APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS: Developing Theory From Practice 9781350109292, 9781350109322, 9781350109308

Citation preview

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Bloomsbury Studies in Systemic Functional Linguistics Series Editors: J. R. Martin, John S. Knox, David Caldwell Among functional approaches to language and related semiotic systems, Systemic Functional theory stands out as an evolving paradigm, constantly developing new systems to accommodate descriptive challenges. Bloomsbury Studies in Systemic Functional Linguistics responds to this ever-developing field, speaking to instances of evolution at the frontier of the discipline. Publishing contemporary, cutting edge research in Systemic Functional Linguistics, this cohesive series unites complementary developments into an integrated multiperspectival whole. Titles focus on specific themes to explore emerging new fields of research in Systemic Functional theory alongside innovations within long established areas of SFL research. Placing emphasis on new voices and directions, Bloomsbury Studies in Systemic Functional Linguistics demonstrates how a disciplinary singular like SFL continues to evolve and subsume its past into possible futures. New and forthcoming titles in the series: Exploring British Sign Language via Systemic Functional Linguistics, Luke A. Rudge Interpersonal Meaning in Multimodal English Textbooks, Yumin Chen Modelling Paralanguage Using Systemic Functional Semiotics, Thu Ngo, Susan Hood, J. R. Martin, Clare Painter, Bradley A. Smith and Michele Zappavigna Critical Lexicography from a Social Semiotic Perspective, Wenge Chen The Language of Business Reasoning and Decision-Making: Analysing Student Writing in Business Higher Education, Eszter Szenes Writing With Students: New Perspectives on Supporting Collaborative Writing in EAP Contexts, Lucy Macnaught

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

DEVELOPING THEORY FROM PRACTICE David Caldwell, John S. Knox, and J. R. Martin

BLOOMSBURY ACADEMIC Bloomsbury Publishing Plc 50 Bedford Square, London, WC1B 3DP, UK 1385 Broadway, New York, NY 10018, USA 29 Earlsfort Terrace, Dublin 2, Ireland BLOOMSBURY, BLOOMSBURY ACADEMIC and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2022 Copyright © David Caldwell, John S. Knox, and J. R. Martin, 2022 David Caldwell, John S. Knox, and J. R. Martin have asserted their right under the Copyright, Designs and Patents Act, 1988, to be identified as Editors of this work. For legal purposes the Acknowledgements on p. xxvii constitute an extension of this copyright page. Cover image © Magnia / Shutterstock All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. Bloomsbury Publishing Plc does not have any control over, or responsibility for, any third-party websites referred to or in this book. All internet addresses given in this book were correct at the time of going to press. The author and publisher regret any inconvenience caused if addresses have changed or sites have ceased to exist, but can accept no responsibility for any such changes. A catalogue record for this book is available from the British Library. A catalog record for this book is available from the Library of Congress. ISBN: HB: 978-1-3501-0929-2 ePDF: 978-1-3501-0930-8 eBook: 978-1-3501-0931-5 Series: Bloomsbury Studies in Systemic Functional Linguistics Typeset by Integra Software Services Pvt. Ltd. To find out more about our authors and books visit www.bloomsbury.com and sign up for our newsletters.

CONTENTS

L ist

of

F igures 

ix

L ist

of

T ables 

xiv

N otes

on

C ontributors 

P reface  A cknowledgements  Introduction: Introducing Appliable Linguistics J. R. Martin, John S. Knox and David Caldwell

xviii xxv xxvii 1

Part I  Innovative Description 1 Graduation in Research Writing: Managing the Dual Demands of Objectivity and Critique Susan Hood

23

2 A Balancing Act: Theorizing Compositional Choices in Photographs Helen Caple

41

3 Developing a Multimodal Rank Scale John S. Knox

55

4 A Hip-Hop Battle: Describing Sound in the Contested Academy David Caldwell

67

vi

CONTENTS

Part II  Innovative Practice 5 ‘There are known unknowns … ’: Turning to Systemic Functional Linguistics to Understand Language Use in a Classified Military Context Elizabeth A. Thomson 6 Designing Pedagogic Registers: Reading to Learn David Rose

83 103

7 Developing a Systemic Functional Grammar of Korean as a Resource for Practical Application Mira Kim

127

8 The Reading to Learn Bilingual Programme: Taking Multilingualism into Genre-based Pedagogic Practices Harni Kartika Ningsih

141

9 Big Data and Managing Multimodal Complexity Kay L. O’Halloran, Sabine Tan, Peter Wignell, Rebecca Lange, Kevin Chai and Michael Wiebrands

157

10 Our Own Double Helix: The Power of Teachers and Students Understanding How Language Works in Schooling Contexts Brian Dare and John Polias

181

11 ‘Appliability’ of SFL in Two Academic Contexts in Argentina: Developing the Teaching of Academic Genres in Spanish Estela Inés Moyano

197

12 Developing an SFL-inspired Four-year Collegiate Foreign Language Programme: Contexts, Constructs, Curriculum Heidi Byrnes

211

13 Putting Systemic Functional Linguistics (SFL) and Sydney School Genre Pedagogy (SSGP) to Work in a Compulsory First-year Arts Subject Shoshana Dreyfus and Trish Weekes

231

Part III  Evolving Theory 14 Modelling Newborn Screening Genetics Diagnosis Texts: Language as a Tool for Research and Treatment of Sickle Cell Disease Giacomo Figueredo and Kelen Cristina Sant’Anna de Lima

251

CONTENTS

15 Re-reading Reading Images Theo van Leeuwen 16 The Appraisal Framework and Analyses of Journalistic Discourse: Objectivity, Subjectivity and Attitudinal Positioning Peter R.R. White 17 Binding and Bonding: A Retrospective and Prospective Gaze Maree Stenglin 18 Affiliation: An Appliable Framework for Exploring Community and Identity in Discourse Lorenzo Logi and Michele Zappavigna 19 Semiotic Description: Grappling with Mathematics Y. J. Doran 20 Revisiting the Modelling of ‘big texts’ in SFL: The Role of Embedding in Construing ‘depth’ in Long Business Reports Eszter Szenes

vii

275

295 309

325 341

355

Part IV  Appliable Linguistics in Progress 21 Re-thinking the Unit of Phase in the Analysis of Classroom Discourse Lucy Macnaught 22 Construing Knowledge through Mandarin Chinese: A Discourse Semantic Perspective Jing Hao 23 Semiotic Software through the Lens of Systemic Functional Theory Emilia Djonov and Theo van Leeuwen 24 Reading to Learn, Learning to Teach: Emergent Bilingual Latina/ o/x Parents Read in English to Their Young Emergent bilingual Children at Home Andrés Ramírez

385

405 421

437

25 Revitalizing an Endangered Genre: Applying Genre Theory and the Teaching and Learning Cycle Anna Crane

457

26 Teaching the Nation: Iconography and Nationalism in English Language Textbooks in Post-War Sri Lanka Namala Tilakaratna

473

viii

CONTENTS

27 Bodies Talk: Modelling Paralanguage in Systemic Functional Linguistics Lorenzo Logi, Michele Zappavigna and J. R. Martin

487

I ndex 

507

LIST OF FIGURES

0.1

Basic English mood options and their structural realizations

4

0.2

Basic Tagalog mood options and their structural realizations

4

0.3

Language strata

5

0.4

Ideational, interpersonal and textual metafunctions

6

0.5

A rank scale for English grammar

7

0.6

Language, register and genre

8

0.7

Register (field, tenor, mode) and genre

9

1.1

A system network for graduation: force: intensification Source: Hood and Zhang (2020: 23)

31

A system network for graduation: force: quantification. Source: Hood and Zhang (2020: 23)

33

A system network for graduation: focus: specification. Source: Hood and Zhang (2020: 23)

34

A system network for graduation: focus: fulfilment. Source: Hood and Zhang (2020: 23)

35

A page from the Sydney Morning Herald demonstrating the use of a single, striking photograph and caption on the printed page (2 June 2003: 11 © Fairfax syndication)

43

Visual representations of the golden mean, rule of thirds and dynamic asymmetry. Source Caple (2009: 67–68)

46

1.2 1.3 1.4 2.1

2.2

x

LIST OF FIGURES

2.3

The balance network. Source: Caple (2009: 168)

48

2.4

An example of an axially composed photograph (Sydney Morning Herald, 15 June 2004: 12 © Fairfax syndication)

50

Class-function diagram showing a rank scale perspective on online newspaper structure, from newspaper to newsbite. Source: Knox (2009: 161)

58

Lay approach to understanding online newspaper website structure

59

3.3

Possible basic system of newspapers. Source: Knox (2009: 165)

63

3.4

Possible basic system of websites. Source: Knox (2009: 167)

63

4.1

System network for syllable for rap and sung voice

72

4.2

Visual representation comparing rap (above) and sung (below) voice from West. Source: Caldwell (2011: 123, 120)

74

4.3

System network for sound act for rap and sung voice

74

5.1

Bi-directionality between language levels and social and cultural contexts. Source: de Silva Joyce and Feez (2012: 29)

88

5.2

Teaching/learning cycle. Source: Rose and Martin (2012: 308)

94

6.1

Realization, instantiation and individuation

106

6.2

Story phases

107

6.3

Curriculum genres configure two registers together

112

7.1

NAATI translation pass ratios. Adapted from Kim (2009: 151)

130

7.2

The system of function marking

135

8.1

Topology of bilingual education programmes. Source: Kartika-Ningsih and Rose (2018)

145

Reading to learn three-layered cycle. Source: Rose and Martin (2012)

148

8.3

The cycle of the R2L bilingual programme

149

8.4

R2L learning exchange. Source: Martin & Rose (2012) © Equinox Publishing Ltd 2012

150

A multimodal mixed-methods approach. Source: O’Halloran et al. (2018: 24)

163

3.1

3.2

8.2

9.1

LIST OF FIGURES

9.2

Screenshot of sample analysis carried out within Multimodal Analysis Image software: Overlays [1]; selected system choice [2]; list of available system choices [3]; annotation nodes [4]; system strips [5]

xi

165

9.3

MMA Visualization App: Results of reserve image search for ‘The Flood’, Dabiq, Issue 2. Source: O’Halloran et al. (2017: 197) 168

9.4

MMA Cluster App: Interactive visualization of text and image clusters

171

MMA Cluster App: Individual results of text and image analysis for an article

172

MMA Cluster App: Tables with results of language and image analysis

173

9.5 9.6 9.7

MMA Cluster App: Vector data for language (and image) analysis 174

9.8

Sankey diagram showing distribution of articles with ISISrelated images in online media, in terms of IBM Watson categories, uClassify’s IAB Taxonomy webpage content topics, and tonality (i.e. corporate (formal) or personal (informal))

174

The multimodal analysis approach. Source: O’Halloran et al. (2020: 106)

175

9.9

11.1 Results of an implementation of the adapted pedagogic proposal (teacher: Prof Yanina García)

206

13.1 Workshop 1 whiteboard notes

234

13.2 Genre analysis spreadsheet

237

14.1 The flowchart of our application

255

14.2 Representation of a text converted into points in language space

261

14.3 Representation of a text with clauses clustered in a dendrogram (tree graph)

262

14.4 The system of speech person in English

263

14.5 Representation of clause points in language space and as a dendrogram (tree graph)

264

14.6 Clustering in language space in BP

266

xii

LIST OF FIGURES

14.7 Dendrograms for manuals in BP and English

267

15.1 Invasion by Warsaw Pact troops in front of the Radio Headquarters, Prague, August 1968. Photo: Joseph Koudelka © Magnum Photos, reproduced with the permission of Ritzau

280

15.2 Research KPIs in an Australian University (detail). Source: Kress and Van Leeuwen 2021. Reproduced by permission of Taylor & Francis Group

281

15.3 The Singing Man (Ernst Barlach, 1928) © MoMA. Reproduced with the permission of Scala Archives

284

18.1 Hashtag with engagement resource implicates alternate coupling

334

20.1 Point of departure for exploring macrogenres: part/part univariate structure. Based on Halliday (1985) and Martin (1996) 359 20.2 Embedded genres realizing the generic structure of the Canada Report

362

20.3 The orbital structure of the Canada Report

368

21.1 Interpersonal systems of meaning organized onto specific ranks in a rank scale

388

21.2 Part of the negotiation system network organized onto the rank of exchange. Adapted from Martin and Dreyfus (2015)

389

21.3 System-structure cycles. Adapted from Hao (2015) and Martin (2013)

391

21.4 The system of collaborative writing

396

21.5 Towards a system of teacher–student communication

398

21.6 A higher and lower rank of pedagogic activity with corresponding systems

399

21.7 System–structure cycles for pedagogic activity

399

21.8 A rank scale for pedagogic activity in contexts of teaching and learning (following Rose’s (2018) terms for pedagogic register)

402

22.1 Field activities organized at different tiers

410

23.1 Slide design styles: (a) a slide from a presentation on copyright law; (b) a slide from a cultural studies lecture

431

24.1 Mothers learn with their children at Parent Power

441

LIST OF FIGURES

xiii

24.2 Chicken-based recipe: Tacos of Chiken breast. I cook chiken breast. Cut in pieces. Cold taco and cut tomatos, lettace in enjoy

443

24.3 Procedural recount (recipe) produced by one of the mothers

444

24.4 Mother recounting the story to the whole class during Preparing to Read

447

24.5 Word recognition exercises with the sentence ‘They zoomed past the school and into the laundromat’

448

24.6 Mothers working together to reorganize the cut-out sentence

449

24.7 Mothers take turns to write the sentence they reorganized

450

24.8 Excerpt from home guide for parents

451

24.9 Nubia holds her daughter’s hand to trace the letter m

452

24.10 Nubia teaching other mothers how to teach their children

453

25.1 Epistemological contrasts

463

26.1 Iconography. Adapted from Tann (2010b)

477

26.2 Topological representation of the legends iconography. Source: Tilakaratna (2016: 122)

478

27.1 YouTube recording of Julia Gillard’s ‘misogyny speech’. Source: Gillard (2012) © ABC News

488

27.2 The genesis of language, paralanguage and body language from protolanguage according to Cléirigh (2011)

489

27.3 Ontogenesis – from multimodal protolanguage (Halliday 2004) via a transitional period to language and paralanguage. Adapted from Matthiessen (2009: 22)

492

LIST OF TABLES

0.1

English Mood Types and Structural Realizations

1.1

Text 1 (from Painter, Martin and Unsworth (2013: 1))

29

1.2

Text 2 (from Painter, Martin and Unsworth (2013: 5–6))

30

5.1

Table of In-Flight Example of a Spoken Information Exchange Using an Aviation Register

91

Table of Meanings and Lexicogrammatical Features of the Information Exchange Text Example

92

Table of the Set of Lexicogrammar and Phrasal Expressions Present in the Information Exchange Text Example

93

5.2 5.3

3

6.1

Stages, Phases and Themes in a Traditional Story

109

6.2

Lesson Stages in Series

115

6.3

Activities within Lesson Stages

116

6.4

Presence and Mass in Curriculum Genres

117

6.5

Preview Paragraph

118

6.6

Elaborating Meanings

119

7.1

Example 1: Experiential Function Marking

133

7.2

Example 2: Textual Function Marking

133

7.3

Example 3: Interpersonal Function Marking

133

7.4

Function Markers in Korean

134

7.5

Example 4: Perspective Function

136

7.6

Example 5: Modal Function

136

LIST OF TABLES

xv

7.7

Example 6: Installer Function

137

7.8

Example 7: Subjoiner Function

137

8.1

Example 1: Bilingual Learning Exchange in Detailed Reading Iterations 1 and 2

151

Example 2: Bilingual Learning Exchange in Detailed Reading Iteration 3

152

Example 3: Bilingual Learning Exchange in Joint Construction Iteration 3. Source: Kartika-Ningsih and Rose (2018)

152

8.4

Example 4: Enveloping: L2 Learning Exchange

153

9.1

Examples of Language and Image systems (adapted from Halliday & Matthiessen 2014; Martin & Rose 2007; O’Toole 2011; Kress & van Leeuwen 2020)

164

Multimodal Framework for Prototype Model

171

8.2 8.3

9.2

11.1 Adapted Genre-Based Pedagogic Proposal. Source: Moyano (2007)

205

13.1 Genres in First-Year Writing in Arts Disciplines at UOW, adapted from Martin and Rose (2008); Szenes (2017)

238

13.2 Parts of Essay Prompts

239

13.3 Functional Parts of Essay Prompts

240

14.1 Corpus for Modelling

258

14.2 Example of Annotation of a Clause

259

14.3 Static Profile of the Corpus for transitivity, mood, theme in BP and English

265

16.1 The System of Judgement, based on Iedema, Feez and White (1994), revised and updated in Feez, Iedema and White (2008)

304

17.1 Challenges Facing Museums of the Twenty-First Century

312

18.1 Positive Evaluation of a Pie Party in a Casual Conversation, adapted from Knight (2010a: 219)

329

18.2 Deferral of a Wrinkling Bond in a Casual Conversation, adapted from Knight (2010a: 219)

329

18.3 An Example of Condemning, adapted from Knight (2010a: 233)

330

20.1 The ISSUE Stage in the Analytical Discussion of the Canada Report

363

20.2 Synoptic View of the Generic Structure of the Analytical Discussion Realizing the Canada Report

364

xvi

LIST OF TABLES

20.3 The Staging of Argument 2 of the PERSPECTIVE 1 Stage of the Canada Report

365

20.4 Embedded Arguing Genres Functioning as Multivariate Stages to Build the Depth of the PERSPECTIVE 1 Stage of the Canada Report

366

21.1 Syntagmatic Structure from the Perspective of Function and Class 390 21.2 Functional Structures in Teacher–Student Talk to Co-create a Text, refined from Macnaught (2015), with terms following Rose and Martin (2012)

392

21.3 Generating Activity during Joint Construction

393

21.4 Gathering and Choosing Activity during Joint Construction

394

21.5 Reworking Activity during Joint Construction

395

21.6 Different Types of Class Structures Co-occurring with a Functional Structure

397

21.7 Classroom Discourse Analysis of Pedagogic Activity at Two Ranks

400

22.1 Example 1: Configuration with an Occurrence and an Entity That Engenders the Occurrence

412

22.2 Example 2: Configuration with an Occurrence and Two Entities

412

22.3 Example 1b: Realization of a Figure through a Clause and the Test with the Aspect le

412

22.4 Example 3: A Figure (with an Engendered Entity) realised through a Nominal Group

413

22.5 Example 4: A Figure (with an Engenderer Entity) realised through a Nominal Group

413

22.6 Examples 5, 6 and 7: Unpacking Figures from the Perspective of Periodicity

414

22.7 Example 8: Activity Entities Realized through Nominal Groups

415

23.1 Software for Media/Text Production and Social Media: Some Differences

425

25.1 Example of Setting Context and Building Field during Deconstruction Phase

464

25.2 Example of Clause, Gloss and Translation from Whiteboard

464

25.3 Excerpt from Exchange between Elder and Ranger during Deconstruction

465

25.4 Jointly Constructed Text with Analysis of Obligatory Meanings

466

LIST OF TABLES

xvii

27.1 Sonovergent and Semovergent Paralanguage

493

27.2 An Example of Hand Movement Enacting Interpersonal Sonovergence. Images from Gillard 2012 © ABC News

494

27.3 Examples of a Gesture Co-extensive with a Tone Group Enacting Textual Sonovergence (Tonality). Images from Gillard 2012 © ABC News

494

27.4 An Example of a Gesture Co-occurring with Tonic Syllables. Images from Gillard 2012 © ABC News

495

27.5 An Example of a Gesture Co-occurring with Salient Syllables. Images from Gillard 2012 © ABC News

496

27.6 Converging Paralinguistic and Discourse Semantic Systems

496

27.7 An Example of Hand Shape and Motion Concurring with an Ideational Event. Images from Gillard 2012 © ABC News

497

27.8 An Example of Hand Shape and Position Delineating an Ideational Entity. Images from Gillard 2012 © ABC News

497

27.9 An Example of Hand Shape Delineating a Semiotic Entity. Images from Gillard 2012 © ABC News

498

27.10 An Example of Gesture Concurring with Ideational Entity of Spatial Location. Images from Gillard 2012 © ABC News

498

27.11 An Example of Evaluation Resonating with Negative Affect. Images from Gillard 2012 © ABC News

499

27.12 An Example of Facial Tension Resonating with Negative Affect. Images from Gillard 2012 © ABC News

500

27.13 An Example of Hand Shapes and Arm Position Resonating with force. Images from Gillard 2012 © ABC News

500

27.14 An Example of Movement Resonating with force. Images from Gillard 2012 © ABC News

501

27.15 An Example of Variations in Deictic Gesture Concurring with focus. Images from Gillard 2012 © ABC News

501

27.16 An Example of Hand and Arm Shape Concurring with Deictic Exophoric Reference ‘He’. Images from Gillard 2012 © ABC News

502

27.17 An Example of Hand Shape and Motion Concurring with Deictic Exophoric Reference ‘We’. Images from Gillard 2012 © ABC News

502

27.18 An Example of Body Motion Concurring with Periodicity. Images from Gillard 2012 © ABC News

503

NOTES ON CONTRIBUTORS

Heidi Byrnes is George M. Roth Distinguished Professor of German Emerita at Georgetown University, USA. Her publications have focused on the development of advanced literacy, including the relationship of language, thought and the construction of knowledge, particularly in a curricular context. SFL has been her language theory of choice. She is a past president of AAAL, past editor of the Modern Language Journal and is the recipient of a number of professional association awards. David Caldwell is Senior Lecturer in English Language and Literacy at the University of South Australia. David is interested in the application of Systemic Functional Linguistics to contemporary language contexts, including hip-hop music, the on-field language of sports people, wordings on printed T-shirts and genre pedagogy in Australian schools. Helen Caple is Associate Professor in Journalism at the University of New South Wales, Australia. Her research interests centre on news photography, text-image relations in new and old media, discursive news values analysis and visual-verbal representations of diversity. Key publications include: Photojournalism Disrupted: The View from Australia (2019); Multimodal News Analysis Across Cultures (2020). Kevin Chai is currently working/seconded as the lead data scientist to the Health Economics and Data Analytics (HEDA) group in the School of Population Health and the Western Australia Health and Artificial Intelligence Consortium (WAHAIC) from the Curtin Institute for Computation (CIC). Kevin’s role involves providing data, computing, analytics, machine learning/

NOTES ON CONTRIBUTORS

xix

artificial intelligence and research expertise in projects with researchers and industry partners. Anna Crane is a linguist and PhD candidate at the University of Sydney, Australia. Her main research interest is the nexus between linguistic description and the priorities of First Nations people working to strengthen and teach their languages. She has worked with Gija people on community and school-based documentation, education and revitalization projects since 2005. Brian Dare is an educational consultant in language and literacy across all levels of schooling. He is a co-author of a number of professional development programmes focusing on the role of language in teaching and learning. He has a particular interest in building a systemic functional metalanguage for teachers across all curriculum areas including languages other than English. In recent years, he has been working as a mentor in secondary schools seeking to improve literacy outcomes for their students. Emilia Djonov is Senior Lecturer in language and literacy education at Macquarie University, Australia. Her research in social semiotics, semiotic technology, and multimodality has been published in journals such as Discourse, Context & Media; Critical Discourse Studies; Social Semiotics; Visual Communication and Text & Talk. Y. J. Doran is Senior Lecturer at the Australian Catholic University, Australia. His research focuses on language, semiosis, knowledge and education, spanning the fields of educational linguistics, discourse analysis, language description and multimodality, and he focuses mainly on English and Sundanese. Shoshana Dreyfus is Senior Lecturer in English Language and Linguistics at the University of Wollongong, Australia. Specializing in Systemic Functional Linguistics and discourse analysis, her research has mostly focused on nonverbal communication and language disorder, discipline-specific tertiary academic literacy and developments in SFL theory. Giacomo Figueredo is Associate Professor of Linguistics at the Federal University of Ouro Preto, Brazil, where he is an investigator in the Laboratory of Language Experimentation, carrying out empirical experimental research on language description, modelling and generation. Jing Hao is Assistant Professor at the Pontifical Catholic University of Chile (PUC). Previously she worked as Postdoctoral Research Fellow at PUC and at the Hong Kong Polytechnic University, following her doctorate in Linguistics

xx

NOTES ON CONTRIBUTORS

at the University of Sydney. Her research interests include knowledge building through English and Mandarin Chinese, its interaction with other semiotic modes and the application in literacy education. Susan Hood is Honorary Associate Professor in the Department of Linguistics at the University of Sydney, Australia. Her primary research focus is academic discourse, with on-going interests in language and disciplinarity and the multimodality of teaching/learning, focusing especially on the role of the semiotic body in the mode of live lecturing. Harni Kartika-Ningsih is Assistant Professor in the Linguistics Department at the University of Indonesia. Harni did a postdoctoral fellowship at the Chinese University of Hong Kong after finishing her doctoral degree at the University of Sydney, Australia. She is currently researching in the area of bilingual education and therapeutic communication. Her research interests include genre pedagogy, multilingualism and education, classroom discourse, and health discourse. Mira Kim is Associate Professor in the School of Humanities and Languages at UNSW Sydney, Australia. Her research interests include translator education, text analysis for translation, systemic functional typology, personalized learning, learner autonomy and international students in higher education. She has co-edited Systemic Functional Linguistics and Translation Studies (2021). John S. Knox is Honorary Senior Lecturer in the Department of Linguistics at Macquarie University, Australia. His research and publications are in the areas of social semiotics and applied linguistics. Rebecca Lange was a data scientist at the Curtin Institute for Computation (Curtin University, Australia) where she applied her research, data analytics and coding skills to support researchers with their computational projects, before she moved to work in industry. Kelen Cristina Sant’Anna de Lima is Professor at the Faculty of Health and Human Ecology (Federal University of Minas Gerais, Brazil), where she teaches medical text production and bioethics. She also works as an Educational Assistant at NUPAD. Kelen finished her postdoctoral studies at the Federal University of Minas Gerais. She has been involved in several projects interfacing Language and Medicine. She has published many papers in the areas of Linguistics, Translation and Human Health. Lorenzo Logi is a casual academic working in Sydney, Australia. He completed his PhD in 2021 at the University of New South Wales, Australia, employing

NOTES ON CONTRIBUTORS

xxi

Systemic Functional Linguistics to research the social semiosis occurring in stand-up comedy performances. His research interests include humour, multimodality and new media, and he is currently investigating how emoji interact with language in social media discourse. Lucy Macnaught is Senior Lecturer (and the Senior Learning Advisor) for the Learning Success team at Auckland University of Technology, New Zealand. She collaborates with faculty to teach academic literacies within coursework and research programmes. Her research interests include: academic literacy development, synchronous online writing instruction, User Experience Design and classroom discourse analysis. J. R. Martin is Professor of Linguistics at the University of Sydney, Australia. His research interests include systemic theory, functional grammar, discourse semantics, register, genre, multimodality and positive discourse analysis (PDA), focusing on English, Tagálog, Korean and Spanish – with special reference to the transdisciplinary fields of clinical linguistics, educational linguistics, forensic linguistics and social semiotics. He was elected a fellow of the Australian Academy of the Humanities in 1998, and was Head of its Linguistics Section from 2010 to 2012. In April 2014 Shanghai Jiao Tong University opened its Martin Centre for Appliable Linguistics, appointing Professor Martin as Director. Estela Inés Moyano gained her PhD at the University of Buenos Aires, Argentina. She works as Senior Lecturer at the National University of General Sarmiento, Argentina, and at the National University of Guillermo Brown, Argentina, and is Researcher and Coordinator of PROLEA-UFLO at the Universidad de Flores, Argentina. Her research, based in Systemic Functional Linguistics, explores the description of Spanish, scientific discourse analysis at different specialization levels and academic literacy at different educational levels. She has created and run academic literacy programmes across the curriculum at university level as well as projects of secondary teacher training in scientific literacy. Kay O’Halloran is Chair Professor and Head of Department of Communication and Media at University of Liverpool, UK. Kay’s research area is multimodal analysis. Her early work involved multimodal approaches to mathematics and in her later research she focused on developing new digital tools and techniques for analysing text, images and videos. More recently, she has been developing mixed-methods approaches for analysis of large multimodal data sets. John Polias is Director of Lexis Education and an international consultant in teaching and learning with a focus on language, especially multilingual learning

xxii

NOTES ON CONTRIBUTORS

contexts. His principal research interest lies in developing relevant crosscurriculum pedagogies that are underpinned by the role of language but which also have a multisemiotic approach. Andrés Ramírez is Associate Professor in the Department of Curriculum and Instruction at Florida Atlantic University, USA. His research explores the economic, cultural and linguistic intricacies implicated in the development of culturally and linguistically responsive pedagogies for Emergent to Advanced Bilingual students (EABs) with special emphasis on Latina/o/x EABs populations. David Rose is Associate of the University of Sydney and Director of Reading to Learn (www.readingtolearn.com.au). His research interests include analysis and design of literacy teaching practices, classroom discourse and teacher professional learning, as well as language typology, language and culture, and social semiotic theory. His books include The Western Desert Code (2001), and, with J. R. Martin, Working with Discourse (2007), Genre Relations (2008) and Learning to Write, Reading to Learn (2012). Eszter Szenes is Marie Skłodowska-Curie Actions Research Fellow at the Department of Sociology and Social Anthropology, Central European University, Austria, and Senior Fellow at the Center for Global Resilience and Security at Norwich University, USA. She holds a PhD in Systemic Functional Linguistics from the University of Sydney, Australia. Her research interests include violent extremism, information warfare, critical thinking and disciplinary literacies. Maree Stenglin’s research interests include multimodal discourse analysis, 3D space, visual communication, social semiotics, Systemic Functional Theory and multiliteracies. For her PhD thesis, she wrote a metafunctionally diversified grammar of 3D space. The spaces analysed in her thesis were all museum exhibitions, both in Australia and abroad. She also chaired the development of Interpretive Strategies for the Indigenous Australian’s exhibition at the Australian Museum work for which she and her colleagues received the Premier’s Inaugural Award. Sabine Tan is Senior Research Fellow at Curtin University, Australia. She has a background in critical multimodal discourse analysis, social semiotics and visual communication. She has applied multidisciplinary perspectives for the analysis of institutional discourses involving traditional and new media. She has worked on interdisciplinary projects involving the development of interactive software for the multimodal analysis of images, videos and 360-degree videos for research and educational purposes.

NOTES ON CONTRIBUTORS

xxiii

Elizabeth A. Thomson is Adjunct Associate Professor in the Division of Learning and Teaching, Charles Sturt University, Australia. She has taught Japanese language, English for academic purposes, English for defence purposes and linguistics. Her research interests and publications are in the fields of descriptive grammar within the tradition of Systemic Functional Linguistics, applied linguistics, language for specific purposes, discourse analysis, language education, sociolinguistics and online learning. Namala Tilakaratna is Senior Lecturer at the Centre for English Language and Communication at the National University of Singapore. She draws on SFL to teach public communication, CLIL and design semiotics and has published on the use of SFL and LCT to create effective pedagogy across a range of disciplines such as social work and nursing. She is currently co-editing a book on demystifying critical reflection. Theo van Leeuwen is Professor of Language and Communication at the University of Southern Denmark. He has published widely in the areas of visual communication, multimodality and critical discourse analysis. His latest books are the third revised edition of Reading Images: The Grammar of Visual Design (with Gunther Kress) and Multimodality and Identity. Trish Weekes is a literacy researcher and academic who specializes in disciplinary literacy and Systemic Functional Linguistics theory and practice in educational contexts. Her university teaching experience includes academic literacy across faculties and pre-service teacher training. She is also the author of the Literacy Works range of SFL-based literacy resources for secondary schools. Peter R. R. White teaches linguistics with a socio-semiotic orientation and journalism studies at the University of New South Wales, Australia. He is one of the architects of the Appraisal framework, an approach to the analysis of evaluative language. His recent work focuses on persuasiveness and the putative reader in mass communicative discourse. He currently performs on saxophone and keyboards with the Travelling Circumstances, a jazz and blues combo of (mostly) linguists. Michael Wiebrands is a software developer and PhD candidate at Curtin University, Australia. His work involves the use of game engines to build educational tools, and create immersive 3D visualizations for research across the university. His own research is investigating methods to enhance realism in virtual reality-based underwater virtual heritage experiences.

xxiv

NOTES ON CONTRIBUTORS

Peter Wignell currently works as a sessional academic at Curtin University, Australia. He has a research and teaching background in Systemic Functional Linguistics, discourse analysis and multimodal analysis. His research has been both theoretical and applied. For example, his work on the role of language in the construction of specialized knowledge systems has informed literacy theory and pedagogy. Recent research has led to publication in a diverse range of applications such as brand semiotics, performance studies, translation, political discourse and online extremist discourse. Michele Zappavigna is Associate Professor at the University of New South Wales, Australia. Her major research interest is in exploring ambient affiliation in the discourse of social media using social semiotic, multimodal and corpusbased methods. She is a co-editor of the journal Visual Communication. Key books include Searchable Talk: Hashtags and Social Media Metadiscourse (2018) and Discourse of Twitter and Social Media (2012).

PREFACE

This collection of papers grew out of a suggestion from Andrew Wardell at Bloomsbury, who felt a volume dedicated to Systemic Functional Linguistics (SFL) as an appliable linguistics was timely (as part of a series of monographs, now emerging as Bloomsbury Studies in Systemic Functional Linguistics). Jim invited David and John to join him for both projects, and our appliable book and monograph series editing team was formed. The editors are much indebted to the authors for their contributions and for their patience and resilience during the editing process. It has been a pleasure working with such an exceptional group of people, and a privilege to be able to bring these chapters together. Many of the challenges faced by the authors in their research are documented in the pages that follow; the challenges of bringing this volume together during 2020–1 and the COVID-19 pandemic are not. But the patience and assistance of the editorial team at Bloomsbury, particularly over this incredibly trying time, are greatly appreciated by the editors – Andrew Wardell, Becky Holland and Morwenna Scott. The editors would also like to thank Kate Leeson for her meticulous, efficient and professional proof/copy editing of the manuscript. And a big shout-out to our loved ones, who so caringly put up with us as we burrowed down. As reflected in this volume, the editors are part of a wonderful mutually supportive community of functional linguistic scholars who are dedicated to the dialectic of theory, description and practice that enables SFL to evolve as a socially responsible paradigm. We acknowledge here the ongoing dialogue, in seminars and in regional and international conferences around the world, that

xxvi

PREFACE

has made this enterprise possible. This volume cannot make up for the lack of face-to-face interaction we are all experiencing now. But we offer it as a point of departure for the camaraderie that the future will inevitably engender for all of us. Adelaide and Sydney, July 2021

ACKNOWLEDGEMENTS

The editors and publisher gratefully acknowledge the permission granted to reproduce the copyright material in this book. Every effort has been made to trace copyright holders and to obtain their permission for the use of copyright material. However, if any have been inadvertently overlooked, the publishers will be pleased, if notified of any omissions, to make the necessary arrangement at the first opportunity.

xxviii

Introduction Introducing Appliable Linguistics J. R. MARTIN, JOHN S. KNOX AND DAVID CALDWELL

SYSTEMIC FUNCTIONAL LINGUISTICS AND SEMIOTICS This collection of papers, as our title suggests, is inspired by what Michael Halliday (2008: 7) articulated as his guiding principle: the search for what I have called an ‘appliable’ linguistics – a comprehensive and theoretically powerful model of language which, precisely because it was comprehensive and powerful, would be capable of being applied to the problems, both research problems and practical problems, that are being faced all the time by the many groups of people in our modern society who are in some way or other having to engage with language. All of our authors have been deeply influenced by Halliday’s perspective on the role that linguistics could play in social life and his foundational contributions to theory, description and practice. Many of them have extended his work to make Systemic Functional Linguistics (SFL) an ever more comprehensive and powerful model of language and to develop Systemic Functional Semiotics (SFS) as an ever more comprehensive and powerful model of semiotic systems (including language). Models of language with a theoretical, descriptive and practical commitment of this kind are not easy to find. In the twentieth century, the only comparable

2

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

model to emerge alongside SFL was Pike’s tagmemics (Pike 1971, 1982, 1988; Pike and Pike 1983), which evolved in support of the bible translation and missionary work of the Summer Institute of Linguistics. As Pike (1988) phrased his guiding principle, ‘I wanted a theory that would allow one to live outside the office with the same philosophy one uses inside it.’ The contrast with Halliday’s avowedly Marxist linguistics as far as ideology is concerned is striking (see Martin 2013). And, as Halliday comments in relation to his collaboration with the British sociologist of education Basil Bernstein, ‘From Bernstein I learnt also, for the second time in my life, that linguistics cannot be other than an ideologically committed form of social action’ (1985: 5). But the two theories in fact share a number of fundamental similarities. We will review a number of these here, because they are foundational to appliable linguistics and bear critically on the work surveyed in this volume. Perhaps first and foremost, an appliable linguistics must be a theory of meaning. Ideally this means eschewing a form–content duality in all its guises and embracing the idea that language makes meaning. Second, an appliable linguistics must engage with the fact that meaning is made in texts. Smaller units of language must of course be recognized, but all have to be interpretable in relation to co-text (i.e. the text and other units in and with which any smaller unit is situated). Third, an appliable linguistics must work with a fully articulated model of context (i.e. the social environment of which a text is an integral part). Ideally this will be a model that treats context as a higher level of meaning realized through language choices. In Hjelmslev’s (1961) terms, this means modelling context as a connotative semiotic system – so that it can be modelled in the same terms as the denotative semiotics that realize it.1 Fourth, an appliable linguistics must adopt a multimodal perspective on texts in context – and model the interaction between language and attendant semiotic systems (e.g. paralanguage, image, music, dance). Finally, in light of the complexity of models dealing with meaning, text, context and multimodality, an appliable linguistics must be fractal. Analytical tools and formalisms have to be designed that work in similar ways across a range of semiotic phenomena. This means that, once linguists and semioticians are apprenticed into a theory (via genre, appraisal, image analysis, language education etc.), they can re-deploy their skills as new tasks come to hand and make contributions to a wholistic model of language and semiosis.

SFL AS AN APPLIABLE LINGUISTICS Meaning, text, context, multimodality, fractality – quite a challenge! How have SFL and SFS evolved to meet this challenge?

INTRODUCTION

3

From Firth ([1957] 1968) SFL inherits what he proposed as the first principle of phonological and grammatical analysis: the distinction between system and structure. Halliday (1985: xxvii) tilts this complementarity towards system (the axis of choice), noting in his canonical description of the grammar of English that a ‘systemic grammar is paradigmatic … structure is … the mechanism for expressing the choices that have been made’. Introductions to SFL often orientate readers to this perspective with reference to the grammar of mood in English (e.g. Martin, Wang and Zhu 2013). At stake here is the difference between declarative, interrogative and imperative moods and how they are realized. The very basic system and structure relations involved are outlined in Table 0.1, drawing on Halliday (1985). As outlined, for declarative mood the Subject function precedes the Finite; for interrogative mood the Finite precedes the Subject and for imperative mood the Subject and Finite functions are absent. The way in which SFL privileges system over structure is reflected in the system network presented in Figure 0.1. The square brackets indicate choice – [indicative] or [imperative], and [declarative] or [interrogative] – and the choice of [declarative] or [interrogative] is dependent on the choice of [indicative]. The structural consequences of the choices are expressed as realization statements indicated by diagonal arrows – if [indicative] then a Subject function and a Finite function are present; if [declarative] then the Subject precedes the Finite (signalled by the caret ‘^’); if [interrogative] then the Finite precedes the Subject, and [imperative] has, in contrast to [indicative], no structural requirement for Subject or Finite. This paradigmatic orientation to linguistic and semiotic analysis has two very significant advantages over the more common syntagmatic perspective. For one thing it affords a relational theory of meaning (drawing on what Saussure called valeur) in which meaning is choice; this means that language and semiotic systems in general make meaning in terms of the relation between what is performed and what is not. This circumvents the pernicious form–content dualities that cripple so many theories in relation to appliable linguistics. For another, this take on the complementarity of system and structure means that paradigmatic relations can be formalized for a wide range of structural realizations (across languages, across semiotic systems, across levels of language and context etc.).

TABLE 0.1  English Mood Types and Structural Realizations Mood type

Structure

Example

Declarative

Subject ^ Finite

They have won.

Interrogative

Finite ^ Subject

Have they won?

Imperative

–

Win.

4

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 0.1  Basic English mood options and their structural realizations.

Consider, for example, the realization of mood in Tagalog (often referred to as Filipino in its guise as the national language of the Philippines). In Figure 0.2 the analysis proposes that for [major] clauses there will be a Predicator function, that for [indicative] clauses the Predicator will be realized by a [finite] verb, that for [imperative] clauses the Predicator will be realized by a [non-finite] verb, and that for [interrogative] clauses the question particle ba will follow the Predicator. The structural differences from English are considerable; but a paradigmatic orientation is sustained. The fact that this strategy can be pursued across so many dimensions of language and semiosis is what gives SFL its key fractal principle – namely deployment of system networks across a wide range of structural phenomena. Once you master the basic principles of system network design, then you can apply them to structures of many kinds – and anyone familiar with this form of description/analysis can access your descriptions via a common paradigmatic metalanguage and attendant symbolism.

FIGURE 0.2  Basic Tagalog mood options and their structural realizations.

INTRODUCTION

5

This approach to axis (i.e. system–structure relations) has led over time to what SFL and SFS refer to as ‘trinocular’ vision. This refers to the multiple perspectives that one can take on any instance of meaning within SFL theory. And these multiple perspectives are possible because of the way in which systems tend to bundle together in groups that are relatively independent of one another. We will not take time to step through the argumentation here (see Martin et al. 2013 for a detailed introduction). In broad terms, what has been suggested is outlined below. One parameter has to do with the way systems bundle according to levels of abstraction in linguistic analysis (i.e. strata) – for language, this involves discourse-semantic systems realized through lexicogrammatical systems realized through phonological ones.2 A general picture of this take on systemic interdependency is outlined in Figure 0.3, using co-tangential circles to represent the realization hierarchy. A model of this kind ensures that meaning is interpreted in relation to co-text.

FIGURE 0.3  Language strata.

6

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

From another simultaneous perspective systems bundle according to the kinds of meaning made (i.e. metafunctions) – for many semiotic systems, as ideational systems construing what is going on or being talked about, as interpersonal systems enacting social relations and as textual systems composing information flow. An outline of this perspective on systemic interdependency is presented in Figure 0.4, adapting yin–yang imaging as a reminder that, in order to make meaning, texts draw on all three metafunctions at the same time. A model of this kind highlights the importance of all kinds of meaning to developing understandings of texts, groups of texts and whole languages. Metafunctions have proven an important jumping off point as far as the evolution of SFS and SFL is concerned (e.g. Kress and van Leeuwen 1990; Painter, Martin and Unsworth 2013; Ngo et al. 2022). From another simultaneous perspective systems bundle according to the size of unit being described (i.e. ranks) – for grammar, clauses made up

FIGURE 0.4  Ideational, interpersonal and textual metafunctions.

INTRODUCTION

7

of groups and phrases, groups and phrases made up of words, and words (depending on the language) made up of morphemes. An outline of this perspective on systemic interdependency is presented in Figure 0.5. A comparable scale for English phonology would cover the units tone group, foot, syllable and phoneme. This kind of system bundling in fact varies across strata, languages and many semiotic systems – sensitive as it is to how many layers of constituency we find. In SFL, trinocular vision of the metafunctional kind functions as a bridge between language and models of social context. The intrinsic functionality of language (i.e. ideational, interpersonal and textual meaning) is mapped onto context in terms of ideational meaning by and large construing field, interpersonal meaning by and large enacting tenor, and textual meaning by and large composing mode. In Martin (1992), this extrinsic functionality (i.e. field, tenor and mode) is referred to by the cover term register; and register is thus modelled as a higher-order connotative semiotic realized through language. Martin’s model includes an additional stratum of context called genre; at this ultimate level of abstraction a culture is mapped as a system of genres – realized through choices in field, tenor and mode in the staging and phasing of unfolding text (Martin and Rose 2008). This stratified model of language (phonology, lexicogrammar and discourse semantics) and context (register and genre) is outlined in Figure 0.6. A model of this kind ensures that meaning is ultimately interpreted in relation to co-text in context. The SFL approach of stratifying language and social context has been extremely influential as far as applications of the theory are concerned – as the chapters in this volume make clear. Its strength in this regard derives from the way it grounds a model of social processes (e.g. register and genre) in language (e.g. via discourse semantics and on through lexicogrammar and phonology). A little more detail about the interconnections in play is offered in Figure 0.7, which interfaces genre and register with discourse semantic systems (Martin 1992; Martin and White 2005; Martin and Rose 2007). In these terms, the discourse-semantic systems of connexion and ideation are seen as construing

FIGURE 0.5  A rank scale for English grammar.

8

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 0.6  Language, register and genre.

the field systems of activity, taxonomy and property, just as negotiation and appraisal are modelled as enacting the tenor systems of status and contact, just as the identification and periodicity are read as composing the mode systems of dialogue/monologue and action/reflection – with all of these engendering genre. We will not push the model further into the detail of lexicogrammar and phonology here. But these connections are crucial to the grounding of the social in language noted above. In this section we have provided a very brief overview of SFL, focusing on the principal parameters that have made it the most powerful model of appliable linguistics developed to this point in the history of linguistics. For a survey of some of its most relevant evolution, see Martin (2014, 2016). For a basic introduction to the model we recommend two bilingual (English and

INTRODUCTION

9

FIGURE 0.7  Register (field, tenor, mode) and genre.

Chinese) publications by the Higher Education Press in Beijing: Matthiessen and Halliday’s Systemic Functional Grammar: A First Step into the Theory (2009) with Martin, Wang and Zhu’s Systemic Functional Grammar: A Next Step into the Theory – Axial Relations (2013). In addition we recommend Martin, Quiroz and Wang’s Systemic Functional Grammar: A Text-based Description of English, Spanish and Chinese (in press). Foundational papers have been collected in Martin and Doran (2015a, 2015b, 2015c, 2015d, 2015e). And Matthiessen, Teruya and Lam (2010) provide a comprehensive glossary of terms. We close this section with a comment on our understanding of the terms theory, description and practice. In this book theory is understood as the basic principles from which descriptions of particular semiotic phenomena

10

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

are derived. SFL theory is founded on its interpretation of axial relations (i.e. its privileging of paradigmatic relations over structural ones) and the parameters of stratification, metafunction and rank which generalize relations among bundles of axial relations across languages and semiotic systems. SFL descriptions draw on these theoretical parameters to produce descriptions. In Figures 0.1 and 0.2, for example, we outlined in simple terms some basic system and structure relation for mood in English and the mood in Tagalog. Theory is general; descriptions are specific. Axis, stratification, metafunction and rank can be drawn on again and again as we move from one language and culture to another and across semiotics systems. But a description is always specific to the phenomenon being described. What we describe as mood in English is not mood in Tagalog, just as what we describe as interpersonal meaning in one or another language is not interpersonal meaning in image or paralanguage. Respecting this complementarity of theory and description is critical in appliable linguistics; otherwise, we run the risk of foisting the description of one linguistic or semiotic phenomenon onto another one – without respecting the distinctive affordances of the alternative phenomenon we are attempting to describe. This is absolutely crucial for practice – as we draw on theory and description in sites where we want to make a change. Making change is, of course, a challenging enterprise. It regularly demands new descriptions, and these demands may in turn demand new theory – an interaction of theory, description and practice often referred to as a dialectical process. In Martin’s work, for example, we can observe a number of these interactions. His discourse-semantic descriptions of English identification and connexion were driven by his concern to explain how psychiatrists diagnosed thought disorder in schizophrenic speech (Rochester and Martin 1979). And these descriptions in turn rebounded on SFL theory – in relation to his stratification of SFL’s content plane as lexicogrammar and discourse semantics. Turning from clinical to educational linguistics, work by Martin and his colleagues to open up access to academic literacy led to descriptions of text types as genres. And this rebounded onto SFL’s theory of context – and led to stratification of SFL’s context plane as register and genre. Comparable stories in the history of SFL abound (in relation to appraisal and SFL’s approach to interpersonal meaning and prosodic structure, for example, or in relation to images and the evolution of Systemic Functional Semiotics). This volume tells many of these tales. Two things we have surely learned: (i) theory is every practitioner’s best friend, and (ii) there is no better way to improve theory and description than trying to use it to change the world. Enjoy our stories; in the next section we give a taste of what is to come.

INTRODUCTION

11

THE CHAPTERS IN THIS VOLUME The chapters in this volume demonstrate the fundamental dialectic between description, practice and systemic functional theory, and we have structured the volume accordingly. As outlined below, we begin with chapters that foreground innovative description, then innovative practice, developments in theory and finally current practice in SF-inspired appliable linguistics. Innovative description The initial section, on ‘innovative description’, presents appliable linguistic studies that have critiqued, developed and in some cases transformed the way in which a particular field has been described. We begin this section with Hood’s account of an early application of the system of appraisal to a study of evaluative strategies in academic research papers. Hood explains that phases of texts construing one or another field were found to associate with distinctly contrastive attitudinal profiles, which in turn led to an elaboration of system options in graduation as both force and focus. Hood’s chapter concludes by returning to the field of enquiry to show how her research has transformed description in academic research papers. Caple then recounts her application of SF theory in the context of the newspaper industry in Australia, and in particular the profession of photojournalism. Her research was conducted in response to a pedagogical problem – how to apply existing SF and multimodal theory to news photographs in teaching journalism/media studies. Her response was to apply the theoretical tool of system networks, in conjunction with insights from gestalt psychology and her professional background in photojournalism. The outcome was a framework for analysing composition – the balance network – which has since been applied and adapted beyond photography to the analysis and description of children’s picture books, television title sequences and cohesion between images, text and their layout. Also in the field of media discourse, Knox’s chapter provides an account of the development of a rank scale in a study of online newspapers. Grappling with novel texts and structures, including home pages, story pages, news stories, photographs and headline-based hyperlinks, Knox begins his recount with a provocation from a colleague who queried his use of rank (or lack thereof) in his analysis of online newspapers. Knox accordingly takes the reader through his application of rank scale to online newspapers, with a particular focus on the home page. Recounting the various obstacles and challenges of describing online newspapers through a constituency-based approach, Knox concludes his chapter with an outline of the various benefits of this approach, including: theoretical creativity, rankshift as a means for understanding website

12

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

structure and home pages, and the generative benefits of a systemic perspective to describing online newspapers. The final chapter in this section is Caldwell’s account of appliable linguistics in the context of vocal sound, and in particular hip-hop culture and rap music. His chapter recounts the application of SF theory to Kanye West’s vocal performances, with the aim to develop a novel description of the rap and sung voice through system networks and visual representations. Throughout the chapter, Caldwell recounts the constraints and affordances that were presented when developing innovative description through SF theory in the contested, interdisciplinary context of popular music studies. Ultimately, his system networks and visual representations present a development to description in that field – foregrounding choice, embodiment and insights from other disciplines – to describe the meaning-making potential of performance voice in hip-hop culture. Innovative practice This section, on ‘innovative practice’, presents studies that have, through the application of SF theory, critiqued, developed and transformed practice in a respective field. We begin this section with Thomson’s work with the military, and specifically the Australian Defence Force. Thomson shows how SF-inspired genre pedagogy provided a strategy to enable design, development and delivery of a new LOTE (languages other than English) programme at the Defence Force School of Languages, in an environment where the target language requirements were classified, and therefore not available to the teachers and curriculum developers. Drawing on the predictive power of genre, Thomson recounts the construction of a curriculum to support military personnel. Her chapter illustrates the appliability of genre analysis to transform an area of language teaching of which little is known, and which is rarely discussed, in the civilian community. Rose’s chapter on the Reading to Learn (R2L) curriculum intervention is especially ambitious. It details his description of a pedagogic register, extending the application of SF tools from denotative to connotative semiotics. Drawing on a traditional story, and registerial phase analysis, Rose presents a close analysis of knowledge genres and the structuring of pedagogic activity in curriculum genres. Rose’s final reflection is particularly critical to this section, as he presents R2L as an exemplar of innovative practice, with implications for the field of education, as well as a set of key future directions for scholars seeking to examine pedagogical practice through SFL. Kim’s chapter opens with an overview of the author’s professional training and experience in translation, and an exploration of SFL as a means to research

INTRODUCTION

13

translation studies and develop teaching practice. It progresses to the description of Korean, and explores how a systemic approach to language description has solved long-standing issues in the description of Korean grammar, which in turn illustrates how the in-progress SF grammar of Korean is a productive tool for the practice of translation and language professionals working with the Korean language. Kartika recounts the implementation of the R2L pedagogy in a multilingual setting in Indonesia. Kartika explains that the R2L pedagogy is of particular relevance in Indonesia, which adopted genre pedagogy in 2004 in their national English-language school curriculum. She recounts a curriculum intervention that involved re-designing an existing pedagogical approach to include a principled progression of language of instruction and classroom interactions (from predominantly L1 to a greater use of the L2). Kartika reflects on the linguistic complexity of her context, and the considerable language-related challenges facing curriculum developers. At the same time, her chapter is an exemplar of innovative practice – mapping a theoretically grounded approach to dealing with the challenges of multilingualism in a national English-language school curriculum. O’Halloran, Tan, Wignell, Lange, Chai and Wiebrands argue that largescale analysis and corpus-based empirical grounding and testing of insights are required to transform the theory and practice of multimodal analysis beyond existing levels of description, generalization and demonstration of concepts. In their chapter, they take up this challenge, and present an approach to multimodal discourse analysis which involves integrating multidisciplinary theories and techniques of multimodal analysis with computer vision, natural language understanding systems and machine learning. They specifically apply their context-based mixed-methods approach to online violent extremist content to examine how the modes of image and language work together to promote or resist a terrorist agenda. Their chapter flags with readers the opportunity to apply and develop their context-based mixed-methods approach, as well as to transform the practice of multimodal discourse analysis generally. In the next chapter, Polias and Dare present an encouraging story of innovative practice which has transformed the field of language teaching and learning in schools. Drawing on their extensive experience as teacher educators, Dare and Polias recount their experience with the early genre-based pedagogy, followed by a description of how they set about building teachers’ capacity to take up SFL metalanguage. Their reflection includes some of the personal and political challenges faced in promoting SF theory to the teaching profession in Australia. Ultimately, their recount is one of success, as they show not only a strong take-up of their professional learning materials, but also significant impact on students’ and teachers’ language practices across the world.

14

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Moyano’s chapter outlines a pedagogical revision of the traditional genrebased teaching-learning cycle and its application to two Argentinian universities, both involving academic Spanish at an undergraduate level. The traditional practice of teachers providing written feedback (including corrections) on students’ written work was explicitly avoided in the development and design of a curriculum, with a preference for scaffolded, in-class pedagogical stages involving both teachers and students in the editing process. This chapter demonstrates that ‘innovative practice’ is often realized as an adaption of traditional and valued pedagogical approaches based on the demands of particular teaching-learning contexts. Byrnes’s chapter provides a historical overview of the intellectual context of second- and foreign-language education in the 1990s, and the polarized situation that existed in the research, teaching practices and dominant professional organizations of the time. In this context, Byrnes and colleagues in the German Department at Georgetown University drew on Halliday’s theories of language, on SFL genre-based approaches to curriculum development, and on current research and practice in language education (notably task-based language teaching). In addition to documenting key issues in the process of curriculum development, Byrnes presents a perspective on learner development that is significantly different from the majority of the literature on language education. She explains why development in instructed language learning is fundamentally a curricular, rather than a pedagogical, issue, and also why research into adult instructed language learning is inherently connected to curriculum, since longitudinal studies reflect the design and implementation of the curriculum as much as they do the developmental trajectory of a learner. The final chapter of this section also explores the field of undergraduate education, this time in Australia. Dreyfus and Weekes detail the process of establishing and developing a compulsory first-year academic literacy course, based on SFL, for students entering a Bachelor of Arts. They discuss the course, the development of the concept for the unit, collaboration with discipline academics, analysis of student writing and assessment task prompts, and design of the course. Like many of the chapters presented in this section, and the volume more broadly, their chapter tells a story of challenges and opportunities when attempting to change the practice of teaching and learning academic literacy, through SFL, in a tertiary context. Evolving theory The third section, on ‘evolving theory’, presents first-hand accounts of both seminal and contemporary contributions to SF theory, arising from its application to a range of fields of practice.

INTRODUCTION

15

This section begins with Figueredo and Lima’s recount of how SFL was applied to a language problem in the field of Newborn Screening Genetics Diagnosis, specifically sickle cell disease. Recounting their text analysis of research articles, technical guides and pamphlets/patient information leaflets, they argue that efficient models of texts and text production proved to increase efficiency in building and transferring knowledge between specialists, technicians and laypeople, which resulted in an improved quality of sickle cell disease management. This included an increase in pamphlet and leaflet production and distribution, resulting in a public health campaign at state level. Ultimately, their chapter also reveals several contributions to SF theory, including a greater understanding of the quantification of meaning-making processes in systems, the placement and ordering of valeur-reactants in the Gleason Scale, displacement in language space and the description of ‘text-asinstance’ as the unit of displacement. Van Leeuwen’s chapter re-visits the seminal publication Reading Images, which he co-authored with Gunther Kress. Providing a historical perspective on the development of multimodality and its relations with linguistics, he in turn re-visits the work of Halliday, and in particular Arnheim. In doing so, limitations in the original version of Reading Images and areas of development in the recent revision of the book are explored. Given the enormous impact of Reading Images, the issues dealt with in this chapter present fundamental questions for the evolution of multimodality and SF theory. Van Leeuwen concludes his chapter by considering the place of multimodality and linguistics in the academy, and society more broadly. In the next chapter, White recounts his work with journalistic discourse and the development of the appraisal framework. This framework aimed to provide theoretically principled and linguistically based input into ‘debates’ around news texts, and the frequently asserted distinction between ‘objective’ and ‘subjective’ reporting. White reports on the theoretical insights which thus became possible, including identifying the genres of news coverage, explaining the interpersonal workings of hard news reporting, describing the different sub-registers or ‘voices’ of news journalism, and providing a linguistic account of how journalistic commentary negotiates both alignment and disalignment with its intended readerships. In many ways, this chapter presents the origin story of the appraisal framework. Working within the field of museum studies, Stenglin revisits her important work in museums, including their varied displays involving objects, images, dioramas and interactives of all kinds. In particular, she recounts her development of two constructs – bonding and binding – analytical tools first developed for exploring museum spaces, and how they work to attract and engage the interest of visitors. Bonding grew from the need to understand resources for community formation in museums, while binding explores how

16

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

3D space can be organized to make visitors feel secure. Recounting three key phases of her research, Stenglin documents the theorizing required to capture the inherent function of a museum – to attract, inform and inspire patrons – and shows how her general theory of 3D space developed from more specific investigation and description. Logi and Zappavigna’s chapter explores the evolution of affiliation – a model for understanding how people negotiate values and form communities. Foregrounding the evolution of this model, and drawing on first-hand experience in its evolution, Logi and Zappavigna trace affiliation theory from earlier work on casual conversation and the concepts of couplings and bonds. The chapter then turns to more recent adaptations of a model for affiliation in registers that ‘do’ dialogue in non-traditional ways, namely social media and Twitter. Here, the authors illustrate the inevitable dialectic between engaging in new fields of practice on the one hand, and the evolution of SFL theory (in this case, the model of affiliation) on the other. In the following chapter, Doran provides an insightful first-hand account of his recent ‘grapplings’ with the field of physics, and in particular the semiotics of mathematical symbolism in physics. Rather than taking the theoretical tools of language as the point of departure, Doran, inspired by parallel contemporary work in SF language typology, pushes back against the theory, and the rich traditions of describing multimodality through an SF theoretical framework. In its place, he presents a semiotic description of mathematics that does not presume metafunction, rank or strata, and instead foregrounds axis and the logical component of mathematical semiosis. In short, Doran’s chapter presents an exemplar of the dialectic between theory, practice and description, in this case, a description of practice which calls for a significant re-thinking of how SF scholars work with the theory in semiotic modes beyond English linguistics. In the final chapter of the ‘evolving theory’ section, Szenes looks at ‘big’ texts, in this case, detailed and complex ‘country reports’ that tertiary students are required to write for a business course. Drawing on the SFL literature on macro genres, Szenes argues that these texts are actually genre simplexes, with embedded genres at different levels that can be modelled in terms of constituent structure and also in terms of orbital structure. Arguing that genre embedding is more common than previously acknowledged in the SFL literature, this chapter shows how re-visiting, applying and questioning the assumptions made in existing theory can enrich our understanding of language and SF theory. Appliable linguistics in progress The final section, on ‘appliable linguistics in progress’, captures the insights of scholars and studies who are currently engaged with an emerging dialectic of practice, description and theory.

INTRODUCTION

17

The first chapter from Macnaught presents cutting-edge research into the question of ‘phase’ in classroom discourse, and where it fits theoretically in the stratification hierarchy of curriculum genre, pedagogic register, discoursesemantics, lexicogrammar and in terms of rank and class-function cycles. Macnaught builds on the work of Rose and colleagues, positioning the notion of phase in the theory in a way that has not been previously done, despite the decades of SFL research that has used this notion, and the decades of SFL research on classroom discourse. In doing so, she demonstrates that even familiar theoretical constructs and familiar contexts can still raise fundamental issues for researchers and practitioners, and for the theory they work with. Hao’s chapter on Chinese in the context of history textbooks in China reveals an emerging dialectic between practice, description and theory. She argues for a shift away from an exclusively ‘bottom-up’ perspective to descriptions of Chinese, and instead presents a ‘top-down’ approach. This approach draws insights from SF approaches to describing various languages, taking a trinocular perspective on Chinese texts. In addition, it draws on recent descriptions of field and discourse semantics to provide a useful and productive method of investigating how disciplinary knowledge is construed in Chinese, with texts as the unit of analysis. In the next chapter, Djonov and van Leeuwen outline their research on software as a semiotic technology. They developed a three-dimensional model for studying semiotic software which they applied to Microsoft PowerPoint presentation software, and to the academic social media platform ResearchGate. On the basis of their research, they identify several key differences between semiotic software and social media platforms, and distinguish between semiotic practices and social practices. In terms of the process of research, they demonstrate the importance of taking a diachronic perspective, and of questioning and adapting the theoretical tools of SF theory according to the object of and the social context of the research, substituting key analytical tools of SF theory (e.g. system networks) for others more suitable for particular semiotic resources, and employing ethnographic methodologies to better understand those semiotic resources and the semiotic and social practices to which they contribute. Ramírez’s chapter returns to R2L to describe a bilingual adaption of that approach involving Spanish-speaking, language-minority mothers enrolled in an ESL family literacy programme in south-east United States. After outlining the linguistic environment, he explains his bilingual adaptation of R2L, and recounts in detail the positive teaching and learning experience had by those in the programme. Of particular significance to this project is the instruction provided to the mothers/students. This instruction enabled them to systematically use the same R2L pedagogical sequences they experienced to effectively engage their young Spanish-speaking children with picture books at

18

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

home. Ramírez concludes by considering more broadly the emerging issues that have come out of his research, with particular reference to the dialectic of R2L, SFL theory and the practice of bilingual education. Crane’s chapter recounts her experiences using SFL, and genre pedagogy specifically, to help revitalize the language of Gija – an endangered language from the Kimberley region of Western Australia. Drawing on her community knowledge, and expertise in SFL, she outlines the challenges faced by linguists, senior expert speakers and Aboriginal Teachers’ Assistants as they work together to develop school-based programmes. She argues for, and demonstrates how, SFL-based language typology and genre pedagogy can help in the process of empowering Aboriginal educators with the best tools and understandings for effective teaching of, in and about Gija. Tilakaratna explores the construal of Sri Lankan national identity in Englishlanguage textbooks produced and disseminated at the end of the thirty-yearlong separatist ethnic conflict in that country. Drawing on the SFL framework of iconography – how people and things are charged with interpersonal meaning – the chapter shows how selected texts inculcate students with a normalized understanding of majoritarian nationalism. The chapter argues that identifying icons and mapping their logogenetic development uncovers how governments can ‘allocate’ semiotic resources in pedagogical contexts. In the final chapter of this section, Logi, Zappavigna and Martin present a framework for exploring the often neglected semiosis of paralanguage. They argue that paralanguage is dependent on language, and realized through both sound quality and body language (including facial expression, gesture, posture and movement). Their model builds on the work of Cléirigh, distinguishing between sonovergent (convergent with prosodic phonology) and semovergent (convergent with semantics/the content plane of language) paralinguistic systems. To illustrate their model, the authors draw on former Australian prime minister Julia Gillard’s renowned ‘misogyny speech’. In addition to Cléirigh, the authors contextualize the framework with reference to previous SFLoriented paralinguistic analyses, in particular, research into New South Wales youth justice conferences, which also included Cléirigh. In this way, the context provided in this chapter is not simply presented as a precursor to the current model, it serves as a further exemplum of an emerging dialectic of practice, description and theory.

CONCLUSION Collectively, the chapters in this volume indicate the appliability of SFL’s theory, and also the extent to which the theory has developed and expanded in response to the range of contexts and problems to which it has been applied.

INTRODUCTION

19

The development of systemic functional theory beyond English,3 beyond grammar and beyond language, shows the adaptability and explanatory power of a theory of meaning that privileges system, while keeping sight of structure, of whole texts and of context. The chapters provide an account of how scholars, practitioners and scholarpractitioners have developed their understanding of social and semiotic phenomena by taking a principled, theory-informed approach to research; and how these explorations have been an essential and integral aspect of the development of the theory. Our vision for this volume was a focus on the processes of doing research and of building theory. We wanted to provide an opportunity for authors to document their experience so that novice and experienced researchers and practitioners alike could gain insight into how and why various theoretical and practical issues have been addressed. This, we believe, is a valuable complement to the ‘outcomes-based’ approach to reporting research, which is typical of the journal articles and book chapters in which this work has already been reported. We hope these stories inspire readers to explore their own contexts, and in doing so develop the theory in ways not yet imagined.

NOTES 1 For Hjelmslev a denotative semiotic is a semiotic system that has its own expression form, whereas a connotative semiotic uses a denotative semiotic as its expression form. 2 In this chapter, we draw on Martin’s approach to systemic functional theory. See e.g. Bartlett and O’Grady (2020) for discussion of different approaches in SFL, all of which rely fundamentally on Halliday. 3 Halliday’s initial work on language description was on Chinese languages. His work on English came later.

REFERENCES Bartlett, T. and G. O’Grady (2020), The Routledge Handbook of Systemic Functional Linguistics, London: Routledge. Firth, J. R. (1968), ‘A Synopsis of Linguistic Theory’, in F. R. Palmer (ed.), Selected Papers of J. R. Firth 1952–1959, 168–205, London: Longman. Halliday, M. A. K. (1985), An Introduction to Functional Grammar, London: Edward Arnold. Halliday, M. A. K. (2008), ‘Working with Meaning: Towards an Appliable Linguistics’, in J. J. Webster (ed.), Meaning in Context: Implementing Intelligent Applications of Language Studies, 7–23, London: Continuum. Hjelmslev, L. (1961), Prolegomena to a Theory of Language, Madison, Wisconsin: University of Wisconsin Press.

20

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Kress, G. and T. van Leeuwen (1990), Reading Images, Geelong: Deakin University Press. Martin, J. R. (1992), English Text: System and Structure, London: Benjamins. Martin, J. R. (2013), Interviews with Michael Halliday: Language Turned Back on Himself, London: Bloomsbury. Martin, J. R. (2014), ‘Evolving Systemic Functional Linguistics: Beyond the Clause’, Functional Linguistics, 1: art. 3. Martin, J. R. (2016), ‘Meaning Matters: A Short History of Systemic Functional Linguistics’, Word, 61 (2): 35–58. Martin, J. R. and D. Rose (2007), Working with Discourse: Meaning Beyond the Clause, London: Equinox. Martin, J. R. and D. Rose (2008), Genre Relations: Mapping Culture, London: Equinox. Martin, J. R. and P. R. R. White (2005), The Language of Evaluation: Appraisal in English, London: Palgrave. Martin, J. R. and Y. J. Doran, eds (2015a), Around Grammar: Phonology, Discourse Semantics and Multimodality, London: Routledge. Martin, J. R. and Y. J. Doran, eds (2015b), Context: Register and Genre, London: Routledge. Martin, J. R. and Y. J. Doran, eds (2015c), Grammatical Descriptions, London: Routledge. Martin, J. R. and Y. J. Doran, eds (2015d), Grammatics, London: Routledge. Martin, J. R. and Y. J. Doran, eds (2015e), Language in Education, London: Routledge. Martin, J. R., B. Quiroz and P. Wang (in press), Systemic Functional Grammar: A Text-based Description of English, Spanish and Chinese, Cambridge: Cambridge University Press. Martin, J. R., P. Wang and Y. Zhu (2013), Systemic Functional Grammar: A Next Step into the Theory – Axial Relations, Beijing: Higher Education Press. Matthiessen, C. M. I. M. and M. A. K. Halliday (2009), Systemic Functional Grammar: A First Step into the Theory, Beijing: Higher Education Press. Matthiessen, C. M. I. M., K. Teruya and M. Lam (2010), Key Terms in Systemic Functional Linguistics, London: Continuum. Ngo, T., S. Hood, J. R. Martin, C. Painter, B. Smith and M. Zappavigna (2022), Modelling Paralanguage using Systemic Functional Semiotics, London: Bloomsbury. Painter, C., J. R. Martin and L. Unsworth (2013), Reading Visual Narratives: Image Analysis of Children’s Picture Books, London: Equinox. Pike, K. L. (1971), Language in Relation to a Unified Theory of the Structure of Human Behavior, The Hague: Mouton. Pike, K. L. (1982), Linguistic Concepts: An Introduction to Tagmemics, Lincoln, NE: University of Nebraska Press. Pike, K. L. (1988), ‘Bridging Language Learning, Language Analysis, and Poetry, via Experimental Syntax’, in Deborah Tannen (ed.), Linguistics in Context: Connecting Observation and Understanding, 221–45, Norwood, NJ: Ablex. Pike, K. L. and E. G. Pike (1983), Text and Tagememe, London: Frances Pinter. Rochester, S. and J. R. Martin (1979), Crazy Talk: A Study of the Discourse of Schizophrenic Speakers, New York: Plenum.

PART I

Innovative Description

22

CHAPTER ONE

Graduation in Research Writing Managing the Dual Demands of Objectivity and Critique SUSAN HOOD

INTRODUCTION As an appliable linguistics (Halliday 1985: 7), the interaction of theory and application has a central place in Systemic Functional Linguistics (SFL), as discussed in the introduction to this volume. Fields of application are valued as grounds for positive social impact but also for their potential to nudge theory forward towards greater explanatory power. Rothery explains this simultaneous process of applying and interrogating theory thus: ‘From a linguistic perspective the lexicogrammar and discourse semantic analyses of texts … tests and challenges the systemic functional model of language in constructive ways. Such analyses enable an ongoing dialectic between the realisation of system in text and the construction of system networks’ (1996: 88). This chapter presents an account of how a relatively early application of the emerging discourse semantic system of appraisal led to elaborations and extensions of the theoretical modelling of graduation. The relevant period of research – a doctoral study undertaken from 2000 to 2004 – functions as

24

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

the hub for reflection. I look back to motivating concerns and the theories of interpersonal meaning available at the time, first in the broader applied linguistics field and then more specifically in the evolving system of appraisal in SFL. I look around to key influences from the insights of others and from shared discussions and comment on contributions emerging from this research process. I then project forward to consider ways in which an expanded system network of graduation has contributed to greater awareness of the dynamic rhetorical strategies deployed in academic research writing (Hood 2010). Along the way I comment on revisions to terminology used in describing couplings of graduation and ideational meanings, as they are indicated in networks. These flow on from recent renovations to modelling field and the discourse semantics as ideation (e.g. Hao 2015, 2020; Doran and Martin 2021). I conclude with a brief note on a current extension to the trajectory of research on graduation – one that describes the invocation of attitude through embodied paralanguage in the spoken discourse of academic lectures (Hao and Hood 2019; see also Ngo et al. 2022).

SOME BACK STORY IN THE FIELD OF ACADEMIC WRITING A central concern raised by academic colleagues and resonating with my own experience was how better to respond to the problems students as novice researchers had in effectively positioning their studies in introductions to their dissertations. To some extent the students were caught in a bind of our making. On the one hand advice came in the form of instructions to de-personalize their discourse; to avoid, for example, explicit subjective projections such as ‘I think … ’ or ‘In my opinion … ’, or the use of personal pronouns for self-reference, as well as lexicogrammatical intrusions from more spoken registers (e.g. ‘Actually’, ‘it might be … ’). On the other hand, and in apparent contradiction to this praise for ‘objectivity’, came frustrated pleas from supervisors for students to take a ‘critical’ stance in reporting on relevant contributions. Clearly didactic responses in either or both terms were an inadequate basis for supporting students’ progressive management of affective rhetorical strategies. What was needed was a much greater understanding of how an evaluative stance is enacted in the writings of valued research papers in relevant fields of study, what resources are deployed, what rhetorical strategies are configured and ultimately how such strategies make sense of the dual demands for objectivity and critique. From such a base we would be in a position to identify in much more meaningful ways the specific challenges faced by students and the resources they need to address them – bases for effective pedagogic intervention. The proposed doctoral study was shaped in these terms.

GRADUATION IN RESEARCH WRITING

25

LEARNING FROM THE PAST: ESTABLISHING A THEORETICAL BASE FOR THE ANALYSIS OF EVALUATION IN ACADEMIC DISCOURSE The research context involved a study of evaluative meaning making in published research papers and in those of novice student researchers. While evaluative meaning in academic discourse is now the focus of a great deal of academic research and practice in the field of applied linguistics, at the commencement of the study it was still a relatively emergent field in the research literature. Early contributions were mainly positioned within the broad field of pragmatics. There, influential contributions could be categorized in terms of where and how they attended to language. Corpusbased studies tended to focus on frequencies of discrete grammatical structures (adverbials, tense etc.), interpreted as in some way contributing to evaluative meaning (see e.g. Biber, Conrad and Reppen 1998; Conrad and Biber 2000). Studies of vague language (e.g. Channell 1994; Myers 1996) focused on particular lexical choices; Myers noted their strategic role in negotiating claims in academic writing. Explorations at the discourse level included studies of the rhetorical impact of various forms of citation (e.g. Swales 1990; Hyland 1999) and resources contributing to the generalized notion of hedging (e.g. Myers 1996; Salager-Meyer 1997; Hyland 1998). While the pragmatics literature served well to draw attention to the intrinsic rhetorical function of academic discourse, there were considerable limitations to be considered. The positioning of pragmatics in relation to the structuring of language (the ‘syntax, semantics and pragmatics’ trilogy) includes a form– function opposition (syntax vs semantics) and positions interpersonal meaning (pragmatics) as something that language does, rather than something that is an intrinsic part of language. In the absence of a coherent theory of the relationship of language choices to interpersonal meaning, research in pragmatics is strongly reliant on intuition for the interpretation of meaning and on justifications made post hoc in relation to syntactic choices. A consequence of this has been a proliferation of relatively loosely defined terminology for generalizing across sets of structural choices (e.g. hedging, politeness). Relatively weakly defined boundaries and intuitive interpretations both challenged the consistent application of concepts in research and seemed to offer little to shape pedagogic interventions concerned with the strategic use of linguistic resources for meaning interpersonally as texts unfold. Systemic Functional Linguistics offered an alternative theoretical approach and the potential for a more comprehensive and systematic account of the diversity of linguistic resources deployed and their interaction in the enactment of stance in written academic texts. A number of basic theoretical concepts distinguish a social semiotic approach to studying evaluative meaning in SFL from that of pragmatics. First SFL

26

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

constitutes a relational theory of meaning rather than a referential one and models meaning as valeur in networks of paradigmatic system-structure choices. The interpretation of meanings in text necessarily references these system networks. Secondly a metafunctional perspective on meaning sees text as simultaneously construing ideational, enacting interpersonal and composing textual meaning. It makes explicit the interpersonal dimension as something that language is, not just something that it does, and at the same time alerts the researcher to relations across metafunctions (i.e. how choices in one metafunction might impact on another). My research focus centred on interpersonal meaning, but an appreciation of the rhetorical impact of evaluative choices required me to attend closely to ideational and textual choices as well. Finally, SFL’s tri-stratal model of language (i.e. phonology, lexicogrammar and discourse semantics) means that choices in discourse semantic systems can be enacted through more than one lexicogrammatical construction. In a study of evaluation this supports the researcher (and hopefully in turn teachers and students of academic writing) to identify recurring abstracted rhetorical strategies in texts. These theoretical foundations will become evident in the descriptive account to follow.

KEEPING IN TOUCH WITH WHAT’S AROUND IN SFL An important impetus for closer consideration of interpersonal meaning in discourse came in the mid-1980s, prompted by Poynton’s (1985) proposals for tenor relations – the simultaneous dimensions of ‘power’, ‘contact’ and ‘affect’, each underpinned by particular principles (see also Poynton 1996). ‘Affect’ implicated the principle of ± amplification, and this constituted a foundation for later elaborations of attitude and graduation in the system of appraisal in Martin (2000), in Martin and Rose (2003) in a chapter on appraisal in the first edition of Working with Discourse, and in Martin and White’s (2005) The Language of Evaluation. Related SFL work on interpersonal meaning in discourse from the 1980s and early 1990s focused on exchange structure (Berry 1981; Ventola 1987; Martin 1992a). This work evolved as the discourse semantic system of negotiation, which models kinds of dialogic interaction and the roles adopted by or assigned to speakers (see Martin and Rose 2007). While the system of negotiation modelled the inter-dimension of the interpersonal, as Martin (2000: 144) notes, it omitted ‘the semantics of evaluation – how the interlocutors are feeling, the judgements they make, and the value they place in the various phenomena of their experience’. This was a further impetus for a complementary theoretical model of the ‘personal’ in interpersonal meaning. Prior to the elaboration of appraisal as attitude in Martin (2000) and extensions into engagement in Martin and White (2005), the most significant contribution to the emergent modelling of evaluative meaning was Martin’s

GRADUATION IN RESEARCH WRITING

27

(1992b) ‘Macro-proposals: Meaning by Degree’. Inspired by earlier contributions of Sapir ([1949] 1994: 123), Martin explores gradable systems of meaning in English, noting that choices within these systems always ‘enter into oppositions concerned with the evaluation of experience’ (1992b: 366). This source was a valuable reference throughout my study and remains significant given that the subsystem of graduation in appraisal received relatively minor attention in Martin and White (2005). Therein Martin’s major interest in the system of attitude and White’s primary focus on engagement left graduation as the runt of the litter. Discussion of graduation in that source is mainly in terms of the grading of inscriptions of attitude. While some examples include instances of couplings with ideational meaning, there is little discussion of its significant role in invoking attitude. During the course of my research there were, of course, many opportunities to engage with other concurrent studies and theoretical developments through seminars, conferences and publications. Other applied studies appearing around this time included those in media discourse (e.g. White 2003) and school discourse (e.g. Rothery and Stenglin 2000; Macken-Horarik 2003). SFL’s interest in academic discourse at the time tended to focus on genre, field and mode rather than tenor and interpersonal meaning. While these studies seemingly held no direct association to my own research, they did at times encourage a fresh perspective on data. One such encounter was with Frances Christie’s work on classroom discourse (e.g. 1997, 1999). This proved especially significant in shaping the directions of my own study. Christie’s analysis of curriculum genres drew on Bernstein’s (1990) work on pedagogic discourse, in particular his concepts of regulative and instructional ‘discourses’. In linguistic terms, Christie interpreted this as a distinction of regulative and the instructional registers of language and proposed that a relationship of projection held from one to the other. As she explained, the relationship of the two is so intimate, it is argued that the regulative register ‘projects’ the instructional register, where the term is used metaphorically from the functional grammar, following Halliday’s advice (1979, 1981, 1982) about the value of thinking grammatically about a text, modelling its organisation on that of the clause. Where a relationship of projection applies, a secondary clause is said to be projected through the primary one (Halliday 1994: 219), so that something either said or thought hitherto is in this sense ‘reinstated’. The metaphor of projection is a useful one to employ for the relationship of the two registers … It accords with Bernstein’s general view about the manner in which a pedagogic discourse takes a discourse from sites elsewhere and reinstates or even ‘relocates’ it for the purposes of the pedagogic activity (Bernstein 1990: 183–5). (Christie 1997: 136–7)

28

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

It was apparent that the notion of ‘instating’ (Halliday 1994) one field of human experience into another applied to the context of research writing. A revelation at the time, and now more widely appreciated, this insight pointed to the need to ‘track … the operation of two registers’ (Christie 1991: 237). Significantly, the ideational discourse of my texts was found to construe two fields – that of the researcher’s object of study and that of the practice of research itself, the latter projecting the former.

HOW THE DATA PUSHED FOR AN EXPANSION OF THE THEORETICAL FRAMING OF GRADUATION What was particular about research writing that prompted a need to expand the modelling of the system of graduation from that available at the time? This is best illustrated with two indicative phases of text. The texts are taken from a relatively recent publication, published as a research monograph: Painter, Martin and Unsworth’s (2013) Reading Visual Narratives: Image Analysis of Children’s Picture Books. Chapter 1 in this book is similar in its structuring to a canonical introduction to a research article, but writ large. The phases of discourse considered below (tabulated by this author) illustrate the dominant patterns that emerged in analyses of my earlier data. (For a more detailed analysis see Hood and Zhang 2020.) Text 1 is the opening phase of the chapter and contributes to establishing the significance of the object of study, in this case children’s picture books. Text 2 appears several pages later and constitutes a phase of discourse in a section that positions the authors’ study in relation to other relevant research and contributions to knowledge. The phases are presented in such a way as to differentiate the fields (the field of research on the left and the object of study on the right); the staggered sequencing points to relations of projection as the discourse unfolds (see Hood 2010: ch. 4 for further examples). Instances of inscribed attitude (including scaled instances) are in bold, and instances of graduation invoking attitude are in italics and underlined. Attention is drawn to many of these instances in the following section. On the basis of field distinctions, patterns of preference in the expression of evaluation are readily evident in the texts referenced above. In Text 1, which reports evaluatively on the object of study, there is a generous deployment of inscribed attitude. This contrasts with the preference for graduation invoking attitude for the ideational construal of the field of research practice. It was the latter that captured my interest and pushed me to further develop graduation systems.

GRADUATION IN RESEARCH WRITING

29

TABLE 1.1  Text 1 (from Painter, Martin and Unsworth (2013: 1)) Projecting field of research practice

Projected field of object of study Children’s picture book stories play a foundational role in the lives and education of young children in several ways. There is, first and foremost, the sheer delight that children take in them, whether reading or being read to, and the fact that these activities

are generally regarded as an essential preparation for the child’s transition into literacy and future school success. As well, picture books in narrative form arguably provide the beginning reader with an entry into the highly valued realm of literature (Meek, 1988), giving these books additional value which is capitalized on in ‘literature based’ reading programs in primary schools. Finally, picture books, most obviously those in narrative form, constitute a significant instrument of socialization, as a source of both overt and covert ‘ideological’ messages about the world and about social values (Stephens, 1992).

Graduation

as force

Force as the adjustment of feelings and values expressed in inscribed attitude had been discussed in a number of publications concurrent with my research. Martin and Rose (2003: 42) analogized the function as turning up or toning down the strength of an attitude. Exemplifying instances included intensified qualities, as in ‘a very important contribution’, or in the quantification of a nominalized value as in, ‘a big problem’. As noted above, my primary interest was not in the amplification of inscriptions of attitude but rather in the grading of ideational meanings. I wanted to understand how contributions to knowledge were being evaluated largely in the absence of inscriptions of attitude. Emerging from the analyses were choices in which ideational meanings were being intensified. For example, there were instances where intensification was grading a quality realized through an epithet, but was arrived at through the

30

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 1.2  Text 2 (from Painter, Martin and Unsworth (2013: 5–6)) Projecting field of research practice

Projected field of object of study

Other writers have attempted to taxonomize the specific relationships possible between word and image, with Schwarcz (1982) offering one of the earliest analyses, suggesting ‘congruency’ and ‘deviation’ as the major types, each having a number of subcategories (cf. Golden 1990: ch. 6). More recently, picture books themselves have been classified into types according to the relations between images and words. For example, Agosto (1999) distinguishes ‘parallel storytelling’ (visual and verbal redundancy) from ‘interdependent storytelling’, suggesting eight subcategories of the latter, grouped within two main types that recall those of Schwarcz: ‘augmentation’ and ‘contradiction’ … While the diversity of picture books is brought into focus in these classification schemes, a general problem with them is that they allow for only one possible kind of relationship between any image and the words it accompanies and/or propose that a single kind of relationship will hold throughout the text.

grading of a classifier, e.g. ‘a more action-oriented study’. This entailed a change in the nominal group structure from Classifier^Thing to Epithet^Thing, since grammatically speaking only epithets can be graded. In other instances, a figure might be intensified by grading the manner of its occurrence. In this case the grading typically had to do with tenacity or resolve – ‘the aim is to look at … ’ – in contrast to the intensified ‘the aim is to explore/ examine … ’. I referred to this general category originally as rigour, a term appropriate in the field of research practice (but not so for other contexts). Martin and White (2005: 146) incorporated this semantic realm in their modelling of graduation as ‘vigour’. In Text 2 we find: reports on close classroom observations …

GRADUATION IN RESEARCH WRITING

31

Figures constituting proposals might also be adjusted in intensity through degrees of modulation in the service of invoking attitude – e.g. ‘the data set should be expanded’/‘the data set must be expanded’. The implied judgement in these instances contrasts with inscription in the lexical realization of obligation in ‘it’s vital the data set be expanded’. In Text 1 we find: these activities are … an essential preparation for the child’s transition into literacy A system network for graduation: force as intensification is presented in Figure 1.1. I add here some changes to terminology in bracketed references to the kinds of ideation that would be implicated in coupling. The revised terminology here and in other networks responds to developments in theorizing field (Martin 2020; Doran and Martin 2021) and in the discourse semantics of ideation (Hao 2015, 2020). This avoids the duplication of lexicogrammatical terminology and maintains stratal consistency as we consider the ideational meanings with which appraisal choices couple. Thus occurrence as a unit of ideation replaces earlier references to process. Analyses of the data revealed that research writers, or the voices they projected, relied more extensively on grading force as quantification than intensification, most notably in terms of amount. An amplified number of research studies or sources frequently implied a degree of significance for a domain of research. Multiple means of expression were deployed. In Text 2 and other similar functioning phases from the same source we find: • Much of the critical literature … • picture books … have been classified into types … For example, Agosto (1999) distinguishes

FIGURE 1.1  A system network for graduation: force: intensification. Source: Hood and Zhang (2020: 23).

32

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

• Other writers have attempted to taxonomize the … relationships possible • these activities are generally regarded as … • a number of studies of … • all of this research … • much of it … In describing the development of ideas or claims, projecting voices also relied on expressions of quantification of extent. Extent as scope in space might imply generalizability or consistency. In Text 2 we find: [the] relationship will hold throughout the text. Extent as scope in (past) time (e.g. ‘for more than two decades’) might also imply positive consistency, but alternatively a negative assessment of ‘time for a change’. Extent as relative distance in time typically implies relative value in terms of relevance. In Text 2 we find: • … offering one of the earliest analyses • More recently, picture books … have been classified into … Extent as distance in space may also imply relative relevance of a study closer to a writer’s context than one more distanced from it along various parameters, including location or methodology. Distance in space is expressed metaphorically in an instance in Text 2: Agosto (1999) distinguishes … two main types that recall those of Schwarcz … Here relative distance in space (proximity) is expressed in the description of one source, ‘Agosto’, as recalling that of another, ‘Schwarcz’. A system network for graduation: force as quantification is presented in Figure 1.2.

Graduation

as focus

In contrast to force, the concept of graduation as focus concerns ‘resources for making something that is inherently non-gradable gradable’ (Martin and Rose 2007: 46). As with force, it has the potential to invoke attitude. These ‘ungradable’ meanings are categorical – they are bounded ideational meanings. So in lieu of force, grading in focus relies on the sharpening or softening of the boundaries of categories. (It is, of course, possible for a category to name an attitude, as in a kind of surprise, in which case this would be taken as the grading of an inscription.)

GRADUATION IN RESEARCH WRITING

33

FIGURE 1.2  A system network for graduation: force: quantification. Source: Hood and Zhang (2020: 23).

The ideational categories attended to in early work on focus were limited to those of entity and quality. The sharpening or softening of boundaries concerned the relative prototypicality with respect to a category, as in, for example, ‘its true colour’; ‘a real audience’; ‘a sort of ethnographic approach’; ‘real ethnography’. These function to invoke an attitudinal interpretation, typically one of appreciation. I treat such choices as options in grading as authenticity. The boundaries of entities can be additionally adjusted in terms of particularity (see Figure 1.3). This option is realized through the sharpening or softening of the boundary of an entity by narrowing or broadening the point of attention. It is exemplified in instances such as ‘the general topic of … ’; ‘discourse particularly from primary schools’. In Text 2 we find: • picture books themselves have been classified into types • the specific relationships possible between word and image • the diversity of picture books is brought into focus in these classification schemes Both authenticity and particularity are options in the system of focus as specificity, as outlined in Figure 1.3. Close attention to the data from the perspective of focus also revealed other units of ideation that could come into play in invoking a value. The boundary of the unit of occurrence (referred to as process in earlier accounts) could also be adjusted with respect to completion in the sense of being un/finished, as in

34

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 1.3  A system network for graduation: focus: specification. Source: Hood and Zhang (2020: 23).

‘try to … ’, ‘manage to … ’, ‘succeed in … ’. In these instances, unfinished is realized through resources of conation in the verbal group. In Text 2 we find: • Other writers have attempted to taxonomize the … relationships possible. Elsewhere we might find nominalized instances such as ‘their attempt to taxonomize … ’; ‘their success in … ’. The boundary of a unit of entity or quality could also be adjusted in terms of completion, in this case in the sense of maximizing out a category, as in ‘partial/total (adherence)’, ‘completely/somewhat (obvious)’. In Text 1 we find an instance in which the inscribed quality ‘delight’ is maximized: • The sheer delight children take in them Finally, instances were identified in which the boundary of a figure constituting a propositional claim could be considered as sharpened or softened with respect to the actualization of the knowledge claimed. The softening of focus was typically realized through congruent or metaphorical enactments of modalization. In Texts 1 and 2 we find: • picture books in narrative form arguably provide … (vs ‘provide’) • one of the earliest analyses, suggesting ‘congruency’ and ‘deviation’ … (vs ‘showing’). As the engagement system was developed, modalized figures as exemplified above were interpreted as instances of heteroglossic expansion, making space for negotiation. As options in graduation: focus as actualization, they may invoke a degree of value with respect to the source of the proposition, or to the reliability of the claim. focus as relative completion – either as un/finished occurrence or non/ maximized category of quality or entity – or focus as relative actualization of a proposition are identified as more delicate features of fulfilment in the system of focus set out in Figure 1.4.

GRADUATION IN RESEARCH WRITING

35

FIGURE 1.4  A system network for graduation: focus: fulfilment. Source: Hood and Zhang (2020: 23).

SO WHAT? Returning to where this research project began and my motivations at the time, where did we get to? Did the findings illuminate ways in which published academic research writers were managing those apparent dual demands of objectivity and critique? And looking forward, what contributions might the study offer to frustrated supervisory academics and their novice research students. How might they support more visible pedagogic practices – including the text deconstruction and joint construction stages of genre pedagogies or the detailed reading practice of Reading to Learn (Rose and Martin 2012). There are a few summary points to make here. • The implicit evaluation of research in the grading of ideational meanings is a significant means by which writers achieve an apparent ‘objectivity’ in their discourse. While novice writers are frequently encouraged to make their writing less personal, the key resource for evaluating research identified in this study involves expressing a subjective, and thus evaluative, position by grading an ‘objective’ ideational meaning. In other words, it involves making the objective more personal, or we might say, ‘subjectifying the objective’. • Space has not permitted discussion of the prosodic structuring of evaluation in this chapter but it is worth adding here that the grading of ideational meanings also plays a significant role in the extension of prosodic domains – meaning that a little explicit inscribed attitude can be made to go a long way in the flow of the text (see Hood 2010 for a more detailed account in relation to Lemke’s notion of propagation). • Inscribed attitude functions to set up a dichotomous choice of positive (in-group) and negative (out-group) alignment, while the grading of ideational meanings positions ‘voices’ on a cline with variations in degrees of similarity or difference. By evaluating other research

36

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

through resources of force and focus, the writer can potentially retain membership of a research community while at the same time establishing difference and hence space for their own research. • Finally, a comprehensive network proposed for graduation in the service of invoking attitude provides a valuable reference for research supervisors, teachers of academic writing and students, in considering the evaluative strategies used in texts to be read and to be written. I conclude with a brief consideration of how research on graduation in language is currently being extended in studies of embodied paralanguage. Hao and Hood (2019) draw on emerging systems of paralinguistic graduation to describe ways in which evaluation is invoked in gestural expression in live lectures. They note, for example, the embodied expression of force as intensification realized through relative muscle tension and/or high frequency of a gestural beat (as illustrated in Hao and Hood 2019: 206–7). force as quantification can be expressed, for example, through the relative size of a gestured entity. This is shown in Hood and Hao (2021: 234) in contrasting images of a gestured entity held in one hand, and one which requires two hands to hold. The sharpening of focus in paralanguage can be expressed through a pinching (precision) gesture which brings the tips of thumb and forefinger/s sharply together. The softening of focus can be expressed through a destabilizing or decentring gesture as in an oscillating hand or an off-centre shrug of the shoulders, as illustrated in Hao and Hood (2019: 211). For an emergent system of options for embodied graduation, see Ngo et al. (2022).

SOME FINAL ADVICE To those setting out on an SFL-informed discourse analytic study I would say: let your data talk to you (mean to you) as you approach it with soft eyes, before your analytical eyes take over. Read widely and deeply of course and engage with other concurrent SFL-informed studies while you explore your data. You never know when the insights of others will spark a new perspective on your own study. The broad SFL community offers many such opportunities, including at its regional and national conferences, local seminars and email lists, and casual encounters with colleagues. ‘There is nothing so practical as good theory,’ said Kurt Levine, and I might add, ‘there is nothing so good for theory as a strong inquisitive and supportive community.’

GRADUATION IN RESEARCH WRITING

37

REFERENCES Bernstein, B. (1990), Class, Codes and Control, Vol. 4: The Structuring of Pedagogic Discourse, London: Routledge. Berry, M. (1981), ‘Systemic Linguistics and Discourse Analysis: A Multi-Layered Approach to Exchange Structure’, in M. C. Coulthard and M. Montgomery (eds), Studies in Discourse Analysis, 120–45, London: Routledge and Kegan Paul. Biber, D., S. Conrad and R. Reppen (1998), Corpus Linguistics: Investigating Language Structure and Use, Cambridge: Cambridge University Press. Channell, J. (1994), Vague Language, Oxford: Oxford University Press. Christie, F. (1991), ‘First and Second Order Registers in Education’, in E. Ventola (ed.), Functional and Systemic Linguistics: Approaches and Uses, 235–56, Berlin: Mouton de Gruyter. Christie, F. (1997), ‘Curriculum Macrogenres as Forms of Initiation into a Culture’, in F. Christie and J. R. Martin (eds), Genre and Institutions: Social Processes in the Workplace and School, 134–60, London: Cassell. Conrad, S. and D. Biber (2000), ‘Adverbial Marking of Stance in Speech and Writing’, in S. Hunston and G.Thompson (eds), Evaluation in Text: Authorial Stance and the Construction of Discourse, 56–73, Oxford: Oxford University Press. Doran, Y. J. and J. R. Martin (2021), ‘Field Relations: Understanding Scientific Explanations’, in K. Maton, J. R. Martin and Y. J. Doran (eds), Teaching Science: Knowledge, Language, Pedagogy, 105–33, London: Routledge. Halliday, M. A. K. (1979), ‘Modes of Meaning and Modes of Expression: Types of Grammatical Structure, and Their Determination by Different Semantic Functions’, in D. J. Allerton, E. Cartney and D. Holdcroft (eds), Function and Context in Linguistic Analysis: Essays Offered to William Haas, 57–79, Cambridge: Cambridge University Press. Halliday, M. A. K. (1981), ‘Types of Structure’, in M. A. K. Halliday and J. R. Martin (eds), Readings in Systemic Linguistics, 29–41, London: Batsford. Halliday, M. A. K. (1982), ‘How Is a Text Like a Clause?’, in S. Allen (ed.), Text Processing: Text Analysis and Generation, Text Typology and Attribution (Proceedings of Nobel Symposium 51), 209–47, Stockholm: Almqvist & Wiksell. Halliday, M. A. K. (1985), ‘Systemic Background’, in J. D. Benson and W. S. Greaves (eds), Systemic Perspectives on Discourse, Volume 1. Selected Theoretical Papers from the 9th International Systemic Workshop, 1–15, Norwood, NJ: Ablex Publishing Corporation. Halliday, M. A. K. (1994), Introduction to Functional Grammar, London: Edward Arnold. Hao, J. (2015), ‘Construing Biology: An Ideational Perspective’, PhD thesis, University of Sydney, Sydney. Hao, J. (2020), Analysing Academic Discourse from a Systemic Functional Linguistic Perspective: A Framework for Exploring Knowledge-Building in Biology, New York: Routledge. Hao, J. and S. Hood (2019), ‘Valuing Science: The Role of Language and Body Language in a Health Science Lecture’, Journal of Pragmatics, 139: 200–15. Hood, S. (2010), Appraising Research: Evaluation in Academic Writing, London: Palgrave Macmillan. Hood, S. and J. Hao (2021), ‘Grounded Learning: Telling and Showing in the Language and Paralanguage of a Science Lecture’, in K. Maton, J. R. Martin and

38

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Y. J. Doran (eds), Teaching Science: Knowledge, Language, Pedagogy, 226–56, London: Routledge. Hood, S. and D. Zhang (2020), ‘Graduation in Play with Other Systems of Meaning in the Enactment of Interpersonal Relations’, Journal of Foreign Languages, 43 (6): 21–41. Hyland, K. (1998), Hedging in Scientific Research Articles, Amsterdam: John Benjamins. Hyland, K. (1999), ‘Academic Attribution: Citation and the Construction of Disciplinary Knowledge’, Applied Linguistics, 20 (3): 341–67. Macken-Horarik, M. (2003), ‘Appraisal and the Special Instructiveness of Narrative’. Text – Special Issue: Negotiating Heteroglossia: Social Perspectives on Evaluation, 23 (2): 285–312. Martin, J. R. (1992a), English Text: System and Structure, Amsterdam: John Benjamins. Martin, J. R. (1992b), ‘Macro-proposals: Meaning by Degree’, in W. C. Mann and S. A. Thompson (eds), Discourse Description: Diverse Linguistic Analyses of a Fund-Raising Text, 359–96, Amsterdam: John Benjamins. Martin, J. R. (2000), ‘Beyond Exchange: Appraisal Systems in English’, in S. Hunston and G. Thompson (eds), Evaluation in Text: Authorial Stance and the Construction of Discourse, 142–75, Oxford: Oxford University Press. Martin, J. R. (2020), ‘Revisiting Field: Specialised Knowledge in Secondary School Science and Humanities Discourse’, in J. R. Martin, K. Maton and Y. J. Doran (eds), Accessing Academic Discourse: Systemic Functional Linguistics and Legitimation Code Theory, 114–47, London: Routledge. Martin, J. R. and D. Rose (2003), Working with Discourse: Meaning beyond the Clause, London: Continuum. Martin, J. R. and D. Rose (2007), Working with Discourse: Meaning beyond the Clause, 2nd edn, London: Continuum. Martin, J. R. and P. R. R. White (2005), The Language of Evaluation: Appraisal in English, London: Palgrave. Myers, G. A. (1996), ‘Strategic Vagueness in Academic Writing’, in E. Ventola and A. Mauranen (eds), Academic Writing: Intertextual and Textual Issues, 3–17, Amsterdam: John Benjamins. Ngo, T., S. Hood, J. R. Martin, C. Painter, B. Smith and M. Zappavigna (2022), Modelling Paralanguage Using Systemic Functional Semiotics: Theory and Application, London: Bloomsbury Academic. Painter, C., J. R. Martin and L. Unsworth (2013), Reading Visual Narratives: Image Analysis of Children’s Picture Books, London: Equinox. Poynton, C. (1985), Language and Gender: Making the Difference, Geelong, VIC: Deakin University Press. Poynton, C. (1996), ‘Amplification as a Grammatical Prosody: Attitudinal Modification in the Nominal Group’, M. Berry, C. Butler, R. Fawcett, Huang, Guowen (eds), Meaning and Form: Systemic Functional Interpretations. Meaning and Choice in Language Studies for Michael Halliday, 211–27, New Jersey: Ablex Publishing Corporation. Rose, D. and J. R. Martin (2012), Learning to Write, Reading to Learn: Genre, Knowledge and Pedagogy in the Sydney School, Sheffield, UK: Equinox. Rothery, J. (1996), ‘Making Changes: Developing an Educational Linguistics’, in R. Hasan and G. Williams (eds), Literacy in Society, 86–123, London: Longman.

GRADUATION IN RESEARCH WRITING

39

Rothery, J. and M. Stenglin (2000), ‘Interpreting Literature: The Role of Appraisal’, in L. Unsworth (ed.), Researching Language in Schools and Communities: Functional Linguistic Perspectives, 222–4, London: Cassell. Salager-Meyer, F. (1997), ‘“I Think That Perhaps You Should”: A Study of Hedges in Written Scientific Discourse’, in T. Miller (ed.), Functional Approaches to Written Text: Classroom Applications, 105–18, Washington, DC: USIA. Sapir, E. (1944), ‘Grading: Study in Semantics’, Philosophy of Science 11: 93–116. [Reprinted in D.G. Mandelbaum (ed.), (1949), Selected Writings of Edward Sapir in Language Culture and Personality, 122–49, Berkley: University of California Press.] Swales, J. M. (1990), Genre Analysis, Cambridge: Cambridge University Press. Ventola, E. (1987), The Structure of Social Interaction: A Systemic Approach to the Semiotics of Service Encounters, London: Pinter. White, P. R. R. (2003), ‘Beyond Modality and Hedging: A Dialogic View of the Language of Intersubjective Stance’. Text – Special Issue: Negotiating Heteroglossia: Social Perspectives on Evaluation, 23 (2): 259–84.

40

CHAPTER TWO

A Balancing Act Theorizing Compositional Choices in Photographs HELEN CAPLE

This chapter accounts for the development of the balance network, a systemic approach to the analysis of compositional choices in photographs/images. While initially created out of and for the analysis of press photographs, the system has wider applications beyond journalism, as will be discussed at the end of the chapter. I begin by establishing the context that motivated this work, and then discuss how my professional practice as a photographer along with theorizing on composition from art and photography influenced how I conceptualize composition in photography. I then demonstrate how the principles of Systemic Functional Theory, e.g. the system network, can be used to explain the compositional aspects of the professional practice of visual news storytelling. Professional journalism in the twenty-first century has been marked by unprecedented disruptions, on an institutional, social and technological level. Every aspect of the production and distribution of news has been affected by these disruptions. Key among these has been the failure of news organizations to stem the outflow of advertising revenue, which has always financed the production of editorial content, away from their own coffers and into those of the tech giants Facebook and Google (Feik 2017: 26). The result has been at best the decimation of staffing levels and at worst the closure of newspapers. In the United States, for example, newspaper employment has fallen by more than 50 per cent (Greenslade 2016; Pope 2018), while in the United Kingdom, the

42

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

net loss of local newspaper titles stood at 228 between 2005 and 2017 (Kakar 2018). In Australia, more than 3,000 editorial jobs have been lost in the decade to 2018 (MEAA 2018: 4), and press photography has been particularly affected by these losses (see Caple 2019). The Australian news media context provides an interesting example of how photojournalism practices have evolved throughout this period, and of how the press photograph, as a vehicle for visual news storytelling, has weathered these disruptions. My interest in tracing and analysing photojournalism practices began in earnest in the early 2000s when news organizations were experimenting with digital distribution platforms (Knox 2010) and circulation figures for print newspapers were starting to fall (Barthel 2018). Indeed, a specific moment of disruption at one Australian metropolitan (and then) broadsheet newspaper, the Sydney Morning Herald, precipitated my research and theorizing of compositional choices in news photography. Following a re-design in 2000, the Sydney Morning Herald (print version) started publishing very large (more than half a broadsheet page) and strikingly beautiful news photography (see Figure 2.1), which was accompanied by minimal verbal text: in the early 2000s, only a caption appeared alongside the photograph (as in Figure 2.1). This new (for the Sydney Morning Herald1) form of visual news storytelling foregrounded the compositional/technical excellence of the photography. The way this interacted with the subject matter of the photographed event resulted in a re-telling of the event as a moment of beauty. The size at which this photography was published on the printed page also suggests that this beauty in the photography was ‘shouted out’ to audiences (much like capitalized screamer headlines do). The photograph was the story. As a former photojournalist myself, I was pleased that photography was being championed in this way in these ‘image-nuclear’ news stories (Caple 2008).2 Inevitably, language eventually (from 2004 onwards) reasserted itself as a partner in this form of news storytelling, in the form of a headline that often engaged in visual-verbal play with the photograph (see Caple 2010a, 2013, and Figure 2.4). However, the exceptional quality of the photography never diminished, and the verbal text never extended beyond the caption. At that time, I was teaching journalism/media studies, and the materials that informed the course development (e.g. Iedema, Feez and White 1994) did not consider the press photograph a full partner in the news storytelling process.3 Yet, the news photography that I and my students were engaging with in newspapers like the Sydney Morning Herald deserved to be analysed with the same rigour as the verbal text had been. Thus, I turned to Kress and van Leeuwen’s (2006) Reading Images for inspiration and a comparable set of tools, also inspired by Systemic Functional Theory (e.g. Halliday 1994), with which to address the meaning potential of news photographs. I was particularly interested in the photograph’s compositional meaning, as I believed, as a former

THEORIZING COMPOSITIONAL CHOICES

FIGURE 2.1  A page from the Sydney Morning Herald demonstrating the use of a single, striking photograph and caption on the printed page (2 June 2003: 11 © Fairfax syndication).

43

44

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

professional photographer, that composition was the source of the beauty in these photographs. In my teaching, I was already drawing on Kress and van Leeuwen’s (2006: 210) systems for compositional meaning to analyse advertisements and magazine stories, and students were confident in explaining the construction of meaning in these texts using the system of information value. However, when it came to examining how news photographs construct meaning, they struggled to assess the placement of elements along the horizontal axis using Kress and van Leeuwen’s (2006: 180) generalization that ‘the left is the side of the “already given”, something the reader is assumed to know already’, while the right is ‘New, the message, the “issue” ’. This given–new structure is drawn from the ‘close similarity between sequential information structure in language and horizontal structure in visual composition’ (Kress and van Leeuwen 2006: 181). This means that image-participants placed in the left-hand side of the image frame would construct meanings such as the ‘agreed-upon point of departure for the message’ or ‘commonsensical’ or ‘self-evident’. Information placed in the right-hand side of the frame would be interpreted as ‘not yet known’, ‘problematic’ or ‘contestable’ (Kress and van Leeuwen 2006: 181). When students did try to interpret the meaning of horizontally polarized compositional structures as given–new they drew on meanings from the verbal news story text or from wider cultural understandings of either the image participants or the types of events being depicted, which resulted in extremely tenuous, and ultimately futile, interpretations of the photographs. This meant that their analysis no longer relied on the ways in which different elements relate to each other within the image frame, but rather how these elements relate to external contextual or cultural understandings. In their writing, Kress and van Leeuwen (2006) do state that not all images make ‘significant use of the horizontal axis’ (180–1, emphasis added), and this suggests that interpretations of given–new are not always appropriate (see Caple 2013: 87–9 for exemplification of these issues). Yet, many (news) photographs do organize elements away from the centre of the frame, often along the horizontal and diagonal axis, and this was especially true for those photographs being championed by the Sydney Morning Herald. Therefore, I set out to develop additional/complementary tools to be able to explain and analyse these polarized compositional choices. I did this not in analogy to language structure, but rather based on the workings of perception and Gestalt psychology as they relate to photography theory (Arnheim 1954; Dondis 1973; Zakia 1997; Altengarten 2004; Präkel 2006). I also drew on my professional practice in photography and on my training in Systemic Functional Linguistics. The study of visual arts incorporates the two broad categories of the fine and applied arts. Fine arts are motivated by the ability to produce an object of beauty, and with this, pleasure in looking at that object, while the applied

THEORIZING COMPOSITIONAL CHOICES

45

arts involve the additional function of utility. I was particularly drawn by this dialectic between aesthetics and utility, as this is what had always been in play in the visual storytelling that I did as a press photographer. On the one hand, I was required to produce photography that told readers something about the news event, its utility function, if you like. On the other hand, the compositional choices that I made in telling a visual story were also motivated by their aesthetic potential to construe the event as beautiful. The latter, I found, impacted on how the photograph was then packaged in relation to the words accompanying it, as well as in relation to the page as a whole. This is because a well-crafted photograph attracts the viewers’ attention and makes them look for longer and, as I have argued elsewhere (Caple 2019: 1), when you look for longer, you begin to feel something. Newspapers want readers to feel something. To look deeper at the relationship between the well-crafted photograph and its aesthetic potential, I turned to Gestalt psychology and photography theory. First, with respect to Gestalt psychology, Arnheim (1954) explores the workings of perception, the quality of individual visual units and the strategies for their unification into a final and complete whole. In his view, the human organism seems to seek harmony in what it views, a state of ease, of resolution, what Zen Buddhists speak of as ‘meditation in supreme repose’ (Dondis 1973: 85). There appears to be a need to organize all stimuli into rational wholes and Gestaltists have established that the eye, and for that matter the brain, ‘will not be deterred in its endless pursuit of resolution or closure in the sensory data it views’ (Dondis 1973: 89). Resolution is achieved through regularity, symmetry and simplicity, that is, absolute balance. In photography, this sense of resolution or balance is achieved instantly because we view the photograph as a whole. Second, in photography theory, concepts such as the golden mean or the rule of thirds explain how compositional balance is achieved in the photograph. The golden mean is based on the Fibonacci sequence, where the ratio of each successive pair of numbers in the series approximates to the golden number (1.618034), identified by the Greek letter phi φ. It is visually represented as a spiral that connects rectangles, joined at the corners, in a ratio of 1:1.618 (Example 1 in Figure 2.2). The golden mean provides a reliable method for dividing a composition into regular, symmetrical, simple or pleasing proportions (Präkel 2006: 22). This is because it is calculated on the basis of the physical proportions of naturally occurring life forms, such as flowers, seashells and even the human body (see Caple 2013: 93–5 for examples). A simpler version of the golden mean more commonly used in photography is the rule of thirds (Example 2 in Figure 2.2). This allows a photographer to organize information in the frame around four points, or ‘hot spots’, that are located away from the centre of the frame. This is a useful aid in composing reliably well-proportioned photographs where information placed on or near

46

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 2.2  Visual representations of the golden mean, rule of thirds and dynamic asymmetry. Source Caple (2009: 67–68).

these hotspots is generally of equal weight, i.e. conventionally balanced. This does not, however, produce very ‘exciting’ photographs. A more challenging and dynamic approach to the organization of information in the frame is exemplified schematically in Example 3 in Figure 2.2. It is challenging because of the potential to unbalance the photograph by placing information on the diagonal axis rather than on the horizontal or vertical axis. Diagonal lines in photography are said to produce more dynamic compositions (Altengarten 2004; Präkel 2006). Placing information in the image frame on or near one of the diagonal hotspots can be counterbalanced by other information on or near the other hotspot on the same diagonal axis. The use of the diagonal creates tension, especially if the opposing hotspot is left ‘empty’. To bring this back to the discussion of Gestalt psychology, this tension stimulates the eye to work harder to resolve the potential imbalance, which it does eventually do. The resulting asymmetrical balance is highly valued in photography as an aesthetically pleasing form of composition (Altengarten 2004; Präkel 2006). To relate this again to my own professional practice, as an experienced press photographer, I understood the narrative impulses driving the photography that is used in news storytelling. These are underpinned by compositional choices that aim to balance different elements in the frame in relation to each other, either creating equal relations between those elements or establishing more dynamic, unequal relations. Photographs that do not achieve this sense of balance between elements in the frame are very difficult to read and understand. There is no sense of closure or resolution in the sensory data deployed in the image frame. Meaning becomes confused or difficult to access. It is also clear, through experiencing how some (but not all) photographs are championed in the newspaper, that it is out of the sense of absolute balance, and the (sometimes deliberate) subversion of symmetry, that press photographs are also able to achieve beauty and aesthetic pleasure. Kress and van Leeuwen (2006) also observe the importance of symmetry, rhythm and balance in composition. However, they relate this to their system for salience (2006: 201–3). They state for example: ‘Being able to judge the

THEORIZING COMPOSITIONAL CHOICES

47

visual weight of the elements of a composition is being able to judge how they “balance” ’ (2006: 202). Composition, they suggest, can be viewed in formal or aesthetic terms (through harmony and balance), in pragmatic terms (through feelings and grabbing the readers’ attention) and in functional terms (as marshalling meaningful elements into coherent texts) (Kress and van Leeuwen 2006: 203). From the perspective of a photographer, I saw the value in bringing these three aspects together: meaning and coherence in a photograph are achieved by bringing the elements within the image frame into compositional balance, which in turn enhances the aesthetic pleasure to be had in viewing the photograph, which in turn grabs and holds the readers’ attention. This marrying of balance, aesthetic and reader attention, I would contend, allows readers to engage more deeply with the content, and thereby to care more deeply about the event depicted. Based on these notions of compositional balance and aesthetic pleasure, I turned my thinking to how to create a tool that would allow analysts (including my students) to systematically analyse the compositional choices being made in a photograph. This had to be a tool capable of producing reliable and replicable results and that also complemented the systems developed by Kress and van Leeuwen (2006), inspired as they are by the powerful theorizing of Systemic Functional Theory (hereafter SFT). The particular orientation of SFT that I drew on in theorizing image composition is the system network. SFT uses system networks to demonstrate the relations that hold between the different features of a semiotic resource (e.g. a photograph). The organizing principle of the system is that of choice, where structure is the outward form taken by systemic choices (Halliday and Matthiessen 2004: 23). In order to understand the choices being made in the construction of a photograph, I collected and analysed 1,000 press photographs that were published in the Sydney Morning Herald between November 2002 and December 2006. All of these photographs were published in the format discussed above, as image-nuclear news stories. The compositional choices, how elements in the image frame were organized and how they related to each other were used to develop the system network for compositional balance – the balance network – that I introduce in more detail in the following paragraphs. A number of observations based on the analysis of these press photographs are worth mentioning at this point. First, the balancing of elements within the image frame was clearly central to the compositional choices made in these photographs. They overwhelmingly focused on only one activity sequence or one set of image participants (elements), thus creating a single visual unit of information (VUI) for each photograph.4 The elements that made up the VUI also engaged mostly in transactional narrative structures (Kress and van Leeuwen 2006: 59), where both the actor and the goal are depicted in the image frame and in relation to each other. I term such image composition

48

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 2.3  The balance network. Source: Caple (2009: 168).

as being endocentrically balanced (Caple 2013: 96). This relates well to Dondis’s (1973: 91) observation that the human perceptual system needs to find resolution in what is viewed, that is, we perceive images as ‘organised configurations rather than as collections of independent parts’ (Stroebel, Todd and Zakia 1980: 164). A very small number of photographs were what I term exocentrically balanced. In such photographs, only the actor is depicted and the goal is cropped out of the image frame. The goal, however, is still implied through the unresolved vectors emanating from the actor and from the reader’s contextual/cultural understanding of the activity sequence depicted. These exocentrically balanced photographs are also balanced, but in a way that challenges perception and the ease with which the eye achieves resolution in what is (partially) viewed. I also found that these photographs focused only on a small number of participants, sometimes singling out individuals as the point of focus in the frame (in what I term isolating configurations) and sometimes showing the image participants in equal relations with each other (in what I term iterating configurations).

THEORIZING COMPOSITIONAL CHOICES

49

The balance network that I developed is presented in Figure 2.3. It is made up of two simultaneous systems. This means that compositional configurations consist of features that are either isolating OR iterating AND either endocentrically OR exocentrically balanced. It has been suggested over recent years that the first system could be made recursive, allowing the analyst to select both isolating and iterating configurations for a single photograph. However, a more compelling argument might be made for clustering elements together into a single element and arguing for how this then balances with other elements within the image frame. I discuss this issue to some degree in Caple (2013: 107–8). Space precludes giving full explanation and exemplification of each of the systemic choices available through the systems shown in Figure 2.3. Instead I refer the reader to Caple (2013: 98–111), where these choices are exemplified with multiple authentic press photographs. Rather, I focus here on how the balance network relates to the discussion earlier in this chapter on Gestalt psychology and photography theory, and through this offers an alternative way to discuss composition in a photograph. Gestaltists point to the need for resolution in the sensory data that the human eye views, and this is achieved most simply through the regular, symmetrical organization of elements in an image frame. This absolute balance is captured in the iterating choice in the balance network, where elements in the frame are organized in mostly equal relation to each other along the horizontal or vertical axis. It is also captured in its simplest form in the isolating:centred options in the upper half of the network, where a single element is centred in the frame. Photographs composed according to these configurations could be said to be well proportioned and conventionally balanced, and in the iterating configurations, relations between image participants are more or less equal. Balance is also achieved through the more dynamic, asymmetrical organization of elements of unequal weight along the diagonal axis. Such compositional configurations create unequal relations that require more effort for the eye to reach a point of resolution. This is captured in the isolating:axial choice in the system network. This is a less conventional method of balancing elements within the image frame, and, as noted above, the resulting asymmetrical balance is highly valued in photography as aesthetically pleasing. The photograph in Figure 2.4 (under the headline ‘Leaning to make a splash’) is a good example of an axially composed photograph where two elements (young divers) of unequal weight are related to each other along the diagonal axis. The extra work needed to balance the relations between these elements stimulates the reader to resolve any potential imbalance, thus creating interest and aesthetic appeal in the photograph. Aesthetic appeal in photographs may be enhanced through the many ways in which the regularity and symmetry of the iterating function in the

50

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 2.4  An example of an axially composed photograph (Sydney Morning Herald, 15 June 2004: 12 © Fairfax syndication).

network can be challenged yet remain within the realms of a balanced photograph. This may be manifested through interrupted symmetry (one image participant may be looking in a different direction to others, or wearing different coloured clothing to other participants), or through cropping the photograph in a way that requires the reader to fill in the rest of the information (e.g. by focusing only on the reflection of image participants). Further, creating playful tension between the isolating and iterating configurations challenges the extent to which elements in the frame can be clustered together and be placed in relation to other elements of unequal visual weight (see Caple 2013: 117 for examples).5 My analysis of 1,000 press photographs using the balance network revealed conventional practices in visual news storytelling. They were overwhelmingly endocentrically balanced (only 10 photographs were exocentrically balanced). balance

THEORIZING COMPOSITIONAL CHOICES

51

This suggests that press photographers focus on representing a single activity sequence where relations between elements in the frame are established (and resolved) within the frame. Only one third of photographs were balanced using iterating configurations, showing equal relations between image participants, and producing reliably well-balanced photographs. Two thirds of photographs were composed using isolating configurations, with a near even split between centred and axially balanced compositions. This again explains how press photographs usually focus on single activities/participants. None of the 1,000 photographs fell outside this composition analysis, which further demonstrates that press photographers focus on the representation of a meaningful whole. Photographers also aim to produce photography that challenges perception and that stimulates the viewer to resolve potential imbalance in a photograph through axial compositional choices. Additionally, my analysis allowed me to draw conclusions about the role and function of press photographs in journalistic practices, specifically in relation to their role as nucleus in the image-nuclear news story genre in the Sydney Morning Herald. Not only do press photographers engage in the compositionally safe photography that is guaranteed to produce a well-balanced photograph, they also produce photographs composed in ways that stimulate the viewer to participate more actively in the aesthetic composition of newsworthy information. This active participation in the construction of meaning leads to deeper engagement with the photograph’s content and creates a lasting memory of the event. The scarcity of verbal text with these stories further focuses attention on the photography. As such, these were not stories that needed to be told; they were stories that needed to be shown. Another outcome of this theorizing of compositional balance and the aesthetic pleasure to be found in composition was the addition of Aesthetic Appeal as a news value, in that a news event may be discursively constructed as beautiful through the photograph’s composition (see Bednarek and Caple 2017). By way of concluding this reflection, I point to the influence that this research has had on other scholars, and how composition has been theorized in other visual practices. In Painter, Martin and Unsworth (2014: 109–20), for example, the focus network, which maps the compositional meaning of visuals in children’s picture books, borrows heavily from the balance network. Bednarek (2014) has applied the balance network to the analysis of television title sequences. Durrani (2020) extends the balance network to examine external cohesion between pictures, verbal text and their layout patterns. This research demonstrates the strengths of this approach to composition and suggests that it can be applied to both still and moving images, to illustration and page layout, and to photography more generally.

52

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Finally, I would like to bring the discussion back to the fate of the press photograph in these times of precarity in employment models (Caple 2019) and to the increasing reliance on online platforms for the distribution of news. Fairfax Media, the publisher of the Sydney Morning Herald (now in tabloid format), has been sold to Nine Entertainment, a commercial television and entertainment company. Its continued existence in print has been a discussion point for several years. The image-nuclear news story has largely disappeared, as have the photographers that once supplied the imagery for such stories. As depressing as this may sound, especially in a market as restricted as the Australian news media market, there should be cause for optimism. Digital platforms now dominate the distribution of news and this should signal a new golden age for the press photograph. There are ample means of championing the photograph, e.g. through the online news gallery (Caple and Knox 2019) and in other forms of multimedia storytelling. However, these often fall short both in sourcing the best quality press photography, and in taking advantage of the visual storytelling potential of sequences of photographs (Caple and Knox 2017). This is another area of professional practice where Systemic Functional Theory has been used to make such shortfalls explicit and to provide advice on how to improve professional practice (see Caple and Knox 2019). Ultimately, however, professional practice must be underpinned by funding models that fully support the production and distribution of news discourse. This begins by recognizing the important role that professional photographers play in news storytelling. It also means recognizing the power of the well-crafted photograph to pull readers into a story, to make them look, and ultimately to make them feel something.

NOTES 1 Other newspapers throughout the UK and Europe had been producing such story formats for a long time. 2 I analysed the generic structure of these stories as ‘image-nuclear’ news stories. However, this is not an aspect of the discussion in this chapter (see Caple 2008, 2010b, 2013). 3 See Caple (2013) for how the press photograph can be incorporated into the functional structure of the news story genre. 4 Painter et al. (2014: 92) reconstrue the visual unit of information as ‘focus groups’ and note that in children’s picture books a spread or page may be made up of a number of visual elements that encourage simultaneous viewing. 5 Compositional balance is just one of a number of ways in which the aesthetic pleasure in a photograph can be enhanced. This can also be achieved through technical means and by exploiting the natural beauty of the environment. Examples are given in Caple (2013: 114–16).

THEORIZING COMPOSITIONAL CHOICES

53

REFERENCES Altengarten, J. (2004), ‘Creativity and the Rule of Thirds’, Photo Composition Articles. Available online: http://photoinf.com/Golden_Mean/Jim_Altengarten/Creativity_ and_the_Rule_of_Thirds.html (accessed 20 February 2006). Arnheim, R. (1954), Art and Visual Perception: A Psychology of the Creative Eye, Berkeley, CA: University of California Press. Barthel, M. (2018), ‘Newspapers Fact Sheet’, Pew Research Centre, Journalism & Media, 13 June. Available online: http://www.journalism.org/fact-sheet/newspapers/ (accessed 12 February 2019). Bednarek, M. (2014), ‘The Televisual Title Sequence: A Visual Analysis of Flight of the Conchords’, in E. Djonov and S. Zhao (eds), Critical Multimodal Studies of Popular Discourse, 36–54, New York: Routledge. Bednarek, M. and H. Caple (2017), The Discourse of News Values: How News Organizations Create Newsworthiness, New York: Oxford University Press. Caple, H. (2008), ‘Intermodal Relations in Image Nuclear News Stories’, in L. Unsworth (ed.), Multimodal Semiotics: Functional Analysis in Contexts of Education, 125–38, London: Continuum. Caple, H. (2010a), ‘Doubling-Up: Allusion and Bonding in Multi-semiotic News Stories’, in M. Bednarek and J. R. Martin (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 111–33, London: Continuum. Caple, H. (2010b), ‘Multisemiotic Communication in an Australian Broadsheet: A New News Story Genre’, in C. Bazerman, A. Bonini and D. Figueiredo (eds), Genre in a Changing World: Perspectives on Writing, 243–54, Fort Collins, CO: WAC Clearinghouse and Parlor Press. Caple, H. (2013), Photojournalism: A Social Semiotic Approach, Basingstoke: Palgrave Macmillan. Caple, H. (2019), Photojournalism Disrupted: The View from Australia, London: Routledge. Caple, H. and J. S. Knox (2017), ‘Genre(less) and Purpose(less): Online News Galleries’, Discourse, Context & Media, 20: 204–17. Caple, H. and J. S. Knox (2019), ‘How to Author a Picture Gallery’, Journalism: Theory, Practice and Criticism, 20 (11): 1440–59. https://doi.org/10.1177/ 1464884917691988 Dondis, D. A. (1973), A Primer of Visual Literacy, London: MIT Press. Durrani, S. (2020), ‘Previewing News Stories: How Contextual Cohesion Contributes to the Creation of News Narratives’, in H. Stöckl, H. Caple and J. Pflaeging (eds), Shifts Towards Image-Centricity in Contemporary Multimodal Practices, 123–45, London: Routledge. Feik, N. (2017), ‘Killing Our Media: The Impact of Facebook and the Tech Giants’, The Monthly, July: 24–33. Available online: https://www.themonthly.com.au/ magazine/july-2017 (accessed 8 April 2022). Greenslade, R. (2016), ‘Almost 60% of US Newspaper Jobs Vanish in 26 Years’, The Guardian, 6 June. Available online: https://www.theguardian.com/media/ greenslade/2016/jun/06/almost-60-of-us-newspaper-jobs-vanish-in-26-years (accessed 16 August 2018). Halliday, M. A. K. (1994), An Introduction to Functional Grammar, 2nd edn, London: Edward Arnold.

54

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Halliday, M. A. K. and C. M. I. M. Matthiessen (2004), An Introduction to Functional Grammar, 3rd edn, London: Arnold. Iedema, R., S. Feez and P. R. R. White (1994), Stage Two: Media Literacy, Sydney: Department of School Education. Kakar, A. (2018), ‘Some 40 UK Local Newspapers Closed in 2017 with Net Loss of 45 Jobs, New Research Shows’, Press Gazette, 27 March. Available online: https:// www.pressgazette.co.uk/some-40-uk-local-newspapers-closed-in-2017-with-netloss-of-45-jobs-new-research-shows/ (accessed 16 August 2018). Knox, J. S. (2010), ‘Online Newspapers: Evolving Genres, Evolving Theory’, in C. Coffin, T. Lillis and K. O’Halloran (eds), Applied Linguistics Methods: A Reader, 33–51, London: Routledge. Kress, G. and T. J. van Leeuwen (2006), Reading Images: The Grammar of Visual Design, 2nd edn, London: Routledge. Media, Entertainment and Arts Alliance (MEAA) (2018), Submission to the Australian Competition and Consumer Commission’s Digital Platforms Inquiry, April. https:// www.accc.gov.au/system/files/Media%2C%20Entertainment%20and%20Arts%20 Alliance%20%28April%202018%29.pdf (accessed 8 April 2022). Painter, C., J. R. Martin and L. Unsworth (2014), Reading Visual Narratives: Image Analysis of Children’s Picture Books, Sheffield: Equinox. Pope, K. (2018), ‘So You Wanna Be a Journalist?’, Columbia Journalism Review, Spring/Summer. Available online: https://www.cjr.org/special_report/journalismjobs.php/ (accessed 8 June 2018). Präkel, D. (2006), Composition, London: AVA. Stroebel, L. D., H. N. Todd and R. D. Zakia (1980), Visual Concepts for Photographers, Boston: Focal Press. Zakia, R. D. (1997), Perception and Imaging, Boston: Focal Press.

CHAPTER THREE

Developing a Multimodal Rank Scale JOHN S. KNOX

BACKGROUND This chapter provides an account of the development of a rank scale in a study of online newspapers. This was one part of my PhD project, which investigated multimodal communication on the home pages of online newspapers (Knox 2009). The project was begun in the early 2000s when online newspapers were still in their infancy and evolving rapidly, and before the partial colonization of news distribution became widespread – first by aggregators (such as Google News) and later by social media (such as Facebook). At that point in the history of news media and the internet, it appeared that online newspapers were ‘the future’ of newspapers (e.g. Boczkowski 2004; Knox 2009: 1–3). In this context, it was important to develop a better understanding of the ways they functioned, the kinds of meanings they made and also the way that the ideologically driven aims of newspaper institutions (e.g. Fowler 1991) could be enacted in a new medium, especially on home pages that featured numerous extremely short texts. The focus of the study was home pages, but something that became apparent in the study was that online newspapers were structured in such a way that each newspaper section had its own home page (initially identified as ‘section pages’ in the study), and so questions about the overall semiotic structure of online newspapers became important. The motivation to look at rank came from a question posed by Alison Moore at a seminar presentation I gave on the work. Having presented a number of

56

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

system networks under development, I was asked by Alison: ‘What rank are these systems at?’ Since I did not have an answer to this question, I decided to find out. In the early stages of the PhD, I had been working on a completely different problem. My supervisor, Jim Martin, encouraged me to read Sinclair and Coulthard’s (1975) work on developing a rank scale for classroom discourse in order to better understand levels of function–class relations (their work was inspired by Halliday’s classic 1961 paper ‘Categories of the Theory of Grammar’). Later, as part of the reading for a Friday afternoon Systemic Functional Linguistics (SFL) course for research students in the Sydney region, Martin also required us to read the work of Halliday (e.g. 1981, 1985), Huddleston (1981) and Hudson (1981), among others. These readings had been foundational to my understanding of axial relations in SFL theory, but had not been directly relevant to my study of home pages until I decided to look at rank. This was also at a time when a number of people were asking questions about rank and multimodality, including whether ranks exist in non-linguistic modes and, if so, what they are (e.g. Kress and van Leeuwen 1990 and subsequent editions had not organized their description of images by rank, whereas O’Toole 1994 had). These in turn led to questions about kinds of interaction among elements in different modes and the senses in which rank is or is not implicated in such interaction (see Kok 2004; O’Halloran 2004; Martinec 2005; Martinec and Salway 2005; Zhao 2010; see also Baldry and Thibault 2006: 144). I was also fortunate to be conducting this research around the same time that two colleagues and collaborators were working on separate but relevant projects. Emilia Djonov (2005) was researching logico-semantic relations in children’s news websites; and Sumin Zhao (2011) was researching children’s digital learning programmes. I was able to draw on their research in constructing the rank scale discussed in this chapter. Zhao’s (2010) chapter on rank and multimodality provided an important perspective on the multimodality literature cited above.

DEVELOPING THE RANK SCALE I began exploring the question of rank only towards the end of my PhD research. But that work on rank, described below, provided a theoretically grounded account of the overall structure of online newspapers that had been missing from the study. As noted above, questions were being asked about rank and multimodality at the time. I had developed system networks for news stories on online newspaper home pages (or newsbites: Knox 2007). However, the extent to which the architecture of a functional theory of language could or should apply to other

DEVELOPING A MULTIMODAL RANK SCALE

57

modalities was not clear (hardly surprising since this remains an ongoing and fundamental question). It was not a ‘given’ that there were ranks involved. Djonov (2005) had outlined perspectives on the structure of websites from the fields of human–computer interface, information architecture and web usability. Importantly, she had also developed an account of the periodic structure of children’s news websites. These structural perspectives were extremely important for the questions she was asking – focusing on navigation in and around such sites. I observed that, from the perspective I was taking, online newspapers had these structures. But they also had constituent structures – since online newspapers consisted of sections (e.g. sport, business), sections consisted of pages, pages consisted of different functional zones (containing, for example, news, navigation menus and advertisements) and zones had texts. This meant that I was working with a number of levels, and (one of) the relation(s) between the levels was constituency. This meant that I was probably working with a rank scale, and that the system networks I had been developing were at the lowest rank of that scale. Constructing the rank scale was a process, first, of identifying potential ranks, second, of identifying constituents at those ranks, and third, identifying the class–function cycle for each rank. To illustrate with language, in SFL, there are class elements (such as a clause), which in turn have structural constituents (such as Participants, Processes and Circumstances). Such a class– function combination is a rank. At the rank below, the functional elements (e.g. Participants) are realized in turn by class elements (e.g. nominal groups), which have their own functional structure (e.g. Deictic, Epithet, Classifier, Thing) and so on. In terms of the complex texts I was dealing with, it was unclear what the relations were between an individual newsbite, the home page on which it was located and the part of the page (e.g. a specific column or horizontal section) in which it was located. Lay descriptions such as a news story being ‘on’ a page, or ‘in’ a particular column, or ‘under’ a particular heading had no theoretical grounding and provided no power to generalize or to explain the semiotic architecture of the news sites, the home pages or the news stories. The identification of class and function elements, and possibly ranks, offered one potential approach to at least partially address this problem. I will discuss the two most challenging aspects of this process: i) The theoretically driven problem of separating class and function (e.g. separating structural elements such as Classifier and Thing from class elements such as noun), and this will be exemplified at the ranks of page and zone. ii) The data-driven problem of accounting for the place of the home page in the website structure.

58

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Figure 3.1 shows the rank scale I developed with most of the key structures detailed. Rather than working top-down or bottom-up, I started at the point with which I was most familiar – the structure of the home page (page rank). The challenge of separating class and function arose early. I had identified common zones on home pages across online newspapers. For example, home pages typically have a header that includes the masthead, menus in navigation zones, a footer with copyright and other administrative information, and marketing zones with advertisements. They also feature many short news texts grouped together in what I termed a news taxonomy, since the layout of the news texts used the grammar of a visual taxonomy (Kress and van Leeuwen 1996; Knox 2009).

FIGURE 3.1  Class-function diagram showing a rank scale perspective on online newspaper structure, from newspaper to newsbite. Source: Knox (2009: 161).

DEVELOPING A MULTIMODAL RANK SCALE

59

I came to realize that the zones on the page were in themselves spaces with no inherent function and were therefore class items. Those class items potentially ‘carried’ various functions in the structure of the home page. The Brand of each newspaper (function element) appeared in the header (class element). The copyright, legal formalities and navigation to administrative functions all appeared in the footer. I collected these various rhetorical and social functions under the function term Signature. I continued in this manner – identifying the function each zone played in terms of the home page structure – and ended up with a page rank with the functional structure Brand + Navigation + News Coverage + Marketing + Signature. Each structural element was realized by a class element at the rank of zone, respectively: header, navigation zone, news taxonomy, marketing zone and footer. I repeated this challenging and time-consuming process of identifying structures, finding appropriate function labels and then finding appropriate class labels for items at the rank below. The most challenging part of this process was at the rank of website section. I had identified that online newspapers had sections – an initially commonsense observation that was supported by analysis of the websites, and by the work of Djonov (2005) mentioned above. There remained, however, a significant problem. The home page of an online newspaper is not a part of any of the sections; metaphorically speaking it is ‘above’ everything else in the website. This is reflected in traditional ‘segmental’ and ‘holistic’ approaches to describing website structure (see Djonov 2005), as illustrated in Figure 3.2. However, I could not put home pages above sections in the rank scale, because home pages consist of zones (Figure 3.1), not of website sections. Home

FIGURE 3.2  Lay approach to understanding online newspaper website structure.

60

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

pages are at the rank of page, below the rank of section, and this appeared to be an intractable problem for some time. Thus, my emerging conclusion was that a rank scale could not account for online newspapers (or for websites). Fortunately, around this time, one of the Thai newspapers that I regularly read resumed using splash pages, featuring a striking image and/or short text (e.g. an image of the Thai King Rama IX’s mother and a short statement about her following her death) and a single link to the home page. Splash pages are a historical anomaly of web design which ‘served no navigational function. Instead, the screen was eye candy for users’ (Cooke 2003: 171). This led me to consider how splash pages function, and where they fit in website structure. On reflection, I realized that splash pages provide a welcoming, introductory function, which is after all one of the functions of home pages. Further, on progressing past a splash page, readers are taken to the home page of a website. This allowed me to reconsider the semiotic structure of the website-as-text, and posit a Portal^Content structure at the rank of website, with the Portal structure realized by a home section (with an optional splash page followed by an obligatory home page), and the Content function realized by a content section (see the discussion of rankshift below for more on this; see also Figure 3.1). When this piece of the puzzle finally ‘fell into place’, the theoretical model of a rank scale began to make sense. Briefly (and returning to Figure 3.1), at the rank of online newspaper, an online newspaper has three functional elements: Portal (bringing the reader in), Content (providing the news) and Administration (dealing with the institutional, legal and other non-news aspects necessary to publishing the newspaper online). The Portal is realized by a home section with a structure of Splash^Orientation (realized in turn at the rank below of splash page followed by home page). The Content function is realized by a content section which has a structure of Briefing (realized by a home page) and Dispatch (realized by a story page). The structure of home pages has already been discussed above and can also be seen in Figure 3.1. The ways in which this theoretical perspective paid dividends is discussed below.

LESSONS LEARNED Constructing a rank scale was rewarding, and provided a theoretical foundation for aspects of my research on online newspapers that were otherwise lacking. Here I discuss a number of benefits from this insight: theoretical creativity, rankshift, website structure and home pages, and a systemic perspective.

DEVELOPING A MULTIMODAL RANK SCALE

61

To begin, something I learned from the process described above is that developing a model is an act of semiotic construction. When the rank scale ‘fell into place’, I did not have a feeling that I had ‘captured’ the structure of online newspapers. Rather, I felt like I had constructed an explanatory model that had sufficient elegance to account for the structure of online newspapers from the perspective of constituency; it was an act of theoretical creativity. Online newspapers do not have ranks, and are not constructed, written or read according to a rank scale (see following section). But we can use the theoretical construct of a rank scale to describe and understand them. It was a useful lesson in understanding what theory is, and what it is not. Another thing I learned was that, armed with a rank scale, I was able to explain the recursive structures in online websites in terms of rankshift, which is the phenomenon whereby a structure from a higher rank is ‘brought into service’ at a lower rank. To exemplify using the English language, in the clause eating chocolate is healthy, the non-finite clause eating chocolate functions as a Participant. Thus we have a clause functioning as a Participant within a clause – it is rankshifted (often also described as being ‘embedded’ in SFL). In online newspapers, rankshift commonly occurs at many ‘locations’ in the rank scale. For example, rankshift can be observed in newsbites. Some newsbites have Tangents (a hyperlink to a separate, related story), which are realized by newsbits (headline-only hyperlinks that function as news stories at the same rank). And some newsbites have Navigator Menus (menus in stories that link to other pages in the newspaper), which are realized by menus (also at the same rank). An example of this can be seen in Knox (2009, p. 200 – freely available online), a figure which could not be included here due to Bloomsbury copyright requirements. Another example is the functional element of Content at the rank of online newspaper (see Figure 3.1), which can be realized by a rankshifted online newspaper (Knox 2009). An example of this is shown in Knox (2009, p. 197 – freely available online), a figure which again could not be included here due to Bloomsbury copyright requirements. In this example, the Sport section has the structure of an online newspaper, with its own home page and content sections such as Cricket, NRL, Union, Beijing Olympics and so on (Knox 2009). As explained in Knox (2009: 196–9), following a hypothesized but likely navigational path shows how rankshift helps the institutional author organize information, and helps the reader navigate the newspaper.1 Home pages function ideationally as successive Orientations as the reader navigates the newspaper, and then as Briefing in the content section.2 From Djonov’s (2005, 2007) periodic perspective on website design, home pages function textually as successive macro-Themes as the reader navigates through them in their role in the hierarchy of macro-Themes of the online newspaper. ‘From the metafunctional perspective of SF theory, these two structural perspectives (periodic and particulate) are simultaneous’ (Knox 2009: 199).

62

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

This means that the recursive structure of websites that can expand (potentially) indefinitely can be explained experientially by a rank scale, and textually by a hierarchy of periodicity. The former gives us a perspective on how the field is construed within the semiotic architecture of a website by a given news institution; the latter gives us a perspective on how readers are guided to the information. In terms of website structure and the place of home pages in that structure, the rank scale provided a genuinely novel perspective on website architecture, with home pages positioned within an overall structure as opposed to somehow ‘standing above’ the rest of the website. Further, by identifying home pages as class elements with an identifiable structure, it became evident that so-called ‘section pages’ (or the home page of a specific section of the news site) had the same structure as the home page of the entire site – which helps account for the recursive structure of these news sites and the use of rankshift as discussed above (see Knox 2014 for discussion and exemplification; see Djonov 2005 for discussion of traditional segmental and holistic approaches to describing websites in other disciplines). The final point to be raised here is in relation to taking a systemic perspective on online newspapers. Having identified a rank scale, the question of what systems are in operation at each rank arose. These questions were beyond the scope of the study; but they offered a way to address an issue raised at the time by John Bateman, who argued that online newspapers could be considered to have more in common with search engine result pages than with print newspapers: ‘Although there are many issues of design and usability to consider here, what is most relevant for our present discussion is the way in which the properties of the online newspaper align with very different sets of cogeneric texts than might have originally been thought on the basis of its informal classification as a “newspaper” ’ (Bateman 2008: 181). The rank scale provides a means by which a systemic approach to investigating such issues can be taken. For instance, should the question be approached from an institutional perspective, classifying print and online newspapers against each other as shown in Figure 3.3? Or do they need to be classified completely separately, and approached from a medium-based perspective (i.e. as one type of website), as shown in Figure 3.4? These remain open questions, but the rank scale provides a principled grounding for exploring such questions systemically. Further, this way into addressing the analytical question of the relations between online newspapers, other news texts and other online texts more generally also leads into the theoretical perspective of viewing ranks axially: that is, not exclusively as class–function cycles, but also as bundles of systems organized by constituency rather than by metafunction or stratum (see Martin 2013).

DEVELOPING A MULTIMODAL RANK SCALE

63

FIGURE 3.3  Possible basic system of newspapers. Source: Knox (2009: 165).

FIGURE 3.4  Possible basic system of websites. Source: Knox (2009: 167).

REMAINING CHALLENGES When videos were first appearing in online newspapers, they tended to open on their own separate page. Now though, video stories can in many cases be watched from a home page, meaning that the function of Briefing and Dispatch in a content section (see Figure 3.1) appear on a single page. The rank scale as it currently stands does not account for this. Introducing elements such as ‘screen’ or ‘window’ in addition to page might be a potential solution. But video and audio (which is clearly not part of what has long been understood as a ‘page’) are now routinely included in online newspapers. Similarly, advertisements might appear as pop-ups, as animations that appear and disappear or that move around the screen. ‘Such elements clearly cannot be considered as parts which combine into whole zones which in turn combine into whole pages in the same

64

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

way as … [other elements discussed above]. This poses fundamental problems for any attempt to describe home pages, and therefore online newspapers using the structural principle of constituency’ (Knox 2009: 202–3). A separate, critical issue in online news that rank cannot address is the social takeover of news by social media. Legacy news institutions still exist, but people come to them via different platforms, and/or encounter their texts embedded in different platforms – so the once-prevalent reading path of home page to story page is far less common than it was in the earlier days of the world wide web (for the development of home pages over time, compare Knox 2009, 2014, 2020). Online newspapers still have their own structure, but they exist in a far broader set of intertextual and hypertextual relations that make the constituent structure of individual news sites less relevant to the reading experience of news consumers than it was when this research was conducted.

CONCLUSION The study this chapter is based on began by examining online newspapers. It quickly had to deal with the rapid transformation of these newspapers and the multimodal design of news during the course of conducting the study, and this rapid change actually became one of the foci (Knox 2009). Interestingly, though, many of the findings regarding website structure (based on the work on rank described above) remain relevant to online newspapers today (cf. Knox 2014, 2020). Had the initial study stuck to description, subsequent work would have been limited to describing the changes in multimodal news design (which are many). But with the benefit of the theoretical work on rank, long-term consistency in the semiotic structure of online news sites can be identified and explained. This consistency is despite differences in how newspapers construe the field of ‘news’, and despite dramatic changes in the visual design of home pages and story pages in the twenty-one years since the work began. The notion of rank is fundamental to SFL, and was conceptually important to some of the early SFL work on multimodality (e.g. O’Toole 1994; O’Halloran 2004). Developing, or attempting to develop, a rank scale forces clarity on the analyst in terms of units of analysis, levels of analysis and relations between units and levels. The identification of class and function, and the fundamental tenet of SFL that there is not a one-to-one correspondence between class and function, means that a rank scale requires formal identification of elements along with a grounded and consistent account of how they function. The extent to which rank scales can be developed for different modes and the value of doing so where it is even possible are open questions. These questions can only be resolved case by case, where analysts take on this particular task of theoretical creativity.

DEVELOPING A MULTIMODAL RANK SCALE

65

NOTES 1 Importantly, it also shows the periodic structure of the same website which can be explained by applying the work of Djonov (2005, 2007), and how a constituency (or rank) perspective and a periodic perspective are complementary. 2 See Knox (2009) for a worked example with more levels of embedding (cf. Knox 2014).

REFERENCES Baldry, A. and P. J. Thibault (2006), Multimodal Transcription and Text Analysis, London: Equinox. Bateman, J. A. (2008), Multimodality and Genre: A Foundation for the Systematic Analysis of Multimodal Documents, Hampshire: Palgrave Macmillan. Boczkowski, P. J. (2004), Digitizing the News: Innovation in Online Newspapers, Cambridge, MA: MIT Press. Cooke, L. (2003), ‘Information Acceleration and Visual Trends in Print, Television, and Web News Sources’, Technical Communication Quarterly, 12 (2): 155–81. Djonov, E. (2005), ‘Analysing the Organisation of Information in Websites: From Hypermedia Design to Systemic Functional Hypermedia Discourse Analysis’, PhD thesis, University of New South Wales, Sydney. Available online: http://handle. unsw.edu.au/1959.4/23915 (accessed 24 July 2021). Djonov, E. (2007), ‘Website Hierarchy and the Interaction between Content Organization, Webpage and Navigation Design: A Systemic Functional Hypermedia Discourse Analysis Perspective’, Information Design Journal, 15 (2): 144–62. Fowler, R. (1991), Language in the News: Discourse and Ideology in the Press, London: Routledge. Halliday, M. A. K. (1981), ‘Structure’, in M. A. K. Halliday and J. R. Martin (eds), Readings in Systemic Linguistics, 122–31, London: Batsford Academic and Educational. Halliday, M. A. K. (1985), An Introduction to Functional Grammar, London: Edward Arnold. Huddleston, R. D. (1981), ‘Rank and Depth’, in M. A. K. Halliday and J. R. Martin (eds), Readings in Systemic Linguistics, 42–53, London: Batsford Academic and Educational. Hudson, R. A. (1981), ‘Constituency in a Systemic Description of the English Clause’, in M. A. K. Halliday and J. R. Martin (eds), Readings in Systemic Linguistics, 103–21, London: Batsford Academic and Educational. Knox, J. S. (2007), ‘Visual-Verbal Communication on Online Newspaper Home Pages’, Visual Communication, 6 (1): 19–53. Knox, J. S. (2009), ‘Multimodal Discourse on Online Newspaper Home Pages: A Social-Semiotic Perspective’, PhD thesis, University of Sydney, Sydney. Available online: http://hdl.handle.net/2123/7696 (accessed 24 July 2021). Knox, J. S. (2014), ‘Online Newspapers: Structure and Layout’, in C. Jewitt (ed.), The Routledge Handbook of Multimodal Analysis, 2nd edn, 440–9, London: Routledge. Knox, J. S. (2020), ‘Multimodal Mobile News: Design and Images in Tablet-Platform Apps’, in H. Stockl, H. Caple and J. Pflaeging (eds), Shifts towards Image-Centricity in Contemporary Multimodal Practices, 233–52, London: Routledge.

66

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Kok, K. C. A. (2004), ‘Multisemiotic Mediation in Hypertext’, in K. L. O’Halloran (ed.), Multimodal Discourse Analysis: Systemic Functional Perspectives, 131–62, London: Continuum. Kress, G. and T. van Leeuwen (1990), Reading Images, Geelong: Deakin University Press. Kress, G. and T. van Leeuwen (1996), Reading Images: The Grammar of Visual Design, London: Routledge. Martin, J. R. (2013), Systemic Functional Grammar: A Next Step into the Theory – Axial Relations, Beijing: Higher Education Press. Martinec, R. (2005), ‘Topics in Multimodality’, in R. Hasan, C. Matthiessen and J. J. Webster (eds), Continuing Discourse on Language: A Functional Perspective, Vol. 1, 157–81, London: Equinox. Martinec, R. and A. Salway (2005), ‘A System for Image-Text Relations in New (and Old) Media’, Visual Communication, 4 (3): 337–71. O’Halloran, K. L., editor (2004), Multimodal Discourse Analysis: Systemic Functional Perspectives, London: Continuum. O’Toole, M. (1994), The Language of Displayed Art, Rutherford: Fairleigh Dickinson University Press. Sinclair, J. and M. Coulthard (1975), Towards an Analysis of Discourse: The English Used by Teachers and Pupils, Oxford: Oxford University Press. Zhao, S. (2010), ‘Rank in Visual Grammar: Some Implications for Multimodal Discourse Analysis (MDA)’, in A. Mahboob and N. K. Knight (eds), Appliable Linguistics, 251–66, London: Continuum. Zhao, S. (2011), ‘Learning through Multimedia Interaction: The Construal of Primary Social Science Knowledge in Web-Based Digital Learning Materials’, PhD thesis, University of Sydney, Sydney. Available online: http://hdl.handle.net/2123/8376 (accessed 24 July 2021).

CHAPTER FOUR

A Hip-Hop Battle Describing Sound in the Contested Academy DAVID CALDWELL

BACKGROUND: ‘WORD IS BORN’ This chapter recounts my PhD research (c 2006–10) which drew on Systemic Functional Linguistics (SFL) and social semiotics to examine Kanye West’s rap music1 (Caldwell 2011; see also Caldwell 2009, 2010, 2014a, 2014b; Caldwell and Zappavigna 2011). A particular aim of that work, and the focus of this chapter, was the development of system networks which described the distinctive semiotic affordances of the vocal sounds of the rap voice in comparison with the sung voice. Rapper Kanye West, and his album The College Dropout (West 2004), was chosen as my data for two reasons. First, West raps and sings, which made for a more valid comparative analysis. I was able to control key variables, such as gender, age and ethnicity, when comparing his rap and sung voice. I was also able to acquire instances in which West rapped and sang the same lyrics. Second, West was considered ‘popular’. He was ‘widely experienced and/or enjoyed’ (Hesmondhalgh and Negus 2002: 2). He also exemplified the ‘popular’ rap music genre: ‘Pop-Rap is a marriage of hip-hop beats and raps with strong melodic hooks, which are usually featured as part of the chorus section in a standard pop-song structure’ (Bogdanov et al. 2003: ix). Accordingly, my research was driven by an exploration of ‘popularity’ in hiphop culture. Why was The College Dropout widely experienced and enjoyed?

68

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

What is popular rap music? What makes a sung chorus ‘hook’ a listener? And how is that ‘hook’ different than for a rap verse? Drawing on SFL and social semiotics, these questions prompted a systematic description of the rap and sung performing voice.2 Of course, this descriptive agenda required me to locate my approach within a range of fields of inquiry, all of which, to varying degrees, had already examined vocal performance, hip-hop music culture and popular music. This descriptive work and my ‘battle’ to advance description in the field of popular music studies are the story of this appliable linguistics chapter.

THE FIELDS: ‘MY CORNER’ To be clear, my PhD thesis was in linguistics, drawing on the theory of SFL and social semiotics. In this way, as noted throughout this volume, my research was conducted in the spirit of Halliday’s conception of appliable linguistics (see Halliday 2003). In particular, I found van Leeuwen’s (2005) definition of social semiotics, informed by Halliday, to be especially helpful. The most defining characteristic of social semiotics is that it aims to describe semiotic systems such as language, sound, images, bodily movement, space and colour in relation to their cultural context. And this does not simply mean ‘contextualizing’ a semiotic reading. For social semiotics, culture is a semiotic phenomenon. The semiotic analysis presented in my research, and in this chapter, was intended to be a description of hip-hop culture, albeit one very specific instance of meaning making of that culture. Van Leeuwen (2005: 1) cites two further key principles of social semiotics of particular relevance to my research: 1. Social semiotics is not pure theory, not a self-contained field. It comes into its own when it is applied to specific instances and specific problems, and it also requires immersing oneself not just in semiotic concepts and methods as such but also in some other field … 2. Social semiotics is a form of inquiry. It does not offer ready-made answers. It offers ideas for formulating questions and ways of searching for answers. These two principles, driven by the systematic description of semiotic systems in/as context, were especially influential on my descriptive work. First, my work was applied to a specific instance of meaning making – Kanye West’s debut album The College Dropout (West 2004) – with the aim to examine a deliberately broad problematic: why is The College Dropout popular? Moreover, I was very much immersed in the field of rap music (e.g. Escher and Rappaport 2009). In this spirit of interdisciplinarity, I also drew heavily on fields outside of SFL and

DESCRIBING SOUND

69

semiotics, including musicology (e.g. Keyes 2002) and African-American studies (e.g. Smitherman 2000). In terms of van Leeuwen’s (2005) second principle, my semiotic analysis did not offer ‘ready-made answers’. The findings did not proclaim to provide definitive explanations as to why The College Dropout was popular. They were generative. In the case of this chapter, they generated a novel description of rap and sung voice. My descriptive work drew on fields that could assist with the close vocal analysis such as phonology and music theory, while at the same time speaking to disciplines such as sociology, media studies and African-American studies. Informing my interdisciplinary thinking at the time was Bernstein’s (1999) distinction between hierarchical and horizontal knowledge structures. For me, the hierarchical knowledge structures (the ‘scientists’ – phonologists, music theorists etc.) would ‘help me’ systematically describe the sounds and meanings of rap and sung voice, while the horizontal knowledge structures (‘social theorists’ – popular music scholars, ethnomusicologists, African-American scholars etc.) would help me explore ‘why they make the meanings they do’ (Martin and Rose 2007: 4). Thus, ‘my corner’ of SFL and social semiotics, applied to performance voice in hip-hop culture, was one of inclusivity. Like my subject Kanye West, my coding orientation was open to be ‘all things to all people’: West’s overall social identity is organic; a post-modern self. In fact, West is hyper post-modern. He has publicly rejected ideological discourses such as social class, gender and ethnicity. He has been outspoken against homophobia in hip-hop culture and has even rejected a music award on the basis that it was in the category of ‘urban’ music which he considered to be stereotyping his music in terms of ethnicity. West’s coding orientation is probably best explained by West himself: ‘You know that saying: you can’t be all things to all people? Well seriously, why not? I want to be all things to all people’ (Tyrangiel 2005: 8). (Caldwell 2009: 74) Social semiotics afforded me an interdisciplinarity and eclecticism, including within linguistics itself (phonology, acoustic phonetics, semiotics, etc.). At the same time, I wanted to find a way (read: ‘all things to all people’) to locate and legitimize my analytical and descriptive work within the large, diverse and often contested worlds of music scholarship (e.g. musicology, music theory, ethnomusicology) and ‘social’ studies (e.g. sociology, media studies, AfricanAmerican studies). For this, I turned to popular music studies, which is generally defined as having no disciplinary centre: ‘the study of popular music is, at its best, a uniquely interdisciplinary area of research, drawing significant contributions from writers within a number of academic fields, including musicology, media and cultural studies, sociology, anthropology, ethnomusicology, folkloristics,

70

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

psychology, social history and cultural geography’ (Hesmondhalgh and Negus 2002: 1). In summary, my ‘corner’ was SFL-informed social semiotics and popular music studies. Later in the chapter, I will recount my attempts to advance, or at least contribute to, the description of rap and sung voice in popular music studies. Before I do, I will recount the development of my novel description of the rap and sung voice.

NOVEL DESCRIPTION OF THE RAP AND SUNG VOICE: ‘MY FLOW’ Despite the number of studies focusing on rap music and hip-hop culture (at the time of conducting my PhD research, The Hiphop University Working Bibliography [Hiphop University n.d.] listed a total of 3,841 publications), no research had compared the semiotic properties of the rap and sung voice. Research had examined the sounds of the rap voice, including close, semiotic analyses (see e.g. Krims 2000). However, this was never compared with the sung performance voice. Moreover, no research had explicitly drawn on principles of SFL and social semiotics to describe the rap and sung voice. Van Leeuwen’s (1999) seminal text Speech, Music, Sound was, of course, especially influential on my thinking, as was Callaghan and McDonald’s (2007) comparative analysis of speech and sung voice. In each case, it was their foregrounding of the paradigmatic axis from SFL that inspired my description of the rap and sung voice into the form of system networks. At the same time, a key agenda in developing description was to follow the lead of scholars in popular music studies (e.g. Middleton 1990; Walser 1995) and move away from more syntagmatic descriptions of sound (based on traditional Western staff musical notation) and attempt to better capture the paradigmatic relations involved: ‘what is meant, against the background of what might have been meant but was not’ (Halliday and Greaves 2008: 94). The construction of the system networks was based on two linguistic units – the syllable and the tone unit or ‘sound act’ (van Leeuwen 1999). In the subsections that follow, I will present my system networks for rap and sung voice according to syllable then sound act. I will define the unit, recount my analysis and outline the networks produced from that analysis. I will then recount some of my attempts to draw on those system networks to examine the comparative meaning-making potential of rap and sung voice. The syllable My definition of syllable followed the phonological rank scale outlined by Halliday and Matthiessen (2004), where a syllable is defined as a unit of sound

DESCRIBING SOUND

71

clusters which is organized around a peak of high sonority, consisting of onset and rhyme, realized by consonant and vowel phonemes respectively. Or as Clark and Yallop explain, ‘a major peak of prominence represents the nucleus of a syllable and that this nucleus will usually be a vowel or vowel-like segment; consonants will generally occur as margins to these peaks, either as onset or coda’ (1995: 60). A key motivation for starting my description with the syllable was van Leeuwen’s (1999) notion of experiential meaning: our experience of what we physically have to do to produce a particular sound creates a meaning potential for that sound. For instance, we know that we can create tense sounds by tensing our articulatory musculature. We also know that we tend to do this in certain kinds of situations (when we are tense or ‘charged up’, or want to appear that way). Hence tense sound quality can come to be associated with aggression, repression, nervousness, excitement and so on. (van Leeuwen 1999: 205) Van Leeuwen (1999) argues that the meaning-making potential of sound is afforded by its materiality; meaning derives from the physical experience of producing a sound. Examining the syllables of the rap and sung voice, particularly the duration of vowels and consonants, was therefore critical to gaining insight into their respective physical production and, in turn, their experiential meaning. As noted in my introduction, I was able to access examples from Kanye West where he both rapped and sung the same lyrics, providing me with ‘minimal pairs’ to compare using the acoustic spectrogram analysis software Praat (Boersma and Weenik 2008). In Caldwell (2009), for example, the monosyllabic word ‘time’ /taɪm/ was presented through Praat. In terms of total duration, the rap voice for time was calculated at 0.287 seconds, compared with the sung voice of 0.777 seconds – a perceivable and measurable difference of 0.490 seconds. More specifically, the syllable nucleus (the diphthong vowel /aɪ/) of the rap voice was a total duration of 0.287 seconds, compared with the sung voice vowel at a total duration of 0.777 seconds. In terms of consonant production, for that same syllable, a duration analysis revealed that the initial plosive consonant /t/ for the rap voice (0.111 seconds) was in fact longer than the consonant equivalent for the sung voice (0.072 seconds), despite the significantly reduced overall duration of the rap syllable noted above. Drawing on this and many other similar examples, I constructed a syllable system network for the rap and sung voice (see Figure 4.1). The system network comprised two simultaneous systems: syllable nucleus and consonant aspiration (with the same entry condition of syllable). For syllable nucleus the network essentially presents a choice for the performance voice of shorter

72

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 4.1  System network for syllable for rap and sung voice.

or longer vowels. For consonant aspiration, the choice is spitting or sliding plosive consonants – the extent to which the vocalist articulates the explosion/ transitions into the vowel. What is critical to note is that these are clined systems, indicated by slanted network brackets. In other words, the choice between short and long syllable nuclei, or spitting and sliding consonants is not either– or. It is gradable. The system comprises a range of options between the two extremes of short–long and spit–slide. Moreover, because this is a comparative system network, these graded choices for the respective performance voice are intended to be read as weighted towards one or other of the performance voices. In other words, the rap voice tends to select shorter vowel length and greater consonant aspiration relative to the sung voice (see performance voice in brackets in Figure 4.1). The sung voice tends to select longer vowel duration and minimal consonant aspiration relative to the rap voice. The sound act Following van Leeuwen (1999), the unit of ‘sound act’ can generally be considered equivalent to a tone unit in SFL phonology (Halliday and Greaves 2008). The sound act was considered a general enough unit to capture both the rap and sung voice, despite the fact that each performance voice deploys rhythm and pitch in distinct ways. Technically, a sound act is framed by a rhythmic phrase of up to seven measures or rhythmic feet: ‘marked off from each other by breaks or changes in the regular rhythm of the pulses’ (van Leeuwen 1999: 211). Moreover, a sound act carries a discernible melodic contour, much like the tone unit. In terms of the phonological rank scale in SFL, a sound act, like a tone group, is identified by a tonic syllable – where the intonation movement is

DESCRIBING SOUND

73

most noticeable. To varying degrees, both the rap and the sung voice produce a comparable tonic syllable. However, the term ‘sound act’ was preferred to ‘tone unit’ because neither performance voice strictly construes English intonation (see Caldwell 2014a). Instead, they both comprise particular ‘musical’ qualities that spoken language does not typically foreground. For instance, an equivalent tonic syllable in the rap voice does not carry the typical intonation patterns of English. And in the case of the sung voice, it is often difficult to even locate an equivalent tonic syllable, given that many syllables are long in duration and carry an audible pitch movement. Nevertheless, both performance voices do tend to include a syllable that marks the end of a melodic phrase. If that syllable is not produced, the sound act feels incomplete (see van Leeuwen 1999). Once the unit of sound act was established, the next challenge was to find an effective and accurate means of describing that larger unit. To this end, I again drew on the acoustic spectrogram analysis software Praat (Boersma and Weenik 2008) to produce novel visual representations for the respective performance voices (see Caldwell 2009 for a detailed overview of the construction of these visual representations). Like the syllable analysis above, Praat meant that I was able to accurately measure the duration of a syllable or musical sound, as well as the duration of silence between those sounds. Given the limitations of an A4 page, each visual representation was limited to around five to six seconds in duration. This turned out to be an ideal total duration; sufficient time to capture a single sound act, while at the same time, producing visual representations that were legible. Moreover, five to six seconds generally equated to eight rhythmic pulses, or two bars of 4/4 metred music. This was ideal given that most rap music is based on quadruple time (Krims 2000: 52–3), and tends to align with van Leeuwen’s (1999) sound act (up to seven measures/feet). Once a sound act was selected in Praat, I constructed an identical horizontal axis using drawing software. The construction of the sound representation began with the percussive musical sounds that articulated the main 4/4 pulse of the music. These percussive sounds were easily located on the spectrogram as a sharp thin black line. I then used an ‘explosion’ symbol from the drawing software to represent the rhythmic pulse. Starting from the beginning of the sound act, I then used Praat to identify the first syllable (identified to fractions of a second). I then returned to the drawing software and produced an ‘oval’ symbol that was identical in duration to the highlighted segment. This process was repeated until all of the vocalist’s syllables were reproduced in the representation. It is worth noting that the oval and explosion symbols were translucent, which meant that the vocals and percussion could overlap, capturing the aural ‘reality’ of vocal and musical interaction. I also used the vertical axis to present any discernible prescribed pitch movement (or melody). Technically, the horizontal line was considered ‘middle C’. And a single, semitone movement up or down the scale was represented by four horizontal ‘clicks’ on the curser.

74

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

By way of illustration, Figure 4.2 compares a sound representation of West’s rapping voice (above) with a sound representation of his singing voice (below). These visual representations were especially illuminating, highlighting both the respective rhythmic and melodic affordances of the rap and sung voice. In the examples below we can observe that, while the rap and sung voice both articulate measured time (syllables aligned with the musical accompaniment), the sung voice tends to use the musical time as a point of departure for the syllable. The rap voice in contrast tends to overlap or synchronize with the rhythm of the percussive musical accompaniment. Another distinguishing feature is the melodically prescribed movement of the sung voice compared with the rap voice. The production of numerous visual representations for the rap and sung voice led to the construction of the system network for the sound act (Figure 4.3).

FIGURE 4.2  Visual representation comparing rap (above) and sung (below) voice from West. Source: Caldwell (2011: 123, 120).

FIGURE 4.3  System network for sound act for rap and sung voice.

DESCRIBING SOUND

75

Like the system network for syllable, the network for sound act comprised two gradable systems: melodiousness and time. The system of melodiousness presents a choice from single to multiple longer syllables per sound act. While the comparative length of syllables is captured in the syllable network above, this network recognizes that the rap voice typically comprises at least one long syllable per sound act, functioning to signal the end of a sound act (see e.g. the final syllable overlapping with pulse 8 in Figure 4.2). In comparison, the sung voice typically comprises multiple long syllables per sound act. The system of time comprises an initial choice between measured and unmeasured time. According to van Leeuwen, measured time has a regular pulse: ‘time you can tap your feet to’ (1999: 207). In contrast, unmeasured time has no discernible regular pulse, and as such, ‘you might at best sway your body to and fro’ (1999: 51). The choice between measured and unmeasured time was integral to the comparative description of rap and sung voice. According to van Leeuwen, there are two kinds of unmeasured time: continuous time, which lacks any form of phrasing and either does not vary in pitch at all, or wavers in pitch in slight and irregular ways, and fluctuating time, which also lacks phrasing, but does shift between pitches, at more or less regular intervals which are, however, too long to produce a clear sense of regular pulse or periodicity, a rhythm listeners could tap their feet to. (1999: 54) Only the sung voice has the potential to produce unmeasured time. The syllables in the rap voice are articulated too frequently to be classified as unmeasured. To be fair, there were very few examples of an unmeasured sung voice in van Leeuwen’s (1999) terms. Yet the sung voice was clearly distinguishable in the visual representations in terms of the extent to which it articulated time (see e.g. Figure 4.2). As noted above, the lengthy sung syllables used the main pulse as a point of departure, rather than articulating the time itself. Moreover, the sung voice produced fewer syllables in a given sound act. In embodied terms, the sung voice I analysed was harder to ‘tap one’s feet to’. In order to capture this reduction in rhythmic articulation, the system network construed the choice between measured and unmeasured along a cline, where the sung voice was considered to have the potential to be less measured than the rap voice. It was important for the system network of time to account for the way in which the measured voice engaged with the regular, metred 4/4 pulse of the musical accompaniment. Accordingly, there was an additional level of delicacy for measured time: a choice between degrees of synchronization. The system of synchronization is analogous to van Leeuwen’s (1999) system of metronomic and non-metronomic time, where synchronization is metronomic

76

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

and non-synchronization is non-metronomic time. According to van Leeuwen, metronomic time is precise and unwavering, whereas non-metronomic timing subverts measured time by anticipating or delaying the beat, or ‘syncopation’ in traditional music terminology. In contrast to van Leeuwen’s (1999) system of metronomic timing, the choice between synchronization and nonsynchronization was considered gradable for my network of time. This is because the system of time is conceptualized as a unit of sound act. It comprises multiple pulses. As such, it is gradable in terms of the extent to which the voice synchronizes with the musical pulses in a given sound act – all pulses, some pulses or no pluses. However, it was not possible to align the rap voice or the sung voice with being more weighted towards synchronized or nonsynchronized. The majority of sound acts for both performance voices were synchronized. And in the marked case of non-synchronization, it was rarely sustained. Both the rap and sung voice would quickly reconcile/synchronize with the musical pulse within a single sound act. Making meaning Having constructed the comparative system networks from Praat acoustic analysis (Boersma and Weenik 2008) and visual representations, I then shifted focus to the meaning-making potential of the respective vocal performances. To do this, I followed van Leeuwen (1999) and argued for an ‘experiential’ motivation to meaning making. What does the physical production of the rapping voice afford? What does the physical production of the sung voice afford? This step was not particularly challenging. From an experiential perspective, the choices identified in the system networks are the meaning-making realizations. For example, the rap voice is more plosive; it ‘spits’, and so it means ‘spit’, as well as any other near synonyms, such as ‘forced’ or ‘pushed’. While the open aperture and vowel length of the sung voice has the potential to mean ‘open’, ‘long’ and ‘unobstructed’. Or, for example, the time of sung voice, in contrast to the rap voice, has the potential to be ‘unmeasured’ – which can equate to experiential synonyms such as ‘free’ or ‘unbound’, compared with the ‘measured’ and ‘fixed’ rap voice (for more examples see Caldwell 2014a: 259). It is important to stress that these assigned meanings are potentials. And they are meant to be read as gradable and relative – that is, the sung voice is more ‘free’, relative to the more ‘bound’ rap voice and so on. In line with the principles of social semiotics outlined in the introduction (van Leeuwen 2005), I then ‘immersed myself’ in other fields in an attempt to develop the ‘social’ side of my semiotic analysis. Of course, the description to this point had already drawn on a range of fields, including phonology, music theory, ethnomusicology, musicology and like (again, broadly categorized under popular music studies). Nevertheless, it was clear that there was significant scope to develop the experiential meanings derived from my system networks.

DESCRIBING SOUND

77

For example, not long after establishing my system networks and novel descriptions, I drew on the sociology of DeNora (2000) to examine the various rap identity tropes expressed through the time system: synchronized and nonsynchronized (Caldwell 2009). Drawing on the analytical methods recounted in this chapter, I identified two distinct types of rappers across Kanye West’s albums: the ‘enforcer’ (rappers who more frequently synchronized) and the trickster (rappers who less frequently synchronized). From here, I drew on Bernstein (1999) and SFL identity theorizing (e.g. Bednarek and Martin 2009) to consider the role of Kanye West as a producer rapper. I concluded that his sampling of both enforcer and trickster tropes was illustrative of a weak coding orientation, in turn enabling West to affiliate with a large and diverse audience. In another study (Caldwell 2014b), I developed van Leeuwen’s notion of experiential meaning to consider how sound is received rather than produced. In other words, what is the physical experience of those who receive the rap voice and sung voice? How do we physically engage with the rap voice compared with the sung voice? Drawing on insights from ethnographies of hip-hop dance (e.g. Hazzard-Donald 1996), I found that the rap voice affords a more restricted physical response, in turn, construing a power dynamic in which the listener either physically follows the rapper, or elects to completely reject (and potentially ‘battle’) that rapper. The sung voice in contrast construes a more open and elaborated physical response, whereby solidarity and closeness is offered. In simple terms, you can more easily ‘sing along’ to the sung voice; in ‘chorus’ so to speak. In terms of dance, rap is individualistic; it does not afford the same solidarity as the sung voice. It is difficult to dance ‘cheek to cheek’ through a rap voice, for example. I also drew on Stenglin’s (2008) work on binding (as applied to museum studies) to explore the rap and sung voice – since ‘material’ sound waves present a more or less ‘bound’ space for an audience receiving that sound. Drawing again on my novel descriptions, I argued that rap’s consonant aspiration, short vowels, minimal pitch movement and regular time materialized as a more bound space. The sung voice on the other hand materialized a more unbound space through increased pitch movement and less regular time. These findings aligned neatly with work in gender studies (e.g. Keyes 2002), where hip-hop culture and rap music is considered an exemplar of hyper-masculinity. I then conceptualized West’s inclusion of the sung voice in The College Dropout (West 2004) as a more feminized soundscape, his popular rap ‘opening up’ the semiotic and somatic space to a broader, non-traditional audience.

DEVELOPING THE FIELD: ‘MY OUTRO’ My PhD research (c. 2006–11) recounted above has helped develop the description of performance voice in rap music. It has contributed to the

78

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

ongoing work in the field of popular music studies, which has long considered traditional music notation limited in accounting for genres outside of Western classical music (e.g. Middleton 1990; Walser 1995). Within that context, the application of principles from SFL and social semiotics (van Leeuwen 1999, 2005) presented an additional development to the field – foregrounding choice (paradigmatic), embodiment and insights from other disciplines – to describe the meaning-making potential of performance voice in hip-hop culture. The extent to which my description helped develop the field, however, is a much more provocative question and provides a neat ‘outro’ to this chapter. For a start, my descriptive work did not change the field of practice of rap music, nor did I expect it to. The more relevant question is the extent to which my work has developed description in a particularly interdisciplinary scholarly context. Drawing on the ‘science’ and the ‘social’, through social semiotics, enabled me to better unpack the meaning making of the performance voice – ‘why they make the meanings they do’ (Martin and Rose 2007: 4). In turn, this work is a contribution to description in SFL and social semiotics, particularly the limited research to date in performance voice and sound semiotics. However, presenting and publishing to/for those outside of social semiotics – the scientists and social theorists in popular music scholarship – was less encouraging. In the case of the music theorists and musicologists, my description was generally considered to be lacking in technicality. Similarly, for those working in social theory, I received feedback that my work needed to be more ‘culturally enriched’, and other similar critiques. By way of illustration, in a seminar to performance studies scholars, I presented sound bites, spectrograms and system networks. At the end of the presentation, I was asked about the extent to which I ‘considered culture’ in my analysis. I fumbled through my answer, only realizing later that my answer should simply be ‘I am mapping culture’. Unfortunately this same resistance extended to the interdisciplinary field of popular music studies, which I had hoped would provide the kind of disciplinary eclecticism necessary to locate my work from SFL and social semiotics. Instead, responses from key journals in that field also indicated a preference for the kind of social theory that my close sound description did not satisfy. In conclusion, and to be especially clear, I do not consider myself marginalized when recounting these experiences. Nor do I consider, for example, the research of musicologists or sociologists to be lacking or limited. My hope was to bring together insights from all these groups to inform my description. And I did just that through, and for, SFL and social semiotics. However, my additional hope was that this description would blur and even transcend discipline boundaries. I wanted to bridge hierarchical and horizontal knowledge structures (Bernstein 1999) – that music theorists, for example, would see value in the paradigmatic; that social theorists would want to examine the musical text more closely; or

DESCRIBING SOUND

79

that popular music studies would find a space for SFL and social semiotics. Perhaps, in time, they will. For now, I can happily say this work has contributed back to sound description in SFL and social semiotics. It is located in scholarship; it has a home. And while it may not be published by those working outside of SFL and social semiotics, it has been read and cited by those scholars. I may not be ‘all things to all people’ but, as Kanye West says, ‘they can’t say I didn’t try.’

NOTES 1 I use the terms ‘rap music’ and ‘rap voice’ to refer to a particular mode of expression that is one feature of hip-hop culture. Hip-hop culture itself comprises a range of semiotic modes, including hip-hop dance, hip-hop visual art, rap music and an ‘aesthetic’ expressed through stylized language, gestures and stylized dress. 2 The PhD research (Caldwell 2011) also included a complementary, comparative discourse semantic analysis of the rap verse lyrics and the sung chorus lyrics.

REFERENCES Bednarek, M. and J. R. Martin (eds) (2009), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, London: Continuum. Bernstein, B. (1999), ‘Vertical and Horizontal Discourse: An Essay’, British Journal of Sociology of Education, 20 (2): 157–73. Boersma, P. and D. Weenik (2008), Praat: Doing Phonetics by Computer Version 5.0.32. Available online: http://www.fon.hum.uva.nl/praat/ (accessed 1 December 2008). Bogdanov, V., C. Woodstra, S. T. Erlewine and J. Bush (2003), All Music Guide to Hip-Hop: The Definitive Guide to Rap and Hip-Hop, San Francisco: Backbeat Books. Caldwell, D. (2009), ‘Making Metre Mean: Identity and Affiliation in the Rap Music of Kanye West’, in M. Bednarek and J. R. Martin (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 59–80, London: Continuum. Caldwell, D. (2010), ‘Making Many Meanings in Popular Rap Music’, in A. Mahboob and N. Knight (eds), Appliable Linguistics, 234–51, London: Continuum. Caldwell, D. (2011), ‘The Rhetoric of Rap: A Social Semiotic Analysis of KanYe West’, PhD thesis, University of Sydney, Sydney. Caldwell, D. (2014a), ‘A Comparative Analysis of Rapping and Singing: Perspectives from Systemic Phonology, Social Semiotics and Music Studies’, in W. Bowcher and B. Smith (eds), Recent Studies in Systemic Phonology Volume I: Focus on the English Language, 235–63, London: Equinox. Caldwell, D. (2014b), ‘The Interpersonal Voice: Applying Appraisal to the Rap and Sung Voice’, Social Semiotics, 24 (1): 40–55. Caldwell, D. and M. Zappavigna (2011), ‘Visualising Multimodal Prosody’, in S. Dreyfus, S. Hood and M. Stenglin (eds), Semiotic Margins: Meaning in Multimodalities, 229–43, London: Continuum.

80

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Callaghan, J. and E. McDonald (2007), ‘A Comparative Study of Spoken and Sung Voice in Performance’, in Proceedings of the Third Conference on Interdisciplinary Musicology (CIM07), Tallinn, Estonia. Available online: http://www-gewi.uni-graz. at/cim07/ (accessed 20 November 2009). Clark, J. and C. Yallop (1995), An Introduction to Phonetics and Phonology, 2nd edn, Oxford: Blackwell. DeNora, T. (2000), Music in Everyday Life, Cambridge: Cambridge University Press. Escher, E. and A. Rappaport (2009), The Rapper’s Handbook: A Guide to Freestyling and Writing Rhymes, New York: Flocabulary. Halliday, M. A. K. (2003), ‘Language in a Changing World’, in J. J. Webster (ed.), On Language and Linguistics: Collected Works of M. A. K. Halliday, 213–31, London: Continuum. Halliday, M. A. K. and C. M. I. M. Matthiessen (2004), Introduction to Functional Grammar, 3rd edn, London: Edward Arnold. Halliday, M. A. K. and W. S. Greaves (2008), Intonation in the Grammar of English, London: Equinox. Hazzard-Donald, K. (1996), ‘Dance in Hip Hop Culture’, in W. E. Perkins (ed.), Droppin’ Science: Critical Essays on Rap Music and Hip Hop Culture, 220–36, Philadelphia: Temple University Press. Hesmondhalgh, D. and K. Negus (2002), ‘Introduction’, in D. Hesmondhalgh and K. Negus (eds), Popular Music Studies, 1–11, London: Arnold. Hiphop University (n.d.), Hiphop University: Working Bibliography. Available online: http://www.hiphoparchive.org/university/bibliography (accessed 1 April 2010). Keyes, C. (2002), Rap Music and Street Consciousness, Chicago: University of Illinois Press. Krims, A. (2000), Rap Music and the Poetics of Identity, Cambridge: Cambridge University Press. Martin, J. R. and D. Rose (2007), Working With Discourse: Meaning beyond the Clause, 2nd edn, London: Continuum. Middleton, R. (1990), Studying Popular Music, Milton Keynes: Open University Press. Smitherman, G. (2000), Talkin That Talk: Language, Culture and Education in African America, New York: Routledge. Stenglin, M. (2008), ‘Binding: A Resource for Exploring Interpersonal Meaning in Three Dimensional Space’, Social Semiotics, 18 (4): 425–47. Tyrangiel, J. (2005), ‘Why You Can’t Ignore Kanye’. Time, Available online: http:// www.time.com/time/prinout/0,8816,1096429,00.html (accessed 10 December 2008). van Leeuwen, T. (1999), Speech, Music, Sound, London: Macmillan. van Leeuwen, T. (2005), Introducing Social Semiotics, London: Routledge. Walser, R. (1995), ‘Rhythm, Rhyme, and Rhetoric in the Music of Public Enemy’, Ethnomusicology, 39 (2): 193–218. West, K. (2004), The College Dropout, CD, Roc-A-Fella/Island Def Jam.

PART II

Innovative Practice

82

CHAPTER FIVE

‘There are known unknowns … ’ Turning to Systemic Functional Linguistics to Understand Language Use in a Classified Military Context ELIZABETH A. THOMSON

INTRODUCTION In 2008, after sixty-five years of general proficiency language training within the Australian Defence Organisation under the purview of the Royal Australian Air Force, the Defence Force School of Languages (DFSL) was moved to the Australian Army. This triggered a number of fundamental changes to the language training programme. These changes were: 1. a requirement to incorporate the curricula into the army’s competencybased Training Management Package Support System (TMPSS); 2. the opportunity to evaluate the workplace relevance of the extant curricula through the army’s Centre for Lessons Learned, ‘whereby

‘There are known knowns. These are things we know that we know. There are known unknowns. That is to say, there are things that we know we don’t know,’ Donald Rumsfeld, 4 February 2004.

84

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

returnee personnel debrief army on their experiences and offer evaluations on equipment, operations, pre-deployment training and mission success’ (Thomson 2015: 20); 3. the directive in the Defence White Paper 2009 (Department of Defence 2009) for language training to be regionalized and converted to a blended learning mode of delivery; and 4. the development and promulgation of a new Language Capability Statement for Languages Other Than English (LOTE) (Department of Defence 2010), a policy which laid out a Language for Defence Purposes (LDP) requirement for language training. This was a fundamental shift away from the general proficiency courses assessed against proficiency scales to competency-based language training with criterion-referenced and standards-based assessment. See the appendix for a synopsis of the framework. These four drivers of change prompted a fundamental overhaul of the curricula at DFSL. This overhaul began with a two-stage language needs analysis study. The language needs analysis was carried out in accordance with the Defence Training Manual (Department of Defence 2006) – in the sense that it followed appropriately the five steps outlined in the ‘analyse phase’, which is the first of five phases in a cyclical loop moving from analyse to design, develop, conduct and evaluate. The analyse phase ‘directs training towards workplace performance needs’ (Department of Defence 2006: 1–1) and is thus oriented to job analysis, duty/task inventories, target student profiles, performance gaps, feasibility analysis and training specifications – all appropriate in a language needs analysis for workplace requirements. However, the manual was not written for workplace language needs analysis specifically, but rather for competency training in general across the trade training requirements of the Australian Defence Force. The fact that the needs analysis process was not tailored to language training meant that two important steps were missing: (i) data collection, involving authentic workplace language as well as workplace tasks – as outlined by Long (2005) and (ii) linguistic analyses of the authentic workplace language samples in order to determine the contextual features of the texts (the field, tenor – including social purpose – and mode) and the features of the texts themselves such as their semantic attributes and lexicogrammatical features – all of which are important (Ramzan and Thomson 2013) in the development of graded and sequenced, specificpurpose, text-based learning and teaching resources, whether they be online or face to face.

LANGUAGE USE IN A MILITARY CONTEXT

85

While these two critical steps were not part of the needs analysis, they were successfully incorporated into the design phase of the curriculum rewrite after discussions with army leadership. Having incorporated these two steps, it was then important to base the collection of authentic texts around workplace tasks identified in the needs analysis and to select an appropriate theory of language and language development on which to base the discourse analysis for teaching and learning purposes. In the process of selecting a theoretical approach, an important constraint imposed by the Language Capability Statement for LOTE needed to be taken into account. This was the Grade 3 category of defence linguist1 as described in the statement, namely ‘Grade 3 – the trade analyst – a soldier “with the Language for Specific Purpose (LSP) skills required by their employment category specifications”  ’ (Department of Defence 2010). This non-specific statement describes the employment category of intelligence analyst in the electronic warfare environment. The employment category specifications for a trade analyst are classified, thus explaining the non-specific nature of the Grade 3 description. In addition, and not unexpectedly, the nature of the trade analysts’ work is not available in the open source space. Nor was it possible for the Grade 3 language needs analysis process to investigate the workplace or investigate the duty/ task inventory of this category of employment. Thus the curriculum writing team was not able to collect authentic workplace language samples, leaving it uninformed as to the language requirement. Despite this, the curriculum writing team was directed to develop an appropriate language training course for the Grade 3 trade analyst category. It was within this curriculum writing context that the curriculum team chose Systemic Functional Linguistics (SFL) as the underpinning model of language and the genre approach (GA) (Rose and Martin 2012) as the appropriate theory and pedagogy on which to base the task of discourse analysis and curriculum development. The team took the position that SFL theory and GA can inform the practice of language learning curricula development in a classified context. It did this because of the particular theoretical characteristics of the model and the construal of additional or second language development. In terms of theoretical characteristics, SFL considers language a form of social behaviour; there is a twoway relation between language and context which is patterned and predictable. This construal of language as a social semiotic is manifested in texts which are grouped in populations or genres (Martin 2001) that share similar purposes, semantics and lexicogrammatical features. In terms of language development, the notion of genre and the learning/teaching cycle (LTC) used to build learners’ control of genre is underpinned by the features of the SFL model of language. Essentially, there is a complementarity between the model of language and the LTC. The cycle foregrounds the social uses of language, situates learning

86

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

in context, uses coherent texts as models and builds mastery through social, collaborative learning events. This chapter will demonstrate the innovative use of SFL and GA and their affordances in this unusual language learning context. The chapter demonstrates how the particular characteristics of SFL and GA, a text-based approach to language learning, guided the curriculum writing process. It will describe innovations in practice, critique the theory and offer some insights into the dialectic between theory and practice gleaned from the writing team’s experience.

THE WORKPLACE CONTEXT Before going further, it is worth considering the electronic warfare environment, to give a sense of the nature of the work of the trade analyst. To do this, a publicly available example from the air force is presented. This is the workplace task of identify friend or foe (IFF): ‘an identification system designed for command and control. It enables military and civilian air traffic control interrogation systems to identify aircraft, vehicles or forces as friend and to determine their bearing and range from the interrogator’ (Wikipedia 2019). Trade analysts play an important role in this task. First, technical analysts, such as air traffic controllers, identify the signature of a target object, be it from an aircraft, drone, missile, etc. The signature could be of various kinds, including its radar cross-sectional area, its electronic footprint or its communication transmission. This signature is triangulated against a database of known signatures. If the target object responds positively and matches a known signature in the database, it is positively identified. If there is no positive identification or positive response from the target object, a trade analyst will interrogate the object further (A. Wennerbom, personal communication, 11 January 2019). Amongst other possible interrogation techniques, a trade analyst with linguistic skills could indeed listen to command and control chatter on the radio communication systems of the target object to determine the object’s identity and intent. This linguistic analysis of the chatter will allow decision-makers to determine an appropriate response. The language training of a trade analyst in this context of IFF is thus about identification and intent of unidentified actors. This is one of the roles of the trade analyst. In a more general sense, the trade analyst functions to understand patterns of life and note when patterns change.

THE KNOWNS OF THE NEW CURRICULUM As mentioned above, the training needs analysis prescribed by the Defence Training Manual was conducted. Despite the Grade 3 trade analyst category

LANGUAGE USE IN A MILITARY CONTEXT

87

being classified, the needs analysis report (Booz & Company 2012) identified the following requirements: 1. The trade analyst work was understood as ‘operations without interaction’; that is, the analysts did not interact with interlocutors of the additional language. 2. Passive language skills, specifically listening to and reading texts that could have implications for operations or national security, were required. 3. Understanding of cultural issues that could impact operations was required. The linguistically trained trade analyst thus contributes to two important tasks in the workplace. First, they understand the oral and written texts of actors in the field (which could be in the air space, cyberspace or the battle space etc.). Second, they report this understanding up the chain of command. Put simply, they analyse language and report their analysis. The new curriculum thus needs to equip the trade analyst to analyse what they hear and read, and then report it. In order to report, the analyst also needs to convey its meaning to others, in effect a translation.2 Armed with SFL as a resource for understanding language in use, these minimally articulated ‘known’ requirements were indeed enough for the curriculum writing leadership team to determine the kinds of content and pedagogy required in the development of new LOTE courses for the trade analyst in the classified defence workplace.

CONTEXT AND TEXT SFL models language as a form of social behaviour in context (Halliday 1978). Language is instrumental in enacting our social relationships and social behaviour (Halliday 1973). Language, instantiated as text,3 is used to get work done in a particular cultural context. This notion of cultural context is described by Malinowski (1923), referring to the cultural assumptions, norms and traditions behind the participants and the kinds of practices of a cooperating social group of people. The relevance of this stance on culture for the curriculum writers relates to developing an understanding of the context of culture of foreign military actors, in order to ensure meaning is interpreted in a culturally appropriate manner. For example, in certain cultures where there are politeness constraints, such as in Japanese around giving superiors explicit instructions (Mizusawa 2008), the analyst would need to understand the nuances of indirect commands up the chain of command in order to interpret the message appropriately. The curriculum thus needs to incorporate an understanding of the target culture and how military life is conducted.

88

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

In tandem with this context of culture, is the context of situation,4 the ‘environment in which a text is actually functioning’ (Halliday and Hasan 1985: 46). In the context of situation, the text has a social function and achieves a social purpose. It consists of field, tenor and mode. Field is the kind of activity – the ‘play’; tenor relates to the roles played by the interactants – the ‘players’; and mode is the ‘part’ language plays and the channel of communication, how the text is being transmitted (Halliday and Hasan 1985: 45). Each of these aspects of the contextual configuration activates different kinds of meaning: field activates ideational meaning; tenor activates interpersonal meanings and mode activates textual meanings. These meanings are realized through lexicogrammatical systems which express the meanings, such as the system of transitivity for ideational meanings; the system of mood for interpersonal meanings; and the system of theme for textual meanings. Importantly, this relationship between context and text is bidirectional (Halliday and Hasan

FIGURE 5.1  Bi-directionality between language levels and social and cultural contexts. Source: de Silva Joyce and Feez (2012: 29).

LANGUAGE USE IN A MILITARY CONTEXT

89

1985: 55). We can discern much of the context from a text, and vice versa, we can discern much of a text from the context: the context of situation, the context in which the text unfolds, is encapsulated in the text, not in a kind of piecemeal fashion, nor at the other extreme in any mechanical way, but through a systemic relationship between the social environment on the one hand and the functional organization of language on the other. If we treat both text and context as semiotic phenomena, as ‘modes of meaning’, so to speak, we can get from one to the other in a revealing way. (Halliday and Hasan 1985: 12) This bidirectional feature of context and text, as illustrated in Figure 5.1, afforded the curriculum writers a strategy to make inferences from the context of situation about the kinds of text(s) used in the classified workplace. For example, the training needs analysis noted contexts relevant to operations.5 The curriculum team thus sought to understand the context of operations. By analysing the context of situation, particular text types can be predicted. Using contextual configuration analysis of the field, tenor and mode of IFF, it is predictable that: 1. Field would traverse topics such as flight path, altitude, speed, position, weather and destination using numerous military and aeronauticalspecific technical terms and codes. 2. Tenor would involve negotiation between personnel of different ranks, authority and responsibility, requiring appropriate terms of address, while maintaining very direct, specific, experiential choices of language to avoid misunderstanding and ensure mission success. 3. Mode would be spoken, via radio following radio transmission protocols. This particular contextual configuration activates a configuration of meanings, known as a register (Halliday and Hasan 1985: 38), which incorporates phonological expressions and lexicogrammatical and phonological features that realize these meanings. The register activated by the IFF context is an example of a restricted register (Firth [1950] 1957), a kind of register in which there is relatively little scope for individuality or creativity. Air crew have to learn the international language of aviation, a fixed language, in order to act as pilots and navigators and to communicate with ground control (Halliday and Hasan 1985: 39). The register of aviation is a predictable variety of language use. There is a limited set of social purposes, and a limited set of expressions, lexicogrammar and phonological features in play. Knowing that the register is restricted and is used for special purposes limits the kinds of texts that could occur in the

90

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

context of situation. The curriculum writing team was thus able to focus on these limitations having applied the notion of context of situation to military operational contexts to develop the language content for the curriculum.

GENRES Arranging texts into populations of texts which achieve the same broad social purpose and share semantic and lexicogrammatical features is to arrange texts into genres (Martin 2001; Martin and Rose 2008): ‘genre is a staged, goaloriented social process. Social because we participate in genres with other people; goal-oriented because we use genres to get things done; staged because it usually takes us a few steps to reach our goals’ (Martin and Rose 2007: 8). In the IFF example, analysts might be listening to air crew needing to exchange information relevant to safe flying, a routine pattern of life. The kind of genre which achieves this is a transactional, spoken information exchange (Burns and Joyce 1997: 25), specifically within the restricted register of aviation. As genres have recognizable staging and semantic attributes which are realized through lexicogrammatical choices, this afforded the curriculum writers a strategy to research and use open-source texts from the spoken genre of information exchange within relevant registers to substitute for and presumably approximate the classified texts of the genre. Knowing the genre and register, it was possible to identify the features of appropriate texts for each genre relevant to the trade analyst and thus incorporate these into the new courses. While the curriculum writing team did not know for sure if a text from the genre of information exchange was indeed the kind of text that a trade analyst would engage with, nonetheless, SFL theory suggests with high probability that they would. We know this because SFL theory models language as text in context with identifiable and predictable semantic and lexicogrammatical patterns.

TEXT STRUCTURE Table 5.1 shows a short, open-source transcript from a flight in the United States which, after takeoff, is handed off from the control tower to the departure controller (tailstrike.com n.d.). The departure controller manages flights as they ascend to cruising altitude. In this short spoken exchange, the first officer pilot greets the departure controller, and repeats their instructions from the control tower in relation to their flight path, which is then confirmed by the departure controller. The exchange signs off with the flight number. This is an example of a routine in-flight communication with the ground controllers. From the wording, it is possible to determine the actions or functions being undertaken through language. There is an exchange of greetings, identification and then information exchange related to heading and altitude.

LANGUAGE USE IN A MILITARY CONTEXT

91

TABLE 5.1  Table of In-Flight Example of a Spoken Information Exchange Using an Aviation Register Interactants

Wording

Language action

Aircraft First Officer

Atlanta Departure, Good morning. Regional Express eight sixty one’s with ya. 190 on the heading out of 2.4 for four thousand

Identify addressee Greet addressee Identify self (your aircraft no.) State information relevant to intended flight path: direction and altitude

Atlanta Departure Controller

Regional Express eight sixty one, Atlanta Departure. Good morning radar contact maintain four thousand.

Identify addressee, Identify self (departure control) Greet Inform aircraft of radar contact Direct flight path: altitude

Aircraft First Officer

Four thousand, eight sixty one

Confirm altitude Identify self (your aircraft) sign off

It is clear from the dialogue that there is a pattern to the exchange. Each interactant opens their first turn using a protocol that identifies whom they are talking to and whom they are and offers a salutation. Once this is complete, information is exchanged which is confirmed to ensure everyone knows what information has been transacted and subsequent action(s) confirmed. The exchange concludes with a sign off. Analysis of numerous such exchanges shows that this pattern, or schematic staging, is predictable. The lexicogrammatical choices which are selected to realize the meanings exchanged are also predictable and patterned. These choices are listed in the second column in Table 5.2. The curriculum writers analysed the lexicogrammatical resources required to express the meanings in this exchange. Further, they knew that this kind of exchange would occur highly frequently using the same generic staging with the same generic set of meanings or functions. Therefore, they could be confident that teaching the language of this information exchange text would set learners up to be able to understand, analyse and report on similar texts of this type. The lexicogrammar in this interaction is very straightforward to learn as the choices are restricted, being limited to the set listed according to lexis, grammar and phrasal expressions in Table 5.3. From the trade analysts’ perspective, it would not be necessary to interact in this dialogue; rather the analyst would only listen to and understand what is said. In relation to the macro-skill of listening, the new course would thus

92

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 5.2  Table of Meanings and Lexicogrammatical Features of the Information Exchange Text Example Language action (functions): meanings being made

Lexicogrammatical choices realizing meanings

1. Identifying addressee 2. Greet addressee 3. Identify self (your aircraft no.) 4. State information relevant to (intended flight path: direction and altitude)

1. Lexical vocatives: proper nouns for ground control entities 2. Interjection: conventional expression depending on time of day 3. Aircraft ID & identifying relational process, declarative mood 4. Identifying relational process (elided), lexis of direction in degrees, lexis of height in feet, navigational term: heading.

5. 6. 7. 8. 9.

5. Lexical vocatives: proper nouns for aircraft 6. Ground control ID & identifying relational process (elided) 7. Interjection: conventional expression depending on time of day 8. Possessive relational process: has/ have (elided) & technical lexis: radar contact 9. Imperative mood with material process: to maintain

Identify addressee Identify self (departure control) Greet Inform aircraft of radar contact direct flight path: altitude

10. Confirm altitude 11. Identify self (your aircraft) sign off

10. Lexis of height in feet 11. Aircraft ID

concentrate on listening in order to understand the context and topic with a critical ear for what is important, observing the patterns of life as either routine or non-routine for reporting purposes. This section has demonstrated the predictive power of the SFL model of language in determining what language to teach when a needs analysis cannot provide that information, in this case, due to the classified workplace. SFL theory provided the curriculum developers with an understanding of: (1) the bidirectional relationship between context and text, (2) the predictability of register which is activated by the particular contextual configurations of field, tenor and mode and (3) the fact that texts of a particular register, with a shared social purpose organize as genres with predictable schematic stages, meanings and lexicogrammatical choices. This knowledge provided the curriculum writers with a strategy for language curriculum development for language use in this classified workplace. Curriculum writers used this approach across a

LANGUAGE USE IN A MILITARY CONTEXT

93

TABLE 5.3  Table of the Set of Lexicogrammar and Phrasal Expressions Present in the Information Exchange Text Example Lexis

Grammar

Phrasal expressions

Numbers: E.g. 2.4, 190, 4000

Circumstances of duration: E.g. for four thousand (feet)

Elided classifiers of numbers related to flight: E.g. 2.4 (nautical miles), four thousand (feet), 190 (degrees)

Circumstances of location: E.g. out of 2.4 (nautical miles)

Interjections: good morning, good afternoon, good night

Identification nouns 1. For aircraft: proper noun + aircraft tail number 2. For ground control: proper noun

Elided identifying relational process in declarative mood: E.g. (This is) Regional Express 861, (This is) Atlanta Departure

Technical terms related to flight: E.g. Heading Radar contact

Circumstances of location: E.g. On the heading Elided possessive relational process in declarative mood: E.g. (We have) radar contact

Verb: E.g. to maintain

Imperative mood: E.g. Maintain 4000 (ft)

range of military operational contexts and macro-skills to decide on language content for the trade analyst curriculum. It was a process of answering the following questions within a military operations context: 1. What is the context of culture? 2. What is the field, tenor and mode of the context of situation? 3. What is the register? 4. What genres are used? 5. What meanings are being made? 6. What lexicogrammatical resources are activated to realize these meanings? Working methodically using this approach, the curriculum writing team built up a corpus of texts for the linguistically trained trade analyst. It was an exercise in theoretically informed prediction of the known unknowns.

94

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

COURSE STRUCTURE AND PEDAGOGY Writing the curriculum for the new courses was more than just working out what to teach; it was about how to teach as well. Course content was about selecting, grading and sequencing texts (de Silva Joyce and Feez 2012), while the selection of a methodology was about identifying an approach which enabled learning through a lens sympathetic to the features of the SFL model of language. GA, a methodology for teaching genres in context (Martin, Christie and Rothery 1987; Rose and Martin 2012; de Silva Joyce and Feez 2012), has been used effectively in literacy and additional language development in national and international contexts in school, vocational training and higher education settings (Joyce 1992; Burns and Joyce 1997; Burns and Coffin 2001; Coffin et al. 2003; Schleppegrell 2004, 2010; Byrnes 2006; Christie and Derewianka 2009; Ramzan

FIGURE 5.2  Teaching/learning cycle. Source: Rose and Martin (2012: 308).

LANGUAGE USE IN A MILITARY CONTEXT

95

and Thomson 2013; de Silva Joyce and Thomson 2015). The approach uses a teaching/learning cycle, represented in Figure 5.2, which introduces learners to texts (of targeted genres) in context, deconstructs the texts to reveal the lexis, the grammar, and the meanings and the stages of the texts so that learners can then jointly (with teachers and learners) and then independently construct their own texts to achieve a similar social purpose (de Silva Joyce and Feez 2012). The trade analysts’ course was to be developed as a text and skills-based, Language for Defence Purposes (LDP) course, assessed as competencies against criterionreferenced and standards-based assessment.6 The genre approach complements and applies the features of SFL theory in literacy and additional language classrooms. Using the teaching/learning cycle, it does this by: 1. situating learning in context; 2. modelling language as genres using coherent texts (instantiations of the genre) as examples; 3. building mastery through social, collaboratively learning using deconstruction, joint construction and independent construction as major steps in language skills development; 4. moving learning from simple through to more complex control of a genre; and 5. assessing the learners’ ability to perform authentic, language-based workplace tasks.

INNOVATIONS IN PRACTICE The trade analysts’ curriculum was developed to be rolled out across multiple languages. The development of the courses transformed the content development and delivery practices of the school in numerous ways. In the first instance, teachers developed new levels of language awareness as they worked in teams to research and predict the genres and identify the semantic and lexicogrammatical features of the genres in their respective languages. While the curriculum specification was written in English, it had to be interpreted and written specifically to accommodate the features of context and text for each language. Teachers followed a common development process based on the textbased curriculum (de Silva Joyce and Feez 2012) with common course learning outcomes and assessment instructions. Teachers became ‘text detectives’, seeking out authentic texts, interrogating their linguistic features and growing in their understanding of LDP. This was a new experience for the teachers of DFSL who had previously not been part of curriculum development, which had been the sole responsibility of the head teacher. It was also the first time the teachers were expected to produce a specific type of course (the trade analysts’

96

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

course in this case) in a specified language which was based on a generalized curriculum specification. As many of the teachers were civilians without military experience, the development teams included military subject matter experts (SMEs) to assist in selecting and or authoring texts for the courses. It became a truly collaborative effort between teachers and military experts. Together they built courses that incorporated the particular contexts of culture and situation of each language. The teachers contributed language awareness, semantic and lexicogrammatical knowledge, curriculum development experience, learning activities development and pedagogy, while the military SMEs contributed a deep understanding of the defence workplace, military doctrine and military tasks. This was transformative for the teachers and SMEs alike. Assessment practice underwent a transformation from general proficiency testing using proficiency scales to criterion-referenced, standards-based language testing situated within predicted workplace scenarios. Learners were assessed against specific criteria using rubrics that describe the standard of performance expected in the workplace. The complexity of the texts used in the assessments was mapped to an appropriate standard of the Australian Defence Language Proficiency Rating Scale (ADLPRS). This was an important innovation in practice as the teachers had to learn to: (1) assess using criteria, (2) write marking rubrics and (3) analyse texts against the standards used for learning and assessment to ensure they were at the appropriate level of complexity as dictated by the ADLPRS. In terms of delivery, teachers had to understand and adapt to the kind of teaching required by the teaching/learning cycle. This meant a concerted effort to reduce ‘teacher talking’ time and reliance on the grammar-translation method. Teachers facilitated learning, which included students working together collaboratively or individually, using the teacher as a resource rather than as an authority. The teachers also had to enable the expertise of the military learner to be part of the language learning process as, like the military SME, they bring knowledge of the workplace to the classroom. The teachers, military SMEs and learners each contributed to the classroom learning process in new and innovative ways. This facilitation role and team collaboration was a transformation of teaching practice. This kind of teaching practice was an example of joint construction, in the sense that the learners would act out a workplace scenario, after which the teacher would suggest additional language choices appropriate to the context and task. The lexicogrammatical features would be captured, listed, discussed, sequenced and practised with guidance from the teacher. The learners would co-create with the teacher, then in groups or individually learners reproduced the language in the context of the workplace task. A new review process was established which required all teachers, management and stakeholders such as workplace leads to evaluate student

LANGUAGE USE IN A MILITARY CONTEXT

97

and teacher feedback and agree on improvements. The courses were evaluated during the delivery and reviewed after each delivery iteration. This cyclical review process was then validated with the workplace to ensure the courses met the workplace requirements. The new courses brought considerable change and innovation into the professional practice of the teachers. These changes in design and development saw teachers become guides, providing ‘guidance through interaction in the context of shared experience’ (Rose and Martin 2012: 207) in the classroom. In terms of assessment, they became assessors of workplace language tasks situated in a military context. In evaluation, they became co-designers with learners and employers to improve each iteration of the course over time. These were significant innovations in teaching practice at the school which were equally embraced and resisted, depending on the individual teacher (an issue I take up below).

OPPORTUNITIES AND CHALLENGES Introducing a change in practice brought with it a number of opportunities and challenges related to: (1) the writing of competency-based curriculum, (2) the development of the content of the courses in specific languages, (3) the delivery of the courses using a facilitating, genre-based pedagogy, (4) the testing of workplace language usage using a criterion-referenced and standards-based assessment regime and (5) the continuous improvement of the curriculum and course content. There was a group of teachers who were ready for a change. They were tired of hearing from students that course content was not ‘based on real operational situations’ (de Silva Joyce and Thomson 2015: 28–9) and were interested in addressing this deficient in curricula: ‘The Indonesian curriculum development team is long in gratitude for taking us on this exciting adventure. The best is yet to come’ (team member, personal communication, 2012). They were therefore open to a new approach which, for the trade analysts’ course, meant that there was a strategy available to research and predict the workplace requirements around which to develop a specific course. These teachers typically were selected as a curriculum writing team. For the teachers who were challenged by the change, fears surfaced relating to their confidence as teachers, the relevance of their qualifications, exposure of their perceived literacy challenges and even fears relating to potential job loss (Thomson and de Silva Joyce 2013: 43). These teachers resisted the change. The resistance was understandable, as ‘people do not naturally resist change, they resist change they do not understand, the values of which they do not see, or the demands of which they cannot meet’ (Herold and Fedor 2008: 141).

98

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

The challenge for the curriculum leadership team was thus multifaceted. It included recognizing the knowing–doing gap of teachers – the ability of an organization to transfer knowledge (Pfeffer and Sutton 2000) – and understanding the post-change performance gap (Herold and Fedor 2008). This gap typically arises when the new skills and knowledge are still emerging for the staff, and so the impact of the change in practice is not immediate. There is a period of underperformance. Essentially, there is a dip in performance during the period of adjustment to change. While these two organizational phenomena are about institutional and management processes, it was nonetheless the teachers who needed systems in place to support them as they changed their professional practice. The institution needed to provide ongoing, systematic professional development addressing: 1. language awareness of English and LOTE, 2. SFL theory, genre pedagogy and metalanguage, 3. skills in discourse analysis, 4. skills in competency-based curriculum writing, 5. the development of task-based and text-based language learning activities for the classroom, 6. the delivery of content using the teaching/learning cycle, criterionreferenced and standards-based assessment and 7. quality assurance practices such as training evaluation, assessment moderation, course revision, instructional standards and workplace validation. Professional development (PD) which was put in place to support the new curriculum writing and course delivery practices consisted of weekly PD seminars, mentoring support for teachers using external consultants, an internal mini teacher training course addressing the key components of the generic curriculum, and the development of language-specific iterations of the curricula, pedagogy, assessment and evaluation. By far the biggest challenge was the adoption of a shared metalanguage for expanding the teachers’ knowledge of language, their ability to talk about the features of the genre-based curriculum and their ability to use metalanguage in the classroom to expand learners’ linguistic repertoires and workplace skills. It was very important not to discourage the teachers as they ‘practised’ their new teaching practices in development and delivery. In order not to overwhelm the staff, the strategy was to write student guides which had a dual role of inducting teachers while providing the learners with the content. The guides addressed: 1. the culture and society of the target languages, which included military culture;

LANGUAGE USE IN A MILITARY CONTEXT

99

2. language learning strategies; 3. language awareness using SFL metalanguage; and 4. criterion-referenced and standards-based assessment. Our motivation here was to demonstrate to the teachers in PD sessions how to effectively use the four student guides in the classroom. This approach indirectly taught staff about culture, context and text, the SFL model of language and introduced SFL metalanguage and learning strategies for the various macroskills required in the courses and assessment. In this manner, the amount of new knowledge and skills that the teachers were coming to terms with was measured. Teachers who resisted the change were sensitively inducted into the new framework via the student guides before entering the classroom. This ensured that teachers had an imperative to understand the guides, as they were to use them immediately in the classroom. Another challenge was the complexity of working across languages – in other words, instantiating a curriculum, which was written in English, in an additional language. While the curriculum was common, the generic staging, meanings and lexicogrammatical choices were language specific in nature. If teachers were not well versed in the structure, meanings and grammar of the additional language, then the development of the course content was challenging. In some languages, there was not a lot of SFL reference material available so it rested with the curriculum writing team to do some significant descriptive work even before thinking about course development. However, within this context, the writing team were relieved to learn that military doctrine is in many respects similar across defence forces. This relates to the historical development of military practices in Europe and modern-day coalition activities. Multinational military activity requires common practices so that all defence forces behave and act within an agreed set of practices, doctrines and rules of engagement. Using commonly understood military doctrine as a way to understand military activity assisted the writing team to construct military tasks in additional language contexts without relying on authentic doctrine from another country, which can be difficult to obtain.

CONCLUSION This chapter has demonstrated the power of SFL as an appliable model of language. It used a case study to illustrate the affordances of the model and the genre approach to teaching and learning in the design of curriculum for additional language use in a classified, military workplace. By modelling language as a social semiotic, motivated by social purpose in context, which activates predictable, patterned meanings and lexicogrammatical choices, teachers are provided with a tool to support them in curriculum work. SFL

100

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

and GA are powerful descriptive and teacher development tools. Their power lies in their appliability across multilingual language training, even in classified contexts. These tools enable the known unknowns to be identified for teaching purposes.

NOTES 1 A defence linguist is a defence member who is trained in an additional language and who has this language skill listed on their personnel file. It does not refer to a person who professes or practises the discipline of linguistics. 2 The text example within this chapter was not taken from the content of the classified language course. All examples have been used to exemplify the affordances of Systemic Functional Linguistics and the genre approach rather than to provide any particular details on the electronic warfare curriculum specification itself. 3 Text is ‘an instance of the process and product of social meaning in a particular context of situation’ (Halliday and Hasan 1985: 11). 4 The context of situation is the ‘environment of the text’ (Halliday and Hasan 1985: 6). 5 ‘Operations’ is the technical term for military activity covering both tactical activity, on the ground in the battle space environment, and operational activity, which is the command, control and coordination of tactical activity. The Australian Army uses an operational conceptual framework called adaptive campaigning (Head Modernisation and Strategic Planning – Army 2009). It has five mutually reinforcing and interdependent lines of operation which enable coalition activity with other defence forces: Joint Land Combat; Population Protection; Information Actions; Population Support and Indigenous Capacity Building (Head Modernisation and Strategic Planning – Army 2009: 28). 6 Criterion-referenced assessment is assessment against specific criteria; while standards-based is assessment which meets defined standards such as competencies and learning outcomes (de Silva Joyce and Feez 2012: 159).

REFERENCES Booz & Company (2012), DFSL Indonesian Training Needs Analysis Report V1.02, Melbourne: Booz & Company. Burns, A. and C. Coffin (2001), Analysing English in a Global Context, London: Routledge. Burns, A. and H. Joyce (1997), Focus on Speaking, Sydney: National Centre for English Language Teaching and Research, Macquarie University. Byrnes, H. (2006), Advanced Language Learning: The Contribution of Halliday and Vygotsky, London: Continuum. Christie, F. and B. Derewianka (2009), School Discourse: Learning to Write across the Years of Schooling, London: Continuum. Coffin, C., M. J. Curry, S. Goodman, A. Hewings, T. M. Lillis and J. Swann (2003), Teaching Academic Writing: A Toolkit for Higher Education, London: Routledge.

LANGUAGE USE IN A MILITARY CONTEXT

101

Department of Defence (2006), The Defence Training Model ADFA 7.0.2, Canberra: Defence Publishing Service. Department of Defence (2009), Defending Australia in the Asia Pacific Century: Force 2030, Defence White Paper, 2009, Canberra: Commonwealth of Australia. Department of Defence (2010), LOTE Capability Statement – Languages Other than English within Defence. Unpublished. de Silva Joyce, H. and S. Feez (2012), Text-Based Language Literacy Education: Programming and Methodology, Sydney: Phoenix Education. de Silva Joyce, H. and E. A. Thomson (2015), Language in Uniform: Language Analysis and Training for Defence and Policing Purposes, Cambridge: Cambridge Scholars. Firth, J. R. ([1950] 1957), ‘Personality and Language in Society’, in J. R. Firth (ed.), Papers in Linguistics (1934–1951), 177–89, Oxford: Oxford University Press. Halliday, M. A. K. (1973), Explorations in the Functions of Language, London: Edward Arnold. Halliday, M. A. K. (1978), Language as Social Semiotic, London: Edward Arnold. Halliday, M. A. K and R. Hasan (1985), Language, Context, and Text: Aspects of Language in a Social-Semiotic Perspective, Geelong: Deakin University Press. Head Modernisation and Strategic Planning – Army (2009), Army’s Future Land Operating Concept (AC FLOC), Canberra: Army Headquarters. Herold, D. and D. B. Fedor (2008), Leading Change Management: Leadership Strategies That Really Work, Palo Alto, CA: Stanford University Press. Joyce, H. (1992), Workplace Texts in the Language Classroom, Sydney: NSW Adult Migrant English Service. Long, M. (2005), ‘Methodological Issues in Learner Needs Analysis’, in M. Long (ed.), Second Language Needs Analysis, 19–76, London: Cambridge University Press. Malinowski, B. (1923), ‘The Problem of Meaning in Primitive Languages’, Supplement 1 in C. K. Ogden and I. A. Richards (eds), The Meaning of Meaning, 146–52, London: Kegan Paul. Martin, J. R. (2001), ‘Language, Register and Genre’, in A. Burns and C. Coffin (eds), Analysing English in a Global Context, 149–66, London: Routledge. Martin, J. R. and D. Rose (2007), Working with Discourse: Meaning beyond the Clause, London: Continuum. Martin, J. R. and D. Rose (2008), Genre Relations: Mapping Culture, London: Equinox. Martin, J. R., F. Christie and J. Rothery (1987), ‘Social Processes in Education: A Reply to Sawyer and Watson (and Others)’, in I. Reid (ed.), The Place of Genre in Learning: Current Debates, 46–57, Geelong: Centre for Studies in Literacy Education. Mizusawa, Y. (2008), ‘Investigating the Directive Genre in the Japanese and Australian Workplaces: A Systemic Functional Approach’, PhD diss., University of Wollongong, Wollongong. Pfeffer, J. and R. I. Sutton (2000), The Knowing–Doing Gap: How Smart Companies Turn Knowledge into Action, Boston, MA: Harvard Business School Press. Ramzan, Y. and E. A. Thomson (2013), ‘Modelling Writing: Using the Genre Approach in the Japanese as a Foreign Language Classroom’, in E. A. Thomson and W. Armour (eds), Systemic Functional Perspectives of Japanese: Descriptions and Applications, 254–84, London: Equinox Publishing. Rose, D. and J. Martin (2012), Learning to Write, Reading to Learn: Genre, Knowledge and Pedagogy in the Sydney School, Sheffield, UK: Equinox Publishing. Schleppegrell, M. (2004), The Language of Schooling: A Functional Linguistics Perspective, Mahwah, NJ: Erlbaum.

102

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Schleppegrell, M. (2010), ‘Supporting a Reading to Write Pedagogy with Functional Grammar’, in C. Coffin (ed.), Language Support in EAL Contexts: Why Systemic Functional Linguistics? – Special Issue of NALDIC Quarterly, 26–31, Oxford: NALDIC. tailstrike.com (n.d.), Cockpit Voice Recorder Database, 2002–2018. Available online: https://www.tailstrike.com/080692.htm (accessed 6 February 2019). Thomson, E. A. (2015), ‘Language at the Pointy End: Understanding the Additional Language Needs of the Australian Defence Force’, in H. de Silva Joyce and E. A. Thomson (eds), Language in Uniform: Language Analysis and Training for Defence and Policing Purposes, 18–40, Cambridge: Cambridge Scholars. Thomson, E. A. and H. de Silva Joyce (2013), ‘Professional Development – A Pillar of Curriculum Change’, Professional and Academic English, 41: 36–44.36. Wikipedia (2019), Identification Friend or Foe. Available online: https://en.wikipedia. org/w/index.php?title=Identification_friend_or_foe&oldid=882881249 (accessed 18 January 2019).

APPENDIX: LANGUAGE USER TYPE DESCRIPTORS Language user grade

Grade description

Grade 1

The culturally enabled user – a user with intercultural competence

Grade 2

The vocational user – ‘with sufficient command of the target language to conduct their routine primary duties in the LOTE but are not able to function as interpreters or translators’

Grade 3

The trade analyst – a soldier ‘with the LSP skills required by their employment category specifications’

Grade 4

The professional user (operational) – ‘with sufficient command of the target language to act as communication facilitators within the restricted settings of routine military tasks and the specialist knowledge encompassed by their specific corps/mustering, trade and rank’

Grade 5

The professional user (strategic) – ‘with sufficient command of the target language to act as communication facilitators within the settings of routine and non-routine military tasks and are able to operate beyond the scope of their specific corps/mustering, trade and rank’

Grade 6

The military language instructor who has ‘achieved language Grades 4 or 5 and [is] selected for posting to DFSL and regional language training centres as instructors’

Grade 7

The professional interpreter – ‘with verbal and written language proficiency equal to target language native speakers, able to conduct professional translating and interpreting tasks to nationally accredited standards’

Source: Department of Defence (2010).

CHAPTER SIX

Designing Pedagogic Registers Reading to Learn DAVID ROSE

INTRODUCTION This contribution attempts to convey the interplay between theory and application in the development of the Reading to Learn literacy methodology. Applications and extensions of SFL genre and register theory are illustrated with the design of a series of reading and writing lessons that are embedded in a curriculum unit on Indigenous Australian culture. This device exemplifies an approach to analysing written genres focused on structures at the level of register, known as phases, that has been developed in the methodology. It then illustrates the approach developed for analysing and designing pedagogic practices, in which knowledge and values are exchanged by teachers and learners. This approach focuses on variables in pedagogic register, including (i) the structuring of pedagogic activities, at the scales of lesson stages and teacher/learner interactions; (ii) the sources of meanings in spoken, written, visual and gestural modalities; and (iii) the roles of teachers and learners in pedagogic interactions. The challenge that led to Reading to Learn (R2L) was to develop a teaching methodology that could effectively democratize engagement with and outcomes of formal education. This long-term action research project began in the 1990s with teachers of Indigenous students in Australia (Rose 2010). Early in the

104

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

project it became apparent that the struggles of these students with school education were experienced to some extent by students from all backgrounds, in all levels and sectors of education. The problem was not with one group of students or another, but with teaching practices that failed to do what they were supposed to: educate students in school. The methodology therefore had to be appliable in all these domains, but also reproducible for teacher education. Existing tools for its design included Martinian genre and register theory (Martin 1992; Martin and Rose 2008), genre-based writing and scaffolded reading pedagogies (Christie and Martin 1997; Rose, Gray and Cowey 1999) and Bernsteinian sociology of education (Bernstein 2000). These theories and applications were drawn on and extended to produce close analyses of curriculum genres that could enable teachers to re-design their practice. Significant theoretical advances have included analysis of the structuring of register in knowledge genres and curriculum genres. Current work includes detailed description of pedagogic registers, extending the application of systemic functional tools from denotative to connotative semiotics. From the 1980s, the focus of genre-based literacy pedagogy had been on improving students’ success through explicit modelling of writing for assessment. This project entailed appliable descriptions of common written genres in school and further education, termed knowledge genres, to provide models for writing, and design of practices for teachers to guide writing, termed curriculum genres. Descriptions of knowledge genres focused on variations in staging that distinguished types of stories, explanations, reports, procedures and arguments (Rose and Martin 2012; Rose 2015). A curriculum genre dubbed the ‘teaching/learning cycle’ (TLC) was similarly described in three stages, as deconstruction, joint construction and independent construction (Rothery 1994). In the 1990s, the Write it Right research project extended descriptions to knowledge genres found in secondary school curricula and a variety of workplaces (Christie and Martin 1997; Martin and Veel 1998). The TLC curriculum genre was refined to include ‘building field’, ‘setting context’ and ‘critical orientation’ through all three stages. As genre descriptions were recontextualized for teacher training, linguistic realization of knowledge genres focused on grammatical syndromes that were considered typical for various genres, particularly variations in process types, participants, circumstances, conjunctions and clause themes. Building on this earlier work, the Reading to Learn pedagogy required several shifts in focus for both linguistic analysis and recontextualization for teacher education. First, a set of further curriculum genres had to be designed for embedding reading in curriculum teaching, for scaffolding detailed comprehension of challenging texts, and for teaching beginning reading and writing for first and other language learners. Second, the role of language

DESIGNING PEDAGOGIC REGISTERS

105

knowledge in pedagogy needed refocusing, from using ideal genre models and typical language features to inform writing, to identifying instantial patterns in reading texts, and using them as models for writing. Third, finer grained analyses of knowledge genres were needed to give teachers and learners frameworks for identifying instantial patterns in reading texts, and borrowing them into their writing. Finally, close analyses of classroom discourse were needed, for teachers to design learning interactions around reading and writing that would involve and support all students equally despite highly variable literacy skills. Driven by the demands of teachers in the R2L professional learning programme, design of these analyses and design of curriculum genres for the classroom and teacher training preceded their theorizing. Over time, models have emerged from this dialogue between action research and SFL theory of relations between genre, register and language, for both curriculum genres and knowledge genres.

ARCHITECTURE FOR DESIGN: REALIZATION, INSTANTIATION AND INDIVIDUATION The interpretation of learning in the R2L model draws on SFL theory, taking into account three systemic hierarchies of realization, instantiation and individuation, each of which emerge from relations in semiotic systems. In broad terms, a system is a set of relations of abstraction and generality. Its features are generalizations based on recurrent similarity of function at higher strata. From this instantial perspective, a system is an accumulation of instances over time scales of phylogenesis and ontogenesis. In order to recognize a system, instances must be perceived recurrently, in association with familiar functions. The association of features with functions is a social relation, as it arises from a community’s shared recognition of recurrent instances. Individuation is a scale of such social recognition, from cultures to communities to personae (Martin 2010). Semiosis arises from these socially recognized couplings of abstraction with recurrent generalizations. From a realization perspective, each feature in a genre system (each genre) is realized structurally by a unique pattern of staging, which is realized at the next stratum of register, by a distinct configuration of field, tenor and mode variables. Register features are realized in language by selections from discourse semantic systems, realized by lexicogrammatical selections, realized by selections in expression systems (phonological or graphological). Register features may also be realized by other modalities, such as image or gesture systems. While each genre is realized by a predictable register configuration, in order for texts to differ there must also be scope for variations within these register

106

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 6.1  Realization, instantiation and individuation.

settings. And this variability must also extend to stratal relations between register, discourse, grammar and expression. All strata instantiate, meaning that texts vary in how each stratum is instantiated, within predictable ranges of mutual expectancy. The totality of meaning in unfolding text is a product of this multi-stratal co-instantiation or logogenesis. All strata also individuate, meaning that readings vary in how much systems at each stratum are shared by speakers and communities. Intersections of realization, instantiation and individuation are schematized in Figure 6.1 (after Martin 2010). Instantiation and individuation intersect at the point of individual readings. One implication for pedagogic design is that newly encountered systems are learnt by experiencing recurrent instances in familiar contexts, suggesting that some degree of expectancy is required to recognize new features. Expectancy may flow from general features to particular instances, or vice versa, or from abstract functions to realizing features, or vice versa. Instances vary independently in genre and register, so that patterns in both, as well as in language, must be recognized to comprehend texts and reproduce their features.

DESIGNING PEDAGOGIC REGISTERS

107

KNOWLEDGE GENRES In order to guide students to read and write knowledge genres, teachers need analyses that can be applied to the instantial variability between texts, as well as their generic similarities. The analysis of knowledge genres in the pedagogy focused on features of register realized as structures termed phases (Rose 2006a; Martin and Rose 2008). Phases tend to be expressed at the scale of paragraphs in written texts. While the stages of each knowledge genre are highly predictable macro-structures, phases within each genre stage are more variable. Types of phases are associated with families of genres. Phase types are therefore a step down the instantiation cline from general systems of genre and register, as they co-instantiate or couple selections in both strata (Martin 2006). For example, there are phases associated with story genres. Story phases function to move the plot forward, contextualize its elements, and engage listeners/ readers by manipulating expectancy, in highly variable patterns. They are basic building blocks of story construction across story genres, in spoken and written modes. They have also been found to be common across cultures (Rose 2019). Figure 6.2 maps story phases relationally, in terms of their functions to either contextualize or sequence the story’s plot. Settings precede other phases, presenting the time, place and activity of characters. Descriptions pause the action to describe qualities or behaviours of a character, thing or place. Comments stop the action for the narrator to explain or evaluate the activity. Reflections are characters’ thoughts about the activities,

FIGURE 6.2  Story phases.

108

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

which may pause or continue the action. Problems create tension through activities that disrupt expectancy. Solutions release tension through activities that restore equilibrium. Expectant relations between phases may be simply a sequence of events or consequential. Results are material consequences, which may be an activity or quality. Reactions are characters’ feelings, expressed as behaviours or qualities. By way of example, Table 6.1 presents a traditional story of the Saibai people of Torres Strait. The Story of Girbar is published as an illustrated children’s book, translated from Kalaw Kawaw language to English (Saibai Community 2011), and is a curriculum text in Australian primary schools. The story genre is narrative: oriented with activities of everyday life, disrupted by a switch to the sacred world of rainmaking spirits, and resolved by a return to the everyday with sacred gifts. The orientation begins with a setting, which identifies the central character Girbar and her activity. Tension is then introduced with a problem that foreshadows the complication, which begins with the appearance of Madhubal rain spirits. Girbar’s fearful reaction invokes the danger associated with the sacred world, and she reflects on the power of the spirits, who then take her on a magical journey from earth to sky. The tension created by this series of unexpected events is released as they hand her sacred gifts of food crops. To this point, the perspective has been Girbar’s, as she hurries, notices, realizes and speaks to the Madhubal, and they to her. In the resolution, the perspective shifts to the Saibai people. The last two phases recount the people’s rainmaking ceremony, and bring the mythic history into the present day, synchronizing the ceremonial activities of the people and the spirits. In terms of textual structure, each of these phase shifts is signalled by a marked theme, underlined in the text, or by Girbar’s mental processes in reaction and reflection phases, shown in italics. The switch from sacred back to everyday worlds is signalled by a thematic switch to the Saibai people. In the published version, each phase is presented as a paragraph on a separate page with an accompanying illustration. The labels for each phase in the left-hand column are metalinguistic terms for generalized types of phases, derived from studying stories across genres, modes and cultures. The labels on the right are more field specific, derived partly from common themes in myths across cultures, such as the switches between everyday and sacred worlds, the protagonist’s vision of the sacred, the magical journey, the sacred gifts, the sacred origins of cultural practices and the synchrony of divine and human ceremonies. In terms of the model outlined above, the generalized register structuring of story phases is co-instantiated with these general mythic themes, and the specific plot of this story. Considering instantiation in terms of generality, from system to text type to text, the story phase system is coupled with the story genre system, across cultures. Myths are then a story type that couples certain

DESIGNING PEDAGOGIC REGISTERS

109

TABLE 6.1  Stages, Phases and Themes in a Traditional Story Phases

Themes

Orientation

Everyday world

Setting

Late one afternoon, a woman named Girbar set out from her home at Ziril to fetch water from the Mag. She was carrying four kusul. There was a drought and the water was low, so she had to climb down the banks to the bottom of the waterhole.

domestic work

Problem

By the time she had collected the water, the sun had set and it was growing dark. In a hurry to cross the swampy country and reach her home before nightfall, Girbar quickly put the kusul filled with water over her shoulder and turned to climb the banks of the waterhole.

foreshadow danger

Complication

Sacred world

Description

As she did so, she noticed the dark reflections of the Madhubal on the surface of the water. Madhubal are rainmakers, rain spirits. They are magical beings who look like humans and wear big headdresses. They live in the clouds in the sky, and they dance and sing to make the rain fall.

Reaction

danger of sacred Girbar realized she had been surrounded by the Madhubal, who were standing around the edge of the waterhole. ‘Why have you come here, and what do you want?’, she asked, feeling frightened. She’d never seen these people before. ‘Girbar’, said the Madhubal, ‘we have come to take you with us to the sky.’

Reflection

Girbar did not wish to go with the Madhubal, but she had no choice. For she was only one and they were many. They had magic and special powers. They were spirits, not humans.

Solution

magical journey In the blink of an eye, they had taken Girbar back to their home in the sky. ‘We’ve brought sacred gifts you here for these three things’, they told her. They gave her three food crops: dhamuway, dhowpay and guru. ‘Go back now, and take these with you’, they said. Then the Madhubal became invisible, and Girbar could no longer see them.

vision of sacred world

power of sacred

110

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Phases

Themes

Resolution

Everyday world

solution

The people of Saibai watched and waited for Girbar for many days. At last, they saw her returning from the sky. She climbed down a coconut palm shaped like a cross at a place called Bagunangulaynga.

Result

sacred origins of In her arms, she had banana, taro and sugar cane, none of which had ever grown on Saibai culture Island before. She came as a messenger with stories about the seasons – how and when to plant the foods according to the stars and tides. People planted the gifts and prepared to sing and dance with the Madhubal. They danced so the rains would come, and their new crops would flourish and spread across the island.

Result

Wearing dhibal, they performed the spirit dance for rain, dancing to open the sky. They sang:

return

origins of prayer

Madhubaal madhubaal ngoeyk ay madhubaal. Madhubaal madhubaal ngoeyk ay madhubaal. Ina motheie motheie a ina kapoeli kapoeli a. Ina motheie motheie a ina kapoeli kapoeli a. Result

Carrying long sticks, they beat the surface of the water, strengthening the power of the rain spirits and calling the rain towards their gardens. Still today, the Madhubal make the rain by singing and dancing in the clouds, and Saibai Island rainmakers sing and dance with them.

synchrony of heaven and earth

field options with story phases, such as the mythic themes above, along with specific characters and places in a non-specific ancient time. These general mythic elements are then coupled with the particular characters, settings and events of this story. With respect to individuation, only the latter reading of the story plot is available to most readers. The mythic themes may only be consciously accessible to senior members of Saibai culture, and perhaps to mythologists as general categories. However, the story also has a pedagogic function, in constructing reading positions for younger Saibai people around the mythic themes. It does so by manipulating expectancy and perspectives, from phase

DESIGNING PEDAGOGIC REGISTERS

111

to phase. Listeners are engaged in characters’ perspectives through their thoughts, feelings, perceptions and speech. Deepening engagement with Girbar’s inner experiences leads the listener to expect and dread harm to her, so the sacred gifts come instead as a marvellous release. By these means, expectancy through phases enacts an exchange of value between the young listeners and their ancestors, mediated by their elders who tell them the story (Rose 2019). Register expectancy thus has dual functions in a story: it is an instantial relation of probability between fields and their elements that drives the plot from phase to phase, and this relation is negotiated between speaker and listener or writer and reader, to construct reading positions. It also intersects instantiation with individuation, as readings depend on repertoires of knowledge and values. In the R2L literacy pedagogy, phasal analysis has proven highly effective for modelling the structuring of all knowledge genres. With guidance and practice, teachers rapidly learn the techniques of analysis, and use them to plan reading lessons and to model genre and register structuring for writing.1 Detailed linguistic knowledge is not required for the pedagogic phasal analyses, as the approach is from register, which is the stratum most accessible to intuitive pattern recognition. Teachers learn to look for shifts in field, guided by the system of story phase types, paragraphing, and textual markers such as marked themes and internal conjunction. A pedagogic metalanguage has been developed for each genre family, described in Martin and Rose (2008), Rose (2020) and Rose and Martin (2012).

CURRICULUM GENRES The structuring of curriculum genres has previously been approached in SFL research from Bernstein’s (2000) sociological description of pedagogic discourse as embedding an ‘instructional register’ in a ‘regulative register’ (Martin 1997; Christie 2002). Extensive study of learning interactions has led to a re-analysis of these dimensions, guided by SFL principles of register variation (Rose 2004, 2018, 2020; Martin 2006; Martin and Rose 2007a, 2012; Kartika-Ningsih and Rose 2018). While knowledge genres configure selections in field, tenor and mode, curriculum genres configure two registers together, a curriculum register of knowledge and values, and a pedagogic register of activities, modalities and teacher/learner relations, through which knowledge and values are exchanged, symbolized in Figure 6.3.2 It is proposed that this model applies to cultural learning in general. As with phase systems associated with knowledge genre families above, curriculum and pedagogic registers are subsystems of overall register systems associated with types of curriculum genres. They are likewise a step down the instantiation cline from general systems.

112

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 6.3  Curriculum genres configure two registers together.

Pedagogic activities are centred on learning tasks undertaken by learners. Learning tasks may be prepared and focused by teachers, who also usually evaluate them, and may elaborate on the learning. Learning tasks can be identified at three broad scales: of lessons and lesson stages of activities within each lesson stage, and of teacher–learner interactions, where the task is often to respond to teacher questions. This tier of pedagogic activity has been termed learning cycles, as it often involves cycles of focus questions, responses and evaluations (widely known as ‘initiation-response-feedback’ or IRF cycles). Pedagogic modalities are the sources of meanings, including teachers’ and learners’ knowledge, recorded texts and images, and the environment, and the means of sourcing them into the classroom discourse, through speaking, gesturing, writing and drawing. Pedagogic relations negotiate the activities of learning, in which teachers and learners take complementary but asymmetric roles to co-construct curricular knowledge and values. Teachers’ roles broadly include presenting knowledge, evaluating learners and directing activities, while learners display knowledge or receive it, and both teachers and learners may solicit knowledge and actions from each other (Rose 2018, 2019). These categories of pedagogic activities, modalities and relations have emerged from diverse studies of learning interactions, informed by register theory and SFL research methods (Rose 2014, 2018). Initially, the structuring of activities was analysed to inform design of teacher–learner interactions in reading instruction (Rose 2004; Martin 2006; Martin and Rose 2007b; Rose and Martin 2012). Analysis of pedagogic modalities was motivated by the need to explain how teachers and learners derive meanings from various sources in classrooms, including texts, images and their own knowledge. Sources

DESIGNING PEDAGOGIC REGISTERS

113

and sourcing are generally readily apparent on close examination of learning interactions, but often switch implicitly and rapidly, revealing the complexity of this dimension of pedagogic practice (Rose 2020, 2021). Interpersonal enactment of pedagogic activities was initially handled at the level of exchange structure in discourse semantics (Martin 2006; Martin and Rose 2007b). However, the pedagogic functions of exchange roles were found to be more diverse and specific in the context of curriculum genres, distinguished by the roles of teachers and learners in interactions. It is proposed that register-level analysis has greater explanatory potential for interpreting pedagogic relations than exchange structure analysis (Martin and Rose 2007b; Zappavigna and Martin 2018). Particularly interesting are the types of conscious acts that are negotiated in pedagogic relations, and the tacit models of learning as negotiated consciousness that their analysis reveals. More specific features found in systems of pedagogic activities, modalities and relations are given in Rose (2018, 2021). The knowledge dimension of curriculum registers includes commonsense and uncommon-sense fields (Bernstein’s 2000 ‘horizontal and vertical discourses’), configured in knowledge genres. Systems of knowledge genres and their realizations may be made more or less explicit. When made explicit these systems are often referred to as ‘knowledge about language’ (Martin’s 1997 ‘social semiotic instructional discourse’). Curriculum knowledge may also include the activities, modalities and relations of pedagogic registers, which may also be made more or less explicit. Knowledge of a pedagogic register may coincide with Bernstein’s recognition rules, as it enables learners to recognize the structuring and relations of the pedagogic register. Knowledge of curriculum registers may be associated with realization rules, as it enables learners to display this knowledge for successful evaluation. Values in curriculum registers are exchanged together with curriculum knowledge and are associated with curriculum fields and with pedagogic registers. Values in curriculum fields position people, things, texts and activities in hierarchies of moral, aesthetic and affective values, as illustrated for Girbar’s story above. These values are central concerns for critical theories and pedagogies, and are sites of struggle over school curricula. However, values in pedagogic registers may be more significant for identities and outcomes in schools, as they position learners in hierarchies of authority, access, inclusion, success and autonomy. These hierarchies correlate with variables in curriculum genres in the following proportions: curriculum knowledge: authority:: pedagogic modalities: access:: pedagogic relations: rapport:: pedagogic activities: autonomy

114

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Learners’ authority derives from evaluations of their knowledge displays, in classroom interactions and assessment tasks. Their access to knowledge depends on control of pedagogic modalities, particularly reading in formal education. Rapport is the degree to which learners participate actively in activities and interactions, and how they are evaluated (Bernstein’s 2000 ‘hierarchy of success and failure’). Autonomy is an outcome of accumulated evaluations that permits more or less independent practice with more or less surveillance. For successful students autonomy grows year by year towards tertiary study and professional roles, but less so for unsuccessful students. These values are thus as much a part of a curriculum register as the values ascribed to curriculum fields. They are exchanged through the activities, modalities and relations of pedagogic registers. Over time they serve to differentiate types of learner identities, and hence naturalize differences in evaluations and progression. They are as much a part of individuation as the knowledge that comprises learners’ repertoires. This analysis of curriculum genres can be applied to both description of learning discourse and design of lessons. It is recontextualized in the Reading to Learn professional learning programme, to guide teachers to analyse their practices and design effective lessons that embed literacy development in curriculum teaching. By way of illustration, a lesson series was designed with teachers using the Girbar story, for a curriculum task that required Year 4 students (approximately 9 years old) to write a review of this story. A lesson series realizes a complex of curriculum genres. In this instance, the two curriculum genres deployed are known as preparing for reading and joint construction. The staging of these genres is given in Table 6.2. Stage labels are names for types of pedagogic activities. Types of sources are also given, as well as trends in interacts that show increasing autonomy, or ‘handover’, through each stage. The task of reading the story is prepared with an overview of the field that is largely presented by the teacher using pictures, maps and book illustrations. In paragraph-by-paragraph reading, the teacher guides the class to highlight and discuss elements of each paragraph. In note making, students dictate the highlighted wordings as other students scribe on the board in turn. Joint construction is prepared with the teacher deconstructing the target genre, labelling a projected model text, while students label their own copies. The teacher then guides the class in jointly writing a new text, referring to the projected model and the notes, as students propose wordings and scribe on the board and their own workbooks. Joint writing is ideally elaborated with individual writing, in which students either continue the joint text or write a new one in their books, and the teacher gives individual support as needed. In this lesson series, the overview stage begins with an orientation to Saibai culture and places, introducing those elements of the story that would be unfamiliar to most students, including the Kalaw Kawaw words used in the

DESIGNING PEDAGOGIC REGISTERS

115

TABLE 6.2  Lesson Stages in Series Curriculum genres

Prepare

Task

Elaborate

Preparing for reading

Overview

Paragraph-byparagraph reading

Note making

sources:

pictures, map, book illustrations

text copies, highlighting

class board, highlighted copies

interacts:

teacher presents

teacher guides class

students dictate as students scribe

Joint construction

Deconstruct model

Joint writing

Individual writing

sources:

projected model text & student copies

notes & text on board & student workbooks

notes & text in student workbooks

interacts:

teacher presents

teacher guides class

students write, teacher supports

story, which students can practise saying. The story plot is then previewed by talking through the illustrations, reiterating the elements discussed previously, now in context of the story sequence. The task is then to listen as the story is read aloud by the teacher. Because the field has been iteratively previewed before reading, in sequence, all students should be able to follow the story with general comprehension. Key meanings can then be elaborated in discussion. Paragraph-by-paragraph reading is a cyclic activity, in which a paragraph is briefly previewed and read aloud. In this lesson stage, students follow the written text as it is read. It is elaborated by guiding students to identify and highlight key wordings in the paragraph, which are discussed in more depth. The cycle is then repeated for the next paragraph. Note making is a cooperative activity, in which students take turns to scribe on the class board, while other students dictate the highlighted wordings from the reading text. The teacher may elaborate meanings as they go up on the board, and follow the activity by labelling the notes with the phases they instantiate. The notes are dictated from highlighted elements of the story’s register. They consist of nominal and verbal groups and clause fragments, but are primarily selected for the lexical items realizing the field. Activities within each stage of the whole lesson series are summarized in Table 6.3. The sequencing principle for these curriculum genres is to accumulate expectancy through iteration with variation. The cumulative effect is that learners will recognize more of the curriculum register in each stage and phase of the lesson, towards independent reading of source texts and writing of target genres.

116

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 6.3  Activities within Lesson Stages Stages

Prepare

Task

Elaborate

Overview

discuss Saibai culture, preview plot

listen as story is read aloud

review key meanings

Paragraph-byparagraph reading

preview paragraph

follow copy as paragraph is read aloud

mark & discuss key wordings, label phases & themes

Note making

students dictate highlighted wordings

students scribe in turns on board & notebooks

discuss key meanings, label phases & themes

Deconstruct model

preview target genre

read model & discuss stages & phases

label stages & phases

Joint writing

prepare wordings using notes & model

students propose wordings

review, adjust & scribe on board & not books

Individual writing

review model & notes

students write in notebooks

teacher circulates & supports

The pedagogic process of accumulation can be described in terms of text/ context relations, or presence (Martin 2020; Martin and Matruglio 2013; Rose 2020). In pedagogic registers, the matter under focus may be iconic with everyday experience or abstracted from it; interacts vary in degrees of negotiability; sources vary in how implicitly or explicitly they are sourced. Conversely, curriculum registers vary in the condensation of meanings, or mass, including degrees of technicality of knowledge, iconization of values and aggregation of meanings as discourse unfolds. Variables of presence and mass in curriculum genres are summarized in Table 6.4. Each task in a lesson series is first prepared with high presence and low mass. Presence then generally decreases while mass increases in task and elaboration phases. In the overview, the story’s field is presented as iconic with students’ experience, using images that may be sourced implicitly by indicating (verbally or gesturally), and negotiated by asking students to identify them in pictures. As a result, in paragraph-by-paragraph reading, learners’ knowledge has increased so that the story’s exotic field is less remote from their experience. Meanings are now more explicitly sourced in the words of the text, which the class is guided to identify. When the field of each paragraph is well understood, its more abstract role in the series of story phases can be discussed and labelled on the text. By the note-making stage, students are familiar with the text’s

DESIGNING PEDAGOGIC REGISTERS

117

TABLE 6.4  Presence and Mass in Curriculum Genres

Presence in pedagogic register

Field

Tenor

Mode

iconicity of field

negotiability of interacts

implicitness of sourcing

iconization of values

aggregation of meanings

Mass in curriculum technicality of knowledge register

field, and sources of meanings are wordings that they have highlighted in their copies, and have then negotiated into shared notes. On this basis, the abstract field of mythic themes can be elaborated and annotated on the notes. Paragraph-by-paragraph reading is analysed at the tier of learning cycles in Tables 6.5 and 6.6. In this presentation, learning cycles are analysed from three perspectives: (1) their structural phases (prepare, focus, task, evaluate, elaborate), and the matter each phase is concerned with; (2) the sources of meanings in texts or images, or in teachers’ or learners’ knowledge; (3) teacher–learner relations, including their interacts and the acts they negotiate. Interacts are italicized in the transcript and named in the first column. The second column names sources and their sourcing, while the third names the phases of learning cycles and their matter. In Table 6.5, the teacher previews and reads the first problem in the story. The preview locates and rephrases the paragraph, summarizing its events. This summary also explains why it is a problem for Girbar, foreshadowing reasoning for the phasal analysis that will follow. The task is focused with have a look, which directs students’ perception, and the paragraph is read. Students are then explicitly asked to reason about Girbar’s reaction to the problem, and one student proposes the property worried. The teacher affirms and elaborates by recasting the text, adding the property dangerous to swampy country, which explains her hurry. Elaborations may thus make explicit aspects of the field that are left implicit in the text. Most interacts here explicitly negotiate conscious acts of perception and reasoning, with have a look, a problem because, how do you think, that’s why. The reasoning is about expectancy, and the series of tasks here are scaffolded by accumulating expectancy. This begins by naming the phase and explaining its problems for Girbar, which frames the task of receiving its meanings as it is read. Expectancy is from the general register function of a story problem, to the series of particular properties and activities that instantiate the problem. These properties and activities then implicitly expect a certain type of problem, potential danger. The focus question then narrows down the task of inferring Girbar’s reaction, to feelings expected by this type of problem. An appropriate

118

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 6.5  Preview Paragraph

T

Interact & act

Sourcing & source

Phase & matter

The next paragraph is a problem for Girbar, because it was getting dark and she had to cross swampy country to get home. But first she had to climb up the banks of the waterhole.

model reasoning

locate & rephrase text

prepare phase

Have a look while I read it to you.

direct perception

focus activity read text

receive phase

inquire reasoning

elicit student knowledge

focus property

infer student knowledge

propose property

recast text

elaborate property

By the time she had collected the water, the sun had set and it was growing dark. In a hurry to cross the swampy country and reach her home before nightfall, Girbar quickly put the kusul filled with water over her shoulder and turned to climb the banks of the waterhole. T

How do you think Girbar might be feeling?

SS

[hands up]

T

S1

S1

She might be worried?

display reasoning

T

Exactly.

praise

The swampy country might be dangerous in the dark. That’s why she was in a hurry.

model reasoning

answer such as worried is thus highly likely, affording a moment of success and affirmation. On this affective basis, implicit expectancy relations in the text are finally made explicit by the teacher, with the property dangerous as the reason for Girbar’s hurry. The task of responding is shared among students with the ‘hands up’ routine. In the model here, the [hands up] gesture functions to invite evaluation by offering to display knowledge. The teacher permits one student to display by

DESIGNING PEDAGOGIC REGISTERS

119

naming and pointing. These routines are shaded in the tables, as they have no direct function in the pedagogic activity except to distribute display tasks. The inferences of danger and worry are pursued in the next three cycles in Table 6.6, as wordings in the paragraph are identified, highlighted and discussed. In the first cycle, the expectancy range of worry is turned back on the text as students are asked to identify two reasons for it in the first sentence. In the second cycle, the identifying task is now focused with anxious, and located precisely at the beginning of the next sentence. Students are then asked to identify activities associated with hurry, again with a precise location in the next words, to do what? The identified wordings are then evaluated as literature, very descriptive, and for their function in the story, sets the mood of danger. The activity of identifying such wordings for the text response task is then made explicit: So we might use it when we write about the language in the story. This elaboration expects the same activities to be repeated with the following paragraphs. Following paragraph-by-paragraph reading and note making, students are sufficiently familiar with the text’s register as it unfolds through the genre that the review genre can be introduced. A model genre can be deconstructed in two

TABLE 6.6  Elaborating Meanings T

Now let’s highlight some important wordings.

direct activity

In the first sentence, two things make Girbar worried.

model reasoning

Can you see what they are?

inquire reasoning

SS

[hands up]

T

S2

S2

sun had set and it was growing dark

display reasoning

T

That’s right.

approve

Let’s highlight sun had set and growing dark.

direct marking

SS

[highlight]

T

Now at the beginning of the next sentence what tells us she is anxious?

SS

[hands up]

T

S3

prepare activity locate text

prepare wording focus property

read text

propose property

mark text inquire reasoning

locate text

focus property

120

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

S3

She’s in a hurry?

display reasoning

T

Exactly right.

approve

T

In a hurry to do what?

S3 T

SS

read text

identify property

inquire perception

locate text

focus activity

to cross the swampy country and reach her home before nightfall

display perception

read text

identify activity

Yes.

approve

Let’s highlight that whole clause up to the comma.

direct marking

locate text

[highlight]

mark text

model That’s very descriptive and it sets the activity mood of danger. So we might use it when we write about the language in the story.

recast text

elaborate clause

ways. One is to define its stages and phases, put headings on the board, and start writing each phase from the notes and discussion. This method teaches from the system rather than the instance, as in traditional language teaching approaches. It abstracts names of features without an actual text, which learners must imagine. More supportive in terms of presence is to project a model review of a text that students are already familiar with, and borrow its instantial patterns into the new text, using the content of the notes. This way, features can be sourced by pointing to the model text, negotiating their meanings with the class, and labelling on the projection and students’ copies. In individual construction, students may write a new text using the same notes and text structuring, or complete a text started in the joint construction. This may be an individual or group activity in class. The aim is to provide supported writing practice as needed by each student, before attempting the assessment task. The assessment task for this curriculum unit would entail independently writing a review of another traditional story, which may have been read and analysed jointly or independently. The lesson series described here should enable all students to complete this task successfully.

REFLECTION Although professional fields are ostensibly learnt through the written mode, all involve a measure of tacit practice that is learnt ostensively. Education possibly tops the list in this regard. In two decades of training many thousands of practising teachers, I have met very few who considered their pre-service

DESIGNING PEDAGOGIC REGISTERS

121

degree studies to have been a major influence on their classroom practice. Overwhelmingly it is their tacit experience as students in school and then as teachers that has shaped the pedagogies they deploy. From the perspective of genre and register theory, this suggests that the curriculum genres of schooling have largely evolved in the institution, through the tacit practice of generations of teachers and students, rather than by technical design and training, as in other professional regions such as medicine or engineering. Recent pedagogic theories appear to have had minimal effect on this institutional syndrome. Rather than designing wholly new practices, they tend to classify and name existing ones, and advocate some over others. Genre writing pedagogy was a notable break in this pattern. Its technical analysis of target knowledge genres was applied to the design of the deconstruction and joint construction stages of the teaching/learning cycle, which were substantially new pedagogic activities. The pedagogic design was based on broad social psychological principles of explicit demonstration and joint practice, but without the degree of technical analysis and associated metalanguage of knowledge genres. A large measure of the practice and its transmission to teachers remains tacit, relying on shared professional experience rather than explicit detailed analysis. An attempt was made in Reading to Learn to model these pedagogic principles more explicitly, in the structuring of pedagogic activities centred on learning tasks, influenced by scaffolded reading practices (Rose, Gray and Cowey 1999). This enabled design of a large set of curriculum genres for scaffolding reading and writing, including carefully designed teacher/learner interactions (Rose 2020). The more explicit analysis also provided a metalanguage that facilitated training of teachers in these designed practices (Rose 2020). Nevertheless, their design and transmission still rely on tacitly shared recognition of effective practice. Pedagogic register analysis is intended to offer explanatory power. It has been designed to answer pedagogic research questions, starting with the structuring of pedagogic activities at the tier of learning cycles. A series of studies explored the relation between this activity structuring and exchange structures in discourse semantics (Martin 2006; Martin and Rose 2007a, 2007b; Rose and Martin 2012, 2013; Rose 2014, 2018; Kartika-Ningsih and Rose 2018; Kartika-Ningsih 2019; Humphrey, Hao and Rose 2020; Jones, Matruglio and Rose 2021). Exchange analysis describes the structuring of speaker roles and moves, but a functional explanation requires a step up in abstraction, to register functions realized by discourse systems, just as exchange roles explain the functions of grammatical mood selections. However, pedagogic register functions are clearly more than activity structuring. While we could empirically show that teachers prepare, focus and elaborate on tasks in learning cycles, interpreting how they do so remains in the realm of intuitive judgement. Classroom discourse analysis generates a mass

122

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

of data that we found could be sorted into register variables of field, mode and tenor, from which emerged more specific variables in curriculum genres, of pedagogic activities, modalities and relations, alongside curriculum knowledge and values. Systemic research methods were deployed to sort the data into systems of features, related by delicacy and co-selection, that eventually appeared to exhaustively encompass options in pedagogic register, exhibited in a diverse corpus of learning interactions (Rose 2014, 2018). Beyond these overt features of classroom discourse, we need to develop descriptive tools for values in pedagogic registers of authority, access, autonomy and rapport. From the perspective of individual learners these values accrue over time, enabling or constraining the acquisition of knowledge, individuating their repertoires and learner identities. They can also be used to measure the effectiveness of pedagogic practices in widening or narrowing these differences. In the R2L lesson series illustrated above, each curriculum genre begins with the teacher’s authority to prepare the field for reading or the target genre for writing. It is then negotiated with students in the reading and writing tasks, and assumed by students for elaborating these tasks with note making and individual construction. Through each stage, access to meanings is ensured by starting with high presence and low mass, and building expectancy from task to task, so that all students comprehend the matter of each activity. Every student is included in each learning activity through carefully designed teacher–learner interactions to prepare tasks, distributing focus questions around the class, and affirming all responses. Continually preparing, engaging and affirming all students collapses the hierarchy of success and failure, redistributing success to all. Along with authority, autonomy builds through each curriculum genre. In note making it is exercised collectively by the class, and may be exercised by groups in individual construction. The goal is effective autonomy for each student in assessment tasks. I have called this type of practice ‘democratising the classroom’ (Rose 2005, 2006b). It should be possible to design fine-grained criteria for measuring all types of pedagogic practice against these values. At this point, the application of SFL theory in the practice of the pedagogy continues to precede its theorizing. Hopefully this contribution has suggested some interesting opportunities for linguistic and pedagogic research.

NOTES 1 The display in Table 9.1 converts the text into a table, with a paragraph to each row, and columns added to label the phases and themes. This type of display can be used by teachers for preparing lessons, and then for jointly deconstructing texts with their students, by projecting the display onto a whiteboard, and guiding students to label their own copies.

DESIGNING PEDAGOGIC REGISTERS

123

2 The relation between ‘regulative’ and ‘instructional’ registers has been modelled with the ideational metaphor of ‘projection’, on the model of verbal projection in grammar (Christie 2002). Here it is proposed that curriculum registers are exchanged through pedagogic registers, between teachers and learners, a bidirectional interpersonal relation.

REFERENCES Bernstein, B. (2000), Pedagogy, Symbolic Control and Identity: Theory, Research, Critique, Oxford: Rowman and Littlefield. Christie, F. (2002), Classroom Discourse Analysis, London: Continuum. Christie, F. and J. R. Martin (eds) (1997), Genre and Institutions: Social Processes in the Workplace and School, London: Cassell. Humphrey, S., J. Hao and D. Rose (2020), ‘Launching Research: A Martinian Perspective on Science Pedagogy’, in M. Zappavigna and S. Dreyfus (eds), Discourses of Hope and Reconciliation: On JR Martin’s Contribution to Systemic Functional Linguistics, 85–97, London: Bloomsbury. Jones, P., E. Matruglio and D. Rose (2021), ‘Investigating Pedagogic Discourse in Late Primary and Junior Secondary English’, in P. Jones, E. Matruglio and C. EdwardsGroves (eds), Transition and Continuity in School Literacy Development, 145–68, London: Bloomsbury. Kartika-Ningsih, H. (2019), ‘Implementing the Reading to Learn Bilingual Program in Indonesia’, in K. Rajandran and S. A. Manan (eds), Discourses of Southeast Asia: A Social Semiotic Perspective, 145–63, Singapore: Springer. Kartika-Ningsih, H. and D. Rose (2018), ‘Language Shift: Analysing Language Use in Multilingual Classroom Interactions’, Functional Linguistics, 5 (1): 9. https://rdcu. be/be2d9 Martin, J. R. (1992), English Text: System and Structure, Amsterdam: Benjamins. Martin, J. R. (2006), ‘Metadiscourse: Designing Interaction in Genre-Based Literacy Programs’, in R. Whittaker, M. O’Donnell and A. McCabe (eds), Language and Literacy: Functional Approaches, 95–122, London: Continuum. Martin, J. R. (2010), ‘Semantic Variation: Modelling System, Text and Affiliation in Social Semiosis’, in M. Bednarek and J. R. Martin (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity and Affiliation, 1–34, London: Continuum. Martin, J. R. (2020), ‘Revisiting Field: Specialized Knowledge in Secondary School Science and Humanities Discourse’, Onomázein, (1): 111–48. [reprinted in J. R. Martin, K. Maton and Y. Doran (eds), 114–47]. Martin, J. R. and E. Matruglio (2013), ‘Revisiting Mode: Context In/Dependency in Ancient History Classroom Discourse’, in H. Guowen, Y. Zhu, Z. Delu and Y. Xinzhang (eds), Studies in Functional Linguistics and Discourse Analysis, 72–95, Beijing, China: Higher Education Press. Martin, J. R. and D. Rose (2007a), ‘Interacting with Text: The Role of Dialogue in Learning to Read and Write’, Foreign Languages in China, 4 (5): 66–80. Martin, J. R. and D. Rose (2007b), Working with Discourse: Meaning beyond the Clause, 2nd edn, London: Continuum. Martin, J. R. and D. Rose (2008), Genre Relations: Mapping Culture, London: Equinox.

124

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Martin, J. R. and D. Rose (2012), ‘Genres and Texts: Living in the Real World’, Indonesian Journal of SFL, 1 (1): 1–21. Martin, J. R. and R. Veel eds (1998), Reading Science: Critical and Functional Perspectives on Discourses of Science, London: Routledge. Rose, D. (2004), ‘Sequencing and Pacing of the Hidden Curriculum: How Indigenous Children Are Left out of the Chain’, in J. Muller, A. Morais and B. Davies (eds), Reading Bernstein, Researching Bernstein, 91–107, London: RoutledgeFalmer. Rose, D. (2005), ‘Democratising the Classroom: A Literacy Pedagogy for the New Generation’, Journal of Education, 37: 127–64. http://dbnweb2.ukzn.ac.za/joe/ joe_issues.htm Rose, D. (2006a), ‘Reading Genre: A New Wave of Analysis’, Linguistics and the Human Sciences, 2 (2): 185–204. Rose, D. (2006b), ‘Literacy and Equality’, in A. Simpson (ed.), Proceedings of Future Directions in Literacy Conference, University of Sydney, 188–203. Rose, D. (2010), ‘Beating Educational Inequality with an Integrated Reading Pedagogy’, in F. Christie and A. Simpson (eds), Literacy and Social Responsibility: Multiple Perspectives, 101–15, London: Equinox. Rose, D. (2014), ‘Analysing Pedagogic Discourse: An Approach from Genre and Register’, Functional Linguistics, 1: 11. https://doi.org/10.1186/s40554-0140011-4 Rose, D. (2015), ‘New Developments in Genre-Based Literacy Pedagogy’, in C. A. MacArthur, S. Graham and J. Fitzgerald (eds), Handbook of Writing Research, 2nd edn, 227–42, New York: Guildford. Rose, D. (2018), ‘Pedagogic Register Analysis: Mapping Choices in Teaching and Learning’, Functional Linguistics, 5 (3): 1–33. http://rdcu.be/HD9G Rose, D. (2019), ‘The Baboon and the Bee: Exploring Register Patterns across Languages’, in Y. Doran, J. R. Martin and G. Figueredo (eds), Systemic Functional Language Description: Making Meaning Matter, 273–306, London: Routledge. Rose, D. (2020), ‘Building a Pedagogic Metalanguage I: Curriculum Genres and Building a Pedagogic Metalanguage II: Knowledge Genres II’, in J. R. Martin, K. Maton and Y. J. Doran (eds), Accessing Academic Discourse: Systemic Functional Linguistics and Legitimation Code Theory, 236–302, London: Routledge. Rose, D. (2021), ‘Doing Maths: (De)constructing Procedures for Maths Processes’, In K. Maton, J. R. Martin and Y. Doran (eds), Teaching Science: Knowledge, Language, Pedagogy, 257–86, London: Routledge. Rose, D., B. Gray and W. Cowey (1999), ‘Scaffolding Reading and Writing for Indigenous Children in School’, in P. Wignell (ed.), Double Power: English Literacy and Indigenous Education, 23–60, Melbourne: National Language & Literacy Institute of Australia. Rose, D. and J. R. Martin (2012), Learning to Write, Reading to Learn: Genre, Knowledge and Pedagogy in the Sydney School, London: Equinox. Rose, D. and J. R. Martin (2013), ‘Intervening in Contexts of Schooling’, in J. Flowerdew (ed.), Discourse in Context: Contemporary Applied Linguistics, Vol. 3, 447–75, London: Continuum.

DESIGNING PEDAGOGIC REGISTERS

125

Rothery, J. (1994), Exploring Literacy in School English (Write It Right Resources for Literacy and Learning), Sydney: Metropolitan East Disadvantaged Schools Program. Saibai Community (2011), The Story of Girbar: A Story from the Saibai Community, Port Melbourne: Pearson Australia. Zappavigna, M. and J. R. Martin (2018), Discourse and Diversionary Justice: An Analysis of Youth Justice Conferencing, London: Palgrave Macmillan.

126

CHAPTER SEVEN

Developing a Systemic Functional Grammar of Korean as a Resource for Practical Application MIRA KIM

CONTEXT Translation has played a critical role in disseminating knowledge and information in human history. However, it is only in the second half of the twentieth century (e.g. Nida 1964; Catford 1965) that translation started to draw serious attention as a subject of research; and it is only in the last few decades that it began to be recognized as an independent discipline, generally referred to as translation studies – following the proliferation of rigorous research published in books and journals (Munday 2012: 10–11). The late development of research on translation has been a major challenge for professional practitioners and translator educators. For example, when I was a translation and interpreting student in the early 1990s in Korea, we developed our translation and interpreting skills almost exclusively based on repeated practice – without much guidance from established research or theory, even though we were trained in the Postgraduate School of Translation and Interpreting at Hankuk University of Foreign Studies, which was the only educational institute that trained professional translators and interpreters in the

128

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

country. As a consequence, I felt an unquenchable need for a well-grounded theory while I was working as a professional translator and interpreter. This was primarily because I was not able to explain what I was doing in a professional manner. I constantly made linguistic choices when translating and interpreting. But sometimes I was not sure if my choices were the most appropriate; and at other times, I thought they were but could not clearly articulate my reasoning when asked. Laypersons’ explanations such as ‘it feels right’ or ‘it sounds awkward’ were the most common ‘justifications’ I could use, as that was the type of feedback we received from our teachers and peers when I was a trainee. While working as a professional translator and interpreter for several years, I could not help asking myself questions about various aspects of translation and constantly felt them creeping back up to me again and again because I could not find any answers. In 2000, I quit my job and moved to Sydney to find answers to these questions. As I did not have any theoretical framework on which I could base my research, I started a master’s programme of applied linguistics at Macquarie University, where I started to teach translation practice courses at the same time. During that time, I was introduced to Systemic Functional Linguistics (SFL). The more I learned about it, the more I was fascinated by it. What appealed to me the most was that SFL sees language as a meaning-making resource and provides a useful set of tools to analyse different kinds of meaning. That motivated me to embark on my first research on a metafunctional analysis of translation errors in my students’ translations of a short (250 words) English source text (Kim 2003). The project started with a high degree of uncertainty because I was not sure if it would be feasible to do it without having any description of Korean from a Systemic Functional Linguistic perspective to base it on. In addition, I was not sure about my capacity to use the theory, which was new knowledge that I was still wrestling with. The study, however, proved it was possible to classify the majority of translation errors metafunctionally and suggested to me that this approach could be used to provide systematic feedback on individual translations in order to empower translators-to-be to think critically about their translation choices. The findings of the study were presented at the first International Association of Translation and Intercultural Studies conference in 2014, which was the first international conference that I attended as a researcher (rather than as an interpreter); the paper was accepted for publication (Kim 2007b) and was reprinted (Kim 2010). I felt very encouraged that my research was recognized by peers in a field where SFL was not well known and has been even denigrated as a theory that is too complex to apply to translation (e.g. Fish 1981: 59–64, cited in Munday 2012: 153). However, what excited me the most was that I found myself explaining different choices of translation in a more systematic way and witnessed a significant improvement in my students’ learning. The tremendous

DEVELOPING AN SFG OF KOREAN

129

joy from feeling empowered to help students be empowered was the main driver for my insistent application of an SFL-based meaning-oriented approach in my English into Korean translation classes. Such a high intensity of joy and satisfaction as a teacher resulted from the fact that I had not been equipped previously with any systematic tools to use. This was not an issue unique to myself, as translation assessment was one of those areas that had not been rigorously researched despite its being a fundamental issue: The area of translation assessment has been under-researched (Cao 1996: 525; Hatim and Mason 1997: 197) and regarded as a problematic area (Bassnett-McGuire 1991; Malmkjaer 1998; Snell-Hornby 1992) primarily due to ‘its subjective nature’ (Bowker 2000: 183). As a consequence, there appears to be a lack of systematic criteria that can be used universally to assess translations (Bassnett-McGuire 1997; Hönig 1998; Sager 1989). This presents an enormous challenge to translation teachers, who need to assess students’ translations for both formative and summative purposes, and provide constructive, detailed feedback on their translations. (Kim 2009: 123) The SFL-based meaning-oriented approach proved, based on both quantitative and qualitative data, to be an efficient formative assessment tool in translator education (Kim 2009). The qualitative data collected from students’ learning journals showed that students were willing to discuss and justify their translation choices. More and more students became aware of different ways of translating at word, group and phrase, clause, clause complex, paragraph and even text levels; and they confidently tried out options and made informed choices without fear of what the teacher would say. Students’ independent learning was enhanced as they learned how to analyse their own translation errors and their translation skills improved substantially over just one semester. Such substantive changes, which are usually hard to observe in translation classes, were possible because they learned how to analyse different translation choices at different ‘ranks’ and understand their potential impact on different aspects of meaning (in context). This kind of learning experience resulted in direct benefits for students. For example, in a translation course that prepared students for the Australian National Authority of Accreditation for Translators and Interpreters (NAATI) translation exam, students in the Korean stream showed a significant increase of the NAATI exam pass ratio from around 10 per cent in 2004 to over 60 per cent in 2007. When the ratios were examined within the stream, it was clear that the ratio was always higher in the semesters when the meaning-oriented approach was used than in the two semesters when it was not (Semesters 1 and 2, 2004 and Semester 1, 2006). This was unprecedented, as no other language

130

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 7.1  NAATI translation pass ratios. Adapted from Kim (2009: 151).

streams showed such a trend. For example, the largest language group, with 50–80 students, showed very stable pass ratios between 30 and 40 per cent, as shown in Figure 7.1 (Kim 2009: 150–1). In addition, a number of students in the Korean stream developed an interest in research. I supervised five master’s dissertations in 2009 alone and one of them went on to complete her PhD in 2013. This is very uncommon as students generally do not want to write a thesis as part of a practice-oriented master’s programme of translation and interpreting.

MOVING FROM TRANSLATOR EDUCATION TO KOREAN LANGUAGE DESCRIPTION Even though the SFL-based meaning-oriented assessment proved itself to be a very effective tool for translation teaching in my own context, there were translation choices that remained controversial due to the lack of an SFL description of Korean. One of them has to do with theme choices. It seems to be widely accepted that 은/는un/nun is a theme or topic marker in Korean (cf. Li and Thomson 1976; Lim 1972) and therefore what the clause is going to be about is indicated by the element marked by the particle 은/는 un/nun. However, this supposition applies only to a limited extent because there are other elements that appear to play a similar thematic role, but are not marked by 은/는un/nun. In addition it is not uncommon to observe clauses without any nominal group marked by 은/는un/nun or 이/가i/ga, which is traditionally

DEVELOPING AN SFG OF KOREAN

131

known as a subject marker in Korean. Another issue was that it was not always straightforward to analyse whether or not a circumstance has any textual role when a participant is elided in the clause, which is highly frequent. Such issues were always heatedly discussed amongst students, most of whom were native speakers of Korean and tended to have strong opinions based on their intuitions but not backed by hard evidence. This practical dilemma motivated me to carry out my first linguistic descriptive work on Korean as part of my PhD. I compiled a corpus of 17 short Korean written texts spanning three text types (expounding, reporting and recreating) and manually analysed 537 clauses in order to understand what the point of departure is in Korean and how it works textually to orient the reader as the text unfolds. The study discovered some commonalities between the theme systems of English and Korean, as well as unique features of theme in Korean. The common features include the fact that both languages can have multiple themes (i.e. topical theme plus interpersonal and/or textual themes), and that what comes at the clause-initial position has textual significance, which was not recognized in the literature at all. The two languages, however, differ in that clause-initial placement is not the sole indicator for themes in Korean. An unmarked theme is realized by a nominal group, prototypically marked by un/nun when information assigned to the theme is predictable within context and is alternatively marked by i/ga when it is newsworthy. And a prototypical unmarked theme tends to be elided when it is recoverable – either because it has been introduced in the preceding discourse or is obvious within the context. The newsworthy unmarked theme can be also elided in similar circumstances in spoken Korean (Shin and Kim 2008). Therefore, it is vitally important to analyse Korean themes with due consideration for thematic progression in a discourse – taking both co-text (cohesion) and context (register and genre) into account. My descriptive work was just a beginning, but it did inspire two PhD projects. Choi, who is also a professional translator and interpreter, undertook a PhD focusing on logical meaning in Korean (Choi 2013); and Park (2013) explored clause rank experiential grammar, studying transitivity in 1,400 clauses from 25 short authentic Korean texts.

COLLABORATIVE WORK TO WRITE AN APPLIABLE GRAMMAR Once these three PhD studies were completed, I felt strongly motivated to write a full systemic functional description of Korean. My colleague Shin was keen to explore the interpersonal grammar of Korean, drawing on his previous studies which were related to this area. I also felt the need to collaborate with an experienced systemic functional linguist with experience in language

132

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

description in relation to theoretical and practical issues. J. R. Martin was the obvious choice as he and I had been running a functional language typological seminar for our PhD students since 2008. We drew specifically on the principles of the linguistic theory that were devised by M. A. K. Halliday and his colleagues in the 1950s and 1960s, as inspired by earlier work by Firth and Hjelmslev in this initial conceptual period (see Martin 2016 for a short history of SFL). Our grammar writing project makes a number of distinctive contributions to Korean linguistics and applied fields such as translation. It also provides an exemplar for any future typological work describing a language using SFL as the informing theory and resisting using English as the basis of descriptions. To begin, whereas existing Korean grammars have been largely concerned with relations among elements within a clause (i.e. syntagmatic structure), our Systemic Functional Grammar of Korean also focuses on the relation of alternative grammatical elements to each other (i.e. paradigmatic relations) – describing how meaning changes when one choice is made rather than another. In other words, this grammar is concerned with the way in which Korean grammar makes meaning. Our Korean grammar is the first of its kind and has been able to shed new light on a number of issues that have been regarded as headaches in Korean linguistics for a long time. One of them is the so-called ‘particles’ of Korean. These have been studied individually or in groups but have never been described from a holistic perspective. Investigating them within texts, from a functional perspective, we were able to categorize them according to their primary functions and treat them as function-marking post-positions. They are divided into three types: experiential function marking (EFM), textual function marking (TFM) and interpersonal function marking (IFM). Each function is realized by a post-positional particle that helps distinguish the role the nominal group plays in clause rank (experientially, interpersonally and textually). This enables us to present Korean examples in a more meaningful way. For instance, examples (1), (2) and (3) are identical except that each contains a different particle. The EFM, 이, i occurring after, 선녀 셋, seonnyeo ses, ‘three nymphs’ in example (1) tells us who descended; the TFM, 은, eun in example (2) flags the orientation to the field by thematizing three nymphs; and the IFM, 만, man in example (3) indicates the speaker’s unsatisfied expectation about the number of nymphs descending. With this information, one can understand how these three examples are different and translate example (3) differently from (1) and (2). The subtle difference between (1) and (2), which has been explained above, cannot be translated in English.

DEVELOPING AN SFG OF KOREAN

133

TABLE 7.1  Example 1: Experiential Function Marking 선녀 셋이 내려왔다 seonnyeo

ses

nymph

three

i

naeryeow-at-da descended

P1

Process

nominal group

verbal group

Thing

Quantity

EFM

noun

numeral

particle

Event

‘Three nymphs descended’

TABLE 7.2  Example 2: Textual Function Marking 선녀 셋은 내려왔다 seonnyeo

ses

nymph

three

eun

naeryeow-at-da descended

P1

Process

nominal group

verbal group

Thing

Quantity

TFM

noun

numeral

particle

Event

‘Three nymphs descended’

TABLE 7.3  Example 3: Interpersonal Function Marking 선녀 셋만 내려왔다 Seonnyeo

sen

man

naeryeow-at-da

Nymph

three

only

descended

P1

Process

nominal group

verbal group

Thing

Quantity

IFM

Noun

numeral

particle

‘Only three nymphs descended’

Event

134

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Our work to date sorts the particles that indicate nominal groups’ clause rank functions, as presented in Table 7.4. The choices for function marking are presented in Figure 7.2. Second, our grammar is based on Korean texts which were collected as the primary data for three PhD theses (Kim 2007a; Park 2013; Choi 2013) and a new spoken corpus used for Shin (2018). The corpora are limited in terms of size. But they have been selected to provide a representative sample of Korean texts across a range of registers and genres including recounts, news stories, narratives, reports, descriptions, explanations, arguments and chat.1 As such TABLE 7.4  Function Markers in Korean Function marker

Examples

Experiential

Participant 1

이/가 i/ga 께서 kkeseo (honorific) 에서 eseo

Participant 2

을/를/ㄹ eul/reul/l

Participant 3

에게 ege 한테 hante 께 kke (honorific)

Circumstances

에 e ‘on, at, in, to’ 더러 deoreo ‘to’ 에서 eseo ‘in, at, on, from’ 에게서 egeseo ‘from’ 한테서 hanteseo ‘from’ 으로/로 ro/euro ‘with’ 에로 ero ‘toward’ 에게로egero ‘toward’ 한테로 hantero ‘toward’ 으로써/로써 rosseo/eurosseo ‘with’ 로서/으로서 roseo/euroseo ‘as’ 보다 boda ‘than’

Interpersonal

도 do ‘also, even’ 만 man ‘only’ 조차 jocha ‘even, as well’ 부터 buteo ‘from’ 까지 kkaji ‘even’ 마저 majeo ‘even’ 이나마/나마 eunama/nama ‘in spite of’ 이나/나 ina/na ‘as many (much) as’ 이라도/라도 irado/rado ‘even’ 이야/야 iya/ya ‘surely’

Textual

은/는 eun/neun ‘as for’ 이란/란 iran/ran ‘as for’

DEVELOPING AN SFG OF KOREAN

135

FIGURE 7.2  The system of function marking.

they have allowed us to explore a comprehensive range of Korean grammatical resources across various text types. Third, our grammar interprets lexicogrammatical features in a way that is sensitive to both discourse semantics (co-text) and context (register and genre). In other words, it helps us see how Korean texts are organized to do what Korean speakers and writers need them to do across a range of communicative tasks. Throughout the process of developing the grammar, we have constantly moved back and forth between our provisional descriptions and discourse to make sure our descriptions do not collapse into an armchair grammar which is hard to apply to the analysis of authentic texts. This is an essential and non-negotiable feature of the research precisely because the project was initiated for practical purposes, as mentioned at the outset of this chapter. Alongside these contributions, our grammar addresses dimensions of Korean that have not been thoroughly explored before (e.g. the structure of the verbal group, structures involving bound nouns) and engages productively with the reasoning SFL deploys to ground paradigmatic relations in syntagmatic ones. This puts us in a stronger position to interpret Korean from a functional perspective as far as typological variation across languages is concerned. Let us take structures involving bound nouns as an example. A bound noun is a short word that behaves like a noun in the sense that it can be immediately followed by a particle; it cannot be used alone but always needs to come with a linker (e.g. -ㄹ-l) when realizing a particular function at the group or clause level. That is why it is called a ‘bound’ noun: it is bound to something else. This interesting element raised challenges in several aspects of our grammar. It

136

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

became important: (a) to clarify if it was operating as part of groups, clauses or clause complexes; (b) to come up with different function labels when it realizes different functions and (c) to analyse its structure at different levels. Our current work shows that bound nouns operate at all three levels (i.e. group, clause and clause complex), involving distinct structural analysis – as introduced below. Some bound nouns such as 앞 ap ‘front’ and 뒤 dwi ‘behind’ function as perspective within nominal groups that realize a circumstance of location in space or time in a clause. This function is used to specify the locative meaning of the culminative EFM particle deployed in these nominal groups. For example, the particles 에 e, or 에서 eseo can both be glossed as ‘in, at, on’ etc.; they are not specific enough to distinguish among different types of location. A perspective function is deployed to make the finer distinctions. This function is realized in the penultimate position in nominal groups through a bound noun – following a noun that realizes the thing, as exemplified in (4). Some bound nouns such as 수 su, and 지 ji, on the other hand, realize the modal function in verbal groups. For example, the probability modal function TABLE 7.5  Example 4: Perspective Function 방 앞에 bang

ap

e

room

front

in

Thing

Perspective

EFM

noun

bound noun

particle

‘in front of the room’

TABLE 7.6  Example 5: Modal Function 죽을 수 있다 jug

–eul

su

die

could

Event

Modal

Verb

word complex

Head

Link

stem

suffix

‘(You) could die (if you do … ).’

bound noun

it

–da

auxiliary verb Head

Exchange Mark

stem

suffix

DEVELOPING AN SFG OF KOREAN

137

is realized at group rank through word complexes, such as (으)ㄹ 수 있다 -(eu)l su it-da ‘could’ or (으)ㄹ 수 없다 -(eu)l su eup-da ‘couldn’t’. This modal function is exemplified in (5). Other bound nouns such as 것, geot, 줄, jul and 지 ji operate at the clause level to realize the installer function, as shown in (6). An important point here is that the verb preceding the bound noun culminates with one of the suffixes, -(으)ㄴ -(eu)n (‘realis’ complete), -는 neun (‘realis’ incomplete) or -(으)ㄹ -(eu)l (‘irrealis’). These suffixes link the down-ranked clause to the bound noun, which in turn positions the embedded clause as the phenomenon in the mental clause. Other bound nouns such as 뒤 dwi, 후 whu, 때문 ttaemum are involved in clause complexing. This is illustrated in (7), where the subjoiner function specifies the relation of a hypotactic clause to the clause it depends on. TABLE 7.7  Example 6: Installer Function 나는 [[사람이 떨어지는 것을]] 보았다. na

Neun [[saram i

tteoreoji-neun

geos

eul]]

be falling

bo-at-da

I

person

saw

P1: Senser

P2: Phenomenon(act)

Process

Ng

[[clause]]

vg

P1: Actor

Process

Installer

EFM

nominal group

verbal group

bound noun

ptcl

Thing

EFM

Event

noun

ptcl

verb Head

Link

stem

suffix

‘I saw a person falling.’

TABLE 7.8  Example 7: Subjoiner Function 지현이가 밥을 한 뒤 Xβ

Jihyeoni Jihyun

ga

bap

eul

ha-n

dwi

ricecooking

do

after

P2

Process

Subjoiner

Xβ P1

138

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

지현이가 밥을 한 뒤 Ng

ng

vg

Thing

EFM

Thing

EFM

Event

Noun

particle

noun

particle

verb

bound noun

Head

Linker

stem

suffix

‘After Jihyun cooked rice, Jihye made soup.’ 지혜가 국을 끓였다. α

Jihye

ga

jihye

Guk

eul

kkeury-eot-da.

soup

cook

P1

P2

Process

Ng

Ng

vg

Α

Thing

EFM

Thing

EFM

Event

noun

particle

noun

particle

verb Head

TM

EM

stem

suffix

suffix

… Jihye made soup.’

In (7), the suffix -n links the clause to the bound noun 뒤 dwi. The bound noun carries some residual lexical meaning (enough to distinguish types of hypotactic enhancement) – mainly having to do with space, time and manner. The bound noun construction is frequently used for certain subcategories of expansion, most commonly for ‘hypotactic enhancement: temporal: different time’. Also, the bound noun construction can culminate with the particle 에, e ‘at, on, in’, providing circumstance function marking.

CONCLUDING REMARKS In concluding this chapter, I would argue that our functional grammar of Korean will serve as an empowering research tool for translation studies in general and translation studies that involve Korean in particular. SFL has been increasingly adopted as a major theoretical framework in translation studies (e.g. Kim et al. 2021) and one of the primary reasons is that SFL provides analytical

DEVELOPING AN SFG OF KOREAN

139

tools with which researchers can interpret meaning from a lexicogrammatical perspective. This is possible because languages are described as a system of choices from a paradigmatic point of view in SFL. This makes SFL a powerful tool for researchers in translation studies, where they can explore translation choices systemically and systematically, instead of relying on personal opinions or preferences. A special issue of the Journal of Translation Studies is an excellent demonstration of how we can use SFL to understand available translation choices and how they impact on meaning. All the papers in the issue employ specific analytical tools suggested to investigate for specific modes of meaning in SFL, analyse data to understand variation in translation and interpret the variation in relation to different modes of meaning. For example, Jing and White (2016) explore the mode of interpersonal meaning expressed through interjections in audiovisual translation. These advances are substantial in translation studies because the arguments put forward are not grounded on arbitrary reasoning but on data-driven findings. In addition, the findings can be validated as the studies can be repeated with expanded data or different sets of data using the same methods clearly explained within the SFL framework. Even though it is always possible that findings may vary slightly depending on how ambiguous methodological issues have been dealt with in each study, they should be comparable as long as each study provides a clear description of how data have been analysed (Kim 2016). In this chapter, I have discussed the motivation for the Systemic Functional Grammar of Korean we are developing, how this grammar has been developed and the contributions it is making to Korean linguistics, and to SFL and its application in translation studies. It is hoped that this discussion inspires many more meaningful collaborations between SFL and the professions, since the dialectic of theory and practice involved is so mutually beneficial.

NOTE 1 The genre categories used in the grammar are taken from Martin and Rose (2008) and Eggins and Slade (1997).

REFERENCES Catford, J. C. (1965), A Linguistic Theory of Translation, London: Oxford University Press. Choi, G. (2013), ‘A Study on Logical Meaning Using SFL and the Implications of This for Translation Studies’, PhD dissertation, Faculty of Arts and Social Sciences, University of New South Wales. Eggins, S. and D. Slade (1997), Analysing Casual Conversation, London: Cassell.

140

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Fish, S. (1981), ‘What Is Stylistics and Why Are They Saying Such Terrible Things about It?’, in D. Freeman (ed.), Essays in Modern Stylistics, 53–78, London: Methuen. Jing, Y. and P. R. R. White (2016), ‘Why Audiovisual Translators Downplay the Interpersonal: The Case of “Interjections” in English-to-Chinese Movie Subtitling’, Journal of Translation Studies, 17 (4): 107–42. Kim, M. (2003), ‘Analysis of Translation Errors Based on Systemic Functional Grammar: Application of Text Analysis in English/Korean Translation Pedagogy’, minor dissertation, Linguistics Department, Macquarie University. Kim, M. (2007a), ‘A Discourse Based Study on Theme in Korean and Textual Meaning in Translation’, PhD dissertation, Linguistics Department, Macquarie University. Kim, M. (2007b), ‘Translation Error Analysis: A Systemic Functional Grammar Approach’, in D. Kenny and K. Ryou (eds), Across Boundaries: International Perspectives on Translation Studies, 161–75, Newcastle upon Tyne: Cambridge Scholars Publishing. Kim, M. (2009), ‘Meaning-Oriented Assessment of Translations: SFL and Its Application to Formative Assessment’, in C. Angelelli and H. Jacobson (eds), Testing and Assessment in Translation and Interpreting, 123–57, Amsterdam: John Benjamins. Kim, M. (2010), ‘Translation Error Analysis: A Systemic Functional Grammar Approach (Revised)’, in C. Coffin (ed.), Applied Linguistics Methods: A Reader, 84–94, London: Routledge. Kim, M. (2016), ‘SFL, an Empowering Tool for Translation Studies’, Journal of Translation Studies, 17 (4): 5–10. Kim, M., J. Munday, P. Wang and Z. Wang (2021), Systemic Functional Linguistics in Translation Studies, London: Bloomsbury. Li, C. N. and S. A. Thompson (1976), ‘Subject and Topic: A New Typology of Language’, in C. N. Li (ed.), Subject and Topic, 457–89, New York and London: Academic Press. Lim, H-B (1972) ‘Kwukeuy cwucehwa yenkwu [The study of topicalization in Korean]’, Kwuke Yenkwu [Korean Research], 28. Martin, J. R. (2016), ‘Meaning Matters: A Short History of Systemic Functional Linguistics’, Word, 62 (1): 35–58. Martin, J. R. and D. Rose (2008), Genre Relations: Mapping Culture, London: Equinox. Munday, J. (2012), Introducing Translation Studies, 3rd edn, London: Routledge. Nida, E. A. (1964), Toward a Science of Translating, Leiden: E. J. Brill. Park, K. (2013), ‘The Experiential Grammar of Korean’, PhD dissertation, Linguistics Department, Macquarie University. Shin, G. (2018), ‘Interpersonal Grammar of Korean: A Systemic Functional Linguistics Perspective’, Functions of Language, 25 (1): 20–53. Shin, G. and M. Kim (2008), ‘A Systemic Functional Analysis of Topic NPs in Korean’, in C. Wu, C. Matthiessen and M. Herke (eds), Proceedings of 35th International Systemic Functional Congress, 213–18, Sydney: Macquarie University.

CHAPTER EIGHT

The Reading to Learn Bilingual Programme Taking Multilingualism into Genre-based Pedagogic Practices HARNI KARTIKA NINGSIH

PRELUDE: CONTEXTUAL BACKGROUND Sydney School genre pedagogy, known locally as the genre-based approach or GBA, has been adopted in the Indonesian national curriculum for the subject English since 2004. Its adoption was partly influenced by a new perspective on the role of English in Indonesia, i.e. being the language of access and participation in global affairs (Musthafa and Hamied 2014; Suherdi 2012). To achieve this demand, subject English teaching needed to be oriented towards literacy education emphasizing the success of high-stakes reading and writing (Emilia 2010). In addition to that, Sydney School pedagogy was chosen because of its successful implementation in Australia, and its principled way of thinking about English language and literacy that could inform the teaching of English in Indonesia. However, years after its implementation, debates and discussions related to aspects of multilingualism in relation to the curriculum have not been resolved. The first issue involves the ongoing debate about the use of the L1 and L2 on the one hand, versus L2 only on the other. Though it is often suggested to use only English during teaching, code switching inevitably occurs. In everyday

142

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

life in Indonesia, at least Bahasa Indonesia as the lingua franca and a regional language are used involving mixing and switching (see Errington 1998; Goebel 2010), and this practice of mixing and switching also takes place in English classrooms. The second issue was to do with the teaching of subjects in the L2. Bahasa Indonesia is predominantly the language of education and is used for the students to learn subject areas. Bahasa Indonesia in these subject areas uses specialized discourse, or vertical discourse, to use a term by Bernstein (1999, 2000), which is different from the everyday and common-sense registers of Bahasa Indonesia that both the teachers and the students use in their daily life. Teaching and learning subject disciplines in Bahasa Indonesia are a challenge, let alone teaching those same subjects in English when it comes to teaching high-stakes literacy. This important challenge led to a pilot project on a Reading to Learn bilingual programme, designed to address the needs of high-stakes literacy involving aspects of multilingualism (Kartika-Ningsih 2016, 2020). Reading to Learn (R2L) (Rose 2016) is the latest generation of genre pedagogy reported for its success across the globe (Rose and Martin 2012). At the beginning of the project, questions arose related to the existing practice, motivating the design of the project. In a multilingual nation where a national language serves as a lingua franca, including in education, how can EFL play a part in advancing students’ literacy competence when a national language is already used? R2L methodology, mainly developed in monolingual classrooms, indicates great potential to be extended in multilingual classrooms. But in environments where languages are in contact, is there a way of building multilingualism into the pedagogy, and developing systematic use of L1 and L2? The project was carried out in two stages – the descriptive stage and the intervention stage – in order to address several questions. The descriptive stage was done to gain understanding of how Sydney School genre pedagogy has been adopted in Indonesia. The results informed the planning in terms of the ways in which multilingualism had been dealt with – what aspects had been addressed, what was lacking and how it could be improved. The intervention stage consisted of the programming of the pedagogy and the implementation. The programming included reframing various bilingual programmes, discussing the development of Sydney School pedagogy and designing the R2L bilingual programme. The implementation stage involved enacting the programme in classrooms.

SYDNEY SCHOOL GENRE PEDAGOGY IN INDONESIA The descriptive stage examined how the Sydney School had been adopted in Indonesian classrooms. It observed two classes of Year 7 (13–15 years old) in junior high schools, and each was taught using genre pedagogy focusing

THE READING TO LEARN BILINGUAL PROGRAMME

143

on English integrated with science. The study indicated the differences between the previous pedagogical approaches and the GBA, including the focus on teaching about language based on traditional grammar, and a strong influence of EFL teaching methods found throughout the teaching/ learning cycle. In general, the teaching/learning cycle of the GBA consists of four stages (Emilia 2011), taken from the Language and Social Power Project (Murray and Zammit 1992) with variant names for some stages: • negotiating field is known as building knowledge of the field (BKOF) • deconstruction is known as modelling of the text (MOT or simply modelling) • joint construction is known as joint construction of the text (JCOT) • independent construction is known as independent construction of the text. In practice, typically the broad outline of the teaching/learning cycle is that BKOF, which entails preparatory activities for reading and writing, is followed by modelling of relevant texts, which is then followed by joint construction of text, which is followed by independent construction of texts. The findings indicated a partial adoption of genre pedagogy, a strong influence of past EFL methods and little attention to multilingual aspects. The GBA approach was used in a rather eclectic fashion, including aspects of traditional grammar and translation teaching method, as well as aspects of audio-lingual approaches. In BKOF, knowledge was built around activities such as labelling ‘language features’, or translation of new words found in audio or reading materials. In MOT, the stages of a genre and its typical grammatical patterns were explicitly shown, usually followed by an exercise of rearranging jumbled paragraphs, or a quiz to test the students’ knowledge about the stages of the genre and grammar. In JCOT, though recommended as the core activity which includes teacher–student collaboration (Emilia 2011), the practice observed was a result of reinterpreting the original joint construction either by adding different activities while maintaining a comparable guiding role, or by replacing it with group activities. The former reflects the practice as suggested, while the latter involves students working in groups with the teacher providing occasional support. The teaching emphasized language in a traditional sense with very little discussion related to relevant scientific fields, though the teaching unit was claimed to be integrated with science. Throughout the teaching, Bahasa Indonesia was used interchangeably with English, usually to refer to knowledge about language (procedure genre, text types, goal etc.), and translation of terms to Bahasa Indonesia and vice versa. Sundanese was also

144

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

used to a limited degree by teachers for commands, and for humour among students. GBA was considered successful in that it pushed the EFL teaching towards a more literacy oriented and written, text-based approach; and away from an approach that focused on teaching/learning at a sentence level, and on oral practice. However, as discussed, the teaching of field and aspects of multilingual classrooms were not being taken into consideration. It was thus timely to develop a bilingual teaching programme which deliberately and strategically involved the use of L1 and, at the same time, involved the teaching of relevant fields of the subject being taught. This chapter now turns to the existing practices of bilingual programmes across the world in order to contextualize the intervention in terms of relevant theoretical influences and contemporary practices of bilingual pedagogic practices, and revisits the development of Sydney School genre pedagogy to likewise contextualize the development of bilingual pedagogic practices in the programme.

INVOLVING MULTILINGUALISM, EXTENDING R2L This section focuses on re-examining various bilingual education programmes and discussing the development of Sydney School genre pedagogy. Bilingual education programmes are considered based on the use of language and the focus of the teaching. The discussion of the Sydney School considers the origin of the pedagogical frameworks to develop the bilingual programme. Re-examination of bilingual education programmes Characterization of bilingual education has often been to do with educational programmes which simply involve two or more languages (e.g. Hamers and Blanc 2000) and those programmes aiming at balanced language and literacy in two languages (e.g. García 2009). Since such characterizations offer little precision specifying the role of bilingualism in teaching/learning, the examination reported on here turned to Halliday’s (1993) threefold perspective on language learning (learning language, learning through language and learning about language) and Bernstein’s (1975) notion of classification. This resulted in a topology of bilingual pedagogy programmes with two clines – the first cline is the use of languages in the teaching or the medium of instruction, and the second cline is the focus of the teaching. The use of languages in bilingual education programmes varies from one programme to another, though it tends to polarize between L1 and L2 use, allowing code switching to be part of the teacher’s talk, or making L2-only

THE READING TO LEARN BILINGUAL PROGRAMME

145

compulsory. Several programmes involve and/or allow the use of L1 during teaching/learning. For example, the grammar-translation method encourages the use of L1 for teaching L2, but communicative language teaching tends to encourage maximal ‘L2 exposure’ (e.g. Spada 2007). The variations in language use in classrooms among programmes are best seen as functioning on a cline. At one end of the cline, termed as ‘enfolding’, L1 is used whenever and for as long as necessary in the move towards L2 proficiency. At the other end of the cline, termed as ‘enveloping’, L2 is the only medium of instruction, excluding the use of L1. The focus of L2 learning in relation to the content of the curriculum can be considered by using Bernstein’s (1975) terms isolated or integrated. Where learning L2 is isolated, L2 is positioned as what is being learned – students learn L2 as a language subject. EFL teaching methods typically focus on teaching this type of content, emphasizing knowledge about language. In the integration

FIGURE 8.1  Topology of bilingual education programmes. Source: Kartika-Ningsih and Rose (2018).

146

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

type, L2 is treated as part of content-focused teaching/learning. Content and language integrated learning (e.g. Llinares, Morton and Whittaker 2012) and immersion programmes (e.g. Genesee and Lindholm-Leary 2013) are a few examples of L2 teaching through other subjects. This contrast sets up another cline with ‘language-focused’ at one end of the cline, where L2 learning is isolated from other subject areas, and ‘content-focused’ at the other end of the cline, since L2 learning is integrated with other subjects. This topology is not meant to be simply read as each quadrant being assigned to a particular language programme. In fact, in an ideal sense it suggests a bilingual teaching programme should be able to guide students to move from enfolding to enveloping, moving up and down the language-focused and content-focused poles. In other words, successful bilingual pedagogy involves students in learning language – i.e. learning to get full control of the L2; learning through language – i.e. learning to access non-language content or ‘field’ in the L2; and learning about language – i.e. learning how the L2 works as a system. After the focus on understanding bilingualism in language programmes, the investigation turned to the pedagogic principles and practices of Sydney School genre pedagogy. As it has been recontextualized in Indonesian classrooms, it is a good chance to revisit the actual frameworks as they informed the design of the bilingual pedagogy programme. The development of the Sydney School Sydney School genre pedagogy offers considerable potential for developing bilingual pedagogy programmes for several reasons. First, the pedagogy is based on a language-based theory of learning, making explicit the relationship between text and the social context and developing explicit teaching procedures for application. Second, a number of successful examples of implementation have served as evidence of the transferability of the pedagogy to different levels of education and language settings across the world (e.g. Brisk 2015; Schleppegrell et al. 2014). Lastly, in relation to multilingualism, the pedagogy has been developed for ESL education, many concerning students from migrant and Indigenous backgrounds in disadvantaged schools in New South Wales, Australia (see Disadvantaged Schools Program 1988). The pedagogy is based on a model of language used in different social contexts (Martin 1992; Martin and Rose 2007, 2008) – or text in context (de Silva Joyce and Feez 2012). The text-in-context model makes explicit the relationship between text and social context. A text consists of three levels of articulation: discourse semantics, lexicogrammar and expression. Discourse semantics deals with patterns of meaning across the whole text; lexicogrammar focuses on patterns of meaning within clauses; and expression attends to patterns of letters, and/or sounds, and/or signs in sign languages. But a text cannot be a text without its social context. The social context embodying

THE READING TO LEARN BILINGUAL PROGRAMME

147

text is two-layered, i.e. register and genre (Martin 1992; Martin and Rose 2008). ‘Register’ is the cover term for the tenor (of social relations), the field (of experience) and mode (of communication); and genre refers to the social purpose of a text, as a ‘staged, goal-oriented, social process’ (Rose and Martin 2012: 54). The relationship between text and social context can be seen through the ways in which each text attempts to achieve a purpose addressing a particular audience (e.g. scientists, children, general public), managing different modalities of communication (spoken or written, visual or verbal etc.), and focusing on particular institutions (school, science etc.). This concept allows for the mapping of genres, or text types (Martin and Rose 2008; Rose and Martin 2012), and for educators to develop teaching procedures designed to control these genres. The teaching procedures are known as genre pedagogy, which have been used in the Indonesian curriculum. They are represented in a teaching–learning cycle, which has been developed through intensive work with teachers in actual classroom situations and on observations of the impact of the pedagogy. Since early 1980, the teaching–learning models have evolved into several teaching– learning cycles, developed as principles of purposeful teaching–learning activities. As discussed above, Indonesian GBA adopted an earlier generation of the teaching–learning cycle from the Language and Social Power Project. The reason for this choice was that the BKOF stage is considered as a scaffolding stage, important for building unfamiliar EFL knowledge (Emilia 2011). Clearly, this is an aspect of multilingualism which influenced this decision. A new generation of genre pedagogy, the Reading to Learn (R2L) programme (Rose 2016), has a three-tier cycle with each layer consisting of different steps and having its own function, targeting different strata of language (see Figure 8.2). It shifts the focus of teaching to incorporate reading alongside writing. Rose (2006) foregrounds reading as the foundation of learning in school. As the name implies, R2L is a reading-centred pedagogy which involves unpacking the language and social context of a text (Rose 2006; Rose and Martin 2012). In principle, R2L is a reading-oriented pedagogy which organizes the teaching and learning steps by focusing on one high-stakes reading text to be deconstructed and reconstructed in one cycle of teaching and learning. In addition to the cycle, ‘learning exchange’ is an important part of the teaching. Learning exchange is teacher-planned talk in the classroom in which the opening and closing of a teacher’s exchanges are designed to enable students to complete a task successfully. To this point, R2L was selected to be extended for the bilingual pedagogy programme for at least two reasons. First, the focus on unpacking reading and choices of intensive strategies for reading and writing in the pedagogy provide a potential platform for recontextualization in multilingual classrooms. Second, the learning exchange offers potential to design the language interplay, planning the involvement of L1 and/or L2 in classroom interactions. In addition to that,

148

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 8.2  Reading to Learn three-layered cycle. Source: Rose and Martin (2012).

a number of successful examples of implementation have served as evidence of the appliability of the pedagogy to different levels of education. The following sections consider the extension of the R2L approach to multilingual settings. This is because those applications of the Sydney School approach involving ESL learners are often largely monolingual, since the main challenge is to assist learners in accessing the curriculum of the mainstream education, which is only in the L2.

READING TO LEARN BILINGUAL PROGRAMME The approach to extending R2L for multilingual classrooms followed the parameters of bilingual programmes based on Halliday’s threefold model of language learning. The programme was extended to include both content and language focus, as well as moving from enfolding towards enveloping (see Figure 8.1). The R2L bilingual programme aimed to teach English and biology, and was designed to involve the use of Bahasa Indonesia in different degrees. It consisted of two main structures: the curriculum cycle and the classroom interactions. The curriculum cycle was enacted in three iterations, each realized in the stages from the R2L three-tier cycle (see Figure 8.2). The classroom interactions were designed based on the R2L learning exchange structure but Bahasa Indonesia (L1) was involved systematically to different degrees depending on the stage of the iteration. Bilingual curriculum cycle The R2L programme is centred on selected high-stakes reading texts appropriate to the curriculum. These reading texts along with the curriculum and students’ needs are the basis on which the teaching steps from the R2L three-layered

THE READING TO LEARN BILINGUAL PROGRAMME

149

cycle are selected. In the intervention programme, this concept was extended by using two high-stakes reading texts in Bahasa Indonesia and one in English, all selected based on genre and register appropriate to the English and biology units that were being taught. At the time of the programme implementation, the Year 8 student participants needed to study descriptive report texts for their subject English, and endangered animals for biology. Thus, three reading texts were selected, entitled ‘Nisaetus bartelsi’, ‘Ninox ios’ and ‘The purplecrowned fairy-wren’. The first two texts, written in Bahasa Indonesia, are about endangered birds from Indonesia, and the latter is in English, about an endangered bird of Australia. The texts were selected based on the comparability of their topics. Each text included written text presented as a list of information and an illustrated image of the bird. Each of these reading texts was unpacked in one iteration of the R2L cycle (Figure 8.3), resulting in three iterations conducted in the programme. Each iteration selected the same four strategies from the R2L three-tier cycle: (1) preparing for reading, (2) detailed reading, (3) joint construction and (4) individual construction (see Figure 8.3). The first two steps aimed to unpack the reading texts in all strata. Preparing for reading focused on field and genre, that is, on what the text was about and how it was organized. Detailed reading focused on discourse and grammar, guiding students to recognize patterns of meaning and wording within and between sentences. The joint construction step aimed to repack the knowledge built in the previous steps. It focused on genre and field, but also built skills in discourse, grammar and graphology. A distinct stage in joint construction called note making played an important role in assisting in re-instantiation. This stage specifically focused on field,

FIGURE 8.3  The cycle of the R2L bilingual programme.

150

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

grammar, graphology and phonology in L1 and L2. Individual construction focused on discourse, grammar and graphology. It was closely related to writing development and assessment, since the teacher marked the students’ individually written texts. The use of L1 reading texts was a vital factor in designing the curriculum in such a way that teacher–student interactions could take advantage of the multilingual contexts. As the iterations started from texts in Bahasa Indonesia, and then moved to English, the L1 was used more dominantly earlier in the curriculum before shifting to L2, while support was still provided in L1. More specific patterns of L1 and L2 shifting will be elaborated in the subsection below. Bilingual classroom interactions The classroom interactions in the R2L bilingual programme were designed to include both peripheral and core phases in the learning exchange structure, as discussed immediately below. Further, this design included selection of languages to be used at different phases in the exchange. The interactions aimed to manage teachers’ talk, ensuring success for all students in accomplishing a learning task. Thus, the strategic selection of languages reflected this aim. The bilingual exchange patterns were built on the phases of learning exchange (Figure 8.4). In R2L, enacting the curriculum cycle includes the design of talk around text, consisting of core and peripheral phases. The core phases are focus – task – evaluate, and the peripheral phases include prepare and elaborate. The prepare, focus, evaluate and elaborate phases are performed by the teacher. The task phase is performed exclusively by the students, and can be realized in different ways, from answering a question to performing an activity. As each step has its own function, there were different patterns of exchange and different combinations of language shifting. Iterations 1 and 2, which used L1 reading texts, involved more L1, and Iteration 3 involved more L2, moving from enfolding towards the enveloping pole (Figure 8.1). The preparing for reading stage (Figure 8.3) aimed to unpack the reading texts in terms of their

FIGURE 8.4  R2L learning exchange. Source: Martin & Rose (2012) © Equinox Publishing Ltd 2012.

THE READING TO LEARN BILINGUAL PROGRAMME

151

genre and field, preparing the students for getting into the detail of the reading texts. In this stage, the teacher made explicit the purpose of the text, the stages and phases built the text through the use of the paragraphs, and the information built in the field of the text. In all iterations, L1 was used most of the time and a shift to L2 was kept to a minimum so as to reduce the students’ learning load. In the detailed reading stage, the use of languages was more varied depending on the reading text. As the step deals with greater detail of the text in terms of its discourse semantics and grammar, more support was provided to ensure the students accessed the key lexical items and grammatical structures in the texts. Throughout the iterations, the teacher’s talk involved both core and peripheral phases in exchanges, but in Iterations 1 and 2 the exchange was initiated by L1 use and closed by L2 use, as exemplified below. By preparing and focusing in L1 (Figure 8.4), the students were guided to identify words in the language that they were familiar with, which assisted in them to successfully complete the task. After this success, the teacher gave positive evaluation in L2 (Figure 8.4), affirming their answer and introducing the use of L2 positively. In Iteration 3 where an L2 reading text was used, the prepare and focus phases involved different languages. The teacher started with selected prepare phases in L2 followed by L1, then continued with a focus phase in L2. As a student answered the question by identifying the word in the text (task: identify), the teacher gave praise in the evaluate phase in L2. Example 2 illustrates this typical scenario. In the joint construction stage (Figure 8.3), the function was to jointly write a new text, repacking the genre, register, discourse semantics and grammar that the students have learned in the previous stages. In all iterations, the classroom interactions involved L1 in each phase. Typically, L2 was used for metalanguage

TABLE 8.1  Example 1: Bilingual Learning Exchange in Detailed Reading Iterations 1 and 2 Speaker

Exchange

Phases

T

Dalam kalimat pertama.

Prepare

(It’s) in the first sentence. Berapa besarkah Elang jawa?

Focus

How big is Elang jawa? S

Enampuluh sentimeter.

Task: identify

Sixty centimetres. T

Fantastic!

Evaluate

152

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 8.2  Example 2: Bilingual Learning Exchange in Detailed Reading Iteration 3 Speaker

Exchange

Phases

T

Now I want you to focus on the first sentence.

Prepare

Yang ini ya. It’s this one (reading the sentence aloud) I want you to find the genus of this species. Focus S

Aku tahu!

Task: identify

I know! Malurus T

Fantastic!

Evaluate

TABLE 8.3  Example 3: Bilingual Learning Exchange in Joint Construction Iteration 3 Speaker

Exchange

Phases

T

Ini kalimatnya sudah mulai ‘the female is like non-breeding male’.

Prepare

The sentence has been started with . T

Kalimat selanjutnya akan menerangkan the female.

Prepare

The next sentence will describe . Source: Kartika-Ningsih and Rose (2018).

and key words used to build the sentence as the teacher and the students wrote the text together. Throughout the iterations, there were patterns of L1 and L2 use, before moving towards the enveloping cline. This can be seen in Example 4 where the learning exchange leaned towards enveloping cline. The teacher prepared by restating the sentence, emphasizing ‘only the male’, and asked the class to interpret its implied meaning ‘What about the female?’ One student proposed ‘doesn’t have’, which the teacher affirmed by repeating, and further elaborated by putting it in a sentence. L2 learning exchange occurred more in Iteration 3 both in detailed reading and joint construction in that the students responded to the teacher’s question in L2 by proposing answers in L2. This reflected the movement from the

THE READING TO LEARN BILINGUAL PROGRAMME

153

TABLE 8.4  Example 4: Enveloping: L2 Learning Exchange Speaker

Exchange

Phases

T

OK it’s purple, purple crowned.

Prepare

Only the male has the purple crown. What about the female?

Focus

S

Doesn’t have.

Propose

T

Doesn’t have.

Affirm

So you can say: ‘The female doesn’t have the purple crown’.

Elaborate

enfolding to the enveloping pole. L1 was used as a scaffold at the beginning of the intervention, and this scaffold was gradually withdrawn as students’ L2 resources expanded.

ENVOI Summary The Reading to Learn bilingual programme is an innovative bilingual pedagogic practice, designed to address the issues of high-stakes literacy teaching/learning in multilingual classrooms. It is innovative since aspects of multilingualism are carefully designed from the teaching stages to the classroom interactions. The multilingual aspects include the use of high-stakes reading texts in L1 and the systematic design of bilingual interactions. Using L1 reading texts in L2 teaching has proven to be useful due to the similarities in terms of generic staging and the assistance it provides students in managing the complexity of the field. The bilingual interactions provide a novel and strategic approach to designing the language interplay in classroom talk in a way that uses the patterns of L1 and L2 classroom talk in a principled, pedagogical manner (see Kartika-Ningsih and Rose 2018). Challenges and reflection Notable challenges in the implementation of the programme included: (1) the consideration of reading texts in Bahasa Indonesia and (2) the challenges for teachers to become familiar with the learning exchange structures. Regarding the first difficulty, generically, the texts in the study were descriptive reports in science, which have similarities with English descriptive reports in terms of generic stages and register. The similarities proved useful in designing the programme: it was possible to exploit the consistency in the generic stages,

154

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

field of science and discourse semantics through the use of technical terms. However, at the grammatical level, a lot of sentence deconstruction done in the interactions relies on intuition. This is because grammatical descriptions of Bahasa Indonesia based on Systemic Functional Linguistics are not fully developed, and often unavailable. Regarding the second difficulty, during the interactions, the teacher needed to follow the plan designed in the learning exchanges, and consciously shift languages based on the phases built. It was difficult to follow the lesson plan in the beginning, and it took continuous awareness in order to be consistent. Several issues remain to be solved in further studies concerning regional languages and the teaching of other genres. The issue of the variety of multilingual backgrounds is especially important. For example, Sundanese, a regional language prominent in the area where the project took place, was not considered in this study. As already noted, most of the students spoke Sundanese, though they learned specialized discourse in Bahasa Indonesia. But if we turn to Sundanese, will the same method of teaching be applicable? Will teaching English involve three languages? Or will teaching English involve two languages, English and Bahasa Indonesia, or the regional language? There are some technical terms and concepts that are not yet lexicalized in many regional languages, adding additional challenges to developing trilingual classroom interactions. Additionally, the lessons in the study involved writing a descriptive report in science in a genre that is relatively simple and shares similarities with that of Bahasa Indonesia. What would happen if other genres were taught? Narrative or argumentative genres have different generic structures in Bahasa Indonesia, and so it may require different strategies in the design and practices of teaching/learning. Apart from these issues, the whole area of classroom interaction requires much more research and analysis. Teacher–student interactions are important in the sense that during the interactions with the teacher knowledge is constructed and interpreted for and by the students, and peer interactions often involve the students talking about the field and about the target language during learning. Thus, a great deal of research remains to be undertaken in order to inform further design and implementation of effective and contextually appropriate language curricula in multilingual settings in the future.

REFERENCES Bernstein, B. (1975), Class, Codes and Control III: Towards a Theory of Educational Transmission, London: Routledge and Kegan Paul. Bernstein, B. (1999), ‘Vertical and Horizontal Discourse: An Essay’, British Journal of Sociology of Education, 20 (2): 157–73. Bernstein, B. (2000), Pedagogy, Symbolic, Control, and Identity: Theory, Research, Critique, rev. edn, Lanham, MD: Rowman & Littlefield.

THE READING TO LEARN BILINGUAL PROGRAMME

155

Brisk, M. E. (2015), Engaging Students in Academic Literacies: Genre-Based Pedagogy for K–5 Classrooms, New York: Routledge. de Silva Joyce, H. and S. Feez (2012), Text-Based Language and Literacy Education: Programming and Methodology, Outney, NSW: Phoenix Education. Disadvantaged Schools Program (1988), Teaching Factual Writing: A Genre-Based Approach, Sydney: Disadvantaged Schools Program. Emilia, E. (2010), Teaching Writing: Developing Critical Learners, Bandung: Rizqi Press. Emilia, E. (2011), Pendekatan Genre-Based dalam Pengajaran Bahasa Inggris: Petunjuk untuk Guru, Bandung: Rizqi Press. Errington, J. (1998), Shifting Languages: Interaction and Identity in Javanese Indonesia, Cambridge: Cambridge University Press. García, O. (2009), ‘Education, Multilingualism and Translanguaging in the 21st Century’, in A. Mohanty, M. Panda, R. Phillipson, T. Skutnabb-Kangas (eds), Multilingual Education for Social Justice: Globalising the Local, 128–45, New Delhi: Orient Blackswan. Genesee, F. and K. Lindholm-Leary (2013), ‘Two Case Studies of Content-Based Language Education’, Journal of Immersion and Content-Based Language Education, 1 (1): 3–33. Goebel, Z. (2010), Language, Migration, and Identity, Cambridge: Cambridge University Press. Halliday, M. (1993), ‘Towards a Language-Based Theory of Learning’, Linguistics and Education, 5 (2): 93–116. Hamers, J. F. and M. H. Blanc (2000), Bilinguality and Bilingualism, Cambridge: Cambridge University Press. Kartika-Ningsih, H. (2016), ‘Multilingual Re-instantiation: Genre Pedagogy in Indonesian Classrooms’, PhD thesis, Sydney University, Sydney. Kartika-Ningsih, H. (2020), ‘Implementing the Reading to Learn Bilingual Program in Indonesia’, in K. Rajandran and S. Abdul-Manaf (eds), Systemic Functional Linguistics: South East Asian Perspectives, 145–63, Singapore: Springer. Kartika-Ningsih, H. and D. Rose (2018), ‘Language Shift: Analysing Language Use in Multilingual Classroom Interactions’, Functional Linguistics, 5: 9. doi:10.1186/ s40554-018-0061-0 Llinares, A., T. Morton and R. Whittaker (2012), The Roles of Languages in CLIL, New York: Cambridge University Press. Martin, J. R. (1992), English Text: System and Structure, Amsterdam: John Benjamins. Martin, J. R. and D. Rose (2007), Working with Discourse: Meaning beyond the Clause, London: Continuum. Martin, J. R. and D. Rose (2008), Genre Relations: Mapping Culture, London: Equinox. Murray, N. and K. Zammit (1992), The Action Pack: Animals (Activities for Teaching Factual Writing), Sydney: Metropolitan East DSP. Musthafa, B. and F. Hamied (2014), ‘Teaching English as a Foreign Language in Indonesian Schools in the Reform Era: What Do Teachers Have to Say?’, The New English Teachers, 8 (2). Rose, D. (2006), ‘Towards a Reading-Based Theory of Teaching’, in L. Barbara and T. B. Sardinha (eds), Proceedings of the 33rd International Systemic Functional Congress, 33–77, Sao Paulo: PUCSP. Available online: https://www.pucsp.br/isfc/ proceedings/Artigos%20pdf/02pl_rose_david_36a77.pdf (accessed 1 November 2014).

156

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Rose, D. (2016), Reading to Learn: Accelerating Learning and Closing the Gap, Gladesville, NSW: David Rose. Rose, D. and J. R. Martin (2012), Learning to Write, Reading to Learn: Genre, Knowledge and Pedagogy in the Sydney School, South Yorkshire: Equinox Publishing. Schleppegrell, M. J., J. Moore, S. Al-Adeimi, C. O’Halloran, A. Palincsar and C. Symons (2014), ‘Tackling a Genre: Situating SFL Genre Pedagogy in a New Context’, in L. de Oliveira and J. Iddings (eds), Genre Pedagogy across the Curriculum: Theory and Application in US Classrooms and Contexts, 25–39, London: Equinox. Spada, N. (2007), ‘Communicative Language Teaching: Current Status and Future Prospects’, in J. Cummins and C. Davison (eds), International Handbook of English Language Teaching, 271–88, Boston, MA: Springer. Suherdi, D. (2012), Rekonstruksi Pendidikan Bahasa: Sebuah Keniscayaan bagi Keunggulan Bangsa, Bandung: Celtics Press.

CHAPTER NINE

Big Data and Managing Multimodal Complexity KAY L. O’HALLORAN, SABINE TAN, PETER WIGNELL, REBECCA LANGE, KEVIN CHAI AND MICHAEL WIEBRANDS

INTRODUCTION: CONTEXT AND ISSUE We have proposed that abstract context-based frameworks for modelling multimodal semiotic resources and analytical methodologies for investigating patterns of human communication are required if multimodality is to become a discipline in its own right (O’Halloran et al. 2020). If this were to happen, we foresee that multimodality has the potential to provide the foundations for the multimodal sciences, in much the same way that linguistics and mathematics led to the development of the language sciences and mathematical sciences respectively. However, in order for this to occur, multimodality needs to move beyond existing levels of description, generalization and demonstration of theoretical concepts derived from the analysis of a limited number of texts to large-scale analysis and corpusbased empirical grounding and testing of insights (e.g. O’Halloran 2015; Bateman 2016). In this way, multimodality would itself become an ‘appliable science’ for investigating human activity (see Michael Halliday’s notion of appliable linguistics in this volume; Halliday 2008; Mahboob and Knight 2010). We first started working towards this goal in the Multimodal Analysis Lab in the Interactive & Digital Media Institute at the National University of Singapore in 2007. This involved establishing interdisciplinary teams of researchers (social

158

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

scientists, computers scientists, mathematicians and scientists) to explore how multimodal theory could be combined with computer-based techniques in order to (a) handle the complexity of multimodal analysis; (b) advance knowledge of how semiotic resources work together to make meaning and (c) trace patterns and trends across large multimodal data sets (e.g. see O’Halloran 2015). Following this, we established a research programme at Curtin University in 2013–19 which involved integrating multidisciplinary theories and techniques of multimodal analysis with computer vision, natural language understanding systems and machine learning in order to tackle the complex problem of big data. In doing so, we sought to track the changes in human communication arising from digital technology and the resultant impact. We have continued this research programme in the Department of Communication and Media in the School of the Arts at the University of Liverpool (2019–present). The research programme at Curtin University involved working with the Curtin Institute of Computation (Curtin 2020) to explore the potential of a context-based mixed methods approach for investigating how violent extremist groups use images and text to attempt to legitimize their views, incite violence and influence potential recruits and supporters in online propaganda materials, and how the images from these materials are re-used and recontextualized across different media platforms to support or resist violent extremism (Curtin University 2018; Drok 2018). Terrorist groups’ use of the internet has rapidly evolved, given the availability of new technological opportunities, the proliferation of social media platforms and other applications, developments in online policing and other factors. In this case, we show how Islamic State (referred to hereafter as ISIS) adapt their communication strategies according to their changing situation, and how other social actors respond by extracting and re-inserting elements of these communications in online contexts (e.g. news media, social media, blogs) in ways that function to support or resist the terrorist group’s agenda. A pilot project in this area provided the grounds for a collaborative project with the Defence, Science and Technology Group in Edinburgh, South Australia, in 2018–19.1 The proof-of-concept methodology developed in the project uses multimodal discourse analysis to enhance the results obtained from computer vision, machine learning and natural language understanding systems, to analyse text and images in the context of online extremism-related communications. The research programme at Curtin University involved: (a) analysing image and text relations in large data sets through the development of a multimodal, mixed methods analytical model; (b) understanding ISIS’s world view and agenda as constructed through image and text relations in their propaganda materials and strategies for enacting these; (c) examining how images from ISIS propaganda materials are recontextualized in different online media

BIG DATA AND MULTIMODAL COMPLEXITY

159

sites and how these images are used in texts which report on ISIS activities; and (d) testing and evaluating results obtained from computer vision, natural language understanding systems and machine learning for integration into a (semi)-automated prototype system for analysing images, text and image–text relations according to context. We developed two prototype software applications to assist us with these research activities: (1) a Multimodal Analysis Visualization App for displaying relations among image and article types in online extremist propaganda magazines and the sites where images from these materials reappear over time (e.g. Tan et al. 2018) and (2) a Multimodal Analysis Cluster App for visualizing similarities and differences in online media articles and measures of those differences in terms of features derived from the analysis with language and image processing tools which are integrated into a multimodal framework for the analysis of text, image and text–image relations. In what follows, we provide an overview of the background and goals of the research programme that aimed to develop multimodal context-based mixedmethods approaches to big data analytics, in this case in the domain of online extremist communications. Following this, we describe the methodologies and outcomes of the project, together with future directions.

OVERVIEW AND BACKGROUND: A MULTIMODAL APPROACH TO ONLINE EXTREMIST COMMUNICATIONS The use of online media to distribute violent extremist messages is well documented by governments, think tanks and large corporations alike. For example, ISIS have used the internet and social media on a constant basis for propaganda purposes. However, many current approaches employing ‘big data’based methodologies such as social network analysis, data mining and other tools for analysing large datasets are grounded in content- and/or platform-focused analyses of messaging and interactions. While such analyses can offer important insights about the meaning potential of extremist propaganda materials and channels of their dissemination, they remain so far insufficient for revealing the multimodal communication strategies employed, and for understanding their effects within a variety of contexts. For example, the key issue of the meanings arising from the integration of language and images needs to be addressed in order to understand terrorist-related communications (see discussion in Wignell, Tan, O’Halloran and Lange 2017; O’Halloran et al. 2019). With this in mind, we aimed to demonstrate the potential of a contextbased mixed-methods approach for analysing how violent extremist groups use images and text to attempt to legitimize their views, incite violence and influence potential recruits and supporters in online propaganda materials, and

160

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

how the images from these materials are re-used and recontextualized across different media platforms to support or resist violent extremism. The proof-ofconcept methodology involved using multimodal discourse analysis to enhance the results obtained from computer vision, natural language understanding systems and machine learning, along with Wikipedia classifications of these results to provide higher-order semantic information for key objects in the images, keywords in the text, and the context (e.g. metadata, the URL). The context-based mixed-methods approach aimed to ‘resolve the gap between highly-detailed, contextualised analyses of small samples of multimodal texts on the one hand, with highly-aggregated, decontextualised big data approaches (e.g. reductive content analysis) on the other’ (O’Halloran et al. 2019: 454). The research activities focused on developing an integrated (semi)automated prototype system for analysing images, text and image–text relations according to context to interpret terrorist-related communications in online media; and investigating key features of the messages which function to support or resist political violence and terrorism. The outcomes included: (a) understanding the communication strategies employed by radicalized actors and the patterns of recontextualization in the spread of violent extremist images in online media; (b) improving discourse analytical techniques for contextenhanced information fusion and situational awareness, where materials are gathered from a variety of sources; (c) an evidence-based assessment of the multimodal strategies (text and images) used by ISIS and how the global community responds in terms of reusing the images for various purposes; and (d) proof of concept of theoretically informed empirical techniques for the automated contextual analysis of text and image relations in extremismrelated communications.

RESEARCH APPROACH AND METHODOLOGY The research programme unfolded in several stages. First, we proposed a multimodal approach to big data analytics (which we later revised: see below). Second, we explored the nature of ISIS propaganda and recontextualizations of images from those online media materials in depth. Third, we developed an integrated (semi)-automatic system for identifying key features of the dataset using multimodal analysis, computer vision, natural language processing and contextual information (e.g. website classification). The prototype uses existing image and language processing tools, semantic classifications and machine learning and neural networks to identify the messages which function to support or resist extremism and terrorism, together with the linguistic, visual and contextual attributes of those messages. Lastly, we investigated how the image and text processing tools and the system as a whole can be improved

BIG DATA AND MULTIMODAL COMPLEXITY

161

using multimodal analysis theory and computational techniques, and how the multimodal analysis system could be trained to identify relevant features of texts, images and their relations. The theoretical approach to multimodal discourse analysis underpinning this work is based on social semiotics, involving the study of how sign systems are used to create meaning contextually (van Leeuwen 2005). Specifically, the approach we adopted is based on Systemic Functional Theory, where language, images and other sign systems (i.e. semiotic resources) are viewed as resources for making meaning (e.g. Martin and White 2005; Martin and Rose 2007; Halliday and Matthiessen 2014). As such, semiotic resources are conceptualized as ‘systems of meaning’ with networks of options from which choices are made in multimodal texts. Following Halliday, the systems are organized according to the functions which the resources have evolved to serve: (a) experiential and logical meaning: to construe experience of the world; (b) interpersonal meaning: to enact social relations and create a stance; and (c) textual meaning: to organize the messages as coherent text. The messages in any communicative situation are characterized in terms of options selected from these systems.

THE DATA SET The research project focused on ISIS propaganda materials published in English. The English language versions of Dabiq and Rumiyah, ISIS’s official online internet magazines, were chosen as they are two of the few original sources of data directly produced by ISIS. The original dataset comprised fifteen issues of Dabiq, released between 5 July 2014 and 31 July 2016, and thirteen issues of Rumiyah, released between 6 September 2016 and 9 September 2017, and downloaded materials from webpages which contain the same or similar images as those used in the ISIS propaganda materials. The URLs of webpages were located using TinEye2 reverse image search, which crawls the web to identify these images. The reverse image search resulted in 130,681 webpages. The repository was cleaned by identifying the English language webpages only, and removing duplicated and no longer available webpages. The cleaned dataset consisted of 39,464 webpages. We used subsets of the cleaned data set for testing the approach and the proof-of-concept methodology, as elaborated below.

RESEARCH ACTIVITIES, OUTCOMES AND FINDINGS The overall goal was to demonstrate the potential of a context-based mixedmethods approach (using multimodal analysis, computer vision, natural language processing and contextual information) in developing an integrated

162

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

semi-automatic system for analysing images, text and image–text relations in terrorist-related communications, and patterns of reuse in online media. The work can be categorized into five themes, each of which was developed simultaneously and sequentially, as follows: (a) developing a multimodal context-based mixed-methods analytical model for analysing text and image relations in large data sets; (b) understanding ISIS’s world view and agenda as they are constructed through image and text relations in their propaganda materials and their strategies for enacting these; (c) examining how images drawn from ISIS propaganda material are recontextualized in different online media and other sites and explore how these images are used in texts which report on ISIS activities; (d) testing and evaluating the results obtained from automated computer vision natural language understanding systems and machine learning. This body of work moves from detailed analysis of a relatively small number of texts to analysis of progressively larger samples of ‘bigger’ data; (e) consolidating prior research activities into a prototype model which integrates multimodal discourse analysis, computer vision, natural language understanding and machine learning in a context-based multimodal framework. The respective outcomes of these research activities are summarized below. Development of a multimodal mixed-methods analytical model We proposed a theoretical prototype for the development of a multimodal analytical model in O’Halloran et al. (2018). In this chapter, we discuss how a digital environment offers new opportunities for transforming qualitative data into quantitative data in order to use data mining and information visualization for mixed-methods research. The digital approach to mixedmethods research is illustrated by a framework (see Figure 9.1) which combines qualitative methods of multimodal discourse analysis with quantitative methods of data mining and information visualization in a multilevel, contextual model that is intended to result in an integrated, theoretically wellfounded, and empirically evaluated technology for analysing large data sets of multimodal texts. The proposed mixed-methods approach involves four stages: (a) compilation of the multimodal dataset; (b) identification of key systems choices; (c) use of machine learning and data mining to analyse large datasets; and (d) exploration

BIG DATA AND MULTIMODAL COMPLEXITY

163

FIGURE 9.1  A multimodal mixed-methods approach. Source: O’Halloran et al. (2018: 24).

of patterns using interactive visualizations (see Figure 9.1). We proposed that the framework would be applicable to situations in which critical information needs to be extracted from geotagged public data: for example, in crisis informatics. We discuss how this model might be applied to the analysis of violent extremist discourse in O’Halloran et al. (2019). We exemplify this approach through a study which analyses how violent extremist groups use language and images to attempt to legitimize their views, incite violence and influence recruits in online propaganda materials, and how the images from these materials are re-used in different media platforms in ways that support and resist violent extremism. We exemplify the multimodal framework and introduce examples of language and image systems, as illustrated in Table 9.1 (linguistic systems are in small caps, following Halliday and Matthiessen 2004; Martin and White 2005). The approach introduces and forecasts the move into interdisciplinary big data analysis that is progressively developed in the research programme. Figure 9.2 shows an example of the multimodal framework applied to the analysis of text and images relations, with overlays corresponding to various system choices. However, we subsequently revised the proposed approach of manually coding the materials and using the results for machine learning and data mining, given the large volume of analysed data which was required and our limited resources. We continued to closely examine our data set of ISIS propaganda materials, as described in the following subsections, while developing an

164

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 9.1  Examples of Language and Image Systems (Adapted from Halliday & Matthiessen 2014; Martin & Rose 2007; O'Toole 2011; Kress & van Leeuwen 2020) Metafunction

Rank

System

Description

Discourse

Ideation

Activity sequences

Clause

Transitivity; Ergativity

Happenings, actions and relations (i.e. processes, participants and circumstances)

Discourse

Appraisal

Attitude, emotion and judgment

Clause

Speech Function; Modality

Exchange of information/ goods & services; Probability, usuality and potentiality

Discourse

Identification

Tracking of participants

Clause

Information Focus

Organization of the message

Work

Narrative Themes; What is portrayed Scenes, Interplay of Episodes; Setting

Episode

Actions; Events; Interplay of Actions

Visual happenings, actions and relations (processes, participants and circumstances)

Figure

Character; Acts/ Stance/Gesture; Clothing; Posture

Participant characteristics

Work

Light, Perspective, Angle; Camera Distance

Visual effects and orientation

Episode

Relative Prominence; Focus; Perspective

Happenings, actions and relations with respect to the whole image

Figure

Gaze-Visual Address; Contrast

Direction of participant’s gaze as internal to image or external to viewer; Contrasting elements

LANGUAGE Experiential

Interpersonal

Textual IMAGE Experiential

Interpersonal

BIG DATA AND MULTIMODAL COMPLEXITY

165

Metafunction

Rank

System

Description

Textual

Work

Compositional Vectors; Proportion; Framing

The organization of the parts as a whole, with visual markings of certain parts

Episode

Relative Placement of Episode; Framing

Relative position of the happenings, actions and relations in relation to the whole image, with visual marking of certain aspects

Figure

Relative Placement in the Episode; Arrangement; Framing

Relative position of figures and the visual marking of certain aspects of those figures

FIGURE 9.2  Screenshot of sample analysis carried out within Multimodal Analysis Image software: Overlays [1]; selected system choice [2]; list of available system choices [3]; annotation nodes [4]; system strips [5].

alternative big data approach which utilized multimodal theory with existing computational tools. Understanding ISIS’s world view and agenda as constructed through image and text relations We developed our initial analysis of ISIS’s world view and values and how these are synthesized through bonding icons in Wignell, Tan and O’Halloran (2017b). This study employs a multimodal social semiotic approach to the

166

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

analysis of text and image relations in ISIS propaganda material published in Dabiq magazine. The study focuses on iconization, where meanings are condensed and interpersonally charged through ‘bonding icons’ (e.g. ISIS flag, AK-47 rifles, ISIS heroes) supported by references to selected Islamic scripture which embody the organization’s world view and values. The ISIS world view is shown to be internally cohesive, based on a narrow, fundamentalist interpretation of selected Islamic scripture and violently opposed to any other world view. This study was followed by a more detailed investigation of the role of bonding icons in materials produced by ISIS, the attraction of ISIS to potential recruits and ‘copy-cat’ jihadist groups and the recontextualization of ISIS propaganda materials in different media platforms. For example, we explore the use of bonding icons in recruitment strategies to entice foreign fighters to join ISIS and other Sunni militant organizations in Wignell, Tan and O’Halloran (2017a). The work on understanding ISIS’s world view and agenda as constructed through text and image relations in their propaganda materials contributed to our understanding of (a) how ISIS propaganda changed over time and (b) the different ways in which ISIS propaganda materials are recontextualized in mainstream online media over time. Changes in ISIS propaganda and examination of recontextualized ISIS propaganda materials in different online media In September 2016, Dabiq magazine was replaced by Rumiyah magazine as ISIS’s principal online propaganda outlet; a change which coincided with a decline in ISIS’s fortunes on the ground. We examined the changes in ISIS’s strategic focus concurrent with the change in magazine titles in Wignell, Tan, O’Halloran and Lange (2017). In this study, we used a mixed-methods approach integrating a qualitative social semiotic discourse analysis approach with quantitative methods of information visualization to examine empirical changes in emphasis and approach in both magazines over time to determine whether the changes were ones of style or substance. We argue that, while ISIS have changed their strategic focus over time in response to its changing circumstances, the organization’s underlying world view, values and ultimate aims remained consistent and unchanged. As explained above, we had shown how ISIS orchestrate combinations of images and text to construct a world view which is internally cohesive, based on a narrow, fundamentalist interpretation of selected Islamic scripture which rejects and is violently opposed to any other world view (Wignell et al. 2020; see also Wignell, O’Halloran et al. 2018). While ISIS present their ideology explicitly in materials they produce, images which encapsulate this ideology are distributed widely on the internet and recontextualized in different patterns in a variety of mainstream and social media. Thus, the next phase involved

BIG DATA AND MULTIMODAL COMPLEXITY

167

investigating the communication strategies employed by terrorist groups and the patterns of recontextualization in the spread of violent extremist images in online media, with a view to developing effective means for understanding and thus countering the effects of online violent extremist messages. Therefore, we developed our work on image and text relations in several further studies which examine image and text relations in their context in Dabiq and Rumiyah and extended the approach to comparing and contrasting the types of image and text relations found when images taken from the ISIS material are recontextualized in other online media outlets. For this purpose, we categorized articles and images in Dabiq and Rumiyah into types and explored relationships between article types and image types (O’Halloran et al. 2017). We found that certain image types were closely associated with certain article types in the two ISIS magazines. We then conducted reverse image searches to determine how widely distributed the images were and in what types of media outlets they were found. This work was facilitated through the development of the prototype Multimodal Analysis Visualization application which offers several different visualizations for displaying image and article type relations in Dabiq and Rumiyah, such as arc graphs for showing image–article type connections, pie charts for showing image–article type distributions, and image–article type frequency charts for showing how extremist groups change their propaganda strategies according to their situation on the ground based on text–image combinations (O’Halloran et al. 2017; Tan et al. 2018). In addition, the MMA Visualization App is used for displaying the results of the reverse image searches conducted with TinEye, which show where images from ISIS materials reappear across different online sites over time (e.g. see Figure 9.3). The app also has tools for displaying an overview of websites in which images from these materials are found by means of an integrated word-cloud tool and for displaying the distribution of image categories before and after the publication date of each magazine using bar graphs. We further developed our study of recontextualized ISIS propaganda materials in different online media in Wignell, O’Halloran et al. (2018) where a systemic functional multimodal social semiotic approach is applied to the analysis and discussion of image and text relations in two sets of data. First, we examine the contextualization patterns of images and text in the online magazines Dabiq and Rumiyah. A second data set consisting of a sample of texts from Western online news and blog sites which include recontextualizations of images found in the first data set was then examined. We discuss a sample of examples involving the use and re-use of images in order to identify patterns of similarity and difference when images and text are recontextualized. We argue that the ISIS material tends to foreground an interpersonal stance which works in combination with textual composition: that is, the focus is on who

168

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 9.3  MMA Visualization App: Results of reserve image search for ‘The Flood’, Dabiq, Issue 2. Source: O’Halloran et al. (2017: 197).

and how. In contrast, the second data set (i.e. online materials with images from ISIS magazines) tends to foreground the content of what was being reported. These inferences indicate that further exploration of a larger data set is worth pursuing. In a study that signals a move into analysis of larger data sets, we used a multimodal mixed-methods approach to explore general recontextualization patterns of violent extremist images in online media (Tan et al. 2018). The study reported preliminary findings of the investigation of patterns in the reuse of a representative sample of 26 violent extremist images which appear in Dabiq and Rumiyah by others across various online media platforms (e.g. news websites, social media, blogs). Using multimodal discourse analysis, data mining and information visualization tools, we addressed questions such as which types of violent extremist images are recirculated most frequently in online media over time, on which types of online media these images reappear, and in which contexts they are used and re-used on these websites. Preliminary findings suggest different recontextualization patterns for certain types of ISISrelated images over time (e.g. images of captives decline over time unless a related event occurs, while images of young jubilant ISIS fighters appear on a

BIG DATA AND MULTIMODAL COMPLEXITY

169

consistent basis). The study also found that the majority of violent extremist images used in the sample analysis appeared to circulate most frequently on Western news and politics websites and news aggregate platforms. In order to determine the context in which these images are used and reused in online media, we explored the use of machine learning algorithms, specifically uClassify’s (2020a) IAB Taxonomy V2 classifier, which categorizes texts into 560 content topics based on the Interactive Advertising Bureau Quality Assurance Guidelines Content Taxonomy V2 (IAB 2020), and uClassify’s (2020b) Tonality classifier, which determines the ‘tonality’ of a text based on the degree of formality of the language: that is, whether the tone of a text is corporate (formal) or personal (informal). This work is described further below. Using computer vision, machine learning and natural language understanding systems for analysing larger samples of recontextualized online materials Shifting focus to the analysis of language and to the use of automated analytical tools, we tested and reported on the application of natural language processing models, such as IBM Watson’s (2020) Natural Language Understanding online demonstration models, to the text of a sample of articles from Dabiq and Rumiyah (Wignell, Chai et al. 2018). For comparison and contrast, we also analysed the text of articles found by reverse image search software which reused the lead images from the original articles in text which either reported on or opposed extremist activities. The aim was to explore what insights the natural language processing models (specifically categories, concepts, entities, keywords, emotion and sentiment) could provide in distinguishing between texts produced as propaganda to incite violent extremism and texts which either reported on or opposed violent extremism. The results showed that some valuable insights can be gained from such an approach and that these results could be improved through integrating automated analyses with a theoretical approach to the analysis of language and images in their immediate and social contexts. We proposed that such an approach could inform the interpretation of results and could be used in training software so that stronger results can be achieved in the future. We also tested the possibility of integrating automated visual recognition tools, which use machine learning and deep neural networks for identifying and captioning objects in images, into a multimodal framework to expedite the analysis of large amounts of visual data. For this purpose, we tested prebuilt models for object detection and labelling available with the free online trial versions of four visual recognition tools: DenseCap (Stanford University 2020), Clarifai,3 Google Cloud Vision (Google 2020) and IBM Watson Visual Recognition (IBM 2020). The results of the analysis of a small sample consisting

170

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

of five extremist terrorist-related images and five political discourse-related images carried out with these tools were then compared and evaluated in terms of their suitability for being integrated into a multimodal framework for image analysis. The results showed that, despite some apparent limitations and inaccuracies in object labelling and detection, certain models offered by visual recognition tools could be feasibly adapted and integrated into a multimodal framework to facilitate the analysis of large sets of visual images from a social semiotic perspective. Proof-of-concept development of an integrated multimodal analysis prototype system The last phase of our research programme at Curtin University involved consolidating earlier work in a prototype model which integrates multimodal analysis, computer vision, natural language understanding and contextual information for identifying patterns in large sets of multimodal data. Reverse image searches conducted for an earlier study which investigated the recontextualization patterns of a sample of 26 violent extremist-related images in online media (Tan et al. 2018) showed that these images were used and re-used on 4,504 English language webpages. From this subset, 985 webpages were downloaded and annotated manually for the purpose of identifying webpages that either support or resist terrorism, and to eliminate unusable webpages that were either too short, or incorrectly formatted. From this subset, articles with more than 10 sentences were selected for further analysis. The final data set comprised 254 English language webpages and 20 different ISIS-related images. The prototype model incorporates (a) the results obtained from language analysis obtained through existing natural language understanding algorithms provided by IBM Watson and (b) the results of the automated image analysis performed with Clarifai. The results of the analyses were integrated in a multimodal framework (see Table 9.2). In order to make sense of the similarities and differences in terms of use of text, images and the text–image relations in the 254 English language webpages, we applied K-mode clustering (Huang 1997, 1998) (which is an extension of K-means, and uses modes instead of means to form clusters of categorical data; Trevino 2016) to the results of the language and image analyses. The results of the cluster analysis are displayed in a web-based application, the MMA Cluster App, an interactive, interrogable visualization for text, image and text–image combinations (Figure 9.4). The spirals in Figure 9.4 show clusters of texts which are grouped according to similarities and differences in their linguistic and visual features. Each cluster has a centroid with a mean value for that cluster, and the clusters are positioned according to their relations to the cluster at the centre of the display. Individual results are displayed in the application as separate webpages, which include the article text, the URL of the source

BIG DATA AND MULTIMODAL COMPLEXITY

171

TABLE 9.2  Multimodal Framework for Prototype Model Rank

Metafunction

Features

Language Analysis (IBM Watson) Register

Experiential

Watson’s categorical values: 12 categories/features

Discourse semantics

Interpersonal

Watson sentiment and emotion analysis for each paragraph: Sentiment: 3 features Emotion: 5 features

Lexicogrammar

Logical

Conjunctive adverb analysis: 2 features

Work

Ideational/field

Clarifai concepts: 195 features

Episode

Interpersonal/tenor

Clarifai focus: 2 features for focus density

Figure

Ideational/field

Clarifai demographics: 7 features

Image Analysis (Clarifai)

article, the associated ISIS-related image (Figure 9.5), and a set of tables with the results of the language and image analyses (Figure 9.6). The data in the tables are mapped from the vector data (i.e. the set of features) displayed below the tables (Figure 9.7).

FIGURE 9.4  MMA Cluster App: Interactive visualization of text and image clusters.

172

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 9.5  MMA Cluster App: Individual results of text and image analysis for an article.

In total, the K-mode cluster analysis resulted in six text-only clusters and nine text/image clusters. Preliminary findings suggest that in terms of the language features investigated there appears to be similarity among texts, with the majority of articles (214) organized into a single text cluster. Closer examination of the text/image clusters further reveals that, while there is overlap with text-only clusters, six out of nine text/image clusters also appear to be organized around one particular image each. These findings were consistent even for clusters with two or three different images, as articles with secondary images in the cluster were few (usually only one or two) and generally furthest from the centroid (i.e. the mean result). In terms of context, the majority of articles in the clusters were assigned the IBM Watson categories of law, government and politics (81 articles), society (78 articles) and religion and spirituality (67 articles). A comparison with the content topics assigned by uClassify’s taxonomy classifier, and the tonality of the texts as defined by uClassify’s tonality classifier further

BIG DATA AND MULTIMODAL COMPLEXITY

173

FIGURE 9.6  MMA Cluster App: Tables with results of language and image analysis.

reveals that the majority of articles with images of bonding icons and ISIS heroes circulate on news and politics-related webpages with a corporate (formal) tone, whereas articles with images of their enemies (e.g. political figures in the West) appear to be (re)distributed more diversely in terms of webpage categories and content topics, and are also associated with a more personal (informal) tone, as shown in the Sankey diagram in Figure 9.8. In summary, the context-based multimodal mixed-methods approach developed in the project (which we have extended to include videos in subsequent research) has four steps (O’Halloran et al. 2020), as displayed in Figure 9.9.

174

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 9.7  MMA Cluster App: Vector data for language (and image) analysis.

FIGURE 9.8  Sankey diagram showing distribution of articles with ISIS-related images in online media, in terms of IBM Watson categories, uClassify’s IAB Taxonomy webpage content topics, and tonality (i.e. corporate (formal) or personal (informal)).

BIG DATA AND MULTIMODAL COMPLEXITY

175

1. Feature extraction: Automated computational techniques are applied (e.g. natural language understanding, image processing and video processing models). 2. Multimodal feature enhancement: The various computational models are integrated within a multimodal analysis framework, so that each model is categorized according to semiotic resource, metafunction (experiential, logical, interpersonal and textual), and rank. This means that the results from the computational models are marked up according to the multimodal theoretical framework, which also indicates the gaps where there is missing information. 3. Feature representation: The various features of the multimodal analysis and the context are represented by a series of vectors. 4. Classification: The multimodal texts are classified using machine learning algorithms which have been trained using previously classified data. Examples of machine learning algorithms include neural networks, decision trees, logistic regression and other statistical methods (e.g. K-modes clustering, as described above).

FIGURE 9.9  The multimodal analysis approach. Source: O’Halloran et al. (2020: 106).

176

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

REFLECTION AND CONCLUDING COMMENTS The research programme resulted in several key achievements. First, we developed a Multimodal Analysis System Prototype, which is an integrated semiautomatic system for identifying key features of the dataset using multimodal analysis, computer vision, natural language processing and contextual information. Second, we theorized the nature of extremist propaganda materials published in English, the language other than Arabic most frequently used by ISIS to reach Western audiences. Lastly and most importantly, we developed a proof-of-concept methodology for theoretically informed empirical techniques for the automated contextual analysis of text, image and video relations in extremism-related communications which is applicable to other domains of human communication. In this way, we have demonstrated that the mixed-methods approach developed and applied, in this case in the context of violent extremist terrorist communications and their uptake in online media, is not only feasible and productive, but potentially applicable to any situation in which critical information needs to be tracked and analysed. For example, we are adapting the approach to the study of online news and social media in our research group at the Centre for Interdisciplinary Research, Bielefeld University in 2020. Likewise, applications developed in this project, such as the Multimodal Analysis Visualization App, have captured the interest of international research communities. For example, we worked on a collaborative project involving the analysis of images of austerity in the UK (Tan et al. 2020). We have continued this research programme at the University of Liverpool (2019–present) by developing a Multimodal Analysis Platform for search, retrieval and analysis of online news and social media (O’Halloran, Pal and Jin 2021). We have taken some small steps towards paving the way for the multimodal sciences for studying human communication in the digital age. This has not proved easy, given the traditional disciplinary boundaries which stand in the way of multidisciplinary research (e.g. in terms of research funding applications). Nonetheless, we hope that it will be increasingly recognized that such collaboration is imperative if we are to understand the means through which thought and reality are structured in the digital age.

NOTES 1 ‘A Context-Based Mixed Methods Approach to Extremist Communications’ with Lead Chief Investigator Kay O’Halloran (Curtin University) and Chief Investigator Dr Lucy Resnyansky, Defence, Science and Technology Group, Department of Defence, Australia. 2 https://www.tineye.com/ 3 https://clarifai.com/

BIG DATA AND MULTIMODAL COMPLEXITY

177

REFERENCES Bateman, J. (2016), ‘Methodological and Theoretical Issues for the Empirical Investigation of Multimodality’, in N.-M. Klug and H. Stöckl (eds), Sprache im Multimodalen Kontext/Language and Multimodality, 36–74, Berlin: De Gruyter Mouton. Curtin University (2018), ‘Analysing Violent Extremist Communications and Their Uptake in the Media’, Curtin Institute for Computation. Available online: https:// computation.curtin.edu.au/research/case-studies/analysing-violent-extremistcommunications-uptake-media/ (accessed 30 October 2021). Curtin University (2020), ‘Computation at Curtin’, Curtin Institute for Computation. Available online: https://computation.curtin.edu.au/ (accessed 30 October 2021). Drok, K. (2018), ‘Getting Clever with Big Data to Understand the Spread of ISIS Propaganda’, Curtin University News and Events. Available online: https://news. curtin.edu.au/stories/getting-clever-big-data-understand-spread-isis-propaganda/ (accessed 31 December 2020). Google (2020), ‘Vision AI’. Available online: https://cloud.google.com/vision/ (accessed 31 December 2020). Halliday, M. A. K. (2008), ‘Working with Meaning: Towards an Appliable Linguistics’, in J. J. Webster (ed.), Meaning in Context: Strategies for Implementing Intelligent Applications of Language Studies, 7–23, London: Continuum. Halliday, M. A. K. and C. M. I. M. Matthiessen (2004), An Introduction to Functional Grammar, 3rd edn, London: Arnold. Halliday, M. A. K. and C. M. I. M. Matthiessen (2014), Halliday’s Introduction to Functional Grammar, 4th edn, London: Routledge. Huang, Z. (1997), ‘Clustering Large Data Sets with Mixed Numeric and Categorical Values’, in H. Motoda and H. Lu (eds), Proceedings of the First Pacific Asia Knowledge Discovery and Data Mining Conference, Singapore, 21–34, Singapore: World Scientific Publishing Co Pte Ltd. Huang, Z. (1998), ‘Extensions to the K-modes Algorithm for Clustering Large Data Sets with Categorical Values’, Data Mining and Knowledge Discovery, 2 (3): 283–304. IAB (2020), ‘Content Taxonomy’. Available online: https://www.iab.com/guidelines/ content-taxonomy/ (accessed 31 December 2020). IBM (2020), ‘Watson Visual Recognition’. Available online: https://www.ibm.com/ cloud/watson-visual-recognition (accessed 31 December 2020). IBM Watson (2020), ‘Natural Language Understanding’. Available online: https:// natural-language-understanding-demo.ng.bluemix.net/ (accessed 31 December 2020). Kress, G. and T. van Leeuwen (2020), Reading Images: The Grammar of Visual Design, 3rd edn, London: Routledge. Mahboob, A. and N. Knight eds (2010), Appliable Linguistics, London: Continuum. Martin, J. R. and D. Rose (2007), Working with Discourse: Meaning beyond the Clause, 2nd edn, London: Continuum. Martin, J. R. and P. R. R. White (2005), The Language of Evaluation: Appraisal in English, London: Palgrave Macmillan. O’Halloran, K. L. (2015), ‘Multimodal Digital Humanities’, in P. Trifonas (ed.), International Handbook of Semiotics, 383–409, Dordrecht: Springer. O’Halloran, K. L., S. Tan, P. Wignell and R. Lange (2017), ‘Multimodal Recontextualisations of Images in Violent Extremist Discourse’, in S. Zhao,

178

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

E. Djonov, A. Björkvall and M. Boeriis (eds), Advancing Multimodal and Critical Discourse Studies: Interdisciplinary Research Inspired by Theo Van Leeuwen’s Social Semiotics, 181–202, New York: Routledge. O’Halloran, K. L., S. Tan, D.-S. Pham, J. Bateman and A. Vande Moere (2018), ‘A Digital Mixed Methods Research Design: Integrating Multimodal Analysis with Data Mining and Information Visualization for Big Data Analytics’, Journal of Mixed Methods Research, 12 (1): 11–30. doi:10.1177/1558689816651015 O’Halloran, K. L., S. Tan, P. Wignell, J. Bateman, D.-S. Pham, M. Grossman and A. Vande Moere (2019), ‘Interpreting Text and Image Relations in Violent Extremist Discourse: A Mixed Methods Approach for Big Data Analytics’, Terrorism and Political Violence, 31 (3): 454–74. doi:10.1080/09546553.2016.1233871 O’Halloran, K. L., S. Tan, P. Wignell, R. Wang, K. Chai and R. Lange (2020), ‘Towards a Discipline of Multimodality: Parallels to Mathematics and Linguistics and New Ways Forward’, in J. Wildfeuer, J. Pflaeging, J. Bateman, O. Seizov and C.-I. Tseng (eds), Multimodality: Disciplinary Thoughts and the Challenge of Diversity, 93–112, Berlin: De Gruyter. O’Halloran, K. L., G. Pal and M. Jin (2021), ‘Multimodal Approach to Analysing Big Social and News Media Data’, Discourse, Context and Media, (40). doi:https://doi. org/10.1016/j.dcm.2021.100467 O’Toole, M. (2011), The Language of Displayed Art, 2nd edn, London: Routledge. Stanford University (2020), ‘DenseCap: Fully Convolutional Localization Networks for Dense Captioning’, Stanford Computer Science. Available online: https:// cs.stanford.edu/people/karpathy/densecap/ (accessed 31 December 2020). Tan, S., K. L. O’Halloran, P. Wignell, K. Chai and R. Lange (2018), ‘A Multimodal Mixed Methods Approach for Examining Recontextualisation Patterns of Violent Extremist Images in Online Media’, Discourse, Context and Media, 21: 18–35. doi:10.1016/j.dcm.2017.11.004 Tan, S., K. L. O’Halloran, P. Wignell and K. Lobinger (2020), ‘Images of Austerity in the British Press and in Online Media’, in T. Griebel, S. Evert and P. Heinrich (eds), Multimodal Approaches to Media Discourses: Reconstructing the Age of Austerity in the United Kingdom, 134–62, London: Routledge. Trevino, A. (2016), ‘Introduction to K-Means Clustering’, Oracle AI & Data Science Blog, 6 December. Available online: https://blogs.oracle.com/datascience/ introduction-to-k-means-clustering (accessed 31 December 2020). uClassify (2020a), ‘IAB Taxonomy V2’. Available online: https://uclassify.com/browse/ uclassify/iab-taxonomy-v2 (accessed 31 December 2020). uClassify (2020b), ‘Tonality’. Available online: https://www.uclassify.com/browse/ prfekt/tonality (accessed 31 December 2020). van Leeuwen, T. (2005), Introducing Social Semiotics, London: Routledge. Wignell, P., S. Tan and K. L. O’Halloran (2017a), ‘Under the Shade of AK47s: A Multimodal Approach to Violent Extremist Recruitment Strategies for Foreign Fighters’, Critical Studies on Terrorism, 10 (3): 429–52. doi:10.1080/17539153.20 17.1319319 Wignell, P., S. Tan and K. L. O’Halloran (2017b), ‘Violent Extremism and Iconisation: Commanding Good and Forbidding Evil?’, Critical Discourse Studies, 14 (1): 1–22. doi:10.1080/17405904.2016.1250652 Wignell, P., S. Tan, K. L. O’Halloran and R. Lange (2017), ‘A Mixed Methods Empirical Examination of Changes in Emphasis and Style in the Extremist Magazines Dabiq and Rumiyah’, Perspectives on Terrorism, 11 (2): 2–20. Available

BIG DATA AND MULTIMODAL COMPLEXITY

179

online: http://www.terrorismanalysts.com/pt/index.php/pot/article/view/592 (accessed 31 December 2020). Wignell, P., K. Chai, S. Tan, K. L. O’Halloran and R. Lange (2018), ‘Natural Language Understanding and Multimodal Discourse Analysis for Interpreting Extremist Communications and the Re-Use of these Materials Online’, Terrorism and Political Violence, advance online publication. doi:10.1080/09546553.2018. 1520703 Wignell, P., K. L. O’Halloran, S. Tan, R. Lange and K. Chai (2018), ‘Image and Text Relations in ISIS Materials and the New Relations Established through Recontextualisation in Online Media’, Discourse & Communication, 12 (5): 535–59. doi:10.1177/1750481318766938 Wignell, P., S. Tan, K. L. O’Halloran, R. Lange, K. Chai and M. Wiebrands (2020), ‘Images as Ideology in Terrorist-Related Communications’, in H. Stöckl, H. Caple and J. Pflaeging (eds), Shifts Towards Image-Centric Practices in the Contemporary Multimodal Practices, 253–74, London: Routledge.

180

CHAPTER TEN

Our Own Double Helix The Power of Teachers and Students Understanding How Language Works in Schooling Contexts BRIAN DARE AND JOHN POLIAS

In this chapter, we chart the moves and shifts, the manoeuvrings and breakthroughs, the hindrances and support we experienced over the last thirty years as educator linguists working with Systemic Functional Theory. In doing so, we attempt to provide some insight into how we have arrived at giving some level of power to individual students and teachers through an understanding of language’s role in teaching and learning. It is noteworthy that, after all those years working with teachers, we still have the same passion for a pedagogy underpinned by a functional model of language. In our former lives as English as a second language (ESL) consultants, we were constantly confronted with the key issue for ESL students: how to speed up their language development so that they could achieve their potential in the same way as their peers. From the outset, we felt that we had found an answer in the work emerging from the ‘Sydney School’ and the Language and Social Power Project in particular (Walsh et al. 1990; Cope and Kalantzis 1993). From those early days, we have remained committed to an explicit language-based pedagogy that provides the means for teachers and students alike to build shared understandings about language, about how it works to make meaning. While it has been a productive journey, it has not unfolded without its tensions and challenges. From the early 1990s, when we first engaged with this

182

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

radical thinking about language and its role in teaching and learning, our major challenge has been to find ways to make such an elaborate and complex model of language accessible to teachers. At the same time, we have been acutely aware of the constant demand to show the efficacy of a pedagogy underpinned by students and teachers having a shared metalanguage. Along the way, particularly in the 1990s in the state of South Australia, we encountered resistance from many corners of the educational community. Despite the scepticism (and at times the hostility), the thirst teachers had for such knowledge and the benefits that teachers could see for their students as they took on this radical approach were, and remain, a constant source of inspiration to continue with the work.

LOOKING BACK TO THE EARLY DAYS When we first met in the early 1990s as consultants working in the ESL field, there was no national curriculum in Australia. Our South Australian state curriculum, like those of other states, had very little to say about the need for teachers or students to have any explicit knowledge about language. As Joan Rothery wrote around that time: ‘The language system has completely disappeared from view in schooling and in most Australian pre-service teacher education courses. So successful has this effacement of language knowledge been that students and educators alike have no idea what might be known about language. The invisibility of language has itself become invisible’ (1996: 86). So not only was knowledge about language invisible but also the means by which that invisibility could be redressed. Pre-service education was itself, with a few exceptions, showing no capacity to improve the situation. The dominant pedagogical approach at that time has been described as ‘postmodernist progressivism’, with its emphasis on a ‘whole-language approach’ and a focus on ‘process’ writing (Cope and Kalantzis 1993; Rose and Martin 2012). Mainstream orthodoxy by and large dismissed the need for any kind of grammatical knowledge, which was not entirely unreasonable given that traditional grammar approaches had shown little promise for education contexts. In addition, since the ESL field in Australia had taken on the communicative competency framework (Burns and De Silva Joyce 2007), there was also considerable resistance to building any real focus on teaching about language, in particular grammar – especially for students who were at the very early stages of learning English. The early impact of the ‘Sydney School’ The early 1990s in Australia saw a burgeoning interest in Halliday’s functional model of language and Martin and Rothery’s genre theory emerging from the ‘Sydney School’ (Cope and Kalantzis 1993; Martin 1999;

OUR OWN DOUBLE HELIX

183

Rose and Martin 2012). In our roles as consultants, we were fortunate to be able to engage with Systemic Functional Linguistics (SFL) and genre theory through SFL-related conferences and training workshops as well as through the SFL literature and resources emerging from the Language and Social Project (Macken et al. 1989; Macken 1989; Christie et al. 1990; Martin 1999). This, coupled with discussions between ourselves and like-minded educators, convinced us that there were enormous benefits both for teachers and for students in having a shared explicit metalanguage – a metalanguage that would extend beyond genre and schematic structure, and include the language system at the lexicogrammar and discourse semantic levels at the very least. One of the interesting and challenging features of our South Australia context at that time was our reliance on the eastern Australian states for the intellectual impetus. South Australia was the systemics-free home of progressivism, having taken up a whole-language pedagogy with greater enthusiasm than most (Cope and Kalantzis 1993). With no support coming from any of the universities, we needed to find our own way of building crucial understandings across the population of South Australian teachers.

WHAT DID WE DO INITIALLY? Promotion through early conferences and professional development sessions The early interest in genre pedagogy was given a significant boost in South Australia in 1992 through John Walsh. Having just completed his role as the principal research officer evaluating the Language and Social Power Project (Cope et al. 1993: 242–4), Walsh was appointed to the position of ESL Curriculum Officer within the Department of Education and Children’s Services (DECS), serendipitously with John Polias as his line manager. Using his contacts in the Language and Social Power Project, Walsh quickly established connections with Brian Dare, an ESL consultant in Catholic education, and drawing on his contacts in the Language and Social Project, South Australia’s first genre conference was organized. All ninety places were subsequently filled, and on the day of the conference some forty non-payers attended to hear Cate Poynton, Clare Painter and Mike Callaghan share their insights into the theory and its application in educational contexts. In 1993, Walsh and Dare organized a much more ambitious conference aimed squarely at teachers and teacher educators. The speakers at the conference, held at the University of South Australia, included Jim Martin, Robert Veel, Pauline Gibbons, Mary Macken-Horarik and Maree Stenglin, among many others. Attracting over 300 participants, it was seen as an outstanding success and was pivotal in creating momentum for the work

184

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

in South Australia. This was followed later in the year by the Australian Systemic Functional Linguistics Association’s (ASFLA) Summer School and Conference, which gave many South Australians the chance to extend their theoretical knowledge. In early 1994, Walsh returned to Sydney with the parting advice to Dare and Polias: ‘You two have to work together.’ And so began a rewarding and productive professional partnership that endures to the present day. Our early ambitions were to extend the theoretical knowledge base of ESL consultants and teachers alike and to build up a critical mass that could carry the work forward and build in some long-term sustainability. To this end, we repeated the summer school model from those early years. However, we saw early on that we would have to take a different approach, one that would allow for greater accessibility to the theory and at the same time connect much more directly to teachers’ classroom contexts. Further influences from the eastern states The plenary presentations at the 1994 Brisbane ASFLA conference from Geoff Williams, Joan Rothery and Ruth French were particularly inspiring for us. There, we encountered some convincing research that showed us what is possible in teaching young children about language, where such teaching was mediated by a linguistically informed teacher. The success the students enjoyed and the joy they experienced engaging in what was often denigrated as ‘boring grammar’ dispelled any doubts we had about both its efficacy and the eagerness for such metalinguistic knowledge on the part of students. In 1997, we brought French to Adelaide as a ‘language educator in residence’, and over a number of days she worked closely with key teachers and consultants. Everyone highly valued her ability to ‘ground’ the theory in practice and appreciated the way she presented the teaching of the grammar as part of a systematic and comprehensive pedagogy in which language took a central focus.

DEVELOPING THE LANGUAGE AND LITERACY COURSE: CLASSROOM APPLICATIONS OF FUNCTIONAL GRAMMAR As a result of our experience working closely with teachers over the previous years, we grew confident that we had developed the basis of an effective language-based pedagogy that would benefit all students. Mainstream teachers were telling us as much, and so we began to think about ways we could take this beyond the margins of the ESL field to the wider literacy field. Of course, doing

OUR OWN DOUBLE HELIX

185

the work under the ‘guise’ of ESL also allowed us to get on with the work in the face of disinterested hostility from the mainstream literacy field, which was at the time essentially comprised of teachers of English literature. Drawing on the support we had from our ESL line managers, we started to imagine what this work could look like for all teachers across schooling and across all learning areas – well beyond the confines of ESL. Moving on to teacher development: High challenge and high support In taking this ambitious view, we knew we had to come up with a more comprehensive and more systematic way of inducting teachers into this work. Drawing on ESL in the Mainstream (Burke et al. 1991), a very successful mainstream professional course developed by DECS’s ESL Program, revised as Teaching ESL Students in Mainstream Classrooms (Custance, Dare and Polias 2006), we started to build up and deliver a series of professional development sessions around the functional model. These sessions became the basis of the first iterations of what would become known as the Language and Literacy course (Dare and Polias 2000). Like the ESL in the Mainstream course, it involved teachers attending a number of input sessions and undertaking practical activities and readings in between the sessions to consolidate and extend their understandings. One of the prominent drivers of the first iteration of the course was the need to maintain high expectations of what it was possible for teachers to know about language. While ever conscious of the complexity of a linguistic theory ‘oriented to extravagance, rather than parsimony’ (Halliday and Martin 1993: 23), we did not back away from providing a rigorous and challenging course that would do justice to the functional model. We wanted to ensure that teachers had a comprehensive overview of the whole model and its broad theoretical framework, and at the same time open up the possibilities for classroom application in their own contexts. We felt that having a more expansive view of the model of language would ensure that teachers taking on ‘a genre approach’ would be focusing on more than just the purpose and schematic structure of any given genre. It became increasingly obvious that, without any explicit teaching about how language works to make meaning both at the grammar level and the discourse semantic level, the pedagogy would be bereft of one of its most powerful mediating resources, that is, knowledge about language. In particular, we felt the lexicogrammar provided a rich affordance that we could not ignore. As Halliday (2005: 59) put it: ‘The powerhouse of a language is its lexicogrammar, the unified stratum of syntax and vocabulary; so thinking about meaning means thinking grammatically.’

186

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

The design of the course Working along the same ‘train-the-trainer’ model as the successful Teaching ESL Students in Mainstream Classrooms course noted above, the Language and Literacy course consisted of 10 three-hour modules structured around SFL’s stratified model of language. The first two modules introduced participants to the notion of genre and register. Previously, little attention had been given to the register stratum in education and we were keen to rectify this situation. MackenHorarik’s (1996) framework provided the basis for our early attempts to deepen our teachers’ understandings of register. In terms of the language system, we took a metafunctional approach, beginning with the experiential. There, we introduced the functional groupings of participant, process and circumstance, borrowing French’s (2013) colour-coding of red for participants, green for processes and blue for circumstances. We then focused on each functional category in turn, which allowed us to explore different process types, alongside patterns in participants and patterns in circumstances. From there, we investigated the nominal group, beginning with its relationship to participants and circumstances, then considered its structure, followed by an exploration of the patterns in the nominal group across different genres and registers. We rounded off with a look at nominalization and its role in spoken and written language. Subsequent modules dealt with clause complexes, so we introduced conjunctions using Martin’s (1992) categorization of linking, binding, cohesive and rhetorical conjunctions and considered different types of dependency relationships that can exist between clauses. In these middle modules, we also looked at other cohesive elements such as the reference system and lexical cohesion (Halliday and Hasan 1976). Moving to the textual metafunction, we considered theme and rheme patterns across a range of genres. Borrowing the notion of ‘orientation’ used to introduce theme, we saw how this could be applied to the paragraph and whole-text level through hypertheme and macrotheme, respectively (Martin 1992). Modules 8 and 9 focused on the interpersonal metafunction, with clause as exchange the starting point for understanding the four major speech functions and how these are realized through typical mood systems. Modality was a must for teachers and one area that they immediately recognized as an area that students struggled with, typically expressing their meanings in absolute positive or negative terms rather than in more nuanced and tempered ways. We limited our discussion of evaluation at this stage since the appraisal system was just emerging in the field. We also took up the notion of interpersonal metaphor and how that might play out in various contexts within and beyond the classroom. The final module consisted of participants considering how they might incorporate all the learnings built up through the course into their own teaching

OUR OWN DOUBLE HELIX

187

practice. Thus, the focus was on the teaching and learning cycle (Martin 1999) as a macro-scaffold for scaffolding language and learning in a unit of work. Once we had the course set out, we began to offer it in two modes – as a five-day course during holiday periods and as a more extended course of one module per week over ten weeks during school terms. There was considerable enthusiasm for both modes, and for a while we had three courses running each week, with participants ranging across the school sectors and year levels. On some occasions we had teachers from initial reception levels through to university lecturers in the same group. Such was the popularity of the course that the ESL Program within DECS created a new position called Language and Literacy Project Training Officer, and appointed John Polias. This reflected the growing commitment of the ESL field across both the state and Catholic sectors to our language-based approach. Moving on While the Language and Literacy course was the obvious professional development vehicle for introducing SFL to South Australia, there were, in the mid-1990s, various policy and curriculum projects at the time that were influencing the field. One of the specific curriculum materials was a science curriculum for intensive English language programmes for newly arrived students with no or low levels of English. The materials, with Polias as the major writer, were published as Teaching ESL Through Science: A Science Curriculum for Intensive English Language Programs (Hiles et al. 1998). The final product was essentially a manual with science instructions accompanied by descriptions of the language resources needed to participate in learning activities. These language descriptions and the activities in the worksheets were informed by the notions of register continua and genre continua (Derewianka 1990; MackenHorarik 1996). These materials and other classroom applications of SFL continually informed our professional development with teachers, with the result that our courses changed their orientation to pedagogical applications of working with language, and this was evident in their titles. For example, the Language and Literacy course became the How Language Works: Success in Literacy and Learning course (Custance, Dare and Polias 2011).

TRANSFORMING PRACTICE: RESEARCH EVIDENCE With growing numbers attending the Language and Literacy course, there was an imperative to demonstrate that this approach was not only possible but in fact made a difference to student outcomes. Working together with teachers, we wanted to encourage spaces for them to reflect on their work and to establish the efficacy of this approach in their contexts. Initially, teachers

188

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

worked in informal, loosely connected networks and shared ways of teaching about language with each other. These were exciting times as teachers across all schooling contexts started to appreciate what was possible to teach about language and, importantly, the benefits of taking such an explicit approach. With time, many of these teachers then joined more formal research projects set up with funding – initially from the National Language and Literacy Institute of Australia and later with funding for two projects from the Spencer Foundation. These projects provided the vehicle for dismissing misconceptions about the explicit teaching of grammar in particular (Polias and Dare 2006). Nineteen of our teachers presented very well-received papers at the 1998 ASFLA conference held in Adelaide, showcasing the outstanding outcomes achieved with their students. Some, such as the transforming work undertaken by Donna Riethmuller and Louise Ferris at St Margaret Mary’s Croydon (Polias and Dare 2006; Derewianka 2012) and Jan Finlayson (Dare and Polias 2001), pushed the boundaries of what is possible to teach even very young children. Both of these contexts involved classrooms with very high proportions of ESL students. We were encouraged to see such positive outcomes emerging from these challenging contexts. If we consider the work done on writing sequential explanations with sevenyear-old students by Riethmuller and Ferris, we can see in the following beforeand-after texts the kind of improvements made over the space of ten weeks of classroom work. Student A’s text prior to the ten-week period is used as the baseline: How milk gets from the cow to us The sun makes the grass grow then the cows eat the grass. After that the cows make milk in the/her body. Next the farmer comes out and milks the cows. Finerly the truck comes to pick up the milk and puts it in the factory. And the factory cleans the milk and puts the milk in the botels then they put the milk in the shop and we buy the milk to drink. The following is Student A’s draft text after the ten-week period: The cow eats grass. The farmer uses suction cups to milk the cow. The raw milk which is milk straight from the cow flows into the vats. The tanker comes and the milk flows to the tanker. The tanker driver tests the milk. The tanker deliveres raw milk to the factory every day. At the factory the raw milk is tested. The tanker is washed inside and outside. The raw milk is now pasteurised which is heating the milk up and homogenized which is spreding the cream. The milk is flavoud. The milk is placed into cartons. The cartons

OUR OWN DOUBLE HELIX

189

are placed into crats and puted into the cold room. Folk-lifs cary the crats to the refrigerated delivery trucks. The truck delivers milk to the supamaket and the deli. This improvement was an outcome of an explicit approach to developing the language of sequential explanations: transitivity (processes, participants and circumstances), active and passive voice, theme and rheme, and interrupting clauses. Comments from Riethmuller and Ferris after this work were, for example: ‘students are now very good at identifying participants as a whole nominal group – they can see how language works in chunks’, ‘we can challenge the children to see aspects of functional grammar in all the reading and writing they encounter’, ‘it is valuable to listen to how easily children can discuss language using the metalanguage and solve challenges set for them’ and ‘it is good to see the confidence that all children developed when they became “expert” in the field knowledge and language used’. While there were many examples of individual teachers enjoying significant gains with their students, we began advocating for a whole-school approach to bring about greater change for more students. This, of course, would demand a greater commitment from schools but would also bring much more enduring benefit in terms of both student outcomes and long-term capacity building within the school. With this increasing demand on resources came an increasing demand to show demonstrable improvement in learning outcomes, where schools put significant resources into whole-school approaches. Over time, a number of key studies have shown that, when such an explicit languagebased approach is taken up with fidelity in all classrooms, significant gains, even in the most challenging of teaching contexts, can be made. One of the earliest sets of data to emerge was from the Islington Borough in the UK, a borough which reflected all the usual indicators of socio-economic disadvantage. As one of the earliest boroughs in that country to implement the Teaching ESL Students in Mainstream Classrooms course with some rigour across a number of schools, it was also one of the first to come up with some compelling long-term evidence that such a language-based approach made a significant difference to learning outcomes for ESL students. One of their research findings was that over a seven-year period from 2006 to 2013 the increase in achievement1 of English as an additional language students in Islington Borough was double the national average, with students moving from 7 points below the national average to 7 points above in that period. Further success with the ESL in the Mainstream course, which was being referred to in the UK as LiLaC (Language in Learning across the Curriculum), emerged from work by the same team from Islington but in the neighbouring borough of Enfield. Its final evaluation report, which was commissioned by the London Schools Excellence Fund (2015), concluded that the training

190

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

undertaken had been successful and should be delivered more widely to support teacher development. More recent qualitative and quantitative research undertaken at a number of schools in Victoria, Australia (Pomagalska 2018, 2020), has clearly shown that where a whole school undertakes an explicit language-based approach that is underpinned by a committed building of knowledge of both the functional model of language and an explicit teaching and learning cycle, then a school can expect significant gains in students’ literacy outcomes. At the Southern Secondary College,2 for example, the school’s mean scaled score3 for writing in the National Assessment Program – Literacy and Numeracy (NAPLAN) shifted from some 48 points below the state mean in 2013 to 42 points above the state mean by 2016. This research captures not only these remarkable improvements in student literacy outcomes but also the significant impacts on the teachers and the whole school community.

ESL SCOPE AND SCALES In 2000, the state of South Australia embarked on developing a new curriculum and assessment tool called the South Australian Curriculum Standards and Accountability framework (SACSA). As part of this new curriculum, the ESL programme was able to argue for developing a resource that gave ESL teachers a statement of what they could teach and also a detailed description of the language needed to engage with the wider school curriculum. This resource became known as the ESL Scope and Scales (Polias 2003; Polias and Dare 2006). It was organized using two main organizers: text in context and language. These were further divided according to genre and register. With its detailed language descriptions across a maximum of fourteen grade levels or scales, the ESL Scales was unique in providing a linguistic description of learning in schooling contexts and thus providing teachers with an understanding of the language needed for success in schooling. In the absence of a literacy continuum in the SACSA, it also became a de facto literacy continuum for non-ESL teachers. This was a major outcome of the ESL Scales – namely the idea that learning and language development were so tied together that both were the concern of all teachers, whether ESL specialists or not.

LEXIS EDUCATION In 2002, we formed Lexis Education and began writing and delivering courses domestically and internationally for DECS. The various courses run by Lexis Education, using the train-the-trainer model, have been, and are being, delivered by over 4,000 tutors in Australia and many countries of the world

OUR OWN DOUBLE HELIX

191

with the major locations being in the UK, the Middle East, Europe, China and South-East Asia. In Anglophone countries, it includes both government and international schools, while in other countries, it is generally international schools. Initially, Lexis Education trained tutors for the ESL in the Mainstream course, since this training had attracted considerable interest in the UK and from the Education Bureau in Hong Kong. Later, tutors were trained in the Teaching ESL Students in Mainstream Classrooms: Language in Learning across the Curriculum course (TESMC) (Dare and Polias 2006) and the Language and Literacy course (LLC) (Dare and Polias 2000). In the UK, various Local Education Authorities, especially in the London boroughs of Hounslow, Islington, Newham and Lewisham, have participated in these courses and together with a large number of international schools, such as the International School of Brussels and the Chinese campuses of Dulwich College, over 600 tutors have been trained since 1998. Between 2000 and 2009 with the Education Bureau in Hong Kong, a total of over 400 teachers in 110 of its 114 secondary schools that used English as the medium of instruction were trained as tutors in the three courses and over 3,000 teachers attended the subsequent teacher courses. In 2006, the ESL in the Mainstream course’s functional underpinnings were significantly strengthened and it was re-badged as TESMC. It enjoyed particular success in the UK, where Paddy Walsh promoted the course and the SFL work with great enthusiasm (Walsh 2006). In Australia, the worldwide success of the ESL in the Mainstream course led DECS to publish and market LLC and it enjoyed a very strong take-up in Queensland, for example, thanks to the support of Christine Ludwig and her like-minded colleagues within the Queensland Department of Education. Ultimately, 114 tutors were trained and at least 1,300 educators completed the course in Queensland. Building on the increasing demand for the courses and drawing on our continuing work with teachers, we updated LLC to the How Language Works: Success in Literacy and Learning course (Custance et al. 2011), a move which we saw as strengthening the classroom applications by looking at genre continua and the register continuum and taking up a stronger focus on discourse semantics by incorporating appraisal analysis. Radical shifts at the national level By 2010, things had changed considerably in Australia with the introduction of a new national English curriculum that took ‘a fairly radical step in placing knowledge about language at the core of classroom practice, thereby raising the issue of an appropriate model of language to inform the Language Strand

192

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

of the Curriculum’ (Derewianka 2012). Its functional orientation was clearly evidenced by the three sub-strands, Language for interaction, Text structure and organisation and Expressing and developing ideas – which are commonsense glosses of Halliday’s metafunctions (Halliday and Matthiessen 2014). The new national curriculum also took the position that all teachers are teachers of literacy. It shifted this responsibility away from English teachers and put pressure on learning area teachers in particular to both understand what literacy meant and how they would go about ‘teaching literacy’. Given the wholesale effacement of knowledge about language over many decades, this meant that these teachers would need considerable support if this mandate were to be taken seriously. Even for English teachers, the new curriculum with its functional underpinnings would challenge their conception of what a practical metalanguage would look like. In response to this growing national discourse around the importance of literacy, we were commissioned by DECS to write what became known as the Literacy for Learning course. We had to juggle a number of constraints, not the least of which was to make ‘grammar knowledge’ and ‘text knowledge’ palatable and accessible to all teachers. The uptake of the course, especially in Australia, suggests that it answered a real need. Initially the course was delivered to large numbers of teachers in South Australia with some 500 tutors trained and over 4,000 teachers attending the teacher course, before the department cut back on its delivery. On a more positive note, it has been more than heartening to see the uptake of our courses across Victoria in recent years. Given that Victoria, like South Australia, was one of the pillars of progressivist pedagogies (Cope and Kalantzis 1993), it has been thrilling to see over 200 Literacy for Learning tutors and some 400 tutors in all carrying the work forward in Victoria.

CONCLUSION In 2017, Lexis Education became sole proprietor of the content and intellectual property of the courses. With the publication of a new course, Teaching Young Children in English in Multilingual Contexts (Dare and Polias 2015), which is aimed at teachers of five- to eight-year-old students, and renewals of the TESMC and LLC courses to now be, respectively, Teaching in English in Multilingual Contexts: Language in Learning across the Curriculum and 3L: Language and Literacy for Learning, we are able to offer a suite of courses – all of which are underpinned by our confidence in an explicit pedagogy based on the functional model of language. With over 5,000 tutors worldwide, covering our four courses, we can safely assume that many educators across the globe share that belief. On reflection, we can see that the departure point for our work is the building up of meaning-making resources for both teachers and students so

OUR OWN DOUBLE HELIX

193

that both can draw on these semiotic tools to become self-reliant and powerful language users, whether as teacher or student. From the very early days we put faith in the train-the-trainer model (Suhrheinrich 2011) to build this capacity within schools rather than have it residing with us, or with academics or other ‘experts’, and to build capacity in a way that speeds up the process. Another guiding principle has been to attempt to enact whole-school change (Luke 2003). This requires a serious commitment by schools in terms of human and financial resources, but we have not resiled from this principle, even in the face of alternative approaches that promise maximum results with minimum input. For us, whole-school change requires setting up and maintaining a consistent, systematic approach to teaching and learning that is underpinned by a functional model of language. We are now beginning to see the very positive outcomes of a whole-school language-based approach in a number of sites in Australia (Pomagalska 2018, 2020) and the UK (Forey and Cheung 2019; Forey 2020). As we have charted above, there have been times when we have met resistance from people at all levels of decision-making for varied reasons. But teachers and schools continue to access our courses and our related work because, as we have seen over and over again, once teachers and students have the language resources they need to make meaning in an ever-increasing range of contexts, then the work prevails.

NOTES 1 The baseline was measured by the number of students achieving five A*–C grades including English and maths at the General Certificate of Secondary Education examinations. 2 Not the actual name of the school. 3 The mean scaled score is the mean score achieved by a cohort of students and then adjusted (scaled) so that any given score represents the same level of achievement. This allows valid comparisons to be made (i.e. the Growth Date) regarding a student’s level of achievement over time.

REFERENCES Burke, D., A. Kay and D. Rees (1991), ESL in the Mainstream, Adelaide, SA: Department of Education and Children’s Services. Burns, A. and J. De Silva Joyce (2007), ‘Adult ESL Programs in Australia’, Prospect, 22 (3): 5–17. Christie, F., B. Gray, P. Gray, M. Macken, J. R. Martin and J. Rothery (1990), Exploring Reports: Teachers Book: Levels 1–4, Sydney: Harcourt Brace Jovanovich. Cope, B. and M. Kalantzis, eds (1993), The Powers of Literacy: A Genre Approach to Teaching Writing, London: Falmer Press.

194

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Cope, B., M. Kalantzis, G. Kress, J. R. Martin and L. Murphy (1993), ‘Bibliographical Essay: Developing the Theory and Practice of Genre-based Literacy’, in B. Cope and M. Kalantzis (eds), The Powers of Literacy: A Genre Approach to Teaching Writing, 242–4, London: Falmer Press. Custance, B., B. Dare and J. Polias (2006), Teaching ESL Students in Mainstream Classrooms: Language in Learning across the Curriculum, Adelaide, SA: Department of Education and Children’s Services. Custance, B., B. Dare and J. Polias (2011), How Language Works: Success in Literacy and Learning, Melbourne: Lexis Education. Dare, B. and J. Polias (2000), Language and Literacy: Classroom Applications of Functional Grammar, Adelaide, SA: Department of Education and Children’s Services. Dare, B. and J. Polias (2001), ‘Learning about Language: Scaffolding in ESL Classrooms’, in J. Hammond (ed.), Scaffolding Teaching and Learning in Language and Literacy Education, 91–110, Sydney: Primary English Teachers Association. Dare, B. and J. Polias (2006), Teaching ESL Students in Mainstream Classrooms: Language in Learning across the Curriculum, Adelaide, SA: Department for Education and Child Development. Dare, B. and J. Polias (2015), Teaching Young Children in English in Multilingual Contexts, Adelaide, SA: Department for Education and Child Development. Derewianka, B. (1990), ‘Rocks in the Head: Children and the Language of Geology’, in R. Carter (ed.), Knowledge about Language and the Curriculum: The LINC Reader, 197–215, Kent, UK: Hodder and Stoughton. Derewianka, B. (2012), ‘Knowledge about Language in the Australian Curriculum: English’, Australian Journal of Language and Literacy, 35 (2): 127–46. Forey, G. (2020), ‘A Whole School Approach to SFL Metalanguage and the Explicit Teaching of Language for Curriculum Learning’, Journal of English for Academic Purposes, 44: art 100822. Forey, G., and L. M. E. Cheung (2019), ‘The Benefits of Explicit Teaching of Language for Curriculum Learning in the Physical Education Classroom’, English for Specific Purposes, 54: 91–109. French, R. (2013), ‘Teaching and Learning Functional Grammar in Junior Primary Classrooms’, PhD thesis, University of New England, Armidale, NSW. Halliday, M. A. K. (2005), ‘On Matter and Meaning: The Two Realms of Human Experience’, Linguistics and the Human Sciences, 1 (1): 59–82. Halliday, M. A. K. and J. R. Martin (1993), Writing Science: Literacy and Discursive Power, London: Falmer Press. Halliday, M. A. K. and R. Hasan (1976), Cohesion in English, London: Longman. Halliday, M. A. K. and C. M. I. M. Matthiessen (2014), Halliday’s Introduction to Functional Grammar, 4th edn, London: Hodder Arnold. Hiles, B., S. Pagonis, P. Lee and J. Polias (1998), Teaching ESL through Science: A Science Curriculum for Intensive English Language Programs, Adelaide, SA: Department of Education and Children’s Services. London Schools Excellence Fund (2015), Self-Evaluation Toolkit: Final Report, London: London Schools Excellence Fund. Luke, A. (2003), ‘Literacy and the Other: A Sociological Approach to Literacy Research and Policy in Multilingual Societies’, Reading Research Quarterly, 38 (1): 132–41.

OUR OWN DOUBLE HELIX

195

Macken, M. (1989), A Genre-Based Approach to Teaching Writing, Years 3–6: Book 2: Factual Writing, Sydney: Literacy and Education Research Network (LERN), Common Ground. Macken, M., G. Kress, M. Kalantzis, J. R. Martin, J. Rothery and W. Cope (1989), An Approach to Writing – 12: Vol. 1 Introduction, Sydney: Literacy and Education Research Network & Directorate of Studies, NSW Department of Education. Macken-Horarik, M. (1996), ‘Literacy and Learning across the Curriculum: Towards a Model of Register for Secondary School Teachers’, in R. Hasan and G. Williams (eds), Literacy in Society, 232–78, London: Longman. Martin, J. R. (1992), English Text: System and Structure, Amsterdam: John Benjamins. Martin, J. R. (1999), ‘Mentoring Semogenesis: “Genre-Based” Literacy Pedagogy’, in F. Christie (ed.), Pedagogy and the Shaping of Consciousness: Linguistic and Social Processes, 123–55, London: Continuum. Polias, J. (2003), ESL Scope and Scales, Adelaide, SA: Department of Education and Children’s Services. Available online: https://lexised.com/resources/ Polias, J. and B. Dare (2006), ‘Towards a Pedagogical Grammar’, in A. McCabe, M. O’Donnell and R. Whittaker (eds), New Directions in Language and Education, 123–43, London: Continuum. Pomagalska, D. (2018), Evaluation of the Whole-School Implementation of a Language-Based Pedagogy at Southern Secondary College, Melbourne: Lexis Education. Available online: https://lexised.com/report-ssc/ Pomagalska, D. (2020), Outcomes of a Functional Grammar-Based Pedagogy in an Intensive English Language School, Melbourne: Lexis Education. Available online: https://lexised.com/report-wels/ Rose, D. and J. R. Martin (2012), Learning to Write, Reading to Learn: Genre, Knowledge and Pedagogy of the Sydney School, Sheffield, UK: Equinox Publishing. Rothery, J. (1996), ‘Making Changes: Developing an Educational Linguistics’, in R. Hasan and G. Williams (eds), Literacy in Society, 86–123, New York: Addison Wesley Longman. Suhrheinrich, J. (2011), Examining the Effectiveness of a Train-the-Trainer Model: Training Teachers to Use Pivotal Response Training, Washington, DC: Society for Research on Educational Effectiveness. Walsh, J., J. Hammond, G. Brindley and D. Nunan (1990), Metropolitan East Disadvantaged Schools Program: Factual Writing Project Evaluation, Sydney: National Centre for English Language Teaching and Research, Macquarie University. Walsh, P. (2006), ‘The Impact of Genre Theory and Pedagogy and Systemic Functional Linguistics on National Literacy Strategies in the UK’, in R. Whittaker, M. O’Donnell and A. McCabe (eds), Language and Literacy: Functional Approaches, 159–76, London: Continuum.

196

CHAPTER ELEVEN

‘Appliability’ of SFL in Two Academic Contexts in Argentina Developing the Teaching of Academic Genres in Spanish ESTELA INÉS MOYANO

INTRODUCTION This chapter outlines the application of Systemic Functional Linguistics (hereafter SFL) to interrelated practices in Spanish. These practices are educational linguistics, discourse analysis and language description in different strata, all of them applied in two educational contexts at university level, one of them with repercussions in secondary school. By SFL as an ‘appliable’ linguistics, Halliday means ‘a socially accountable linguistics, and this in two distinct though related senses: that it put language in its social context, and at the same time it put linguistics in its social context, as a mode of intervention in critical social practices’ ([1993] 2003: 223, emphasis added). In this chapter, I take the second sense in which SFL as ‘appliable’ linguistics is understood. Halliday (2008: 203) also explains that an ‘appliable’ linguistics is ‘a theory and a body of methods that can be drawn upon in those

198

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

areas of human activity in which language is critically involved’, that is, where people ‘engage consciously with language’. The main practice that is addressed here is educational linguistics in the area of academic discourse, which first of all has the objective of enhancing students’ abilities in reading and writing disciplinary texts to develop their capacity to manage abstract and technical discourse. Second, it is considered here as a practice with the purpose of developing in future teachers the ability to teach this kind of discourse and the genres involved. Educational linguistics in the field of sciences requires not only learning the use of oral and written language – in this case Spanish as scientific and specialized language – but also learning specialized content – that is, scientific content in different disciplines – through language. These practices demand learning about language, reflecting on available choices in different strata and, especially, learning about genres (Halliday [1991] 2007; Rose and Martin 2012). To accomplish these goals, the practice of educational linguistics needs to rely on discourse analysis, especially of scientific and technical texts in Spanish, which are ‘able to play a central part in the creation and transmission of knowledge’ (Halliday 2008: 193) in that language. In turn, discourse analysis in a specific language, such as Spanish, relies on the target language description in different strata and metafunctions. Finally, when the practice of teaching is involved, another knowledge is required, the knowledge of a social activity, that is, as a genre: a pedagogic model. I will not address these topics in this chapter, but it is important to say that they are required research to perform the pedagogic work.

CONTEXTS Two contexts of educational linguistics are addressed here: the first one (hereafter Context I) involves a secondary teaching training degree in language; the second (hereafter Context II), a programme across the university curriculum in various degrees. Context I is an undergraduate programme developed in the Universidad Nacional de General Sarmiento, Argentina, in which two pedagogic subjects are included. One of them, called Enseñanza de la Lengua y la Literatura (Teaching Language and Literature) is oriented to teach future teachers how to develop a pedagogic sequence with two different goals: teaching the use of language in different cultural spaces, including academic literacy, and teaching secondary students to read, analyse and comment on literary texts. In Teaching Language, the students first approach different theories about reading, writing and teaching these practices and then are oriented to learn how to apply a genre-based model informed by SFL in order to develop reading and writing skills of non-literary texts in secondary students. The students also have to

TEACHING ACADEMIC GENRES IN SPANISH

199

study a second undergraduate subject that lasts two semesters, in which they are guided in their first attempts as teachers in what are called ‘teaching practices’. The students have to do these teaching practices for two months in two secondary schools subjects, separately: Language Practices and Literature. I will focus here on the language orientation (as opposed to literature) of both university undergraduate subjects. This context requires the training of future teachers to enable them to help secondary students develop skills for reading and writing texts instantiating journalistic and academic genres. The journalistic genres are those that appear as news stories, chronicles and opinion texts in news publications. The academic genres are those that appear in school textbooks and those that the students have to produce in order to answer the teachers’ questions and learn content in different subjects. For students who are beginning secondary school, these genres are a novelty. Those extracted from the school textbooks and those from popular science journals are especially difficult because they require specialized language recontextualized for the school context (see Bernstein 1990). This language and the structuring of the texts defy students’ capacities in reading and writing, so it is necessary that not only science teachers but also language teachers take the challenge of guiding the students in learning the recontextualized discourse of science and its genres. Context II is an across-the-curriculum programme focused on teaching academic and professional literacy in university degrees at the Universidad de Flores in Buenos Aires, Argentina. The programme, which is called Programa de Lectura y Escritura Académica (PROLEA) (Academic Literacy Program), includes literacy activities conducted by a language academic inside the classes of selected subjects across the curriculum of all the undergraduate degrees. PROLEA, in a process of negotiation with the director of each degree, is included in different subjects, one per year. In this context, a language academic works with the professor of the selected subject in applying the pedagogic model to teach how to produce the written genres used to evaluate the learning of the specialized content and/or professional activities. In order to do this, both academics negotiate the implementation of the different stages of the genre-based teaching model in a number of classes of the subject, under the guidance of the language academic. The role of the subject professor is to attend to the students’ questions related to the content of the discipline when necessary. Context II is concerned with students who need to learn scientific, academic and professional texts, in order to learn the content of different disciplines and professional practices. To do this they need to read and write texts that instantiate the genres of these social activities. These genres are highly specialized, so the students have no previous experience with them. Consequently, they need to be guided in approaching these genres and developing abilities to read and

200

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

write them for different purposes, such as for learning, for assessment and for engaging in scientific, academic and professional practices in the future.

REQUIRED INITIAL INVESTIGATIONS In order to prepare future schoolteachers and university language academics to accomplish these pedagogic activities, a number of requirements were identified at the beginning of a long-term study that was conducted to start solving the problems posed in both contexts. First of all, a genre-based pedagogic model was needed to teach academic literacy. This model had to be appliable to different educational contexts, such as those discussed above and had to have two purposes: to teach one particular genre at a time and to provide the students a procedure to access unknown genres. Second, the pedagogic model needed to focus on the purpose and the structure of the target genres and on the language of the different disciplines in each context of use, in order to provide resources to read and write the genres at stake. These demands imply the necessity to describe the genres in question in both contexts, including the language features used to construct knowledge through them. Finally, it was necessary to find a consistent way to frame the specific pedagogical approach in each context, in order to give principled answers to the questions that arose from them. The theory that emerged as a resource to solve the problems presented by the teaching contexts described was SFL, because it provides a comprehensive description of language as a social semiotic resource – a resource for producing meaning in context (Halliday and Hasan 1985; Halliday [1991] 2007; Matthiessen and Halliday 2009); a theory of genre that suits the contexts and the problems faced (Martin 1992, 1994; Martin and Rose 2008); a description of language systems at discourse-semantic level (Martin 1992; Martin and Rose 2007) and their realization in lexicogrammar (Halliday and Matthiessen 2014); characterizations of the language of different groups of disciplines (Halliday and Martin 1993; Rose 1998; Wignell 1998, 2007) and a genre-based pedagogic proposal to teaching literacy (Martin 1999). At the same time that the approaches were developed, several questions emerged: • Was the pedagogic proposal designed in the frame of SFL suitable for the contexts under consideration? • What genres needed to be described for each context? Which are the relevant genres in the school textbooks of different subjects in secondary schools in Argentina? What academic and professional genres do the undergraduate students of different degrees need to write to be evaluated in their subjects?

TEACHING ACADEMIC GENRES IN SPANISH

201

• What are the key linguistic features of those genres for pedagogical purposes? • Is it possible to apply or adapt descriptions of genres, discoursesemantic and lexicogrammatical resources developed for English to texts in Spanish?

APPLYING SFL TO EDUCATIONAL CONTEXTS IN SPANISH The contexts of application addressed in this chapter demanded several research actions to respond to the questions posed above. In this section I will discuss what aspects of the theory were applied and what new resources were (and still are) needed. A first step was the adaptation of the teaching learning cycle (TLC) outlined by Martin (1999). To do this, it was necessary to apply SFL genre theory to the analysis of texts and identification of genres in Spanish (see examples below). However, as discourse-semantic systems had been described primarily for English but not for Spanish, it was necessary to study if and how they apply to this language (e.g. Oteíza 2010, 2017; Oteíza and Pinuer 2012; Moyano 2016). Another area that needed to be studied was the description of lexicogrammatical resources of Spanish (Quiroz 2015, 2017; Moyano 2016). Finally, a brief description of the language of disciplines in Spanish needed to be conducted. Due to space limitations and the focus of this volume, the detailed linguistic analysis is not included in this chapter. Genres in Contexts I and II Applying the genre-based pedagogic proposal to educational contexts in Argentinian Spanish requires knowledge about the genres to work with. It is necessary to identify the genres involved in each context and to describe them. The relevant genres in Context I are journalistic genres and school-level academic genres. As a first step, the research presented here has been working with the academic genres. To identify them, we draw upon the work done in Australia by Martin and colleagues in genre and register theory (e.g. Martin 1989, 1992, 1993a, 1993b, 1994; Christie and Martin 1997; Martin and Rose 2008). We started the project by collecting texts from popular science journals as well as from textbooks of different disciplines, both sources in Argentinian Spanish. Taking the work cited before as a point of departure, we recognized and analysed genres. The hypothesis underlying this decision is that the practices of conducting and recontextualizing science are largely consistent in Argentina and other post-industrial societies, including Australia, so it is highly probable that the same genres that appear in English-language contexts would

202

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

also appear in the Argentinian Spanish one, although some small variants could appear. What we found in early unpublished work confirmed this assumption. So, we use the analysed texts to teach in the first period of this action-research project in a previous context, a writing course at the beginning of undergraduate studies (Moyano 2007). During this first period, the necessity of introducing the developments in understanding discourse made by Martin (1992) and then by Martin and Rose (2007) as a tool for genre analysis was recognized. So that, in a project financed by the ANPCyT1 (PICTO-MECyT Nº 36533), working with secondary teachers in service, which ultimately feeds Context I, we collected texts from Argentinian school textbooks and analysed them (Moyano 2010a, 2013a; Giudice and Moyano 2011; Giudice 2013; Escobar and Jovenich 2013). The results of this research showed that the genres described in English are present also in Argentinian school textbooks written in Spanish. Some of the texts are macro-genres and some others are hybrid texts. By hybrid texts, we mean texts that accomplish more than one purpose at a time (see e.g. Moyano 2010a). We also found that characteristics of the language of science (e.g. Halliday and Martin 1993; Rose 1998; Wignell 1998, 2007) were recontextualized in these texts (Giudice 2013; Moyano 2013b). Deciding on the genres needed to teach in Context II requires negotiation with the subject professors of each subject with which the programme is associated (Moyano 2010b, 2017, 2018). The professors in charge of each subject decide which genres the students need to produce in order to be evaluated. These include genres called ‘case analysis’, ‘literature review’ and many other academic genres, as well as professional genres found in different degrees (e.g. law, architecture, engineering, psychology). The work done in this context by our group of researchers as well as other work done by the group at the Pontificia Universidad Católica de Chile is emerging (Vidal Lizama, 2020, Vidal Lizama ad Leiva, 2021, Serpa 2021; Acebal 2021; forthcoming; Vidal Lizama and Montes, forthcoming). However, we have found in many cases that these genres received a name from the community of use, but that the texts under the same name may have different structures. Many of them are macro-genres that combine different elementary genres to achieve a global purpose. All of them use the language of the discipline involved. Although the research article is not a genre (properly a macro-genre) taught at the undergraduate level of education, it is worth mentioning the analysis of the discussion section made by Moyano (2013b, 2015, 2019, 2021), which shows the appliability of SFL genre theory and discourse-semantic resources to analysis of scientific-academic texts in Spanish. It is highly relevant to continue this line of research in both Context I and II and to publish as soon as possible the results found at this stage.

TEACHING ACADEMIC GENRES IN SPANISH

203

Adapting the genre-based pedagogic proposal As a first step in the process of applying SFL to the contexts at stake here, the Sydney School’s TLC (Martin 1999) was adapted to be applied in other educational contexts before being applied in the two contexts discussed above (Moyano 2007). The TLC was attractive as a resource for teaching academic literacy for several reasons. To begin, it puts text analysis at the centre of the work with students, considering text purpose, schematic structure and relevant linguistic features. Second, it proposes a principled sequence of purposeful pedagogical stages: deconstruction, joint construction and independent construction. Another virtue of the proposal is that it takes the concept of scaffolding (Wood, Bruner and Ross 1976) that progresses from activities conducted jointly between the teacher and the students, to independent activities on the part of the students. Finally, the proposal uses the concept of explicit pedagogy (Bernstein 1990), which makes clear to the students what they are doing during the process and for what purposes. These theoretical resources are useful to make the students conscious of the process they are following in order to learn how to manage to produce a genre they approach for the first time. However, in our experience, the texts that the students produce by themselves, even after the joint construction, show various problems – at all strata – that need to be solved by a process of editing, as expert writers do. Traditionally, the teacher marks the students’ text, identifies the problems and, frequently, suggests how to solve them. But often, students do not understand these suggestions clearly and consider that the teacher does not understand what they mean. This practice also has a negative connotation: the teacher does not give positive responses to the students (Martin and Rose 2005; Rose and Martin 2012) but rejects what they have done. Consequently, the students often ignore the corrections made by the teacher or misinterpret them, the process resulting in a waste of time and instances of unsuccessful learning and teaching. To solve this problem, two decisions were made in order to reform the TLC. First, a sub-stage called ‘text design’ was introduced before the ‘construction’ stage; second, a new stage was introduced, which was called ‘editing’. The ‘text design’ stage has the purpose of make a plan for the text, showing how the contents will be arranged in the schematic structure. The ‘editing’ stage proposes that the students can find the problems posed by the first version of the text following what they have learnt through the deconstructed model. An important point to note here is that all the stages are realized as joint work between the teacher and the students in order to teach how to do the work in each stage (modelling) and offering scaffolding. In subsequent iterations, the students do the work by themselves, first in small groups and after that individually, in order to remove the scaffolding gradually.

204

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Table 11.1 shows the adapted genre-based pedagogic proposal and the way in which it was reorganized. The table has three columns. The third one illustrates the scaffolding, which is removed gradually, as explained before. The first column illustrates how the negotiation of the field of the text permeates the different stages of the proposal, as in the original model. The column in the middle illustrates how the stages in the earlier model of the TLC have been reorganized, with the introduction of the modifications already mentioned. The first stage is ‘deconstruction’. It involves discovering the relevant characteristics of the text as an instance of a genre through interaction between the teacher and the students. These characteristics are the schematic structure and the main discourse-semantic resources that accomplish the purpose of the genre. Another goal of this stage is that the students understand the meaning of the text so that, while identifying the main linguistic resources, the teacher and the students reflect on the meaning they produce. The language is understood as a meaning resource. The second stage is the ‘construction’ of a new text of the same genre. In Context I, or in writing courses such as those located at the beginning of undergraduate studies, the first construction is made in interaction with the teacher in secondary school classrooms. In Context II it is more difficult to conduct joint construction due to the length of the texts and the short time the professor has to do the work in the subject with which the programme is associated. However, it is recommended to perform joint construction when teaching grammatical metaphor (Halliday 1998) or resources of engagement of the system of appraisal (Martin and White 2005), both relevant features of the discourse of disciplines. The work in the ‘construction’ stage starts with the sub-stage ‘text design’, in which the students and the teacher reflect on how to arrange the content the students need to include in the text they are going to write, jointly or independently, considering the schematic structure of the genre. The content is provided by the teacher in the form of different diagrams or infographics in Context I, but provided by the students from the literature provided by the subject professor in Context II. They also discuss with the language teacher, in both contexts, which main discourse-semantic resources they need to construct the text, considering what they have learnt in the ‘deconstruction’ stage. The teacher needs to make specific reference to what they have learnt about the genre in the deconstruction stage to make it possible for the students to apply this knowledge. After writing their texts, the students give them to the teacher, who selects and anonymizes a number of texts based on factors such as the best written and others which need changes. Then, the ‘editing’ stage takes place, first as ‘joint editing’ of the selected texts. The teacher presents the texts one by one to be discussed with the group of students in order to determine if they have the

TEACHING ACADEMIC GENRES IN SPANISH

205

TABLE 11.1  Adapted Genre-Based Pedagogic Proposal – Joint reading – Joint reading in small groups – Individual reading

Deconstruction Field negotiation

Text design

Editing

Construction

– Joint writing – Joint writing in small groups – Individual writing – Joint editing – Joint editing in small groups – Individual editing

Source: Moyano (2007).

proper schematic structure (or if the student modified it for a good reason) and if the language resources chosen are an effective way to construe the content the author wants to communicate and are consistent with the characteristics of the language of the discipline. Again, the reference point is what they have discovered and learnt at the ‘deconstruction’ stage. This way, it is not the teacher who suggests changes for a new version of the text (as with traditional ‘feedback’ discussed earlier): the suggestions emerge as a result of the discussion of all the students in the class. The teacher guides the process, showing what the writer needs to pay attention to in order to improve their own text. After this process, every student does a first ‘independent edit’ of their text, applying what they have learnt. Figure 11.1 shows, as an example, the results of this process in a class from Context II, the programme across the curriculum. The horizontal axis (x) represents the students, while the vertical (y) shows the grade each student got in the first and in the last version of their text. The grades from 1 to 5 represent respectively: insufficient, regular, good, very good and excellent. The language academic grades the first and the last version of the texts based on a rubric constructed ad hoc but following the same indicators as the rest of the teacher research group. Figure 11.1 shows that eleven of the first versions of the students’ texts obtained less than 2 points (i.e. they were insufficient); three obtained approximately 2.5 (i.e. regular), and three approximately 3.5 (i.e. good). After the process of joint editing and independent editing, the last version of the fourteen students’ texts obtained grades between 4 and 5 points (i.e. at least very good), and the three that had been scored as 3.5 in the first version were scored as 5 points (i.e. excellent). This indicates that all the students did

206

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 11.1  Results of an implementation of the adapted pedagogic proposal (teacher: Prof Yanina García).

improve their texts from the first version to the last one, and also that the gap between students was reduced. If the students have the chance to repeat this process several times with different genres, they will also learn the writing process to transform themselves into ‘experienced writers’.

REFLECTIONS This chapter has outlined an approach to teaching academic literacy in Spanish to students in two contexts, both using SFL as the informing theory: training future teachers to teach academic literacy (Context I) and teaching academic literacy in different disciplines across the curriculum at the undergraduate level (Context II). During the process of working in this area, it was possible to observe that genre theory and discourse-semantic analysis are productive resources for analysing academic texts at different levels of education and in different areas of expertise in Spanish. Also, this project showed the need for more work in the

TEACHING ACADEMIC GENRES IN SPANISH

207

description of Spanish at different strata and across the metafunctions based on the language typology methods proposed by SFL. More work on Spanish academic genres is also needed in order to map genres from school textbooks of different subjects and genres used in academic undergraduate settings in different disciplines. Very important here is work on genre and macro-genres, discourse-semantic analysis and the language of science, not only to develop knowledge about these genres but to develop content and materials for teaching students explicitly in order to develop their abilities to read and write target texts independently. Both contexts required a reworking of the well-established TLC in order to improve the performance of the students. The results shown here, in Moyano (2007) and in unpublished work confirm that the modifications made to the TLC were effective for the pursued ends. They also show that the modified model is suitable for both contexts of implementation. More publications on this project are needed in order to disseminate the results to other educational linguists. To accomplish these goals, collaborative work is required in research groups. Consequently, a critical issue here is the need to teach these theoretical resources to researchers who understand the value of an appliable linguistics such as SFL: one that intervenes in critical social practices as educational linguistics.

NOTE 1 Acronym for the Spanish Agencia Nacional de Promoción de Ciencia y Técnica (National Agency for the Promotion of Science and Technology).

REFERENCES Acebal, M. (2021). De las guías de escritura al macrogénero. La relevancia del conocimiento lingüístico en los Programas de Lectura y Escritura. Rev. De Signos y Sentidos, 22 (2021): 193–217). Acebal, M. (forthcoming b), ‘El posicionamiento actitudinal y las metáforas gramaticales. La inscripción de los significados interpersonales en las respuestas a un parcial universitario de “Psicología Evolutiva”’, in E. Moyano and M. Vidal Lizama (eds), Centros y Programas de Escritura en América Latina: opciones teóricas y pedagógicas para la alfabetización académica. Colorado: The WAC Clearinghouse. Bernstein, B. (1990), Class, Codes and Control, Vol. IV: The Structuring of Pedagogic Discourse, London: Routledge. Christie, F. and J. R. Martin, eds (1997), Genre and Institutions: Social Processes in the Workplace and School, London: Continuum. Escobar, S. and M. I. Jovenich (2013), ‘Un acercamiento a los géneros explicativos en Geografía: el caso de la Explicación histórica y la Explicación Factorial’, in E. I. Moyano (ed.), Aprender ciencias y humanidades: una cuestión de lectura y escritura.

208

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Aportes para la construcción de un programa de inclusión social a través de la educación lingüística, 333–63, Los Polvorines: UNGS. Giudice, J. (2013), ‘La historia contemporánea en manuales de ciencias sociales argentinos: tipo de lenguaje y elección genérica’, in E. I. Moyano (ed.), Aprender ciencias y humanidades: una cuestión de lectura y escritura. Aportes para la construcción de un programa de inclusión social a través de la educación lingüística, 297–332, Los Polvorines: UNGS. Giudice, J. and E. I. Moyano (2011), ‘Género y formación de ciudadanos: la reconstrucción del período 1976–1983 en manuales argentinos para la escuela primaria’, in T. Oteiza and D. Pinto (eds), En (re)construcción: Discurso, identidad y nación en los manuales escolares, 205–68, Santiago de Chile: Cuarto Propio. Halliday, M. A. K. (1998), ‘Things and Relations: Regrammaticising Experience as Technical Knowledge’, in J. R. Martin and R. Veel (eds), Reading Science: Critical and Functional Perspectives on Discourses of Science, 185–235, London: Routledge. Halliday, M. A. K. ([1993] 2003), ‘Language in a Changing World’, in J. J. Webster (ed.), Collected Works of M.A.K. Halliday, Vol 3: On Language and Linguistics, 213–31, London: Continuum. Halliday, M. A. K. ([1991] 2007), ‘The Notion of “Context” in Language Education’, in J. J. Webster (ed.), Collected Works of M.A.K. Halliday, Vol 9: Language in Education, 269–90, London: Continuum. Halliday, M. A. K. (2008), Complementarities in Language, Beijing: Commercial Press. Halliday, M. A. K. and R. Hasan (1985), Language, Context, and Text: Aspects of Language in a Social-Semiotic Perspective, Geelong, VIC: Deakin University Press. Halliday, M. A. K. and J. R. Martin (1993), Writing Science: Literacy and Discursive Power, Pittsburgh, PA: University of Pittsburgh Press. Halliday, M. A. K. and C. M. I. M. Matthiessen (2014), An Introduction to Functional Grammar, 4th edn, London: Arnold. Martin, J. R. (1989), Factual Writing: Exploring and Challenging Social Reality, Oxford: Oxford University Press. Martin, J. R. (1992), English Text: System and Structure, Amsterdam: Benjamins. Martin, J. R. (1993a), ‘A Contextual Theory of Language’, in B. Cope and M. Kalantzis (eds), The Power of Literacy: A Genre Approach to Teaching Writing, 116–36, London: Falmer Press. Martin, J. R. (1993b), ‘Literacy in Science: Learning to Handle Text as Technology’, in M. A. K. Halliday and J. R. Martin (eds), Writing Science: Literacy and Discursive Power, 166–202, Pittsburgh, PA: University of Pittsburgh Press. Martin, J. R. (1994), ‘Macro-genres: The Ecology of the Page’, Network, 21: 29–52. Martin, J. R. (1999), ‘Mentoring Semogenesis: “Genre-Based” Literacy Pedagogy’, in F. Christie (ed.), Pedagogy and the Shaping of Consciousness: Linguistic and Social Processes, 123–55, London: Cassell. Martin J. R. and D. Rose (2005), ‘Designing Literacy Pedagogy: Scaffolding Asymmetries’, in J. Webster, C. Matthiessen and R. Hassan (eds), Continuing Discourse in Language, 251–80, London: Continuum. Martin J. R. and D. Rose (2007), Working with Discourse: Meaning beyond the Clause, London: Continuum. Martin J. R. and D. Rose (2008), Genre Relations: Mapping Culture, London: Equinox. Martin, J. R. and P. P. R. White (2005), The Language of Evaluation: Appraisal in English, London: Palgrave.

TEACHING ACADEMIC GENRES IN SPANISH

209

Matthiessen, C. M. I. M. and M. A. K. Halliday (2009), Systemic Functional Grammar: A First Step into the Theory, Beijing: Higher Education Press. Moyano, E. I. (2007), ‘Enseñanza de Habilidades Discursivas en Español en Contexto Pre-universitario: Una Aproximación desde la LSF’, Revista Signos, 40 (65): 573–608. Moyano, E. I. (2010a), ‘Aportes del análisis de género y discurso a los procesos de enseñanza y aprendizaje escolar: las ciencias biológicas y la historia’, Discurso & Sociedad, 4 (2): 294–331. Moyano, E. I. (2010b), ‘Escritura académica a lo largo de la carrera: un programa institucional’, Revista Signos, 43 (74): 465–88. Moyano, E. I. (2013a), ‘Géneros y discurso en los manuales de biología: la construcción del conocimiento y la actividad científica’, in E. I. Moyano (ed.), Aprender ciencias y humanidades: una cuestión de lectura y escritura. Aportes para la construcción de un programa de inclusión social a través de la educación lingüística, 229–96, Los Polvorines: UNGS. Moyano, E. I. (2013b), ‘El lenguaje de las disciplinas y los géneros de su recontextualización escolar. Una aproximación desde la lingüística sistémicofuncional’, in E. I. Moyano (ed.), Aprender ciencias y humanidades: una cuestión de lectura y escritura. Aportes para la construcción de un programa de inclusión social a través de la educación lingüística, 31–78, Los Polvorines: UNGS. Moyano, E. I. (2015), ‘La sección Discusión del artículo científico como género: Construcción del nuevo conocimiento y construcción del autor’, PhD thesis, Universidad de Buenos Aires, Buenos Aires. Moyano, E. I. (2016), ‘Theme in English and Spanish: Different Means of Realization for the Same Textual Function’, English Text Construction, 9 (1): 190–220. Moyano, E. I. (2017), ‘Diseño e implementación de programas de lectura y escritura en el nivel universitario: principios y estrategias’, Revista Lenguas Modernas, 50: 47–72. Moyano, E. I. (2018), ‘La enseñanza de la lectura y la escritura académicas mediante un programa a lo largo del curriculum universitario: opción teórica, didáctica y de gestión’, Revista DELTA, 34 (1): 235–67. Moyano, E. I. (2019), ‘Knowledge Construction in Discussions of Research Articles in Two Disciplines in Spanish: The Role of Resources of appraisal’, Journal of Pragmatics, 139: 231–46. Moyano, E. I. (2021), ‘Descripción de géneros para su enseñanza en un programa de escritura académica. Hacia la Deconstrucción conjunta’, in Ávila, N. (ed.), Contribuciones multilingües a la investigación en escritura: Hacia un intercambio académico igualitario, 289–310, Colorado: The WAC Clearinghouse. Oteíza, T. (2010), ‘De la conspiración de silencio al reconocimiento de voces alternativas. Las violaciones a los derechos humanos en Chile según el Informe Valech (2003)’, Revista Latinoamericana de Estudios del Discurso (ALED), 9 (1): 87–112. Oteíza, T. (2017), ‘Escritura en la historia: Potencial de los reursos lingüísticos interpersonales e ideacionales para la construcción de la evidencia’, Lenguas Modernas, 50: 193–224. Oteíza, T. and C. Pinuer (2012), ‘Prosodia valorativa: construcción de eventos y procesos en el discurso de la historia’, Discurso y Sociedad, 6 (2): 418–46. Quiroz, B. (2015), ‘La cláusula como movimiento interactivo: Una perspectiva semántico-discursiva de la gramática interpersonal del español’, DELTA: Documentação e Estudos em Linguística Teórica e Aplicada, 31 (1): 261–301.

210

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Quiroz, B. (2017), ‘The Verbal Group’, in T. Bartlett and G. O’Grady (eds), The Routledge Handbook of Systemic Functional Linguistics, 301–18, London: Routledge. Rose, D. (1998), ‘Science Discourse and Industrial Hierarchy’, in J. R. Martin and R. Veel (eds), Reading Science: Critical and Functional Perspectives on Discourses of Science, 236–65, London: Routledge. Rose, D. and J. R. Martin (2012), Learning to Write, Reading to Learn: Genre, Knowledge and Pedagogy in the Sydney School, London: Equinox. Serpa, C. (2021), ‘La enseñanza de la escritura en la educación superior: una experiencia didáctica en torno al macrogénero “Descripción de Flujograma” ’, Íkala, Revista de Lenguaje y Cultura, 26 (1): 77–96. Vidal Lizama, M. (2020), ‘Aproximación al ensayo académico como género de formación en Ciencias Sociales: el caso de Sociología’, D.E.L.T.A., 36 (4): 1–26. Vidal Lizama, M. and N. Leiva (2021), ‘Aproximación discursiva a la ‘solemne’ en Derecho desde la LSF: una aplicación del sistema de figura’, Literatura y Lingüística, 14: 345–75. Vidal Lizama, M. and S. Montes (forthcoming), ‘Escritura académica en la formación de estudiantes de Arte: continuidades y cambios a lo largo de la carrera’, in E. I. Moyano and M. Vidal Lizama (eds), Centros y Programas de Escritura en América Latina: opciones teóricas y pedagógicas para la alfabetización académica, Colorado: The WAC Clearinghouse. Wignell, P. (1998), ‘Technicality and Abstraction in Social Science’, in J. R. Martin and R. Veel (eds), Reading Science: Critical and Functional Perspectives on Discourses of Science, 297–326, London: Routledge. Wignell, P. (2007), On the Discourse of Social Science, Darwin: Charles Darwin University Press. Wood, D., J. Bruner and G. Ross (1976), ‘The Role of Tutoring in Problem Solving’, Journal of Child Psychology and Psychiatry, 17: 89–100.

CHAPTER TWELVE

Developing an SFL-inspired Four-year Collegiate Foreign Language Programme Contexts, Constructs, Curriculum HEIDI BYRNES

INTRODUCTION In his prepared remarks for the opening plenary at the 2010 annual meeting of the American Association for Applied Linguistics in Atlanta, GA, Michael Halliday offered these concluding thoughts: An appliable linguistics is a way of thinking about language; that is its immediate scope and context of application. But to be appliable to reallife situations and real-life tasks, it has to be good to think with: that is, a resource for investigating these various domains and spheres of activity – not just their forms of discourse, important as these are, but their essential nature as systems-&-processes of meaning. (Halliday 2010b) It is one of numerous statements that lays out Halliday’s particular take on the inseparable, reciprocal and dialectic relationship between what he considered

212

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

worthwhile linguistic theorizing and attempts to solve problems in real life where language plays a central part. In this section of the handbook on appliable linguistics I will explore this intricate interrelationship in the domain of educational linguistics, fundamentally a language-driven and complex societal activity. Specifically, I will investigate how an entire group of educators, faculty and graduate students in my home department, the German Department at Georgetown University (GUGD), came to ‘think with language’ in the problem space of delivering adult non-English foreign language instruction at the college level. Yet more focused, my inquiry will explore how Systemic Functional Linguistics (SFL) as a comprehensive functional theory of language eventually came to provide both the indispensable theoretical frame of reference to which the educators of the programme turned for continued inspiration and, at the same time, how their setting and the practices they chose to realize their educational vision highlighted particular aspects of the theory. In that gradually evolving ‘knowingby-doing’ dialectic the programme’s educators continuously stretched, sculpted and solidified the theory’s range of applicability. At the same time, they subtly repositioned its theoretical reach and strengthened its already formidable claims to serve as a language theory of choice for the language-related challenges that present themselves to societies around the world in the age of multilingualism, migration and technologized forms of communication (see e.g. the discussion in Douglas Fir Group 2016). It is a fortuitous time to engage in such reflections as the programme has just celebrated twenty years of successfully delivering educative work in adult language teaching and learning that has gained national and international attention for its unique contribution to enabling its instructed learners of German to attain advanced levels of multi-literate competence (for a monograph-length treatment, see Byrnes, Maxim and Norris 2010; a list of publications can be found at https://german.georgetown.edu/bibliography). My goal in this chapter is to offer a critical retrospective on the intricate give and take between SFL and a particular instance of situated educational practice in order to recognize both the unique contributions of SFL theorizing and the particular opportunities and realizations of an appliable linguistics that distinguish this project from other applications of SFL.

PROBING AND DISCERNING THE EDUCATIONAL CONTEXT: ON THE WAY TO FRAMING THE QUESTION One way to understand the ‘domains and spheres of activity’ that created a rationale for the GUGD’s efforts to embark, in the spring of 1997, on a comprehensive rethinking of its educative work is to consider external as well

COLLEGIATE FOREIGN LANGUAGE PROGRAMME

213

as internal factors that contributed to how ‘the real-life situations and real-life tasks’ came to be framed by the department’s educators. Dynamics of the external context With regard to the external context, despite the undisputed conceptual appeal and practical prominence of communicative language teaching – sharpened in the US context by the performative focus of proficiency-oriented language teaching and assessment – by the 1990s unease was building up in the higher education foreign language community about the deeper implications of that reorientation. Increasingly, voices were heard that duly recognized these frameworks’ admirable focus on language use in social context, as contrasted with an earlier focus on knowledge about language structures and forms epitomized in concerns about ‘grammar’ and relentless insistence on formal accuracy. At the same time, these voices questioned their suitability for bolstering the argument for the intellectual-academic merit of collegiate FL learning, a defence that was needed in light of the increasingly instrumentalist and utilitarian approach to language study encapsulated in the derogatory appellation of an advancing ‘Berlitz-ification’. The resulting deep unease points to two divergent sets of epistemological and ontological assumptions and foci about the nature of language and, by extension, the goals of instructed language learning. The first privileged an immanent, largely ahistorical social science–oriented understanding of interactive, mostly oral language use as the best way to foster language learning portrayed as socially situated, mostly quotidian communication that would engage the learner creatively in his/her learning processes (the operative catchwords being ‘real-world communication’ and ‘learner-centeredness’). This direction aligned with much psycholinguistically motivated second language acquisition (SLA) research, its findings and pedagogical recommendations. The second affirmed a way of engaging with language that had long characterized the humanities, namely its focus on inquiry into oral and written texts in the other language in order to develop sophisticated understandings of the complexity of the human condition, both individual and societal, as that culture had come to experience and express it in its texts. In the majority of collegiate language departments in the United States, particularly the influential graduate departments, those simultaneous and yet disparate interests were manifested in starkly bifurcated curricula. They would offer largely content-less, culturally shallow, proficiency-driven foreign language learning classes in the first two years of study, followed by the demand for sophisticated textually oriented literary-cultural engagement with the other languaculture in the remaining two years and, further on, into graduate studies. That split was deepened by the two pertinent professional organizations, the American Council on the Teaching of Foreign Languages

214

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

(ACTFL) and the humanities-oriented Modern Language Association (MLA) that provided the discursive justification and professional activities for one or the other approaches. The prominence of the so called Standards movement in K–12 education, bolstered by the 2002 No Child Left Behind legislation, with its emphasis on accountability, transparency, benchmarking of educational outcomes and outcomes assessment along proficiency lines favoured the ACTFL model and created a spill-over effect into the first two years of collegiate language teaching (see Scott 2010). That was possible because higher education simply continued to assert the superior value of its humanistic approach to language study while offering few intellectual arguments and, more important, few convincing on-the-ground educational resources in support of its preferred educational vision. Eventually, however, a widely heralded report by the MLA (2007: 237) seemed to address matters head on when it proposed ‘replacing the two-tiered language-literature structure with a broader and more coherent curriculum in which language, culture, and literature are taught as a continuous whole, supported by alliances with other departments and expressed through interdisciplinary courses’. Despite this seemingly compelling solution, ultimately, the called-for profession-wide change did not occur (for a critical discussion, see e.g. Allen 2010; Byrnes et al. 2010; Byrnes 2012b; Paesani, Allen and Dupuy 2015). That raises the following question: What factors might explain why the GUGD faculty, unlike the vast majority of US collegiate foreign language departments, chose to take bold, even if initially unsure steps towards reworking its entire programme? Identifying an internal rationale for change: Asking the question A first answer to the previous question is this: At the time, the department’s faculty could draw on considerable expertise regarding the politics and substance of higher education language studies on both sides of the dividing line. That meant that, once the decision to undertake this project was made, it was relatively straightforward to identify core programme goals: its linguistic adult learners, many of whom were third language learners, should be given the opportunity to learn German and engage with literary-cultural content about the German-speaking area as efficiently and effectively as possible. They were to ‘become competent and literate users of German who can employ the language in a range of intellectual and professional contexts and who can also draw from it personal enrichment and enjoyment’ (for the full goals statement, see https://german.georgetown.edu/page/1242716542088.html). Such goal setting recognized that in a PhD-granting programme the privileged educational ethos had to be that of the humanities with its strong text and meaning orientation in cultural context, a choice that was further

COLLEGIATE FOREIGN LANGUAGE PROGRAMME

215

corroborated by Georgetown’s commitment to a liberal arts education in the Jesuit tradition. At the same time, among the university’s outstanding programmes were professional schools like the School of Foreign Service (SFS) whose students were a prominent constituency in language departments. As a consequence, advanced-level proficiency, as operationalized by the SFS and regularly assessed as part of a formal graduation requirement, provided a crucible for both faculty and students regarding the programme’s actual learning outcomes in a performative exam that typically lasted 20–25 minutes: a student’s ability to read, understand, summarize succinctly and discuss critically with two departmental faculty examiners an article in an influential newspaper or magazine on a pertinent contemporary topic. In other words, linking language and content in a culturally informed, multiperspectival way, doing so at a reasonably high level of literacy in German and, most importantly, finding ways to enable students to attain those ability levels in a short period of time were driving forces for the faculty to initiate this project. They took on new urgency at the time, first because increasingly in the United States German had become a less commonly taught language; that is, just like Arabic, Chinese or Japanese, German was unlikely to be taught at the secondary level. As a consequence, university programmes needed to be prepared to enrol ab initio learners of German alongside students who had previously had the opportunity to learn or study German and to enable all of them to attain worthwhile ability levels beyond highly contextualized transactional language use. Second, persuasive evidence for the intellectual merit and successful pursuit of studying German would have to be delivered in a remarkably short period of time, namely the four-year period of the typical undergraduate programme. Even more challenging, ideally the programme should move decisively towards that goal after only four semesters of study in order to enable even those students from various disciplines (e.g. economics, government, theology, philosophy) who had begun their German studies at Georgetown to participate in the much desired junior year abroad with direct enrolment at a German-speaking university. Finally, in order to sustain itself, the programme needed to be able to attract and retain students from diverse disciplines across the university who, for whatever reasons, were interested in studying German, not only German majors and minors who had always been a relatively small cohort. To sum up, it was the faculty’s publicly stated willingness and ability to draw on shared professional ethics and a sense of responsibility towards facilitating all their students’ culturally and linguistically meaning-oriented learning at a high functional level that turned out to be the decisive driver behind this unique educational initiative in American higher education. It also turned out to be an extraordinary challenge – and an exciting journey for all stakeholders, faculty, graduate students and the programme’s learners.

216

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

INVESTIGATING THE ISSUE: CONSIDERING OPTIONS The department began by first laying out possible academic-administrative options that might be used for attaining its stated goals within existing educational practices at Georgetown University, along with considering broad consequences for the kind of faculty engagement that would be required over an agreed-to initial three-year period of implementation (e.g. collaborative and open-minded participation by all in curriculum-related activities, adjustments to the faculty reward system, graduate student involvement, new course development, materials development, implications for assessment; see e.g. Byrnes 2001; Byrnes et al. 2010). More important for this discussion was the next step, a search in the professional literature – theoretical, empirical and pedagogical-practical – whether the issues we had laid out for ourselves might already have been addressed by others. What we uncovered, however, was that remarkably little substantive engagement with, much less evidence-based recommendations for, such programmatic thinking over an extended sequence of instruction were available anywhere. To be sure, by the 1980s and 1990s, diverse educational reform efforts had been underway. In one form or another they all grappled with how instruction should handle the relationship between a focus on content/meaning and a focus on the formal features of language. Given the perceived – and real – absence of meaningful cultural content in the pivotal (typically requirement) language courses, the thrust of those efforts went towards bringing content into language teaching. Examples are the language across the curriculum movement, which in the 1980s sought to overcome the isolation of FL departments from the remainder of a university’s educational enterprise, particularly in the humanities by involving faculty in non-language disciplines in the work of language departments (see Allen, Anderson and Narváez 1992); long-standing and still ongoing initiatives towards realizing content-based instruction in language departments themselves (Byrnes 2005; for early arguments, see Brinton, Snow and Wesche 1989; for contemporary justifications, see Tedick and Wesely 2015 for US K–12 education, and Cammarata 2016 for K–16 considerations); and the cautionary, though ultimately form-oriented, conclusions drawn from the bilingual and immersion education efforts in Canada (see e.g. Harley et al. 1990) about the appropriate mix of focus on content or on form. As intimated earlier on, how advanced-level learning of the foreign language was to be fostered in the so-called content courses constituted the fateful educational lacuna (Byrnes 2011). Not only were the educational settings from which these findings arose at a considerable distance from our own, they were also in need of interpretation regarding their relevance. Ironically, their real shortcoming – and this was an insight that crystallized only gradually in the numerous departmental

COLLEGIATE FOREIGN LANGUAGE PROGRAMME

217

workshops and extensive curricular and pedagogical discussions to which we had committed ourselves – were their constant affirmations of an inextricable link between form and meaning, which, in their own ways, pointed to the fundamental shortcomings of a structuralist understanding of the system of language as an assemblage of more or less general, but always arbitrary, formalist rules that learners needed to apply flawlessly in order to communicate. At best, it could present an additive notion, rather than a functionally integrated, ‘natural’ link between socially situated meaning and form; in reality it erected a nearly insurmountable hurdle to successful language learning ‘beyond the basics’. Just as seriously, other options mentioned in the literature seemed diversionary, even escapist. For example, the department had never subscribed to the lure of ‘interdisciplinarity’ as the solution to what was surely a languagerelated, that is, a discipline-specific, problem; nor had we felt that ‘outsourcing’ the issue by appealing to cooperation with other departments and their content areas was intellectually sufficiently honest. Then, too, we had long focused on texts in context, rather than sentence-level formal features, the scope of so much of the professional preoccupation with beginning and intermediate performance levels, though we grappled with its challenges. Also, at least aspirationally, we had continued to uphold that ordinary faculty would teach at all levels of the programme – including most especially ‘language’ courses – and not only in their specialty areas. And, finally, our educational philosophy had never subscribed to the deficit notions of what could be accomplished in instructed learning that characterized the monolingually oriented research literature (for a contemporary discussion, see Ortega 2019). Instead, we had consistently affirmed – because we regularly experienced it in our programme – that advanced ability levels were indeed attainable for learners under certain educational-theoretical-pedagogical conditions. In short, while we had glimpses of preferred directions and preferred educational parameters, we were looking to gain a better understanding of the dynamic nature of that facilitative nexus – and with that a more sophisticated and better-substantiated way of realizing it throughout our programme. The sought-after breakthrough occurred when, by way of content-based instruction and, even more so, the English for specific purposes (ESP) literature, we became increasingly aware of genre-oriented approaches to teaching and learning. Once more, the educational settings and the teaching and learning interests being described differed appreciably from ours. For example, the focus was on English, mostly for international graduate students who had experienced years of instruction in their home countries, which meant that the programmes’ key challenge was to enable them to attain academic abilities in various disciplinary areas, rather than imagining and implementing an entire developmental progression from beginning to advanced performance. Also, in

218

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

many cases, these were highly restricted programmes of a few semesters or intense summer offerings and often located at the fringes of their universities’ work, in contrast with the educational expectations that a degree-granting programme like ours needed to meet as it aspired to claim a valued place in an institution’s academic profile. Finally, they tended to treat language use modalities separately – targeting the enhancement of speaking or reading or writing – in contrast with the inherent necessity in a programme such as ours to integrate all modalities, not least due to its short duration. At the same time, the comprehensiveness, whole text, content and meaning orientation that characterized this research was immediately appealing, all the more so as its major conceptual construct of genre was entirely familiar to the department’s literature faculty. Not only did this facilitate the creation of a new shared discursive platform that removed the oftentimes off-putting, because ultimately mechanistic, ‘input and output’ metaphors of psycholinguistically oriented SLA research; precisely because the understanding of genre in the ESP literature differed from conceptualizations, theorizations and uses prominent in literary studies, it created the need to negotiate a meeting ground that was new to all educators in the department, faculty and graduate students, as they endeavoured to gauge the language and cultural learning possibilities that a genre orientation might offer for the GUGD programme, no matter their specific teaching assignments. The decisive step in those early years – and one that since then has indelibly shaped the entire project – came about when, three years into its comprehensive curricular renewal project, the department applied for and received a two-year grant under the Spencer Foundation’s ‘Practitioner research communication and mentoring grants’. Three research groups, involving a total of twelve faculty and graduate students and their faculty mentors, chose the following areas for in-depth study and practical engagement: first, the socialization and professional development of graduate teaching assistants in the revised GU curriculum with its articulated four-year sequence, which we called the Multiple Literacies curriculum; second, the development of materials that would address the desired link of content and language acquisition on all levels of instruction towards the goal of advanced ability levels; and, third and most pertinent to this chapter, an in-depth exploration that sought to develop shared descriptions and working definitions of the construct of genre as well as narrative, both of which had become central to the curricular and pedagogical materials that, in their first iterations, had been developed in the early years of the project.

LEARNING TO WORK WITH AN APPLIABLE SFL Not surprisingly, it was the genre group’s inquiry in particular that, over time, allowed the entire department to find in SFL a theory of language that was ‘good

COLLEGIATE FOREIGN LANGUAGE PROGRAMME

219

to think with’ for a myriad of issues, many of which revealed their significance and far-reaching consequences for what we valued in our educational work only when we learned to see and express them in light of major concepts and processes that the theory makes available. Departmental faculty have addressed a number of such phenomena and have engaged in oftentimes extended – even iteratively revisited – consideration in light of SFL constructs. As mentioned earlier, they have done so in joint publications by faculty and graduate students, in collaborations of literature and linguistics faculty and, most extensively, in their educational practice. From among many other possibilities, in this segment I highlight five particularly noteworthy issues. Linking SFL and a task-based approach to language learning and teaching (TBLT) Perhaps the most consequential adjustment made within the GUGD project to what is commonly described as the Sydney School of genre-oriented teaching within SFL is its deliberate linking to one of the most influential theoretical, empirical and pedagogical re-orientations in language learning and teaching of the past decades, namely a task-based approach (TBLT; Byrnes 2002, 2012a, 2015; Byrnes et al. 2006). Detailed in numerous publications pertaining to the GUGD, that link can be seen as both opportunistic and strategic. Two decades ago, when SFL had yet to attain any noteworthy recognition, much less credibility, within the North American applied linguistics scene, not to mention serious consideration for meeting the formidable challenges facing US collegiate FL programmes as I have previously recounted them, such an explicit alliance facilitated positioning our project, with its defining language use and content and meaning orientation, at the cutting edge of applied linguistics research and practice. At the same time, it offered to the construct ‘task’ something that has repeatedly been critiqued as lacking in the advocacy of TBLT, namely a well-theorized foundation that would guard against seemingly limitless interpretations of what constituted a task both theoretically and pedagogically beyond eschewing a focus on ‘grammar’. Specifically, by linking ‘task’ to the notion of genre, both oral and written, as well as theorized within the conceptual architecture of SFL as a functional theory of language, two consequential contrasts were highlighted: first, the contrast to ‘genre’ prevalent in new rhetoric, where it was sociologically motivated and thus less well suited to establishing the critically important link between language use and the system of language itself (see Hyon 1996); and second, the repositioning of task away from being grounded primarily in psycholinguistic, sentence-oriented processing considerations towards concerns of textuality and literacy as they are inherent to the notion of genre. By anchoring pedagogical

220

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

tasks in diverse textual genres, themselves amenable to the kinds of selection and sequencing decisions that are at the heart of curriculum building, it was possible to begin to develop systematic and principled ways of imagining both multiyear extended curricula and the pedagogies that might be used to implement them. Furthermore, inasmuch as genres themselves were defined by the kinds of lexicogrammatical resources that were highly likely to be deployed in their realization, such an approach had the added benefit of linking the development of these resources in a manner that could be presumed to facilitate language learning over extended time periods (Byrnes 2012a). Finally, by linking genrebased tasks, particularly for writing development but also for speaking, to assessment, faculty gained considerable insights into the kind of specificity necessary – and possible – both for valid assessment practices and for arriving at statements about realistic and at the same time challenging learning goals at different instructional levels whose trajectory pointed towards advanced ability levels (Norris 2016). In sum, the deliberate linking of task and genre began to give intellectual heft and much-needed detail to how one might address several central concerns that had motivated the GUGD project in the first place. Put another way, the truism that language learning to advanced ability levels takes time – that is, the imperative for a longitudinal perspective if worthwhile approaches to language learning are to be pursued – gradually became translatable into educational action that was researchable and adaptable. Foregrounding the developmental axis of SFL constructs The previous discussion foregrounded a position that has been implicit all along, namely a favouring of those concepts in SFL that have a strong developmental thrust (Halliday 2007; Byrnes 2019b). Assuming some level of professional competence on the part of teachers, learning should occur in any language classroom. However, that is not the same thing as development, particularly when development towards useful levels of competence is desired within an extraordinarily restricted time frame. In light of that challenge, what guiding considerations might be used to foster it carefully and unapologetically towards advanced ability levels? Here the GUGD educators worked with and extended a number of SFL constructs that were developmental in their conceptualization. Among them is a set of concurrent continua that bolstered ways of imaging an integrated extended programme, did not depend on a formal background in linguistics, and yet were seen as plausible by the non-linguist majority of the departmental faculty. First, the oral–literate continuum, well described for its major implications with regard to likely lexicogrammatical resources realizing one or the other avenue of communication (see particularly, Halliday 2002), along

COLLEGIATE FOREIGN LANGUAGE PROGRAMME

221

with their enormous consequences for a linguistic theory of learning and development as an educational linguistics requires it. Second, the continuum of dialogicality, which asserted in its metafunction of tenor that all language use presumed an Other to whom the language was directed and with whom a relationship was being realized. For the kind of oral language use that tends to dominate beginning levels of language learning that dialogicality is overt and directly anchored in tangible parameters of space and time; for written language, however, with its increased distance between lived experience and language use, that dialogicality must now be negotiated through an ever richer set of linguistic resources that facilitate the kind of complex positioning that characterizes our relationship with others (see e.g. Ryshina-Pankova 2010). The third continuum is the semiotic continuum, exquisitely theorized in SFL as a movement from congruent to metaphorical semiosis (Halliday and Matthiessen 1999) that provides one of the most convincing demonstrations of the linguistic-cum-cognitive nature of human expressiveness as it gradually evolves, particularly when appropriately supported by educational practices. Finally, the programme’s engagement with genre and its literary-cultural focus that aspires to historical depth found Bakhtin’s notion of a continuum of the language system, such that there is a dialogical interrelationship between synchronic and diachronic manifestations of language use, a good way to make manifest to learners the ‘presence of the past’ in all cultural negotiations. Privileging an explicit trajectory towards advancedness Within the programme’s focus on advancedness, particularly as we learned to express it in terms of an expanded repertoire of registers and genres and an increasing facility with deploying metaphorical forms of semiosis in both oral and written language use, no concept has had greater descriptive, pedagogical and explanatory power than that of grammatical metaphor (GM). It provided the much-needed counterweight to the dominant interpretation in SLA research that language learning was best captured in terms of increasing complexity, accuracy and fluency of performance, all form-oriented criteria that have no link to meaning making and little textual reach (Norris and Ortega 2009; Ryshina-Pankova 2015b). In the crucially important shift towards more academic language use, the intricately interrelated consequences of GM realized as nominalizations in particular have become not only an indispensable optic for uncovering a treasure trove of insights about learner language. More important, they enabled faculty to appreciate with ever greater specificity the enormity of the cognitive-linguistic shifts that learners need to engage in order to become advanced learners. If ever a long-term, developmental view was in order, it

222

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

was here, as all participants in this educational challenge learned to notice and interpret its diverse tell-tale forms of realization, beginning at intermediate levels of performance and extending all the way into graduate study, where the latter proved particularly revelatory to our non-native graduate students with regard to their own academic language use (see Byrnes, Crane and Sprang 2002). Continuous observations and continuing faculty dialogue have led to ever more refined practices with regard to the entire spectrum of language learning and teaching activities: to support that shift pedagogically through well-chosen texts and tasks; to illustrate to learners what new ways of meaning making these shifts in resources facilitated; to provide useful feedback to them as each embarked on their individual journey of longue durée and variability even when within a shared instructional context; to temper teacher expectations as to what, realistically, was achievable when – to mention only some of the most salient considerations. Beyond such core issues, perhaps nowhere else were the well-developed literacy capacities in the students’ L1 and the aspiration to attain comparable meaning-making capacities in the L2 more palpable. As a consequence, as feedback over many years has indicated, perhaps nowhere else could students gain the kind of immensely motivating sense of achievement than in this crucial passage towards advanced multiple literacies (Pfeiffer and Byrnes 2009). Integrating all modalities Earlier on, I portrayed the integration of all modalities as a matter of expediency in light of our constricted learning context. However, over the years, we have come to appreciate – and use deliberately – a much more substantive justification for an integrated approach. It shows considerable affinities to Rose’s (2016) description of the three phases in the genre-based literacy pedagogy of the Sydney School, whereby the ‘Reading to Learn’ phase, in particular, explored relations between reading and writing as key institutional practices that required certain pedagogical practices, here realized in terms of deconstruction, joint construction and independent construction. Such integration of modalities is enriched by several additional perspectives. A central feature of our pedagogical work is to incorporate well-chosen instances of genre texts to model all aspects of textuality and lexicogrammar. This meant using the guided reading that characterized such modelling for highly targeted text-mining activities that, in turn, provided the core of diverse in-class speaking activities intended to direct learners to attend to the genre’s particular qualities. In addition, model texts not only provided the basis for genre-based writing tasks that students accomplished over a series of drafts and revisions; those writing tasks were elevated into further classroom discourse at

COLLEGIATE FOREIGN LANGUAGE PROGRAMME

223

a yet higher performance level, thus modelling the performance reach, even in oral language, of more academic language use. Unsurprisingly, this practice recommended itself especially in the crucial phase of reaching towards advanced ability levels, whereby the nexus of joint reading–directed joint speaking–individual writing with feedback and revisions–communal as well as individual speaking tasks using the benefit of cognitive-linguistic rehearsal facilitated through the in-depth engagement in writing turned out to be one of the most advantageous ways in which adult learners can draw on their existing L1 literacy capacities in order to advance their L2 abilities in both writing and speaking. Once this became apparent, the programme’s already strong focus on writing tasks was further increased, overriding arguments in its disfavour that recognize it as a labour-intensive form of engagement with language for both teachers and learners. Not only was L2 writing ability enhanced, an expected outcome given a practice effect; especially gratifying and confirming the acquisitional and developmental quality of an integrated literacy-oriented approach was students’ outstanding performance in officially validated speaking assessment that far outpaced the learning outcomes at US colleges (Norris and Pfeiffer 2003). Realizing curricular thinking My final point is both a reprise of the previous discussion and a way to highlight what might well be regarded as the unique quality of the GUGD project, namely its translation of its educational interests, now considerably sharpened under the influence of the conceptual, analytical and research resources made available by SFL, into an integrated, genre-based and task-oriented articulated four-year curriculum at the undergraduate level, with implications for graduate study. The majority of publications pertaining to the GUGD refer to the importance of this curricular stance, some more and some less explicitly, but typically from an educational standpoint. Beyond that, more recent publications have argued for a curricular perspective not merely as a way to deliver language programmes, but more deeply from a theoretical perspective if we wish to understand instructed language acquisition as a phenomenon in its evolving ‘systems-&-processing of meaning’ (e.g. Byrnes 2015, 2018, 2019a, 2019b, 2020). In other words, what merits particular emphasis within the overall argument for SFL as an appliable linguistics is that development of a curricular sequence is not merely an educational-administrative exercise in order to give structure to a programme. Instead, it creates the necessary context for making theoretical statements about the nature of instructed learning. That turns it into an intellectual enterprise whose validity and value sit at the interface of the practice–research–practice interface in the domain of adult-instructed foreign language learning. Ideally laid out by all the faculty of a programme as

224

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

a statement about how they imagine major constructs and stages in the overall trajectory towards advanced ability levels, a curriculum is a kind of distillation of what we ‘know’ that has a strong anchor in theory and research and shows high practical plausibility. Plausibility is not certainty, with the implication that such a proposal will need to be scrutinized for the validity of its underlying assumptions – the real point of educational research. Depending on findings, adjustments may be in order. But without the existence of a curricular proposal in the first place it will essentially be impossible to interpret research findings and make them actionable towards improving learning outcomes. In other words, it is not longitudinal study in and of itself, by now ritually called for in applied linguistics in order to overcome the obvious shortcomings of the prevailing cross-sectional studies, that will yield what we really need to know. Rather, only if ‘longitudinal studies are located within a carefully developed curriculum can we expect truly insightful statements about the nature of adult instructed L2 learning’ (Byrnes 2002: 434–5). Given the remarkably limited validated – as contrasted with speculative – knowledge about how instructed learning actually evolves, for now building up well-documented insights from richly described case studies regarding the consequences for learning of situated thought and thoughtfulness for L2 learning may need to take precedence over premature generalizing conclusions (Norris and Manchón 2012). Beyond such concerns stand eminently practical challenges. To date, the language studies field seems to have amassed ample evidence for mediocre learning outcomes when learners have to fend for themselves without a coherent educational philosophy as a curriculum instantiates it. What we do not have evidence for is how successful they might actually be when their learning is appropriately supported through a well-considered curricular progression. That means we also have no way of judging whether ‘a particular performance was or was not expected, is about average, or is unusual within the instructional programme in which it arose, is considered good or less good’ (Byrnes 2015: 207), all key considerations in any educative work, both internally for its own health and externally in order to meet societal expectations. And a final point: As the GUGD experience amply demonstrates, to be worthwhile – and that means to be realizable in practice by teachers and students alike – a curricular proposal will require a natural functional theory of language that builds on the assumption of non-arbitrary relationships between the meaning side of language and its expressive resources, where those relationships, furthermore, lend themselves to motivated decisions regarding the selection and sequencing of educational activities. At present, SFL is the only serious contender for that distinction (contrary to the characterization in Tyler 2010). That requirement has the enormously important consequence of facilitating assessment practices that not only are intricately linked to languagefocused instructional practices; it facilitates another aspect of being appliable,

COLLEGIATE FOREIGN LANGUAGE PROGRAMME

225

namely a kind of rigour that has heretofore eluded language studies as a humanistic, meaning-oriented enterprise (see e.g. Ryshina-Pankova 2015a).

CONCLUDING REFLECTIONS Building on the preceding discussion, I offer these brief concluding reflections. First, an appliable linguistics that is ‘good to think with’ in an educational setting is most likely to fulfil its promise the more educators consider the situatedness of their educational efforts. A sober awareness of that ‘context of education’ will need to address and continue to monitor its external and internal dynamics, its possibilities and constraints, its unmistakable demands and its particular challenges. It is here that we may locate faculty development and graduate student socialization, which are all about who continues to do the thinking in line with an increasingly refined educational vision and assures that a theory is, in fact, appliable. In other words, far from ‘one size fits all’, even if one works with a functional theory of language like SFL, and far from ‘once an issue has been addressed it is settled’, such reflection will provide critically important guidance as well as ongoing impetus for exploring how a programme might best meet the threepronged challenge of language education: language as substance, in this case learning a particular foreign language on the part of a particular learner group; language as instrument, part of knowledge construction about another cultural area and in a cultural area where decisions regarding selection of what should and can be included are unavoidable; and language as object, which foregrounds its system and semiotic resource quality that awaits realization for instances of meaning making in actual communication. Regarding this latter, overarching demand, the key question to be answered is to what extent and how the texts that a particular programme makes available over its entire duration will enable learners to construe both the system of language and the context of culture within which such texts reside or, as Halliday (1999: 22) put it, ‘to predict the text from the context, and to predict the context from the text’. I have argued that this is first and foremost a curricular, not a pedagogical challenge, as the preponderance of the literature regularly portrays it. To be sure, curriculum is realized through pedagogies; therefore, enabling faculty to develop a rich palette of appropriate pedagogical resources is crucial. But pedagogies, if they are not to turn into simplistic templates, are extraordinarily situated – and varied – and must be so if they are to enable teachers to foster particular instances of learning. They do not readily provide a window on what we know about long-term instructed language learning, and in that sense they do not constitute the core of a linguistics that is appliable in an educational setting (Byrnes 2019a).

226

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Second – and an indispensable complement to the previous comment – a curriculum that endeavours to realize meaning-oriented language learning must build on the conceptual privileging of notions of choice, variation and probabilities. In the first instance, the centrality of situated meaningdriven choice – or, if you will, agency (for an excellent discussion from a chaos/complexity perspective, see Larsen-Freeman 2019) – is a local-level phenomenon that challenges learners to privilege, as best as they can, the paradigmatic axis over the syntagmatic axis of structural chaining. Unlike the customary foregrounding of rule-based syntax, it is this choice-oriented focus that is crucial for enabling them to develop a nimble facility with using the lexicogrammatical resources of a language in a trinocular way – above, below and round-about within the stratified system of language. Third, curricular thinking that is infused with the pivotal role of choice and, therefore, variation, will subscribe to an emergentist ontology. To be sure, it will lay out a trajectory according to which instructed learning might be facilitated and, in broad outline, might actually take place. But its underlying assumption is that of a complex emergence along different time frames, all of which must be held simultaneously – logogenetic, ontogenetic and phylogenetic – where the latter recognizes the increasing role of multilingual language use and language shift that characterizes contemporary societies (Matthiessen 2009). Fourth, such a stance takes seriously the possibility as well as the ethical responsibility that comes with the nature of language as a collective meaningmaking resource that is on offer for individual meaning making, including and quite specifically by L2 learners as language users. From that perspective, not only must the learning environment of the language classroom be treated as just as real as what happens outside of it; it must be realized in its ethical dimension towards engaging the whole person as a languaging, thinking and acting person. To me, it is this expanded sense that Halliday references when he states that an appliable linguistics ‘means a way of engaging with language that is theoretically robust and at the same time serviceable’ (2010a: 19). In the end, what has transpired in the GUGD programme is perhaps best described as a re-weighting of emphases, a privileging of certain perspectives, and an advantageous linking to other theories and practices that offer additional supporting evidence or facilitate worthwhile extensions. In that regard, the project continues a tradition in the now considerable history of SFL (see Matthiessen 2007) of providing more detailed maps and the conceptual and descriptive vocabulary for exploring certain regions of the overall topography of language use – here foreign language education – even as the overarching systemic quality of the theory, expanded early on into a Systemic Functional Theory that builds on three metafunctions, is sustained.

COLLEGIATE FOREIGN LANGUAGE PROGRAMME

227

REFERENCES Allen, H. W. (2010), ‘In Search of Relevance: The Role of the Standards in the Undergraduate Foreign Language Curriculum’, in V. M. Scott (ed.), Principles and Practices of the Standards in College Foreign Language Education, 38–52, Boston: Heinle Cengage Learning. Allen, W., K. Anderson and L. Narváez (1992), ‘Foreign Languages across the Curriculum: The Applied Foreign Language Component’, Foreign Language Annals, 25 (1): 11–19. Brinton, D. M., M. A. Snow and M. Wesche (1989), Content-Based Second Language Instruction, New York: Newbury House. Byrnes, H. (2001), ‘Reconsidering Graduate Students’ Education as Teachers: “It Takes a Department!”’, Modern Language Journal, 85 (4): 512–30. Byrnes, H. (2002), ‘The Role of Task and Task-Based Assessment in a ContentOriented Collegiate FL Curriculum’, Language Testing, 19 (4): 419–37. Byrnes, H. (2005), ‘Content-Based Foreign Language Instruction’, in C. Sanz (ed.), Mind and Context in Adult Second Language Acquisition: Methods, Theory, and Practice, 282–302, Washington, DC: Georgetown University Press. Byrnes, H. (2011), ‘Reconsidering Graduate Students’ Education as Scholar-Teachers: Mind Your Language!’, in H. W. Allen and H. H. Maxim (eds), Educating the Future Foreign Language Professoriate for the 21st Century, 17–42, Boston: Heinle Cengage Learning. Byrnes, H. (2012a), ‘Conceptualizing FL Writing Development in Collegiate Settings: A Genre-Based Systemic Functional Linguistic Approach’, in R. M. Manchón (ed.), L2 Writing Development: Multiple Perspectives, 190–218, Berlin: De Gruyter Mouton. Byrnes, H. (2012b), ‘Of Frameworks and the Goals of Collegiate Foreign Language Education: Critical Reflections’, Applied Linguistics Review, 3 (1): 1–24. Byrnes, H. (2015), ‘Linking “Task” and Curricular Thinking: An Affirmation of the TBLT Educational Agenda’, in M. Bygate (ed.), Domains and Directions in the Development of TBLT: A Decade of Plenaries from the International Conference, 193–224, Amsterdam: John Benjamins. Byrnes, H. (2018), ‘Advanced-Level Grammatical Development in Instructed SLA’, in P. A. Malovrh and A. G. Benati (eds), The Handbook of Advanced Proficiency in Second Language Acquisition, 133–56, Malden, MA: Wiley-Blackwell. Byrnes, H. (2019a), ‘Affirming the Context of Instructed SLA: The Potential of Curricular Thinking’, Language Teaching Research, 23 (4): 514–32. Byrnes, H. (2019b), ‘Applying SFL for Understanding and Fostering Instructed Second Language Development’, in G. Thompson, W. L. Bowcher, L. Fontaine and D. Schönthal (eds), The Cambridge Handbook of Systemic Functional Linguistics, 512–36, Cambridge: Cambridge University Press. Byrnes, H. (2020), ‘Toward an Agenda for Researching L2 Writing and Language Learning: The Educational Context of Development’, in R. M. Manchón (ed.), Writing and Language Learning: Advancing Research Agendas, 73–94, Amsterdam: John Benjamins. Byrnes, H., C. Crane and K. A. Sprang (2002), ‘Non-native Teachers Teaching at the Advanced Level: Challenges and Opportunities’, ADFL Bulletin, 33 (3): 25–34. Byrnes, H., H. H. Maxim and J. M. Norris (2010), Realizing Advanced Foreign Language Writing Development in Collegiate Education: Curricular Design,

228

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Pedagogy, Assessment. Modern Language Journal, 94 (Supplement S-1): iv–vi, 1–235. Byrnes, H., C. Crane, H. H. Maxim and K. A. Sprang (2006), ‘Taking Text to Task: Issues and Choices in Curriculum Construction’, ITL: International Journal of Applied Linguistics, 152: 85–110. Cammarata, L., ed. (2016), Content-Based Foreign Language Teaching: Curriculum and Pedagogy for Developing Advanced Thinking and Literacy Skills, New York: Routledge/Taylor & Francis. Douglas Fir Group (2016), ‘A Transdisciplinary Framework for SLA in a Multilingual World’, Modern Language Journal, 100 (Supp 2016), 19–47. Halliday, M. A. K. (1999), ‘The Notion of “Context” in Language Education’, in M. Ghadessy (ed.), Text and Context in Functional Linguistics, 1–24. Philadelphia/ Amsterdam: John Benjamins. Halliday, M. A. K. (2002), ‘Spoken and Written Modes of Meaning’, in J. J. Webster (ed.), On Grammar, 323–51, London: Continuum. Halliday, M. A. K. (2007), ‘A Language Development Approach to Education (1994)’, in J. J. Webster (ed.), Language and Education, 368–82, London: Continuum. Halliday, M. A. K. (2010a), ‘Pinpointing the Choice: Meaning and the Search for Equivalents in a Translated Text’, in A. Mahboob and N. K. Knight (eds), Appliable Linguistics, 13–24, London: Continuum. Halliday, M. A. K. (2010b), ‘Putting Linguistic Theory to Work’, Plenary presented at the 2010 AAAL Conference, Atlanta, GA, 6 March. Halliday, M. A. K. and C. M. I. M. Matthiessen (1999), Construing Experience through Meaning: A Language-Based Approach to Cognition, London: Continuum. Harley, B., P. Allen, J. Cummins and M. Swain, eds (1990), The Development of Second Language Proficiency, Cambridge: Cambridge University Press. Hyon, S. (1996), ‘Genre in Three Traditions: Implications for ESL’, TESOL Quarterly, 30 (4): 693–722. Larsen-Freeman, D. (2019), ‘On Language Learner Agency: A Complex Dynamic Systems Theory Perspective’, Modern Language Journal, 103 (Supp 2019), 61–79. Matthiessen, C. M. I. M. (2007), ‘The “Architecture” of Language According to Systemic Functional Theory: Developments since the 1970s’, in R. Hasan, C. M. I. M. Matthiessen and J. J. Webster (eds), Continuing Discourse on Language: A Functional Perspective, Vol. 2, 505–61, London: Equinox. Matthiessen, C. M. I. M. (2009), ‘Meaning in the Making: Meaning Potential Emerging from Acts of Meaning’, Language Learning, 59 (Supp 1): 206–29. MLA Ad Hoc Committee on Foreign Languages (2007), ‘Foreign Languages and Higher Education: New Structures for a Changed World’, Profession, 2007: 234–45. Norris, J. M. (2016), ‘Language Program Evaluation’, Modern Language Journal, 100 (Supp 2016): 169–89. Norris, J. M. and L. Ortega (2009), ‘Towards an Organic Approach to Investigating CAF in Instructed SLA: The Case of Complexity’, Applied Linguistics, 30 (4): 555–78. Norris, J. M. and P. C. Pfeiffer (2003), ‘Exploring the Use and Usefulness of ACTFL Oral Proficiency Ratings and Standards in College Foreign Language Departments’, Foreign Language Annals, 36 (4): 572–81. Norris, J. M. and R. M. Manchón (2012), ‘Investigating L2 Writing Development from Multiple Perspectives: Issues in Theory and Research’, in R. M. Manchón

COLLEGIATE FOREIGN LANGUAGE PROGRAMME

229

(ed.), L2 Writing Development: Multiple Perspectives, 219–45, New York: De Gruyter Mouton. Ortega, L. (2019), ‘SLA and the Study of Equitable Multilingualism’, Modern Language Journal, 103 (Supp 2019), 23–38. Paesani, K., H. W. Allen and B. Dupuy (2015), A Multiliteracies Framework for Collegiate Foreign Language Teaching: Theory and Practice in Second Language Classroom Instruction, New York: Prentice Hall. Pfeiffer, P. C. and H. Byrnes (2009), ‘Curriculum, Learning, and the Identity of Majors: A Case Study of Program Outcomes Evaluation’, in J. M. Norris, J. McE. Davis, C. Sinicrope and Y. Watanabe (eds), Toward Useful Program Evaluation in College Foreign Language Education, 183–208, Honolulu, HI: National Foreign Language Resource Center. Rose, D. (2016), ‘New Developments in Genre-Based Literacy Pedagogy’, in C. A. MacArthur, S. Graham and J. Fitzgerald (eds), Handbook of Writing Research, 2nd edn, 227–42, New York: Guildford Press. Ryshina-Pankova, M. (2010), ‘Towards Mastering the Discourses of Reasoning: Use of Grammatical Metaphor at Advanced Levels of Foreign Language Acquisition’, Modern Language Journal, 92 (2): 181–97. Ryshina-Pankova, M. (2015a), ‘Foreign Language Curriculum as a Means of Achieving Humanistic Learning Goals: Assessment of Materials, Pedagogy, and Learner Texts’, in J. M. Norris and J.McE. Davis (eds), Student Learning Outcomes Assessment in College Foreign Language Programs, 221–46, Honolulu, HI: University of Hawai’i National Foreign Language Resource Center. Ryshina-Pankova, M. (2015b), ‘A Meaning-Based Approach to the Study of Complexity in L2 Writing: The Case of Grammatical Metaphor’, Journal of Second Language Writing, 29 (1): 51–63. Scott, V. M., editor (2010), Principles and Practices of the Standards in College Foreign Language Education, Boston: Heinle Cengage Learning. Tedick, D. J. and P. M. Wesely (2015), ‘A Review of Research on Content-Based Foreign/Second Language Education in US K–12 Contexts’, Language, Culture, and Curriculum, 28 (1): 25–40. Tyler, A. E. (2010), ‘Usage-Based Approaches to Language and Their Applications to Second Language Learning’, Annual Review of Applied Linguistics, 30: 270–91.

230

CHAPTER THIRTEEN

Putting Systemic Functional Linguistics (SFL) and Sydney School Genre Pedagogy (SSGP) to Work in a Compulsory First-year Arts Subject SHOSHANA DREYFUS AND TRISH WEEKES

BACKGROUND This chapter reports on the development of a compulsory first-year academic literacy subject for students entering the Bachelor of Arts programme at the University of Wollongong (UOW). The opportunity to develop and deliver a compulsory, programme-wide, credit-bearing, academic literacy subject based on Systemic Functional Linguistics (SFL) (Halliday and Matthiessen 2014) and Sydney School Genre Pedagogy (SSGP) (Rose and Martin 2012; Dreyfus et al. 2016) is a great achievement, as many colleagues have struggled for years to get SFL-based academic literacy adopted by faculties and subject lecturers. This chapter describes the process in order to share the successes and challenges and how we have dealt with some of them.

232

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

The story begins in 2017 when I, Shooshi, started as Senior Lecturer in English Language and Linguistics at UOW. A couple of years before this, the linguistics programme had been downgraded from a major to a minor, which meant linguistics was potentially in danger of being erased altogether. In conjunction with colleague Alison Moore, I decided that one of my aims would be to try to get linguistics reinstated as a major and was told I had to make a ‘business case’. In effect this means ‘putting bums on seats’ – that is, increasing the number of students and showing that there are enough students to warrant a major in a context where the number of arts students is declining globally. With this in mind, I looked for opportunities to build awareness of linguistics and promote its application across the university. This chapter details the process of establishing this new compulsory academic literacies subject for first-year arts. It details the development of the concept for the unit, collaboration with discipline academics, analysis of student writing and assessment task prompts, and design of the subject. The chapter concludes with a discussion of some of the theoretical issues that have arisen from the project and explores some possible ways forward.

GETTING THE SUPPORT OF THE HEAD OF SCHOOL (HOS) One of the elective subjects we teach in linguistics at first-year level is an academic literacy subject named Effective Academic Writing. Based on SFL and SSGP, it was originally designed in the 1980s by Susan Feez, Sally Humphrey and Elizabeth Thomson, and has been continually updated by others. It was based on the considerable work done within the DSP project Write it Write, which focused on mapping disciplinary knowledge and discourse practices in high schools in NSW, Australia (e.g. Rothery 1994; Christie and Martin 1997; Rose and Martin 2012). Each time I met with our HoS, I talked about this subject and its importance and value for students. Given the general decline in arts students and the firstyear attrition rate, there is a lot of pressure on universities to improve the ‘firstyear experience’ – and one way that this can be achieved is through improving academic skills. I knew that the HoS would be under pressure, particularly in a regional university with a large international student population. Additionally, academics in my university often talk about students’ poor writing and literacy skills. In addition to the Effective Academic Writing subject, for the past four years there has been a compulsory first-year arts subject called LHA 101, which was a generic introduction to university study – including how to use Moodle and the library, as well as some wellbeing, critical thinking, group work and cohort building, and academic skills tuition. A number of students had complained

SFL AND GENRE IN FIRST YEAR ARTS

233

about this subject and I thought it would be more productive if students had an academic literacy subject as their compulsory first-year subject. I thus proposed to the HoS that we design a second compulsory subject, with a focus on building academic literacy skills specific to the arts and humanities. She told me that we could not have two compulsory subjects as students did not like the fact that they already had one. Nevertheless, at the beginning of 2019, she decided to change the current compulsory unit to make it primarily focus on building discipline-specific academic literacy skills and asked me to develop it. She gave me funds for a research assistant and teaching relief.

THE PROCESS The process began of building the subject. It had to connect to all the disciplines in the school at first-year level for roll-out the next year (2020). The first step was to choose a colleague to work with. I did not want a junior but someone with equal or even more experience in teaching generic and specific academic skills across a variety of disciplines and who was experienced in developing SFLand SSGP-based courses, materials and resources. I thus asked Trish Weekes. Trish is an experienced literacy educator in the SFL and SSGP tradition, who works in secondary and tertiary contexts. There were several processes that ran together over the course of the year: collecting and analysing data, mapping discourse practices and running workshops with discipline colleagues, based on developing with them an SFL framework and metalanguage for us to collectively draw on. The brief from the HoS was that the new subject had to relate to all first-year subjects across the ten arts disciplines: cultural studies, English, Indigenous studies, international studies, history, linguistics, philosophy, politics, science and technology studies, and sociology. The HoS led the project and gave directives to the discipline lecturers to provide examples of essay prompts and student work, and to participate in my workshops. Collecting data The next step was to collect the subject outlines and assignment information for every first-year subject. With the help of faculty admin support, we created a spreadsheet with every subject name and associated assignments. I asked lecturers to send me all assignment prompts/questions, as well as examples of high-scoring student writing responding to these. Engaging with discipline colleagues I knew from my previous research projects in SFL, such as the SLATE project (see Dreyfus et al. 2016), and the experiences of colleagues that working with

234

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

discipline specialists can be tricky. I also knew from my work in disability that to get buy-in and support for a project from the people involved, you have to let them be the expert and not come in and tell them how you think they should do things. These ideas informed my approach to the whole process. I decided to run a series of three workshops with each discipline. The foci of the workshops were student writing, essay prompts and marking criteria. Workshop 1: Deconstructing high-scoring essays  It was clear that we needed to get academics to reveal what they valued in the student texts they scored highly. Thus, the first workshop was a group discussion of one of their first-year high-scoring student texts which they said quintessentially represented good first-year disciplinary writing. I asked them what features they valued highly and why the student had received good marks. As lecturers commented, I wrote on the whiteboard and organized their ideas according to genre and register, as shown in Figure 13.1. My aim was to build useful SFL-based metalanguage with staff as I went, by saying things like, ‘This is what we call tenor.’ As can be seen in Figure 13.1, I used the SFL concepts of register (field, tenor and mode) and genre to organize lecturers’ ideas. During this process, the metafunction framework emerged as organic to what lecturers valued, and

FIGURE 13.1  Workshop 1 whiteboard notes.

SFL AND GENRE IN FIRST YEAR ARTS

235

they could see that I was not imposing an external framework but actually reorganizing their comments in a more systematic way. The first workshop created an opportunity to discuss important questions with their discipline colleagues such as: • What is (valued) first-year discourse practice? • What are we looking for in student writing? • What do we expect from students? • What do we want them to produce? • What concepts are foundational and what should be saved until second year? These kinds of discussions had never been held before in any of the disciplines. Interestingly, the HoS had only planned to attend the first workshop in each of the series; but she ended up attending them all. She told me that she could not keep herself away because she found them fascinating and illuminating about what was going on inside the disciplines. When there was disagreement among discipline lecturers, the HoS urged them to continue the discussion in their own discipline meetings. Workshop 1 showed that among lecturers there were some commonalities. All: • value a well-organized text; • want students to include a text preview in their introduction (even if they did not use that metalanguage directly); • like a confident student writer; • want students to weave external voices into their texts (but there are disciplinary differences in which voices are valued); • want the field to be rendered in discipline-specific ways (although sometimes lecturers could not agree on what that was). Workshop 2: Essay prompts  The aim of the second workshop was to help lecturers see that how they write essay prompts matters and that prompts point students towards writing particular genres. This workshop aimed to make visible the connection between essays and questions/prompts and lecturers’ actual intentions. Before the second workshop, we collected and analysed every essay prompt in first year. In this workshop, I introduced discipline lecturers to an analysis tool (developed from analysis of the essay prompts) and asked them to analyse their own essay prompts. We discussed how they were construing the field in those questions. I asked them if the students were responding to the prompts in expected ways and, if not, how the prompts could be improved. Workshop 2 was a revelation for discipline lecturers, who often write prompts without being conscious of the language used to construct them.

236

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Following the workshop, some discipline lecturers asked me to review their prompts to ascertain whether they were clear and met their intended purpose. Workshop 3: Marking criteria  The intention of workshop 3 was to link student writing (workshop 1) and essay questions (workshop 2) to the marking guidelines/rubrics created by the lecturers. I collected all the marking guides and rubrics from first-year assignments. Marking guides tell students what they will be marked against, with criteria such as: ‘has topic sentences that link to the text preview’. Rubrics provide details about levels of achievement (e.g. excellent, very good) and what is required to achieve that level, such as, for a pass level, ‘some topic sentences link to the text preview’. With the help of my colleague, Emily Purser, I sorted the materials into marking guides and rubrics. From our preliminary analysis of all the collected marking guides/rubrics, it was clear there is very little consistency across the school; there are vast differences among the types of information given to students about how their work is marked. This is evidently an area that needs more work in the interest of improving student outcomes. Data analysis The data included all the information about first-year writing practices: subjects, assessment tasks/prompts and high-scoring student responses. Organizing the data  First, we created a large spreadsheet containing all subject names, assignment questions/prompts and notes about our genre mapping. Genre analysis  Together Trish and I analysed high-scoring student responses, identifying their genre from key language features at both discourse and lexicogrammatical strata. Given the amount of data we could not undertake a comprehensive itemized analysis but pursued more a holistic approach based on our experience in text analysis. The genre of some texts was easy to identify, e.g. analytical exposition (Martin and Rose 2008). Some texts were harder to identify because they had many features of different well-recognized genres, e.g. combining explanation with arguments. For these texts, we conducted more detailed analysis of their language features to try to work out what was going on in the text. We analysed high-scoring student work for every assignment prompt in every subject across: • 10 disciplines • 21 first-year subjects • 28 different extended writing tasks, often with numerous questions and parts, most of which are called an ‘essay’.

SFL AND GENRE IN FIRST YEAR ARTS

237

FIGURE 13.2  Genre analysis spreadsheet.

We then created Table 13.1, which maps these according to genre. We found nine individual genres plus genre complexes, where a text comprises a number of separate genres joined to form a whole text, and macrogenres, where several genres are embedded in a text at the service of an overall higher-order genre (see Martin 1992, 1994, 2002; Martin and Rose 2008; Szenes 2017). Implications of the genre analysis and lecturer workshop on genre  The analysis found a wide array of genres that students have to write in first-year arts. We also found that lecturers ask students to produce a range of genres and often neither the lecturers nor the students are very aware of these. The genre analysis gave us an appreciation of this range and a better appreciation of disciplinarity discourses in the arts/humanities, as made visible by our colleagues in both school contexts (e.g. Christie and Martin 1997; Christie and Derewianka 2008; Martin and Rose 2008) and tertiary contexts (e.g. Dreyfus et al. 2016). Our disciplinary map of genres contributes to SFL understandings of disciplinary knowledge and discourse practices. In the workshop on student texts (workshop 1), it was clear that lecturers do not have a shared metalanguage for talking about language and text.

238

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 13.1  Genres in First-Year Writing in Arts Disciplines at UOW Social purpose

Genre family

Genre

Subject number

To describe and classify phenomena

report

long report

Aust102

definition

Aust101

To show how or why something/s happen/ed/s

explanation

causal/factorial

Aust101, Aust102, HIST120*

consequential

INTS100, SOC103

historical account

HIST111, HIST120*, INTS100

To interpret meanings in a text

response

interpretation

ENGL120, ELL110, INTS121, ENGL131

To persuade someone to think or do something

argument

one-sided:a. Hortatory exposition

STC100, INDS150

one-sided:b. Analytical exposition

Aust101, CST120, HIST111, HIST112, HIST120*, INTS121, PHIL106, PHIL107, POL150, SOC103, STC100

genre complex

INTS100 (factorial+conseque ntial+argument) PHIL151 (report+argument) INTS121 (report+argument) POL150 (report+argument)

macrogenre

POL150 (argument with embedded historical recount)

Source: Adapted from Martin and Rose (2008); Szenes (2017).

Assignment prompts  The next source of data was the assignment prompts from first-year arts. We analysed the different parts of every assignment prompt from every first-year arts subject, including the number and contents of parts. We built on research from the University of Sydney Learning Centre (e.g. Webb 1991) about parts of essay prompts and further developments by Trish Weekes (2019). We found that essay prompts range from a one-part question to five or more parts. As we can see, some prompts attempt to provide more scaffolding for students than others.

SFL AND GENRE IN FIRST YEAR ARTS

239

TABLE 13.2  Parts of Essay Prompts Parts

Prompts

Discipline

1 part

Why were the Portuguese so successful in taking over the spice trade?

History

2 parts

How and why did the first wave feminist movement emerge?

History

3 parts

How reliant is the world on fossil fuels and how did we become ever more reliant on them as a primary energy source? What hope is there for renewable energy in the 21st century?

International Studies

4 parts

In the film Memento, the protagonist says that ‘memory is just an interpretation, not a record of events’. Gaut argues that the film itself shows us that this claim about memory is true. Is Gaut right? Explain.

Philosophy

5 parts

Outline the main strengths and weaknesses of Copernicus’s theories from (as far as possible) the standpoint of an educated person of that time (mid16th Century). Points to consider when writing the essay: What were the main features of the Ptolemaic/ Aristotelian world view? What were the main features of Copernicus’s theories? What difference it might have made if you were sympathetic to Neo-Platonist ideas? How significant were possible religious objections?

Society, Culture and Technology

In analysing types of parts, we started with existing classifications (e.g. Webb 1991; Weekes 2019), but subsequently developed a more nuanced analysis based on field and genre, as summarized in Table 13.3. Task is an obligatory stage that tells students what to do, and is realized by a command verb or question word (either Wh- or polar interrogatives). • e.g. the ‘command’ verb ‘Discuss’ in: If it were not for the leadership of Otto von Bismarck, Germany would not have unified to become a nation. Discuss. • e.g. the Wh- interrogative: What was so new about Enlightenment thought?

240

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 13.3  Functional Parts of Essay Prompts 2 stages

Types

Function/composition

Task

Tells students what to do (command verb or question word) (either polar or Wh-) Points to genre

Scope Focus of (identifies the study field)

Identifies the phenomenon to write about (can also be instantiated as a statement, orientation, scenario or data display)

Lens

Tells how to examine the phenomenon

A range of command verbs were found across disciplines with little evidence of any disciplinary patterning. The most common command verb was ‘discuss’, which occurred twenty-three times. During workshops with discipline lecturers, understandings of the meaning of ‘discuss’ varied. For example, in history, discuss meant to consider the arguments of both sides and come to a conclusion at the end (similar to the understanding of discussion genres in SFL/SSGP). For example, in history, ‘discuss’ meant to consider the arguments of both sides and come to a conclusion at the end (similar to the understanding of discussion genres in SFL/SSGP). However, in cultural studies, ‘discuss’ indicated that the genre was open to student choice. The other most common command verbs were ‘explain’ (four instances), ‘outline’ (three instances), ‘argue’, ‘choose’ and ‘evaluate’ (two instances), and the following only had one instance: identify, describe, provide, consider, pick, assess, find, analyse, present, use. Regarding patterns of task words, how and what are the most commonly found in the humanities, with 24 instances of what and 22 instances of how. In some cases, task words pointed students towards a particular genre for their written responses. For example, explain and how appear to direct students to write an explanation genre. However, the link between the prompt and the genre students were expected to write was not always clear. Nominalized versions of how such as implications and consequences were often coupled with task words like what or discuss, meaning the target genre was somewhat opaque. Consider: • STC 100 • Mark Zuckerberg said privacy is no longer a social norm. Is he right? And if so, what are the consequences of this for society? In this question, Is he right? points students towards an argument – as they need to take a position on whether or not Zuckerberg is right. It is implied

SFL AND GENRE IN FIRST YEAR ARTS

241

that he is right, because the second part of the question begins with and if so. This is an example of giving an implied position to the students for an argument. The last part of the prompt And if so, what are the consequences for society? seems to require an explanation of consequences. It is unclear if students are meant to integrate these two aspects into one high-level argument text response (macrogenre), or whether they should write two separate parts (genre complex). Further, if students choose to disagree with the statement, suggesting Zuckerberg is not right, then how can they possibly explain it (if it is ‘not so’)? Another issue we found was that some instances of how and why questions do not point to an explanation genre. This can be seen in the following question from international studies: How reliant is the world on fossil fuels and how did we become ever more reliant on them as a primary energy source? What hope is there for renewable energy in the 21st century? In this question, the second how suggests an explanation; but the first how is really a to what extent question, thus an argument, and not a true how question. Scope  The scope is another obligatory part of a prompt. It covers the field of the question, that is, the content, issue or topic for the assignment. The Scope can be realized by nominal groups. e.g. How does the text in question present an ambivalent attitude to nation or national literature? (English) In this example, the nominal groups are the text in question and an ambivalent attitude to nation or national literature. The Scope can also be shown in disciplinary processes, as in this question from history: How did the Manchus successfully invade and colonize China? In this example, Scope is also realized by couplings of ideational and interpersonal meanings, as shown in the processes invade and colonize combined with the appraisal resource successfully. When we analysed the data, there was a lot more going on with Scope than simply ‘content and topic’, as so often described in the literature. It was clear that Scope needed more analysis. The following prompt from cultural studies illustrates how we developed a new model of Scope: Australia Day is an example of an ‘invented tradition’. Discuss.

242

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Using previous models of analysis, the Scope would have included Australia Day and an example of an invented tradition. This does not differentiate between what students are examining and how they are supposed to look at it. Scope is better analysed as containing two parts, which we label Focus of study and Lens. The Focus of study is the phenomenon being examined and could potentially be examined by any discipline. The Lens provides the angle on that Scope, as valued by the discipline. In the example above, Australia Day is the Focus of study. In contrast, the Lens is the way students need to look at Australia Day – that is, as an invented tradition. In the discipline of cultural studies, taken-for-granted social practices such as Australia Day are problematized and re-examined as invented traditions. We did find that in some questions there is no explicit Lens – such as in this history question: If it were not for the leadership of Otto von Bismarck, Germany would not have unified to become a nation. Discuss. In this question, it is unclear whether the Focus of study is the unification of Germany or Otto von Bismarck. The history lecturer suggested that the Focus of study is the unification of Germany and the Lens is the leadership of Otto von Bismarck. Implications of analysis of question prompts  The analysis of question prompts enabled us to learn more about the parts of an essay question, particularly in adding greater delicacy of analysis for Scope. We have also highlighted the importance of bringing prompts into discussions about genre. For students, there is no doubt that questions and prompts are complex, and the demands of an assignment can be tacit. From lecturers’ perspectives, the project as a whole and workshop 2 in particular (which covered the prompts) brought to consciousness the need for intentional language use in writing prompts. The next section covers the design of the new first-year subject in light of the analysis presented above.

DESIGN OF THE SUBJECT LHA 101 Theoretical underpinnings for the subject design A decision was made early on in the process to design the unit focusing on key genres students have to write in arts using SSGP/SFL and deploying a spiral curriculum (Bruner [1960] 1977), continually reinforcing simpler tasks as a foundation for more difficult ones. We began with genres that can be written as shorter texts and language resources that can be developed and built on before moving on to more demanding genres of factorial and consequential

SFL AND GENRE IN FIRST YEAR ARTS

243

explanations and finally expositions. Another approach informing design was backward mapping, where we start with the end result (e.g. a student’s final assignment in the unit) and then work backwards to how to help students reach that final goal. There are many theories which use the term ‘genre’ (Hyons 1996), but in a more general sense of the purpose of the text (Swales 1990; Starfield 2019), rather than from an SFL perspective. Much of the research work in an SFLinformed approach genre has been at the school level (Christie and Derewianka 2008; Martin and Rose 2008) with the DSP work being conducted almost thirty years ago (see Rose and Martin 2012); and while there has been some SFL research into genres at university (Drury and Webb 1991; Byrnes, Maxim and Norris 2010; Nesi and Gardner 2012; Humphrey and Economou 2015) and disciplinary study of genre in particular faculties such as business (Szenes 2017) and cultural studies (Hood 2010), there has not been a lot of cross-disciplinary research on arts in tertiary contexts. Our aim for the subject is not to cover every single arts genre, but to build awareness of discipline-specific tertiary language use. We wanted to equip students with the skills to decode assignment prompts and know how these relate to common genres of writing in arts/humanities. We also wanted to build a students’ awareness of field, tenor and mode in the academic domain and what it means to write in discipline-specific ways – like a philosopher, or a sociologist or a historian – and to understand similarities and differences. We believe these are critical parts of an academic apprenticeship (Woodward-Kron 2004; Christie and Maton 2011; Dreyfus and Weekes 2019). The unit was built around a sequence of learning activities following the SSGP teaching and learning cycle (Rothery 1994; Rose and Martin 2012). This included building the field through reading and lectures; modelling and deconstruction of texts in lectures and tutorials; joint construction of texts during tutorials; and group construction of texts before independent construction in assessments. Developing the unit Trish and I met for a series of whole-day sessions where we developed the subject’s frameworks and content. In the first session we covered the aims, outcomes and overall shape of the twelve-week unit and the assessment tasks. We decided to focus on three genres within the stipulated constraint of students writing 6,000 words in total per subject. Based on our analysis, we decided to focus on three of the most common genres: definition (sub-type of information report), explanation (both factorial and consequential types) and analytical exposition. We decided that the analytical exposition would be the final assignment, modelled as a macro-genre,

244

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

with an embedded definition and explanation. The definition would be taught first as it was short and fairly straightforward, followed by explanations in the middle of the unit. For each of the three genres, we developed a 3×3 matrix identifying key discourse features in ideational, interpersonal and textual meanings at the level of whole text, discourse semantics and lexicogrammar (after Humphrey et al. 2010; Dreyfus et al. 2016). We then selected the most salient features to teach. Next we sequenced the learning content for the unit, week by week, increasing complexity as we went on. It should be noted that we also had to teach reading and referencing, so these aspects were added at the most logical points in the unit. The backward mapping process and SSGP involved writing model texts, annotating them for generic structure, discourse and language features. Next, we wrote model texts in the target genres. This was a challenging task, as we needed to choose topics that were relevant to humanities students in all disciplines and did not require too much lecture time for field building. For the definition genre, we drew on the disciplines that students would be studying at university. The relevant readings that we used were the subject outlines on the university website and a reading about disciplines and the ‘three cultures’ by Peters (1985). The explanation genre proved to be the most demanding and challenging to write: we tried and failed to write a model factorial one on several topics, e.g. to respond to the question: How do the soft skills in the humanities help a graduate get ahead in their career? However, we realized that our text was too close to an argument, as it ended up presenting the benefits of soft skills rather than explaining the factors that result in a phenomenon. In the end, we wrote a factorial explanation prompt about the causes of the 2019 Australian bushfires and a consequential explanation prompt on the impacts of Indigenous burning practices. Interestingly, when analysing explanations and arguments, we also found that some prompts had a clear ideological position, while some did not. Some prompts, such as polar interrogatives, explicitly point students towards an argument essay, asking them to take a position, e.g. Is he right? Do you agree? Others, such as Wh- questions, contain the argument within the prompt without overtly declaring that, such as in this question from Indigenous studies: How is colonialism ongoing in the present day? In response to this, students are expected to write an argument that supports this position, without being explicitly told by the lecturer that that is the case. This whole process deepened our understanding of the discipline-specific nature of genres and confused us regarding the genre typology framework that we had previously worked with. The clear demarcation between explanations and expositions, which we had thought we understood, now seems less clear. We have been led to wonder whether, at the tertiary level, every text that

SFL AND GENRE IN FIRST YEAR ARTS

245

students write is in some sense arguing a case. In this context, genres could be arranged more topologically rather than typologically. We wonder whether our understanding of the difference between these genres is misled by the terms and will continue to pursue this line of thought in our analysis and thinking about this project. For each assessment task in the new unit, we developed the question, wrote the answer and analysed it. In addition, we wrote a joint construction text that was similar and would be used in tutorials. And we prepared detailed marking criteria that matched what was taught in the lectures and tutorials – aligning genre, structure, discourse and language features. Each assignment requires that students annotate their text for the features we teach them – such as genre stages and phases, previews and themes, and expanded nominal groups. Almost half the marks for each assignment are allocated to these annotations. In our approach to this unit, learning how to analyse one’s own writing is as much part of the learning as writing the text. At the time of writing, the subject has been taught twice (both autumn and spring sessions of 2020), with around 300 students in autumn and 100 in spring. We can report that many students find the analysis difficult and so we are in the process of designing many more practice analysis exercises to be embedded in Moodle as ‘learning objects’. We also know that many students learned a lot from the subject as evidenced by the following testimonial: The amount of work that you have put into the subject workbook astounds me. There is a wealth of information in there that I know I will refer to throughout my studies. When I first started, I looked through the book and never thought I’d manage to finish, let alone understand the work and enjoy it. As I said in the survey, the gentle introduction of each topic leading to the next was so well done, I didn’t even realise along the way that I was learning. Having exercises to do throughout to cement the information also made it fun and interesting. As a mature aged student with absolutely no knowledge of what was required regarding academic writing, this course has not only provided me with the tools to learn this skill, it has given me the confidence to know that I’m capable of achieving something. Without this course, and had I not done it as my first subject at university, I know I would have floundered and given up. This subject should be the first subject done in the first year for every student studying at university! Each week’s content was included in a handbook for students. This included: • lecture notes • model texts • deconstruction activities

246

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

• joint construction activities • readings • links to relevant materials • assessment tasks, marking guides and marking criteria. A handbook for tutors was also prepared, containing all the answers for worksheets and activities, tutorial teaching guides and joint construction task models. Finally, Covid-19 has also had an impact on the rollout of the subject as, within a two-week period, we had to turn the subject into an online version of itself.

IMPLICATIONS The development of the new compulsory academic literacies unit for firstyear arts has been an interesting process and it is ongoing. We have raised consciousness about the nature of disciplinary writing in first-year arts for all concerned. And we have built significant relationships with our colleagues, the discipline lecturers. With respect to collaboration, the support of the HoS was essential for the success of this project. Discipline lecturers are open to discussions about writing, but they lack a common metalanguage for these conversations. By focusing directly on the actual work that they ask their students to write and on the essay prompts that they themselves construct, we were able to engage lecturers in the workshop process. It is hoped that the interdisciplinary conversations at the university can be continued this year and beyond, as the unit is taught. The perennial problem of dealing with busy and protective academics persists, as in any university setting. This has all been amplified by Covid-19. More data from other universities and faculties are needed in order to develop disciplinary genre maps that can inform theory and practice in academic skills for first-year students. Three issues are of particular theoretical interest to us: the concept of a genre typology/topology. First, while some work has been done to map the genres of academic writing at university, more needs to be done in this area. As explained above, we found that explanations and arguments are quite close in the humanities at tertiary level, though it has been argued that we were confused by terminology rather than texts. We intend to pursue this line of research to trying to gain more clarity about tertiary arts/humanities genres. A second issue involves the disciplinary values in assignment prompts, and whether or not students are required to take an ideological position. The third issue is around the link between assignment prompts and the genres students are expected to write in relation to these. We need to know more about these

SFL AND GENRE IN FIRST YEAR ARTS

247

links and find ways to make the requirements of assignment prompts and the demands of academic writing accessible and available for both students and their disciplinary lecturers. We hope that more linguists can collaborate with disciplinary colleagues on projects like this that aim to build knowledge about language at university.

REFERENCES Bruner, J. ([1960] 1977), The Process of Education, Cambridge, MA: Harvard University Press. Byrnes, H., H. H. Maxim and J. M. Norris (2010), ‘Realizing Advanced Foreign Language Writing Development in Collegiate Education: Curricular Design, Pedagogy, Assessment’, Modern Language Journal, 94 (Supp 1): 1–235. Christie, F. and B. Derewianka (2008), School Discourse, London: Continuum. Christie, F. and J. R. Martin (1997), Genre and Institutions: Social Processes in the Workplace and School, London: Cassell. Christie, F. and K. Maton, eds (2011), Disciplinarity: Functional Linguistic and Sociological Perspectives, London: Continuum. Dreyfus, S., S. Humphrey, A. Mahboob and J. R. Martin (2016), Genre Pedagogy in Higher Education: The SLATE Project, Basingstoke: Palgrave Macmillan. Dreyfus, S. and T. Weekes (2019), ‘Are You the Next Apprentice?’, Universe, UOW, November. Available online: https://universe.uow.edu.au/education/are-you-thenext-apprentice/ (accessed 18 December 2020). Drury, H. and C. Webb (1991), ‘Teaching Academic Writing at the Tertiary Level’, Prospect, 7 (1): 7–21. Halliday, M. A. K. and C. M. I. M Matthiessen (2014), An Introduction to Functional Grammar, 4th edn, London: Routledge. Hood, S. (2010), Appraising Research: Evaluation in Academic Writing, London: Palgrave Macmillan. Humphrey, S. and D. Economou (2015), ‘Peeling the Onion – A Textual Model of Critical Analysis’, Journal of English for Academic Purposes, 17: 37–50. Humphrey, S., J. R. Martin, S. Dreyfus and A. Mahboob (2010), ‘The 3x3: Setting Up a Linguistic Toolbox for Teaching and Assessing Academic Writing’, in A. Mahboob and N. Knight (eds), Appliable Linguistics: Texts, Contexts and Meanings, 185–99, London: Continuum. Hyons, S. (1996), ‘Genre in Three Traditions: Implications for ESL’, TESOL Quarterly, 30 (4): 693–722. Martin, J. R. (1992), ‘Macro-proposals: Meaning by Degree’, in W. C. Mann and S. Thompson (eds), Discourse Description: Diverse Analyses of a Fundraising Text, 359–95, Amsterdam: Benjamins. Martin, J. R. (1994), ‘Modelling Big Texts: A Systemic Functional Approach to Multigenericity’, Network, 21: 29–52. Martin, J. R. (2002), ‘From Little Things Big Things Grow: Ecogenesis in School Geography’, in R. Coe, L. Lingard and T. Teslenko (eds), The Rhetoric and Ideology of Genre: Strategies for Stability and Change, 243–71, Cresskill, NJ: Hampton Press. Martin, J. R. and D. Rose (2008), Genre Relations: Mapping Culture, London: Equinox.

248

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Nesi, H. and S. Gardner (2012), Genres across the Disciplines: Student Writing in Higher Education, Cambridge: Cambridge University Press. Peters, P. (1985), Strategies for Student Writers: A Guide to Writing Essays, Tutorial Papers, Exam Papers and Reports, Milton, QLD: John Wiley & Sons. Rose, D. and J. R. Martin (2012), Learning to Write, Reading to Learn. Genre, Knowledge and Pedagogy in the Sydney School, Sheffield: Equinox. Rothery, J. (1994), Exploring Literacy in School English, Sydney: Metropolitan East Disadvantaged School Program. Starfield, S. (2019), ‘Student Writing in Higher Education’, in X. Gao (ed.), Second Handbook of English Language Teaching, 1–22, Cham: Springer. Swales, J. (1990), Genre Analysis, Cambridge: Cambridge University Press. Szenes, E. (2017), ‘The Linguistic Construction of Business Reasoning: Towards a Language-Based Model of Decision-Making in Undergraduate Business’, PhD thesis, University of Sydney, Sydney. Webb, C. (1991), Writing an Essay in the Humanities and Social Sciences, Sydney: Learning Assistance Centre, University of Sydney. Weekes, T. (2019), Literacy Works for Secondary School Subjects, Sydney: Literacy Works. Woodward-Kron, R. (2004), ‘“Discourse Communities” and “Writing Apprenticeship”: An Investigation of These Concepts in Undergraduate Education Students’ Writing’, Journal of English for Academic Purposes, 3 (2): 139–61.

PART III

Evolving Theory

250

CHAPTER FOURTEEN

Modelling Newborn Screening Genetics Diagnosis Texts Language as a Tool for Research and Treatment of Sickle Cell Disease GIACOMO FIGUEREDO AND KELEN CRISTINA SANT’ANNA DE LIMA

INTRODUCTION1 In this chapter, we report on an application of linguistic analysis and modelling to solve a problem in the field of the Center for Newborn Screening and Genetics Diagnosis (NUPAD) and the diagnosis and treatment of sickle cell disease (SCD) (Lima 2013). In particular, we detail how Systemic Functional Linguistics (SFL) was deployed to create text templates (Figueredo 2014; Figueredo and Figueredo 2020) which were used to produce more adequate research articles, technical manuals and pamphlets on the diagnosis and treatment of SCD. SCD is a chronic genetic condition present in newborn children. Thus, research on SCD and its treatments involves researchers, doctors and nurses (specialists), hospital personnel (technicians) and parents and extended family (laypeople). Training better professionals and educating

252

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

the family on daily healthcare is fundamental to the well-being of patients (Kikuchi 2007). However, this poses a challenge for building and sharing knowledge on treatment and care – which places a number of demands on public policy, education and awareness raising (Gomes et al. 2013). For the field of linguistics in particular, this creates the challenge of how to produce texts appropriate for dealing with the knowledge problems of specialists, technicians and laypeople (Lima 2013). In the context of our application (in the state of Minas Gerais, Brazil), this problem was tackled in a joint project between the Laboratory for Experimentation in Translation (LETRA) and NUPAD in the Federal University of Minas Gerais. In the course of this project, we found that our contribution could only be made possible by developing templates of texts which could then be written by following a model (pipeline). A ‘pipeline’ (Guseva et al. 2020) is a technical term taken from Natural Language Processing which includes all steps needed to process a text (chunking, parsing, tagging, counting probability and sequencing categories) and to use it to produce another document (annotation, text, protocol). So we began by developing the idea that efficient models of texts should be an important part of SCD management. Our goal is to use text production to increase efficiency in building and transferring knowledge between specialists, technicians and laypeople (Lima 2013) – for example, by writing papers that are accepted by well-regarded journals or writing pamphlets that help people adhere to treatment for longer.

CONTEXTUALIZING OUR APPLICATION Managing sickle cell disease: A problem for healthcare Sickle cell disease (SCD) is the most prevalent genetic disorder in the world (Aygun and Odame 2012). In this condition, a genetic mutation causes red blood cells to re-shape into the form of a sickle (instead of their characteristic rounded shape). Sickle-shaped red blood cells do not flow as smoothly in the bloodstream and tend to cluster together. This may cause episodes of vase obstruction, acute pain, inflammation and organ damage. SCD is a chronic condition and patients need lifelong care. Because patients are born with the condition, the family is a fundamental part of care since they help maintain patients’ good general physical and psychosocial health. Alongside family, professional monitoring on a regular basis is also essential. Primary care in particular is extremely important, thus ‘training programs for primary care professionals are imperative’ (Gomes 2015: 9).

NEWBORN GENETICS DIAGNOSIS TEXTS

253

Our application was carried out in Minas Gerais, Brazil. In Minas Gerais, the incidence of SCD is 1 case for every 1,400 newborn children. Accordingly, it is considered a public health issue. In Minas Gerais, management of SCD involves different groups of people on a regular basis. In broad terms, these people can be divided into healthcare professionals on the one hand and the families and community of SCD patients on the other. Healthcare professionals include researchers, doctors, nurses and hospital staff. Families include parents and members of the extended family; the community may include teachers, neighbours and friends. Family members need to be able to care both physically and psychosocially for patients. For example, they need to learn how to avoid dehydration and intense cold or heat, prevent infections, manage pain and regularly follow-up on medical consultations (Kikuchi 2007). They also need to help with psychosocial problems that acute episodes of SCD may cause, like missing school or lapsing social relationships. However, families may have problems with learning how to identify symptoms of SCD. They also have to battle against the stigma of SCD patients as being academic and social failures (Marques et al. 2015) and the lack of awareness the general public has about SCD. Healthcare professionals need training to diagnose symptoms, provide treatment and follow up on their patients’ history. Most importantly, they need training on how to establish a strong social bond with patients and families (Weis et al. 2013). However, several studies show that this is not always the case. Gomes et al. (2013), for example, show that less than 75 per cent of specialists (doctors and nurses) and 65 per cent of technicians correctly answered questionnaires on SCD diagnosis and treatment. Managing sickle cell disease: A demand on appliable linguistics In the context of SCD diagnosis and lifelong care, family members need knowledge about daily care; and trained professionals need to be able to provide quality health services for patients. This context shapes SCD management – which, in turn, relies on building knowledge and educating people. In the context of our application, this poses an important problem for linguistics – namely, finding a way to produce texts that can effectively create and store knowledge on SCD and educate professionals and families. We approached this problem from the vantage point of modelling. The challenge for us lied in developing a model for analysis which should be comprehensive enough to account for meaning making in different areas. Developing a carefully detailed methodology can address this challenge, since it may serve as the basis for applications. In particular, we devoted ourselves to the design of a specific methodology for SFL-based modelling.

254

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Drawing on the seminal works of Halliday (2008) and Mahboob and Knight (2010), we adopt the concept of ‘appliable linguistics’ as (i) a way of doing linguistics so it can always be applied and (ii) as a method of using linguistic knowledge to help solve problems in different areas. Halliday (2008) posits that Systemic Functional Linguistic (SFL) theory is a resource for solving problems, thus ‘appliable’. He states that as an appliable linguistic theory, SFL is ‘a body of method that can be drawn upon in those areas of human activity in which language is critically involved’ (2008: 203). In this sense ‘appliable’ involves more than a direct application of linguistics to other areas. It is, as it were, an ethical programme ranging from theory development to end use in the community. A key challenge has to do with developing an underlying principle by which theory can be ‘applied’ in such a way that it is transparent and reproducible – in other words, how the method of application can become part of the theory in appliable linguistics. More specifically, our interest lies in developing enabling methodologies addressing problems in different fields which depend on semiotic systems knowledge to solve problems. Our chapter will outline the steps involved in the design of this application.

APPLYING LINGUISTICS TO SCD MANAGEMENT IN THE CONTEXT OF MINAS GERAIS Our application is the result of a joint effort between two research teams at the Federal University of Minas Gerais (UFMG). The partnership began with Lima’s (2013) PhD dissertation. The healthcare team is based at NUPAD, which is part of the UFMG Medical School. NUPAD is responsible for developing Minas Gerais state’s first newborn screening programme. Today it is internationally recognized as a reference service for newborn screening (Fernandes et al. 2013). A number of research, teaching and extension programmes are developed by NUPAD. Most of them focus on children and women’s health. NUPAD also manages projects on humanization of care in health services for the Brazilian Universal Health Care Network. The appliable linguistics team works at LETRA, which is part of the UFMG Language School. LETRA develops experimental research projects on language, in particular on the field of expertise in translation. Most of the relevant research output of LETRA is related to mapping the expert profile of translators – which involves analysis of translators’ rhythm (Alves 2005), performance and text production. The main goal of producing such output is to inform application. In particular, the research informs appliable projects on translation training, text modelling and expertise in translation and text production. Our application is a direct result of the collaboration between the two teams. NUPAD had a need for efficient text production due to the challenge of building

NEWBORN GENETICS DIAGNOSIS TEXTS

255

knowledge and educating users – as reflected in the international partnerships between NUPAD and other research groups from around the world. Text production in Brazilian Portuguese (henceforth BP) was needed along with translations into different languages enabling multilingual text production. LETRA’s role in this work was to provide linguistic solutions by modelling texts and creating a template so future texts could be produced in a pipeline – the ‘pipeline’ (Guseva et al. 2020) is technically defined as the steps needed to process a text and to use it to produce another document (annotation, text, protocol). The notion of efficiency was also included in the modelling design, and it would be obtained by including texts recognized by expert people in the SCD field as good texts to be consumed by the layperson. The modelling included BP and English. Lima’s (2013) PhD project, as well as other applications that followed, aimed at developing models of text production and translation in multilingual environments. Following discussions between NUPAD and LETRA, a brief for the application was produced. After detailing NUPAD’s demands in relation to the linguistic problem posed by SCD management – that is, to identify which text types they needed modelling the most – our next step was to write an application brief. The brief included the guidelines for the application, as well as the steps that needed to

FIGURE 14.1  The flowchart of our application.

256

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

be taken and the goals for each step (Figure 14.1). The brief consisted of four steps, showed on the flowchart: 1) Determine the types of text to be modelled in Brazilian Portuguese and English. 2) Compile a representative corpus for each text type in BP and English. 3) Carry out a systemic profile of the text types. 4) Create a template of text production for each text type for BP and English. The development of the brief is detailed below.

DESIGNING THE APPLICATION PART 1: THE CORPUS Text types to be modelled The decision concerning the text types to be modelled was based on contextual variables of situations involving SCD. This made us list important field and tenor characteristics that our texts had to realize. Since mode is textual and in a sense enables the other metafunctions (Matthiessen 1992) it was not included in this step. In broad terms, the context of situation of SCD management shapes texts with less behaviour changes in field when compared to tenor. In our application, field is more stable because most texts are about SCD. Accordingly, sequences of activity are related to building knowledge about some aspect of diagnosis, symptoms, treatment, genetics and biochemistry, and communicating the knowledge to concerned groups. The generalized aspects of field we needed to consider for any text type in our sample have as the domain of experience Medicine: SCD: Screening. The social purpose was always to report on scientific research. The social activity was to expound on and communicate knowledge. The most important contextual variable for us is tenor – because configurations are more varied when compared to field. This is due to the fact that experts (i.e. the people mostly responsible for building and communicating knowledge on SCD) need to establish interpersonal relations with (a) other experts, (b) technicians in healthcare and (c) laypeople (including families and the general public, assuming they are not experts in SCD). The generalized aspects of tenor we needed to consider for any text types in our sample had to be different for each kind of relationship. For the expert– expert relationship, there is a horizontal relation in terms of status, established both by institutional hierarchy and by level of expertise. For the expert– technician relationship, there is a vertical relation due to different institutional

NEWBORN GENETICS DIAGNOSIS TEXTS

257

positions and level of expertise. For the expert–laypeople relationship, there is a vertical relation in terms of level of expertise, but not in terms of institutional hierarchy. Considering status in the community in general, there is a vertical relation between health professionals and the general public. A discussion was conducted among researchers at NUPAD and LETRA to determine which types of text would be most representative of these generalized contextual variables. Because field remained more stable, tenor was crucial at this point. In addition, the texts should be those most relevant in everyday life of the SCD management community (cf. Lima 2013). Ultimately, three types of text were chosen: (1) research articles (specialist–specialist interaction), (2) technical manuals (specialist–technician interaction) and (3) pamphlets and patient information leaflets (specialist–layperson interaction). After deciding on the text types, the next step in the brief was to build the corpus. Corpus compilation The corpus of our application was first compiled and analysed by Lima (2013). It was composed of texts in the field of medicine, more specifically on newborn screening for SCD. Texts were collected from NUPAD’s virtual library. Using the software PHL (Personal Home Library), NUPAD’s library is constantly growing as new publications are added. All publications in NUPAD’s library are indexed according to the DeCS (the Latin American health sciences descriptors). The method of compilation began by inserting a list of relevant descriptors in the library’s PHL. The descriptors were ‘newborn’, ‘screening’, ‘sickle cell disease’ and ‘sickle cell anaemia’. Every text indexed with these descriptors was collected. This collection formed our corpus of reference. At this point, we confirmed our expectation that there are many more texts than we were capable of analysing. So, we needed to generate a representative sample of the texts retrieved from the corpus of reference. This sample became our corpus for modelling purposes. The corpus for modelling was compiled following the steps developed in Biber (1990) and Neumann (2005). First, we divided the text into samples of roughly 1,000 tokens. This is the average number of tokens considered to be representative of most language categories present in a text. Moreover, it is also sufficiently representative that a text type can be separated out from other types of text (cf. Biber 1990; Nunes 2010). We used a randomization programme to select samples of texts manually, labelling them as articles, manuals or pamphlets. Finally, a header was added to each sample. The header included relevant information about the sample. For instance:

258

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

The header covers the following information: 1) Author: KIKUCHI, Berenice. 2) Place of publication, publisher and year: São Paulo. OPAS/OMS. 2012. 3) Name of sample: Anemia Falciforme, manual para trabalhadores de saúde e educação nas Américas [Sickle Cell Anaemia, A Manual for Health and Education Professionals in the Americas]. 4) Number of pages: 53–61. 5) Type of text: Manual técnico [technical manual] 6) Type of sample: Parcial [excerpt] 7) Language: português brasileiro [Brazilian Portuguese] The corpus for modelling contained eighteen samples divided among the three types of text.

DESIGNING THE APPLICATION PART 2: THE PROFILE The modelling we present in this chapter was guided by three criteria. First, the modelling of our corpus can be extrapolated to the language. Second, there is a degree of variation among texts of the same type and the model takes into account the most probable variables. Third, the model is considered effective only after it is suitable for application and meets the demands of our client. A systemic profile is divided into two parts: a static text profiling and a dynamic modelling of unfolding text.

TABLE 14.1  Corpus for Modelling Sample #

Type of text

Field

Tenor

Tokens

1 through 6

Article

SCD

Expert–expert

6,000

6 through 12

Manual

SCD

Expert–technician

6,000

13 through 18

Pamphlet

SCD

Expert–laypeople

6,000

NEWBORN GENETICS DIAGNOSIS TEXTS

259

Static text profiling Static text profiling matches features of the language to actual occurrences in the corpus. This provides a description of language systems used in a text in terms of their frequency – leading to a profile of most probable use. In SFLbased quantitative studies there is not a precise definition of ‘most probable’ based on frequencies of occurrence. Technically, any probability over 50 per cent is the most probable. However, we arbitrarily decided to define as ‘most probable’ any expectancy of occurrence that is equal to or greater than 67 per cent. The calculation we made was a simple probability for all occurrences annotated in our corpus.



(1)

For the text profiling of our corpus, we analysed different discourse semantic and lexicogrammtical systems. In the grammar, we analysed clause and group grammar – including a focus on nominal groups and downranking, since these are foregrounded resources in science discourse. We also carried out lexical analyses of token distribution and its correlation with systemic features. For the present chapter we will focus only on clause grammar systems. For our model we analysed transitivity & circumstantiation (ideational metafunction); mood & modality (interpersonal metafunction) and theme & predication (textual metafunction). For BP the description in Figueredo (2011, 2014) was deployed and Halliday and Matthiessen (2014) was used for English. The texts were divided into units of analysis – for the analysis reported

TABLE 14.2  Example of Annotation of a Clause Clause

System

Feature annotation

Universal newborn screening has significantly decreased mortality in young children with SCD.

transitivity

Material: transformative: extending & transitive: operative

circumstance

Non-selected

mood

Clause: major: free: positive & minus assessment & non-interactant & indicative: declarative

assessment

Comment: propositional: qualificative

theme

Subject theme: declarative: identified & minus interpersonal & minus textual

predication

Non-selected

260

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

here, the ‘clause’. Systemic features were identified and annotated as outlined in Table 14.2. Dynamic modelling of unfolding text We then built dynamic graphs of our data. A dynamic graph is a representation of features selected for each clause. Because several features are selected from several systems simultaneously for each clause, they are brought together and the co-selection is represented as a single point in the dynamic graph. In addition, the flow of discourse (dynamics) is represented as movement in space as the distance between two points in the graph (Figueredo and Figueredo 2020). This representation in the dynamic graph enabled us to decide which features of our corpus were more similar and which were more probable choices in transitivity, mood and so on for a given clause. In order to design dynamic graphs of our corpus we had to take the following steps. First, we needed to analyse the texts in our corpus according to the SFL categories that are relevant to our application. When we analysed a corpus, we ended up with a table with a list of labels and annotations for each text (see Table 14.2). This kind of analysis output posed a practical problem in that it was difficult to visualize relevant data for each text and compare them to other texts. However, most importantly, it did not show the dynamics of text production – that is, how a choice of a given feature in a text unfolded in subsequent choices. Our solution to these two problems is to represent language as a space. Text units – more specifically clauses – are then modelled as points in this space. This means that any given choice made in a given clause was marked as a point; then, a subsequent choice in the next clause was marked as a different point and so on. By doing this, we were able to ‘convert’ our annotation tables into graphs. Graphs are easier to visualize and they capture ‘movement’ as a text unfolds since they show how choices are made clause by clause (Figure 14.2). As we can see in Figure 14.2, clauses 1, 2, 5, 6, 9, 10, 11 and 12 are closer to each other, forming a cluster in the low left-hand corner of the graph of language space. By the same token clauses 3, 4, 7, 13, 15 and 17 also form a cluster (in the mid left-hand corner). The other clauses (14, 16, 18, 19 and 20) do not form clusters and are scattered throughout the language space. Clustering indicates that their internal organization and the selection of features among the clustered clauses are similar. Consequently, they perform comparable semiotic work in the text (e.g. explaining what SCD is, or describing the history of SCD for parents, or training a technician on a hospital routine – in other words, patterns typical of explanations, histories and procedure genres). Clusterings like these ultimately reveal recurrent configurations of meaning, which in turn realize genres. This is a very important step for our modelling,

NEWBORN GENETICS DIAGNOSIS TEXTS

261

FIGURE 14.2  Representation of a text converted into points in language space.

but also for description of language in context and for the theory as a whole, since we approach language patterning through instantiation – differently from most research bound by stratification and metafunction alone, and never getting detailed descriptions of meaning making as it unfolds as ‘movement through time’. Language space graphs like those in Figure 14.2 are constructed with the help of statistical packages. For this specific application we used R-statistics. After annotating data on tables and having the calculations for distance, we used the following list of commands in R to generate the graphs (for a detailed explanation, see Lima 2013). require(scatterplot3d) dado=read.table("Tabela Artigo inglês.txt",T) attach(dado) x=Modo y=Transitividade z=Tema scatterplot3d(x, y, z, type="p",highlight.3d=F, col.axis="blue",col.grid="lightblue", main="", pch=20, xlab="Modo", ylab="Transitividade",zlab="Tema")

262

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 14.3  Representation of a text with clauses clustered in a dendrogram (tree graph).

The clustering tells us which regions of the language space are more associated with the specific semiotic labour a text needs to accomplish. They do not, however, tell us exactly which configuration is the most frequent – thus most probable. In order to do this we need to build dendrograms for our data. A dendrogram is a tree graph that represents a hierarchy of categories and shows which categories are more similar and how they cluster together. Accordingly, it also shows how distant clusters are from one another.2 Similarly, the calculation of distance between clauses was done with the help of R-statistics. The command for dendrograms we used is: ArtigoPortuguês=read.table("Tabela Artigo Português.txt",T) ArtigoPortuguês=ArtigoPortuguês[,–1] ArtigoPortuguês=ArtigoPortuguês[,–4] Artigo_Português=dist(ArtigoPortug uês) cluster.s=hclust(Artigo_Português,method="complete") plclust(cluster.s,ylab="Distância") At this point, we need to explain how we derive the values for clauses to place them in the space graph and their distance in the dendrogram. Locating a clause in space – that is, how we determine the coordinates of a point – means assigning a numerical value for the valeur of features in a system. We will briefly explain the calculations here (for the detailed version, see Figueredo 2014; Oliveira et al. 2017; Figueredo and Figueredo 2020).

NEWBORN GENETICS DIAGNOSIS TEXTS

263

FIGURE 14.4  The system of speech person in English.

The first element of dynamic modelling is to understand language as a system. This means that axial relations (choices in paradigm realized by structures in syntagm) are foregrounded. We can describe a system as set of choices related by the concept of valeur (see Saussure 1966). The value of a choice is relative to other choices when they are contrasted because choices are simultaneously (a) somewhat different, (b) but also partially the same. The basis of all paradigms is formed around a system of choices (Martin 2013). Thus, all phenomena in language can be represented paradigmatically as a system. For example, the system of speech person3 in the verbal group of English is described, as shown in Figure 14.4. The system of speech person has [personal pronoun] as entry condition. Here [speaker] and [addressee] are features – i.e. the items available to be chosen. The feature [speaker] is realized in structure by the word rank items I and we; [addressee] is realized by the item you. The valeur of each feature is given by two operations in the system. (1) The first operation is the contrast between features. Consequently, [speaker] means ‘choosing not-addressee’ at the same time that [addressee] means ‘choosing not-speaker’. (2) The second operation is the increase in detail and specificity from [personal pronoun] to [speaker/addressee]. This contrastive property of valeur is called agnation. Agnation (Gleason 1965) is the operating contrast. Agnation is the factor making features ‘somewhat different’ – although both are speech persons, [speaker] is not the same as [addressee].

264

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 14.5  Representation of clause points in language space and as a dendrogram (tree graph).

The detailing property of valeur is called delicacy. Delicacy (Halliday, McIntosh and Strevens 1964) is the factor of valeur operating specificity. It is the factor making features ‘partially the same’ – although [speaker] is not the same as [addressee], both are personal pronouns and both operate as choices in speech person. We determine the factors of agnation and calculate the amount of agnation for a system. Concomitantly, we determine the factors of delicacy and calculate the amount of delicacy for a system. Finally, we multiply agnation by delicacy and calculate the amount of valeur for a system. When we annotate a text, we can then convert our annotation into the amount of valeur. The amount of valeur provides the coordinates to place a point in language space. After locating all clause points in the dynamic graph of language space, we combine them with the dendrogram analysis and obtain the following results for the data in Figure 14.2 above. Finally, we interpret metafunctions as frequent co-occurring systems to build a prototypical metafunctional profile of each type of text and develop a model to serve as a basis for the production of these text types. As we deployed this methodology to our application, it allowed us to characterize register variation as conditioned probabilities in language dynamics. On this basis we developed templates of the three types of text in our corpus and used them to guide professionals in NUPAD when producing texts related to SCD management.

MODELLING TEXT In this section we present the results of static profiling and dynamic unfolding for our corpus. Due to the limited scope of this chapter we will focus on the most important results and expand further when needed.

theme

mood

53.32

rel. attrib.

1.59

Verbal

84.13 15.87 – –

theme prominent

theme unmarked

theme marked

100

theme default

non–interpersonal

–

61.9

non–textual

interpersonal th.

38.10

–

Imperative

textual theme

–

interrogative

100

–

behav.

Declarative

–

Existential

7.93

6.68

rel. ident.

Mental

30.16

Material

transitivity

BP %

Clause type

System

Article

4.08

95.92

–

–

100

–

86

14

–

–

100

–

–

2.04

–

41.04

5.90

51.02

ENG %

–

–

14.10

85.90

–

–

88.30

11.70

–

–

100

–

–

1.28

–

70.53

6.40

21.79

BP %

Manual

TABLE 14.3  Static Profile of the Corpus for transitivity, mood, theme in BP and English

–

100

–

–

–

–

88.44

11.56

–

–

100

–

–

9.62

–

76.91

3.78

9.62

ENG %

–

–

10.53

89.47

98.25

1.75

78.94

21.06

–

–

100

1.72

–

–

–

79.32

6.90

12.07

BP %

Pamphlet

4.00

96.00

–

–

88.89

11.11

87.83

12.17

2

–

98

–

–

–

–

74.00

4.00

22.00

ENG %

266

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

After analysing all systems for our corpus and finding the most probable configurations, we were able to bring the results together as models for articles, manuals and pamphlets. Each model is composed by a set of three representations: (1) the most probable systemic configurations for the text type (we define ‘most’ as a frequency of 67 per cent or more); (2) the clustering and dynamic development of choices for features through the text; and (3) the most probable distribution of features clause by clause. The first representation is related to static profile; the second and third representations are related to dynamic unfolding. The results for static profiling are shown in Table 14.3. The profiling was done both for BP and English. The numbers show the most frequent percentage of occurrence for each feature. This is important not only to show how the texts types are configured, but also to separate texts out into different types. In addition, these configurations may reveal important aspects of discursive and extrinsic functionality – as they show the preferred choices of each type of text for any given feature. Such information turned out to be very important when delivering our product to our client as we were able to explain the reasons why articles, manuals and pamphlets were configured the way they were. It is important to make clear that there was language comparison among them. The results in Table 14.3 were then complemented from our dynamic perspective. This involved the clustering of clauses in the language space and the clause-by-clause distribution of features – since a frequency percentage does not tell us how dispersed the features are throughout a text. The clustering was effected by placing clauses in the language space and identifying points where similar clauses clustered. Both the clustering and the distribution are features of the corpus. They do not belong in any one text in particular; rather, they are the outcome of the higher probability for the clusters and how they were distributed as texts unfold. As an illustration, the patterns for different text types in BP is shown in Figure 14.6. The distribution in dendrograms for manuals in BP and English is illustrated in Figure 14.7.

FIGURE 14.6  Clustering in language space in BP.

NEWBORN GENETICS DIAGNOSIS TEXTS

267

FIGURE 14.7  Dendrograms for manuals in BP and English.

USING THE MODEL TO CREATE TEXTS After modelling articles, manuals and pamphlets, we were able to control the production of these text types along the following lines: i. The most probable configuration for each text type. This works as a measure to determine whether a text is an article, a manual or a pamphlet. For example, in the ideational configuration material clauses alone is an important marker because articles have on average 30.16 per cent of them, whereas manuals have 21.79 per cent and pamphlets have 12.07 per cent. As a result, if a text has roughly 12 per cent material clauses, there is a higher probability it is a pamphlet. ii The grammatical configuration for each individual clause. Every clause in the corpus was modelled. Thus, we know that the average configuration of, for example, the second clause of an article in English is: declarative/relational: attributive/textual and unmarked. Example 1: Second clause of an article (2) However, a comprehensive review of the literature for the prevention and management of paediatric SCD-related complications is lacking.

268

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

In addition, the clause configurations show the sequence in which different configurations appear – i.e. how the texts develop from one type of configuration to the next. Because probabilities are conditioned by previous choices, we can also see whether or not the configuration of the second clause follows the average or not, depending on choices made for the first clause. iii The way clauses work together to create more complex meanings. Clustering shown in graphs and branches of dendrograms explains how sequences of clauses form discourse patterns. In English, articles form 10 clusters, manuals form 6 and pamphlets form 7. For example, in the model for manuals in English, clauses 3, 7, 42 and 44 form a cluster of declarative/existential/unmarked. Example 2: a cluster in manuals (3) There is increasing and excruciating pain which require medical intervention and hospitalization. (7) There are a number of short- and long-term complications associated with sickle cell disease. (42) There is no specific treatment for individuals with sickle cell disease. (44) There are national clinical guidelines for acute and community management of children and adults with sickle cell disease. iv The length of a text. Knowing the average number of clauses per text reveals how long a text type is on average. In our models, for example, a manual has 78 clauses whereas a pamphlet has 57 clauses. v Recommendations for elements an efficient text needs to have. Linguistically speaking, the efficiency of a text is very hard to measure. But using our model we can gain some insight into the efficiency of a text by comparing it to the model. As mentioned above, we modelled texts our client considered efficient by their own standards. For example, on average, the third clause of an article should be configured as declarative/relational: identification/unmarked. Suppose, however, that our client produces a declarative/material/marked configuration. In this case the model can be used to ‘tweak’ the clause to make it fit more ‘predictably’ into the text. Example 3: adjusting a clause to make it ‘fit’. • declarative/material/marked: For the care provided to children with SCD this article is doing a comprehensive review of the literature. • declarative/relational: identification/unmarked: This article presents a comprehensive review of the literature for the care provided to children with SCD.

NEWBORN GENETICS DIAGNOSIS TEXTS

269

vi The improvement of non-efficient texts by locating exactly where they miss the most probable configurations. As a direct consequence of (v), we can find all clauses that do not fit the average, most probable configuration and ‘tweak’ them so they can fit the model. This point is illustrated below by an excerpt from a pamphlet. As mentioned above, the social purpose of a pamphlet is to educate the general public and help families and friends to understand SCD symptoms and act accordingly. Example 4 below was written without the model. Its configuration is closer to that of a manual than a pamphlet. Example 4: excerpt from a pamphlet without the model. When no haemoglobin A is produced the condition is known as sickle beta zero thalassaemia, sometimes written as SBETA Thal. This behaves more like sickle cell anaemia. For more information about beta thalassaemia contact the Thalassaemia Society or one of the specialist Centres. SBETA Thal occurs if your child has inherited a haemoglobin gene from one parent and a persisting haemoglobin gene from the other parent. It is sometimes written, HbS/HPFH and very rarely causes any health problems. Sickle haemoglobin SD disease is a more unusual form of sickle cell disease and occurs if your child has inherited haemoglobin_S from one parent and haemoglobin D from the other parent, sometimes written, HbSD. This form can be as serious as sickle cell anaemia but could be milder. Example 5 shows the text after being corrected to be more efficient according to our model (see Braga 2013 for the modelling of these corrections). It is important to mention that the corrections to make the text more efficient followed the clause grammar model we presented above, but also group grammar, discourse semantics and lexis models that we, unfortunately, do not have time to show in this chapter (for extra material on modelling and this application see Braga 2013; Lima 2013; Kogut 2017; Oliveira 2018). Example 5: excerpt from a pamphlet ‘tweaked’ to be efficient by following the model. Our blood is formed by millions of red cells called erythrocytes, also called red blood cells. Red blood cells are rounded, flattened and elastic, and contain a pigment called hemoglobin, that provides the red color to the blood. The hemoglobin is responsible to take the oxygen from lungs to all the body through the bloodstream so that all the organs work properly. Hemoglobinopathies are genetic diseases that affect the hemoglobin. The sickle cell disease is the most common example of hemoglobinopathies. As it is a genetic alteration, the person with sickle cell disease produces an

270

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

altered type of hemoglobin, the hemoglobin S (Hb S) and not the normal hemoglobin, the hemoglobin A (Hb A), found in adults. The transport of oxygen for all the body, a function of the hemoglobin, is not done in a satisfactory way when the person has the sickle cell disease. Under certain occasions, Hb S makes the red blood cells stop being flexible and becomes more rigid. In these cases, the red blood cells become half-moon or sickled shaped (there comes the name sickle cell disease). As can be seen, the ‘tweaked’ text has much less grammatical intricacy, less clause complexing, organization of the nominal group complexity and almost no explicit connexion. According to the model, this is what a pamphlet should look like because it has to be easy to read and the information should be quickly accessible by most people.

CONCLUSION We began our chapter by introducing a problem of healthcare related to newborn screening and sickle cell disease. In particular we explained briefly the relationship between health professionals and families of SCD patients and how language plays a fundamental role in building knowledge about diagnosis and treatment as well as communicating and popularizing the knowledge. If language is not used efficiently in this process, then research, professional training and family support may face difficulties. These problems were approached as a challenge for linguistics, which led to a collaboration between a medical research centre (NUPAD) and a language modelling laboratory (LETRA). The collaboration established a brief for determining relevant text types and modelling them. The models, in turn, would be used as a template to build new texts and correct efficiency problems. The notion of efficiency was determined by successful texts (e.g. published articles) and judgements by healthcare professionals (e.g. widely used manuals). We used SFL as the base theory for our model. SFL was used to analyse texts grammatical (clause and group) and discourse semantics configurations. Here we were able to discuss only clause grammar results due to the scope of our chapter. The SFL analysis was complemented by our modelling methodology, which includes (i) calculations of frequencies for all configurations; (ii) establishing the most probable configurations; (iii) using these probabilities to create average, most expected texts; then (iv) finding out the distribution of these configurations as a sequence in time (represented as points in the language space) and clause-by-clause distribution (represented by dendrograms).

NEWBORN GENETICS DIAGNOSIS TEXTS

271

Our design for the application had to, first, find commonality in probability to group texts grouped types. We found that three types of texts are mainly involved: (1) research articles (specialist–specialist interaction), (2) technical manuals (specialist–technician interaction) and (3) pamphlets and patient information leaflets (specialist–layperson interaction). We then had to interpret frequent co-occurring systems to build a prototypical metafunctional profile of each type of text and develop a model to serve as a basis for the production of these text types. Finally, we deployed the notion of register as conditioned probabilities in language (Halliday 1992) to develop templates of these three types of text and used them to guide professionals in NUPAD when producing texts. The linguistic product presented to our client was able to be used effectively in modelling texts, measuring efficiency and correcting problems in text production. Moreover, text production was proven to increase efficiency in building and transferring knowledge between specialists, technicians and laypeople, which resulted in improved quality of sickle cell disease management, such as an increase in pamphlet production and distribution, resulting in a public health campaign at the state level. In Minas Gerais, the modelling was adopted by NUPAD personnel to write their texts; and state health department official pamphlets were produced using these models. This is estimated to have a positive impact on the quality of sickle cell disease management. Testimonies from our clients stated that new texts based on the modelling allowed the layperson to have information about their condition more easily due to an increase in the new pamphlets’ production and distribution. This application enabled us to consolidate language modelling as application and a reliable methodology for SFL-based dynamics modelling. Such developments motivated our deeper understanding of the quantification of meaning-making process in systems, the placement and ordering of valeur reactants in the Gleason Scale, the displacement in language space and the description of ‘text-as-instance’ as the unit of displacement. In addition, this enabled us to recognize reconfigurations of meaning between stages of genres through empirical data, as we departed from the instantiation dimension to understand language patterns.

NOTES 1 This study was financed by the National Council for Scientific and Technological Development – CNPq. 2 A more familiar example of a dendrogram is the ‘tree of life’ in evolutionary biology. 3 All systems names are conventionally written in small caps; system of speech person, system of mood and so on.

272

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

REFERENCES Alves, F. (2005), ‘Ritmo cognitivo, meta-reflexão e experiência: parâmetros de análise processual no desempenho de tradutores novatos e experientes’, in A. Pagano, C. M. Magalhães and F. Alves (eds), Competência em tradução: cognição e discurso, 90–122, Belo Horizonte: Editora UFMG. Aygun, B. and I. Odame (2012), ‘A Global Perspective on Sickle Cell Disease’, Pediatric Blood Cancer, 59 (2): 386–90. Biber, D. (1990), ‘Methodological Issues Regarding Corpus-Based Analyses of Linguistic Variation’, Literary and Linguistic Computing, 5 (4): 257–69. Braga, A. (2013), A popularização da popularização da ciência: gramaticalização da interação, avaliação, representação da experiência, construção e organização discursive, Mariana: Departamento de Letras, Universidade Federal de Ouro Preto. Fernandes, A. P. P. C., A. P. S. Batista, A. L. F. Dias, H. C. Torres, J. B. Silva, J. N. Januario, K. C. S. L. Lima, L. G. Fonseca, L. N. F. Meira, M. Murao, P. Q. S. Menezes, R. S. F. Silva and M. B. Viana (2013), ‘Process of Disseminating Knowledge on Sickle Cell Disease to Primary Health Care Professionals in the State of Minas Gerais, Brazil’, Paper presented at Sickle Cell Disease in the Post-Genomic Era, 7th Annual Sickle Cell Disease Research and Educational Symposium and 36th National Sickle Cell Disease Scientific Meeting, Miami, FL, 14–17 April. Figueredo, G. (2011), ‘Introdução ao perfil metafuncional do português brasileiro: contribuições para os estudos multilíngues’, PhD thesis, Faculdade de Letras, Universidade Federal de Minas Gerais, Belo Horizonte. Figueredo, G. (2014), ‘Uma metodologia de perfilação gramatical sistêmica baseada em corpus’, Letras & Letras, 30 (2): 17–45. Figueredo, G. and G. P. Figueredo (2020), ‘A Systemic Dynamics Model of Text Production’, Journal of Quantitative Linguistics, 27 (4): 291–320. https://doi.org/ 10.1080/09296174.2019.1567301 Gleason, H. A. (1965), Linguistics and English Grammar, New York: Holt, Rinehart & Winston. Gomes, L. (2015), Avaliação da efetividade de uma intervenção educativa no conhecimento de profissionais de saúde que acompanham pessoas com doença falciforme, Belo Horizonte: Faculdade de Medicina da UFMG. Gomes, L., M. Vieira, T. Reis, T. Andrade-Barbosa and A. Caldeira (2013), ‘Understanding of Technical Education Level Professionals Regarding Sickle Cell Disease: A Descriptive Study’, Online Brazilian Journal of Nursing, 12 (2): 482–90. Guseva, A. I., I. A. Kuznetsov, P. V. Bochkaryov, S. A. Filippov and V. S. Kireev (2020), ‘Development of Text Data Processing Pipeline for Scientific Systems’, in A. Samsonovich (ed.), Biologically Inspired Cognitive Architectures 2019: Proceedings of the Tenth Annual Meeting of the BICA Society, 124–36, Cham: Springer. Halliday, M. A. K. (1992), ‘How Do You Mean?’, in M. Davies and L. Ravelli (eds), Advances in Systemic Linguistics: Recent Theory and Practice, 20–35, London: Pinter. Halliday, M. A. K. (2008), Complementarities in Language, Beijing: Commercial Press. Halliday, M. A. K. and C. Matthiessen (2014), An Introduction to Functional Grammar, 4th edn, London: Edward Arnold. Halliday, M. A. K., A. McIntosh and P. Strevens (1964), The Linguist Sciences and Language Teaching, London: Longmans. Kikuchi, B. (2007), ‘Assistência de enfermagem na doença falciforme nos serviços de atenção básica’, Rev Bras Hematol Hemoter, 29 (3): 331–8.

NEWBORN GENETICS DIAGNOSIS TEXTS

273

Kogut, L. (2017), ‘A Construção do Texto Argumentativo no RPG de Mesa’, Master’s thesis, Universidade Federal de Ouro Preto. Lima, K. (2013), ‘Caracterização de registros orientada para a produção textual no ambiente multilíngue: um estudo baseado em corpora comparáveis’, PhD thesis, Faculdade de Letras, Universidade Federal de Minas Gerais, Belo Horizonte. Mahboob, A. and N. Knight (2010), ‘Appliable Linguistics: An Introduction’, in A. Mahboob and N. Knight (eds), Appliable Linguistics, 1–12, London: Continuum. Marques, L. et al. (2015), ‘Living with Sickle Cell Disease: Adolescents’ Perception’, Rev Ter Ocup Univ São Paulo, 26 (1): 109–17. http://dx.doi.org/10.11606/ issn.2238-6149.v26i1p109-117. Martin, J. (2013), Systemic Functional Grammar: A Next Step into the Theory–Axial Relations, Beijing: Higher Education Press. Matthiessen, C. (1992), ‘Interpreting the Textual Metafunction’, in M. Davies and L. Ravelli (eds), Advances in Systemic Linguistics, 37–82, London: Pinter. Neumann, S. (2005), Corpus Design, Deliverable no. 1, DFG project STE 840/5-1, Saarbrücken: Universität des Saarlandes. Nunes, L. (2010), ‘As conjunções but e mas em textos ficcionais originais e traduzidos: uma abordagem tridimensional com base na Linguística Sistêmico-Funcional’, Master’s thesis, Faculdade de Letras, Universidade Federal de Minas Gerais, Belo Horizonte. Oliveira, F. (2018), ‘Modelagem do ambiente multilíngue de produção de introduções de artigos acadêmicos da área das ciências da saúde em inglês e português brasileiro’, Master’s thesis, Faculdade de Letras, Universidade Federal de Minas Gerais, Belo Horizonte. Oliveira, F., T. Campolina and G. Figueredo (2017), ‘Estudo comparativo do perfil sistêmico de manuais de instrução em inglês/português’, Domínios de Linguagem, 11 (2): 418–47. Saussure, F. de, C. Bally, A. Sechehaye, A. Riedlinger and W. Baskin (1966), Course in General Linguistics, New York: McGraw-Hill Book Co. Weis, M., M. Barbosa, R. Bellato, L. Araújo and A. Silva (2013), ‘A experiência de uma família que vivencia a condição crônica por anemia falciforme em dois adolescentes’, Saúde Debate, 37 (99): 597–609.

274

CHAPTER FIFTEEN

Re-reading Reading Images THEO VAN LEEUWEN

INTRODUCTION Along with the work of Michael O’Toole (1994), Reading Images (Kress and Van Leeuwen 1990) was, in its time, one of the first applications of Systemic Functional Linguistics to modes of communication other than language. Nevertheless, even in its first edition, published by Deakin University Press in 1990, Kress and I stressed that ‘although this book would not have been possible without the achievements of linguistics, we do not simply import the theories and methodologies of linguistics into the domain of the visual’ (Kress and Van Leeuwen 1990: 4) and in the preface of that book we stressed that, in writing it, we did not just aim to learn how to analyse the whole of texts in which the visual plays a role, but also ‘to understand language better’ (1990: 1). At the time we admitted that we were not yet ready to do so. Our ‘visual grammar’ was still in its early stages. But almost thirty years later, it may be possible to re-read Reading Images in the light of two more general questions – whether, or to which degree, linguistic methods can be applied to visual analysis (and by extension to the analysis of other semiotic modes) and what new perspectives this may open for the study of language. To do so, I will first review linguistically inspired approaches to multimodality which preceded Reading Images. What did they, and what did they not, take from linguistics, and why? Did they envisage that linguistics might, in turn, learn from multimodality studies and, if so, how and why? I will then use this as a background for asking the same questions with regard to multimodal extensions of Systemic Functional Linguistics.

276

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Next, I will focus on two issues in particular. The first is a re-reading of Arnheim, whose work was as influential on Reading Images as the work of Halliday, something which we may not always have sufficiently acknowledged. I will argue that, in hindsight, our reading of Arnheim was selective, focusing on aspects that integrated well with Systemic Functional Grammar. Bringing out what we did not include will foreground ways in which the visual is quite unlike language. Other non-linguistically inspired literature on visual communication can also help identify areas in which we overlooked the specificities of the visual, for instance in relation to the concept of modality. The second is a rereading in the light of post-Reading Images developments in the social semiotic study of visual communication, in particular the concept of parametric systems which emerged in the study of colour, typography and visual texture, and the concept of ambient visuality. Both are aspects of visuality which had been ignored in Reading Images, which fit less easily into its systemic-functionally inspired framework, and which have at least the potential to challenge some of the givens of that framework.

LINGUISTICS AND MULTIMODALITY STUDIES IN THE TWENTIETH CENTURY Prior to Reading Images, linguists studied non-linguistic modes of communication in three different contexts: the Prague School, founded in 1926 and continuing through the 1930s and 1940s; the structuralist-influenced Paris School, especially during the 1960s; and linguistically inspired North American work on non-verbal communication in the late 1950s and 1960s. In Prague, Mathesius, one of the founders of the Prague Linguistics Circle, had reacted against the Neo-Grammarian emphasis on absolute laws of sound change and advocated focusing on language use and language variation, including the variation between everyday language and verbal art. This led to an interest, not only in verbal art, but also in the theatre, music, film and, in the work of Bogatyrev, to folk music and folk costume. The theatre received special attention because of its multimodality, which combines dialogue, prosody, body language, dance, objects, sets and music. According to Honzl, who was a theatre director as well as a linguist, all these modes should be studied in their interaction, rather than separately, and they should be seen as tied together, not by language, but by action (Honzl [1941] 1976). What the Prague School approach took from linguistics was mainly its general semiotic principles, especially the concept of the sign, with its relation between the signans (signifier), signatum (signified) and denotatum (referent) and the idea of communicative functions. Mukařovský ([1936] 1976: 272) argued that the model of the sign could apply to other semiotics, in all of which ‘the artefact functions as an external signifier’, while the signified can be found

RE-READING READING IMAGES

277

in the ‘collective consciousness’ of the culture and society, rather than in the semantics of language. However, Mukařovský also maintained that ‘Language is the most fundamental set of signs (though by no means the simplest) in view of the fact that the entire realm of signs is the tendency to express everything in verbal terms’ ([1936] 1976: 273). With this not all Prague School scholars agreed. Veltruszký, for instance, writing about the theatre, said (quoted in Matejka 1976: 281–2): Words cannot be fully translated into gestures, pictures, music, the meaning of a picture cannot be fully conveyed by language, music, the play of facial muscles, etc. Each of these types of signs is entirely different, each has its own unique ability to refer to certain kinds and certain aspects of reality, and each is deficient in some respects. Language, for instance, is inferior to painting in indicating colours, the location of objects or people within a certain situation, the distance between them, etc. The second key idea the Prague School took from linguistics was functionality, which had been introduced by Bühler (1990 [1934])), who distinguished the representational function (what is spoken about), the expressive function (related to the speaker) and the appellative function (related to the person or persons spoken to). Mukařovský ([1936] 1976: 157) argued that these functions could be applied to all semiotic modes: ‘We are easily able to distinguish traces of the three basic functions in any informational discourse,’ and he added the aesthetic function, later re-baptized ‘poetic’ by Jakobson (1960). The Prague School also emphasized the social. Mukařovský ([1936] 1976) strongly criticized the Russian Formalist scholar Sklovskij (1925), who had argued that analysts of verbal art should focus only on the material used and the structure of the text, using textiles as an analogy – all that mattered for text analysis was the yarn and the techniques of weaving, not the international wool market. To which Mukařovský replied that techniques of weaving necessarily reflect the needs and pressures of the market and the laws of supply and demand. The study of verbal art should include ‘structures such as science, politics, economy, social stratification, language, ethics, religion’ (quoted in Matejka 1976: 271), and the interaction between them. Paris School structuralist semiotics emerged in the late 1950s, with Roland Barthes as a pivotal figure. Rather than on the arts, it focused on popular media and popular culture, which had grown in importance as a result of the spread of television. Its linguistic inspiration came primarily from Saussure and Hjelmslev, as well as from French linguists such as Martinet, Benveniste and Mounin. But where Saussure saw linguistics as part of semiotics, Barthes saw semiotics as part of linguistics. In contrast to the Prague School, Barthes saw ‘the world of signifieds as none other than that of language’ (Barthes 1967: 10)

278

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

and language as ‘not only as a model, but also a component, relay or signified’ (1967: 10). Like the Prague School, the Paris School drew mostly on general semiotic concepts from linguistics, ‘analytical concepts which we think a priori to be sufficiently general to start semiological research on its way’ (1967: 12). For Barthes, these were langue and parole, signifier and signified, syntagm and system, and denotation and connotation. He did, however, also say that applying these concepts to other modes of communication might require adapting them: ‘It is not presupposed that semiology will always be forced to follow the linguistics model closely’ (1967: 11). Like Barthes, Christian Metz, in his earlier work on the semiotics of the cinema, also focused on general principles of linguistic analysis: ‘commutation, analytical breakdown, strict distinction between the signifier and the signified’ (1974: 107). Although Systemic Functional Linguists would later analyse images in terms of ranks of different magnitude (O’Toole 1994, Boeriis and Holsanova 2012), for Metz ‘the cinema has no phonemes, nor does it, whatever one may say, have words’ (1974: 65). It only has shots and these, he thought, make meaning at a level analogous to the sentence. But unlike sentences, shots cannot be analysed into ‘parts of speech’. They are ‘blocks of reality’, ‘actualized in their total meaning’ (1974: 11). It is only on the level of discourse that patterns and regularities can be found, such as those Metz identified in his ‘grande syntagmatique’ (1974: 108–46), a typology of sequences primarily based on the conjunctive relations between the shots in these sequences. Structuralism has often been criticized for being ‘structuralist’ and for its neglect of the role of the viewer and social and ideological context (cf. Bateman and Schmidt 2012: 38). However, Barthes’ semiotic analyses in Mythologies (1973) and elsewhere were always also ideological critiques. For Barthes, semiotics only came into its own in relation to ‘systems where the sociological significance is more than superficial’ (1967: 10). As Norris (2016) rightly stresses in her four-volume anthology of multimodality papers, in the late 1950s North American scholars had begun to focus on the multimodality of spoken language – facial expression, gesture and posture. This had been foreshadowed by Sapir, who had described gesture as ‘an elaborate and secret code that is written nowhere, known by none, and understood by all’ (1949: 556). Scholars like Birdwhistell (1970) followed the model of the American descriptive linguistics of the period quite closely, positing ‘kinemes’ as non-verbal phonemes, and developed intricate modes of transcription. But Birdwhistell also stressed the social and cultural dimension of body language, arguing against universalist psychological accounts of body language as well as against the idea of non-verbal communication as ‘paralanguage’. This he referred to as the ‘modifier temptation’, which holds ‘that all other modes of communication are to be studied as subsystems subordinate to it [i.e. language]’ (1970: 175), and he showed that there are many forms of kinesic behaviour

RE-READING READING IMAGES

279

in which verbalization is absent, and that the modes in the multimodal mix of everyday interaction are ordered in complementary rather than hierarchical ways.

APPLYING SYSTEMIC FUNCTIONAL GRAMMAR TO IMAGE ANALYSIS Prior to Reading Images, linguistically inspired ways of analysing images focused for the most part on analysing the people, places and things represented in images, with the aid of concepts such as denotation and connotation, metaphor and other rhetorical figures (Durand 1970) or, in art theory, iconographical and iconological symbolism (cf. Van Leeuwen 2001). The internal composition of images was usually discussed in formalistic and aesthetic terms, in terms of the quite specific conventions of, for instance, biblical genres such as the Nativity, the Pietà, the Last Supper and so on, or in terms of ad hoc symbolisms (van Leeuwen 2001: 115–17). In other words, the visual was treated as what Martin (1992: 15) called a ‘lexese’, a bi-stratal system. And, in terms of function, although several schools of linguistics and pragmatics have, ever since Bühler, included the interpersonal function, or even entirely focused on it (e.g. speech act theory and conversation analysis), image analysis remained almost exclusively focused on the ideational, and to some extent on the ‘poetic’, function. Halliday’s grammar made it possible to attempt a theory of visual composition as ‘grammar’, as a set of meaning-making systems realized by specific visual configurations. If it was not possible, as Metz (1974) had said, to analyse images into formal constituents in the way clauses can be analysed into nominal groups, verbal groups and so on, it might be possible, Kress and I argued, to follow Halliday in analysing images into functional elements such as processes and participants, provided these can be linked to specific visual signifiers. For this we relied on Arnheim’s (1982) theory of visual composition. Unlike Metz, Arnheim did analyse images into their constituents: ‘volumes’ and ‘vectors’, both defined on formal grounds. ‘Volumes’ or ‘masses’ were said to have a distinct ‘weight’ or ‘gravitational pull’ and to form a tonal contrast with their environment (today they can be identified automatically, for instance by edge detection; cf. Ziou and Tabbone 1998). Vectors were the dynamic, usually diagonal, elements that link the volumes. Volumes therefore realize what Halliday called ‘participants’, the people or things in images, and vectors the processes, the actions that link them together in what we called ‘narrative images’, images that, abstractly or figuratively, represent actions and events, or, in Arnheim’s words (1982: 154), ‘We shall distinguish between volumes and vectors, between being and acting.’ On this basis, we were able to show that images, including photographic images, are constructions rather than ‘blocks

280

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

of reality’, as Metz had claimed. Consider the difference between Koudelka’s famous photo of the Prague uprising (Figure 15.1) in which the student on the left defiantly gestures and shouts at the Russian soldiers on their tank, and Stuart Franklin’s equally famous photo of the ‘tank man’ standing in front of a tank with his shopping bags on Tiananmen Square in 1989. In the former the student is the actor and his gesture the main vector; in the latter the tank is the actor and its barrel the main vector. As a result, these images, though both containing the same principal components, presented very different interpretations of what may be seen as two very similar events. Along these lines, we were able, in Reading Images, not only to describe a range of compositional systems, but also to link systemic functional concepts to forms and configurations described in the literature on visual perception and art and design theory. We regard multidisciplinarity of this kind as fundamental to multimodality studies. Systemic Functional Grammar alone is not enough. As Metz stressed decades earlier (1974: 121): ‘Semiotics as we now understand it must always rest on a double support: on the one hand upon linguistics, and, on the other hand upon the theory peculiar to the field under consideration.’ To give one more example, our concept of salience rested not only on linguistic theories of intonation and rhythm but also on the psychology of perception (e.g. Gregory 1970): the eye is especially sensitive to contrasts in intensity and colour (e.g. red/green and blue/yellow), and the way viewers

FIGURE 15.1  Invasion by Warsaw Pact troops in front of the Radio Headquarters, Prague, August 1968. Photo: Joseph Koudelka © Magnum Photos, reproduced with the permission of Ritzau.

RE-READING READING IMAGES

281

assess salience is both ‘controlled’ by a ‘bottom-up, fast primitive mechanism with variable selection criteria which directs the “spotlight of attention”’ and ‘cognitive, volitional control’ (Itti and Koch 2000: 1490). As Barthes had anticipated, linguistic models cannot always be followed closely. In the course of our work we came to question to which degree Halliday’s metafunctions can be realized in every semiotic mode, as we assumed they would in Reading Images, where we drew parallels between linguistic and visual systems, for instance between the linguistic mood system and the system of the gaze in images. When depicted people (or other participants with eyes) look at the viewer, we said, there is a ‘demand’, analogous to Halliday’s ‘demand of goods and services’, in language prototypically realized by the imperative mood. When they do not look at the viewer, the image realizes the function of ‘offering information’, in language prototypically realized by the indicative mood. But it has to be acknowledged that, to realize such a demand, someone must be represented as looking at the viewer. In other words, the interpersonal must piggyback on the ideational. It can never be as direct as in face-to-face communication. Second, in actual instances of multimodal communication there may be a metafunctional division of labour between the different modes. Figure 15.2 reproduces a document prepared by the Vice-Chancellor, Research, of an Australian university to explain the criteria for setting faculty research income KPIs – and at the same time for imposing these KPIs on faculty deans. Tables of this kind contain many words, but hardly any clauses. The role of language is by and large only lexical, realized by nouns and nominal groups. It is language, rather than the visual, which here becomes a ‘lexese’. The grammar is visual, realized by spatial relations – in this case the spatial structure of the

FIGURE 15.2  Research KPIs in an Australian University (detail). Source: Kress and Van Leeuwen 2021. Reproduced by permission of Taylor & Francis Group.

282

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

table which creates meaningful links between the information in the boxes, classifying it along the vertical axis and specifying these classes in terms of various attributes along the horizontal axis (the horizontal axis in fact fuses attribution with a timeline, as the table is horizontally also divided in a ‘historical performance’ and a ‘future performance’ section). As there are no verbs, there is no mood or modality, and hence language loses most of its interpersonal dimension, because it is mood and modality which, in Halliday’s words (1985: 70), make language ‘something that can be argued about – something that can be affirmed or denied, and also doubted, contradicted, insisted on, accepted with reservation, qualified, tempered, regretted and so on’. The table also lacks linguistically realized logical connectivity. A (partial) translation in linear prose might read something like this: The Key Performance Indicator ‘Grant Application Success’ is the responsibility of the PVC (Research). Success is measured as the percentage of national competitive grants received by a particular agency or scheme. To set a target historical performance should be taken into account and the previous year’s results taken as a baseline. If, say, the percentage of ARC Discovery grants was 1.44 and the percentage of ARC linkage grants 3.89 in 2004, and if the percentages were 2.01 and 4.00 in 2005, then the 2006 thresholds could be 2.00 and 4.00. As shown by the italics, conjunctions like ‘(in order) to’ and ‘if … then’ create logical connections between the items of information displayed in the table. But tables have no resources for expressing these kinds of connection. They create their own, different kinds of logical connections. They objectify the information, creating a kind of filing cabinet which organizes the information in a neat classification system. For this reason, texts of this kind become more context dependent. No longer being able to be read aloud, they need the spoken word for their completion, in this case a meeting in which the Deputy Vice Chancellor explained the KPIs to the faculty deans. It is here that the interpersonal will return and that the ideational take a backseat, becoming, for the most part, a matter of deictic reference to the information on the chart. Context dependency of this kind is characteristic of what elsewhere I have called ‘the new writing’ (Van Leeuwen 2008). It also characterizes PowerPoint slides, for instance, where the dot points often just list items of information and lack mood, modality and direct address. It is then up to the presenter to supply the explanations and the interpersonal elements, so as to attune the slides to the specifics of the occasion where, and the audience to which, they are presented (Djonov and Van Leeuwen 2014). In the next section I will discuss some further ways in which the linguistic model cannot always be followed closely.

RE-READING READING IMAGES

283

RE-READING ARNHEIM Discussions with my colleague Morten Boeriis brought out that Kress and I had been somewhat selective in the way we had used Arnheim’s (1982) work in Reading Images, focusing on aspects that integrated well with the Hallidayan model and neglecting others. Re-reading Arnheim’s work on composition we noted his emphasis on the multiplicity of vectors in images, on the way in which, in many images, ‘sheaves of concentric radii emerge from a centre or converge towards it’ (Arnheim 1982: 155). We applied this, for instance, to the well-known logo of Starbucks, where the vectors formed by the double tail of the mermaid suggest Starbuck’s global expansion through a multiplicity of what, in Reading Images, we called ‘non-transactional actions’, or, alternatively, customers moving to Starbucks from all directions, through a multiplicity of what, in Reading Images, we called ‘events’ (Boeriis and Van Leeuwen 2017b). In the same way Arnheim interpreted the converging vectors of central perspective as ‘a flow oriented towards a specific end … a happening in time, a directed sequence of events’ (1974: 298). In the Middle Ages, he explained (1974: 297), only God was infinite. In the Renaissance, the world became infinite, and this was, at the time, an uncomfortable thought for many. Hence ‘artists tended to avoid spelling out the vanishing point’ (1974: 297). Its location could only be guessed, but was blocked from view, as, for instance, in Leonardo’s Last Supper, where the figure of Christ overlaps the vanishing point, so that Christ, rather than a distant and unknowable future, dominates the centre of the image. In Koudelka’s photograph of the Prague uprising (Figure 15.1) the most salient vector is formed by the defiant gesture of the young student as he raises his arm towards the group of soldiers sitting on the tank and looking at him. I have often used this image in my teaching, suggesting it functions like a clause (‘The student defies the soldiers’) with the student as actor and the soldiers as goal. But there are many more vectors in Koudelka’s image. There are the soldiers’ boots pointing towards the student. There are the clasped hands of one of the soldiers. There is the soldier, high up, looking at the student with a sullen, hostile expression, and finally there is, in the background, the future created by the converging lines of the trees that line the street and by the tram rails and overhead wires. For the moment this future is obscured by the student’s gesture and the sense of hope it conveys. But what lies ahead is a dense cloud of gun smoke and the crushing of the Prague Spring. All this, in all its complexity, is simultaneously present in a complex network of vectors, and taken in at a glance. Another key concept from Arnheim (1982) we more or less ignored in Reading Images was the concept of ‘node’. Visual ideation not only draws on a wider range of processes than we foresaw at the time; it is also a great deal more complex than linguistic ideation, which does not have the equivalent

284

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

FIGURE 15.3  The Singing Man (Ernst Barlach, 1928) © MoMA. Reproduced with the permission of Scala Archives.

of the ‘node’, the point at which vectors intersect. Arnheim’s key example is reproduced in Figure 15.3. Here, in slightly abridged form, is his brilliant analysis: In seated figures, the main balancing centre is close to the floor, producing a solidly fastened focus, from which the various elements sprout upward in different directions. This is the basic constellation of vectors formed by the figure as a whole … Within the vertical/horizontal framework the body rises as a powerful diagonal, kept by the bracing arms from falling backward, yet not reposing in itself. It hovers between rise and fall. Song, the subject of the work, is an action, and it is an upward-directed action as is revealed by the posture of the head, which is raised like that of a singing bird. This mighty sounding, sent from the ground to heaven, is reflected in the opening and spreading shapes of the figure as a whole, the wide spaces between the legs and between legs and torso. And we notice that under the impact of the rousing song the right foot lifts from the ground, as though the column of the vertical were pulled up and soaring. But this is only half the story …

RE-READING READING IMAGES

285

although wholeheartedly given over to this song, our man is not addressing anybody. His eyes are tightly closed so that is attention is focused on the music from within. This counter-theme of concentration, indicated in the face, is spelled out more thoroughly in the other secondary node, the folded hands. They lock the brace of the arms, without whose grip the figure would fall apart. Severely contracted and safely tied to the ground, the sturdy figure thus lets forth the powerful music without exploding or being torn from its mooring. (Arnheim 1982: 169–71) The essential elements of a semiotic analysis are all present here – the signifying actions (bracing the arms, lifting the foot, raising the arms), whether they are transactional or not (‘our man is not addressing anybody’), what kind of actions they are (the material action of bracing the arms versus the mental action of concentration, signified by the closure of the eyes). But most importantly, Arnheim analyses the work as a network of nodes, a uniquely visual set of simultaneous volume–vector relations, rather than a clause-like, linear structure. The increasing importance of the network as a figure of visual ideation, capable of representing complexities that are well beyond what can be linguistically expressed, is now evident everywhere, and one of the sources of the power and influence of contemporary visualization (Lima 2011). That new developments in communicative practices can and should lead to new grammars was already foreseen by Halliday (1985: xxxv, emphasis added): ‘What kinds of grammar can we imagine, that would be different from those we have? … we can learn a lot by constructing the semantic system that lies behind some of the texts produced by small children and perhaps some of those produced by computers when they are being programmed for text generation’. Finally, Arnheim discussed facial vectors (1982: 163). ‘Can a face be described as a node, a constellation of vectors?’, he asked, and his answer was yes: ‘It can if, as we must, we perceive the facial features not as lifeless shapes but as vectors’ (Arnheim 1982: 163). The eyebrows, the mouth, and the wrinkles and folds of the smiles and frowns that can contract or open up the face all clearly delineate vectors, resulting from the coordinated tension of the facial muscles. If seen as processes, they differ from the visual ‘narrative structures’ we described in Reading Images in a number of ways. They do not clearly have actors and/or goals, as they do not meaningfully emanate from a volume or point at some other volume. And they are centred around the graded distinction between opening up (raised eyebrows, wide eyes, open mouth etc.) and closing down (knitted eyebrows, closed or half-closed eyes, pursed lips etc.), all this in various combinations (e.g. pursed lips and open eyes). This, too, is significant. Too often the representation of facial expression in images is studied on the basis of universalist psychological literature on facial expression, ignoring not only the cultural dimensions of everyday facial expression, but

286

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

also the fact that the vectors formed by the mouths, eyebrows and wrinkles in drawings (especially simple drawings like emojis and simple animations) do not necessarily reproduce real-life facial expressions, and may begin to form new vocabularies of vectors that need to be understood on the basis of attributes such as their extent, directionality and conspicuousness, rather than in terms of Darwinian ‘basic emotions’.

VISUAL MODALITY One of the systemic functional grammatical concepts we applied in Reading Images was modality, and our basic reason for doing so remains unchanged: all semiotic modes will have resources for expressing ‘how valid’ a representation is to be taken. But visual modality, as we described it in Reading Images, differs not only at the level of expression but also at the level of content. In linguistic modality validity is based on the concepts of probability and frequency. Visual modality, on the other hand, is based on the concept of ‘realism’, and ‘realism’ is not quite the same thing as ‘probability’ or ‘frequency’. Concepts of this kind have particular cultural histories. Ever since the Renaissance, visual modality has been based on perceptual criteria – on how much an image looks like what we think we would see if we saw what it depicts in reality. This is then expressed by the way in which the image represents visual detail, depth, gradations of colour and light and shade, and so on. For this reason, in the new edition of Reading Images, we use the term ‘validity’ rather than ‘modality’, adopting a term which is not only applicable to all semiotic modes, but also resonates with the socially and historically grounded way in which Habermas (1981) has used this term, and hence with modality as clearly grounded in the cultural, social and historical context. The Renaissance developed technologies of enhanced naturalistic observation and recording – in science (microscopes, telescopes), as well as in the visual arts, where Alberti’s theory of perspective soon led to the use of the camera obscura, allowing artists to project landscapes on a flat surface and measure its dimensions, and where artists stopped debating the symbolic meanings of colour and instead began to study meteorological, physical and optical phenomena – the sun, the clouds, raindrops, the reflection and refraction of light, and more. All of this eventually led to the invention of photography, which seeks, not to represent, but to record what is in front of the camera. Today, however, scientists as well as artists seek to construct rather than to (re)present perceptual reality, increasingly working with models, schemas for understanding reality, an approach which also underlies digital visualization (Lima 2011). Despite this, the ‘naturalistic’ validity criterion still reigns in many areas of visual production. Computer games and Hollywood computer animations are constantly praised for their ‘photorealism’. On the other hand, the quest for ‘photorealism’ has long

RE-READING READING IMAGES

287

been challenged in abstract art, in contemporary visualizations that represent reality in ways that no longer bear any traces of naturalism, and in the way Photoshop, Instagram and other similar technologies invite users to manipulate naturalistic validity, for instance, by changing photographs into drawings of different styles, or by adding or subtracting texture and lighting effects (Boeriis and Van Leeuwen 2017a; see also Ravelli and Van Leeuwen 2018). Not long after the introduction of Photoshop in 1989, all this engendered a debate about the validity of photography, which had traditionally been founded on the indexicality of the photograph, the idea that photography records reality without human intervention, by means of a chemical rather than a semiotic process. In a celebrated article, Fred Ritchin (1990: 28) warned that ‘the new malleability of the image may eventually lead to a profound undermining of photography’s status as an inherently truthful pictorial form’, and argued that the credibility of photo journalism would henceforward have to be guaranteed by the reputation of the photographer and the institution for which he or she works, rather than by the inherent truthfulness of photography. Others, such as Rosler (1991), argued for what, in Reading Images, we called abstract modality – ‘a truer truth, one closer to conceptual adequacy’, which does not rely on the ‘illusionistic elements in the surface of (and even definition of) reality’ (58).

PARAMETRIC SYSTEMS In Reading Images, we followed Systemic Functional Linguistic practice in representing our grammatical systems as system networks of essentially binary choices (although some simultaneous choices are of course possible in system networks). Just as Systemic Functional Linguistics described, for instance, the mood system of English in terms of binary distinctions between basic speech functions such as offering and demanding information and offering and demanding goods and services, so Reading Images described one of its key interpersonal systems in terms of a binary choice between offering and demanding: if someone looks at the viewer from within an image, we argued, something is ‘demanded’ from the viewer (sympathy, respect, erotic attraction etc.). If this is not the case, the image simply ‘offers information’ about what is represented. This binary choice can then lead to further binary choices, specifying further kinds of demands, for instance, realized by facial expressions and/or gestures. A number of key aspects of visual communication, however, were not discussed in Reading Images, especially the more abstract qualities of images, such as colour, texture and the graphic shape of abstract elements and letter forms, for instance, in logos. In studying the semiotic mode of colour, Kress and I (2002) did not find it possible to construct a satisfactory system network. All

288

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

choices were not only graded, but also simultaneously rather than sequentially selected. To model this, we drew on Jakobson and Halle’s (1956) ‘distinctive feature’ theory, in which speech sounds are characterized by a number of simultaneously present features. A [p], for instance, is unvoiced and labial and plosive, among other things. We then presented a framework for interpreting colour, based on a range of simultaneous features – not only hue, but also ‘value’ (how light or dark colours are), ‘saturation’ (how intense or pale they are), purity (to which degree they are mixed), luminosity, lustre and luminescence (to which degree they glow from within, shine or are themselves sources of light), modulation (to which degree they are flat or variegated with varied tints and shades) and differentiation (to which degree the overall palette or colours is varied or not, with monochrome as the extreme on one end). But we also diverged from ‘distinctive feature’ theory. For Jakobson, who had worked with Trubetzkoy, distinctive features function only to distinguish phonemes, and hence words, from each other. They do not add meaning. For us, expression makes meaning in its own right, not only in the case of colour, but also in speech (articulation and voice quality) and in writing (graphology and typography). A speaker’s voice, for instance, might be high-pitched or low-pitched and rough or smooth and tense or lax, and so on, and all these qualities bring meaning potentials with them (van Leeuwen 2014). The letter forms of written or printed language will be rounded or angular, and bold or light, and regular or irregular, and so on, and such features, too, carry meaning potentials (Van Leeuwen 2006). In Kress and van Leeuwen (2001) we called such ‘expression modes’ media, using the term the way artists do, to indicate the materiality of their signifiers. The kinds of meaning such ‘media’ create, we argued, and the way they create them, are specific. ‘Media’ create style, and style signifies identity, saying something about the speaker or writer, about the organization for which s/he works, or about the communicative genre s/he is engaged in, and they do so in ways that that have aesthetic and emotive resonance (Van Leeuwen 2015). It is no accident that style has become increasingly important in the age of ‘branding’, which today is no longer restricted to advertising products, but also used in expressing the identities of all kinds of organizations, commercial and non-commercial, and even the identities of individuals who seek to style themselves in accordance with lifestyle values and attributes that are ultimately market driven. At the same time, ‘media’ also do textual work, visually identifying the different elements of texts and emphasizing the similarities of others, or visually separating foreground from background, title from main copy, bullet points from hyperlinks and so on (cf. van Leeuwen 2011a). To make meaning with ‘media’ is to work directly, and often creatively, with materiality. This is even the case with technologies such as the older ‘mass media’ and the newer internet media, which today no longer only serve

RE-READING READING IMAGES

289

to record, preserve and distribute texts, leaving their meaning unchanged, but become ‘media’ of their own, as we already described in Kress and Van Leeuwen (2001). To describe how such meaning making happens, we drew on the metaphor theory of Lakoff and Johnson (1980) for whom meaning derives from metaphors based on common, material experience, whether bodily experience, experience in the handling of materials or experience of interactions with people. The experiential meaning potential of light and dark, for instance, rests on our experience of night and day, which is a source of a wide range of metaphors, both in everyday language and in visual communication. Precisely how this wide range is narrowed down in a given instance then depends on the context. Light colours may signify light-heartedness in Hollywood comedies or divine revelations in paintings by Rembrandt, and these are only two of the many possibilities. As for the handling of tools and materials, experience tells us why graphic shapes, including letter forms, may be irregular – for instance because they are produced by children who have not yet mastered writing or drawing, because of inadequate tools, or because a writer is intoxicated or unable to see what s/he is writing. In this way irregularity can become a metaphor for childish innocence or playfulness, as well as for rebelliousness or deviance, to mention just two possibilities (Johannessen and Van Leeuwen 2018a). The multiple affordances of media allow for much leeway and creativity, both in the creation and the interpretation of identity meaning. Over time, however, configurations of colours or shapes may become conventional in specific historical and cultural contexts, and can then be recognized as meaningful on that basis. A combination of irregularity, high colour saturation and differentiation and low colour modulation, for instance, is today instantly recognized as ‘childish’ and commonly used on the covers of children’s books and in the design of many other products for children. Such configurations become ‘myths’ in the sense in which Barthes (1973) used that term. Their meaning is based on provenance (historical, geographical, cultural, contextual etc.), and they carry the associations which viewers have with the ‘place’ from where they come, for instance, childhood. Finally, our work on parametric systems made us aware of another aspect of visual communication we had ignored in Reading Images. Visual communication is not only about images, it is also about the use of layout, colour and typography in the ‘new writing’, and about decoration (van Leeuwen 2011b) – the shapes of everyday objects, the patterns of the carpets in airports or the cladding of new buildings, the design of the page furniture and backgrounds provided by PowerPoint, and so on. To these forms of ‘environmental’ visual communication, we do not pay the kind of focused attention we pay to ‘texts’, but that does not mean we do not see them and it does not mean they do not make meaning. To date semioticians have not yet paid much attention to this, but artists such as Rothko or Olafur Eliasson have long moved to creating

290

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

environments rather than objects, designers have announced that ‘pattern is back’ (Cole 2007), cultural studies scholars have begun to study ‘atmosphere’ (Edensor and Sumartojo 2015) and contemporary psychoanalysts such as Bion explore the ways in which we ‘read yet do not remember’ (1967: 163) and the thought processes of which we are not consciously aware but which nevertheless ‘translate into something we do know and can communicate’ (1989: 54). Given their importance in branding, which now extends to the branding of environments (Aiello and Dickenson 2014), semioticians would do well to pay attention.

CONCLUSION My conclusion can be brief. Semiotic approaches to visual communication can, and should, draw on the achievements of relevant schools of linguistics, as they have done successfully throughout the twentieth and into the twenty-first centuries. Systemic Functional Linguistics has, in the last few decades, been especially fruitful in this regard, as has cognitive metaphor theory (Forceville and Urios-Aparisis 2009). However, semiotic approaches must also draw on other traditions of inquiry into the visual arts and visual communication, and engage with the physical affordances (Gibson 1986) and concrete experiences of materiality (cf. Johannessen and van Leeuwen 2018b) and with the cultural contexts in which the semiotic resources of visual communication developed, for instance, the histories of the kind of systems we described in Reading Images. Only the cultural context can explain, rather than just describe, why these systems are the way they are. As Halliday has often said, language is the way it is because of the functions it must serve in social life, and as society changes so will language. This applies equally to non-linguistic modes of communication, and to the ways in which they combine in texts (Van Leeuwen 2017). For all these reasons, multimodality studies, and linguistics itself, should remain firmly anchored in the humanities, even when it engages, as it should, with psychology and neuroscience. It should resist the universalist and scientistic tendencies which have, from time to time, tempted linguists, as did Mathesius in the early days of linguistically inspired multimodality studies, when he critiqued the linguistics of his day for treating language like a biological organism, and argued that linguistics ‘must struggle against excessive, mechanistic simplifications of language phenomena’ (quoted in Matejka 1976: 266).

ACKNOWLEDGEMENT Thanks to Emilia Djonov for valuable comments and suggestions.

RE-READING READING IMAGES

291

REFERENCES Aiello, G. and G. Dickenson (2014), ‘Beyond Authenticity: A Visual-Material Analysis of Locality in the Global Redesign of Starbucks Stores’, Visual Communication, 13 (3): 303–22. Arnheim, R. (1974), Art and Visual Perception, Berkeley, CA: University of California Press. Arnheim, R. (1982), The Power of the Centre, Berkeley, CA: University of California Press. Barthes, R. (1967), Elements of Semiology, New York: Hill and Wang. Barthes, R. (1973), Mythologies, St Albans: Paladin. Bateman, J. A. and K.-H. Schmidt (2012), Multimodal Film Analysis – How Films Mean, London: Routledge. Bion, W. R. (1967), Second Thoughts, London: Heinemann. Bion, W. R. (1989), Two Papers: The Grid and Caesura, London: Karnac. Birdwhistell, R. L. (1970), Kinesics in Context: Essays on Body Motion Communication, Philadelphia, PA: University of Pennsylvania Press. Boeriis, M. and J. Holsanova (2012), ‘Tracking Visual Segmentation: Connecting Semiotic and Cognitive Perspectives’, Visual Communication, 11 (3): 259–81. Boeriis, M. and T. Van Leeuwen (2017a), ‘Towards a Semiotics of Film Lighting’, in J. Wildfeuer and J. A. Bateman (eds), Film Text Analysis – New Perspectives on the Analysis of Filmic Meaning, 24–45, London: Routledge. Boeriis, M. and T. Van Leeuwen (2017b), ‘Vectors’, in O. Seizov and J. Wildfeuer (eds), New Studies in Multimodality – Conceptual and Methodological Elaborations, 15–35, London: Bloomsbury. Bühler, K. (1990 [1934]), The Theory of Language: The Representational Function of Language, Amsterdam: John Benjamins. Cole, D. (2007), Patterns – New Surface Designs, London: Laurence King. Djonov, E. and T. Van Leeuwen (2014), ‘Bullet Points, New Writing and the Marketization of Public Discourse: A Critical Multimodal Perspective’, in E. Djonov and S. Zhao (eds), Critical Multimodal Studies of Popular Discourse, 232–50, London: Routledge. Durand, J. (1970), ‘Rhétorique et image publicitaire’, Communications, 15: 70–96. Edensor, T. and S. Sumartojo (2015), ‘Designing Atmosphere’, Visual Communication, 14 (3): 251–65. Forceville, C. and E. Urios-Aparisi, eds (2009), Multimodal Metaphor, Berlin: De Gruyter. Gibson, J. J. (1986), The Ecological Approach to Visual Perception, Hillsdale, NJ: Laurence Erlbaum. Gregory, R. L. (1970), The Intelligent Eye, New York: McGraw-Hill. Habermas, J. (1981), The Theory of Communicative Action, Vol. 1: Reason and the Rationalization of Society, Boston, MA: Beacon Press. Halliday, M. A. K. (1985), An Introduction to Functional Grammar, London: Arnold. Honzl, J. ([1941] 1976), ‘Dynamics of the Sign in the Theatre’, in L. Matejka and I. R. Titunik (eds), Semiotics of Art – Prague School Contributions, 74–93, Cambridge, MA: MIT Press. Itti, L. and C. Koch (2000), ‘A Saliency-Based Search for Overt and Covert Shifts of Visual Attention’, Vision Research, 40: 1489–506. Jakobson, R. (1960), ‘Closing Statement: Linguistics and Poetics’, in T. A. Sebeok (ed.), Style in Language, 351–77, Cambridge, MA: MIT Press.

292

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Jakobson, R. and M. Halle (1956), Fundamentals of Language, The Hague: Mouton. Johannessen, C. M. and T. Van Leeuwen, T. (2018a), ‘(Ir)regularity’, in C. M. Johannessen and T. van Leeuwen (eds), The Materiality of Writing – A TraceMaking Perspective, 175–92, London: Routledge. Johannessen, C. M. and T. Van Leeuwen, eds (2018b), The Materiality of Writing – A Trace-Making Perspective, London: Routledge. Kress, G. and T. Van Leeuwen (1990), Reading Images, Geelong: Deakin University Press. Kress, G. and T. Van Leeuwen (2001), Multimodal Discourse – The Modes and Media of Contemporary Communication, London: Arnold. Kress, G. and T. Van Leeuwen (2002), ‘Colour as a Semiotic Mode: Notes for a Grammar of Colour’, Visual Communication, 1 (3): 343–69. Lakoff, G. and M. Johnson (1980), Metaphors We Live By, Chicago: University of Chicago Press. Lima, M. (2011), Visual Complexity – Mapping Patterns of Information, New York: Princeton Architectural Press. Martin, J. R. (1992), English Text – System and Structure, Amsterdam: John Benjamins. Matejka, L. (1976) ‘Postscript. Prague School Semiotics’, in Matejka, L. and I. R. Titunik (eds), Semiotics of Art – Prague School Contributions, 265–90, Cambridge, MA: MIT Press. Metz, C. (1974), Film Language – A Semiotics of the Cinema, New York: Oxford University Press. Mukařovský, J. ([1936] 1976), ‘Poetic Reference’, in L. Matejka and I. R. Titunik (eds), Semiotics of Art – Prague School Contributions, 155–63, Cambridge, MA: MIT Press. Norris, S., editor (2016), Multimodality: Vol I–IV, London: Routledge. O’Toole, M. (1994), The Language of Displayed Art, Leicester: Leicester University Press. Ravelli, L. J. and T. Van Leeuwen (2018), ‘Modality in the Digital Age’, Visual Communication, 17 (3): 277–97. Ritchin, F. (1990), ‘Photojournalism in the Age of Computers’, in C. Squiers (ed.), The Critical Image, 28–37, Seattle: Bay Press. Rosler, M. (1991), ‘Image Simulations, Computer Manipulations, Some Considerations’, Ten-8, 2 (2): 52–63. Sapir, E. (1949) Selected Writings of Edward Sapir, Berkeley and Los Angeles: University of California Press. Sklovskij, V. (1925), Theory of Prose, Elwood Park, IL: Dalkey Archive Press. Van Leeuwen, T. (2001), ‘Semiotics and Iconography’, in T. van Leeuwen and C. Jewitt (eds), Handbook of Visual Analysis, 92–118, London: Routledge. Van Leeuwen, T. (2006), ‘Towards a Semiotics of Typography’, Information Design Journal, 14 (2): 139–55. Van Leeuwen, T. (2008), ‘New Forms of Writing, New Visual Competencies’, Visual Studies, 23 (2): 130–5. Van Leeuwen, T. (2011a), The Language of Colour: An Introduction, London: Routledge. Van Leeuwen, T. (2011b), ‘The Semiotics of Decoration’, in K. L. O’Halloran and B. A. Smith (eds), Multimodal Studies – Exploring Issues and Domains, 115–27, London: Routledge.

RE-READING READING IMAGES

293

Van Leeuwen, T. (2014), ‘Parametric Systems: The Case of Voice Quality’, in C. Jewitt (ed.), The Routledge Handbook of Multimodal Analysis, 76–85, London: Routledge. Van Leeuwen, T. (2015), ‘Looking Good: Aesthetics, Multimodality and Literacy Studies’, in J. Rowsell and K. Pahl (eds), The Routledge Handbook of Literacy Studies, 426–39, London: Routledge. Van Leeuwen, T. (2017), ‘A Social Semiotic Theory of Synesthesia? – A Discussion Paper’, HERMES – Journal of Language and Communication in Business, 55: 105–19. Ziou, S. D. and S. Tabbone (1998), ‘Edge Detection Techniques: An Overview’, Journal of Pattern Recognition and Image Analysis, 8 (4): 537–59.

294

CHAPTER SIXTEEN

The Appraisal Framework and Analyses of Journalistic Discourse Objectivity, Subjectivity and Attitudinal Positioning PETER R.R. WHITE

THE INITIAL CHALLENGE: EXPLICATING NOTIONS OF JOURNALISTIC SUBJECTIVITY Linguists with an interest in journalistic discourse have long been challenged by the apparently popular view that at least some types of news reporting can and should be ‘objective’. This, of course, is the notion that news journalism should confine itself to the ‘facts’ of the matter and ‘accurately’, ‘impartially’ and ‘disinterestedly’ report what has happened and been said. That this notion is often taken for granted is evidenced by the fact that the legislation which dictates the functions of Australia’s national broadcaster, the Australian Broadcasting Corporation, states that its board must ‘ensure that the gathering and presentation by the Corporation of news and information is accurate and impartial according to the recognized standards of objective journalism’ (Australian Broadcasting Corporation Act 1983 (Cth), s 8). What is entailed in the notion of ‘objectivity’ is, of course, widely contested in academic settings, with many holding the view that the concept is of little value, that

296

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

it misleadingly asserts an impossible separation of the observer from her/his individual subjectivity. Certainly, within journalism studies there have been forthright challenges to the notion of ‘objective journalism’. Thus Schudson noted that ‘objectivity in journalism … came to be looked upon as the most insidious bias of all. For “objective” reporting reproduced a vision of social reality which refused to examine the basic structures of power and privilege. It was not just incomplete … it was distorted. It represented collusion with institutions whose legitimacy was in dispute’ (1981: 160). In similarly rejecting the notion of journalistic objectivity, Dennis and Merrill contend that every news report is always ‘judgemental, value-loaded, incomplete, and distorted as to reality. That is the nature of journalism. That is the nature of any kind of communication’ (1984: 106). The need to deal in a theoretically principled way with the notion of journalistic objectivity, and its counterpart, journalistic subjectivity, arose when I was involved in the ‘Write It Right’ research project for the New South Wales Disadvantaged Schools Program in the early 1990s. I was part of a team, with Rick Iedema, Susan Feez and James Martin, given the task of developing linguistics-based media literacy resources for secondary school media studies teachers. The stated objectives of the project were as follows: to research the nature of literacy demands within and across industrial sectors [in this case the demands of the news media industry] and relate these findings to literacy demands in the NSW secondary curriculum Key Learning Areas. The primary goal of the project was to give secondary teachers the tools to analyse and teach explicitly the language associated with their subject areas. (Iedema, Feez and White 1994: 5) Given the prominence of the notion of ‘objectivity’ in community debates about the nature and standing of the news media, it was obviously one of the concepts which the project team would need to deal with. We might, of course, have simply followed the lead of scholars such as Dennis and Merrill (as cited above) and basically proposed that teachers entirely discount the possibility of an ‘objective’ news report, that is, simply proposed that news reports are ‘judgemental, value-loaded, incomplete’ and probably ‘distorted as to reality’. But this might well not have resonated with teachers, given that it is so widely taken for granted (even in legislation) that there is ‘objective journalism’. More importantly, to do so would be to ignore the fact that there appear to be clear linguistic differences, at least in some cases, between the supposedly ‘objective’ reporting of ‘hard news’ and the ‘subjective’ interpretation, evaluation and argumentation of journalistic commentary and analysis. Although, as linguists, we might broadly agree with Dennis and Merrill that, like all texts, all news

APPRAISAL AND JOURNALISTIC DISCOURSE

297

reports are ‘judgemental, value-laden and incomplete’, this did not mean that we wanted to simply lump together all news journalism as equally ‘subjective’. Yes, we did recognize that news reports could be ‘persuasive’, could function to position readers to adopt a particular interpretation of events. But at the same time, we did not want to forgo the opportunity to provide insights into the potentially distinctive linguistic properties associated with such texts being held to be ‘factual’, ‘accurate’ and ‘impartial’. Our plan was to explore patterns in the choice of language in journalistic texts by which certain meanings are ‘favoured’ or ‘disfavoured’ (appearing more often or less often than in other settings) and to determine whether there were any correlations between these patterns of semantic favouring/disfavouring and the sub-classifications deployed by the news media itself, for example, ‘hard news’, ‘analysis’, ‘commentary/opinion’ and ‘human interest’. In terms of Systemic Functional Linguistics (SFL) theory, was ‘hard news’ reporting a register (or sub-register) of journalistic language which differed from, for example, the register/sub-register of journalistic commentary in ways which might account for the perception that ‘hard news reporting’ was the domain of journalistic ‘factuality’, ‘impartiality’ and ‘disinterestedness’? The available linguistics-informed journalistic discourse analysis literature available at the time focused largely on the potential of the news report to influence public opinion and perception. This literature typically was concerned, for example, with exposing the ‘slantedness’ of particular news items, with uncovering the ideological workings by which selected news reports sustained social hierarchies and inequalities, naturalized prejudicial stereotypes, covertly promoted the interests of the wealthy and powerful, and so on (see e.g. Fowler et al. 1979; van Dijk 1988; Fowler 1991; Fairclough 1995). While this was, of course, important work, it tended, as indicated, to be directed at uncovering the ideological underpinnings of individual news items. We needed to be able to offer more general understandings of the persuasive potential of news reporting as a verbal style or register, while at the same time also offering some suggestions as to why it might, nevertheless, continue to be held to be ‘objective’.

EXPLORING A LINGUISTIC BASIS FOR THE ‘OBJECTIVE JOURNALISM’ VERSUS ‘SUBJECTIVE JOURNALISM’ DICHOTOMY This meant that, in our estimation, there was little prior literature on which we could draw for our purpose of providing teachers with an account of those linguistic properties which underlie the perception that certain news items are ‘objective’ even while those news items can be shown to favour particular

298

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

value-laden interpretations and attitudinal positionings. We also needed to be able to account in a systematic way for the differences in the communicative working of news reports versus the explicitly subjective modes of journalism, that is, journalistic commentary. Over several months the project team started to outline elements of an account of the resources of evaluative meaning making in English which would come, initially, to be known as ‘appraisal theory’ and subsequently as the ‘appraisal framework’. As systemic functional linguists, our discussions were located in the account provided by the theory of those meaning-making resources associated with what is termed the ‘tenor’ of a text. Tenor is one of the three parameters of variation which, according to SFL theory, condition the language options taken up by speakers/writers in any given communicative event: tenor, field and mode (Halliday and Hasan 1987: 26). Tenor is a matter of who is involved in the communication, the relationships which hold between participants, their social roles, their statuses, their identities, their degree of intimacy and so on. Naturally, therefore, we viewed the question of the basis on which a text might be viewed as more or less ‘objective’ or more or less ‘subjective’ as likely to be associated with tenor, with those meanings which serve this interpersonal function. If the ‘subjectivity’ of a text is a matter of how the speaker/writer is revealed as personally involved in the communication, the perspective or personal orientation they bring to the text, then presumably a text viewed as ‘objective’ might be one in which mechanisms are deployed by which the ‘who’ of the text is obscured, by which the personal perspective they bring to the text is backgrounded in some way. Specifically, we turned to the then recent work by Poynton (1989) in which she had proposed that a key parameter of variation in the ‘tenor’ of a text (the ‘who’ of the text) was what she termed ‘affect’ – referring to what Halliday (1978: 33) had described as the ‘degree of emotional charge’ in the relationship between participants. Poynton was proposing that affect, as construed in language, can be either positive or negative and permanent or transient. At the time of our project, Poynton’s account of affect had recently been developed further by Martin (1992: 533–6) into a more delicate taxonomy of ‘affections’, in which affectual responses, as construed through language, are classified as either positive or negative, as directed at self or other, as a ‘surge’ of emotion or a ‘disposition’ to an emotion state, and then as one of the following subtypes: ‘discord’ (negative) versus ‘satisfaction’ (negative); ‘insecurity’ (negative) versus ‘security’ (positive); ‘frustration’ (negative) versus ‘fulfillment’ (positive) (Martin 1992: 534). In this we saw a potentially useful line of enquiry, a line which might focus on what was central to both Poynton and Martin’s accounts – that one key aspect of tenor (the ‘who’ involved in the text) was a matter of choices speakers/ writers make in conveying positive or negative attitudes. Accordingly, it seemed

APPRAISAL AND JOURNALISTIC DISCOURSE

299

fruitful to explore ways in which journalistic texts functioned to convey positive or negative attitudes, to position the reader to take a positive or negative view of events and those involved in these events. Might it be that different subtypes of news journalism deploy in different ways meanings by which such attitudinal assessments are conveyed, or make use of different types of attitudinal meaning? In developing some answers to these and related questions, the project team laid the foundations for the appraisal theory/framework mentioned above.

PATTERNS OF POSITIVELY/NEGATIVELY ATTITUDINAL MEANING MAKING IN JOURNALISTIC DISCOURSE We were concerned, therefore, with investigating what might be typical patterns in the way news journalism texts positioned the reader to adopt positive or negative viewpoints of events and participants in those events. We surveyed a wide range of texts from Australian newspapers and news radio broadcasts, spanning those which presented as ‘straight’ or ‘hard news’ reports, those which were labelled ‘analysis’, and those which were clearly labelled commentary or opinion. The following extracts from the texts which we collected for analysis are illustrative of the differences we immediately observed in the ways in which different news items positioned readers/listeners attitudinally. Extract 1 [Radio 2UE, 11/8/1992, 8 a.m. news bulletin] The Government and the ACTU [Australian Council of Trade Unions] will meet in Canberra today to thrash out a new wages agreement before next week’s budget. Lyndel Curtis reports: There’s been speculation the wages package will involve a rise of $9 or $10 a week, for those workers who haven’t had rises through enterprise bargaining deals. However the amount hasn’t been finalized and the unions say the discussions will also involve social wages such as Medicare and child care. As well the unions are expected to seek a commitment on varying tariff cuts in the textile, clothing and footwear industry, if employment is hit hard by the changes. [continues] Extract 2 [Radio 2UE, 11/8/1992, 8 a.m. news bulletin] The Chamber of Commerce says it’s ridiculous but the federal government and the ACTU are to agree on a national pay rise when the country is going through the worst recession in 60 years. It’s expected both sides will agree today to a wage rise of between $8 and $10 a week for Australia’s

300

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

7.7 million workers. The Chamber’s chief economist, Brent Davis, says we simply can’t afford a wage rise at the moment. [continues] Extract 3 [Sydney Morning Herald, 10/18/92, p. 11] Spotlight on Bush’s House Party By Pilita Clark, Herald Correspondent HOUSTON, Monday: All the glitzy balloons and brass bands ready to be unleashed for this week’s Republican National Convention suddenly looked rather small and dowdy yesterday as attention was diverted to the showdown looming between President Bush and Iraq’s President Saddam Hussein. And that is a pity, because this year’s convention was shaping up to be one of the most interesting Republican gatherings in a long time. Usually it is the Democrats who strain desperately, and unsuccessfully, to disguise their inner tensions and divisions at convention time. For the past two presidential elections, for instance, the Democratic nominee’s attempts to placate the Rev Jesse Jackson have provided lively sport for onlookers. But this year the two major parties appear to have swapped roles. The Democrat’s convention in New York last month was notable for its general harmony and unified support for its nominee, Mr Bill Clinton. [continues] Extract 4 [Sydney Morning Herald, 19/10/92, p. 4] A Cheap Slur on Childless by Adele Horin Dr Hewson’s [then Leader of the Opposition in Australia’s federal Parliament] insensitive remarks imply that the significant proportion of Australians who are childless cannot be trusted. They are somehow morally inferior to parents, and the men in particular are not ‘full-blooded’. He also assumes that just because people do not have their own children they do not like children. [continues] Extracts 3 and 4 obviously include meanings which position readers to adopt positive or negative viewpoints. Extract 3 was labelled ‘analysis’ by the newspaper and extract 4 was located in the paper’s op-ed (i.e. commentary) section. They are both texts in which the journalist author is overtly attitudinal in announcing her own personal evaluative assessments of events and people. Thus Horin, in extract 4, positions the reader to take a negative view of ‘Dr Hewson’ on the basis of her estimation that he was ‘insensitive’ in his remarks and Clark, in extract 3, positions the reader to take a negative view of the

APPRAISAL AND JOURNALISTIC DISCOURSE

301

Republican convention on the basis of her estimation that it came to appear ‘small and dowdy’. She also opines that, in the past, the Democrats have been ‘unsuccessful’ in ‘disguising their inner tensions’. In extract 2, the attitudinal positioning and the journalist author’s role in this is rather less straightforward. Certainly, the author does offer an explicit attitudinal assessment of the country’s economic circumstances as ‘the worst recession in 60 years’, but the negative assessment of the wage rise being agreed to by the government is attributed to external sources – to the ‘Chamber of Commerce’ who reportedly assesses it to be ‘ridiculous’ and to the chamber’s chief economist who reportedly assesses it to be unaffordable. As well, the author juxtaposes the government agreeing to the pay rise with the observation that it is taking place during ‘the worst recession in 60 years’ – a juxtaposition which presumably has some potential vis-à-vis the reader’s view of the pay rise. There is some complicating, therefore, of the journalist author’s role in any attitudinal positioning of the reader in the sense that it is an external ‘expert’ source who passes judgement on the pay rise, not the author. Similarly, while the author does juxtapose the pay rise with the current adverse economic circumstances, this is not, of itself, an explicit authorial assessment by the author of the pay rise decision. If the reader does end up forming a negative view of the pay rise and those involved (the government and trade unions), this is arguably the outcome of the reader inferring certain negative conclusions, and is not the result of any negative assessments being overtly advanced by the author. Extract 1 seems free of any attitudinal positioning on the part of the journalist author (or anyone else) vis-à-vis the wage rise and those involved in agreeing to it. It would seem to be unproblematically ‘factual’ in reporting the details of the wage agreement, its timing, to whom it will apply and so on. There is perhaps a hint of an authorial attitude in the use of the term ‘thrash out’ to describe the negotiations between the government and the trade unions. Perhaps there is a suggestion in this use of this term that such negotiations will be difficult or fraught (so presumably not a positive experience for those involved), but this does not lead to any obvious positive or negative assessment of the wage rise decision itself or those involved in this ‘thrashing out’. In making such observations, we came to the view that it would be useful in analysing news journalism texts to track patterns with respect to the extent that attitudinal assessments were (1) explicitly asserted by the journalist author, (2) attributed to external sources and (3) activated indirectly through observations which invite the reader to make attitudinal inferences. In the following I will consider what flowed from these observations first in terms of the project’s concern with the subjectivity–objectivity dichotomy and second in terms of the subsequent, continued development of appraisal theory. On the basis of news items exemplified by extract 1, it is possible to conclude that some news items do not in any obvious way act to position the reader

302

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

attitudinally vis-à-vis the events being described nor any of the participants involved in those events. This is with respect to the journalist author’s own observations and assertions, the observations/assertions of any quoted sources and with respect to any content which might have the potential to trigger attitudinal inferences on the part of the reader. Thus extract 1 reports the decision by the government to approve a wage rise without any positioning of the reader to view the decision or those involved in a positive or negative light. There is thus a basis for regarding such a report as ‘objective’, to the extent that it is free of any attitudinal positioning vis-à-vis the event under consideration. This is not, however, to suggest that the version of events offered is in some way value free or that it is not reflective of a particular value-laden way of understanding and interpreting the world. It is ‘subjective’ in being one of many ways of reporting this news – but one of any number of possible ‘versions of events’. In contrast, texts of the type exemplified by extract 2 could hardly be treated as ‘neutral’ or ‘impartial’ in the manner of the type of text exemplified by extract 1. As discussed, it does have the potential to position the reader attitudinally – to take a negative view of the wage rise and those engaged in agreeing to it. On the other hand, as also noted, the journalist author does not present as personally advocating this, refraining from explicitly stating a position on the matter. Thus, the attitudinal positioning is more indirect. What conclusions then should be drawn here in terms of notions of journalistic objectivity and subjectivity and, in turn, what are the implications for developing those parts of the SFL theory of language which are concerned with subjectively evaluative meanings and the construing of interpersonal relations? Our solution with respect to outlining for teachers the communicative workings of such texts was to propose that they entail what can be termed a ‘strategic impersonalization’. This is achieved by the author refraining from overtly contributing to the attitudinal workings of such texts and presenting as ‘just’ reporting the views of ‘relevant’ external spokespersons and reminding the reader of the ‘fact’ that at that time Australia was experiencing a serious economic downturn. In this, of course, items exemplified by extract 2 stand in obvious contrast with texts exemplified by extracts 3 and 4, where there is no such ‘impersonalization’. Thus, we could point to the need to attend to the potential of these two key aspects of this type of reporting (quoting and reporting of attitudinal charged ‘facts’) to function attitudinally, even while the journalistic author can remain in the background, so to speak. In the materials the project team prepared for teachers, the attitudinal functionality of these two features was obviously noted and discussed. With respect to instances where ‘factual’ reporting of actions and events may function to imply an attitudinal assessment, we proposed the notion of the attitudinal ‘token’. We explained such ‘tokens’ in the following terms.

APPRAISAL AND JOURNALISTIC DISCOURSE

303

Society establishes a set of connections between certain behaviours and certain [positive/negative attitudes]. Physical violence, intentionally misleading others, for example, are classed as wrong, while donating to charity or going to the rescue of people in danger are associated with positive values. By describing such events, the author attaches the corresponding [attitudinal assessments] to the text. (Feez, Iedema and White 2008: 213) Thus ‘attitudinal tokens’ are meanings which are not of themselves lexically attitudinal but which, on account of the cultural frames on which they rely, have the potential to imply or activate an attitudinal assessment. For example, by reporting that the government has agreed to grant workers a pay rise during a serious economic downturn, extract 2 invites the inference by the reader that the government has acted irresponsibly. Whether or not the reader reaches such a conclusion will be, of course, dependent on her/his reading position – the particular ideologically charged cultural frame s/he brings to the text. In the materials the team prepared for teachers, we made some progress in accounting for these ‘tokens’: the linguistic mechanisms by which attitudinal assessments may be ‘invoked’ in this way. This account has been developed substantially in subsequent years in many publications (see e.g. White 2006, 2016; Hood and Martin 2007; Hood 2010; Don 2016, 2018). The observations the team developed re the potential for attribution to function attitudinally were relatively limited. These were developed in subsequent work on what is now termed the ‘engagement’ system in the appraisal literature – an account of the resources by which the speaker/writer engages dialogistically with prior utterances on the current topic and anticipates responses to what is being asserted (see White 1998, 2000, 2003, 2010, 2020, 2021). In this work proposals are developed as to how the journalistic author may ‘neutrally’ attribute viewpoints and versions of events to external sources, may attribute so as to ‘distance’ the authorial voice from such quoted material, or may indicate its alignment with that material via attributions which ‘endorse’ attributed propositions. In later work, I have explored additional mechanisms by which attributed material can be ‘favoured’ or ‘disfavoured’ so that the reader is positioned to view attributed content as credible and well-founded or, alternatively, to question its credibility and well-foundedness (White 2012; see also Paula 2011). Along similar lines, Hao and Humphrey (2012) developed an account of the mechanisms by which the quoted sources themselves may be ‘burnished’ or ‘tarnished’. In terms of issues around the notion of ‘objective’ and ‘subjective’ journalism, texts exemplified by extracts 3 and 4 were predictably unproblematic. The authorial presence associated with explicit authorial attitudes provides obvious grounds for treating them as ‘subjective’. They represented the ‘opinion’ side

304

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

of the ‘facts versus opinion’ dichotomy so widely alluded to in discussions of journalistic ‘objectivity’. In terms of the development of appraisal theory/appraisal framework, the meanings we observed in these types of texts were crucial in the development of the account of attitudinal meanings now available in the appraisal literature. The types of attitudinal assessment we observed here made it obvious that the notion of ‘affect’ as formulated by Poynton would need to be wider than the category of ‘emotion’. There are no meanings in the above texts which in an obvious or direct way report the emotional responses or states of either the journalist author or of participants in the news events under consideration. ‘Affect’ (or Halliday’s notion of ‘emotional engagement’) would need to be extended to incorporate positive/negative assessments of human behaviour and character, since evaluations of this type are obviously central to this type of journalism. Along similar lines, Martin’s formulation of affect did not obviously encompass the kinds of attitudinal meanings operating in these extracts, oriented as it was to characterizing and taxonomizing emotional states and reactions. As a consequence, we proposed a second system of positively/ negatively attitudinal meanings, to stand beside Martin’s system of affect, what TABLE 16.1  The System of Judgement Judgements to do with social admiration or contempt Social esteem

Positive (admire/be captivated by)

Negative (hold in contempt/pity)

Normality (usuality)

normal, outstanding, lucky, remarkable

peculiar, odd, eccentric, unlucky, abnormal

Capacity (ability)

competent, powerful, weak, incompetent, graceful, witty, fashionable stupid, foolish, incapable, unfashionable

Tenacity (inclination)

plucky, heroic, curious, resolute, self-reliant

cowardly, rash, apathetic, obstinate, vexatious, lazy, servile, complacent

Social sanction

Institutional praise

Institutional blame

Veracity (probability)

honest, frank, real, genuine, credible

deceitful, fake, bogus, dishonest, deceptive

Propriety (obligation)

right, good, ethical, kind, generous, loyal, forgiving

wrong, evil, sinful, mean, cruel, greedy, arrogant, corrupt

Source: Based on Iedema, Feez and White (1994), revised and updated in Feez, Iedema and White (2008).

APPRAISAL AND JOURNALISTIC DISCOURSE

305

we chose to term the ‘judgement’ system: meanings where human behaviour and character are positively/negatively evaluated by reference to social norms. The taxonomy, as set out in the report published at the conclusion of the project, was as follows. This taxonomy has undergone some refinement and modification in later formulations (see e.g. Martin 2000; White 2002; Martin and White 2005). Of note here is that, as indicated in the above table, we divided the types of judgement into two broad categories: ‘social sanction’ versus ‘social esteem’. Our positing this division arose from the observation that, while the full array of subtypes might be observed in an opinion piece and commentary (where the author presents as the unrestrained adjudicator of human behaviour and character), there is a subset of journalistic items where the author seems only to offer assessments of social esteem. Extract 3 exemplifies texts of this latter type (explicit authorial assessment confined to values of social esteem). These were often those items which were labelled ‘analysis’ or where the author was identified as a ‘correspondent’, that is, a journalist with expertise in a given subject area or ‘round’. We observed that the attitudinal assessments advanced in these ‘analysis’ or ‘correspondent’ pieces involved evaluation of the competence, strength or skills of human agents (‘capacity’), their psychological disposition (‘tenacity’), and whether their behaviour was to be viewed positively or negative for being usual or unusual (‘normality’), but not assessments as to whether the human agent was a ‘good’ or a ‘bad’ person, or as to whether the behaviour was ‘right’ or ‘wrong’. That is to say, these ‘correspondent’ texts avoided having the author pass judgement as to the honesty of the human agent (‘veracity’) or their ethics (‘propriety’). It seemed, therefore, that the journalistic authors of commentary/opinion pieces (e.g. extract 4) are ‘authorized’ to make assessments of a type which have the highest interpersonal charge, that is, assessments where the social norms referenced are typically codified in either secular or religious laws/codes. In contrast, the assessments typically deployed in the ‘correspondent’ pieces have a slightly lower interpersonal charge, assessments of the laudability/illaudibility of behaviour rather than assessment of ethical ‘rights and wrongs’. In observing the patterns of use of attitudinal meanings in texts exemplified by the four extracts, we were thus in a position to propose what we termed the system of ‘journalistic voice’ by which we grouped texts according to whether (1) there was no or infrequent explicit authorial judgement (‘reporter voice’), (2) explicit authorial judgement was largely confined to the social esteem values of tenacity, capacity or normality (‘correspondent voice’), and (3) explicit author judgement was unconstrained, ranging across values of both social esteem and the social sanction values of veracity and propriety. This account has subsequently been elaborated (White 1998; Martin and White 2005), with additional systems of evaluative making being seen to be implicated. This line of

306

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

enquiry, in which texts are classified by reference to patterns of use of different types of evaluative meanings, has been taken up and modified for analyses of ‘voice’, ‘style’ and discursive persona in a range of discourse domains, including, to name just a few, the discourse of history (Coffin 2009), the discourse of academic English (Hood 2010), the discourse of online discussion groups (Don 2007), the discourse of real estate advertisements (Pounds 2011) and the discourse of film reviewing (White 2008). Additionally, the framework we developed initially for English has been applied to journalistic discourse in a range of languages (see e.g. Thomson and White 2008). It is of interest to note that in the materials produced by the project, in our analysis of extract 3, we classified ‘small and dowdy’ as an evaluation of ‘this week’s Republican National Convention’ as an instance of the judgement value of capacity, and the assessment of ‘this year’s convention’ as ‘interesting’ as an instance of judgement/normality. Today we would classify these as instances of the attitudinal sub-system of appreciation (Martin and White 2005: 56). The appreciation system encompasses positive/negative assessment of entities, situations and processes by reference to aesthetics and social value – contra the system of judgement which involves assessments of human behaviour. Appreciation was not recognized as the third sub-system of attitudinal meaning until a few years later, through the work of Joan Rothery on the language of the visual arts (see Martin 2000: 175). It is telling that, not having identified appreciation as a sub-system, we managed to cram an aesthetic assessment (dowdy) and an assessment of perceptual impact (interesting) into a system designed to deal with assessments of human behaviour. We managed this via a line of reasoning by which, for example, to characterize the Republican’s convention as ‘dowdy’ was to evaluate its organizers as lacking the ability to produce something more vital and engaging. Subsequently we would recognize this as a negative appreciation of the Republican convention which, in turn, might function to imply a negative assessment of the capacity of those who organized it. Thus, it was subsequently recognized that an explicit appreciation might function as a ‘token’ of judgement.

CONCLUSION In this chapter I have endeavoured to demonstrate how insights from Systemic Functional Linguistics could be applied to the challenge of providing media studies teachers with a more nuanced understanding of the widespread notion that there are two types of news journalism – the objective and the subjective. I demonstrated that it was possible to account for this belief via an understanding of the language features of the different modes of news coverage, and specifically by a linguistic account which can recognize a regime of ‘strategic impersonalization’. I demonstrated how, in tackling this challenge, the Write It

APPRAISAL AND JOURNALISTIC DISCOURSE

307

Right project provided the impetus for subsequent thoroughgoing explorations of the resources for evaluative meaning making, with outcomes not just for understandings of the communicative workings of journalistic discourse but for discourse studies more widely.

REFERENCES Coffin, C. (2009), Historical Discourse: The Language of Time, Cause and Evaluation, London: Bloomsbury. Dennis, E. E. and J. C. Merrill (1984), ‘Journalistic Objectivity’, in E. E. Dennis and J. C. Merrill (eds), Media Debates: Issues in Mass Communication, 103–18, New York: Macmillan. Don, A. (2007), ‘An Approach to the Analysis of Textual Identity through Profiles of Evaluative Disposition’, in M. Zappavigna and C. Cloran (eds), Bridging Discourses: Proceedings of the Australian Systemic Functional Linguistics Association Annual Congress, 1–17, Wollongong, NSW: ASFLA. Available online: https://www. researchgate.net/publication/237449723_An_approach_to_the_analysis_of_textual_ identity_through_profiles_of_evaluative_disposition (accessed 16 April 2021). Don, A. (2016), ‘“It is Hard to Mesh All This”: Invoking Attitude, Persona and Argument Organisation’, Functional Linguistics, 3: 9. Don, A. (2018), ‘The Value of Intertextual Associations: How GM Technologies Are Given Value through Association’, in R. Page, B. Busse and N. Nørgaard (eds), Rethinking Language, Text and Context: Interdisciplinary Research in Stylistics in Honour of Michael Toolan, 269–85, London: Routledge. Fairclough, N. (1995), Media Discourse, London: Edward Arnold. Feez, S., R. Iedema and P. R. R. White (2008), Media Literacy, Sydney: New South Wales Adult Migrant Service. Fowler, R. (1991), Language in the News: Discourse and Ideology in the Press, Abingdon, UK: Routledge. Fowler, R., R. I. V. Hodge and G. Kress et al. (1979), Language and Control, London: Routledge & K. Paul. Halliday, M. A. K. (1978), Language as a Social Semiotic: The Social Interpretation of Language and Meaning, London: Edward Arnold. Halliday, M. A. K. and R. Hasan (1987), Language, Context, and Text: Aspects of Language in a Social-Semiotic Perspective, Oxford: Oxford University Press. Hao, J. and S. L. Humphrey (2012), ‘Burnishing and Tarnishing in Academic Literacy’, in J. Knox (ed.), To Boldly Proceed: Papers from the 39th International Systemic Functional Congress, 15–21, Sydney: International Systemic Functional Linguistics Association. Hood, S. (2010), Appraising Research: Evaluation in Academic Writing, Houndmills, UK: Palgrave Macmillan. Hood, S. and J. R. Martin (2007), ‘Invoking Attitude: The Play of Graduation in Appraising Discourse’, in R. Hasan, C. M. I. M. Matthiessen and J. Webster (eds), Continuing Discourse on Language: A Functional Perspective, Volume 2, 739–64, London: Equinox. Iedema, R., S. Feez and P. R. R. White (1994), Media Literacy, Sydney: Department of Education.

308

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Martin, J. R. (1992), English Text: System and Structure, Philadelphia, PA: John Benjamins. Martin, J. R. (2000), ‘Beyond Exchange: APPRAISAL Systems in English’, in S. Hunston and G. Thompson (eds), Evaluation in Text: Authorial Stance and the Construction of Discourse, 142–75, Oxford: Oxford University Press. Martin, J. R. and P. R. R. White (2005), The Language of Evaluation: Appraisal in English, London: Palgrave Macmillan. Paula, J. (2011), ‘Appraising through Someone Else’s Words: The Evaluative Power of Quotations in News Reports’, Discourse & Society, 22 (6): 766–80. Pounds, G. (2011), ‘“This Property Offers Much Character and Charm”: Evaluation in the Discourse of Online Property Advertising’, Text & Talk, 31 (2): 195–220. Poynton, C. (1989), Language and Gender: Making the Difference, Oxford: Oxford University Press. Schudson, M. (1981), Discovering the News: A Social History of American Newspapers, New York: Basic Books. Thomson, E. and P. R. R. White (2008), Communicating Conflict: Multilingual Case Studies of the News Media, London: Continuum. van Dijk, T. A. (1988), News as Discourse, Hillsdale, NJ: Lawrence Erlbaum Associates. White, P. R. R. (1998), ‘Telling Media Tales: The News Story as Rhetoric’, PhD thesis, Department of Linguistics, University of Sydney, Sydney. White, P. R. R. (2000), ‘Dialogue and Inter-Subjectivity: Reinterpreting the Semantics of Modality and Hedging’, in M. Coulthard, M. Cotterill and F. Rock (eds), Working With Dialog, 67–80, Berlin: Max Niemeyer Verlag. White, P. R. R. (2002), ‘Appraisal – The Language of Evaluation and Stance’, in J. Verschueren, J.-O. Östman and J. Blommaert (eds), The Handbook of Pragmatics, 1–23, Amsterdam: John Benjamins. White, P. R. R. (2003), ‘Beyond Modality and Hedging: A Dialogic View of the Language of Intersubjective Stance’, Text, 23: 259–84. White, P. R. R. (2006), ‘Evaluative Semantics and Ideological Positioning in Journalistic Discourse’, in I. Lassen (ed.), Image and Ideology in the Mass Media, 45–73, Amsterdam: John Benjamins. White, P. R. R. (2008), ‘Praising and Blaming, Applauding and Disparaging – Solidarity, Audience Positioning, and the Linguistics of Evaluative Disposition’, in Antos, G. & Ventola, E. (eds), Handbook of Interpersonal Communication, 542–67, Berlin & New York: De Gruyter Mouton. White, P. R. R. (2010), ‘Taking Bakhtin Seriously: Dialogic Effects in Written, Mass Communicative Discourse’, Japanese Journal of Pragmatics, 12: 37–53. White, P. R. R. (2012), ‘Exploring the Axiological Workings of “reporter voice” News Stories – Attribution and Attitudinal Positioning’, Discourse, Context & Media, Elsevier 1 (2–3): 57–67. White, P. R. R. (2016), ‘Evaluative Contents in Verbal Communication’, in A. Rocci and L. de Saussure (eds), Verbal Communication, Vol 3, Handbooks of Communication Sciences, 77–96, Berlin: De Gruyter Mouton. White, P. R. R. (2020), ‘The Putative Reader in Mass Media Persuasion – Stance, Argumentation and Ideology’, Discourse & Communication, 14: 404–23. White, P. R. R. (2021), ‘Textual Anticipation and the Putative Reader in Persuasive Discourse’, Journal of Foreign Languages, 44 (1): 120.

CHAPTER SEVENTEEN

Binding and Bonding A Retrospective and Prospective Gaze MAREE STENGLIN

INTRODUCTION In 2004, I developed a metafunctionally diversified grammar of threedimensional space (Stenglin 2004). Inspired by Halliday’s (1978, 2004) account of metafunctions, it explored all three communicative functions of space: the interpersonal, the ideational and the textual. This chapter focuses on the interpersonal metafunction only; and how it has been theorized, applied and extended in my work over the past twenty years – using the social semiotic tools of binding and bonding. Binding is concerned with the interpersonal relationship between a space and its occupant(s), with a particular focus on how the organization of space can help people feel secure or insecure. Bonding is concerned with ways of aligning users into communities of belonging. It explores resources for building solidarity, inclusiveness and affiliation. The two tools work together since feeling safe in a space (binding) makes possible related feelings – such as developing a sense of emotional engagement with the institution and feeling secure enough to explore ideas that may be exciting and challenging with like-minded others (bonding). Several discourses, public events and public displays galvanized the theorization of how interpersonal meanings are made in three-dimensional space. Extant discourses included museology and multimodality, Systemic Functional Linguistic Theory, especially appraisal (Martin 1997, 2000; White

310

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

1997, 1998; Martin and White 2005) and lexical metaphors, together with public events including the Sydney 2000 Olympic Torch Relay. Field work was also pivotal to the investigation; it involved the analysis of public displays, not only in museums but also in the Collins Bookstore, Sydney and the Cotton Store, Darling Harbour. All of these informed the theorization of space, and each contributed to the emergence of different aspects of binding and bonding. Once the tools were theorized, binding and bonding were applied and refined during museum field trips in Australia, the United States, Canada and Europe. They were also applied to research into the phylogenesis of domestic architecture, which demonstrated that the tools had wider applicability than museum exhibition spaces only (Stenglin 2011). Since then, the application of binding and bonding has been extended to include other spatial types such as universities (Ravelli and Stenglin 2008; Stenglin 2012b), birth spaces (Stenglin and Foureur 2013), restaurants (Stenglin 2008b) and even virtual spaces (Stenglin and Djonov 2010). Word constraints, however, mean this chapter will focus on museums only. To this end, the next section will begin by addressing the initial problems motivating the research and the three key challenges identified. These challenges frame the rest of the chapter as the focus is on exploring the theories that helped develop solutions; the key was appraisal analysis – an invaluable tool for probing and opening up interpersonal meanings in three-dimensional space (Martin 1997, 2000; White 1997, 1998; Martin and White 2005).

THE ISSUE, PROBLEM AND CHALLENGES The need for binding and bonding was driven by my involvement in developing the Indigenous Australians exhibition at the Australian Museum during the 1990s. There I worked on the development of the exhibition with a multidisciplinary team, including anthropologists, designers, conservators, marketers and public relations staff. Despite a shared vision for the exhibition, our greatest obstacle was the fact our multidisciplinary team did not have a shared metalanguage for discussing design. This meant our ability to participate in the meaning-making processes of exhibition development was severely curtailed. In particular, a shared metalanguage would have allowed each of us to make more informed contributions to the conceptualization, interpretation and design stages of the exhibition. A shared metalanguage would also have helped us deconstruct the exhibits we visited for inspiration before our exhibition planning had begun, and articulate much more clearly the design elements that appealed to us alongside explanations of why we felt they were powerful.

BINDING AND BONDING

311

Motivated by this experience, I accepted an APA(I) scholarship. The specific research questions being pursued, though, were situated within the broader context of Western museums at the end of the twentieth century, and the enormous economic pressure on them to expand their visitor base or perish (Bentley 2009). To become inclusive, museums in the 1990s were striving to understand the barriers to participation and much valuable research was done in this area. In particular, seminal research into the socio-cultural factors impacting on European art museum attendance done by Bourdieu and Darbel (1991) raised awareness of the strong correlation between education, social class and museum visitation. Merriman (1989) and Hood (1995) reported similar findings in England and the United States, respectively, while Bennett (1994) found social participation in Australian museums also strongly reflected class and educational differences. Most recently, ABS figures have foregrounded the importance of socialization, since those who visited a museum or art gallery as a child are 200 per cent more likely to continue doing so later in life (Gan 2019). Given these findings, making museums available to a diverse audience base was, and continues to be, an important but incredibly difficult task. This is especially true if the museum is tapping into new audiences. One of the first steps in expanding a visitor base thus involves identifying a potentially new community. Much research has been done in this area by marketers, museum visitor studies units, sociologists and global advertising companies. Saatchi and Saatchi (2000: 13), for instance, surveyed the Australian population and found that 35 per cent of Australians thought museums and the arts were ‘OK’ but not relevant. Establishing the relevance to people’s daily lives appears to be crucial to achieving this goal. Second, careful thought needs to be given to how to help newcomers, who may be unfamiliar with museum spaces and social practices, feel safe and secure as they move through the building. This is a critical challenge because visitors who do not feel comfortable tend to leave the building. Once visitors begin to feel at ease in the exhibition spaces, museums need to consider ways of building semiotic bridges that serve two functions: first, tap into the prior knowledge and understandings visitors bring with them; second, open up new knowledge for their consideration, learning and enjoyment. The challenge of exploring new knowledge needs to be negotiated in ways that the visitor will find absorbing, inspiring and challenging. The final challenge museums face is how to create a lasting disposition of regular visitation. Rather than turning to economic rationalism to justify this need, a much more important argument is offered by museologist Elaine Heumann Gurian, who says, ‘museums should welcome all because they house the collective memory of all’ (2006: 2). Table 17.1 summarizes all the challenges facing museums.

312

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 17.1  Challenges Facing Museums of the Twenty-First Century Problem (divided into key phases)

Specific challenges to overcome

Phase 1: Reach new communities

Encourage diverse audiences into the museum Establish the relevance of the museum to different types of communities

Phase 2: Help new communities engage with museum exhibitions and programmes

Keep people feeling at home in the museum’s spaces after they have crossed the threshold into the building Engage and inspire visitors

Phase 3: Create a lasting disposition to museum visiting

Entice visitors back to the museum again and again

After a brief prelude, each phase of the problem and its associated challenges will be explored and, as mentioned earlier, these three phases provide the scaffold for this paper.

PRELUDE: THE GENESIS OF THE POSTMODERN MUSEUM Using Martin’s (1997) model of social context, different configurations of genre and register that typify three seminal moments in the museum’s history were probed. This involved engaging with the discourses of museology to chart the evolution of museums and provided invaluable insights into the trials facing the museum of today – the ‘hybrid’ museum. They are hybrid because today’s museum has two clear social purposes: education and entertainment. With funding linked to visitor numbers, the task of reconciling entertainment with education, postcolonial scholarship and ethically inspired social change created some conflicting challenges that are not simple to resolve (see Stenglin 2004, 2009b for a fuller account). Charting points of continuity and change in this way was invaluable as it foregrounded curiosity – the lasting characteristic hybrid museums of today share with their predecessors, the Renaissance cabinets. Curiosity is an integral part of museums’ attraction as it is the portal to engagement with the objects they house and display, the meanings they make, and the pathways for interaction they open up for visitors. Since the turn of the twenty-first century, moreover, museums have been so focused on serving community needs that their audiencecentred mission is now driving the current global redefinition of museums to reflect this new dynamic (ICOM 2021). With these thoughts in mind, I turn to the first phase of the problem I needed to tackle: how to reach new communities.

BINDING AND BONDING

313

FIRST PHASE: REACHING NEW COMMUNITIES In the research, two tools were developed to help the problem of reaching new communities. Both are aspects of bonding: hybridization and bonding icons. Hybridization The first tool developed to address the challenge of reaching new audiences was hybridization. Ravelli’s (2000) analysis of the Olympic store documented the inclusion of Olympic medals, swimming lanes and podiums in the retail store. These functioned to evoke values – in this instance, the values of Olympism. Building on Ravelli’s (2000) observations, I explored how such values align visitors into a community. The starting points for this exploration were the concepts of values and ‘value transfer’. To examine them, I undertook extensive field work in two spaces: the Collins Bookstore in Broadway, Sydney and the Cotton Store in Darling Harbour, Sydney. Each of these contained simulations, but even more importantly, each site was designed to serve multiple functions. Collins Bookstore, for instance, was not just a retail space – it was a coffee shop, a meeting space for book clubs, a library, a collector’s corner and so on. Spaces with multiple functions like these open up multiple choices for interaction, for example, browsing, buying, having a coffee with a friend, reading while eating or drinking and so on. These observations highlighted the first interpersonal resource for developing the concept of bonding: hybridization. Hybridization involves designing a space with many functions. A wonderful example of hybridization is the Museum of Old and New Art (MONA) in Hobart, which opened in 2011. It offers a museum, a library, a cinema, a hotel, a winery, a brewery, an outdoor performance space with restaurants, cafes, bars and shops; in addition, it hosts two annual festivals. In short, it offers something for everyone. At one level, such hybridization provides multiple fields of the culture for people to engage with. For example, a person does not have to visit the museum and view art to be part of MONA’s community. Instead, they could go on a wine tour, eat in a restaurant, attend an outdoor concert or do research in the library. But MONA does have one interesting and overarching feature: art is integrated into each space of the museum precinct, including the ferries that carry visitors there. In this way, all MONA visitors are immersed in art. This establishes some familiarity with art, especially sculpture, which has the potential to pique interest and curiosity. At another level, hybridization recontextualizes the values of one field or activity onto another. For instance, weekend outdoor concerts at MONA, such as the performances by the Southern Gospel Choir in August 2019 (Mona 2021), mean values associated with popular culture (in this instance the strong freedom message of Mandela) are being evoked and transferred

314

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

to this ‘eccentric’ art museum. Similarly, the prestige associated with the subversiveness of this specific museum is transferred onto the concert-going audience. To ground hybridization theoretically, lexical metaphors provided a way of understanding what was happening because lexical metaphors (imageladen language) function in much the same way (see Martin 2001, 2004; Stenglin 2004 for more discussion of this). In terms of the impact hybridization can have on potential audiences, especially with regard to museums, journalist Riley Fitzgerald (2016) writes of MONA: The opening of eccentric millionaire David Walsh’s ‘subversive adult Disneyland’ the Museum of Old and New Art (MONA) in 2011 accompanied by the museum’s biannual MONA FOMA and DARK MOFO festivals, has not only garnered the attention of creative minds domestically and abroad but also opened the city up to a new wave of cultural tourism. It’s beginning to rival the traditional hubs of Melbourne and Sydney as a cultural tourism destination. Although the subversiveness of MONA is an intriguing aspect of this museum’s appeal, MONA is currently ranked on Tasmania’s official website (Discover Tasmania n.d.) as their number two attraction. It thus appears that museums wanting to expand their reach have much to think about in relation to hybridization. Not only does the provision of multiple options for engaging with a diverse range of cultural activities and programmes help to lure non-museum goers, the act of constantly offering them something new, enticing and intriguing them is also important. This combination appears to have implications for creating a more lasting disposition to visiting a cultural precinct and offers multiple ways to serve new and existing communities. In the case of MONA, indirect exposure to art in these ways, over time, may also provide the impetus for further engagement. Bonding icons The challenge of reaching new audiences also has much to gain from the second tool developed during this research: bonding icons. The galvanizing stimulus for research into bonding icons was the Olympic Torch Relay, held in Australia in 2000. The immense communing power of the Sydney Olympic Games had its nucleus in the Olympic flame. This flame was transported around the entire continent before the games began and it had an enormous impact. It broke down boundaries between young and old, elite athletes and non-athletes, urban and rural dwellers, land, sea and air. It also transcended political, social,

BINDING AND BONDING

315

economic and cultural divisions. It was actually the first bonding icon identified in this research. The powerful way it transcended divisive boundaries meant that its relevance to my research into museums could not be ignored. Cynics, for instance, were expecting the flame to be greeted in Sydney as an ‘anticon’, especially in the ‘big city’ – with protests linked to its Nazi origins as well as the corruption of the International Olympic Committee and the elitism of the games. The reality was very different. As the city’s main broadsheet newspaper commented: ‘Despite its fascist origins, the concluding stages of the torch relay are bringing together Australians in a wonderfully democratic way. Australians have forgotten Carl Diem but will remember the light the torch brought to their lives’ (SMH, 15 September 2000: 5). Comparable appreciations appeared elsewhere: ‘The cynical big city [Sydney] has embraced the torch with enthusiasm and tears’ (Huxley 2000). Not only did the torch become ‘a light’, but it created social harmony and fostered inclusiveness by connecting Australians to one another through the shared activity of gathering and witnessing its passage. The overall impact was effectively captured in the words of one suburban resident: I tried to be cynical! I tried so hard! But last night I stood with my neighbours as the torch passed my address in Kirrawee. What a wonderful night! For all the corruption, elitism and excesses of the IOC, you have to appreciate an event that draws the community on to the streets, that engenders a carnival atmosphere and fills people with joy. (Gillman 2000) In terms of this research, it was clear that the torch was able to rally disparate groups of people around shared communal ideals. Theoretically, appraisal (Martin 1997, 2000; White 1997, 1998; Martin and White 2005) helped illuminate exactly how it came to have this effect. The appraisal analysis of countless newspaper reports and letters to the editor, like the one below, revealed an important pattern: the torch was moving people through all three attitudinal systems (affect, judgement, appreciation): What a wonderful night [+ appreciation: valuation]! For all the corruption, elitism and excesses [judgement: ethics] of the IOC, you have to appreciate [+appreciation: valuation] an event that draws the community on to the streets [+appreciation: valuation], that engenders a carnival atmosphere [+ appreciation: reaction: quality] and fills people with joy [+affect: happiness: cheer]. Anchored in communal responses to the 2000 Olympic flame, bonding icons emerged. They are evocative symbols of belonging that can rally a specific

316

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

community or communities of people in favour of, or against, the values they represent. They do this by crystallizing strong interpersonal attitudes to ideational meanings. The attitudes they evoke can be either positive (bonding icons) or negative (anticons). Given the strong evidence during the 2000 Torch Relay and the games proper of the power of bonding icons to unite disparate groups of people into oneness, they became another tool to help museums lure diverse communities into their spaces. They often have two fields: one concrete (e.g. a flame) and one abstract (e.g. the Olympic spirit – friendship, fair play, solidarity). For further discussion of bonding icons and a fuller analysis of the Olympic flame see Stenglin (2008a: 60–6); as well as Stenglin (2012a) for an analysis of the ‘Croata’ tie. Returning to museums, bonding icons can help explain, for instance, why an exhibition called Speedway was able to draw large numbers of young, working-class males into the Casula Powerhouse Museum in Sydney. So keen was this group of young men to see the exhibition, they were knocking on the doors long before opening hours (Peach 2001). Why? First and foremost, this exhibition tapped into the human obsession with speed. In a literal sense, a speedway is a special track for the fast racing of cars or motorbikes without brakes. This involves ‘bravery’ (+ judgement: tenacity). Interpersonally, speed also evokes joy (+ affect), excitement (+ appreciation), thrills and adventure (+ appreciation), alongside positive appreciation of vehicles built with technological advances for high-speed performance such as turbo chargers, semi-automatic transmission and so forth (Ashish 2020). By tapping into this powerful bonding icon, the Casula Powerhouse Museum not only drew large numbers of people through its doors but rallied a group of people who, for the most part, feel that museums are irrelevant (Bourdieu and Darbel 1991). It is this rallying potential of bonding icons that makes them an invaluable resource – for they do, indeed, have the potential to counter the single key obstacle to museum participation, namely irrelevance. But their deployment requires careful thought and selection. As we have just seen, the interpersonal nature of the problems being addressed by this research meant that appraisal theory offered the best theoretical foothold. Not only was it being used to analyse evaluation in language; it was also providing a bridge from the semiosis of language to the semiosis of space. In a similar vein, other semioticians began applying it to semiotics modes such as visual images (Martin 2001; Caple 2006; Economou 2006; Painter et al. 2013). In fact, appraisal theory would continue to make a substantial contribution to solving the problems identified in this research, especially in the next stage where it offered an invaluable axiological tool for understanding bonding in museums.

BINDING AND BONDING

317

SECOND PHASE: HELPING NEW COMMUNITIES ENGAGE WITH THE MUSEUM, ITS EXHIBITIONS AND PROGRAMMES This section will explore how bonding and binding can help both new and established communities engage with the museum, its exhibitions and its programmes. Bonding is a multidimensional resource concerned with how people connect and develop a sense of belonging, and two of its tools, hybridization and bonding icons, have already been discussed in relation to the first challenge. Bonding also involves epistemic design (Stenglin 2017) and alignment – that is, the alignment of people into communities of affiliation through shared attitudes and experiences (Stenglin 2008a, 2012a, 2012b). Complementing this focus on social interaction and community building is binding, a tool concerned with the affectual disposition between an occupant and a space. As walking into a museum involves entering and moving through a series of institutional spaces, let us begin with binding. Binding The challenge of helping new audiences feel comfortable and secure inside the institutional spaces of the museum has much to gain from binding. The power of exhibitions to evoke strong emotions like fear and insecurity was first identified in White’s (1994) analysis of the Shark exhibition. Strong feelings of insecurity are important as they may motivate some people to leave, some to hide and others to lash out. Understanding how such feelings can be evoked by the design of space is vital for museums. Accordingly, emotion in three-dimensional space was investigated. The starting point was O’Toole’s framework for analysing architecture (1994: 96). But as it did not explore feelings, the next port of call was appraisal, especially affect: in/security. From this point of departure, the concept of binding evolved. This section provides a brief introduction to the work on binding (for further discussion see Stenglin 2002, 2007, 2008b; Stenglin 2004: chapters 3, 4). The interpersonal experience of three-dimensional space is tied to how firmly a space encloses a person. Spaces that are too restricted create a smothering relationship with users (too bound) while spaces that are too unrestricted create relationships of vulnerability with users (too unbound). These extreme choices evoke strong feelings of insecurity such as claustrophobia or agoraphobia. In addition, there are two choices for spatial security: bound and unbound spaces. Firm levels of enclosure, especially by walls and ceilings, can make a person feel comfortable and secure (bound). When such enclosures are loosened, more space opens up around the occupant (unbound) and they feel free and unencumbered but still secure.

318

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Theoretically, binding falls within the range of interpersonal meaning related to the expression of attitude, in particular, affect: in/security. Since feelings are gradable, binding is a topological resource. This means the four choices for binding were mapped on a continuum. During the 2000 field trip to American and Canadian museums, it was found that some spaces felt more or less bound. Accordingly, gradation was built into the scale. As the two extreme points, too bound and too unbound, cannot be intensified further, gradation can only be applied to the median choices. Spaces could thus be strongly, moderately or minimally bound or minimally, moderately or strongly unbound. In addition, since buildings unfold in both space and time as users move through them, binding needed a dynamic (logogenetic) dimension. This was first explored in relation to the Law Court in Melbourne (Stenglin 2004: 189). This analysis demonstrated the way small shifts in binding can maintain a relationship of security with users as they walk through an unfolding sequence of spaces (see Stenglin 2004, 2008b). The logogenetic dimension of binding was later refined as the notion of ‘binding contours’ (McMurtrie 2012). From the museum’s perspective, if visitors feel secure, comfortable and welcome in the institution’s spaces, they are more likely to stay, explore, engage, muse and interact. Feeling secure also increases the likelihood of future return visits, which is vital to museums from a long-term perspective. Moreover, the patterning of bound and unbound spaces can help address the problem of ‘museum fatigue’ (physical and mental tiredness) first identified by Gilman (1916). Importantly, research into binding does not mean that the extreme choices of the binding scale should always be avoided. Recently, for instance, the architects of the Islamic Museum of Australia, desypher, deliberately created an entrance that was too bound in order to simultaneously encourage immersion in a new culture as visitors open themselves to exploring and engaging with new perspectives. Their website’s interpretation of binding is presented below (desypher n.d.); the binding analysis [too bound, bound, unbound] is my own. The museum is a place that beckons the visitor to enter but it is not without its implicit challenges. The entry itself is but negative space between the corten veil and the stone prism, it is also dark and upon entering the path it is unclear [too bound]. A ninety degree turn emphasises the need to reorientate or refocus, as if to say ‘look at what is to come from a different viewpoint’, before subsequently leading one to a naturally lit space that is the entry foyer [bound]. The theme remains throughout museum: darkened thresholds [bound] leading to lit spaces [unbound], bridges actual and metaphoric, lines of sight that do not correspond with pathways and doorways that are spaces between creased walls. These are all designed to maintain intrigue through gradual unravelling, even the way in is not the way out.

BINDING AND BONDING

319

Not only does the entrance space encourage visitors to refocus and adopt new ways of seeing the world, the patterning of too bound, bound and unbound spaces throughout the museum functions to evoke enlightenment: the central metaphor in the Islamic Museum of Australia. Bonding: Attitudinal alignment For museums, the challenge of how to facilitate bonding through deep emotional engagement and ongoing participation, not just a one-off visit, is a crucial issue. One important lead in this research came from Hodge and D’Souza’s (1979) exploration of simulations. They argued that simulations can evoke values, positive or negative, and thus impact on the visitor experience. These claims were further supported by museum focus group research (Kelly 1998) which showed that, in addition to values, simulations evoke strong emotions and these emotions are remembered by visitors long after their visit – and can even impact on people’s behaviour. In fact, Kelly (1998) documents that some visitors were so moved by an exhibition they participated in social justice campaigns following their visit. One bonding tool for exploring deeper engagement with displays such as simulations was field-appraisal alignment, especially in relation to attitude. Field refers to ‘what’s going on’ (Martin 1992) in an exhibit and involves consideration of both activity and the item(s) involved in them. Interpersonally, field has the potential to invoke appraisal, especially the feelings and evaluations at the heart of attitude. Such feelings are vital to affiliation, as Martin (2001) clearly demonstrated. In particular, he showed how shared affect aligns readers around empathy, shared judgement aligns them around shared principles while shared appreciation aligns them around shared tastes. This coarticulation of field/attitude was extended to the exploration of bonding in museums as it allowed investigation of the way visitors are invited into different communities. Over the past two decades, two key types of alignment have emerged from social semiotic analyses of museum exhibitions: alignment that facilitates ethically inspired attitudinal change (Martin and Stenglin 2007; Stenglin 2013) and alignment that fosters a deeper appreciation of an individual’s or a society’s social, scientific, technological and artistic accomplishments (Stenglin 2007, 2009a, 2017). Alignment oriented to ethically motivated social change tends to be oriented to themes like reconciliation, environmental conservation and the combating of violence, racism and bigotry. This type of alignment was first explored in relation to the Indigenous Australians exhibition (Stenglin 2004, 2013). Crucially, access to focus group research (Kelly 1998) made it possible to chart the emotional journey of visitors through key simulations and track how communal alignment was negotiated, noting that it is ultimately the visitor’s choice to accept or reject the institution’s discursive positioning (Stenglin 2004).

320

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

In brief, in the Indigenous Australians exhibition most simulations evoked negative judgements, especially impropriety, of the behaviours of white Australians towards Indigenous peoples (e.g. attempted cultural genocide, racial discrimination, the negligence driving deaths in custody). Alongside these judgements sit negative feelings of affect (the unhappiness, including tears, experienced by some visitors at the plight Aboriginal people have suffered) together with strong positive appreciation of the survival of Indigenous people and their cultural heritage (Stenglin 2013). In the exhibition, such appraisals were evoked prosodically and cumulatively as visitors moved through a variety of simulations (e.g. a cave, a church, the freedom bus, a cell, the Aboriginal tent embassy). Some focus group participants reported feeling so moved by the experience they turned to social justice activism after the museum visit (Kelly 1998). Although it is impossible to know if this experience prompted return visits, it did facilitate deep levels of emotional engagement. Moreover, it was not just the simulations that played a pivotal role in facilitating the attitudinal engagement and change. The strongest emotional response reported was in relation to a black-and-white photograph of Aboriginal men in chains (Kelly 1998); many visitors wept when they saw it. This seems to indicate that in general objects – photographs, artefacts, specimens, artworks – can be pivotal to visitor alignment, not just simulations. To understand more about how objects can foster a deeper appreciation of a society’s artistic and technological accomplishments and belief systems, I turn to one of the seminal stories of my museum career: taking an extremely reluctant visitor on a guided tour of the Pharaohs exhibition at the Australian Museum. To make the tour as enjoyable as possible, I offered to showcase my ten favourite objects from the exhibition. The first was a strange curved wooden object on a stand; the second was an opaque oval shaped piece of bronze with a handle. We stopped in front of the first artefact and I asked if the visitor knew what it was. She shook her head. I asked her to describe her pillow. Her surprise was palpable, her curiosity stimulated. She inundated me with a tsunami of questions that continued as we engaged with the next artefact, the bronze piece. Once again, I asked her to first guess what it was; and then describe her mirror. At the end of the tour of my ten favourite objects, the visitor purchased an audio guide and spent another hour in the space musing on her own but leaving me desperately wanting to understand how and why this change of heart had taken place. In grappling to understand how these artefacts had fascinated and intrigued the visitor, I began to explore the grammatical concepts of form (shape) and function (purpose). The first object, the wooden headrest, appears to be ‘unfamiliar,’ but it is not – it functions in exactly the same way as a pillow. Once this commonality is established, it opens up a space for a multitude of questions. Similarly, with the second object. It is not immediately obvious how

BINDING AND BONDING

321

this bronze piece can function as a mirror. In fact, the function remains obscure until the capacity of bronze to be polished and offer reflection is pointed out. By arousing her curiosity, the tension between form and function was the portal that opened the door to engagement for this reluctant visitor. The next four objects on the tour were familiar in both function and form (comb, ruler, sandals, earrings) so the discussion was more oriented to appreciation than curiosity. Subsequently, I chose objects with familiar forms (canopic jars and a sarcophagus) but unfamiliar functions. This provided a springboard for exploring new knowledge (applied, theoretical and reflexive) about ancient Egyptian cosmology (see Stenglin 2017 for an analysis of all the objects used in this tour). By stimulating the visitor’s curiosity in this way, the object choices were indeed the portal to her engagement; one that opened up an alignment pathway which helped this visitor develop a deeper appreciation of life and cosmology in a fascinating ancient civilization. To understand even more about both attitudinal social change and deeper cultural appreciation, we need an intermodal theory of semiosis – one that can account for the meanings made in different semiotic modes (objects, the written language of labels and text panels, the spoken language of the guide, maps, diagrams, films) and how these modalities co-articulate to align the visitor as s/he moves through an exhibition. Importantly, most museum visits are self-guided. This means most people do not have access to the informed object selections I was able to make and may not avail themselves of the information I was able to provide. There is also a whole dimension to bonding in pedagogical institutions like museums that we have not touched on yet – one that involves learning theories. Two, in particular, have been invaluable in my work: Bernstein’s (1975) notions of classification and framing and Macken-Horarik’s (1996) domains of knowledge. But it is beyond the scope of this chapter to explore those.

THE FINAL PHASE: CREATING A LASTING DISPOSITION TO VISITATION In addressing the final challenge, we turn to the tandem tools: binding and bonding. Perhaps the simplest explanation of how they can create a lasting disposition to museum visiting is found in the reflection of one visitor to Casula Powerhouse Museum, Sydney (binding and bonding comments added are mine): I was a youth member at PCYC in Liverpool, just making art and shit, and my friend was like, ‘hey come to this meeting’ so I thought, ‘why not?’ From there, I just wanted to stay here [binding] and make art [bonding], and I became part of the family [bonding], now I’m running it. (Phu, member of Casula Powerhouse Arts Centre Youth)

322

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

As this reflection indicates, binding can and does open the door for bonding – for joining ‘the family’ – becoming part of a community, part of the institution by interacting with others and sharing in the core activities, knowledge and understandings being offered. The impact of this can be such that when one walks away, one feels inspired, uplifted, stimulated. When complemented with hybridization, there is indeed much to lure visitors back. Museums thus have enormous potential to fascinate and enthral audiences of all ages as well as to build metaphorical bridges and heal longstanding divisions. They just need to find the right ways in. I do hope this toolkit can help them to unlock some of those portals.

REFERENCES Ashish (2020), ‘Why Do We Feel so Thrilled by Speed?’, Science ABC. Available online: https://www.scienceabc.com/pure-sciences/why-do-we-feel-so-thrilled-byspeed.html (accessed 1 February 2021). Bennett, T. (1994), The Reluctant Museum Visitor: The Study of Non-goers to History Museums and Art Galleries with International Comparisons, Redfern, NSW: Australia Council. Bentley, P. (2009), ‘Putting a Value on Museums’, Museum Matters, 18 (2): 1–4. Bernstein, B. (1975), Towards a Theory of Educational Transmissions: Class, Codes and Control Volume 3, 2nd edn, London: Routledge and Keegan. Bourdieu, P. and A. Darbel (1991), The Love of Art: European Art Museums and Their Public, Stanford, CA: Stanford University Press. Caple, H. (2006), ‘Nuclearity in the News Story – The Genesis of Image Nuclear News Stories’, paper presented at the Australian & New Zealand Communication Association International Conference, Adelaide, 5–7 July. desypher (n.d.), Islamic Museum of Australia. Available online: https://desypher.com. au/index.php/projects/islamic-museum/ (accessed 1 February 2021). Discover Tasmania (n.d.), Top Ten Attractions. Available online: www. discovertasmania.com.au/about/top-ten (accessed 1 February 2021). Economou, D. (2006), ‘The Big Picture’, in I. Lassen, J. Strunck and T. Vestergard (eds), Mediating Ideology through Image and Text, 112–234, Amsterdam: Benjamins. Fitzgerald, R. (2016), ‘MONA and the State of Hobart’s Underground Music Scene’, Howl and Echoes, 29 October. Available online: https://howlandechoes. com/2016/10/mona-hobart-music-scene/ (accessed 1 February 2021). Gan, L. (2019), ‘Museums: They’re Good for Your Health’, Pursuit, 28 February. Available online: https://pursuit.unimelb.edu.au/articles/museums-they-re-good-foryour-health (accessed 1 February 2021). Gillman, A. (2000), ‘Letter to the Editor’, Sydney Morning Herald, 15 September: 11. Gilman, B. I. (1916), ‘Museum Fatigue’, Scientific Monthly, 2 (1): 62–74. Halliday, M. A. K. (1978), Language as Social Semiotic, London: Arnold. Halliday, M. A. K. and C. M. I. M. Matthiessen (2004), An Introduction to Functional Grammar, 3rd edn, rev. London: Arnold. Heumann Gurian, E. (2006), Civilising the Museum: The Collected Writings of Elaine Heumann Gurian, London: Routledge.

BINDING AND BONDING

323

Hodge, R. and W. D’Souza (1979), ‘The Museum as a Communicator: A Semiotic Analysis of the Western Australian Museum Aboriginal Gallery, Perth’, Museum International, 31 (4): 251–67. Hood, M. (1995), ‘Audience Research Tells Us Why Visitors Come to Museums – And Why They Don’t’, paper presented at the Evaluation and Visitor Research in Museums Towards 2000 Conference, Powerhouse Museum, Sydney, 16–19 March. Huxley, J. (2000), ‘Sydney Embarks on Its 15 Minutes of Flame’, Sydney Morning Herald, 15 September: 5. International Council of Museums (ICOM) (2021), International Council of Museums. Available online: http://icom.org (accessed 1 February 2021). Kelly, L. (1998), Indigenous Australians Focus Group Report, Sydney: Centre for Evaluation and Audience Research, Australian Museum. Macken-Horarik, J. (1996), ‘Construing the Invisible: Specialised Literacy Practices in Junior Secondary English’, PhD thesis, University of Sydney, Sydney. Martin, J. R. (1992), English Text: System and Structure, Amsterdam: John Benjamins. Martin, J. R. (1997), ‘Analysing Genre: Functional Parameters’, in F. Christie and J. R. Martin (eds), Genre and Institutions: Social Processes in the Workplace and School, 3–39, London: Cassell. Martin, J. R. (2000), ‘Beyond Exchange: appraisal Systems in English’, in S. Hunston and G. Thompson (eds), Evaluation in Text, 142–75, Oxford: Oxford University Press. Martin, J. R. (2001), ‘Fair Trade: Negotiating Meaning in Multimodal Texts’, in P. Coppock (ed.), The Semiotics of Writing: Transdisciplinary Perspectives on the Technology of Writing, 311–38, Turnhout, Belgium: Brepols. Martin, J. R. (2004), ‘Mourning – How We Get Aligned’, Discourse & Society, 15 (2/3): 321–44. Martin, J. R. and M. Stenglin (2007), ‘Materialising Reconciliation: Negotiating Difference in a Post-colonial Exhibition’, in T. Royce and W. Bowcher (eds), New Directions in the Analysis of Multimodal Discourse, 215–38, Mahwah, NJ: Lawrence Erlbaum Associates. Martin, J. R. and P. R. R. White (2005), The Language of Evaluation: Appraisal in English, London: Palgrave. McMurtrie, R. J. (2012), ‘Feeling Space Dynamically: Variable Interpersonal Meanings in High-Rise Apartment Complexes’, Visual Communication, 11 (4): 511–34. Merriman, N. (1989), ‘Museum Visiting as Cultural Phenomenon’, in P. Vergo (ed.), The New Museology, 149–71, London: Reaktion Books. MONA (2021), Southern Gospel Choir: Love Theory. Available online: https://mona. net.au/stuff-to-do/southern-gospel-choir (accessed 1 February 2021). O’Toole, M. (1994), The Language of Displayed Art, London: Leicester University Press. Painter, C., J. R. M. Martin and L. Unsworth (2013), Reading Visual Images: Image Analysis of Children’s Picture Books, London: Equinox. Peach, R. (2001), Speedway Racing, Liverpool, NSW: Liverpool Regional Museum. Ravelli, L. (2000), ‘Beyond Shopping: Constructing the Sydney Olympics in ThreeDimensional Text’, Text, 20 (4): 489–515. Ravelli, L. R. and M. Stenglin (2008), ‘Feeling Space: Interpersonal Communication and Spatial Semiotics’, in E. Ventola and G. Antos (eds), Handbook of Applied Linguistics, Volume 2, Interpersonal Communication, 355–93, Berlin: De Gruyter Mouton.

324

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Saatchi & Saatchi (2000), Australians and the Arts: What Do the Arts Mean to Australians?, Surry Hills, NSW: Australia Council. Stenglin, M. (2002), ‘Comfort and Security: A Challenge for Exhibition Design’, in L. Kelly and J. Barrett (eds), UNCOVER: Volume 1, 17–22, Sydney: Australian Museum. Stenglin, M. (2004), ‘Packaging Curiosities: Towards a Grammar of ThreeDimensional Space’, PhD thesis, University of Sydney, Sydney. Stenglin, M. (2007), ‘Making Art Accessible: Opening Up a Whole New World’, Visual Communication, 6 (2): 202–13. Stenglin, M. (2008a), ‘Olympism: How a Bonding Icon Gets Its “Charge”’, in L. Unsworth (ed.), Multimodal Semiotics: Functional Analysis in the Contexts of Education, 50–66, London: Continuum. Stenglin, M. (2008b), ‘Binding: A Resource for Exploring Interpersonal Meaning in 3D Space’, Social Semiotics, 18 (4): 425–47. Stenglin, M. (2009a), ‘Space and Communication in Exhibitions: Unravelling the Nexus’, in C. Jewitt (ed.), Routledge Handbook of Multimodal Analysis, 272–83, Oxford: Routledge. Stenglin, M. (2009b), ‘From Musing to Amusing: Semogenesis and Western Museums’, in E. Ventola and A. J. M. Guijarro (eds), The World Shown and the World Told, 245–65, Houndsmills: Palgrave Macmillan. Stenglin, M. (2011), ‘Spaced Out: An Evolving Cartography of a Visceral Semiotic’, in S. Dreyfus, S. Hood and M. Stenglin (eds), Semiotic Margins: Meaning in Multimodalities, 73–100, London: Continuum. Stenglin, M. (2012a), ‘“Glocalisation”: Exploring the Dialectic between the Local and the Global’, in W. L. Bowcher (ed.), Multimodal Texts from around the World: Cultural and Linguistic Insights, 123–45, Houndsmills: Palgrave Macmillan. Stenglin, M. (2012b), ‘Transformation and Transcendence: Bonding through Ritual’, closing plenary given at the 39th International Systemic Functional Congress, University of Technology Sydney, Sydney, 16–20 July. Stenglin, M. (2013), ‘Interpersonal Meaning in 3D Space: Binding Creates Hope for Bonding’, in E. Montagna (ed.), Readings in Intersemiosis and Multimedia, 319–37, Como: Ibis. Stenglin, M. (2017), ‘Building Bridges: Design, Emotion and Museum Learning’, in P. Goodyear, M. de Laat and L. Carvelho (eds), Place-Based Spaces for Networked Learning, 131–41, London: Routledge. Stenglin, M. and E. Djonov (2010), ‘Unpacking Narratives in a Hypermedia Art Adventure for Children’, in C. R. Hoffman (ed.), Narrative Revisited: Telling a Story in the Age of New Media, 185–212, Amsterdam: John Benjamins. Stenglin, M. and M. Foureur (2013), ‘Designing Out the Fear Cascade to Increase the Likelihood of Normal Birth’, Midwifery, 29 (8): 819–25. White, P. (1994), ‘Images of the Shark: “Jaws”, Goldfish or Cuddly Toy? An Analysis of the Australian Museum’s Shark Exhibition from a Communicative Perspective’. Unpublished monograph, Department of Linguistics, the University of Sydney. White, P. (1997), ‘Death, Disruption and the Moral Order: The Narrative Impulse in Mass-Media “Hard News” Reporting’, in F. Christie and J. R. Martin (eds), Genre and Institutions: Social Processes in the Workplace and School, 101–33, London: Cassell. White, P. (1998), ‘Telling Media Tales: The News Story as Rhetoric’, PhD thesis, University of Sydney, Sydney.

CHAPTER EIGHTEEN

Affiliation An Appliable Framework for Exploring Community and Identity in Discourse LORENZO LOGI AND MICHELE ZAPPAVIGNA

INTRODUCTION Understanding how feelings and stances are shared or contested in interactions is crucial to interpreting how people forge social relationships. This is of critical concern across many areas of applied linguistics including interactional sociolinguistics, conversation analysis and pragmatics (see, for instance, Locher and Graham 2010 for an overview of studies concerning interpersonal relationships in the field of pragmatics). Interest in affiliation within SFL reflects Firth’s (1964: 112) insight that the ‘promotion, establishment, and maintenance of communion of feeling is perhaps four-fifths of all talk’, as well as his focus on studying users of language to understand how linguistic resources are deployed to perform identity and construe community (Firth 1957). SFL work on affiliation has also been motivated by an interest, particularly keen at the turn of the century, in forging a greater focus on interpersonal meaning in linguistic research given the tendency to background this dimension in work at the time (Poynton 1991). Interpersonal meaning in language, according to a social semiotic perspective, is concerned with how relationships are enacted (Halliday 1978). It is thus key to understanding how values are shared and negotiated in discourse, and

326

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

how ‘persona [align] themselves into sub-cultures, configuring master identities and constituting a culture’ (Martin 2010: 24). Accordingly, the study of how individual interactants propose and negotiate feelings and opinions allows linguists to infer the development and nature of their individual and social identities, thus providing a crucial theoretical and analytical bridge between SFL’s extravagant elaboration of the system and structure of language and exploring real-life social semiosis. SFL work on affiliation has focused on the discourse semantic and context strata, analysing texts across varied genres and modalities, and has evolved from its seminal form to describe a variety of medium-specific models. From Knight’s (2010b) work mapping the role of humour and laughter in negotiating social bonds in casual conversation among friends to Zappavigna’s (2018) work exploring the role of hashtags in online discourse, the study of affiliation has sought answers to questions of how context-specific linguistic resources function to enact the broader sociolinguistic phenomena of solidarity and alignment. Work on affiliation also progresses SFL theory on individuation and instantiation (Martin 2010), elaborating a model for how language construes identities out of cultural reservoirs and individual repertoires of meaning (to use Bernstein’s 2000 terms), and how communities form around shared values. The affiliation model foregrounds the social dimension of social semiosis, providing a framework for mapping how the semiotic resources of language bundle and interact to form our social reality.

UNITS OF ANALYSIS: ‘COUPLINGS’ AND ‘BONDS’ In order to elaborate a social semiotic theory of affiliation, a link was needed between linguistic resources identifiable at the discourse semantic stratum and the interpersonal phenomena occurring at the contextual stratum. Within discourse, the key unit of analysis underlying the approaches to affiliation described in this chapter is ‘coupling’, that is the co-selection of discursive features,1 in particular the coupling of ideational and attitudinal meanings. The term was first proposed by Martin (2000: 164) to explain humour ‘as involving discordant couplings – either between appraisal selections and what is being appraised, or among the appraisal variables themselves’. Evaluative couplings, based on the systems of ideation and attitude (as presented in Martin and White 2005), have since formed the core analytical unit in research spanning casual conversation (Knight 2013), academic discourse (Hao and Hood 2017), business writing (Szenes 2016), social media (Zappavigna 2018) and youth justice conferencing (Zappavigna and Martin 2018).

EXPLORING AFFILIATION IN DISCOURSE

327

An example of an ideation–attitude coupling is realized in the following extract from a casual conversation about holiday food among young female university students (a dataset used by Knight (2010a: 150) that we will return to in the next section of this chapter): This was an awesome pie party guys. This can be analysed as construing the following coupling: [ideation: pie party/attitude: + appreciation] This coupling is presented in the notational format2 proposed by Martin et al. (2013) and may be generalized as: [ideation: / attitude: ] The square brackets indicate that ideational and attitudinal meaning are fused in the coupling and can thus be negotiated by interlocutors as a whole. The plus (+) or minus (–) signs denote positive or negative polarity of attitude resources. Consider, for instance, the following text of two YouTube viewers commenting on a video: Viewer 1: Watching this the night before my GCSEs. I’ve always thought this is one of the best ASMR3 videos ever made. Viewer 2: Yep. I’m subscribed to her channel, but I favorited this one because it’s such a classic. In this example, Viewer 1’s description of the video as ‘one of the best ASMR videos ever made’ construes the coupling of [ideation: ASMR video/ attitude: + appreciation]. This combination of ideation and attitude is then repeated in Viewer 2’s comment that ‘it’s such a classic’ (where ‘it’ refers to the ASMR video), signalling that the coupling is being jointly negotiated by the interactants. Viewed through the lens of affiliation, couplings are considered linguistic evidence for the social semiotic unit of bonds (as conceptualized by Stenglin 2008; Martin 2004, 2010; Martin and Stenglin 2007). Martin explains this relationship: ‘the coupling of experience with evaluation, when shared by interlocutors, creates a bond … It is these bonds which form the basic building blocks of the individuation hierarchy, clustering into the sub-cultures and master identities to which community members subscribe’ (2010: 26). As such, the affiliation model provides an appliable rubric for identifying the linguistic resources (couplings construed in discourse) that serve to table the social semiotic bonds which can then be negotiated among interactants.

328

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

DIALOGIC AFFILIATION IN FACE-TO-FACE CONTEXTS The model of dialogic affiliation was developed by Knight (2008, 2010a, 2010b, 2013) to account for how social bonds are realized interactively in dialogic communication. The model grew out of Knight’s research into how interpersonal meaning functions in the register of casual conversation among friends discussing everyday experiences, and combines work on coupling and bonding to build on Poynton’s (1984, 1991) description of how spoken texts enact social solidarity. Excerpts of Knight’s data are shown in Tables 17.2 to 18.4. These extracts give an indication of the kind of unfolding prosodies of attitude that are often produced in casual conversation and which were identified repeatedly in Knight’s larger dataset. According to Knight’s model, there are three affiliation strategies that can be deployed in proposing and reacting to a bond: communing, laughing and condemning: • Communing: sharing a bond or rallying around a bonding icon (Stenglin 2008) as per rallying/epideictic discourse; • Laughing: deferring an unshared potential bond; communing around an implicated bond, thereby creating a ‘wrinkle’ tension, as per humorous discourse; • Condemning: rejecting an unshareable potential bond to commune around a shared bond, creating a ‘violation’ tension, as per gossip discourse (Knight 2010: 217). Communing affiliation describes the type of affiliation occurring when ‘participants come together to share a bond’ as in epideictic or rallying discourses (Knight 2010a: 217). For example, the interlocutors show collective support for what Knight terms a ‘fun pie party’ bond by rallying around the [ideation: pie/attitude: + t-appreciation] coupling in an act of communing affiliation (Table 18.1). Knight interprets the unfolding affiliation in Table 18.2 as the three conversational interlocutors communing ‘around a bond realised by a coupling of intensified positive appreciation for a pie party that they regularly participate in together’ (Knight 2010a: 219). Conversely, laughing affiliation occurs when a bond is tabled that is unshareable in the context of the interaction and wrinkles with a shared, implicated bond, and involves discharging the ensuing tension through laughter. The ‘wrinkle’ interrupts the bonding process, creating social tension because it is ‘a potential bond that cannot be shared by the conversational participants and which contrasts in a laughable way with an implicated bond that they share together’ (Knight 2010a: 208). As we can see in Table 18.2, the friends repeatedly table a ‘happy fatness’ bond, instantiated as the coupling [ideation: eating/attitude: + appreciation]. This wrinkles with a shared, implicated

EXPLORING AFFILIATION IN DISCOURSE

329

TABLE 18.1  Positive Evaluation of a Pie Party in a Casual Conversation Speaker

Talk

Coupling

A

Party’s over. For me;

B

No more pie party ==

C

== This was an awesome pie party guys

[ideation: pie party/ attitude: + appreciation]

B

I love pie parties

[ideation: pie party/ attitude: + t-appreciation4]

C

I can’t wait to have another one

[ideation: pie party/ evaluation: + t-appreciation]

Bond (deferred or shared) Fun pie party

Source: Adapted from Knight (2010a: 219).

TABLE 18.2  Deferral of a Wrinkling Bond in a Casual Conversation Speaker

Talk

Coupling

Bond (deferred or shared)

A

== Yeah I saw like my family and friends … I ate well ((laughs))

[ideation: eating/attitude: + appreciation]

Happy fatness Beautiful thinness

B

We all ate well.

[ideation: eating/attitude: + appreciation]

all

((all laugh))

B

Dude we all ((laughing)) ate good pie!

all

((continuous laughing))

B

Yes I agree. ((continuous laughing)) On a diet now.

all

((all laugh))

[ideation: pie/attitude: + appreciation] [ideation: self/attitude: – judgement5]

Source: Adapted from Knight (2010a: 219).

‘beautiful thinness’ bond, commensurate with other body-conscious values that they share, and the tension generated by this wrinkle is discharged through laughter. Similarly to laughing affiliation, condemning occurs when a tabled bond is in tension with a shared bond. However, in the case of condemning the tabled bond is unshareable in any context, violates a shared bond and generates tension

330

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

TABLE 18.3  An Example of Condemning Speaker

Talk

Coupling

Bond (rejected or shared)

A

It was the way she would act sometimes would be very … kinda negative. [pause 2 secs]

[ideation: she/attitude: – appreciation]

And we’d – we’d try like to keep a positive environment … (at the tech),

[ideation: keeping positive/ attitude: + appreciation] [ideation: motivation/ attitude: + appreciation]

destructive criticism motivating encouragement destructive criticism

you can motivating be but it’s not very motivating to have someone yelling at you for stuff, and

[ideation: yelling/attitude: –t-appreciation

B

Mm hm

Source: Adapted from Knight (2010a: 233).

that cannot be discharged through laughter. The tension is instead resolved by rejecting the tabled bond and communing around the shared bond it violates;6 this can be seen in the rejection of the ‘destructive criticism’ bond in favour of a ‘motivating encouragement’ bond (Table 18.3). Following Knight’s early work, affiliation theory was employed in much of the research published in the chapters in Martin and Bednarek (2010), which presented new approaches in SFL to exploring identity and affiliation. Analysing texts ranging from rap music (Caldwell 2010), front pages of Thai newspapers (Knox, Patpong and Piriyasilpa 2011), image-nuclear news stories (Caple 2010), bilingual picture books (Tian 2010), history text books (Tann 2010), educational hypertexts (Zhao 2010), the paralanguage of youth justice conferencing (Zappavigna et al. 2010) and corpora of television drama dialogue (Bednarek 2010), the authors demonstrate that affiliation can be explored across media and genres and propose application-specific linguistic tools for achieving this.

IDENTITY AS ‘COUPLING DISPOSITION’ In pursuing a social semiotic exploration of affiliation, Knight’s work has also informed research into identity, which can be seen as inversely correlated with affiliation in that any act of affiliating into a shared bond community is also an act of construing the self (as a textual persona). In this vein, Zappavigna (2014c)

EXPLORING AFFILIATION IN DISCOURSE

331

conceptualized identities as characterized by particular ‘coupling dispositions’ which constitute a tendency to construe particular patterns of values. Zappavigna (2014b) analysed Twitter discourse about coffee, an everyday bonding icon, to identify two general tendencies in coupling patterns suggesting two broad identities enacted in relation to coffee. The first coupling disposition, termed the ‘coffee connoisseur’, was defined by evaluating the sensory/aesthetic quality of coffee. This identity was typified by the coupling pattern [ideation: coffee/ attitude: + reaction], for example: Ditto! RT @User: Just had an Allagash Black. Fantastic! Notes of roasted coffee, chocolate, and fruity Belgian yeast. (Zappavigna 2014b: 153) This identity related to that of wine and chocolate connoisseurs, as well as the generalized ‘foodie’ or ideologically charged ‘food snob’, characterized by couplings of positive evaluation (reaction) with the ideational target, a ‘fine life’. The second coupling pattern identified in the dataset was that of the ‘coffee addict’ persona, who is ‘bound to consuming coffee in order to survive the day’ (Zappavigna 2014b: 153). This persona was construed through repeated couplings such as [ideation: coffee/attitude: desire] in hashtags such as #needcoffee: I hate waking up at 6am everyday. Ughh #needcoffee. (Zappavigna 2014b: 154) From a methodological perspective, Zappavigna (2014b) demonstrates how combining close, qualitative manual coding of texts for couplings with quantitative methods of finding patterns in corpora can offer insight into the bonds that construe communities and, conversely, textual persona. Whilst this mixed-methods approach would require repetition and possibly refinement for other genres, it nonetheless offers a promising example of how affiliation theory might be used to analyse the values that comprise communities and how they are construed in larger-scale studies.

AMBIENT AFFILIATION IN DIGITAL CONTEXTS Digital communication, in particular social media communication, has become an important way in which people relate to one another. However, affiliation theory designed to account for the negotiation of values in face-toface communication cannot simply be directly ported to the digital domain. While Knight’s model accounts for affiliation in dialogic registers where there is a direct interaction between interlocutors, we also need to account for the ‘ambient’ or ‘broadcast’ dimension of affiliation where interactants are

332

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

separated by physical and/or temporal distance. One important domain in which social bonding is increasingly occurring is on social media platforms. ‘Ambient affiliation’ is a framework aimed at investigating bonding in these kinds of digital domains. This framework was originally developed for considering affiliation in microblogging, focusing on the microblogging platform Twitter (Zappavigna 2011). It has since been used to explore bonding in relation to Twitter discourses about motherhood (Zappavigna 2014c), everyday life (Zappavigna 2014a, 2014b), food (Drasovean and Tagg 2015), language learning (Solmaz 2017), celebrity (Zappavigna 2013), organic farming (Kozaki and Akoumianakis 2014), feminism (Han 2015), Brexit (Zappavigna 2017, 2019a) and politics (Persson 2017; Zappavigna 2014d, forthcoming). It has also been used to interpret self-representation practices in selfies across visual social media such as Instagram and Tumblr (Zappavigna 2016; Zappavigna and Zhao 2017, 2020; Zhao and Zappavigna 2017, 2018a, 2018b). For an overview of SFL research into social media and affiliation see Zappavigna (2019b). An important reason why existing models of ‘offline’ or ‘face-to-face’ affiliation cannot simply be transposed to digital communication is that, due to the affordances of digital platforms, some of the assumptions underlying direct interactional discourse cannot be made. Tabling a coupling via a social media platform usually means construing a value in the absence of paralanguage and, often, repartee. In addition, the concept of turn taking or an explicit dialogic exchange cannot be assumed in platforms where interactions can be multiparty, temporally ambiguous (sometimes synchronous; sometimes asynchronous), overlapping, and never necessarily reaching completion (often additional messages are possible even years after an initial encounter) (Zappavigna 2019a). Another important dimension of affiliation that digital communication problematizes is the assumption that direct interaction is necessary for bonding to occur. While many forms of interaction are possible via social media platforms, often social media posts do not attract a direct reply from another user (e.g. most tweets do not receive a reply; Liu, Kliman-Silver and Mislove 2014). Nevertheless these texts are attempting to forge bonds with the ambient audience and our model of affiliation needs to account for this semiotic practice. Users may be contributing to mass semiotic practices such as using a shared hashtag or contributing to an iteration of an internet meme (Zappavigna, forthcoming), neither of which necessarily require direct communication with another user, yet both of which are participations in collective shared feeling. Communing affiliation: Convoking, promoting and finessing a coupling when there is no dialogue Affiliation in modes where participants do not interact directly has been most extensively explored in relation to hashtag7 use (Zappavigna 2018; Zappavigna

EXPLORING AFFILIATION IN DISCOURSE

333

and Martin 2018). In social media discourse, hashtags serve to commune around particular couplings without negotiating them between interactants, as would occur in a conversational exchange. As such, hashtags comprise one of a suite of resources (including emoticons/emoji, etc.) that can augment written language in digital modes in lieu of paralanguage and clearly defined dialogic exchange. Exploring how they are used therefore sheds insight into how interactants elaborate different resources for expressing meaning and negotiating ideation– attitude couplings in contexts such as social media platforms which afford ambient affiliation. An augmented affiliation framework was thus developed by Zappavigna (2018) in order to account for three systems of meaning which are relevant to ambient communion of the kind possible on social media platforms: • convoking – mustering community around a coupling, for instance via naming or classifying, e.g. #TrumpTrain • finessing – heteroglossically positioning a coupling in relation to other potential couplings, e.g. via engagement resources such as negation, e.g. #NotMyPresident • promoting – interpersonally emphasizing a coupling, for instance via graduation or typographic emphasis, e.g. #EVIL. These systems of meaning describe an interactant’s ‘alignment to the coupling’ (Han 2015: 30). The convoking system invites a specific community to bond around a coupling by ‘calling together’ a group through the use of a hashtag and implies strong alignment from the author of the tweet and the convoked community to the coupling. For example, Zappavigna (2018) documented use of the #MAGA (Make America Great Again) hashtag for convocation which acted as a marker for a Trump-supporter identity and thereby invited communion from the proTrump community. For instance, in this example the #MAGA tag was used to convoke the pro-Trump community around the coupling [ideation: media/ attitude: -propriety]: The liberal media is all hypocrisy. #maga ☕ (Zappavigna 2018: 134) Finessing is used to position the tabled coupling in relation to other potential stances in the social media discourse landscape. Finessing modulates how potential bonders will respond to a tabled coupling by referencing alternative voices. As such, it relates to the discursive system of engagement described by Martin and White’s (2005) appraisal framework. engagement describes dialogic resources that construe alternate value positions and textual personae. Where engagement is concerned with the patterning of attitudinal resources, finessing

334

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

accounts for how ideation–evaluation couplings are modulated. For instance, consider the following posts tabling couplings of negative propriety (evil) with Trump, his administration or an act of his administration: Trump’s EO undoing Obama’s protection for Women on #EqualPayDay is so gross. 45 is pure effing evil. #NotMyPresident #GoToJail (Zappavigna 2018: 137) Here, the hashtag #NotMyPresident contains heteroglossic contraction (Not) interacting with the ideational target (MyPresident), which affords negative judgement in order to critique Trump. The tabled bonds of negative evaluation position themselves as opposed to an implicit ‘mypresident’ alternative ascribed to a putative ambient other (shown in grey in Figure 18.1, based on Han’s (2015) diagramming of a similar pattern in #feminism tweets). Thus the coupling in the hashtag implicated this potentially opposing voice, which finesses the [ideation: Trump/attitude: - propriety] couplings in the posts by positioning them in a landscape of alternative perspectives. Beyond the convoking and finessing functions identified above, hashtags can interpersonally promote a coupling in a tweet. Promoting performs a similar upscaling or intensifying function to resources in the system of graduation,

FIGURE 18.1  Hashtag with engagement resource implicates alternate coupling.

EXPLORING AFFILIATION IN DISCOURSE

335

‘whereby feelings are amplified and categories blurred’ (Martin and White 2005: 35). With promoting, however, the target of the upscaling or intensifying is not only an attitude but an entire coupling. This is most evident when the coupling itself is contained in the hashtag, for instance, #TrumpIsEvil [ideation: Trump/attitude: - propriety]: Trump and the GOP are going to destroy our country and maybe the world. #AntiChrist #TrumpIsEvil #TrumpResistance [GIF of a nuclear bomb explosion] (Zappavigna 2018: 140) Another resource used to foreground a coupling is typographic emphasis via all CAPS (e.g. #EVIL), such as: and #TrumpCriminal just can’t wait 2 start more wars There’s one word for #Trump & the @GOP & that word is #EVIL #impeachTrump #VoteThemOut (Zappavigna 2018: 140) Further to the strategies outlined above, it should be noted that, because the multimodal affordances of hashtags render a tweet more easily findable and therefore more bondable, any instance of a hashtag will highlight the values it expresses. Hashtag memes8 offer yet another avenue for promoting couplings in tweets. Given the linguistic creativity manifest in online communication, there are likely still many more paralinguistic resources that impact how communable a coupling will be and how much engagement it will receive in ambient environments such as social media platforms.

CONCLUSION AND FUTURE DIRECTIONS This chapter has explored the evolution of affiliation as it has been applied to diverse genres and modalities. It began by tracing the evolution of affiliation theory from earlier work on interpersonal meaning and social semiosis in the form of couplings and bonds, through Knight’s (2010b) proposal of the affiliation model and description of three general affiliation strategies in casual conversation. The chapter then considered subsequent adaptations of the model to fields of practice, such as communication on social media platforms, that need to account for affiliation in non-dialogic registers. Renovations to the affiliation model arising from its application continue to advance SFL theorization of individuation and instantiation and make inroads into elaborating a framework that spans strata and metafunctions to explore how identity and community are construed in language. Affiliation remains a fecund field for further research. Its application in many modes and registers remains unexplored, as does the role alternate forms of

336

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

couplings might play in construing bonds. In addition the role of various linguistic devices such as irony and sarcasm needs to be clarified. Another area that merits deeper investigation is the role paralanguage plays in affiliation. Martin et al. (2013), Zappavigna et al. (2010) and Zappavigna and Martin (2017) have made promising inroads in this area, showing that paralanguage can serve to ‘support the proposal of bonds in discourse’ (Martin et al. 2013: 489). For example, they have suggested how paralanguage may function to foreground certain couplings over others, as well as construe alignment between interactants (through eye contact or hand gestures) and enact responses to bonds. However, a multimodal model for affiliation is yet to be proposed. Such a model would dramatically expand the analytic powers of the affiliation framework, and open the way for even more comprehensive and fine-grained analysis of how social semiosis is construed. Methodologically, the current reliance on manual coding has precluded incorporating quantitative analysis into affiliation work. The addition of quantitative techniques might offer insight into identities and communities in longer texts or across corpora. The prospect that affiliation analysis might one day be facilitated by computational tools is tantalizing and would open new horizons for the application and validation of the framework.

NOTES 1 At its most generalized definition, coupling refers to the interaction of meanings across a variety of semiotic dimensions ‘as pairs, triplets, quadruplets or any number of coordinated choices’ (Martin 2008: 39); it concerns a ‘relation of “with”: variable x comes with variable y’ (Zhao 2011: 144). This coordination can occur across ranks, metafunctions, strata or modalities, and will result in a combined discursive function. 2 Other studies have used a yin-yang symbol to show that ideational and attitudinal meanings are combining (see for instance, Hood 2010: 144; Szenes 2016). 3 Autonomous sensory meridian response (ASMR) videos are a popular type of YouTube video aimed at creating sounds that elicit a pleasant physical response from viewers. 4 The t-notation here is used to indicate that the affect (love) is a ‘token’ of appreciation; that is, it is a good pie party. 5 Knight (2010a: 305) analyses this as invoked attitude where negative judgement is flagged by counterexpectancy. 6 This is the case for affiliative discourse such as gossip; where a bond is shared by one interactant and rejected by another, the result is disaffiliation, characteristic of conflictual discourse. 7 A hashtag is a form of social tagging indicated by the # symbol followed by a word or phrase, e.g. #SystemicFunctionalLinguistics. 8 Memes are ‘shareable media where users contribute iterations on a theme or template, mostly as displays of wit’ (Zappavigna 2018: 13). Hashtag memes are

EXPLORING AFFILIATION IN DISCOURSE

337

marked by a hashtag and proliferate on social media platforms. Political hashtag memes are often phrases from political discourse that are used to counter particular viewpoints, e.g. #AlternativeFacts used US presidential adviser Kellyanne Conway’s phrase to mock discourse (see Zappavigna 2019b for an analysis of this meme).

REFERENCES Bednarek, M. (2010), ‘Corpus Linguistics and SFL: Interpersonal Meaning, Identity and Affiliation in Popular Culture’, in J. R. Martin and M. Bednarek (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 237–61, New York: Continuum. Bernstein, B. B. (2000), Pedagogy, Symbolic Control, and Identity: Theory, Research, Critique, Lanham, MD: Rowman & Littlefield. Caldwell, D. (2010), ‘Making Metre Mean: Identity and Affiliation in the Rap Music of Kanye West’, in J. R. Martin and M. Bednarek (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 59–79, New York: Continuum. Caple, H. (2010), ‘Doubling-Up: Allusion and Bonding’, in J. R. Martin and M. Bednarek (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 111–33, New York: Continuum. Drasovean, A. and C. Tagg (2015), ‘Evaluative Language and Its Solidarity-Building Role on TED.com: An Appraisal and Corpus Analysis’, Language@Internet, 12, article no. 1. Firth, J. R. (1957), Papers in Linguistics, 1934–1951, London: Oxford University Press. Firth, J. R. (1964), The Tongues of Men, and Speech, Oxford: Oxford University Press. Halliday, M. A. K. (1978), Language as Social Semiotic: The Social Interpretation of Language and Meaning, London: E. Arnold. Han, J. (2015), ‘“feminism is Not a Dirty Word”: Axiology, Ambient Affiliation and Dialogism in Discourses Surrounding Feminism in Microblogging’, Honours thesis, University of Sydney, Sydney. Hao, J., and S. Hood (2017), ‘Valuing Science: The Role of Language and Body Language in a Health Science Lecture’, Journal of Pragmatics, 139: 200–15. doi:10.1016/j.pragma.2017.12.001 Hood, S. (2010), Appraising Research: Evaluation in Academic Writing, New York: Palgrave Macmillan. Knight, N. K. (2008), ‘“Still Cool … and American Too!”: An SFL Analysis of Deferred Bonds in Internet Messaging Humour’, in N. Nørgaard (ed.), Systemic Functional Linguistics in Use, 481–502, Odense, Denmark: Institut for Sprog og Kommunikation, Odense Universitet. Knight, N. K. (2010a), ‘Laughing Our Bonds Off: Conversational Humour in Relation to Affiliation’, PhD thesis, University of Sydney, Sydney. Knight, N. K. (2010b), ‘Wrinkling Complexity: Concepts of Identity and Affiliation in Humour’, in M. Bednarek and J. R. Martin (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 35–58, New York: Continuum. Knight, N. K. (2013), ‘Evaluating Experience in Funny Ways: How Friends Bond through Conversational Humour’, Text & Talk, 33 (4–5): 553–74.

338

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Knox, J. S., P. Patpong and Y. Piriyasilpa (2011), ‘ขาวหน าหน (Khao naa nung): A Multimodal Analysis of Thai-language Newspaper Front Pages’, in M. Bednarek and J. R. Martin (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 80–110, New York: Continuum. Kozaki, S. and D. Akoumianakis (2014), ‘Ambient Affiliation on Twitter: A Case Study on Organic Farming’, Paper presented at 2014 International Conference on Telecommunications and Multimedia, Heraklion, Crete, Greece, 28–30 July. Liu, Y., Kliman-Silver, C., and Mislove, A. (2014), ‘The Tweets They Are a-Changin’: Evolution of Twitter Users and Behavior’, Proceedings of the International AAAI Conference on Web and Social Media, 8 (1): 305–14. Locher, M. A. and S. L. Graham (2010), Interpersonal Pragmatics, Berlin: Walter De Gruyter. Martin, J. R. (2000), ‘Beyond Exchange: Appraisal Systems in English’, in S. Hunston and G. Thompson (eds), Evaluation in Text: Authorial Stance and the Construction of Discourse, 142–75, Oxford: Oxford University Press. Martin, J. R. (2004), ‘Mourning: How We Get Aligned’, Discourse & Society, 15 (2–3): 321–44. doi:10.1177/0957926504041022 Martin, J. R. (2008), ‘Tenderness: Realisation and Instantiation in a Botswanan Town’, Odense Working Papers in Language and Communication, 29: 30–58. Martin, J. R. (2010), ‘Semantic Variation – Modelling Realisation, Instantiation and Individuation in Social Semiosis’, in J. R. Martin and M. Bednarek (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 1–34, New York: Continuum. Martin, J. R. and M. Bednarek, eds (2010), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, New York: Continuum. Martin, J. R. and M. Stenglin (2007), ‘Materializing Reconciliation: Negotiating Difference in a Transcolonial Exhibition’, in T. D. Royce and W. Bowcher (eds), New Directions in the Analysis of Multimodal Discourse, 215–338, Mahwah, NJ: Erlbaum. Martin, J. R. and P. R. R. White (2005), The Language of Evaluation: Appraisal in English, Basingstoke: Palgrave Macmillan. Martin, J. R., M. Zappavigna, P. Dwyer and C. Cléirigh (2013), ‘Users in Uses of Language: Embodied Identity in Youth Justice Conferencing’, Text & Talk, 33 (4–5): 467–96. doi:10.1515/text-2013-0022 Persson, G. (2017), ‘Love, Affiliation, and Emotional Recognition in kämpamalmö: The Social Role of Emotional Language in Twitter Discourse’, Social Media + Society, 3 (1). doi:10.1177/2056305117696522 Poynton, C. (1984), ‘Names as Vocatives: Forms and Functions’, Nottingham Linguistic Circular, 13: 1–34. Poynton, C. (1991), ‘Address and the Semiotics of Social Relations: A SystemicFunctional Account of Address Forms and Practices in Australian English’, PhD thesis, University of Sydney, Sydney. Solmaz, O. (2017), ‘Autonomous Language Learning on Twitter: Performing Affiliation with Target Language Users through Hashtags’, Journal of Language and Linguistic Studies, 13 (2): 204–20. Stenglin, M. (2008), ‘Interpersonal Meaning in 3D Space: How a Bonding Icon Gets Its “Charge”’, in L. Unsworth (ed.), Multimodal Semiotics: Functional Analysis in Contexts of Education, 50–66, London: Continuum. Szenes, E. (2016), ‘The Linguistic Construction of Business Reasoning: Towards a Language-Based Model of Decision-Making in Undergraduate Business’, PhD thesis, University of Sydney, Sydney.

EXPLORING AFFILIATION IN DISCOURSE

339

Tann, K. (2010), ‘Imagining Communities: A Multifunctional Approach to Identity Management in Texts’, in J. R. Martin and M. Bednarek (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 163–94, New York: Continuum. Tian, P. (2010), ‘Playing with “Femininity”: An Intermodal Analysis of the Bilingual Picture Book The Ballad of Mulan’, in J. R. Martin and M. Bednarek (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 134–62, New York: Continuum. Zappavigna, M. (2011), ‘Ambient Affiliation: A Linguistic Perspective on Twitter’, New Media & Society, 13 (5): 788–806. doi:10.1177/1461444810385097 Zappavigna, M. (2013), ‘“If You Do It Too Then RT and Say idoit2”: The Copatterning of Contingency and Evaluation in Microblogging’, in M. Taboada and R. Trnavac (eds), Nonveridicality and Evaluation: Theoretical, Computational and Corpus Approaches, 188–213, Leiden: Brill. Zappavigna, M. (2014a), ‘Ambient Affiliation in Microblogging: Bonding around the Quotidian’, Media International Australia, 151 (1): 97–103. Zappavigna, M. (2014b), ‘Coffeetweets: Bonding around the Bean on Twitter’, in P. Seargeant and C. Tagg (eds), The Language of Social Media: Communication and Community on the Internet, 139–60, London: Palgrave. Zappavigna, M. (2014c), ‘Enacting Identity in Microblogging’, Discourse and Communication, 8 (2): 209–28. Zappavigna, M. (2014d), ‘Enjoy Your Snags Australia … Oh and the Voting Thing Too ausvotes auspol: Iconisation and Affiliation in Electoral Microblogging’, Global Media Journal: Australian Edition, 8 (2). https://www.hca.westernsydney. edu.au/gmjau/?p=1139 Zappavigna, M. (2016), ‘Social Media Photography: Construing Subjectivity in Instagram Images’, Visual Communication, 15 (3): 271–92. Zappavigna, M. (2017), ‘“Had Enough of Experts”: Intersubjectivity and Quotation in Social Media’, in E. Friginal (ed.), Studies in Corpus-Based Sociolinguistics, 321–43, London: Routledge. Zappavigna, M. (2018), Searchable Talk: Hashtags and Social Media Metadiscourse, London: Bloomsbury. Zappavigna, M. (2019a), ‘Ambient Affiliation and #brexit: Negotiating Values about Experts through Censure and Ridicule’, in V. Koller, S. Kopf and M. Miglbauer (eds), Discourses of Brexit, 48–68, London: Routledge. Zappavigna, M. (2019b), ‘Language and Social Media: Enacting Identity through Ambient Affiliation’, in G. Thompson, W. Bowcher, L. Fontaine and J. Y. Liang (eds), The Cambridge Handbook of Systemic Functional Linguistics, 714–37, London: Cambridge University Press. Zappavigna, M. (forthcoming). ‘“And Then He Said … No One Has More Respect For Women Than I Do”: Intermodal Relations and Intersubjectivity in Image Macros’, in H. Stöckl, H. Caple and J. Pflaeging (eds), Image-Centric Practices in the Contemporary Media Sphere, 204–25, London: Routledge. Zappavigna, M. and J. Martin (2017), Discourse and Diversionary Justice: An Analysis of Youth Justice Conferencing, London: Springer. Zappavigna, M. and J. R. Martin (2018), ‘Communing Affiliation: Social Tagging as a Resource for Aligning Around Values in Social Media’, Discourse, Context & Media, 22: 1–64. Zappavigna, M. and S. Zhao (2017), ‘Selfies in “Mommyblogging”: An Emerging Visual Genre’, Discourse, Context & Media, 20: 239–47.

340

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

Zappavigna, M. and S. Zhao (2020), ‘Selfies and Recontextualisation: A Social Semiotic Perspective on the Visual Structure of Instagram Images’, in M. Miles and E. Welch (eds), Photography and Its Publics, 207–30, London: Bloomsbury. Zappavigna, M., C. Cléirigh, P. Dwyer and J. R. Martin (2010), ‘The Coupling of Gesture and Phonology’, in J. R. Martin and M. Bednarek (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 219–36, New York: Continuum. Zhao, S. (2010), ‘Intersemiotic Relations as Logogenetic Patterns: Towards the Restoration of the Time Dimension in Hypertext Description’, in J. R. Martin and M. Bednarek (eds), New Discourse on Language: Functional Perspectives on Multimodality, Identity, and Affiliation, 195–218, New York: Continuum. Zhao, S. (2011), ‘Learning through Multimedia Interaction: The Construal of Primary Social Science Knowledge in Web-based Digital Learning Materials’, PhD thesis, University of Sydney, Sydney. Zhao, S. and M. Zappavigna (2017), ‘Beyond the Self: Intersubjectivity and the Social Semiotic Interpretation of the Selfie’, New Media & Society, 20 (5): 1735–54. Zhao, S. and M. Zappavigna (2018a), ‘Digital Scrapbooks, Everyday Aesthetics and the Curatorial Self: Social Photography in Female Visual Blogging’, in F. Forsgren and E. S. Tønnessen (eds), Multimodality and Aesthetics, 218–35, London: Routledge. Zhao, S. and M. Zappavigna (2018b), ‘The Interplay of Technologies and Genre: The Case of the Selfie’, Social Semiotics, 28: 665–82.

CHAPTER NINETEEN

Semiotic Description Grappling with Mathematics Y. J. DORAN

BACKGROUND Why are physics and mathematics hard? In broad terms, this is the question that drove the research I will discuss in this chapter. As an undergraduate student of physics and mathematics at the University of Sydney, I enjoyed studying them, but at times struggled with making sense of how they worked and how to do them (leading me to drop mathematics at the time I was enjoying it most). At the same time, as a student of linguistics, I had an interest in combining these three areas. In my limited understanding of the world of research, this left for me the option of going ‘small’ and exploring acoustic phonetics (the physics and mathematics of language) or going ‘big’ (relatively speaking) and looking at the discourse of physics and mathematics. What made this decision relatively simple was my more political interest in education, which connected with the ideological world view of Sydney School pedagogy as I knew it at the time. So, I went ‘big’. This chapter reflects upon the process of grappling with the mathematics used in physics and how this led to exploring broader theoretical questions in semiotics. In particular, it explores the process of describing mathematical symbolism in semiotic terms, the influence it had from areas as diverse as multimodality, sociology of education and language description, and the implications this had for understanding theoretical dimensions such as

342

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

metafunction in semiosis (Doran 2018a, 2018b). This description was by no means the only component of the broader research into physics, but it functions as a relatively self-contained exploration that moves from practical considerations to theoretical questions and contributions (for SFL). In the following section, I will first position my research within the broader context of educational linguistics at the time, before stepping through the process of grappling with mathematics as a semiotician.

KNOWLEDGE BUILDING This work was situated in a broader concern for how knowledge is built across disciplines. One of the key learnings of Sydney School pedagogy through the 1980s and 1990s was that a generic focus on literacy that did not take into account particular practices of disciplines was likely going to be ineffective. Amongst other things, this led to an emphasis on embedded literacy whereby the reading and writing is integrated into the ‘content’ area teaching, whether this be in topic areas in primary school, or the more disciplinary-specialized subjects of secondary school or tertiary education. This was in contrast to literacy programmes that taught generic literacy skills as a focus in and of themselves, with little link to any disciplinary practice. To teach embedded literacy across schooling, however, required that the literacy practices of each discipline be understood. Most directly relevant to my study of physics, foundational work on scientific discourse in SFL/social semiotics had been developed for many decades (e.g. Lemke 1990; Rose, McInnes and Korner 1992; Halliday and Martin 1993; Martin and Veel 1998; Halliday 2004), which gave a solid grounding for the work. At the same time, what was most clear about the later years of secondary physics and early years of tertiary physics was the enormous emphasis on mathematics and, to a slightly lesser extent, on images (backed up by corpus studies of Lemke 1998; Parodi 2012). Although key descriptions of mathematics (Lemke 2003; O’Halloran 2005) and images (in particular Kress and van Leeuwen 2006) had been developed, their role in disciplinary knowledge building had not yet been made clear, especially for physics. Embedded literacy programmes also forced the issue that not only were the particular literacy practices in use – genres, linguistic and multimodal resources – different across the disciplines, but the ways of teaching them were too. It was not a simple issue where once the literacy practices had been mapped for a discipline, they could be inserted into the same generic teaching sequence for every classroom in every discipline (David Rose’s (2021) exploration of mathematics teaching and learning in comparison to other disciplines given in Rose and Martin (2012) is an excellent illustration of this). The literacy practices and ways of teaching across schooling differ as the knowledge of the

SEMIOTIC DESCRIPTION AND MATHEMATICS

343

disciplines differ (and vice versa). The questions then became: Why are these different? And how can this be understood? A key influence for understanding this work came in the form of Bernstein’s (1999) late work distinguishing different types of knowledge structure. In very broad terms, the sciences were said to exemplify a ‘hierarchical knowledge structure’ that developed through integrating an ever-expanding range of phenomena into more and more general propositions, while the humanities were said to exemplify a ‘horizontal knowledge structure’ that developed through generating new sets of languages for understanding phenomena, with little generalizing and integrating capacity. Physics, from this perspective, was often positioned as the archetypal hierarchical knowledge structure. But what this actually meant for classroom teaching and learning was not clear. In this vein, then, to understand how physics worked would likely shine a light on how other hierarchical knowledge structures worked. So at the outset of the research into physics, the problem posed was that embedded literacy teaching in physics required an understanding of both the specific literacy practices of physics as well the ways it organized its knowledge. It was suggested to me that each of these seemed dependent on the highly multimodal nature of physics discourse, especially its use of mathematics. This suggestion came in the form of the seemingly innocuous but frustratingly difficult question to answer: Why does physics use mathematics? Or why can it not just use language? It was to this that I first turned.1

ANALYSING MATHEMATICS To understand why mathematics was used in physics, a linguistic or semiotic approach needs a way of grasping mathematical symbolism, just as it needs a way to grasp language. For this task, I was able to build upon O’Halloran’s (2005) foundational systemic functional description of mathematics. This description opened the way for mathematics to be understood on comparable terms to language, enabling texts to be analysed. In particular, this description offered an insightful exploration of the interactions between symbols such as 2, x, y and z, and their operators such as + in equations like (1):



(1)

O’Halloran describes these through a system of transitivity, where operators such as +, – and ÷, etc. were said to realize a distinct type of process not used in language called ‘operative processes’. The = sign, on the other hand, could be considered a type of relational identifying process, similar to that of English (Halliday and Matthiessen 2014), enabling one side of the equation, such as

344

APPLIABLE LINGUISTICS AND SOCIAL SEMIOTICS

2x + y in the above equation, to be identified as symbolizing the other, such as z. This intuitive analysis highlighted the highly ‘relational’ nature of such equations, where distinct symbols or combinations of symbols from potentially quite disparate areas of physics could be identified as the same. In terms of the structural configuration of equations, the relational identifying analysis implicated a Token and a Value structure, such as in:



(2)

This interpretation of equations as relational clauses offered a crucial insight that gave a glimpse into how mathematics is able to link a wide range of phenomena into a relatively integrated whole. But when it came to analysing texts, it was not always clear how this grammar could be applied. For example, it was not clear what to do with equations such as (3) or (4) where there was more than one equals sign occurring at the time:





(3)

(4)

For such an equation, what should be analysed as Token and what as Value? And with only two functions to play with, it meant that there needed to be multiple iterations of these functions, as in (5):



(5)

Compounding this, it was not clear how one could decide which side of the equation was Token and which was Value. Should the left side in 2x + y = z be the Token or should the right? And if the ordering was swapped to become z = 2x + y, did this mean that the Token^Value analysis was swapped too? If so, why? Or if not, why not? Finally, although the description made intuitive sense for equations using the equals sign =, it was not clear what should be done with statements that used other relators such as >,