
out as a discourse analyst in Birmingham (e.g. Hoey, 1983), then become addicted to corpus
linguistic methods, but then realized where he had come from – his origins as a discourse ana-
lyst. So he advocated adopting what we have called complementary methods.
It can be very productive to work with a large corpus – ideally one of “situated discourse” –
and investigate the whole corpus by means of automated analysis while also choosing what
appear to be representative shorter and therefore manageable samples and analyse them manually.
The choice of the starting point will depend on various considerations, including the register(s)
of the corpus and the research goals. But doing a pilot involving manual analysis of short passages
can be very productive. (1) On the one hand, this can reveal interesting patterns to pursue by
means of automated analysis of “low level” manifestations of these patterns in the whole corpus.
For instance, a manual analysis of short passages based on the system of transitivity may indicate
particular verbs serving as the Event of the verbal groups realizing the Processes of the transitivity
configuration, and these verbs can then be explored through automated analysis of the whole
sample – or combinations of verbs and nouns (e.g. drop + song [and names of other semiotic
artefacts]) or verbs and adverbs (e.g. want + badly), taking the investigation into the territory of
collocations and colligations, or constructions. (2) On the other hand, it will reassure the analysts
that they really understand what they are doing and have the expertise and skills to undertake
high-level analysis – giving them insights that they can draw on when they interpret the findings
produced by the automated, large-scale analysis. However, it is obviously also possible to use
automated analysis as part of a kind of fishing expedition, trawling through vast volumes of text.
The findings of the analysis may suggest patterns of interest, e.g. clusters or N-grams – patterns
worth investigating by means of higher-level manual analysis. Thus the system set out in Figure
4.14 specifying the choice between ‘manual analysis’ and ‘automated analysis’ should probably
be cycled through more than once in every text-analytic research project.
As an aside, but a very important one, we should note that there is considerable variation with
respect to the challenge involved in automated different kinds of analysis. We have indicated that
one very general indicator is the “level” of the analysis (the stratum, and within a given stratum,
the rank): the “lower” it is, the easier it is to automate. But there are also other considerations,
centrally ones related to the spectrum of metafunction and to the cline of delicacy. For example,
within lexicogrammar, it is still quite hard to automate experiential analysis (i.e. transitivity
analysis), but interpersonal lexical analysis is within the reach of automatic analysis, as evidenced
by the proliferation of computational approaches to “sentiment analysis” (see Whitelaw, Garg
and Argamon, 2005; O’Donnell, 2017: 567) – roughly the attitudinal part of the resources of
appraisal (e.g. Martin and White, 2005).
For tools for recording manual analysis and tools for conducting automated analysis, see
Section 9.1.3.

4.4.6 State of text as data: raw text vs. annotated text

The state of text as data is clearly of fundamental empirical significance. The general principle is
this: any decision about how text is to be represented as data is a theoretical one, and “raw text”
should preserve as much information as possible so that this information can be accessed when it
turns out to be needed (just as in the case of “raw” versions of photographs). Thus any annota-
tion added in order to indicate semantic, lexicogrammatical, phonological or graphological unit

231
 Application guides

boundaries should be such that it does not interfere with the “raw text” (cf. the development of
stand-off mark-up schemes).
Annotation is always a value addition to text as data. This is obvious when we are concerned
with “meta-data”, including demographics; but it actually applies equally to the data. As just
noted, the key issue is actually not whether text has been annotated or not but rather how the
annotation is represented. For example, in the published version of the London-Lund corpus
(Svartvik and Quirk, 1980) – a truly pioneering corpus of spoken English, there is no “raw”
transcript; the transcript has already been annotated prosodically, basically following Crystal’s
(e.g. 1969, 1975) approach (see also Crystal and Quirk, 1964). The same has tended to be the
case with conventions for transcription in Conversation Analysis. This may be good for certain
specific investigations, but for many others, the researcher will need to remove the annotations
first in order to make this unique corpus generally useful and searchable. The fundamental les-
son is this: annotations must be possible to separate from the transcript, and the transcript should
ideally be such that it can serve as data for different analytical approaches. Thus any concern that
annotation will interfere with the raw text (cf. Sinclair, 1992) is unwarranted; it should always
be possible to access the raw text without the added annotation. In the Lancaster tradition of
corpus linguistics, there has been considerable emphasis on the development of corpus tools
(e.g. McEnery and Hardie, 2012), and there have been many sophisticated contributions from
researchers in NLP (Natural Language Processing).
Text may of course be either pre-annotated, and then made available as tagged text, or
by another step, syntactically analysed text, as in the case of “tree banks”, or annotated “on
the fly” using one of the many annotation tools that are now available (e.g. Wu, 2009; Teich,
2009; Sharoff, 2017; Bateman et al., 2019). (Syntactic “tree banks” can be further augmented
with discourse annotation, a process that may need manual input: see e.g. Prasad et al., 2008.
And they can also serve as the input for further automated analysis, as shown by Honnibal,
2004.)
The issue of how “raw text” is represented is of course a fundamental one. Halliday (2002c)
notes that corpora have tended to be represented in such a way that they support lexical research
more easily than grammatical research, and by another stratal step, this also applies to the task
of semantic analysis. But this is also a question of what computational tools are available for the
automatic analysis of corpora.

4.4.7 Phase of text: text as product vs. text as process

Text analysis can focus on different phases of text unfolding in context at the instantial pole
of the cline of instantiation. The basic choice is between focussing on text as process – text
unfolding through time – or text as product – the final phase of text unfolding through time.
Throughout most of the history of text analysis (or “discourse analysis”), the focus has been
on text as product, a written text or a transcript of a spoken text; it is illustrated by most pub-
lished studies of systemic functional text analysis – virtually all of the studies listed in Table
4.10 earlier.
However, the need to engage also with text as process has long been recognized, foreshad-
owed in Halliday’s (1961) interpretation of “language as patterned activity”, an activity which
“takes place in time”. Halliday’s (1977b) characterization of text as semantic choice in context
opened the possibility of analysing it as a process of successive choices, later taken up by Matthies-
sen in representations of texts as scores of systemic choices (e.g. Matthiessen, 1995a, 2002b), the
term text score being a translation of Weinreich’s (1972) Textpartitur. Martin (1985a) raised the
issue of synoptic and dynamic perspectives on text, referring also to the research by Eija Ventola,

232
 Application guides

reported in Ventola (1987); and O’Donnell (1990) presented a dynamic model of exchange (see
also O’Donnell and Sefton, 1995) and then of context (O’Donnell, 1999). And dynamic con-
siderations were also foregrounded in Bateman’s (1985) emphasis of intersubjective achievement
in the study of dialogue (see also Thibault, 2005). At the same time, researchers developing the
Penman text generation system under Bill Mann’s direction also had to grapple with process
accounts of text, more specifically with the process of text generation. Here text needed to be
interpreted as emergent – again as a process of successive choices. This computational modelling
fed into work on logogenesis and processes of instantiation in SFL starting around 1990 (Mat-
thiessen, 1993b, 1995b). Illustrations of how logogenetic patterns can be visualized are provided
by e.g. Matthiessen (1995a, 2002b), Zappavigna (2011), Caldwell and Zappavigna (2011), and
Leong (2019). When lexicogrammatical analysis is visualized logogenetically, it is usually pos-
sible to detect patterns that can be interpreted semantically, and also contextually; for example,
elements of contextual structure may show up as lexicogrammatical chords in the lexicogram-
matical text scores. Viewed in terms of the completion of the unfolding of a text, these patterns
will be reflected in the character of the text (cf. Halliday, 1981/2002d).
Within his “communication linguistics” (Gregory, 1985), working with Elissa Asp and
Karen Malcolm, Michael Gregory introduced phasal analysis as a way of bringing out the
phases of the unfolding of text (see e.g. Gregory, 1984, 2002; Stillar, 1991; Cloran, Stuart-
Smith and Young, 2007; Malcolm, 2010; Bartlett, 2016; and cf. also for multimodal analysis,
Baldry and Taylor, 2002; Taylor Torsello and Baldry, 2005). Gregory (1984: 9) characterizes
phase as follows:

It is a construct which relates language activity which is dynamic to the synoptic lan-
guage code. Phase characterizes stretches within discourse (which may be discontinu-
ously realized in text) exhibiting their own significant and distinctive consistency and
congruity in the selections that have been made from the language’s codal resources: its
semological, morphosyntactic and phonological resources for encoding and decoding
ideational, interpersonal and textual meaning.

Phasal analysis is thus concerned with the process of instantiation and will bring out logogenetic
patterns in text (as Weinreich’s, 1972, Textpartitur does), covering all metafunctions and cutting
across strata and ranks within a given stratum.
Advances in technology have helped researchers in various traditions concerned with text
analysis track features in the unfolding of texts in contexts. The technologies that have played a
particularly “game-changing” role are key-stroke logging and eye movement tracking, but
also researchers can increasingly get access to brain imaging techniques with sufficient temporal
and spatial resolution. One area where the techniques have been used in the analysis of text as
process is TPR, translation process research (e.g. Jakobsen, 2014, 2017), and this strand of
research has now been combined with systemic functional text analysis (e.g. Alves et al., 2010;
Serbina, Niemietz and Neumann, 2015).

4.4.8 Multisemiosis: mono vs. multi (analysis of

multisemiotic texts (“MDA”))
The focus of text analysis is clearly text in context – i.e., language functioning in context (cf.
Halliday and Hasan, 1976). Here any given text is thus located at the instance pole of the
cline of instantiation, instantiating the meaning potential of some particular language. This
kind of text analysis is mono-semiotic in the sense that deals with one semiotic system, viz.

233
 Application guides

language, and most classic examples of systemic functional text analysis have been mono-
semiotic in this sense (though with recognition of other semiotic systems, as in Halliday,
1978) – Steiner (1988) being an important early example of linguistic and musical analysis of
a multi-semiotic text, a folk ballad (for Firthian, pre-SFL analysis, see Robin and McLeod’s,
1956a, 1956b of Yurok songs).
Mono-semiotic text analysis dealing with spoken language may of course include pro-
sodic analysis, i.e. intonation and rhythm, which have been part of the description of language
from the start (e.g. Halliday, 1967). The analysis of tone is an essential aspect of interpersonal
analysis, of both the lexicogrammatical system of mood and the semantic system of speech
function. And this may be extended to voice quality, as in Wan Yau Ni’s (2010) analysis of call
centre service encounters. Similarly, the analysis of tonicity and tonality should be included in
textual analysis, in particular in the analysis based on the system of information. The prosodic
analysis of spoken text together with lexicogrammatical and semantic considerations is explained
and illustrated by Halliday and Greaves (2008). Martinec (2018) demonstrates the insights gained
from rhythmic analysis as part of the semantics of text.
Mono-semiotic text analysis may, obviously, be focussed on semiotic systems other than
language – e.g. analysis of gesture, gaze, proxemics; analysis of choreography; analysis based on
pictorial systems (image analysis drawings, paintings, photographs, video); analysis based on math-
ematics and logic; analysis of sculpture, of architecture, of museum exhibitions. See Section 3.4,
and Table 3.19, for accounts and references. However, the analysis of “texts” engendered by a
semiotic system other than language is often undertaken in combination with other semiotic
systems – multi-semiotic analysis.
In multi-semiotic analysis of texts, often referred to as multimodal discourse analysis
(MDA, or MMDA), “texts” are analysed in terms of two or more semiotic systems, which
is typically the case in film analysis, as shown by Baldry and Thibault (2006), Bateman and
Schmidt (2012) and Bateman, Wildfeuer and Hiippala (2017). But multi-semiotic analysis of
interactants in face-to-face interaction, or in computer-mediated interaction, is of course also
highly desirable, and increasingly possible, thanks to both technological and descriptive advances
(cf. Martin and Zappavigna, 2019; Ngo et al., 2021). Naturally, the same tools and techniques
used in film analysis can be used in such analysis, including platforms such as ELAN for tran-
scription and analysis of different semiotics tracks, and techniques such as eye-tracking (e.g.
Tseng, Laubrock and Bateman, 2021). In both cases, one challenge is to identify descriptions of
the semiotic systems under focus in the analysis, so that the analysis can be systemic in the sense
that it is grounded in the description of the relevant semiotic potential (in terms of both content
and expression), as illustrated by Martinec (2004) for gestures that operate together with spoken
language. For a helpful account coming from outside SFL of research methods in dealing with
“multimodal data”, see Norris (2019); for a collection of papers drawing on different theories,
including SFL, see Maiorani and Christie (2014).
Multi-semiotic analysis of written text accompanied by other semiotic systems, typically ones
expressed visually (as opposed to say music or sound effects), has so far been given more atten-
tion in SFL, as is evident from Table 3.19, and the same is true of other approaches to “mul-
timodal discourse analysis” in various traditions; cf. the examples given by Kress (2012) – one
problem clearly being how to present examples from face-to-face interaction in writing without
technical representation. But multimodal interaction analysis has, naturally, focussed on face-to-
face interaction (e.g. Norris, 2004, 2019).
To round off the discussion of the multi-semiotic option, we present an adapted version of
O’Halloran et al.’s (2016) “examples of text and image systems” (noting that their “image sys-
tems” are based on O’Toole, 1994): see Table 4.12.

234
 Table 4.12 Adaptation of O’Halloran et al.’s (2016) “examples of text and image systems”, with further lexicogrammatical details

Semiotic system Metafunction Rank System Description

language (text) logical clause logico-semantic type & taxis: construal of a sequence of figures (contributing to the
(complex) [parataxis:] 1 2 3 . . . and [hypotaxis:] representation of episodes, methods and other more
αβγ... extensive semantic patterns)
experiential clause transitivity: process + participants + construal of a figure (doing-&-happening, behaving,
circumstances sensing, saying, being-&-having, existing)
interpersonal clause mood: Mood + Residue enactment of a proposition or proposal as a dialogic
move, realizing speech functions
modal assessment: mood and comment enactment of assessment of proposition or proposal, or
Adjuncts; lexical connotations act of exchange
textual clause theme: Theme + Rheme presentation of experiential and interpersonal
meanings as message organized into local context for
(orientation) and the remainder of the message

Application guides
information information: Given + New presentation of experiential and interpersonal meanings
unit as message organized into information presented as
given and information presented as newsworthy
235

pictorial semiotic (picture) experiential work narrative motif; representation; setting nature of the scene
episode process + participants + circumstances visual construal of (doing-&-happening, behaving,
sensing, saying, being-&-having, existing)
figure posture; dress characteristics of the participants
interpersonal work angle; camera distance; lighting visual effects
episode proportion in relation to the whole happenings, actions and relations with respect to the
image; focus; perspective whole image
figure gaze – visual address direction of participant’s gaze as internal to image or
external to viewer
textual work compositional vectors; framing the organization of the parts as a whole, with the visual
marking (e.g. framing) of certain parts
episode relative placement of episode; framing position of the happenings, actions and relations in
relation to the whole image, and the visual marking
of certain aspects
figure relative placement of the figure within position of figures in relation to happenings, actions and
the episode; arrangement; framing relations, and the visual marking of certain aspects of
those figures
 Application guides

4.4.9 Analysis and interpretation

In a very real sense, the analysis of text is not the final step in the process of engaging with it; it
needs to be followed up by an interpretation of the results of the analysis. The interpretation is
a kind of 2nd-order analysis; it takes the results of the 1st-order analysis and transforms them into
answers addressing the original research questions in the project that is based on text analysis.
A number of the studies listed in Table 4.10 illustrate how the results of text analysis can be
interpreted to yield a rich account of how texts in context “mean” – adapted to the register
or registers that they instantiate. For example, as Hasan (1985b) shows, in the interpretation of
verbal art based on text analysis, one would strive to discern major motifs, revealing the symbolic
articulation of the text illuminated by analysis.
Thompson (2014: 117–119) provides a helpful discussion of how to interpret the results of
transitivity analysis. In general, there are a number of steps that may help researchers to move
from analysis to interpretation, guided by the semiotic dimensions of the systemic functional
architecture of language in context:

• Frequency of instantiation of systemic options: what are the common systemic selec-
tions? Even some simple counts of absolute numbers and relative frequencies can be very
helpful when one is looking for patterns, and the next step of trying out charts that will
visualize the quantitative information can lead to breakthroughs in teasing out patterns
that are significant. An obvious example is the selection of marked spatio-temporal topical
Themes serving to indicate episodic shifts in a narrative text, but depending on the text,
there may be patterns of selection that give the text a special instantial character, beyond
its registerial characteristics (Halliday, 1981/2002d).
• Alternative systemic options: how could the text be different? The analysis specifies the
systemic selections, and the function structures, that occur in the text, but these actual selections
can be seen and interpreted against the background of the selections that could have been made
according to the systemic description used in the analysis, thus looking for the cloud of shadow
texts surrounding the actual text. Here it can be very helpful to compare and contrast one text
or set of texts with another, and one can learn a great deal about texts by revising aspects of
them. This is certainly the case in translation studies based on text analysis, but the use of shadow
texts is helpful in many research contexts, including of course language education.
• Higher-stratal systemic options – stratal ascent: what are the interpretations of selec-
tions at the highest stratum of analysis at the stratum above it? Can the patterns of selections
within systems (logogenetic patterns) at one stratum be interpreted by reference to stages at
the stratum above? (A “local” version of this within a stratum is ascent in rank; for example,
if a text has been analysed automatically by means of computational tools, the output is likely
to be low-ranking patterns that can be interpreted at higher ranks, perhaps pointing to pas-
sages of texts that can be analysed manually to complement the automatic low-rank results.)

These questions all involve exploring the environment of selections recorded in the analysis of
texts, moving along one dimension or another: shunting along semiotic dimensions can yield
insights supporting the process of interpretation the output of analysis. It is also helpful to look
for patterns of features occurring frequently together:

• Frequency of combinations: in simultaneous systems, are certain combinations more

likely to occur together rather than other ones? If there is information about the patterns in
the linguistic system in general, are these combinations consistent with the general tenden-
cies or do they depart from them? (If we think of selections of systemic options as notes in

236
 Application guides

a text score, combinations can be thought of as chords; see Matthiessen, 2021b). Selections
in different systems may “conspire” to produce a certain effect, as in a personal account of
a student’s experience of Covid-19 where causative selections in verbal group complexing
(realizing ‘effective’ agency) and high selections of modulation: obligation together create a
sense of a highly constrained world during the pandemic.
• Syndromes: are there more extensive combinations of selections, involving a wider range
of systems? Halliday (1990b/2003b: 150) characterizes syndromes as “assemblies of features
from different parts of the grammar which typically co-occur, in part reinforcing each other
and in part contradicting each other, but constituting as a whole a distinctive meaning
style”; we can of course identify syndromes of features from any strata within language in
context. For example, is it possible to detect metafunctional resonances, e.g. consistencies
between what is construed ideationally and what is enacted interpersonally? Martin (1999b)
brings out the construal and enactment of freedom through the analysis of multiple systems
from different metafunctions. (On metaphorical syndromes, see e.g. Halliday and Matthies-
sen, 1999: 250–255; Halliday, 2009b: 245. For an example of lexical metaphor in relation
to appraisal, see Fuoli, Littlemore and Turner, 2021.)

The questions just listed are general ones that can be asked of a wide range of systems, but they
can be supplemented by questions concerned with the semiotic work done by particular sys-
tems, as illustrated in Table 4.13.

Table 4.13 Examples of questions to ask to develop interpretations of the results of analysis

Metafunction System Questions Examples of studies

with interpretations

experiential transitivity “world view”: in what participant roles are entities Halliday (1971);
construed? Do they form a hierarchy of potency Matthiessen
(e.g. who can do what to whom?) (2015f)
interpersonal mood “dialogic control”: in what speech function roles are
interactants enacted? Are they distributed evenly,
or do certain interactants tend to take up particular
roles with respect to other interactants? Who tends
to initiate, who tends to respond?
modality modulation: “constraints”: to what extent is the world Halliday (1982)
constrained by necessities, responsibilities and
requirements and enabled by abilities?
textual theme “method of development”: what relations are Halliday (1990c)
marked by textual Themes? Are there recurrent
circumstantial types among marked topical Themes?
Do unmarked topical Themes make certain
references or lexical cohesive relations prominent?
information “main point”: do the selections of (the Focus
of) New in a spoken text or the Culminative
elements in a written text form a certain pattern,
like a cumulative pattern of positive or negative
assessments in a persuasive text or frontiers of new
knowledge in an expounding text?

237
 Application guides

There are obviously myriads of questions about systemic details. For example, after analysing
a spoken story, one might ask if the “narrative tense” is ‘past’ or ‘present’, or if there are shifts
between the two, and if the shifts can be interpreted as a manipulation of the immediacy story
by the story teller, as has been suggested in the literature. To get further inspiration about how to
go about interpreting the results of analysis, one can compile studies of texts relevant to the task
at hand, and identify the interpretative steps taken. Naturally, the interpretation of the analysis
results will have to be informed by the research questions asked at the beginning of the project,
although it is entirely possible that researchers will find that they need to adjust the questions
once the results start coming in.

4.5 Language development, language and education

One of the first areas of application of SFL was education, going back to projects in the UK in
the 1960s (e.g. Mackay, Thompson and Schaub, 1970; Pearce, Thornton and Mackay, 1989).
Since the 1960s, the study of language in education emerged as a distinct field of study and
application – educational linguistics has been central in SFL research and application since the
1960s (e.g. Halliday, 1988d, 1990a, 2007a; Hasan and Martin, 1989; Christie, 1994; Rothery,
1996; Christie and Unsworth, 2005). In terms of approaches coming from, or relevant to, edu-
cation, key scholars have included Basil Bernstein, Lev Vygotsky, Jerome Bruner and Colwyn
Trevarthen – and see also Feez (2010) on Maria Montessori.
Seminal contributions were made in Australia, including those by members of J.R. Martin’s
group, which came to be known as the “Sydney School” (cf. Rose and Martin, 2012) and also
parallel strands (e.g. Williams, 1994, 2005a, 2016, 2019; French, 2013; Derewianka and Jones,
2016), but educational linguistics based on SFL is now a worldwide undertaking. McCabe
(2017) provides an overview dealing specifically with language teaching, and McCabe (2021)
chronicles all the phases in language development that have been studied in SFL. So far, most
work in educational linguistics in SFL has been devoted to the mother tongue education, to
learning through language in different school subjects (e.g. Christie and Derewianka, 2008) and
also more recently certain university disciplines (e.g. Gardner and Nesi, 2013, on the corpus of
British Academic Written English; Hao, 2020, on biology; Martin, Maton and Doran, 2020a,
on academic discourse).
However, while there are pioneering contributions (e.g. Byrnes, Maxim and Norris, 2010),
SFL has so far been less involved in second and foreign (“L2”) education, as McCabe (2017)
points out (despite Halliday’s own early participation in the area and despite the input of SFL
into communicative language teaching in the 1970s). There are various reasons for this. The
metaphor of “acquisition” in “Second Language Acquisition” (SLA) gained popularity, but was
misleading, as Halliday has noted, and has actually done damage, as argued forcefully outside
SFL by Larsen-Freeman (2011, 2015). The cognitive orientation of SLA originally left no room
for socio-interactive consideration; this was the mainstream approach at the time, not the later
contributions informed by Vygotsky’s theory. Studies of L2 development tended to be cross-
sectional rather than longitudinal (cf. Ortega and Byrnes, 2008), which has meant that we don’t
actually know anywhere near enough about L2 learners’ paths as they set out to mean in a new
language under different conditions. However, the situation is changing: outside SFL, scholars
have noted that the socio-interactive turn is now following the cognitive turn, that longitu-
dinal studies are essential, and that the metaphor of “acquisition” needs to be replaced with
development (cf. Perret, 2000). The academic context should thus be much more conducive to
systemic functional contributions, and McCabe (2017) notes the changing trend. One recent
contribution nurturing this trend is a special issue of the journal System devoted to L2 education

238
 Application guides

informed by SFL involving different languages in varied institutional and national settings: see
Troyan, Herazo and Ryshina-Pankova (2022).
Like a number of other areas of application, educational linguistics can be characterized in
institutional terms: the role of language (and other semiotic systems) in institutions of education –
obviously including the family as a crucial site for pre-school learning (e.g. Williams, 1995;
Painter, 1999), kindergartens and pre-schools, primary schools (e.g. Williams and French, 2016),
secondary schools and universities (cf. Alsop and Nesi, 2009; Heuboeck, Holmes and Nesi,
2010; Gardner and Nesi, 2013), but also institutions whose primary focus is not educational
but in-service education takes place (e.g. Woodward-Kron et al., 2011; cf. Slade et al., 2015,
Chapter 7, on strategies for achieving effective communication in hospitals). While educational
linguistics is naturally centred on institutions of education, it has a much wider coverage; this
reflects both the fact that education begins before we enter institutions of formal education and
the fact that it is a life-long undertaking (cf. McCabe, 2021). Focussing on literacy in particular,
de Silva Joyce and Feez (2016) cover the full range of what they call lifespan literacies.

4.5.1 Learning how to mean: phases

SFL studies in educational linguistics have in fact been concerned with the different phases of
people’s lifespans or lifelines. If we recognize that ontogenesis spans all these phases, including
the end of (semiotic) life, we can represent these phases schematically as in Figure 4.17. The
figure is organized to indicate the timeline horizontally and the expansion of semiotic resources
vertically – the point being that as we go through life, if given the appropriate opportunities,
we can continue to “grow” our semiotic resources, enabling us to take on an increasing range
of institutional roles and engage in the fields of activity associated with them. Sadly, we may
experience shrinkage towards the end of life (senescence), but the diagram does not reflect this
common but not inevitable stage of the journey through life.9
In terms of institutions, this means that SFL studies have covered a range of institutions that
we are likely to meet as we proceed through life (cf. the introductory comments and references
earlier), starting with the family, continuing through institutions of formal education and then
with workplaces and other post-educational institutions (keeping in mind that teaching and
learning is a life-long prosody, finding settings in a wide range of institutions).
In terms of language and other semiotic systems, this means that SFL studies have covered
the period of “early childhood” (as in Halliday, 2003a) – pre-language > protolanguage > tran-
sition from protolanguage to language (mother tongue or tongues) > mother tongue(s) – and
then subsequent developments once meaners have begun to develop the meaning potential or
their mother tongue(s). The language of “early childhood” has been investigated in a number
of longitudinal case studies (the foundational ones being Halliday, 1975a; Painter, 1984; Torr,
1997; for overviews, see Painter, Derewianka and Torr, 2007; Torr, 2015; Painter, 2009, 2017;
and see also Williams, 2019, who also includes a summary of the study of codal variation
in language development). Longitudinal case studies of later developments are even harder to
undertake, with considerable practical challenges (cf. Ortega and Byrnes, 2008, on longitudinal
studies of second/foreign language learning), so they are rarer (e.g. Byrnes, 2009; Xuan, 2015),
but Derewianka (1995) has contributed a crucial case study of one child making the semiotic
transition from childhood to adolescence by gradually learning how to mean in the metaphori-
cal mode as he moves through the educational system. And it is possible to undertake quasi-
longitudinal studies, as done by e.g. Ryshina-Pankova (2011) and Quam (2020).
While the schematic diagrammatic summary in Figure 4.17 is based on phases of learning
language, it also includes indications of changes in how we learn through language (keeping in

239
 Application guides

mind that one special aspect of learning through language is actually learning about language
through language). As our linguistic and other semiotic resources keep expanding – as we grow
our personal meaning potentials, we are also increasingly empowered to learn through lan-
guage (and other semiotic systems). In terms of the nature of the knowledge that we construe
for ourselves as we move through semiotic life, this means a gradual accumulation of differ-
ent kinds of knowledge: folk knowledge (commonsense knowledge) > educational knowledge
(pedagogically tailored uncommonsense knowledge)10 > scientific knowledge (uncommonsense
knowledge).
In terms of learning how to mean, four phases can be distinguished (Halliday, 1975a, 2003a):
pre-language > protolanguage > transition (from protolanguage to post-infancy language, the
mother tongue(s)) > post-infancy, adult language (the mother tongue(s)). As SFL studies have
shown, children learn through their semiotic resources throughout – learning through proto-
language is crucial (cf. Halliday, 1984b, 1992c), but it is of course Phase III language – post-
infancy, adult language – that will serve as our resource throughout life after early childhood
(Halliday, 1993b; Painter, 1999). Here SFL has produced crucial insights into how a person’s
meaning potential keeps growing, as indicated in Figure 4.17, and how the expanding meaning
potential empowers learners to construe new (forms of) knowledge, as documented by Painter
(1999), and enact new relationships. The gradual addition of the semogenic strategy of gram-
matical metaphor expands the meaning potential of school children very considerably, as shown
by Derewianka (1995) in her longitudinal case study of one young learner. Another aspect of
the expansion of the meaning potential is the growth of people’s registerial repertoires through-
out formal education but also after graduation, in workplaces, recreational institutions – in any
institutions where they assume meaning roles. Based on research in Australian schools, Christie
and Derewianka (2008) paint a unique picture of this period in people’s lives. They show that
throughout their school years, students also keep adding to their lexicogrammatical resources
(e.g. the lexicogrammar for construing proofs), becoming increasingly empowered meaners (if
they succeed, of course). The focus on language education throughout the school years is fol-
lowed up by Christie (2012); see also McCabe (2021).
Based on Figure 4.17, we can understand that researchers concerned with the language of
early childhood and with language and education have adopted different research methods,
including different ways of collecting data. The studies of early language development have
all been longitudinal case studies based on the diary method – so authentic natural text,
which had been established before Halliday’s (1975a) pioneering study, notably by Leopold
(1939–1949, 1948). When Halliday undertook his study, audio recording would have been too
intrusive, and video-recording even more so; now of course the technology is much smarter
and less obtrusive, but there are obviously ethical issues that need to be addressed. As an indica-
tion of what has now become technologically possible, we can cite a study outside SFL, Roy et
al. (2006). They report on a very large-scale longitudinal project video-recording one child in
his home throughout the day: “To date, we have collected 24,000 hours of video and 33,000
hours of audio recordings representing approximately 85% of the child’s waking experience.
Over the course of the three-year study this corpus will grow six-fold.” To cope with the mas-
sive audio-visual corpus they amassed over the years (truly an instance of “big data”), they had
to deploy sophisticated automated computational tools for data mining and data visualization.
The findings are impressive, but at the same time they illustrate what happens when a kind of
“corpus-driven” approach is adopted with minimal guidance by linguistic theory and models
of learning how to mean. The combination of Halliday’s (1975a) framework and the extraor-
dinary resources Roy et al. (2006) had at their disposal would have yielded ground-breaking
results.11

240
 Application guides

Figure 4.17 Ontogenesis throughout life

The diary method seems best suited to studies of early language development since the
challenge of keeping a diary during later stages becomes nearly insurmountable, but the longi-
tudinal approach has been applied to later stages of language development – for the mother
tongue, notably by Derewianka (1995), and for the learning of a second/foreign language by
Xuan (2015). It is an approach that has been promoted outside SFL in the last decade and a half

241
 Application guides

in the study of second/foreign language development since the large volume of cross-sectional
studies actually give us very little or no information about learner paths; they give us no insight
into the extended process of learning how to mean in a second/foreign language (e.g. Ortega
and Byrnes, 2008; Byrnes, 2009).

4.5.2 The beginning of educational linguistics in Britain

Turning now to educational linguistics, we can begin by noting the connection to the studies
of early language development. One reason for undertaking the studies of early language devel-
opment was precisely because educators needed to have much more information about where
their primary pupils were coming from linguistically, and, importantly, what resources they
brought with them as they began their learning paths towards increasingly structured, educa-
tional knowledge (Halliday, 1975a). Thus Painter’s (1999) study sheds light on how young pre-
school children learn through language, and, crucially, how their expanding meaning potentials
empower them to learn in new ways and to expand their “knowledge bases”. For example, once
he had mastered the resources of ‘identifying’ clauses, one of the two children in her study was
able to construe abstractions by means of definitions.
Educational linguistics in the 1960s – or, perhaps to be more accurate, what was to become
“educational linguistics” – was first developed in the UK, during Halliday and Hasan’s British
period. Halliday (2005d: 133) writes:

When I first worked together with primary and secondary teachers, during the 1960s
in the UK, there was no term “educational linguistics”. I don’t know when this term
first came to be used – perhaps in the late 1970s; but when it did, I think it signalled
a significant moment in people’s thinking about language and learning: a recogni-
tion that, if you wanted to intervene in the processes of language education (literacy,
writing skills, mother tongue and second/foreign language teaching, language across
the curriculum), it was not enough just to “take language into account”; you needed
a language-theoretic approach, a coherent model of language in which notions like
“grammar” and “word” and “meaning” were integrated into some overall conceptual
framework. Then – but only then – you could start to explore which aspects of lan-
guage as system-&-process would be relevant as “abstract tools” for helping with this
or that specific educational task – to highlight the appropriate resources, and turn them
from knowledge base into resources for research and classroom practice.

(According to Google’s Ngram viewer, mentions of “educational linguistics” begin to appear

in the late 1960s and then rise steadily through the 1970s. According to Spolsky, 2008, “edu-
cational linguistics” was named by him as a field in 1974.) The work was first centred on
Edinburgh and then when Halliday moved to London he initiated and got involved in activi-
ties focussed on language and education – or language in education. Geographically, the next
area of development building on and expanding the projects in Britain was Australia, where
Halliday and Hasan moved in 1976 – joined by J.R. Martin in 1977. The first step towards
establishing Australia as a base for educational linguistics were taken in the second half on
the 1970s, and in the 1980s and beyond research and educational implementational activities
branched out (see e.g. Hasan and Martin, 1989; Martin, 1998; Rose and Martin, 2012; Wil-
liams, 2016). Particularly since the late 1990s, educational linguistics has been taken up and
been developed in many parts of the world, and it is now a global undertaking with many
centres of activity.

242
 Application guides

In many ways, Halliday, McIntosh and Strevens (1964) set the scene linguistically, and rep-
resented advances in educational linguistics, and applied linguistics more generally, during Hal-
liday’s time in Edinburgh. At over 300 pages, their book, The Linguistic Sciences and Language
Teaching, provides quite a comprehensive overview of “the linguistic sciences and language
teaching”. It is divided into two parts, “Part I The linguistic sciences” and “Part II The linguis-
tic sciences in relation to language teaching and language learning”. Part I covers linguistics and
phonetics, dialect variation and register variation, and comparison and translation. Part II covers
the role of linguistics in relation to language teaching and learning, studying the native language,
learning foreign languages, and English at home and abroad. The book is full of insights and
still worth consulting (cf. Strevens, 1987); theoretically it has not really aged – not because it is
“theory neutral” or “eclectic” but because it is theory empowered and the theory is not one of
the kind characterized by built-in obsolescence but rather one designed to be expanded over
time in terms of its coverage of phenomena. Still, of course, it also has to be read in its historical
context, as illustrated by the following passage (p. 268):

The grammar-translation method is being replaced by other teaching procedures, all

of which concentrate more on practical performance, especially in the spoken lan-
guage in the early stages, and many of which use electro-mechanical aids. These aids
range from the gramophone and tape recorder at one end of the case of complexity
and cost to language laboratories, audio-visual techniques and teaching machines at
the other.

During the 1960s a couple of important lines of investigation were undertaken in “educational
linguistics”. One was concerned with the teaching of English as the mother tongue, including
literacy education, and was directed by Halliday at University College London, starting with the
“Nufeld Project”, funded by the Nufeld Foundation, in the mid-1960s (e.g. Hasan, 2015:
117–118; Halliday in Martin, 2013: 182), followed by the “Schools Council Project on Linguis-
tics and English Teaching” initiated by Halliday. It was centred on early literacy and led to the
“breakthrough to literacy” (e.g. Pearce, Thornton and Mackay, 1989; Mackay, Thompson and
Schaub, 1970). This programme includes an innovation having to do with materiality; the mate-
rial challenge for children of mastering writing (graphetics) was separated from their mastering
of writing as a system (graphology).
Another line of investigation – or really extended research programme – was based on the
interaction between Basil Bernstein’s sociology of language and SFL (e.g. Hasan, 1999b). So far
there have been about three phases of the extended dialogue, although Maton and Doran (2017)
provide a more detailed historical overview (in terms of SFL, it is presented from the vantage
point of the “Sydney School”, e.g. Rose and Martin, 2012; for a chronology of this dialogue,
see Martin, Maton and Doran, 2020b), suggesting five phases – differentiating the third phase
discussed herein to three phases.
(i) The first phase took place mainly in Britain, although follow-up work continued in
Australia. This started with the dialogue between Bernstein and Halliday, Hasan and other
systemic functional linguists, prominently Geoffrey Turner; it included research at Bern-
stein’s SRU (Sociological Research Unit): see e.g. Bernstein (1971, 1973), with systemic
functional contributions: Halliday (1973), Hasan (1973), Hawkins (1973), Turner (1973);
Pickvance and Turner (1971). Central to this research was Bernstein’s theory of code, also
in relation to class and cultural reproduction, against the background of the societal problem
of educational failure and how to address it. To Halliday, Bernstein stood out as a social theo-
rist who took language seriously and gave it a central place in his theory. And SFL was as it

243
 Application guides

were primed for dialogue with social theory, going back to J.R. Firth (and his dialogue with
Bronislaw Malinowski). Reviews of the research during this period include Turner (1987)
and Halliday (1994d).
But the research was extended on a large scale by Hasan in Australia in the 1980s, based
on a corpus of dialogues between mothers and their young children from two social groups,
which can be referred to as working class and middle class (but which were very carefully
defined by Hasan in terms of autonomy): e.g. Hasan (1986, 1989, 2009b); Cloran (1989);
Hasan and Cloran (1990). Hasan’s research project is methodologically very important as a
model showing how such research can be conducted empirically. In addition to the carefully
compiled corpus, she and her team also demonstrated how the dialogues between mothers
and their young children could be analysed by means of semantic networks that she had
developed (e.g. Hasan, 1996b; Hasan et al., 2007) and then analysed statistically by means of
principal components analysis (see Section 4.2.3 earlier). This research was undertaken and
documented after the earlier debates around Bernstein’s work that grew out of misunder-
standings of his work by scholars from different fields, including sociolinguists (cf. Halliday,
1994d). Bernstein’s social theory from this period is illuminated by Atkinson (1985). He sets
it out in an accessible way, and he also deals with the (sometimes seemingly wilful, ideologi-
cally based and biased) misinterpretations of Bernstein’s work, e.g. the mistaken notion that
Bernstein’s code theory is a theory of deficit rather than a theory of difference and the point
that he draws on the original Durkheim rather than later British developments. In a recent
article, Ivinson (2018) also deals with misunderstandings of Bernstein’s “restricted code”,
and drawing on a range of studies, she suggests “re-imagining restricted codes as relational
assemblages”.
(ii) The second phase of dialogue between Bernstein and SFL focussed on Bernstein’s theo-
ries of types of pedagogy, of different organizations of discourse, of different knowledge struc-
tures, of framing and classification, and other related insights brought together by Bernstein
(2000). This was a very significant contribution to the systemic functional research programme
of analysing the discourses of different school subjects, and also university disciplines. This
dialogue started in the 1990s (Bernstein was invited to two ISFCs as a keynote speaker in the
1990s, to Ghent in 1994 and to Melbourne in 1999), and has continued since; one good source
is Christie and Martin (2007), which includes systemic functional insights based on Bernstein’s
distinction between horizontal and vertical discourse, e.g. Christie and Macken-Horarik (2007),
Wignell (2007b).
Perhaps the largest-scale empirical study demonstrating the significance of Bernstein’s dis-
tinction between horizontal and hierarchic knowledge structures in the construction of dif-
ferent fields of knowledge is the research project led by Giovanni Parodi (2010) at PUCV in
Valparaíso (for further developments, extended to seven disciplines, see Parodi, 2015). They
compiled a large corpus of undergraduate reading material assigned to undergraduate stu-
dents over four years in four different university disciplines. While they did not interpret their
findings in reference to Bernstein’s categories, it is very clear that the two disciplines within
material sciences are dominated by registers (“discourse genres”) characteristic of hierarchic
knowledge structure but the two disciplines within the human-social sciences are dominated
by registers characteristic of horizontal knowledge structure. In terms of field of activity, the
centre of registerial gravity of the material sciences can be located in ‘expounding’ contexts,
whereas the centre of registerial gravity of the human-social sciences can be located in ‘explor-
ing’ contexts. Parodi’s research project was concerned with readings assigned to students, and
we can also consider the complementary focus on written output by students sampled and
compiled into BAWE, the corpus of British Academic Written English, described by Nesi

244
 Application guides

and Gardner (2012). Their account of the distribution of registers, or “genres”, across different
disciplinary groupings (see Gardner and Nesi, 2013, on the “classification of genre families in
university student writing”, informed both by the Genre Model of the “Sydney School” in
SFL and the North American tradition). For the interpretation of these two research projects
in terms of Bernstein, see Matthiessen (2021c).
The second phase also included engagement with Bernstein’s account of different types of
pedagogy (1990, 2000). Martin (1998: 419) elaborates on Bernstein’s (1990: 185) representation
of distinctions in approaches to pedagogy as a topology where he locates four types of peda-
gogy, viz. progressive pedagogy, behaviourist pedagogy, critical pedagogy and social/psycho-
logical pedagogy. In Figure 4.18, we have complemented this topological representation with
a typological one, using a system network to show options available in pedagogic choices. The
intersections of the two sets of options define the regions in the topology representing the four
types of pedagogy.
(iii) The third phase includes the continuation of the work drawing on Bernstein (2000) in
dialogue with SFL. But on the one hand, it is characterized in a very pronounced way by the
different dialogic dyads that began to emerge in the 1990s, during Bernstein’s lifetime. On the
other hand, since Bernstein passed away in 2000, it has involved sociologists developing his work
in different parts of the work in exchange with SFL.
Maton and Doran (2017) chronicle and describe one of the dialogues that was between the
“Sydney School” of SFL (Rose and Martin, 2012) and Karl Maton’s version of the developments
on Bernstein’s foundations, and detailed studies are provided by Martin, Maton and Doran
(2020a). The overarching theme of their book is “teaching science”, and it deals with different
aspects of scientific knowledge (i.e. ideational meanings located within the field parameter of
context) and of the role the designed semiotic of mathematics plays. The studies are under-
pinned by the combination of SFL and Legitimation Code Theory.

Figure 4.18 Types of pedagogy – as system of pedagogic options (typology) and as a pedagogic stance
(topology, adapted from Martin, 1998: 419)

245
 Application guides

4.5.3 Knowledge in education

Systemic functional contributions within educational linguistics cover many facets of education –
contextually ranging across field, tenor and mode, metafunctionally across ideational, inter-
personal and textual meanings, and in terms of the cline of instantiation, they extend from the
meaning potential in the institution of education (and other institutions as well where teaching
and learning take place, like the family) via the registers of particular subjects and disciplines to
particular texts in their contexts of situation.
One of the many topics that has been illuminated by systemic functional work in educa-
tional linguistics is how learners construe knowledge in different domains of experience,
building up the organization as they engage with different texts. “Knowledge” is actually
ideational meaning embedded within the field parameter of context relative to different
tenor relations between learners and the people they learn from, like parents, older siblings
and teachers (Halliday and Matthiessen, 1999/2006). Learners begin to construe knowledge
through language in early childhood before formal education (Painter, 1999), when they
construe folk models of the world – commonsense knowledge, typically in spoken dialogic
interaction.
As they enter institutions of formal education, they continue to construe knowledge, first
in school subjects and later in university disciplines. Their commonsense knowledge is gradu-
ally replaced – or rather complemented – by educational versions of scientific, uncommonsense
knowledge (see e.g. Wignell, Martin and Eggins, 1989) and later, in late secondary and tertiary
institutions of education by straight scientific, uncommonsense knowledge. This involves a meta-
phorical reconstrual of what they learned during their pre-school and early primary school years
(e.g. Derewianka, 1995, 2003). They continue to learn through spoken language, but also begin
to learn through written language – the medium of instruction being either their mother tongue
or a “new” language of the educational system or a combination of the two, and increasingly they
also learn through other semiotic systems, like branches of mathematics and designed visual repre-
sentations such as maps, taxonomies and flowcharts (e.g. Mohan, 1986). As learners move through
the educational system, the knowledge they have to construe becomes increasingly specialized into
different fields of expertise.
As we noted earlier, knowledge is actually ideational meaning: it constitutes our cumulative
collective process of construing our experience of the world around us and inside us. It is thus
comprehensive in the sense that it embodies the totality of our collective experience – all expe-
rience is construed as ideational meaning. At the same time, it is distributed across the meaners
of a meaning group, the individual “knowers” – all of whom are trustees (or “stake-holders”
in the collective ideational meaning potential). No single person knows the whole body of
“knowledge”, as Malinowski (1922) pointed out a century ago in relation to the collective
knowledge of the kula exchange system in the Trobriand Islands (p. 64):

Yet it must be remembered that what appears to us an extensive, complicated,

and yet well ordered institution is the outcome of ever so many doings and pur-
suits, carried on by savages, who have no laws or aims or charters definitely laid
down. They have no knowledge of the total outline of any of their social structure.
They know their own motives, know the purpose of individual actions and the
rules which apply to them, but how, out of these, the whole collective institution
shapes, this is beyond their mental range. Not even the most intelligent native has
any clear idea of the Kula as a big, organised social construction, still less of its

246
 Application guides

sociological function and implications. If you were to ask him what the Kula is,
he would answer by giving a few details, most likely by giving his personal experi-
ences and subjective views on the Kula, but nothing approaching the definition
just given here. Not even a partial coherent account could be obtained. For the
integral picture does not exist in his mind; he is in it, and cannot see the whole
from the outside.

Ideational models of domains of experience (“knowledge”) are thus distributed across the
persons who serve together in complementary roles within the relevant tenor role network.
Learners are apprenticed into such networks. At the same time, ideational models are richly
diversified across diferent fields of experience, e.g. classified educationally and academically
into subjects and disciplines (cf. Bernstein’s, 2000, distinction between vertical and horizon-
tal knowledge structures, reviewed and revised by Martin, Maton and Doran, 2020a) and
diferentially deployed within diferent fields of activity. In accounts of branches of science,
fields of experience treated scientifically are usually viewed in terms of scale – from micro via
meso to macro. This works up to a point for the sciences of material phenomena (in physical
and biological systems), e.g. quantum physics to astrophysics. However, to include immaterial
phenomena, we need to introduce another dimension, one that complements the dimension
of (material) scale. This is the dimension of abstraction rather than size – of the ordering of
systems operating in phenomenal realms of increasing complexity (see Section 1.3.2): see
Figure 4.19.
The fields of experience differentiated in a highly schematic way in Figure 4.19 combine
with different fields of activity (expounding, reporting, recreating, sharing, doing, enabling, rec-
ommending, exploring). In principle, all possible combinations could be expected to occur in
school subjects and university disciplines. However, research has shown that there is considerable

Figure 4.19 Domains of experience construed in science, differentiated according to scale and abstraction
Source: Left part of figure, from Wikipedia https://upload.wikimedia.org/wikipedia/commons/7/75/
The_Scientific_Universe.png

247
 Application guides

variation across subjects and disciplines. In certain subjects and disciplines, the most highly val-
ued field of activity is to ‘expound’ phenomena, characterizing and explaining them, whereas in
other subjects and disciplines, the most highly valued field of activity is to ‘explore’ alternative
accounts of phenomena, arguing in favour of one over other alternatives or simply enacting
value by authority.
Systemic functional investigations of knowledge in education covers both field-based knowl-
edge structures and the registers (genres) where knowledge is construed, transmitted and negoti-
ated (e.g. Mohan, 1986; Wignell, Martin and Eggins, 1989; Unsworth, 1995; O’Halloran, 2005;
Coffin, 2006; Wignell, 2007a; Christie and Derewianka, 2008; Christie and Martin, 2007;
Heuboeck, Holmes and Nesi, 2010; Zhao, 2011; Doran, 2016; Ferreira and Zappa-Hollman,
2019; Martin, Maton and Doran, 2020a; Martin, 2020; Unsworth et al., 2022). Mohan (1986)
is a pioneering study showing the pedagogic value of representing knowledge domains by means
of different semiotic systems (e.g. verbal text and flow charts or taxonomies), and of teaching
learners to translate among them.
In the work on representing knowledge, educational SFL overlaps with computational SFL,
as outlined by Matthiessen in Matthiessen et al. (2022: 139) in response to a question about the
influence of computational SFL on SFL in general:

One very interesting candidate [for productive collaboration between educational

and computational SFL] is clearly subject or disciplinary knowledge – field-classified
domains within the ideation base [Halliday and Matthiessen, 1999/2006]. In edu-
cational linguistics, researchers have contributed a fascinating range of studies of
“domain knowledge”, e.g. in geography, history and physics; and in computational
SFL, researchers have modelled domain knowledge quite explicitly, also including
multimodal considerations. So there is a considerable potential for productive dia-
logue; I can envisage very exciting collaboration in this area. I think it could really
be of tremendous benefit to our 21st century students. I dream of workbenches for
students from primary school, or at least secondary school, onwards informed both
by educational SFL and computational SFL. This kind of learning environment could
really constitute a quantum leap – and it could include the insights from the research
that Mohan (1986) reports on.

4.5.4 Options in educational linguistics

Educational linguistics covers an extensive number of research activities and applications; many
areas have been added since the work in the 1960s on literacy. We can characterize the options
that have been taken or can be taken in terms of three simultaneous primary systems‚ the role
of language in educational processes, the site of research or application and the selection of
learners (and teachers). These systems are set out in Figure 4.20, with additional more delicate
systems, and studies taking up the different options are exemplified in Table 4.14. The systems
are characterized briefly here:

role of language: this system captures Halliday’s well-known characterization of the role
of language in teaching-learning processes: ‘teaching-learning language’/‘teaching-
learning through language’/‘teaching-learning about language’. (1) The term ‘teaching-
learning language’ means that language is the object; here learners are learning how
to mean, either for the first time through the phases leading up to and including the
mother tongue, or learning how to mean in a second or foreign language. (2) The term

248
 Application guides

‘teaching-learning through language’ means that language is a medium for teaching-

learning but it extends far beyond the notion of language as the “medium of instruction”
since it also involves the construction of “knowledge” in language, i.e. knowledge is con-
strued as ideational meaning (cf. Halliday and Matthiessen, 1999/2006). Painter’s (1999)
account demonstrates very clearly how as children expand their linguistic potentials,
they can construe new kinds of “knowledge”. (3) The term ‘teaching-learning about
language’ means that knowledge about language (“KAL”) is the object, an option that has
been discussed in the context both of the mother tongue, e.g. Williams (2004, 2005a),
Macken-Horarik (2012), Williams and French (2016), and second/foreign languages,
e.g. Teruya (2006, 2009) for Japanese and Caffarel (2006a) for French.
site: this system is concerned with the choice of research and application; we have differen-
tiated ‘educational policy’ and ‘teaching-&-learning’. (1) The term ‘educational policy’
refers to research and application concerned with education – as institution and as pro-
cesses of teaching-&-learning – located within institutional sites where policy about edu-
cation is formulated or executed, like departments of education and education bureaus.
Such policy may of course also interact with language policy, e.g. regarding official
languages, languages of instruction – cf. Yang and Wang (2017). (2) The term ‘teaching-
&-learning’ covers different institutional settings, not only those of formal education but
also other institutions such as those of family and friendship, of recreation, of workplaces
(e.g. in service teaching-&-learning in institutions of healthcare: Woodward-Kron et al.,
2011; Slade et al., 2015).
selection: this system is concerned with the principle of selection of the domain of
research and application in terms of the people involved (learners and also teachers,
whether professional ones or members of other institutions such as the family). (1) The
term ‘cross-sectional’ has been the unmarked one in a great deal of work outside SFL,
including cross-sectional studies of “language acquisition”. However, there are few if any
in SFL because the conception of language learning as learning how to mean favours
longitudinal case studies. (2) The term ‘longitudinal’ has characterized a growing num-
ber of systemic functional studies since Halliday (1975a), including both early childhood
studies and studies focussed on learning in primary and secondary schools (e.g. Derewi-
anka, 1995, 2003). Encouragingly, the value of longitudinal studies (as opposed to cross-
sectional ones) has been recognized more generally (see e.g. Ortega and Byrnes, 2008);
it can be argued that a number of scholars have recognized that all the cross-sectional
studies undertaken of learners learning a second/foreign language actually tell us very
little about learning how to mean in a second/foreign language. Longitudinal studies
include quasi-longitudinal ones, where samples of learners of different ages are taken as
representative of developmental paths (e.g. Quam, 2020; Ryshina-Pankova, 2011; Chris-
tie and Derewianka, 2008; Xuan and Matthiessen, forthcoming).

As indicated earlier, certain combinations of options are more likely than others; for example, all
or virtually all systemic functional studies of learning language have been longitudinal rather than
cross-sectional. There are, of course, relevant considerations that are not captured in Figure 4.20,
including the approach to data in educational studies, e.g. sampling periods in longitudinal stud-
ies and criteria used in the compilation of corpora of educational materials (e.g. Painter, 2000;
Torr, 2015; de Silva Joyce and Feez, 2016), and also approaches to language assessment (e.g. Mat-
thiessen, Slade and Macken, 1992; Gardner, 2010; Byrnes, Hiram and Norris, 2010; Shrestha,
2011; Shrestha and Cofn, 2012). Some key considerations are, obviously, the same as in work
based on text analysis in general such as the selection of semiotic systems considered in studies.

249
 Application guides

Figure 4.20 Choices open to educational linguists

Unsworth (2000) includes a number of contributions on how to carry out language-based

research in schools, but also in the community outside schools. In addition, there are several
sources providing guidance in research in educational linguistics and studies of ontogenesis,
including:

• Literacy studies: de Silva Joyce and Feez (2016) devote a chapter to methodology in
literacy research – which they characterize as “a methodological map” (Chapter 6). They
cover methods that are also used elsewhere in educational linguistics: action research, case
studies, text analysis (of written and spoken text, including “classroom discourse” and also
multimodal analysis), ethnographic research, verbal protocols and meta-analysis, educa-
tional contexts.
• First language development studies: Painter (2000) presents a detailed account of
considerations involved in researching early childhood language development, including
data collection (pp. 66–71), where she contrasts the preferred method in SFL studies,
naturalistic observation and recording in case studies undertaken by participant-observers,
with “experimental studies” and “semi-structured observations”, and discusses the chal-
lenge of dealing with time in longitudinal studies. She also deals with transcription and
the annotation or coding of naturally occurring texts (pp. 71–75). Importantly, she dis-
cusses how to frame developmental studies, including the issue of research questions advice
about research strategies based on how to frame research contextually, semantically and/or
lexicogrammatically. (pp. 75–82). Torr (2015: 248–250) offers a concise overview of data
collection and method of analysis, summarizing and contextualizing the approach taken by
Halliday (1975a).

To round of our review of guiding principles in work on language development and edu-
cational linguistics, we would like to draw attention to the benefit of exploring educational

250
 Table 4.14 Examples of educational linguistic studies

System Terms Further delicacy References

role of learning/teaching mother tongue case studies: Halliday (1975a, 2003a); Painter (1984); Qiu (1985); Torr
language language (1997, 1998); Hoyt, Torr and Degotardi (2014)
overviews (including methodology): Painter (2000); Torr (2015)
Halliday in relation to Vygotsky and Bernstein (also L2): Foley (1991)
second/foreign Hasan and Perret (1994)
language EFL: Guo (2014), Xuan (2014)
methodology: Perrett (2000)
curriculum and syllabus: Feez (1998); Byrnes, Maxim and Norris (2010)
learning/teaching grammatics: Williams (2004, 2005a); Macken-Horarik (2012); Williams
about language and French (2016)

Application guides
knowledge about language (KAL) in general, broadly conceived: Thwaite
(2015)
251

learning/teaching pre-school: Painter (1999);

through language the language of schooling: Schleppegrell (2004)
genres (registers in context) in primary and secondary school: Christie
and Martin (1997); Martin and Veel (1998); Rose and Martin (2012);
Christie and Derewianka (2008) – history: Coffin (2006); mathematics
– O’Halloran (2005) (multimodal analysis)
site educational policy Yang and Wang (2017)
teaching and educational setting classroom: Sinclair and Coulthard (1975); Lemke (1985, 1990); Christie
learning institutions (2002); Gibbons (2007); Thwaite and Rivalland (2009); Hammond
(2011); Amundrud (2017)
body language in face-to-face teaching: Hood (2011)
materials books about animals for pre-school children: Koutsikou and Christidou
(2019) (multimodal analysis)
textbooks: Schleppegrell (2004: 139–142)
ESL textbooks: Guo (2014) (multimodal analysis); Guo and Yao (2021)
(Conitnued)
 Table 4.14 (Continued)

System Terms Further delicacy References

teacher Gebhard (2019)
education
level primary overview of all aspects: Christie (2005)
literacy: [breakthrough to literacy:] Mackay, Thompson and Schaub
(1970); [literacy at various levels:] Halliday (1991d); Humphrey (1996)
[in school geography]; Christie and Misson (1998); Thwaite (2007)
critical school literacy: Macken-Horarik (1998)
reading: Goodman, Fries and Strauss (2016)
secondary Unsworth (1995, 1999); Veel (1998); O’Halloran (2005); Coffin (2006);
Doran (2016)

Application guides
primary and Schleppegrell (2004); Christie and Derewianka (2008), Christie (2012)
secondary
252

tertiary academic discourse: Coffin et al. (2003); Gardner and Nesi (2013);
Martin, Maton and Doran (2020a); social science: Wignell (2007a,
2007b); biology: Hao (2020); Ferreira and Zappa-Hollman (2019)
other institutions case study family and Painter (1999)
friendship
workplace Woodward-Kron et al. (2011); Slade et al. (2015)
selection longitudinal institution: family case study Halliday (1975a), Painter (1984, 1999), Torr (1997);
institution: formal case study Derewianka (1995, 2003)
education
group, Christie and Derewianka (2008); Christie (2012)
materials Xuan (2014); Guo (2014)
cross-sectional (few if any cross-sectional SFL studies)
 Application guides

Figure 4.21 Educational contexts – the meta-context of learning/teaching how to mean in a second/
foreign language characterized in terms of field, tenor and mode

contexts in terms of the field, tenor and mode of educational processes. As an illustration, we
gloss the context of learning/teaching a second/foreign language: see Figure 4.21. The figure
shows that the context of learning/teaching as educational processes are meta- in relation to
the contexts in which learners learn how to mean in a second language, e.g. learning how to
share personal experiences and opinions in casual conversation, how to transact commodities in
a service encounter. The meta-context includes features that have become fashionable outside
SFL such as learner motivation.

4.6 Translation studies

4.6.1 Phases of development of SFL-informed studies of translation

Like education, translation was an early focus for systemic functional research and application;
in fact, the first contribution to appear was Halliday’s (1956a) proposal for the use of a thesaurus
in machine translation. This paper was, of course, pre-SFL, but it was systemic in orientation –
a thesaurus being a systemic representation of the lexical resources of a language12 in contrast
with the item-based dictionary view of lexis (cf. Matthiessen, 1991a, 1995a). In the 1960s,
there were a number of further systemic contributions to translation studies – notably, Catford’s
(1965) quite influential book presenting a systemic linguistic theory of translation, and Halliday,
McIntosh and Strevens (1964) devote one chapter to translation, Chapter 5 “Comparison and
translation”; we can also note the location of the study of translation in terms of Ellis’ (1966)
conception of general comparative linguistics. These early contributions were based on the
“scale-&-category” theory stage of SFL (essentially the theory presented by Berry, 1975/1977),
and approached translation using rank, system and other scales categories, also foregrounding
the importance of varieties of language. Importantly, they maintained a connection between the
comparison of systems and the translation of texts (cf. Matthiessen, Teruya and Wu, 2008, on
multilingual studies in general). They included the well-known demonstration that the potential
for effective translation between two languages increases as translators move up the rank scale
from morphemes or words to clauses, and complexes of clauses.

253
 Application guides

Having appeared by the mid-1960s, these contributions did not include later developments
in the decade – centrally, Halliday’s theory of metafunction, and beyond, like the development
of accounts of semantics and of grammatical metaphor and of context. Matthiessen (2001) was
written as an attempt to “update” Catford (1965) and other contributions from that period in the
light of the theoretical developments since the mid-’60s. (By the turn of the century, research into
translation and applications of findings had increasingly become institutionalized as a separate dis-
cipline, “translation studies” – sometimes with “descriptive” added as a pre-modifier; the linguis-
tic conception of that area of scholarship sketched by Catford and other scholars in the 1960s with
the connection to linguistics in general and to other multilingual activities within linguistics such
as comparison and typology [as generalized comparison] had been attenuated or even lost. Hence
the call for a unified framework of “multilingual studies” in Matthiessen, Teruya and Wu, 2008.)
As far as publications are concerned, the period from the second half of the 1960s to the sec-
ond half of the 1980s was a quiet one in terms of SFL and translation: the additions to the theory
that turned it into systemic functional theory had to be developed, consolidated and metabolized,
and the compilation of the cumulative descriptions of the lexicogrammar and then the semantics
of English had to be given enough time (meanwhile, the second round of descriptive work on
Chinese had been initiated, and new languages came into the descriptive picture: see Table 3.11).
Then, in the second half of the 1980s, the EUROTRA machine translation project was
undertaken with a systemic component thanks to Erich Steiner (e.g. Steiner, 1986; Steiner and
Winter-Thielen, 1988; Steiner et al., 2022), and researchers began to draw on the comprehen-
sive “thumb nail” description of the lexicogrammar of English presented by Halliday (1985b),
either directly or as a model of investigations of other languages. The value of SFL in translation
studies was brought out by a number of publications in the 1990s, notably Baker (1992), who
drew mainly on the textual metafunction in her widely used textbook, Taylor (1993), who
provided an SFL-informed guide for translators moving between Italian and English, and House
(1997), who contributed an updated version of her model for translation quality assessment
(House, 1977), taking the contextual parameters of field, tenor and mode and their linguistic
patterns of realization into consideration (see also House, 2015).
The beginning of the new century was marked by the publication of a number of contribu-
tions to SFL-informed translation studies in a volume edited by Erich Steiner and Colin Yallop
(Steiner and Yallop, 2001). It was based on a colloquium that they had organized as part of ISFC
1998, where they had brought together a number of SFL and SFL-informed scholars to take
stock of translation and multilingual studies. The book includes a very valuable chapter by Yallop
(2001) on “translation equivalence” based on a study of an adaptation and translation of Alice in
Wonderland into Pitjantjatjara. While many translation studies involve languages that are linguis-
tically and culturally within the same realm of standard languages spoken by large communities
in modern nation states, Yallop’s study explores issues arising when “source” and “target” are
much further apart linguistically and culturally (in the non-SFL literature, cf. Grace, 1981). In a
way, Kunz et al. (2014) is a sequel to Steiner and Yallop (2001); it is a Festschrift for Erich Steiner
on the occasion of his 60th birthday, and it includes a number of contributions to the study of
translation. Kim et al. (2021) bring together a number of recent studies of translation informed
by SFL, and Wang and Ma (2022) include interviews with systemic functional scholars who have
made leading contributions to the study of translation.

4.6.2 Overview of SFL-informed studies of translation

In the last two decades, SFL-based and SFL-informed have not only grown in number and range
of languages, but also expanded into investigations of additional aspects of translation, using new

254
 Application guides

research methods: see Table 4.15. The table is not exhaustive – it does not include many studies
published in languages other than English – but it gives an indication of richness of the field, and
it includes comparative studies based on comparable texts.
Introductions to translation and translation studies including material on SFL or couched in
terms of SFL have been growing steadily in number since the early 1990s: Bell (1991), Baker (1992,
2018), Malmkjaer (2005), Munday (2016), House (2016), Kunz and Teich (2017), Hill-Madsen
(2020); Wang and Ma (2021, 2022), Kim et al. (2021); and Munday (2016) also covers SFL among
other contributions to translation studies. Over this period, groups and centres developing systemic
functional translation studies have emerged in a number of different places around the world, includ-
ing China, South Korea, Indonesia, India, Iran, Australia, Germany and the UK, and Brazil (see
Matthiessen, Arús and Teruya, 2021); Vasconcellos (2009) provides an overview of work in Brazil.

Table 4.15 Systemic functionally informed work on translation since 2000 (including some contributions
to the study of comparable texts)

Focus Types Publications

tools for translators Taylor and Baldry (2001)

translation in terms of textual Baker (1992, 2018: Chapters 5 and 6),
phenomena metafunction Munday (2000), Kim (2007a, 2007c),
Espindola (2010, 2016); Wang (2014);
Kim and Matthiessen (2017)
interpersonal Wang (2004), Munday (2012); Espindola
and Wang (2015); Yu and Wu (2016,
2017)
experiential Mason (2003)
Choi (2013)
in terms of process Alves et al. (2010); Serbina, Niemietz
instantiation and Neumann (2015); cf. also Munday
(2012) on the translator’s decision-
making process
probabilities in Jesus and Pagano (2006); cf. also outside
translation SFL Toury (2004) drawing on Halliday’s
probabilistic conception of language
variation: register Teich (1999b, 2013); Hansen and Teich
(1999); Lavid (2000); Murcia-Bielsa
(2000); Hansen, Klaumann and
Neumann (2002); Neumann (2003a,
2003b); Steiner (2004); Neumann
(2013, 2014); Hansen-Schirra,
Neumann and Steiner (2012); control
language: Hartley and Paris (1997)
particular registers [detective stories:] Wang Yan (2015);
[drama:] Wang (2017), Wang and Ma
(2021); [poetry:] Ma and Wang (2021c)
in terms of axis paradigmatic Halliday (2012), Matthiessen (2014b)
(choice)
(Continued)

255
 Application guides

Table 4.15 (Continued)

Focus Types Publications

in terms of mode subtitling Taylor (2003); Baldry and Taylor (2002);

Espindola and Vasconcellos (2006);
Espindola (2010, 2012, 2016); Matielo,
Vasconcellos and Espindola Baldissera
(2015)
audiovisual Taylor (2013, 2014); audio description:
Taylor (2014)
in terms of explicitation, Teich (2003); Hansen (2003); Neumann
properties of simplification, and Hansen-Schirra (2005); Hansen-
translated texts normalization, Schirra, Neumann and Steiner (2006,
and texts to be source language 2012); Wang and Tang (2014); cf. also
translated shining through Kunz et al. (2021)
evaluation/ House (1977, 1997, 2001a, 2001b, 2015);
assessment Vasconcellos (2004); Kim (2007b);
(quality) Steiner (1996)
control language Hartley and Paris (1997)
methods for corpus methods Baldry and Taylor (2002), Teich (2003),
studying Hansen-Schirra and Neumann (2003),
translation Neumann and Hansen-Schirra (2005),
phenomena Neumann (2013); Hansen-Schirra,
Neumann and Steiner (2012); Kunz
and Teich (2017); Kunz et al. (2021);
Hasselgård (forthcoming); and cf. also
Johansson (2007)
in combination Lavid, Arús and Moratón (2009); cf. also
with Davidse and Heyvaert (2004)
comparison
of comparable
original texts
keystroke logging Alves et al. (2010); Serbina, Niemietz and
and eye tracking Neumann (2015)
brain imaging cf. García (2019)

4.6.3 Research methodology adopted in SFL-informed studies of translation

The research methodology established already during the early period of SFL-informed research
into translation is that of systemic text analysis, i.e. analysis of source and target text or texts
referencing systemic descriptions of the languages involved. In terms of the options in research
methodology set out in Figure 4.4 earlier, the considerations have been as follows:

qual/quant: studies have been qualitative in nature, but have been supplemented by quan-
titative information – counts of selections of terms in the systems in focus, and also
increasingly with the growing application of corpus linguistic tools and techniques,
counts of patterns detected through automatic analysis (e.g. Hansen-Schirra, Neumann
and Steiner, 2012; Kunz et al., 2021).

256
 Application guides

naturalness: studies have been largely natural (non-experimental) in orientation, based on

naturally occurring source and target texts. But more experimental studies will arguably
be necessary as research into translation continues to focus also on translation processes
and on observations “from below”, using observational techniques such as keystroke log-
ging, eye tracking (Alves et al., 2010; Serbina, Niemietz and Neumann, 2015) and brain
imaging (cf. García, 2019).
authenticity: studies have been based largely on authentic material. The possibility exists,
of course, of “eliciting” translations by setting up a project where one or more people
are given a translation task. These people may be professional translators, translators in
training, multilingual experts on the field of discourse (domain of experience) central to
the translation task.
order of data: studies have been based on primary data, i.e. on text in context – translated
texts, often also corresponding source texts. Here the methodological issues inherent in
the selection of texts to analyse are central, and Hansen-Schirra, Neumann and Steiner
(2012) address and model the compilation of multilingual corpora for translation research.
But to the extent that findings based on the analysis of primary data are produced and
presented in comparable formats, the potential for the effective use of secondary data is
considerable, especially in view of the fact that translation studies based on primary data
tend to involve just a small number of translation pairs, and descriptive generalizations
must be based on a much more extensive database. (For a research synthesis overview of
SFL translation studies, see Chen, Xuan and Yu, 2022.)
automation: studies have been either manual or automated, and some have included both
methods. Early studies were manual – manual identification of systemic and functional
patterns in descriptions of languages, and this tradition has continued (see e.g. Wang and
Ma, 2021) because manual analysis can reach “higher” than automated analysis in terms
of level of analysis (i.e. strata, and rank within lexicogrammar). This reflects the gen-
eral considerations concerning methodological trade-offs in text analysis, as represented
diagrammatically in Figure 4.15. At the same time, an increasing number of studies are
based on large corpora of parallel (or comparable) texts analysed automatically by means
of computational tools (e.g. Teich, 2003; Hansen-Schirra, Neumann and Steiner, 2012;
Kunz and Teich, 2017; cf. also Kunz et al., 2021).
direction: studies have largely been in the direction of analysis, although they have been
supplemented by some work relevant to the understanding of translation in machine
translation and multilingual text generation (e.g. Steiner, 1992; Bateman, 1997; Bate-
man, Matthiessen and Zeng, 1999; Arús and Lavid, 2001).

As the preceding comments suggest, diferent options can productively be combined in research
investigating complementary aspects of translation. For example, Alves (2003) provides a col-
lection of studies illustrating the value of triangulation in translation research, which is highly
relevant to the multiperspectival approaches to the investigation of translation in all its manifes-
tations made possible by SFL (cf. Alves et al., 2010) – which of course invites interdisciplinary
collaboration (e.g. Hansen, 2003).
One consideration not included in Figure 4.4 is the choice of language (or other semiotic
system) as data. Past studies in SFL have tended to be concerned with pairs of “major” languages
and of pairs commonly involved in translation (an important exception being Yallop, 2001), as is
generally the case in translation studies. Therefore there is a serious need to expand the sample of
language pairs based on various considerations, including the need to include “minor” languages
and also the need to take typological consideration into account (cf. Grace, 1981). In addition,

257
 Application guides

there are interesting issues to consider in relation to the translation of varieties – dialectal, regis-
terial and codal (cf. outside SFL, Weber and Mann, 1981, on dialect adaptation).
At the same time, while research methodology has been fine-tuned for the investigation of
language pairs, there are clearly research questions of fundamental importance having to do with
translation into a wider range of languages, as happens in the case of highly valued texts, includ-
ing ones consider sacred, of literary value or simply of popular commercial value. This approach
can also be used in the service of text-based language comparison and typology, as in investiga-
tions of the lexicogrammar of motion (cf. Matthiessen, Arús and Teruya, 2021).

4.7 Institution of medicine (healthcare)

The institution of medicine, or of healthcare, is a crucial aspect of modern nation states; activities of
care and healing that may have been part of other institutions in traditional societies have become a
centre of clinics, hospitals, care homes and other manifestations of the institution of healthcare with
intricate networks of professional roles. This institution has increasingly become a site for linguistic
research and application, as have institutions that it forms part of an interactive network with, e.g.
the institutions of the family, of friendship, of the workplace and of administrative and governing
bodies such as departments of health. For example, from the point of view of persons who play
institutional roles as patients, their roles in other institutions that constitute their “lifeworld” are
crucial in a holistic framing of illness, disease and well-being.

4.7.1 Clinical linguistics and healthcare communication studies

Thus research and application concerned with the institution of healthcare often involve other
institutions as well, at least by implication. In systemic functional work, as in work located
within other traditions of linguistics, and also discourse analysis, there are two concerns that
are in principle distinct: work focussed (i) on language as phenomenon – as part of the clinical
picture of signs and symptoms, and (ii) on language as medium – as a resource for healthcare
practitioners, patients and also their family members or friends (cf. Matthiessen, 2013).
(i) The concern with language as phenomenon was the first to develop, and is often known as
clinical linguistics (for a recent overview, see Asp and de Villiers, 2019). (ii) The concern with
language as a resource in the practice of healthcare has gained attention more recently, although
Mishler (1984) was an early forerunner (cf. also Mishler, 1986) using cohesive analysis in a study
of medical interviews. This is usually known as healthcare communication studies. Contri-
butions to this area are the focus of a recent survey, Moore (2019), also with information relevant
to clinical linguistics. These two concerns shade into one another under certain circumstances,
importantly in psychotherapy where language is both phenomenon and resource.
Both concerns very often involve teams whose members have complementary areas of expertise
and experience. The complementarity may be purely academic in nature, with members from
different disciplinary areas such as linguistics and medicine, but it often also involves teams with
academics and healthcare professionals and possibly also patients. Both areas should provide good
opportunities in the future for healthcare professionals with a research degree in linguistics; Kealley’s
(2007) systemic functional linguistic study of nursing depends on this kind of background, and in
this respect the situation is similar to the development of educational linguistics.

4.7.2 Clinical linguistics: language as clinical phenomenon

In clinical linguistics, language is in one way or another an aspect of what ails patients, and
this has become an important area of research and application in SFL (e.g. Halliday, 2005d;
258
 Application guides

Armstrong et al., 2005; Fine, 2006; Armstrong, 2009; Ferguson and Armstrong, 2009; Asp
and de Villiers, 2010, 2019; and also Moore, 2019). The term “clinical linguistics” originated
with Crystal (1981), in the context of British linguistics. He starts his book, on p. 1, with an
initial definition of the term: “Clinical linguistics is the application of linguistic science to
the study of communication disability, as encountered in clinical situations”; then he contin-
ues immediately with the following observation: “Unfortunately, almost every term in this
definition requires further discussion, in order to identify the orientation and scope of the
subject.” However, in the 40 years that have passed since he wrote the book, the field has
become well-established, and the demands made on linguistics (including phonetics) have
become quite clear.
In their overview of systemic functional contributions to clinical linguistics, Asp and de Vil-
liers (2019: 590) state the appliability of SFL to clinical linguistic task succinctly:

The calls from Crystal [viz. (2002, 2013)] – for clinical linguists to engage in more
systematic descriptions of discourse, to take on semantics, grammar, and lexical analyses
more extensively, and especially to consider the choices speakers make in discourse as
indices for understanding ‘why’ speakers exhibit specific patterns – suggest that SFL
models of language and descriptive techniques should provide excellent frameworks and
a toolkit for clinical linguistics. SFL’s functional grounding, its focus on text/discourse
as the object of linguistics, and its organization of grammars as system networks that
characterize semiotic potential as choices that may be probabilistically weighted for
demographic, contextual, and textual factors recommend it for clinical work.

Their overview of systemic functional contributions to clinical linguistics shows that these con-
tributions have drawn on comprehensive systemic functional descriptions in the first instance,
being less focussed on theoretical modelling. Based largely on their overview, we have designed
Table 4.16 using their clinical categories as row headings and locating studies according to the
metafunctional “home” of the systems they have been concerned with. We have added informa-
tion about “mental” health in the area of psychotherapy.
As can be seen from Table 4.16, the system of cohesion within the textual metafunction has
been the target of the largest number of studies, some of which have dealt with cohesive har-
mony (as developed by Hasan, 1984b; see e.g. Taboada, 2019: 320–321), and the interpersonal
metafunction has also provided resources for a number of studies, usually the semantic system of
speech function. A number of studies have also dealt with contextual structure, taking account
of Hasan’s (e.g. 1978, 1984a), Generic Structure Potential (GSP).
Research into language disorders is, naturally, multidisciplinary in nature, drawing on both
different areas of disciplinary expertise and disciplinary methodology. Thus language disorders
can be viewed in terms of the ordered typology of systems (see Section 1.3.2); they can be
observed, recorded and analysed biologically, socially and semiotically.
Focussing on the semiotic order of systems, 4th-order systems, we can now draw on
a number of guides or handbooks that are informed by SFL, dealing with language as a
phenomenon in clinical contexts:

• Fine (2006), Language in Psychiatry: A Handbook of Clinical Practice, deals specifically with the
field of psychiatry. It addresses clinicians (in training) and is designed to help with diagnosis
and treatment but also with research. Thus, the view that might be adopted by clinicians
is foregrounded: prominent in the organization of the book are chapters dealing with dif-
ferent kinds of disorder (e.g. attention deficit hyperactivity disorders, psychotic disorders,
mood disorders, personality disorders).
259
 Table 4.16 Clinical linguistic studies in SFL – analytical focus
Domain Ideational Interpersonal textual various
Logical Experiential theme, information cohesion
psychotherapy narrative Muntigl (2004a,
therapy 2004b, 2007);
Muntigl et al.
(2013)
conversation Henderson-Brooks
model (2006); Butt,
Moore and
Henderson-Brooks
(2012); Butt, Fahey
and Henderson-
Brooks (2003)

Application guides
depression Nagar and
Fine (2013):
260

clause
complexes
schizophrenia Cohen (2011): Rochester and Martin
speech function, (1979)
appraisal
neurodevelopmental Autism Fine et al. Baltaxe and D’Angiola
disorders Spectrum (1991): (1992); Fine et al.
information (1991)
structure and
intonation
Specific Thomson Adams and Bishop
Language (2005): theme (1989)
Impairment
Attention Mathers Tannock et al. Tannock et al. (1995) Mathers (2006)
Deficit (2006): (1995): speech
Hyperactivity clause function
Disorder complexes
 neurocognitive Alzheimer’s Mortensen Müller and Wilson Lock and Armstrong
disorders Disease (1992): (2008): speech (1997); Ripich,
transitivity function; Müller Carpenter, and Ziol
and Mok (2012); (2000); Dijkstra et al.
Mok and Müller (2004)
(2013)
Acquired Brain Armstrong Togher (2001); Coelho et al. (1991a,
Injury – (2005); Togher et al. 1991b, 1994); Davis
Traumatic Rigaudeau- (1997a, 1997b, and Coelho (2004);
Brain Injury McKenna 2004, 2009, Mentis and Prut-
(2005) 2013); Togher ting (1987); Hartley
and Hand and Jensen (1991);
(1998); Guo and Jorgensen and Togher
Togher (2008); (2009); Liles et al.
Jorgensen and (1989); Coelho (2002);
Togher (2009) McDonald (1993)

Application guides
Acquired Brain Armstrong Ferguson (1992): Piehler and Holland
Injury – (2001, speech function; (1984); Glosser and
261

Aphasia 2000, Armstrong Deser (1990); Coelho

2005a); and Ulatowska et al. (1994); Lock and
Armstrong, (2007); Armstrong (1997);
Fox and Armstrong et al. Armstrong (1987,
Wilkinson (2011, 2012); 2000); Olnes and
(2013); Armstrong Ulatowska (2011);
Mortensen (2005b); Andreetta, Cantagallo
(2005) Armstrong, Fox and Marini (2012);
and Wilkinson Bottenberg, Lemme
(2013): appraisal and Hedberg (1985)
Childhood- transitivity modality theme Spencer et al. (2005)
Onset Fluency
Disorder
(Stuttering)
complexity modality theme Spencer et al. (2009)
Lee et al. (2015,
2016a, 2016b):
appraisal
 Application guides

• Ferguson and Armstrong (2009), Researching Communication Disorders, provide the back-
ground information and research methodology needed to do research focussed on com-
munication disorders, in both children and adults. Their book is informed by SFL and also
takes into account both psycholinguistic and sociolinguistic considerations. Their goal is
“to provide an overview of communication disorders for researchers interested in exploring
this area, providing both theoretical and practical perspectives on this endeavour”. One part
of the book is specifically concerned with the planning and execution of research projects.
The book also includes case studies that can be used in guiding future research.
• Asp and de Villiers (2010), When Language Breaks Down: Analysing Discourse in Clinical
Contexts, is concerned with “clinical discourse analysis”, “the term we use to describe the
analysis of language behaviour observed in clinical contexts”, showing how this analysis
can be used to diagnose problems (complementing other diagnostic tools) and to monitor
patients over time (detailed in Chapter 7). They provide background needed in the field of
language disorders, including ones associated with the autism spectrum and neurogenera-
tive disorders, and give a brief overview of different approaches to discourse analysis. They
describe key resources helpful in the analysis of discourse in clinical contexts, including
features of conversation covered in Conversation Analysis, intonation as described in SFL
by Halliday and others, grammar (covering interpersonal, experiential and textual systems,
based to a large extent on Michael Gregory’s account within SFL) and context (where they
also refer to work within AI to deal with frames, scripts, etc.). In terms of the approach to
discourse analysis, they introduce Michael Gregory’s phasal analysis, illustrated in detail for
different cases in their Chapter 9 (see also e.g. Gregory, 2002; Cloran, Stuart-Smith and
Young, 2007). They devote one chapter (Chapter 6) to “study design”, dealing with “nar-
rative and conversation tasks” in terms of “linguistic structure, narrative and memory”.

(As an important and relevant aside: Jonathan Fine was a remarkable systemic functional scholar.
He was a pioneer in empowering work in clinical settings through his extensive expertise in
SFL. While he contributed to other areas of scholarship from the 1970s to his premature death
in the 2010s, e.g. Cohen and Fine, 1978; Fine, 1978, 1985, 1988a, 1988b, most of his contribu-
tions illuminated the role of language as a phenomenon in clinical settings – language disorders
and disorders where language can serve as a gateway in diagnosis and research.)

4.7.3 Healthcare communication studies: language as healthcare resource

The study of communication in healthcare is relatively recent in SFL, as in other traditions as
disciplines, although one frequently cited study regarded as a pioneering contribution, Mishler
(1984), drew on the description of cohesion presented by Halliday and Hasan (1976). This line
of research is distinct from that of clinical linguistics, although there are clearly areas of overlap
such as psychotherapy and mental health generally where language may be both phenomenon
under investigation and medium of diagnosis and treatment. In contrast, in healthcare commu-
nication studies, the focus is on language (and also other semiotic systems like gesture, gaze, facial
expression and proxemics in face-to-face consultation) as a resource in healthcare – whether this
is in the delivery of healthcare by clinicians as professionals, in the processing of health issues
by a group of friends or family members or in public health campaigns. While the institutional
centre is the institution of medicine, or healthcare, other institutions are also involved. Studies
informed by SFL are exemplified in Table 4.17, and can serve as models for future expansions
of systemic functional research into healthcare communication. Pun et al. (2017) are concerned
specifically with the methodology of discourse analysis in healthcare communication studies.

262
 Table 4.17 Examples of systemic functional studies of communication in healthcare14

General description Field Tenor Mode Data Analysis References

medical recommending: doctor-patient spoken & dialogic text cohesion Mishler (1984)
consultation advising: consulting
GP recommending: doctor-patient spoken & dialogic text contextual structure Halliday (1998a)
advising: consulting
recommending: doctor-patient spoken & dialogic text context: tenor ~ Thompson (1999)
advising: consulting interpersonal: mood;
textual: cohesion
recommending: doctor-patient spoken & dialogic text multimodal analysis: gaze; Thwaite (2015)
advising: consulting intonation

Application guides
recommending: doctor-patient; spoken & dialogic text appraisal Pounds (2011)
advising: consulting empathy
263

recommending: clinicians-patient; spoken & dialogic text tenor: patient-initiated Eggins (2014)
advising: consulting empathy humour
family planning recommending: doctor-patient spoken & dialogic text & post- contextual (“generic”) Slade et al. (2009); de
advising: consulting consultation structure; tenor: Silva et al. (2015)
interviews empathy & rapport;
(doctors and interpersonal semantics:
patients) speech function
HIV recommending: doctor-patient spoken & dialogic text lexicogrammatical: Moore, Candlin and
advising: consulting cohesion, transitivity; Plum (2001)
implicature
recommending: doctor-patient spoken & dialogic multi-stratal; semiotic Moore (2004, 2005,
advising: consulting: agency 2021).
joint decision
making

(Continued)
 Table 4.17 (Continued)

General description Field Tenor Mode Data Analysis References

emergency recommending: doctor-patient spoken & dialogic text ethnographic observation Slade et al. (2008);
departments advising: consulting & interviews, context Matthiessen (2013);
notes, interpersonal Slade et al. (2015);
analysis Fung (2018)
nurse-patient spoken & dialogic text ethnographic observation Chandler et al. (2015)
&
interviews, context
notes, interpersonal
analysis
oncology recommending: oncologist-patient spoken & dialogic text context: contextual Karimi, Moore and

Application guides
advising: consulting system networks Lukin (2018)
recommending: oncologist-patient spoken & dialogic text context, semantics, Brown (2014)
advising: consulting lexicogrammar
264

reporting: chronicling/ patients spoken & interviews Jordens et al. (2001);

sharing: illness monologic Jordens (2002)
narratives
recommending: doctor-patient Lobb et al. (2006)
advising:
consulting, with
communication aid
patient journey recommending: doctor-patient; spoken & dialogic text, interviews, context + interpersonal Matthiessen (2013);
through advising: consulting nurse-patient ethnographic systems Slade et al. (2015)
emergency + supporting observation
department activities
clinical handover various kinds reporting clinician-clinician spoken/written text, interviews, Slade et al. (2015);
of handover, & dialogic/ ethnographic Eggins and Slade
including monologic observation (2016)
bedside
handover
 operating theatre doing: conducting doctors-nurses spoken text Cartmill et al. (2007)
surgery
nursing the language of varied: nursing nurse-colleagues, spoken text & dialogic & S. Candlin (2000);
nursing activities patients facilitative Kealley (2007);
Chandler et al.
(2015)
nurse-(ethnic spoken interviews discourse (CDA), Lassen and Strunck
minority)- identifying “positive (2011)
patient discourse” (Martin and
Rose, 2003)
patient blog the language of patients (as written text & dialogic McDonald (2016);
community patients as a community McDonald and
community members) Woodward-Kron
(2016)

Application guides
palliative care advice websites recommending: written text lexicogrammar: Driscoll (2012)
about terminal advising transitivity
265

illness
oncologist’s role recommending: oncologist-patient spoken text & dialogic context: contextual Karimi, Moore and
in relation advising networks Lukin (2018)
to patient
approaching
end of life
pamphlet: nurse-relatives of written text & monologic Kealley, Smith and
information patients Winser (2004)
for relatives
public health public health recommending: health authority- written + image posters & multimodal analysis, RST Zhang Peija (2018)
information posters in NY advising/promoting general public monologic analysis
and and HK
campaigns public health health authority- written text & monologic lexicogrammar: Cheung (2018)
information general public interpersonal: modality
during
epidemics and
pandemics
 Application guides

There are a number of good reasons for conducting studies of communication in healthcare,
relating to both field and tenor within context:

• field-based goals concern the processes and procedures of healthcare, ensuring that they run
without problems, so they include the accurate and secure transfer of information such as
patient records;
• tenor-based goals concern the roles and relationship among those taking part in healthcare,
critically between patients and healthcare practitioners, so they include sociometric roles
or considerations of affect – establishing rapport between patients and practitioners and the
practitioners’ capacity for empathy.

While field-based goals can be related to the need for fault-tolerant healthcare systems
where some part of the system can fail without the system breaking so that risk is minimized,
and tenor-based goals can be derived from the emphasis on patient-centred care, and also
under certain conditions on relationship-centred care, and concerns with joint decision
making and informed consent, it turns out that both sets of goals must be pursued together. If
tenor-based goals are ignored, this can have a severe impact on field: see e.g. Slade et al. (2016).
For example, if patients aren’t listened to and aren’t given opportunities to “tell their stories”,
doctors may miss information vital to field concerns of diagnosis and treatment, information
that may not be anticipated by their diagnostic questions.
The concern with tenor in terms of conceptions of care goes back at least to the 1960s in
healthcare (see Balint, 1969). In terms of the ordered typology of systems (Section 1.3.2), the
development of patient-centred care and relationship-centred care can be interpreted as a move
from the focus of the clinical gaze on patients as biological organisms to include also social inter-
pretations of patients as persons and, by another step, of patients as meaners: see Figure 4.22 and
Figure 4.23. As both persons and meaners, patients take on many roles in different social groups
and meaning groups, both within healthcare and in other institutions they take part in. This is
important because it shows how linguistic studies, and more generally semiotic studies, can play
a significant role in studies of healthcare designed to diagnose problems and produce results on
which solutions can be based, and it also shows how semiotic studies can complement studies
focussed on the lower systemic orders.
Healthcare can be studied longitudinally, as when a patient’s journey through an institution is
tracked. Examples include following a patient through counselling sessions, as investigated by Muntigl
(2004b) to determine how a patient’s story may change in the course of interaction between coun-
sellors and clients in couples counselling, and also patient journeys through hospitals. The latter is
illustrated Figure 4.24, which is on a study of the Accident and Emergency Department of a large
hospital in Hong Kong (for discussion, see e.g. Matthiessen, 2013; Pun et al., 2016; Slade et al., 2016;
Fung, 2018); research in Australia produced a similar picture: see Slade et al. (2015).
As they move through an Accident and Emergency Department, patients engage in a
sequence of situation types, starting with triage, and ending with either discharge or admission
to another hospital department. Along the way, they need to interact with different healthcare
professionals, centrally including different doctors and nurses, and maps of the networks of
people they interact with show why this may be bewildering to patients (cf. Slade et al., 2015).
From the point of view of the hospital, there is also a considerable challenge to ensure that
information is passed on accurately and effectively through hand over so that the complex team
can deliver co-ordinated continuity of care. Both the patient view and the hospital view can be
interpreted in terms of the intersection of field of activity and institutional role: the division of
labour in healthcare according to institutional role.

266
 Application guides

Figure 4.22 Views on patients according to different systemic orders within the ordered typology of systems –
as organisms (biological), as persons (social) and as meaners (semiotic)

Figure 4.23 The move from the clinical gaze in doctor-centred care oriented towards field concerns
towards relationship-centred care embodying both field and tenor concerns

267
 admission

situation type triage consultation examination diagnosis treatment discharge

institutional doctor
roles professionals

Application guides
nurse
268

patient

KEY:
spoken Cantonese/
Mandarin

written English

Figure 4.24 A patient’s journey through an Accident and Emergency Department in a large Hong Kong hospital
 Application guides

The journey represented schematically in Figure 4.24 illustrates the general notion of trac-
ing the movements of persons through institutions, engaging in particular institutional roles in
a succession of situation types. This illustration can also serve as a model for the investigation of
other institutions, like the institution of the family, the school, the workplace and the law. We
can chart people’s journeys through these institutions during a certain period of time, like a day
of studies or a day of work. If we focus on a particular person, we will of course find that they
move from one institution to another, taking up different institutional roles in situation types
located within different situation types. Such meanderings in meaning can be recorded in a
discourse diary or, if the timeframe is longer, a discourse biography.
As an alternative to such a longitudinal account following one person, or a set of persons,
over time, we can keep the situation type constant and map it in terms of all the roles involved
(tenor) and the activities they take part in (field). This mapping would be a record of the division
of social and of semiotic labour in a situation type, or set of related situation types, as illustrated
in Figure 4.25. This highly schematic representation shows persons in different institutional roles

Figure 4.25 Division of labour in an institution as a mapping between institutional role (tenor) and activity
(field), with examples from healthcare

269
 Application guides

engaging in different fields of activity (i.e. socio-semiotic processes). Such analyses are likely to
show that while doctors are focussed in field-based protocols, they leave the very important work
of enacting empathy and establishing rapport – tenor-based concerns – to nurses. In view of the
increasing awareness over the last few decades of the importance of patient-centred care, and even
relationship-centred care, this is an area that needs considerable further research (cf. Karimi, 2017).
In applications concerned with healthcare, we come back to the significance of institutions
as the critical isolates of culture located along the cline of instantiation, as shown in Figure 4.25.
In terms of the cline of instantiation, they are located within the region intermediate between
culture – the contextual potential of a community, and contexts of situation – the contextual
instances of a community. Approached from the potential pole, from culture, institutions can
be interpreted as sub-cultures, and we can see that any given culture is simply an open dynamic
adaptive aggregate of inter-related institutions. Approached from the instance pole, from con-
texts of situation, institutions can be interpreted as open dynamic adaptive aggregates of situation
types. Any given institution can be characterized in terms of the intersection of field, tenor and
mode ranges of values found in the situation types that make up the institution, as illustrated very
schematically for field and tenor in Figure 4.26.
In terms of the hierarchy of stratification characteristic of semiotic systems, institutions are
realized by functional varieties of language, i.e. by registers. Thus we can approach institutions
from below within the semiotic order of systems, and map them in terms of the ranges of regis-
ters that operate within them – functional varieties of language, but also of course the resources
of other denotative semiotic systems. Taking a step back, we can also locate institutions in terms
of the ordered typology of systems. Institutions are, fundamentally, semiotic constructs, but they
are enacted within the social order as subsystems of social organization, and they are embodied

Figure 4.26 The location of “institution” in the overall architecture of systems of different orders and the
cline of instantiation

270
 Application guides

as niches in ecosystems, and by another step manifested as physical habitats. As far as context of
situation is concerned (at the instance pole of the cline of instantiation), it is materialized (bio-
logically embodied and physically manifested) as what Ruqaiya Hasan has called the “material
situational setting”. But this is just a special case – the instantial case – of the materialization of
semiotic and social patterns.
Thus an institution of healthcare such as a hospital is simultaneously a sub-cultural system of
meanings, a domain of social organization, a niche in an ecosystem and a building.
Systemic functional research into communication in healthcare has often had application and
intervention as objectives, and there are a number of examples of training and of guidelines,
e.g. training based on research findings: Woodward-Kron et al. (2011); Woodward-Kron et al.
(2014); guidelines and recommendations for care: Brown et al. (2004); Clayton et al. (2007);
Slade et al. (2009); Slade et al. (2015: Chapter 5–7); Eggins and Slade (2016); cf. also Brown
(2014: 623–624) on future directions.
Let us round off this section by quoting from a case study dealing with communication in
healthcare, Pun et al. (2017):

Using ethnographic discourse analysis in an Emergency Department in Hong Kong,

this study explored the features of doctor-patient interactions in a hospital setting.
By audio-recording 10 patient journeys, from triage to disposition, we analysed
the complexity of turn-taking patterns in spoken interactions between patients and
doctors, as well as the subsequent complexities in this communication process. In
particular, we traced the flow of communication surrounding the patients’ medical
conditions at different stages of their journeys (e.g., taking patient history, making
diagnosis and translating medical information in a bilingual environment). Commu-
nication in this Emergency Department, as in all Emergency Departments in Hong
Kong, involves repeated translation from spoken Cantonese interactions to the writ-
ten English patient notes and vice versa. For this study, the ethnographic discourse
analysis includes different layers of detailed language diagnoses of the observed inter-
actions (e.g., turn-taking strategies, speech functions and exchange structures). In
this analysis, we examined the strategies that doctors used to transfer medical knowl-
edge to their patients and with other clinicians; this research illustrated how a series
of contextual factors (e.g., time pressure, staff shortages) were linked with the quality
of doctor–patient communication. To illuminate the path for future research, we
developed a dual-goal communication framework focusing on both medical and
interpersonal aspects of the doctor–patient relationship. We strongly recommend
the application of this framework for training medical students, junior clinicians and
clinicians in practice.

4.8 Summary
In this chapter, we have been concerned with applications of SFL as a resource with the poten-
tial and power to support a great and expanding variety of applications – as appliable linguis-
tics. Precisely because SFL was developed from the start (even since the pre-SFL activities in the
1950s) as an appliable kind of linguistics, there is now a wide range of applications empowered
by systemic functional theory and descriptions, and we have only been able to discuss a few of
them, noting that there are many more actual applications. Behind the actual applications that
have been explored so far, there is the vast potential for application inherent in SFL as an
appliable kind of linguistics.

271
 Application guides

We began by specifying the tasks of system description and text analysis, locating them within
the overall space of engagement with languages, and we presented the methodology associated with
these tasks as part of the overall research potential of SFL – as networks of options systemic func-
tional linguists can choose among when they undertake language description and text analysis. Lan-
guage description and text analysis form the foundation of most applications of SFL: pre-systemic
text analysis supports the development of rich comprehensive descriptions, and such descriptions
enable systemic text analysis, i.e. text analysis based on descriptions of the resources that engender
texts. Most applications involve text analysis, which is one reason it was part of the systemic func-
tional research agenda from the start (Halliday, 1964a). (And quite a few applications in applied
linguistics that are not normally informed by the analysis of authentic texts but are based on elicited
data by means of instruments such as questionnaires would benefit greatly from analysis of corpora
of authentic texts, and thus show what people actually say rather than what they say that they say.)
We went on to focus on three fields of application, viz. educational linguistics, translation
studies and healthcare communication studies. The first two have been part of the agenda of
applying SFL to solve problems since the 1960s, whereas the third has emerged in the last
couple of decades and is certainly a new growth area. Like all applications of SFL, the ones
we have discussed in some detail here can be viewed “ecologically” in reference to the meta-
context of application, i.e. in reference to the field of application, the tenor of application
and the model of application:

field – the activity of applying SFL and the domain of application:

activities include diagnosis of problems (e.g. finding out why people fail in education
or healthcare), intervention with solutions (e.g. new protocols; revised curricula,
in-service training programmes), consciousness raising about problematic practices
(e.g. intrusive job interviews, false or misleading claims in election debates), advo-
cacy (e.g. in favour of institutional support of small community languages), model-
ling (e.g. of language and other semiotic systems in knowledge construction and
management)
domains include a wide array of institutions and institutional settings – not only ones that
have been prominent so far such as education, healthcare, professional institutions like
translation and administration – but other institutions that collectively form a culture
(e.g. family, friendship, charity, recreation, religion, clubs and cabals).
tenor – the people taking part in the activity of applying SFL within different domains:
their institutional roles as members of the team, e.g. academic roles, professional roles,
“subject” roles of any people being studied, and the complementarity of these roles in
the pursuit of the goals of the project being undertaken
their power (status) roles as members of the team: project leaders – research associates,
research assistants, student helpers
their familiarity with one another (e.g. length of experience working together) and dia-
logue interfaces (in particular in larger teams: who talks to whom?)
mode – the role played by semiotic and social activities in the pursuit of the goals of the
project:
for example, what semiotic systems are deployed – language, databases or spreadsheets,
branches of mathematics such as statistics, and related software applications
rhetorical mode: is the orientation towards field goals or tenor goals; is the mode heuris-
tic, didactic, persuasive, preceptive, homiletic

272
 Application guides

mode: is the project undertaken in, or the work presented in, spoken mode, written
mode, or some combination involving new technologies of the channel
channel: what are the channels of communication among team members, e.g. face-to-
face team meetings or virtual ones using some collaborative platform

The preceding lists under the headings of field, tenor and mode can be used as ways of exploring,
designing and executing research projects of application in diferent institutional settings.
It is, of course, always helpful to search for previous similar projects that can serve as models
for new applications being proposed or undertaken; and in view of the growing array of applica-
tions of SFL, it makes strategic and tactical sense to go beyond one’s own disciplinary home –
for example, across education and healthcare, or across education and computational SFL. This
is yet another instance of the methodological principle of shunting in SFL. In this way, it will
be possible to get more mileage out of SFL as an appliable kind of linguistics with a vast and
expanding potential for application.
The holistic architecture of language (and other denotative semiotic systems) in context pro-
vided by SFL invites us to think systemically, shunting along the different dimensions of the archi-
tecture. For example, drawing on the work of Critical Linguistics by the “East Anglia” group,
Fairclough and scholars developed Critical Discourse Analysis, CDA. The focus on discourse was
natural since a great deal of the goings-on in a society that must be examined critically based on
a commitment to social accountability can be investigated by analysing the texts or discourses
central to these goings-on. However, this essentially locates the concerns at the instantial pole
of the cline of instantiation – albeit with implications for the system engendering the discourses
analysed critically; we can also imagine – and we urgently need – the complementary focus on
other regions of the cline of instantiation, importantly on the potential pole itself.
In short, we also need critical system description (suggested acronym: CSD). The signifi-
cance of the institutional development of such an area of activity is demonstrated by Halliday’s
(1990b) challenge to applied linguistics. This challenge turned out to energize the development
of ecolinguistics, an undertaking that will hopefully involve a critical stance towards all regions
along the cline of instantiation, not only of discourse at the instance pole. One of Halliday’s
points was precisely that by examining linguistic systems we can detect a certain inertia reflect-
ing the conditions of agrarian societies, and we can then view such systems critically against the
background of contemporary conditions.
The cline of instantiation is one dimension of the “architecture” of language in context
where shunting is imperative and will reveal fundamental complementarities. Let us just refer to
one more of these complementarities, viz. the complementarity of the comparison of descrip-
tions of systems in different languages and the comparison of original and translated texts: see
e.g. Lavid, Arús and Moratón (2009) and Matthiessen, Teruya and Wu (2008) on the general
framing of multilingual studies (cf. also Ellis, 1966, 1987b).
In view of the many splendoured applications of SFL, it seems appropriate to end this chapter
with a reminder of Bohm’s (1980) warning against the fragmentation of knowledge. However,
as long as we keep returning to, and replenishing, the collective and shared resources of SFL
(instead of only remaining in our own fields of activity and tenor relations), the very nature of
SFL as appliable linguistics can ensure that our knowledge is not fragmented. In the interest of
avoiding the fragmentation of knowledge, we would like to appeal to members of the com-
munity to avoid acronyms that rely on researchers and users being inducted into a particular
domain of application – e.g. CDA, CL, GSP, MDA, SF-MDA, SFTS, LCT, LT, GM, TPR. They
tend to emerge and be used for good Zipfian principles, but they can be off-putting or at least
opaque to members of other communities; so Zipfian principles must be balanced against tenor

273
 Application guides

considerations. And as electronic publishing is increasing its footprint, space-saving considerations

become less important.

Notes
1 There are many discussions and representations of Dürer’s “drawing machine”; this reproduction
comes from: https://anamorphicart.wordpress.com/2010/04/23/perspective/
2 For an example of the use of questionnaires in a meta-linguistic research project comparing and
contrasting functional-cognitive approaches to language, see Gonzálvez-García and Butler (2014)
discussed earlier in Section 2.6.2.1. Their methodology is not systemic functional, but their study
sheds light on how users of SFL answered their questionnaire.
3 Halliday (1994d/2007b: 240) comments in relation to work on Bernstein’s codal variation involving elici-
tation: “But these are all contexts where the language being used is under the subject’s own attention,
closely self- monitored as it progresses, whereas (as Boas used to stress in relation to his ethnographic work
a century ago) language is perhaps the most unconscious aspect of human behaviour, and may therefore be
considerably displaced in being reflected on. There is a wide gulf between what people say and what they
think they say – let alone what they think they ought to say; and sensitive investigations of this kind require
a database of authentic, natural and unselfconscious speech.”
4 One field of research where SFL has been involved where triangulation has been developed as a meth-
odology is translation studies – see Alves (2003) for the general notion – and this has enabled research
combining SFL with methods developed in translation process research (TPR), as discussed by Alves
et al. (2010).
5 However, it is interesting to note Linde and Labov’s (1975) famous classic study of how people “sur-
vey” the layout of their apartments; it turns out that a virtual walking tour is the most popular strategy.
In their study of the lexicogrammatical construal of space in English, Matthiessen and Kashyap (2014)
show how the uses of the resources vary across registers (see also Kashyap and Matthiessen, 2017,
2019).
6 Outside SFL, “discourse analysis” is of course sometimes or even often undertaken without any ref-
erence to a description of the system instantiated by text or discourse. But this then turns out to be
merely hopefully informed commentary on (“close reading” of) the texts being “analysed” in the tradi-
tion of explication de texte – and this actually applies to “texts” in semiotic systems other than language
(cf. Halliday, McIntosh and Strevens, 1964: 199; Halliday, 1982; Halliday, 1984b: 4; also Halliday,
2007b: 210, on Sacks’ commentary on conversation).
7 Allowing for text complexes – “macro texts”.
8 In their pioneering corpus-based reference grammar of English, Biber et al. (1999) provide quantita-
tive information from four registers, but these registers are arguably too general to reveal significant
patterns in the deployment of grammatical resources across registers.
9 Life as a journey is of course a very common (lexical) metaphor, and it is complemented by notions
like learner paths and study paths.
10 And we must include the important role of “popularized knowledge”, noting that systemic func-
tional research has shown that it may be construed in fairly different ways: see Fuller (1995, 1998).
11 The studies of child language development cited by Roy et al. (2006) in their presentation of the
“human speech-home project” are all US-based: Bloom, Brown, Bruner, MacWhinney, Tomasello.
Of course, the team have produced other reports on their amazing project, including a TED Talk by
Deb Roy, but their effacement of the Halliday tradition is not at all unique.
12 As originally conceived and developed by Mark Roget and as represented in the first edition of
Roget’s Thesaurus (1852/1992) and subsequent editions following his organization. But the term
“thesaurus” has also come to be used for synonym dictionaries.

274
 5
GUIDE TO ALTERNATIVES,
QUESTIONS, ISSUES AND
DEBATES

5.1 Choices among metalinguistic options

SFL conceptualizes language and other semiotic systems as resources and models them as system
networks. And SFL is of course itself a semiotic resource – a metalanguage for engaging with
semiotic system, providing us with a range of options in theorizing, comparing, describing and
analysing (e.g. Matthiessen and Nesbitt, 1996; Halliday and Matthiessen, 1999/2006; Matthies-
sen, 2007b). The choices of many options are straightforward, but there are other options where
questions have arisen, and different researchers explored and argued for alternative options.
(The variants of SFL that have emerged over time in pursuit of different agendas of research
and application can, to a large extent, be interpreted as dialectal and registerial variants of the
systemic functional metalanguage. But there are, naturally, also alternative accounts within the
same metalinguistic dialect and register, and we may have to choose among them although they
may in fact turn out to be complementary.)
In this chapter, we will identify a number of areas where issues have arisen, and outline the
alternative positions that researchers have taken. The format has often been one of thesis ^
antithesis, leading to debates. The outcome of the debates may be a synthesis, or a recognition
that the choice between thesis and synthesis depends on yet other considerations (cf. Halliday,
1964a, on syntax and the consumer, and Martin, 1998, on linguistics and the consumer). We
provide examples of issues from the SFL literature in Table 5.1. We characterize the issues as
either theoretical or descriptive; we could of course also include issues arising in the analysis of
particular texts – which often figure in sysfling messages, but they are too numerous to survey
since almost every text will provide material for discussion of different analytical options (cf.
Section 4.4; Chapter 7).
Some issues have been subject to recent reviews.

• Under the heading of “Continuing issues in SFL”, O’Donnell (2019) discusses two issues in
some detail, one being “To what degree are the grammatical categories of SFL determined
on notional grounds (mirroring extra-linguistic organization), rather than on regularities of
form?” – “how ‘semantic’ is the grammar” for short, and the other being “Where does genre
belong in relation to other components of the model?” (See also O’Donnell’s notes on criti-
cism of SFL available at: www.isfla.org/Systemics/Topics/index.html)

275 DOI: 10.4324/9781315675718-5

 Table 5.1 Examples of issues, positions, alternatives and debates in SFL

Area Issues Thesis Antithesis (criticisms, alternatives) Replies, debates

context theoretical the potential stratification of one stratum intersecting with context stratified: ideology Hasan (1996b, 1999);
context cline of instantiation > genre > register (e.g. Martin (2010); cf.
(e.g. Halliday, 2002a) Martin, 1992a) O’Donnell (2019)
descriptive the interpretation of parameters with mode part of field network (e.g.
“constitutive” vs. (e.g. Halliday, 1978) Hasan, 1999a)
“facilitative”
content plane: theoretical/ the stratal status of system lexicogrammatical system semantic system networks: cf. O’Donnell (2019)

Guide to alternatives, questions, issues and debates

semantics/ descriptive networks of the clause networks Fawcett (e.g. 1980)
lexicogrammar
theoretical the number of metafunctions three metafunctions (e.g. 6–8 metafunctions – Cardiff
Halliday, 1976a, 1978) – model (e.g. Fawcett,
ideational, interpersonal 1973b, 1980, 2000; Schulz
and textual, with ideational and Fontaine, 2019); for
differentiated into two modes discussion, see also Gregory
276

of construal (logical and (1982)

experiential)
lexicogrammar theoretical the status of rank: rank-based rank-based constituency: Matthews (1966); Halliday (1966a)
constituency (or not) Halliday (1961) for a chronology, see Fawcett
(2000: Appendix C)
descriptive the status of ‘bound’ clauses: hypotactically dependent: embedded (as adverbial they realize satellites
hypotactic or embedded Halliday (1985a) clause): the traditional in nucleus-satellite
position relations: Matthiessen
and Thompson (1988),
Matthiessen (2002a)
descriptive the status of verbal items in serving as elements of the verbal no verbal group: Fawcett review: Quiroz (2017:
a clause group: Halliday (1976a, 1985a: (1980, 2000) 306–310)
Ch. 6) central typological variable:
Matthiessen (2004a)
 descriptive the extent of Theme: up to and including first Berry (1996); always including evidence from information
including first experiential experiential element: Halliday Subject (in a declarative structure: Halliday and
element or beyond (1967/1968, 1985a) clause): Martin and Rose Greaves (2008)
[English, but also other (2003, 2007)
languages]
descriptive the payoff in positing Subject Subject essential in description no need for Subject: Quiroz
in Spanish of Spanish: longstanding (2008)
position, e.g. Lavid, Arús and

Guide to alternatives, questions, issues and debates

Zamorano-Mansilla (2010)
descriptive the experiential status of it in they do not serve participant they have experiential
meteorological clauses and roles: Halliday (1967/8, 1985a) reference: Davidse (1999)
there in existential clauses
descriptive the status of behavioural ‘behavioural’ as one of six not a distinct process type:
clauses in English primary process types: Halliday Banks (2016)
(1985a) onwards
277

descriptive the boundary between Halliday (1985a) onwards Fawcett (1987) Halliday and Matthiessen
material and relational (1999)
clauses
descriptive the distinction between experiential distinction: Halliday no experiential distinction: Davidse (1999); Halliday
attributive and identifying (1967/8, 1985a) Fawcett (1987) and Matthiessen (1999)
clauses
phonology descriptive intonation Halliday (1967) Brazil (1997) [in the [Halliday and Greaves
Birmingham School]; cf. (2008)]
O’Grady (2010) in SFL
 Guide to alternatives, questions, issues and debates

• Approaches taken in the “Cardiff Grammar” contrasting with the line of development orig-
inated by Halliday and continued by him and scholars building on his work are presented in
recent overviews by Schulz and Fontaine (2019) on “functional syntax”, Huang (2017) on
theme, Neale (2017) on transitivity. There is, of course, an extensive literature documenting
the Cardiff approaches to both theory and description going back to Fawcett (1980), with
key contributions being reviewed in the publications just cited.

Since SFL provides a holistic theory of language in context, it can accommodate positions that
might at first seem to be opposing alternatives as complementarities (see Section 1.4 earlier) such
as the synoptic ~ dynamic complementarity (see Martin, 1985a), as shown and illustrated for
system ~ instance, grammar ~ lexis and speaking ~ writing by Halliday (2008b). For example,
the grammatical image of the resource of wording and the lexical one can be reconciled as
complementary perspectives on the cline of delicacy within the stratum of lexicogrammar. This
does not mean that there are no problems to be solved; there certainly are interesting challenges,
such as phenomena explored in terms of “phraseology”, explored by Tucker (e.g. 2007). But it
does mean that it is possible to shed light on the problems, at least meta-theoretically.
As in other disciplines, it has often turned out that issues can be resolved by the introduc-
tion of, or consideration of, an additional semiotic dimension, as in the case of solving the
problem of accounting for the relation between clause and text (see Halliday, 1981). The history
of theory development in physics provides interesting examples, as shown by Kaku (1994) as
he chronicles the addition of dimensions since the 19th century in the theoretical construal of
physical phenomena.

5.2 Theoretical issues

We will focus here on theoretical issues that have arisen within SFL, but it can also be helpful
to examine issues where SFL differs from other traditions that are fairly close. It turns out that
as with issues internal to SFL, it is often the case the issues revolve around choice of semiotic
dimension. For example, how is the relationship between clause and text (or sentence and text)
to be interpreted? In SFL, it has been interpreted in terms of the hierarchy of stratification, as
emphasized already by Halliday and Hasan (1976) and again by Halliday (1981), but in Tagme-
mics it has been interpreted in compositional terms, based on the Tagmemic analogue of the
rank scale. Similarly, we would learn a good deal in meta-theoretical terms by comparing the
various proposals within Stratificational Linguistics for stratification of language (e.g. Lamb,
1966; Lockwood, 1972) with the systemic functional account. While the number of strata in the
modelling of language has varied in different versions of Stratificational Linguistics, they have
remained fairly stable in SFL, with the exception that as more attention was given to the expres-
sion plane, it was clearly articulated into both phonology and phonetics, not only phonology.1
However, here we will consider only issues internal to SFL.
We will consider theoretical issues under the following headings:

• the theoretical architecture of language in context (Section 5.2.1);

• context: stratified and/or extended along the cline of instantiation (Section 5.2.2);
• content plane: the stratal status of system networks (Section 5.2.3);
• content plane: the number of metafunctions (Section 5.2.4);
• content plane: rank (Section 5.2.5);
• exploration of the relationship between individual meaners and their speech fellowships
(Section 5.2.6).

278
 Guide to alternatives, questions, issues and debates

5.2.1 The theoretical architecture of language in context

Starting with Halliday’s (1961) presentation of the theory of grammar, proto-SFL (includ-
ing “scale-&-category theory”) and SFL have offered an architecture of language in con-
text based on semiotic dimensions and the relations that have them as their domains. The
early version (Halliday, 1961) specified levels, later called strata ([substance] phonic/graphic
substance – phonology/orthography – [form] grammar, lexis – context – [situation] extra-
textual features), in the “global” organization of language in context; this version is repro-
duced here as Figure 5.1. As SFL was developed by Halliday in collaboration with different
colleagues during different periods, he added (the spectrum) of metafunction and (the cline
of) instantiation as “global” semiotic dimensions. These are represented in Figure 1.2 ear-
lier, and specified together with their relations in Table 1.2. In addition, each stratum in
the stratified model is organized locally by dimensions that were already part of Halliday’s
(1961) account, viz. delicacy, rank and axis, the major change being the prioritization of the
paradigmatic axis as the primary axial mode of organization (Halliday, 1966c).
There are alternative versions of this multi-dimensional theoretical mode – principally, the
version representing the work by Michael Halliday, Ruqaiya Hasan and their colleagues and stu-
dents, and the version developed by J.R. Martin, first documented in detail by Martin (1992a)
and then further elaborated to take account of additional phenomena, e.g. by Martin (2010).
(These two variants also come with descriptive diferences in the accounts of English and other
languages, but it is important to maintain the distinction between theory and description.)
We will touch on some key differences between the two versions later, in particular in rela-
tion to the modelling of context, but against the background of strands within SFL that have
developed since the 1970s, the major difference is between the multi-dimensional “architecture”
characteristic of both versions and the kind of modular or componential architecture proposed
by Robin Fawcett and developed by him and his colleagues and students in and around Cardiff
University – now including scholars in different parts of the world.
The Cardiff architecture is much more like the models found in generative linguistics and
computational linguistics than like the multi-dimensional and relational model just referred to.
Versions of this architecture have appeared in a number of publications; we have reproduced
“The major components of a communicating mind” from Fawcett (2013) in Figure 5.2. The
orientation towards computational modelling is quite natural since this has been a major area of
activity in Cardiff under Fawcett’s direction. As the figure shows, “components” such as “belief

Figure 5.1 Halliday’s (1961) model of the levels of language in context

279
 Guide to alternatives, questions, issues and debates

Figure 5.2 Fawcett’s (2013: 132) Figure 6.2 “The major components of a communicating mind”

280
 Guide to alternatives, questions, issues and debates

system” and “sentence generator” are differentiated from states, “input/output”, such as “basic
logical form” and “enriched logical form”, and these are distinguished from “message flow”
and “consultation”. Some of the “components” appear to be purely declarative representations
of resources (e.g. “belief system”) whereas other components also involve processes, reflecting
the different directions of generation and interpretation (e.g. “sentence generator” and “sen-
tence interpreter”). This makes the task of translating the “components” into regions within
the multi-dimensional SFL architecture somewhat challenging; in this architecture, all processes
involve processes of instantiation, i.e. moves along the cline of instantiation.

5.2.2 Context: stratified and/or extended along

the cline of instantiation
Issues in the interpretation and modelling of context – as a connotative semiotic system – have
largely to do with which semiotic dimension is best suited to the theoretical account of the orga-
nization of context – the hierarchy of stratification or the cline of instantiation or both. There
is essentially agreement about the functional diversification of context although, as we shall see,
there is one crucial issue, viz. how to deal with what has been called by a few different names,
including “functional tenor” and “rhetorical mode”. Key contributions outlining different posi-
tions on context include Hasan (1999a, 2009a) and Martin (1992a, 1999a, 2013c, 2014); see also
Matthiessen (2015c), Gardner (2017), Bartlett (2015, 2017).

5.2.2.1 Terminology
Before we identify some key issues that have emerged in the SFL community regarding the
interpretation, theorization and modelling of context, we will offer a note on terminology. The
different uses of the terms we will include are, obviously, related to different theoretical framings
of context so we will come back to them, but it will be helpful to mention different uses of three
terms first: (1) context, (2) genre and (3) register.

(1) CONTEXT

In the early model of the levels (later “strata”) of language in context, Halliday (1961: Section 1.7/
2002d: 244) used the term “context” for ‘semantics’ and “situation” for ‘context’: see Figure 5.1.
In the article where he presented this model, he was concerned with “form” – grammar and
lexis (later, he included “phonology” and “graphology” under form as well), and while seman-
tics was called “context”, it already had the status of an “interlevel” between “form” and “situ-
ation” – later, between “lexicogrammar” (as form) and “context” (see e.g. Halliday, 1973). So
as he developed his account, “context” was renamed “semantics” and “situation” was included
under the general category of “context”.2
Terminologically, the situation (pun intended!) was soon sorted out, and “context” came to
refer to what Malinowski (e.g. 1923, 1935) and Firth (e.g. 1957c) understood by the term; and
“situation” was used as shorthand for “context of situation”.

(2) GENRE

The term “genre” is arguably a real headache. Like “discourse”, it has been used in so many
ways that it is hard to use it in a way that does not invite misinterpretations. That was in fact a
consideration in the 1960s, when Halliday and his colleagues decided to avoid the term because

281
 Guide to alternatives, questions, issues and debates

it came with too much “semantic slime” (one of Halliday’s jocular technical terms). In particu-
lar, “genre” was associated with classifications of literature (and of other semiotic forms of art,
including painting and film), and they used the term “register” instead, taken from Reid (1956).
In order to locate what “genre” traditionally referred to, Halliday (1978: 145) clarified its place
in his account of “mode”:

The concept of genre discussed above is an aspect of what we are calling the ‘mode’.
The various genres of discourse, including literary genres, are the specific semiotic
functions of text that have social value in the culture. A genre may have implications
for other components of meaning: there are often associations between a particu-
lar genre and particular semantic features of an ideational or interpersonal kind, for
example between the genre of prayer and certain selections in the mood system. Hence
labels for generic categories are often functionally complex: a concept such as ‘ballad’
implies not only a certain text structure with typical patterns of cohesion but also a
certain range of content expressed through highly favoured options in transitivity and
other experiential systems-the types of process and classes of person and subject that are
expected to figure in association with the situational role of a ballad text. The ‘fable’ is
a category of a similar kind.

But the attempt to stay clear of the term “genre” to avoid confusion did not work; it (re-)
appeared in various contributions to SFL, notably in Hasan’s (1984a) description of “the nurs-
ery tale as a genre”, in the “Sydney School’s” “Genre Model”, originally developed with close
attention to educational applications (e.g. Martin, 1992a; Martin and Rose, 2008; Rose and
Martin, 2012; Gardner, 2017; Tann, 2017), and in Bateman’s (e.g. 2008b) GeM model. So users
of SFL must continue to read the literature carefully, having glossaries within easy reach. And
outside SFL, “genre” continues to be used in a wide variety of ways (cf. Hyon, 19963 – followed
up by e.g. Martin and Rose, 2008: 20; Swales, 2009; Malavska, 2016; Gardner, 2017).

(3) REGISTER

The term “register” has been used in different senses in linguistics, in reference to phenomena
within both the content plane of language and the expression plane. The original use in SFL
of “registers” was as the technical term for ‘functional variety of language’ – variety of language
according to the context of use, the term having been taken from Reid (1956); see e.g. Hal-
liday (2008b: 114), Halliday, McIntosh and Strevens (1964), Hasan (1973), Halliday (1978),
Lukin et al. (2008) and Matthiessen (1993a, 2015b, 2019). As just noted, the term “register” was
adopted instead of the traditional term “genre” because the latter was felt to be limiting, and
too closely associated with literary genres. There was also a continuity with J.R. Firth’s notion
of “restricted languages” (see Butt, 2019); register was simply a generalization of this notion to
embrace functional variation in general.
However, beginning with the work by J.R. Martin and his group in Sydney – later to be
known as the “Sydney School” (cf. Rose and Martin, 2012) – the term “register” was, as it were,
moved up from language to context, due to a misunderstanding of Halliday’s use of the term.
Martin (1992a: 501–502) documents this stratal move:

Before beginning however, it is important to note that English Text extends the use
of the term register as defined by Halliday. Halliday uses the term simply to refer
to language as context’s expression plane – the linguistic meanings (entailing their
expressions) at risk in a given situation type. English Text extends the notion to cover
282
 Guide to alternatives, questions, issues and debates

in addition part of context’s content plane; register is used in other words to refer to
the semiotic system constituted by the contextual variables field, tenor and mode. As
outlined above, in the model of context developed here, register is the name of the
metafunctionally organised connotative semiotic between language and genre. This
means that instead of characterising context of situation as potential and register as
(context’s) actual, English Text treats register as a semiotic system in its own right,
involving notions both of system and process.

(The interpretation of Halliday’s use of “the term [register] simply to refer to language as con-
text’s expression plane” is based on Martin’s account of context as a connotative semiotic.)
Martin’s sense of “register” is roughly equivalent to “situation type”. In a footnote (number 7),
Martin (1992a: 589) adds a comment on his use of the term “register”:

This was originally simply a misunderstanding on Martin’s part of Halliday’s model (cf.
Thibault, 1987: 610); since it has now appeared in so many publications, it seemed
more appropriate to extend Halliday’s notion than undo the misinterpretation here.

In a generous jocular fashion, Halliday and Martin discussed on one occasion whether this
misunderstanding was due to Martin being a bad student (his suggestion) or Halliday being
a bad teacher (his counter-suggestion). It can of course be related to the natural (rather than
conventional, or “arbitrary”) relationship between context and semantics. But, in our view, the
insistence on using “register” along the lines of Martin’s “misunderstanding” has haunted SFL
publications for decades now and keeps causing fundamental problems in the understanding of
contributions to SFL. This is superficially a terminological issue, but it is in fact also related to
a fundamental choice in the theoretical interpretation of context, and of context in relation to
language.

5.2.2.2 Context: interpretation in terms of the hierarchy of stratification

and/or the cline of instantiation
The development of the interpretation and theoretical modelling of context is traced in Section
3.1.1 from Malinowski (1923, 1935) via Firth (e.g. 1957c) to its gradual re-interpretation in
SFL, first the parameterization of context in terms of field, tenor and mode starting in the 1960s
(e.g. Halliday, McIntosh and Strevens, 1964), and then the semioticization of context as a higher
order of meaning in the 1970s (e.g. Halliday, 1978) – a connotative semiotic system (e.g. Martin,
1992a). All along within SFL, context was theorized not as an optional extra or separate phenom-
enon but rather as integral to the overall account – a phenomenon that could be, and had to be,
related to language in terms of the semiotic dimensions posited as part of the general theory of
language in context. SFL has thus always embraced an ecological approach to language, interpret-
ing it in its semiotic environment (and, by further steps, in reference to systems of other orders).
Throughout the SFL period, there has been general agreement about the functional diver-
sification of context into field, tenor and mode, or some variant thereof (the interpretation of
“functional” tenor being a hinge), but two different approaches have emerged in the interpre-
tation of other aspects of the organization of context, relying on either of the two other global
semiotic dimensions – either on the hierarchy of stratification or on the cline of instantiation.
As we have shown in Section 3.1.1, having interpreted context as a “connotative semiotic”
in Hjelmslevian terms, as articulated by Martin (1992a), Halliday worked out the implications
of extending it along the cline of instantiation, key references being Halliday (1991d, 2002a) –
both originally published as not very easy to access publications. His stratification-instantiation
283
 Guide to alternatives, questions, issues and debates

matrix, reproduced in Figure 3.5 earlier in this volume, brings out the phases of context (and
of the linguistic strata) along the cline of instantiation very clearly. The phases are, moving
from potential to instance along the cline of instantiation: context of culture [potential] –
cultural domain or institution [subpotential]/situation type [instance type] – context of situation
[instance]. The mid-region along the cline of instantiation can be approached from either pole.
For example, Halliday (1991d) characterizes ‘school’ from both vantage points, having drawn an
analogy between culture and situation with climate and weather:

And likewise with “culture” and “situation”: a school, for example, is clearly a cultural
institution, a matrix of social practices governed by cultural norms and values. Be we
can also look at it as an assembly of situations: it consists of regular events called “lessons”
in which people in certain role relationships (teachers and pupils) take part in certain
forms of interaction in which certain kinds of meanings are exchanged. We can look at
it as system (this is what we mean by education: the school considered systemically), or
as text, repetitive instances of the processes of teaching and learning. We may choose to
look at this phenomenon from either end; but it is still a single phenomenon, not two.

Other contextual patterns can similarly be located along the cline of instantiation, including
ideological patterns. (That is, ideological formations are interpreted by reference to the cline
of instantiation in the first instance, not as a separate contextual stratum within the hierarchy of
stratification.)
The alternative was to explore the organization of context in terms of stratification rather
than in terms of instantiation, which is what Martin and his group did throughout the 1980s
(Martin and Rose, 2008: 16–18),4 initially within educational contexts of research and applica-
tion – but the need to accommodate ideology within the theoretical model was also an impor-
tant consideration (cf. Martin, 1986). Martin (1992a: 493–495) introduces Halliday’s (1978)
conception of context in relation to language and then moves on to Bakhtin’s (1986) work on
speech genres. Comparing the two contributions, he writes (1992a: 495):

The tension between these two perspectives will be resolved in this chapter by including
in the interpretation of context two communication planes, genre (context of culture) and
register (context of situation), with register functioning as the expression form of genre, at
the same time as language functions as the expression form of register. Register can then
itself be organised with respect to field, tenor and mode, reflecting metafunctional diver-
sity in its expression form, leaving genre to concentrate on the integration of meanings
engendered by field, tenor and mode as systemically related social processes. This three
plane model can be outlined as in Figure 7.2 [reproduced here as Figure 5.2 (a)].

(It is worth noting “processes” in “social processes”; in terms of the systemic functional architec-
ture of language in context, processes are associated with instantiation – the cline of instantia-
tion – rather than with stratification. See Section 3.2.3.) Having stratified context into genre and
register, Martin (1992a: 495–496) then takes one additional step to account for another aspect
of contextual phenomena:

Any configuration of this kind then needs to be qualified with respect to cultural diver-
sity (cf. dialogism and heteroglossia in Bakhtin, 1982). Clearly, meaning potential is not
evenly distributed across a culture (any more than material resources are). Access to genre,
register and language as semiotic resources is mediated through discourses of ethnicity,

284
 Guide to alternatives, questions, issues and debates

class, gender and generation, which discourses are in a continual process of negotiation
with each other (see Cranny-Francis, 1990; Kress, 1985/1989). Not only is this process of
negotiation manifest in all text, but it functions as well as the source of semogenesis, both
contextual and linguistic. It is for this reason that a fourth communicative plane, ideology,
will be articulated here, with genre, and hence register and language Is its expression form.

An alternative form of projection, incorporating this fourth plane, is presented as Figure 7.3.
(from Martin and Matthiessen, 1991) [reproduced here as Figure 5.3]. In this projection metare-
dundancy (Lemke, 1984) is reflected through the metaphor of concentric circles,5 with larger
circles recontextualising smaller ones; the size of the circles also reflects the fact that the analysis
tends to focus on larger units as one moves from phonology to ideology. Thus the tendency
at the level of phonology to focus on syllables and phonemes, at the level of lexicogrammar to
focus on the clause, at the level of discourse semantics to focus on an exchange or “paragraph”,
at the level of register to focus on a stage in a transaction, at the level of genre to focus on whole
texts and at the level of ideology to focus on discourses manifested across a range of texts. More
in the spirit of Firth than Hjelmslev, this projection lends itself to a reading whereby meaning is
constructed on all levels, backgrounding the form/content dualities structuring Figures 7.1 and
7.2 [reproduced here as Figure 5.2 and Figure 5.3].

Genre

Register

Language

(a)

connotative semiotic
stratified context plane expression form

tenor

genre field

mode

discourse lexico- phonology/

semantics grammar graphology

stratified content form expression

form
denotative semiotic
(b)

Figure 5.3 (a) “Stratifying context as language’s content plane”, from Martin (1992a: 495); (b) from
Martin (1996a: 40)

285
 Guide to alternatives, questions, issues and debates

Martin’s account of context as stratified into ideology, genre and register has been widely cited
and used extensively in educational linguistics but also in other domains of application. Fur-
ther key theoretical exploration of context include Martin (1999a, 2001, 2010). Tann (2017)
chronicles the development of the model of context in what he calls the “Sydney architecture”
(with “Sydney” in the specific sense of the “Sydney School”, as in Rose and Martin, 2012) and
details the inclusion of the cline of instantiation and the construal of individuation as a “scale
of identities”, to be noted immediately below. The location of genre as a contextual stratum
above “register” in the Sydney School sense is intended to make it possible to show how field,
tenor and mode are coordinated by genre and to explain combinations of field, tenor and mode
values but also lack of combinations – gaps in the culture. However, there is arguably a need
for extensive descriptive work documenting the realization of particular genres by field, tenor
and mode values. This interstratal relationship is not part of the account provided by Martin
and Rose (2008); reviewing their account of genre under the heading of “Genre in a functional
model of language and social context”, they note the absence of accounts of field, tenor and
mode in their book (p. 231):

The first thing we need to do in response to a query about the limits of genre is to place
our work on genre within the functional model of language and social context in which
it evolved. As we noted in Chapter 1, in this model (see Figure 1.9) genre is positioned
as an abstract level of analysis co-ordinating field, mode and tenor (known collectively as
register), and register is realised in turn through language (discourse semantics, lexico-
grammar and phonology/graphology). This picture means that of course there is more
to genre than the descriptions in Chapter 2–6 entailed. Our treatment of linguistic reali-
sations there was necessarily sketchy and exemplary; and as we apologised in Chapter
1, serious consideration of field, mode and tenor was beyond the scope of this volume.

In many applications of the Genre Model, it seems that “level-skipping” is common: the
stages of the schematic structure of a given genre are characterized in lexicogrammatical terms,
skipping register and semantics.
One decade into the 21st century, the cline of instantiation has been added quite explicitly
to this account and is intersected with the stratal model shown in Figure 5.4. However, context
is still stratified but now into two strata, genre and register, rather than three; ideology is now
no longer interpreted stratally as a higher stratum within context but as pervasive orientations
within the whole system. Tann (2017: 452–453) summarizes this theoretical move succinctly:

Martin (1992a: 507) proposes another stratum beyond genre that constrains these
choices and calls it ‘ideology’, defined as ‘a system of coding orientations constituting
a culture’. This initial attempt to account for heterogeneity in speech communities
along the lines of generation, gender, ethnicity and class was, however, abandoned sub-
sequently, because it was later felt that the notion of ideology as formulated by Martin
does not so much constrain the language system top-down via context, as suggested by
stratification, but rather is a variable that influences the different strata in specific ways
(for example Banks, 2009). The distribution of semiotic resources in communities and
the tensions between different discourses are subsequently modelled in the Sydney
architecture as distinct dimensions from stratification.

The hierarchy of stratification and the cline of instantiation are modelled as intersecting, just as in
the theory due to Halliday (e.g. Halliday, 1991d, 2002a; Halliday and Matthiessen, 1999/2006,

286
 Guide to alternatives, questions, issues and debates

ideology

genre

register

semantics

grammar

phonology

Figure 5.4 The stratification of context as the “semiotic environment” of language, from Martin (1992a:
496), taken from Martin and Matthiessen (1991)

2014; Matthiessen, 2007b, forthcoming a), sketched in Section 1.3.3 and represented diagram-
matically in Figure 1.5. Martin (2010: 21–22) introduces Figure 5.5 as a way of rounding of the
account of the hierarchy of stratification and the cline of instantiation (or “hierarchy of instantia-
tion” as he calls it). He comments:

Before turning to consideration of a further hierarchy, individuation, in the next sec-

tion, let’s pause for a moment here to reinforce the complementarity of realisation (a
scale of abstraction) and instantiation (a scale of generalization), and reflect on the fact
that all strata instantiate, and all texts are sub-potentializations of systems. Figure 1.9
[Figure 5.5] has been designed to flag this complementarity, since the illusion of reali-
sation conflating with instantiation (or with axis for that matter) is so strong.

(The “illusion of realisation conflating with instantiation” was reflected in Firth’s term “expo-
nence”, as Halliday, 1992c, points out.)
In addition to stratification and instantiation, the model now also includes another “dimen-
sion”, viz. individuation. Working with identity and affiliation in the interpretation of humour,
Knight (2010a: 36) writes:

Individuation is a relatively new concept in SFL theory that was originally coined by
Matthiessen (2003, cited in Martin, 2007). This theoretical parameter presents a per-
spective on identity that is informed by notions of ideology and relates specifically to
how individuals may display differential access to the linguistic resources of the culture.
As Martin (2006: 276) explains, individuation “interprets the relation of system to

287
 Guide to alternatives, questions, issues and debates

Figure 5.5 Language in stratified context intersected with the cline of instantiation between system and
text, from Martin (2010: 22)

individual (of cultural reservoir to individual repertoires in Bernstein’s terms; e.g. Ber-
nstein, 2000: 158)”. He further notes that for Bernstein, this relationship is mediated
by coding orientation, or the speech codes that regulate a child’s verbal communica-
tion and so shape his/her social identity (cf. Bernstein, 2000/1996). To date, the work
of Hasan (e.g. Hasan, 1992c, 1995a, 1996a, 2005) and her colleagues (e.g. Cloran,
1989, 1999, 2000; Hasan and Cloran, 1990; Williams, 2001) on coding orientation in
particular contexts is the only substantial body of work in SFL that informs our notions
of individuation (Martin, 2008: 36).

Martin and Rose (2008: 232) characterize individuation briefly in relation to genre:

Another important respect in which there is more to genre than canvassed here has to
do with what Matthiessen, 2003 has called ‘individuation’. In Bernstein’s terms, this
has to do with the relationship between the reservoir of meanings in a culture and the
repertoire a given individual can mobilise.

Martin (2010: 30) adds individuation and instantiation to a stratal representation of language in
context in the figure reproduced here as Figure 5.6.
The overviews of context in the two SFL handbooks published by Routledge (Bartlett and
O’Grady, 2017) and Cambridge (Thompson et al., 2019) touch on issues in the theoretical
modelling of context and descriptive work. Tann (2017) contrasts the “Sydney architecture”
with the Halliday and Hasan line of development, attempting to show the advantages of the
former where there are differences. Bartlett (2017) reviews aspects of the approach to context
in SFL, starting with the pre-systemic functional contributions by Malinowski and then Firth,

288
 Guide to alternatives, questions, issues and debates

Figure 5.6 Martin (2010: 30), “Realisation, instantiation and individuation in relation to genesis”

while also relating the discussion to Vygotsky (cf. also Bartlett, 2015). He concludes his review
with “interconnected critiques”, previewed at the beginning of his chapter as follows:

• Halliday’s classical Marxist orientation entailed an intellectual programme to correlate

variations in linguistic behaviour with variations in social structure (specifically, class and
situation type), which has resulted in an overly deterministic/congruent modelling of the
relationship between contextual and linguistic features, and which underplays divergence,
diversity and the unique socialisation of the individual; and
• some of the core elements of the SFL theory of context have not developed in pace with
the thinking that has surrounded them, with the result that such terms no longer have
the same meaning, or valeur, as in their original context and cannot interact as previously
theorised.

At the end of the chapter, he sketches his own solutions to the problems he sees, referring to e.g.
Bartlett (2012, 2013). The second point can only be discussed in reference to his more detailed
critique, but the first, which is arguably an ad-hominem point, seems to be contradicted by the
publications and activities by Halliday – and by Hasan and those working with them. Halliday
(2015: 95–96) documents the concerns of the Communist Party Linguistics Group:

More important, in our view, was the task of using linguistics to tackle current prob-
lems of social and political life. The 1950s was a period of rapid decolonization, when

289
 Guide to alternatives, questions, issues and debates

newly independent countries were having to work out their language policies and find
resources for developing a new national language, (or more than one) and making it
acceptable, and accessible, to the population in general; while at the same time main-
taining an international language, usually that of the former colonial power (in practice
mainly English or French), as the channel of communication and of ongoing interac-
tion with all the rest of the world. This imposed a huge strain on their limited avail-
able resources; and this could easily be made worse by wrongheaded decisions, such
as setting up planning agencies to design changes in the national language (inventing
new vocabulary for administration and for science and technology) without first trying
to understand how new registers evolve in general, and how the particular language
of their own creates new meanings when simply left to itself. [Footnote: reference to
Halliday (1990b)].
In more general terms, we felt that a marxist linguistics should give value to languages,
and varieties of language, that were usually regarded as of little value, and often ignored
altogether: minority languages, unwritten languages, languages of hybrid origin (cre-
oles), non-standard, or non-literary, varieties, spoken languages (especially casual speech),
trade languages, underworld languages and so on. Language planning meant working out
policies and priorities, particularly educational policy, to determine how limited funding
and energy should be distributed among competing agencies and institutions.

In his writings in general and in his work on education and society in particular, Halliday
has emphasized diversity, identity, tension, contestation, including particular phenomena like
children’s experience in the middle school years (Halliday, 1977c) and his classic study of what
he called “anti-languages” (Halliday, 1976c) and very general ones like the tension in societies
between old and new ways of meaning (Halliday, 1990b) and the overlay of modes of meaning
in semogenesis (Halliday, 2010).
For example, discussing the experiences children meet in their middle school years, Halliday
(1977c/2007a: 53–54) highlights their memberships in different meaning groups – the family
(which is “essentially adult-oriented”), the peer group (“child-oriented”) and the school (“child-
centred but adult-oriented”), and documents the different demands made on their language in
these different institutional settings in terms of register and the rhetorical skills associated with
them. He notes (p. 51) that children in middle school are “often adept at dialect switching” but
“it is not until adolescence, around the age range 13–18, that they learn the social significance of
dialect variation: the way in which adults use language as an index of social background, level of
formality, and so on”. Children have to negotiate membership in the different meaning groups
they operate in (p. 54): “By the time they are in school most children have been developing
their language not only quantitatively, by enlarging the total potential, but also qualitatively, by
learning to use language in new ways – in the service of different realities, so to speak.”

5.2.2.3 The nature of ideology

The nature of ideology has been being explored explicitly in SFL since the 1980s, and research-
ers have tried alternative ways of interpreting this very complex phenomenon theoretically; for
example:

1 in terms of the hierarchy of stratification: drawing on explorations in the 1980s (e.g.

Martin, 1986), Martin (1992a) suggests that ideology can be interpreted as the highest stra-
tum within contexts, as already noted earlier and shown in Figure 5.4;

290
 Guide to alternatives, questions, issues and debates

2 in terms of the cline of instantiation: this was where Halliday thought it would have to
be interpreted, as varied but consistent patterns of instantiation within the overall system
of language in context of culture,6 which of course relates to the interpretation of different
coding orientations within a culture (see Section 3.2.2, Figure 3.12 and Figure 3.1) and to
Hasan’s (1986) exploration of the ontogenesis of ideology;
3 in terms of the functional diversification of context, as an aspect of tenor: reviewing
approaches to ideology and studying different samples of scientific discourses, Banks (2009)
suggests this interpretation as a possibility.

The first option has been tried, but then ultimately rejected by Martin, as noted earlier (see e.g.
Martin, 2010; Tan, 2017). The second option is still part of the picture. In the most extensive
study of ideology, Lukin (2019: 70) notes the importance of the cline of instantiation (see also
Lukin, 2017, on the power of the systemic functional architecture of language in the interpreta-
tion of ideology):

The concept of the cline of instantiation helps us understand how ideological mean-
ings gather force, and become naturalized, by reiteration. Ideologies are produced
and reproduced in innumerable exchanges of meaning. Official ideologies, despite
their many available mechanisms of reinforcement, require constant cultivation. Even
in situations where power relations are at their most asymmetrical, the ideology that
defends the asymmetry requires constant fortification.
(Qabani, 2018)

This is consistent with the notion that ideologies are developed “subliminally”, as discussed by
Matthiessen (2015f). The third option would foreground the interpretation of ideology as value
system – an axiological interpretation, but this needs to be supplemented by ideology in the
sense of world view, a cosmology – which suggests that field is an equal partner in the functional
interpretation of ideology, and it seems likely that mode would also play an essential part since
ideologies difer in their divisions of labour between semiotic and social systems and among dif-
ferent semiotic systems.

5.2.3 Content plane: the stratal status of system networks

In the “reference” account of SFL that we have presented here – based on the work by
Michael Halliday and those elaborating his work without taking other forks in the road, as it
were – the axial distinction between paradigmatic organization and syntagmatic organization
permeates all stratal subsystems of language. That is why it has been called a fractal principle
of organization (e.g. Matthiessen, 1995a, 2007b): axial organization or axiality is manifested
in the different stratal and rank environments throughout the semiotic complex of language
in context.
This seems generally accepted in descriptions of the phonological systems of particular lan-
guages, and it is also clear in “Hallidayan” accounts of their semantic and lexicogrammatical
systems. However, some systemic functional linguists have treated system networks within the
content plane as inherently semantic and structures as inherently syntactic – following the posi-
tion first articulated in detail by Fawcett (1980) and then followed up in various publications
including, importantly, Fawcett (2000) and recently Fawcett (2017). Our interpretation of Faw-
cett’s position is set out in Table 5.2. Fawcett’s exploration of the modelling possibilities is based
on the basic mode set out in Figure 5.7, taken from Fawcett (2000: 36).

291
 Guide to alternatives, questions, issues and debates

Table 5.2 The relationship between stratification and axiality in Fawcett’s (e.g. 1980, 2000) work

Content domain Paradigmatic organization Syntagmatic organization

semantics system network (semantics)

lexicogrammar structure (syntax)

Figure 5.7 From Fawcett (2000: 36); see also Fawcett, Tucker and Lin (1993); Fawcett (2017: Figure 5.1)

Table 5.3 Planes in language intersected with form vs. substance

Plane Form Substance

content content form = lexicogrammar [wording] content substance = semantics [meaning]

expression content form = phonology content substance = phonetics
[sounding] [manifested sounding]

One issue here is terminological. Most leading work in the “Hallidayan” tradition of SFL –
including that of Martin’s “Sydney School” – draws on the Hjelmslevian insight into the distinc-
tion between “form” and “substance” on the one hand and “content” and “expression” on the
other, as specified in Table 5.3 (Fawcett, 2000: 2, only mentions Hjelmslev once, listing him
among “antecedents”). So in this interpretative framework, the contrast between “meaning” and
“form” makes no sense; they belong to diferent dimensions of diferentiation.7 But, of course,
the next step in comparison would be to try to transcend these terminological diferences.
Thus Fawcett (1980: 39) makes explicit the stratal location he assigns to system networks such
as those of transitivity and theme (capitals as in the original):

I want here to suggest that systemic relationships of transitivity, polarity, theme, etc. are
not simply a REFLECTION of the semantics, but ARE the semantics. And, more-
over, I would like to consider the possibility of A GRAMMAR WHICH USES SYS-
TEM NETWORKS ONLY IN THE SEMANTICS. Such a model reflects more
strongly than any possible alternative Halliday’s view of ‘language as a meaning poten-
tial, a . . . range of options in meaning [that] constitute the functional components of
the semantic system’.
(Halliday, 1975a: 17)

292
 Guide to alternatives, questions, issues and debates

Figure 5.8 Fawcett’s (1980: 45) “systemic functional model of language”

After giving detailed arguments for his position, Fawcett (1980: 45) presents an overview dia-
gram of his “systemic functional model of language”. (Note that this model has appeared in
many publications, including more detailed versions of it “embedded” in the complete archi-
tecture of language.)
One way of taking the comparison of Halliday’s and Fawcett’s theories of stratification in
relation to axis further would be to examine how Halliday’s stratal interpretation of a number
of phenomena fundamental to language are handled by Fawcett’s theory. These include the
double agnation within the content plane (i.e. in terms of both semantic and lexicogrammatical
systems), the nature of lexicogrammatical metaphor, the ontogenesis of the stratification of the
content plane as children continue to expand their meaning potentials, the interface nature of
the semantic system contrasting with lexicogrammar as a language-internal system, the scope
of the semantic system beyond direct realization through lexicogrammar (“form”) and “intona-
tion”. The nature of the semantic system as an interface to what lies beyond language includes its
potential for “transforming” meaning in other semiotic systems into linguistic meaning – mean-
ing in the connotative semiotic system of context, in denotative semiotic systems other than
language and also in bio-semiotic systems (sensorimotor systems; see Halliday and Matthiessen,
1999/2006). In this rich interpretation of the semantic system, the need for semantics to sup-
port reasoning has also been emphasized. In contrast, in Fawcett’s model (Figure 5.2) reasoning
is handled by a separate “planner” and “reasoner”, accessing the model’s “belief system”. The
rich interpretation of the semantic system is likely to be further illuminated by the growing
body of work in neurosemiotics refencing SFL, e.g. García and Ibáñez (2017) (cf. Matthiessen,
forthcoming e, and references therein).

5.2.4 Content plane: the number of metafunctions

Like everything else in SFL, the spectrum of metafunctions needs to be viewed trinocularly – a
point made already by Halliday (1978: 130–131). His visualization of this trinocular approach is
reproduced here as Figure 5.9. He comments on the different views (130–131):

293
 Guide to alternatives, questions, issues and debates

Figure 5.9 The spectrum of metafunctions viewed trinocularly (Halliday, 1978: 131)

The grouping of semantic components differs according to the perspective from which
we look at them.
From the standpoint of their realization in the lexicogrammatical system (i.e. ‘from
below’), the logical component is the one that stands out as distinct from all the others,
since it alone is, and always is, realized through recursive structures.
From the standpoint of the functions of the linguistic system in relation to some
higher-level semiotic that is realized through the linguistic semiotic (i.e. ‘from above’),
it is the textual component that appears as distinct, since the textual component has an
enabling function in respect of the other components: language can effectively express
ideational and interpersonal meanings only because it can create text. Text is language
in operation; and the textual component embodies the semantic systems by means of
which text is created.
From the point of view of the organization within the semantic system itself (i.e.
‘from the same level’), the experiential and the logical go together because there is
greater systemic interdependence between these two than between other pairs. This
shows up in various places throughout the English semantic system (the general pat-
tern may well be the same in all languages, though the specifics are different): for
example, the semantics of time reference, of speaking (‘X said – ’), and of identifying
(‘A = B’) all involve some interplay of experiential and logical systems. To illustrate
this from the semantics of speaking, the process ‘say’ is an option in the transitivity
system, which is experiential; whereas the relation between the process of saying and
what is said – the ‘reporting’ relation – is an option in the logical system of inter-
clause relations.

The view “from below” has also been explored in a number of subsequent publications concerned
with metafunctional modes of expression, including Halliday (1979a), Halliday (1981), Matthies-
sen (1991c, 2004a), Martin (1996a). In addition, we can note that Halliday (1975a, 2003a) has pro-
vided a developmental account of the emergence of the metafunctional organization of language
in the course of ontogenesis: microfunctions > macrofunctions > metafunctions.

294
 Guide to alternatives, questions, issues and debates

Halliday’s trinocular view of the spectrum of metafunction suggests that there may be differ-
ent arrangements if one view is foregrounded over another. Fawcett (1980: 28) presents a figure
showing “types of choice in meaning”, where he aligns his proposal for “functional compo-
nents” with Halliday’s (1973: 40, 144):8 see Figure 5.10. Fawcett (1987: 27) raises the issues of
the number of “functional components”:

Just how many functional components are there? There is no simple answer to this ques-
tion, because the answer depends on the criteria that are used in judging. In the rest of
this section I shall introduce a range of types of meaning, all very different from each
other, and yet all capable of being realised simultaneously in the one unit of the clause.

So at this meta-theoretical level, there is a general consensus: the answer depends on the crite-
ria. Halliday (1978) had already suggested that the answer will depend on the criteria based on
trinocular vision – not only the number but also the groupings. Seen “from above”, there are
two, viz. extrinsic and enabling, which can by another step be related to his first and second
order field, tenor and mode categories (see Table 3.2). After raising the question, Fawcett (1980:
27–38) goes on to discuss “functional components”, setting out “criteria for recognising func-
tional components” (pp. 34–38).

Fawcett (1980) Metafunctional

specifications based on
IFG
# clause nominal group Fawcett Halliday (1973:
40, 144)
1 TRANSITIVITY: he NUMBER: pins experiential experiential experiential
fell down a pin
he was pushed down
2 TAXIS: he fell down TAXIS: pins logical logical logical
if he fell down pins and needles relationships
3 POLARITY: he fell DETERMINATION: pins negativity ↕? interpersonal
down no pins
he didn’t fall down
interpersonal
4 MOOD: he fell down PERSON: pins interactional
did he fall down? you
5 MODAL ASSESSMENT: ATTITUDINAL affective
he fell down EPITHESIS: nice pins
unfortunately he fell lousy pins
down
6 MODALITY: he fell POST- modality
down DETERMINATION:
he may have fallen (pins
down possible pins)
7 THEME: he fell down (a rather nice pin thematic textual textual
down he fell rather a nice pin)
8 he fell down pins informational
yes they
x MODAL ASSESSMENT: MODAL ASSESSMENT: inferential? ? interpersonal
he fell down pins
he even fell down only pins
9 MODAL ASSESSMENT: POST- metalingual ?
he fell down, DETERMINATION: (the
he fell down, as it pins
were the so-called pins)
10 he fell down ? discourse ? textual
firstly, he fell down organisational

Figure 5.10 Fawcett’s (1980: 28) comparison of his functional components with Halliday (1973), with
addition of system names and the metafunctional assignments in IFG (e.g. Halliday and
Matthiessen (2014)

295
 Guide to alternatives, questions, issues and debates

Since this early critique of Halliday’s theory of metafunctions, Fawcett has followed it up in
various publications, including not only Fawcett (2000) but also elsewhere. By the same token,
systemic functional researchers working with Halliday’s theory of metafunction have continued
to contribute to our understanding of it. The extension from accounts of grammatical units to
semantic ones (cf. Halliday, 1981; Martin, 1992a, 1996a) has been very significant, partly because
it has shown that the metafunctions engender not only particular lexicogrammatical systems
such as transitivity, mood and modality but also semantic ones. Perhaps even more significant has
been the growing body of descriptions of an increasing range of languages, which has demon-
strated (among other things) that notions like “modality” (in Fawcett’s column) may be, or are,
too specific to particular languages and that “negativity” may be a metafunctional “floater” or
“swing system”, in that different languages may handle it in metafunctionally distinct ways. At
the same time, such descriptive studies together with systemic-functionally informed readings
of the literature on language typology reinforce the point that the metafunctions are part of the
general theory of language, not of the description of “types of meaning” in a particular language
such as English: see e.g. Matthiessen (2004a).
This last point can, in a way, be related to the success researchers have had in identifying
functional modes of meaning comparable to Halliday’s metafunctions in their accounts of semi-
otic systems other than language, including of course the pioneering studies by O’Toole (1994,
2011) and Kress and van Leeuwen (1990, 1996, 2006). Such studies do not, obviously, settle the
issue of the number and nature of metafunctions in our theory of language, but they certainly
shed considerable comparative light on them, and they provide further indication that metafunc-
tions should not be interpreted too concretely as “functional components” and cannot be tied to
systems of particular languages such as “negativity” or “modality”.

5.2.5 Content plane: rank

The treatment of the compositional units in SFL and also in the pre-systemic period leading up
to SFL is one of the earliest issues, stimulating a debate that started in the mid-’60s. For a chro-
nology of the “rank debate” through the decades, see Fawcett (2000: Appendix C, 309–338).
This was originally a debate between Halliday and Matthews. Halliday (1961) had put forward
the theory that composition in language is rank-based and other linguists adopted and developed
linguistic descriptions in terms of rank. Matthews (1966) criticized the concept of rank as it had
been specified by Halliday (1961), and Halliday (1966a) replied to his criticism. (The opening of
Matthew’s article gives an indication of his overall stance: “The present writer is convinced that
the so-called ‘Neo-Firthian’ grammatical theory (the theory of grammar advanced by M.A.K.
Halliday and his associates) deserves less attention than it has received in recent years.” A great
deal more attention was, of course, to be paid to ongoing development of this theory!)
Since the exchange between Matthews and Halliday in the mid-1960s, rank-based constitu-
ency has been further illuminated in a number of ways, importantly by the introduction of the
metafunctional spectrum and with it the distinct functional patterns engendered by the different
metafunctions, as first put forward in some detail by Halliday (1979b). By then, Halliday had
also sketched the ontogenesis of rank in his case study of one child learning how to mean (Hal-
liday, 1975a): the simple model of clauses consisting of words is expanded into one with groups/
phrases as an intermediate rank, and words are deconstructed into morphemes. The division of
grammatical labour across ranks thus emerges as the child’s system increases in complexity and
power to mean.
Further, descriptions of a growing number of languages have shown how languages vary
considerably in where they do the work in terms of ranked units, and grammaticalization

296
 Guide to alternatives, questions, issues and debates

studies – largely outside SFL – have shown how this changes as languages evolve, one aspect
of grammaticalization being the drift of grammatical items down the rank scale (cf. Matthies-
sen, 2004a). Thus items operating as elements of the clause may get more grammaticalized
in stages – and thus more highly systemicized and also phonologically reduced: first they are
cliticized to other elements of the clause, which constitutes a kind of intermediate position on
the rank scale, then they are incorporated into the structure of the unit serving as the element
that they are cliticized to, and then, moving further down, they are gradually reinterpreted as
bound morphemes. Such processes of grammaticalization are illustrated schematically in Figure
10.3 in Matthiessen (2004a: 563).
More recently, questions about rank have arisen within SFL – not about the usefulness of the
rank scale in theory and description but rather about the number and nature of ranks. There has
been some variation within treatments of lexicogrammar, primarily having to do with whether
sentences as grammatical domains are interpreted as a rank above clauses or as complexes of
clauses. Also, within treatments of phonology, there has been explorations of the possibility of
ranks higher than the tone group; Halliday (1961/2002d: 78) suggests “paraphone” as a possible
phonological rank higher than that of the tone group, but Tench (2011: Ch. 12) adopts the term
“paratone” as the analogue of a written “paragraph” – a “phonological paragraph” (p. 182).
However, the theoretical issues really emerge in work on the semantic systems of different
languages – actually, so far, primarily English, and they are directly related to the frontiers in
the descriptions of the semantic systems of particular languages. In treatments of semantics,
systemic functional linguists have also explored the ranking of units of meaning. One funda-
mental issue is whether there is a single semantic rank scale in a given language or different
rank scales differentiated in terms of register or in fact different types compositional scale (e.g.
rank-based constituency vs. hierarchies of perdiodicity). Working with Hasan’s account of
semantics, Cloran (e.g. 1994, 1995) has proposed a ranking unit intermediate between text and
message (in Hasan’s sense). Alternatively, the semantic composition the “rhetorical unit” has
been postulated to deal with can be explored in terms of logico-semantic complexing of the
kind articulated by Rhetorical Structure Theory: see Cloran, Stuart-Smith and Young (2007)
for a comparison that also includes phasal analysis. Cloran’s “rhetorical unit” can also be viewed
in relation to the work by Robert Longacre (see e.g. Longacre, 1970, 1979, 1996; Longacre
and Hwang, 2012).

5.2.6 Exploration of the relationship between individual meaners

and their speech fellowships
Perhaps one of the most difficult theoretical issues to address is conception and modelling of the
relationship between individual meaners and the meaning groups and speech fellowships that
they take part in. The phenomena to be sorted out, observed and interpreted are complex in
themselves, but then in addition they are overlaid by the different traditions of psychological/
cognitive approaches and social ones; now neuroscientific studies are beginning to shed more
light on the interpretation of the individual in interaction with others.
As far as the phenomena are concerned, two central considerations are (i) the systemic order
involved in the interpretation and (ii) the relationship between individuals and collectives and
the weight given to either as a starting point. (i) In terms of systemic orders, individuation
emerges as a property of biological systems (2nd-order systems) – individuation not being a
property of physical systems – and it is also manifested in the two higher orders of system, social
systems and semiotic systems; but, at these higher orders, an individual is more complex – being
an aggregate of repertoires of social roles in social groups and of semiotic roles in meaning

297
 Guide to alternatives, questions, issues and debates

groups (cf. Halliday, 1978). An added source of complexity in the study of these systems is that
4th-order systems have been interpreted both as semiotic systems and as cognitive systems – a
complementarity that is foregrounded in Halliday and Matthiessen (1999). Since the first exten-
sive statement of the Cardiff version of SFL, Fawcett (1980) has emphasized both social and
cognitive interpretations, whereas Halliday and Matthiessen (1999) argue for the added insights
the semiotic approach give us over the cognitive approach.
(ii) In terms of the relationship between individuals and collectives, the interpretation clearly
depends on the systemic order – and on the interpretation of 4th-order systems as either semi-
otic or cognitive. In classical mainstream cognitive science (which was dominant at least up
through the mid-1980s), the focus was almost entirely on the individual as an isolated mind – a
view that turns up in cognitive psychology, AI, philosophy of the mind and early cognitive lin-
guistics. This is where notion such as knowledge bases and belief systems come from.
One way of exploring angles of approach to the relationship is to compare and contrast
interpretations of meaners engaging in dialogic exchanges of meaning (see Bateman, 1985, for
an early significant but often overlooked study) – compare e.g. speech act theory emerging from
philosophy and adopted in AI and the Birmingham School approach to exchange structure
taken over by some lines of investigation within SFL (though by no means all).
In contributions to Bednarek and Martin (2010), a number of contributors interpret the
relationship between individual and collective by reference to a distinct semiotic dimension, the
cline of individuation (with references to a talk given by Matthiessen at ASFLA in Adelaide in
2003). This, then, constitutes one possible way of theorizing the relationship.
However, as far as empirical research goes, by far the most significant studies in SFL
illuminating the conception of individuals in relation to collectives are the child language
development studies – pioneered by Halliday (1975a) in his longitudinal case study of one
child, Nigel, learning how to mean, and followed up in several other similar case studies,
e.g. Painter (1984, 1999) and Torr (1997). These and comparable studies must form the
evidential basis of the development of theories in SFL of the relationship between individu-
als and collectives. Here contributions from computational SFL are also highly relevant in
the development of explicit models since individual speakers and listeners must be modelled
and computational modelling of language in general has been greatly influenced by accounts
of planning and goal pursuit (resonant with speech act theory and more generally with the
philosophy of action).
In a recent contribution, Williams (2015) explores “the mind” and “the brain” framed by sys-
temic functional linguistics. He proposes a conceptual framework for “embodied and embedded
cognition” based on SFL – drawing on the semiotic dimensions that constitute the architecture
of language and semiotic systems in general according to SFL and emphasizing the contribution
they make in enabling us to sort out levels of analysis. As part of his contribution, he also reviews
and critiques “previous conceptual frameworks in cognitive science”, in particular philosophical
accounts involving supervenience, classical approaches and also Marr’s three-level framework,
connectionist approaches and the “4E theories of cognition: Embodied, Embedded, Extended,
and Enactive cognition” (Chapter 2). His work can thus feed into dialogue between SFL and
strands within cognitive science, but it can also provide a systemic functional interpretation of
phenomena highlighted in cognitive science.

5.3 Descriptive issues

Descriptive issues abound, and we can only touch on a small number of them (a subset of the
ones identified in Table 5.1). Descriptive issues include questions that emerge as part of the

298
 Guide to alternatives, questions, issues and debates

description of a particular language, but they also include questions that arise in the analysis of
text based on an existing description. Such issues often appear on the systemic functional email
groups sysfling and sysfunc – often in reference to English, but also involving other languages.
In fact, such issues inspired O’Donnell, Zappavigna and Whitelaw (2008) to undertake their
study of the analysis of examples in reference to the system of process type, which in turn led
Gwilliams and Fontaine (2015) to follow up with a related study.
The descriptive issues covered here are all located within the content plane and organized
according to metafunction. As far as the expression plane is concerned, O’Grady (2017)
provides an overview not only of systemic functional descriptions of intonation, but also of
differences among descriptions offered by Halliday (1967) and Halliday and Greaves (2008) on
the one hand and contributions by Tench (1990, 1996) and by O’Grady (2010). In addition,
he reviews studies investigating the claim that English has foot-timed rhythm (rather than
syllable-time rhythm), and he compares the systemic functional descriptions of intonation
with the ToBi framework. For a recent relevant account of rhythm in relation to meaning,
see Martinec (2018).

5.3.1 Interpersonal
Descriptive issues that can be located within the interpersonal metafunction occur within both
semantics and lexicogrammar – and between them in the sense of issues having to do with indu-
bitably interpersonal systems within semantics and/or lexicogrammar.

5.3.1.1 Interpersonal systems: negotiation

Within the interpersonal metafunction, one region that has been explored through different
proposals is quite naturally that of dialogic interaction, accounted for in terms of exchange,
negotiation and speech function (Halliday, 1984b). (i) On the one hand, there have been
descriptive issues in the border area between SFL and the Birmingham School, with a good deal
of productive dialogue along the way (cf. Section 2.6.1.1 earlier in this volume). (And further
away from SFL, there have been other approaches coming from philosophy – speech act theory,
sociology – Conversation Analysis (CA), computational linguistics and AI – e.g. dialogue games.
Cf. Matthiessen and Slade, 2011). (ii) On the other hand, there have been alternative descrip-
tions within SFL (obviously related to the dialogue with and influence from Birmingham).
Differences here have included the issue of whether “exchange” constitutes a dialogic unit in its
own right or is a pattern that emerges through speech functional choices move by move and the
issue of how to handle exchanges that interactants open up while already involved in exchanges,
suspending them to deal with clarifications, challenges and the like that arise contingently –
“exchange dynamics” in Martin’s (1992a: 66) terms.
A number of key issues were already clear by the end of the 1980s, and Martin (1992a:
76–78) provides a helpful summary under the heading of “Consensus and debate”, beginning
with the following orientation:

Consensus and debate

Before illustrating the analytical framework outlined above with respect to three texts,
it is important to re-contextualise the model. To begin, as will be obvious to the
reader, the discussion is systematically incomplete in at least three major respects.
First, it has ignored work on speech act theory, which is clearly relevant to a more
delicate categorisation of negotiation and speech function and their realisation in

299
 Guide to alternatives, questions, issues and debates

lexicogrammar (see however Fawcett, van der Mije and van Wissen, 1988; Hasan,
1988, forthcoming for relevant bridge-building in this area). Second, it has not dealt
with conversational analysis as practised by the ethnomethodologists, and so is espe-
cially weak in the area of turntaking (although less so in other areas as will be pointed
out below). And thirdly it has dealt only minimally with intonation; whether this is
treated as a meaning making resource in its own right following El-Menoufy (1988),
as most delicate mood following Halliday (1967), or as a direct encoding of discourse
structure following Brazil (1981), this omission is a serious one. By way of apology
one can only point out that English Text is being written as a reasonably compre-
hensive model of discourse semantics within the framework of systemic functional
linguistics, and so could not afford to devote more than a single chapter to conversa-
tional structure.

Regarding speech act theory, we can note that it has of course been influential both in pragmat-
ics within linguistics outside SFL and in computational modelling of dialogue. But a comment
by a fellow traveller outside SFL, Willis Edmonson, is important to recognize; Edmonson (1981:
22–23) writes:

what Searle characterises and classifies are not units of conversational behaviour, but
concepts evoked by a set of lexical items in English – illocutionary verbs. The criterion
for the existence of an illocutionary set of a specific kind is essentially the existence in
English of a term which may be used in an explicit performative formula in the alleged
performance of that act.

After further insightful discussion of speech act theory of the Austin and Searle vintage, Edmon-
son (1981: 25) arrives at the following assessment:

The conclusion to be drawn from this discussion is that illocutionary categories in

Searle’s theory are common-sense, not technical categories (cf. Schegloff, 1977:
82–83). Searle is a philosopher, not a discourse analyst. For the characterisation of
verbal behaviour in spoken discourse we require technical terms.

We can add that Searle seems to have been guided by the way that dialogic interaction is con-
strued ideationally rather than by how it is enacted interpersonally – the ideational construal
involving the logical grammar of projection and the experiential grammar of verbal clauses. And
this is perhaps not surprising because that is how dialogue is represented for us in various regis-
ters and because the interpersonal system of negotiation has in fact co-opted ideational resources
to expand its meaning potential – interpersonal metaphors of mood, or “indirect speech acts”
(cf. Halliday, 1984b; Martin, 1992a; Halliday and Matthiessen, 2014: Ch. 10). Another issue
with speech act theory from the point of view of SFL has to do with interpersonal meaning in
relation to context, more specifically the tenor parameter of context: see Section 3.1.2.4, where
Halliday’s (1975c/2003b: 79–80) discussion of speech act theory is quoted at some length.
There is, of course, an extensive body of work drawing on speech act theory, not only in
philosophy, but also in pragmatics/linguistics and computational linguistic/AI. Here the con-
tributions from computational linguistics and AI are alternatives to SFL accounts that can be of
interest, because they tend to be more explicit than philosophical and linguistic accounts in the
sense that they provide models designed to be part of systems for text understanding and text
generation (of interest in computational SFL), and since the early foundational contributions to

300
 Guide to alternatives, questions, issues and debates

AI, there are now also applications within “applied AI”, e.g. concerned with dialogic interac-
tion between people and computational systems. They also provide a bridge to general work
of potential interest in accounts of goal pursuit and planning. Pioneering contributions include
Bruce (1975) – an exploratory brief conceptual sketch (usually credited as the first contribution
to make the connection); Cohen and Perrault (1979) – probably the most cited foundational
paper relating speech act theory to models of planning; Cohen, Morgan and Pollack (1990);
and Allen and Perrault (1980). Contributions along these lines can be read against Bateman’s
(1985) systemic functional account, developed in the context of the computational work but
foregrounding insights into “intersubjective achievement”. (The focus of speech act theory
could arguably have been on authentic dialogue as exchanges of meaning – “speech interaction
theory” in an alternative academic universe where the philosopher’s armchair corresponds to
the field worker’s mobile deck chair as a perch for observing, or also taking part in, naturally
occurring dialogue in context.)
While the development of speech act theory within computational linguistics and AI is prob-
ably the best-known approach to dialogic exchanges in this field, William C. Mann’s theory of
“dialogue games” (e.g. Mann, 1979, 1988, 2002; Levin and Moore, 1977) is important in this
context – not only because he saw the value of SFL in the development of text generation sys-
tems but also because he created the initial conditions for this very productive line of research
through a succession of projects producing the “Penman” text generation system. His inter-
pretation of exchanges in dialogue as “dialogue games” was inspired by the second intellectual
incarnation of Ludwig Wittgenstein. His model of “dialogue games” is, clearly, closer to the
general systemic functional approach to dialogic exchanges than is the philosophical approach of
speech act theory, so this is one reason for including it here as an alternative to SFL – but one
that may be worth exploring further. Another reason is that Mann’s model of dialogue games has
fed into the development of the model of dialogue put forward by Pickering and Garrod (e.g.
2004, 2006, 2021). Like Fawcett in his “Cardiff ” version of SFL (Fawcett, 1980, onwards), they
combine social and cognitive considerations in their model: interactants develop dialogue jointly
in a shared workspace, where they achieve alignment – shared representations (cf. the use of the
notion of alignment in contributions to Bednarek and Martin, 2010).

5.3.1.2 Interpersonal systems – appraisal: stratal location within content plane

The interpersonal resources for enacting various types of evaluation cover systems extended
along the cline of delicacy from the grammatical zone to the lexical zone within lexicogram-
mar – with their semantic correlates and prosodic phonological realizations (including accom-
panying paralinguistic features such as voice quality) or their graphological analogues. The first
subsystem to be described in detail was the highly grammaticalized system of modality in
English, first presented by Halliday (1970a) and then revised by him for successive versions of
IFG. There have been other proposals within SFL, and many outside SFL, and there are systemic
functional descriptions of languages other than English. However, as far as English is concerned,
there seems to be general consensus about Halliday’s description. In his function-rank matrices
(e.g. Halliday, 1970a, 1976a, 1978) he also includes “connotation” in reference to interpersonal
lexis (complementing experiential “denotation”, and also textual lexical cohesion).
The study of interpersonal lexis – or rather of interpersonal meanings that tend to be realized
lexically in English – was advanced by Poynton’s (e.g. 1985, 1990, 1996) contributions, and her
insights fed into the development of the account of the resources of appraisal by Martin (e.g.
2000), Martin and White (2005) and represented in a growing number of publications (see e.g.
Oteíza, 2017). Most of the contributions have been descriptions of the English resources or use

301
 Guide to alternatives, questions, issues and debates

of these descriptions in the analysis of a wide variety of texts. The description of appraisal cov-
ers engagement, attitude and graduation. Among these, attitude is the region that tends to be
realized lexically, either by purely interpersonal items (which may be close to grammar along
the cline of delicacy) or by lexical items that fuse experiential denotation and interpersonal con-
notation. It is in this area in particular that interesting issues arise having to do with the location
of the description within the content plane of English. Accounts of appraisal have interpreted
the resources as part of the description of “discourse semantics” (cf. Martin, 1992a), but there
are two related crucial theoretical issues that then arise:

• how are terms in appraisal systems realized lexicogrammatically? Descriptions that have
been published tend not to include explicit realization statements but rather lists of exam-
ples of lexical items that realize different options within attitude;
• what is the realization between interpersonal lexis and interpersonal grammar? There has
been extensive discussion of the relationship between experiential grammar and experi-
ential lexis – both theoretical and descriptive (cf. Section 3.1.4, on lexis as most delicate
grammar), but what is the interpersonal analogue (cf. Halliday, 2008b: 49–53)?

If the description of interpersonal lexis is “filled out” to account for attitude in particular,
then we could begin to see a cline from more grammaticalized regions of interpersonal (or
modal) assessment – those realized by the Finite, mood Adjuncts and comment Adjuncts in
English, and more open-ended lexical resources, possibly along the lines sketched systemically
by Martin and White (2005), and the semantic description of the interpersonal region realized
by attitudinal lexis could be taken in the direction of a more strategic account (that is, more
strategic and less taxonomic). (And, very importantly methodologically, it would show how
computational tools for doing automated “sentiment analysis” can be deployed to reveal pat-
terns involving attitudinal lexis.) Such a strategic account would bring out the semantic strate-
gies open to language users for assessing diferent ideational meanings, covering both lexical
realizations and grammatical ones. This would show that a number of options are comparable
to those found in modality (see Matthiessen, 2007c), like the contrast between ‘subjective’ and
‘objective’ orientation, and it brings out the semantic deployment of systemic contrasts in the
grammar, like the contrast between ‘mental’ and ‘relational’ clauses. This approach is illustrated
by Slade (1996) in her description of the semantic strategies in the register of gossip for judging
a workmate who’s transgressed the value system of the in-group (see also Matthiessen, 2007c).
Her account is register-specific, like Halliday’s (1972) description of regulatory semantics for
maternal control of a young child, but this is one helpful method making it easier to develop
more strategic descriptions of semantics – strategic in relation to recurrent types of contextual
settings of field, tenor and mode. (For an excellent general overview of approaches to attitudi-
nal meaning and evaluation, see Holmberg, 2002.)

5.3.1.3 Interpersonal systems – subject: the role of subject in the description

of different languages
In the description of English and of many other languages – both ones belonging to the Indo-
European family and ones belonging to other families, the category of Subject has been posited
as a functional element in the interpersonal structure of the clause as a move in dialogue. Being
a descriptive category rather than a theoretical one, it is like all other descriptive categories: it
must be established in the description of each language based on empirical evidence (see Sec-
tion 4.3) – critically, on naturally occurring dialogue involving negotiation (since Subject is

302
 Guide to alternatives, questions, issues and debates

interpreted as interpersonal). It is most definitely not a “universal” category, and its value in the
description of any given language must emerge from trinocular considerations; it should not
be assumed. In other words, it would be misleading to approach a “new” language looking for
Subjects. Rather, “Subject” will be used to interpret clusters of interpersonal properties that are
comparable to those characterized in other languages as “Subject”, but the evidence should be
internal to the language under description in the first instance.
The interpretation of Subject as an interpersonal category is a unique contribution made by
Michael Halliday. Other linguists have interpreted it in textual terms as something like topical
theme or in experiential terms as something like actor or doer; or else, they have assumed that it
is as it were a purely syntax-internal function serving to mediate between (in our terms) textual
and experiential functions, as in the system of voice in English. (Seuren, 1998, summarizes the
subject-predicate debate among linguists from the second half of the 19th century to around
1920, when linguists turned to other topics. To avoid the potential confusion about the nature
of subject, some linguists have abandoned the term, and instead introduced “pivot” with a more
specific sense.) Even Halliday flirted in the 1960s with the possibility that Subject could be
interpreted in logical terms (e.g. Halliday, 1967/8 [2005a: 148]), but that clearly didn’t provide
any real insights and he noted that it could be derived from the interpersonal interpretation of
the clause. A key reason Halliday arrived at the interpersonal interpretation of Subject in English
was that he worked with naturally occurring dialogic text (also in his study of language develop-
ment; cf. Halliday, 1984b). So Subject emerged as the “nub of the argument” in the negotiation
of propositions and proposals in dialogues that he analysed, and he interpreted interpersonally in
terms of “modal responsibility”.
It so happens that the interpersonal nature of Subject in English is fairly overt, e.g. in the way
it is picked up in “question tags” together with the Finite, the other function that constitutes
the Mood element of the clause as a move in English, as in You could study literature in a foreign
language, couldn’t you? This strategy for eliciting a response to declaratives (and also to impera-
tives), with a mirror image of the Mood element, appears to be unique to English, an “exotic”
feature, but there are many other pieces of evidence that Subject is a function that language
users operate with and one that is interpersonal. At the same time, Subject has more covert
properties – reactances in the grammar that take more time to arrive at in description through
careful probing. These include the orientation of modalities of the modulation type and also
of certain types of modal assessment (where the focus of the assessment is on the Subject rather
than on the proposition as a whole, as in Julien foolishly gives him his credit card just to get rid of him,
which is agnate with Julien is foolish to give . . . – Halliday and Matthiessen, 2014: 191–192), but
also some of the properties that non-SFL typologically oriented linguists drew attention to dur-
ing the discussion in the second half of the 1970s, where Austronesian languages played a central
role (Schachter, 1976, 1977; Keenan, 1976a, 1976b) – properties like “triggering” reflexives
(as indicators of the identity of the modally responsible element of the clause, the “nub of the
argument” and presupposition in tactically related clauses (i.e. preserving the nub of the argu-
ment). While these linguists did not put forward interpersonal characterizations of subject, such
properties can be thought of as interpersonal reactances (as can even the choice of Subject in
imperatives; cf. Dixon, 1994: 133–134, on “Subject” in imperative clauses in ergative languages) –
drawing on Whorf ’s (1956) insights into cryptotypes and reactances, usually or always illustrated
with examples from the experiential clause grammar. Thus in exploring the interpersonal clause
grammar of Japanese, Teruya (2007) shows that there are clusters of properties and reactances
that can productively be interpreted in terms of Subject as the modally responsible element of
the clause as a proposition or proposal – an element that is as it were interpersonally elevated,
interacting with the system of honorification.

303
 Guide to alternatives, questions, issues and debates

At the same time, systemic functional linguists have demonstrated that the clause grammar
of negotiation may be organized structurally in ways that are significantly different from those
of languages where the Subject has traditionally been used in descriptive statements (and even
English can dilute the modal responsibility of the Subject through impersonal constructions). In
particular, Martin (1990, 2004b) shows that the negotiation of clauses in Tagalog is not orga-
nized around a nodal element upon which the argument rests. His account reinforced the doubts
among SFL descriptivist of the value of anything like the category of Subject in the description
of various other languages as well, and Quiroz (2008, 2013) even argues against it as a category
relevant to the description of (Chilean) Spanish – a language where it would appear that Subject
is “coded” overtly as well as being implicated by interpersonal reactances (cf. Lavid, Arús and
Zamorano-Mansilla, 2010, who posit Subject in their description of Spanish). As is often the
case with traditionally posited categories like Subject, we may need to move through descriptive
phases of thesis, anthesis and then synthesis.
As noted earlier, when we observe and interpret interpersonal patterns in a language as part
of the development of a description of the language, we must adopt trinocular vision – viewing
such patterns “from below” in terms of patterns of realization, “from above” in terms of the
interpersonal semantics of the clause as a move in dialogic exchanges of propositions and propos-
als, and “from roundabout” in terms of the interpersonal systems of the clause as a move – the sys-
tem of mood, of course, but also other interpersonal systems such as those of modal assessment
and of “speech level” (as in Korean and Japanese). If these three angles on the patterns resonate
sufficiently, we may have reason to posit Subjects as part of the interpretation of them. This will
then be a thick description of Subject (cf. Halliday and Matthiessen, 2014: 147–150) – one
where it is justified relationally along the various relevant semiotic dimensions that we can see by
adopting trinocular vision. Neither Subject nor any other descriptive categories are “things in
themselves”; they are simply nodes in the multidimensional relational interpretation of the lan-
guage under description (cf. Halliday’s, 2014, illustration of explanation in SFL, with reference
to Subject). So they must be explored and justified in terms of multiple relationships that they
enter into.

5.3.2 Textual systems – theme: the boundary between Theme and Rheme
Here we will note some variation within the IFG tradition of textual descriptions. The “Cardiff ”
grammar approach to the description of theme differs from this tradition in various respects,
e.g. in the grouping of certain constructions as “enhanced theme”: Huang (2017) provides an
overview of the Cardiff description, including pointers to differences between this approach and
the IFG tradition.
Forey and Sampson (2017: 138–142) discuss different views on where to draw the boundary
between Theme and Rheme in English. Since Halliday’s original account to the thematic orga-
nization of the clause in Halliday (1967/8), which is elaborated in the successive version of his
Introduction to Functional Grammar, another view has been put forward: see Table 5.4.
In Halliday’s account of the extent of Theme, represented in successive versions of IFG,
thematic analysis is, of course, complemented by Subject analysis; and the two give different
angles on text, as illustrated forcefully by Halliday (1992a) in the lexicogrammatical analysis of
a fundraising letter. The complementarity of the analysis of Theme and of Subject will be even
more obvious in the analysis of dialogic texts.
Issues around the description of the system of theme, and the analysis of Theme in text,
also include the treatment of dependent clauses preceding the dominant clauses they qual-
ify. In Halliday’s IFG, they are treated as thematic within the clause nexus they operate in

304
 Guide to alternatives, questions, issues and debates

Table 5.4 Alternative views of the boundary between (topical) Theme and Rheme in descriptions of
English

Topical theme Up to and including the first experiential element of Up to and including the subject of the clause
the clause

references Halliday’s IFG; Halliday and Greaves (2008); Berry (1995, 1996); Martin and Rose
Matthiessen (1995a) (2007); Forey (2009); Huang (2017)
arguments argument based on potential break between argument based on thematic progression,
Theme and Rheme in terms of information with lexico-referential chains often
units running through Subject

(Halliday and Matthiessen, 2014: 549–553): the thematic principle is seen at work within
different grammatical domains, not only that of the clause but also those of the clause nexus
and groups (both nominal and verbal ones). And the manifestation of the thematic principle
extends beyond grammatical domains to semantic ones, where it is manifested in the orga-
nization of text, as suggested by Halliday (1981) and taken further e.g. by Martin (1993a,
1996a). In certain other accounts pre-posed dependent clauses are treated as marked topical
Theme within the clause that they qualify, as in Martin and Rose (2007). It is worth noting,
however, that such pre-posed dependent clauses often scope over the whole remainder of the
clause complex they are part of, and even beyond (as shown for dependent purpose clauses
by Thompson, 1985).

5.3.3 Experiential systems – TRANSITIVITY (including process type)

The experiential system of the grammar of the English clause – transitivity – has been being
described since the early 1960s (cf. the “Bloomington Grammar” in Halliday, 1976a). Halliday
has presented successive versions, and we can see a step-wise refinement of the description,
making it both more comprehensive and more semantically “transparent” (cf. Matthiessen,
2018a). Let us track this development not only because it is helpful when we consult different
versions of his description but also because it helps us see where other scholars have intro-
duced variants of his account.9 Naturally, descriptions of the transitivity systems of languages
differ from the description of English that we track here; for some overview, see Matthiessen
(2004a).

5.3.3.1 The development of Halliday’s description of the system

of TRANSITIVITY
The first extended published account of the system of transitivity was Halliday (1967/8),
where he introduced his systemic description of the experiential system of transitivity together
with his descriptions of the textual systems of theme and information. He characterized tran-
sitivity as follows (1967/8/2005a: 7):

Transitivity is the name given to a network of systems whose point of origin is the
major clause, the clause containing a predication; it is thus simultaneous at the point of
origin with other networks such as those of mood and theme (Halliday, 1964b). The
transitivity systems are concerned with the type of process expressed in the clause, with

305
 Guide to alternatives, questions, issues and debates

the participants in this process, animate and inanimate, and with various attributes and
circumstances of the process and the participants.

While he included system networks and a chapter on process types in Halliday (1976a: 110–114;
159–173), the next extended presentation of his description of transitivity did not appear until
Halliday (1985a); he devotes Chapter 5 of his Introduction to Functional Grammar (IFG 1) to this
system, and returns to it in Chapter 10, where he deals with metaphors of transitivity. In Chapter
5, he gives an extended characterization of transitivity (p. 101); let us quote the revised version
from the second edition of IFG, IFG 2, Halliday (1994c: 106):

In Chapter 4 we were looking at the clause from the point of view of its interpersonal
function, the part it plays as a form of exchange between speaker and listener. In this
chapter, by contrast, we shall be concerned with the clause in its experiential func-
tion, its guise as a way of representing patterns of experience. Language enables human
beings to build a mental picture of reality, to make sense of what goes on around them
and inside them. Here again the clause plays a central role, because it embodies a gen-
eral principle for modelling experience – namely, the principle that reality is made up
of processes.
Our most powerful impression of experience is that consists of ‘goings-on’ – hap-
pening, doing, sensing, meaning, and being and becoming. All these goings-on are
sorted out in the grammar of the clause. Thus as well as being a mode of action, of
giving and demanding goods-&-services and information, the clause is also a mode of
reflection, of imposing order on the endless variation and flow of events. The gram-
matical system by which this is achieved is transitivity. The transitivity system con-
strues the world of experience into a manageable set of process types.

The descriptions presented in IFG 1 and IFG 2 are basically structural rather than systemic –
although they are, of course, derived from the systemic description Halliday had been devel-
oping since the first half of the 1960s. The successive published systemic descriptions that had
appeared by the time Halliday published IFG 1 are summarized in Table 5.5, beginning with the
transitivity network from the 1964 grammar (Halliday, 1964a).

Table 5.5 Halliday’s successive systemic descriptions of transitivity in English

Period Description

1964a

306
 Guide to alternatives, questions, issues and debates

Period Description

1967/8

1969

1977

307
 Guide to alternatives, questions, issues and debates

The systemic description closest to the one underpinning IFG 1 appeared in Halliday (1973),
and then in Halliday’s (1977b) systemic and structural analysis of a text, an analysis that served to
illustrate the conception of text as semantic choice. This systemic description was only elabo-
rated where it needed to be in order to take account of the text being analysed. It differs in a
couple of places from the later version that informed IFG 1: in IFG 1, ‘behavioural’ clauses are
introduced as a process type distinct from ‘material’ and ‘mental’ ones (for a different position,
see Banks, 2016), and ‘existential’ is treated as a primary type of process, not grouped together
with ‘relational’ (as they were in e.g. Matthiessen, 1995a, which was an expansion of the descrip-
tion of the English clause Halliday had developed for a computational project: see Halliday,
2005c: 268–28410). This description was shown systemically in the third edition of IFG, IFG 3,
when Halliday and Matthiessen introduced system networks to accompany the description of
the major systems of English (Halliday and Matthiessen, 2004): see Figure 5.11. Here the system
of process type is shown explicitly with six terms: ‘material’/‘behavioural’/‘mental’/‘verbal’/‘
relational’/‘existential’.
We will say a little bit more about revisions of the description of the system of process type
in the next subsection, but let us first note one area of descriptive alternatives in the overall
account of transitivity in English offered by different scholars. For identification and discus-
sion of alternative descriptions, see also Davidse (2017), who also includes references to work
on languages other than English, and Neale (2017), who provides a brief overview of different
accounts of transitivity in SFL and then focusses on the “Cardiff Grammar” in comparison with
Halliday’s account.
This has to do with the way in which the ergative and transitive models are reconciled in
the overall description of the system. In Halliday’s work, they constitute two complemen-
tary perspectives on the system of transitivity, and represented in terms of the systems of
agency and of process type in Figure 5.11; the details are provided in Section 5.8 of Hal-
liday and Matthiessen (2014). According to this account, the ergative perspective constitutes
a generalization about transitivity based on cause-&-effect, and the transitive perspective is
embodied in the description of particular process types, the traditional notion of intransitive
vs. transitive being represented most clearly by the grammar of ‘material’ clauses – but as
Halliday has shown since Halliday (1967/8) the ergative model has increased in descriptive
power in the account also of ‘material’ clauses over the last 500 years or so in the evolution
of English.
Scholars have explored alternative ways of bringing the ergative and transitive patterns
together in the description of English transitivity.11

5.3.3.2 The system of PROCESS TYPE in particular

As the successive systemic descriptions in Table 5.5 show, Halliday gradually worked out the
nature of the system of process type, showing more and more clearly that natural relationship
between the experiential nature of the clause in lexicogrammar and its semantic correlate, the
figure, and this development of the description was also accompanied by continuing analysis of
text, an early significant and much-cited example being Halliday’s (1971) analysis of passages
from Golding’s The Inheritors. The analyses by Halliday and other scholars covered an increasing
volume of text, providing ongoing opportunities to test and refine the description.
In the 1964 description, the basic distinction was between ‘intensive’ clauses and ‘extensive’
ones, and this was also Halliday’s (1967/8) point of departure. ‘Intensive’ clauses are clauses of the
ascriptive type of ‘relational’ clauses, with examples such as (from the 1964 account) Mary seemed
happy; Mary made a good wife; Mary made John a good wife, and (from Halliday, 1967/8/2005a: 7)

308
 Guide to alternatives, questions, issues and debates

Figure 5.11 The summary system network of transitivity in Halliday and Matthiessen (2004: 302)

309
 Guide to alternatives, questions, issues and debates

Mary looked happy. By the time Halliday (1967/8) had reviewed the description of transitivity
from the point of view of the ergative model, he had identified ‘action’ clauses, ‘mental’ clauses
(including also ‘verbal’ ones), and introduced ‘relational’ clauses as a grouping of ‘intensive’ and
‘equative’ clauses (Halliday, 1967/8/2005a: 136–138):

Such clauses [sc. action clauses] have been distinguished, in the previous discussion,
from clauses of ascription (“intensive” clauses) and, latterly, from clauses of mental pro-
cess. However, many of these other types can be accounted for as displaying the same
features as are associated with action clauses. If instead of “intensive” a more general
feature “relational” is postulated as a third process type contrasting with action and
mental process, this can be extended to cover also the equative type. . . .
Intensive and equative clauses may thus be grouped together as relational and share
the features of other clauses as follows:
(1) intensive (ascription): middle or non-middle, descriptive; attributive, depictive or
resultative
(2) equative (a) decoding: middle, nuclear (be) or descriptive (others); range, range-
active or range-receptive

By the end of his Notes on transitivity and theme, Halliday had thus arrived at the three-term sys-
tem of process type, ‘action’/‘mental’/‘relational’, that has formed the basis of all later descrip-
tive developments, where the term “action” was replaced by “material”, yielding the contrast
‘material’/‘mental’/‘relational’. These later developments included considerable further elabora-
tion of the primary process types in delicacy, but also refinements in the treatment of the primary
process types themselves: the developments are summarized in Table 5.6. The second column
represents the version of the description from 1980 developed by Halliday for a computational
implementation of the grammar, originally for a project at UC Irvine for a project directed by
Benjamin Colby but then taken over in our work on the Penman projected, directed by William
C. Mann at the Information Sciences Institute, and given the name “the Nigel grammar” (e.g.
Matthiessen, 1983; Mann and Matthiessen, 1985).
As can be seen from Table 5.6, a number of the changes in the description of transitivity are
concerned with the delicacy of the differentiation of systemic terms in the system of process
type; for example, are ‘material’ and ‘behavioural’ distinct primary types of process or second-
ary subtypes? and, similarly, are ‘relational’ and ‘existential’ distinct primary types or secondary
subtypes? (see Figure 5.12). Halliday (1985b/2002d: 277) offers a useful tabular display that
groups the six primary process types into three broad areas, (i) ‘doing’, (ii) ‘sending/saying’
and (iii) ‘being’: see Table 5.7. In other words, it stands to reason that there has been variation
in the description of the system of process type: as more facts have been uncovered since the
descriptive work began, we can make a more informed decision about where to move in when
we postulate the first or primary system (cf. Martin, 1996b). However, there are still likely to be
different possibilities depending on the relative weight given to different considerations and the
particular task at hand; so linguists are likely to continue to argue about the “right” description –
until they realize that the indeterminacy is a feature of language!12
As far as systemic functional descriptions of the system of process type as described in Hal-
liday’s IFG are concerned, alternatives that have appeared since the first edition of IFG include:

• Fawcett (1987) draws the boundary between ‘material’ and ‘relational’ clauses in a different
place, which is discussed by Halliday and Matthiessen (2006);

310
 Table 5.6 Refinements in the description of the system of process type from the 1960s (also reflected changes of some of the names of participant roles, e.g. Cognizant
> Senser; Attribuant > Carrier)

Guide to alternatives, questions, issues and debates

1967/8 1980 [Nigel] (Also LexCart) 1985 [IFG 1] 2004 [IFG 3] 2014 [IFG4]

action effective/nuclear/ material [& effective] creative/dispositive material material creative

descriptive [& middle] eventive/ transformative
behavioural
– behavioural behavioural
mental reaction mental reactive mental affective mental lower-order emotive
311

desiderative
perception perceptive perceptive higher-order perceptive
cognition cognitive cognitive cognitive
verbalization verbal verbal verbal
relational ascriptive [intensive]/ relational intensive relational intensive relational intensive
equative
possessive possessive possessive
circumstantial circumstantial circumstantial
existential existential existential
 Guide to alternatives, questions, issues and debates

Figure 5.12 The choice of where to locate the primary cut in delicacy along the cline of delicacy in the
description of the system of process type

Table 5.7 Reproduction of Halliday’s (1985b/2002d: 277) Figure 5.10

Pro-verb Direction Participants Unmarked Verb
(active) PresentTense Accent
medium other

I Doing present in +
1 material do, happen, one thing thing present
2 behavioral do to, do one conscious
with do being
II Sensing/ present +/−
saying
3 mental (do to) two: conscious thing or fact
a likes x, x being
pleases a
4 verbal one signal source
III Being present -
5 relational - one thing or fact
Attributive class of thing or fact
Identifying identity of thing
or fact
6 existential one thing or fact

312
 Guide to alternatives, questions, issues and debates

• Fawcett (1987) does not recognize the distinction between ‘identifying’ and ‘attributive’ (or
‘ascriptive’) relational clauses as one of transitivity, which is discussed by Davidse (1996a)
and by Halliday and Matthiessen (2006);
• Banks (2016) does not treat ‘behavioural’ clauses as one of the primary types of process
types, which is of course an area where descriptions have varied over the years;
• Systemic functional descriptions of English have varied in the treatment of ‘existential’
clauses, either taking them to be one of the six primary process types or interpreting them
as a subtype of ‘relational’ clauses.

Neale (2017) updates the comparison of the “Cardif Grammar” description of process types
(and transitivity in general), comparing it with the IFG description in her Table 12.1. Difer-
ences of the kind listed here can usually be characterized by reference to Figure 5.11. This does
not mean that the descriptive tensions have been resolved, but they have at least been diagnosed.
See also Martin (1996b: 329–332) for a discussion of diferent classifications of process types, and
the whole chapter for an exploration of approaches to transitivity and contextualization of Halli-
day’s approach, also with reference to Whorf ’s important notions of cryptotypes and reactances.

5.3.4 Logical
The logical mode of the ideational metafunction construes our experience of the flow of events
as chains or series; this logical mode contrasts with the experiential of construing experience
as organic configurations of unit wholes (see e.g. Halliday, 1965, 1979a; Matthiessen, 1995a:
Chapter 3; Martin, 1996a; Matthiessen, 2002a, 2004a; Halliday and Matthiessen, 2006, 2014:
Chapters 6 through 8; Butt and Webster, 2017 – and Table 9.2). Within the stratum of lexico-
grammar, we can take the system of clause complexing – taxis, logico-semantic type and the
option of systemic recursion – as the most elaborated prototype of logical organization. Other
logical domains – complexing below the rank of the clause and logical organization of groups
(representing them as word complexes) – are in some sense reduced versions of the potential for
clause complexing; they are lower-ranking manifestations of the same basic type but constrained
by their grammatical environment (see Matthiessen, 1995a: 90–93, on the logical prototype as a
fractal system manifested in different grammatical environments).
In the description of English (and by implication many other languages), the advance Hal-
liday made was to establish a system of taxis, with a tactic distinction between ‘parataxis’ and
‘hypotaxis’, and to distinguish ‘hypotaxis’ from the downranking of clauses or “embedding”
(cf. Matthiessen and Thompson, 1988), and sort out the types of links by which clauses can be
related in his description of the system of logico-semantic type, making a primary distinction
between ‘projection’ and ‘expansion’. Crucially, he set these two systems up as simultaneous
showing that all intersections are possible, though with different probabilities attached to the
intersections (cf. Nesbitt and Plum, 1988; Matthiessen, 2002a, 2006): see Table 5.8.
In Table 5.8, we have also included categories used in traditional descriptions of the grammar
of English (and so also of the grammars of other languages). As the table shows, these descrip-
tions tended not to cover all the possible combinations of parataxis/hypotaxis and of projection/
expansion. And they typically did not distinguish hypotaxis from downranking, treating both
as “subordination” as opposed to “coordination”. These traditional categories have tended to
seep into modern formal treatments so have come to constitute a “main stream” or assumed
default position, and, within SFL, the “Cardiff Grammar” has arguably stayed closer to the
“main stream” not only in the treatment of clause combining but also in other areas of the gram-
mar. It is Halliday who has represented radical descriptive departures from the “main stream”.

313
 Guide to alternatives, questions, issues and debates

Table 5.8 The intersection of the systems of taxis and logico-semantic type with indication of traditional/
current formal treatment (in double quotes)

Approach TAXIS Downranking (embedding)

LOGICO- IFG Parataxis Hypotaxis

SEMANTIC “main stream” “coordination” “subordination”
TYPE
projection IFG quoting reporting [[fact-clauses]]
1 “2, 1 ’2 α “β, α ‘β
“main stream” → “complement clauses”
(as object)
expansion elaborating IFG 1=2 α=β [[restrictive relative
[non-restrictive clauses]]
relative clauses]
“main stream” “apposition” → “relative clauses”
extending IFG 1+2 α+β
“main stream” “coordination”
enhancing IFG 1×2 α×β
“main stream” “adverbial clauses”

Importantly, Halliday’s account of the clause complex has brought out quite clearly the natural
relationship between clause complexing in lexicogrammar and logico-semantic, or rhetorical-
relational, complexing in semantics: see e.g. Matthiessen and Thompson (1988), Matthiessen
(1995a, 2002a), Stuart-Smith (2001).
As just noted, one of Halliday’s descriptive innovations was to treat taxis and logico-
semantic types as independently variable systems – which is an example of a fundamental
descriptive strategy. (In other words, if one tentatively treats systemic variables as independent,
one is much more likely to discover rare or overlooked intersections than if one collapses dis-
tinctions; cf. Section 4.3.3 earlier.)

Notes
1 This theoretical quadri-stratal model of language has now been around for quite a long time, but
it hasn’t yet completely permeated the SFL literature; there are still even fairly recent contributions
around that characterize the model of language as tri-stratal. But “phonetic substance” and “graphic
substance” were already part of the picture presented by Halliday (1961).
2 In personal communication, Halliday called his early use of “context” as the term for semantics silly
or unfortunate. But it is important to note that in some sense he was struggling with the terminol-
ogy at the time and with Firth’s insistence that meaning permeated the system. Compare Halliday’s
(1961) comments on “formal meaning” and “contextual meaning”. In fact, Halliday, McIntosh and
Strevens (1964: 18) offer a version of the diagram, distinguishing between “level (general)” and “level
(specific)”. In their diagram, “context” appears at “level (general)” and is characterized as “relation of
form and situation”, and it corresponds to “semantics” at “level (specific)”.
3 Hyon (1996) is a very helpful contribution, but the fact that she was focussed on the term “genre”
rather than on the phenomenon of functional variation and that she was oriented towards applied
linguistics meant that she excluded the extensive body of work on linguistics on text typology, includ-
ing work in Europe documented in languages other than English (e.g. Werlich, 1975; Gülich and
Raible, 1977) and work in and around Tagmemic Linguistics (e.g. Longacre, 1974, 1996; Longacre
and Hwang, 2012). Text or genre typology has also been important in other areas, e.g. translation
studies. Unfortunately, Hyon’s focus on “genre” has sometimes meant that scholars relying on her
overview have not taken account of important contributions that fall outside the grid of her overview.

314
 Guide to alternatives, questions, issues and debates

4 Martin and Rose (2008: x) specify the origin of the split between “genre” and “register”: “Joan Rothery
and Guenter Plum first came up with the idea of distinguishing register from genre circa 1980–1981.”
5 Martin has pointed out elsewhere that these circles are “co-tangential”.
6 Personal communication, in discussion with Matthiessen.
7 Fawcett (2017) writes: “Meanings are realized in forms: i.e. in items (words and morphemes), syntax,
and intonation or punctuation.” (Elsewhere he glosses it as “wording”.) So “form” here is quite dif-
ferent from the Hjelmslevian notion of form often used in SFL. The pairing of “form” and “mean-
ing” has of course been used widely in linguistics; see e.g. Bybee and Beckner’s (2010) discussion of
constructions.
8 As Fawcett (1980) points out, he had increased the number of “functional components” from six in
Fawcett (1973b).
9 The following account has been adapted from Matthiessen (2018a).
10 “Systems of the English clause: a trial grammar for the Penman text generation project, Information
Sciences Institute, University of Southern California.” See e.g. Matthiessen and Bateman (1991).
11 Outside SFL, Dixon (1994: 18–22) has criticized Halliday’s use of the term “ergative”. In addition,
he makes a distinction between “lexical semantics” and “grammar” as far as ergative patterns are con-
cerned. Noting uses of verbs in Fijian as either “transitive” or “intransitive” – comparable to English
open in he opened the door and the door opened, he comments: “This is entirely a matter of lexical seman-
tics, and does not relate to the grammatical characterisation of a language as ergative or accusative.”
This position does not make sense, of course, in SFL for two reasons: (i) the continuum between
grammar and lexis within lexicogrammar, and (ii) the natural (rather than conventional or arbitrary)
relation between semantics and lexicogrammar. Dixon (1994: xiii) does, however, note that it was
Halliday who characterized Dyirbal patterns as ergative for him: “When I first went out to Australia
to study an indigenous language, in 1963, the word ‘ergative’ wasn’t in my linguistic vocabulary. The
Australian Institute of Aboriginal Studies suggested working in the Cairns Rain Forest region and I
chose Dyirbal for my major focus of study simply because it was the language with the most speakers
(perhaps 100 fluent speakers, at that time). When I returned to London and explained the structure
of Dyirbal to M. A. K. Halliday he told me that these unusual-looking grammatical patterns I had
uncovered were ‘ergative’.”
12 And until they realize that indeterminacy is a positive feature of evolved – and evolving – systems. In
describing the system of process type, we can try to take account of the fact that it embodies change
in progress. For example, when we draw the line between ‘verbal’ and ‘relational’ in the region of ‘say-
ing’ and ‘symbolizing’ (cf. Figure 5–12 below), we should try to take account of the evolution in Eng-
lish of senses of ‘proving’ for a number of verbs of saying such as tell, indicate, show, demonstrate, prove,
suggest (Halliday, 1988a) and the analogous ontogenetic trajectory in secondary education (Christie
and Derewianka, 2008). One way of making sense of this change is precisely to interpret it as the
development of impersonal ‘proving’ modelled in ‘relational’ clauses out of personal saying modelled in
‘verbal’ clauses (cf. the analysis of behavioural research suggests that monolinguals and bilinguals differ in how
they manage within-language phonological competition when listening to language as an ‘identifying relational’
clause of symbolization in Figure 5.12.).

315
 6
SFL TERM GUIDES
(METALANGUAGE)

As with any “metalanguage”, an important part of SFL is its terminological resources – its
scientific or technical terms, including not only technical terms not in common use such
as morpheme, phoneme, grammatics, metafunction, protolanguage and semogenesis but also “techni-
calized” everyday lexical items such as text, context and word. Where necessary, Halliday has
introduced new terms like grammatics and polarity and appliable (linguistics), but on the whole
wherever possible he has adapted existing terms, re-purposing them as it were. This strat-
egy contrasts with the one adopted by Louis Hjelmslev’s approach, as in his “Prolegomena”
(Hjelmslev, 1943): his Glossematics can be difficult to access because of the sheer number
of new terms.
The use of terms in SFL has been characterized as “idiosyncratic” on various occasions,
e.g. in van Dijk’s (2008) critique, but such characterizations are arguably part of the “dis-
missal genre” discussed by Martin (e.g. 1992c),1 and idiosyncrasy is surely in the eye of the
beholder, a reflection of their membership in academic cultures. The challenge of terms in
linguistics as the science of language and of glossing them in language is identified by Hal-
liday (1984a): he notes that grammatical categories in particular are “ineffable”, explaining
that this is a feature of language serving as its own metalanguage. This ineffability is impor-
tant to keep in mind when we struggle with the task of glossing terms like “Theme”, “Sub-
ject”, “Medium”, “present” and “perfective”; it always helps not to treat them in isolation
but in relation to closely agnate terms, e.g. “present” vs. “past” vs. “future” and “perfective”
vs. “imperfective”.2 More generally, terms must be viewed trinocularly as nodes in networks
of relationships; by considering closely agnate terms, we view them from roundabout, but
we also need to take other relationships into consideration along all the relevant semiotic
dimensions.
It is also important to note that as more introductions are published for readers at different
educational levels, the selection of terms will be adjusted to make the texts accessible to the
intended readers. Interestingly, work on “grammatics” in school, even upper primary, has shown
that young students have no problem with technical terms; in fact, they take enthusiastically to
the terminology, gaining analytical skills and a sense of achievement and contributing new terms
themselves (cf. Williams, 2005a; Williams and French, 2016).

DOI: 10.4324/9781315675718-6 316

 SFL term guides (metalanguage)

6.1 SFL terminology

There are a number of accounts of SFL terminology, including those set out in Table 6.1. They
range from lists of terms to extended entries, as in Matthiessen, Teruya and Lam (2010). The
book by de Joia and Stenton (1980) is a very valuable early contribution because it includes
quoted passages from Halliday’s publications; thus it also serves as a term-based index into his
earlier work.
As the table shows, there are now glossaries for a number of the languages in which SFL work
is published, including Chinese, Danish, Norwegian, Swedish and Spanish. In 1994, Halliday
and Matthiessen led a workshop dealing with the development of SFL terminologies in different
languages at ISFC in Ghent, and there have certainly been considerable advances since then.
In addition to published compilations, it is of course possible to extract bilingual or multilin-
gual lists of terms from the growing number of translations of key SFL works into a range of
languages, but it is perhaps even more important to extract terms from original publications in
different languages. It would be possible to compile lists of technical terms also from publications
in e.g. Japanese, Vietnamese, Arabic, Portuguese and Italian.
Interesting issues will arise for each language in which SFL contributions are published. In
English, technical-scientific terms tend to draw on Latin and Greek sources, the Greek ones
tending to be used for more abstract notions such as metaphor (but there are often alternative
terms, like coordination/subordination and parataxis/hypotaxis, and there may be mixtures,
as Halliday has pointed out that metafunction is). There is a certain tendency for Latin and
Greek borrowings to be used as terms in the description of lexicogrammar, and for “native”
terms to be used in the description of semantics. In Japanese, learned vocabulary tends to
draw on Chinese sources, so in presenting systemic functional work in Japanese, scholars will

Table 6.1 Technical terms used in SFL

Nature of glossary Language Reference

terms used in SFL with English de Joia and Stenton (1980)

quoted passages from the
work by M.A.K. Halliday
terms used in SFL: English Halliday and Martin (1981: 337–346);
glossaries Matthiessen (1995a: 772–797); Halliday
(2009b: 229–253; Matthiessen and
Halliday (2009: 92–97)
Spanish Ghio and Fernández (2005: 155–162),
based on Matthiessen and Halliday (2009)
Portuguese Cabral et al. (2021)
Danish (~ English) Andersen, Petersen and Smedegaard
(2001: 338–341)
Danish, Swedish, Norwegian Holmberg and Karlsson (2006: 209–218)
Chinese Zhu and Yan (2001): list of English-
Chinese equivalents
key terms used in SFL with English Matthiessen, Teruya and Lam (2010)
extended entries

317
 SFL term guides (metalanguage)

have a choice between native and Sinitic vocabulary, and they have made different choices,
one principle being that native Japanese terms are more accessible to users outside specialized
academic areas.
Glossaries are typically presented as dictionary views on systemic functional terminology, and
they need to be supplemented by thesaurus views on technical terms – i.e. views that bring
out the lexico-semantic relationships among the terms, representing them as an onomatologi-
cal resource, as discussed by Matthiessen, Teruya and Lam (2010). Halliday (2009b) includes a
network diagram for each of the key terms in his glossary, illustrated here for the term metaphor:
see Figure 6.1, and for function later: Figure 6.2.
The thesaurus view of SFL terms illustrated in Figure 6.1 and Figure 6.2 brings out the fun-
damental point that in SFL technical terms are defined relationally – i.e. by the relationships that
they enter into. These relationships must be viewed trinocularly along the relevant dimensions,
i.e. “from above”, “from roundabout” and “from below”. This is illustrated for:

• metafunctions in Figure 5.9;

• the clause in Figure 7.3.

Thus all the attempts to define the clause (or sentence) as a thing in itself over the decades
need to be replaced by a trinocular view of the relationships that it enters into (for discussion,
see Section 7.2.1). Dictionary glosses are not a substitute for a relational account of the category
named by a given technical terms. This is important to emphasize and then to re-emphasize
because there is always the temptation to argue about dictionary glosses of terms rather than
about the location of the categories they refer to in the overall network of relations.3 One obvi-
ous example is “Theme” and another is “Subject” (cf. Section 5.3.1.3 on debates about Subject).
See further Halliday (1984a) on the “ineffability of grammatical categories”.
In addition to compiling glossaries and thesauri, we can use computational tools to produce
rough lexical profiles of SFL publications – centrally to get a sense of the fields they are con-
cerned with (cf. Benson and Greaves, 1992). As an illustration, we have produced “word clouds”
for a couple of systemic functional publications: see Figure 6.2.

Figure 6.1 Example of lexico-semantic network from Halliday’s (2009b) key terms

318
 SFL term guides (metalanguage)

(a) Halliday (1978) – language as social semiotic

(b) Martin (1992a) – text in context

Figure 6.2 Word clouds for a few SFL publications

The word clouds in Figure 6.2 give an indication of motifs in the diferent publications rep-
resented. For example, in Halliday (1978), “language” stands out as most frequent, in Martin
(1992a) “text” and in Halliday and Matthiessen (2014) “clause”. In contrast, in Fawcett (2000),
“structure” and “syntax” are of high frequency as technical terms. Such word clouds are, of

319
 SFL term guides (metalanguage)

(c) Fawcett (2000) – syntax

(d) Halliday and Matthiessen (2014) – Halliday’s IFG

Figure 6.2 (Continued)

course, very crude measures of motifs; they represent the frequency of orthographic words, not
lexical items, so an orthographic word such as “process” would include both the technical sense
of Process as an element in the transitivity structure of a clause and non-technical senses – which
would likely be brought out by an examination of collocations. Also, of course, word clouds do
not represent collocational patterns, nor do they show changes in frequency from the beginning
to the end of a document.

320
 SFL term guides (metalanguage)

Even if we strive to supplement glossaries with thesaurus views of SFL technical terms – a
very important project, it is still the case that technical terms are only the most exposed aspect
of systemic theory, description and application. We will need to drill deeper to reveal crypto-
grammatical and semantic patterns, and to identify registers that have been common in SFL
and ones that haven’t yet been (or “genres”; cf. Martin, 1992c). It is our impression that the
registerial profile of SFL is unlike that of “mainstream” linguistics; it is more like that of natural
sciences than like that of social sciences based on the profiles Parodi (2010), and his team identi-
fied in four university disciplines. But large-scale investigations of the SFL register profile remain
to be undertaken (one of the methods that could be used being the data-mining approach to
register characterization presented by Teich et al., 2016).
Still, it is certainly the case that because SFL is committed to the development of holistic
theory – a holistic theory of semiotic systems in context in relation to systems of lower orders (cf.
Matthiessen, 2007b, forthcoming d) – and of comprehensive descriptions of particular languages,
systemic functional linguists have needed to use registers (genres) that support such develop-
ments, essentially developments concerned with “building field” in Martin’s (1993b) sense. Con-
sequently, scholars have written many expanded taxonomic reports expounding the resources of
different languages. Such reports are needed both for instructional texts setting out procedures
for undertaking linguistic tasks such as analysing texts or describing systems and for argumentative
expositions and debates concerned with alternative interpretations and representations.

6.2 “Tricky” terms

6.2.1 Theoretical terms

6.2.1.1 Overview the “tricky” theoretical terms
Out of all the SFL terms in common use, about a dozen can be “tricky” in the sense that they
might invite confusion because they are polysemous, being used in more than one sense in SFL,
like system and register, or in linguistics in general, like level.4 We have listed these terms in Table
6.2. For more detail on context, genre and register, see Section 5.2.1.1 earlier in this volume.
Some terms that could have been misunderstood have been replaced, as in the case of Firth’s
exponence being replaced by realization and instantiation since it embodied both senses (Halliday,
1992b). We can also mention the term situation here as a tricky one since it may be used in a
way that obscures the distinction between situation type and context of situation, but if the term is
expanded with type or context of as appropriate, the location along the cline of instantiation will
be clear.5 In the table, the potentially tricky term is listed in the left-most column, followed by
the different uses (senses) of the term, each in a different row, and by subtypes, and then con-
trasting terms and alternative or closely related terms for a given sense, and finally references.
Some of the terms listed in Table 6.2 are “tricky” because they are used in a variety of ways in
linguistics in general, sometimes rather informally. Other terms are tricky because they have come
to be used in different ways within SFL; we have noted context, genre and register, which we discuss
also in Section 5.2.1.1 since their distinct uses are associated with theoretical alternatives within SFL.

6.2.1.2 “Function”
Let us take two of the tricky terms, the ones that appear in the name of SFL: systemic – functional (see
also Asp, 2017, for a discussion of “system” and “function”). Halliday (1985d: 15–16) distinguishes
between two basic senses of “function” – to use Martin’s (1991) terminology, extrinsic functions

321
 Table 6.2 “Tricky” technical terms in SFL

Term Senses (uses) Subtypes Contrasting with Alternative or closely Reference

related term

configuration experiential mode of structure (transitivity configuration) prosody, pulse (wave), constituency Halliday (1979a)
series
Hasan’s term for settings of setting Halliday and Hasan (1985:
field, tenor and mode values: Chapter 4)
“contextual configuration” (CC)
element element of structure class structural function Halliday (1961)
type of phenomenon (in description process, participant, sequence, figure Halliday and Matthiessen
of ideational semantics) circumstance, relator (2006)

SFL term guides (metalanguage)

form organization of strata internal to lexicogrammar, phonology substance (interface internal stratum Halliday (1992c); Martin
language stratum) (1992a); Matthiessen
(2007a, forthcoming d)
level of form: “a systemic functional meaning Fawcett (e.g. 2000); also
322

(SF) model of language that has non-SFL approaches

the two levels of ‘meaning’ and where “meaning and
‘form’” (Fawcett, 2000: xix) form” are treated as a
pair, e.g. Bybee and
Beckner (2010)
function structural function univariate vs. multivariate constituent unit in element of Halliday (1966c)
functions syntagm structure
function in the organization of the microfunction, macrofunction, extrinsic functions Halliday (1975a, 1976a,
meaning potential of language metafunction 1978); Martin (1991)
(intrinsic function)
use of language in context (extrinsic Bühler’s (1934) person- intrinsic functions use (of language) Halliday (1985d); Martin
function) system-based functions (1991)
of language, extended by
Jakobson (1960): referential
(denotative, cognitive), emotive
(expressive), conative, phatic,
metalingual, poetic (see also
Leech, 1974: 40 ff.)
 genre rhetorical mode didactic, persuasive, medium, channel, turn [functional tenor; Halliday (1978)
performative . . . Gregory, 1967;
Martin, 1992a]
stratum within context explanation, report, recount, ideology, register register Martin (1985b, 1992a),
exposition . . . (= functional Gardner (2017); cf. also
variety of Bateman (2008a)
language)
(many uses also outside SFL in Hyon (1996), Swales
the sense of ‘type’, not only in (2009), Malavska
(applied) linguistics but also in (2016)
other disciplines)

SFL term guides (metalanguage)

grammar phenomenon: part of the Dagaare grammar, Bajjika lexis (within morphosyntax Caffarel, Martin and
lexicogrammatical system of a grammar, English grammar lexicogrammar); Matthiessen (2004)
language semantics (within
content plane)
323

study of phenomenon: theory scale-&-category grammar, grammatics Halliday (1984a, 1996a);

and description of part of the systemic grammar, systemic Williams (2004, 2005a)
lexicogrammatical system of a functional grammar,
language Functional Unification
Grammar
in some traditions outside SFL, linguistic system “grammar” in generative
extended to include aspects of the linguistics, e.g.
linguistic system in addition to Chomsky (1965)
grammar (= lexicogrammar)
level stratum: early term for stratum, now e.g. “grammatical level”, stratum “level” used by e.g. Halliday
largely replaced by “stratum” (cf. “contextual level” (1957, 1961, 1966b);
Firth’s “level of analysis” where “stratum” in Reference
the levels were not ordered to Stratificational
stratally) Linguistics: Halliday
(1966a)
roughly equivalent to rank, e.g. in rank
Tagmemic Linguistics
(Continued)
 Table 6.2 (Continued)

Term Senses (uses) Subtypes Contrasting with Alternative or closely Reference

related term

meaning property of semiotic systems in value, life; matter Halliday (1996a), Halliday
general (i.e. 4th-order systems) and Matthiessen
(1999/2006: Ch. 13);
Halliday (2005b)
property of the content plane of expression content Hasan (1985a) shows the

SFL term guides (metalanguage)

language (‘content’) complementarity of
signification and valeur;
Taverniers (2019)
identifies four senses of
“meaning” (shown in
324

her Figure 3.1)

property of the higher of the two [metafunctional modes of wording (content2) content1 Halliday (1979a, 1992c)
content strata (i.e. semantics) meaning:] ideational
meaning, interpersonal
meaning, textual meaning
message the textual semantic correlate of a move, figure (quantum of Halliday and Matthiessen
clause information) (2014: Ch. 3)
semantic unit in semantic rank scale text, rhetorical unit, Hasan (1995b); Cloran
in Hasan’s description; “message message, message (1994); Hasan et al.
semantics” component or text (2007); Cloran, Stuart-
radical Smith and Young
(2007)
move move in an exchange in dialogue Martin (1992a); Eggins
and Slade (1997, 2005)
 (closely related:) the interpersonal according to the system of figure, message Halliday (1984b); Halliday
correlate of a clause speech function: initiating/ and Matthiessen (2014:
responding, giving/ Ch. 4)
demanding, information/
goods-&-services
register functional variety of language (subtypes according to contexts dialectal variety, codal diatypic variety, Halliday, McIntosh and
(variety according to context of of use: mid-region along variety functional Strevens (1964), Hasan
use, diatypic variety) the cline of instantiation variety (1973), Halliday (1978),
between potential and Matthiessen (1993a),

SFL term guides (metalanguage)

instance) Lukin et al. (2008)
context: field, tenor and more, genre Martin (1992a)
roughly equivalent to situation
type (along the cline of
instantiation)
325

phonology, phonetics: vocal register, Abercrombie (1967:

pitch register, register of voice 99–100); Catford
(from music) (1977: 101, 103, 109)
semantic linguistic variation: linguistic codal variation, registerial dialectal variation Halliday (1994d)
variation variation located in the semantic variation
system in the first instance
linguistic variation: codal variation (Bernstein:) elaborated code, registerial variation (as codal variation Williams (2005c); Hasan
[term used by Ruqaiya Hasan, restricted code semantic variation); (2009b)
with reference to William Labov] dialectal variation
system semiotic system, system of meaning primary semiotic system, social system, Halliday and Matthiessen
higher-order semiotic system biological system, (1999/2006); Halliday
physical system (2005b); Matthiessen
(2007b, forthcoming a)
(Continued)
 Table 6.2 (Continued)

Term Senses (uses) Subtypes Contrasting with Alternative or closely Reference

related term

organization at the potential pole of context of culture, semantic subsystem, instance Halliday (1991d,
the cline of instantiation system, lexicogrammatical type, instance 2002a); Matthiessen
system, phonological system (2007b); Halliday and
Matthiessen (1999,
2014)
axis: organization along structure Halliday (1966b, 1976a,
paradigmatic axis ~ system (of 1994a); Matthiessen

SFL term guides (metalanguage)

options, terms) in system network (2021)
word rank on grammatical rank scale (word classes, “parts of clause, group, Halliday (1961); Halliday
speech”) morpheme and Matthiessen (2014)
orthographic word Halliday (1985f)
326

lexical item grammatical item Halliday (1966b)

 SFL term guides (metalanguage)

of language, or uses of language, and intrinsic functions of language, i.e. functions as organizing
principles internal to language. Halliday starts with extrinsic functions, or uses of language:

In the simplest sense, the word “function” can be thought of as a synonym for the word
“use”, so that when we talk about the functions of language, we may mean no more than
the way people use their language, or their languages if they have more than one. Stated in
the most general terms, people do different things with their language; that is, they expect
to achieve by talking and writing, and by listening and reading, a large number of differ-
ent aims and different purposes. We could attempt to list and classify these in some way
or other, and a number of scholars have attempted to do this, hoping to find some fairly
general framework or scheme for classifying the purposes for which people use language.
There are a number of familiar classifications of linguistic functions: for example,
that put forward by Malinowski, which is associated with his work on situation and
meaning referred to earlier. Malinowski (1923) classified the functions of language
into the two broad categories of pragmatic and magical. As an anthropologist, he was
interested in practical or pragmatic uses of language on the one hand, which he further
subdivided into active and narrative, and on the other hand in ritual or magical uses
of language that were associated with ceremonial or religious activities in the culture.
A quite different classification is that associated with the name of the Austrian psychol-
ogist Karl Bühler (1934), who was concerned with the functions of language from the
standpoint not so much of the culture but of the individual. Bühler made the distinction
into expressive language, conative language, and representational language: the expressive
being language that is oriented towards the self, the speaker; the conative being language
that is oriented toward the addressee; and the representational being language that is
oriented towards the rest of reality – that is, anything other than speaker or addressee.
Bühler was applying a conceptual framework inherited from Plato: the distinction
of first person, second person, and third person. This in turn is derived from gram-
mar (its source was in the rhetorical grammar that came before Plato) – based on the
fact that the verbal systems in many European languages (including ancient Greek) are
organised around a category of person, comprising first person, the speaker; second
person, the addressee; and third person, everything else. On this basis, Bühler recog-
nised three functions of language according to their orientation to one or other of the
three persons. His scheme was adopted by the Prague School and later extended by
Roman Jakobson (1960), who added three more functions: the poetic function, ori-
ented towards the message; the transactional function, oriented towards the channel;
and the metalinguistic function, oriented towards the code.

Halliday then adds a couple of other frameworks detailing uses of language – those developed by
Britton and by Morris, and goes on to comment on their activities:

What such scholars were doing was essentially constructing some kind of conceptual
framework in non-linguistic terms, looking at language from outside, and using this
as a grid for interpreting the different ways in which people use language. In all these
interpretations of the functions of language, we can say that function equals use.

Thus the functions he has surveyed up to this point are extrinsic functions, or uses of language
viewed from outside language itself. Halliday then explains why we need to turn to intrinsic
functions, functions that are inherent in the organization of language.

327
 SFL term guides (metalanguage)

But in order to pursue our own investigations, we have to take a further step: a step that
interprets functional variation not just as variation in the use of language, but rather as
something that is built in, as the very foundation, to the organisation of language itself,
and particularly to the organisation of the semantic system.
In other words, function will be interpreted not just as the use of language but as
a fundamental property of language itself, something that is basic to the evolution of
the semantic system. This amounts to saying that the organisation of every natural
language is to be explained in terms of a functional theory.

Embarking on this investigation, Halliday introduces the metafunctions of language one

by one – experiential, interpersonal, logical and textual. But we will stop quoting him at
this point because in our attempt to clarify the diferent ways in which “function” is used,
we have now separated extrinsic functions, or uses of language, from intrinsic functions.
Importantly, extrinsic functions, or uses of language, are of course taken account of in
SFL; they are grounded explicitly in the account of context – the semiotic environment in
which language (and other denotative semiotic systems) operate (see Section 3.1.2 earlier).
The context-based account of the uses of language are much more sophisticated, elaborated
and multidimensional than those ofered by Bühler, Malinowski, Jakobson, Britton and
Morris. And as Halliday (1978) suggests, they can actually be shown to “resonate” with
the functional organization of language itself (a realizational relationship often referred to
as the “[context-] metafunctional hookup hypothesis”; e.g. Halliday, 1994a; Hasan, 2014;
Clarke, 2012).
In referring to the intrinsic functions of language, “function” is of course associated with
metafunction in the first instance, but it also includes the developmental phases of func-
tions leading up to metafunction – microfunction and macrofunction, and the interpre-
tation of an element of structure in a linguistic unit as a structural function. These different
but related senses of “function” are captured by Halliday (2009b: 253), reproduced here as
Figure 6.3.

Figure 6.3 “Function” in Halliday (2009b: 253) – the different but related senses of intrinsic functions

328
 SFL term guides (metalanguage)

Metafunction has also been referred to as “functional component”, as in Halliday (1973) and
also in Fawcett’s (e.g. 1980) alternative interpretation (see Section 5.2.3), but while Halliday
first “discovered” the metafunctions when he observed componential clustering of systems in
system networks and developed his theory of the metafunctional organization of language based
on these “components”, the term “functional component” can arguably lead to an interpreta-
tion of metafunctions in terms that are somewhat too concrete and do not capture the diferent
trinocular views set out in Figure 5.8 (taken from Halliday, 1978). (We write “discovered” the
metafunctions as shorthand for discovered patterns that he interpreted by means of his theory
of metafunction.)
In any case, “functional” in “systemic functional linguistics” means ‘metafunctional’ in the
first instance, but it really indicates that function is pervasive in the SFL engagement with lan-
guage. Interestingly, a term has emerged to cover approaches to language that are in one way
or another based on text – viz. “usage-based theory”, a term that was introduced by Ronald
Langacker. Bybee and Beckner (2010: 827) characterize usage-based theory as follows:

Usage-based theory takes language to be an embodied and social human behavior and
seeks explanations in that context. As the name indicates, this theoretical perspective
incorporates the basic insight that usage has an effect on linguistic structure. It thus
contrasts with the generative paradigm’s focus on competence to the exclusion of per-
formance and rather looks to evidence from usage for the understanding of the cogni-
tive organization of language. Thus usage patterns, frequency of occurrence, variation,
and change are all taken to provide direct evidence about cognitive representation. No
relevant methods for gaining evidence about language are excluded; studies of cor-
pora, large and small, diachronic data, psycholinguistic experiments, cross-linguistic
comparison and child language development all provide essential data for constructing
a comprehensive theory of language.

In discussing approaches of this kind, they mention West-Coast Functionalists (including

linguists working on grammaticalization), Variationists, John Sinclair as a representative of
corpus linguistics, Construction Grammarians, and work on “language acquisition”. One
linguist they do not mention is M.A.K. Halliday, but he is in good company: they do not
mention any other systemic functional linguists either (as it happens, their chapter appears
immediately after Cafarel, 2010, in the same handbook). And yet, their characterization of
“usage-based theory” would seem to apply to a great deal of systemic functional work since
the 1960s.

6.2.1.3 “System”
What about the other technical term in the name “SFL” – “systemic”? Like “functional” it is also
“tricky”. One reason is superficial but nonetheless significant: people often mistake “systemic”
for “systematic”, so they refer to “systematic (functional) linguistics”.6 This is understandable;
while “systemic” is already more or less technical, “systematic” is not – it is widely used, and
can of course be readily associated with scientific activities. But while all modern academic
approaches to language are (hopefully!) systematic, only very few are “systemic” in the central
sense, which can be related to Firth’s “system-structure theory” – which was pre-systemic but
an important source, where the paradigmatic (“system”) and the syntagmatic (“structure”) were
given equal status as axes of organization.

329
 SFL term guides (metalanguage)

Giving priority to the paradigmatic axis as systemic organization was one of the fundamental
changes brought about by Halliday in the 1960s in what turned out to be the development of
SFL (a key publication being Halliday, 1966a, and also – as a short but powerful example, Hal-
liday, 1969). This is “system” in the sense of “the organizing concept for modelling paradigmatic
relations in language” (see the following).
In his documentation of key terms in SFL, Halliday (2009b: 232–233) distinguishes two
senses of “system”:

System (in the sense of system1) is the organizing concept for modelling paradigmatic
relations in language. A paradigm is a set of forms which share a common environ-
ment, like the set of English finite verbal operators can/could/may/might/shall/should/
will/would/must/ought/need/dare. Underlying these are several systems of contrasting
features (“terms”), such as value: high (must . . .)/median (will . . .)/low (may . . .);
orientation: subjective (must . . .)/objective (certainly . . .), and others, together making
up the system network of modality; and also features from the system network of tense.
...
The term “system” is extended to be applied to language as a whole (the linguis-
tic system), in the opposition (language as) system/(language as) text. System2 and
text are related by instantiation: the text is the observable instance of the underlying
systemic potential.
In this sense, system relates to general systems theory. Language can be characterized
as a complex dynamic system, one that persists through constant change in interaction
with its (eco-social) environment. It belongs to the class of semiotic systems (systems
of meaning), in contrast with systems of other kinds, physical, biological and social.
System1 and system2 are in fact the same concept, though operating on a differ-
ent scale. In both cases the system is the (representation of the) potential that inheres
in a given set of phenomena. These systems have all evolved; evolved systems con-
trast, in turn, with designed systems, which have been brought into being to explain
(and sometimes to control) some realm of human experience. A scientific theory is a
designed system of this kind.
Usually, “system” is used to cover system-&-process: the system, together with the
processes that derive from (or “realize”) it in real or virtual time.

The diferent but related senses of “system” in SFL are set out in Figure 6.4. This figure draws
on Halliday’s (2009b) two diagrams for his distinction between system1 and system2, but while
it incorporates his two diagrams in one, it leaves out some details.
So to return to the name “systemic functional linguistics”, we can relate SFL to other func-
tional traditions that have emerged in the scientific (or at least scholarly) engagement with
language, including European continental functionalist traditions and US American West-Coast
functionalism. See further Section 2.6 earlier.
In terms of system in Halliday’s sense of system1, SFL can of course be related to other tradi-
tions where the paradigmatic axis has been given a status equal to that of the syntagmatic axis.
Interestingly, there are few other systemic traditions – if any, in fact – in the sense of engage-
ments with language where the paradigmatic axis of organization is given priority. While Euro-
pean (as opposed to US American) structuralists emphasized paradigmatic organization – as did
Firth in his system-structure theory and (in the US context, influenced by Hjelmslev) Lamb
in his Stratificational Linguistics, SFL is probably unique in giving priority to the paradigmatic
or systemic mode of organization (Halliday, 1966a), which can be derived directly from the

330
 SFL term guides (metalanguage)

Figure 6.4 The different but related senses of “system” in SFL

conception of language as a resource for making meaning (cf. Halliday, 1977a). In this context,
we can note Heller and Macris (1967), possibly one of the few studies outside SFL that really
explores the power of paradigmatic modelling in different domains.7
In terms of system in Halliday’s sense of system2, SFL can be related to systems thinking in
other disciplines and as a transdisciplinary motif (cf. Halliday and Matthiessen, 1999/2006;
Cléirigh, 1998; Matthiessen, 2009b), ranging from early initiatives in the mid-20th century by
Ludwig von Bertalanffy, who coined the term “General Systems Theory” (GST), to current
research into complex adaptive systems, network science and other forms of holistic approaches
to phenomena in different realms. See Section 1.2 and Section 2.5 earlier.

6.2.1.4 “Meaning”
The term “meaning” is of course notoriously “tricky”, and Ogden and Richard’s (1923) classic
The Meaning of Meaning is very well known (see also e.g. Lyons, 1977: 1–4). Malinowski related
meaning to use, an approach that Wittgenstein later became famous for suggesting (see Samp-
son, 1980), and Firth saw meaning as dispersed throughout all the levels of linguistic analysis.
Firth’s position has often puzzled linguists, but it could be illuminated through an exploration of
“meaning” in SFL. Here we will just make the following observations.
(1) In the ordered typology of systems operating in different phenomenal realms, meaning is
the property of systems of the 4th order – semiotic systems (see Table 1.1 earlier). It is thus the
defining property of semiotic systems, distinguishing them from social systems, and it is made
possible through the emergence of stratification of semiotic systems into content and expres-
sion, overlaying the organization of social systems, which are characterized by “value” (or valeur).
(2) In higher-order semiotic systems – language being the prototype – the content plane is fur-
ther stratified into two content strata, semantics and lexicogrammar. The relationship between
them is natural rather than conventional (or “arbitrary”), so the lexicogrammatical construction

331
 SFL term guides (metalanguage)

of content is semantically transparent. These two content strata are characterized as “meaning”,
now within semantics, and “wording”, within the domain of lexicogrammar. The split of con-
tent into two content strata makes metaphor possible (both grammatical metaphor and lexical
metaphor). (3) Of the two content strata, semantics is the interface to what lies beyond language
as far as content is concerned – to meaning in context, to meaning in denotative semiotic
systems other than language and to meaning in bio-semiotic systems. (4) With the emergence
of higher-order semiotic systems, meaning is transformed into simultaneous metafunctional
modes of meaning – ideational, interpersonal and textual; and processes of transforming what
lies outside language into meaning are metafunctionally differentiated – ideational: construing
experience as meaning; interpersonal: enacting roles and relations as meaning (with the textual
metafunction enabling these two modes of meaning as a flow of text). (5) Meaning thus perme-
ates language and other semiotic systems; they are organized to carry or create meaning. This
is brought out more clearly when stratification is interpreted in terms of metaredundancy (Hal-
liday, 1992c).
Three of these observations are represented diagrammatically in Figure 6.5.

6.2.2 Descriptive terms

The tricky terms we have listed and discussed are all theoretical in nature. However, in descrip-
tions there are also tricky terms, perhaps largely because of differences in use among linguistic
traditions. There are quite a number of them – including very common ones such as “topic”,
“theme”, “subject”, “mood”, “goal” and “voice” – but we will just discuss a few that are very
general, and even on the borderline between theory and description: viz. “text” and “discourse”
and “clause” and “sentence”.

Figure 6.5 Meaning as (1) property of 4th-order systems, semiotic, systems in general, and (2) as property
of the semantic systems within the stratal organization of language – (3) in relation to meaning
in other (quasi-) semiotic systems

332
 SFL term guides (metalanguage)

6.2.2.1 “Text” and “discourse”

If we survey different frameworks and approaches in linguistics and in discourse analysis, we will
find a number of uses of “text” and “discourse”. In the history of the study of language since
the 1960s, we can note that the term used in continental Europe, text linguistics, has tended to be
replaced by discourse analysis (Webster, 2015c). From the point of view of SFL, neither denotes a
distinct branch within linguistics since the engagement with text, or discourse, is part and par-
cel of the engagement with language. However, in certain traditions, the two terms have been
used to refer to different phenomena, as in Fairclough’s Critical Discourse Analysis, discussed
in Section 2.6.1.3 (cf. also e.g. Kress, 2012, on multimodal discourse analysis). In SFL, the two
terms do not refer to different phenomena but may reflect different views on one and the same
phenomenon, according to Halliday (2011: 55):

Text, discourse, and discourse analysis

Some while back I was asked about the terms ‘discourse’ and ‘text’: were they the
same thing, and if not, how would I distinguish between them? They refer to the same
entity, I suggested; but looked at from different points of view. ‘Text’ is discourse that
is being viewed as a linguistic process (hence ‘texts’ are pieces of language), while ‘dis-
course’ is text that is being viewed as a socio-cultural process (so ‘discourses’ are kinds
of language) (and cf. Halliday, 2008b: 77–78). This means that ‘discourse’ is likely to
refer to texts of more than minimal length; apart from that, any passage of wording may
be referred to in either way.

A “text” is language functioning in context (see e.g. Halliday and Hasan, 1976; Halliday, 1981).
In terms of stratification, it is located within semantics (unlike approaches in linguistics such as
Tagmemics treating texts as “macro-sentences” in syntax; Halliday, 1981), in terms of composi-
tion within semantics, it is taken to be the highest rank (cf. Section 5.2.5), and in terms of the
cline of instantiation, it is located at the instance pole of the cline: see Figure 6.6 (and for more
detail, see Figure 4.13). Text is thus like all other categories in systemic functional linguistics: it
is not defined as a “thing in itself ” but rather as a node in a network of relations along diferent

Figure 6.6 Text, defined relationally by reference to its location in terms of the hierarchy of stratification as
language functioning in context of situation at the semantic stratum, the cline of instantiation
as instance of the meaning potentially, and the semantic rank scale as the highest-ranking unit

333
 SFL term guides (metalanguage)

semiotic dimensions. If text is approached “from above”, from context of situation, it might be
referred to as “discourse”. In some approaches outside SFL, it sometimes seems that the combi-
nation of context of situation and text is called “discourse”.

6.2.2.2 “Clause” and “sentence”

As happens in many disciplines, the same terms are used in different ways by different scholars
during different periods of time and by scholars using different frameworks. In accounts of
grammar, there has been variation in the use of two central terms, “clause” and “sentence”. Tra-
ditional grammars tend to offer fairly confused accounts, using terms like “complex sentences”
and “compound sentences” to compound the confusion about this complex area.
One of the contributions made by M.A.K. Halliday is to clear up the terminological confu-
sion by using “clause” and “sentence” in consistent ways according to clear definitions – as in
Chapter 1 of IFG: cf. Figure 6.7. We have summarized the uses of “clause complex”, “clause”,
“sentence” and “sub-sentence” in Table 6.3. The use of these terms, set out in the first edition
of Halliday’s IFG, has remained stable in SFL. The earlier use of the term “sentence” as a rank
above that of the clause has been abandoned, as Halliday realized that there was no need in the
description of English to posit a unit ranking above the clause: all grammatical patterns extend-
ing beyond the clause can be handled by the account of the clause complex, i.e. the account of
tactic combinations of clauses forming complexes.8
A clause complex realizes a semantic sequence, and, in writing, it is in turn realized by an
orthographic sentence, and, in speech, by a particular sequence of tone groups, as illustrated in
Figure 6.8.

Table 6.3 Clause complex, clause and sentence

Grammar Graphology
[wording] [writing]
“bigger” clause complex sentence (i.e. orthographic sentence)
“smaller” clause sub-sentence (if marked by punctuation)

Figure 6.7 Diagram from IFG showing the relationship between units and unit complexes in (1) grammar,
(2) graphology, (3) phonology and (4) poetry

334
 SFL term guides (metalanguage)

Figure 6.8 Clause complexes (grammar) and realized by sentences (graphology – i.e. orthographic sentences)

Notes
1 There is no shortage of such dismissals, but let us refer to one in the context of applied linguistics [AL],
since SFL has been seen to be successful and thus become a target. In his excellent history of recent
applied linguistics, de Bot (2015) quotes William Grabe: “On the basis of comments made by the infor-
mants, a spectacular growth of SFG in AL is not expected, though Tim McNamara feels that the United
States is ready for SFG. William Grabe disagrees: ‘SFG is not the solution. There is not enough empirical
evidence. The theory is arcane, the terminology complex and the texts are often painful to read.’ When
asked why Halliday never took off in the United States, he remarked jokingly: ‘Because he moved to
Australia!’” The combination of Grabe’s comments on SFG and attempt at a joke is very telling, and
characteristic of dismissal strategies, although surely more familiar from political discourse than academic
discourse. And degree of reading pain (“painful to read”) is certainly in the eyes of the reader. In contrast,
to William Grabe, Tim McNamara has throughout the years been a thoughtful non-SFL applied linguist
ready to engage in constructive dialogue.
2 For users familiar with more than one language, it can be very helpful to check terms in the other
language(s) they know. For example, one source of difficulty in English is that the terms for the primary
tenses that are in common use are the same as the terms for the temporal notions. Thus the past tense
refers to past time, which can be a source of confusion. In other languages, there may be different sets of
terms. For example, in Swedish, past time is dåtid and the tense is preteritum (cf. occasional uses in English
of the term preterit); and future time is framtid but the tense is futurum.
3 In a way, this is part of the same picture as the temptation to approach interpersonal patterns in dialogic
interaction through the folk theory embodied in the ideational construal of speech events: see Section
5.3.1.1.
4 To these terms, we could add Sydney in the characterization of strands or streams within SFL. Robin
Fawcett has contrasted the “Sydney Grammar” with his own “Cardiff Grammar”; by “Sydney Gram-
mar” he refers to the theoretical framework presented in successive versions of Halliday’s Introduction to
Functional Grammar (cf. Matthiessen, 2007b) and the description of English used as an illustration. Later
scholars outside SFL started referring to the work by J.R. Martin and his group as the “Sydney School”,

335
 SFL term guides (metalanguage)

and they have adopted this term in many publications – now also simply the “Sydney Architecture” in
Tann (2017). This use of “Sydney” is associated with the “Genre Model”, the stratified account of con-
text, the description of appraisal and other related contributions, but it excludes other work in Sydney,
notably the ongoing development of the Halliday-Hasan strand.
5 As noted in Section 1.3.3, the relationship between the functional parameters of context, field, tenor and
mode, and the metafunctions of language, ideational, interpersonal and textual, extends along the cline
of instantiation from the potential pole to the instance pole.
6 This seems to have occurred regularly over the years. For example, at the Systemic Functional event in
Canterbury in 1986, there was a sign pointing to the venue saying “systematic”. Search engines will turn
up examples of “systematic” instead of “systemic”. In fact, in a recent conference paper, Jonathan Web-
ster addressed the problem head-on: “Why Systemic (not Systematic) Functional Linguistics?”: https://
scholars.cityu.edu.hk/en/publications/publication(490c1428-8c67-4ece-940c-82274d367336).html
7 Heller and Macris’s (1967: 12) concept of parameter is comparable to that of a system: “A linguistic
parameter, as defined in this monograph, is any variable which differentiates two categories in a lan-
guage either (1) on the functional plane (such a parameter would be a functional distinction) or (2) on
the signalling plane (such a parameter would be a manifesting distinction). We avoid the term distinctive
feature because recent usage has generally restricted its meaning to the phonological, and occasionally to
the morphological level.”
8 Some other linguists operating with something like a rank scale have taken the position that “sentence”
is a rank, or “level”, above the clause, Longacre being a notable proponent of this view, as in Longacre
(1970, 1985).

336
 7
ANALYSIS GUIDE
English lexicogrammar

In this chapter, we provide resources compiled to help analysts as they analyse texts in English by
means of the description of the lexicogrammar of English as presented in Halliday’s IFG (most
recent version: Halliday and Matthiessen, 2014), introductions to it (e.g. Thompson, 2014), the
IFG workbook (Martin, Matthiessen and Painter, 2010/1997) and Matthiessen (1995a). They
include templates and charts guiding analysts in sorting out the patterns of grammatical units –
clauses in particular, and tabulated lists of items – both grammatical items (“function words”)
and lexical items (“content words”) that can prove difficult in analysis, often because they have
evolved two or more senses.

7.1 Lexicogrammatical analysis of text

While analysis of texts in contexts can involve any of the strata of language in context and there
are partial descriptions of all the strata (see: context: Table 3.2, Table 3.3, Table 3.4; semantics:
Table 3.6), publications providing guides to text analysis have on the whole focussed on lexi-
cogrammar, an early contribution being Halliday (1985b). The analysis of texts always involves
relating instantial patterns in texts to the description of the system of lexicogrammar. It is thus
a move along the cline of instantiation, between text (instance) and system (potential). This is
one fundamental way in which systemic text analysis differs from commentary on texts in the
tradition of explication de texte or a close reading of texts.

7.1.1 Workbooks and models

The information we provide here complements other guides in the SFL literature supporting
texts analysts. Most lexicogrammatical analysis of English texts draws on one of the editions
of Halliday’s Introduction to Functional Grammar, the most recent edition being Halliday and
Matthiessen (2014) – “IFG4”. Two editions of a workbook designed to accompany Halli-
day’s Introduction to Functional Grammar, Martin, Matthiessen and Painter (2010/1997) include
“trouble shooting” sections dealing with problems that are known to arise in the analysis of
texts. We have listed the trouble shooting sections in Table 7.1, and added references to pas-
sages in IFG4 that provide trouble shooting kind of help, and also references to our material
in this chapter.

337 DOI: 10.4324/9781315675718-7

 Analysis guide

Table 7.1 Overview of “trouble shooting” sections in Deploying Functional Grammar (DFG), Martin,
Matthiessen and Painter (2010), with additional references to IFG 4 and the present book

DFG IFG 4 reference Section, table in

this Guide
Rank Metafunction Section Section, pages

clause textual § 3.1 Picking out the § 1.1 Text and grammar § 7.2.4
unit of analysis grammatical, graphological & Thematic
phonological constituency analysis
Figure 7.2
Chapter 2 § 3.2 Identifying the
Theme
§ 3.3 Identifying the § 3.4 Textual, interpersonal and
metafunction topical Themes
§ 3.4 Subject it as § 4.7 Clause as Subject
unmarked topical
Theme
§ 3.5 There’s a . . .
‘existential’ clauses
§ 3.6 Recognising § 3.2 Group/phrase complexes Figure 7.4
longer Theme as Theme; thematic equatives
units
§ 3.7 Hypotactic § 7.6 The clause complex as
clause as Theme textual domain
interpersonal § 3.1 The unit of § 7.2.5 Modal
analysis analysis
Chapter 3 § 3.2 Identifying § 4.2 The Mood element
structural elements § 4.3 Other elements of Mood
structure
§ 3.3 Ambiguous pp. 170–172: freedom: free/ Figure 7.9
mood type bound clauses
§ 3.4 Ambiguous
polarity
experiential § 3.1 A topology of § 7.2.6
processes Transitivity
analysis
Chapter 4 § 3.2 General probes p. 354, Table 5–45 Criteria § 7.3.2 Verbs
helpful in analysing for distinguishing process
for transitivity types
§ 3.3 One process or § 8.9 Logical organization:
two? complexes at clause and group/
phrase structure, and groups
§ 3.4 Material clauses § 5.2 Material clauses:
processes of
doing-&-happening
§ 3.5 Mental clauses § 5.3 Mental clauses: processes
of sensing

338
 Analysis guide

DFG IFG 4 reference Section, table in

this Guide
Rank Metafunction Section Section, pages

§ 3.6 Relational § 5.4 Relational clauses:

clauses processes of being-&-having
pp. 267–268: ‘attributive’ vs.
‘identifying’
pp. 274–276: ‘relational’ vs.
‘mental’
§ 3.7 Verbal clauses § 5.5.2 Verbal clauses
§ 3.8 Behavioural § 5.5.1 Behavioural clauses
clauses
§ 3.9 Discriminating § 5.6 Circumstantial elements
circumstance type Table 5–28 Types of
circumstantial element
§ 3.10 Distinguishing p. 158 types of Adjunct; §
between 5.6.3.2 Some difficulties in
circumstances and identifying circumstantial
other elements elements
group/ (all metafunctions) § 3.1 Identifying p. 363, Table 6–1 Group and
phrase Chapter 5 groups/phrases phrase classes in relation to
clause function
§ 3.2 Assigning a § 6.6 Word classes and group
function to a word functions
in a group Table 6–19 Word classes and
their typical functions in
groups
§ 3.3 Different § 7.3.4
meanings for Prepositions
prepositions
§ 3.4 Assigning a
function to a phrase
or a clause: ranking
or embedded?

When undertaking lexicogrammatical analysis of text, it is also very helpful to consult reliable
examples of such analysis in the SFL literature, including (see also Table 4.9 earlier):

• Halliday (1997) is an account of text as semantic choice, but it includes a lexicogrammatical

analysis of a modern fable, Thurber’s The Lover and his Lass. The description referenced in
the analysis is somewhat different from that presented in Halliday (1985a), but the analysis
still serves as a very useful model, including details like the textual analysis of verbal quoting
clauses following quoted clauses.
• Halliday (1985a), IFG, includes an appendix with a detailed analysis of the “Silver Text”,
related directly to the description of the lexicogrammar of English in the book, with a

339
 Analysis guide

characterization of the context of the text. The appendix is also part of IFG2, but had to be
removed from subsequent editions due to page constraints.
• Halliday (1985b) provides an example of the lexicogrammatical analysis of a conversation
between an adult and children. The analysis covers context (pp. 52–53), all metafunctions
at clause rank (pp. 32–46) and group/phrase analysis (pp. 45–49), including the textual pat-
terns of cohesion (pp. 49–51) and grammatical metaphor (pp. 51–52), and intonation and
rhythm (pp. 31–32).
• Hasan (1985b) provides examples of the analysis of verbal art, including poetry. She presents
a scaler of potency of participant roles in the transitivity structure of clauses, “-er” roles and
“-ed” roles.
• Halliday (1992a) provides a fairly comprehensive lexicogrammatical analysis of a US American
fund-raising letter, the Zero Population Growth text, including lexical cohesion, lexical clichés,
grammatical metaphor and an intonation analysis of a reading of the text. This example is valu-
able also because it can be viewed together with a couple of other systemic functional analyses
of the text, Martin’s (1992b) analysis of the text as a macro-proposal, and Benson and Greaves’
(1992) analysis of the text in terms of collocation and field, and also an RST analysis by Mann,
Matthiessen and Thompson (1992) that can be interpreted as a logico-semantic analysis of the
text (cf. Matthiessen, forthcoming b). In addition, Mann and Thompson (1992) include analyses
by scholars representing a number of other linguistic frameworks, including Tagmemics.
• Halliday (1994b) is also a contribution to a volume dealing with one text approached from
different theoretical vantage points, Grimshaw’s (1994) Multiple Analysis Project (MAP).
Halliday’s contribution includes prosodic transcription (pp. 200–206), tabulation of meta-
functional clause analysis (pp. 207–229): taxis; theme, mood and modality, transitivity. An
excerpt from his presentation of the analysis in tabular form is given in Table 7.2.
• Matthiessen (1995a) provides an overview of the lexicogrammar of English based on the
systemic description of the grammar: system networks are used to map the grammar, and the
description is organized in terms of system networks located within the function-rank matrix.
Each system, with its systemic terms, is characterized, and each term is illustrated in use by a
range of authentic examples. All the systems are also included in a tabular index at the end of
the book. Analysed texts are presented for each major system, and there is an appendix with
texts analysed in terms of multiple systems. The book also includes typological outlooks on
languages other than English.

The 2nd volume of Halliday’s collected works includes the examples of lexicogrammatical text
analysis by him except for the “Silver Text” in Halliday (1985a) and the dialogue analysed in
Halliday (b). The final chapter of Martin, Matthiessen and Painter (2010) is a guide to lexico-
grammatical text analysis. There are, of course, many more lexicogrammatical analyses of texts
of many diferent kinds, but the ones we have listed can be treated as model analyses, with accu-
rate analyses consistent with the description in Halliday’s IFG.

7.1.2 Annotation, presentation and tools

7.1.2.1 Boundary annotation (chunking)

In preparing a text for lexicogrammatical analysis, one can begin by annotating is in terms
of clause boundaries, indicating boundaries between both ranking clauses and downranked
(embedded) ones. Here is a passage of text annotated for clause boundaries (for the conventions,
see Section 9.3):

340
 Table 7.2 Example of tabular presentation of analysis, adapted from Halliday (1994b: 207–229)

Turn Speaker Clause Clause Rankshift Theme MOOD POLARITY MODALITY TRANSITIVITY
complex textual interpersonal topical Modalization Modulation PROCESS PROCESS Congruent
TYPE TYPE
Value Orientation Value Orientation

1 A 1 so you declar pos vbl

Analysis guide
2 J 2 minor pos
3 A [3] is that int: pos rel attr/int
341

y/n
3 that decl neg rel attr/int ment:
cog
4 J 4 I decl pos bhv
5 S 5 [what] what int: pos median subj/exp rel attr/poss ment:
wh cog
6 A 6 a well I guess I decl low subj/imp mat vbl
7 ‘b that I decl
 Analysis guide

||| There was once a velveteen rabbit. ||| He was fat and bunchy, || his coat was spotted
brown and white, || and his ears were lined with pink sateen. ||| On Christmas morn-
ing, when he sat wedged in the top of the Boy’s stocking, with a sprig of holly between
his paws, || the effect was charming. |||
||| For at least two hours the Boy loved him, || and then, in the excitement of [[looking
at all the new presents]] the Velveteen Rabbit was forgotten. |||
||| For a long time he lived in the nursery. ||| He was naturally shy, || and some of the more
expensive toys snubbed him. |||
||| The mechanical toys were very superior, || and pretended || they were real. ||| The
model boat caught the tone || and [0] referred to the rigging in technical terms. ||| Even the
jointed wooden lion put on airs. ||| The only person [[who was kind to him at all]] was the
Skin Horse, || who had lived longer in the nursery than any of the others. |||
||| “What is REAL?” || the Rabbit asked the Skin Horse one day. ||| “Does it mean [[having
things [[that buzz inside you]] and a stick-out handle]]?” |||
|||“Real isn’t [[how you are made]],” || said the Skin Horse. ||| “It’s a thing [[that happens to
you]]. ||| When a child loves you for a long, long time, || you become Real. ||| Generally, by
the time you are Real, || most of your hair has been loved off, || and you get very shabby.” |||
||| “I suppose you are Real?” || said the Rabbit. |||
||| “The Boy’s uncle made me Real many years ago,” || said the Skin Horse. ||| Once you
are Real, || it lasts for always. |||

Key
||| = boundary between (ranking) clause complexes
|| = boundary between (ranking) clauses
[[[ ]]] = boundaries (left and right) between rankshifted clause complexes
[[ ]] = boundaries (left and right) between rankshifted clauses

When a text has been marked up in this way for boundaries, it can be converted into a format
for analysis in some software tool, as in SysFan (e.g. Wu, 2000). Alternatively, the text can be
entered using such a format directly, as when O’Donnell’s UAM Tools are used. If the text is
to be analysed in a word processing program or a spread sheet application, it can be converted
into a table where each unit of analysis such as a clause is represented by a row (as in Table 7.2).

7.1.2.2 Analysis representation

In publications, the lexicogrammatical analysis is presented either in the form of box diagrams
or in the form of tabular displays, both being illustrated in IFG4.
(i) Box diagrams can be used to represent the univariate analysis of complexes of units and the
multivariate analysis of simple units. Each row represents a distinct strand of analysis, either function
structure or syntagm (sequence of classes), as illustrated in Figure 7.1. Box diagrams have the advan-
tage that they represent all the strands of analysis of a complex of units or of a unit together so that it
is easy to see how they relate to one another. However, they do not accommodate systemic analysis
as easily although the systemic analysis can be included in a column to the left of the first structural
element, and they do not make it easy to detect patterns in a text emerging complex by complex,
unit by unit, so tabular displays can be used as an alternative or complement.
(ii) Tabular displays are usually designed with column headings representing elements
of structure and names of systems and rows representing units of analysis such as clauses, as

342
 Analysis guide

(a) clause complex (univariate)

Go off the San Diego circle right and turn right onto going about one-half
Freeway at the Wilshire down the Wilshire, mile to San Vicente
Boulevard-West offramp, offramp, westbound, Boulevard.

1 ×2 × 3α ×3×β

clause: free clause: free clause: free clause: bound

(b) clause (multivariate)

There ’s a sunny little country south of Mexico

Theme Rheme

Subject Finite Complement Adjunct

Mood Residue

Process Existent Place

nominal group verbal group nominal group prepositional phrase

Figure 7.1 Box diagrams representing function structures and syntagms (sequences of classes)

illustrated by Table 7.2, adapted from Halliday (1994b). Other examples can be found in vari-
ous publications, including Halliday and Matthiessen (2014), Matthiessen (1995a). Such tabular
displays can be represented using tables in a word processing program or spreadsheets.
They can thus be used to record systemic analysis as well as structural analysis. Unlike box
diagrams, tabular displays are fairly variable in format and are adapted to the needs of particu-
lar analytic tasks. They help analysts scan for patterns, and even if they start life as tables in a
word processing program such as MS word, they can easily be saved as tab-separated texts, and
imported into a spreadsheet program or a database program (see e.g. Wu, 2009). In practical
terms, it will often make sense to start with a spreadsheet or records/lists in a database since they
are not limited by the size of pages in word processing programs and can later be exported in
part for inclusion in articles, chapters or books.

7.1.2.3 Tools, standards

We have mentioned that lexicogrammatical analysis can be recorded in tables, spreadsheets
or databases, but there are more powerful options available. The lexicogrammatical analysis
may be supported by “coding” tools designed to support systemic functional analysis such as
SysFan (Wu, 2000), UAM Corpus Tools (O’Donnell, 2012, 2017: 565–566; Sharoff, 2017:
541; Bateman et al., 2019: 564) or general linguistic coding tools such as ELAN (Wittenburg
et al., 2006) and Anvil, designed to support “multimodal” analysis (see Bateman, Wildfeuer
and Hiippala, 2017: 156–157). Two important general considerations nowadays are that the
analytical annotation should be separable from the text being analysed and should be repre-
sented, or exportable, in a notation such as XML that can be used by various tools. This will

343
 Analysis guide

ensure that the analysed texts do not get “trapped” in a proprietary representation within a
commercial software application.
The notation used is important also to ensure that analysed text can be added to repositories
and shared by members of the community. In other words, the increasing emphasis on data
sharing by scientific bodies and funding agencies applies also to the sharing of annotated data
produced by analysis. Such annotated data are usually produced by larger scale projects than has
typically been possible in SFL, even efforts involving research students from different institutions
contributing their data and their annotated data over long periods of time. If the annotation is
standardized, this will also support meta-analysis of the kind that has been practiced for a long
time now in medical and social science, and been introduced into applied linguistics by John
Norris and Lourdes Ortega (Norris and Ortega, 2006; Ortega, 2015) and into SFL by Winfred
Xuan and his collaborators (e.g. Mwinlaaru and Xuan, 2016; Xuan and Chen, 2019).

7.1.3 The choice of what to include in the analysis

Systems of choices relevant to any kind of text analysis are set out in Figure 4.14 and character-
ized briefly in Table 4.10; these options are relevant to text analysis undertaken within any of the
stratal subsystems of language in context. As far as the system of lexicogrammar is concerned,
the selection of lexicogrammatical systems to focus on in the course of the text analysis can be
made in consultation with the synoptic overview provided by a lexicogrammatical function-
rank matrix, as illustrated for English in Table 3.9. Such matrices show the systems that make
up the lexicogrammatical resources of a language, organized in terms of metafunction (column
headings) and rank and primary class (row headings).
Most often, it makes sense to start with systems at clause rank. Sometimes analysts will have
some specific objective that can be pursued by analysis at a lower rank, like the construal of
process time by means of tense at group rank or the enactment of assessment by means of deriva-
tion at word rank. However, it usually turns out that starting at clause rank will ensure that no
lexicogrammatical considerations are left out of the analysis since this is the ecological approach
to lexicogrammatical analysis – starting with the most extensive grammatical environment, one
that is also the grammatical “gateway” to text as a semantic unit. For example, time is construed
not only through tense at group rank by also by means of phase in verbal group complexing and
temporal circumstantiation at clause rank. And at clause rank, time is not only construed but also
enacted through mood Adjuncts (e.g. still, already) and presented through conjunctive Adjuncts
(e.g. meanwhile, later).
So starting at clause rank will ensure a maximally informed selection of systems to consider –
of course against the background of semantic and contextual considerations. But there is another
major choice to make – having to do with the metafunctions represented in Table 3.10. Should
the selection include logical systems, experiential systems, interpersonal systems and/or textual
systems? There is always a temptation to start with a selection of systems that are – or appear to
be! – easy in terms of analysis, favourites being theme and appraisal (if the latter is approached
from the vantage point of lexicogrammar as interpersonal lexis). But while doing a thematic
analysis of a text is always a good way of getting acquainted with it – even as a preparation for
carrying out a semantic analysis, it is helpful to consider if there are criteria that can inform the
selection of systems to focus on in the analysis.
Such criteria can usually be derived “from above”. If the objectives of the analysis have been
articulated with reference to the functional parameters of context – that is, if they can be speci-
fied in terms of field, tenor and mode features such as those listed in Table 3.3, then it is easier
to make informed selections of lexicogrammatical systems to conduct the analysis in terms of:

344
 Analysis guide

field: logical systems of taxis and logico-semantic type; experiential systems of transitivity
(agency, process type, circumstantiation), thing type, classification, epithesis, qualifica-
tion, and so on;
tenor: interpersonal systems of mood, polarity, modal assessment; person, attitudinal epith-
esis; connotation;
mode: textual systems of cohesion, and of theme, information; determination, and so on.

For example, we can contrast Halliday’s (1971) selection of the experiential system of transitiv-
ity with his (1982) selection of the interpersonal system of modality:

• Halliday (1971/2002e: 108) selects the system of transitivity in response to the following
objective: “I propose to examine an aspect of the linguistic resources as they are used first to
characterize the people’s world and then to effect the shift of world-view.” In other words,
his objective can be located within the contextual parameter of field, which includes world-
view or cosmology, and field implicates ideational systems, including the experiential clause
system of transitivity.
• Halliday (1982/2002e: 128) is concerned with the de-automatization of lexicogrammar in
relation to semantics, the potential for the deployment of grammatical systems to construe
or enact meanings above and beyond automatically realizing semantic patterns. To illustrate
this phenomenon, he chooses J.B. Priestley’s play An Inspector Calls, and starts with his
interpretation of the themes of the play: The themes that concern us here will be two that
are clearly among the predominant ones in his work. One is interdependence and social
responsibility: “no man is an island”. The other is time: “time is not ticking our lives away”.
To investigate the first theme by means of lexicogrammatical analysis, he selects to focus on
the interpersonal system of modality.

Clearly the selectivity based on objectives specified in terms of – or at least with reference to –
field, tenor and mode depends on the possibility of increasing the delicacy of the specification
of field, tenor and mode values in focus in the specification of field, tenor and mode values.
For example, while Halliday’s (1982/2002e) objective can certainly be located within tenor,
we can see that it is possible to take it a couple of systems further: it is concerned with the enact-
ment of value systems, more specifically values related to social responsibility. Therefore we can
then see that the system of modality is likely to be a resource involved in the enactment of tenor
since there is a link between social responsibility and the modality of obligation. If we take into
consideration developments in the last 40 years, since Halliday analysed passages from An Inspec-
tor Calls, we might think that the system of appraisal (e.g. Martin and White, 2005) would also
be implicated, more specifically of the subtype ‘judgement’ within ‘attitude’; and if we begin to
examine the play with our appraisal spectacles firmly in place, we would probably attend to the
opening exchanges of the play –

BIRLING: Giving us the port, Edna? That’s right. [He pushes it towards ERIC.] You ought to like
this port, Gerald. As a matter of fact, Finchley told me it’s exactly the same port your father
gets from him.
GERALD: Then it’ll be all right. The governor prides himself on being a good judge of port. I
don’t pretend to know much about it.

Here Mr. Birling’s potential son-in-law, Gerald, teases his own father about “being a good judge
of port”, and as an analyst one might then hypothesize (i.e. guess!) that this signals an aspect

345
 Analysis guide

of a theme in the play, and tentatively add judgement to the systems in focus in the course of
analysis. And this will turn out to be helpful also in dealing with the region between grammar
and lexis where the grammatical system within modality of modulation shades into judgement,
as in the following passage in Act 2 analysed by Halliday:

INSPECTOR [massively]: Public men, Mr Birling, have responsibilities as well as privileges.

BIRLING: Possibly. But you weren’t asked to come here to talk to me about my responsibilities.

The general point is this: the selection of lexicogrammatical systems to focus on in text analysis
is very likely the outcome of a “dialogue” between objectives stated in terms of the contextual
parameters of field, tenor and mode and the ideational, interpersonal and textual systems likely
to resonate with values within those parameters; that is, as analysts we keep shunting between
the two sets of system, contextual and linguistic (lexicogrammatical or semantic).
In selecting lexicogrammatical systems to focus on in text analysis, we also need to take practi-
cal considerations into account:

• the effort involved in the analysis will determine what systems to focus on and also how
much text must, or can, be analysed;
• here there is a related consideration: we need to consider the degree of difficulty involved
in different kinds of analysis, and also the degree to which the analysis can be automated;
• for example, doing manual analysis, analysts usually find that thematic analysis is easier
and faster than transitivity analysis, and that appraisal analysis is easier and faster (at least in
certain respects) than modality analysis (to a large extent because of differences in poly-
semy between lexical appraisal systems – in particular, attitude – and grammatical modality
systems);
• and analysts will be able to use tools for automated analysis of some systems but must rely
on manual analysis for other systems; for example, tools for sentiment analysis can provide
crude support for appraisal analysis but transitivity analysis remains hard to automate.

These considerations all play a role in decisions both about the selection of systems to focus on
and decisions about the volume of text to analyse. Decisions about the volume of text obvi-
ously also depend on other factors, perhaps primarily whether the text serves as an artefact or
a specimen (Figure 4.14) and whether the analysis is qualitative, quantitative or both. If it’s
quantitative, analysts must consider statistical measures concerning representativeness and reli-
ability. Analysts should also try to develop a sense of their own analytic speed – how long it takes
them to undertake analysis in reference to diferent systems for a particular quantum of texts,
using one kind of tool or another. This is one reason why pilot studies are very useful.

7.1.4 Interpretation of analysis

Once a text has been analysed lexicogrammatically, the next step is to interpret it in the light of
the objective or objectives for undertaking text analysis. These objectives may be closely related
to the interpretation; they are often concerned with the contextual significance of the text in
terms of field, tenor or mode. Thus, as noted earlier, Halliday (1971/2002e: 108) starts with
field – “world view”, and Halliday (1982/2002e: 128) begins with tenor – “interdependence
and social responsibility”.
Interpretation involves going beyond the analysis within the lexicogrammatical stra-
tum in order to identify lexicogrammatical patterns but it also typically means transcending

346
 Analysis guide

lexicogrammar into semantics and also context in order to suggest linguistic and contextual
meanings. The examples of publications based on text analysis listed in Section 7.1.1 illustrate
the move from analysis to interpretation. In the case of verbal art, there has naturally been spe-
cial emphasis placed on interpretation, and Hasan (1985b) presents a framework for articulating
higher-order meanings in verbal art – a framework that has been widely used; overviews include
Lukin and Butt (2009), Lukin (2015), Miller (2017, 2019) and examples can be found in e.g.
Birch and O’Toole (1988), Lukin and Webster (2005), Webster (2015b).
As a brief example, we can refer to Matthiessen’s (2015f) study of a retelling for children of
the Genesis account of the great flood. The analysis includes a transitivity analysis of all the
clauses of the text.1 Moving on to the interpretative stage, he then identifies patterns of charac-
ters interacting in different participant roles. Clause by clause, a distinctive world view emerges,
construed by clauses of the different process types. For example, there is a ‘material’ world order
where God acts on natural forces, and the natural forces destroy the world. This is construed by
successive combinations of Actor + Process + Goal. Putting these patterns together, Matthiessen
arrives at the world order sketched in Figure 7.2: in terms of potency of participation, the world
is order as follows: God > natural forces > humans > non-humans (animates and inanimates).
Transcending lexicogrammar, Matthiessen suggests that this is an ideational construal of the
world that instantiates the Old Testament construed by many texts. By another stratal step, we
arrive at the contextual parameter of field, where we can interpret the patterns as realizing Old
Testament cosmology (which is fundamentally different from the cosmology construed in Hal-
liday’s, 1984a, North Star text represented in Figure 3.4).
Based on the interpretation we arrive at, we can then examine the text in its context sys-
temically exploring shadow texts – alternative versions of the text within the linguistic and

Figure 7.2 Interpretation of lexicogrammatical analysis: (i) identification of lexicogrammatical analysis in

terms of participant interaction hierarchies and (ii) stratal ascent

347
 Analysis guide

contextual systems that engendered it. For example, what if trees were construed not only as
Goals in destructive ‘material’ clauses (e.g. The very next morning Noah and his sons went to the
cypress forest to cut down the tallest trees for timber.) but also as Actors in creative ‘material’ clauses
(see Halliday, 1990b, and cf. a 7-year-old boy explaining the value of trees to his father: If there
weren’t trees on the earth, we would all be dead, cause there wouldn’t be oxygen; trees make oxygen, so we
can breath . . .). Thus we can adopt a critical stance, identifying problems and suggesting systemic
alternatives that can help us move towards solutions (e.g. in institutions of education). For a
recent contribution to this kind of agenda, see Moya-Guijarro and Ventola (2022).

7.2 Clause analysis guides

Since the clause is the grammatical “gateway” to the text as a semantic unit and since the clause
serves as the environment for all other, and lower-ranking, grammatical units, we shall focus on
guides supporting clause analysis here (for guidance in group/phrase analysis, see Martin, Mat-
thiessen and Painter, 2010). This is directly related to the insights gained from taking a trinocular
view of the clause.

7.2.1 Trinocular view of the clause

All units of language are defined by the relationships that they enter into; that is, they are not
“things in themselves” but rather nodes in the overall relational network in terms of which
language is organized. In other words, they are defined relationally in terms of all the relations
they enter into. As we try to identify the relationship that a given category enters into, the most
general guiding principle is that of trinocular vision: we view the category “from above”,
“from roundabout” and “from below” in terms of the relevant semiotic dimensions – stratifica-
tion, axis, rank and so on. Thus when we approach the clause in analysis, we keep in mind that
it is a “node” in a network of relationships along the dimensions of stratification, rank, and axis,
as shown in Figure 7.3. So the clause is characterized trinocularly as follows:

• stratification
• “from above”, the stratum of semantics: a (ranking) clause realizes a message (textual),
a move (interpersonal) and a figure (experiential);
• “from below”, the stratum of phonology/graphology: a (ranking) clause is often real-
ized by tone group (phonology) or a sub-sentence (graphology);
• “from roundabout”, the stratum of lexicogrammar: a clause is the highest-ranking
unit of the grammatical rank scale, and it is the point of origin of the textual system of
theme, the interpersonal system of mood and the experiential system of transitivity
• rank
• “from above”: the clause is the highest-ranking grammatical unit;
• “from below”: the clause is structured functionally, and its function structure is realized
by a syntagm of classes of the rank immediately below, i.e. of groups and phrases
• axis
• “from above” and “from roundabout”: the clause is the point of origin of the textual
system of theme, the interpersonal system of mood and the experiential system of
transitivity; and “from below” the options in these systems are realized by patterns
along the syntagmatic axis.
348
 Analysis guide

Figure 7.3 Trinocular view of the clause: (i) from above – the clause realizing a message, a move and a
figure; (ii) from below – the clause realized by a syntagm of groups and phrases, and (in terms of
stratification) the (ranking) clause (often) realized by a tone group; and (iii) from roundabout – the
clause as the point of origin of the textual system of theme, the interpersonal system of mood,
and the experiential system of transitivity

The clause is thus located at the intersection of these dimensions. This approach to the char-
acterization of the clause is strikingly diferent from traditional definitions of the clause (or
often the “sentence”) as expressing “a complete thought” or as “consisting of subject and
predicate”. Such definitions represent very partial views, and do not attempt to provide a
rich understanding of clauses in language. But what about practical guidance; how do we
proceed in analysing clauses? The general answer is that clause analysis is an exercise in
weighing considerations “from above”, “from roundabout” and “from below” against one
another, trying to arrive at a balanced picture. A prototypical clause will be easy to identify,
but as soon as clauses are at some remove from the prototypical clauses – for example because
they are downranked (embedded) – we may have to compromise. Another consideration – a
practical one – is what method of analysis we opt for (see Figure 4.14 earlier). If we proceed
349
 Analysis guide

by means of manual analysis, we can easily factor in considerations “from above”, but if we
rely on automated analysis, factors “from below” will necessarily be foregrounded, as in stan-
dard parsers. One option here is to use one of the standard parsers that’s available, and then
do a second round of analysis of the output from that parser, trying to add more functional
annotation, as in an early exploratory study by Honnibal (2004).
The move “from below” in the analysis of clauses involves the identification of groups and
phrases that serve as constituent elements of the clause, so we will summarize probes that are
helpful in identifying groups and phrases in the next subsection. It is helpful to note that in
SFL many probes are based on “exploiting” patterns of agnation: we shed light on a particular
example by examining closely agnate variants (“shadows” of the example under investigation).
For example, to determine whether will is temporal or modal in the following example:

The boat will be salvaged once the weather permits, said the officer.

We can see if it is agnate with a version with secondary future is going to, in which case it is
‘temporal’, or with, say, an agnate variant of modal will such as probably:

The boat is going to be salvaged once the weather permits, said the officer. – Yes, this is an
agnate version.
The boat is probably salvaged once the weather permits, said the officer. – No, this is not
an agnate version.

This probe based on agnation indicates that the example is ‘temporal’ rather than ‘modal’ (of course,
we would need to continue our investigation, probing the other kinds of modality as well, but none
of the following ‘modal’ variants make sense: [usuality] the boat is often salvaged once the weather permits,
said the ofcer; [readiness: inclination] the boat is keen to be salvaged once the weather permits, said the ofcer;
[obligation:] the boat is expected to be salvaged once the weather permits, said the ofcer). (Probes based on
patterns of agnation have also been used outside SFL, for example characterized as “paraphrase”, as
in Hermerén’s, 1978, study of modality in the London-Lund corpus of spoken English.)
In the investigation of spoken text, the clausal analysis discussed here can, and should, be
complemented by the analysis of information units to identify the assignment of Given and New
information, and of Focus of information within New. The most detailed treatment is Halliday
and Greaves (2008, especially Chapter 5).

7.2.2 Constituent parts of clauses

7.2.2.1 The view “from below”

Clauses consist of the units of the rank next below on the rank scale (see Figure 7.3 earlier); the
elements of a clause are groups (nominal, verbal, adverbial or conjunction groups) or phrases
(prepositional phrases). For example, in the following clause from one of Johnny Cash’s songs,
Highway Patrolman,

I chased him through them county roads

there are four groups/phrases:

I | chased | him | through them county roads

350
 Analysis guide

Figure 7.4 Rank-based analysis of a clause

And we can represent them in a box diagram as in Figure 7.4.

But how do we know that the groups and phrases I, chased, him, through them county roads
function as units serving as elements in the clause? The simplest answer is: when we “manipu-
late” the clause, they behave as units; when we “move” them around, when we focus on them
using the “it is . . . that” construction [theme predication, cleft construction], when we use
thematic equative, when we replace them, they behave as units:

I chased him through them county roads :

Him I chased through them county roads :
Through them country roads I chased him ::
He was chased by me through them county roads :
By me he was chased through them county roads :
Through them county roads he was chased by me ::
It was me who chased him through them county roads :
It was him that I chased through them county roads :
It was through them county roads that I chased him ::
What I did was chase him through them county roads :
Chase him through them county roads is what I did
etc.

(And there are other possibilities, e.g. chase him through country roads I did. This is highly marked,
having Residue as Theme, but it is possible.) The important point is that you must learn to
manipulate clauses (and other grammatical units), and the most important strategy for manipu-
lating them is to look for agnate alternatives within the textual metafunction. In other words,
you can use the resources of the textual metafunction to probe the organization of clauses.
This is a natural consequence of the nature of the textual metafunction as an enabling resource,
imposing discursive patterns on experiential and interpersonal wordings, and even co-opting
experiential constructions to assign textual statuses to grammatical elements, as in the case of
theme predication (cf. Halliday, 1978; Matthiessen, 1992). Some of the manipulated versions
of the clause are set out in Figure 7.5, and the thematic versions of the “active” voice version of
the clause are set out as box diagrams in Figure 7.6.

7.2.2.2 Types of probe

The probes that can be used in identifying groups and phrases serving as elements of the clause are
summarized in Figure 7.7. The model example is the clause the ancient Mayas lived in present-day
southern Mexico, Guatemala, Honduras, and western El Salvador, and it can be probed in different ways.

351
 Analysis guide

Figure 7.5 Textual “manipulations” of I chased him through them country roads

I chased him through the county roads

Theme Rheme
topical
Subject

Through the county roads I chased him

Theme Rheme
topical
Adjunct

Him I chased through the county roads

Theme Rheme
topical
Complement

Figure 7.6 Topical Theme as Subject (unmarked), Adjunct (marked) or Complement (marked)

(1) We can deploy probes focussed on the composition of potential groups/phrases –

their typical sequences of words of different classes – or pronominal or pro-adverbial variants.
For example, in the clause the sequence in present-day southern Mexico, Guatemala, Honduras,
and western El Salvador can be replaced with the pro-adverb there: the ancient Mayas lived there;

352
 Analysis guide
353

Figure 7.7 Probes for identifying groups and phrases serving as constituent elements of clauses
 Analysis guide

and it can be probed by the interrogative adverb where: where did the ancient Mayas live? (and by
another step, we can construct a thematic equative with where: where the ancient Mayas lived was
in present-day southern Mexico, Guatemala, Honduras, and western El Salvador). These probes indi-
cate that in present-day southern Mexico, Guatemala, Honduras, and western El Salvador is a unit –
an organic whole – that serves as an element of the clause. Similarly, the ancient Mayas can be
“replaced” with they and interrogated by who?
In addition, we can consider the sequences of classes to determine if they constitute likely
syntagms of a unit functioning as an element of the clause. For example, the ancient Mayas is a
sequence of classes – determiner (the) + adjective (ancient) + noun (Mayas) – that is likely to
realize the structure of a nominal group – Deictic ^ Epithet ^ Thing, but Mayas + lived would
only be one if lived could be interpreted as a v-en (past/passive participle), as in Mayas born in
Guatemala.
Further, we can probe potential groups/phrases to see if they can be contracted or expanded
and still retain their integrity as units, e.g. the ancient Mayas > the admired ancient Mayas > the
admired ancient Mayas who built many pyramids; the ancient Mayas > the Mayas; and similarly: lived >
may have lived > may have been going to live.
(2) At the same time, we can also deploy probes that manipulate the status of poten-
tial groups and phrases within the clause, giving them a different status within the clause.
One important class of such probes is based on textual systems. For example, can a sequence
of classes be “moved” together to be given the status of Theme (“replacing” the actual Theme
in an agnate variant of the clause). This is clearly the case with in present-day southern Mexico,
Guatemala, Honduras, and western El Salvador; it can be “moved” to the beginning of the clause,
displacing the ancient Mayas as Theme:

in present-day southern Mexico, Guatemala, Honduras, and western El Salvador the ancient Mayas
lived: or possibly
in present-day southern Mexico, Guatemala, Honduras, and western El Salvador lived the ancient
Mayas.

Another textual probe is to use the highlighting strategy of theme predication to determine
whether a sequence of classes can be focussed on in a theme predication structure such as: it
was . . . that . . .:

it was the ancient Mayas who lived in present-day southern Mexico, Guatemala, Honduras, and
western El Salvador; or
it was in present-day southern Mexico, Guatemala, Honduras, and western El Salvador that the
ancient Mayas lived; or
it was live that the ancient Mayas did in present-day southern Mexico, Guatemala, Honduras, and
western El Salvador.

(Theme predication also distinguishes between experiential elements of the clause and inter-
personal and textual ones since only experiential elements can be highlighted through theme
predication.)
In clear cases, the probes will give us the same results. However, they may point us in differ-
ent directions or they may allow for more than one version, suggesting complementary lines of
analysis. One interesting example from English is the analysis of phrasal verbs such as put off, look
up, look into, put up with. Halliday has shown that they need to be given two metafunctional anal-
yses that are different in their functional constituency: experientially, the phrasal verb is simply

354
 Analysis guide

part of the verbal group functioning as Process, functioning as the Event of that verbal group;
but interpersonally, the phrasal verb is split between two functions, viz. Predicator and Adjunct.
This “incongruity” between the experiential and interpersonal functional structural analyses
helps explain the different textual possibilities: see Halliday and Matthiessen (2014: 413–419).
In the general linguistic and computational linguistic literature, there are various examples
of ambiguities in the analysis of the function of groups and phrases.2 One of the most discussed
ones is the problem of “prepositional phrase attachment”: in our terms, a prepositional phrase
serves as an Adjunct in a clause – typically a circumstantial one, but there are also prepositional
phrases serving as conjunctive and modal Adjuncts. However, prepositional phrases may alter-
natively be downranked to serve as post-modifiers in groups – most commonly as Qualifiers
in nominal groups. So the problem of prepositional phrase attachment is the choice between
the analysis of a prepositional phrase as ranking, so serving as an Adjunct in a clause, or down-
ranked, so serving as a Modifier in a group. Using a pair of examples from Madhyastha, Carreras
and Quattoni (2017), I went to the restaurant by the Hudson and I went to the restaurant by bike, we
can give the most likely analyses that would be assigned in manual analysis in Figure 7.8.
While difficulties may arise in manual analysis, human analysts can usually choose between
the alternatives based semantic considerations. And, therefore, importantly, they can perform
the probes we have just illustrated. Compare the plausibility of marked Theme selection: the

Figure 7.8 Pair of examples with prepositional phrases attached as either Adjunct or Qualifier

355
 Analysis guide

restaurant by the Hudson I went to with the implausibility of the restaurant by bike I went to; and also
of highlighting by theme predication: it was the restaurant by the Hudson that I went to vs. it was the
restaurant by bike that I went to. (In contrast, I went to the restaurant by the river would be genuinely
ambiguous: ‘I went to the restaurant located by the river’ vs. ‘I went to the restaurant by way of
the river’.) However, difficulties often arise in automated analysis, and there is an extensive lit-
erature in computational linguistics/Natural Language Processing on methods for dealing with
prepositional phrase attachment. For example, Gelbukh and Calvo (2018) devote a chapter to
the task of “prepositional phrase attachment disambiguation” based on the approach their book
is concerned with, viz. “selectional preferences”.

7.2.3 Correspondences across metafunctional

function structures of the clause
The clause is a multi-metafunctional construct; it simultaneously realizes a message (tex-
tual), a move (interpersonal) and a figure (experiential). These three metafunctional strands of
function structure are mapped onto one another by conflation; for example, topical Theme
may be conflated with Subject, and in turn Subject may be conflated with a participant such as
Agent. When we analyse clauses exhaustively – i.e. covering them in their three metafunctional
guises, we need to sort out the relationships between the metafunctional layers or strands. Let us
first summarize these relationships schematically, and then mention some issues that may arise
in the course of analysis.
The correspondences across the metafunctional structures are set out in Figure 7.9. This sche-
matic representation brings out the fact that the interpersonal strand of the clause – the modal
function structure – “mediates” between the textual strand – the thematic function structure –
and the experiential strand – the transitivity function structure. (In this chapter, we are con-
cerned with English; the “unification” of metafunctional contributions in the grammar of the
clause is an important typological variable: see Matthiessen, 2004a.)
Thus from the point of view of function structure, the textual system of theme selects among
elements of the clause in their modal functions:

• textual Theme: if present in the modal structure of the clause, conjunctive Adjuncts (real-
ized by cohesive conjunctions) may or may not be given prominence as textual Theme (e.g.
thematic first in First, I divided the presidents between bibliophiles and those to whom books were
more or less alien territory.);

Figure 7.9 Correspondences across metafunctional function structures of the clause

356
 Analysis guide

• interpersonal Theme: if present in the modal structure in the clause, modal Adjuncts (both
comment Adjuncts and mood Adjuncts) may or may not be given prominence as interper-
sonal Theme (e.g. thematic sometimes in Sometimes the guards punish them for minor violations of
rules which have never been explained to them), and the same applies to Vocative elements (e.g.
thematic Jamie in Jamie, you’re awfully negative today.); in contrast, the Finite element in a ‘yes/
no’ interrogative clause and the Wh-element in a ‘wh-’ interrogative clause are almost always
given the status of interpersonal Theme (e.g. thematic Finite are in Are you talking about a
unifying spiritual vision?, and thematic Wh why in Why did you decide to start writing novels, par-
ticularly historical novels?), the Wh-element also serving as unmarked topical Theme;
• topical Theme: the topical Theme is always present (unless structurally ellipted or oth-
erwise “supressed”), and it is an experiential element serving as Subject or Complement
(a participant) or as circumstantial Adjunct (a circumstance) – its status as ‘unmarked’ or
‘marked’ topical Theme depending on the mood type of the clause (see Table 7.5); e.g.
‘declarative’: ‘unmarked’; The little rabbit lay very still for a long time vs. ‘marked’ For a long
time the little Rabbit lay very still; ‘interrogative’: ‘yes/no’: ‘unmarked’ Is the role of ice mimicked
by other aerosols such as volcanic dust, at lower latitudes where there are no stratospheric ice crystals?
vs. ‘marked’ At lower latitudes where there are no stratospheric ice crystals, is the role of ice mimicked
by other aerosols such as volcanic dust?; ‘interrogative’: ‘wh-’: ‘unmarked’ what is the relationship
between the fiction and the non-fiction in terms of the vision in your head? vs. ‘marked’ in terms of the
vision in your head, what is the relationship between the fiction and the non-fiction?; ‘imperative’:
‘unmarked’ turn left at George Street vs. ‘marked’ at George Street turn left.

While the thematic and modal structures are, as it were, mutually exhaustive – i.e. all thematic/
rhematic statuses correspond to modal elements (with the exception of structure markers serving
as structural Theme) and all modal elements have a thematic/rhematic status, only certain ele-
ments of the modal structure correspond to experiential elements in the transitivity structure of
the clause: Subject and Complements to participants, Finite and Predicator to the process, and
circumstantial Adjuncts to circumstances.
And while the system of theme “mediates” between the thematic and modal structures, the
relationship between the modal and transitivity structures is partly determined by the transitiv-
ity pattern of the clause and partly mediated by the system of voice, which determines which
participant will serve as Subject (and, by implication, Complement or agentive Adjunct):

• in a clause with no participant (i.e. a ‘meteorological’ clause), the Subject has no experien-
tial analogue (e.g. it will soon begin to rain; contrast: showers of spears and arrows rained down);
in a clause only one participant (i.e. a non-ranged middle clause, e.g. my wife works in the
pediatrics ward of a hospital) or two participants only one of which can serve as Subject (i.e. a
ranged middle relational clause of the attributive type, e.g. he was a vegetarian who abstained
from alcohol; it lasted 2500 years), the Subject is invariable the Medium (the exception being
prototypical ‘existential’ clauses, e.g. there was a very colorful captain of that turtle boat I was on,
where the Subject is there, as shown by the nature of the ‘tagged’ version, there was . . . wasn’t
there?, and the Medium/Existent is a very colorful captain of that turtle boat I was on);
• in a clause with two or more participants where all of them have the potential to serve as
Subject, the mapping is determined by the system of voice:
• in a ranged middle clause: active – Subject/Medium (e.g. they often enact the death
penalty in vengeance); passive – Subject/Range (e.g. the death penalty is often enacted in
vengeance);

357
 Analysis guide

• in an effective clause: active – Subject/Agent (e.g. developers were destroying the wild
places in many parts of the world); passive – Subject/Medium (e.g. the wild places were being
destroyed in many parts of the world) (with Beneficiary as an alternative if the clause is a
benefactive effective one, e.g. In 1998, Epstein was given a career achievement award by the
Los Angeles Film Critics Association).

(For the metafunctional status of the system of voice, see e.g. Matthiessen, 1995a: 590–603.)
Having summarized the metafunctional correspondences – the conflations – between the
thematic and modal elements of the clause and the modal and transitivity elements of the clause,
we will now note some issues that arise in the functional interpretation of syntagm of groups
and phrases that make up the clause.
(1) Polysemous conjunctions/adverbs. A number of grammatical items can serve in
different senses as conjunctive Adjuncts, comment Adjuncts, mood Adjuncts or circumstantial
Adjuncts, as shown in Figure 7.10; a number of them are listed further below in Table 7.3
(which also includes uses of adverbs as (sub-)modifiers). The main alternatives are as follows:

• either as the Head of a conjunction group serving as a conjunctive Adjunct (e.g. of course in
Of course, Tove Jansson is a friend to them all.) or as the Head of an adverbial group serving as
mood Adjunct (e.g. of course in The clergy’s concern was, of course, still spiritual): actually, at least,
eventually, finally, indeed, in fact, of course, particularly, soon, still, yet;
• either as Head of an adverbial group serving as a comment Adjunct (e.g. honestly in Well
look, honestly, Mrs Finney, my suggestion to you would be that if you want to read English honours
you should spend a year in solid preparation for it and then reapply.) or as the Head of an adverbial
group serving as a circumstantial Adjunct (e.g. honestly in But many do it very self-consciously,
very honestly, and even very constructively.): frankly, happily, honestly, hopefully, naturally, plainly,
stupidly, wisely, wrongly;
• as the Head of a conjunction group serving as a conjunctive Adjunct (e.g. Soon the ark began
to take shape.), or as the Head of an adverbial group serving as mood Adjunct (e.g. It will
soon begin to rain.) or as the Head of an adverbial group serving as a circumstantial Adjunct
(e.g. and get in the right lane fairly soon): finally, soon.

The kind of polysemy noted here is not accidental in the language. In an important contribution
to the literature on grammaticalization, Traugott (1997) shows how indeed (in deed) and in fact have
expanded metafunctionally from their experiential origins, adding interpersonal and textual uses.
To determine which function an instance of a given item serves, we can construct closely
agnate versions with wordings that are not polysemous, e.g.: frankly: frankly speaking as comment
Adjunct vs. in a frank manner as circumstantial Adjunct of Manner: quality. And circumstantial
Adjuncts, unlike conjunctive and modal ones, can be highlighted using theme predication and
thematic equative. Thus, given the example he departed happily, we might analyse happily either as
a comment Adjunct ‘I’m happy that he departed’ or as a circumstantial Adjunct ‘he departed in a
happy way’; but it is only the former that can be highlighted thematically by theme predication
(it was happily that he departed: circumstantial Adjunct) or thematic equative (how he departed was
happily: circumstantial Adjunct).
(2) Phrasal verbs. Phrasal verbs such as look up, look into and put up with may introduce ten-
sion in the organization of the clause in that they may be somewhat indeterminate in terms of
constituency:

The government shouldn’t force people to put up with ugly boxes.

358
 Analysis guide
359

Figure 7.10 Polysemous items and different types of Adjunct

 Analysis guide

The government shouldn’t force people to put up with ugly boxes.

Inducer Pro- Senser -cess Phenomenon
Subject Finite Predi- Complement -cator Adjunct Adjunct

Figure 7.11 Experiential and interpersonal analyses of clause with a phrasal verb (“put up with”)

Experientially, they are “multiword” realizations of the Event in the verbal group that functions
as the Process in the transitivity structure of the clause, but interpersonally, they are split between
the Predicator of the clause and an Adjunct, the latter being realized either by an adverbial group
(in the case of phrasal verbs consisting of verb + adverbial particle) or by a prepositional phrase
(in the case of phrasal verbs consisting of verb + preposition). Thus the example above can be
analysed as in Figure 7.11. Textual probes involving the systems of theme and voice provide
evidence for the dual analysis of put up with:

Theme
With ugly boxes the government shouldn’t force people to put up
Ugly boxes the government shouldn’t force people to put up with

Predicated theme
It is with ugly boxes that the government shouldn’t force people to put up
It is ugly boxes that the government shouldn’t force people to put up with

But not
Up with ugly boxes the government shouldn’t force people to put

Compare also
Ugly boxes shouldn’t be put up with by people
The government shouldn’t force people to put involuntarily up with ugly boxes.

There are many examples where tension arises between the diferent metafunctional strands of
analysis, and such tensions may be resolved in the course of evolution.

7.2.4 Thematic analysis

Thematic analysis of the clause treats it as a message (cf. Figure 7.3) – a quantum of information –
in the flow of information created by texts unfolding in their contexts of situation. It is probably
the type of lexicogrammatical analysis that is most commonly undertaken (if we set aside aspects
of appraisal analysis that can be interpreted as interpersonal lexical analysis) – partly because
thematic analysis tends to reveal interesting patterns beyond the clause having to do with the
“method of development” of text (as shown by Fries, 1981, in his pioneering study of Theme
selection in relation to the organization of text), and partly because thematic analysis tends to be
easier and faster than other types of clausal analysis – certainly than transitivity analysis.
Out of the different metafunctional strands of the clause, it is the textual one that lends itself
most obviously to a left-to-right “parse” of the clause, a sequential parse. This is not surpris-
ing, since thematic prominence is realized iconically in English as positional prominence at the

360
 Analysis guide

Figure 7.12 The potential for multiple Themes, and the thematic parse of the clause from left to right

beginning of the clause. If a clause includes thematic contributions from all three metafunc-
tions, the first phase of the Theme is typically textual Theme followed by interpersonal Theme,
although we can readily find instances where interpersonal Theme precedes textual Theme
(Thankfully, however, Combined Services Entertainments was formed); but topical Theme always con-
cludes the thematic part of the clause. In any case, thematic analysis can proceed from left to
right, as represented schematically in Figure 7.12. In this left to right parse of the clause, we first
check if there is a textual Theme, then if there is an interpersonal Theme and finally we identify
the topical Theme as the first experiential element of the clause. In the case of the example in
the figure, the sequence is thus Theme: nevertheless, unfortunately, many of these strategies Rheme:
utilize harmful organic solvents.
In carrying out the thematic analysis, we can refer to the “checklist” of thematic candidates
set out in Figure 7.13, and then in more detail in Table 7.3 for clauses without Wh-elements
and Table 7.4 for clauses with Wh-elements.
The thematic analysis of the clause can of course be supplemented by the textual analysis of
the intonation unit in spoken English, a good source for details and illustrations being Halliday
and Greaves (2008: Section 5.1).

7.2.5 Modal analysis of the clause

Modal analysis of the clause treats the clause as a move (cf. Figure 7.3) – a quantum of
interaction – in the exchanges that make up dialogue. It encompasses all the interpersonal
systems of the clause – centrally the system of mood, but also the closely related systems
of polarity, deicticity, modal assessment (including modality) and address (vocative). In
spoken English, the potential of the interpersonal clause grammar is expanded thanks to the

361
 Analysis guide

Figure 7.13 Thematic analysis chart

362
 Table 7.3 Elements capable of serving as textual, interpersonal and topical Theme (1): non-Wh-elements

Textual Interpersonal Topical

1 2 3 1 2 3

Function: Conjunctive adjunct Modal adjunct Vocative Finite Participant, circumstance or process

Class: Continuative Structural conjunction, Cohesive conjunction Modal adverb Nominal group Verbal operator Nominal group, adverbial group,
conjunctive prepositional phrase, verbal group
preposition

continuatives structural conjunctive modal Adjunct Vocative finite verb

now, well, oh, conjunction Adjunct: – mood Adjunct: nominal group – auxiliary
ah, yes, – linker: e.g., i.e., cohesive mood adverb (group) (proper ± title, do, does, did, is,
yeah, no viz.; and, or, but; conjunction surely, certainly, probably, role, term of am, are, was,
so, then for example, in perhaps, maybe; loaded affect were, has,
– binder: while, other words, in always, never, usually, often, (positive/ have, had,

Analysis guide
whereas; when, summary, thus, sometimes, occasionally, rarely negative)) will; must,
as, since, before, specifically; also, – comment Adjunct: For example: shall, ought
363

after, if, although, in addition, comment adverb (group) Ms Finney, to, should,
because, since; that, moreover, naturally, of course, obviously, Eliza, my will, would,
whether alternatively, in clearly, doubtless, unsurprisingly, dear may, might,
(conjunctive contrast, however; predictably, surprisingly, can, could
preposition, therefore, evidently, allegedly, supposedly, – lexical
serving as a consequently, arguably, presumably, luckily, verb in
binder: as well as, thus, nevertheless, fortunately, hopefully, sadly, ‘relational’
instead of, with, later, afterwards, unfortunately, amusingly, clause that
on, after, before, meanwhile, funnily, importantly, can serve
because of, despite) firstly, significantly; without “do
secondly . . . truly, honestly, admittedly, certainly, support” am,
finally similarly, actually, in fact, generally, is, are, was,
likewise broadly, roughly, on the whole, were; [chiefly
legally, academically, morally, in BE:] has,
frankly, candidly, honestly, have, had
confidentially, between you and
me, personally, truly, strictly,
tentatively, hypothetically
 Analysis guide

Table 7.4 Elements capable of serving as textual, interpersonal and topical Theme (2): Wh-elements

Textual Interpersonal Topical

1 2 3 1 2 3

function: Wh-element [relative] Wh-element Wh-element

<also topical> [interrogative] <also [relative or
topical> interrogative]
class: – nominal group with – nominal group – nominal
relative pronoun or with interrogative group
determiner pronoun or
[pronoun] that; who, whom, determiner
which; whatever & other [pronoun] who,
indefinite -ever forms; whom, what, which;
[determiner] whose + noun, [determiner] what +
(rare: which + noun) noun, which + noun,
whose + noun
– adverbial group with – adverbial group – adverbial
relative adverb with interrogative group
when, where, why, how adverb
when, where, whither, why,
how, how long, how far
– prepositional phrase – prepositional phrase – prepositional
with interrogative with interrogative phrase
nominal group nominal group
e.g. with whom, instead of e.g. with whom, instead of
which or who . . . with, what or who . . . with,
which . . . instead of what . . . instead of

expressive resources of the phonological system of tone: see Halliday and Greaves (2008).
The maximal modal structure is shown schematically in the move row in Figure 7.14, with
examples given below. It covers:

the structure of the clause enacting it as a proposition or a proposal, viz. Mood +

Residue. Mood consists of Subject and Finite, and potentially also a mood Adjunct.
Residue consists of Predicator, Complement(s) and circumstantial Adjunct(s);
the structure of the clause relating to the move exchanging this proposition or proposal, viz.
a specification of the addressee (Vocative), comments on the proposition or proposal or the
nature of the move itself (comment Adjuncts) and conjunctive items linking the move to
other moves (conjunctive Adjuncts).

7.2.5.1 Modal analysis: the mood element

In undertaking modal analysis, one can use the fact that the unmarked Theme in English
always reveals the interpersonal key signature of the clause, as shown in Table 7.5. However,
the Theme may be marked, and to do a maximally informed modal analysis of an English
clause, we need to locate the Mood element. This involves a prosodic parse of rather than a
sequential, as in the case of the thematic analysis: we need to identify the places in the clause

364
 Analysis guide

Figure 7.14 Maximal modal structure of the clause, with thematic and transitivity correspondences

Table 7.5 The identity of unmarked Themes in ‘free’ clauses (i.e. clauses selecting for mood)

MOOD TYPE Interpersonal theme Topical theme

indicative declarative non-exclamative Subject

exclamative [Wh] Wh
interrogative yes/no Finite Subject
wh [Wh] Wh
imperative (Finite) Predicator

where there is possible variation between Subject ^ Finite and Finite ^ Subject, which is like
the contrast between falling tone and rising tone, and the potential tail in the form of a Mood-
tag: Tagfinite ^ Tagsubject, e.g.

According to David Niven, the moon is/is the moon a balloon, isn’t it?

365
 Analysis guide

In a ‘free’ clause, we can locate the Mood element by checking an agnate tagged version of the
clause with a Moodtag since the Moodtag picks up the Subject and the Finite of the Mood ele-
ment in reverse sequence (though it doesn’t include mood Adjuncts). This works for ‘declara-
tive’ and ‘imperative’ clauses (and for ‘yes/no’ interrogatives in Australian English):

‘declarative’: the moon’s a balloon – the moon’s a balloon,isn’t it? so: the moon’s a balloon
‘imperative’: bring on the empty horses – bring on the empty horses, won’t you? so: no explicit
Mood element here; but:
don’t you bring on the empty horses – don’t you bring on the empty horses, will you? so: don’t you
bring on the empty horses
let’s bring on the empty horses – let’s bring on the empty horses, shall we? so: let’s bring on the
empty horses
don’t let’s bring on the empty horses – don’t let’s bring on the empty horses, shall we? so: don’t let’s
bring on the empty horses

In the case of ‘declarative’ clauses, another strategy is to check the agnate ‘yes/no’ interrogative
version, since it will reveal Subject and Finite in reverse sequence:

the moon’s a balloon – is the moon a balloon? so: the moon’s a balloon

And this may pick up mood Adjuncts as well:

the moon’s really a balloon – is the moon really a balloon? so: the moon’s really a balloon

While it is not obvious from the sequence, the orientation of the modal assessment is a clue: ‘do
you really say the moon’s a balloon’.
If the clause is ‘interrogative’ (or indeed a bound finite one), the best strategy is to check the
agnate ‘declarative’ clause:

what’s the moon – the moon’s something, isn’t it? so: what’s the moon ?
if the moon is a balloon – the moon’s something, isn’t it? so: if the moon is a balloon

The two can be combined if we construct a passage of dialogic exchange:

DAVID NIVEN: The moon’s a balloon.

INTERVIEWER: Is the moon really a balloon?
NIVEN: Yes, it is.
INTERVIEWER: But surely it can’t be, can it?

The point is that just as in the analysis of other areas of the grammar, we use patterns of agnation
captured in the systemic description to probe the unit being analysed by examining its systemic
neighbourhood. If the clause is a non-finite bound one, there are more degrees of separation
in agnation from a declarative version; but we can still relate the moon being a balloon in Niven
considered the possibility of the moon being a balloon to Niven thought the moon was a balloon and then
Niven said: “The moon is a balloon”, and now we can relate the moon is a balloon to the moon is a
balloon, isn’t it? and is the moon a balloon?
In the same way, exploring the system in the analysis of an example, we can use patterns of
agnation to indicate whether an interpersonal Adjunct is a mood Adjunct, i.e. part of the Mood
element or not:

366
 Analysis guide

the moon’s perhaps a balloon – the moon may be a balloon so, suggesting: the moon’s per-
haps a balloon

And a consideration of the ‘yes/no’ interrogative version supports this analysis: is the moon perhaps
a balloon means ‘in your opinion, is it?’. In other words, there is a contrast in orientation of the
modal assessment from (declarative) orientation towards speaker and (interrogative) orientation
towards addressee.

7.2.5.2 Modal structure of the clause in relation to thematic

and transitivity structure; maximal modal structure
The modal structure of the clause as a move in dialogic exchanges maps onto the experiential struc-
ture of the clause as a figure and the textual structure of the clause as a message. The mapping is set
out schematically in Figure 7.14, with examples constructed to illustrate different patterns. Func-
tions in the modal structure of the clause correspond to textual and experiential functions as follows:

modal (interpersonal) – experiential

Subject and Complement(s) are conflated with participant functions; one participant func-
tion is given the status of actual Subject, the other(s) the status of potential Subject, i.e.
Complement;
Adjuncts are conflated with circumstantial functions – circumstantial Adjuncts; they are not
potential Subjects, and are thus further removed than Complements from having modal
responsibility, from the status of the nub of the argument (Subject);
Finite and Predicator together conflate with the Process function, which is thus split when
Finite and Predicator are split (as with has and completed in has she completed the experi-
ment; she has probably completed the experiment). When the clause has no Finite operator,
the Predicator alone corresponds to the Process, as in the unmarked kind of addressee-
oriented imperative clause (complete the experiment!) and in non-finite bound clauses (com-
pleting the experiment [she asked her class if they had any questions]).
The Wh-element can be conflated either with a participant role (as Subject or Comple-
ment) or with a circumstantial role (as Adjunct).

The following modal functions have no experiential role:

interpersonal Adjuncts (comment Adjuncts, mood Adjuncts);

the Vocative element;
conjunctive Adjuncts

modal (interpersonal) – textual: the potential of modal functions to serve as interpersonal

Theme or topical Theme depends on the mood type of the clause and is represented for
the unmarked case in Table 7.5. Unless they serve as the Wh-element, Complements and
circumstantial Adjuncts are marked as topical Theme, Complements being more marked
than Adjuncts. Unless the clause is imperative, the Predicator is highly marked as Theme:
it may be given thematic status as part of the Residue, as in the following example with
Residue as Theme and Mood as Rheme:
Just because e-books are available on better (the new Kindle) and more (all cell phones) devices
doesn’t mean people will read them. But mark my words, read them they will.

367
 Analysis guide

Figure 7.14 represents the modal structure of the clause as move between the transitivity struc-
ture of the clause as figure and the thematic structure of the clause as message, and the modal
structure does in fact serve to “mediate” between the transitivity structure and the thematic
structure since the nature of the unmarked topical Theme depends on the mood type of the
clause. This relates directly to the system of voice: the choice among participants which will be
mapped onto the Subject of the clause.

7.2.5.3 Freedom: free vs. bound clauses

The distinction between ‘free’ and ‘bound’ clauses has to do with the nature of the contribution
that clauses make to the creation of meaning in text, and as a consequence also with options that
are open to free clauses but not to bound ones. Free clauses contribute to the creation of mean-
ing in text directly, having the full range of options in textual, interpersonal and experiential
systems. Bound clauses contribute indirectly, as it were; they contribute by qualifying another
clause (as hypotactically dependent clauses) or by being downranked to serve as an element of
another clause, either in the structure of the clause or in the structure of a group or phrase serv-
ing within the clause.
Let us begin giving an illustration from a dialogue: see Example 7–1. Here most clauses are
‘free’, contributing directly to the dialogic exchanges, either as question + answer sequences, or
as clauses expanding on questions or answers, as illustrated in Figure 7.15. The bound clauses
make a different kind of contribution to the dialogue. (1) The dependent bound clauses qualify
other clauses by enhancing them in terms of time (until my mother put a stop to it; while I was at
Yale; when my first short story, “Sadie” won the Atlantic Price), reason (because I came back there) or
purpose (to teach writing my first year out of college) or by elaborating on them (which was very useful).
(2) The embedded (downranked) bound clauses serve as postmodifying Qualifiers in nominal
groups, defining (restricting) them (various gunnery training devices [[used during World War II]]; a
letter [[[that said || . . . .]]]).

Key for the examples

underlining = dependent bound clauses;
italics = embedded bound clauses;
plain = free clauses. Colour coding;
double underliningg = linkers (structural conjunctions introducing paratactically related
clauses);
bold = binders (structural conjunctions introducing bound clauses, dependent or embedded);
borderedd = cohesive conjunctions (including linkers used cohesively).

Example 7–1: Excerpt from an interview with Peter Matthiessen

INTERVIEWER: ||| Can you tell us a little about your early life? ||| You’re a veteran, || I under-
stand.3 |||
MATTHIESSEN: ||| Yes, || I served in World War II and then I went to Yale. ||| I took my junior
year abroad at the Sorbonne. |||
INTERVIEWER: ||| Did you grow up in the city? |||
MATTHIESSEN: ||| I was born right here in New York City, Madison Avenue and 65th Street,
|| I think4 it was, at a little lying-in hospital. ||| I went to school in New York City || and
then we lived up on the Hudson for a while, || then moved to Connecticut. ||| So I’ve been
around the New England coast all my life. |||

368
 Analysis guide

INTERVIEWER: ||| What did your father do? |||

MATTHIESSEN: ||| He was an architect. ||| Then he went in the Navy || and helped design vari-
ous gunnery training devices [[used during World War II]]. ||| He didn’t really want to go back
to architecture. ||| After the war, he became a fund-raiser and spokesman for conservation
groups, the Audubon Society and Nature Conservancy, that kind of thing. |||
INTERVIEWER: ||| So you would credit your family life then, with your interest in nature, the
environment? |||
MATTHIESSEN: ||| We had a wonderful piece of property in Connecticut, back up in the hills,
|| and my brother and I were both very interested in snakes and birds. ||| We had a big
copperhead den on our property. ||| Until my mother put a stop to it, || we had a lot of
copperheads in cages. ||| I went on to birds || – starting with my mother’s feeder – || and my
brother became a marine biologist. ||| I think we kind of taught my father – || it was the
other way around. |||
INTERVIEWER: ||| What did you study in school? |||
MATTHIESSEN: ||| I was an English major, || but I took courses in biology and ornithology. |||
I began writing in boarding school, smart-alec articles about this and that. ||| T Then , w with a
friend, I did a column for the Yale Daily News on hunting and fishing. ||| I started writing
short stories || while I was at Yale, || and I was still there || when my first short story, “Sadie”
– <<it’s in On The River Styx>> – won the Atlantic Prize, || which was very useful ||
because I came back there || to teach writing my first year out of college. ||| I didn’t last very
long as a teacher, just one term, || but the publication was a big help. ||| The Atlantic took a
second story, || and I got an agent. ||| Then I star t my first novel || and sent off about four
started
chapters || and waited by the post office || for praise to roll in, calls from Hollywood, every-
l my agent sent me a letter [[[that said || “Dear Peter, James Fenimore Cooper
thing. ||| Finally
wrote this a hundred and fifty years ago, || only he wrote it better. || Yours, Bernice.”]]] ||| I probably
needed that; || it was very healthy. |||

As Example 7–1 illustrates, there is thus a scale from ‘free’ clauses via dependent ‘bound’ clauses
to embedded ‘bound’ clauses: see Figure 7.15. Free clauses are the most negotiable clauses, while
non-finite downranked bound clauses are the least.

Figure 7.15 Scale of clausal “freedom” – from free to downranked (embedded) bound clause (with clause metaphorically
re-represented by nominal group with nominalized verb as Head/Thing as one additional step)

369
 Analysis guide

Figure 7.16 Dialogic exchange – initiating ^ responding moves realized by free clauses selecting for mood type

In fact, the system of mood type is the crucial characteristic of ‘free’ clauses, as shown in Figure
7.16. It is precisely this system that enables speakers to use ‘free’ clauses in dialogue to assume
a speech role and to assign a complementary one to their addresses, as in the question-answer
sequences in Example 7–1 earlier. So we can use the possibility of contrasting mood selections
characteristic of ‘free’ clauses as a way of probing clauses to determine whether they are ‘free’ or
‘bound’. Thus the clause I was at Yale is a ‘free’ one because it contrasts with was I at Yale?, where
was I?, be at Yale!

• declarative: I was at Yale (wasn’t I?)

• interrogative: yes/no: was I at Yale?
• interrogative: wh – where was I?
• imperative: be at Yale (won’t you?)

But these options in mood are not open to the ‘bound’ clause while I was at Yale; for example,
we cannot say while was I at Yale? nor while where was I?; and we cannot tag this bound clause –
we cannot say while I was at Yale, wasn’t I? Thus even though finite bound clauses have the same
Mood element as a declarative free clause, i.e. Subject ^ Finite, there is no contrast with other
Mood elements. The sequence of Subject ^ Finite is simply the interpersonal default in English,
as it were. Semantically, while a declarative clause realizes a statement – a gift of information –
that can be negotiated, a bound clause makes no such contribution: what is presented is to be
assumed, or taken for granted.
The system of mood type is thus a key property of free clauses, and we can include it at the
top of a list of properties differentiating free and bound clauses: see Table 7.6.
Note that in projected bound clauses, the mood type of the equivalent free clause is
represented, but it is not an interpersonal contribution to the ongoing dialogic interaction:
Table 7.7.

370
 Analysis guide

Table 7.6 Freedom: some properties of free vs. bound clauses

Property Free Bound

can select for mood Yes: No

type mood type: indicative [declarative/
interrogative]/imperative
negotiable Yes No
(challengeable)
A: David was surprised by the fact [[that A: David was surprised by the fact [[that the
the moon was a balloon]] – moon was a balloon]] –
B: No, he wasn’t. B: [Not possible as a rejoinder:] No, it isn’t.
taggable Yes, if declarative or imperative No
A: David was surprised by the fact [[that A: David was surprised by the fact [[that the
the moon was a balloon]], wasn’t he? moon was a balloon, wasn’t it?]].
finiteness Always finite Either finite or non-finite
The moon is a balloon because the moon is a balloon/
because of [the moon] being a balloon

Table 7.7 Straight mood in free clauses, and projected mood in bound clauses (of projection: reporting)

Straight mood Projected mood

indicative declarative
The moon is a balloon (he said) that the moon was a balloon
interrogative: yes/no
Is the moon a balloon? (he asked) whether the moon was a balloon
interrogative: wh-
What is the moon? (he asked) what the balloon is
imperative Bring on the empty horses! (he told them) to bring on the empty horses

In the excerpt from the spoken dialogue in Example 7–1, most clauses are ‘free’ rather than
‘bound’, and the majority of ‘bound’ clauses are dependent rather than downranked (embed-
ded). The situation is likely to be fairly diferent in a formal written monologic text, as is illus-
trated by the excerpts in Example 7–2 and Example 7–3.

Example 7–2: Excerpt from a written opinion text (coding as for the first example)
||| Trump’s presidency is unravelling. ||| B
But h
he won’t fall without a push |||
by Gary Younge, theguardian.com March 15, 2018 11:38 AM
||| The administration is mired in incompetence and calamity. ||| So why aren’t his enemies
presenting a hopeful alternative? |||
Illustration by Ben Jennings Photograph: Ben Jennings
||| Even by Donald Trump’s standards, Tuesday was extraordinary. ||| First came the tweet
[[that he had fired his secretary of state Rex Tillerson]]. ||| T
Then a state department spokes-
man issued a statement [[[claiming || Tillerson was “unaware of the reason” for his dismissal, ||
and had heard about it on Twitter]]]. ||| A few hours later the spokesman had been fired too.

371
 Analysis guide

||| Meanwhile
M the lawyer of porn actor Stephanie Clifford (stage name: Stormy Daniels),
<<who allegedly had an affair with Trump,>> warned || the country to “buckle up” ||
as Clifford sought to extract herself from her non-disclosure agreement || so she could
“publish any materials, such as text messages, photos and/or videos [[relating to the presi-
dent]] [[that she may have in her possession]]”. ||| Back in Washington, the Trump team
announced || it would be hiring John McEntee, Trump’s former personal assistant, as a
senior adviser for campaign operations. ||| The day before, McEntee had been escorted
from the White House || because he is under investigation by the Department of Home-
land Security for serious financial crimes. |||
||| While all this was going on, || voters in south-west Pennsylvania’s 18th district went to
the polls in a byelection, in a district [[Republicans have held for the past 15 years]]. ||| It was so
safe [[that Democrats didn’t even bother contesting the last two elections]]. ||| Trump trounced Hill-
ary Clinton there by about 20 points. ||| It should have been a shoo-in for the Republicans.
||| By the end of the night Democrats were celebrating a wafer-thin victory, || though this
may yet be challenged. |||

Example 7–3: Excerpt from the introduction to Gass and Selinker’s textbook second
language acquisition
1 Introduction
1.1 The study of second language acquisition
||| Second language acquisition (SLA) is a relatively young field. ||| We would be hard-
pressed to state a “beginning” date, || but it is probably fair [[[to say || that the study of SLA
has expanded || and developed significantly in the past 40–45 years]]]. ||| This is not [[[to say ||
that there wasn’t interest in the fields of language teaching and learning before then]]], || for
surely there was. ||| It is [[[to say, however , || that since that time the body of knowledge of
the field has seen increased sophistication]]]. |||
||| We are far from a complete theory of SLA, || but there is progress. ||| By approaching
SLA from a variety of disciplinary perspectives, || as we will see in this chapter and in the
remainder of this book, || we have come a long way from pure descriptive studies to research
[[that connects with other disciplines]]. |||
||| What is the scope of SLA? ||| What does the study of SLA consist of ? ||| It is the study of
[[how second languages are learned]]. ||| In other words , it is the study of the acquisition
of a non-primary language; || that is, the acquisition of a language beyond the native lan-
guage. ||| It is the study of [[how learners create a new language system with only limited exposure
to a second language]]. ||| It is the study of [[[what is learned of a second language || and what is
not learned]]]; || it is the study of [[[why most second language learners do not achieve the same
degree of knowledge and proficiency in a second language || as they do in their native language]]]; || it
is also the study of [[why only some learners appear to achieve native-like proficiency in more than
one language]].
A
||| Additionally , second language acquisition is concerned with the nature of the hypotheses
(whether conscious or unconscious) [[that learners come up with regarding the rules of the second
language]]. ||| Are the rules like those of the native language? ||| Are they like the rules of
the language being learned? ||| Are there new rules, like neither language, being formed?

372
 Analysis guide

||| Are there patterns [[that are common to all learners regardless of the native language and regard-
less of the language [[being learned]] ]]? ||| Do the rules [[created by second language learners]] vary
according to the context of use? ||| Do these rules and patterns vary more in individuals in
a second language [[than they vary in the native language]]? ||| Given these varied questions,
|| the study of second language acquisition draws from || and impacts many other areas
of study, among them linguistics, psychology, psycholinguistics, sociology, sociolinguistics,
discourse analysis, conversational analysis, and education, || to name a few. |||

7.2.5.4 Subject
The Subject functions is part of the Mood element, so in principle, it will be identified as part
of the modal analysis when the Mood element is identified, as indicated above. Thus we will
only add some general comments here, and point to a few cases that may need special attention.
Like all functions of the clause, and of other grammatical units, the Subject is characterized
by its role as a node in relations with the different semiotic dimensions of the architecture of
language in context. In other words, it is not a “thing in itself ”. Instead, it can be interpreted,
apprehended and identified by reference to the relationships it enters into. As in the case of all
other structural functions, it needs to be viewed trinocularly in reference to the relevant semiotic
dimensions – that is, “from above”, “from below” and “from roundabout”. This is brought out
in IFG4, where Subject is characterized in terms of a “thick description” (pp. 165–166). Here
we will present just a few of these views:

In terms of the hierarchy of stratification, seen “from above”, from the point of view
of the proposition or proposal being negotiated by the clause as a dialogic move, the
Subject is that element which is held modally responsible for the success of the propo-
sition or proposal – its validity is vested in the Subject. Thus it is that element in terms
of which the proposition or proposal is argued (negotiated) – the nub of the argument.
This is pushed to humorous absurdity in Monty Python’s sketch The Argument Clinic,
which essentially consists of arguments about the polarity of propositions:
MAN: Is this the right room for an argument?
OTHER MAN: (pause) I’ve told you once.
MAN: No you haven’t!
OTHER MAN: Yes I have.
M: When?
O: Just now.
M: No you didn’t!
O: Yes I did!
M: You didn’t!
O: I did!
M: You didn’t!
O: I’m telling you, I did!
M: You didn’t!

In terms of the hierarchy of rank within grammar, the grammatical rank scale, seen “from
below”, the Subject is typically realized by a nominal group. If this nominal group shows a
distinction in case, it will be in the ‘nominative’ case rather than the ‘oblique’ one, except
for non-finite clauses, where it is in the ‘oblique’ case (e.g. me [rather than I] in me to do it:
OK, I’ll do it now; if you desperately want me to do it.), which is in fact a reflection of the point

373
 Analysis guide

that such clauses are enacted as non-challengeable (assumed) rather than challengeable. In
addition, to the extent concord (agreement) is present, Subject and Finite agree in person
and number; this can be interpreted as a person and number prosody with the Mood ele-
ment at its domain (picked up, of course, in the Moodtag). However, both these properties
seen from below are fairly marginal in Modern English (as opposed to earlier versions of the
language, and many other Indo-European languages).
In terms of the hierarchy of axis, seen “from above”, the Subject is determined by options
in the interpersonal system of mood with respect to its presence, nature and relative sequence
within the Mood element, coming either before or after the Finite. In ‘free’ clauses, it is present
in clauses of the ‘indicative’ mood but not in the most unmarked type of ‘imperative’ mood; in
‘bound’ clauses, it is present in ‘finite’ clauses, but typically not in ‘non-finite’ ones (although it
may be). Two of the subject person options have evolved special forms of Subject in ‘impera-
tive’ clauses, as shown in Table 7.8, viz. ‘speaker’ lemme (or let me) and ‘speaker-plus’ let’s; they
are picked up in tagged versions by the regular pronouns: lemme . . ., shall I? and let’s . . ., shall
we?, as in well, then let’s continue, shall we? When the polarity is ‘negative’, this Subject may be
preceded by Finite don’t, as negative imperatives in general: don’t let’s go to the dogs tonight.
In terms of the spectrum of metafunction, seen “from roundabout”, the Subject has cer-
tain characteristic textual and experiential functions, as noted in Section 7.2.3 earlier. For
example, if the clause is ‘declarative’ or ‘interrogative: yes/no’ in mood, the Subject is the
topical Theme in the unmarked case. And if the clause is ‘middle’ in agency, the Subject is the
Medium in the unmarked case; but it if is ‘effective’, it is the Agent. There are two excep-
tions, viz. metrological ‘material’ clauses and typical ‘existential’ clauses, both of which have
Subjects that are not conflated with participant roles but are realized by nominal groups with
non-personal pronouns as Head (Halliday and Matthiessen, 2014: 369), it and there respec-
tively. In meteorological clauses, the non-personal pronoun serving as Subject is ‘specific’, it,
as in the more it snows (Tiddely pom), the more it goes on snowing, whereas in ‘existential’ clauses, it
is ‘non-specific’, there, as in There once was a man from Nantucket Who kept all his cash in a bucket.
These non-personal pronouns as Subject still do their interpersonal duty in the realization of
options in the system of mood, e.g. it snowed : did it snow? :: there was : was there? and it snowed,
didn’t it? : there was, wasn’t there? That these Subjects do not conflate with participant roles is
shown by the fact that there are no agnates with other nominal groups e.g. with lexical Heads
(we can’t say the sky is snowing, except jocularly), by the fact that they cannot be queried in
wh-interrogatives (we cannot say what snowed?, where was once a man from Nantucket?), by the
fact that they cannot be highlighted by thematic equatives (we cannot say what snowed was it,

Table 7.8 Different types of subject person in declarative and imperative clauses

Subject person Declarative Imperative

interactant speaker I have brought on the empty horses Let me bring on the empty horses (shall I?)
(haven’t I?)
speaker-plus We have brought on the empty horses Let’s bring on the empty horses (shall we?)
(haven’t we?)
addressee You have brought on the empty horses Bring on the empty horses (will you?)
(haven’t you?)
You bring on the empty horses (will you?)
non-interactant They have brought on the empty horses (marginal)
(haven’t they?)

374
 Analysis guide

where once was a man from Nantucket was there) or theme predications (we cannot say it was it that
snowed, it was there that was once a man from Nantucket).

7.2.6 Transitivity analysis of the clause

Transitivity analysis of the clause treats it as a figure (cf. Figure 7.3) – a quantum of change in
our experience of the flow of events. The grammar models a quantum of change as a configu-
ration of a process, participants involved in it and attendant (or indirectly involved) circum-
stances: process + participant(s) + circumstance(s). This structure is represented schematically
in the figure row of Figure 7.14.
The process is always present as the Process of the transitivity structure of the clause (the
experiential structure of the clause), unless it is ellipted for textual reasons or the clause is special
case of a non-finite relational clause (e.g. with the cat away in with the cat away, the mice will play,
where being is left structurally implicit).
There are one to three participants in different participant roles, except in meteorological
clauses such as it’s snowing and in cases where the participant is ellipted as Subject or Comple-
ment for textual reasons. Participants are inherent in the process, so they can in principle be
predicted by the process type of the clause.
The combination of process + participant(s) may be augmented circumstantially. Most fre-
quently across the process types, there are no circumstantial augmentations, i.e. zero circum-
stances; but there may be augmentations by different kinds of circumstances – usually no more
than three, but this is a matter of probability.
In the transitivity analysis of the clause, it is helpful to try to identify the Process first. While
the textual parse of a clause is sequential and the interpersonal parse of a clause is prosodic, the
experiential parse of a clause is a configurational parse, and the nucleus of the experiential
configuration is the Process. Once the Process has been identified, this will provide analysts
with information about the number and nature of participants to expect and look for. This
part of the analysis can be guided by the information about clauses of different process types in
Table 7.10. The information about verbs serving as the Event in the verbal groups functioning
as Process in the Vocabulary Guide in Section 7.3.2 is designed to help, in particular in the case
of very common verbs since they are likely to have evolved different senses corresponding to
different process types. Here looking up verbs in a good dictionary can also be very helpful (as
illustrated for suggest in Figure 7.17 in Section 7.3.1 later in this chapter): we can try to match
the verb senses identified in the dictionary to different process types (as illustrated for suggest in
Table 7.11), and identify the sense corresponding to, or closest to, the sense of the verb in the
clause being analysed. One way to differentiate senses is to check synonyms or near synonyms:
each sense is likely to correspond to a different set of synonyms (cf. Table 7.12 in Section
7.3.2 listing common polysemous verbs). This is a manifestation of paradigmatic probing –
i.e. the general principle of probing examples being analysed paradigmatically, checking them
against closely agnate versions. (Another related method is to check the translation equivalent in
another language since different verb senses often have different translation equivalents.)
While the Process is nuclear in the transitivity configuration of the clause as a figure, other tran-
sitivity roles are obviously also relevant. In particular, it is helpful to consider the Process together
with the most centrally involved participant, the Medium (i.e. the participant which is the medium
through which the process is manifested). Thus Process: break + Medium: glass is a configuration in
a ‘material’ clause, while Process: break + Medium: heart is a configuration in a ‘mental’ clause; and
only the latter can be configured with a ‘fact’ clause as Agent, as in it breaks my heart [[that we might

375
 Analysis guide

suggest | sə(g)’jest |
verb [reporting verb]

put forward for consideration: [ with clause ] : I suggest that we wait a day or two | [ with direct speech
] : “Maybe you ought to get an expert,” she suggested | [ with obj. ] : Ruth suggested a vacation.

[ with obj. ] cause one to think that (something) exists or is the case: finds of lead coffins suggested
a cemetery north of the river | [ with clause ] : the temperature wasn’t as tropical as the bright
sunlight may have suggested.

state or express indirectly: [with clause] : are you suggesting that I should ignore her? | [ with obj. ]
: the seduction scenes suggest his guilt and her loneliness.

[ with obj. ] evoke: the theatrical interpretation of weather and water almost suggests El Greco.

(suggest itself) (of an idea) come into one’s mind.

DERIVATIVES

suggester noun

ORIGIN

early 16th cent.: from Latin suggest- ‘suggested, prompted,’ from the vern suggerere, from sub- ‘from
below’ + gerere ‘bring.’

Figure 7.17 Dictionary entry for “suggest” (Apple’s OS X Dictionary)

have to abandon this dream]], where it . . . [[that we might have to abandon this dream]] serves as the Agent.
The Range element may also be criterial, in particular when the lexical verb is a very general one
(a “vector verb”) such as take, have, make, do. For example, Process: take + Range: drive is material, as
in I also took a drive out to his ranch, and Process: take + Range: look is ‘behavioural’, as in I took a look
around, while Process: take + Range: umbrage is ‘mental’, as in Feminists took umbrage at the suggestion
that the under-representation of females in the scientific community might be due to anything other than male
oppression, and Process: take + Range: object is relational (with the sense of ‘have’, ‘go with’), as in
Some verbs must always take a direct object, some never take a direct object. Such examples also illustrate the
fact that it can be helpful to identify collocations such as take + shape/place, take + a look/listen, take +
umbrage/offence, take + object.
Table 7.10 provides information about the possible configurations of participants in
clauses of the different process types, e.g. Senser or Senser + Phenomenon in ‘mental’
clauses. It also provides information about other properties such as unmarked present tense,
pro-verb and ability to project a report or a quote. These can also be used to probe a clause
being analysed paradigmatically by examining close agnates. For example, if we meet clauses

376
 Analysis guide

in the ‘past’ tense, we can probe them by examining an agnate version in the unmarked
‘present’ tense, as with:

Going into his house he cut slices of bread and spread honey upon them. (Anderson,
Winesburg Ohio: Hands)

The unmarked ‘present’ tense version of he cut slices of bread would be he is cutting slices of bread (in
response to what’s happening?) rather than he cuts slices of bread, which might occur in the stage directions
of a play or a cooking demonstration (or if primed to be interpreted as a habit: His morning routine
is always the same. He cuts slices of bread, then spreads honey upon them.). Similarly, we can examine the
agnate thematic equative, e.g. he spreads honey upon them : what he does with honey is spread it upon them,
which suggests that honey serves as Goal in a ‘material’ clause. The general principle is that we can
analyse a clause by probing it paradigmatically to determine its systemic neighbourhood.
For each process type, there will be cases where an alternative analysis might seem tempting
or even plausible. For example:

Is On one side, Woodard Road winds 10 miles from U.S. 17 outside Windsor a ‘material’
clause of motion through space or a ‘relational’ one of location in space? – Arguably,
‘relational’: the present tense selection is the simple ‘present’, not present-in-present
(contrast: Woodard Road is winding 10 miles from . . .); and the clause does not actually
construe movement – the Process: verbal group winds represents the shape of the road,
not its movement (a reminder that we must deal with lexical metaphor).
Is My name escapes me a ‘material’ clause or a ‘mental’ one? – Arguably, ‘mental’: again,
the present tense (my name escapes me rather than my name is escaping me), but also the fact
that agnates are ‘mental’ rather than ‘material’, e.g. I forget my name, the answer to the question
escapes me and that it can be configured with a project clause, e.g. (in the sense of ‘I don’t
understand’) Like you, it completely escapes me that there were enough people to carry Romney
as far as he got and I have reserves about the accuracy. And probing further, we can begin to
explore a field of lexical metaphor where cognition is mapped onto movement away from:
it eludes me, it evades me, it slips my mind, and the opposite: it comes to me, it comes to my mind,
it springs to my mind, it enters my mind. (The wording My name escapes me was used by Alec
Guiness as the title of one of his autobiographical volumes, with the subtitle The diary of a
retiring actor. Compare the title of James Mason’s autobiography Before I forget.)

(And there are cases that are systemic in the sense that they reflect diferences in where linguists
draw descriptive lines, as in the case of ‘material’ vs. ‘relational’ in Halliday’s description vs.
Fawcett’s, e.g. Halliday and Matthiessen, 1999, 2014, and Fawcett, 1987; for a more general
comparison, see Bartley, 2018.)
Perhaps the most discussed case in the SFL literature is that of ‘behavioural’ clauses – their sys-
temic status as a primary process type or as a more delicate subtype of another process type, typi-
cally ‘material’, and their boundaries based on their properties. They were identified as one of the
six primary process types for the first time in Halliday’s first edition of his Introduction to Functional
Grammar, so in terms of his successive refinements of the descript of the system of process type in
English, this constituted an advance, and subsequent editions have retained their status as one of the
six primary process types, although a minor one rather than a major one. In Matthiessen (1995a),
they are treated as a subtype of ‘material’ clauses, although this clearly does not reflect their value in
the overall system, simply because that’s how they were described in the “Nigel grammar”, origi-
nally due to a description prepared by Halliday in the late 1970s. Banks (2016) surveys different

377
 Analysis guide

accounts (see also Matthiessen, 2018a), suggesting that ‘behavioural’ clause should be interpreted
as a subtype of ‘material’ clauses rather than as distinct primary process type, but his contribution
is concerned with alternative descriptions in the first instance rather than with the properties of
‘behavioural’ clauses (whether overt or covert). Complementing Banks (2016) and Matthiessen
(2018a), we have provided an overview of a number of criteria for viewing ‘behavioural’ clauses
against the background of prototypical ‘material’ and ‘mental’ clauses in Table 7.9 (and we could
actually add ‘verbal’ clauses to the table). As always in text analysis, two key considerations are the
consistency in the analysis of instances and the potential to reassign the analysis of these instances in
the view of new insights into the description of the system of process type. Thus if ‘behavioural’
clauses are recorded as a distinct category in the analysis of a text, it will still be possible at a later
stage to represent them as one of the six primary process types or as a subtype of ‘material’ process
clauses. Somewhat informally: this is no big deal!
In the transitivity structure of the clause, the configuration of process + participant(s) – the
domain of the system of process type – may be augmented by circumstantial elements. This is
the system of circumstantiation, circumstantial transitivity. Unlike participants, circumstances
are not inherence in clauses of a given process type; they are not directly involved in the process,
but rather indirectly involved in it or “attendant” on it. The indirect involvement of circum-
stances is reflected in the fact that they are realized by prepositional phrases (e.g. Time: in 1791,

Table 7.9 Behavioural clauses contrasted with material and mental ones

Material Behavioural Mental

key participant role Actor Behaver Senser
constraints – + conscious
manipulable 2nd do to/with Goal –
participant
what she did to/with the what she did to/with the what she did to/with
ball was throw it bird was watch it the bird was see it
configured with act- – √
clause as Range
participant
she watched [[the bird she saw [[the bird
eating the crumbs]] eating the crumbs]]
projection of idea – √
she threw → that the she watched → that the she thought → that
birds had eaten the birds had eaten the the birds had eaten
crumbs crumbs the crumbs
unmarked present present-in- she’s throwing the ball she’s laughing
tense present
(simple) she laughs she sees the bird
present
present tense close – √ [perceptive]
to modulation:
readiness: potential
she’s throwing the she is watching the she sees the bird ≠
ball ≠ bird ≠ she can see the bird
she can throw the ball she can watch the bird

378
 Table 7.10 Summary of characteristic of terms in the system of process type
PROCESS Approximate Approximate Meaning Participants Properties Examples of uses Examples
TYPE frequency proportion of
verbs

material 40% 62% doing-&-happening: Actor; unmarked present event line in Tropical Cyclone Haima
actions, activities, Actor + Goal; tense: present- stories and in will move across the
events Actor + Goal + in-present; recounts, event northern part of Luzon.
Recipient or pro-verb: happen line in sequential One day, his father gave
Client; (to); do + to/ explanations, him a huge bag of nails.
Actor + Scope with procedures, Throughout history, power
direction and has more often than not
collaboration in been wielded through
‘doing’ contexts terror.
(language in action) They’ll continue training
troops in the USA.

Analysis guide
behavioural 3.5% 10% behaving: Behaver; unmarked present characterization in they’ll watch the nuns
physiological or Behaver + tense: present- stories in terms [[that got away]]
379

mental activity Behaviour; in-present; of activities (e.g. The secretary watched

Behaver + act-clauses as accompanying [[him disappear into the
Phenomenon Phenomenon speaking in dramatic night]] with a smile of not
dialogue) unkindly irony
mental 11% 9% sensing: perceiving, Senser; unmarked present inner life of characters You hear the screams from
knowing, wanting, Senser + tense: (simple) in narratives, paradise
feeling Phenomenon present self-representation Mrs. Hamlyn felt [[the
act-clauses as in conversation, hot tears flow down her
Phenomenon; subjective cheeks]].
can project ideas attitudes in casual I admire the fact [[that he’s
(typically conversation and still out there hitting the
reported); in reviews, beliefs ball]].
emotive clauses and hypotheses in She had loved him dearly
often come academic discourse He suspects - you’re
in pairs (like ~ innocent.
please)

(Continued)
 Table 7.10 (Continued)
PROCESS Approximate Approximate Meaning Participants Properties Examples of uses Examples
TYPE frequency proportion of
verbs
verbal 8% 6% saying Sayer; unmarked present quoting in dramatic The boy replied, - “a hole
Sayer + Verbiage; tense: (simple) dialogue in stories, in the fence!”
Sayer + Receiver; present; quoting and You mean to tell me -
Sayer + Receiver can project reporting in news, your keen perceptive eyes
+ Verbiage locutions and in academic have only just figured out
(either quoted discourse after all these years?
or reported) Historians of terrorism
may point out - that
the word “terror” applies
to the state terror of the
French Revolution.
relational 36% 14% being-&-having Carrier + unmarked present taxonomies The parents of the little boy
Attribute; tense: (simple) (classifications), were very depressed due

Analysis guide
Identified + present; characterizations, to his bad temper.
Identifier verb serving topographic reports The Occipital Lobe is the
380

as Process (verbal maps), visual processing center

typically descriptions in of the brain. It contains
unaccented stories, definitions; most of what is referred
(e.g. be, have); evaluations in gossip, to as the “visual cortex”.
always two opinion texts, Its primary target is the mind.
participants reviews Terrere means “to make
(the ones in the tremble” in Latin.
relationship) No, I haven’t any close
relations.
existential 2% 2% being (existing) Existent unmarked present presentation of There’s a sunny little
tense: (simple) characters in stories, country south of Mexico.
present; the introduction of There will be frequent
most common places of interest squalls and heavy rain.
version being in topographic there shouldn’t be any
there + be + procedures, pain.
Existent representation of
certain ambient
conditions
1
For frequency and proportion of verbs, see Matthiessen (2014c), “Extending the description of process type within the system of transitivity in delicacy based on Levinian verb classes” Functions of
Language 21(2): 139–175.
 Analysis guide

Place: in Sydney and Role: as a member of . . . in In 1791 John Macarthur arrived in Sydney as a
member of the New South Wales Corps) or adverbial groups (e.g. Manner: quality: successfully in
soon he was successfully growing vegetables, grain crops, fruit and vines) rather than by nominal groups
(with the exception of nominal groups of measurement serving as temporal or spatial extent, e.g.
Distance: thousands of miles in they can be carried thousands of miles by gentle currents). A prepositional
phrase is like a nominal group mediated by a minor process (the preposition) in its function in
the transitivity structure of the clause. In fact, prepositional phrases can be interpreted as mini-
clauses on the model of Process: preposition + Range: nominal group. This interpretation helps
us see that by another paradigmatic step, they can be related to non-finite clause condition-
ing other clauses in hypotactic clause nexuses. Adverbial groups can be interpreted as oblique
nominal groups with Epithet as Head. The different types of circumstance can be interpreted
in terms of the fractal types of ‘expansion’ (elaborating, extending, enhancing) and projection,
and are summarized in Table 5–28 in Halliday and Matthiessen (2014: 313–314) together with
Wh-items that can be used as probes and common prepositional realizations. In the analysis of
prepositional phrases, Table 7.15 in Section 7.3.4 can be consulted. It specifies different senses of
prepositions corresponding to different types of circumstance (and participant, where relevant, as
in the case of Agent). For the analysis of adverbial groups, see Table 7.13 in Section 7.3.3. This
table lists adverbs that can function, in different senses, as conjunctive, modal or circumstantial
Adjuncts, or as (Sub-)Modifiers.

7.3 Vocabulary guides

7.3.1 Introduction
In undertaking lexicogrammatical analysis, we work trinocularly (as for the clause displayed
schematically in Figure 7.3), of course, taking into consideration information “from below” –
from the rank or ranks below within grammar and also from stratum below, from phonol-
ogy of graphology, “from above” – from semantics, and from roundabout, from the point
of view of the system or systems of the grammatical unit being analysed. In terms of the
cline of delicacy, these systems can be approached from the grammatical end of the cline.
Here they will most likely have been described systemically, so we can access systems in the
system network that they are part of and the realization statements associated with terms
in these systems. At the same time, the systems can also be approached from the lexical
end of the cline of delicacy. But here it is most likely that they haven’t yet been described
systemically, with certain exploratory exceptions such as Hasan (1987) and Tucker (1998);
at best, we will have references to lexical sets. So instead, we will have to rely on standard
dictionaries and also thesauri, and any other lexical accounts that can be helpful such as
Levin’s (1993) description English verb classes.
Dictionaries and thesauri are most likely to provide useful information for lexical items; they
often fail to provide clear entries for grammatical items. For lexical items, experiential senses
or denotations are likely to be foregrounded; although they will take account of interpersonal
senses of purely interpersonal items and of connotations of lexical items that fuse experiential
and interpersonal senses, it will be helpful also to consider lists of lexical items provided by Mar-
tin and White (2005) in their description of the system of appraisal in English.
When we conduct lexicogrammatical analysis of texts, we will have to determine what classes
lexical and grammatical items belong to and what functions they serve in the function structures
of different units. We will meet these items in their graphological manifestation – the nature
of which will depend on the writing system used for the language we are dealing with, e.g.

381
 Analysis guide

Table 7.11 Classification of dictionary subentries for “suggest” according to process type

Entry: gloss Verbal Relational

“put forward for [with clause] [α:] I suggest [“β:] that we wait a –
consideration” day or two
[with direct [“1:] “Maybe you ought to get an –
speech] expert,” [2:] she suggested.
[with object] [Sayer:] Ruth [Process:] suggested –
[Verbiage:] a vacation.
“cause one to think [with clause] – [Token:] the bright sun
that (something) [Process:] suggested
exists or is the case” [Value:] that the temperature
might be tropical1
[with object] – [Token:] finds of lead coffins
[Process:] suggested
[Value:] a cemetery north of
the river
“state or express [with clause] [α:] are you suggesting [“β:] that I –
indirectly” should ignore her?
[with object] – [Token:] the seduction scenes
[Process:] suggest [Value:]
his guilt and her loneliness.
“evoke” [with object] – [Token:] the theatrical
interpretation of weather and
water [<Adjunct:>] almost
[Process:] suggests [Value:]
El Greco.
1
We have revised the example to make it a clearer illustration of the intended sense of suggest.

orthographic words in English or Arabic and characters (essentially orthographic morphemes)

in Chinese. If a dictionary or dictionaries have been produced for the language, we can look
up this item, e.g. suggest in English: see Figure 7.17. Here we find that the orthographic word
suggest represents a lexical item of the grammatical class of verb, and that this verb has four dis-
tinct senses (setting suggest itself aside).5 In terms of process type, they can be assigned to ‘verbal’
clauses or ‘relational’ ones: see Table 7.11.
The fact that suggest has senses that are either ‘verbal’ or ‘relational’ in nature is no isolated
coincidence. Rather it reflects an area of overlap between these two process types, an area
having to do with semiosis and the models provided by the grammar of many languages –
semiosis and saying and semiosis as being, or (more specifically) indication. Thus a number of
verbs in English have both ‘verbal’ and ‘relational’ senses, e.g.

demonstrate, indicate, prove, show, suggest, tell

Interestingly, Christie and Derewianka (2008) have demonstrated that the relational senses tend
to be learned much later than the verbal ones; they are likely to appear well into the years of

382
 Analysis guide

frown | froun |
verb [ no obj. ]

furrow one’s brow in an expression of disapproval, displeasure, or concentration: he frowned as he

reread the letter.

(frown on/upon) disapprove of: the old Russian rural system frowned on private enterprise.

noun

a facial expression or look characterized by a furrowing of one’s brows: a frown of disapproval.

DERIVATIVES

frowner noun

frowningly adverb

ORIGIN

late Middle English: from Old French froignier, from froigne ‘surly look,’ of Celtic origin.

Figure 7.18 Dictionary entry for “frown” (Apple’s OS X Dictionary)

high school as part of the discourses of particular subjects. But such verbs are still, of course,
likely to slow text analysts down since they will have to decide on which interpretation is more
appropriate in a given passage of text – ‘verbal’ or ‘relational’.
The orthographic word suggest can be compared with the orthographic word frown; the dic-
tionary is shown in Figure 7.18. Unlike suggest, frown can realize lexical items of either the class
of verb or the class of noun; it is an example of the common phenomenon of “zero derivation”
in English (cf. smile, laugh, grimace, cry). Compare frown verb : frown noun with suggest verb :
suggestion noun. However, unlike suggest, the verb frown has only one dictionary sense belong-
ing to one process type: it serves as the verb in ‘behavioural’ clauses. Like the verbal sense of
suggest (but unlike the relational suggest), it can occur in ‘ranged’ constructions as a noun with a
very general lexical verb: suggest : make a suggestion (cf. make a statement, make an announcement),
frown : give a frown (cf. also give a smile, give laugh, give a cry); and unlike suggest it can occur in
the “cognate object” variant of this construction, e.g. frown a morose frown.
In text analysis, a lexical item such as suggest will thus pose more problems to the analyst than
a lexical item such as frown. While frown can be either verb or noun, these will be syntactically
quite distinct in their patterning even though they are morphologically the same, and the more
significant challenge is differentiating among verb senses belonging to different process types, as

383
 Analysis guide

is the case with suggest. It is probably no coincidence that dictionaries differentiate more verb
senses for suggest than for frown: the more frequent an item is, the more senses it is likely to have
developed. Thus the verb suggest is many times more frequent than the verb frown, as is shown by
the graph produced by Google’s Ngram Viewer in Figure 7.19. This is true even if we include
frown as a noun; even though it is much more frequent than frown as a verb, suggest is still much
more frequent.
Based on the tendency for more frequent items to develop more senses, we can make a prin-
cipled selection of very frequent ones known to cause problems in text analysis: in this chapter,
we have included a small sample of such frequent items.
Before we provide a guide to a sample of very frequent lexical and grammatical items to help
differentiate senses in text analysis, let us give a systemic illustration of the potential polysemy
of orthographic words, using will as an example: see Figure 7.20. As always, one way of deter-
mining the relevant sense of a word (i.e. an orthographic word) is to probe it paradigmatically,
identifying close agnates. For example:

noun: (1) ‘legal document’ testament, bequest; (2) ‘mental faculty’ determination, resolve; desire,
wish, intention
verb: lexical: (1) ‘bequeath’ bequeath, leave, give, transfer; (2) ‘desire’ want, wish, desire
verb: grammatical (operator): (1) ‘temporal operator’ be going to; (2) ‘modal operator’ (2.1)
‘probability’ probably; (2.2) ‘usuality’ usually; (2.3) ‘inclination’ be determined to

7.3.2 Verbs
In carrying out experiential analysis of a clause in terms of the system of transitivity, analysts will
find it helpful to focus on the Process early on in the analysis (see Section 7.2.6), and to determine
whether it is realized by a simple verbal group or a verbal group complex, either a hypotactic one or
a paratactic one. This obviously involves identifying the lexical verb serving as the Event of each ver-
bal group. While transitivity analysis depends on the configuration of elements serving in transitivity
roles – in particular, to begin with on the configuration of Process + Medium, and also of Process +
Range – rather than on the Process alone (cf. Matthiessen, 1995a), identifying the possible process
type based on the sense or senses of the lexical verb is an important aspect of the analysis.
Perversely (but very naturally!), common lexical verbs in a language are very likely to have
evolved a number of different senses (cf. the example of suggest earlier), and the different senses
often derive from different process types, and also from verbs serving as “catenatives” in hypotactic
verbal group complexes. We have listed 136 verbs in Table 7.12, classifying their (main) diction-
ary senses according to their grammatical uses as auxiliary verb or as lexical verb, and for lexical
verbs, as catenative or as lexical verb reflecting one of the six process types. For example, admit has
five different uses, all as lexical verb: (1) as a catenative in hypotactic verbal group complexes of
projection; (2) as a verb of doing in ‘material’ clauses with the sense of ‘allow (someone) to enter
a place’ (with various special senses like admitting a person to hospital); (3) as a verb of sensing in
‘mental’ clauses with the sense of ‘accept as valid’; (4) as a verb of saying in ‘verbal’ clauses with
the sense of ‘confess to be true or to be the case, typically reluctantly’, ‘acknowledge’; and (5) as
a phrasal verb of being (admit of) in ‘relational’ clauses with the sense of ‘allow for the possibility
of ’. Serving as the Event in the verbal group (complex) realizing the Process of a clause, admit
will occur in different configurations in the different uses, so probing the relevant sense of admit
includes checking its grammatical environment in the clause it serves in. For example, in

384
 Analysis guide
385

0.00800%
0.00700% suggest_VERB
0.00600%
0.00500%
0.00400%
0.00300%
0.00200%
0.00100%
frown_NOUN
0.00000% frown_VERB
1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000

Figure 7.19a,b Frequency of (orthographic words) “suggests”, “suggested” and “frowns”, “frowned” 1940–2000 according to Google’s Ngram Viewer
 Analysis guide

Figure 7.20 From orthographic word to grammatical word – class assignment, sense and function

||| I can’t possibly admit || that Captain Raib was based on him, || because if I ever do
go to Grand Cayman, || he’d put me in jail. |||

the verb admit functions as Event in the verbal group [Finite] can’t . . . [Event:] admit; and the
clause that this verbal group operates in, I can’t possibly admit, occurs in a clause nexus of projection,
with the projected clause that Captain Raib was based on him, which indicates that I can’t possibly
admit is a ‘verbal’ clause and the sense of admit is ‘confess to be true or to be the case, typically
reluctantly’. This is backed up by the fact that the nominal group I can be interpreted as Sayer,
and a nominal group interpretable as Receiver could be added: I can’t possibly admit to you. In
addition, we can probe examples paradigmatically by considering closely related verbs in the
various senses of polysemous verb. Here acknowledge would be an alternative, but the paraphrase
allow to enter is not. So it is helpful to examine the “synsets” (in the WordNet sense) of each sense
of the verb being analysed.

7.3.3 Adverbs and conjunctions

Most conjunctions and many adverbs are close on the cline from grammatical items to lexical
items to the grammatical pole of the cline, and have evolved more than one sense. In fact, some
items have uses either as adverbs or as conjunctions. Common polysemous adverbs and conjunc-
tions are listed in Table 7.13.

386
 Table 7.12 Common polysemous verbs

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

add √ ‘join to sth else’; √ ‘say as a further

‘put together to remark’
calculate total
value’
admit √ [projection: √ ‘allow (sb) to √ ‘accept as √ ‘confess to be
proposition: enter a place’ valid’ true or to be the
locution] case, typically
reluctantly’;
‘acknowledge’
agree √ [projection: √ ‘have the same √ ‘concur’ agree with √ ‘be

Analysis guide
proposal: opinion’ consistent with’
consenting]
387

allow √ [modulation: √ ‘permit sb to have √ ‘concede’ √ ‘admit as legal’

agency & sth’; ‘set aside
causative] sth for a specific
purpose’
anticipate √ [projection: √ ‘expect, √ ‘be a forerunner
proposal: predict’; or precursor of ’
intention] ‘look forward
to’
appear √ [reality-phase: √ ‘come into sight; √ ‘seem; give the √ ‘come into
apparent] become visible impression of sight; become
or noticeable, being’ visible or
typically without noticeable,
visible agent or typically
apparent cause’ without visible
agent or
apparent cause’
(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

ask √ [projection: √ ‘say to get an

proposal: answer’; ‘request’
demanding]
aspire √ [projection: √ ‘direct hopes
proposal: or ambitions
expectations] towards’
assure √ ‘cover (a person) √ ‘tell sb sth √ ‘make (sth)
with life positively or certain to
insurance’ confidently to happen’
dispel any doubts
they may have’
attempt √ [conation: √ make an effort

Analysis guide
conative] ‘make to achieve or
an effort to complete (sth,
388

achieve or typically a
complete (sth, difficult task or
typically a action) ; ‘try to
difficult task or climb’
action)’
avoid √ [conation: √ ‘try not to meet’;
conative] ‘not go through’;
‘prevent’
be √ [tense: √ be like ‘say’ √ ‘having the state, √ ‘exist’, ‘be
secondary quality, identity, present’
present be nature, role,
v-ing] etc., specified’;
√ [voice: passive ‘occur, take
be v-en] place’
become √ ‘begin to be’;
‘suit, look
good on’; ‘be
appropriate or
suitable to’
 begin √ [time-phase: √ ‘start; perform or √ ‘start speaking by
inceptive] undergo the first saying’
part of (an action
or activity)’;
‘originate’
begin by √ [modulation:
time] ‘do first’
believe √ ‘accept (sth) as √ [assignment:
true’; ‘feel sure projection]
of the truth
of’; ‘hold
(sth) as an
opinion; think
or suppose’;
‘feel sure that
(sb) is capable

Analysis guide
of a particular
action’
389

break √ {many!} √ (sb’ heart)

break in √ ‘force entry to a √ ‘interrupt’
building’
cause √ ‘make (sth, √ ‘make (sth, √ ‘make (sth,
typically sth typically sth bad) typically sth
bad) happen’ happen’ bad) happen’
cease √ [time-phase: √ ‘bring or come to
inceptive] ‘come an end’
to an end, stop’
choose √ [projection: √ ‘pick out or select
proposal: (sb or sth) as
intention] being the best or
‘decide on most appropriate
a course of of alternatives’
action, typically
after rejecting
alternatives’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

circle √ ‘move all the way √ ‘be around’,

around (sb or ‘form a ring
sth), especially around’
more than once’;
‘draw a line
around’
claim √ [projection: √ ‘make a demand √ ‘state or assert √ ‘cause the loss
proposition: for (money) that sth is the of ’
claiming] under the terms case, typically

Analysis guide
of an insurance without providing
policy’ evidence or proof ’
390

come √ [modulation: √ ‘move toward the √ ‘occur to’ √ ‘become’; ‘be √ ‘appear’
concession] speaker’; ‘occur, available’;
‘do contrary to happen’ ‘extend to’;
expectation’ ‘be in a certain
position’
consent √ [projection: √ ‘give permission
proposal: for sth to happen’
consenting]
consider √ [projection: √ ‘look attentively √ ‘look √ ‘believe, think’ √ ‘regard’
proposal: at’ attentively
intention] at’
‘think about
and be drawn
towards (a
course of
action)’
 contain √ ‘control or √ ‘have or hold
restrain (oneself (sb or sth)
or a feeling)’; within’
‘prevent (a
severe problem)
from increasing
in extent or
intensity’
continue √ [time-phase: √ ‘persist in an √ ‘carry on speaking
durative] ‘persist activity or after a pause or
in an activity or process’; ‘carry interruption’
process’ on with (sth

Analysis guide
that one has
begun’; ‘carry on
391

traveling in the
same direction’;
‘recommence
or resume after
interruption’
contrive √ [conation: √ ‘create’
conative]
create √ ‘bring into √ ‘invest (sb) with
existence’ a new rank or
title’; ‘cause
(sth) to happen
as a result of
one’s actions’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

cross √ ‘go or extend √ ‘anger’; √ ‘be across’; ‘pass

across or to the ‘oppose or in an opposite
other side of stand in the or different
(a path, road, way of (sb)’ direction;
stretch of water, intersect’
or area)’; ‘pass in
an opposite or
different direction;
intersect’; ‘cause
(two things) to
intersect’; ‘draw

Analysis guide
a line or lines
across’; ‘mark
392

with a cross’;
‘cause (an animal
of one species,
breed, or variety)
to interbreed with
one of another
species, breed, or
variety’
decide √ [projection: √ ‘make a choice √‘come to a
proposal: from a number of resolution
intention] alternatives’ as a result of
‘come to a consideration’;
resolution ‘come to a
as a result of decision
consideration’ about (sth)’
√ [projection: √ ‘cause (sb) to
proposal: come to a
intention & resolution’
causative]
 decline √ [projection: √ ‘politely
proposal: refuse’; ‘move
consenting downward’;
(negative)] ‘become smaller’
demand √ [projection: √ ‘ask √ ‘require, need’
proposal: authoritatively
demanding] or brusquely’;
‘insist on
having’
demonstrate √ ‘take part √ ‘give a practical √ ‘clearly show
in a public exhibition and the existence or
demonstration’ explanation of truth of (sth) by
(how a machine, giving proof or’
skill, or craft

Analysis guide
works or is
393

performed)’;
‘show or express
(a feeling or
quality) by one’s
actions’
desire √ [projection: √ ‘strongly wish
proposal: for, want’
desideration]
determine √ ‘ascertain √ ‘cause (sth)
or establish to occur in a
exactly, particular way;
typically as be the decisive
a result of factor in’
research or
calculation’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

develop √ ‘grow or cause

to grow’; ‘start
to exist’; ‘treat
(a photographic
film) with
chemicals to
make a visible
image’
do √ √ many senses, √ ‘be suitable,
including acceptable,
substitute do and usable’

Analysis guide
“vector verb” do
394

(i.e. in the very

general sense of
‘perform’ + a
nominal group
specifying the
lexical content
serving as Scope/
Range)
drip √ ‘let fall or be so √ ‘be bad’
wet as to shed
small drops of
liquid’; ‘cause or
allow (a liquid)
to shed small
drops’; ‘display a
copious amount
or degree of a
particular quality
or thing’
 elect √ ‘choose to do’ √ ‘choose in √ [assignment]
preference to’ ‘choose as’
elude √ ‘evade, escape √ ‘fail to be
from’; ‘fail to be understood
attained by’ or
remembered’
enable √ [potentiality: √ ‘make operational’ √ ‘make possible’;
potential & ‘cause’
causative]
encourage √ [conation: conative √ ‘help develop’ √ ‘give √ ‘cause’,
& causative] confidence’ ‘stimulate’
end up (by) √ [modulation: √ ‘eventually come
time] ‘do last’ to’
ensue √ ‘happen or √ ‘happen
occur afterward or occur

Analysis guide
or as a result’ afterward or as
395

a result’
erupt √ ‘(of a volcano) √ ‘give vent √ ‘(of an object)
become active to anger, explode with
and eject lava, ash, enthusiasm, fire and noise
and gases’; ‘be amusement, resembling an
ejected from an or other active volcano’
active volcano’; feelings in
‘(of an object) a sudden
explode with and noisy
fire and noise way’
resembling an
active volcano’
escape √ ‘break free’; √ ‘fail to be
‘elude’; ‘avoid’; noticed by’;
‘leak’ ‘fail to be
remembered
by’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

exceed √ ‘become greater √ ‘be more than’,
in number or size ‘be greater
than (a quantity, in number
number, or other or size than
measurable thing)’; (a quantity,
‘go beyond what number,
is allowed or or other
stipulated by (a set measurable
limit, especially of thing)’

Analysis guide
one’s authority)’;
‘surpass’
396

exist √ ‘live, especially √ ‘have objective

under adverse reality or
conditions’ being’
expect √ [projection: √ ‘regard as likely
proposal: to happen’;
expectation] ‘believe that
sb or sth
will arrive
soon’; ‘think,
suppose’
fail √ [conation: √ ‘break down’;
reussive] ‘not succeed’
fear √ [projection: √ ‘be afraid’
proposal:
fearing]
 feel √ ‘examine √ ‘be aware of √ ‘experience (an
or search (a person emotion or
by touch’ or object) sensation)’
through
touching
or being
touched’;
‘experience (an
emotion or
sensation)’;
‘have a belief or
impression,
especially
without an
identifiable
reason’, ‘hold
an opinion’
find √ ‘reach, arrive √ ‘ascertain √ ‘become √ ‘perceive to

Analysis guide
at’; ‘succeed by study’; aware of ’; be the case’;
397

in obtaining’; ‘discover ‘perceive by ‘declare to be

‘summon up after a chance’ the case’
with an effort’ deliberate
search’
flourish √ ‘(of a person, √ ‘develop
animal, or other rapidly and
living organism) successfully’
grow or develop
in a healthy or
vigorous way,
especially as the
result of a parti-
cularly favourable
environment’;
‘(of a person)
wave (sth) around
to attract the
attention of
others’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

follow √ ‘go after’, ‘go or √ ‘understand’ √ ‘be after’, ‘be √ ‘appear after’
come after (a after in time
person or thing or order’;
proceeding ‘happen after
ahead)’; ‘move (sth else) as a
or travel behind’; consequence’
‘come after in ‘be according to’
time or order’;
‘act according to
(an instruction or
precept)’
force √ [modulation: √ ‘break open by √ ‘achieve

Analysis guide
agency & force’; ‘push into by effort’,
causative] a position’; ‘make e.g. force a
398

do against will’ smile

forget √ [modulation: √ ‘forget to bring’ √ ‘fail to
reason] ‘not do remember’;
according to ‘cease to
intention’ think of ’
get √ [voice: √ [time-phase: √ ‘receive’; ‘go’; √ ‘understand’; √ ‘become’
passive: inceptive]; ‘catch’; ‘strike’; ‘baffle’;
mutative] √ [modulation: ‘punish’ ‘annoy’;
agency & ‘amuse’
causative]
go on √ [time-phase: √ ‘start working’; √ ‘talk at √ ‘continue
durative] ‘happen’; great speaking’
‘continue doing’ length’
grow √ [time-phase: √ ‘undergo natural’; √ ‘become
inceptive] ‘become larger gradually or
‘come to feel or or greater over a increasingly’
know sth over period of time’;
time’ ‘increase’
 happen √ [modulation: √ ‘occur, take place’ √ ‘come about
reason] ‘do by by chance’ it
chance’ happens that …
hasten √ [modulation: √ ‘move quickly’;
manner] ‘do ‘cause to happen
quickly’ sooner’
hate √ [projection: √ ‘feel intense
proposal: dislike for’
desideration]
have √ [tense: √ [modulation: √ ‘undergo’; √ ‘perform’ √ ‘possess’; ‘hold’
secondary: agency & ‘perform’; have a look
past have causative] ‘receive’ etc.
v-en]
help √ [modulation: √ ‘assist’; ‘serve sb’
accompaniment]

Analysis guide
‘do together with’
hesitate √ [modulation: √ ‘pause before [√ have + idea, √ ‘pause before
399

manner] ‘do doing’ doubt, fear saying’

reluctantly’ etc.]
hold √ [many!] √ ‘have or √ ‘(of a judge or √ ‘contain’, ‘contain
adhere to court) rule; or be capable
(a belief or decide’ of containing
opinion)’ (a specified
amount)’;
‘have in one’s
possession’;
‘be or remain valid
or available’;
√ [assigned]
‘consider (sb) to
be responsible
or liable for
a particular
situation’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

hope √ [projection: √ ‘want sth to

proposal: be the case’
expectation]
hurt (see √ ‘cause pain’; √ ‘distress’, √ ‘suffer pain’, ‘be
Halliday ‘undergo pain’ ‘cause distress’ painful’
1998a)
include √ ‘make part of a √ ‘comprise or
whole or set’ contain as part
of a whole’
indicate √ ‘point to’; ‘signal √ ‘mention’ √ ‘be a sign
an intention to of ’, ‘strongly
change lanes or suggest’;
turn using an ‘register a

Analysis guide
indicator’ reading of ’
400

insist on √ [modulation: √ ‘demand forcefully’

manner] ‘do
perversely’
intend √ [projection: √ ‘plan that’;
proposal: ‘mean’
intention]
keep √ [time-phase: √ ‘cause to continue √ ‘have or retain
durative]; [~ & in a certain way’; as possession’;
causative] ‘provide for’; ‘cause to
‘guard’, ‘protect’; continue to be’
‘support’; ‘honour’
know √ [potentiality: be √ ‘be aware √ ‘regard as’, ‘be
able to] (know of ’, ‘have called’
how to do) knowledge
(of)’, ‘be
familiar with’,
‘have good
command of ’,
‘have personal
experience of ’
 learn √ [potentiality: √ ‘acquire skill’ √ ‘commit to √ ‘come to
become able to] memory’ know’
let √ [modulation: √ ‘allow to pass’;
agency & ‘rent’
causative]
like √ [projection: √ ‘find
proposal: agreeable’;
desideration] ‘wish for,
want’
long √ [projection: √ ‘have a strong
proposal: wish or
desideration] desire’
look √ ‘direct √ ‘have the
one’s gaze appearance
toward sb or give the
or sth or in impression of

Analysis guide
a specified being’
401

direction’
make √ [modulation: √ ‘create’; ‘draw up’; √ ‘constitute’;
agency & ‘cause to exist’; ‘appoint’;
causative] ‘gain, earn’; ‘act ‘cause to be’
as if ’ etc.
manage √ [conation: √ ‘control’; ‘serve
reussive] as manager of ’;
‘succeed against
odds’
mean √ [projection: √ ‘intend to √ ‘represent’;
proposal: convey’; ‘be ‘have as a
intention] of importance consequence’
to’
mind √ [projection: √ ‘take care of √ ‘be annoyed
proposal: temporarily’ by’; ‘regard as
desideration] ‘be important’
reluctant to do’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

move √ ‘go in a specified √ ‘provoke √ ‘propose for

direction or a strong discussion and
manner’; ‘change feeling, resolution
position’; ‘change especially of at a meeting
or cause to sorrow or or legislative
change from one sympathy’ assembly’
state, opinion,
sphere, or activity
to another’;
‘make progress;
develop in a

Analysis guide
particular manner
or direction’
402

need √ [projection: √ ‘require sth

proposal: because it is
needing] essential’
oblige √ [modulation: √ ‘do as sb asks or √ ‘please’
agency & desires’
causative]
permit √ [modulation: √ ‘authorize’;
agency & ‘make possible’;
causative] ‘present no
obstacle’
plan √ [projection: √ ‘design’, ‘arrange’ √ ‘decide in
proposal: advance’
intention]
play √ [many!] √ ‘represent (a
character) in
a theatrical
performance
or on film’
 possess √ ‘(of a demon or √ ‘have as
spirit, especially belonging to
an evil one) have one’; ‘own’
complete power
over (sb) and
be manifested
through their
speech or actions’
practise √ [potentiality: √ ‘perform to
become able to] improve’; ‘per-
form regularly’;
‘pursue, be
engaged in’
pretend √ [projection: √ ‘simulate’ √ ‘speak so as to
proposition: make it appear’;

Analysis guide
pretending] ‘lay claim to’
profess √ [projection: √ ‘received into a √ ‘claim openly but
403

proposition: religious order often falsely’;

claiming] under vows’ ‘affirm one’s faith
in or allegiance to’
promise √ [projection: √ ‘pledge to marry’ √ ‘assure sb’; √ ‘give good
proposal: ‘announce as grounds for’
promising] being expected to
happen’
prove √ [reality-phase: √ ‘(of bread dough) √ ‘demonstrate √ ‘be seen or
realized] become aerated the truth or found to be’;
by the action of existence of (sth) ‘demonstrate by
yeast’; ‘rise’ by evidence or evidence or
argument’ argument (sb or
sth) to be’
provide √ ‘make available √ ‘stipulate in a will
for use’ , ‘supply’ or other legal
document’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

refuse √ [projection: √ ‘not give’; ‘not √ ‘indicate or show

proposal: jump’ that one is not
consenting willing to do sth’
(negative)]
√ [conation:
reussive]
regret √ [modulation: √ ‘feel sad,
manner] ‘do sadly’ repentant’
remain √ ‘continue √ ‘continue to
to possess exist’
a particular

Analysis guide
quality or fulfil
a particular
404

role’; ‘stay in
the place that
one has been
occupying’; ‘be
left over after
others or other
parts have been
completed, used,
or dealt with’
remember √ [modulation: √ ‘convey greetings √ ‘recall’; ‘keep √ ‘convey greetings
reason] ‘do from one person in mind (as to’
according to to another’ important)’
intention’ (remember sb to
sb); ‘recover
one’s manners’
(remember oneself)
remind √ [modulation: √ ‘cause to √ ‘bring to the
reason & remember’; attention of ’
causative] ‘cause to think’
 request √ [projection: √ ‘politely or
proposal: formally ask for’
demanding]
require √ [modulation: √ ‘need’; ‘specify
agency & as compulsory’
causative]
√ [projection:
proposal:
needing]
run √ ‘move at a speed √ ‘become’; ‘cost’
faster than walk’;
‘flow or cause to
flow’; ‘make a
regular journey’;
‘manage’; ‘be
or cause to be
in operation’;

Analysis guide
‘smuggle’
405

say √ [projection: √ ‘assume’ (let’s √ ‘indicate’; ‘report’

proposition: say that …)
hearsay] be said
see √ ‘meet’; ‘escort’ √ ‘perceive √ ‘ensure’ (e.g. see
(e.g. see sb out) with eyes’; to it that)
‘understand’
seem √ [reality-phase: √ ‘give the impres- √ ‘appear to exist’
apparent] sion or sensation
of being sth or
having a par-
ticular quality’;
‘(it seems or it
would seem)
used to suggest
in a cautious,
guarded, or polite
way that sth is
true or a fact’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

show √ ‘offer sth for √ ‘display or allow to √ ‘be visible’;

inspection’; be perceived’ ‘represent
‘put on display’; in art’;
‘present for ‘demonstrate,
viewing’; prove’
‘become visible’
smell √ ‘emit an odour √ ‘sniff at’, √ ‘perceive or √ ‘emit an odour
or scent of a ‘sniff at (sth) detect the or scent of a
specified kind’ in order to odour or specified kind’,
perceive or scent of (sth)’ ‘have a strong
detect its or unpleasant
odour or odour’

Analysis guide
scent’
406

sound √ ‘emit or cause to √ ‘convey a

emit sound’ specified
impression
when heard’
start √ [time-phase: √ ‘come into being’; √ ‘give a
inceptive]; ‘begin or be small jump
[~ & causative] reckoned from a or make a
particular point sudden
in time or space’; jerking
‘cause (an event movement
or process) to from
happen’ surprise or
alarm’
stink √ ‘emit smell’; √ ‘have a strong
[stink a place up] unpleasant
‘fill a place with a smell’; ‘be very
strong unpleasant unpleasant,
smell’ contemptible,
or scandalous’
 stop √ [time-phase: √ ‘(of an event,
inceptive]; [~ & action, or
causative] process) come to
an end’; ‘cease
to happen’;
‘cause (an action,
process, or event)
to come to an
end’; ‘block or
close up (a hole
or leak)’
strive √ [conation: √ ‘try to achieve,
conative] obtain’
succeed √ [conation: √ ‘achieve result’; √ ‘follow’, ‘come

Analysis guide
reussive] ‘take over from’; after and take
‘attain fame, the place of ’
407

wealth or status’
suck √ ‘draw into the √ ‘be very bad,
mouth by disagreeable, or
contracting the disgusting’
muscles of the
lip and mouth
to make a partial
vacuum’; ‘draw
in a specified
direction by
creating a vacuum’
supply √ ‘make (sth needed √ ‘be adequate
or wanted) to satisfy (a
available to sb’; requirement or
‘provide’ demand)’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

take √ ‘lay hold of sth √ ‘understand √ ‘regard or

with one’s hand’; or accept view in a
‘reach for and as valid’; specified way’;
hold’; ‘remove ‘tolerate, ‘accommodate’;
from a particular stand’; ‘require as part
place’; ‘consume’ ‘assume’ (take of ’ [grammar]
etc.; ‘carry or it that)
bring with one’;
‘accept or receive’
etc.; ‘perform (an
action of task)’
taste √ ‘eat or drink a √ ‘sample or √ ‘perceive or √ ‘have a specified

Analysis guide
small portion of ’ test the experience flavour’
408

flavour of the flavour

(food or of ’; ‘have
drink) by experience
taking it into of ’
the mouth’
teach √ [potentiality: √ ‘instruct’; ‘work √ ‘cause to √ ‘show, explain’ √ ‘encourage the
achieval & as a teacher’ know or acceptance of
causative] learn’ sth as a fact or
principle’
tell √ ‘decide or √ ‘communicate √ ‘reveal
determine information, facts, information’,
correctly or or news to sb’; ‘indicate’,
with certainty’; ‘order, instruct, ‘convey’
‘distinguish or advise (sb) to
(one person do sth’; ‘narrate
or thing) from or relate (a tale
another’; or story)’; ‘reveal
‘perceive (the (information) to sb
difference) in a nonverbal way’
 between one
person or thing
and another’
tend √ [modulation: √ ‘go or move
time] ‘do in a particular
typically’ direction ‘;
[to, towards]
‘be liable to
possess or display
(a particular
characteristic)’
thank √ ‘express gratitude
to sb’; ‘request
of command,
implying reproach
or annoyance’

Analysis guide
threaten √ [projection: √ ‘seem likely to √ ‘state one’s √ ‘cause to be
409

proposal: occur’ intention to take vulnerable’,

promising] hostile action’ ‘endanger’
throw √ [many!] √ ‘have’ (throw √ ‘surprise’,
a tantrum, ‘disconcert’;
fit) ‘confuse’
try √ [conation: √ ‘test’; ‘subject to √ ‘make severe
conative] trial’ demands on
√ [modulation: (a person or a
purpose] ‘do as quality, typically
means to an end’ patience)’
turn √ ‘move or cause to √ ‘become’,
move in a circular ‘change in
direction wholly nature, state,
or partly around age, form, or
an axis or point’; colour’
‘shape (sth) on a lathe’

(Continued)
 Table 7.12 (Continued)

Item Auxiliary Catenative Lexical verb — PROCESS TYPE

Material Behavioural Mental Verbal Relational Existential

understand √ ‘perceive the

intended
meaning of ’;
‘infer sth from
information
received’;
‘assume to be
the case, take
for granted’
undertake √ [projection: √ ‘take on’; ‘catch √ ‘guarantee or

Analysis guide
proposal: up with and pass’ affirm’; ‘give as a
promising] formal pledge’
410

upset √ ‘knock (sth) over’; √ ‘make (sb)

‘cause disorder in unhappy,
(sth)’; ‘disrupt’ disappointed,
or worried’
venture √ [modulation: √ ‘dare to do’; √ ‘dare to say’
manner] ‘do ‘expose to the
tentatively’ risk of loss of ’
vow √ [projection: √ ‘solemnly promise’
proposal:
promising]
want √ [projection: √ ‘desire to √ ‘be short of,
proposal: possess, wish lack’; ‘desire
desideration] for’ to be’
wish √ [projection: √ ‘desire, hope √ ‘desire sb to have
proposal: for’ (success etc.)’
desideration]
 Table 7.13 Common polysemous adverbs and conjunctions

item textual interpersonal experiential interpersonal / experiential

as continuative as conjunctive Adjunct as comment Adjunct as mood Adjunct as circumstantial Adjunct as (Sub-)Modifier

actually √ [elaborating: clarification: √ [intensity:

verifactive] ‘in fact’ counterexpectancy:
You grew up in St. Louis, exceeding] ‘really’
Missouri, went to Vassar as But you might like this
an undergraduate, and then one ‘cos I remember
came back to Iowa for your you’ve eaten it in
graduate work. Actually, the past and you’ve
there was a year in there actually quite liked
where after I finished it.
Vassar I went to Europe
with my then husband

Analysis guide
and we hitchhiked around,
411

wondering what to do.

almost √ [intensity: degree: √ ‘not quite’, ‘very
high] nearly’
The story was so well The bill would make further
written paragraph demarcation of Indian
to paragraph [[that lands almost impossible.
it almost made me
angry]].
You are almost setting
yourself up as a
laughingstock

(Continued)
 Table 7.13 (Continued)

item textual interpersonal experiential interpersonal / experiential

as continuative as conjunctive Adjunct as comment Adjunct as mood Adjunct as circumstantial Adjunct as (Sub-)Modifier
at least √ [elaborating: clarification: √ [intensity] ‘if √ ‘not less than’, ‘at a
corrective] ‘anyway’ nothing else’ minimum’
At the 360-K surface, the She is at least trying ... Amnesty International said
calculated ozone loss is and always too busy. it calculated that U.S.
somewhat less than the prisons for adults also
observed loss. At least we hold at least 3,500
can say that above about child convicts in violation
the 400-K level, there does of an international
seem to be enough ClO to convention on civil rights.
explain the observed ozone
loss.

Analysis guide
doubtless √ [propositional: on
412

whole: asseverative:
sure] ‘presumably
or very probably’
Events will doubtless
open up windows
of opportunity for
stepwise advances
in support and
understanding
even √ [intensity: √ ‘all the more’, ‘still’
counterexpectancy: In fact, shares of the new
exceeding] ‘indeed’ AT&T Wireless Group
People convicted for also fell Tuesday even as
crimes committed as three major brokerages
children can even face initiated coverage of
the death penalty the stock with a “buy”
recommendation.
‘at the very same time as’
 eventually √ [enhancing: temporal: √ [temporal: future &
simple: conclusive] remote] ‘in the end’
‘finally’ Given all the trouble
Then they instructed the we had to make Pat’s
enemy forces to look the story into a film (The
other way as their marines Patricia Neal Story)
performed amphibious back-whenever- which
landings. Eventually, Van eventually had to be
Riper got so fed up with all made for television
this cheating that he refused instead because no
to play any more. one would touch it
as a theatrical film, I
would say there isn’t
much chance.
finally √ [enhancing: temporal: √ [temporal: future & √ [manner: quality]
simple: conclusive] remote] ‘in such a way as
[1] external ‘as the last in ‘after a long time, to put an end to
a series of related events typically when doubt and dispute’
or items’ there has been

Analysis guide
Finally my agent sent me a difficulty or delay’
413

letter You did finally take the

[2] internal ‘used to two Masters Degrees,
introduce a final point or the MA and the
reason’ MFA.
Finally, a point that should
not be overlooked is that
this notion of connections
is achieving more and more
significance for management
and policy.
happily √ [desirability: √ [manner: quality] ‘in
desirable] a happy way’
‘fortunately’, ‘it is Her married life glided
fortunate that’ happily on.
Happily, one effect
of all this quality
consciousness will be
an upgrading of the
average paper that
gets published.

(Continued)
 Table 7.13 (Continued)

item textual interpersonal experiential interpersonal / experiential

as continuative as conjunctive Adjunct as comment Adjunct as mood Adjunct as circumstantial Adjunct as (Sub-)Modifier
hopefully √ [desirability: √ [manner: quality] ‘in
desirable] ‘it is to a hopeful manner’
be hoped that’ Bloomfield hopefully
Hopefully, one also suggested that ‘the
recognizes passages methods of linguistics’
[[that are good]]. ‘resemble those of a
natural science, the
domain in which
science has been the
most successful’ (BL
509)

Analysis guide
indubitably √ [propositional: √ ‘unquestionably’
414

on whole: This is an indubitably

asseverative: sure] creative and original
‘unquestionably’ book.
Hartley Fudges, by
this measure, was
indubitably a
gentleman.
inevitably √ [propositional: on √
whole: asseverative: leading to certain sickness
natural] ‘as is and death, not to
certain to happen’ mention the inevitably
Inevitably, the book’s unpleasant appearance
terrain and feline and odor
star will bring to
mind Matthiessen’s
1978 National Book
Award winner, The
Snow Leopard.
 indeed √ [elaborating: clarification: √ [intensity:
verifactive] counterexpectancy:
And yet that new world, so exceeding]
violently announced in Hennig did, indeed,
the revolutionary decade create a “general
of 1789-1799, was reference scheme of
not quickly installed. biology.”
Indeed, the first half of
the nineteenth century
was an era of turmoil and
confusion.
in fact √ [elaborating: clarification: √ [intensity:
verifactive] counterexpectancy:
The disappointing forecast, exceeding]
which came as AT&T Gradually, they outgrow
posted first-quarter results their baby shoes —

Analysis guide
that met most expectations, if the expression is
415

dampened the enthusiasm pardoned, as Snufkin

created by last week’s is in fact the only
initial public offering of one of them who uses
$10.6 billion worth of footwear at all.
stock in the company’s
wireless business. In fact,
shares of the new AT&T
Wireless Group also fell
Tuesday even as three major
brokerages initiated coverage
of the stock with a “buy”
recommendation.

(Continued)
 Table 7.13 (Continued)

item textual interpersonal experiential interpersonal / experiential

as continuative as conjunctive Adjunct as comment Adjunct as mood Adjunct as circumstantial Adjunct as (Sub-)Modifier
in particular √ [elaborating: clarification: √ [intensity:
particularizing] counterexpectancy:
‘specifically’ limiting] ‘especially’
Green has long been involved The Navy and the Air
in studies related to Force in particular
biochemical phenomena are experiencing
that take place within retention gaps with
the body when muscles their first, second, and
contract and develop force. third term enlisted
In particular, he’s looking members.
at the role of high-energy
chemicals in skeletal
muscles, and in how cells

Analysis guide
adapt to protect the skeletal
muscles’ energy states.
416

just √ [temporal: non- √ ‘exactly’

future & near] ‘very Occasionally, in winter,
recently’ tropical cyclones
from which I have just [[traversing the South
returned China Sea]] pass to the
√ [intensity: south of Hong Kong ||
counterexpectancy: just when a monsoon is
limiting] ‘simply’, affecting the coastal areas
‘only’; ‘barely’ of south China.
She might have just
fallen over
naturally √ [propositional: on √ ‘in a normal manner’ √ ‘by nature’
whole: asseverative: We found || that they He was naturally shy.
natural] ‘as may be all seem to arise
expected’ naturally in an
Irwin Shaw was on that embodied way from
ship, in first class, forms of well-being
naturally, - health, wealth,
uprightness, light,
wholeness, cleanliness,
and so on.
 obviously √ [propositional: on
whole: asseverative:
obvious] ‘clearly’
of course √ [tonic] ‘you √ [elaboration: clarification: √ [propositional: on
should have verifactive] ‘actually, as whole: asseverative:
known that a matter of fact, in fact, obvious, natural]
already’ (Halliday indeed’ This of course makes
& Hasan, 1976: Of course any culture which the biggest difference
269); allows significant numbers in the poorest
“Everything’s just as of its membership to feel neighbourhoods.
it was!” — “Of unwanted — this society is
course it is”, said looking for trouble.
the Queen.
[non-tonic] ‘I
accept the fact;

Analysis guide
you must accept
417

the fact’ (ibid.:

269)
They were going
to come to the
meeting. Of
course they may
have changed their
minds.
only √ [extending: variation: √ [intensity: √ ‘no more than’
subtractive] ‘except that’, counterexpectancy: Only about four out of
‘but’ limiting] ‘solely’, every 10 residents
[structural:] ‘merely’ affected even know their
Dear Peter, James Fenimore I wrote only fiction for a new number
Cooper wrote this a hundred long time.
and fifty years ago, only he
wrote it better.

(Continued)
 Table 7.13 (Continued)

item textual interpersonal experiential interpersonal / experiential

as continuative as conjunctive Adjunct as comment Adjunct as mood Adjunct as circumstantial Adjunct as (Sub-)Modifier
particularly √ [elaborating: clarification: √ [intensity: √ ‘to a higher degree
particularizing] counterexpectancy: than is usual or
‘specifically’ limiting] average’
It is to be noted that glass, ‘specifically’ In November 1796 an Irish
plastic, or the like can They just mechanically convict, Francis Morgan,
be used to form a sensor applied Chinese who had been found
chip used of an SPR strategy, emulating guilty of a particularly
measurement. Particularly, particularly the brutal murder, was rowed
glass is suitably used. ideological activities out to the island
of the Chinese Red
Guards during the
Cultural Revolution.
plainly √ [propositional: on √ [manner: quality]
whole: asseverative: ‘in a style that is

Analysis guide
obvious] ‘used to not elaborate or
state one’s belief luxurious; simply’
418

that something No, I’m not hinting. I’m

is obviously or saying very plainly
undeniably true’ that the Yankees are
Plainly, ‘English is better equipped than
not a finite state we.
language’.
quite √ [intensity: degree: √ ‘absolutely’,
high] ‘really’ ‘completely’
and you’ve actually quite The language is quite
liked it. plain, and it was
hypnotic in a way.
really √ [intensity: degree: √ ‘very’
high] For me, as a writer, that
There wasn’t really was a really important
a literature yet in technical way of looking
written Ibo. at things.
simply √ [intensity: counter- √ [manner: quality] ‘in √ ‘absolutely’,
expectancy: limiting] a straightforward or ‘completely’
‘merely’, ‘just’ plain manner’ it’s simply incredible
It would simply become It can be stated simply
no history at all. enough
 soon √ [enhancing: temporal: √ [temporal: near & √ [time] ‘early’
complex: interrupted] future] ‘in or after a Others arrive there
Soon the ark began to take short time’ sooner.
shape. It will soon begin to rain.
still √ [enhancing: causal- √ [temporal: positive & √ ‘even’
conditional: since] ‘up until the And flying into Hanoi, as
specific: concessive] present or a specified opposed to Ho Chi Minh
‘nevertheless’, ‘all the or implied time; by City (i.e. Saigon), made the
same’ now or then’ sense of retro transgression
Still, I get itchy. and I was still there still more piquant
yet √ [extending: addition: √ [temporal: negative √ ‘even’
adversative] & by] ‘up until The second may be yet
√ [enhancing: causal- the present or a more subtle
conditional: specific: specified or implied
concessive] ‘in spite of time’, ‘by now or
that’, ‘nevertheless’ then’

Analysis guide
I’ve always thought that my I haven’t hit my stride
real writing was the fiction, yet.
419

which seems odd, since I’ve

done over twice as many
non-fiction books as fiction
books. Yet I really haven’t
changed my view.
wisely √ [propositional: on √ [manner: quality]
Subject: wisdom & ‘in a way that
positive] shows experience,
The British wisely knowledge, and
moved it to York good judgement’
We have to choose
wisely
wrongly √ [propositional: on √ [manner: quality] ‘in a
Subject: morality & way that is incorrect or
negative] mistaken’, ‘in an unjust,
Rightly or wrongly, or immoral way’
at least mercifully, he The questions are
slid into a coma wrongly put.
 Analysis guide

Certain sets of distinct uses are quite common, like the use of an adverb as the Head of an
adverbial group that can serve either as a circumstantial Adjunct of manner: quality or as a com-
ment Adjunct, frankly and hopefully being well-known examples. Some of these alternatives are
diagrammed and illustrated in Figure 7.10 earlier. As always, probes include testing for close
agnates. For example, comment Adjuncts may be expanded with speaking . . ., speaking frankly,

Table 7.14 Items serving as conjunctions according to type of expansion

item structural (str.)/ elaborating extending enhancing
cohesive
actually cohesive clarifying:
verificative
after str. hypotactic spatio-temporal: simple:
following (since)
after a while cohesive spatio-temporal: complex:
interrupted (soon)
afterwards cohesive spatio-temporal: simple:
following (then, next)
also cohesive additive: positive
(furthermore,
moreover)
alternatively cohesive varying: alternative
(or [else])
and str. para additive: positive
(also, moreover)
anyway cohesive clarifying:
dismissive (in any
case, leaving that
aside)
apart from that cohesive varying: subtractive
(except for that)
as str. hypotactic spatio-temporal: simple:
simultaneous (while);
causal-conditional: causal:
reason (because)
as to that cohesive matter [respective]:
positive (here, there, in
that respect)
at least cohesive elaborating:
clarifying:
corrective (or
rather, to be more
precise)
at once cohesive spatio-temporal: complex:
immediate (thereupon)
at that time cohesive spatio-temporal: complex:
durative (meanwhile)

420
 Analysis guide

item structural (str.)/ elaborating extending enhancing

cohesive
at this moment cohesive spatio-temporal: complex:
punctiliar
because str. hypotactic causal-conditional: causal:
general (since)
because of this cohesive causal-conditional: causal:
general (so, therefore;
consequently)
before str. hypotactic spatio-temporal: simple:
preceding
before then cohesive spatio-temporal: simple:
preceding (previously)
besides str. hypotactic additive: positive
[prep.] (as well as)
besides cohesive additive: positive
(furthermore,
moreover)
briefly cohesive clarifying:
summative (in
short, to sum up)
but 9 str. paratactic adversative concessive ‘nevertheless’
‘however’;
varying: replacive
‘instead’
by the way cohesive clarifying:
distractive
(incidentally)
consequently cohesive causal-conditional: causal:
general (so, therefore;
because of this)
elsewhere cohesive matter [respective]:
negative (in other
respects)
except for that cohesive varying: subtractive
(apart from that)
finally cohesive spatio-temporal: simple:
conclusive & internal
for example cohesive appositive:
exemplifying
(for instance, to
illustrate)
for instance cohesive appositive:
exemplifying
(for example, to
illustrate)

(Continued )

421
 Analysis guide

Table 7.14 (Continued)

item structural (str.)/ elaborating extending enhancing
cohesive
for that purpose cohesive causal-conditional:
causal: specific:
purpose
for that reason cohesive causal-conditional:
causal: specific:
reason
furthermore cohesive additive: positive
(also, moreover)
hence cohesive causal-conditional:
causal: general (so,
then, therefore)
here cohesive matter: positive (as to
that)
however cohesive adversative (but,
yet, on the other
hand)
in a different way cohesive manner: comparative
(similarly, likewise)
in any case cohesive clarifying:
dismissive
(anyway)
in case str. hypotactic causal-conditional:
conditional: positive
(if)
in that case cohesive causal-conditional:
conditional: positive (in
such an event)
in short cohesive clarifying:
summative
(briefly, to sum
up)
in the same cohesive manner: means
manner
incidentally cohesive clarifying:
distractive (by the
way)
instead cohesive varying: replacive
(on the other
hand)
lastly cohesive spatio-temporal: simple:
conclusive & internal
(finally)
likewise cohesive manner: comparative
(similarly, in a different way)

422
 Analysis guide

item structural (str.)/ elaborating extending enhancing

cohesive
meanwhile cohesive spatio-temporal:
complex: durative
(at that time)
moreover cohesive additive: positive
(also, furthermore)
next cohesive spatio-temporal: simple:
following & external
(then)
next day cohesive spatio-temporal:
complex: specific &
external (that
morning)
next time cohesive spatio-temporal:
complex: repetitive &
external
nevertheless cohesive causal-conditional:
conditional: concessive
(yet, still)
nor str. paratactic additive: negative
or str. paratactic appositive: varying: alternative
expository (in (alternatively, or
other words, that is) else)
otherwise cohesive causal-conditional:
conditional: negative
(if not)
previously cohesive spatio-temporal:
simple: preceding
(before then)
similarly cohesive manner: comparative
(likewise)
since str. hypotactic spatio-temporal:
simple: following
(after);
causal-conditional: causal:
reason (because)
so str. paratactic; causal-conditional:
cohesive causal: general (then,
therefore)
still cohesive causal-conditional:
conditional: concessive
(yet, though)
subsequently cohesive spatio-temporal: simple:
following (next)

(Continued)

423
 Analysis guide

Table 7.14 (Continued)

item structural (str.)/ elaborating extending enhancing
cohesive
then cohesive spatio-temporal: simple:
following (next);
causal-conditional:
[causal: general (so,
therefore)/
conditional: positive (in
that case)]
there cohesive matter: positive (here, in
that respect)
therefore cohesive causal-conditional:
causal: general (so,
then, hence)
thereupon cohesive spatio-temporal: complex:
immediate (at once)
though cohesive causal-conditional:
conditional: concessive
(still, yet)
until then cohesive spatio-temporal: complex:
terminal
while str. hypotactic adversative [1] concessive
[2] temporal
yet cohesive adversative causal-conditional:
(however) conditional: concessive
(nevertheless, still,
though)

or may be agnate with ‘mental’ clauses, as in I hope, whereas circumstantial Adjuncts may have
alternative realizations by prepositional phrases, e.g. with hope.
The uses that a particular item has evolved over time may cover two or all three metafunc-
tions, usually with an experiential point of origin, as documented for in fact and indeed by
Traugott (1997).
Conjunctions, functioning as the Head of conjunction groups, are listed separately in
Table 7.14, with the full “systemic gloss” derived from the system of conjunction in Halliday
and Matthiessen (2014: Chapter 9).

7.3.4 Prepositions
Prepositions are (close to being) grammatical items on the cline between grammatical and lexical
items, and they have usually evolved different senses, often with spatial ones being the source.
Some prepositions realize only one type of circumstance. For instance, instead of, as well as and
in addition to only serve to mark Accompaniment. However, a number of common prepositions
have several different circumstantial uses, and a few may also mark participants: see Table 7.15.
Prepositions function either (i) in prepositional phrases or (ii) in other environments:

424
 Table 7.15 Prepositions according to type of expansion and projection

preposition expansion projection participant

enhancement extension elaboration

Location Extent Cause Manner Accompaniment Role Matter Angle

about (a)round on, of,
she walked concerning
about the old they talked about
neighbourhood old times
according in the manner of in the opinion of
to they built it according to the paper,
according to they’ve reached an
plan agreement

Analysis guide
against towards [purpose:
425

she leant against negative]

the wall
as note: as a like for, in the
result of role of, by
way of
beside at/by the (older use)
side of besides
she sat down who would you
beside him like to invite
beside Mary
by before; via; at with Agent
for [intended [duration, [behalf, as Client
destination] distance] purpose] I’ll build a
he left for the they walked she called up the garage for you
coast for miles agency for him

(Continued)
 Table 7.15 (Continued)

preposition expansion projection participant

enhancement extension elaboration

Location Extent Cause Manner Accompaniment Role Matter Angle

from [source] [reason] (because)

he flew from of
Miru Guinness died from
liver cancer
in [place, time] [quality] Verbiage
he worked in the note: in + way/ they said in
attic manner French ...
he left in a hurry;

Analysis guide
he spoke in
426

a slow voice;
he spoke in a
deliberate way
into [place] [resultative]
turn right into in material
Pitt Street clause
cut into
cubes
of 10 (because) of on, about,
he died of cancer concerning
they talked of old
times
on [place, time] [quality] walk on about, of,
he sat on the all fours concerning
roof she meditated on
nothingness
 out of [place] [reason]
he walked out of because of
the house he did it out of pity
to [place, [resultative] Recipient,
time] she scared Receiver,
he sent it to him to Beneficiary
Paris death (rel.)
he gave it to her;
he said it to
her; he owed
it to her
with [place] [reason] [means] [comitative] (Agent) by
for, because of he proofed it with she danced with she was pleased
he shook with fear a spell-checker Fred with/by the

Analysis guide
result
427

without [means] [comitative]

he proofed it she danced
without a without
spell-checker Fred
 Analysis guide

(i) in prepositional phrases as Minor-process, e.g. in in the prepositional phrase in the city, function-
ing as a circumstance of Place in Did you grow up in the city? and in the prepositional phrase on the
deck, functioning as a Qualifier in the nominal group the hole on the deck in They could have fallen
through the hole on the deck. Cf. Section 7.2.2.2 above on “prepositional phrase attachment”.
(ii) in other environments
(1) as conjunctive preposition (e.g. with, on, after, before, because of), serving as binders in
non-finite ‘bound’ clauses, e.g. with the PSCs persisting through October, including non-
finite clause with the Process implicit (the explicit agnate version would be being), as in
Politically, the state was reforming, with the rudiments of modern bureaucracy in place and with
the tendency toward centralization of authority already evident in a country as important as France.
(2) as the “phrasal” part of phrasal verbs (Halliday and Matthiessen, 2014: Section 6.3.6)
consisting of verb + preposition, e.g. look after, look into, as in if I couldn’t look after myself
and he could be looked after.

In clauses, prepositional phrases usually function as circumstances, but a few prepositions can
serve to mark participants under certain textual conditions: Range: on, at, over, up; Range/Ver-
biage: in; and Attribute in ‘circumstantial’ ‘relational’ clauses.
A number of items that are used as prepositions are also used as adverbs, e.g. in, on, up.
The prepositional and adverbial senses are differentiated in good dictionaries. Jespersen (e.g.
1933) grouped prepositions and adverbs together as “particles”, and some linguists have analysed
adverbs that can also be used as prepositions as prepositions in intransitive prepositional phrases.

Table 7.16 Grammatical items serving as continuatives

Continuative Non-continuative

Item Tonic Non-tonic

anyway dismissive, “no matter “cohesion with the preceding conjunction: extending:
under which or sentence by simply brushing it adversative.
what circumstances” aside”; “let’s get on with the
(Halliday & Hasan job” (H & H, 1976: 270)
[H & H], 1976: 270)
no phonologically weak as mood Adjunct: as statements,
either in answer to a question,
in acknowledgement to a
statement, in undertaking of a
command or in acceptance of an
offer — tonic (so also salient),
or salient (but not tonic).
Or as minor clause.
now non-continuative —> indicates a shift in the text tonic: deictic time reference,
(“opening up a new stage in experiential time interpersonally
the communication”, H & H, anchored. Serves as Location.
1976: 268)
of course “you should have “I accept the fact; you must interpersonal adverb: presumptive
known that already” accept the fact” (H & H, (clearly, no doubt, obviously). Serves
(H & H, 1976: 269) 1976: 269) as mood Adjunct.

428
 Analysis guide

Continuative Non-continuative

Item Tonic Non-tonic

surely non-continuative --> “am I right in my understanding interpersonal adverb: modal.

of what’s just been said; you “invites the hearer to assent to the
can’t have meant ...?” (H & proposition being enunciated”
H, 1976: 270) (H & H, 1976: 270)
well “I acknowledge the “What I’m about to say is a experiential adverb (adverbial form
question, and will response to your move”; if of good); experiential adjective
give a considered the previous move was by the (not ill)
answer” (H & H, same speaker, it introduces an
1976: 269) explanatory comment. (H &
H, 1976: 269)
yes phonologically weak as mood Adjunct: as statements,
either in answer to a question,
in acknowledgement to a
statement, in undertaking of a
command or in acceptance of an
offer — tonic (so also salient),
or salient (but not tonic).

Or as minor clause.

7.3.5 Continuatives
All items that can be used as continuatives (Halliday and Hasan, 1976: Section 5.8; Martin,
1992a: 230–234; Halliday and Matthiessen, 2014: 107) also have non-continuative uses as
conjunctions or adverbs: see Table 7.16.
The grammatical items yes and no have evolved different uses and can be a challenge to anal-
yse: see Halliday and Matthiessen (2014: 175–176).

Notes
1 For general guidance in the interpretation of the patterns emerging from transitivity analysis, see
Thompson (2014).
2 For a provisional list of possible ambiguities encountered in lexicogrammatical analysis, see Matthies-
sen (1995a: 811–817).
3 The clause I understand seems to serve as an explicitly subjective interpersonal assessment given by the
Interviewer, indicating the status of the proposition you’re a veteran. (We can interpret it as an evidential
marker; cf. I hear.) This is reflected in the Moodtag; we would say you’re a veteran, I understand, aren’t you?
rather than you’re a veteran, I understand, don’t I? And it is indeed the proposition you’re a veteran that Peter
Matthiessen replies to: Yes [I am a veteran], and then expands on I served in World War II . . .
4 The clause I think serves as an explicitly subjective modal assessment of the proposition it was at a little
lying-in hospital.
5 It would be appropriate to amend the entry to draw attention to the fact that the v-en form of suggest,
i.e. suggested, often serves as an Epithet in a nominal group, as in A suggested course of study in American
government for senior high school social studies; Outline of Suggested Specifications for Purchasing Processed
Fruits and Vegetables.

429
 8
CONCLUSION

Since this book is designed as a reference book to be consulted as the need arises rather than as a
macro-report, a macro-narrative, a macro-explanation or a macro-exposition, there is in a sense
no need for a conclusion (or indeed a place) – final wrap-up as a kind of macro-New. But it may
be helpful to take stock and look ahead.
Taking stock of what SFL offers and can offer, we can note that there are many regions
within the theoretical “map” of language in context and within the descriptive maps of particu-
lar languages in their particular contexts that need a great deal of further work, even if we are
only to achieve low delicacy accounts across all regions. (And here “language” can be taken as
short-hand for “denotative semiotic systems” [Martin, 1992a]; the engagement with such sys-
tems in SFL has become second nature, or even first nature. This engagement was foreshadowed
by Halliday’s, 1978, characterization of “social semiotic” and in fact also in his account of the
protolinguistic phase of how young children learn how to mean since the protolanguages of
early childhood are multimodal from the start, in terms of the expression plane resources infants
deploy – Halliday, 1975a.) While scholars in the humanities tend to leave the choice of research
fields, questions and objectives to their research students – unlike many of their colleagues
around campus engaged in sciences of material (physical and biological) realms of phenomena,
the time has probably come to provide research students with more guidance in the form of
agendas of (urgent) research projects. (As a start, it could be helpful and productive to maintain a
collective list of urgent research topics, one updated regularly – say annually – as a complement
to the new SFL theses added to the international treasure trove of systemic functional documen-
tation of research.)
We can begin by identifying gaps interpreted in terms of single semiotic dimensions; the
global ones are the hierarchy of stratification, the cline of instantiation and the spectrum of
metafunction (cf. Matthiessen, 2009c, 2010):

• Stratification: while a fair amount of work has been done to develop the descriptions of
the lexicogrammars of a growing number of languages (Table 9.5) up to a certain point
in delicacy (usually before they reach the lexical zone of lexicogrammar), the descriptions
of semantics and context are more partial and largely focussed on English. The work on
semantics as an interface to other (denotative) semiotic systems and biosemiotics systems
remains exploratory (e.g. Bateman et al., 2010, on the “ontology” of space; Matthiessen,

DOI: 10.4324/9781315675718-8 430

 Conclusion

forthcoming e). The descriptions of the expression plane systems are limited to the pro-
sodic systems of English and a few other languages and to “model” accounts of articula-
tory phonology for a few languages; and comprehensive systemic functional descriptions
of graphological (and graphetic) systems are also high on the agenda, and arguably within
reach since they overlap with “multimodal” accounts of page layout and similar phenomena
(the same also applies to the expression plane of sign languages of deaf communities). – In
terms of theory, inter-stratal relations need more attentions – much more attention, in fact;
while they have all been characterized as forms of realization, this relation still needs con-
siderably more work, also in explicit modelling, and the realizational relationship between
context and semantics has been problematized and given special attention. The work that
is needed includes explicit modelling of lexicogrammatical metaphor – that is, both gram-
matical metaphor, which has received a great deal of attention in SFL, and lexical metaphor
(referred to as “conceptual metaphor” in cognitive linguistics).
• Instantiation: the preceding comments on the stratal coverage of descriptions apply to
the potential pole of the cline of instantiation. Such systemic descriptions have been used
extensively in text analysis, and this analytic work has informed the description of reg-
isters (text types, or genres), with ones operating in the institution of education having
been covered most extensively. Registerial maps of other institutions are also being cre-
ated, e.g. of institutions of healthcare. But the project of registerial cartography is clearly a
long-term undertaking, and needs broadening to many additional languages and, building
on contributions since the 1990s, to semiotic systems other than language. – In terms of
theory, the notion that language in context is a probabilistic system is in place, and systemic
probabilities have been studied through the frequency of instantiation of terms (options) in
systems, but conditional probabilities – both paradigmatic and syntagmatic ones – need to
be theorized further so that they can be investigated in text. More generally, the theory of
processes of instantiation needs to be developed considerably, building on but transcending
computational modelling since the early 1980s. There are great opportunities for engage-
ment with research outside SFL involving (unsupervised) machine learning. This research
can be thought of as constituting a rediscovery of discovery procedures outside SFL (e.g.
Lee, 2018).1
• Metafunction: the full spectrum of metafunctional resources has been described for Eng-
lish and also for a growing number of other languages (this is obviously a biased unfor-
tunate way of putting it – cf. the term LOTE; on occasion, we have suggested jocularly,
this the term should be LOTS, “Languages Other Than Swedish”); however for languages
other than English, the descriptions tend to be confined to the lexicogrammatical stratum
of the content plane, with much less descriptive work having been done on semantics
(and on the expression plane). For example, while there have been studies comparing the
speech functional selections in different registers (pioneered by Teich, 1999b; Lavid, 2000;
Murcia-Bielsa, 2000) and typological comparisons of mood systems (Teruya et al., 2007;
Mwinlaaru, Matthiessen and Akerejola, 2018), there is as yet no comprehensive account
of the systems of speech function in a significant sample of typologically fairly different
languages. In the last two decades, interpersonal systems have been given a great deal of
attention – in particular appraisal, both in the description of English (Martin and White,
2005) and in the analysis of texts from many different registers. In terms of text analysis,
it could be argued that the tendency remarked on by Halliday (1988c) to analyse texts in
terms of one metafunction is still predominant, leaving room for more combinations –
transitivity and appraisal being a potent one. – In terms of theory, the metafunctional
spectrum has been clearly articulated: case studies have demonstrated the move from the

431
 Conclusion

microfunctions of protolanguage via the generalized macrofunctions of the transition into

the metafunctions of post-infancy adult language, the metafunctional modes of meaning
and the modes of expression have been clearly documented (and “tested” for a range of
languages), and the resonance or “hook-up” between the contextual parameters of field,
tenor and mode and the ideational, interpersonal and textual metafunctions has continued
to inform productive research and application. There is still a great deal of theoretical work
to be done across all the metafunction. For example, the modes of expression remain some-
what figurative in nature, and need further work on forms of representation.

Within each stratal subsystem, we can then examine the local semiotic dimensions, the hierarchy
of axis and the hierarchy of rank (the rank scale):

• Axis: given the priority given in SFL to the paradigmatic axis over the syntagmatic axis
as (cf. Halliday, 1966c) gaps that are local to stratification, instantiation and metafunction
are mostly helpfully characterized systemically (paradigmatically) in the first instance rather
than structurally (syntagmatically): we need to identify systems that have not been taken
account of, and then move on to the structural realizations of terms in these systems that
need to be documented. At the same time, it is also important to strengthen the specifica-
tion of structural specification to the point where they can support automatic analysis (cf.
Bateman, 2008b).
• Rank: given the ecological approach to the description of stratal subsystems in SFL, the
highest-ranking unit of each subsystem is usually taken as the starting point for the descrip-
tion of a given language (in contrast with traditional grammar, where the word was the
starting point) and the higher-ranking units are thus given more descriptive attention.
Therefore, descriptions of different languages often need to be expanded downwards in
terms of rank, extending the descriptions of units of lower ranks in the higher-ranking
environments that they operate in (cf. Matthiessen, 2015d on morphology, and 2021a on
phonology).

These observations are stated at the stratum of theory in the stratified organization of the systemic
functional metalanguage. As we move down the stratal organization of the metalanguage from
theory to theoretical representation and then to computational representation (in the research
context of computational SFL), there will be gaps to fill – domains within the theory that have
not yet been represented by means of theoretical representation or modelled in terms of
computational representation. Filling such gaps is important for many reasons, e.g. in order to
make advances in automated text analysis.
Taking a step further back to view semiotic systems in the ordered typology of systems, we
can identify the frontier of research relating semiotic systems to social systems, to biological
systems and to physical systems. The grand project of investigating the relationship between
semiotic systems and social ones has, of course, been part of the Malinowski-Firth-Halliday
flow of ideas and of research projects all along; and systemic functional linguists continue to
contribute to what we might call Hasan’s (e.g. 2009b) long-term programme concerned with
the development of a kind of sociolinguistics informed by SFL. Here we need studies engag-
ing with languages other than English, and more generally with languages other than “world
languages”, operating in a much wider range of social systems. The next frontier has received
much less attention over the decades – the frontier between semiotic (and social) systems and
biological ones; but this has begun to change rather dramatically, and the notion of neurosemiot-
ics promoted by Adolfo García and his colleagues provides, e.g. a foundation for future research

432
 Conclusion

empowered by the systemic functional understanding of language and other semiotic systems
in addition to insights from Relational Network Theory (as a new phase of Stratificational
Linguistics). García, Franco-O’Byrne and Ibáñez (2020: 93) write: “neurosemiotics aims to
study communication from a perspective that is both social and biological, going far beyond the
abstract conceptualization of signs”.
As we view semiotic systems in the environment of systems of lower orders in the ordered
typology of systems operating in different phenomenal realms, we can also look ahead towards
more investigations of systems in general – investigations of the kind that have been undertaken
in variations on the theme of general systems science, including the study of complex adap-
tive systems associated with the Santa Fe Institute (www.santafe.edu) and network science (e.g.
Barabási, 2016). Such efforts resonate with the holistic nature of systemic functional theory,
and the “systems thinking” embodied in it, e.g. in the exploration of syndromes and of fractal
patterns, and also with the positioning of SFL as inherently permeable and oriented towards
dialogues across traditions and disciplines.
As an appliable kind of linguistics, SFL will continue not only to engage with problems in
well-established areas such as educational linguistics but also to contribute to more recent efforts
such as ecolinguistics and to take on new ones e.g. related to the continued expansion of the
volume of texts and the challenge of assessing bias and veracity.
Across all these areas of activity, it is likely that SFL will develop along lines similar to those
that characterized the development of more established sciences of lower-order systems –
adding more technological support and tending more towards research carried out by teams in
labs, now in an increasingly “virtual” space.

Note
1 Here Halliday’s (1961) comments on Chomsky’s (1957) criticism of “discovery procedures’ is worth
recalling: “Most, if not all, of the points made in this section can be brought together under Chom-
sky’s observation that “a linguistic theory should not be identified with a manual of useful procedures,
nor should it be expected to provide mechanical procedures for the discovery of grammar”. The
point is a familiar one to British linguists, who have for some time stressed the theoretical, as opposed
to procedural, character of their own approach. But is it true that “it is unreasonable to demand of
linguistic theory that it provides anything more than a practical evaluation procedure for grammars”?
This it must do. But it can be asked to do more: to provide a framework of logically interrelated
categories (so that it can be evaluated as a theory, and compared with other theories) from which can
be derived methods of description, whether textual, exemplificatory or transformative – generative,
which show us something of how language works.”

433
 9
APPENDIX

In the Appendix, we present information about online sources, conventions in the representa-
tion of system networks and realization statements and annotations of texts, tabular overviews
of accounts of SFL and descriptions of English and other languages, and modes of displaying
text analysis.

9.1 Online sources

There are a growing number and range of online SFL sources. An overview is presented by
O’Donnell (2009), soon to appear in a revised version.

9.1.1 Websites
SFL websites include ones maintained by the international association, ISFLA, and also by
regional associations (www.isfla.org/Systemics/Associations/index.html). The ISFL website is a
good starting point since it has links to regional associations and academic events:

International Systemic Functional Association (ISFLA): www.isfla.org

Australian Systemic Functional Association (ASFLA): https://asfla.net
Japan Association of Systemic Functional Linguisitics (JASFL): www.jasfl.jp
Chinese Association of Functional Linguistics: www.isfla.org/Systemics/Associations/CAFL.
html
Latin American Systemic Functional Linguistics Association (ALSFAL): www.facebook.
com/alsfaloficial/
North American Systemic Functional Linguistics Association (NASFLA): www.facebook.
com/groups/1728860180716525/group/
Systemic Functional Linguistics Association of Tunisia (SYSFLAT): www.syflat.tn
European Systemic Functional Association (ESFLA): www.esfla.org
Nordisk forening for SFL og socialsemiotik: www.sdu.dk/da/om_sdu/institutter_centre/
isk/forskning/forskningsprojekter/nordisk_sfl.aspx

DOI: 10.4324/9781315675718-9 434

 Appendix

Peter White’s web site, including information about appraisal

www.prrwhite.info
www.prrwhite.info/prrwhite%2C%202015%2C%20Appraisal%20theory%2C%20
Wiley%20Encylopedia.pdf

Alvin Leung provides resources for studying SFG (the major systems of the clause, theme, mood
and transitivity, and clause complexing) and appraisal.

www.alvinleong.info/sfg/

Systemic functional books, like academic books in general, now often have web companion
sites.
Many systemic functional linguistics and also projects have accounts with ResearchGate and
Academia, and may make publications and other materials available for download.

9.1.2 Presentations, YouTube channels

There are now quite a few presentations informed by SFL available on the web, for example as
part of (online) conferences, and there are also a number of contributions by individual scholars,
including:
Christian M.I.M. Matthiessen maintains a YouTube Channel with presentations informed by
SFL, including lectures from courses, talks, and short presentations on particular topics:

www.youtube.com/channel/UCOFu2gXnwpH1MA2ZroXJrTQ

Martin J. Tilney provides a “Stylistics Masterclass: Using [systemic functional] linguistics to

interpret literature”:

www.academia.edu/courses/9rW8x6?tab=0&trigger=personalized-courses-ping&v=
6D9G1E

9.1.3 Tools
Mick O’Donnell has created and made freely available a suite of tools supporting systemic func-
tional linguists, his UAM Corpus Tool, “text annotation for the 21st century”:

www.corpustool.com

In addition, he has also created an image analysis tool: www.wagsoft.com/ImageTool/

There is a wide range of (free) tools for linguists in general, in particular designed for (mul-
timodal) text (discourse) analysis; they are not specifically systemic functional, but can be very
useful:

• Praat: www.fon.hum.uva.nl/praat/
• Books Ngram Viewer: https://books.google.com/ngrams
• Google Trends: https://trends.google.com/trends/?geo=ES

435
 Appendix

• Voyant Tools: https://voyant-tools.org

• Sentiment analysis tools, e.g. https://blog.hubspot.com/service/sentiment-analysis-tools;
AFINN: https://darenr.github.io/afinn/
• Linguistic Inquiry and Word Count (LIWC, “Luke”), a digital humanities workbench:
https://www2.fgw.vu.nl/werkbanken/dighum/tools/tool_list/liwc.php
• Sketch engine, “tool to explore how language works”: www.sketchengine.eu
• Wmatrix, “a software tool for corpus analysis and comparison” (Lancaster University):
https://ucrel.lancs.ac.uk/wmatrix/
• ELAN is “an annotation tool for audio and video recordings”, with a very large user-base,
supporting varied textual annotations: https://archive.mpi.nl/tla/elan

In addition to tools, there are many online corpora that can provide data for systemic functional
analysis and description for a variety of languages.

9.2 Conventions
This section presents conventions in SFL that are commonly used in the graphic and algebraic
representations of systems networks, including realization statements associated with systemic
terms.

9.2.1 System networks

System networks represent the theory of axial organization in SFL where the paradigmatic
axis is treated as primary. Thus system networks capture the paradigmatic organization, fore-
grounding paradigmatic relationships, and syntagmatic specifications are always given in the
context of paradigmatic specifications, i.e. terms in systems. For example, the sequence of
Subject before Finite in a ‘declarative’ clause in English is specified as a realization statement
associated with the systemic term ‘declarative’, the realizational relationship being represented
by an arrow, ↘: ‘declarative’ ↘ Subject ^ Finite, which says “the systemic terms ‘declarative’
is realized by the sequence Subject before Finite”. The term ‘declarative’ contrasts with the
term ‘interrogative’ in the system of indicative type, whose entry condition is ‘indicative’, so
through ‘declarative’ it is possible to relate the term and its realization statement to all other
specifications in the system network, as illustrated by the fragment of the network shown in
Figure 9.1.
In terms of the stratal organization of the systemic functional metalanguage (see Section
1.3.1 in Chapter 1), system networks are located within the stratum of theoretical repre-
sentation, below the stratum of theory (the theory of paradigmatic organization) and above
the stratum of computational representation. In the literature on computational SFL, there
are contributions of modelling of the realization of system networks by means of computa-
tional forms of representation on the path to computational implementation (see references
in Section 1.3.1).
Systems in system networks may be represented either algebraically or graphically, as shown
in Figure 9.2:

(i) algebraically by means of the conventions set out in Table 9.2 and in Figure 9.1, e.g.:
‘clause: indicative/imperative’, which says that there is a system whose terms are repre-
sented by the features ‘indicative’ and ‘imperative’ and whose entry condition is ‘clause’;

436
 Appendix

Figure 9.1 System network with realization statements

(ii) graphically by means of graphs, as shown in Figure 9.3. Systemic terms may be fur-
ther specified by the addition of their probability of instantiation, e.g. ‘positive
0.9’/‘negative 0.1’.

The line leading from the entry condition to the terms that constitute a system should have
an arrowhead pointing to the system – out of respect for the inventor of system networks and
the graphic representation, M.A.K. Halliday, but also for theoretical reasons stated in Halliday
(2013b), “Meaning as choice”.
Note that the point of origin of a system network (i.e. the entry condition of its least deli-
cate system) should always be specified, otherwise the location (“semiotic address”) of the system
network within the total system of language (or of any other semiotic system) has not be given
and the network floats in the air, as it were.
Note that names of systems are useful labels, but they are not a formal part of a system (unlike
the entry condition and the terms), so they should not be represented graphically as part of the
“logic” of the system network but rather placed above (and below as well, if they are long) the
arrow pointing to systems. For example, in graphic representation of the system of modality,
there are two simultaneous systems with the term ‘modal’ as their entry condition (“input”).
They are thus represented by means of a curly bracket leading from ‘modal’, and names of the
system are specified above (and below, if necessary) the arrows leading from ‘modal’ to the two
systems of alternative terms: modality type and modality value. These two system names are
helpful in referring to the systems, but they are not part of the paradigmatic relations shown by
the system network.
Each parallel system can be represented by a dimension in a table or matrix and intersecting
features define cells in this table. For instance, if we just tabulate up to the first step in delicacy
in the system network in Figure 9.4, we get the paradigm set out in Table 9.1.
Creating tables such as Table 9.1 is a useful way of displaying the paradigm defined by a sys-
tem network by filling the cells with examples, and also of probing paradigms. For example, in

437
 Appendix

Figure 9.2 Graphic conventions and algebraic equivalents

438
 Appendix

Figure 9.3 Graphic representation of system

Figure 9.4 Conventions for placement of names of systems in system networks

439
 Appendix

Table 9.1 Paradigm of the intersection of the systems of modality type and modality value, with an
additional distinction between ‘subjective’/‘objective’ orientation

Modalization Modulation

Probability Usuality Obligation Readiness:

potentiality

median that’ll be the she’ll play tennis he should finish his –

postman/that’s on Tuesdays : homework/he is
probably the she usually plays expected to
postman tennis on Tuesdays finish his
homework
outer high that must be the she must play tennis he must finish his –
postman : that’s on Tuesdays/she homework/he
certainly the always plays tennis is required to
postman on Tuesdays finish his
homework
low that may be the she can play tennis he may finish his he can finish his
postman/that’s on Tuesdays/she homework/he homework/he is
perhaps the sometimes plays is allowed able to finish his
postman tennis on Tuesdays to finish his homework
homework

constructing Table 9.1 using constructed examples to fill the cells, we find that certain combi-
nations do not occur. Thus ‘potentiality’ only combines with ‘low’ modality value, not with
‘median’ and ‘high’ (unlike the other type of readiness, i.e. inclination). We may also get the
sense that certain combinations are more likely in the ‘objective’ orientation than in the ‘subjec-
tive’ one, specifically in the case of ‘usuality’, and this then suggests an extensive investigation of
natural examples in a sizeable corpus. However, while tables are heuristically useful, it is easy to
see that tables are much more limited than system networks in their representational power – it
is hard to show more than two or three dimensions and disjunctions (such as tagging above in
Figure 9.3) create problems.

Graphic conventions and algebraic equivalents

Conventions

Table 9.2 Conventions used in algebraic representation of system networks

‘indicative’ Systemic feature [i.e. term in a system] (all lower case)

mood type System name (all small caps)
/ Systemic disjunction
& Systemic conjunction
† *; *->, ->* Marking convention: ‘if, then’
° unmarked feature (contrasting with marked feature(s))
{} Selection expression

440
 Appendix

9.2.2 Realization
As noted earlier, systemic terms may be realized syntagmatically, and such realizations are speci-
fied by means of realization statements that are associated with the systemic terms they realize,
as illustrated by the fragment of the system of mood in Figure 9.4.
A realization statement consists of a realization operator, e.g. Preselect, Order, Conflate,
and one more realization operands, either features or functions. Each type of operator is
represented by a symbol, e.g. Preselect “:”, Order “^” and Conflate “/”. For example, Subject
^ Finite means that Subject precedes Finite. (While the names of realization operators all sug-
gest operations, they can all be interpreted declaratively as constraints.) The following types of
realization statement are used (see Table 9.3).

(i) structuring
(1) Presence of functions in the structure: the presence of a function in a function struc-
ture is specified by inserting the function into the structure; the operation of insertion
is symbolized by “+”; e.g. + Subject, + Mood, etc.
(2) Functional constituency relations: two functions may be related by constituency and to
specify this constituency relationship in the function structure one function is expanded
by the other; the expansion is symbolized by putting the expanding constituent func-
tion within parenthesis, e.g. Mood (Subject), which means that Mood is expanded to
have Subject as a constituent function. A function may be expanded by more than one
other function, e.g. Mood (Subject, Finite).
(3) Relative ordering of functions and ordering relative to unit boundaries: two functions
may be ordered relative to one another in the function structure and this relative order-
ing is symbolized by “^”; e.g. Subject ^ Finite, Mood ^ Residue. The ordering may
also be relative to the left or right boundary of a grammatical unit (represented by #),
e.g. # ^ Theme and Moodtag ^ #.
(ii) layering
(4) Conflation of one function with another: one function from one perspective is con-
flated with a function from another perspective, i.e. the two functions are specified as
different layers of the same constituent – they are identified with one another. Confla-
tion is symbolized by “/”; for example, Subject/Agent means that Subject (interper-
sonal) and Agent (ideational) apply to the same constituent.
(iii) inter-rank realization
(5) Realization of a function in terms of features from the rank below: the realization of a
function in a function structure is stated by preselecting one or more features from the
unit realizing it; preselection is symbolized by “:”, e.g. Subject: nominal group, Finite
and Predicator: verbal group, etc.
(6) Realization of a function by means of an item from a rank below: the realization of a
function in a function structure is stated by preselecting an item, grammatical or lexical,
from the unit realizing it; this preselection by item is symbolized by “::”, e.g. Finite:: do.

As noted earlier, since some of these operators suggest processes, like “insertion” and “confla-
tion”, it is important to note that they can all be interpreted as “declarative constraints”. For
example, “insertion” simply indicates the constraint that a function is present in the structure of
a unit. They are thus realization statements, not realization rules.

441
 Appendix

Table 9.3 Tabular summary of types of realization statements

Major type Operator Operand 1 Operand 2 Example

(i) structuring insert + function – + Subject

expand ( ) function function Mood (Subject)
order ^ function function Subject ^ Finite
(ii) layering conflate / function function Subject/Agent
(iii) inter-rank realization preselect : function term(s) Subject: nominal group
preselection by item :: function item Finite:: do

9.2.3 Structural representation

In multivariate structures, the elements operating in the function structure of a unit are repre-
sented by labels with initial upper case, e.g. Theme, Rheme, Subject, Finite, Process, Medium;
Pretonic, Tonic, Onset, Rhyme. In univariate structures, the elements are represented either
by Arabic numerals (parataxis), e.g. 1, 2, 3, or lower case Greek letters (hypotaxis), e.g. α,
β, γ. Superscripts after Arabic numerals or small Greek letters indicate the type of logico-
semantic relation, e.g. 1 × 2 means 1 paratactically enhanced by 2. These conventions are set
out in Table 9.4. An arrow - indicates an interdependency relation in a univariate structure,
e.g. 1 - 2.

9.2.4 Semantic structure

The preceding conventions for representing structure apply to all strata. Although they have
been used most widely in representations of grammatical structure, they apply equally to pho-
nological structure. In the representation of semantic structure, systemic functional linguists
have designed and adapted conventions enabling them to capture and show syntagmatic patterns
extending over passages of texts and whole texts. These include:

semantic reticula: drawing on the work by Hartford stratificationalists, Martin (1992a)

uses reticular to represent syntagmatic patterns within different systems of his “discourse
semantics”, including conjunction and reference.
RST diagrams of the logico-semantic structure of texts: in the analysis of texts as logico-
semantic complexes, the diagrammatic conventions of RST (Rhetorical Structure The-
ory) have been used extensively in systemic functional studies. They are presented e.g. in
Matthiessen (1995a), Bateman (2008a), Halliday and Matthiessen (2014) and Matthiessen
(forthcoming c).

In the representation of experiential semantics, systemic functional linguists have used versions
of frame-based inheritance network knowledge representation systems, typically in the context
of explicit computational modelling of language: see e.g. Matthiessen (1987b), Bateman (1990),
Halliday and Matthiessen (2006).

9.3 Transcription and annotation

Texts, whether written or spoken, can be represented in different ways depending on how the
representations are to be used. It is always advisable to retain a “raw” copy of a text, as close to

442
 Appendix

Table 9.4 Structural representation

univariate taxis hypotaxis α and other small Greek letters. Elements of

structure hypotactic interdependency structure.
parataxis 1 and other Arabic numerals. Elements of
paratactic interdependency structure.
logico- projection location “
semantic idea ‘
type
expansion elaborating =
extending +
enhancing ×
- interdependency relation.
multivariate Actor and other terms with initial capitals.
structure Function labels in multivariate
structures. Note that this convention
applies to all function labels,
including Mood and Residue.

the original version as possible, alongside any annotations made for the purpose of analysis. If the
text is written, the “raw” version may include graphological information that can be useful in
graphological analysis, for example of the layout. If the text is spoken, the “raw” version can be
used in acoustic analysis using an application such as Praat. For the purpose of analysis, spoken
texts may be transcribed, and the approach to the transcription will depend on the nature of the
analysis that it should support; importantly, in SFL, this is likely to include prosodic analysis so
the transcription must support this mode of analysis.
The “raw” text can be annotated to specify the units of analysis by the insertion of bound-
ary markers. The boundary markers used in SFL in grammatical analysis and in phonological
analysis are summarized in Table 9.5. (Additional boundary markers can, of course, be used, e.g.
“#” for boundaries between words, “+” for boundaries between morphemes in words, and “.”
for boundaries between syllables in feet.) There is (as yet) no general set of boundary markers
for semantic analysis. See also Section 9.7, “Displays of text analysis”. These conventions were
developed by M.A.K. Halliday in the 1960s, and designed to be used with a typewriter. In cur-
rent approaches to the annotation, or mark-up, of text, analysts take care to separate the annota-
tion from the text being annotated, which may be achieved by using an analysis platform such
as ELAN. In SysFan, a text marked up with the boundary markers shown in Table 9.5 can be
chunked automatically into records for the analysis of different units: see Wu (2009).
The SFL literature includes presentations dealing with the transcription and annotation of
texts, and the following contributions can be consulted:

Channel: oral-auditory
• transcription of spoken language: Halliday and Plum (1985)
• prosodic annotation: Halliday (1967), Halliday and Greaves (2008)
• phonological transcription and annotation: Tench (2011)

Channel: graphic-visual
• layout: Bateman, Wildfeuer and Hiippala (2017)

443
 Appendix

Table 9.5 Grammatical and phonological boundary markers

Grammar ||| Clause complex, boundary markers

<<< >>> Included clause complex
|| Clause (not rankshifted), boundary markers
<< >> Included clause
[[ ]] Rankshifted (embedded) clause, boundary markers
[[[ ]]] Rankshifted (embedded) clause complex, boundary markers
| Phrase, group, boundary markers
<> Included phrase, group
[] Rankshifted group/phrase, boundary markers
Phonology // Tone group, boundary markers
/ Foot, boundary markers

9.4 Summary of SFL overviews

As we have noted in Chapter 3, there are now many overviews of SFL in the form of introduc-
tions, companions, handbooks, encyclopaedias, thematic volumes, thematic book series and
journals publishing SFL-informed articles. New contributions are added regularly, so the best
way of obtaining up-to-date information is to carry out web searches. Nevertheless, in Table
9.6, we provide an overview of resources that are available two decades into the 21st century.
We have organized the table according to different domains in the study of language within SFL,
based on the semiotic dimensions that define the “architecture” of language in context.

9.5 Summary of descriptions of the lexicogrammar of English

SFL has been used as a guide in the description of a fair range of languages, and the range has been
increasing steadily in the last three decades, as can be seen from the overview in Section 9.6. Never-
theless, English remains the language given the most attention in description and also the language
where most work has taken place drawing on existing descriptions. And most attention has been
devoted to the description of the lexicogrammar of English. Therefore we focus on an overview of
descriptions of the lexicogrammar of English here, providing it in the form of Table 9.6.

9.6 Key descriptions of different languages

Descriptions of languages framed or informed by SFL are steadily growing in number and
coverage, and are presented not only in English but also an increasing range of languages,
including Chinese, Japanese, Vietnamese, Spanish, Portuguese, German, Danish and Swed-
ish. For overviews, see e.g. Caffarel, Martin and Matthiessen (2004), Teruya and Matthiessen
(2015), Mwinlaaru and Xuan (2016), Kashyap (2019). Here we can only present an overview
of seminal and comprehensive descriptions in Table 9.7. We note that there are languages not
covered here but represented in the KPML multilingual grammar system (e.g. Bateman, 1997).

9.7 Displays of text analysis

In Section 9.3 earlier, we set out the conventions used in the transcription of text and anno-
tation of texts in preparation for text analysis. In Section 9.2, we detail the representation of

444
 Table 9.6 Descriptions of the lexicogrammar of English with reference to chapters and sections

Rank Metafunction System IFG 1–4 LexCart DFG Thompson Fontaine Halliday Other sources
IFG 1–3 (2013) and Greaves
(2008)

clause (and [overview] Halliday (1985a, Matthiessen Chapter 1 Thompson Chapters 1–2 Eggins (1994); Gerot and Wignell
below) 1994c); (1995a) (1996) (1994); Bloor and Bloor (1995);
Halliday and Berry (2019)
Matthiessen
(2004, 2014)
ideational/ grammatical Chapter 10 Chapter 9 Ravelli (1985); Halliday and Martin
interpersonal metaphor (1993); Halliday and Matthiessen
(1999/2006: Ch. 6); Simon-
Vandenbergen, Taverniers and Ravelli
(2003); for work in China, see Zhang
et al. (2005) (Chapter 2: Section 4);

Appendix
Yang (2015)
445

clause: logical taxis Chapter 7 Chapter 3 Chapter 6 Chapter 7 Section 7.3 Ellis (1987a); Martin (1988); Nesbitt
complex and Plum (1988); Matthiessen and
Thompson (1988); Matthiessen
(2002a); Butt and Webster (2017)
clause experiential transitivity Chapter 5 Chapter 4 Chapter 4 Chapter 5 Chapter 4 Davidse (1991, 1992a, 1992b, 1996a,
1996b, 1999, 2017); Halliday (1998a);
Hasan (1987); Matthiessen (1999,
2014c); Morley (2000); Steiner (1985)
interpersonal mood Chapter 4; Chapter 5 Chapter 3 Chapter 5 Halliday (1984b); Thibault (1995)
Chapter 10,
Section 10.4
key IFG 3–4: Ch. 4 Chapter 5 Chapter 3
modality Chapter 4; Phillips (1986); Torr (1998)
Chapter 10,
Section 10.3
polarity Chapter 4 Halliday and James (1993)
(Continued)
 Table 9.6 (Continued)

Rank Metafunction System IFG 1–4 LexCart DFG Thompson Fontaine Halliday Other sources
IFG 1–3 (2013) and Greaves
(2008)

modal Chapter 4: Thompson and Zhou (1999)

assessment Halliday and
Matthiessen
(2004: 125–
132; 2014:
183–193)
vocation Chapter 4, Poynton (1984)
pp. 159–160
textual theme Chapter 3 Chapter 6 Chapter 2 Chapter 6 Chapter 6 Bäcklund (1992); Collins (1991a, 1991b,
1992); Downing (1991); Fawcett
and Huang (1997); Fries (1981,

Appendix
1994, 1995, 2002); Francis (1989);
446

Ghadessy (1995); Ravelli (1995);

Gosden (1993, 1996); Berry (1995,
1996); Hasan and Fries (1995);
Huang (1996, 2002); Kies (1988);
Martin (1992b); Thomas and Hawes
(1997); Whitaker (1995); Ghadessy
(1995, 1999); Gómez-González
(2001)
cohesion Chapter 9 Chapter 8 Taboada (2019)
conjunction Chapter 9 Chapter 8 Oshima (2004: 89)
group: logical modulation: Chapter 8 Chapter 7
complex cause
(verbal)
 group: logical tense Chapter 6 Chapter 7 Chapter 5 Halliday (1980); Matthiessen (1983,
verbal 1984, 1996); Halliday and James
(1993)
experiential tense;
aspect
event type
interpersonal modality, (see above under
polarity “clause”)
textual voice Downing (1996)
group: [general] Chapter 3
nominal logical modification Fries (1986); Tucker (1997, 1998)
experiential thing type Tucker (1996a)
epithesis Fries (1986); Tucker (1997, 1998);
Halliday and Matthiessen (1999:

Appendix
189–205)
447

interpersonal attitude Poynton (1984, 1996)

clause, textual cohesion Hasan (1984a); Fine (1994); Parsons
group (1995, 1996); for work in China,
see Zhang et al. 2005 (Chapter 2:
Section 3)
information textual information IFG 3–4: Chapter 6 Halliday and Prakasam (1985: Ch. 7); Davies (1996);
unit Chapter 3 Greaves Geluykens (1990); Bloor and Bloor
(2008) (1992); Martinec (1995: 157–168);
 Table 9.7 Seminal and comprehensive descriptions of different languages in (proto) systemic functional terms

Family Language 1950s 1960s 1970s 1980s 1990s 2000s, 2010s, 2020s

Sino-Tibetan Sinitic Chinese Halliday A series of MA Halliday Halliday & McDonald

[Mandarin] (1956a, thesis supervised (1992d, (2004); Halliday
1959) by MAKH 1993d) (2006); Li (2007)
at Sydney
University
Chinese Tam (2004)
[Cantonese]
Tibetan Wang (2020)
Austroasiatic Vietic Vietnamese Hoang (1997); Thai (2004); Hoang
Thai (1998) (2012)
Tai-Kadai Tai Thai Patpong (2005)
(Kra-Dai)
Mongolic Mongolian Zhang (2020a,b)

Appendix
448

(isolate) Korean Kim (2007a); Park

(2013); Kim et al.
(2023)
Japanese Japanese Teruya (1998) Teruya (2004, 2007)
languages
(Japonic)
Austronesian Philippine Tagalog Martin (1988) Martin (1990, Martin (2004b)
1996c)
Malayo-Polynesian Indonesian Wirnani (1985),
Made (1988)
Dravidian South-Central Telugu Prakasam (1972) Prakasam (1982b) Prakasam Prakasam (2004)
(1992)
“Australian” Pama-Nyungan Western Desert Rose (2001)
Non-Pama- Gooniyandi McGregor
Nyungan (1990)
Papuan Trans-New Weri Boxwell (1990)
Guinea
 Afro-Asiatic Semitic Arabic (Modern Bardi (2008)
Standard)
Cushitic Beja Hudson (1973)
Niger-Congo Atlantic-Congo: Mbembe Barnwell
Benue-Congo (1969)
Atlantic-Congo: Nzema Mock
Kwa (1969)
Akan Matthiessen
(1987a,b)
Atlantic-Congo: Gur Dagaare Mwinlaaru (2017)
Atlantic-Congo: Oko Akerejola (2005)
Volta-Niger
Yoruba Bamgbos.e
(1966)
Indo-European Indo-Aryan Bajjika Kumar (2009); Kashyap
(forthcoming)
Germanic English Halliday Halliday [see Halliday (1985a) Martin

Appendix
[see Halliday, (1992a);
449

Halliday, 2005a]; Matthiessen

2005a] Halliday & (1995a)
Hasan (1976)
German Steiner & Teich (2004)
Danish Andersen et al. (2001)
Swedish Holmberg & Karlsson
(2006)
Romance French Huddleston Caffarel (2004, 2006a)
& Uren
(1969)
Spanish Lavid, Arús &
Zamorano-Mansilla
(2010); Quiroz
(2008, 2013)
Uralic Finnic Finnish Shore (1992,
1996)

This table is of course not exhaustive. We have included Ph.D. theses only where they provide comprehensive overviews of the stratal subsystem of a language (typically
the clause grammar) that have not subsequently been published in book form. For certain languages there is now a fairly extensive literature, including a good deal of
work presented in languages other than English (e.g. the literature on Chinese in Chinese).
 Appendix

systemic and structural specifications. Here we provide a brief overview of displays used in the
presentation of text analysis, supplementing the account of text analysis as an application of SFL
in Chapter 4, Section 4.4. As we noted in that section, the choice of type of display will depend
on what patterns we want to bring out, for example, to support the interpretation of analysis.
Halliday and Greaves (2008: 201–211) provide an analysis guide that while focussed on intona-
tion also includes grammatical and semantic information.

9.7.1 Chunking of texts: unit boundaries

For the annotation of texts chunked into phonological and grammatical units, see Section 9.3
earlier.

9.7.2 Phonology: prosodic analysis

In the domain of articulatory phonology, systemic functional linguists use the International
Phonetic Alphabet (IPA) unless there are language-specific conventions for transliteration. In the
domain of prosodic phonology, Halliday had to introduce new annotations in his description of
English intonation and rhythm in the early 1960s (e.g. Halliday, 1967). This scheme is still in
use, helpful examples being Halliday (1985b/2002d: 264) and Halliday and Greaves (2008): see
Figure 9.5.
The use of the system of annotation in prosodic text analysis can be illustrated by means of
the following passage of dialogue taken from Halliday and Greaves (2008). They represent it first
in ordinary orthography as might be represented as a passage of dramatic dialogue in a novel or
short story:

M: Sit up and finish your lunch! [child obeys] Do they sit – they make you sit up straight like that at kinder, do they?
C: No.
M: I remember sitting up like that when I was at school. The teacher’d say ‘Sit up straight!’, and everyone sat up
looking like that. [stiffening her body]
C: Why did they?
M: I don’t know; ’cause they thought it made them look straighter, I suppose.
C: Why did the teacher at your school say ‘Sit up straight!’?

They then show the version annotated for phonological units (with tone group [//] and foot
[/] boundaries), and analysed in terms of primary tones (but still with ordinary orthography):

M: // 1 – sit up and/finish your */ lunch //

// do they/sit – //1they/make you/sit up/straight like/that at*/kinder *// 2 do they //
C: *//1no//
M: // 1 I remember/sitting up like/that when/I was at */ school the
// 1 teacher’d say/sit up */ straight and // 2 everyone/sat up/looking like */ that //
C: //5why*/did they//
M: *// 2 I don’t */ know ‘cause they // 13 thought it/made them look/*/straighter I
sup*/pose //
C: // 5 why did the/teacher at/your school say/sit up */ straight //

Here the annotation is included in the text that has been analysed prosodically. As we have
noted elsewhere, while this makes the analysis accessible to readers, the annotation should be
represented in such a way that it can be separated from the text being analysed (cf. the practice

450
 Appendix

Category phonology grammar

A. Boundary markers // tone group
in text
/ foot (rhythm group)
B. Tone symbols 1 tone one: falling ‘polarity known’: statement, wh-
question, command
1+ wide tonic strong key
1. medium tonic neutral key
1- narrow tonic weak key
.1 steady pretonic neutral
…1 listing pretonic listing (enumerating)
-1 broken pretonic forceful
2 tone two: rising ‘polarity unknown’: yes/no question
2. rise (only) tonic neutral question
2 fall-rise query-point question
.2 high pretonic neutral (British)
-2 low pretonic ‘surely’ question; neutral (U.S.)
3 tone three: level (low ‘non-committal’: secondary, incomplete,
rising) tentative
.3 neutral pretonic neutral
-3 low pretonic casual
4 tone four: falling-rising ‘known • unknown’: reservation,
contrast, hypothesis
4. high tonic, steady pretonic neutral key
4 low tonic, broken pretonic strong key
5 tone five: rising-falling ‘unknown • known’: override,
assertion, exclamation
5. high tonic neutral key
5 low tonic (breathy) exclamatory key
13 tone one three: falling + statement plus secondary information
level
53 tone five three: rising- override plus secondary information
falling + level
C. Other symbols ^ silent beat

… pause (sufficient to disrupt

rhythm)
bold tonic accent (nucleus) realizing information focus (culmination
type of New)

Figure 9.5 Prosodic analysis of text (Halliday, 1985b/2002d: 264); see also Halliday and Greaves (2008:
210–211)

of stand-of mark-up or annotation), which can be achieved by means of analysis platforms such
as ELAN.
Alternatively, the prosodic analysis can be represented by means of box diagrams of the kind
used in grammatical analysis, as illustrated in Figure 9.6. The information shown in the box
diagram can be derived automatically from the prosodic annotation of the example. It consists

451
 Appendix

Figure 9.6 Box diagram display of the prosodic annotation in // 5 why did the/teacher at/your school say/
sit up */ straight //

of one tone group, Pretonic ^ Tonic, which has selected for tone 5 (primary tone), and con-
sists of five feet. The feet could be further analysed functionally into Ictus ^ Remiss. Informa-
tion about secondary tones can also be added (Halliday and Greaves, 2008: Chapter 7).

9.7.3 Lexicogrammar
Since comprehensive lexicogrammatical analysis covers more strands of analysis than prosodic
analysis, the standard form of display is the box diagram. However, if only a single strand of
analysis has been selected, it may be displayed as annotations of running text – with the proviso
that it should be stored as a separate “stand-off ” line. For example:

[α] I remember sitting up like that [× β] when I was at school.

[Theme] I [Rheme] remember sitting up like that.

Box diagrams are appropriate for the analysis of grammatical structure, as we will illustrate later
in this Appendix, but analysts need other forms of display for non-structural, cohesive patterns.

9.7.3.1 Cohesive analysis

Cohesive analysis was introduced by Halliday and Hasan (1976), and their Chapter 8 deals with
the analysis of cohesion including a coding scheme (pp. 333–339). They also offer seven sample
texts (pp. 340–355). The coding scheme was designed to be compact, for analysis either by hand
or typewriter. Similarly, the tabular displays of the coding are quite condensed, as illustrated by
their analysis of the excerpt from Alice in Figure 9.7. This form of presentation of cohesive
analysis is not very accessible, and it is hard to discern patterns (contrast Martin’s, 1992a, reticular
displays). Halliday (1985b/2002d: 280–281) has proposed a simpler (but less informative) version
of cohesive analysis. When not restricted to the printed page, analysts may use colour coding and
extended tables, e.g. using spreadsheets.

9.7.3.2 Univariate analysis

Univariate structure is engendered by logical systems; it is serial in nature, like chains formed by
links – the links being logico-semantic relations. It is characteristic of complexes of units of dif-
ferent ranks and classes as also of groups, like the modification structure of nominal and adverbial
groups and the auxiliary structure of verbal groups in English. In both complexes and groups it
can include internal nesting, as in the examples analysed in Figures 9.8 and 9.9: see e.g. Halliday

452
 Appendix

Figure 9.7 Example of cohesive analysis of a text in Halliday and Hasan (1976: 340)

and Matthiessen (2014: 450–451). When there is a change in logico-semantic types in a tactic
sequence, this always opens up internal nesting – an important point even advanced analysts may
not have mastered. Thus the following clause complex (from an interview with Chinua Achebe)

||| It is a metaphor for [[what we have all observed]], || that a very weak father will
generally produce a very strong son, || and a very strong father will generally produce
a weak son. |||

453
 Appendix

is a paratactic sequence. However, the structure is not simply 1 ^ 2 ^ 3 but rather, since there is a
switch in logico-semantic type, it is 1 ^ = 21 = 2 + 2, with internal nesting within the elaborat-
ing continuing clause nexus:

1 It is a metaphor for [[what we have all observed]],

= 21 that a very weak father will generally produce a very strong son,
= 2 + 2 and a very strong father will generally produce a weak son.

Univariate structure can be displayed by means of box diagrams, as in Figure 9.8. Internal nest-
ing is shown algebraically or by boxes within boxes (see also Halliday and Matthiessen, 2014:
449–450). Alternatively, to bring out internal nesting in the diagram, we can opt for a vertical
tabular display, as in Figure 9.9. Here the depth in nesting is represented by columns, each col-
umn representing an additional step in the depth of nesting. For alternative representations of
the logical analysis of clause complexes, including ones using the graphic conventions of RST
(Rhetorical Structure Theory), see also Halliday and Matthiessen (2014: 456–459).

9.7.3.3 Multivariate analysis

Multivariate structure is engendered by experiential, interpersonal and textual systems, and
when they are analysed structurally as strands of syntagmatic patterning within one unit – proto-
typically the clause, each metafunctional strand can be represented by one or more rows within
a box diagram, as illustrated in Figure 9.10. The first two rows below the example show the
textual thematic analysis, the next two rows show the interpersonal modal analysis, the next row
shows the experiential transitivity analysis, and the final row represents the syntagm of classes of
units realizing the elements of the function structure. Note that the “phases” of the Theme (tex-
tual, interpersonal, topical) are spelt with lower case since they are simply phases of the Theme

A magnitude-6 quake can cause severe damage if it is centered under a populated area.
α ×β
free clause bound clause

To restore the original it was necessary [[to heat the residue of the so that the freed phlogiston
combustible substance, combustion with something [[that burned might again combine with the
easily]] ]], ashes.
×β αα α×β
bound clause free clause bound clause

To restore the original it was necessary [[to heat the residue of the so that the freed phlogiston
combustible substance, combustion with something [[that burned might again combine with the
easily]] ]], ashes.
×β α
α ×β
bound clause free clause bound clause

Figure 9.8 Box diagram representation of the univariate structure of clause complexes

454
 Appendix

×β To restore the original combustible substance,

α α it was necessary [[to heat the residue of the combustion with something [[that burned easily]] ]],
× β so that the freed phlogiston might again combine with the ashes.

1 This was explained by the supposition [[[that the more combustible a substance was || the more
phlogiston it contained]]],
×2 ×β and since free phlogiston sought always to combine with some suitable substance,
α α it was only necessary [[[to mix the phlogisticating agents, such as charcoal, phosphorus, oils,
fats, etc., with the ashes of the original substance, || and heat the mixture]]],
× β the phlogiston << thus freed>> uniting at once with the ashes.
× γ << thus freed>>

Figure 9.9 Vertical table diagram representation of the univariate structure of clause complex with internal
nesting

Figure 9.10 Box diagram showing the multiple metafunctional strands of the multivariate structure of a
clause

element, not separate elements. (In both examples analysed, the Residue is discontinuous in the
modal [interpersonal] structure of the clause. Here the discontinuity is represented simply by
splitting “Residue” into “Resi-” and “-due”. Such discontinuous functional elements do occur,
in English typically due to either textual or interpersonal patterns.)
The elements of the strands within the function structures are spelt with an initial cap, but the
classes of the units realizing those elements are spelt with initial lower case. The initial cap should

455
 Table 9.8 An example of tabular analysis

¶ cc # cl # MOOD TYPE Mood PROCESS TYPE Process Agent Medium Range circumstances

Subject Finite Adjunct

¶ Language of mathematics
¶ 1 1 decl. & The language is relational: is The language of mathematics
temporal of identifying & is the system [[[used by
mathematics intensive & mathematicians || to
middle communicate mathematical
ideas among themselves]]].
2 1 decl. & This language consists relational: consists of This a substrate This language consists of a
temporal attributive & language … substrate of some natural
possessive & language (for example
middle English) [[[using technical

Appendix
terms and grammatical
456

conventions [[that are peculiar

to mathematical discourse (see
Mathematical jargon)]], ||
supplemented by a highly
specialized symbolic notation
for mathematical formulas]]].
¶ 3 1 decl. & discourse [[ can material & can discourse Manner: Like natural languages in general,
modal using …]] effective employ [[…]] like discourse [[using the language
natural of mathematics]] can employ a
language scale of registers.
in general
4 1 declarative research are relational: are research sources Research articles in academic
& articles … identifying & articles for … journals are sources for
temporal intensive & … detailed theoretical discussions
middle about ideas [[concerning
mathematics and its
implications for society]].
 Appendix

be used also for Mood and Residue; sometimes they are spelt with all caps, but this should be
avoided since all caps are used for names of systems (alternatively small caps) and the use of all
caps thus invites confusion. The functional elements Mood and Residue are sometimes referred
to as “blocks”; but they are actually not “blocks” at all but more like prosodies running through
the elements of the modal (interpersonal) structure of the clause. Such prosodies may be realized
by concord (agreement).

9.7.4 Tabular analysis displays

Box diagrams are good for displaying the structure of units and unit complexes one by one, for
example bringing out how the textual, interpersonal and experiential functional elements of the
clause are mapped onto one another (conflated, unified). However, when we want to scan for
patterns that emerge from one unit or unit complex to another as a text unfolds, it is helpful
to use a tabular display where each unit is represented in a row and columns are used to show
functional elements in the structure of units and also selections of systemic terms as in Table
9.8 – see also e.g. Appendix 6 of Matthiessen (1995a).

457
 REFERENCES

Abbi, Anvita. 2021. Voices from the lost horizon: Stories and songs of the Great Andamanese. Delhi: Niyogi
Books.
Abercrombie, David. 1967. Elements of general phonetics. Edinburgh: Edinburgh University Press.
Abercrombie, David. 1968. Paralanguage. British Journal of Disorders of Communication 3: 55–59.
Adams, Catherine and Dorothy V. Bishop. 1989. Conversational characteristics of children with semantic-
pragmatic disorder I: Exchange structure, turntaking, repairs and cohesion. British Journal of Disorders of
Communication 24: 211–239.
Adejare, Oluwole. 1992. Language and style in Soyinka: A systemic text-linguistic study of a literary idiolect.
Ibadan: Heinemann Educational.
Akerejola, Ernest. 2005. A systemic functional grammar of Òkó. Ph.D. thesis. Macquarie University, Sydney.
Albrow, Kenneth H. 1972. The English writing system: Notes towards a description. London: Longman (Reis-
sued London: Schools Council, 1981).
Allen, James F. and C. Raymond Perrault. 1980. Analyzing intention in utterances. Artificial Intelligence 15:
143–178.
Allwood, Jens. 1981. On the distinction between semantics and pragmatics: Studies presented to Manfred
Bierwisch. In Wolfgang Klein and Willem Levelt (eds.), Crossing boundaries in linguistics. Ordrecht, Bos-
ton and London: D. Reidel. 177–189.
Aloni, Maria and Paul Dekker (eds.). 2016. The Cambridge handbook of formal semantics. Cambridge: Cam-
bridge University Press.
Alsop, Sian and Hilary Nesi. 2009. Issues in the development of the British Academic Written English
(BAWE) corpus. Corpora 4 (1): 71–83.
Alves, Fabio (ed.). 2003. Triangulating translation: Perspectives in process oriented research. Amsterdam: John
Benjamins.
Alves, Fabio, Adriana Pagano, Stella Neumann, Erich Steiner and Silvia Hansen-Schirra. 2010. Translation
units and grammatical shifts: Towards an integration of product- and process-based translation research.
In Gregory M. Shreve and Erik Angelone (eds.), Translation and cognition. Amsterdam: John Benjamins.
109–142.
Amano, Masa-chiyo (ed.). 2006. Multimodality: Towards the most efficient communications by humans. Nagoya:
Nagoya University.
Amundrud, Thomas M. 2017. Analyzing classroom teacher-student consultations: A systemic-multimodal perspec-
tive. Ph.D. thesis. Macquarie University, Sydney.
Andersen, Thomas Hestbaek. 2017. Interpersonal meaning and the clause. In Tom Bartlett and Gerard
O’Grady (eds.), Chap. 8. 115–130.
Andersen, Thomas Hestbaek, Morten Boeriis, Eva Maagerø and Elise Seip Tønnessen (eds.). 2015. Social
semiotics: Key figures, new directions. London: Routledge.
Andersen, Thomas Lund, Uwe Helm Petersen and Flemming Smedegaard. 2001. Sproget som ressource:
Dansk systemisk funktionel lingvistik i teori og praksis. Odense: Odense Universitetsforlag.

458
 References

Andreetta, Sara, Anna Cantagallo and Andrea Marini. 2012. Narrative discourse in anomic aphasia. Neu-
ropsychologia 50: 1787–1793.
Argyle, Michael, Adrian Furnham and Jean Ann Graham. 1981. Social situations. Cambridge: Cambridge
University Press.
Armstrong, Elizabeth. 1987. Cohesive harmony and its significance in listener perception of coherence. In
Robert H. Brookshire (ed.), Clinical aphasiology. Minneapolis: BRK Publishers. 210–215.
Armstrong, Elizabeth. 1992. Clause complex relations in aphasic discourse: A longitudinal case study. Jour-
nal of Neurolinguistics 7(4): 261–275.
Armstrong, Elizabeth. 2000. Aphasic discourse analysis: The story so far. Aphasiology 14(9): 875–892.
Armstrong, Elizabeth. 2001. Connecting lexical patterns of verb usage with discourse meanings in aphasia.
Aphasiology 15(10–11): 1029–1045.
Armstrong, Elizabeth. 2005a. Language disorder: A functional linguistic perspective. Clinical Linguistics and
Phonetics 19(3): 137–153.
Armstrong, Elizabeth. 2005b. Expressing opinions and feelings in aphasia: Linguistic options. Aphasiology
19(3–5): 285–295.
Armstrong, Elizabeth. 2009. Clinical applications. In M.A.K. Halliday and Jonathan J. Wester (eds.),
143–153.
Armstrong, Elizabeth. 2016. Systemic functional linguistics. In Jack S. Damico and Martin S. Ball (eds.),
The Sage encyclopaedia of communication sciences and disorders. Thousand Oaks, CA: Sage.
Armstrong, Elizabeth, Natalie Ciccone, Erin Godecke and Betty Kok. 2011. Monologues and dialogues in
aphasia: Some initial comparisons. Aphasiology 25(11): 1347–1371.
Armstrong, Elizabeth, Alison Ferguson, Lynne Mortensen and Leanne Togher. 2005. Acquired language
disorders: Some functional insights. In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan J.
Wester (eds.), 383–412.
Armstrong, Elizabeth, Sarah Fox and Ray Wilkinson. 2013. Mild aphasia: Is this the place for an argument?
American Journal of Speech-Language Pathology 22(2): 268–278.
Armstrong, Elizabeth, Lynne Mortensen, Natalie Ciccone and Erin Godecke. 2012. Expressing opinions
and feelings in a conversational setting. Seminars in Speech and Language 33(1): 16–26.
Armstrong, Elizabeth and Hanna Ulatowska. 2007. Making stories: Evaluative language and the aphasia
experience. Aphasiology 21(6–8): 763–774.
Arús, Jorge and Julia Lavid. 2001. The grammar of relational processes in English and Spanish: Implications
for machine-aided translation and multilingual generation. Estudios Ingleses de la Universidad Complutense
9: 61–79.
Arús Hita, Jorge. 2017. Theme in Spanish. In Tom Bartlett and Gerard O’Grady (eds.), Chap. 13. 194–212.
Arús Hita, Jorge, Kazuhiro Teruya, Mohamed Ali Bardi, Abhishek Kumar Kashyap and Isaac N. Mwin-
laaru. 2018. Quoting and reporting across languages: A system-based and text-based typology. Word
64(2): 69–102.
Asp, Elissa. 2013. The twin paradoxes of unconscious choice and unintentional agents: What neurosciences
say about choice and agency in action. In Lise Fontaine, Tom Bartlett and Gerard O’Grady (eds.), Chap.
8. 161–178.
Asp, Elissa. 2017. What is a system? What is a function? A study in contrasts and convergences. In Tom
Bartlett and Gerard O’Grady (eds.), Chap. 3. 27–43.
Asp, Elissa and Jessica de Villiers. 2010. When language breaks down: Analysing discourse in clinical contexts.
Cambridge: Cambridge University Press.
Asp, Elissa and Jessica de Villiers. 2019. Clinical linguistics. In Geoff Thompson, Wendy L. Bowcher, Lise
Fontaine and Jennifer Yameng Liang (eds), Chap. 23. 587–619.
Atkinson, Paul. 1985. Language, structure and reproduction: An introduction to the sociology of Basil Bernstein.
London: Methuen.
Bäcklund, Ingegerd. 1992. Theme in English telephone conversation. Language Sciences 14(4): 545–565.
Baker, Mona. 1992. In other words: A coursebook in translation. London and New York: Routledge.
Baker, Mona. 1995. Corpora in translation studies: An overview and some suggestions for future research.
Target 7(2): 223–243.
Baker, Mona. 2018. In other words: A coursebook in translation. 3rd edition. London and New York: Routledge.
Baker, Paul, Costas Gabrielatos, Majid Khosravi Nik, Michal Krzyzanowski, Tony McEnery and Ruth
Wodak. 2008. A useful methodological synergy? Combining critical discourse analysis and corpus
linguistics to examine discourses of refugees and asylum seekers in the UK press. Discourse and Society
19(3): 273–306.

459
 References

Bakhtin, Mikhail Mikaǐlvich. 1982. The dialogic imagination: Four essays. Austin, TX: University of Texas
Press.
Bakhtin, Mikhail Mikaǐlvich. 1986. The problem of speech genres. In Caryl Emerson and Michael Holquist
(eds.), Speech genres and other late essays, translated by Vern W. McGee. Austin: University of Texas Press.
Baklouti, Akila and Lise Fontaine (eds.). 2018. Perspectives from systemic functional linguistics. London:
Routledge.
Baldry, Anthony and Christopher Taylor. 2002. Multimodal corpus authoring system: Multimodal corpora,
subtitling and phasal analysis. In Mark Maybury and Jean-Claude Martin (eds.), Multimodal resources and
multimodal systems evaluation, LREC congress, post-conference workshop papers. Las Palmas: LREC. 45–51.
Baldry, Anthony and Paul J. Thibault. 2006. Multimodal transcription and text analysis: A multimedia toolkit and
coursebook. London and Oakville: Equinox.
Balint, Enid. 1969. The possibilities of patient-centred medicine. Journal of the Royal College of General
Practice 17: 269–276.
Baltaxe, Christiane A.M. and Nora D’Angiola. 1992. Cohesion in the discourse interaction of autistic,
specifically language-impaired and normal children. Journal of Autism and Developmental Disorders 22:
1–21.
Bamgboṣe, Ayọ. 1966. A grammar of Yoruba (West African Language Monograph Series 5). Cambridge:
Cambridge University Press.
Banks, David. 1991. Some observations concerning transitivity and modality in scientific writing. Language
Sciences 13(1): 59–78.
Banks, David (ed.). 2004. Text and texture: Systemic functional viewpoints on the nature and structure of text. Paris:
L’Harmattan.
Banks, David. 2005. Emerging scientific discourse in the late seventeenth century: A comparison of New-
ton’s Opticks, and Huygens’ Traité de la lumière. Functions of Language 12(1): 65–86.
Banks, David. 2008. The development of scientific writing, linguistic features and historical context. Sheffield:
Equinox.
Banks, David. 2009. The position of ideology in a systemic functional model. WORD 60(1): 39–63.
Banks, David. 2016. On the (non)necessity of the hybrid category behavioural process. In Donna Miller
and Paul Bayley (eds.), 21–40.
Banks, David. 2019a. A systemic functional grammar of English. Milton Park: Routledge.
Banks, David. 2019b. SFL and diachronic studies. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine
and David Schöntal (eds.), Chap. 16. 410–432.
Barabási, Albert László. 2016. Network science. Cambridge: Cambridge University Press.
Barbara, Leila and Sara Cabral (eds.). 2013. Teoria sistêmico-funcional para brasileiros (Systemic functional
theory for Brazilians). PPGL: Programa de Pós-Graduação em Letras. Santa Maria, Brazil: Universidade
Federal de Santa Maria – UFSM.
Bardi, Mohamed Ali. 2008. A systemic functional description of the grammar of Arabic. Ph.D. thesis. Macquarie
University, Sydney.
Barnwell, Katharine Grace Lowry. 1969. A grammatical description of Mbembe (Adun Dialect): A cross river
language. Ph.D. thesis. University of London, London.
Bartlett, Tom. 2012. Hybrid voices and collaborative change: Contextualising positive discourse analysis. Routledge
critical studies in discourse. London and New York: Routledge.
Bartlett, Tom. 2013. ‘I’ll manage the context’: Context, environment and the potential for institutional
change. In Lise Fontaine, Tom Bartlett and Gerard O’Grady (eds.), 342–364.
Bartlett, Tom. 2015. Mode as a troublesome category: Social expectations and/or the construal of textual-
ity? In Wendy L. Bowcher and J.Y. Liang (eds.), Essays in honour of Ruqaiya Hasan: Society in language,
language in society. London and New York: Palgrave. 166–183.
Bartlett, Tom. 2016. Phasal dynamism and the unfolding of meaning as text. English Text Construction 9(1):
143–164.
Bartlett, Tom. 2017. Context in systemic functional linguistics: Towards scalar supervenience? In Tom
Bartlett and Gerard O’Grady (eds.), Chap. 23. 375–390.
Bartlett, Tom. 2019. Models of discourse in systemic functional linguistics. In Geoff Thompson, Wendy L.
Bowcher, Lise Fontaine and Jennifer Yameng Liang (eds.), 285–310.
Bartlett, Tom and Gerard O’Grady (eds.). 2017. The Routledge handbook of systemic functional linguistics. Mil-
ton Park: Routledge.
Bartley, Leanne Victoria. 2018. Putting transitivity to the test: A review of the Sydney and Cardiff models.
Functional Linguistics 5(4).

460
 References

Bateman, John A. 1985. Utterances in context: Towards a systemic theory of the intersubjective achievement of
discourse. Ph.D. thesis. Department of Artificial Intelligence/School of Epistemics. University of Edin-
burgh, Edingurgh.
Bateman, John A. 1988. Aspects of clause politeness in Japanese: An inquiry semantic treatment. The 26th
Annual Meeting of the Association for Computational Linguistics. Buffalo, NY: Association for Computa-
tional Linguistics. 147–154.
Bateman, John A. 1989. Dynamic systemic-functional grammar: A new frontier. Word 40(1–2): 263–287.
Bateman, John A. 1990. Upper modeling: Organizing knowledge for natural language processing. The Fifth
International Workshop on Natural Language Generation, June 1990, Pittsburgh, Pittsburgh. 54–61.
Bateman, John A. 1992. The theoretical status of ontologies in natural language processing. Proceedings of
the Workshop on ‘Text Representation and Domain Modelling: Ideas from Linguistics and AI’, The Technical
University Berlin, 1991. KIT Report 97, edited by Susanne Preuß and Birte Schmitz.
Bateman, John A. 1997. Enabling technology for multilingual natural language generation: The KPML
development environment. Natural Language Engineering 1(1): 1–42.
Bateman, John A. 2007. Towards a grande paradigmatique of film: Christian Metz reloaded. Semiotica
167(1/4): 13–64.
Bateman, John A. 2008a. Multimodality and genre: A foundation for the systematic analysis of multimodal docu-
ments. London and New York: Palgrave Macmillan.
Bateman, John A. 2008b. Systemic functional linguistics and the notion of linguistic structure: Unanswered
questions, new possibilities. In Jonathan J. Wester (ed.), Meaning in context: Implementing intelligent applica-
tions of language studies. London and New York: Continuum. 24–58.
Bateman, John A. 2014. Text and image: A critical introduction to the visual/verbal divide. London and New
York: Routledge.
Bateman, John A. 2017. The place of systemic functional linguistics as a linguistic theory in the twenty-first
century. In Tom Bartlett and Gerard O’Grady (eds.), Chap. 2. 11–26.
Bateman, John A., Joana Hois, Robert Ross and Thora Tenbrink. 2010. A linguistic ontology of space for
natural language processing. Artificial Intelligence 174: 1027–1071.
Bateman, John A., Thomas Kamps, Jörg Kleinz and Klaus Reichenberger. 2001. Towards constructive text,
diagram, and layout generation for information presentation. Computational Linguistics 27(3): 409–449.
Bateman, John A., Daniel McDonald, Tuomo Hiippala, Daniel Couto-Vale and Eugeniu Costetchi. 2019.
Systemic functional linguistics and computation: New directions, new challenges. In Geoff Thompson,
Wendy L. Bowcher, Lise Fontaine and David Schöntal (eds.), 561–586.
Bateman, John A., Christian M.I.M. Matthiessen and Zeng Licheng. 1999. Multilingual language genera-
tion for multilingual software: A functional linguistic approach. Applied Artificial Intelligence: An Interna-
tional Journal 13(6): 607–639.
Bateman, John A. and S. Momma. 1991. The nondirectional representation of systemic functional gram-
mars and semantics as typed feature structures. GMD/Institut für Integrierte Publikations- und Informations-
systeme, Darmstadt und Institut für Maschinelle Sprachverarbeitung, Universität Stuttgart.
Bateman, John A. and Mick O’Donnell. 2015. Computational linguistics: The Halliday connection. In
Jonathan J. Wester (ed.), 453–466.
Bateman, John A. and Karl-Heinrich Schmidt. 2012. Multimodal film analysis: How films mean. London and
New York: Routledge.
Bateman, John A. and Klaas J. Rondhuis. 1997. Coherence relations: Towards a general specification.
Discourse Processes 24: 3–49.
Bateman, John A. and Francisco O.D. Veloso. 2013. The semiotic resources of comics in movie adaptation:
Ang Lee’s Hulk (2003) as a case study. Studies in Comics 4(1): 135–157.
Bateman, John A., Francisco O.D. Veloso, Janina Wildfeuer, Felix Hiu Laam Cheung and Nancy Songdan
Guo. 2016. An open multilevel classification scheme for the visual layout of comics and graphic novels:
Motivation and design. Digital Scholarship in the Humanities 32(3): 476–510.
Bateman, John A., Janina Wildfeuer and Tuomo Hiippala. 2017. Multimodality: Foundations, research and
analysis: A problem-oriented introduction. Berlin: Mouton de Gruyter.
Bateson, Mary Catherine. 1963. Kinesics and paralanguage conference. Fellow Newsletter: 2–3. https://
anthrosource.onlinelibrary.wiley.com/doi/pdf/10.1111/an.1963.4.2.3
Bazell, Charles Ernest, John Cunnison Catford, M.A.K. Halliday and Robert Henry Robins (eds.). 1966.
In memory of J.R. Firth. London: Longman.
Beckner, Clay, Richard Blythe, Morten H. Christiansen, William Croft, Nick C. Ellis, John Holland,
Jinyun Ke, Diane Larsen-Freeman and Tom Shoenemann. 2009. Language is a complex adaptive

461
 References

system: position paper. Language Learning 59: Supplement 1, Language as a complex adaptive system, edited
by Nick C. Ellis and Diane Larsen-Freeman: 1–26.
Bednarek, Monika and Helen Caple. 2012. News discourse. London and New York: Continuum.
Bednarek, Monika and J.R. Martin (eds.). 2010. New discourse on language: Functional perspectives on multimo-
dality, identity, and affiliation. London: Continuum.
Bell, Philip and Theo van Leeuwen. 1994. The media interview: Confession, contest, conversation. Sydney:
University of New South Wales Press.
Bell, Roger. 1991. Translation and translating: Theory and practice. London and New York: Longman.
Benítez-Castro, Miguel-Ángel and Encarnación Hidalgo-Tenorio. 2019. Rethinking Martin and White’s
affect taxonomy: A psychologically-inspired approach to the linguistic expression of emotion. In J.
Lachlan Mackenzie and Laura Alba-Juez (eds.), Emotion in discourse. Amsterdam: Benjamins. Chap. 12.
301–331.
Benson, James D., Meena Debashish, William S. Greaves, Jennifer Lukas, Sue Savage-Rumbaugh and Jared
Taglialatela. 2004. Mind and brain in apes: A methodology for phonemic analysis of vocalizations of
language competent bonobos. Language Sciences 26: 643–660.
Benson, James D., Peter H. Fries, William S. Greaves, Kazuyoshi Iwamoto, E. Sue Savage-Rumbaugh and Jared
Taglialatela. 2002. Confrontation and support in bonobo-human discourse. Functions of Language 9(1): 1–38.
Benson, James D. and William S. Greaves. 1984. You and your language: Styles and dialects. Oxford: Pergamon
Press.
Benson, James D. and William S. Greaves. 1992. Collocation and field of discourse. In William C. Mann
and Sandra A. Thompson (eds.), Discourse description: Diverse analyses of a fund-raising text. Amsterdam:
Benjamins. 397–410.
Benson, James D. and William S. Greaves (eds.). 2009. Functional dimensions of ape-human discourse. London:
Equinox.
Benson, James D. and Paul Thibault. 2009. Language and other primate species. In M.A.K. Halliday and
Jonathan J. Wester (eds.), Chap. 6. 104–112.
Bernstein, Basil. 197l. Class, codes and control, Vol. 1: Theoretical studies towards a sociology of language. London:
Routledge and Kegan Paul.
Bernstein, Basil (ed.). 1973. Class, codes and control: Applied studies towards a sociology of language. Volume 2.
London: Routledge and Kegan Paul.
Bernstein, Basil. 1990. Class, codes and control 4: The structuring of pedagogic discourse. London: Routledge.
Bernstein, Basil. 2000 [1996]. Pedagogy, symbolic control and identity: Theory, research, critique. London: Taylor
and Francis.
Berry, Margaret. 1975/7. Introduction to systemic linguistics 1 and 2. London: Batsford.
Berry, Margaret. 1981. Systemic linguistics and discourse analysis: A multi-layered approach to exchange
structure. In Malcolm Coulthard and Michael Montgomery (eds.), Studies in discourse analysis. London:
Routledge and Kegan Paul. 120–145.
Berry, Margaret. 1995. Thematic options and success in writing. In Mohsen Ghadessy (ed.), Thematic devel-
opment in English texts. London: Pinter. 55–84.
Berry, Margaret. 1996. What is theme? A(nother) personal view. In Margaret Berry Christopher Butler,
Robin Fawcett and Guowen Huang (eds.), Meaning and form: Systemic functional interpretations. Nor-
wood, NJ: Ablex. 1–65.
Berry, Margaret. 2017. Stratum, delicacy, realisation and rank. In Tom Bartlett and Gerard O’Grady (eds.),
Chap. 4. 42–55.
Berry, Margaret. 2019. The clause: An overview of the lexicogrammar. In Geoff Thompson, Wendy L.
Bowcher, Lise Fontaine and David Schöntal (eds.), 92–117.
Berry, Margaret, Christopher Butler, Robin Fawcett and Guowen Huang (eds.). 1996. Meaning and form:
Systemic functional interpretations. Volume 2 of Meaning and choice in language: Studies for Michael Hal-
liday. Norwood, NJ: Ablex.
Bertalanffy, Ludwig von. 1968. General system theory: Foundations, development, applications. New York:
Braziller.
Bhatia, Vijay K. 2004. Worlds of written discourse. London and New York: Continuum.
Biber, Douglas. 1988. Variation across speech and writing. Cambridge: Cambridge University Press.
Biber, Douglas. 1994. An analytical framework for register studies. In Douglas Biber and Edward Finegan
(eds.), Sociolinguistic perspectives on register. Oxford: Oxford University Press. 31–56.
Biber, Douglas. 1995. Dimensions of register variation: A cross-linguistic comparison. Cambridge: Cambridge
University Press.

462
 References

Biber, Douglas. 2010. Corpus-based and corpus-driven analyses of language variation and use. In Bernd
Heine and Heiko Narrog (eds.), The Oxford handbook of linguistic analysis. Oxford: Oxford University
Press. 169–191.
Biber, Douglas and Susan Conrad. 2009. Register, genre, and style. Cambridge: Cambridge University Press.
Biber, Douglas and Edward Finegan (eds.). 1994. Sociolinguistic perspectives on register. Oxford: Oxford Uni-
versity Press.
Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad and Edward Finegan. 1999. The Longman
grammar of spoken and written English. London: Longman.
Birch, David and Michael O’Toole (eds.). 1988. Functions of style. London: Frances Pinter.
Birdwhistell, Ray L. 1952. Introduction to kinesics: An annotated system for analysis of body motion and gesture.
Washington, DC: Dept. of State, Foreign Service Institute.
Birdwhistell, Ray L. 1970. Kinesics and context: Essays on body-motion communication. London: Allen Lane
the Penguin Press.
Bloomfield, Leonard. 1933. Language. London: Allen and Unwin.
Bloor, Meriel and Thomas Bloor. 1992. Given and new information in the thematic organization of
text: An application to the teaching of academic writing. Occasional Papers in Systemic Linguistics 6:
33–43.
Bloor, Meriel and Thomas Bloor. 2007. The practice of critical discourse analysis: An introduction. London:
Routledge.
Bloor, Thomas and Meriel Bloor. 1995. The functional analysis of English: A Hallidayan approach. London:
Edward Arnold.
Bloor, Tom and Meriel Bloor. 2004. Functional analysis of English. 2nd edition. Hodder and Stoughton
Educational.
Bloor, Tom and Meriel Bloor. 2013. Functional analysis of English. 3rd edition. London and New York:
Routledge.
Boas, Hans C. and Ivan A. Sag (eds.). 2010. Sign-based construction grammar. Stanford, CA: Center for the
Study of Language and Information.
Bod, Rens, Jennifer Hay and Stefanie Jannedy (eds.). 2003. Probabilistic linguistics. Cambridge, MA: MIT
Press.
Bohm, David. 1980. Wholeness and the implicate order. London and New York: Routledge.
Bottenberg, Donna, Margaret Lemme and Natalie Hedberg. 1985. Analysis of oral narratives of normal
and aphasic adults. In Robert H. Brookshire (ed.), Clinical aphasiology: conference proceedings. Minneapolis:
BRK Publishers. 241–247.
Boulding, Kenneth. 1956. General systems theory: The skeleton of a science. Management Science 2(3):
197–208. Reprinted in E:CO 6(1–2), (Fall 2004): 127–139.
Bowcher, Wendy L. 1999. Investigating institutionalization in context. In Mohsen Ghadessy (ed.), Text and
context in functional linguistics. Amsterdam: John Benjamins Publishers. 141–176.
Bowcher, Wendy L. 2001. Play by play talk on radio: An enquiry into some relations between language and context.
Ph.D. thesis. University of Liverpool, Liverpool.
Bowcher, Wendy L. 2003. Speaker contributions in radio sports commentary. Text 23(4): 445–476.
Bowcher, Wendy L. 2004. Theme and New in play-by-play radio sports commentating. In David Banks
(ed.), Text and texture: Systemic functional viewpoints on the nature and structure of text. Paris: L’Harmattan.
455–493.
Bowcher, Wendy L. 2007. Field and multimodal texts. In Ruqaiya Hasan, Christian M.I.M. Matthiessen
and Jonathan J. Wester (eds.), 620–646.
Bowcher, Wendy L. 2010. The history and theoretical development of context of situation in systemic func-
tional linguistics. Annual Review of Functional Linguistics 2: 64–93.
Bowcher, Wendy L. (ed.). 2012. Multimodal texts from around the world: Cultural and linguistic insights. Basing-
stoke, Hampshire: Palgrave Publishers.
Bowcher, Wendy L. 2014. Issues in developing unified systems for contextual field and mode. Functions of
Language 21(2): 176–209.
Bowcher, Wendy L. 2017. Field, tenor and mode. In Tom Bartlett and Gerard O’Grady (eds.), Chap. 24.
391–403.
Bowcher, Wendy L. 2019. Context and register. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine
and David Schönthal (eds.), 142–170.
Bowcher, Wendy L. and Meena Debashish. 2019. Intonation. In Geoff Thompson, Wendy L. Bowcher,
Lise Fontaine and David Schönthal (eds.), 171–203.

463
 References

Bowcher, Wendy L. and Jennifer Yameng Liang (eds.). 2016. Society in language, language in society: Essays in
honour of Ruqaiya Hasan. Houndsmills, Basingstoke: Palgrave Macmillan.
Bowcher, Wendy L. and Bradley Smith (eds.). 2014. Systemic phonology: Recent studies in English. Sheffield
and Bristol: Equinox.
Boxwell, Maurice. 1990. Co-referentiality through nominal elements in Weri. Ph.D. thesis. Macquarie Univer-
sity, Sydney.
Boxwell, Maurice. 1995. Nothing makes sense in Weri: A case of extensive ellipsis in nominals in a Papuan
language. In Ruqaiya Hasan and Peter H. Fries (eds.), On subject and theme: A discourse functional perspec-
tive. Amsterdam: Benjamins. 123–151.
Brazil, David. 1975. Discourse intonation. Birmingham: English Language Research.
Brazil, David. 1997. The communicative value of intonation in English. Cambridge: Cambridge University
Press.
Bresnan, Joan, Ash Asudeh, Ida Toivonen and Stephen Wechsler. 2016. Lexical-functional syntax. 2nd edi-
tion. Hoboken, NJ: Wiley Blackwell.
Brezina, Vaclav. 2018. Statistics in corpus linguistics: A practical guide. Cambridge: Cambridge University Press.
Brisk, María Estela and Mary. J. Schleppegrell. 2021. Language in action: SFL theory across contexts. Sheffield:
Equinox.
Brown, Richard F. 2014. Physician: Patient communication about cancer clinical trials. In Heidi Hamilton
and Wen-ying Sylvia Chou (eds.), Routledge handbook of language and healthcare communication. London:
Routledge. 615–628.
Brown, Richard F., P. Butow, D. Butt, A.R. Moore, and M. Tattersall. 2004. Developing ethical strategies
to assist oncologists in seeking informed consent to cancer clinical trials. Social Science and Medicine 58:
379–390.
Brown, Roger and Albert Gilman. 1960. The pronouns of power and solidarity. In Thomas A. Sebeok
(ed.), Style in language. Cambridge, MA: MIT Press. 253–276.
Bruce, Bertram C. 1975. Generation as social action. Theoretical Issues in Natural Language Processing 1:
64–67. Reprinted in Grosz, Barbara J., Karen Sparck Jones and Bonnie Lynn Webber (eds.). 1986.
Readings in natural language processing. Los Altos, CA: Morgan Kaufman. 419–422.
Bruner, Jerome. 1975. The Ontogenesis of speech acts. Journal of Child Language 2(1): 1–19.
Bruner, Jerome. 1986. Actual minds, possible worlds. Cambridge, MA: Harvard University Press.
Bühler, Karl. 1934. Sprachtheorie. Die Darstellungsfunktion der Sprache. Stuttgart, New York: G. Fischer 1982.
Transl.: Theory of language: The representational function of language. Amsterdam and Philadelphia: John
Benjamins.
Burns, Anne. 1999. Collaborative action research for English language teachers. Cambridge: Cambridge Univer-
sity Press.
Burns, Anne. 2010. Doing action research in English language teaching: A guide for practitioners. London and
New York: Routledge.
Burns, Anne. 2015. Action research. In Brian Paltridge and Aek Phakiti (eds.), Research methods in
applied linguistics: A practical guide. 2nd edition. London and New York: Bloomsbury Academic.
241–256.
Burns, Anne, Mira Kim and Christian M.I.M. Matthiessen. 2009. Doctoral work in translation studies as an
interdisciplinary mutual learning process: How a translator, teacher educator, and linguistic typologist
worked together. The Interpreter and Translator Trainer 3(1): 107–128.
Burton, Deidre. 1980. Dialogue and discourse. London: Routledge and Kegan Paul.
Butler, Christopher S. 1982. The directive function of the English modals. Ph.D. thesis. University of Notting-
ham, Nottingham.
Butler, Christopher S. 1988. Politeness and the semantics of modalised directives in English. In James D.
Benson, Michael J. Cummings and William S. Greaves (eds.), Linguistics in a systemic perspective. Amster-
dam: Benjamins. 119–154.
Butler, Christopher S. 2005. Focusing on focus: A comparison of functional grammar, role and reference
grammar and systemic functional grammar. Language Sciences 27: 585–618.
Butler, Christopher S. 2019. SFL in context. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine and
David Schöntal (eds.), 259–282.
Butler, Christopher S. and Francisco Gonzálvez-García. 2014. Exploring functional-cognitive space. Amster-
dam and Philadelphia: John Benjamins.
Butt, David G. 1984. The relationship between theme and lexicogrammar in the poetry of Wallace Stevens. Ph.D.
thesis. Macquarie University, Sydney.

464
 References

Butt, David G. 2001. Firth, Halliday and the development of systemic theory: History of the language sciences.
Volume 2. Sylvain Auroux, E.F.K. Koerner and Kees Versteegh (eds.). Berlin and New York: Walter de
Gruyter. 1806–1838.
Butt, David G. 2004. Parameters of context: On establishing the similarities and differences between social processes.
Unpublished mimeo. Sydney: Macquarie University.
Butt, David G. 2008. The robustness of realizational systems. In Jonathan J. Wester (ed.), Meaning in context:
Implementing intelligent applications in language studies. London and New York: Continuum. 59–83.
Butt, David G. 2015. The ‘history of ideas’ and Halliday’s natural science of meaning. In Jonathan J. Wester
(ed.), 17–61.
Butt, David G. 2019. Firth and the origins of systemic functional linguistics: Process, pragma, and polysys-
tem. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine and David Schöntal (eds.), 11–34.
Butt, David G., Rhondda Fahey and Caroline Henderson-Brooks. 2003. Outer and inner weathers. In
Russell Meares and Patrick Noolan (eds.), The self in conversation. Volume 2. Sydney: Australia and New
Zealand Association of Psychotherapy.
Butt, David G., Rhondda Fahey, Susan Feez, Sue Spinks and Colin Yallop. 2000. Using functional grammar:
An explorer’s guide. 2nd edition. Sydney: Macquarie University, NCELTR (National Centre for English
Language Teaching and Research).
Butt, David G., Rhondda Fahey, Sue Spinks and Colin Yallop. 1995. Using functional grammar: An explorer’s
guide. Sydney: Macquarie University, NCELTR (National Centre for English Language Teaching and
Research).
Butt, David G. and Jennifer Yameng Liang. 2015. In her own words: An interview with Ruqaiya Hasan.
In Wendy L. Bowcher and Jennifer Yameng Liang (eds.), 381–411.
Butt, David G., Annabelle Lukin and Christian M.I.M. Matthiessen. 2004. Grammar: The first covert
operation of war. Discourse and Society 15(2–3): 267–290.
Butt, David G., Alison R. Moore and Caroline K. Henderson-Brooks. 2012. Discourse correlates of the
therapeutic method and patient progress. In Russell Meares, Nick Bendit, Joan Haliburn, Anthony
Korner, Dawn Mears and David Butt (eds.), Borderline personality disorder and the conversational model: A
clinician’s manual. New York: Norton and Company. 267–290.
Butt, David G. and Jonathan J. Webster. 2017. The logical metafunction in systemic functional linguistics.
In Tom Bartlett and Gerard O’Grady (eds.), Chap. 7. 96–114.
Butt, David G. and Rebekah Kate Ardley Wegener. 2007. The work of concepts: Context and metafunc-
tion in the systemic functional model. In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan
J. Wester (eds.), 589–618.
Bybee, Joan L. and Clay Beckner. 2010. Usage-based theory. In Bernd Heine and Heiko Narrog (eds.), The
Oxford handbook of linguistic analysis. Oxford: Oxford University Press. 828–855.
Bybee, Joan and Paul Hopper (eds.). 2001. Frequency and the emergence of linguistic structure. Amsterdam and
Philadelphia: John Benjamins.
Byrnes, Heidi (ed.). 2006. Advanced instructed language learning: The complementary contribution of Halliday and
Vygotsky. London and New York: Continuum.
Byrnes, Heidi. 2009. Emergent L2 German writing ability in a curricular context: A longitudinal study of
grammatical metaphor. Linguistics and Education 20(1): 50–66.
Byrnes, Heidi. 2019. Applying SFL for understanding and fostering instructed second language develop-
ment. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine and David Schöntal (eds.), 512–536.
Byrnes, Heidi, Hiram H. Maxim and John Norris. 2010. Realizing advanced foreign language writing
development in collegiate education: Curricular design, pedagogy, assessment. The Modern Language
Journal 94(Supplement).
Cabral, Sara Regina Scotta, Orlando Vian Jr., Cristiane Fuzer and Karen Andresa Teixeira Santorum.
2021. Lista de termos da linguística sistêmico-funcional em português brasileiro: Léxico-gramática. (List
of terms of systemic functional linguistics in Brazilian Portuguese: Lexicogrammar). Organon 36(71):
483–495.
Caffarel, Alice. 1992. Interacting between a generalized tense semantics and register-specific semantic tense
systems: A bi-stratal exploration of the semantics of French tense. Language Sciences 14(4): 385–418.
Caffarel, Alice. 2004. Metafunctional profile of the grammar of French. In Alice Caffarel, J.R. Martin and
Christian M.I.M. Matthiessen (eds.), 77–137.
Caffarel, Alice. 2006a. Learning advanced French through SFL: Learning SFL in French. In Heidi Byrnes
(ed.), Advanced instructed language learning: The complementary contribution of Halliday and Vygotsky. London
and New York: Continuum. 204–224.

465
 References

Caffarel, Alice. 2006b. A systemic functional grammar of French: From grammar to discourse. London and New
York: Continuum.
Caffarel, Alice. 2010. Systemic functional grammar and the study of meaning. In Bernd Heine and
Heiko Narrog (eds.), The Oxford handbook of linguistic analysis. Oxford: Oxford University Press.
797–825.
Caffarel, Alice. 2017. The verbal group in French. In Tom Bartlett and Gerard O’Grady (eds.), Chap. 19.
319–337.
Caffarel, Alice, J.R. Martin and Christian M.I.M. Matthiessen (eds.). 2004. Language typology: A functional
perspective. Amsterdam and Philadelphia: John Benjamins.
Caldas-Coulthard, Carmen Rosa. 2003. Cross-cultural representation of ‘otherness’ in media discourse. In
Gilbert Weiss and Ruth Wodak (eds.), Critical discourse analysis: Theory and interdisciplinarity. Houndmills:
Palgrave Macmillan. 272–296.
Caldas-Coulthard, Carmen Rosa and Malcolm Coulthard (eds.). 1995. Texts and practices: Readings in critical
discourse analysis. London: Routledge.
Caldas-Coulthard, Carmen Rosa and Rosamund Moon. 2010. ‘Curvy, hunky, kinky’: Using corpora as
tools for critical analysis. Discourse & Society 21(2): 99–133.
Caldas-Coulthard, Carmen Rosa and Theo van Leeuwen. 2002. Stunning, shimmering, iridescent: Toys as
the representation of gendered social actors. In Lia Litosseliti and Jane Sunderland (eds.), Gender identity
and discourse analysis. Amsterdam: John Benjamins. 91–108.
Caldwell, David. 2007. The rhetoric of rap: A challenge to dominant forces? Bridging Discourses: ASFLA
2007 Online Proceedings. Woollongong: University of Woollongong. 1–15.
Caldwell, David. 2008. Affiliating with rap music: Political rap or gangsta rap. Novitas-ROYAL 2(1): 13–27.
Caldwell, David. 2010. Making metre mean: Identity and affiliation in the rap music of Kanye West. In
Monika Bednarek and J.R. Martin (eds.), 59–79.
Caldwell, David. 2014a. The interpersonal voice: Applying appraisal to the rap and sung voice. Social
semiotics 24(1): 40–55.
Caldwell, David. 2014b. A comparative analysis of rapping and singing: Perspectives from systemic pho-
nology, social semiotics and music studies. In Wendy Bowcher and Brad Smith (eds.), Recent studies in
systemic phonology volume I: Focus on the English language. London: Equinox. 235–263.
Caldwell, David and Michele Zappavigna. 2011. Visualizing multimodal patterning. In Shoshana Dreyfus,
Susan Hood and Maree Stenglin (eds.), Semiotic margins: Meaning in multimodalities. London and New
York: Continuum. 229–242.
Candlin, Sally. 2000. New dynamics in the nurse: Patient relationship? In Srikant Sarangi and Malcolm
Coulthard (eds.), Discourse and social life. Harlow: Pearson Educational. 230–245.
Caple, Helen. 2008. Reconciling the co-articulation of meaning between words and pictures: Explor-
ing instantiation and commitment in image nuclear news stories. In Ahmar Mahboob and Naomi K.
Knight (eds.), Questioning linguistics. Newcastle: Cambridge Scholars Publishing. 77–94.
Capra, Fritjof. 1996. The web of life. New York: Doubleday.
Cartmill, John, Alison Moore, David Butt, L. Squire. 2007. Surgical teamwork: Systemic functional lin-
guistics and the analysis of verbal and non verbal meaning in surgery. ANZ Journal of Surgery 77(s1):
A79.
Caruso, Assunta, Antonietta Folino, Francesca Parisi and Roberto Trunfio. 2014. A statistical method for
minimum corpus size determination. JADT 2014: 12es Journées internationales d’analyse de données tex-
tuelles. Paris. 135–146.
Catford, J.C. 1965. A linguistic theory of translation. London: Oxford University Press.
Catford, J.C. 1969. J.R. Firth and British linguistics. In Archibald A. Hill (ed.), Linguistics today. New York
and London: Basic Books, Inc. 218–228.
Catford, J.C. 1977. Fundamental problems in phonetics. Indiana: Indiana University Press.
Catford, J.C. 1985. ‘Rest’ and ‘open transition’ in a systemic phonology of English. In James D. Benson and
William S. Greaves (eds.), Systemic perspectives on discourse. Norwood NJ: Ablex. 333–348.
Chandler, Eloise, Diana Slade, Jack Pun, Graham Lock, Christian M.I.M. Matthiessen, Elaine Espindola
and Carol Ng. 2015. Communication in Hong Kong accident and emergency departments: The clini-
cians’ perspectives. Global Qualitative Nursing Journal 2: 1–11.
Chen, Jing. 2019. On re-instantiation of literary dialects: A systemic functional approach. Theory and Prac-
tice in Language Studies 9(6): 706–711.
Chen, Shukun, Winfred Wenhui Xuan and Hailing Yu. 2022. Applying systemic functional linguistics in
translation studies: A research synthesis. Babel, International Journal of Translation. 68(4): 517–545.

466
 References

Cheung, Germaine. 2018. Public health and risk communication: The experience of the severe acute respiratory
syndrome (SARS) epidemic and the swine influenza pandemic. Ph.D. thesis. The Hong Kong Polytechnic
University, Hong Kong.
Chik, S.Y. Sonya. 2018. The language of privacy in Japanese and English written discourse: A systemic functional
perspective. Ph.D. thesis. The Hong Kong Polytechnic University, Hong Kong.
Choi, Gyunghee. 2013. A study on logical meaning using SFL and the implications of this for translation studies.
Ph.D. thesis. University of New South Wales, Sydney.
Chomsky, Noam. 1957. Syntactic structures. The Hague: Mouton.
Chomsky, Noam. 1965. Aspects of the theory of syntax. Cambridge, MA: MIT Press.
Chomsky, Noam and Morris Halle. 1968. The sound pattern of English. New York: Harper and Row.
Chouliaraki, Lilie and Norman Fairclough. 1999. Discourse in late modernity: Rethinking critical discourse analy-
sis. Edinburgh: Edinburgh University Press.
Christie, Frances. 1985. Language education. Geelong, Vic.: Deakin University Press.
Christie, Frances. 1990. Curriculum genres in early childhood education: A case study in writing development. Ph.D.
thesis. The University of Sydney, Sydney.
Christie, Frances. 1994. Developing an educational linguistics for English language teaching: A systemic
functional linguistic perspective. Functions of Language 1(2): 95–127.
Christie, Frances. 2002. Classroom discourse analysis. London: Continuum.
Christie, Frances. 2005. Language education in the primary years. Sydney: University of New South Wales
Press.
Christie, Frances. 2012. Language education throughout the school years: A functional perspective. Oxford:
Wiley-Blackwell.
Christie, Frances and Beverley Derewianka. 2008. School discourse: Learning to write across the years of schooling.
London and New York: Continuum.
Christie, Frances and Mary Macken-Horarik. 2007. Building verticality in subject English. In Frances
Christie and J.R. Martin (eds.), Language, knowledge and pedagogy: Functional linguistic and sociological per-
spectives. London and New York: Continuum. 156–183.
Christie, Frances and J.R. Martin (eds.). 1997. Genre and institutions: Social processes in the workplace and
school. London: Cassell.
Christie, Frances and J.R. Martin (eds.). 2007. Language, knowledge and pedagogy: Functional linguistic and
sociological perspectives. London and New York: Continuum.
Christie, Frances and Ray Misson (eds.). 1998. Schooling and literacy. London and New York: Routledge.
Christie, Frances and Len Unsworth. 2005. Developing dimensions of an educational linguistics. In Ruqa-
iya Hasan, Christian M.I.M. Matthiessen and Jonathan J. Wester (eds.), 217–250.
Clarke, Ben. 2017. Cohesion in systemic functional linguistics: A theoretical reflection. In Tom Bartlett
and Gerard O’Grady (eds.), Chap. 25. 404–417.
Clarke, Benjamin. 2012. Do patterns of ellipsis in text support systemic functional linguistics’ ‘context‐metafunction
hook‐up’ hypothesis? Ph.D. Cardiff University, Cardiff.
Clayton, J., K.M. Hancock, P.N. Butow, M.H. Tattersall, D.C. Currow, et al. 2007. Clinical practice
guidelines for communicating prognosis and end-of-life issues with adults in the advanced stages of a
life-limiting illness, and their caregivers. MJA 186: 76–108.
Cléirigh, Chris. 1998. A selectionist model of the genesis of phonic texture: Systemic phonology and universal Dar-
winism. Ph.D. thesis. The University of Sydney, Sydney.
Cloran, Carmel. 1989. Learning through language: The social construction of gender. In Ruqaiya Hasan
and J.R. Martin (eds.), Language development: Learning language, learning culture. New York: Ablex.
111–151.
Cloran, Carmel. 1994. Rhetorical units and decontextualisation: An enquiry into some relations of context, meaning
and grammar. Nottingham: School of English Studies, Nottingham University.
Cloran, Carmel. 1995. Defining and relating text segments: Subject and theme in discourse. In Ruqaiya
Hasan and Peter Fries (eds.), 361–403.
Cloran, Carmel. 1999. Contexts for learning. In Fran Christie (ed.), Pedagogy and the shaping of consciousness.
London: Cassell. 31–65.
Cloran, Carmel. 2000. Socio-semantic variation: Different wordings, different meanings. In Len Unsworth
(ed.), Researching language in schools and communities: Functional linguistic perspectives. London: Cassell.
152–183.
Cloran, Carmel, Virginia Stuart-Smith and Lynne Young. 2007. Models of discourse. In Ruqaiya Hasan,
Christian M.I.M. Matthiessen and Jonathan J. Wester (eds.), 645–668.

467
 References

Coelho, Carl A. 2002. Story narratives of adults with closed head injury and non-brain-injured adults:
Influence of socioeconomic status, elicitation task and executive functioning. Journal of Speech Language
and Hearing Research 45(6): 1232–1248.
Coelho, Carl A., Betty Z. Liles and Robert J. Duffy. 1991a. The use of discourse analyses for the evaluation
of higher level traumatically brain-injured adults. Brain Injury 5(4): 381–392.
Coelho, Carl A., Betty Z. Liles and Robert J. Duffy. 1991b. Analysis of conversational discourse in head-
injured patients. Journal of Head Trauma Rehabilitation 6(2): 92–99.
Coelho, Carl A., Betty Z. Liles, Robert J. Duffy, Janine V. Clarkson and Deanne Elia. 1994. Longitudinal
assessment of narrative discourse in a mildly aphasic adult. Clinical Aphasiology 22: 145–155.
Coffin, Caroline. 1997. Constructing and giving value to the past: an investigation into secondary school
history. In Frances Christie and J.R. Martin (eds.), Genre and institutions: Social processes in the workplace
and school. London: Cassell. 196–230.
Coffin, Caroline. 2006. Historical discourse. London and New York: Continuum.
Coffin, Caroline, Mary Jane Curry, Sharon Goodman, Ann Hewings, Theresa Ann Lillis and Joan Swann.
2003. Teaching academic writing: A toolkit for higher education. London: Routledge.
Coffin, Caroline, Jim Donohue and Sarah North. 2009. Exploring English grammar: From formal to functional.
London: Routledge.
Coffin, Caroline and Kieran O’Halloran. 2010. Describing English. In Janet Maybin and Joan Swann
(eds.), The Routledge companion to English language studies. London and New York: Routledge.
11–41.
Cohen, Andrew D. and Jonathan Fine. 1978. Reading history in English: Discourse analysis and the expe-
rience of native and non-native readers. Working Papers in Bilingualism 16: 55–74.
Cohen, Inbal. 2011. The expression of schizophrenia through interpersonal systems at the level of discourse semantics.
PhD thesis. Bar-Ilan University, Tel Aviv.
Cohen, Philip R., Jerry Morgan and Martha E. Pollack (eds.). 1990. Intentions in communication. Cam-
bridge, MA: The MIT Press.
Cohen, Phillip R. and C. Raymond Perrault. 1979. Elements of a plan-based theory of speech acts. Cogni-
tive Science 3(3): 177–212.
Colby, Benjamin N. 1966. Cultural patterns in narrative. Science 151: 793–798.
Colby, Benjamin N. and Lore M. Colby. 1981. The daykeeper: The life and discourse of an Ixil Diviner. Cam-
bridge, MA and London: Harvard University Press.
Collins, Peter J. 1991a. Pseudo cleft and cleft constructions: A thematic and informational interpretation.
Linguistics 29: 481–519.
Collins, Peter J. 1991b. Cleft and pseudo-cleft constructions in English. London and New York: Routledge.
Collins, Peter J. 1992. Cleft existentials in English. Language Sciences 14(4): 419–435.
Cooper, Harris. 2017. Research synthesis and meta-analysis: A step-by-step approach. 5th edition. London:
SAGE.
Cooper, Harris and Larry V. Hedges. 2009. Research synthesis as a scientific process. In Harris and Val-
entine (eds.), 3–16.
Cooper, Harris and Jeffrey C. Valentine (eds.). 2009. The handbook of research synthesis and meta-analysis. 2nd
edition. New York: Russell Sage Foundation.
Coulthard, Malcolm. 1985. An introduction to discourse analysis. 2nd edition. London: Longman.
Coulthard, Malcolm (ed.). 1992. Advances in spoken discourse analysis. London: Routledge.
Coulthard, Malcolm (ed.). 1994. Advances in written text analysis. London: Routledge.
Coulthard, Malcolm and Alison Johnson (eds.). 2010. The Routledge handbook of forensic linguistics. London
and New York: Routledge.
Coulthard, Malcolm, Alison Johnson and David Wright. 2017. An introduction to forensic linguistics: Language
in evidence. 2nd edition. London and New York: Routledge.
Covington, Michael A. 1984. Syntactic theory in the high middle ages: Modistic models of sentence structure. Cam-
bridge: Cambridge University Press.
Cranny-Francis, Anne. 1990. Feminist fiction: Feminist uses of generic fiction. London: Polity.
Cranny-Francis, Anne and J.R. Martin. 1991. Contratextuality: The poetics of subversion. In Fran Christie
(ed.), Literacy in social processes: Papers from the inaugural Australian Systemic Linguistics Conference, held at
Deakin university, January 1990. Darwin: Centre for Studies in Language in Education, Northern Ter-
ritory University. 286–344.
Croft, William. 2001. Radical construction grammar: Syntactic theory in typological perspective. Oxford: Oxford
University Press.

468
 References

Cross, Marilyn. 1992. Choice in lexis: Computer generation of lexis as most delicate grammar. Language
Sciences 14(4): 579–607.
Crystal, David. 1969. Prosodic systems and intonation in English. Cambridge: Cambridge University Press.
Crystal, David. 1975. The English tone of voice. London: Edward Arnold.
Crystal, David. 1981. Clinical linguistics. Vienna: Springer.
Crystal, David. 2002. Clinical linguistics and phonetics’ first 15 years: An introductory comment. Clinical
Linguistics and Phonetics 16(7): 487–489.
Crystal, David. 2013. Clinical linguistics: Conversational reflections. Clinical Linguistics and Phonetics 27(4):
236–243.
Crystal, David and Randolph Quirk. 1964. Systems of prosodic and paralinguistic features in English. The
Hague: Mouton.
Cummings, Michael. 1995. A systemic functional approach to the thematic structure of the Old English
clause. In Hasan and Fries (eds.), 275–316.
Cummings, Michael J. and Robert Simmons. 1983. The language of literature: A stylistic introduction to the
study of literature. Oxford: Pergamon Press.
Dalton-Puffer, Christiane. 2007. Discourse in content and language integrated (CLIL) classrooms. Amsterdam:
Benjamins.
Daneš, František. 1964. A three-level approach to syntax. Travaux Linguistiques de Prague 1: 225–240.
Daneš, František (ed.). 1974a. Papers on functional sentence perspective. Prague: Academia.
Daneš, František. 1974b. Functional sentence perspective and the organisation of the text. In Daneš (ed.),
106–128.
Davidse, Kristin. 1986. M.A.K. Halliday’s functional grammar and the Prague School. In René Dirven and
Vilém Fried (eds.), Functionalism in linguistics. Amsterdam: Benjamins. 39–79.
Davidse, Kristin. 1991. Categories of experiential grammar. Ph.D. thesis. Catholic University of Leuven,
Leuven.
Davidse, Kristin. 1992a. Transitive/ergative: The Janus-headed grammar of actions and events. In Martin
Davies and Louise Ravelli (eds.), Advances in systemic linguistics. London: Pinter. 105–135.
Davidse, Kristin. 1992b. A semiotic approach to relational clauses. Occasional Papers in Systemic Linguistics
6: 99–131.
Davidse, Kristin. 1992c. Existential constructions: A systemic perspective. Leuven Contributions in Linguistics
and Philology 81: 71–99.
Davidse, Kristin. 1996a. Turning grammar on itself: Identifying clauses in linguistic discourse. In Chris-
topher Butler, Margaret Berry, Robin Fawcett and Guowen Huang (eds.), Meaning and form: Systemic
functional interpretations. Norwood, NJ: Ablex. 367–393.
Davidse, Kristin. 1996b. Ditransitivity and possession. In Carmel Cloran, David Butt and Ruqaiya Hasan
(ed.), Functional descriptions: Theory in practice. Amsterdam: Benjamins. 85–144.
Davidse, Kristin. 1999. Categories of experiential grammar. Nottingham: University of Nottingham.
Davidse, Kristin. 2017. Systemic functional linguistics and the clause: The experiential metafunction. In
Tom Bartlett and Gerard O’Grady (eds.), Chap. 4. 79–95.
Davidse, Kristin and Tine Breban. 2019. A cognitive-functional approach to the order of adjectives in the
English noun phrase. Linguistics 57(2): 327–371.
Davidse, Kristin and Liesbet Heyvaert (eds.). 2004. Functional linguistics and contrastive description: Special issue
of languages in contrast 4:1 (2002/03). Amsterdam and Philadelphia: Benjamins.
Davies, Martin. 1996. Theme and information until Shakespeare. In Christopher Butler, Margaret Berry,
Robin Fawcett and Guowen Huang (eds.), Meaning and form: Systemic functional interpretations. Nor-
wood, NJ: Ablex. 113–150.
Davis, Albyn G. and Carl A. Coelho. 2004. Referential cohesion and logical coherence of narration after
closed head injury. Brain and Language 89: 508–523.
Deacon, Terrence. 1992. Brain-language coevolution. In John A. Hawkins and Murray Gell-Mann (eds.),
The evolution of human languages. Redwood City, CA: Addison-Wesley. (Proceedings Volume XI, Santa
Fe Institute Studies in the Sciences of Complexity). 49–85.
Deacon, Terrence. 1997. The symbolic species: The co-evolution of language and the human brain. Harmond-
sworth: Penguin Books.
Degand, Liesbeth. 1996. Causation in Dutch and French: Interpersonal aspects. Ruqaiya Hasan, Carmel
Cloran and David Butt (eds.), Functional descriptions: Theory into practice. Amsterdam: Benjamins. 207–237.
Delin, Judith L. and John A. Bateman. 2002. Describing and critiquing multimodal documents. Document
Design 3(2): 140–155.

469
 References

Denzin, Norman K. and Yvonna S. Lincoln (eds.). 2018. The SAGE handbook of qualitative research. 5th
edition. London: SAGE.
Derewianka, Beverly. 1995. Language development in the transition from childhood to adolescence: The role of gram-
matical metaphor. PhD thesis. Macquarie University, Sydney.
Derewianka, Beverly. 2003. Grammatical metaphor in the transition to adolescence. In Simon-Vandenber-
gen, Taverniers and Ravelli (eds.), 185–220.
Derewianka, Beverly and Pauline Jones. 2016. Teaching language in context. 2nd edition. Melbourne: Oxford
University Press.
de Beaugrande, Robert. 1991. Linguistic theory: The discourse of fundamental works. London: Longman.
de Beaugrande, Robert. 1997. The story of discourse analysis. In van Dijk (ed.), 35–62.
de Bot, Kees. 2015. A History of applied linguistics: From 1980 to the present. London and New York:
Routledge.
de Joia, Alex and Adrian Stenton (eds.). 1980. Terms in systemic linguistics: A guide to Halliday. London:
Batsford.
de Silva Joyce, Helen and Susan Feez. 2016. Exploring literacies: Theory, research and practice. New York:
Palgrave Macmillan.
de Silva Joyce, Helen, Diana Slade, Deborah Bateson, Hermine Scheeres, Jeannette McGregor and
EcGregor Weisberg. 2015. Patient-centred discourse in sexual and reproductive health consultations.
Discourse and Communication 9: 275–292.
de Villiers, Jessica, Jonathan Fine, Gary M. Ginsberg, Liezanne Vaccarella and Peter Szatmari. 2007. A scale
for rating conversational impairment in autism spectrum disorder. Journal of Autism and Developmental
Disorders 37(7): 1375–1380.
de Villiers, Jessica and Robert J. Stainton (eds.). 2001. Communication in linguistics: Papers in honour of Michael
Gregory. Toronto: Editions du Gref.
Diamond, Jared. 2013. The world until yesterday: What can we learn from traditional societies? Harmondsworth:
Penguin Books.
Dijkstra, Katinka, Michelle S. Bourgeois, Rebecca S. Allen, and Louis D. Burgioc. 2004. Conversational
coherence: Discourse analysis of older adults with and without dementia. Journal of Neurolinguistics
17(4): 263–283.
Dik, Simon. 1978. Functional grammar. Amsterdam: North-Holland.
Dinneen, Francis P. 1967. An introduction to general linguistics. Washington, DC: Georgetown University
Press.
Dixon, R.M.W. 1994. Ergativity. Cambridge: Cambridge University Press.
Djonov, Emilia and Sumin Zhao. 2013. From multimodal to critical multimodal studies through popular
discourse. In Emilia Djonov and Sumin Zhao (eds.), Critical multimodal studies of popular culture. London:
Routledge. 1–16.
Doran, Yaegan, John. 2016. Knowledge in physics through mathematics, image and language. PhD thesis. The
University of Sydney, Sydney.
Downing, Angela. 1991. An alternative approaches to theme: A systemic-functional perspective. Word
42(2): 119–144.
Downing, Angela. 1996. Discourse-pragmatic distinctions of the past in present in English and Spanish.
In Christopher Butler, Margaret Berry Robin Fawcett and Guowen Huang (eds.), Meaning and form:
Systemic functional interpretations. Norwood, NJ: Ablex. 509–533.
Dreyfus, Shoshana J., Susan Hood and Maree Stenglin (eds.). 2011. Semiotic margins: Meaning in multimodali-
ties. London and New York: Continuum.
Dreyfus, Shoshana J., Sally Humphrey, Ahmar Mahboob and James R. Martin. 2016. Genre pedagogy in
higher education: The SLATE project. New York: Palgrave Macmillan.
Dreyfus, Shoshana J. and Sandra C. Jones. 2010. Construing sports starts: Appliable linguistics and the lan-
guage of the media. In Ahmar Mahboob and Naomi K. Knight (eds.), Appliable linguistics. London and
New York: Continuum. 114–129.
Driscoll, Jeni. 2012. The representation of terminally ill cancer patients: A transitivity analysis of advice and interview
texts. Ph.D. Thesis. University of Liverpool, Liverpool.
Drury, Helen and Janet Jones. 2010. Transforming the teaching of report writing in science and engi-
neering through an integrated online learning environment, WRiSE (Write reports in science and
engineering). In Caroline H. Steel, Mike J. Keppell, Philippa Gerbic and Simon Housego (eds.), Cur-
riculum, technology and transformation for an unknown future: Proceedings ascilite Sydney 2010. Brisbane: The
University of Queensland. 313–323.

470
 References

Dubreuil, Laurent and Sue Savage-Rumbaugh. 2019. Dialogues on the human ape. Minneapolis: University
of Minnesota Press.
Dunbar, Robin. 1996. Grooming, gossip, and the evolution of language. London: Faber and Faber.
Edmonson, Willis. 1981. Spoken discourse: A model for analysis. London: Longman.
Eggins, Suzanne. 1990. Conversational structure: A systemic-functional analysis of interpersonal and logical meaning
in multiparty sustained talk. Ph.D. thesis. The University of Sydney, Sydney.
Eggins, Suzanne. 1994. An introduction to systemic functional linguistics. London: Pinter.
Eggins, Suzanne. 2004. An introduction to systemic functional linguistics. 2nd edition. London and New York:
Continuum.
Eggins, Suzanne. 2014. Hospital humor: Patient-initiated humor as resistance to clinical discourse. In
Elke Stracke (ed.), Intersections: Applied linguistics as a meeting place. Newcastle: Cambridge Scholars.
43–66.
Eggins, Suzanne and J.R. Martin. 1997. Genres and registers of discourse. In Teun van Dijk (ed.), Dis-
course as structure and process: Discourse studies: A multidisciplinary introduction. Volume 1. London: Sage.
230–256.
Eggins, Suzanne and Diana Slade. 1997. Analysing casual conversation. London: Cassell.
Eggins, Suzanne and Diana Slade. 2005. Analysing casual conversation. London: Equinox.
Eggins, Suzanne and Diana Slade. 2016. Resource: Communicating effectively in bedside nursing hando-
ver. In Suzanne Eggins, Diana Slade and Fiona Geddes (eds.), Effective communication in clinical handover:
From research to practice. Berlin: De Gruyter. 115–125.
Eggins, Suzanne, Peter Wignell and J.R. Martin. 1993. The discourse of history: Distancing the recover-
able past. In Ghadessy (ed.), 75–109.
Ekman, Paul. 1992. An argument for basic emotions. Cognition and Emotion 6(3–4): 169–200.
Ekman, Paul, Wallace V. Friesen and Silvan S. Tomkins. 1971. Facial affect scoring technique: A first valid-
ity study. Semiotica 3: 37–58.
Ellegård, Alvar. 1953. The auxiliary ‘do’: The establishment and regulation of its use in English. Stockholm:
Almqvist och Wiksell.
Ellis, Jeffrey D. 1966. Towards a general comparative linguistics. The Hague: Mouton.
Ellis, Jeffrey D. 1987a. The logical and textual functions. In M.A.K. Halliday and Robin P. Fawcett (eds.),
New developments in systemic linguistics: Theory and description. London: Pinter. 107–130.
Ellis, Jeffrey D. 1987b. Some ‘dia-categories’. In Ross Steele and Terry Threadgold (eds.), Language topics:
Essays in honour of Michael Halliday. Volume 2. Amsterdam: Benjamins. 81–94.
El-Menoufy, Afaf. 1969. A study of the role of intonation in the grammar of English. Ph.D. thesis. University of
London, London.
El-Menoufy, Afaf. 1988. Intonation and meaning in spontaneous discourse. In James D. Benson, Michael J.
Cummings and William S. Greaves (eds.), Linguistics in a systemic perspective. Amsterdam: Benjamins. 1–27.
El-Rabbat, Amin Hussein. 1978. The major clause types of Egyptian colloquial Arabic: A participant-process
approach. Ph.D. thesis. University of London, London.
Espindola, Elaine. 2010. Illuminated the analysis of the translation is: Systemic functional linguistics strikes Yoda
back. Ph.D. thesis. Universidade Federal de Santa Catarina, Florianópolis.
Espindola, Elaine. 2012. Systemic functional linguistics and audiovisual translation studies: A conceptual
basis for the study of the language of subtitles. D.E.L.T.A. 28: Especial, 495–513.
Espindola, Elaine. 2016. A systemic functional translation analysis of thematic structure: Directing attention
to Yoda’s linguistic manifestation. Word 62(1): 22–34.
Espindola, Elaine and Maria Lúcia Vasconcellos. 2006. Two facets in the subtitling process: Foreignisation
and/or domestication procedures in unequal cultural encounters. Fragmentos 30: 43–066.
Espindola, Elaine and Yan Wang. 2015. The enactment of modality in regulatory texts: A comparative
study of tenancy agreements. Journal of World Languages 2(2–3): 106–125.
Evans, Nicholas. 2010. Dying words: Endangered languages and what they have to tell us. Oxford:
Wiley-Blackwell.
Evans, Nicholas and Stephen C. Levinson. 2009. The myth of language universals: Language diversity and
its importance for cognitive science. Behavioral and Brain Sciences 32: 429–492.
Fairclough, Norman. 1989. Language and power. London: Longman.
Fairclough, Norman. 1992. Discourse and social change. Cambridge: Polity Press.
Fairclough, Norman. 2003. Analysing discourse: Textual analysis for social research. London: Routledge.
Fairclough, Norman. 2010. Critical discourse analysis: The critical study of language. London and New York:
Routledge.

471
 References

Fang, Yan and Canzhong Wu (eds.). 2009. Challenges to systemic functional linguistics: Theory and practice.
Proceedings of the Conference, ISFC 36, 14–18 July. Beijing: Tsinghua University.
Faulkner, Adam Robert. 2014. Automated classification of argument stance in student essays: A linguistically moti-
vated approach with an application for supporting argument summarization. PhD thesis. The City University
of New York, New York.
Fawcett, Robin P. 1973a. Generating a sentence in systemic functional grammar. In M.A.K. Halliday and
James R. Martin (eds.), 1981, Readings in systemic linguistics. London: Batsford. 146–183.
Fawcett, Robin P. 1973b. Systemic functional grammar in a cognitive model of language. University College
London, mimeo.
Fawcett, Robin P. 1974–1976. Some proposals for systemic syntax 1–3. MALS Journal 1–2.
Fawcett, Robin P. 1980. Cognitive linguistics and social interaction. Exeter and Heidelberg: University of
Exeter and Julius Groos.
Fawcett, Robin P. 1987. The semantics of clause and verb for relational processes in English. In M.A.K.
Halliday and Robin P. Fawcett (eds.), New developments in systemic linguistics: Theory and description. Lon-
don: Pinter. 130–183.
Fawcett, Robin P. 2000. A theory of syntax for systemic functional linguistics. Amsterdam: Benjamins.
Fawcett, Robin P. 2008. Invitation to systemic functional linguistics: The Cardiff grammar as an extension and
simplification of Halliday’s systemic functional grammar. Sheffield: Equinox.
Fawcett, Robin P. 2011. Problems and solutions in identifying processes and participant roles in discourse
analysis, part 1: Introduction to a systematic procedure for identifying processes and participant roles.
Annual Review of Functional Linguistics 3: 34–87.
Fawcett, Robin P. 2013. Choice and choosing in systemic-functional grammar. In Lise Fontaine, Tom
Bartlett and Gerard O’Grady (eds.), 115–134.
Fawcett, Robin P. 2017. From meaning to form in the Cardiff Model of language and its use. In Tom
Bartlett and Gerard O’Grady (eds.), Chap. 5. 56–76.
Fawcett, Robin P. and Guowen Huang. 1997. Enhanced theme in English: Towards a functional explanation of
the ‘it-cleft’ construction. London: Cassell.
Fawcett, Robin P., Gordon Tucker and Lin Yuen. 1993. How a systemic-functional grammar works. In
Helmut Horacek and Michael Zock (eds.), New concepts in natural language: Planning realisation and sys-
tems. London: Pinter. 114–186.
Fawcett, Robin P., Anita van der Mije and Carla van Wissen. 1988. Towards a systemic flowchart model for
discourse structure. In Robin P. Fawcett and David Young (eds.), New developments in systemic linguistics,
Vol. 2: Theory and application. London: Pinter. 116–143.
Feez, Susan. 1998. Text-based syllabus design. Sydney: Macquarie University, NCELTR (National Centre for
English Language Teaching and Research).
Feez, Susan. 2010. Montessori and early childhood: A guide for students. London: SAGE.
Ferguson, Alison. 1992. Interpersonal aspects of aphasic conversation. Journal of Neurolinguistics 7(4):
277–294.
Ferguson, Alison and Elizabeth Armstrong. 2009. Researching communication disorders. New York: Palgrave
Macmillan.
Ferguson, Alison, Elizabeth Spencer and Elizabeth Armstrong. 2017. Systemic functional linguistics and
clinical linguistics. In Tom Bartlett and Gerard O’Grady (eds.), 491–505.
Fernandino, Leonardo, Colin J. Humphries, Lisa L. Conant, Mark S. Seidenberg and Jerry R. Binder.
2016. Heteromodal cortical areas encode sensory-motor features of word meaning. Journal of Neurosci-
ence 36(38): 9763–9769.
Ferreira, Alfredo A. and Sandra Zappa-Hollman. 2019. Disciplinary registers in a first-year program: A
view from the context of curriculum. Language, Context and Text 1(1): 148–193.
Fillmore, Charles J. 1968. The case for case. In Emmon Bach and Robert T. Harms (eds.), Universals in
linguistic theory. New York: Holt, Rinehart and Winston. 1–88.
Fillmore, Charles J. 1977. The case for case reopened. In Peter Cole and Jerrold M. Sadock (eds.), Syntax
and semantics, volume 8: Grammatical relations. New York: Academic Press.
Fillmore, Charles J. 1988. The mechanisms of ‘construction grammar’. Berkeley Linguistic Society 14: 35–55.
Fillmore, Charles J. 1999. Inversion and constructional inheritance. In Gert Webelhuth, Jean-Pierre Koe-
nig and Andreas Kathol (eds.), Lexical and constructional aspects of linguistic explanation. Stanford, CA: CSLI
Publications.
Fine, Jonathan. 1978. Conversation, cohesive and thematic patterning in children’s dialogues. Discourse
Processes 1: 247–266.

472
 References

Fine, Jonathan. 1985. What do surface markers mean? Towards a triangulation of social, cognitive and
linguistic factors. In James D. Benson and Williams S. Greaves (eds.), Systemic perspectives on discourse:
Selected applied papers. New York: Ablex.
Fine, Jonathan. 1988a. Cognitive processes in context: A systemic approach to problems in oral language
use. In Robert Veltman and Erich H. Steiner (eds.), Pragmatics, discourse and text: Some systemically-
inspired approaches. London: Pinter. 171–181.
Fine, Jonathan. 1988b. Second language discourse: A textbook of current research. Norwood, NJ: Ablex.
Fine, Jonathan. 1991. The static and dynamic choices of responding: Toward the process of building social
reality by the developmentally disordered. In Eija Ventola (ed.), Functional and systemic linguistics. The
Hague: Mouton DeGruyter. 213–234.
Fine, Jonathan. 1994. How language works: Cohesion in normal and nonstandard communication. Norwood, NJ:
Ablex.
Fine, Jonathan. 2006. Language in psychiatry: A handbook of clinical practice. London: Equinox Publishing.
Fine, Jonathan, Gampierro Bartolucci, Gary M. Ginsberg and Peter Szatmari. 1991. The use of into-
nation to communicate in pervasive developmental disorders. Journal of Child Psychiatry 32(5):
771–782.
Firbas, Jan. 1992. Functional sentence perspective in written and spoken communication. Cambridge: Cambridge
University Press.
Firth, J.R. 1948. Sounds and prosodies. Transactions of the Philological Society 127–152.
Firth, J.R. 1957a. Papers in linguistics 1934–1951. London: Oxford University Press.
Firth, J.R. 1957b. Ethnographic analysis and language with reference to Malinowski’s views. In R. Firth
(ed.), 93–118.
Firth, J.R. 1957c. A synopsis of linguistic theory, 1930–1955. In Studies in linguistic analysis (Special issue of
the Philological Society). London: Blackwell. 1–31. Reprinted in Firth, 1968, 168–205.
Firth, J.R. 1968. Selected papers of J. R. Firth 1952–59. F.R. Palmer (ed.). Bloomington, London and Har-
low: Longman.
Firth, Raymond (ed.). 1957d. Man and culture: An evaluation of the work of Bronislaw Malinowski. London:
Routledge and Kegan Paul.
Foley, Joseph. 1991. Vygotsky, Bernstein and Halliday: Towards a unified theory of L1 and L2 learning.
Language, Culture and Curriculum 4(1): 17–42.
Fontaine, Lise. 2013. Analysing English grammar: A systemic functional introduction. Cambridge: Cambridge
University Press.
Fontaine, Lise. 2017. The English nominal group: The centrality of the Thing element. In Tom Bartlett
and Gerard O’Grady (eds.), Ch. 17. 267–283.
Fontaine, Lise, Tom A.M. Bartlett and Gerard N. O’Grady (eds.). 2013. Systemic functional linguistics: Explor-
ing choice. Cambridge: Cambridge University Press.
Fontaine, Lise and David Schöntal. 2019. The rooms of the house: Grammar at group rank. In Geoff
Thompson, Wendy L. Bowcher, Lise Fontaine and David Schöntal (eds.), 118–141.
Forey, Gail. 2009. Marked interpersonal themes: Projecting clauses in workplace texts. In Gail Forey and
Geoff Thompson (eds.), Text type and texture: In honour of Flo Davies. London: Equinox. 151–174.
Forey, Gail and Jane Lockwood (eds.). 2010. Globalization, communication and the workplace: Talking across the
world. London: Continuum.
Forey, Gail and Nicholas Sampson. 2017. Textual metafunction and theme. In Tom Bartlett and Gerard
O’Grady (eds.), Chap. 9. 131–145.
Fowler, Roger. 1991. Language in the news: Discourse and ideology in the press. London: Routledge.
Fowler, Roger. 1995. The language of George Orwell. London: Macmillan.
Fowler, Roger, Bob Hodge, Gunther Kress and Tony Trew. 1979. Language and control. London: Routledge
and Kegan Paul.
Francis, Gill. 1989. Thematic selection and distribution in written discourse. Word 40(1–2): 201–223.
François, Jacques. 2018. The stance of systemic functional linguistics amongst functional(ist) theories of
language and its ‘systemic’ purpose. In Akila Baklouti and Lise Fontaine (eds.), 6–25.
French, Ruth. 2010. Starting points in teaching grammatics: Children learning about verbs. In T. Hays
(ed.), Bridging the gap between ideas and doing research. Proceedings of the 4th Annual Postgraduate Research
Conference, Faculty of the Professions, University of Armidale, Armidale NSW. Armidale: University of
New England. 79–106.
French, Ruth. 2013. Teaching and learning functional grammar in junior primary classrooms. Ph.D. thesis. Uni-
versity of New England, Armidale.

473
 References

Fries, Peter H. 1970. Tagmeme sequences in the English noun phrase. Norman: Summer Institute of Linguistics
of the University of Oklahoma.
Fries, Peter H. 1981. On the status of theme in English: Arguments from discourse. Forum Linguisticum
6(1): 1–38. Reprinted in Janos Petöfi and E. Sözer (eds.), Micro and macro connexity of texts. Hamburg:
Helmut Buske Verlag. 116–152.
Fries, Peter H. 1985. C.C. Fries’ view of language and linguistics. In Peter H. Fries and Nancy M. Fries
(eds.), Toward an understanding of language: Charles Carpenter Fries in perspective. Amsterdam: Benjamins.
63–83.
Fries, Peter H. 1986. Toward a discussion of the ordering of adjectives in the English noun phrase. In
Benjamin F. Elson (ed.), Language and global perspective: Papers in honor of the Summer Institute of Linguistics,
1935–1985. Dallas, TX: Summer Institute of Linguistics. 123–134.
Fries, Peter H. 1994. On theme, rheme and discourse goals. In Malcolm Coulthard (ed.), Advances in written
text analysis. London: Routledge and Kegan Paul. 229–249.
Fries, Peter H. 1995. Themes, methods of development and texts. In Ruqaiya Hasan and Peter Fries (eds.),
317–359.
Fries, Peter H. 2001. Systemic functional linguistics: A close relative of French functional linguistics? La
Linguistique 37: 89–100.
Fries, Peter H. 2002. Theme and new in written advertising. In Huang Guowen and Wang Zongyan (eds.),
Discourse and language functions. Shanghai: Foreign Language Teaching and Research Press. 56–72.
Fries, Peter H. and Nancy M. Fries (eds.). 1985. Toward an understanding of language: Charles Carpenter Fries
in perspective. Amsterdam: Benjamins.
Fries, Peter H. and Michael J. Gregory (eds.). 1996. Discourse in society: Systemic functional perspectives, mean-
ing and choice in language: Studies for Michael Halliday. Norwood, NJ: Ablex.
Fuller, Gillian. 1995. Engaging cultures: Negotiating discourse in popular science. Ph.D. thesis. Sydney University,
Sydney.
Fuller, Gillian. 1998. Cultivating science: Negotiating discourse in the popular texts of Stephen Jay Gould.
In J.R. Martin and Robert Veel (eds.), 35–62.
Fung, Andy. 2015. Hasan’s semantic networks revisited: A Cantonese systemic functional approach. In
Wendy Bowcher and Jennifer Yang (eds.), Society in language, language in society: Essays in honour of
Ruqaiya Hasan. Berlin: Springer. 115–140.
Fung, Andy and Francis Low. 2019. Semantic networks. In Geoff Thompson, Wendy L. Bowcher, Lise
Fontaine and David Schönthal (eds.), 333–357.
Fung Ka Chun, Andy. 2018. Analysing cantonese doctor-patient communication: A semantic network approach.
Ph.D. thesis. The Hong Kong Polytechnic University, Hong Kong.
Fuoli, Matteo, Jeannette Littlemore and Sarah Turner. 2021 Sunken Ships and Screaming Banshees: Metaphor
and evaluation in film reviews. English Language and Linguistics 26(1): 75–103.
García, Adolfo M. 2019. The neurocognition of translation and interpreting. Amsterdam: Benjamins.
García, Adolfo M., Daniel Franco-O’Byrne and Aqustín Ibáñez. 2020. Neurosemiotics: Blurbing a field
beyond the ‘two cultures divide’. In Ludmila Lacková, Claudio Rodríguez and Kalevi Kull (eds.), Gath-
erings in biosemiotics XX. Tartu: University of Tartu Press. 92–97.
García, Adolfo M. and Agustín Ibáñez. 2017. Processes and verbs of doing, in the brain:
Theoretical implications for systemic functional linguistics. Functions of Language 23(3): 305–335.
García, Adolfo M. and Agustín Ibáñez (eds.). forthcoming. The Routledge handbook of neurosemiotics. Lon-
don: Routledge.
García, Adolfo M., William Sullivan and Sarah Tsiang. 2017. An introduction to relational network theory: His-
tory, principles, and descriptive applications. London: Equinox.
Gardner, Sheena. 2010. SFL: A theory of language for dynamic assessment of EAL. NALDIC Quarterly
8(1): 37–41.
Gardner, Sheena. 2017. Systemic functional linguistics and genre studies. In Tom Bartlett and Gerard
O’Grady (eds.), Chap. 29. 473–488.
Gardner, Sheena and Hilary Nesi. 2013. A classification of genre families in university student writing.
Applied Linguistics 34(1): 25–52.
Garvin, Paul (ed.). 1964. A Prague School reader on esthetics, literary structure and style. Washington, DC:
Georgetown University Press.
Gazdar, Ewan Klein, Geoffrey K. Pullum and Ivan A. Sag. 1985. Generalized phrase structure grammar. Cam-
bridge, MA: Harvard University Press.
Gazdar, Gerald. 1979. Pragmatics: Implicature, presupposition, and logical form. New York: Academic Press.

474
 References

Gebhard, Meg. 2019. Teaching and researching ELLs’ disciplinary literacies: Systemic functional linguistics in action
in the context of U.S. school reform. London and New York: Routledge.
Geertz, Clifford. 1973. The interpretation of cultures: Selected essays. New York: Basic Books.
Gelbukh, Alexander and Hiram Calvo. 2018. Automatic syntactic analysis based on selectional preferences. Cham:
Springer.
Gell-Mann, Murray. 1994. The quark and the jaguar: Adventures in the simple and the complex. New York: W.H.
Freeman and Company.
Geluykens, R. 1990. Information structure in English conversation: The given-new distinction revisited.
Occasional Papers in Systemic Linguistics 3: 129–147.
Gengogaku Kenkyukai (ed.). 1983. Nihongo bunpo: Rengoron (data). (Japanese grammar: Collocation (data)).
Tokyo: Mugi Shobo.
Gerot, Linda and Peter Wignell. 1994. Making sense of functional grammar: An introductory workbook. Cam-
meray, NSW: Anti-podean Educational Enterprises.
Ghadessy, Mohsen (ed.). 1988. Registers of written English: Situational factors and linguistic features. London:
Pinter.
Ghadessy, Mohsen (ed.). 1993a. Register analysis: Theory and practice. London and New York: Pinter.
Ghadessy, Mohsen. 1993b. On the nature of written business communication. In Mohsen Ghadessy (ed.),
Register analysis: Theory and practice. London and New York: Pinter. 149–164.
Ghadessy, Mohsen (ed.). 1995. Thematic development in English text. London: Pinter.
Ghadessy, Mohsen (ed.). 1999. Text and context in functional linguistics. Amsterdam: Benjamins.
Ghadessy, Mohsen. 2003. Comments on Douglas Biber, Susan Conrad, Randi Reppen, Pat Byrd, and
Marie Helt’s Speaking and writing in the university: A multidimensional comparison: A reader reacts.
TESOL Quarterly 37(1): 147–150.
Ghadessy, Mohsen and Yanjie Gao. 2001. Small corpora and translation: Comparing thematic organization
in two languages. In Mohsen Ghadessy, Alex Henry and Robert L. Roseberry (eds.), Small corpus studies
and ELT. Amsterdam and Philadelphia: Benjamins. 335–362.
Ghio, Elsa and María D. Fernández. 2005. Manual de lingüística sistémico funcional: El enfoque de M.A.K. Hal-
liday y R. Hasan aplicaciones a lengua Española. Santa Fe: Universidad Nacional del Litoral.
Gibbons, John and Victoria Markwick-Smith. 1992. Exploring the use of a systemic semantic description.
International Journal of Applied Linguistics 2(1): 36–51.
Gibbons, Pauline. 2002. Scaffolding language, scaffolding learning: Teaching second language learners in the main-
stream classroom. Portsmouth, NH: Heinemann.
Gibbons, Pauline. 2007. Bridging discourses in the ESL classroom: Students, teachers, and researchers. London:
Continuum.
Gil, José María. 2013. A neurocognitive interpretation of systemic functional choice. In Lise Fontaine, Tom
Bartlett and Gerard O’Grady (eds.), Chap. 9. 179–204.
Givón, Talmy. 1979a. On understanding grammar. New York: Academic Press.
Givón, Talmy (ed.). 1979b. Syntax and semantics: Discourse and syntax. Volume 12. New York: Academic Press.
Givón, Talmy. 1982. Syntax: A functional typological introduction. Amsterdam: Benjamins.
Givón, Talmy. (ed.). 1983. Topic continuity in discourse. Amsterdam: Benjamins.
Gleason, H.A. 1965. Linguistics and English grammar. New York: Holt, Rinehart and Winston.
Glosser, Guila and Toni Deser. 1990. Patterns of discourse production among neurological patients with
fluent language disorders. Brain and Language 40: 67–88.
Goldberg, Adele E. 1995. Constructions: A construction grammar approach to argument structure. Chicago: The
University of Chicago Press.
Goldberg, Adele E. 2003. Constructions: A new theoretical approach to language. Trends in Cognitive Sci-
ences 7(5): 219–224.
Goldberg, Adele E. 2006. Constructions at work: The nature of generalizations in language. Oxford: Oxford
University Press.
Gómez-González, María de los Ángeles. 2001. The theme-topic interface: Evidence from English. Amsterdam
and New York: Banjamins.
Gonzálvez-García, Francisco and Christopher S. Butler. 2006. Mapping functional-cognitive space. Annual
Review of Cognitive Linguistics 4: 39–96.
Gonzálvez-García, Francisco and Christopher S. Butler. 2014. Exploring functional-cognitive space. Amster-
dam: Benjamins.
Goodenough, Ward H. 1956. Componential analysis and the study of meaning. Language 32(1):
195–216.

475
 References

Goodman, Ken, Peter H. Fries and Steven L. Strauss. 2016. Reading: The grand illusion: How and why people
make sense of print. London and New York: Routledge.
Gosden, Hugh. 1993. Discourse functions of subject in scientific research articles. Applied Linguistics 14(1):
56–75.
Gosden, Hugh. 1996. A genre-based investigation of theme: Product and process in scientific research articles written
by NNS novice researchers. Monographs in systemic linguistics, Number 7. University of Nottingham.
Grace, George W. 1981. An essay on language. Columbia, SC: Hornbeam Press.
Greaves, William S. 2007. Intonation in systemic linguistics. In Ruqaiya Hasan, Christian M.I.M. Mat-
thiessen and Jonathan J. Wester (eds.), 979–1025.
Gregory, Michael J. 1967. Aspects of varieties differentiation. Journal of Linguistics 3: 177–198.
Gregory, Michael J. 1982. The nature and use of metafunctions in systemic theory: Current concerns. In
The Eighth LACUS Forum. Columbia South C.: The Hornbeam Press.
Gregory, Michael J. 1984. Phasal analysis within communication linguistics: Two contrastive discourses.
Prepared for the Proceedings of the Second University Symposium on Linguistics and Semiotics: Text Semantics and
Discourse Semantics, from a Presentation to the Symposium, February 11.
Gregory, Michael J. 1985. Towards communication linguistics: A framework. In James D. Benson and Wil-
liam S. Greaves (eds.), Systemic perspectives on discourse. Norwood, NJ: Ablex. 119–134.
Gregory, Michael J. 2002. Phasal analysis within communication linguistics: Two contrastive discourses. In
Michael Cummings, Peter H. Fries and David Lockwood (eds.), Relations and functions within and around
language. London and New York: Continuum. 316–345.
Gregory, Michael J. and Susanne Carroll. 1978. Language and situation: Language varieties and their social con-
texts. London: Routledge and Kegan Paul.
Grimes, Joseph E. 1975. The thread of discourse. The Hague: Mouton.
Grimshaw, Allen D. (ed.). 1994. What’s going on here: Complementary studies of professional talk. Norwood,
NJ: Ablex.
Grishman, Ralph and Richard Kittredge (eds.). 1986. Analyzing language in restricted domains: Sublanguage
description and processing. Hillsdale, NJ: Lawrence Erlbaum Associates.
Gross, Maurice. 1979. On the failure of generative grammar. Language 55(4): 859–885.
Gu, Yueguo. 2002. Sampling situated discourse for spoken Chinese corpus. Available from https://cupdf.com/
document/compiling-a-spoken-chinese-corpus-of-situated-discourse-linguistics-although-con-
structed.html?page=1
Guerra Lyons, Jesús David. 2021. Making meaning throughout writing trajectories: Towards a social semiotic account
of language change in scholars’ theory construction. Ph.D. thesis. The Hong Kong Polytechnic University,
Hong Kong.
Gülich, Elisabeth and Wolfgang Raible. 1977. Linguistische textmodelle: Grundlagen und Möglichkeiten.
München: Wilhelm Fink Verlag.
Guo, Songdan Nancy. 2014. The ontogenesis of multiliteracy scaffolding in textbooks: Multimodal analysis of
English language teaching textbooks of different grade. Ph.D. thesis. The Hong Kong Polytechnic University,
Hong Kong.
Guo, Songdan Nancy and Yuan Yao. 2021. Language demands of textbooks for learning English in Hong
Kong: A multi-stratal analysis. Linguistics and Education 63.
Guo, Yiting Emily and Leanne Togher. 2008. The impact of dysarthria on everyday communication after
traumatic brain injury: A pilot study. Brain Injury 22(1): 83–97.
Gutwinski, Waldemar. 1976. Cohesion in literary texts: A study of some grammatical and lexical features of English
discourse. The Hague: Mouton.
Gwilliams, Laura and Lise Fontaine. 2015. Indeterminacy in process type classification. Functional Linguistics
2(8): 1–19.
Hajičová, Eva. 2019. Michael Alexander Kirkwood Halliday (1925–2018). Linguistica Pragensia 29(2): 243–245.
Hall, Edward T. 1959. The silent language. Garden City, NJ: Doubleday.
Hall, Edward T. 1966. The hidden dimension. New York and Toronto: Anchor Books.
Hall, Edward T. and George L. Trager. 1953. The analysis of culture. Washington, DC: Foreign Service
Institute, Department of State.
Hall, Phillip. 2004. Prone to distortion? Undue reliance on unreliable records in NSW Police Service for-
mal interview model. In John Gibbons, V. Prakasam, K.V. Tirumalesh and Nagrajan (eds.), Language in
the law. New Delhi: Orient Longman Private Limited. 44–81.
Halliday, M.A.K. 1956a. The linguistic basis of a mechanical thesaurus, and its application to English prep-
osition classification. Mechanical Translation 3(3): 81–88. Reprinted in M.A.K. Halliday, 2004b, 6–19.

476
 References

Halliday, M.A.K. 1956b. Grammatical categories in Modern Chinese. Transactions of the Philological Society
55(1): 177–224. Reprinted in M.A.K. Halliday, 2006, 209–248.
Halliday, M.A.K. 1957. Some aspects of systematic description and comparison in grammatical analysis.
In Studies in Linguistic Analysis (Special Volume of the Philological Society). Oxford: Basil Blackwell.
54–67. Reprinted in M.A.K. Halliday, 2002d, 21–36.
Halliday, M.A.K. 1959. The language of the Chinese secret history of the Mongols. Oxford: Blackwell. (Publica-
tions of the Philological Society 17.) Reprinted in M.A.K. Halliday, 2006, 3–171.
Halliday, M.A.K. 1961. Categories of the theory of grammar. Word 17(3): 242–292. Reprinted in Halliday,
M.A.K. 2002d, 37–94.
Halliday, M.A.K. 1963a. Intonation in English grammar. Transactions of the Philological Society 62(1): 143–169.
Halliday, M.A.K. 1963b. The tones of English. Archivum Linguisticum 15(1): 1–28.
Halliday, M.A.K. 1964a. Syntax and the consumer. In C.I.J.M. Stuart (ed.), Report of the Fifteenth Annual
(First International) Round Table Meeting on Linguistics and Language. Washington, D.C.: Georgetown
University Press. 11–24. Reprinted in M.A.K. Halliday and J.R. Martin, 1981, 21–28. Reprinted in
Halliday, M.A.K. 2003b, 36–49.
Halliday, M.A.K. 1964b. English system networks (mimeographed). Published in Halliday, 1976a.
Halliday, M.A.K. 1966 [1960]. General linguistics and its application to language teaching. In M.A.K. Hal-
liday and Angus McIntosh (eds.), Patterns of language: Papers in general, descriptive and applied linguistics.
London: Longman. Reprinted in M.A.K. Halliday, 2007a, 135–173.
Halliday, M.A.K. 1966a. The concept of rank: A reply. Journal of Linguistics 2(1): 110–118.
Halliday, M.A.K. 1966b. Lexis as a linguistic level. In Charles Ernest Bazell, John Cunnison Catford,
M.A.K. Halliday and Robert Henry Robins (eds.), 148–162. Reprinted in M.A.K. Halliday, 2002d,
158–172.
Halliday, M.A.K. 1966c. Some notes on ‘deep’ grammar. Journal of Linguistics 2(1): 57–67. Reprinted in
M.A.K. Halliday, 2002d, 106–117.
Halliday, M.A.K. 1967. Intonation and grammar in British English. The Hague: Mouton.
Halliday, M.A.K. 1968 [1967]. Notes on transitivity and theme in English 1–3. Journal of Linguistics 3(1):
37–81, 3(2): 199–244, 4(2): 179–215. Reprinted in M.A.K. Halliday, 2005a, 110–153.
Halliday, M.A.K. 1969. Options and functions in the English clause. Brno Studies in English 8: 81–88.
Reprinted in M.A.K. Halliday, 2005a, 154–163.
Halliday, M.A.K. 1970a. Functional diversity in language, as seen from a consideration of modality and
mood in English. Foundations of Language 6: 322–361. Reprinted in M.A.K. Halliday, 2005a, 164–204.
Halliday, M.A.K. 1970b. A course in spoken English: Intonation. London: Oxford University Press.
Halliday, M.A.K. 1970c. Phonological (prosodic) analysis of the new Chinese syllable (modern Pekingese).
In Frank R. Palmer (ed.), Prosodic analysis. London: Oxford University Press.
Halliday, M.A.K. 1970d. Language structure and language function. In John Lyons (ed.), New horizons in
linguistics. Harmondsworth: Penguin Books. 140–165.
Halliday, M.A.K. 1971. Linguistic function and literary style: An enquiry into the language of William
Golding’s ‘the inheritors’. In Seymor Chatman (ed.), Literary style: A symposium. New York: Oxford
University Press. 330–368. Reprinted in M.A.K. Halliday, 2002e, 88–125.
Halliday, M.A.K. 1972. Towards a sociological semantics. Working papers and prepublications (series C, no.
14). Urbino: Università di Urbino, Centro Internazionale di Semiotica e di Linguistica. Reprinted
in M.A.K. Halliday, 1973, Explorations in the functions of language. London: Edward Arnold. 64–94.
Reprinted in M.A.K. Halliday, 2003b, 323–354.
Halliday, M.A.K. 1973. Explorations in the functions of language. London: Edward Arnold.
Halliday, M.A.K. 1974a. The place of ‘functional sentence perspective’ in the system of linguistic descrip-
tion. In František Daneš (ed.), Papers on functional sentence perspective. Prague: Academia. 43–53.
Halliday, M.A.K. 1974b. Discussion. In Herman Parret (ed.), Discussing language. The Hague: Mouton.
81–120.
Halliday, M.A.K. 1975a. Learning how to mean: Explorations in the development of language. London: Edward
Arnold.
Halliday, M.A.K. 1975b. Some aspect of sociolinguistics. In Interactions between linguistics and mathemati-
cal education, report on a symposium sponsored by UNESCO-CEDO-ICMI, Nairobi, September 1974,
ED-74/CONF. 808. 64–73. Reprinted as Aspects of sociolinguistic research in M.A.K. Halliday,
2007b, 203–222.
Halliday, M.A.K. 1975c. The context of linguistics. In Francis P. Dinneen (ed.), Report of the twenty-fifth
annual round table meeting on linguistics and language study. Washington, DC: Georgetown University Press.

477
 References

Halliday, M.A.K. 1976a. System and function in language, edited by Gunther Kress. London: Oxford Uni-
versity Press.
Halliday, M.A.K. 1976b. ‘The teacher taught the student English’: An essay in applied linguistics. In Peter
A. Reich (ed.), The second LACUS forum. Columbia: Hornbeam Press. 344–349. Reprinted in M.A.K.
Halliday, 2005a, 297–305.
Halliday, M.A.K. 1976c. Anti-languages. American Anthropologist 78 (3): 570–584.
Halliday, M.A.K. 1977a. Ideas about language. In M.A.K. Halliday (ed.), Aims and perspectives in linguistics,
no 1. Brisbane: Applied Linguistics Association of Australia. 32–49. Reprinted in M.A.K. Halliday,
2003b, 92–115.
Halliday, M.A.K. 1977b. Text as semantic choice in social contexts. In Teun van Dijk and Janos Petöfi
(eds.), Grammars and descriptions. Berlin: Walter de Gruyter. 176–225. Reprinted in M.A.K. Halliday,
2002e, 23–81.
Halliday, M.A.K. 1977c. Some thoughts on language in the middle school years. English in Australia 42:
3–16. Reprinted in M.A.K. Halliday, 2007a, 49–62.
Halliday, M.A.K. 1978. Language as social semiotic: The social interpretation of language and meaning. London:
Edward Arnold.
Halliday, M.A.K. 1979a. Modes of meaning and modes of expression: Types of grammatical structure and
their determination by different semantic functions. In David J. Allerton, Edward Carney and David
Holdcroft (eds.), Function and context in linguistic analysis: A festschrift for William Haas. Cambridge: Cam-
bridge University Press. 57–79. Reprinted in M.A.K. Halliday, 2002d, 196–218.
Halliday, M.A.K. 1979b. Some notes on ‘deep’ grammar. Journal of Linguistics 2(1): 57–67. Reprinted in
M.A.K. Halliday, 2002d, 106–117.
Halliday, M.A.K. 1980. On being teaching. In Sidney Greenbaum et al. (eds.), Studies in English linguistics:
For Randolph Quirk. London: Longman. 61–64.
Halliday, M.A.K. 1981. Text semantics and clause grammar: Some patterns of realization. seventh LACUS
forum. Columbia: Hornbeam Press. 31–59. Reprinted as Text semantics and clause grammar: How is a
text like a clause? in M.A.K. Halliday, 2002d, 219–260.
Halliday, M.A.K. 1981 [1965]. Types of structure. In M.A.K. Halliday and J.R. Martin (eds.), Working paper
for the O.S.T.I. programme in the linguistic properties of scientific English. 29–41.
Halliday, M.A.K. 1982. The de-automatization of grammar: From Priestley’s ‘an inspector calls’. In John
M. Anderson (ed.), Language form and linguistic variation: Papers dedicated to Angus McIntosh. Amsterdam:
Benjamins. 129–159. Reprinted in M.A.K. Halliday, 2002e, 126–148.
Halliday, M.A.K. 1984a. On the ineffability of grammatical categories. In Alan Manning, Pierre Martin
and Kim McCalla (eds.), Tenth LACUS forum. Columbia: Hornbeam Press. 3–18. Reprinted in M.A.K.
Halliday, 2002d, 291–322.
Halliday, M.A.K. 1984b. Language as code and language as behaviour: A systemic-functional interpreta-
tion of the nature and ontogenesis of dialogue. In M.A.K. Halliday, Robin P. Fawcett, Sydney Lamb
and Adam Makkai (eds.), The semiotics of language and culture. Volume 1. London: Frances Pinter. 3–35.
Reprinted in Halliday, M.A.K. 2003b, 226–250.
Halliday, M.A.K. 1984c. Listening to Nigel. Sydney University Linguistics Department. Mimeo. Published
as part of M.A.K. Halliday, 2003a.
Halliday, M.A.K. 1984d. Linguistics in the university: The question of social accountability. In James E.
Copeland (ed.), New directions in linguistics and semiotics. Houston, Texas: Rice University Studies. 51–67.
Halliday, M.A.K. 1985a. An introduction to functional grammar. London: Edward Arnold.
Halliday, M.A.K. 1985b. Dimensions of discourse analysis: Grammar. In Teun A. van Dijk (ed.), Hand-
book of discourse analysis. Volume 2. New York: Academic Press. 29–56. Reprinted in M.A.K. Halliday,
2002d, 261–286.
Halliday, M.A.K. 1985c. Systemic background. In James D. Benson and William S. Greaves (eds.), Systemic
perspectives on discourse. Norwood, NJ: Ablex. 1–15. Reprinted in M.A.K. Halliday, 2003b, 185–198.
Halliday, M.A.K. 1985d. The relationship of the text and its context of situation. In M.A.K. Halliday and
Ruqaiya Hasan (eds.), Language, context and text: Aspects of language in a social-semiotic perspective. Geelong,
VIC: Deakin University Press.
Halliday, M.A.K. 1985e. English intonation as a resource for discourse. Beiträge zur Phonetik und Linguistik
48. Festschrift in Honour of Arthur Delbridge: 111–117.
Halliday, M.A.K. 1985f. Spoken and written language. Geelong, VIC: Deakin University Press.
Halliday, M.A.K. 1985g. It’s a fixed word order language is English. ITL Review of Applied Linguistics 67–68:
91–116. Reprinted in M.A.K. Halliday, 2005a, 213–231.

478
 References

Halliday, M.A.K. 1987a. Spoken and written modes of meaning. Rosalind Horowitz and S. Jay Samuels
(eds.), Comprehending oral and written language. New York: Academic Press. 55–82.
Halliday, M.A.K. 1987b. Language and the order of nature. Nigel Fabb, Derek Attridge, Alan Durant and
Colin MacCabe (eds.), The linguistics of writing: Arguments between language and literature. Manchester:
Manchester University Press. 135–154. Reprinted in M.A.K. Halliday, 2003b, 116–138.
Halliday, M.A.K. 1988a. On the language of physical science. In Mohsen Ghadessy (ed.), Registers of writ-
ten English: Situational factors and linguistic features. London and New York: Pinter Publishers. 162–178.
Reprinted in M.A.K. Halliday, 2004a, 140–158.
Halliday, M.A.K. 1988b. Poetry as scientific discourse: The nuclear sections of Tennyson’s ‘in memoriam’.
In David Birch and Michael O’Toole (eds.), Functions of style. London: Pinter. 31–44. Reprinted in
M.A.K. Halliday, 2002e, 149–167.
Halliday, M.A.K. 1988c. Foreword. In David Birch and Michael O’Toole (eds.), Functions of style. London:
Pinter. vi–vii.
Halliday, M.A.K. 1988d. Some basic concepts of educational linguistics. In Verner Bickley (ed.), Proceed-
ings of the ILE International Seminar on Languages in Education in a Bilingual or Multi-lingual Setting. Hong
Kong: The Institute of Language in Education. Reprinted in M.A.K. Halliday, 2007a, 341–353.
Halliday, M.A.K. 1990a. On the concept of ‘educational linguistics’. In Rod Giblett and John O’Carroll
(eds.), Discipline-dialogue-difference: Proceedings of the language in education conference, December 1989,
Murdoch University. Murdoch: 4D Duration Publications. Reprinted in M.A.K. Halliday, 2007a,
354–367.
Halliday, M.A.K. 1990b. New ways of meaning: A challenge to applied linguistics: Greek Applied Linguis-
tics Association. Journal of Applied Linguistics 6 (Ninth World Congress of Applied Linguistics Special
Issue): 7–36. Reprinted in M.A.K. Halliday, 2003b, 139–174.
Halliday, M.A.K. 1990c. The construction of knowledge and value in the grammar of scientific discourse:
With reference to Charles Darwin’s the origin of species. In Clotilde de Stasiom Maurizio Gotti and Ros-
sana Bonadei (eds.), La rappresentazione verbale e iconica: Valori estetici e funzionali. Bergamo: Atti del XI
Congresso Nazionale dell’A.I.A., 24–25 October 1988. Guerini Studio, 57–80. Reprinted in M.A.K.
Halliday, 2002e, Linguistic studies of text and discourse, 168–192.
Halliday, M.A.K. 1991a. Linguistic perspectives on literacy: A systemic-functional approach. In Frances
Christie (ed.), Literacy in social processes: Papers from the inaugural Australian systemic linguistics conference,
Deakin University, January 1990. Darwin: Centre for Studies of Language in Education, Northern Ter-
ritory University. Revised version in M.A.K. Halliday, 2007a, 97–129.
Halliday, M.A.K. 1991b. Towards probabilistic interpretations. In Eija Ventola (ed.), Trends in linguistics stud-
ies and monographs 55: Functional and systemic linguistics approaches and uses. Berlin: Mouton de Gruyter.
39–61.
Halliday, M.A.K. 1991c. Corpus linguistics and probabilistic grammar. In Karin Aijmer and Bengt
Altenberg (eds.), English corpus linguistics: Studies in honour of Jan Svartvik. London: Longman. 30–43.
Reprinted in M.A.K. Halliday, 2004b, 63–75.
Halliday, M.A.K. 1991d. The notion of ‘context’ in language education. In Thao Le and Mike McCaus-
land (eds.), Interaction and development: Proceedings of the international conference, Vietnam, 30 March–1 April
1991. Tasmania: University of Tasmania, Language Education. 1–26. Reprinted in M.A.K. Halliday,
2007a, 269–290.
Halliday, M.A.K. 1992a. Some lexicogrammatical features of the zero population growth text. In Sandra
A. Thompson and William C. Mann (eds.), Discourse description: Diverse analyses of a fund-raising text.
Amsterdam: Benjamins. 327–358.
Halliday, M.A.K. 1992b. The history of a sentence: An essay in social semiotics. In Vita Fortunait (ed.),
La cultura italiana e le letterature straniere moderne. Bologna: Longo Editore [for University of Bologna].
29–45. Reprinted in M.A.K. Halliday, 2003b, 355–374.
Halliday, M.A.K. 1992c. How do you mean? Martin Davies and Louise Ravelli (eds.), Advances in systemic
linguistics: Recent theory and practice. London: Pinter. 20–35. Reprinted in M.A.K. Halliday, 2002d,
352–368.
Halliday, M.A.K. 1992d. A systemic interpretation of Peking syllable finals. In Paul Tench (ed.), Studies in
systemic phonology. London: Pinter. 98–121.
Halliday, M.A.K. 1993a. Systemic grammar and the concept of a ‘science of language’. In Zhu Yongsheng
(ed.), Language, text, context. Beijing: Tsinghua University Press. 1–22.
Halliday, M.A.K. 1993b. Towards a language-based theory of learning. Linguistics and Education 5(2):
93–116. Reprinted in M.A.K. Halliday, 2003a, 327–352.

479
 References

Halliday, M.A.K. 1993c. Quantitative studies and probabilities in grammar. In Michael Hoey (ed.), Data,
description, discourse: Papers on the English language in honour of John McH. Sinclair. London: Harper Col-
lins. 1–25. Reprinted in M.A.K. Halliday, 2004b, 130–156.
Halliday, M.A.K. 1993d. The analysis of scientific texts and Chinese. In M.A.K. Halliday and J.R. Martin
(eds.), 137–146. Reprinted in Halliday, 2006, 334–45.
Halliday, M.A.K. 1994a. Systemic theory. In R.E. Asher (ed.), The encyclopedia of language and linguistics.
Volume 8. Oxford: Pergamon Press. 4505–4508. Reprinted as Appendix in M.A.K. Halliday, 2003b,
433–441.
Halliday, M.A.K. 1994b. So you say ‘pass’ . . . thank you three muchly. In Allen D. Grimshaw (ed.), What’s
going on here: Complementary studies of professional talk. Norwood, NJ: Ablex. 175–229.
Halliday, M.A.K. 1994c. An introduction to functional grammar. 2nd edition. London: Edward Arnold.
Halliday, M.A.K. 1994d. Language and the theory of codes. In Alan Sadovnik (ed.), Knowledge and peda-
gogy: The sociology of Basil Bernstein. Norwood, NJ: Ablex. 124–142. Reprinted in MA.K. Halliday,
2007b, 231–246.
Halliday, M.A.K. 1995. On language in relation to the evolution of human consciousness. In Sture Allén
(ed.), Of thoughts and words: Proceedings of Nobel Symposium 92. The relation between language and
mind, Stockholm, 8–12 August 1994. Singapore: River Edge N.J. and London: Imperial College Press.
45–84. Reprinted in M.A.K. Halliday, 2003b, 390–432.
Halliday, M.A.K. 1996a. On grammar and grammatics. In Ruqaiya Hasan, Carmel Cloran and David Butt
(eds.), Functional descriptions: Theory into practice. Amsterdam: Benjamins. 1–38. Reprinted in M.A.K.
Halliday, 2002d, 384–417.
Halliday, M.A.K. 1996b. Literacy and linguistics: A functional perspective. In Ruqaiya Hasan and Geoff
Williams (eds.), Literacy in Society. London: Longman. 339–376. Reprinted in M.A.K. Halliday, 2007a,
97–129.
Halliday, M.A.K. 1997. Linguistics as metaphor. In Kristin Davidse, Dirk Noel and Anne-Marie Simon-
Vandenbergen (eds.), Reconnecting language: Morphology and syntax in functional perspectives. Amsterdam:
Benjamins. 3–27. Reprinted in M.A.K. Halliday, 2003b, 248–270.
Halliday, M.A.K. 1998a. On the grammar of pain. Functions of Language 5(1): 1–32. Reprinted in M.A.K.
Halliday, 2005a, 306–337.
Halliday, M.A.K. 1998b. Grammar and daily life: Concurrence and complementarity. In Peter H. Fries
and David G. Lockwood James E. Copeland (eds.), Functional approaches to language, culture and cognition
(Festschrift for Sydney Lamb). Amsterdam and Philadelphia: Benjamins. Reprinted in M. A. K. Hal-
liday, 2002d, 369–383.
Halliday, M.A.K. 1998c. Things and relations: Regrammaticizing experience as technical knowledge. In
J.R. Martin and Robert Veel (eds.), 49–101.
Halliday, M.A.K. 2000. Phonology past and present: A personal retrospect. Folia Linguistica 34(1–2):
101–111.
Halliday, M.A.K. 2001a. Is the grammar neutral? Is the grammarian neutral? In Jessica de Villiers and
Robert J. Stainton (eds.), Communication in linguistics, vol. 1: Papers in honour of Michael Gregory. Toronto:
Editions du Gref. Reprinted in M.A.K. Halliday, 2003b, 271–294.
Halliday, M.A.K. 2001b. On the grammatical foundations of discourse. In Ren Shaozeng, William Guth-
rie and Ronald I.W. Fong (eds.), Grammar and discourse: Proceedings of the international conference on dis-
course analysis. University of Macau (in conjunction with Tsinghua University, China), 16–18 October
1997. Macau: University of Macau Publication Centre. 47–58. Reprinted in M.A.K. Halliday, 2006,
346–367.
Halliday, M.A.K. 2002a. Computing meanings: Some reflections on past experience and present prospects.
In Guowen Huang and Zongyan Wang (eds.), Discourse and language functions. Shanghai: Foreign Lan-
guage Teaching and Research Press. 3–25. Reprinted in M.A.K. Halliday, 2004b, 239–267.
Halliday, M.A.K. 2002b. Applied linguistics as an evolving theme. Presented at AILA 2002, Singapore.
Published in M.A.K. Halliday, 2007a, 1–19.
Halliday, M.A.K. 2002c. The spoken language corpus: A foundation for grammatical theory. In Karin
Aijmer and Bengt Altenberg (eds.), Proceedings of ICAME 2002: The theory and use of corpora, Göteborg,
22–26 May. Amsterdam: Editions Rodopi. Reprinted in M.A.K. Halliday, 2004b, 157–189.
Halliday, M.A.K. 2002d. On grammar. Volume 1 of the collected works of M.A.K. Halliday, edited by
Jonathan J. Webster. London and New York: Continuum.
Halliday, M.A.K. 2002e. Linguistic studies of text and discourse. Volume 2 of the collected works of M.A.K.
Halliday, edited by Jonathan J. Webster. London and New York: Continuum.

480
 References

Halliday, M.A.K. 2003a. The language of early childhood. Volume 4 of the collected works of M.A.K. Halliday,
edited by Jonathan J. Webster. London and New York: Continuum.
Halliday, M.A.K. 2003b. On language and linguistics. Volume 3 of the collected works of M.A.K. Halliday,
edited by Jonathan J. Webster. London and New York: Continuum.
Halliday, M.A.K. 2003c. Introduction: On the architecture of human language. In M.A.K. Halliday, 2003b,
1–29.
Halliday, M.A.K. 2004. The language of science. Volume 5 in the collected works of M.A.K. Halliday, edited
by Jonathan J. Webster. London and New York: Continuum.
Halliday, M.A.K. 2005a. Studies in English language. Volume 7 of the collected works of M.A.K. Halliday,
edited by Jonathan J. Wester. London and New York: Continuum.
Halliday, M.A.K. 2005b. On matter and meaning: The two realms of human experience. Linguistics and the
Human Sciences 1(1): 59–82.
Halliday, M.A.K. 2005c. Computational and quantitative studies. Volume 6 of the collected works of M.A.K.
Halliday, edited by Jonathan J. Webster. London and New York: Continuum.
Halliday, M.A.K. 2005d. A note on systemic functional linguistics and the study of language disorders.
Clinical linguistics and phonetics 19(3): 133–135.
Halliday, M.A.K. 2006. Studies in Chinese language. Volume 8 of the collected works of M.A.K. Halliday,
edited by Jonathan J. Webster. London and New York: Continuum.
Halliday, M.A.K. 2007a. Language and education. Volume 9 of the collected works of M.A.K. Halliday,
edited by Jonathan J. Webster. London and New York: Continuum.
Halliday, M.A.K. 2007b. Language and society. Volume 10 of the collected works of M.A.K. Halliday, edited
by Jonathan J. Webster. London and New York: Continuum.
Halliday, M.A.K. 2008a. Working with meaning: Towards an appliable linguistics. In Jonathan J. Webster
(ed.), Meaning in context: Implementing intelligent applications of language studies. London and New York:
Continuum. 7–23.
Halliday, M.A.K. 2008b. Complementarities in language. (Halliday Centre series in Appliable Linguistics).
Beijing: The Commercial Press.
Halliday, M.A.K. 2009a. Methods-techniques-problems. In M.A.K. Halliday and Jonathan J. Wester (eds.),
Chap. 3. 59–86.
Halliday, M.A.K. 2009b. Keywords. In M.A.K. Halliday and Jonathan J. Wester (eds.), Chap. 13. 229–253.
Halliday, M.A.K. 2010. Language evolving: Some systemic functional reflections on the history of mean-
ing. Manuscript of plenary given at ISFC 37, UBC Vancouver, Canada, July. Published in M.A.K.
Halliday, 2013a, 237–254.
Halliday, M.A.K. 2011. On text and discourse, information and meaning. Choice and text group, Institute
of Language and Communication, University of Southern Denmark, Odense, Denmark. Reprinted in
M.A.K. Halliday, 2013a, 237–254.
Halliday, M.A.K. 2012. Pinpointing the choice: Meaning and the search for equivalents in a translated text.
In M.A.K. Halliday (ed.), 2013a, 143–154.
Halliday, M.A.K. 2013a. Halliday in the 21st century. Volume 11 of the collected works of M.A.K. Halliday,
edited by Jonathan J. Webster. London: Bloomsbury Academic.
Halliday, M.A.K. 2013b. Meaning as choice. In Gerald N. O’Grady, Tom A.M. Bartlett and Lise M. Fon-
taine (eds.), Chap. 1. 15–36.
Halliday, M.A.K. 2014. On explanation in systemic functional theory. Linguistics and Human Sciences 10(1):
13–27.
Halliday, M.A.K. 2015. The influence of Marxism. In Jonathan J. Webster (ed.), The Bloomsbury companion
to M.A.K. Halliday. London: Bloomsbury Academic. 94–100.
Halliday, M.A.K. 2016. Aspects of language and learning, edited by Jonathan J. Webster. Berlin and Heidel-
berg: Springer.
Halliday, M.A.K. 2017. Contributions of linguistics in China to the development of SFL. In Jonathan J.
Webster and Xuanwei Peng (eds.), Applying systemic functional linguistics: The state of the art in china today.
London: Bloomsbury Academic. 3–13.
Halliday, M.A.K. and William S. Greaves. 2008. Intonation in the grammar of English. London: Equinox.
Halliday, M.A.K. and Ruqaiya Hasan. 1976. Cohesion in English. London: Longman.
Halliday, M.A.K. and Ruqaiya Hasan. 1980. Text and context: Aspects of language in a social semiotic perspective.
Sophia Linguistica VI. Tokyo: Sophia University Press.
Halliday, M.A.K. and Ruqaiya Hasan. 1985. Language, context and text: A social semiotic perspective. Geelong,
VIC: Deakin University Press.

481
 References

Halliday, M.A.K and Zoe L. James. 1993. A quantitative study of polarity and primary tense in the English
finite clause. John M. Sinclair, Michael Hoey and Gwyneth Fox (eds.), Techniques of description: Spo-
ken and written discourse (a festschrift for Malcolm Coulthard). London and New York: Routledge. 32–66.
Reprinted in Halliday, M.A.K., 2004b, 93–129.
Halliday, M.A.K and J.R. Martin (eds.). 1981. Readings in systemic linguistics. London: Batsford.
Halliday, M.A.K. and J.R. Martin. 1993. Writing science: Literacy and discursive power. London: Falmer.
Halliday, M.A.K. and Christian M.I.M. Matthiessen. 2004. Halliday’s introduction to functional grammar. 3rd
edition. London and New York: Routledge.
Halliday, M.A.K. and Christian M.I.M. Matthiessen. 2006. Construing experience through meaning: A
language-based approach to cognition. London and New York: Continuum. [Originally published in
1999].
Halliday, M.A.K. and Christian M.I.M. Matthiessen. 2014. Halliday’s introduction to functional grammar. 4th
edition. London and New York: Routledge.
Halliday, M.A.K. and Edward McDonald. 2004. Metafunctional profile of Chinese. In Alice Caffarel, J.R.
Martin and Christian Matthiessen (eds.), 305–396.
Halliday, M.A.K., Angus McIntosh and Peter Strevens. 1964. The linguistic sciences and language teaching.
London: Longman.
Halliday, M.A.K. and Guenter A. Plum. 1985. On casual conversation. In Ruqaiya Hasan (ed.), Discourse
on discourse. Wollongong, NSW: ALAA.
Halliday, M.A.K. and Jonathan J. Wester (eds.). 2009. Continuum companion to systemic functional linguistics.
London and New York: Continuum.
Hammond, Jenny. 2011. Classroom discourse. In Ken Hyland and Brian Paltridge (eds.), Continuum com-
panion to discourse analysis. London: Continuum. 291–305.
Han, Hee Jeung and David Kellog. 2019. A story without SELF: Vygotsky’s pedology, Bruner’s construc-
tivism and Halliday’s construalism in understanding narratives by Korean children. Language and Educa-
tion 33(6): 506–520.
Hansen, Silvia. 2003. The nature of translated text: An interdisciplinary methodology for the investigation of the
specific properties of translations. Saarbrücken: DFKI/Universität des Saarlandes.
Hansen, Silvia, Mary Klaumann and Stella Neumann. 2002. How to overcome registerial translation prob-
lems: A corpus-based approach. Revista Brasileira de Lingüística Aplicada 2(2): 15–23.
Hansen, Silvia and Elke Teich. 1999. Kontrastive Analyse von Übersetzungs-korpora: ein funktionales
Modell. In Hrsg J. Gippert (ed.), Sammelband der Jahrestagung der GLDV 99. Prag: Enigma Corporation.
311–322.
Hansen-Schirra, Silvia and Stella Neumann. 2003. The challenge of working with multilingual corpora.
In Stella Neumann and Silvia Hansen-Schirra (eds.), Proceedings of the workshop on multilingual corpora:
Linguistic requirements and technical perspectives. Corpus Linguistics, Lancaster, UK, 27 March. Saarland:
Saarland University.
Hansen-Schirra, Silvia, Stella Neumann and Erich Steiner. 2006. Cohesion and explicitation in an English-
German translation corpus. Pre-proceedings of the SPRIK Conference 2006: Explicit and Implicit Information
in Text: Information Structure across Languages, Oslo, Norway, 8–10 June. 45–50.
Hansen-Schirra, Silvia, Stella Neumann and Erich Steiner. 2012. Cross-linguistic corpora for the study of trans-
lations: Insights from the language pair English-German. Berlin: de Gruyter.
Hao, Jing. 2020. Analysing scientific discourse from a systemic functional linguistic perspective a framework for explor-
ing knowledge-building in biology. New York and London: Routledge.
Harman, Chris. 1999. A people’s history of the world: From the stone age to the new millennium. London and
New York: Verso.
Harman, Ruth (ed.). 2018. Bilingual learners and social equity: Critical approaches to systemic functional linguistics.
Berlin: Springer.
Harris, Zellig S. 1952. Discourse analysis. Language 28(1): 1–30.
Harrison, K. David. 2007. When languages die: The extinction of the world’s languages and the erosion of human
knowledge. Oxford: Oxford University Press.
Hartley, Anthony and Cécile Paris. 1997. Multilingual document production from support for translating
to support for authoring. New Tools for Human Translators 12(1/2): 109–129.
Hartley, Leila L. and Paul Jensen. 1991. Narrative and procedural discourse after closed head injury. Brain
Injury 5(3): 267–285.
Hasan, Ruqaiya. 1964. A linguistic study of contrasting features in the style of two contemporary English prose writers.
Ph.D. thesis. University of Edinburgh, Edinburgh.

482
 References

Hasan, Ruqaiya. 1968. Grammatical cohesion in spoken and written English, part II: Nuffield programme in lin-
guistics and English teaching. Mimeo.
Hasan, Ruqaiya. 1973. Code, register and social dialect. In Basil Bernstein (ed.), Class, codes and con-
trol: Applied studies towards a sociology of language. Volume 2. London: Routledge and Kegan Paul.
253–292.
Hasan, Ruqaiya. 1975. The place of stylistics in the study of verbal art. In Hakan Ringbom (ed.), Style and
text. Amsterdam: Skriptor.
Hasan, Ruqaiya. 1978. Text in the systemic-functional model. In Wolfgang Dressler (ed.), Current trends in
text linguistics. Berlin: de Gruyter. 228–246.
Hasan, Ruqaiya. 1979. On the notion of text. In Janos Petöfi (ed.), Text versus sentence: Basic questions of text
linguistics. Papers in Text Linguistics 20. Hamburg: Buske. 369–390.
Hasan, Ruqaiya. 1984a. The nursery tale as a genre. Nottingham Linguistic Circular 13. Reprinted in Ruqaiya
Hasan, 1996, Ways of saying: Ways of meaning: Selected papers of Ruqaiya Hasan, edited by Cloran, Carmel,
David Butt and Geoffrey Williams. London: Cassell. 51–72.
Hasan, Ruqaiya. 1984b. Coherence and cohesive harmony. In James Flood (ed.), Understanding reading
comprehension. Newark: International Reading Association. 181–219.
Hasan, Ruqaiya. 1985a. Meaning, context and text: Fifty years after Malinowski. In James D. Benson and
William S. Greaves (eds.), Systemic perspectives on discourse. Norwood, NJ: Ablex. 16–50.
Hasan, Ruqaiya. 1985b. Linguistics, language and verbal art. Geelong, VIC: Deakin University Press.
Hasan, Ruqaiya. 1985c. Lending and borrowing: From grammar to lexis. Beiträge zur Phonetik und Linguis-
tik 48: 56–67.
Hasan, Ruqaiya. 1985d. Offers in the making: A systemic functional approach. Mimeo.
Hasan, Ruqaiya (ed.). 1985e. Discourse on discourse. Wollongong, NSW: ALAA.
Hasan, Ruqaiya. 1986. The ontogenesis of ideology: An interpretation of mother-child talk. In Elizabeth
A. Grosz, Terry Threadgold and M.A.K. Halliday (eds.), Semiotics-ideology-language. Sydney: Sydney
Association for Studies in Society and Culture. 125–146.
Hasan, Ruqaiya. 1987. The grammarian’s dream: Lexis as most delicate grammar. In M.A.K. Halliday and
Robin P. Fawcett (eds.), New developments in systemic linguistics: Theory and description. London: Pinter.
184–211.
Hasan, Ruqaiya. 1988. The analysis of one poem: Theoretical issues in practice. In Birch and O’Toole
(eds.), 45–73.
Hasan, Ruqaiya. 1989. Semantic variation and sociolinguistics. Australian Journal of Linguistics 9: 221–275.
Hasan, Ruqaiya. 1992a. Meaning in sociolinguistic theory. In Kingsley Bolton and Helen Kwok (eds.),
Sociolinguistics today: International perspectives. London: Routledge. 80–119.
Hasan, Ruqaiya. 1992b. Rationality in everyday talk: From practice to theory. In Jan Svartvik (ed.), Direc-
tions in corpus linguistics. Berlin: Mouton de Gruyter.
Hasan, Ruqaiya. 1992c. Speech genre, semiotic mediation and the development of higher mental func-
tions. Language Sciences 14(4): 489–528.
Hasan, Ruqaiya. 1995a. On the social conditions for semiotic mediation: The genesis of mind in society.
In Alan R. Sadovnik (ed.), Knowledge and pedagogy: The sociology of Basil Bernstein. Norwood, NJ: Ablex,
171–196.
Hasan, Ruqaiya. 1996a [1995b]. Semantic networks: A tool for the analysis of meaning. In Ruqaiya Hasan
(ed.), 104–131.
Hasan, Ruqaiya. 1996a. Ways of saying: Ways of meaning: Selected papers of Ruqaiya Hasan, edited by Cloran,
Carmel, David Butt and Geoffrey Williams. London: Cassell.
Hasan, Ruqaiya. 1996b. Semantic networks: A tool for the analysis of meaning. In Ruqaiya Hasan (ed.),
51–72.
Hasan, Ruqaiya. 1999a. Speaking with reference to context. In Mohsen Ghadessy (ed.), Text and context in
functional linguistics: Systemic perspectives. Amsterdam and Philadelphia: John Benjamins. 219–328.
Hasan, Ruqaiya. 1999b. Society, language and the mind: The meta-dialogism of Basil Bernstein’s theory.
In Fran Christie (ed.) Pedagogy and the shaping of consciousness: Linguistic and social processes. London and
New York: Continuum. 10–30.
Hasan, Ruqaiya. 2004. The world in words: Semiotic mediation, tenor and ideology. In Geoff Williams
and Annabelle Lukin (eds.), 158–181.
Hasan, Ruqaiya. 2005. Language, society and consciousness. In Jonathan J. Webster (ed.), Language, society
and consciousness: Learning and teaching in society. The collected works of Ruqaiya Hasan. Volume 1.
London: Equinox. 106–129.

483
 References

Hasan, Ruqaiya. 2007. Private pleasure, public discourse: Reflections on engaging with literature. In
Donna Miller and Monica Turci (eds.), Language and verbal art revisited: Linguistic approaches to the study of
literature. London and Oakville: Equinox. 13–40.
Hasan, Ruqaiya. 2009a. The place of context in a systemic functional model. In M.A.K. Halliday and
Jonathan J. Wester (eds.), 166–189.
Hasan, Ruqaiya. 2009b. Semantic variation: Meaning in society and sociolinguistics. Volume 2 of the Collected
Works of Ruqaiya Hasan, edited by Jonathan J. Wester. London: Equinox.
Hasan, Ruqaiya. 2014. Towards a paradigmatic description of context: Systems, metafunctions, and seman-
tics. Functional Linguistics 1: 9.
Hasan, Ruqaiya. 2015. Systemic functional linguistics: Halliday and the evolution of social semiotics. In
Jonathan J. Webster (ed.), 101–134.
Hasan, Ruqaiya. 2019. Describing language: Form and function. Volume 5 of the collected works of Ruqaiya
Hasan, edited by Carmel Cloran and Jonathan J. Wester. London: Equinox.
Hasan, Ruqaiya. forthcoming. Verbal art: A social semiotic perspective. Volume 7 of the collected works of
Ruqaiya Hasan, edited by Jonathan J. Webster and David G. Butt. Sheffield: Equinox.
Hasan, Ruqaiya and Carmel Cloran. 1990. A sociolinguistic interpretation of everyday talk between moth-
ers and children. In M.A.K. Halliday, John Gibbons and Howard Nicholas (eds.), Learning, keeping and
using language: Selected papers from the eighth world congress of applied linguistics, Sydney. Philadelphia: John
Benjamins, 16–21 August. 67–99.
Hasan, Ruqaiya, Carmel Cloran, Geoff Williams and Annabelle Lukin. 2007. Semantic networks: The
description of linguistic meaning in SFL. In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jona-
than J. Wester (eds.), 697–738.
Hasan, Ruqaiya and Peter Fries (eds.). 1995. On subject and theme: A discourse functional perspective. Amster-
dam and Philadelphia: Benjamins.
Hasan, Ruqaiya and J.R. Martin (eds.). 1989. Language development: Learning language, learning culture. Mean-
ing and choice in language. Norwood, NJ: Ablex.
Hasan, Ruqaiya, Christian M.I.M. Matthiessen and Jonathan J. Wester (eds.). 2005. Continuing discourse on
language: A functional perspective. Volume 1. London: Equinox Publishing.
Hasan, Ruqaiya, Christian M.I.M. Matthiessen and Jonathan J. Wester (eds.). 2007. Continuing discourse on
language: A functional perspective. Volume 2. London: Equinox Publishing.
Hasan, Ruqaiya and Gillian Perrett. 1994. Learning to function with the other tongue: A systemic func-
tional perspective on second language teaching. In Terrence Odlin (ed.), Perspectives on pedagogical gram-
mar. Cambridge: Cambridge University Press. 179–226.
Hattie, John A.C. 2009. Visible learning: A synthesis of over 800 meta-analyses relating to achievement. London
and New York: Routledge.
Hawkins, P.R. 1973. Social class, the nominal group and references. In Basil Bernstein (ed.), Class, codes and
control Volume 2: Applied studies towards a sociology of language. London and Boston: Routledge.
Hays, David G. 1964. Dependency theory: A formalism and some observations. Language 40(4):
511–525.
He, Hengxing. 2019. Past achievements and future possibilities of systemic-functional discourse/text analy-
sis: In memory of M. A. K. Halliday. International Journal of Language and Linguistics 7(1): 32–41.
Heller, Louis G. and James Macris. 1967. Parametric linguistics. The Hague: Mouton.
Henderson, Eugénie J.A. 1987. J.R. Firth in retrospect: A view from the eighties. In Ross Steele and Terry
Threadgold (eds.), Language topics. Essays in honour of Michael Halliday. Amsterdam: Benjamins. 57–69.
Henderson, Monika and Michael Argyle. 1985. Social support by four categories of work colleagues:
Relationships between activities, stress and satisfaction. Journal of Occupational Behaviour 6(3): 229–239.
Henderson-Brooks, Caroline Kay. 2006. ‘What type of person am I, Tess?’: The complex tale of self in psycho-
therapy. Ph.D. thesis. Macquarie University, Sydney.
Hengeveld, Kees and J. Lachlan Mackenzie. 2008. Functional discourse grammar: A typologically-based theory of
language structure. Oxford: Oxford University Press.
Henrici, Alick. 1981 [1965]. Notes on the systemic generation of a paradigm of the English clause. In
M.A.K. Halliday and James R. Martin (eds.), Readings in systemic linguistics. London: Batsford. 74–98.
Henriques, Gregg. 2003. The tree of knowledge system and the theoretical unification of psychology.
Review of General Psychology 7(2): 150–182.
Henriques, Gregg. 2011. A new unified theory of psychology. Heidelberg and New York: Springer.
Hermerén, Lars. 1978. On modality in English: A study of the semantics of the modals. (Lund Studies in English
53). Lund: CWK Gleerup.

484
 References

Heuboeck, Alois, Jasper Holmes and Hilary Nesi. 2010. The BAWE corpus manual for the project entitled ‘an
investigation of genres of assessed writing in British Higher Education’. Version III, June. Available from www.
coventry.ac.uk/globalassets/media/global/08-new-research-section/current-projects/bawemanual-v3.
pdf
Hiippala, Tuomo. 2014. Modelling the structure of a multimodal artefact. Ph.D. thesis. University of Helsinki,
Helsinki.
Hiippala, Tuomo. 2016. The structure of multimodal documents: An empirical approach. New York and Abing-
don: Routledge.
Hill, Trevor. 1966. The technique of prosodic analysis. In Bazell, Charles Ernest, John Cunnison Catford,
M.A.K. Halliday and Robert Henry Robins (eds.), 198–226.
Hill-Madsen, Aage. 2020. SFL and descriptive translation studies: Systemic-functional grammar as a frame-
work for the analysis of shifts in translation. Nordisk SFL i praksis og teori, Special Issue of Globe: A Journal
of Language, Culture and Communication 10: 143–169.
Hjelmslev, Louis. 1943. Omkring sprogteoriens grundlæggelse. København: Akademisk Forlag. (English ver-
sion. 1961 Prolegomena to a theory of language. Madison, Wisconsin: University of Wisconsin Press).
Hoang, Van Van. 1997. An experiential grammar of the Vietnamese clause: A functional description. Ph.D. thesis.
Macquarie University, Sydney.
Hoang, Van Van. 2012. An experiential grammar of the Vietnamese clause. Ha Noi: Vietnam Education Pub-
lishing House.
Hobbs, Jerry R., Mark E. Stickel, Douglas E. Appelt and Paul Martin. 1993. Interpretation as abduction.
Artificial Intelligence 63: 69–142.
Hodge, Robert. 2017. Discourse analysis. In Tom Bartlett and Gerard O’Grady (eds.), 520–532.
Hodge, Robert and Gunther Kress. 1988. Social semiotics. Ithaca: Cornell University Press.
Hoey, Michael. 1983. On the surface of discourse. London: George Allen and Unwin.
Hoey, Michael. 1999. Persuasive rhetoric in linguistics: A stylistic study of some features of the language of
Noam Chomsky. In Susan Hunston and Geoff Thompson (eds.), Evaluation in text: Authorial stance and
the construction of discourse. Oxford: Oxford University Press. 28–37.
Hoffmann, Thomas and Graeme Trousdale (eds.). 2013. The Oxford handbook of construction grammar.
Oxford: Oxford University Press.
Holmberg, Per. 2002. Emotiv betydelse och evaluering i text. (Emotive meaning and evaluation in text). Göte-
borg: Acta Universitatis Gothoburgensis.
Holmberg, Per and Anna-Malin Karlsson. 2006. Grammatik med betydelse: En introduktion till funktionell
grammatik. Uppsala: Hallgren and Fallgren.
Honeybone, Patrick. 2005. Firth, J.R. (John Rupert). In Siobhan Chapman and Christopher Routledge
(eds.), Key thinkers in linguistics and the philosophy of language. Edinburgh: Edinburgh University Press.
80–86.
Honnibal, Matthew. 2004. Adapting the Penn Treebank to systemic functional grammar: Design, creation and use of
a metafunctionally annotated corpus. BA honours thesis. Macquarie University, Sydney.
Hood, Susan. 2010. Naming and negotiating relationships in call centre talk. In Gail Forey and Jane Lock-
wood (eds.), 88–105.
Hood, Susan. 2011. Body language in face-to-face teaching: A focus on textual and interpersonal meaning.
In Shoshana Dreyfus, Susan Hood and Maree Stenglin (eds.), Semiotic margins: Meaning in multimodalities.
London and New York: Continuum. 31–52.
Hood, Susan. 2019. Appraisal. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine and David Schöntal
(eds.), 382–409.
Hood, Susan and Gail Forey. 2008. The interpersonal dynamics of call-centre interactions: Co-construct-
ing the rise and fall of emotion. Discourse and Communication 2(4): 389–409.
Hood, Sue and J.R. Martin. 2007. Invoking attitude: The play of graduation in appraising discourse. In
Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan J. Webster (eds.), 739–764.
Hopper, Paul J. (ed.). 1982. Tense-aspect: Between semantics and pragmatics. Amsterdam: John Benjamins.
Hopper, Paul J. and Sandra A. Thompson. 1980. Transitivity in grammar and discourse. Language 56:
251–299.
Hopper, Paul and Sandra A. Thompson (eds.). 1982. Studies in transitivity. (Syntax and Semantics, 15). New
York: Academic Press.
Hopper, Paul and Sandra A. Thompson. 2001. Transitivity, clause structure, and argument structure: Evi-
dence from conversation. In Joan Bybee and Paul Hopper (eds.), Frequency and the emergence of linguistic
structure. Amsterdam: Benjamins. 27–60.

485
 References

Hopper, Paul and Elizabeth Closs Traugott. 1993. Grammaticalization. Cambridge: Cambridge University
Press.
House, Juliane. 1977. A model for translation quality assessment. Tübingen: Narr.
House, Juliane. 1997. Translation quality assessment: A model revisited. Tübingen: G. Narr.
House, Juliane. 2001a. How do we know when a translation is good? In Erich Steiner and Colin Yal-
lop (eds.), Beyond content: Exploring translation and multilingual text production. Berlin and New York: de
Gruyter. 127–160.
House, Juliane. 2001b. Translation quality assessment: Linguistic description versus social evaluation. Meta:
Translators’ Journal 46(2): 243–257.
House, Juliane. 2015. Translation quality assessment: Past and present. London and New York: Routledge.
House, Juliane. 2016. Translation as communication across languages and cultures. London and New York:
Routledge.
Hoyte, Frances, Jane Torr and Sheila Degotardi. 2014. The language of friendship: Genre in the conversa-
tions of preschool children. Journal of Early Childhood Research 12(1): 20–34.
Huang, Guowen. 1996. Experiential enhanced Theme in English. In Margaret Berry, Christopher Butler,
Robin Fawcett and Guowen Huang (eds.), Meaning and form: Systemic functional interpretations. Nor-
wood, NJ: Ablex. 65–113.
Huang, Guowen. 2002. Cleft sentences as grammatical metaphors. In Huang Guowen and Wang Zongyan
(eds.), Discourse and language functions. Shanghai: Foreign Language Teaching and Research Press. 34–41.
Huang Guowen. 2011. Challenges of developing SFL in China. Contemporary Foreign Language Studies 12:
15–23.
Huang, Guowen. 2017. Theme in the Cardiff Grammar. In Tom Bartlett and Gerard O’Grady (eds.), Chap.
11. 163–177.
Huang, Guowen, Chang C, and Dai Fan (eds.). 2010. Functional linguistics as appliable linguistics. Guangzhou:
Sun Yat-sen University Press.
Huang, Guowen and Yumin Chen, annotated. 2016. Key terms in systemic functional linguistics. In Christian
M.I.M. Matthiessen, Kazuhiro Teruya and Wenjie Lin. 2010. Beijing: Foreign Language Teaching and
Research Press.
Huddleston, Rodney D. 1965. Rank and depth. Language 41: 574–586. Reprinted in M.A.K. Halliday and
J.R. Martin (eds.), 1981, 42–53.
Huddleston, Rodney D. 1971. The sentence in written English: A syntactic study based on an analysis of scientific
texts. London: Cambridge University Press.
Huddleston, Rodney D. and Ormond Uren. 1969. Declarative, interrogative, and imperative in French,
Lingua 22: 1–26.
Huddleston, Rodney D., Richard A. Hudson, Eugene Winter and Alick Henrici. 1968. Sentence and clause
in scientific English: Final report of O.S.T.I. programme. London: University College London, Communica-
tion Research Centre.
Hudson, Richard A. 1964. The grammatical study of Beja. Ph.D. thesis. London University, London.
Hudson, Richard A. 1971. English complex sentences. Amsterdam: North Holland.
Hudson, Richard A. 1973. An item-and-paradigm approach to Beja syntax and morphology. Foundations
of Language 9: 504–548.
Hudson, Richard A. 1976. Arguments for a non-transformational grammar. Chicago: Chicago University Press.
Hudson, Richard A. 1984. Word grammar. Oxford: Oxford University Press.
Hudson, Richard A. 1987. Daughter dependency theory and systemic grammar. In M.A.K. Halliday and
Robin P. Fawcett (eds.), New development in systemic linguistics. London: Frances Pinter.
Hudson, Richard A. 2010. An introduction to word grammar. Cambridge: Cambridge University Press.
Hudson, Richard A. 2021. Word grammar. In Xu Wen and John R. Taylor (eds.), The Routledge handbook
of cognitive linguistics. New York and London: Routledge. 111–126.
Hui, Jon S.Y. 2010. ‘I was so angry. It was unbelievable . . .’: A comparison of written and spoken customer
service complaints. In Gail Forey and Jane Lockwood (eds.), 59–87.
Humphrey, Sally. 1996. Exploring literacy in school geography. Erskineville NSW: Metropolitan East Disad-
vantaged Schools Program.
Humphrey, Sally, J.R. Martin, Shooshi Dreyfus and Ahmar Mahboob. 2010. The 3×3: Setting up a lin-
guistic toolkit for teaching academic writing. In Ahmar Mahboob and Naomi Knight (eds.), 185–199.
Hunston, Susan. 1993. Evaluation and ideology in scientific writing. In Mohsen Ghadessy (ed.), Register
analysis: Theory and practice. London and New York: Pinter. 57–73.
Hunston, Susan. 2010. Corpus approaches to evaluation: Phraseology and evaluative language. London: Routledge.

486
 References

Hunston, Susan. 2011. Corpus approaches to evaluation: Phraseology and evaluative language. London: Routledge.
Hunston, Susan and Gill Francis. 2000. Pattern grammar: A corpus-driven approach to the lexical grammar of
English. Amsterdam: Benjamins.
Hunston, Susan and Hang Su. 2017. Patterns, constructions, and local grammar: A case study of ‘evalua-
tion’. Applied Linguistics 40(4): 567–593.
Hunston, Susan and John Sinclair. 1999. A local grammar of evaluation. In Susan Hunston and Geoff
Thompson (eds.), 74–101.
Hunston, Susan and Geoff Thompson (eds.). 1999. Evaluation in text: Authorial stance and the construction of
discourse. Oxford: Oxford University Press.
Hunston, Susan and Geoff Thompson (eds.). 2006. System and corpus: Exploring connections. London and
Oakville: Equinox:
Hu Zhuanglin, Zhu Yongsheng and Zhang Delu. 1989. A survey of systemic functional grammar [in Chinese].
Changsha: Hunan Educational Publishing House.
Hyland, Ken and Carmen Sancho Guinda (eds.). 2012. Stance and voice in written academic genres. New York:
Palgrave Macmillan.
Hymes, Dell H. 1967. Models of the interaction of language and social setting. Journal of Social Issues 23(2):
8–38.
Hyon, Sunny. 1996. Genre in three traditions: Implications for ESL. TESOL Quarterly 30: 693–722.
Iedema, Rick. 1995. Administrative literacy. Sydney: Metropolitan East Disadvantaged Schools Program.
Iedema, Rick. 1996. The language of administration. (Write it right industry research report no. 3). Sydney:
NSW, Department of Education, Disadvantaged Schools Program Metropolitan East.
Iedema, Rick. 1997. The language of administration: Organizing human activity in formal institutions. In
Frances Christie and J.R. Martin (eds.), Genre and institutions: Social processes in the workplace and school.
London: Cassell. 73–100.
Iedema, Rick (ed.). 2007. The discourse of hospital communication: Tracing complexities in contemporary health
organizations. London and New York: Palgrave Macmillan.
Iedema, Rick, Susan Feez and Peter White. 1994. Media literacy. (Write it right industry research report
no. 2). Sydney: NSW, Department of Education, Disadvantaged Schools Program Metropolitan East.
Ivinson, Gabrielle. 2018. Re-imagining Bernstein’s restricted codes. European Educational Research Journal
17(4): 539–554.
Jackendoff, Ray. 1972. Semantic interpretation in generative grammar. Cambridge, MA: MIT Press.
Jackendoff, Ray. 1983. Semantics and cognition. Cambridge, MA: MIT Press.
Jackendoff, Ray. 1988. Conceptual semantics. In Umberto Eco, Marco Santambrogio and Patrizia Violi.
(eds.) Meaning and mental representations. Bloomington, IN: Indiana University Press. 81–99.
Jackendoff, Ray. 2012. A user’s guide to thought and meaning. Oxford: Oxford University Press.
Jakobsen, Arnt Lykke. 2014. The development and current state of translation process research. In Elke
Brems, Reine Meylaerts and Luc van Doorslaer (eds.), The known unknowns of translation studies. Amster-
dam: John Benjamins. 65–88.
Jakobsen, Arnt Lykke. 2017. Translation process research. In John W. Schwieter and Aline Ferreira (eds.),
The handbook of translation and cognition. Hoboken, NJ: Wiley-Blackwell. 19–49.
Jakobson, Roman. 1949. On the identification of phonemic entities. Recherches Structurales, Travaux du
Cercle Linguistique de Prague V: 205–213.
Jakobson, Roman. 1960. Linguistics and poetics. In Thomas A. Sebeok (ed.), Style in language. Cambridge,
MA: MIT Press. 350–377.
Jakobson, Roman, C. Gunnar, M. Fant and Morris Halle. 1951. Preliminaries to speech analysis: The distinctive
features and their correlates. Cambridge, MA: The MIT Press.
Jesus, Silvana Maria de and Adriana Silvina Pagano. 2006. Probabilistic grammar in translation. In Leila
Barbara and Tony Berber Sardinha (eds.), Proceedings of the 33rd International Systemic Functional Congress.
São Paulo: PUCSP. 428–448.
Jewitt, Carey (ed.). 2009. The Routledge handbook of multimodal analysis. London and New York:
Routledge.
Jewitt, Carey (ed.). 2014. The Routledge handbook of multimodal analysis. 2nd edition. London and New
York: Routledge.
Jindrich, Toman. 1995. The magic of a common language: Jakobson, Mathesius, Trubetzkoy, and the Prague Circle.
Cambridge, MA: MIT Press.
Johansson, Stig. 2007. Seeing through multilingual corpora: On the use of corpora in contrastive studies. Amsterdam
and Philadelphia, PA: John Benjamins.

487
 References

Johnson, Allen W. and Timothy Earle. 2000. The evolution of human societies: From foraging group to agrarian
state. Stanford: Stanford University Press.
Johnston, Trevor. 1992a. The realization of the linguistic metafunctions in a sign language. Language Sciences
14(4): 317–353.
Johnston, Trevor. 1992b. The realization of the linguistics metafunctions in a sign language. Social Semiotics
2(1): 1–43.
Jones, Carys and Eija Ventola (eds.). 2008. From language to multimodality: New developments in the study of
ideational meaning. London: Equinox.
Jordens, Chris. 2002. Reading spoken stories for values: A discursive study of cancer survivors and their professional
careers. Ph.D. thesis. The University of Sydney, Sydney.
Jordens, Chris, J. Little, K. Paul and E. Sayers. 2001. Life disruption and generic complexity: A social lin-
guistic analysis of narratives of cancer illness. Social Science and Medicine 53(9): 1227–1236.
Jorgensen, Mikaela and Leanne Togher. 2009. Narrative after traumatic brain injury: A comparison of
monologic and jointly-produced discourse. Brain Injury 23(9): 727–740.
Joshi, Aravind K. and Yves Schabes. 1997. Tree-adjoining grammars. In Grzegorz Rozenberg and Arto
Salomaa (eds.), Handbook of formal languages. Volume 3: Beyond words. Berlin and Heidelberg: Springer.
69–123.
Jurafsky, Daniel. 1992. An on-line computational model of human sentence Interpretation: A theory of the rep-
resentation and use of linguistic knowledge. In American Association for Artificial Intelligence (eds.), Pro-
ceedings of the National Conference on Artificial Intelligence (AAAI-92). Cambridge, MA: MIT Press. 302–308.
Kachru, Braj B. 1980. Socially realistic linguistics: The Firthian tradition. Studies in the Linguistic Sciences
10(1): 85–111.
Kaku, Michio. 1994. Hyperspace: A scientific odyssey through parallel universes, time warps, and the tenth dimen-
sion. New York: Anchor Books.
Kamp, Hans and Uwe Reyle 1993. From discourse to logic: Introduction to modeltheoretic semantics of natural
language, formal logic and discourse representation theory. Berlin: Springer.
Kaplan, Ronald and Joan Bresnan. 1982. Lexical-functional grammar: A formal system for grammatical
representation. In Joan Bresnan (ed.), The mental representation of grammatical relations. Cambridge, MA:
The MIT Press. 173–281.
Karimi, Neda. 2017. Patient-centred advanced cancer care: A systemic functional linguistic analysis of oncology con-
sultations with advanced cancer patients. PhD thesis. Macquarie University, Sydney.
Karimi, Neda, Alison Moore and Annabelle Lukin. 2018. Cancer care as an integrated practice: Consulta-
tions between an oncologist and patients with advanced, incurable cancer. In Akila Sellami-Baklouti
and Lise Fontaine (eds.), 315–337.
Kashyap, Abhishek Kumar. 2012. The pragmatic principle of agreement in Bajjika verbs. Journal of Pragmat-
ics, 44(13): 1868–1887.
Kashyap, Abhishek Kumar. 2019. Language typology. In Geoff Thompson, Wendy L. Bowcher, Lise Fon-
taine and David Schöntal (eds.), 767–792.
Kashyap, Abhishek Kumar. forthcoming. A functional grammar of Bajjika: An SFL perspective. Leiden: Brill.
Kashyap, Abhishek Kumar and Christian M.I.M. Matthiessen. 2017. Figure and ground in the construal of
motion: A registerial perspective. Word 63(1): 62–91.
Kashyap, Abhishek Kumar and Christian M.I.M. Matthiessen. 2019. The representation of motion in dis-
course: Variation across registers. Language Sciences 72: 71–92.
Kasper, Robert. 1988. Systemic grammar and functional unification grammar. In James D. Benson and
William S. Greaves (eds.), Systemic functional approaches to discourse. Norwood: Ablex. 176–199.
Kay, Martin. 1979. Functional grammar. Proceedings of the fifth annual meeting of the Berkeley Linguistics Society.
142–158.
Kealley, Dorothy Jill. 2007. ‘I can’t find a pulse but that’s OK’: Nursing in context: A systemic functional linguistic
examination of nursing practice. Ph.D. thesis. University of South Australia, Adelaide.
Kealley, Jill, Colleen Smith and Bill Winser. 2004. Information empowers but who is empowered? Com-
munication and Medicine 1(2): 119–129.
Keenan, Edward L. 1976a. Remarkable subjects in Malagasy. In Charles Li (ed.), Subject and topic. New
York: Academic Press. 247–301.
Keenan, Edward L. 1976b. Towards a universal definition of “subject”. In Charles Li (ed.), Subject and topic.
New York: Academic Press. 303–333.
Kempson, Ruth, Wilfred Meyer-Viol and Dov Gabbay. 2001. Dynamic syntax: The flow of language under-
standing. Oxford: Blackwell.

488
 References

Kies, Daniel. 1988. Marked theme with and without pronominal reinforcement: Their meaning and dis-
tribution in discourse. In Erich H. Steiner and Robert Veltman (eds.), Pragmatics, discourse and text: Some
systemically-inspired approaches. London: Pinter. 47–75.
Kilpert, Diana. 2003. Getting the full picture: A reflection on the work of M.A.K. Halliday. Language Sci-
ences 25(2): 159–209.
Kim, Mira. 2007a. A discourse based study on THEME in Korean and textual meaning in translation. Ph.D. thesis.
Macquarie University, Sydney.
Kim, Mira. 2007b. Translation error analysis: A systemic functional approach. In Dorothy Kenny and Kyo-
ngjoo Ryou (eds.), Across boundaries: International perspectives on translation studies. Newcastle upon Tyne:
Cambridge Scholars Publishing.
Kim, Mira. 2007c. Using systemic functional text analysis for translator education: An illustration with a
focus on textual meaning. The Interpreter and Translator Trainer 1(2): 223–246.
Kim, Mira, J.R. Martin, Gi-Hyun Shin and Gyunghee Choi. in preparation. A systemic functional grammar
of Korean. Cambridge: Cambridge University Press.
Kim, Mira and Christian M.I.M. Matthiessen. 2017. Ways to move forward in translation studies: A textual
perspective. In Jeremy Munday and Meifang Zhang (eds.), Discourse analysis in translation studies. Amster-
dam: John Benjamins. 11–26.
Kim, Mira, Jeremy Munday, Zhenhua Wang and Pin Wang (eds.). 2021. Systemic functional linguistics and
translation studies. London: Bloomsbury Academic.
Kinneavy, James L. 1971. A theory of discourse: The aims of discourse. New York and London: Prentice-Hall.
Kittredge, Richard. 1987. The significance of sublanguage for automatic translation. In Sergei Nirenburg
(ed.), Machine translation: Theoretical and methodological issues. Cambridge: Cambridge University Press.
59–67.
Kittredge, Richard and John Lehrberger (eds.). 1982. Sublanguage: Studies of language in restricted semantic
domains. Berlin: de Gruyter.
Knain, Erik. 2015. Scientific literary for participation: A systemic functional approach to analysis of school science
discourses. Rotterdam: Sense Publishers.
Knight, Naomi K. 2010a. Wrinkling complexity: Concepts of identity and affiliation in humour. In Mon-
ika Bednarek and J.R. Martin (eds.), 34–58.
Knight, Naomi K. 2010b. Laughing our bonds off: Conversational humour in relation to affiliation. Ph.D. thesis.
The University of Sydney, Australia.
Kobayashi, Ichiro. 1995. A social system simulation based on human information processing. Ph.D. thesis. Tokyo
Institute of Technology, Tokyo.
Kobayashi, Ichiro and Michio Sugeno. 1994. An approach to social system simulation based on linguistic
information: An application to the forecast of foreign exchange rate changes. Journal of Japan Society for
Fuzzy Theory and Systems 6(4), August. (In Japanese).
Körner, Henrike. 2000. Negotiating authority: The logogenesis of dialogue in common law judgments. Ph.D. thesis.
The University of Sydney, Sydney.
Korta, Kepa. 2008. Malinowski and pragmatics: Claim making in the history of linguistics. Journal of Prag-
matics 40: 1645–1660.
Koutsikou, Maria and Vasilia Christidou. 2019. The interplay between interpersonal and compositional
meanings in multimodal texts about animals for young children. Punctum 5(1): 114–137.
Kress, Gunther. 1985. Linguistic processes in sociocultural experience. Geelong, VIC: Deakin University Press.
Kress, Gunther. 2010. Multimodality: A social semiotic approach to contemporary communication. Milton Park and
New York: Routledge.
Kress, Gunther. 2012. Multimodal discourse analysis. In James Paul Gee and Michael Handford (eds.), The
Routledge handbook of discourse analysis. London and New York: Routledge. 35–50.
Kress, Gunther and Robert Hodge. 1979. Language as ideology. London: Routledge and Kegan Paul.
Kress, Gunther and Theo van Leeuwen. 1990. Reading images. Geelong, VIC: Deakin University Press.
Kress, Gunther and Theo van Leeuwen. 1996. Reading images: The grammar of visual design. London:
Routledge.
Kress, Gunther and Theo van Leeuwen. 2001. Multimodal discourse: The modes and media of contemporary
communication. London: Arnold.
Kress, Gunther and Theo van Leeuwen. 2006. Reading images: The grammar of visual design. 2nd edition.
London: Routledge.
Kronenfeld, David. 1996. Plastic glasses and church fathers: Semantic extension from the ethnoscience tradition.
(Oxford Studies in Anthropological Linguistics). Oxford: Oxford University Press.

489
 References

Kumar, Abhishek. 2009. A systemic functional description of the grammar of Bajjika. Ph.D. thesis. Macquarie
University, Sydney.
Kunz, Kerstin and Elke Teich. 2017. Translation studies. In Tom Bartlett and Gerard O’Grady (eds.),
547–560.
Kunz, Kerstin, Ekaterina Lapshinova-Koltunski, José Manuel Martínez Martínez, Katrin Menzel and Erich
Steiner. 2021. GECCo – German-English contrasts in cohesion insights from corpus-based studies of languages,
registers and modes. Berlin: De Gruyter Mouton.
Kunz, Kerstin, Elke Teich, Silvia Hansen-Schirra, Stella Neumann and Peggy Daut (eds.). 2014. Caught
in the middle: Language use and translation: A festschrift for Erich Steiner on the occasion of his 60th birthday.
Saarland: Saarland University Press.
Kuper, Adam. 1983. Anthropology and anthropologists: The modern British school. London: Routledge and
Kegan Paul.
Ladefoged, Peter. 1988. Hierarchical features of the International Phonetic Alphabet. Proceedings of the
Fourteenth Annual Meeting of the Berkeley Linguistics Society. Berkeley, CA: Berkeley Linguistics Society.
124–141.
Ladefoged, Peter. 1997. Instrumental techniques for linguistic phonetic fieldwork. In William J. Hardcastle
and John Laver (eds.), The handbook of phonetic sciences. Oxford: Blackwell. 137–166.
Ladefoged, Peter. 2004. Phonetics and phonology in the last 50 years. Paper Presented at From Sound to Sense,
June 11–13. Cambridge, MA: MIT.
Laffut, An. 2006. Three-participant constructions in English: A functional-cognitive approach to caused relations.
Amsterdam: Benjamins.
Lakoff, George. 1987. Women, fire and dangerous things. Chicago: Chicago University Press.
Lakoff, George and Mark Johnson. 1980. Metaphors we live by. Chicago: University of Chicago Press.
Lakoff, George and Mark Johnson. 1999. Philosophy in the flesh: The embodied mind and its challenge to western
thought. New York: Basic Books.
Lam, Marvin. 2015. Interfacing field with tenor: Hasan’s notion of personal distance. Wendy L. Bowcher
and Yameng Liang (eds.), 206–226.
Lamb, Sidney M. 1966. Outline of stratificational grammar. Washington DC: Georgetown University Press.
Lamb, Sydney M. 1999. Pathways of the brain: The neurocognitive basis of language. Amsterdam: Benjamins.
Lamb, Sydney M. 2013. Systemic networks, relational networks, and choice. In Lise M. Fontaine, Tom
A.M. Bartlett and Gerard N. O’Grady (eds.), 137–160.
Langacker, Ronald W. 1990. Concept, image, and symbol: The cognitive basis of grammar. Berlin: Mouton de
Gruyter.
Langacker, Ronald W. 2005. Construction grammars: Cognitive, radical, and less so. In Francisco J. Ruiz
de Mendoza Ibáñez and M. Sandra Peña Cervel (eds.), Cognitive linguistics: Internal dynamics and interdis-
ciplinary interaction. Berlin and New York: Mouton de Gruyter. 101–159.
Langacker, Ronald W. 2008. Cognitive grammar: A basic introduction. Oxford: Oxford University Press.
Langacker, Ronald W. 2009. Investigations in cognitive grammar. Berlin: Mouton de Gruyter.
Langacker, Ronald W. 2013. Essentials of cognitive grammar. Oxford: Oxford University Press.
Lantolf, James. 2010. Minding your hands: The function of gesture in L2 learning. In Robert Batstone
(ed.), Sociocognitive perspectives on language use and language learning. Oxford: Oxford University Press.
131–150.
Lappin, Shalom and Chris Fox (eds.). 2015. The handbook of contemporary semantic theory. 2nd edition.
Oxford: Wiley Blackwell.
Larsen-Freeman, Diane. 2011. Saying what we mean: Making a case for ‘language acquisition’ to become
‘language development’. Plenary at AILA, Beijing.
Larsen-Freeman, Diane. 2015. Saying what we mean: Making a case for ‘language acquisition’ to become
‘language development’. Language Teaching 48(4): 491–505.
Larsen-Freeman, Diane and Lynne Cameron. 2008. Complex systems and applied linguistics. Oxford: Oxford
University Press.
Lassen, Inger and Jeanne Strunck. 2011. ‘I think Danish patients would feel the same’: Counter-discourses
emerging in the Danish health sector. Communication and Medicine 8(3): 223–233.
Lavid, Julia. 2000. Cross-cultural variation in multilingual instructions: A study of speech act realisation
patterns. In Eija Ventola (ed.), Discourse and community: Doing functional linguistics. Tübingen: Günter
Narr Verlag. 71–85.
Lavid, Julia, Jorge Arús and Lara Moratón. 2009. Comparison and translation: Towards a combined meth-
odology for contrastive corpus studies. International Journal of English Studies Special Issue, 159–173.

490
 References

Lavid, Julia, Jorge Arús and Juan Rafael Zamorano-Mansilla. 2010. Systemic functional grammar of Spanish: A
contrastive study with English. London and New York: Continuum.
Law, Lok Hei Locky. 2017. House M.D. and creativity: A corpus linguistic systemic functional multimodal discourse
analysis approach. Ph.D. thesis. The Hong Kong Polytechnic University, Hong Kong.
Layzer, David. 1990. Cosmogenesis: The growth of order in the universe. New York and Oxford: Oxford Uni-
versity Press.
Leckie-Tarry, Helen. 1995. Language and context: A functional linguistic theory of register. London: Pinter.
Lee, Amanda, Ondene van Dulm, Michael Robb and Tika Ormond. 2015. Communication restriction in
adults who stutter. Clinical Linguistics and Phonetics 29(7): 536–556.
Lee, Amanda, Michael Robb, Ondene van Dulm and Tika Ormond. 2016a. Communication restriction
in adults who stutter: Part II. Clinical Linguistics and Phonetics 30(7): 546–567.
Lee, Amanda S., Michael Robb, Ondene van Dulm and Tika Ormond. 2016b. Communication restriction
in adults who stutter: Part III. Clinical Linguistics and Phonetics 30(11): 911–924.
Lee, Jackson L. 2018. On the discovery procedure. In Diana Brentari and Jackson L. Lee (eds.). Shaping
phonology. Chicago: The University of Chicago Press. 223–233.
Lee, Penny. 1996. The Whorf theory complex: A critical reconstruction. Amsterdam and Philadelphia: John
Benjamins.
Leech, Geoffrey. 1974. Semantics: The study of meaning. 2nd edition. Harmondsworth: Penguin.
Lemke, Jay L. 1984. The formal analysis of instruction. In Jay L. Lemke (ed.), Semiotics and Education,
Working Papers and Prepublication, No. 2. Victoria University. Toronto: Toronto Semiotic Circle
Monographs. 94–149.
Lemke, Jay L. 1985. Using language in the classroom. Geelong, VIC: Deakin University Press.
Lemke, Jay L. 1987. The topology of genre: Text structures and text types. MS.
Lemke, Jay L. 1990. Talking science: Language, learning and values. Norwood, NJ: Ablex.
Lemke, Jay L. 1995. Textual politics: Discourse and social dynamics. London and Bristol, PA: Taylor and
Francis.
Lemke, Jay L. 1998. Multiplying meaning: Visual and verbal semiotics in scientific text. In J.R. Martin and
Robert Veel (ed.), 87–113.
Lemke, Jay L. 2012. Multimedia and discourse analysis. In James Paul Gee and Michael Handford (eds.),
The Routledge handbook of discourse analysis. London and New York: Routledge. 79–89.
Lemmens, Maarten. 1988. Lexical perspectives on transitivity and ergativity: Causative constructions in English.
Amsterdam: Benjamins.
Léon, Jacqueline. 2007. From linguistic events and restricted languages to registers: Firthian legacy and
corpus linguistics. The Henry Sweet Society Bulletin 49: 5–26.
Leong, Alvin Ping. 2019. Visualizing texts: A tool for generating thematic-progression diagrams. Functional
Linguistics 6(1). https://doi.org/10.1186/s40554-019-0069-0
Leopold, Werner F. 1939–1949. Speech development of a bilingual child: A linguist’s record. 4 Volumes. Evanston
and Chicago: North-Western University Press.
Leopold, Werner F. 1948. The study of child language and infant bilingualism. Word 4(1): 1–17.
Levin, Beth. 1993. English verb classes and alternations: A preliminary investigation. Chicago and London: The
University of Chicago Press.
Levin, James A. and James A. Moore. 1977. Dialogue‐games: Metacommunication structures for natural
language interaction. Cognitive Science: A Multidisciplinary Journal 1(4): 395–420.
Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press.
Li, Eden Sum-hung. 2007. Systemic functional grammar of Chinese: A text-based analysis. London and New
York: Continuum.
Li, Eden Sum-hung. 2003. A text-based study of the grammar of Chinese from a systemic functional approach.
Ph.D. thesis. Macquarie University, Sydney.
Li, Eden Sum-hung. 2017. The nominal group in Chinese. In Tom Bartlett and Gerard O’Grady (eds.),
Chap. 21. 338–353.
Li, Wei. 2020. System remembers Gunther Kress: Gunther Kress’ contribution to the study of language,
technology, and learning. System 93.
Liles, Betty, Carl A. Coelho, Robert J. Duffy and Mary Rigdon Zalagens. 1989. Effects of elicitation
procedures on the narratives of normal and closed head-injured adults. Journal of Speech and Hearing
Disorders 54: 356–366.
Linde, Charlotte and William Labov. 1975. Spatial networks as a site for the study of language and thought.
Language 51 (4): 924–939.

491
 References

Lobb, E., P. Butow, A.R. Moore, A. Barratt, K. Tucker, C. Gaff, J. Kirk, T. Dudding and D.G. Butt. 2006.
Development of a communication aid to facilitate risk communication in consultations with women
from high risk breast cancer families. Journal of Genetic Counselling 15: 393–405.
Lock, Sarah and Linda Armstrong. 1997. Cohesion analysis of the expository discourse of normal, fluent
aphasic and demented adults: A role in differential diagnosis? Clinical Linguistics and Phonetics 11(4):
299–317.
Lockwood, David G. 1972. Introduction to stratificational linguistics. New York: Harcourt Brace Jovanovich.
Lockwood, David G. 2003. Syntactic analysis and description: A constructional approach. London: Continuum.
Longacre, Robert E. 1964. Grammar discovery procedures: A field manual. The Hague: Mouton.
Longacre, Robert E. 1970. Sentence structure as a statement calculus. Language 46: 783–815.
Longacre, Robert E. 1974. Narrative vs other discourse genres. In Ruth Brend (ed.), Advances in tagmemics.
Amsterdam: North-Holland. 357–376.
Longacre, Robert E. 1979. The paragraph as a grammatical unit. In Talmy Givón (ed.), Syntax and seman-
tics: Discourse and syntax. Volume 12. New York: Academic Press. 115–134.
Longacre, Robert E. 1985. Sentences as combinations of clauses. In Timothy Shopen (ed.), Language typol-
ogy and syntactic descriptions: Volume III complex constructions. Cambridge: Cambridge University Press.
235–287.
Longacre, Robert E. 1996. The grammar of discourse. 2nd edition. New York: Plenum.
Longacre, Robert E. and Shin Ja J. Hwang. 2012. Holistic discourse analysis. Dallas, TX: SIL International
Publications.
López, Néstor Nápoles. 2017. Automatic harmonic analysis of classical string quartets from symbolic score. MSc
thesis. Universitat Pompeu Fabra.
Lucassen, Jan. 2021. The story of work: A new history of humankind. New Haven and London: Yale University
Press.
Lucy, John A. 1992. Language diversity and thought: A reformulation of the linguistic relativity hypothesis. (Studies
in the social and cultural foundations of language 12). Cambridge: Cambridge University Press.
Lukin, Annabelle. 2003. Examining poetry: A corpus based enquiry into literary criticism. Ph.D. thesis. Macquarie
University, Sydney.
Lukin, Annabelle. 2010. ‘News’ and ‘register’: A preliminary investigation. In Ahmar Mahboob and Naomi
K. Knight (eds.), Appliable linguistics. London and New York: Continuum. 92–113.
Lukin, Annabelle. 2015. A linguistics of style: Halliday on literature. In Jonathan J. Wester (ed.),
348–366.
Lukin, Annabelle. 2017. Ideology and the text-in-context relation. Journal of Functional Linguistics 4(16).
Lukin, Annabelle. 2019. War and its ideologies: A social-semiotic theory and description. Singapore: Springer.
Lukin, Annabelle and David B. Butt. 2009. Literary stylistics. In M.A.K. Halliday and Jonathan J. Wester
(eds.), 190–215.
Lukin, Annabelle, Alison Moore, Maria Herke, Rebekah Wegener and Wu Canzhong. 2008. Halliday’s
model of register revisited and explored. Linguistics and the Human Sciences 4(2): 187–243.
Lukin, Annabelle and Jonathan J. Webster. 2005. SFL and the study of literature. In Ruqaiya Hasan, Chris-
tian M.I.M. Matthiessen and Jonathan J. Wester (eds.), 413–456.
Lyons, John. 1977. Semantics. 2 volumes. Cambridge: Cambridge University Press.
Ma, Yuanyi and Wang Bo. 2021. Translating Tagore’s stray birds into Chinese: Applying systemic functional linguis-
tics to Chinese poetry translation. London: Routledge.
Macdonald, Kathleen. 2019. Construing musical discourses: Axial reasoning for a contrastive description of habitual
ideational resources in English and Korean, with reflection on translation. Ph.D. thesis. The Hong Kong Poly-
technic University, Hong Kong.
Mackay, David, Brian Thompson and Pamela Schaub. 1970. Breakthrough to literacy. London: Longman.
Macken-Horarik, Mary. 1998. Exploring the requirements of critical school literacy: A view from two
classrooms. In Frances Christie and Ray Misson (eds.), Literacy and schooling. London: Routledge.
74–103.
Macken-Horarik, Mary. 2012. Why school English needs a ‘good enough’ grammatics (and not more
grammar). Changing English: Studies in Culture and Education 19(2): 179–194.
Made, Sutjaja I Gusti. 1988. The nominal group in Bahasa Indonesia. Ph.D. thesis. The University of Sydney,
Sydney.
Madhyastha, Pranava Swaroop, Xavier Carreras and Ariadna Quattoni. 2017. Prepositional phrase attach-
ment over word embedding products. Proceedings of the 15th International Conference on Parsing Technolo-
gies, Pisa, Italy. 32–43.

492
 References

Mahboob, Ahmar, Shoshana Dreyfus, Sally Humphrey and J.R. Martin. 2010. Appliable linguistics and
English language teaching: The scaffolding literacy in adult and tertiary environments (SLATE) project.
In Ahmar Mahboob and Naomi Knight (eds.), 25–43.
Mahboob, Ahmar and Naomi Knight (eds.). 2010. Appliable linguistics. London and New York: Continuum.
Maiorani, Arianna and Christine Christie (eds.). 2014. Multimodal epistemologies: Towards an integrated frame-
work. London: Routledge.
Makkai, Adam and David Lockwood. 1973. Readings in stratificational linguistics. Alabama: University of
Alabama Press.
Malavska, Valerija. 2016. Genre of an academic lecture. LLCE 3(2): 56–84.
Malcolm, Karen. 1985. Communication linguistics: A sample analysis. In James D. Benson and William S.
Greaves (eds.), Systemic perspectives on discourse. Norwood, NJ: Ablex. 136–151.
Malcolm, Karen. 2010. Phasal analysis: Analyzing discourse through communication linguistics. London:
Bloomsbury.
Maley, Yon and Rhondda Fahey. 1991. Presenting the evidence: Constructions of reality in court. Interna-
tional Journal for the Semiotics of Law 4(1): 3–17.
Malinowski, Bronislaw. 1922. Argonauts of the Western Pacific: An account of native enterprise and adventure in the
Archipelagoes of Melanesian New Guinea. London: Routledge and Kegan Paul.
Malinowski, Bronislaw. 1923. The problem of meaning in primitive languages. In Supplement I to C.K.
Ogden and I.A. Richards (eds.), The meaning of meaning. New York: Harcourt Brace and World.
296–336.
Malinowski, Bronislaw. 1935. Coral gardens and their magic: A study of the methods of tilling the soil and of agri-
cultural rites in the Trobriand Islands. 2 Volumes. London: George Allen and Unwin.
Malinowski, Bronislaw. 1944. A scientific theory of culture and other essays. Chapel Hill: University of North
Carolina Press.
Malkiel, Yakov. 1991 [1990]. Review of Nerlich, Brigitte. Change in language: Whitney, Bréal and Wegener.
London and New York: Routledge. 221–228.
Malmkjaer, Kirsten. 2005. Linguistics and the language of translation. Edinburgh: Edinburgh University
Press.
Mann, William C. 1979. Dialogue games. Marina del Rey, CA: USC Information Sciences Institute, ISI/
RR-79-77.
Mann, William C. 1988. Dialogue games: Conventions of human interaction. Argumentation 2: 511–532.
Mann, William C. 2002. Dialogue macrogame theory. Proceedings of the Third SIGdial Workshop on Discourse
and Dialogue, July 2002. Philadelphia, PA: Association for Computational Linguistics. 129–141.
Mann, William C. and Christian M.I.M. Matthiessen. 1985. Demonstration of the Nigel text generation
computer program. In James D. Benson and William S. Greaves (eds.), Systemic functional approaches to
discourse. Norwood: Ablex. 50–83.
Mann, William C., Christian M.I.M. Matthiessen and Sandra A. Thompson. 1992. Rhetorical structure theory
and text analysis. USC/ISI Report. Also in William C. Mann and Sandra A. Thompson (eds.), Discourse
description: Diverse linguistic analyses of a fund raising text. Amsterdam: Benjamins. 39–78.
Mann, William C. and Sandra A. Thompson (eds.) 1992. Discourse description: Diverse linguistic analyses of a
fund-raising text. Amsterdam and Philadelphia: Benjamins.
Mann, William C. and David J. Weber. 1979. Prospectus for computer-assisted dialect adaptation. Notes on
linguistics special publication 1. Dallas: Summer Institute of Linguistics.
Manning, Christopher D. and Hinrich Schütze. 1999. Foundations of statistical natural language processing.
Cambridge, MA: The MIT Press.
Martin, J.R. 1983. Participant identification in English, Tagalog and Kate. Australian Journal of Linguistics
3(1): 45–74.
Martin, J.R. 1985a. Process and text: Two aspects of human semiosis. In James D. Benson and William S.
Greaves (eds.), Systemic perspectives on discourse. Norwood, NJ: Ablex. 248–274.
Martin, J.R. 1985b. Factual writing: Exploring and challenging social reality. Geelong, VIC: Deakin University
Press.
Martin, J.R. 1986. Grammaticalizing ecology: The politics of baby seals and kangaroos. In Elizabeth A.
Grosz, Terry Threadgold and M.A.K. Halliday (eds.), Semiotics-ideology-language. Sydney: Sydney Asso-
ciation for Studies in Society and Culture. 225–267.
Martin, J.R. 1988. Grammatical conspiracies in Tagalog: Family, face and fate – with reference to Benjamin
Lee Whorf. In Michael J. Cummings, William S. Greaves and James D. Benson (eds.), Linguistics in a
systemic perspective. Amsterdam: Benjamins. 243–300.

493
 References

Martin, J.R. 1990. Interpersonal grammaticalisation: Mood and modality in Tagalog. Philippine Journal of
Linguistics 21(1): 2–51.
Martin, J.R. 1991. Intrinsic functionality: Implications for contextual theory. Social Semiotics 1(1):
99–162.
Martin, J.R. 1992a. English text: System and structure. Amsterdam: Benjamins.
Martin, J.R. 1992b. Macro-proposals: Meaning by degree. In William C. Mann and Sandra A. Thompson
(eds.), Text description: Diverse analyses of a fund-raising text. Amsterdam: Benjamins. 359–395.
Martin, J.R. 1992c. Theme, method of development and existentiality: The price of reply. Occasional Papers
in Systemic Linguistics 6: 147–184.
Martin, J.R. 1993a. Genre and literacy: Modelling context in educational linguistics. Annual Review of
Applied Linguistics 13: 141–172.
Martin, J.R. 1993b. Life as a noun: Arresting the universe in science and humanities. In M.A.K. Halliday
and J.R. Martin (eds.), 242–293.
Martin, J.R. 1996a. Types of structure: Deconstructing notions of constituency in clause and text. In Edu-
ard Hovy and Donia Scott (eds.), Burning issues in discourse: A multidisciplinary perspective. Heidelberg:
Springer. 39–66.
Martin, J.R. 1996b. Metalinguistic diversity: The case from case. In Ruqaiya Hasan, Carmel Cloran and
David Butt (eds.), Functional descriptions: Theory into practice. Amsterdam: Benjamins. 323–375.
Martin, J.R. 1996c. Transitivity in Tagalog: A functional interpretation of case. In Christopher Butler,
Margaret Berry, Robin Fawcett and Guowen Huang (eds.), Meaning and form: Systemic functional inter-
pretations. Norwood, NJ: Ablex. 229–296.
Martin, J.R. 1997. Analysing genre: Functional parameters. In Fran Christie and J.R. Martin (eds.), Genre
and institutions: Social processes in the workplace and school. London: Cassell. 3–39.
Martin, J.R. 1998. Linguistics and the consumer: The practice of theory. Linguistics and Education 9(4):
411–448.
Martin, J.R. 1999a. Modelling context: A crooked path of progress in contextual linguistics. In Mohsen
Ghadessy (ed.), Text and context in functional linguistics. Amsterdam: John Benjamins Publishers.
25–61.
Martin, J.R. 1999b. Grace: The logogenesis of freedom. Discourse Studies 1(1): 29–56.
Martin, J.R. 2000. Beyond exchange: Appraisal systems in English. In Susan Hunston and Geoff Thomp-
son (eds.), 142–175.
Martin, J.R. 2001. A context for genre: Modelling social processes in functional linguistics. In Robert J.
Stainton and Jessica de Villiers (eds.), Communication in linguistics. Toronto: GREF (Collection Theoria).
1–41.
Martin, J.R. 2002. Blessed are the peacemakers: Reconciliation and evaluation. In Christopher N. Candlin
(ed.), Research and practice in professional discourse. Hong Kong: City University of Hong Kong Press.
187–227.
Martin, J.R. 2003. Making history: Grammar for interpretation. In J.R. Martin and Ruth Wodak (eds.),
Re/reading the past: Critical and functional perspectives on time and value. Amsterdam and Philadelphia:
Benjamins. 19–57.
Martin, J.R. 2004a. Positive discourse analysis: Power, solidarity and change. Revista Canaria de Estudios
Ingleses 49: 179–200.
Martin, J.R. 2004b. Metafunctional profile of the grammar of Tagalog. In Alice Caffarel, J.R. Martin and
Christian M.I.M. Matthiessen (eds.), 255–304.
Martin, J.R. 2006. Genre, ideology and intertextuality: A systemic functional perspective. Linguistics and
the Human Sciences 2(2): 275–298.
Martin, J.R. 2007. Multimodality: Some issues. Plenary address at the semiotic margins: Reclaiming meaning
workshop, 10–12 December, The University of Sydney.
Martin, J.R. 2008. Innocence: Realization, instantiation and individuation in a Botswanan Town. In
Ahmar Mahboob and Naomi Knight (eds.), Questioning linguistics. Newcastle: Cambridge Scholars
Publishing. 32–76.
Martin, J.R. 2009. Discourse studies. In M.A.K. Halliday and Jonathan J. Wester (eds.), 154–165.
Martin, J.R. 2010. Semantic variation: Modelling realisation, instantiation and individuation in social semi-
osis. In Monika Bednarek and J.R. Martin (eds.), 1–34.
Martin, J.R. 2013a. Systemic functional grammar: A next step into the theory – axial relations. (Chinese transla-
tion and extensions by Wang Pin and Zhu Yongsheng). Beijing: Higher Education Press.
Martin, J.R. (ed.). 2013b. Interviews with M.A.K. Halliday. London: Bloomsbury Academic.

494
 References

Martin, J.R. 2013c. Modelling context: Matter as meaning. In Carlos Gouveia and Marta Filipe Alexandre
(eds.), Languages, metalanguages, modalities, cultures: Functional and socio-discursive perspectives. Lisbon: BonD
and ILTEC. 10–64.
Martin, J.R. 2014. Evolving systemic functional linguistics: Beyond the clause. Functional Linguistics 1(3).
Martin, J.R. 2015a. Cohesion and texture. In Deborah Schiffrin, Deborah Tannen and Heidi E. Hamilton
(eds.), The handbook of discourse analysis. 2nd edition. Oxford: Wiley Blackwell. 61–81.
Martin, J.R. 2015b. Halliday the grammarian: Axial foundations. In Jonathan J. Wester (ed.), 257–290.
Martin, J.R. 2017. Revisiting field: Specialized knowledge in secondary school science and humanities
discourse. Onomázein Número especial LSF y TCL sobre educación y conocimiento. 111–148.
Martin, J.R. 2020. Ideational semiosis: A tri-stratal perspective on grammatical metaphor. D.E.L.T.A.,
36–3: 1–27. 2020360304.
Martin, J.R., Y.J. Doran and Giacomo Figueredo (eds.). 2020. Systemic functional language description: Making
meaning matter. New York and London: Routledge.
Martin, J.R., Karl Maton and Y.J. Doran (eds.). 2020a. Accessing academic discourse: Systemic functional linguis-
tics and legitimation code theory. London: Routledge.
Martin, J.R., Karl Maton and Y.J. Doran. 2020b. Academic discourse: An inter-disciplinary dialogue. In
J.R. Martin, Karl Maton and Y.J. Doran (eds.), Accessing academic discourse: Systemic functional lin-
guistics and legitimation code theory. London: Routledge. 1–31.
Martin, J.R. and Christian M.I.M. Matthiessen. 1991. Systemic typology and topology. In Frances Christie
(ed.), Literacy in social processes: Papers from the Inaugural Australian Systemic Functional Linguistics Conference,
Deakin University, January. Darwin: Centre for Studies of Language in Education, Northern Territory
University. 345–383. Reprinted in J.R. Martin (2010), SFL theory. Volume 1 in the collected works
of J.R. Martin, edited by Wang Zhenhua. Shanghai: Shanghai Jiao Tong University Press. 167–215.
Martin, J.R. and Christian M.I.M. Matthiessen. 2011. Interview with Prof. J.R. Martin and Prof.
Christian Matthiessen. Profs. James Martin and Christian Matthiessen sat down to talk about
genre-based teaching with Andrea. Available from http://blog.tesol.org/interview-with-professors-
james-martin-and-christian-matthiessen/
Martin, J.R. and Christian M.I.M. Matthiessen. 2014. Modelling and mentoring: Teaching and learning
from home through school. In Ahmar Mahboob and Leslie Barratt (eds.), English in a multilingual con-
text: Language variation and education. Berlin: Springer Verlag. 137–163.
Martin, J.R., Christian M.I.M. Matthiessen and Clare Painter. 1997. Working with functional grammar. Lon-
don: Edward Arnold.
Martin, J.R., Christian M.I.M. Matthiessen and Clare Painter. 2010. Deploying functional grammar. Exten-
sively revised, new edition of 1997 edition. Shanghai: Commercial Press.
Martin, J.R., Beatriz Quiroz and Giacomo Figueredo. 2021. Interpersonal grammar: Systemic functional linguis-
tic theory and description. Cambridge: Cambridge University Press.
Martin, J.R., Beatriz Quiroz, Wang Pin and Zhu Yongsheng. in press. Systemic functional grammar: Another
step into the theory: Grammatical description. Beijing: Higher Education Press.
Martin, J.R. and David Rose. 2003. Working with discourse: Meaning beyond the clause. London and New
York: Continuum.
Martin, J.R. and David Rose. 2005. Designing literacy pedagogy: Scaffolding democracy in the classroom.
In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan J. Wester (eds.), 251–280.
Martin, J.R. and David Rose. 2007. Working with discourse: Meaning beyond the clause. 2nd edition. London:
Continuum.
Martin, J.R. and David Rose. 2008. Genre relations: Mapping culture. London and Oakville: Equinox.
Martin, J.R. and Robert Veel (ed.). 1998. Reading science: Critical and functional perspectives on discourses of
science. London: Routledge.
Martin, J.R. and Peter R.R. White. 2005. The language of evaluation: Appraisal in English. London and New
York: Palgrave Macmillan.
Martin, J.R. and Ruth Wodak (eds.). 2003. Re/Reading the past: Critical and functional perspectives on time and
value (Discourse approaches to politics, society and culture, 8). Amsterdam and Philadelphia: Benjamins.
Martin, J.R. and Michele Zappavigna. 2019. Embodied meaning: A systemic functional perspective on
paralanguage. Functional Linguistics 6(1): 1–33.
Martin, J.R., Michele Zappavigna, Paul Dwyer and Chris Cléirigh. 2013. Users in uses of language:
Embodied identity in youth justice conferencing. Text and Talk 33(4–5): 467–494.
Martin, Samuel E. 1992. A reference grammar of Korean: A complete guide to the grammar and history of the Korean
language. Rutland, Vermont and Tokyo: Charles E. Tuttle.

495
 References

Martinec, Radan. 1995. Hierarchy of rhythm in English speech. Ph.D. thesis. The University of Sydney, Sydney.
Martinec, Radan. 2004. Gestures that co‐occur with speech as a systematic resource: The realization of
experiential meanings in indexes. Social Semiotics 14(2): 193–213.
Martinec, Radan. 2005. Topics in multimodality. In Ruqaiya Hasan, Christian M.I.M. Matthiessen and
Jonathan J. Wester (eds.), 157–181.
Martinec, Radan. 2018. Linguistic rhythm and its meaning: Rhythm waves and semantics fields. Linguistics
and the Human Sciences 14(1–2): 70–98.
Martínez-Insua, Ana Elina. 2013. There-constructions as a choice for coherence in the recent history of
English. In Lise Fontaine, Tom Bartlett and Gerard O’Grady (eds.), 207–225.
Mason, Ian. 2003. Text parameters in translation: Transitivity and institutional cultures. In Eva Haji-
cova, Peter Sgall, Zuzana Jettmarova, Annely Rothkegel, Dorothee Rothfuß-Bastian and Heidrun
Gerzymisch-Arbogast (eds.), Textologie und Translation (Jahrbuch Übersetzen und Dolmetschen 4/2).
Tübingen: Narr. Reprinted in Lawrence Venuti (eds.), 2004, The translation studies reader. 2nd edition.
London: Routledge. 470–481.
Mathers, Margaret E. 2006. Aspects of language in children with ADHD: Applying functional analyses to
explore language use. Journal of Attention Disorders 9(3): 523–533.
Mathesius, Vilém. 1911. O potenciálnosti jevů jazykových. (On the potentiality of the language phenom-
enon). Reprinted in Josef Vachek, 1964, 1–32.
Mathesius, Vilém. 1928. On linguistic characterology with illustrations from modern English. Actes du Ier
Congrès International des Linguistes, à la Haye, du 10–15 Avril. 56–63.
Mathesius, Vilém. 1975. A functional analysis of present day English on a general linguistic basis, edited by Josef
Vachek. The Hague: Mouton.
Matielo, Rafael, Maria Lúcia Barbosa de Vasconcellos and Elaine Espindola Baldissera. 2015. Subtitling
words or omitting worlds? A metafunctionally-oriented analysis. Revista de Estudos da Linguagem, Belo
Horizonte 23(2): 363–388.
Maton, Karl and Yaegan J. Doran. 2017. Systemic functional linguistics and code theory. In Tom Bartlett
and Gerard O’Grady (eds.), Chap. 38. 605–618.
Matthews, Peter H. 1966. The concept of rank in ‘neo-Firthian’ grammar. Journal of Linguistics 2: 101–110.
Matthiessen, Christian M.I.M. 1983. Choosing primary tense in English. Studies in Language 7(3): 369–430.
Matthiessen, Christian M.I.M. 1984. Choosing tense in English. USC/ISI Report: ISI/PR. 84–143.
Matthiessen, Christian M.I.M. 1987a. Notes on Akan lexicogrammar: A systemic interpretation. To appear
in Kazuhiro Teruya, Kaela Zhang and Diana Slade (eds.), forthcoming, Description of English and Akan.
Volume 3 of collected works of Christian. Sheffield and Bristol: Equinox.
Matthiessen, Christian M.I.M. 1987b. Notes on the organization of the environment of a text generation
grammar. In Gerard Kempen (ed.), Natural language generation. Dordrecht: Martinus Nijhof. 253–278.
Matthiessen, Christian M.I.M. 1987c. Notes on Akan phonology: A systemic interpretation. To appear
in Kazuhiro Teruya, Kaela Zhang and Diana Slade (eds.), forthcoming, Description of English and Akan.
Volume 3 of collected works of Christian. Sheffield and Bristol: Equinox.
Matthiessen, Christian M.I.M. 1988a. Semantics for a systemic grammar: The chooser and inquiry frame-
work. In Michael J. Cummings and William S. Greaves James D. Benson (eds.), Linguistics in a systemic
perspective. Amsterdam: Benjamins. 221–242.
Matthiessen, Christian M.I.M. 1988b. Representational issues in systemic functional grammar. In James D.
Benson and William S. Greaves (eds.), Systemic functional perspectives on discourse. Norwood, NJ: Ablex.
136–175.
Matthiessen, Christian M.I.M. 1990. Two approaches to semantic interfaces in text generation. COL-
ING-90, Helsinki, August. 322–329.
Matthiessen, Christian M.I.M. 1991a. Lexico(grammatical) choice in text-generation. In Cécile Paris,
William Swartout and William C. Mann (eds.), Natural language generation in artificial intelligence and
computational linguistics. Dordrecht: Kluwer. 229–247.
Matthiessen, Christian M.I.M. 1991b. Language on language: The grammar of semiosis. Social Semiotics
1(2): 69–111.
Matthiessen, Christian M.I.M. 1991c. Metafunctional complementarity and resonance in syntagmatic
organization. MS. In Matthiessen, Christian M.I.M. in press, Systemic functional linguistics Part 2. Volume
2 of Christian M.I.M. Matthiessen’s Collected works, edited by Kazuhiro Teruya and Diana Slade.
Sheffield and Bristol: Equinox.
Matthiessen, Christian M.I.M. 1992. Interpreting the textual metafunction. In Martin Davies and Louise
Ravelli (eds.), Advances in systemic linguistics: Recent theory and practice. London: Pinter. 37–82.

496
 References

Matthiessen, Christian M.I.M. 1993a. Register in the round: Diversity in a unified theory of register analy-
sis. Mohsen Ghadessy (ed.), Register analysis: Theory and practice. London: Pinter. 221–292.
Matthiessen, Christian M.I.M. 1993b. Instantial systems and logogenesis. Written version of Paper presented at
the Third Chinese Systemic-functional symposium, Hangzhou University, Hangzhou, June 17–20. In Chris-
tian M.I.M. Matthiessen, forthcoming d.
Matthiessen, Christian M.I.M. 1993c. The object of study in cognitive science in relation to its construal
and enactment in language. Language as Cultural Dynamic, edited by M.A.K. Halliday, special issue of
Cultural Dynamics 6(1–2): 187–243.
Matthiessen, Christian M.I.M. 1995a. Lexicogrammatical cartography: English systems. Tokyo: International
Language Sciences Publishers.
Matthiessen, Christian M.I.M. 1995b. Theme as an enabling resource in ideational knowledge construction.
In Mohsen Ghadessy (ed.), Thematic developments in English texts. London and New York: Pinter. 20–55.
Matthiessen, Christian M.I.M. 1995c. Fuzziness construed in language: A linguistic perspective. Proceedings
of FUZZ/IEEE, Yokohama, March. Yokohama. 1871–1878.
Matthiessen, Christian M.I.M. 1996. Tense in English seen through systemic-functional theory. In Chris-
topher Butler, Margaret Berry, Robin Fawcett and Guowen Huang (eds.), Meaning and form: Systemic
functional interpretations. Norwood, NJ: Ablex. 431–498.
Matthiessen, Christian M.I.M. 1999. The system of transitivity: An exploratory study of text-based pro-
files. Functions of Language 6(1): 1–51.
Matthiessen, Christian M.I.M. 2000. Organising principles and expansion of the systemic functional
model: A conversation with Christian Matthiessen. Revista Canaria des Estudios Ingleses 40: 251–256.
Matthiessen, Christian M.I.M. 2001. The environments of translation in exploring translation and multi-
lingual text production. In Erich Steiner and Colin Yallop (eds.), Exploring translation and multilingual text
production. Berlin: De Gruyter Mouton. 41–124.
Matthiessen, Christian M.I.M. 2002a. Combining clauses into clause complexes: A multi-faceted view.
In Joan Bybee and Michael Noonan (eds.), Complex sentences in grammar and discourse: Essays in honor of
Sandra A. Thompson. Amsterdam: Benjamins. 237–322.
Matthiessen, Christian M.I.M. 2002b. Lexicogrammar in discourse development: Logogenetic patterns
of wording. In Guowen Huang and Zongyan Wang (eds.), Discourse and language functions. Shanghai:
Foreign Language Teaching and Research Press. 91–127.
Matthiessen, Christian M.I.M. 2003. Language, social life and discursive maps. Plenary for Australian Sys-
temic Functional Linguistics Conference, Adelaide.
Matthiessen, Christian M.I.M. 2004a. Descriptive motifs and generalizations. In Alice Caffarel, J.R. Martin
and Christian Matthiessen (eds.), 537–673.
Matthiessen, Christian M.I.M. 2004b. The evolution of language: A systemic functional exploration of
phylogenetic phases. In Geoff Williams and Annabelle Lukin (eds.), 45–90.
Matthiessen, Christian M.I.M. 2006. Frequency profiles of some basic grammatical systems: An interim
report. In Susan Hunston and Geoff Thompson (eds.), 103–142.
Matthiessen, Christian M.I.M. 2007a. Lexicogrammar in systemic functional linguistics: Descriptive and
theoretical developments in the ‘IFG’ tradition since the 1970s. In Ruqaiya Hasan, Christian M.I.M.
Matthiessen and Jonathan J. Wester (eds.), 765–858. Reprinted in Christian M.I.M. Matthiessen,
2021b, 14–88.
Matthiessen, Christian M.I.M. 2007b. The ‘architecture’ of language according to systemic functional
theory: Developments since the 1970s. In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan
J. Wester (eds.), 505–561. Reprinted in Christian M.I.M. Matthiessen, 2021b, 89–134.
Matthiessen, Christian M.I.M. 2007c [2006]. The lexicogrammar of emotion and attitude in English. Elec-
tronic Proceedings of the Third International Congress on English Grammar (ICEG 3), Sona College, Salem,
Tamil Nadu, India, January 23–27.
Matthiessen, Christian M.I.M. 2009a. Multisemiotic and context-based regi11ster typology: Registerial
variation in the complementarity of semiotic systems. In Eija Ventola and Arsenio Jesús Moya-Guijarro
(eds.), The world shown and the world told. Basingstoke: Palgrave Macmillan. 11–38.
Matthiessen, Christian M.I.M. 2009b. Meaning in the making: Meaning potential emerging from acts of
meaning. In Anniversary Issue of Language Learning 59 (Supplement 1): 211–235.
Matthiessen, Christian M.I.M. 2009c. Ideas and new direction. In M.A.K. Halliday and Jonathan J. Wester
(eds.), 12–58. Reprinted in Christian M.I.M. Matthiessen, 2021b, 135–183.
Matthiessen, Christian M.I.M. 2010. Systemic functional linguistics developing. Annual Review of Func-
tional Linguistics 2: 8–63. Reprinted in Christian M.I.M. Matthiessen, 2021b, 184–221.

497
 References

Matthiessen, Christian M.I.M. 2012. Systemic functional linguistics as appliable linguistics: Social account-
ability and critical approaches. D.E.L.T.A. (Revista de Documentação de Estudos em Lingüística Teórica e
Aplicada) 28: 437–471.
Matthiessen, Christian M.I.M. 2013. Applying systemic functional linguistics in healthcare contexts. Text
and Talk 33(4–5): 437–466.
Matthiessen, Christian M.I.M. 2014a. Appliable discourse analysis. Fang Yan and Jonathan J. Webster
(eds.), Developing systemic functional linguistics: Theory and application. London: Equinox. 135–205.
Matthiessen, Christian M.I.M. 2014b. Choice in translation: Metafunctional consideration. In Kerstin
Kunz, Elke Teich, Silvia Hansen-Schirra, Stella Neumann and Peggy Daut (eds.), 271–333.
Matthiessen, Christian M.I.M. 2014c. Extending the description of process type in delicacy: Verb classes.
Functions of Language 21(2): 139–175.
Matthiessen, Christian M.I.M. 2015a. Halliday’s conception of language as a probabilistic system. In Jona-
than J. Webster (ed.), The Bloomsbury companion to M.A.K. Halliday. London and New York: Blooms-
bury Academic. 203–241.
Matthiessen, Christian M.I.M. 2015b. Register in the round: Registerial cartography. Functional Linguistics
2(9): 1–48.
Matthiessen, Christian M.I.M. 2015c. Modelling context and register: The long-term project of registerial
cartography. Estudos sistêmico-funcionais: Desdobramentos e interfaces. Revista Letras, Santa Maria 25(50),
january/june. 15–90.
Matthiessen, Christian M.I.M. 2015d. Systemic functional morphology: The lexicogrammar of the word.
In Edson Rosa de Souza (ed.), Estudos de descrição funcionalista: Objetos e abordagens. LINCOM Studies in
Theoretical Linguistics 55. München: LINCOM. 150–199.
Matthiessen, Christian M.I.M. 2015e. The language of space: Semiotic resources for construing our expe-
rience of space. Japanese Journal of Systemic Functional Linguistics 8, May: 1–64.
Matthiessen, Christian M.I.M. 2015f. Subliminal construal of world order clause by clause: Hierarchy of
control in Noah’s Ark. Linguistics and the Human Sciences 11 (1–2): 250–283.
Matthiessen, Christian M.I.M. 2015g. Halliday on language. In Jonathan J. Webster (ed.), The Bloomsbury
companion to M.A.K. Halliday. London and New York: Bloomsbury Academic. 137–202. Reprinted in
Christian M.I.M. Matthiessen, 2021b, 222–287.
Matthiessen, Christian M.I.M. 2017. Language use in a social semiotic perspective. In Anne Barron,
Gu Yueguo and Gerard Steen (eds.), The Routledge handbook of pragmatics. London: Routledge.
459–489.
Matthiessen, Christian M.I.M. 2018a. Transitivity in systemic functional linguistics: Achievements and
challenges. In Sara Regina Scotta Cabral and Leila Barbara (eds.), Estudos de transitividade em linguística
sistêmico-funcional. (Transitivity studies in systemic functional linguistics). Santa Maria, Rio Grande do
Sul, Brasil: Programa de pós-graduação em letras – PPGL UFSM. 14–108.
Matthiessen, Christian M.I.M. 2018b. The notion of a multilingual meaning potential: A systemic explora-
tion. In Akila Baklouti and Lise Fontaine (eds.), 90–120.
Matthiessen, Christian M.I.M. 2019. Register in systemic functional linguistics. Register Studies 1(1):
10–41.
Matthiessen, Christian M.I.M. 2021a. The architecture of phonology according to Systemic Functional
Linguistics. In Christian M.I.M. Matthiessen (ed.), 2021b, 288–338.
Matthiessen, Christian M.I.M. 2021b. Systemic functional linguistics Part 1. Volume 1 of collected works of
Christian M.I.M. Matthiessen, edited by Kazuhiro Teruya, Diana Slade and Wu Canzhong. Sheffield
and Bristol: Equinox.
Matthiessen, Christian M.I.M. 2021c. Register cartography and Giovanni Parodi’s research: Registerial
profiles of school subjects and university disciplines. Special Issue of Revista Signos in Honour of the Memory
of Giovanni Parodi 54(107): 799–841.
Matthiessen, Christian M.I.M. 2022. Theory: A resource for engaging with language. In Kazuhiro Teruya
(ed.), Imi ga yokku wakaru yooni naru tame no gengogaku: Taikei kinoo gengogaku eno shootai (Linguistics for a
better understanding of meaning: An introduction to systemic functional linguistics). Tokyo: Koroshio
Publishers. 1–49.
Matthiessen, Christian M.I.M. forthcoming a. The architecture of language according to systemic functional lin-
guistics. Book MS.
Matthiessen, Christian M.I.M. forthcoming b. The system of logico-semantic relations. Book MS.
Matthiessen, Christian M.I.M. forthcoming c. Rhetorical system and structure theory: The semantic system of
rhetorical relations. First volume in a series, in discussion with Routledge.

498
 References

Matthiessen, Christian M.I.M. forthcoming d. Systemic functional linguistics Part 2. Volume 2 of collected
works of Christian M.I.M. Matthiessen, edited by Kazuhiro Teruya and Diana Slade. Sheffield and
Bristol: Equinox.
Matthiessen, Christian M.I.M. forthcoming e. The architecture of language according to SFL: Some reflec-
tions on implications for neurosemiotics. Systemic functional linguistics, Part 2. Volume 2 of collected
works of Christian M.I.M. Matthiessen, edited by Kazuhiro Teruya and Diana Slade. Sheffield and
Bristol: Equinox.
Matthiessen, Christian M.I.M., Jorge Hita Arús and Kazuhiro Teruya. 2021. Translations of representations
of moving and saying from English into Spanish. Word 67(2): 188–207.
Matthiessen, Christian M.I.M. and John A. Bateman. 1991. Systemic linguistics and text generation: Experiences
from Japanese and English. London: Frances Pinter.
Matthiessen, Christian M.I.M., Wang Bo, Yuanyi Ma and Isaac N. Mwinlaaru. 2022. Systemic functional
insights on language and linguistics. Berlin: Springer.
Matthiessen, Christian M.I.M. and M.A.K. Halliday. 2009. Systemic functional grammar: A first step into
the theory. Bilingual edition, with introduction by Huang Guowen. Beijing: Higher Education
Press.
Matthiessen, Christian M.I.M. and Abhishek Kumar Kashyap. 2014. The construal of space in different
registers: An exploratory study. Language Sciences 45: 1–27.
Matthiessen, Christian M.I.M., Annabelle Lukin, David G. Butt, Chris Cleirigh and Christopher Nesbitt.
2005. Welcome to Pizza Hut: A case study of multistratal analysis. Australian Review of Applied Linguistics
19: 123–150.
Matthiessen, Christian M.I.M. and Christopher Nesbitt. 1996. On the idea of theory-neutral descrip-
tions. In Ruqaiya Hasan, Carmel Cloran and David Butt (eds.), Functional descriptions: Theory in practice.
Amsterdam: Benjamins. 39–85.
Matthiessen, Christian M.I.M. and Diana Slade. 2011. Analysing conversation. In Ruth Wodak, Barbara
Johnstone and Paul Kerswill (eds.), The SAGE handbook of sociolinguistics. Los Angeles, London, New
Delhi, Singapore and Washington DC: SAGE. 375–395.
Matthiessen, Christian M.I.M., Diana Slade and Mary Macken. 1992. Language in context: A new model
for evaluating student writing. Linguistics and Education 4(2): 173–195.
Matthiessen, Christian M.I.M. and Kazuhiro Teruya. 2015a. Registerial hybridity: Indeterminacy among
fields of activity. In Donna Miller and Paul Bayley (eds.), 205–239.
Matthiessen, Christian M.I.M. and Kazuhiro Teruya. 2015b. Grammatical realization of rhetorical relations
in different registers. Word 61(3): 232–281.
Matthiessen, Christian M.I.M. and Kazuhiro Teruya. 2015c. Rhetorical relations and their lexicogram-
matical realizations in different registers. Word 61(3): 232–281.
Matthiessen, Christian M.I.M., Kazuhiro Teruya and Marvin Lam. 2010. Key terms in systemic functional
linguistics. London and New York: Continuum.
Matthiessen, Christian M.I.M., Kazuhiro Teruya and Marvin Lam. 2016. Xitong gongneng yuyuanxue zhong
de guanjian shuyu. (Key terms in systemic functional linguistics), translated by Huang Guowen and Chen
Yumin. Beijing: Foreign language teaching and research press.
Matthiessen, Christian M.I.M., Kazuhiro Teruya and Canzhong Wu. 2008. Multilingual studies as a multi-
dimensional space of interconnected language studies. In Jonathan J. Wester (ed.), 146–221.
Matthiessen, Christian M.I.M. and Sandra A. Thompson. 1988. The structure of discourse and ‘subor-
dination’. In John Haiman and Sandra A. Thompson (eds.), Clause combining in grammar and discourse.
Amsterdam: Benjamins. 275–329.
McCabe, Anne. 2017. Systemic functional linguistics and language teaching. In Tom Bartlett and Gerard
O’Grady (eds.), 591–604.
McCabe, Anne. 2021. A functional linguistic perspective on developing language. London: Routledge.
McCabe, Anne, Mick O’Donnell and Rachel Whittaker (eds.). 2007. Advances in language and education.
London and New York: Continuum.
McCawley, James D. 1998. The syntactic phenomena of English. 2nd edition. Chicago: The University of
Chicago Press.
McDonald, Daniel. 2016. Linguistic change in an online support group. Ph.D. thesis. University of Melbourne,
Melbourne.
McDonald, Daniel and Robyn Woodward-Kron. 2016. Member roles and identities in online support
groups: Perspectives from corpus and systemic functional linguistics. Discourse and Communication 10(2):
157–175.

499
 References

McDonald, Edward. 2017. Form and function in groups. In Tom Bartlett and Gerard O’Grady (eds.),
Chap. 15. 251–266.
McDonald, Skye. 1993. Pragmatic language skills after closed head injury: Ability to meet the informa-
tional needs of the listener. Brain and Language 44(1): 28–46.
McEnery, Tony and Andrew Hardie. 2012. Corpus linguistics. (Cambridge Textbooks in Linguistics). Cam-
bridge: Cambridge University Press.
McEnery, Tony and Andrew Hardie. 2013. The history of corpus linguistics. In Keith Allan (ed.), The
Oxford handbook of the history of linguistics. Oxford: Oxford University Press.
McGregor, William B. 1990. A functional grammar of Gooniyandi. Amsterdam and Philadelphia: John
Benjamins.
McQuown, Norman A., Gregory Bateson, Ray L. Birdwhistell, Henry W. Brosin and Charles F. Hockett
(eds.). 1971. The natural history of an interview. Microfilm collection of manuscripts on cultural anthropol-
ogy, 15th series. Chicago: University of Chicago Joseph Regenstein Library, Dept of photoduplication.
Mehler, Alexander, Serge Sharoff and Marina Santini (eds.). 2010. Genres on the web: Computational models
and empirical studies. Heidelberg: Springer.
Melrose, Robin. 2006. Protolanguage, mirror neurons, and the front-heavy brain: Explorations in the evo-
lution and functional organization of language. Linguistics and the Human Sciences 2(1).
Mentis, Michelle and Carol A. Prutting. 1987. Cohesion in the discourse of normal and head injured
adults. Journal of Speech and Hearing Research 30: 88–98.
Metz, Christian. 1966. La grande syntagmatique du film narratif. Communications, 8, 1966. Recherches sémi-
ologiques: L’analyse structurale du récit. 120–124.
Michel, Jean-Baptiste et al. 2011. Quantitative analysis of culture using millions of digitized books. Science
331: 176–182.
Mickan, Peter. 2019. Language and education: Learning how to mean. In Geoff Thompson, Wendy L.
Bowcher, Lise Fontaine and David Schöntal (eds.), 537–560.
Miller, Carolyn R. 1984. Genre as social action. Quarterly Journal of Speech 70: 151–167.
Miller, Donna. 2017. Language as verbal art. In Tom Bartlett and Gerard O’Grady (eds.), 506–519.
Miller, Donna. 2019. Language and literature. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine and
David Schöntal (eds.), 690–714.
Miller, Donna and Monica Turci (eds.). 2007. Language and verbal art revisited: Linguistic approaches to the study
of literature. London and Oakville: Equinox.
Mishler, Elliot G. 1984. The discourse of medicine: Dialectics of medical interviews. Norwood: Ablex.
Mishler, Elliot G. 1986. Research interviewing: Context and narrative. Cambridge, MA: Harvard University
Press.
Mitchell, T.F. 1957. The language of buying and selling in Cyrenaica: a situational statement. Hesperis
44: 31–71. Reprinted in T.F. Mitchell, Principles of neo-Firthian linguistics. London: Longman, 1975.
167–200.
Miyajima, Tatsuo. 1994. Goiron kenkyuu. (Studies of lexicology). Tokyo: Mugi Shobo.
Mock, Carol C. 1969. The grammatical units of the Nzema language: A systemic analysis. Unpublished Ph.D.
thesis. University of London, London.
Mohan, Bernhard A. 1986. Language and content. Reading, MA: Addison-Wesley.
Mok, Zaneta and Nicole Müller. 2013. Staging casual conversations for people with dementia. Dementia
13(6): 834–853.
Monaghan, James. 1979. The Neo-Firthian tradition and its contribution to general linguistics. Tübingen:
Niemeyer.
Moore, Alison Rotha. 2004. The discursive construction of treatment decisions in the management of HIV disease.
Ph.D. Thesis. Macquarie University, Sydney.
Moore, Alison Rotha. 2005. Modelling Agency in HIV Decision-making. Australian Review of Applied
Linguistics 19: 103–122.
Moore, Alison Rotha. 2017. Register analysis in systemic functional linguistics. In Bartlett, Tom and
Gerard O’Grady (eds.), Chap. 26. 418–437.
Moore, Alison Rotha. 2019. Language and medicine. In Geoff Thompson, Wendy L. Bowcher, Lise Fon-
taine and David Schöntal (eds.), Chap. 25. 651–689.
Moore, Alison Rotha. 2020. Progress and tensions in modelling register as a semantic configuration. Lan-
guage, Context and Text 2(1): 22–58.
Moore, Alison Rotha. 2021. Pills, life, agency: HIV treatment decisions as language in social context. Sheffield:
Equinox.

500
 References

Moore, Alison Rotha, Christopher N. Candlin and Gunther A. Plum. 2001. Making sense of HIV-related
viral load: One expert or two? Culture Health and Sexuality 3: 229–250.
Morley, David G. 2000. Syntax in functional grammar: An introduction to lexicogrammar in systemic linguistics.
London and New York: Continuum.
Morris, Charles W. 1938. Foundations of the theory of signs. In Otto Neurath, Rudolf Carnap and Charles
Morris (eds.), International encyclopedia of unified science 1(2), 1–59. Chicago, IL: The University of Chi-
cago Press.
Morris, Charles W. 1946. Signs, language and behavior. New York: Prentice-Hall.
Mortensen, Lynne. 1992. A transitivity analysis of discourse in dementia of the Alzheimer’s type. Journal of
Neurolinguistics 7(4): 309–321.
Mortensen, Lynne. 2005. Written discourse and acquired brain impairment: Evaluation of structural and
semantic features of personal letters from a systemic functional linguistic perspective. Clinical Linguistics
and Phonetics 19(3): 227–247.
Moya-Guijarro, A. Jesús and Eija Ventola (eds.). 2022. A multimodal approach to challenging gender stereotypes
in children’s picture books. London: Routledge.
Müller, Nicole and Zaneta Mok. 2012. Applying systemic functional linguistics to conversations with
dementia: The linguistic construction of relationships between participants. Seminars in Speech and Lan-
guage 33(1): 5–15.
Müller, Nicole and Brent T. Wilson. 2008. Collaborative role construction in a conversation with dementia:
An application of systemic functional linguistics. Clinical Linguistics and Phonetics 22(10–11): 767–774.
Munday, Jeremy. 2000. Using systemic functional linguistics as an aid to translation between Spanish and
English: Maintaining the thematic development of the ST. Revista Canaria de Estudios Ingleses 40: 37–58.
Munday, Jeremy. 2012. Evaluation in translation: Critical points of translator decision-making. London: Routledge.
Munday, Jeremy. 2016. Introducing translation studies: Theories and applications. 4th edition. London: Routledge.
Munday, Jeremy and Zhang Meifang (eds.). 2017. Discourse analysis in translation studies. Amsterdam: John
Benjamins.
Muntigl, Peter. 2004a. Modelling multiple semiotic systems: The case of gesture and speech. In Eija Ven-
tola, Cassily Charles and Martin Kaltenbacher (eds.), Perspectives on multimodality. Amsterdam: Ben-
jamins. 31–50.
Muntigl, Peter. 2004b. Narrative counselling: Social and linguistic processes of change. Amsterdam: Benjamins.
Muntigl, Peter. 2006. Macrogenre: A multiperspectival and multifunctional approach to social interaction.
Linguistics and the Human Sciences 2: 233–256.
Muntigl, Peter. 2007. A metapragmatic examination of therapist reformulations. In Wolfram Bublitz and
Axel Hübler (eds.), Metapragmatics in use. Amsterdam: Benjamins. 235–262.
Muntigl, Peter, Naomi Knight, A. Watkins, A. Horvath and L. Angus. 2013. Active retreating: Person-
centered practices to repair disaffiliation in therapy. Journal of Pragmatics 53: 1–20.
Murcia-Bielsa, Susana. 2000. The choice of directives expressions in English and Spanish instructions: A
semantic network. In Eija Ventola (ed.). Discourse and community: Doing functional linguistics. Language in
Performance 21. Tübingen: Gunter Narr Verlag. 117–146.
Mwinlaaru, Isaac Nuokyaa Ire. 2016. In Marquis Who’s Who (ed.), Who’s who in America 2016. 70th edi-
tion. Berkeley Heights, NJ: Marquis Who’s Who LLC. Credo.
Mwinlaaru, Isaac Nuokyaa-Ire. 2017. A systemic functional description of the grammar of dagaare. Ph.D. thesis.
The Hong Kong Polytechnic University, Hong Kong.
Mwinlaaru, Isaac N., Christian M.I.M. Matthiessen and Ernest Akerejola. 2018. A system-based typology
of MOOD in African languages. In Augustine Agwuele and Adam Bodomo (eds.), Handbook of African
languages. London: Routledge. 93–117.
Mwinlaaru, Isaac N. and Winfred Wenhui Xuan. 2016. A survey of studies in systemic functional language
description and typology. Functional Linguistics 3(8). https://doi.org/10.1186/s40554-016-0030-4
Nagar, Revital and Jonathan Fine. 2013. Clause complex manifestation in depression. Text and Talk 33(4–
5): 595–615.
Nanri, Keizo. 1993. An attempt to synthesize two systemic contextual theories through the investigation of the process
of the evolution of the discourse semantic structure of the newspaper reporting article. Ph.D. thesis. The University
of Sydney, Sydney.
Neale, Amy. 2002. More delicate transitivity: Extending the process type system networks for English to include full
semantic classifications. Ph.D. thesis. Cardiff University, Cardiff.
Neale, Amy. 2006. Matching corpus data and system networks: Using corpora to modify and extend the
system networks for transitivity in English. In Susan Hunston and Geoff Thompson (eds.), 143–163.

501
 References

Neale, Amy. 2017. Transitivity in the Cardiff Grammar. In Bartlett, Tom and Gerard O’Grady (eds.), Chap.
12. 178–193.
Nekula, Marek. 1999. Vilém Mathesius J. Verschueren, J.-O. Östman, J. Blommaert and Ch. Bulcaen
(eds.), Handbook of pragmatics. Amsterdam/Philadelphia: John Benjamins Publishing Company. 1–14.
Nerlich, Brigitte. 1990. Change in language: Whitney, Bréal and Wegener. London and New York: Routledge.
Nesbitt, Christopher N. and Guenter Plum and. 1988. Probabilities in a systemic grammar: The clause
complex in English. In Robin P. Fawcett and David Young (eds.), New developments in systemic linguistics,
vol. 2: Theory and application. London: Frances Pinter. 6–39.
Nesi, Hilary and Sheena Gardner. 2012. Genres across the disciplines: Student writing in higher education. Cam-
bridge: Cambridge University Press.
Nettle, Daniel and Suzanne Romaine. 2000. Vanishing voices: The extinction of the world’s languages. Oxford:
Oxford University Press.
Neumann, Stella. 2003a. Textsorten und Übersetzen. Eine Korpusanalyse englischer und deutscher Reiseführer.
Frankfurt am main: Peter Lang Verlag. Reihe SABEST Band 3.
Neumann, Stella. 2003b. Exploitation of an SFL-annotated multilingual register corpus. In Anne Abeillé,
Silvia Hansen-Schirra and Hans Uszkoreit (eds.), Proceedings of the 4th international workshop on linguisti-
cally interpreted corpora (LINC-03), 13–14 April. Budapest: Association for Computational Linguistics.
85–92.
Neumann, Stella. 2013. Contrastive register variation: A quantitative approach to the comparison of English and
German. Trends in Linguistics. Studies and Monographs. Berlin, Boston: de Gruyter Mouton.
Neumann, Stella. 2014. Cross-linguistic register studies: Theoretical and methodological considerations.
Languages in Contrast 14(1): 35–57.
Neumann, Stella and Silvia Hansen-Schirra. 2005. The CroCo Project: Cross-linguistic corpora for the
investigation of explicitation in translations. Proceedings from the Corpus Linguistics Conference Series 1(1).
Available from www.corpus.bham.ac.uk/PCLC.
Neumann, Stella, Rebekah Wegener, Jennifer Fest, Paula Niemietz and Nicole Hutzen (eds.). 2017. Chal-
lenging boundaries: Systemic functional perspectives. Frankfurt: Peter Lang.
Ngo, Thu, Susan Hood, J.R. Martin, Clare Painter, Bradley A. Smith and Michele Zappavigna. 2021.
Modelling paralanguage using systemic functional semiotics: Theory and application. London: Bloomsbury.
Nord, Christiane. 2005. Text analysis in translation: Theory, methodology, and didactic application of a model for
translation-oriented text analysis. Amsterdam: Rodopi.
Norris, Sigrid. 2004. Analyzing multimodal interaction: A methodological framework. London: Routledge.
Norris, Sigrid. 2019. Systematically working with multimodal data: Research methods in multimodal discourse
analysis. Oxford: Wiley-Blackwell.
Norris, John M. and Lourdes Ortega (eds.). 2006. Synthesizing research on language learning and teaching.
Amsterdam: Benjamins.
Ochi, Ayako. 2013. A text-based study of the interpersonal grammar of modern Japanese: mood, modality and
evidentiality. Ph.D. thesis. Macquarie University, Sydney.
Ochs, Elinor, Emanuel A. Schegloff and Sandra A. Thompson (eds.). 1996. Interaction and grammar (Studies
in Interactional Sociolinguistics 13). Cambridge: Cambridge University Press.
O’Donnell, Michael. 1990. A dynamic model of exchange. Word 41(3): 293–328.
O’Donnell, Michael. 1994. Sentence analysis and generation: A systemic perspective. Ph.D. thesis. The Univer-
sity of Sydney, Sydney.
O’Donnell, Michael. 1999. Context in dynamic modelling. In Mohsen Ghadessy (ed.), Text and context in
functional linguistics. Amsterdam: Benjamins. 63–99.
O’Donnell, Michael. 2012. UAM CorpusTool: Version 2.8 user manual. http://www.corpustool.com/Docu-
mentation/UAMCorpusToolManualv28.pdf
O’Donnell, Michael. 2013. A dynamic view of choice in writing: Composition as text evolution. In Lise
Fontaine, Tom Bartlett and Gerard O’Grady (eds.), 247–268.
O’Donnell, Michael. 2017. Interactions between natural-language processing and systemic functional lin-
guistics. In Tom Bartlett and Gerard O’Grady (eds.), 561–574.
O’Donnell, Michael. 2019. Continuing issues in SFL. In Geoff Thompson, Wendy L. Bowcher, Lise Fon-
taine and David Schöntal (eds.), 204–229.
O’Donnell, Michael and John A. Bateman. 2005. SFL in computational contexts. In Ruqaiya Hasan,
Christian M.I.M. Matthiessen and Jonathan J. Wester (eds.), 343–382.
O’Donnell, Michael and Peter Sefton. 1995. Modelling telephonic interaction: A dynamic approach. Jour-
nal of Applied Linguistics 10(1): 63–78.

502
 References

O’Donnell, Mick. 2009. Resources and courses. In M.A.K. Halliday and Jonathan J. Webster (eds.),
216–228.
O’Donnell, Mick, Michele Zappavigna and Casey Whitelaw. 2008. A survey of process type classification
over difficult cases. In Jones Carys and Eija Ventola (eds.), 47–64.
Ogden, C.K. and I.A. Richards. 1923. The meaning of meaning. New York: Harcourt Brace and World.
O’Grady, Gerald. 2010. A grammar of spoken English discourse: The intonation of increments. London: Continuum.
O’Grady, Gerard. 2017. Intonation and systemic functional linguistics: The way forward. In Bartlett, Tom
and Gerard O’Grady (eds.), Chap. 10. 146–162.
O’Grady, Gerard. 2019. SFL and critical discourse analysis. In Geoff Thompson, Wendy L. Bowcher, Lise
Fontaine and David Schöntal (eds.). Chap. 18. 462–484.
O’Grady, Gerard, Tom Bartlett and Lise Fontaine. 2013. Choice in language: Applications in text analysis.
London: Equinox.
O’Halloran, Kay L. (ed.). 2004. Multimodal discourse analysis. London and New York: Continuum.
O’Halloran, Kay L. 2005. Mathematical discourse: Language, symbolism and visual images. London and New
York: Continuum.
O’Halloran, Kay L., K.L.E. Marissa and Sabine Tan. 2015. Multimodal semiosis and semiotics. In Jonathan
J. Wester (ed.), 386–411.
O’Halloran, Kay L., Gautam Pal and Minhao Jin. 2021. Multimodal approach to analysing big social and
news media data. Discourse, Context and Media, 40.
O’Halloran, Kay L., Sabine Tan and K.L.E. Marissa. 2015. Multimodal analysis for critical thinking. Learn-
ing, Media and Technology 42(2): 147–170.
O’Halloran, Kay L., Sabine Tan, Peter Wignell. 2019. SFL and multimodal discourse analysis. In Geoff
Thompson, Wendy L. Bowcher, Lise Fontaine and David Schöntal (eds.), 433–461.
O’Halloran, Kay L., Sabine Tan, Peter Wignell, John A Bateman, Duc-Son Pham, Michele Grossman and
Andrew Vande Moere. 2016. Interpreting text and image relations in violent extremist discourse: A
mixed methods approach for big data analytics. Terrorism and Political Violence 31(3): 454–474.
Okuda, Yasuo. 1985. Kotoba no kenkyuu: Josetsu (Studies of Language: An Introduction). Tokyo: Mugi Shobo.
Okuda, Yasuo and Taro Kokubun’ichi (eds.). 1974. Yomikata kyooiku no riron (Theory of teaching reading).
Tokyo: Mugi Shobo.
Olness, Gloria S. and Hanna K. Ulatowska. 2011. Personal narratives in aphasia: Coherence in the context
of use. Aphasiology 25: 1393–1413.
Ortega, Lourdes. 2015. Research synthesis. In Brian Paltridge and Aek Phakiti (eds.), Research methods in
applied linguistics: A practical guide. 2nd edition. London and New York: Bloomsbury Academic.
Ortega, Lourdes and Heidi Byrnes (eds.). 2008. The longitudinal study of advanced L2 capacities. New York
and London: Routledge.
Osgood, Charles E. 1964. Semantic differential technique in the comparative study of cultures. American
Anthropologist 66: 171–200.
Oteíza, Teresa. 2017. The appraisal framework and discourse analysis. In Tom Bartlett and Gerard O’Grady
(eds.), 457–473.
O’Toole, Michael. 1982. Structure, style and interpretation in the Russian short story. London and New Haven:
Yale University Press.
O’Toole, Michael. 1994. The language of displayed art. London: Leicester University Press (Pinter).
O’Toole, Michael. 2011. The language of displayed art. 2nd edition. London and New York: Routledge.
O’Toole, Michael. 2015. Halliday’s three functions and their interaction in the interpretation of painting
and music. In Jonathan J. Wester (ed.), 369–385.
Oshima, Makoto. 2004. Functions and meanings of the English language viewed from systemic functional grammar.
Tokyo: Eihoosha.
Oteíza, Teresa. 2017. The appraisal framework and discourse analysis. In Tom Bartlett and Gerard O’Grady
(eds.), 457–473.
Otto, Jespersen. 1933. Essentials of English grammar. London: Routledge.
Ouyang, Xiaoqing. 1986. Clause complex in Chinese. M.A. thesis. The University of Sydney, Sydney.
Ovadia, R. and Jonathan Fine. 1995. A functional analysis of intonation in Asperger’s syndrome. In Jurg
Siegfried (ed.), Therapeutic and everyday discourse as behavior change: Towards a micro-analysis in psychotherapy
process research. Norwood, N.J.: Ablex. 491–510.
Painter, Clare. 1984. Into the mother tongue: A case study in early language development. London: Frances
Pinter.
Painter, Clare. 1999. Learning through language in early childhood. London: Cassell.

503
 References

Painter, Clare. 2000. Researching first language development in children. In Len Unsworth (ed.), Research-
ing language in schools and communities: Functional linguistic perspectives (Re-issued 2005). London: Cassell.
65–86.
Painter, Clare. 2003. Developing attitude: An ontogenetic perspective on appraisal. Text 23(2): 183–209.
Painter, Clare. 2009. Language development. In M.A.K. Halliday and Jonathan J. Webster (eds.), 87–103.
Painter, Clare. 2017. Learning how to mean: Parent-child interaction. In Tom Bartlett and Gerard O’Grady
(eds.), 619–633.
Painter, Clare, Beverly Derewianka and Jane Torr. 2007. From microfunctions to metaphor: Learning lan-
guage and learning through language. In Ruqaiya, Hasan, Christian M.I.M. Matthiessen and Jonathan
J. Webster (eds.), 563–588.
Painter, Clare, J.R. Martin and Len Unsworth. 2014. Reading visual narratives: Image analysis of children’s
picture books. London: Equinox.
Park, K.H. 2013. The experiential grammar of Korean: A systemic functional perspective. Ph.D. thesis. Macquarie
University, Sydney.
Parodi, Giovanni (ed.). 2010. Discourse genres in Spanish: Academic and professional connections. Amsterdam:
Benjamins.
Parodi, Giovanni. 2015. Variation across university genres in seven disciplines: A corpus-based study on
academic written Spanish. International Journal of Corpus Linguistics 20(4): 469–499.
Parsons, Gerald H. 1995. Measuring cohesion in English texts: The relationship between cohesion and coherence.
Ph.D. thesis. Nottingham University, Nottingham.
Parsons, Gerald H. 1996. The development of the concept of cohesive harmony. In Margaret Berry, Chris-
topher Butler, Robin Fawcett and Guowen Huang (eds), 585–599.
Patpong, Pattama. 2005. A systemic functional interpretation of Thai grammar: An exploration of Thai narrative
discourse. Ph.D. thesis. Macquarie University, Sydney.
Patrick, Jon. 2008. The Scamseek project: Using systemic functional grammar for text categorization. In
Jonathan J. Webster (ed.), 221–233.
Patten, Terry. 1988. Systemic text generation as problem solving. Cambridge: Cambridge University Press.
Pearce, John, Geoffrey Thornton and David Mackay. 1989. The programme in linguistics and English
teaching, University College London, 1964–1971. In Hasan Ruqaiya and J.R. Martin (eds.), 329–383.
Peng Xuanwei, Alex. 2015. Halliday in China: Legacies and advances from Luo, Wang and beyond. In
Jonathan J. Webster (ed.), 2015a. 62–71.
Perrett, Gillian. 2000. Researching second and foreign language development. In Len Unsworth (ed.),
87–110.
Phillips, Joy. 1986. The development of modality and hypothetical meaning: Nigel 1;7 1/2–2;7 1/2. Work-
ing Papers no. 3. Linguistics Department. The University of Sydney.
Pickering, Martin J. and Simon Garrod. 2004. Toward a mechanistic psychology of dialogue. Behavioral and
Brain Sciences 27: 169–190.
Pickering, Martin J. and Simon Garrod. 2006. Alignment as the basis for successful communication.
Research on Language and Computation 4: 203–228.
Pickering, Martin J. and Simon Garrod. 2021. Understanding dialogue: Language use and social interaction.
Cambridge: Cambridge University Press.
Pickvance, R. and Geoffrey J. Turner. 1971. Social class differences in the expression of uncertainty in five-
year-old children. Language and Speech 14(4): 303–325.
Piehler, Margaret F. and Audrey L. Holland. 1984. Cohesion in aphasic language. In Robert H. Brook-
shire (ed.), Clinical aphasiology: Conference proceedings. Volume 14. Seabrook Island: BRK Publishers.
208–214.
Pike, Kenneth L. 1945. The intonation of American English. Ann Arbor: The University of Michigan Press.
Pike, Kenneth L. 1948. Tone languages. Ann Arbor: Michigan University Press.
Pittenger, Robert E., Charles F. Hockett and John J. Danehy. 1960. The first five minutes: A sample of micro-
scopic interview analysis. Ithaca, NY: Paul Martineau.
Plug, Leendert. 2008. J.R. Firth: A new biography. Transactions of the Philological Society 106(3): 337–374.
Plum, Guenter A. 1988. Textual and contextual conditioning in spoken English: A genre-based approach. Ph.D.
thesis. The University of Sydney, Sydney.
Pollard, Carl and Ivan A. Sag. 1993. Head-driven phrase structure grammar. Chicago and London: University
of Chicago Press.
Pounds, Gabrina. 2011. Empathy as ‘appraisal’: A new language-based approach to the exploration of clini-
cal empathy. Journal of Applied Linguistics and Professional Practice 7: 139–62.

504
 References

Poyatos, Fernando. 1993. Paralanguage: A linguistic and interdisciplinary approach to interactive speech and sounds.
Amsterdam: Benjamins.
Poynton, Cate. 1984. Forms and functions: Names as vocatives. Nottingham Linguistic Circular 13: 1–34.
Poynton, Cate. 1985. Language and gender: Making the difference. Geelong, VIC: Deakin University Press.
Poynton, Cate. 1990. Address and the semiotics of social relations: A systemic-functional account of address forms and
practices in Australian English. Ph.D. thesis. The University of Sydney, Sydney.
Poynton, Cate. 1996. Amplification as a grammatical prosody: Attitudinal modification in the nominal
group. In Christopher Butler, Margaret Berry, Robin Fawcett and Guowen Huang (eds.), Meaning and
form: Systemic functional interpretations. Norwood, N.J.: Ablex. 211–229.
Prakasam, V. 1972. A systemic treatment of certain aspects of Telugu phonology. D.Phil. thesis. University of York,
York.
Prakasam, V. 1976. A functional view of phonological features. Acta Linguistica Hungarian 26(1–2): 77–88.
Prakasam, V. 1977. An outline of the theory of systemic phonology. International Journal of Dravidian Lin-
guistics 6: 24–42.
Prakasam, V. 1979. Aspects of sentence phonology. Archivum Linguisticum 10, New Series: 57–82.
Prakasam, V. 1982a. Functional stylistics: Theory and practice. Patiala: Indian Institute of Language Studies.
Prakasam, V. 1982b. The system of length in Telugu. Pakha Sanjam 15: 349–56.
Prakasam, V. 1985. The linguistics spectrum. Patiala, India: Punjabi University.
Prakasam, V. 1987. Aspects of word phonology. In M.A.K. Halliday and Robin P. Fawcett (ed.), New devel-
opments in systemic linguistics: Theory and description. London: Pinter. 272–289.
Prakasam, V. 1992. Length in Telugu. In Paul Tench (ed.), 70–76.
Prakasam, V. 2004. Metafunctional profile of Telugu. In Alice Caffarel, J.R. Martin and Christian Mat-
thiessen (eds.), 433–478.
Prasad, Rashmi, Nikhil Dinesh, Alan Lee, Eleni Miltsakaki, Livio Robaldo, Aravind Joshi and Bonnie
Webber. 2008. The Penn Discourse TreeBank 2.0. In Proceedings, 6th International Conference on Language
Resources and Evaluation, Marrakech, Morocco.
Pun, Jack K.H., E. Angela Chan, Kristen A. Murray, Diana Slade and Christian M.I.M. Matthiessen. 2016.
Complexities of emergency communication: Clinicians’ perceptions of communication challenges in a
trilingual emergency department. Journal of Clinical Nursing 26 (21–22): 3396–3407.
Pun, Jack, Christian M.I.M. Matthiessen, Geoff Williams and Diana Slade. 2017. Using ethnographic
discourse analysis to understand doctor-patient interactions in clinical settings. SAGE Research Methods
Cases. http://dx.doi.org/10.4135/9781473979697
Qabani, Abdullah. 2018. Language, power and the ‘Arab Spring’: Three case studies. Ph.D. thesis. Macquarie
University, Sydney.
Qiu, Shijin. 1985. Early language development in Chinese children. MA honours thesis. Department of Linguis-
tics. The University of Sydney, Sydney.
Quam, Justin E. 2020. Helping language learners align with readers through narrative: Insights into the breadth,
targets, and explicitness of evaluation from appraisal studies of second language German writers. Ph.D. thesis.
Georgetown University, Georgetown.
Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik. 1985. A comprehensive grammar of
the English language. London: Longman.
Quiroz, Beatriz. 2008. Towards a systemic profile of the Spanish MOOD. Linguistics and the Human Sciences
4(1): 31–65.
Quiroz, Beatriz. 2017. The verbal group. In Tom Bartlett and Gerard O’Grady (eds.), Chap. 19. 301–318.
Quiroz, Beatriz. 2020. Experiential cryptotypes: Reasoning about process type. In J.R. Martin, Y.J. Doran
and Giacomo Figueredo (eds.), 102–128.
Quiroz Olivares, Beatriz Enriqueta. 2013. The interpersonal and experiential grammar of Chilean Spanish:
Towards a principled systemic-functional description based on axial argumentation. Ph.D. thesis. The University
of Sydney, Sydney.
Rajandran, Kumaran. 2012. Us and them: The portrayal of Malaysians and British in Malaysian history
textbooks. Journal of Asian and African Studies 48(3): 313–331.
Rajandran, Kumaran. 2014. The discursive representation of companies in Malaysian business magazines.
Pertanika Journal of Social Science and Humanities 22(2): 489–505.
Rajandran, Kumaran and Shakila Abdul Manan (eds.). 2019. Discourses of Southeast Asia: A social semiotic
perspective. Singapore: Springer.
Rajandran, Kumaran and Fauziah Taib. 2014. The representation of CSR in Malaysian CEO statements: A
critical discourse analysis. Corporate Communications: An International Journal 19(3): 303–317.

505
 References

Ravelli, Louise J. 1985. Metaphor, mode and complexity: An exploration of co-varying patterns. B.A. honours
thesis. The University of Sydney, Sydney.
Ravelli, Louise J. 1995. A dynamic perspective: Implications for metafunctional interaction and an under-
standing of theme. In Ruqaiya Hasan and Peter H. Fries (eds.), On Subject and Theme: A discourse func-
tional perspective. Amsterdam: Benjamins. 187–235.
Ravelli, Louis J. 2003. Renewal of connection: Integrating theory and practice in an understanding of
grammatical metaphor. In Anne-Marie Vandenbergen, Miriam Taverniers and Louise Ravelli (eds.),
37–64.
Rébori, Victoria. 2002. The legacy of J. R. Firth: A report on recent research. Historiographia Linguistica
29: 165–190.
Reid, Thomas B.W. 1956. Linguistics, structuralism, philology. Archivum Linguisticum 8: 28–37.
Rigaudeau-McKenna, Bernadette. 2005. Towards an analysis of dysfunctional grammar. Clinical Linguistics
and Phonetics 19(3): 155–174.
Ripich, Danielle N., Brian D. Carpenter, and Elaine W. Ziol. 2000. Conversational cohesion patterns in
men and women with Alzheimer’s disease: A longitudinal study. International Journal of Language & Com-
munication Disorders 35(1): 49–64.
Robins, R.H. 1961. John Rupert Firth: Obituary. Language 37: 191–200.
Robins, R.H. 1964. General linguistics: An introductory survey. London: Longman. 4th edition (1988) pub-
lished in 2014 by Routledge.
Robins, R.H. and Norma McLeod. 1956a. A musical and textual analysis. In Honour of J.R. Firth. Bulletin
of the School of Oriental and African Studies, University of London 18(3): 592–609.
Robins, R.H. and Norma McLeod. 1956b. Five Yurok songs: A musical and textual analysis. In Honour
of J.R. Firth. Bulletin of the School of Oriental and African Studies, University of London 18(3): 592–609.
Rochester, Sherry and J.R. Martin, J.R. 1979. Crazy talk: A study of the discourse of schizophrenic speakers.
New York: Plenum.
Roget, Peter Mark. 1852/1992. Roget’s thesaurus of English words and phrases: Facsimile of the first edition 1852.
London: Bloomsbury.
Rose, David. 2001. The western desert dode: An Australian cryptogrammar. Canberra: Pacific Linguistics.
Rose, David. 2004. Metafunctional profile of the grammar of Pitjantjatjara. In Alice Caffarel, J.R. Martin
and Christian Matthiessen (eds.), 479–536.
Rose, David. 2005. Narrative and the origins of discourse: Construing experience in stories around the
world. Australian Review of Applied Linguistics 19: 151–173.
Rose, David and J.R. Martin. 2012. Learning to write, reading to learn: Genre, knowledge and pedagogy in the
sydney school (Equinox Textbooks and Surveys in Linguistics). London: Equinox.
Rothery, Joan. 1996. Making changes: Developing an educational linguistics. In Ruqaiya Hasan and Geoff
Williams (eds.), Literacy in society. London: Longman. 86–123.
Roy, Deb, Rupal Patel, Philip DeCamp, Rony Kubat, Michael Fleischman, Brandon Roy, Nikolaos
Mavridis, Stefanie Tellex, Alexia Salata, Jethran Guinness, Michael Levit and Peter Gorniak. 2006.
The human speechome project. The 28th Annual Conference of the Cognitive Science Society, 26–29 July.
Vancouver: Cognitive Science Society.
Royce, Terry D. and Wendy L. Bowcher (eds.). 2006. New directions in the analysis of multimodal discourse.
New York: Routledge.
Ruppenhofer, Josef, Michael Ellsworth, Miriam R. L. Petruck, Christopher R. Johnson, Collin F. Baker
and Jan Scheffczyk. 2016. FrameNet II: Extended theory and practice. https://framenet2.icsi.berkeley.edu/
docs/r1.7/book.pdf
Ryle, Gilbert. 1971. Collected papers volume 2: Collected essays 1929–1968. Michigan: Hutchinson. Repub-
lished 2009. London and New York: Routledge.
Ryshina-Pankova, Marianna. 2011. Developmental changes in the use of interactional resources: Persuad-
ing the reader in FL book reviews. Journal of second language writing 20(4): 243–256.
Salkie, Raphael. 1995. Text and discourse analysis. London and New York: Routledge.
Sampson, Geoffrey. 1980. Schools of Linguistics. Stanford, CA: Stanford University Press.
Sapir, Edward. 1921. Language: An introduction to the study of speech. New York: Harcourt, Brace,
Jovanovich.
Saville-Troike, Muriel. 2002 [1982]. The ethnography of communication: An introduction. 3rd edition. Oxford:
Basil Blackwell.
Schachter, Paul. 1976. The subject in Philippine languages: Topic, Actor, Actor-topic, or none of the
above. In Charles Li (ed.), Subject and topic. New York: Academic Press. 491–518.

506
 References

Schachter, Paul. 1977. Reference-related and role-related properties of subjects. In Peter Cole and Jerrold
M. Sadock (eds.), Grammatical relations. New York: Academic Press. 279–306.
Schleppegrell, Mary J. 2004. The language of schooling: A functional linguistics approach. Mahwah, NJ: Law-
rence Erlbaum.
Schleppegrell, Mary J. 2012. Systemic functional linguistics. In James Paul Gee and Michael Handford
(eds.), The Routledge handbook of discourse analysis. London and New York: Routledge. 21–34.
Schleppegrell, Mary J. and M. Cecilia Colombi (eds.). 2002. Developing advanced literacy in first and second
languages: Meaning with power. Mahwah, N.J.: Lawrence Erlbaum.
Schultz, Anke and Lise Fontaine. 2019. The Cardiff model of functional syntax. In Geoff Thompson,
Wendy L. Bowcher, Lise Fontaine and David Schöntal (eds.), 230–258.
Schwabe, Kerstin and Susanne Winkler (eds.). 2007. On information structure, meaning and form. Amsterdam
and Philadelphia: Benjamins.
Schwartz, Theodore. 1962. Review of Pittenger, Robert E., Charles F. Hockett and John J. Danehy. 1960.
The first five minutes: A sample of microscopic interview analysis. Ithaca, NY: Paul Martineau. American
Anthropologist 64: 1313–1316.
Schwartz, Vanessa Santiago and Laura Hamman-Ortiz. 2020. Systemic functional linguistics, teacher edu-
cation, and writing T outcomes for U.S. elementary English learners: A review of the literature. Journal
of Second Language Writing 49: 1–12.
Searle, J. 1965. What is a speech act? In Max Black (ed.), Philosophy in America. London: Allen and Unwin;
Ithaca, NY: Cornell University Press.
Sefton, Peter M. 1990. Making plans for Nigel or defining interfaces between computational representations of linguis-
tic structure and output systems: Adding intonation, punctuation and typography systems to the Penman system.
BA honours thesis. The University of Sydney, Sydney.
Senft, Gunter. 2005. Bronislaw Malinowski and linguistic pragmatics. In Piotr Cap (ed.), Pragmatics today.
Frankfurt am Main: Peter Lang. 139–155.
Senis, Angela. 2016a. The contribution of John Rupert Firth to the history of linguistics and the rejection
of the phoneme theory. Proceedings of ConSOLE XXIII, 2016. 273–293.
Senis, Angela. 2016b. Le phonème face à la théorie du langage de J.R. Firth. (The phoneme and J. R.
Firth’s theory of language) 74, document 5, mis en ligne le 02 août 2017, consulté le 05 décembre 2019.
Available from http://journals.openedition.org/ml/2027
Serbina, Tatiana, Paula Niemietz and Stella Neumann. 2015. Development of a keystroke logged transla-
tion corpus. In Claudio Fantinuoli and Federico Zanettin (eds.), New directions in corpus-based translation
studies. Berlin: Language Science Press. 11–34.
Seuren, Pieter A.M. 1998. Western linguistics: An historical introduction. Oxford: Blackwell.
Sharoff, Serge. 2006. How to handle lexical semantics in SFL: A corpus study of purposes for using size
adjectives. In Geoff Thompson and Susan Hunston (eds.), 184–205.
Sharoff, Serge. 2017. Corpus and systemic functional linguistics. In Tom Bartlett and Gerard O’Grady
(eds.), 533–546.
Shore, Susanna. 1992. Aspects of a systemic functional grammar of Finnish. Ph.D. thesis. Macquarie University, Sydney.
Shore, Susanna. 1996. Process types in Finnish: Implicate order, covert categories and prototypes. In Ruqa-
iya Hasan, Carmel Cloran and David Butt (eds.), Functional descriptions: Theory in practice. Amsterdam:
Benjamins. 237–265.
Shrestha, Prithvi. 2011. Dynamic assessment of academic writing for business studies. EdD doctoral thesis. Milton
Keynes: The Open University.
Shrestha, Prithvi N. 2020. Dynamic assessment of students’ academic writing: Vygotskian and systemic functional
linguistic perspectives. Cham: Springer.
Shrestha, Prithvi and Coffin, Caroline. 2012. Dynamic assessment, tutor mediation and academic writing
development. Assessing Writing 17(1): 55–70.
Simon-Vandenbergen, Anne-Marie. 2003. Lexical metaphor and interpersonal meaning. In Anne-Marie
Simon-Vandenbergen, Miriam Taverniers and Louise Ravelli (eds.), 223–255.
Simon-Vandenbergen, Anne-Marie, Miriam Taverniers and Louise Ravelli (eds.). 2003. Grammatical meta-
phor: Views from systemic functional linguistics. Amsterdam: John Benjamins.
Sinclair, John M. 1966. Beginning the study of lexis. In C.E. Bazell, J.C. Catford, M.A.K. Halliday and
R.H. Robins (eds.), In memory of J.R. Firth. London: Longmans.
Sinclair, John M. 1991. Corpus, concordance, collocation. Oxford: Oxford University Press.
Sinclair, John McH. 1992. Trust the text. In Martin Davies and Louise Ravelli (eds.), Advances in systemic
linguistics: Recent theory and practice. London: Pinter. 5–19.

507
 References

Sinclair, John McH. and Malcolm Coulthard. 1975. Towards and analysis of discourse: The English used by
teachers and pupils. London: Oxford University Press.
Sinclair, John McH., Susan Jones and Robert Daley. 2004. English collocational studies: The OSTI Report.
London and New York: Continuum.
Skyttner, Lars. 1996. General systems theory: An introduction. Houndsmills: Macmillan.
Skyttner, Lars. 2001. General systems theory: Ideas and applications. Singapore, London, Hong Kong: World
Scientific.
Slade, Diana M. 1996. The texture of casual conversation in English. Ph.D. thesis. The University of Sydney,
Sydney.
Slade, Diana, Marie Manidis, Jeannette McGregor, Hermine Scheeres, Eloise Chandler, Jane Stein-Par-
bury, Roger Dunstan, Maria Herke and Christian M.I.M. Matthiessen. 2015. Communication in hospital
emergency departments. Berlin: Springer.
Slade, Diana, Christian M.I.M. Matthiessen, Graham Lock, Jack Pun and Marvin Lam. 2016. Patterns of
interaction in doctor-patient communication and their impact on health outcomes. In Lourdes Ortega,
A. Tyler, H.I. Park and M. Uno (eds.), The usage-based study of language learning and multilingualism.
Washington, DC: Georgetown University Press. 235–254.
Slade, Diana, Hermine Scheeres, Helen de Silva-Joyce, Jeannette McGregor, Nicole Stanton, Maria
Herke, Edith Weisberg and Deborah Bateson. 2009. Developing effective communication between doctors
and clients: Sexual and reproductive health consultations. Research report to family planning NSW.
Slade, Diana, Hermine Scheeres, Marie Manidis, Christian M.I.M. Matthiessen, Rick Iedema, Maria
Herke, Jeannette McGregor, Roger Dunston and Jane Stein-Parbury. 2008. Emergency communica-
tion: The discursive challenges facing emergency clinicians and patients in hospital emergency depart-
ments. Discourse and Communication 2(3): 289–316.
Slade, Diana, Marie Manidis, Jeannette McGregor, Hermine Scheeres, Eloise Chandler, Jane Stein-Par-
bury, Roger Dunstan, Maria Herke and Christian M.I.M. Matthiessen. 2015. Communication in hospital
emergency departments. Berlin: Springer.
Smith, Bradley A. and William S. Greaves. 2015. Intonation. In Jonathan J. Wester (ed.), 2015a,
291–313.
Snell-Hornby, Mary. 1995. Translation studies: An integrated approach. 2nd edition. Amsterdam: Benjamins.
Spencer, Elizabeth, Ann Packman, Mark Onslow and Alison Ferguson. 2005. A preliminary investigation
of the impact of stuttering on language use. Clinical Linguistics and Phonetics 19(3): 191–201.
Spencer, Elizabeth, Ann Packman, Mark Onslow and Alison Ferguson. 2009. The effect of stuttering on
communication: A preliminary investigation. Clinical Linguistics & Phonetics 23(7): 473–488.
Spencer, John and Michael J. Gregory. 1964. An approach to the study of style. In John Spencer (ed.),
Linguistics and style. London: Oxford University Press.
Spolsky, Bernard. 2008. Introduction: What is educational linguistics? In Bernard Spolsky and Francis M.
Hult (eds.), The handbook of educational linguistics. Oxford: Blackwell. 1–9.
Stainton, Robert J. and Jessica de Villiers (eds.). 2001. Communication in linguistics. Toronto: GREF (Col-
lection Theoria).
Starc, Sonja. 2010. Textual patterning and information flow (Theme ^ Rheme) in the generic evolution
of 19th century Slovene newspaper advertisements. In Elizabeth Swain (ed.), Thresholds and potentiali-
ties of systemic functional linguistics: Multilingual, multimodal and other specialised discourses. Trieste: Edizioni
Università di Trieste. 133–157.
Starc, Sonja. 2015. The difference in text structure between advertisements and classified advertisements
from a diachronic perspective. In David Banks (ed.), Aspects linguistiques de la ‘petite annonce’. Paris:
L’Harmattan. 25–35.
Starc, Sonja, Carys Jones and Arianna Maiorani (eds.). 2015. Meaning making in text multimodal and multilin-
gual functional perspectives. Basingstoke: Palgrave Macmillan.
Steele, Ross and Terry Threadgold (eds.). 1987. Language topics: Essays in honour of Michael Halliday. 2 Vol-
umes. Amsterdam: Benjamins.
Steels, Luc. 1998. Synthesizing the origins of language and meaning using coevolution, self-organization and
level formation. In James R. Hurford, Michael Studdert-Kennedy and Chris Knight (eds.), Approaches
to the evolution of language: Social and cognitive bases. Cambridge: Cambridge University Press. 384–404.
Steels, Luc and Katrien Beuls (eds.). 2019. Case studies in fluid construction grammar: The verb phrase. Amster-
dam: Benjamins.
Steiner, Erich. 1983. Die Entwicklung des Britischen Kontextualismus. Heidelberg: Groos.

508
 References

Steiner, Erich. 1985. Working with transitivity: System networks in semantic-grammatical descriptions. In
James D. Benson and William S. Greaves (eds.), Systemic perspectives on discourse. Norwood, NJ: Ablex.
163–184.
Steiner, Erich. 1986. Generating semantic structures in Eurotra-D. In COLING. Bonn: Universität Bonn,
Institut für Kommunikationsforschung und Phonetik. 304–306.
Steiner, Erich. 1988. Language and music as semiotic systems: The example of a folk ballad. In James D.
Benson, Michael J. Cummings and William S. Greaves (eds.), Linguistics in a systemic perspective. Amster-
dam: Benjamins. 393–441.
Steiner, Erich. 1991. A functional perspective on language, action and interpretation. Berlin and New York:
Mouton de Gruyter.
Steiner, Erich. 1992. Some remarks on a functional level for machine translation. Language Sciences 14(4):
623–659.
Steiner, Erich. 1996. An exercise in translation evaluation. In Angelika Lauer (ed.), Perspectives on translation
evaluation (Möglichkeiten der Übersetzungsevaluierung). Duisburg: LAUD papers in Linguistics.
Steiner, Erich. 2004. Translated texts: Properties, variants, evaluations. Frankfurt am Main: Peter Lang.
Steiner, Erich. 2005a. Hallidayan thinking and translation theory: Enhancing the options, broadening
the range, and keeping the ground. In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan J.
Wester (eds.), 481–500.
Steiner, Erich. 2005b. The heterogeneity of individual languages as a translation problem. In Harald Kit-
tel, Armin Paul Frank, Norbert Greiner, Theo Hermans, Herbert Koller, José Lambert and Fritz Paul
(eds.), Translation: An international encyclopedia of translation studies. Berlin and New York: Walter de
Gruyter. 446–454.
Steiner, Erich. 2015. Halliday’s contributions to a theory of translation. In Jonathan J. Webster (ed.),
412–426.
Steiner, Erich. 2019. Theorizing and modelling translation. In Geoff Thompson, Wendy L. Bowcher, Lise
Fontaine and David Schöntal (eds.), 739–766.
Steiner, Erich and Teich, Elke. 2004. Metafunctional profile of the grammar of German. In Alice Caffarel,
J.R. Martin and Christian Matthiessen (eds.), 139–184.
Steiner, Erich, Bo Wang, Christian M.I.M. Matthiessen and Yuanyi Ma. 2022. Bridging boundaries
between systemic functional linguistics and translation studies. In Bo Wang and Yuanyi Ma (eds.), Key
themes and new directions in systemic functional translation studies. Abingdon and New York: Routledge.
52–74.
Steiner, Erich and Jutta Winter-Thielen. 1988. On the semantics of FOCUS phenomena in EUROTRA.
In COLING. Budapest: Association for Computational Linguistics. 630–635.
Steiner, Erich and Colin Yallop (eds.). 2001. Beyond content: Exploring translation and multilingual text produc-
tion. Berlin and New York: de Gruyter.
Stillar, Glenn. 1991. Discerning the discerning traveller: Phasal analysis and ideology. Social Semiotics 1(2):
112–122.
Strevens, Peter. 1987. ‘The Linguistic Sciences and Language Teaching’ Revisited. In Ross Steele and Terry
Threadgold (ed.), Language Topics. Volume I. Amsterdam: Benjamins. 79–84.
Stuart-Smith, Virginia. 2001. Rhetorical structure theory as a model of semantics: A corpus-based analysis from a
systemic-functional perspective. Ph.D. thesis. Macquarie University, Sydney.
Suzuki, Shigeyuki. 1972. Nihongo bunpoo: Keitairon. (Japanese grammar: Morphology). Tokyo: Mugi Shobo.
Svartvik, Jan and Randolph Quirk. 1980. A corpus of English conversation. Lund: Gleerup.
Swadesh, Morris. 1934. The phonemic principle. Language 10(1): 117–129.
Swales, John M. 1990. Genre analysis: English in an academic research setting. Cambridge: Cambridge Uni-
versity Press.
Swales, John M. 2009. Worlds of genre: Metaphors of genre. In Charles Bazerman, Adair Bonini and
Débora Figueiredo (eds.), Genre in a changing world. West Lafayette, IN: Parlor Press. 147–157.
Taboada, Maite. 2019. Cohesion and conjunction. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine
and David Schöntal (eds.), 311–332.
Takahashi, Taro. 2005. Nihongo no bunpo. (Grammar of Japanese). Tokyo: Mugi Shobo.
Tam, Roy Hoi-Sang. 2004. A systemic functional interpretation of Cantonese clause grammar. Ph.D. thesis. The
University of Sydney, Sydney.
Tann, Ken. 2017. Context and meaning in the Sydney architecture of systemic functional linguistics. In
Tom Bartlett and Gerard O’Grady (eds.), Chap. 27. 438–456.

509
 References

Tannock, Rosemary, Jonathan Fine, Tracey Heintz and Russell J. Schachar. 1995. A linguistic approach
detects stimulant effects in two children with attention-deficit hyperactivity disorder. Journal of Child and
Adolescent Psychopharmacology 5(3): 177–189.
Tarski, Alfred. 1936. Grundlegung der wissenschaftlichen Semantik. Actes du Congrès international de phi-
losophie scientifique, Sorbonne, Paris 1935, vol. III, Langage et pseudo-problèmes. Paris: Hermann. 1–8. New
English translation of Polish original with introduction by Magda Stroinska and David Hitchcock,
2002, On the concept of following logically. History and Philosophy of Logic 23: 155–196.
Taverniers, Miriam. 2017. Grammatical metaphor. In Tom Bartlett and Gerard O’Grady (eds.), Chap. 22,
354–372.
Taverniers, Miriam. 2019. Semantics. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine and David
Schöntal (eds.), 55–91.
Taylor, Christopher. 1993. Systemic linguistics and translation. Occasional Papers in Systemic Linguistics 7: 87–103.
Taylor, Christopher. 2003. Multimodal transcription in the analysis, translation and subtitling of Italian
films. The Translator 9(2): 191–205.
Taylor, Christopher. 1998. Language to language: A practical and theoretical guide for Italian/English translators.
Cambridge: Cambridge University Press.
Taylor, Christopher. 2013. Multimodality and audiovisual translation. In Yves Gambier and Luc van
Doorslaer (eds.), Handbook of translation studies. Volume 4. Amsterdam: John Benjamins. 98–104.
Taylor, Christopher. 2014. The ADLAB project: And some ideas on audio description. In Kerstin Kunz,
Elke Teich, Silvia Hansen-Schirra, Stella Neumann and Peggy Daut (eds.), 387–399.
Taylor, Christopher. 2017. Reading images (including moving ones). In Tom Bartlett and Gerard O’Grady
(eds.), 575–590.
Taylor, Chris and Anthony Baldry. 2001. Computer assisted text analysis and translation: A functional
approach in the analysis and translation of advertising texts. In Erich Steiner and Colin Yallop (eds.),
Beyond content: Exploring translation and multilingual text production. Berlin and New York: de Gruyter.
277–305.
Taylor Torsello, Carol. 1996. On the logical metafunction. Functions of Language 3(2): 151–183.
Taylor Torsello, Carol and Anthony Baldry. 2005. SFL in text-based, web-enhanced language study. In
Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan J. Wester (eds.), 311–342.
Teich, Elke. 1995. A proposal for dependency in systemic functional grammar: Metasemiosis in computational systemic
functional linguistics. Ph.D. thesis. Universität des Saarlandes, Saarbrücken.
Teich, Elke. 1999a. Systemic functional grammar in natural language generation: Linguistic description and computa-
tional representation. London: Cassell.
Teich, Elke. 1999b. Contrastive linguistics and translation studies revisited. In Alberto Gil, Heidrun Ger-
zymisch-Arbogast and Erich Steiner (eds.), Papers from the Saarbrücken Symposium on Translation Studies,
Saarbrücken, November 98. Saarbrücken: University of Saarland.
Teich, Elke. 2003. Cross-linguistic variation in system and text: A methodology for the investigation of translations
and comparable texts. Berlin and New York: Mouton de Gruyter.
Teich, Elke. 2009. Computational linguistics. In M.A.K. Halliday and Jonathan J. Wester (eds.), 113–127.
Teich, Elke. 2013. Choices in analysing choice: Methods and techniques for register analysis. Lise Fontaine,
Tom Bartlett and Gerard O’Grady (eds.), 417–431.
Teich, Elke, Stefania Degaetano-Ortlieb, Peter Fankhauser, Hannah Kermes and Ekaterina Lapshinova-
Koltunski. 2016. The linguistic construal of disciplinarity: A data-mining approach using register fea-
tures. Journal of the Association for Information Science and Technology 67(7): 1668–1678.
Teich, Elke, Catherine I. Watson and Cécile Pereira. 2000. Matching a tone-based and tune-based approach
to English intonation for concept-to-speech generation. Proceedings of the 18th Conference on Computa-
tional Linguistics. Volume 2. Morristown, NJ: Association for Computational Linguistics. 829–835.
Tench, Paul. 1990. The roles of intonation in English discourse. Frankfurt am Main: Peter Lang.
Tench, Paul (ed.). 1992a. Studies in systemic phonology. London: Pinter.
Tench, Paul. 1992b. From prosodic analysis to systemic phonology. In Paul Tench (ed.), 1–17.
Tench, Paul. 1996. The intonation systems of English. London: Cassell.
Tench, Paul. 2011. Transcribing the sound of English: A phonetics workbook for words and discourse. Cambridge:
Cambridge University Press.
Tench, Paul. 2017. The phoneme and word phonology in systemic functional linguistics. In Tom Bartlett
and Gerard O’Grady (eds.), Chap. 15, 233–250.
Teruya, Kazuhiro. 1998. An exploration into the world of experience: A systemic functional interpretation of the
grammar of Japanese. Ph.D. thesis. Macquarie University, Sydney.

510
 References

Teruya, Kazuhiro. 2004. Metafunctional profile of Japanese. In Alice Caffarael, J.R. Martin and Christian
Matthiessen (eds.), Language typology: A functional perspective. Amsterdam: Benjamins. 185–254.
Teruya, Kazuhiro. 2006. Grammar as resource for the construction of language logic for advanced language
learning in Japanese. In Heidi Byrnes (ed.), Advanced instructed language learning: The complementary con-
tribution of Halliday and Vygotsky. London and New York: Continuum. 109–133.
Teruya, Kazuhiro. 2007. A systemic functional grammar of Japanese. 2 volumes. London: Continuum.
Teruya, Kazuhiro. 2009. Grammar as a gateway into discourse: A systemic functional approach to subject,
theme and logic. Linguistics and Education 20(1), Special issue edited by Heidi Byrnes, Instructed foreign
language acquisition as meaning-making: A systemic-functional approach: 67–79.
Teruya, Kazuhiro. 2010. Projection in Japanese: Ideational and interpersonal manifestations. Proceedings of
the 37th International Systemic Functional Linguistics. Beijing. 367–370.
Teruya, Kazuhiro. 2017. Mood in Japanese. In Tom Bartlett and Gerard O’Grady (eds.), Chap. 14. 213–229.
Teruya, Kazuhiro (ed.). 2022a. Imi ga yoku wakaru yooni naru tame no gengogaku: Taikei gengogaku heno shootai.
(Linguistics for a better understanding of meaning: An introduction to systemic functional linguistics). Tokyo:
Kuroshio Publishers.
Teruya, Kazuhiro. 2022b. Imizukuri no risoosu: Nihongo no taikei kinoo bunpoo to gengo bunseki
(Resource for meaning-making: Systemic functional grammar of Japanese and linguistic analysis). In
Kazuhiro Teruya (ed.), Tokyo: Kuroshio Publishers. 51–98.
Teruya, Kazuhiro, Ernest Akerejola, Thomas Hestbæk Andersen, Alice Caffarel, Julia Lavid, Christian
M.I.M. Matthiessen, Uwe Helm Petersen, Pattama Patpong and Flemming Smedegaard. 2007. Typol-
ogy of MOOD: A text-based and system-based functional view. In Ruqaiya Hasan, Christian Mat-
thiessen and Jonathan Webster (eds.), Continuing discourse on language: A functional perspective. Volume 2.
London: Equinox. 859–920.
Teruya, Kazuhiro and Christian M.I.M. Matthiessen. 2015. Halliday in relation to language comparison
and typology. In Jonathan J. Webster (ed.), 427–452.
Tesnière, Lucien. 1959. Éléments de syntaxe structurale. Paris: Librairie C. Klincksieck.
Thai, Minh-duc. 1998. A systemic functional interpretation of Vietnamese grammar. Ph.D. thesis. Macquarie
University, Sydney.
Thai, Minh-duc. 2004. Metafunctional profile of the grammar of Vietnamese. In Alice Caffarel, J.R. Mar-
tin and Christian M.I.M. Matthiessen, 397–431.
Thibault, Paul J. 1987. An interview with Michael Halliday. In Ross Steele and Terry Threadgold (eds.),
Language topics: Essays in honour of Michael Halliday. Amsterdam: Benjamins.
Thibault, Paul J. 1988. Knowing what you’re told by the agony aunts: Language function, gender differ-
ence and the structure of knowledge and belief in the personal columns. In David Birch and Michael
O’Toole (eds.), Functions of Style. London: Pinter. 205–233.
Thibault, Paul J. 1991. Social semiotics as praxis: Text social meaning making and Nabokov’s ‘Ada’. Minneapolis:
University of Minnesota Press.
Thibault, Paul J. 1995. The interpersonal grammar of mood and the ecosocial dynamics of the semiotic
exchange process. In Ruqaiya Hasan and Peter H. Fries (eds.), On subject and theme: A discourse functional
perspective. Amsterdam and Philadelphia: Benjamins. 51–89.
Thibault, Paul J. 2004. Brain, mind and the signifying body: An ecosocial semiotic theory. London and New York:
Continuum.
Thibault, Paul J. 2005. The interpersonal gateway to the meaning of mind: Unifying the inter- and intra-
organism perspective on language. In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan J.
Wester (eds.), 117–156.
Thomas, Margaret. 2011. Fifty key thinkers on language and linguistics. London: Routledge.
Thomas, Sarah and Thomas Hawes. 1997. Theme in academic and media discourse. University of Nottingham,
Monographs in systemic linguistics, Number 8. Nottingham: University of Nottingham.
Thomason, Richard H. (ed.). 1974. Formal philosophy: Selected papers of Richard Montague. New Haven and
London: Yale University Press.
Thompson, Geoff. 1996. Introducing functional grammar. London: Hodder Education.
Thompson, Geoff. 1999. Acting the part: Lexicogrammatical choices and contextual factors. In M.
Ghadessy (ed.), Text and context in functional linguistics. Amsterdam: Benjamins. 101–124.
Thompson, Geoff. 2004. Introducing functional grammar. 2nd edition. London: Hodder and Stoughton
Educational.
Thompson, Geoff. 2007. Unfolding theme: The development of clausal and textual perspectives on theme.
In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan J. Webster (eds.), 671–696.

511
 References

Thompson, Geoff. 2014. Introducing functional grammar. 3rd edition. London and New York: Routledge.
Thompson, Geoff, Wendy L. Bowcher, Lise Fontaine and David Schöntal (eds.). 2019. The Cambridge
handbook of systemic functional linguistics. Cambridge: Cambridge University Press.
Thompson, Geoff and Susan Hunston (eds.). 2006. System and corpus: Exploring connections. London: Equinox.
Thompson, Geoff and Jianglin Zhou. 1999. Evaluation and organization in text: The structuring role of
evaluative disjuncts. In Susan Hunston and Geoff Thompson (eds.), Evaluation in text: Authorial stance
and the construction of discourse. Oxford: Oxford University Press. 121–141.
Thompson, Sandra A. 1985. Grammar and written discourse: Initial vs. final purpose clauses in English.
Text 5(1/2): 55–84.
Thompson, Sandra A. 1990. Information flow and dative shift in English discourse. In Jerold A. Edmond-
son, Crawford Feagin and Peter Mühlhäusler (eds.), Development and diversity: Language variation across
time and space: A festschrift for Charles-James N. Bailey. Summer Institute of Linguistics and the University
of Texas at Arlington Publications in Linguistics 93. Dallas: Summer Institute of Linguistics and the
University of Texas at Arlington. 239–253.
Thompson, Sandra A. 2002. Object complements and conversation towards a realistic account. Studies in
Language 21(1): 125–163.
Thomson, Elizabeth A., Matoko Sano and Helen de Silva Joyce (eds.). 2017. Mapping genres, mapping cul-
tures: Japanese texts in context. Amsterdam: Benjamins.
Thomson, Elizabeth and Peter R.R. White (eds.). 2008. Communicating conflict: Multilingual case studies of the
rhetoric of the news media. London: Continuum.
Thomson, Julie. 2005. Theme analysis of narratives produced by children with and without specific lan-
guage impairment. Clinical Linguistics and Phonetics 19(3): 175–190.
Thwaite, Anne. 2007. Inclusive and empowering discourse in an early childhood literacy classroom with
indigenous students. The Australian Journal of Indigenous Education 36: 21–31.
Thwaite, Anne. 2015. Using the multimodal analysis video program for register analysis: A preliminary
study. TESOL International 10(1): 110–125.
Thwaite, Anne. 2019. Halliday’s view of child language learning: Has it been misinterpreted? Australian
Journal of Teacher Education 44(5): 42–56.
Thwaite, Anne and Judith Rivalland. 2009. How can analysis of classroom talk help teachers reflect on their
practices? Australian Journal of Language and Literacy 32(1): 38–54, South Australia.
Togher, Leanne and Linda Hand. 1998. Use of politeness markers with different communication partners:
An investigation of five subjects with traumatic brain injury. Aphasiology 12(7–8): 755–770.
Togher, Leanne, Linda Hand and Chris Code. 1997a. Analysing discourse in the traumatic brain injury
population: Telephone interactions with different communication partners. Brain Injury 11(3): 169–189.
Togher, Leanne, Linda Hand and Chris Code. 1997b. Measuring service encounters in the traumatic brain
injury population. Aphasiology 11(4–5): 491–504.
Togher, Leanne, Skye McDonald, Chris Code and Susan Grant. 2004. Training communication partners of
people with traumatic brain injury: A randomised controlled trial. Aphasiology 18(4): 313–35.
Togher, Leanne, Skye McDonald, Carl A. Coelho and Lindsey Byom. 2013. Cognitive communication
disability following TBI: Examining discourse, pragmatics and executive function.” In McDonald, Skye,
Leanne Togher and Chris Code (eds.). 2013. Social and communication disorders following traumatic brain
injury. London: Psychology Press. 89–118.
Togher, Leanne, Skye McDonald, Robyn Tate, Emma Power and Rachel Rietdijk. 2009. Training com-
munication partners of people with traumatic brain injury: Reporting the protocol for a clinical trial.
Brain Impairment 10(2): 188–204.
Tognini-Bonelli, Elena. 2001. Corpus linguistics at work. Amsterdam: Benjamins.
Toolan, Michael. 1989. Narrative: A critical linguistic introduction. New York: Routledge.
Toolan, Michael. 2009. Narrative progression in the short story: A corpus stylistic approach. Amsterdam: Benjamins.
Torr, Jane. 1997. From child tongue to mother tongue: A case study of language development in the first two and
a half years. Monographs in systemic linguistics, Number 9. Nottingham: University of Nottingham.
Torr, Jane. 1998. The development of modality in the pre-school years: Language as a vehicle for under-
standing possibilities and obligations in everyday life. Functions of Language 5(2): 157–178.
Torr, Jane. 2015. Language development in early childhood: Learning how to mean. In Jonathan J. Webster
(ed.), Chap. 9, 242–256.
Torr, Jane and Alyson Simpson. 2003. The emergence of grammatical metaphor: Literacy-oriented expres-
sions in the everyday speech of young children. In Anne-Marie Simon-Vandenbergen, Miriam Taver-
niers and Louse Ravelli (eds.), 169–183.

512
 References

Toury, Gideon. 2004. Probabilistic explanations in translation studies: Welcome as they are, would they
qualify as universals? In Anna Mauranen and Pekka Kujamäki (eds.), Translation universals: Do they exist?
Amsterdam and Philadelphia: Benjamins. 15–32.
Trager, George L. 1958. Paralanguage: A first approximation. Studies in Linguistics 13: 1–12.
Trager, George L. 1961. The typology of paralanguage. Anthropological Linguistics 3(1): 17–21.
Traugott, Elizabeth C. 1997. The role of the development of discourse markers in a theory of gram-
maticalization. Paper presented at ICHL XII, Manchester 1995, Version of 11/97. Published as Le rôle
de l’évolution des marqueurs discursifs dans une théorie de la grammaticalization. In M.M. Jocelyne
Fernandez-Vest and Shirley Carter-Thomas (eds.), Structure informationnelle et particules énonciatives: Essai
de typologie. Paris: L’Harmattan. 295–333.
Trevisan, Piergiorgio and Adolfo M. García. 2019. Systemic functional grammar as a tool for experimental
stimulus design: New appliable horizons in psycholinguistics and neurolinguistics. Language Sciences 75:
35–46.
Trew, Tony. 1979. Theory and ideology at work. In Roger Fowler, Bob Hodge, Gunther Kress and Tony
Trew (eds.), Language and control. London: Routledge and Kegan Paul. 94–116.
Triandis, Harry C. 1977. Interpersonal behavior. Monterey, CA: Brooks/Cole Publishing Company.
Troyan, Francis John, José David Herazo and Marianna Ryshina-Pankova. 2022. SFL pedagogies in lan-
guage education: Special issue introduction. System 104(2022): 102694.
Trubetzkoy, N.S. 1939. Grundzüge der Phonologie. Travaux au Cercle Linguistique de Prague 7. Prague.
Tseng, Chiao-I. 2013. Cohesion in film: Tracking film elements. Basingstoke: Palgrave Macmillan.
Tseng, Chiaoi and Bateman, John A. 2012. Multimodal narrative construction in Christopher Nolan’s
memento: A description of method. Journal of Visual Communication 11(1), 91–119.
Tseng, Chiao-I, Jochen Laubrock and John A. Bateman. 2021. The impact of multimodal cohesion on
attention and interpretation in film. Discourse, Context and Media 44: 100544.
Tsui, Amy. 1986. A linguistic description of utterances in conversation. Ph.D. thesis. University of Birmingham,
Birmingham.
Tsui, Amy B.M. 1989. Systemic choices and discourse processes. Word 40(1–2): 163–188.
Tucker, Gordon H. 1996a. Cultural classification and system networks: A systemic functional approach to
lexical semantics. In Margaret Berry, Christopher Butler, Robin Fawcett and Guowen Huang (eds.),
533–566.
Tucker, Gordon H. 1996b. So grammarians haven’t the faintest idea: Reconciling lexis-oriented and gram-
mar-oriented approaches to language. In Ruqaiya Hasan, Carmel Cloran and David Butt (eds.), Func-
tional descriptions: Theory into practice. Amsterdam: Benjamins. 145–179.
Tucker, Gordon H. 1997. A functional lexicogrammar of adjectives. Functions of Language 4(2), 215–250.
Tucker, Gordon H. 1998. The lexicogrammar of adjectives: A systemic functional approach to lexis. London:
Cassell.
Tucker, Gordon H. 2002. ‘Getting our heads around it’: Semantic and syntactic tension in the transitiv-
ity analysis of metaphorically derived multiword verbs. Studi Italiani di Linguistica Teorica e Applicata 2:
303–315.
Tucker, Gordon. 2005. Extending the lexicogrammar: Towards a more comprehensive account of partially
clausal and non-clausal expressions. Language Sciences 27(6).
Tucker, Gordon. 2007. Between grammar and lexis: Towards a systemic functional account of phraseology.
In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan J. Wester (eds.), 953–977.
Tucker, Gordon H. 2016. Hybridity in transitivity: Phraseological and metaphorically derived processes in
the system network for transitivity. In Donna Miller and Paul Bayley (eds.), 41–63.
Tucker, Gordon H. 2017. The adjectival group. In Tom Bartlett and Gerard O’Grady (eds.), Chap. 18.
284–300.
Tucker, Gordon, Guowen Huang, Lise Fontaine and Edward McDonald (eds.). 2020. Approaches to systemic
functional grammar: Convergence and divergence. Sheffield: Equinox.
Turner, Geoffrey J. 1973. Class and children’s language of control. In Basil Bernstein (ed.), Class, codes and
control vol. 2: Applied studies towards a sociology of language. London: Routledge and Kegan Paul.
Turner, Geoffrey J. 1987. Sociosemantic networks and discourse structure. In M.A.K. Halliday and Robin
P. Fawcett (eds.), New developments in systemic linguistics. London: Frances Pinter. 64–93.
Turner, Jonathan H. 2000. On the origins of human emotions: A sociological inquiry into the evolution of human
affect. Stanford, CA: Stanford University Press.
Turner, Jonathan H. and Alexander Maryanski. 1979. Functionalism. Menlo Park, CA: The Benjamin/
Cummings Publishing Company.

513
 References

Unsworth, Len. 1995. How and why: Recontextualizing science explanations in school science books. Ph.D. thesis.
The University of Sydney, Sydney.
Unsworth, Len. 1999. Explaining school science in book and CD ROM formats: Using semiotic analy-
ses to compare the textual construction of knowledge. International Journal of Instructional Media 26(2):
159–179.
Unsworth, Len (ed.). 2000. Researching language in schools and communities: Functional linguistic perspectives.
(Re-issued 2005.). London: Cassell.
Unsworth, Len, Russell Tytler, Lisl Fenwick, Sally Humphrey, Paul Chandler, Michele Herrington and
Lam Pham. 2022. Multimodal literacy in school science: Transdisciplinary perspectives on theory, research and
pedagogy. London and New York: Routledge.
Urbach, Claire. 2013. ‘Choice’ in relation to context: A diachronic perspective on cultural valeur. In Lise
Fontaine, Tom Bartlett and Gerard O’Grady (eds.), 300–317.
Ure, Jean N. 1971. Lexical density and register differentiation. In G.E. Perren and J.L.M. Trim (ed.),
Applications of linguistics: Selected papers of the 2nd International Congress of Linguistics, Cambridge 1969.
Cambridge: Cambridge University Press.
Ure, Jean N. and Jeffrey Ellis. 1977. Register in descriptive linguistics and linguistic sociology. In Oscar
Uribe-Villegas (ed.), Issues in sociolinguistics. The Hague: Mouton. 197–243.
Vachek, Josef. 1960. Dictionnaire de linguistique de l’École de Prague. Utrecht/Anvers: Spectrum Éditeurs.
New English edition, Dictionary of the Prague School of Linguistics. Translated from the French, German
and Czech original sources by Aleš Klégr, Pavlína Šaldová, Markéta Malá, Jan Čermák and Libuše
Dušková. Edited by Libuše Dušková. Amsterdam and Philadelphia: John Benjamins.
Vachek, Josef. 1964. A Prague School reader in linguistics. Prague: Academia.
van Dijk, Teun. 1997a. Discourse as structure and process: Discourse studies: A multidisciplinary introduction. Vol-
ume 1. London: SAGE Publications.
van Dijk, Teun. 1997b. The study of discourse. In van Dijk (ed.), 1–34.
van Dijk, Teun. 2008. Discourse and social context. Cambridge: Cambridge University Press.
van Leeuwen, Theo. 1996. The representation of social actors. In Carmen Rosa Caldas-Coulthard and Mal-
colm Coulthard (eds.), Text and practices: Readings in critical discourse analysis. London: Routledge. 32–70.
van Leeuwen, Theo. 1999. Speech, music, sound. London and New York: Palgrave Macmillan.
van Leeuwen, Theo. 2008. Discourse and practice: New tools for critical discourse analysis. (Oxford studies in
sociolinguistics). Oxford: Oxford University Press.
van Leeuwen, Theo and Carey Jewitt (eds.). 2001. Handbook of visual analysis. London: Sage.
Vandelanotte, Lieven. 2009. Speech and thought representation in English: A cognitive-functional approach. Berlin
and New York: Mouton de Gruyter.
Vasconcellos, Maria Lucia Barbosa. 2004. Re-textualizing Joyce’s eveline: A systemic-functional approach
to translation quality assessment. In Tradução e comunicação. Volume 13. São Paulo: Unibero. 83–118.
Vasconcellos, Maria Lucia Barbosa. 2009. Systemic functional translation studies (SFTS): The theory travel-
ling in Brazilian environments. DELTA: Documentação de estudos em lingüística teórica e aplicada (PUCSp.
Impresso) 25: 585–607.
Veel, Robert. 1998. The greening of school science: Ecogenesis in secondary classrooms. In J.R. Martin
and Robert Veel (ed.), 114–151.
Veloso, Francisco O.D. 2006. Never awake a sleeping giant . . .: A multimodal analysis of post 9–11 comic books.
Ph.D. thesis. Universidade Federal de Santa Catarina, Florianópolis.
Veloso, Francisco O.D. 2012. Comic books and the construction of reality: A critical approach in the class-
room. In Christina DeCoursey (ed.), Language arts in Asia: Literature and drama in English, Putonghua and
Cantonese. Cambridge: Cambridge Scholars Publishing. 83–102.
Veloso, Francisco O.D. 2015. Comicbooks as cultural archeology: Gender representation in Captain Amer-
ica during WWII. Journal of Linguistics and Human Sciences 11(2–3): 284–299.
Veloso, Francisco O.D. and John A. Bateman. 2013. The multimodal construction of acceptability: Marvel’s
Civil War comic books and the PATRIOT Act. Critical Discourse Studies 10(4): 1–17.
Veloso, Francisco O.D., John A. Bateman and Lau Yan Ling. forthcoming. Page design as a medium of
communication: A corpus-based analysis of visual style in comics and graphic novels. Language and
Communication. SAGE. SJR H Index: 34.
Ventola, Eija. 1987. The structure of social interaction: A systemic approach to the semiotics of service encounters.
London: Frances Pinter.
Ventola, Eija. 1988. The logical relations in exchange. In James D. Benson and Williams S. Greaves (eds.),
Systemic functional approaches to discourse. Norwood, NJ: Ablex. 51–72.

514
 References

Ventola, Eija, Cassily Charles and Martin Kaltenbacher (eds.). 2004. Perspectives on multimodality. Amster-
dam: Benjamins.
Ventola, Eija and Arsenio Jesús Moya-Guijarro (eds.). 2009. The world shown and the world told. Basingstoke:
Palgrave Macmillan.
Ventola, Eija and A. Jesús Moya-Guijarro (eds.). 2021. Challenging gender stereotypes and the traditional family
unit in children’s picture books: A multimodal analysis. London: Routledge.
Verstraete, Jean-Christophe. 2001. Subjective and objective modality: Interpersonal and ideational func-
tions in the English modal auxiliary system. Journal of Pragmatics 33: 1505–1528.
Vygotsky, Lev Semenovich. 1962. Thought and language. Edited and translated by Eugenia Hanfmann and
Gertrude Vakar. Cambridge, MA: The M.I.T. Press.
Xuan, Wenhui Winfred. 2015. A longitudinal study of Chinese high school students learning English based on
systemic functional text analysis. Ph.D. thesis. The Hong Kong Polytechnic University, Hong Kong.
Xuan, Wenhui Winfred and Christian M.I.M. Matthiessen. forthcoming. System networks as a resource in L2
writing education. Manuscript Submitted.
Wan Yau Ni, Jenny. 2009. Call centre communication: An analysis of interpersonal meaning. Ph.D. thesis. The
Hong Kong Polytechnic University, Hong Kong.
Wan Yau Ni, Jenny. 2010. Call centre discourse: Graduation in relation to voice quality and attitudinal
profile. In Gail Forey and Jane Lockwood (eds.), 106–124.
Wang, Bo. 2014. Theme in translation: A systemic functional linguistic perspective. International Journal of
Comparative Literature and Translation Studies 2(4): 54–63.
Wang, Bo. 2017. Lao She’s Chaguan (Teahouse) and its translations: A systemic functional perspective on drama
translation. Doctoral thesis. the Hong Kong Polytechnic University, Hong Kong.
Wang, Bo and Ma Yuanyi. 2020. Lao She’s Teahouse and its two English translations: Exploring Chinese drama
translation with systemic functional linguistics. London: Routledge.
Wang, Bo and Ma Yuanyi. 2021. Systemic functional translation studies: Theoretical insights and new directions.
Sheffield: Equinox.
Wang, Bo and Yuanyi Ma (eds.). 2022. Key themes and new directions in systemic functional translation studies.
Abingdon and New York: Routledge.
Wang, Constance and Liling Tang. 2014. Explicit Holmes: A diachronic investigation of explicitness and
explicitation in Chinese translations of detective stories. In Kerstin Kunz, Elke Teich, Silvia Hansen-
Schirra, Stella Neumann and Peggy Daut (eds.), 417–428.
Wang, Peng. 2004. Harry Potter and its Chinese translation: An examination of modality system in systemic func-
tional approach. Ph.D. thesis. Sun Yat-sen University, Guangzhou.
Wang, Pin. 2020. Axial argumentation and cryptogrammar in interpersonal grammar: A case study of clas-
sical tibetan mood. In J.R. Martin, Y.J. Doran and Giacomo Figueredo (eds.), 73–101.
Wang Yan, Constance. 2015. A systemic perspective on the translation of detective stories. Ph.D. thesis. The Hong
Kong Polytechnic University.
Watt, David L.E. 1992. An instrumental analysis of English nuclear tones. In Tench (ed.), 1992a, 135–160.
Watt, David L.E. 1994. The phonology and semology of intonation in English: An instrumental and systemic perspec-
tive. Bloomington, IN: Indiana University Linguistics Club Publications.
Weber, David J. and William C. Mann. 1981. Prospectus for computer-assisted dialect adaptation. American
Journal of Computational Linguistics 7(3): 165–177.
Webster, Jonathan J. (ed.). 2008. Meaning in context: Implementing intelligent applications of language studies.
London and New York: Continuum.
Webster, Jonathan J. (ed.). 2015a. The Bloomsbury companion to M.A.K. Halliday. London: Bloomsbury Academic.
Webster, Jonathan J. 2015b. Understanding verbal art: A functional linguistic approach. Berlin: Springer.
Webster, Jonathan J. 2015c. Text linguistics. In Jonathan J. Webster (ed.), 314–326.
Webster, Jonathan J. 2019. Key terms in the SFL model. In Geoff Thompson, Wendy L. Bowcher, Lise
Fontaine and David Schöntal (eds.), 35–54.
Webster, Jonathan J. and Xuanwei Peng (eds.). 2017. Applying systemic functional linguistics: The state of the art
in China today. London: Bloomsbury Academic.
Wegener, Philipp. 1885. Untersuchungen über die Grundfragen des Sprachlebens. Halle: Max Niemeyer.
Wegener, Rebekah. 2011. Parameters of context: From theory to model and application. Ph.D. thesis. Macquarie
University, Sydney.
Weinreich, Harald. 1972. Die Textpartitur als heuristische Methode. Der Deutschunterricht 24(4): 43–60.
Weiss, Gilbert and Ruth Wodak (eds.). 2003. Critical discourse analysis: Theory and interdisciplinarity. London:
Palgrave Macmillan.

515
 References

Wells, Gordon. 1994. The complementary contributions of Halliday and Vygotsky to a ‘language-based
theory of learning.’ Linguistics and Learning 6: 41–90.
Werlich, Egon. 1975. Typologie der Texte: Entwurf eines textlinguistischen Modells zur Grundlegung einer Text-
grammatik. Heidelberg: Quelle and Meyer.
Wertsch, James V. 1985. Vygotsky and the social formation of mind. Cambridge, MA: Harvard University Press.
Whitaker, Rachel. 1995. Theme, processes and the realization of meanings in academic articles. In
Mohsen Ghadessy (ed.), Thematic developments in English texts. London and New York: Pinter.
105–128.
White, Peter R.R. 2015. Appraisal theory. In Karen Tracy, Cornelia Ilie and Todd Sandel, The international
encyclopedia of language and social interaction. Place: John Wiley and Sons, Inc.
Whitelaw, Casey, Navendu Garg and Shlomo Argamon. 2005. Using appraisal groups for sentiment analy-
sis. Proceedings of the 14th ACM International Conference on Information and Knowledge Management. New
York: Association for Computing Machinery. 625–631.
Whorf, Benjamin Lee. 1956. Language, thought and reality: Selected writings of Benjamin Lee Whorf, edited by
John B. Carroll. Cambridge, MA: The MIT Press.
Wignell, Peter. 1997. Making the abstract technical: On the evolution of the discourse of social science. Ph.D. thesis.
The University of Sydney, Sydney.
Wignell, Peter. 1998. Technicality and abstraction in social science. In J.R. Martin and Robert Veel (eds.),
297–326.
Wignell, Peter. 2007a. On the discourse of social science. Darwin: Charles Darwin University Press.
Wignell, Peter. 2007b. Vertical and horizontal discourse and the social sciences. In Frances Christie and
J.R. Martin (eds.), Language, knowledge and pedagogy: Functional linguistic and sociological perspectives. Lon-
don and New York: Continuum. 184–204.
Wignell, Peter, J.R. Martin and Suzanne Eggins. 1987. The discourse of geography: Ordering and
explaining the experiential world. Working papers in Linguistics 5 – Writing project – Report 1987.
Department of Linguistics. Sydney. University of Sydney: 66–116. Reprinted as Wignell, Peter, J.R.
Martin and Suzanne Eggins. 1993. The discourse of geography: Ordering and explaining the experi-
ential world. In Halliday, M.A.K. and J.R. Martin, Writing science: Literacy and discursive power. London:
Falmer. 136–165.
Wignell, Peter, J.R. Martin and Suzanne Eggins. 1989. The discourse of geography: Ordering and explain-
ing the experiential world. Linguistics and Education 1(4): 359–391.
Wildfeuer, Janina. 2014. Film discourse interpretation: Towards a new paradigm for multimodal film analysis. Lon-
don: Routledge.
Wildfeuer, Janina and John A. Bateman. 2017. Film text analysis: New perspectives on the analysis of filmic
meaning. London: Routledge.
Williams, Geoff. 1994. Using systemic grammar in teaching young learners: An introduction. Melbourne:
Macmillan.
Williams, Geoff. 1995. Joint book reading and literacy pedagogy: A socio-semantic examination. Ph.D. thesis.
Macquarie University, Sydney.
Williams, Geoff. 1999. The pedagogic device and the construction of pedagogic discourse: A case example
in early literacy education. In Frances Christie (ed.), Pedagogy and the shaping of consciousness: Linguistic
and social processes. London: Cassell. 88–122.
Williams, Geoff. 2001. Literacy pedagogy prior to schooling: Relations between social positioning and
semantic variation. In Ana M. Morais, Isabel Neves, Brian Davies and Harry Daniels (eds.), Towards a
sociology of pedagogy: The contribution of Basil Bernstein to research. New York: Peter Lang. 17–45.
Williams, Geoff. 2004. Ontogenesis and grammatics: Functions of metalanguage in pedagogical discourse.
In Geoff Williams and Annabelle Lukin (eds.), 241–267.
Williams, Geoffrey. 2005a. Grammatics in schools. Ruqaiya Hasan, Christian M.I.M. Matthiessen and
Jonathan J. Wester (eds.), 281–310.
Williams, Geoff. 2005b. Language, brain, culture. Linguistics and the Human Sciences 1(3): 147–150.
Williams, Geoff. 2005c. Semantic variation. In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jona-
than J. Webster. Continuing discourse on language: A functional perspective. Volume 1. London: Equinox.
457–480.
Williams, Geoff. 2015. Halliday as an international educator. In Jonathan J. Webster (ed.), 327–347.
Williams, Geoff. 2016. Language socialization: A systemic functional perspective. In Patricia A. Duff and
Nancy H. Hornberger (eds.), Language socializaion: Encyclopedia of language and education. 3rd edition.
Cham: Springer. 1–16.

516
 References

Williams, Geoff. 2019. Language development. In Geoff Thompson, Wendy L. Bowcher, Lise Fontaine
and David Schöntal (eds.), 487–511.
Williams, Geoff and Ruth French. 2016. Teaching and learning grammatics. In Helen de Silva Joyce and
Susan Feez (eds.), 295–299.
Williams, Geoff and Annabelle Lukin (eds.). 2004. The development of language: Functional perspectives on spe-
cies and individuals. London and New York: Continuum.
Williams, Malcolm Paston. 1989. A comparison of the textual structures of Arabic and English written texts: A
study in the comparative orality of Arabic. Ph.D. thesis. The University of Leeds, Leeds.
Winograd, Terry. 1968. Linguistics and the computer analysis of tonal harmony. Journal of Music The-
ory 21: 2–49. Reprinted in M.A.K Halliday and J.R. Martin (eds.), 1981, 257–270. Reprinted in
Stephan Schwanauer and David Levitt (eds.), 1993, Machine models of music. Cambridge, MA: MIT
Press. 113–153.
Winograd, Terry. 1983. Language as a cognitive process: Syntax. Reading: Addison Wesley.
Wirnani, Indah. 1985. Verbal, mental and behavioural processes in Indonesian. M.A. thesis. Department of Lin-
guistics. The University of Sydney, Sydney.
Wittenburg, Peter, Hennie Brugman, Albert Russel, Alex Klassmann and Han Sloetjes. 2006. ELAN:
A professional framework for multimodality research. Proceedings of LREC 2006, Fifth International
Conference on Language Resources and Evaluation. Genoa: European Language Resources Association.
1556–1559.
Wodak, Ruth. 2015. Von Wissensbilanzen und Benchmarking: Die fortschreitende Ökonomisierung der
Universitäten. Eine Diskursanalyse. In Rainer Diaz-Bone and Gertraude Krell (eds.), Diskurs und Öko-
nomie Interdisziplinäre Diskursforschung. 2nd edition. Berlin: Springer. 367–388.
Wodak, Ruth and Michael Meyer. 2009. Critical discourse analysis: History, agenda, theory, and meth-
odology. In Ruth Wodak and Michael Meyer (eds.), Methods for critical discourse analysis. 2nd edition.
London: Sage. 1–33.
Wong, May. 2019. Multimodal communication: A social semiotic approach to text and image in print and digital
media. Cham: Palgrave Macmillan.
Woodward-Kron, Robyn, Catriona Fraser, John Pill and Eleanor Flynn. 2014. How we developed doctors
speak up: An evidence-based language and communication skills open access resource for international
medical graduates. Medical Teacher 37(3): 31–33.
Woodward-Kron, Robyn, Diana Slade, Eleanor Flynn, Jane Stein-Parbury, Jacquie Widin, Lisa Smith and
Hermine Scheeres. 2011. Multimedia learning and teaching resources. Communication for Health in Emergency
Contexts (CHEC): Teaching and learning resource for emergency department communication. Media learning
and teaching resources, Final report. Available from www.chec.meu.medicine.unimelb.edu.au/
Wu, Canzhong. 2000. Modelling linguistic resources. Ph.D. thesis. Macquarie University, Sydney.
Wu, Canzhong. 2009. Corpus-based research. In M.A.K. Halliday and Jonathan J. Wester (eds.), 128–142.
Wu, Canzhong, Christian M.I.M. Matthiessen and Maria Herke (eds.). 2008. Proceedings of the 35th Interna-
tional Systemic Functional Linguistics Congress: Voices around the world, 19–25 July 2008. Sydney: Macquarie
University.
Xuan, Wenhui Winfred. 2014. A longitudinal study of Chinese high school students learning English based on
systemic functional text analysis. Ph.D. thesis. The Hong Kong Polytechnic University, Hong Kong.
Xuan, Wenhui Winfred and Chen Shukun. 2019. A synthesis of research on grammatical metaphor: Meta-
data and content analysis. Word 65(4): 213–233.
Xuan, Wenhui Winfred and Chen Shukun. 2020. Taking stock of accumulated knowledge in projection
studies from systemic functional linguistics: A research synthesis. Functional Linguistics 7(1). https://doi.
org/10.1186/s40554-019-0070-7
Yallop, Colin. 2001. The construction of equivalence. In Erich Steiner and Colin Yallop (eds.), Beyond
content: Exploring translation and multilingual text production. Berlin and New York: de Gruyter. 229–246.
Yan, Bingjun and Wang Rui. 2017. Language policy: A systemic functional approach. London and New York:
Routledge.
Yang, Bingjun and Wang Rui. 2017. Language policy: A systemic functional approach. London and New York:
Routledge.
Yang, Yanning. 2015. Grammatical metaphor in Chinese. Sheffield: Equinox.
Young, Lynne and Claire Harrison (eds.). 2004. Systemic functional linguistics and critical discourse analysis:
Studies in social change. London and New York: Continuum.
Young, Michael W. 2004. Malinowski: Odyssey of an anthropologist, 1884–1920. New Haven: Yale University
Press.

517
 References

Yu, Hailing and Wu Canzhong. 2016. Recreating the image of Chan master Huineng: The roles of mood
and modality. Functional Linguistics 3(4): 1–21.
Yu, Hailing and Wu Canzhong. 2017. Recreating the image of Chan master Huineng: The role of personal
pronouns. Target 29(1): 64–86.
Zappavigna, Michele. 2011. Visualizing logogenesis: Preserving the dynamics of meaning. In Shoshana
Dreyfus, Susan Hood and Maree Stenglin (eds.), Semiotic margins: Meaning in multimodalities. London and
New York: Continuum. 211–229.
Zappavigna, Michele. 2012. The discourse of Twitter and social media. London: Continuum.
Zappavigna, Michele. 2014. Enacting identity in microblogging through ambient affiliation. Discourse and
Communication 8(2): 209–228.
Zappavigna, Michele, Chris Cléirigh, Paul Dwyer and J.R. Martin. 2010. The coupling of gesture and
phonology. In Monika Bednarek and J.R. Martin (eds.), New discourse on language: Functional perspectives
on multimodality, identity, and affiliation. London: Continuum. 219–236.
Zappavigna, Michele and Shoshana Dreyfus (eds.). 2020. Discourse of hope and reconciliation: On J.R. Martin’s
contribution to systemic functional linguistics. London: Bloomsbury Academic.
Zhang, Daozhen Russell. 2020a. Language through translation: Exploring Alice in Chao Yuen-Ren’s Chinese
‘wonderland’. Cambridge: Cambridge Scholars Publishing.
Zhang, Delu, Edward McDonald, Fang Yan and Huang Guowen. 2005. The development of systemic
functional linguistics in China. In Ruqaiya Hasan, Christian M.I.M. Matthiessen and Jonathan J. Wester
(eds.), 15–36.
Zhang, Dongbing. 2020b. Negotiating interpersonal meaning in Khorchin Mongolian. Ph.D. thesis. The Uni-
versity of Sydney, Sydney.
Zhang, Dongbing. 2020c. Axial argumentation below the clause: The verbal group in Khorchin Mongo-
lian. In J.R. Martin, Y.J. Doran and Giacomo Figueredo (eds.), 35–72.
Zhang Peija, Kaela. 2018. Public health education through posters in two world cities: A multimodal corpus-based
analysis. Ph.D. thesis. The Hong Kong Polytechnic University, Hong Kong.
Zhang Ruihua, Flora. 2016. The systems and functions of identifying processes: A semiotic perspective. Ph.D. thesis.
The Hong Kong Polytechnic University, Hong Kong.
Zhao, Sumin. 2011. Learning through multimedia interaction: The construal of primary social science knowledge in
web-based digital learning materials. Ph.D. thesis. The University of Sydney, Sydney.
Zhao, Sumin, Emilia Djonov, Anders Björkvall and Morten Boeriis (eds.). 2018. Advancing multimodal and
critical discourse studies: Interdisciplinary research inspired by Theo van Leeuwen’s social semiotics. London and
New York: Routledge.
Zhu, Yongsheng and Yan Shiqing. 2001. Reflections on systemic functional linguistics. Shanghai: Shanghai For-
eign Language Education Press.

518
 INDEX

1st-order systems: physical system 13 Arabic (Modern Standard) 449

2nd-order systems: biological systems 13 arbitrary 117, 125, 134, 331
3rd-order system: social systems 13 architecture of language 9–10, 27, 278
4th-order systems: semiotic systems 14 arguing 205, 216
artefact 199, 346
abduction 138, 202 articulatory 22
abductive 222 articulatory phonology 134
abstraction 113 aspect 28
acquisition 245 assembly of registers 144
action clause 310 Attention Deficit Hyperactivity Disorder 260
action research 183 aural 113
activity theory 17 authentic 188
adult language 240 authentic data 195
adverb 386 authenticity 257
advising 215–216 authentic (naturally occurring) observing 191
affect 113 Autism Spectrum 260
affective 293 automated analysis 189, 191, 225, 228, 230
Akan 211, 449 automation 257
alternative systemic option 236 autosegmental phonology 78
Alzheimer’s Disease 261 axiality 9, 20, 172
American anthropological linguistics 37–38 axial order 331
analysis 190–191, 236 axiology 72
analysis of text 185 axis 7, 100, 348, 432
annotated text 225, 231
annotation 442 Bajjika 211, 449
anthropology 54 behavioural 277
anti-language 290 behavioural clauses 308, 377
antithesis 275 behavioural process 379
Anvil 343 behaviourist pedagogy 245
Aphasia 261 Beja 449
appliable 87, 198 Bernstein’s sociology 54
Appliable Discourse Analysis 72, 179 Bernstein’s work 244
appliable linguistics 5, 50, 65, 73, 179, 183, 271 big data 240
Appliable System Description (ASD) 72 biological systems 5
applications 1 bio-semiotic systems 118, 136–137
appraisal 301–302 Birmingham School 59, 61, 298
appraisal 345 Bloomington grammar 107

519
 Index

bodily environment 101 complementary methods 187, 190, 192, 230

boundaries 340, 450 complementary models 28
boundary markers 443–444 complementary research methods 192
bound clause 276, 368–369 complex adaptive systems 17, 331, 433
box diagram 342, 452 component part 2
breakthrough to literacy 243 composition 9
British Academic Written English Corpus 244 compositional scale 165
Brown Corpus 76 comprehensive description 4
butterfly collection approach 202 comprehensiveness of analysis 227
computational linguistics 53, 154
calligraphic potential 137 computational representations 11
Cardiff architecture 279 conative 327
Cardiff Grammar 35n3, 50, 163, 176n5, 335n4 conceptual metaphor 118
Cartesian Analysis 2, 30–31, 205 concordancer 228
categorizing 214, 216 configuration 322, 375
channel 219, 443 configurational parse 375
character 233 conflate 442
charactery 137 conflation 359
Chinese (Cantonese) 448 conjunction 386, 420
Chinese (Mandarin) 211, 448 connotation 301
Chomskyan linguistics 23 connotative semiotic 283, 285
chooser-&-inquiry framework 117 connotative semiotic system 19, 102, 104
chord 237 conspiracies 38
chronicling 214, 216 construction 83
chunking 450 Construction Grammar 74
circumstance 375 construing 162
clause 334 contact 113
clause as figure 359 content 14, 292
clause as message 359 content plane 7, 117, 331
clause as move 359 context 41, 118–119, 281, 283, 314n2, 320
clause complex 313, 334 context of culture 104, 114, 176n4, 284
cline 9 context of situation 36n11, 102, 104–105, 176n4, 284
cline of delicacy 83 contextualism 37
cline of instantiation 1, 7, 8, 18, 20, 104, 118, contextual parameter 114
138–140, 143, 224, 284, 291 continuative 428
clinical handover 264 conventional 117, 125, 134, 331
clinical linguistics 258 conventions 436, 440
clinical linguistic studies 260 Conversation Analysis (CA) 33, 76, 98n18
codal variation 141–142, 145 corpus 195, 212
code 243 corpus annotation 76
coding orientation 288 corpus-based 221
co-evolution 16, 144 corpus-based studies 60
cognitive approaches to meaning 89 corpus-driven 221
Cognitive Construction Grammar 84 corpus linguistics 76
Cognitive Grammar 52, 84 cosmogenesis 6, 13
cognitive linguistics 74 cosmology 291, 347
cognitive turn 238 Critical Discourse Analysis (CDA) 33, 59, 64, 273
cohesion 259 Critical Linguistics 59, 64, 66, 273
cohesive analysis 223, 452 critical pedagogy 245
collaborating 215–216 critical system description 273
collective 6 cross-sectional 250, 252
commonsense knowledge 241, 246 cryptogrammar 38
communication disorders 262 culture 14, 19
Communication Linguistics 97n6 culturomics 54
Communist Party Linguistics Group 289
companions and handbooks 43 Dagaare 211, 449
complementarities 24, 26–27 Danish 449

520
 Index

data collection 195 effect of verbal action 103

data sharing 344 ELAN 234, 343
Daughter Dependency Grammar 49 element 121, 322
debate 275 elicit 193
deduction 138 elicitation 185
deductive 22 elicited 188, 191
delicacy 26, 100 embedded 369
delicacy of focus 205 embedded cognition 299
delicate 162 embodied cognition 299
denotation 301 emergence 6
denotative semiotic 285 empathy 270
denotative semiotic system 118 enabling 215–216, 220
dependency grammar 83 enacting 120
descriptions 1 engendering 120
descriptions of particular languages 22 English 449
descriptive 221, 276–277 ergative 308, 315n11
descriptive category 4, 203 ethnosemantics 217
descriptive issue 298 European functionalism 37
descriptive term 332 EUROTRA 254
development 152 evaluation procedures 198
diagnostic tool 262 evaluative discourse 60
dialect adaptation 204 events 32
dialectal variation 141 evolution 6, 152
dialogue game 299, 301 exchange 299
diary method 240 exchange dynamics 299
diatypic variation 90 existential clause 308
dictionary 381 existential process 380
dictionary view of lexis 253 exosomatic semiotic systems 170
dimensions 23 expand 442
directing 215–216 expansion 313, 425
direction 257 experiential 49, 293, 328
discontinuity 455 experiential mode 162
discourse analysis 60, 222, 333 experimental 187, 191
discourse biography 269 expertise 108
discourse diary 32, 195, 196, 269 explaining 214, 216
Discourse Representation Theory 89 explication de texte 274n6, 337
discovery procedure 198, 433 exploring 215–216, 220
disorder 259 exponence 9, 287
distinctive feature 335n7 expository 213
division of labour 266 expounding 214, 216, 220
doctor-centred 267 expression 14, 292
doing 215–216, 220 expression plane 7, 131
domain knowledge 248 expressive 327
double agnation 293 (Extended) Standard Theory 51
downranked 369 extensive clauses 308
drama 213 extrinsic function 321
dramatizing 216 extrinsic functionalism 37
eye movement tracking 233
East Anglia Group 72
ecolinguistics 73, 98n16, 273 familiarity 108, 217
ecological 3 family of construction grammars 84
ecologism 3 fault-tolerant healthcare system 266
eco-social environment 101 festschrift 45
educational contexts 253 field 12, 107–108, 110, 113, 220, 269, 272
educational linguistics 44, 180, 238, 242, 248 field of activity 213, 216, 270
educational linguistic studies 251 field of experience 217, 247
educational policy 251 field work 185

521
 Index

figure 121, 375 genesis 154

film 42 genre(s) 90, 213, 281, 321, 323
film theory 54 Genre Model 90–91, 105, 245, 282, 336n4
Finnish 449 geographic regions 44
first language development studies 250 German 449
first-order category 109 gesture 42
Firthian linguistics 37 Glossematics 22
Firthian tradition 22 Gooniyandi 448
flexi theory 50 gradual approximation 205
flow of discourse 157 grammar 323
flow of information 157 grammar-translation method 243
Fluid Construction Grammar 84 grammaticalization 154
folk knowledge 241, 246 grammatical metaphor (GM) 95, 241
forensic linguistics 60 grammatics 181
form 292, 322 graphetics 131, 137
formal, model-theoretic semantics 78 graphology 131, 137
formal approaches to meaning 89 grid 181
formal grammar 47 growth 152
formalism 3 guide 2
fractal 9, 20
fractal patterns 4, 166 Hallidayan account 291
fractal principle 291 Hallidayan tradition 292
FrameNet 84 Halliday’s work 51
Frankfurt School 97n14 handbook 2
freedom 368 Hartford Stratificationalists 220
French 449 Head-Driven Phrase Structure Grammar (HPSG) 82
French and Dutch Functionalism 73 healthcare 258
frequency of combination 236 healthcare communication studies 180, 258
frequency of instantiation of systemic option 236 helical 24
function 321–322 helical move 28
functional approaches to meaning 88–89 helical progression 30
functional-cognitive space 75, 83, 89 hierarchic knowledge 244
functional component 250 hierarchy 9
Functional Discourse Grammar 83 hierarchy of stratification 1, 7, 18, 20, 99, 224, 290
functional diversification 19, 20, 155 hierarchy of the stratification of metalanguage 12
functional diversification of context 291 higher-order content plane 14
functional grammar 47 higher-order semiotic system 15, 331
Functional Grammar (FG) 83 historical linguistics 154
functionalism 73 Hjelmslevian 104
functionalism and contexualism in anthropology 38 holistic 3
functional linguistics 5 holistic account 7
functional tenor 281, 283 holistic approach 2, 23, 31, 205
functional traditions 330 horizontal knowledge 244
Functional Unification Grammar 52 hortatory 213
functional variation in context 92 human bodily potential 135
function-rank matrix 21, 23, 126–127, 197 human speech-home project 274n11
function-stratification matrix 7, 23, 116, 166 hypotaxis 313, 442
fuzzy logic 123
fuzzy theory 53 ideational 328
ideational base 120
gaps 430 ideational experiential 71
Generalized Phrase Structure Grammar (GPSG) 82 ideational grammatical metaphor 163
General System(s) Theory (GST) 17, 331 ideational lexical metaphor 163
general template 203 ideational: logical 71
general theory of language 22 ideational metaphor 167
generative linguistics 13 ideology 284
Generic Structure Potential (GSP) 259 IFG 126, 304

522
 Index

image 172 language as healthcare resource 262

immaterial 270 language as ideology 62
immaterial in nature 5 language as knowledge 56
immaterial system 14 language as resource 3
individual 6 language as rule 3
individual meaner 297 language as system 56
individuation 6, 100, 286, 287–288, 297 language-based theory of learning 55
Indonesian 448 language development 238
induction 138, 199 language diary 195
inductive 222 language evolution 154
ineffable 47, 316 language functioning in context 42
informational 293 language in context 32, 56
insert 442 language learning 243
instance 6, 119 language planning 181
instance type 119 language teaching 243
instantial character 236 languaging 32
instantiation 6, 26, 100, 197, 270, 431 langue 17
institutional role(s) 108, 217, 266, 268 layering 441
institution(s) 109, 144, 180, 239, 284 learning/teaching about language 251
instructing 215–216 learning/teaching language 251
intensive clause 308 learning/teaching through language 251
interactional 293 Legitimation Code Theory 245
interaction base 120 lens 181
interdisciplinarity 53 level 323
interface 118 level formation 18
interlevel 120, 136 level of analysis 227
interpersonal 71, 328 level-skipping 286
interpersonal grammatical metaphor 162 LexCart 126
interpersonal lexical metaphor 162 Lexical Functional Grammar (LFG) 52, 82
interpersonal metaphor 167 lexical metaphor 118
interpersonal reactances 303 lexical profile 318
interpersonal systems 160 lexicogrammar 6, 7, 41, 117, 119, 125
interpersonal Theme 357, 363 lexicogrammatical analysis 223
interpretation 236, 346 lexicogrammatical analysis of text 337, 339
inter-rank realization 441 lexicogrammatical chords 233
inter-stratal realization 100 lexicogrammatical function-rank matrix 9
interview 45 lexicogrammatical metaphor 87, 167
intonation 60 life 13
intrinsic function 327–328 lifespan literacies 239
introduction to SFL 39 lifeworld 258
inventorying 214, 216 line 334
invisible pedagogy 245 linguistic research method 186
iteration 113 linguistic universal 22
literacy 30, 137
Japanese 211, 448 literacy education 243
Japanese Functionalism 74 literacy studies 250
literature review 93
key-stroke logging 233 local dimension 7
Khorchin Mongolian 211 logical 49, 313, 328
kinesics 170 logical mode 162, 164
knowers 246 logical relationship 293
knowledge 246, 248 logico-semantic type 313
Korean 448 logogenesis 155
logogenetic time frame 152
language acquisition 153–154 London School 37–38
language as art 56 longitudinal 250, 252, 266
language as behaviour 56 longitudinal approach 241

523
 Index

longitudinal case study 240 modal responsibility 303

longitudinal study 238 modal structure 367
lower-level variation 142 mode 13, 107–108, 110, 113, 219, 272
models 194, 220
macrofunction(s) 294, 328 modes of meaning 120, 168
macro texts 274n7 modules 51, 126
macro-thematic 31 Mongolian 448
magical 327 mono-semiotic 233
Malinowski’s anthropological theory 102 Mood element 373
manipulate 351 mood type 370
manual 2 morphology 125
manual analysis 189, 191, 225, 228, 230 morphosyntax 126
map 126 mother tongue(s) 241
maps of SFL 2 move 121, 324
Marxism 39 multilingual studies 254, 273
Marxist 289 multi-metafunctional construct 356
material 270 multimethod research 192
material action 113 multimodal 90
material clause 378 Multimodal Analysis tool 77
material in nature 5 Multimodal Discourse Analysis (MDA) 64
material process 379 multimodality 64
material system 14 multimodal (multisemiotic) studies 44, 167
matrices 7, 23 multiple Themes 361
Mbembe 449 multisemiotic 233
meaners 169 multi-semiotic analysis of text 234
meaning 14, 88, 168, 292, 324, 331 multi-semiotic system 42
meaning group 290 multivariate 164
meanings at risk 142, 144 multivariate analysis 454
medical consultation 263 multivariate structure 442
medium 108, 219, 375
mental clause 310, 378 narrating 216
mental process 379 narrative 213
message 121, 324 natural 117, 125, 331
message semantics 177n12 natural grammar 87
meta-analysis 93, 94, 189, 193, 344 naturalistic (non-experimental) 191
meta-context 12 naturalness 257
meta-context of application 272 natural (non-experimental) 187
metafunction(s) 9, 100, 197, 293, 294, 328, 431 navigational help 2
metafunctional 90 negativity 293
metafunctional clause analysis 223 negotiation 299
metafunctional hookup 19, 328 network science 17, 55, 433
metafunctional modes of meaning 6 neuroscience 54
metafunctional spectrum 158 neurosemiotics 17, 293, 432
metalanguage 9, 10, 275, 316 New Rhetoric 91
metalinguistic function 327 NLP (Natural Language Processing) 76
metalinguistic strata 12 non-transformational 35n3
metaphor 118, 166 non-verbal action 103
metaredundancy 100 Nuffield Project 243
metastability 144 number of metafunctions 276
method of development 159 nursing 265
methodological strategy 24
methodological trade-off 227 object 248
microfunction(s) 294, 328 observe 193
Minimalist Program 75 Oko 449
mixed methods 192 Okudian functional linguistic 46
modal analysis 361 ontogenesis 117, 155, 241, 294
modality 26, 293, 345 ontogenetic time frame 152

524
 Index

order 442 Praat 136

order of data 257 pragmatic 327
ordered typology of systems 5, 10, 61, 266, 331 pragmatics 124
overall dimension 100 Prague School 37–39, 59, 98n22
pre-language 240–241
palliative care 265 preposition 424
paradigm 440 prepositional phrase attachment 355
paradigmatic 329 preselect 442
paradigmatically-based 87 presentation 435
paradigmatic axis 21, 121, 330 pre-systemic 225
paradigmatic organization 12 pre-systemic analysis 229
paradigmatic orientation 168 pre-systemic literature 37
paradigmatic probing 375 pre-systemic text analysis 229
paradigm probing 202 primary data 189, 191
paralanguage 42, 170 primary semiotic system 15
parameter 336n7 primatology 54
parameterization of context 107 principal components analysis 187
paraphone 297 probabilistic linguistics 75
parataxis 313, 442 probe(s) 351, 352, 354
paratone 297 procedural 213
parole 17 process 375
parser 228 process type 305
participant 375 process type 308, 310–312, 315n12, 379, 382
particular dimensions 100 programming language 12
patient 267 progressive pedagogy 245
patient-centred 267 projection 113
patient-centred care 266 projection 313, 425
patient journey 264 promoting 215–216
pattern grammar 85 proportionality 24–25
periodicity 159 prosodic 22
person 14 prosodic analysis 38–39, 78, 223, 450–451
personae 14 prosodic parse 364
phasal analysis 233 prosodic phonology 134, 157
phenomenal order 331 protolanguage 101, 167, 240–241
phenomenal realm 16, 176n9 proto-pragmatics 799
phonemes 134 proxemics 170
phonetics 7, 131, 135 psychiatry 259
phonetic yoga 32 public health 265
phonology 6, 7, 26, 41, 131, 135, 450
phrasal verb 358 quadri-stratal model of language 314n1
phraseology 49, 84 quadristratal system 101
phylogenesis 153 qualitative 185, 191, 346
phylogenetic time frame 152 qual/quant 256
physical system 5 quantitative 185, 191, 346
pivot 303
Placement 118 Radical Construction Grammar 84
poetic function 327 rallying 215
poetry 334 Range 376
polysemous 387 rank 7, 20, 100, 348, 432
polysemous conjunctions/adverb 358 rank debate 296
Positive Discourse Analysis 64, 72 rank scale 9, 21
POS tagger 227 rap 134
post-infancy 240 rapport 270
potential 6, 119, 135 raw text 225, 231
potential for application 271 reactance 24, 303
power 108 reading 138
power (status) role 218 realization 7, 9, 26, 100, 441

525
 Index

realization operands 441 semiotic 89

realization operator 441 semiotic dimension 7, 51, 197, 278
realization statement 441 semioticization 135
reasoning 123 semiotic space 23
recommending 215–216, 220 semiotic system 1, 5, 7, 103, 167, 239
recording: record 191, 193 semogenesis 152
recreating 214, 216, 220 sensorimotor 137
reference corpus 200 sentence 334
reference grammar 200 sentiment analysis 97n13
register 143, 144, 201, 208, 212, 282–283, 321, 325 sequence 120, 335
registerial cartography 145 sequential parse 360
register studies 146 service encounters 105
register variation 90, 141 SFG 51
regulating 215–216 SFL websites 434
regulatory semantics 115, 119, 146, 302 shadow text 236
relational 51 sharing 216, 220
relational process 380 sharing experiences 216
relationship-centred 267 sharing values 216
relationship-centred care 266 shunting 24
renewal of connection 33 Sign-Based Construction Grammar 84
replicate 194 signification 125
reporting 193, 214, 216, 220 situation 19
representational 327 situation type(s) 144, 268–269, 284
research potential 185 social accountability 65
research potential of SFL 272 social action 108
research synthesis 93–94, 189 social actors 17
resonance 107, 118 social and biological and physical 5
resource 32, 131, 185 social/psychological pedagogy 245
resource for making meaning 32 social semiotics 59, 61, 63–64
reviewing 215–216 social system 5
rhetorical mode 108, 219, 281 socio-interactive turn 238
Rhetorical Structure Theory (RST) 159, 297 sociology 54
rhetorical unit 297 sociometric role 218
risk 266 socio-semiotic system 136
Role and Reference Grammar (RRG) 83 sociosemantic system network 119, 146
role structure 108 somatic semiotic system 169
RST diagrams 442 song 134
song-and-rap 134
scale-&-category 23, 32 sounding potential 131
scale-&-category theory 9, 253 Spanish 449
schizophrenia 260 specimen 346
scientific knowledge 241 spectrum of functional diversification 20
secondary data 189, 191 spectrum of metafunction 1, 7, 155
second/foreign language(s) 241 speech act theory 114, 299, 300
Second Language Acquisition (SLA) 238 speech community 14
second-order categories 109 speech fellowship 14, 297
self-replication 6 speech function 259
semantic component 294 speech level 304
semantic networks 121 speech role 108
semantic rank scale 121 speech synthesis 134
semantic reticula 442 sphere of action 113
semantics 7, 41, 119 stage 183
semantic slime 282 stanza 334
semantic structure 442 status 113
semantic system 160 Stoic-Saussurean theory of the sign 36n8
semantic variation 141, 325 stratal 89
semiosis 5 stratal ascent 236

526
 Index

stratal location of semantics 117, 118–119 system network 21, 172, 291, 436–440
stratal organization of language in context 101 system science 17
stratal tension 166 systems thinking 3, 30
stratification 6, 100, 197, 348, 430 system-structure theory 329
Stratificational Linguistics 52, 58–59
stratification-function matrix 197 tabular analysis 456–457
stratification-instantiation matrix 7, 21, 23, tabular display 342
113, 139 Tagalog 448
stratified 10 tagged text 232
stratum 71 Tagmemic Linguistics 59
stratum of computational representation 12 Tagmemics 50
stratum of systemic functional theory 10 taxis 313
stratum of theoretical representation 10 teaching about language 250
structural organization 100 teaching language 250
structuring 441 teaching and language (education) 250–251
style 107 teaching policy (meta-education) 250
Subject 277, 302–303, 373 teaching through language 250
subject matter 108 Telegu 448
Subject person 374 tenor 13, 107–108, 110, 113, 217, 220, 269, 272
subliminal 291 tense 28
subpotential 119 terminological ambiguity 181
sub-sentence 334 text 19, 120, 333
substance 292 text alone (monomodal) 225
substance strata 101 text analysis 41–42, 222
surveying 214, 216 text as artefact 225, 229–230
Swedish 449 text as process 225, 232
Sydney Architecture 335n4 text as product 225, 232
Sydney Grammar 50, 335n4 text as specimen 225, 229, 230
Sydney School 105–106, 176n5, 243, 245, 282, 286 text base 120
syllable 134 text in context 185
symbolic organization 108 text linguistics 333
syndrome 38, 237 Textpartitur 232
syntagmatic 329 text + other semiotic “text” (multimodal) 225
syntax 125 texts 42
synthesis 190–191 text score 232, 237
SysConc 77 text type 143–144, 213
SysFan 343 text typology 91, 212, 314n3
system 19, 25, 325, 329, 336n7 textual 48, 71, 90, 157, 328
system-&-process 330 textual metafunction 351
system-&-structure theory 39 textual metaphor 178n23
systematic 329 textual Theme 356, 363
systematic literature reviews 189 Thai 211, 448
systematic reviews 94 thematic 293
system-based text analysis 202 thematic analysis 360, 362
system description 196 thematic approach 31
systemic 89, 329 theme 357, 304
systemic coverage 206 theoretical 276
systemic functional grammar 52 theoretical guidance 203
systemic functional linguistics (SFL) 1, 61, 82 theoretical issue 278
systemic functional metalanguage 10 theorization of context 105
systemic functional semiotic theory 181 theory 1, 48
systemic functional theory 254 theory-neutral 182
systemic organization 100 theory of codes 142
systemic text analysis 229 thesaurus 274n12, 381
systemic traditions 330 a thesaurus in machine translation 253
systemic variation 141 thesaurus view 318
system in SFL 331 thesis 275

527
 Index

thick description 4, 304, 373 universalist fallacy 23

ToBi 134 usage-based theories/theory 75, 329
tonal harmony 167
tone group 134 valeur(s) 25, 125, 135, 289
tools 435 validity of research finding 193
topical Theme 357, 363 valuation 218
topology 123 value 14
transactional function 327 variation 50, 141, 167
transcription 442 vectors 100
transdisciplinarity 53 verb 384
trans-disciplinary approach 3 verbal action 103
a transdisciplinary motif 331 verbal action 113
transfer comparison 204 verbal art 177n21
transformational grammar 35n3 verbal clause 310
transition 240–241 verbal process 380
transitive 308 verse 334
transitivity 26, 305, 345 Vietnamese 448
translation equivalence 254 visible pedagogy 245
translation process research (TPR) 187, visual 113
233, 274n4 vocabulary 381
translation studies 44, 180, 253 voice quality 134
transmission 113, 245 volume of texts analysed 227
Tree Adjoining Grammar (TAG) 82
tree banks 232 wave of information 159
triangulation 192 Weri 448
tricky term 321 West-Coast Functionalism 74–75
trinocular 293 West-Coast Functionalist tradition 4
trinocular perspectivizing 24 Western Desert 448
trinocular view 349 whole texts 208
trinocular vision 304, 348 word 326
tri-stratal 314n1 word clouds 319
turn 219 Word Grammar 35n3
turn-taking 108 wording at risk 216
typological guidance 203 wording potential 126
working paradigm 207, 209
UAM Corpus tool(s) 77, 343, 435 writing system 29
unification 82 written language 29
unification-based grammar 82 written medium 29
Unification Construction Grammar 84 written systems 29
unified theories 3 written world 29
unified theory 4, 7
univariate 164 XML 343
univariate analysis 452
univariate structure 422 Yoruba 449

528