TEACHING SCIENCE

Science has never been more important, yet science education faces serious chal-
lenges. At present, science education research only sees half the picture, focusing on
how students learn and their changing conceptions. Both teaching practice and what
is taught, science knowledge itself, are missing.This book offers new, interdisciplinary
ways of thinking about science teaching that foreground the forms taken by science
knowledge and the language, imagery and gesture through which they are expressed.
This book brings together leading international scholars from Systemic
Functional Linguistics, a long-established approach to language, and Legitimation
Code Theory, a rapidly growing sociological approach to knowledge practices. It
explores how to bring knowledge, language and pedagogy back into the picture of
science education but also offers radical innovations that will shape future research.
Part I sets out new ways of understanding the role of knowledge in integrating
mathematics into science, teaching scientific explanations and using multimedia
resources such as animations. Part II provides new concepts for showing the role of
language in complex scientific explanations, in how scientific taxonomies are built,
and in combining with mathematics and images to create science knowledge. Part
III draws on the approaches to explore how more students can access scientific
knowledge, how to teach professional reasoning, the role of body language in sci-
ence teaching, and making mathematics understandable to all learners.
Teaching Science offers major leaps forward in understanding knowledge, language
and pedagogy that will shape the research agenda far beyond science education.

Karl Maton is the creator and architect of Legitimation Code Theory.

J. R. Martin is a world-leading authority in Systemic Functional Linguistics.

Y. J. Doran is a leading young scholar combining both frameworks in research.

All three are members of the LCT Centre for Knowledge-Building.

Legitimation Code Theory
Knowledge-building in research and practice
Series editor: Karl Maton
LCT Centre for Knowledge-Building

This series focuses on Legitimation Code Theory or ‘LCT’, a cutting-edge approach

adopted by scholars and educators to understand and improve their practice. LCT
reveals the otherwise hidden principles embodied by knowledge practices, their
different forms and their effects. By making these ‘legitimation codes’ visible to be
learned or changed, LCT work makes a real difference, from supporting social jus-
tice in education to improving design processes. Books in this series explore topics
across the institutional and disciplinary maps of education, as well as other social
fields, such as politics and law.

Accessing Academic Discourse

Systemic Functional Linguistics and Legitimation Code Theory
Edited by J. R. Martin, Karl Maton and Y. J. Doran
Building Knowledge in Higher Education
Enhancing Teaching and Learning with Legitimation Code Theory
Edited by Chris Winberg, Sioux McKenna and Kirstin Wilmot
Turning Access into Success
Improving University Education with Legitimation Code Theory
Sherran Clarence
Teaching Science
Knowledge, Language, Pedagogy
Edited by Karl Maton, J. R. Martin and Y. J. Doran
Decolonizing Knowledge and Knowers
Struggles for University Transformation in South Africa
Edited by Mlamuli Nkosingphile Hlatshwayo, Aslam Fataar, Hanelie Adendorff, Paul
Maluleka and Margaret A. L. Blackie
Enhancing Science Education
Exploring Knowledge Practices with Legitimation Code Theory
Edited by Margaret A. L. Blackie, Hanelie Adendorff and Marnel Mouton
Legitimation Code Theory
A Primer
Karl Maton

For a full list of titles in this series, please visit:

https://www.routledge.com/Legitimation-Code-Theory/book-series/LMCT
TEACHING SCIENCE
Knowledge, Language,
Pedagogy

Edited by Karl Maton, J. R. Martin and

Y. J. Doran
First published 2021
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2021 selection and editorial matter, Karl Maton, J. R. Martin and Y. J. Doran; individual chapters, the
contributors
The right of Karl Maton, J. R. Martin and Y. J. Doran to be identified as the authors of the editorial
material, and of the authors for their individual chapters, has been asserted in accordance with sections
77 and 78 of the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by
any electronic, mechanical, or other means, now known or hereafter invented, including photocopying
and recording, or in any information storage or retrieval system, without permission in writing from
the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used
only for identification and explanation without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record has been requested for this book

ISBN: 978-0-815-35576-2 (hbk)

ISBN: 978-0-815-35575-5 (pbk)
ISBN: 978-1-351-12928-2 (ebk)

Typeset in Bembo
by SPi Global, India
CONTENTS

List of Figures vii

List of Tables xi
List of Contributors xiii

1 The teaching of science: New insights into knowledge,

language and pedagogy 1
Y. J. Doran, Karl Maton and J. R. Martin

PART I
Knowledge-building in science education 21

2 Targeting science: Successfully integrating mathematics into

science teaching 23
Karl Maton and Sarah K. Howard

3 Constellating science: How relations among ideas help build

knowledge 49
Karl Maton and Y. J. Doran

4 Animating science: Activating the affordances of multimedia in

teaching 76
Karl Maton and Sarah K. Howard
vi Contents

PART II
Language in science education 103

5 Field relations: Understanding scientific explanations 105

Y. J. Doran and J. R. Martin

6 Building taxonomies: A discourse semantic model of entities

and dimensions in biology 134
Jing Hao

7 Multimodal knowledge: Using language, mathematics

and images in physics 162
Y. J. Doran

PART III
Pedagogy in science education 185

8 Widening access in science: Developing both knowledge

and knowers 187
Karen Ellery

9 The relationship between specialized disciplinary knowledge

and its application in the world: A case study in engineering
design 205
Nicky Wolmarans

10 Grounded learning: Telling and showing in the language

and paralanguage of a science lecture 226
Susan Hood and Jing Hao

11 Doing maths: (De)constructing procedures for maths processes 257

David Rose

Index 287
FIGURES

Figure 2.1 The autonomy plane (Maton and Howard 2018: 6) 29

Figure 2.2 Examples of two autonomy pathways 30
Figure 2.3 Shift from sovereign code to exotic code 36
Figure 2.4 Shift from sovereign code to exotic code 40
Figure 2.5 Shift from exotic code to introjected code 41
Figure 2.6 Shift from introjected code to sovereign code 43
Figure 2.7 Autonomy tour with graphing results of an experiment 44
Figure 3.1 Taurus – some stars before constellating; stars constellated
into Taurus; overlaid with image of bull (Northern
hemisphere view) 51
Figure 3.2 Aldebaran in constellations from Inuit, Korean and Maori
cultures (hemisphere view is that of each culture) 52
Figure 3.3 Example of constellating and condensing movements in
ballet teaching (Lambrinos 2020: 91) 53
Figure 3.4 Schematic constellation of the tides explanations across
textbooks 57
Figure 3.5 Schematic constellation for tides – base and supplementary
clusters 58
Figure 3.6 Pedagogic constellation for tides – independents and
base cluster 60
Figure 3.7 Pedagogic constellation for tides – supplementary clusters 62
Figure 3.8 Schematic constellation of the seasons explanations across
textbooks 63
Figure 3.9 Schematic constellation for seasons – base and supplementary
clusters 65
viii Figures

Figure 3.10 Pedagogic constellations for seasons – ‘daylight length’ and

‘sunlight angle’ routes 67
Figure 3.11 Pedagogic constellations for seasons – ‘base cluster’ and
combined routes 70
Figure 4.1 The autonomy plane (divided into 16 modalities) 82
Figure 4.2 Moon animation: First orbit, without rotation (at 14 seconds) 84
Figure 4.3 Moon animation: Second orbit, overly fast rotation
(at 53 seconds) 84
Figure 4.4 Moon animation: Sun’s shadow simulation (at 1:53 minutes) 85
Figure 4.5 Autonomy analysis (with heuristic heat map) of
Moon animation 86
Figure 4.6 Seasons animation: Showing Earth’s tilt (at 18 seconds) 91
Figure 4.7 Seasons animation: Different angles of sunlight hitting the
Earth (at 51 seconds) 91
Figure 4.8 Seasons animation: Effects of angle and concentration of
sunlight – winter (at 1:23 minutes) 92
Figure 4.9 Seasons animation: Summer/Winter (at 1:41 minutes) 93
Figure 4.10 Visual elements of the seasons animation 94
Figure 4.11 Audio elements of the seasons animation 95
Figure 4.12 Teacher narration of the seasons animation 97
Figure 5.1 Strata of language in Systemic Functional Linguistics 107
Figure 5.2 Classification taxonomy of the seasons 108
Figure 5.3 Composition taxonomy of the solar system 108
Figure 5.4 Network for a static perspective on field 109
Figure 5.5 Momenting the activity inflammation (three tiers)
(Greenwood and Allen 2004: 118–119) 111
Figure 5.6 Network for a dynamic perspective on field 114
Figure 5.7 Dynamic and static perspectives on field 115
Figure 5.8 Basic parameters of field 117
Figure 5.9 Network of property, arraying and gauging 118
Figure 5.10 Picture of the flame robin showing its various properties.
Reproduced from Menkhorst et al. (2017: 473) with
permission from CSIRO Publishing. Original colour
illustrations by Peter Marsack 119
Figure 5.11 Graph of resistance and temperature, showing two gauged
properties 120
Figure 5.12 Map showing spatio-temporal property of the flame
robin in Australia. Reproduced from Menkhorst et al. (2017:
472) with permission from CSIRO Publishing 122
Figure 5.13 Network of property 122
Figure 5.14 Network for interdependency 129
Figure 5.15 Network of field 129
 Figures ix

Figure 6.1 Stratification and metafunctions in SFL theory 136

Figure 6.2 A simplified birdwatchers’ taxonomy of birds of prey
(adapted from Wignell et al. 1993: 156) 138
Figure 6.3 A simplified scientific taxonomy of birds of prey
(adapted from Wignell et al. 1993: 157) 139
Figure 6.4 Distinctive terminologies across strata 142
Figure 6.5 Choices across strata 144
Figure 6.6 System of entity type (1) 149
Figure 6.7 Visual representation of an item in undergraduate textbook
(Reece et al. 2014: 11) 150
Figure 6.8 A visual representation of a trained gaze entity – sea urchin
(Reece et al. 2014: 683) 151
Figure 6.9 Visual representation of Chytrids (Reece et al. 2014: 655) 151
Figure 6.10 Visual representation of chemical reaction (Reece et al.
2014: 146) 152
Figure 6.11 System of entity type (2) 152
Figure 6.12 Categorization of dung fungus 157
Figure 6.13 Categorization in relation to locust 158
Figure 6.14 Structure of locust 158
Figure 7.1 The semantic plane (Maton, 2016 p. 16) 165
Figure 7.2 Movements in the semantic plane through the language
of pushing and pulling 172
Figure 7.3 Movements in the semantic plane through mathematics 176
Figure 7.4 Diagram of an experimental apparatus 177
Figure 7.5 Graph and questions in a senior secondary school
examination 179
Figure 7.6 Movements in the semantic plane through graphs 181
Figure 8.1 The specialization plane (Maton 2016: 12) 189
Figure 8.2 Specialization codes enacted in the ISCM course 200
Figure 9.1 Idealized parking structure (S1) 217
Figure 9.2 Effect of recontextualization by weakening OSD (S1)
or DSD (U1) 217
Figure 9.3 Inferential reasoning: U1 (Bikeshare scheme) 218
Figure 9.4 Inferential reasoning: S1 (Parking structure) 220
Figure 10.1 Stratification and metafunction in systemic
functional linguistics 228
Figure 10.2 Two hands sculpting a paralinguistic entity 234
Figure 10.3 One hand sculpting a paralinguistic entity 234
Figure 10.4 A paralinguistic entity convergent with ‘box’ 235
Figure 10.5 A paralinguistic entity convergent with ‘form into three
dimensions’ 235
Figure 10.6 The taxonomy of matter configured in text (A) 236
x Figures

Figure 10.7 Equivalence in the paralinguistic depiction of co-class

entities in (9) – viewed from right to left 238
Figure 10.8 Equivalence in the depiction of an exhaustive set of
co-related paralinguistic entities (in 10) – viewed from
right to left 239
Figure 10.9 Equivalence in the depiction of a non-exhaustive set of
co-subclass entities (in 11) – viewed from left to right 240
Figure 10.10 Equivalence in the depiction of a non-exhaustive set of
co-class entities (in 12) – viewed from left to right 241
Figure 10.11 Paralinguistic entities concurrent sequentially with (16)
and (17) – viewed from right to left 243
Figure 10.12 A paralinguistic expression depicting a quality as an entitied
occurrence concurrent with ‘how sticky is it’ 244
Figure 10.13 The paralinguistic occurrence convergent with
‘being formed’ (in 19) 246
Figure 10.14 Paralinguistic expressions convergent with ‘solvent’ and
‘pull apart our solid particles’ 247
Figure 10.15 An implication sequence in language concurrently
depicted as sequence in space/time only 250
Figure 10.16 Types of ‘reaction’ established in the lecture data 251
Figure 10.17 The paralinguistic entity concurrent with ‘crystallization’ 252
Figure 10.18 The paralinguistic expression concurrent with of ‘repulsion’ 253
Figure 11.1 Hierarchy of lessons stages, lesson activities and
learning cycles 264
Figure 11.2 Nuclear and marginal phases of learning cycles 265
Figure 11.3 Multimodal teacher demonstration 266
Figure 11.4 Labelling the diagram 271
Figure 11.5 Writing the solution 277
Figure 11.6 Scribing the procedure 278
TABLES

Table 2.1 Generic translation device (Maton and Howard 2018: 10) 31
Table 2.2 Simplified specific translation device for this analysis 32
Table 2.3 Table provided by teacher for activity 34
Table 4.1 Generic translation device, first two levels (adapted from
Maton and Howard 2018: 10) 80
Table 4.2 Translation device for analysis in this chapter 81
Table 4.3 Specific translation device for Moon animation task 83
Table 4.4 Specific translation device for seasons animation task 90
Table 5.1 Field reconstruals 125
Table 5.2 Field relations in an explanation of the seasons 127
Table 6.1 Field types in relation to activity sequences and taxonomies
(c.f. Martin 1992: 545, Martin 2017) 137
Table 6.2 Linguistic definition of diffusion 140
Table 6.3 Kinds of entities (Martin and Rose 2007: 114) 141
Table 6.4 Entity types in undergraduate biology texts
(c.f. Hao 2015, 2020b) 143
Table 6.5 Realizing taxonomic relations through co-elaborations and
dimensions 154
Table 6.6 Dimensionality of entities (c.f. Hao 2015) 154
Table 6.7 Grammatical realizations of dimensions in biology texts 156
Table 7.1 Elements of field in a primary school text 171
Table 8.1 Specific translation device for knowledge code in ISCM
(adapted from Ellery 2016, 2017a) 195
Table 8.2 Specific translation device for knower code in ISCM
(adapted from Ellery 2016, 2018) 197
xii Tables

Table 9.1 Units of analysis 211

Table 9.2 Semantic gravity categories of analysis 212
Table 9.3 Semantic density categories of analysis 213
Table 9.4 Summary of analysis U1: Bikeshare scheme 219
Table 9.5 Summary of analysis S1: Parking structure 219
Table 10.1 Paralinguistic expressions of the relation of solid to matter
as subclass to class 238
Table 10.2 Paralinguistic expressions of co-subclass relations among
solid, liquids and gases 239
Table 10.3 Convergent expressions of occurrences 248
Table 11.1 Average % scores in topic tests in Year 8 maths class 259
Table 11.2 Lesson plan for R2L maths process genre 262
Table 11.3 Teacher demonstration, Step 2: writing the ratios 267
Table 11.4 Teacher demonstration, Step 4: writing the ratios 268
Table 11.5 First guided practice, Step 2: labelling the diagram 270
Table 11.6 Writing the ratios in first guided practice 272
Table 11.7 Second guided practice, Step 4: writing the ratios 274
Table 11.8 Second guided practice, Step 8: solving the problem 275
Table 11.9 Joint construction of procedure steps 279
Table 11.10 Growth in topic test scores using R2L maths process genre 282
Table 11.11 Learning cycle phases 282
Table 11.12 Record sources 283
Table 11.13 Spoken sources 283
Table 11.14 Recording 283
Table 11.15 Acts and interacts 284
CONTRIBUTORS

Y. J. Doran is a researcher at the University of Sydney who focuses on language,

semiosis, knowledge and education, spanning the interdisciplinary fields of educa-
tional linguistics, multimodality, and language and identity. He works primarily on
English and Sundanese, and from the perspectives of Systemic Functional Linguistics
and Legitimation Code Theory. His book, The Discourse of Physics: Building knowledge
through language, mathematics and image, was published by Routledge in 2018. He
co-edited (with J. R. Martin and Karl Maton) Accessing Academic Discourse (2020) for
the Routledge LCT series.

Karen Ellery is a Senior Lecturer in the Science Extended Studies Programme

of the Centre for Higher Education Research, Teaching and Learning at Rhodes
University, South Africa. Her research centres on enabling epistemological access
in the sciences in a higher education context, focusing specifically on curricu-
lum structures, pedagogic practices and student learning. Recent journal publi-
cations include ‘Congruence in knowledge and knower codes: The challenge of
enabling learner autonomy in a science foundation course’ (Alternation, 2019) and
‘Legitimation of knowers for access in science’ (Journal of Education, 2018).

Jing Hao is an Assistant Professor at the Pontificia Universidad Católica de Chile

(PUC). Previously she worked as Postdoctoral Research Fellow at PUC and at The
Hong Kong Polytechnic University, following her doctorate in linguistics at the
University of Sydney. Her research explores knowledge-building through English
and Mandarin Chinese and its interaction with other semiotic modes. Her book
Analysing Scientific Discourse From a Systemic Functional Linguistic Perspective was pub-
lished by Routledge in 2020.
xiv Contributors

Susan Hood is an Honorary Associate Professor in the Department of Linguistics at

the University of Sydney. Her research draws on systemic functional linguistics in
studies of academic discourse that explore disciplinary differences in the expression
of knowledge and values, the multimodal cooperation of language and embodied
paralanguage in lectures, and storytelling as a research practice. Recent publications
include Appraising Research: Evaluation in academic writing (2010, Palgrave) and editing
Semiotic Margins: Meaning in multimodalities (with S. Dreyfus and M. Stenglin, 2012,
Continuum). She also co-edited (with Karl Maton and Suellen Shay) Knowledge-
building: Educational studies in Legitimation Code Theory (2016) for the Routledge
LCT series.

Sarah K. Howard is an Associate Professor of Digital Technologies in Education,

SMART Infrastructure Facility Education Group Leader, and a member of the Early
Start Research Institute at the University of Wollongong. Her research focuses on
the use of new methodological approaches to explore teacher change and classroom
practice, specifically related to technology adoption and integration in learning.
As part of this agenda, she uses Legitimation Code Theory to explore cultural and
individual factors of teachers’ digital technology use, educational change and learn-
ing design, with particular interest in subject-area change and underlying principles
of teaching and learning.

J. R. Martin is Professor of Linguistics at the University of Sydney, where he is also

Deputy Director of the LCT Centre for Knowledge-Building. Recent publications
include a book on teaching academic discourse on-line (Genre Pedagogy in Higher
Education, Palgrave Macmillan 2016), with Shoshana Dreyfus, Sally Humphrey and
Ahmar Mahboob; a book on Youth Justice Conferencing (Discourse and Diversionary
Justice, Palgrave Macmillan 2018), with Michele Zappavigna; and a special issue
of Functions of Language (2018) focusing on interpersonal grammar. Eight volumes
of his collected papers (edited by Wang Zhenhua) have been published in China
(2010, 2012; Shanghai Jiao Tong University Press). In April 2014, Shanghai Jiao
Tong University opened its Martin Centre for Appliable Linguistics, appointing
Professor Martin as Director.

Karl Maton is Professor of Sociology at the University of Sydney and Director of the
LCT Centre for Knowledge-Building, as well as Visiting Professor at the University
of the Witwatersrand (South Africa) and Visiting Professor at Rhodes University
(South Africa). Karl’s book Knowledge and Knowers: Towards a realist sociology of educa-
tion is the founding text of Legitimation Code Theory (Routledge, 2014). A col-
lection of studies illustrating how to enact Legitimation Code Theory in research,
Knowledge-building: Educational studies in Legitimation Code Theory (K. Maton, S.
Hood & S. Shay, eds), was published by Routledge in 2016. Karl co-edited Accessing
Academic Discourse (2020, Routledge) with J. R. Martin and Y. J. Doran, illustrating
the productive dialogue between Legitimation Code Theory and systemic func-
tional linguistics.
 Contributors xv

David Rose is Director of Reading to Learn, a literacy program that trains teachers
across school and university sectors in Australia, Africa, Asia and Europe (www.
readingtolearn.com.au). He is an Honorary Associate in the Dept of Linguistics
at the University of Sydney. His research includes analysis and design of classroom
discourse, pedagogic design for embedding literacy in curriculum learning and for
beginning literacy, professional training for teachers in pedagogy and language, lan-
guage typology, language evolution and social semiotic theory. Publications include
The Western Desert Code: An Australian cryptogrammar (2001, Pacific Linguistics),
Genre Relations: Mapping culture (with J. R. Martin, 2008, Equinox) and Learning
to Write, Reading to Learn: Genre, knowledge and pedagogy in the Sydney School (with
J. R. Martin, 2012, Equinox).

Nicky Wolmarans is a Senior Lecturer in the Department of Civil Engineering at

the University of Cape Town. She holds a PhD from the University of Cape Town
and an MSc in Mechanical Engineering. She teaches on the undergraduate pro-
gram as well as having responsibility for academic development. She also teaches
on the postgraduate programs specializing in Engineering Education Research.
Her research interests focus mainly on understanding the nature of professional
knowledge from a realist perspective.
1
THE TEACHING OF SCIENCE
New insights into knowledge, language
and pedagogy

Y. J. Doran, Karl Maton and J. R. Martin

Introduction
Science is significant. Faced with the climate emergency, global pandemics and
proliferating threats to life on Earth, that significance should go without saying.
However, science is under sustained attack by irrationalism in politics, the news
and social media. The need for science education has never been more urgent.
Yet, in many advanced societies, science is struggling to attract and retain students
through school and university. One problem facing attempts at addressing these
issues is a tendency for studies to obscure the knowledge and language that com-
prise science. The subfields of education research dedicated to science disciplines
have contributed greatly to our understanding of scientific ways of knowing. Studies
explore the cognitive resources, perceptions and judgements of students and the
processes of learning. They are extensively examining the conceptions, motivations
and dispositions that students bring to learning science and the ways of thinking
exhibited by students when learning science. However, what students are learning
when they study science, the nature of the scientific knowledge itself, receives far less
attention. Moreover, this dominant focus on how students learn science has been
accompanied by neglect of the teaching of science. Pedagogy is too often reduced
to an afterthought of findings about how students think. Without understanding
how scientific knowledge and language may help shape the best ways of teaching
that knowledge, we have only part of the picture. This volume aims to help fill this
gap by exploring the knowledge practices and multimodal discourses of science
teaching.
The collection extends two approaches that make pedagogic discourse a cen-
tral object of study in education: Legitimation Code Theory (LCT) and Systemic
Functional Linguistics (SFL). These approaches bring to light the meaning-making
activities of social actors, in complementary ways. LCT is a sociological framework that
foregrounds the knowledge practices of science education, revealing features such as
2 Y. J. Doran, Karl Maton and J. R. Martin

their complexity, context-dependence, boundedness and specialized procedures. SFL

explores the language and other semiotic resources, including mathematics, images and
gesture, through which these knowledge practices are expressed. In recent years schol-
ars and educators using these two frameworks have been closely working together,
generating a fast-growing body of work that draws on both sides of this genuinely
interdisciplinary dialogue (e.g. Martin and Maton 2013, Martin et al. 2020b).This book
builds on this collaboration to offer cutting-edge developments in both approaches
that will generate major advances in not only how science education is understood but
also how knowledge, language and pedagogy are conceived more widely.
Teaching Science is organized into three parts. Part I draws on LCT to explore
how science teaching can support knowledge-building. A series of innovative stud-
ies focus on the integration of mathematics into science (Chapter 2), the building
of scientific explanations (Chapter 3) and the integration of multimedia such as
animations into science teaching (Chapter 4).These studies introduce new concepts
and new methods in LCT, including ‘autonomy tours’ (Chapter 2), ‘constellation
analysis’ (Chapter 3) and ‘epistemic affordances’ (Chapter 4).
Part II greatly extends SFL to explore the multimodal discourses that underpin
knowledge-building in science classrooms. These chapters articulate the expansive
range of meanings involved in explaining complex scientific phenomena (Chapter
5), the ways in which deep scientific taxonomies are built (Chapter 6), and the essen-
tial interdependence of language, mathematics and images in scientific knowledge
(Chapter 7). They advance the modelling of meaning-making in SFL into new areas
and articulate extensions to concepts that will kickstart new forms of research into sci-
ence discourse, including ‘field’ (Chapter 5), ‘ideational discourse semantics’ (Chapter
6) and the multimodal analysis of language, mathematics and image (Chapter 7).
Part III explores the practical implications that LCT and SFL analyses can deliver
for teaching science. Chapters discuss how access to scientific knowledge can be wid-
ened to a greater diversity of students (Chapter 8), how knowledge is transformed to
address real-world problems in engineering design (Chapter 9), the meanings com-
municated in live lectures that are rarely taken into account in debates over peda-
gogy (Chapter 10), and how mathematics can be taught effectively to all students
through the influential pedagogic program ‘Reading to Learn’ (Chapter 11).
In this opening chapter we introduce the book by outlining the traditions of
studies enacting the frameworks of LCT and SFL to examine science education.
First, we locate this volume in the long-standing body of work using SFL to under-
stand talking, writing and reading science. Second, we turn to the more recent but
growing body of research and practice enacting LCT to examine and shape teaching
and learning science. Finally, we introduce the chapters of this book, highlighting
how they offer new ways of understanding, analyzing and shaping science teaching.

The language of science

Systemic Functional Linguistics (SFL) is a major linguistic approach that has devel-
oped worldwide since the 1960s. SFL explores how language is used in social life
 The teaching of science 3

to organize our interpersonal relations, manage the content meanings we want to

express and bring this together to make coherent text. Scholars using the approach
consistently emphasize the need to study language in a way that contributes to
improving the world. This emphasis on ‘appliable’ linguistics means SFL engages
with a very wide range of contexts and uses of language. One of its longest engage-
ments has been with scientific discourse. At first this was not to examine science
itself but rather to understand English grammar (Huddleston et al. 1968). Since the
1980s, however, SFL engaged more deeply with scientific texts educationally, from
the perspective of teaching literacy in science. At the time, researchers were dissatis-
fied with prevailing approaches to literacy that emphasized writing without explic-
itly modelling the language patterns that students need to learn. In response, SFL
scholars began developing ways of explicitly teaching differences in writing across
subject areas, an approach which became known as ‘genre pedagogy’ (Rothery
1989, Rose and Martin 2012).
However, to teach the specific language patterns used in different subjects, edu-
cators needed to know those patterns. This necessitated a large-scale descriptive
effort to map language features across the disciplinary spectrum, much of which
was funded by the Metropolitan East Disadvantaged Schools Program in Sydney,
Australia. A key focus of this work was to analyze the language of science using
concepts from across the SFL framework. An early concern was with text types
and text structures of science, conceptualized in terms of genre. This work is illus-
trated by the book Factual Writing in which Martin (1985) presented descriptions of
different genres in primary school scientific texts, including reports that generalize
experiences about things, procedures that focus on how to make things happen, and
explanations that explain things.1
Through the late 1980s and 1990s these models of genre in science were greatly
expanded as SFL researchers engaged with later primary school, secondary school
and workplaces associated with science and technology (e.g. Rose et al. 1992,
Unsworth 1997a, 1997b, 1997c, Veel 1992, 1997). This work formed part of major
innovations in SFL as a framework that transformed linguistic modelling of not
only scientific discourse but also language more broadly. Many of these innova-
tions are brought together in a series of landmark books on talking, writing and
reading science. Talking Science (Lemke 1990) engaged in depth with language and
meaning-making in science education for the first time. Writing Science (Halliday
and Martin 1993) offered significant advances in the grammatical modelling of sci-
entific language and insights into the nature of scientific meaning and technicality
from the perspective of the register variable field. Reading Science (Martin and Veel
1998) collected papers significantly expanding the reach of SFL research into sci-
ence, including popular science, environmental discourse, and both pedagogic and
workplace settings, as well as exploring features of scientific discourse such as gram-
matical metaphor and multimodality. These books laid foundations for subsequent
SFL research into science and fed back into the development of pedagogic programs
through SFL. Many of these programs have aimed at designing classroom practices
and resources for improving the teaching of both the content of any particular
4 Y. J. Doran, Karl Maton and J. R. Martin

subject area as well as the literacy practices so necessary for organizing this content.
This has included developing frameworks for teaching students language about lan-
guage (‘metalanguage’) so they can more easily talk and reflect upon their reading
and writing, and designing classroom practices that enable a gradual handover of
scientific ways of reading and writing from teachers to students (Unsworth 2001,
Christie and Derewianka 2008, Rose and Martin 2012, Derewianka and Jones 2016,
Dreyfus et al. 2016, Humphrey 2017). These programs have been internationally
influential, underpinning among other things, the current Australian literacy curric-
ulum. Underpinning this book, the interaction between SFL research into scientific
discourse and its pedagogic development highlighted three key areas of ongoing
exploration: the grammar of scientific language, technical meanings in science and
the multimodal nature of science.

The grammar of scientific language

In Writing Science, Halliday (1993a) highlighted significant differences in the lan-
guage used in science compared to that of other registers. Scientific language, he
summarized, involves:

• interlocking definitions, where technical terms are mutually defining;

• technical taxonomies, where terms are ordered into complicated layers of clas-
sification and composition;
• special expressions, where large multi-word constructions are technical, rather
than simply individual words;
• lexical density, where there is a relatively high number of lexical words per
clause;
• grammatical metaphor, where a grammatical structure is ‘mismatched’ with its
meaning, such as when the clause ‘how quickly cracks in glass grow’ is recon-
strued as the nominal group ‘glass crack growth rate’;
• syntactic ambiguity, where grammatical forms produced by grammatical meta-
phor lead to ambiguity in their meaning; and
• semantic discontinuity, where steps in the logic of reasoning or specific links con-
necting technical meanings are presumed or not made explicit.

These individual language features, Halliday points out, are regular occurrences in
scientific texts and perform distinct functions in the organization of technical mean-
ing. Through the history of scientific discourse, these grammatical features evolved
in conjunction with scientific knowledge. To construe new meanings of scientific
thought, scholars such as Chaucer, Newton, Maxwell and Darwin required new
forms of English. Over time, these ‘new forms of English’ became the dominant
motif for scientific texts (Halliday 1993b, 1993c, Banks 2008, 2017). However, when
a number of these language features are used together in the same text, they can
make scientific language difficult for learners – they are ‘grammatical problems’.
 The teaching of science 5

In terms of linguistic modelling, the grammar of science raised questions regard-

ing the nature of language in general. One key issue in particular was the nature of
grammatical metaphor (Simon-Vandenbergen et al. 2003). Grammatical metaphors
involve meanings being reconstrued in an ‘incongruent’ manner. For instance, in
the example given by Halliday noted above, ‘how quickly cracks in glass grow’ is
reconstrued as a grammatical metaphor through the nominal group ‘glass crack
growth rate’. Metaphors such as these were shown to be key in moving into the
language of school science and other subjects (Derewianka 2003). But how one
can show that something is a grammatical metaphor rather than a ‘normal’ realiza-
tion of a meaning was less clear (Halliday 1998). For example, in biology the word
‘phagocytosis’ is a noun, yet it seems to construe an ‘event’ in some sense because
it names the process whereby pathogens in the body are engulfed and destroyed.
However, unlike ‘glass crack growth rate’, it cannot easily be turned back into a
more ‘congruent’ form – it is rare, if indeed it happens at all, for biology textbooks
to talk about ‘phagocytosing’ as a verb.These more theoretical and descriptive ques-
tions led to a refocus on how SFL could model semantics in relation to grammar.
Among other things they influenced the highly elaborated model of ideational
semantics put forward by Halliday and Matthiessen (1999), which has in turn laid
a platform for the discourse semantic modelling of Hao (2020b, Chapter 6 of this
volume) and has shed further light on how grammatical metaphors work (Hao
2020a, Martin 2021).
For educational research, this work gave a sense of the language patterns students
needed to write. However, it was clear that many subject areas valued getting the
technical ‘content’ meanings correct over other linguistic patterns. Although con-
cerns for genre and grammar were important, this was only to the extent that they
were used in service of the technical meanings of science. This raised the question
of what a ‘successful’ text in science education is, and more broadly how to model
scientific meanings in general.

Technical meanings in science

To understand how technical meanings are organized in science, the grammatical
and genre perspectives on scientific language in SFL were coupled with a perspec-
tive of field within SFL. Field is a variable of register (following the stratal model of
Martin, 1992) which is realized by ideational meaning in language and notionally
organizes the ‘content’ meanings of discourse. In the 1980s, one of the key issues for
SFL scholars was the large taxonomies of technicality in science with which stu-
dents must engage (a field perspective on one of Halliday’s ‘grammatical problems’).
This required renovating existing SFL descriptions of field to make a series of key
distinctions. First, taxonomies were modelled as being either of classification, relat-
ing things in terms of type–subtype, or of composition, relating things in terms of
part–whole relations. Different text types emphasized different types of taxonomy,
and as Wignell et al. (1993) argued in Writing Science, different fields could also be
6 Y. J. Doran, Karl Maton and J. R. Martin

distinguished by exploring their taxonomies. For example, in the specialized field of

bird watching, taxonomies of birds tend to be more elaborated than those typically
used in everyday life. However, these taxonomies share the fact they both tend to
be based on observable characteristics of birds, such as colour, markings, size, habi-
tat, etc. In contrast, the scientific taxonomies of ornithology, in addition to being
significantly more elaborated, are based on shared lineage shown through genetic
comparison and requiring specialist equipment, rather than observable characteris-
tics.This renovated model enabled texts to be explored in terms of the field-specific
meanings they presented. However, a key issue that remained was how language is
used to build these taxonomies. Chapter 6 of this volume (Hao) takes a major step
forward by laying out an extended model of ‘ideational discourse semantics’ for
understanding how highly elaborated taxonomies are built in biology.
A second distinction concerned the intricate sequences of events known as activi-
ties that often occur in explanations and procedures. In contrast to many fields,
activities in scientific fields are often linked by ‘implication’ where an activity nec-
essarily implicates another activity. An example of this from Wignell et al. (1993)
concerns convective uplift:
Convective uplift
Air in contact with a warm surface will become heated and expand causing it
to rise. Dew point will be reached, condensation will take place and convec-
tional clouds will form.

This text presents a series of activities that are related by implication, where one
necessarily follows another (here ^ indicates sequence):

Air in contact with a warm surface will become heated

^
and expand
^
causing it to rise
^
Dew point will be reached
^
condensation will take place
^
and convectional clouds will form

Each step in this series necessarily occurs because of the previous step. This con-
trasts with series of activities in genre such as stories where events tend to occur
in a less definite fashion – things can ‘go wrong’ and unexpected events may occur
(indeed this is one of the key features of stories, Martin and Rose 2008). As the text
above shows, these activities may include technical terms (dew point, condensation,
convectional clouds) that may be positioned within the taxonomies of scientific
discourse, while the whole sequence itself is named technically as ‘convective uplift’.
Such naming of activities, alongside elaborated taxonomies and extensive use of
 The teaching of science 7

grammatical metaphor, was shown in Writing Science to underpin the enormous

technicality of science.
This model of field has informed decades of research into scientific discourse
and has proven both a relatively intuitive means of understanding technical meaning
and one that can be linked with language patterns. However, it still did not account
for many areas of scientific meaning, including the various gradable properties (size,
weight, force, etc.) that often distinguish items in taxonomies and in later years of
secondary school are often realized through mathematics and graphs, nor how the
various aspects of field (taxonomy and activity) could be integrated into a coherent
‘whole’ to explain or describe phenomena. Taking these issues as its point of depar-
ture, Doran and Martin (Chapter 5, this volume) directly build upon this model
of field to encapsulate a wider range of scientific meaning-making that has been
highlighted by research in recent years.

Multimodality in science
Running parallel to studies of grammatical, genre and field-based attributes of sci-
ence has been a focus on the multimodal nature of its discourse. Books such as
Reading Images (Kress and van Leeuwen 1990) opened the way for linguists to
engage with semiotic resources beyond written language. The work of Kress and
van Leeuwen (1996) on scientific diagrams, for example, showed that images were
organized in highly conventional ways that could be analyzed in terms comparable
to both the field and grammatical modelling being developed for language. This
highlighted the importance of multimodal meaning making in science education
and in broader schooling contexts, and heavily influenced the turn towards ‘mul-
tiliteracies’ in education. School literacy is no longer considered just reading and
writing language but also viewing, drawing and organizing images, and a range of
other meaning-making systems (New London Group 1996, Unsworth 2001, Kress
et al. 2001). Understanding how individual semiotic resources enable this to happen
became a key question for SFL multimodality researchers. Work on images along
these lines has continued steadily since the 1990s, accelerating recently in relation to
the role graphs play in science (Doran 2018a, 2019, Chapter 7 of this volume), the
significant degree of meaning organized into scientific diagrams (Unsworth 2020,
Martin et al. 2021, Doran 2019), the use of certain types of image as technical for-
malisms across disciplines (Yu in press, Doran 2020), and how images are dynamized
into animations for teaching science (He 2020).
In addition to images, the advent of multimodality encouraged further develop-
ments for other semiotic resources. For science a key step forward in this regard was
O’Halloran’s (2005) work on mathematics, which began to fill a major gap in our
understanding of scientific discourse. Along with work by Lemke (1998, 2003), this
work emphasized that multimodal texts were more than the ‘sum of their parts’.
Although each semiotic resource organized its own meaning, these meanings were
‘multiplied’ (Lemke 1998) when used with other semiotic resources in a text to
make new meanings not available to any individual resource. O’Halloran’s study
put forward for the first time an operationalizable description of how different
8 Y. J. Doran, Karl Maton and J. R. Martin

semiotic resources of language, mathematical symbolism and images worked not

only individually but also in terms of their interactions or ‘intersemiosis’. This focus
on intersemiosis has laid the groundwork for ongoing work in systemic functional
work on science education. For example, Doran’s (2018a, 2018b) model of math-
ematical symbolism built on O’Halloran’s work to develop a fully systematized
grammatical description of mathematical symbolism used in secondary school and
university physics. It also explored the interaction between language and mathemat-
ics in terms of the genres these two resources realized together. This in turn raised
theoretical questions about the modelling of semiotic resources outside of language.
In particular Doran questioned the general assumption that the three metafunctions –
ideational, interpersonal and textual – occur for all semiotic resources (e.g. Kress and
van Leeuwen 1996, O’Halloran 2005). Doran (2018b) argued that when descrip-
tions did not assume it, there was no evidence for identifying an interpersonal
metafunction in mathematical symbolism. This work on symbolisms is currently
being extended by Yu (in press), focusing on chemistry discourse, which is helping
to open space for engagement with the highly technical texts of upper secondary
school science. A number of chapters in this volume directly extend SFL work
on multimodality, including Chapter 7 (Doran) on the interaction of mathemat-
ics, images and language in physics, Chapter 10 (Hao and Hood) on interactions
between body language and language in chemistry lectures, and Chapter 11 (Rose)
on teaching mathematics.
Alongside the aforementioned advances in understanding grammar and tech-
nicality, this move towards multimodality helped transform the SFL understanding
of science between the late 1980s and early 2000s. New models of language and
semiosis were generated that opened the way for new objects of study and drove
new pedagogic applications. More recently, a further transformation has been tak-
ing place in how SFL approaches science education, one sparked by an ongoing
dialogue between SFL and Legitimation Code Theory.

The knowledge of science

Legitimation Code Theory (LCT) is a framework for analyzing and shaping
practices.2 LCT integrates and extends insights from a range of influences but
most explicitly builds on the sociological frameworks of Pierre Bourdieu and
Basil Bernstein (see Maton 2014, 2018). Like SFL, LCT is concerned with the
meaning-making activities of social actors. As its name implies, LCT provides tools
for exploring the bases of legitimacy or ‘rules of the game’ in social fields, in ways
that reveal the organizing principles underlying practice. Metaphorically, LCT gets
at the DNA of practice.
Reflecting its sociological foundations, LCT views society as comprising rela-
tively autonomous social fields of practice (such as law, medicine, education, etc.)
that have distinctive resources and forms of status. In each social field, actors coop-
erate and struggle, both for more of what is viewed as signs of success and over
what defines success in that social field. Their practices thus embody messages as to
 The teaching of science 9

what should be dominant measures of achievement. This is to highlight that there

is more to what we say or do than what we say or do. For example, if a teacher has
a class undertake a practical experiment, they are teaching not only whatever sci-
entific content the experiment imparts but also that engaging in concrete, tangible
activities is important and that students discovering results by themselves is impor-
tant. LCT reveals these kinds of messages by analyzing the organizing principles of
practice.
These organizing principles can be manifold. Any set of practices has a diverse
range of characteristics, such as their complexity, context-dependence, emphasis on
specialized knowledge or personal experience, boundedness from other practices,
and so forth. The organizing principles underlying practices are conceptualized by
LCT as different species of legitimation code. The conceptual framework is structured
into a series of dimensions or sets of concepts that each explore a distinctive species
of legitimation code. There are currently three active dimensions – Specialization,
Semantics and Autonomy – centred on exploring specialization codes, semantic
codes and autonomy codes, respectively.3 Put simply, Specialization explores how
knowledge and knowers are articulated within practices, Semantics explores ques-
tions of context and complexity, and Autonomy explores where contents and
purposes of practices come from. (On Specialization, see Maton 2014, 2020; on
Semantics, see Maton 2013, 2014 and Maton and Chen 2020; on Autonomy, see
Maton and Howard 2018, 2020, and Chapters 2 and 4 of this volume. For how LCT
concepts relate together, see Maton 2016). These different dimensions do not refer
to different sets of empirical practices but rather offer ways of revealing different
organizing principles underlying the same set of practices. How many and which
dimensions are drawn on in empirical research depends on the problem-situation
(specific questions concerning a particular object of study).

LCT and science education

Scholars and educators are enacting LCT to examine and shape practices across
the disciplinary map and in all kinds of educational institutions, as well as beyond
education (e.g. Maton et al., 2016a; Winberg et al. 2021). The framework is widely
applicable and studies of topics far beyond one’s own substantive areas of concern
can offer insights. For example, in developing the concept of ‘constellations’ to ana-
lyze the teaching of scientific explanations, Maton and Doran (Chapter 3, this vol-
ume) found constellation studies of ballet lessons, History courses and ethnopoetics
all highly valuable. Thus, insights into science education offered by LCT research
are not confined to studies expressly focused on that topic. Nonetheless, there is a
growing body of work dedicated to exploring the teaching and learning of science
using different dimensions of LCT either separately or in combination.
The dimension of Specialization is proving particularly valuable for exploring
how knowledge and knowers come together in science education. For example,
concepts from Specialization have opened up new ways of thinking about support-
ing ‘epistemological access’ to science knowledge for students from diverse social
10 Y. J. Doran, Karl Maton and J. R. Martin

backgrounds (Ellery 2018, Chapter 8 of this volume). They are also providing prac-
tical tools for supporting engagement with decolonization in science education
in ways that respect different ways in which knowledges and knowers are valued
(Adendorff and Blackie 2021). Focusing on the ‘knowledge’ side, an interlocking
series of studies led by Karin Wolff (e.g. Pott and Wolff, 2020, Wolff 2021) draw
on the epistemic plane (Maton 2014: 171–195) to explore how teaching and learn-
ing emphasize in different ways the specialized procedures of engineering and/
or the phenomena for which they are used. Combined with studies of issues from
work-integrated learning (Winberg 2012) to student design projects for ‘real world’
problems (Wolmarans, Chapter 9, this volume), this work is building a sophisticated
picture of engineering education.
The dimension of Semantics directly resulted from two major studies of teach-
ing practices in secondary school science (see Maton 2020). Specifically, this work
introduced the notion of semantic waves, which describes recurrent shifts in the
context-dependence and complexity of knowledge, and the method of semantic pro-
filing those changes over time (Maton 2013, Macnaught et al. 2013). These ideas are
being successfully enacted across science education, including to support cumulative
learning in chemistry (Blackie 2014), student transition from school to university
biology (Mouton and Archer 2019), project-based learning in biology (Mouton
2019) and in chemistry (Veale et al. 2017), real-world applications in chemical engi-
neering (Dorfling et al. 2019), student design practices in engineering (Wolmarans
2016), problem-solving in physics (Conana et al. 2021), and student assessments in
chemistry (Rootman-le Grange and Blackie 2018) and physics (Georgiou et al.
2014; Steenkamp et al. 2019). Extending these concepts into exploring multimodal-
ity is also a burgeoning focus, such as the role of language, mathematics and image
in physics (Doran 2018a).
Studies of science education have also played a crucial role in developing new
concepts in the dimension of Autonomy. Research into teaching in science class-
rooms helped generate the notion of autonomy tours (Maton and Howard 2018),
which shows how to successfully integrate science knowledge with other content
and purposes, such as everyday experiences, metaphors, analogies and knowledge
from other subjects. These are providing new ways of understanding how to suc-
cessfully integrate mathematics and multimedia objects in science teaching (Maton
and Howard, Chapters 2 and 4 of this volume).

Dialogue with SFL

A genuinely inter-disciplinary dialogue between LCT and SFL has been underway
since the turn of the century. This collaboration built on and intensified previous
discussions between Basil Bernstein (whose theory is a foundational framework
for LCT) and Michael Halliday and Ruqaiya Hasan. This dialogue went through
a series of phases (see Maton and Doran 2017). In the late 1990s, Bernstein’s ideas
were inspiring SFL scholars to think about knowledge structures in education
(Martin et al. 2020a). The emergence of LCT in the early 2000s offered to SFL,
 The teaching of science 11

among other things, a means of engaging more empirically with knowledge prac-
tices, in both research and teaching. This began a series of more intense phases of
direct inter-disciplinary collaboration between LCT and SFL scholars. One result
is a rapidly growing number of papers, books and doctoral theses that use both
frameworks together to generate greater explanatory power (e.g. Martin et al. 2020b;
Maton et al. 2016). In this volume, for example, Doran (Chapter 7) engages with
both the LCT dimension of Semantics and the SFL register variable field to examine
how language, image and mathematics come together in school physics. Another
result is an ongoing series of innovations in each framework. For example, the LCT
concepts of ‘semantic gravity’ and ‘semantic density’, which explore the context-
dependence and complexity of knowledge practices, were in part stimulated by
dialogue with SFL scholars. In turn, these LCT concepts provoked the development
of new SFL concepts of ‘presence’ (Martin and Matruglio 2020) and ‘mass’ (Martin
2020), which bring together the many ways these issues are manifested in language.
These new concepts have begun to show great promise in gathering together mani-
fold linguistic resources in studies of science education (Hood and Hao, Chapter
10, this volume).
More generally, this dialogue with LCT has evoked in SFL research a grow-
ing interest in the role of linguistic and semiotic resources in knowledge-building.
This focus is resulting in major theoretical advances. One recent example is Hao’s
development of ideational discourse semantics (2018, 2020a, 2020b), as illustrated
by Chapter 6 of this volume. Building on Halliday and Matthiessen’s (1999) descrip-
tion of ideational semantics, Hao has developed a model for exploring the ways that
technical knowledge is built through unfolding discourse. This addresses an issue
that had long vexed SFL scholars, with previous attempts at understanding discourse
semantics regularly becoming blurred into either grammatical description or field-
based description. Hao’s model (Chapter 6) offers a clear and distinct descriptive
level that enables a view of scientific language on its own terms. A second example,
which complements Hao’s work, is a newly extended model of field by Doran and
Martin, as shown in Chapter 5 of this volume. Taken together with the expansive
grammar of English put forward by Halliday (Halliday and Matthiessen 2014) and
the models of genre developed for science (see Martin 1985, Martin and Rose 2008),
this new period of research has enabled SFL to greatly expand our understanding of
the rich and multifaceted language and multimodality inherent in science.

On Teaching Science
This volume offers major steps forward in how both LCT and SFL conceive sci-
ence and science education. The title consciously echoes past landmark works in
SFL, with Teaching Science capturing the principal focus of and stimulus for this
new work. The subtitle, Knowledge, language, pedagogy, sets out the organization of
the book into three main parts that explore: knowledge-building in teaching sci-
ence; the multimodal discourses that underpin this knowledge-building; and how
to improve the teaching and learning practices of science.
12 Y. J. Doran, Karl Maton and J. R. Martin

Part I comprises three chapters that develop new ideas in LCT to understand
major issues in teaching science: the integration of mathematics (Chapter 2), the
teaching of scientific explanations (Chapter 3) and the use of multimedia digital
resources (Chapter 4). These chapters build on one another. Chapter 2 (Maton
and Howard) addresses the vexing question of how mathematics can be success-
fully integrated in science teaching. That students often struggle with mathematics
in science lessons, even when they have little difficulty with those ideas in math-
ematics lessons, has been a long-running concern for educators. One reason this
issue remains unsolved is that existing approaches cannot systematically distinguish
‘mathematics’ knowledge from ‘science’ knowledge. Maton and Howard introduce
cutting-edge tools from the LCT dimension of Autonomy that enable knowledge
practices to be distinguished without lapsing into either essentialist definitions that
neglect how these bodies of knowledge differ between contexts or relativist claims
that they are nothing but endless flux. The concepts are illustrated through detailed
analyses of real-world classroom practices that show a key attribute of success-
ful integration of mathematics in science teaching to be autonomy tours that shift
between different contents and purposes in particular ways. The ideas outlined here
are poised to have a major impact on both research and practice in education, far
beyond science teaching.
Chapter 3 (Maton and Doran) focuses on the role played by relations among
ideas in teaching scientific explanations. Reflecting their knowledge-blindness,
dominant approaches to researching science education neglect the ways in which
ideas are connected to create explanations. Maton and Doran introduce the method
of constellation analysis from LCT as a way of revealing these relations among ideas.
This innovative method is used to analyze explanations of the tides and seasons. In
each case the logic of explanations presented in school textbooks is analyzed and
compared to how the explanation is taught in a classroom. These analyses show
that explanations of seemingly similar kinds of phenomena differ in terms of how
ideas are related together and that the logic of these relations impacts on how they
are taught in classrooms. Constellation analysis offers a new analytic method with
huge potential as a practical tool for researchers, curriculum designers, educators
and students.
Chapter 4 (Maton and Howard) examines how multimedia such as anima-
tions can be integrated into teaching science. Existing research overwhelmingly
focuses on developing principles for designing multimedia that support cognitive
processing of information. This chapter meets an urgent need to foreground teach-
ing and knowledge as a first step towards developing pedagogic principles for teach-
ing multimedia that support learning science. To do so, the chapter extends existing
limited uses of the notion of ‘affordances’ to examine how the knowledge practices
expressed by multimedia relate to those central to specific classroom tasks. These
epistemic affordances are revealed through an innovative form of autonomy analysis
that shows how the diverse elements of complex multimedia objects relate to the
contents and purposes of specific classroom tasks. In-depth analyses of two con-
trasting examples of science teaching with animations show the pedagogic work
 The teaching of science 13

required of teachers to integrate such multimedia and how LCT offers a way of
getting to grips with these complex objects in real-world contexts.
Part II comprises detailed studies of the language and multimodal discourse of
science teaching using newly-developed tools from Systemic Functional Linguistics.
These chapters articulate the expansive range of meanings involved in explaining
complex scientific phenomena (Chapter 5), the ways deep scientific taxonomies are
built (Chapter 6), and the interdependence of language, mathematics and images in
building scientific knowledge (Chapter 7).These three chapters push the modelling
of language and meaning-making in SFL into new territory and articulate theoreti-
cal principles that will enable new forms of research into science discourse. They
complement the LCT analyses of Part I to significantly expand our understanding
of scientific knowledge and semiosis and how these are taught.
Chapter 5 (Doran and Martin) introduces an evolving model of the register
variable ‘field’ for understanding the intricate explanations built in science. This
explores how science can view phenomena from a static perspective in terms of
elaborated taxonomies or a dynamic perspective in terms of unfolding activities.
Doran and Martin also explore how large swathes of gradable properties can be
arrayed and measured, and how all of these can be reconstrued and interconnected
for any particular ‘topic’ in science. This description takes a big step forward in
building upon decades of modelling of field in SFL and works to be generalizable
across language, image, mathematics and a wide range of other semiotic resources.
This new model of field is being developed in dialogue with a new model of
ideational discourse semantics that informs Chapter 6 (Hao). This explores scien-
tific meaning as it unfolds in text to create an intricate discourse semantic frame-
work for grasping the elaborate taxonomies built in biology, and the various entities
they marshal. Chapter 6 forms part of a larger discourse semantic model developed
by Hao (2018, 2020a, 2020b) that significantly pushes forward knowledge of how
scientific language organizes its technical meaning. Together with Chapter 5, this
offers for the first time an expanded resource for modelling ideational meaning that
links lexicogrammar, discourse semantics, field and genre.
Chapter 7 (Doran) explores scientific meaning as inherently multimodal.
Focusing in particular on the interaction of mathematics, images and language in
physics, Doran shows how each builds technical meaning in ways that can ‘hand over’
this meaning to others. Drawing on the SFL model of field developed in Chapter
5 and the dimension of Semantics from LCT, this chapter shows how semiotic
resources are brought together to move from relatively common-sense meanings to
technical, while at the same time moving between the empirical and the theoretical.
Part III exemplifies the close connection between research and practice that
characterizes LCT and SFL research into science teaching. These chapters draw on
LCT and/or SFL to explore how more students can be supported to access scien-
tific knowledge (Chapter 8), the vexed problem of simplifying ‘real world’ problems
to teach professional reasoning in engineering education (Chapter 9), the role of
body language in face-to-face chemistry lectures (Chapter 10), and methods for
teaching mathematics to enable all students to learn (Chapter 11).
14 Y. J. Doran, Karl Maton and J. R. Martin

Chapter 8 (Ellery) examines a science foundation course that is intended to

support more students to successfully engage with learning science at university.
Enacting the Specialization dimension of LCT to explore curriculum, staff beliefs
and student experiences, Ellery shows that the course involves two different bases
of achievement. One basis requires students to demonstrate their understanding of
scientific knowledge, the other requires students to be a particular kind of scientific
knower. Moreover, students must be the right kind of knower in order to access the
right kinds of knowledge. Ellery uses LCT to dig beneath the surface of curricular
intentions and show how educational practices may contradict those intentions.The
chapter poses a significant challenge to current thinking about how more students
can be supported to access science in education.
Chapter 9 (Wolmarans) problematizes the widely-held view of professional
education, such as engineering, as learning how to apply disciplinary science
knowledge to ‘real world’ problems. Enacting concepts from the Semantics and
Specialization dimensions of LCT, Wolmarans analyzes design projects in a civil
engineering course that are intended to authentically mimic professional prob-
lems. Such projects require simplification of the problem for students. The chapter
pushes LCT analysis to distinguish between reduced complexity of the knowledge
required by students to solve a problem and reduced complexity of the problem
itself. Crucially, the analysis shows that simplifying ‘real world’ projects requires care-
ful negotiation between how the problem and the specialized scientific knowledge
are simplified or else problems ensue for teaching professional reasoning. In short,
the chapter shows how teaching engineering is more than simply ‘applying’ science
knowledge to ‘real world’ problems.
Chapter 10 (Hood and Hao) explores the rich meanings made by body lan-
guage in face-to-face lectures, drawing on a developing model of paralanguage in
SFL (Cléirigh 2011, Hood 2011, Martin and Zappavigna 2019, Ngo et al. 2021).
Focusing on university lectures in chemistry, Hood and Hao show the extensive
means through which technical meaning is distributed across language and body
language. In particular, they draw on Hao’s model of ideational discourse semantics
(Chapter 6, this volume) to detail how the scientific construction of the world in
spoken language is regularly coupled with a more ‘common-sense’ construction in
body language, thereby grounding the technical meanings of science. Coming at
a time where classes are being increasingly moved online, this chapter offers key
insights into what can be lost when we move away from face-to-face teaching.
Chapter 11 (Rose) extends the internationally influential SFL pedagogic pro-
gramme ‘Reading to Learn’ into the teaching of mathematics. Rose begins from the
straightforward yet significant observation that, when teaching, the symbols of math-
ematics are typically written down on the board but the procedure of how to do the
mathematics is spoken aloud. The procedure is thus potentially lost to student at the
moment of being taught. To address this, Rose lays out a principled pedagogic pro-
gramme that makes explicit the procedures involved in solving mathematical problems,
exemplified through a secondary school lesson on trigonometry. This involves a series
of tasks that progressively build written procedures for doing mathematics and gradually
 The teaching of science 15

hands over control from teacher to students. As with all aspects of the Reading to Learn
programme, this method is developed to ensure that all students have access to math-
ematical knowledge in ways that are accessible and deliverable in the classroom.
The chapters of this collection will launch new research agendas in understand-
ing knowledge, language and pedagogy in science. They offer major leaps forward
in both LCT and SFL, with implications far beyond teaching science.

Notes
1 This book also laid out models of a range of argumentative expository genres, recounts
and descriptions, which are more typically found in the humanities and social sciences.
2 For LCT papers, blogs and events, see: www.legitimationcodetheory.com.
3 A fourth dimension, Temporality, is under fundamental redevelopment.

References
Adendorff, H. and Blackie, M. (2021) ‘Decolonizing the science curriculum: When good
intentions are not enough’, in C. Winberg, S. McKenna and K. Wilmot (eds) Building
Knowledge in Higher Education, London: Routledge, 237–54.
Banks, D. (2008) The Development of Scientific Writing: Linguistic features and historical context,
London: Equinox.
Banks, D. (2017) The Birth of the Academic Article: Le Journal des Sçavans and the Philosophical
Transactions, 1665–1700, London: Equinox.
Blackie, M. (2014) ‘Creating semantic waves: Using Legitimation Code Theory as a tool to
aid the teaching of chemistry’, Chemistry Education Research and Practice, 15: 462–9.
Christie, F. and Derewianka, B. (2008) School Discourse, London: Continuum.
Cléirigh, C. (2011) ‘Gestural and postural semiosis: A systemic functional linguistics approach to “body
language”’, Unpublished manuscript.
Conana, H., Marshall, D. and Case, J. (2021) ‘A semantics analysis of first year physics teach-
ing: Developing students’ use of representations in problem-solving’, in C. Winberg,
S. McKenna and K. Wilmot (eds) Building Knowledge in Higher Education, London:
Routledge, 162–79.
Derewianka, B. (2003) ‘Grammatical metaphor and the transition to adolescence’, in
A. Simon-Vandenbergen, M. Taverniers and L. J. Ravelli (eds) Grammatical Metaphor,
Amsterdam: John Benjamins, 185–220.
Derewianka, B. and Jones, P. (2016) Teaching Language in Context, Melbourne: Oxford
University Press.
Doran, Y. J. (2018a) The Discourse of Physics: Building Knowledge through Language, Mathematics
and Image, London: Routledge.
Doran, Y. J. (2018b) ‘Intrinsic functionality of mathematics, metafunctions in systemic func-
tional semiotics’, Semiotica, 225: 457–87.
Doran, Y. J. (2019) ‘Building knowledge through images in physics’, Visual Communication,
18(2): 251–77.
Doran, Y. J. (2020) ‘Academic formalisms: Toward a semiotic typology’ in J. R. Martin,
Y. J. Doran and G. Figueredo (eds) Systemic Functional Language Description, London:
Routledge, 331–58.
Dorfling, C.,Wolff, K. & Akdogan, G. (2019) ‘Expanding the semantic range to enable mean-
ingful real-world application in chemical engineering’, South African Journal of Higher
Education, 33 (1): 42–58.
16 Y. J. Doran, Karl Maton and J. R. Martin

Dreyfus, S., Humphrey, S., Mahboob, A. and Martin, J. R. (2016) Genre Pedagogy in Higher
Education:The SLATE project, Basingstoke: Palgrave Macmillan.
Ellery, K. (2018) ‘Legitimation of knowers for access in science’, Journal of Education, 71: 24–48.
Georgiou, H., Maton, K. and Sharma, M. (2014) ‘Recovering knowledge for science educa-
tion research: Exploring the “Icarus effect” in student work’, Canadian Journal of Science,
Mathematics, and Technology Education 14(3): 252–68.
Halliday, M.A. K. (1993a) ‘Some grammatical problems in scientific English, in M.A. K. Halliday
and J. R. Martin (eds) Writing Science, London: Falmer, 76–94.
Halliday, M. A. K. (1993b) ‘On the language of physical science’ in M. A. K. Halliday and
J. R. Martin (eds) Writing Science, London: Falmer, 59–75.
Halliday, M. A. K. (1993c) ‘The construction of knowledge and value in the grammar of
scientific discourse’, in M. A. K. Halliday and J. R. Martin (eds) Writing Science, London:
Falmer, 95–116.
Halliday, M. A. K. (1998) ‘Things and relations: Regrammaticizing experience as technical
knowledge’, in J. R. Martin and R.Veel (eds) Reading Science, London: Routledge, 185–236.
Halliday, M. A. K. and Martin, J. R. (1993) Writing Science: Literacy and Discursive Power,
London: Falmer.
Halliday, M. A. K. and C. M. I. M. Matthiessen (1999) Construing Experience Through Meaning:
A Language-Based Approach to Cognition, London: Cassell.
Halliday, M. A. K. and C. M. I. M. Matthiessen (2014) Halliday’s Introduction to Functional
Grammar, London: Routledge.
Hao, J. (2018) ‘Construing scientific causality in published research articles in biology’, Text
and Talk, 38(5): 520–50.
Hao, J. (2020a) ‘Nominalisation in scientific English: A tristratal perspective’, Functions of
Language, 27(2): 143–73.
Hao, J. (2020b) Analysing Scientific Discourse from a Systemic Functional Perspective, London:
Routledge.
He, Y. (2020) ‘Animation as a Semiotic Mode: Construing Knowledge in Science Animated Videos’,
unpublished PhD Thesis, University of Sydney.
Hood, S. (2011) ‘Body language in face-to-face teaching: A focus on textual and interpersonal
metafunctions’, in S. Dreyfus, S. Hood and M. Stenglin (eds) Semiotic Margins, London:
Continuum, 31–52.
Huddleston, R. D., Hudson, R. A., Winter, E. O. and Henrici, A. (1968) Sentence and Clause in
Scientific English, London: UCL.
Humphrey, S. (2017) Academic Literacies in the Middle Years, London: Routledge.
Kress, G. and van Leeuwen, T. (1990) Reading Images, Geelong: Deakin University Press.
Kress, G. and van Leeuwen, T. (1996) Reading Images, 2nd edition, London: Routledge.
Kress, G., Jewitt, C., Ogborn, J., and Tsatsarelis, C. (2001) Multimodal Teaching and Learning:
Rhetorics of the Science Classroom, London: Continuum.
Lemke, J. L. (1990) Talking Science: Language, Learning, and Values, Norwoord, N.J.: Ablex.
Lemke, J. L. (1998) ‘Multiplying meaning: Visual and verbal semiotics in scientific text’, in
J. R. Martin and R.Veel (eds) Reading Science, London: Routledge, 87–113.
Lemke, J. L. (2003) ‘Mathematics in the middle: Measure, picture, gesture, sign and word’, in
M. Anderson, A. Sanez-Ludlow, S. Zellweger and V. Cifarelli (eds) Educational Perspectives
on Mathematics as Semiosis, Ottawa: Legas, 215–34.
Macnaught, L., Maton, K., Martin, J.R. and Matruglio, E. (2013) ‘Jointly constructing seman-
tic waves: Implications for teacher training’, Linguistics and Education, 24(1): 50–63.
Martin, J. R. (1985) Factual Writing: Exploring and Challenging Social Reality, Geelong: Deakin
University Press.
Martin, J. R. (1992) English Text: System and Structure, Amsterdam: John Benjamins.
 The teaching of science 17

Martin, J. R. (2020) ‘Revisiting field: Specialized knowledge in secondary school science and
humanities discourse’, in J. R. Martin, K. Maton and Y. J. Doran (eds) Accessing Academic
Discourse, London: Routledge, 114–47.
Martin, J. R. (2021) ‘Ideational semiosis: A tri-stratal perspective on grammatical metaphor ’,
DELTA Documentação de Estudos em Linguistica Teorica e Aplicada 36:3.
Martin, J. R. and Maton, K. (2013) (eds) ‘Cumulative knowledge-building in secondary
schooling’, Special Issue of Linguistics and Education, 24(1): 1–74
Martin, J. R. and Matruglio, E. (2020) ‘Revisiting mode: Context in/dependency in Ancient
History classroom discourse’, in J. R. Martin, K. Maton and Y. J. Doran (eds) Accessing
Academic Discourse, London: Routledge, 89–113.
Martin, J. R. and Rose, D. (2008) Genre Relations: Mapping culture, London: Equinox.
Martin, J. R. and Veel, R. (eds) (1998) Reading Science: Critical and Functional Perspectives on
Discourses of Science, London: Routledge.
Martin, J. R. and Zappavigna, M. (2019) ‘Embodied meaning: A systemic functional perspec-
tive on paralanguage’, Functional Linguistics, 6(1), 1–33.
Martin, J. R., Maton, K. and Doran, Y. J. (2020a) ‘Academic discourse: An inter-disciplinary
dialogue’, in J. R. Martin, K. Maton and Y. J. Doran (eds) Accessing Academic Discourse,
London, Routledge, 1–31.
Martin, J. R., Maton, K. and Doran, Y. J. (eds) (2020b) Accessing Academic Discourse: Systemic
functional linguistics and Legitimation Code Theory, London: Routledge.
Martin, J. R., Unsworth, L. and Rose, D. (2021) ‘Condensing meaning: Imagic aggregations in
secondary school science’, in G. Parodi (Ed.) Multimodality, Berlin: Peter Lang.
Maton, K. (2013) ‘Making semantic waves: A key to cumulative knowledge-building’,
Linguistics and Education, 24(1): 8–22.
Maton, K. (2014) Knowledge and Knowers:Towards a realist sociology of education, London: Routledge.
Maton, K. (2016) ‘Starting points: Resources and architectural glossary’, in K. Maton, S. Hood
and S. Shay (eds) Knowledge-building, London: Routledge, 233–43.
Maton K. (2018) ‘Thinking like Bourdieu: Completing the mental revolution with
Legitimation Code Theory’, in J. Albright, D. Hartman and J. Widin (eds) Bourdieu’s Field
Theory and the Social Sciences, London: Palgrave Macmillan, 249–68.
Maton, K. (2020) ‘Semantic waves: Context, complexity and academic discourse’, in
J. R. Martin, K. Maton and Y. J. Doran (eds) Accessing Academic Discourse, London,
Routledge, 59–85.
Maton, K. and Chen, R. T.-H. (2020) ‘Specialization codes: Knowledge, knowers and stu-
dent success’, in J. R. Martin, K. Maton and Y. J. Doran (eds) Accessing Academic Discourse,
London, Routledge, 35–58.
Maton, K. and Doran,Y. J. (2017) ‘SFL and code theory’, in T. Bartlett and G. O’Grady (eds)
Routledge Systemic Functional Linguistic Handbook, London: Routledge, 605–18.
Maton, K. and Howard, S. K. (2018) ‘Taking autonomy tours: A key to integrative knowledge-
building’, LCT Centre Occasional Paper 1: 1–35.
Maton, K. and Howard, S. K. (2020) Autonomy tours: Building knowledge from diverse
sources, Educational Linguistic Studies, 2: 50–79.
Maton, K., Hood, S. and Shay, S. (eds) (2016a) Knowledge-building: Educational Studies in
Legitimation Code Theory, London, Routledge.
Maton, K., Martin, J. R. and Matruglio, E. (2016b) ‘LCT and systemic functional linguistics:
Enacting complementary theories for explanatory power’, in K. Maton, S. Hood and S.
Shay (eds) Knowledge-building, London, Routledge, 93–113.
Mouton, M. (2019) ‘A case for project based learning to enact semantic waves: Towards
cumulative knowledge building’, Journal of Biological Education, DOI: 10.1080/00219266.
2019.1585379.
18 Y. J. Doran, Karl Maton and J. R. Martin

Mouton, M. & Archer, E. (2019) ‘Legitimation Code Theory to facilitate transition from high
school to first-year biology’, Journal of Biological Education, 53(1): 2–20.
New London Group (1996) ‘A pedagogy for multiliteracies’, Harvard Educational Review,
66(1): 60–93.
Ngo, T., Hood, S., Martin, J. R., Painter, C., Smith, B. and Zappavigna, M. (2021) Modelling
Paralanguage Using Systemic Functional Semiotics, London: Bloomsbury.
O’Halloran, K. (2005) Mathematical Discourse: Language, symbolism and visual images, London:
Continuum.
Pott, R. and Wolff, K. (2020) ‘Using Legitimation Code Theory to conceptualize learning
opportunities in fluid mechanics’, Fluids, 4(203): 1–13.
Rootman-le Grange, I. & Blackie, M. (2018) ‘Assessing assessment: In pursuit of meaningful
learning’, Chemistry Education Research and Practice, 19: 484–90.
Rose, D. and Martin, J. R. (2012) Learning to Read, Reading to Learn: Genre, knowledge and
pedagogy in the Sydney School, London: Equinox.
Rose, D., McInnes, D. and Körner, H. (1992) Scientific Literacy, Sydney: Metropolitan East
Disadvantaged Schools Program.
Rothery, J. (1989) ‘Learning about language’, in R. Hasan and J. R. Martin (eds) Language
Development: Learning language, learning Culture, Norwood, NJ: Ablex, 199–256.
Simon-Vandenbergen, A., Taverniers, M. and Ravelli, L. (eds) (2003) Grammatical Metaphor:
Views from Systemic Functional Linguistics, Amsterdam: John Benjamins.
Steenkamp, C., Rootman-le Grange, I. and Müller-Nedebock, K. (2019) ‘Analysing assess-
ments in introductory physics using semantic gravity’, Teaching in Higher Education, DOI:
10.1080/13562517.2019.1692335
Unsworth, L. (1997a) ‘Scaffolding reading of science explanations: Accessing the grammatical
and visual forms of specialized knowledge’, Reading, 31(3): 30–42.
Unsworth, L. (1997b) ‘Explaining explanations: Enhancing scientific learning and literacy
development’, Australian Science Teachers Journal, 43(1): 34–49.
Unsworth, L. (1997c) ‘“Sound” explanations in school science: A functional linguistics per-
spective on effective apprenticing texts’, Linguistics and Education, 9(2): 199–226.
Unsworth, L. (2001) Teaching Multiliteracies Across the Curriculum, Buckingham: Open
University Press.
Unsworth, L. (2020) ‘Intermodal relations, mass and presence in school science explanation
genres’ In M. Zappavigna and S. Dreyfus (eds) Discourses of Hope and Reconciliation: J. R.
Martin's Contribution to Systemic Functional Linguistics. London: Bloomsbury, 131–152.
Veale, C., Krause, R. and Sewry, J. (2017) ‘Blending problem-based learning and peer-led
team learning, in an open ended “home-grown” pharmaceutical chemistry case study’,
Chemistry Education Research and Practice, 19(1): 68–79.
Veel, R. (1992) ‘Engaging with scientific language: A functional approach to the language of
school science’, Australian Science Teachers Journal, 38(4): 31–35.
Veel, R. (1997) ‘Learning how to mean – scientifically speaking: Apprenticeship into sci-
entific discourse in secondary school’, in F. Christie and J. R. Martin (eds) Genre and
Institutions, London: Cassell, 161–95.
Wignell, P., Martin, J. R. and Eggins, S. (1993) ‘The discourse of geography: Ordering and
explaining the experiential world’, in M. A. K. Halliday and J. R. Martin (eds) Writing
Science, London: Falmer, 151–83.
Winberg, C. (2012) ‘“We’re engaged”: Mechanical engineering and the community’, in P.
Trowler, M. Saunders, and V. Bamber (eds), Tribes and Territories in the 21st Century, London:
Routledge, 142–55.
 The teaching of science 19

Winberg, C., McKenna, S. and Wilmot, K. (eds) (2021) Building Knowledge in Higher Education:
Enhancing teaching and learning with Legitimation Code Theory, London: Routledge.
Wolff, K. (2021) ‘From principle to practice: Enabling theory-practice bridging in engineer-
ing education’, in C. Winberg, S. McKenna and K. Wilmot (eds) Building Knowledge in
Higher Education, London: Routledge, 180–97.
Wolmarans, N. (2016) ‘Inferential reasoning in design: Relations between material product
and specialised disciplinary knowledge’, Design Studies, 45(A): 92–115.
Yu, Z. (in press) ‘Chemical formalisms in secondary school chemistry: Toward a semiotic
typology’, Semiotica.
PART I

Knowledge-building in
science education
2
TARGETING SCIENCE
Successfully integrating mathematics into
science teaching

Karl Maton and Sarah K. Howard

… like trying to hit a bullet with a smaller bullet, whilst wearing a blindfold, riding a horse.
– ‘Scotty’, in Star Trek, 2009

Introduction
In his Opus Majus of 1267, Roger Bacon described mathematics as ‘the door and
key’ to science. This has become an axiom of educational research into science and
its constitutive disciplines.1 Mathematics is widely heralded as the ‘backbone’ of sci-
ence (Bing and Redish 2009: 1) and ‘deeply woven’ into its practice and teaching
(Redish and Kuo 2015: 562). Science textbooks are shown to exhibit greater use of
mathematics than those of other disciplines (Lemke 1998, Parodi 2012). Learning
the ‘appropriate application’ of mathematical skills is said to be ‘a key part of the
hidden curriculum in science’ (Quinnell et al. 2013: 814). Accordingly, the ability
to integrate mathematical and scientific knowledge is viewed as an important sign
of student progress (Redish 2017). In short, teaching and learning mathematical
knowledge is a central issue for science education. However, just as widely acknowl-
edged is that integrating mathematics into science teaching poses persistent prob-
lems. Studies regularly proclaim that many students struggle with mathematics in
science lessons and consequently become discouraged from continuing in science
(Meli et al. 2016). Even students who have chosen further studies in science ‘often
seem to see “maths” as a separate subject, a necessary evil … rather than an integral
part of the discipline’ (Quinnell et al. 2013: 811).
A key aspect of the problem is said to lie with fundamental differences between
the two disciplines. Scholars emphasize that mathematics when used within sci-
ence has ‘a different purpose – representing meaning about physical systems rather
than expressing abstract relationships. It even has a distinct semiotics … from pure
mathematics.’ (Redish and Kuo 2015: 563). Integrating mathematics into science
24 Karl Maton and Sarah K. Howard

education is thus not a simple matter of adding mathematical content to science les-
sons. Indeed, differences between the disciplines are such that students encounter-
ing a concept located within both bodies of knowledge often fail to recognize that
they are exploring similar ideas in different contexts (e.g. Planinic et al. 2012). As a
result, students may succeed in mathematics classes but ‘fail to use those same tools
effectively’ in science classes, leaving educators ‘distressed and confused’ (Redish
2017: 25).
A challenge facing science education is thus to identify and develop teaching
practices that select, recontextualize and integrate mathematical knowledge in ways
that support the learning of scientific knowledge. Unsurprisingly, this has been the
subject of a significant body of research. The resulting work has generated sugges-
tive ideas for learning specific mathematical skills for particular scientific problems.
However, there is no overarching or integrating model for successful pedagogic
integration. The problem ‘remains unsolved’ (Redish and Kuo 2015: 561). This
chapter contributes towards the creation of such a model. It does so through by
offering a fresh approach that complements existing frameworks by bringing to
light issues that have hitherto been sidelined.
We begin by highlighting how research into science education offers significant
insights into how students learn scientific ways of knowing but either neglects the
forms of knowledge being taught or, where knowledge is discussed, treats ‘science’
and ‘mathematics’ as self-evident and unchanging.We argue that these constructions
of the problem help underpin its persistence by ignoring knowledge, background-
ing teaching, and glossing over how ‘science’, ‘mathematics’ and their interrelations
vary across contexts and change through the course of education. We then outline
a framework that can complement existing approaches by bringing these issues into
the picture. We introduce concepts from the Autonomy dimension of Legitimation
Code Theory that conceptualize one aspect of the forms taken by knowledge prac-
tices and allow research to capture changing relations between changing forms of
knowledge. Specifically, autonomy codes reveal the organizing principles underlying
different knowledge practices, autonomy pathways trace changes in relations between
knowledge practices over time, and targets embrace the contextual nature of what
is viewed as ‘science’ and ‘mathematics’. We illustrate the value of these concepts in
analyses of science classrooms drawn from a major study of secondary schooling.
These show that different autonomy pathways taken by teachers enable or constrain
the integration of mathematics into classroom science. We conclude by reflecting
on the potential of the concepts of autonomy codes, pathways and targets to bring
knowledge into the picture and to connect specific instances of pedagogic practice
together within a general model of pedagogic integration.

Studying mathematics in science: blind spots

To integrate mathematics into classroom science requires teaching practices that
appropriately select ideas from one body of knowledge (mathematics) and recon-
textualize that selection within a second selection of ideas from another body of
 Targeting science 25

knowledge (science). A key issue is thus how different teaching practices shape
the forms taken by ideas from those bodies of knowledge when they are brought
together. Put simply, the question is: what teaching practices enable or constrain the
integration of mathematical knowledge into scientific knowledge? These statements
may seem unnecessary. However, studies of science education typically sideline both
teaching practices and changing forms of knowledge. These blind spots arise from
three assumptions that pervade the field: that knowledge equates to knowing, that
education equates to learning, and that ‘science’ and ‘mathematics’ are self-evident.

Knowing and learning

The first assumption is that ‘knowledge’ comprises mental processes of under-
standing that reside ‘in the heads of persons’ (von Glasersfeld 1995: 1). Reflecting
this ‘subjectivist doxa’ (Maton 2014: 3–14), research focuses almost exclusively
on cognitive and affective ways of knowing. Knowledge as an object of study in
its own right – one taking particular forms which have effects for bringing that
knowledge together with other forms – is left out of the picture. Put another way,
the assumption is that to analyze knowledge one must analyze ways of knowing.
Rather than distinguish between students’ dispositions and what they are learning,
as a precursor to exploring relations between knowing and knowledge, the only
concern is the former. This assumption that knowledge is nothing but knowing
is typically accompanied by a second assumption: that education is nothing but
learning. When studying ‘ways of knowing’, research overwhelming focuses on
student interactions, such as when solving a scientific problem. Teaching is rarely
centre stage, if considered at all. Each of these assumptions thereby takes part of the
picture for the whole.
This focus on learning and ways of knowing is illustrated by studies using the
‘resources framework’ (e.g. di Sessa 1993, Hammer 2000, Redish 2014, 2017), an
influential approach to physics education research. The framework explores ‘how
our students think’ (Redish 2014), such as ‘the student’s perception or judgement
(unconscious or conscious) as to what class of tools and skills is appropriate to bring
to bear in a particular context’ (Bing and Redish 2009: 1). The concern is how
‘cognitive resources’ are ‘activated in response to a perception and interpretation
of both external and internal contexts’ (Redish and Kuo 2015: 573). Thus student
perceptions are central – what their perceptions may be about, the forms taken by
knowledge itself, is not analyzed. This subjectivism is thoroughgoing – everything
is psychological. For example, the term ‘epistemology’ is used to refer not to inter-
subjectively shared field-level knowledge practices but rather to personal frames of
individual understanding (e.g. di Sessa 1993, Hammer and Elby 2002). Accordingly,
disciplines such as mathematics are viewed as comprising ‘ways of knowing’ (Redish
2017) and studies of mathematics in science explore how students solve problems in
order ‘to model their thinking’ (Bing and Redish 2009: 2).
Research using the ‘resources framework’ offers valuable insights into how stu-
dents learn ways of knowing. However, learning is not the sum of education, and
26 Karl Maton and Sarah K. Howard

ways of knowing are not the sum of disciplines. Much remains missing from the
equation. Such studies could provide a powerful basis for understanding the inte-
gration of mathematics into science lessons if they were complemented by analyses
of teaching and analyses of knowledge. However, frameworks for bringing these
issues into the picture are lacking in the wider field. Currently, whatever approach
they are using, studies of science education tend to reduce knowledge to knowing
and education to learning, generating blind spots in the overall field of vision.2 For
example, studies of mathematics in science education draw on such frameworks as
‘thinking dispositions’ to suggest that attributes such as curiosity can help students
shift from ‘rigidity of mind’ to ‘fluid thinking’ that ostensibly supports successful
integration (Quinnell et al. 2013). Similarly, studies adopting a ‘cognitive blend-
ing framework’ analyze how students draw on ‘mental spaces’ when combining
physical and mathematical knowledge (Bing and Redish 2007). Thus, everything
lies in the mind of the beholder. Similarly, teaching is sidelined. The implications
of studies of student learning for how mathematics should be taught in science
often take the form of afterthoughts, as if teaching is merely an epiphenomenon
of learning. Typically, such implications simply comprise calls for the integration
of mathematics into science in teaching (e.g. Meli et al. 2016; Planinic et al. 2012),
leaving unsaid what teaching practices would support that integration. In short,
the widely shared focus on ways of knowing (rather than also knowledge) and on
learning (rather than also teaching) limits current understanding of how pedagogic
practices enable or constrain the integration of mathematics into science within
classrooms.

Knowledge as self-evident and invariant

Knowledge is not entirely absent from discussions of mathematics in science. As
mentioned earlier, differences between their purposes and ‘semiotics’ are said to
contribute to student difficulties.Typically, science is described as condensing addi-
tional meanings into numbers and symbols and as requiring a different approach
to interpreting mathematical results, reflecting its relationship with the exter-
nal world (e.g. Bing and Redish 2009, Redish and Kuo 2015). Such attributes
are also highlighted in discussion of the ‘affordances’ of mathematical represen-
tation for science (e.g. Fredlund et al. 2012). In systemic functional linguistics,
work has shown how mathematics ‘multiplies’ meanings in relation with language
and images (Lemke 1998), offering additional resources for construing scientific
knowledge (O’Halloran 2010). More recently, Doran (2018) has brought together
systemic functional linguistics with Legitimation Code Theory (LCT) to identify
key mathematical genres in school physics and explore the role these play along-
side language and images as physics progresses through school. Studies using LCT
on its own have also explored forms of scientific knowledge (including math-
ematical symbols) involved in teaching and assessment in terms of differences in
their complexity and context-dependence (Georgiou et al. 2014; Blackie 2014;
Conana et al. 2016).
 Targeting science 27

What is left open, however, is the question of identifying ‘mathematics’ and ‘sci-
ence’.To analyze integration in a classroom requires determining which knowledge
is ‘mathematics’, which knowledge is ‘science’, and when a specific idea or practice
has been recontextualized from one into the other. However, existing studies typi-
cally describe content, such as the formula or problem being discussed by students,
and simply state what is ‘mathematical’ and what is ‘scientific’ (e.g. Meli et al. 2016),
as if self-evident. Alternatively, discussions of the nature of ‘science’ or ‘mathematics’
make claims about each discipline as a whole, as if homogeneous and unchang-
ing. That these ways of constructing knowledge are problematic flows from two
uncontentious commonplaces. First, however distinctive the bodies of knowledge
populating their intellectual fields might be (and this is debatable), the manifestations
of ‘science’ or ‘mathematics’ in a specific classroom cannot be assumed. Which knowl-
edge from an intellectual field is selected, recontextualized and enacted as curricula,
and which knowledge from a curriculum is selected, recontextualized and enacted
in classroom pedagogy varies geographically, institutionally and through the stages
of education.3 Put simply, the knowledge practices comprising ‘science’ and ‘math-
ematics’ are not necessarily the same in two classes in a school, let alone in different
schools, years of study, states or countries. Second, at what stage of education specific
knowledge practices from ‘mathematics’ are integrated into and become ‘science’ is
not universal. What has already been integrated into ‘science’ in one classroom may
remain separate ‘mathematics’ in another classroom. In short, what is ‘science’ var-
ies between classroom contexts and changes through education, the ‘mathematics’
drawn on when teaching ‘science’ varies and changes, and the degree to which that
‘mathematics’ has been transformed into ‘science’ also varies and changes. Classroom
‘science’ and ‘mathematics’ are two variable and changing bodies of knowledge
whose interrelations are themselves situational and mutable. They are anything but
self-evident, homogeneous or unchanging – thus our liberal use of quote marks. To
simply state that specific ideas are ‘science’ or ‘mathematics’ is to beg the question
of how that is determined.
Addressing the question is not easy. It requires avoiding a false dichotomy
between essentialism and relativism that continues to bewitch education research
(Maton 2014: 1–22). On the one hand, universalizing claims about ‘science’ and
‘mathematics’ without a limiting context (such as ‘science in this classroom’) can
lead to essentialism that treats their properties as homogeneous and invariant. On
the other hand, insistence on the contextual limits of any definitions (or offering no
more than ‘science in this classroom, at this moment’) can slide into relativism that
treats ‘science’ or ‘mathematics’ as an endless flux. The former generates overgener-
alized claims that are unhelpful for analyzing empirical data; the latter leads to the
banal conclusion that subject areas are constructed, contested and fluid, paralyzing
the possibility of analysis. As yet, research into science education has not steered a
course between this Scylla and Charybdis. However, without facing squarely the
question of distinguishing knowledge practices, empirical studies can only continue
creating a series of context-bound models of specific instances. The wider issue of
pedagogic integration will remain unsolved.
28 Karl Maton and Sarah K. Howard

Seeing into the blind spots

Assuming that knowledge is only knowing, that education is only learning, and
that disciplinary knowledges are self-evident generates blind spots. It is difficult to
develop pedagogic practices that integrate mathematics into science lessons so long
as teaching practice and both forms of knowledge are not analyzed. Existing insights
into how students learn ways of knowing thus need to be complemented by: (i)
studies of teaching practice; (ii) concepts that make visible the forms of knowledge
practice being taught and learned; and (iii) a means of enacting those concepts
that captures the variant and contextual nature of ‘science’ and ‘mathematics’.4 The
first requires a shift of empirical focus. The second can be addressed by drawing
on Legitimation Code Theory, a framework that reveals the organizing principles
of knowledge practices. The third is trickier – it is akin to how Scotty described
transwarp beaming in the motion picture Star Trek: ‘like trying to hit a bullet with a
smaller bullet, whilst wearing a blindfold, riding a horse’. It needs to account for two
changing phenomena (what is ‘science’ and what is ‘mathematics’) whose relations
are also changing (through different degrees of separation and integration).We now
turn to a means for doing so.

Autonomy
Legitimation Code Theory or ‘LCT’ is a framework for researching and shaping
practice. It begins from the notion that there is more to what we say or do than
what we say or do. In other words, the meanings of practices are not exhausted by
their content; practices are also ‘languages of legitimation’ or criteria for measur-
ing achievement (Maton 2014). In short, what we say or do express principles
of legitimacy or ‘legitimation codes’. LCT comprises several dimensions or sets of
concepts that explore different aspects of legitimacy (Maton 2016). Central to each
dimension are concepts for analyzing the organizing principles underlying practices,
dispositions and contexts as a particular species of ‘legitimation code’. In terms of
our needs here, these concepts bring knowledge into the picture by revealing the
organizing principles generating its various forms.The dimension most directly rel-
evant to exploring integration is Autonomy, which focuses on relations between sets
of practices (such as subject areas) and conceptualizes their organizing principles as
autonomy codes. We shall first define the concepts, then discuss how they are enacted
using translation devices and targets. For reasons that become clear, we begin rather
abstractly, before concretizing the meanings of concepts.

Autonomy codes
The dimension of Autonomy begins from the simple premise that any set of prac-
tices comprises constituents that are related together in particular ways. Constituents
may be actors, ideas, institutions, machine elements, body movements, etc.; how
they are related together may be based on explicit procedures, tacit ways of working,
 Targeting science 29

mechanisms, unstated orthodoxies, etc. The concepts of ‘autonomy codes’ explore

how practices distinguish their constituents and their ways of relating from those
of other practices. Put another way, the concepts examine how practices establish
different degrees of insulation around their constituents and the ways those con-
stituents are related together. These are analytically distinguished as:

• positional autonomy (PA) between constituents positioned within a context or

category and those positioned in other contexts or categories; and
• relational autonomy (RA) between the relations among constituents of a con-
text or category and the relations among constituents of other contexts or
categories.

Each may be stronger (+) or weaker (−) along a continuum of strengths, where stron-
ger represents greater insulation and weaker represents lesser insulation. Stronger
positional autonomy (PA+) indicates that constituents of a context or category
are relatively strongly delimited from constituents associated with other contexts
or categories (strongly insulated positions); and weaker positional autonomy (PA–)
indicates where such distinctions are less demarcated (weakly insulated positions).
Stronger relational autonomy (RA+) indicates that the ways of relating constituents
together are relatively specific to a set of practices (autonomous principles), and
weaker relational autonomy (RA–) indicates that the ways of relating may be drawn
from or shared with other sets of practices (heteronomous principles).
As shown in Figure 2.1, positional autonomy and relational autonomy are visual-
ized as axes of the autonomy plane. Varying their strengths independently (PA+/−,
RA+/−) generates four principal autonomy codes:

FIGURE 2.1 The autonomy plane (Maton and Howard 2018: 6)

30 Karl Maton and Sarah K. Howard

• sovereign codes (PA+, RA+) of strongly insulated positions and autonomous

principles, where constituents are associated with the context or category and
act according to its specific ways of working;
• exotic codes (PA−, RA−) of weakly insulated positions and heteronomous prin-
ciples, where constituents are associated with other contexts or categories and
act according to ways of working from other contexts or categories;
• introjected codes (PA−, RA+) of weakly insulated positions and autonomous
principles, where constituents associated with other contexts or categories are
oriented towards ways of working emanating from within the specific context
or category; and
• projected codes (PA+, RA−) of strongly insulated positions and heteronomous
principles, where constituents associated with the specific context or category
are oriented towards ways of working from elsewhere.
These concepts help address the need to make visible the forms of knowledge
being taught and learned. Put simply, the four codes state that what matters are:
internal practices and principles (sovereign codes); other practices and principles
(exotic codes); other practices turned to intrinsic purposes (introjected codes); and
internal practices turned to other purposes (projected codes). To explore processes
that occur through time, such as classroom practice, one can analyze the different
pathways traced around the plane by successive autonomy codes. There is an unlim-
ited number of potential pathways (see Maton and Howard 2018). In this chapter
we discuss the two pathways illustrated in Figure 2.2: one-way trips that begin in one
code and end in another code; and tours that begin in one code, move through other
codes, and return to their originating code. We shall show that autonomy tours in
teaching practice enable, and one-way trips constrain the integration of ‘mathemat-
ics’ into ‘science’. However, before doing so there remains the question of defining

FIGURE 2.2 Examples of two autonomy pathways

 Targeting science 31

‘science’ and ‘mathematics’ in a way that systematically embraces their variant and
contextual nature. This is achieved through translation devices and targets.

Translation devices and targets

Thus far we have described ‘autonomy codes’ in abstract terms and without specific
examples. This allows the concepts to be enacted across a wide range of diverse
phenomena. Doing so offers the possibility of reaching beyond descriptions of
specific instances of classroom practice to generate a general model of pedagogic
integration. However, it also means one must be clear how the concepts are mani-
fested within a specific object of study. In LCT this is achieved through ‘transla-
tion devices’ that relate concepts to data (Maton and Chen 2016). Table 2.1 is a
generic translation device that relates autonomy codes to all forms of data. The device
divides the continua of strengths for positional autonomy and relational autonomy
into categories of progressively finer-grained levels of delicacy, from categories for
stronger/weaker (target/non-target) through subcategories, use of which depends on
the analysis.
To activate the device one asks: what constituents (practices, beliefs, ideas, actors,
etc.) and what principles (purposes, aims, ways of working, etc.) are considered con-
stitutive of this context or category, here, in this space and time, for these actors? This
gives a ‘target’. As shown in Table 2.1, target constituents embody stronger posi-
tional autonomy and all other, non-target constituents embody weaker positional
autonomy; similarly, target principles embody stronger relational autonomy and all
other, non-target principles embody weaker relational autonomy. These categories
can be divided into subcategories by asking which target constituents and principles
are considered core and which ancillary to the context or category, and which non-
target constituents and principles are considered associated or unassociated with the
target. Asking the same basic questions again generates a third level comprising inner
and outer forms of core and ancillary targets, and near and remote forms of associated
and unassociated non-targets.

TABLE 2.1 Generic translation device (Maton and Howard 2018: 10)

PA/RA 1st level 2nd level 3rd level

target core inner
outer
ancillary inner
outer
non-target associated near
remote
unassociated near
remote
32 Karl Maton and Sarah K. Howard

Activating the device allows us to aim directly at the problem of defining ‘sci-
ence’ and ‘mathematics’. Rather than shying away from knowledge practices vary-
ing across contexts, the notion of ‘target’ makes that the starting point for analysis.
This is best shown by a concrete example. Here we draw on a major study of how
secondary school teachers select, assemble and enact knowledge in their classroom
practice when teaching at Stage 4 (Years 7–8) and Stage 5 (Years 9–10) in three
secondary schools in New South Wales, Australia.5 Data comprised videorecordings
of lessons across whole units of study (6–8 hours each), interviews with teachers, all
teaching materials, and student artefacts. In this chapter we shall discuss examples of
classroom practice by two science teachers in Year 7 of secondary schooling.6
To enact ‘target’ first one considers whose view of the target to begin from, as other
agents in a context (such as students in a classroom) may have different targets. A
target is always someone’s or something’s conception of what makes a context distinc-
tive and thus in our analyses always accompanied by possessives (e.g. his/her/their
targets). Here, reflecting our concern with teachers’ practices, we focus on their tar-
gets. Second, one must consider what level of their targets to examine. LCT concepts
can be enacted at all levels of analysis; for example, a teacher’s targets may include
an entire curriculum, a unit of study, a lesson, a task, and so forth. Reflecting our
concern with how teachers attempt to integrate mathematics into science to meet
the needs of the Stage 4 curriculum, our specific translation device has curriculum
stage as its first level and unit of study as its second level.7 As summarized in
Table 2.2, in interviews and pedagogic materials the teachers identified their tar-
get content (PA+) as the Stage 4 science syllabus in New South Wales and their
target purpose (RA+) as teaching students that content. Put simply, here: positional
autonomy conceptualizes where the ideas expressed in classroom practice are drawn
from, the Stage 4 science syllabus (PA+) or elsewhere (PA–); and relational autonomy
conceptualizes the purposes for which they being expressed, teaching and learning
that science syllabus (RA+) or other purposes (RA–). Interviews and pedagogic
materials further identified the teachers’ core targets (++) as the specific science unit
being taught, with other units in Stage 4 science considered ancillary targets (+).

TABLE 2.2 Simplified specific translation device for this analysis

PA/RA 1st level In this analysis: 2nd level In this analysis:

New South Wales core specific unit in target

target Stage 4 syllabus ancillary other topics or years in
for subject area target
associated other educational
other contents knowledge
non-target
or purposes unassociated knowledge from beyond
education
 Targeting science 33

(Their inner-core targets comprised the content points they created for each specific
lesson). In terms of non-targets, teachers viewed other educational knowledge (such
as other subjects or other Stages and levels of education in science) as associated (–)
with their target, and knowledge from beyond education as more distanced or unas-
sociated (– –).
By ‘targeting’ analysis we can identify ‘science’ as it is constructed in the spe-
cific context under study, avoiding universalizing essentialism. By translating that
particular set of empirical ideas and practices into ‘autonomy codes’, we can move
beyond context-bound, endlessly varying descriptions of difference, avoiding rela-
tivism. We can both embrace the specificities of each context and compare practices
across different contexts, capturing the endless forms most wonderful that are ‘sci-
ence’. Moreover, targeting ‘science’ allows us to analyze the movement of ideas and
practices between subject areas as they are recontextualized and integrated. Given
that the ‘target’ depends on the object of study, no single idea, practice, belief, etc. is
always and everywhere the same code. A practice may be moved around the plane;
for example, in our discussion of an autonomy tour below, ‘graphing’ is successively
constructed as an exotic code (as mathematics content for mathematical purposes),
an introjected code (mathematics content for learning science), and a sovereign
code (science content for learning science).
We can now begin to explore what teaching practices enable or constrain inte-
gration of ‘mathematics’ into ‘science’. To illustrate how, we shall analyze classroom
practices by two teachers (mentioned above) from Year 7 schools teaching the same
unit from the same state curriculum. The difference between the examples lies in
the autonomy pathways traced by their teaching practice. In the first example, the
teacher fails to integrate mathematics into science. He leads students on a one-way
trip out of ‘science’ into an activity he describes as ‘maths’ that remains segmented
from his target knowledge. In the second example, a different teacher takes students
on an autonomy tour that integrates non-target ‘mathematical’ knowledge about cre-
ating graphs into her target ‘science’ knowledge about Earth’s seasons.

One-way trip from science ‘to do some maths’

Our example of teaching that fails to integrate ‘mathematics’ into ‘science’ com-
prises a distinct phase of activity spanning an entire lesson of over 50 minutes. The
teacher’s core target for the wider unit, as later described in an interview, is:

to teach them [students] about the universe and our solar system and what’s
beyond Earth. Some of them didn’t quite understand the relationships in the
universe so we have to make them clearer for them.… How we get night and
day or how you get the different seasons.

This reflects a ‘sub-strand’ of the state curriculum for Year 7 science entitled ‘Earth
and space sciences’, which is ‘concerned with Earth’s dynamic structure and its place
in the cosmos’ (NESA).8 In the lesson discussed here, the teacher tells students they
34 Karl Maton and Sarah K. Howard

are going to make sense of the scale of the solar system, but quickly shifts classroom
practice into using numbers to calculate percentages, detached from learning about
the science content.This takes students on a one-way trip from the teacher’s sovereign
code into an exotic code, a pathway that does not return to his target content or
target purpose. After 52 minutes he draws the ‘maths’ activity to a close by declaring:
‘I know it’s confusing’.

Pathway into ‘a lot of numbers’

The teacher begins the lesson by showing students a short YouTube video entitled
‘The smallest to the biggest thing in the universe’. Starting from hypothesized enti-
ties in quantum physics (such as strings), the video zooms outwards through ever-
larger phenomena to end with the known galaxy. He then segues to the activity that
will consume the rest of the lesson:

TEACHER So, as you saw, some of those distances and some of those sizes don’t
really mean a lot to us, because we just can’t fathom the distances involved,
okay? So some of the other distances, especially in our solar system, are the
same. So what we’re going to do is, we’re going to put the distances and the
sizes relative to Earth. Okay? So we’re going to put all the planets and the dis-
tance to the Sun and we’re going to make them relative to the Earth.

At this point, the intended classroom practice is to explore content about the solar
system (stronger positional autonomy) for the purpose of understanding the solar
system (stronger relational autonomy); i.e. within the teacher’s core target – deep
inside his sovereign code.
The teacher then directs students to ‘draw up a table’ of ‘seven columns and 10
rows’ and shows a PowerPoint slide of a table, to which he adds two column titles
by hand on the whiteboard, reproduced here as Table 2.3. He tells the class to ‘copy
down this information if you haven’t already got it’. After reminding students they

TABLE 2.3 Table provided by teacher for activity

Radius Distance from Time to Time taken to Diameters Distance

(km) the sun (km) orbit around turn once on its as % of as % of
the sun axis Earths Earths
The Sun 695800
Mercury 2439.7 57910000 88Ed 58d15h30m
Venus 6052 108200000 224.7Ed 116d18h0m
Earth 6371 149600000 365.25Ed 1d
Mars 3390 227900000 686.97Ed 1d0h40m
Jupiter 69911 778500000 12Ey 9h56m
Saturn 58232 1433000000 29Ey 10h39m
Uranus 25362 2877000000 84Ey 17h14
Neptune 26422 4503000000 165Ey 16h06m
 Targeting science 35

had written down diameters of planets and their distances from the sun in a previ-
ous lesson, he explains:

TEACHER I want you to add this information, these two [points to third and
fourth titled columns], because these two are relative to the Earth. … Since
we’ve already got the information just do four columns, because we’re going
to do some maths.

That the teacher declares ‘we’re going to do some maths’ does not by itself indicate
a shift beyond ‘science’ into non-target knowledge. For example, ‘maths’ could refer
to procedures or ideas he has previously integrated into his target – already scien-
tized mathematics, so to speak. Similarly, calculating percentages is not necessarily
beyond his target. As we shall see in our second example, no practice is always a
specific code. To identify the autonomy code, one must leave aside assumptions of
what is ‘science’ or ‘maths’ and begin from the teacher’s target. This he described,
and teaching materials revealed, as the Stage 4 science syllabus. A strand of this
syllabus entitled ‘science inquiry skills’ includes for Year 7: ‘Summarise data, from
students’ own investigations and secondary sources, and use scientific understanding
to identify relationships and draw conclusions based on evidence’.9 So, calculating
diameters and distances from the sun of planets as percentages of those of Earth
could potentially sit within the teacher’s target. However, the syllabus emphasizes
that such ‘science inquiry skills’ give students ‘the tools they need to achieve deeper
understanding of the science concepts’ – they must be ‘closely integrated’ with
learning the ‘science knowledge’ outlined in a strand entitled ‘Science understand-
ing’.This strand includes having ‘students view Earth as part of a solar system, which
is part of a galaxy, which is one of many in the universe, and explore the immense
scales associated with space’. Thus, whether calculating percentages lies within the
teacher’s target depends on whether he integrates its content or purpose with view-
ing Earth as part of a solar system and exploring the immense scales of space. As
the teacher stated at the outset, this was his intention. However, as we shall see, in
practice he does not relate the activity to any such ‘science understanding’.
Instead, as illustrated by Figure 2.3, the teacher quickly shifts the task into an
exotic code in which non-target content is used for non-target purposes. He weak-
ens positional autonomy by disconnecting the contents of the table from his target
topic. For example, he describes its contents as ‘information’ five times in just the
first minute: ‘copy down this information … just use your information … add
this information … you’ve already got the information … we’ve already got the
information’. At the same time he weakens relational autonomy by describing the
purpose as ‘to do some maths’ without relating this either to procedures previously
integrated into ‘science’ or to learning new syllabus content.10 For example, when
responding to questions from students, he states:

TEACHER Just do the last two columns and then add two more because we’re
going to do some maths in the last two.
36 Karl Maton and Sarah K. Howard

FIGURE 2.3 Shift from sovereign code to exotic code

Following his instructions, some students copy numbers from the table while a
series of students ask the teacher which numbers they should copy. This concern
with numbers as only numbers continues when, after eight minutes, he addresses
the whole class:

TEACHER Alright, do we know how to work out the percentage for ….? So
which one do we have to divide by? [Student name], what do you think? To
work out Mercury, the percentage compared to Earth? What do you think we’d
have to do?
STUDENT Divide it by a hundred?
Mercury
TEACHER No, no, no. Alright. What we do [draws on whiteboard: ×
Earth
100]. Alright, so distance percentage [pointing to last column title] is the dis-
tance from the sun. Okay? So to work out the percentage, you divide each of
the planets by the Earth’s diameter.

Over the next 15 minutes the teacher repeats similar instructions to a series of
individual students, each time describing what ‘information’ must be multiplied
or divided to ‘give you a percentage’. When he mentions the names of planets, the
teacher is referring to specific empty cells in the table – shown by physically point-
ing to the cell – rather than to planets. More often, the teacher refers to the content
as ‘information’ or ‘number’, such as:

TEACHER Divide that number [pointing to the table] by that number [pointing]
for the distance; that number [pointing] by that number [pointing] for the
diameter. Alright? That’s what you’re supposed to be doing.
 Targeting science 37

The content is from neither the teacher’s target of the syllabus nor his core target
of learning about the solar system; the purpose is to ‘work out the percentage’ or
‘do maths’. The knowledge being expressed thus remains within an exotic code
(Figure 2.3).
During the course of the activity some students ask questions that represent
opportunities to relate the activity to the issue of grasping the scale of the solar
system. For example, one student asks whether people are made of ‘planks’, another
asks what is smaller than a ‘string’ (both mentioned in the earlier video), and a third
asks ‘Are we made of stardust?’. The teacher’s responses – ‘No’, ‘Didn’t you watch
the video?’, and ‘What do you think?’, respectively – do not connect to target con-
tent or turn the questions to his target purpose.
After 38 minutes the teacher asks students to call out numbers for cells in the
column ‘Diameters as % of Earths’. He then raises the question: ‘What do these
percentages actually mean?’. This is an opportunity to strengthen relational auton-
omy by turning these numbers to the purpose of viewing Earth as part of a solar
system or exploring the immense scales involved, and an opportunity to strengthen
positional autonomy by connecting the ‘information’ to knowledge about the solar
system. A student suggests ‘A lot of numbers’, an answer that accurately reflects
the exotic code characterizing the activity. The teacher leaves his own question
unanswered. The class then repeats the pattern outlined above: students make cal-
culations (for the ‘Distances’ column), the teacher repeats similar instructions to
students, and numbers are solicited from the class. Classroom practice stays in an
exotic code. The activity is ended after 52 minutes by the teacher saying ‘I know
it’s confusing’ and announcing that they will look at ‘day and night’ in the next
lesson.

Summary: ‘That’s maths!’

The autonomy pathway traced by this lesson represents a one-way trip out of the
teacher’s target of ‘science’ in order ‘to do some maths’. As portrayed by Figure 2.3,
the knowledge expressed in classroom practice shifts from a fleeting sovereign code
to a very long stay in an exotic code. As shown by the times given above, almost
the entire ‘science’ lesson is ‘maths’. The teacher could have chosen to conduct this
activity inside his sovereign code by closely integrating the numeric activity with
his syllabus target. Instead, he chooses to project the activity as beyond his target,
as doing ‘maths’ to ‘work out the percentage’. As we discuss below, this code shift
is not necessarily antithetical to integrating this ‘maths’ into ‘science’. At any point
during the lesson, the teacher could strengthen positional autonomy by connect-
ing to his target content or strengthen relational autonomy by turning non-target
content (calculating percentages) to his target purpose. Instead, he keeps classroom
practice in the exotic code: the content remains numbers and calculations, and the
purpose remains using numbers to calculate other numbers. Thus, the shift to an
exotic code does not integrate ‘maths’ into ‘science’. Late in the lesson, in response
to a student declaring ‘This is hard, sir’, the teacher replies ‘That’s maths! We still
38 Karl Maton and Sarah K. Howard

have to do maths in science’. However, this ‘maths’ is not ‘in science’ and so knowl-
edge of calculating percentages remains strongly segmented from knowledge of the
solar system.

Autonomy tour integrating ‘mathematics’ into ‘science’

To illustrate how ‘mathematics’ can be integrated into ‘science’ we turn to a dif-
ferent teacher at a different school but teaching the same unit (‘Earth and space
sciences’) at the same level (Year 7 secondary school). The example begins in the
second lesson of a unit on the causes of Earth’s seasons, as students transform their
results from a practical experiment into graphs. In the first lesson students had con-
ducted an experiment to explore the effect on temperature of the angle at which
sunlight strikes the Earth’s surface. In groups, students used a lamp to represent the
sun, and a wooden block to represent the Earth. Varying the angle of the lamp to
the block (15, 30, 60 and 90 degrees), they recorded the temperature of the block
at different times (initial, 2.5 minutes, 5 minutes) from an attached thermometer.
Prior to the experiment each student had written a hypothesis of whether increas-
ing the angle would increase, decrease or have no effect on the temperature. The
second lesson directly builds on this activity. The teacher begins by setting out her
(inner-core) target:

TEACHER What we will be doing today is looking at those results, graphing the
results and then talking about what it is that we were actually trying to model.

Over the next 35 minutes the teacher leads students on an autonomy tour through
those activities: from her sovereign code (discussing their results), through an exotic
code (recapping ‘graphing rules’), and an introjected code (applying those rules to
graph their results), before returning to her sovereign code (by relating the resulting
graphs to Earth’s seasons). As a result, the graphing activity becomes integrated into
‘science’.

A tour through ‘graphing’

The teacher begins by recounting the experiment and then solicits students’ overall
findings:

TEACHER So looking at your results there, who can give me a statement about
what their results did?
STUDENT As the angle of the block increased, the temperature increased.
TEACHER Fantastic. I love that. That’s a really great statement. Did someone get
something different in their results?

The teacher thus begins deep inside her sovereign code. Both content (experiment
modelling a factor in Earth’s seasons) and purpose (to learn about the results) are
 Targeting science 39

located within her inner-core target for the lesson. After discussing the findings of
several students, she announces:

TEACHER Here [on the whiteboard] is your table that you should have had drawn
up from the last lesson. We are going to graph… I want you to think about the
graphing rules and start getting yourself ready for graphing.

As we emphasized, no activity is intrinsically a specific autonomy code. ‘Graphing’ is

not necessarily non-target – ‘graphing’ can be mathematical or scientific. To deter-
mine autonomy codes we must consider the teacher’s target: the Stage 4 science
syllabus. A strand entitled ‘science inquiry skills’ includes for Year 7: ‘Construct and
use a range of representations, including graphs, keys and models to represent and
analyze patterns or relationships in data’.11 Thus, graphing is potentially within the
teacher’s target. However, as discussed in the previous example, the syllabus describes
‘science inquiry skills’ as giving students ‘the tools they need to achieve deeper under-
standing of the science concepts’ by being ‘closely integrated’ with learning the ‘sci-
ence knowledge’ outlined in the syllabus strand ‘Science understanding’. This strand
includes ‘how changes on Earth, such as day and night and the seasons, relate to Earth’s
rotation and its orbit around the sun’.Thus, whether graphing lies within the teacher’s
target depends on whether she integrates its content or purpose with learning about
Earth’s seasons. Here, we shall show that she begins by separating graphing in terms
of both content and purpose, then turns it to purpose, before connecting its content.
This tour begins with the teacher recapping her ‘graphing rules’ separately from
Earth’s seasons. Continuing on from the preceding classroom quote, she says:

TEACHER So, who can remind me about what the rules are for graphing?
STUDENT Y versus X.
TEACHER Y versus X. How do we know which one goes where?
STUDENT The independent variable goes on one side.
TEACHER The independent variable goes on one of them.Yes, that’s good.
STUDENT And the dependent variable …
TEACHER … goes on the other one. The thing that is the most regular, which
is usually your IV [independent variable], goes on the X, and your DV
[dependent variable] goes on the Y.

This recap embodies: weaker positional autonomy (PA–), as these ‘graphing rules’
are not related to Earth’s seasons; and weaker relational autonomy (RA–), as the
purpose is recapping the ‘graphing rules’ rather than learning about Earth’s seasons.
As portrayed in Figure 2.4, the teacher has shifted from deep inside her sovereign
code to just inside an exotic code.12
Thus far, classroom practice traces the same pathway as the previous example.
However, where that teacher remained within an exotic code for the entire lesson,
this teacher does not stay for long. She quickly shifts the class into a third code by
repurposing the knowledge of ‘graphing rules’:
40 Karl Maton and Sarah K. Howard

FIGURE 2.4 Shift from sovereign code to exotic code

TEACHER Now, in this experiment, who can tell me – there’s a little problem.
Have a look at our data. Can you tell me which one goes on the X and which
one goes on the Y?

One student suggests ‘the angle’ should go on the Y-axis, another suggests the X-axis,
and the teacher asks students for the locations of ‘temperature’ and ‘time’. After a
student exclaims ‘Wait! What?’, the teacher explains the problem:

TEACHER So in this experiment we’ve got three sets of data, okay? So, this one’s
going to kind of break the rules a tiny bit. The easiest way for us to do this is
that you’re going to have […] ‘time’ on the X, ‘temperature’ on the Y, and four
different lines. The four lines you’re going to draw is one line for 15 degrees,
one line for 30 degrees, one line for 60 degrees and one line for 90.

The content of discussion – rules about locating variables on axes – remains weakly
integrated with what the experiment reveals about Earth’s seasons and so embod-
ies weaker positional autonomy. However, the purpose is to create a graph that
can show this knowledge, embodying stronger relational autonomy. As portrayed in
Figure 2.5, this shifts classroom practice into an introjected code.13
This introjected code is maintained throughout the graphing activity. While
students apply the adapted ‘graphing rules’ to their results, the teacher alternates
between addressing the whole class and advising individual students; for example,
to the class:

TEACHER All right! Along the X-axis, there will be three values: the X-axis has
your time on it. There will be a time for five minutes, there will be a time for
 Targeting science 41

two and a half, and there will be a time for ‘initial’, which we can call zero, zero
minutes. Okay? …

Then (continuing straight on), she looks at a student’s workbook and asks:

TEACHER Why is this word here?

STUDENT ‘Angle’.
TEACHER We are not doing ‘angle’ like that.
STUDENT Oh, whoops!
TEACHER Just follow what’s going on here. This is the X.
STUDENT Okay.
TEACHER Okay? So ‘temperature’ does not belong there. X along here is ‘time’.Y
along here is ‘temperature’.

Discussion continues along these lines for the next 12 minutes while students draw
their graphs. As these quotes illustrate, the content of discussion involves locating
variables on axes, setting ranges for variables, sizing the graph, using symbols, label-
ling, evenly spacing intervals, creating a key for symbols, and avoiding overlapping
lines. Content is thus not related to what the results of the experiment might reveal
about Earth’s seasons: weaker positional autonomy. However, the purpose is to cre-
ate graphs which help show what the experiment might reveal about Earth’s season:
stronger relational autonomy. As the teacher explains to the whole class: ‘This is a
better way of presenting the data than it is to look at a table. … Straight away when
you look at this graph, you can see which one has increased in temperature fastest.’
Graphing thus manifests here as an introjected code (Figure 2.5).
Once students have completed graphing, the teacher shifts classroom practice
back to her sovereign code. Students write in their workbooks a ‘conclusion’ of

FIGURE 2.5 Shift from exotic code to introjected code

42 Karl Maton and Sarah K. Howard

what their graphs show and whether this supports or refutes their previous hypoth-
esis and a ‘discussion’ of whether their results were consistent and expected, what
errors occurred, how they could improve their experimental design, and what they
could do to test the idea further. The teacher then leads a discussion of the graphs
that continues this concern with what they reveal about the focus of the experiment:

TEACHER What did we learn? […]

STUDENT We learned that the steeper the angle, the hotter the temperature.
TEACHER Good. The steeper the angle of the block, we got a greater increase in
our temperature. Who can tell me why? Why did it get hotter? …
STUDENT Because the core of the block is closer to the light.
TEACHER Good. The middle part of the block, as you increase the degrees, makes
it closer to the light. Good.
STUDENT Because it’s getting more direct rays when it’s on a higher angle as
opposed to when it’s on …
TEACHER Good. When we have a higher angle, we have more of those light rays
striking the block, and those light rays then can heat up the block more effec-
tively than the ones that are just skimming over the top.

This shifts the content and purpose to exploring the results of the experiment.
The graphing activity has now been integrated with the experiment. The teacher
then consolidates this sovereign code to integrate the experiment into discussion
of Earth’s seasons. First, she emphasizes her target purpose – stronger relational
autonomy:

TEACHER Okay, but what’s the point in doing this? Are we really interested in
whether or not blocks can heat up with a lamp?
STUDENT No!
TEACHER No? Who can remember the word I used to describe what this experi-
ment was? Starts with an ‘m’.
STUDENT A model?
TEACHER A model. Fabulous. This was a model. It was a model of the Earth and
the sun.

Second, she explains differences between the model and reality and how those
differences shape the experience of heat on Earth, content that embodies stronger
positional autonomy:

TEACHER Does the Earth change its angle?

STUDENTS Yeah.
STUDENTS No!
STUDENTS It rotates.
TEACHER It rotates – good. When the Earth rotates the angle changes. … When
the Earth rotates, we change the angle that the sunlight is striking the Earth.
 Targeting science 43

FIGURE 2.6 Shift from introjected code to sovereign code

The teacher then segues to an animation that shows the Earth rotating, sunlight
striking the surface, and how this creates night and day. As portrayed in Figure 2.6,
the teacher has shifted to content from the syllabus for the purpose of learning that
syllabus – her sovereign code. The experiment and graphing activity are now inte-
grated into the wider discussion of Earth’s seasons.

Summary: separate, repurpose, integrate

Analysis of workbooks from this lesson suggests that students successfully trans-
lated their tables of experimental results into graphs and their graphs into conclu-
sions about the effects of the angle of sunlight on temperature. This is no simple
feat. Studies of science education widely report that many secondary school and
university students struggle with understanding and interpreting graphs (Planinic
et al. 2012). Indeed, the students here were creating graphs from data. Moreover, the
teacher is also laying foundations for students’ future learning. As she highlighted
in an interview, her Year 7 students ‘have no experience with graphing for science
or they’ve got no experience with drawing tables for science – we’ve really got to
teach that stuff in the beginning, because then we expect them to follow it through’
subsequent years of school science.
In terms of the knowledge involved, this learning was supported by teaching
which traced an autonomy tour from ‘science’ through graphing and back to ‘sci-
ence’. As shown by Figure 2.7, classroom practice went through:

(1) the teacher’s sovereign code by discussing results of the experiment

(2) an exotic code when discussing ‘graphing rules’
44 Karl Maton and Sarah K. Howard

FIGURE 2.7 Autonomy tour with graphing results of an experiment

(3) an introjected code when adapting those ‘rules’ to graphing results of the
experiment
(4) the teacher’s sovereign code when translating graphs into conclusions about what
was modelled by the experiment

Given that graphing could be discussed in ways that locate it either within or
beyond the teacher’s target, one question this pathway raises is why she chose to
leave her sovereign code. Students had learned the ‘graphing rules’ in previous les-
sons, so the teacher could have treated them as part of her target – as already inte-
grated into ‘science’ – by relating them directly to the experiment’s result. However,
this strategy would not have captured the distinct nature of this experiment – that
there are three variables. By constructing the ‘graphing rules’ as an exotic code, the
teacher keeps that knowledge separate from the specific ‘science’ content, allowing
her to connect to ideas that students have already learned in a way that highlights
how graphing will be different here. By separating this ‘mathematics’ from the ‘sci-
ence’, she is able to select ideas from the ‘graphing rules’, repurpose those ideas, and
integrate their use into her target knowledge. Separation comprises a shift to an
exotic code. Repurposing involves strengthening relational autonomy by turning
the ‘rules’ to the purpose of graphing results from this experiment – an introjected
code. Integration involves strengthening positional autonomy to translate the resul-
tant graphs into knowledge about Earth’s seasons – a sovereign code. In our previ-
ous example, leaving ‘science’ for ‘maths’ was a one-way trip that failed to integrate
the ‘maths’ back into ‘science’. Here, leaving ‘science’ was a precursor to successful
integration through an autonomy tour.
 Targeting science 45

Conclusion
A key challenge faced by research into science education is developing teaching
practices that select, recontextualize and integrate mathematical knowledge to sup-
port the learning of scientific knowledge. We argued that existing approaches offer
insights into student learning of ways of knowing but typically sideline teaching
practice and the forms of knowledge being taught and learned. To help address
these blind spots we offered a complementary approach centred on autonomy codes,
pathways, and targets. The concepts of autonomy codes foreground the forms taken by
knowledge practices expressed in classroom discourse, focusing on a key relevant
feature for integration: their relations with other knowledge practices. Analyzing the
pathways traced by successive codes reveals how different teaching practices enable
or constrain the integration of ‘mathematics’ into ‘science’. Using these concepts, we
analyzed contrasting examples of classroom practice from secondary school lessons
in science that suggest autonomy tours support and one-way trips obstruct the integra-
tion of ‘mathematics’ into ‘science’.
One implication of the analysis is that pathways to successful integration are not
necessarily direct. In the tour example, the separation of ‘mathematical’ ideas from
‘scientific’ knowledge was an important precursor to its subsequent integration.
This speaks to an issue highlighted by science education research: student difficul-
ties with recognizing they are exploring mathematical ideas when presented in
scientific contexts (e.g. Planinic et al. 2012). Constructing the ‘mathematics’ knowl-
edge as separate (exotic code) is an opportunity to highlight the specific constel-
lations of meanings within which that knowledge is located and which underpins
its ‘mathematical’ nature. Turning those ideas to a ‘scientific’ purpose (introjected
code) and connecting those repurposed ideas to ‘scientific’ knowledge (sovereign
code) recontextualizes the ideas within a new constellation of meanings. A tour thus
offers the possibility of making explicit those constellational differences. It makes
the knowledge visible to students. As a growing body of research is showing, not all
tours involve this specific combination of autonomy codes, but all involve departing
and returning. If classroom discourse remained within a sovereign code through-
out, these constellational differences would not be made visible; and if classroom
discourse did not return, the recontextualization of ideas between constellations
would not be possible.
While ‘autonomy codes’ help bring knowledge practices into the picture, we
should emphasize that the concepts are not limited to that focus – one can also ana-
lyze students’ dispositions, interactions and changing understandings. In short, the
concepts can be enacted to examine both forms of knowledge practices and ways of
knowing. This would enable ‘matches’ and ‘clashes’ to be identified, supporting the
development of appropriate pedagogic practices.
Autonomy codes are, of course, not the only feature of knowledge practices, and
autonomy tours are not the only factor in integration. As mentioned earlier, a key
issue highlighted by physics education research is that science involves ‘learning
46 Karl Maton and Sarah K. Howard

to blend physical meaning into mathematical representations and use that physi-
cal meaning in solving problems’ (Redish 2017: 25). This can be traced through
autonomy pathways: the second teacher condenses ‘mathematical’ ideas with empir-
ical meanings when repurposing ‘graphing rules’ (from exotic code to introjected
code) and when relating the resultant graphs to explaining Earth’s seasons (from
introjected code to sovereign code). In contrast, the ‘maths’ of the first teacher
remains disconnected from empirical referents. However, this changing attribute is
not directly conceptualized by autonomy codes. For this one can draw on the LCT
dimension of Specialization to conceptualize relations with empirical referents in
terms of epistemic relations and to reveal that integration of ‘mathematics’ into ‘sci-
ence’ involves processes of ontic condensation (Maton 2014: 175–84; Wolff 2017).
Nonetheless, autonomy codes offer a valuable start for bringing knowledge prac-
tices into view and autonomy tours may represent a key to pedagogically integrating
mathematics into science learning. Using the notion of targets to enact these con-
cepts resolves a major obstacle to generating a general model of pedagogic inte-
gration: the problem of defining ‘mathematics’ and ‘science’ when the constitutive
features of each subject, and relations between them, vary across contexts.‘Targeting’
these constitutive features in translation devices addresses the contextual and chang-
ing nature of whether practices are constructed as ‘mathematics’ or as ‘science’.
‘Targeting’ also allows studies to translate the specificities of each empirical context
into concepts capable of generating a general model. We can examine, for example,
the role of ‘autonomy tours’ in integration, howsoever ‘science’ and ‘mathemat-
ics’ are defined. By targeting science in this way, we can indeed hit a bullet with a
smaller bullet, while blindfolded and riding a horse.

Notes
1 Research into science education is divided into ‘science education research’ for schooling
and disciplinary specialisms (such as ‘physics education research’) at university level. In
our empirical examples from secondary schooling, ‘science’ is taught, but our argument
is not limited to one discipline or level of education.
2 Work discussing ‘disciplinary discourse’ (e.g. Airey and Linder 2009) points towards
knowledge but reduces this ‘discourse’ to representations of more fundamental ‘ways of
knowing’, leading again to studies of student ‘fluency’ in ways of knowing – subjectivism
returns.
3 In LCT this is to say that production fields, recontextualization fields, and reproduction fields
have their own distinctive logics (Maton 2014: 47–52).
4 We must emphasize ‘complemented’: to replace studies of ways of knowing with analysis
of forms of knowledge would be to continue taking part of the picture for the whole.
Both knowledge and knowing are significant.
5 This study was funded by the Australian Research Council (DP130100481) and led by
Karl Maton, J. R. Martin, Len Unsworth and Sarah K. Howard.
6 These examples were introduced in Maton and Howard (2018) and more extensively
analysed here.
7 In Chapter 4 of this volume we focus on how teachers integrate a multimedia object into
a specific task, so each teacher’s target is the lesson and their core target is the specific task.
 Targeting science 47

8 During data collection the curriculum authority was the New South Wales Board of
Studies. Though renamed the New South Wales Education Standards Authority (NESA),
its syllabus remains the same at the time of publication. All ‘NESA’ quotes are from
https://syllabus.nesa.nsw.edu.au/stage-4-content/.
9 All quotes in this paragraph are from NESA.
10 In Figure 2.3 the shift is to just inside the exotic code because content and purpose con-
cern educational knowledge or associated non-targets (PA–, RA–).
11 All quotes in this paragraph are from NESA.
12 Both content and purpose may be non-target but still concern educational knowledge,
so embody an associated exotic code.
13 The pathway in Figure 2.5 shifts to the far right, indicating extremely strong relational
autonomy, because creating a graph for the experiment’s results is within the teacher’s
inner-core target purpose.

References
Airey, J. and Linder, C. (2009) ‘A disciplinary discourse perspective on university science
learning: Achieving fluency in a critical constellation of modes’, Journal of Research in
Science Teaching, 46(1): 27–49.
Bing,T. J. and Redish, E. F. (2007) ‘The cognitive blending of mathematics and physics knowl-
edge’, American Institute of Physics Conference Proceedings, 883(26): 26–29.
Bing, T. J. and Redish, E. F. (2009) ‘Analyzing problem solving using math in physics:
Epistemological framing via warrants’, Physical Review Special Topics – Physics Education
Research, 5(020108): 1–15.
Blackie, M. (2014) ‘Creating semantic waves: Using Legitimation Code Theory as a tool to
aid the teaching of chemistry’, Chemistry Education Research and Practice, 15: 462–69.
Conana, H., Marshall, D. and Case, J.M. (2016) ‘Exploring pedagogical possibilities for trans-
formative approaches to academic literacies in undergraduate physics’, Critical Studies in
Teaching and Learning, 4(2): 28–44.
di Sessa,A. (1993) ‘Toward an epistemology of physics’, Cognition and Instruction, 10(2/3): 105–225.
Doran,Y. J. (2018) The Discourse of Physics, London: Routledge.
Fredlund, T., Airey, J. and Linder, C. (2012) ‘Exploring the role of physics representations: An
illustrative example from students sharing knowledge about refraction’, European Journal
of Physics, 33: 657–66.
Georgiou, H., Maton, K. and Sharma, M. (2014) ‘Recovering knowledge for science educa-
tion research: Exploring the ‘Icarus effect’ in student work’, Canadian Journal of Science,
Mathematics, and Technology Education, 14(3): 252–68.
Hammer, D. (2000) ‘Student resources for learning introductory physics’, American Journal of
Physics, 68(7 PER Suppl.): S52–S59.
Hammer, D. and Elby, A. (2002) ‘On the form of a personal epistemology’, in B. K. Hofer and
P. R. Pintrich (eds) Personal Epistemology, Mahwah, NJ: Erlbaum, 169–90.
Lemke, J. (1998). ‘Multiplying meaning: Visual and verbal semiotics in scientific text’, in
J. R. Martin and R.Veel, Reading Science, London: Routledge, 87–113.
Maton, K. (2014) Knowledge and Knowers: Towards a Realist Sociology of Education, London:
Routledge.
Maton, K. (2016) ‘Starting points: Resources and architectural glossary’, in K. Maton, S. Hood
and S. Shay (eds) Knowledge-Building, London: Routledge, 233–43.
Maton, K. and Chen, R. T.-H. (2016) ‘LCT in qualitative research: Creating a translation
device for studying constructivist pedagogy’, in K. Maton, S. Hood and S. Shay (eds)
Knowledge-Building, London: Routledge, 27–48.
48 Karl Maton and Sarah K. Howard

Maton, K. and Howard, S. K. (2018) ‘Taking autonomy tours: A key to integrative knowl-
edge-building’, LCT Centre Occasional Paper 1 (June): 1–35.
Meli, K., Zacharos, K. and Koliopoulos, D. (2016) ‘The integration of mathematics in physics
problem solving: A case study of Greek upper secondary school students’, Canadian Journal
of Science, Mathematics and Technology Education, 16(1): 48–63.
O’Halloran, K. (2010) ‘Mathematical and scientific forms of knowledge: A systemic func-
tional multimodal grammatical approach’, in F. Christie and K. Maton (eds), Language,
Knowledge and Pedagogy, London: Continuum, 205–36.
Parodi, G. (2012) ‘University genres and multisemiotic features: Accessing specialized knowl-
edge through disciplinarity’, Forum Linguistico, 9:4: 259–82.
Planinic, M., Milin-Sipus, Z., Katic, H., Susac, A. and Ivanjek, L. (2012) ‘Comparison of stu-
dent understanding of line graph slope in physics and mathematics’, International Journal of
Science and Mathematics Education, 10: 1393–414.
Quinnell, R., Thompson, R and LeBard, R.J. (2013) ‘It’s not maths; it’s science: Exploring
thinking dispositions, learning thresholds and mindfulness in science learning’, International
Journal of Mathematical Education in Science and Technology, 44(6): 808–16.
Redish, E. F. (2014) ‘Oersted lecture: How should we think about how our students think?’,
American Journal of Physics, 82: 537–51.
Redish, E. F. (2017) ‘Analysing the competency of mathematical modelling in physics’, in,
T. Greczyło and E. Debowska (eds) Key Competences in Physics Teaching and Learning,
Switzerland: Springer, 25–40.
Redish, E. and Kuo, E. (2015) ‘Language of physics, language of math: Disciplinary culture
and dynamic epistemology’, Science and Education, 24: 561–90.
von Glasersfeld, E. (1995) Radical Constructivism, Washington, DC: Falmer.
Wolff, K. (2017) ‘A language for the analysis of disciplinary boundary crossing: Insights from
engineering problem-solving practice’, Teaching in Higher Education, 23(1): 104–19.
3
CONSTELLATING SCIENCE
How relations among ideas help build
knowledge

Karl Maton and Y. J. Doran

Comprising a simple explanation of a complex phenomenon,

namely complex explanations of seemingly simple phenomena.

Introduction
Science is complex. This is not simply a matter of the extraordinarily large number
of ideas, practices and beliefs that comprise science; it also concerns the manifold
ways in which they are related together. These complex relations pose challenges
for teaching science. Even after isolating a topic and deciding the level of detail to
teach, questions arise of where to begin and how to proceed through the diverse
connections among ideas that constitute a scientific explanation. To address these
questions requires an understanding of different kinds of relations among ideas.Yet,
there remains a need for concepts that reveal and analyze those relations, thanks to
knowledge-blindness and atomism.
‘Knowledge-blindness’ (Maton 2014) describes the way the forms taken by
knowledge remain unseen by research. Most approaches instead explore either kinds
of knowers or ways of knowing. Sociologically-inflected approaches emphasize that
‘knowledge’ reflects dominant interests and focus on whose knowledge is taught and
learned (Ellery 2017), a concern with kinds of knowers that is increasingly salient
in calls to ‘decolonize’ science (Adendorff and Blackie 2020). More common in
science education are psychologically-inflected approaches that assume ‘knowledge’
comprises mental processes and so focus on ‘conceptions’ and ways of thinking
(see Georgiou et al. 2014). For example, the notion of ‘threshold concepts’ appears
to highlight relations among ideas: they are ‘concepts that bind a subject together’
(Land et al. 2005: 54). However, the nature of those relations is not analysed. A
‘threshold concept’ is defined as ‘opening up a new and previously inaccessible way
of thinking about something’ (Meyer and Land 2003: 1). It is identified by being
50 Karl Maton and Y. J. Doran

‘transformative’, ‘integrative’, ‘irreversible’, and often ‘troublesome’ for students and

distinguished from a ‘core concept’ that ‘progresses understanding of the subject’
without transforming students’ ways of thinking (Meyer and Land 2003: 4). Leaving
aside their ill-defined nature (Salwén 2019), these concepts focus entirely on ways
of knowing. The forms taken by ‘core’ and ‘threshold’ concepts are not part of the
picture. Taken together, that research is concerned with kinds of knowers or ways
of knowing means that knowledge as an object of study in its own right, one that
takes different forms, forms which have effects for all kinds of issues, is left out of
the picture.
The second reason, atomism, describes how knowledges (and ways of knowing)
are viewed atomistically, as if theories, explanations, etc. are simply collections of
individual ideas. For example, diSessa (1993) defines ‘phenomenological primitives’
or ‘p-prims’ as knowledge in physics that students intuitively believe to be irreduc-
ible features of reality. As well as focusing on knowing, this notion illustrates atom-
ism: how these ideas relate to other ideas to form an explanation remains unclear. As
diSessa states, the approach views ‘knowledge in pieces’ (1993: 111). This is akin to
listing ingredients but not describing the recipe, as if how ingredients are combined
does not affect what is created. Such atomism also characterizes typologies – such
as Shulman’s ‘PCK’, ‘TPCK’, and Bloom’s taxonomy – which list different kinds of
‘knowledge’ but do not conceptualize relations among those types.
Given these tendencies to knowledge-blindness and atomism, there remains a
need for concepts to grasp different relations among ideas. This issue was helpfully
opened up by systemic functional linguists exploring explanations in science (e.g.
Rose et al. 1992, Unsworth 1997). Such work showed that ideas in explanations are
connected in language through a range of relations of condition, cause and time.
However, the issue of how relations among ideas affect how they are explained
remains unclear. In this chapter, we draw on constellations from Legitimation Code
Theory (Maton 2014) to conceptualize and visualize relations among ideas. We
focus on analyzing explanations, a key genre in science education (Unsworth 1997).
Our aim is to illustrate how constellation analysis can show the significance of rela-
tions among ideas for science teaching.
We begin by introducing the notion of constellations.We then make several simple
conceptual distinctions relevant to our analyses in this chapter. These concepts are
enacted in analyses of scientific explanations of tides and seasons, focusing on Year
7 of secondary school science in New South Wales, Australia. Analyzing textbooks
created for this curriculum, we examine the logical relations among ideas in each
explanation. Then, we examine how each explanation is taught by the same teacher
in the same unit of study. From these analyses we argue that relations among ideas
in the logic of explanations may affect how those explanations are taught in a school
classroom. We conclude by considering how constellation analysis can provide edu-
cators with a pedagogic tool for making visible relations among the ideas they
are teaching, and researchers with an analytic tool for making visible how knowl-
edge changes when recontextualized between research, curriculum and classroom
practice.
 Constellating science 51

Constellations
A way of grasping relations among ideas is provided by constellation analysis,
which forms part of cosmological analysis from Legitimation Code Theory (LCT).
Cosmological analysis describes any set of stances (e.g. ideas, beliefs, practices, etc.)
as a selection from a larger set of possible stances that has been arranged into a par-
ticular pattern or constellation, condensed with meanings, and charged with valuations,
according to a particular cosmology or worldview (Maton 2014: 148–70).

Analogy from astronomy

As a way into the approach, consider the notion of ‘constellations’ in astronomy. On
a clear night without light pollution one can see an enormous number of stars. As
an example, the left image in Figure 3.1 shows some of the stars visible in a small
part of the sky in the northern hemisphere. Of those stars, a small number have been
selected and arranged into a pattern that is the constellation of ‘Taurus’ (middle in
Figure 3.1). There may also be smaller clusters; for example, ‘Pleiades’ is a cluster of
stars that lie within Taurus. As well as being constellated, the stars are condensed with
meaning. For example, since the ancient Mesopotamians,Taurus has been associated
with the image of a bull (right in Figure 3.1). These meanings are charged positively,
neutrally or negatively, to varying degrees. For example, in modern astrology Taurus
signals such attributes as ‘creativity’, ‘affectionate’, and ‘grasping’.
The selection, arrangement, meaning and valuation of constellations may
vary across place and over time. For example, some stars are visible only from
specific locations, but this may change – ‘The Southern Cross’ was visible in
the northern hemisphere before the fifth century. The meanings of a constel-
lation may also change. For example: the Zuni people of New Mexico call the
Pleiades cluster ‘seed stars’, as their position was traditionally used to determine
the time of the year to plant seeds; and Pleaides is the logo of the car manu-
facturer Subaru, whose advertising attempts to associate the symbol with such
notions as ‘reliability’. The constellations themselves also vary, reflecting differ-
ences in cosmology or worldview. For example, Figure 3.2 shows Aldebaran, a star
in Taurus, located within constellations from (left to right) Inuit, Korean and
Maori cultures.1

FIGURE 3.1 Taurus – some stars before constellating; stars constellated into Taurus; over-
laid with image of bull (Northern hemisphere view)
52 Karl Maton and Y. J. Doran

FIGURE 3.2 Aldebaran in constellations from Inuit, Korean and Maori cultures (hemi-
sphere view is that of each culture)

LCT concepts
There are limits to any analogy, but the astronomy example offers a way into under-
standing the LCT approach to analyzing practices.2 Cosmological analysis is centred
on the five words italicized above:

• clusters are groupings of nodes (such as ideas, beliefs or practices);

• constellations are larger grouping of nodes that may include clusters;
• condensation is how nodes, clusters and constellations are imbued with meanings;
• charging describes the valuations given to nodes, clusters and constellations; and
• cosmology refers to the organizing principles underlying the selection, arrange-
ment and valuation of nodes in a constellation, which are revealed by analyzing
their legitimation codes.

Reflecting its sociological nature, LCT holds that a constellation has coherence
from a particular point in social space and time, to actors with a particular cosmol-
ogy, and that cosmologies (and so constellations) are subject to contestation, vary
across contexts, and change over time (Maton 2014: 148–70).
This description is intentionally abstract, to reflect the wide applicability of cos-
mological analysis. The constellation being analyzed may be a scientific theory, reli-
gion, political system, ideology, sport, dance, song, machine, etc. – its nodes may be
ideas, beliefs, institutions, body movements, sounds, machine parts, etc. Similarly,
analysis may focus on many different issues about these constellations, using dif-
ferent cosmological concepts. Here we shall explore relations among stances using
only the concepts of clusters and constellations. We shall not explore the nature of the
meanings being related (using condensation and charging) nor reveal the underlying
principles generating, maintaining and changing stances (cosmologies). Our focus is
a constellation analysis of how stances are related together, rather than a cosmological
analysis of the organizing principles underlying those relations.3
Constellation analysis can itself be used in different ways. For example, Maton
(2014: 148–70) illustrates a synchronic analysis of ideas. Analyzing claims by advo-
cates of constructivism, Maton shows they construct two constellations of stances
on curriculum, pedagogy, assessment and many other issues, that are given such
labels as ‘student-centred’ and ‘teacher-centred’. Relations within each constellation
 Constellating science 53

are constructed as essential: choosing one stance is viewed as choosing all other
stances in the same constellation. Relations between the constellations are con-
structed as oppositional: one cannot select stances from both. The two constel-
lations are also portrayed as the only options available. Maton argues that these
relations have effects, such as narrowing what is viewed as possible in education
by excluding stances either outside or spread between the constellations. Similarly,
Glenn (2016) analyzes beliefs about climate change, showing how different groups
of people constellate different ideas together and charge those ideas differently, in
ways which mean they are more or less open to scientific evidence for climate
change.
Another way of using constellation analysis is to explore how stances are selected,
linked and given meaning over time to build practices. For example, Lambrinos
(2020) reveals how ballet teachers bring together sets of behaviours, dispositions and
movements to teach both dance and how to be a dancer. Such analysis can show
how clusters or even whole constellations are condensed into a new node, which
can then be constellated with more nodes. Figure 3.3, for example, illustrates how
a ballet teacher teaches an exercise called ‘springs’ by linking the instruction ‘jump’
with other instructions (‘sink’, ‘feet in 1st position’, etc.), where ‘jump’ (solid black
in Figure 3.3) has itself been condensed with meanings (‘powerful’, ‘straight legs’,
etc.). In this way, Lambrinos shows how words, gestures and movements are brought
together to create complex constellations. Other studies focus on the building of
values in texts. For example, Tilakaratna and Szenes (2020) show how successful
student ‘reflective writing’ assignments cluster and condense meanings to align with
disciplinary values, and Doran (2020) reveals the rhetorical strategies of a highly
influential text that cluster, condense and charge values to effect change within an
intellectual field.

FIGURE 3.3 Example of constellating and condensing movements in ballet teaching

(Lambrinos 2020: 91)
54 Karl Maton and Y. J. Doran

However, constellation analysis is still in its infancy. One area for development is
identifying different kinds of relations within a constellation. Thus far, studies have
mostly focused on differences among constellations in terms of their cosmologies
or underlying legitimation codes (Maton 2014: 148–70). As yet few studies explore
different relations within a constellation.4 In this chapter we contribute to exploring
this issue.

Three simple distinctions

To illustrate the significance of different relations among ideas, we shall make three
simple distinctions relevant to our analyses in this chapter, between ‘independent’
and ‘dependent’ links, ‘base’ and ‘supplementary’ clusters, and ‘assembling’ and
‘aggregating’ when building constellations.
First, we distinguish between:

• independent links, where the meanings of a node or cluster are independent of

other nodes or clusters in the constellation; and
• dependent links, where the meanings of a node or cluster depend on relations to
other nodes or clusters in the constellation.

For example, in the explanation of tides (below), the node ‘The Earth is divided into
water and solid components’ does not depend on other nodes in the explanation,
generating independent links with those nodes. It does not depend, for example,
on the idea that ‘The Moon has gravity, which is stronger the closer things are to
the Moon’. This is not to suggest that the ideas expressed by such nodes are always
independent; they will be related to ideas in other constellations, such as explanations
of the Earth, planets, etc. However, this highlights where nodes create independent
links with other nodes within the constellation in question. In contrast, in the tides
explanation a node that describes the different gravitational pulls of the Moon on
the water and solid components of the Earth links to two other nodes in the expla-
nation – ‘The Earth is divided into water and solid components’ and ‘The Moon has
gravity, which is stronger the closer things are to the Moon’ – by describing their
implications when combined. It thereby generates dependent links with those two
nodes within the constellation.5
Second, we distinguish between:

• a base cluster or set of nodes that creates a basic version of the constellation; and
• supplementary clusters or sets of nodes that serve to elaborate, augment and refine
the base cluster.

For example, in the explanations of tides, the textbooks and teaching we analyze
draw on four nodes to create a basic explanation of tides (the base cluster), and
elaborate on this explanation by employing other clusters of nodes to explain daily
variation in tides, and ‘spring’ and ‘neap’ tides (two supplementary clusters).Together
 Constellating science 55

all three clusters constitute the explanation, but one cluster offers a ‘base’ on which
the other clusters elaborate. This is not a distinction between fundamental and
peripheral, essential and inessential nodes, but rather highlights two roles played by
clusters of nodes in the explanations we analyze in this chapter.
Third, we distinguish between two ways that constellations unfold over time:

• assembling, where nodes and clusters develop in a linear and incremental fash-
ion; and
• aggregating, in which nodes or clusters are developed separately, in a multilinear
fashion.

In assembling the constellation is likely to grow from one origin; in aggregating the
constellation involves multiple separate parts, each of which may be assembled on
its own before being combined.6 As we shall illustrate, the tides explanation appears
to lend itself to assembling nodes and clusters in a linear manner, while the seasons
explanation has a range of potential ways of aggregating nodes and clusters together.
These three distinctions are not the only relations within constellations created
by scientific explanations, let alone within constellations generally. Our aim is not to
conceptualize all relations among ideas but to make intentionally simple distinctions
that demonstrate a simple point: different relations among ideas matter. Conversely,
not all constellations involve these distinctions: they may comprise only dependent
or independent nodes, have no base or supplementary clusters, and remain static.
The distinctions are thus not generic characteristics – they reflect our specific focus
and aim.

Analyzing two explanations: Of tides and seasons

Our focus is on analyzing two scientific explanations. Our aim is to show that,
though they appear similar, the explanations involve different relations among ideas
that may be significant for how they are taught. Both are core topics in Year 7
of secondary school science in New South Wales (NSW), Australia. Both concern
widely-known phenomena in the natural world: the tides and the seasons on Earth.
Both phenomena appear prima facie simple: water reaches higher and lower through
the day (tides) and the temperature goes higher and lower over the year (seasons).7
However, we shall show that explanations of tides and seasons involve complex
constellations of ideas brought together in distinctive ways.
We analyze each explanation in two ways. First, we create schematic constellations
of the logic of each explanation according to textbooks aimed at Year 7 secondary
school science in NSW. This is not analyzing how textbooks sequence their expla-
nations or build constellations; it does not describe a specific textual or pedagogic
expression. Rather, a ‘schematic constellation’ is a synchronic representation of key
nodes and how they are logically related – a snapshot of the logic of the explanation.
To create the constellation, we identify nodes and links that the textbooks contain
or assume in order to make sense.8 The resulting constellation diagrams are akin
56 Karl Maton and Y. J. Doran

to transit maps that show not specific journeys but rather the roads, stations, con-
nections, routes, etc. They are a composite from analyses of textbooks known as:
Oxford Insight (Zhang et al. 2013), Nelson iScience (Bishop et al. 2011), Pearson
(Rickard 2011), Core Science (Arena et al. 2009), and Science World (Stannard and
Williamson 1995).
Second, we create pedagogic constellations showing how each explanation is taught
by the same teacher during the same unit of secondary school science.9 Our focus
is to explore whether the logic might shape the pedagogic, that is whether relations
among ideas in the logic of an explanation might affect how that explanation is
taught. To return to the metaphor, these pedagogic constellations describe specific
journeys across the terrain shown by the schematic constellations.
This analysis does not, of course, offer a comprehensive account of how tides and
seasons are explained in schooling. As stated above, constellations often vary across
time and space. In other contexts, different explanations may include more or fewer
nodes and different links. The analysis is thus limited to our empirical examples.
Moreover, our examples are not intended to demonstrate best practice and whether
the explanations are accurate or complete is not our concern. Our aim is simply to
explore how relations among ideas are expressed and their potential significance.
In summary, we shall argue that the schematic constellation for tides is less com-
plex than that for seasons. One effect of this difference is, we suggest, shown by
pedagogic constellations: the tides explanation lends itself to an assembling form
of teaching that proceeds in a linear fashion through successive clusters, while the
seasons explanation lends itself to an aggregating form of teaching that proceeds in
a more patchwork fashion. We conjecture that the forms taken by relations among
ideas in the logic of an explanation (schematic constellation) may shape how the
explanation is likely to be taught (pedagogic constellation). To reach these conjec-
tures, we analyze tides and of seasons, in turn.

Explaining tides
Schematic constellation of textbooks
From analyses of sections on ‘tides’ in the five textbooks, we developed a composite
of key nodes in their explanations. We begin by summarizing these nodes as simply
as possible, in our own words. In bold are words included in the diagrams that follow.

A. The Earth is divided into water and solid components.

B. The Moon has gravity, which is stronger the closer things are to the Moon.
C. Together node A (water and solid) and node B (Moon’s gravity) mean that the
Moon’s gravitational pull is strongest for the water on the Earth closest to the
Moon, less strong for the Earth’s solid, and weakest for the water on the Earth
furthest from the Moon.
D. Node C produces bulges of water on the parts of the Earth that are closest and
furthest away from the Moon, which we experience as ‘high tides’, and no
 Constellating science 57

bulges on the parts of the Earth that are neither closest nor furthest, which we
experience as ‘low tides’.
E. The Earth rotates.
F. Earth’s rotation (node E) combined with the bulges of water created by the
Moon’s gravity (node D) leads to the experience of daily variation of tides as
the Earth moves through the bulges.
G. The Moon orbits the Earth.
H. The Sun’s gravity pulls on the water and solid components of the Earth.
I. Combining the Moon’s orbit of the Earth (node G) and the Sun’s gravitational
pull (node H) with the daily variation of tides (node F) leads to variation in the
size of tides. When the Sun and Moon line up, the tides vary the most (highest
highs and lowest lows), which is known as the ‘spring tide’; and when the Sun
and Moon are perpendicular, the tides vary the least (with the lowest highs and
highest lows), which is known as the ‘neap tide’.

Figure 3.4 represents this description as a constellation diagram. Nodes which gen-
erate independent links are in squares and nodes which generate dependent links are in

FIGURE 3.4 Schematic constellation of the tides explanations across textbooks

58 Karl Maton and Y. J. Doran

circles. Dependent links are shown by lines with arrows that indicate the direction
of implication.

Links, clusters and form

Though ‘tides’ may seem a simple empirical phenomenon, its explanation is rela-
tively complex, involving different links and clusters. In terms of links, the sche-
matic constellation is characterized by both independent and dependent relations
among nodes. As Figure 3.4 shows, the schematic explanation begins with two
nodes that create independent links: the Earth’s components of water and solid
and the Moon’s gravity are not dependent on other nodes in this explanation. In
contrast, how the Moon’s pull affects Earth’s components comprises implications
of bringing together those two nodes, generating dependent links (shown by the
arrow). The node introducing high and low tides outlines the implications of the
Moon’s pull for creating bulges of water. This reaches the notion of ‘tides’. That the
Earth rotates is independent. The daily variation of tides creates a dependent
link by describing implications of bringing together the Earth’s rotation with
high and low tides. That the Moon orbits the Earth and the Sun’s gravity are
both independent.To reach the notions of spring and neap tides, these two nodes
are combined with daily variation, creating dependent links with all three nodes.
In summary, a series of independent nodes establish factors or phenomena and
dependent nodes relate those phenomena together.
In terms of clusters, the schematic constellation for the textbooks’ explanation of
tides comprises a base cluster and two supplementary clusters. As shown in Figure 3.5a, the
top four nodes form a base cluster that creates a basic explanation of tides. The two
supplementary clusters are successive in that each progresses from a logically preceding
node. As shown by Figure 3.5b, the cause of high and low tides is the starting point

FIGURE 3.5 Schematic constellation for tides – base and supplementary clusters
 Constellating science 59

for a supplementary cluster with the Earth’s rotation that augments the basic expla-
nation to reach daily variation in tides. As shown by Figure 3.5c, daily variation
is part of a second supplementary cluster with the Moon orbits and Sun’s gravity
that elaborates the preceding explanation to describe spring and neap tides.
In terms of its form, the schematic constellation does not show how the expla-
nation unfolds over time. However, it does reveal that its logical links are relatively
linear: the simple explanation of the base cluster is augmented through successive
supplementary clusters. There are no parallel clusters elaborated or large numbers
of separate nodes described which then require integration.The implications of this
relative simplicity will, however, only become apparent in contrast with our analysis
of ‘the seasons’, further below.

A pedagogic constellation for tides

We now turn to how an explanation of tides was taught in a secondary science
classroom in comparison to this schematic constellation. Our example is part of a
lesson within a unit on ‘Earth and space sciences’ from Year 7 of secondary school,
NSW, that explores such topics as ‘day and night’, ‘the seasons’ and ‘tides’. We focus
on a lesson phase in which the topic of ‘tides’ is introduced by the teacher and a
video is played that details an explanation.

Introducing the topic

Immediately prior to addressing ‘tides’, the teacher shows a video about the Moon and
the Earth. She then asks the class what might cause tides.A student suggests ‘the Moon’s
gravitational pull’, which was mentioned in the video, and the teacher elaborates:

Yes, very good. So, because the Moon has gravity it actually pulls the mass of
water towards itself. Wherever the Moon is, that’s where the high tide will be
and wherever the Moon isn’t, that’s where the low tide will be.

So, the teacher links the Moon’s gravity and the Earth’s water and solid compo-
sition by how the Moon’s pull affects the water and thence to a simple description
of high and low tides. She thereby provides a succinct, Moon-focused equivalent
to the base cluster outlined above (Figure 3.5a) before starting the video on tides. As
this highlights, pedagogic constellations may form part of a series in which teach-
ers and student build on preceding discussions or look ahead to future topics. They
venture onto the terrain shown by the transit map from where they have just been,
in this case discussing the Moon.

A video explanation
The teacher plays a video entitled ‘Watching the tides’ in which an astronomer
explains the causes of tides.10 The video begins by summarizing key factors:
60 Karl Maton and Y. J. Doran

Tides are caused by the gravitational pull of the Moon and the Sun and the
rotation of the Earth. The Earth is not a solid sphere like we think. It’s actually
kind of squishy. Especially this layer of water we have on the outside – the
oceans.

This is accompanied by an animation showing the Moon orbiting the Earth. It thus
begins by highlighting: the Moon’s gravity, the Sun’s gravity, the Earth’s rota-
tion, the Earth’s composition as water and solid, and the Moon’s orbit. Thus,
as shown by Figure 3.6a, the video begins by introducing the five independent
nodes.11
The video then begins creating dependent nodes that link these independent
nodes together:

So the gravity of the Sun and the Moon actually kind of squeeze or stretch the
Earth and its oceans out into a couple of bulges. One under the Moon, one on
the other side of the Earth.

This links ‘the Earth and its oceans’ (water and solid) with the Sun’s gravity and
the Moon’s gravity to explain that the oceans are pulled into bulges, while the
accompanying animation shows bulges on the parts of the Earth closest to and fur-
thest from the Moon, creating a simple version of the node stating that the Moon
pulls on Earth’s components differently. The video continues:

FIGURE 3.6 Pedagogic constellation for tides – independents and base cluster
 Constellating science 61

And as the Earth rotates over the course of the day, you, standing on the surface
of the Earth, move along with the Earth’s surface into these bulges. And we
experience that as the rising and lowering tides.

This is a simple version of high and low tides. As illustrated by Figure 3.6b, the
video has now outlined a base cluster echoing that of the schematic constellation.
There are two main differences. First, the Sun’s gravity has been recruited verbally
into the base cluster, though the animation does not include the Sun and its role is
not discussed. Second, the key attribute of the Moon’s gravity – that it is stronger
the closer things are – has not yet been mentioned. This issue is, though, immedi-
ately discussed (emphasis added):

Now it’s easy to see how on the side facing the Moon or the Sun you can get
this bulge of ocean. You can imagine the gravity pulling the oceans up into a
bob or a bubble. But it’s not as easy to understand why there’s a bulge on the
other side as well. And the easiest way to describe that is: the Moon’s gravity is
stronger, of course, the closer you get to it. So, on the side of the Earth close to the
Moon, the Moon has a stronger pull. So while the oceans on the Moon side get
pulled more strongly than the general Earth does, on the other side it’s kinda
opposite. The pull on the oceans on the far side are less than the pull on the
Earth. So that far bulge actually gets created … think of it as the Earth being
pulled out from under the oceans, a little bit.

The video thereby completes the base cluster (Figure 3.6b). The attribute of the
Moon’s gravity (in italics) has been added and brought together with the Earth’s
composition as water and solid to describes implications for the Moon’s pull on
those components, and so the creation of bulges of water or high and low tides.
In short, the video first creates a preliminary, intuitive version of the base cluster
and then uses the counter-intuitive nature of the Moon’s gravity creating a bulge
on the far side of the Earth as a way of completing that basic explanation. Next, the
video links high and low tides with the Earth’s rotation to describe implications
for daily variation of tides:

You get two high tides a day because as the Earth rotates, we rotate through
these two bulges.

As shown in Figure 3.7c, this echoes the first supplementary cluster of the schematic
constellation.The video then creates the second supplementary cluster – Figure 3.7d.
The role of the Sun’s gravity is shown in the accompanying animation and linked
with the Moon’s orbit to introduce ‘spring tide’ and ‘neap tide’:

Both the Moon and the Sun play a part in tides. Each one pulls. And when the
Sun and the Moon combine their forces – that is, when they’re both acting
together – we get much stronger tides than usual. Higher highs and lower
62 Karl Maton and Y. J. Doran

FIGURE 3.7 Pedagogic constellation for tides – supplementary clusters

lows – we call these ‘spring tides’. The name ‘spring tide’ doesn’t have anything
to do with the season of spring, but we get them about twice a month at new
Moon and at full Moon, when the Earth, the Moon and the Sun are all lined
up and the gravity of the Sun and the Moon are acting together. ‘Neap tide’ is
when the tidal effects of Sun and Moon are kind of cancelling each other out
or making each other not as extreme. And that happens around first and third
quarter phases of the Moon. The Sun is in one part of the sky and the Moon is
ninety degrees around and they’re kind of pulling in different directions. So
you get lower highs and higher lows during the neap tide.

This is the end of the video’s explanation of tides, completing the constellation.
The pedagogic constellation thereby contains the same nodes, links and clusters
as the schematic constellation. The form of constellation building here is what we
defined as assembling. After introducing the independent nodes, the video selected
and brought together, through dependent nodes, a subset to create a base cluster.
It then added another independent node and drew out its implications through a
dependent node, repeating this move to complete the explanation. It set this out in
a linear and incremental fashion, accreting new nodes and linking them to existing
nodes. A further attribute is how closely the pedagogic ordering of the nodes and
clusters in the video matches the logic of the schematic constellation. There are
many good reasons why teaching may differ from the logic of an explanation. For
example, an educator may choose to build on what they have been discussing (as the
teacher did in her introduction) or to wait before introducing an attribute in order
to begin from shared experiences or intuitive common sense (as the video did here).
 Constellating science 63

However, in this case the ordering remains remarkably similar. As we shall see, this
contrasts with the unfolding of the explanation of seasons.

Explaining seasons
Schematic constellation of textbooks
From analyses of explanations of ‘seasons’ in the five textbooks, we developed a
composite description of their key nodes, which we again state as simply as possible,
in our own words, and with bold indicating nodes in constellation diagrams (start-
ing with Figure 3.8).

A. The Earth is tilted on its axis at an angle of 23.4 degrees.

B. The Earth is divided into northern and southern hemispheres.

FIGURE 3.8 Schematic constellation of the seasons explanations across textbooks

64 Karl Maton and Y. J. Doran

C. The Earth orbits the Sun.

D. The Earth’s tilt (node A), division into hemispheres (node B), and orbit of the
Sun (node C) together mean that the Earth’s northern and southern hemi-
spheres point towards or away from the Sun at different times of the year.
E. The Earth receives solar energy from the Sun’s rays.
F. That hemispheres point towards or away from the Sun (node D) and that the
Earth receives solar energy from the Sun (node E) mean that sunlight hits
each hemisphere at different angles through the year: when a hemisphere
is pointing towards the Sun the angle of its rays are more direct; and when a
hemisphere is pointing away from the Sun the angle of its rays are less direct.
G. Different angles of sunlight (node F) means that the concentration of light
changes in each hemisphere through the year: when a hemisphere is pointing
towards the Sun, the more direct sunlight it receives is concentrated in a smaller
area; and when a hemisphere is pointing away from the Sun, the less direct
sunlight it receives is concentrated in a larger area.
H. Differences in concentration of sunlight (node G) means that the amount of
heat radiation in each hemisphere varies through the year.
I. The Earth rotates.
J. That hemispheres point towards or away from the Sun (node D), the Earth
receives solar energy (node E) and the Earth rotates (node I) together mean that
different parts of the Earth experience different length of daylight at different
times of the year.
K. Variations in heat radiation (node H) and/or daylight length (node J) leads to
variations in temperature in each hemisphere through the year that are called
‘seasons’. This can also be more simply put as: that hemispheres point towards
or away from the Sun (node D) leads to variations in temperature in each
hemisphere through the year that are called ‘seasons’.

Links, clusters and form

Figure 3.8 represents the description as a schematic constellation. Like that for
tides, the constellation involves nodes that generate both independent (squares) and
dependent (circles) relations with other nodes. Also like tides, it exhibits a base
cluster and two supplementary clusters. However, there are significant differences in
how the seasons constellation relates together its constitutive ideas.
The schematic constellation begins with three independent nodes describ-
ing Earth’s tilt, hemispheres and orbit. These nodes are brought together to
describe their implications for hemispheres pointing towards or away from the
Sun, generating dependent links with all three. From here one can proceed directly
to implications for variations in temperature in hemispheres through the year,
reaching seasons through a dependent link. As shown in Figure 3.9a, this repre-
sents a base cluster. There are then two supplementary clusters that develop this
explanation. At this point the constellation becomes more complex than that of
tides, in two ways.
 Constellating science 65

FIGURE 3.9 Schematic constellation for seasons – base and supplementary clusters

First, the supplementary clusters offer parallel routes through the logic of the
explanation.What we term the ‘sunlight angle cluster’ recruits the nodes highlighted
in Figure 3.9b. This brings together the node describing hemispheres pointing
towards or away from the Sun with the independent node that the Earth receives
solar energy from the Sun’s rays to describe their implications for variations in
the angle of sunlight received by hemispheres. It then elaborates implications of
these variations in sunlight angle for the concentration of sunlight, the implica-
tions of variations in concentration for heat radiation, and in turn the implication
of variations in heat radiation for temperatures, which reaches seasons. A second
supplementary, which we term the ‘daylight length cluster’, is shown in Figure 3.9c.
This too brings together the nodes on pointing towards or away from the Sun
and solar energy but this time adds the independent node that the Earth rotates
to describe the implications of these three together for variations in the length of
daylight in hemispheres through the year. It then draws out the implications of
these variations in daylight length for temperatures, reaching seasons. Crucially,
these two clusters are not successive but parallel: each cluster reaches ‘seasons’ sepa-
rately. The logic allows seasons to be explained through either cluster or both clus-
ters together.
A second difference with tides is that the supplementary clusters do not elaborate
the base cluster in the same way. In tides, each supplementary cluster adds to the logi-
cally preceding cluster, proceeding from ‘tides’ to ‘daily variation in tides’ to ‘spring
and neap tides’. In contrast, here the supplementary clusters ‘unpack’ the implica-
tions of hemispheres pointing towards or away from the sun for creating seasons.
They are akin to focusing in on the long dashed arrow in Figure 3.9a and provid-
ing more detailed explanations of that relation. Put another way, the supplementary
clusters for tides add relations to new destinations along one route (tides – daily
66 Karl Maton and Y. J. Doran

variation – spring/neap) while the supplementary clusters for seasons clarify and
deepen an existing relation by adding new routes to the same destination (seasons).
These routes show the complexity latent within the link between ‘towards or away’
and ‘seasons’.
In short, the seasons constellation is less linear and successive than that of tides,
offering more options for navigating its logic to create an explanation.We now turn
to explore how this difference might be reflected in how explanations of the seasons
are taught in a classroom.

A pedagogic constellation for seasons

The example we discuss comprises a lesson on ‘seasons’ in the same Year 7 second-
ary school classroom, immediately preceding the lesson on tides analyzed above. The
teacher begins the topic by showing how ripe the topic is for misunderstandings.12 She
plays a ‘vox pop’ video in which adults are asked for causes of seasons and suggest such
mistaken beliefs as the equator, changing distance from the Sun, and Earth’s elliptical
orbit. The teacher states this shows that ‘You have to think carefully about what we’re
doing’. Unlike ‘tides’, she thus begins by highlighting the complexity of the constella-
tion. This complexity, we argue, is reflected in the number of times she takes the class
on different routes to reach an explanation. Specifically, she takes the class through: (a)
a ‘daylight length’ route; (b) a ‘sunlight angle’ route; (c) a ‘base cluster’ route; and (d) a
composite route that includes ideas from all three clusters. These routes are discrete:
they are separated by class discussions of related ideas (such as the names of longest
and shortest days) or by student questions (such as whether leap years affect the Earth’s
orbit) that are not woven into explaining the seasons.We shall go through each expla-
nation briefly, using the schematic constellation as a basis for comparing these routes.

(a) A ‘daylight length’ route

The teacher begins from where the previous lesson ended, with changes in the
length of days through the year:

In the summer, days are longer, and in the winter, days are shorter. When the
days are longer, that means there is more time for the Sun to heat the Earth. So
that means that the temperature is warmer. Okay?

In terms of the schematic constellation, the teacher begins with the effects out-
lined as ‘daylight length’ and ‘seasons’. She then refers to an experiment students
conducted in a recent lesson that involved heating a wooden block with a lamp to
model the effect of the Sun’s rays on the Earth:

So remember that experiment we did when we put the thermometer in the

block? As the time at which we kept that light on the block increased the tem-
perature. It’s the same thing that happens with days. When the days become
shorter, we have overall lower temperatures. Because we have long days in
 Constellating science 67

summer and short days in winter, we get higher temperatures, we get lower
temperatures.

This brings in solar energy from the Sun.The teacher asks: ‘why do we have some
long days and some short days?’. A student answers ‘because the Earth is tilted’, and
the teacher responds:

Good. So it’s got to do with the tilt of the Earth and depending on which part is
tilted closer to the Sun will determine which will have longer and shorter days.

This is to say that depending on which of Earth’s hemispheres is brought by

Earth’s tilt to be pointing towards or away from the Sun will determine the
daylight length in that hemisphere.
Figure 3.10a, like all constellation diagrams in this analysis, shows the ideas and
relations among ideas of each explanation rather than the exact sequencing of nodes
in the teacher’s discourse. As the Figure illustrates, the teacher offers here a simple
‘daylight length’ route through the explanation. Taken as a whole, it brings together
Earth’s tilt and hemispheres to reach ‘towards or away’, which, bringing in the

FIGURE 3.10 Pedagogic constellations for seasons – ‘daylight length’ and ‘sunlight angle’
routes
68 Karl Maton and Y. J. Doran

aforementioned solar energy, reaches daylight length and thence seasons. At

this stage, the teacher did not include Earth’s orbit and Earth’s rotation.

(b) A ‘sunlight angle’ route

In the next version of the explanation, the teacher shows an animation entitled
‘What causes Earth’s seasons’, muting the sound and adding her own commentary
(to which we have added the nodes discussed in brackets):13

… what we’re looking at here is obviously the Earth spinning on its axis [Earth
rotates] and you can see that that axis is about 23.5 degrees from what could
be the “theoretical midline” of the Earth [Earth’s tilt]. We know that that axis
holds itself. … Now, when the Sun’s rays [solar energy] hit the Earth, the
Earth has that also other “theoretical midline”, the equator, that breaks it into
half: northern hemisphere and southern hemisphere [hemispheres]. Okay?
We’re in the southern hemisphere and at some times the sunlight will strike, in
this case, the southern hemisphere at that 90 degree angle [sunlight angle]. All
right? And here, this is what we can sometimes call an ‘oblique’ … or ‘glancing’
[sunlight angle].

The first node, Earth rotates, is mentioned to direct students to watch the anima-
tion and not woven into the explanation. The rest – Earth’s tilt, solar energy,
hemispheres and sunlight angle – are introduced first as factors to be related. She
then brings them together:

Now, when this moves around the Sun [Earth’s orbit], and holds its axis
[Earth’s tilt], because it doesn’t … the axis won’t change around like that,
when this moves to the other side of the Sun, we will see the other side of the
Earth [towards or away]. … now we’re on the back side of the Sun.We’ve got
spring and autumn [seasons]. Now we’re on the opposite side. Now the north-
ern hemisphere has summer and the southern hemisphere has winter [sea-
sons]. Now, we come out to the “fourth side” of the Sun, we get autumn and
spring split around [seasons].

The teacher then sums up this explanation with a PowerPoint slide showing the
Earth (with equator and axis shown), the Sun and its rays hitting the Earth: ‘So
we know the Sun is hitting the Earth [solar energy]. We know that some parts
of the Earth [hemispheres] will be getting the full force and some will getting
those glancing rays [sunlight angle]’ and then emphasizes that ‘The thing that will
change is our position in relationship to the Sun’ [towards or away].
As illustrated by Figure 3.10b, taken as a whole this offers a ‘sunlight angle’ route
through the explanation. In combination with the animation, the teacher shows
that Earth’s tilt, hemispheres and orbit mean that different hemispheres point
towards or away from the Sun which, when struck by solar energy, means there
are differences in sunlight angle that create seasons.
 Constellating science 69

(c) A ‘base cluster’ route

The third explanation comprises three activities: discussing a video animation, stu-
dents writing in their workbooks, and students drawing a diagram. First, the teacher
shows a video animation of the Earth and the Sun and highlights that the Earth
is ‘moving around the Sun’ (Earth’s orbit), that Earth’s tilt is not changing, and
that these together mean the hemispheres change their relative position to the
Sun (towards or away). The teacher sums this up as: ‘This is why we end up with
opposite seasons. Because of that tilt in our axis puts us in different positions in
relationship to the Sun’. As Figure 3.11c shows, this echoes the base cluster of the
schematic constellation.
Second, the teacher asks students ‘to write one or two sentences that explains
how the Sun and the Earth create seasons’. She plays the animation again while
students write for several minutes, before soliciting their answers. The first student
answer is that ‘The tilt of the Earth on a 23.5 degree angle and the orbit around
the Sun makes the Earth have seasons’, which is to state that Earth’s tilt and
Earth’s orbit lead to seasons. This does not include hemispheres or pointing
towards or away from the Sun and the teacher responds: ‘It doesn’t quite explain,
though, why we get the different seasons’. The next two answers provide what she
is seeking:

STUDENT The Earth is always on a 23 ½ degree tilt. This tilt remains the same as
we orbit the Sun but the area facing the Sun is different [towards or away].
This causes the seasons.
TEACHER Fantastic. I like that.Very good. Who else has one?
STUDENT The seasons are created by the Earth’s 23.5 degree tilt.When the north-
ern hemisphere is tilted towards the Sun, it is summer. As the Earth orbits the
Sun the tilt stays the same but the side [hemispheres] that’s tilted towards the
Sun changes [towards or away], making it winter in the northern hemisphere
because it’s furthest away.
TEACHER Fantastic, I like that. That’s a good one. All right. Hopefully yours
says something similar to that.

The first answer adds the node of pointing towards or away from the Sun and
the second answer adds both that node and hemispheres. Once these responses
complete the constellation that she had set out, the teacher moves on.
Finally, a simpler version of the constellation is repeated again while the stu-
dents draw a diagram. The teacher shows a simulation in which she can move the
Earth to different positions around the Sun. Students are told to draw a diagram
of the Sun and the Earth for one season. ‘The key point here’, she emphasizes, ‘is
to make this an accurate diagram’. Through questions, the teacher solicits from
students ‘the things we cannot be sloppy on in this diagram’: 23.5 degrees tilt and
axis, lines for the equator to show hemispheres, and an elliptical orbit around
the Sun. Thus, the independent nodes of the ‘base cluster’ are again emphasized as
key factors.
70 Karl Maton and Y. J. Doran

FIGURE 3.11 Pedagogic constellations for seasons – ‘base cluster’ and combined routes

(d) A composite route

The final explanation comprises a summary spoken by the teacher:

We know that because the Earth’s tilted, that different parts of the Earth will
face the Sun at different parts of the year. That difference then in the sunlight
striking the Earth will lead to different seasonal variations in temperature,
depending on how much the Earth has heated up during the day. The length
of our day will also then depend on how far north or south you are from the
equator. And that explains why temperatures over the year will change at dif-
ferent locations.

As Figure 3.11d shows this states that Earth’s tilt means different hemispheres –
routinely called ‘parts of the Earth’ through the lesson – will point towards or
away from the Sun, which means that solar energy will strike the Earth in
different ways (sunlight angle), which leads to variations in temperature (sea-
sons), but that this also depends on how much the Earth has heated during the
day due to variations in daylight length. Taken together, this all explains the
seasons. So, this explanation includes the ‘base cluster’ (minus Earth’s orbit),
 Constellating science 71

her ‘sunlight angle’ route and her ‘daylight length’ route. In short, it provides
a summarizing composite of the three routes she has taken the class on when
creating explanations.

Aggregating explanations
Our concern has been not to evaluate this teaching but to explore what it can tell
us about relations among ideas. Here, explaining the seasons seems to offer a variety
of potential pedagogic constellations. During an hour-long lesson the teacher takes
the class on four different routes, focused on ‘daylight length’, ‘sunlight angle’, the
‘base cluster’ (on which she checks students’ understanding), and a composite of
all three. This involves what we termed aggregating: separate clusters are assembled,
each on its own, before being combined. Moreover, this aggregating involves node
options: some nodes from the schematic constellation were not included: sunlight
concentration, heat radiation and Earth rotates. This is, though, not a gap in
teaching.The schematic constellation is not a list of essential ideas but rather a com-
posite of all nodes found in the textbooks we analyzed; not all textbooks include
all those nodes. There are likely to be sound pedagogic reasons for employing a
certain level of complexity and not adding further nodes that may overcomplicate
the explanation.

Conclusion
Science is complex – seemingly simple empirical phenomena may require complex
explanations. Our argument has been simple: relations among ideas are one aspect
of this complexity and constellation analysis offers a way of seeing these relations
and analyzing the roles they play in building knowledge. Our analysis of the logic of
explanations of tides and seasons that are offered by textbooks aimed at Year 7 sci-
ence in NSW, Australia, made the simple point that relations among ideas differ even
between otherwise seemingly similar sets of ideas. Our analysis of how those expla-
nations were taught in the same classroom highlighted the simple point that differ-
ences in how sets of ideas are taught may be related to differences in relations among
their ideas. By analyzing their pedagogic constellations in relation to their schematic
constellations, we were not arguing that the logic of explanations is an ideal against
which teaching should be measured. As we emphasized, there are many sound rea-
sons for teaching knowledge in different sequences to, or in different degrees of
detail than shown by schematic constellations. (Indeed, each textbook unfolds its
explanation differently to the composite schematic constellation). Rather, our aim
was to show that relations among ideas in an explanation may affect how those
explanations are taught – in short, the logic may help shape the pedagogic.
To demonstrate this point, we focused on examining simple distinctions
between two kinds of links, clusters, and forms of constellation-building. Using
these distinctions we showed that the schematic constellation for tides involves a
72 Karl Maton and Y. J. Doran

base cluster that is elaborated successively by supplementary clusters, while that

for seasons involves parallel supplementary clusters that ‘unpack’ part of the base
cluster and offer different ways of reaching ‘seasons’. Put simply, the elaboration of
clusters for tides was akin to ‘A then B then C’, where for seasons it was akin to ‘A
and/or B and/or C’.
Using the same distinctions we then analyzed how each explanation was taught
by the same teacher during the course of a single unit of study. The pedagogic
explanation of tides echoed the logic of its schematic constellation: linear and suc-
cessive clusters of ideas that build on preceding ideas. The pedagogic explanation of
seasons also echoed the logic of its schematic constellation: the teacher set out four
different selections and arrangements of ideas that offered differing versions. So the
two explanations differed in how they were taught by the same teacher. From this
analysis we suggested that ‘seasons’ appears more amenable to variation of which
ideas are selected and how they are brought together. In short, these assembling and
aggregating forms of building constellations differed in ways that echoed relations
among ideas within the logic of the explanations.
Beyond our modest aims, these insights suggest that constellation analysis may
shed light on relations between research, curriculum and pedagogy. By showing
how knowledge changes as it is transformed into curricula or taught in classrooms,
constellation analysis could open the ‘black box’ of recontextualization. Following
Bernstein (1990), LCT distinguishes between the logics underlying production fields
that create ‘new’ knowledge, recontextualizing fields that create curriculum, and repro-
duction fields or sites of teaching and learning (Maton 2014: 43–64). Movements
of knowledge between fields are held to involve ‘recontextualization’ – selection,
arrangement and enactment of ideas – that restructures that knowledge. As yet lit-
tle light has been shed on differences in recontextualization. Constellation analysis
offers a way of analyzing how a set of ideas is structured one way in research, differ-
ently in a curriculum, and differently again when taught and learned in classroom
practice. Comparative analysis could show that some constellations are more ame-
nable to restructuring than others, providing insight into the nature and value of
different recontextualizations. Analysis could also reveal when and how it is valuable
for the pedagogic unfolding of a constellation to differ from the schematic constel-
lation of its logic.
Constellation analysis may also be practically useful for educators. It offers a way
of mapping out lesson plans and teaching designs for the content to be taught and
learned. Constellation diagrams could help educators make the knowledge being
taught more explicit to students by highlighting key ideas and relations among
them. They could be used to map progress through the sequencing of content and
make visible how different issues come together. They could also serve as a way for
students to demonstrate their understanding. Comparing students’ diagrams of, for
example, an explanation of tides to the teacher’s diagram could help make clear
what has been learned and what nodes, links and clusters need revisiting. In this way,
constellation analysis offers a means of connecting studies of knowledge and studies
of student conceptions, the dominant preoccupation in science education research.
 Constellating science 73

Another area for future development is exploring different kinds of relations

among ideas. Links may take many forms. One can draw here on other LCT con-
cepts to examine their attributes. For example, enacting the concept of epistemologi-
cal condensation (specifically its ‘translation device’ for clausing, see Maton and Doran
2017) would show how different links – classifying, composing, causing, correlating,
etc. – add meaning to constellations at different rates. Similarly, analysis of charging
would show how different valuations of nodes and clusters suffuse constellations.
The possibilities offered by such work are yet to be explored. This chapter was but
a first step: the future may be written in the stars.

Acknowledgements
We wish to sincerely thank Claire Flanagan for creating Figure 3.1, Greg Rusznyak
for creating Figure 3.2 (inspired by http://www.datasketch.es/may/code/nadieh/),
and David Fergusson and Kylie Wynne for insights into science textbooks.

Notes
1 See MacDonald (1998) and data available at www.stellarium.org.
2 Rusznyak (2020) goes further by using the constellation of Orion to additionally illus-
trate the concepts of semantic density and condensation.
3 Though these concepts were first introduced using concepts from the Semantics dimen-
sion of LCT (Maton 2014: 148–70), they do not belong to any dimension. Constellation
and cosmological analyses can be conducted with concepts from any dimension.
4 Exceptions include: Lambrinos (2020), which distinguishes relations among nodes in
dance teaching by how much complexity they add; and Doran (2020), which explores
relations in terms of rhetorical moves that show how values are constellated.
5 For a complementary view of relations among ideas from systemic functional linguistics,
see Doran and Martin, Chapter 5 of this volume.
6 Aggregating may involve separate clusters or separate nodes. In this chapter, teaching
about Earth’s seasons involves several clusters of ideas that are built separately before
being brought together. In other fields aggregating may involve a large number of indi-
vidual nodes. For example, in History lessons the discussion of a period may involve
aggregating a large number of facts (Maton 2015).
7 Both are also epistemological constellations, in which stances are formal definitions (such as
concepts and empirical descriptions). For discussion of axiological constellations, in which
stances are affective, aesthetic, ethical, moral or political, see Maton (2014: 148–70),
Martin et al. (2010), Doran (2020), and Tilakaratna and Szenes (2020).
8 We used the same method on the same textbooks for both tides and seasons. We expand
on the process for constructing ‘schematic constellations’ in a future publication.
9 We draw on a major study funded by the Australian Research Council (DP130100481).
10 The video (https://www.youtube.com/watch?v=QcbN9SVkqYU) is by the US public
service broadcaster KQED; the excerpt is narrated by Ben Burress, staff astronomer at the
Chabot Space and Science Center in Oakland, California.
11 As stated earlier, our pedagogic constellation diagrams show nodes and links in comparison
to the schematic constellations because of our specific aims here. Constellation diagrams of
the teaching as it unfolds would differ.
74 Karl Maton and Y. J. Doran

12 The two preceding lessons included mentions of Earth’s tilt, orbit, rotation, and how
sunlight heats the Earth. However, these ideas were nodes in other constellations: explain-
ing ‘day and night’ and explaining different temperatures on Earth. In neither case was
the content constellated into an explanation of seasons. This highlights that no content
is locked into a specific constellation. Mentions of, say, ‘tilt’ or ‘orbit’ are not necessarily
discussions of ‘seasons’.
13 The video is available at: https://www.teachertube.com/videos/what-causes-seasons-
on-earth-657. Maton and Howard (Chapter 4, this volume) examine the teacher’s use
of this video in detail, as an example of using multimedia in science teaching, explaining
why she replaced its audio with her own commentary.

References
Adendorff, H. and Blackie, M. A. L. (2020) ‘Decolonizing the science science: When good
intentions are not enough’, in C. Winberg, S. McKenna and K. Wilmot (eds) Building
Knowledge in Higher Education, London: Routledge, 237–54.
Arena, P., Warnant, P., Burrows, K., Everygreen, M. J. and Lofts, G. (2009) Core Science: Stage 4
Complete Course, Milton: John Wiley and Sons.
Bernstein, B. (1990) Class, Codes and Control, Volume IV: The Structuring of Pedagogic Discourse,
London: Routledge.
Bishop, S., Bass, G., Champion, N., Gregory, E., McKenna, E. and Walker, K. (2011) Nelson
iScience for the Australian Curriculum,Year 7, South Melbourne: Nelson Cengage Learning.
diSessa, A. (1993) ‘Toward an epistemology of physics’, Cognition and Instruction, 10(2/3):
105–225.
Doran, Y. J. (2020) ‘Cultivating values: Knower-building in the humanities’, Estudios de
Lingüística Aplicada, 37(70): 169–98.
Ellery, K. (2017) ‘Conceptualising knowledge for access in the sciences’, Higher Education,
74(5): 915–31
Georgiou, H., Maton, K. and Sharma, M. (2014) ‘Recovering knowledge for science educa-
tion research: Exploring the “Icarus effect” in student work’, Canadian Journal of Science,
Mathematics, and Technology Education, 14(3): 252–68.
Glenn, E. (2016) ‘From clashing to matching: Examining the legitimation codes that underpin shift-
ing views about climate change’, unpublished PhD thesis, University of Technology Sydney,
Australia.
Lambrinos, E. (2020) ‘Building ballerinas: Developing dance and dancers in ballet’, unpublished
PhD thesis, University of Sydney, Australia.
Land, R., Cousin, G. and Meyer, J. H. F. (2005) ‘Threshold concepts and troublesome knowl-
edge (3): Implications for course design and evaluation’, in C. Rust (Ed.) Improving Student
Learning Diversity and Inclusivity, Oxford: OCSLD, 53–64.
MacDonald, J. (1998) The Arctic Sky: Inuit Astronomy, Stare Lore, and Legend, Ontario: Royal
Ontario Museum.
Martin, J. R., Maton, K. and Matruglio, E. (2010) ‘Historical cosmologies: Epistemology
and axiology in Australian secondary school history discourse’, Revista Signos, 43(74):
4330–463.
Maton, K. (2014) Knowledge and Knowers: Towards a Realist Sociology of Education, London:
Routledge.
Maton, K. (2015) ‘Legitimation Code Theory: Past, present, future’, Opening Address, LCT1 –
First International Legitimation Code Theory Conference, Cape Town, South Africa, June.
 Constellating science 75

Maton, K. and Doran,Y. J. (2017) ‘Condensation: A translation device for revealing complex-
ity of knowledge practices in discourse, part 2 – clausing and sequencing’, Onomázein
Special Issue, March: 77–110.
Meyer, J. and Land, R. (2003) ‘Threshold concepts and troublesome knowledge’, ETL Project
Occasional Report 4, http://www.etl.tla.ed.ac.uk/docs/ETLreport4.pdf
Rickard, G. (2011) Pearson Science 7: Student Book, Frenchs Forest: Pearson Education.
Rose, D., McInnes, D. and Korner, H. (1992) Scientific Literacy, Sydney: Metropolitan East
Disadvantaged Schools Program.
Rusznyak, L. (2020) ‘Supporting the academic success of students through making knowl-
edge-building visible’, in C.Winberg, S. McKenna and K.Wilmot (eds) Building Knowledge
in Higher Education, London: Routledge, 90–104.
Salwén, H. (2019) ‘Threshold concepts, obstacles or scientific dead ends?’, Teaching in Higher
Education, DOI: 10.1080/13562517.2019.1632828
Stannard, P. and Williamson, K. (1995) Science World 7, South Yarra: Macmillan.
Tilakaratna, N. and Szenes, E. (2020) ‘(Un)critical reflection: Uncovering disciplinary values
in social work and business reflective writing assignments’, in C. Winberg, S. McKenna
and K. Wilmot (eds) Building Knowledge in Higher Education, London: Routledge, 105–25
Unsworth, L. (1997) ‘“Sound” explanations in school science: A functional linguistic perspec-
tive on effective apprenticing texts’, Linguistics and Education, 9(2): 199–226.
Zhang, J., Alford, D., McGowan, D. and Tilley, C. (2013) Oxford Insight Science: Australian
Curriculum for NSW, Stage 4,Year 7, South Melbourne: Oxford University Press.
4
ANIMATING SCIENCE
Activating the affordances of multimedia
in teaching

Karl Maton and Sarah K. Howard

Integration is more than selection; affordances are more than interactions.

Introduction
Multimedia objects that combine diverse visual and audio elements – text, pictures,
moving images, speech, sounds, etc. – are a common feature in classrooms (Li et al.
2019). Increased access to the Internet and rapid growth in online ‘educational’
materials has made such multimedia as animations evermore available to teachers
as classroom resources (Berney and Bétrancourt 2016). These are particularly of
interest to science education as ways of displaying complex explanations (Ploetzner
and Lowe 2012). Moreover, studies claim a ‘multimedia effect’ whereby students
learn better through words and pictures together (Mayer 2003). However, not all
multimedia are created equal – some are more suited to classroom practice than
others. Moreover, even if designed for a particular curriculum, a multimedia object
is unlikely to match the needs of a specific task in a specific lesson in a specific
classroom. Thus, questions of how teachers integrate such ready-made objects into
classroom practice are increasingly significant (e.g. Jenkinson 2018).Yet, these ques-
tions are not the concern of most education research into multimedia. Instead, the
principal concern is exploring cognitive processes of student learning and generating
principles for designing multimedia that support those processes (e.g. Mayer 2014b,
Mutlu-Bayraktar et al. 2019). Questions of how pedagogic practices by teachers support
integrating multimedia into classroom tasks are marginal. This chapter contributes to
foregrounding and addressing these questions in the context of science teaching.
We begin by arguing that to understand integration requires bringing into the
picture both teaching and the knowledge practices being taught. To begin meeting
this need we draw on the notion of ‘affordances’ (e.g. Bower 2017), which high-
lights that technology such as multimedia objects differ in their abilities to meet
 Animating science 77

the demands of different classroom tasks. However, we argue that current uses of
‘affordances’ are too limited. Studies overwhelmingly examine affordances for inter-
actions, especially among students, and ignore affordances for teaching and learn-
ing specific knowledge practices. To address this ‘knowledge-blindness’ (Howard
and Maton 2011, 2013), we introduce concepts from the Autonomy dimension
of Legitimation Code Theory. These concepts reveal one aspect of the knowledge
practices expressed by resources and their epistemic affordances for teaching and learn-
ing. Specifically, positional autonomy conceptualizes relations between the content
required by a task and that expressed by a resource, relational autonomy conceptual-
izes relations between the purposes for which that content is needed by a task and
those for which it is used in a resource, and target places the specificities of each
classroom task at the centre of analyzing these two relations. Enacting these con-
cepts, we explore both the affordances of multimedia objects (and other resources)
for building knowledge in tasks and how teachers select and activate those epistemic
affordances in classroom practice.
Specifically, we conduct two in-depth analyses of integrating multimedia in sec-
ondary school science classrooms. We focus on the use of animations, a central
concern of multimedia research (Li et al. 2019). In both examples, a science teacher
selects an animation whose affordances are well suited to their task, both in terms
of interactions enabled and knowledge practices expressed. However, while one
teacher activates its epistemic affordances, the other fails to do so. These analy-
ses illustrate that integration does not end with selection and that pedagogic work
by teachers is required to activate the epistemic affordances of resources. In short,
they show how understanding success in teaching science with multimedia requires
seeing not only the multimedia but also the teaching practice and the science
knowledge being taught.

Integrating multimedia into teaching

The cognitive–learning–design axis
Multimedia combine words (spoken or text) and pictures (such as still or moving
images). It is widely acknowledged that multimedia resources are becoming more
accessible for education (Berney and Bétrancourt 2016). How, then, can teachers
best integrate such multimedia objects as animations into classroom practice to sup-
port teaching science? This is our focus in this chapter. However, this is not the focus
of the vast majority of research into multimedia (see Li et al. 2019). Most studies
focus on learning rather than teaching, designing rather than integrating multime-
dia, and computer-based contexts rather than classrooms (e.g. Mayer 2014b). This is
not to dismiss this body of work but rather to highlight that dominant approaches
to the educational potential of multimedia are not concerned with integration.
Instead the focus of most research lies on a cognitive–learning–design axis.
Commonly used models, such as ‘cognitive theory of multimedia learning’ (Mayer
2014a) and ‘cognitive load theory’ (Sweller et al. 2019), examine such issues as how
78 Karl Maton and Sarah K. Howard

the brain processes information, audio and visual channels, and kinds of memory.
Implications are drawn for how to combine visual and auditory information in
order to, for example, manage ‘cognitive load’ (e.g. Mayer and Moreno, 2003). This
provides insights into design principles, such as the ‘spatial congruity principle’ that
text and visual content should be proximate (Mayer and Fiorella 2014). However,
this cognitive–learning–design focus obscures integration in two principal ways.
First, teachers and teaching are treated as if limited to designing ‘environments’
or selecting resources that meet design principles. This leaves aside almost all real-
world classroom pedagogy. Second, the forms taken by the knowledge practices to
be taught and learned are absent or reduced to the status of ‘topic’, ‘subject matter’
or ‘information’.This ignores that learning involves learning something and that the
forms taken by the knowledge practices expressed by multimedia may affect their
integration into teaching and learning of that something. Thus, while valuable for
developing generic principles for designing multimedia, these frameworks are not
well suited to questions of integrating multimedia into teaching specific knowledge.
We thus need to bring teaching and knowledge into the picture.

The affordances of ‘affordances’

A fruitful starting point is the notion of ‘affordances’ which is typically used in edu-
cation research to explore the capacities of technologies for enabling actions (Bower
2008, Hammond 2010, Antonenko et al. 2017). For classroom practices, this is to say
that a resource may be more suited to some tasks than others. Affordance ‘frame-
works’ list ‘abilities’ of technology; Bower (2017), for example, includes read-ability,
write-ability, playback-ability, share-ability, among others. The aim is for ‘designers’
of ‘learning environments’ to select resources that offer the abilities required by
tasks. The notion of ‘affordance’ thus offers the potential to foreground classroom
tasks and so integration.
However, this affordance of ‘affordance’ is often not activated by research, which
is animated by the cognitive–learning–design axis. For example, the notion is said to
help show ‘how tools interplay with cognition and hence how to best design educa-
tional systems that meet the learning requirements of tasks’ (Bower 2008: 4; empha-
ses added). Studies also share the blind spots of teaching and knowledge practices.
Consider the proclaimed affordances of ‘affordance’. Hammond (2010: 211), for
example, states: ‘This is its major value as a concept: it is not the tool, it is not the
person, it is the interaction of tool and person’. Focusing on interactions that a ‘tool’
enables, both between ‘tool and person’ and among people, is indeed the principle
way that ‘affordances’ is used. However, the people studied are only learners and
not teachers, the ‘interaction’ is only learning and not teaching, and the role played
by what is to be taught and learned through the interaction is absent. This focus
on student interactions is underscored by the huge field of ‘Computer-Supported
Collaborative Learning’, whose gravitational weight pulls attention away from real-
world classrooms and most forms of pedagogy. For example, affordances are often
clustered together as supporting, for example, ‘static/instructive’ or ‘collaborative/
 Animating science 79

productive’ interactions (Bower 2017), where the latter focus on student interac-
tions garners most research.
In short, existing research focuses on what we shall term ‘interactive affordances’.1
This underestimates the affordances of the notion as an animating metaphor for
research. To help examine integration, we shall focus on epistemic affordances – how
knowledge practices expressed by a resource offer opportunities for teaching and
learning – and analyze how teachers can activate those epistemic affordances in
teaching science.2 To do so, we draw on Legitimation Code Theory.

Legitimation Code Theory: Autonomy

Legitimation Code Theory (LCT) is a framework for researching and shaping all
kinds of practice (Maton 2014) but has proven particularly valuable for exploring
knowledge practices (e.g. Howard and Maton 2011). LCT is also widely enacted
to analyze and inform teaching.3 The framework comprises several dimensions that
reveal different aspects to knowledge practices. For example, concepts from the
Semantics dimension (Maton 2020) can, among other things, reveal how the con-
text-dependence of meanings expressed by a resource can support a task, such as
affording the desired degree of concreteness.4 Concepts from the Specialization
dimension (Maton 2014) can reveal, among other things, how the representational
fidelity of images afford possibilities for teaching and learning specific ideas. Indeed,
LCT affords far more possibilities for analysis than can be enumerated here.Thus, we
shall focus on the broader issue of the content and purpose of knowledge practices
expressed by resources. We draw on the Autonomy dimension, which is particularly
suited to exploring integration (Maton and Howard 2018, chapter 2 of this volume).

Autonomy
The dimension of Autonomy explores what makes practices distinctive. It begins
from the simple premise that any set of practices comprises constituents that are
related together in particular ways. Autonomy explores how practices establish dif-
ferent degrees of insulation around their constituents and around the ways they are
related together. Here we can describe constituents as the content of practices and how
they are related together as the purpose to which they are put.We can then distinguish:

• positional autonomy (PA) between content within a context or category and

content from other contexts or categories; and
• relational autonomy (RA) between the purposes to which content is put within
a context or category and purposes from other contexts or categories.

Each may be stronger or weaker along a continuum of strengths. The stronger the
positional autonomy, the more insulated are contents from those of other contexts
or categories; and the stronger the relational autonomy, the more insulated are pur-
poses from those of other contexts or categories.
80 Karl Maton and Sarah K. Howard

TABLE 4.1 Generic translation device, first two levels

(adapted from Maton and Howard 2018: 10)

PA/RA 1st level PA/RA 2nd level

stronger target ++ core
+ ancillary
weaker non-target – associated
–– unassociated

The ‘contexts or categories’ in these definitions are held open to embrace the
manifold diversity of practices. To enact the concepts, one needs to develop trans-
lation devices that show how they are realized within specific problem-situations
(Maton and Chen 2016). As shown in Table 4.1, Autonomy offers a generic translation
device as a first step. This device is activated by asking: what contents and purposes
are considered constitutive of this context or category, here, in this space and time,
for these actors? This gives a ‘target’ against which to compare practices. If their con-
tents are from within this target, they embody stronger positional autonomy, and if
from elsewhere, they embody weaker positional autonomy; and the same holds for
purposes and strengths of relational autonomy. A more fine-grained analysis may
then ask: what of the target is considered core and what ancillary, and what of other
contents and purposes are considered associated with or unassociated from the target?
As Table 4.1 shows, this generates four strengths for positional autonomy and for
relational autonomy.
This device is, though, still ‘generic’.To translate to specific data we must decide
whose target to examine and at what level. First, a target is always someone’s con-
ception of what makes a context or category distinctive; whose target is chosen
to ground analysis depends on the problem-situation. We shall analyze two exam-
ples of teaching, so we focus on each teacher’s targets (and use such possessives
throughout the chapter).5 Second, targets can describe ‘contexts or categories’ of
all kinds. For example, in our analysis of integrating mathematics into science
(chapter 2, this volume), each teacher’s target is a syllabus stage and their core
target is a unit of study. These broad categories capture relations between two
subject areas across whole lessons. Here our questions are more fine-grained: how
a specific multimedia object relates to specific tasks in specific lessons. Thus, we
analyze each teacher’s target as the lesson and their core target as the task in that
lesson. Using the translation device of Table 4.2, we can revisit the concepts to
state that here:

• strength of positional autonomy conceptualizes whether contents expressed by a

multimedia object match the teacher’s core target content for a task (PA++),
match their target content for the wider lesson (PA+), come from beyond the
lesson (PA–) or come from beyond the unit (PA– –); and
 Animating science 81

TABLE 4.2 Translation device for analysis in this chapter

PA/RA 1st level In this analysis: 2nd level In this analysis:

++ target lesson core task
+ ancillary rest of lesson
– non-target other contents associated other unit lessons or
or purposes related to task
–– unassociated knowledge from
beyond unit

• strength of relational autonomy conceptualizes whether the purposes of a mul-

timedia object match the teacher’s core target purposes for the task (RA++),
match their target purposes for the lesson (RA+), come from beyond the lesson
(RA–) or come from beyond the unit (RA– –).

More succinctly, positional autonomy will show relations between the content
expressed by a multimedia object and that required by a teacher’s task, and relational
autonomy will show relations between the purposes for which that content is used
in a multimedia object and the purposes needed by a teacher’s task.
We shall illustrate the usefulness of these concepts through analyses of two teach-
ers using video animations in science classrooms.6 Both examples are from Year
7 of secondary schooling in New South Wales, Australia. Both are from a unit of
study called ‘Earth and space sciences’. We develop specific translation devices for
each analysis, showing contents and purposes for each task. Given that multime-
dia objects are complex sets of diverse elements, we analyze their key elements
separately. We map these elements on an autonomy plane, as shown in Figure 4.1,
divided into the 16 modalities generated by combining the strengths of positional
autonomy and relational autonomy outlined above. By locating elements of each
animation in relation to the teacher’s targets, we show its epistemic affordances for
the task, the pedagogic work required by the teacher to integrate the animation
into the task, and whether the teacher succeeds.7 By ‘pedagogic work’ we mean
here changing the positions of elements on the plane.This can involve: integration or
connecting to content (strengthening PA, or moving up on Figure 4.1); disintegra-
tion or disconnecting from content (weakening PA or moving down); introjection or
turning to purpose (strengthening RA or moving right); and projection or turning
to another purpose (weakening RA or moving left). We can also describe addition,
where new contents or purposes are added, subtraction, where they are removed, and
substitution.Through such changes, teachers and students can more closely match the
knowledge practices of multimedia and tasks.
As we shall show, both teachers select animations whose interactive affordances
match the interactional needs of their tasks and whose epistemic affordances largely
82 Karl Maton and Sarah K. Howard

FIGURE 4.1 The autonomy plane (divided into 16 modalities)

match the knowledge practices they intend to teach. They are well suited but, as
is typical for classroom practice, not perfect. In our first example, the animation’s
epistemic affordances need pedagogic work to activate their potential to support
the task but the teacher does not provide that integrative pedagogy. In the second
example, the teacher’s pedagogic work makes the animation a closer match to the
knowledge needs of her task by substituting the animation’s audio elements with
her own commentary. Both animations have limited interactive affordances but
plentiful epistemic affordances, highlighting that these are not confined to interac-
tions and require activating. The analyses thereby bring knowledge and teaching to
the fore when examining integration.

Failing to activate epistemic affordances

The teacher’s targets: Shooting for the Moon
In our example of teaching that fails to activate epistemic affordances, the teacher
declares his target for the lesson at its outset: ‘today we are going to continue on
with the Moon’. For the task in question, the content of his core target concerns the
Moon’s rotation and its purpose is for students to answer two questions:

I’m going to pose a question for you … [reading PowerPoint slide] ‘Rotation
of the Moon’. So write that down: ‘Rotation of the Moon’ … That’s our topic.
And then I want you to answer: does the Moon rotate on its axis?. If so, how
fast does it rotate?
 Animating science 83

TABLE 4.3 Specific translation device for Moon animation task

target The Moon in ++ core Explanation in task, involving: Moon’s

this lessson orbit, length of orbit, rotation, length of
rotation, synchronous rotation, facing
side and ‘dark side’
+ ancillary Other science knowledge concerning
the Moon expressed in lesson (e.g.
phases and eclipses)
non- Other – associated Other science knowledge related to the
target knowledge Moon not included in lesson (e.g. tides)
or topics in other lessons of the unit
(e.g. seasons)
–– unassociated Knowledge from beyond the unit

As summarized in Table 4.3, the teacher’s core target concerns ‘synchronous rota-
tion’; in other words, that the time the Moon takes to rotate on its axis is almost
the same as the time it takes to orbit the Earth, so we always see the same side.
The rest of the lesson focuses on phases of the Moon and eclipses – these form
the teacher’s ancillary target. Other science knowledge related to his target, such as
the Moon’s role in tides, and from other topics in the same unit, such as seasons,
form his associated non-target. Knowledge from beyond the unit is unassociated
non-target.

The Moon animation

The task has two parts: discussion of a collage of static photographs and a video ani-
mation of 2:15 minutes length called ‘Synchronous rotation’.8 The animation com-
prises captions and two-dimensional images of entities, of which one (the Moon)
moves. There is no sound. The animation has two main parts: simulations of differ-
ent rotations of the Moon while orbiting the Earth; and a simulation of the chang-
ing shadow on the Moon created by sunlight during its orbit.
In the first main part a half-red and half-white Moon travels around an ellipse
four times under the same top caption: ‘Do we always see one side of the Moon?’.
As shown by Figure 4.2, the first orbit is sub-captioned ‘Moon without any rota-
tion…’ and the division between its red and white halves remains vertical during its
orbit. After the orbit, the sub-caption is extended by: ‘we see all sides in this case’.
The second orbit is sub-captioned ‘Moon with rotation’ and the Moon rotates
very quickly. As shown in Figure 4.3, after this orbit the words ‘How fast is it rotat-
ing?’ appear and a sub-caption states: ‘This isn’t actually what the Moon does…’.
The third orbit is also sub-captioned ‘Moon with rotation…’ and the Moon spins
so that its red side is always facing Earth. Central captions then appear: ‘How fast is
it rotating? It rotates on its axis once in the SAME time it takes to orbit us once!’.
84 Karl Maton and Sarah K. Howard

FIGURE 4.2 Moon animation: First orbit, without rotation (at 14 seconds)

FIGURE 4.3 Moon animation: Second orbit, overly fast rotation (at 53 seconds)

To ‘Moon with rotation…’ is added ‘we only see the red coloured side!!’.The Moon
then repeats that orbit.
In the second main part, illustrated by Figure 4.4, the Sun is represented on
the left. The top caption changes to ‘Now again with the Sun’s shadow’ and a
central caption states ‘THE ROTATION PERIOD OF THE MOON IS
EXACTLY THE SAME AS ITS PERIOD OF REVOLUTION!’. As it orbits,
the right-hand side of the Moon remains darker. Another central caption appears:
‘SYNCHRONOUS ROTATION! WILL EVENTUALLY HAPPEN TO THE
EARTH TOO! (IT’S [sic] ROTATION RATE WILL SLOW DOWN TO
EVENTUALLY MATCH ITS REVOLUTION RATE AROUND THE SUN)’.
Finally, a sub-caption states: ‘We always see the red coloured side! Position of the
Sun’s shadow does not move’.
 Animating science 85

FIGURE 4.4 Moon animation: Sun’s shadow simulation (at 1:53 minutes)

Affordances for the task

In terms of interactive affordances, the animation is relatively limited. It is freely
accessible online, shareable, watchable, viewable (though not listenable), and can be
played back. The animation is not easily manipulated, beyond pausing, repeating
or jumping ahead. It does not lend itself to being redrawn or support actions such
as drawing, writing or recording. These interactive affordances are associated by
Bower (2017) with ‘static/instructive’ rather than ‘collaborative/productive’ interac-
tions. Though limited in interactivity, the animation can support the teacher’s stated
purpose: students answering two questions about the topic.9

Epistemic affordances
To examine its epistemic affordances, Figure 4.5 presents the animation’s elements on
the autonomy plane, using the translation device of Table 4.3. (Positions within each
modality do not denote differences in strengths).To give an indication of prominence,
the shading offers a heuristic heat map of relative duration overall, from almost ever-
present (darkest) through prolonged and briefer to momentary (lightest).10
As Figure 4.5 summarizes, the animation visually represents the teacher’s core
target knowledge: key factors – such as lengths of orbit and rotation – are animated
at length to explain synchronous rotation (PA++, RA++). However, the anima-
tion also spends prolonged periods outside the teacher’s target. The first minute
comprises orbits with no rotation and with overly fast rotation, both of which are
inaccurate. In terms of content, they are not part of the explanation (PA–).Whether
they serve the purpose of learning about synchronous rotation depends on how
this inaccuracy is presented. For the first orbit with no rotation, its inaccuracy is not
made explicit.The caption asks whether we always see one side of the Moon, rather
than stating that as a fact, and there is no indication that the viewer should judge
whether, given that fact, the animated rotation is correct or not. Indeed, the caption
implies the opposite: that the viewer should use the animated rotation to answer
the question of whether we only see one side of the Moon. This orbit is thus not
86 Karl Maton and Sarah K. Howard

FIGURE 4.5 Autonomy analysis (with heuristic heat map) of Moon animation

serving the purpose of teaching synchronous rotation; it exhibits both non-target

content and non-target purpose (PA–, RA–). The second orbit, with overly fast
rotation, exhibits the same non-target content and purpose (PA–, RA–): only after
the orbit ends and a caption asks how fast it is rotating does the animation make its
inaccuracy explicit.
The third orbit represents the Moon’s rotation accurately. Captions then ask and
answer the teacher’s second question of ‘how fast’ the Moon rotates: core target
knowledge (PA++, RA++). During the final simulation showing the shadow on
the Moon created by the Sun, this knowledge is further underlined by captions.
The changing shadow itself is part of the teacher’s wider target for the lesson: phases
of the Moon (which the shadow demonstrates) is the next topic he discusses –
ancillary target knowledge (PA+, RA+). Further from his target lies the caption
about the Earth’s fate, which is not part of the unit (PA– –, RA– –).
In summary, the animation’s interactive affordances can support the teacher’s
‘instructive’ task of having students consider two questions. The animation also
exhibits significant epistemic affordances for the task. On both counts, then, it
appears well selected. However, integration is more than selection. Our autonomy
analysis shows that activation of its epistemic affordances requires pedagogic work
by the teacher, specifically through additions. First, he needs to establish a shared
understanding that the same side of the Moon is always seen from Earth and to
 Animating science 87

clarify that the question is whether, given that fact, the rotation being simulated
is accurate or not.11 These additions would support the introjection or turning to
purpose of the first two (inaccurate) orbits, strengthening their relational autonomy
(and so moving them right on the plane to become PA–, RA++). Second, he could
explicitly relate the simulation of the Sun’s shadow (PA+, RA+) to his next task
of discussing phases of the Moon. This addition would support both the introjec-
tion (turning to purpose, moving right) and the integration (connecting to content,
moving up) of these elements into his core target (PA++, RA++). Doing so would
expand his task to include setting up his subsequent discussion of Moon phases,
making the task less discrete and the lesson overall more cumulative.

The dark side of the Moon

The teacher does not undertake the pedagogic work needed to activate the anima-
tion’s epistemic affordances. He begins by asking students to answer his two ques-
tions by looking at a collage of pictures of the Moon from Earth:

So, there are some pictures taken from all around the world, different days,
alright? See if you can work out whether or not the Moon spins on its axis or
not, and if it does, how fast it spins. So talk to the person next to you.

A student asks whether ‘fast’ means ‘How many kilometres an hour?’, but is not
answered. Another student asks ‘How are we supposed to know how fast it rotates if
we’ve only got pictures?’, to which the teacher’s responds ‘Well, if you think it turns
once every night….’. Several minutes later, he tells the whole class:

I’ll give you a clue: the pictures, if you look very closely, and I’ve chosen these
because they all show it, you just have to look very closely, it will answer the first
part of the question for you. But you have to be careful and I’ll explain why.

When he asks students to report back, several respond by stating that the Moon does
spin. The teacher claims that a ‘dot’ appears on the Moon in every picture (though
cannot point to it on several) and says this ‘dot’ shows that it spins. Thus what could
establish that we always see the same side of the Moon – that pictures ‘taken from
all around the world, different days’ have the same ‘dot’ – is instead used to state that
the Moon rotates. On ‘how fast’ it spins, a student answers: ‘Is it the same speed as
the spin on its orbit? … It turns the same as the speed … like the speed of when the
Moon turns, it’s the same as the speed… the Moon rotating’. The teacher responds:

Yeah, alright. Did everybody understand? … Now, let’s watch a video to explain
why it spins on its axis once every twenty-nine and a half days.12

So, the teacher has not added to the animation that: the same side of the Moon
is always seen from Earth; this fact should be used by students to judge whether
88 Karl Maton and Sarah K. Howard

the animation’s simulations are accurate; and several simulations will be inaccurate.
Instead he plays the first (inaccurate) orbit and says: ‘Now, it’s just a visual with
some … but you actually have to watch it, okay? There’s no sound, just watching’.
After the first orbit he pauses the video and says ‘Alright, so that’s what happens.
We’ll play that again. Alright? We’ll play that again’. He thus presents the inaccurate
orbit as accurate.
The teacher returns to the start and states that the red ‘side’ of the Moon is
‘the dark side’ and ‘the side that you actually see’. He asks ‘what happens if it
doesn’t spin?’ and plays the first orbit again. These additions still do not establish
that only one side is visible from Earth and that simulations may be inaccurate.
The students remain unusually silent and the teacher answers his own question:
‘we can see the opposite side of the Moon… but we don’t ever see the opposite
side of the Moon, so that can’t work’. To this a student loudly exclaims ‘What?!’.
The teacher then plays the second orbit (fast rotation) while saying ‘Now, so
again you’d see all sides of the Moon, so that’s not the case’. A student responds
hesitantly: ‘I don’t think… does the Moon spin like that?’. As the third (correct)
orbit plays, a student adds ‘It seems the red side is always facing us’, to which the
teacher confirms ‘The red side is always facing the Earth … So yeah, it takes –
every time it orbits the Earth, it spins once on its axis’. Whether by ‘red side’ he
means in the animation or reality is not clear. After the teacher says this is what
the student was saying in his answer prior to the animation, another student
loudly exclaims: ‘There’s only one person understanding! There’s only one person
understanding!’.
The teacher pauses the animation and tells students to write down a caption
as their answer: ‘It rotates on its axis once in the SAME time it takes to orbit us
once!’. After two minutes he continues the animation, reading out a caption that
calls the phenomenon ‘synchronous rotation’.The teacher ignores the Sun’s shadow
simulation and eschews the opportunity to connect the animation with phases of
the Moon. He then opens a new PowerPoint slide and begins discussing that topic
separately.
In summary, the animation offers much that is needed for the teacher’s task. He
activates some of the animation’s limited affordances for interactivity, principally
pausing and replaying. However, while doing so he deactivates the animation’s
epistemic affordances. Instead of adding knowledge that could turn to purpose the
simulations of its first part, introjecting ‘no rotation orbit’ and ‘fast rotation orbit’
rightwards on Figure 4.5 (to become PA–, RA++), he suggests the first orbit is
correct and within his core target, replaying the orbit as if it was accurate. What
is and is not core target content is thus unclear. As illustrated by the outbursts of
students expressing confusion, it is likely many remained in the dark about the
Moon’s synchronous rotation. Second, he does not add to the second part (on the
Sun’s shadow) anything to activate its affordance for connecting to his next topic.
Thus the task remains segmented from other knowledge he discusses in the same
lesson.
 Animating science 89

Activating epistemic affordances through pedagogic work

The teacher’s targets: The seasons
To illustrate how a multimedia animation can be successfully integrated into teach-
ing a scientific explanation, we turn to a different teacher at a different school but
teaching the same unit (‘Earth and space sciences’) in the same curriculum (New
South Wales) at the same level (Year 7 secondary school). The teacher makes her
target for the lesson explicit at the outset:

The last time I saw you, we were talking about day and night, we were talking
about the tilt of the Earth on its axis, and we had started to discuss seasons.
That’s what we’re going to be looking at today.

As she tells students, the core target content for the task is how ‘the tilt of our axes
and our position around the Sun’ shape Earth’s seasons and her core target purpose
is to support their understanding of that explanation:

If you’d like to make some notes of things that you might think are important,
it’s a good time. Later on, we’re going to ask some questions to see how much
you understand.

Scientific explanations can become extremely complex. Not all factors in an expla-
nation may be taught in a particular syllabus, year of schooling, unit or task. ‘The
seasons’ is one such explanation – not everything on the topic necessarily lies within
this teacher’s targets. A ‘constellation analysis’ of this lesson by Maton and Doran
(chapter 3, this volume) shows the factors involved in this specific task to be: Earth’s
axis is tilted, Earth has hemispheres, Earth orbits the Sun, that bringing those three
together means the hemispheres point towards or away from the Sun through the
year, the Earth receives energy from sunlight, that pointing towards or away the
Sun means the angle of that sunlight varies, and that this creates the variations in
temperature through the year known as ‘seasons’. As summarized in Table 4.4, this
summarizes the teacher’s core target, both its content (those factors and relations) and
purpose (teaching those factors and relations) – it shows what she means by ‘the tilt
of our axes and our position around the Sun’.
Maton and Doran (Chapter 3) also show that the teacher discusses other factors
in different explanations of the seasons she gives during the rest of the lesson, specifi-
cally: that the Earth rotates, that pointing towards or away from the Sun while the
Earth rotates leads to variations in the length of daylight, and that changes in daylight
length helps explain the seasons.These factors form her ancillary target (see Table 4.4).
Her associated non-target comprises other topics in the unit, such as the Moon, or fur-
ther factors explaining the seasons not discussed in the lesson, such as how variations
in sunlight angle create variations in the concentration of sunlight in each hemi-
sphere.13 Finally, unassociated non-target knowledge is anything else outside her target.
90 Karl Maton and Sarah K. Howard

TABLE 4.4 Specific translation device for seasons animation task

target Explanations of ++ core Explanation in task, involving:

the seasons in Earth’s tilt, hemispheres, orbit,
this lesson hemispheres point towards or away
from Sun, solar energy, variations
in angle of sunlight, variations in
temperature called ‘seasons’
+ ancillary Factors for seasons in other
explanations in lesson (e.g. Earth
rotation and length of daylight).
non- Other knowledge – associated Other science knowledge related to
target seasons not covered in lesson (e.g.
sunlight concentration) or topics
in other lessons of the unit (e.g. the
Moon)
–– unassociated Knowledge from beyond the unit
(e.g. door shutting sound effect)

The seasons animation

The multimedia object is a two-minute long animation called ‘What causes Earth’s
seasons?’ that includes entities, movements, visual effects, captions, labels, spoken
narration (a female human voice with American accent), music and sound effects.14
The animation can be distinguished into four main explanatory stages: (i) focusing
the question of what causes Earth’s seasons on the role of axial tilt; (ii) outlining an
explanation; (iii) showing how this leads to seasons changing through the year; and
(iv) summing up.
The animation begins by asking as voice and title text: ‘What causes Earth’s
seasons?’. Electronic, ‘jazzy’ music continues throughout the animation. Patterned
panels leave the title screen to the sound of a heavy door shutting. A photoreal-
istic Earth rotates (to whirring sounds) and becomes abstracted, as illustrated by
Figure 4.6, to display continents, latitude and longitude lines, and an axis line. At
the same time the voice answers itself: ‘Earth’s seasons are caused by Earth’s tilt on
its axis. Instead of going straight up and down [a green vertical line appears briefly
through the Earth], Earth’s axis tilts 23.5 degrees’. Then the Earth orbits the Sun
(while orbited by the Moon), as the voice asks: ‘Can you see why the tilt causes
Earth to have seasons throughout the year? Let’s find out!’.
Second, over a static, rotating Earth, as illustrated in Figure 4.7, the narration
outlines how tilt affects the angle of sunlight:

Here’s the energy of the Sun [rumbling sound as yellow band extends from Sun
to Earth]. Notice that it hits the lower half of Earth [cymbals sound as equator
glows; lower half turns yellow], called the southern hemisphere, most directly
 Animating science 91

FIGURE 4.6 Seasons animation: Showing Earth’s tilt (at 18 seconds)

[hemispheres labels appear].The Sun’s energy hits the northern hemisphere too
[sunlight hitting top half turns blue], but look at how the northern hemisphere
is pointed away from the Sun [arrow upwards through Earth’s axis]. Also, the
light hits at an angle [two arrows appear in sunlight band] that causes the
energy to be spread out over a greater area. So, at this spot in Earth’s revolution
it receives less of the Sun’s energy.

The narration then describes the effects of different sunlight concentration on

hemispheres:

FIGURE 4.7 Seasons animation: Different angles of sunlight hitting the Earth (at 51
seconds)
92 Karl Maton and Sarah K. Howard

FIGURE 4.8 Seasons animation: Effects of angle and concentration of sunlight – winter
(at 1:23 minutes)

When more of the Sun’s energy hits the southern hemisphere it causes the
temperature to go up [sun patterns and ‘Summer’ appear in yellow band]. It’s
summertime! [Sounds of children playing] More sunlight, longer days! With
less of the Sun’s energy hitting the northern hemisphere, it gets a lot colder
[snowflake patterns and ‘Winter’ appear in blue band; see Figure 4.8]. Put on
your winter coat! [Sounds of wind howling] Less sunlight, shorter days!

Third, the animation shows the seasons through the year: ‘Let’s see how the seasons
change as Earth revolves around the Sun’. The caption ‘December’ appears (to door
sound) and the animation proceeds through months of the year (signalled by a keyboard
tap sound) as the Earth orbits the Sun.At ‘March’ the Earth is labelled with ‘Spring’ with
green patterns on the top half and ‘Autumn’ and brown patterns on the bottom half; at
‘June’, as illustrated by Figure 4.9, ‘Summer’ and ‘Winter’ with yellow and blue patterns;
and at ‘September’, ‘Autumn’ and ‘Spring’ with brown and green patterns. Month cap-
tions move around the screen, with the door sound when they stop.
Finally, the voice sums up:‘So why do seasons change? Because the tilt of Earth’s axis
causes the hemispheres to receive different amounts of the Sun’s energy’. The Earth
orbits the Sun as month captions are tapped through again. The door sound marks
the appearance of a photograph of a green hillside next to text stating ‘Seasons occur
because of Earth’s tilt!’, as the voice says breathily: ‘That’s some cool science!’.

Affordances for the task

The animation offers limited interactive affordances. It is accessible, shareable,
watchable, viewable, listenable, and can be played back. One can choose to only
listen or only watch, but the animation is not easily manipulated; for example,
 Animating science 93

FIGURE 4.9 Seasons animation: Summer/Winter (at 1:41 minutes)

the voice cannot be altered, specific audio elements (such as the music) cannot be
turned off and specific content (such as discussion of particular factors) cannot eas-
ily be removed. Nonetheless, these affordances for ‘instructive’ interactions can meet
the teacher’s purpose of students making notes of key issues explaining the seasons.

Epistemic affordances
The animation’s complex mix of elements exhibit different degrees of match to
the task. So we shall discuss visual and audio elements in turn, using the translation
device of Table 4.4.
Visual elements include text (title, labels, captions), entities (e.g. Earth), move-
ment (e.g. rotating), a photograph and various effects (e.g. yellow band for sunlight).
Figure 4.10 locates these elements on the autonomy plane: most are at the far top
(PA++) and/or far right (RA++). Overall, the animation thereby visually matches
the teacher’s core target content and/or purpose.Taking each kind of visual element
in turn, text appears frequently, though each briefly. All text – summarized in Figure
4.10 for reasons of space as ‘all captions’, ‘all labels’ and ‘titles’ – is situated in the
teacher’s core target (PA++, RA++). Entities and movements, such as Earth and its
orbit, are also largely in her core target. The exceptions are: Earth’s rotation and the
Moon and its orbit. Both, however, were previously discussed in the class: Earth’s
rotation was a factor in an explanation discussed earlier in the lesson (PA+, RA+);
and the Moon and its orbit was discussed in the previous lesson (PA–, RA–). Other
visual elements are in the teacher’s core target: axis line, sunlight band and sunlight
arrows are part of the explanation and students’ ability to use these symbols in dia-
grams is checked by the teacher later in the lesson. Other visual effects are mostly
brief and located at the bottom right of Figure 4.10.While their contents – colours,
patterns, glowing – are far from the seasons (and so coded as unassociated non-
target content or PA– –), they all serve the teacher’s core target purpose (RA++) by
94 Karl Maton and Sarah K. Howard

FIGURE 4.10 Visual elements of the seasons animation

emphasizing factors such as different hemispheres and seasons. The one exception
is only momentary: patterns framing the static images at the start and end (PA– –,
RA– –).
Audio elements include a spoken narration, voice tone, music and various sound
effects (e.g. howling wind). The almost ever-present narration, shown as non-italics
in Figure 4.11, covers all factors and relations that comprise the explanation of
seasons in the teacher’s core target (PA++, RA++).15 It would thus appear ide-
ally suited to the task. However, the narration also includes two further factors:
daylight length is discussed in another explanation during the lesson but not in
this task (PA+, RA+); and concentration of sunlight, though related to the seasons,
is not addressed in the lesson (PA–, RA–). The narration also exudes an excitable,
breathy tone, especially in playful phrases (e.g. ‘Put on your winter coat!’) – these
are located at the bottom left of the plane, far from the teacher’s core target (PA– –,
RA– –). Other audio effects (italics in Figure 4.11) are frequent but individu-
ally brief. Sounds of rumbling, children playing, etc. represent non-target content
(PA– –) but serve the teacher’s core target purpose (RA++) by emphasizing fac-
tors such as solar energy and seasons. Exceptions are: whirring for Earth’s rotation
(PA– –, RA+); and the recurrent door shutting sound and ever-present background
music (PA– –, RA– –).
In summary, the animation’s interactive affordances can support the teacher’s
‘instructive’ task of students making notes of key issues. The animation also exhibits
 Animating science 95

FIGURE 4.11 Audio elements of the seasons animation

significant epistemic affordances for the task. Visually, what does not lie within the
teacher’s core target is either part of the lesson and unit of study or only a momen-
tary feature of the animation. Aurally, the spoken narration includes all the factors
and relations that the teacher wishes to convey. In short, the animation is well
selected for both interactive and epistemic affordances. However, as we empha-
size, multimedia objects are unlikely to perfectly match a specific task in a specific
classroom. In this case, two issues could potentially require pedagogic work. First,
factors in the narration that the teacher does not wish to include (daylight length
and concentration of sunlight) are not easily subtracted. Second, the ever-present
voice tone and music, if not wanted, can only be subtracted by omitting all audio
elements. Given that the visual elements do not by themselves provide an explana-
tion, this requires adding a commentary. The success of such substitution – replacing
non-target with core target elements – would depend on the teacher’s ability to
discuss core target knowledge while selecting and turning to purpose the anima-
tion’s visual elements.

Seasons in the Sun

The teacher undertakes the pedagogic work required to maximize the animation’s
epistemic affordances for supporting her task. She does so by taking advantage of
96 Karl Maton and Sarah K. Howard

its watchability and listenability, muting the animation and substituting the audio
with her own narration over the visual elements. In an interview, she highlighted
its terminology, music and voice tone as reasons – epistemic rather than interactive
affordances:

I didn’t use the sound in that one because I wanted to be able to use the lan-
guage that I was using in class, that the sound in that animation was a very hard
American voice, almost computerised.There was this tinkly music in the back-
ground. I thought, ‘This is going to be distracting’.

As we state above, such substitution depends on a teacher’s ability to add audio

content while activating an animation’s visual elements. In this case, the teacher is
highly adept. Without pausing the animation, she highlights, through her speech or
gestures, specific entities, movements, visual effects and text while introducing fac-
tors and explaining how they relate together to create the seasons. The teacher also
asks questions of students that are timed so that the appearance of captions confirms
their responses. In short, the animation and teacher narration form an integrated
multimedia and multimodal performance.
To show the pedagogic work this involves, we shall discuss the task using the
same four stages as the animation’s voiceover (above). As outlined earlier, her core
target is to explain how ‘the tilt of our axes and our position around the Sun’ shape
Earth’s seasons. First, she highlights the key issue of tilt:

So [reading out title] what causes Earth’s seasons? When we have a little look at
this animation, what we’re looking at here is obviously the Earth spinning on its
axis and you can see that that axis [hand gesture matching its axis line] is about
23.5 degrees [points to axis line] from [hand traces vertical line] what could be
the [air quotes gesture] “theoretical midline” of the Earth. We know [repeatedly
points to Earth as it orbits Sun] that that axis holds itself. Can you see [pointing
to Earth] how it’s holding its 23.5 degrees as it moves around the Sun?

Figure 4.12 plots key elements of her narration as a whole.16 As the above shows,
she begins by quickly turns from describing what is being shown (Earth rotates) to
the issue of axial tilt, selecting visual elements relevant to her core target: Earth’s axis,
tilt and orbit (PA++, RA++). The teacher’s regular gestures are non-target content
(PA– –), because hand movements are far from her target content, but they serve her
core target purpose (RA++) by, for example, highlighting tilt.
Second, the teacher set out an explanation of how tilt leads to differences in the
angle of sunlight in different hemispheres:

Now, [pointing to yellow band from Sun] when the Sun’s rays hit the Earth, the
Earth has that also other [air quotes] “theoretical midline”, the equator [hand
traces equator line], that breaks it into half: [pointing to labels] northern hemi-
sphere and southern hemisphere. Okay? We’re [pointing to herself] in the
 Animating science 97

FIGURE 4.12 Teacher narration of the seasons animation

southern hemisphere and at some times [tracing horizontal line in yellow band]
the sunlight will strike, in this case, the southern hemisphere at that 90 degree
angle. All right? And here [hand tracing angled arrow in sunlight band], this is
what we can sometimes call an ‘oblique’.You heard that person in the vox pop
talk about it. Yeah? Or ‘glancing’. So [hand tracing where angled arrow had
been] these are like the 15 degrees, okay? We’re getting sunlight bouncing off,
and here [hand tracing where horizontal arrow had been] it’s hitting directly.

Through this explanation the teacher’s gestures continue linking visual elements to
her core target knowledge: the sunlight band is recruited into discussing the Sun’s
rays, the equator line helps identify hemispheres, and the arrows are mirrored by
gestures that indicate sunlight angles. In comparison to the muted narration, she
offers a simpler explanation by not discussing the effects of sunlight angle on day-
light length or on sunlight concentration and their effects, in turn, on temperatures.
Instead, these elements are substituted by: locating herself and her students in the
southern hemisphere; providing more information about sunlight angles, specifi-
cally the terms ‘oblique’ and ‘glancing’; and explicitly connecting that information
to a ‘vox pop’ video they watched earlier in the lesson. Everything she adds serves
her core target purpose (far right on the Figure, or RA++) and helps integrate the
task and animation into the rest of the lesson and thus the classroom experiences
shared by students.
98 Karl Maton and Sarah K. Howard

Third, the teacher checks student understanding of the effects of sunlight angle
on seasons and, emphasizing the Earth’s tilt during its orbit of the Sun, discusses how
seasons change through the year:

So, in this case, what season does the southern hemisphere have? [Students say
‘summer’, as label appears]. And what about the northern hemisphere? [Students
say ‘winter’]. Winter! Good. Now, when this moves [gesturing to Earth] around
the Sun, and holds its axis, because it doesn’t … the axis won’t change around
like that, when this moves to the other side of the Sun, we will see the other side
of the Earth … so [pointing to screen showing Earth with ‘spring’ and ‘autumn’
labels] now we’re on the back side of the Sun. We’ve got spring and autumn.
Now we’re on the opposite side [Earth with ‘summer’ and ‘winter’ labels]. Now
the northern hemisphere has summer and the southern hemisphere has winter.
Now, we come out to the [air quotes] “fourth side” of the Sun [Earth with
‘autumn’ and ‘spring’ labels], we get autumn and spring split around.

Here the questions to students are not part of the explanation (PA–) but turned to
her core target purpose (RA++), as shown on Figure 4.12. She then recruits visual
entities (the Earth and the Sun), movements (Earth orbiting) and labels (e.g. ‘win-
ter’) to discuss the changing seasons, remaining throughout inside her core target.
Finally, the animation continues as a student queries the description of ‘sides’ and
the teacher explains that ‘sides’ is a way of emphasizing the change of seasons by
referring to quarters of Earth’s orbit: non-target content turned to her core target
purpose (PA–, RA++). The animation’s final textual summary is not referred to by
the teacher. Instead, after it concludes she sums up: ‘So, we know the Sun is hitting
the Earth. We know that some parts of the Earth will be getting the full force and
some will be getting those glancing rays.’ She asks the students whether the Sun
changes and emphasizes: ‘typically the strength of the Sun never changes. The thing
that will change is our position in relationship to the Sun … We get seasons because
of being tilted towards or away’. This concludes the task; the teacher then moves
to a task involving an animated simulation that shows the seasons change through
the year.
In summary, the teacher uses the animation’s ability to be watched and muted
in order to reshape its epistemic affordances. Her substitution of audio elements
with her own narration and gestures adds elements that are almost entirely on the
far right of Figure 4.12 – everything she says or does is turned to her core target
purpose. Indeed, Figure 4.12 underplays the extent to which her performance as
a whole, and not just her audio narration, is replete with core target knowledge
because, for reasons of space, we summarized ‘gestures to visual elements’ rather than
included all those elements she activates in this way. Returning to the plane of visual
elements in Figure 4.10, she draws extensively on entities, movements and labels
located within her core target: axis line, Earth, Sun, orbit, sunlight band, arrows,
titles, captions and labels. In contrast, she ignores those visual elements outside her
target.Those which could have served her purpose – the visual effects at the bottom
 Animating science 99

right of Figure 4.10 – were substituted by hand gestures bringing attention to key
factors. So, the teacher selects the animation’s visual elements that match her task
and integrates those elements through gesture with an audio narration that provides
a simpler, more focused explanation that is, in addition, more integrated into the
wider lesson and classroom experience. Her strategy of substitution thereby not
only activates the epistemic affordances of the animation but more closely integrates
the multimedia object into the task and lesson.

Conclusion
Multimedia objects such as animations are a common part of teaching science.
Questions of how teachers can integrate such ready-made objects into classroom
tasks are urgent. However, such questions are sidelined by the cognitive–learn-
ing–design axis dominating research into multimedia learning and technology
affordances. To address questions of integration requires studies of actual classroom
practice and a framework that does not reduce teaching to designing and what is
being taught and learned to homogeneous information. In this chapter we extended
the metaphor of ‘affordances’ beyond its current focus on enabling students interac-
tions to embrace the epistemic affordances of knowledge practices. We drew on the
LCT dimension of Autonomy to bring one aspect of these knowledge practices
into view. The concepts of positional autonomy and relational autonomy helped fore-
ground relations between the contents and purposes of multimedia and those of
classroom tasks. The concept of targets helped ensure that specific classroom prac-
tices remained the centre of analysis, thereby avoiding the slide into generic descrip-
tions of context-free ‘abilities’ towards which affordance frameworks tend. These
concepts provided a platform for seeing that teaching is more than design, that
integration is more than selection, and that affordances are more than enabling
interactive practices among students.
We illustrated the value of seeing epistemic affordances with two examples of
teaching science with animations. Both teachers selected animations whose limited
interactive affordances matched their ‘instructive’ needs – from the conventional
perspective of ‘affordances’ there is little to distinguish between them.Yet they offer
contrasting examples of teaching science with multimedia. The first teacher did not
undertake the pedagogic work required to activate the epistemic affordances of the
Moon animation for supporting his task (teaching about synchronous rotation) and
left that task segmented from the rest of the lesson. The second teacher undertook
pedagogic work by substituting the audio elements in the seasons animation that did
not match her specific task with her own narration and connecting this knowledge
to its visual elements through gestures. From the viewpoint of epistemic affordances
for teaching scientific explanations, then, these examples are radically different.
By bringing teaching and knowledge practices into view, autonomy analysis
emphasizes that integration does not end with design or selection. While multime-
dia are unlikely to perfectly match the knowledge needs of a specific classroom task,
that is not necessarily an obstacle to their integration. Elements beyond a teacher’s
100 Karl Maton and Sarah K. Howard

core target offer possibilities for teaching and pedagogic work by teachers and stu-
dents can ‘connect to content’ (strengthen PA) and ‘turn to purpose’ (strengthen
RA) those elements of a resource that lie beyond their core targets. Thus, ready-
made multimedia objects can be integrated through classroom practice. Moreover,
the epistemic affordances of a multimedia object may differ wildly from one task to
the next. The coding of an object is in relation to targets that are highly specific: a
particular task in a particular lesson. The same multimedia objects analyzed in rela-
tion to other teacher’s targets, to other tasks of these teachers or to students’ targets
will result in different codings. Epistemic affordances depend on the specific knowl-
edge practices being taught and learned. Thus, design and selection are not the end
of the matter: integration is a practice by teachers and students in classrooms.
Of course, seeing teaching, knowledge practices and epistemic affordances is but
a small step forward. One cannot expect teachers to undertake time-consuming
autonomy analyses of tasks and multimedia. The next step requires more practice-
oriented outcomes. Autonomy analyses of classroom practices that integrate math-
ematics into science (Maton and Howard, chapter 2, this volume) and everyday
experiences into History teaching (Maton and Howard 2018) have developed the
pedagogic practice of ‘autonomy tours’, which is already having an impact on teach-
ing practice. Similar pedagogic principles are required for integrating multimedia
into teaching. Nonetheless, this chapter has illustrated that LCT offers the potential
for addressing such issues of integration. All we have to do now is to activate those
affordances.

Notes
1 This holds for models of both ‘technological’ affordances and also ‘social’ and ‘educa-
tional’ affordances (Kirschner et al. 2004) – all concern interactions.
2 We use ‘epistemic affordances’ to encompass all knowledge practices, whether involving
epistemological constellations of concepts and empirical descriptions or axiological con-
stellations of affective, moral, ethical or political meanings (Maton 2014: 148–70).
3 See: www.legitimationcodetheory.com.
4 There is a growing body of work using Semantics to study multimedia in science, such as
student-generated digital products (Georgiou 2020, Georgiou and Nielsen 2020). From a
complementary perspective, He (2020) explores the semiotic resources of science anima-
tions with systemic functional linguistics.
5 Students, other teachers, etc. may have different targets, allowing comparative analysis of,
for example, learning experiences.
6 We draw on research funded by the Australian Research Council (DP130100481), led by
Karl Maton, J. R. Martin, Len Unsworth and Sarah K. Howard.
7 The concepts can be enacted in different ways. In Maton and Howard (2018, chapter 2,
this volume), we focus on changes in knowledge practices over time and so plot ‘auton-
omy pathways’ between different ‘autonomy codes’ on the plane. Here we do not name
autonomy codes, as our focus is on 16 rather than four modalities, and we plot a syn-
chronic analysis of positions, as our focus is relating task and object rather than changes
in strengths through time.
 Animating science 101

8 The animation is freely available on the Internet.The earliest we discovered was uploaded
to YouTube in 2009 by ‘astrogirlwest’, about whom no further information is available.
9 For both animations analysed in this chapter, it is easy to imagine different resources that
could make the tasks more collaborative for students. However, that would not diminish
their ability to support ‘instructive’ interactions and thus match the teachers’ core target
purposes.
10 Relative duration is, of course, but one indicator of prominence – one could also exam-
ine relative size, position, and other attributes.
11 One uploader to YouTube added their own starting caption: ‘This is a video explaining
why we always see the same side of the moon all the time’; see https://www.youtube.
com/watch?v=6vkVxu04DcE.
12 The teacher confuses the duration between New Moons (29.5 days) with the Moon
completing a revolution on its axis (27.3 days). However, accuracy here would not
change the degree to which he integrates the animation.
13 These factors are highlighted in a constellation analysis of secondary school science text-
books by Maton and Doran (chapter 3, this volume).
14 The animation is dated 2006, accredited to Ignite! Learning, and freely available at
https://www.teachertube.com/videos/what-causes-seasons-on-earth-657.
15 We have included the logic of its explanation by using ‘+’ and ‘=>’ to refer to links
between nodes; see Maton and Doran (chapter 3, this volume).
16 We have highlighted the logic of her explanation with ‘+’ and ‘=>’. Maton and Doran
(chapter 3, this volume) discuss this logic in greater detail.

References
Antonenko, P., Dawson, K. and Sahay, S. (2017) ‘A framework for aligning needs, abilities and
affordances to inform design and practice of educational technologies’, British Journal of
Educational Technology, 48(4): 916–27.
Berney, S., and Bétrancourt, M. (2016) ‘Does animation enhance learning?: A meta-analysis’,
Computers & Education, 101: 150–67.
Bower, M. (2008) ‘Affordance analysis – matching learning tasks with learning technologies’,
Educational Media International, 45(1): 3–15.
Bower, M. (2017) Design of Technology-Enhanced Learning: Integrating Research and Practice,
Bingley: Emerald Publishing.
Georgiou, H. (2020) ‘Characterising communication of scientific concepts in student-gener-
ated digital products’, Education Sciences, 10(1): https://doi.org/10.3390/educsci10010018.
Georgiou, H. and Nielsen, W. (2020) ‘New assessment forms in higher education: A study
of student generated digital media products in the health sciences’, in C. Winberg, S.
McKenna and K.Wilmot (eds) Building Knowledge in Higher Education, London: Routledge,
55–75.
Hammond, M. (2010) ‘What is an affordance and can it help us understand the use of ICT
in education?’, Education and Information Technologies, 15(3): 205–17.
He, Y. (2020) ‘Animation as a Semiotic Mode: Construing Knowledge in Science Animated Videos’,
unpublished PhD thesis, University of Sydney, Sydney.
Howard, S. K. and Maton, K. (2011) ‘Theorising knowledge practices: A missing piece of the
educational technology puzzle’, Research in Learning Technology, 19(3): 191–206.
Howard, S. K. and Maton, K. (2013) ‘Technology & knowledge’, AERA 2013 SIG-Computer and
Internet Applications in Education, San Francisco, USA: AERA, 1–8.
102 Karl Maton and Sarah K. Howard

Jenkinson, J. (2018) ‘Molecular biology meets the learning sciences: Visualizations in educa-
tion and outreach’, Journal of Molecular Biology, 430(21): 4013–27.
Kirschner, P., Strijbos, J.-W., Kreijns, K. and Beers, P. J. (2004) ‘Designing electronic collabora-
tive learning environments’, Educational Technology Research & Development, 52(3): 47–66.
Li, J., Antonenko, P. D. and Wang, J. (2019) ‘Trends and issues in multimedia learning research
in 1996–2016: A bibliometric analysis’, Educational Research Review, 28; 100282
Maton, K. (2014) Knowledge and Knowers: Towards a Realist Sociology of Education, London:
Routledge.
Maton, K. (2020) ‘Semantic waves: Context, complexity and academic discourse’, in J. R.
Martin, K. Maton and Y. J. Doran (eds) Accessing Academic Discourse, London, Routledge,
59–85.
Maton, K. and Chen, R. T-H. (2016) ‘LCT in qualitative research: Creating a translation
device for studying constructivist pedagogy’, in K. Maton, S. Hood and S. Shay (eds)
Knowledge-building: Educational studies in Legitimation Code Theory, London: Routledge,
27–48.
Maton, K. and Howard, S. K. (2018) ‘Taking autonomy tours: A key to integrative knowl-
edge-building’, LCT Centre Occasional Paper 1: 1–35.
Mayer, R. E. (2003) ‘The promise of multimedia learning: Using the same instructional
design methods across different media’, Learning and Instruction, 13(2): 125–39.
Mayer, R. E. (2014a) ‘Cognitive theory of multimedia learning’, in R. E. Mayer (Ed.)
The Cambridge Handbook of Multimedia Learning, 2nd edition, Cambridge: Cambridge
University Press, 43–71.
Mayer, R. E. (2014b) (Ed.) The Cambridge Handbook of Multimedia Learning, 2nd edition,
Cambridge: Cambridge University Press.
Mayer, R. E. and Fiorella, L. (2014) ‘Principles for reducing extraneous processing in multi-
media learning’, in R. E. Mayer (Ed.) The Cambridge Handbook of Multimedia Learning, 2nd
edition, Cambridge: Cambridge University Press, 279–315.
Mayer R. E. and Moreno, R. (2003) ‘Nine ways to reduce cognitive load in multimedia learn-
ing’, Educational Psychologist, 38(1): 43–52.
Mutlu-Bayraktar, D., Cosgun, V. and Altan, T. (2019) ‘Cognitive load in multimedia learning
environments: A systematic review’, Computers & Education, 141: 103618.
Ploetzner, R. and Lowe, R. (2012) ‘A systematic characterisation of expository animations’,
Computers in Human Behaviour, 28: 781–94.
Sweller, J., van Merriënboer, J. and Paas, F. (2019) ‘Cognitive architecture and instructional
design: 20 years later’, Educational Psychology Review, 31: 261–92.
PART II

Language in science
education
5
FIELD RELATIONS
Understanding scientific explanations

Y. J. Doran and J. R. Martin

Making sense
If you live away from the equator, the seasons are clear. Summer is hot, winter is cold
and spring and autumn are somewhere in between, harbingers of what’s to come. If
you are a student of secondary-school science, this often becomes the starting point
for a considerably more academic discussion: why do seasons occur? And the result-
ing explanation is far removed from everyday experience. It involves the rotation
of the earth, its tilt, its division into northern and southern hemispheres, its orbit
around the sun, the sun’s emission of light and a number of effects resulting from
some combination of these. Our everyday understanding of hot and cold through-
out the year quickly transforms into a large complex of scientific ‘facts’.
An explanation oriented to a popular audience, for example, explains: 1

What Causes Seasons on Earth?

Seasons happen because Earth’s axis is tilted at an angle of about 23.4 degrees
and different parts of Earth receive more solar energy than others.
Because of Earth’s axial tilt (obliquity), our planet orbits the Sun on a slant
which means different areas of Earth point toward or away from the Sun at dif-
ferent times of the year.
Around the June solstice, the North Pole is tilted toward the Sun and the
Northern Hemisphere gets more of the Sun’s direct rays. This is why June, July
and August are summer months in the Northern Hemisphere.

Opposite Seasons
At the same time, the Southern Hemisphere points away from the Sun, creat-
ing winter during the months of June, July and August. Summer in the Southern
Hemisphere is in December, January, and February, when the South Pole is
tilted toward the Sun and the Northern Hemisphere is tilted away.
106 Y. J. Doran and J. R. Martin

Such an explanation pushes us far away from our everyday experience. But for
students to be successful in science, they must be able to both read and write
explanations of this kind regularly. Indeed explaining why seasons occur is a
key topic in the first year of secondary school science in New South Wales,
Australia.
In recent years, dialogue between Systemic Functional Linguistics (SFL) and
Legitimation Code Theory (LCT) has been exploring the nature of such knowl-
edge and the discourse used to organize it (Martin 2011, Maton et al. 2016, Maton
and Doran 2017, Martin et al. 2020). A goal of this research across disciplines is to
develop discipline-sensitive pedagogy and curriculum to improve the learning of
subject-specific knowledge. As part of this endeavour, research has focused on dif-
ferent kinds of knowledge, how knowledge develops over time and how we can
model the knowledge we find in spoken and written texts and across the vast range
of multimodal semiotic resources used in school.
In terms of Legitimation Code Theory, for students to master the scientific
knowledge involved in explaining the seasons, they need to be able to understand
and manipulate intricate constellations of meaning (Maton 2014: 149–159; Maton
and Doran, Chapter 3 of this volume). These constellations organize large networks
of meaning into specific arrangements according to the discipline, the year level
and the phenomenon being explained. In science, what is foregrounded are what
LCT terms epistemological constellations which focus on the ‘content’ of disciplines –
discipline-specific configurations of causal relations, taxonomies and scientific
procedures, along with methods for investigating the world. This is in contrast to
axiological constellations that emphasize specific dispositions, political and aesthetic
stances, morals and ethics, often found for example in the humanities (Martin et al.
2010, Maton 2014, Doran 2020).
From an epistemological constellation perspective, the key components in the
constellation underpinning why seasons occur in this explanation can be synthe-
sized from the text as follows:

Earth’s axial tilt is 23.4 degrees

The amount of solar energy received by different parts of the earth at different times of
the year varies.
The earth orbits around the sun
The earth is divided into hemispheres
Seasons occur

These components are not simply a bunch of isolated ‘facts’ about seasons. In order
to explain seasons, they need to be brought together in specific configurations. For
example, the variation in the amount of solar energy received by different parts
of the earth that underpins seasons is caused by the particular combination of the
Earth’s 23.4 degree tilt, its orbit around the sun, its division into hemispheres and
the fact that the sun emits light that hits the earth. Without any of these compo-
nents, seasons would not occur.
 Field relations 107

The questions then for this chapter are: what are the relations that underpin
scientific knowledge? And how do we see these relations in texts, both through lan-
guage and other semiotic resources? Put another way, we are concerned with what
it is that students are expected to learn when they ‘do’ science.
As far as SFL is concerned, one vantage point from which to explore this is
through the register variable field. In SFL, field is one component of register, along
with tenor and mode, and is concerned with what educators consider the con-
tent of language and semiosis. Seen in terms of SFL, field is a more abstract level
of meaning positioned above the ideational meanings construed through language
at the levels of discourse semantics, lexicogrammar and phonology/graphology.
These levels of abstraction, called strata, are represented as co-tangential circles in
Figure 5.1 (Martin 1992).
Over a number of years, work on field and ideational meaning has provided
a highly productive lens through which to view scientific understandings of the
world (e.g. Lemke 1990, Halliday and Martin 1993, Martin and Veel 1998, Halliday
2004, Martin 2020). More recently, the model of field has been renovated, largely
in response to SFL’s dialogue with LCT (Martin and Maton 2013, Martin et al.
2017, Martin et al. 2020), the development of ideational discourse semantics by Hao
(2015, 2018, 2020a, 2020b, 2020c, Chapter 6 of this volume), and investigations of
a range of semiotic resources used in science, including mathematics, graphs, dia-
grams, animations, other formalisms and body language (Doran 2017, 2018, 2019,
Chapter 7 of this volume; Unsworth 2020; He 2020; Hood and Hao, Chapter 10 of
this volume; Martin et al. in press).

FIGURE 5.1 Strata of language in Systemic Functional Linguistics

108 Y. J. Doran and J. R. Martin

FIGURE 5.2 Classification taxonomy of the seasons

This paper outlines these evolving understandings. It will introduce each com-
ponent of the new model, drawing on a range of linguistic texts and semiotic
resources. It will then use the model to explore seasons as they are taught in second-
ary school in New South Wales, Australia. In doing so, the chapter illustrates how
field can make visible the knowledge teachers need to teach and students need to
learn in order to achieve success in school science.

Field relations: static and dynamic perspectives

The model developed here expands on that of Martin (1992). At its most general
level, field can be described as a resource for construing phenomena either stati-
cally as relations among items or dynamically as activities oriented to some global
institutional purpose. Beginning with the static perspective, this orientation to field
views phenomena as items organized into particular taxonomies. For example, our
common-sense understanding distinguishes four seasons: winter, summer, autumn
and spring. This can be modelled as a classification taxonomy, as in Figure 5.2.
Classification views relations between items in terms of types and sub-types
(class and sub-class). In terms of seasons, the items spring, autumn, winter and summer
are all sub-classes in relation to the more general item of seasons and co-classes in
relation to each other.
An alternate static perspective on phenomena is through composition – the
part-whole relations among items. In terms of explaining the earth’s seasons, com-
position relations are used to divide the solar system into the sun and the earth, and
the earth in turn into the northern and southern hemispheres. The relevant com-
position taxonomy of the solar system for explaining seasons is shown in Figure 5.3.
Our solar system of course contains many different components aside from the
earth and the sun, and the seasons can be divided into many other configurations

FIGURE 5.3 Composition taxonomy of the solar system

 Field relations 109

(northern Australians may well wonder where the wet season or the dry season or
the build-up are in the classification introduced above). But the important point
here is that for the particular problem of explaining how seasons work in junior
high school, ethnocentric though it may be, these are the composition and classi-
fication relations that are relevant. In LCT terms, these meanings form part of the
specific constellation comprising knowledge of the seasons.
As far as our model of field is concerned, taxonomies may be indefinitely wide
or deep. For example, the composition taxonomy introduced above indicates two
parts for each level, whereas the classification taxonomy distinguishes four seasons.
And in principle, there may be any number of subtypes or parts depending on what
one is looking at. Similarly, the composition taxonomy of the solar system shows
a slightly deeper hierarchical arrangement than the classification taxonomy, with
three levels of the composition (solar system → earth → northern hemisphere, for
example). Again, this may be expanded indefinitely, with any number of parts and
wholes relevant for a given field. Indeed the expansion of taxonomies is one of the
key features of scientific knowledge. This is one respect in which science differs
from common-sense understandings whose utility tends to demand less width and
depth of classification and composition (Wignell et al. 1989).
We can begin mapping the types of relations underpinning fields with the net-
work in Figure 5.4. This network says that from a static perspective, a field may
involve a single item or multiple items arranged into taxonomies of either compo-
sition or classification. The n above taxonomy indicates it may be indefinitely wide
and/or deep.
An alternate perspective on field involves construing phenomena dynamically in
terms of activities. Activity involves some sort of change that is oriented to some
global everyday, professional or institutional purpose. One activity associated with
the seasons, for example, is:

The sun warms the earth

This example gives a single, isolated activity, specified lexicogrammatically by the

Process warms, and involving two items: the sun and the earth. In longer explanations,
rather than construing just a single undivided activity, it is common for an activity
to be construed as a series of smaller activities. For example, the above activity could
be reconstrued as a series of activities as follows:

FIGURE 5.4 Network for a static perspective on field

110 Y. J. Doran and J. R. Martin

The sun emits light

^
which hits the earth
^
and warms it.

Here we have a momented activity, where a single activity (the sun warms the earth)
is construed as three separate activities, all oriented to the more general activity of
warming.2
Each activity can be momented into any number of further activities within the
limits of a field. For example, in Figure 5.5 below, from a secondary school biology
textbook (Greenwood and Allen 2004), the unmomented activity inflammation is
momented through various subheadings into four stages:

Increased diameter and permeability of blood vessels

^
Phagocyte migration
^
Phagocytosis
^
Tissue repair

Each of these activities can in turn be momented into another series of activities.
For example, at the bottom of the figure, phagocytosis is momented as:

Detection
^
Ingestion
^
Phagosome forms
^
Fusion with lysosome
^
Digestion
^
Discharge

This produces three tiers of activity shown in Figure 5.5.

An unfolding series of activities can be related in one of two ways, through impli-
cation or expectancy. Implication relations describe series of activities where one activ-
ity necessarily entails another (if one activity happens, then another always does).3
In the following momented activity from a year seven science classroom, the
activity tides is described as a necessary result of the gravitational pull of the moon and
the sun and the rotation of the earth, through the Process causes:
 Field relations 111

FIGURE 5.5 Momenting the activity inflammation (three tiers) (Greenwood and Allen
2004: 118–119)

‘The gravitational pull of the moon and the sun, and the rotation of the earth
causes tides.’4

Implication series are most common in scientific explanations where events tend
to be described in terms of causal or conditional relations of entailment. In terms
of the ideational discourse semantic model developed by Hao (2020a), the activities
gravitational pull of the moon and the sun and the rotation of the earth are realized by
occurrence figures in discourse semantics that are in turn realized metaphorically
through nominal groups, while the activity tides is realized by an activity entity (see
Hao, Chapter 6 of this volume). The implication relation linking these activities is
realized discourse semantically by a causal connexion and lexicogrammatically by
the Process cause.5 This gives the implication series:
112 Y. J. Doran and J. R. Martin

The gravitational pull of the moon and the sun, and the rotation of the earth
^ (causes)
tides

Analyzed for multiple strata, the realization of this momented activity in discourse
semantics and lexicogrammar is:

The gravitational pull of the sun and the moon and the rotation of the earth causes tides

field momented implication activity

disc. sem. sequence
lexicogram. clause

Viewed in terms of the realizations of individual activities, this example is:

The gravitational pull of the sun and the moon and the rotation of the earth causes tides
activity activity activity
occurrence figure connexion occurrence figure connexion activity entity6
nominal group nominal group verbal group nominal group

If we unpack the grammatical metaphor in this example, we can see the implica-
tion series more clearly:

The moon and the sun are pulled by gravity

and the earth rotates,
^ (so)
tides occur.

The use of the consequential connexion (causes or so) indicates that the tides neces-
sarily arise due to the other two activities. As Hao (2018) has shown, in scientific
explanations implication series can also be realized by temporal connexions. For
example, both series of activities below are related through implication (where one
activity entails another), although one uses conditional connexion (if… then) and
the other involves temporal connexion (when):

If you move an electron from one orbit to another then energy is either
absorbed or released.
When you move an electron from one orbit to another, energy is either
absorbed or released.

Whereas scientific explanations tend to be concerned with entailment, texts

orienting toward how to do science (such as experimental procedures, procedural
recounts or protocols, or stories of scientific discoveries) tend to construe activity
 Field relations 113

in terms of expectancies rather than as logical necessities (Hood 2010, Hao 2015,
2020a). In these texts, temporal connexions are not used to realize a series of activi-
ties related by implication, but rather by expectancy. For example, in the following
description from a video shown in a high-school science classroom of how data is
collected to measure tides, the gathering and uploading are expected to occur together
in a temporal sequence (linked by and then), but one activity does not necessarily
entail the other. This kind of text records scientific activity rather than offers a sci-
entific explanation of phenomena.

‘The data is gathered from the station and then is uploaded via the satellite
every 6 minutes’

One of the key features of expectancy relations is that the series of activities can be
interrupted or go against what is expected (as any science teacher or observer of a
classroom experiment can attest). An interruption of an expectancy series is often
shown through concessive connexion – for example but in The data was gathered from
the station but was not uploaded via the satellite. The ability to play with expectancy
in momented activities in this way is the basis of many story genres, where some
unexpected and often dramatic activity occurs to establish the main complication
of the story (Martin and Rose 2008).
Individual activities can take a number of forms. In the broadest terms, all activi-
ties involve some sort of change. As exemplified by the orbit and revolve activities in
the explanation of seasons given above, this change can often be cyclical:

Our planet orbits the sun on a slant.

Earth revolves around the sun

Similarly, in the following brief explanation of tides, rotation indicates a cyclical

activity:

The gravitational pull of the moon and the sun, and the rotation of the earth causes tides.

These cyclical activities involve an event that recurs an indefinite number of times.
Put another way, if rotation or orbit were to be momented, the cycle would repeat
indefinitely. Indeed, this is made explicit in a teacher’s description of the rotation of
the earth, where the cyclical activity spinning is momented into it’s moving day night,
day night:

The earth is spinning on its axis, happening every day. Every time it spins, it’s moving
day night, day night.

In contrast, many activities involve linear unfolding, without recurrence. Such lin-
ear activities may involve some sort of culmination, as in the following when the
motion of light ends at the earth:
114 Y. J. Doran and J. R. Martin

FIGURE 5.6 Network for a dynamic perspective on field

The sun’s light hits earth

Alternatively there may be no indication of an end point:

The sun emits light

These options are pulled together into a network in Figure 5.6. This network says
that if a dynamic perspective is chosen, there are two sets of options available (the
curly bracket { indicates simultaneous choices): if dynamic (activity), then choose
from both momenting and activity type). Within momenting, an activity can be
unmomented (given as a single whole), or it can be divided on another tier into
multiple activities through the choice of momented. The superscript n indicates
that an activity may be momented an indefinite number of times. If an activity is
momented, then the series of activities which moment it may be related through
implication or expectancy. The activity type system indicates that each activity
may be either cyclical or linear, and if linear, may be either culminative or unending.
The dynamic activity and static item options offer alternative but complementary
perspectives on phenomena. Although in any particular field one perspective will tend
to be emphasized over the other, in principle all phenomena can be viewed in either
way. For example, the cardiovascular system within the biological sciences can be
viewed statically as a composition taxonomy of constituent items, including the heart,
lungs, veins and arteries; or it can be viewed dynamically as the circulation of blood
and transportation of nutrients, oxygen and the like to nourish, help fight disease and
stabilize temperature. To show these alternate perspectives, the network in Figure 5.7
brings together the options for both a dynamic and static perspective on field.
As far as explanations of seasons are concerned, we can use this network to show
how particular sets of activities and taxonomies form its basic building blocks. The
 Field relations 115

FIGURE 5.7 Dynamic and static perspectives on field

following passage from a year seven science teacher comes towards the end of a
number of lessons introducing the key components underpinning the seasons. Here,
the teacher is speaking over an animation of the earth orbiting the sun. The impor-
tant activities for explaining the seasons are in italics:

Now when the sun’s rays hit the earth, the earth has that also other theoretical
midline, the equator, that breaks it into half. Northern hemisphere and south-
ern hemisphere.
…
Ok here we go. The earth is spinning on its axis, happening every day. Every time it
spins, it’s moving day night, day night…. And then the earth is also moving around the
sun.You will notice at this point the northern hemisphere is closest to the sun,
but when we started the southern hemisphere was closest to the sun. This is
why we end up with opposite seasons. Because of that tilt in our axis puts us in
different positions relationship to the sun.

The activities relevant here are to do with the Earth’s rotation, both unmomented:

the earth is spinning;

it spins

and momented:

it’s moving day night, day night

116 Y. J. Doran and J. R. Martin

The passage also deals with the earth’s orbit around the sun:

the earth is also moving around the sun

And it notes the emission of light from the sun to the earth:

the sun’s rays hit the earth

These activities are crucial components for eventually showing that there are differ-
ent temperatures at different times of the year and thus that there are seasons. This
however does not capture the additional fact that different parts of the earth are
affected differently. For this, the teacher additionally partitions the earth into a com-
positional taxonomy of northern and southern hemispheres, divided by the equator:

the earth has that also other theoretical midline, the equator, that breaks it into half.
Northern hemisphere and southern hemisphere.

The activities of orbit, rotation and light emission, and the decomposition of earth
into northern and southern hemispheres provide most of the information needed to
explain the seasons. But the final sentence mentions one other crucial component
not covered by the activity and item options in Figure 5.7, namely the tilt of the
earth’s axis. This tilt is crucial for any thorough explanation, as it accounts for the
fact that different amounts of light hit different parts of the earth throughout the
year. Indeed the following explanation by a student, which emphasizes tilt, is very
positively evaluated by the teacher:

STUDENT The seasons are created by the earth’s 23.5 degree tilt.When the northern
hemisphere is tilted towards the sun it is summer. As the earth orbits the sun the
tilt stays the same, the side that’s tilted towards the sun changes, making it winter
in the northern hemisphere because it’s furthest away.
TEACHER Fantastic, I like that. That’s a good one. All right. Hopefully yours says
something similar to that.

Note that tilt is not an activity – it is not unfolding in any way and cannot be
momented. Nor is tilt a part of the earth (we cannot say, for example, the earth’s tilt is
a part of the earth); nor is it a type of the earth. In order to account for tilt, we need
to expand our model of field to introduce properties. To do this, we will take a step
away from seasons into other areas of science and then return to how this affects the
explanation of seasons considered above.

Property
In addition to activities and items, fields may be construed in terms of properties.
Properties, in broad terms, organize potentially gradable qualities or positions that
 Field relations 117

enable rich descriptions of phenomena. They often underpin distinctions between

items and activities and are vital components of fields in themselves.They may char-
acterize items, such as the earth is tilted or the negatively charged particle; or they may
qualify activities as in the electrons oscillate rapidly, or inflation is high.
If characterizing items, they can provide the criteria organizing taxonomies. For
example, in nuclear physics, neutrons and protons are different types of nucleon
that are distinguished primarily by their charge: protons are positively charged, while
neutrons are neutrally charged. Similarly, though different bands of electromagnetic
radiation such as visible light, ultraviolet light, x-rays and radio waves are all oscillat-
ing light, it is their wavelength that distinguishes them: visible light waves are around
390-700 nanometres long, while ultraviolet waves are between 10-390 nanometres long.
In a similar fashion, properties may be key features distinguishing moments in an
activity, for example, the activity of water freezing involves it getting colder, which
means its particles move more slowly, and eventually they lock together to become
ice.7
As properties can optionally occur for both activities and items, we can set up a
basic network as in Figure 5.8.
Properties can take many forms. They may involve qualitative descriptions
(Everest is tall, it is climbed slowly); or they may offer some spatio-temporal position
(Everest is in Asia, Everest’s current height, they trudged through the snow). In addition,
these properties may be graded and potentially ordered into arrays in relation to
other properties (Everest is the tallest mountain on earth, The Himalayas stretch from
China to Pakistan through Bhutan, Nepal and India); this in turn opens the way for
properties to be measured or quantified, what we will call gauged (Everest is 8,848 m
tall, Everest is 27.99° N, 86.93° E).
Adding to our network, Figure 5.9 indicates that properties may be arrayed or
not, and if arrayed, they may be gauged.

FIGURE 5.8 Basic parameters of field

118 Y. J. Doran and J. R. Martin

FIGURE 5.9 Network of property, arraying and gauging

Qualitative properties
We will explore properties through the following descriptive report from a field
guide to Australian birds (Menkhorst et al. 2017: 472). In this text, properties per-
meate the description in order to evocatively describe the flame robin’s physical
qualities including its colour, shape and size. The properties not only allow birders
to recognize flame robins, but also to distinguish between males and females, adults
and juveniles, and other robins such as the dusky robin (bold and italics in original).

Flame Robin Petroica phoenicea

Wing 73-83 mm Bill 13-15 mm Wt 11-15g
Largest, most slender-looking Petroica robin; long-bodied appearance accentuated
by smallish head and longish wings. Ad ♂ distinctive; note slaty upperparts; small
white forehead spot; underparts bright orange-red from chin to belly. Ad ♀ differ from
Scarlet ♀ by sandier brown upperparts, small white forehead spot (sometimes tinged
buff or wholly absent) and light grey-brown breast grading to whitish chin and belly.
Some have diagnostic, but inconspicuous, traces of orange-yellow to orange on
breast and belly.White eye-ring and bold wing-bars rule out Dusky Robin in Tas.
Juv like Juv Scarlet with slightly finer white streaks on upperparts.

Voice: Contact call, a single note tlip, is sweeter than Scarlett Robin and sel-
dom given on the non-breeding grounds. Musical warbling song, more com-
plex and piping than other Petroica robins, often consists of 3 sets of 3 notes ‘you
may come, if you wish, to the sea’.
Notes: Breeds mainly in upland eucalypt forests and woodlands, especially
with open understory or small clearings; readily, but temporarily, colonizes
cleared or burnt areas. Most leave the high country in autumn, wintering in
more open habitats in lowlands including grasslands, farmland, and open forests
and woodlands with grassy cover. Mostly seen singly or in pairs during the
breeding season. Often seen in loose groups of up to 20 birds at other times, the
only Petroica to form flocks.

We will focus initially on the first paragraph. This paragraph is primarily concerned
with the physical appearance of the flame robin. As such it describes its properties
such as colour, shape and size, and those of its various body parts (qualitative proper-
ties underlined):
 Field relations 119

Largest, most slender-looking Petroica robin

long-bodied appearance
smallish head
longish wings
slaty upperparts
small white forehead spot
underparts bright orange-red from chin to belly
sandier brown upperparts
sometimes tinged buff or wholly absent
light grey-brown breast
whitish chin and belly
traces of orange-yellow to orange on breast and belly
White eye-ring
bold wing-bars
slightly finer white streaks on upperparts

These precise verbal descriptions of colour (e.g. slaty, whitish, sandier brown), size
(largest, long-bodied, longish) and shape (most slender-looking) are often supplemented
by pictorial representations such as in Figure 5.10, to make it easier to recognize the
bird in the wild (the original image is in colour, showing among other things, the
distinct orange breast and belly of the adult male).
As this text shows, properties like colour, size and shape can be graded through
arraying. In terms of field, arraying like this establishes degrees of a property in com-
parison to other more or less explicitly specified instances. For example, the size and
shape of the flame robin is arrayed as:

Largest, most slender-looking Petroica robin

FIGURE 5.10 Picture of the flame robin showing its various properties. Reproduced
from Menkhorst et al. (2017: 473) with permission from CSIRO Publishing. Original
colour illustrations by Peter Marsack
120 Y. J. Doran and J. R. Martin

Similarly, many of the colours are graded in terms of their brightness, or other
qualities:

underparts bright orange-red from chin to belly

sandier brown
light grey-brown breast

If more precision is required, arrays can be gauged. The sub-heading shows this by
gauging the length of the flame robin’s wing and bill in terms of millimetres (mm)
and its weight in terms of grams:

Wing 73-83 mm Bill 13-15 mm Wt 11-15g

As students move through schooling, precise measurements of phenomena become

more strongly emphasized. This is particularly the case for sciences such as physics
where these measurements are regularly presented in graphs or long mathematical
texts called quantifications (Lemke 1998; Parodi 2012; Doran 2018, Chapter 7 of
this volume). The graph in Figure 5.11 is from a senior high-school physics exam.
This graph arrays two properties of a wire on the vertical and horizontal axes,
resistance and temperature, and gauges them in ohms (Ω) and degrees Celsius (°C)
(resistance and temperature are in fact itemized properties, discussed below). A series of
points (marked by X) and a trend line have been plotted by a student, giving specific
measurements for both its resistance and temperature.
In a sense, properties, which may be arrayed and gauged, establish an ideational
perspective on gradable meanings that have generally been explored through the
interpersonal system of graduation within appraisal (Hood and Martin 2005, Martin
and White 2005).8 By putting forward the property network, we are suggesting that

FIGURE 5.11 Graph of resistance and temperature, showing two gauged properties
 Field relations 121

the instances of grading shown above are not necessarily geared towards making
evaluative meanings but are simply organizing the field of the petroica robin. It is
likely that there will be some indeterminacy in this area, since properties supply an
arena for attitudinal meanings invoked through graduation; indeed if we look at the
second paragraph of the flame robin descriptive report, we see a handful of proper-
ties that could be read as invoked appreciations of the flame robin’s call (graduated
attitude, either inscribed or invoked, combined with arrayed properties underlined):

Voice: Contact call, a single note tlip, is sweeter than Scarlett Robin and sel-
dom given on the non-breeding grounds. Musical warbling song, more com-
plex and piping than other Petroica robins, often consists of 3 sets of 3 notes ‘you
may come, if you wish, to the sea’.

Spatio-temporal properties
A second type of property is concerned not with some quality of an item or activ-
ity, but rather with an item or activity’s location in either space or time – termed
spatio-temporal properties.
In our flame robin text, spatio-temporal properties are used extensively when detail-
ing the flame robin’s habitat and the time of year they can be observed (underlined):

Notes: Breeds mainly in upland eucalypt forests and woodlands, especially

with open understory or small clearings; readily, but temporarily, colonizes
cleared or burnt areas. Most leave the high country in autumn, wintering in
more open habitats in lowlands including grasslands, farmland, and open forests
and woodlands with grassy cover. Mostly seen singly or in pairs during the
breeding season. Often seen in loose groups of up to 20 birds at other times, the
only Petroica to form flocks.

Like qualitative properties, spatio-temporal properties can occur for both activi-
ties and items. Examples of spatio-temporal properties of activities from the above
excerpt include:

Breeds mainly in upland eucalypt forests and woodlands

Most leave the high country in autumn
wintering in more open habitats
Mostly seen singly or in pairs during the breeding season
Often seen in loose groups of up to 20 birds at other times

Examples of spatio-temporal properties of an item from earlier in the text, include:

slightly finer white streaks on upperparts9

Some have diagnostic, but inconspicuous, traces of orange-yellow to orange on breast and
belly
122 Y. J. Doran and J. R. Martin

FIGURE 5.12 Map showing spatio-temporal property of the flame robin in Australia.
Reproduced from Menkhorst et al. (2017: 472) with permission from CSIRO Publishing

Just like qualitative properties such as colour, size and shape, spatio-temporal prop-
erties can be arrayed by ordering their spatial or temporal positions in some way. In
the initial description the underparts were described as:

bright orange-red from chin to belly.

Here, the chin and belly are arrayed as outerpoints that the bright-orange red colour
occurs between. Such arraying of spatio-temporal properties is often more easily
shown through diagrams and maps, such as in Figure 5.12 showing the areas of
Australia the flame robin can be observed.
Finally, like all properties, spatio-temporal properties can be gauged. An example
of this is the use of latitude and longitude, such as 34.0386° S, 151.1407° E. Or
more commonly in everyday discourse, this is regularly seen when giving the time:

Be there 9 o’clock.

From the description so far, we can flesh out our network of properties as Figure 5.13.

FIGURE 5.13 Network of property

 Field relations 123

Complex fields
Activity, taxonomy and property systems provide the basic resources for constru-
ing field. But before we can return to mapping the explanation of seasons this
chapter began with, there is one more system we need to introduce. This system
makes room for interdependent variables that organize the complex constellations
of meaning in technical discourses such as those of science.There are two basic ways
of building interdependency: reconstruing and interrelating.

Reconstruing variables
Beginning with the reconstrual of field variables, we can return to the tilt of the earth
that we mentioned is crucial for explaining seasons. In one sense, we can now readily
account for this in our expanded model – tilt is a property of the earth and so can be
arrayed and gauged. Indeed, our original text explaining the seasons does just that:

Seasons happen because Earth’s axis is tilted at an angle of about 23.4 degrees

In this form, where the Earth’s axis is specified as being tilted, this analysis is
unproblematic. However, in the next sentence, the property realized verbally as tilted
above is nominalized as tilt:

Earth’s axial tilt

In addition, the nominalized tilt is classified by axial. This positions tilt in a clas-
sification taxonomy; axial tilt is a type of tilt. This is problematic for the outline of
field resources canvassed thus far as classification is a key feature of items, not prop-
erties. Moreover, in terms of Hao’s model of ideational discourse semantics (2020a,
Chapter 6 of this volume), the nominalized axial tilt is not a quality, which is the
typical realization of property. Rather, it is a measured dimension of the earth that
can be reorganized lexicogrammatically as a Focus^Thing structure:

Axial tilt of Earth

To reconcile these two seemingly conflicting analyses – tilt as a property and tilt
as an item – what we will suggest here is that the property tilted is being reconstrued
as an item tilt. This we will refer to as an itemized property. This analysis is based on
the fact that axial tilt shows many of features of both properties and items: it can be
arrayed and gauged like properties and can be taxonomized like items. And from the
perspective of discourse semantics, it is not realized by a quality, which is the pro-
totypical realization of qualitative properties, nor by an entity, which is the standard
realization of items, but rather by a dimension of an entity.
Such itemized reconstruals are regularly used to ‘name’ broader sets of properties.
Indeed, we have used a number of these throughout this paper to group together
various properties. In the following examples, the underlined itemized properties
are ‘names’ of the bolded properties:
124 Y. J. Doran and J. R. Martin

The colours (of the flame robin) are slaty, whitish and sandier brown.
The shape (of the flame robin) is described as slender-looking.

Itemized properties comparable to these are the basis for symbolic variables in
mathematics (see Doran, Chapter 7 of this volume). For example, in the following
formula the acceleration of a moving body is symbolized as a, the mass of that body is
m and the force on the body is F:

F = ma
Like all properties, each of these can be numerically gauged – in this case through
Newtons (N), metres per second squared (m/s2) and kilograms (kg):

F = 4N
a = 2 m / s2
m = 2 kg

But like items, they can all be taxonomized:

angular acceleration, linear acceleration;

centripetal force, electrostatic force;
relativistic mass, rest mass.

Reconstruals of properties as items is a regular feature of many fields. Under this

interpretation, field is a resource not just for construing items, activities and proper-
ties, it is also a resource for reconstruing meanings. It enables multiple overlapping
perspectives on phenomena to be realized in a single instance. This is emphasized
by the fact that in addition to being itemized, properties can also be reconstrued as
activities. Similarly, items can be dynamized by being reconstrued as activities, and
activities can in turn be itemized. Each of these options have typical realizations in
discourse semantics, as introduced below.
When properties are reconstrued as activities, such activated properties enable a
dynamic unfolding of a property:
It gets hotter

In terms of property, this indicates an array of temperature (degrees of heat). But

in terms of activity, it can be used to moment a larger activity:

It gets hotter
^
And then eventually it melts.

This type of activation moves us into the realm of ‘becoming’, where properties are,
in effect, dynamized.
Activating items very commonly involves them being positioned in a taxonomy.
The following excerpt from a university physics textbook shows an example of this
for composition (Young and Freedman 2012: 742):
 Field relations 125

… now we let the ball touch the inner wall… The surface of the ball becomes part of the
cavity surface.

Here, the part whole relation between the surface of a ball and the cavity surface of
a wall is activated, and becomes a moment in a larger activity:

We let the ball touch the inner wall

^
The surface of the ball becomes part of the cavity surface.

Finally, just as items can be activated, activities can be itemized.10 This regularly
happens in the process of technicalization, as scientific terms are distilled as activity
entities (see Hao 2020a and Chapter 6 of this volume for discussion of the discourse
semantics of activity entities). One of our initial examples of a momented activity in
fact showed a series of itemized activities that moment phagocytosis – itself an item-
ized activity; the itemized activities involved are underlined below:

Detection
^
Ingestion
^
Phagosome forms
^
Fusion with lysosome
^
Digestion
^
Discharge

Like all items, these itemized activities can enter into a taxonomy. For example,
phagocytosis is one type of endocytosis, along with potocytosis and micropinocytosis, which
all contrast with exocytosis.
Table 5.1 shows various reconstruals and some typical discourse semantic realiza-
tions (from Hao 2020a). Note here that the order in which the reconstrual takes
place is significant – an itemized activity is different from an activated item.

TABLE 5.1 Field reconstruals

Field reconstruals Typical discourse semantic realization Example

itemized property measured or perceived dimension The colour of skin
occurrence figure It heats up
activated property
state figure It gets hotter
itemized activity activity entity Phagocytosis
activated item state figure You become part of the team
126 Y. J. Doran and J. R. Martin

Interrelating fields and the explanation of seasons

We are now in a position to characterize each of the main elements of our initial
explanation of the seasons, replayed here:

What Causes Seasons on Earth?

Seasons happen because Earth’s axis is tilted at an angle of about 23.4 degrees and
different parts of Earth receive more solar energy than others.
Because of Earth’s axial tilt (obliquity), our planet orbits the Sun on a slant
which means different areas of Earth point toward or away from the Sun at dif-
ferent times of the year.
Around the June solstice, the North Pole is tilted toward the Sun and the
Northern Hemisphere gets more of the Sun’s direct rays. This is why June, July
and August are summer months in the Northern Hemisphere.

Opposite Seasons
At the same time, the Southern Hemisphere points away from the Sun, creating
winter during the months of June, July and August. Summer in the Southern
Hemisphere is in December, January and February, when the South Pole is tilted
toward the Sun and the Northern Hemisphere is tilted away.

For our discussion here, the main field components needed for this explanation are:

Earth’s 23.4 degree axial tilt

Division of earth into northern and southern hemispheres
Earth’s orbit of the sun
Earth receipt of solar energy
Different parts of the earth receive more solar energy than others at different times of the year
Seasons of summer and winter

Each of these components is classified with reference to field resources in Table 5.2.
For ease of reference, each of the components is named in small caps (tilt, hemi-
spheres, etc.).
The first thing to note is that far more than a common sense understanding of
the seasons is involved. For a full explanation, students need to attend to both the
classification and composition of items, two types of activity and both gauged and
arrayed properties. In addition, they need to conceptualize three itemized proper-
ties, and a set of items ordered into two separate arrays. This is already a conceptual
challenge for most junior secondary school students.
But there is more going on. As set out in Table 5.2, more than one factor is
needed for the explanation, however we have not yet specified the relations among
 Field relations 127

TABLE 5.2 Field relations in an explanation of the seasons

Factors in explanation of seasons Field relation

Tilt
Earth’s 23.4 degree axial tilt classified itemized gauged property
Hemispheres
Division of earth into northern and southern composition taxonomy of items
hemispheres
Orbit
Earth’s orbit of the sun itemized cyclical activity
Receipt of solar energy
Earth receives solar energy culminative activity
Variation in solar energy
Different parts of the earth receive more composition\ taxonomy (different parts of the
solar energy than others at different times earth)
of the year ordered into an array of an itemized qualitative
property (more solar energy)
ordered into another array of an itemized spatio-
temporal property (different times of the year)
Seasons
Seasons of summer and winter classification taxonomy of items

them. In order to explain seasons, specific types of logical interdependency have

to be established. To model this, we will introduce the second way of linking field
variables – interrelating (complementing reconstruing as introduced above).
Interrelating is concerned with how different elements of field are associated
with each other. In broad terms, there are two means of interrelating elements: they
can be positioned as relatively independent of one another or they can be posi-
tioned as in some sense dependent on each other (although not precisely the same,
this closely relates to the LCT distinction between dependent and independent
links in constellations, introduced in Maton and Doran, Chapter 3 of this volume).
Let’s focus first on the relatively independent relation. In the seasons explanation,
the tilt, hemispheres, orbit and receipt of solar energy are not dependent on
each other in any way. One can vary or not exist at all without affecting the others.
The fact that the earth is tilted, for example, has no bearing on the fact that the sun
emits solar energy that hits the earth. Similarly, the fact that there are hemispheres
on earth has no impact on whether it orbits the sun.
To describe this relation, we can borrow from Halliday and Matthiessen’s
(2014) logico-semantic relations for English clause complexing and call this exten-
sion (signified by a +). Here we are analogizing from an ‘and’ relation, where
multiple elements are coordinated but are not ordered in any way (ideationally
speaking). The extending factors in this relationship can be usefully laid out in
parallel, as follows:
128 Y. J. Doran and J. R. Martin

TILT + HEMISPHERES + ORBIT + RECEIPT OF SOLAR ENERGY

itemized composition of cyclical culminative activity
property items activity

In contrast, some elements of field are dependent on others. For example, the
variation in solar energy factor is the result of the combination of the tilt, orbit,
hemispheres and receipt of solar energy factors. Without each of these factors,
the precise form of the variation in solar energy would not occur. This in part
accounts for the complexity of seasons explanations. variation in solar energy
takes the compositional taxonomy from the hemispheres, orders it into an array of
solar energy involved in the receipt of solar energy, reinterprets the orbit of the
sun as an array of time through the year and, through a couple of unspecified steps
resulting from the tilt, additionally arrays the variation in solar energy at different
parts of the earth according to the time of the year. Analogizing again from Halliday
and Matthiessen’s logico-semantic relations, we will call this dependency relation
enhancing (signified by an x).
Finally, the seasons reorganizes the variation in solar energy from an array
of temperatures to a classification taxonomy of items – summer, winter, autumn,
spring. This in effect names components of these variations and distils its mean-
ing into the technical term ‘seasons’. Following our analogy with Halliday and
Matthiessen’s logico-semantic relations, we will call this naming relation elaboration
(signified by =).
Pulling these interrelation types together with our previous description, we
can visualize the field variables underpinning the explanation of the seasons as
follows (with enhancing factors laid out vertically in relation to what they depend
upon):

TILT + HEMISPHERES + ORBIT + RECEIPT OF SOLAR ENERGY

itemized composition of cyclical culminative activity
property items activity
×
VARIATION IN SOLAR ENERGY
compositional taxonomy on an arrayed itemized property on another arrayed itemized
property
=
SEASONS
classification taxonomy of items

This outlines how the relatively independent components of the tilt, hemi-
spheres, orbit and receipt of solar energy together produce the variation in
solar energy, which in turn produces the seasons. The mapping effectively dis-
plays the complexity underpinning a scientific explanation of a phenomenon we all
experience in everyday terms.
 Field relations 129

FIGURE 5.14 Network for interdependency

As far as our broader model of field is concerned, this completes the description.The
options for building interdependency in fields produces the network in Figure 5.14.
Pulling this together with the rest of the field network gives the basic systems of
field shown in Figure 5.15.This network indicates that construals of field can adopt
either a dynamic or a static perspective, can be optionally propertied and can be
related to other aspects of a field.

Making uncommon sense

From the perspective of educational linguistics, our model of field as a resource for
construing phenomena reveals the complexity of science fields at even the lower
levels of secondary school. The explanation that we focused on for this paper is a
relatively simple one in the broader scheme of things. It did not take into account

FIGURE 5.15 Network of field

130 Y. J. Doran and J. R. Martin

the effect of variation in the length of the day arising from the rotation of the
earth; nor did it go into detail about how the earth’s tilt actually produces variation
in light intensity in different part of the earth at different times of the year; nor did
it conceptualize the hemispheres as a continuous array of latitude (affecting how
‘summery’ or ‘wintery’ it is depending on how far you are from the poles); nor did
it consider this explanation in relation to other meteorological and astronomical
phenomena. And it definitely did not consider how this conception of seasons
is related to seasonal change in parts of the world where a wet-dry distinction is
more relevant (and in doing so can be critiqued from a post-colonial perspective).
As scientific knowledge develops through schooling, more and more of these addi-
tional elements are interrelated and then presumed. For an educational program
that aims to develop a discipline-specific pedagogy based on the varying ways
knowledge is built, managing this complexity such as through a model of this kind
is crucial.
As far as functional linguistics is concerned, our model of field builds on Hao’s
modelling of ideational discourse semantics (2020a) and opens up the possibility of
explicitly differentiating and linking ideational meaning resources across strata (reg-
ister, discourse semantics and lexicogrammar – specifically, field realized through
ideation and connexion, and ideation and connexion realized through clause com-
plexing, transitivity and nominal group structure). It has done so in a way that is
proving productive not just for language, but for a range of semiotic resources used
in science (see Doran, Chapter 7 of this volume). This places functional linguistics
and semiotics in a far stronger position to manage the distinctive complexity of
knowledge building across disciplines, and within disciplines across modalities of
communication. Making sense of uncommon sense depends on robust modelling
of this kind.

Notes

1 https://www.timeanddate.com/astronomy/seasons-causes.html
2 A little closer to everyday life, following Barthes (1977: 101), we could alternatively
moment the activity having a drink, into the series ordering a drink, obtaining it, paying for it,
drinking it.
3 Previously, series of activities were termed activity sequences. However in light of Hao’s
work on ideational discourse semantics (2015, 2018, 2020a), series will be used here for
strings of activities in field and sequences will be reserved for strings of figures in discourse
semantics.
4 This and a number of other examples in this chapter come from a study examining class-
room practice in secondary schools in New South Wales, Australia, led by Karl Maton,
J. R. Martin, Len Unsworth and Sarah Howard and funded by the Australian Research
Council (DP130100481).
5 We will again follow Hao (2018, 2020a) here in using connexion for the discourse
semantic relation between figures, previously known as conjunction in Martin (1992)
and Martin and Rose (2007).This is in order to distinguish the discourse semantic system
from the word-class conjunction within lexicogrammar.
 Field relations 131

6 Here tides is an activity entity that is caused by the previous occurrence figures. However,
under Hao’s model (2020a: 94), it is in fact part of a presented state figure within the
larger sequence, alternatively realised as The gravitational pull of the sun and the moon and the
rotation of the earth causes (the forming of) tides (Hao personal communication). Hao argues
that sequences such as this are one means through which activity entities like tides can be
presented in scientific texts.
7 Thanks to Dragana Stosic, Sally Humphrey and Jing Hao for this example.This raises the
broader point that all perspectives on field (activity, taxonomy, property) can be used as
criteria for distinguishing and defining all other perspectives. For any particular field, its
interlocked networks of field relations are such that its technical meanings will be mutu-
ally defining.
8 A complementary ideational perspective on gradable meanings from the perspective of
discourse semantics is given by Hao (2015, 2020a) and Hood and Hao (chapter 10, this
volume), in terms of qualities of entities.
9 This is not saying that the upperparts are white (i.e. it is not attributing a qualitative prop-
erty to the upperparts). Rather is locating the white streaks as being on the upperparts.
10 Note that we do not need to account for propertied items or propertied activities in this sec-
tion as this is already accounted for in our initial system in Figure 5.8, where either an
activity or an item can take a property.

References

Barthes, R. (1977) Image, Music,Text, London: Fontana.

Doran, Y. J. (2017) ‘The role of mathematics in physics: Building knowledge and describing
the empirical world’, Onomázein Special Issue, March: 209–26.
Doran, Y. J. (2018) The Discourse of Physics: Building Knowledge through Language, Mathematics
and Image, London: Routledge.
Doran, Y. J. (2019) ‘Building knowledge through image in physics’, Visual Communication,
18(2): 251–77.
Doran, Y. J. (2020) ‘Cultivating values: Knower-building in the humanities’, Estudios de
Lingüistica Aplicada, 37(70): 169–98..
Greenwood, T. and Allen, R. (2004) Year 12 Biology 2004: Student resource and activity manual,
Hamilton: Biozone.
Halliday, M. A. K. (2004) The Language of Science: Volume 4 in the Collected Works of M. A. K.
Halliday, London: Continuum.
Halliday, M. A. K. and Martin, J. R. (1993) Writing Science: Literacy and Discursive Power,
London: Falmer.
Halliday, M. A. K. and Matthiessen, C. M. I. M. (2014) Halliday’s Introduction to Functional
Grammar, London: Routledge.
Hao, J. (2015) ‘Construing Biology: An Ideational Perspective’, Unpublished PhD thesis,
Department of Linguistics, University of Sydney.
Hao, J. (2018) Reconsidering “cause inside the clause” in scientific discourse – from a dis-
course semantic perspective in systemic functional linguistics. Text and Talk, 38(5): 520–50.
Hao, J. (2020a) Analysing Scientific Discourse from a Systemic Functional Perspective: A Framework
for Exploring Knowledge Building in Biology, London: Routledge.
Hao, J. (2020b) ‘Construing relations between scientific activities through Mandarin Chinese’,
in J. R. Martin,Y. J. Doran and G. Figueredo (eds) Systemic Functional Language Description,
London: Routledge, 238–72.
132 Y. J. Doran and J. R. Martin

Hao, J. (2020c) ‘Nominalisation in scientific English: a tristratal perspective’, Functions of

Language, 27(2): 143–73.
He, Y. (2020) ‘Animation as a Semiotic Mode: Construing Knowledge in Science Animated Videos’,
unpublished PhD thesis, University of Sydney.
Hood, S. (2010) Appraising Research: Evaluation in Academic Writing, New York: Palgrave
Macmillan.
Hood, S. and Martin, J. R. (2005) ‘Invoking attitude: The play of graduation in appraising
discourse’, Revista Signos, 38(58): 195–220.
Lemke, J. L. (1990) Talking Science: Language, Learning and Values, Norwood, NJ: Ablex.
Lemke, J. L. (1998) ‘Multiplying meaning: Visual and verbal semiotics in scientific text’, in
J. R. Martin and R.Veel (eds) Reading Science, London: Routledge, 87–113.
Martin, J. R. (1992) English Text: System and Structure, Amsterdam: John Benjamins.
Martin, J. R. (2011) ‘Bridging troubled waters: Interdisciplinarity and what makes it stick’, in
F. Christie and K. Maton (eds) Disciplinarity, London: Routledge, 35–61.
Martin, J. R. (2020) ‘Revisiting field: Specialized knowledge in secondary school science and
humanities discourse’, in J. R. Martin, K. Maton and Y. J. Doran (eds) Accessing Academic
Discourse, London: Routledge, 114–47.
Martin, J. R. and Maton, K. (2013) (eds) ‘Cumulative knowledge-building in secondary
schooling’, Special Issue of Linguistics and Education, 24(1): 1–74.
Martin, J. R. and Rose, D. (2007) Working with Discourse: Meaning beyond the Clause, London:
Continuum.
Martin, J. R. and Rose, D. (2008) Genre Relations: Mapping Culture, London: Equinox.
Martin, J. R. and Veel, R. (eds) (1998) Reading Science: Critical and Functional Perspectives on the
Discourse of Science, London: Routledge.
Martin, J. R. and White, P. R. R. (2005) The Language of Evaluation: Appraisal in English,
Basingstoke: Palgrave Macmillan.
Martin, J. R., Maton, K. and Doran,Y. J. (2020) Accessing Academic Discourse: Systemic functional
linguistics and Legitimation Code Theory, London: Routledge.
Martin, J. R., Maton, K. and Matruglio, E. (2010) ‘Historical cosmologies: Epistemology and
axiology in Australian secondary school history discourse’, Revista Signos, 43(74): 433–63.
Martin, J. R., Maton, K. and Quiroz, B. (eds) (2017) ‘Systemic functional linguistics and
Legitimation Code Theory on education and knowledge’, Onomázein Special Issue, March:
1–242.
Martin, J. R., Unsworth, L. and Rose, D. (in press) ‘Condensing meaning: Imagic aggregations
in secondary school science’, in G. Parodi (ed) Multimodality, London: Bloomsbury.
Maton, K. (2014) Knowledge and Knowers: Towards a Realist Sociology of Education, London:
Routledge.
Maton, K. and Doran,Y. J. (2017) ‘SFL and code theory’, in T. Bartlett and G. O’Grady (eds)
The Routledge Handbook of Systemic Functional Linguistics, London: Routledge, 605–18.
Maton, K., Martin, J. R. and Matruglio, E. (2016) ‘LCT and systemic functional linguistics:
Enacting complementary theories for explanatory power’, in K. Maton, S. Hood and
S. Shay (eds) Knowledge-Building, London: Routledge, 93–113.
Menkhorst, P., Rogers, D., Clarke, R., Davies, J., Marsack, P. and Franklin, K. (2017) The
Australian Bird Guide, Clayton South: CSIRO Publishing.
Parodi, G. (2012) ‘University genres and multisemiotic features: Accessing specialized knowl-
edge through disciplinarity’, Fórum Linguístico, 9(4) 259–82.
 Field relations 133

Unsworth, L. (2020) ‘Intermodal relations, mass and presence in school science explana-
tion genres’, in M. Zappavigna and S. Dreyfus (eds) Discourses of Hope and Reconciliation,
London: Bloomsbury, 131–52.
Wignell, P., Martin, J. R. and Eggins, S. (1989) The discourse of geography: Ordering and
explaining the experiential world, Linguistics and Education, 1(4): 359–91.
Young, H. D. and Freedman, R. A. (2012) Sear’s and Zemansky’s University Physics with Modern
Physics, 13th ed., San Francisco: Addison Wesley.
6
BUILDING TAXONOMIES
A discourse semantic model of entities
and dimensions in biology

Jing Hao

Introduction
Studying science at the undergraduate level represents a critical stage of appren-
ticeship into scientific disciplines. Undergraduate students who are training in the
biological sciences learn to observe biological phenomena, conduct experiments,
engage with academic literature, record observations and provide reasoning largely
through written texts. Through this work students make a transition from knowl-
edge ‘reproduction’ to knowledge ‘production’ (Bernstein 2000), including a shift
from building knowledge that is recontextualized in pedagogic texts to engaging
with knowledge that is published in peer-reviewed research articles. In Australian
universities, students who have been successfully apprenticed into undergraduate
course work can choose a pathway towards postgraduate research. This chapter
explores one aspect of knowledge development during this transition, that of build-
ing taxonomies.
As introduced in Doran, Maton and Martin (Chapter 1, this volume), under-
standing scientific knowledge building has been a longstanding object of study in
both Systemic Functional Linguistics (SFL) and Legitimation Code Theory (LCT).
SFL conceptualizes scientific taxonomies by and large from an ideational perspec-
tive, in terms of ideational meanings construing the register variable field. Scientific
taxonomies are characterized as a key dimension of ‘technicality’ in science (Martin
1992; Halliday and Martin 1993). Technical taxonomies are distinguished from
those used in other fields, such as everyday life and recreational activities. Their dif-
ferences are determined largely by different recognition criteria in different fields.
Technical taxonomies (e.g., cells, pathogen, bacteria) are identified based on crite-
ria that are typically learned in the institutional settings, such as through scientific
measurements.Taxonomies in the recreational fields are identified based on special-
ized usage and functionality (e.g., shuttlecock, screwdriver, piano); and in everyday
 Building taxonomies 135

life taxonomies are identified based on our daily and domestic experience (e.g.,
shoes, toothbrush, bags). The differences of fields recognized in SFL resonates with
Bernstein’s (2000) sociological distinction between ‘horizontal discourse’ and ‘verti-
cal discourse’ (i.e. common sense vs. uncommon sense), and in vertical discourse
between ‘horizontal knowledge structures’ (e.g., the humanities) and ‘hierarchical
knowledge structures’ (e.g., sciences). Both SFL’s field typology and Bernstein’s con-
cepts provide useful ways to think about the nature of different disciplines. However,
when it comes to analyzing knowledge structures in practice, neither of them pro-
vide sufficient analytical tools. As Maton (2014) argues, Bernstein’s concepts of dis-
courses and knowledge structures are ‘good to think with but less useful to analyze
with’ (Martin and Maton 2017: 28).
LCT extends Bernstein’s concepts and offers a rigorous multi-dimensional con-
ceptual toolkit for analyzing disciplinary knowledge building (Maton 2014, Maton
et al. 2016). Building scientific taxonomies, for instance, can be analyzed through
the theoretical dimensions of both Specialization and Semantics (Maton 2020,
Maton and Chen 2020). With respect to Specialization, biological science can be
seen as a knowledge code, emphasizing stronger epistemic relations between scientific
claims and the biological phenomena under study, and sfvweaker social relations
between scientific claims and biologists themselves. In terms of Semantics, semantic
density, particularly its form of epistemological condensation, offers a way of concep-
tualizing the degree to which meanings are condensed in building taxonomies.
Meanings at stake include the role of taxonomies in biological activities (Martin
and Maton 2013).
This study is an attempt to make knowledge structure visible from the per-
spective of SFL. A key question is how to reveal disciplinary knowledge (i.e. field)
by analyzing language patterns in texts. The language features concerned are par-
ticularly associated with analyses of unfolding meaning across texts – a discourse
semantics perspective, which addresses meaning made in and beyond the clause.
Descriptions of one discourse semantic unit, known as entity, have been used to
identify the language of taxonomies in field (Martin and Rose 2007; Martin 1997).
However, the classification of entities has by and large drawn on distinctions in the
register variable field (Martin 1992). This limits the usefulness of discourse seman-
tics as a tool for examining the breadth and depth of taxonomies.
To address both theoretical and descriptive challenges, this chapter reports on
further development of an SFL framework for analyzing scientific taxonomies. I
first introduce the theoretical principle known as ‘trinocularity’, whereby language
choices are viewed from three simultaneous perspectives (Matthiessen and Halliday
2009; Martin 2013). I then draw on the trinocular principle to describe a range of
choices in the system network of entities, including the systems of entity type and
dimensionality. Note that since the categories have emerged from undergraduate
biology texts (including pedagogic materials and students’ experimental reports)
(Hao 2015), the description does not aim to provide generalized categories that can
be used for all scientific discourse. However, the method through which the choices
are identified is appliable and can be generalized across studies. After introducing
136 Jing Hao

the systems, I illustrate how the framework outlined can be used for analyzing
texts. An excerpt from a high-graded student research report produced at the final
undergraduate year is selected for the illustrative analysis. The analysis reveals both
the diversity and depth of taxonomies that are developed at the end of undergradu-
ate study.

Foundations
Theoretical principles of language description
A critical understanding of language from an SFL perspective is that language is a
primary means for construing knowledge (together with other semiotic resources
such as images and body language, see Hood and Hao this volume). ‘Construe’
means that knowledge is not only expressed by language (and other semiotics), but
its organization is also influenced by how language is used (Halliday 2004 [1998]).
A linguistic understanding of scientific taxonomies involves recognizing language
resources that construe taxonomies.
SFL theorizes the language of disciplinary knowledge from a multi-stratal and
multi-functional perspective. The theoretical dimensions of stratification and meta-
function are modelled visually in Figure 6.1.
Stratification conceptualizes the inherent relationship between language and con-
text, by recognizing patterns of meaning across different levels of abstraction known
as strata.The strata within language include phonological/graphological systems
organizing sound/script; lexicogrammatical systems organizing meanings in a
clause; and discourse semantic systems organizing meaning unfolding in a text.
The strata are simultaneously organized from the perspective of three metafunctions,

FIGURE 6.1 Stratification and metafunctions in SFL theory

 Building taxonomies 137

which relate naturally to three contextual register variables – ideational meanings

construing what is going on in a field; interpersonal meanings enacting social
relations (tenor); and textual meanings composing mode (e.g., spoken vs. writ-
ten discourse). Interactions among field, tenor and mode enable the performance
of a range of staged and goal-oriented social purposes in our culture – i.e. genres
(Martin and Rose 2008).
Critical to understanding both stratification and metafunction is the idea that
choices in language are not made in isolation but are based on: 1) meaning that is
construed at a higher stratum, 2) meaning at a lower stratum that is a realization
of the meaning at a given stratum, and 3) meanings in the other metafunctions at
a given stratum. This simultaneous consideration of meaning-making from ‘above’,
‘below’ and ‘around’ is in essence a tri-stratal perspective. Specific concerns are
determined by specific descriptive tasks. If our standpoint is ideational discourse
semantics, of particular concern are meanings construed at the level of field (look-
ing from ‘above’), their realizations through experiential grammar (looking from
‘below’), and the interactions of ideational with interpersonal and textual meanings
in the discourse (looking from ‘around’). For further work adopting a tri-stratal
perspective, see (Hao 2020a) on ‘nominalization’ and (Hao 2018) on ‘causality’.
In what follows, I review previous work on scientific taxonomies. I argue that
previous descriptions were by and large based on field and grammar – reasoning
from above and below – but not from around. There is accordingly a need to bring
discourse semantic resources into the picture.

Previous SFL description of scientific taxonomies

Taxonomies at the level of field
From the perspective of the register variable field, Martin (1992) models taxonomy
as one aspect of field organization, along with activity sequences the taxonomies
enter into. Distinctive configurations of activity sequences and taxonomies allow
several field types to be identified, as shown in Table 6.1.
Science is seen as an example of a ‘technical’ field (Martin 2017), character-
ized by ‘technical’ taxonomies and ‘implication sequences’. The term ‘implication’

TABLE 6.1 Field types in relation to activity sequences and taxonomies (c.f. Martin 1992:
545, Martin 2017)

activity sequences taxonomies

domestic (guidance) implicit ‘natural’
specialized (participation) manuals utilitarian (tools)
e.g. sport, hobby, trades
administration (cooperation) procedures pragmatic (subjects)
exploration / technical (instruction) implication technical (things)
e.g. humanities, social sciences and science sequence
138 Jing Hao

emphasizes that in scientific explanations one activity is determined by what has

gone before. The term ‘technical’ emphasizes that scientific taxonomies have dis-
tinctive criteria for classification (superordinate and subtypes) and composition
(parts and whole) – different from those used in everyday and specialized fields.
Wignell et al. (1993) exemplify this difference by comparing taxonomies of birds of
prey in science with those recognized in the specialized field of bird watching. They
show that birdwatchers’ taxonomy relies largely on observable physical character-
istics (e.g., colour and tail-shape for Black Kite vs. Square-tailed Kite in Figure 6.2),
whereas scientific taxonomies for birds of prey considers their differences with
respect to chromosomes and genes. Scientific taxonomies also often involve sup-
plementing vernacular names with binomial Latin terminology (e.g., Acciptridae,
Milvus, Lophoictinia in Figure 6.3).
Complementing Martin’s (1992) field typology, Hood (2010) recognizes that
in academic texts, specifically research articles, two fields can be identified based
on their distinctive taxonomies and activity sequences: one is the field of the
‘object of study’ that is explored by the discipline (e.g. the scientific phenomena
observed by scientists); the other is the field of ‘research’ referring to ‘the process
of enquiry and knowledge building’ (e.g. the scientific methods practised by sci-
entists) (2010: 121). This understanding helps us to make an important distinction
between biology as a body of knowledge (i.e. the object of study) and biology
as an academic discipline (involving both the object of study and research). This
chapter adopts Hood’s distinction, considering taxonomies for both the object of
study and research.

FIGURE 6.2 A simplified birdwatchers’ taxonomy of birds of prey (adapted from Wignell
et al. 1993: 156)
 Building taxonomies 139

FIGURE 6.3 A simplified scientific taxonomy of birds of prey (adapted from Wignell et al.
1993: 157)

Nominal realizations of taxonomies

Alongside the field perspective discussed above, previous studies have also adopted
a lexicogrammatical perspective on scientific taxonomies. It was recognized
that taxonomies are realized nominally, typically through a Thing (e.g. cell) or a
Classifier^Thing (e.g. cell wall) structure. Halliday suggests that this is because expe-
riential meanings construed by nominal groups are ‘more stable’ and ‘likely to be
subcategorized’ in comparison to other word classes (Hao 1998: 197). For this rea-
son, realizing taxonomies nominally implicates the transcategorization potential of
grammar – especially deriving nouns from other word classes (i.e. nominalization).
Naming a person who runs as a runner or a person who cooks as a cook both involve
transcategorizing a verb as a noun.
In addition to its vocabulary building function, transcategorization affords the
possibility of remapping the realization of discourse semantics in lexicogrammar.
For example, a discourse semantic figure (Halliday and Matthiessen 1999; Hao
2020b), which is typically realized through a clause (e.g., the salt diffused in the water),
can be remapped as a nominal group (e.g., the diffusion (of salt in the water)). The
remapping of a figure as a nominal group creates stratal tension between discourse
semantics and lexicogrammar, known as grammatical metaphor – here, to be
more specific, an experiential metaphor (Halliday 1994; Martin 2008).1
As scientific knowledge develops, what was initially an experiential metaphor
can be gradually ‘distilled’ (Martin 1993: 191) to become a conventional way of
talking about activities in a particular field; it thus loses stratal tension, becoming a
‘dead metaphor’ (Halliday 1998a). In science, distilled technical terms are typically
introduced by definition. For example, diffusion, which we treated above as a gram-
matical metaphor, can be defined as in Table 6.2. A definition is structured through
an elaborating relation of some kind – commonly via a Token^Value structure real-
izing a relational identifying clause.
140 Jing Hao

TABLE 6.2 A linguistic definition of diffusion

Term Definition
diffusion is the process whereby a substance in high concentration
moves to a place of low concentration
Token Process: intensive Value
identifying

The linguistically defined technical terms are often involved in construing tax-
onomies of activities (e.g. Halliday 1998a; Wignell et al. 1993; Martin and Rose
2007).This understanding connects meanings at the level of field with grammar.We
now turn to how resources in between field and grammar – i.e. discourse semantic
systems – were considered in previous studies.

Previous discourse semantic descriptions of entities

At the level of discourse semantics, language for realizing taxonomies has been pre-
viously discussed in relation to the discourse semantic system of ideation (Martin
1992). The description related to the realization of taxonomies is the classification
of entity types presented in Martin and Rose (2007), reproduced below in Table 6.3.
While the entity types in Table 6.3 are intended to cover a diversity of realiza-
tions, this categorization is problematic in many ways. First, most of the categories
are motivated largely from the perspective of field. The more delicate entity types
(i.e. everyday, specialized, technical and institutional) are in an one-to-one corre-
spondence to the field types presented in Martin (1992) (i.e. domestic, specialized,
administrative and exploration). Based on these categories, entities used in scientific
discourse (e.g. cells, gene, fungi, inflation) are seen as ‘technical’, under the general cat-
egory of ‘abstract’.This creates a number of problems in the application of these cat-
egories. For example in biological science, items are often tangible and observable,
with concrete physical presence (e.g. Birds of prey exemplified in Wignell et al. 1993).
Generalizing these as ‘abstract’ meanings is misleading with respect to the tangibility
of observable scientific phenomena. In addition, the one-to-one mapping of field
types and entity types creates uncertainty when determining whether the identified
‘entities’ represent discourse semantic features or field types.
Second, the category ‘metaphoric’ treats grammatical metaphors as ‘entities’,
which contradicts the concept of grammatical metaphor as stratal tension between
discourse semantics and lexicogrammar, rather than a meaning on a particular stratum.
This creates serious confusion in relation to the discourse semantic meanings at stake.
Third, while not explicitly specified in Table 6.3, technical entities also include
elements referring to activities (e.g., inflation). No criteria, however, are offered as
to how the nominalizations which function as ‘technical’ entities are to be distin-
guished from those that are experiential metaphors (e.g. exposure and humiliation in
Table 6.3). And no clear distinction is made with respect to the technical entities
referring to scientific items (e.g., gene, cells).
 Building taxonomies 141

TABLE 6.3 Kinds of entities (Martin and Rose 2007: 114)

Indefinite pronouns some/any/nothing/one

concrete everyday man, girlfriend, face, hands, apple, house, hill
specialized mattock, lathe, gearbox
abstract technical inflation, metafunction, gene
institutional offence, hearing, applications, violation, amnesty
semiotic question, issue, letter, extract
generic colour, time, manner, way, kind, class, part, cause
metaphoric process relationship, marriage, exposure, humiliation
quality justice, truth, integrity, bitterness, security

Furthermore, the instances of ‘generic’ entity are often realized through the Focus
function in a nominal group structure, in form of an embedded nominal group (e.g.
kind in the kind of; part in the part of). By identifying the meaning construed by Focus
group as a distinctive entity type over-privileges the perspective from grammar and
is inconsistent with Martin’s (1992: 314) suggestion that a Focus^Thing structure
as a whole enters into lexical cohesion.2 So in his terms, a Focus^Thing structure
realizes one discourse semantic unit, rather than two.
As a result of these inconsistencies, the framework shown in Table 6.3 is not
readily applicable to the analysis of the discourse semantics of ideational meaning.
A more appliable discourse semantic description is needed in order to understand
how taxonomies are construed by language. In the following sections, I introduce
a refined discourse semantic system of entities for analyzing taxonomies, including
types of entities and their associated dimensions (cf. Hao 2015).

A description of discourse semantic entities

The discourse semantic description of entities presented here is based on a range
of scientific texts used in undergraduate biology, including pedagogic materials
such as textbooks and laboratory manuals, and high-graded students’ written assess-
ments across undergraduate years (including 26 laboratory reports and six research
reports). I consider entities from the perspective of field, discourse semantics and
lexicogrammar by way of establishing criteria for their classification.
Looking from above, I draw on the description of field presented in Martin
(1992) as well as Doran and Martin (this volume). Field is re-articulated in Doran
and Martin (this volume) as knowledge constructed from two complementary
perspectives: a static perspective concerning classification and composition among
items and a dynamic perspective considering the unfolding of activities as either
expectancy or implication activities (i.e. as one activity ‘expecting’ what would fol-
low or as one activity ‘implicating’ what must follow). This distinction resonates
with previous descriptions of activity sequences and implication sequences (Martin
1992). Doran and Martin also consider qualitative and spatio-temporal properties
potentially associated with activities and items.
142 Jing Hao

Looking from around, I consider the interaction between entities and other dis-
course semantic systems, including in particular the interpersonal system appraisal
(Martin and White 2005) and the textual system identification (Martin 1992).
Due to space limitations, familiarity with these systems is assumed (see Martin 1992,
Martin and White 2005 for details).Within discourse semantics, I consider meanings
in terms of how entities are differentiated from and related to meanings construed
by figures and sequences of figures (Hao 2015; cf. Halliday and Matthiessen 1999).
Looking from below, of particular concern are the grammatical realizations of
entities at both clause and group ranks. For this analysis Halliday and Matthiessen’s
functional grammar of English (2014) is assumed.
It is important to note that in order to clearly distinguish meanings at different
strata, in what follows, the terms ‘taxonomy’ and ‘taxonomic relations’ are used only
to refer to meanings at the level of field; the discourse semantic resources construing
taxonomies will be introduced, including entities, co-elaborations and dimen-
sions. Figure 6.4 illustrates distinguishing terminologies across strata that we will
attend to in the following discussion.
For the purpose of illustration, I present here in Table 6.4 the relevant entity
types. The initial identification of entity types in Hao (2015, 2020b) includes
broader choices – i.e. source, thing, activity, semiotic, place and time entities. This
chapter focuses only on thing entities and activity entities (and their subtypes). It
is important to note that it is not the naming of a category that is important, but
rather the criteria the naming is based on. Each of the distinctions in Table 6.4 will
be explored in the following sections.

Thing entity vs. activity entity

From the static perspective at the level of field, a scientific item (e.g. a prokaryotic
cell) can be named through a discourse semantic entity, such as prokaryotes in (1).

FIGURE 6.4 Distinctive terminologies across strata

 Building taxonomies 143

TABLE 6.4 Entity types in undergraduate biology texts (c.f. Hao 2015, 2020b)

Types of entity Examples

thing instrumental ostensively defined plate, container, microscope, pipette
linguistically defined: glycerol medium, sodium carbonate
trained gaze
observational/ling. trained gaze insects, gut, sea urchin, herbivore
defined tech enhanced gaze fungal spore, eukaryotic cells,
inferable pathogen, enzyme, cytoplasm
activity enacted/ostensively experiment, method, treatment
defined
observed/ling. trained gaze maceration, peristaltic movement
defined tech enhanced gaze germination, fungal spore dispersal
inferable enzymic digestion, gene expression

The item can be potentially classified (e.g. prokaryotes vs. eukaryotes) or de/composed
(e.g. the plasma membrane and cytoplasms are parts of prokaryotes).

(1) Prokaryotes [thing entity] were seen.

From a dynamic perspective, phenomena are considered in terms of activities.

Activities can be either momented or unmomented. When an activity is unmo-
mented, it can be realized through a figure such as in (2). Here a figure construing
an activity is referred to as an occurrence figure, which is realized grammatically
through a material process.

(2) A cell engulfs food.

When activities are momented into a series of activities, they can be realized through
a sequence of figures in the discourse, as shown in (3).

(3) After a cell engulfs food by phagocytosis or pinocytosis,

newly formed food vacuoles fuse with lysosomes.
And then this mixes the food with the enzymes… (Reece et al. 2011: 880)

The figures in (3) are connected through temporal connexions (cf. Martin 1992)
(after; and then).3 The connexions are realized both within and across sentence bound-
aries.The connexion after is realized between the clauses in a clause complex, whereas
the connexion And then is realized between two sentences. Connexions between fig-
ures can also be implicit. For example, the connexions in (3) can be realized as in (4),
where the relation needs to be abduced from the surrounding discourse.

(4) A cell engulfs food by phagocytosis or pinocytosis.

(and then) Newly formed food vacuoles fuse with lysosomes.
(and then) This mixes the food with the enzymes.

In addition to figures and sequences, language allows activities to be construed

in a static way. To illustrate this, we consider example (5), which is selected from
144 Jing Hao

the co-text of excerpt (3). Here a term intracellular digestion is used to name the
momented activities realized in (3).

(5) The hydrolysis of food inside vacuoles is called intracellular digestions. (Reece
et al. 2011: 880)

By naming the momented activities through intracellular digestion, language enables

the field activities to be encapsulated into one term. This language resource is
coined as an activity entity, distinguished from the thing entities naming items
such as Prokaryotes. Both thing entities and activity entities can enter into lexical
cohesion in the discourse. Based on this discourse semantic understanding, Doran
and Martin (this volume) suggest that activity entities allow field activities to be
‘re-construed’ from a static perspective, becoming ‘itemized activities’. Like items,
an itemized activity can enter into classification and composition taxonomies. In
example (6), the activity entities (underlined) construe part/whole composition
taxonomic relations among itemized activities – i.e. food processing is the whole of
four different parts.

(6) The main stages of food processing are ingestion, digestion, absorption, and
elimination. (Reece et al. 2011: 880)

Figure 6.5 consolidates the meaning choices across the strata and their typical
interstratal relations.
At the lexicogrammatical level (looking from ‘below’), the realization of activ-
ity entities can often take the form of nominalization. For example, digestion is a

FIGURE 6.5 Choices across strata

 Building taxonomies 145

nominalization derived from the verb digest. However, this is not always the case.
Activity entities used in different fields can also draw on different lexical resources.
This will be elaborated later when different types of activity entities are concerned.

Thing entities: instrumental vs. observational

Among thing entities, a further distinction can be made between instrumental thing
entities, referring to items such as tools and apparatuses that are used in the labora-
tory during experiments (e.g. pipette, glycerol medium) and observational thing entities
referring to biological items that are observed during laboratory experiments (e.g.
prokaryotic organism). This distinction recognizes that the object of study and ‘doing
biology’ (Hood 2010) are two different sets of disciplinary knowledge which impli-
cate different sets of taxonomies.
To distinguish instrumental and observational thing entities, we can begin by
looking from ‘around’ in the discourse semantics. Of particular relevance are the
subtypes of appreciation in the attitude system (Martin and White 2005; Hao
and Humphrey 2012), including valuation concerning the worthiness; composition
concerning the organization; and reaction orienting to the significance.The distinc-
tion among these three types of attitude resonates with type of mental processes;
that is to say, valuation is motivated by cognition (‘thinking’), composition by per-
ception (‘seeing’), and reaction by affection (‘feeling’) (Martin and White 2005: 57).
In the scientific discourse, perception of composition tends to be associated with
measurable properties (cf. Hao and Humphrey 2012).
The two types of thing entities, instrumental and observational types, have dis-
tinctive ways of interacting with interpersonal meanings. Observational entities are
typically evaluated through attitudinal choices of valuation, such as play a crucial role
in (7); instrumental entities on the other hand are usually evaluated through com-
position such as imprecise in (8).

(7) Mandibles [observational entity] play a crucial role in the digestive process
of the locust.
(8) The balance [instrumental entity] used (in the experiment) was imprecise.

In addition, from a textual perspective, observational and instrumental entities are

identified differently in students’ experimental reports. While both observational
and instrumental entities are identified through specific reference, only instrumental
entities are presumed exophorically, by referring to items ‘outside’ the text in the
experimental setting – such as the instrumental entity balance in (8) (for identifica-
tion resources, see Martin 1992, Martin and Rose 2007).
Grammatically, the different participant roles through which instrumental and
observational entities are realized further confirm their opposition. In the students’
experimental reports, instrumental entities are typically realized by a Goal in a mate-
rial process, with people entities realized as Actors (typically implicitly in receptive
clauses, as in (9)); observational things on the other hand tend to be realized through
a Phenomena in mental processes, as in (10).
146 Jing Hao

(9) In this experiment a Finnpipette [instrumental entity, Goal] was calibrated

[material Process] (by us [people entity, Actor]).
(10) Both prokaryotic and eukaryotic organisms [observational entity, Phenomenon]
were seen [mental Process] (by me/us [people entity, Senser]).

In addition to Participant roles, instrumental entities are also often realized in

Circumstances, as either location (e.g. container in (11)) or manner (e.g. sodium car-
bonate in (12)).

(11) Set amount of water was pipetted (by us) into a container [instrumental entity,
Location].
(12) The reaction was stopped (by us) with sodium carbonate (6.9mM) [instrumen-
tal entity, Manner].

The distinction between instrumental and observational entities is sensitive to the

different activities they enter into, which leads us to distinguish between two types
of activity entities.

Activity entities: enacted vs. observed activity entities

From the perspective of field, activity entities are further categorized through their
relationship with types of activities – i.e. expectancy activity vs implication activity
(Martin and Doran this volume). Expectancy activities in undergraduate biology are
realized in the discourse through an enacted activity entity, including those referring
to scientists’ experimental methods, as those underlined in (13). Implication activi-
ties, typically those of the naturally occurring scientific phenomena, are realized
through observed activity entities, such as digestion and ingestion in (14). Both activity
entity types allow field activities to enter in the discourse without being momented.

(13) In this experiment a Finnpipette was calibrated, using three methods– weight-
of-water, spectrophotometry and radioactivity [enacted activity entities].
(14) The viability of fungal spores may be determined by the effect of the physical and
chemical processes involved in ingestion and digestion [observed activity entities].

In discourse analysis, distinguishing enacted and observed activity entities requires

identifying their agnate construal as sequences of figures. Sequences agnate to enacted
activity entities tend to be found in procedures such as laboratory manuals and recounts
such as those named as ‘Method’ in student’s experimental report (e.g. excerpt (15)).

(15) Set amounts of dye were pipetted into 1mL cuvettes,

And water was added
to give a total volume of 1mL.
Each solution was mixed,
and absorbances were read,
using a spectrophotometer.
 Building taxonomies 147

In contrast, sequences agnate to observed activity entities typically occur in explana-

tions. For instance, the temporal sequencing of figures in excerpt (3), reproduced
here in (16), agnates to the activity entity intracellular digestion, a type of digestion.

(16) After a cell engulfs food by phagocytosis or pinocytosis,

newly formed food vacuoles fuse with lysosomes.
And then this mixes the food with the enzymes…. (Reece et al. 2011: 880)

When recognizing sequences in the discourse, the difference between expectancy

and implication activities is reflected in different logical connexions used in the
sequence. Both temporal and causal connexions can be found; however, the con-
nexions are used differently for construing different activity types. We consider first
an example of expectancy activity, which can be itemized through an enacted activ-
ity entity, spectrophotometry in (13), and momented through a sequence in excerpt
(17). This sequence involves explicit and implicit temporal [successive] connexions
(italicized) between figures (i.e. and then), and manner (by (means of)) or purpose (in
order to); they indicate a facilitating relationship between the figures.

(17) Set amounts of dye were pipetted into 1mL cuvettes,

And water was added
(in order) to give a total volume of 1mL.
(And then) Each solution was mixed,
and absorbances were read,
(by) using a spectrophotometer.

In contrast, the fact that implication activities are both ‘chronological and logical’
(Barthes 1975) provides a distinctive criterion for identifying sequences agnate to
observed activity entities. This means that causal and temporal connexions between
figures are usually interchangeable. This can be exemplified by comparing the tem-
poral connexions in (16) and the paraphrased causal ones in (18).

(18) If a cell engulfs food by phagocytosis or pinocytosis,

then newly formed food vacuoles fuse with lysosomes.
So then this mixes the food with the enzymes…

A second set of recognition criteria between enacted and observed activity enti-
ties arises in the discourse through their interaction with interpersonal and textual
meanings (i.e. a perspective from ‘around’). Interpersonally, as exemplified in (19)
and (20), both enacted and observed activity entities can be coupled with apprecia-
tion [valuation] (in bold).

(19) Symbiotic relationships [observed activity entity] between termites and cellu-
lose-producing gut fungi may be beneficial to both insects and fungi.
(20) …making such a study [enacted activity entity] ecologically realistic and
important.
148 Jing Hao

However, in addition to valuation, enacted activity entities can also be evaluated

through appreciation [composition], such as in (21), which appears not to occur
with observed activity entities.

(21) This method [enacted activity entity] was time consuming.

Textually, both enacted and observed activity entities can be presented as generic
meanings. However, their generic presentation tends to draw on different lexico-
grammatical resources. Enacted activity entities involve the non-specific determin-
ers a/an, as in a biocontrol in (22). But observed activity entities are mostly (if not
always) represented without a determiner, e.g. symbiosis in (23).

(22) If entomopathogens are to be developed towards a biocontrol [enacted activity

entity] …
(23) Members of the Chytridiomycota may be involved in symbiosis [enacted activ-
ity entity] with the Echinoidea.

In addition, enacted activity entities differ from the observed ones in that they can
also be presumed by tracking a specific meaning in the preceding text – such as
this method in (21), pointing back in the text where the method was recounted in
Method stage. However, observed activity entities are not normally presumed.
Grammatically, while nominalization is a common lexicogrammatical realiza-
tion of both enacted and observed activity entities, enacted activity entities do not
always involve nominalizations – especially in more common sense (i.e. party and
game) and specialized fields (e.g. volley and overhead in tennis (Martin 2017)). By con-
trast, observed activity entities appear regularly as nominalizations, often drawing
on Latin and Greek etymology (White 1998) (e.g. vaccination from the Latin vacca,
cow; phagocytosis from the Greek phagein, to eat). When activity entities are realized
nominally, they can be confused with experiential metaphors, particularly realizing
figures metaphorically through nominal groups (e.g. ingestion [activity entity] is a
stage of food processing vs. the ingestion [experiential metaphor] of algae was successful).
A key distinction between the two is that activity entities are generalizable across
texts, whereas the use of experiential metaphor is specific to the unfolding of a par-
ticular text. For an in-depth discussion on the distinction between activity entities
and experiential metaphors, see Hao (2020a).
Identifying enacted and observed activity entities has enabled us to reveal the
discourse semantic resources construing field activities from a static perspective.
Before taking a further step, we take stock in Figure 6.6 of the oppositional
choices of entities that have been so far identified.

Linguistically defined vs. ostensively defined entities

A further distinction simultaneous with the thing and activity entities distinction
arises in terms of how these entities interact with textual meanings composing
 Building taxonomies 149

FIGURE 6.6 System of entity type (1)

mode. Based on studies in language development (Painter 1999), Martin suggests

that meanings in more common sense fields (including everyday and specialized
fields) are learned ostensively in spoken language as they are pointed out based on
one or another dimension of sensory experience (to see, touch, taste, hear and/or
smell) (2017: 117). By contrast, educational knowledge such as science relies heav-
ily on learning through language in written modes, in institutional settings (Martin
2007: 40–41).This interaction of field with mode motivates a further entity distinc-
tion between those which are ostensively defined in relation to sensuous experience,
and those which are linguistically defined through definitions in written language.
This opposition is reflected in the way entities are textually identified through
choices in identification. An ostensively defined entity used in an everyday dis-
course typically refers exophorically to a meaning ‘outside’ the text, as in example
(24) (selected from Painter (1999: 84)).

(24) Child: What’s that?

Mother: That [exophoric pointing; Token] ’s my belly button [specific;Value].

Through identification of this kind, an item enters into the discourse as an entity.
Building on the naming of such instances, young children can subsequently develop
an orientation to classes – e.g. dolphins are mammals, with both entities being ‘named’.
In the written texts of undergraduate biology, identification is more complex.
While some instances of instrumental things tend to combine with exophoric ref-
erence and implicate ostensively defined meanings (i.e. the balance in (7) above), the
majority of entities are introduced and tracked within the text and they are typically
provided with linguistic definitions in pedagogic materials.
Linguistic definitions explicitly establish relationships between entities. In the
definition of lysosome below, lysosome is related to other entities including membra-
nous sac, hydrolytic enzymes and animal cell.

(25) A lysosome [Token] is a membranous sac of hydrolytic enzymes that an animal cell
uses to digest all kinds of macromolecules [Value] (Reece et al. 2011: 106)

In addition to linguistic definitions, presentations of ‘how things look’ in pedagogic

texts rely heavily on multimodal resources, including photographs, illustrations and
150 Jing Hao

FIGURE 6.7 Visual representation of an item in undergraduate textbook (Reece et al.

2014: 11)

diagrams. These images present instances of entities. For example, in Figure 6.7, the
caption describes one of the distinctive features of an item introduced as a generic
entity (Kingdom Fungi).The specific reference to the image – this mushroom – exem-
plifies the generic category.
Imagic exemplification of linguistically defined entities suggests a difference
between characterizing phenomena through ostensive definition and doing so in
terms of their tangibility. This will be the focus of the next section.

Tangibility of linguistically defined entities

The tangibility of linguistically defined entities allows us to identify three further
types: trained gaze entities referring to items and activities can be observed with
human senses, tech-enhanced gaze entities referring to items and activities whose
observation requires technology, and inferable entities referring to items and activi-
ties whose physical existence is inferred rather than directly seen and observed.
The tangibility of linguistically entities can be revealed in terms of how they
interact with images. A trained gaze entity, such as a sea urchin, is often represented
in pedagogic texts through realistic photographs such as the image in Figure 6.8.
This realistic image illustrates the tangible observable nature of trained gaze entities.
The caption here maintains the description generically (e.g., sand dollars, sea stars,
marin animals, a sea urchin).
Like trained gaze entities, tech-enhanced gaze entities also refer to relatively
tangible items; but they need to be perceived with technology. For instance, we can
observe the presence of Chytrids, a kind of fungi, using a microscope. Its presence can
be realistically captured and visually represented, such as in Figure 6.9. This obser-
vation would have been impossible with our naked eyes. The image provides an
instance of the entities which are presented generically in the caption (e.g., Chytrids,
multicellular, branched hyphae).
 Building taxonomies 151

FIGURE 6.8 A visual representation of a trained gaze entity – sea urchin (Reece et al.
2014: 683)

FIGURE 6.9 Visual representation of Chytrids (Reece et al. 2014: 655)

In contrast to the interaction with realistic imagic representation, inferable enti-

ties tend to be made ‘visible’ through illustrations and diagrams, which Kress and
van Leeuwen (2006) refer to as analytical diagrams. In the visual representation in
Figure 6.10, both the inferable activity entity chemical reaction and the inferable things
involved (i.e. glucose molecules, simpler molecules) are represented through illustrations.
The interaction between entities and images suggests that different types of lin-
guistically defined entities have different choices of visual representation.4
To summarize, we represent the entity types for scientific taxonomy building in
Figure 6.11.
These entity types are helpful for thinking about the diversity of taxonomies
in biological sciences. However, in order to fully understand taxonomy building
including diversity as well as its depth, we need to further explore the relationships
between entities of the same type.

Dimensionality
In this section we explore the depth of taxonomies by describing the relationships
between entities of the same type. The resources for differentiating one entity from
152 Jing Hao

FIGURE 6.10 Visual representation of chemical reaction (Reece et al. 2014: 146)

FIGURE 6.11 System of entity type (2)

 Building taxonomies 153

another (of the same type) are conceptualized here as discourse semantic augmenta-
tion of entities, referred to as dimensions. Below are some examples of dimensions
(in italics) augmenting entities (underlined).

the size of the cell

the shape of sea urchin
the kind of birds
parts of a cell

The system of dimensions, dimensionality, is simultaneous with the system of

entity types in the entity system. In other words, all entities can be dimensioned.
The more an entity is dimensioned, the more in-depth knowledge of the item is
developed. The symbol ‘<’ and ‘>’ are used to annotate the augmenting relationship
between entity and dimension, depending on their order in the text. The arrow
points always towards the entity (e.g. the size of > the cell; the cell’s < size).
Looking from above, dimensions allow properties and relationships among items and
activities to be catalogued and named through language. Properties, which are typically
realized through qualities in the discourse, such as big in (26), are ‘itemized’ (Martin and
Doran this volume) through dimensions such as size in (27). Notice that Focus^Thing
structure of a nominal group (e.g., the size of the cell; for Focus group analysis see Martin
et al. 2010) provides a canonical grammatical realization of dimension>entity.

(26) The cells [entity] are very small [quality].

(27) The size [dimension] of > the cell [entity] was measured.

In building taxonomic relations, entities in the discourse are co-related with one
another. As the text unfolds, each instance in the discourse is understood not in iso-
lation but in relation to its superordinate, co-classes and subclasses, and to its whole,
co-parts and parts. The relationship between entities is ‘co-elaborative’ (annotated
as ‘entity=entity’) (e.g. dolphins are mammals [dolphins = mammals]), in the sense that
two related entities both indicate the classification and composition with respect to
each other. Co-elaboration can be either realized lexicogrammatically, through a
range of grammatical resources, or abduced from the co-text. For example in (28),
the co-elaboration between B-galactosidase and enzyme is realized through an elabo-
rating nominal group, their further co-elaboration to protein is subsumed in the text.

(28) The activity of proteins can be controlled… In this experiment the activity of
B-galactosidase, an enzyme which breaks down lactose, was studied.

While co-elaborations construe taxonomic relations as the text unfolds (see Hood
and Hao this volume), dimensions can provide explicit naming of taxonomic rela-
tions established, such as kind in (29).

(29) enzyme [entity] is a kind [dimension] of > protein [entity].

154 Jing Hao

TABLE 6.5 Realizing taxonomic relations through co-elaborations and dimensions

Co-elaboration without dimension Co-elaboration with dimension

Thing^Qualifier parts of > the sea urchins;
the coelomic cavity and digestive tract of = the sea urchins
possessive Deictic^Thing part of > rainforest;
rainforest’s = canopy
elaborating nominal group complex a kind of >irregular urchin;
Chytridiomycota were present in the digestive system of the superordinate of > Echinocardium
1 the irregular urchin, = 2 Echinocardium cordatum cordatum
relational identifying process a kind of > enzyme;
B-galactosidase is = an enzyme the superordinate of > B-galactosidase
which breaks down lactose.
relational attributive process a kind of > food source
Glucose is = a preferred simpler food source.

Table 6.5 illustrates a range of co-elaborations construing taxonomic relations

which can be potentially named with dimensions.
Before drawing perspectives from around and below, we present an overview
of the types of dimension identified in undergraduate biology texts (Table 6.6) –
categorized, structured, measured and perceived dimensions (cf. Hao 2015).
Categorized dimensions offer a way of talking about how items are interrelated
as superordinate to subtype, or class to instance; structured dimensions include
resources for relating items as parts to wholes; measured dimensions provide ways of
talking about properties which are otherwise realized in the discourse through qual-
ities (e.g., high, small, long) and quantities (e.g., 5 meters); and perceived dimensions
name properties based on our sensory experience (to see, smell, touch, hear, etc.).
When looking ‘from around’ at the level of discourse semantics, entity and
dimension as a whole interact with other discourse semantic meanings. For exam-
ple in (30), Members of Chytridiomycota as a whole is an evaluated target (ecologi-
cally important). Similarly in (31) what is being measured and recorded is ‘measured
dimension>entity’ as a whole.

(30) Members of > Chytridiomycota [categorized dimension>observational entity]

are ecologically important [valuation].

TABLE 6.6 Dimensionality of entities (c.f. Hao 2015)

Types of dimension Examples

categorized type, kind, category, species…
structured part, component, layer, level, stage…
measured size, weight, height, length…
perceived color, shape, texture, smell….
 Building taxonomies 155

(31) The weight of > the water [measured dimension>instrumental entity] was mea-
sured and recorded.

Instances of dimensions can sometimes appear with an elided entity, such as the weight
was measured. Such instances occur only after the entity has been introduced in the
preceding text. As exemplified in (32), the omitted entity DNA, which is elaborated
with a measured dimension quantity, can be recovered based on its presentation in
the previous clause. The recovered choices are annotated with superscript.

(32) Isolating DNA and determining the quantity > of DNA extracted…

It is also possible to elide a dimension. As exemplified in (33) and (34), the elided
perceived dimension shape and categorized dimension kind/instance can be readily
inferred, shown in the superscript.

(33) The shape of > the sea urchin is irregular.

(34) This kind/instance of > irregular urchin is called Echinocardium cordatum.

Given the diverse range of linguistically defined entities in biology texts, dimen-
sions of linguistically defined entities are also found to be ‘field-specific’ in biology.
Scientists have developed ‘field-specific’ ways of naming the relationships among
items of organisms, such as domain, kingdom, phylum, class, order, family, genus and
species. Field-specific dimensions typically also require linguistic definition. For
example, species is defined in the textbook as following: each of these forms of life (i.e.
plants, animals, fungi and bacteria in the proceeding text) is called a species (Reece et al.
2011: 4).
A further characteristic of field-specific dimensions is that they may originate
from grammatical metaphors. This is particularly the case in biology for measured
dimensions, as part of the distillation of technicality. For example, a number of field-
specific measured dimensions in (35) such as activity are associated with inferable
entities (proteins, B-galactosidase):

(35) The activity of > proteins can be controlled… In this experiment the activity of
> B-galactosidase, an enzyme which breaks down lactose, was studied.

The nominalization activity here is derived from the adjective active. However,
instead of representing an experiential metaphor, the term activity used here is
defined in the textbook as how efficiently the enzyme functions and the measurement
of activity (of protein/enzyme) is based on a mathematical equation of quantifica-
tion: Activity of enzyme = moles of substrate concerted per unit time = rate × reaction
volume.
Grammatically, dimensions are realized either canonically through a Focus group
in Focus^Thing structure, or through a Thing when the entity is elided, or being
rendered as other nominal group structures as outlined in Table 6.7.
156 Jing Hao

TABLE 6.7 Grammatical realizations of dimensions in biology texts

Lexicogrammatical realizations of Examples

dimensions
Thing type; parts; shape; weight
Focus^Thing a kind of >sea urchin;
members of > the geofungi
Classifier^Thing regular sea urchin < species
Deictic^Thing sea urchin’s < shape
Thing^Qualifier mandibles < of different sizes

I have now illustrated the choices for discourse semantic entities and their associ-
ated dimensions in undergraduate biology texts.The description identifies resources
that connect taxonomies of items and activities at the level of field with a range of
realizations at the level of lexicogrammar. In the next section, I demonstrate how
this description enables a thorough exploration of taxonomies demonstrated in an
excerpt of research report produced at the final undergraduate year.

Analyzing taxonomy in a research report

The selected text for the illustrative analysis is a student research report produced at
the final year of undergraduate coursework. The text reports on an investigation of
insects interacting with fungal spores. As an illustration, I will focus on the Abstract
previewing the report. The excerpt (36) presents this annotated text. The entities
are underlined; the dimensions are in italics; and the elided entities, dimensions and
their substitutions are made explicit and annotated as superscripts.

(36) The viability of > fungal spores after ingestion and passage through the gas-
trointestinal tract of an insect may be determined by the effect of physi-
cal and chemical processes involved in the ingestion and digestion of food.
In particular, mandibular maceration could damage fungal spore < integrity
and result in spores losing their (fungal spores’) < viability. The purpose of this
study was to determine whether the size of > fungal spores was a factor in
affecting spore < viability after passage through the gastrointestinal tract of
the Australian plague locust, Chortichocetes terminifera. The effect of spore
< size on viability > of spore was tested using five genera > of dung fungi –
Absidia, Isaria, Penicillin, Phycomycetes and Podospora – whose spores were
fed to either second or fifth instar C. terminifera. Absidiaspores, Isariaspores and
Penicillin spores were recovered from both the faecal and gut < samples from
secondinstars and fifth instars, following feeding by C. terminifera on wheat
inoculated with fungal spores. Phycomyces was not recovered from faecal
material obtained from second instars but was present in all other samples >
of gut. Podospora spores, 20um in diameter, were not recovered from any of

the samples > of gut.

 Building taxonomies 157

The analysis outlines the complex taxonomies construed in this text both in terms
of their diversity and depth. The discussion breaks down the analysis based on dif-
ferent entity types.
To begin, the text demonstrates significant use of thing entities, including both
tech-enhanced things and trained gaze things. Tech-enhanced realizations refer
mostly to the item dung fungus and its part fungal spore.The entities are co-elaborated
in relation to one another. For example, dung fungi as a superordinate is co-elabo-
rated by the subtypes Absidia, Isaria, Penicillin and Phycomycetes. The classification is
named through the categorized dimension genera. The co-elaborations associated
with dung fungus construe the taxonomy displayed in Figure 6.12. The diagram also
includes the item Zygomyce appearing in the subsequent text (in Discussion section)
of the research report. The items unmentioned in the report are left implicit in the
diagram.
In addition to the categorized dimension, measured dimensions augmenting the
entity fungal spore are also salient in this text, including both field-neutral measure-
ments (e.g., size) and field-specific ones (e.g., integrity, viability). These measured
dimensions construe properties of the item from a static perspective. The itemized
properties allow for the re-construal of the properties that are typically realized
through qualities and quantities in the discourse. In example (37), selected from the
Discussion of the report, a property the item spores is realized through a quality larger
and a quantity 14-20um. In the Abstract section the property is itemized through
the measured dimension size, as shown in (38).

(37) The larger spores of Podospora (14–20um) did not retain their viability.
(38) The purpose of this study was to determine whether the size of > fungal spores
was a factor in affecting spore viability

The multiple dimensions augmenting fungal spore indicate that establishing proper-
ties of the item is a prominent feature of this text, reflecting an in-depth knowledge
of the item.
The other salient thing type in this excerpt is trained gaze things (e.g., insects,
gut and Australian plague locust). These entities are co-elaborated around the entity
locust, construing both its classification and composition. For example, Chortichocetes
terminifera as a kind of locust is indicated by an elaborating nominal group complex

FIGURE 6.12 Categorization of dung fungus

158 Jing Hao

FIGURE 6.13 Categorization in relation to locust

FIGURE 6.14 Structure of locust

(Australian plague locust, = Chortichocetes terminifera). Its further elaboration to higher

level category (i.e. locust = insect) is subsumed in the text. Similarly, the part/
whole co-elaboration between the gastrointestinal tract and locust is realized through
a Thing^Qualifier structure (e.g., the gastrointestinal tract = of an insect). While an
instance of categorized dimension naming classification was found (e.g., sample relat-
ing to class to instance), no structured dimensions naming composition were found.
Figure 6.13 and Figure 6.14 provide an overview of the classification and compo-
sition in relation to locust established throughout the whole text of the research report.
Along with the tech-enhanced things, corresponding activity entities (e.g., man-
dibular maceration, ingestion) were also found. The activity entities are co-elaborated
by connecting activities both at different tiers (e.g., physical and chemical processes
= ingestion/digestion), and at the same tier (e.g., digestion = ingestion). However, the
co-elaborations are not named as dimensions (e.g., a stage of digestion; stages of food
processing), indicating that classification and composition of activities are not the
focus of the study.
Few instances of enacted entities were identified in the excerpt. This suggests
that its object of study focuses on the biological knowledge (i.e. ‘what is observed’),
rather than research methodology (i.e. ‘how to observe’).
In summary, the preview of research report reveals relative broad and deep tax-
onomies, referring to tangible and observable items and activities. The most elabo-
rated and augmented entities are those of trained gaze and tech-enhanced ones.
This suggests that demonstrating static knowledge of items and activities at different
levels of tangibility is an important task in this text.
 Building taxonomies 159

Onward and outward

In this chapter I have illustrated a discourse semantic framework for analyzing tax-
onomies in undergraduate biology. Importantly, the chapter illustrates a ‘tri-stratal’
method for describing discourse semantic meanings, particularly entity types and
their dimensions. With this method, highly technical meanings of entities and
dimensions can be made transparent. The approach can be used to explore how
language construes taxonomies across disciplinary fields.
As far as undergraduate biology is concerned, the analysis of taxonomies in the
excerpt of a student’s research report reveals a diverse range of entities referring to
relative tangible items (observed either through naked eyes or through microscope).
This suggests that being able to establish taxonomies of items and activities at differ-
ent levels of tangibility is significant for the knowledge development of biological
sciences.
Dialogue with LCT, in particular with respect to knowledge building, has
encouraged the development of further work on the ‘language’ of disciplinary
knowledge. SFL’s perspective on entities and dimensions can be related to many
aspects of knowledge building in LCT, including the ‘epistemic-semantic density’
concepts presented in Maton and Doran (2017a, 2017b). This paper is conceived as
a contribution to further productive interdisciplinary conversation.

Notes
1 Grammatical metaphors have been distinguished as experiential, interpersonal and logi-
cal metaphors (Halliday 1994, Martin 2008). This chapter will only be concerned with
experiential metaphors.
2 Note that the use of Focus group here follows the terminology in Martin, Matthiessen
and Painter (2010). Focus group is referred to as Pre-Deictic in Martin (1992).
3 Connexion is used to substitute Martin’s (1992) discourse semantic ‘conjunction’. This
reserves ‘conjunction’ as a grammatical term as used in Halliday and Matthiessen (2014).
4 It is important to note that while some items are usually inferred in biological experi-
ments, in chemistry they can be shown through technology such as the sophisticated
imaging instrument PET scanner, which can monitor chemical processes. In other words,
the distinction between a tech-enhanced gaze entity and an inferable entity in a field
depends on the different technologies deployed.

References
Barthes, R. (1975) ‘Introduction to the structural analysis of narrative’, New Literary History,
6(2): 237–72.
Bernstein, B. (2000) Pedagogy, Symbolic Control and Identity, London: Rowman and Littlefield.
Halliday, M. A. K. (1994) An Introduction to Functional Grammar, London: Edward Arnold.
Halliday, M. A. K. (1998a) ‘Things and relations: Regrammaticizing experience as techni-
cal knowledge’, in J. R. Martin and R. Veel (eds) Reading Science, London: Routledge,
185–236.
Halliday, M. A. K. (1998b) ‘Language and knowledge: The “unpacking” of text’, in J. Webster
(Ed.) The Language of Science, London: Continuum, 24–48.
160 Jing Hao

Halliday, M. A. K. and Matthiessen, C. M. I. M. (1999) Construing Experience Through Meaning:

A language-based approach to cognition, London: Cassell.
Halliday, M. A. K. and Martin, J. R. (1993) Writing Science: Literacy and discursive power, London:
Falmer Press.
Hao, J. (2015) ‘Construing Biology: An Ideational Perspective’, unpublished PhD thesis, University
of Sydney.
Hao, J. (2018) ‘Reconsidering “cause inside the clause” in scientific discourse – from a
discourse semantic perspective in systemic functional linguistics’, Text and Talk, 38(5):
520–50.
Hao, J. (2020a) ‘Nominalisation in scientific English: A tristratal perspective’, Functions of
Language, 27(2): 143–73.
Hao, J. (2020b) Analysing Scientific Discourse from a Systemic Functional Linguistic Perspective:
A Framework for Exploring Knowledge Building in Biology, London: Routledge.
Hao, J., and Humphrey, S. (2012) ‘The role of ‘coupling’ in biological experimental reports’,
Linguistics and the Human Sciences, 5(2): 169–94.
Hood, S. (2010) Appraising Research: Evaluation in Academic Writing, New York: Palgrave
Macmillan.
Kress, G. R. and van Leeuwen,T. (2006) Reading Images:The grammar of visual design, New York
and London: Routledge.
Martin, J. R. (1992) English Text: System and Structure, Amsterdam: Benjamins
Martin, J. R. (1993) 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, London: The
Falmer Press, 166–202.
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, London and New York:
Continuum, 3–39.
Martin, J. R. (2007) Construing knowledge: a functional linguistic perspective, in F. Christie
and J. R. Martin (eds) Language, Knowledge and Pedagogy: Functional linguistic and sociological
perspectives, London and New York: Continuum, 34–64.
Martin, J. R. (2008) ‘Incongruent and proud: De-vilifying ‘nominalization”, Discourse and
Society, 19(6): 801–10.
Martin, J. R. (2013) Systemic Functional Grammar: A Next Step into the Theory – Axial Relations,
Beijing: Higher Education Press.
Martin, J. R. (2017) ‘Revisiting field: Specialized knowledge in Ancient History and Biology
secondary school discourse’, Onomázein Special Issue, March: 111–48.
Martin, J. R. and Maton, K. (2013) (eds) ‘Cumulative knowledge-building in secondary
schooling’, Linguistics and Education 24(1), 1–74.
Martin, J.R. and Maton, K. (2017) ‘Systemic functional linguistics and Legitimation Code
Theory on education: Rethinking field and knowledge structure’, Onomázein Special Issue,
March: 12–45.
Martin, J. R. and Rose, D. (2007) Working with Discourse: Meaning beyond the clause, New York:
Continuum.
Martin, J. R. and Rose, D. (2008) Genre Relations: Mapping culture, London: Equinox.
Martin, J. R. and White, P. R. R. (2005) The Language of Evaluation: Appraisal in English,
London: Palgrave.
Martin, J. R., Matthiessen, C. M. I. M. and Painter, C. (2010) Deploying Functional Grammar,
Beijing: The Commercial Press.
Maton, K. (2014) Knowledge and Knowers: Towards a Realist Sociology of Education, London:
Routledge.
 Building taxonomies 161

Maton, K. and Chen, R. T.-H. (2020) Specialization codes: Knowledge, knowers and stu-
dent success, in J. R. Martin, K. Maton and Y. J. Doran (eds) Accessing Academic Discourse,
London: Routledge, 35–58.
Maton, K. (2020) ‘Semantic waves: Context, complexity and academic discourse’, in J. R.
Martin, K. Maton and Y. J. Doran (eds) Accessing Academic Discourse, London: Routledge,
59–85.
Maton, K. and Doran,Y. J. (2017a) ‘Semantic density: A translation device for revealing com-
plexity of knowledge practices in discourse, part 1 – wording’, Onomázein Special Issue,
March: 46–76.
Maton, K. and Doran,Y. J. (2017b) ‘Condensation: A translation device for revealing complex-
ity of knowledge practices in discourse, part 2 – Clausing and sequencing’, Onomázein
Special Issue, March: 77–110.
Maton, K., Hood, S. and Shay, S. (eds) (2016) Knowledge-building: Educational studies in
Legitimation Code Theory, London: Routledge.
Matthiessen, C. M. I. M. and Halliday, M. A. K. (1997/2009) Systemic Functional Grammar: A
first step into the theory, Beijing: Higher Education Press.
Painter, C. (1999) Learning through Language in Early Childhood, London: Cassell
Reece, J. B., Urry, L. A., Cain, M. L.,Wasserman, S. A., Minorsky, P.V., and Jackon, R. B. (2011)
Campbell Biology (9th Edition), San Francisco: Pearson.
Reece, J. B., Urry, L. A., Cain, M. L.,Wasserman, S. A., Minorsky, P.V. and Jackon, R. B. (2014)
Campbell Biology (10th Edition), San Francisco: Pearson.
White, P. (1998) ‘Extended reality, proto-nouns and the vernacular: Distinguishing the tech-
nological from the scientific’, in J. R. Martin and R. Veel (eds) Reading Science, London:
Routledge, 167–97.
Wignell, P., Martin, J. R. and Eggins, S. (1993) ‘The discourse of geography: Ordering and
explaining the experiential world’, in M. A. K. Halliday & J. R. Martin (eds) Writing
Science, London: Falmer, 151–83.
7
MULTIMODAL KNOWLEDGE
Using language, mathematics and images
in physics

Y. J. Doran

Introduction
What knowledge do students need in order to be successful in science? Such a
question pushes at the heart of educational programmes that aim to develop a
comprehensive pedagogy that reaches across disciplines. In one sense, the answer
is simple. The knowledge students need is specified in the syllabus, detailed in the
textbook and explained in the classroom. From this perspective it is simply a matter
of reading the textbook, listening to the teacher and writing down the knowledge
in the exam or in an assignment. However, few in educational research would argue
for such a simplistic model. For one, it is clear from decades of research into literacy
across the curriculum that the processes of reading, listening and writing are far
from unproblematic. From an educational linguistic perspective, there is a range of
highly specific literacies at stake; students must be able to read and write a wide
range of scientific genres, and interpret and make use of highly intricate scientific
language (Martin 1985, Lemke 1990, Rose et al. 1992, Halliday and Martin 1993,
Christie and Martin 1997, Martin and Veel 1998, Unsworth 1997, 2001a, Halliday
2004, Martin and Rose 2008, Martin and Doran 2015, Hao 2020). In the classroom
they need to be able to listen and interpret what the teacher is saying, engage with
them in dialogue and successfully reconstrue scientific meanings at the teacher’s
bidding (Rose 2004, 2014, 2020, chapter 11 of this volume, Christie 2002, Rose
and Martin 2012).These literacy demands involve not just language but also extend
to the multiliteracies inherent in science schooling – where language, mathematics,
images, specialized symbolic formulae, animations and demonstration apparatus all
need to be ‘read’ as one and reorganized where necessary in assignments and exams
(Lemke 1998, 2003, Kress et al. 2001, Unsworth 2001b, O’Halloran 2005, Parodi
2012, Doran 2017, 2018, 2019, Doran and Martin, Chapter 5 of this volume).
To learn the knowledge of a discipline, one must be able to grasp the way it is
 Multimodal knowledge 163

construed through language and other semiosis; and to show you have the knowl-
edge, you must be able to marshal these specialized linguistic and semiotic resources
to reconstrue these meanings.
Such mastery of a wide range of literacy demands is in many ways the crux of
many issues in learning how to do science. But simply being able to read and write
scientific language and the particular text types needed across the curriculum is not
enough. Students need to understand where and when each particular text type,
semiotic resource or particular linguistic resource is appropriate. They need to be
able to interpret new situations and new demands and organize the meanings they
have learnt in an appropriate manner.That is, they need to understand the principles
underpinning the selection and application of particular meanings and why they are
used. At stake here is a way of seeing the world. Students must develop a scientific
gaze that allows them, among other things, to shift between the knowledge of the
empirical world and that of abstract theory, and between everyday understandings
and technical conceptions. They need to be able to make connections between
phenomena that at first glance may seem disparate and carry out specialized proce-
dures and protocols to investigate prescribed phenomena. And they must be able to
marshal particular literacies to do this at the appropriate time.
This chapter explores how knowledge is organized multimodally in science,
focusing in particular on physics as it is taught in schooling. It will explore two main
concerns: the technical meanings construed through the key resources of language,
image and mathematics, and the underlying principles that organize a scientific
gaze and enable students to understand when to use particular technical mean-
ings and semiotic resources. To do this, it will view the meanings made from two
perspectives. First it will consider scientific meanings through the register variable
field from Systemic Functional Linguistics (SFL) (Doran and Martin, chapter 5 of
this volume). Second it will explore these meanings through the sociological frame-
work of Legitimation Code Theory (LCT), which is being widely used alongside
SFL (Maton and Doran 2017a, Martin et al., 2020), specifically its dimension of
Semantics (Maton 2014, 2020).
The perspective from field in SFL will explore how particular technical meanings
are organized by different resources – language, mathematics and image. Here we will
be concerned with whether technical meanings are presented statically as set of items
positioned in taxonomies (either of classification – type/subtype, or of composition –
part/whole), or whether they are presented dynamically as series of events (known
as activities) given in more or less detail. In addition, we will be concerned with
whether these items or activities involve particular properties that can be measured
and to what degree all of these meanings are placed in large, interdependent net-
works of meaning. Each dimension of field will be introduced in more detail as they
become relevant, but this analysis will show that particular semiotic resources tend to
specialize in particular ways of organizing meaning, while backgrounding others. In
this sense, this chapter will show that the organization of content in many disciplines
is inherently multimodal, as particular components of its technical array tend to be
given full expression through particular semiotic resources.
164 Y. J. Doran

Complementing this view from SFL we will also consider knowledge in physics
from the perspective of Semantics in LCT (Maton 2014, 2020). This will give an
insight into two organizing principles that characterize knowledge practices. The
first is known as semantic gravity (SG), which conceptualizes the degree of context-
dependence of meaning. Stronger semantic gravity (SG+) indicates meanings are
more dependent on their context; weaker semantic gravity (SG–) indicates mean-
ings are less dependent on their context. For example, in a primary school physics
text that we return to below, the concepts of pushing and pulling are introduced by
listing a series of examples, such as Michael pushes the keys to a tune, which refers to
an image of a child playing a keyboard (Riley 2001). In this case, we would say that
the pushes here involves relatively strong semantic gravity (SG+) as it is dependent
on a particular instance of pushing – it has relatively high context-dependence. In
contrast, later on in the book, the text generalizes in order to introduce the concept
of force, using the sentence A push is a force. Here push does not make reference to
any particular instance of pushing, rather can be applied to any number of instances.
In this case, we can describe push as indicating weaker semantic gravity (SG–) than
the first instance as it has less context-dependence. Shifts in semantic gravity are
crucial to the organization of knowledge in physics (and indeed all sciences) and are
organized through language, mathematics and image.
The second organizing principle at the centre of Semantics in LCT is semantic
density (SD). Semantic density refers to the complexity of meaning. Stronger seman-
tic density (SD+) indicates more complex meanings; weaker semantic density (SD–)
indicates less complexity. In the sentence a push is a force, the term force is a technical
term in physics. As students move through the years, this term becomes more cen-
trally integrated into the vast network of meaning organizing physics knowledge.
For example, it underpins Newtown’s laws of motion, which organize the field
of mechanics; it underpins different types of field in both nuclear physics and in
relativity; and it forms a variable within innumerable mathematical equations at all
levels. In this sense, force exhibits relatively strong semantic density (SD+) as it has
a relatively complex set of meanings associated with it. In contrast, the term push
exhibits significantly weaker semantic density (SD–), as it does not resonate out to
the same degree range of technical meaning in the field.1
Semantic gravity and semantic density can vary independently. On this basis, the
semantic plane in Figure 7.1 (Maton 2016: 16) shows the possibilities for variation –
with semantic gravity and semantic density as axes and each quadrant representing
a different semantic code.
Rhizomatic codes (top-right quadrant) are meanings that maintain weaker seman-
tic gravity (SG–) and stronger semantic density (SD+). In common-sense terms, this
quadrant indicates meanings that are relatively ‘generalized’ and ‘complex’, as often
exemplified by technical theory. In contrast, prosaic codes (bottom-left quadrant) are
meanings with stronger semantic gravity (SG+) and weaker semantic density (SD–);
meanings that are ‘concrete’ and ‘simple’, and often illustrated with common-
sense everyday knowledge. Worldly codes (bottom-right quadrant) exhibit stron-
ger semantic gravity and stronger semantic density; meanings that are both highly
 Multimodal knowledge 165

FIGURE 7.1 The semantic plane (Maton, 2016, p. 16)

context-dependent and very complex. This quadrant is often illustrated by knowl-

edge in areas such as vocational education, and practically oriented fields and crafts,
where projects are often embedded in a specific instance or case (such as building
a particular bridge over a particular river) but involve a complex knowledge-base
to achieve their goals. Finally, rarefied codes (top-left quadrant) are meanings with
weaker semantic gravity and weaker semantic density; meanings that are relatively
context independent but simple.
The semantic plane enables us to see the possible combinations and gradations
available across the disciplinary map. Focusing on physics, it is tempting to imme-
diately position it within the top-right quadrant as a rhizomatic code – as being
both ‘abstract’ (weaker semantic gravity) and ‘technical’ (stronger semantic density).
Indeed if we follow Biglan’s (1973) characterization of physics as a relatively ‘pure’
discipline (i.e. non-applied) and a relatively ‘hard’ science, or Kolb’s (1981) descrip-
tion of physics as both a relatively reflective (non-applied) but also a particularly
‘abstract’ discipline, then a classification of physics as a rhizomatic code makes sense.
However, once we look at the actual practices of any field, it becomes clear very
quickly that no discipline fits into a single box. There is immense variation across
sub-disciplines (is it nuclear physics? classical mechanics? electromagnetism? astro-
physics?); across year level (elementary? secondary? tertiary? which year in each?),
across individual classes and across research, practice and education. LCT is invalu-
able here: the semantic codes are not boxes – the semantic plane is a topological
space with indefinite variation within each code.
As we will see, one of the key shifts that occurs in physics texts from secondary
education onwards is the ability to move between abstract theory and empirical
instances while still using highly complex knowledge. Put another way, a key pat-
tern is the varying of semantic gravity while maintaining relatively strong semantic
166 Y. J. Doran

density – shifts between rhizomatic codes and worldly codes. Similarly, when teach-
ing new concepts, teachers and textbooks will regularly shift between ‘everyday’,
‘concrete’ knowledge of examples (prosaic codes) and the abstract theoretical
knowledge typically considered the realm of physics (rhizomatic codes). In short,
the key to doing physics is not just in being able to understand highly technical,
abstract knowledge, but to vary the abstraction and technicality as required by the
situation and to utilize the particular semiotic resources that organize this.
The following sections will step through the ways language, mathematics and
image are used to organize both the technical meanings of physics and the shifts in
semantic gravity and semantic density that allow students to bridge between ‘theory’
and the ‘everyday’, the ‘empirical’ and the ‘abstract’, and the ‘concrete’ and the ‘com-
plex’. First, it will focus on the developments in language that move knowledge from
common-sense to technical understandings. Then it will turn to the move from
language to mathematics in early secondary school and the reorganization of knowl-
edge this entails. Finally, we will consider how the range of images used in physics
complement the technical meanings and shifts in Semantics seen in mathematics.
In sum, we will see that the three semiotic resources work together to organize the
knowledge of physics in a way that allows students to build a scientific gaze.

Knowledge through language

Language and activity
An early stepping stone into physics is exemplified by a book for elementary (pri-
mary) school students, focusing on pushing and pulling (Riley 2001).This book gives
a series of examples of pushing and pulling with very large photos of each action:

When do you push?

You push a pram
You push a swing

When do you pull?

You pull a brush through your hair
You pull a book from your bag.

This very concrete set of examples illustrates a series of physical events. In terms
of the SFL register variable field, these events construe activities (see Doran and
Martin, chapter 5 this volume). Activities such as these offer the child a dynamic
perspective on the world as happenings. To build a picture of the activities involv-
ing pushing and pulling, the text lists a long series of examples with large pictures
of each activity:

A digger pushes rocks into a heap.

A toy car needs a push to make it go.
 Multimodal knowledge 167

Michael pushes the keys to play a tune. [referring to playing the keyboard]
Alex pushes the ball when he kicks it.
…
An engine pulls a train.
A tractor pulls a trailer.
A dog pulls on its lead.
Tim pulls on a jumper.

From the perspective of more advanced physics, this listing of examples appears rel-
atively simple. However, it performs a crucial early role in building an uncommon-
sense understanding of the world. By listing a series of other quite different activities
(kicking, playing with a car, piling rocks in a heap, putting on a jumper, driving a
train) as ‘pushing’ and ‘pulling’, the text emphasizes their similarity and categorizes
them as instances of the more general activity of pushing and pulling. In addition,
the book illustrates that a wide range of items may do the pushing or pulling or be
pushed or pulled. This is an important move into scientific knowledge as it means
first that all examples may be discussed along similar lines, and second that pushes
and pulls can be found almost everywhere. The student is learning that this wide
range of activities, in some sense, all do the same thing.
In terms of LCT Semantics, this weakens the semantic gravity of knowledge. The
individual examples describe specific events that are relatively context-dependent. But
as more examples are listed, the differences between them are generalized in a way
that emphasizes their similarities as pushes and pulls. That is, the text weakens semantic
gravity so that it can eventually discuss pushes and pulls as events of their own (dis-
cussed below). Similarly, in terms of semantic density, by listing this series of examples as
pushes and pulls, it relates these otherwise distinct events together.This adds connections
between their meanings, slightly adding to their complexity, and thus strengthening
their semantic density. Although these are relatively small shifts in semantic gravity and
semantic density in comparison to higher level physics, they are important first steps in
learning this knowledge which will repeated regularly as students move through school.
Looking further into the text, we see it also illustrates that pushing and pulling has
effects on the world (‘^’ indicates that one activity follows or is entailed by the other):

You push on clay to squash it flat.

You push on clay

^ (in order to)
squash it flat
…
You pull an elastic band to stretch it.

You pull an elastic band

^ (in order to)
stretch it
168 Y. J. Doran

…
Paul pushes the pedals on his bicycle.The wheels turn round.

Paul pushes the pedals on his bicycle

^
The wheels turn round.

Here, the text presents squashing, stretching and turning as resulting from the
pushing or pulling in the previous clause. Series of events such as these where
one activity implicates or results from another are a key feature of scientific dis-
course (Wignell et al. 1993). From the perspective of field, they show chains of
activity that are implicated by one another.2 Although at this stage each impli-
cation relation includes only two activities, later on in schooling these series
become long and intricate, involving a range of possible dimensions of condi-
tionality and causation (Rose 1998) and underpinning key genres in science
such as explanations (Unsworth 1997, Martin and Rose 2008). Interpreting
this from the perspective of LCT Semantics once more, these implication rela-
tions work to connect activities together, further strengthening their semantic
density.

Activities and items

By generalizing a range of common-sense activities as ‘pushing’ and ‘pulling’, and
linking them up with other activities to form series of implication activities, this
primary school book has already taken some key steps in the development of an
uncommon-sense scientific field. But the book does not stop here. After introducing
a number of examples, the text nominalizes the activities of pushing and pulling as
pushes and pulls, through a resource known as ‘grammatical metaphor’ (Halliday 1998):

A toy car needs a push to make it go

…
What pushes and pulls can you find as you play?
…
Can you think of three more pulls?
…
Emma gives a weak push to her car.
Ben gives a strong push to his car.

In doing so, the text reconstrues the activity of pushing and pulling as an item push
and pull; it turns an ‘event’ into a ‘thing’.This is arguably one of the most significant
moves in the transition to uncommon-sense knowledge (Halliday and Martin 1993,
Halliday 1998), as it enables much greater possibilities for meaning than if it were
dealing with just an activity, or just an item.
 Multimodal knowledge 169

In this first instance, itemizing these activities enables them to ‘do’ other activities
themselves. For example:

A push can squash something

…
A pull can stretch something

Here, the push and pull have been abstracted from any particular thing doing the
pushing and pulling. This establishes a more generalized series of activities whereby
one activity (the push or pull) leads to another (squashing or stretching). This further
weakens semantic gravity, as the pushes and pulls are no longer tied to any particular
instances:

Push
^
squash something
and
Pull
^
stretch something

This text does not take the next step of also generalizing the squashing and stretch-
ing and putting the whole series of implication in one clause, such as in squashes are
caused by pushes (a specific type of grammatical metaphor known as a logical meta-
phor; Hao 2018, Halliday 1998). But through this initial grammatical metaphor,
the book establishes the basic building blocks that students need for a scientific
construal of experience.
Second, by itemizing pushing and pulling as push and pull, the text enables them
to enter into relations normally reserved for items. This is most clearly seen towards
the end of the book, when it specifies that:

A push is a force
…
A pull is a force

Here, the pushes and pulls are classified as types of force. In doing so, the book estab-
lishes a classification taxonomy. As a wide range of previous events have already been
described as pushes and pulls, this means they are all in turn classified as types of force.
The term force then resonates out to a wide range of common-sense experiences
into a single technical term (Wignell et al. 1993). Again in terms of LCT Semantics,
this strengthens its semantic density, establishing a relatively large series of intercon-
nections emanating from the technical term force (see Maton and Doran 2017b). As
students move through later years of school, force will become increasingly central to
the field of physics as it distils more and more meaning.
170 Y. J. Doran

Language and properties

There is one final step this text takes in construing its scientific view of the world.
This can be seen in the following example where the text notes that pushes and
pulls can be stronger or weaker:

Emma gives a weak push to her car.

Ben gives a strong push to his car.
Ben’s car goes further than Emma’s car.

In the terms of Doran and Martin (Chapter 5 of this volume) model of field, weak
and strong are properties of the pushes. Properties are meanings attached to an item or
an activity that can be graded as more or less. As this instance shows, by attributing
the property of strength to the activities, the two pushes can be ordered as more or
less strong (or strong and weak) – what is referred to as an array of strength (Doran
2018). Importantly, this array of strengths can also have effects. In this case, the dif-
ferences in strength lead to a difference in the distance each of the cars go. One
array, the strength of the push, leads to another array, the distance of the movement,
shown by Ben’s car goes further than Emma’s car.
This potential for giving items and activities a property, ordering these properties
into arrays, and setting up chains of dependencies and causation between differ-
ent arrays, opens a further avenue for building connections between meanings and
strengthening semantic density.

Building the scientific gaze

Although at first glance the meanings being made at this level may appear relatively
simple, this text has introduced each of the main basic building blocks that students
will need throughout their scientific study. From the perspective of field, the text
has explored a dynamic perspective on phenomena as activities of pushing and pull-
ing, complemented with a static perspective of pushes and pulls as items, and added
gradable properties of strength to the pushes and pulls. In addition, the text has
introduced small sets of relations between each of these meanings. It has brought
together activities into relations of implication, where one activity entails another
(such as pushing leading to squashing); it has brought together items into a small clas-
sification taxonomy whereby one set of items are positioned as types of another (i.e.
pushes and pulls as types of force); and it has ordered sets of properties in relation to
each other along an array (where pushes and pulls are ordered as more or less strong).
Table 7.1 synthesizes the field-specific meanings made through language to this
point.
These elements of field underpin the content knowledge developed in science
from primary school onward. As students move through schooling, implication
series, uncommon-sense taxonomies and arrays will be extended and integrated;
and as we will see in the following sections, they will be greatly elaborated through
 Multimodal knowledge 171

TABLE 7.1 Elements of field in a primary school text

Elements of field Examples

activity single activity An engine pulls a train.
momented as an implication activity You push on clay to squash it flat.
item single item What pushes can you find as you play?
items in a (classification) taxonomy A push is a force.
property single property Emma gives a weak push to her car.
properties ordered into an array Ben’s car goes further than Emma’s car.

the use of images and mathematics. However, the introduction of these meanings
does more than simply lay out the specific types of technical meaning needed in
science. They also provide early steps in shifts that underpin the scientific gaze. In
particular, they show a pathway through which students can move from relatively
everyday concrete meanings to more uncommon-sense meanings. More technically
in terms of LCT, they model a pathway from relatively weak semantic density to
slightly stronger semantic density by raising the complexity of meanings, and from
relatively strong semantic gravity to slightly weaker semantic gravity by lessening
the context-dependence of meanings.
In terms of semantic density, we have seen that the text establishes relations
between examples first by likening them together as examples of pushing and
pulling, and second by connecting them into classification taxonomies, arrays
and relations of implication. This takes relatively distinct common-sense mean-
ings and begins to establish more elaborated networks of meaning. The final step
in strengthening semantic density in this text comes through the introduction of
the term force. By classifying pushes and pulls as forces, the text condenses all of
the meanings established through the book into a single term. This positions it as
a key instance of technicality in the field of physics (Halliday and Martin 1993),
which exhibits relatively strong semantic density (Maton and Doran 2017b).
Although the field of physics at this stage is not particularly elaborated, the term
force is the first to position the entire discussion within the technical domain of
physics.
At the same time, the text weakens the semantic gravity of its knowledge. It
generalizes a series of distinct examples as the ‘same’ events – pushing and pulling –
thereby moving them further away from any particular instance. It then reconstrues
these events of pushing and pulling as items: pushes and pulls. This has the effect of
extricating pushes and pulls from any particular things doing the pushing and pulling,
and ensures they can be discussed as items in themselves. Extending this, by tech-
nicalizing pushes and pulls as forces, the text further lessens its context-dependence,
weakening semantic gravity. As Doran and Maton (2021) argue, positioning mean-
ings within more elaborated networks of meaning within a particular field tends to
stabilize them. It makes them more dependent on other meanings in the field than
on any particular context in which it is used, and in doing so weakens its semantic
gravity.
172 Y. J. Doran

FIGURE 7.2 Movements in the semantic plane through the language of pushing and
pulling

These movements in semantic gravity and semantic density are mapped in

Figure 7.2.As the plane indicates, a number of the linguistic resources used have effects
on both semantic gravity and semantic density. This means that for students reading
this text, the scientific knowledge to be learned is two-fold. On the one hand, students
are to learn the specific content meanings associated with pushing, pulling and force –
considered here through the field relations organizing scientific meanings. On
the other hand, underpinning this, students are to learn that scientific knowledge
involves reconstruing everyday situations into less context-dependent and more
complex meanings that can resonate out to a range of other situations. In terms of
LCT, they are learning that the scientific gaze involves being able to weaken seman-
tic gravity and strengthen semantic density across a range of situations.

Knowledge through mathematics

Language occurs across all levels of science. The patterns shown in primary school
where common-sense meanings are reconceptualized into uncommon-sense, tech-
nical meanings recur innumerable times throughout schooling. With each passing
year, the expanse of these meanings increases, with deeper and more integrated rela-
tions in field being built, significantly strengthening semantic density. All the while,
everyday meanings with stronger semantic gravity are reconceptualized in terms of
more generalized meanings with weaker semantic gravity. For students successful in
accessing the knowledge of science, this pattern instils a gaze that values movements
from prosaic code (SG+, SD–) conceptions of the world to rhizomatic code (SG–,
SD+) reconstruals.
 Multimodal knowledge 173

From the perspective of activity, the later years of schooling build much longer
implication series in order to explain and predict phenomena. These are comple-
mented by activities associated with experimental procedures and recounts, which
involve expectancy series. In these activities, relations tend to unfold through time
in terms of what is expected to happen, rather than via definite entailment (Hao
2020, Doran 2018). In terms of taxonomy, larger classification taxonomies are com-
plemented by compositional taxonomies that construe parts to wholes (such as a
nucleus to an atom).
In addition one of the major changes in the move from primary school to sec-
ondary school tends is often an increased emphasis on properties. As mentioned
above, properties construe gradable phenomena – as in the strength of a push. These
properties may be ordered into arrays – as in one push is stronger than another. Such
arrays enable science to construe the world not just in terms of categorical distinc-
tions, but also as involving infinitely variable gradations of more or less. This allows
science to technicalize the ‘fuzziness’ of the physical world, where things are often
distinguished by degree rather than by discrete oppositions.
From secondary school onwards, a regular resource for organizing properties is
mathematical symbolism (Doran 2017). Each symbol in a mathematical statement
realizes a property.3 For example the equation V = IR describes properties of elec-
tricity going through a conductor (known as Ohm’s law). This equation involves
symbols where there may be more or less voltage (V), more or less current (I) or
more or less resistance (R). The equation thus construes variation that occurs in the
world.
Importantly, placing these symbols into an equation indicates that these prop-
erties are not independent. For any particular instance, not every possible value
is available for each symbol. Rather, the equation specifies a set of relationships
between each of these symbols whereby if there is a change in the value of one
symbol, it will likely affect all the others. For example, in V = IR, if I (the current)
was to increase, then either V (voltage) would also need to increase, or R (resistance)
would need to decrease, or both. Similarly, if V were to increase, then either I or R
or both would also need to increase; and if R were to increase then either V would
increase or I decrease or both (see Doran 2018: 88–96). As far as field is concerned,
Doran and Martin (chapter 5, this volume) term these relations between proper-
ties interdependency relations. These interdependencies greatly expand the meaning
potential of science. In the first instance, by involving properties, they allow scien-
tific fields to construe degrees of gradation in ways that activity and taxonomy can-
not. Secondly, by specifying definite interdependencies between these properties,
they offer a means for describing the effects of a change in any particular property.
Put another way, through the equation V = IR, physics is able to account for there
being more or less voltage, more or less current and more or less resistance, and it
is also able to precisely describe the effect of a change in voltage on the current
or resistance of a conductor. By virtue of establishing interconnections between
properties, mathematics also contributes to strengthening the semantic density of
knowledge.
174 Y. J. Doran

Mathematical equations can be organized into two key genres that deal with dif-
ferent aspects of scientific knowledge. These genres are known as derivations, which
develop new relations between symbols, and quantifications which work to quantify
symbols (Doran 2017, 2018). Derivations are concerned with deriving new equa-
tions from previously known ones. In New South Wales, Australia, they tend to
appear in physics in the later years of secondary school. For example, in the follow-
ing derivation from a secondary school student exam response, the student derives a
new equation for W (work) (crudely, the energy associated with a force) in response
to a question asking about the work needed to move a satellite from earth to an
orbit altitude. Note that in the final equation, the W on the left side of the = is
elided by convention in mathematical texts (Doran 2018: 69–72):

work  GPE  GPE f  GPEi

Gm1m2  Gm1m2 
W    
rf  ri 
1 1 
 G m1 m 2   
 ri r f 
The student begins by stating an equation that is assumed at this level, namely
that work is equal to the change in gravitational potential energy (∆GPE, meaning
roughly the change in energy associated with moving between different positions in
the gravitational field of earth). From this initial relation, the student then inserts sets
of other known relations – such as the fact that the change in gravitational potential
energy (∆GPE) is equal to the difference between the final gravitational potential
energy (GPEf) and the initial gravitational potential energy (GPEi), and that gravita-
tional potential energy itself is equal to Gm1m2 (glossed as the gravitational constant
r
G, multiplied by the mass of the satellite (m1), multiplied by the mass of the earth
(m2), all divided by the distance between the satellite and the centre of the earth (r)).
In the final line, the student rearranges the equation to simplify it for calculations
further along in the text. This derivation brings together a complex network of
technical meanings and relates them precisely in terms of their interdependencies.
The specific relations in this derivation are not important for our discussion.
What is important is that through the manipulation of mathematical symbolism the
student is able to move from one set of relations, work = ∆GPE, to a new set of
1 1 
relations work  Gm1m2    . In effect, the student has established new sets of
 ri rf 
interdependencies in the field; they have established relations not previously assumed
(at this level, or at least logogenetically in this text), and in doing so, they have more
tightly integrated the technical meanings of physics. In terms of LCT Semantics,
derivations function to strengthen the semantic density of the discipline. They
enable new relations to be established, which integrate an increasingly wide range
of technical meanings.This means that any particular meaning specified in the deri-
vation can now resonate out to all other meanings mentioned. But more than this,
 Multimodal knowledge 175

each term can be linked to other, unstated technical meanings by virtue of their
relation any other symbol that is mentioned.This is a powerful means of strengthen-
ing semantic density, and one that appears to be increasingly relied upon as students
move into higher levels of physics.
A little before derivations are introduced, students become familiar with another
genre of mathematics known as a quantification. Where derivations build new, pre-
viously unknown relations, quantifications measure a particular property by numer-
ically quantifying it. In terms of field, Doran and Martin (Chapter 5 of this volume)
refer to the numerical measurement of properties as gauging. An example of this
can be seen in the student exam response immediately following the derivation
discussed above. Drawing on the final result of the derivation, the student inserts
a set of numbers given in the question to calculate the work done in a particular
situation:

1 1
W   Gm1m2  r  
rf 
 i

 1 1 
 6.67  10 11  1428.57  5.97  1024   
 6 380 000 6380000  355000 
 4699722327
 4.70  109 J  3 sig fig 

Here, through numerous steps (detailed in Doran 2018), the student concludes that
the work done is approximately 4.70 × 109 Joules (~ 4.7 billion Joules).What is sig-
nificant about this text is that the quantification has enabled the student to relatively
precisely measure a specific instance of a property. It has, in other words, taken the
relatively generalized relations given in the symbolic equations and applied them to
a specific situation. The quantification genre has moved the text from the relatively
weak semantic gravity descriptions of general relations between properties in phys-
ics, to relatively strong semantic gravity measurements of an individual property tied
to a very specific situation.
The two mathematical genres, derivation and quantification, thus enable students
to strengthen semantic density and strengthen semantic density – to build theory
and to link this with the empirical world. Symbolic equations such as V = IR are
not tied to any particular instance. Rather, they describe abstracted relations that
encapsulate an innumerable number of instances. In this sense, they maintain rela-
tively weak semantic gravity. At the same time, they realize a definite set of technical
relations between highly technical symbols and so involve relatively strong semantic
density. Symbolic equations can thus be positioned in the rhizomatic code (SG–,
SD+). Derivations progressively pull together more and more relations between an
ever-widening number of technical symbols, but they do so in a way that is not
heavily tied to any particular context or situation – they remain relatively general-
ized. In this sense, derivations strengthen semantic density of the field, while having
176 Y. J. Doran

little effect on semantic gravity.4 Derivations can therefore be described as strength-

ening semantic density within the rhizomatic code (moving rightwards on the
semantic plane). On the other hand, quantifications move texts from relatively weak
semantic gravity relations that describe a wide range of phenomena to
measurements tied specifically to a single instance. This means they significantly
strengthen semantic gravity of the text. At the same time, they do this without
1 1 
decreasing the complexity of this knowledge. Moving from W   Gm1m2    to
 ri rf 
W  4.70  109 J  3 sig fig  does not move us into the realm of everyday, com-
mon-sense meanings. The meanings maintain relatively strong semantic density by
virtue of the fact that the meanings being quantified (i.e. the units of measurement
and the organization of this measurement) are all highly technical and resonate out
to the much wider disciplinary knowledge of physics. But they do so while signifi-
cantly strengthening semantic gravity. In this sense, quantifications enable mathe-
matics to move from rhizomatic codes into worldly codes – to move from theoretical to
empirical (rather than from theory to the ‘everyday’). This shows that the descrip-
tion of physics as a relatively ‘abstract’ discipline (e.g. Kolb 1981) misses a key com-
ponent of its knowledge, namely that it enables increasing complexity of knowledge
while shifting from ‘abstract’ theory to grounded empirical measurements. These
movements in semantic gravity and semantic density through mathematics are
shown in Figure 7.3.
For students learning this knowledge, the use of mathematics complements
the trained gaze developed in language. In addition to the movements shown in
language from a prosaic code (SG+, SD–) to a rhizomatic code (SG–, SD+), the
discipline emphasizes the importance of both continually strengthening semantic
density within the rhizomatic code and being able to reach down to the empirical

FIGURE 7.3 Movements in the semantic plane through mathematics

 Multimodal knowledge 177

(worldly code or SG+, SD+), without moving into the common-sense. This has
two effects. First, it emphasizes regular movement: the scientific gaze it is exhibiting
is not one of static technicality or abstraction but of constant movement between
the everyday, the theoretical and the empirical. Second, it emphasizes the utility of
theoretical knowledge in terms of its ability to reach down and make precise predi-
cations about and descriptions of the empirical world. Such movements are vital for
conceptualizing new situations from a scientific viewpoint.
This raises the question of how empirical meanings (SG+, SD+) can reach back
to the theoretical meanings (SG–, SD+). We have seen that mathematics enables
students to reach towards empirical description by quantifying individual properties
but we have not yet seen how physics can use empirical measurements to change
theory. In terms of semantic gravity, mathematics enables physics to strengthen
semantic gravity (and to move from rhizomatic codes to worldly codes) but it offers
no way at this stage to weaken semantic gravity again (to move back from worldly
codes to rhizomatic codes). To deal with this issue, physics brings in images –
specifically graphs.

Knowledge through image

Images occur throughout science (Parodi 2012, Lemke 1998). They enable a large
number of meanings to be brought together and presented in a single, synoptic snap-
shot (Doran 2019, Martin et al. 2021). We can see this in the diagram in Figure 7.4,
from the same student exam response as the mathematical examples above. In this
question, the student was asked to Draw a labelled diagram of the vacuum tube used by
Thomson to calculate the q/m ratio of electrons.
From the perspective of field, the diagram depicts a compositional taxonomy,
in which each component is positioned as a part of the apparatus (the whole).
However, the diagram does not just present the various parts without any sense of

FIGURE 7.4 Diagram of an experimental apparatus

178 Y. J. Doran

how they come together to make the whole, what Kress and van Leeuwen (2006)
call an unstructured analytical image. Rather, the components are spatially arranged
in relation to one another – known as a spatially structured analytical image. In terms
of the field relations of Doran and Martin (chapter 5, this volume), this means each
component of the image realizes two distinct meanings: its part within a composi-
tional taxonomy and its spatial position in relation to all the other parts in a spatial
array.5 The image also depicts activity through the dotted line labelled centre read
cathode rays. Although there is no arrow or other overt indicator of a vector, the lin-
ear arrangement of the terms cathode and anode – as technical meanings in this field –
are enough for an informed reader to understand the movement of the cathode rays
is from left to right.
This diagram thus integrates three distinct dimensions of field: a compositional
taxonomy incorporating each part of the apparatus; a spatial array that arranges the
parts of the apparatus in relation to one another; and an activity involving a cathode
ray that moves through the apparatus. It is difficult for either language or mathemat-
ics alone to bring together this wide range of meaning in one snapshot. Indeed, this
is one of the key affordances of images: they are able to pull together a large number
of complementary meanings in a synoptic ‘eyeful’. Diagrams in this sense are a key
resource for packaging up multiple meanings to be read and viewed together. In
terms of LCT, they offer the potential for significantly stronger semantic density in
a way that complements that of language and mathematics.
As students move further into physics study, these diagrams are typically comple-
mented by graphs. Graphs offer a relatively unique affordance in physics in compari-
son to the other resources we have seen so far and appear to be especially prevalent
as students engage more deeply with experimental investigations. In terms of their
knowledge-building potential, graphs complement mathematical quantifications by
enabling a more flexible interaction between empirical measurements and more
abstract theory. We can see this by exploring Figure 7.5. This figure displays a ques-
tion from the same high-school student exam response as the diagram and the
mathematical examples above. The prompt includes a table of measurements for
Resistance and Temperature of a wire.The question asks the student to: (1) plot these
values on the graph (shown by the ‘x’s on the graph); (2) draw a line of best fit (that
generalizes across the ‘x’s, shown by the line labelled LOBF); and (3) use this line
of best fit to estimate the resistance of the wire at 30°C (shown below the graph).
From the perspective of field, the graph is organized around two properties pre-
sented on the axes (Resistance and Temperature) which are given measured values in
the table above the graph.The first task of the student is to take these measurements
in the table and plot them on the graph (shown by the ‘x’s on the graph). From the
perspective of LCT Semantics, these plotted measurements show relatively strong
semantic gravity. Each point describes a very precise instance that cannot be gen-
eralized to any other. But, like the numerical measurements in quantifications, they
also exhibit relatively strong semantic density as they signify the intersection of two
technical properties and are measured with the particular units of these properties
(Ohms for Resistance, Degrees Celsius for Temperature).
 Multimodal knowledge 179

FIGURE 7.5 Graph and questions in a senior secondary school examination

The next step for the student is to draw a ‘line of best fit’. This involves general-
izing across the plotted points to draw a line that ‘best fits’ each of the values. In this
case, this is a straight line drawn from about six degrees Celsius on the tempera-
ture scale to the left, up to about 37 degrees in the top right. Like mathematical
equations, this line realizes a general interdependency between properties. It shows
that when resistance increases, so does temperature (and vice versa). Similarly, this
line has significantly weaker semantic gravity than the individual plotted points.
Rather than being tied to any actual instance, the line develops a more generalized
description. Indeed, the line does not intersect precisely with any of the measured
180 Y. J. Doran

points. Drawing a line of best fit weakens semantic gravity such that the relations it
describes are not dependent on any particular instance or context.
By shifting between measured points and generalized lines of best fit, graphs
can be used to shift from stronger semantic gravity to weaker semantic gravity
while maintain relatively strong semantic density. They thus complement math-
ematical quantifications by allowing for students to shift from the worldly code
back into the rhizomatic code. They enable a shift from the ‘empirical’ back up to
the ‘theoretical’.
An important feature of images is that they do not insist on a definite reading
path (Bateman 2014). In the case of graphs, this means students do not just have to
move from measured points to generalized lines (from stronger to weaker semantic
gravity). They can also use the generalized line to predict specific instances. Indeed,
this is the third task the student is asked to do, when question (a) says to estimate
the electrical resistance of wire at 30° C. To answer this question, the student would
have to find where the line of best fit is at 30° C, and ‘read off ’ the value of the
Resistance at this point. In this example, the student has done this by drawing two
dotted lines, one vertical line at 30° Celsius, and then one horizontal line at around
.1285 Ohms Resistance. Below the graph, the student then rounds this up to 0.129
Ohms. In terms of semantic gravity, this displays a movement back from weaker
semantic gravity to stronger semantic gravity. The student begins with the general-
ized description shown by the line of best fit and uses this to make a prediction of
a specific measurement. Although this is an estimation, the prediction is strongly
dependent on a context where the wire is 30°C, and so it displays relatively stronger
semantic gravity.
This example illustrates how graphs enable movements back and forth between
stronger semantic gravity and weaker semantic gravity. This affordance makes it
a key tool for organizing the knowledge of physics (and many other disciplines).
Figure 7.6 illustrates this movement. Where mathematical derivations enable move-
ments within the rhizomatic code, and quantifications organize one-way trips from
the rhizomatic code to the worldly code, graphs give opportunities for shifts back
and forth – what Maton and Howard (2018) term ‘return trips’.

Multimodal knowledge
We can now return to the original question of this chapter: what knowledge do
students need to learn in order to be successful in science? In the first instance,
they need to understand the technical content meanings of the discipline. From the
perspective of field, this means marshalling various items, activities, properties, and
their respective relations when necessary. As this chapter has shown, this necessarily
implicates multiple semiotic resources. Mathematics is able to bring together rich
interdependencies between properties and measure these quantitatively; graphs are
able to establish multiple arrays upon which measurements can be ordered; and
diagrams and language can integrate a wide range of different types of meaning,
 Multimodal knowledge 181

FIGURE 7.6 Movements in the semantic plane through graphs

including taxonomy, activity and property. The content meanings of science are
organized through the multimodal resources of science.
In the second instance, students need to learn a particular way of seeing the
world. This way of seeing the world does not simply involve technical discus-
sions of abstract phenomena. More importantly, it involves moving between the
‘common-sense’ and the ‘technical’, the ‘theoretical’ and the ‘empirical’. In terms
of LCT Semantics, this means being able to move between a prosaic code (SG+,
SD–) and a rhizomatic code (SG–, SD+), and between a rhizomatic code and a
worldly code (SG+, SD+). Or, to put this another way, students need to be able to
strengthen and weaken semantic gravity and semantic density independently and
together. They need a way of seeing the world from various perspectives and shift-
ing between these perspectives as needs arise. As mentioned above, what is at stake
here is a trained scientific gaze. Being able to read and write particular the semiotic
resources and field-specific meanings of science is not enough. Students need to
be able to select and arrange these resources appropriately in potentially new sce-
narios. This means they need a way of understanding the organizing principles of
the knowledge in any particular situation. For physics (and science more broadly),
this necessarily involves using multiple semiotic resources. Language enables the
common-sense to be repackaged as uncommon-sense (and vice versa); mathemat-
ics enables a steady increase in the complexity of meaning while also moving from
the theoretical to the empirical; and images enable an integration of a wide range
of meanings to increase complexity and movement back and forth between the
empirical to the theoretical. In all, students must learn that scientific knowledge is
multimodal.
182 Y. J. Doran

Notes
1 The semantic density and semantic gravity explored in this chapter are those associated
with ‘content’ meanings – technical procedures, logical reasonings, taxonomies, etc.This is
known as epistemic–semantic density and epistemic–semantic gravity (Maton 2014, Maton and
Doran 2017b).This contrasts with axiological–semantic density and axiological–semantic gravity
that explore meanings associated with values, political judgements, morals, aesthetics, etc.
2 These were previously called implication sequences (e.g. Martin 1992). However, follow-
ing Hao (2015, 2020) ‘sequence’ is reserved for relations between figures in discourse
semantics, and ‘implication’ will be reserved for one type of relation between activities in
a field. The relationship between a larger activity and the smaller activities that constitute
it is known as momenting (discussed in Doran and Martin, Chapter 5 of this volume).
3 More precisely, each symbol realizes an itemized property – a property reconstrued as an
item. This is because they may be graded (like properties), but they can also be classi-
fied (like items). For example, one may have more or less E (energy) (interpreted as a
property), but one may also distinguish between different types of energy, such as initial
energy, Ei, final energy Ef and average energy Eav (interpreted as items in a classification
taxonomy). See Doran and Martin, chapter 5 of this volume for discussion.
4 It is possible that particular derivations may strengthen or weaken semantic gravity by
either deriving a set of relations only applicable to more precise situations or conversely
deriving more generalized relations to cover a wider set of phenomena. This would be a
function of the particular derivation under study rather than all derivations in general.
5 More precisely, each element realizes a position along two spatial arrays, one vertical and
the other horizontal.

References
Bateman, J. (2014) Text and Image: A Critical Introduction to the Visual/Verbal Divide, London:
Routledge.
Biglan, A. (1973) ‘The characteristics of subject matter in different academic areas’, Journal of
Applied Psychology, 57(3): 195–203.
Christie, F. and Martin, J. R. (1997) (eds) Genre and Institutions: Social Processes in the Workplace
and School, London: Continuum.
Christie, F. (2002) Classroom Discourse Analysis: A functional perspective, London: Continuum.
Doran, Y. J. (2017) ‘The role of mathematics in physics: Building knowledge and describing
the empirical world’ Onomázein Special Issue, March: 209–26.
Doran, Y. J. (2018) The Discourse of Physics: Building Knowledge through Language, Mathematics
and Image, London: Routledge.
Doran, Y. J. (2019) ‘Building knowledge through images in physics’, Visual Communication,
18(2), 251–77.
Doran,Y. J. and Maton, K. (2021) ‘Context-dependence: Epistemic-semantic gravity in English dis-
course’, LCT Centre Occasional Paper 2
Halliday, M. A. K. (1998) ‘Things and relations: Regrammaticizing experience as technical
knowledge’, in J. R. Martin & R.Veel (eds) Reading Science, London: Routledge, 185–235.
Halliday, M. A. K. (2004) Language of Science, London: Continuum.
Halliday, M. A. K. and Martin, J. R. (1993) Writing Science: Literacy and Discursive Power,
London: Falmer.
Hao, J. (2015) ‘Construing biology: An ideational perspective’, Unpublished PhD thesis,
Department of Linguistics, University of Sydney.
 Multimodal knowledge 183

Hao, J. (2018) ‘Reconsidering “cause inside the clause” in scientific discourse – from a
discourse semantic perspective in systemic functional linguistics’, Text and Talk, 38(5):
525–50.
Hao, J. (2020) Analysing Scientific Discourse from a Systemic Functional Linguistic Perspective,
London: Routledge.
Kress, G., Jewitt, C., Ogborn, J. and Charalampos, T. (2001) Multimodal Teaching and Learning:
The Rhetorics of the Science Classroom, London: Continuum.
Kress, G. and van Leeuwen, T. (2006) Reading Images: The Grammar of Visual Design, London:
Routledge.
Kolb, D. A. (1981) ‘Learning styles and disciplinary differences’, in A. Chickering (Ed.) The
Modern American College, San Francisco: Jossey-Bass, 232–55.
Lemke, J. L. (1990) Talking Science: Language, Learning and Values, Norwood, NJ: Ablex.
Lemke, J. L. (1998) ‘Multiplying meaning: Visual and verbal semiotics in scientific text’, in J.
R. Martin and R.Veel (eds) Reading Science, London: Routledge, 87–113.
Lemke, J. L. (2003) ‘Mathematics in the middle: Measure, picture, gesture, sign, and word’, in
M. Anderson, A. Saenz-Ludlow, S. Zellweger and V. Cifarelli (eds) Educational Perspectives
on Mathematics as Semiosis, Ottawa: Legas, 215–34.
Martin, J. R. (1985) Factual Writing: Exploring and Challenging Social Reality, Geelong: Deakin
University Press.
Martin, J. R. (1992) English Text: System and Structure, Amsterdam: John Benjamins
Martin, J. R. and Doran,Y. J. (2015) (eds) Language in Education, London: Routledge.
Martin, J. R., Maton, K. and Doran, Y. J. (2020) (eds) Accessing Academic Discourse: Systemic
Functional Linguistics and Legitimation Code Theory. London: Routledge.
Martin, J. R. and Rose, D. (2008) Genre Relations: Mapping Culture, London: Equinox.
Martin, J. R. and Veel, R. (1998) (eds) Reading Science: Critical and Functional Perspectives on
Discourses of Science, London: Routledge.
Martin, J. R., Unsworth, L. and Rose, D. (2021) ‘Condensing meaning: Imagic aggregations in
secondary school science’, in G. Parodi (ed.) Multimodality, London: Bloomsbury.
Maton, K. (2014) Knowledge and Knowers: Towards a realist sociology of education, London:
Routledge.
Maton, K. (2016) ‘Legitimation Code Theory: Building knowledge about knowledge-
building’, in K. Maton, S. Hood and S. Shay (eds) Knowledge-Building, London: Routledge,
1–23.
Maton, K. (2020) ‘Semantic waves: Context, complexity and academic discourse’, in Martin,
J. R., Maton, K. and Doran, Y. J. (eds) Accessing Academic Discourse, London, Routledge,
59–85.
Maton, K. and Doran,Y. J. (2017a) SFL and code theory, in Bartlett, T. and O’Grady, G. (eds)
Routledge Systemic Functional Linguistic Handbook, London, Routledge, 605–18.
Maton, K. and Doran,Y. J. (2017b) ‘Semantic density: A translation device for revealing com-
plexity of knowledge practices in discourse, part 1 – wording’, Onomázein Special Issue,
March: 46–76.
Maton, K. and Howard, S. K. (2018) ‘Taking autonomy tours: A key to integrative knowledge-
building’, LCT Centre Occasional Paper 1: 1–35.
O’Halloran, K. L. (2005) Mathematical Discourse: Language, Symbolism and Visual Images,
London: Continuum.
Parodi, G. (2012) ‘University genres and multisemiotic forms of knowledge: Accessing spe-
cialized knowledge through disciplinarity’, Fórum Linguístico, 9(4): 259–82.
Riley, P. (2001) Ways into Science: Push and Pull, London-Sydney: Franklin Watts.
Rose, D. (1998) ‘Science discourse and industrial hierarchy’, in J. R. Martin and R.Veel (eds)
Reading Science, London: Routledge, 236–65.
184 Y. J. Doran

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, London: Routledge, 91–107.
Rose, D. (2014) ‘Analysing pedagogic discourse: An approach from genre and register’
Functional Linguistics, 2(11): 1–32.
Rose, D. (2020) ‘Building a pedagogic metalanguage I: Curriculum genres’, in J. R. Martin,
K. Maton and Y. J. Doran (eds) Accessing Academic Discourse, London: Routledge. 236–67.
Rose, D. and Martin, J. R. (2012) Learning to Write, Reading to Learn: Genre, Knowledge and
Pedagogy in the Sydney School, London: Equinox.
Rose, D., McInnes, D. and Korner, H. (1992) Scientific Literacy (Literacy in Industry Research
Report: Stage 1), Sydney: Metropolitan East Disadvantaged Schools Program.
Unsworth, L. (1997) ‘“Sound” explanations in school science: A functional linguistic perspec-
tive on effective apprenticing texts’, Linguistics and Education, 9(2): 199–226.
Unsworth, L. (2001a) ‘Evaluating the language of different types of explanations in junior
high school science texts’, International Journal of Science Education, 23(6): 585–609.
Unsworth, L. (2001b) Teaching multiliteracies across the curriculum: Changing contexts of text and
image in classroom practice, Buckingham: Open University Press.
Wignell, P., Martin, J. R. and Eggins, S. (1993) ‘The discourse of geography: Ordering and
explaining the experiential world’, in M. A. K. Halliday and J. R. Martin (eds) Writing
Science, London: Falmer, 151–83.
PART III

Pedagogy in science
education
8
WIDENING ACCESS IN SCIENCE
Developing both knowledge and knowers

Karen Ellery

Introduction
Historically universities served the interests of an élite few from the middle and
upper classes of society. Thanks to the global economic and social justice drivers,
massification of universities in the last century has resulted in major transforma-
tions across the higher education sector. A significant feature of this transformation
has been greater diversity of the student body in terms of social and educational
backgrounds. However, for many universities the fundamental ‘culture’ of teaching,
learning and research has remained remarkably constant (Longden 2006), resulting
in an inequitable system that favours certain social groups over others (Archer et al.
2003, Arum et al. 2012). Physical access without concomitant success is meaningless,
and the emotional (Pym and Kapp 2011) and the financial toll (Council on Higher
Education Report 2013) on individuals and families makes equity of outcomes an
urgent moral imperative. The broader-scale implications of perpetuating a system
with a grossly unequal distribution of power and access to educational, social and
economic resources in any society add impetus to this imperative.
This chapter is particularly interested in access for success in the sciences. New
scientific knowledge is constantly being developed and a typical pedagogic response
in the higher education context has been to maintain the focus on knowledge but
at the same time increase the volume and pacing of curriculum material. Muller
(2015: 410) suggests this favours students from privileged educational backgrounds
as they are ‘better equipped by virtue of being educated in cognitively rich envi-
ronments by better qualified teachers to respond to the increased volume of novel
material’. There are, however, increasing calls from knowledge-based fields, such as
the sciences and engineering, to consider ontological aspects of student ‘becom-
ing’ and ‘being’ and agency in promoting learning (Barnett 2007, Dall’Alba and
Barnacle 2007). In an in-depth study of student learning in engineering, Case
(2013) makes an argument for student agency as a central part of engagement
188 Karen Ellery

with, and learning of, requisite knowledge. As such, a student’s social, cultural and
educational background would strongly influence her/his engagement with a cur-
riculum, and Case suggests that while engineering (or science) curricula must focus
on the knowledge, there needs to be better accommodation of student identity
development and agency. In short, she calls for more focus on both the knowledge
and the knower.
The main question that this chapter addresses is how access for success can be
enabled or constrained in a science curriculum.1 To explore these issues, access for
success is encapsulated in the term ‘epistemological access’ which Morrow (2009:
77) defines as learning ‘how to become a participant in an academic practice’ and
that this requires learning ‘the intrinsic disciplines and constitutive standards of
the practice’. By invoking ‘disciplines’ and ‘standards’ of practice he is suggesting
an underpinning framework, but his work does not provide the necessary analyti-
cal tools to unpack them. For this I turn to Legitimation Code Theory, which
is a conceptual framework that provides a means for identifying the organizing
principles that constitute curriculum practices. This chapter draws on a single
case study of a higher education science access course that has been designed to
accommodate students from traditionally less privileged education backgrounds.
The findings suggest that the curriculum of the course involves two main bases for
achievement, one which emphasizes the possession of knowledge, skills or proce-
dures and another which emphasizes the need to be a particular kind of knower.
Moreover, the course requires students to be the right kind of knower in order
to then access the right kinds of knowledge. These findings are used to develop a
conceptual model of epistemological access that could inform curriculum trans-
formation processes in the sciences in this age of diversity and difference in higher
education.

Conceptual and analytical framework

Legitimation Code Theory (LCT) is a conceptual framework that allows ‘knowl-
edge practices to be seen, their organizing principles to be conceptualized, and
their effects to be explored’ (Maton 2014: 3). The notion of legitimation is central
to LCT: when actors engage in a practice, they are making a claim of legitimacy
for the basis of that practice. What actors say or do can therefore be described as
‘languages of legitimation’, and the principles that underpin those practices are
conceptualized as ‘legitimation codes’ (Maton 2016: 10). These codes represent
the means by which success can be achieved in the practice. In this chapter, the
knowledge practice that is the focus of our attention is the curriculum of an access
course, and the legitimation codes that will help explore the role of constructions
of knowledge and knowers in this curriculum are from the LCT dimension of
Specialization (Maton 2014).
Specialization explores a set of organizing principles underlying practices as
‘specialization codes’. These concepts begin from the simple point that prac-
tices are ‘oriented towards something and by someone’ (Maton 2016: 12), which
 Widening access in science 189

highlights an analytic distinction: epistemic relations (ER) between practices and

their object (that part of the world towards which they are oriented); and social
relations (SR) between practices and their subject, author or actor (who is enacting
the practices). The relative strengths of epistemic relations and social relations can
vary from stronger (+) to weaker (–) along a continuum of strengths. Both refer to
the basis of knowledge practices. For example, if your knowledge of environmental
conditions (soil-type, nutrients, temperature, light conditions) promoting optimal
growth of a particular plant was based on scientific knowledge, skills and proce-
dures, it would represent stronger epistemic relations (ER+). In contrast, under-
standing that the plant grows best against the sunny wall of your house because
it did when you tried it would represent relatively weaker epistemic relations
(ER–): there is relatively less emphasis here on specialist knowledge, skills or pro-
cedures. Conversely, that experiential basis for deciding where to place your plant
represents stronger social relations (SR+): it emphasizes your own experiences
and opinions. The specialist scientific knowledge of optimal growth conditions
downplays such personal opinions and experiences, representing weaker social
relations (SR–). In identifying the relative strengths of epistemic relations and
social relations, the basis upon which success in the practice is achieved is relevant
as opposed to the sometimes more obvious focus. For example, student activities
in a plant growth experiment may focus on recording plant size, leaf colour and
flower production, but the basis for achievement in the final report may instead be
the correct presentation of data.
By plotting the relative strengths of epistemic relations and social relations on a
two-dimensional plane as in Figure 8.1, four principal codes are identified: knowl-
edge codes, élite codes, knower codes and relativist codes (Maton 2014). When a
practice has relatively strong epistemic relations and relatively weak social relations,

FIGURE 8.1 The specialization plane (Maton 2016: 12)

190 Karen Ellery

it legitimates a knowledge code (ER+, SR–). That is, the basis for specialization and
legitimacy in the practice depends on possession of determinate and specialized
knowledge and practices (ER+), and the attributes of knowers, such as their dis-
position and ‘gaze’, are downplayed (SR–). Science and science disciplines, such
as physics and zoology, are typically considered to be dominated by knowledge
codes. Educational practices characterized by a knower code (ER–, SR+) have rela-
tively weak epistemic relations, but relatively strong social relations. Here, special-
ized knowledge is downplayed (ER–) and legitimacy is based on the disposition or
‘gaze’ of the knower (SR+). The humanities and many of their disciplines, such as
sociology and anthropology, are often considered to be dominated by knower codes.
Educational practices characterized by an élite code (ER+, SR+) have relatively
stronger epistemic relations and relatively stronger social relations. In other words,
both specialist knowledge and knower dispositions are equally valued. Professional
levels of classical music, for example, require both depth of knowledge and personal
talent – an élite code (Lamont and Maton 2008). Finally, practices characterized by
a relativist code (ER–, SR–) have relatively weak epistemic relations and relatively
weak social relations. Legitimacy here requires neither specialist knowledge nor
knower dispositions.This is unlikely to be illustrated by subject areas in a specialized
context such as education but can be found during periods of unfocused class-
room activity or potentially during periods of ‘brainstorming’ when all ideas are
welcomed.
It is worth emphasizing that these examples are simply illustrative: most sets of
practices are likely to involve more than one code, and the particular code char-
acterizing science, for example, depends on the context. While science is typically
considered to be dominated by knowledge codes, in a specific classroom it may
involve a range of codes, of which the dominant code may not be a knowledge
code. Maton and Howard (2016), for example, found that a significant proportion
of students’ perceptions of secondary school science in New South Wales, Australia,
exhibited a relativist code. The point is that these concepts enable us to explore
these issues, rather than allocate subjects in advance to particular labels.
The specialization code exhibited by a curriculum may be expressed in explicit
ways and therefore be obvious to participants, but may also be tacit, especially when
underpinned by poorly articulated norms and values. Building on Bernstein’s dis-
tinction between ‘recognition’ and ‘realization’ (2000), LCT suggests that for a stu-
dent to be successful in a context they need to not only recognize its dominant
code but must also be able to realize or enact practices that match that code. LCT
therefore not only allows for identification and explicit articulation of the codes, but
also opens up opportunity to see possible points of conflict. For example, if a science
curriculum requires the use of empirical data to develop an argument, a knowledge
code is being legitimated. However, if instead personal opinion is provided, the
student is invoking a knower code. The resultant ‘code clash’ (Lamont and Maton
2008) can help explain poor student achievement. This study identifies specializa-
tion codes of a science course curriculum and invokes codes clashes to explain dif-
ficulties students have in their studies.
 Widening access in science 191

Context of study
Students’ backgrounds
The study was located within a South African higher education science access
programme. In order to enter any of the mainstream disciplines in science, stu-
dents in the programme are required to pass three year-long foundation courses:
Mathematical Foundations, Computer Skills for Science, and Introduction to
Science Concepts and Methods (henceforth ‘ISCM’). ISCM is the focus of this
study. Because of a range of entrance selection criteria, students in the programme
have achieved only slightly lower school marks than those entering mainstream
directly, seldom have their home language as English, regularly require government
funding to help pay for their tertiary studies, and attended poorly resourced and
performing schools. They enter university with a range of literacy, numeracy and
learning practices that are very different from the open-minded, analytical, creative,
critical and independent approaches required when working with the high levels of
conceptual knowledge in the sciences in higher education.

Study site
The purpose of ISCM, a multidisciplinary, integrated science foundation course,
is to enable epistemological access to mainstream science by introducing students
to scientific concepts, methods and literacies within the context of specific sci-
ence disciplines, as well as through developing academic learning competencies and
practices (Science Extended Studies Programme Review Report 2011). Currently
ISCM draws on physics, chemistry, life sciences (human kinetics and ergonomics)
and earth sciences (geology) for disciplinary input by mainstream staff from each
of their respective disciplines. The focus in the two disciplinary lectures and one
practical session per week is therefore on developing understanding of disciplinary
concepts. Two permanently employed access programme lecturers facilitate the rest
of the ten weekly contact periods, in what are termed ‘literacies’ pedagogic inter-
actions.2 While there is much overlap in the work of these two staff members, the
‘language-related literacies’ facilitator focuses mainly on reading and writing in the
sciences. In contrast, the scientific-related literacies facilitator (myself) focuses on
building foundational science concepts (such as spatial and temporal scales, hierar-
chies and connections, diversity) and procedures (such as working accurately and
precisely, solving problems, working with empirical data) and considering how sci-
entific knowledge is constructed. At least two of the weekly pedagogic interac-
tions, particularly in the first semester, focus primarily on student learning in which
strategies and approaches to learning are explicitly voiced, and appropriate practices
related to student learning are modelled. As a coordinator of the course, I attend
and observe or participate in all disciplinary interactions and some of the literacies
interaction taught by my colleagues.
ISCM therefore has two main themes in its teaching: scientific and disciplinary
concepts and methods; and student learning competencies. While these themes
192 Karen Ellery

are often addressed separately in particular pedagogic interactions, the curriculum

has been developed as a coherent whole and there is much integration across
themes.

Methodology
The overall approach in the study was a single, in-depth case study. In order to sur-
face the specialization codes that underpin the ISCM curriculum, data for epistemic
relations and social relations analysis were obtained from course documents (hand-
outs, resource materials, assessment tasks, student answers in assessment tasks), infor-
mal observations of pedagogic interactions, and interviews with four disciplinary
and one literacies lecturer in which their perceptions of the purpose of ISCM were
discussed. In order to bridge the gap between the empirical data and the abstract
specialization concepts, a ‘translation device’ (Maton and Chen 2016) was devel-
oped through iterative movement between theory and data. In the original analysis,
the strengths of epistemic relations and of social relations were examined indepen-
dently of each other (Ellery 2017a, 2018) and thereafter specialization codes were
identified (Ellery 2017b).What follows is the final code categorizations. In addition,
in order to understand student response to codes, their answers in assessment tasks
were analyzed and follow-up interviews with 17 volunteers of 47 students probed
their approach to their studies (Ellery 2016, 2017b).

Specialization codes of ISCM

Because ISCM has an explicit dual focus of introducing students to disciplinary and
scientific concepts, methods and literacies, as well as developing academic learn-
ing competencies and practices, two distinct codes were recognized: a science-related
knowledge code and an academic practices-related knower code. These are now discussed
in turn.

Science-related knowledge code

Since science and science disciplines are often associated with knowledge codes,
it is not unexpected that such a code is highly visible in the ISCM curriculum.
However, because the nature of the work done by the disciplinary and literacies
lecturers in the course is so different, and because the strengths of epistemic relations
for these two sets of work also vary, they are considered separately below.

Disciplinary epistemic relations and social relations

Evidence for the disciplinary component of the science-related knowledge code
relates primarily to teaching done by mainstream disciplinary lecturers. Their focus
on specialized and distinct concepts and procedures linked to their own disciplines,
with little overlap between the work of the different lecturers, indicates relatively
 Widening access in science 193

very strong epistemic relations (ER++). Table 8.1 outlines examples of disciplinary
concepts identified in ISCM course documents, such as the geological processes
of river formation, the concept of a mole in chemistry, the process of radioactive
decay in physics, and factors that affect range of joint motion in human kinetics
and ergonomics. Linking with these respective examples of disciplinary concepts,
students are required to engage with specialized procedures such as plotting a river
course on a topographic map, titrating a solution to determine molar concentration,
calculating a radioactive decay constant and measuring range of joint motion using
a goniometer. Assessment tasks that draw on these concepts and procedures consist
of both low-order questions such as ‘State the three main processes involved in
river formation’ and high-order questions such as ‘Using a well-annotated diagram,
explain the process of subduction in plate tectonics’.
While epistemic relations are strongly legitimated, social relations in the disci-
plinary component of the science-related knowledge code are not. Student dispo-
sitions, behavioural attributes and opinions are downplayed, representing weaker
social relations (SR–). Typical examples, as indicated in Table 8.1, are where student
opinion is sought on the effect of genetically modified organisms on the environ-
ment, or the use of nuclear energy as a form of energy generation, but students are
judged on the coherence of their scientific argument, not on their opinion. Another
example relates to the geological field trip in which the stated primary purpose is to
enable ‘careful, rigorous and systematic observation’, which could represent stronger
social relations linked to a particular way of working, but marks are allocated instead
for correct identification of certain features and correct use of geological descriptors
of rock characteristics, representing weaker social relations.

Scientific literacies epistemic relations and social relations

Evidence for the scientific literacies component of the science-related knowledge
code relates mainly to work done by the two literacies lecturers. Since this work is
underpinned by specialized science concepts, it represents stronger epistemic rela-
tions, but because many of the principles in this category can apply across a number
of science disciplines, they are not as strong as the disciplinary-based category. Table
8.1 presents some examples of scientific literacies concepts that are visible in the
ISCM course documentation. These include developing an understanding of how
scientific knowledge is generated (for example through careful measurement and
observation, inductively through looking for patterns in nature, deductively through
generating hypotheses and conducting controlled experiments, and making predic-
tions) as well as the basis upon which knowledge claims are made (for example,
through using empirical data, recognizing the tentative nature of science, recog-
nizing the idealization of many science laws, and knowing that understanding is
often based on models of reality). Specialized procedures are closely linked to these
concepts and relate to, among other things, designing and conducting experiments,
developing coherent arguments, working with data (collecting, analysis, interpret-
ing, presenting), writing scientifically and evaluating sources. In an independent
194 Karen Ellery

research project, students are expected to design, conduct and present research in
which they examine the effect of an environmental factor of their choice on plant
growth. Most draw on home-based practices of growing plants, and their projects
usually include the use of grey water or some form of organic waste. Assessment
criteria for the proposal include such aspects as using an appropriate experimental
design (taking into account randomization, replication and control of local condi-
tions) and developing a logical argument (with the hypothesis being supported by
information from the literature). These criteria represent stronger epistemic rela-
tions (ER+).
These scientific literacies concepts and procedures have relevance in the disci-
plinary work as well. For example, the principles of experimental design could apply
equally in chemistry and human kinetics and ergonomics experiments. However,
the analysis indicates that they are addressed initially in ISCM tutorials in fairly
generic ways, and only later drawn on specifically in the disciplinary work.
In terms of social relations, the required dispositional attributes and values associ-
ated with this experimental work are closely linked to the specialized procedures.
Students are expected to work in rigorous, reliable, accurate, precise, honest, curi-
ous, logical, analytical, critical and evaluative ways, all attributes required of scien-
tists (Matthews 2015). While these attributes might at first glance suggest stronger
social relations, they in fact represent weaker social relations (SR–): their focus may
be knower attributes but their basis for achievement are the epistemic relations
outlined above. For example, as indicated in Table 8.1, honesty in collecting and
presentation of data is emphasized in tutorial discussions, but since this is difficult
to ascertain students are instead awarded marks based solely on the presented out-
comes of their experiment. Furthermore, students are generally expected to work
objectively in the sciences, which means they must suspend personal biases and sub-
jectivity in favour of what the empirical data tells them. This valuing of objectivity
in the sciences therefore also represents weaker social relations (SR–).

Student response to the knowledge code

In terms of the science-related knowledge code, when students exhibit poor realiza-
tion in the disciplinary component their answers are usually conceptually incorrect,
too generalized or oversimplified (see Ellery 2017b). In other words, they recog-
nize that a knowledge code is necessitated but are unable to produce the required
cognitive or abstract standard. Similar code recognition but poor code realization
is evident in the scientific literacies work and is usually linked to, amongst other
things, students not being acceptably accurate or precise in their measurements,
logical in their argument or objective in gathering their data. Therefore, although
social relations are relatively weaker in this code, it appears that students need to pay
some attention to the valuing of procedures as they influence being able to produce
the requisite text.The student interviews indicate that educational background may
play a role in poor realization of the science-related knowledge code at a university
level. As one student commented:
TABLE 8.1 Specific translation device for knowledge code in ISCM (adapted from Ellery 2016, 2017a)

Code Indicators Empirical evidence from observations, document analysis and staff interviews
Science- Disciplinary ER++ Concepts (from lecture handouts) Procedures (from practical handouts)
related component Geology: processes of river formation Plot river course on topographic map
knowledge Disciplinary Chemistry: concept of a mole Titrate a solution to determine molar
code conceptual Physics: process of radioactive decay concentration
and procedural HKE: factors that affect range of joint motion Calculate a decay constant.
knowledge is Measure range of joint motion using a goniometer
emphasized Assessment (test): State main river formation processes; explain process of subduction in plate
while tectonics
dispositions SR– Dispositional attributes and opinions
and opinions Assessment (tutorial discussion): Opinion sought on nuclear energy as a form of energy
are downplayed generation, but answers need scientific base
Assessment (field trip) Stated focus is careful, rigorous and systematic observation, but marks
allocated for correct identification of geological features
Scientific literacies ER+ Concepts (from tutorial handouts) Procedures (from tutorial handouts)
component How scientific knowledge is generated Design and conduct experiments
Scientific literacies (measurement, observation, inductively, Develop coherent arguments

Widening access in science 195

conceptual deductively, developing hypotheses, Work with data
and procedural conducting experiments, making predictions) Write scientifically
knowledge The basis upon which knowledge claims are Evaluate sources
valued while made (use of empirical data and models
dispositions of reality; recognizing tentative nature of
and values are science and idealization of laws)
downplayed Assessment criteria in a proposal: Experimental design (principles of randomization, replication,
control of conditions); Logical argument (hypothesis supported by literature)
SR– Dispositional attributes
Tutorial discussion: On the value of and need for honesty in reporting in experiment, but
assessment criteria based on final outcome of experiment
196 Karen Ellery

I won’t lie, it’s the first subject that I can say it was quite difficult for me ... last
year when I was doing my matric, I didn’t see how deep is science, but in ISCM
I saw how deep is science ... in terms of being critical and put your understand-
ing towards your work.3
(Kanelo)

Academic practices-related knower code

Epistemic relations and social relations
The second code legitimated in ISCM is an academic practices-related knower
code. The term ‘academic practices’, which has social connotations and infers
underpinning values as opposed to the more acontextual ‘study skills’, is used here
to describe this work. It relates to what staff perceive to be the main purpose of
the course: enabling students to become effective learners in a higher education
science context. This purpose arises from a concern that once students leave the
access programme and enter the mainstream, support for learning is either greatly
reduced or non-existent. As one staff member stated in an interview: ‘When they
leave us students need to be able to get on with the job ... work on their own’
(Lecturer 1). Since there are no underpinning specialized concepts, but instead
students are required to engage in and develop learning practices appropriate
for a higher education context, it embodies relatively weaker epistemic relations
(ER–).
As indicated in Table 8.2, epistemic relations here are primarily procedural and
linked to particular learning practices:

• organizational practices that relate to issues such as managing time and organiz-
ing notes;
• technical practices that are linked to their accessing information and taking
good lecture notes from which they can learn;
• study practices that require students to prepare for and attend lectures, ask ques-
tions, review and consolidate work, practice calculations, and engage with and
respond appropriately to feedback; and
• assessment techniques that include managing time in assessment and unpacking
questions.

Working effectively in these academic practices is contingent on understanding

the contextual ‘rules’, as outlined in Table 8.2. For example, in order to develop the
practice of reviewing and consolidating lectures, students need to know that the
ISCM context requires them to work constructively and independently outside
class, otherwise it is unlikely they will be successful. Likewise, answering test ques-
tions effectively requires understanding that verbs of instruction (describe, explain,
evaluate) determine the kind of answer, and that the supporting text and mark allo-
cation signal the scope of answer required.
TABLE 8.2 Specific translation device for knower code in ISCM (adapted from Ellery 2016, 2018)

Code Indicators Empirical evidence from observations, document analysis and staff interviews
Academic practices- Dispositions and behavioural ER– Contextual ‘rules’ (from Procedures (from tutorial handouts)
related knower attributes for shaping own tutorial interactions) Organizational practices: manage time, file
code learning are emphasized Organizing principles notes
and valued of library, Internet, Technical practices: access information and
textbooks, dictionary, take lecture notes
words (suffix, prefix), Study practices: prepare for and attend
learning context, lectures, ask questions, review and
assessment questions consolidate work, practice calculations,
work with feedback
Assessment techniques: manage time, unpack
questions

Widening access in science 197

SR+ Dispositional attributes
Staff interview quotes: ‘realising they don’t know; study effectively and
independently; developing deeper and better understanding; right kind of
academic level; willingness to engage; active participation’
198 Karen Ellery

It is through these academic practices that a particular kind of learner is being

legitimated. In interviews, as indicated in Table 8.2, staff articulate the norms and val-
ues they perceive as necessary to become an effective learner in ISCM in particular
and in a higher education context in general. All five staff interviewed spoke about
learner independence. As one stated: ‘[They] need to become capable as students to
study effectively and independently as we cannot always be here to support them’
(Lecturer 5). In this regard another said: ‘[They should] not always rely on someone to
teach them everything nor rely on someone to check whether they have understood’
(Lecturer 1). Most staff emphasized the need for students to develop proper under-
standing. As one mentioned: ‘It’s not just surface content … it’s about developing a
deeper and better understanding … [they] cannot just rote learn all the time’ (Lecturer
3). They also spoke about engaging at a higher level than at school. As one suggested:
‘[Staff] need to bring them to the right kind of academic level in terms of the scientific
reasoning approach’ (Lecturer 3). Most spoke about developing metacognitive, reflec-
tive and reflexive understanding. In this regard one stated: ‘It’s metacognitive thinking,
you know. Realizing that they don’t know what’s going on, thinking about how they
learn, and what they are going to learn, and why they are learning it’ (Lecturer 2). Other
aspects that staff spoke about were seeking help when needed, a willingness to engage
and be challenged and active participation. In short, they view the academic practices
work as engendering independence in learning and development of depth understand-
ing, which is conceptualized here as students becoming and being autonomous learners.
This invokes learners underpinned by relatively stronger social relations (SR+).
Autonomy in learning is enabled in ISCM through tutorial interactions in which
appropriate learning practices such as preparatory reading for a topic, consolidating
lectures and practical sessions, and active reflection on feedback for improvement,
are supported and modelled in order that students can later do such activities inde-
pendently of lecturer input. Outputs from these tutorial activities are not assessed
directly, but many disciplinary and scientific literacies assessment tasks in ISCM
are instead designed to not only assess knowledge of concepts and procedures but
also to test autonomy in student engagement. Therefore, what may appear to be a
question exhibiting distinctly stronger epistemic relations, such as the identification
of a limiting reagent in chemistry, in fact has stronger learning-context social rela-
tions embedded in the task, based on how the topic has been addressed in class. If
students have not heeded the cue to watch the necessary YouTube video, practiced
calculating molar mass and number of moles and developed a proper understanding
in order to answer a question that is phrased differently to that in class, it is highly
unlikely that they will be successful in the task. In other words, the basis for achieve-
ment of the academic practices knower code is not measured directly, but rather
indirectly through assessment tasks in the science-related knowledge code.

Student response to the knower code

When asked about poor performance in knowledge-code assessment tasks many
students recognize the need to engage with their studies in particular ways, as
 Widening access in science 199

demanded by the knower code, but are less successful at realizing such practice
(see Ellery 2017b). In this regard students speak about the need to learn more:
‘I just did not learn enough for this test, I will be honest’ (Liwa). Along with
insufficient engagement, how students engage is also an issue. Some appreciate
the need to learn for understanding: ‘I think I just did it [learning] for marks not
for knowledge’ (Khuselwa). Poor engagement often arises from a misreading of
context requirements: ‘I didn’t know [learn] this because when she [the lecturer]
answered that question she didn’t spend a lot of time with it, she just browsed
through it’ (Anele). Students also recognize the need to practice technical tasks
more: ‘I understand the concept and method of calculating but I made a confusion
in the test … I should practise more often to familiarize myself with the conver-
sions’ (Mbuyiselo).
In summary, students signal poor realization of the academic practices-related
knower code in interviews when they speak of, amongst other things, ignoring
difficult work, poor engagement with resources (including feedback), inconsistent
study habits, reliance on rote learning and surface understanding, and dependence
on an authority figure to direct their learning, which they acknowledge originates
from their school contexts. These habits suggest unspecialized learning-context
dispositions that exhibit weaker social relations (SR–). When combined with the
weaker epistemic relations (ER–) of the generic procedures or skills being taught,
this suggests that students enact this component of the course as a relativist code
(ER–, SR–), instead of a knower code (ER–, SR+). There is thus a code clash
for the learning context. Despite explicit articulation and strong support for the
knower code in ISCM, students are not easily making the necessary shift. In this
regard, students mention the challenge of changing entrenched study practices that
have ensured success previously:

I did well at school, very well, I was often first or close in my class ... my pattern
was to study very hard just before tests – sometimes for hours ... [but here at
university] I always run out of time ’coz there are tests and assignments and
homework and stuff and it all takes too much time ... I guess I should be work-
ing more consistently like every day like you tell us [laughs].
(Mandisa)

Epistemological access in ISCM

Figure 8.2 indicates that ISCM legitimates two distinct codes. In summary, the
science-related knowledge code focuses mainly on students understanding of dis-
ciplinary and scientific concepts and procedures and forms the primary basis for
success in ISCM. While the disciplinary and scientific literacies components can be
plotted separately on the specialization code plane based on their different strengths
of epistemic relations, they still form part of the same code in which epistemic
relations are emphasized and social relations downplayed. The academic practices-
related knower code focuses primarily on student learning and relates to students
200 Karen Ellery

FIGURE 8.2 Specialization codes enacted in the ISCM course

becoming and being independent learners responsible for their own knowledge
and understanding, which represents weaker epistemic relations and stronger social
relations.
Because of the way the course and assessment practices are structured in ISCM,
there is a close relationship between the two codes. In other words, if students
do not work independently and develop deep understanding, as required by the
academic practices-related knower code, they seldom can provide the depth that
science answers require by the science-related knowledge code. This raises the key
finding of this study that realizing the science-related knowledge code is contingent
on students realizing the academic practices-related knower code. This hierarchy of
access to these two codes has major implications for students’ success and forms the
basis for the more generalized conceptual model of epistemological access in the
following section.

Epistemological access in higher education science

In serving its purpose of enabling access to the sciences ISCM has a dual focus
of developing students’ science knowledge as well as their dispositions as learners.
ISCM is therefore unrepresentative of science higher education courses in general
in that it has an explicit focus on developing students as learners through overt
articulation, support, modelling and scaffolding of appropriate learning practices.
However, many science-related higher education courses do, in fact, legitimate the
kind of learner similar to that valued in ISCM (see Case 2013, Ellery 2017c). In
other words, by working at a high volume and pace and addressing most concepts
only once in class, they are legitimating a self-regulated learner who will work
independently to ensure their own understanding. However, provision of support
 Widening access in science 201

for such work is little or non-existent (Ellery 2017c). It appears that in current
content-laden science-related curricula it is generally assumed that students will
invoke learning approaches necessary for their success.
Since many science courses likely legitimate both a knowledge code (explicitly)
and a knower code (implicitly), two levels of access are required for success in the
sciences in general.The first level represents access to the specialized knowledge and
procedures associated with stronger epistemic relations and is referred to as science-
context access. This is the core of any science curriculum. However, since academic
success is influenced by both the mastery of specialized knowledge and the ‘capacity
to be in control of one’s own learning’ (Edwards 2015: 14), access at the science-
context level may well be constrained, as this study suggests, by poor access at the
learning-context level.The focus here on becoming a particular kind of learner in a
particular academic context is referred to as learning-context access. Without access at
both levels, students are unlikely to be successful.
Access to two codes, with different strengths of epistemic relations and social rela-
tions, has implications for curriculum and pedagogy. The science context with its
stronger epistemic relations poses certain structural constraints in any curriculum.
To give an example, understanding the process of evolution as described in many
higher education science curricula requires a prior understanding of base concepts
such as the structure of DNA and the process of gene recombination. In contrast, the
learning context with its stronger social relations demands a certain kind of learner
as perceived relevant by staff. Aspects of the curriculum associated with the learning
context are thus influenced by the purpose of the course and the values of the lectur-
ers and are much more negotiable than those of the science context. In this regard,
because ISCM is a science foundation course that is preparing students for success in
the mainstream, an independent and self-regulated learner is legitimated. However,
studies in engineering courses have shown that the ability to integrate multidisci-
plinary knowledge from a range of different contexts (Wolff and Luckett 2013) and
the capacity to work with large volumes and at a high pace (Case 2013) may also
(or instead) be key components of the learning context. Because there is more room
to adjust expectations in learning-context aspects of the curriculum, I suggest this is
where future effort needs to be focused to better accommodate students’ backgrounds.
Prior educational experiences will influence access at both levels. Nonetheless,
poor learning-context access in science-related disciplines is increasingly recog-
nized as problematic (Adendorff and Lutz, 2009, Case 2013) and several studies have
shown that prior educational experiences can limit uptake of, and access to, a new
code (Hoadley 2007, Chen 2010, Pearce et al. 2015). It is suggested that the larger
the articulation gap between secondary and higher education requirements, the
more challenging the uptake (Scott 2012). The difficulty of changing dispositions
and attitudes based on entrenched habits needs to be recognized, accommodated
and supported in our educational practices (Pym and Kapp, 2011), without which
certain groups of students will continue to be alienated and excluded from the sys-
tem. From a social justice perspective, there is an obvious and urgent need to better
support all students entering the higher sector today.
202 Karen Ellery

Conclusion
The contribution of this study, through enacting the concepts of specialization codes,
has been the recognition of two levels of concern in a curriculum: the science (or
disciplinary) context and the learning context.These two sets of concerns are always
present in an educational field of practice, but empirical studies usually focus on
either one or the other. By examining them together, these concepts allow a more
holistic and nuanced analysis of curriculum and the different bases of achievement
that are articulating within the educational context. As such, this approach offers the
necessary subtleties to identify a key site of student difficulties with success – at the
level of the learning-context knower. This highlights a somewhat unusual finding
for education in the sciences, namely that learning-context social relations are key
to enabling access to the highly valued science knowledge that will ensure success
in academia.This emphasizes the need to significantly rethink science higher educa-
tion to take into account methods for teaching both the knowledge of science and
how students can be apprenticed into becoming science learners in a higher educa-
tion context, thereby enabling epistemological access. This is particularly relevant
for students whose prior socialization at home and school does not match well with
expectations at university.

Notes
1 For the larger study to which this chapter relates, see Ellery (2016).
2 The literacies approach in ISCM is based on Street’s (2006) ideological model that
assumes there are multiple literacies and that these are acquired in socio-cultural contexts.
3 Pseudonyms are used for student quotes.

References
Archer, L., Hutchings, M. and Ross, A. (2003) Higher Education and Social Class: Issues of
Exclusion and Inclusion, London: Routledge.
Adendorff, H. and Lutz, H. (2009) ‘Factors influencing the learning process in first-year
Chemistry’, in B. Leibowitz, A.Van der Merwe, and S.Van Schalkwyk (eds) Focus on First-
Year Success, Stellenbosch: Sun Media, 167–80.
Arum, R., Gamoran, A. and Shavit,Y. (2012) ‘Expanded opportunities for all in global higher
education systems’, in L. Weis and N. Dolby (eds) Social Class and Education: Global
Perspectives, New York: Routledge, 15–36.
Barnett, R. (2007) A Will to Learn: Being a Student in an Age of Uncertainty, Maidenhead,
Berkshire: SRHE and Open University Press.
Bernstein, B. (2000) Pedagogy, Symbolic Control and Identity:Theory, research, critique, revised edi-
tion, Oxford: Rowman and Littlefield.
Case, J. M. (2013) Researching Student Learning in Higher Education: A social realist approach,
London: Routledge.
Chen, R. T. (2010) ‘Knowledge and Knowers in Online Learning: Investigating the Effects of Online
Flexible Learning on Student Sojourners’, unpublished PhD thesis, University of Wollongong.
Council on Higher Education Report. (2013) A Proposal for Undergraduate Curriculum Reform
in South Africa:The Case for a Flexible Curriculum Structure, Pretoria: CHE.
 Widening access in science 203

Dall’Alba, G. and Barnacle, R. (2007) ‘An ontological turn for higher education’, Studies in
Higher Education, 32(6): 679–91.
Edwards, A. (2015) ‘Designing tasks which engage learners with knowledge’ in I. Thompson
(Ed.) Designing Tasks in Secondary Education: Enhancing subject Understanding and Student
Engagement, Abingdon, Oxon: Routledge, 13–27.
Ellery, K. (2016) ‘Epistemological Access in a Science Foundation Course: A Social Realist Perspective’,
unpublished PhD thesis, Rhodes University.
Ellery, K. (2017a) ‘Conceptualising knowledge for access in the sciences: Academic develop-
ment from a social realist perspective’, Higher Education, 74: 915–31.
Ellery, K. (2017b) ‘A code theory perspective on science access: Clashes and conflicts’, South
African Journal of Higher Education, 31(3): 82–98.
Ellery, K (2017c) ‘Framing of transitional pedagogic practices in the sciences: Enabling access’,
Teaching in Higher Education, 22(8): 908–24.
Ellery, K. (2018) ‘Legitimation of knowers in science: A close-up view’. Journal of Education,
71: 24–38.
Hoadley, U. K. (2007) ‘The reproduction of social class inequalities through mathematics ped-
agogies in South African primary schools’, Journal of Curriculum Studies, 39(6): 679–706.
Lamont, A. and Maton, K. (2008) ‘Choosing music: Exploratory studies into the low uptake
of music GCSE’, British Journal of Music Education 25(3): 267–82.
Longden, B. (2006) ‘An institutional response to changing student expectations and their
impact on retention rates’ Journal of Higher Education Policy and Management, 28(2):
173–87.
Matthews, M. (2015) Science Teaching:The Contribution of History and Philosophy of Science (2nd
edn.), New York: Routledge.
Maton, K. (2014) Knowledge and Knowers: Towards a Realist Sociology of Education, London:
Routledge.
Maton, K. (2016) ‘Legitimation Code Theory: Building knowledge about knowledge-
building’ in K. Maton, S. Hood and S. Shay (eds) Knowledge-building: Educational Studies in
Legitimation Code Theory, London: Routledge, 1–23.
Maton, K. and Chen, R. (2016) ‘LCT and qualitative research: Creating a translation device
for studying constructivist pedagogy’, in K. Maton, S. Hood and S. Shay (eds) Knowledge-
Building: Educational Studies in Legitimation Code Theory, London: Routledge, 27–48.
Maton, K. and Howard, S. K. (2016) ‘LCT in mixed-methods research: Evolving an instru-
ment for quantitative data’, in K. Maton, S. Hood & S. Shay (eds) Knowledge-Building:
Educational Studies in Legitimation Code Theory, London: Routledge, 49–71.
Morrow, W. (2007) Learning to Teach in South Africa, Pretoria: HSRC Press.
Morrow, W. (2009) Bounds of Democracy: Epistemological Access in Higher Education, Cape Town:
Human Science Research Council Press. Retrieved from http://www.hsrcpress.ac.za
Muller, J. (2014) ‘Every picture tells a story: Epistemological access and knowledge’, Education
as Change, 18(2): 255–69.
Muller, J. (2015) ‘The future of knowledge and skills in science and technology higher educa-
tion’, Higher Education, 70: 409–16.
Pearce, H., Campbell, A., Craig, T. S., le Roux, P., Nathoo, K. and Vicatos, E. (2015) ‘The
articulation between mainstream and extended degree programmes in engineering at
the University of Cape Town: Reflections and possibilities’, South African Journal of Higher
Education, 29(1): 150–63.
Pym, J. and Kapp, R. (2011) ‘Harnessing agency:Towards a learning model for undergraduate
students’, Studies in Higher Education, 38(2): 272–84.
Science Extended Studies Programme Review Report (2011) Curriculum Review: Science
Extended Studies Programme, Rhodes University, unpublished report, Rhodes University.
204 Karen Ellery

Scott, I. (2012) ‘Alternative Access to University: Past, Present and Future’, in R. Dhunpath and
R.Vithal (eds) Alternative Access to Higher Education: Underprepared Students or Underprepared
Institutions?, Cape Town: Pearson, 26–50.
Street, B. (2006) Autonomous and ideological models of literacy: approaches from New Literacy Studies.
Retrieved from http:/www.philbu.net/media-anthropology/street_newliteracy.pdf.
Wolff, K. and Luckett, K. (2013) ‘Integrating multidisciplinary engineering knowledge’,
Teaching in Higher Education, 18(1): 78–92.
9
THE RELATIONSHIP BETWEEN
SPECIALIZED DISCIPLINARY
KNOWLEDGE AND ITS
APPLICATION IN THE WORLD
A case study in engineering design

Nicky Wolmarans

Introduction
Traditionally, professional education has been about the acquisition of a body of
knowledge that graduates are expected to ‘apply’ in their professional practice after
graduation. In this chapter I argue that a model of professions as the application
of disciplinary knowledge is inadequate because it fails to take into account the
complexity of the ‘real world’ to which the knowledge is applied. This in turn has
led to curriculum design choices that fail to prepare graduates to work dialectically
between the complexity of specialized disciplinary knowledge and the complexity
of the world in which it is employed.
Engineering provides a rich case study. It is a profession that has positioned itself
as a science-based discipline, founded on a canon of well-defined disciplinary sub-
jects. Engineering curricula have tended to focus on the transmission and acquisi-
tion of scientific concepts and the relations between concepts, culminating in a final
‘capstone’ design project. This capstone design project is intended to integrate the
specialized disciplinary knowledge acquired throughout the curriculum for applica-
tion in a single ‘real world’ project. Most projects are set up to mimic the sorts of
projects that engineering graduates are likely to encounter in professional practice
(Froyd et al. 2012, Harris et al. 1994). They are intended to bridge the gap between
engineering science and engineering practice. However, many students lack the
skills to design when confronted with these design projects for the first time, even
when they have successfully completed their engineering science courses (see Kotta
2011 for a detailed study of students’ experiences of senior design projects).
206 Nicky Wolmarans

Over the last century of engineering education reform (Froyd et al. 2012, Grinter
1955, Mann 1918), employers have been calling for improved interpersonal and
enabling skills, a strong foundation in the fundamental sciences, and the centrality
of design in the curriculum. Recent studies of employer perceptions of graduate
engineers report an improvement in, for example, teamwork, communication skills
and management compared to the past (J. King 2007, R. King 2008). However,
many engineering graduates still appear to be unable to apply scientific knowledge
to solve professional problems; as J. King (2007: 7) reports:

Although industry is generally satisfied with the current quality of graduate

engineers it regards the ability to apply theoretical knowledge to real industrial
problems as the single most desirable attribute in new recruits. But this ability
has become rarer in recent years

In this chapter I address the question of what it means to ‘apply theoretical knowl-
edge’ to ‘real industrial problems’ emergent from the complexity of the world. Using
engineering design projects in the curriculum as a proxy for real professional prob-
lems, I present an analysis of the relationship between material artifacts and the
abstract concepts used to analyze and mathematically model them for the purpose
of design. The chapter draws on part of a larger PhD study (Wolmarans 2017a)
in which 17 engineering design projects located three engineering streams where
investigated. Only two of the projects are presented here, both introductory design
projects. One is the first project in a sequence of civil engineering projects, the
other is the first project in a sequence of structural engineering projects. A com-
parison of the two projects shows that different ways of simplifying ‘real’ projects for
the purpose of learning have vastly different effects on the nature of the required
reasoning.
I bring together concepts from the Semantics and Specialization dimensions
of Legitimation Code Theory (Maton 2013, 2014, 2020), because of their capac-
ity when integrated to analyze relationships between objects of knowledge and
knowledge of objects. The concept of semantic gravity provides a lens to analyze the
significance of knowledge of the object of design, while the concept of semantic
density provides a lens to analyze the complexity of the reasoning. By coding the
knowledge requirements of each step in the design thinking process, I am able
to show shifts between knowledge of complex things (using the concept of ontic
relations) and knowledge of complex theoretical concepts (using the concept of
discursive relations).The analysis contributes to building a more robust model of pro-
fessional reasoning which has implications for learning. The study provides insight
into the limitations of certain types of tasks when they constrain the complex-
ity of thinking about the ‘things’ being analyzed. In short, privileging specialized
knowledge over knowledge of objects effectively distorts the complex dialectical
relations involved in professional reasoning. Although the case presented in this
chapter is that of engineering, these findings have implications for a range of dif-
ferent professions.
 Disciplinary knowledge and its application 207

Theoretical framework
The theoretical backdrop to this study begins from Bernstein’s distinction (2000)
between ‘singulars’ and ‘regions’. He described “singulars as bodies of disciplinary
educational knowledge that are strongly bounded from other educational knowl-
edge and from external concerns in the world beyond education. When discussed
in broad terms, disciplines such as Physics, Mathematics and History might be
described as singulars, each with its own specialized concepts and rules of con-
ceptual relations among concepts. Bernstein described ‘regions’ as the interface
between singulars and the practical concerns of the world – they involve selection
of ideas from a range of singulars and their application to a field of practice beyond
education. Professions are usually described as regions. This model of knowledge
assumes singulars precede regions which in turn project that knowledge onto the
world (see Smit 2017 for a detailed critique of this notion of singulars and regions
in a study of thermodynamics in science and engineering).
Bernstein’s model is useful in that it does recognize a distinction between
the structure and organization of specialized disciplinary knowledge (e.g., fluid
mechanics) and the structure and organization of knowledge in engineering prac-
tice where that knowledge is used in relation to the design of complex systems.
However, the model has significant shortcomings. For one thing, it does not address
the nature of the relationship between knowledge inside the discipline (abstract
fluid mechanics) and knowledge outside the discipline (the design of a slipway for
a new yacht basin) – see, for example, Wolmarans (2017b). Based on Bernstein’s
model of regions, scholars have characterized professional education in terms of
mastery of specialized disciplinary knowledge prior to its application to problems in
the world – the ‘word’ before the ‘world’ (Beck 2002, Beck and Young 2005). This
model tacitly assumes that if graduates have learned the concepts and legitimate
rules of combination among concepts within a disciplinary specialization, then such
theoretical knowledge can be unproblematically ‘applied’ to external problems. It is
a model of knowledge that tacitly underpins many engineering programs. But one
of the under-developed aspects of this notion of ‘regions’ is that it leaves us blind to
the significance of the contextual detail inherent in professional problems emergent
from the world. It fails to take account of the need to translate knowledge struc-
tured by the logic of internal conceptual coherence into a logic structured by the
external realities of the world. In short, it fails to explore the nature of the external
problem itself.
Legitimation Code Theory (LCT) includes a number of conceptual tools that
offer a more nuanced way of viewing knowledge practices (Maton 2014). The
Semantics dimension of LCT is a particularly attractive analytic tool for investi-
gating professional knowledge in terms of the relations between ‘abstract’ special-
ized disciplinary knowledge and ‘contextually embedded’ professional problems.
Semantics centres on exploring the organizing principles underlying practices in
terms of semantic gravity (SG), which conceptualizes degrees of context-dependence,
and semantic density (SD), which conceptualizes complexity. Both semantic gravity
208 Nicky Wolmarans

and semantic density describe a range of relative strengths from stronger to weaker.
For example, when SG is weaker, the underlying practices are relatively less depen-
dent on any specific context, while stronger SG means that the underlying prac-
tices are relatively more dependent on a specific context. In the case of SD, more
complex practices exhibit relatively stronger SD and less complex practices exhibit
relatively weaker SD.
For the purposes of this study:

• stronger semantic gravity (SG+) means that making sense of the problem itself
and the specialized knowledge recruited to develop a solution are strongly
dependent on the specifics of the problem;
• weaker semantic gravity (SG–) suggests that the specialized knowledge is gen-
eralizable across multiple contexts;
• stronger semantic density (SD+) indicates projects requiring more complex
reasoning; and
• weaker semantic density (SD–) indicates a simpler problem.

SG and SD vary independently of each other and can shift strength through the
duration of any activity. SG↑ indicates a process of strengthening semantic gravity,
in this case increasing the specificity of a project. SD↓ indicates weakening semantic
density, here simplifying either relations among specialized concepts or the aspects
of the object of design to be considered.
Semantic codes have been used productively to show the importance of shifting
between more abstracted theory (SG–) and more concrete examples (SG+), both in
driving cumulative knowledge-building and to identify tacit evaluative criteria. See,
for example, Blackie (2014) on Chemistry, Georgiou et al. (2014) on Physics, Maton
(2013) on Biology and History, and Shay and Steyn (2016) on Design. In terms of
recruiting specialized knowledge to solve professional problems, semantic gravity
can provide a means of describing the relation between the ‘abstract’ theoretical
knowledge in its academic form and the more ‘concrete’ problems that emerge from
practice. When looking at professional education, semantic density offers a way to
look at progression through a curriculum based on increasing levels of complexity,
or strengthening semantic density. Shay and Steyn (2016) used semantic codes to
analyze a sequence of projects in an introductory design course. Building from their
results, they were able to redesign the sequence of tasks into a coherent trajectory of
increasing complexity and to identify the link between complexity and ‘concrete-
ness’ of the project.
I have also used Semantics as an analytical lens. However, semantic density has not
yet been used in LCT studies to distinguish between i) the complexity of the object
of study or the artifact of design (the thing, what it is and how it works and ii) the
complexity of the knowledge recruited to do the analysis (the specialized disciplin-
ary knowledge). In order to make this distinction I draw on the LCT dimension of
Specialization, specifically the distinction that Maton (2014) makes between ontic
relations between knowledge and its objects of study and discursive relations between
 Disciplinary knowledge and its application 209

knowledge and other knowledges. For my study, this can be enacted to distinguish
between those concepts with which we make sense of things in and of the world
(ontic relations) and those concepts defined within any specific discipline that con-
form to disciplinary rules (discursive relations). For example, the way in which we
understand a bicycle, what it is used for, how to ride it, what it looks like, requires
knowledge of a bicycle in terms of its physicality and one’s experience of seeing or
using a bicycle (ontic relations). Modelling the dynamics of motion, the principles of
friction between road and tire, and the strength of the frame requires specialized dis-
ciplinary knowledge (discursive relations). In this study I have analytically separated:

• the complexity of the disciplinary knowledge recruited in the task – semantic

density of discursive relations or DSD; and
• the complexity of the knowledge of the object of design – semantic density of
ontic relations or OSD.

Put simply, I have separately analyzed the complexity of the object (OSD) and the
complexity of the specialized knowledge recruited (DSD).

The case study

The data presented in this chapter draws on a PhD study of ‘The nature of profes-
sional reasoning’ (Wolmarans, 2017a). The study was based on an analysis of three
sequences of engineering design projects located in two engineering degree pro-
grams. Engineering was chosen as the case study because it is a profession that is
founded on a well-established scientific knowledge base.The knowledge tends to be
more explicit and clearly bounded than in some of the newer professions or those
based on social sciences and humanities. The reason for investigating engineering
design projects located in a curriculum rather than in practice is twofold. First, the
assessment requirements in an educational task require that the design reasoning is
elaborated and made more explicit than might be the case in professional practice
where professionals may draw more tacitly on their specialized knowledge. Second,
because the data was collected in an educational context, engineering design projects
were selected because engineering design is seen as the bridging subject between
knowledge and practice. Engineering design projects are usually intended to mimic
professional engineering projects, making them a reasonable proxy for investigating
professional knowledge in action.
Some of the projects analyzed in the broader study include, for example – the
design of a culvert to attenuate floodwaters on a particular watercourse; the design
of a multistory parking garage; and the specification of the requirements for a power
station. Although all the projects in the study were identified as ‘design’ projects by
the lecturers concerned, they did take different forms with different educational
objectives and privileged different forms of knowledge. This provided an opportu-
nity to investigate the effect that different ways of simplifying a professional project
(recontextualizing choices) have on the nature of the required reasoning.
210 Nicky Wolmarans

In this chapter I present a comparison of two of the projects: the design of

a bikeshare scheme (U1: Bikeshare scheme) and the analysis of the loading on a
structure (S1: Parking structure). Both are introductory projects, the first in a learn-
ing sequence of five design projects. Both are intended to mimic aspects of the
sorts of projects that professional engineers encounter in their professional practice.
However, because the projects are located early in the curriculum, they require
substantial simplification in comparison to ‘real’ engineering projects. The projects
need to be constructed so that they are appropriate for students not quite midway
through their curriculum, with limited exposure to the range and complexity of
the engineering sciences. The projects are further constrained by the limited time
available in the curriculum.These two projects were selected because they represent
recontextualization based on very different principles of recontextualization. In the
case of U1 (Bikeshare scheme) the disciplinary knowledge needed to understand
the project requirements was reduced (although it was reintroduced later in the
project). In the case of S1 (parking structure) the structure itself was simplified.
The analysis is relevant to understanding the consequences of teaching engineer-
ing sciences in relation to simplified idealized objects and illustrates some of the
unintended consequences of the choices made in terms of how design tasks are
simplified.

Data
The data collected for each design project included the design brief and a design
solution. The design brief is provided to students at the beginning of the project
and lays out the project requirements and task instructions. Briefs tend to identify
a problem scenario or client need and describe or refer to a context in which the
need or problem arises and in which the proposed design solution must operate.
The design solution describes the proposed artifact and documents evidence of
the proposed artifact’s performance in context. The design solution collected was
in the form of either a solution memorandum (a sample solution prepared by the
lecturer, a ‘model answer’) or a student design report. Because design rarely has a
‘model answer’ and marking rubrics are inadequate to show the details of design
reasoning, in most cases student solutions were used as a proxy for a solution
memorandum. In these cases, a ‘good’ student design solution was selected for
analysis. The basis of selection was on the grade achieved by the student for the
particular project.
Four units of analysis were identified for each project. The design brief typically
prescribed an artifact to be designed (artifact prescribed) and described or identified a
context in which the artifact needed to perform (context described). The design solu-
tion was analyzed in terms of the resultant artifact proposal (solution specified) as well
as the process of reasoning required to move from the brief to the final solution
(inferential reasoning). These four units of analysis – the artifact prescribed, the context
described, the solution specified, and the inferential reasoning – were coded for each
design project. For clarity, the four units of analysis described above are summarized
 Disciplinary knowledge and its application 211

TABLE 9.1 Units of analysis

Data Unit Description of each unit of analysis

Design Brief Artifact prescribed Each design brief usually prescribed an artifact to be
designed. In some cases, the required artifact may
be left to the discretion of the students, emergent
from the design purpose.
Context described All artifacts function in a context, and the context
places requirements and limitations on the design
of the artifact.
Design Inferential reasoning The inferential reasoning refers to the ideas
Solution generated, analytical models of potential
performance and decisions made during the
process of design, from brief to solution.
Solution specified At the end of the process of design a final solution
artifact is specified. Typically, a design solution
would be in the form of a set of technical
drawings detailing the artifact proposed as a
solution to the design problem.

in Table 9.1. The results of the analysis of each of the units of analysis for the two
projects are shown in Table 9.4 (U1: Bikeshare scheme) and Table 9.5 (S1: Parking
structure), further below.
The two projects presented, U1 (Bikeshare scheme) and S1 (Parking structure)
are both considered to be the first design project in a sequential trajectory of design
projects. These projects are compared because they illustrate two significantly dif-
ferent recontextualizing principles evident in the briefs. The parking structure (S1)
is recontextualized by simplifying the artifact significantly (OSD↓). The bikeshare
scheme (U1) is recontextualized by reducing the disciplinary knowledge required
(DSD↓).

Analysis and discussion

LCT Semantics was used to investigate the nature of the reasoning in engineering
design projects. Semantic gravity (SG) was used to analyze relations between ideas
and the object that they describe, and semantic density (SD) was used to investigate
the relation between concepts. In terms of semantic density, it became necessary to
distinguish relations between formal theoretical concepts defined within specialized
disciplines from more ‘everyday’ concepts used to make sense of the object of design
and the context in which it was intended to operate. This resulted in the distinc-
tion between semantic density of discursive relations (DSD) and semantic density
of ontic relations (OSD).1
The analysis is presented below in three stages. The first stage describes the
development and explanation of how semantic gravity and semantic density were
212 Nicky Wolmarans

TABLE 9.2 Semantic gravity categories of analysis

SG+/– Code description Examples

Generic or idealized, described in terms imposed from a body of disciplinary knowledge.
SG– – Meaning resides in general laws or S1 solution: A compressive force
concepts that transcend contexts; which is transferable across all
it does not require a concrete contexts and is applicable to all
reference to make sense. bodies as a result of gravity.
SG– Meaning is imposed from a S1 artifact: The structure is idealized
disciplinary body of conceptually but also unrealistic and impractical.
coherent knowledge, but the It is designed to elicit analytic
general law/s or concept/s are techniques rather than to perform a
specialized for application to an material functional.
object or class of object.
Specific and detailed, described in terms emergent from the context
SG+ Meaning relates to (originates in) a U1 artifact: A type of system; the brief
type of object or system; it refers does not prescribe the specifics of
to real objects, but abstracted from any particular system.
a specific, unique instance to a
class/type of object.
SG++ Meaning relates directly to a specific U1 context: A specific campus
instantiation of an object, accessible to students as part of
described in rich detail specific the design project. Familiarity
to a unique case/situation. with the specifics of the campus is
required.

operationalized in this study. The tables summarizing the categories that were used
to analyze the data (Table 9.2 and 9.3) make reference to examples from the data.
The examples are presented in the second stage where analysis of the two projects
is compared. The process of reasoning required to develop a solution to the task
is further elaborated in the third stage, where each step in reasoning is linked in a
chain or network of inferential steps.

Operationalizing semantic gravity and semantic density

The analytical categories used for semantic gravity are shown in Table 9.2. At the
first level stronger semantic gravity is distinguished from weaker semantic gravity
based on whether meaning emerges from an understanding of the contextual or
material detail (SG++ and SG+) or appears to be imposed from a specialized body
of knowledge (SG– – and SG–). Within stronger semantic gravity I differentiate
between specific or unique artifacts or contexts (SG++) and somewhat more gen-
eral types or classes of artifacts or contexts (SG+). Within weaker semantic gravity
I distinguish between generalized laws or principles that transcend contexts (SG– –)
and generalized theories specialized and imposed on the artifact or context in order
to describe it (SG–).
 Disciplinary knowledge and its application 213

TABLE 9.3 Semantic density categories of analysis

SD+/– Description Examples

Integrated or condensed into a coherent whole
SD++ Multiple interdependent DSD++: As entry projects neither
concepts/components required students to identify and select
integrated into a coherent relevant disciplines from others.
whole. Theoretical antecedents/ OSD++ U1 context: Students are
causal interdependencies are referred to the UCT campus; a
embedded and not identified complex context with emergent
explicitly characteristics.
SD+ Multiple interdependent DSD+ S1 inference: The calculations
concepts/components integrate geometric considerations,
integrated into a coherent mathematical techniques and structural
whole. However, relevant analysis.
disciplines/concepts/objects OSD+ U1 inferences: Simultaneous
are explicitly identified while consideration of surveyed route,
retaining simultaneous gradients, chosen bicycles, inexpert
interdependencies users, and shared user spaces.
Separated or elaborated into constituent parts
SD– Relevant concepts/components are DSD– S1 artifact: The structure is
identified and separated into a designed to demonstrate a sequence of
linear sequence of prescribed analytical techniques at the expense of
relations. functionality.
OSD– U1 solution: The solution is
presented as a sequence of features
listed in the report.
SD– – A single concept/component DSD– – U1 context: Located in a
is identified as relevant and is surveying course identifies surveying
isolated and dislocated from knowledge as the only relevant
its disciplinary/contextual discipline.
relations. OSD– – S1 inference: Each structural
element is analyzed in isolation from
the others.
SD0 No specialized concepts or DSD0 U1 artifact: Students are referred
knowledge of the object are to Wikipedia for a description of a
required. bikeshare scheme.
OSD0 S1 context: The structure is
stripped of interaction with a real
context.

The analytical categories used for semantic density are shown in Table 9.3. Ontic
semantic density (OSD) here refers to concepts used to make sense of the ‘things’
of the design, where coherence among concepts tends to be held together by the
contextual details of the material artifacts – what they are and how they work.
Discursive semantic density (DSD) here refers to formal concepts defined within
a specialized body of knowledge and the formal conceptually coherent relations
214 Nicky Wolmarans

among them. As described previously, OSD requires knowledge of the physicality

and experience of an objects, while DSD requires knowledge of the formal special-
ized knowledge used to analyze the object.
The strength of semantic density was categorized in relation to the level of detail,
number of concepts and number of relations between the concepts. It relates to the
nature of those conceptual relations needed to complete the design task. Although
OSD was differentiated from DSD, the principles that define the coding categories
are the same. In both cases stronger semantic density (SD++ or SD+) implies inte-
gration as the main principle of categorization. Weaker semantic density (SD– – or
SD–) implies separation or elaboration as the main principle of categorization.
Within stronger semantic density, SD++ means that integrated interdependen-
cies are embedded but not necessarily obvious. For example, students may need to
identify appropriate disciplinary knowledge or concepts within a discipline without
being given direction, or students may need to distinguish between those aspects
of an artifact which are relevant to the design from those that are incidental to the
task. SD+ indicates that the brief or instructions are given in such a way that the rel-
evant components or concepts have been identified for students, although the parts
retain necessary simultaneous interdependence. For example, the brief may instruct
students to design a gearbox, including selecting a motor, bearings and power trans-
mission elements and designing the shaft. This identifies separate components for
students, but each component interacts interdependently with the others. All design
decisions interact with other design decisions.
Within weaker semantic density, SD– means that each component or concept
has been identified for students and retain a sequential relation. In other words,
each relevant concept is identified and one concept links sequentially to the next
concept. SD– – indicates that a relevant component or concept has been identified
for students but separated to the point of dislocation from other components or
concepts. For example, students may need to size a bearing dislocated from other
machine components or students may be required to do a single calculation dislo-
cated from the complex conceptual network of meaning defined within a discipline.

Comparative analysis of the two introductory design projects

The two projects compared in this chapter both represent design projects at the
beginning of a sequence of design projects. At this point in the curriculum stu-
dents have little to no experience in design and have limited proficiency in the
engineering sciences but are expected to be relatively proficient in mathematics
and the basic sciences. Thus, design projects need to be significantly recontextual-
ized in comparison to ‘real’ professional projects, to align with the expertise that
students are expected to bring. The first design project in the ‘civils’ stream in civil
engineering required students to design a bikeshare scheme for the University of
Cape Town Campus (U1: Bikeshare scheme). The project was a two-week block
course in surveying. The project illustrates recontextualization that leaves the con-
textual and material details (ontic relations) intact, while reducing the disciplinary
 Disciplinary knowledge and its application 215

knowledge (discursive relations) to a single discipline with a sequence of special-

ized procedures prescribed. The first design project in the ‘structures’ stream in civil
engineering required students to complete a loading analysis of a generic parking
structure (S1: Parking structure). The project was integrated into the first structural
engineering course and ran parallel to the engineering science content taught in the
course. The project illustrates recontextualization based on stripping the contextual
and material details (ontic relations) in order to develop specialized conceptual rela-
tions (discursive relations).
The analysis of the projects first compares the four units of analysis for each
project, namely the artifact prescribed, the context described in the brief and the solution
specified and the inferential reasoning required to develop the solution. The analysis
illustrates the effect of recontextualizing choices on the complexity of understand-
ing required to engage in the task. The categorization of the inferential reasoning is
elaborated in the second analytical section. It shows the inferential chains of reason-
ing required of students as they develop an adequate design solution proposal. The
analysis of the detailed inferential reasoning is particularly interesting in that it shows
how significant stronger semantic density of ontic relations (OSD) are to the devel-
opment of professional reasoning.

U1: Bikeshare scheme

The artifact prescribed in the brief was a ‘bikeshare’ scheme, a type of system (SG+).
Initially the artifact can be understood in ‘common sense’ ways without recourse
to specialized disciplinary concepts (DSDo); students are referred to Wikipedia for a
generic description of a bikeshare scheme. The information provided by Wikipedia
functions to assist students to identify the relevant elements of the system and
recognize how they interact as a system (OSD+).
The context described in this project is integral to the design. Rather than describ-
ing the context, students are referred to the UCT campus in its current form,
including the terrain and usage: a congested campus with limited parking and nar-
row pathways shared by motor vehicles, busses and pedestrians. The climate is dry
hot summers with strong winds and cool wet winters. The campus is built on the
slope of a mountain and has many staircases and steep inclines.These details (SG++)
are not specified in the brief and it is expected that the students will draw on their
own experiences of the campus in the design. The design brief specifically identi-
fies the campus and does make reference to some of the aspects of the campus
context that might be considered, such as drawing attention to the importance of
surrounding buildings and their usage to determine potential demand; potential
interaction with the student bus service; and implied relations to existing roads for
access. However, much of the detail is left to students to elaborate from their own
everyday and embedded experience of the campus. Students need to either identify
as significant or discard as irrelevant to the design details from their experience.The
context is thus richly detailed and complex but without much guidance in terms of
what to consider and what to ignore (OSD++). Initially students are able to engage
with and make sense of the context without specialized disciplinary knowledge,
216 Nicky Wolmarans

although the location of the course in a surveying course does imply the impor-
tance of a single discipline over others that may be relevant to this project (DSD– –).
The solution specified has been analyzed in two parts because of the differences in
the nature of the parts. The description bikeshare system (U1A: Bikeshare descrip-
tion) included bicycle selection, safety equipment, exchange logistics, storage and
maintenance plans specific to this particular context (SG++). It was descriptive in
nature and retained its unspecialized format (DSD– –). Each element was identified
and described sequentially (OSD–) without much evidence of simultaneous inter-
ferences evident in the specification of the solution.
However, the real focus of the project was on the survey of the route and the
modifications required to accommodate its purpose (U1B: Bikeshare route). This
part of the solution was presented in the form of a digital elevation model (DEM),
a specialized representation of a particular terrain (SG++). A number of vertical
profiles at significant points along the route needed to be modified. Modifications
were required along the route to accommodate, for example, inexpert users and
interaction between multiple modes of transport. Although the process of devel-
oping the design (inferential reasoning) required networks of simultaneous consid-
erations, the presentation of the solution is simplified. Each part of the map can
be read as a sequence of positions along a contour map, and a sequence of vertical
profiles (DSD–). And the route is a significant simplification of the campus, with
each modification understood in relation to the terrain in one sense and the users
in another (OSD–).
The inferential reasoning on the other hand required simultaneous consideration
of the contextual details of the route in relation to specialized surveying knowl-
edge. The network of inferences involved a range of different strengths of both
OSD and DSD. Although a sequence of steps for the survey was provided in the
brief, each step involved interrelations between measurements with specialized
equipment, conversion into specialized surveying representations using multiple
mathematical relations, and the separation of vertical and horizontal interactions
and conversion into a map (DSD+). Once identified, the route itself could be
considered sequentially, but at each point in terms of interactions between terrain
and a range of users, including expected expertise of cyclists, interaction between
different modes of users (cyclists pedestrians and motorized transport) (OSD+).
The network of simultaneous reasoning shown in Figure 9.3 illustrates the nature
of OSD+, DSD+. Table 9.4 is a summary of the analysis above and is illustrated
in Figure 9.2.

S1: Parking structure

The artifact prescribed is a generic structure, stripped of all functional elements (ramps,
lifts, parking and vehicle flow arrangements). Figure 9.1 shows the structure reduced
to a collection of columns, beams and slabs configured in such a way as to offer a
range of different analytical challenges. That is, disciplinary analytical requirements
are imposed on the structure at the expense of functionality (SG–). As a result, each
structural element can be treated independently of the others (OSD– –). As with
 Disciplinary knowledge and its application 217

FIGURE 9.1 Idealized parking structure (S1)

FIGURE 9.2 Effect of recontextualization by weakening OSD (S1) or DSD (U1)

218 Nicky Wolmarans

FIGURE 9.3 Inferential reasoning: U1 (Bikeshare scheme)

U1 (Bikeshare scheme) the project is located in a disciplinary course in structural

analysis, signaling structural analysis as the only significant discipline. However, the
description of the structure does include a list of symbolic markers used to link the
elements through a sequence of disciplinary calculations (DSD–).
The context in which the structure operates is stripped from the project to the
point of being completely general (SG– –) and replaced by a single symbolic repre-
sentation of a ‘live load’ (variable load imposed under operation, specified for typi-
cal loading types). Even this level of understanding of a live load is not required to
complete the task (OSDo). The parking load is quantified as a uniformly distributed
load given in the brief as 4kN/m2 (DSD– –).
The solution specified was in the form of four compressive forces operating at the
base of each of the four bottom columns. The form of the solution is completely
transferable across all contexts, and in an abstract sense is applicable to all bodies in
that all bodies on earth interact with a surface by a compressive force at their base as
a result of gravity (SG– –). As with the context description given in the brief, the answer
is a single symbolic term dislocated from its relation to other terms disciplinary terms
(DSD– –) and stripped of any interaction in the world (OSDo). Students are able to
present the answer without engaging at all with what it means in the world.
 Disciplinary knowledge and its application 219

TABLE 9.4 Summary of analysis U1: Bikeshare scheme

Unit of analysis Semantic code

Artifact prescribed Bikeshare Scheme: SG+, OSD+, DSD0
Context described University campus: SG++, OSD++, DSD– –
Solution specified Bikeshare description: SG++, OSD–, DSD– –
Bikeshare Route (DEM): SG++, OSD–, DSD–
Inferential reasoning Bikeshare description: SG++, OSD+, DSD– –
(dominant mode) Bikeshare Route (DEM): SG++, OSD+, DSD+

TABLE 9.5 Summary of analysis S1: Parking structure

Unit of analysis Semantic code

Artifact prescribed Generic structure: SG–, OSD– –, DSD–
Context described Uniform distributed load: SG– –, OSD0, DSD– –
Solution specified Discrete compressive load: SG--, OSD0, DSD– –
Inferential reasoning Prescribed sequence of procedural calculations:
(dominant mode) SG–, OSD– –, DSD+

The details of the analysis of the inferential reasoning are shown in Figure 9.4. The
reasoning remains discursive, imposed on the structure for the purpose of struc-
tural analysis (SG–). Each element is considered sequentially but can be considered
independently of the other elements (OSD– –), linked only through a sequence of
calculations prescribed in the brief. Although the inferential steps are a prescribed
sequence of calculations, each calculation does integrate a number of structural
engineering principles that work together interdependently (DSD+).
The analysis summarized in Tables 9.4 and 9.5 is illustrated in Figure 9.2. on
a plane showing semantic density (OSD and DSD). There are three different pat-
terns of reasoning evident. The bikeshare description (U1A) shows the design of
the overall bikeshare scheme: bicycle selection, storage and exchange, the position
of bike shelters for storage with consideration of access for maintenance, and the
prescription of required safety gear. The bikeshare route (U1B) shows the disci-
plinary component of the design, the survey and generation of the DEM of the
proposed modifications to the route taking into account the bikeshare description
(U1A), users and terrain. The parking structure (S1) shows the loading analysis on
the structure.

Discussion of the effects of recontextualizing choices

in the design brief
The basis of recontextualization of the brief in U1 (Bikeshare scheme) is the weak-
ening of DSD: both context and artifact can be adequately understood without
requiring any specialized disciplinary knowledge. On the other hand, students need
220 Nicky Wolmarans

FIGURE 9.4 Inferential reasoning: S1 (Parking structure)

to construct considerable understanding of the context and artifact to begin the

design task: the strength of OSD is retained. By comparison, the basis of recontex-
tualization of the brief in S1 (Parking structure) is the weakening of ontic relations
OSD: both context and artifact can be adequately understood without engaging
meaningfully with the context or the functionality of the structure. Here, the struc-
ture is constructed such that students need to engage with a number of different
analytical approaches defined within the specialization of structural engineering: the
strength of discursive relations DSD is retained.
Both principles of recontextualization have consequences. In the case of the
bikeshare description (U1A), students are able to develop a solution based solely on
common sense knowledge and experience without recourse to disciplinary knowl-
edge. The science of logistics of the scheme, the structural analysis of the associ-
ated storage structures, formal principles of usage analysis were not required in the
recontextualized project. Projects of this nature do not help students to learn to
reason using specialized disciplinary knowledge in relation to the complex reality
 Disciplinary knowledge and its application 221

of the world. Instead students are able to develop an adequate solution based solely
on common sense. However, by locating the project in a surveying course and
providing a requirement to survey the route and propose modifications in terms of
gradient changes and vertical elevation points for the bikeshare route design (U1B),
disciplinary knowledge was introduced into the project and students were required
to engage simultaneously with both with the complexity of the world (stronger
OSD) and with disciplinary knowledge (stronger DSD).
In contrast, the recontextualized brief presented in S1 forces students to engage
with structural engineering principles and procedures (stronger DSD), but in this
case without any significant engagement with the complexity of the world (weaker
OSD). Again, projects of this nature do not help students to learn to reason using
specialized disciplinary knowledge in relation to the complex reality of the world.
In this case students are able to develop an adequate solution based solely on disci-
plinary knowledge and procedures, without engaging with the complexity of a real
object functioning in context.
The analysis of inferential relations that was described above was determined by
the nature of the relations between concepts through the process of design. The
details of the analysis of the inferential relations is elaborated in the section that fol-
lows. At this point suffice to say that the categorization of the inferential relations
shown in Figure 9.2 is based on the pattern of reasoning between context, artifact
and each step in the process of design from brief to final solution. In the case of
bikeshare route design (U1B) the strength of the ontic relations of the artifact and
context were retained, and certain discursive relations were required. Each step in
the prescribed sequence of tasks then drew interdependently on the artifact, context
and prior steps. In the case of the parking structure loading analysis (S1), the ontic
relations were stripped from the context and weakened for the artifact. Although
the discursive relations remained strong, the prescribed inferential steps became
effectively independent of artifact and context, resulting in a linear sequence of
reasoning.
In summary, weakening the semantic density of ontic relations risks dislocat-
ing the reasoning from the complexity of the world, resulting in a linear sequence
of reasoning. Retaining the complexity of ontic relations without introducing a
disciplinary knowledge component risks severing the links to professional knowl-
edge. This is particularly a risk early in the curriculum when students have limited
proficiency with specialized knowledge. In order to develop the skills needed to
reason between the world (ontic relations) and specialized disciplinary knowledge
(discursive relations), it is critical to retain or introduce both ontic and discursive
components in the requirements of the design project. The design of the bikeshare
route (U1B) provides one example of how this might be done.

The effect of recontextualizing choices on inferential reasoning

This section elaborates the categorization of the inferential relations presented
above. The analysis is based on a list of sequential steps provided in an addendum
222 Nicky Wolmarans

to each brief. Each step was analyzed in terms of its own relative complexity and
the relations to other steps in the process. Figures 9.3 and 9.4 show a comparison
between the steps involved in the design of the bikeshare route (U1B) and the
analysis of the parking structure (S1), respectively. What is immediately evident is
the difference in the patterns of reasoning. Figure 9.3 (U1B) shows a network of
inferences required in the development of a solution. Figure 9.4 (S1) shows a chain
of inferences.
S1 (Parking structure), shows how weakening OSD to the point of dislocation
resulted in a single input point (○) at the start of the chain of reasoning. With
the context (☐) stripped, it plays no role in the reasoning. What results is a linear
sequence of potentially procedural calculations as the load is calculated at each level.
With the retention of stronger OSD in U1B (Bikeshare route), both the artifact
(○) and the context (☐) are relevant to the development of the solution (✖). The
interdependence between steps is retained and the resultant inferences resemble
a network rather than a chain. For example, Step 4 in U1B requires proposals for
modifying the terrain of the route. This draws on the interaction between artifact
and context: the analysis of the campus (Step 1), the survey of the selected route
(Step 2), and issues emergent from the description of the bikeshare scheme (Step 3).
Decisions such as modification of the maximum operational gradient for inexpert/
casual cyclists or the introduction of wider lanes to accommodate multimodal trans-
port rely on a network of interdependent reasoning.
A comparison of the analysis of each project suggests that when ontic relations
are stripped – as in S1 (Parking structure) – the resulting reasoning tends to be lin-
ear. On the other hand, despite stripping the discursive relations in the brief for U1
(Bikeshare scheme), the retention of the complexity of the ontic relations resulted
in a more complex network of reasoning. However, without prescribing require-
ments for using disciplinary knowledge – surveying knowledge and procedures in
the case of U1B (Bikeshare route) – the risk is that the semantic density of discursive
relations will remain weak, as illustrated in the bikeshare description (U1A).

The influence of semantic gravity

One final point to make refers to the relationship between the semantic density
of ontic relations (OSD) and semantic gravity (SG). The complexity of the ontic
relations in U1 (Bikeshare scheme) is retained (OSD+), while in S1 (Parking struc-
ture) the ontic relations have been simplified to the point of dislocation (OSD– –)
or striped of relevance completely (OSDo). This tends to correspond with the
strength of semantic gravity. The specificity of a context or artifact (SG↑) depends
on the level of detail of ontic relations (OSD↑). On the other hand, as ontic detail is
stripped (OSD↓) the object becomes more generic (SG↓).This suggests that retain-
ing stronger OSD requires the brief to refer to more specific contexts and artifacts.
But that leaves the question of redundancy of OSD. If the strength of OSD corre-
sponds with the strength of SG, is it not redundant to introduce OSD? With refer-
ence to the analysis of the inferential reasoning in U1B (Bikeshare route) presented
 Disciplinary knowledge and its application 223

in Figure 9.3, the diagram shows a wide range of OSD, however the reasoning
always relates to the specific solution of a specific campus (SG++). While there is
some correspondence between OSD and SG, it is not inevitable. Even if a specific
object is referenced (SG++) that object may be more or less complex in its own
right (OSD can vary between OSD++ through OSD– – through a design). The
context (university campus) referred to in the case of U1 (Bikeshare scheme) is
extremely complex (OSD++), but the analysis of a particular slope on the selected
route is significantly simplified (OSD–). Both relate to a specific context or part of
the context (SG++).

Concluding remarks
Real artifacts and real contexts are extremely complex in their contextually embed-
ded state. Making sense of them and simplifying them in order to make design deci-
sions and to predict performance requires making sense of the contextual details, both
in terms of unspecialized ‘everyday’ familiarity, and informed by disciplinary insights.
Projects or ‘problems’ encountered in the ‘real world’ are specific, contextually embed-
ded and emergent. In their emergent form they would typically be coded SG++,
OSD++, DSDo. While many ‘problems’ encountered in the world can be resolved
without recourse to specialized knowledge, those projects or problems that fall into
the preserve of any particular profession do assume the recruitment of the specialized
knowledge associated with that profession. Those projects require the professional to
strengthen the semantic density of discursive relations significantly (DSD↑) as they
identify appropriate theoretical principles from the canon of professional knowledge
(DSD++). In order to theoretically model any potential solution, the material com-
plexity of the emergent problem (ontic relations) needs to be simplified; a process
of weakening the semantic density of ontic relations (OSD↓) by identifying relevant
aspects of the problem and discarding those that are not relevant. But this process is
not independent of discursive relations; OSD↓ is likely to be informed by introduc-
ing principled insights (discursive relations), which involves DSD↑.
Students learning to become engineers or other professionals need to learn to
identify and recruit appropriate specialized disciplinary knowledge to solve profes-
sional problems. Students should not be expected to develop this expertise without
an explicit introduction, at an appropriate level of complexity. This is especially so
early in the curriculum before students become proficient in multiple disciplines,
or even have much familiarity with the objects of their profession (what they are
and how they work). It is therefore necessary to simplify those projects intended to
mimic professional projects to an appropriate level of complexity for any particu-
lar point in a learning trajectory. The two projects I have contrasted in this chap-
ter appear at the start of the design learning trajectory and so required significant
recontextualization. The analysis shows the way in which the artifact prescription and
context description are recontextualized in the design brief affects the relationship
between ‘concrete’ object and ‘abstract’ theory, modifying the inherently dialectical
relation between them.
224 Nicky Wolmarans

The argument I am making is that when the semantic density of ontic relations
(OSD) remains stronger, the required reasoning is more complex. It is likely to
result in a simultaneous network of inferences between ontic relations and discur-
sive relations rather than a sequential chain of inferences. Of course, if the discursive
relations are not relevant, the reasoning is also simplified, but in a different way,
the reasoning never requires the semantic density of discursive relations (DSD)
to be strengthened. The suggestion is that professional reasoning requires dialecti-
cal reasoning between ontic relations and discursive relations. If ontic relations are
stripped then the reasoning becomes more procedural, while if discursive relations
are stripped the range of semantic density is reduced, and the significance of special-
ized disciplinary knowledge is compromised.
I have not suggested any prescription of how one ‘should’ simplify design proj-
ects; there are many ways to do that. But if one is aware of the potential conse-
quences of the recontextualization choices made, then one can be more intentional
about designing coherent learning trajectories. In typical engineering curricula,
based on learning decontextualized sciences, there is a tendency to weaken OSD
to the point of dislocation. Unless one takes seriously the complexity inherent in
objects in the world and the effect that they have on the complexity of reasoning
required to analyze them we will continue to be surprised when students cannot
‘just’ apply theory to the sorts of complex, contextually emergent problems that
professionals encounter in practice.
In addition to insights into the empirical challenge of educating professionals,
this study offers insights for building a better model of professional knowledge.
Models of professions based on what Bernstein (2000) called ‘regions’ argue that
specialized knowledge should be taught first and can then be ‘applied’ to external
problems. But regions face both internally to specialized bodies of knowledge and
externally to their application in fields of professional practice. This suggests that
both internal relations of coherence (conceptual – discursive relations) and external
relations of coherence (contextual – ontic relations) need to be held simultane-
ously. There is a dialectical relationship between the internally defined disciplinary
knowledge and the externally emergent problem in the world. It is thus not merely
an application of specialized knowledge onto an emergent problem. Professional
reasoning shifts continuously between the internal specialized knowledge and the
external concerns of the problem.
This study contributes to understanding ‘regions’, an underdeveloped concept
in code theory, and shows how LCT Semantics can provide a lens into the exter-
nal objects of analysis. Moreover, the categorization framework offers an analytical
language that may be useful to other professional education research and develop-
ment studies. As education moves in the direction of regionalization and education
moves towards a focus on application of knowledge in the world, the complexity
of the world needs to be taken into account. Otherwise we run the risk of sliding
into issues of personal attributes and generic skills. We risk losing the power of spe-
cialized disciplinary knowledge because very few students learn to use specialized
knowledge in the contextually complex situations that are the workplace.
 Disciplinary knowledge and its application 225

Note
1 For an earlier version of the analytical distinction between ontic relations and discursive
relations, see Wolmarans (2014).

Reference
Beck, J (2002) ‘The sacred and the profane in recent struggles to promote official pedagogic
identities’, British Journal of Sociology of Education, 23(4): 617–26.
Beck, J. and Young, M. (2005) ‘The assault on the professions and the restructuring of aca-
demic and professional identities: A Bernsteinian analysis’, British Journal of Sociology of
Education, 26(2): 183–97.
Bernstein, B. (2000) Pedagogy, Symbolic Control, and Identity: Theory, Research, Critique, Oxford:
Rowman and Littlefield.
Blackie, M. A. (2014) ‘Creating semantic waves: Using Legitimation Code Theory as a tool
to aid the teaching of chemistry’, Chemistry Education Research and Practice, 15(4): 462–69.
Froyd, J. E., Wankat, P. C., and Smith, K. A. (2012) ‘Five major shifts in 100 years of engineer-
ing education’, IEEE, 100: 1344–60.
Georgiou, H., Maton, K. and Sharma, M. (2014) ‘Recovering knowledge for science educa-
tion research: Exploring the ‘Icarus effect’ in student work’, Canadian Journal of Science,
Mathematics, and Technology Education, 14(3): 252–68.
Grinter, L. E. (1955) ‘Report on evaluation of engineering education’, Journal of Engineering
Education, 46(1): 25–63.
Harris, E. M. D. L., Grogan, W. R., Peden, I. C., and Whinnery, J. R. (1994) ‘Journal of
Engineering Education round table: Reflections on the Grinter report’, Journal of
Engineering Education, 83(1): 69–94.
King, J. (2007) Educating Engineers for the 21st Century, London: Royal Academy of Engineering.
King, R. (2008) Engineers for the Future: Addressing the Supply and Quality of Australian engineer-
ing Graduates for the 21st Century, Epping, NSW: Australian Council of Engineering Deans.
Kotta, L. (2011) ‘Structural Conditioning and Mediation by Student Agency: A Case Study of Success
in Chemical Engineering Design’, unpublished PhD thesis, University of Cape Town.
Mann, C. R. (1918) ‘A study of engineering education’, Bulletin, 11.
Maton, K. (2013) ‘Making semantic waves: A key to cumulative knowledge-building’,
Linguistics and Education, 24(1): 8–22.
Maton, K. (2014) Knowledge and Knowers: Towards a Realist Sociology of Education, London:
Routledge.
Maton, K. (2020) ‘Semantic waves: Context, complexity and academic discourse’, in J. R.
Martin, K. Maton and Y. J. Doran (eds) Accessing Academic Discourse: Systemic Functional
Linguistics and Legitimation Code Theory, London, Routledge, 59–85.
Shay, S., and Steyn, D. (2016) ‘Enabling knowledge progression in vocational curricula: Design
as a case study’, in K. Maton, S. Hood, and S. Shay (eds) Knowledge-Building: Educational
Studies in Legitimation Code Theory, London: Routledge, 138–57.
Smit, R. (2017) ‘The Nature of Engineering and Science Knowledge in Curriculum: A Case Study in
Thermodynamics’, unpublished PhD thesis, University of Cape Town.
Wolmarans, N. (2014) ‘Exploring the role of disciplinary knowledge in engineering when learning to
design’, paper presented at Design Thinking Research Symposium, West Lafayette, IN, USA.
Wolmarans, N. (2017a) ‘The nature of professional reasoning: An analysis of design in the
engineering curriculum’, unpublished PhD thesis, University of Cape Town.
Wolmarans, N (2017b) ‘Flexible curricula:Addressing the transition from engineering science to engineering
design’, paper presented at SASEE 2017, Cape Town, South Africa, 14–15 June, pp 349–57.
10
GROUNDED LEARNING
Telling and showing in the language and
paralanguage of a science lecture

Susan Hood and Jing Hao

Introduction
The case study presented in this chapter is one of a series which draws on systemic
functional linguistics (SFL) and more broadly systemic functional semiotics (SFS)
to explore the potential of different semiotic modes to support the building of
disciplinary knowledge and values in live lectures (Hood 2020; Hao and Hood
2019; Hood and Lander 2016). The object of study in this instance is the teaching
of molecular chemistry in a face-to-face undergraduate lecture. As is typical of this
pedagogic context, interaction is mediated across multiple semiotic modes includ-
ing spoken and written language and visual images (static and dynamic) on pre-
sentation slides. In addition, there is the often-overlooked semiotic potential of the
lecturer’s embodied meaning-making. Student access to this latter semiotic mode
is largely restricted to participation in live lectures. This is certainly so to the extent
that the lecturer’s body language is a feature of interaction with visible co-present
others (Wilkin and Holler 2011: 294). However, to date this mode of pedagogic
interaction has received relatively little research attention.This study aims to explore
how technical scientific knowledge is built in the intersemiosis of the oral mode of
telling and the visual embodied mode of showing. We pursue this work with a goal
to inform pedagogic policies and practices for supporting students as they move
into new uncommon-sense understandings of their disciplinary fields, particularly
in environments which increasingly promote online learning.

Functional perspectives on the discourses of science

Current attention to discourses of science, especially in educational contexts,
builds on a long history in SFL. Notable early contributions include a trilogy of
titles foregrounding modes of interaction: Lemke’s Talking Science (1990), Halliday
 Grounded learning 227

and Martin’s Writing Science (1993) and Martin and Veel’s Reading Science (1998).
Significant recent additions have taken a more subject-specific approach, privileging
variation within the generalized field of science (e.g. Doran 2018 on the language
of mathematics in physics education; Hao 2020a on the discourse of undergraduate
biology; see also Chapters 6 and 7 this volume). At the same time these recent works
have extended theoretical modelling of knowledge from an SFL perspective, with
contributions informing this study. There has also been a refocusing of earlier inter-
est in modes of interaction, expanding research beyond spoken and written texts to
consider the interplay with other kinds of semiosis. In discourses of science these
include images in textbooks (e.g. Ge et al. 2018), animations (He and van Leeuwen
2020) and embodied expressions of value (Hao and Hood 2019). Here we continue
to explore the cooperation of spoken language and body language, complement-
ing work to date with a focus on the intersemiotic construal of technicality in the
context of a lecture on molecular chemistry.

Stratification and metafunction in language

Intrinsic to an SFL approach is firstly a relational theory of meaning. In other words,
verbal or visual signs are not understood as symbolizing a reality ‘outside’ of semiosis
but as bringing ‘reality’ into being – as making meaning (Halliday 1978). Following
Saussure (1974), meaning is understood as valeur, as emergent in the relationship
which holds between any two choices in a semiotic system of choices.
Secondly SFL’s concept of meaning is metafunctional. Ideational meaning con-
strues a world around us and inside us. Interpersonal meaning enacts personal and
social relationships with others. Textual meaning has a composing role, organizing
other metafunctional choices as a coherent flow of meaning and assigning rela-
tive prominence (Halliday 1994; Halliday and Matthiessen 2014). This metafunc-
tional perspective on language is complemented by SFL’s modelling of language as
a tri-stratal system of systems – namely, discourse-semantic systems focusing on
discourse structure, lexicogrammatical systems focusing on clause, group/phrase
and word structure, and phonological and graphological systems of expression
(Martin 1992; Martin and Rose 2007). The intrinsic metafunctionality of language
across strata correlates with the extrinsic functionality of context as register; the
register variables of field, tenor and mode correlate respectively with ideational,
interpersonal and textual meanings in language. These aspects of the theory are
captured in Figure 10.1.
Our focus on the building of disciplinary knowledge in molecular chemistry
orients this study primarily towards ideational meaning. From a tri-stratal perspec-
tive on language (discourse semantics – lexicogrammar – phonology/graphology),
we approach the question of convergence of meanings in speech and embodied
expression at the stratum of discourse semantics (Figure 10.1). This offers a level
of abstraction, a “view from above” (Matthiessen 2007: 2) that can account for co-
expression across different semiotic systems. Here we focus on intersemiotic rela-
tions in spoken language and paralanguage only, space preventing inclusion of for
228 Susan Hood and Jing Hao

FIGURE 10.1 Stratification and metafunction in systemic functional linguistics

example visual and written verbal semiosis in presentation slides. We analyze first
the spoken discourse of the lecture, then explore concurrent embodied expressions
with reference to choices in the discourse system of ideation.
Before proceeding to more detailed discussion of ideational meaning-making in
spoken language and the body, we introduce briefly a systemic functional modelling
of body language as paralanguage.

Body language as paralanguage

As with language a social semiotic perspective on body language is metafunctional
(e.g. Martinec 2000, 2001, 2002; Cléirigh 2011; Zappavigna et al., 2010; Hood 2011;
Martin and Zappavigna 2019; Ngo et al. 2021). A brief account of the modelling of
the semiotic potential of the body follows, with an explanation of why it is hence-
forth referred to as paralanguage in this chapter.
Setting aside protolinguistic expression (discussed in Ngo et al. 2021), SFL stud-
ies of paralanguage have identified two distinct systems (Cléirigh 2011; Martin and
Zappavigna 2019). The first is one in which speech must be co-present for body
movements to be meaningful. Here a full interpretation of textual prominence
involves the synchronicity of body movements with the phonological rhythm and
stress of speech (Halliday and Greaves 2008; Ngo et al. 2021) and a full interpreta-
tion of interpersonal tone involves body movements (e.g. of eyebrows, hands) cor-
responding with pitch movement (e.g. rising/falling). Cléirigh (2011) refers to this
system as ‘linguistic’ body language. Following Martin and Zappavigna (2019) we
refer to these systems as sonovergent paralanguage, using terminology which
makes explicit the necessity of accompanying speech. Sonovergent paralanguage
can express textual or interpersonal (but not ideational) meaning (Martin and
 Grounded learning 229

Zappavigna 2019). For ideational meaning we need to explore a second sys-

tem, one which can occur with or without immediately co-present (convergent)
speech. Cléirigh (2011) refers to this as epilinguistic body language. We refer to
these resources as semovergent paralanguage (Hao and Hood 2019; Martin and
Zappavigna 2019; Ngo et al. 2021).
Semovergent paralanguage relates meaningfully to discourse semantics across all
three metafunctions (Martin 1992; Martin and Rose 2007). Interpersonally, it can
realize options in the system of appraisal – through facial expression of attitude
as affect, through tension or size of gestures expressing graduation, and through
postures and positionings of the body expressing expansion or contraction of space
for negotiating values in engagement (Painter et al. 2013; Ngo et al. 2021; Hao
and Hood 2019). Textually, embodied paralanguage can realize choices in identi-
fication, expressed in kinds of pointing gestures; it also converges with choices in
periodicity through longer wavelengths of whole-body movement and position-
ing in space (Martinec 2002; Hood 2011; Ngo et al. 2021). In this chapter we focus
on the potential for expressing ideational meanings concurrently in speech and
paralanguage.1
The systems of sonovergence and semovergence are both theorized as para- to
language in that whenever speech is co-present there is synchrony with the pro-
sodic phonology and in periods of absent speech, the rhythm of prior speech is
typically maintained. Furthermore when speech is not immediately co-present,
embodied expressions are interpreted as meaningful in ways dependent on the
flow of spoken discourse (Martin and Zappavigna 2019; Hao and Hood 2019; Ngo
et al. 2021).

The study
Our data come from a video recording of a 50-minute live lecture in molecular
chemistry to a class of about 180 undergraduate students. Multiple semiotic modes
are deployed and interact to different degrees as the discourse unfolds. Throughout,
the lecturer deploys presentation slides displaying written language, static images
and video clips. The slides are projected onto two large screens set behind and
above the head of the lecturer. The lecturer takes up a position around a lectern
in a center space at the front of the room. From there he faces the general student
body, while periodically re-orienting himself to identify aspects of visual or verbal
information on slides or to interact with individual students. The lecturer’s spoken
language and any audible student talk were transcribed and reviewed along with
corresponding slides to identify the textual structuring of the content. Phases of
discourse which share a similar function in building the scientific field and in which
paralinguistic expressions accompany language were identified. The process mod-
elled in this chapter is first to analyze linguistic expressions of scientific knowledge
with reference to the discourse system of ideation and then to explore patterns of
convergence with paralanguage. Both to foreground the contribution of the body
and for reasons of length we limit the scope to the modalities of spoken language
230 Susan Hood and Jing Hao

and paralanguage, while acknowledging the relevance of written text and images on
slides in the multimodal discourse of knowledge building.

Building the field of molecular chemistry in language

and paralanguage
Interest in how the field of molecular chemistry is built intersemiotically in a live
lecture privileges an ideational perspective and explores choices in the discourse
system of ideation.To contextualize this analysis, we first take account of the mean-
ing potential at the higher stratum of register – focusing on field (Figure 10.1). In
a pedagogic context such as a live lecture multiple fields can be involved in the
discourse. One is the field of pedagogy which organizes classroom interactions and
planning. This is realized for example through the underlined wording in Let’s talk
about solid chemistry, or in the lecturer’s relatively frequent use of personal pronouns
like we/us/our. At times pedagogic discourse is drawn on to present disciplinary
knowledge of molecular chemistry in more commonsense terms as ‘doing’, as in
we can also get covalent network solids, or ‘having’ as in we have solids, liquids and gases.
Where helpful in analyses we have added bracketed paraphrasing which reinterprets
from pedagogic to scientific discourse, as in we can also get (there are also…).
Setting aside the pedagogic discourse, we tease apart two specific topics involved
in disciplinary knowledge building in this lecture. The first is that of established
(observed) knowledge. In our data this includes taxonomies and activities in the field
of molecular chemistry.The second is that of experimentation, the (observing) pro-
cedures which underpin established knowledge (Hao 2020a; Hood 2010). Each of
these topics can be explored from two simultaneous orientations, one static and the
other dynamic (Doran and Martin this volume). A static perspective models field as
items, their properties, and their organization into taxonomies. A dynamic per-
spective models field as activities, their properties and how activities develop as
time unfolds. Specific sciences are distinguished by distinctive types of items, activi-
ties and properties and the relationships which hold between them (Martin 1992,
Hao 2020a, this volume, Doran and Martin this volume).
We elaborate briefly on the basic discourse semantic units of ideation
which realize field. From a static perspective, relevant units include entities
(e.g. molecules, atoms, sodium chloride), qualities which extend those entities (e.g.
sodium is smaller) and dimensions which are dependent on those entities (e.g. a
type of solid).2 Entities, their qualities and dimensions are presented and related
in discourse in what are referred to as state figures (e.g. sodium is smaller; one
type of solid is ionic solid). From a dynamic perspective the basic discourse units
are occurrences (e.g. the lattice expands), occurrence figures which configure
occurrences and entities (e.g. one chlorine ion is attracted to one sodium ion), and
finally sequences of figures (e.g. as our solution is being formed … our solvent begins
to pull apart our solid particles, and surround them). These units of ideational meaning
are introduced and explained in turn with reference to a number of indicative
phases of the lecture’s talk.
 Grounded learning 231

From analyses of ideation in spoken language we explore convergent expressions

in the lecturer’s embodied paralanguage. We consider relations in meaning between
what is told and what is shown through embodied expressions. The theorizing of
meaning-making in intersemiotic convergences continues to present a consider-
able challenge to the broad field of multimodality (Zhao 2010; Martin 2011). The
approach taken here applies the concepts of instantiation and commitment intro-
duced below (and explained in more depth in Martin (2011),Painter, et al. (2013)).
Instantiation refers to a hierarchy of generalization in a system of meaning.
The upper point of the cline embraces the full meaning potential of a system – a
potential progressively constrained by selections of genre and register, through types
of texts to actual instances of discourse (Martin 2008, 2011).This concept is relevant
to explorations of how much meaning potential is committed from the system in
instances of discourse (Martin 2008; Hood 2008).
Commitment offers a framework for comparing instances within and across
semiotic modes. In this chapter we examine the relative commitment of ideational
meaning in language and paralanguage by considering the degree to which activi-
ties and items of a given field construe a technical reality or an everyday com-
monsense one. Technicality condenses a considerable mass of ideational meaning
through relations to other phenomena, within a given text and in the field generally
(Martin 2017). Much more ideational meaning potential is committed in naming
technical phenomena than in naming everyday commonsense things and events.
This variation strongly differentiates the two modalities of language and paralan-
guage in the data. The verbal discourse of the disciplinary field is highly technical
while the paralanguage expresses an everyday commonsense ideational reality, as is
made evident in comparative instances throughout the chapter.
Variations in how ideational meanings are instantiated in discourse can also be
considered from the perspective of textual meaning and mode. The focus then is
on the role which language plays in an interaction, with language accompanying
action at one end of a continuum and language as reflection at the other. These
variations are associated with ‘more written-like’ or ‘more spoken-like’ discourse
(Martin 1992; Martin and Rose 2007).
In comparing the relative commitment of ideational meaning in instances of lan-
guage and paralanguage, we also consider the mode of expression as relatively con-
gruent or experientially metaphorical. This contrast is exemplified below for
the discourse semantic systems of ideation and connexion (Martin and Matruglio
2020; Hao 2018, 2020a; Hood 2020).
For ideation, the expressions in focus are underlined below.

(we) added more and more sodium acetate

vs.
the addition of one single sodium acetate crystal…

The first instance (added) is a congruent realization of Process in a material clause.

The second (addition) is experientially metaphorical in that it construes two layers
232 Susan Hood and Jing Hao

of meaning – a Participant in the clause and an occurrence figure in the discourse

semantics.
For connexion, a contrast between a congruent and a logically metaphorical
expression of a causal connexion is exemplified in:

Water molecules huddle around the ions and so the lattice begins to break apart
vs.
The action of the water molecules leads to the break-up of the lattice.

In the first, the causal connexion between two figures is construed congruently
through a conjunction. In the second, the causal connexion is construed metaphori-
cally through a grammatical process.
While language allows us to shift between more or less congruent and meta-
phorical construals of an ideational reality – variations we associate with ‘more
written-like’ or ‘more spoken-like’– paralanguage does not. We can only gesture
an entity metaphorically construed in languages as if it were a congruent entity. In
other words, paralanguage always construes a ‘congruent’ construal of an ideational
‘reality’ – one grounded in a here-and-now. In this it may be convergent or diver-
gent with accompanying verbal text.3
In the sections to follow we progressively explore the lecturer’s spoken lan-
guage and concurrent paralanguage in terms of the relative congruence that they
instantiate. Several indicative extracts of lecturer talk are analyzed for choices in
specific units of ideation. Concurrent paralinguistic expressions are then described
and interpreted.We consider how the two modalities collaborate in the construal of
ideational meaning, and how this supports students as they move into new uncom-
mon-sense understandings of their disciplinary fields.

Entities and their relation in discourse

The ideation unit entity is differentiated into thing entities naming items (such
as molecules, atoms), activity entities naming activities (the technical naming of a
sequence of occurrences as in crystallization) and semiotic entities naming mean-
ing (e.g. an example; the periodic table) (Hao this volume). The first two types are of
particular relevance in this study. We begin with a focus on thing entities, their role
in a static perspective on field knowledge, and their convergence with paralinguistic
expressions. Activity entities and concurrent paralinguistic expressions are explored
later where a dynamic perspective is adopted.
The first episode of talk for analysis of ideation comes at about the four-minute
mark in the 50-minute lecture. It follows a brief review of the previous lecture
and sets up the focus of a segment to come. This phase is densely packed with
technical entities. In the notation system deployed in text (A) and other phases of
analysis, entities are shown in bold. As our focus is on disciplinary knowledge, the
discourse of the pedagogic field is bracketed off – as in [But now let’s talk about]. In
occasional superscripts we insert meanings elided in talk but recoverable from the
 Grounded learning 233

co-text. As will become evident, this annotation supports the identification of rela-
tions which hold between entities in the phase of text. In round brackets we offer
some paraphrasing from pedagogic into scientific discourse and substitutes of highly
generalized terms such as something and thing, as in so something (/examples) like solid
sulphur).

text (A)

[But now let’s talk about] solid chemistry, [just for a little bit]. [So] even though
solid is just a state of matter – [so we have] (there are) solids, liquids, and
gases – the internal structure of [our] solids can be different depending on
what atoms [we have] (are) present. [So] there are a few different types of solid
[and we’re going to go through them in the next few slides]. The one type of solid
[we’re all fairly acquainted with at this point] is ionic solids. [So] the main
example of ionic solid is sodium chloride. [We can also get] (there are also) molecu-
lar solids, so something (/examples) like solid sulphur. [We can also get] (there
are also) covalent network solids, so these are things (/examples) like dia-
mond, sand, graphite. Metallic solids is [our] (the) last one type of solid.

As in any discourse semantic system, semantic choices may be realized in more than
one kind of grammatical construction (Hao this volume). Our analysis of language
focuses primarily on choices at the discourse semantic stratum.

Entities in paralanguage
While the grammatical realization of entities in discourse varies, all except one of
those bolded in text (A) (atom) are concurrently expressed in paralanguage; and for
all of these the embodied entity is sculpted through the shaping of the palms of both
hands, oriented to each other to some degree as in Figure 10.2.
Figure 10.2 illustrates a default expression of a sculpted paralinguistic entity in
our data – i.e. one which generally requires minimal effort to express, with relaxed
curved hands, palms oriented to each other, with forearms extended out from the
body from waist to chest height and at torso width.Variations on this shape, size and
position commonly associate with the depiction of particular qualities of an entity.
In some cases an entity is expressed through a single-hand shape – as in the
upwardly (supine) curved palm in Figure 10.3, and later, in Figure 10.12, an inverted
(prone) palm. We hereafter refer to these as paralinguistic entities.
Elsewhere in the data paralinguistic entities express more defined shapes, typically
requiring two hands. Concurrent with the entity box in (1) is a paralinguistic entity
in which straightened hands more sharply define the boundaries, as in Figure 10.4.
Relative to the default expression (Figure 10.2), the paralinguistic entity in Figure
10.4 commits more meaning potential – here associated with the straight-sided
quality of the entity box.
234 Susan Hood and Jing Hao

FIGURE 10.2 Two hands sculpting a paralinguistic entity

(1) it’s a bit like taking a 1cm by 1cm box out of our ionic solid (…)

FIGURE 10.3 One hand sculpting a paralinguistic entity

Concurrent with the underlined dimension in (2) is a sculpted paralinguistic

entity in which digits of both hands depict a kind of pyramid shape, as in Figure 10.5.
In relation to Figure 10.2, the paralinguistic expression commits more meaning
potential – here associated with the quality of a 3D entity.
 Grounded learning 235

FIGURE 10.4 A paralinguistic entity convergent with ‘box’

(2) it (the inner lattice) [wants to] form into three dimensions

FIGURE 10.5 A paralinguistic entity convergent with ‘form into three dimensions’

An alternative means for depicting a paralinguistic entity is to ‘draw’ an object

in the air with one or more fingers or hands. Sculpted or drawn entities can vary
in size, shape and location in the gestural space. A smaller entity for example can
be sculpted in a single hand (Figure 10.3 above), and the depiction in Figure 10.2
might be expanded or reduced in size to commit more meaning of big-ness or
small-ness. The depiction in Figure 10.3 might be duplicated with entities in both
236 Susan Hood and Jing Hao

the left and right hand. This visual depiction might co-commit a meaning of two
entities in concurrent verbal text. However, as this is the maximum number of enti-
ties able to be expressed (statically) in the two-handed body, the gesture can at times
under-commit a meaning of lots of (i.e., more than two) in a verbal text.
With reference to the spoken text (A), the technical entities of the disciplinary
field of molecular chemistry (solids; atoms; sodium chloride; etc.) each commit a con-
siderable mass of ideational meaning, infused as each is with a complex set of rela-
tions to other technicality in the field (Martin 2017).The commitment of ideational
meaning in these instances is considerably greater than in concurrent paralinguistic
entities. At the same time the concurrent paralanguage construes the meaning of
entities in commonsense terms, depicting each as a hand-held thing. This estab-
lishes an important complementarity of technical telling and everyday showing. It
grounds the technicality in commonsense.

Relations between entities in language

The discourse of text (A) not only construes multiple technical thing entities, but
also explicitly construes the ways entities relate to one another. A detailed set of rela-
tions among entities is established in this relatively short phase. The related entities in
Figure 10.6 all construe relations of classification building from the general entity matter.
In text (A) each entity is meaningful not in isolation but in relation to its super-
ordinate category, its co-classes and subclasses. All the entities in Figure 10.6 co-
classify each other. For example, matter and solid co-classify as superordinate and
subordinate respectively, and ionic solids and molecular solids co-classify as co-subtypes.
The co-classification between entities can be grammaticalized in a number of ways:
for example in a relational process, as in (3).
(3) solid is just a state of matter
Another possibility is a nominal group complex as in (4).

FIGURE 10.6 The taxonomy of matter configured in text (A)

 Grounded learning 237

(4) [We can also get] molecular solids, [so] something like solid sulphur

Additionally, relations between entities in the spoken discourse can be abduced

(interpreted from the co-text), in some cases across extended stretches of text. For
example, in text (A), molecular solids and covalent network solids are introduced as co-
subclasses of the category of solids. The clues supporting this abduction include the
distinguishing Classifiers of the solids (underlined) and the textual reiteration of We
can also get, bracketed in (5).
(5) [We can also get] molecular solids ... [We can also get] covalent network
solids ...

A further discourse semantic choice of ideation which supports the construal

of taxonomic relations among entities is that of dimensions (Hao this volume).
Dimensions explicitly name those taxonomic relations. The instances of dimension
underlined below, type and example, explicitly name the relation between the entities
involved as one of classification. The superscript notation shows that the relevant
entities are elided, having been introduced in preceding text.
(6) [So] there are a few different types of solid
(7) the main example of ionic solid is sodium chloride.

Relations between entities in paralanguage

A taxonomic relation of classification between entities in text (A) is almost always
concurrently expressed paralinguistically through paralinguistic entities which
depict relative size and/or shape and/or location. Several instances exemplify this.
Two kinds of classification relation are construed in the immediately sequential
state figures in (8) and (9).
(8) even though solid is just a state of matter …
(9) we have solids, liquids, and gases
In (8), there is the classification relation of the class matter with the subclass solid.
In (9), the tshree entities solids, liquids, and gases are related to each other as co-
subclasses to matter.
In (8), a class-subclass relation is made explicit in the dimension state of the entity
matter. In convergent paralanguage, one entity is expressed concurrently with solid
and another with matter.That convergent with solid differs in size, shape and location
from that with matter, as described in Table 10.1.
Correspondences evident in Table 10.1 generalize to the data as a whole. When
the verbal construal of a class-subclass classification relation converges with the
expression of paralinguistic entities, they are depicted differently with respect to
one or more of the features of size, shape and location. Non-equivalence on a
taxonomic hierarchy converges with non-equivalence in depicted entities. In some
instances size is a variant, but more consistently there is a variation in shape, in line
238 Susan Hood and Jing Hao

TABLE 10.1 Paralinguistic expressions of the relation of solid to matter as subclass to class

Language
entity solid matter
relation Subclass Class
Paralanguage
size smaller than default (median) larger than default (median)
shape sculpted: palms facing; thumbs sculpted: more open curved palms;
up; defined boundaries with undefined boundaries
straightened fingers
location located at chest height located below chest height

FIGURE 10.7 Equivalence in the paralinguistic depiction of co-class entities in (9) –

viewed from right to left

with that described in Table 10.1. The paralinguistic entity convergent with class is
depicted as sculpted in soft curved palms (akin to Figure 10.2), and that convergent
with subclass is depicted with more sharply defined borders (akin to Figure 10.4).
This suggests that a paralinguistic entity with more defined boundaries can not only
commit a specific quality of an entity (as having straight sides as in ‘box’) but can
more generally express a ‘kind of ’ taxonomic relation.
In (9), the state figure (we have solids, liquids, and gases) presents a different classifica-
tion relation. Here, the three entities are related to each other as co-subclasses of matter.
The co-subclass relation establishes them as equivalent on a classification hierarchy.This
equivalence is concurrently expressed in the depiction of three paralinguistic entities,
each showing equivalence in size, shape and location (Figure 10.7 and Table 10.2).
The taxonomic relations between the entities in (9) (solids, liquids and gases) are
committed differently in each semiotic mode. In the spoken text, their equal status
in a taxonomic hierarchy has to be inferred from verbal co-text.The paralanguage is
more explicit in its depiction of three entities equivalent in size, shape and location.
In this instance a higher level of ideational commitment is construed in the showing
than in the telling of the relations between the entities.
 Grounded learning 239

TABLE 10.2 Paralinguistic expressions of co-subclass relations among solid, liquids and gases

Language
entity solids liquids gases
relation co-subclass co-subclass co-subclass
Paralanguage
size equivalent in size: median
shape equivalent in shape: sculpted in two hands; relatively defined sides
location equivalent in location: along same horizontal plane; in front of
body just above waist height
[Note: torso rotates from the lecture’s left to right with each
paralinguistic entity. From the students’ view the entities move
from right to left in space with equal distance between them]

A further instance of co-subclass relations is shown in (10).

(10) How do we go from (/distinguish) a primitive cubic to a body-centred cubic to a

face-centered cubic?

The concurrent paralanguage shows a similar pattern of intersemiotic convergence

to Figure 10.7. The bolded entities in (10) construe three co-subclasses of cubic
(structure). Again, the co-relations are expressed through paralinguistic entities of
equivalent size, with relatively defined boundaries, and located equidistant from one
another along a horizontal plane (Figure 10.8).
In (9) and (10) the language construes the co-subclass entities as an exhaustive
set of kinds of matter in (9) and kinds of cubic in (10). In each case the set of concur-
rent paralinguistic entities are depicted with more sharply defined boundaries, with
relatively straightened palms.
In other instances of similar relations, the entities are construed as non-exhaustive
sets, in other words as some of, or examples of a set of co-subclass entities, as in (11).

FIGURE 10.8 Equivalence in the depiction of an exhaustive set of co-related paralinguis-

tic entities (in 10) – viewed from right to left
240 Susan Hood and Jing Hao

(11) [We can also get] covalent network solids, [so] these are things like diamond,
sand, graphite
In (11) an occurrence figure followed by a state figure construes diamond, sand,
graphite as a set of co-subclass entities to the class covalent network solids. Their status
as examples only is construed in the generalized reference to more such entities in
things like. The non-exhaustive status of this set corresponds to a difference (from
Figures 10.7 and 10.8) in the depiction of paralinguistic entities in Figure 10.9.

FIGURE 10.9 Equivalence in the depiction of a non-exhaustive set of co-subclass entities

(in 11) – viewed from left to right

In contrast to the exhaustive sets of co-related entities depicted in Figures 10.7

and 10.8 as paralinguistic entities with relatively defined boundaries, the non-
exhaustive set depicted in Figure 10.9 is sculpted in single hands, but more signifi-
cantly in hands with an open, curved palm. There is also a difference in how the
paralinguistic entities are located in relation to each other in the gestural space. In
Figures 10.7 and 10.8 they are spaced out equidistant from one another along a
horizontal plane. In Figure 10.9 they are located in the same general space in front
of the body, depicted in alternating left, right then left hands.
The general pattern of convergence in (11) and Figure 10.9 is repeated in rela-
tion to (12). The text again construes a non-exhaustive set of co-related entities,
with pure and compound construed as a non-exhaustive set of kinds of solution.

(12) [We spoke about] solutions, and pure, compound, all those sorts of things, [way
back in the first lecture]

Their shared sub-class status of pure and compound is construed in the dimension sorts
of. Their status as a non-exhaustive set is construed in the generalized reference to
more entities of this kind in all those sorts of things.
Instance (12) is similar to (9), (10) and (11) in that the concurrent paralinguis-
tic entities for pure, compound, and generalized others in all those sorts of things are
depicted as equivalent in size, shape and relative location (Figure 10.10).
 Grounded learning 241

FIGURE 10.10 Equivalence in the depiction of a non-exhaustive set of co-class entities

(in 12) – viewed from left to right

Each paralinguistic entity in Figure 10.10 is expressed as more or less equivalent

in shape and size, with the proviso that in the first (‘pure’), the lecturer had lowered
his left hand to notes on the desk. Here the right hand sculpts the paralinguistic
entity palm down over the lowered left hand. The left hand is then freed to come
into play to sculpt the other entities (compound and those sorts of things) as two-
handed. Each entity is located more or less in the same space in front of the torso.
A rotation in the angle of the two sculpting hands shifts the location of each para-
linguistic entity a little to the right or left side of the torso.
The instances of intersemiotic convergence in Figures 10.7, 10.8, 10.9 and 10.10
all depict co-relations within sets of paralinguistic entities through equivalence in
their depictions. However, where the set of co-related entities is exhaustive, the
paralinguistic entities are sculpted in hands which depict more defined boundaries
and are viewed as placed more or less equidistantly along a horizontal plane (Figures
10.7 and 10.8). Both features reinforce a stronger categorical perspective on the
co-relations between entities concurrent with the verbal text. Where the set of co-
related entities are non-exhaustive, the paralinguistic entities are sculpted with less
defined boundaries in softer curved palms. They are viewed as located more or less
in the same location in front of the body (Figures 10.9 and 10.10). Concurrent with
the verbal text these features foreground commonality or likeness over categorical
distinctions in entity relations.

Entities and related qualities

Qualities of entities in language
A focus on the qualities of entities shifts our perspective from relations of classifica-
tion to those of composition. At about eight minutes into the lecture, after establish-
ing the taxonomic relations in text (A), the lecturer narrows the focus to one type of
solid – ionic solids, and discusses its delicate sub-classes based on various composi-
tions.The critical knowledge to be established in this part of the lecture is not only a
242 Susan Hood and Jing Hao

recognition of particular types of composition, but importantly an understanding of

why they are different. To do this the lecturer describes qualities of the ions which
compose them. This phase is presented as text (B), with entities in bold, qualities in
bold and italics, and dimensions of entities underlined.

text (B)

[So] the structure of our internal solids are dependent on two things – the first
one is charge of relevant atoms and the second one is size of relevant atoms. [So] let’s look at
size of relevant atoms first. [So] we’ve got sodium chloride, and cesium chloride. [So]
sodium is very small, and cesium is very big, further down on our period table.
These are what their ionic structures look like (pointing momentarily to images on
PPT slide). We have the sodium chloride from the previous example. Each
sodium has six chlorines, each chlorine has six sodiums. But because the size of
sodium is smaller it (/sodium chloride)’s able to have that six to one ratio. The size
of Cesium is bigger. So cesium takes up more space than sodium. So it (/cesium)’s

going to come into contact with more ions. (…) [So] because it (cesium)’s in
contact with more ions, its (cesium chloride’s) structure is going to be
different.

Text (B) begins by introducing two dimensions of entities, those of size and charge,
with size the focus of the phase. Dimensions in text (A) mostly name classification
relations (type of; example of). In text (B) the dimension of size names a measurable
property of an item in field (Doran and Martin this volume). Properties of items are
realized in discourse by qualities. In (13), qualities such as very small, very big, smaller
and bigger extend the meanings of the entities sodium and cesium.

(13) … the size of sodium is smaller (…) The size of Cesium is bigger.
The qualities in (13) are realized grammatically through Attributes in relational
attributive clauses, in which the Head of the nominal group is an Epithet. Other
grammatical realizations of qualities follow in later instances.

Qualities of entities in paralanguage

In spoken language a quality is always ascribed to an entity, either in the grammar or
in the co-text. In paralanguage too, an entity is required for the depiction of a quality.
The entity and the quality may be expressed as separate paralinguistic entities (one
convergent with the entity and one with the quality).This is the case for the state fig-
ures in (14) and (15). Each underline indicates concurrence of a paralinguistic entity.

(14) sodium is smaller

(15) cesium is bigger
 Grounded learning 243

While not illustrated here, the quality smaller is convergent with a sculpted paralin-
guistic entity which is smaller in size than that convergent with bigger.

In other instances, state figures which construe an entity and a quality can be con-
currently depicted in a single paralinguistic entity.This is the case for both (16) and (17).

(16) sodium is very small

(17) cesium is very big
The state figures in (16) and (17) occur sequentially in text (B).The concurrent paralan-
guage for the sequence of figures is shown in Figure 10.11.

Concurrent with (16), in the faded depiction on the right, a small paralinguis-
tic entity is sculpted in a single hand. The lecturer’s torso is oriented to his left.
Concurrent with (17) a larger paralinguistic entity is sculpted with two hands. The
lecturer’s torso is oriented to his right. The meaning of relative small-ness or big-
ness is co-committed in language and paralanguage. The contrasting locations of
the paralinguistic entities as the torso rotates left then right additionally commit the
contrastive relation of the qualities of small and big.

FIGURE 10.11 Paralinguistic entities concurrent sequentially with (16) and (17) – viewed
from right to left

In another instance the quality sticky of the entity adhesive force is realized verbally
through the grammatical Attribute bolded in (18).

(18) The adhesive force is the intermolecular forces that bind a substance to some-
thing else – how sticky is it to another surface.

Concurrent with ‘how sticky is it’ a paralinguistic entity is expressed dynamically as an

entity in motion – an ‘entitied occurrence’ in the discourse semantics as explained
shortly (Figure 10.12).
244 Susan Hood and Jing Hao

FIGURE 10.12 A paralinguistic expression depicting a quality as an entitied occurrence

concurrent with ‘how sticky is it’

The paralinguistic entity in Figure 10.12 is sculpted in one-hand, with a prone

(down-facing) palm and fingers extended. The occurrence is depicted dynamically
in the lifting of the paralinguistic entity in a restrained motion, with the extended
fingers coming together. In the flow of spoken language, the lecturer’s construal of a
technical field in adhesive force [as] intermolecular forces that bind a substance to something
else is reinterpreted into a commonsense one in how sticky is it. The technicality of
the former commits considerably more ideational meaning potential than does the
commonsense of the latter. The paralanguage (Figure 10.12) co-commits the com-
monsense not the technical construal. The dynamic perspective on field introduced
in this instance is the focus for the next section, but we’ll briefly consolidate the static
perspective on intersemiosis in building disciplinary knowledge taken to this point.
The spoken language of the lecture on molecular chemistry construes multiple
technical entities, qualities which extend their meanings, and relations which hold
between them. An understanding of all is required in appreciating the relevant estab-
lished knowledge of the disciplinary field. The required technical load is consider-
able, both in terms of the mass of meaning condensed in individual entities (Martin
2017), and also in the density of technical entities in the discourse. We note the
means by which this mass of technicality is complemented in paralanguage through
the construal of commonsense entities, qualities and relations between entities.
While not discussed to this point, in all instances the paralinguistic showing of
ideational meanings also construes them as congruent – as realizing phenomena in
the here-and-now of a shared material space. The intersemiosis of spoken language
and paralanguage in the lecturer’s discourse bridges the technical and the everyday.
It grounds meanings in a congruent material ideational reality in support of build-
ing knowledge of specialized items and properties in the field.
 Grounded learning 245

Dynamic construal of knowledge in language

Taking a dynamic perspective on disciplinary knowledge, the focus shifts to sci-
entific activities and their unfolding through time. We begin by considering how
activities in the field of molecular chemistry are realized in the discourse of spoken
language and then in the intersemiosis with concurrent paralanguage.

Occurrences and occurrence figures in language

In language, activities are expressed in the ideational unit of an occurrence figure,
which configures an occurrence and associated entities. Occurrence figures can
also unfold in a sequence linked through logical relations. This ideational unit is
explored with reference to a number of phases of the lecturer’s talk.

Occurrence in language
At around 30 minutes into the lecture, having built knowledge of different types of
solid, their qualities and internal structures, the lecturer begins a stage in which he
explains an activity which relates to ‘intermolecular forces’ and the formation of a
solution. The lecturer uses the example of salt dissolving in water, as presented in
text (C). Here and elsewhere in this section occurrences are bolded and underlined
and associated entities bolded without underlining.

text (C)

And as our solution is being formed – [so we think about] dissolving salt – our
solvent begins to pull apart our solid particles and surrounds them (solid par-
ticles). [So our depiction of] salt dissolving [is we] start with our lattice. (…) Those
are our water molecules huddling around our different ions to break it (lattice)
apart. And it (lattice) will be uniformly dispersed throughout the solution.

Occurrences in text (C) are realized grammatically by Processes in material clauses.

They represent generalized scientific phenomena – phenomena observable in dif-
ferent locations in time and space.

Occurrence figures in language and paralanguage

An occurrence figure is a larger ideational unit which construes an activity. It
configures an occurrence with one or more associated entities. These entities may
function in one of two roles: they may precipitate an occurrence, or they may
be precipitated by one. In (19), the single entity solution is the precipitated. In
(20), solvent is the precipitator of the occurrence pull apart and solid particles are the
precipitated.
246 Susan Hood and Jing Hao

FIGURE 10.13 The paralinguistic occurrence convergent with ‘being formed’ (in 19)

(19) … as our solution is being formed

(20) our solvent begins to pull apart our solid particles.
Almost all instances of occurrence in text (C) are concurrently construed in
paralanguage. The paralinguistic occurrences are realized through motion, typi-
cally expressed in a hand and forearm or arm. Potential variations include the tra-
jectory of the motion through the gestural space (linear, curved or circular), the
kind of flow depicted (relatively restrained or accelerated) and whether the move-
ment is iterated or not (Ngo et al. 2021). For example, concurrent with the bold
and underlined occurrence in (19), the lecturer rotates both hands and forearms
around each other in a clockwise direction outwards from the body at chest height
(Figure 10.13).
This circular, iterative, and relatively accelerated motion depicts an occurrence as
unfolding in time. The gesture co-commits with the verbal being formed a meaning
of ongoing-ness.
In a second instance, in (20), the occurrence figure construes two entities, one
(solvent) functioning as precipitator and the other (solid particles) as precipitated.
These are configured with the occurrence (pull apart).The concurrent paralinguistic
expressions are illustrated in Figure 10.14.
In Figure 10.14, in the faded image on the right and concurrent with solvent, a para-
linguistic entity is sculpted in two hands. Subsequently on the left and concurrent with
the occurrence pull apart, two paralinguistic entities, each sculpted in a single hand,
move away from each other in contrary directions, depicting an entitied occurrence.
In the spoken language in (20), the entities (solvent; solid particles) are techni-
cal. The concurrent paralanguage depicts entities in commonsense terms, as things
which can be held in the hand/s. The meaning potential is much greater in the
 Grounded learning 247

FIGURE 10.14 Paralinguistic expressions convergent with ‘solvent’ and ‘pull apart our
solid particles’

verbal text than in the concurrent paralanguage.4 However the occurrence, pull
apart, is construed in commonsense terms. It does not name a specialized activity
in the field. Here there is a co-commitment of the commonsense meaning in both
language and concurrent paralanguage.
The entities which configure into occurrence figures in text (C) are all con-
strued as technical things (solution, salt, solvent, solid particles, lattice, water molecules,
ion), and in the concurrent paralanguage the paralinguistic entities are all construed
as everyday commonsense things – things held in the hand. The spoken language
commits much more ideational meaning of entities than does the concurrent
paralanguage. However, the occurrences in text (C) include those which are spe-
cialized (dissolving, uniformly dispersed, being formed) and others are construed as
more everyday and commonsense (pull apart; surrounds, huddling around; break apart).
In the discourse of text (C), which concerns the building of knowledge of scien-
tific activities, the lecturer moves back and forth between the two. The concurrent
paralanguage configures both the entities and the occurrences in everyday terms.
In this it functions both convergently and divergently with the spoken language.

Intersemiotic construal of research

As well as the established (and observed) scientific knowledge discussed to this point,
a disciplinary field such as chemistry also involves a sub-field of research, which
deals with activities of experimentation conducted by scientists. This is exemplified
in the language of text (D) where the lecturer gives an account of activities involved
in an experiment called ‘the hot ice experiment’. For clarity, entities which are pro-
nominally identified are named within round brackets.
248 Susan Hood and Jing Hao

text (D)

That experiment there is what we refer to as the hot ice experiment. And [so]
what we have there – and what this animation here is showing – is that we have
sodium acetate that we’ve added to water, and we’ve got it to a saturated
point. We then heated that solution more. [So] we heated that solution,
added more and more sodium acetate, until it (sodium acetate)’s completely
dissolved. We can then let that (the solution) sit at room temperature for ever
how long we want. And it’s only after the addition of one single sodium ace-
tate crystal that it (solution) forms this crystallization.

Activities verbally realized though occurrences in (D) are concurrently expressed in

paralanguage as motion (as in Figure 10.14). Given the speed of some paralinguistic
expressions, it is not always clear whether the rotating hands depict a paralinguis-
tic occurrence figure (of entities and occurrence) or just an occurrence. Unless
made clear they are assumed to depict paralinguistic occurrences only. Some such
instances are described in Table 10.3 with convergent speech underlined.

Sequencing occurrence figures in language and paralanguage

From a dynamic perspective, in both the subfield of established disciplinary knowl-
edge (text (C)) and that of research (text (D)), we are concerned with how activities
are related to one another. In the discourse semantics, occurrence figures are related
through either causal or temporal logical connexions (and then; if… then; so that)
(Hao 2020a; Martin 1992).
Returning briefly to text (C), activities involved in ‘dissolving salt’ are explained
through a series of logically connected occurrence figures. The sequence in (21) is
divided into three figures with the connexions linking them shown in small caps.

(21) Those are our water molecules huddling around our different ions
(in order) to break it (lattice) apart.

TABLE 10.3 Convergent expressions of occurrences

Spoken language Paralanguage

We have sodium acetate that we’ve added two rotations of the left hand & forearm
to water clockwise away from the body
We’ve then heated that solution more five rotations of both hands & forearms one
over the other away from the body
added more and more (/kept adding) two rotations of the right hand & forearm
sodium acetate away from the body
 Grounded learning 249

And (then) it (lattice) will be uniformly dispersed throughout the solution.

The first link is a causal-purposive one – in order to. The second is a temporal one –
and then. The sequence in (21) unfolds by ‘implication’, meaning that one activity
implies what has gone before (Doran and Martin this volume). In the discourse
semantics, both temporal and causal connexions can be used interchangeably for
construing implication relations (Hao 2020a). This is illustrated in the rephrasing of
(21) above as (22) below.

(22) Those are our water molecules huddling around our different ions
and then (they) break it (lattice) apart.
So that it (lattice) will be uniformly dispersed throughout the solution.

Text (D) construes the procedure of the hot ice experiment. Here, in addition to
implication, we also have activities organized through ‘expectancy’, i.e. one activity
expects what potentially follows (Doran and Martin this volume). In example (23),
the occurrence figures are related through temporal connexions in small caps.

(23) we heated that solution,

and then (we) add more and more sodium acetate
We can then let that (the solution) sit at room temperature for however long
we want

In contrast to the implication activities, temporal connexions construing expec-

tancy are not interchangeable with causal ones. The rephrasings marked * in (24)
are not possible.

(24) we heated that solution,

*so that (we) add more and more sodium acetate
*so that we can let that (the solution) sit at room temperature for however
long we want

Language provides resources which can account for a diversity of temporal and causal
connexion – resources which allow us to make explicit both expectancy and impli-
cation relations in a disciplinary field. In contrast to language, paralanguage cannot
differentiate kinds of connexion in a sequence. It can only commit the meaning of
connexion as sequence in space/time.This is evident in the relation of the underlined
spoken language in text (E) and concurrent paralanguage.

text (E)

So the structure of our internal solids is dependent on charge and size – what
effect that has on our physical properties – the higher our charges, the greater
the electrostatic attraction between positive and negative, which means we’re
going to need to put more energy in there to break it. Which means we have a
stronger bond. So, higher our charge, stronger bond, high melting point.
250 Susan Hood and Jing Hao

FIGURE 10.15 An implication sequence in language concurrently depicted as sequence

in space/time only

The underlined summative encapsulation of an implication sequence in text (E),

elides all but the grammatical attributes:

So, (the) higher our charge (is), (the) stronger bond (is), (the) high(er the) melting
point (will be)

The lecturer’s concurrent paralanguage in Figure 10.15 converges with this elided
summative encapsulation. It depicts a sequence of three paralinguistic entities
located along a horizontal axis in space/time. Potentially these paralinguistic enti-
ties, convergent as they are with higher our charge etc., might be interpreted as con-
struing qualities, or as semiotic entities realizing a sequence of propositions, (the)
higher our charge (is) etc.

Activity entities in language and paralanguage

In a final step we return to the discourse semantic system of entities to consider one
further type, that of activity entities (Hao 2020a, 2020b, this volume). Scientists
need to explain and record scientific activities step by step through the sequencing
of occurrence figures as in text (E). But they also need to talk about activities in
more consolidated terms. They do so by naming activities.
We illustrate the use of activity entities in the lecturer’s discourse by referring
back to text (C) on ‘dissolving salt’. There he explains that dissolving salt in water
can make a highly saturated solution, one which reacts to a single added sodium
acetate crystal. He illustrates this process in a YouTube video clip which visually
displays the reacting process but does not explain it. After the viewing the lecturer
consolidates the displayed dynamic field knowledge in the spoken text transcribed
in text (F).
 Grounded learning 251

text (F)

it’s only after the addition of one single sodium acetate crystal that it forms this
crystallization. [So] we’re just seeing that reaction. And why we refer to that
reaction as the hot ice reaction, it’s because – as we saw in the video – it looks
like ice crystals being formed, or feathers, as they so astutely put it in the video.
But it’s actually an exothermic reaction. (…) [So] it’s a really useful reaction,
and it looks kind of cool.

FIGURE 10.16 Types of ‘reaction’ established in the lecture data

In text (F), the lecturer uses four activity entities to refer to the process seen in the
video: crystallization, hot ice reaction, exothermic reaction, and reaction. These activity
entities are related to one another through classification (Figure 10.16).
The entity reaction is a superordinate entity and can be categorized into subtypes.
Exothermic reaction is one subtype. The entity hot ice reaction paraphrases exothermic
reaction but is not a scientifically defined term. A more delicate subtype of exothermic
reaction is crystallization.
Since activity entities allow for classifying activities, they share similar features
with items in field (referred to as itemized activities; Doran and Martin this vol-
ume). In classifying activities, activity entities (like technical thing entities) commit
a considerable mass of ideational meaning potential.
The inherent relationship between an activity entity and field activities and items
makes for interesting interactions with paralanguage. In (25) from text (F), the activ-
ity entity in bold, crystallization, is concurrently expressed as a paralinguistic entity
in Figure 10.17.

(25) it’s only after the addition of one single sodium acetate crystal that it forms this
crystallization.

In text (G), in an explanation of field activities, the lecturer construes two of

these as activity entities – repulsion and attraction.
252 Susan Hood and Jing Hao

text (G)

…this is what we know with ionic solids. They pack themselves into a certain
way in order to maximize attractions between the positive and negative and
minimize any repulsion between same-charged ions.

FIGURE 10.17 The paralinguistic entity concurrent with ‘crystallization’

Concurrent with the activity entities of attraction and repulsion are paralinguistic
occurrence figures expressing entities in motion. In our data the paralanguage
concurrent with repulsion always depicts two sculpted entities (one in each hand)
alternating rapidly into and out from the body (Figure 10.18). The paralanguage
concurrent with attraction is always expressed in the two sculpted entities in rapid
horizontal alternating movements from left to right in front of the lecturer’s body.
Activity entities in language assign a technical name to field activities.They con-
strue both an activity and an item at the same time.5 The ‘itemized activity’ in field
can be depicted as either a paralinguistic entity (crystallization in Figure 10.17) or
a sequence of motion (repulsion in Figure 10.18). In either expression this highly
technical phenomenon can only have a concrete representation in an everyday,
commonsense reality.
 Grounded learning 253

FIGURE 10.18 The paralinguistic expression concurrent with of ‘repulsion’

Conclusion
Readily observed in the data is the extent to which the semiotic mode of paralan-
guage is in play throughout the lecture. In convergence with language, meanings are
committed in two modes, one supporting the other in meaning-making potential.
However, our interest extends to how convergent meanings are being expressed in
each mode and the relations which hold between them – the potential offered in
the intersemiosis.
Narrowing our focus to concurrence in expressions of ideational meaning (ide-
ational semovergence), two consistent patterns of cooperation and complementarity
emerge. The first has to do with the relative commitment of ideational meanings
relevant to the established knowledge of the disciplinary field and to building that
knowledge in pedagogic encounters.The construal of the field of molecular chemis-
try in the lecturer’s talk is unsurprisingly highly technical overall, although there are
shifts into more commonsense from time to time. In the technicality which names
entities and occurrences, a heightened mass of ideational meaning is committed.
This is complemented in paralanguage in the construal of a commonsense reality,
one which commits apparent concrete and tangible entities, entities in motion, and
motions across gestural space.
A second perspective on the complementarity is from mode. This points us to
textual resources for identifying entities in discourse. Of particular relevance here
is the distinction between generic and specific identification. The technicality of
the spoken language refers to phenomena in categorical terms which generalize
across contexts, while the paralanguage depicts specific instances of phenomena in
254 Susan Hood and Jing Hao

the immediate shared material reality of the lecture theatre.The intersemiosis of the
lecturer’s language and paralanguage thus bridges the technical and the everyday,
a generalizing perspective and a specific material reality – all in support of build-
ing knowledge of the specialized field. It is in both these senses that we discuss the
paralanguage as grounding learning.
Finally, we note the rapidly evolving environment in which undergraduate stu-
dents are introduced to new uncommonsense disciplinary knowledge. A growing
reliance on online platforms and social media technologies for pedagogic interac-
tion sees correspondingly reduced access to face-to-face lectures for many students.
Justifications for the pace and extent of these changes in pedagogic modes rarely
engage with an adequate research base to inform and critique practices, particularly
one which takes account of how modes of interaction participate in knowledge
building. Our aim has been to contribute insights to inform policies and pedagogic
practices which support students to move into new uncommonsense understand-
ings of their disciplinary fields. As we have shown, lecturers’ paralanguage in face-
to-face teaching/learning offers intuitive yet significant support for students’ access
to uncommonsense knowledge. We propose that this mode of communication be
taken into serious consideration in pedagogic practices.

Notes
1 Comparable observations on embodied meanings are found within the broader gesture
literature from other theoretical and disciplinary perspectives. Key sources for other per-
spectives include McNeill (1992, 2000), Kendon (2004) and Calbris (2011).
2 An additional discourse semantic unit of quantity is not discussed in this paper as it was
not foregrounded in our selected segments of data.
3 In referring to a congruent construal of meaning in paralanguage we avoid using the
term iconicity which carries a very different meaning in SFS from that typical in
other descriptive accounts of gesture. In the latter it labels a commonsense association
of gestures with things they look like in a material world (Wilkin and Holler 2011:
299, McNeill 1992). In SFS, a reference to paralanguage as iconic would be making
special reference to the way in which everyday commonsense language construes a
sense of a world around us as a commonsense ideational reality. In other words, it
would necessarily be interpreted through meaning-making choices in language (Ngo
et al., 2021).
4 In (21) we infer from the co-text that solid particles refer to very many entities, but the
commitment of quantity in the paralanguage is limited to two entities, one in each hand.
5 Note this is not grammatical metaphor as there is no possible unpacking at the grammati-
cal level.

References
Calbris, G. (2011) Elements of Meaning in Gesture, Philadelphia: John Benjamin.
Cléirigh, C. (2011) Gestural and Postural Semiosis: A Systemic Functional Linguistics Approach To
‘Body Language’, Unpublished manuscript.
 Grounded learning 255

Doran, Y. J. (2018) The Discourse of Physics: Building Knowledge through Language, Mathematics
and Image, London: Routledge.
Ge, Y.-P., Unsworth, L., Wang, K.-H., and Chang, H.-P. (2018) ‘Image design for enhancing
science learning’, in K.-S. Tang and K. Danielsson (eds) Global Developments in Literacy
Research for Science Education, Dordrecht: Springer, 237–58.
Halliday, M. A. K. (1978) Language as Social Semiotic, London: Arnold.
Halliday, M. A. K. (1994) An Introduction to Functional Grammar, London: Arnold.
Halliday, M. A. K. and Greaves, W. S. (2008) Intonation in the Grammar of English, London:
Equinox.
Halliday, M. A. K. and Martin, J. R. (1993) Writing Science, London: Falmer Press.
Halliday, M. A. K. and Matthiessen, C. M. I. M. (2014) Halliday’s Introduction to Functional
Grammar, London: Routledge.
Hao, J. (2018) ‘Reconsidering “cause inside the clause” in scientific discourse – from a dis-
course semantic perspective in systemic functional linguistics’, Text and Talk, 38(5): 520–50.
Hao, J. (2020a) Analysing Scientific Discourse from a Systemic Functional Linguistic Perspective,
London: Routledge.
Hao, J. (2020b) ‘Nominalisation in scientific English: A tristratal perspective’, Functions of
Language, 27(2): 143–73.
Hao, J. and Hood, S. (2019) ‘Valuing science: The role of language and body language in a
health science lecture’, Journal of Pragmatics, 139, 200–15.
He,Y. and van Leeuwen,T. (2020) ‘Animation and the remediation of school physics – a social
semiotics approach’, Social Semiotics, 30(5), 665–84.
Hood, S. (2008) ‘Summary writing: Implicating meaning in processes of change’, Linguistics
and Education 19, 351–65.
Hood, S. (2010) Appraising Research, New York: Palgrave Macmillan.
Hood, S. (2011) ‘Body language in face-to-face teaching: A focus on textual and interpersonal
metafunctions’, in S. Dreyfus, S. Hood and M. Stenglin (eds) Semiotic Margins, London:
Continuum, 31–52.
Hood, S. (2020) ‘Live lectures: The significance of presence in building disciplinary knowl-
edge’, in J. R. Martin, K. Maton and Y.J. Doran (eds) Accessing Academic Discourse, London:
Routledge, 211–35.
Hood, S. and Lander, J. (2016) ‘Technologies, modes and pedagogic potential in live versus
online lectures’ International Journal of Language Studies 10(3), 23–42.
Kendon, A. (2004) Gesture, Cambridge: Cambridge University Press.
Lemke, J. L. (1990) Talking Science, Norwood, NJ: Ablex.
Martin, J. R. (1992) English Text: System and Structure, Amsterdam: John Benjamins.
Martin, J. R. (2008) ‘Tenderness: Realisation and instantiation in a Botswanan town’, Odense
Working Papers in Language and Communication, 30–62.
Martin, J. R. (2011) ‘Multimodal semiotics:Theoretical challenges’, in S. Dreyfus, S. Hood and
M. Stenglin (eds), Semiotic Margins, London: Continuum, 243–70.
Martin, J. R. (2017) ‘Revisiting field: Specialized knowledge in Ancient History and Biology
secondary school discourse’, Onomázein Special Issue, March: 111–48.
Martin, J. R. and Matruglio, E. (2020) ‘Revisiting mode: Context in/dependency in Ancient
History classroom discourse’, in J. R. Martin, K. Maton and Y. J. Doran (eds) Accessing
Academic Discourse, London: Routledge, 72–95.
Martin, J. R., Rose, D. (2007) Working with Discourse, New York: Continuum.
Martin, J. R. and Veel, R. (1998) Reading Science, London: Routledge.
Martin, J. R. and Zappavigna, M. (2019) ‘Embodied meaning: A systemic functional perspec-
tive on paralanguage’, Functional Linguistics, 6 (1), 1–33.
256 Susan Hood and Jing Hao

Martinec, R. (2000) ‘Rhythm in multimodal texts’, Leonardo 33 (4), 289–97.

Martinec, R. (2001) ‘Interpersonal resources in action’, Semiotica, 135 (1/4), 117–45.
Martinec, R. (2002) ‘Rhythmic hierarchy in monologue and dialogue’, Functions of Language,
9(1), 39–59.
Matthiessen, C. M. I. M. (2007) ‘The multimodal page: A systemic functional perspective’, in
T. Royce and W. Bowcher (eds) New Directions in the Analysis of Multimodal Discourse, New
Jersey: Erlbaum, 1–62.
McNeill, D. (1992) Hand and Mind, Chicago: University of Chicago Press.
McNeill, D. (2000) Language and Gesture, Cambridge: Cambridge University Press.
Ngo, T., Hood, S., Martin, J. R., Painter, C., Smith, B. and Zappavigna, M. (2021) Modelling
Paralanguage Using Systemic Functional Semiotics, London: Bloomsbury.
Painter, C., Martin, J. R. and Unsworth, L. (2013) Reading Visual Narratives, London: Equinox.
Saussure, F. (1974) Course in General Linguistics, London: Peter Owan.
Wilkin, K. and Holler, J. (2011) ‘Speakers’ use of “action” and “identity” gestures with definite
and indefinite references’, in G. Stam and M. Ishino (eds) Integrating Gesture, Philadelphia:
John Benjamins, 293–307.
Zappavigna, M., Cléirigh, C., Dwyer, P. and Martin, J. R., (2010) ‘The coupling of gesture and
phonology’, in M. Bednarek and J. R. Martin (eds) New Discourse on Language, London:
Continuum, 237–66.
Zhao, S. (2010) ‘Intersemiotic relations as logogenetic patterns:Towards the restoration of the
time dimension in hyper-text description’, in M. Bednarek and J.R. Martin (eds) New
Discourse on Language, London: Continuum, 195–218.
11
DOING MATHS
(De)constructing procedures for maths processes

David Rose

Introduction
Reading to Learn (R2L) is both a teaching methodology and a professional learning
program that trains teachers in the methodology (Rose 2017). The methodology
is designed to enable teachers to support all students in their classes to continually
succeed with learning tasks at appropriate levels for their age, grade and curriculum.
Central to the methodology is a close analysis of the texts that students are expected
to read and write (knowledge genres) and of the classroom activities through which
students learn the curriculum (curriculum genres). A primary focus is on guiding
students to read curriculum texts, and then use what they have learnt from reading
in their writing. Mathematics differs from other curriculum subjects as its writ-
ten texts are generally brief and ancillary to the major curriculum genre of the
discipline – which is in fact the intermodal modelling of maths processes by teach-
ers, followed by individual student problem solving using the modelled process.
While R2L offers strategies for teaching the written maths genres of definitions,
explanations and ‘word problems’, the primary focus of the professional learning
program and this chapter is on the curriculum genre modelling maths processes.
The term ‘maths process’ is commonly used by teachers for mathematical genres
that Doran (2018) describes technically as quantifications, which find a numerical
result for a maths problem, and derivations, which derive relations among technical
symbols. We will use the name maths process modelling genre for the curriculum genre
that models these mathematical genres.
Following the introduction, the chapter describes this curriculum genre as it
has been re-designed in the R2L methodology. The re-designed process mod-
elling genre has three stages before students are expected to solve problems for
practice and assessment. The first is a teacher demonstration of the process using a
detailed lesson plan. The second is a series of guided practice cycles with increasing
handover of control to students. The third is a joint teacher/student construction
258 David Rose

of the procedural steps for the process. Each of these stages is described by analyz-
ing extracts from a secondary school maths lesson. Lessons are analyzed from three
perspectives: the structuring of pedagogic activities, teacher/learner interactions,
and the modalities through which meanings are exchanged in the discourse. This
approach is known as pedagogic register analysis (Rose 2014, 2018).
The primary social concerns of R2L are with the inequity of educational out-
comes and with enabling teachers to narrow the achievement gap between their
students. This inequity is particularly apparent in maths outcomes. At the level of
systems, there seems to have been little or no growth in maths outcomes in recent
generations (Leigh and Ryan 2008, Meeks, Kemp and Stephenson 2014). Concern
is regularly expressed in the media about the diminishing rate of candidates for
science-based degrees, all of which depend on expertise in maths. At the classroom
level, Table 11.1 presents the averaged outcomes of maths topic tests in one Sydney
secondary maths class. While the top group is consistently successful, the middle
group barely passes, and the botton group consistently fails. Maths teachers in the
R2L professional program report that this scale of difference in achievement levels
is typical. The proportion of students achieving success in maths may vary between
classes and suburbs, but the majority of students are still not successful enough to
qualify for maths dependent disciplines. For the bottom group to achieve passable
scores would require an average growth rate of over 60%. For the middle group to
reach the top group average would require over 30% growth rate.This was the chal-
lenge for designing effective maths curriculum genres in R2L.
Teachers in the R2L professional program regularly express three specific con-
cerns about their students’ maths learning. One is with the ‘maths word problems’
that are supposed to contextualize students’ calculation tasks in their everyday expe-
rience. Text 1 is an example from the trigonometry section of a Year 9 maths text-
book (Barnes et al. 2015: 188).

Text 1: Maths word problem

A rescue helicopter spots a missing surfer drifting out to sea on his damaged board. The
helicopter descends vertically to a height of 19m above sea level and drops down an
emergency rope, which the surfer grips. Due to the wind the rope swings at an angle of
27° to the vertical, as shown in the diagram.What is the length of the rope?

A major concern of teachers is the effect on weaker students of all these irrelevant
words, which distract students from the abstract data and operations they need to
solve the problems. It appears that this pervasive practice may actually have the
opposite effect of its purported function, to facilitate maths learning. As students
are assessed on their success with problem solving, the emotional effect on weaker
students is to create and maintain learner identities as ‘no good at maths’. At the
level of schools’ socioeconomic function to reproduce educational hierarchies (the
assessment ‘bell curve’), it may be surmised that this too is an actual social function
of maths word problems. In response, R2L has designed a simple analysis of maths
 Doing maths 259

TABLE 11.1 Average % scores in topic tests in Year 8 maths class

Student %
Samim 90
Syed 89
Rafah 85
Usaama 75
Timothy 70
Blake 61
Daniel 61
Top group average 76%
James 59
Dylan 57
Sameer 56
Luke 55
Ryan 54
Jack 53
Marcus 50
Zack 50
Middle group average 54%
Daniel 47
Airrison 41
Monroe 39
Matthew 36
Cory 33
Steven 28
Thomas 27
Brendan 25
Bottom group average 34%

word problems and a procedure for applying it in the classroom that enables stu-
dents to disentangle the data, solution and operations from their wordings. Briefly,
the three-step method is to first identify the data presented literally by mathematical
items in the question, e.g.

• height of 19m
• angle of 27° to the vertical

The second step is to infer the required solution, encoded in lexical relations among
wordings. In this case, this is the length of the hypotenuse, which must be inferred
from lexical relations between height 19m above sea level <–> rope at angle 27° to the
vertical <–> length of rope. The third step is to choose the maths processes required
to solve this question, in this case, calculating the hypotenuse from sine of 27° with
opposite side of 19m.
260 David Rose

Another concern of teachers is with students’ inability to comprehend, remem-

ber and apply technical terminology in maths. One source of this problem is the
complex technicality of mathematical texts, such as the definitions and explanations
that traditionally introduce each topic in maths textbooks. A typical example is Text
2, from the Year 9 trigonometry curriculum.

Text 2: Technical definition

In any right angled triangle, for any angle:
The sine of the angle = the length of the opposite side
the length of the hypotenuse.
The cosine of the angle = the length of the adjacent side
the length of the hypotenuse.
The tangent of the angle = the length of the opposite side
the length of the adjacent side.

Where students are weak readers, one strategy of maths teachers is avoid reading
such highly technical texts in textbooks, and instead present definitions and expla-
nations orally, in the course of modelling relevant maths processes on the board.This
in fact seems to be the standard practice in the majority of Australian classrooms.
A recent trend in textbooks is to delete many technical texts, and instead present
technical terms in lists. A problem that remains with either approach is that written
questions in problem-solving activities are equally technical, so that weaker students
still struggle to read them.
In response, R2L provides teachers with a method known as Detailed Reading
and Rewriting. In these activities, teachers guide students to identify each meaning
in a short passage of text and discuss their meanings in depth. Detailed Reading
enables all students to read a text with comprehension, no matter what their inde-
pendent reading abilities. The information from the text is typically then written
on the board as notes, and the class then rewrites the notes as a new text. When
maths teachers regularly apply these strategies to key texts, students rapidly learn to
independently read mathematical technicality. Detailed Reading and Rewriting are
designed curriculum genres, described in Rose and Martin (2012).

The maths process modelling genre

A third teacher concern is the inability of many students to independently solve
problems, following teacher modelling of the relevant process. This is perhaps the
most common curriculum genre in maths classrooms the world over. In a typical
practice, the teacher demonstrates each step of the process by writing a worked
example on the board, explaining what they are doing, and engaging the class by
asking for appropriate information at each step. Close observation shows this prac-
tice to be a highly complex mix of modalities and interactions between teachers and
students. This pedagogic complexity is additional to the process itself, which may
 Doing maths 261

involve many steps and choice points, along with technical terms and knowledge
that students must have mastered in previous lessons and school stages.
For this reason, some teachers choose to model difficult processes with more
than one worked example and involve students as much as possible, before they
are expected to solve problems individually. Nevertheless, all teachers participating
in R2L professional learning report that a minority of students in their classes can
generally solve problems accurately following the modelling. Other students have
more difficulty with problem solving, while others cannot do the tasks without
individual support from the teacher. Although teachers spend most time support-
ing these students during problem solving activities, they continue to struggle with
maths. The pedagogic function of problem solving activities is to practise applying
maths processes. Hence successful students receive most benefit from the practice,
and weak students achieve least benefit. Furthermore, the practice function is con-
flated with evaluation, so that weaker students experience continual failure – the
emotional impact of which further reduces their learning capacities and learner
identities.
The source of these disparities in students’ capacity to apply modelled processes
to problem solving is not hard to see. Students are more likely to succeed with prob-
lem solving if they are able to follow teachers’ explanations in each step of a process,
understand the technical terms they are using, recognize the relations between the
verbal explanation and the symbolic text and diagrams that are being written on
the board, and remember each step. If students miss any of this complex web of ele-
ments, they will find problem solving more difficult.
The R2L program addresses this problem by redesigning the maths process mod-
elling genre, to provide more support for all students to learn processes effectively.
The goal of this redesigned genre is for students to record the steps in a maths pro-
cess as a written procedure, which they can then apply to problem solving. To this
end, teachers plan how they will model the process in close detail, including the
approximate wordings they will use to explain each step, the questions they will ask
students, and the worked example they will write on the board. This detailed plan-
ning impels teachers to carefully consider how they present maths processes, and the
wordings they use to do so.Table 11.2 shows such a lesson plan for the trigonometry
lesson analyzed in this chapter.
The first column names eight steps that this teacher considers necessary for solv-
ing trigonometry questions at this school grade. These steps will be named orally as
the process is demonstrated in the lesson.The second column includes one or more
questions with which students’ participation may be engaged. These questions are
designed to ensure an appropriate answer, derived from information written on the
board or shared knowledge from previous lessons.The third column shows how the
teacher will elaborate these responses as a series of notes, diagram and equations on
the board.
Teachers then use such procedural plans to model the maths process.This model-
ling is similar to their standard practice, demonstrating and explaining each step of
a process, except that it has been carefully planned. The process is then modelled
262 David Rose

TABLE 11.2 Lesson plan for R2L maths process genre

Steps Questions Example

1 Read the question. What does it say? In a triangle LMN, angle
M = 90°. Side LN is 9.2m
and side LM is 8.2m. Find the
angle L to the nearest degree.
2 Make a note of the What is the first The triangle is a right angled
important information important triangle
in the question. information?
What is the next Side LM is 8.2
important Side LN is 9.2
information? Right angle at M
3 Draw a labelled diagram What are the angles
with this information. called?
What are the lengths of
the sides?
4 Label the angle to be Which angle is ϴ?
found as ϴ

5 Write down the three What is the first opposite side MN

trigonometric ratios trigonometric ratio? Sin ϴ = =
hypotenuse 9.2
related to the right What is the next ratio?
adjacent side 8.2
Cos ϴ = hypotenuse = 9.2
angled triangle. What is the next ratio?

opposite side MN
Tan ϴ = adjacent side = 8.2

6 Find the right ratio to Which ratio gives both Cos ϴ

solve the question. sides?
7 Use this ratio to solve What are the steps to 
cos  8.2
the question. simplify the ratio? =
1 9.2
9.2 × Cos ϴ = 8.2
What is the value of ϴ? 8.2
ϴ = Cos−1
9.2
8 Use a calculator to What is the answer? 8.2
calculate the angle. ϴ = Shift Cos
9.2
0
ϴ = 27

again with a different example, but this time the teacher asks students to name each
step, and students take turns to write the worked example on the board, with the
help of the teacher and class.This guided practice is then repeated, typically once for
secondary students and up to five or six times for younger students, with increasing
 Doing maths 263

handover of control to students in each iteration. Finally, students take turns to

record each step of the process, as a written procedure on the board, with the help
of teacher and class, and students write the procedure in their maths books for later
reference. Results to date include 100% improvements within three months for fail-
ing students, and an average of 20% improvements within two months in controlled
research projects (Lövstedt and Rose 2015, Table 11.10 below).
In terms of genre and register theory, each of these iterations – demonstration,
guided practices and joint construction – are stages in the R2L maths process cur-
riculum genre. Each step in the lesson plan is a phase, involving a series of activi-
ties, within each genre stage. Stages are structural units at the stratum of genre, that
realize one or another genre. Phases are structural units at the stratum of register,
realizing features in field, tenor and mode systems (Martin and Rose, 2007, Rose
2006, 2020).
The methodology is freely available for teachers and researchers in a set of vid-
eoed demonstration lessons and resources, produced by this author for the NSW
Education Standards Authority (NESA 2018). Extracts from these videoed lessons
are analyzed here to describe the stages of the curriculum genre.

Pedagogic register analysis

As previewed earlier, the lesson analyses have three functional strands.They describe
the structuring of pedagogic activities, that are negotiated in pedagogic relations
between teachers and learners, and presented through pedagogic modalities of
speaking, writing, signing, drawing, viewing, gesturing and other somatic activity.
This is an analysis at the level of register. In broader terms, pedagogic activities are
the field, pedagogic relations are the tenor, and pedagogic modalities are the mode
of a pedagogic register.
The institutional function of a pedagogic register is to exchange knowledge and
values which constitute a curriculum register. Learners construe, or abduce, cur-
riculum knowledge through the pedagogic register, which also enacts curricular
values. Curricular values include learners’ degrees of access to knowledge through
pedagogic modalities, the authority they derive from this access, the affirmation
they receive, their inclusion in the learning conversation, and the learning auton-
omy they acquire. Each learner’s abduction of curriculum knowledge is condi-
tioned by their positioning in these value hierarchies, in relation to their peers.1
While curriculum knowledge is often an explicit focus of lessons, curricular values
tend to be enacted tacitly.
A curriculum genre is an unfolding configuration of features in pedagogic regis-
ter and curriculum register. In Bernstein’s terms (1990: 62), the function of curricu-
lum genres is ‘cultural reproduction or transformation’. The purpose of pedagogic
register analysis is to describe how these social processes unfold at the level of indi-
vidual learners and classes; to get at how learning actually occurs, by closely observ-
ing pedagogic exchanges using the research tools of systemic functional theory
(Rose 2014, 2018).
264 David Rose

In the lesson analyzed here, the curriculum register includes knowledge of the
procedure to solve trigonometry questions. Its values are carefully oriented towards
equality of access, authority, affirmation and inclusion. The goal is equal auton-
omy in the problem-solving practice that follows. In terms of instantiation theory
(Martin 2008), the procedure presented in the lesson re-commits a mathematical
quantification genre (Doran 2018) as a series of verbally named steps. Steps 1–3 are
concerned with the Situation stage of this genre (the question and its data), steps
4–7 with the Substitution stage (substituting numerical values in symbolic formu-
lae), and step 8 with the Numerical Result.
In terms of individuation, curricular values may unequally allocate resources to
learners to abduce and apply the procedure, hence their readings of the procedure
may be divergent. The goal of this lesson is to allocate access, authority, affirmation
and inclusion more equally, to facilitate convergent readings of the procedure, and
hence more equal success with autonomous problem solving.
The global structuring of pedagogic activity comprises a hierarchy of segments,
from whole lessons, to lesson activities, to learning cycles at the micro-level of
teacher/learner interactions. Units at each of these levels may be iterated in series.
The lesson analyzed here is one in a series within the curriculum topic of trigo-
nometry. Each stage of the lesson comprises a series of lesson activities named as
steps in the lesson plan (Table 11.2). Each of these lesson activities comprises a series
of learning cycles that are concerned with a specific micro-activity, such as writing
each trigonometric ratio in step 5. This hierarchy is illustrated for the trigonometry
lesson in Figure 11.1. In this curriculum genre, the central stage of Guided practice
may also be complexed into series, indicated in the figure by superscriptn.

FIGURE 11.1 Hierarchy of lessons stages, lesson activities and learning cycles
 Doing maths 265

FIGURE 11.2 Nuclear and marginal phases of learning cycles

The structure of units at each of these levels is centred on a learning task, through
which knowledge is abduced by learners. Tasks may involve an active display by
learners, to identify meanings in a text or propose from their own knowledge.
Or tasks may be receptive, to receive meanings verbally, or perceive them visually.
Learning tasks are typically focused (specified) and then evaluated by a teacher.
In addition, teachers may first prepare learners to succeed with the task, and the
knowledge they construe through the task may then be elaborated. These five
structural elements are termed Prepare, Focus, Task, Evaluate and Elaborate phases
(Martin and Rose 2007, Rose and Martin 2012).Their orbital structuring as nuclear
and marginal phases is diagrammed in Figure 11.2. Analyses may also name the
pedagogic matter that each phase is concerned with. This enables the analyst to
identify relations between the matter of the learning Task, and Prepare, Focus and
Elaborate phases.
Pedagogic relations include the roles of teachers and learners in interactions,
and the pedagogic acts that they negotiate. Pedagogic roles are termed interacts in
the analyses, as the term role is used within discourse semantics, for generalized
speaker roles in exchange structures. Interacts specify the pedagogic functions of
exchange roles at the connotative stratum of register. For teachers, potential inter-
acts include presenting knowledge, evaluating learners and directing the activ-
ity. Learners may display knowledge and both learners and teachers may solicit
acts from each other. Acts include various pedagogic behaviours and acts of con-
sciousness such as attention, perception, knowledge, reasoning, engagement and
anticipation.
Pedagogic modalities are the sources of meanings in the learning discourse, and
the means of sourcing them into the discourse. Sources of meanings include the
environment, spoken knowledge of teachers and learners, and records such as writ-
ten texts, graphic images and video recordings. Each of these source types has a set
of options for sourcing them into the discourse. For example, written texts may
be read aloud, but meanings in images must be recast as wordings. Meanings in
the environment and in records may also be indicated by naming or pointing to
them, by words or gestures. Spoken teacher/learner knowledge may be individual
or shared, and may be presented or elicited by teachers, and recalled or inferred by
learners. In addition, meanings may be recorded during the lesson, and then become
sources. They may be recorded by writing verbal or symbolic texts, or annotating a
266 David Rose

FIGURE 11.3 Multimodal teacher demonstration

text or image, or by drawing pictures or diagrams. Many of these options in modali-

ties are visible in Figure 11.3, a still from the videoed lesson (NESA 2018), in which
the teacher points at the hypotenuse in the diagram, while saying the trigonometric
ratio, sine theta is equal to opposite divided by hypotenuse.Values in pedagogic activities,
relations and modalities, that are applied in the analyses here, are set out as tables in
the Appendix to this chapter.

Teacher demonstration stage

Table 11.3 is an extract from step 2 in the teacher demonstration, in which key
information from the question is written as notes on the board. At this point the
teacher has read out the question which is written on the board (step 1).This extract
illustrates how students may be engaged in a maths demonstration by asking for
their perception, knowledge or reasoning. Where possible, realizations of interacts
are italicized in the transcript.
The extract includes two learning cycles, marked by horizontal lines in the
table. The first cycle begins with the teacher preparing the step, by directing the
activity to come, you write down all the important points from the question. The task is
for students to identify items in the question, with the property all the important
points. This requires them to reason about which items are important. To this end,
the focus phase invites students to reason, Who can tell me what’s the first important
point.
This focus phase includes the ‘hands-up’ classroom ritual, marked by dashed
lines.The command Put your hand up insists that students display for teacher evalu-
ation, the hand-up gesture invites evaluation, and one student is then permitted
to display. The curricular values of this ritual are inclusion and affirmation of
 Doing maths 267

TABLE 11.3 Teacher demonstration, Step 2: writing the ratios

phase interact
1 T Step 2 says you write down all the prepare step direct activity
important points from the
question.
T Who can tell me what’s the first focus property invite reasoning
important point?
Put your hand up insist display
Ss [hands up] invite evaluation
T Yes, S1 permit display
S1 Right angle triangle identify property display reasoning
T First it’s a right angle triangle. affirm repeat
[writes It is a right angled triangle] elaborate step
It is a right angle triangle.
First important point.
2 T Next one? focus item inquire reasoning
Ss [hands up] invite evaluation
T [points to S2] permit display
S2 Angle M is ninety degrees identify item display reasoning
T Angle M is ninety degrees. affirm repeat
[writes ∠M = 90° ] elaborate step

students’ authority. By commanding the whole class, the teacher here is counter-
ing a widespread problem where only a few students in a class are willing to invite
evaluation.
Here, Student 1 displays reasoning by identifying the item, Right angle triangle.
The teacher affirms by repeating this display, and then elaborates by writing the item
on the board, reading it back, and restating its value as the first important point. This
cycle combines curriculum goals for knowledge and values. The student’s response
instantiates the procedural step with a particular example. In so doing, he models the
reasoning task for the whole class to see. By repeating the display, the teacher both
affirms the student and reinforces the item. It is further reinforced by writing it and
reading it again, which models the activity, and is again valued as first important point,
which reinforces the reasoning task in this procedural step.
Cycle 2 reiterates the same pattern more briefly. The focus simply inquires rea-
soning, Next one? Student 2 is permitted to display reasoning and is affirmed by
repeating the display, Angle M is ninety degrees. This time the elaboration recasts this
verbal sentence by writing a mathematical equation, a symbolic text, ∠M = 90°.
These symbols will be used to label the diagram in step 3 of the trigonometry
process.
Table 11.4 is an extract from step 4 of the demonstration stage, in which
trigonometric ratios are written using values from the diagram. This monologic
268 David Rose

TABLE 11.4 Teacher demonstration, Step 4: writing the ratios

phase sourcing
1 OK, Step 4. prepare step new teacher
[writes Step 4] write note
2 Since this is a trigonometry question classify text
we’re going to
write down all the trig ratios related to new teacher
the question.
[points to diagram] point diagram
3 We’ll write down sine theta, cos theta restate lesson
and tan theta.
[lists on fingers] symbolize text
4 So, looking from the sine of theta, focus expression locate diagram
[writes Sin ϴ = ] perceive expression write expression
sine theta is equal to receive expression recast expression
[points to diagram] perceive part point diagram
opposite receive expression restate lesson
[points to opposite line] perceive part point diagram
divided by hypotenuse. receive expression restate lesson
[points to hypotenuse] perceive part point diagram
5 Therefore, MN receive expression recast diagram
[writes Sin ϴ = MN] perceive expression write expression
divided by receive expression recast expression
hypotenuse receive expression restate lesson
[points to hypotenuse] perceive part point diagram
which is 9.2. receive expression recast diagram
[writes Sin ϴ = MN/9.2] perceive expression write expression

demonstration involves a series of rapid shifts in modality, from speaking to writ-

ing words and mathematical expressions, to indicating the diagram, to speaking and
writing again.
Learning tasks in this cycle are receptive, switching back and forth from perceiv-
ing written expressions and parts of the diagram to receiving the teacher’s spoken
knowledge. In the table, each sentence in the monologue is numbered, and dashed
lines mark couplings of speaking and writing or gesturing.2 Where possible, sources
and sourcing are bolded in the transcript.
In sentences 1–3, the teacher prepares the step, first by stating Step 4, and
writing it as a note on the board, then by classifying the question, a trigonometry
question, and presenting new teacher knowledge about the step, to write down all
the trig ratios related to the question, and pointing to the diagram. He then specifies
these ratios as sine theta, cos theta and tan theta, which restates shared knowledge
from prior lessons.
 Doing maths 269

In Sentence 4, the teacher reviews the elements of the first ratio, the sine of theta,
that is shared knowledge from prior lessons, while pointing to the relevant elements
in the diagram. The focus locates the ratio in the diagram, looking from the sine of
theta. The students’ first tasks are then to perceive this as a symbolic expression
Sin ϴ =, to receive it rephrased in wordings, sine theta is equal to, and perceive
the relevant part of the diagram as the teacher points to it. Their next tasks are to
receive the technical term opposite, while perceiving this part of the diagram as the
teacher points.The next tasks are to receive the expression divided by hypotenuse, and
perceive this part of the diagram as teacher points.
In Sentence 5, the teacher specifies these elements of the ratio with values from
the diagram. The first task is to receive the expression Therefore MN, which recasts
labels from the diagram. (M and N are labelled angles in the diagram, so MN is the
side of the triangle between these angles, opposite the angle ϴ.) The next tasks are
to perceive this expression as the teacher writes it, including the underline MN,
that symbolizes division, and receive this operator as words, divided by. The final
tasks are to receive the term hypotenuse, and to perceive this part of the diagram
as the teacher points to it, then to receive its value, which is 9.2, and perceive the
whole ratio as the teacher finishes writing it, Sin ϴ = MN/9.2.
The intricate complexity of this intermodal activity illustrates a major source of
difficulties that many students experience with maths processes. It should be pointed
out that the teacher is an expert practitioner. He carefully reviews and builds on
knowledge that has been accumulated in previous lessons and uses it to contex-
tualize the worked example with which he illustrates the new process. The rapid
modality shifting is designed to facilitate this contextualization and is undoubtedly
effective for many learners. It maximizes their access to the curriculum knowledge,
by recommitting the procedure and its content through multiple modalities.
Nevertheless, if a student has a weak command of any of the spoken, written, sym-
bolic and visual elements, or relations between them, they may struggle to maintain
the close attention needed to follow the teacher’s modelling. Consequently, teach-
ers must often re-explain the process to such students during the individual prob-
lem solving activity. This pervasive practice attempts to repair the ineffectiveness
of standard curriculum genres, after students have failed to acquire the knowledge
they need, and simultaneously evaluates them as failing. The R2L methodology
reverses this normal but inequitable practice, by ensuring that all students are pre-
pared to succeed with learning tasks before they are evaluated. In the re-designed
maths process genre, this is accomplished with a series of guided practices of the
process, and a joint construction of the procedure. It does not matter if any students
did not get a perfect grasp of the procedure from the teacher demonstration, as it
is now repeated with increasing handover of authority, towards autonomy for all.

First guided practice

Table 11.5 is an extract from Step 2 of the first guided practice in this lesson. There
are now few prepare phases in learning cycles, as the tasks have been prepared in the
 270 David Rose
TABLE 11.5 First guided practice, Step 2: labelling the diagram

phase sourcing interact

1 T What is the next step? focus step remind lesson inquire knowledge
[hands up] invite evaluation
S1 permit display
S1 Draw the diagram propose step recall lesson display knowledge
T Draw the diagram. affirm repeat
OK. Let’s see how you can draw the diagram. elaborate step invite display
S1 [comes to board, draws triangle] draw diagram display knowledge
2 T Who can tell him where to put B? focus place read text invite reasoning
[hands up] invite evaluation
S2 permit display
S2 Bottom left propose place locate diagram display reasoning
T Bottom left. affirm repeat
S1 [labels angle B] elaborate step label diagram display knowledge
3 T Where would C go? focus place read text invite reasoning
[hands up] invite evaluation
S3 permit display
S3 Bottom right propose place locate diagram display reasoning
T Bottom right. affirm repeat
S1 [labels angle C] elaborate step label diagram display knowledge
 Doing maths 271

FIGURE 11.4 Labelling the diagram

prior demonstration. Instead of the teacher demonstrating, authority is handed to

the students, who now take turns to come to the board to scribe notes, expressions,
diagrams and labels. In Cycle 1, the focus reminds students of knowledge presented
in the demonstration, What is the next step? Student 1 proposes this step by recalling
the demonstration, Draw the diagram. The teacher affirms by repeating, and invites
the student to display drawing skills, Let’s see how you can draw the diagram. Drawing
the right angle triangle recasts this wording as a graphic image, and recalls the dem-
onstration, further displaying the student’s knowledge.
In Cycles 2 and 3, student 1 starts labelling the diagram from notes on the board
(Figure 11.4), guided by the class. The teacher’s first focus is a place in the diagram,
where to put B. Student 2 locates this place in the diagram, bottom left, by reasoning
from the note right angle at B, to the right angle in the diagram. Cycle 3 repeats this
pattern with angle C, as do the following cycles until the diagram is fully labelled
with values given in the question.
Table 11.6 analyses step 4, to show how this complex intermodal activity is
negotiated as students take more authority. The first focus invites students to display
their recall of the step, writing the ratios. Student 1 first recalls the formula from the
demonstration and prior lessons, Cos theta equal to adjacent over the hypotenuse, and
then recasts it with values from the diagram, which is BC over 30.The teacher affirms
by directing him to ‘come up’ to the board, where he recasts his answer as a written
equation, Cosϴ = BC/30.
Cycle 2 follows a similar pattern, but student 2 proposes only the type of ratio,
sine. Consequently, the teacher prepares cycle 3 by restating this move, and then
focuses the required expression, Sine theta over what? Student 2 then proposes these
values from the diagram, 22 over 30. Again, the teacher affirms by directing him to
‘go ahead’ to the board, where he recasts the values as a written equation, Sinϴ =
22/30. Cycle 4 is now even more abbreviated. The focus simply names the next
TABLE 11.6 Writing the ratios in first guided practice

272 David Rose

phase sourcing interact
1 T Who can come up and write all the ratios at the same focus step remind lesson invite display
time?
Ss [hands up] invite evaluation
T S1 permit display
S1 Cos theta equal to adjacent over the hypotenuse propose formula recall lesson display knowledge
which is BC over 30. recast diagram display reasoning
T Come up. affirm direct display
S1 [comes to board, writes Cosϴ = BC/30] elaborate equation write expression display knowledge
2 T Who can do the next ratio? focus expression remind lesson invite display
Ss [hands up] invite evaluation
T S2 permit display
S2 Sine propose expression recall lesson display knowledge
OK affirm approve
3 T S2 says sine ratio. prepare expression restate move repeat display
Sine theta over what? focus expression inquire reasoning
S2 That’s 22 over 30 propose expression recast diagram display reasoning
T Go ahead. affirm direct display
S2 [comes to board, writes Sinϴ = 22/30] elaborate equation write expression display knowledge
4 T Tan theta? focus expression remind lesson inquire reasoning
Ss [hands up] invite evaluation
T S3 permit display
S3 Tan theta propose equation recall lesson display reasoning
which is 22 over BC recast diagram
T Tan theta is 22 over BC. affirm repeat
S3 [comes to board, writes Tanϴ = 22/BC] elaborate equation write expression display knowledge
 Doing maths 273

ratio, Tan theta? Student 3 proposes the values 22 over BC from the diagram, the
teacher repeats, and the student comes to the board without being directed, to write
the equation, Tanϴ = 22/BC.
In this manner, the intermodal activity of recalling the formula for each ratio,
identifying values in the diagram, and writing these as an equation, is negotiated
interactively between teacher and students. Students are invited to display their
authority and are consistently affirmed.
In this first guided practice, the teacher chooses students who are most likely to
recall the formulae and reason accurately about the relevant values in the diagram,
following his demonstration. However, the whole class experiences this previously
modelled activity reinstantiated as an interactive process involving their peers. In
this stage around half the students in the class have an opportunity to propose
expressions or write on the board. The second guided practice now includes stu-
dents who are typically less successful at maths.

Second guided practice

Table 11.7 analyses writing the ratios in the second guided practice, in which the
teacher uses more interpersonal scaffolding to engage weaker students, along with
the support provided by repeating the activity. Cycle 1 is similar to cycles in the first
guided practice. Student 1 recalls the first ratio, adding values from the diagram, Cos
theta equal to 17 over PR, and is directed to write it on the board.
However, in Cycle 2, the teacher first prepares by directing students to reason
about the next ratio, warning them to be ready, I’m going to ask you about the next
ratio. He then names student 2, a weaker student, without the hands-up ritual. S2
correctly proposes the ratio with values from the diagram. To affirm S2, the teacher
hands authority to the whole class, by asking them to concur with the answer, He
said sine theta is equal to PQ over 21. Is he right? This is a strategy for inclusion in the
learning community. S2 then correctly writes the equation on the board, Sinϴ =
PQ/21cm. Following all this scaffolding, Cycle 3 abbreviates the interaction. The
teacher names student 3, who proposes merely the values from the diagram, PQ over
17cm, and then writes the whole equation, Tanϴ = PQ/17cm.
The power of repeated guided practice, in which all students participate actively
in constructing the maths process, is illustrated in Table 11.8, the solution phase of
the process. In cycle 1, multiple students propose the next step, solve the equation.
The whole class then solicits the weakest student 1 to become the scribe for this
step. The teacher checks and asks the student whether he concurs, Really? You want
to go there? S1 at first demurs by shaking his head, but the other students encourage
him by applauding, and he comes to the board and writes the first expression of
the solution, Cos ϴ =, displaying his knowledge of the correct ratio to solve the
problem. In cycle 2, the teacher asks for the values to ‘end up’ the equation. Student
2 proposes 17cm over 21, from the notes on the board.The teacher repeats this to S1,
who writes it as an expression, completing the equation, Cos ϴ = 17/21.
 274 David Rose
TABLE 11.7 Second guided practice, Step 4: writing the ratios

phase source
1 T Who can remember the first ratio? focus expression remind lesson
Ss [hands up]
T S1
S1 Cos theta propose equation recall lesson
equal to 17 over PR recast diagram
T OK. affirm
Then write down that here. elaborate equation
S1 [comes to board, writes Cos ϴ = 17cm/21cm] write expression
2 T Think about the next ratio while he’s writing. prepare step remind lesson
I’m going to ask you about the next ratio.
S2, what’s the next ratio? focus expression
S2 Sine theta propose equation recall lesson
equals PQ over 21 recast diagram
T He said sine theta is equal to PQ over 21. affirm
Is he right?
SS Yes
T Go and write on the board. elaborate equation
S2 [comes to board, writes Sin ϴ = PQ/21cm] write expression
3 What’s the next ratio? S3 focus expression remind lesson
S3 PQ over 17 cm propose expression recast diagram
T Can you put that on the board there? affirm
S4 [comes to board, writes Tan ϴ = PQ/17cm] elaborate equation write expression
TABLE 11.8 Second guided practice, Step 8: solving the problem

phase interact
1 T What is the next step? focus step inquire knowledge
Ss Solve the equation propose step display knowledge
Ss [name and point to S1 (weak student)] invite display
T Really? check
You want to go there? inquire accordance
S1 [shakes head, smiling] demur
Ss [applaud] insist display
S1 [comes to board, writes Cos ϴ = ] elaborate expression display knowledge
T OK affirm approve
2 T What is he going to end up? S2 focus expression inquire reasoning
S2 17cm over 21 propose expression display reasoning
T OK affirm approve
He said 17 over 21. elaborate equation direct display
S1 [writes Cos ϴ = 17/21] display knowledge
T OK approve
3 T What is the next step? focus step inquire reasoning
S3 Cos and then to the power of minus 1 propose expression display reasoning
4 T Anything on the left hand side? focus expression inquire knowledge
S3 Yes, and then theta is equal to propose expression display knowledge
T OK affirm approve

Doing maths 275

He’s saying put down theta is equal to. elaborate expression direct display
S1 [writes ϴ = ] display knowledge
5 T What’s he saying? focus expression inquire reception
S4 Cos to the power of minus 1 propose expression display reception

(Continued)
 276 David Rose
TABLE 11.8 (Continued)

phase interact
T Cos to the power of minus 1. affirm repeat
S1 [writes ϴ = cos-1 (17/21)] elaborate equation display reasoning
T Is he right? affirm inquire accordance
Ss Yes concur
T OK approve
6 T What is the next step? S5 focus step inquire knowledge
S5 Using a calculator to calculate the equation propose step display knowledge
T Using calculator to calculate the equation. affirm repeat
S5 [uses calculator] elaborate step
The answer is 35 degrees, 57 minutes display perception
T 35 degrees, 57 minutes. direct display
S1 [writes ϴ = 35° 57’] display knowledge
T Give him a round of applause. affirm praise
Ss [applaud] concur
 Doing maths 277

FIGURE 11.5 Writing the solution

Cycles 3–5 negotiate the next equation in the solution. Student 3 proposes the
expression Cos and then to the power of minus 1. The teacher asks for the left-hand
term in the equation, and S3 proposes theta is equal to.The teacher repeats this to S1,
who writes the correct expression ϴ =. In Cycle 5, the teacher asks for the right-
hand value to be restated, What’s he saying?3 Student 4 restates Cos to the power of
minus 1, which the teacher repeats to S1, who completes the equation as ϴ = cos−1
(17/21). This displays mathematical reasoning, as cos-1 is multiplied by brackets
enclosing 17/21. To affirm S1, the teacher then hands authority to the whole class,
Is he right?
Cycle 6 negotiates the last step in the solution. Student 5 uses his calculator to
calculate the value of cos-1 (17/21) and reads the answer as 35 degrees, 57 minutes.
The teacher repeats this to S1, who recasts it correctly in symbols, ϴ = 35° 57’
(Figure 11.5). By means of this collaborative scaffolding, a reluctant student is
actively included in modelling the trigonometry process, displays his own authority
and is affirmed by his peers and teacher. His achievement is finally affirmed by a
round of applause from the class.

Joint construction of procedure

The final stage in the R2L maths process genre is a joint construction of the pro-
cedure for the process. Each step in the procedure distils students’ experience of the
series of activities that they have just observed and participated in. As in the guided
practice stages, students take turns to scribe the steps on the board, with the support
of their peers and the teacher (Figure 11.6).
Table 11.9 illustrates the negotiation of the joint construction. Each step is nego-
tiated in a single learning cycle, in which the teacher focuses the step and elaborates
278 David Rose

FIGURE 11.6 Scribing the procedure

a student’s response, and the student writes the step on the board. Cycle 1 begins
with preparing by directing the activity, while reminding of the previous lesson
stage, We are going to review the steps.The focus inquires the first step, student 1 recalls
You read the question, and he is directed to the board, where he writes the step. Cycle
2 follows the same pattern, minus the preparation phase. Here the teacher expands
student 2’s answer, but the student writes his own version. Cycle 3 again follows this
pattern. Here the teacher again expands student 3’s answer, and the student writes
the expanded version.
The joint construction is completed in four more steps that follow these
variations. Students copy the procedure into their maths books as it is being writ-
ten, to refer to during individual problem solving. Text 3 shows the completed
procedure.

Text 3: Jointly constructed procedure

Trigonometry
Subtopic: Right angle triangle
Step 1: Read the question
Step 2: Write the important information
Step 3: Draw a labelled diagram
Step 4: Label the angle ϴ
Step 5: Write down the three trigonometric ratios
Step 6: Use process of elimination on 3 trig ratios
Step 7: Use the right ratio to find ϴ
TABLE 11.9 Joint construction of procedure steps

phase sourcing interact

1 T We are going to review the steps that you’re going to use to solve prepare activity remind lesson direct activity
the right angle triangle.
Who can tell me, what is the first thing you do? focus step remind lesson inquire knowledge
S1 You read the question propose step recall lesson display knowledge
T Very good. affirm praise
Come and put it on the board. elaborate text direct display
S1 [comes to board, writes Step 1: Read the question] write text display knowledge
T OK affirm approve
2 T What is the next step you take in order to solve this problem? focus step remind lesson inquire knowledge
S2 Write the important information propose step recall lesson display knowledge
T Write down the important information from the question. elaborate text rephrase move repeat
Can you put it on the board? direct display
S2 [comes to board, writes Step 2:Write the important information] write text display knowledge
3 T What is the third thing you do? focus step remind lesson inquire knowledge
S3 Draw a diagram propose step recall lesson display knowledge

T Draw a diagram. affirm repeat

Doing maths 279

Draw a labelled diagram. elaborate text rephrase move
S3 [comes to board, writes Step 3: Draw a labelled diagram] write text display knowledge
280 David Rose

Note that the wordings in Text 3 are not identical with the steps in the teacher’s
lesson plan (Table 11.1), as Text 3 is jointly constructed with contributions from
students as well as teacher. The joint construction with students scribing is a criti-
cal activity for students to take control of the procedure, and remember how to
reproduce it in their problem solving. Although each step in the procedure is only
a synopsis of the activity it stands for, students’ participation in the guided activi-
ties and recasting them as the procedure ensures that this knowledge becomes
their own.

Conclusion
Mathematics is the archetypal intellectual discipline, the exercise of pure reason
applying theory. But its practices do not occur just in disembodied minds; they are
spoken in words and written in symbols, in order to be shared, extended and accu-
mulated over generations. Moreover, learning mathematics is an emotional expe-
rience, and the sources of emotion are almost always found in our relations with
others. In this regard, the function of affirmation in pedagogic practice cannot be
overstated. Bernstein tells us that “evaluation condenses the meaning of the whole
[pedagogic] device” (2000: 50). Where the social function of this institution is to
create and maintain inequality, it uses unequal evaluation to create unequal inclu-
sion and learner autonomy. Hence the curriculum registers of mathematics are
more than mathematical knowledge such as trigonometry formulae and quantifica-
tion genres; equally important are curricular values of access, authority, affirmation,
inclusion and autonomy.
Pedagogic relations are not simply a knowledge transfer of transmission and
acquisition. Rather learners abduce knowledge and values from the totality of an
unfolding pedagogic register, and this abduction inevitably varies from learner to
learner. The degrees to which learners are able to abduce elements of culturally
accumulated knowledge such as trigonometry are directly related to the curricu-
lar values they experience. By middle secondary school, successful learners have
internalized continual affirmation. They can learn autonomously, not only because
they possess the requisite field knowledge (authority) and reading skills (access), but
because they are self-affirming.
Such students are able to attend to the rapid modality switching in teacher dem-
onstrations, illustrated in Table 11.4 above, and reproduce the activity in their inde-
pendent problem solving. This targeted repetitive practice with continual success
automatizes their procedural knowledge of the process so that it can be applied with
ease to complex problems that require meta-reasoning about procedures. The neu-
rological basis of such procedural learning and reasoning is described by Edelman
(1992), who explains how perceptual categorization is necessarily charged by emo-
tional values from the brain’s limbic system. More abstract concepts are formed by
the brain’s perceptions of its own categories, again charged with emotional values.
Automatization results from repeated, emotionally charged perception and concept
 Doing maths 281

formation, that becomes the background for further perceptual categorization and
concept formation.
Here is another concern of teachers, and a criticism of the R2L curriculum
genre described above. Teachers report that struggling students are unable to inter-
pret the solutions required by complex maths problems, even when they know the
processes involved. The constructivist position, that currently dominates western
teacher education, holds that learners must first understand the maths ‘concept’,
from which they can find their own way to solve problems. From these positions,
the R2L genre appears to be constraining rather than enabling, as it teaches a pre-
defined set of steps that cannot help students’ creative problem solving. What is
invisible to these positions is the automatized procedural knowledge possessed by
successful students, that enables them to focus attention on meta-procedures to
interpret complex problems.
The R2L curriculum genre is designed to provide access to this procedural
knowledge for all students in ‘mixed-ability’ classes (i.e. all classes). Its informing
principles include the model of learning cycles, in which learners are prepared for
learning tasks so that they are continually successful and affirmed, and ready for the
elaborations that follow. Second is the principle of repetition with increasing hando-
ver of authority from teacher to learners. Third is the need to ensure that every
student in a class is equally included and affirmed in the learning activity. Fourth
is the intermodal strategy of recasting spoken knowledge as written texts, and vice
versa, in both teacher planning and jointly constructed procedures.
The goal is to maximize the autonomy of all students to practise individual
problem solving successfully, and so benefit from its pedagogic functions. To this
end, R2L principles are also applied to problem solving activities. Teachers select
representative problems and guide the class to solve them jointly – identifying data,
solution and operations – before individual practice with the remaining problems.
This strategy further minimizes potential struggles for weaker students and maxi-
mizes success and automatization for all. For teachers, the close analysis of learning
tasks, and of their own teaching practice, becomes an automatic part of their prac-
tice. It can then be applied to analyzing and explicitly teaching the meta-procedures
involved in complex problem solving, rather than mystifying them as exotic cogni-
tive processes. For students, the kinds of results that can be expected from consistent
application of the R2L maths process genre are illustrated in Table 11.10. This table
shows the growth in average scores in maths topic tests over two school terms, for
the same Year 8 students as Table 11.1. Significantly, the average growth rates for
each cohort are close to the necessary growth rates predicted by the achievement
gaps in Table 11.1.

Appendix: Pedagogic Register Analysis

Table 11.11–11.15 set out the values in pedagogic activities, modalities and relations
that are applied in analyses above. See Rose (2018) for further discussion.
282 David Rose

TABLE 11.10 Growth in topic test scores using R2L maths process genre

Student Pre-R2L average After 2 terms of Percentage

scores R2L maths growth
Samim 90 93 3
Syed 89 95 7
Rafah 85 90 6
Usaama 75 82 9
Timothy 70 80 14
Blake 61 83 36
Daniel 61 68 11
Average growth of top group 12%
James 59 68 15
Dylan 57 55 0
Sameer 56 74 32
Luke 55 73 32
Ryan 54 87 61
Jack 53 89 68
Marcus 50 54 8
Zack 50 63 26
Average growth of middle group 30%
Daniel 47 51 8
Airrison 41 53 29
Monroe 39 67 46
Matthew 36 48 33
Cory 33 66 100
Steven 28 55 96
Thomas 27 39 44
Brendan 25 67 128
Average growth of bottom group 60%

TABLE 11.11 Learning cycle phases

Nuclear focus guided explicit criteria

Phase unguided criteria implicit
task manual ...
semiotic displaying identify in text
propose from knowledge
receptive receive verbally
perceive visually
evaluate affirm
reject (if not affirmed)
Marginal prepare with explicit criteria before focus
Phase elaborate monologic teacher knowledge
dialogic negotiated with learners
 Doing maths 283

TABLE 11.12 Record sources

Record record modality

sources verbal visual
record graphic verbal text picture
type record symbolic text diagram
recording audio recording video recording
record individual
access shared display
copy
Record sourcing restate repeat e.g. read aloud
sourcing mode diverge summarize
rephrase
recast
indicate pointing verbal locate
gesture point
describing verbal compare
class/part
gesture imitate
symbolize
sourcing same as record
language other language

TABLE 11.13 Spoken sources

Spoken individual knowledge teacher knowledge

sources learner knowledge
shared knowledge prior lesson
prior move
Spoken teacher speaking present new
Sourcing restate
elicit remind
enquire
learner speaking recall
infer

TABLE 11.14 Recording

Write writing mode

wordings symbols
writing write text constructed text equation
type notes expression
annotate record
Draw picture
diagram
mark record
TABLE 11.15 Acts and interacts

284 David Rose

Interacts teacher roles instructing presenting model
knowledge impart
evaluating check (before evaluating)
learners evaluate affirm repeat
approve
praise
reject implicit qualify
ignore
explicit negate
admonish
directing direct
suggest
permit
learner roles display (for evaluation)
accord concur4
demur
teacher/learner invite
roles inquire
insist
Acts behavioural learners behaviour
acts display
accordance
teacher evaluation
teacher/learners activity
conscious perceptive attention
acts perception
reception
cognitive knowledge
choice
reasoning
conception
affective attitude
engagement
anticipation
 Doing maths 285

Notes
1 Hierarchies in curricular values expand Bernstein’s ‘hierachy of success and failure’ (2000:
11), in terms of curriculum genres, where access:pedagogic modalities :: affirmation and
inclusion:pedagogic relations :: authority:curriculum knowledge :: autonomy:control
over curriculum genres.
2 In terms of discourse semantics, written sentences in lesson transcripts here are moves in
exchanges (Martin and Rose 2007, Rose 2018). Each role of teacher and learners in the
exchange may be complexes of moves, or single items such as OK. Each exchange move
has a register level value in pedagogic relations. Learning cycle phases may be realized by
one or more exchange moves. Pedagogic modalities tend to switch from move to move,
but can also switch rapidly within moves, or be constant across several moves. Linguistic
labelling of exchange moves is not included in the analyses here, to reduce the complex-
ity of the presentation and save space in tables.
3 Doran (2018) describes the ‘left-hand’ and ‘right-hand’ of equations as Theme and
Articulation of mathematical statements.
4 Learners have options to concur or demur from teachers’ or learners’ acts, whereas teach-
ers have the institutional authority to evaluate learner acts by affirming or rejecting
(Bernstein 1990, Rose 2018).

References
Barnes, M., J. Bradley, L. Elms and R. Cahn (2015) Maths Quest 9 for the Australian Curriculum,
2nd edition, Brisbane: Jacaranda Press
Bernstein, B. (1990) The Structuring of Pedagogic Discourse, London: Routledge.
Doran,Y. J. (2018) The Discourse of Physics: Building knowledge through language, Mathematics and
Image, London: Routledge.
Edelman, G. (1992) Bright Air, Brilliant Fire: On the matter of the mind, London: Harper
Collins.
Leigh, A. and Ryan, C. (2008) ‘How has School Productivity Changed in Australia?’ Canberra:
Australian National University http://econrsss.anu.edu.au/~aleigh/.
Lövstedt, A-C and Rose, D. (2015) Reading to Learn Maths: A teacher professional development
project in Stockholm, https://www.readingtolearn.com.au/wp-content/uploads/2016/01/
Reading-to-Learn-Maths-A-teacher-professional-development-project-in-Stockholm.
pdf.
Martin, J. R. (2008) ‘Tenderness: realization and instantiation in a Botswanan town’, Odense
Working Papers in Language and Communication, 30–62.
Meeks, L., Kemp, C. and Stephenson, J. (2014) ‘Standards in literacy and numeracy:
Contributing factors’ Australian Journal of Teacher Education (Online), 39(7), 106.
NESA (2018) Planning for Success in Secondary Mathematics, Sydney: NSW Educational
Standards Authority, http://educationstandards.nsw.edu.au/wps/portal/nesa/k-10/
learning-areas/mathematics/planning-for-success-in-secondary-mathematics.
Martin, J.R. and Rose, D. (2007) Working with Discourse: Meaning beyond the Clause, London:
Continuum.
Rose, D. (2006) ‘Reading genre: a new wave of analysis’, Linguistics and the Human Sciences
2(2), 185–204.
Rose, D. Analysing pedagogic discourse: an approach from genre and register. Functional
Linguist. 1, 11 (2014). https://doi.org/10.1186/s40554-014-0011-4.
Rose, D. (2017) ‘Languages of schooling: Embedding literacy learning with genre-based ped-
agogy’ European Journal of Applied Linguistics, 5(2), 1–31.
286 David Rose

Rose, D. (2018) ‘Pedagogic register analysis: Mapping choices in teaching and learning’
Functional Linguistics, 5(3), 1–33.
Rose, D. (2020) ‘The baboon and the bee: Exploring register patterns across languages’, in
J. R. Martin, Y. J. Doran and G. Figuedero (eds) Systemic Functional Language Description,
London: Routledge, 273–306.
Rose, D. and Martin, J. R. (2012) Learning to Write, Reading to Learn: Genre, knowledge and
pedagogy in the Sydney School, London: Equinox.
INDEX

Note: concepts from Legitimation Code Theory or Systemic Functional Linguistics are
indicated as LCT or SFL

activity (SFL) 6–7, 13, 109–17, 123–30, autonomy codes (LCT) 9, 24, 28–31, 33,
131n7, 131n10, 137–51, 153, 156, 35, 39, 45–6, 100n7; see also exotic codes;
158–9, 163, 166–71, 173, 178, 180–1, introjected codes; positional autonomy;
182n2, 230, 245, 247; cyclical 113–14; projected code; relational autonomy;
expectancy 110, 113–14, 141, 146–7, sovereign codes
173, 249; implication 6, 110–12, 114, autonomy pathways (LCT) 24, 30, 33, 34,
137–8, 141, 146–7, 168–71, 173, 182n2, 37, 39, 44–6, 100n7; see also autonomy
249–50; itemized 125, 144, 147, 169; tours; one-way trips
linear 113–14; momented 110, 112–14, autonomy plane (LCT) 29, 81, 82, 85, 93;
116, 147, 171, 182n2 see also addition; disintegration;
addition (LCT) 81, 86–8, 98; see also integration; introjection; projection;
autonomy plane; disintegration; subtraction; substitution; targets
integration; introjection; projection; autonomy tours (LCT) 2, 10, 12, 30, 33,
subtraction; substitution 38–46, 100; see also autonomy pathways
affordance 12, 26, 76–9, 81–2, 85–9, axiological constellations (LCT) 73n7,
93–100, 178, 180; see also epistemic 100n2, 106
affordance (LCT) axiological–semantic density
aggregating (LCT) 54–6, 71–2, 73n6; (LCT) 182n1
see also assembling; constellations axiological–semantic gravity
animations 115; see also integrating (LCT) 182n1
animations into science teaching
appraisal (SFL) 120–1, 142 Bacon, Roger 23
array (SFL) see property, arrayed ballet 9, 53
assembling (LCT) 55, 56, 62, 72; see also Bernstein, Basil 72, 134–5, 190, 207, 224,
aggregating; constellations 263, 280; and LCT 8, 10
atomism (LCT) 49–50 biology 5–6, 10, 13, 110–11, 134–59, 227
attitude (SFL) 121, 145, 147–8, 229, 284; birds 6, 118–22, 138–40
see also appraisal Bloom’s taxonomy 50
Autonomy (LCT) 9–11, 24, 28–33, 79–82; body language 8, 13, 14, 107, 136, 226–54;
see also autonomy codes; autonomy plane; see also paralanguage
dimensions (LCT); targets Bourdieu, Pierre 8
288 Index

charging (LCT) 51–3, 73 dimension (SFL) 141–2, 151–9, 230, 237,

chemistry 8, 13, 14, 159n4, 191, 193–5, 198; 240, 242; categorized 154–8; measured
and LCT 10, 208; and paralanguage 226, 123, 125, 154–8; perceived 125, 154–6;
227, 229–30, 233, 236, 244, 245, 247, 253 structured 154–6, 158
clusters (LCT) 51–73; base 54; discourse semantics 136, 141; ideational
supplementary 54 (SFL) 2, 5–6, 11, 13–14, 107, 11, 123,
code clash (LCT) 190, 199 130, 130n3
code shift (LCT) 37 discursive relations (LCT) 206, 208–9, 211,
co-elaboration of entities (SFL) 142, 153, 215, 220–4, 225n1
154, 157, 158 diSessa, A. 50
cognitive blending framework 26 disintegration (LCT) 81; see also autonomy
cognitive-learning-design axis 77–8; plane; integration; introjection;
see also knowledge-blindness projection; subtraction; substitution
cognitive load theory 77–8 distillation (SFL) see technicality
cognitive theory of multimedia learning Doran,Y. J. 7–9, 11–13, 26, 53, 73, 73n4,
77–8 73n5, 73n7, 89, 101n13, 101n15, 101n16,
commitment (SFL) 231, 236, 238, 247, 253, 134, 141, 144, 159, 170, 171, 173, 175,
254n4 178, 182n3, 257, 285n3
complexity see semantic density (LCT) see
mass (SFL) élite codes (LCT) 189, 190; see also
Computer-Supported Collaborative specialization codes
Learning 78–9 engineering 2, 10, 14, 187–188, 201,
condensation (LCT) 46, 51–53, 73, 73n2, 205–225
135, 171, 213; see also epistemological entity (SFL) 13, 131n8, 134–59, 230;
condensation activity entity 111–12, 125, 131n6, 142,
congruent (SFL) 5, 231–2, 244, 254n3 144–8, 151, 158, 232, 250–2; enacted
conjunction (SFL) see connexion entity 146–8, 158; inferable entity 150–1,
connexion (SFL) 111–13, 143, 147, 155; instrumental entity 143, 145–6, 149,
159n3, 232 155; linguistically defined entity 148–50;
constellation diagrams (LCT) 55, 56, 63, 67, observational entity 143, 145–6, 150,
72, 73n11 154; observed entity 146–8; ostensively
constellations (LCT) 2, 9, 11, 50, 51–55, defined entity 148–9; in paralanguage
89, 100n2, 101n13, 106, 109, 123, 233–6; semiotic entity 232; tech-
127; enacted to analyse textbooks and enhanced entity 150–1, 156–7; thing
teaching 55–73; see also pedagogic entity 142–5, 147, 232, 236, 251; trained
constellations; schematic constellations gaze entity 150–1, 156–7; types of entity
context-dependence see presence (SFL) 140–3, 149
see semantic gravity (LCT) epistemic affordance (LCT) 2, 12, 77, 79,
cosmological analysis (LCT) 51, 52, 73n3 81, 85–9, 93–100, 100n2
cosmologies (LCT) 51–2, 54, 73n3 epistemic plane (LCT) 10
curriculum genre (SFL) 257, 260–4, 281 epistemic relations (LCT) 46, 135, 189–90,
192–4, 196, 198–201
dead metaphor (SFL) see activity entity epistemic-semantic density (LCT) 159, 182n1
decolonization 10, 49 epistemological access 10, 188, 191,
dependent links (LCT) 54, 57, 64, 127; see also 199–202
clusters; constellations; independent links epistemological condensation
derivation (genre) (SFL) 174–6, 180, (LCT) 73, 135
182n4, 257 epistemological constellations (LCT) 73n7,
description (genre) (SFL) 15n1 100n2, 106
diagrams 7, 94, 107, 122, 150, 151, 178–9, essentialism 12, 27, 33
261, 265, 271 ethnopoetics 9
dimensions (LCT) 9–10, 28, 79, 135; exotic codes (LCT) 29, 30, 33–8, 38–41,
see also Autonomy; Semantics (LCT); 43–6, 47n10, 47n12; see also autonomy
Specialization; Temporality codes
 Index 289

experiments 9, 38–44, 47n13, 66, 113, 134, inflammation 110–11

143, 145–6, 159n4, 177–8, 189, 193–5, image 2, 7–8, 10–11, 13, 26, 77, 79, 83, 94,
247–9 107, 136, 150–1, 162–4, 166, 171, 177–81,
explanations 3, 6, 11–12, 105–6, 251, 261; 226–7, 229–30, 242, 246, 265, 271
analysed as constellations (LCT) 49–73; individuation (SFL) 264
and field (SFL) 105–33; teaching with inferential relations (SFL) 221–2
animations (LCT) 82–101 instantiation (SFL) 231, 264
integrating animations into science teaching
field (SFL) 2, 5–7, 11, 13, 107–30, 135, 7, 11, 75–102; and ‘affordances’ 78–9;
137, 141, 162–82, 227, 230–3, 236, 242, analysis of failing to integrate 82–8;
244–54, 263, 280; dynamic see activity; analysis of successful integration 89–99;
everyday field 5–6, 135, 149; field of and multimedia learning approaches
object of study 138, 145, 158; field of 77–8; value of LCT in examining 79–82,
research 138, 145, 247–8; recreational 99–100
field 5–6, 135; specialized field 149; static integrating mathematics into science
see item see taxonomy; types of field 137 teaching 23–48; analysis of failure to
figure (SFL) 139, 142 see also occurrence integrate 33–8; analysis of successful
figure see also state figure integration 38–44; science education
flame robin 118–22 research and 23–7; value of LCT in
examining 28–33, 45–6
genre (SFL) 3, 8, 11, 13, 137, 162, 260–2; integration (LCT) 81, 87; see also addition;
see also curriculum genre; quantification autonomy plane; disintegration;
genre introjection; projection; subtraction;
genre pedagogy (SFL) 3 substitution
generic translation device (LCT) 31, 80; see interdependency of field (SFL) 123–30,
also specific translation device; translation 173–4, 179–80; see also interrelating of
device fields; reconstrual of fields
Glenn, Eleanor 53 interpersonal metafunction (SFL) 8, 120–1,
grammar (SFL) see lexicogrammar 137, 228
grammatical metaphor (SFL) 3–6, 112, interrelating of fields (SFL) 126–30, 173–4,
139–40, 155, 168–9, 182n5, 231–2; 179–80; elaboration 128; enhancing 128;
experiential 139, 148 extension 127–8
graphology 107, 227 intersemiosis (SFL) 8, 226, 244–5, 253
graphs 7, 33, 38–44, 46, 47n13, 107, 120, introjected codes (LCT) 30, 33, 38–46; see
177, 178–81 also autonomy codes
guided practice 269–77 introjection (LCT) 81, 87, 88; see also
addition; autonomy plane; disintegration;
Halliday, M. A. K. 4–5, 10, 11, 127–8, 139, integration; projection; subtraction;
142, 226 substitution
Hasan, R. 10 item (SFL) 108–17, 123, 131n10, 157, 163,
Hao, Jing 5, 11, 13, 14, 111, 112, 123, 130, 167–71, 180, 182n3, 230; activated 125
130n3, 130n5, 131n6, 131n7, 131n8, itemization (SFL) see activity, itemized see
142, 182n2; see also discourse semantics; property, itemized see reconstrual of fields
ideational
Hood, Susan 14, 138 joint construction (SFL) 263, 269, 277–80
horizontal discourse 135
humanities 15n1, 106, 135, 137, 190, 209 knower code (LCT) 189, 190, 192, 196–201
knowledge-blindness (LCT) 12, 25–6,
ideational metafunction (SFL) 5, 107, 49–50, 77; see also ways of knowing
120–1, 130, 135, 137, 140, 227–9, 223, knowledge code (LCT) 135, 189, 190,
232–3 192–5, 198–201
identification (SFL) 142, 145, 149–50, 229, knowledge structure (LCT) 10, 135
233, 253; exophora 145, 149–50
independent links (LCT) 54, 57, 127; see also Lambrinos, Elena 53, 73n4
clusters; constellations; dependent links languages of legitimation (LCT) 28, 188
290 Index

learning cycle (SFL) 264–6, 269, 277, 281, nominalization 123, 137, 139–40, 144–5,
282, 285n2 148, 155, 168; see also activity entity;
legitimation codes (LCT) 9, 28, 52, 54, 188 grammatical metaphor
Legitimation Code Theory (LCT) 1–2, non-targets (LCT) see targets
8–12, 24, 26, 28, 50, 51–4, 77, 79, 134,
135, 163, 188, 206, 207; dialogue occurrence figure (SFL) 111–12, 125,
with SFL 10–11, 106–7; and science 131n6, 143, 230, 245; in paralanguage
knowledge 8–10; see also Autonomy; 245–6
Semantics; Specialization; Temporality one-way trips (LCT) 30, 33–8, 44, 45, 180;
Lemke, J. L. 7, 226 see also autonomy pathways
lexical density (SFL) 4 ontic condensation (LCT) 46
lexicogrammar (SFL) 3–8, 11, 13, 107, 109, ontic relations (LCT) 208–9, 213–14, 224
111–12, 123, 130n5, 136–7, 139–48, 153,
155–6, 159n3, 227, 232, 233, 236, 242–3, paralanguage (SFL) 14, 226–56
245, 250, 254n5 PCK (pedagogic content knowledge) 50
literacy 3–4, 7, 162–3 pedagogic constellations 56, 59–63,
66–71, 73n11
Martin, J. R. 3, 7, 11, 13, 134, 137–8, 140–1, pedagogic register (SFL) 263–6, 281–4;
144, 149, 170, 173, 175, 178, 227–8, 260 pedagogic activity 264–5; pedagogic
mass (SFL) 11, 244, 251, 253 modalities 265–6, 280–1; pedagogic
mathematics 2, 7–8, 10–14, 107, 120, 124, relation 265, 280
155, 162–5, 171–81, 182n3, 257–85; pedagogic work (LCT) 12, 77, 81–2, 86, 87,
mathematics education 257–259, 280; 89–99, 100
see also derivation genre; integrating phenomenological primitives 50
mathematics into science teaching; phonology 107, 227, 229
quantification genre physics 8, 10–11, 13, 26, 34, 50, 117, 120,
Matthiessen, C. M. I. M. 5, 11, 127, 124–5, 162–82, 190, 191, 193, 195, 207–8
128, 142 positional autonomy (LCT) 29–35,
Maton, Karl 9, 12, 46n5, 46n6, 52–3, 74n13, 37, 39–42, 44, 77, 79–81, 99; see also
89, 100n6, 100n7, 101n13, 101n16, 127, autonomy codes; relational autonomy;
130n4, 134, 135, 159, 171, 180, 190, 208; targets
see also the son and the heir of a shyness presence (SFL) 11
that is criminally vulgar procedural recount (SFL) 112
metafunction (SFL) 8, 136–7, 227–9; see also procedure (SFL) 3, 6, 14, 112, 173, 137, 146,
ideational metafunction; interpersonal 173, 230, 249, 257–86
metafunction; textual metafunction production fields (LCT) 46n3, 72, 135
metalanguage (SFL) 4 projected codes (LCT) 30; see also
Metropolitan East Disadvantaged Schools autonomy codes
Program 3 projection (LCT) 81; see also addition;
mode (SFL) 107, 137, 149, 227, 231, autonomy plane; disintegration; integration;
253, 263 introjection; subtraction; substitution
Moon, synchronous rotation with the Earth property (SFL) 7, 13, 116–30, 131n7,
82–8; textbook accounts of role in tides 131n10, 141, 157, 163, 170–1, 173–81,
56–9; teaching of role in tides 59–62 182n3, 230, 242; activated 124–5;
multiliteracies 7, 162 arrayed 117–23, 170–1, 173, 178, 180,
multimedia 10–11; multimedia effect 76; 182n5; gauged 117–18, 120, 122–3, 175;
see also integrating animations into itemized 120, 123, 125, 153, 182n3;
science teaching qualitative 117–21, 123; spatio-temporal
multimodality 3, 7–8, 13, 162–82 117, 121–2
prosaic codes (LCT) 164, 166, 172, 176, 181
New South Wales Education Standards protocol (SFL) 112
Authority (NESA) 47n8, 47n9,
47n11, 263 quality (SFL) 123, 131n8, 141, 153, 157,
nominal group (SFL) 111–12, 139–41, 163, 230, 233–4, 238, 242–4; in paralanguage
157–8 242–4
 Index 291

quantification (genre) (SFL) 120, 174–6, semantics (SFL) see discourse semantics
178, 180, 257, 264, 280 semantic waves (LCT) 10
semovergent paralanguage (SFL) 229
rarefied codes (LCT) 165 sequence (discourse semantics) (SFL)
Reading to Learn 2, 14–15, 257 112, 142, 144, 146–7, 230, 249; in
Reading Science (book) 3, 227 paralanguage 249–50
realization (rules) 191, 194, 198–9 sequence (field) (SFL) see activity,
recognition (rules) 191, 194, 199 momented
recontextualization fields (LCT) 46n3, 72 Shark Park 122
reconstrual of fields (SFL) 123–5, 144, 157, singulars 207
168, 171, 182n3 social field of practice (LCT) 8–9
recount (SFL) 15n1, 112, 146, 173 social relations (LCT) 135, 137, 189–90,
register (SFL) 3–5, 11, 13, 107, 130, 192–4, 196–202
134–5, 137, 163, 166, 227, 230, 231, 258, sonovergent paralanguage (SFL) 228
263–84, 285n2; see also field; mode; tenor sovereign codes (LCT) 29, 30, 33–4, 36–46;
regions 207, 224 see also autonomy codes
relational autonomy (LCT) 29–32, 34, special expressions (SFL) 4
35, 37, 39–44, 47n13, 77, 79–81, 87, Specialization (LCT) 9–10, 14, 46, 79,
99; see also autonomy codes; positional 135, 188–9, 206, 208; see also epistemic
autonomy; targets relations; social relations; specialization
relativism 27, 33 codes
relativist codes (LCT) 190, 199 specialization codes (LCT) 9, 188–90, 200,
report (SFL) 3, 118, 121, 135–6, 141, 145–6, 202; see also epistemic relations; social
156–9 relations
reproduction fields (LCT) 46n3, 72, 135 specialization plane (LCT) 189
resources framework 25 specialized field (SFL) 6–7, 138, 148–9, 253
return trips (LCT) 180; see also autonomy specific translation device (LCT) 32, 81, 83,
pathways 85, 90, 93, 195, 197, 213; see also generic
rhizomatic codes (LCT) 165, 166, 172, translation device; translation device
175–7, 180, 181 Star Trek 23, 28
rules of the game (LCT) 8 state figure (SFL) 125, 131n6, 230
Rusznyak, Lee 73n2 story (genre) (SFL) 6, 112–13
stratal tension see grammatical metaphor
schematic constellations (LCT) 56–66, 69, stratification (SFL) 107, 130, 136–7, 227–8
71–2, 73n8, 73n11; see also constellations subjectivism 25, 26, 46n2
seasons, explanations of 11, 105–29; substitution (LCT) 81, 82, 95–9; see also
analysed as constellations (LCT) 63–71; addition; autonomy plane; subtraction
and field (SFL) 123–9; use of animations subtraction (LCT) 81, 95; see also addition;
in teaching of 89–99 autonomy plane; substitution
semantic codes (LCT) 9, 164, 165, 208, 219 syntactic ambiguity (SFL) 4
semantic density (LCT) 11, 73n2, 135, 159, Systemic Functional Linguistics (SFL) 1–15,
164–6, 167–81, 182n1, 206–9, 211–15, 103–84, 226–84; dialogue with LCT
219, 221–4; see also axiological-semantic 11–12, 106–7, 159
density; epistemic-semantic density Szenes, Eszter 53, 73n7
semantic discontinuity (SFL) 4
semantic gravity (LCT) 11, 164–7, 169, Talking Science (book) 3, 226
171–2, 175–81, 182n1, 182n4, 206–208, targets (LCT) 24, 28, 31–5, 37–9, 42,
211–12, 222; see also axiological-semantic 44–6, 77, 80–3, 85–90, 93–100, 100n5;
gravity; epistemic-semantic gravity ancillary target 31–2, 80, 81, 83, 86, 89,
semantic plane (LCT) 164–5, 172, 176, 181 90; associated non-target 31–3, 47n12,
semantic profiling (LCT) 10 47n13, 80, 81, 83, 89, 90; core target
Semantics (LCT) 9–11, 14, 79, 135, 162–82; 31–4, 37–9, 46n7, 47n13, 80–100, 101n9;
see also dimensions (LCT); semantic code; non-targets 31–3, 35, 37, 39, 47n10,
semantic density; semantic gravity 47n12, 80–3, 85–6, 89, 94–6; target/
292 Index

non-target content 32, 34, 35, 37, 80, 86, thinking dispositions 26
88, 89, 93, 94, 96, 98; target/non-target threshold concepts 49–50
purpose 32, 34, 35, 37, 42, 47n13, 81, 86, tides: analysed as constellations (LCT)
89, 94, 96, 98, 101n9; unassociated 56–63; explanations of 11, 110–11,
non-target 31–3, 80, 81, 83, 89, 90, 94; 113, 50
see also Autonomy; autonomy plane Tilakaratna, Namala 53, 73n7
Taurus 51 transitivity (SFL) 130
taxonomy (SFL) 4–7, 13, 108–9, 123–30, translation device (LCT) 28, 46, 73, 80,
131n7, 134–59, 163, 169–71, 173, 177, 81, 192, 192; and targets 31–3; see also
178, 181, 182n3, 236; and classification generic translation device; specific
4–5, 108–9, 123, 137–8, 141, 156–7, 163; translation device
and composition 4–5, 108–9, 114, 116, trinocular vision (SFL) 135
141, 156–7, 163
teacher demonstration 266–9 verbal group (SFL) 112
technicality (SFL) 5, 7, 8, 134, 125, 155, 166, vertical discourse 135
171, 177, 227, 231, 236, 244,
253, 260 ways of knowing 1, 24–6, 28, 45, 46n2,
Temporality (LCT) 15n3; see also 46n4, 50; see also knowledge-blindness
dimensions (LCT) work-integrated learning 10
tenor (SFL) 107, 137, 227, 263 worldly codes (LCT) 164, 166, 176–7,
textual metafunction (SFL) 137, 148–50, 180, 181
228, 263 Writing Science (book) 3–5, 7, 227